From paul at boddie.net  Wed Apr 17 11:39:36 2002
From: paul at boddie.net (Paul Boddie)
Date: 17 Apr 2002 08:39:36 -0700
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019036193.29349.python-list@python.org>
Message-ID: <23891c90.0204170739.ec18eb1@posting.google.com>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> wrote in message news:<mailman.1019036193.29349.python-list at python.org>...
> On Tue, 16 Apr 2002, P. Alejandro Lopez-Valencia wrote:
> >
> > And while you are at it... You have not compiled python itself as PIC
> > code. No, gcc -shared *doesn't* work in Solaris unless you are using GNU
> > Binutils instead of the native linker and assembler, something *not
> > recommended at all*. Solaris' linker can't bild a dynamic loading
> > library/module using object code compiled in position dependent mode.

I first came up against this with GNU Bayonne on Solaris 7. However,
wxPython also brought up some similar issues in the past, too:

  http://groups.google.com/groups?th=610b65907f740e28

> I ran configure.  I have tried running configure with --enable-gnu-ld
> and this didn't help, even with the latest binutils, so I have used the
> sun ld.

As far as I remember, you really do need to use binutils if you start
to experience the linking problems that you described. Before, that
didn't seem to be necessary with Python, but I'd guess that this has
somehow become the case in the past two years - I haven't really used
Solaris for about a year now, however.

Paul


From scarblac at pino.selwerd.nl  Thu Apr  4 02:09:43 2002
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 4 Apr 2002 07:09:43 GMT
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <Xns91E5CE0E5919Dcliechtigmxnet@62.2.16.82> <eppstein-055127.14142303042002@news.service.uci.edu>
Message-ID: <slrnaanv0j.vt3.scarblac@pino.selwerd.nl>

David Eppstein <eppstein at ics.uci.edu> wrote in comp.lang.python:
> In article <Xns91E5CE0E5919Dcliechtigmxnet at 62.2.16.82>,
>  Chris Liechti <cliechti at gmx.net> wrote:
> 
> > if d.has_key(key):
> >      foo = stuff(d[key])
> > else:
> >      foo = otherStuff(key)
> 
> I've seen a couple responses suggesting this.
> 
> Is there a reason why in modern Python versions one should still use 
> has_key?  I like the syntax "if key in d" better.
 
Are you sure you have the newest Python everywhere you want to use it? I'm
not. Why make your programs incompatible with such a trivial change?

-- 
Remco Gerlich


From johnroth at ameritech.net  Sat Apr 27 20:15:25 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 27 Apr 2002 20:15:25 -0400
Subject: inline assignments in conditionals
References: <2143388891-BeMail@sugoi> <mailman.1019937169.5421.python-list@python.org>
Message-ID: <ucmfnkt9ounsac@news.supernews.com>

"Michael Gilfix" <mgilfix at eecs.tufts.edu> wrote in message
news:mailman.1019937169.5421.python-list at python.org...
> On Sat, Apr 27 @ 12:33, jaf wrote:
> > Writing things that way is really unnatural, at least for English
> > speakers.  It makes the intent of the code harder to understand (at
> > least for me).  For example, if you were talking to a friend, you
> > woudn't say:
> >
> >    "If today is Sunday, I'm going to go shopping"
> >
> > but rather:
> >
> >    "If Sunday is today, I'm going to go shopping"
> >
> > and in any case, writing the constant first only protects in the
case
> > where one of the compared values is a constant.  If you are doing
> >
> >    if (a == b):
> >
> > then there is no way to protect yourself against actually writing
'='.

Unnatural and partial as it is, it is a recommended best
practice to protect yourself from accidentally writing '='
when you meant '==' in C or C++.

I don't have any objection to implementing assignment
in expressions. I have a real objection to doing it in such
a way that it makes silly mistakes easier, however.

John Roth




From clubdelpelosuperfluo at yahoo.it  Tue Apr 23 03:29:14 2002
From: clubdelpelosuperfluo at yahoo.it (Stefano)
Date: 23 Apr 2002 00:29:14 -0700
Subject: Numeric Python, COM and Eigenvalues
References: <9e1f166b.0204220437.57a087de@posting.google.com> <aa1ft2$mn7$2@peabody.colorado.edu>
Message-ID: <9e1f166b.0204222329.6f5685a6@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aa1ft2$mn7$2 at peabody.colorado.edu>...
> Stefano wrote:
> 
> > I am working with Numpy and trying to create a COM server.
> > I managed to call this server by Excel and get results like Inverse
> > matrix or Cholesky decomposition but the server doesn't work as I ask
> > for Eigenvalues/vectors (Excel does not respond anymore).
> > Has anyone heard of a problem of this kind?
> 
> Well, with zero example code and details provided that's going to be a hard 
> one to get help on ;)
> 
> Do you know for a fact you're feeding it a diagonalizable matrix? Does numpy 
> compute the eigenvals/vects nicely if you test it with the same matrix on its 
> own (not via excel)? Start testing with a symmetric 2x2 real matrix for which 
> you can hand-compute the eigenvalues trivially, that might tell you where the 
> problem is.
> 
> Good luck,
> 
> f.

I have already tried with a trivial matrix via Shell.
The same matrix makes Excel freeze.

I have walked two path to use numeric via com
1. making a client that talks with Mark Hammond's doubletalk (do you
know?)
2. making a server and a client by my own (actually with the help of
Poulsen)
I have problems only with eigenvalues (I can invert, i can decompose ?
la Cholesky...) with both the codes!

1.
doubletalk server: http://starship.python.net/crew/mhammond/ppw32/ppw32_samples.zip
chapter 6
the client:
Const txtStartupScript = ""
Public BookServer As Object
'    public methods
                'double',
                'interpretString',  # tries exec, then eval
                'execFile',
                'importFile',
                'beginTrappingOutput',
                'endTrappingOutput',
                'getStandardOutput',
                'getViewData',
                'count',
                'createTransaction',
                'getTransaction','getTransactionString',
                'getOneLineDescription',
                'load', 'loadFromText', 'loadFromString',
                'save', 'saveAsText', 'asString',
                'edit','add',
                'getAccountDetails',
                'getAccountList',
                'drawAccountChart',
                'doDelphiCallbackDemo'



Sub InitCOMServer()
    Dim startupScript As String
    
    'called when the program starts
    On Error GoTo InitCOMServer_error
    
    Set BookServer = CreateObject("Doubletalk.BookServer",
RemoteMachine)
    
    On Error GoTo 0
    
    'tell it to capture output for the console
    BookServer.beginTrappingOutput
    
    'if there is an init script, run it
    If txtStartupScript <> "" Then
        On Error GoTo InitCOMServer_StartupScriptError
        BookServer.execFile txtStartupScript
        On Error GoTo 0
    End If
    
    'grab any standard output for the console
    
    Exit Sub
    
InitCOMServer_error:
    Dim msg As String
    msg = "There was an error trying to initialize the BookServer." +
_
            "Please check that it is properly registered and try the
Python " + _
            "test functions first.  The program will now abort."
    MsgBox msg
    End
    
InitCOMServer_StartupScriptError:
    MsgBox "An error occurred running your startup script '" + _
        txtStartupScript + "'. " + _
        "Please check it is present and correct."
End Sub

Sub CloseCOMServer()
    Set BookServer = Nothing
End Sub

Sub TestCOMServer()
    'just to check it is alive
    Dim hopefully_four As Integer
    hopefully_four = BookServer.Double(2)
    MsgBox "2 x 2 = " & hopefully_four & ", so your server is alive"
End Sub

Sub ScriptRun(FileName)
    BookServer.execFile FileName
    On Error GoTo mnuScriptRun_Error:
    Exit Sub
    
mnuScriptRun_Error:
    MsgBox "Error running script:" + vbCrLf + vbCrLf + Err.Description
End Sub
Sub ScriptImport(FileName)
    BookServer.importFile FileName
End Sub
Function ProcessPythonCode(expr As String) As String
    'passes a chunk of text to the server's interpretString
    'method, and returns the result.  If an error occurs,
    'returns the error message.
    
    On Error GoTo mnuProcessPythonCode_Error
    ProcessPythonCode = BookServer.interpretString(expr)
    Exit Function

mnuProcessPythonCode_Error:
    ProcessPythonCode = Err.Description
End Function

Sub ProcessCommand(txtInput As String)
    'gets a one-line string from the input text box.
    'if there is a result, appends it to the output box.
    'Also copy the input command to output to keep a command history
    Dim strInput As String
    Dim strResult As String
    Dim strStandardOutput As String
    Dim strOutText As String  'for building the final textbox data
    
    'VB puts CRLF pairs in the input for multi-line
    'statements; Python just wants line feeds
    strInput = Replace(txtInput, vbCrLf, vbLf)
    
    'multi-line statements need a final line-feed; this is optional
    'fopr one liners.  WE'll add it anyway.
    If Right(strInput, 1) <> vbLf Then strInput = strInput + vbLf
           
    'process it
    strResult = ProcessPythonCode(strInput)
    Debug.Print strResult
    'get output
    strStandardOutput = BookServer.getStandardOutput()
    strStandardOutput = Replace(strStandardOutput, vbLf, vbCrLf)
    Debug.Print strStandardOutput
    
End Sub

Sub InitializeNumeric()
    Dim s As String
    s = "from LinearAlgebra import *"
    ProcessPythonCode s
    s = "from Numeric import *"
    ProcessPythonCode s
End Sub
Sub testNumerical()
    Call InitCOMServer
    Call InitializeNumeric
    Dim s As String
    s = "a = array(([2,1],[1,2]))"
    ProcessCommand s
    s = "inverse(a)"
    ' s = "eigenvalues(a)" Excel freezes
    ProcessCommand s
    CloseCOMServer
End Sub

2.
Python server (assigning the _reg_clsid_ with pythoncom, imported and
then registered with win32com.server.register.UseCommandLine...)
# PythonNumeric.py 
 
class PythonNumeric:
    _reg_clsid_ = "{E34C200E-D95A-49C6-A0F9-F29FF8FF903A}"
    _reg_desc_ = "Python Numeric COM Server"
    _reg_progid_ = "PythonNumeric.Utilities" #for external calling
    _reg_class_spec_ = "PythonNumeric.PythonNumeric"
    _public_methods_ = [ 'InvertMatrix', 'Eigenvalues' ]
    
       
    # Use "print pythoncom.CreateGuid()" to make a new one.
    
    
    def InvertMatrix(self, val):
        import Numeric
        import LinearAlgebra
                        
        return (LinearAlgebra.inverse(Numeric.array(val))).tolist()
 
    def Eigenvalues(self, val):
        import Numeric
        import LinearAlgebra
                        
        return (LinearAlgebra.eigenvalues(Numeric.array(val))).tolist()



VBA Client
Function NumpyMatInverse(mynum As Range)
    Dim numpy As Object
    Dim matrice As Variant
    Range2Array2 matrice, mynum
    Set numpy = CreateObject("PythonNumeric.Utilities")
    NumpyMatInverse = numpy.InvertMatrix(matrice)
    Set numpy = Nothing
End Function
Function NumpyMatEigen(mynum As Range)
    Dim numpy As Object
    Dim matrice As Variant
    Range2Array2 matrice, mynum
    Set numpy = CreateObject("PythonNumeric.Utilities")
    NumpyMatEigen = numpy.Eigenvalues(matrice)
    Set numpy = Nothing
End Function
Sub Range2Array2(arr As Variant, r As Range, Optional base As Integer)
Dim n(1 To 2) As Integer, i As Integer, j As Integer, k As Integer

n(1) = r.Rows.Count + (base - 1)
n(2) = r.Columns.Count + (base - 1)

Select Case n(2)
    Case 0
        ReDim arr(base To n(1))
        For i = base To n(1)
             arr(i) = r.Cells(i - base + 1, 1 - base + 1)
        Next i
    Case Else
        ReDim arr(base To n(1), base To n(2))
        For i = base To n(1)
            For j = base To n(2)
                 arr(i, j) = r.Cells(i - base + 1, j - base + 1)
            Next j
        Next i
End Select

End Sub


From jessica at sonic.net  Sun Apr 21 15:44:28 2002
From: jessica at sonic.net (jessica at sonic.net)
Date: Sun, 21 Apr 2002 19:44:28 GMT
Subject: Jessica Felix Designs - Art and all that Jazz
Message-ID: <comp.lang.python.ac17b7f8@tempest.sonic.net>

I am mercilessly ugly, so I improve you.

"ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
that incorporates the two passions of its owner, Jessica Felix -
art and jazz.

Jessica is an artist whose medium is jewelry. She creates what she calls
"spirited art," specializing in images from indigenous cultures around the world.

http://www.artandallthatjazz.com

119A Plaza Street
Healdsburg, CA  95448
1-707-433-7900

Who will we open after Bernadette kicks the pathetic star's boat?






We fear the polite gardner.



From SBrunning at trisystems.co.uk  Tue Apr  9 11:50:14 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 9 Apr 2002 16:50:14 +0100 
Subject: yet another "Here's to 1.5.2" post
Message-ID: <31575A892FF6D1118F5800600846864DCBCF5B@intrepid>

> From:	Syver Enstad [SMTP:syver-en+usenet at online.no]
> Am I the only one in the entire world that thinks '\t'.join(mylist) is
> elegant and natural?
 
Nope. Took a *little* getting used to, but it seems as natural as breathing
to me now.

Much like a number of Python's *other* idiosyncratic features, past, present
and future...

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




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




From Allen.Ingling at nyu.edu  Tue Apr  2 16:50:06 2002
From: Allen.Ingling at nyu.edu (Allen W. Ingling)
Date: Tue, 02 Apr 2002 16:50:06 -0500
Subject: how do I find a syntax error reported by Mac BuildApplet ?
Message-ID: <3CAA278D.F484DB34@nyu.edu>

How do I find a syntax error reported by the Mac Python BuildApplet
utility ?

I have the Python 2.2 distribution for Mac.  I am running it on  OS
9.2.1.  It includes the BuildApplet utility.   I have written a python
script  which executes in the Python interpreter without error.
However, when I try to build an applet using BuildApplet I received a
"syntax error" message and my script does not compile.  BuildApplet
does not report where in the script the error lies.   Is there any
fast way to automatically identify the line in my script which causes
applet builder to report a syntax error ?

I expect that there really is an error, and that its not reported in
the interpreter because it lies along a conditional branch chain which
is rarely executed.   Having to examine  the entire script myself to
find a syntax error which the compiler detects but does not report
feels incorrect.  What don't I know, or what am I doing wrong here ?

Best,

Allen





From fredrik at pythonware.com  Tue Apr  9 05:20:27 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 09 Apr 2002 09:20:27 GMT
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk>
Message-ID: <vnys8.12$p56.2197@newsb.telia.net>

> So perhaps Guido need a little work with the people skills ;-) and the
> rest of us need to act properly, which includes not getting upset when
> others don't.

"It's all about doing what's best for the little ones."

http://www.theonion.com/onion3811/child-rearing_advice.html

</F>




From msoulier at nortelnetworks.com_.nospam  Tue Apr  9 17:31:30 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 9 Apr 2002 21:31:30 GMT
Subject: shutil.copytree - wishlist
Message-ID: <slrnab6nf1.1du.msoulier@pmerd071.ca.nortel.com>

    Hey people. 

    So, I was just using the very helpful shutil.copytree function, documented
like so:

copytree(src, dst[, symlinks])
    Recursively copy an entire directory tree rooted at src. The destination
directory, named by dst, must not already exist; it will be created.
Individual files are copied using copy2(). If symlinks is true, symbolic links
in the source tree are represented as symbolic links in the new tree; if false
or omitted, the contents of the linked files are copied to the new tree.
Errors are reported to standard output.

    I had symlinks in the original directory that I copied. With the symlinks
boolean turned on, this function created symlinks to the symlinks in the
original directory. 

    As a wishlist feature, I was wondering if an option of whether or not to
perform a readlink of the symlinks can be done prior to creation of those
symlinks, so that instead of pointing to the original directory, the symlinks
would reference what the originals referenced instead. 

ie. 

A -> B

copy A, we get C -> A -> B

Optionally, it would be nice to have C -> B. 

    Is there an official way to make such a wishlist request?

    Thanks as always,

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From maciej at maciejsobczak.com  Thu Apr 11 17:48:05 2002
From: maciej at maciejsobczak.com (Maciej Sobczak)
Date: Thu, 11 Apr 2002 23:48:05 +0200
Subject: Communication between a Python process and a Python cgi script
References: <3CB5E224.4E3DC5BA@olen.to>
Message-ID: <a9508q$1q2$1@sunsite.icm.edu.pl>

Hi,

"Joonas Paalasmaa" <joonas at olen.to> wrote in message
news:3CB5E224.4E3DC5BA at olen.to...
> I have a Python program started by root that does some magic stuff
> every 10 minutes. So the program in an infinite loop of sleep, do stuff,
> sleep, do stuff... How can I make a cgi-script to end the process's
> sleep phase and make it to "do stuff" immediately.

For this kind of interprocess communication, you can try

http://www.maciejsobczak.com/prog/yami/

and in particular:

http://www.maciejsobczak.com/prog/yami/impl/index.html#python

It can be used not only for communication, but also for synchronization.

> I tried signals, but
> it is propably too risky because the process is run as root. Any
> suggestions?

It's rather hard to come up with a scheme that will be really safe here.
FIFO with rights properly set up would probably do, but this adds some
complexity.

Cheers,

--
Maciej Sobczak
http://www.maciejsobczak.com/





From nospam-readonly at getsoft.com  Sun Apr 21 18:27:36 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 08:27:36 +1000
Subject: A simply newbie question about ndiff
Message-ID: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au>

Hi, I've just started trying to do some simple things with Python. I've read
the Tutorial and other Doc's but I'm missing something obvious somewhere.

All I want to do right now is run Tim Peter's ndiff program on some files
and see it's output. I'm using Pyton 2.1.2 Jan 15 2002 on Windows XP using
the Python command line.

I enter 'import difflib' which I assume is necessary and get no complaints.
Then if I type 'ndiff' I get: NameError: name 'ndiff' is not defined. If I
type: 'ndiff file1 file2' I get: SyntaxError: invalid systax.

I've also tried:
>>> s = SequenceMatcher( None, "abcd", "bcde" )
and I get:
NameError: name 'SequenceMatcher' is not defined.

Thanks for your help with these most basic of questions.




From tchur at optushome.com.au  Sat Apr  6 18:06:27 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sun, 07 Apr 2002 09:06:27 +1000
Subject: Pattern for permutations?
Message-ID: <3CAF7F73.9C27F19D@optushome.com.au>

This may be an elementary question but the solution is
not obvious to me.

I want to define a function which takes as its argument a
list which contains one or more sequences, and then does stuff
with every permutation of the elements of those sequences. For
example, a function which works for a list containing three
sequences:

def nestedloops(list_of_sequences):
	for element0 in list_of_sequences[0]:
		for element1 in list_of_sequences[1]:
			for element2 in list_of_sequences[2]:
				# do stuff with each permutation
				print element0, element1, element2

nestedloops([(1,2,3),('one','two','three'),('isa','dalawa','tatlo')])

How does one generalise such a function so that it handles any number 
of sequences in the passed list of sequences i.e. a dynamic number of
levels of nesting? 

It occurred to me that the function could emit dynamic code for the 
"for" loops and then eval() that, but that seems a bit messy and
possibly slow.

Is there a better way?

Tim C




From boud at valdyas.org  Fri Apr 12 08:46:41 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Fri, 12 Apr 2002 14:46:41 +0200
Subject: ANN: Source code for GUI Programming with Python Using the Qt Toolkit
References: <3cb60427$0$71652$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <3cb6d785$0$85458$e4fe514c@dreader3.news.xs4all.nl>

Boudewijn Rempt wrote:

> 
> http://www.xs4all.nl/~bsarempt/python/pyqt_qt2.tgz
> 
>      Source for Python 2.x, Qt 3.x and PyQt 3.x. (Grab this
>      if you're current with the PyQt from
>      http://www.riverbankcomputing.co.uk
> 

As Alex pointed out -- this is wrong. The right place is:

http://www.xs4all.nl/~bsarempt/python/pyqt_qt3.tgz

Sorry for the inconvenience...

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


From cliechti at gmx.net  Wed Apr  3 14:06:04 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 20:06:04 +0100
Subject: Bastion Documentation Error?
References: <a8dsj3$1n4j$1@news.idiom.com>
Message-ID: <Xns91E5CCA838FCDcliechtigmxnet@62.2.16.82>

jimd at vega.starshine.org (Jim Dennis) wrote in 
news:a8dsj3$1n4j$1 at news.idiom.com:
>  I'm a bit confused by the Bastion.Bastion() function.
>  From the documentation I read:
> 
>  """
>  Protect the object object, returning a bastion for the object. Any
>  attempt to access one of the object's attributes will have to be
>  approved by the filter function; if the access is denied an 
>  AttributeError exception will be raised. 
>  """

it says also:
"""
 It must always be used with the rexec module, in order to allow 
restricted-mode programs access to certain safe attributes of an object, 
while denying access to other, unsafe attributes. 
"""

so you must use it within rexec to make it work...

chris

>  ... but playing with it I find that it raises an AttributeError on 
>  any attempt to access any class or instance variables.  

-- 
Chris <cliechti at gmx.net>



From ak at silmarill.org  Sat Apr 20 16:52:26 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 20 Apr 2002 20:52:26 GMT
Subject: Disappearing global variables
Message-ID: <slrnac3lc2.3hm.ak@ak.silmarill.org>

Hello,

When I run book.read() method, instances that were in global namespace
disappear. I made a small function loc() that prints 0 if 'local' is in
global namespace (the name has nothing to do with namespaces - it just
means local list of books stored on hard drive). I run it right before
calling book.read() and it prints 1, I run it from inside book.read()
right after calling it, and it prints 0. I'm using python2.1.

def loc():
    print 'loc',
    print 'local' in globals().keys()

# method of Book class
    def read(self):
        """Read the book."""
        loc()
        arg = ""
        if self not in local.lst:
            # to scroll past gutenberg licence stuff..
            if viewer.endswith("vim"): 
                arg = "+/\*END\*"
            self.download()
        if last.lst and self != last.lst[-1]:
            last.compress()
        if self.gzip:
            self.uncompress()
        last.add(self)
        if viewer.endswith("vim"): 
            # -N not fully vi-compat; -u <vimrc> use vimrc
            arg = " -N -u %s %s " % (VIMRC, arg)
        os.system(viewer + arg + self.l_fname)

You can look at full program at silmarill.org/pybook/pybook.py - it's 
a little over 500 lines.

I have to say I'm really puzzled. Any hints?

 - Andrei

-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From emile at fenx.com  Fri Apr  5 10:02:13 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 05 Apr 2002 15:02:13 GMT
Subject: newbie: declare function before definition?
References: <Pine.LNX.4.44.0204050953040.17812-100000@kramer.thekramers.net>
Message-ID: <V%ir8.172253$Yv2.56285@rwcrnsc54>

David Kramer
> It appears to me that in Python, functions need to be defined before
they
> are called (physically in the file, I mean).  This means any piddly
little
> utility functions have to be before my main routine, and the reader
has to
> wade through them before getting to the main routine.
>
> In some other languages, one can declare a function, and define it
later,
> and the compiler is happy.  Is there a way to do this in Python?

How about:

#!/usr/remote/bin/python

def MyFuncBefore ():
    print "In Before"

def main():
    print "Before"
    MyFuncBefore()
    print "After"
    MyFuncAfter()

def MyFuncAfter ():
    print "In After"

if __name__ == '__main__':
    main()

--

Emile van Sebille
emile at fenx.com

---------



From benji_york at cknainc.com  Fri Apr 26 15:22:43 2002
From: benji_york at cknainc.com (Benji York)
Date: 26 Apr 2002 12:22:43 -0700
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com>
Message-ID: <b10fcc9a.0204261122.5ad30b96@posting.google.com>

Fernando Rodr?uez <frr at easyjob.net> wrote in message news:<olvacuoe4oo1aa5gjj6sa4vsimdom3pabv at 4ax.com>...

> This cgi shouldn't return anything
> and shouldn't change the contents of the browser. Is this possible? O:-)  

An excerpt from http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html:

[Status Code] 204 No Content

The server has fulfilled the request but does not need to return an
entity-body, and might want to return updated metainformation. The
response MAY include new or updated metainformation in the form of
entity-headers, which if present SHOULD be associated with the
requested variant.

If the client is a user agent, it SHOULD NOT change its document view
from that which caused the request to be sent. This response is
primarily intended to allow input for actions to take place without
causing a change to the user agent's active document view, although
any new or updated metainformation SHOULD be applied to the document
currently in the user agent's active view.

The 204 response MUST NOT include a message-body, and thus is always
terminated by the first empty line after the header fields.

(end excerpt)

This looks like what you want.

---
Benji York


From pyth at devel.trillke.net  Wed Apr 17 12:09:58 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 17 Apr 2002 18:09:58 +0200
Subject: Simulating multi-dim array problem - OR - reference confusions
In-Reply-To: <rzev8.12773$hl2.1966@atlpnn01.usenetserver.com>
References: <3CBC991C.7090202@nyu.edu> <mailman.1018999655.9328.python-list@python.org> <rzev8.12773$hl2.1966@atlpnn01.usenetserver.com>
Message-ID: <20020417160958.GA17361@devel.trillke>

> > means that the "name 'b' is bound to 'a[0]'". Internally
> > the reference count of the object a[0] is incremented.
> >
> Well, newbies might avoid mis-perceptions by reading the Python as "name 'b'
> is bound to the object currently bound to 'a[0]'", to avoid any belief that
> anything in a changes as a result of the assignment. In strict terms, a[0]
> doesn't *have* a reference count, as far as I can tell without reading the
> interpreter code, though both a and the object bound to a[0] will have a
> reference count.

ok, so  

- 'a' is a name bound to a list object. 

- 'b=a[0]' binds 'b' to the object that 'a[0]' returns

  (because 'a[0]' is actually a method-call to the list object
   returning (a reference to) the first object of the list)

No matter how the reference counts actually work, 
'b' is a name. And this name is bound to an object.
by issueing 'b=...' you don't change this object but just the
binding. 

The best way to explore this is the prompt. by issueing

>>> locals()

you get the local name bindings. I must say that i
*extremly* appreciate that for all variables,
modules, packages, classes, functions etc ... you have
means of binding a name to these objects. 

sorry if this was redundant and everyone knew this :-)

	holger




From opengeometry at NOSPAM.yahoo.ca  Tue Apr 16 11:03:58 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 16 Apr 2002 15:03:58 GMT
Subject: ideas for the best way to do some math, please??
References: <693d5681.0204160516.541c0bfe@posting.google.com>
Message-ID: <a9hegt$36sns$1@ID-99293.news.dfncis.de>

Ben Carrington <bjc32 at hermes.cam.ac.uk> wrote:
> i am a complete beginner with python, please be gentle :I
> 
> i am trying to evaluate a reatively simple equation but i keep getting
> this error:
> 
> overflowError: math range error
> 
> ...pointing to math.log(p)
>        ...
>        #part3
>        part3 = 1 - math.e**part2
>        p = abs(part3)
> 
>        #part4
>        part4i = math.log(p)
>        part4ii = k*T*part4i

Well, as you know, log(x) -> -infinity as x -> 0.  So, 'p' and 'part2' are
probably 0.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8 CPU cluster, NAS, (Slackware) Linux, Python, LaTeX, Vim, Mutt, Tin


From bokr at oz.net  Wed Apr 24 16:02:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Apr 2002 20:02:52 GMT
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <aa731c$q53$0@216.39.172.122>

On 24 Apr 2002 09:41:22 -0700, zaka07 at hotmail.com (Rad) wrote:

>I am preparing myself to work on extracting data from 4 text files
>(fixed width format) which combined size is about 80GB.  Considering
How wide a fixed with? I.e., how many records? How much duplicate or
unused fluff is there in the records? How much will wind up in your output?
Is this a one-time thing?

>deadlines, costs, and my limited programming knowledge I thought using
>Python/Windows for the job would be the best option for me.  However,
>I am worried about the speed in which Python (me and my hardware) will
>be able to deal with these massive data sets but I am hoping that this
>is still a quicker route then learning C.
>I still haven't received above mentioned files so I can't test the
Are you going to receive a hard disk with the files on it, or tapes,
(or ~125 CDs ;-) ? Is the data pre-sorted in any way that might help you?
What OS are you running? How much disk space and RAM do you have? What CPU?
How many disks on your system? SCSI or IDE? If they give you a HD, will it
be compatible physically & file system wise? Will your power supply hack it?

>time needed to (for example) read a 15GB "file1", filter by few
>variables, and write a resulting subset as a "sub_file1".  Things
>would afterwards get more complicated cause I will have to pullout
>ID's from "sub_file1", remove duplicate ID's create
>"no_dup_sub_file1", match those to ID's in remaining 3 main files and
How many total unique IDs of what type/form? How much memory do you have or
can you expand to? E.g., if the IDs would not make too big a dictionary
(using all the same value of None for every key), then you might be able
to eliminate duplicates as they happen in one pass, and create "no_dup_sub_file1"
without first creating "sub_file1".

>pullout data linked with those ID's.
>
Does this imply random access to data in those three files? Simultaneously?
I.e., you might want to make an index of ID/file positions. And/or you might want
to sort your access IDs by position so you can get everything in one pass. (OTOH,
what ordering is required in your output? What you do might depend.)

But whether you can do all this easily will depend on how much memory you have and
the number of IDs.

All this stuff will be re-inventing data base stuff, so you might want to consider
going that route, as others have suggested. A flat fixed-width source sounds like an
easy importing/loading problem. I haven't tried MySQL on windows, but on Linux is was
not hard to get going, and it might be faster than a fully transactional DB like
postgresql (I don't know any recent benchmarks). I would guess that you could beat
the disk space requirements by writing a special-purpose program, though.

>I have a few weeks to prepare myself before data arrives and my
>question is: am I going the right way about the project, is Python
>(with humanly written code) capable of doing this kind of staff
>relatively quickly?
>
It doesn't sound that complex from what you've described. Just big.
So it will be important to know how much state you have to accumulate
to do the processing you need, and whether that will fit in memory.
E.g., if you have a billion unique IDs, a naive directory approach
is not likely to fly for detecting duplicates.

How you then go about getting end results efficiently will depend
on what the problem really is ;-)

Your best bet on getting help with that (as well as solving it yourself)
is to specify it as precisely and unambiguously as you can (while assiduously
relegating pre-conceived solutions to footnotes or appendices).

If you post a precise definition of the problem, someone might bite.
Also, if you have any options on the way the data is delivered, you might
want to get some opinions on that before it's too late.

Regards,
Bengt Richter


From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 14:30:37 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 20:30:37 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net>
Message-ID: <a9kf12$gi0$1@nets3.rz.RWTH-Aachen.DE>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:a9kebg$ub9$1 at slb5.atl.mindspring.net...
> Philipp Lenssen:
> >can I use something like
> >    verboseGender = (gender == 'm') ? 'male' : 'female'
> >in Python?
>
> No.  Search the archives for "ternary operator" for plenty of
> past discussion.  Google lists 65 posts for comp.lang.python.
>

Sorry. I actually did search but typed "trinary" instead of "ternary". Since
I didn't know what either word means I got confused, and now that I checked
an online translator to german I'm not even sure the word "trinary" exists.
Oh well.

> Besides, for this specific case you should have some error checking,
> because there are genders other than 'male' and 'female' (eg,
> asexual bacteria and hermaphradite earthworms).
>
>   verboseGenderTable = {
>      "m": "male",
>      "f": "female",
>   }
>   verboseGender = verboseGenderTable[gender]
>

(I suppose while the bacteria get too frustrated to use my application, the
hermaphrodites are busy playing with themselves.)




From matt at greenroom.com.au  Wed Apr 10 00:03:29 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Wed, 10 Apr 2002 14:03:29 +1000
Subject: Is Python Sticky?
References: <3cb3a8dd$1@nntp01.splitrock.net>
Message-ID: <3CB3B991.607@greenroom.com.au>

ball at righty.eiu-newman.org wrote:

> Hello!
> 
> For a while "Learn Python" has been on my list of things to do.  Right
> now I have an application that could use some glue between a web based
> interface (in its simplest form a single page with a couple of buttons
> on it), and a shell script (I'm running NetBSD).  Would Python be
> suitable for this kind of thing?  Are there other pieces of the puzzle
> that I'll need?  Any pointers would be most welcome.
> 
> Regards,
>   - Andy Ball.
> 

It sounds to me like python would not be neccessary at all if most of 
the work is done with the shell script. If this is the case then my 
advice would be to simply use a cgi interface through apache, or 
equivalent which would dramatically increase execution speed as the 
python interpreter would not need to be loaded unnecessarily.

Good luck!

Cheers,
   matt



From westernsam at hotmail.com  Mon Apr 15 14:55:55 2002
From: westernsam at hotmail.com (sam)
Date: 15 Apr 2002 11:55:55 -0700
Subject: upgrading to python2.2
References: <292c8da4.0204150521.35d4a842@posting.google.com>
Message-ID: <292c8da4.0204151055.6eb252e4@posting.google.com>

WOW!!

Six hours after reporting a bug it is fixed!!!

You would never get that kind of response from a commercial software vendor

Open Source rules

Thanks a lot Roger

Sam


From max at alcyone.com  Mon Apr  1 14:39:14 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 01 Apr 2002 11:39:14 -0800
Subject: Root access with python
References: <d906b2c1.0203250237.60f6ad12@posting.google.com> <7x663krjtp.fsf@ruckus.brouhaha.com> <a7ndu9$bme$1@panix1.panix.com> <20020325123855.A1716@ibook.distro.conectiva> <mailman.1017071438.28406.python-list@python.org> <Ou4o8.17920$Ii4.1702@atlpnn01.usenetserver.com> <a88h4i$1v8n$1@news.idiom.com>
Message-ID: <3CA8B762.D17E7E4A@alcyone.com>

Jim Dennis wrote:

>   As for your overall "advice."  It's of little value to tell
>   someone: "don't do that" without offering any suggestion about
>   what he or she SHOULD try.
> 
>   Whenever discussing computer system security I avoid
>   proscriptive pronouncements like: "you can't do that" or
>   "no one should ..."

You have a point, but it is still the case that there _are_ solutions
and approaches which are so ill-advised that you are better off stopping
them in their tracks before they go any further down the roads.  Some
approaches are so bad -- i.e., they have caused endless problems over
the years -- that an expert, much less a novice, should avoid them like
the plague.  I'd say setuid scripts is pretty much up there.

A better approach to just, "Don't do that," would be to discourage the
current approach by suggesting that they go back to basics -- instead of
asking how to do the approach they've come to, ask them what fundamental
problem they're trying to solve.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From johnroth at ameritech.net  Tue Apr 30 07:30:41 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 30 Apr 2002 07:30:41 -0400
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk> <just-9E8624.12281026042002@news1.xs4all.nl> <ucjicvlrf7djd3@news.supernews.com> <b0f083db.0204290751.52c145f4@posting.google.com> <Io2j9JAgacz8EwST@jessikat.fsnet.co.uk>
Message-ID: <uct01lso59lu7d@news.supernews.com>

"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:Io2j9JAgacz8EwST at jessikat.fsnet.co.uk...

> the one thing I think is that other languages eg pascal/modula have a
> better/simpler grasp of what humans see in programs. Lexical scope ie
> higher on the page and further left means precede is a 'good' rule.
The
> 'local' idea that anything that is assigned to is 'local' is much
harder
> since that can be realised very distantly. I assume there are reasons
> for the current 'oddness', but they seem to be hidden from us poor
> lusers.
> --
> Robin Becker

If you do that, you can get some very non-intuitive behavior.

For example:

c = 2
def foobar(a):
    b = c
    c = a

In Python, the first assignment fails because 'c' is an
undefined local variable. I've worked in languages
where 'c' would be resolved as 2, and then suddenly
change in the middle of the function to be resolved
as the parameter 'a'. It's a very fertile source of bugs.

Consider this loop:

a = 2
for x in foobar:
    c = a
    a = x

The first time through the loop, 'c' is set to a
value global to the loop, the rest of the time,
it's set to the loop variable (on the preceding pass!)
This is also very non-intuitive behavior.

However, I suspect the primary motivation for this
is optimization. The variable 'dictionary' in functions
is not a dictionary, but a structure that allows faster
access. Given that, the 'if it's assigned to, its a local
variable' rule is quite reasonable.

In any case, any function or method that's big
enough to cause a problem is probably too big.
The 'one page rule' was in existence thirty years
ago when we had to look at programs on paper!

John Roth




From paiarunk at yahoo.com  Mon Apr 22 04:36:04 2002
From: paiarunk at yahoo.com (Arun KUmar)
Date: 22 Apr 2002 01:36:04 -0700
Subject: Making Tk() as modal dialog
Message-ID: <5e430dc.0204220036.3d25f9da@posting.google.com>

hai,

i want to make the Tkinter.Tk() instance as modal window :
 i.e:
     1. It should not be resizable
     2. It should not be movable on the screen
     3. it should not be able to minimize or iconify
    

how to do this?
if any body has already got answers , please help me.

bye
arun


From jb at cascade-sys.com  Tue Apr 30 13:46:21 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 10:46:21 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <kNvz8.12234$xy4.11902@atlpnn01.usenetserver.com>
Message-ID: <3CCED86D.C3173778@cascade-sys.com>

Steve Holden wrote:

> If you're talking about "from __future__ ...", these can only appear before
> any other source lines, and so (bugs notwithstanding) they aren't as
> difficult to detect as you believe. Remember that some imports from
> __future__ modify how the compiler acts, so you can't just switch them on
> and off at random.

Good point.  Thanks for the clarification.

Nevertheless, having to look at the global context for basic things like how
divison works is really screwey and such changes should not be employed
gratuitiously.

> But don't underestimate the cruft that has to be carried forward into new
> implementations either. I appreciate what you're trying to say, but under
> the hood there have been some radical changes in the last two releases. It's
> amazing that such a small amount of code was broken as a result.

I think there are two different things here: (a) bits of cruft carried forward
and (b) necessary radical new features.

The former is in the category of tempting, possibly irresistable but
inconsequential nits that in all likelihood more trouble than it's worth.  Plus
they tend to be highly controversial, like removing map from builtins.  We
should never use "from future" for mere "cruft".

The latter (with a few exceptions (like true division)) are highly desirable,
well thought out new features, e.g., lexical scoping, which correct a SERIOUS
shortcoming in the original language.  Although a radical change that cannot
help but break code, it's well worth the pain to move ahead.

FWIW

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From pinard at iro.umontreal.ca  Mon Apr  8 20:05:34 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 08 Apr 2002 20:05:34 -0400
Subject: Nested scopes hitch
In-Reply-To: <MPG.171bbba6ee73df8c98969f@news.nwlink.com>
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl>
	<1018044224.815568@cswreg.cos.agilent.com>
	<b0f083db.0204080837.625d270f@posting.google.com>
	<1018304015.46066@cswreg.cos.agilent.com>
	<MPG.171bbba6ee73df8c98969f@news.nwlink.com>
Message-ID: <oq1ydprb1t.fsf@titan.progiciels-bpi.ca>

[Jeff Shannon]

> If you want access to variables inside of a function, you should 
> pass those specific variables as arguments.  If you've got lots 
> of arguments in common between lots of functions, assemble the 
> lot into a class.  If neither of these is appropriate, then you 
> should redesign -- you're using a very ugly approach.

I beg to disagree, a bit strongly.

Something is not ugly just because you prefer something else.  I have very,
very neat applications which use globals() and locals(), and they will
loose much if they use arguments all over, or get assembled into classes.

It is possible that the evolution of Python forces me away of these
nice ways, and I'll happily make necessary adjustments in the name of a
greater good.  But I resist the idea that we were wrong or ugly, before.

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




From Jeff.Dubois at westgroup.com  Mon Apr 15 14:37:20 2002
From: Jeff.Dubois at westgroup.com (Dubois, Jeff)
Date: Mon, 15 Apr 2002 13:37:20 -0500
Subject: Regular Expression Pattern Matching "State" Object
Message-ID: <C5723FF8C56CD5119606000347417D8901E17862@eg-msgmbx-a12.int.westgroup.com>

I am looking for a regular expression object\algorithm that does something
like the following:

    1. Initialize  the object with a regular expression pattern.
    2. After initializing the object with a regular expression pattern you
are able to "feed" the object a sequence of characters, one character at a
time, where, after each character is fed to the object, the object indicates
whether or not it is still in a "continuing to match" state, "not a match"
state, or a "match complete" state.  

I am thinking that such a utility may be useful when searching for strings
across DOM nodes for example.

Thanks!

Jeff






From noah at noah.org  Sun Apr 14 14:44:09 2002
From: noah at noah.org (Noah)
Date: Sun, 14 Apr 2002 11:44:09 -0700
Subject: Pseudo-TTY (pty module) help!
Message-ID: <PAENINBOOODFDEPOKPKCOECOCLAA.noah@noah.org>

Does anyone have any experience or samples of use of the pty module?

What I would like to do is fork a child that echos back every
character, but converted to upper-case. The parent would write
out characters then read back and print the results on STDOUT.

I tried using pty.fork. Here is the behavior that I observed.
I successfully get a Master (parent) and Slave (child).
The Master can read and write, but everything it writes comes
back in the read. The Slave can write, but blocks on all reads.
1. The Master gets a file descriptor (fd) for communication with the Slave.
2. The Master can write data to the fd.
3. The Slave can write to its STDOUT.
4. The Master can read the data on fd written by the Slave.
5. The Master alread reads all the data on fd that it wrote.
    it appears to be echoing back, but not from the slave.
    Even if I disable the slave from any writes, the Master's
    data still enchos back.
6. The Slave cannot read from STDIN at all. It blocks.

I use os.write and os.read for all file descriptor operations.
For the slave I use os.write(sys.stdout.fileno(), foo) which I
think is the most correct way.

I have been experimenting with pty, but to no avail.
The documentation could only be briefer if it was absent.
The test_pty.py that comes with the Python installation
also does not fully test all the features, so it is hard
to use that as starter code.

Maybe I could try pty.spwan, but it isn't clear how to use it.
I'm not quite sure how to use the master_read and stdin_read
callback functions.

Yours,
Noah





From hancock at anansispaceworks.com  Mon Apr 29 13:44:02 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 29 Apr 2002 10:44:02 -0700
Subject: Cookie encoding
Message-ID: <3CCD8662.D8E542DC@anansispaceworks.com>

Hi,
I'm trying to put a builtin Python object into an HTTP
cookie for saving state information in a Zope web app.

Zope provides a simple means of doing the actual Cookie
setting (RESPONSE.setCookie()), but it fails silently
if the cookie contains illegal characters (or so it seems).

The Zope docs are pretty hostile towards the idea of using
general picklers in web apps, for security reasons, but
there's a "MiniPickle" module that will only pickle
"safe" data -- i.e. only a select set of Python objects,
like lists and dictionaries are allowed. This seems to
work fine, but the strings generated are apparently
illegal as HTTP cookie values.

So I need to encode the result into something safer. I've
tried the uu.encode, base64.encode, etc. but I'm just
taking shots in the dark (and these didn't work) -- I
figure there has to be a standard answer to this
problem.  There is a "Cookie" module, which looks
interesting, but it seems to want to make the whole
HTTP set-cookie line, not just the legal value.  So
I'm not sure if I can make that play well with Zope.
I picked through the Cookie.py code, and indeed there
is an encoding/quoting function inside it, but it's
a "private" method (begins with "_"), so I gather
it would be unwise to use that directly.

I'm sure I could actually write a custom encoder that
would work, but I'd prefer to do it "right".

Anyway, thanks for any advice on this!

Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
------------------------------------------------------




From eppstein at ics.uci.edu  Sat Apr  6 19:25:56 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 06 Apr 2002 16:25:56 -0800
Subject: a problem with subclassing a dict
References: <a8o0ro$1qoa@r02n01.cac.psu.edu>
Message-ID: <eppstein-683203.16255606042002@news.service.uci.edu>

In article <a8o0ro$1qoa at r02n01.cac.psu.edu>,
 Rajarshi Guha <rxg218 at psu.edu> wrote:

>   I was playing with some code from a tutorial by GvR 
...
> Traceback (most recent call last):
>   File "p2.py", line 22, in ?
>     print md[2]
> KeyError: 2

Your modified __getitem__ function appears not to be called even for the 
successful accesses (e.g. if you put a print statement before the try, 
it never gets printed).  The tutorial mentions a problem like this for 
local and global scope dictionaries, but doesn't say that it happens for 
other kinds of dictionaries too...

Someone more expert than I can tell you why this doesn't work, but
if you really need to define your own dictionary-lookalike class, you 
might also want to look at
<http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/117236>

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From hgs at dmu.ac.uk  Mon Apr 15 09:02:16 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Mon, 15 Apr 2002 14:02:16 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <lkn0w8hqzm.fsf@pc150.maths.bris.ac.uk>
Message-ID: <Pine.GSO.4.40.0204151358340.11562-100000@neelix>

On Sat, 13 Apr 2002, Michael Hudson wrote:

> martin at v.loewis.de (Martin v. Loewis) writes:
>
> > Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
> >
> > > ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/Modules/mathmodule.so: symbol PyExc_OverflowError: referenced symbol not found
> > >
> > > Could someone suggest how to proceed, please?
> >
> > Can you please report the command line option that was used to link
> > the python binary? (delete it, and re-invoke make) PyExc_OverflowError
> > ought to be exported from the Python binary, and there ought to be a
> > gcc command line option to achieve that.
>
> But this is really strange; mathmodule.so must have imported at some
> point, or setup.py would have deleted it...

Hmm.  I don't know.  Anyway, here is the result of re-making the
Python, but not the make test, after removing that file...

Script started on Mon Apr 15 13:57:01 2002
neelix hgs 1 %> gmake
gcc -shared  Modules/mathmodule.o   -o Modules/mathmodule.so
case $MAKEFLAGS in \
*-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py build;; \
esac
running build
running build_ext
building 'regex' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/regexmodule.c -o build/temp.solaris-2.7-sun4u-2.2/regexmodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/regexpr.c -o build/temp.solaris-2.7-sun4u-2.2/regexpr.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/regexmodule.o build/temp.solaris-2.7-sun4u-2.2/regexpr.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/regex.so
WARNING: removing "regex" since importing it failed
building 'pcre' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/pcremodule.c -o build/temp.solaris-2.7-sun4u-2.2/pcremodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/pypcre.c -o build/temp.solaris-2.7-sun4u-2.2/pypcre.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/pcremodule.o build/temp.solaris-2.7-sun4u-2.2/pypcre.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/pcre.so
WARNING: removing "pcre" since importing it failed
building '_hotshot' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/_hotshot.c -o build/temp.solaris-2.7-sun4u-2.2/_hotshot.o
/home/hgs/Python-2.2.1/Modules/_hotshot.c: In function `get_version_string':
/home/hgs/Python-2.2.1/Modules/_hotshot.c:1453: warning: subscript has type `char'
gcc -shared build/temp.solaris-2.7-sun4u-2.2/_hotshot.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/_hotshot.so
WARNING: removing "_hotshot" since importing it failed
building 'strop' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/stropmodule.c -o build/temp.solaris-2.7-sun4u-2.2/stropmodule.o
/home/hgs/Python-2.2.1/Modules/stropmodule.c: In function `strop_atoi':
/home/hgs/Python-2.2.1/Modules/stropmodule.c:769: warning: subscript has type `char'
gcc -shared build/temp.solaris-2.7-sun4u-2.2/stropmodule.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/strop.so
WARNING: removing "strop" since importing it failed
building 'rotor' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/rotormodule.c -o build/temp.solaris-2.7-sun4u-2.2/rotormodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/rotormodule.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/rotor.so
WARNING: removing "rotor" since importing it failed
building 'audioop' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/audioop.c -o build/temp.solaris-2.7-sun4u-2.2/audioop.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/audioop.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/audioop.so
WARNING: removing "audioop" since importing it failed
building 'imageop' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/imageop.c -o build/temp.solaris-2.7-sun4u-2.2/imageop.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/imageop.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/imageop.so
WARNING: removing "imageop" since importing it failed
building 'rgbimg' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/rgbimgmodule.c -o build/temp.solaris-2.7-sun4u-2.2/rgbimgmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/rgbimgmodule.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/rgbimg.so
WARNING: removing "rgbimg" since importing it failed
building 'gdbm' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/gdbmmodule.c -o build/temp.solaris-2.7-sun4u-2.2/gdbmmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/gdbmmodule.o -L/usr/local/lib -lgdbm -o build/lib.solaris-2.7-sun4u-2.2/gdbm.so
WARNING: removing "gdbm" since importing it failed
building '_curses' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/_cursesmodule.c -o build/temp.solaris-2.7-sun4u-2.2/_cursesmodule.o
In file included from /usr/include/curses.h:23,
                 from /usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.3/include/curses.h:5,
                 from /home/hgs/Python-2.2.1/Include/py_curses.h:8,
                 from /home/hgs/Python-2.2.1/Modules/_cursesmodule.c:116:
/usr/include/widec.h:42: warning: `getwc' redefined
/usr/include/wchar.h:326: warning: this is the location of the previous definition
/usr/include/widec.h:43: warning: `putwc' redefined
/usr/include/wchar.h:329: warning: this is the location of the previous definition
/usr/include/widec.h:44: warning: `getwchar' redefined
/usr/include/wchar.h:327: warning: this is the location of the previous definition
/usr/include/widec.h:45: warning: `putwchar' redefined
/usr/include/wchar.h:330: warning: this is the location of the previous definition
gcc -shared build/temp.solaris-2.7-sun4u-2.2/_cursesmodule.o -L/usr/local/lib -lcurses -ltermcap -o build/lib.solaris-2.7-sun4u-2.2/_curses.so
WARNING: removing "_curses" since importing it failed
building '_curses_panel' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/_curses_panel.c -o build/temp.solaris-2.7-sun4u-2.2/_curses_panel.o
In file included from /usr/include/curses.h:23,
                 from /usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.3/include/curses.h:5,
                 from /home/hgs/Python-2.2.1/Include/py_curses.h:8,
                 from /home/hgs/Python-2.2.1/Modules/_curses_panel.c:15:
/usr/include/widec.h:42: warning: `getwc' redefined
/usr/include/wchar.h:326: warning: this is the location of the previous definition
/usr/include/widec.h:43: warning: `putwc' redefined
/usr/include/wchar.h:329: warning: this is the location of the previous definition
/usr/include/widec.h:44: warning: `getwchar' redefined
/usr/include/wchar.h:327: warning: this is the location of the previous definition
/usr/include/widec.h:45: warning: `putwchar' redefined
/usr/include/wchar.h:330: warning: this is the location of the previous definition
gcc -shared build/temp.solaris-2.7-sun4u-2.2/_curses_panel.o -L/usr/local/lib -lpanel -lcurses -ltermcap -o build/lib.solaris-2.7-sun4u-2.2/_curses_panel.so
WARNING: removing "_curses_panel" since importing it failed
building 'fpectl' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/fpectlmodule.c -o build/temp.solaris-2.7-sun4u-2.2/fpectlmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/fpectlmodule.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/fpectl.so
WARNING: removing "fpectl" since importing it failed
building 'zlib' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/zlibmodule.c -o build/temp.solaris-2.7-sun4u-2.2/zlibmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/zlibmodule.o -L/usr/local/lib -lz -o build/lib.solaris-2.7-sun4u-2.2/zlib.so
WARNING: removing "zlib" since importing it failed
running build_scripts
not copying /home/hgs/Python-2.2.1/Tools/scripts/pydoc (up-to-date)
neelix hgs 2 %> exit
neelix hgs 3 %>
script done on Mon Apr 15 13:58:14 2002

        Hugh





From phr-n2002a at nightsong.com  Thu Apr 25 07:44:28 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 04:44:28 -0700
Subject: Faster md5/1-way encryption?
References: <20020424215752.GC15371@teich.Garten.DigitalProjects.com> <3CC7317B.3040208@rogers.com> <mailman.1019718106.22365.python-list@python.org>
Message-ID: <7xhem0vw83.fsf@ruckus.brouhaha.com>

Alexander Skwar <ASkwar at DigitalProjects.com> writes:
> > 	myResults = string.join( myResults, os.linesep )
> 
> Yes, that makes sense.  Thanks!
> 
> > Which would save you creating and destroying huge numbers of (at the end 
> > of processing) very large strings.  If you can avoid that last step 
> 
> Yes, I also noticed, that it gets slower and slower over the time.

I don't understand this.  Do you really want to accumulate a string
containing 100,000 hash results (1.6 megabytes of hashes)?  Or do you
want to hash all the lines together and get one result?

If you want to hash everything together, use

  d = md5.new()
  ...
     (in loop):
     d.update(zeile)

  result = d.digest()


From JamesL at Lugoj.com  Sat Apr  6 00:52:24 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 05:52:24 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAE6DF8.9E655008@engcorp.com>
Message-ID: <Xns91E7DF1A19529JamesLLugojcom@216.148.53.84>

Peter Hansen <peter at engcorp.com> wrote:
> Let me see if I'm clear on this.  You are complaining about the entire
> history of Python releases because of a problem you had with _Unicode_?

_I_ didn't have any problem with Unicode. Python did. Python caused 
_forward_ incompatibilities, not just backward ones.

> When was Unicode added to Python?  

That appears to be an irrelevant rhetorical question. One of the people 
using my program runs a Linux system running 1.5.2 and others seem to run 
2.x versions on Windows. The Windows versions would get e-mail containing 
UTF-8 sequences and after being displayed in a Tkinter text widget, would, 
as far as I can tell, come out as Unicode. Whatever the case, the text 
couldn't be e-mailed on to anyone else because the POP module choked on 
trying to send it over a socket.

> Did you really have such serious compatibility problems in other areas
> that you were forced to abandon the language?

There are other compatibility issues that gave me some problems.

>  Or was it just personal
> distaste at the pace of change?  Don't condemn the fast release cycle
> just because you got screwed on one incompatibility.

I condemn the growing number of versions that are proliferating. More 
versions invite a large probability of forward and backward compability 
problems. That is resulting in a Balkanization of the language. So I have 
concluded the fault is the fast release cycle. I can't change it, I didn't 
get any respectful responses when complaining about it a few months ago and 
still can't get any respectful responses this time around, so why is it so 
surprising that I should put aside the language after 4 some years?


From cliechti at gmx.net  Mon Apr  1 15:04:58 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Apr 2002 21:04:58 +0100
Subject: readlines problems - Newbie
References: <k_1q8.146744$Yv2.44641@rwcrnsc54>
Message-ID: <Xns91E3D6A2BD74Ccliechtigmxnet@62.2.16.82>

"John Purser" <NO_SPAM_jmpurser2 at attbi.com> wrote in
news:k_1q8.146744$Yv2.44641 at rwcrnsc54: 
> I've got a little function:
> def FindLine(Record):
>      im1_Dif.seek(0)
>      im1_Dif.readlines(Record - 1)
>      test = im1_Dif.readline()
>      print test
> 
> When I call it I wind up with the same record in "test" whether I pass
> it 1 or 10.  The readlines line reads a huge chunk of text with no
> regard for lines at all.  I have no idea what is causing this.  Where
> can I find the documentation on the file methods?  I can't find them
> in the documentation. Can anyone tell me what's going on here?

in the docs:
http://www.python.org/doc/current/lib/bltin-file-objects.html

you can see that this is a limit in bytes if even implemented.

you can read all the lines in a buffer:
>>> all = im1_Dif.readlines()
and just work with the last one
>>> test = all[-1]

chris
-- 
Chris <cliechti at gmx.net>



From python at rcn.com  Mon Apr 29 03:16:26 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 03:16:26 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <j4elgz9e9a.fsf@informatik.hu-berlin.de>
Message-ID: <aais2f$l5f$1@bob.news.rcn.net>

"Martin v. L?wis" <loewis at informatik.hu-berlin.de> wrote in message
news:j4elgz9e9a.fsf at informatik.hu-berlin.de...
> "Raymond Hettinger" <python at rcn.com> writes:
>
> >     The module additions would go into Python 2.3.  All six
> > functions would be left intact and raise deprecation warnings in
> > Python 2.3 and 2.4.  The six functions would be removed in Python
> > 2.5 (having allowed two versions and a full year for code updates).
>
> I believe the phasing out of the old names should be even slower. In
> 2.3, introduce the new names; leave the old names intact unchanged. In
> 2.4 and 2.5, raise DeprecationWarnings; in 2.6, remove them.

Agreed.





From huaiyu at gauss.almadan.ibm.com  Fri Apr 26 20:47:36 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 27 Apr 2002 00:47:36 +0000 (UTC)
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to do  math?)
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com> <slrnacjga3.2uv.huaiyu@gauss.almadan.ibm.com> <3CC9C88A.9B447B18@alcyone.com>
Message-ID: <slrnacjt98.4sg.huaiyu@gauss.almadan.ibm.com>

Erik Max Francis <max at alcyone.com> wrote:
>Huaiyu Zhu wrote:
>
>> It appears that another display level is needed, somewhat between repr
>> and
>> str.  Let's call it 'disp' here.  It is intended to be used for
>> display in a
>> controlled way, esp for nested structures.
>
>Just derive a new class and override __str__ and/or __repr__ to do what
>you want.
>
>Adding _another_ stringifying builtin function doesn't seem anywhere
>near worth it.

That's fine if the objects are user defined, but not when you build up
structures from basic building blocks.

- list, tuple and dict default to repr, so __str__ is out of the picture.
- Overriding __repr__ deprives a handy means for identity-based display.
- __repr__ for floats sucks for the purpose of display.

There is no obvious way to define a stringifying function that automatically
recurses over substructures (unlike str), works on builtins, and does not
override __repr__.  If we override __repr__, what is the best way to print
the identity of an *arbitrary* object (not just id)?

Huaiyu


From emile at fenx.com  Sat Apr  6 08:32:14 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 06 Apr 2002 13:32:14 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <%_Br8.20276$nt1.1661286@newsread2.prod.itd.earthlink.net>
Message-ID: <yNCr8.179198$Yv2.61305@rwcrnsc54>

"robin and jim" <robinjim at earthlink.net> wrote in message
news:%_Br8.20276$nt1.1661286 at newsread2.prod.itd.earthlink.net...
>
> Gee, I thought it would be obvious why I was asking such a question
<wink>!
>
> As unit tests run, using the unittest module shipped with Python 2.2,
the
> default behavior is to display a one-line description of each test.
The
> default implementation displays the first line of the test method's
> **docstring**, if available; otherwise the name of the test method
(e.g.,
> "test_my_regular_expression") is displayed.

Yes.  This comes from the __doc__ attribute of the method.

>
> Well, yours truly thought he would be clever and add some additional
> information to this docstring as it became available at run time via
the %
> placeholder mechanism.  Alas, the result is a single-quoted string
that
> unittest does not recognize as a docstring so it displays the test
method's
> name (e.g., "test_my_regular_expression") instead. I, naively, think
that if
> the single quoted result can be re-converted to a triple-quoted
string, it
> will be recognized by unittest.
>

You could change the __doc__ attribute of the method before testing:

import unittest

class IntegerArithmenticTestCase(unittest.TestCase):
    def testAdd(self):  ## test method names begin 'test*'
        self.assertEquals((1 + 2), 3)
        self.assertEquals(0 + 1, 1)
    def testMultiply(self):
        self.assertEquals((0 * 10), 0)
        self.assertEquals((5 * 8), 40)

if __name__ == '__main__':
    testvals = [3,7,4,5]
    IntegerArithmenticTestCase.__dict__['testAdd'].__doc__ = 'testing
with %s' % testvals
    unittest.main()

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From rich at worldsinfinite.com  Wed Apr  3 12:12:05 2002
From: rich at worldsinfinite.com (Rich Harkins)
Date: Wed, 3 Apr 2002 12:12:05 -0500
Subject: How best to write this try/except block?
In-Reply-To: <xIGq8.25783$VQ2.8306517@twister.socal.rr.com>
References: <a8fba8$ln8$1@panix2.panix.com> <xIGq8.25783$VQ2.8306517@twister.socal.rr.com>
Message-ID: <200204031713.g33HDM605101@cvg-65-27-138-13.worldsinfinite.com>

On Wednesday 03 April 2002 12:10 pm, Jeff Youel wrote:
> "Roy Smith" <roy at panix.com> wrote in message
> news:a8fba8$ln8$1 at panix2.panix.com...
>
> > I've got a dictionary d, which may or may not have a value I'm looking
> > for.  If it does, I want to do a few things.  If it doesn't, I want to
> > do a few other things.  I could write:
>
> <snip>
>
> You can avoid the exception handler altogether with:
>
>     value = d.get(key, None)
>     if value:
>         foo = stuff(value)
>     else:
>         foo = otherStuff(key)
>
>
> Jeff

One tweak on this if you want None to be allowed in your dictionary then do 
something like the following:

def fn(d):
	MYBLANKVALUE='MYBLANKVALUE'
	obj=d.get(key,MYBLANKVALUE)
	if obj is MYBLANKVALUE:
		# Do something positive with obj
	else:
		# Do something else...

Because of the is comparison even if "MYBLANKVALUE" is in d then this code 
will still work.  Just don't intern whatever "MYBLANKVALUE" is (I frequently 
use undefined="UNDEFINED").

Rich




From mwh at python.net  Wed Apr 10 06:12:58 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 10 Apr 2002 10:12:58 GMT
Subject: big ints & elliptic curves
References: <a8v338$4sl$1@watserv3.uwaterloo.ca>
Message-ID: <ud6x7amdu.fsf@python.net>

"John" <none at none.com> writes:

> Is there any good library for multi-precision arithmetics in Python? Also
> looking for elliptic curve implementations.

In my experience you'll have a hard time getting acceptable
performance for elliptic curve calculations using Python.  I ended up
using Haskell, I think...

Cheers,
M.

-- 
  ... Windows proponents tell you that it will solve things that
  your Unix system people keep telling you are hard.  The Unix 
  people are right: they are hard, and Windows does not solve 
  them, ...                            -- Tim Bradshaw, comp.lang.lisp


From roy at panix.com  Mon Apr  8 10:47:02 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 08 Apr 2002 10:47:02 -0400
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk> <3CAFAAA4.41EA7CF9@python.org> <+CcOX7A2gWs8EwPR@jessikat.fsnet.co.uk> <3CB1846B.5050109@geneva-link.ch>
Message-ID: <roy-995FAF.10470208042002@news1.panix.com>

Boris Borcic <borcis at geneva-link.ch> wrote:
> I'd rather note the typical elegance and economy of the BDFL's designs;
> did you notice how he wrote "made my kill file" and not "made it to my
> kill file" ?

If he wanted to be truly elegant and economical, he'd have simply written 
"<ploink>".


From jimb417 at totalNOSPAMspeed.net  Sat Apr 20 21:35:47 2002
From: jimb417 at totalNOSPAMspeed.net (Jim)
Date: Sat, 20 Apr 2002 19:35:47 -0600
Subject: IDLE Help problem
Message-ID: <a9t4ui0cnm@enews4.newsguy.com>

Hello to all,

I'm running IDLE with Python 2.2.1. When I click on Help|Python 
Documentation, I get the following error:

Exception in Tkinter callback
Traceback (most recent call last):
  File "/usr/local/lib/python2.2/lib-tk/Tkinter.py", line 1292, in __call__
    return apply(self.func, args)
  File "/usr/local/Python-2.2.1/Tools/idle/EditorWindow.py", line 304, in 
python_docs
    webbrowser.open(self.help_url)
  File "/usr/local/lib/python2.2/webbrowser.py", line 43, in open
    get().open(url, new, autoraise)
  File "/usr/local/lib/python2.2/webbrowser.py", line 38, in get
    raise Error("could not locate runnable browser")
Error: could not locate runnable browser

I have Opera, Konquerer, and Netscape available. What do I need to do to 
get rid of this error. I'm using Mandrake 8.0. Thanks in advance.

Jim

-- 
Registered Linux User #269187
http://counter.li.org/


From tdelaney at avaya.com  Sun Apr 21 18:48:30 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 22 Apr 2002 08:48:30 +1000
Subject: The Python Manual Index
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A279@natasha.auslabs.avaya.com>

> From: Steve Holden [mailto:sholden at holdenweb.com]
> 
> "Philip Swartzleonard" <starx at pacbell.net> wrote in message
> news:Xns91F5997F18E9ARASXnewsDFE1 at 130.133.1.4...
> >
> [Thomas Heller's new search interface]
> > >>
> > DeepBleu || Fri 19 Apr 2002 08:16:23a:
> >
> > > WOW!
> > > Great Job!
> >
> > Indeed. Something like this needs to be added to the python.org
> > interface... i've been wondering for a long time why they 
> don't have a
> > document search function =).
> >
> Well, all they really needed to do was link to
> 
>     http://web.pydoc.org/

Not really - I just tried to use both systems to find the string formatting
specifiers (something which I always have trouble finding in the standard
docs). web.pydoc.org didn't find it for me (using "string format",
"formatting", "format", etc). Thomas' system found it for me immediately
using "formatting".

Thomas - this is wonderful :)

Tim Delaney




From mwh at python.net  Mon Apr 15 07:38:24 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 15 Apr 2002 11:38:24 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org> <7xr8lic2yb.fsf@ruckus.brouhaha.com> <a9b9a9$iam$0@216.39.172.122>
Message-ID: <lku1qdgpc1.fsf@pc150.maths.bris.ac.uk>

bokr at oz.net (Bengt Richter) writes:

> Apparently 1.5.2 and 2.1.x and 2.2.x are semi-official baseline
> versions. Is it/could it be official policy that
> #!/usr/bin/python(15|21|22) links should exist if a site has the
> identified version?

If you run "make install" you should have
$(prefix)/bin/python$(VERSION) where $(VERSION) is x.y (eg. 2.1, 1.5).

Cheers,
M.

-- 
  The Internet is full.  Go away.
                      -- http://www.disobey.com/devilshat/ds011101.htm


From theller at python.net  Tue Apr 16 13:13:36 2002
From: theller at python.net (Thomas Heller)
Date: Tue, 16 Apr 2002 19:13:36 +0200
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr>
Message-ID: <a9hm4h$3eg92$1@ID-59885.news.dfncis.de>

"Ante" <abagaric at rest-art.hr> wrote in message news:a9hhbs$1ksn$1 at as201.hinet.hr...
> from file 'dummy.py'
> ----------------------------
> num = 0
>
> def dummy():
>     num += 1
>
> dummy()
> ----------------------------
>
> now, in the interpreter:
>
> >>> import dummy
> Traceback (most recent call last):
>   File "<string>", line 1, in ?
>   File "C:\downloads\python2.2.1\dummy.py", line 6, in ?
>     dummy()
>   File "C:\downloads\python2.2.1\dummy.py", line 4, in dummy
>     num += 1
> UnboundLocalError: local variable 'num' referenced before assignment
>
>
> Why does this happen?
> If I replace num += 1 with print num, it prints 0 without an error.
> I guess it scans the whole function and check if there's an assignment to a
> name,
> and if there is, assumes it's a local variable. But += isn't supposed to
> move the
> reference but instead change the same variable. Or what did I get wrong?
>
> Ante.
>

Take a look:

C:\>python -OO
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> num = 1
>>> def f():
...   num += 100
...
>>> f()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in f
UnboundLocalError: local variable 'num' referenced before assignment
>>> import dis
>>> dis.dis(f)
          0 LOAD_FAST                0 (num)
          3 LOAD_CONST               1 (100)
          6 INPLACE_ADD
          7 STORE_FAST               0 (num)
         10 LOAD_CONST               0 (None)
         13 RETURN_VALUE
>>>

Behind the scenes, the value bound to 'num' is retrieved,
100 is added 'inplace', and the result will be stored back.
It wouldn't work otherwise - integers are immutable in Python.

Thomas




From mlh at vier.idi.ntnu.no  Mon Apr  8 03:47:07 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 8 Apr 2002 07:47:07 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk> <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk>  <mailman.1018248742.29568.python-list@python.org>
Message-ID: <slrnab2inr.8qu.mlh@vier.idi.ntnu.no>

In article <mailman.1018248742.29568.python-list at python.org>, Lulu of
the Lotus-Eaters wrote:
[snip]
>Of course, I'm sure I'll be justly condmened for preferring:
>
>    fs = (can_be_moved, can_be_sized, can_be_stacked,
>          can_be_coloured, can_be_hidden)
>    major = len([1 for f in fs if f(o)]))
>
>to either :-).

Why not...

from operators impor add
fs = ...
major = reduce(add, [f(o) for f in fs])

OTOH, if you can write the tuple in the fs-assignment, writing a
similar sum shouldn't be that hard ;)

>Yours, Lulu...

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From mwh at python.net  Wed Apr 17 10:37:16 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 14:37:16 GMT
Subject: Counterintuitive Python behavior
References: <mailman.1019052695.4161.python-list@python.org> <lksn5uqtsv.fsf@pc150.maths.bris.ac.uk>
Message-ID: <lkofgiqteo.fsf@pc150.maths.bris.ac.uk>

Michael Hudson <mwh at python.net> writes:

> > The next time a co-worker asks a assignment/binding/references
> > question I'm going to refer them to this message.
> 
> I was hoping more people than the OP would find it useful.  I'll
> probably knock up a version for my web pages (when the starship
> reappears...).

I should say that Alex's answer is well worth referring to too.  I
wonder if he would mind me incorporating it into aforementioned
webpage...

Cheers,
M.

-- 
  If a train station is a place where a train stops, what's a
  workstation?                            -- unknown (to me, at least)


From brueckd at tbye.com  Fri Apr 12 12:57:21 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 12 Apr 2002 09:57:21 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <23328581.1018597365@[192.168.1.102]>
Message-ID: <Pine.LNX.4.33.0204120932060.27572-100000@ns0.tbye.com>

On Fri, 12 Apr 2002, David Eppstein wrote:

> On 4/12/02 8:20 AM -0700, brueckd at tbye.com wrote:
> > The implication is that tuples group bits of often dissimilar but related
> > information that, grouped together, represent a unique entity - sort of a
> > poor man's object instance.
>
> But there's nothing preventing a list from holding inhomogeneous items,
> either.

Yes (and, once again, in that case the usefulness of comprehensions break
down), but in the general sense they are most powerful when there *is*
some common feature to the items in the list. Obviously the common ground
can range from quite specific (a list of integers) to quite broad (a list
of whatevers that this function is capable of processing). If the items
don't have this sort of relation, why do you have them in a list anyway?

>  So I don't see how this is a difference between tuples and lists,
> unless it's just one of standardized style.

The essence of this discussion is that it *is* more than a difference of
just standardized style - tuples were built for one thing and lists
another, albeit with a certain amount of overlap. Tuples work well in the
record-like or dictionary key scenario, while lists are more general
purpose, for use when you need to apply similar processing to a lot of
"stuff".

If tuples don't seem to do all you want it *might* indicate that an
enhancement to the language is in order, but maybe it means that a
different data type would be more appropriate.

Spoons and forks solve similar problems, and in many cases it doesn't
matter which you choose to use even though they were designed to solve
slightly different problems. If your ice cream refuses to cooperate with
your fork it *might* mean you need a better fork, but maybe it means you
should use a spoon (or a spork[1]).

-Dave

[1] http://www.sonic.net/~ian/Spork/spork.faq.html





From anthony at interlink.com.au  Wed Apr 10 09:55:14 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 10 Apr 2002 23:55:14 +1000
Subject: Easter Eggs 
In-Reply-To: Message from Kalle Svensson <kalle@gnupung.net> 
   of "Wed, 10 Apr 2002 00:35:50 +0200." <20020409223550.GC16263@proton.lysator.liu.se> 
Message-ID: <200204101355.g3ADtEZ25492@mbuna.arbhome.com.au>

>>> Kalle Svensson wrote
> [Lulu of the Lotus-Eaters]
> > Who can tell me what's in there.  Someone on c.l.py recently let drop
> > the hint about finding the Zen of Python at the interactive prompt in
> > 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
> > 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
> > magic command was.  Can someone remind me, please?
> "import this", IIRC.

It's in 2.1.2, 2.1.3, 2.2, and 2.2.1.

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From jon+usenet at unequivocal.co.uk  Fri Apr 19 22:41:53 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sat, 20 Apr 2002 02:41:53 -0000
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org>
Message-ID: <slrnac1lbh.nma.jon+usenet@snowy.squish.net>

In article <mailman.1019254077.7316.python-list at python.org>, holger krekel wrote:
> 	 re.sub('\{\{(.*?)\}\}',lambda x: dict[x.group(1)], str)

Need a raw string there, i.e.

  r"\{\{(.*?)\}\}"

Plus 'str' is perhaps not the best choice of variable name ;-)


From marklists at mceahern.com  Thu Apr 11 17:22:44 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 11 Apr 2002 16:22:44 -0500
Subject: emacs python mode completion
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEJPKKAA.marklists@mceahern.com>

Python mode in emacs has the nice syntax coloring and all, but what would be
really great is if I could press tab or something and have it complete
statements.  I found this thread where Guido replied and pointed out
rlcompleter and the like for the interactive interpreter:


http://groups.google.com/groups?hl=en&selm=199801111855.NAA21082%40eric.CNRI
.Reston.Va.US

Is there a tool for emacs that does Python command completion?

Thanks,

// mark





From bellman at lysator.liu.se  Sat Apr  6 21:41:16 2002
From: bellman at lysator.liu.se (Thomas Bellman)
Date: Sun, 7 Apr 2002 02:41:16 +0000 (UTC)
Subject: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org> <Xns91E898777C949JamesLLugojcom@216.148.53.83>
Message-ID: <a8obkc$p3b$1@news.island.liu.se>

James Logajan <JamesL at Lugoj.com> writes:

> Sorry, I meant to use "major" here as in "there are new features or 
> semantic changes in this release that if you use in your code, it wont run 
> at all or as you expect on previous releases". Doesn't matter what you call 
> it, if a 2.1 script can't run at all or as expected on 2.0, that is a major 
> change in the _colloquial_ sense. I consider the nature of the changes 
> between so-called "minor" releases and their frequency a major problem to 
> _me_. I don't think you can realistically "argue" those problems away. I am 
> hoping that you will try to accommodate my experiences into your Python 
> world view before brushing them off.

I think your world-view is broken.  If a program uses new features
in a release, or even just assumes the bug-fixes in that release
to be present, then the program *obviously* won't work the way it
was intended if you run it in a release old enough to *lack* those
features or bug-fixes.  It doesn't matter if those new features
involve the introduction of static typing, if it's a new library
module not present in the older release, a new function, or just
fixing a bug so that Python behaves as documented.  The program
which is written assuming those new features, won't work in a
Python version old enough to not have them, irregardless of how
small the changes are.

If your definition of a minor release is that no program written
for that new release will behave differently when run on the
previous release, then it is a useless release, because it changes
*nothing*, it does not even fix bugs.

100% *forward* compatibility can only be achieved by not changing
anything at all.


-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"God is real, but Jesus is an integer."      !  bellman @ lysator.liu.se
                                             !  Make Love -- Nicht Wahr!


From rjgruet at yahoo.com  Thu Apr 18 16:59:49 2002
From: rjgruet at yahoo.com (Richard Gruet)
Date: Thu, 18 Apr 2002 22:59:49 +0200
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it>
Message-ID: <a9nbuq$916$1@aphrodite.grec.isp.9tel.net>

Alex

Well, this is a very impressive and learned argumentation and I don't doubt
that using operators and built-in functions offer a better polymorphism.
I just meant that to my human eyes, it is easier to memorize when everything
is consistently object oriented and there are only methods (like in Java -
and I prefer Python to Java :-), or the contrary, only functions. Then I
have not to remember which functions are methods and which ones are
built-in. I think this is because a mixed model is more complicated than a
non-mixed one.

And BTW, people would not to have implement any __len__ method if all the
sequence types would derive from a common ancestor (sequence), in which the
(in)famous method would be implemented. Would it be completely nonsensical ?
I have been told that starting from 2.2, new style classes (including most
built-in types) all derive from a common base class 'object'.

NB: I am NOT setting a dispute, but just expressing my feeling from a user
perspective.

Richard


"Alex Martelli" <aleax at aleax.it> wrote in message
news:zwFv8.36236$vF6.1097456 at news2.tin.it...
> Richard Gruet wrote:
>
> > Hi all
> >
> > I wonder why there is no 'length' (or 'len') method - or attribute - in
> > the sequences types. Having to use the built-in function len() doesn't
> > sound very Object-Oriented to me. I understand that the built-in
function
> > len() must be kept for compatibility reasons, but why not to add also
the
> > method len() to the sequence types ? Is there a reason ?
>
> I would be a silly burden on people who want to implement a sequence
> type, to make them define a method called len that is an exact synonym
> for the existing special method __len__.  And it would be an intolerable
> burden, conceptually and in terms of performance, if looking up an
> attribute named 'len' performed strange black-magic contortions (even
> worse if those contortions were specialized for only the cases in which
> said attribute is meant to be called -- remember Python does not draw
> deep distinctions between attributes meant to be called, and others).
>
> It may not SOUND "very Object-Oriented", but having special methods
> "clothed" by means of built-ins and operators is a clever and subtle
> set-up, as I tried explaining in the other post to this thread.  And
Python
> has always preferred the steak to the sizzle, whenever choice was needed:
> part of what makes it such a great language.
>
>
> Alex
>




From phr-n2002a at nightsong.com  Sat Apr 13 13:52:20 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 13 Apr 2002 10:52:20 -0700
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net> <3cb843d0$0$4068$9b622d9e@news.freenet.de> <82bscn4nvz.fsf@acropolis.localdomain>
Message-ID: <7xr8lja3l7.fsf@ruckus.brouhaha.com>

Patrick W <xnexau at yahoo.com.au> writes:
> Do you have a technical reason to actually want interfaces, or are you
> simply experimenting with the concept?
> 
> As far as I can see, which is admittedly not that far, interfaces
> don't really serve a useful purpose in Python. If you want
> polymorphism in statically typed OO languages like C++ and Java, you
> need to design your functions and methods to accept only objects
> descended from a known type. But in Python, the type need not
> (actually cannot) be specified at design time, so what purpose would
> an interface serve? Wouldn't it be better to just inherit
> _implementation_ when necessary, rather than to inherit pure
> interfaces?
> 
> I ask this not to be a prick, but to find out if I'm missing something
> important.

The purposes of interfaces in java is mostly to provide a form of multiple
inheritance.  Java doesn't support multiple base classes, so interfaces
are a workaround for that.  Python supports multiple base classes directly
so it doesn't need a workaround.


From starx at pacbell.net  Tue Apr  2 20:04:03 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 3 Apr 2002 01:04:03 GMT
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de> <MPG.1713c8b646c14e7098968e@news.nwlink.com>
Message-ID: <Xns91E4ADCD75B59RASXnewsDFE1@130.133.1.4>

Jeff Shannon || Tue 02 Apr 2002 01:55:39p:

> In article <3ca9d7d1$1_2 at news.uni-bielefeld.de>, mario at bio58.uni-
> bielefeld.de says...
> 
>> wxPython looked good, but it is so damn slow on my machine 
>> (PPro 200) even on very small programs (hello - world).
> 
> Hm, wxPython seemed to perform adequately for me, on a (Win95) 
> P200 with only 64MB RAM.  
> 
> (Yeah, it's slow, but so is everything *else*, too! ;) )
> 

I use WX too, and it seems that you pretty much pay a fixed price for 
importing the modlue, no matter how much you need or how big your program 
is. I'd think most other frameworks would have the same problem, an easy 
way to tell would be something like:

print "Loading wxPython...",
from wxPython.wx import *
print "done"

... I bet most of your execuation time for a hello world program would be 
between those prints.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From cepl.m at neu.edu  Fri Apr 26 19:28:03 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Fri, 26 Apr 2002 23:28:03 +0000 (UTC)
Subject: iis: comparsion of two lists
In-Reply-To: <20020426224924.GA1318@rehor.surfbest.net>
References: <20020426224924.GA1318@rehor.surfbest.net>
Message-ID: <20020426232802.GA1927@rehor.surfbest.net>

On Fri, Apr 26, 2002 at 10:49:24PM +0000, Matej Cepl wrote:

>    erase = ()
>    for msg in proc_old:
>    		if not(msg in loc_old):
>    			erase.append(msg)

Oh, boy,

I an answer myself. Sorry, for that. Am I correct, that mere
	erase = ()
	erase = filter (lambda m: return not(m in loc_old), proc_old)

would fly?

	Thanks
	
		Matej

-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
I believe in the power of prayer. It's been said: "I would rather
stand against the cannons of the wicked than against the prayers
of the righteous." The prayers of a friend are one of life's most
gracious gifts.
	-- George W. Bush at the 2001 National Prayer Breakfast



From jdhunter at nitace.bsd.uchicago.edu  Mon Apr 22 22:11:46 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Mon, 22 Apr 2002 21:11:46 -0500
Subject: flatten tuple
References: <mailman.1019523041.12884.python-list@python.org>
Message-ID: <m2znzvmae5.fsf@mother.paradise.lost>

>>>>> "Delaney," == Delaney, Timothy <tdelaney at avaya.com> writes:

    >> I want to be able to create a new tuple, which is the expanded
    >> and flattened version x, so that if an element of x is a key of
    >> m, then the new list has that value instead of the key value .
    >> In the example

    Delaney> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/121294
    Delaney> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/118845

Thanks for the tips -- these look helpful.  

It appears that I can not use the yield features of generators with
2.1.3.  I tried 'from __future__ import generators' but still got a
Syntax Error on each of the yield lines.  Is this feature only
supported in the 2.2 line, even with the __future__ ?

It may be time to upgrade my python soon...

Thanks,
John Hunter




From duncan at NOSPAMrcp.co.uk  Tue Apr 30 05:11:50 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 09:11:50 +0000 (UTC)
Subject: Cookie encoding
References: <mailman.1020145372.5420.python-list@python.org>
Message-ID: <Xns920066C44F22Eduncanrcpcouk@127.0.0.1>

Terry Hancock <hancock at anansispaceworks.com> wrote in 
news:mailman.1020145372.5420.python-list at python.org:

> I'm trying to put a builtin Python object into an HTTP
> cookie for saving state information in a Zope web app.

Can you not use SESSION.set() which would save your object locally and just 
put a session id in the cookie?

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From s328940 at student.uq.edu.au  Tue Apr 30 00:42:54 2002
From: s328940 at student.uq.edu.au (Alejandro Dubrovsky)
Date: Tue, 30 Apr 2002 14:42:54 +1000
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <3CCC730A.FC60D13F@alcyone.com> <3ccd1ed9$0$15477$afc38c87@news.optusnet.com.au> <837kmqrvxl.fsf@panacea.canonical.org> <7xit6aqdz5.fsf@ruckus.brouhaha.com>
Message-ID: <3cce1e79$0$15473$afc38c87@news.optusnet.com.au>

On Tue, 30 Apr 2002 09:31:10 +1000, Paul Rubin wrote:


>> Or you can just use named functions, especially if you don't have
>> closure data.
> 
> I thought a named function was just a variable whose value was a lambda:
> 
>   def square(x): return x*x
> 
> is the same as
> 
>   square = lambda x: x*x
> 
> Of course the syntax is different (you can't have control statements in
> a lambda) but I'd have thought if you can't marshal a lambda, you also
> can't marshal a named function.
 
well, that's right, you can't.  that's why i was pickling functions by
storing their .__name__

Alejandro


From aahz at pythoncraft.com  Sat Apr  6 00:06:21 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 00:06:21 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84>
Message-ID: <a8lvod$589$1@panix1.panix.com>

In article <Xns91E7CD754DB67JamesLLugojcom at 216.148.53.84>,
James Logajan  <JamesL at Lugoj.com> wrote:
>Guido van Rossum <guido at python.org> wrote:
>>
>> You could have typed the code in to Python 2.2 and gotten
>> your answer. 
>
>Therein lies the problem: I typed it into Python 2.1 and naturally it 
>didn't work. I loaded 2.1 onto my laptop mid-year last year to be "up to 
>date" and I see that less than a year later my version is out of date. (By 
>comparison, my Linux machine still has 1.5.2 so I can do version checks.)

Python 2.1 (particularly 2.1.2) is a fine release.  There's no
compelling reason to upgrade unless you care about the type/class
unification.  But it's foolish at best to stick your nose into a thread
about language changes without finding what the current state of the
language is.

You seem to make a habit of irresponsible posts.  (Remember your actions
in the comp.distributed brouhaha?)

>Seriously Guido, how often will I need to check back to www.python.org to 
>keep up with the changes you are making? Do you have a schedule or some 
>constraints in place on how fast you are releasing major versions?

The current standard is approximately one release every six months.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From eppstein at ics.uci.edu  Wed Apr 10 20:10:41 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 10 Apr 2002 17:10:41 -0700
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu>
Message-ID: <eppstein-40A0D7.17104110042002@news.service.uci.edu>

In article <a92hl1$1v4c at r02n01.cac.psu.edu>,
 Rajarshi Guha <rxg218 at psu.edu> wrote:

> I got this code to work:
> 
> liter = iter(l)
> while(1):
>   try:
>      print liter.next(), liter.next()
>   except StopIteration:
>      break
> 
> Works fine - is it an efficient way to handle this situation?
> (I could arrange the data into tuples if necessary)

Does it work fine?  What if the list has an odd number of items -- the 
last item won't be printed.

In one situation I had a list of items that I wanted to handle three at 
a time, except that, if the number of items in the list was not a 
multiple of three, I wanted to handle the last two as a pair, or the 
first two and last two both as pairs.  I ended up writing the part that 
handled pairs or triples as a separate function, calling it once for an 
initial pair (if any), then having a separate loop for the triples and 
final pair:

      if len(L) % 3 == 1:
         loopbody(L[:2])
         L = L[2:]
      while len(L) > 0:
         loopbody(L[:3])
         L = L[3:]

this worked ok in my situation but is unsatisfactory as a general 
technique because all the slicing can be slow for long lists -- I guess 
you could use an index into your list instead:

      i = 0
      while i < len(L):
          print L[i:i+2]
          i += 2

Unfortunately due to the non-functionalness of print, I don't see a way 
of writing one print statement that will give the same results as
"print L[i],L[i+1]" without sometimes throwing an exception.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From cnd at ecn.purdue.edu  Tue Apr  2 09:40:32 2002
From: cnd at ecn.purdue.edu (Christopher N. Deckard)
Date: Tue, 02 Apr 2002 09:40:32 -0500
Subject: verifying ssl certificates
References: <3CA88D15.140053DC@ecn.purdue.edu> <7x7knrdv9k.fsf@ruckus.brouhaha.com>
Message-ID: <3CA9C2E0.EA8A2B15@ecn.purdue.edu>

Paul Rubin wrote:
> 
> "Christopher N. Deckard" <cnd at ecn.purdue.edu> writes:
> > I was wondering if anyone has some sample code, or could point me at
> > some documentation, to where I can see how to verify SSL
> > certificates after making an SSL connection via sockets.  Is it as
> > simple as comparing the response of the server to that of something
> > in a file?
> 
> Usually you configure the SSL stack to do this for you.  If the
> certificate doesn't verify, the SSL connection won't even open,
> so your application will never get to see an invalid cert.

In other words:

import socket

cert = open('my_cert')
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect(('host', 'port'))
ss = ssl(s, cert, None)

That last line being the key.  Does that do the certificate
validation?

-Chris


From andersjm at dancontrol.dk  Wed Apr 24 05:02:53 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Wed, 24 Apr 2002 11:02:53 +0200
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net> <3cc6036f$1@news.nz.asiaonline.net> <mailman.1019617724.32183.python-list@python.org>
Message-ID: <3cc67564$0$11927$edfadb0f@dspool01.news.tele.dk>

<jepler at unpythonic.net> wrote:
> On Wed, Apr 24, 2002 at 12:57:02PM +1200, Colin Brown wrote:
> > Thanks guys
> >
> > I did not think to look up this  "documented feature", sorry.

Not your fault, Colin, after all it's in the reference manual which the docs
refers to as "for language lawyers".  A lot of (most?) newbies trip over
this.  (Well I did anyway ;-)

>
> Well, how else would you suggest that "raw" strings should work?  For
> instance, if you want to let
>     r'\' == '\\'
> how will you write the raw string that equals "\\'"?

Python raw strings are delineated like this:
  A raw string ends at the first matching quote character which is not
  preceded by an odd number of consecutive backslashes.

It would be great if I could have written this instead:
  A raw string ends at the first matching quote character.

Much simpler, every bit as useful and - dare I say it - more pythonic
than the current semantics.

The paragraph Andrew quoted from the reference manual could then be
shortened to:

"When an `r' or `R' prefix is present, the backslash character is not
subject to any kind of special interpretation."

A raw string that equals "\\'"?  I would write that r"\'".

I-just-hope-MSOE-will-let-me-post-this-unmangled-ly-y'rs, Anders





From mlh at vier.idi.ntnu.no  Mon Apr 15 07:42:05 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 15 Apr 2002 11:42:05 +0000 (UTC)
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp>  <7xelhhmq5e.fsf@ruckus.brouhaha.com>
Message-ID: <slrnablf4d.5k9.mlh@vier.idi.ntnu.no>

In article <7xelhhmq5e.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
[snip]
>The non-GPL-ness of Python is a main reason I don't have much interest
>in doing significant development work on the Python implementation
>(though I'm willing to make smaller enhancements to fill my own
>requirements, and send in the patches).  On the other hand, Python's
>non-GPL may have attracted some corporate development money, so it's a
>trade-off.  And of course some volunteers do contribute heavily.  It's
>a decision everyone has to make for themselves based on their own
>priorities.

Well -- the Python license is GPL-compatible, so couldn't you just
release a GPL'ed version of it, if that would make you happy? Of
course, the original would still be freely available without the added
burden of the GPL restrictions, but if you're really interested in
freedom, that shouldn't bother you...

<duck and cover>

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From tdelaney at avaya.com  Mon Apr 22 20:49:14 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 23 Apr 2002 10:49:14 +1000
Subject: flatten tuple
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A288@natasha.auslabs.avaya.com>

> From: John Hunter [mailto:jdhunter at nitace.bsd.uchicago.edu]
>
> I want to be able to create a new tuple, which is the expanded and
> flattened version x, so that if an element of x is a key of m, then
> the new list has that value instead of the key value .  In the example

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/121294
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/118845

Always check the cookbook
http://aspn.activestate.com/ASPN/Python/Cookbook/

Tim Delaney




From gtalvola at nameconnector.com  Fri Apr  5 12:38:43 2002
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Fri, 5 Apr 2002 12:38:43 -0500 
Subject: Newbie's 1st Web App decisions!
Message-ID: <BFFEB31BC23CCD44B8924E139E1D87E4124213@MAILBOX.nameconnector.com>

Scrumpy wrote:
> I do like to separate style, layout, app and data and will 
> avoid using PSP, 
> Cheetah etc. completely or at least in ways which integrate 
> app code within 
> HTML pages.

Webware's PSP certainly doesn't require you to mix app code into your pages.
I use PSP extensively, but all of the app logic is contained in other Python
modules -- either WebKit Servlets which are used as base classes for the
PSP, or just as regular old reuseable Python modules.  The only code I
include in the PSP is simple presentation-level stuff like if statements,
for loops, etc. that you need to render the page properly based on the data.
It's all very natural and easy with Webware's PSP.

I think that the *SP model got a bad reputation from Microsoft's Active
Server Pages, where the methods for splitting out your app logic from your
presentation logic require more labor-intensive techniques like writing the
app logic into "middle-tier" compiled Visual Basic COM objects or writing
"scriptlets" in a combination of VBScript and an unusual XML-ish wrapper.
So many people wound up just including the app logic directly in the ASP
pages because it was easier.  But really, Webware's PSP's make it so easy
and natural to separate out your app logic into Python modules that you
can't use that as an argument against PSP.

A better argument against PSP is that it isn't web-designer-friendly.  That
was a primary motiviation for creating Cheetah -- because its syntax is
considered better for non-programmers than PSP's syntax.

- Geoff




From jiba at tuxfamily.org  Tue Apr  2 06:21:32 2002
From: jiba at tuxfamily.org (Lamy Jean-Baptiste)
Date: Tue, 02 Apr 2002 13:21:32 +0200
Subject: Romance object sharing
Message-ID: <mailman.1017766022.7230.clpa-moderators@python.org>

Hi everyone !

I want to introduce Romance.

Romance is an heterogeneous vision of programming / hacking. It relies on
a light dynamic object sharing protocol for sharing between different
programmic languages and programs (such as desktop applications).

In the future Romance will be more than that. It may provide a full
library, including even a graphic toolkit !

Romance is a free (GPL'ed) alternative to .NET, CORBA / Bonobo or
XML-RPC. It allows to share objects between languages, but keeps the
each language's interpreter (contrary to .NET) and look-and-feel (you
code Python, not C#).

Romance currently support Python, Guile (Scheme) and C. Future versions
will support other scripts languages (Perl, Ruby, ...), Lisp / Scheme
(ELisp, ...) and other dynamic languages (Smalltalk, OCaml)...

Romance is available at :
http://oomadness.tuxfamily.org/romance/en/index.html
http://savannah.gnu.org/projects/romance


The use of Romance in Python is totally transparent. Here is an example
of a Romance server :

 # Step 1: import and init Romance !
import romance

# Step 2: create our stuff : the Person factory
class Person:
	def __init__(self, name, proof):
		self.name  = name
		self.proof = proof
    
	def get_description(self):
		return "%s is a %s." % (self.name, self.proof)
  
# Step 3: make it available to any Romance app as "romance.Person"
romance.Person = Person

# Step 4: wait for client !
romance.serve_forever()


And the corresponding client :

# Step 1: import and init Romance !
import romance

# Step 2: access to Romance lib and data (and modify them if you want !!!)
#         Here, create me as a person !
jiba = romance.Person("Jiba", "hacker")

print "My name is :", jiba.name
print "More precision :", jiba.get_description()
print
print "Server is in %s" % jiba.__lang__()

=> My name is : Jiba
=> More precision : Jiba is a hacker.
=>
=> Server is in python

The client is in Python too (because comp.lang.python !), but it can be Guile, or whatever "romantic"
language ;-)

Jiba






From netzapper at magicstar.net  Thu Apr  4 13:10:58 2002
From: netzapper at magicstar.net (A. Jones)
Date: Thu, 04 Apr 2002 18:10:58 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com> <IbXq8.48892$pT1.1388420@news1.tin.it>
Message-ID: <3cac943f.132176618@mammoth.usenet-access.com>

On Thu, 04 Apr 2002 11:56:24 GMT, Alex Martelli <aleax at aleax.it>
wrote:

>You cannot use ANY identifier safely after that exec
>statement, as you have no idea any more what the
>identifier can refer to.  As a bonus, your code crawls,
>since the Python compiler knows it does not know, and
>therefore executes a full identifier look-up at runtime
>rather than recognizing local variables at compile time.
>
>What kind of useful code can you write without being
>able to use ANY identifier safely?  WHY would you ever
>WANT to trample all over your namespace to ensure every
>identifier becomes an utter and total mystery?  Beats me.

I agree with you if it's an arbitrary string being passed to your exec
call .  However, if you have some idea of what the form of the string
is going to be, it doesn't seem so dangerous.  For instance:

for x in range(0, 100):
	exec foo + `x` + " = " + "myClass()"

allows you to easily make a large list of sequentially named
variables, with no chance of stepping on a system call.

Furthermore, you can wrap it inside a class.

class test:
	def wrongHeaded(self, name, value):
	    exec "self."+name+'='+repr(value)
	    print vars()

this produces no exceptions, and doesn't cause any problems, since
vars is never defined... only self.vars is.

Netzapper/Aubrey Jones


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From mats at laplaza.org  Thu Apr  4 11:41:23 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Thu, 04 Apr 2002 16:41:23 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <3cac7f2d.137279317@news.laplaza.org>

On Wed, 03 Apr 2002 17:32:52 -0500, Guido van Rossum
<guido at python.org> wrote:

:I believe this problem should be solved by education (the Zen master
:hits the student on the head with a stick, and the student gets
:enlightenment) rather than by holding back what I believe will be a
:useful feature. 

As a sometime Python educator, I have a problem with this approach: it
has become rather hard to carry the stick on airplanes, if necessary
to fly to the venue.

I'll have to wander down to the neighborhood stupa and ask the lama
for alternative suggestions.

grin :-)

Mats


Mats Wichmann



From cnd at ecn.purdue.edu  Mon Apr  1 11:38:45 2002
From: cnd at ecn.purdue.edu (Christopher N. Deckard)
Date: Mon, 01 Apr 2002 11:38:45 -0500
Subject: verifying ssl certificates
Message-ID: <3CA88D15.140053DC@ecn.purdue.edu>

I was wondering if anyone has some sample code, or could point me at
some documentation, to where I can see how to verify SSL
certificates after making an SSL connection via sockets.  Is it as
simple as comparing the response of the server to that of something
in a file?

Thanks,
-Chris

-- 
--------------------------------------------------------------------
    Christopher N. Deckard      |     Lead Web Systems Developer
      cnd at ecn.purdue.edu        |    Engineering Computer Network
  http://www.ecn.purdue.edu/    |         Purdue University 
---- zlib.decompress('x\234K\316Kq((-J)M\325KM)\005\000)"\005w') ---


From gcash at luncheonmeat.cfl.rr.com  Fri Apr 26 21:51:03 2002
From: gcash at luncheonmeat.cfl.rr.com (gcash)
Date: Sat, 27 Apr 2002 01:51:03 GMT
Subject: modifying classes used by other classes
Message-ID: <1yd129kx.fsf@cfl.rr.com>

OK, here's a rough question in OOP that I've been tangling with...

I've tried googling about it, but I'm not even really sure of my
terminology, so that's not been much help.

So you've got a module that defines 2 classes: spam is the main class to be
used by apps, and it instantiates a whole bunch of "eggs".

class eggs:
   def __init__:
      ...

   def scramble:
      ...

   def drop:
      ...

class spam:
   def __init__:
      ...

   def ham:
      ...
      <instantiate a whole bunch of eggs as sub-objects>
      ...

The problem is you want to override methods of eggs.  You can subclass eggs
all you want, but since spam doesn't use your subclass, you're out of luck.
You're reduced to ugly things like setting methods of eggs equal to
functions you define.

Should I pass the class for eggs as an argument to spam's __init__, which
defaults to the eggs defined by the module?  That seems rather clumsy in a
lot of ways.

A more concrete example: a tkinter tree widget class that creates a
structure of a whole bunch of node instances, each node representing a
folder in the tree.

What would the best way for a user to alter the node class so it binds a
certain mouse click to it's symbol when instantiated?

Another example: a GPS class which creates lists of track, route, and
waypoint instances - how do I best override the __repr__ for the track
class, so it prints the way I want?

What's the best OOP term for classes like "eggs", anyways?  "Helper class"?
It's definitely not a mix-in, obviously.  "Helper class" didn't show up
much on google, though.

-gc

-- 
It was pretty sad, really. When I heard him start to explain 56K, it was
like watching someone try to explain quantum physics to a goat.  At that
point, though, he had no choice it was like the Titanic after the iceberg,
just a matter of time until the people started screaming.


From mlh at vier.idi.ntnu.no  Tue Apr 30 14:35:51 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 30 Apr 2002 18:35:51 +0000 (UTC)
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org>  <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com>
Message-ID: <slrnactp07.pv.mlh@vier.idi.ntnu.no>

In article <fjbz8.7570$M63.5673 at atlpnn01.usenetserver.com>, Steve Holden wrote:
[snip]
>> > Why use type({}) when you can use dict?
>>
>[ ... good reasons ...]
>>
>> 3. And 'dict' is something i often use in local definitions
>>    when working with dictionaries.
>>
>Yeah, well we all have to cut that out and start using "adict" or "thisdict"
>pretty sharpish. This one bites now.

How so? It only "bites" if you actually need access to the dict type
in the same scope where you bind your variable... I've got plenty of
code with stuff like

  file = open('foo.bar')

and it doesn't bother me one bit (except in the cases where 'file'
should have been something more descriptive, I guess ;)

Thes names aren't reserved in any way...

>regards
> Steve

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From aleax at aleax.it  Thu Apr 11 11:44:54 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 15:44:54 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <a941v1$1dr$1@bagan.srce.hr>
Message-ID: <Wbit8.8361$vF6.239936@news2.tin.it>

Ante Bagaric wrote:
        ...
> BTW, i never wrote a single line in Python but it looks darn attractive
> from what I have read so far. Just that this print behaviour brings some
> reminiscence to Fortran output formatting where some thing just happen and
> there's nothing you can do about it :)

print is not really meant for "production-level" output, such as
reporting and so on.  It's meant to be handy for quick & dirty
"let's see what this look like" things, and it is.

> Sure in most cases this feature helps, you dont need zillion " " strings
> between your variables :), but wouldnt it be nice if there was
> python-simple way to just get around it?

There IS -- it's sys.output.write (or more generally the write method
of any file object).  Does no formatting -- it emits a string, and
that's all.

The formatting you do by preparing the string.  The % operator is
most often what you want there.  You don't get full control since
it relies to some extent on the underlying C runtime library, for
example how many digits you get for the exponent in exponential
format is not directly controllable.  However, Python is very good
at string manipulation, so, once you've gotten "more or less" what
you want, typically with operator %, you can (in the relatively
rare cases where you need that) fine-tune things pretty easily.

Formatting a string is one job; putting strings out to some file,
another.  They happen together frequently enough to warrant some
shortcut for simple cases where nothing much is needed, which is
just about what the print statement gives you.  But for the fine
tuning needs, working on strings (for example via regular
expressions, but not necessarily) is a better idea than making
the print statement more complicated in its "joint job".


Alex



From kzaragoza at attbi.com  Mon Apr  1 17:38:39 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: Mon, 01 Apr 2002 22:38:39 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <3CA62CD8.6ABD44D0@accessforall.nl> <m3adsp8zie.fsf@mira.informatik.hu-berlin.de> <3CA6F012.E2631003@accessforall.nl> <m3adspm1kl.fsf@mira.informatik.hu-berlin.de> <3CA79401.CFA41B65@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain> <a8akqh$8bk$0@216.39.172.122>
Message-ID: <slrnaahoic.nsu.kzaragoza@localhost.localdomain>

In article <a8akqh$8bk$0 at 216.39.172.122>, Bengt Richter wrote:
> Of course the 'hidden property' is only rhetoric here, but why not
> make this a clear and controllable unified mechanism instead of a
> collection of special effects? I would like to see an overridable
> __bool__ method (not actually a programmatic property) for the base
> object class, perhaps making deprecation of __nonzero__ possible.

Now this is an idea with some merit.  I can see some real advantages
to being able to define the truth value of a given object, or at least
control its conversion to a bool.  But then, Guido's PEP already shows
this as possible when creating a bool.  If you look at the __new__()
constructor, it checks the truth value of the object when deciding
which global value to return.  If this is how things are implemented
in the C code (and I haven't checked), then __nonzero__() (or
__len__()) is called if defined.  This gives us the control we're
looking for.

> Well, what do you think of the above?

It certainly clarifies matters a bit, but I still have to wonder
whether it's worth the effort.  The way I see it, the mechanism for
the customizations you describe is already present, albeit badly
named.

I certainly don't object to implementing the bool type.  I guess I
just don't see it as a significant advantage.

Kris

-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From doru-cat at ifi.uio.no  Wed Apr 24 04:41:34 2002
From: doru-cat at ifi.uio.no (Doru-Catalin Togea)
Date: Wed, 24 Apr 2002 10:41:34 +0200
Subject: Art and all that Jazz:  are you ugly, I mean, attacking through
  outer powders
In-Reply-To: <3CC62A08.77F6AC93@bellatlantic.net>
References: <comp.lang.python.d42583a5@agozgi.sc.us> <3CC5E850.3080903@ix.netcom.com> <3CC62A08.77F6AC93@bellatlantic.net>
Message-ID: <Pine.LNX.4.21.0204241030190.14813-100000@draslar.ifi.uio.no>

To begin with, I thought the postings from Jesica Felix were commercial
adds from herself. As some have pointed out, her website claims that
somebody is spamming in her name. Her website does indeed look
professional, I really doubt it is a fake.

However, as these messages keep coming in, I began to think that maybe she
is spamming herself, poseing as a victim in order to, maybe,  draw
attention to her business. 

But who can tell? The world is full of lunatics, and the Internett is only
providing another avenue for them to work through. I would rather believe
the story told on her page, unless proven otherwise, that somebody is
spamming in her name. For my own sanity's sake, if nothing else. :-)


Catalin



	<<<< ================================== >>>>
	<<     We are what we repeatedly do.      >>
	<<  Excellence, therefore, is not an act  >>
	<<             but a habit.               >>
	<<<< ================================== >>>>




From claird at starbase.neosoft.com  Sun Apr 28 16:21:28 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 28 Apr 2002 15:21:28 -0500
Subject: Scripting lanuages for laboratory tests
References: <aagkfm$6kf$1@news5.svr.pol.co.uk> <YHRy8.72400$vF6.2161044@news2.tin.it>
Message-ID: <E8135636558E9D45.F00F0E8BA53F58CA.8FF66118989DDBBC@lp.airnews.net>

In article <YHRy8.72400$vF6.2161044 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
			.
			.
			.
>> calling an arbitrary dll
>> 
>> Neither Tcl nor Python supports calling an arbitrary dll.  Tcl supports
>
>http://www.nightmare.com/~rushing/dynwin/calldll-2001-05-20.zip
>
<URL: http://wiki.tcl.tk/ffidl >
-- 

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


From sholden at holdenweb.com  Tue Apr 23 08:27:14 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 08:27:14 -0400
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com>
Message-ID: <Uscx8.93527$T%5.72611@atlpnn01.usenetserver.com>

"Philipp Lenssen" <phil at mrinfo.de> wrote in message
news:aa3fbk$pfb$1 at swifty.westend.com...
> "Petr Prikryl" <Answer.via.news.please.prikrylp at nospam_skil.cz> wrote in
> message news:aa3dti$n0$1 at news.nextra.cz...
> >..
> > In other words, should I prefer writing:
> >     s1 = 'first string'
> > or
> >     s2 = "second string"
> >..
>
> For XHTML string output, I prefer single quotes because I don't have to
> escape double-quotes within the string, e.g.:
> s += '<img src="..." alt="" />'
>

For XHTML and HTML output you would find the triple-quoted forms much
easier. Then you can use EITHER quotes in your output, and include newlines
as well, as in:

s = """%s
<img src="..." alt="A comment">""" % s

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From max at alcyone.com  Sat Apr 13 15:45:24 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 13 Apr 2002 12:45:24 -0700
Subject: ANN: Munch 0.2 -- Embed Python in template text as markup (was ANN: 
 Munch 0.1 ...)
References: <3CB75C38.757A849D@alcyone.com>
Message-ID: <3CB88AD4.F29D4F6E@alcyone.com>

A new version is available:  It supports specifying the prefix character
to something other than @, and properly refusing to expand a None as an
expression evaluation so that proper procedures (functions evaluated for
side effects not return value) can be used.  The full information
follows.

. 

Summary

    Munch is a system for embedding Python expressions and statements
    in template text (or any other unexecuted content); it takes
    source files, processes them, and produces output.  These are done
    via expansions, which are snippets of Python code delimited by
    special characters involving a special character (by default the @
    at sign), and are substituted in the output with the return value
    (for expressions) or with any in-Python printing to stdout that
    takes place (for both expressions and statements).  If the value
    of an evaluated expression is None, nothing is expanded (that is,
    it acts like the interactive Python interpeter).  Textual data not
    explicitly delimited in this way is sent unmolested to the output
    file, allowing Python to be used effectively as a markup language.
    Explicit expansion within Python code is possible.  Also supported
    are m4-style diversions, where output can be systematically
    deferred and recalled.

    Expressions are embedded in text with the '@(...)' notation; as a
    shortcut, simple variables and expressions can be abbreviated as
    '@variable', '@object.attribute', '@array' [index], or
    '@function(arguments)'.  Full-fledged statements are embedded with
    '@{...}'.


Getting the software

    The software is available in a tarball here:
    http://www.alcyone.com/pyos/munch/munch-0.2.tar.gz.

    The official URL for this Web site is
    http://www.alcyone.com/pyos/munch/.


-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From borcis at geneva-link.ch  Thu Apr 11 09:10:39 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 11 Apr 2002 15:10:39 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <slrnab9s50.6g8.mlh@vier.idi.ntnu.no> <3cb57b6b$0$78750$edfadb0f@dspool01.news.tele.dk> <3CB5829C.4030305@geneva-link.ch>
Message-ID: <3CB58B4F.9060908@geneva-link.ch>

>>
>>>
>>>> [] and {} are objects,
>>>>
>>> Yes, as are None and 0.
>>>

> IOW you forgot the empty sequence () and the empty string "".


Raises the interesting question of a locus of functionality of immutable 
objects that would be to dictionaries much like tuples are to lists.

Mmmm...

Maybe "name munching rules" could grow to first-class values, and unify 
with both function parameter binding "action objects" and with (compiled 
code for ?) destructuring-restructuring assignment statements (such as 
a,b = b,a). Some kind of algebra is clearly possible, and would justify 
these being values.

Guess it means knowing the VM is part of a complete python education :(

Boris
--
"Art is a trigram"





From fakeaddress at nowhere.org  Wed Apr 24 23:50:59 2002
From: fakeaddress at nowhere.org (Bryan Olson)
Date: Thu, 25 Apr 2002 03:50:59 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com>
Message-ID: <3CC77D30.7030603@nowhere.org>

Paul Rubin wrote:

 > Note that there's a trivial O(2**64) attack on my authentication since
 > I'm truncating the MAC to 8 bytes.  So if a security difference
 > between HMAC and what I'm doing needs more than O(2**64) work to
 > exploit, it's not really useful to an attacker.  Is that a reasonable
 > way to think of this?

I don't think it's reasonable, no.  The 2**64 attack is to try sending
messages until the defender accepts one.  It cannot be done off-line
because the attacker cannot tell when he has the right MAC.  It's not
comparable to a 2**64 off-line work factor.

--Bryan



From aleax at aleax.it  Tue Apr  9 12:43:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 16:43:01 GMT
Subject: Bitwise/Binary operations in Python < 2.1
References: <slrnab2sb5.j99.petro@lists.bounty.org> <Kgfs8.59392$pT1.1709038@news1.tin.it> <a8si1g$bn7$0@216.39.172.122> <slrnab5nor.p0c.petro@lists.bounty.org>
Message-ID: <pSEs8.382$b62.8262@news1.tin.it>

petro wrote:
        ...

>>>def extractPriorityAndFacility(single32bitQuantity):
>>>    priority = single32bitQuantity & 7
>>>    facility = (single32bitQuantity >> 3) & 0xFFFFFFF
>>Depending on the meaning of "top" ...   ^-- might need 4 here??
>>>    return priority, facility
>>Just a nit, but I suppose it could make a difference.
> 
>     3 gives the right answers:
>     
>     <22> is generated by postfix, and gives 6 (info) for priority, and 2
> (SMTP) for facility. <62> is generated by Leafnode, and gives 6 for
> priority, 7 (news) for facility.
> 
>     Thanks, that did the trick.
> 
>     Could one of you that to me? I don't really understand how that
> worked.

Operator >> shifts a number rightwards (looking at the number as
a string of bit), and operator & applies a "mask" (a bit-by-bit
AND).  Are you familiar with bit manipulations at all?


Alex



From borcis at geneva-link.ch  Mon Apr 15 06:52:39 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 15 Apr 2002 12:52:39 +0200
Subject: Py needs cust nwsclt for mngg eg nntp xrefs ?
Message-ID: <3CBAB0F7.8020608@geneva-link.ch>

Claims :

(1) the PEP process augments the contrast along an axis
of variation of purpose of c.l.p. messaging : pertaining
to PEP-proposal-grade changes of the design of Python,
or not. Some messages do better along that scale, others
do less well.

(2) such discussion relates typically many points, which
need to be cross referenced for facilitating review of
arguments.

(3) part, but part only, of this cross-referencing is
provided by the NNTP thread protocol as supported by
news clients.

(4) with batteries included, python is powerful enough to
provide out-of-the-box a news client and special protocol
tuned to needs of its communal development.

(5) for particular discrete installation of such a special
protocol, an implementation by a custom "transparent firewall"
could be provided, providing eg local through-the-web interface
along with further features; backside of course interfacing
with current-world python "people's protocols".

(6) *the* particular feature of that client protocol, we
would like to have, is facilitating the construction of
button interfaces to earlier messages, eg back-references,
*not* having to follow from the NNTP threading protocol,
even for people using news client UI designed under the
opposite assumption.

(7) the main claim being that *treedagging* is a natural
primitive operation on reasonable contributions to pep
processing. Treedagging is just a portmanteauname we construct
to express an operation that takes forests as purest initial
input (nntp threads form forests) and transputs/outputs DAGs -
Directed Acyclic Graphs.

(8) the purpose is that given such an uniformly distributed
power to knot thread and thread branches, natural dagging experts
will shine well enough, to a "recording representation" s.t.
when ever the BDFL comes to the community as has been the
case with PEP 285 ("I have enough feedback to last me for a
lifetime") he will find it easier to demonstrate his
decision not leaving people with the feeling that he missed
a point.

(9) as a consequence of purpose (8), people eg "ego-driven
by the will of making themselves heard or approved by group
hotshots", will themselves learn that "the system" proportions
their visibility to the quality of their contribution to
diagonal cross-referencing of newsgroups threads, and thus
will themselves save a better attention to that factor.

(10) a further python project would be that of the ngfaqbot.
This means the perfected UI would allow newbies to gain
system credit/weight by risking themselves to assert that point
such-and-such is "#N by the ngfaqbot". This seems to mean that
we should expect the PEP-list to map to a FAQ-list that
would perhaps inherit some of its ordinals.

(11) newbies would be competing with a standard distribution
python AI and site, itself managing the mapping-to-faq of
elementary interventions on the newsgroup using AI software
technology implemented in standard python. Competing newbies
would be rewarded any time an expert confirms that their
intervention saved from a potential or actual AI error and
deserves mention as a motivation to update the latter.

(12) Couldn't this turn into a diatribe on it being below
Python to see it's BDFL using so crude and unsubtle a kill
file technology ? Kill files analogues are everybody's problem
today ! It's Python's mission to provide the newbie user and
the py community expert worker and designer as well, with
a design for newsgroup effective bandwidth enhancement that's
more up to date !

;-)

Boris Borcic
--
"Art is a trigram"



From jeff at ccvcorp.com  Tue Apr 23 13:49:42 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 23 Apr 2002 10:49:42 -0700
Subject: What is an import and where is it on the hard disk
References: <3CC4A51A.A7D770ED@sympatico.ca>
Message-ID: <MPG.172f3e906b1e876f9896c3@news.nwlink.com>

In article <3CC4A51A.A7D770ED at sympatico.ca>, 
francis049 at sympatico.ca says...
> I receive an error message like :
> 
>     Traceback (innermost last) :
>     File : /usr/sbin/up2date, line 9, in ?
>       import rpm
>     Import error : No module named rpm
> 
> 
> Anyone can tell me why is this error happening and give me some details
> ... please ?
 
As Geoff Gerrietts said, it's hard to tell exactly *why* this 
happened when there's no detail given about when or where this 
error occurred.

For an explanation of what import is, and what it means, read the 
Python tutorial chapter on modules, here:

    http://www.python.org/doc/current/tut/node8.html

The error you got implies that there is no rpm.py file in any of 
the directories in your sys.path.  This could be because the file 
is missing, or it could be because your path is bad.
 
Red Hat uses Python 1.5.2 for a number of its system utilities, 
and expects to be able to access that version from 
/usr/bin/python.  Perhaps the most likely scenario here, is that 
you've upgraded Python so that /usr/bin/python is now Python 2.x, 
and the new version of Python doesn't search the old version's 
site-lib for Red Hat's tools.  (Some of those tools are broken 
under 2.x anyhow...)  You can verify this as Geoff suggested, by 
checking the version of python with this shell command:

    $ python -c "import sys; print sys.version"

If this does not show version 1.5.2, then you'll want to remove 
the current version of python, restore 1.5.2 (which may be as 
simple as deleting /usr/bin/python and recreating it as a link to 
/usr/bin/python15, or may require reinstalling the RPM), and then 
installing the new version of python from Red Hat's latest 
python2 RPMs.  This will create the new version as 
/usr/bin/python2, and both can coexist (though you'll need to run 
your scripts as 'python2 myscript.py').

-- 

Jeff Shannon
Technician/Programmer
Credit International


From max at alcyone.com  Thu Apr 25 18:39:58 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 25 Apr 2002 15:39:58 -0700
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <3CC885BE.28611B5@alcyone.com>

Chris Spencer wrote:

>         Okay, I might not be a math genius, but this is most
> definately WRONG.
> What gives?  Anyone have insight into this little "feature" of Python
> floating
> point operations?

It's a feature of _all_ floating pointer operations.  It has nothing to
do with Python specifically.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From aleax at aleax.it  Mon Apr  8 06:24:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 10:24:59 GMT
Subject: newbie question about event based programming
References: <mailman.1018257016.24202.python-list@python.org>
Message-ID: <%des8.59197$pT1.1704440@news1.tin.it>

ashimali at btinternet.com wrote:

> Hello All,
> I'm wondering if there is a way to put an "event listener" (sorry for the
> Java terminology), on a directory. What I'm trying to do is have a python
> script be invoked on files whenever new file/s are  copied/moved to a
> particular directory. Any tips gratefully recieved! Cheers Adam

That depends, first of all, on the functionality of your underlying
operating system -- for example, Windows lets you do that (if your
event loop is coded in a powerful enough way, but that's another
issue).  I don't believe there is any cross-platform way -- it's not
up to the language, alas.  The best you can achieve in most platforms
would be to "poll", say every second, to see if the directory was
modified.  Even just getting a function of yours called every second
is hard to arrange in a cross-platform way.  Fine if you're using
some given GUI toolkit, such as Tkinter (it has an .after method to
schedule a function call, and you need to call it again in the
scheduled function to get periodic repetitions), but there's no way
that works across platforms and toolkits.

Event-driven programming is a great thing, but it's not supported
as well as it could and should be across every platform and toolkit
in the world:-(.


Alex



From distrex at hotmail.com  Sun Apr  7 15:56:42 2002
From: distrex at hotmail.com (distrex)
Date: 7 Apr 2002 12:56:42 -0700
Subject: python newbie
Message-ID: <d9ea2696.0204071156.4284b1f1@posting.google.com>

I'm trying my first script with a module and I can't get it to work
right.  I'm using the time module to make part of the game random and
i just want the number to be at a whole number (no decimals) and I
can't figure out how to make it do that. If someone needs me to post
the script I'd be happy to but if you can help me just from memory
that'd be great too. Thanx in advance.


From sabren at manifestation.com  Tue Apr  2 10:55:58 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Tue, 2 Apr 2002 10:55:58 -0500 (EST)
Subject: ANN: Twisted 0.16.0
In-Reply-To: <3CA83409.8080401@itamarst.org>
Message-ID: <Pine.LNX.4.44.0204021043340.31094-100000@mercury.sabren.com>

On Mon, 1 Apr 2002, Itamar Shtull-Trauring wrote:
 
> For more information on how Twisted helps you run licensed, unpirated
> Open-Source software, visit:
> http://twistedmatrix.com/documents/howto/register

Okay, I see this was posted yesterday (April Fools day), but
the site's still up today. 

Is this for real?

If so, why? How can piracy be a problem for an open source
project?  How does this licensing scheme prevent it? 

Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Give your ideas the perfect home: http://www.cornerhost.com/
 cvs - weblogs - php - linux shell - perl/python/cgi - java
------------------------------------------------------------




From hancock at anansispaceworks.com  Tue Apr 30 16:43:34 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Tue, 30 Apr 2002 13:43:34 -0700
Subject: Python vs. Perl, which is better to learn?
References: <20020501052506.30164.5146.Mailman@mail.python.org>
Message-ID: <3CCF01F6.8A812758@anansispaceworks.com>

I was intrigued to find this was a serious discussion and
not a flame war. :-)

Cool -- I just learned a bit of Perl, which is probably
the reverse of most people's experience.  Note that
Perl naivety is to be expected, therefore.  However, I
narcissistically suppose that a reverse perspective
might be useful to somebody...

My frustration with Perl was that "arrays" and "hashes"
were very like "lists" and "dictionaries" except for
one critical point -- they don't allow nesting!  (There
are more subtle differences too -- apparently Perl does
pretty much everything "by value" instead of
"by reference").  I'm not quite experienced enough
with either to fully appreciate the consequences of
this difference, but I know it must have interesting
consequences.  However, I've started calling dictionaries
"hashes" in my head now -- I suppose that's not good,
but it seems more intuitive to me.  "Dictionary" has
too strong a connection with the thing I use to look
up words in (whether paper or electronic).

Also, although Perl does apparently have "objects",
they aren't so easy as in Python.  All of this has
led me to conclude that I will not be "switching" to
Perl anytime soon.

However, Perl does fill its "duct tape" roll pretty
well -- it's possible to compactly write simple
pieces of code to do quick fixes for various string
and file manipulations.  It also does some common
tasks in nicely compact ways (such as the "process
all lines in all files mentioned on the command line,
or failing that, all the lines coming from stdin",
which is basically what "<>" means).  Basically, it
seems to me that if all you ever do is write scripts
of 200 lines or less, Perl may serve you very well.

But if you're like me, and almost every program you
write quickly grows beyond 1000 lines, because you
can't resist adding new bells and whistles, or because
you are actually fusing a lot of your programs into
one or -- more often -- fusing a lot of other people's
programs (i.e. modules / "the batteries") to do a job,
then you'll quickly learn to love the objects, modules,
and superior data structures of Python.  (Actually
Perl may have these structures,  but they aren't obvious
enough to have been taught to me in learning basic Perl --
whereas in Python you learn and use them right away).

Which wording makes it clear that I took a class on Perl.
This was the only way for me to learn it. I learned Python
by myself in about a week, after which I was able to write
useful (if not elegant) programs.  With Perl I tried this
and couldn't get over a fundamental resistence to learning
what seemed like arcane and annoying syntax.  I finally
punted and signed up for a class, which I almost never
do.  I suppose Perl is now the first programming language
I ever learned this way.

Lists in Python finally made me appreciate what Lisp
programmers were on about. As a long time number-crunching/
scientific applications programmer, in the Basic, Fortran,
C, Pascal tradition, I had little understanding of the
utility of things like lists, hashes or dictionaries,
and so forth.  Python really sold me on them, though.

The regular expressions are nice too. There's a lot
of useful tasks that can be solved much more quickly
using them, "line noise" or no.  Since I don't plan
to switch languages for any major project, the main
impact has been to increase my interest in using
Python's "re" module to solve similar problems.

Unfortunately, Python documenters seem to have a bit
of a chip on their shoulder about regular expressions,
whereas the main promoters of it have been long accustomed
to Perl. So most of the RE explanations have a definite
Perl bias.  Python's implementation is a bit different,
so it can be tricky to learn (i.e. using Perl-centric
sources to learn Python's REs).  Thus, while I'm sure
Python's RE implementation is "good" it may not be
"easy" by dent of being contrary to the only "texts"
I can find to learn it from (but see comments below).
 
Christopher Browne <cbbrowne at acm.org> wrote:
> Mind you, I have found that the "More Pythonic Way" of constructing
> complex regexes by building components, like the following, to have
> merit:
> 
>   digit = '[0-9]'
>   date = '\(' + digit + digit + digit + digit + '/' + digit + digit + '/'
>   date = date + digit + digit + '\)'
> 
> By building the regex out of something with named components, I don't
> have the horridness of line noise like:
>    if ($line =~ /.*\s+\d+.\d+\s+\d+.\d+\s+.\d+.\d+/) {
> 
> Ultimately, the answer is that both approaches have their own merits
> and demerits.  The "Perl thing" tends to be shorter, while the "Python
> thing" tends to be more readable.

Thanks for the example -- I think there is some definite
merit to this.  It occurs to me that you might be both more
clear and more compact (and equally pythonic?) to say:

digit = '[0-9]'
date = '\(' + 4*digit + '/' + 2*digit + '/' + 2*digit + '\)'

for the regular expression, so you don't have to count
the digits.  That's starting to get kind of elegant, I
like it.   Somehow it always helps me to be able to
pronounce my code. ;-D  

At the risk of being terribly Python-centric, I think
Python feels like the "second implementation" of a
solution, where you have the benefit of hindsight to
help you design a simple, elegant, and powerful design
from the start, which you got by using the old, less
elegant solution which was simply the accumulation of
patches to your old problems -- which is what Perl
feels like.  Given that Perl was indeed earlier, and
that experience with it has no doubt driven Python's
design, this may be no slight to Perl, in fact.

My Perl teacher described Perl as being "like English",
full of idioms and synonyms and ambiguities resolved
by context -- an organic, naturally evolved solution
to programming problems. If so, Python is probably like
Esperanto -- artificial, clean, regular, and easy to
learn, though perhaps not as creatively expressive in
itself.

One other odd observation -- I've noticed that Python
is very popular among computer graphics/graphic
artists, whereas Perl remains very popular among
musicians. This is a very unscientific, unsupported
observation -- but I noticed it when trying to find
both for a project I was working on.  I have some
weird theories about why this might be so, but they're
probably wrong. :-)

Given that no one speaks computer languages natively,
the huge difference in learning curve between Esperanto,
which was designed for it, and English (I have only
pity for those who are forced to learn it as a foreign
language -- I've tried to do penance by studying Japanese,
but frankly even that is more logical), is surely
an argument in Python's favor.  (In all fairness, I
should admit that I've never actually tried to learn
Esperanto, which is not so popular in America. Also,
it will always be associated with Red Dwarf and
Rimmer for me).

IMHO, of course...

Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
------------------------------------------------------




From guido at python.org  Fri Apr 12 14:49:17 2002
From: guido at python.org (Guido van Rossum)
Date: Fri, 12 Apr 2002 18:49:17 GMT
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
	 <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com>
	 <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no> <mailman.1018592665.14227.python-list@python.org>
Message-ID: <3CB72CE6.DFAD5C10@python.org>

"David Mertz, Ph.D." wrote:
> Unfortunately, there's more to it than this.  There still is--and
> presumably always will be--a difference between "instances" and other
> types of things.  For example, a list is a mutable sequence that has
> some items in it.  In contrast, an instance object--whether new style or
> old style (and w/ or w/o __slots__)--is a thing that has some
> attributes.  The basic shape of the data will always be different.  The
> type/class unification doesn't change that.

Hm...

So after

    class C(list):
        def appendmany(self, *args):
            for arg in args:
                self.append(arg)

Do you consider a C instance to be a list or an instance?
You might look at the presence of a __dict__ attribute
storing instance attributes as the distinguishing factor.
But then what would you do after

    class D(list):
        __slots__ = ["foo"]

List or instance?  The shape of a D instance is the same
as the shape of a list, but with an extra slot added.

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


From gcordova at hebmex.com  Mon Apr 15 16:31:22 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 15 Apr 2002 15:31:22 -0500
Subject: String interpolation question
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062337A9E10@mail.hebmex.com>

> Hi group,
> 
> [..snip..]
> 


You could use:

print "The value is: %(value)s" % { "value":y[i] }


-gustavo




From ajs at ix.netcom.com  Tue Apr  9 09:51:01 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 9 Apr 2002 09:51:01 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000001c1dfce$4a2ffb00$0334fea9@carol>

 > I do not think the PEP *did* damage - it only revealed that >our
community has problems. I don't understand what >problems.

In my view, something about the exception to the rule
syndrome.

Guido has expressed some welcome self-awareness and
humility on a number of occasions by publicly admitting
frustration at his own inability to articulate the basis for
his design intuitions.

But then seems to go off the wall frustrated when he
opens things up for discussion based on those articulations
and finds that folks don't get it.

My contribution to the div operator discussion - and
I tried to make that clear - was "I have *no* opinion on
how a div operator should behave and understand myself
unqualified to discuss the issue in general terms. I am *not* a peer."

But - I felt very very strongly that the justifications
being articulated did not hold water.  Because the justifications
were largely about "non-programmers*, *newbies* , etc and
etc and pointed to evidence from specific projects about
which I happened to have been quite familiar, the issue was *totally*
outside the realm of the technical, I had the basis to consider myself a
peer in that portion of the discussions and was *convinced* the evidence
cited did provide the basis to draw the conclusions that were being drawn.
So now a discussion of "/" is becoming in my mind an insult to scientific
thinking.

I still have *no* objection to the div operator decision and
am at the same time more convinced than ever that a major articulated basis
for it does not hold water.

Where my own sensibilities happen to explode is when I perceive a good
portion of the community defending Guido's
position when it is based on thin evidence, is non-technical, and represents
no more than one guy's take based on the amount of attention he happens to
have been able to give the matter, as they do when it is based on an
articulated position on technical matters - where I would agree the burden
of proof most definitely shifts.

And when I find there was a Zen masters' wand at our disposal all along.  I
certainly cannot discern *any* basis to understand when we are allowed to
resort to it, and when not.

So I grow to distrust not so much Guido, as the community.

And at this point will be happy enough just to read the release notes.

Art







From boud at valdyas.org  Thu Apr 18 10:58:44 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 18 Apr 2002 16:58:44 +0200
Subject: PyQt QTable
References: <a9mksg$275d$1@news6.isdnet.net>
Message-ID: <3cbedf88$0$305$e4fe514c@dreader1.news.xs4all.nl>

Michael Melchert wrote:

> Hello,
> 
> I am having some difficulties using QTable (PyQt3.0-2.3.1)
> in a python QDialog:
> basically all I want to do is to connect the QTable.clicked() signal to
> a python method [
> self.connect(qtb,SIGNAL('clicked(....)'),self.the_method)] my problem
> arises from the parameters which are defined for
> QTable.clicked(int,int,int,QPoint&). what would be the correct way to
> write the connect() call, especially regarding the QPoint& reference? And
> how would the connected python method be defined?
> 

Haven't tested this -- busy studying for a blasted Java exam -- but
you should be able to something like
QObject.connect(qtb, SIGNAL("clicked(int, int, int, const QPoint &"),
        self.theMethod)

Note the 'const' -- that's needed.

You need to have either five parameters or a variable-number argument
list to theMethod:

def theMethod(self, row, col, button, point):
    bla

or

def theMethod(self, *args):
    bla

or even (but I like to have at least an explicite self):

def theMethod(*args):
    bla
-- 
Boudewijn Rempt | http://www.valdyas.org


From lumber at jack.com  Mon Apr 22 02:29:05 2002
From: lumber at jack.com (Lumberjack)
Date: 22 Apr 2002 06:29:05 GMT
Subject: Programming as literature. Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com>
Message-ID: <Xns91F7EED51FA51lumberjack@216.148.53.83>

Tim Daneliuk <tundra at tundraware.com> wrote:
> Carl Banks wrote:
>> Computer programming is not science.
> 
> Agreed - it is craft.

Computer programming has all the properties of literature, including 
subjective analysis, the existence of "schools of thought," and genres, 
among other things.

Me, I miss the beat era of programming. The bohemians really knew how to 
code some cool prose man.

"The fact was I had the vision...I think everyone has...what we lack is the 
method." - Jack Kerouac


From gcordova at hebmex.com  Thu Apr 11 15:44:22 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 11 Apr 2002 14:44:22 -0500
Subject: Communication between a Python process and a Python cgi scrip
	t
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFBA9@mail.hebmex.com>

> I have a Python program started by root that does some magic stuff
> every 10 minutes. So the program in an infinite loop of 
> sleep, do stuff,
> sleep, do stuff... How can I make a cgi-script to end the process's
> sleep phase and make it to "do stuff" immediately. I tried 
> signals, but
> it is propably too risky because the process is run as root. Any
> suggestions?

?How can it be risky?

If it's risky, and you're trying to activate it with a CGI,
then maybe you shouldn't use a CGI to activate it.

On the other hand, the process could wait for a SIGALRM (alarm).
SO, now, your program can:

1. Setup alarm to signal after X seconds.
2. Sleep forever.
3. Catch alarm, do stuff.
4. Goto 1.

Now, your CGI simply sends a SIGALRM to the process,
which starts the cycle again.

Kinda neato.

-gustavo




From jozef at jozef.joz  Sat Apr 13 17:28:02 2002
From: jozef at jozef.joz (Jozef)
Date: Sat, 13 Apr 2002 23:28:02 +0200
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it>
Message-ID: <a9a7sa$qd3$1@reader05.wxs.nl>

"Alex Martelli" <aleax at aleax.it> wrote:
 
> You're clearly using Python 2.1.something and don't have a:
> 
>     from __future__ import nested_scopes
> 
> at the very top of your module.

Right! So my problem was already solved :-)

> I like FP a lot, and I practise it -- but I eschew lambda in Python in
> most cases: it's limited and goes much against the grain of the
> language.  All it boils down to is, give your functions a name -- in
> most cases a well-chosen name will clarify your code.  E.g.,
> 
> def relatively_prime(a, b): return gcd(a, b) == 1

Thats funny, I wanted to do that but I don't know how to use filter with a
function that takes 2 variables. E.g., to get a list if all numbers less
than 10 relatively prime to 3, I would write

	filter(relatively_prime, range(1, 10)) # aaah, where does the 3 go???

So I learned the lambda way. I also looked at Haskell recently, and  there
(relatively_prime 3) is a valid function that takes only 1 argument, so no
problem there. Just a thought: a nice way to get rid of lambda's and save
a little typing would be if for example

	filter(is_even, alist)
and
	filter(is_even(_), alist)

meant the same thing, and in

	filter(relatively_prime(3, _), alist)

the list elements would be inserted at the '_'. It would of course just be
short for

	filter(lambda x: relatively_prime(3, x), alist)

I'm using Python for about a month so I know I can't be making any sense
to real programmers, but anyway :-)

And thx for the advice


From offer at sgi.com  Wed Apr 17 10:31:01 2002
From: offer at sgi.com (Nhi Vanye)
Date: 17 Apr 2002 14:31:01 GMT
Subject: 2.0*4 ==> Bus error on IRIX 6.5
References: <mailman.1019052034.30806.python-list@python.org>
Message-ID: <a9k0v5$s0anj$1@fido.engr.sgi.com>

$ from stever at bic.mni.mcgill.ca -#195338 | sed "1,$s/^/> /"
>
>
>Hello,
>
>I just built python 2.1.3 on an IRIX 6.5 machine (Origin 200) using MIPSpro
>compilers (Version 7.30) and the test suite fails.
>
>It turns out that multiplying a float by an integer causes a
>bus error:
>
>  $ ./python
>  Python 2.1.3 (#2, Apr 17 2002, 09:37:30) [C] on irix646
>  Type "copyright", "credits" or "license" for more information.
>  >>> 2.0*4 
>  Bus error (core dumped)
>
>
>[Oddly, it works if you multiply two floats or two integers]
>
>From the core dump, I see that the fault happens here:
>
>  #0  0x1004fd7c in float_mul () at floatobject.c:382
>  382             CONVERT_TO_DOUBLE(v, a);
>
>
>I looked on the python.org web pages, FAQs, etc and didn't
>find any similar problem reported.  I couldn't find the way to
>search the python bug tracker, so I had a quick look at the
>summary lines.  There was one about alignment that caught my eye
>(#472568) but the workaround suggested there does not ameliorate
>the situation.
>
>Help?

2.1.1 from May release of Irix Freeware seems to work okay

[ morgaine 6.5.16f-Apr01 ] python2
Python 2.1.1 (#1, Mar 25 2002, 11:44:01) [C] on irix6-n32
Type "copyright", "credits" or "license" for more information.
>>> 2.0*4
8.0
>>> 

There are problems with the Feb release (some files were installed in
the wrong places---seems to only affect the development of extensions),
so that is also likely to be okay.

If you really need the May version right now, mail me and I'll see if
I can get a copy of the latest build to you...

richard.

-- 
-----------------------------------------------------------------------
Richard Offer                         Technical Lead, Trust Technology.
"Specialization is for insects"
__________________________________________http://reality.sgi.com/offer/



From paul.osman at sympatico.ca  Tue Apr 16 12:18:12 2002
From: paul.osman at sympatico.ca (Paul Osman)
Date: Tue, 16 Apr 2002 12:18:12 -0400
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr>
Message-ID: <3CBC4EC4.4080401@sympatico.ca>

Ante,

do some reading about scope... add the following line to dummy.py:

<code>
num = 0

def dummy():
    global num
    num += 1

dummy()
</code>

now in the interpretor:
 >>> import dummy
 >>> print dummy.num
1
 >>> dummy.dummy()
 >>> print dummy.num
2

HTH,

-Paul Osman

Ante wrote:

>from file 'dummy.py'
>----------------------------
>num = 0
>
>def dummy():
>    num += 1
>
>dummy()
>----------------------------
>
>now, in the interpreter:
>
>>>>import dummy
>>>>
>Traceback (most recent call last):
>  File "<string>", line 1, in ?
>  File "C:\downloads\python2.2.1\dummy.py", line 6, in ?
>    dummy()
>  File "C:\downloads\python2.2.1\dummy.py", line 4, in dummy
>    num += 1
>UnboundLocalError: local variable 'num' referenced before assignment
>
>
>Why does this happen?
>If I replace num += 1 with print num, it prints 0 without an error.
>I guess it scans the whole function and check if there's an assignment to a
>name,
>and if there is, assumes it's a local variable. But += isn't supposed to
>move the
>reference but instead change the same variable. Or what did I get wrong?
>
>Ante.
>
>
>






From logiplexsoftware at earthlink.net  Tue Apr 16 16:18:52 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 16 Apr 2002 13:18:52 -0700
Subject: A little advice please? (Convert my boss to Python)
In-Reply-To: <3CBB8ABB.57BAA189@engcorp.com>
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
	<3CBB8ABB.57BAA189@engcorp.com>
Message-ID: <20020416131852.3a61b3ac.logiplexsoftware@earthlink.net>

On Mon, 15 Apr 2002 22:21:47 -0400
Peter Hansen wrote:

> Most of the time when someone concludes Python is not fast
> enough, they have a situation where something is expected
> to run once or twice a day, and it might take ten minutes
> with Python and one minute with C, and they conclude that
> is too slow...  meanwhile the C version is buggy, took 
> four times longer to develop, and nobody likes to maintain
> it because the code is five times longer than the Python
> one would be.  

Especially when the money saved by developing in Python versus C could easily
cover the cost of a faster PC (or two).  It seems so bizarre that there are
people who would spend $10,000 (or more) in developer time to optimize an
in-house program rather than simply spend $1000 on a new PC.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mhammond at skippinet.com.au  Fri Apr 26 21:41:59 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 27 Apr 2002 11:41:59 +1000
Subject: Python.NET?  Jython.NET?  Ni!
In-Reply-To: <3CC9EC6E.1030108@ActiveState.com>
Message-ID: <LCEPIIGDJPKCOIHOBJEPIEMKFIAA.mhammond@skippinet.com.au>

[David Ascher]

> >Is ActiveState's Python for .NET dead?  I can find no reference
> to it on the
> >ActiveState page, other than an old research build for .NET Beta 2.
> >
> >Seems to me the easiest place to start for a .NET version of
> Python would be
> >with Jython.  I haven't looked at the Jython source code, but can someone
> >who has done so comment?
> >
> "Dead" sounds so harsh, in this age of Buffy.  The truth is that no one
> is working on it from ActiveState, and as far as I know Mark isn't
> either.  There are a few people who are interested, however, so there's
> always hope.

I'm not doing all that much on it at the moment, but am keeping it up to
date.  I recently uploaded a version that should work with the release build
of .NET.

http://starship.python.net/crew/mhammond/dotnet is where the latest can be
found.

Mark.





From stephen at xemacs.org  Mon Apr 15 01:11:59 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 15 Apr 2002 14:11:59 +0900
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org>
Message-ID: <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Chris" == Chris Gonnerman <chris.gonnerman at newcenturycomputers.net> writes:

    >> A public domain software can be modified by anyone, and the
    >> result can be made available to someone else under a binding
    >> license.  So, that software was indeed free, but that quality
    >> of being "free" was totally fragile...

    Chris> Exactly.  MIT (X11) and BSD licenses exhibit similar
    Chris> fragility.

Only if you identify modified forms of the software with the original.
The original is forever free.  You can't make the usual arguments
about the non-invasiveness of GPL, and at the same time argue that the
freedom of PD/MIT/BSD code is somehow "fragile", without being wide
open to charges of sophistry.

    >> The GPL wants to help programs at being free in more permanent
    >> ways.

The original code, as a unit, is still free even if it is not
copyleft.  Nothin "unpermanent" about it.  It is available for anyone
who wants to use it.  The "problem", when one exists, is that nobody
wants it, as something better or more compatible is available.  But
that doesn't make the code unfree.  So PD/MIT/BSD is good enough
protection of freedom for many developers.

And the (rms-deprecated!) LGPL is surely good enough for anybody who
only cares about freedom of their own code.  So the GPL is an advocacy
tool, pure and simple.  It's all about affecting licensing decisions
for code that others write, it has nothing to do with protecting
freedom of yours.

    Chris> There are those (uniformly programmers) who hate the GPL
    Chris> because it restricts their ability to "hide" the source if
    Chris> they incorporate GPL'd code in their non-GPL programs.

Consumers also dislike the GPL because it removes their freedom to
sell their privilege of inspection of enhanced derivative software, a
privilege most will never wish to exercise.  They don't hate it the
way Bill Gates does, of course; they just would rather not pay for it
in higher prices and reduced functionality.  (Whether an "intelligent"
consumer would wish to sell the privilege in practice is another
question, although it's trivial to demonstrate that in theory open
source licensing can destroy the market for a given software product.)

Aladdin Ghostscript shows that the set of programs where non-GPL
licensing results in clear benefit to all concerned is non-empty.
With the exception of those who would like to use Aladdin's code to
harm Aladdin's interests without paying for it.  Everybody else gets
Ghostscript for free, including the most recent source code.  (I
understand why neither the OSD nor the FSF admits APFL as a free
license.  But I have never seen a coherent argument that the world
would be a better place if Aladdin Ghostscript, and only Aladdin
Ghostscript, had its licensing policy changed to GPL.)

On the other hand, there are almost certainly other software packages
which would make a much greater social contribution under AFPL terms
than under the GPL---because their authors would actually choose to
publically distribute them under the AFPL, while they keep them
private under the GPL.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From Tom_Good1 at excite.com  Fri Apr 12 18:03:34 2002
From: Tom_Good1 at excite.com (Tom Good)
Date: 12 Apr 2002 15:03:34 -0700
Subject: How to avoid mutual imports
References: <slrnabeai4.lrg.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ac677656.0204121403.2bf3c8d8@posting.google.com>

huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote in message news:<slrnabeai4.lrg.huaiyu at gauss.almadan.ibm.com>...
> I need advice on how to handle the following situation:
> 
> utils1.py:
>     a
> 	b
>     c
>     if __name__ == '__main__':
> 	   from interface import data_initialization  # circular
[snip]

>From your description I'm not sure what the problem is.  That import
is not  actually circular because when utils1.py is imported from
another module, ( _name__ == '__main__' )  will be false, so the
import statement after it will not be executed.

Am I missing something?


Tom


From tjreedy at udel.edu  Wed Apr 24 16:57:49 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 24 Apr 2002 20:57:49 GMT
Subject: Stackless goes Limbo
References: <mailman.1019662503.1913.python-list@python.org>
Message-ID: <h%Ex8.184144$K5.15424154@bin5.nnrp.aus1.giganews.com>

"Christian Tismer" <tismer at tismer.com> wrote in message
news:mailman.1019662503.1913.python-list at python.org...
> Hi all,
>
> after some time of consideration,
> I have decided to move the interface to Stackless
> Python's multitasking towards the ideas behind
> the Limbo Language.
>
> Besides the tasklets, there are "channel" objects,
> which allow for communication between tasklets.
> This approach is simple and powerful. Blocking of
> tasklets takes place on writes when no tasklet
> is listening, or on reads, when no data is available.

Whatever else you do, please consider adding an iterator interface for
reading channels:

for item in channel: whatever()

Also nice, but probably much harder, would be to hijack 'yield' to
write to a channel (for tasklets that write to only one channel, or
which have a default output channel).

Terry J. Reedy





From tony-clpy at lownds.com  Wed Apr 17 22:16:04 2002
From: tony-clpy at lownds.com (Tony Lownds)
Date: 17 Apr 2002 19:16:04 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz>
Message-ID: <ad053a43.0204171816.80fccb0@posting.google.com>

> In theory, it would seem that you could install Unix python
> on Mac OS X together with the Aqua version of Tcl/Tk, and
> run Tkinter stuff.
> 

You sure can! AquaTk is very far along, probably 90%+

> When I tried this recently, however, it didn't work. I
> got a Tkinter window, but it was "dead" -- it didn't
> respond to mouse clicks. Tcl applications run using
> wish worked, but it wouldn't work from Python for
> some reason.
> 

With the second alpha file release of the AquaTk stuff on
http://tcl.sf.net,
this won't happen.

> I'd be interested if anyone happens to know how to
> get this to work!

Essentially, build Python.app according to Python-2.2/Mac/OSX after
adding this to Python/Modules/Setup.local:

_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
        -framework Tcl -framework Tk \
#       -DTK_FRAMEWORK \
        -I/Library/Frameworks/Tcl.framework/Headers/ \
        -I/Library/Frameworks/Tcl.framework/Versions/Current/PrivateHeaders/
\
        -I/Library/Frameworks/Tk.framework/Headers/ \
        -I/Library/Frameworks/Tk.framework/Versions/Current/PrivateHeaders/
\
        -I/usr/X11R6/include/ 

The headers in AquaTk don't work with Darwin's way finding of finding
them. If that gets fixed, then just the -DTK_FRAMEWORK option will be
needed and not the -I options.


-Tony


From jon+usenet at unequivocal.co.uk  Sun Apr 28 11:34:32 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sun, 28 Apr 2002 15:34:32 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no>
Message-ID: <slrnaco5k7.67e.jon+usenet@snowy.squish.net>

In article <aagr7h$5c9$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> I previously saw a posing named "Python vs Php" and sys atht
> "... there are many packages you can download..."
> 
> However, only one package is mentioned: jonpy

Ah, that would be because I wrote that article and I wrote jonpy ;-)

> What is the most popular and promising webpackages for python ?

Obviously, I think mine is, but you can see a list of other packages
at:

  http://thor.prohosting.com/~pboddie/Python/web_modules.html

There is an updated list at:

  http://www.boddie.org.uk/python/web_frameworks.html

but that web site is currently down.


From fredp at mygale.org.nospam  Fri Apr 12 12:32:10 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 12 Apr 2002 16:32:10 GMT
Subject: CGIHTTPServer, revisited
Message-ID: <Xns91EEBD0068607PaCmAnRDLM@161.48.128.20>

Well, seeing there was little in response to my previous post ("Baffled by 
CGI cross-browser issues"), I went back to what I should have started with 
: doing my homework on Google Groups. And what do you know, I'm not alone : 
there's been over a hundred _threads_ on this topic just in the past two 
years.

I read through quite a few of them, and although there was a lot of 
information to be found, I didn't feel there was a definitive answer to the 
question : is CGIHTTPServer somewhat broken under Win9x, even in Python 
2.2? Some say it _does_ work, others it _should_ work, but many seem to be 
stuck just like me.

I've tried a few of the hints found here and there : running scripts with 
'python -u', hacking CGIHTTPServer.py to remove 'fork' and 'popen' 
detection, forcing it to run scripts 'in-process'... Although I'm still 
wondering what Tim Peters-ly-meant by 'having w9xpopen.exe in all the right 
places' :-)

Right now, after upgrading to 2.2.1, I'm still left with the following 
unsatisfactory result : under Win98SE the very minimal script below (and 
its many variants) work fine in Gecko-based browsers (Mozilla, K-meleon), 
always generates a "Network error - connection reset by peer" in Netscape 
4.x (with subsequent output sometimes truncated), and always fails totally 
on submit in MSIE 5.0 and 5.5 at least (server not found etc.).

I'd really like to understand how/why this happens. Can't be just a 
browser-side issue, can it ?...

#-----------------------------
# test.py
import cgi   
form = cgi.FieldStorage()
s=''
for name in form.keys():
    s = s + "Input: " + name + " --> " + form[name].value + "<p>\n"

print """
<head><title>Test form</title></head><body>
<FORM METHOD=POST action="/cgi-bin/test.py">
<INPUT TYPE=text NAME=spam>
<br>
<INPUT TYPE=RADIO NAME="page_type" VALUE="front" CHECKED>Front
<INPUT TYPE=RADIO NAME="page_type" VALUE="back">Back
<br>
<input type=submit name="sub_go" value="Go">
<p>
%s
</body></html>
""" % s
#-----------------------------

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


From op73418 at mail.telepac.pt  Mon Apr  1 14:30:33 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 01 Apr 2002 20:30:33 +0100
Subject: None not None in Excel COM?
Message-ID: <s4dhaukcq7v1v7gdgi9vcb9a46tiu6kv20@4ax.com>

Maybe someone can help me out here.

Im using Excel via Python COM and have already run makepy on its type
library.

The problem is the following:

Let sheet be a reference to an Excel worksheet and suppose
sheet.Cells(row, 1) returns None for a given row > 1. The problem is
that if I do

sheet.Cells(row, 1) == None
sheet.Cells(row, 1) is None

both return 0. What am I doing wrong?

Thanx for all the help,
Gon?alo Rodrigues


From Sibylle.Koczian at Bibliothek.Uni-Augsburg.de  Wed Apr 10 08:09:26 2002
From: Sibylle.Koczian at Bibliothek.Uni-Augsburg.de (Sibylle Koczian)
Date: Wed, 10 Apr 2002 14:09:26 +0200
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au> <a8v0mo$j50$1@newsg1.svr.pol.co.uk>
Message-ID: <a914t4$g70$1@rzmux02.rz.uni-augsburg.de>

>I'm a newbie myself, so this is probably wrong, but wouldn't this be *even*
>better?
>
>x = 2
>def wrong(num):
>    x = x+1
>    print x
>
>wrong(x)
>wrong(x)
>
>
Just tested with Python 2.2:

>>> x = 2
>>> def wrong(num):
...  x = x + 1
...  print x
...
>>> wrong(x)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "<interactive input>", line 2, in wrong
UnboundLocalError: local variable 'x' referenced before assignment

I think the exception is new (result of nested scopes?), but it's quite
probable you meant to write

x = 2
def wrong(num):
    num = num + 1
    print num

Result:

wrong(x)
3
wrong(x)
3

for the reason already stated: call by value, which works on a copy of the
parameter inside the function and doesn't change the parameter itself.
(Would be the same in C, by the way.)

So the "return num + 1" is really crucial, if you want your x to change.

HTH,
Koczian

--
Dr. Sibylle Koczian
Universitaetsbibliothek , Abt. Naturwiss.
D-86135 Augsburg
e-mail : Sibylle.Koczian at Bibliothek.Uni-Augsburg.DE





From neal at metaslash.com  Sat Apr 20 18:59:49 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Sat, 20 Apr 2002 18:59:49 -0400
Subject: Problem embedding python on Red Hat Linux 7.2
References: <a9q49n$tq7$1@news.datasync.com> <m3662mngkr.fsf@yakisoba.forte-intl.com>
Message-ID: <3CC1F2E5.11886F77@metaslash.com>

Ronald Cole wrote:
> 
> *newbie alert*
> 
> I have a program that calls Py_Initialize() and I cannot compile it
> because I cannot seem to find the python library (-lpython?) in
> /usr/lib to link it with.  Here's what Red Hat figured I needed:
> 
> $ rpm -qa | grep python
> python-xmlrpc-1.5.1-7.x.3
> python-1.5.2-35
> python-devel-1.5.2-35
> rpm-python-4.0.4-7x
> python-popt-0.8.8-7.x.2
> 
> and I've scanned distribution cd's 1 and 2 looking for it but can't
> seem to find any sign of a lib*.{a,so} in the rpms.
> 
> Please, someone take pity and clue me!

$ locate libpython

should tell you it's in /usr/lib/python1.5/config/libpython1.5.a
That's where mine is.

Neal


From jb at cascade-sys.com  Tue Apr 30 15:59:06 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 12:59:06 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no> <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>  <mailman.1020033011.19748.python-list@python.org> <slrnactoru.pv.mlh@vier.idi.ntnu.no>
Message-ID: <3CCEF78A.44CD7849@cascade-sys.com>

Magnus Lie Hetland wrote:

> James J. Besemer wrote:
> >
> >Made me look.
>
> Gullible, anyone? <wink>

Yup nailed me good on that one.

I realized just AFTER pressing send.  A little less coffee that day and I may have
spared myself that minor embarrassment.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From dsavitsk at e-coli.net  Tue Apr 23 14:55:23 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 23 Apr 2002 18:55:23 GMT
Subject: dictionary method question
Message-ID: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com>

for some dictionary d is there any relationship between the returns from
d.keys() and d.values()?  that is, to get them in some arbitrary but
matching order, do i need to loop over d.items() or can i count on a
similarity of the above? in small tests this seems to be the case.

thanks,

-doug




From skip at pobox.com  Mon Apr  1 09:00:44 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 1 Apr 2002 08:00:44 -0600
Subject: PyOS -- was: Re: ANN: cage 1.0 -- Cellular automata engine in Python
In-Reply-To: <20020401100837.GA1536@blackscarab>
References: <3CA5194E.D5AB00E1@alcyone.com>
        <20020401021435.GF333@lilith.hqd-internal>
        <20020401100837.GA1536@blackscarab>
Message-ID: <15528.26636.26194.935490@12-248-41-177.client.attbi.com>

    sasoft> there was (the link is dead): "spyLinux" at
    sasoft> http://dilbert.physast.uga.edu/~andy/minilinux.html

Try:

    http://bluesine.com/Projects/spylinux

Courtesy of a little googling.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)



From bbollenbach at shaw.ca  Thu Apr 11 21:52:57 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 01:52:57 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <a941v1$1dr$1@bagan.srce.hr> <Wbit8.8361$vF6.239936@news2.tin.it>
Message-ID: <Z5rt8.16020$Kq4.678935@news2.calgary.shaw.ca>

In article <Wbit8.8361$vF6.239936 at news2.tin.it>, Alex Martelli wrote:
> Ante Bagaric wrote:
>> Sure in most cases this feature helps, you dont need zillion " " strings
>> between your variables :), but wouldnt it be nice if there was
>> python-simple way to just get around it?
> 
> There IS -- it's sys.output.write (or more generally the write method
> of any file object).  Does no formatting -- it emits a string, and
> that's all.

Not to mention, sys.stdout.write() is considerably faster than print:

_printspeed.py_

#!/usr/bin/python
for i in range(1000000):
    print "hello, world"

_writespeed.py_

#!/usr/bin/python
import sys
faster = sys.stdout.write
for i in range(1000000):
    faster("hello, world\n")
    
$ time ./printspeed.py

[snip 1,000,000 lines of output]

real    1m11.301s
user    0m23.140s
sys     0m6.620s

$ time ./writespeed.py

[snip 1,000,000 lines of output]

real    0m54.578s
user    0m12.250s
sys     0m6.730s

You probably won't care about this unless you working on large amounts
of data, but it's worth keeping in mind.


--
Brad Bollenbach


From chris.gonnerman at newcenturycomputers.net  Sat Apr 13 10:51:50 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 13 Apr 2002 09:51:50 -0500
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <005001c1e2aa$3d554280$0101010a@local> <20020413052610.GA5561@lilith.my-fqdn.de>
Message-ID: <005a01c1e2fa$d44249e0$0101010a@local>

----- Original Message -----
From: "Gerhard H?ring" <gerhard at bigfoot.de>


> * Chris Gonnerman <chris.gonnerman at newcenturycomputers.net> [2002-04-13
00:15 -0500]:
> > Another reason often given (by the misinformed) is that GPL code can
> > somehow "infect" non-GPL free code (BSD let's say); it goes something
> > like this: Bob writes an excellent library and releases it under the
> > BSD license.  George has his previously mentioned GPL'd library, which
> > is also most cool.  Fred comes along, and writes a program which uses
> > both Bob's and George's libraries, and for some reason Fred actually
> > incorporates the library source in his project rather than expecting
> > you to separately install it.  Fred must release his program under the
> > GPL (he has no choice in this matter).
>
> This is wrong. Fed must release his program under a Free Software
> license (see the FSF page I mentioned in a previous posting for a
> listing of those). So Fred can release his program under a
> BSD/Python-like license, too.

If Fred *incorporates* George's GPL code in his program, he must make
the rest of his code GPL too... the license is pretty clear this way.
Alternately, if he makes a binary release the linkage rule may require
him to GPL the program.

<from the gpl>

  10. If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission.  For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this.  Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.

</from the gpl>

The point is, Bob's license can't be changes unilaterally by Fred.

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





From varial at gmx.de  Fri Apr  5 17:13:21 2002
From: varial at gmx.de (Robert Spielmann)
Date: 5 Apr 2002 22:13:21 GMT
Subject: crc32 function
Message-ID: <Xns91E821BC599Fvarialgmxde@130.133.1.4>

Hi,

I have a program where I need to calculate the crc32 values of fibonacci 
numbers, is there a way to get the crc32 of int variables or do I need to 
first write the number to a file, then read it back as a string to do the 
crc32 stuff?

Thanks for any help
Robert Spielmann


From gerhard at bigfoot.de  Tue Apr  9 12:06:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 9 Apr 2002 16:06:17 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no>
Message-ID: <slrnab64bp.m9.gerhard@lilith.my-fqdn.de>

Syver Enstad wrote in comp.lang.python:
> "Fredrik Lundh" <fredrik at pythonware.com> writes:
> 
>> Sean 'Shaleh' Perry wrote:
>> > " ".join(mylist) # is this perl or python?
>> 
>> *why* string objects have join method has been explained
>> over and over and over again, and should also be obvious for
>> anyone with some knowledge how strings and sequences work
>> in Python.
>> 
>> > whoever decided this is the new style should be shot, a lot
> 
> Am I the only one in the entire world that thinks '\t'.join(mylist) is
> elegant and natural?

Of course not. It's only that there's an unusual amount of complaining
going on on c.l.p recently. Those who are happy with the development
of the language, just use the new features and don't complain
publicly. OTOH other users complain why 2.2 is so different from 1.5.2
instead of just sticking with 1.5.2 or even 1.5.2 features if they're
happy with it. I fail to see the logic of that, if there is any.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From hgg9140 at seanet.com  Fri Apr  5 20:12:13 2002
From: hgg9140 at seanet.com (Harry George)
Date: 05 Apr 2002 17:12:13 -0800
Subject: Rule Engine in Python
References: <eb612768.0203101940.5565b282@posting.google.com> <m3it82d32l.fsf@wilma.localdomain> <I45r8.34886$0T6.10550048@e3500-atl1.usenetserver.com> <m3u1qq3ma8.fsf@wilma.localdomain> <B4nr8.36171$ku5.3121156@e3500-atl2.usenetserver.com>
Message-ID: <m3zo0hiqaa.fsf@wilma.localdomain>

"Robert Oschler" <Oschler at earthlink.net> writes:

> Harry George <hgg9140 at seanet.com> wrote in message
> news:m3u1qq3ma8.fsf at wilma.localdomain...
> > A few lines of python doing
> > regualr expressions and some text manipulation takes the palce of a
> > page or so of prolog.  That was all with ALS -- I've used SWI for a
> > few simple examples, but never compared them for heavy processing.
> >
> 
> I don't want to misinterpret anything, but are you indicating that if you
> had your druthers you'd replace the prolog code wth python code?
> 
> thx
> 

Yes, at least for our context.  Out of perhaps 100 prolog modules,
only 1 or 2 really have declarative rules which take advantage of
backtracking.  The others are procedural (usually enforced with cuts).
For that python is a better choice.  Either way, the heavy lifting for
3D graphics is in the underlying C/C++/OpenGL libraries.  Prolog/Pyton
are there to add intelligence.

> 
> 

-- 
Harry George
hgg9140 at seanet.com


From hgg9140 at seanet.com  Mon Apr 22 20:26:43 2002
From: hgg9140 at seanet.com (Harry George)
Date: 22 Apr 2002 17:26:43 -0700
Subject: Lisp (sexpr) parsing?
Message-ID: <m3n0vvtg3g.fsf@wilma.localdomain>

I've used the old sexpr v0.2 from starship for years, and recently saw
an sexpr v0.2 from sourceforge.  The old one was incomplete and I've
yet to build the new one.  I've also noticed an emacs library ("elib"
I think).

What do folks recommend for parsing lisp code?  Just parsing data
structures would be enough, but need to handle those oddities like
maybe-a-number(3.14159XYZ), single-quote quoted lists, vectors, etc.

-- 
Harry George
hgg9140 at seanet.com


From shalehperry at attbi.com  Fri Apr 26 19:53:44 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 26 Apr 2002 16:53:44 -0700 (PDT)
Subject: Removing dupes from a list
In-Reply-To: <20020426212006.GY21205@teich.Garten.DigitalProjects.com>
Message-ID: <XFMail.20020426165344.shalehperry@attbi.com>

On 26-Apr-2002 Alexander Skwar wrote:
> ?Sean 'Shaleh' Perry? sagte am 2002-04-26 um 13:56:30 -0700 :
>> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560
> 
> Thanks, but connections to activestate.com simply time out :(  However,
> a traceroute to aspn.activestate.com doesn't show any unusual long hops.
> But AS is reachable from other networks.  Strange.
> 
> Anyhow, seems like uniquer() from the 1st comment is what I'm looking
> for, because unique() also looses order just like my two "functions".
> 

look in the comments, Alex Martelli discusses keeping the order.




From duncan at NOSPAMrcp.co.uk  Tue Apr 23 09:01:04 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 23 Apr 2002 13:01:04 +0000 (UTC)
Subject: flatten tuple
References: <m2adrvnxiv.fsf@mother.paradise.lost>
Message-ID: <Xns91F98E705186Dduncanrcpcouk@127.0.0.1>

John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote in
news:m2adrvnxiv.fsf at mother.paradise.lost: 

> The only hitch is that I want to be able to do this recursively, so
> I could have for example
> 
> m = {
>     'where': ('was', 'somekey'),
>     'somekey': 'someval',
>     'someval': ('asking', 'really', 'silly', 'questions')
>     }
> 
> and end up with 
> 
> 
> y = ( 'John', 'Hunter',  'was',  'asking', 'really', 'silly',
> 'questions') 
> 

Without checkig for infinite recursion, here is one way to do it:
>>> m = {
    'where': ('was', 'somekey'),
    'somekey': 'someval',
    'someval': ('asking', 'really', 'silly', 'questions')
    }
>>> x = ( 'John', 'Hunter', 'where' )
>>> def expand(aList, aDict):
	if isinstance(aList, tuple):
		aList = list(aList)
	result = []
	while aList:
		head = aList.pop(0)
		if aDict.has_key(head):
			head = [aDict[head]]
		if isinstance(head, tuple) or isinstance(head, list):
			aList[:0] = list(head)
		else:
			result.append(head)
	return tuple(result)

>>> print expand(x, m)
('John', 'Hunter', 'was', 'asking', 'really', 'silly', 'questions')
>>> 

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From gieSPAm at free.fr  Thu Apr  4 05:29:27 2002
From: gieSPAm at free.fr (jm)
Date: 04 Apr 2002 10:29:27 GMT
Subject: ANN: SLiP and SLIDE
References: <mailman.1017879911.8562.python-list@python.org>
Message-ID: <3cac2b07$0$21495$626a54ce@news.free.fr>

Scott Sweeney <ssweeney at idiosync.net> writes:
> Announcing...
> 
> SLiP 0.1: a "Sorta Like Python" shorthand syntax for XML
> SLIDE 0.1: a "Sorta Like an IDE" lightweight windows exe for SLiP editing
> SLiPpy 0.1: a SLiP package for Python, including a SLiP->XML library
> (sliplib.py) and source for SLIDE (slide.py)
> 
> 	http://www.scottsweeney.com/projects/slip

How different is Slip from SOX ?

http://www.langdale.com.au/SOX/

--
 jm


From nhodgson at bigpond.net.au  Thu Apr 25 20:08:27 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 26 Apr 2002 00:08:27 GMT
Subject: CallDLL Package
References: <d9773b18.0204251218.3e96bf7@posting.google.com>
Message-ID: <%T0y8.50035$o66.156124@news-server.bigpond.net.au>

Rajat Chopra:

> I am using MinGW32 (particularly GCC and the BINUTILS package) to
> rebuild the Python library file. When I run the script to rebuild the
> library file, however,  I continuously get the error of "undefined
> reference to LoadLibraryA, FreeLibrary, GetModuleHandleA, and
> GetProcAddress functions".

   ...

   Try the -mwindows option to dllwrap. One peculiarity of the GNU linker is
that the order of libs is significant so sometimes to satisfy all
(particularly circular) dependencies you get to write things like:
dllwrap x.o -lkernel32 -lmine -lkernel32

   Neil




From JamesL at Lugoj.com  Sun Apr 14 02:26:16 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 14 Apr 2002 06:26:16 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org>
Message-ID: <Xns91EFEE502A532JamesLLugojcom@216.148.53.85>

Tim Peters <tim.one at comcast.net> wrote:
> A crushing amount of debate over "stability" has gone by on Python-Dev
> the last week, and the one thing that strikes me over and over is that,
> with just a few minor exceptions, nobody volunteers to *do* anything
> except tell other people what to do.  What are you willing to do to
> make your desires a reality?  If it doesn't involve contributing work,
> time or money, you just want a free ride.  But nobody owes anyone a
> free ride, and open source can't change that fact of life either.

Not sure, but I think you are close to describing what has been called the 
tragedy of the commons.

Back when I was paid to do Python work, I contributed all of US$50 (I 
think; maybe it was $20) to the PSA (I think it was called) and tried to 
get the company I was consulting for to kick in a couple thousand (or a 
couple hundred at least). They never got around to that, alas.

Unfortunately I'm not sure what would be gained by contributions to a 
strategy that I have specifically stated I don't agree with. Further, it is 
unclear how I can "contribute" to the slower pacing of releases that I'd 
like to see.

I've overstayed my welcome here, but before I go I'd like to thank you for 
your efforts to advance Python and I wish you well in your endeavors.


From elian at inbrief.net  Mon Apr 22 11:45:55 2002
From: elian at inbrief.net (Edmund Lian)
Date: Mon, 22 Apr 2002 11:45:55 -0400
Subject: Overriding methods in classes that use property()
Message-ID: <OFDC6A67B7.8A3300B0-ON85256BA3.00560517@juscom.com>


On Mon, 22 Apr 2002, Denis Otkidach wrote:


>EL> class C(A):
>EL>     def setVal(self, value):
>EL>         self._val = 'Huh?'
>
>Simply add one line here:
>        val = property(getVal, setVal)

Hmm... I tried this, and I get a NameError because getVal is not defined.
However, things work if I redefine class C to:

class C(A):
    def setVal(self, value):
        self._val = 'Huh?'

    def getVal(self):
        return super(C, self).getVal()

    val = property(getVal, setVal)

However, this is messy since one would have to do this for every property()
attribute in the base class that is not being overridden in a subclass!
This is *not* intuitive, *not* expected, and doesn't seem Pythonic! Seems
more like a bug to me!!!

...Edmund.





From marklists at mceahern.com  Wed Apr  3 16:35:09 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 15:35:09 -0600
Subject: File opening
In-Reply-To: <a8frho$qqu$1@netnews.upenn.edu>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEBBKJAA.marklists@mceahern.com>

[Salim Zayat]
> But if you wanna import the module, then the path to the open statement 
> needs to be full path to the file.  Is there a fix to this?

Put it in a package and use __path__.

Consider:

/site-packages/
  /a/
	__init__.py

# __init__.py
pass

This means you can do this:

import a
print a.__path__[0]

// mark




From ods at fep.ru  Mon Apr 22 11:41:44 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Mon, 22 Apr 2002 19:41:44 +0400 (MSD)
Subject: Overriding methods in classes that use property()
In-Reply-To: <Pine.LNX.4.44L.0204221923330.2654-100000@ods.pp.ru>
Message-ID: <Pine.LNX.4.44L.0204221940340.2654-100000@ods.pp.ru>

On Mon, 22 Apr 2002, Denis S. Otkidach wrote:

DSO> EL> class A(object):
DSO> EL>     def __init__(self):
DSO> EL>         self._val = None
DSO> EL>
DSO> EL>     def setVal(self, value):
DSO> EL>         self._val = value
DSO> EL>
DSO> EL>     def getVal (self):
DSO> EL>         return self._val
DSO> EL>
DSO> EL>     val = property(getVal, setVal)
DSO> EL>
DSO> EL> class B(A):
DSO> EL>     pass
DSO> EL>
DSO> EL> class C(A):
DSO> EL>     def setVal(self, value):
DSO> EL>         self._val = 'Huh?'
DSO>
DSO> Simply add one line here:
DSO>         val = property(getVal, setVal)

Sorry, getVal is not in current scope. The following should be
right:
             val = property(A.getVal, setVal)





From robin at jessikat.fsnet.co.uk  Thu Apr 11 03:49:26 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 11 Apr 2002 08:49:26 +0100
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx>
 <mailman.1018391839.6535.python-list@python.org>
 <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
 <mailman.1018481805.1192.python-list@python.org>
Message-ID: <SARlHIAGAUt8Ewit@jessikat.fsnet.co.uk>

In article <mailman.1018481805.1192.python-list at python.org>, Cliff Wells
<logiplexsoftware at earthlink.net> writes
>On Thu, 11 Apr 2002 00:17:40 +0100
>Robin Becker wrote:
>> Am I the only person that finds the whole concept of Easter eggs a slap
>> in the face from the developers. 
>
>Maybe not, but probably the only one who got up on the wrong side of the bed
>today.
>
>
well I think I wrote that before going to bed and perhaps the beer
wasn't good or something.
-- 
Robin Becker


From phr-n2002a at nightsong.com  Thu Apr  4 09:45:01 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 06:45:01 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xbsd0qpmn.fsf@ruckus.brouhaha.com> <3CABABCA.EB446FEE@python.org> <7x7kno5isl.fsf@ruckus.brouhaha.com> <3CAC6205.39B5E01B@python.org>
Message-ID: <7x4rirjzf6.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
> The array object is an old-style type object.  That's not the same as an
> old-style class (it's totally unrelated).  But new-style type objects are
> the same as new-style classes (that's what new-style classes are all
> about) and it's easy enough to convert an old-style type to a new-style
> type.  Then the new-style type will allow subtyping, so you can write a
> subtype that implements bit elements.

Thanks.  I still don't fully understand this, but I'm probably best off
reading the code before asking further questions about it here.

> > Anyway, I'll put it on my list, but don't hold your breath.
> 
> I won't, but I'll remind you that if you give, you'll get more back.

Fair enough.



From m.faassen at vet.uu.nl  Thu Apr 25 18:44:46 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Apr 2002 22:44:46 GMT
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <aaa0su$b86$1@newshost.accu.uu.nl>

Chris Spencer <clspence at one.net> wrote:
>        I tested this using the Python command interpreter on a Windows XP box
> running:
> Python 2.1 (#15, Apr 16 2001, 18:25:49) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
>>>> float(5)/float(100)
> 0.050000000000000003
>>>> 5.0/100.0
> 0.050000000000000003

>        Okay, I might not be a math genius, but this is most definately WRONG.
> What gives?  Anyone have insight into this little "feature" of Python floating
> point operations?

Heh. The timbot will be with you shortly. If you're impatient, do a 
google groups search and find things like this:

http://groups.google.com/groups?hl=en&selm=000501be0ac4%2433c3f2a0%2418a22299%40tim&rnum=4

http://groups.google.com/groups?q=IEEE+group:comp.lang.python+author:Tim+author:Peters&start=10&hl=en&scoring=d&selm=mailman.991512844.1883.python-list%40python.org&rnum=11

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From starx at pacbell.net  Wed Apr 10 02:32:00 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 10 Apr 2002 06:32:00 GMT
Subject: Problem with packages and modules
References: <117880a1.0204090542.258f0f51@posting.google.com> <3CB2FA14.1050708@mxm.dk> <117880a1.0204092207.7e51ff80@posting.google.com>
Message-ID: <Xns91EBEF972B8BFRASXnewsDFE1@130.133.1.4>

J?rgen Hansen || Tue 09 Apr 2002 11:07:37p:

> Max M <maxm at mxm.dk> wrote in message news:<3CB2FA14.1050708 at mxm.dk>...
>> J?rgen Hansen wrote:
>> 
>> > This is an example of directory hierachy:
> <snip example hierachy>
[ Snip some other stuff ]
>> > Further if I typed in:
>> >    >>> import company
>> > 
>> > I would assume that I could access all subpackages and their
>> > modules with a dotted name (company.plot.figure.Figure() for
>> > instance), but this is obviously not the case.
>> >
>> 
>> You would need to import the modul in company/__init__.py:
>> import plot
>> 
>> to do:
>> 
>> import company
>> fig = company.plot.figure.Figure()
>> 
>> 
>> Hmm... I hope this helps a bit.
> 
> It does. Thanks a lot. After I wrote, I got thinking, that my approach
> would have meant that I would import the entire hierachy. Not very
> smart if the hierachy, where larger.
> 
> Well thanks again, for the prompt answer.
> 
> regards
> Jorgen
> 

No one's mentioned this particular permutation; if you want to use 
'plot' as a logical module in your code, you can just do 'from company 
import plot' =).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From martin at v.loewis.de  Fri Apr 12 16:20:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 12 Apr 2002 22:20:32 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018632996.10456.python-list@python.org>
Message-ID: <m3d6x4fz3j.fsf@mira.informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/Modules/mathmodule.so: symbol PyExc_OverflowError: referenced symbol not found
> 
> Could someone suggest how to proceed, please?

Can you please report the command line option that was used to link
the python binary? (delete it, and re-invoke make) PyExc_OverflowError
ought to be exported from the Python binary, and there ought to be a
gcc command line option to achieve that.

If you know what nm(1) is, please verify that the symbol is exported.

Regards,
Martin



From db3l at fitlinxx.com  Wed Apr 17 23:27:57 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 17 Apr 2002 23:27:57 -0400
Subject: listobject.c:101: bad argument to internal function
References: <a9l0o9$hdg$1@web1.cup.hp.com>
Message-ID: <uelhdadoi.fsf@ctwd0143.fitlinxx.com>

"Jim Henry" <jamespp35 at yahoo.com> writes:

> I have gone over the C code, and it is VERY simple.  (...)
(...)
> Any ideas? (please!)

Post the code.  It's hard to say without seeing something concrete.
Are you consistently checking for errors on all of your Python library
calls?

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From timr at probo.com  Sat Apr  6 23:53:09 2002
From: timr at probo.com (Tim Roberts)
Date: Sat, 06 Apr 2002 20:53:09 -0800
Subject: [Q] Python CGI Q
References: <8Far8.36437$tg4.432342@vixen.cso.uiuc.edu>
Message-ID: <o3kvauc4u8i6cp803df18s5924qq0s7mlm@4ax.com>

"John" <nobody at nobody.com> wrote:
>
>Hi, I'm wondering if I can write a CGI that loads/edits/saves files on
>server side.. (I don't need to worry about security concerns for this
>application) Any starting point or open-source?

EVERYTHING in a CGI script happens on the server side.  You can create a
<TEXTAREA> and copy the contents of your file inside it, then have a second
script wait for the response and write the new contents out.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From ark at research.att.com  Wed Apr 10 15:34:12 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 19:34:12 GMT
Subject: Binding a reference to a variable (solved?!?)
References: <E16v1a2-0005a6-00@mail.python.org> <mailman.1018463002.9069.python-list@python.org>
Message-ID: <yu99bscrz6tn.fsf@europa.research.att.com>

Patrick> I'll likely submit the following as a ASPN recipe, though it
Patrick> clearly needs a better description of motivation.  Just
Patrick> because you CAN do a thing doesn't mean you SHOULD do a
Patrick> thing.

I'll say.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From sholden at holdenweb.com  Tue Apr 30 14:56:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 14:56:37 -0400
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org>  <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no>
Message-ID: <aQBz8.2627$Au5.1454@atlpnn01.usenetserver.com>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnactp07.pv.mlh at vier.idi.ntnu.no...
> In article <fjbz8.7570$M63.5673 at atlpnn01.usenetserver.com>, Steve Holden
wrote:
> [snip]
> >> > Why use type({}) when you can use dict?
> >>
> >[ ... good reasons ...]
> >>
> >> 3. And 'dict' is something i often use in local definitions
> >>    when working with dictionaries.
> >>
> >Yeah, well we all have to cut that out and start using "adict" or
"thisdict"
> >pretty sharpish. This one bites now.
>
> How so? It only "bites" if you actually need access to the dict type
> in the same scope where you bind your variable... I've got plenty of
> code with stuff like
>
>   file = open('foo.bar')
>
> and it doesn't bother me one bit (except in the cases where 'file'
> should have been something more descriptive, I guess ;)
>
> Thes names aren't reserved in any way...
>

I understand that. They just scream out to be overridden in the
module-global namespace, making the builtin version invisible for the
duration of that (dynamic) scope. The problem is when you've overridden
them, and then want to use the built in. I understand exsisting code won't
break: I'm just suggesting that modifications to such code might have
unforeseen consequences.

not-that-you-or-i-would-ever-do-that-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From shredwheat at attbi.com  Sun Apr 14 20:17:00 2002
From: shredwheat at attbi.com (Pete Shinners)
Date: Mon, 15 Apr 2002 00:17:00 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <3CBA1C89.4030407@attbi.com>

Raymond Hettinger wrote:
> The help and tutor maillists indicate that the use of input() rather than
> raw_input() is widespread.  I think this function is so harmful and so
> likely to be used that it should be deprecated hasta pronto (right away).
> When really needed, eval(raw_input()) is easy enough to code and explicit
> about what it does.
> 
> <no wink -- this function needs to go away>

i've always wondered why input's default behavior was like this. a 
better function name would be "execute_arbitrary_code". if the user 
wants a certain type of data, 10 out of 10 times they can just convert 
it to the desired type...

input = raw_input #BLARGH
float_val = float(input('Enter A Float'))
int_val = int(input('Enter an Int'))




From tinu at email.ch  Wed Apr 17 06:28:38 2002
From: tinu at email.ch (tinu)
Date: 17 Apr 2002 12:28:38 +0200
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu> <861ydg3ff3.fsf@haggis.aae.uiuc.edu> <3CBC5686.BCBCF663@noaa.gov> <a9holn$5t1$2@peabody.colorado.edu>
Message-ID: <m3elheippl.fsf@yak.ethz.ch>

Fernando P?rez <fperez528 at yahoo.com> writes:

> This is another one for which weave can be a godsend. The blitz module in 
> weave can for many numeric expressions generate equivalent C++ code which 
> avoids all the intermediate copies. I don't want to sound like a cheerleader 
> (I'm not a developer of weave, only a  user), but weave is really amazing in 
> how easily it allows you to integrate Numeric pyhton code with C snippets.
> 
> If anyone needs, I have a bunch of little examples lying around which I wrote 
> when I was learning weave and which I can pass along, just drop me a  line.

Yes, please, these examples would be very helpful. Maybe you post an URL to this
list or add the examples to the scipy CVS?

Thanks a lot

Martin

-- 
Martin L?thi                     VAW Glaciology, ETH Z?rich, Switzerland
                                 mel: luthi at vaw.baug.ethz.ch


From peter at engcorp.com  Wed Apr 10 23:50:21 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 23:50:21 -0400
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3CB4BB04.1B8DD2A6@bgb-consulting.com>  <3CB4E083.7EC2EDC@engcorp.com> <slrnaba0ig.efc.mlh@vier.idi.ntnu.no>
Message-ID: <3CB507FD.E630AC98@engcorp.com>

Magnus Lie Hetland wrote:
> 
> In article <3CB4E083.7EC2EDC at engcorp.com>, Peter Hansen wrote:
> >Don Garrett wrote:
> >> "M. J. Mix" wrote:
> >> > Is anyone using Python as a spacecraft ground system scripting
> >> > language, or onboard (embedded) language?
> >> >
> >>   I'm not aware of ANY scripting languages that have been developed in a manor
> >> suitable for life critical fail-safe situations. I like Python a lot, and find
> >> it very reliable for day to day stuff, but I'd hate to have it embedded in my
> >> pace maker.
> >
> >Agree.  Unsuitable for pacemakers, or any other life-critical
> >application, at this time.  Not inherently unsuitable mind you.
> 
> There was a similar discussion a while ago about Air Traffic Control.
> Lots of important points came up. Perhaps take a look at it before
> repeating all of them in this thread?
> 
> Just a suggestion :)

I did.  Have it bookmarked in fact.  I also participated at the time.

I also thought I had contributed something new this time around, phrased
somewhat more succinctly than I had before.

Maybe I erred in that judgement. :-(

-Peter


From sholden at holdenweb.com  Mon Apr  1 07:25:28 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 07:25:28 -0500
Subject: List problem
References: <a82mlc$ja1$1@slb6.atl.mindspring.net> <xU6p8.28161$VJ1.2411615@bin3.nnrp.aus1.giganews.com> <a84eu0$g8t$1@slb3.atl.mindspring.net> <a87970$i9q$1@slb2.atl.mindspring.net>
Message-ID: <VlYp8.107343$Ou.55255@atlpnn01.usenetserver.com>

"Jon Cosby" <jcosby at mindspring.com> wrote in message
news:a87970$i9q$1 at slb2.atl.mindspring.net...
> Is a line from a file an ordinary  string?
>
Yes, it's indistinguishable from a string created any other way.

regards
 Steve






From bokr at oz.net  Tue Apr 16 19:04:48 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 23:04:48 GMT
Subject: Brackets in brackets, braces in braces
References: <2e1a250.0204160730.637139f1@posting.google.com>
Message-ID: <a9iamg$93h$0@216.39.172.122>

On 16 Apr 2002 08:30:05 -0700, ahorvath at szgti.bmf.hu (Horvath, Arpad) wrote:

>How can I find strings like this with regular expressions:\something
>[something1] {something2} {something3}Where in the braces can be also
>another braces. For example:\frac{1\times10^{-16}}{4}   can be at the
>place, where something2 is.(braces in pairs)\interval [J]
>{\frac{1\times10^{-16}}{4}} {1..5}My half solution is at
>www.szgti.bmf.hu/~ahorvath/regexp.txt.

Depending on what your real goal is, you might want to play with
using split and making an actual nested list structure.
E.g., I posted this for a similar question in the past (it just does one
kind of brackets, i.e, braces):

> >>> s = "A line of text with {multiple and {nested} braces} to {confuse} re!"
> >>> sl = eval('["'+'"], "'.join('", ["'.join(s.split('{')).split('}'))+'"]')
> >>> sl
> ['A line of text with ', ['multiple and ', ['nested'], ' braces'], ' to ', ['confuse'], ' re!']

>then walk the tree as desired.

The above uses eval to do the nest counting. You'd have to split on '[' and ']' first, and then
walk the tree and do '{' and '}' processing similarly to expand the first tree's elements, if
you had both kinds of bracketing. Plus you'd want to preserve some kind of indication of what
kind of bracket your final nested lists represented.

Or you could write a recursive parser. Might be fun ;-)
Or you could use a Python flex workalike to do it, I suppose. I haven't tried one yet.

Regards,
Bengt Richter


From noah at noah.org  Wed Apr 24 10:14:59 2002
From: noah at noah.org (Noah Spurrier)
Date: Wed, 24 Apr 2002 07:14:59 -0700
Subject: variable substitution
References: <e2e9e807.0204240519.688cf3cb@posting.google.com>
Message-ID: <103d01c1eb9a$6b424280$0200a8c0@saysyou>

It looks like you are trying to do Perl :-P
You're close, but you need to use a dictionary.
You still must follow the string with a dictonary that
contains the values to be substituted. So it takes the form
    'My string substitution %(my_key_name)s'  %  my_dictionary
The substitutions in the form  %(name)s  are not variable names.
They are key names into the dictionary.

In your case something like this would work:

    foo = {'sessionId':123456789}
    print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
Type</a>' % foo

What may have confused you in the examples in the URL you gave
is that the string formats were defined and assigned to variables before
they were used.
In your case it might look like this:

    foo = {'sessionId':123456789}
    format = '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
Type</a>'

    # Stuff happens... Many lines of code later...
    print  format  %  foo

See the end of this section 7.1:

http://www.python.org/doc/current/tut/node9.html#SECTION00910000000000000000
0

Yours,
Noah

----- Original Message -----
From: "Hugh" <h.e.w.frater at cs.cf.ac.uk>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, April 24, 2002 6:19 AM
Subject: variable substitution


> How do I get the %(variable)s thing to work. I need to substitute a
> variable into a print statement so I can send links with sessionId's
> attached back to the browser. I'm trying this at the moment, and It
> says invalid sysntax.
> print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
> Type</a>'
> I saw this in a document written by Guido that I found here:
> http://www.w3j.com/6/s3.vanrossum.html
>
>
> hugh
>





From mwh at python.net  Wed Apr 17 09:41:15 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 13:41:15 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <lkbscisakh.fsf@pc150.maths.bris.ac.uk>

dominikush at yahoo.com writes:

> one thing I like very much about Python is that statements
> work like you would expect them to work.

Well, Python works very much as I expect it, but it's not clear if
this says more about me or Python <wink>.

At the end of your email, you say:

> Who is wrong here: my intuition or Python (2.2)? If it's
> my intuition, how can I train my thinking about Python's
> execution model, so that my intuition get's better ;-)

It's you :) As I can't read my email at the moment[1], I have no
better way of wasting my time to hand than drawing you some ascii art.

First, some terminology.  Actually, the very first thing is some
anti-terminology; I find the word "variable" to be particularly
uphelpful in a Python context.  I prefer "names", "bindings" and
"objects".

Names look like this:

    ,-----.
    | foo |
    `-----'

Names live in namespaces, but that's not really important for the
matter at hand as the only namespace in play is the one associated
with the read-eval-print loop of the interpreter.  In fact names are
only minor players in the current drama; bindings and objects are the
real stars.

Bindings look like this:

    ------------>

Bindings' left ends can be attached to names, or other "places" such
as attributes of objects and entries in lists or dictionaries.  Their
right hand ends are always attached to objects[2].

Objects look like this:

    +-------+
    | "bar" |
    +-------+

This is meant to be the string "bar".  Other types of object will be
drawn differently, but I hope you'll work out what I'm up to.

> Take for example the use of dict.values() for dictionaries: If you
> store the result of dict.values(), and change the dictionary after-
> wards, the previously stored result remains untouched.
> 
> >>> dict = {'a':1,'b':2}

After this statement, it would seem appropriate to draw this picture:

    ,------.       +-------+
    | dict |------>|+-----+|     +---+
    `------'       || "a" |+---->| 1 |
                   |+-----+|     +---+
                   |+-----+|     +---+
                   || "b" |+---->| 2 |
                   |+-----+|     +---+
                   +-------+

> >>> list = dict.values()

Now this:

    ,------.       +-------+
    | dict |------>|+-----+|             +---+
    `------'       || "a" |+------------>| 1 |
                   |+-----+|             +---+
                   |+-----+|              /\
                   || "b" |+-----.    ,---'
                   |+-----+|     |    |
                   +-------+     `----+----.
                                      |    |
    ,------.       +-----+            |    \/
    | list |------>| [0]-+------------'   +---+
    `------'       | [1]-+--------------->| 2 |
                   +-----+                +---+

> >>> list
> [1, 2]

Which is of course, no surprise.

> >>> dict['a']=3

Now this:


    ,------.       +-------+
    | dict |------>|+-----+|             +---+
    `------'       || "a" |+-.           | 1 |
                   |+-----+| |           +---+
                   |+-----+| |            /\
                   || "b" |+-+---.    ,---'
                   |+-----+| |   |    |
                   +-------+ |   `----+----.
                             |        |    |
    ,------.       +-----+   |        |    \/
    | list |------>| [0]-+---+--------'   +---+
    `------'       | [1]-+---+----------->| 2 |
                   +-----+   |            +---+
                             |            +---+
                             `----------->| 3 |
                                          +---+


> >>> list
> [1, 2]
> >>> dict
> {'a': 3, 'b': 2}

These should also come as no surprise; just chase the arrows
(bindings) above.

> However, if a dictionary has lists as value entries, I get
> a counterintuitive behavior (which, just recently, broke my
> code): If you change the dict, the list you previously
> created via dict.values() gets automagically updated. A nice
> feature, but nothing I would have expected!

That's because you're not thinking in terms of Names, Objects and
Bindings.

> >>> dict = {'a':[1],'b':[2]}

    ,------.       +-------+
    | dict |------>|+-----+|     +-----+   +---+
    `------'       || "a" |+---->| [0]-+-->| 1 |
                   |+-----+|     +-----+   +---+
                   |+-----+|     +-----+   +---+
                   || "b" |+---->| [0]-+-->| 2 |
                   |+-----+|     +-----+   +---+
                   +-------+

> >>> list = dict.values()

    ,------.       +-------+
    | dict |------>|+-----+|             +-----+   +---+
    `------'       || "a" |+------------>| [0]-+-->| 1 |
                   |+-----+|             +-----+   +---+
                   |+-----+|               /\
                   || "b" |+-----.    ,----'
                   |+-----+|     |    |
                   +-------+     `----+-----.
                                      |     |
    ,------.       +-----+            |     \/
    | list |------>| [0]-+------------'   +-----+   +---+
    `------'       | [1]-+--------------->| [0]-+-->| 2 |
                   +-----+                +-----+   +---+

> >>> list
> [[1], [2]]

Again, no surprises here.

> >>> dict['a'].append(3)

                                                    +---+
    ,------.       +-------+                     ,->| 1 |
    | dict |------>|+-----+|             +-----+ |  +---+
    `------'       || "a" |+------------>| [0]-+-'
                   |+-----+|             | [1]-+-.
                   |+-----+|             +-----+ |  +---+
                   || "b" |+-----.         /\    `->| 3 |
                   |+-----+|     |    ,----'        +---+
                   +-------+     |    |
                                 `----+-----.
    ,------.       +-----+            |     \/
    | list |------>| [0]-+------------'   +-----+   +---+
    `------'       | [1]-+--------------->| [0]-+-->| 2 |
                   +-----+                +-----+   +---+

> >>> dict
> {'a': [1, 3], 'b': [2]}
> >>> list
> [[1, 3], [2]]

And now these should not be surprising either.

> Looks like that in the first case a copy is returned while
> in the latter case list references are returned. Ok, but
> according to Python's philosophy I shouldn't mind if I work
> with lists in the dictionary or anything else. If the
> behavior depends on the knowledge of the type of values I
> put into a dictionary, I find that somehow counterintuitive.

If you haven't realised where you're misconceptions come from from the
above pictures, I'm not sure more prose would help.

Cheers,
M.
[1] Does anyone know where the starship's gone?
[2] Anyone mentioning UnboundLocalError at this point will be shot.

-- 
  A.D. 1517: Martin Luther nails his 95 Theses to the church door and
             is promptly moderated down to (-1, Flamebait).
        -- http://slashdot.org/comments.pl?sid=01/02/09/1815221&cid=52
                                        (although I've seen it before)


From cliechti at gmx.net  Fri Apr 12 14:56:41 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 12 Apr 2002 20:56:41 +0200
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu> <Xns91EDE505F5BD7cliechtigmxnet@62.2.16.82> <a959s9$s08$1@peabody.colorado.edu>
Message-ID: <Xns91EED544B2F00cliechtigmxnet@62.2.16.82>

Fernando P?rez <fperez528 at yahoo.com> wrote in
news:a959s9$s08$1 at peabody.colorado.edu: 

> Chris Liechti wrote:
> 
>> Rajarshi Guha <rxg218 at psu.edu> wrote in
>> news:a94qk5$22mg at r02n01.cac.psu.edu:
>> 
>>> On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de
>>> Maria wrote:
>>> 
>>>> Is there a way to save in a file the status, objects, functions of
>>>> interpreter
>>>> in order to restore it after the interpreter has been closed?
>>>> 
>>>> If it is possible, one can work on a project stop and then continue
>>>> later adding functions, objects and so on.
>> 
>> thats a question for an IDE with such functionality. i don't know one
>> but i think it should be easy to store all objects in the global
>> namespace of the interpreter and load it again on next startup.
>> the textual history of the interpreter with all the outputs must be
>> saved separately.
>> 
> 
> well, I've tried (not terribly hard) and failed. My ipython project is
> a fancy interpreter (http://www-hep.colorado.edu/~fperez/ipython/) and
> I wanted exactly that kind of functionality. It turns out pickle cant
> quite pickle _everything_ you throw at it, so there's bound to be
> things in your namespaces that you can't pickle.

i think PYRO.sf.net has solved some of the issues. Irmen says that he 
transmits missing code objects/class definitions to the remote machine
if it misses them. but i don't know if it works only with pyc files or with 
any class including those that were defined interactively. (the marshal 
module helps transfering pyc files (but code objects - well it seems so 
when i look at the docs?))
 
> I'd like to know if there's a way to make this work, because as I said
> my first attempts didn't get me anywhere. My kludgy semi-solution was
> to implement session logging/restore by replaying all input, but
> that's slow, error prone and has many other limitations.

in editors like PythonWin we could just replace the text in the interactive
console and unpickle/unmarshal the namespace.
maybe some special treatment is needed so that only code object are 
marshalled and instances are pickled.

i have not yet used pickle or marshal for real apps but its worth a try :-)
 
chris

-- 
Chris <cliechti at gmx.net>



From borcis at geneva-link.ch  Tue Apr  9 09:01:11 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 09 Apr 2002 15:01:11 +0200
Subject: True/False story... (PEP 285)
References: <a8u99o$2lcu$1@news.nextra.cz> <vHys8.62333$pT1.1792330@news1.tin.it>
Message-ID: <3CB2E617.8050600@geneva-link.ch>

Alex Martelli wrote:


> there IS a natural mapping of booleans onto a subset
> of integers, just as there is a natural mapping of
> integers onto a subset of reals (or rationals, etc).


Could be argued that there is no single natural mapping, for an 
alternate to the current one we could have

True     == 1
False    == -1
not x    == -x
x and y  == min(x,y)
x or y   == max(x,y)
x xor y  == -x*y

A nice feature of that one (to my aesthetics), is that

1j == sqrt(False)

It also meshes well with the cyclotomic fields.

Boris
--
"Hope achieves the square root of the impossible"



From tejarex at yahoo.com  Wed Apr  3 09:29:49 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 03 Apr 2002 14:29:49 GMT
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
Message-ID: <xlEq8.211910$2q2.18796288@bin4.nnrp.aus1.giganews.com>

"Jan Kybic" <kybic at ieee.org> wrote in message
news:lbhpu1hhwnb.fsf_-_ at xihu.inria.fr...
> On the other hand, all the calculation that you cannot express using
> high-level primitives (using matrices, vectors,...) and has to be
> performed with loops element by element, is painfully slow. Equally
> slow in Python, as in IDL, Matlab,... It is so slow that it is
> unusable, it is very difficult to develop an algorithm if it takes
> days to process any reasonably sized input. My solution is to
rewrite
> the critical part of the code in C/C++, I did this for Matlab,
> Python+Numeric, as well as IDL, but this is awkward - it forces you
to
> work simultaneously with two languages and to deal besides the
algorithm
> itself with a lot of implementation and interface details.
>
> In fact, it has become such a burden, that I am seriously
considering
> switching to another language. The only problem is that I still have
> not found anything high-level, elegant, fast, and free at the same
time.
>
> It would be so much easier if I could just compile parts of the
Python
> code into the machine code. This would enable to use Python both for
> the prototyping and for the production code.

I imagine that it should be possible to pretty easily translate a
very restricted, special purpose, subset of Python into C for
compilation.  A module of strictly numeric functions is the most
obvious example.  Ie, assume all vars except loop indexed are floats
and arrays thereof.  I do not know what has been done in this
direction, though.

I believe there is a recently announced program (weave?) for embedding
C within Python.

Terry J. Reedy





From gmuller at worldonline.nl  Tue Apr 30 15:34:18 2002
From: gmuller at worldonline.nl (GerritM)
Date: Tue, 30 Apr 2002 21:34:18 +0200
Subject: PSA/PSF, pyZine international payments
Message-ID: <aamrko$a8v$1@reader1.tiscali.nl>

Living in Europe it is still amazingly difficult to transfer money from a
local bank account to an USA account (Note that despite the introduction of
the Euro even transfers crossing borders is still difficult). It would be
nice to have some easy way to pay PSF (PSA?) and order pyZine from Europe.

For Dutch readers would the PSF be seen as charity and therefor as tax
deducable?

regards Gerrit

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




From simon at incutio.com  Fri Apr 12 05:16:26 2002
From: simon at incutio.com (Simon Willison)
Date: Fri, 12 Apr 2002 10:16:26 +0100
Subject: Problems accessing the new Google Web APIs from Python
Message-ID: <3CB6A5EA.9010804@incutio.com>

Google have just launched a new service - Google Web APIs - which allows 
developers to query the google search engine via SOAP for non-commercial 
use (you're allowed up to 1000 queries a day). To use the service you 
have to sign up for a license key (all they ask for is an e-mail address 
and password) and download the API package:

http://www.google.com/apis/

Anyway, I've been trying to get this working with Python (having never 
used SOAP before) and I'm having problems, but I'm not sure if it's 
their service going down or a problem with my code. I've grabbed the 
SOAP module from here:

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pywebsvcs/SOAPpy/

And I'm running the following code in the Python interactive prompt:

#######
import SOAP

key = 'my google license key here'
query = 'python'

s = SOAP.SOAPProxy("http://api.google.com/search/beta2")
r = s.doGoogleSearch(key, query, 0, 10, "false", "", "false", "", 
"latin1", "latin1")

Which results in the following error:

Traceback (most recent call last):
  File "<pyshell#19>", line 1, in ?
    r = s.doGoogleSearch(key, query, 0, 10, "false", "", "false", "", 
"latin1", "latin1")
  File "C:\Python22\lib\SOAP.py", line 3607, in __call__
    return self.__r_call(*args, **kw)
  File "C:\Python22\lib\SOAP.py", line 3627, in __r_call
    self.__hd, self.__ma)
  File "C:\Python22\lib\SOAP.py", line 3536, in __call
    config = self.config)
  File "C:\Python22\lib\SOAP.py", line 3471, in call
    raise HTTPError(code, msg)
HTTPError: <HTTPError 500 Internal Server Error>
#######

Does this mean their web server is dead (from the load of all the 
enthusiastic developers trying out the new service) or am I just getting 
my SOAP usage wrong?

The example for using Perl that comes with the Web API package suggests 
getting the Perl SOAP::Lite package to load in the WSDL description file 
for the Web API before doing anything. As far as I could tell the Python 
SOAP module doesn't include support for WSDL (I may be wrong). Is this a 
likely cause of my problem?

Cheers,

Simon Willison





From loewis at informatik.hu-berlin.de  Mon Apr 29 03:01:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 29 Apr 2002 09:01:03 +0200
Subject: GC and resource finalization
References: <mailman.1020038390.20821.python-list@python.org>
Message-ID: <j4it6b9efk.fsf@informatik.hu-berlin.de>

"Ian Kjos" <ikjos at email.uophx.edu> writes:

> In the slightly more general case, when (if ever) can the
> interpretter prove that it is safe to close that file? Will the file
> be closed?

In your example, the file will be closed when the cyclic garbage
collection is invoked. This happens at an unspecified time, but can be
triggered with gc.collect().

HTH,
Martin


From phr-n2002a at nightsong.com  Fri Apr  5 18:37:17 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 15:37:17 -0800
Subject: The Topy project
References: <mailman.1018039992.15896.python-list@python.org> <Xns91E7EF76BA86Acliechtigmxnet@62.2.16.82> <mailman.1018046786.6596.python-list@python.org>
Message-ID: <7xit75wwcy.fsf@ruckus.brouhaha.com>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:
> The current meta-representation is surprisingly simple, at least so far.
> It seems sufficient to convey most of Scheme (I have nothing to represent
> lazy evaluation, nor continuations).  Some reasonable trickery is needed to
> represent nested scopes and `let' loops: I found nothing intractable yet,
> and no real problem will pop up.  Plans are still opened about `letrec';
> hopefully, I'll find something not too ugly. :-)

What do you do about mutable strings?


From quinn at vomit.ugcs.caltech.edu  Mon Apr  1 03:17:34 2002
From: quinn at vomit.ugcs.caltech.edu (Quinn Dunkan)
Date: 1 Apr 2002 08:17:34 GMT
Subject: list comprehension question
References: <mailman.1017132281.21793.python-list@python.org>
Message-ID: <slrnaag5su.qe.quinn@vomit.ugcs.caltech.edu>

On Tue, 26 Mar 2002 03:43:15 -0500, Tim Peters <tim.one at comcast.net> wrote:
>    A bot may injure a human being, or, preferably, through inaction,
>    allow a human being to come to harm, although laughing about
>    either in the hearing of humans is MACNAM-017B3^H.

All right then Dr. Memory, if you've got a brain the size of a planet, why
does the porridge bird lay his egg in the air?


From dalke at dalkescientific.com  Wed Apr  3 01:29:48 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 2 Apr 2002 23:29:48 -0700
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net> <mailman.1017810728.14199.python-list@python.org>
Message-ID: <a8e7gv$fut$1@nntp9.atl.mindspring.net>

David Mertz:
>From the other responses, it seems like a lot of users don't know the
>"Python static trick."  I use this moderately often... it is a slightly
>strange spelling, but not hard once you learn it:
>
>    >>> def counter(msg, cnt=[0]):
>    ...   print cnt[0], msg

Well, I do know about this trick, but I don't like using it except
in limited circumstances.  The problem is that people might see
that as part of the external API and think it indicates an optional
parameter.  In your case, you show that it can be used to change
the counter value, but then it's strange to take list instead of
a simple number.

Sometimes I will use it for an extra tiny bit of performance, since
it's a fast lookup.

A couple of times I've done

def _counter(msg, cnt = [0]):
  ...
def counter(msg, cnt = None):
  if cnt is None:
    return _counter(msg)
  return _counter(msg, [cnt])

but for this case it isn't any better than having another module
variable.

What I didn't think about was Gerhard's "counter.cnt" mechanism.
(My mind's still stuck in the 1.4 days :)  I really like that, and
think I'll use it more often in the future.

                    Andrew
                    dalke at dalkescientific.com





From Chris.Barker at noaa.gov  Tue Apr  2 12:10:53 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 09:10:53 -0800
Subject: PyArg_BuildValue, how to return a PyString object?
References: <5a4226f0.0204012004.24bb7b00@posting.google.com> <lkelhya3rg.fsf@elios.maths.bris.ac.uk>
Message-ID: <3CA9E603.F950315@noaa.gov>


Michael Hudson wrote:

> No.  Rememebr to DECREF "O" args after calling Py_BuildValue --
> something I'm always forgetting.

Can't you use "N" instead? That seems to have worked for me.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From aahz at pythoncraft.com  Thu Apr  4 16:31:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 16:31:55 -0500
Subject: Anonymous functions (and classes)
References: <mailman.1017937990.14515.python-list@python.org>
Message-ID: <a8igob$3el$1@panix1.panix.com>

In article <mailman.1017937990.14515.python-list at python.org>,
Hellen Claggett  <hellen at claggetts.net> wrote:
>
>Here is a relativly simple feature that I would like to use while
>programming in Python. First, I did make a (brief) effort to determine
>if this specific request had been made and found nothing obvious in the
>PEPs or in this year's python-list archives.
>
>I would like to be able to define anonymous functions and classes by
>omitting the function or class name normally found after the def or
>class keyword. 

Feel free to write a PEP so that this abomination can be rejected once
and for all.  ;-)  (I don't think there have been any large threads on
this in 2002, but there certainly were at least two or three in 2001.)

(Yes, I can be quite certain it will be rejected, given Guido's regret
over including lambda in the first place.  Python has rightly included
many useful bits from functional programming languages, but Python is
also emphatically *NOT* a functional language itself.)

If you don't want functions cluttering up a namespace, put them in
another module.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From nospam-readonly at getsoft.com  Mon Apr 22 00:12:56 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 14:12:56 +1000
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net> <3cc3765f$0$15478$afc38c87@news.optusnet.com.au> <ddLw8.50678$HH5.4019@rwcrnsc51.ops.asp.att.net>
Message-ID: <3cc38db1$0$15477$afc38c87@news.optusnet.com.au>

Thanks for that Emile.

One thing I've noticed now is for larger files (73,000 lines) ndiff can take
several minutes on my 1G PC, but that is another issue again. I assume the
fact that it is being interpreted has quite a bit to do with this.


"Emile van Sebille" <emile at fenx.com> wrote in message
news:ddLw8.50678$HH5.4019 at rwcrnsc51.ops.asp.att.net...
> Neville Franks
> > Ok thanks. I've installed 2.2 and ndiff is available but I'm still
> confused.
> > I thought that Tim's original ndiff compared files, but the one in
> difflib
> > compares lists of strings. Does this mean I somehow need to read the
> files
> > into lists of strings and then call ndiff. A short example of how to
> use
> > ndiff to compare two files would be greatly appreciated.
> >
>
> David pointed you to python22/tools/scripts/difflib.py.  In it you'll
> find this example of how to use it:
>
> def fopen(fname):
>     try:
>         return open(fname, 'r')
>     except IOError, detail:
>         return fail("couldn't open " + fname + ": " + str(detail))
>
> # open two files & spray the diff to stdout; return false iff a problem
> def fcompare(f1name, f2name):
>     f1 = fopen(f1name)
>     f2 = fopen(f2name)
>     if not f1 or not f2:
>         return 0
>
>     a = f1.readlines(); f1.close()
>     b = f2.readlines(); f2.close()
>     for line in difflib.ndiff(a, b):
>         print line,
>
>     return 1
>
>
> You could then say something like:
>
> fcompare(r'c:\CrunchyFrog.txt', r'c:\CrunchyFfruug.txt')
>
> Assuming-you-were-casting-Peter-Sellers-in-a-skit-ly y'rs,
>
>
> --
>
> Emile van Sebille
> emile at fenx.com
>
> ---------
>




From cliechti at gmx.net  Sat Apr  6 15:43:40 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 21:43:40 +0100
Subject: beginners syntax question
References: <3caf47a5@news1.homechoice.co.uk> <3CAF4E4B.ECE28871@ntlworld.com>
Message-ID: <Xns91E8DD351888Acliechtigmxnet@62.2.16.82>

tony clarke <a.clarke11 at ntlworld.com> wrote in 
news:3CAF4E4B.ECE28871 at ntlworld.com:

> Hi
> 
> Try listing the objects  l=["object1','object2'..]
> then
> for i in l:
>  i.update()

almost, leave away the quotes:
l=[object1, object2]
 
>> I'm trying to call the update() method of several objects ie:
>>
>> object1.update()
>> object2.update()
>> etc
>>
>> I'd really like to do
>>
>> for x in range(7):
>>   "object"+str(x).update
>>
>> but can't figure outthe syntax, is this possible?

yes it is, but there are better ways.
why are you not storing the objects in a list
instead of naming them with numbers?

calling methods on objects in a list is shown above.

chris

-- 
Chris <cliechti at gmx.net>



From starx at pacbell.net  Wed Apr 10 17:48:07 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 10 Apr 2002 21:48:07 GMT
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <Xns91EC96C0C8E9CRASXnewsDFE1@130.133.1.4>

Joel Bender || Wed 10 Apr 2002 01:35:10p:

> Hi,
> 
> 
> I would like to process the elements of a list that match a condition, 
> then remove them.  I'll use integers in my example, but my list items 
> are actually objects.
> 
> What's the "right" way to do this?  For example:
> 
>     >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>     >>> dothese = [i for i in lst if i < 5]
>     >>> lst = [i for i in lst if i >= 5]
>     >>> for item in dothese:
>     ...     print item
> 
> I'd rather not scan the list twice, particularly if there's nothing to 
> do.  I'd rather not build another double-linked list.
> 
> Is there something like this?
> 
>     >>> for item in lst:
>     ...     if (item < 5):
>     ...         print item
>     ...         del lst[ xxxx ]
> 
> But the "trick" is figuring out what xxxx should be.  Using lst.remove
() 
> would work if the internals used "is" and not "is equal to" and the 
> removal didn't modify the list where it would break the loop somehow.
> 
> I could rebuild the list for things that should not be processed:
> 
>     >>> newlst = []
>     >>> for item in lst:
>     ...     if (item < 5):
>     ...         print item
>     ...     else
>     ...         newlst.append(item)
>     >>> lst = newlst
> 
> But these seems like a lot of work when there's nothing in the list to 
> process.

Two approaches come to mind, both using while instead (Because we need 
to use the number instead of just the item):

i = 0
while i < len(list)
    	if list[i] < 5:
    	    	print list[i]
    	    	del list[i]
    	i += 1

and

i = 0
used = []
length = len(list)
while i < length
    	if list[i] < 5:
    	    	print list[i]
    	    	used.append(i)
    	i += 1
for x in used
    	del list[x]

If you want to not modify the list in the using loop, which requries 
checking the new length each time (don't know how expensive this is in 
py)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From Oschler at earthlink.net  Mon Apr  1 00:07:29 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 00:07:29 -0500
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <a886hv$st6$1@panix1.panix.com> <J4Op8.61588$Q7.48627836@e3500-atl1.usenetserver.com> <mailman.1017628802.8651.python-list@python.org>
Message-ID: <xORp8.63088$Q7.49964592@e3500-atl1.usenetserver.com>

Gerhard,

So what then do you think?  BlueFish, Quanta, WebSphere Homepage builder?

thx

Gerhard H?ring <gh_pythonlist at gmx.de> wrote in message
news:mailman.1017628802.8651.python-list at python.org...
> > I didn't know Zope also had a WYSIWIG editor for HTML pages.
>
> It doesn't.
>






From ftobin at neverending.org  Mon Apr 29 12:40:00 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 29 Apr 2002 16:40:00 GMT
Subject: comparing str's with ints
Message-ID: <20020429115856.R1840-100000@palanthas.neverending.org>

This makes sense (arithmetic comparison):
>>> 3 > 5
0


This makes sense (lexigraphic comparison):
>>> 'a' > 'b'
0


BUT, *this* does not make sense:
>>> '5' > 3
1

What in the world is supposed to be the semantics of comparing a string
with an integer?  Seems like a type trap...

-- 
Frank Tobin		http://www.neverending.org/~ftobin/



From brueckd at tbye.com  Thu Apr 11 12:52:38 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 11 Apr 2002 09:52:38 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <BIEJKCLHCIOIHAGOKOLHOEJFCPAA.tim.one@comcast.net>
Message-ID: <Pine.LNX.4.33.0204110946060.26004-100000@ns0.tbye.com>

On Thu, 11 Apr 2002, Tim Peters wrote:

> [brueckd at tbye.com]
> > ...
> > I'd *love* to have dictionary comprehensions!
>
> Try this on for size:
>
> >>> dict([(i, i**2) for i in range(10)])
> {0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25, 6: 36, 7: 49, 8: 64, 9: 81}
> >>>
>
> That is, 2.2 dict() accepts an iterable object producing iterable objects
> each producing two values.  A list of 2-tuples is one instance of that, and
> one particularly easy to create with a listcomp.  More possibilities open if
> you ponder other kinds of iterable objects; e.g.,
>
> >>> def squarepairs(n):
> ...     for i in range(n):
> ...         yield i, i**2
> ...
> >>> dict(squarepairs(10))
> {0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25, 6: 36, 7: 49, 8: 64, 9: 81}
> >>>
>
> is one way to get the same end result without building a list in one gulp
> first.

Woo-hoo! "Upgrader's delight" is Python's antithesis of "buyer's remorse";
you upgrade to get some new feature only to go back later and discover
additional gems as well.

Thanks for the tip!

-Dave





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 22:40:49 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 19:40:49 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com> <a9fcut$e9l$0@216.39.172.122>
Message-ID: <3cbb8f83@news.mhogaming.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:a9fcut$e9l$0 at 216.39.172.122...
| On Mon, 15 Apr 2002 00:06:34 -0700, "Ken Peek"
<Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
|
| >
| >"Martin v. Loewis" <martin at v.loewis.de> wrote in message
| >news:m34ridmq00.fsf at mira.informatik.hu-berlin.de...
| >
| >| That's what I'm trying to tell you all the time: the
| >| '>>>' operator is meaningless if it is defined as
| >| "fill in zeroes". How does it know where to start
| >| inserting zeroes?
| >
| >No, it isn't meaningless:
| >
| >The 'long' class has an internal representation for the
| >long number.  The number of bytes that are currently
| >being used to contain the number are known (internally)
| >to the object.  The zeroes get shifted into the high bit
| >of the number, no matter how many bytes are used to
| >contain the number.
| >
| Actually the longs may not be byte-based at all. They could
| be based on an array of shorts with 15-bit values. You don't
| want to see the actual bit pattern of that, do you? At least,
| not as a general purpose thing.

What I want, is an object that "looks and feels" like a 'binary twos-complement
integer', that can be extended a byte at a time.  I don't _CARE_ how the machine
represents it internaly.  It might even be in BCD for some weird CPU (for all we
know)-- but the interface that _I_ see, is a 'binary twos-complement integer'
made up of a finite number of bytes.  I don't want to know HOW the machine
STORES it-- I just want a consistent interface to it.

When I print it out (in decimal)-- I want it to appear the way it does now--
sign and magnitude.

When I print it out in hexadeciaml, I want the bit representation of a 'binary
twos-complement' number, of a size (in bytes) that will represent the number
without truncation (regardless of HOW the number is stored inside the machine.)

| >An int type simply shifts a zero into bit 31, so that
| >this works the same as it will on a long.

Let me clarify the above-- the '>>' should behave in a manner such that the sign
bit is COPIED to the high bit during a right shift.  So-- if the number starts
out as a positive number, it STAYS a positive number after the shift.  The
proposed '>>>' operator forces a zero into the high bit, having no effect on the
sign bit of a positive number, but forcing a negative (twos-complement) number
to now be positive after the shift.  Again-- I don't care HOW the number is
represented internally-- this is the way it should always appear to the
programmer...

| >The '>>' operator should also work the same for a long
| >as it does for an int.
| It can't, because the long does not represent a fixed-width
| bit pattern, it represents an infinite bit pattern. The infinite
| comes from extending the sign bit. Thus shifting in zeroes at
| the 'top' of a negative number implies shifting them in at bit
| position infinity. You can say that we don't actually use an
| infinite number of bits in the internal representations, but
| the/a point of numeric unification is to hide that representation.
| All you can legitimately refer to is the implied abstraction, which
| has indefinitely extended sign bits.
|
| If you do choose some top bit in the indefinite extension as 'the'
| sign bit, it should be done consistently, and what you are doing
| should be clear. For example, if you scanned down from the left until
| you found two adjacent differing bits (or a single bitif you reach the
| bottom) you could say the ms bit of those was 'the' sign bit. Now you have
| a width, and this can be computed for any integer, positive or negative.
|
| On the basis of that width, you could shift in zeroes. I'm not sure how
| useful that is, but I would say it would be more consistent that picking
| an arbitraryly occuring k*4 or k*8 or k*15 bit width value that you might
| get from the current state of an internal representation.
|
| In any case, you can define a function to whatever you actually want.
| (I suspect you'll wind up specifying width explicitly). But if
| you want a >>> operation, what width should it use? Remember, different
| platforms may use different internal long representations.

You CANNOT do this in a portable way from OUTSIDE of the object.  Only the
OBJECT "knows" how many bytes in an equivalent twos-complement binary number are
required to represent the number contained in the object.  Even if the internal
representation is in BCD, or a funny "17-bit per word" [an old IBM machine did
this] format, the programmer should always "see" an array of BYTES that
represent a twos-complement binary integer.  If the long class overloads the
'>>>' operator, and it "knows" how many "pseudo-bytes" are needed to represent
the number, then it can damn well "do the good thing" when this method is called
(as an operator on the long.)  YOU however, are not supposed to be peeking
inside the object (OOP rule #1-- encapsulation)-- let the object figure out how
to do the '>>>' (and the '>>' and '<<' for that matter.)

| >The '<<' operator should also work the same for a long
| >as it does for an int.
| >
| That one's easy ;-)
|
| >The 'hex()' method should work the same for a long as
| >it does for and int.
| >
| I think that's agreed too, just not the details. I agree with you that
| sign/magnitude is not very useful if you're interested in bits. (I might
| be referring to different bits than you though).

See my comments above-- IF we ALWAYS have the "look and feel" of a binary
twos-complement integer that is ALWAYS made up of an array of BYTES (no matter
HOW it is represented internally to the machine), then we can ALWAYS print a
hexadecimal in the twos-complement form, without any silly sign characters...
Example:

>>> a = 0x43210986

# a is an 'int' now, (on most machines,
# but that is none of our business)

>>> b = a << 1

# b is now a long, equal to: 0x008642130c, or the
# equivalent of 5 bytes (no matter HOW it is
# represented internally.)  This is because we need
# the extra byte to convey the sign.  We could
# shift this left 7 more times, without having to
# add another byte...

>>> print b
2252477196

>>> b = -b  # change sign

>>> print b
-2252477196

>>> print hex(b)

0xff79bdecf4     # no 'L' when version >= 2.3.x!!

# THIS is what _I_ expect--  the number of "bytes"
# (or at least the equivalent) is still the same,
# and we can see that the number is now negative,
# and the number of "bytes" is still 5 (no matter
# HOW it is represented internally), so we now we
# print 10 digits for 5 bytes...

# Now, if we type:

>>> d = 0x8642130c # d is a 'long' on most machines

>>> print type(d)
<type 'long'>

# when we reach Python version >= 2.3.x:
>>> d = d >> 1  # d is NOW an 'int' on most machines
>>> print type(d)
<type 'int'>

>>> print d
0x43210986

# ===============================

And now, to address the so-called "type casts" 'int()'
and 'long()'-- well, these just aren't needed anymore
because the system handles this automatically now, so
they are deleted from Python...

You see? Maybe I should have provided some examples.

This is how _I_ believe things should work.





From logiplexsoftware at earthlink.net  Wed Apr 10 19:47:57 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 16:47:57 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>
References: <mailman.1018408041.12582.python-list@python.org>
	<Xns91EBD90A7449Dlumberjack@216.148.53.83>
	<slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de>
	<a91mdd$ck$1@panix1.panix.com>
	<slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>
Message-ID: <20020410164757.57b760b5.logiplexsoftware@earthlink.net>

On 10 Apr 2002 23:42:36 GMT
Gerhard H?ring wrote:

> Aahz wrote in comp.lang.python:
> > In article <slrnab7hn9.4tj.gerhard at lilith.my-fqdn.de>,
> > Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
> >>
> >>Why would anybody care if you aren't man/woman enough to post with
> >>your real name?
> > 
> > Because many people, respectable people, *do* post to Usenet without
> > using their "real name".
> 
> Yes, there are such people, but they are rare. For one of those, there
> come a hundred trolls, 1337 hax0rz and their like.
> 
> > What the heck is a real name, anyway?
> 
> In western contries, firstname + lastname.
> 
> IMO using your real name is polite and whatever you say is not worth
> being said if you can't put your name under it.
> 
> Gerhard

And you expect us to believe that "Gerhard" is a *real* name?  And that your
parents decided to use Unicode when they named you?  Ha.  Not buying it,
whoever you are.

;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From max at alcyone.com  Tue Apr  9 14:01:39 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 09 Apr 2002 11:01:39 -0700
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>
Message-ID: <3CB32C83.AB67AFFC@alcyone.com>

Magnus Lie Hetland wrote:

> That's not how I saw it. The main point I extracted was: Until now,
> the Pythonic "truth values" have been "something" and "nothing" (i.e.
> the empty list, zero, None, etc.) With the bool type this is still
> true, but we get a type which will give newbies the impression that it
> isn't.

So explain it to them.  It only takes a sentence or two.  The bool type
just provides one more value that's a "nothing," and one more value
that's a "something," for cases where what you're interested in is truly
nothing but a truth functional value.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From geiseri at yahoo.com  Tue Apr  9 08:51:44 2002
From: geiseri at yahoo.com (ian reinhart geiser)
Date: Tue, 9 Apr 2002 12:51:44 -0000
Subject: my own sip/pyqt bindings on win32
Message-ID: <E16uv6X-0000H4-00@mail.python.org>


 
Greetings
 
I am trying to get my python bindings from linux to run on windows.  I am 
 
getting a few linker errors though.  I looked and see that I need the -
 
DSIP_USE_DLL and -DQTDLL flags.  Has anyone else done this before?  I am 
 
basicly writeing a wrapper arround a widget I wrote, so it I have to link in 
 
both libsip.lib and libqtc.lib.  Are there funky -D flags for libqtc.lib?  I 
 
cannot find any.
 

 
Thanks
 
-ian reinhart geiser
 
-- 
 
Ian Geiser <geiseri at yahoo.com>
 
http://geiseri.myip.org:8080/~geiseri/
 

 


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From robin at jessikat.fsnet.co.uk  Fri Apr 26 06:02:53 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 26 Apr 2002 11:02:53 +0100
Subject: Is it me or a python wart
Message-ID: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>

With 2.2.1 I'm puzzled why the first returns the local x and the second
the global I suppose x=x rather than z=x is the problem, but don't
understand why. I opposed local scopes originally as I liked the two
scope rule, but this seems more complex than I imagined.

>>> x=2
>>> def bingo(x):
...     class A:
...             z=x
...             def get(self):
...                     return self.z
...     return A().get()
... 
>>> bingo(4)
4
>>> def bango(x):
...     class A:
...             x=x
...             def get(self):
...                     return self.x
...     return A().get()
... 
>>> bango(4)
2

-- 
Robin Becker


From gmcm at hypernet.com  Tue Apr 30 08:43:25 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 30 Apr 2002 12:43:25 GMT
Subject: Access violation trying to run Python over Win NT network.
References: <ucrs88tkj2ttc7@corp.supernews.com>
Message-ID: <Xns920058DCEAB82gmcmhypernetcom@199.171.54.215>

Patrick wrote:

> I'm trying to run Python from a network drive.  On most machines it
> works, but on some machines it generates an access violation about 60%
> of the time, (just by typing 'python' with no arguments).  If I install
> Python locally it runs fine on all the machines.

One possibility is that the install is updating some MS redistributable
dll. This goes into the server's Windows\System. The client is still
using an out of date copy from the client machine's Windows\System.

That would explain why it doesn't fail on Win2K boxes (as you mention
in a later post).

-- Gordon
http://www.mcmillan-inc.com/


From geoff at gerrietts.net  Mon Apr 29 14:03:40 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 29 Apr 2002 11:03:40 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
In-Reply-To: <93d52e82.0204290618.349a0e5f@posting.google.com>
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <aaevge$emr$1@peabody.colorado.edu> <93d52e82.0204290618.349a0e5f@posting.google.com>
Message-ID: <20020429180340.GA15013@isis.gerrietts.net>

The use of "pass" in the example doesn't actually do anything. I think
what you're looking for is "continue". I can't explain the dramatic
speedup you got out of adding them. It defies logic.

The "pass" keyword is a no-op. It's a placeholder. The idea is that
you might want to do something like this:

class Spam:
  pass

In this example, "pass" is a way to indicate that you don't want
anything special -- your Spam class can use a default constructor, and
needs no special methods. But you can't just write:

class Spam:

That's a syntax error. So "pass" allows you to say "do nothing
special". This is also a way you can insert a virtual method:

class Spam:
  def subclassed_operation(self):
    pass

But "pass" isn't always the preferred approach. Lots of people have
commented that better form might be:

class Spam:
  def subclassed_operation(self):
    """ this operation must be overridden by the subclass if it is to
        do anything.
    """

The doc string counts as a "statement" for the purposes of the
compiler, making this valid syntax. Meanwhile, it provides explicit
explanation of your intent.

Still others will prefer both: have a doc string AND a pass.

The "continue" keyword says "let's go back to the top of the loop and
do the next item". The "break" keyword says "let's exit the loop right
now, even if there's more work pending".

Those are your loop-control statements.

Luck,
--G.


-- 
Geoff Gerrietts             "If life were measured by accomplishments, 
<geoff at gerrietts net>     most of us would die in infancy." 
http://www.gerrietts.net/       --A.P. Gouthey




From huaiyu at gauss.almadan.ibm.com  Fri Apr  5 19:04:40 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 6 Apr 2002 00:04:40 +0000 (UTC)
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050112.11d4ef89@posting.google.com>
Message-ID: <slrnaaseso.99d.huaiyu@gauss.almadan.ibm.com>

On 5 Apr 2002 01:12:46 -0800, Alan Kennedy <alanmk at hotmail.com> wrote:
>brian d foy <comdog at panix.com> wrote:
>
>> for our April 2002 issue (April Fool!), The Perl Review replaced its
>> Perl content with Python articles, so Python folks might want to read
>> this issue:
>> 
>>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf
>

This is an interesting read.  Unfortunately, the first paragraph of Paul
Prescod's article contains a glaring trivial error of Fibonacci numbers.  It
must have been calculated by hand, as his program is correct.  Wonder what
impression that will make on the Perl audience.  

Huaiyu



From lac at strakt.com  Sat Apr  6 11:33:32 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 18:33:32 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from "John Roth" <johnroth@ameritech.net> 
   of "Sat, 06 Apr 2002 10:14:28 CDT." <uau3rsge16bp1e@news.supernews.com> 
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> 
Message-ID: <200204061633.g36GXWdj017461@ratthing-b246.strakt.com>

+> 
> "Alex Martelli" <aleax at aleax.it> wrote in message
> news:gJDq8.55255$S52.1904212 at news2.tin.it...
> > John Roth wrote:
> >         ...
> > > Maybe my post in response didn't propagate? As I said in that
> > > post, the root of her problem is a poorly chosen name for the
> > > function. If she paid attention to constructing the name properly,
> >
> > She adopted the name *Guido* had given to the function right in
> > his pep (actually clarifying it a bit from 'isleap' in his PEP
> > to 'isLeapYear', more readable, in her post).
> 
> Sorry I'm late in responding to both posts - I had a problem
> with my newsgroup vendor and I'm just back on.
> 
> I thought the original example was very bad, but I didn't
> say so at the time because I hadn't figured out what I didn't
> like about it.
> 
> The original example was something like
> 
> number of days in year = 365 + isLeap(year).
> 
> To put it bluntly, that's a very confusing hack that depends
> on conceptual True being 1 and conceptual False being
> zero.
> 
> Again, to put it very bluntly, Laura swallowed a camel
> while straining at a gnat. She had a great deal to say about
> the difficulty of teaching that example, and I agree completely.
> It would be very difficult to teach, but the reason is that
> the name is simply wrong. The fact that Guido himself
> used that as an example doesn't make it right, either.
> 
> If she had corrected the name, then her entire basis for
> the objection would have evaporated.
> 
> John Roth
> >
> >
> > Alex
> >

Still haven't found the time to read my original post, notice that
I _did_ correct the name in the original post, and understand what
I was trying to say, I see.  I will correct your misconception
because you might confuse somebody else.

I am objecting to the fact that isleap, whatever the hell it is
called, returns a boolean.  I am rejecting the existence of
booleans in the language because it is too hard to teach how
to use them properly.  Most people never learn.  Most code that
uses booleans should not.

Proper use of booleans requires not only that you have a 
function that returns one of two values, but THAT YOU NEVER
EVER WANT THAT FUNCTION TO EVER BE MODIFIED TO RETURN ONE OF THREE,
(or more).  What I tell my students is that you really want 
a hole to open in the ceiling and a Warner Brothers safe to fall
on the head of somebody who attempts to extend, subclass, or 
in any other way change the function to return one of three values.

And, by God, if the function that returns _the number of leap days
in the year_ isn't safe, at 0 or 1, then damn near close to nothing
that my students are going to be tempted to use 'True' and 'False'
for is.  This example makes my course notes, under Why using booleans 
is bad, and you should never do it.  (unless you are a mathematician
who wants to do real 2-valued algebra, or somebody interfacing with
some other piece of software that needs those sort of bools, or 
some other, EXTREMELY RARE,  person.  Don't you guys interrupt here.
This is hard enough to teach as it is.)

It does not help that it appears to the outside that Guido just went
through all the python functions that returned 1 or 0 and changed
them to return True or False.  I know that Guido would not be
so stupid, and that there is a reason why he wants, for instance,
isinstance to return only 0 or 1 for evermore, but I can't see it.
I would like an isinstance that returned 0 for not an instance of
this class, 1 for a direct instance, and 2 for a subclass.  There
must be a lot of code out there I have never seen that relies on
this never happening, and now, alas, I have to find out why.

Because my students will ask, of course.  If True and False are
good enough for Guido, then it's good enough for them.  I would
love it if a huge comment was added to python files that now
return booleans saying why it was necessary to restrict the
return values to one of 2 forever more, so people are not confused
into thinking that returning booleans is a good idea in general,
but that may be too much to ask for.  I would add the comments
myself, but I am having problems seeing what the reasons are.
Maybe after I get some more sleep, it will come clearer to me.

I truly liked telling students, especailly ones that have never used
Python while we are using some other language,   'You don't need 
True and False. Look at Python.'  

Now I have to say, to the students who are using
Python: 'Python has True and False.  It is not for you little
         worms, and if I catch you using it I will make you do 
         the assignment over.'

I know in all the Python code I have ever written I have only 5 times
wanted a boolean True or False -- all times when I was talking to a
database that required it.  The rest of the time I wanted an integer,
0 or 1.  Even if I couldn't see a way that somebody was likely to
extend it to return 3, there most certainly was never ever going to be
any code that required that it only have one of 2 values, so now I
will have to change all my True and False there to be 'Yes' or 'No' or
something.  (grumble, now I wish I had lost the argument to the person
who wanted the constants we picked to be TRUE and FALSE).

Teaching people to not use booleans is hard.  They grab for them
whenever they have a function that returns one of two values.  It
was a lot easier to teach this concept when the fool things were not in
the language.

Note:  don't tell me that because python bools are just ints, it
       doesn't matter -- people who were using bools for
       DEMOCRAT and REPUBLICAN can now add 
       THE_DAMN_VOTE_WASNT_PUNCHED_THROUGH - (the integer 3)
       when to their horror they discover 2 values was not enough.
       I need to teach students to not use bools in languages which
       are not so forgiving.

Laura Creighton




From logstx at bellatlantic.net  Thu Apr  4 17:14:12 2002
From: logstx at bellatlantic.net (logistix)
Date: Thu, 04 Apr 2002 22:14:12 GMT
Subject: Anonymous functions (and classes)
References: <mailman.1017937990.14515.python-list@python.org>
Message-ID: <Ue4r8.2400$WQ4.329@nwrddc02.gnilink.net>

>
> I would like to be able to define anonymous functions and classes by
> omitting the function or class name normally found after the def or
> class keyword. For example, the following two function definitions would
> be equivalent:
>

That's why god created nested scopes (or one of the reasons at least):

>>> def factory():
...  def foo():
...   print "bar"
...  class bar:
...   def foo(self):
...    print "baz"
...  instance = bar()
...  return foo, instance
...
>>> a,b = factory()
>>> a()
bar
>>> b.foo()
baz
>>> foo()
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
NameError: name 'foo' is not defined
>>> x = bar()
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
NameError: name 'bar' is not defined
>>>


And if you want to clobber factory() from the namespace:

>>> del factory
>>> factory()
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
NameError: name 'factory' is not defined




From matt at greenroom.com.au  Tue Apr  9 11:28:44 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Wed, 10 Apr 2002 01:28:44 +1000
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au> <a8v0mo$j50$1@newsg1.svr.pol.co.uk>
Message-ID: <3CB308AC.5040405@greenroom.com.au>

? wrote:

> Matthew Baulch <matt at greenroom.com.au> wrote in message
> news:3CB3039C.3040108 at greenroom.com.au...
> 
>>I think the code you are looking for could be:
>>
>>x = 2
>>
>>def wrong(num):
>>   return num + 1
>>
>>x = wrong(x)
>>print x
>>
>>x = wrong(x)
>>print x
>>
>>
> 
> I'm a newbie myself, so this is probably wrong, but wouldn't this be *even*
> better?


> x = 2
> def wrong(num):
>     x = x+1
>     print x
> 
> wrong(x)
> wrong(x)
> 


Not wrong, just different!



From jon+usenet at unequivocal.co.uk  Wed Apr 24 17:05:20 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 21:05:20 -0000
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <slrnacag2g.1u3.gerhard@lilith.my-fqdn.de> <aa63ju$nva$1@news.ilk.de> <7uxx8.44918$0x3.9466@atlpnn01.usenetserver.com>
Message-ID: <slrnace7gg.hoe.jon+usenet@snowy.squish.net>

In article <7uxx8.44918$0x3.9466 at atlpnn01.usenetserver.com>, Steve Holden wrote:
>> curs.execute("""insert into table (col1, col2,...) values ('%s',
>>         '%s',...)""" % (val1, val2, ...)
> 
> Your version (which I used to suggest myself, so it's not like I'm accusing
> you of a crime <wink>) generates a unique SQL statement which is guranteed
> to be different from the next, and the one after that, depriving the SQL
> implementation of the chance to optimize.

... it is also always a bug, because some characters such as ' will
cause syntax errors, and also, in many situations (for example, web
applications) it is a HUGE SECURITY HOLE since it means the Evil
Hacker can send arbitrary SQL queries to your database server.


From roy at panix.com  Mon Apr  1 07:48:57 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 01 Apr 2002 07:48:57 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net>
Message-ID: <roy-502972.07485701042002@news1.panix.com>

Stefan Schwarzer <s.schwarzer at ndh.net> wrote:
> Example: "(val1, val2)[0]" is pretty clear, but "(val1, val2)[False]"
> looks quite strange.

Does it look any stranger than "False : val1 ? val2" ?


From amckay at merlinsoftech.com  Mon Apr 22 20:06:01 2002
From: amckay at merlinsoftech.com (Andy McKay)
Date: Mon, 22 Apr 2002 17:06:01 -0700
Subject: Using a Python script in Zope
In-Reply-To: <yi1x8.52892$ro5.4759478@bin6.nnrp.aus1.giganews.com>
References: <yi1x8.52892$ro5.4759478@bin6.nnrp.aus1.giganews.com>
Message-ID: <200204230002.g3N02qq23474@mail.merlinsoftech.com>

Use an external method, python scripts cannot read from the files because of 
security issues.

On April 22, 2002 04:47 pm, Steven James wrote:
> I have a script that I wrote in Python to parse a text file and return a
> dictionary of the values that I get from it. I used file.readline()
> extensively to get the values from the file, but I can't seem to get this
> function to work in Zope, where I intend to use the script. Anyone have
> ideas on another way to do it, or how to make this work?

-- 
  Andy McKay




From pyth at devel.trillke.net  Mon Apr 29 14:49:35 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 20:49:35 +0200
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <aak1uq$bgt$1@bob.news.rcn.net>
References: <aak1uq$bgt$1@bob.news.rcn.net>
Message-ID: <20020429184935.GO16217@devel.trillke>

On Mon, Apr 29, 2002 at 02:03:02PM -0400, Raymond Hettinger wrote:
> From the comments, I've learned:
> 
> 1. map, filter, divmod, and pow may be immovable and should
>     be left alone.
> 2. reduce has fewer friends, but should probably also be
>     left alone
> 
> 3. input() has almost no supporters and should probably be
>     deprecated.  it is security risk, a trap for new users, easily
>     replaced with eval(raw_input), and is named in a way that
>     doesn't suggest that eval is occurring.
> 
> 4. apply() was repeatedly suggested as another candidate for
>     deprecation since the * and ** forms make it obsolete
> 
> 5.  oct() is another candidate based on the rarity of usage
> 
> 6. command line options and environment variables are
>     universally hated, so this idea is dropped
> 
> 7. the language needs some means of deprecating so that clutter
>     does accumulate.  that means needs to be well documented,
>     have a long, slow phase-in, and have a mechasism for
>     restoring old behavior if it is ever needed.
> 
> 8. there is some lumberjack person who is very angry
> 
> 9. putting new functional tools in a separate module was
>     well received.

Thanks for the good summary (especially 8)

    Holger




From skip at pobox.com  Tue Apr  9 18:26:21 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 9 Apr 2002 17:26:21 -0500
Subject: Managing extension module build using automake & friends?
Message-ID: <15539.27277.799749.785291@12-248-41-177.client.attbi.com>

I have a small library whose configuration and build process is managed with
the GNU config tools (autoconf, automake, and libtool).  In one subdirectory
is a straightforward Python extension module that wraps the library.  Before
I go reinventing wheels, does someone already have the Makefile.am magic
necessary to build an appropriately linked .so file?

Thanks,

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From loewis at informatik.hu-berlin.de  Tue Apr  2 10:08:36 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Apr 2002 17:08:36 +0200
Subject: file.write[lines]() weirdness
References: <3ca96e30.52835380@mammoth.usenet-access.com>
Message-ID: <j4adsm2l57.fsf@informatik.hu-berlin.de>

netzapper at magicstar.net (A. Jones) writes:

> Any insight would be much appreciated.

Seeking and text mode does not really mix well. When you write to a
text file, the C library may expand or collapse newline characters,
and perform other modifications. So if you write n bytes to a text
file, that doesn't mean that n bytes end on disk.

I recommend to open the file in binary.

Regards,
Martin



From flyguysyl at hotmail.com  Sat Apr 13 20:12:00 2002
From: flyguysyl at hotmail.com (flyguy)
Date: 13 Apr 2002 17:12:00 -0700
Subject: textarea to a python sript
Message-ID: <c73096a5.0204131612.700cd7ae@posting.google.com>

How can I get a <TEXTAREA> to a python script?


From aleax at aleax.it  Mon Apr 29 12:08:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 16:08:59 GMT
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com> <13285ea2.0204290619.3563152f@posting.google.com>
Message-ID: <veez8.31957$8D3.895347@news1.tin.it>

J.Jacob wrote:
        ...
>     quote = ''
>     if not quoteslist:
>         for s in quotesfile.readlines():
>             if s[0] == '%':
>                 if quote:
>                     quoteslist.append(quote)
>                     quote = ''
>             else: quote = quote + s
>         if quote: quoteslist.append(quote)

It appears to me that:

    quoteslist = quotesfile.read().split('\n%')

should be roughly equivalent to this -- except
for removing the trailing newline of each item
of quoteslist; some variation on this split,
depending on what else you want to have on
those lines starting with %, perhaps with a
following cleanup step if those lines starting
with % have to allow arbitrary comments to be
removed, might in any case be faster & handier,
as long as you're building the quoteslist in
memory anyway.


Alex



From sholden at holdenweb.com  Mon Apr 22 19:20:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 19:20:39 -0400
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019510980.7097.python-list@python.org> <3CC48EDD.2BEE5215@tundraware.com>
Message-ID: <qX0x8.82827$T%5.64632@atlpnn01.usenetserver.com>

"Tim Daneliuk" <tundra at tundraware.com> wrote ...
> arthur.siegel at rsmi.com wrote:
> >
> > >>>Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
> > >>>would-be applications to the natural sciences. I think it was P.
Erdos
> > >>>who remarked once that all natural science is conceivably finite -
e.g.
> > >>>there would be a time when there would be nothing more to learn about
> > >>>the universe - but mathematics is literally infinite.
> > >
> > >>I'm not so sure about the latter assertion, and I'm *definitely* not
> > >>certain about the assertion that natural science is finite.
> >
> > >The first assertion by Erdos is indeed problematic. The second is not.
> >
> > My sense, intution is of mathemetics as infinite only in the sense of
allowing
> > infinite homologous descriptions of/perspectives on  the One Thing.
> >
> > The perspectives are invented.
> >
> > The Thing is not.
>
> This is entirely cogent and on-the-mark, IMO, but it still sidesteps the
> question of whether or not mathematics can operate on things which do
> not exist in the natural realm.  My (unproven/unprovable) claim is that
> mathematics is capable of operating on a larger problem set than
> that posed by natural science no matter how large/finite/infinite natural
> science, in fact, is.
>

I would still have a problem with the Godelian limits of formal systems and
the intellect's ability to comprehend things more complex than itself. Can
methematics do things the human intellect cannot comprehend? Mathematics
*is*. Humans *do* mathematics (discovered or invented, it has to be used by
an intellect).

In the end you can draw limits around natural science that allow you to
answer your own question either way. Maybe Cambridge had it right in
including both mathematics and science under "Natural Philosophy".

While interesting, however, this is all a rather long way from Python
["spoilsport!!", shouted the crowd]. Next: "How many angels can dance on the
head of a pin?"

regards
 Steve
--
home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From jeff at ccvcorp.com  Fri Apr 19 14:21:40 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 11:21:40 -0700
Subject: GNU GPL & Python stand-alone packages from py2exe
References: <i0kit6nn3l4.fsf@Thales.MI.Uni-Koeln.DE> <j4hem7u2ol.fsf@informatik.hu-berlin.de>
Message-ID: <MPG.1729fff0170354d59896b8@news.nwlink.com>

In article <j4hem7u2ol.fsf at informatik.hu-berlin.de>, 
loewis at informatik.hu-berlin.de says...
> Ulrich Goertz <u at g0ertz.de> writes:
> > 2) Would I have to put the source code of the Python interpreter
> > on the CD, too? 
> 
> Upon reading the GPL, that indeed appears to be the case:
[...]
> So the GPL applies to the work 'as a whole', and therefore, the same
> source distribution clause applies.

Correct me if I'm wrong, but... I don't think that the GPL 
requires that the source be distributed on the same media, just 
that it be "available".  In which case, simply having the 
appropriate pointers to www.python.org and/or Sourceforge should 
be sufficient.  But this *does* depend on the exact terms of the 
GPL, which you should verify for yourself.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From borcis at geneva-link.ch  Mon Apr  8 08:17:17 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 08 Apr 2002 14:17:17 +0200
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org> <eppstein-4311CC.17190504042002@news.service.uci.edu> <3CB13147.89F02706@cosc.canterbury.ac.nz> <eppstein-F7B726.23232707042002@news.service.uci.edu>
Message-ID: <3CB18A4D.1080602@geneva-link.ch>


David Eppstein wrote:

> In article <3CB13147.89F02706 at cosc.canterbury.ac.nz>,
>  Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> 
>>The point of the example isn't to show how to find prime numbers 
>>quickly. It's just a representative CPU-intensive task that 
>>demonstrates what Pyrex can achieve. Using a more efficient algorithm 
>>wouldn't make that point any more effectively.
>>
> 
> But I think your example makes a different point even more effectively:


If such was true, the different point would not need explicitation.
Frankly I feel you are being overly pedantic. While the sieve is 
efficient, it's neither secret magic nor particularly readable as an 
expression of the definition of prime numbers. And I do feel Greg's 
efforts deserve more relevant feedback.

Regards, Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 11:19:01 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 08:19:01 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cbaefb5@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3lmbpoyai.fsf at mira.informatik.hu-berlin.de...

| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
|
| > 2) Secondly, when we are printing integers in the hex format,
| > we are not interested in the sign-- we are looking at the
| > bit pattern.
|
| Who is this "we" in this statement? When I print numbers in hex, I'm
| interested in the hexadecimal representation if the number. For
| *printing* it, I usually don't care what the internal representation
| is.

That would be most people that need to get inside and take a look at the bit
representation of a number-- when doing things like developing new random nubmer
generators, and encryption techniques.  There are other examples of course, but
you should get the idea.  When _I_ want to se a number in hex, it is because I
am interested in the bit pattern.  I'm sure many others feel the same way about
it.  I have only been programming for 30 years, (professionally for 25)-- so
maybe I don't have enough experience to give you the right answer to this.

| > 3) With unification, the long and the int types should not
| > print differently when calling the 'hex()' method.
|
| And indeed, they won't.
|
| > (The 'int' version of this method NEVER has, and shouldn't print a
| > minus sign, and _I_ don't think the long version of 'hex()' should
| > either.  This _is_ what needs to happen to have proper unification.
|
| The PEP currently says differently:
|
| # In most cases the long int semantics will prevail; however, the
| # trailing 'L' from long int representations will be dropped.
|
| This clearly says that the long int semantics will prevail.

Yes-- I read the PEP too--  when this happens, I will consider it a Python
"_WART_", and will set about inventing ways around it for my own use.  I have
forgotten more computer languages than most people will ever learn-- an I have
NEVER found printing a hexadecimal number in sign and magnitude format to be
useful...

I will say it again: when I am printing a number in hexadecimal, I am interested
in the _BIT_PATTERN_ -- _NOT_ the numeric _VALUE_.  I'm sure many others feel
this way as well.

| > 4) We are not using 1.5.2 anymore.  We are on 2.2.x now.
| > You need to imprint on your brain that there now should be
| > _NO_ visible difference (to the programmer) between the long
| > and the int type.  In fact, the int and long representations
| > are now just an implementation detail, and should be hidden,
| > and transparent to the programmer.
|
| I completely agree.
|
| > Martin, why do you resist the unification of these two
| > types for _ALL_ operations and methods?
|
| Why did you conclude that I resist? I'm just saying that the long
| semantics will prevail, not the int semantics.

Again-- when this change is made, I think it will be a Python "_WART_".
Numbers should only print the sign and magnitude when printing decimal numbers.
When printing hexadecimal numbers, the bit pattern is what is needed-- not the
sign and magnitude.

wart- Wart- WART- WWAAARRTT!!!!!





From jeff at ccvcorp.com  Thu Apr 25 15:03:50 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 25 Apr 2002 12:03:50 -0700
Subject: ActiveX events in wxPython not working
References: <6ddcff55.0204250625.1992ce1b@posting.google.com>
Message-ID: <MPG.1731f2f1a9e379e79896ca@news.nwlink.com>

In article <6ddcff55.0204250625.1992ce1b at posting.google.com>, 
perry at subdimension.com says...
> Hi,
> 
> I'm using wxPython as an ActiveX (GUI) container. The OCX draws a
> gantt
> chart from supplied data. Everything works fine, appart events.
[...]
> class MyGanttPanel(wxPanel):
>   def __init__(self, parent):
>      wxPanel.__init__(self, parent, -1)
> 
>      sizer = wxBoxSizer(wxVERTICAL)

If you only have one control, there's no need to use a sizer.  
wxPython will, by default, automatically resize a single control 
to fill its parent window (in this case the panel).  But this has 
nothing to do with your problem...  :)

>         
>      GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
>      gantt.VcGantt = GanttClass(self, -1, wxPoint(0,0),
> style=wxMAXIMIZE)
>      gantt.VcGantt.EnableSupplyTextEntryEvent = 1

Okay, you're setting your OCX as an attribute of this gannt 
object... but you haven't showed how you're creating or storing 
the gantt object itself.  If this is something that's local to 
the method, then it goes out of scope when __init__() finishes, 
and your OCX is destroyed as soon as it starts!  (Hmmm... 
actually, your sizer may be holding a reference for you, and thus 
preventing it from being garbage-collected, but this doesn't seem 
like a very safe thing to count on...)

Unfortunately I'm not familiar with the details of ActiveX 
containment, so I can't offer any more specific help.  You might 
do better to ask this question on the wxpython-users mailing list 
(http://lists.wxwindows.org/mailman/listinfo/wxpython-users).

-- 

Jeff Shannon
Technician/Programmer
Credit International


From davidccarson at hotmail.com  Thu Apr 18 23:07:55 2002
From: davidccarson at hotmail.com (David Carson)
Date: 18 Apr 2002 20:07:55 -0700
Subject: raw_input bug only in cygwin python
Message-ID: <12257ec4.0204181907.7a79b15b@posting.google.com>

Here is a snippet of code:

01: #!/usr/bin/env python
02: 
03: list = []
04: while 1:
05:     try:
06:         stuff = raw_input("[^D to exit]: ")
07:     except EOFError:
08:         print "got EOF"
09:         break
10:     list.append(stuff)
11: 
12: raw_input("bang! ")


This works on every Python environment I've tried but cygwin (Win2000
-- I haven't got NT to try it):
  Python 2.2 (#1, Dec 31 2001, 15:21:18)
  [GCC 2.95.3-5 (cygwin special)] on cygwin

In this environment, the EOFError is caught.  Then the "bang! " prompt
for the raw_input() outside the loop is printed, but immediately the
program exits with an uncaught EOFError.

This code works with:
- ActiveState Python, 2.1 on Win2000
- python 2.2 on Mandrake
- python 1.5.2 on RedHat
and everywhere else, I'm sure.

Is this a known problem with cygwin?

Cheers,
David


From bokr at oz.net  Mon Apr  8 04:54:30 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 08:54:30 GMT
Subject: Function question
References: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu>
Message-ID: <a8rls6$335$0@216.39.172.122>

On Sun, 7 Apr 2002 23:55:47 -0500, "John" <nobody at nobody.com> wrote:

>
>Hi I have a question:
>
>Let's say there is a function with two arguments.
>def add(a,b): return a+b
>
>What I want to do is:
>I want to get 'a' dynamically..
>But once a value is decided, I want a separate function (say, new_add) that
>does
>new_add(b) (equals to) add(a,b)
>
>That is, I want the function definition to be dynamic.
>
>Would that be possible? Any comment would be greatly appreciated.
>
If you want to store initialized state to be used by a function,
one way is to define a class whose constructor takes the initial
parameters and stores them (or state computed based on them, etc.)
in the data of the created instance. If you supply a __call__ method,
you can call it with syntax that looks just like a function call using
the instance as the function:

 >>> class AddK:
 ...     def __init__(self, k):
 ...         self.k = k
 ...     def __call__(self, v):
 ...         return self.k + v
 ...
 >>> add2 = AddK(2)
 >>> addten = AddK(10)
 >>> add2(5)
 7
 >>> add2(3)
 5
 >>> addten(5)
 15

I find this simple and clear, and I would expect it
to be faster than unoptimized generalized currying as
described in another post (though it is nice).
The function-returning function in another post might
be about the same speed. You'd have to test if speed is
a concern.

The nice thing about the object approach is that you have
an open door to further customization of your computing gizmo.
You can add methods to query it, modify it, have it present
itself as string, as true or false, etc. etc.

Regards,
Bengt Richter


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 19:14:21 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 16:14:21 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cb8bc26@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3sn5z1axf.fsf at mira.informatik.hu-berlin.de...
| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
|
| > Since Python does not have unsigned numbers, I think
Python should
| > have a (Java-like) '>>>' operator.  The '>>>' means
| > 'logical-shift-right', where a zero is copied into the
high bit.
|
| Since the Python int and long int types are going to be
merged, I
| doubt that this will be added to the language.

The merging of the 'long int' and the 'int' type is
precisely WHY there _should_ be a '>>>' operator.  Did you
read ALL of my post?

| One might question whether the bitwise operators should
have been
| there in the first place - functions can do the same
things just as
| well.

You _ARE_ joking about this, aren't you?

Let's say I want to work with a 64-bit number:

>>> a = 0x8000000000000000
>>> print hex(a >> 1)
0x4000000000000000L

# (I expected the above)...

>>> b = -a
>>> print hex(b >> 1)
-0x4000000000000000L

# !!! SURPRISE !!!
# (I expected -0xC000000000000000L !!!)

I think this should work the same way an 'int' does-- but I
don't know how to do that without breaking older code...

Very_bit_diddle'ly yours,
--Ken Peek





From neuhauser at mail.cz  Mon Apr 29 08:49:12 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Mon, 29 Apr 2002 14:49:12 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <Hd6z8.74979$vF6.2240539@news2.tin.it>
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it>
Message-ID: <20020429124912.GC325@roman.mobil.cz>

> From: Alex Martelli <aleax at aleax.it>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Mon, 29 Apr 2002 07:01:59 GMT

> Raymond Hettinger wrote:
> Take, for example, 'apply'.  What need does it serve, ever since the
> forms *args and **kwds started applying to actual arguments (i.e., at
> function-call time) as well as to formal arguments (i.e., at function
> definition time)?  I.e., since 2.0 (or 1.6, I don't recall).  Indeed
> that would have been my first candidate for built-in deprecation (also
> for alphabetical reasons:-).
> 
> abs, buffer, chr, cmp, coerce, id, intern, oct, ord, are a few others
> that might be worth thinking about in terms of frequency of need --
> they're all needed some of the time, of course, but how often?  Often
> enough to be warranted as built-ins?  I surely use, e.g., map, more
> often than, e.g., cmp, coerce, or id.  Not to mention intern.  So why
> should map go away and the others remain?  This is not meant to be a
> rhetorical question, but to prompt active reflection/discussion/study.

    __builtin__.chr(int) -> int.chr()
    __builtin__.abs(int) -> int.abs()
    __builtin__.ord(str) -> str.ord()
    etc

    There's quite a few places in python that break the POLA rule, and
    the seemingly randomly chosen builtins are a prominent example.

-- 
FreeBSD 4.4-STABLE
2:43PM up 5:09, 6 users, load averages: 0.01, 0.10, 0.21




From dominikush at yahoo.com  Wed Apr 17 10:28:44 2002
From: dominikush at yahoo.com (dominikush at yahoo.com)
Date: Wed, 17 Apr 2002 16:28:44 +0200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <3CBD869C.54BC7BF1@yahoo.com>

Dear Michael, dear Alex,

you are excellent teachers!!!

Michael, you helped me really getting the point with your
drawings. Thanks a lot for your art work!

Alex, the anecdote about the statue pointing to Hotel Belfiore
made my wrong intuition so obvious! I like that and will never
ever forget it anymore! Thanks for your answer!

I think, today I learnt a lot on my way becoming a real Pythoniac!

Best regards,

Dominikus


From grante at visi.com  Sun Apr 14 00:12:42 2002
From: grante at visi.com (Grant Edwards)
Date: Sun, 14 Apr 2002 04:12:42 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com>
Message-ID: <slrnabi0f4.15a.grante@tuxtop.visi.com>

In article <3cb8b86a at news.mhogaming.com>, Ken Peek wrote:

> (not a) ^ (not b)
> 
> IMHO, neither the above, nor the future v2.3.x idiom:
> 
> bool(a) ^ bool(b)
> 
> is as clear as:
> 
> a xor b

I agree.  Since the other boolean operators are "not" "and"
"or", it would be a bit of a surprise that one doesn't use
"xor". And having to convert a boolean to a "chunk of bits"
first and do a bitwise xor on it is just plain ugly.

OTOH, when I want to do exclusive or, it's pretty much always
the bitwise operation I want.

-- 
Grant Edwards                   grante             Yow!  If I had a Q-TIP, I
                                  at               could prevent th' collapse
                               visi.com            of NEGOTIATIONS!!


From imbosol at vt.edu  Sat Apr  6 20:41:37 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 6 Apr 2002 20:41:37 -0500
Subject: Pattern for permutations?
References: <mailman.1018131433.4336.python-list@python.org>
Message-ID: <h48o8a.qq1.ln@127.0.0.1>

Tim Churches wrote:
> This may be an elementary question but the solution is
> not obvious to me.

When you start doing stuff to lists in a general and arbitrarily deep
way, recursion is your friend.  Learn to think recursively first, then
iteratively second.


> I want to define a function which takes as its argument a
> list which contains one or more sequences, and then does stuff
> with every permutation of the elements of those sequences. For
> example, a function which works for a list containing three
> sequences:
> 
> def nestedloops(list_of_sequences):
>        for element0 in list_of_sequences[0]:
>                for element1 in list_of_sequences[1]:
>                        for element2 in list_of_sequences[2]:
>                                # do stuff with each permutation
>                                print element0, element1, element2
> 
> nestedloops([(1,2,3),('one','two','three'),('isa','dalawa','tatlo')])
> 
> How does one generalise such a function so that it handles any number 
> of sequences in the passed list of sequences i.e. a dynamic number of
> levels of nesting? 

Here's the basic idea:

Write a function that accepts a stub list, and a list of sequences
that you want to permute.  Suppose your function takes the following
arguments:

stub=[], seqs=[(1,2,3),('one','two','three')]

What it should do is recursively call itself once for each element of
seqs[0].  It should call itself with that element added to the end of
the stub list, and the first subsequence removed from seqs.  Thus,
your function should make these three calls:

stub=[1], seqs=[('one','two','three')]
stub=[2], seqs=[('one','two','three')]
stub=[3], seqs=[('one','two','three')]

When called with these arguments, the function, since it is the same
function, will do the same thing: call itself for each element of
seqs[0], with the element added to stub, and the seqs[0] removed.
When called, the function would have the arguments (for the first case
above):

stub=[1,'one'], seqs=[]
stub=[1,'two'], seqs=[]
stub=[1,'three'], seqs=[]

In this case, seqs is an empty list.  That is the signal that the stub
list contains the entire permutation.  So the function must test
whether seqs is empty, and if it is, do the desired operation.

This technique is a cliche of functional programming, and it is very
useful when you have arbitrarily nested stuff.


Here is the code:

    def permute_op (op, seqs):
        "Apply function op to each permutation of the sequences in seqs."
        _sub_permute_op (op, [], seqs)

    def _sub_permute_op (op, stub, seqs):
        if seqs:
            for p in seqs[0]:
                _sub_permute_op (op, stub+[p], seqs[1:])
        else:
            op (cseq)


You can do this iteratively, however.  The iterative method might be
easier on the computer but is harder on the programmer.

    import operator

    def permute_op (op, seqs):
        "Apply function op to each permutation of the sequences in seqs."

        nseqs = len(seqs)
        seqrange = range(0,nseqs)
        perm = [None]*nseqs
        lengths = map(len,seqs)
        nperms = reduce(operator.mul,lengths)

        for i in xrange(0,nperms):
            for j in seqrange:
                perm[j] = seqs[j][i%lengths[j]]
                i = i / lengths[j]
                #### Careful--the above won't work as of Python 3.0 ####
            op (perm)


> It occurred to me that the function could emit dynamic code for the 
> "for" loops and then eval() that, but that seems a bit messy and
> possibly slow.
> 
> Is there a better way?

Unless you're programming in LISP, there's always a better way than
self-modifying code.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From smithj at nsidc.org  Tue Apr  9 10:30:47 2002
From: smithj at nsidc.org (Jeff Smith)
Date: Tue, 9 Apr 2002 08:30:47 -0600
Subject: Newbie Problem Building 2.2 On SGI
Message-ID: <Pine.SGI.4.10.10204090829040.218574-100000@graupel.colorado.edu>

Apologies to everyone. The problem is solved. I won't embaress myself
further by detailing what I did wrong :).

Jeff Smith






From news.ASkwar at DigitalProjects.com  Mon Apr 22 15:43:59 2002
From: news.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Mon, 22 Apr 2002 21:43:59 +0200
Subject: Checking if class is derived of some other class
References: <pan.2002.04.22.19.11.59.254871.5745@teich.garten.DigitalProjects.com>
Message-ID: <pan.2002.04.22.19.43.59.567352.5745@teich.garten.DigitalProjects.com>

On Mon, 22 Apr 2002 21:11:59 +0200, Alexander Skwar wrote:

> How can I check if myObject is either directly instantiated from class a
> or is if of a class which inherited from class a?

Okay, figured it out myself:

if issubclass(myObject, a):
	print "myObject has some connection to a"
else:
	print "no connection between myObject and a"

However, reading
http://www.python.org/doc/current/lib/built-in-funcs.html#l2h-33 , 
I would have thought that the appropriate method would have been
isinstance().  But the following prints "No":

class a:
	pass

class b(a):
	pass

class c(b):
	pass

myObject = c

if isinstance(myObject, c):
	print "yes"
else:
	print "No"

Why is that?

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 12 hours 46 minutes


From carmstro at twistedmatrix.com  Fri Apr 19 14:38:14 2002
From: carmstro at twistedmatrix.com (Christopher Armstrong)
Date: Fri, 19 Apr 2002 13:38:14 -0500
Subject: Twisted 0.17.2 Released
Message-ID: <20020419183814.GA4412@twistedmatrix.com>

Terry Reedy wrote:
> After reading the 'license' page in Apr 2, I looked at every page I
> could find for something about the RPG engine but could not find much
> at all.  Is there a page I missed or must I download and look for doc
> or source which will tell me whether or not I wanted to download?
> 
> Terry

Twisted Reality was recently broken out of the main distribution, and
probably won't have an official release till it's been hacked on for a bit.
If you want, you can get it from anonymous CVS, CVSROOT being
:pserver:anon at twistedmatrix.com:/cvs, module "Reality".

Reality's been in a state of disrepair for a while, because all of the game
projects that were using it stagnated. Some of the developers are getting
re-interested in it again, though, so we can pray that it'll get some
devel-time =)

--
Christopher Armstrong
<carmstro at twistedmatrix.com>
http://twistedmatrix.com/users/carmstro.twistd/




From robin at jessikat.fsnet.co.uk  Tue Apr 16 14:17:46 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 16 Apr 2002 19:17:46 +0100
Subject: Python 2.2 unwise.exe
References: <mailman.1018727067.4980.python-list@python.org>
 <upu104ym0.fsf@ctwd0143.fitlinxx.com> <GSYATKAqUGv8EwRX@jessikat.demon.co.uk>
Message-ID: <5SgPTIAKrGv8Ew3Q@jessikat.demon.co.uk>

In article <GSYATKAqUGv8EwRX at jessikat.demon.co.uk>, Robin Becker <robin at jessikat.fsnet.co.uk>
writes
>On the office machine well I installed Python 2.2.1 last Friday evening.
>My unwise.exe is now gone as of Tuesday 10:00 am.
>
>I had a hard crash Monday morning.
One further thing is that I notice that UNWISE.EXE is listed in the registry 

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\AppCompatibility\UNWISE.EXE]
"x"=hex:0c,00,00,00,00,00,00,00,06,00,00,00,00,00,00,00
"DllPatch-x"=""


I have no idea what this means, but perhaps it might
have something to do with the problem.
-- 
Robin Becker


From mwh at python.net  Tue Apr 23 07:04:59 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 23 Apr 2002 11:04:59 GMT
Subject: Interesting topics
References: <mailman.1019468028.28617.python-list@python.org> <Xns91F85897F9A6DRASXnewsDFE1@130.133.1.4> <20020422165134.GD25778@andy.tynant.ftech.net> <mailman.1019502847.26728.python-list@python.org>
Message-ID: <lkwuuypt8e.fsf@pc150.maths.bris.ac.uk>

Andy Gimblett <gimbo at ftech.net> writes:

> On Mon, Apr 22, 2002 at 05:51:35PM +0100, Andy Gimblett wrote:
> 
> > ... in mutt I find Ctrl-D (delete entire thread) and Meta-D (delete
> > subthread) _indispensable_ around here.  :-)
> 
> And they're getting lots of usage on the maths thread right now...

(setq gnus-use-adaptive-scoring t) <wink>

Cheers,
M.

-- 
  Screaming 14-year-old boys attempting to prove to each other that
  they are more 3133t than j00.
         -- Reason #8 for quitting slashdot today, from
            http://www.cs.washington.edu/homes/klee/misc/slashdot.html


From heiland at ncsa.uiuc.edu  Mon Apr  8 14:31:38 2002
From: heiland at ncsa.uiuc.edu (Randy Heiland)
Date: Mon, 08 Apr 2002 13:31:38 -0500
Subject: html parsing
Message-ID: <3CB1E20A.B3300AA4@ncsa.uiuc.edu>

I'm a newbie at parsing html in Python... could someone offer a snippet
of code to:
1) specify a URL
2) parse the html and return the hyperlink associated with a unique
anchor string

thanks (and could you cc me directly),
--Randy



From mwh at python.net  Fri Apr 12 07:38:03 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 11:38:03 GMT
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk> <mailman.1018608627.2395.python-list@python.org>
Message-ID: <lkvgaxhyuu.fsf@pc150.maths.bris.ac.uk>

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

> On Fri, Apr 12, 2002 at 10:33:40AM +0000, Michael Hudson wrote:
> > > case $MAKEFLAGS in \
> > > *-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py -q build;; \
> > > *) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py build;; \
> > > esac
> > > running build
> > > running build_ext
> > > building 'struct' extension
> > > Segmentation Fault - core dumped
> > 
> > Yikes!  Can you dig into this?  I can't, not having access to a
> > relavent machine... (you did run make clean, didn't you?)
> 
>    I'll try. What is worse, Python 2.2 failed the same way :( It seems
> upgrading gcc from 2.8.0 to 2.95.3 did more harm than good :(((

Oh dear.  I apologise for encouraging you to break you system...

-- 
6. The code definitely is not portable - it will produce incorrect 
   results if run from the surface of Mars.
               -- James Bonfield, http://www.ioccc.org/2000/rince.hint


From news.ASkwar at DigitalProjects.com  Mon Apr 22 17:44:17 2002
From: news.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Mon, 22 Apr 2002 23:44:17 +0200
Subject: Checking if class is derived of some other class
References: <pan.2002.04.22.19.11.59.254871.5745@teich.garten.DigitalProjects.com> <pan.2002.04.22.19.43.59.567352.5745@teich.garten.DigitalProjects.com> <aa1qjg$g16$1@panix1.panix.com>
Message-ID: <pan.2002.04.22.21.44.16.998218.5745@teich.garten.DigitalProjects.com>

On Mon, 22 Apr 2002 22:08:16 +0200, Aahz wrote:

> Try
> 
> myObject = c()

Aargl.  Yes, it surely helps if it's done right.  Thanks!
 
Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 14 hours 57 minutes


From jriveramerla at yahoo.com  Mon Apr  8 22:52:59 2002
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 8 Apr 2002 19:52:59 -0700
Subject: Compatibilty table between BOA, wxPython, Python?
References: <11e94203.0204062147.5d2281b9@posting.google.com> <3CB0245A.3010808@csi.com> <11e94203.0204071640.103c4bb7@posting.google.com> <3CB20573.6060006@csi.com>
Message-ID: <11e94203.0204081852.2796226d@posting.google.com>

Thank you... that was the solution... to delete the intializers of POS
and SIZE to the StatusBar.

Thanks to guide me to that list of BOA, I did'nt know it existed.

Thanks.


From gmcm at hypernet.com  Tue Apr  2 08:47:54 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 02 Apr 2002 13:47:54 GMT
Subject: The Python Way
References: <5515F24370BCB753.DA7CF91464346EF6.E3A577CFA6882D82@lp.airnews.net> <j2m1au4se21mc222877l06690vei117jc3@4ax.com> <a7qqhr$m5$1@panix1.panix.com> <jjb5-B81262.13422827032002@newsstand.cit.cornell.edu> <15522.13124.505385.871392@beluga.mojam.com> <mailman.1017748504.20611.python-list@python.org>
Message-ID: <Xns91E459912A1E5gmcmhypernetcom@199.171.54.214>

Andy Gimblett wrote:

[snip]
 
> So, is Aahz/Tim saying that Queue.Queue is "better" in general than
> threading.RLock ?  If so, why?  

It's not that Queue.Queue is "better" than RLock (Queue was written
before threading.py, but if it had been written after, I would
expect it to use a Condition which uses an RLock.)

What's "better" is keeping the number of "shared" (really, "contended")
resources to a minimum. If your threads only meet through Queues,
you have virtually no chance of ending up deadlocked.

It's really a design philosophy: each resource should be owned
by one and only one thread.

-- Gordon
http://www.mcmillan-inc.com/


From mkc+dated+1021607974.93924d at mathdogs.com  Wed Apr 17 00:03:04 2002
From: mkc+dated+1021607974.93924d at mathdogs.com (Mike Coleman)
Date: 16 Apr 2002 23:03:04 -0500
Subject: Why don't strings share data in Python?
References: <mailman.1018937610.11880.python-list@python.org>
Message-ID: <87n0w37z0n.fsf@mathdogs.com>

Tim Peters <tim.one at comcast.net> writes:
> There are many reasons, but one of them is invisible to Python users:
> because Python is intended to play nice with C libraries, all Python strings
> end with a \0 byte.

Well, that's certainly a bug in the ointment.

I almost thought I might be able to use NumPy arrays, as it looks like they do
share data.  They're not hashable, though, unfortunately, which kills the
usage I had in mind.

Maybe I'll just buy more RAM...  ;-)

Mike


From aleax at aleax.it  Tue Apr 16 10:16:18 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 14:16:18 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no> <s4zu8.20814$vF6.690664@news2.tin.it> <slrnabobad.8ho.mlh@vier.idi.ntnu.no>
Message-ID: <SmWu8.26250$vF6.834282@news2.tin.it>

Magnus Lie Hetland wrote:
        ...
>>I think the best name for what's now raw_input might be something
>>like getline.
> 
> Or perhaps readline (? la file.readline)?

I slightly prefer getline, because raw_input differs from a
somefile.readline -- the latter does not strip off the '\n',
while raw_input does -- each of the two slightly different
behaviors being optimal in its context.  I think it would be
harder to teach if the names coincided.


Alex



From siegfried.gonzi at kfunigraz.ac.at  Wed Apr 10 03:59:58 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 10 Apr 2002 09:59:58 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk>
Message-ID: <3CB3F0FE.B5382A7C@kfunigraz.ac.at>

phil hunt wrote:

> I disagree. K&R has a mixture of clarity and concision that makes
> it a delight to read.
> 
> IMO, anyone for whom K&R isn't to their taste, probably won't make a
> very good programmer anyway.

The last thing which I am  bothered with: whether to become a good
programmer or not. For example, I did never use a struct in C; that
means my C code is more or less a big spaghetti one. I do not care as
long as I get the correct result.

And I have really, really not the intention to waste my life as a
professional programmer (there I think one should adapt to a better
programming style). Though programming is sometimes interesting and
indispensible.


S. Gonzi


From maxm at mxm.dk  Fri Apr 12 07:02:48 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 12 Apr 2002 13:02:48 +0200
Subject: Looking for doh!.. simple sequence type
Message-ID: <3CB6BED8.2000502@mxm.dk>

Hmm ... I am trying to create a simple sequence type that I cant iterate 
over like (Python 2.1)::

lst = Lst()
for i in lst:
     print i

And I thought I could do it like:

class Lst:

     def __len__(self):
         return 4

     def __getitem__(self, idx):
         return 42

But I have to write it like below to avoid an endless loop..

lst = Lst()
for i in range(len(lst)):
     print i,

 >>>42 42 42 42

I am shure it's so simple tha I will go Doh! .. But perhaps I haven't 
had enough cofee yet. 'Cause I cannot seem to figure it out.

regards Max M



From Lutz.Schroeer at kybernetik-manufaktur.de  Fri Apr 26 14:48:48 2002
From: Lutz.Schroeer at kybernetik-manufaktur.de (Lutz Schroeer)
Date: 26 Apr 2002 18:48:48 GMT
Subject: Simple bar graphs
Message-ID: <Xns91FCD43535458Latzikatz@139.174.2.56>

I'm looking for a _pure_ Python module which generates some very simple bar 
graphs and outputs them as gif or jpg. I looked through Parnassus and other 
sources but could not get enough information to make a decision.

Thanks
Lutz


From hfoffani at yahoo.com  Sat Apr 20 08:55:30 2002
From: hfoffani at yahoo.com (Hernan Foffani)
Date: Sat, 20 Apr 2002 14:55:30 +0200
Subject: Python 2.2 unwise.exe
References: <mailman.1019259099.12047.python-list@python.org>
Message-ID: <3CC16542.70705@yahoo.com>

Tim Peters wrote:
 > Wise agreed, and quickly built a new version of UNWISE.EXE
 > that doesn't have this problem.  This leaves one other
 > small problem:  how do we manage to magically replace a
 > hundred thousand or so vulnerable UNWISE.EXE's already
 > installed in Python directories on Windows boxes?
 > It's times like this I'm glad Python installs secret
 > backdoors on Windows boxes so Guido can control them
 > from the comfort of home.

I was going to ask if you can upload it to python.org, but
since you're going to patch the 100K PCs all by yourself...

 > it's-typing-in-all-your-names-that's-the-bitch-ly y'rs  - tim
one less to type: Hernan Foffani / IP: 217.126.144.119
;-)




From stephen at xemacs.org  Fri Apr 12 03:14:04 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 12 Apr 2002 16:14:04 +0900
Subject: emacs python mode completion
References: <mailman.1018589425.10772.python-list@python.org>
Message-ID: <871ydl75j7.fsf@tleepslib.sk.tsukuba.ac.jp>

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

    Mark> Is there a tool for emacs that does Python command
    Mark> completion?

Although it's not exactly what you what, something that works in all
modes is dabbrev (bound to M-/ in XEmacs by default).  This completes
words according to context.  So if I type M c M-/ here I get "McEahern."

This means the you can automatically complete long variable names and
the like, as well as reserved words.  There is also a static (ie,
explicitly defined) abbreviation mode.  I don't know of anything that
specifically understands Python syntax, though.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From jeff at ccvcorp.com  Tue Apr 23 14:04:21 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 23 Apr 2002 11:04:21 -0700
Subject: Classes: Old or New and How?
References: <Xns91F8BDF3CF983RASXnewsDFE1@130.133.1.4> <aa2sh6$2o7v$1@ulysses.noc.ntua.gr>
Message-ID: <MPG.172f41f3c3983209896c4@news.nwlink.com>

In article <aa2sh6$2o7v$1 at ulysses.noc.ntua.gr>, 
rs96057 at hotmail.com says...
> 
> "Philip Swartzleonard" <starx at pacbell.net> wrote in message
> news:Xns91F8BDF3CF983RASXnewsDFE1 at 130.133.1.4...
> > [...] When i start a class like
> > this, should I bother explicetly deriving from Object, or should I set
> > that flag that makes everything like that, or should I just not bother?
>
> I think that, new style is some kind of an extension of the old one and that
> in later versions of Python the default style will change to the new (maybe
> a unified one).

Yes, the intent of new-style classes is that they will be the 
default in a future version of Python.  The reason behind the 
change is so that classes can be unified with types.  The 
question at hand is, whether during this transitional period, it 
makes sense to make an effort to use the new style, or to simply 
keep using the old (and still default) style.

I would say that it depends upon two things -- if the subtle 
differences in behavior work for or against your intent for the 
class, then it makes the most sense to use whichever style is 
most suited.  Of course, the vast majority of trivial classes, it 
won't make any difference.

In those cases, I'd look at the probable lifetime of the code.  
If you think you might still be using it when the next version of 
Python comes out, you're probably best to use new-style classes 
wherever practical.  That way, when new-style becomes the 
default, you won't get any nasty surprises, because you'll have 
been coding to that for some time, and any effect that the change 
would have, will already have been accounted for.  If you're 
writing throwaway code, though, then it doesn't matter, and you 
might as well use the default old-style classes.

Of course, it might not be a bad idea to try to use new-style for 
everything (except where it breaks code), so that you'll be 
thoroughly used to the semantic differences when it *does* become 
the default...

(Spoken as someone who's still using Python 2.1 ...)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From cash4u at email.com  Mon Apr 15 08:08:29 2002
From: cash4u at email.com (cash4u at email.com)
Date: Mon, 15 Apr 2002 20:08:29 +0800
Subject: Double your traffic + Earn 20% of your referrals!!
Message-ID: <E16x5E4-0005Ia-00@mail.python.org>

Dear Webmaster:

We have 2 stirring programs:

1 Free Pop-behind Exchange---DOUBLE your traffic without any effort.

2 Earn 20% traffic of your referrals----Get millions of visitors just by a textlink.

Join Now!! 

Please visit http://www.2traffic.net









---------------------------------------------------------------
   ??????????????????????SMTP????????????????????????????????

   ??????????????????PC????????SMTP??????????????????????????

   ??????????????????????????????????????????????????????????

   ??????????????????????????????????????????????????????????

   ????????????????      http://www.seekeasysoft.net
---------------------------------------------------------------




From jeff at ccvcorp.com  Fri Apr 19 18:22:12 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 15:22:12 -0700
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com>
Message-ID: <MPG.172a385937afc0db9896bc@news.nwlink.com>

In article <pan.2002.04.19.16.48.06.120768.1546 at jonesj.com>, 
jeff at jonesj.com says...

> I use html templates for all of my CGI programming, and everything that needs
> to be replaced is contained within double curly braces ( {{value}} ).  In
> perl I use a hash to keep all of the variables that are going to replace
> the placeholders in the template....

Well, I'm not familiar with Perl, but perhaps this will help you?

>>> mydict = { 'name': 'Jeff', 'food': 'Spam' }
>>> template = 'Hi, my name is %(name)s, and my favorite food is 
%(food)s!'
>>> print template % mydict
Hi, my name is Jeff, and my favorite food is Spam!
>>> 

Your template contains '%(var)s' as placeholders.  When you 
provide the string formatting operator (%) with a dictionary, it 
will then go through your template and substitute every '%(var)s' 
with the value of mydict[var].  

-- 

Jeff Shannon
Technician/Programmer
Credit International


From rjones at ekit-inc.com  Mon Apr  1 17:54:09 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 2 Apr 2002 08:54:09 +1000
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <20020401195130.B13742@phd.pp.ru>
References: <A7B5C22F20EE3C4E95744E2F03F1887F5612@MAYOR.retzlaff.com> <20020401195130.B13742@phd.pp.ru>
Message-ID: <200204012254.WAA29114@crown.off.ekorp.com>

On Tue, 2 Apr 2002 01:51, Oleg Broytmann wrote:
> On Sun, Mar 31, 2002 at 11:28:29PM -0800, Jimmy Retzlaff wrote:
> > > Richard Jones [mailto:rjones at ekit-inc.com] wrote
> > > Does anyone have a copy of kjbuckets.c that they could send me?
> >
> > Oleg Broytmann has it on a page mirrored in a couple of places. Look at
> > the bottom of either:
> >
> > http://phd.by.ru/Software/Python/
> > or
> > http://phd2.chat.ru/Software/Python/
>
>    I am not a maintainer, just a holder. I rely on the help of people who
> update the source and provide binaries.


Thanks to all (especially Oleg), I've now got the module code.


    Richard



From tejarex at yahoo.com  Wed Apr 24 11:47:13 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 24 Apr 2002 15:47:13 GMT
Subject: Subclass of int and initialisation with not integer values
References: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr>
Message-ID: <5sAx8.162510$GS6.15531047@bin3.nnrp.aus1.giganews.com>

"Boris Boutillier" <boris.boutillier at lip6.fr> wrote in message
news:pan.2002.04.24.17.09.52.840883.17049 at lip6.fr...
> I've got a little problem with subclass of int.
...
> In some way it seems that the int initialisation is called before
__init__

You can verify this, to be sure, by sprinkling toto.__init__() with
print statements.

TJR





From donn at u.washington.edu  Thu Apr 11 15:39:49 2002
From: donn at u.washington.edu (Donn Cave)
Date: 11 Apr 2002 19:39:49 GMT
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
References: <mailman.1018549044.28363.python-list@python.org>
Message-ID: <a94oq5$m06$1@nntp6.u.washington.edu>

Quoth "Noah" <noah at noah.org>:
| Damn... The only problem I see with this explaination is
| that this would imply that popen is almost totally useless.
| Why bother opening a pipe with both in and out streams
| if can't actually read and write from both streams?
| I could have sworn that I have seen interactive programs
| such as telnet and ftp scripted from Python.

Pipes are fine for what they're normally used for, but for
circular structures, why bother indeed, as they will fail
in most cases.

| I looked at the pty stuff, but I was going to implement that
| AFTER I got pipes working. Pipes seem to be the most portable
| and intuitive way to solve this problem.
|
| Damnit! If I can do it in Perl I'm sure I can make it work in Python!

If you can explain how you do it in Perl, maybe someone can
tell you what you need in Python.  The buffering problem I was
talking about really has nothing to Python, it would be in ftp.
If ftp won't write its buffered output, then Perl, Python and C
are all in the same boat.

	Donn Cave, donn at u.washington.edu


From tejarex at yahoo.com  Tue Apr  9 00:16:36 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 09 Apr 2002 04:16:36 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8sf0k$2a5$0@216.39.172.122> <ahjs8.73608$S52.2597704@news2.tin.it> <a8sn8u$sj7$0@216.39.172.122> <Eeos8.74753$S52.2651788@news2.tin.it>
Message-ID: <EWts8.59900$%i.5713939@bin5.nnrp.aus1.giganews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:Eeos8.74753$S52.2651788 at news2.tin.it...
> Bengt Richter wrote:
>         ...
> > Well, I guess that depends on how you define "assignment" in
Python.
>
> I don't have to, as the Language Reference sees to that:
> http://www.python.org/dev/doc/devel/ref/assignment.html

This defines an 'assignment statement', which uses an 'assignment
operator'.  'Assignment can be read as a more generic concept whose
definition we do not necessarily agree on.

> It's a serious logical fallacy to reason "assignment statements may
> (among other things) rebind names, therefore what rebinds a name
> is an assignment".  Many bindings occur other than by assignment.
> Would you call a for statement, a def statement, a class statement,
> or an import statement "assignments", because they also rebind
> names?!

In the broad sense, yes.  Realizing that import, def, and class are
syntactic sugar for an underlying assignment process that actually or
potentially could have been written as an explicit assignment
statement was a milestone in my understanding of Python.  All three
are run time functions rather than compile time declarations (which
def and class initially looked like to me).  To wit:

import a #and
a = __import__('a')

do the same thing. Writing 'a' just once, without being quoted, is
handier and allows the 'from' and 'as' extensions.  There is a builtin
class constructor function (I forget the details).  And

def fname(p1, ..., pk): suite # is much nicer than the hypothecal
equivalent
fname = makefunc('fname', ('p1', ..., 'pk'), '''suite''')

For is more complicated since it combines repeated assignment and
looping:

for i in iterable: suite # is now equivalent to something like this:

__seqgen = iter(iterable)
try:
    while 1:
        i = __seqgen.next() # break loop on exception
        suite
except StopIteration: pass

Terry J. Reedy






From bokr at oz.net  Tue Apr 23 16:24:36 2002
From: bokr at oz.net (Bengt Richter)
Date: 23 Apr 2002 20:24:36 GMT
Subject: flatten tuple
References: <m2adrvnxiv.fsf@mother.paradise.lost>
Message-ID: <aa4fu4$5lt$0@216.39.172.122>

On Mon, 22 Apr 2002 18:06:48 -0500, John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote:

>
>I have a tuple of strings and a map from strings to either a string or
>a tuple of strings
>
>tuple of strings:
> 
>  x = ( 'John', 'Hunter', 'where' )
>
>
>map:
>
>  m = {
>       'where': ('was', 'here'),
>       'somekey': 'someval'
>      }
>
>
>I want to be able to create a new tuple, which is the expanded and
>flattened version x, so that if an element of x is a key of m, then
>the new list has that value instead of the key value .  In the example
>above, the expansion would give
>
>y = ( 'John', 'Hunter',  ('was', 'here') )
>
>and the flattening would give
>
>y = ( 'John', 'Hunter',  'was', 'here' )
>
>The only hitch is that I want to be able to do this recursively, so
>I could have for example
>
>m = {
>    'where': ('was', 'somekey'),
>    'somekey': 'someval',
>    'someval': ('asking', 'really', 'silly', 'questions')
>    }
>
Depending on your space/time tradeoffs, and when/how you build m
(i.e., is it built once and then used unchanged?) seems like you
could keep m flat, or flatten it once (assuming no cycles). E.g.,

m = {
    'where': ('was', 'asking', 'really', 'silly', 'questions'),
    'somekey': ('asking', 'really', 'silly', 'questions'),
    'someval': ('asking', 'really', 'silly', 'questions')
    }

(Where somekey and someval might even be eliminated if they can't
appear in x tuples)

then
    x = ( 'John', 'Hunter', 'where' )
could become
    y = ( 'John', 'Hunter', 'was', 'asking', 'really', 'silly', 'questions')
in one non-recursive lookup for each word, e.g.,

 >>> def sub(t):
 ...     result=[]
 ...     for w in t:
 ...         result.extend(m.get(w,[w]))
 ...     return tuple(result)
 ...
 >>> x
 ('John', 'Hunter', 'where')
 >>> y = sub(x)
 >>> y
 ('John', 'Hunter', 'was', 'asking', 'really', 'silly', 'questions')

Note that all the m[k] values have to be tuples and x[i] have to be strings
for this example.


>and end up with 
>
>
>y = ( 'John', 'Hunter',  'was',  'asking', 'really', 'silly', 'questions')
>
>I see that I expose myself to the dangers of infinite recursion which
>I should need to guard against, but I am willing to put the burden on
>the user not to define recursive tuple/map combinations if checking
>against recursion is too difficult.
>
>I would like to use tuples rather than lists because I would like
>these tuples to be keys in a hash.
>
It's hard to suggest useful stuff without the big picture.

Regards,
Bengt Richter


From aahz at pythoncraft.com  Thu Apr 25 15:59:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Apr 2002 15:59:45 -0400
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org>
Message-ID: <aa9n7h$4m9$1@panix1.panix.com>

In article <mailman.1019758546.27031.python-list at python.org>,
James J. Besemer <jb at cascade-sys.com> wrote:
>Christian Tismer wrote:
>>
>> There is some nice introduction to Limbo at
>> http://www.vitanuova.com/inferno/papers/descent.html
>> written by Brian W. Kernighan, one of the inventors
>> of the C language.
>
>Just a nit:
>
>Dennis Ritchie invented C.  http://cm.bell-labs.com/cm/cs/who/dmr/chist.html
>
>Brian helped invent the AWK language, a precursor to Python.  AWK stands for
>Aho Weinberger and Kernighan.  Brian wrote a lot of books about C but he was
>not one of the inventors.

Ewww!  Where did you get the idea that awk is a precursor of Python?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From r.demaria at tiscalinet.it  Thu Apr 11 13:58:18 2002
From: r.demaria at tiscalinet.it (Riccardo de Maria)
Date: 11 Apr 2002 10:58:18 -0700
Subject: Persistence
Message-ID: <90cc01ee.0204110958.6945c776@posting.google.com>

I hope it is not a faq.
Is there a way to save in a file the status, objects, functions of
interpreter
in order to restore it after the interpreter has been closed?

If it is possible, one can work on a project stop and then continue
later adding functions, objects and so on.

Thanks
Riccardo de Maria


From barry at zope.com  Wed Apr  3 11:35:38 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Wed, 3 Apr 2002 11:35:38 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
	<yu99wuvovkmm.fsf@europa.research.att.com>
Message-ID: <15531.12122.601076.455659@anthem.wooz.org>

>>>>> "AK" == Andrew Koenig <ark at research.att.com> writes:

    AK> Specifically: If I want to write a function that answers a
    AK> yes/no question, I have lots of possible ways of spelling yes
    AK> (1, 2, "yes", and so on) and lots of possible ways of spelling
    AK> no (0, {}, None, and so on).  There isn't a single preferred
    AK> way.

This is a strong pro (IMO) for PEP 285.  When I code a flag, do I use
0 and 1, or None and 1, or ...what?  When you see code that says
"x.isthing = 1" what does that tell you about how this attribute will
be used?  I like the bool because it provides clarity when coding flag
values, and you still maintain the (IMHO beautiful) feature of being
able to say "this variable has a value, or it's never been set".
I.e. flag=None, flag=True, flag=False.

The strongest con for PEP 285 seems to be the natural inclination for
people to want to write "if cond == True:".  Heck, /I/ certainly wrote
that the first time I thought about bools years ago (and even made
that "work" in my early pure-Python prototypes).  But it doesn't make
sense since "if cond:" and "if not cond:" are better than comparison
against the boolean constants.

There's two reasons we teased out about why people may tend to
want to compare against True and False.  People may think that the
comparison is somehow part of the syntax of the if statement.  Also,
using the comparison can sometimes be more aligned with English
grammar.  E.g. this reads well on its own:

    if year.isLeapYear():

Note the use of `is' which tends to be a common prefix on predicates.
But if you don't use that prefix, then you are maybe inclined to add
the test:

    if flag == True:

so that it reads somewhat like "if flag is True".  (Why don't we tend
to write exactly that?  Maybe because as Python programmers, we're
trained not to use identity tests except in very limited situations?).

-Barry




From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 03:20:16 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 09:20:16 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE> <mailman.1018991313.16792.python-list@python.org> <a9i4ms$ece$1@nets3.rz.RWTH-Aachen.DE> <mailman.1018995393.20362.python-list@python.org>
Message-ID: <a9j7o6$4tg$2@nets3.rz.RWTH-Aachen.DE>

"Geoff Gerrietts" <geoff at gerrietts.net> wrote in message
news:mailman.1018995393.20362.python-list at python.org...
>..
> I'd give some serious consideration to Andrew Dalke's solution, too.
> That may be a quicker route to exactly where my complicated route gets
> you -- I think that when I wrote this, I'd tried eval(string, {}, {})
> and never even considered replacing the __builtins__ namespace with an
> empty dictionary.
>
> I'm considering refactoring, myself....
>..

OK. In the end I will probably try to play around with the different
approaches, and your code, until I understand what's happening (that's
always good). Then I might be able to rebuild a small solution from scratch.




From tim at vegeta.ath.cx  Fri Apr  5 01:13:58 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Fri, 05 Apr 2002 06:13:58 GMT
Subject: I'm glad (was: I like python)
References: <mailman.1017984852.25463.python-list@python.org>
Message-ID: <slrnaaqgkc.h4d.tim@vegeta.ath.cx>

> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> _________________________________________________________
> Do You Yahoo!? 
> ??????????????????????????????????????????????
> http://cn.worldcup.yahoo.com/
> 
> 

Tim Hammerquist
-- 
...the word "hack" is used as a verb to indicate
a massive amount of nerd-like effort.
    -- Harley Hahn, "A Student's Guide to Unix"


From phr-n2002a at nightsong.com  Thu Apr 18 06:02:20 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 18 Apr 2002 03:02:20 -0700
Subject: Pure python crypto
References: <mailman.1019024313.11744.python-list@python.org> <7x3cxubvgb.fsf@ruckus.brouhaha.com> <mailman.1019053415.9589.python-list@python.org> <7x662qqb3n.fsf@ruckus.brouhaha.com> <mailman.1019100335.10929.python-list@python.org>
Message-ID: <7xofghe34j.fsf@ruckus.brouhaha.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
>   Well, I'll have to post it when I return home to Canada. Not a good
> idea to distribute it in the US :) I can send you a copy if you're in
> the US if you email me directly.

Might as well.  My address is above.  What's the encryption speed?

> > FWIW, I coded RC4 in Python a while back and if I remember correctly,
> > got about 100k/sec throughput on a P3-750.  That's sort of marginal
> > for (say) sftp over ADSL.  Using the SHA-OFB hack that I described,
> > I got about 400-500k/sec.  
> 
>  Do you still have a copy of this? I would be nice to assemble a suite
> that can be used to create a pure python implementation of SSH (which
> could also use wrappers to C implementations that exist as well).  I
> wish I had a better idea of what was out there module-wise though.

I'll take a look, I probably still have it.  I only spent a few
minutes coding it anyway, so can do it again if needed.  If coding RC4
is an obstacle to you though, I don't think you're going to finish
implementing SSH ;-).  Anyway, after coding RC4, I saw how much slower
it was than SHA-OFB, so I stopped caring about it and may have trashed
the implementation.

With luck, future versions of Python should include a DES/3DES module,
so an SSH package can use that.


From fperez528 at yahoo.com  Mon Apr 29 19:29:50 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 29 Apr 2002 17:29:50 -0600
Subject: ANN: IPython 0.2.10
Message-ID: <aakl63$8d$1@peabody.colorado.edu>

Announcing an update to IPython, an enhanced interactive Python shell. 

* WHERE: http://www-hep.colorado.edu/~fperez/ipython


* NEW since last public release (ChangeLog included with full details):

- Improvements and fixes to the embedding facilities.

- Integration with pdb, the python debugger. You can now test code (full 
programs) within IPython, have pdb automatically called at an uncaught 
exception, fix the program and re-run it without leaving IPython.

- Use of the readline library is now configurable by the user. Sensible 
defaults are provided,  but anything which readline.parse_and_bind() can 
process can be set by the user in the config file.

- New 'tutorial' profile which allows direct pasting of code with '>>>' and 
'...' in it at the IPython command line (such as commonly found in the Python 
tutorial or on this newsgroup). 

- Improvements to the Gnuplot access system. Now you can open a gnuplot shell 
inside IPython, and go back and forth between the two. Something similar 
could be implemented for any plotting system which supports two-way 
communications.

- Screenshots added to webpage at: 
http://www-hep.colorado.edu/~fperez/ipython/screenshots/index.html

- Many bugfixes, minor improvements and documentation updates.


* NOTE to users of previous versions: after installing the new version, please 
run IPython as 'ipython -up' the first time. This will update all the files 
in the configuration directory so that all new options work correctly. It 
makes a backup of any overwritten files.


* WHAT is IPython? (this is just for context. See the webpage for full 
details). IPython tries to:

1. Provide an interactive shell superior to Python's default. IPython has many 
features for object introspection, system shell access, and its own special 
command system for adding functionality when working interactively. It tries 
to be a very efficient environment both for Python code development and for 
exploration of problems using Python objects (in situations like data 
analysis).

2. Serve as an embeddable, ready to use interpreter for your own programs. 
IPython can be started with a single call from inside another program, 
providing access to the current namespace. This can be very useful both for 
debugging purposes and for situations where a blend of batch-processing and 
interactive exploration are needed.

3. Offer a flexible framework which can be used as the base environment for 
other systems with Python as the underlying language. Specifically scientific 
environments like Mathematica, IDL and Mathcad inspired its design, but 
similar ideas can be useful in many fields.


* Portability: Linux (and other unices, including Mac OSX), Windows XP/2k, 
Win9x (tested on 98, don't really know about 95)


* Python version: requires 2.1 or newer.


* License: LGPL (a few files from third parties carry MIT licenses).


Cheers,

Fernando Perez.



From starx at pacbell.net  Fri Apr 19 18:04:04 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 19 Apr 2002 22:04:04 GMT
Subject: The Python Manual Index
References: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de> <uc0d5d68i0vhf1@corp.supernews.com>
Message-ID: <Xns91F5997F18E9ARASXnewsDFE1@130.133.1.4>

Thomas Heller || --- 
>> Hi, folks. 
>>
>> A new Python-powered search engine is online at 
>> http://starship.python.net/crew/theller/pyhelp.cgi 
>>
>> It searches for keywords in the Python manuals at 
>> http://www.python.org/doc/. 
>>
>> The script can also be downloaded, and can be configured to search 
>> HTML manual pages installed locally. The keyword index is build on
>> the fly. It can also (probably) be easily integrated in your
>> editor/ide to provide context sensitive help. 
>>
>> Mark Hammond has integrated an interface to this keyword search in 
>> his cool Mozilla/Netscape sidebar he recently announced: 
>> http://starship.python.net/crew/mhammond/mozilla 
>>
>> Regards, 
>>
>> Thomas 
>>
>>
DeepBleu || Fri 19 Apr 2002 08:16:23a: 

> WOW!
> Great Job!

Indeed. Something like this needs to be added to the python.org 
interface... i've been wondering for a long time why they don't have a 
document search function =).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com 


From bkelley at wi.mit.edu  Tue Apr  2 15:29:09 2002
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Tue, 02 Apr 2002 15:29:09 -0500
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de> <uajueoqkg3f044@corp.supernews.com>
Message-ID: <3CAA1495.7080001@wi.mit.edu>

> Take a look at this tutorial which I found very useful:
> 
> http://www.icon.co.za/%7Ezapr/Project1.html
> 
> Glade is a Gnome application for designing graphical interfaces. You will
> then need libglade to parse the file and use pygtk to attach actions to
> the events generated by the user when they do something.

I'll also plug my tutorial from python 9

http://www.bioreason.com/~brian/GladeBase.html

The page hasn't been updated since I left bioreason though...

Brian Kelley



From kseehof at neuralintegrator.com  Sat Apr  6 15:32:13 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 6 Apr 2002 12:32:13 -0800
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204061633.g36GXWdj017461@ratthing-b246.strakt.com>
Message-ID: <MEENLMFCHGAOOICLMCHHKECPDLAA.kseehof@neuralintegrator.com>

 
> Still haven't found the time to read my original post, notice that
> I _did_ correct the name in the original post, and understand what
> I was trying to say, I see.  I will correct your misconception
> because you might confuse somebody else.
> 
> I am objecting to the fact that isleap, whatever the hell it is
> called, returns a boolean.  I am rejecting the existence of
> booleans in the language because it is too hard to teach how
> to use them properly.  Most people never learn.  Most code that
> uses booleans should not.

I agree with you, but I would say that the name isleap() should not
exist.  The problem is not with booleans, but with poor function
naming.  I'd prefer leapdays(year) that returns int.  If someones
makes the mistake of writing a function named isleap, they should
keep there promise implicit in the name and always return boolean.
The name is very much to the point.

> Proper use of booleans requires not only that you have a 
> function that returns one of two values, but THAT YOU NEVER
> EVER WANT THAT FUNCTION TO EVER BE MODIFIED TO RETURN ONE OF THREE,
> (or more).  What I tell my students is that you really want 
> a hole to open in the ceiling and a Warner Brothers safe to fall
> on the head of somebody who attempts to extend, subclass, or 
> in any other way change the function to return one of three values.

Proper use of booleans requires that the function evaluates
whether or not a proposition is true.  It is not sufficient that
the fuction merely returns two values.

Even if you are returning int, the safe will fall if the name of the
function implies a 0 or 1.  For example:

  if isleap(year):
      y = y+1

The above code is quite reasonable given the name of the function.
However, it will break if isleap returns a 2.  This is the fault
of the person who wrote the function, and especially the fault of
the person who foolishly extends the function to return a 2 without
renaming the function and going through all of the code that uses
the function.  If isleap is in an API, there is no good solution!

  if leapdays(year):
      y = y+1

This code has the same effect, but now the blame is on the person
who wrote the above code, not the person who wrote or modified the
function.  The difference is in the name (and perhaps the docs).
It is clear from the name of the function that the proper code
should be:

  y = y+leapdays(year)

On the other hand, the following code is ugly no matter what type
is returned:

  y = y+isleap(year)

But this looks ok to me:

  y = y+int(isleap(year))

> And, by God, if the function that returns _the number of leap days
> in the year_ isn't safe, at 0 or 1, then damn near close to nothing
> that my students are going to be tempted to use 'True' and 'False'
> for is.  This example makes my course notes, under Why using booleans 
> is bad, and you should never do it.  (unless you are a mathematician
> who wants to do real 2-valued algebra, or somebody interfacing with
> some other piece of software that needs those sort of bools, or 
> some other, EXTREMELY RARE,  person.  Don't you guys interrupt here.
> This is hard enough to teach as it is.)

How about a section "When not to use booleans"?

A function leapdays(year) that returns the number of leap days in the
year should return an integer.

A function isleap(year) that specifies whether or not the year is
a leap year should return a bool.  (better not to write this anyway)

I would tend to prefer to implement the former.  If someone foolishly
implements a boolean isleap function and the world government in
2024 decides to declare a double-leap year for political reasons, at
least you will -know- that you have to double check all of the code
that uses your function, since your self-documenting code implies a
promise that isleap will never return 2 or -1.  I would call it a bug
and fix it by renaming the function to leapdays returning integer,
then go through every usage of isleap and replace it with leapdays
while making sure that the code handles values other than 0 and 1
correctly.  Note that I would do have to go through this procedure
whether or not python has booleans.

Even without booleans in the language it would be inappropriate to
name a function isleap if there is some possibility that the value
will ever not be 0 or 1 since the name implies a promise never to make
such a change.  Booleans merely make such promises explicit.

By distinguishing between bools and ints, you motivate your students
to learn to write functions like leapdays(year) rather than
isleap(year).

> It does not help that it appears to the outside that Guido just went
> through all the python functions that returned 1 or 0 and changed
> them to return True or False.  I know that Guido would not be
> so stupid, and that there is a reason why he wants, for instance,
> isinstance to return only 0 or 1 for evermore, but I can't see it.
> I would like an isinstance that returned 0 for not an instance of
> this class, 1 for a direct instance, and 2 for a subclass.  There
> must be a lot of code out there I have never seen that relies on
> this never happening, and now, alas, I have to find out why.

I think that changing any function that used to return 0 or 1 to
return other values is likely to break code.  The desire for that
kind of flexibility is a seriously perverse interpretation of
polymorphism.

> Because my students will ask, of course.  If True and False are
> good enough for Guido, then it's good enough for them.  I would
> love it if a huge comment was added to python files that now
> return booleans saying why it was necessary to restrict the
> return values to one of 2 forever more, so people are not confused
> into thinking that returning booleans is a good idea in general,
> but that may be too much to ask for.  I would add the comments
> myself, but I am having problems seeing what the reasons are.
> Maybe after I get some more sleep, it will come clearer to me.

Any case that requires such documentation should not return bool.

I can think of several that don't need such documentation:
__eq__, __gt__, ...
hasattr
is_foo

> I truly liked telling students, especailly ones that have never used
> Python while we are using some other language,   'You don't need 
> True and False. Look at Python.'  
> 
> Now I have to say, to the students who are using
> Python: 'Python has True and False.  It is not for you little
>          worms, and if I catch you using it I will make you do 
>          the assignment over.'
> 
> I know in all the Python code I have ever written I have only 5 times
> wanted a boolean True or False -- all times when I was talking to a
> database that required it.  The rest of the time I wanted an integer,
> 0 or 1.  Even if I couldn't see a way that somebody was likely to
> extend it to return 3, there most certainly was never ever going to be
> any code that required that it only have one of 2 values, so now I
> will have to change all my True and False there to be 'Yes' or 'No' or
> something.  (grumble, now I wish I had lost the argument to the person
> who wanted the constants we picked to be TRUE and FALSE).

Why?  True and False will still behave like 1 and 0 if my understanding
of the PEP is correct.

Also, I don't believe you about only using booleans 5 times.  Every
time you use an 'if' statement you are using a boolean concept and
your code follows exactly two paths.  Every time you use >, =, < you
are using boolean concepts with corresponding expectation of two
possible results.  You probably mean that you only created variables
or functions with boolean values on five occaisions.  That is quite
reasonable.  Boolean varables and high level boolean functions are not
as common as boolean library functions and certain kinds of methods.
 
> Teaching people to not use booleans is hard.  They grab for them
> whenever they have a function that returns one of two values.  It
> was a lot easier to teach this concept when the fool things were not
> in the language.

Okay, you may have a point there.  However, IMO time spent gaining 
a clear concept of booleans is time very well spent.  In fact you
have come up with several excellent examples of when not to use
booleans.  You should add them to your lesson plan.  Every time you
see a student innappropriately returning bool to represent a pair
of values, you have an excellent opportunity to teach them to think
clearly about types in general.  I believe this lesson is so general
that it could help avert the Y2Ks of the future.

> Note:  don't tell me that because python bools are just ints, it
>        doesn't matter -- people who were using bools for
>        DEMOCRAT and REPUBLICAN can now add 
>        THE_DAMN_VOTE_WASNT_PUNCHED_THROUGH - (the integer 3)
>        when to their horror they discover 2 values was not enough.
>        I need to teach students to not use bools in languages which
>        are not so forgiving.
> 
> Laura Creighton

Perfect!  Bool should not be used to represent DEMOCRAT and
REPUBLICAN because party is not a boolean concept even if you
believe that there are only two parties!  Booleans are meant to
represent things that are true or not true.  Any value that I would
even be tempted to use a boolean for would certainly not be extended
to an integer.  Perhaps a float would be more likely, if I wanted
to used fuzzy logic :-)  Hmm.  Fuzzy logic.  Hmm.  Maybe the 'if'
statement should be extended to use fuzzy bools and automatically
fork weighted threads...  But I digress.  My point is that boolean
values are conceptually distinct from integers, even the set {0,1}.

- Ken Seehof






From paul at boddie.net  Fri Apr  5 04:56:51 2002
From: paul at boddie.net (Paul Boddie)
Date: 5 Apr 2002 01:56:51 -0800
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk>
Message-ID: <23891c90.0204050156.1840d913@posting.google.com>

Scrumpy <reply at in_newsgroup.tks> wrote in message news:<a8i8gn$rbe$1 at hfc.pacific.net.hk>...
> I want to learn Python for use in creating web apps and possibly to create 
> stand-alone apps with wxPython.

[...]

> There seem to be various projects which provide web app 
> servers/components/development frameworks. Quite a few seem to overlap ;-) 
> Here are one's I've looked at:
> Slither http://slither.toolsofcomputing.com/

I hadn't seen this one before, but it looks rather interesting. A
number of people have already asked about the Web modules overview
which I had, and in the replacement document(s) I should attempt to
cover Slither as well as the other major frameworks.

> Webware for Python http://webware.sourceforge.net/
> Skunkweb http://skunkweb.sourceforge.net/about.html
> OSE http://ose.sourceforge.net/

OSE looks powerful, but I haven't heard of many people using it for
Web applications.

> There are also various projects which 'integrate' Python with 
> Apache/AOLserver:
> Mod_Python
> Mod_Snake
> PyApache
> Cheetah
> PyWX
> 
> It look like I'm spoilt for choice and it's hard to decide where I should 
> direct my efforts.

Cheetah is a templating system, but you should not let the other
options listed confuse you too much - whether one uses mod_python,
mod_snake or PyApache should really only be a deployment and/or
performance consideration.

> As a side note, some of the comments made by Paul Boddie ( 
> http://mail.python.org/pipermail/python-list/2001-September/064133.html ) 
> seem justified.

I wish to change my mind regarding some of them, though. :-) Actually,
I think those comments still stand as an accurate, if subjective
record of Zope history from the point of view of a casual
experimenter, but I did get to try out certain things with Zope 2.5
recently and certain things are as easy as they always were (or had
the potential to be), and certainly easier than they are with other
frameworks. (Notably the use of Zope Components and their persistence
- this is a lot easier than with other frameworks.)

[...]

> Webware seems to be one of the more respected projects and I get the 
> impression that it improves the productivity/time to completion for people 
> using it.

There are live sites running Webware (such as http://www.vorbis.com)
and whilst, in my opinion, there are various issues that arise about
support for certain things which one might have believed to have been
covered by now, the development community is very active and very
eager to improve the framework. Reading the list shows how hard it is
to write a general Web framework, and I would recommend anyone
thinking of writing their own framework to skim the archives (and view
the Wiki) at the very least.

> On the surface, Slither seems to cover more web app server bases than 
> Webware. I am interested in persistance with backend db's but am more drawn 
> to using Slither with mod_snake than mod_python. So far though, I have not 
> found any independant info on Slither (and nothing at all on Slither with 
> mod_snake) so input from users with experience of using Slither and Webare 
> would be most welcomed.

There are actually many persistence modules in Python, as it turns
out, partly because persistence is a bit like Web page templating -
everyone has their own idea about how it should be done - but it's
also a subject which lends itself to different approaches. (In the
article you referenced, I still agree with those comments on
templating systems.)

> What I would be most appreciative of, is your helpful suggestions on how to 
> start in Python Web app development and on relevant books.
> 
> Finally (at last you cry!), although Zope is one way to go, as this is for 
> me primarily a learning experience, I would prefer to stay closer to Python 
> and 'perhaps' enjoy better web app performance. Zope does seem to be better 
> suited for creating CMS apps but I am not creating a CMS app :)

Having experimented with Zope, before becoming disappointed or
concerned about development, deployment and management issues, and
finally looking into Zope again after seeing interesting applications
like Squishdot in widespread production use, I think Zope is very
suitable for a number of application types. Moreover, through various
features, Zope specially addresses areas that the other frameworks are
less likely to address "out of the box", such as the way applications
can be configured by creating objects at appropriate places in the
object database - this is very intuitive (and quite remarkable) once
the concepts are understood.

> Bows and leaves the stage,
> Scrumpy :)

Applause? ;-)

Paul


From bokr at oz.net  Mon Apr  8 14:24:30 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 18:24:30 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8q0g1$qu5$1@panix1.panix.com> <a8qqf7$p4h$0@216.39.172.122> <Epes8.59265$pT1.1705378@news1.tin.it> <a8sf0k$2a5$0@216.39.172.122> <ahjs8.73608$S52.2597704@news2.tin.it>
Message-ID: <a8sn8u$sj7$0@216.39.172.122>

On Mon, 08 Apr 2002 16:09:42 GMT, Alex Martelli <aleax at aleax.it> wrote:

>Bengt Richter wrote:
>    ... 
>>>>>Assignments in expressions will never happen.
>    ...
>> But why "identical" ;-) If I can find another spelling that works
>> everywhere (and it wouldn't surprise me if you had one in mind ;-)
>> wouldn't that be interesting?
>
>http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66061
>
>for example.  But, like your attempt, it's based on NOT being
>an assignment (so Aahz's quote above remains perfectly right).

Well, I guess that depends on how you define "assignment" in Python.
ISTM there really is no such thing, in the sense of updating a named
storage place with new content, so what goes by the name "assignment"
needs some explanation. If the statement a=b is an assignment,
then binding the name 'a' to object b is the meaning of Python
assignment as I see it. And the expression

    locals().__setitem__('a','bee') and 0 or a

did result in binding 'a' to 'bee' in the context I used it, so
I feel justified in calling it an assignment (side effect) in an
expression ;-) I'm not sure what that side effect should be called
if not assignment, since it caused the identical binding that
a='bee' would in that context.

If it binds like a duck... ;-)

>
>Not all "stashing a reference to this value off somewhere" are
>assignments (some of the many who aren't can correctly appear
>within expressions -- others can't).
>
My criterion for judging whether it was an assignment was whether
it accomplished the same name re/binding as a normal assignment,
and I think it did, in the demonstrated context. So IMO Aahz's
quote doesn't quite remain perfectly right ;-)

Regards,
Bengt Richter


From gmcm at hypernet.com  Thu Apr 11 09:28:46 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 11 Apr 2002 13:28:46 GMT
Subject: Include zlib.dll in McMillan Installer
References: <a93ld3$jb4$03$1@news.t-online.com>
Message-ID: <Xns91ED608AADBDEgmcmhypernetcom@199.171.54.213>

Harald Schneider wrote:

> is there a possibility to include zlib.dll into the single file
> distribution of the installer?
> E.g. PIL uses zlib and complains about missing it on some systems.

In ancient days (1.5.2), zlib.pyd loaded zlib.dll. But
since then, zlib.pyd has changed to statically link in
zlib, so zlib.dll should not be needed (in fact, it
doesn't exist in the distribution).

Installer will always include zlib[.pyd/.so], unless 
your installation doesn't have it, or you turn it off
by editting config.dat. Windows comes with zlib.pyd,
so I expect it's there (you can check with ArchiveViewer.py
to see if it's in the exe).

So more details are needed.

-- Gordon
http:/www.mcmillan-inc.com/




From brian at sweetapp.com  Sat Apr 27 13:39:41 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 27 Apr 2002 10:39:41 -0700
Subject: Python's Lisp heritage
In-Reply-To: <tyfsn5h13mh.fsf@pcitapi22.cern.ch>
Message-ID: <002201c1ee12$82e9ffc0$445d4540@Dell2>

Jacek wrote:
> What? you object to my obtuse avoidance of Python's nifty infix
> synonyms for these operators which cunningly call the functions
> without recourse to parentheses?  In that case you surely allow me to
> object to obtuse avoidance of such possibilities in Lisp.

The Lisp version is still more obtuse. Ask 100 people to evaluate the
expressions "sin(0/2 + pi/2)" and "#i(sin(0/2 + pi/2))". I bet 90% of
people won't know what sin is, 10% will be able to evaluate the Python
version and 1% will be able to evaluate the Lisp version.

Cheers,
Brian





From ark at research.att.com  Wed Apr 10 14:18:58 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 18:18:58 GMT
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <mailman.1018385900.7252.python-list@python.org> <u8z7valyh.fsf@python.net> <yu99662zmufg.fsf@europa.research.att.com> <rxZs8.254$p56.69633@newsb.telia.net>
Message-ID: <yu99vgazifhp.fsf@europa.research.att.com>

Fredrik> you could of course use a RE-like approach:

Yes, that's one possibility.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From sholden at holdenweb.com  Tue Apr 16 07:46:33 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 07:46:33 -0400
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>
Message-ID: <KcUu8.22993$Ji7.1840@atlpnn01.usenetserver.com>

"sameer" <sameer_ at email.com> wrote ...
> I am trying to run multiple threads in a wxFrame class to instantiate
> different panels within it on Win 2k,  and well, it's a little hairy.
> Anyone got a good web source that discusses python threading in depth?
>
> More so, my question is, how can I get the return value of a function
> that I run on a seperate thread?  I don't see anything for this in the
> threading module.

The canoncial reference was written by Aahz, whe keeps threatening to update
it. See

    http://starship.python.net/crew/aahz/OSCON2001/index.html

The Queue module might help you with passing values between threads, as it's
specifically designed to be thread-safe.

regards
 Steve






From francis.meyvis at sonycom.com  Tue Apr 30 13:18:47 2002
From: francis.meyvis at sonycom.com (Francis Meyvis)
Date: Tue, 30 Apr 2002 17:18:47 GMT
Subject: remove read only flag
Message-ID: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>

Hoi,

How can one remove the read-only flag from a file
(e.g. c:\temp\file.txt) using python in win32.

Tx
--
---
Kind regards, best wishes,
francis meyvis




From aahz at pythoncraft.com  Tue Apr  2 13:17:10 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Apr 2002 13:17:10 -0500
Subject: using (was Re: The Python Way)
References: <5515F24370BCB753.DA7CF91464346EF6.E3A577CFA6882D82@lp.airnews.net> <jjb5-0A3399.14320327032002@newsstand.cit.cornell.edu> <a80nej$jlp$1@panix1.panix.com> <mvkq8.43706$pT1.1223399@news1.tin.it>
Message-ID: <a8csj6$9ca$1@panix1.panix.com>

In article <mvkq8.43706$pT1.1223399 at news1.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>finally:
>    <guaranteed termination stuff>
>
><code>
>
>has the same semantics as:
>
>try:
>    <code>
>finally:
>    <guaranteed termination stuff>

Good idea.  Now rework it to have a Pythonic connection between your
finally block and the code it applies to.  ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From bokr at oz.net  Tue Apr 16 04:39:24 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 08:39:24 GMT
Subject: Does Python need a '>>>' operator?
References: <mailman.1018919132.7451.python-list@python.org>
Message-ID: <a9gnvs$42p$0@216.39.172.122>

On Mon, 15 Apr 2002 21:03:45 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[Bengt Richter]
>> ...
>> I just think the signed hex is not terribly useful.
>
>In practice, if I'm working with 300-bit bitsets represented as longs, I
>never do hex(x), but hex(x & mask) instead, where mask = ~(-1L << 300) (a
>string of 300 low-order bits).  Regardless of x's sign, that "shows the
>bits" in a 1-means-on 0-means-off sense.  The key is that I *know* the width
>I have in mind, much as hex(int) knows the width of a platform C long today,
>so can pull the same trick for regular ints.
>
Sure. But you can also pick a width for minus_sign-less-but-not-sign_bit-less
hex purposes by defining a normalized width that encompasses just enough bits
to show a hex 0 or F as the first digit, knowing that all to the left are
redundant duplication. (Credit to Greg Ewing for the suggestion of making sure
it's always a reader-friendly leading 0 or F to indicate sign, vs. my ensuring
just an msb of 0 or 1 within the ms hex digit).

The current hex output also normalizes out leading duplicate sign bits, in effect.
what I am suggesting is just a differently normalized hex representation, and you
don't need to specify width nor create a positive number just to see the bits.

I recognize backwards compatibility problems with using 0xF9 to mean -7, so I
have been saying we could us 0h in place of 0x, so that -0x7 (unified L semantics)
can live along side the equivalent 0hF9.

BTW, if I wanted several looks at 300-bit numbers in hex, I'd probably use '%075x'.
Any plans for modifying this when there's no implied int widths at all?:

 >>> '%016X' % -3
 '00000000FFFFFFFD'
 >>> '%016X' % -3L
 '-000000000000003'

I'd like the unified result here (where we _are_ specifying a width and prefixing after all ;-)
to be:

 >>> '%016H' % -3
 'FFFFFFFFFFFFFFFD'

and if we don't specify width, it could be normalized down to

 >>> '%H' % -3
 'FD'

(I faked those, of course).
Regards,
Bengt Richter


From dominikush at yahoo.com  Wed Apr 17 02:53:55 2002
From: dominikush at yahoo.com (dominikush at yahoo.com)
Date: Wed, 17 Apr 2002 08:53:55 +0200
Subject: Counterintuitive Python behavior
Message-ID: <3CBD1C02.8AB2CF62@yahoo.com>

Hello,

one thing I like very much about Python is that statements
work like you would expect them to work. Take for example
the use of dict.values() for dictionaries: If you store the
result of dict.values(), and change the dictionary after-
wards, the previously stored result remains untouched.

>>> dict = {'a':1,'b':2}
>>> list = dict.values()
>>> list
[1, 2]
>>> dict['a']=3
>>> list
[1, 2]
>>> dict
{'a': 3, 'b': 2}

However, if a dictionary has lists as value entries, I get
a counterintuitive behavior (which, just recently, broke my
code): If you change the dict, the list you previously
created via dict.values() gets automagically updated. A nice
feature, but nothing I would have expected!

>>> dict = {'a':[1],'b':[2]}
>>> list = dict.values()
>>> list
[[1], [2]]
>>> dict['a'].append(3)
>>> dict
{'a': [1, 3], 'b': [2]}
>>> list
[[1, 3], [2]]

Looks like that in the first case a copy is returned while
in the latter case list references are returned. Ok, but
according to Python's philosophy I shouldn't mind if I work
with lists in the dictionary or anything else. If the
behavior depends on the knowledge of the type of values I
put into a dictionary, I find that somehow counterintuitive.

Who is wrong here: my intuition or Python (2.2)? If it's
my intuition, how can I train my thinking about Python's
execution model, so that my intuition get's better ;-)

Best regards,

Dominikus




From trentm at ActiveState.com  Thu Apr 18 12:27:12 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Thu, 18 Apr 2002 09:27:12 -0700
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
In-Reply-To: <20020418092251.K28973@ActiveState.com>; from trentm@ActiveState.com on Thu, Apr 18, 2002 at 09:22:51AM -0700
References: <mailman.1019071474.3193.python-list@python.org> <Xns91F4680F3221cliechtigmxnet@62.2.16.82> <20020418092251.K28973@ActiveState.com>
Message-ID: <20020418092712.M28973@ActiveState.com>

[Trent Mick wrote]
> 
> (Reply privately be mistake in first response.)

My also try english speak in most responces.

Phew,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From ianb at colorstudy.com  Mon Apr 29 01:40:07 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 29 Apr 2002 00:40:07 -0500
Subject: Python Webpages
In-Reply-To: <aai5oo$99l$1@maud.ifi.uio.no>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> 
	<aai5oo$99l$1@maud.ifi.uio.no>
Message-ID: <1020058807.2285.267.camel@localhost>

On Sun, 2002-04-28 at 19:56, Johan Fredrik ?hman wrote:
> Just realized that these Servlet based development models really limits the
> usage of software like Dreamweaver.  Since all code is in the program these
> utilities are almost worthless.  Is there a good way to combine Webware with
> templates.  I think of templates as pages where certain %words% are substituted
> etc...

Yes, you can use PSP like a template (putting most of your code in
Python servlets and modules, with your layout in the PSP files), or you
can use a templating language of some sort -- Cheetah is one, and there
are many others.  It's not even that difficult, AFAIK, to use ZPT like
this (without Zope).

  Ian






From aleax at aleax.it  Mon Apr 15 07:46:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 11:46:32 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no>
Message-ID: <s4zu8.20814$vF6.690664@news2.tin.it>

Magnus Lie Hetland wrote:

> In article <a9cv63$qhu$1 at bob.news.rcn.net>, Raymond Hettinger wrote:
>>The help and tutor maillists indicate that the use of input() rather than
>>raw_input() is widespread. I think this function is so harmful and so
>>likely to be used that it should be deprecated hasta pronto (right away).
>>When really needed, eval(raw_input()) is easy enough to code and explicit
>>about what it does.
>>
>><no wink -- this function needs to go away>
>>
> 
> But (in the (very?) long run) I think that the name "input" should
> then be used for what is now raw_input. I know -- breakage, and all
> that... I just think it fairly odd to have *only* raw_input, since the
> "raw" qualification then becomes fairly meaningless.

I think the best name for what's now raw_input might be something
like getline.  That's what it does, net of line ending issues -- get
one line (from stdin).  'input' is too vague.  And with a zillion
existing books explaining 'input', we can never reuse that name
for anything else, IMHO, without utterly confusing future newbies
who happen to have an old book or website or &c at hand.


Alex



From jeffbi at jetsoft.com  Mon Apr  1 18:40:48 2002
From: jeffbi at jetsoft.com (Jeff Bienstadt)
Date: Mon, 1 Apr 2002 15:40:48 -0800
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net>
Message-ID: <3ca8f000$1@news.nwlink.com>

"Mindbomb" <Mindbomb at localhost.net> wrote in message
news:RHSZTyiDU6p8EwiF at nym.alias.net...
> In message <3ca4a322 at news.nwlink.com>, Jeff Bienstadt
> <jeffbi at jetsoft.com> writes
> >I've just recently started working on a newsreader in Python (it was the
> >underlying reason behind my earlier "IntSpan" posts).  It's not much
> >further along than the idea stage at this point, since I'm trying to
> >sandwich it in between several other projects, but I have a few lines of
> >experimental code that uses nntplib.
> >
> >I'm doing this primarily as a way to learn Python (why start small :->),
> >but also so that I can have a news reader that
> >   a) works
> >   b) runs on both my Linux and Windows boxen
> >   c) isn't Outlook Express
>
> If you carry on with this, I'd for one feel there is a gaping hole of
> newsreaders which actually make it east to post anonymously. If it could
> use anonymous re-mailers and interface with PGP/GPG it'd be even better!
>
>
> >If anybody is interested in joining me on this, or just interested in
> >hearing some of my thoughts on the subject, e-mail me --- I love
> >collaborative projects.
>
> I'd love to get involved in something like this, but I fear my Python
> skills would just show me up. I lurk in here, reading interesting
> threads and have dabbled a little with Python on my Linux box, but never
> actually sat down and played with the language enough to really teach
> myself.
>
> I'd be interested in hearing about your design & thoughts though - maybe
> this could be there very thing to hold my interest enough to get me to
> write some decent code, seeing as I have a vested interest!
>
> --
> Mindbomb

Regarding anonymous postings, I don't use that myself so I don't
grok the advantage other than spamming newsgroups.  Or is it a
mechanism for hiding from spam?  If it's useful, I'd be happy to
include it, but not if it's just a spam enabler.

As far as your Python knowledge goes, I wouldn't worry about it.
My Python experience is currently measured in weeks, although
I've been doing the coding thing (mostly C/C++) for more years
than I care to admit.  I'm finding it surprisingly easing to learn;
the hard part is giving up my hard-earned C++ habits :-)

My goals for this project (personally speaking) are, in this order:

   1) Learn Python
   2) build a newsreader that I'm happy with
   3) build a newsreader that others are happy with :-)

C'mon down!

---jkb





From skip at pobox.com  Tue Apr 16 17:57:35 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 16 Apr 2002 16:57:35 -0500
Subject: Unicode name questions
In-Reply-To: <009901c1e58b$28a5bd90$445d4540@Dell2>
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com>
        <009901c1e58b$28a5bd90$445d4540@Dell2>
Message-ID: <15548.40527.385680.204020@12-248-41-177.client.attbi.com>

    Brian> Your spelling of the first two isn't quite right (at least
    Brian> according to the Unicode guys). Remove the "B" from "LAMBDA".

"Lambda" has been spelled with a "b" as long as I can remember.  I see both
"LAMBDA" and "LAMDA" in the comments in the encodings modules though.  I'll
give it a try.  Note that

    http://www.w3.org/TR/REC-html40/sgml/entities.html

spells it with a "b".

    Brian> And change "GREEK SMALL LETTER THETA SYMBOL" to "GREEK SMALL
    Brian> LETTER THETA"

Note that there are both '&theta;' and '&thetasym;' HTML entities.  I found
it here:

    http://www.htmlhelp.com/reference/html40/entities/symbols.html

but also saw it here:

    http://www.w3.org/TR/REC-html40/sgml/entities.html

where it is marked "new".  I suspect perhaps Python's codecs just haven't
caught up yet.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From python at rcn.com  Tue Apr 23 16:38:04 2002
From: python at rcn.com (Raymond Hettinger)
Date: Tue, 23 Apr 2002 16:38:04 -0400
Subject: dictionary method question
References: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com> <3CC5BB74.9070009@alcyone.com>
Message-ID: <aa4gpc$7ma$1@bob.news.rcn.net>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CC5BB74.9070009 at alcyone.com...
> dsavitsk wrote:
>
> > for some dictionary d is there any relationship between the returns from
> > d.keys() and d.values()?  that is, to get them in some arbitrary but
> > matching order, do i need to loop over d.items() or can i count on a
> > similarity of the above? in small tests this seems to be the case.
>
> I presume they would return values in the same object order, but I
> wouldn't count on it.

Erik is absolutely correct on both counts.

The C source for the .keys() and .values() is identical except for the
structure element returned.  So, if given the SAME dictionary, the two
routines will loop of the dictionary in the same order.   HOWEVER, the
dictionary is not guaranteed to remain the same.  A resize operation will
reorder the keys.  Because resize removes dummy entries for previously
deleted items, the order may change even if  the contents of the dictionary
remain the same.

>The .items method is designed expressly for that
> purpose anyway; you're better off using .items since it's potentially
> faster than zipping the two lists together yourself, and it's clearer
> anyway.

Excellent advice.


Raymond Hettinger





From aleax at aleax.it  Fri Apr  5 17:16:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 22:16:52 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAE24A2.7ECBB78@accessforall.nl>
Message-ID: <onpr8.64923$S52.2260694@news2.tin.it>

Ype Kingma wrote:
        ...
>> Even if having to modify the __getitem__, __setitem__, __delitem__ (at
>> the least) of a zillion user-defined types out there was feasible (and
>> it's not), I think this would not fully exhaust, as I indicated.
> 
> The users writing such code were knowingly taking a risk by using
> a name that starts with a double underscore. For these cases I think
> deprecation is quite acceptable.

You completely misunderstand the role of special names in Python.
It doesn't mean "I'm taking a risk": it means "I'm following fully
documented, STABLE interfaces".

> (Off course one could simply iirc :1,$s/\<__//g in all concerned code,
> but that would cause too many other problems.)

It would make every someobject[x] mention stop working.  It's
therefore a totally absurd suggestion.


> When the object passed to your __getitem__ method does not
> have .start and .stop attributes you could simply call int()
> on it from within your __getitem__ implementation.

Incompatible with built-in lists, which raise an exception when
indexed with complex or float numbers for example.  There are
good reasons for that, and it's unwise to suggest breaking the
mold with as little understanding of those reasons as your posts
are consistently suggesting.

> But you would only need to do this in certain cases, see below.

In fact, doing it in most other cases would be very unadvisable.


>> > Indexing a dict with both int and bool keys might involve some
>> > confusion when True is not 1 anymore. This case should be rare,
>> > however.
>> 
>> Why should it be rare?  All hashables are used uniformly as keys
>> into dictionaries.
> 
> It should be rare because the confusion would only occur when both
> True and 1 are used as keys and they are intended to be the same key.

Any existing code doesn't know about any distinction between True and 1.
There is no distinction between them up to the current release.  Therefore,
any existing code will become utterly fragile if and when a distinction 
should suddenly appear out of the blue.  From the point of view of code
implementing a mapping, there's no "intended to be the same" -- as there
is no notion of any distinction between two cases (there is only one case
today).  Similarly for client-code, code that USES a mapping (most often
polymorphically).  What "intended to be" can it mean (or fail to mean)
between objects that's always been identical?  For example, using a dict
as a set has always yielded identical results for:
    - testing membership with aset.has_key(item),
or:
    - indicating membership with aset[item] = 1, then
    - testing membership with aset.get(item, 0)

Why would you think this is "rare"?  It's common currency.  How many tens 
of thousands of lines of python have you written, reviewed, or maintained, 
that you presume to teach me what is common and what is rare in python
"as she is spoken", pray?

> That should only happen when True and 1 are used interchangeably,
> which might happen by programmer mistake or by using
> old (1/0) and new (True/False) style code together.

Old-style only will do perfectly well, since lots of extremely common 
idioms therein (a<b, d.has_key(k), x in y, and so on) change return
values silently.  Neither any mistake nor any 'together' is needed,
unless you're somehow trying to imply that anybody coding some kind
of "advanced" (?!) construct such as "a<b" was using "new style code"
unwittingly years ago...


> The slowdown would be there for backwards compatibility only.

In yoyr proposal the slowdown would be there in every dict access.

> In the pure case an int() call would not be needed at all:

But to check whether it's needed or not, a type-check would be
needed, and that IS a substantial slow-down.  You don't seem to
realize the very elementary facts of Python life, yet keep presuming
to "teach your grandmother to suck eggs".  Try.  Get python 2.2
plain source and insert an always-failing typecheck on each access
to all dicts, then measure (e.g.) pystones before and after.  I haven't
done this specific absurd-measurement, but, like every spunky new
Pythonista presuming HE ('shes' are more sensible) can of course
improve over years of effort by the likes of Tim and Guido, I did try
all sorts of funky "improvements" in fundamentals such as dicts and
lists.  Of course I wasn't quite foolish enough to POST about them
(mostly) -- that's the nice thing about open source -- you can be
stupid in private, and save some modest amount of face.  Most are.
You seem to be something of an exception -- suggesting changes
to crucial fundamentals without a smidgeon of try-and-see.  Give
your suggestions a try... a dry run suffices... and have a laugh.

> When adding the int() call to the __setitem__ method of the standard
> dicts for backward compatibility the dict shown above would only a have
> a single key 1 and the value would be the last value put there.

...and pystones would become...?


>> "only hope" is not acceptable in this context.  Fortunately, we can
>> do better than that (and I'm confident that the unacceptability of
>> slowing down dictionaries and lists is quite obvious to decision
>> makers in this case, so I don't fear it will happen).
> 
> I'm still putting my hopes on deprecation (as above) after PEP 285 enters
> into force. During the deprecation phase with non int bools such cases
> could be warned for eg. by a special verion of int() that would also cause
> a slowdown, and after the deprecation a pure version with normal speed
> would remain.

We have a saying in Italy -- "chi vive sperando muore cantando".  "He
who lives hoping, dies singing".  I suggest you practice gorgheggio -- not
a popular singing technique any more, alas, but should you ever find
yourself cast in some (e.g.) Salieri or Gluck (Italian-style) it might come 
in handy.  And if you don't understand a word I'm saying, may I suggest
remedying that with a course in opera buffa -- its 17-th century and
earlier roots, its 18-th century apogee (you'll find debate on that timing,
but that's the school I subscribe to), its 19-th century long, slow decline
(ditto).  With any luck, such a course might just happen to show you the
career you were always destined to, and distract you from subjects you
appear to be unsuited for, such as programming language design and
implementation.  One can always hope (recursive call to the beginning of
this paragraph).


> P.S. I havn't followed the change of behaviour of the / operator in
> detail, but wasn't/isn't this done using a deprecation period with
> warnings, too?

And commandline switches (see -Q, still alive and kicking) and so
on.  It's not a matter of "following the change", you just need some
modest amount of studying the current Python release.

Now a surprise-test: how many divisions are there in a typical piece
of Python, versus, how many dictionary lookups...?  Fair warning: this
IS a trick question.  Only sensible way to answer it is, instrument the
Python compiler.  If you find this a problem, then you're not very
qualified to comment on language implementation strategy, are you...?


Alex



From vinay_sajip at yahoo.co.uk  Wed Apr 10 13:54:48 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: Wed, 10 Apr 2002 18:54:48 +0100
Subject: PEP 282: A Logging System
References: <Pine.LNX.4.44L.0204101618300.13708-100000@ods.pp.ru>
Message-ID: <000501c1e0b8$d1136cc0$652b6992@alpha>

Denis> First it's very slow. Here are timings obtained from my test script
Denis> (http://ods.pp.ru/logging/logtest.py):
Denis> ---
Denis> simplelog (log): 0.59
Denis> simplelog (ignore): 0.07
Denis> logging (log): 212.88
Denis> logging (ignore): 0.42
Denis> logging (log, no inspect/thread): 3.25
Denis> logging (ignore, no inspect/thread): 0.42
Denis> ---
Denis> That's mainly due to unnecessary inspect calls. Vinay said that he
fixed
Denis> this problem in upcomming release, but the module is still slow. I'm
not
Denis> sure what we realy need: not-so-complicated but fast logging system
or
Denis> extremly flexible and a bit slow? I guess it should be very fast at
Denis> least when logging is disabled.

The inefficiency in the way I called inspect was *major*. I ran Denis'
script on my machine - on Windows - so I had to change "/dev/null" to "nul"
:-). I ran with logging module V0.4.1 (same as Denis), with Python 2.1.1.

- time-0.4.1----------------------------------
simplelog (log): 0.18
simplelog (ignore): 0.04
logging (log): 125.07
logging (ignore): 0.17
logging (log, no inspect/thread): 1.06
logging (ignore, no inspect/thread): 0.16
----------------------------------------------

The performance was, I agree, pretty poor! But then I ran the same test
against the logging module V0.4.2 (not yet released - but will be in a day
or two). The timings are much better:

- time-0.4.2----------------------------------
simplelog (log): 0.18
simplelog (ignore): 0.04
logging (log): 5.59
logging (ignore): 0.16
logging (log, no inspect/thread): 1.08
logging (ignore, no inspect/thread): 0.16
----------------------------------------------

The speedup is a factor of over 22.4, just with this one optimization. (Note
that the bulk of this time is in calls to inspect and thread modules. I
think most of this is in calls to inspect. I've commented on threading
below.)

Now, this may still not be fast enough for some people. It's horses for
courses - there's often a trade-off between flexibility and performance.
Denis' example shows a very simple scenario - one that doesn't use different
logging channels, nor logging to different sinks. To me, the power of this
basic architecture (and I take no credit for this - it's log4j's) is that
you can turn logging verbosity up and down, in different parts of the
application, at different times, without modifying the source code -
likewise, you can arrange for the events to end up for consumption on the
other side of the world, if you want - again without modifying the
application's source code. This flexibility is worth paying some performance
penalty for, but exactly how much overhead is bearable?

I did another benchmark, which I think shows logging in a more realistic
scenario - as a marginal part of the application rather than hogging the
limelight (and the CPU). My benchmark is an existing application which uses
about a minute of CPU time. The application scans PDF files; I added a
single
.debug() call to the root logger whenever a PDF page is processed. This
function was called 266 times in my test program. I ran on the same data,
with 4
scenarios:

A - The .debug() logging call was commented out (this is the baseline for
comparison).
B - logger level set to DEBUG, and the .debug() call was enabled.
C - logger level set to INFO, and the .debug() call was still enabled.

Here's what I found, over one single test run per scenario. The figures are
from the Python profiler.

A - No logging calls -  2634576 function calls (2632640 primitive calls) in
48.997 CPU seconds
B - Logging call made with logger level set to DEBUG - 2652937 function
calls (2651001 primitive calls) in 49.761 CPU seconds
C - Logging call made with logger level set to INFO - 2635388 function calls
(2633452 primitive calls) in 49.508 CPU seconds

These figures indicate that with logging calls enabled but filtered out by
level, for this application the logging overhead (with no useful work done)
is 1%. With useful work done (i.e. actual output occurring), the overhead is
1.55%. Admittedly, one call in the application is hardly a strenuous test.
(But then it's not a big application.) More work is needed in this area. Can
I have comments, please? Over to you, fellow Pythoneers! (If anyone is
interested, I can send them the actual profiler output so they can see the
relative times spent in the application, inspect, logging and other library
modules).

Denis> May we should use lazy evaluation of some time-consuming properties
Denis> (filename/lineno, asctime)?

The asctime formatting only happens at the point at which a handler is about
to emit a record - it is deferred till the last reasonable moment. Some
marginal optimization can be done here - for example, currently, if the same
record is passed to multiple handlers each having the same formatter, the
asctime formatting is done again without caching (assuming that the format
string has asked for asctime).

The filename/lineno information needs to be obtained when the call is made,
because the log record may be sent to a remote machine for presenting to
users. The logging system has no knowledge, when a call is made, what
handlers are configured (even if it could tell for the basic handlers
provided, user-defined handlers could do anything with the message). Hence
this information is always obtained - a little expensively, as you can see.

If it's felt that this information is only of use some of the time, I could
make a module-level setOptions(**kwargs) available. One of the kwargs might
be "want_caller" to say that inspect should be called to get source
file/lineno info (the default could be that this info is not collected).

Denis> Default behavior of logging is to log all messages: if no handlers
Denis> found on first call, default is installed.  That's strange.  I guess
it
Denis> should produce no log messages by default, i.e. no default handlers
Denis> should be installed.

Only the module-level convenience functions call basicConfig() if no handler
is installed. These are intended for "casual" use of logging [in short
scripts], or for users new to logging. If you just execute the statements

import logging
logging.getLogger('').debug('Hello')

then no output would occur except for a one-off message stating that no
handlers were found (this is there only to indicate that there may be
something wrong with the logging configuration).

Denis> I dislike the idea to import (and use for locking) thread module even
Denis> when application is not threaded. This is a general problem with many
Denis> modules. May be we need a atexit-like registry of functions that will
be
Denis> called on first thread import?

The logging module is not designed to use threading, or for thread safety.
The thread module is only called to find out what the calling thread ID is
(for inclusion in the logging message). Again, this information will not be
available at a later time for "lazy evaluation" to pick up. It will only be
useful in certain contexts (e.g. Webware applications) so perhaps it needs
to be made optional through the setOptions(want_thread=0) function. Feedback
please!

Denis> The filter interface (object with filter method) is defined so that
even
Denis> simple filters cannot defined inline. It can be changed to require
any
Denis> callable object (instance with __call__ method or just an inlined
Denis> lambda). Vinay said that this make it less readable for newbies, but
I'm
Denis> not sure he is right.

While having no problems with lambda or __call__ myself, I feel that this
idiom is less common and a little less readable. Pythoneers, please disabuse
me of this notion if you think otherwise, Denis already has :-) The other
point about filters is that you can use any object which has a method
filter(self, record). It may be that this is some other application class
(which contains useful application context to determine whether a record
should be logged or not). Now, if I had such a class, I might choose to use
the __call__ functionality to represent something more directly related to
my application's functionality. If I (as app developer) want to do so, I
won't take too kindly to logging "taking over" the semantics of __call__.

Denis> The most usefull filter (after simple setting the minimum log level)
I guess should be a name filter. Something like this (I've replaced filter
method with __call__):
Denis> def NameFilter:
Denis>     def __init__(self, name):
Denis>         self.name = name
Denis>         self.length = len(name)
Denis>     def __call__(self, record):
Denis>         return record.name.startswith(self.name) and \
Denis>                    (len(record.name)==self.length or
Denis>                     record.name[self.length:self.length+1]==',')

I think this is a useful filter (I assume the "," in the code above should
be a ".") - I've taken to Denis' idea and propose to incorporate this
functionality into the basic Filter class (will be released in 0.4.2).
Filter will now have an initializer with an optional name parameter - if a
false value, Filter will behave as before. Otherwise, we use it as a
hierarchy position to filter out all records which do not fall below a
certain section in the hierarchy. Is this a good idea?

Looking forward to python-list comments,


Regards

Vinay Sajip



_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From aleax at aleax.it  Thu Apr 11 12:17:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 16:17:24 GMT
Subject: Tuple comprehension
References: <mailman.1018540222.3383.python-list@python.org>
Message-ID: <oGit8.8539$vF6.244299@news2.tin.it>

Tim Peters wrote:
        ...
> ...     for i in range(n):
        ...
> is one way to get the same end result without building a list in one gulp

Actually, range DOES "build a list in one gulp" (but I guess you might
use xrange instead:-).


Alex



From max at alcyone.com  Tue Apr  2 23:45:30 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 02 Apr 2002 20:45:30 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>              <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com> <uakogq1r0i0524@news.supernews.com>
Message-ID: <3CAA88EA.194CC0B2@alcyone.com>

John Roth wrote:

> I absolutely agree. For me, the entire point is providing
> a **standard** way of expressing **in the code** the
> notion that a particular variable only takes two values:
> True and False. This is such a common case that it deserves
> to be promoted to be a part of the language.

Indeed, and it being standard is important.  I can certainly make my own
names for True and False (or true and false, or TRUE and FALSE) but they
may not be the same name -- or even have the same definitions! -- as
something other module writers have used.  When you have to interact
with a bunch of different Boolean types from a bunch of different
modules and authors, it sorts of defeats the self-documenting aspect.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From printers at sendme.cz  Wed Apr  3 03:33:31 2002
From: printers at sendme.cz (A)
Date: Wed, 3 Apr 2002 10:33:31 +0200
Subject: How to check whether email /domain exists
Message-ID: <3CAADA7B.25443.509AE1@localhost>

Hello,
What is the simplest way how to check 
 whether a given email address( or domain for this email) exists?
Thanks for help

Ladislav





From aleax at aleax.it  Mon Apr  1 10:11:38 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 15:11:38 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de>
Message-ID: <KM_p8.41075$pT1.1129125@news1.tin.it>

Martin v. Loewis wrote:
        ...
>> b. The hoped-for uniformity of notation won't happen -- we've already
>> seen on these threads people who want every constant to be uppercased
>> (to make it 'stand out') and so will use TRUE=True etc anyway.
> 
> That won't happen. People don't dare rename builtins, normally, and

Maybe it won't happen, but one poster on this thread _has_ said he
will do it -- if you claim he's lying, take it up with him.  Note that we're
not talking about "renaming" built-ins, just adding "synonyms" (binding
_extra_ names to the same object, without removing previous ones).


>> iff they're used to them from previous languages -- I've never seen
>> people used to (e.g.) APL, good old (uncommon?-) Lisp, or (dare I say it)
>> Perl, feel sad and deprived by the lack of booleans in their respective
>> languages.
> 
> That does not prove the claimed conceptual benefits as
> illusions. 

No proof (has any _proof_ ever been brought on this thread -- or just
about any?!), just an indication.  Note that those people are hardly
monolingual (I've met _once_ a person whose _only_ programming
language was APL, and I still recall it as a surprise, for example) -- they
are also quite likely to have experience of other languages such as
Fortran-IV, which did have 'LOGICAL'.

> Programmers that are used to booleans from previous
> languages might well find their lifes easier. Even people that you

In the same sense as they might well find their lifes easier if suddenly
they had to start decorating their Python programs with deuced braces
or BEGIN/END to indicate block structure -- they might find the
familiarity comforting.  It would still be a substantial mistake for Python
to give up to that perennial pressure, despite the minor advantages
(familiarity for Python-newbies coming from other languages, easier
life when embedding Python source in environments that may not respect 
whitespace or e.g. reading Python sources on the phone) because the
disadvantages would dominate.

> have not seen sad might enjoy this enhancement - they just did not
> know what they were missing :-)

This presupposes monolingual programmers, ones without exposure
to e.g. Fortran IV or Pascal.  I don't think that's typical at all of users
of languages such as APL or Perl.


>> Which leads me right onto disadvantages of the PEP -- because that's
>> the main one I see: extra conceptual load in teaching the language.  "0
>> can be spelled either 0 or False" with subtly (but generally
>> irrelevantly:-) different consequences...
> 
> 0 cannot be spelled False. 0 is spelled 0 (or perhaps 00, or 0x0). 0
> is a integral number, False is a boolean. These are different things.

No they're not: they're DISTINCT (is not) but NOT *different* (!=) -- not
even as 'diverse' as, say, 0 and 0.0, since there are quite a few cases
where you may use the former and not the latter (such as indexing) but
part of the PEP's claims is that (string representations apart) there will
be essentially no cases in which 0 is acceptable and False isn't (so, by
the "walk like a duck" principle, the distinction becomes rather vague --
and by that much harder to teach, and to use appropriately).

The point is: given any function named iswhatever, you may get from
it two different results when the question (predicate) is answered in
the negative -- False, if it's a function from the Python libraries, a
recently-coded function which respects the new 'bool', or an old function
which happens to use Python library functions (and operators) whose
result type has been changed -- 0, for an old function which happens
to have an explicit return 0.  No matter, because -- except for string
representations, basically -- the distinction is almost guaranteed not to 
make a difference.  So, you do have to learn about the distinction -- it
IS an extra conceptual load in teaching the language -- because you
MIGHT come upon a string-representation and need to be aware of
what will come up; but for almost all uses the distinction won't matter.

I think teaching this as alternate ways of *spelling* equal (but distinct: 
==, and essentially interchangeable in use, but NOT 'is') values will be
the least of evils.  The focus on spelling is sensible, because the string
representation (either str or repr or both -- I'm still not sure of how
Guido will decide that) makes up most of the difference.  There is, as
you say, a type distinction too -- but it's carefully designed to be as
irrelevant as possible... existing isinstance won't show that distinction,
only horrid type(...) == type(...) kinds of constructs will.


> In the past, untruth could be spelled in many ways (0, None, (), ...);

0 is (2.2 is still present, not yet past) the canonical way to spell 
untruth: it's what 'not', 'is', 'in', comparisons, and all standard library 
iswhatever's return -- including operator.truth, of course.

> with the PEP, you can say "untruth is spelled as False". You still
> need to explain that other things "count as false" also, but the
> proper spelling becomes "False" (whereas 0 served this position only
> by convention).

Exactly the same set of objects (plus the new object False, that's ==0
anyway and can be used just about wherever 0 can be) "count as
untrue".  The "by convention" status of the canonical spelling for untrue
still holds, except that you want to change the convention to the new
object which exists solely for the value of changing that canonical
spelling, more or less.  That canonical spelling is still the value returned
by 'not', 'is', 'in', and so forth -- exactly the same 'happenstance' (ha)
which gives its present canonical spelling its current status.


>> I've wasted too much of my life trying to explain to people why they must
>> NOT code "if some_condition==True:" or the equivalent thereof in various
>> languages, but just "if some_condition:".  The "== True" appendix,
>> useless at best, often damaging, seems one of the most natural errors by
>> newbies.
> 
> That is a problem. Perhaps comparing booleans and non-booleans for
> equality should cause a warning (for that matter, perhaps even
> comparing booleans and booleans should cause that warning...)

'if x == y:' should cause no warning, of course -- it's a general, perfectly
good polymorphic usage.  Why should the user have to silence warnings
or break the smoothness of polymorphism to ensure against x, or y, or
both being of this newfangled bool type?

> More realistically, the *parser* should emit a warning if it sees the
> name True used in such a context.

...assuming the parser can determine at parse time that True stands
for the built-in value at that point -- or do you want to give True an
exhalted semi-keyword status, beyond that of, say, None, a dignity that 
will produce frequent warnings if you rebind that identifier e.g. in a local
namespace?  I know 'slippery slope' arguments are disliked around
here, but both this exchange and the various voices raised in clamor
towards 'stronger, purer booleans' suggest to me that PEP 285's own
slope may be slippery indeed.


>> In Python I was blessedly spared that waste of teaching time.  If the PEP
>> is approved, I'll be back to wasting that time and effort (perhaps more,
>> given how seductive "is True" will appear to many newbies...!).
> 
> Fortunately, "is True" won't be much worse than "== True" (except when
> you have 1 on the other side).

No, but it's one more warning-case you have to propose (and how else
will one check for the wonderful new objects' identities if 'is True' 
warns?! by twists and turns of isinstance checking for bool then int()
and == checking for 1...?!).  And 'is True' is very newbie-alluring: a
true 'attractive nuisance' (in more than one meaning of the phrase).


>> And what happens to other forms of output, such as marshaling and
>> pickling -- does the bool or int nature of a 0 or 1 get preserved or
>> lost in a round trip?
> 
> That's a good point: booleans should round-trip when pickling. I guess
> it is a bug in the PEP that this hasn't been spelled out.

If they do round-trip, then, it seems to me, we get a 'gratuitous'
incompatibility of pickles produced by 2.3 wrt 2.2 programs trying to
read them -- one more annoyance (and not a very attractive one either,
this time) to weigh against the claimed benefits (miniscule when at
all present, it still seems to me).  And of course old pickles will give
0 and 1 so that much work will be needed to reconstruct them into
False and True when appropriate *IF* it matters at all if something
is True or 1.

And if it doesn't matter at all, then why bother introducing PEP 285
at all?  I suspect a mixture of objects, some recently loaded from old
pickles and not updated yet so their attributes are 0 and 1, some
newly built so their attributes are False and True, and some in a
half-way stage (recently loaded or deepcopied from old pickles, but
partly modified/updated) with some 'boolean' attributes actually at
0 or 1, others at False or True -- all of such soup will probably work
pretty well apart from utter display anomalies if some of the objects
are ever converted into string representations.  So _why_ pay the
price -- not a huge, world-racking one, but still one nevertheless?


Alex



From paulo.pinto at altitude.com  Thu Apr 11 08:43:11 2002
From: paulo.pinto at altitude.com (Paulo Pinto)
Date: Thu, 11 Apr 2002 12:43:11 GMT
Subject: 4Suite for Python 2.2
Message-ID: <zxft8.37$Mh5.1049@nreader1.kpnqwest.net>

Hello,

does anyone know when 4Suite for Python 2.2
would be available ?

I can only find support up to version 2.1 :(

Thanks in advance,
Paulo Pinto




From news.ASkwar at DigitalProjects.com  Tue Apr 23 18:53:31 2002
From: news.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Wed, 24 Apr 2002 00:53:31 +0200
Subject: Passing arguments to threading objects
References: <pan.2002.04.23.22.39.14.487844.11225@teich.garten.DigitalProjects.com> <aa4o6l$1ce$1@panix1.panix.com>
Message-ID: <pan.2002.04.23.22.53.31.124002.11225@teich.garten.DigitalProjects.com>

On Wed, 24 Apr 2002 00:45:41 +0200, Aahz wrote:

> Override __init__().  See my home page for more info.

Wow, that was a fast answer ;)

Thanks for the hint.  I hoped that I would not have to override the
constructor, but that's easy enough.
 
Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 17 hours 7 minutes


From cliechti at gmx.net  Sat Apr  6 16:31:06 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 22:31:06 +0100
Subject: global instance question (newbie)
References: <a8njqq$phs$1@wanadoo.fr>
Message-ID: <Xns91E8E54158554cliechtigmxnet@62.2.16.82>

"manouche" <emmanuel at onda-latina.com> wrote in
news:a8njqq$phs$1 at wanadoo.fr: 

> hello,
> 
> How can I create a global object that I can be used in all the files
> of my project?
> Using "global" I can only access to the instance in the file where it
> was created.

put it in a separate module and import it where you need it.
of course you can add and modify the values on that module.

there is only one module instance in the interpreter, no matter how much 
you import it (singelton is the keyword).

chris

-- 
Chris <cliechti at gmx.net>



From lac at strakt.com  Tue Apr  2 08:40:46 2002
From: lac at strakt.com (Laura Creighton)
Date: Tue, 02 Apr 2002 15:40:46 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Your message of "Sat, 30 Mar 2002 08:35:33 GMT."
             <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> 
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> 
Message-ID: <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>

Dear Guido:

I would first like to thank you for giving us an opportunity to respond.
I have spent most of the weekend thinking and writing a reply to this,
and I think that this has made me a better teacher.  For this I am
grateful.  I realise that this is rather long, but I have condensed
it substantially from my efforts earlier this weekend.   Thank you
for taking the time to read this.

Laura Creighton

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

I am opposed to the addition of the new proposed type on the grounds
that it will make Python harder to teach both to people who have never
programmed before and to people who _have_.  If they have no
preconceived idea of Booleans, then I do not propose to need to teach it
to them in an introductory lesson.  There is a time to learn symbolic
logic but while trying to learn how to program for the first time is not
it.  If, on the other hand, they _do_ have some preconceived idea of
Booleans then Python will not have what they want.  They will want 
stricter booleans that 'behave properly'.

Maybe some day we should give these people such a type.  People who
do symbolic logic and make push-down automatons all day long will love
it. If we implement this, not out of objects, but out of bits, and have
something really sparse in memory consumption, we will cause jubilation
in the NumPy community as well.  But I don't want to discuss this here,
I only want to discuss the new type which is proposed in this PEP.  And
I believe that this half-way thing, int-in-hat,  that is called bool 
but which does not implement what a mathematician would call truth values
will make it much harder for me to teach the Python language and programming
in general.  Not only do I not need, but I actively do not want this
change which is the integer 0 and the integer 1 that have some
odd printing properties.


        1) Should this PEP be accepted at all.

No.

        2) Should str(True) return "True" or "1": "1" might reduce
           backwards compatibility problems, but looks strange to me.
           (repr(True) would always return "True".)

Now we have a teaching problem.  If str(True) returns anything but
'True' then I am going to have to explain this to newbies, really early
on. I can't see myself claiming that 1 is the string representation of
True.  I can see myself explaining that int(True) is 1, or that bool(1)
is True.  If I say that the string representation of True is 1, then I
must assert that True is just a fancier, prettier way of writing 1.

But this breaks the more common practice where str is used for the
prettier way of writing things and repr is used for the uglier one.  And
I guarantee my students will notice this, especially if they have heard
me explain why their floating point numbers are not printed the way that
they expect. I don't want to have to explain that eval(repr(object)) is
supposed to generate the object whenever possible, for the last thing I
want newbies to be thinking about is eval().  I guess I am going to have
to say that it is a wart on the language, and that we have it this way
so as not to break too much exisiting code.

I think this wart is far uglier than the lack of a half-way-but-not-
quite Boolean Truth value.  But I am all for this wart rather than break
the exisiting code.

        3) Should the constants be called 'True' and 'False'
           (corresponding to None) or 'true' and 'false' (as in C++, Java
           and C99).

I would rather the constants, if we have them, be called anything other
than True or False.  It is not the things that you don't know that hurt
you the most in learning a language -- it is 'the things that you know
that ain't so'.  Learning C, from PL/1, my first problem was 'how do I
write a procedure'?  You see, I aleady _knew_ that _by definition_
functions returned values and procedures didn't, and so, since I didn't
want to return a value, I didn't want to write a function.  Learning
that your basic defintions are wrong generally requires a poke from the
outside.  The same thing happened to me when learning Python.  I knew,
by definition, that attributes were not callable.  Using getattr() to test
whether  a class had a method did not occur to me, despite pouring over
python docs for 2 days looking for such a thing.  Finally I went after
it out of the __class__ -- and immediately posted something to
python-list, saying 'This is ugly as sin.  What do real people do?'  I
never got around to questioning whether 'attributes by definition, may
not be a callable' in precisely the same way I never got around to
questioning whether 'you could write a function that did not return a
value'.

The problem is that _everybody_ has some conceptual understanding of
True and False.  And whatever that understanding is, it is unlikely to
be the one used by Python.  Python does not distinguish between True and
False -- Python makes the distinction between something and nothing.
And I think that this is one of the reasons why well-written Python
programs are so elegant.  And this is what I am trying to teach people.

So I out-and-out tell people this.  {} is a dictionary-shaped nothing.
[] is a list-shaped nothing. 0 is an integer-shaped nothing.
0.0 is a float shaped nothing.

I want to save them from the error of writing

if (bool(myDict) == True):

and if they start out believing that python only distinguishes between
Something and Nothing, they mostly are ok.  And to rub this point in you
can do this:

>>> False = 1
>>> True = 0
>>> if False:

...     print 'Surprise!'
...
Surprise!

>>> if True:

...     print 'I am True'
... else:
...     print 'Surprise Again!'
...
Surprise Again!

This is a very nice eye-opener.  It is a true joy.  Watch the minds go
*pop*! and the preconceived notions disappear. (You then let them know 
exactly what you will think of them if they ever do this 'for real', of
course.)

        4) Should we strive to eliminate non-Boolean operations on bools
           in the future, through suitable warnings, so that e.g. True+1
           would eventually (e.g. in Python 3000 be illegal).  Personally,
           I think we shouldn't; 28+isleap(y) seems totally reasonable to
           me.

This is not an argument for allowing non-Boolean operations on bools();
this is an argument for not writing functions that return Booleans.  Make
them return numbers instead, so that you can use them as you did.  Last
month we discussed why in 1712 February had 30 days in Sweden.  (See:

http://groups.google.com/groups?q=leap+year+Sweden+group:
comp.lang.python.*&hl=en&selm=Xns91D08815184B2svenaxelssonbokochwe%
40212.37.1.234&rnum=1

if you care, and you missed it.)

I live in Sweden.  Assigning students the problem of calculating whether
or not a given year is a leap year in Sweden appeals to me. 

But I know students.  I guarantee that I will get an isleap that returns 
a True or a False under the proposed new regime.  And this is precisely
what I do not want, which I will try to teach by assigning, next week,
a program that calculates how many days are in a given year.  I predict
that I will get a lot of answers like this:

if year == 1712:
    days = 367
elif isSwedishLeap(year):
    days = 366
else:
    days = 365

There are many things wrong with this code.  The bizzare special case of
1712 belongs in the SwedishLeap function, with the rest of the
weirdnesses. Thus I will have to convince my students that it is better
to write a function that does not return True or False.  And this is
despite the fact that I originally asked for 'is Y a leap year or not'.

This will be a necessary exercise.  If True and False are in the
language, I am going to have to work especially hard to teach my
students that you (mostly) shouldn't use the fool things.  There is
almost always some value that you would like to return instead.

Having renamed SwedishLeap and fixed it to return how many leap days,
instead of a bool, I have now made a different problem for myself and 
my students.

The new improved solution for 'is this year a leap year' will be:

if bool(SwedishLeap(year)) == True:  
    # the better students will say 'is' instead of '=='
    print 'yes.'
else:
    print 'no.'

Aaargh!
I already see too much code like this.  It's mostly written by people
who come from other languages.  They define their own True and False so
they can do this.  (And they mostly have an extra set of () as well).
Right now I have the perfect fix for this.  I just say 'Python does not
care about True and False, only nothing or something'.  You have just
stolen my great weapon.

What am I going to say?

attempt 1.

Python pretends to have bools, but they are just ints in fancy hats.  So
you are making more comparisons than are necessary.

smart student:

But you said it is better to be explicit than implicit!  And here I am
explicitly performing the type coercion rather than let it happen
implicitly! (or PyChecker _warned_ me that I was making an implicit 
conversion!) I put in the cast so that people will know exactly what
is happening!

attempt 2.

But they are _really_ ints 'under the hood'.  I was not kidding about the
fancy hats!  if SwedishLeap(year): is precisely what you want.  You
don't want to test against True at all!

smart student:

Comparisons yield boolean values.  Therefore they _want_ a Boolean
value.  You are just being lazy because it saves typing.  In the bad old
days before we had Booleans this was ok, but now that we have them we
should use them!  Otherwise what good are they?  What should I be using
them for if not for this?

attempt 3.

Got an hour? I'd like to explain signature-based polymorphism to you ...

smart student:

ha! ha! ha!

        5) Should operator.truth(x) return an int or a bool.  Tim Peters
           believes it should return an int because it's been documented
           as such.  I think it should return a bool; most other standard
           predicates (e.g. issubtype()) have also been documented as
           returning 0 or 1, and it's obvious that we want to change those
           to return a bool.

I think that operator.truth(x) should return an int because about the
only thing I use it for is operator.truth(myDict).  I really want the
integer value.  How do you propose I get it if you change things?  Why
make me go through the gyrations of int(bool(myDict))?

By the way, look at the list of those things that would be changed
to return a bool.  Most of them are python implementations of ANCIENT
C functions.  They date from the time before we invented exceptions!
The primitive old days when we had to test for every possible time
you wouldn't want to run your code before you actually got to run it.

I don't want to go back to the days of
					
if aflag or (bflag and (cflag or dflag)):

either.

And all Truth testing (unless you are doing symbolic logic) reeks
this way to me, of the old style I am trying to stamp out.  Truth
testing is just another form of type testing, and just as ugly.

    Rationale

        Most languages eventually grow a Boolean type; even C99 (the new
        and improved C standard, not yet widely adopted) has one.

        Many programmers apparently feel the need for a Boolean type; most
        Python documentation contains a bit of an apology for the absence
        of a Boolean type.

So fix the docs, don't change the code! <wink>.  I think the fact that
in python control flow structures distinguish between Something and
Nothing is one of the beauties and glories of the language, and you
should delete any documentation that says otherwise.

Under the proposed new scheme you will have to trade apologies for the
lack of bools, for apologies for not producing real bools, only this
int-in-a-new-hat hack that pretends to be a bool.  This is hardly
progress.

        I've seen lots of modules that defined
        constants "False=0" and "True=1" (or similar) at the top and used
        those.  The problem with this is that everybody does it
        differently.  For example, should you use "FALSE", "false",
        "False", "F" or even "f"?  And should false be the value zero or
        None, or perhaps a truth value of a different type that will print
        as "true" or "false"?  Adding a standard bool type to the language
        resolves those issues.

So would adding True and False to the __builtins__, and probably
operator.truth as well, and then modifying PEP 8 saying to use the
things if you actually have a need for True and False.  Then you could
also get a much needed word in edgewise discouraging

if bool(x) == True:

or actually using True and False much, because there is usually a better
more pythonic way to do what people used to other languages are accustomed
to doing with booleans.  This is precisely what some people have said
here: 'When I started using Python, I made True and False, but once I
stopped trying to program in some other language using python, I 
stopped needing these things'.  (see the recent post by Don Garrett
<garrett at bgb-consulting.com> in this thread for an example.)

This is what I have observed as well.  And I fear if you add these new
types to the langauge people will never take this step.  The existence 
of the types in the language will discourage them from thinking that using
True and False all the time is not pythonic.  It is nice that people are
puzzled, wondering how all those python programmers live without a
boolean type.  Eventually they puzzle it out.  This is not a bug, but
a feature <wink>.

        Some external libraries (like databases and RPC packages) need to
        be able to distinguish between Boolean and integral values, and
        while it's usually possible to craft a solution, it would be
        easier if the language offered a standard Boolean type.

I'm one of the people who build interfaces to databases that need to
distinguish this.  For what it's worth, can you please not add this
feature to the language?  Don't do it for me ....

        The standard bool type can also serve as a way to force a value to
        be interpreted as a Boolean, which can be used to normalize
        Boolean values.  Writing bool(x) is much clearer than "not not x"
        and much more concise than

            if x:
                return 1
            else:
                return 0

Conciseness for its own sake is no virtue.

        Here are some arguments derived from teaching Python.  When
        showing people comparison operators etc. in the interactive shell,

        I think this is a bit ugly:
            >>> a = 13
            >>> b = 12
            >>> a > b

            1

        If this was:
            >>> a > b

            True

        it would require one millisecond less thinking each time a 0 or 1
        was printed.

This is the basis of our disagreement.  I think that it is very, very
important that much more than a millisecond be spent on this.  This is a
fundamental python concept, which I want to teach.  This is precisely
where you learn that python distinguishes between Something and Nothing,
and if you have a problem seeing why this implies  a > b printing as 1,
then you probably have a problem with the whole concept.  And making it
return True is precisely what I never, ever, ever want a python learner
to see.

People who come from staticly declared languages have a terrible burden
to overcome when they meet python.  They are not used to the fact that
everything is an object.  They have rigid barriers in their heads
between 'control statements' and 'data'.  This is precisely where such
conceptual barriers first begin to crumble.  This is precisely the
experience I _want_ my students to have.  It is precisely how I train
people to give up their old ideas.  You have just made my teaching job
harder, not easier.

I've never had any trouble teaching anybody that if 1: means do it, and
if 0: means don't.  Ever.  Who is it that has had such trouble?  I fear
they may be new to teaching and are confusing 'this is taking the
students a while to learn because it is something new that they have
never seen before' with 'this is taking the students a while to learn
because the language is broken'.  Learning that python distinguishes
between Something and Nothing is a completely new idea that you are
giving these people, outside of their experience.  This is going to take
a while to get used to.  But it is going to take _longer_ to get used
to, if there is all this confusing stuff about symbolic logic, truth
values, George Boole, and why the math majors in the class are all
snickering and saying 'Python is a sucky language.  Its implementation
of booleans is so, so, lame...' thrown in as well.

If you already know what a boolean is, then chances are Python's bools
are not going to behave the way you expect them to.  If you don't know
what a bool is, then I feel morally obligated to teach you the
difference between real truth values and this int-in-a-hat kludge.  In
either case, I now have to spend a bunch of time teaching about
booleans, something I had no desire to do before today. I _want_ to teach 
that if 1: means do it and if 0: means don't.  I want to teach that python
makes a distinction between Something and Nothing. And I can teach that a
_lot_ faster than An Introduction to Boolean Algebra ...

        There's also the issue (which I've seen puzzling even experienced
        Pythonistas who had been away from the language for a while) 
        that if you see:

            >>> cmp(a, b)
            1
            >>> cmp(a, a)
            0

        you might be tempted to believe that cmp() also returned a truth
        value.  If ints are not (normally) used for Booleans results, this
        would stand out much more clearly as something completely
        different.

I don't think that people are confused about this because they think 
that cmp is returning True or False.  I've had people go on believing 
that cmp should return one of 2 values even as I was telling them that
it returned one of 3.  The desire for cmp to be a two-valued comparison 
runs deep in many souls, and is tied to a passionate belief that comparisons 
are binary, not that they return True and False.

        I don't see this as a problem, and I don't want evolve the
        language in this direction either; I don't believe that a stricter
        interpretation of "Booleanness" makes the language any clearer.

I think that a really strict boolean might be nice to have.  And if you
ever write one, you will curse the day you let this hack be named bool.
Replacing this int-with-a-hat with real bools will break so much
code. 

        Other languages (C99, C++, Java) name the constants "false" and
        "true", in all lowercase.  In Python, I prefer to stick with the
        example set by the existing built-in constants, which all use
        CapitalizedWords: None, Ellipsis, NotImplemented (as well as all
        built-in exceptions).  Python's built-in module uses all lowercase
        for functions and types only.  But I'm willing to consider the
        lowercase alternatives if enough people think it looks better.

I'd really like something other than True and False altogether.
Something Nothing or Empty Full or Yin Yang, __anything__ which people
will not believe they already understand.  It is so much easier to teach
something which people know is brand new rather than something which is
brand new but looks like something subtly different that people already
believe they know.

I have this problem a lot.  It is hard to teach people what floating
point numbers are because their grade school teachers have taught them
what fixed point decimals are so well (except for the name).  If
floating point numbers were traditionally printed xxx#yyy instead of
xxx.yyy I do not think that I would have such difficulties.  In any case
the people who do not understand floating point would be aware that
there is something that they do not understand, rather than blinding 
going off an using them to represent money.

I don't want to have to teach the pythonic meaning of True and False to
people who already believe they know what True and False is.  This is
going to be hell.  It is about as hard a thing as there is in teaching,
getting around people's previous conceptions.

        It has been suggested that, in order to satisfy user expectations,
        for every x that is considered true in a Boolean context, the
        expression x == True should be true, and likewise if x is
        considered false, x == False should be true.  This is of course
        impossible; it would mean that e.g. 6 == True and 7 == True, from
        which one could infer 6 == 7.  Similarly, [] == False == None
        would be true, and one could infer [] == None, which is not the
        case.  I'm not sure where this suggestion came from; it was made
        several times during the first review period.  For truth testing
        of a value, one should use "if", e.g. "if x: print 'Yes'", not
        comparison to a truth value; "if x == True: print 'Yes'" is not
        only wrong, it is also strangely redundant.

This suggestion came from somebody's preconceived idea of 'what is True'
and 'what does it mean for something to be True'.  Now that you have
posted this to python-list, you have found a whole lot more.  Have you
ever had such a response to a PEP?  Everybody thinks that they know what
True and False means, and that Python should do it his or her own way.
This is what everybody's classroom looks like as well.  And this is
why sane teachers do not want to discuss the meaning of True if they
can help it.

I want to keep python out of the True and False business.  Python cares
about whether a value is Something or Nothing.  This is beautiful, and
_better_ than what the other languages do.

Once again, Thank you for reading this and giving me a chance to write up my
objections.  You have made me a better teacher because of this.

Laura Creighton



From news at recombinant.demon.co.uk  Sat Apr 13 08:21:08 2002
From: news at recombinant.demon.co.uk (Stephen D Evans)
Date: Sat, 13 Apr 2002 13:21:08 +0100
Subject: Python 2.2 unwise.exe
References: <ZkjnuIABLAu8EwYC@jessikat.fsnet.co.uk>
Message-ID: <1018700471.13557.0.nnrp-01.d4e54907@news.demon.co.uk>

I tried to uninstall 2.2 for Win32 on NT4 and had the same problems. I
installed 2.2.1 over the top of the existing 2.2. It works.

I had no problems uninstalling Python 2.2 on Windows 2000.

Stephen D Evans


"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message
news:ZkjnuIABLAu8EwYC at jessikat.fsnet.co.uk...
> Anyone else had a problem un-installing 2.2 for win32? When I tried to
> un-install 2.2 prior to installing 2.2.1 I got an error saying it had
> already been un-installed.
>
> Tried the link in start menu and it said no unwise.exe was there.
>
> Luckily I had a copy of unwise.exe on another machine. Copied into the
> right place and it seems to have done the trick.
> -Flat cats are not better than nested birds-ly yrs.-
> Robin Becker




From aahz at pythoncraft.com  Tue Apr  9 18:25:41 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 18:25:41 -0400
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <a8vpp5$67d$1@panix1.panix.com>

In article <yu994rikfzh5.fsf at europa.research.att.com>,
Andrew Koenig  <ark at research.att.com> wrote:
>
>I would like to be able to write an expression that yields an
>object that I can subsequently use to rebind a name that I mention
>only in that expression.  How do I do it?
>
>Here's a function definition:
>
>        def f(v):
>                global x
>                x = v
>
>Now, f is an object that I can use to rebind the name x.  For
>example, if I execute f(42), that sets x to 42.

You really don't want to do this in Python.  Your best bet is to pass
around mutable objects (namely class instances and modules) and use
setattr().  You could be grotesque and also permit lists and
dictionaries:

    def updateObject(obj, key, value):
        try:
            obj.setattr(key, value)
        except AttributeError:
            obj.setitem(key, value)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From cliechti at gmx.net  Wed Apr 24 16:24:10 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Apr 2002 22:24:10 +0200
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6q9k$sf6$1@peabody.colorado.edu>
Message-ID: <Xns91FAE4234656cliechtigmxnet@62.2.16.82>

Fernando P?rez <fperez528 at yahoo.com> wrote in
news:aa6q9k$sf6$1 at peabody.colorado.edu: 
> Rad wrote:
> 
>> I am preparing myself to work on extracting data from 4 text files
>> (fixed width format) which combined size is about 80GB.  Considering
> 
> May I suggest that you also pick up some basic (4 hrs worth) grep/awk 
> knowledge? Sometimes I find that for quickly extracting a few tagged
> data fields from a large file, awk is much faster than python. I'll
> make the awk call in my python code and then keep operating for the
> more complex stuff in python.

i agree with that. i use awk to extract records from 4GB and larger files 
an then do the post processing in python.
it's hard to beat awk's speed but such a pre-processing is only useful if 
you can reduce the dataset. if you can only drop away 20% with post- 
processing it might have been faster to process all data with python.

note that awk is not very suitable for binary file but very powerful on 
text files. you should realy get a data sample file to play with.

chris


-- 
Chris <cliechti at gmx.net>



From ianb at colorstudy.com  Mon Apr 15 14:58:11 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 15 Apr 2002 13:58:11 -0500
Subject: Deprecate input()
In-Reply-To: <slrnablens.5k9.mlh@vier.idi.ntnu.no>
References: <a9cv63$qhu$1@bob.news.rcn.net> 
	<slrnablens.5k9.mlh@vier.idi.ntnu.no>
Message-ID: <1018897091.1149.281.camel@localhost>

On Mon, 2002-04-15 at 06:35, Magnus Lie Hetland wrote:
> But (in the (very?) long run) I think that the name "input" should
> then be used for what is now raw_input. I know -- breakage, and all
> that... I just think it fairly odd to have *only* raw_input, since the
> "raw" qualification then becomes fairly meaningless.

Why not just make input into what it should be, a moderately-evaluating
raw_input... like:

def input(prompt=''):
    val = raw_input(prompt)
    for converter in [int, float, complex]:
        try:
            return converter(val)
        except ValueError:
            pass
    return val


Of course, it gets more complex if you allow lists, dictionaries, and
tuples... which you probably should.  But that's certainly doable while
still being safe.

  Ian






From cliechti at gmx.net  Mon Apr  1 15:15:34 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Apr 2002 21:15:34 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Xns91E3D86E3841Dcliechtigmxnet@62.2.16.82>

martin at v.loewis.de (Martin v. Loewis) wrote in
news:m3elhz8sbs.fsf at mira.informatik.hu-berlin.de: 
>> I've wasted too much of my life trying to explain to people why they
>> must NOT code "if some_condition==True:" or the equivalent thereof in
>> various languages, but just "if some_condition:".  The "== True"
>> appendix, useless at best, often damaging, seems one of the most
>> natural errors by newbies. 
> 
> That is a problem. Perhaps comparing booleans and non-booleans for
> equality should cause a warning (for that matter, perhaps even
> comparing booleans and booleans should cause that warning...)
> 
> More realistically, the *parser* should emit a warning if it sees the
> name True used in such a context.

then it should warn for False too, cause 'var == False' i.e. '{} == False' 
is evil too.
i think thats more of a problem for newbees than the == True redundacy in 
an 'if'. they will use the form above instead of the correct 'not var'.

chris


-- 
Chris <cliechti at gmx.net>



From marklists at mceahern.com  Wed Apr  3 13:14:27 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 12:14:27 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.33.0204031118110.16190-100000@server1.lfw.org>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEAEKJAA.marklists@mceahern.com>

[Ka-Ping Yee]
> Consider this:
>
>     >>> a = [0, False, 1, True]
>
>     >>> print a
>     [0, 0, 1, 1]
>
>     >>> for x in a: print x
>     0
>     0
>     1
>     1

Actually, if you use the sample code provided by the PEP and run this you
get:

  >>> a = [0, False, 1, True]
  >>> print a
  [0, False, 1, True]

  >>> for x in a: print x
  0
  False
  1
  True





From kseehof at neuralintegrator.com  Sun Apr 14 00:36:41 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 13 Apr 2002 21:36:41 -0700
Subject: copy protection
Message-ID: <EFEFIHAINIINNELLHFAIMEBKCBAA.kseehof@neuralintegrator.com>

Does anyone know of anyone who provides copy protection services for python?
What I'm looking for is something based on bytecode obfuscation.

Thanks,

Ken Seehof





From dsavitsk at e-coli.net  Fri Apr 26 17:32:18 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 26 Apr 2002 21:32:18 GMT
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <CHjy8.9301$d7.3015094@newssrv26.news.prodigy.com>

my search of sourceforge turned this up
http://pypedit.sourceforge.net/

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:aacdeg$7g$1 at nets3.rz.RWTH-Aachen.DE...
> Is there a simple, open source text-editor project available in Python?
>
>




From ods at fep.ru  Thu Apr 11 11:52:46 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 11 Apr 2002 19:52:46 +0400 (MSD)
Subject: ANN: Logging Module v0.4.2 released
In-Reply-To: <2e37dc1.0204110722.311fe4b0@posting.google.com>
Message-ID: <Pine.LNX.4.44L.0204111947391.13708-100000@ods.pp.ru>

On 11 Apr 2002, Vinay Sajip wrote:

VS> A new version of the proposed Python standard logging module
VS> (as per
VS> PEP 282) has been released. You can get all the information

I've tried to use logging in a module used by CGI script and ...
it now dies with misterious message: unexpected end of header.
The reason is the message 'No handlers could be found for logger
"qq"' that goes to stdout.  I guess logging should not lo
messages by default and certanly it must not send warnings to
stdout.





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 11:08:09 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 08:08:09 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <dnbe9a.jfl.ln@cvs>
Message-ID: <3cbaed2b@news.mhogaming.com>

"Max Ischenko" <max at malva.com.uaREMOVE.IT> wrote in message
news:dnbe9a.jfl.ln at cvs...
|
|  Ken Peek wrote:
|
| > 5) These two types (int and long) _CAN_ be unified for
| > _ALL_ operators and methods-- I know they can, because the
| > language known as "Ruby" does a very good job of it.  In
| > Ruby, there is simply no difference (at least visibly) to
| > the programmer _at_all_.  The 2 types are automatically
| > converted back and forth without the knowledge or consent
| > of the programmer.  If Ruby does it, so can Python.
|
| Not quite:
| irb(main):001:0> 1222.class
| Fixnum
| irb(main):002:0> 12232333333333332.class
| Bignum
|
|
| For me, Common Lisp's numeric types look like a good source of
| inspiration.
|
|
| --
| BOFH excuse #71:
| The file system is full of it

I'm not suggesting that you shouldn't be able to peek inside an object and find
out the internal representation.  But-- rather-- in common use, it should be
invisible.  You give a poor example, because you are deliberately ASKING Ruby to
give you the type, (which it does dutifully, and correctly.)





From djc at object-craft.com.au  Mon Apr  1 23:34:04 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 02 Apr 2002 14:34:04 +1000
Subject: Python embedded like PHP
References: <mailman.1015988122.18668.python-list@python.org> <359f85cd.0203191203.2a8e4bd@posting.google.com> <mailman.1016619871.27287.python-list@python.org> <7xk7s7mpiy.fsf@ruckus.brouhaha.com> <mailman.1016626151.9950.python-list@python.org> <slrna9haud.e6b.akuchlin@ute.mems-exchange.org> <slrnaa3bi8.lam.jon+usenet@snowy.squish.net> <23891c90.0203290655.25bbcab9@posting.google.com> <slrnaacob1.bu2.jon+usenet@snowy.squish.net> <23891c90.0203310820.60696fb1@posting.google.com>
Message-ID: <m3pu1ivhvn.fsf@ferret.object-craft.com.au>

>>>>> "Paul" == Paul Boddie <paul at boddie.net> writes:

Paul> The next problem, though, is coordinating the output with the
Paul> input.  One can invent a "custom tag" (or corresponding feature
Paul> in one's framework of choice) which emits an HTML "input" field,
Paul> but where is the code that checks the value returned by the user
Paul> through that field, and how can we be sure that such code is
Paul> validating and manipulating the same kind of data as that
Paul> described by the "input" field? For example, I could invent a
Paul> "custom tag" like this:

Paul>   <myapp:specialInput name="someField"/>

Paul> This could result in the following HTML being generated:

Paul>   <input name="someField" size="30" type="text">

Paul> The intention might be that the contents of "someField" are in a
Paul> particular format. Moreover, such data may be stored in a
Paul> database system, and we might want to keep the "size" or
Paul> "length" of a piece of data synchronized with the data type used
Paul> in the database - it's not uncommon for such things to evolve as
Paul> the requirements of an application change, but it would be very
Paul> inconvenient to need to change HTML form descriptions,
Paul> validation code, template system code, all in a number of
Paul> different places.

Paul> Template systems typically don't address these issues, whereas
Paul> stand-alone validation systems address the nature of the input
Paul> data but not how such data is "described" to the browser (and
Paul> thus to the user). A coordinated approach which unifies input,
Paul> output, and validation seems to be the logical next step in any
Paul> significantly big or complex application.

Albatross has features which come close to what you are requesting.
With a little bit of sweat you could make it do exactly what you are
talking about.

The built in behaviour for forms creates a Python list of all input
fields in a form and then pickles (and MD5 signs) that to a hidden
field in the form.  When the browser request comes back the list is
restored and used for the basis of merging the browser request into
the "execution context".  What this currently gives you is that any
field which is not filled in and submitted by the browser will be
detected and the corresponding variable will be set to None.

The standard behaviour is performed by a combination of the <al-input>
tag and the NameRecorderMixin class.

There are two ways that you could change the current Albatross
behaviour.  You could implement your own special input tag to
construct your special input tags with a special form recorder, or you
could subclass one of the Application classes and override the
validate_request() method.

1) A custom tag is pretty easy to build - check this out:

   http://www.object-craft.com.au/projects/albatross/albatross/cust-ref.html

   By implementing your own RecorderMixin you could store the field
   type with the name for automatic validation on request merging.

   This is the documentation for the default recorder behaviour as
   implemented by NameRecorderMixin:

   http://www.object-craft.com.au/projects/albatross/albatross/mixin-rec-name.html

   To complete the loop you would then subclass one of the application
   classes which inherit from Application and override the
   merge_request() method which takes fields out of the browser
   request in place them into the execution context.

   http://www.object-craft.com.au/projects/albatross/albatross/pack-app.html

2) This is the easy but less general option.  You could subclass an
   application object and provide a validate_request() method which
   field names with embedded type information (hungarian notation) to
   determine the type of the field.  By returning FALSE from
   validate_request() you suppress request processing.

   As part of the field validation in your validate_request() you
   could set error messages which were picked up as part of the form
   display.

At some stage we are going to start building up a cookbook for
Albatross.  This is a good cookbook example.

- Dave

-- 
http://www.object-craft.com.au


From merman at freenet.de  Mon Apr 15 18:00:19 2002
From: merman at freenet.de (merman)
Date: Tue, 16 Apr 2002 00:00:19 +0200
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org> <3cbb355a$0$166$9b622d9e@news.freenet.de> <KwHu8.23421$vF6.768479@news2.tin.it>
Message-ID: <3cbb4b3d$0$146$9b622d9e@news.freenet.de>

"Alex Martelli" <aleax at aleax.it> schrieb im Newsbeitrag
news:KwHu8.23421$vF6.768479 at news2.tin.it...
> merman wrote:
>         ...
> > Ok Brian, and how can I write a disutils setup script? I don't know
that!
>
> Try http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66509 --
> the last part of the recipe gives a setup.py script that should be easy to
> adapt to your needs, I hope.
>
>
> Alex

THANX.

o-o

thomas




From tomchance at gmx.net  Tue Apr 30 10:46:03 2002
From: tomchance at gmx.net (Tom Chance)
Date: Tue, 30 Apr 2002 14:46:03 +0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net> <23891c90.0204300033.16549c52@posting.google.com>
Message-ID: <3CCEAE2B.8000300@gmx.net>

Sorry if I'm coming in at a tangent to this conversation, or repeating 
something that's been said before, but I've been using Perl and the 
HTML::Template module for ages now and it seems to me to be the natural 
way to do things. You basically create a series of HTML templates, and 
put fake HTML tags into them like:

<TMPL_VAR NAME=TITLE>

And then in your perl scripts you call a new HTML::Template object, give 
  it the values to put into all of the fake HTML tags, and tell 
HTML::Template to print out the page, like so:

use HTML::Template;
my $template = HTML::Template->new(filename => 'template.html');
$template->param(TITLE => 'Hello there!');
print $template->output();

This lets you keep the HTML seperate from the logic of the script 
(something which I find preferable to PHP's system, especially if the 
script is vey complicated), which is handy both if you have seperate 
people doing the HTML and the Perl code, and if like me you don't want 
to be fiddling around with HTML every time you want to edit the logic of 
the script. It lends to very neatly laid out HTML and Perl. Perhaps 
Python merely needs an equivalent of this?

Tom



From python at rcn.com  Fri Apr 12 23:58:21 2002
From: python at rcn.com (Raymond Hettinger)
Date: Fri, 12 Apr 2002 23:58:21 -0400
Subject: try/except/else flexibility...
References: <slrnabcon3.g5r.mlh@vier.idi.ntnu.no>
Message-ID: <a98aek$32h$1@bob.news.rcn.net>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnabcon3.g5r.mlh at vier.idi.ntnu.no...
> Just a thought: Given the prevalence of try/except/else in the "don't
> look before you leap" way of doing things, could it be useful to make
> the statement more flexible? For instance, instead of writing
>
>   try: foo + ''
>   except: pass
>   else: bar()
>
> (assuming I'm sloppy enough not to specify TypeError on the second
> line) one could simply write:
>
>   try: foo + ''
>   else: bar()

Hmm, trading explicitness for brevity and flexibility.  How about
meeting you halfway:

ignore TypeError:  foo + ''
else:  bar()

>
> and if I want something done only if it causes no exceptions, for
> instance pop off an element of a list unless it is empty, I could just
> do:
>
>   try: somelist.pop()

ignore IndexError:  somelist.pop()



Raymond Hettinger




From aleax at aleax.it  Wed Apr 17 08:20:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 12:20:58 GMT
Subject: Safe eval?
References: <Xns91F2ED00B6997cliechtigmxnet@62.2.16.82> <a9jjc5$9g6$1@thorium.cix.co.uk>
Message-ID: <KMdv8.22884$b62.642941@news1.tin.it>

gbreed at cix.compulink.co.uk wrote:

> Chris Liechti wrote:
> 
>> you can disable the builtin functions you want and populate the
>> namespace with save functions and classes (see Bastion for that)
> 
> I've never worked out how to do the second bit.  Bastion you say?  The
> documentation for that's all about restricting access to classes.  How do
> I get such a class into the namespace for an reval?  Every time this comes
> up, somebody says "use reval" but nothing about how you supply locals to

Nice overview at:
http://py-howto.sourceforge.net/txt/rexec.txt

The key point you ask about:

"""
module = r_env.add_module(module_name)
mod_dict = module.__dict__    Despite its name, the add_module() method 
actually only adds the
   module if it doesn't already exist; it returns the corresponding
   module object, whether or not the module had to be created.
  
   Most commonly, you'll insert variable bindings into the __main__ or
   __builtins__ module, so these will be the most frequent values of
   module_name.
  
   Once you have the module's dictionary, you need only insert a
   key/value pair for the desired variable name and value. For example,
   to add a username variable:
  
mod_dict['username'] = "Kate Bush"    Restricted code will then have access 
to this variable.
"""


Alex



From borcis at geneva-link.ch  Mon Apr 15 07:47:06 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 15 Apr 2002 13:47:06 +0200
Subject: Does Python need an 'xor' operator?
References: <mailman.1018819588.582.python-list@python.org>
Message-ID: <3CBABDBA.5010807@geneva-link.ch>


Tim Peters wrote:

> I doubt that a new "xor" operator will ever get in (compelling use cases are
> conspicuous by absence), but anyone pushing for it had better push for a new
> "eqv" operator too (that is, if you're going to claim that it's ugly that
> x^y may not return one of its argument, surely it's just as ugly that the
> same holds for x==y).


Indeed, x == y should behave like (x == y) and (x or not y) !

Boris
--
assert 304 in 340343, ParadiseMismatchError



From sholden at holdenweb.com  Thu Apr 11 17:57:53 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 11 Apr 2002 17:57:53 -0400
Subject: Sound programming?
References: <3CB50DBC.1B0E533F@alum.mit.edu>
Message-ID: <XHnt8.5263$F43.3946@atlpnn01.usenetserver.com>

"Will Ware" <wware at alum.mit.edu> wrote in message
news:3CB50DBC.1B0E533F at alum.mit.edu...
> Somebody was asking me this evening if there's an easy way to
> play a tone of a specified frequency and duration in Python.
> I didn't know off the top of my head. I've since poked around
> the web a bit and it appears like the Snack Sound Toolkit
> (http://www.speech.kth.se/snack/) might be a good way to go.
> One of its virtues is being cross-platform, which I suspect is
> among his unstated wishes. Are there any other options?

Not cross-platform, but a real simple way is available on Windows:
win32api.Beep()

>>> import win32api
>>> win32api.Beep(440, 120)
>>> win32api.Beep(440, 1200)
>>> win32api.Beep(4400, 1200)

regards
 Steve






From ajs at ix.netcom.com  Sun Apr  7 05:08:28 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 7 Apr 2002 05:08:28 -0400
Subject: ACCEPTED: PEP 285
Message-ID: <000001c1de14$29f94a00$0334fea9@carol>

>I don't buy that stability is an absolute good; neither new >features;
overall I've been very happy with Guido's balancing >of these tradeoffs in
its effects on my own code.

Surprisingly perhaps, considering my rep here as an anti,
so have I.

And most changes -  some much, much more fundamental than
the div operator or bool changes - have passed into circulation
extremely quietly.

But there has been some really process problems and a sense
of being strong armed on those certain changes.

Guido asks "should the PEP be accepted"?

Many folks reacted that - no, it is unnecessary and
poorly timed and took the time to elaborate as intelligently
and thoroughly as they could the basis for that conclusion.

Only to find that the question Guido was *really* asking was
whether anyone saw a *killer* reason not to adopt.

Unnecessary and poorly timed may well be the right answer
but is not a  *killer* reason. You lose.

The fact as to who's football we happening to be playing
with now being a little shoved in our face.  Guido seemed
willing to accept the sense of the community, along as the
community told him what he wanted to hear. When
it didn't, he simply retroactively rephrased his question.

A reaction should be no surprise.

Words matter.

And straight is better than crooked.

Art









From vinay_sajip at yahoo.co.uk  Sun Apr 14 10:23:40 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 14 Apr 2002 07:23:40 -0700
Subject: ANN: Logging Module v0.4.3 released
Message-ID: <2e37dc1.0204140623.5e5923d2@posting.google.com>

A new version of the proposed Python standard logging module (as per
PEP 282) has been released. You can get all the information from

http://www.red-dove.com/python_logging.html

There are "download" and "recent changes" links at the top of that
page. The new stuff includes thread-safety additions and the ability
to change the logging configuration of an application while it is
running! An example is provided in the distribution.

As always, your feedback is most welcome (especially bug reports,
patches and suggestions for improvement).

Cheers

Vinay Sajip
Red Dove Consultants Ltd.

Changes since the last version:
=================================
Bug fix one-off error message to go to sys.stderr rather than
sys.stdout.
logrecv.py fix TCP for busy network.
Thread safety - added locking to Handler and for shared data in
module, and log_test16.py to test it.
Added socket listener to allow on-the-fly configuration and added
log_test17.py to test it.


From tismer at tismer.com  Tue Apr 30 09:09:02 2002
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 30 Apr 2002 15:09:02 +0200
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org> <mailman.1020076925.25849.python-list@python.org> <b0f083db.0204291328.e22664e@posting.google.com>
Message-ID: <3CCE976E.9030908@tismer.com>

Jeremy Hylton wrote:
> Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1020076925.25849.python-list at python.org>...
> 
>>What we are talking about is a way to support PyZine,
>>while at the same time providing a missing structure
>>for Starship.
> 
> 
> Why are these two connected?
 > Couldn't people support PyZine by subscribing?

I connected them. By the idea to support one through the other.

-chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From fbartletFIXIT at optonline.net  Thu Apr 18 21:24:40 2002
From: fbartletFIXIT at optonline.net (Frederick H. Bartlett)
Date: Fri, 19 Apr 2002 01:24:40 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
Message-ID: <3CBF71D4.5668CF1A@optonline.net>

I was recently introduced to OmniMark. One of our exercises was to take
a plain text file of Hamlet and convert it to SGML.

So I did it in Python, too. But the best time I could get from Python
was .57 sec, while OmniMark came in at .20 sec. What's the most
efficient technique for Pythonesque regex-based text processing?

My best time came from using a single rather large regex and findall; I
also tried smaller regexes and scan and match.

Has anyone else compared OmniMark and Python?

Thanks, all.

Fred


From aahz at pythoncraft.com  Sat Apr 13 13:38:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Apr 2002 13:38:01 -0400
Subject: Thread-safe sockets?
References: <slrnabg28s.g9m.mlh@vier.idi.ntnu.no>
Message-ID: <a99qdp$ao$1@panix1.panix.com>

In article <slrnabg28s.g9m.mlh at vier.idi.ntnu.no>,
Magnus Lie Hetland <mlh at vier.idi.ntnu.no> wrote:
>
>I'm working with a ThreadedTCPServer and the following question came
>up: Is it safe for two different threads to use the
>StreamRequestHandler's rfile and wfile simultaneously? (I.e. one
>thread reads and another writes -- no simultaneous reading or
>writing.) These file-like objects are created with the socket.makefile
>method.
>
>Since the two operations seem so unrelated, it would also seem that
>this is safe -- but I'm worried that there may be (perhaps
>paltform-dependent) thread-unsafe interference between the two at some
>lower level. Does anyone know?

I don't know -- but I wouldn't do it.  Unless rfile and wfile are
physically two separate sockets, I would expect cross-platform problems
even if I saw it working somewhere.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From jeff at ccvcorp.com  Tue Apr  2 19:15:43 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 16:15:43 -0800
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net> <3ca8f000$1@news.nwlink.com> <JLcfwNF0zPq8EwJm@nym.alias.net> <3ca9e2bd@news.nwlink.com> <MPG.1713c1ae2a7ecc1698968d@news.nwlink.com> <3caa39b1$1@news.nwlink.com>
Message-ID: <MPG.1713e97d384949ac989690@news.nwlink.com>

In article <3caa39b1$1 at news.nwlink.com>, jeffbi at jetsoft.com 
says...
> 
> I had heard of wxPython (seen it mentioned in various posts here)
> but I hadn't made the connection (doh!) --- I didn't know it was a
> wrapper around wxWindows, I just assumed it was another Win32
> implementation of Python.
> 
> Thanks again,
> 

Glad to help.  :)  If you're at all familiar with wxWindows, then 
wxPython should make you feel right at home.  It's mostly a 
pretty thin wrapper, with some differences due to language 
incompatibilities (methods return tuples instead of using 
reference parameters, multiple method names stand in for 
overloaded versions, and plain functions stand in for static 
methods), though some aspects of wxWindows that are already 
present in Python are not wrapped (threads, for example).

The wxPython-users mailing list is also *very* informative, and 
Robin Dunn (the maintainer of wxPython) puts an amazing amount of 
effort into helping people out.  

-- 

Jeff Shannon
Technician/Programmer
Credit International


From peter at engcorp.com  Thu Apr 18 23:36:29 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 18 Apr 2002 23:36:29 -0400
Subject: traceback output to a network connection
References: <a9npmv$en6$05$1@news.t-online.com>
Message-ID: <3CBF90BD.497F38D4@engcorp.com>

Henning Peters wrote:
> 
> hi NG,
> 
> for debugging a network server i would like to show the parser's traceback
> output on my network client. I figured out that the built-in traceback
> module would be a good choice, but it seems, that the traceback-methods are
> limited to write only to open file handlers.
> 
> is there a known way to relay this stuff to an open network socket
> connection?
> thanks...

You might be interested in socket.Socket.makefile() which can give
you a file handle from a socket.

-Peter


From lists.ASkwar at DigitalProjects.com  Thu Apr 25 19:38:02 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 01:38:02 +0200
Subject: shipping a python app
In-Reply-To: <20020425231732.GJ21205@teich.Garten.DigitalProjects.com>
References: <a9js5c$2m9$1@panix2.panix.com> <upfv8.14124$hl2.9868@atlpnn01.usenetserver.com> <20020425231732.GJ21205@teich.Garten.DigitalProjects.com>
Message-ID: <20020425233802.GK21205@teich.Garten.DigitalProjects.com>

?Alexander Skwar? sagte am 2002-04-26 um 01:17:32 +0200 :
>      ^
> SyntaxError: invalid syntax
> 
> Well, why is that?

Hm, seems like the packages aren't cross platform.  When I use the zip
package, I get the error message.  When I use the .tar.gz file, I don't
get any errors on my Linux machine.

Can I use the Installer to create a cross-platform binary on Linux which
will run on Windows for a wxPython application?  Which requirements do I
have to meet to be able to do so?

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 18 hours 36 minutes




From abagaric at rest-art.hr  Tue Apr 16 11:49:19 2002
From: abagaric at rest-art.hr (Ante)
Date: Tue, 16 Apr 2002 17:49:19 +0200
Subject: Newbie question
Message-ID: <a9hhbs$1ksn$1@as201.hinet.hr>

from file 'dummy.py'
----------------------------
num = 0

def dummy():
    num += 1

dummy()
----------------------------

now, in the interpreter:

>>> import dummy
Traceback (most recent call last):
  File "<string>", line 1, in ?
  File "C:\downloads\python2.2.1\dummy.py", line 6, in ?
    dummy()
  File "C:\downloads\python2.2.1\dummy.py", line 4, in dummy
    num += 1
UnboundLocalError: local variable 'num' referenced before assignment


Why does this happen?
If I replace num += 1 with print num, it prints 0 without an error.
I guess it scans the whole function and check if there's an assignment to a
name,
and if there is, assumes it's a local variable. But += isn't supposed to
move the
reference but instead change the same variable. Or what did I get wrong?

Ante.





From logiplexsoftware at earthlink.net  Wed Apr 24 17:13:37 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 24 Apr 2002 14:13:37 -0700
Subject: Trinary operator?
In-Reply-To: <87k7qwrf0v.fsf@pvv.org>
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
	<29e28c51.0204201335.6e0856da@posting.google.com>
	<87k7qwrf0v.fsf@pvv.org>
Message-ID: <20020424141337.3b74ec04.logiplexsoftware@earthlink.net>

On 24 Apr 2002 22:57:20 +0200
Oyvind Gronnesby wrote:

> * Cimarron Taylor
> |
> | Will this do what you want?
> | 
> | verboseGender = ('female','male')[gender=='m']
> 
> Wouldn't this break in a PEP 285 Python world?

No.  That was the whole point of subclassing bool from int.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From PoulsenL at capecon.com  Tue Apr 30 11:06:19 2002
From: PoulsenL at capecon.com (PoulsenL at capecon.com)
Date: Tue, 30 Apr 2002 11:06:19 -0400
Subject: Remove masked elements from an array
Message-ID: <72F73B808A78D511A73600034771D9FF717CDC@dc_exchange2.howrey.com>

Just wanted to express my gratitude for everyone's input.  This community is
the main reason why I love to use Python.

Loren




From mgilfix at eecs.tufts.edu  Mon Apr 29 14:11:20 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 29 Apr 2002 14:11:20 -0400
Subject: Python sous UNIX
In-Reply-To: <aak1q3$fgl$1@wanadoo.fr>; from infocap@wanadoo.fr on Mon, Apr 29, 2002 at 08:01:11PM +0200
References: <aak1q3$fgl$1@wanadoo.fr>
Message-ID: <20020429141120.B3267@eecs.tufts.edu>

  Je suggere visiter le "Vaults of Parnassus":
http://www.vex.net/parnassus.  Ce site contiens beaucoups de petits
programmes pour toutes sortes de situations. Bonne chance :)

          -- Mike

On Mon, Apr 29 @ 20:01, infocap wrote:
> Bonjour ? tous,
> 
> Je travaille pour une entreprise d'?dition mais voil? que depuis quelques
> temps je m'int?resse beaucoup ? l'informatique et particuli?rement au
> langage de programmation PYTHON  sous UNIX et j'aimerai approfondir mes
> connaissances au travers de petits programmes int?ressants, alors, si
> quelqu'un peut satisfaire ma demande, je le remercie d'avance.
> 
> Cordialement
> David
> 
> 
> 
> --
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (infocap)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From nospam at bigfoot.com  Fri Apr  5 13:45:50 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 5 Apr 2002 20:45:50 +0200
Subject: RPC for ZODB via https
References: <3CAD9F25.8010708@thomas-guettler.de>
Message-ID: <a8kqtu$1d59$1@norfair.nerim.net>

xmlrpclib *is* object oriented, especially when used with Zope

s = xmlrpclib.Server(http://url_to_a_zope_object)
data = s.method_of_zope_object(param)

But the error handling is not easy because Zope replies with a huge HTML
error message using the standard page rendering. :(

In example, put this one line script (call it "allFolders") at the root of
Zope

return context.objectIds('Folder')

then in a client python interpreter

>>> s = xmlrpclib.Server(http://path_to_any_zope_folder)
>>> print s.allFolders()

Standard web services support (SOAP, WSDL...) is expected for next version
(certainly 2.6) of Zope

"Thomas Guettler" <st-newsgroups at thomas-guettler.de> a ?crit dans le message
news: 3CAD9F25.8010708 at thomas-guettler.de...
> I interested in how I can access ZODB via an RPC mechanism
> which uses https.
>
> Possible solutions:
>
> +use Zope and XML-RPC via https.
>
> +use xmlrpclib
>
> But xml-rpc is not object-oriented. Has someone other ideas?
>
>   thomas
>




From jimd at vega.starshine.org  Mon Apr  1 04:50:24 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 1 Apr 2002 09:50:24 GMT
Subject: Data type ideas
References: <mailman.1017475907.10586.python-list@python.org> <c76ff6fc.0203301338.1cf07eb0@posting.google.com>
Message-ID: <a89ah0$2oki$1@news.idiom.com>

In article <c76ff6fc.0203301338.1cf07eb0 at posting.google.com>, John Machin wrote:
>"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in message news:<mailman.1017475907.10586.python-list at python.org>...

>>> So what next?  Any ideas that I can use?  
>> look into pickles.  Act like dictionaries but live on the hard drive.

>Do you mean "shelves" instead of "pickles"? Even so, how do they help
>the OP with his problem?

 If your data doesn't fit in memory, use disk (or even tape).
 So the real question is, how can you structure files on disk to
 partition this data into usefully smaller subsets that can each
 be processed.  
 
 For example you might create a subdirectory (for the whole 
 processing job) and a subdirectory for each group or person
 (or a pair of directories, persons and groups).  Now you can search
 through the original file, parsing parts of it that do fit into 
 memory and splitting out the subsets (appending persons to the 
 group files, and groups to the person files).  Now you should be
 able to do a merge sort to get the information you want, organized
 according to your needs.

 Alternatively you could import this data into a set of database tables
 and then query the database to sort and gather the data.

 Which of these approaches is better depends quite a bit on your 
 background and your needs.  If you'll have frequent need of this data,
 or you'll have frequent large updates of additional data, the database
 approach might be worth the extra investment of your time and energy.
 If you were particularly experienced with RDBMSes (in general) and
 had one (like PostgreSQL or MySQL) already installed then it might
 be easier to do that than to cook up your own filenaming and 
 directory searching scheme.

 From what little I saw of your data it seems to require a 
 straightforward multi-multi schema (use threee tables, one for
 person_id, and person name, one for group_id, and group name, and
 a third to relate person_ids to group_ids).  Any elementary text on
 SQL database design will cover a "third normal form" multi-multi scheme 
 usually with a description involving authors and titles (since that's
 the most obvious real world example of entities that can have a
 multi-multi relationship, at least to the authors of these texts).
 Obviously a title/book can have multiple authors, and any author might
 write multiple works.  The first "normalization" rule (E.F. Codd) prohibit 
 us from putting multiple attributes in any single column (and require
 that all rows be of fixed numbers and types of columns) and the 
 second normal form requires that we put non-dependent attributes in
 independent tables.

 However, it seems likely that you would have already thought of the 
 RDBMS approach if you were familiar with using SQL and database tables.
 Since you aren't, the learning curve may not be worth it -- especially if
 this is just a oneshot.


From vjovanov at stevens-tech.edu  Tue Apr 30 17:15:50 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Tue, 30 Apr 2002 17:15:50 -0400
Subject: cPickle problem
References: <Ssiz8.63$bK3.1993791@newnews.cc.stevens-tech.edu> <FAiz8.78905$vF6.2347612@news2.tin.it>
Message-ID: <bQDz8.64$y34.1997856@newnews.cc.stevens-tech.edu>

Thanks, I understand this a bit better.  I changed my implementation of foo
like this

>>> class foo:
 def __init__(self):
  self.y=4
  self.e={'a':'self.y',
   'b':'self.a+4'}
 def __getattr__(self,arg):
  if self.e.has_key(arg) == 0:
   raise AttributeError,  "foo instance has no attribute '" + arg + "'"
  try:
   exec 'result =' + self.e[arg]
   return result
  except:
   return '#U'
 def addequation(self,attr,value):
  self.e[attr] = value


>>> k=foo()
>>> k.e
{'b': 'self.a+4', 'a': 'self.y'}
>>> k.a
4
>>> k.b
8

>>> k.ddd
Traceback (most recent call last):
  File "<pyshell#121>", line 1, in ?
    k.ddd
  File "<pyshell#105>", line 8, in __getattr__
    raise AttributeError,  "foo instance has no attribute '" + arg + "'"
AttributeError: foo instance has no attribute 'ddd'

As you can see I added a new method 'addequation' that works like this

>>> k.addequation('c','self.a+self.b')
>>> k.e
{'b': 'self.a+4', 'c': 'self.a+self.b', 'a': 'self.y'}
>>> k.c
12

and now I have another question.

can I replace addequation with __setattr__ so that I can type

>>>k.c='self.a+self.b'

to achieve the same effect.
>>> k.e
{'b': 'self.a+4', 'c': 'self.a+self.b', 'a': 'self.y'}
>>> k.c
12

My attempts with __setattr__ were unsuccessful.

Thanks for you help

Vin






From aleax at aleax.it  Thu Apr 11 09:42:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 13:42:53 GMT
Subject: appeding string to file
References: <mailman.1018528523.30799.python-list@python.org>
Message-ID: <xpgt8.6669$b62.161350@news1.tin.it>

Brian Lee wrote:

> I tried follow test code to append string to a file.
> When list t is size 10000, second time is slower fourth
> But list t is size 30000, second time is slower twenty three
> times than first time.
> 
> Why does second test is more slower? Is the load to copy
> string to a variable too high?

Little to do with files, the issue is with how you build strings:
        ...
> for x in t:
>     tt += x + '\n'

Strings are immutable, so this needs to build and then immediately
throw away longer and longer strings.

Just do one tt = '\n'.join(t) and see it soar (with exactly
the same operations on files).


Alex



From spahievi at vega.bg  Tue Apr  9 16:57:14 2002
From: spahievi at vega.bg (Niki Spahiev)
Date: Tue, 9 Apr 2002 23:57:14 +0300
Subject: wxPython / Your Experiences in large project ???
In-Reply-To: <a8rp2u$ht8$02$1@news.t-online.com>
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <1037168086.20020409235714@vega.bg>

08.4.2002, 12:55:24, Harald Schneider wrote:

HS> Hi,

HS> what are your experiences, using wxPython/wxWindows in large projects?

We at Vintech Ltd. have moderate CAM system based on Python and
wxPython. Using Win 9x is not a problem if you watch for GDI
resources. Using Insteller and Inno setup produces very good
distributive.

-- 
Best regards,
 Niki Spahiev


______________________________________
VEGA(TM) Internet Service Provider
      <http://www.vega.bg>
Scanned and protected by Inflex
Tova pismo e provereno i zashtiteno
ot programite "Inflex/Antivir".






From pyth at devel.trillke.net  Sun Apr 28 07:52:21 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 28 Apr 2002 13:52:21 +0200
Subject: Update locals()
In-Reply-To: <0204280830290C.16995@arthur>
References: <20020427195725.GA16217@devel.trillke> <ySFy8.28416$8D3.759092@news1.tin.it> <20020427233027.GE16217@devel.trillke> <0204280830290C.16995@arthur>
Message-ID: <20020428115221.GF16217@devel.trillke>

Thanks for the explanations. 

There is no question that C++-like templates often are
not needed in python. I regret mentioning C++ because
it is only a very distant analogy as you pointed
out precisely. 

> Please give ONE example which, in your opinion, is best handled
> with 'exec' without explicit dictionaries rather than in any other way.

i try to make it short and answer questions (if any of you gurus
ever have any :-) as needed. Obviously you can always pass explicit 
dictionaries to exec (globals, locals). That can't be the point. 

I cite some of my code.

class filename:
    """ Abstraction for filename ... """
    # ...

    for name in filters._stateful_all:
        doc=filters.__dict__.get(name,undoc).__doc__.strip()+'\n'
        exec indentedcode("""

    def %(name)s (self,*args,**kargs):
        '''%(doc)s'''
        return filters.%(name)s (*args,**kargs) (self)
    """ % locals()) 

    # ...

# 'indentedcode' basically prepends 'if 1:' so that i
# don't have to write the template without indenting.

Generally i have lots of nice callable objects 
in the 'filters' module. Some must be initialized, some
are stateless. 

#module 'filters' contains *for example*

class access:
    """(stateful) checks filename for a given access-mode.
       the mode string contains a combination
       of 'r' 'w' 'x'. the filter afterwards accepts
       a filename or a string if it is accessable 
       with the given mode"""
    def __init__(self, mode='r'):
        flags=0
        if 'r' in mode: flags|=os.R_OK
        if 'w' in mode: flags|=os.W_OK
        if 'x' in mode: flags|=os.X_OK
        self.flags=flags

    def __call__(self, fn):
        return os.access(str(fn),self.flags) and fn

In this case of defining templated methods in a class with
tiny adapters to make use of functions somewhere else
it seems natural to stay in the same exec-context (this case 
class definition), whatever it may be. That's why
i wrote 'indentedcode' to stress the fact.

With the prescribed python-templates you can adapt 
'syntax' to achieve a richer calling interface and 
avoid redundancies in coding. I want to define 
stateful/stateless filters in one place and let 
the filename class offer adapted methods and inherit 
the documentation.  Also see an example at the end on how
you this can be put to use.

Note three things:

- I guess that there are other methods like
  Currying which *might* help here. I don't
  think they provide the same ease, though.

- in the prescribed type of templates 
  i think there is no performance penalty
  as the compiler sees the complete 'def .. : '

- i am a maniac at avoiding redundancies.
  that's the reason i looked for transfering
  matches in re-patterns into something
  useable without noise.

If you really know an easy minimally redundant way 
of expressing the same semantics (above) i am very 
interested.

But i even have two other examples :-) 

regards,

    holger


--

# real world example using the precscribed semantics 
# 
base=filename('/usr/src/cpp')

count = Counter()
resultfilters = AND(isfile, nolink, fnmatch('data*'), count)
recursefilter = access('wx')

for dir, fnlist in base.filterwalk( resultfilters, recursefilter ):
    for fn in fnlist:
        print "  ",fn.relativeto(base)
        ...

print 'processed %d occurences of writable data files' % count.count

# filterwalk is a generator 
# filterwalk_now returns the complete list in one go










From rahvee at yaahhoo.com  Wed Apr 10 17:13:11 2002
From: rahvee at yaahhoo.com (Ned Harvey)
Date: Wed, 10 Apr 2002 17:13:11 -0400
Subject: Indenting lines 400..457 help
Message-ID: <a929tk$ik2$1@pcls4.std.com>

For the life of me, I can't find any easy way to indent (or unindent) lines
numbered 400..457 or whatever.  Does anyone have any good suggestions?

I'm usually a vi user, but I'll happily learn a command of some other
format.  I've tried
    :400,457s/(.*)/   \1/g
and a ton of variations of the same idea, but it seems that the
search/replace doesn't do what I expect it to.

Thanks for your help.




From matt at greenroom.com.au  Wed Apr  3 08:17:08 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Wed, 3 Apr 2002 23:17:08 +1000
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <a8eve0$kuo$1@perki.connect.com.au>

"vio" <vmilitaru at sympatico.ca> wrote in message
news:3CAAC372.792A0F9A at sympatico.ca...
> Greetings,
>
> I need to translate a (rather voluminous 100MB) java program to python.
After a
> few searches on google, apparently the best tools for this task are
Lex-Yacc
> (Flex-Bison on linux). While I am working on the Lex-YACC HOWTO, I would
be very
> interested in hearing of others' successful experiences/feedback at this.
Any
> pointers of interest would similarily be appreciated.
>
> Many thanks,
> Vio

I'm inclined to wonder why you would be better off with python rather than
java for such a large project and, whether the added cost of using java sums
up to the HUGE (well, maybe) cost of converting to python.

If it is completely necessary however, i'd suggest reviewing the design,
adapting it to work with python and re-write it rather than translating
class for class, method for method and line for line.

Regards,
  matt




From ben at wblogan.net  Mon Apr  8 12:11:55 2002
From: ben at wblogan.net (Ben Logan)
Date: Mon, 8 Apr 2002 12:11:55 -0400
Subject: Reading/writing binary data.
Message-ID: <20020408121155.A16826@wblogan.net>

Hello,

I've been searching the Python (1.5) docs, but am unable to find any
functions to read/write binary data from/to a file.  Under the "open"
function I see that you can append a "b" to the mode to specify binary
mode, but that doesn't help me any.

For example, if I want to take a list of numbers like [64,66,67] and
write them to a file as binary, how would I do that?  I also need to
be able to control whether the data is written in big endian or little
endian format.

BTW, I'm running Redhat 7.1 on an Intel processor.

Thanks,
Ben

-- 
Ben Logan: ben at wblogan dot net
OpenPGP Key KeyID: A1ADD1F0

Overdrawn?  But I still have checks left!




From krissepu at vip.fi  Wed Apr 10 07:36:21 2002
From: krissepu at vip.fi (pekka niiranen)
Date: Wed, 10 Apr 2002 11:36:21 GMT
Subject: Win32 and Excel question
References: <3CB3F220.2B8E668F@ihug.co.nz>
Message-ID: <3CB423F9.1080907@vip.fi>

Try this link

    http://examples.oreilly.com/pythonwin32/

-pekka-

Matthew wrote:

>Hi,
>I'm playing with win32COM with a little success. I have a large XL table
>with 8 columns and about 3500 rows. What I want to do is search only
>column B for a value and return the row number where this value is
>found. So, I have 2 questions:-
>1) Howto construct a Win32COM statement to do this? So far I have this:-
>
>found=myxl.ActiveSheet.Range("B1:B3500").Find(<value_i'm_looking_for>
>However, this search finds the first match in any column between B1 and
>B3500. Yep, their are similar values in other cells of the s/s
>
>>>>found.Row
>>>>
>25
>
>>>>found.Column
>>>>
>1 (ie: Column 'A')
>Help please.
>2) Howto return a tuple of all values in only column 'B' and then use
>python to search the tuple for the value I'm after. It strikes me that
>this is probably a faster approach?
>Thanks for helping. matthew.
>BTW: Thanks Mark Hammond for a v.cool and useful tool.
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020410/07a1fe85/attachment.html>

From aboufer at atlsci.com  Fri Apr 19 11:26:49 2002
From: aboufer at atlsci.com (A. B.)
Date: Fri, 19 Apr 2002 11:26:49 -0400
Subject: European Win32. Python weirdness
Message-ID: <3CC03738.2725CC26@atlsci.com>

Greetings all,

I am having problems with a module of python, mainly on Win32 platform
using European languages. In some of the code we wrote, python does not
seem to recognise a
number as a number. More specificcaly, when python reads a string of
digits, 'he' does not
recognise it as a number but as a string. This of course causes the
reste of the processing to fail!

Does anybody ever seen such a weird behavior?

Thanks a lot.



From t_list at vtnet.ca  Thu Apr  4 10:22:10 2002
From: t_list at vtnet.ca (trevor)
Date: Thu, 04 Apr 2002 10:22:10 -0500
Subject: Deeply embedded Python still alive?
References: <mailman.1017868878.1845.python-list@python.org>
Message-ID: <Ua_q8.1359$cC5.576756@newsfeed.slurp.net>

in the following i assume you're using gcc-3.0.4. i also assume you're 
using uClibc-0.9.10, that you've setup your install to be put at 
/usr/local/device/ and that you're building your target filesystem at 
/home/me/device (if you look at uClibc you'll know what i'm talking about 
if this isn't clear right now).

[uClibc-0.9.10]
in uClibc's configuration file i've set:
        KERNEL_SOURCE
        DO_C99_MATH = true [not sure if needed or not]
        HAS_LONG_LONG = true
        DOLFS = true [this one is important]
        INCLUDE_THREADS = true
        turned on the shared stuff
        DEVEL_PREFIX = /usr/local/device/$(TARGET_ARCH)-linux-uclibc
# make
# make install
# make PREFIX=/home/me/device install_target

[python-2.2]
unpack python-2.2 somewhere, create a build directory. cd into that build 
directory.
# PATH=/usr/local/device/i386-linux-uclibc/bin:$PATH
# export PATH
# ../path/to/python/source/configure --prefix=/some/where/else
edit pyconfig.h:
        - change "#define LARGEFILESUPPORT 1" -> 0
        - change "#define _FILE_OFFSET_BITS 64" -> 32
edit Modules/Setup
        - uncomment "_socket socketmodule.c" to compile socket w/o SSL
edit source code: Modules/getaddrinfo.c:
*** start of diff ***
diff -urN Python-2.2/Modules/getaddrinfo.c 
SmallPython-2.2/Modules/getaddrinfo.c
--- Python-2.2/Modules/getaddrinfo.c    Sun Dec  2 05:15:37 2001
+++ SmallPython-2.2/Modules/getaddrinfo.c       Wed Apr  3 13:28:18 2002
@@ -198,7 +198,7 @@
 
 #define ERR(err) { error = (err); goto bad; }
 
-char *
+const char *
 gai_strerror(ecode)
        int ecode;
 {
*** end of diff ***
edit Makefile
        search for LIBSUBDIRS string at start of line
        i removed the test* and tk* directories
# make
# make install

you'll notice that some of the modules won't build and other that will 
build will be discarded since they can't be loaded. if you wanted the expat 
module to build (for xml stuff) for example, you'll need to download and 
install (--prefix=/usr/local/device/i386-linux-uclibc) expat-1.95.2.

now you need to go into where you're creating your target device's 
filesystem and re-create the python directories (i.e. 
/home/me/device/some/where/else) you need to re-create this file layout 
since the python executable will look for it when it tries to load modules 
(i don't know how to change this behaviour).

starting at /home/me/device/some/where/else create "bin" and "lib" 
directories. put the python executable into "bin" and create a "python2.2" 
directory in "lib".

you'll notice that the /some/where/else/lib (the place you installed your 
complete python build to) directory is massive. you don't need to copy all 
the modules found there to your target device's filesystem, only the ones 
you need. through experimentation i've discovered that the following 
minimum are required to be able to run the interpreter and do some basic 
stuff. the following are put into the $TARGET/some/where/else/lib/python2.2 
directory:
        create a directory called lib-dynload
        os.py
        UserDict.py
        posixpath.py
        site.py
        stat.py

note that i'm only copying the *.py files. this was done to save disk space 
since on average the *.py files are smaller than the *.pyc or *.pyo files. 
however, the first time these files are opened, python will convert them 
into *.pyc files giving you a really slow first-time startup time. in this 
environment disk space is really tight (4MB). the *.pyc files will be 
created in the RAM filesystem and will live there until the next reboot. if 
you can spare the space then you only need to copy the *.pyc files to the 
above modules.

as you write programs that require different modules you'll need to include 
them. starting python with the "-v" option will tell you which ones you're 
missing.

hope this helps. best regards,
        trevor
-- 
remove underscore if you want to email me


From eric.brunel at pragmadev.com  Thu Apr 11 06:00:09 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 11 Apr 2002 10:00:09 +0000
Subject: Phyton and SQL
References: <3CB5295A.4070803@hrz.tu-chemnitz.de>
Message-ID: <a93fht$nkv$1@norfair.nerim.net>

Chris Burkert wrote:
> I just started yesterday and took a look at the Phyton-Manual the first
> time. It's a very interesting and powerful language. But what I didn't
> find was a comfortable way to communicate with Databases like Oracle or
> MySQL through the SQL language.

That's only because Python does not come "out of the box" with database 
support. But there are additional modules allowing to communicate via SQL 
almost with any database engine you can think of. There's in fact a "SIG" 
(Special Interest Group) entirely devoted to database communication. You'll 
find it here:
http://www.python.org/topics/database

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From chris at cmb-enterprises.com  Mon Apr  8 01:25:08 2002
From: chris at cmb-enterprises.com (Chris)
Date: Mon, 08 Apr 2002 01:25:08 -0400
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4> <chris-CFFF94.21350207042002@corp.supernews.com> <3CB106AA.1AA3A5BF@engcorp.com>
Message-ID: <chris-B48101.01250808042002@corp.supernews.com>

In article <3CB106AA.1AA3A5BF at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> Chris wrote:
> > 
> > In article <Xns91E989929996CRASXnewsDFE1 at 130.133.1.4>,
> >  Philip Swartzleonard <starx at pacbell.net> wrote:
> > 
> > > def percent_chance( percentage ):
> > >       "Assumes percentage as number 0-100"
> > >       if random.random() > (percentage / 100.0):
> > >               return 1
> > >       else:
> > >               return 0
> > 
> > Might more concisely be written as:
> > 
> > def percent_chance( percentage ):
> >      return random.random() > (percentage / 100.0)
> 
> But this is Python, and a newbie.  I think Philip's
> effort was more suited to the audience, as it is
> more readable. 

It's redundant, and "random.random() > (percentage / 100.0)" is no more 
readable with a "if" in front of it than it is with a "return".

I'm not trying to be rude, it just seems silly to do the work twice.

And I commend anyone for learning Python.  The more languages the better!

(Though I wouldn't mind if a few more of them looked like Ruby :-)


From geh at itga.com.au  Mon Apr 22 20:01:48 2002
From: geh at itga.com.au (Gary Hughes)
Date: Tue, 23 Apr 2002 10:01:48 +1000
Subject: problem with extension library
Message-ID: <3CC4A46C.10706@itga.com.au>

Hi,

I have written an extension library in C++, it works fine when built and 
  run on a solaris 2.8 machine with python 2.1.1.

When I build and attempt to use the same library on a solaris 2.6 
machine with python 2.1.1 I get the following error.

Traceback (most recent call last):
   File "./t_py_configuration.py", line 31, in ?
     from configuration import *
ImportError: ld.so.1: /opt/pythontest/bin/python: fatal: relocation 
error: file ./.libs/configuration.so: symbol __ti7ostream: referenced 
symbol not found

The library is build with the same version of g++ (2.95.3) on both 
platforms and the library and python are built with PIC.

I have seen a lot of problems similar to this however none quite the same.

Can anyone shed any light on this.

thanks.

Gary.



From fakeaddress at nowhere.org  Thu Apr 25 00:39:38 2002
From: fakeaddress at nowhere.org (Bryan Olson)
Date: Thu, 25 Apr 2002 04:39:38 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com>
Message-ID: <3CC7889D.2080002@nowhere.org>

Paul Rubin wrote:

 > ARGGH!  I knew about that before too, but forgot about it while writing
 > the code.

I figured you were testing us.

 > I will replace SHA(block) with SHA(key + block) and release
 > a new version tonight, with the MAC improved too.  Sound ok?


Let me see if I understand.  We start with a shared secret key called
"key", which may be of any length.  Each message requires it's own nonce
of 16 bytes.

 From this we form a session encrption key, called "enc_key" as:

     enc_key = _hash("enc" + key + nonce)

Now we need to intialize the variable "seed", and we need a new way
to do it.  Perhaps:

     seed = _hash("seed" + enc_key)

Then we form the running key, called "xkey", which is a list of
20-byte blocks.

     xkey = []
     for i in range(n_blocks_needed):
         seed = _hash(enc_key + seed)
         xkey.append(seed)

The ciphertext is the exclusive or of the plaintext and the running key.
The nonce travels with the ciphertext.


Now I have a few suggestions.

The system presents a cipher, a MAC, key derivations, and formats. I
think these should be defined individually.  I agree with defining how
to encrypt-and-sign with a master key, but I think the smaller elements
should also be defined and exposed on their own.

I would separate the IV from the ciphertext, and make computing the MAC
a separate function.  I see no need to include the nonce in computing
the auth_key. In some protocols, such as SSL/TLS, we would send a stream
in many sections.

Here's a candidate reference implementation.  Each function does one
thing that I believe should be individually defined.  The code is
written for lucidity and simplicity, not speed. I'm using Paul Rubin's
latest (at the time of this writing) MAC, but really I recommend going
with HMAC.


########################################################

import sha

class CryptError(Exception): pass

MACLEN = 8
IVLEN = 16


def session_cipher(enc_key, text):
     seed = _hash("seed" + enc_key)
     result_text = []
     while text:
         (block, text) = (text[:20], text[20:])
         seed = _hash(enc_key + seed)
         block = _xor_strings(block, seed[:len(block)])
         result_text.append(block)
     return "".join(result_text)


def session_sign(auth_key, text):
     k1 = _hash('auth1' + auth_key)
     k2 = _hash('auth2' + auth_key)
     return _hash(k1 + _hash(k2 + text))[:MACLEN]


def derive_enc_key(master_key, nonce):
     assert len(nonce) == IVLEN
     return _hash('enc' + master_key + nonce)


def derive_auth_key(master_key):
     return _hash('auth' + master_key)


def encipher(master_key, plaintext):
     nonce = _generate_nonce_somehow(master_key + plaintext)
     enc_key = derive_enc_key(master_key, nonce)
     return (nonce, session_cipher(enc_key, plaintext))


def decipher(master_key, nonce, ciphertext):
     enc_key = derive_enc_key(master_key, nonce)
     return session_cipher(enc_key, ciphertext)


def sign(master_key, text):
     session_key = derive_auth_key(master_key)
     return session_sign(session_key, text)


def verify(master_key, text, signature):
     if sign(master_key, text) != signature:
         raise CryptError, "invalid key or ciphertext"


def encipher_and_sign(master_key, plaintext):
     (nonce, ciphertext) = encipher(master_key, plaintext)
     signature = sign(master_key, ciphertext)
     return (nonce, ciphertext, signature)


def verify_and_decipher(master_key, nonce, ciphertext, signature):
     verify(master_key, ciphertext, signature)
     return decipher(master_key, nonce, ciphertext)


#############################################################
#   Utility functions

def _hash(x):
     return sha.new(x).digest()

def _xor_strings(s1, s2):
     return "".join(map(lambda x, y: chr(ord(x) ^ ord(y)), s1, s2))

_state = ""
def _generate_nonce_somehow(entropy=""):
     global _state
     import time
     _state = _hash(`time.time()` + _state + entropy)
     return _state[:IVLEN]


#############################################################
#   Not a great test but....

if __name__ == "__main__":
     master_key = "Big time secret key"
     plaintext = "No matter what happens, somebody just knew it would."
     nonce, ciphertext, sig = encipher_and_sign(master_key, plaintext)
     deciphertext = verify_and_decipher(master_key, nonce, ciphertext, sig)
     print deciphertext
     assert(deciphertext == plaintext), "Darn, it failed."





From gh_pythonlist at gmx.de  Fri Apr 19 22:39:24 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 20 Apr 2002 04:39:24 +0200
Subject: OT: ViM compiled against Python 2.2.1 available?
In-Reply-To: <3CBFCA44.8010200@n0spam_xs4all.nl>
References: <3CBFCA44.8010200@n0spam_xs4all.nl>
Message-ID: <20020420043923.A7940@gargamel.hqd-internal>

Le 19/04/02 ? 09:41, iwk ?crivit:
> Hi there,
> 
> I've installed Python 2.2.1 on a new Windows 2000 install. The "problem" 
> is that ViM 6.1 is compiled against Py2.1 so I cannot use any of the 
> integration features I've gotten used to. On Linux I would compile it 
> myself, but I don't have access to a compiler on Windows.

Sure you do. Go here: http://www.mingw.org/ where you can find the Best
Compiler [tm] (gcc) for producing native win32 executables.

> Are there versions of ViM6.1 availabe for download which are compiled aginst
> Py2.2.1?

I've just built it (VIM 6.1.18 linked against Python 2.2.1) and made it
available here:

http://www.cs.fhm.edu/~ifw00065/vim/index.html

And yes, I only changed two lines in the Mingw Makefile included in Vim's
source distribution and compiled the whole thing.

For making the installer, a little more was required (the VisVim and GvimExt
stuff only has MSVC makefiles).

I'm using NSIS bzip2 compression again so the whole download is only 2.3 megs.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.0 ?C      Wind: 2.0 m/s




From dmmcf at uiuc.edu  Mon Apr 15 15:55:28 2002
From: dmmcf at uiuc.edu (D. Michael McFarland)
Date: Mon, 15 Apr 2002 14:55:28 -0500
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu>
Message-ID: <861ydg3ff3.fsf@haggis.aae.uiuc.edu>

Fernando P?rez <fperez528 at yahoo.com> writes:

> Have you looked at the possibility of rewriting some of the tight
> inner loops in C? These days using weave (http://scipy.org/) it's
> perfectly reasonable to get large speedups with very little work. I
> don't know your code so I don't know how realistic this is (and you
> guys seem to be under tight deadlines), but you can look at this
> pdf: http://www-hep.colorado.edu/~fperez/python_misc/python-c.pdf
> for a simple example where the speedup is a factor of 25.

I've looked at doing something like this, but in the end elected not
to because I'm afraid of introducing subtle bugs into something we've
developed some confidence in.

I have profiled the Python code pretty carefully, and was able to
achieve a reduction in run time of a little better than 50% compared
to my first, naive implementation.  The bottlenecks now seem to be in
memory management, which I'm sure Python does better than I would in
C, and numerical linear algebra, which I'm sure Numeric does better
than I would.

Once I understand Weave and friends better, and the stakes are a
little lower, I'll try that approach.  For now, though, the safest
route seems to be to throw CPU time at the existing code.

Thanks for your interest and the pointer.

Cheers,
Michael


From shalehperry at attbi.com  Fri Apr 26 16:56:30 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 26 Apr 2002 13:56:30 -0700 (PDT)
Subject: Removing dupes from a list
In-Reply-To: <20020426203946.GV21205@teich.Garten.DigitalProjects.com>
Message-ID: <XFMail.20020426135630.shalehperry@attbi.com>

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560

The result of typing

unique list python

in google (-:




From phr-n2002a at nightsong.com  Tue Apr 30 21:29:58 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 30 Apr 2002 18:29:58 -0700
Subject: crypto
References: <mailman.1020206574.19040.python-list@python.org>
Message-ID: <7x4rhshcyx.fsf@ruckus.brouhaha.com>

"Mark McEahern" <marklists at mceahern.com> writes:
> 1.  Generate public/private key pairs.
> 2.  Encrypt data with the public key.
> 3.  Decrypt it with the private key.
> 
> Step 1 is likely to be manual.  (You're probably thinking, "No shit,
> sherlock.")
>
> I'm using Linux and I've seen how I can create a key with gpg.  I'm just not
> sure how to use that properly from Python.  I suppose I could write things
> out to files and use os.popen() or something, but that seems like it would
> increase my risk, not decrease it.  ;-)

Do you want to interoperate with some particular program like gpg?

If not, I have a Python library (pre-alpha, but you can try it out if
you want) that does what you're asking.  Preliminary docs are at:

  http://www.nightsong.com/phr/crypto/crypto.txt

It uses fairly boneheaded algorithms and mostly bends over backwards
to keep your keys secure, a failing of most crypto libraries.

If you have general crypto questions, sci.crypt is a better place to
ask than comp.lang.python.


From doyle at acrossthebigpond.com  Sun Apr  7 19:34:33 2002
From: doyle at acrossthebigpond.com (Kevin Doyle)
Date: Sun, 07 Apr 2002 23:34:33 GMT
Subject: Get Button Text
Message-ID: <3CB0D763.3090901@acrossthebigpond.com>

I have code that creates x amount of Buttons on the fly.  Each buttons 
will have a unique Text displayed on it.

I want to be able to get the text vale of the button when it is pressed.

I also want to know is it possible to create just one function and bind 
this one function to all the buttons, and call the function with a 
reference to the button clicked as an argument.

Any help would be great as I am banging my head against a wall trying to 
get this working.

Thanks

Kevin




From aleax at aleax.it  Thu Apr 11 09:06:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 13:06:24 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <3CB58302.5090200@mxm.dk>
Message-ID: <kTft8.6500$b62.159068@news1.tin.it>

Max M wrote:
        ...
> def newJoin(self, items):
>      return self.join([str(i) for i in items])
> 
> Is there any case where it isn't applicable at all?

Could be a disaster if any item was Unicode:

>>> print ''.join(('mapp','erch',u'\u00e8')).encode('latin-1')
mapperch?

vs:

>>> print ''.join([str(x) for x in 
('mapp','erch',u'\u00e8')]).encode('latin-1')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII encoding error: ordinal not in range(128)
>>>

but presumably you could have something subtler that used
str() or unicode() as appropriate (maybe not trivial to
determine what's "appropriate" in each given case).


Alex



From gerhard at bigfoot.de  Fri Apr 26 17:03:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 26 Apr 2002 23:03:34 +0200
Subject: cgi module question
In-Reply-To: <3CC9A3E2.9060803@nmt.edu>
References: <3CC9A3E2.9060803@nmt.edu>
Message-ID: <20020426210333.GA474@lilith.my-fqdn.de>

* Jeff Shipman <shippy at nmt.edu> [2002-04-26 13:00 -0600]:
> I'm using Python 1.5.2 and the cgi module. I'm
> wondering if there's any way to recover the string
> that was passed to the cgi module during a POST
> operation (normally comes in through stdin)? It's
> very important that I get the string in the exact
> form that it was posted to the CGI.

Simply don't use the cgi module, but read sys.stdin yourself instead.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From logiplexsoftware at earthlink.net  Wed Apr  3 15:49:53 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 12:49:53 -0800
Subject: Python Version Strategy?
In-Reply-To: <3c94aaea.161405128@news.laplaza.org>
References: <xIYj8.16613$_L.7649@news1.bloor.is>
	<mailman.1016094984.20818.python-list@python.org>
	<3c94aaea.161405128@news.laplaza.org>
Message-ID: <20020403124953.46b5e1f1.logiplexsoftware@earthlink.net>

On Wed, 03 Apr 2002 19:42:26 GMT
Mats Wichmann wrote:

> Red Hat is trying to at least /appear/ to be stabilising things for
> their customers, and they won't roll key software except at major
> release time.  Python is used "internally" so it qualifies on this
> front. That means that the entire 7.x series continues on Python
> 1.5.2, and the sad news is, the next Red Hat release will be 7.3 so
> figure that Red Hat will be stuck on Python 1.5.2 through the end of
> this year.  You can now have a more modern version as long as you're
> willing to call it python2.
> 
> The sad thing is while Red Hat is sticking to their guns on this
> issue, they keep breaking other stuff for folks, making one wonder why
> they bother standing on the Python issue so hard.

Yeah, I had to revert to an older version of gnucash when I moved from 7.0
to 7.1, and then again when I moved to 7.2, only this time it was an all
day event as it would no longer build on the new system.  I was unable to
balance my checkbook for an entire weekend.  But gnucash is an application
that people would probably come to depend on so it's probably okay to move
to an incompatible version in a point release, versus changing versions of
Python which most people wouldn't even notice.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From woodsplitter at rocketmail.com  Mon Apr 15 20:14:28 2002
From: woodsplitter at rocketmail.com (David Rushby)
Date: 15 Apr 2002 17:14:28 -0700
Subject: Python webmail packages?
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com>
Message-ID: <7876a8ea.0204151614.138853be@posting.google.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote in message news:<7xr8lhb54s.fsf at ruckus.brouhaha.com>...
> Does anyone have a favorite webmail system written in Python?
> Something like Squirrelmail (www.squirrelmail.org) but written in
> Python and hopefully with fewer bugs ;-).  It would be good if it
> talks to an IMAP server like Squirrelmail does, but that's not mandatory.

For what it's worth, Rocketmail (purchased by Yahoo! and transformed
into Yahoo! Mail) was originally written in Python CGI.  I don't know
what portions of the service, if any, are still written in Python.

NotMail is a Zope-based webmail system:
http://www.zope.org/Members/Benno/NotMail2

BoboMail:
http://bobomail.sourceforge.net/

Shicks (I think the web interface is for admin, not accessing the mail
itself):
http://www.p-nand-q.com/shicks.htm

I've never used any of these; I just Googled.


From cyberlync at yahoo.com  Tue Apr  2 22:55:20 2002
From: cyberlync at yahoo.com (Eric Merritt)
Date: Tue, 2 Apr 2002 19:55:20 -0800 (PST)
Subject: C Extension
In-Reply-To: <XFMail.20020402194110.shalehperry@attbi.com>
Message-ID: <20020403035520.99162.qmail@web13002.mail.yahoo.com>

I hope you guys dont mind another stupid question. I
have created a new C extension to python (Its my first
one so I expect issues). This extension contains two
types. In general these types seem to work just fine.
However when I enter a value of one of these types on
a line by itself python segfaults. I have no clue as
to what function is getting called at that particular
point. At first I thought it was the repr function, or
perhaps there string function. But I stuck a couple of
print strings there to verify that and those dont seem
to be getting called. Could one of you tell me which
function is called when a value of one of these types
is on a line by itself? If I could get that debugging
it shouldnt be a big deal.

Thanks
Eric

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/



From phr-n2002a at nightsong.com  Tue Apr 16 15:26:12 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 16 Apr 2002 12:26:12 -0700
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <8MVu8.26112$vF6.828988@news2.tin.it> <x7elhfpmvc.fsf@ra.iago.org> <mailman.1018979071.18732.python-list@python.org>
Message-ID: <7xofgje97v.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Yes, but a GUI thread can't do that either -- nor can a network one;
> they must be ready for user-input events and netword-data-arriving
> events too.  The problem is how to seamlessly merge waiting for
> THOSE kinds of things and for Python-thread-stuff kinds...

What's really needed is a way to raise events asynchronously in other
threads.  However this seems technically difficult.  There's some
discussion of it in one of the sourceforge bug reports; I don't have
the number handy but the title is something like "add a way to make
the program completely exit" (i.e. by killing all the threads).


From feldspar at ix.netcom.com  Wed Apr 24 18:39:59 2002
From: feldspar at ix.netcom.com (Antaeus Feldspar)
Date: Wed, 24 Apr 2002 19:39:59 -0300
Subject: Art and all that Jazz:  are you ugly, I mean, attacking through   outer powders
References: <comp.lang.python.d42583a5@agozgi.sc.us> <3CC5E850.3080903@ix.netcom.com> <3CC62A08.77F6AC93@bellatlantic.net>
Message-ID: <3CC7343F.7030900@ix.netcom.com>

I missed them, unfortunately.  I only followed the links in the spam 
itself and nothing tipped me off that it was anything but what it 
appeared to be:  a spammer basing their business model on no one being 
able to stop them from stealing bandwidth.

I do wonder why she doesn't put a link to those pages explaining that 
she is being attacked *on* the page that the spam is publicizing, 
though.  It would seem to be the obvious countermove.

David Lees wrote:

> I may have missed some of the discussion, but as a naive (with respect
> to hipcrime) newsgroup reader I find an explanation at:
> http://www.artandallthatjazz.com/spam.shtml which claims that the spam
> was actually an an attack on the artist and her business.  Are you
> claiming that those pages are a fake or did you miss them?
> 
> David Lees
> 
> 
> Antaeus Feldspar wrote:
> 
>>This Jessica Felix really is a moron, isn't she?
>>
>>I mean, she steals newsgroup space for her commercial spam (the fact
>>that she's using HipCrime tools to spam *shows* that she knows she's in
>>the wrong) and expects that we'll *want* to do business with her after
>>she's shown herself to be contemptuous of courtesy and law?
>>
>>Jessica Felix wrote:
>>
>>
>>>Some durable cups are bizarre and other brave dusts are clean, but will Abu dream that?
>>>
>>>"ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
>>>that incorporates the two passions of its owner, Jessica Felix -
>>>art and jazz.
>>>
>>>Jessica is an artist whose medium is jewelry. She creates what she calls
>>>"spirited art," specializing in images from indigenous cultures around the world.
>>>
>>>http://www.artandallthatjazz.com/index-2.shtml
>>>
>>>119A Plaza Street
>>>Healdsburg, CA, USA 95448
>>>1-707-433-7900
>>>
>>>Occasionally, Kenny never attempts until Frank smells the think cloud tamely.
>>>
>>>They are sowing near sweet, under clean, with sick jugs.
>>>
>>>
>>>
>>>
>>>



From gtaylor at lowebroadway.com  Tue Apr 30 05:05:36 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 30 Apr 2002 02:05:36 -0700
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com>
Message-ID: <f0fd5987.0204300105.51529004@posting.google.com>

Chris <chris at cmb-enterprises.com> wrote in message news:<chris-4E0B1A.01281230042002 at corp.supernews.com>...
> In article <df30afd4.0204292111.31e5914a at posting.google.com>,
>  xiaotsing_79 at hotmail.com (David) wrote:
> 
> > I have a lot of experience in ANSI C/C++, I want to learn a new
> > language and which it can help me in the work.
> > I just want to clear which(python or perl) is more comfort for me?
> 
> In terms of how it looks, Perl might look more comfortable, but Python 
> is likely to function more similarly to C++, while C will have more in 
> common with Perl.
> 
> Learn both.

I use Perl at work day in day out, and use Python for my own personal
stuff, I would say that Python is better for pretty much everything,
particularly larger programs, where it's more structured syntax makes
code easier to read. I quite like Perl for CGI scripting, and little
1-page scripts, but would never contemplate using it for a big
project.

Learning both is an ideal solution, but personally I wish I started on
Python far earlier than I did.

Garry


From mlh at vier.idi.ntnu.no  Mon Apr 15 07:36:30 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 15 Apr 2002 11:36:30 +0000 (UTC)
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
References: <mailman.1018755873.29647.python-list@python.org>  <3CBA6595.104ED599@cosc.canterbury.ac.nz>
Message-ID: <slrnablepu.5k9.mlh@vier.idi.ntnu.no>

In article <3CBA6595.104ED599 at cosc.canterbury.ac.nz>, Greg Ewing wrote:
>Tim Peters wrote:
>> 
>> As
>> originally posted, they were the "20 Pythonic Theses".  They never need
>> revision, because I only gave 19 -- the 20th (or perhaps the first -- they
>> weren't numbered) was left open for Guido to fill in.
>
>Perhaps it should be "19 Pythonic theses is sufficient."

Or: "Number 20: There is *no*..... Number 20."

- Magnus "G'day Bruce" Hetland

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From logiplexsoftware at earthlink.net  Fri Apr  5 18:13:22 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 5 Apr 2002 15:13:22 -0800
Subject: PEP 285: Adding a bool type
In-Reply-To: <Npor8.8281$q63.818@atlpnn01.usenetserver.com>
References: <mailman.1017466974.6751.python-list@python.org>
	<2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk>
	<mailman.1017755045.4308.python-list@python.org>
	<uakpepcn1iir15@news.supernews.com>
	<mailman.1017865089.20676.python-list@python.org>
	<Npor8.8281$q63.818@atlpnn01.usenetserver.com>
Message-ID: <20020405151322.058b2471.logiplexsoftware@earthlink.net>

On Fri, 5 Apr 2002 16:09:41 -0500
Steve Holden wrote:

Steve, please be a bit more careful.  Your post sort of made it look like
you were responding to things *I* had said (which I've snipped here) which
were in fact said by John Roth.  I was even confused for a second =)

[snip]

> "Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
> news:mailman.1017865089.20676.python-list at python.org...
> > So what you are saying fits perfectly with what Laura asserted.  If you
> > have a function numberOfLeapDays() why on earth would you need
> > isLeapYear()?  The solution is simple: don't return a boolean when
another
> > type can not only express true/false but also convey other information
as
> > well.
> >
> It seems plain (now I've read so much of the discussion) that if book
isn't
> going to behave as a full-blown Boolean object it would be better to
leave
> well enough alone.

Well, it's been proven arguable ;)

> > I didn't necessarily agree with all of Laura's assertions regarding the
> > difficulty of understanding how booleans would fit into Python, but she
> > raised several compelling and thoughtful arguments.  I totally agree
that
> > bools as return values of functions are less than useful.  Booleans
have
> > the lowest information density of any type and thus would tend to
increase
> > code verbosity if their use were to be encouraged.
> >
> I too thought Laura made a thoughtful contribution to the discussion,
that
> deserved to be considered more respectfully.

I agree and in fact I've discussed this offlist with Guido.  He assures me
that he did carefully consider Laura's arguments.  Regardless, he didn't
agree and didn't feel it was necessary to respond directly to them as so
many other people already had.

> > Frankly Guido's dismissal of Laura's arguments as FUD was not only a
bit
> > rude, but IMHO a bit short-sighted as well (given the fact that
someday,
> > there may be a desire for a real boolean type).  Maybe Tim Peters has
> > borrowed his time-machine and not returned it ;)
> >
> However it happened, it left me feeling a little embarrassed that I had
> originally defended Guido's inegrity and suggested that despite his
> authorship the PEP's acceptance was not a foregone conclusion. I hope I
> wasn't wrong, but that's not how it looks right now. Oh well.

While I disagree with Guido's decision, I don't really think it has
anything to do with his integrity (although it certainly could give that
appearance).  I was fairly irritated at how this change went about (enough
to contact Guido personally about it) and despite not reaching agreement on
the decision, I do feel certain that the issue was properly considered on
his part and that he simply came to a different conclusion than many others
(myself included).

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mlh at vier.idi.ntnu.no  Wed Apr 10 22:28:57 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 02:28:57 +0000 (UTC)
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>  <3CB32C83.AB67AFFC@alcyone.com>
Message-ID: <slrnab9t79.6g8.mlh@vier.idi.ntnu.no>

In article <3CB32C83.AB67AFFC at alcyone.com>, Erik Max Francis wrote:
>Magnus Lie Hetland wrote:
>
>> That's not how I saw it. The main point I extracted was: Until now,
>> the Pythonic "truth values" have been "something" and "nothing" (i.e.
>> the empty list, zero, None, etc.) With the bool type this is still
>> true, but we get a type which will give newbies the impression that it
>> isn't.
>
>So explain it to them.  It only takes a sentence or two.  The bool type
>just provides one more value that's a "nothing," and one more value
>that's a "something," for cases where what you're interested in is truly
>nothing but a truth functional value.

Sure. I was only trying to clarify Laura's description of Something
and Nothing (as I saw it), because I like the description.

Although I think the new bool type was completely unnecessary, it
won't cause great problems for me, and I don't think I'll have any
problems explaining how it works to newbies (though I may have some
problems explaining why it exists).

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From Kaisa-Maija.Saarimaa at nokia.com  Fri Apr 26 08:05:17 2002
From: Kaisa-Maija.Saarimaa at nokia.com (Kaisa-Maija.Saarimaa at nokia.com)
Date: Fri, 26 Apr 2002 15:05:17 +0300
Subject: Python training in Finland
Message-ID: <B162EEC71A513847970BB6EC1C7F07B80164E0A0@trebe001.NOE.Nokia.com>

Hi,
Does  anyone know who offers python training in Finland? Contact information?

Thanks for the help!

********************************************************

Kaisa-Maija Saarimaa

Product Manager
Nokia Learning Center Network Europe 1

P.O.Box 779/Hatanp??nkatu 1A
33101 TAMPERE, Finland

tel. +358 (0)7180 74555
mobile +358 40 513 4119
fax +358 (0)7180 75700
mailto:kaisa-maija.saarimaa at nokia.com
********************************************************






From dsavitsk at e-coli.net  Wed Apr 10 20:47:30 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 11 Apr 2002 00:47:30 GMT
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <%34t8.36244$CU.41218@rwcrnsc51.ops.asp.att.net>
Message-ID: <C25t8.5982$d7.1714559@newssrv26.news.prodigy.com>

I guess this proves TMTOWTDI.

"Emile van Sebille" <emile at fenx.com> wrote in message
news:%34t8.36244$CU.41218 at rwcrnsc51.ops.asp.att.net...
> Rajarshi Guha
> >   I have a list like this:  l = [1,'s', 2,'t', 3,'d']
> >
> [snip]
> > whereas I want
> >
> > 1 s
> > 2 t
> > 3 d
> >
>
> >>> for x,y in [(x,y) for x,y,z in zip(l,l[1:],[1,0]*len(l)) if z]:
> ...     print x,y
>
>
> Have fun...  ;-)
>
> --
>
> Emile van Sebille
> emile at fenx.com
>




From dalke at dalkescientific.com  Sun Apr  7 20:31:47 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 7 Apr 2002 18:31:47 -0600
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net> <ub1mp3ssjc7da5@news.supernews.com>
Message-ID: <a8qog4$900$1@slb1.atl.mindspring.net>

John Roth replied to a comment of mine.  As my response was even
more off-topic, I sent email to him, which bounced, saying:

] The following destination addresses were unknown (please check the
] addresses and re-mail the message):
]
] <johnroth at ameritech.net>
]
] Please reply to techsupport at ameritech.net
] if you feel this message to be in error.


John?  Did you know this?  (It doesn't look spam protected, after
all.)  If you want the reply from me, could you send me a valid
email address?

                    Andrew
                    dalke at dalkescientific.com





From loewis at informatik.hu-berlin.de  Wed Apr 17 11:26:34 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Apr 2002 17:26:34 +0200
Subject: minidom and =?iso-8859-15?q?=E5=E4=F6?= once again :-P
References: <WXdv8.2962$ZN4.5053@news2.bredband.com> <9sev8.3366$ZN4.5195@news2.bredband.com>
Message-ID: <j43cxubb2t.fsf@informatik.hu-berlin.de>

Magnus Heino <magnus.heino at pleon.sigma.se> writes:

> If i replace ??? below with codecs.utf_8_encode('???')[0], then
> parseString is happy.

I'm not sure what precisely you are replacing. In the code, those
three characters only appear inside of a string literal.

> toxml() doesnt produce a u'' string though, so I have to do 
> unicode(doc.toxml()) and then feed that into parseString.

Yes, that's a bug in StringIO of Python 2.2.

Regards,
Martin


From gcordova at hebmex.com  Fri Apr 26 19:09:36 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 26 Apr 2002 18:09:36 -0500
Subject: Argh, I give up.
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339F6A5D@mail.hebmex.com>

Hi y'alls.

I've been playing around with Bloodshed C++, which is an IDE
for C and C++ for Win32, which comes with the MingW32 version
of GCC and friends.

The thing is, I can't get it to make a valid DLL (I'm playing
around with the TkIcon module I found on the net... you know
who you are :-)

It creates an .o file, ready for linking, but Python 2.2's .lib
files seem to be undigestible to the linker, which simply complains
that it can't find any python libraries.

It does come with a lot of it's own libs, which it can find
perfectly. So the question is, What can I do to make MingW32
properly understand python's *.lib files, which are needed for
linking?

Your help is much appreciated.

-gustavo




From phd at phd.pp.ru  Tue Apr  2 03:31:42 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 2 Apr 2002 12:31:42 +0400
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
In-Reply-To: <LNBBLJKPBEHFEDALKOLCGEOIOJAA.tim.one@comcast.net>; from tim.one@comcast.net on Tue, Apr 02, 2002 at 12:48:58AM -0500
References: <qzbq8.3238$tW2.2164@nwrddc04.gnilink.net> <LNBBLJKPBEHFEDALKOLCGEOIOJAA.tim.one@comcast.net>
Message-ID: <20020402123142.D25705@phd.pp.ru>

On Tue, Apr 02, 2002 at 12:48:58AM -0500, Tim Peters wrote:
> Someone would.  Someone always does.  Some people even fight to keep gross
> bugs intact, although not yet to preserve segfaults.

   One does not need to preserve them - there are enough operating systems
that do it "bigger, better, faster" than any interpreted programming
language can achieve ;-) Ah, the joys and wonders of programming...

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



From tim.one at comcast.net  Fri Apr 12 23:21:15 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 12 Apr 2002 23:21:15 -0400
Subject: hex and long
In-Reply-To: <8N9+ELA0u5t8Ew2n@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEMIONAA.tim.one@comcast.net>

[Robin Becker]
> In 2.2 I find the following peculiarity
>
> >>> 0xffffffff
> 68719476735L

That looks an error in transcription; I expect you dropped an "f" character
in the input.

> >>> 0xffffffff
> -1
> >>>
>
> clearly the second is being treated as a 32 bit int and the first as a
> long.

Yes.  For more, see PEP 237:

    http://www.python.org/peps/pep-0237.html

> Do I have to be careful with bit 31?

Yes, and that's always (not unique to 2.2) been true on 32-bit boxes.

> What happens on 64 bit machines?

It depends on which 64-bit machine you're talking about.  On one where
sizeof(long)==8, both literals above are positive Python ints; on one where
sizeof(long)==4, same as on a 32-bit box (the first a positive long, the
second a negative regular, int).  If Python has any users left by the time
PEP 237 is fully implemented, they'll enjoy platform-independent results;
however, that can't be achieved without changing the platform-dependent
results currently exposed.

In the meantime you can force interpretation as a postive integer literal
across all platforms by appending "L" to your hex and octal literals.





From elimli at yahoo.com  Sun Apr 28 13:47:33 2002
From: elimli at yahoo.com (eli li)
Date: Sun, 28 Apr 2002 19:47:33 +0200
Subject: cookies ???
Message-ID: <aah926$k1a$1@news.iucc.ac.il>

I "open" http sites using urllib2 library. Method read() retunes content
part only, in other words html code is returned. How can I retrieve cookies?




From ale_xp at yahoo.com  Sat Apr  6 06:03:14 2002
From: ale_xp at yahoo.com (alex)
Date: 6 Apr 2002 03:03:14 -0800
Subject: help:problem using the RawPen class/turtle module.
Message-ID: <8374ee10.0204060303.3333260a@posting.google.com>

Hi all, I'm a Python beginner, and I'm trying to use
the module "turtle" to do some turtle graphics.
But I have a problem when I try to use the turtle
class RawPen in order to use it with a Canvas. 
In particular when I try to run the following code: ...

...
root = Tkinter.Tk()
c       =   Tkinter.Canvas(root,heigh=200,width=300)
c.pack()
mypen   =   turtle.RawPen(c)
... 

I have the following error message:   

  File "l-turtle2.py", line 56, in ?
    mypen=turtle.RawPen(c)
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 16, in __init__
    self.reset()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 42, in reset
    self.clear()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 53, in clear
    self._draw_turtle()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 262, in _draw_turtle
    self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
AttributeError: 'NoneType' object has no attribute 'create_line' 

The Python Library Reference documentation section 16.4 (Turtle)
shows that the RawPen class has to be called with a canvas as argument.
I don't know if there is other documentation on this topic.

thank you in advance.

Alex


From wilk at flibuste.net  Tue Apr 30 03:19:16 2002
From: wilk at flibuste.net (William Dode)
Date: Tue, 30 Apr 2002 09:19:16 +0200
Subject: Python sous UNIX
References: <aak1q3$fgl$1@wanadoo.fr>
Message-ID: <20020430091916.3b6933b9.wilk@flibuste.net>

Le Mon, 29 Apr 2002 20:01:11 +0200
"infocap" <infocap at wanadoo.fr> ?crivait:

> Bonjour ? tous,
> 
> Je travaille pour une entreprise d'?dition mais voil? que depuis
> quelques temps je m'int?resse beaucoup ? l'informatique et
> particuli?rement au langage de programmation PYTHON  sous UNIX et
> j'aimerai approfondir mes connaissances au travers de petits programmes
> int?ressants, alors, si quelqu'un peut satisfaire ma demande, je le
> remercie d'avance.
En fait, j'ai cherch? ?galement, il n'y a pas grand chose...
Pourquoi ? tout simplement parce qu'en python il est beaucoup plus rapide
d'?crire soit-m?me les petits programmes plut?t que de les chercher ;-)
Par exemple, hier soir je me suis fait un petit visualiseur de photo avec
rotation/supression/sauvegarde apr?s avoir cherch? pendant des heures et
des heures un ?quivalent (sous linux). Note que j'ai apris le langage il y
a quelques semaines seulement.

En fran?ais, rejoint nous sur fr.comp.lang.python et sur la mailing-list 
http://www.aful.org/mailman/listinfo/python

Puisque tu es dans le domaine de l'?dition je te conseille d'aller jeter
un oeuil sur http://www.reportlab.com, une librairie pour g?n?rer du pdf. 
J'ai ?galement mis un exemple sur mon site http://flibuste.net/pythreries
qui utilise des notions de pagetemplate et frame assez proche des
principes de l'?dition.

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From pobrien at orbtech.com  Wed Apr  3 15:14:44 2002
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Wed, 3 Apr 2002 14:14:44 -0600
Subject: [Python-Dev] PEP 285: Adding a bool type
In-Reply-To: <200204031946.g33JktP25857@odiug.zope.com>
Message-ID: <NBBBIOJPGKJEKIECEMCBOEFHMLAA.pobrien@orbtech.com>

[Guido van Rossum]
> My thoughts haven't changed at all. :-)
>
> The first round of discussion (on python-dev) brought up some
> contentious issues where people disagreed strongly with particular
> details proposed in the PEP, and I put those contentious issues at the
> top of the review list.  If you go to the latest version of the PEP on
> the web (http://www.python.org/peps/pep-0285.html) you'll see that
> I've added some other issues that were brought up in the more recent
> discussion.

I apologize for not having read the web version. I wasn't aware that it was
being updated during the debate. I have a feeling I'm not alone in this
mistake, so I'll take this opportunity to encourage others to read the web
version also.

> There's an educational issue that I'm just getting my head around.  It
> appears that for various psychological and linguistic reasons, newbies
> naturally write
>
>     if b == True: ...
>
> when they first encounter a bool variable and have to test whether
> it's true.  I believe that this is most of Laura's issue.  The only
> reasonable solution for this is that the Zen master slaps them over
> the head with a stick, and they learn to write
>
>     if b: ...

Yes. And I think describing Python's handling of truth values as "something
vs. nothing" has real merit as a learning aid. It will certainly keep me
from falling into the if b == True: trap.

In summary, I agree with your clarifications and I'm +1 on the PEP.

---
Patrick K. O'Brien
Orbtech





From ianb at colorstudy.com  Fri Apr 26 18:34:58 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 26 Apr 2002 17:34:58 -0500
Subject: cgi module question
In-Reply-To: <3CC9BC7F.5020305@nmt.edu>
References: <3CC9A3E2.9060803@nmt.edu> <kud6wmb6wt.fsf@lasipalatsi.fi> 
	<3CC9BC7F.5020305@nmt.edu>
Message-ID: <1019860498.2058.1458.camel@localhost>

On Fri, 2002-04-26 at 15:45, Jeff Shipman wrote:
> Uh, no, not really because it's consumed by
> FieldStorage(). If I consume it before that, then
> there's FieldStorage() can't grab it. I still
> need to use the cgi module, but I also need
> to get the POST query string.

You can read in the POST string, and then pass it to the cgi module (as
a file-like object with StringIO).  There's an optional argument to
FieldStorage -- look at the source to find it.

  Ian






From claird at starbase.neosoft.com  Wed Apr 24 10:43:07 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2002 09:43:07 -0500
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at>
Message-ID: <C242536810264D16.451F61A772009873.DAA94F10F7E2E70F@lp.airnews.net>

In article <a9rn4i$mm4$1 at newsreader1.netway.at>,
Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
			.
			.
			.
>It is correct that they are not Algol-like; but map, apply,lambda are
>typical for functional languages like Clean, Haskell,...
			.
			.
			.
These--lambda, ...--were relatively late developments in Python's
accumulation of features.
-- 

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


From todd at ppgweb.com  Fri Apr 12 18:18:31 2002
From: todd at ppgweb.com (Todd Chapman)
Date: Fri, 12 Apr 2002 22:18:31 GMT
Subject: PyQt Installation Problem...
References: <3CB749CB.10701@slideburn.com>
Message-ID: <3CB75DA5.9000709@ppgweb.com>

Hold your horses....did a make uninstall followed by another make 
install and it worked...hmmmmm

todd

Todd Chapman wrote:

> Hello,
> 
> Just installed SIP and PyQt and have come across a slight twist.  I'm on 
>  Red Hat 7.2 and after compiling, PyQt installs in the directory where 
> it was extracted and not in the site-packages subdirectory of my python 
> installation.  Looks like the files that were generated in the cong 
> stage got located the right python version and directory, it just didn't 
> go there (/usr/lib/python2.2/site-packages).  Any suggestions?
> 
> Todd
> 



From dsavitsk at e-coli.net  Sun Apr  7 18:51:01 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 07 Apr 2002 22:51:01 GMT
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com>
Message-ID: <p34s8.5367$d7.1394275@newssrv26.news.prodigy.com>

for win32, there is a wrapper to the VB printer object at
www.e-coli.net/software

-d


"Jim Williams" <jimw1956 at cs.comnojunk> wrote in message
news:20020407103654.19127.00002640 at mb-fm.news.cs.com...
> Hi all...I'm new to Python and really enjoying it so far, but I am
somewhat
> perplexed that as far as I have seen it has no facilities for printing to
a
> printer?  Surely this is possible. Someone please enlighten me...I have
done a
> good bit of research and searching and can find no reference to how to do
this.
>
>




From shower at wowlinux.com  Tue Apr  2 00:29:09 2002
From: shower at wowlinux.com (Jinsub Sin)
Date: Tue, 2 Apr 2002 14:29:09 +0900
Subject: is ok
Message-ID: <00c601c1da07$5431ca60$1f57b5d2@shower>

confirm 612609

From huaiyu at gauss.almadan.ibm.com  Tue Apr 16 13:09:47 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 16 Apr 2002 17:09:47 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabm5cp.nkd.huaiyu@gauss.almadan.ibm.com> <87ofgjwzl5.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <slrnabommq.q79.huaiyu@gauss.almadan.ibm.com>

Stephen J. Turnbull <stephen at xemacs.org> wrote:

>And I admit a slight theoretical possibility I could be wrong.  But go
>look at the license lists---they're the same.  Except for the original
>Perl "Artistic License", which is OSI-certified but in the "non-free"
>group at GNU.  So look close---it's a technical DQ for being _too
>vague_!  Ie, it's not that it's not free---it's that it's not really a
>license according to the FSF.  The exception that proves the rule.

The Apple Public Source License (APSL) version 1.2 is listed as "Non-Free
Software License" on http://www.gnu.org/licenses/license-list.html, mainly
because of "Possibility of revocation at any time".

It is "Open Source Initiative Approved License" listed on
http://www.opensource.org/licenses/index.html.

To me, one counterexample is enough, and one with a clear explanation is
very satisfying.

OTH, I agree with you that to the general public the difference will never
look as big as RMS would want it to be.

Huaiyu


From skip at pobox.com  Mon Apr 22 18:47:33 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 22 Apr 2002 17:47:33 -0500
Subject: using automake/autoconf to build extension module?
In-Reply-To: <m34ri34d5w.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1019509602.30761.python-list@python.org>
        <m34ri34d5w.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15556.37637.923236.338797@12-248-41-177.client.attbi.com>

    >> I'd prefer not to use distutils if possible, since the rest of the
    >> project is using automake/autoconf/libtool for build configuration.

    Martin> That does not necessary contradict each other. You can easily
    Martin> embed Python fragments into a configure.in, to determine the
    Martin> appropriate command line options.

I'll take your word for it, but it irks me to no end that distutils can't do
dependencies.  Having that means I have to hack things to force files to be
deleted I shouldn't have to.  If I can use make as the build tool I'd much
prefer that.

Skip




From dsavitsk at e-coli.net  Thu Apr  4 00:05:05 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 04 Apr 2002 05:05:05 GMT
Subject: Newbie: Alarm at end of program
References: <es5HyN42BHA.1996@cpimsnntpa03> <a8gfr8$h4u$1@maud.ifi.uio.no>
Message-ID: <5aRq8.3571$d7.1106160@newssrv26.news.prodigy.com>

on win32 also try

>>> import winsound
>>> winsound.Beep(440,250)

the first number is the frequency, the second is the duration.

-d


"Trond Gaarder" <trondga at -NOSPAM-ifi.uio.no> wrote in message
news:a8gfr8$h4u$1 at maud.ifi.uio.no...
> Don't know if this is the best/correct way of doing it in Python, but it
> works for me (on a PC):
>
> print "\a"
>
> This should emit a beep from the pc. What it sounds like is highly
> machine-dependent, however...
>
>
> -Trond
>
>
> "billludden" <billludden at email.msn.com> wrote in message
> news:es5HyN42BHA.1996 at cpimsnntpa03...
> > Hi,
> > I have written a script that runs from a few minutes to several hours
> before
> > it finishes.  How would I make the computer play a 'beep' to signal that
> it
> > is finished?
> > Thank you,  Bill
> >
> >
>
>




From duncan at NOSPAMrcp.co.uk  Tue Apr 30 12:01:02 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 16:01:02 +0000 (UTC)
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1> <mailman.1020172854.26653.python-list@python.org>
Message-ID: <Xns9200AC87B7A86duncanrcpcouk@127.0.0.1>

Roman Neuhauser <neuhauser at mail.cz> wrote in 
news:mailman.1020172854.26653.python-list at python.org:

>> I prefer writing
>>    str.join(' ', aList)
>> rather than
>>    ' '.join(aList)
>> 
>> When the string is a literal I find the former reads much better. YMMV
> 
>     IMO that's because join being a string method is just awkward. How
>     does this read:
>     aList.join(' ')
>     After all, you tell the list to give you it's concatenation, using
>     the argument as the glue.

That doesn't work if you want to join a tuple, or an iterator or any other 
sequence. Dont forget you can even join a string:
>>> print str.join(',', 'abc')
a,b,c



>> Sometimes though I take advantage of it being a method:
>>    joinlines = '\n'.join
>>    ...
>>    s = joinlines(aList)
> 
>     This is not a method. You masked it as a global function.

No that is taking advantage of the fact that methods can be bound and the 
bound method then called just like any other callable object. If join 
wasn't a method it wouldn't be possible to bind it to a particular string 
in this way.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From mlh at vier.idi.ntnu.no  Tue Apr  2 16:50:21 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 2 Apr 2002 21:50:21 +0000 (UTC)
Subject: Streaming XML-RPC?
References: <slrna9ua49.74f.mlh@vier.idi.ntnu.no>        <3C9F51EE.5000203@jerf.org>        <3C9FAF96.5DE9A0@bgb-consulting.com>  <mailman.1017105342.20070.python-list@python.org> <slrnaa1gb9.87e.mlh@vier.idi.ntnu.no>  <a81584$2gr2$1@news.idiom.com> <slrnaaf8bk.ok0.mlh@vier.idi.ntnu.no>  <a88mdk$2pcv$1@news.idiom.com>
Message-ID: <slrnaak9st.s4j.mlh@vier.idi.ntnu.no>

In article <a88mdk$2pcv$1 at news.idiom.com>, Jim Dennis wrote:
[snip]
> IANAL, but:
>
> I thought that PSF and Guido came up with a GPL compatible license.

IANAL either, but:

Yup. But there's a difference between the Python license being GPL
compatible (i.e. you can distribute Python and friends under the GPL)
and the GPL being compatible with the Python license (i.e.
distributing GPL'ed stuff under the Python license.)

The GPL makes requirements that the Python license does not.

> Since I'm not releasing any code that mixes GPL and Python code
> I haven't scrutinized the licenses in detail.  So I don't know which
> versions of Python are covered by GPL compatible licenses, nor am I
> sure that the two licenses *are* actually compatible.

I think the crucial point is the way in which they are compatible.
(And they are not equivalent.)

E.g. you can distribute software under the Python license (or
derivatives) as closed-source commercial stuff, without providing an
open-source alternative. AFAIK, that is not allowed under the GPL.
Thus, releasing a GPL'ed package under the Python license would be
impossible (unless you were the owner, of course).

> However, it seems that the LGPL should be freely linkable even to 
> proprietary or other non-GPL code.

Yes.

> I was under the impression that
> this was the whole point of the *L*GPL.

Yes. But it still makes the same basic requirements as the GPL
(AFAIK). I couldn't redistribute the code itself under the Python
license (or any other non-(L)GPL open source license, I should
imagine).

At least this is my impression of how things work (and why so many
people rewrite GPL code and release non-GPL versions so that they can
be used in contexts such as the Python standard library). I may be
wrong, of course.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From cimarron+google at taylors.org  Fri Apr  5 02:56:50 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 4 Apr 2002 23:56:50 -0800
Subject: "Don't Kill Opera" documentation release
References: <mailman.1017709803.3793.python-list@python.org> <a8j443$omncm$1@fido.engr.sgi.com>
Message-ID: <29e28c51.0204042356.21bc1d32@posting.google.com>

pj at engr.sgi.com (Paul Jackson) wrote in message news:<a8j443$omncm$1 at fido.engr.sgi.com>...
> Could you explain easily what you had to change to the Docs
> in order to satisfy Opera 6.01?  As an Opera user, I am
> curious what was the bug in Opera that you were working
> around (or at least I assume that's the situation).

It may have to do with the loading of jpg images.  I reported a
similar problem a while ago.  http://www.taylors.org/cim/OperaRich.gif
shows you what I get when I visit http://www.dost.com with Opera 6.01
without images toggled off.

Cim


From ASkwar at DigitalProjects.com  Thu Apr 25 03:00:08 2002
From: ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Thu, 25 Apr 2002 09:00:08 +0200
Subject: Faster md5/1-way encryption?
In-Reply-To: <3CC7317B.3040208@rogers.com>
References: <20020424215752.GC15371@teich.Garten.DigitalProjects.com> <3CC7317B.3040208@rogers.com>
Message-ID: <20020425070008.GA10855@teich.Garten.DigitalProjects.com>

?Mike C. Fletcher? sagte am 2002-04-24 um 18:28:11 -0400 :
> Something that hits me right off the bat:
> 
> >             md5_zeile = md5.new(zeile).hexdigest()
> >             self_Inhalt += md5_zeile + linesep
> 
> Would likely be much better as:
> 
> 	myResults = []
> 	...
> 	myResults.append( md5.new(zeile).hexdigest() )
> 	...
> 	myResults = string.join( myResults, os.linesep )

Yes, that makes sense.  Thanks!

> Which would save you creating and destroying huge numbers of (at the end 
> of processing) very large strings.  If you can avoid that last step 

Yes, I also noticed, that it gets slower and slower over the time.

> can replace your 'i' counter in that approach (won't be a noticable 
> savings, I just hate counters :) ).

Sure, nice idea.

> accurate result as to current position.  Unless you really need that 
> accuracy, consider just accumulating the length of the lines you're 
> processing and adding 1 or 2 each time for the length of os.linesep.

Yes, of course.  Good idea here as well.

> MD5 can process gigabytes of information pretty quickly, so I'd be 
> surprised if a 100,000 line file is a huge problem.

Yes, that's right, but in the case of processing gigabytes of data,
you're just invoking the digest generation once and not over and over
again; and you're also creating the object once and not 100,000 times.

Thanks for your hints!

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 1 hour 56 minutes




From chris at psychofx.com  Wed Apr 10 13:53:34 2002
From: chris at psychofx.com (Chris Miles)
Date: Thu, 11 Apr 2002 03:53:34 +1000
Subject: Help: Python for a big commercial project?
In-Reply-To: <3b091a1c.0204100926.4ba2f13b@posting.google.com>; from cmkleffner@gmx.de on Wed, Apr 10, 2002 at 10:26:35AM -0700
References: <mailman.1018437980.14078.python-list@python.org> <3b091a1c.0204100926.4ba2f13b@posting.google.com>
Message-ID: <20020411035334.A2828@psychofx.com>

I successfully built and use Python 2.2 on HPUX 11.00 with little effort (using
gcc 2.95.2).  My application relies on threading and works fine, so I would be
confident in developing a Python app on HPUX.

CM

On Wed, Apr 10, 2002 at 10:26:35AM -0700, cmkl wrote:
> I agree: IT IS PAINFUL learn to build Python on HPUX but nevertheless it is 
> worth doing. DONT'T use the prebuilt buggy python versions of the 
> HPUX porting Center !! Build python yourself, it is then much easier to get 
> working C-extensions like database modules, numeric python ...
> 
> threads is an issue on HPUX platforms. With HPUX11 I was sucessful to build
> thread enabled python (tests sucessful) but I was not able to build a working 
> multithreaded database application with a multithreaded enabled Informixdb
> module. But it may be possible. It is a matter of time for trying and testing
> different settings at compile time. 

-- 
Chris Miles
chris at psychofx.com
http://www.psychofx.com/chris/




From claird at starbase.neosoft.com  Thu Apr 11 10:39:12 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 11 Apr 2002 09:39:12 -0500
Subject: Help: Python for a big commercial project?
References: <F194tIpqqJgSt51NDa200002509@hotmail.com> <mailman.1018464082.15807.python-list@python.org> <roy-EB08BB.20470710042002@news1.panix.com>
Message-ID: <60F7BB039BFDD68F.7FE6889DFED7EA98.93812C99068FD653@lp.airnews.net>

In article <roy-EB08BB.20470710042002 at news1.panix.com>,
Roy Smith  <roy at panix.com> wrote:
			.
			.
			.
>I think one of the keys to a successful large-scale python project is 
>understanding its strengths and weaknesses.  Take advantage of the former, 
>and don't try to brute-force your way past the latter.
>
>For example, python is not very good at low-level character manipulation.  
>If you're parsing text, for example, and are used to typical C-style 
>tokenizer loop:
>
>   while (c = getchar()) {
>      switch (c):
>         blah, blah
>
>and try to translate that directly into python, you'll probably end up with 
>something which is slow as a dog (DAMHIKT).  The key is to figure out which 
>little bits of your code would benefit most from re-writing them as a C 
>module, and doing that.  You may discover that writing 2% of your code in C 
>will give you something that's 10 times faster than the pure python 
>implementation.  Without having looked at the code, I'm guessing that the 
			.
			.
			.
Let me reinforce this message.

Python is indeed "not very good at low-level character
manipulation" (and other operations) *in "typical C-style"*.
It FREQUENTLY is the case, though, that rewrites in good
Python style result in code which is more maintainable, only
a fraction as large in lines-of-source-code, and, remarkably,
at least as fast as the hand-coded C.  Why and how this latter
is possible have been explained here often in the past; I'll
abbreviate the details by just saying that Python's run-time
library is itself written in C (at least for conventional 
CPython), and supports many of the actions programmers most
need.
-- 

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


From psage at ncaustin.com  Thu Apr 11 16:50:07 2002
From: psage at ncaustin.com (Paul Sage)
Date: Thu, 11 Apr 2002 15:50:07 -0500
Subject: Python Mega Widgets: Counter
Message-ID: <03F53405C6ED434A986221919785A6A78C8E44@nca-postal.ncaustin.com>

Does anyone know if it is possible to change the value in a Counter
Entry field without it missing a beat?  I could get the value difference
between what I want and tell it to increment or decrement the
difference, but that seems awfully complex, though maybe a bit safer.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020411/f6818864/attachment.html>

From rjones at ekit-inc.com  Tue Apr  2 20:22:58 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 11:22:58 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <3CAA562E.C5D592A2@engcorp.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com> <3CAA562E.C5D592A2@engcorp.com>
Message-ID: <200204030123.BAA15136@crown.off.ekorp.com>

On Wed, 3 Apr 2002 11:09, Peter Hansen wrote:
> Paul Rubin wrote:
> > David Goodger <goodger at users.sourceforge.net> writes:
> > > Here's a serious proposal, safe to post now that April Fool's is over.
> > > Please read & comment.
> >
> > Man, I thought it was an April fools joke when I saw it last night.
> > All those `different` :kinds: 'of' "quoting" are much more confusing
> > than explicit markup.  So I think it's better to use explicit markup.
>
> I had the same impression.  I was reading it for the first time,
> as I showed it to a coworker who was interested.  By the time we'd
> scrolled through the ReStructured Text documentation I had to say
> to him, "Hmm... that's so bloody complicated it might be easier
> writing XML tags directly.  Sort of defeats the purpose of structured
> text, doesn't it?"
> 
> When structured text becomes so complicated to type that one has to
> actually work at it, and it needs a large manual to use, it has
> become too unwieldy to be useful.

Are you sure you aren't confusing comprehensive (and, unlike most STX-alike 
documentation formats out there, well-designed) with complicated?

For most people, the first quarter or third of the quick reference (inline 
markup, sections, lists, literals) is plenty enough documentation.

   http://structuredtext.sf.net/docs/quickref.html

The rest of the spec is for advanced use only.


> My suggestion would be go with a simpler Structured Text for the 80%
> of documentation needs it, and allow an escape to DocBook or
> something for more complicated docs...

So we'd have to learn DocBook, or some other [SG|X|HT]ML layout? I'd 
guarantee that the STX version is going to be much easier to edit and 
maintain (without purchasing Frame, that is ;)


    Richard



From jaf at lcsaudio.com  Wed Apr 24 03:34:07 2002
From: jaf at lcsaudio.com (jaf)
Date: Wed, 24 Apr 2002 00:34:07 -0700 (-0700)
Subject: StringIO vs array.tofile() ?
Message-ID: <301735654963-BeMail@sugoi>

Hi all,

I'm seeing some behaviour in Python that I didn't expect.  Is it a bug
in the language, or am I doing something wrong?
Here is an example program to demonstrate the problem:

-----------------------------------------------------------------
import array
import StringIO

a  = array.array('l', [1])
io = StringIO.StringIO()
a.tofile(io)
print io.getvalue()  # we never get here
-----------------------------------------------------------------

I would expect this program to print out the binary representation of
the array (a) and exit.
Instead, I get this TypeError exception:

$ python test.py
Traceback (most recent call last):
  File "test.py", line 9, in ?
    a.tofile(io)
TypeError: arg must be open file

Why is that?  Isn't a StringIO a valid "open file"?

Thanks for any advice,
Jeremy

ps I tried with a cStringIO too, it made no difference





From erno-news at erno.iki.fi  Wed Apr 17 01:01:48 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 17 Apr 2002 08:01:48 +0300
Subject: Why don't strings share data in Python?
References: <mailman.1018937610.11880.python-list@python.org> <87n0w37z0n.fsf@mathdogs.com>
Message-ID: <kuy9fmgbpf.fsf@lasipalatsi.fi>

In article <87n0w37z0n.fsf at mathdogs.com>, Mike Coleman
<mkc+dated+1021607974.93924d at mathdogs.com> writes:

| I almost thought I might be able to use NumPy arrays, as it looks
| like they do share data.  They're not hashable, though,
| unfortunately, which kills the usage I had in mind.

reminds me of the buffer object. you can have a buffer object
representing 'oba' if you do buffer('foobar', 2, 3). i believe this
does what you awnt memory management wise. the re module can work on
buffer objects, maybe some other things can too.

  -- erno


From alanmk at hotmail.com  Thu Apr  4 16:47:42 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 4 Apr 2002 13:47:42 -0800
Subject: problem with http post
References: <cf43c01d.0204022342.73e5c8fa@posting.google.com>
Message-ID: <f72dac0d.0204041347.3b71bd26@posting.google.com>

Warning: I haven't tested any of this.

mongiaamit at yahoo.com (Amit Mongia) wrote:
> See i have a small problem.

...

> But the following code gives me an "Invalid number of arguements"
> error.
> 
> encdata = 'myaddress=addressvalue1'
> conn = httplib.HTTP("karlulbrich.com")
> conn.putrequest("POST", "/address/accept/")
> conn.putheader('Content-type', 'application/x-www-form-urlencoded')
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 
This declaration is telling the server that the parameters are
"url-encoded", so it will expect to see something like this :-

http://karlulbrich.com//address/accept/?myaddress=addressvalue1

But you're sending the form data as a part of the http message body,
so the server might be being confused by this.

If you want to send the data as a part of the message body, you should
set the content type to "multipart/form-data"

So you could try this

conn.putrequest('POST', '/address/accept/')
conn.putheader('Content-type', 'multipart/form-data')
conn.putheader('Accept', 'text/plain') 
conn.endheaders()
conn.send(encdata)

You might want to have a read of the section in the HTML 4 spec about
Form Submission. I guarantee it will clear matters up for you (a
little, at least).

http://www.w3.org/TR/REC-html40/interact/forms.html#submit-format

HTH,

Al.


From maric.michaud at cirec.com  Wed Apr 10 09:44:51 2002
From: maric.michaud at cirec.com (Maric MICHAUD)
Date: Wed, 10 Apr 2002 15:44:51 +0200
Subject: Problem with PIL
Message-ID: <27E327AB8201B54495D54105FBBBB1485D130D@SRVMSG.GERLAND2.local>

I tried ro include tif image in a pdf document with reportlab and Python
2.0, PIL 1.1.3.
This work pretty well with uncompressed and 'Pack Bits" compressed TIFF
files, but not with "CCITT group 4" that I need.
Anyone has a suggestion for this ?




From viking_kiwi at yahoo.com  Tue Apr 30 18:57:24 2002
From: viking_kiwi at yahoo.com (Rod Stephenson)
Date: 01 May 2002 00:57:24 +0200
Subject: Opening UNC style filenames with tkFileDialog... 
Message-ID: <wk6628pzfu.fsf@yahoo.com>

I've installed an application together with some data files on a
network server, but having problems with users on the network opening
the data files using askopenfile() from the tkFileDialog Tkinter
module. The application starts in the directory on the server, and
succesfully opens some hardcoded data files in the application directory.

However on trying to open a user selected file when running the app
from a machine on the network, I get the error message
...
File "tkFileDialog.pyc", line 108, in askopenfile
IOError: [Errno 2] No such file or directory: '//server/drrod/test1.dat'

It's odd since I can: 

--from a remote machine issue:
>>> open("//server/drrod/test1.dat")
at the python prompt and open the file OK.

--from the server, call up the file dialog and open the file either as
  a local filename, or through the network neighborhood on that machine.

--from the remote machine the Tcl script "open [tk_getOpenFile]" works
  perfectly happily as well. 

The application is compiled to a standalone .exe using py2exe; I may
resort to installing python on the server and seeing if this solves
the problem, though first I will probably first try sticking some
debugging code into tkFileDialog.py. But any other suggestions as to
what the problem might be would be appreciated.

-- 
Rod Stephenson


From tejarex at yahoo.com  Mon Apr  8 01:29:10 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 08 Apr 2002 05:29:10 GMT
Subject: Function question
References: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu>
Message-ID: <GU9s8.35530$l7.3694482@bin7.nnrp.aus1.giganews.com>

"John" <nobody at nobody.com> wrote in message
news:zl9s8.37442$tg4.444316 at vixen.cso.uiuc.edu...
> Let's say there is a function with two arguments.
> def add(a,b): return a+b
>
> What I want to do is:
> I want to get 'a' dynamically..
> But once a value is decided, I want a separate function (say,
new_add) that
> does
> new_add(b) (equals to) add(a,b)

Simple method, if you have source code for func:

>>> def adder(a):
...     def func(b, a=a): return b+a
...     return func
...
#In 2.1x with 'from future import nested scopes'
# and in 2.2, 'a=a' should not be necessary

>>> f = adder(2)
>>> f(5)
7

Terry J. Reedy






From a.clarke11 at ntlworld.com  Sat Apr  6 14:36:43 2002
From: a.clarke11 at ntlworld.com (tony clarke)
Date: Sat, 06 Apr 2002 20:36:43 +0100
Subject: beginners syntax question
References: <3caf47a5@news1.homechoice.co.uk>
Message-ID: <3CAF4E4B.ECE28871@ntlworld.com>

Hi

Try listing the objects  l=["object1','object2'..]
then
for i in l:
 i.update()

Another newbie, so treat with caution:-]
Tony

Tony wrote:

> Hi,
>
> I'm trying to call the update() method of several objects ie:
>
> object1.update()
> object2.update()
> etc
>
> I'd really like to do
>
> for x in range(7):
>   "object"+str(x).update
>
> but can't figure outthe syntax, is this possible?
>
> Thanks
> Tony



From marklists at mceahern.com  Fri Apr 26 10:44:45 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 26 Apr 2002 09:44:45 -0500
Subject: configuring mxCGIPython
In-Reply-To: <3CC90ABD.73350628@lemburg.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOELMKMAA.marklists@mceahern.com>

[M.-A. Lemburg]
> Hmm, interesting. Looks like sys.path is empty at this point.
> I wonder whether this is a bug in the frozen cgipython built
> or just a consequence of freezing the interpreter.
>
> Could you try setting PYTHONPATH to the dir where cgipython
> lives ?

I tried that.  The first error related to site.py no longer occurs, but I
still get the second error.  Output appended at the end of the message.  I
don't really know what line 27 of cgipython.py is?

> Since the stuff currently done in site.py doesn't really
> make all that much sense here, I think cgipython should
> probably use an empty site.py file instead of the standard
> one (should enhance startup times even more).

Sadly, I don't know enough to really judge this.

> Good question. It probably doesn't even start looking for
> it since it's already preregistered in the built-in
> modules.

I'm building my own fresh copy of cgipython to see if that makes any
difference.  I'll let you know.

Again, I appreciate your feedback and help.

Cheers,

// mark

<output>

cgipython exists
import site # frozen
import os # frozen
import posix # builtin
import posixpath # frozen
import stat # frozen
import UserDict # frozen
Python 2.2 (#1, Feb 19 2002, 17:44:05)
[GCC 2.95.2 20000220 (Debian GNU/Linux)]
Copyright (c) 2001 Python Software Foundation.
All Rights Reserved.

Copyright (c) 2000 BeOpen.com.
All Rights Reserved.

Copyright (c) 1995-2001 Corporation for National Research Initiatives.
All Rights Reserved.

Copyright (c) 1991-1995 Stichting Mathematisch Centrum, Amsterdam.
All Rights Reserved.
import __main__ # frozen
Traceback (most recent call last):
  File "../../CGIPython/cgipython.py", line 27, in ?
IOError: [Errno 2] No such file or directory
# clear __builtin__._
# clear sys.path
# clear sys.argv
# clear sys.ps1
# clear sys.ps2
# clear sys.exitfunc
# clear sys.exc_type
# clear sys.exc_value
# clear sys.exc_traceback
# clear sys.last_type
# clear sys.last_value
# clear sys.last_traceback
# restore sys.stdin
# restore sys.stdout
# restore sys.stderr
# cleanup __main__
# cleanup[1] signal
# cleanup[1] site
# cleanup[1] posix
# cleanup[1] exceptions
# cleanup[2] stat
# cleanup[2] posixpath
# cleanup[2] UserDict
# cleanup[2] os
# cleanup[2] os.path
# cleanup sys
# cleanup __builtin__
# cleanup ints: 3 unfreed ints in 1 out of 3 blocks
# cleanup floats





From neal at metaslash.com  Tue Apr  9 13:10:16 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 09 Apr 2002 13:10:16 -0400
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk> <13285ea2.0204090806.4a2d987b@posting.google.com>
Message-ID: <3CB32078.381C3AB1@metaslash.com>

"J.Jacob" wrote:

> Is there a way to peek into the latest python-dev messages? I could
> not find a link at http://www.python.org/psa/MailingLists.html
> Maybe if you subscribe to that mailing list (have to find their
> address first) you get too much email to handle so being able to just
> have a look at the lastest postings would be nice.

http://mail.python.org/pipermail/python-dev/

or

http://aspn.activestate.com/ASPN/Mail/Browse/Threaded/python-dev


From paul at boddie.net  Thu Apr 18 09:34:27 2002
From: paul at boddie.net (Paul Boddie)
Date: 18 Apr 2002 06:34:27 -0700
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019123256.19883.python-list@python.org>
Message-ID: <23891c90.0204180534.5b3c8542@posting.google.com>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> wrote in message news:<mailman.1019123256.19883.python-list at python.org>...
>
> What else does it look for?  If I symling /usr/local/bin/ld to gld will
> configure find it?

You might want to look at config.log to see what is going on "behind
the scenes". It might help even more if you were to pass the -v flag
to configure. In the log, I think it should indicate at which line the
configure script is testing for an ld executable, and if this is the
case, you should be able to find what configure is doing at that
point.

I just hope that the whole linking thing isn't some
dark-side-of-the-force libtool-determined configuration, otherwise you
may need to go in even deeper than I've dared to in the past. However,
it doesn't usually take much to persuade configure to look for things
in other places, set non-standard flags and so on, and if all else
fails, you can at least hack the configure script itself.

Paul

P.S. If only I'd bought that Solaris workstation off my defunct former
employer...


From martin at v.loewis.de  Wed Apr 24 03:23:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Apr 2002 09:23:13 +0200
Subject: Exceptions in Extensions: weird problem
References: <yv21yd5fvh1.fsf@lionsp093.lion-ag.de>
Message-ID: <m3vgahh866.fsf@mira.informatik.hu-berlin.de>

Harald Kirsch <kirschh at lionbioscience.com> writes:

> How can step (4) initialize/augment the error traceback with the
> traceback information generated in step (1)?

If you can guarantee that no other Python exception occurs between
step 1 and step 4, just don't perform PyErr_Clear() in step 2
(i.e. don't "catch" the exception); then use PyErr_Fetch in step 4 to
get the exception object and traceback. You might even just return
NULL, in which case the pending exception will continue to unwind the
Python stack.

HTH,
Martin


From frithiof.jensen at removethis.ted.ericsson.dk  Thu Apr  4 07:45:16 2002
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 4 Apr 2002 14:45:16 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CAB16BA.EEED07DE@sympatico.ca> <j48z84zp5z.fsf@informatik.hu-berlin.de> <3CAB60E7.8BEF40DC@sympatico.ca> <1017910237.53325@robert> <lzVq8.48605$pT1.1381394@news1.tin.it>
Message-ID: <1017924338.86233@robert>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:lzVq8.48605$pT1.1381394 at news1.tin.it...

> But I couldn't let this aspersion on f2c stand unchallenged...!-)

:-)

I use f2c a lot, although unconsciously: Every time a Fortran program -
usually som tricky numeric algorithm from Mathlib - is build on my Linux
box, f2c is silently invoked before gcc which performs the actual
compilation. One does not need to know really - so f2c does work well.

I just mentioned f2c as an illustration of what kind of code you will get -
it is fine for a procedural program but once you get different object &
class conventions (not to mention types) then things get really hairy.

When I worked at JET Joint Undertaking (UK) they converted millions of lines
of Fortran to C with surprisingly few problems - the code was *ugly* but it
allowed the migration from a dead harware platform from Norsk Data to SUN
Sparc.




From peter at engcorp.com  Mon Apr  1 20:03:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 01 Apr 2002 20:03:08 -0500
Subject: Most important extensions?
References: <mailman.1017557628.22081.python-list@python.org> <a877p1$sns$1@panix1.panix.com> <20020331112707.353$Hl@news.newsreader.com> <a87gso$eh9$1@panix1.panix.com> <a898r6$2e6r$1@news.idiom.com>
Message-ID: <3CA9034C.3D8FC7F5@engcorp.com>

Jim Dennis wrote:
> 
> In article <a87gso$eh9$1 at panix1.panix.com>, Aahz wrote:
> >In article <20020331112707.353$Hl at news.newsreader.com>,
> >Hernan M. Foffani <hfoffani at yahoo.com> wrote:
> >>"Aahz" <aahz at pythoncraft.com> escribi? en el mensaje
> 
> >>> Pmw does, but that's the only one.  Sorry for not making clear that I
> >>> was specifically referring to 3rd-party extensions.
> 
> >> I'd think win32all may apply then.  But you said OSCON, so.... ;-)
> 
> > I'm adding it to my list of proposed extensions (and it'll likely be one
> > of them, in the end).  OSCON is pro-Open Source, not anti-Windows.
> > (There are plenty of presentations on using Open Source effectively under
> > Windows.)
> 
>  Rpy, Medusa, PythonCard.

I agree about Medusa, but in 2.2 asyncore is part of the standard
distribution so it's probably not eligible given Aahz's criterion.


From frr at easyjob.net  Tue Apr 23 05:31:10 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 11:31:10 +0200
Subject: Trouble with Myqsl
Message-ID: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com>

I'm using CompatMysqldb from http://sourceforge.net/projects/mysql-python with
Python 2.2

Whenever I try to insert some data into a mysql db, I get this error:

>>> curs.execute("insert into customer_info (email, url, referrer, first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
Traceback (most recent call last):
  File "<pyshell#26>", line 1, in ?
    curs.execute("insert into customer_info (email, url, referrer,
first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com',
'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
  File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line 227, in
execute
    af =af+self.__res.affectedrows()
AttributeError: 'NoneType' object has no attribute 'affectedrows'
>>> 

The data does get inserted... What am I doing wrong? O:-)

TIA





-----------------------
Fernando Rodriguez


From noah at noah.org  Tue Apr 23 23:07:07 2002
From: noah at noah.org (Noah)
Date: Tue, 23 Apr 2002 20:07:07 -0700
Subject: stdin - seek() issue on Linux.
In-Reply-To: <284b44c2.0204231816.67c01ca8@posting.google.com>
Message-ID: <PAENINBOOODFDEPOKPKCAEFLCMAA.noah@noah.org>

It does work like a real file object -- it _IS_ a real file object.
The documentation could be more clear on that point, but
I don't think it's wrong (second paragraph):
	http://www.python.org/doc/current/lib/bltin-file-objects.html
It does say that some file operations may throw an IOException for
	"situations where the operation is not defined for some reason, 
	like seek() on a tty device or writing a file opened for reading."
File-like objects don't have to support all the File methods,
they just have to handle all the methods in some way.

On the other hand, perhaps the File object could be more consistent.
In some cases seek() will throw an IOError exception, but in
other cases it will do nothing (no-op). For example if you
call seek on disk files opened for writing in append mode (mode 'a') 
then seek will be do nothing (no-op)... Hmmm... 
I wonder why it throws an exception in some cases, but does nothing in others.

I prefer the IOError exception. It is actually a good way to handle 
the problem since it alerts you quickly to the semantics of the problem.
Otherwise it could have just failed mysteriously or tried to
fake seek() in some non-intuitive way. For example, they could
have made seek() simply do nothing (no-op), but that would be 
confusing and would require more special cases in the documentation.

Yours,
Noah


-----Original Message-----
>From: python-list-admin at python.org
>[mailto:python-list-admin at python.org]On Behalf Of Srihari Vijayaraghavan
>Sent: Tuesday, April 23, 2002 7:17 PM
>To: python-list at python.org
>Subject: Re: stdin - seek() issue on Linux.
>
>OK. IMHO sys.stdin doesn't work as stated in PLR (Python Library
>Reference), i.e., it doesn't work like a _real_ file object,
>particularly not supporting seek() related operations. While your
>technical explanation sounds OK to me, but I am still not convinced
>why PLR advertises that sys.stdin is treated as a file object. Please
>correct me if I am wrong here (as I am still a newbie user :-).
>
>I have fixed the problem by using cStringIO.StringIO(sys.stdin.read())
>method for now, which I think may not be the most optimised method.
>(as I guess it might read the entire file into the memory, OTOH using
>seek() method I could read only specific part of the file (sys.stdin)
>that I want)
>
>Thanks for your time.
>
>Hari,
>harisri at bigpond.com





From tim.one at comcast.net  Sat Apr  6 01:13:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 06 Apr 2002 01:13:11 -0500
Subject: Booleans, integer division,
 backwards compatibility; where is Python  going?
In-Reply-To: <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEHBOLAA.tim.one@comcast.net>

[Paul Rubin]
> ...
> C is not stagnant (C implementations keep improving) but it is mature
> and stable.  If I download an interesting new C program today, I can
> probably build and run it on my 5 year old Debian 1.x system without
> upgrading my C compiler.

That's because vendors have been incredibly sluggish in implementing the
current C standard.  Once they do, you'll have about as much chance of
compiling new C programs with your current compiler as you have of compiling
"ANSI C" programs under a K&R C compiler.  It could be users and vendors
will kill C99 through indifference; I'm not sure should be a fan of that
outcome, since you're one who seems to believe its new features in support
of portable floating-point numerics are important.  OTOH, if you use its new
features, you can forget about passing your code out to people without a C99
implementation.

"stable"-in-your-sense-is-akin-to-planned-obsolescence-ly y'rs  - tim





From peter at engcorp.com  Wed Apr 10 20:16:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:16:31 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at> <a92533$4bs$1@peabody.colorado.edu>
Message-ID: <3CB4D5DF.4912EFE4@engcorp.com>

Fernando P?rez wrote:
> 
> Siegfried Gonzi wrote:
> 
> > And I have really, really not the intention to waste my life as a
> > professional programmer
> 
> You might want to consider that in a programming group such as c.l.p, there's
> bound to be more than one person who will consider 'to waste my life as a
> professional programmer' to be a rather offensive way of asking for
> suggestions on any topic.

I chuckled to read this, mainly because as I read Siegfried's comment
I was not in any way offended, and I am a professional programmer.
It's not like we contribute much to world peace or solving world
hunger or anything.

On the other hand, I wonder what Siegfried does with his time which
is so much more respectable and less wasteful... <grin>

-Peter


From gerhard at bigfoot.de  Mon Apr  8 00:04:41 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 8 Apr 2002 06:04:41 +0200
Subject: Win32 .pyc icon suggestion
In-Reply-To: <439969A9369AE4439637588ADB9B434767861B@weblink73.weblink.com.au>
References: <439969A9369AE4439637588ADB9B434767861B@weblink73.weblink.com.au>
Message-ID: <20020408040441.GA31262@lilith.my-fqdn.de>

* Peter Aylett <petera at weblink.com.au> [2002-04-08 13:49 +1000]:
> Hi all,
> 
> Just a small idea:
> I find that there is very little distinction between the icon for the
> .py files and the .pyc files, which slows down the process of opening
> and managing my program files.
> 
> Here's a replacement icon I suggest for the .pyc files that makes life
> a bit easier, if anyone's interested.  (basically coloured yellow
> instead of green).

Would you mind uploading this as a patch to Sourceforge, so it doesn't
get forgotten? If you don't have an idea (I don't either), where in the
Python distribution this fits, just open up a new patch and only upload
the icon file instead of a full-blown patch.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From sholden at holdenweb.com  Mon Apr  1 17:27:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 17:27:56 -0500
Subject: getting infos from a website
References: <3CA61187.C0548FD5@cam.org> <3CA614B8.2C5C25E4@cam.org> <mailman.1017521988.813.python-list@python.org>
Message-ID: <Ka5q8.118260$Ou.95714@atlpnn01.usenetserver.com>

"Zutroi Zatatakowski" <abou at cam.org> wrote in message
news:mailman.1017521988.813.python-list at python.org...
>
>
> Zutroi Zatatakowski wrote:
> >
> > By the way, I can c.write() without any problem into the file though...
> > That's really puzzling me, I can write but cannot read. I know it's a
> > newbie problem but eh, so I am. :)
>
> Ok, it was pretty stupid, I was just not putting 'print' in front of it.
> But another thing... Now that I can capture a website html and output it
> into a file, I have to remove all html tags (I guess replacing '<>' by '
> ') or, but I don't know if it's possible, instead of capturing the HTML
> source of the page, could I retrieve only the text, like basic ASCII
> copy/paste?
>
You've got to the stage where you need to consider parsing the HTML. Take a
look at htgmllib abd sgmllib to see if you can work out how to do that. It
isn't as complicated as it seems once you've done it once ...

regards
 Steve






From kseehof at neuralintegrator.com  Wed Apr 10 09:49:30 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Wed, 10 Apr 2002 06:49:30 -0700
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool
 type)
In-Reply-To: <OyUs8.2398$b62.59604@news1.tin.it>
Message-ID: <EFEFIHAINIINNELLHFAIKEIBCAAA.kseehof@neuralintegrator.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Alex Martelli
> Sent: Wednesday, April 10, 2002 3:34 AM
> To: python-list at python.org
> Subject: Re: static, dynamic, strong and weak (was RE: PEP 285: Adding a
> bool type)
> 
> 
> Ken Seehof wrote:
>         ...
> > as a float, int, longint, or bool, depending on context.  All numbers
> > of equal value are pretty much interchangable (this becomes more true
> > as we add more numeric unification features).  In c/c++ numbers are
> > more strongly typed, in that there are many ways to get error messages
> > for using the wrong type of number without explicit conversion.
> 
> I disagree.  For example, in C I can use a float to index into an
> array and get it converted to an int implicitly and silently; in
> Python, using a float to index into a list is a TypeError.  What
> kind of compensating examples do you have in mind, where C would
> be stricter than Python "for using the wrong type of number"?

Not with my compiler...

  float x = 1.0;
  int a[4];
  a[x] = 4;

  error C2108: subscript is not of integral type

Hmm.  I thought you could do that in python, but I was wrong too...
Actually I was expecting the exact opposite of what you suggested.

Sequence indexing fails, but dictionary indexing works:

a = {}
a[2.0]=5
a[2]
5

But then this isn't a good example since C doesn't have dictionaries.
Hmm.  When I really think about it, I can't think of any.

You win.

> > The fact that an object's type doesn't actually change seems like an
> > implementation detail to me.  Numbers in python -feel- as if they are
> > weakly typed.
> 
> Not to me.  Even in Haskell I can write 1+2.0 - and Haskell is the
> strongest-typed and feeling-strongest-typed GOOD language I know
> (let's PLEASE forget that unnamed language where you must use + only
> to add integers and +. to add floats... have they changed it yet?-).
> 
> The first time I (accidentally) used a float as a list index and
> got a TypeError has impressed Python on me as strongly (albeit 
> dynamically)
> typed, better than any other study or experience so far.
> 
> 
> Alex

Well, after the grand numeric unification, (python 3.0 right?) you
will be able to index by float if the value is exactly equal to
an integer.  But C/C++ still will not permit it.

- Ken





From siegfried.gonzi at kfunigraz.ac.at  Tue Apr 23 04:04:25 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Tue, 23 Apr 2002 10:04:25 +0200
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> 	 <mailman.1019299357.11513.python-list@python.org> 	 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> 	 <Xns91F67C40EB0E0lumberjack@216.148.53.83> 	 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> 	 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> 	 <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at> <vcbk7qzezty.fsf@beta13.sm.luth.se>
Message-ID: <3CC51589.B9D9658B@kfunigraz.ac.at>

Torkel Franzen wrote:
> 
> Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:
> 
> > Wittgenstein dissmised the pure behavior of mathematicians: proof n, do
> > not rest and proof n+1, do not rest and proof n+2,...
> 
>   What comments of Wittgenstein's do you have in mind here?

No comments of Wittgenstein, but even a quick internet search (I could
look after it at home, but it is possible that I forget the thread)
leads to (sorry I am not a philosopher and do not easily have  any
access to a peer reviewed reference, but nonetheless):

http://www.tu-harburg.de/rzt/rzt/it/sofie/node52.html

"Wittgenstein hielt Mengenlehre und mathematische Logik f?r entbehrliche
Teile der Mathematik."

Which means: "Wittgenstein says that we don't need any set-theory and
logic".

Wittgenstein was more inclined to the so called "Intuitionismus" (e.g.
J. Brouwer), where he searched for a connection between math and common
sense.

I am sure the above does not really discuss whether Wittgenstein meant
do not concentrate on n,n+1,n+2. But for me Wittgenstein did mean with
his language introduction: meta-level explanation.

 
>   Wittgenstein and Godel never met.

I wrote an email to the official Austrian association for "Kurt Goedel"
in order to figure out whether Goedel met Wittgenstein or not.
Wittgenstein was in Vienna (and many other times) from 1924 to 1928 and
he even had contact with the Wiener Kreis (due to Schlick) but it
remains unclear for me whether he really met Goedel or not. Maybe an
official will answer my question.


S. Gonzi


From tim.one at comcast.net  Tue Apr 16 20:42:59 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 20:42:59 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <a9ids1$j54$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCIENHOOAA.tim.one@comcast.net>

> For anyone playing along, this particular way of provoking the problem
> appears stoppable via inserting step 2a:
>
> 2a: Delete this line from near the end of \Python22\INSTALL.LOG:
>
>    Delete in-use files: On

[Bengt Richter]
> That's good info, but you're not suggesting that as a solution, are you?

It's the only workaround I have, so sure.

> ISTM that would be interfering with a normal and legitimate function of
> installers/uninstallers.

It's no big deal.  It's *intended* to clean up in-use DLLs (like
python22.dll), but they're normally not in use at uninstall time, and
without that line in INSTALL.LOG Wise should visibly complain to the user
that it can't delete an in-use file if one is in fact in use.  I believe
it's a Wise bug that it both does manage to clean up UNWISE.EXE (with or
without that line), yet arranges to do a delete-on-reboot dance on
UNWISE.EXE anyway.

Note that we never had a report of a missing UNWISE.EXE before we moved from
Wise 5 to Wise 8.  One thing that changed along with that was including the
"Delete in-use files" line starting with Wise 8.

> ...
> At least that's my theory.
> Why do you think windows always wants you to reboot after
> installing, after all?

I know all that stuff, but it doesn't solve the problem.  Getting rid of the
"delete in-use files" gimmick does.  I don't know whether that's the *only*
problem here, though (Robin Becker seems to think not, but I don't believe
everything I read on Usenet <wink>).





From Ajay.Kapoor at cosinecom.com  Wed Apr  3 20:53:13 2002
From: Ajay.Kapoor at cosinecom.com (Ajay Kapoor)
Date: Wed, 3 Apr 2002 17:53:13 -0800 
Subject: Probelm in marshaling Any with typecode/value of struct
Message-ID: <69BCCDDC980B4641BFC908D7BF95F18401FC4267@exchsrv-eng>

 


 


Hi , 

      I am trying to pass a structure as CORBA.Any. I am constructing the
TypeCode using the stringified Fnorb id. 
      But at the runtime while sending the corba request I get the
marshaling error for CORBA.Any 
     Here is the code snippet for what I am doing 

    idl definition 
----------------------------------- 
     module agent { 

             interface GenericAgent 
                { 

                      struct NameValuePair 
                        { 
                         string name; 
                          any value; 
                    }; 
                struct Subnet 
                { 
                    string address;  /* 32 bit mask  */ 
                    short mask;      /* 0..32        */ 
                }; 
           } }; 

Now I am trying this 

  # got the corba object reference some how 
     agent =  reference 
     sn_tc =CORBA.TypeCode("IDL:cosinecom/agent/common/Subnet:1.0")  #
string is the Fnorb generated ID 
     srcSubnet =
agent.NameValuePair("sourceSubnet",CORBA.Any(sn_tc,Subnet("3.3.3.3",24))) 
...... 
     # making the request 
      agent.createObject(srcSubnet) 



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

I have also tried using 
tc=  CORBA.TypeCode(CORBA.id(Subnet) )  to construct the typecode and use it
with Any . But that also failed. 

I am getting the following error in all the cases. 

 File
"/vobs/orion/cosine/agent_cli/generated/fnidl/cosinecom/agent/__init__.py",
line 85, in createObject 
    apply(request.invoke, args, kw) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/DII.py", line
160, in invoke 
    (forwarded, self.__results) = client.synchronous(self, args) 
  File
"/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/GIOPClient.py", line
117, in synchronous 
    self.__marshal_parameters(cursor, request.inputs(), parameters) 
  File
"/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/GIOPClient.py", line
461, in __marshal_parameters 
    typecode._fnorb_marshal_value(cursor, parameters[i]) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 2121, in _fnorb_marshal_value 
    self._element_type._fnorb_marshal_value(cursor, element) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 1476, in _fnorb_marshal_value 
    member.type._fnorb_marshal_value(cursor, dict[py_member_name]) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 794, in _fnorb_marshal_value 
    value._fnorb_marshal(cursor) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/Any.py", line
77, in _fnorb_marshal 
    self.__typecode._fnorb_marshal(cursor) 
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 216, in _fnorb_marshal 
    cursor.marshal('L', self._kind.value()) 
AttributeError: 'string' object has no attribute 'value' 


Can somebody help me. 

Thanx 
Ajay Kapoor 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020403/1ae74274/attachment.html>

From duncan at NOSPAMrcp.co.uk  Mon Apr 29 09:40:18 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 29 Apr 2002 13:40:18 +0000 (UTC)
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org>
Message-ID: <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1>

holger krekel <pyth at devel.trillke.net> wrote in 
news:mailman.1020082610.7467.python-list at python.org:

> Btw, what is the best/shortest way to express 'filter' and 'reduce' with 
> list comprehensions?

 filter(fn, lst)
becomes:
 [ v for v in lst if fn(v) ]

 filter(None, lst)
becomes:
 [ v for v in lst if v ]

If the second argument to filter is a string or tuple then you have to 
convert the resulting list back to the appropriate type:

e.g.
 filter(fn, aString)
 str.join('', [ v for v in aString if fn(v) ])

reduce doesn't return a list, so a list comprehension is not an appropriate 
replacement.

 reduce(fn, seq, initial)
becomes:
 res = initial
 for v in seq: res = fn(res, v)

All of these replacements are big winners if they remove a use of the 'l' 
word. I don't think any of them lose in clarity even if there are no 'l's 
around.
 
-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From ljohnson at resgen.com  Tue Apr  2 09:27:24 2002
From: ljohnson at resgen.com (Lyle Johnson)
Date: Tue, 2 Apr 2002 08:27:24 -0600
Subject: Swig/Python: data type problem
References: <9d8d2e93.0204020314.362621a4@posting.google.com>
Message-ID: <uajfhvcudhk5cb@corp.supernews.com>

> But in C++, we expect a data type:
>
> void MyObject::read(String* namefile)
>
> I tryed with 3 data types: char / string / string*
> Each time, I have the same error:
>
> File "c:\python21\Module1.py", line 172, in read
>     val = apply(Interfc.RoadNetwork_lecture,args)
> TypeError: Type error. Expected _p_String
>
> Do you have any suggestion?

Well, the short answer is that you need to specify a typemap for String
pointers in your SWIG interface file. Since I'm not familiar with your
library's String class, let's just assume that the String class has a
constructor that takes a const char * for the string's contents. In that
case you could write an "in" typemap:

%typemap(in) String * {
    $1 = new String(PyString_AsString($input));
}

For more details, please read the SWIG user manual (especially the stuff
about typemaps).

Hope this helps,

Lyle




From klausman-un030402 at tuts.net  Wed Apr  3 15:03:13 2002
From: klausman-un030402 at tuts.net (Tobias Klausmann)
Date: 3 Apr 2002 20:03:13 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de> <3cab55f4.61190667@news.laplaza.org>
Message-ID: <a8fn61$s1nk7$1@ID-41653.news.dfncis.de>

Mats Wichmann <mats at laplaza.org> wrote:
> On 3 Apr 2002 08:08:18 GMT, Tobias Klausmann
><klausman-un030402 at tuts.net> wrote:
> 
>:Cliff Wells <logiplexsoftware at earthlink.net> wrote:
>:> On 01 Apr 2002 21:56:44 -0500
>:> nbecker at fred.net wrote:
>:> 
>:>> Keep lobbying Redhat.  The correct solution is for them to put
>:>> 
>:>> #! /usr/bin/python1.5
>:>> 
>:>> If they need a specific version.  I argued this with them already.
>:>> Keep up the pressure.  Maybe they will get it.
>:> 
>:> Absolutely agree.  Or better #!/usr/bin/python.rh with a link python.rh ->
>:> python1.5 so they can eventually upgrade without changing every shebang.
>:
>:In a related matter: how common is /usr/bin/env ?
>:I really like that tool to run python wherever it is: 
>:
>:#!/usr/bin/env python
>:
>:As it is part of sh-utils, I guess it's pretty widely used.
>:
>:Anybody know of specific environments where it is not commonly
>:available?
> 
> Windows :-)

Bah, Windows!

No honest, I don't want to spend the money I would have to if
I'd test my software with WinX. I try to stay as portable as
possible, but if it breaks, I have to rely on users' feedback.

> The thing to remember with env is that it uses the current environment
> to lookup the thing named python.  Which is great if you can precisely
> control this in all cases that matter.  The environment /is/ going to
> be different between your interactive shell and some other program
> started in a different context, such as "cron" (which could be asked
> to run a Python script for you), or a webserver (as in using Python
> for cgi scripts).   That's fine as long as you remember to check what
> "env" will find in non-shell contexts, and make any necessary
> adjustments.

So far, I've only written libraries/modules (no shebang) or console
apps that are at least semi-interactive. But if I'd write some such
utility that might run cronned, I'd surely stumble across the system
where cron doesn't know about /usr/local/bin/python15 =)

Greets,
Tobias


From graeme.matthew at unite.com.au  Wed Apr 17 09:14:05 2002
From: graeme.matthew at unite.com.au (Graeme Matthew)
Date: Wed, 17 Apr 2002 23:14:05 +1000
Subject: Classes Again !
Message-ID: <3cbd7527$1@news.comindico.com.au>

Hi all  I am going mad again, but making progress, been on a problem for 1
hour already and cannot see the cause ! please could someone help

The getSubDocument method in my template class is failing suddenly , i dont
have a clue why ! thanks a mil

ERROR:

Traceback (most recent call last):
  File "C:\www\cgi-bin\main.cgi", line 10, in ?
    data = t.getSubDocument('C:/www/templates/template_org_query.html')
AttributeError: Template instance has no attribute 'getSubDocument'

CGI SCRIPT BELOW

#! C:/python22/python.exe

import sys
sys.path.insert(0,'C:/www/classes/')
import Template

t = Template.Template('C:/www/templates/template_main.html')
t.loadTemplate()
data = "hello mom"
data = t.getSubDocument('C:/www/templates/template_org_query.html')
t.replaceDirective('workbench',data)

print "Content-type: text/html\n\n";
print t.document()

CLASS BELOW:

import re

class Template:

    def __init__(self, filename):

        self.filename = filename
        self.subdoc = none

    def loadTemplate(self):
        f = open(self.filename,'rb')
        self.data = f.read()
        f.close

    def replaceDirective(self, directive, data):
        prog = re.compile('<!--\s*' + directive + '\s*-->.*?<!--\s*' +
directive + '\s*-->',re.DOTALL | re.IGNORECASE)
        self.data = prog.sub(data,self.data)

    def document(self):
        return self.data

    def getSubDocument(self, fil):

        f = open(fil,'rb')
        self.subdoc = f.read()
        f.close

        prog = re.compile('<body>(.*)</body>',re.DOTALL,re.IGNORECASE)
        result = prog.match(self.subdoc)

        if result != none:
            self.subdoc = result.group(0)
        else:
            self.subdoc = '<parse error: no data, is it the correct file ?'

        return self.subdoc





From erno-news at erno.iki.fi  Thu Apr  4 10:14:37 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 04 Apr 2002 18:14:37 +0300
Subject: Py_BuildValue() and C++ arrays
References: <3cac101d$0$25808$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <kuofgzlcma.fsf@lasipalatsi.fi>

In article
<3cac101d$0$25808$6e365a64 at newsreader02.highway.telekom.at>, "Matthias
Stern" <mstern at aon.at> writes:

| I want to return the values "numElem" (int), "unit" (char[250]) and the
| array "data" (float,10000).
| What is the correct syntax for "Py_BuildValue( ??? )" ?

look in <URL: http://www.python.org/dev/doc/devel/ext/buildValue.html>.
note that you don't have to use Py_BuildValue, it's just
a helper for creating various python objects.

| The first two parameters work e.g. this way:
| ---
|      :
|      Py_BuildValue("(si)", unit, numElem);
|      :
| ---

| But how can i send back the data array?!? I cant't write it like this:
| ---
|      Py_BuildValue("(iiiiiiiii... (10.000 time)... iiiii)",  data);
| ---

that would create a tuple, like the Py_BuildValue docs say:

> "(items)" (tuple) [matching-items]
>    Convert a sequence of C values to a Python tuple with the same
>    number of items.

so you can use the PyTuple_* functions.

i don't think the array module exposes a c api, so for arrays you need
to use PyObject_CallObject on array.array and so on (i think).

  -- erno


From jeff at ccvcorp.com  Tue Apr  2 16:02:04 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 13:02:04 -0800
Subject: How to deal with relational data?
References: <3CA8D205.3C9E742B@computer.dot.org>
Message-ID: <MPG.1713bc279a57d2e498968c@news.nwlink.com>

In article <3CA8D205.3C9E742B at computer.dot.org>, 
Steven.dot.Suchting. at computer.dot.org says...

> How do we Python programmers typically deal with creating and using
> objects in which the source data is stored in a relational data base?
 
I am about as far from being database-savvy as a programmer can 
be, but... AFAICS, a database table should be more-or-less 
representable as a dictionary in Python, shouldn't it?  If you 
read each group of objects into a dictionary, keyed on the 
table's index value, then references to another object become the 
key by which that object is stored in its appropriate dictionary.  

Of course, if you have many deeply-nested cross-referencing 
tables, this could become unwieldy.  But it could help out.  If 
nothing else, you can cache some of your lower-level leaf objects 
in dictionaries.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From Patrick.Bussi at space.alcatel.fr  Mon Apr 15 11:00:25 2002
From: Patrick.Bussi at space.alcatel.fr (Patrick.Bussi at space.alcatel.fr)
Date: Mon, 15 Apr 2002 17:00:25 +0200
Subject: Newbie question: strange Re behaviour
Message-ID: <C1256B9C.0052A493.00@vzmta01.netfr.alcatel.fr>

Could someone help me understand what's wrong with this very simple regex, whose
initial purpose was to extract the value of an HTML field. For demonstration, I
have oversimplified the code below:

-----start-----
[pat]$ python
Python 2.0 (#1, Apr 11 2001, 19:18:08)
[GCC 2.96 20000731 (Linux-Mandrake 8.0 2.96-0.48mdk)] on linux-i386
Type "copyright", "credits" or "license" for more information.
>>>
>>> import re
>>> s='[^value=].*[^>]'
>>> T='value=baa>'
>>> re.compile(s).search(T).group()
'baa'
>>> T='value=aba>'
>>> re.compile(s).search(T).group()
'ba'
>>> T='value=aab>'
>>> re.compile(s).search(T).group()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'None' object has no attribute 'group'
>>> T='value=bba>'
>>> re.compile(s).search(T).group()
'bba'
-----end -----

Therefore, it seems that re.compile is sensitive to a leading 'a' character,
which is hard to believe.

Any help would be appreciated.

Regards.

---
Patrick Bussi
patrick.bussi at space.alcatel.fr


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






From aleax at aleax.it  Wed Apr 17 16:49:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 20:49:24 GMT
Subject: preg_replace?
References: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <odlv8.24535$b62.683481@news1.tin.it>

Philipp Lenssen wrote:

> I am porting a program that I already wrote in VBScript and PHP.
> My current task: I need to replace everything inbetween square brackets
> with "0". In VBScript I did this "by hand" (looping, inString and so on).
> In PHP somebody helped me with the following solution:
> 
>     $s = preg_replace("%\[[^\]]*\]%", "0", $s);
> 
> Even though I've been reading through regex-tutorials in the meantime, I'm
> still not familiar with it. But is there a simple rule how to convert the
> above into Python code? Not only would it solve my current problem but I
> already got some helpful regex's for PHP and could reuse them in the

Python 2.2.1 (#1, Apr 15 2002, 17:55:14)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> ret=r'\[[^\]]*\]'
>>> import re
>>> cre=re.compile(ret)
>>> a='ba[pop]ple[pup]'
>>> cre.sub('0',a)
'ba0ple0'
>>>

I'm not sure what's wrong with your code, but this works for me.


Alex



From tim.one at comcast.net  Mon Apr 22 12:44:29 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 22 Apr 2002 12:44:29 -0400
Subject: A simply newbie question about ndiff
In-Reply-To: <3cc3a6cd$0$15475$afc38c87@news.optusnet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEBCPAAA.tim.one@comcast.net>

[Neville Franks]
> Actually my entire reason for getting into Python right now was
> only to look at your ndiff code to see the sort of results it
> delivered and also get a feel for its space/time usage.

It reads both files into RAM, and builds a dict mapping each non-junk line
in "the second file" to a list of all indices in the second file at which
the line appears.  So space is proportional to the file sizes.  Time
complexity is extremely data dependent.

> I only discovered your ndiff over the weekend, which is a little
> surprising as I've also been working on Diff's for quite a few
> years now as part of my programmer's editor, ED for Windows.

Cool.  ndiff will do a fine job for you, but if you're slinging giant files
you'll need to compromise.

> I've been working on a major rewrite of ED4W over the past 3+ years
> and I did look at replacing my diff code with something else which
> would hopefully be a bit better (results/time/memory),

ndiff is after results first, memory second, time last.

> but didn't find anything that fitted the bill. I did have a look at
> fcomp which produced good results quickly for small files, but managed
> to use all available memory and time for larger files. Hence the
> interest in ndiff.
> ...
> I wonder how it would perform in C++. From what you've said it
> doesn't sound like their would be much of performance gain.

You're welcome to try it <wink>.

> FYI on the 73,000 line file I diffed there was only one line different
> in the two files. ED's diff took approx. 16 seconds and ndiff I think
> around 2.5 minutes.

There's one trick you can probably afford that ndiff won't do:  many diff
algorithms start by stripping the longest common prefix and suffix, applying
the workhorse algorithm only to what remains after that.  ndiff does not,
for reasons explained somewhere in a comment.  Instead it gets the effect of
this (specifically, SequenceMatcher does):

    for every element X in sequence 1:
        for every (non-junk) element Y in sequence 2:
           find the longest stretch of equal elements starting at X and Y
           if that's the longest stretch so far, remember it

Now SequenceMatcher uses a much cleverer implementation than that
brute-force triply-nested loop (which, as written, has worst-case cubic
time), but it can still take substantial time, and 70K lines is ridiculously
out of the range of sizes ndiff was intended to handle.  70K is in the range
of a very long novel; the ndiff algorithm was first used to create
human-readable patches for source code files, and later for plain-ASCII
design documents.  Those typically take a few seconds.

For giant files, stripping common prefix and suffix first shouldn't hurt
quality much, and would speed things enormously.  For example,

# Return (i, m), (i, n) s.t.
#     x[:i] == y[:i]
# and
#     x[m:] == y[n:]
# IOW, find the indices of the common prefix and suffix.
# x[i:m] and y[i:n] are the mismatching parts.

def chop(x, y):
    m, n = len(x), len(y)
    smaller = min(m, n)
    i = 0
    while i < smaller and x[i] == y[i]:
        i += 1
    while m > i and n > i and x[m-1] == y[n-1]:
        m -= 1
        n -= 1
    return (i, m), (i, n)

If you fed your 70K-line files thru that first, it should identify the
1-line mismatching regions in a fraction of a second (even leaving the code
in Python).  Etc <wink>.





From grante at visi.com  Mon Apr 15 11:41:44 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 15 Apr 2002 15:41:44 GMT
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <3CBABA71.3040909@geneva-link.ch>
Message-ID: <YwCu8.32367$vm6.5103753@ruti.visi.com>

In article <3CBABA71.3040909 at geneva-link.ch>, Boris Borcic wrote:

> For the shortcut semantics, we'd like to have :
>      a xor b
> 
> behave exactly like :
>      (not b and a) or (not a and b)

I don't understand how that's a shortcut (assuming that means
the same thing as short-circuit).  You have to evaluate both
operands of an xor operator.  There is no possibly way to
short-circuit the expression so that only one is evaluated.

-- 
Grant Edwards                   grante             Yow!  I'm protected by
                                  at               a ROLL-ON I rented from
                               visi.com            AVIS...


From starx at pacbell.net  Sat Apr 13 05:50:47 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 13 Apr 2002 09:50:47 GMT
Subject: New classes, __slots__, __dict__
References: <mailman.1018639826.18289.python-list@python.org> <just-2F729A.10381713042002@news1.xs4all.nl>
Message-ID: <Xns91EF1D24C67C3RASXnewsDFE1@130.133.1.4>

Just van Rossum || Sat 13 Apr 2002 01:38:17a:

> In article <mailman.1018639826.18289.python-list at python.org>,
>  Tim Peters <tim.one at comcast.net> wrote:
> 
>> Another link:
>> 
>>     http://www.python.org/sf/520644
>> 
>> That's bug report "[520644] __slots__ are not pickled".  The short
>> course is that someclass.__slots__ is unreliable because mutable, and
>> so how to find all the actual slots in all the classes (direct and
>> base) by magic is something of a puzzle and/or PITA.  Unless/until
>> that does get done by magic, you need to define a __getstate__ to
>> spell out which slots do and don't want to be pickled.
> 
> Is making __slots__ immutable no solution? Seems almost too easy...

Er... I thought the fact that it WAS mutible was an error somehow...

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From martin at v.loewis.de  Sat Apr 13 18:39:34 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 00:39:34 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com>
Message-ID: <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> I think an 'xor' operator is needed.  The 'xor' operator means
> 'logical-exclusive-or', and returns a 1 or 0.

In Python 2.3, you will be able to write

   bool(a) ^ bool(b)

Seems pretty clear to me.

Regards,
Martin


From jb at cascade-sys.com  Mon Apr 29 14:51:53 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 11:51:53 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>
Message-ID: <3CCD9649.98AED86F@cascade-sys.com>

Brian Quinlan wrote:

> > Huh? The "core" language wouldn't be measurably shrunk by this
> proposal
> > fella. Your rationale is built on air. Show us some genuine
> _advantage_
> > that is worth the huge _cost_ in time and effort caused by your ill-
> > considered proposal. How about you and other PEP writers take some
> > responsibilities for your fantasies for once? You up to helping
> thousands
> > rewrite their code? I thought not.
>
> Why the hostility?

I don't interpret this strongly worded and slightly sarcastic critique to
be hostile.

If you call this hostile I don't think you know the meaning of the word.

> I think that every PEP is a good one. Even if the PEP is ultimately
> rejected, it promotes thought about some aspect of the language. Taking
> responsibility sounds like stifling discussion to me.

And I don't see how his offering a strong opinion stiffles discussion.

I happen to agree with just about everything he said.

This PEP is angles in the clouds stuff.  A lot of work for everybody and
virtually zero practical benefit to anybody.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From mgilfix at eecs.tufts.edu  Fri Apr 19 11:27:39 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 19 Apr 2002 11:27:39 -0400
Subject: why no 'length' method in sequences ?
In-Reply-To: <23891c90.0204190608.11c70ee4@posting.google.com>; from paul@boddie.net on Fri, Apr 19, 2002 at 07:08:35AM -0700
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <E4Iv8.2972$8D3.78501@news1.tin.it> <just-2DF829.09350419042002@news1.xs4all.nl> <23891c90.0204190608.11c70ee4@posting.google.com>
Message-ID: <20020419112738.B28086@eecs.tufts.edu>

On Fri, Apr 19 @ 07:08, Paul Boddie wrote:
> Just van Rossum <just at xs4all.nl> wrote in message news:<just-2DF829.09350419042002 at news1.xs4all.nl>...
> > 
> >     len(o)
> 
> [vs]
> 
> >     o.keys()
> > 
> > This seems quite arbitrary, even to me, after maybe 6 years of working 
> > with Python. If o.__len__() were spelled o.len(), we could easily do 
> > without the len() builtin. (Note that I don't argue we could do without 
> > functions: I wouldn't want to miss them for the world...)
> 
> I suppose that len is considered near-universal, although len(123)
> reveals a notable exception, whereas very few objects bear the notion
> of having "keys". Therefore, near-universal "protocols" (as I believe
> someone once called them) are supported through built-in functions,
> and other "protocols" are supported more or less exclusively through
> methods.
> 
> Of course, providing len as a method on all appropriate objects would
> be more consistent, but as far as Python is concerned right now, it
> seems redundant and potentially confusing: "But I thought len was a
> function!" exclaimed the beginner.

  I think that comes closest to touching on the reason. Of course this
is all philosophical but here's my shot :) Length is something that is
intrisic to object. If we visualize objects as being something in the
real world, then they would have some sort of dimension. __len__ is
an extension of that dimension in some sense and it makes sense that
I have to use an action to obtain that length.  In the case of keys,
keys are a special attribute of dictionaries only and I can think of
the action of retrieving keys as opening up the dict and pulling them
out, a member function.

  Just as you say, these near universal protocols are just ways of
describing how objects behave in the word under certain actions but if
you implements an __add__ function, you still need to invoke the '+'
action to elicit the behavior.

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From dominikush at yahoo.com  Wed Apr 17 10:31:46 2002
From: dominikush at yahoo.com (dominikush at yahoo.com)
Date: Wed, 17 Apr 2002 16:31:46 +0200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <3CBD7406.3050909@mxm.dk> <3CBD80B2.91A7D596@yahoo.com> <vyfv8.30552$vF6.951821@news2.tin.it>
Message-ID: <3CBD8752.11FE745F@yahoo.com>

> > I understand that! But my main objection remains: Why should
> > I care about the type of object being handled by a dictionary?
> 
> You shouldn't: whatever types of objects are the dictionary's values,
> method .values() will ALWAYS return a new list object whose items
> refer to EXACTLY the same objects as the dictionary's values.

I got it after I read your and Michaels posts!
 
> > This could not only break unit tests but also make maintenance
> > of ``old'' code somewhat painful.
> 
> I don't understand what you're talking about.  My sentence above
> was true at least since Python 1.5.2 (probably earlier, but that's
> when I hopped aboard).
> 
> Alex

Now, I don't understand it myself anymore ;-)

Dominikus


From kfrederick at bigfoot.com  Tue Apr 23 22:23:36 2002
From: kfrederick at bigfoot.com (Kevin)
Date: 23 Apr 2002 19:23:36 -0700
Subject: Killing all child processes?
Message-ID: <59aa3aef.0204231823.77932129@posting.google.com>

Hi all,

I can't seem to get my python program to kill all the child prcoesses
that it may fork. Is this possible?

I've researched usenet+the web and read subproc.py from
ftp.python.org, but still no go.

My latest attempt was a python script that forks itself, uses setpgid,
then the child does an execl(). The execl'd program creates some
children. Meanwhile, the parent decides time is up on the child and
tries to kill everything, but  my os.kill(pid,9) only succeeds in
killing the child aka the execvp'd process.

There's another bug that I haven't yet addressed about the parent
waiting around when the child doesn't write to either of the pipes.
But that'd not my focus now, so don't worry about that; I just want
all the children to die.

My full code it at the bottom. Hints, anyone? This is on Linux and
FreeBSD.

Thanks,
Kevin

---Full code follows---
---- pcontrol.py below
#!/usr/bin/python
#
# pcontrol.py
#

import os
import sys
import signal

_alarm = "request timed out"

time = 5

cmd="sh1.sh"

###########################################################

def alarm_handler(a , b):
    global _alarm
    signal.alarm(0)
    raise _alarm

###########################################################

# Okay, let's do this the hard way: create two unnamed
# pipes for stdout/err. Then, fork this process, and the
# child becomes the new command and shoves their output
# onto the pipes, while the parent waits for the child to die.

# Create two named pipes
child_out_r, child_out_w = os.pipe()
child_err_r, child_err_w = os.pipe()

# Let's fork a child that gets replaced with the test process - krf
child_pid = os.fork()

if (child_pid != 0):
    # We're the parent

    os.close(child_out_w) # the parent won't read the pipes
    os.close(child_err_w)

    os.setpgid(child_pid, 0)

    # Now do the timing...
    signal.signal(signal.SIGALRM, alarm_handler)
    signal.alarm(time)

    try:
        child_pid, child_status = os.waitpid(child_pid, 0); # wait for
child
    except OSError:
        print("OSError... did the executable time out?")
        child_status = -1 # this is the same as a timeout for now
    except _alarm:
        print("Timed out")
        # kill it!
        os.kill(child_pid, 9)
        child_status = -1 # timeout

    signal.alarm(0) # clean up alarm

else:
    # We're the child
    os.setpgid(0, 0) # now the child is it's group leader (?)

    # Redirect error to our pipes
    os.close(child_out_r) # the child won't read the pipes
    os.close(child_err_r)
    os.dup2(child_out_w, sys.stdout.fileno())
    os.dup2(child_err_w, sys.stderr.fileno())

    # replace the process
    os.execl(cmd, cmd) #strange, but execl->execv needs a second
arg...
    # the child is no more... (assuming success)
    
print "Status = ", child_status
print "Child stdout [",os.read(child_out_r,999),"]"
print "Child stderr [",os.read(child_err_r,999),"]"


------- sh1.sh below
#!/bin/sh

echo "This is processs (1) $$."
sh2.sh &
sh3.sh &
sleep 600
echo "Process (1) $$ exiting."

------- sh2.sh below
#!/bin/sh

echo "This is processs (2) $$."
sleep 600
echo "Process (2) $$ exiting."

------- sh3.sh below
#!/bin/sh

echo "This is processs (3) $$."
sleep 600
echo "Process (3) $$ exiting."


From pj at engr.sgi.com  Mon Apr  8 20:04:22 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 9 Apr 2002 00:04:22 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84> <a8lvod$589$1@panix1.panix.com> <Xns91E87D184B4BCJamesLLugojcom@216.148.53.83>
Message-ID: <a8tb66$pm65o$1@fido.engr.sgi.com>

|> If that rate is maintained, then I suspect that in a few
|> years I will have a lot more company.

Likely so.  Because it is likely that a certain modest percentage
of any group of users will agree with you.  As Python becomes
more popular, the overall group gets larger, hence the percentage
in agreement with you gets larger.

But somehow I don't think that was the point you were trying
to make ...
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From aleax at aleax.it  Sun Apr 14 02:59:04 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 14 Apr 2002 06:59:04 GMT
Subject: 2.1 vs. 2.2
References: <mailman.1018751366.25497.python-list@python.org>
Message-ID: <YM9u8.14076$b62.391028@news1.tin.it>

Tim Peters wrote:
        ...
> (although, in the case of languages, they're likely to find that the
> eventual standard is carefully worded to allow all variations that already
> exist).

Not what I saw happen in the small set of languages whose standardization
I had to follow.  With C++, in particular, the standard was so far from
"allowing all variations that already existed" that, AFAIK, still today, 
years later, no commercial implementation claims 100% standard compliance
(something that I still find rather amazing).  Of course, the usefulness
of that is as dubious as that of a standard which did allow everything.


Alex



From duncan at NOSPAMrcp.co.uk  Tue Apr 30 09:40:05 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 13:40:05 +0000 (UTC)
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <199334c5.0204260645.3af87dc5@posting.google.com> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <Xns9200935841739duncanrcpcouk@127.0.0.1>

Ralf Juengling <juenglin at informatik.uni-freiburg.de> wrote in 
news:xduu1pt1iuf.fsf at leto.informatik.uni-freiburg.de:

> Here's a proposal, that is 'more elegant' in my eyes. If the builtin 
> 'list' type was an *iterable* factory function and we'd borrow the 
> function 'take' from a more functional language, say
> 
> def take(n, i):
>     if not n:
>         return []
>     else:
>         i = iter(i)
>         return [i.next()] + take(n-1, i)
> 
> then 
> 
> l = take(n, list)
> 
> would do it.
> 
> BTW: It is not easy to add __iter__ to list's meta-type, is it?

How about writing a generator to give you a list of any type:?

def listof(sometype=list):
    while 1: yield sometype()

def take(n, i):
    i = iter(i)
    return [ i.next() for j in range(n) ]

take(6, listof(list))
take(99, listof(dict))

It occurs to me that it ought to be possible to produce a type 
(pseudolist/lazylist?) that can be constructed from a sequence and has all 
the methods of a list, but which only accesses the argument sequence when 
it absolutely has to. e.g.:

    lst1 = lazylist(listof())[:6]
    lst2 = lazylist(file(name))[10:20]

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From cliechti at gmx.net  Tue Apr  2 17:06:51 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Apr 2002 23:06:51 +0100
Subject: style Q: global vars in CAPS?
References: <slrnaak4m7.u14.ak@ak.silmarill.org>
Message-ID: <Xns91E4EB4DCD862cliechtigmxnet@62.2.16.82>

Andrei Kulakov <ak at silmarill.org> wrote in 
news:slrnaak4m7.u14.ak at ak.silmarill.org:
> What do you think about global variables spelled in caps? Is that what
> most people do?

i think most people write constants in all caps no matter if global if 
local vars. at least i do :-)

maybe your interested in Guido's thoughts on style in general:
http://www.python.org/doc/essays/styleguide.html
http://python.sourceforge.net/peps/pep-0008.html

-- 
Chris <cliechti at gmx.net>



From Chris.Barker at noaa.gov  Fri Apr 19 12:01:43 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 19 Apr 2002 09:01:43 -0700
Subject: Slightly OT Linux newbie question
References: <3cbe8b02_1@news2.newsgroups.com>
Message-ID: <3CC04D25.564AC728@noaa.gov>

jb wrote:
> When I worked under MS-Win everything was very simple and comfortable:
Personally,. I find windows very uncomfortable: it's all what you are
used to. Anyway, the only difference is that the way you installed
Python on windows apparently added stuff to the Start menu. I don't
think anyone has botherred to make a Python installer that adds stuff to
the KDE mwnues foir you , though it certainly could be done.

>There
> was a menu

Read the KDE docs, you can add these menu items yourself.

> where I could select the IDLE or the documentation or the Python
> prompt, whatever I wanted.

The docs, if you downloaded the HTML version, are somewhere on your
system. Find them (try using "locate" at teh command line, it's very
handy, just remember that it doesn't know about newly installed files)
you can then just put a bookmark in for the docs in Konqueror, as it is
a browser and file manager in one.

To start IDLE, hit alt+F2, and type "idle", or use the K-menu item: run
command, or start a terminal and tyupe idle in that. For the python
prompt, type python in a terminal. You can create a KDE "link to
application" that opens a terminal and runs Python for you. right click
somewhere and under new yo'll find what you need.

While you might wonder why someone didn't just build an installer that
put all this in the menue for you, I would say that that is not
desireable. The are mjany hundreds of command line programs on Linux I
wouold NOT want every one of them in the menus. Many folkd install
python to use for CGI scripting, or to run other peoples scripts (RedHat
sysadmin stuff, among others), and have no desire to access teh prompt
themselves. Learn to set up the apps you want in KDE, and you'll have a
nice compact, custom system.

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From jacek.generowicz at cern.ch  Sat Apr 27 16:26:00 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 27 Apr 2002 22:26:00 +0200
Subject: Python's Lisp heritage
References: <mailman.1019922475.8361.python-list@python.org> <tyfsn5h13mh.fsf@pcitapi22.cern.ch> <ZADy8.70845$vF6.2115921@news2.tin.it>
Message-ID: <tyf1yd028iv.fsf@pcitapi22.cern.ch>

Alex Martelli <aleax at aleax.it> writes:

> Jacek Generowicz wrote:

> >   that you would really like to extend the set of infix operators,
> >   with say @, to allow you to call splark without having to type those
> >   infernal parentheses. Is it going to be easier in Python or in Lisp?
> 
> Haskell -- it has definition of new infix operators as a built-in.

Yes, it's cute. I've not done any serious programming in Haskell
though, so I have not had the opportunity to form an opinion on
whether this is a desirable feature or not. I suspect it might be more
trouble that it is worth.

> But THAT, IMHO, is a design error: languages with mutable syntax are
> spending far too much complexity on syntax sugar -- and not to good
> overall effect either.  The language's "new improved" syntax may
> LOOK spiffier when you first see a new op introduced in some article
> or tutorial about the language -- but when the rubber hits the road,
> some poor devil of a programmer is going to have a very hard time
> decrypting expressions using the new op -- where does he or she go
> look up precedence and associativity, and how much work does it take
> to parse them mentally...?

... exactly.


From hst at empolis.co.uk  Wed Apr 24 11:54:44 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Wed, 24 Apr 2002 16:54:44 +0100
Subject: Probably I'm just too stupid, but...
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220B82@hendrix.empolisuk.com>

redflag wrote:
> Subject: Probably I'm just too stupid, but...
> 
> 
> >>> def c():
> 	for i in range(10):
> 		print i
> 
> 		
> >>> c
> <function c at 0x00AA93B0>
> >>> for i in range(10):
> 	print i
> 
> 	
> 0
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
> >>> 
> 
> why didn't "c" work as I imagined?
> -- 
because it should have been
c()

to invoke the function.

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From jgoerzen at complete.org  Mon Apr 22 09:34:54 2002
From: jgoerzen at complete.org (John Goerzen)
Date: 22 Apr 2002 08:34:54 -0500
Subject: DBAPI
Message-ID: <87adrvambl.fsf@complete.org>

Hi,

I have some DB-API questions.

I have used things like JDBC for Java and DBI for Perl.  To me, DBAPI
for Python looks rather week.  For instance, if the code is to be
truly portable across databases, I'd have to check the paramstyle for
the driver and write each query 5 times!

Is there any "meta-DBAPI" module that I can use that would present me
with a standardized interface, and then (for instance) convert my
queries to whatever format the backend wants?

-- John


From sholden at holdenweb.com  Mon Apr  8 08:13:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 08:13:38 -0400
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <200204061913.g36JDcdj017764@ratthing-b246.strakt.com> <mailman.1018122494.10925.python-list@python.org> <roy-9E63B6.15273306042002@news1.panix.com>
Message-ID: <xRfs8.26529$Sw5.26056@atlpnn01.usenetserver.com>

"Roy Smith" <roy at panix.com> wrote in message
news:roy-9E63B6.15273306042002 at news1.panix.com...
> Paul Sidorsky <paulsid at shaw.ca> wrote:
> > The proper course of action here would be to add a second check,
> > isPaid(), to wherever the new information is needed.
>
> Or, instead of isValid() and isPaid(), have a single method, getStatus(),
> which could return invalidAccount, paymentOverdue, or accountUpToDate.
Not
> all things fit into a boolean, but those that don't shouldn't be accessed
> by methods whose names begin with "is".  I guess the die-hard lispers
would
> say the same about names ending in "p" :-)
>
> The most fascinating (and often, frustrating) example I know of "almost
> boolean" is SQL's concept of NULL.  In what other language can "A == B"
and
> "A != B" both be false?  The problem is that they tried to cram two
> distinct concepts ("are the operands valid?" and "are the operands
equal?")
> into a single conceptual slot.

I don't think that's true. I think "they" required a value which
specifically indicated that data were unavailable, or irrelevant in certain
semantic contexts (such as the hourly rate of a salaried employee). The
result of the comparisons you quote is *neither* true *nor* false, it's
NULL.

Given the existence of such a value, the decision about what to return as a
result of operations on it (the same such value) appears reasonable and
well-founded to me.

The problem is not with SQL,but with our habit of thinking in two-valued
logic terms. Which is why, when I'm teaching SQL, I tell students that NULL
is an indispensable part of the language that they should avoid until they
understand why :-)

regards
 Steve






From alexandre.herbiniere at caramail.com  Tue Apr 16 09:18:53 2002
From: alexandre.herbiniere at caramail.com (alexandre)
Date: 16 Apr 2002 06:18:53 -0700
Subject: need help to choose an appropriate database
Message-ID: <fac97694.0204160518.6f6d79b7@posting.google.com>

I work in a company who wants to register informations concerning
customers in a datebase. they suggest me to use CGI and Python to
capt data. I must send these data in a database and i would like to
know which ODBC is the most appropriate to that situation ?

thank you


From sholden at holdenweb.com  Fri Apr 19 07:57:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 19 Apr 2002 07:57:06 -0400
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <I6Fv8.36158$vF6.1093396@news2.tin.it> <mailman.1019162916.19664.python-list@python.org> <7cIv8.36928$vF6.1122050@news2.tin.it> <mailman.1019184878.1733.python-list@python.org>  <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com> <mailman.1019212237.13240.python-list@python.org>
Message-ID: <NETv8.45693$eS3.10263@atlpnn01.usenetserver.com>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1019212237.13240.python-list at python.org...
> > Thanks Alex and thanks to all who discussed my problem. I am honoured to
> > participate in such a group.
> >
> > The 'Alex solution' worked but it does lead on to another problem -
> > Reference material!
> >
> > I bought 'Learning Python' by Mark Lutz, First Edition. Read it when I
> > bought it a couple of years back, but I usually get by with the online
> > documentation that comes with python.
> >
> > With all the changes in Python since then I need a new reference work.
Would
> > anyone recommend a  book that includes info on metaclasses?
> >
Unfortunately the [book] publishing world has a problem tracking
developments that are so recent. For metaclasses and the like, added to the
language in 2.2 as part of the effort towards type/classunification, you
will have to wait a while - probably at least another three to six months
before anything as recent as this is covered.

I know from my own experiences writing "Python Web Programming" that no
matter how hard you try it's very difficult to be completely up-to-date. I
*did*, thanks to Andrew Kuchling, manage to include an appendix covering 2.1
and 2.2, but you can get the complete version of his material on the web at

    http://www.amk.ca/python/2.2

It also includes some useful references to other web sources, the
authoritative source for metaclasses (if there is one) being

    http://www.python.org/peps/pep-0253.html

>
> If you are willing to wait, Alex's own Python in a Nutshell will be
> published by O'Reilly soon.  In this particular case, I think the
> Nut is more of a coconut than a pistachio nut :-).  Its not PP2E sized,
and
> no doubt Alex will complain about all the wonderful things he had
> to _cut_, but nonetheless ...
>
I'm sure Alex's book is going to be an excellent contribution to the
language and its growth. While I was disappointed that it had to line up
behind "The Python Cookbook", it will probably be better for the wait. (I
can imagine Alex gnashing his teeth whenever new features requiring changes
are introduced, though. When you are writing, sometimes you just want to
shout "Stop!" at the developers ;-)

regards
 Steve
--
home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From gyromagnetic at excite.com  Tue Apr 16 11:49:06 2002
From: gyromagnetic at excite.com (gyromagnetic)
Date: 16 Apr 2002 08:49:06 -0700
Subject: Where should 'import' statement be put?
Message-ID: <4620daca.0204160749.1f68a62a@posting.google.com>

Hi,
I was wondering what the 'best practice' is for locating 'import'
statements. If I have a module comprising several classes, each of
which has several functions, should the import statements be grouped
- at the top of the module?
- at the top of the classes in which they are relevant? 
- at the top of the functions in which they are relevant?

I believe imports are done only once regardless of the number of times
the statement appears, but are there any other considerations here? I
realize there may be scoping issues as well.

Thanks for the advice.

-gyro


From cliechti at gmx.net  Mon Apr  1 15:49:15 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Apr 2002 21:49:15 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de> <KM_p8.41075$pT1.1129125@news1.tin.it> <a8a0gt$roc$1@panix1.panix.com> <l_1q8.49099$S52.1668112@news2.tin.it>
Message-ID: <Xns91E3DE242622Fcliechtigmxnet@62.2.16.82>

Alex Martelli <aleax at aleax.it> wrote in
news:l_1q8.49099$S52.1668112 at news2.tin.it: 

> Aahz wrote:
> 
>> In article <KM_p8.41075$pT1.1129125 at news1.tin.it>,
>> Alex Martelli  <aleax at aleax.it> wrote:
>>>
>>>0 is (2.2 is still present, not yet past) the canonical way to spell
>>>untruth: it's what 'not', 'is', 'in', comparisons, and all standard
>>>library iswhatever's return -- including operator.truth, of course.
>> 
>> Just to throw in a datapoint, I generally use None to indicate
>> falsehood. 
> 
> ...and you can presumably continue to indulge in this quirk, if you so
> choose, after 285 passes.  'return 0' is obviously faster (a
> LOAD_CONST opcode, rather than a LOAD_GLOBAL for either 'return False'
> or explicit 'return None', where the latter always has to check
> globals() first and then proceed to the built-ins), but just as
> obviously the difference is so minute as not to matter -- a matter of
> 10% slowdown on my machine for 100,000 calls to a function whose body
> is just return None, versus one whose body is just return 0 (0.45 vs
> 0.41 secs for 100,000 calls including timing overhead, python -O,
> time.clock).  An _implicit_ return None (fall-off-the-end) would be
> just as fast as return 0, but I would consider it seriously obfuscated
> (while 'return None' strikes me as only moderately so -- just because
> of the difference wrt what all of the builtin and library
> equivalents). 

the above argument about speed seems to be stronger one against the change. 
it would mean that the python core must be adapted so that booleans are 
fast enough. there are enough (sometimes needless but still) questions 
about speed, don't put oil in the fire.

i agree that writing True instead of 1 clarifies code and it could be an 
advantage for RPC etc. but not in cost of speed and a hacky version:
- if str(True) prints '1' its still confusing for newbees at the prompt.
- the problems with 'var == True' and 'var == False' etc.
- backward problem with pickle and other serializing libs
- speed

i'm reading in the thread pros and cons, regarding the problems for such a 
small benefit of having True and False i vote for inluding it later in 
Python 3000, if ever.

chris

-- 
Chris <cliechti at gmx.net>



From Alex.Adrian at gmx.de  Wed Apr 24 06:50:27 2002
From: Alex.Adrian at gmx.de (Alex Adrian)
Date: Wed, 24 Apr 2002 12:50:27 +0200
Subject: xml question?
References: <aa6121$bgf$03$1@news.t-online.com> <j4it6hjsjb.fsf@informatik.hu-berlin.de>
Message-ID: <aa62jq$p0g$07$1@news.t-online.com>

"Martin v. L?wis" <loewis at informatik.hu-berlin.de> schrieb im Newsbeitrag
news:j4it6hjsjb.fsf at informatik.hu-berlin.de...
> "Alex Adrian" <Alex.Adrian at gmx.de> writes:
>
> > i would like to make this string xml conform? is there an easy
> > filterfunction in python to convert characters like "<>??" .... to xml
> > conform characters??
>

thanks martin,

great, it works,
gruss,
alex




From chris at cmb-enterprises.com  Tue Apr  9 12:55:35 2002
From: chris at cmb-enterprises.com (Chris)
Date: Tue, 09 Apr 2002 12:55:35 -0400
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au> <a8v0mo$j50$1@newsg1.svr.pol.co.uk>
Message-ID: <chris-9CB264.12553409042002@corp.supernews.com>

In article <a8v0mo$j50$1 at newsg1.svr.pol.co.uk>,
 "?" <simon at piranhaflt.freeserve.co.uk> wrote:

> I'm a newbie myself, so this is probably wrong, but wouldn't this be *even*
> better?
> 
> x = 2
> def wrong(num):
>     x = x+1
>     print x
> 
> wrong(x)
> wrong(x)

The problem with this, in a dynamically typed language like Python, is 
that I could do the following:

>>> x = 2
>>> def wrong():
>>>    global x
>>>    x = x + 1
>>>    print x
>>>
>>>
>>> wrong()
>>> x
3
>>>
>>> x = "hello"
>>> wrong()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in foo
TypeError: cannot concatenate 'str' and 'int' objects
>>>


From jacob at boris.cd.chalmers.se.cd.chalmers.se  Wed Apr 10 12:25:08 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 10 Apr 2002 16:25:08 GMT
Subject: ANN: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <a91p14$dfq$1@nyheter.chalmers.se>

In article <3ca8a8fe$0$45639$e4fe514c at dreader3.news.xs4all.nl>,
Boudewijn Rempt  <boud at valdyas.org> wrote:
>Well, I'd like to finally officially announce that my book, _GUI 
>Programming with Python using the Qt Toolkit_ is now officially available 
>on paper. Whew!
>
>You buy it directly from the publisher's website: http://www.opendocs.org,
>
>or, preferably, at your favourite local or on-line bookshop. For instance, 
>Amazon lists it as:
>
>http://www.amazon.com/exec/obidos/ASIN/0970033044/qid=1017685892/sr=8-1/ref=sr_8_1_1/103-3219259-3855046
>
>(And I wonder how that's going to come across.) The ISBN is 0-97003300-4-4.

This ISBN is wrong, as is the one on the website. The correct one is
0-9700330-4-4. Amazon has it right.

Jacob Hall?n

-- 


From phr-n2002a at nightsong.com  Wed Apr  3 03:55:50 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 00:55:50 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com>
Message-ID: <7xit79yxd5.fsf@ruckus.brouhaha.com>

Jeff Shannon <jeff at ccvcorp.com> writes:
> > I am opposed to the addition of the new proposed type on the grounds
> > that it will make Python harder to teach both to people who have never
> > programmed before and to people who _have_.  
> 
> At first glance, I was marginally in favor of this PEP.  After 
> reading Alex Martelli's arguments, I was marginally opposed to 
> it.  I think that Laura's comments are *very* convincing, and I'm 
> now fairly solidly opposed to it.  Well said, Laura.

I agree, I missed Laura's post earlier but it's terrific.


From mwh at python.net  Fri Apr 12 05:38:22 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 09:38:22 GMT
Subject: python on cygwin
References: <mailman.1018591376.23499.python-list@python.org>
Message-ID: <lkhemhjiyr.fsf@pc150.maths.bris.ac.uk>

"eric" <eric at scipy.org> writes:

> Hello,
> 
> I'm trying to build python from scratch on cygwin and am running into "remap"
> errors such as:
> 
>     C:\cygwin\tmp\scipy_test\cygwin\Python-2.2.1\python.exe: *** unable
>     to remap C:\cyggdbm.dll to same address as parent -- 0x420000
[...]
> Can anyone point out the error in my ways, or point me to a fix?

Try linking _socket statically.  Doesn't the README say this?

Cheers,
M.
(recent NT escapee -- so no more cygwin help, I'm afraid, Jason)

-- 
112. Computer Science is embarrassed by the computer.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From logiplexsoftware at earthlink.net  Mon Apr  1 14:15:09 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 11:15:09 -0800
Subject: GUI programming
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F062334CFB33@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F062334CFB33@mail.hebmex.com>
Message-ID: <20020401111509.0101f46d.logiplexsoftware@earthlink.net>

On Mon, 1 Apr 2002 12:49:03 -0600
Gustavo Cordova wrote:

> > [..references to building wxGTK problems..]
> > All in all, this can be very frustrating to new users.
> >
> 
> Extremely.
> 
> I gave up after the n-th iteration.
> 
> And let's not talk about enabling GL. :-(

What OS are you using?  On RedHat (and I assume any other RPM-based
distro):

rpm -i wxGTK-2.3.2-1.i386.rpm
rpm -i wxPython-2.3.2.1-1-Py22.i386.rpm
rpm -i wxPython-gl-2.3.2.1-1-Py22.i386.rpm

(substituting appropriate version numbers) seems to work just fine and has
for some time.  Under Windows it's even easier (run setup.exe).  I haven't
tried a Debian-based system, but I believe deb files are available as well.

If for some reason you do have to build from source, I'm sure someone on
wxpython-users at lists.wxwindows.org would be happy to help.

> By far, Tkinter is the defacto standard precisely
> because it's ubiquitous. Not because it's the
> fasters, nicest, easiest (although it *is* quite
> simple to learn), efficientest, skinnablest,
> or whatever. But because it's the easiest to
> obtain (on most platforms), and ubiquitous.

I'd have to agree with this statement.  wxPython is a snap on Windows and
Linux, but Mac support is lagging and even on other Unices it often
requires compiling, whereas Tkinter pretty much works everywhere.

That said, if all you care about is Windows and Linux (or are willing to
put some work in to build on your target platform), wxPython is the way to
go.  It's far more comprehensive than Tk and has a native look and feel on
whatever platform you're targeting.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From phr-n2002a at nightsong.com  Fri Apr 12 19:21:14 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 16:21:14 -0700
Subject: Worldwide Lexicon project, call for volunteers
References: <e2f36e18.0204111716.78d6f0a6@posting.google.com>
Message-ID: <7xsn60h5at.fsf@ruckus.brouhaha.com>

Why yet another protocol instead of using DICT?


From pyth at devel.trillke.net  Sat Apr 27 19:30:27 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 28 Apr 2002 01:30:27 +0200
Subject: Update locals()
In-Reply-To: <ySFy8.28416$8D3.759092@news1.tin.it>
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <mailman.1019939112.20825.python-list@python.org> <ySFy8.28416$8D3.759092@news1.tin.it>
Message-ID: <20020427233027.GE16217@devel.trillke>

On Sat, Apr 27, 2002 at 10:45:50PM +0000, Alex Martelli wrote:
> holger krekel wrote:
>         ...
> > Uff. Thanks for the pointer although it is dissappointing. I always
> > liked to assume that python does this decision at run time. Doing
> > it at compile time may have performance benefits but it prevents
> > some nice uses.
> 
> It only prevents some terrible hacks that Python is MUCH better
> off forbidding.  Python doesn't go out of its way to stop terrible
> hacks, but it's sure nice when a crucial optimization has such nice
> side effects. 
> As you've noticed, it's easy to dress up dictionary
> lookup by constant keys into attribute-access syntax -- so easy,
> that you and I came up with two totally different 3-liner classes
> for the same purpose.  

speaking of it, why is it possible (read: works) to say

class dictlookup(dict):
    __getattr__ = dict.get

but not

class dictlookup(dict):
    __getattr__ = dict.__getitem__

am i missing something? 

> why ever should the dirty, terrible hacks be allowed 

i would answer "for use with exec" but you already said that
this works. You are probably going to say that "exec" is the
next candidate for evil hacks. That may be true but it certainly
has its good uses and i appreciate that assigning to locals from
within "exec" is allowed. When you want to have e.g. templated 
parts of your program (see c++) this really helps. Or would
you go as far as to say that templating is not really needed
with python?

thanks,

    holger




From mcfletch at rogers.com  Sat Apr 27 16:59:58 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 27 Apr 2002 16:59:58 -0400
Subject: vCard, vCalendar handling modules?
Message-ID: <3CCB114E.1060509@rogers.com>

Wondering if anyone has (BSD or LGPL-licensed) Python modules for 
handling these two formats?  I'm considering adding support for them to 
a program I'm writing, and would rather not re-invent this particular wheel.

What I'm looking for:

	reading into memory, preferably as nodes/objects with properties, though 
a low-level representation would work

	writing to streams/strings, preferably in the plain-text format commonly 
used in Internet Communications and defined in the respective 
specifications (http://www.imc.org/pdi/pdiproddev.html).


If someone has such a beast, please let me know.

Enjoy yourselves,
Mike
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From donn at drizzle.com  Wed Apr 24 23:06:32 2002
From: donn at drizzle.com (Donn Cave)
Date: Thu, 25 Apr 2002 03:06:32 -0000
Subject: stdin - seek() issue on Linux.
References: <mailman.1019617785.428.python-list@python.org> <1019625593.731162@yasure> <aa71nq$tukai$1@fido.engr.sgi.com>
Message-ID: <1019703987.36802@yasure>

Quoth pj at engr.sgi.com (Paul Jackson):
| The seek semantics are lower than the stdio package.  They are
| found in the semantics of the kernel.  Look at the open(2)
| and lseek(2) man pages on most any Unix or Linux or Posix
| emulation thereof.

True, and it's probably not a coincidence, since C and UNIX
are historically related.  But C is much more widely supported
than just UNIX or POSIX 1003.  Whatever seek means in Python,
it inherits that meaning from C fseek (or whatever 64-bit
alternative the platform supports, fseeko etc.)

	Donn Cave, donn at drizzle.com


From gcordova at hebmex.com  Wed Apr  3 10:35:35 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 3 Apr 2002 09:35:35 -0600 
Subject: How to check whether email /domain exists
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB4E@mail.hebmex.com>

> 
> On a more serious note, about the only thing you can really do
> is to look up the MX record and see if the host it points to
> has an SMTP server by connecting to port 25.
> 

And after that sending the message to the intended recepient(s),
if the (supposed) smtp server replies with an error, then
you know that it *probably* doesn't exists (it may be misconfigured),
and if it replies with a success code, then it means that *maybe*
the address does exist, or the server is configured to just dump
any email to nonexistant addresses and ignore any errors.

So, in short, just send the message. You can't really know with
any degree of certainty beforehand if the address you're sending
to exists or not.

Salutations.

-gustavo




From NO_SPAM_jmpurser2 at attbi.com  Mon Apr  1 13:50:56 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Mon, 01 Apr 2002 18:50:56 GMT
Subject: readlines problems - Newbie
Message-ID: <k_1q8.146744$Yv2.44641@rwcrnsc54>

Hello,

I've got a little function:
def FindLine(Record):
     im1_Dif.seek(0)
     im1_Dif.readlines(Record - 1)
     test = im1_Dif.readline()
     print test

When I call it I wind up with the same record in "test" whether I pass it 1
or 10.  The readlines line reads a huge chunk of text with no regard for
lines at all.  I have no idea what is causing this.  Where can I find the
documentation on the file methods?  I can't find them in the documentation.
Can anyone tell me what's going on here?

John Purser




From jeff at ccvcorp.com  Wed Apr  3 14:55:53 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 3 Apr 2002 11:55:53 -0800
Subject: wxpython: problem with events and checklistbox
References: <3caa2be5.48482250@news.easynews.net> <Xns91E534FF3A655rcamesz@amesz.demon.nl> <3caabb5e.1387250@news.easynews.net>
Message-ID: <MPG.1714fe2624ccad5d989693@news.nwlink.com>

In article <3caabb5e.1387250 at news.easynews.net>, stojek at part-
gmbh.de says...
> 
> Believe it of not, i changed the design ;-)
> Anyway I think there should be a way to distiguish between events
> triggered by the user (check with mouseclick) and triggered by a
> function (List.Check(i,true)). At least there should be a way to kill
> an event right after it had been raised by a function.
 
But 95% of the time that you trigger an event like that, it's 
because you *want* it to simulate user interaction.  If you set a 
checkbox, don't you want it to have the same effect as if the 
user set the checkbox?  If not, then why are you setting it to 
begin with?

Admittedly, restoring a serialized object is a special case.  The 
checkbox event handler should modify the state of the object, but 
a serialized object already *has* that modified state, and you're 
trying to set the checkbox to be in sync with that.  Typically, 
though, if you need your handler to respond differently to user 
interaction than to programmatic control, it's because your 
design is poor.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From f.onorati at tiscalinet.it  Wed Apr 10 18:29:42 2002
From: f.onorati at tiscalinet.it (Abacuc)
Date: Thu, 11 Apr 2002 00:29:42 +0200
Subject: PDCurses
Message-ID: <a92e9v$bun$1@pegasus.tiscalinet.it>

How can i use PDCurses onWin32 with Python?




From bh at intevation.de  Tue Apr 30 08:07:09 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 30 Apr 2002 14:07:09 +0200
Subject: comparing str's with ints
References: <20020429115856.R1840-100000@palanthas.neverending.org> <83adrmrwge.fsf@panacea.canonical.org> <aakhth$fme$1@slb6.atl.mindspring.net>
Message-ID: <6qpu0hbdaq.fsf@abnoba.intevation.de>

"Andrew Dalke" <dalke at dalkescientific.com> writes:

> Kragen:
> >Thus, heterogeneous collections are sortable.  This might be a wart;
> >once, couldn't raise exceptions,

cmp could raise exceptions. It's just that for builtin types it didn't.

In 1.5.2:
>>> class C:
...     def __cmp__(self, other):
...             raise TypeError
... 
>>> cmp(C(), C())
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in __cmp__
TypeError
>>> s = [C(), C()]
>>> s.sort()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 3, in __cmp__
TypeError


> and now we're stuck with its old
> >behavior in this case for compatibility.  It's kind of nice that
> >heterogeneous collections are sortable, though.
> 
> Nice perhaps, but not true.
> 
> >>> a = [1, 1j]
> >>> a.sort()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: cannot compare complex numbers using <, <=, >, >=
> >>>

It used to be true, though. Python 1.5.2:

>>> a = [1, 1j]
>>> a.sort()
>>> a
[1j, 1]

Complex numbers used to be sorted by their real parts (resp. their
imaginary parts when the real parts were equal), IIRC.

Also IIRC: the requirement that comparisons between builtin types are
always possible was lifted because you can't reasonably compare a
unicode object with a string without knowing the string's encoding.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From spam at yourself.com  Sun Apr 28 13:04:34 2002
From: spam at yourself.com (Ingo Blank)
Date: Sun, 28 Apr 2002 19:04:34 +0200
Subject: fileinput.input() fails on Windows 2K
Message-ID: <aaha32$heu$05$1@news.t-online.com>

Hi,

this tiny script gives a traceback under Windows 2K if/when input is piped
in to stdin, while it runs as expected under Linux.

## MakeDef.py
## Read ouput of DUMPBIN /EXPORTS <file.DLL> and create a .DEF file

from string import *
import fileinput

print "EXPORTS"
hit = 0
for line in fileinput.input():
 words = split(line)
 if hit and len(words) == 4:
  print "\t%s" % words[3]
 hit |= find(line,"RVA") >= 0

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

C:>DUMPBIN /EXPORTS libjpeg.dll | MakeDEF.py

Traceback (most recent call last):
  File "C:\MakeDEF.py", line 6, in ?
    for line in fileinput.input():
  File "E:\Python21\lib\fileinput.py", line 181, in __getitem__
    line = self.readline()
  File "E:\Python21\lib\fileinput.py", line 262, in readline
    self._buffer = self._file.readlines(self._bufsize)
IOError: [Errno 9] Bad file descriptor


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

Am I missing something, or is this a bug in fileinput.py ?
I'm using ActiveState Python 2.1.



--ingo
voy NG arktb QBG qr [ rot13 ]

PGP PublicKey
http://keys.pgp.com:11371/pks/lookup?op=get&search=0x18B44974





From sholden at holdenweb.com  Mon Apr  8 17:54:28 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 17:54:28 -0400
Subject: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org> <3CB21B26.8000201@hccnet.nl>
Message-ID: <6mos8.3235$vT6.3113@atlpnn01.usenetserver.com>

"Jaap Spies" <j.spies at hccnet.nl> wrote in message
news:3CB21B26.8000201 at hccnet.nl...
> Tim Peters wrote:
> > [James Logajan]
> >
> >>...
> >>Seriously Guido, how often will I need to check back to www.python.org
> >>to keep up with the changes you are making?
> >
[ ... ]
>
> Most car companies release an new model every year.
> James, I suppose you want to drive always the latest model.
>
Not at all. James only wants to drive the latest rocket-powered car if all
the old cars also run off the new rocket fuel. Otherwise he feel the car
manufacturers have broken backwards compatibilty -- even if the new
rocket-powered cars will also run on gasoline.

regards
 Steve





From lellinghaus at yahoo.com  Thu Apr 18 14:28:50 2002
From: lellinghaus at yahoo.com (Lance Ellinghaus)
Date: Thu, 18 Apr 2002 13:28:50 -0500
Subject: problems with Perl; changing to Python
References: <MPG.17291f2bfcb85301989684@news.rz.uni-karlsruhe.de> <20020418140021.I3177@eecs.tufts.edu>
Message-ID: <003101c1e706$ef147580$3201040a@angel1>

> > Are there any applications written in Python?
> > Am I right here with Python or will I have to stick to a hybid C++/Java
> > solution?
>
>   Sure. Go to sourceforge.net and search for python apps. I have
> a 5000 GUI ftp program myself (ftpcube.sourceforge.net) that uses
> pygtk. Has some outdated screenshots if you want to check it out.
> Doesn't run on windows yet since the windows port of GTK doesn't
> support threading but runs nicely on all *nixes. I have some others
> that I've written for research - particularly a graphing program using
> wxPython.
>

You can also look at Yahoo (almost all of their site) and Geocities. They
are written in Python. There are a lot of large systems written in Python
now. You would not be alone.

Lance Ellinghaus





From phr-n2002a at nightsong.com  Sat Apr 13 04:26:30 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 13 Apr 2002 01:26:30 -0700
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it>
Message-ID: <7xvgaw80nd.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> I like FP a lot, and I practise it -- but I eschew lambda in Python
> in most cases: it's limited and goes much against the grain of the
> language.  All it boils down to is, give your functions a name -- in
> most cases a well-chosen name will clarify your code.  E.g.,
> 
> def relatively_prime(a, b): return gcd(a, b) == 1
> 
> and so on...

Here's a fragment from a calculator app that I wrote to try out
Tkinter (the app is at http://www.nightsong.com/phr/python/calc.py).

It's a pair of tables saying what the various calculator buttons
should do to the displayed number when you push the buttons, possibly
preceded by pushing the "inverse" button (pressing "inv sin" gives you
the arcsin, etc).  Using lambdas for the functions where there weren't
ready-made built-ins seemed natural and obvious to me.  Would you really
have written separate definitions for those functions?

    unops = {'sqrt': math.sqrt,
             'sin': math.sin,
             'cos': math.cos,
             'tan': math.tan,
             'ln': math.log,
             'log': lambda x: math.log(x)/math.log(10),
             'clr x': lambda x: 0
             }

    inv_unops = {'sqrt': lambda x: x*x,
                 'sin': math.asin,
                 'cos': math.acos,
                 'tan': math.atan,
                 'ln': math.exp,
                 'log': lambda x: 10.**x,
                 'clr x': unops['clr x']
                 }


From newsfroups at jerf.org  Tue Apr 30 11:47:39 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Tue, 30 Apr 2002 15:47:39 GMT
Subject: OT: Happy Birthday Larry Niven
References: <3cce9f3a$1_6@goliath.newsgroups.com>
Message-ID: <3CCEBD17.5040903@jerf.org>

Brad Clements wrote:
> I think Python is a language that Motie Engineers would use...

Oh goodness, no. They'd definately *prefer* perl; Python is too damned 
clean and very, very human.

Even perl's probably a little on the clean side, for them. Motie 
engineers would go beyond "There's more then one way to do it", they'd 
want "It does more then one thing at a time." They'd probably like a 
language where every *char* can do at least two or three things. My 
guess is assembler with a suitably kooky CPU.

This is a Motie programmer at work: 
http://tuxedo.org/jargon/html/The-Story-of-Mel.html . See, he uses the 
same opcodes as both code *and* data at some points... just as an example.



From logiplexsoftware at earthlink.net  Wed Apr  3 16:07:41 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 13:07:41 -0800
Subject: Threading problem
In-Reply-To: <uit78wl07.fsf@ctwd0143.fitlinxx.com>
References: <a8f8f6$vdu$1@news.rchland.ibm.com>
	<uit78wl07.fsf@ctwd0143.fitlinxx.com>
Message-ID: <20020403130741.5b5e4fe8.logiplexsoftware@earthlink.net>

On 03 Apr 2002 16:05:44 -0500
David Bolen wrote:
 
> Ignoring the specific coding error that has been addressed by other
> responses, it should also be pointed out that if you're trying to get
> effective utilization of your pair of CPUs, this probably won't do it.
> 
> The Python global interpreter lock requires that even with native
> threading, only one thread is actually interpreting Python code at any
> moment in time.  The net result of this is that while threading is
> excellent for overlapping I/O, it won't efficiently distribute CPU
> amongst processors.  That's because if you are CPU bound, then you are
> stuck in Python code execution, and only one processor can be doing
> that at any given moment in time.
> 
> This can be improved somewhat if you use extensions that release the
> GIL, but in most cases that's done during I/O and not processing time,
> particularly if the latter will require access to Python structures.
> But in theory a purely CPU bound extension could release the GIL while
> performing computations that it knew didn't require any interface to
> Python.

Ah.  That's pretty much what I thought.  Of course the solution would be to
write the system as two or more separate Python programs (perhaps started
by a single process) that use some form of IPC.  There's the added
complexity/overhead of the IPC mechanism, but you'd get full advantage of
SMP.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From timr at probo.com  Wed Apr 10 23:44:46 2002
From: timr at probo.com (Tim Roberts)
Date: Wed, 10 Apr 2002 20:44:46 -0700
Subject: Python server from inetd?
References: <iml4bukknqf9nqe13h7dp0qco3qj2r86rt@4ax.com> <kuwuvhmoiy.fsf@lasipalatsi.fi>
Message-ID: <mb1abucnutggsi6v22mbo03qisn1l8agm3@4ax.com>

Erno Kuusela <erno-news at erno.iki.fi> wrote:

>Tim Roberts <timr at probo.com> writes:
>
>|   #! /usr/bin/python
>|   import socket
>
>|   client = socket.fromfd( 0, socket.AF_INET, socet.SOCK_STREAM )
>|   while 1:
>|     bfr = client.read(1024)
>|     if not bfr:
>|       break
>|     ...
>
>| But this does not work.  The "read" call returns 0 immediately, over and
>| over, until the client exits.  I never get any data.
>
>it shouldn't work, since socket objects don't have a read method, only
>recv.  getting 0 sounds really odd, since both s.recv nor f.read only
>return strings. are you sure you're interpreting this correctly?
>can you post the real code?

You're right.  That's what I get for posting from memory instead of from a
file.  I did figure out that I needed to use stdin and skip the
translate-to-socket model.

Here's the code in its most recent incarnation.  This prints "Got empty
buffer" forever.

import os
import sys
import time

log = open('/tmp/tinetd.log','a')
sys.stderr = log
sys.stderr.write( 'Hello\n' )

# Suck all data from the socket until a line ends with Ctrl-D.

bfr = ''
while 1:
  bfr1 = sys.stdin.readline()
  if bfr1:
    log.write( "Got %d\n", len(bfr1) )
    bfr = bfr + bfr1
    if bfr[-1] == chr(4):
      log.write( "Got ctrl-d\n" )
      break
  else:
    log.write( "Got empty buffer\n" )

log.write( repr(bfr) )
log.write( '\n' )
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From mfranklin1 at gatwick.westerngeco.slb.com  Mon Apr 22 07:08:14 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Mon, 22 Apr 2002 11:08:14 +0000
Subject: strip/trim a string
In-Reply-To: <3CC3D89E.9115E1A5@itea.ntnu.no>
References: <3CC3D89E.9115E1A5@itea.ntnu.no>
Message-ID: <200204221008.g3MA8mr24797@helios.gatwick.geco-prakla.slb.com>

On Monday 22 Apr 2002 9:32 am, Bj?rn OveGr?tan wrote:
> yes, I'm a newbie at python.
>
> and I need to strip/trim a string... e.g. I have a string of unknown
> length, and only care
> for the last five(5) characters.
>
> An example string:
> xxxNOGBJ  94873
>
> where 94873 is the only part of the string I want to keep


If it is allways the last five then :


s="xxxNOGBJ  94873"

s[-5:]


at the python int.:
$ python
Python 2.0 (#11, Nov  6 2000, 09:01:06)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> s="xxxNOGBJ  94873"
>>> s[-5]
'9'
>>> s[:-5]
'xxxNOGBJ  '
>>> s[-5:]
'94873'
>>>

take a look at:-

http://www.python.org/doc/current/tut/node5.html#SECTION005120000000000000000



>
> best regards
>
> Bj?rn Ove




From sholden at holdenweb.com  Tue Apr 30 08:45:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 08:45:51 -0400
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org> <200204292311.57363.bhan@andrew.cmu.edu>
Message-ID: <Bowz8.12964$xy4.4358@atlpnn01.usenetserver.com>

"Benjamin Han" <bhan at andrew.cmu.edu> wrote ...
> On Monday 29 April 2002 10:02 pm, Cliff Wells wrote:
> > On 29 Apr 2002 18:54:12 -0700
> >
> > Fortepianissimo wrote:
> > > This question is so fundamental that I strongly suspect it's already
> > > answered - ig that's the case, please forgive this poor newbie, for he
> > > has done some work searching over Python FAQ and Google.
> > >
> > > Basically I want to have a subclass of list, like
> > >
> > > class MyList (list):
> > >     ...
> > >
> > > then I want to
> > >
> > > l=MyList()
> > > ... do stuff to fill l
> > > s=" ".join(l)
> >
> > s = " ".join([str(i) for i in l])
>
> This surely looks nicer. I'm just wondering how far Python has fared in
terms
> of this concept of OO? This example seems to show a very different
character
> of OO in Python, in that __str__ is not automatically called, which seems
> rather counter-intuitive.
>
I'm not sure I'd agree - polymorphism doesn't necessarily imply type
conversion, which is what's required here. I agree it *is* annoying that the
.join() method will only join sequences of strings, and perhaps there is a
case for having it call the __str__() or __repr__() methods of the objects
it is dealing with (since the result is always going to be a string).

> Without private members, and with seemingly complex ways of dealing with
class
> members and methods - so what's the state of OO in Python? Is it still
> rapidly evolving? How much is an effort in converting Python into C++?
Anyone
> would like to comment on some of these questions, or point to some
article,
> FAQ (esp. helpful from a C++ programmer's point of view)?
>
The implementation of OO is currently in a state of rapid development. Until
2.2 the user-defined classes were entirely separate from the built-in types,
but they have moved much closer together, with a view to eventually making
everything a first-class object and allowing sensible subclassing of int,
dict and the other built-ins.

These changes have been made without introducing any significant backward
incompatibilities (for some value of significant). This has not stopped the
user community expressing collective nervousness, despite the remarkably
short list of actual code breakages triggered by relatively recent change.

Python is pretty much WYSIWYG in object terms. It has enough introspection
to make fully dynamic programming and the creation of IDEs possible and
natural. Of course, this makes it easier for the adventurous programmer to
shoot zieself (zerself? Aahz?) in the foot. By and large Pythonistas are
prepared to forego the protection of privacy as implemented in (say) Java or
C++ for the flexibility of being able to do more interesting stuff.

As far as I am aware there is no existing effort to translate Python into
C++. A Python to C translator was, I believe, abandoned and languishes. Some
interesting work has been done into using compiling techniques on Python
functions (Google for Psyco, I think), and gainsays some of the conventional
wisdom about Python's dynamic typing not being well-suited to compilation.

A C++ translator that worked and produced measurable speedups over existing
interpreters would win its implementor the undying affection of many Python
programmers. The remaining 115% would spend their time complaining that the
language still didn't implement the ternary choice operator :0)

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From cliechti at gmx.net  Tue Apr 30 19:31:45 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 May 2002 01:31:45 +0200
Subject: telnet
References: <5cFz8.127$ke3.35@newsread2.prod.itd.earthlink.net>
Message-ID: <Xns9201FD27582Dcliechtigmxnet@62.2.16.82>

"Billy Ng" <kwokng at earthlink.net> wrote in
news:5cFz8.127$ke3.35 at newsread2.prod.itd.earthlink.net: 
> tn = telnetlib.Telnet(host)
> tn.read_until("Please enter your id")
> tn.write(userid+"\r\n")
> 
> From here, tn will return either "Please enter your password" or
> "Wrong id, please enter your id again".  My question is how can I get
> the returning string before I do the next tn.write()

can't you go for a newline? often such messages print on a separate line, 
so you can just "readline" and use string methods to find out if it was 
positive.

chris

-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Wed Apr 17 16:49:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 16:49:54 -0400
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr><mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <mailman.1019073733.27071.python-list@python.org>
Message-ID: <Sflv8.6806$Ip3.2223@atlpnn01.usenetserver.com>

"Ian Bicking" <ianb at colorstudy.com> wrote ...
>
> But that's weird...
>
> q1 += q3
> is not equivalent to:
> q1 = q1 + q3
>
It may be weird, but it's Python. Back when augmented assignment was being
discussed, I suggested that this might confuse newcomers, who would assume
that augmented assignment was merely a syntactic shortcut.

In many cases it is, but equally sometimes it isn't. The implementation of
any particular mutable object is free to perform in-place update for
augmented assignments, thus optimizing certain aspects of performance in
return for a modest amount of confusion.

While it might seem strange, rest assured that such is the language
designers' intent.

regards
 Steve






From claird at starbase.neosoft.com  Fri Apr 19 09:36:38 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Apr 2002 08:36:38 -0500
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com>
Message-ID: <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net>

In article <3CBF9168.AAE0FC46 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>"Frederick H. Bartlett" wrote:
>> 
>> I was recently introduced to OmniMark. One of our exercises was to take
>> a plain text file of Hamlet and convert it to SGML.
>> 
>> So I did it in Python, too. But the best time I could get from Python
>> was .57 sec, while OmniMark came in at .20 sec. What's the most
>> efficient technique for Pythonesque regex-based text processing?
>
>Hmmm... how fast do you need it to be?  Sounds to me like 0.57 seconds
>is pretty darned fast.
>
>Do you have specific goals, or are you just on a search for 
>something faster?  Remember, "better is the enemy of good"
>and the grass is always greener.
>
>(See http://www.seds.org/~chrisl/akin.html )
>
>-Peter

I was counting on Peter to write this.

Because it's correct, of course.  Moreover, you should
know that Peter has payroll responsibilities.  He writes
from a more practical perspective than other acquaintances
might afford you.

Suppose, for a moment, that we make a serious attempt to
investigate "the most efficient technique for Pythonesque
regex-based text processing".  While your question might
seem a reasonable one, in fact it leaves too many degrees
of freedom to allow for a precise response.  First, if I
truly needed speed in big SGML work from Python, I'd exer-
cise the freely-available SGML-specific modules available.
Next, I'd determine whether my test examples are indeed
regex-bound (it might well be I/O which constrains your
performance).  After that ... well, part of the charm of
regex-s for some people is that they're so flexible that
different techniques are superior in different circumstances.
-- 

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


From sholden at holdenweb.com  Sat Apr 27 06:47:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 27 Apr 2002 06:47:50 -0400
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org>
Message-ID: <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote at almost interminable length
...

[disclaimer: this time I really will not be following up, unless I can do so
in less than twenty words...]
>
> "James J. Besemer" wrote:
>
> > Steve Holden wrote:
> >
> >> Well, on this particular group I feel you *are* in the minority.
> >
> > As I acknowledged up front.
> >
> >> Many c.l.py
> >> readers are *not* "professional programmers", but chemists,
> >> mathematicians,
> >> biophysicists, and a host of others for whom programming is a means
> >> to an
> >> end. Many others are, like me, professional programmers who attempt
> >> to
> >> increase Python's popularity by assisting those less familiar with
> >> programming concepts.
> >
> > Some of the better "professional programmers" that I have had the
> > pleasure to work with or employ had formal training in chemistry,
> > physics or math rather than "computer science".  Some of the worst
> > programmers had CS degrees from big name schools and couldn't design
> > their way out of a paper bag.  I think the scientists tend to come to
> > the subject of programming with more maturity, discipline and focus
> > than some with formal [sic.] training in CS.
> >
Perhaps because they see programming as a means to an end and not just an
end in itself.

> > I think some people have a natural aptitude for programming and some
> > people do not.  It has nothing to do with educational background.  The
> > ones with the aptitude excel and the ones without tend to average or
> > sub average work.  I also know that basic programming tasks may be
> > completed by children, so the basic skills are not at all hard to
> > master.
> >
> > I also have had experience coaching newbies in Python and other new
> > languages and also I have spent time teaching programming to middle
> > school kids.  At one point in my life, I worked supporting a bunch of
> > EE researchers who were not programmers, so I think I actually relate
> > to this community better than it may appear on the surface.
> >
> > I guess my main point, which I don't want to get lost is that we DON'T
> > have to dumb down the language in order to help beginning users.
> >
Well, the "future division" discussiosn seemed to see most resistance to the
change from long-term programmers, though really it's probably no big deal.
And you must know (or not) that Guido has interests in "Computer Programming
for Everybody"?

> >> Furthermore, you express exactly the kind of attitude that gets
> >> "professional programmers" a bad name. Let's remember our social
> >> skills here
> >> ;-). Remember, Python is *not* just for professional programmers, so
> >> you
> >> shouldn't write as though it were "your" language.
> >
> > I apologize for any offense I may have given.  I'm new here and I'm
> > the type of person that tends towards direct talk.  I see in this
> > social setting it distracts from my message rather than (as I am used
> > to) lending clarity.  Please forgive me for coming on too strong.
> >
That's OK. I'm from Yorkshire myself, and I've had to consciously reduce my
directness level somewhat since I moved to the USA. It's probably not your
fault you got up my nose ;-)

> > I certainly make no presumption that Python is "my" language.  If I
> > were king, there is a very short list of relatively minor things I'd
> > change immediately, adding a ternary operator being one of them.  And
> > if a topic I care about comes up on the list (I did not start the
> > thread) I feel I have every right to chime in with my opinion, humble
> > or otherwise.
> >
Right, right.

> > More importantly, I don't think we have to make a choice to optimize
> > only for "pros" or only for newbies.  Seems to me already Python is a
> > large and in places a complicated language.  To it's credit, it's way
> > easier to learn and use than Perl, which is the closest practical
> > alternative around today.  There are and always will continue to be
> > advanced topics that will take newbies time to absorb (possibly
> > forever).  Given the overall complexity of the system, seems to me
> > that little items such as a ternary operator are way down in the
> > noise.
> >
> > Does anybody honestly think the addition of a ternary operator would
> > be more confusing to newcomers than the business with lexical scope
> > changing behavior over time?  Iterators?  Generators?  List
> > comprehensions? Unifying types and classes?  Sockets, Select, Polling
> > Objects, multi-threading and synchronization of concurrent
> > programming?
> >
> > No.  Arguing that comparatively trivial extensions such as ternary
> > choice will hurt beginners is pure rhetoric, simply superstitious
> > behavior.  The conclusion is foregone and you use this specious
> > argument to rationalize your prejudice.
> >
Well, like some of the features you mention above, a ternary operator could
easily be ignored. But in that case why have it in the language. I simply
tried to assert my own opinion (as someone whe's been programming a long
time and doesn't much like C) that an if: then: else: combination is
clearer.

There are reasons (mentioned in the FAW) why Python is not an
expression-based language such as Icon, and the philosophy of ternary ? is
kind of against that.

Ultimately we are probably simply debating our own prejudices, though, and I
have no objection to you preferring yours ...

> > For the most part, beginners are focused on getting their own programs
> > to work.  Like with Lexical scope, they can simply ignore the issue
> > for the most part until they gain sufficient experience with the
> > language (along with larger applications).  If they run into questions
> > about constructs they do not understand, then they can look things up
> > in the doc.  But seriously, is your typical newbie going to have more
> > trouble understanding a simple inline choice operator or some code
> > that relies on some subtle aspect of dynamic scoping?
> >
> > Seriously, would something like
> >
> >     x = ( if a>b : a else: b )    # trivial example, would be more
> > complex in real life
> >
> > truly be all that much more confusing for your average newibe than
> >
> >     y = [ func( i ) for i in collection ]    # trivial example, would
> > be more complex in real life
> >
> > Wouldn't the second, existing construct actually be MORE confusing?
> >
> > Ha!  I bet your mythical newbie would be LESS confused by the proposed
> > ternary operator than by the recommended substitute:
> >
> >     x = ( a > b and [ a ] or [ b ])[ 0 ]
> >
> > It'll confuse almost everybody.
> >
> > We already added an "inline FOR", what could possibly be so bad about
> > an "inline IF"?
> >
Well, apart from the if's in list comprehensions such as [x for x in
range(5) if x % 2], I presume you mean. Note that the list comprehension
doesn't include any colons, however. Patching up the syntax might not be a
simple matter, but I haven't seen any suggestions I like.

> > Finally, many the sophisticated concepts mentioned earlier I expect
> > were dreamed up by people with specialized training in software.  Not
> > to put anyone down but people with specialized training and experience
> > in the field of programming and programming languages often are better
> > able to articulate new features like this and figure out how best to
> > fit them into the language.  Just a factual statement, no offense
> > intended to anyone trained in different fields.
> >
> >> > So I am not swayed by arguments that :? confuses people.
> >>
> >> So, if one is confused, one should leave the job of programming to
> >> the
> >> professionals? That will help the programmer shortage.
> >
> > That's not at all what I said.
> >
> > Rather, (a) the average user is smart enough to NOT be confused, and
> > (b) there likely is something to be learned from people with a lot of
> > technical experience and expertise in the matter.
> >
> > You refer to all the scientists who use Python.  I'm sure they've all
> > taken calculus, linear algebra and undoubtedly mastered lots more
> > complicated subjects.  I find it impossible to believe that this
> > august group could not in time come to master the syntax and semantics
> > of such a simple, useful and not all that unfamiliar of a construct.
> > If I was one of these mythical users, I'd be insulted that you presume
> > I'm so stupid.
> >
> > I sense that this is a religious issue, so both parties will resort to
> > elaborate arguments, well out of scope of the actual issue.  All
> > bullshit aside, do you really and sincerely in your heart believe that
> > this feature alone would "bloat" the language, actually harm the user
> > community and materially detract from Python's growing "following?"
> >
Not really. I believe this discussion has appeared so many times before that
we can reliably predict it's not going to happen. Just trying to save you
some time ...

> >> > But I think part of the problem
> >> > with inline choice is that people who use it do not properly
> >> > parenthesize and indent it to properly highlight what's going on.
> >> > I always split the test and the two alternatives on separate
> >> lines,
> >> > with each of the latter two starting with the special operators.
> >> >
> >> >     def abs( n ):
> >> >         return (( n > 0 )
> >> >                       ? ( n  )
> >> >                       : ( -n ))
> >> >
> >> I take it you're using some sort of "desirable" extension syntax
> >> here?
> >
> > Yes.  Here, I'm attempting to employ C's :? operator within Python.
> > I'm sorry, previous posts seemed to assume that people generally knew
> > what it meant, so I apologize for relying on that evidently
> > presumptuous presumption.
> >
> >> I completely fail to see how this is preferable to the instantly
> >> comprehensible
> >>
> >>     def abs(n):
> >>         if n > 0:
> >>             return n
> >>         else:
> >>             return -n
> >
> > In this case you are right there is little advantage.  But I find
> > cases frequently arise where the decision appears nested in a larger
> > expression (other than the entire body of a function) in which return
> > would not be a legal or meaningful alternative.  [For a change of
> > pace, and a test of your intellectual honesty, you give an example and
> > I'll translate it to existing Python.]
> >
Not sure I like the flavor of that last sentence, but we'll overlook it for
now.

> > In any language, you always can get by with additional declarations
> > (functions and/or auxiliary variables).  To some, this improves the
> > readability of the program.  To others, like myself, it obfuscates the
> > meaning.  In any case, creating additional steps creates extra,
> > unnecessary work for the computer and there are cases in inner loops
> > where such differences in performance may be significant.  Since the
> > extra program steps furthermore are unnecessary, it's doubly
> > irritating to have to spell it out.  I.M.H.O.
> >
So, you tend towards the functional. Nothing wrong with that, but you should
be aware that Guido's on record as regretting allowing the modest functional
constructs the language already has, and presumably leaves them in to avoid
the anathema of code breakage.

> > And to say there's already some other way to do it in Python is doubly
> > disingenuous.  You can say that about just about half the features in
> > the language -- Generators, Iterators and List Comprehensions.  Toss
> > 'em out, no need.  Easily done another way.  Hell, by that rationale
> > we can get rid of FOR and IF, as there's way to express everything
> > they can do in terms of WHILE.
> >
By that rationale, we can do all our programming on a simulated Turing
machine, which can compute all computable functions. We are, after all,
arguing about what we see as "best" for the language. Our opinions differ.

> >> By the way, please note that there *is* a style guide for Python,
> >> which you
> >> will, find at
> >>
> >>     http://www.python.org/peps/pep-0008.html
> >>
> >> It unequivocally states that spaces should not be used after opening
> >>
> >> parentheses or before closing ones. So, while we're talking about
> >> "proper"
> >> usages, let's get them *all* right, shall we?
> >
> > Thanks for picking this tiniest of nits from my fur.
> >
Tee, hee. Sorry, that was indeed a cheap shot. But, at the same time, you
should consider it - style is  important in communication about a languge,
and there is a recommended style. I hope you aren't suggesting that
stylistic recommendations should be ignored?
.
> > I actually am familiar with the style guide and even how it's changed
> > over time.  IIRC, it used to be 8 column tabs were god's inviolate
> > creation (which I still tend to favor), with 4 ok if you did penance.
> > Now 4 is standard and god's own are the also ran.  Who knows?
> > Someday, spaces like I use them might become standard.  (Yeah, right
> > ;o)
> >
> > For the record, while I whole heatedly agree with most of the things
> > in the style guide, including many of Guido's personal prejudices, I
> > do happen to disagree on a few points.  And, similarly, after over 25
> > years of writing software for a living there is no point in trying to
> > convince me there's a better way.  Besides, "A Foolish Consistency is
> > the Hobgoblin of Little Minds" and "sometimes the style guide just
> > doesn't apply."  Right?  Seems to me, little example snippets in an
> > email is last place where the rules need to be strictly enforced
> > (unless you're one of those Hobgoblins).
> >
While I don't consider myself a hobgoblin it's probably for others to say.

> > Philosophically, I feel the clarity of written code is of paramount
> > importance (aside from actually working as intended).  I adhere to
> > Cornell CS Professor David Gries' advice to liberally use white space
> > to improve the clarity of your code.  In one effective presentation he
> > admonished his audience of professional programmers to "open up your
> > code and and let it breathe."  He accompanied his advice with a very
> > dramatic slight of hand with transparencies on the overhead projector
> > where a compact snipped of sample code opens up with white space and
> > indentation throughout.  It's always stuck with me and I have found
> > that the extra spaces inside parentheses and square brackets do
> > noticeably improve comprehension of existing code.  It's more work to
> > put them there but to me it's worth it.  Quite simply it more clearly
> > delineates the inside from the outside.  In ten years when all y'all's
> > vision starts to fade, you may finally appreciate the wisdom of this
> > advice.
> >
This is the same David Gries whose book on compiler technology used ASCII
art for the diagrams? I cursed him in my twenties (I went to college late),
but not because the book was flawed in any other way. Those diagrams stank,
so I suppose he has learned better since.

Again, you are free to adopt the style that suits you. I prefer, in my
written communications, to use a commonly-accepted style which (by and
large, hobgoblins excepted) will find most acceptance in the broad Python
community. It's a courtesy to my readers.

> > Things I particularly agree with in the style guide are (a) the
> > importance of consistency varies depending on scope; (b) you generally
> > need to apply some common sense; and (c) if you're working on someone
> > else's code you should respect the style already present (unless
> > you're willing to redo the whole thing properly).
> >
> >> > Thus although you can use and/or as a partial substitute, there
> >> are
> >> problems with the hack:
> >>
> >> Indeed. An experienced (professional?) programmer would, of course,
> >> consult
> >> the FAQ. See
> >>
> >>     http://www.python.org/cgi-bin/faqw.py?query=ternary?req=search
> >>
> >> for further discussion on this topic.
> >
> > The FAQ admits the very "flaw" and "ugly"ness (FAQ's term) with the
> > alternatives.  I didn't need to look because I already was well
> > acquainted with the issue.  Even the sample function they give as an
> > alternative is not a complete equivalent.  The very fact that this
> > topic is a Frequently Asked Question and one with such a lame answer I
> > think emphasizes that it's a true WART on the language.
> >
> >> I agree with this. But don't try to suggest that there aren't
> >> problems with
> >> the ternary construct in C, most notably the omission of an "=" in
> >> the test,
> >> turning a test into an assignment.
> >
> > If you are referring to the confusion between "=" and "==" -- that
> > problem is completely unrelated to the ternary construct and
> > absolutely would not apply to Python extended only with a ternary
> > operator.  Rather it is a problem with C's concept that assignment is
> > an operator that may appear in any expression rather than a
> > statement.  So long as assignment remains a statement, then there is
> > no problem with omitted "=" that I am aware of.  Python could diagnose
> >
> >     a = if a = b : 0 else: 1        # syntax error!
> >
> > as an error just the same as
> >
> >     a = ( b = c )     # syntax error!
> >
> > is today.
> >
But once the ternary was in ...

> > Then too, I'm crazy enough to count assignment operators as another
> > one of Python's few serious omissions.  I wish to hell I could simply
> > write:
> >
> >     while data = socket.recv( 1024 ):
> >         process( data )
> >
Aha! So you are a closet "Python should be C" adherent? You should (since
you are thoroughly familiar with the above FAQ entry) know from FAQ 6.30
this is a definite no-no. Indeed, once you allow assignment operators you
bring in as allowable the very case you say above that Python would be able
to detect easily. The FAQ unambiguously states """The reason for not
allowing assignment in Python expressions is a common, hard-to-find bug in
those other languages, caused by this construct:
    if (x = 0) {
        ...error handling...
    }
    else {
        ...code that only works for nonzero x...
    }

"""

> > The recent addition of file iterators helps some but that's only one
> > case.  Are we going to add iterators to sockets?  If so, should it
> > default to recvline() or some other operation?
> >
There can be only one?

> > While it is true that assignment operators do cause some confusion, it
> > nevertheless again is a simple concept, readily mastered by even
> > novice programmers.  It's not brain surgery.  I've taught non
> > programmers how to program in C and they tend to pick up on these
> > several idioms right away.  Arrays, arrays vs. pointers and
> > multi-dimensional arrays seemed to be the bigger language problem,
> > followed quickly by tools and libraries particular to a particular lab
> > environment.
> >
This just ain't gonna happen, and these discussions can therefore be
considered a waste of bandwidth. I have used expression-based languages
myself. I thought Icon was much preferable to C, taking the whole thing to
further extremes and including generators.

> > Furthermore, to the extent there is risk, there are express remedies.
> > I worked for a while with a dialect of AWK.  AWK, like C, treated
> > assignment as an operator.  To protect people from the '=' vs. '=='
> > confusion, the AWK compiler adopted a convention that it would issue a
> > warning for any top level assignments that appeared within an
> > expression.  To suppress the warning, an additional level of
> > parentheses was required.  This actually worked out quite well,
> > stylistically, as the extra parentheses gave the reader a clear,
> > visible hint that something unusual was going on.  It's a good
> > convention which I retain even in C/C++, where it means nothing to the
> > compiler but aids humans reading the code later on.
> >
So a crappy language with warnings is better than a clean one without them?

> >> > (b) Since and/or lacks clarity it arguably is more likely to
> >> > people than inline choice.  And whereas :? confuses beginners,
> >> > and/or (since it's an idiom rather than a language construct)
> >> > is more likely to confuse more advanced users.
> >>
> >> [I presume that should have been "...likely to confuse people ..."].
> >
> > Yes, thanks.
> >
> >> Frankly, I don't think a language that "confuses beginners" is
> >> likely to
> >> develop a huge following, since everyione is a beginner at the
> >> start. I
> >> think confusion of *any* kind is to be avoided, and Python does as
> >> well as
> >> most languages in this respect. I particularly dislike your
> >> implication that
> >> if anyone is to be confused it should be the beginners and not the
> >> "professional programmers".
> >
> > Of course, that's not what I meant.  I'm saying that the and/or idiom
> > likely will confuse more people (INCLUDING experts) than an explicit
> > ternary operator that's a documented standard part of the language.
> > Don't you agree?
> >
Well, given that I suggest the existing if: then: else: is to be preferred,
yes, I agree. But the FAQ entry specifically points out it's for those who
would wish a ternary ? on the language.

> >> > To me clarity of expression is the most important quality of a
> >> well
> >> > written program.
> >
> >> To me, clarity of expression is the most important quality of a well
> >> written
> >> program.
> >
> > I find it very encouraging that if you go high enough up the tree,
> > there is some common ground; in fact we are absolutely agreed.
> >
> > Even if we disagree about the details and outcomes we both have the
> > same objective in mind, which is a big head start.
> >
Who are you calling a big head?

> >> Binary conditional evaluation ala ? is completely unnecessary. It's
> >> eminently reasonable for programmers to feel a need for such an
> >> addition to
> >> Python, but once they have been reeducated the feelings will go
> >> away.
> >
> > "Reeducated" raises for me the specter of the Khmer Rouge.  Violent
> > revolution and anarchy.  Kill all the professionals and "reeducate"
> > anybody who resists.  I'm sure that's not what you meant.
> >
Well, actually it was. But in a gently ironic sense, rather than an "I'ma
get my gun" sense. I realise not everyone agrees with me all the time.

> > Too, I am fairly confident this feeling will not go away.  The several
> > gratuitous departures from C are a daily annoyance with each new
> > Python module I write.
> >
> >> I disagree. The biggest issue is the bloat this will introduce into
> >> the
> >> language, and programs written in it, without the least advantage.
> >> YAGNI.
> >
> > I think it's unreasonable to refer to this tiny, easily understood
> > feature as 'bloat', especially given the scope of Python now and the
> > myriad other features that already spilled over the dam.
> >
> >> Blerch. This is supposed to be easy to understand? In which
> >> universe. I
> >> submit it's completely unpythonic.
> >
> > Just an alternative.  I thought using the ":" to separate the question
> > from the answers made a kind of sense.  But I'll have to defer to your
> > expertise on what is or is not "Pythonic".
> >
No you won't, that's a game *anyone* can play. As you pointed out (and I
agree), it's not my language either.

> >> Well, there were problems with Algol 60's conditional expressions,
> >
> > Oh?  Something other than the dangling else ambiguity (which does not
> > apply in this case)?
> >
no, nothing else.

> >> but I
> >> agree that *in a free-form language* they are useful and fairly
> >> natural.
> >
> > Wow, so you don't simply hate them altogether.
> >
Not at all.

> >> Python, of course, is not a free-form language, and so inline
> >> conditionals
> >> of any kind are a far less natural fit with the rest of the
> >> language.
> >
> > Fascinating.  I'm not sure I'm getting this but it sounds like this is
> > a fundamental point that makes a ternary operator good or bad.
> >
Not really. Ternary good or bad isn't the issue (for me). The issue is
ternary in or out. The situation now is it's out. I am suggesting it will
stay out, is all.

> > I learned yesterday that Guido says Python is an Algol derivative.
> > Isn't that "free form"?
> >
> > Please define "free-form language", with emphasis on why Python is not
> > one.  Then please explain how free-formed-ness has any bearing either
> > way on allowing a proposed construct, which you otherwise admit may be
> > "useful and fairly natural."
> >
> > I am truly eager to learn, master.
> >
Snide. I simply meant that when indentation is used to indicate block
structure it becomes more difficult to detrmine an appropriate layout for
expressions of this nature. You yourself have referred to simple one0line
constructs, but suggest it would be more natural to spread a ternary over
three lines. I simply suggeested that given that preference if: then: else:
isn't much worse.

> >> This isn't the first time this topic has been dragged out of
> >> retirement, and
> >> I don't suppose it will be the last.
> >
> > I expect it'll be a continual nuisance unless/until Python adopts the
> > changes.  The majority of SW development done on Unix and on PCs is by
> > C/C++ programmers and a large fraction of college grads come
> > pre-programmed to know one or both languages.  To that huge body of
> > prospective users, the several gross and unnecessary inconsistencies
> > with C appear to be glaring errors.  I know a lot of people with that
> > background who reject Python out of hand (in favor of, Bletch, Perl)
> > simply because of those perceived inconsistencies.  Sad and unfair but
> > true.
> >
Well, should we also introduce a COMPUTE statement so the COBOL programmers
will feel at home?

> > I accept I'm probably pissing in the wind on all this but I think it's
> > an important consideration when trying to attract a substantial
> > "following".
> >
The worst thing Python's developers could do is "try to attract a
substantial following", and I think it's to Guido's credit that he expects
the excellence of the language to dictate its eventual popularity rather
than trying to advertise it.

> >> It's probably obvious that your post
> >> got up my nose -- I'm not normally so crotchety, so I guess I should
> >>
> >> apologise for the cranky nature of my reply.
> >
> > Well, you thought I was being uncivil when I did not intend to be and
> > then your broke Miss Manners' first rule that rudeness (perceived or
> > otherwise) does not justify rudeness in turn.  No problem.  I have no
> > problem with criticisms about my ideas, or just about anything else
> > short of personal insults.
> >
I have no problem with this either, and indeed chose to apologise rather
than rewrite when I got to the end of the post.

> > I have nothing but an abundance of goodwill for Python and it's users
> > base.  That's why I'm here.  Any suggestions I may make are sincerely
> > well intentioned, whether it may seem that way or not.
> >
> >> I shall not be contributing more to this thread.
> >
> > Ah, the old, chicken shit, take some pot shots and run away.....
> >
> > Worse: you wait until NOW to tell me!   You shoulda put this
> > disclaimer at the top of your note -- woulda saved me a lotta time.
> > ;o)
> >
So you saw the disclaimer at the top of this one, then? I hate that you made
a liar out of me, but since you had the foresight to email your reply as
well as posting it I felt this was the appropriate forum for the response.

> > But then too, in retrospect, I suppose you could do nothing but agree
> > with everything I say in this reply, so I suppose there's really
> > nothing more for you to add.   ;o)
> >
> > Seriously, I'm interested to hear your further views on the matter.
> >
No, you aren't. Do you want the $5 argument or the $10 argument?

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/





From martin at v.loewis.de  Mon Apr 15 17:29:47 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 23:29:47 +0200
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <m3wuv8iras.fsf@mira.informatik.hu-berlin.de>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:

> Is it possible to use XPath in Python, if so are there any small samples? 

Yes, there is an XPath library in 4Suite, which is been partially
ported to PyXML 0.7.

Regards,
Martin


From tim.one at comcast.net  Fri Apr 26 18:14:13 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 18:14:13 -0400
Subject: String handling bug in Python
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEFKPBAA.tim.one@comcast.net>

> ...
> What do you do in your scheme if you want a z-string to contain a
> single double-quote?
>

[Bjorn Pettersen]
> I always thought that could be spelled r'"' or r'''"''',

Sure, but note the question is asking about the OP's z-string notation,
which supposedly has no difficulties.  You're talking about the r-string
notation.

> but I guess there would still be a problem if you wanted all of ", ',
> ''', and """ in a string.

There is with r-strings, yes.

> I would think that would occur less frequently than \ at the end of a
> string though (especially with windows path names).  I feel I must be
> missing something?

The good news is that I also feel that you are <wink>.  r-strings were
primarily intended to ease writing regular expressions (which can never end
with an odd number of backslashes), and people using \ in Windows pathnames
are usually missing an obvious alternative (like the forward slash --
Windows API functions couldn't care less which kind of slash you use; and
portable code should avoid any literal slashes via os.path.join()).





From jon+usenet at unequivocal.co.uk  Thu Apr  4 08:11:15 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Thu, 04 Apr 2002 13:11:15 -0000
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk>
Message-ID: <slrnaaok7j.4ei.jon+usenet@snowy.squish.net>

In article <3CAC1665.8040209 at mxm.dk>, Max M wrote:
>> relative_url("http://foo/a/b/c", "http://foo/a/d")
>>  --> return: "../d"
> 
> I hope you mean that the result should be "../../d" ??

No, he means "../d".

http://foo/a/b/c + d       == http://foo/a/b/d
http://foo/a/b/c + ../d    == http://foo/a/d
http://foo/a/b/c + ../../d == http://foo/d



From martin at v.loewis.de  Sun Apr 14 05:06:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 11:06:43 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122>
Message-ID: <m3662uy7h8.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> >bool(a) ^ bool(b)
> >
> Hm, ^ should probably be != to return bools instead of integers?

It turns out that ^ returns bool if both arguments are bool.

Regards,
Martin


From max at alcyone.com  Fri Apr  5 16:53:12 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 05 Apr 2002 13:53:12 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <mailman.1017757086.9920.python-list@python.org> <KBor8.8537$q63.3756@atlpnn01.usenetserver.com>
Message-ID: <3CAE1CC8.7D18F4A9@alcyone.com>

Steve Holden wrote:

> *My* point, however badly I originally made it, was that writing
> 
>     if ret == None:
> 
> is *sometimes* like writing
> 
>     if (a == 3) == True:
> 
> assuming this new wart appears in Python.

Comparing to None (preferably via is rather than ==, since == can be
overridden to do strange things) is perfectly legitimate when an API
declares that a function or method returns "an object or None."  After
all, the __nonzero__ method can be overridden to make perfectly valid
objects evaluate to false in a truth test.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From fperez528 at yahoo.com  Mon Apr 15 15:10:21 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 13:10:21 -0600
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCMEPOONAA.tim.one@comcast.net> <mailman.1018763429.14619.python-list@python.org>
Message-ID: <a9f8km$lef$1@peabody.colorado.edu>

Mike C. Fletcher wrote:

> Rather than confuse Tim further with abstract samples of where I want to
> define a "NULL" default argument, here's a particular example of where
> the "NULL pattern" I've described is used:
> 

[snip]

This is one of the few perlisms I still miss after having transitioned to 
python. In perl you have the 'if defined($x)' idiom for checking whether a 
variable is even defined (deeper than truth testing). This allows checking 
for variable names which may be undefined without triggering an error. I'd 
love to have a similar option in python:

def foo(x,y=Undef,z=10):
  if y is not Undef:
        ...

The use would be that an explicit
x=Undef
assignment is equivalent to 'del x', because you make a variable undefined via 
it. But in function prototypes, such a definition simply means 'this argument 
is optional, can be given either positionally or by name, but it has no 
default value because I want to explicitly check whether it was passed or 
not'.

I know this can be implemented with special patterns, so maybe my rambling 
here isn't too justified and 'pythonic' :). Need more coffee today, it's 
Monday.

cheers,

f.


From philh at comuno.freeserve.co.uk  Wed Apr 10 12:33:41 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 10 Apr 2002 17:33:41 +0100
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at>
Message-ID: <slrnab8qb5.ec2.philh@comuno.freeserve.co.uk>

On Wed, 10 Apr 2002 09:59:58 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
>phil hunt wrote:
>
>> I disagree. K&R has a mixture of clarity and concision that makes
>> it a delight to read.
>> 
>> IMO, anyone for whom K&R isn't to their taste, probably won't make a
>> very good programmer anyway.
>
>The last thing which I am  bothered with: whether to become a good
>programmer or not. For example, I did never use a struct in C; that
>means my C code is more or less a big spaghetti one. I do not care as
>long as I get the correct result.

But writing bad code means more bugs, so you are less likely to get 
the correct result. And as your badly-written program gets bigger, 
it gets more and more bug-ridden and hard to understand.

>And I have really, really not the intention to waste my life as a
>professional programmer

That's a fair enough attitude if all you are going to do is write 
the odd 200-line script.

If you are writing larger programs, it rapidly becomes 
self-defeating.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From tim.one at comcast.net  Wed Apr 10 20:01:00 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 20:01:00 -0400
Subject: What has PEP 285 done to us?
In-Reply-To: <000101c1e067$76d35820$0334fea9@carol>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEAMONAA.tim.one@comcast.net>

[Arthur Siegel]
> While we're on the subject of Aahz...

Could you two just get married and fight in private <wink>?

> It was and is a piece of a advice that  - in the
> interests of better group dynamics - a message be
> sent that Aahz never has and never will speak for
> the core development team.

Nobody speaks for Guido.  Aahz speaks for himself.  I speak for me, and
somtimes for Guido, but only when it happens that what I said turns out to
be what he would have said had he spoken.  This doesn't apply to Aahz, or to
anyone else, even if they channel him 100% correctly (which, while it's
better than I do, doesn't count, because I channel that Guido will say so).

> A peek in at dev indicates things moving in the
> direct opposite direction.

You're a sweetheart in person, Art.  Focusing on personalities just doesn't
work in this medium.

> Some-people-don't-know-who-their-friends-arely-yours,

yet-i'm-told-that-some-do-ly y'rs  - tim





From just at xs4all.nl  Wed Apr 17 13:23:43 2002
From: just at xs4all.nl (Just van Rossum)
Date: Wed, 17 Apr 2002 19:23:43 +0200
Subject: Making a C extension compatible with 2.2
References: <915a998f.0204170914.66aee851@posting.google.com>
Message-ID: <just-1CC2DE.19234217042002@news1.xs4all.nl>

In article <915a998f.0204170914.66aee851 at posting.google.com>,
 hamish_lawson at yahoo.co.uk (Hamish Lawson) wrote:

> Must anything be done to make a C extension compatible with Python
> 2.2? I have an extension (not my own) that ran fine under Python 2.1
> on Solaris 2.6. I have since recompiled it for Python 2.2, but
> sometimes I now get tuple indexing errors - and not necessarily at a
> point in my Python code where the loaded extension is actually being
> called upon.

Does this exception make sense at all in that place? If not, that's a 
fairly sure sign the extension is setting an error somewhere yet doesn't 
return NULL. The exception will the pop up a little bit later at a 
seemingly random place.

> Running the same code under Python 2.1 is OK. Would I be
> right in thinking it may be related to the changes in Python 2.2 for
> the new iterator protocol?

Without seeing more details that seems an awfully long shot...

Just


From roy at panix.com  Tue Apr  9 08:14:45 2002
From: roy at panix.com (Roy Smith)
Date: Tue, 09 Apr 2002 08:14:45 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1018339899.2981.python-list@python.org>
Message-ID: <roy-7658EF.08144409042002@news1.panix.com>

 tanzer at swing.co.at (Christian Tanzer) wrote:
> There are several dimensions to typing, including:
> - static vs. dynamic
> - strong vs. weak

Putting it another way...

In Perl, if I write "$x = 1;", x can act as either an integer or a string, 
depending on how I use it.

In Python, if I write "x = 1", x can only act as an integer.  In that 
sense, it's strongly typed.  On the other hand, it's also true that 
sometimes later, I can say "x = '1'", and now x can only act as a string.  
At any particular point in time, its type is strongly defined, but the type 
can also (dynamically) change from time to time.

In Python, the variable (x) can change type, but the value itself (1) 
cannot.  In Perl, the value itself can change type as coerced by the 
context.


From bvdpoel at uniserve.com  Sat Apr 13 20:50:12 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Sat, 13 Apr 2002 17:50:12 -0700
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com> <a9a7ls$gu4$0@216.39.172.122>
Message-ID: <3CB8D244.5BCCDCF4@uniserve.com>

Bengt Richter wrote:
> 
> On Sat, 13 Apr 2002 12:19:35 -0700, bvdpoel at uniserve.com wrote:
> 
> >
> >Yeah, I thought this would be easy...
> >
> >I have a list like this:
> >
> >       [ [0, x, [0, y, [4, n] ...]
> >
> Is this what it appears to be? I.e., a linked list of [offset,event,link]
> rooted in a [] ? Thus:

Not quite. I missed a ']' and it should be [ [0,x], [0,y], [4,n]...].
Each list
element is a pair.

>         [ + ]
>           |
>           +[0, x, +]
>                   |
>                   +[0, y, +]
>                           |
>                           +[4, n *?*]
> 
> (*?* what goes here, or are you using len(element)<3 to indicate eol?)
> 
> If it is a linked list, do you need that for fast insertions on the fly, or what?
> If not, why not [ [0,x], [0,y], [4,n] ...] or [ (0,x), (0,y), (4,n) ...] ?
> 
> The sorting is going to be different, depending ;-)
> BTW, is there no cheap way to maintain desired order as you build/modify the list,
> so you won't need a sort per se?
>

I changed to a dict with the key the offset. So now I have things like:

	dict = [ 0:[x,y], 4:[n] }

Of course this is all built on the fly. Nice thing about the dict method
is that
I don't have to worry about the sort problem, and internally it is
probably more
efficient.

Thanks.


-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel


From tdelaney at avaya.com  Tue Apr 30 22:53:46 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 1 May 2002 12:53:46 +1000 
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2BC@natasha.auslabs.avaya.com>

> From: huaiyu at gauss.almadan.ibm.com 
> 
> John Roth <johnroth at ameritech.net> wrote:
> >I strongly disagree about apply(). The only use I've found for it
> >varies the function which it calls, not the parameter lists! Having
> >variable parameter lists doesn't hack it - I need the variable
> >function.
> 
> >>> from math import *
> >>> a = (2.3,)
> >>> for f in [sin, cos, sqrt]: print apply(f, a)
> ... 
> 0.745705212177
> -0.66627602128
> 1.51657508881
> >>> for f in [sin, cos, sqrt]: print f(*a)
> ... 
> 0.745705212177
> -0.66627602128
> 1.51657508881

There can be real advantages to apply being a first-class object though,
which you don't get with the syntax f(*a, **b). apply can be passed around
and used just as any callable.

There is no way I would deprecate apply - esp. if you want code to work on
1.5.2, 2.0, 2.1, 2.2, 2.3 without change (which is still possible for most
code). I appreciate the f(*a, **b) syntax, but I think of it as an
alternative to apply, not a replacement.

Tim Delaney




From garrett at bgb-consulting.com  Wed Apr 10 18:21:10 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Wed, 10 Apr 2002 22:21:10 GMT
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com>
Message-ID: <3CB4BB04.1B8DD2A6@bgb-consulting.com>


"M. J. Mix" wrote:
> 
> Is anyone using Python as a spacecraft ground system scripting
> language, or onboard (embedded) language?
> 
> Pros and cons when compared with TCL or JAVAScript.  Just
> some real life, honest opinions.

  I'm not aware of ANY scripting languages that have been developed in a manor
suitable for life critical fail-safe situations. I like Python a lot, and find
it very reliable for day to day stuff, but I'd hate to have it embedded in my
pace maker. I also don't think it's suitable for serious real-time processing.

  Fail-safe and Real-time processing are both rather specialized and arcane
arts. They are often associated with things like embedded systems in critical
applications. Embedded systems in spacecraft often fall into these categories
(just like pacemakers). Standard Python is just not suitable for these tasks,
though specialized versions could be developed (given the types of
cost/expense usually associated with such systems). Neither are any TCL or
Javascript engines that I am aware of.

  Maybe you could be more specific about the types of tasks you are trying to
perform?  

> The usual issues: reliability and maintenance effort in applications.
> Also, use as a scripting language by non-programmers. (I saw the post
> on indenting, so I'm concerned about that.)

  For typical applications it's an excellent all around tool. Reliable, easy
to use, and easy to learn.

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 17:24:22 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 23:24:22 +0200
Subject: preg_replace?
References: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE> <odlv8.24535$b62.683481@news1.tin.it>
Message-ID: <a9kp6q$qch$1@nets3.rz.RWTH-Aachen.DE>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:odlv8.24535$b62.683481 at news1.tin.it...
>..
> I'm not sure what's wrong with your code, but this works for me.
>..

Thanks, and sorry. I wrote
        pattern.sub('0', s)
without ever reassigning to "s".
Actually I didn't focus on anything but the regular expression, which is why
I didn't catch it. Now that I'm told the regex is correct another look makes
it seem obvious.




From wesc at deirdre.org  Tue Apr  2 02:23:16 2002
From: wesc at deirdre.org (Wesley J. Chun)
Date: 1 Apr 2002 23:23:16 -0800
Subject: is there gonna be a Core Python Programming 2e ?
References: <codq9uko6ueumep852po13m6pbtfs3r2ts@4ax.com> <a7jkm6$cnt$1@panix1.panix.com>
Message-ID: <fcd9edc1.0204012323.4a88c6fe@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<a7jkm6$cnt$1 at panix1.panix.com>...
> In article <codq9uko6ueumep852po13m6pbtfs3r2ts at 4ax.com>,
> John Warney  <john at yahoo.com> wrote:
> >
> >I was wondering if there was plans on CPP 2e ? (no its not c++, core
> >python programming, lol) If so i would ask for a better binding than
> >what is on CPP.  I would even pay more than what the current CPP is at
> >now.
> 
> I know that Wes is gearing up to work on the second edition; I'll pass
> this on to him (he doesn't read c.l.py regularly these days).  However,
> I'll have to note that this is more up to the publisher than the author.

john,

as aahz has pointed out, i've been toying with the ideas for a
CPP 2e.  i apologize for the binding, however, it is partially
my fault -- the same goes for the page count.  :-)  i gave the
impression to Prentice Hall that i was not going to deliver
very much material, and this may have led to the paper thick-
ness, the binding, the page count, etc.

the goal for 2e would be to add the many missing chapters, i.e.,
SQL/databases, XML, Win32 Programming, etc., and chop the book
into a pair, just like Core Java, which has a "core" for Vol. 1
and an advanced topics for Vol. 2.  in fact, sometime in the
not too distant future, i will be posting to CLP asking the
community for their feedback on how to improve 1e and features
they'd like to see in 2e.

i will not be able to devote the resources to it for a little
while longer, perhaps 2.4 timeframe.  in the meantime, i will
definitely suggest a more user-friendly binding for any future
releases.

thanks for your interest!

-wesley

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://deirdre.org/baypiggies

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/


From martinb at otenet.gr  Wed Apr 17 11:38:32 2002
From: martinb at otenet.gr (Martin Barnard)
Date: Wed, 17 Apr 2002 18:38:32 +0300
Subject: Pure python crypto
In-Reply-To: <20020417021703.H16954@eecs.tufts.edu>
References: <20020417021703.H16954@eecs.tufts.edu>
Message-ID: <02041718383201.00171@toad>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wednesday 17 April 2002 09:17, you wrote:
>   Would anyone have any use for a pure python implementation of
> Blowfish? Or any other crypto algorithms for that matter.  As a matter
> of interest, I found myself implementing the blowfish module recently
> but was unsure if any audience would be interested. No C compiler
> is required, which is nice. Could use some testing though.
>
>                -- Mike
I'd be interested in it.
Where can I download it from?


Martin
- -- 
You look like a million dollars.  All green and wrinkled.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iEYEARECAAYFAjy9lvgACgkQpgoq4z9IM6hwEgCgiOGAbHrPrc4VO1R+2m5n/kHy
tu4AnAmeYMVHw9XKWNQaUJ9E27sHKmvr
=cF7Q
-----END PGP SIGNATURE-----




From matt at mondoinfo.com  Sun Apr 14 15:50:02 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sun, 14 Apr 2002 19:50:02 GMT
Subject: playing with Tkinter, expert needed ?
References: <13285ea2.0204140737.55dc49f1@posting.google.com>
Message-ID: <K3lu8.30956$vm6.4776221@ruti.visi.com>

On 14 Apr 2002 08:37:42 -0700, J.Jacob <joost_jacob at hotmail.com> wrote:

Dear Joost,

>I have been playing with Tkinter to make a widget that is portable and
>usable from Python version 1.5.2 on.  The widget is for emulation of
>very old-fashioned character based computer screens.

Your project sounds very cool. I don't need it right now but I've
bookmarked it because I'm sure that it will come in handy.

>a) How can I have a 'fat' cursor blinking instead of the small one
>when waiting for input?

There might be a better way to do it but here's what I'd try: first,
hide the vertical bar cursor by setting the text widget's insertwidth
to 0. Then you'll need to create the character-cell cursor. You could
do that by creating a tag that's at the same spot as the automatic
mark "insert" and play with its background and foreground colors. You
could use Tkinter's after() call to change them at appropriate
intervals.

I think the annoying part will be moving the cursor appropriately.
You'd probably want to bind to the <Key> (or perhaps <KeyRelese>)
event.  If Tkinter will let you see "insert" after it has been
updated, you could just move the tag. Otherwise you may have to insert
the characters yourself or predict where Tkinter will move the
insertion point based on the character that the user has typed.  The
after() function may be useful here too.

If you haven't yet run into it, Fredrik Lundh's excellent An
Introduction to Tkinter ought to make doing things like that as
painless as possible. Which is not to say entirely painless. It's at:

http://www.pythonware.com/library/tkinter/introduction/index.htm

Regards,
Matt


From dkuhlman at rexx.com  Sat Apr 27 18:10:54 2002
From: dkuhlman at rexx.com (kuhlman)
Date: Sat, 27 Apr 2002 15:10:54 -0700
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <aaf7kl$a8p54$1@ID-139865.news.dfncis.de>

Take a look at pyeditor.  It's built on Scintilla and Gtk.  You 
will need pyscintilla and PyGtk in order to build it.  You can find 
pyeditor at http://www.rexx.com/~dkuhlman.  (Note: Do not use 
WxEditor at that same location.  It doesn't work so good.  It's an 
interesting experiment that does not seem to me to be good enough 
to be usable.)

Also, I have not yet figured it out yet, but ... cooledit may be 
worth looking at.  It is not written in Python, but is extensible 
in Python, if that is what you actually want.  There is a link to 
cooledit at http://www.python.org/editors/ .

  - Dave

Philipp Lenssen wrote:

> Is there a simple, open source text-editor project available in
> Python?

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From magnus.heino at sigma.pleon.se  Thu Apr  4 13:57:36 2002
From: magnus.heino at sigma.pleon.se (Magnus Heino)
Date: Thu, 04 Apr 2002 20:57:36 +0200
Subject: minidom.parseString and =?ISO-8859-1?Q?=E5=E4=F6?=
Message-ID: <Bm1r8.1138$bB4.399@news2.bredband.com>

Hi.

What is the best way to solve this?

[magnus at daysleeper magnus]$ python2.2
Python 2.2 (#1, Mar  1 2002, 14:43:56)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-104)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> mystring='<foo>???</foo>'
>>> import xml.dom.minidom
>>> xml.dom.minidom.parseString(mystring)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 965, 
in parseString
    return _doparse(pulldom.parseString, args, kwargs)
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 952, 
in _doparse
    toktype, rootNode = events.getEvent()
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/pulldom.py", line 256, 
in getEvent
    self.parser.feed(buf)
  File "/usr/lib/python2.2/site-packages/_xmlplus/sax/expatreader.py", line 
143, in feed
    self._parser.Parse(data, isFinal)
UnicodeError: UTF-8 decoding error: invalid data
>>>


--

  /Magnus


From jepler at unpythonic.net  Mon Apr 29 11:00:39 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 29 Apr 2002 10:00:39 -0500
Subject: Tkinter Text widget reading input (how to do overwrite mode?)
In-Reply-To: <13285ea2.0204290624.4fbe8cba@posting.google.com>
References: <13285ea2.0204290624.4fbe8cba@posting.google.com>
Message-ID: <20020429100034.A12526@unpythonic.net>

I have implemented an 'overwrite mode' in the Text widget in tcl.
Conversion to Python is left as an exercise to the reader.
Make your Text widget an OverwriteText by calling 'MakeOverwriteText'
on it, which inserts OverwriteText in the list of bindtags.
Tested (minimally) on tk8.4 CVS, not on earlier versions.

proc TextOverwrite {w s} {
    set len [string length $s]
    $w delete insert insert+${len}c
}

# The overwrite flag is stored as a tag, if it's present
# overwrite mode is active
proc TextIsOverwrite {w} {
    expr {[lsearch [$w tag names] overwrite_flag] != -1}
}

proc TextSetOverwrite {w} {
    $w tag add overwrite_flag 0.0 end
}

proc TextSetInsert {w} {
    $w tag delete overwrite_flag
}

proc TextToggleOverwrite {w} {
    if {[TextIsOverwrite $w]} {
	TextSetInsert $w
    } else {
	TextSetOverwrite $w
    }
}

bind OverwriteText <KeyPress> {
    if {[TextIsOverwrite %W]} { TextOverwrite %W %A }
    # fall through to Text's <KeyPress> binding
}
bind OverwriteText <Insert> {
    TextToggleOverwrite %W
    break; # apparently <Insert> pastes selection, don't fall through
}

proc MakeOverwriteText { w } {
    set bt [bindtags $w]
    set t [lsearch $w Text]
    bindtags $w [linsert $bt $t OverwriteText]
}

#text .t
#pack .t
#MakeOverwriteText .t




From ianb at colorstudy.com  Tue Apr 16 17:51:58 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 16 Apr 2002 16:51:58 -0500
Subject: Safe eval?
In-Reply-To: <a9i41d$dqo$1@nets3.rz.RWTH-Aachen.DE>
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>
	<7xhemb744m.fsf@ruckus.brouhaha.com>  <a9i41d$dqo$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <1018993918.16807.380.camel@localhost>

On Tue, 2002-04-16 at 16:10, Philipp Lenssen wrote:
> Now. In the ASP/ VBScript version I keep all session data on the server side
> and simply submit the session ID forth and back. The untrusted user on the
> client side can alter the session ID, but not the actual session data.
> However in the PHP version, and how I'm currently laying it out in the
> Python port, the session data is put into hidden form variables. These can
> be altered before send back by anybody knowing how to edit the XHTML (and
> change the HTTP header if I'd check the referrer.) I still like this concept
> currently because it allows me to not be dependent on HTTP header cookies
> when I use the language internal session methods.

What you should do is sign your hidden variables.  That is, have a
secret server-side key, and hash the key concatenated with the
expression (using the md5 library, for instance).  When they resubmit
the data you can confirm that the data originated from your server by
comparing the hash they passed you with the hash you reconstruct.  

  Ian






From shagshag13 at yahoo.fr  Tue Apr 23 05:04:18 2002
From: shagshag13 at yahoo.fr (Shagshag)
Date: 23 Apr 2002 02:04:18 -0700
Subject: Python, search engine and NLP...
Message-ID: <409a56e2.0204230104.16d60f52@posting.google.com>

Hello, 

I'm interested in Pyhon but not sure it can fit my needs :

I have a huge search engine to write with many XML input/output, it
must be able to index and process something like 3 GO of data, many
text files, parse text and other natural language processing, not be
too slow, and so on...

- Do you think Python can help me ???

- As by now i don't know Python, do you think that it will not be a
waste of time even if I had to learn it ???

- Do you know about some Python's natural language processing module
???

- No, as there are many Java NLP library, I should definitively do it
with Java...

Thanks in advance,

S13.


From spivey at ieee.org  Wed Apr 17 23:38:50 2002
From: spivey at ieee.org (Gary Spivey)
Date: Thu, 18 Apr 2002 03:38:50 GMT
Subject: Hostname Resolution not working in Explorer
Message-ID: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net>

I have windows XP Home, and IE6. I have a zope server running on my box.
The server is running fine, and I can access if it I access

http://127.0.0.1:8080/mange

However, if I use

http://localhost/mange

the page isn't found. The same holds true if I use my hostname rather than
localhost - also, the actual IP address works.

I have made sure that the name entries exist in the
window/system32/drivers/etc/HOSTS file, and I can ping both localhost and my
hostname from a command window (with the correct address resolution).

So - why can I not get the resolution to work from within Internet Explorer?

Cheers
Gary




From imbosol at vt.edu  Sat Apr 20 18:05:49 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sat, 20 Apr 2002 18:05:49 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com>
Message-ID: <tnos9a.de.ln@127.0.0.1>

Grant Edwards wrote:
> In article <Xns91F67C40EB0E0lumberjack at 216.148.53.83>, Lumberjack wrote:
>> aahz at pythoncraft.com (Aahz) wrote:
>>> What's amazing about Lisp is that it's still in many ways on the cutting
>>> edge of computer science.
>> 
>> Perhaps you should name these "many ways". Taken at face value, your 
>> statement pretty much claims that computer science hasn't progressed in 40 
>> some years. I can at least agree with that statement. Of course the 
>> "science" in "computer science" is pure bovine excrement. If you can't or 
>> wont measure it, it isn't science. "Computer religion" or "computer cults" 
>> would better capture the reality.
> 
> It's part math, part crafstmanship.  I agree it isn't science.
> As a general rule, no field with "science" in its name is.

Computer programming is not science.

The study of algorithms, their properties, etc., is science.
Implementing a quicksort algorithm is not science, but studying its
properties is.

As long as we restrict computer science to the study of computer
algorithms and such, then computer science is science.  If we use
computer science as a sexy synonym for computer programming, then it
isn't.


Don't even get me started on whether is computer engineering is
engineering.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From aleax at aleax.it  Tue Apr  9 13:02:18 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 17:02:18 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no> <usn64ame1.fsf@python.net> <3CB31335.2040501@mxm.dk>
Message-ID: <u8Fs8.462$b62.9923@news1.tin.it>

Max M wrote:

> Michael Hudson wrote:
>> Syver Enstad <syver-en+usenet at online.no> writes:
> 
>>>Am I the only one in the entire world that thinks '\t'.join(mylist) is
>>>elegant and natural?
>> 
>> No.
> 
> Well I would have found mylist.join('\t') more natural, but I can live
> with the other.

joiner.join(sequence) is better: sequence is accessed through the
iterator protocol (now formally, once informally) so you get the
ability to play well with join for ANY kind of sequence of strings
as long as you support iteration (no need to do any work just for
join!) -- AND you can override join specifically for peculiar
joining needs, e.g. (rough -- needs fully indexable nonempty
sequence -- bit more work to refine for any iterable:-):

class mixedJoiner:
    def __init__(self, primary, last):
        self.primary = primary
        self.last = last
    def join(self, sequence):
        return self.primary.join(sequence[:-1]
            ) + self.last + sequence[-1]
    
print mixedJoiner(', ', ' and ').join('wine women song'.split())


joiner.join(sequence) was an EXCELLENT design choice...


Alex



From peter at engcorp.com  Wed Apr 10 00:33:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 00:33:55 -0400
Subject: Help with Python advocacy and marketing
References: <mailman.1018400660.13179.python-list@python.org>
Message-ID: <3CB3C0B3.E01CC333@engcorp.com>

"Stephan R.A. Deibel" wrote:
> 
> * We are collecting user-side advocacy quotes for use in press
>   releases announcing future versions of Python.  If you have
>   a compelling statement about how Python has helped your
>   project or business, this is a great way to help Python and
>   get some media exposure in the process.  We can't guarantee
>   your quote will be used in a press release, but with your
>   permission it will be included on a Python advocacy website
>   that is under development.

Some companies might have difficulty allowing their name to 
be used by "just anyone" in the context of Python advocacy.
Can you describe more about the way in which this information
will be distributed, who is represented (is this a Zope Corp
or PythonLabs sponsored endeavour?), and so on?  My company
(not engcorp.com) has so far remained in the closet about 
its use of Python and although it would be nice to get 
permission to help with a little "XXXX company uses Python
in the following ways..." I know I couldn't begin to get
permission without more background.

> Please direct discussion of this topic to the marketing-python
> mailing list, which can be joined with the URL given above.

Not wishing to subscribe to another list at this time, I'm
hoping you'll respond here instead.

-Peter


From donnal at donnal.net  Fri Apr  5 13:22:51 2002
From: donnal at donnal.net (Donnal Walter)
Date: 5 Apr 2002 10:22:51 -0800
Subject: location of a module's directory
References: <a8k1cb$s9v$1@news6.svr.pol.co.uk>
Message-ID: <918bc22f.0204051022.7a895d4d@posting.google.com>

"Dave Berkeley" <dave at rotwang.freeserve.co.uk> wrote in message news:<a8k1cb$s9v$1 at news6.svr.pol.co.uk>...
> Is there a way to discover where a module was loaded from?
> 
> I want to include a text configuration file with a .dll, and place them in
> the same directory. If the module knows where it was loaded from it will be
> able to find the config file.
> 
> If you do (Python 2.1.1)
> 
> >import sys
> >print sys.modules
> 
> you can get a dictionary of loaded modules, and a description, which
> includes the path the module was loaded from. However, I can find no way of
> accessing this information programmatically.

Will something like this work?

import os
def Main():
    try:
        myPath = os.path.dirname(__file__)
        print myPath
    except:
        pass


Donnal Walter
Arkansas Children's Hospital


From jriveramerla at yahoo.com  Sun Apr  7 20:40:56 2002
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 7 Apr 2002 17:40:56 -0700
Subject: Compatibilty table between BOA, wxPython, Python?
References: <11e94203.0204062147.5d2281b9@posting.google.com> <3CB0245A.3010808@csi.com>
Message-ID: <11e94203.0204071640.103c4bb7@posting.google.com>

The problem is that I made a program with BOA 0.0.5, it worked ok...
and now reports the next error, so I guess that the problem may be the
compatibility with wxPython 2.3.2.1

Traceback (most recent call last):
  File "wxapp1.py", line 24, in ?
    main()
  File "wxapp1.py", line 20, in main
    application = BoaApp(0)
  File "C:\TOOLS\Python22\lib\site-packages\wxPython\wx.py", line
1587, in __ini
t__
    _wxStart(self.OnInit)
  File "wxapp1.py", line 11, in OnInit
    self.main = wxFrame1.create(None)
  File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 11, in
create
    return wxFrame1(parent)
  File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 161, in
__init__
    self._init_ctrls(parent)
  File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 115, in
_init_ctrls
    self.statusBar1 = wxStatusBar(size = wxSize(685, 20), id =
wxID_WXFRAME1STAT
USBAR1, parent = self, name = 'statusBar1', style = 0, pos =
wxPoint(0, 347))
  File "C:\TOOLS\Python22\lib\site-packages\wxPython\stattool.py",
line 55, in _
_init__
    self.this = apply(stattoolc.new_wxStatusBar,_args,_kwargs)
TypeError: 'pos' is an invalid keyword argument for this function
6:33:39 PM: Debug: c:\Projects\wx\src\msw\app.cpp(535):
'UnregisterClass(canvas)
' failed with error 0x00000584 (la clase todav?a tiene ventanas
abiertas.).
6:33:39 PM: There were memory leaks.
6:33:39 PM: ----- Memory dump -----
6:33:39 PM: wxFrame at $8C8690, size 332
6:33:39 PM: wxMenu at $172E768, size 112
6:33:39 PM: wxMenuItem at $1768970, size 124
6:33:39 PM: wxMenuItem at $1768A20, size 124
6:33:39 PM: wxMenuItem at $1768E10, size 124
6:33:39 PM: wxMenuItem at $1768EE0, size 124
6:33:39 PM: wxMenuItem at $1768FB0, size 124
6:33:39 PM: wxMenuItem at $1769080, size 124
6:33:39 PM: wxMenuItem at $1769150, size 124
6:33:39 PM: wxPyCallback at $1768180, size 12
6:33:39 PM: wxObject at $1717EB0, size 28
6:33:39 PM: wxPyCallback at $1769260, size 12
6:33:39 PM: wxPyCallback at $1769310, size 12
6:33:39 PM: wxMenu at $17693C0, size 112
6:33:39 PM: wxMenuItem at $1769488, size 124
6:33:39 PM: wxMenuItem at $1769558, size 124
6:33:39 PM: wxMenuItem at $17696C0, size 124
6:33:39 PM: wxMenuItem at $1769790, size 124
6:33:39 PM: wxPyCallback at $1769860, size 12
6:33:39 PM: wxPyCallback at $17698E0, size 12
6:33:39 PM: wxMenuBar at $8DA6F0, size 312
6:33:39 PM: wxMenu at $1769AE0, size 112
6:33:39 PM: wxMenuItem at $1769BA8, size 124
6:33:39 PM: wxMenuItem at $1769C78, size 124
6:33:39 PM: wxMenuItem at $1769EA8, size 124
6:33:39 PM: wxMenuItem at $1769FF0, size 124
6:33:39 PM: wxMenuItem at $176A148, size 124
6:33:39 PM: wxMenuItem at $176A290, size 124
6:33:39 PM: wxMenuItem at $176A420, size 124
6:33:39 PM: wxMenuItem at $176A550, size 124
6:33:39 PM: wxPyCallback at $176A670, size 12
6:33:39 PM: wxPyCallback at $176A6F0, size 12
6:33:39 PM: wxPyCallback at $176A7A0, size 12
6:33:39 PM: wxPyCallback at $176A850, size 12
6:33:39 PM: wxPyCallback at $176A900, size 12
6:33:39 PM: wxPyCallback at $176A9B0, size 12
6:33:39 PM:
6:33:39 PM:
6:33:39 PM: ----- Memory statistics -----
6:33:39 PM: 1 objects of class wxMenuBar, total size 312
6:33:39 PM: 1 objects of class wxObject, total size 28
6:33:39 PM: 11 objects of class wxPyCallback, total size 132
6:33:39 PM: 19 objects of class wxMenuItem, total size 2356
6:33:39 PM: 3 objects of class wxMenu, total size 336
6:33:39 PM: 1 objects of class wxFrame, total size 332
6:33:39 PM:
6:33:39 PM: Number of object items: 36
6:33:39 PM: Number of non-object items: 0
6:33:39 PM: Total allocated size: 3496
6:33:39 PM:
6:33:39 PM:


From peter at engcorp.com  Fri Apr  5 22:39:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 05 Apr 2002 22:39:36 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python 
 going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84>
Message-ID: <3CAE6DF8.9E655008@engcorp.com>

James Logajan wrote:
> 
[snip history of most Python versions since 1.0]
> 
> > Seriously though, if this is a big concern for you (and it might well
> > be), target the 1.5.2 language spec and the chances of your code not
> > working even in 2.2 are slim, so I'm not sure what would be compelling
> > enough to make you abandon Python.
> 
> I tried that and that strategy _failed_. The chances of failure was 100%.
> The failure was due specifically to how (as far as I can tell) Tkinter
> handles UTF-8 sequences in Unicode aware versions of Python. 

Let me see if I'm clear on this.  You are complaining about the entire
history of Python releases because of a problem you had with _Unicode_?
When was Unicode added to Python?  

Did you really have such serious compatibility problems in other areas
that you were forced to abandon the language?  Or was it just personal
distaste at the pace of change?  Don't condemn the fast release cycle
just because you got screwed on one incompatibility.

I suspect part of the fast pace is because of the desire to get rapid
feedback to make sure the project is "on track" with its customers.
Traditionally change came across multiple years, but in today's world
many are trying to do more iterations less far apart so that they
can benefit from rapid feedback rather than having a project die
a slow death from lack of releases (cf Netscape).

-Peter


From eppstein at ics.uci.edu  Sat Apr  6 13:51:19 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 06 Apr 2002 10:51:19 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <a8km9i$7t6$1@panix1.panix.com> <slrnaaucmq.3j0.philh@comuno.freeserve.co.uk>
Message-ID: <eppstein-B782F3.10511906042002@news.service.uci.edu>

In article <slrnaaucmq.3j0.philh at comuno.freeserve.co.uk>,
 philh at comuno.freeserve.co.uk (phil hunt) wrote:

> What ambiguities? There is nothing ambiguous about current behavour 
> of truth values. Furthermore, there is nothing that, in practise, 
> hampers programmers: a programmer can easily learn to use 1 and 0 
> for true and false.
> 
> The only advantage for the new scheme, as I see it, is it looks 
> prettier and is a bit easier for newbies. Ditto for integer 
> division.

I think the integer division case is quite different.
With bool, there is little change to actual code behavior, it is more 
for the benefit of human readability.

The integer division thing, I think, detracts from readability and may 
be harder for newbies, but is necessary to preserve Python's philosophy 
of polymorphism: functions should not require their inputs to be of a 
specific type, they should behave "the same" as long as they are given 
inputs that obey the protocols that the function is expecting. E.g. if 
you are iterating over something it usually doesn't matter whether it is 
a string, tuple, list, or simple generator.  In order to behave 
appropriately when dividing by a number that could be 2L or could be 
2.0, the function needs to more clearly specify whether integer or 
"exact" (meaning approximate floating point unless/until we get 
rationals) division was intended.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From llchen223 at hotmail.com  Tue Apr  9 16:48:53 2002
From: llchen223 at hotmail.com (Lei Chen)
Date: 9 Apr 2002 13:48:53 -0700
Subject: Simple question! urgent
References: <mailman.1018353258.20452.python-list@python.org>
Message-ID: <de6c123c.0204091248.5da0fbe3@posting.google.com>

"Vincent Fiack" <vfiack at webmails.com> wrote in message news:<mailman.1018353258.20452.python-list at python.org>...
> > x = [1,2,3,4]
> >  
> >  How to get NUMBER OF ITEMS in array? I need it for checkin nr of
>  items
> >  in argv[]
> 
> len() is your friend.

Or you can try:

  reduce(lambda x, y: x + y, map(lambda x: 1, x))

Just Kidding :)  len() works fine.


From sholden at holdenweb.com  Mon Apr  8 18:18:21 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 18:18:21 -0400
Subject: changing local dictionnary for functions
References: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr>
Message-ID: <vIos8.3287$vT6.670@atlpnn01.usenetserver.com>

"Boris Boutillier" <boris at cantal.lip6.fr> wrote in message
news:pan.2002.04.08.18.15.31.115122.19669 at cantal.lip6.fr...
> Is there a way to pass a local dictionnary to a function as it is
> possible to do with a file with the execfile statement ?
>
> In the python reference manual, in the section 4.1 : Code
> Blocks,execution frames and namespace, there is a table that say that
> function body uses globals namespace of the containing block and a new
> locals namespace. And there is a note that say global and local can be
> overrident, how is it possible ? I think it would be the answer to my
> first question.

I suspect this is an arror in the documentation, since the namespaces
provided to a function call are (unless I am mistaken) determined solely by
the context of the call. Sorry.

If  I'm wrong, you can guarantee that three different people will point this
out within three minutes of my hitting the "send" key :-)

regards
 Steve






From borcis at geneva-link.ch  Wed Apr 10 10:22:01 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Wed, 10 Apr 2002 16:22:01 +0200
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <3CB49ADD.89B242A4@sympatico.ca>
Message-ID: <3CB44A89.F74EA8F8@geneva-link.ch>

Colin J. Williams wrote:
> 
[...]
> > Note this would introduce three-valued logic into Python, which would be
> > helluva confusing to beginners, who tend to understand the law of the
> > excluded middle implicitly.
> 
> I wonder about the confusion.  It would perhaps add clarity to an aspect of
> Python.
> 
> Let's look at some examples:

[snip over the instance traces]

> Logical operations
> 0 or  None, 1 or  None: None 1
> 0 and None, 1 and None: 0 None
>               not None: 1
> 
> With a three value scheme, the results would
> be the same for the first two, whereas the
> result of `not None` would itself be None.
> 
> I would appreciate comments.

Beautiful, and so what ? :) And there's indeed my own
use of None over Laura.

I'd say indeed that we should value Laura's intervention
not exactly for what she appears to have meant to mean,
but as an true alarm bell... I mean that if I take to the
psychic gesture to believe her, to make it plausible I have to
evoke an older idea of mine.

Tadah, OIOM :

[b] Python is unlike [a] GatesDesigns(tm) because :

Grammar is valued like a hidden beauty,
holding promises of evolution, not yet
to be fathomed, and this in the service
(choice where they differ) [a] of growing
GatesDesign(tm)'s mass of consumership
[b] of not wasting grammar pockets-of-untapped
-dynamics to silly-purposes-according-to
-better-common-interest.

IOW GvR is serious where BG isn't and LW too clever.

OK, now my current theory of Laura is that she's there
like an Except: clause that fired over the thread of PEPS
because at PEP 285 exactly Sharon invaded Palestine while
"the border of the inner square" of "aesthetic manna" Laura
"was guarding" underwent a "boundaries violation" because of
"quasifascist boolean pseudowisdom taking over python family
jewels in eg falsety gathering."

Should Python grow restartable interrupts to fully
solve the case ?

Should PEPS become mythified ? This needs order. Someone
proposed "chinese torture", I prefer (while I did not look)
to positivise the work, saying "in the ideal" the PEP process
should look like "a linear run of mutually orthogonal bytecodes
of what we'd need to call a 'crazy diagonal VM'".

Sorry, I see I'll have to write a couple PEPS, with no chance
to be taken seriously at having them see pass.

Mmmhhh. This gives me an idea. If we could efficiently market
Python as "God's very own choice of programming language" to
credulous masses, then a market geometry for adaptation of
Python patterns, metapatterns and minipatterns would spring
to existence :)

And this in turn would make it become natural for Python
to grow to the niche of being the only programming language
endowed with a so-called "boolean style sheets" with 
standard styles of the distribution being Style152 or
Style22.285 or StyleCyclotomicF or StyleIconIdol or
StyleDonJim :)

"Boolean style sheets" can certainly be sold as "refactoring
from __future__ statements"

Maybe something as simple as

"from __future__ StyleDonJim import booleans"

Mmmhh. This seems to lead us to somewhere. Three digits releases
tend to be bugfix releases closing their generation. Naked code
runs on them with at least an implied

from __future__ Style3DVersion import defaults

Where 3DVersion stands for the 3 digits of the version :)

> 
> Colin W.
> 

Cheers, Boris "30215" B.
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")


From andy at reportlab.com  Thu Apr 18 03:18:04 2002
From: andy at reportlab.com (Andy Robinson)
Date: 18 Apr 2002 00:18:04 -0700
Subject: pyRXP Memory Likage???
References: <c5f48de3.0204161113.14919d25@posting.google.com>
Message-ID: <956003ae.0204172318.5644e26c@posting.google.com>

chema at imdea.com (Jose Manuel Esteve) wrote in message news:<c5f48de3.0204161113.14919d25 at posting.google.com>...
> I have been playing with pyRXP and is incredibly fast as compared with
> minidom, also the memory footprint is smaller for one file, however if
> you loop for parsing a lot of xml strings the memory usage gets higher
> and higher as compared to minidom
> 
> import pyRXP
> xml='<start><tag1>text here</tag1></start>'
> P=pyRXP.Parser()
> for i in range(0,100000):
>     tup=P.parse(xml)
> 
> will show the behaviour

I confirm this with the latest source.  We are investigating
and wil post back here.  Thanks!

- Andy Robinson
CEO/Chief Architect,
ReportLab Inc.


From kseehof at neuralintegrator.com  Wed Apr 10 08:45:03 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Wed, 10 Apr 2002 05:45:03 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <Xns91EBD90A7449Dlumberjack@216.148.53.83>
Message-ID: <EFEFIHAINIINNELLHFAIOEHPCAAA.kseehof@neuralintegrator.com>

Lumberjack flamed:
> Tim Peters <tim.one at comcast.net> wrote:
> > PEP 285 taught us that people don't agree, and that some people
> > disagree in disagreeable ways.  Frankly, that wasn't news,
>
> Amazing to see so much slow learning taking place. So why don't
> you all do
> it again some time. Like the next PEP and the next one and the next one,
> and so on ad infinitum. Just like previous PEPs. The old Chinese water
> torture would be more humane than continuing these unending so-called
> "enhancements" to the language. Drip drip drip... Neither you nor Guido
> have a clue; unless this is his secret plan to destroy the appeal
> of Python
> by setting Python programmers at each other's throats, or they begin to
> associate emotional discord with the language and decide to keep their
> distance. If so, the plan is well executed.
>
> People aren't going to change, so Guido needs to adjust to the reality of
> the political consequences of his current strategy.

Thank you for your excellent self-referential example of disagreeable
disagreement.

The Chinese water torture image does have the virtue of being humorous at
least.  Pep... Pep... ... PEP... pep pep... Pep... Pep... Pep...

I don't subscribe to the belief that PEPs are causing people to behave
abusively.  I think that that phenomenon is the responsibility of those
who engage in such behaviour.  It is quite possible to express ones
opinions without resorting to personal attacks.  Also it is sometimes
even helpful to acknowledge your opponents point of view, since nearly
every argument has at least some merit.

It is probably true that more change creates more opportunity for debate,
and therefore more opportunity for discord, even if the actual changes are
beneficial.  IMO, this is not in itself sufficient reason to avoid change.

I understand the desire to make all of the following changes:

1. The overall rate of changes to python needs improvement.
1a) Don't make any changes to python ever.
1b) Decrease the rate of change.
1c) Increase the rate of change.
1d) Hurry up and ship 3.0

2. The rate of new releases needs improvement.
2a) Spread out changes over more releases
2b) Combine changes into fewer releases

3. The version numbering system needs improvement.
3a) There are a variety of possible changes to the present system
    although I can't think of any at the moment :-)

All of these changes have their respective advantages and disadvantages.
They all have large numbers of people in favor and against them.
Although I find myself frequently disagreeing on specific PEPs, I like
the process as a whole and would not be in favor of any of these changes.

Interestingly, I have reversed my opinion on each of the implemented PEP's
that I had previously disliked and argued against.

I am glad we are not still using python 1.1 because on the whole, python
2.2 really is better.  Although I see the potential for so called "chinese
water torture" to cause people problems, I for one have found the whole
process (since version 1.5.2 for me) extremely smooth, especially when
compared to Microsoft Visual C++.  I attribute this to the fact that Guido
has taken great care to make sure that changes have minimum impact on
backward compatibility.

It is not logical to argue for significantly slower changes to python
without taking the position that the value of improvements since version
1.5.2 integrated over the population of python users since 1.5.2 and into
the forseeable future is less than the pain caused by dealing with changes
to python in the present time.  It would be nice if python could keep
getting better without changing. :-)

Note that I have made my opinions known without resorting to petty personal
attacks on Lumberjack.  I have even refrained from using words like
"childish" or "petty" to describe his personal attack on "clueless" Guido
and the entire python community of "slow learners".

<sarc> Those of you who have suddenly developed such negative opinions about
Guido should consider finding a different language.  There must be several
languages out there written by people of significantly higher moral
character.
</sarc>

I, on the other hand would like to thank Guido for creating such an
excellent
programming language.  This is not hero worship, it is simple politeness and
gratefulness.

it's-impossible-make-the-world-a-better-place-without-pissing-some-people-of
fly yrs,

- Ken Seehof





From tonyl at homechoice.co.uk  Sat Apr  6 14:04:21 2002
From: tonyl at homechoice.co.uk (Tony)
Date: Sat, 06 Apr 2002 20:04:21 +0100
Subject: beginners syntax question
Message-ID: <3caf47a5@news1.homechoice.co.uk>

Hi, 

I'm trying to call the update() method of several objects ie:

object1.update()
object2.update() 
etc

I'd really like to do 

for x in range(7):
  "object"+str(x).update

but can't figure outthe syntax, is this possible?

Thanks 
Tony





From duncan at NOSPAMrcp.co.uk  Wed Apr 17 04:33:44 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 17 Apr 2002 08:33:44 +0000 (UTC)
Subject: pre-PEP for optional 'pass'
References: <mailman.1018972676.30534.python-list@python.org>
Message-ID: <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1>

<brueckd at tbye.com> wrote in
news:mailman.1018972676.30534.python-list at python.org: 

>> You should always head each method and class with a docstring.
> 
> No! You should head each method and class with a docstring *only if
> the docstrings would add value*. Otherwise you're doing more work now
> AND later (during refactoring) without producing any real benefit,
> cluttering the code in the process.

I kind of agree with you while standing by my original statement. In the 
context of the original message, where the poster was proposing writing a 
function that only contained 'pass', they could always use a good docstring 
instead. Even if the docstring simply says that the function is 
unimplemented this provides useful information to anyone who does not have 
the original source in front of them.


> 
>> class MyClass:
>>    '''A simple class'''
>>    def __init__(self, x, ,y, z): '''MyClass constructor'''
>>    def getTwiddle(self): '''Get a twiddle'''
>>    def setTwiddle(self, twiddle): '''Set a twiddle'''
> 
> The above code snipped contains exactly 4 examples of docstrings that
> should not exist - all of them either contain no useful information or
> no information that isn't already painfully obvious.

Yes, but since I have no idea what any of these methods actually does, I 
couldn't be bothered to guess at a useful docstring. Actually that isn't 
true: I know exactly what these methods do (zilch), but I have no idea why 
anyone might have written them.

> A good rule of thumb is that if the docstring (or comment) could have
> been auto-generated by your editor, you shouldn't include it.

Or you should edit it so it contains some useful content.
Here is another go at the class above

class MyClass:
   '''A completely useless class with no obvious purpose in life'''
   def __init__(self, x, ,y, z):
      '''Parameters x, y, and z are all dropped in the bit bucket.
         Perhaps we are passed them for compatibility with something
         else in which case we could say what...'''
   def getTwiddle(self):
    	 '''Despite the implications of its name, this method actually
         returns None. It also has no side effects and should therefore
         be refactored out.'''
   def setTwiddle(self, twiddle):
      '''Whatever a twiddle may be, this method actually drops the
         argument and does nothing useful. Returns None. Another
         refactoring candidate.'''

I would say that all of these comments could be automatically generated, 
but if I look at the output of (say) pydoc, they add useful information as 
I might otherwise be misled into thinking that the class and methods did 
something.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From h_schneider at marketmix.com  Tue Apr  9 04:03:04 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Tue, 9 Apr 2002 10:03:04 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com> <3CB274D2.A7554C79@engcorp.com>
Message-ID: <a8u6tj$quj$02$1@news.t-online.com>

Thats the fact! If the community gives back the experience, people like me
would stop thinking about using
Delphi or VB ... we soon would have an excellent and up-to-date
documentation.

What about posting them to
http://wiki.wxpython.org/index.cgi/wxPython_20Cookbook :)

Bye,
Harald


"Peter Hansen" <peter at engcorp.com> schrieb im Newsbeitrag
news:3CB274D2.A7554C79 at engcorp.com...
> Chris Spencer wrote:
> >
> > 4.  Did I mention that the documentation is both incomplete and odious?
I
> > should not have to constantly do a "find in files" to find out what
parameters a
> > damned method call takes because the way wxPython implemented a method
is
> > different than the wxWindows way and someone forgot to make a note of it
in the
> > rubbish passed off as a "manual".
>
> Although I share your pain, I don't complain much about it since I'm
> well aware of the effort required to put together something like this
> and maintain it, and I don't believe the authors would appreciate
> my condemnation as much as they would my assistance.
>
> Maybe you could contribute some of what you learned back to the project
> and offer to help improve the documentation?
>
> -Peter




From wortelslaai at yahoo.com  Mon Apr 22 20:14:10 2002
From: wortelslaai at yahoo.com (Jaco Schieke)
Date: 22 Apr 2002 17:14:10 -0700
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu>
Message-ID: <b5ddd682.0204221614.25009f09@posting.google.com>

> 
> Nice! I didn't think to look there as I assumed this part was happening in 
> the C part of readline. Do you care to submit a bug report with this one-line 
> patch to SourceForge? I'd really like to have rlcompleter in 2.3 have this 
> fixed (it already has a patch I sent in earlier for another problem) That way 
> I could stop shipping a separate patched version with my own project.
> 
> f.

Did a bug report at
https://sourceforge.net/tracker/index.php?func=detail&aid=547176&group_id=5470&atid=105470

but if you think a patch is better, then you can submit the patch and
I'll retract the bug report.  Send me an email at
schieke at cae.trachea.wisc.edu [remove the organ 4 spamprotection].

Jaco


From boris at cantal.lip6.fr  Fri Apr 12 04:18:03 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Fri, 12 Apr 2002 10:18:03 +0200
Subject: frameobject.c bug ? No call to PyFrame_LocalsToFast.
References: <mailman.1018591127.18785.python-list@python.org>
Message-ID: <pan.2002.04.12.10.18.02.982610.14362@cantal.lip6.fr>

This is not completely true, as we can generate some code with
CO_NEWLOCALS removed, with new.code for example.

I'll give here an example of what can be done with this initialisation
(that by the way just need to be done if CO_NEWLOCALS = 0 so no overhead
for common case, see and of post for the diff).

I've written a GenFunction class that is used as follow :

def f(toto,tata,*args,**kwds):
	print 'args :',toto,tata,args,kwds
	print 'value :',value
	value = 5
	print 'gbl :',gbl

f = GenFunction(f)
g = {'gbl':3}
l = {'value':4}
gbl = 12

f('toto','tata','a1','a2','a3',k='k1',l='k2',globals=g,locals=l)

>>args : toto tata ('a1', 'a2', 'a3') {'k': 'k1', 'l': 'k2'}
>>value : 4
>>gbl : 3

In fact I've also written a GenDict class, inherited from dict to which
you give a creation function to use when searched key is not in dict,
with the previous GenFunction class this allow to use globals and locals
that create automatic variables.

So I don't understand why not a call to PyFrame_LocalsToFast when Locals
is not null or empty, the overhead is just for the special case (the test
is already done in the code.

I give here the diff between the python2.2.1c2 version of frameobject.c
and my version :

diff Objects/frameobject.c ../Python-2.2.1c2/Objects/frameobject.c 
277,282d276
<       
<       while (--extras >= 0)
<               f->f_localsplus[extras] = NULL;
<       
<       f->f_nlocals = code->co_nlocals;
<       
293d286
<               f->f_locals = locals;
299,300d291
<               f->f_locals = locals;
<               PyFrame_LocalsToFast(f,1);
302c293
<       
---
>       f->f_locals = locals;
307d297
< 
311a302
>       f->f_nlocals = code->co_nlocals;
315a307,308
>       while (--extras >= 0)
>               f->f_localsplus[extras] = NULL;



Boris


From bokr at oz.net  Sun Apr 14 02:58:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 06:58:17 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org> <7xr8lic2yb.fsf@ruckus.brouhaha.com>
Message-ID: <a9b9a9$iam$0@216.39.172.122>

On 13 Apr 2002 21:35:24 -0700, Paul Rubin <phr-n2002a at nightsong.com> wrote:

>Tim Peters <tim.one at comcast.net> writes:
>> It's in the nature of open source development that volunteers scratch their
>> own itches, then share what they've done because it does them good too to
>> have their work widely adopted.  If you can ride that curve, you can get an
>> enormous amount of high-quality work for free.  If not, you'll have to pay
>> for it or go without.
>
>A lot of stuff gets done because it needs to get done, rather than
>scratch an itch.  
>
What's a need but a very insistent itch? ;-)

>> A crushing amount of debate over "stability" has gone by on Python-Dev the
>> last week, and the one thing that strikes me over and over is that, with
>> just a few minor exceptions, nobody volunteers to *do* anything except tell
>> other people what to do.  What are you willing to do to make your desires a
>> reality?  If it doesn't involve contributing work, time or money, you just
>> want a free ride.  But nobody owes anyone a free ride, and open source can't
>> change that fact of life either.
>
>In that particular debate, a lot of people were asking for certain new
>features to be LEFT OUT of Python.  I don't understand how someone can
>volunteer to contribute work, time, or money toward leaving something out.

Maintaining older baseline releases, where the features are already left out?
I don't think that work is zero, so should the BPWIs[1] be doing that, and who should say?

The trick is what happens when various versions are adopted out in the real world
without asking those whose scripts are expected to continue to run.
Can one distribute a script with first line #!/usr/bin/python21 and expect the
link to be there wherever python21 (or fully compatible other version) is available?

Otherwise new versions will potentially make extra work for me, which I'd rather
volunteer in other ways ;-)

Apparently 1.5.2 and 2.1.x and 2.2.x are semi-official baseline versions. Is it/could it
be official policy that #!/usr/bin/python(15|21|22) links should exist if a site
has the identified version? If /usr/bin/python is going to be redirected every six months
forever, I think I'd rather indicate a specific version (and it wouldn't make me unhappy to).

OTOH, if someone is motivated to consolidate all so they can vector all to the same latest
interpreter on their site, there's nothing to prevent it. It's their time and responsibility
fixing breakage and validating. If they share some validated upgrades, no doubt there will be
appreciation, but (ideally) no one's dragged into any work they didn't plan on.

[1] Benevolent Programmers While Interested ;-)

Regards,
Bengt Richter


From tim.one at comcast.net  Thu Apr 11 00:09:23 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 11 Apr 2002 00:09:23 -0400
Subject: Easter Eggs
In-Reply-To: <on6t8.206591$Yv2.67543@rwcrnsc54>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEBMONAA.tim.one@comcast.net>

[Tim]
> We take t-shirts very seriously at PythonLabs.
>
> they're-all-backward-compatible-too-ly y'rs  - tim

[Emile van Sebille]
> I don't know about that... mine seems too big, backward compatibly
> speaking...

No, no, no, Emile -- you're just a user.  They're all backward compatible
with my body.  That's what really counts.

you're-welcome-ly y'rs  - tim





From news.ASkwar at DigitalProjects.com  Tue Apr 23 18:39:14 2002
From: news.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Wed, 24 Apr 2002 00:39:14 +0200
Subject: Passing arguments to threading objects
Message-ID: <pan.2002.04.23.22.39.14.487844.11225@teich.garten.DigitalProjects.com>

Hello.

I'm currently learning Python 2.2 (on Linux) and just started tinkering
around with threads.  I've got the following code:

import threading
import random
import time

def mache(name, max):
	for i in range(max):
		print `name` + ': ' + `i`
		time.sleep(random.random())

class myThread(threading.Thread):
	def run(self, name, max):
		for i in range(max):
			print `name` + ': ' + `i`

for t_num in range(10):
#	t = threading.Thread(target = mache, args = (t_num, 10))
	t = myThread(args = (t_num, 10))
	t.start()

As you can see, I created a subclass "myThread" of threading.Thread.
There, I'd like to overwrite the run method and pass two parameters into
it.  This obviously does not work.

However, when I use the base class as in the commented out line and set a
target which also requires two arguments, I get what I'd expect to get.

When I subclass threading.Thread, how can I access the args parameter
passed to the constructor?

Thanks,

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 16 hours 48 minutes


From tim.one at comcast.net  Tue Apr 16 21:07:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 21:07:49 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <200204161511.g3GFBIl31849@mbuna.arbhome.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCEENJOOAA.tim.one@comcast.net>

[Tim]
> Guido is very reluctant to break code, but it's not an absolute for him:

[Anthony Baxter]
> Note that this is a "in new major releases" statement. Breaking code in
> a point release isn't something that happens. I'm not aware of any code
> changes in 2.1.2 or 2.1.3 that broke someone's 2.1.1 code. If someone
> can point me to these changes that caused problems, I'd love to see them.

All true, and all true by intent, and we've seen no report of any breakage
yet in a micro release (although we eventually will).

> If instead this whole discussion is just usenet noise, with no real
> point or purpose, then ignore the previous paragraph. Needless to say,
> I wait with baited breath for the collection of shocking
> incompatibilities to roll in.

Most of the discussion hasn't been about micro releases.

> (is it a fact of usenet that over time, all newsgroups become like
> comp.lang.perl.misc?)

All of life is like comp.lang.perl.misc, if you live long enough <wink>.





From tim.one at comcast.net  Fri Apr 26 21:17:04 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 21:17:04 -0400
Subject: Killing a process under Windows NT/2K
In-Reply-To: <f67cb2b7.0204261646.20f9595@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGCPBAA.tim.one@comcast.net>

[Tavis Rudd, on
]       win32process.TerminateProcess(handle, 1)

> Is this a hard kill (like kill -9)

Yes.  Quoth the docs:

    The TerminateProcess function is used to unconditionally cause a
    process to exit.  Use it only in extreme circumstances.  The state
    of global data maintained by dynamic-link libraries (DLLs) may be
    compromised if TerminateProcess is used rather than ExitProcess.

    TerminateProcess causes all threads within a process to terminate,
    and causes a process to exit, but DLLs attached to the process are
    not notified that the process is terminating.

> or a soft kill that triggers the appropriate signal handler
> inside the process?

This is Windows -- Unixish signals aren't popular there.  So, sure, it
triggers all the signal handlers a typical Windows program has <wink>.

Windows doesn't offer a safe-in-all-cases way for forced external
termination.  See, e.g.,

     http://www.wd-mag.com/articles/1999/9907/9907c/9907c.htm

for a discussion of some of the issues.  The heart of the method used there
is to create a new thread inside the process to be killed, and have it call
the orderly ExitProcess() from within the process to be killed.  However,
the CreateRemoteThread() API needed to do this didn't exist before Win98.





From ksmith at truesan.com  Thu Apr 11 12:38:47 2002
From: ksmith at truesan.com (Kevin F. Smith)
Date: 11 Apr 2002 09:38:47 -0700
Subject: jython getpass equivalent?
Message-ID: <e40fa75d.0204110838.7f0594cd@posting.google.com>

Does anyone have any suggestions for requesting input in Jython
without having the text echoed to the screen?

I appreciate any ideas.

--kevin smith


From elian at inbrief.net  Mon Apr 22 11:49:47 2002
From: elian at inbrief.net (Edmund Lian)
Date: Mon, 22 Apr 2002 11:49:47 -0400
Subject: Overriding methods in classes that use property()
Message-ID: <OFC355F6DD.0C582BB7-ON85256BA3.0056B244@juscom.com>

On Mon, 22 Apr 2002, Denis Otkidach wrote:

>Sorry, getVal is not in current scope. The following should be
>right:
>             val = property(A.getVal, setVal)

Ah, this works. I'm not quite sure that this is intuitive either though.
One would think/expect that val = property(getVal, setVal) would to just
resolve to A.getVal using the inheritance hierarchy when a subclass does
not override getVal.

...Edmund.





From aahz at pythoncraft.com  Fri Apr  5 09:47:10 2002
From: aahz at pythoncraft.com (Aahz)
Date: 5 Apr 2002 09:47:10 -0500
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? [LONG!])
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com> <20020405014732.GB17563@isis.gerrietts.net> <mailman.1017996733.25069.python-list@python.org>
Message-ID: <a8kdde$12t$1@panix1.panix.com>

In article <mailman.1017996733.25069.python-list at python.org>,
Andy Gimblett  <gimbo at ftech.net> wrote:
>
>Really?  I've used Singleton a fair bit (probably too much in fact)
>for stuff like configuration information.  I'm curious as to how else
>I could/should do this.

Use a module.  Don't use a class unless you need a class.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From hfoffani at yahoo.com  Fri Apr 26 04:39:09 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Fri, 26 Apr 2002 10:39:09 +0200
Subject: AS/400
References: <1e140f34.0204250934.2fee1f26@posting.google.com>
Message-ID: <20020426043917.616$Uf@news.newsreader.com>

"Terje Johan Abrahamsen"
> ....
> The Excel part I have no trouble with. The AS/400 part is a little
> worse. Do you know some additions I have to install with Python? Where
> can I find some info about the functions that moves around on the
> AS/400 display and so forth. ...

Rumba has an API, no? And I bet that there are 3rd parties ActiveX for
it out there.
Your work will just(*) be glue both Excel/Rumba with Python/win32com.

Regards,
-Hernan

(*) :-)





From hgs at dmu.ac.uk  Fri Apr 19 08:41:16 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Fri, 19 Apr 2002 13:41:16 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <m31ydc2lxl.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204191323550.22138-100000@neelix>

On 19 Apr 2002, Martin v. Loewis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > test test_largefile crashed -- exceptions.IOError: [Errno 27] File too large
> >
> >     How much space does it want?  I don't know this one anyway.
>
> It creates a file with 2500000000L. Depending on the file system you
> are using, that may or may not be possible (it tries to create it with
> a hole in it, so it normally takes only little disk space).

OK, I won't worry about that for now.
>
> > test test_nis skipped -- No NIS master found for any map
> >
> >     I didn't expect that.  We use nisplus....
>
> You may want to run the test verbosely:
>
> ./python Lib/test/regrtest.py -v test_nis

I'll come back to this one...
>
> > test test_socket_ssl skipped -- Use of the `network' resource not enabled
> >
> >     What network resouce?
>
> You need to explicitly give regrtest.py permission to use certain
> resources; see 'regrtest.py -h' for details. In the case, the default
> 'make test' doesn't run the test to avoid triggering a dial-up
> connection.

Thanks, I'll come back to this one too.
>
> > neelix hgs 11 %> ./python
> > Python 2.2.1 (#2, Apr 18 2002, 18:23:04) [C] on sunos5
> > Type "help", "copyright", "credits" or "license" for more information.
> > >>> import tkine ter
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > ImportError: No module named tkinter
>
> That should have been either Tkinter, or _tkinter - not tkinter.

I remembered: I forgot to uncomment all the places where the Tk stuff
lives.



# *** Always uncomment this (leave the leading underscore in!):
_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
	-L/usr/local/lib \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
	-I/usr/local/include \
# *** Uncomment and edit to reflect where your X11 header files are:
#	-I/usr/X11R6/include \
# *** Or uncomment this for Solaris:
	-I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#	-DWITH_TIX -ltix8.1.8.2 \
# *** Uncomment and edit for BLT extension only:
#	-DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#     (See http://www.pythonware.com/products/pil/ for more info)
#	-DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
#	-DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
	-ltk8.3 -ltcl8.3 \
# *** Uncomment and edit to reflect where your X11 libraries are:
#	-L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
	-L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
#	-lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#	-lld \
# *** Always uncomment this; X11 libraries to link with:
	-lX11


Now the build gives:


Script started on Fri Apr 19 11:57:36 2002
neelix hgs 11 %> gmake
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H  -o Modules/python.o Modules/python.c
        [...]
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H  -DPYTHONPATH='":plat-sunos5:lib-tk"' \
	-DPREFIX='"/usr/local"' \
	-DEXEC_PREFIX='"/usr/local"' \
	-DVERSION='"2.2"' \
	-DVPATH='""' \
	-o Modules/getpath.o ./Modules/getpath.c
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H  -o Modules/main.o Modules/main.c
        [...]
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H   -DWITH_APPINIT -I/usr/local/include -I/usr/openwin/include -c ./Modules/_tkinter.c -o Modules/_tkinter.o
In file included from /usr/local/include/tk.h:83,
                 from ./Modules/_tkinter.c:49:
/usr/openwin/include/X11/Xlib.h:149: warning: function declaration isn't a prototype
        [more similar]
/usr/openwin/include/X11/Xlib.h:2063: warning: type defaults to `int' in declaration of `XSetTransientForHint'
        [more similar]
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H   -DWITH_APPINIT -I/usr/local/include -I/usr/openwin/include -c ./Modules/tkappinit.c -o Modules/tkappinit.o
In file included from /usr/local/include/tk.h:83,
                 from ./Modules/tkappinit.c:16:
/usr/openwin/include/X11/Xlib.h:149: warning: function declaration isn't a prototype
        [more similar]
/usr/openwin/include/X11/Xlib.h:2063: warning: type defaults to `int' in declaration of `XSetTransientForHint'
        [more similar]
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H   -c ./Modules/xxsubtype.c -o Modules/xxsubtype.o
if test -f buildno; then \
	expr `cat buildno` + 1 >buildno1; \
	mv -f buildno1 buildno; \
else echo 1 >buildno; fi
gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H  -DBUILD=`cat buildno` -o Modules/getbuildinfo.o ./Modules/getbuildinfo.c
rm -f libpython2.2.a
ar cr libpython2.2.a Modules/getbuildinfo.o
ar cr libpython2.2.a Parser/acceler.o Parser/grammar1.o Parser/listnode.o Parser/node.o Parser/parser.o Parser/parsetok.o Parser/tokenizer.o Parser/bitset.o Parser/metagrammar.o Parser/myreadline.o
ar cr libpython2.2.a Objects/abstract.o Objects/bufferobject.o Objects/cellobject.o Objects/classobject.o Objects/cobject.o Objects/complexobject.o Objects/descrobject.o Objects/fileobject.o Objects/floatobject.o Objects/frameobject.o Objects/funcobject.o Objects/intobject.o Objects/iterobject.o Objects/listobject.o Objects/longobject.o Objects/dictobject.o Objects/methodobject.o Objects/moduleobject.o Objects/object.o Objects/rangeobject.o Objects/sliceobject.o Objects/stringobject.o Objects/structseq.o Objects/tupleobject.o Objects/typeobject.o Objects/weakrefobject.o Objects/unicodeobject.o Objects/unicodectype.o
ar cr libpython2.2.a Python/bltinmodule.o Python/exceptions.o Python/ceval.o Python/compile.o Python/codecs.o Python/errors.o Python/frozen.o Python/frozenmain.o Python/future.o Python/getargs.o Python/getcompiler.o Python/getcopyright.o Python/getmtime.o Python/getplatform.o Python/getversion.o Python/graminit.o Python/import.o Python/importdl.o Python/marshal.o Python/modsupport.o Python/mystrtoul.o Python/mysnprintf.o Python/pyfpe.o Python/pystate.o Python/pythonrun.o Python/structmember.o Python/symtable.o Python/sysmodule.o Python/traceback.o Python/getopt.o Python/dynload_shlib.o  Python/thread.o Python/thread.o
ar cr libpython2.2.a Modules/config.o Modules/getpath.o Modules/main.o Modules/gcmodule.o
ar cr libpython2.2.a Modules/threadmodule.o  Modules/signalmodule.o  Modules/posixmodule.o  Modules/_sre.o  Modules/newmodule.o  Modules/symtablemodule.o  Modules/_tkinter.o Modules/tkappinit.o  Modules/xxsubtype.o
ranlib libpython2.2.a
gcc  -Xlinker --export-dynamic -o python \
		Modules/python.o \
		libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread -L/usr/local/lib -ltk8.3 -ltcl8.3 -L/usr/openwin/lib -lX11   -lm
case $MAKEFLAGS in \
*-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py build;; \
esac
ld.so.1: ./python: fatal: libtk8.3.so: open failed: No such file or directory
Killed
gmake: *** [sharedmods] Error 137
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Fri Apr 19 12:08:12 2002

But

find /usr/local/lib -name "libt[ck]*" -print

gives

/usr/local/lib/libtcl8.2.so
/usr/local/lib/libtclstub8.2.a
/usr/local/lib/libtk8.2.so
/usr/local/lib/libtkstub8.2.a
/usr/local/lib/libtk8.3.so
/usr/local/lib/libtcl8.3.so
/usr/local/lib/libtclstub8.3.a
/usr/local/lib/libtkstub8.3.a
>
> > gcc -v -shared build/temp.solaris-2.7-sun4u-2.2/_tkinter.o build/temp.solaris-2.7-sun4u-2.2/tkappinit.o -L/usr/openwin/lib -L/usr/local/lib -ltk8.3 -ltcl8.3 -lX11 -o build/lib.solaris-2.7-sun4u-2.2/_tkinter.so
        [...]
>
> It seems that you have not put grabbed enough commands: there
> definitely *was* a command to build both _tkinter.o and tkappinit.o.
> When setup.py decides that building the module failed, it removes not
> only the resulting .so, but also all involved object files.

I think there should be enough data there now.
>
> Regards,
> Martin
>
>
        Hugh





From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Apr 16 19:18:41 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Wed, 17 Apr 2002 01:18:41 +0200
Subject: [ANN] Pyro 2.6 available
Message-ID: <mailman.1019071684.5095.clpa-moderators@python.org>

You can get the latest Pyro version (2.6) at http://pyro.sourceforge.net

This is mostly a bug-fix release. 
Fixed mobile code (agents), documentation enhancements, 
fixed python-2.2 dependency that had crept in, 
Event Server clients will now correctly terminate,
various other bug fixes.


What is Pyro?
Pyro is an acronym for PYthon Remote Objects. It is a basic Distributed
Object Technology system written entirely in Python.

It is extremely easy to implement a distributed system with Pyro, because
all network communication code is abstracted and hidden from your
application. You just get a remote Python object and invoke methods
on the object on the other machine.

Pyro offers you a Name Server, an Event Service, mobile objects, remote
exceptions, dynamic proxies, remote attribute access, automatic
reconnection, a detailed manual, and many examples to get you started
right away.

Irmen de Jong





From dysfate at gmx.de  Fri Apr 19 03:33:59 2002
From: dysfate at gmx.de (Johannes Stiehler)
Date: 19 Apr 2002 09:33:59 +0200
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net>
Message-ID: <a9oh95$9g8$1@wsc10.lrz-muenchen.de>

If speed is very (very, very ;-) important and you don't mind a bit of 
special syntax, then you could perhaps use mxTextTools from the mx-Base-
Package (which also provides nice date/time functions and so on)

http://www.egenix.com/files/python/eGenix-mx-Extensions.html


Regards,
Johannes
-- 
Johannes Stiehler				 stiehler at cis.uni-muenchen.de
CIS der LMU M?nchen                              089/2180-9706
Systemverwaltung & Anwendungsbetreuung	


From pyth at devel.trillke.net  Mon Apr 29 22:16:58 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 04:16:58 +0200
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <20020429190205.1cd7d1bc.logiplexsoftware@earthlink.net>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <20020429190205.1cd7d1bc.logiplexsoftware@earthlink.net>
Message-ID: <20020430021658.GX16217@devel.trillke>

On Mon, Apr 29, 2002 at 07:02:05PM -0700, Cliff Wells wrote:
> On 29 Apr 2002 18:54:12 -0700
> Fortepianissimo wrote:
> 
> > This question is so fundamental that I strongly suspect it's already
> > answered - ig that's the case, please forgive this poor newbie, for he
> > has done some work searching over Python FAQ and Google.
> > 
> > Basically I want to have a subclass of list, like
> > 
> > class MyList (list):
> >     ...
> > 
> > then I want to
> > 
> > l=MyList()
> > ... do stuff to fill l
> > s=" ".join(l)
> 
> s = " ".join([str(i) for i in l])

or you can use:

" ".join(map(str, l))

#map applies the function 'str' to each element and returns the results
#in a list 

have fun,

    holger




From op73418 at mail.telepac.pt  Mon Apr 22 12:56:55 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 17:56:55 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <MqVw8.47908$vm6.7506303@ruti.visi.com> <mailman.1019489191.17192.python-list@python.org>
Message-ID: <gcf8cust4b9f3absbbpaq3g9a4m9j8kbj8@4ax.com>

On Mon, 22 Apr 2002 09:10:20 -0700 (PDT), <brueckd at tbye.com> wrote:

>On Mon, 22 Apr 2002, Grant Edwards wrote:
>
>> In article <mailman.1019483442.32389.python-list at python.org>, brueckd at tbye.com wrote:
>> >
>> >> > You should understand that this is a philosophical viewpoint. The
>> > [snip]
>> >> To put this another way, ask yourself where you stand on this
>> >> question: is mathematics discovered or invented?
>>
>> Very well put.
>>
>> > Is there a difference?
>>
>> Yes.  Science verifies it's theories by comparing them to the
>> physical world which they are attempting to describe.  It
>> doesn't matter how elegant, or internally consistent Science
>> is.  If it doesn't successfully describe the physical world,
>> it's wrong.
>>
>> Science has an external reference point. Mathematics does not.
>> Internal consistency is the only thing which mathematics can
>> attempt to verify.  Mathematics is not an attempt to describe
>> the physical world.
>
>First, please back up, I was asking if there's a difference between
>discovery and invention, which you don't mention anywhere in your
>response.
>
>Second, math *is* an attempt to describe the physical world - that's what
>makes it useful. Case in point: calculus, which was invented/discovered
>specifically to deal with and describe the motion of physical bodies. Had
>it not been useful in describing such motion, it would have been tossed
>out, or at least not so widely accepted. Math is useful because of its
>relevance to the world around us.

Math *is not* an attempt to describe the physical world. It may have
been so a long ago but not anymore. It is true that calculus was first
invented by the needs in physics (by Newton) but I don't think that
Cauchy, Riemann, Weierstrass and others were thinking about "reality"
(one of the few words that without "" means nothing) when making it
rigorous.

And while I cannot speak for every mathematician I can say for sure that
a large body of them strongly disagrees with "Math is useful because of
its relevance to the world around us."

>
>Finally, in the more general sense, formal mathematical proofs and whatnot
>might not require an external reference point, but the foundation upon
>which they are built certainly does. Indeed, the fact that we've gotten as
>far as something like abstract algebra is largely due to the fact that the
>underlying building blocks *are* verifiable and applicable to our
>experiences outside of math and many of the "advances" in mathematics have
>been due to intuition or hunches provoked by real world phenomena.

I strongly disagree with this and also the modern history of
mathematics. There is a very strong connection between physics and
mathematics but many mathematicians live happily without ever thinking
of physics, or, for that matter, any applications their work might have.

>
>-Dave
>
>

best,
Gon?alo Rodrigues


From v.wehren at home.nl  Thu Apr 25 16:11:57 2002
From: v.wehren at home.nl (v.wehren)
Date: Thu, 25 Apr 2002 22:11:57 +0200
Subject: does readlines stop at 255?
Message-ID: <MABy8.74596$58.4854121@zwoll1.home.nl>

It appears that readlines() stops at 255...when reading a line..

How can I work around this if my file has lines longer than 255 characters
(and making them shorter is not an option...)??

Cheers!

Vincent





From jeff at ccvcorp.com  Mon Apr 22 14:49:04 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 11:49:04 -0700
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <mailman.1019488539.12450.python-list@python.org>
Message-ID: <MPG.172dfae3908dc4049896be@news.nwlink.com>

In article <mailman.1019488539.12450.python-list at python.org>, 
mgilfix at eecs.tufts.edu says...
>   While I'm not commenting on whether this is really useful in python
> (because I get along fine without it), the python equiv would be:
> 
>   def both_ways (arg, dir):
> 
>     if dir:
>       arg1 = blah
>       arg2 = some_blah
>     else:
>       arg1 = something_else
>       arg2 = some_other_thing
> 
>     another_func (arg1, arg2)

Hm, well, I'd be likely to write this as:

    def both_ways(arg, dir):
        if dir:
            another_func(blah, some_blah)
        else:
            another_func(something_else, some_other_thing)

Yes, it could be argued that this isn't as good because I'm 
duplicating the reference to another_func.  But then, you're 
duplicating the references to arg1 and arg2.  And besides, 
practicality beats purity.  :)

I would also argue that my example is a *lot* clearer than the 
ternary operator version, despite being less concise:

    def both_ways(arg, dir):
        another_func( (dir ? blah : something_else), 
                      (dir ? some_blah : some_other_thing) )

Apparently, though, this is a matter of taste.  I'm just happy 
that Guido's taste seems to be in favor of if/else.  <wink>

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mlh at vier.idi.ntnu.no  Sun Apr 14 09:38:30 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 14 Apr 2002 13:38:30 +0000 (UTC)
Subject: Checking whether an async* dispatcher is "alive"
References: <slrnabggo4.o80.mlh@vier.idi.ntnu.no>  <Oj5u8.34280$%A.2510727@news1.west.cox.net>  <slrnabj0ru.rf.mlh@vier.idi.ntnu.no>
Message-ID: <slrnabj1im.2vn.mlh@vier.idi.ntnu.no>

In article <slrnabj0ru.rf.mlh at vier.idi.ntnu.no>, Magnus Lie Hetland wrote:
[snip]

After anohter glance at the asyncore API, it seems that simply
overriding handle_close to do some extra cleanup is just what I need.

Thanks for your help anyway :)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From ak at silmarill.org  Tue Apr  2 15:19:45 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 02 Apr 2002 20:19:45 GMT
Subject: style Q: global vars in CAPS?
Message-ID: <slrnaak4m7.u14.ak@ak.silmarill.org>

Hello,

What do you think about global variables spelled in caps? Is that what
most people do?


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From kevin at cazabon.com  Wed Apr  3 12:36:47 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 3 Apr 2002 09:36:47 -0800
Subject: Buffer Interface, can't get PyObject_AsReadBuffer to work?
References: <5a4226f0.0204022145.49010b2e@posting.google.com> <m3sn6dntiy.fsf@mira.informatik.hu-berlin.de>
Message-ID: <5a4226f0.0204030936.58788ef5@posting.google.com>

According to the Python documentation, 0 means error, 1 means
success...?

"""On success, returns 1, sets buffer to the memory location and
buffer_len to the buffer length. Returns 0 and sets a TypeError on
error"""

???

Kevin.

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3sn6dntiy.fsf at mira.informatik.hu-berlin.de>...
> kevin at cazabon.com (Kevin Cazabon) writes:
> 
> >   // Convert the pInBuffer from PyObject to a PyBuffer (readable),
> > THIS DOESN'T WORK (the Python object type that is passed in as
> > inBuffer is a 'B' array)
> >   if (PyObject_CheckReadBuffer(inBuffer)) {
> >     //this ALWAYS returns 0, meaning unsuccessful... what's wrong???
> >     ok = PyObject_AsReadBuffer(inBuffer, &pInBuffer, &p4size);
> 
> 0 means success, -1 means error. Use the source, Luke.
> 
> HTH,
> Martin


From eppstein at ics.uci.edu  Wed Apr 10 21:11:44 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 10 Apr 2002 18:11:44 -0700
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu> <eppstein-40A0D7.17104110042002@news.service.uci.edu> <3CB4E282.AAD468E4@engcorp.com>
Message-ID: <eppstein-5BEE17.18114410042002@news.service.uci.edu>

In article <3CB4E282.AAD468E4 at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> If the customer wanted to handle lists with odd numbers of items
> he would have specified the desired behaviour.  Or he will now. :-)

I don't know, silently dropping the last item seems like it could be 
nonobvious enough to cause trouble later.  You don't think it should at 
least have a sanity check, causing more obvious breakage if things 
aren't as expected?  Something like:

    i = 0
    while i < len(L):
        print L[i],L[i+1]
        i += 2

which will throw an exception rather than just ignoring the last item...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From nobody at nobody.com  Sat Apr  6 14:25:41 2002
From: nobody at nobody.com (John)
Date: Sat, 6 Apr 2002 13:25:41 -0600
Subject: Textarea
References: <OjHr8.36744$tg4.437615@vixen.cso.uiuc.edu>
Message-ID: <cVHr8.36748$tg4.437829@vixen.cso.uiuc.edu>

Figured it out. Thanks.

"John" <nobody at nobody.com> wrote in message
news:OjHr8.36744$tg4.437615 at vixen.cso.uiuc.edu...
>
> Can we read file to display inside <textarea> and allow users to modify
the
> content and overwrite, using Python CGI? Thanks!
>
>




From python at rcn.com  Sat Apr 27 00:08:59 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sat, 27 Apr 2002 00:08:59 -0400
Subject: Removing dupes from a list
References: <20020426203946.GV21205@teich.Garten.DigitalProjects.com> <XFMail.20020426135630.shalehperry@attbi.com> <mailman.1019856418.32215.python-list@python.org>
Message-ID: <aad8as$adq$1@bob.news.rcn.net>

"Alexander Skwar" <ASkwar at DigitalProjects.com>
> Thanks, but connections to activestate.com simply time out :(  However,
> a traceroute to aspn.activestate.com doesn't show any unusual long hops.
> But AS is reachable from other networks.  Strange.

So I'm not the only one having timeout problems with ActiveState's website.


Raymond Hettinger




From mikael.lexen at stop.spam.volvo.com  Fri Apr 19 05:05:06 2002
From: mikael.lexen at stop.spam.volvo.com (Mikael =?iso-8859-1?Q?Lex=E9n?=)
Date: Fri, 19 Apr 2002 11:05:06 +0200
Subject: chopping csv-files (newbie)
References: <mailman.1019131355.20461.python-list@python.org>
Message-ID: <3CBFDDC2.153C8B2F@stop.spam.volvo.com>


Simon Brunning wrote:

> > From: Mikael Lex?n [SMTP:mikael.lexen at stop.spam.volvo.com]
> > I get the following when I try to import your module
> >
> > PythonWin 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on
> > win32.
> > Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) -
> > see
> > 'Help/About PythonWin' for further copyright information.
> > >>> import csv
> > Traceback (most recent call last):
> >   File "<interactive input>", line 1, in ?
> > ImportError: DLL load failed with error code 193
> >
> > What's wrong? I'am on NT4.
>
> Did you download the pyd file as well as the package? If not, then that's
> the problem.
>

Just the pyd. I shouldn't need the package should I?. Just put the pyd in the
dll dir.

>
> If you did, did you download the right on for your version of Python?
>

Yes I did. I moved to a different machine where I have MVS 6.0 and I looked in
the pyd and it says to be a "MS-DOS Executable files". So I downloaded the
package and made a "python setup.pu instal" and it compiled and installed a new
pyd (cool) and that one works ok. This one is a "Windows executable dynamic link
library file". The two files differ in size

/Mikael




From johahn at delta.telenordia.se  Sun Apr 14 05:09:00 2002
From: johahn at delta.telenordia.se (Johan Hahn)
Date: Sun, 14 Apr 2002 11:09:00 +0200
Subject: C API Q I : calling Python methods from C
Message-ID: <a9bl5j$2aq$1@green.tninet.se>

Hi

I can't figure out how to call methods written in python from C using the
low level API.

I can call functions though. So for now I create the instance in python when
importing the module and provide a helper function to call the methods. It
looks like this:

/* b4 module */
import code
iii = code.InteractiveInterpreter()
def runsource(code):
    iii.runsource(code)


I manage to import a module, read the modules dictionary, get a class from
the dictionary and create an instance of that class, but I can't find the
right way to get a callable object out of that.
You are supposed to read the class' dictionary for methods, convert a method
to a function and call the function with the instance as the first
parameter, right?
So if anyone can post an example on how to do that it would be of enormous
help!


thanks beforehand,
Johan






From DSmithSprint05 at earthlink.net  Sat Apr 20 20:59:04 2002
From: DSmithSprint05 at earthlink.net (David R. Smith)
Date: Sun, 21 Apr 2002 00:59:04 GMT
Subject: Tkinter.Text.index() puzzlement
References: <3CC0B271.C76D3E95@nonesuch.com> <qB2w8.46153$vm6.6102555@ruti.visi.com>
Message-ID: <3CC20EA2.1385A0E5@earthlink.net>

Matthew, thanks for your response.

I'm trying to add a textual read-eval-print cycle onto a program which is
mainly GUI-driven.  I thought I'd set a marker at the end of the
computer-generated output, e.g.,

    marker = text.index(END)

and then when the <Return> event strikes, pick up the text between marker
and text.index(END) as of that time, and send it off to eval.  This keeps
resulting in null strings, even though I entered  "123+456<Return>".  Then
the string might belatedly come through.  Is it hopeless to use
text.index() to do useful work?

    David Smith




From SBrunning at trisystems.co.uk  Fri Apr 26 05:06:55 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 26 Apr 2002 10:06:55 +0100
Subject: AS/400
Message-ID: <31575A892FF6D1118F5800600846864DCBD08C@intrepid>

> From:	terjej at mailandnews.com [SMTP:terjej at mailandnews.com]
> I am trying to write some Python code that moves info from Excel into
> AS/400 throuhgt a display program called Rumba. (www.walldata.com)
> Basically what I want it to do (short version) is that I go to a cell
> in Excel, copy that cell, changes to the Rumba window, goes to
> coordinate x,x pastes the number there. Then goes back to Excel, gets
> a new number, goes to Rumba, changes screens by typing F12, enters
> info from Excel again in coordinate x,x. It is more to it than this,
> but if I can figure out this, I can figure out the rest.
> 
> The Excel part I have no trouble with. The AS/400 part is a little
> worse. Do you know some additions I have to install with Python? Where
> can I find some info about the functions that moves around on the
> AS/400 display and so forth. I have searched in google and on
> python.org but haven't found much. I can't be the only one that needs
> to do something like this? Thanks in advance for all answers and all
> pointing in the right direction so I can find out more info...
 
One possibility - I believe that Rumba is COM scriptable. If so, there
shouldn't be any problem driving it with win32all. (If Rumba *isn't* COM
scriptable, check out Reflection from WRQ, which is *definitely* COM
scriptable.)

But that's not how *I'd* do it. I'd bypass 5250 altogether. I'd write the
data direct to the database on the iSeries if at all possible. If you have
an ODBC driver for your iSeries, then you could write the data up using
mxODBC.

An alternative would be to use the Java based Jython rather than CPython.
This way you could make use of the *excelent* JTOpen Java package. This
package includes a JDBC driver, native database access, and classes to drive
just about anything that you can think of on the iSeries. Jython can't use
win32all, so you might have trouble with the Excel file, but if you saved
your data in CSV format, you could read that OK.

If you *can't* write direct to the database (perhaps you have to put your
data through a complex RPG program), you could make the calls via JTOpen.

Another thing to look at - Per Gummedal's port of Python to the iSeries at
<http://home.no.net/pgummeda/>.

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




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




From mgash at trisystems.co.uk  Wed Apr 10 10:50:20 2002
From: mgash at trisystems.co.uk (Mark Gash)
Date: Wed, 10 Apr 2002 15:50:20 +0100
Subject: how to read hardware info on Win NT
Message-ID: <31575A892FF6D1118F5800600846864DCC77B5@intrepid>

>Thanks for the __winreg hint. Looks like I have to get some infos
>about the registry because I have no idea how to find the right key.
>
>If anyone knows, please tell me.
>
>I need the id of my ethernet card for a license documentation tool we
>want to write in Python. As many commercial softwares are using this
>id for generating the license keys etc. I thought there should be
>module in Python.  Maybe there is a modul for commercial distribution
>of Python software?

Marcus,
With further investigation, I cannot find where the MAC address of the card
is in registry (I assume that you require the MAC address when you say that
you want the id of your Ethernet card).

I am also assuming that you are using a win32 machine.

There are two other ways that I can think of in order to get your info:-

1. Run the command 'ipconfig /all' and pipe the results into a text file
that you can then search i.e.
	'ipconfig /all > c:\ipdetails.txt'

2. The nbtstat command will report the MAC address of a card, so that for
example, you could use the command 'nbtstat -a <machinename>' and pipe the
results out to a text file, i.e.
	'nbtstat -a mypc > c:\nbtdetails.txt' 

Once the correct command has been tested, then from your python script wrap
it up in an os.system call.

So, we will end up with something along the lines of :-

import os
os.system("ipconfig /all > c:\ipdetails.txt")

or 

import os
os.system("nbtstat -a mypc > c:\nbtdetails.txt")

Then you will have to add some string handling/searching to find the MAC
address and process it as required.

Hope this helps.

Kind Regards,

Mark Gash
Consultant
TriSystems Ltd.
020 7264 0440
mgash at trisystems.co.uk





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




From robin at jessikat.fsnet.co.uk  Sat Apr 13 14:05:34 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 19:05:34 +0100
Subject: cStringIO write
References: <QHxVGIALCDu8EwLs@jessikat.fsnet.co.uk>
 <krYt8.13007$b62.354393@news1.tin.it>
Message-ID: <fTIWTKAuNHu8Ewq$@jessikat.fsnet.co.uk>

In article <krYt8.13007$b62.354393 at news1.tin.it>, Alex Martelli
<aleax at aleax.it> writes
>Robin Becker wrote:
>
>> Is it deliberate that 2.2 cStringIO only has a write method when created
>> uninitialised? StringIO doesn't have this 'feature', but no mention is
>> made of this in the docs.
>
>Very deliberate, I'd say -- they're different types:
>
>[alex at lancelot cb]$ python
>Python 2.2 (#1, Dec 23 2001, 20:09:01)
>[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> import cStringIO
>>>> readonly = cStringIO.StringIO('ciao')
>>>> writeonly = cStringIO.StringIO()
>>>> type(readonly)
><type 'cStringIO.StringI'>
>>>> type(writeonly)
><type 'cStringIO.StringO'>
>>>> print type(writeonly).__doc__
>Simple type for output to strings.
>>>> print type(readonly).__doc__
>Simple type for treating strings as input file streams
>>>>
>
>
>And this is indeed covered in the docs:
>"""
>The following data objects are provided as well:
>
>InputType
>The type object of the objects created by calling
>  StringIO with a string parameter.
>OutputType
>The type object of the objects returned by calling
>  StringIO with no parameters.
>"""
>
>Maybe you could post to sf, either a tiny patch to the docs clarifying this 
>(I think a "(read-only)" or something like that in the brief description of 
>InputType might be enough?), or a vastly richer and more complex patch if
>you think this isn't a documentation problem but rather than these types
>should be merged into a single, more complicated one.
>
>
>Alex
>
strange in my docs I see no mention of these special types. I see for
StringIO 

This module implements a file-like class, StringIO, that reads and
writes a string buffer


and indeed StringIO.StringIO('aaa') and StringIO() apear to be the same
type. I see that cStringIO does indeed behave as you say, but the module
docs don't say anything about the differences other than mention non-
subclassibility.

Indeed a writeable cStringIO.StringIO is made readable using seek so the
readonly/writeonly distinction is moot.

I need only instances of either so I think I can get away with

def getStringIO(s=None):
    try:
        from cStringIO import StringIO
    except ImportError:
        from StringIO import StringIO
    f = StringIO()
    if s is not None:
        f.write(s)
        f.seek(0)
    return f

but this makes the assumption that initialised StringIO's are to be read
from and not written again.
-- 
Robin Becker


From martin at v.loewis.de  Fri Apr  5 02:23:09 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Apr 2002 09:23:09 +0200
Subject: Selecting cipher suites with socket.ssl
References: <84cff98e.0204041830.7c0d0171@posting.google.com> <mailman.1017978613.9042.python-list@python.org>
Message-ID: <m3y9g21uea.fsf@mira.informatik.hu-berlin.de>

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

> It's not possible. For everything but the most simple client-side SSL, I
> recommend you use a different SSL implementation for Python. m2crypto
> seems to be quite mature, but there are others, too:

My recommendation would be slightly different: Please modify the
existing SSL support to better suite your needs, and contribute your
changes. I believe this will help Python users better in the long run.

Regards,
Martin



From just at xs4all.nl  Thu Apr 18 03:39:57 2002
From: just at xs4all.nl (Just van Rossum)
Date: Thu, 18 Apr 2002 09:39:57 +0200
Subject: unhandled exceptions without a stack-trace
References: <mailman.1019111684.8904.python-list@python.org>
Message-ID: <just-209240.09395718042002@news1.xs4all.nl>

In article <mailman.1019111684.8904.python-list at python.org>,
 Paul Rodger <paul at paulrodger.com> wrote:

> Is there any way I can:
> 
> 1) raise an exception ...
> 2) not catch that exception (I want python to abort) ...
> 
> without displaying the "Traceback (most recent call last):" stack trace?

Catch the exception anyway, then do sys.exit() or raise SystemExit.

Just


From pyth at devel.trillke.net  Sun Apr 28 10:24:17 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 28 Apr 2002 16:24:17 +0200
Subject: Update locals()
In-Reply-To: <20020428085009.A983@unpythonic.net>
References: <20020427195725.GA16217@devel.trillke> <ySFy8.28416$8D3.759092@news1.tin.it> <20020427233027.GE16217@devel.trillke> <0204280830290C.16995@arthur> <20020428115221.GF16217@devel.trillke> <20020428085009.A983@unpythonic.net>
Message-ID: <20020428142416.GG16217@devel.trillke>

On Sun, Apr 28, 2002 at 08:50:13AM -0500, jepler at unpythonic.net wrote:
> > class filename:
> >     """ Abstraction for filename ... """
> >     # ...
> > 
> >     for name in filters._stateful_all:
> >         doc=filters.__dict__.get(name,undoc).__doc__.strip()+'\n'
> >         exec indentedcode("""
> > 
> >     def %(name)s (self,*args,**kargs):
> >         '''%(doc)s'''
> >         return filters.%(name)s (*args,**kargs) (self)
> >     """ % locals()) 
> > 
> >     # ...
> 
> I don't quite see what you're doing here, but you might try something like
> (given nested scopes) (untested):

i dynamically define functions as filename-methods. Assume
name is 'access' you get

    def access(self,*args,**kargs):
        ''' checks filename for a given access-mode... '''
        return filters.access (*args,**kargs) (self)

defined right in your class scope. After you imported 
the module the methods will be there...
 
>     class filename:
> 	def __getattr__(self, attr):
> 	    if not hasattr(filters, attr): raise AttributeError, attr
> 	    filter_func = getattr(filters, attr)
> 	    ret = lambda *args, **kargs: filter_func(*args, **kargs)(self)
> 	    ret.__doc__ = getattr(filter_func, __doc__, undoc)
> 	    #setattr(self, attr, ret)
> 	    return ret

I am working a lot interactively (like many others with rlcompleter). Dynamic 
constructions like this won't show up. So the ret.__doc__ does not make much sense
here IMHO. help ('filename') won't show it either.  Otherwise your version 
should work. I don't know if you could handle working with
inspect.getargspec to construct the precise signature, though. This can
be done with the above 'templating' (i didn't bother so far).

> My version will also work properly if "filters" is extended at runtime with
> new attributes... It's reasonbly efficient to construct a function at
> runtime with 'lambda', and you can trade time for space by caching the
> attribute on the instance once it's been computed once

My construction was not about extending filters at run time but about
avoiding redundancies and expressing an adapter pattern at compile
time.

> Of course, given that you always immediately call the instance of the
> filters. function, I don't see why you can't simply write

quoting from the example at the end of my posting:

#...
resultfilters = AND(isfile, nolink, fnmatch('data*'), count)
recursefilter = access('wx')

for dir, fnlist in base.filterwalk( resultfilters, recursefilter ):
#...

Note especially that 

    recursefilter = access('wx')

does not map to

>     class filename:
> 	def access(self, mode='r'):

You would need an unbound partially initialized class-method.
You could try to use Currying (see ASPN-Cookbook) 
but it probably gets difficult if you look at the 'resultfilters'-
line. The AND is a stateful-filter as well and you can because
of the templating also do 

fn.AND(isfile,nolink,...) 

> There's probably some reason which isn't as apparent in this example as
> some of the more complex ones you've written, though...

The problem is i don't want to explain my filename+filters module
in boring details. I tried to extract the basic templating mechanism
and still provide some context to answer Alex Martellis question
about "Give me ONE use..." (for exec without explicit dictionaries).

I am not saying at all that there is no other solution (see your
own solution). Given the context, my 'templating' does still make a lot of sense
to me. Although it might be considered unusual or even unpythonic
but that's a matter of taste to some degree :-)

    holger




From mwh at python.net  Thu Apr 18 07:40:50 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 11:40:50 GMT
Subject: Readline/command history w/ interactive prompt
References: <mailman.1019061574.15837.python-list@python.org> <a9kcml$69k$1@news.wrc.xerox.com> <mailman.1019117616.20456.python-list@python.org> <lk662pmi8v.fsf@pc150.maths.bris.ac.uk> <mailman.1019126863.10557.python-list@python.org>
Message-ID: <lksn5tqlha.fsf@pc150.maths.bris.ac.uk>

Andy Gimblett <gimbo at ftech.net> writes:

> On Thu, Apr 18, 2002 at 10:04:18AM +0000, Michael Hudson wrote:
> > Andy Gimblett <gimbo at ftech.net> writes:
> > 
> > > On Wed, Apr 17, 2002 at 05:51:17PM +0000, Mark Jackson wrote:
> > > 
> > > > Try hitting control-r and typing a short string that last appeared
> > > > in the line desired ("my" or even "m" would work in this case).
> > > 
> > > _Nice_!  Thanks for that.  One question though: where's this feature
> > > documented?  :-)
> > 
> > "man bash", "man readline".  Not too hard...
> 
> Hmmm...  Personally, I can't imagine making the conceptual leap from
> "where's the python interpreter documentation" to "I know, I'll look
> at the bash man page" without prompting.

Well, it was you who mentioned bash in your original post.  I'm pretty
sure readline was developed as part of bash before being released on
it's own, but if you're not aware of the history (or I'm wrong), then
this isn't obvious.

> man readline, I admit, was more obvious.

Yup.

> Thanks for the pointers,

Hope they help!

I think I only use about 20% of what's possible with readline in bash;
but as my Mac came with only tcsh & zsh learning more readline now is
probably counter-productive.

Cheers,
M.

-- 
  It's a measure of how much I love Python that I moved to VA, where
  if things don't work out Guido will buy a plantation and put us to
  work harvesting peanuts instead.     -- Tim Peters, comp.lang.python


From philh at comuno.freeserve.co.uk  Sun Apr 14 13:59:57 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 18:59:57 +0100
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <m34rif19hj.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnabjgst.741.philh@comuno.freeserve.co.uk>

On 14 Apr 2002 01:09:44 +0200, Martin v. Loewis <martin at v.loewis.de> wrote:
>"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
>
>> You didn't read my post well enough-- we can already _DO_
>> that with Python v2.2.x:
>
>I read your post. You said that the idiom you bring up is not clear to
>you; I agree - adding a "not" just to get a 0/1 value is not clear.
>
>To me, using a function called "bool" is very clear.

To me, the idea that ^ means xor is totally unclear.

What's wrong with xor meaning xor?

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From sjmachin at lexicon.net  Fri Apr  5 20:01:49 2002
From: sjmachin at lexicon.net (John Machin)
Date: 5 Apr 2002 17:01:49 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018031177.21308.python-list@python.org>
Message-ID: <c76ff6fc.0204051701.68c63ecb@posting.google.com>

Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1018031177.21308.python-list at python.org>...
> [phil hunt, on language change]
> > ...
> > But is this changing going to be continual? Are the language
> > developers constantly going to tinker with the language, out of a
> >
> > Is there a final destination in sight, which these changes are
> > leading to?
> 
> It's possible it's the same final destination that C, Fortran, Perl and Java
> (for examples) are approaching with their own never-ending changes, but, if
> so, I doubt we'll live long enough to celebrate arrival.  If you want a
> change-free language, you're best off looking for a dead or dying language.

Indeed. However with all due (i.e. heaps) respect to the eminent 
personages involved: consolidation, defence & supply of the territory
gained has to be given a reasonably priority compared with that
assigned to attacking the next target -- in particular, some of us may
have preferred (a) docs on new-style classes (b) continuation of what
__slots__ seems to promise to have a higher priority than adding a
bool type :-)

> For example, Algol 60 hasn't changed in decades.

Another example: Sather -- a very good language (in its category) but
now
maintained (in some sense of the word) by a lone prophet crying in the
wilderness -- compare c.l.py and c.l.sather

> For that matter, Python
> 1.5.2 hasn't changed in years.  WRT any technology, you can jump off the
> train any time you like.  Unless you pick your exit point very carefully,
> though, you may end up alone in the middle of nowhere.
> 
> or-in-a-small-town-with-a-crate-full-of-8-track-tapes-ly y'rs  - tim

My crates are in the museum but AFAIR the mag tapes were 7-track or
9-track and the paper ones were 5-track [or was there a (6th) parity
track?]; what's this 8-track story?

Follow-ups-to-alt.fan.Friden.flexowriter-ly y'rs
John


From loewis at informatik.hu-berlin.de  Sun Apr 14 08:39:12 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 14 Apr 2002 14:39:12 +0200
Subject: C API Q II : __main__.__dict__.keys() not same as dir()
References: <a9bl5l$2aq$2@green.tninet.se>
Message-ID: <j4r8lipi8f.fsf@informatik.hu-berlin.de>

"Johan Hahn" <johahn at delta.telenordia.se> writes:

> I figured that since the InteractiveConsole hangs after calling
> interact() you have to have atleast 2 threads to use that, right?

Unless it is acceptable that it hangs: yes.

> I thought the documentation was a little vague on that.

In what respect? The interact documentation says:

  "Closely emulate the interactive Python console."

So it is pretty clear to me that it is an infinite loop - just as the
interactive Python console.

> Where does the user imported modules (and all other objects too for that
> matter) hide??

They don't hide... When you import a module, two things happen:

1. the module is loaded and run
2. the resulting module object is put into sys.modules
3. the resulting module object is added to the namespace that
   executed the import statement. If you do it in the interactive
   prompt, it is added to globals(), which is __main__.__dict__.
[as any good list announced with two items, I have three of them]

Regards,
Martin


From aseem at neurobs.com  Tue Apr  2 15:58:36 2002
From: aseem at neurobs.com (A M)
Date: 2 Apr 2002 12:58:36 -0800
Subject: Zope Q: no restart button & shutdown problem
Message-ID: <ed6d1c1a.0204021258.53585416@posting.google.com>

Hi all,
I am running  (Zope 2.5.0 (binary release, python 2.1, win32-x86),
python 2.1.2, win32) on Win2K. Everythin seems to be running fine. My
only problem is that in the Control Panel I only get the Shutdown
button and there is no restart button. I just installed the binary and
did as was told in the docs, but I cant seem to find an answer or
explantion of this behaviour. Any help would be greatly appreciated.

Also when I shutdown I get the error message:

 Zope has exited normally.

Traceback (innermost last):
  File F:\KANDIN~1\lib\python\ZPublisher\Publish.py, line 150, in
publish_module
  File f:\kandinsky\lib\python\Products\Localizer\__init__.py, line
65, in new_publish
  File F:\KANDIN~1\lib\python\ZPublisher\Publish.py, line 114, in
publish
  File F:\KANDIN~1\lib\python\ZPublisher\Publish.py, line 98, in
publish
  File F:\KANDIN~1\lib\python\ZPublisher\mapply.py, line 88, in mapply
    (Object: manage_shutdown)
  File F:\KANDIN~1\lib\python\ZPublisher\Publish.py, line 39, in
call_object
    (Object: manage_shutdown)
  File F:\KANDIN~1\lib\python\App\ApplicationManager.py, line 342, in
manage_shutdown
    (Object: Control_Panel)
SystemExit: 0


but I gess SystemExit: 0 means everythings OK. Can anybody shed some
light on that??

Thanks in advance.
Aseem


From jeremy at alum.mit.edu  Mon Apr 29 23:39:36 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 29 Apr 2002 20:39:36 -0700
Subject: cPickle problem
References: <Ssiz8.63$bK3.1993791@newnews.cc.stevens-tech.edu> <FAiz8.78905$vF6.2347612@news2.tin.it>
Message-ID: <b0f083db.0204291939.3e0b1ce8@posting.google.com>

Alex's comments should straighten you out.  In addition, I thought
this tip might be helpful in the future: If you get an error with
cPickle that you don't understand, try again with pickle.  The
exception raised by pickle will contain many layers of Python function
calls.  You can often use this extra information to figure out exactly
what object is causing a problem.

Jeremy


From see at below  Mon Apr 22 19:59:45 2002
From: see at below (Paul Foley)
Date: 23 Apr 2002 11:59:45 +1200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com>
Message-ID: <m2n0vvxp1q.fsf@mycroft.actrix.gen.nz>

On Mon, 22 Apr 2002 13:37:39 GMT, Brett g Porter wrote:

> "Paul Foley" <see at below> wrote in message
> news:m2wuv0xr1u.fsf at mycroft.actrix.gen.nz...
>> On Sun, 21 Apr 2002 02:32:29 -0700, James J Besemer wrote:
>> 
>> > That being said, I guess I just don't get the strong reaction to this
> secondary
>> > point.  I'm new here but I get the impression that "lisp" is a naughty
> word,
>> > for some unfathomable reason
>> 
>> A lot of people think "Lisp" means "Scheme".
>> Scheme sucks, so they think "Lisp" sucks.
>> 
>> [Scheme isn't Lisp.  Scheme are Lisp are only distantly related, and
>> Scheme is about as closely related to Algol as to Lisp...]
[...]
> In what ways are Scheme and Lisp only distantly related?

Look up last month's "Why is Scheme not a Lisp?" thread on Google's
comp.lang.lisp archive.

-- 
In modern physics, the more logical you are, the more wrong you are.
                                                           -- Lama Govinda
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From bkc at Murkworks.com  Tue Apr  9 10:28:01 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 9 Apr 2002 10:28:01 -0400
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net>
Message-ID: <3cb2f8c9_7@goliath.newsgroups.com>

Neither a pic nor an MC68HC11 has enough power to run Python.

--
Novell DeveloperNet Sysop #5

_
"akhar" <akhar at videotron.ca> wrote in message
news:_lvs8.33252$lU2.1209375 at wagner.videotron.net...
> Has anyone ported Python to embedded devices? I heard of the deeply
embedded
> python porject but the link is dead. What would it take to port python to
an
> embedded platform (whether be it the mc68hc11 or PICs)?
>
> Regards
>
> Stephane
>
>




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From goodger at users.sourceforge.net  Tue Apr  2 22:12:50 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Wed, 03 Apr 2002 03:12:50 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204030123.BAA15136@crown.off.ekorp.com> <3CAA60A5.F88697D5@engcorp.com> <mailman.1017799677.14060.python-list@python.org>
Message-ID: <B8CFDD87.20D34%goodger@users.sourceforge.net>

Richard Jones wrote:
> On Wed, 3 Apr 2002 11:53, Peter Hansen wrote:
>> I would suggest an even shorter version of the quickref, focused on
>> only those items needed in 80% of embedded documentation.  It does
>> _look_ complicated at first.
> 
> I agree - I think the quickref definitely needs:
> 
> 1. reorganising to filter the most common, simple markup to the top, and
> 2. much higher status/metioning on the ReST website.

You're both talking about a "Gentle Introduction" document, I think.
There's certainly room for one -- any volunteers?

I like the quickref as it is though, a true "Quick Reference".  It should be
reasonably complete, for quick lookups (that's why I added the table of
contents).  It's not really suitable as a primer though, and that document
doesn't exist yet.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From dfan at harmonixmusic.com  Tue Apr  9 17:43:07 2002
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 09 Apr 2002 17:43:07 -0400
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de>
Message-ID: <uvgb0eefo.fsf@turangalila.harmonixmusic.com>

Thomas Guettler <pan-newsreader at thomas-guettler.de> writes:

| Up to now I was happy with emacs because I never used
| to programm it with lisp. Now I need to and find it difficult.
| 
| Is there an emacs-like editor which is scriptable with python?

Pymacs allows you to script Emacs with python.  I haven't used it, though.

http://www.iro.umontreal.ca/~pinard/pymacs/

Dan

-- 
http://www.dfan.org


From max at alcyone.com  Wed Apr  3 21:06:50 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 18:06:50 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com>
Message-ID: <3CABB53A.86D7D79F@alcyone.com>

djw wrote:

> Pardon my ignorance (I've only been working with Python for a short
> while), but I don't understand why an idiom like:
> 
> if x == True:
> 
> would be/should be an error (or warning initially).

Because in Python, the canonical way to test for truth, regardless of
what x might be, is:

	if x:
	    ...

Even with PEP 285, there are more values which qualify as true than just
True.  Explicit testing against one of these true values excludes all
others.

> I see C/C++ code
> like this all the time:
> 
> bool x;
> x = Foo(Bar);
> if (x==true)
> {...

This is C++, not C.  In C++ also, the proper way to write this is:

	if (x)
	    ...

Since the if conditional does an implicit conversion to bool, the
explicit test is unnecessary.  The following is completely legal code,
and evaluates the truth clause:

	int x = 1;
	if (x)
	    ...

In C89, which lacks an explicit Boolean type (although one was added to
C99), the above is also the proper way to test for truth.  Just as in
Python, in C++ (and C99) there are more values considered true than just
`true'.  In C89 there isn't even an explicit Boolean type.

> Of course, this could also be written in the more abbreviated form as
> well. But, why force that on the programmer? I really prefer the more
> explicit  comparison of x with true rather than the if(x) idiom.

Because you're leaving out a whole swatch of values that evaluate to
true in a conditional, they just don't have the explicit value true (of
type bool).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From patmiller at llnl.gov  Wed Apr 10 14:21:59 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Wed, 10 Apr 2002 11:21:59 -0700
Subject: Binding a reference to a variable (solved?!?)
References: <E16v1a2-0005a6-00@mail.python.org>
Message-ID: <3CB482C7.DD49378B@llnl.gov>

Here's a scary way to do what Andrew was looking for.  It has
the proper form and fewer (but still significant) restrictions.

>>> x = 7
>>> set(x,99)
>>> print x

I'll likely submit the following as a ASPN recipe, though
it clearly needs a better description of motivation.  Just
because you CAN do a thing doesn't mean you SHOULD
do a thing.

#**************************************************************************#
#* FILE   **************        globally.py        ************************#
#**************************************************************************#
#* Author: Patrick Miller April 10 2002                                   *#
#**************************************************************************#
#* Real hack to allow update of global variables by value, not name       
#*
#* The ''set'' function needs a dictionary within which to look for values
#* (defaults to the current global scope).  The ''setAnywhere'' function
#* looks in all the global module scopes.
#*        
#* Note that this only works for items with unique identifiers.
#* For instance, small integers 'a = 3; b = 3' have shared python
#* objects (id(a) == id(b)).  The method works OK for lists
#* and dictionaries and the like because python creates a new unique
#* address for each instance (unlike int's which look to share)
#**************************************************************************#

x = 3
y = [1,2,3,4,]
z = {"something": 0}


def set(lhs,rhs,G=globals()):
    address = id(lhs)
    identifiers = map(lambda x: id(x), G.values())
    try:
        index = identifiers.index(address)
    except ValueError:
        raise ValueError,lhs
    label = G.keys()[index]
    G[label] = rhs
    return rhs

def setAnywhere(lhs,rhs):
    import sys
    for module in sys.modules.values():
        try:
            dict = module.__dict__
            try:
                set(lhs,rhs,module.__dict__)
                break
            except ValueError:
                pass
        except:
            pass
    else:
        raise ValueError,lhs
    return rhs

print
print 'before',repr(x),repr(y),repr(z)
set(x,99)
set(y,'something new')
set(z,42)
print 'after',repr(x),repr(y),repr(z)

import math
print 'math.pi',math.pi
setAnywhere(math.pi,3) # see http://www.snopes2.com/religion/pi.htm
print 'math.pi',math.pi
        

-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

Adults are obsolete children.
-- Dr. Seuss, humorist, illustrator, and author (1904-1991)




From fperez528 at yahoo.com  Tue Apr 16 13:56:15 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 16 Apr 2002 11:56:15 -0600
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu> <861ydg3ff3.fsf@haggis.aae.uiuc.edu> <3CBC5686.BCBCF663@noaa.gov>
Message-ID: <a9holn$5t1$2@peabody.colorado.edu>

Chris Barker wrote:

> "D. Michael McFarland" wrote:
>> The bottlenecks now seem to be in
>> memory management, which I'm sure Python does better than I would in
>> C,
> 
> better than me too, I'm sure, but NumPy code often ends up making a lot
> more copies of data that similar C code would, so that might very well
> get better in C too.
> 

This is another one for which weave can be a godsend. The blitz module in 
weave can for many numeric expressions generate equivalent C++ code which 
avoids all the intermediate copies. I don't want to sound like a cheerleader 
(I'm not a developer of weave, only a  user), but weave is really amazing in 
how easily it allows you to integrate Numeric pyhton code with C snippets.

If anyone needs, I have a bunch of little examples lying around which I wrote 
when I was learning weave and which I can pass along, just drop me a  line.

Cheers,

f.


From python at rcn.com  Sat Apr 13 22:15:09 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sat, 13 Apr 2002 22:15:09 -0400
Subject: 2.1 vs. 2.2
References: <mailman.1018739908.9275.python-list@python.org> <roy-56DE95.20201113042002@news1.panix.com>
Message-ID: <a9aop2$p7h$1@bob.news.rcn.net>

"Roy Smith" <roy at panix.com> wrote in message news:roy-
> ISO certification is a wonderful thing.  It lets people who worry about
> that sort of stuff check off the "ISO certification" checkbox and be
happy.
> Other than that, it's not clear it has any value.

Yes.  Double what Roy said.
Standards have value when they are able to pull together multiple competing
variations of the same language.  Python doesn't really have that issue.
Jython and Pippy are a little different but only in ways that make sense for
their platform.  IOW, Python is its own standard.


Raymond Hettinger




From aahz at pythoncraft.com  Fri Apr 26 17:43:00 2002
From: aahz at pythoncraft.com (Aahz)
Date: 26 Apr 2002 17:43:00 -0400
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org> <aa48nq$a61$1@panix1.panix.com> <mailman.1019640045.27779.python-list@python.org>
Message-ID: <aachl4$ltf$1@panix2.panix.com>

In article <mailman.1019640045.27779.python-list at python.org>,
Graham Ashton  <gashton at cmedltd.com> wrote:
>On Tue, 2002-04-23 at 19:21, Aahz wrote:
>> In article <mailman.1019552082.8541.python-list at python.org>,
>> Graham Ashton  <gashton at cmedltd.com> wrote:
>>>
>>>>>> class A(AbstractSingleton, object):
>>>...   pass
>> 
>> Maybe you should tell us a bit more about what you're trying to do so
>> that you needn't rely on crufty internal stuff to get the correct result.
>
>Okay, good point. I'm using the Alex's Borg pattern in place of a
>singleton, as shared state will do quite nicely. I'm also trying to
>protect myself from getting caught out by new style classes (in other
>words it doesn't work -- in general -- for classes derived from, for
>example, dict).

What do you need Borg for?  Alex himself has agreed that using a module
as a singleton is the correct answer 80-90% of the time.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From jeff at ccvcorp.com  Mon Apr  1 19:05:14 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 1 Apr 2002 16:05:14 -0800
Subject: runtime location discovery
References: <mailman.1017453478.23509.python-list@python.org>
Message-ID: <MPG.17129563ab74116c989686@news.nwlink.com>

In article <mailman.1017453478.23509.python-list at python.org>, 
geoff at gerrietts.net says...
 
> The other night, I tried to put in a chunk of code that allowed the
> program to be executed from an arbitrary location (like a user's home
> directory) without previously knowing where we were installed
> (allowing the app to be installed wherever it seems most natural to
> the system administrator).
> 
> The code I wrote did this:
>     root_location = os.path.dirname(sys.argv[0])

This also suffers from the problem that, if you cd to the 
directory that the script is in and then run it, sys.argv[0] will 
not contain path information.

One solution that I've used, is to look at the __file__ attribute 
of a module that I've imported, like so:

import MyModule
installdir = os.path.dirname(MyModule.__file__)

This will tell you the directory that MyModule.py is located in, 
which you can arrange to be the directory that you're interested 
in.  ;)  Be aware, though, that in your original script (with 
__name__ == '__main__'), the __file__ attribute is undefined 
(which is why I grab it from a module, instead of directly in the 
executing script).

-- 

Jeff Shannon
Technician/Programmer
Credit International


From johnroth at ameritech.net  Wed Apr  3 06:34:01 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:34:01 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <uaknnsdgu514a6@news.supernews.com> <B8CFEB59.20D3E%goodger@users.sourceforge.net>
Message-ID: <ualpqv2jufaace@news.supernews.com>

"David Goodger" <goodger at users.sourceforge.net> wrote in message
news:B8CFEB59.20D3E%goodger at users.sourceforge.net...
> John Roth wrote:

>
> 3. Are you objecting to extended, comprehensive docstrings, that may
be
>    unsuitable for ``print my_object.__doc__``, and may take up space
in
>    the running program?  Then the "Additional Docstrings" proposal
from
>    PEP 258 may reassure you.

It's this one. PEP 258 does not reassure me, until I see it
in the same, or an earlier, version of Python than this goes into.
Unless that happens, the reference is pie in the sky snake oil.

> If those don't address your concern, please elaborate.
>
> > Some other mechanism should exist for extensive documentation that
is
> > intended for translation to some other format for eventual printing
> > or display independent of the program.
>
> That's the whole point of docstring extraction tools.  This PEP is
merely
> proposing a standard for the format of those docstrings.
>
> > I much prefer not to have this stuff in the compiled module.
>
> Then don't put it in.  Some don't want inline documentation.  Many do,
and
> this PEP is for them.

I think you've missed the point. If you **need** markup, then
it's too complex for the quick help that docstrings were originally
intended for, and that many people routinely include in their programs.

I'm not against including documentation in the **source**, nor am
I against a supported syntax for doing so. What I'm against is
coopting a mechanism intended for something else, when at least
some people want **both**.

John Roth
>
> --
> David Goodger    goodger at users.sourceforge.net    Open-source
projects:
>  - Python Docstring Processing System:
http://docstring.sourceforge.net
>  - reStructuredText: http://structuredtext.sourceforge.net
>  - The Go Tools Project: http://gotools.sourceforge.net
>




From gh_pythonlist at gmx.de  Mon Apr 29 06:10:38 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 29 Apr 2002 12:10:38 +0200
Subject: Python newbie
In-Reply-To: <ab96a890.0204282321.6c22697e@posting.google.com>
References: <ab96a890.0204131438.5cce8f66@posting.google.com> <slrnabmlpd.a8q.ak@ak.silmarill.org> <ab96a890.0204282321.6c22697e@posting.google.com>
Message-ID: <20020429101038.GA23223@lilith.my-fqdn.de>

* slack <slackness at caramail.com> [2002-04-29 00:21 -0700]:
> [...] I would like to know one more thing... Is it possible to use
> Python to program for Windows CE devices??

Yes. Btw. here's a list of Python ports:

http://dmoz.org/Computers/Programming/Languages/Python/Ports/

If anybody knows one that's not already there, please submit it ("add
URL" at the top of the page).

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 13.0 ?C      Wind: 7.2 m/s




From matthewm at ihug.co.nz  Wed Apr 10 04:04:48 2002
From: matthewm at ihug.co.nz (Matthew)
Date: Wed, 10 Apr 2002 20:04:48 +1200
Subject: Win32 and Excel question
Message-ID: <3CB3F220.2B8E668F@ihug.co.nz>

Hi,
I'm playing with win32COM with a little success. I have a large XL table
with 8 columns and about 3500 rows. What I want to do is search only
column B for a value and return the row number where this value is
found. So, I have 2 questions:-
1) Howto construct a Win32COM statement to do this? So far I have this:-

>>>
found=myxl.ActiveSheet.Range("B1:B3500").Find(<value_i'm_looking_for>
However, this search finds the first match in any column between B1 and
B3500. Yep, their are similar values in other cells of the s/s
>>> found.Row
25
>>> found.Column
1 (ie: Column 'A')
Help please.
2) Howto return a tuple of all values in only column 'B' and then use
python to search the tuple for the value I'm after. It strikes me that
this is probably a faster approach?
Thanks for helping. matthew.
BTW: Thanks Mark Hammond for a v.cool and useful tool.




From st-newsgroups at thomas-guettler.de  Fri Apr  5 07:57:09 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Fri, 05 Apr 2002 14:57:09 +0200
Subject: RPC for ZODB via https
Message-ID: <3CAD9F25.8010708@thomas-guettler.de>

I interested in how I can access ZODB via an RPC mechanism
which uses https.

Possible solutions:

+use Zope and XML-RPC via https.

+use xmlrpclib

But xml-rpc is not object-oriented. Has someone other ideas?

  thomas



From db3l at fitlinxx.com  Mon Apr 29 20:14:00 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 29 Apr 2002 20:14:00 -0400
Subject: String handling bug in Python
References: <mailman.1019884787.11509.python-list@python.org>
Message-ID: <u4rhurqk7.fsf@ctwd0143.fitlinxx.com>

Tim Peters <tim.one at comcast.net> writes:

> Forward slashes are easier for almost everyone to type than backward
> slashes, and work fine on Windows.  Paths using forward slashes also run
> 13-17 times faster in Python code, due to Python's Unix heritage <wink>.

Just as an FYI, forward slashes don't actually work *everywhere* on
Windows - there are notable UNC names that must use \\ as a prefix and
won't accept // in the API.  It's not that common but I've been bitten
by it in the past.

But they definitely work in any local file API.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From max at alcyone.com  Wed Apr  3 14:40:22 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 11:40:22 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff><yu99wuvovkmm.fsf@europa.research.att.com><00af01c1db2a$e7316d10$ced241d5@hagrid><200204031619.g33GJXm02577@europa.research.att.com><016b01c1db31$8da51010$ced241d5@hagrid> <yu99ofh0vg4x.fsf@europa.research.att.com> <mailman.1017858013.30956.python-list@python.org>
Message-ID: <3CAB5AA6.2B9A4E16@alcyone.com>

Fredrik Lundh wrote:

> (Python docs tend to use "true" to mean anything which isn't "false",
> and
> where operator.truth returns the right thing.  this won't change
> post-PEP
> -- if you start interpreting "return true" as "return True", you'll be
> asking
> for trouble)

But these functions return Boolean values but do not guarantee that they
will be 0 or 1, just something which evaluates as the appropriate truth
value.  So the 0 and 1 is not "more than just a convention"; even
standard libraries don't guarantee their usage.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From richard at electrophobia.com  Wed Apr 24 05:29:00 2002
From: richard at electrophobia.com (Richard Parker)
Date: Wed, 24 Apr 2002 09:29:00 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com>
Message-ID: <B8EBC8E9.EFA0%richard@electrophobia.com>

Paul Rubin at phr-n2002a at nightsong.com wrote on 4/24/02 1:41 AM:

> Richard Parker <richard at electrophobia.com> writes:
>> I just took a quick look at your Python code.  I'd encourage you to not use
>> the secret prefix method to construct a MAC from a hash function, i.e.
>> MAC(x) = H(K || x).  This method is generally considered to be insecure.
>> Use the HMAC construction instead.
> 
> Thanks.  What's the real deal with this?  Is an insecurity known, when
> K and X are the output of encryption functions, and not chosen by
> attackers?  

The secret prefix method is vulnerable to an "appending" attack.  An
attacker given a message-MAC pair can compute, without the key, the correct
MAC for a new message that has the original message as a prefix.  In the
case where you are using a MAC as an authentication code you are applying
the MAC to the ciphertext, so the new message will verify as valid, but once
decrypted it will likely have a garbage suffix.  Many applications will
detect this modification, but it could well be a problem for some
applications.  The fact that you truncate your authentication code provides
resistance to this attack, however it still a good idea to use a MAC
construction that is not vulnerable to this kind of attack.

> I know that HMAC tries to be robust under unknown weakness
> of the underlying hash function, but since the OFB mode encryption
> depends on the hash function being strong, HMAC likely doens't help
> that much.  

The portions of your code that use H() just as a hash function don't need to
use HMAC.

> Do you think it matters in practice, given that this is
> running in an interpreted language on a general purpose PC?

Yes.  I think it is an unnecessary limitation.  If a client of your library
uses it to protect something valuable, an attacker is unlikely to have the
same limitations.  The authentication code is an important part of the
security of an encryption scheme, don't neglect it.

-Richard



From rjones at ekit-inc.com  Mon Apr  1 23:07:51 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 2 Apr 2002 14:07:51 +1000
Subject: Extending Python on Windows
In-Reply-To: <20020402045919.GC1247@lilith.hqd-internal>
References: <3CA90BE1.9010703@cfl.rr.com> <20020402045919.GC1247@lilith.hqd-internal>
Message-ID: <200204020407.EAA11781@crown.off.ekorp.com>

On Tue, 2 Apr 2002 14:59, you wrote:
> * Gregory Propf <gpropf at cfl.rr.com> [2002-04-01 20:39 -0500]:
> > [trouble compiling extension modules]
>
> Solution: distutils
>
> A good starting point is Alex Martelli's recipe:
>
> http://groups.google.com/groups?q=group:comp.lang.python+alex+distutils+rec
>ipe&hl=en&selm=9v57kr02q3i%40enews4.newsguy.com&rnum=1

Note that there's optional bits in that URL. All you actually need is the 
message id:

  http://groups.google.com/groups?selm=9v57kr02q3i%40enews4.newsguy.com


     Richard



From dalke at dalkescientific.com  Wed Apr 17 15:51:26 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 17 Apr 2002 13:51:26 -0600
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com>  <a9jaah$493b$1@as201.hinet.hr> <mailman.1019068235.12319.python-list@python.org>
Message-ID: <a9kjh1$qc5$1@slb1.atl.mindspring.net>

Ian Bicking:
>+= is different from .extend()... for instance:
>
>q1 = [1, 2, 3]
>q2 = q1
>q3 = [4, 5, 6]
>
>q1 += q3   # then...
>q2 == [1, 2, 3]

Really?

$ python
Python 2.2b2+ (#10, Dec 12 2001, 04:25:14)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>
>>> q1 = [1, 2, 3]
>>> q2 = q1
>>> q3 = [4, 5, 6]
>>>
>>> q1 += q3
>>> q1
[1, 2, 3, 4, 5, 6]
>>>

So your statement isn't correct.  For lists, += is the same as .extend
In listobject.c you can even see that 'extend' is defined as

static PyObject *
listextend(PyListObject *self, PyObject *b)
{

 b = PySequence_Fast(b, "list.extend() argument must be iterable");
 if (!b)
  return NULL;

 if (listextend_internal(self, b) < 0)
  return NULL;

 Py_INCREF(Py_None);
 return Py_None;
}

while += is defined as

list_inplace_concat(PyListObject *self, PyObject *other)
{
 other = PySequence_Fast(other, "argument to += must be iterable");
 if (!other)
  return NULL;

 if (listextend_internal(self, other) < 0)
  return NULL;

 Py_INCREF(self);
 return (PyObject *)self;
}

The code is almost identical, except for that return self part.
(I don't know why that's needed.)

                    Andrew
                    dalke at dalkescientific.com





From aahz at pythoncraft.com  Wed Apr 10 15:53:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 15:53:04 -0400
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <a8vpp5$67d$1@panix1.panix.com> <yu99pu18lai2.fsf@europa.research.att.com>
Message-ID: <a92570$kop$1@panix1.panix.com>

In article <yu99pu18lai2.fsf at europa.research.att.com>,
Andrew Koenig  <ark at research.att.com> wrote:
>>> Now, f is an object that I can use to rebind the name x.  For
>>> example, if I execute f(42), that sets x to 42.
>
>Aahz> You really don't want to do this in Python.
>
>One thing that fascinates me is how the nature of a language affects
>the preferred idioms therein.  To help me further my understanding,
>would you mind explaining to me why this particular idea is undesirable
>in Python?  I can think of several possible reasons, but I don't know
>which reason is most plausible to the Python community.

You've already gotten a couple of answers; from my perspective, the best
I can say is that is that Python supports mutable objects but not
mutable names.  I think you know Python well enough to understand what I
mean, but if it's not clear, I'll try expanding it.

To repeat some other posts, give us more context for what you're trying
to do, and we'll see if we can come up with Pythonic idioms for it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From jb at cascade-sys.com  Fri Apr 26 13:04:27 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 26 Apr 2002 10:04:27 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <tyf1yd25zgs.fsf@pcitapi22.cern.ch>
Message-ID: <3CC9889A.1DAB6B6F@cascade-sys.com>

Jacek Generowicz wrote:

> No, as Paul said, they are used for "function calls". Admittedly, in
> mathematical expressions function call often coincide with
> sub-expressions, but it does depend on the sub-expression in
> question.

Thanks for the clarification.  Now I get it.

> Compare
>
>    (+ 1 2 3 4 5 6 7 8 9 10)
>
> with
>
>    1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10
>
> Gosh, what a lot of typing I have to do in Python, compared to lisp.

So that begs the question: is it more common to have lengthy "sub expressions"
like your example or is my example with a mixture of operations more typical?

Now we're down the realm of statistics, which means it'll never be resolved.
(I shoulda seen that one comin'.)

> But, if you are
> dealing with mathematical expressions, then, obviously, you will use
> some sort of infix notation macro:

I am not familiar with Common Lisp and did not know that was common.  Though I
suppose I should have guessed that anything's possible.

> Incidentally, it would be interesting to see how many people would
> correctly work out in their heads (on first attempt) what 1+2*3/4-5
> evaluates to (leaving aside the fact that in Python it evaluates to
> -3[*]) . . . and then compare it to the success rate when considering
> the equivalent (+ 1 (/ (* 2 3) 4) -5). Or maybe how long it takes them
> to get the right answer in each case.

Hard to say.  Do the survey and lemme know.

Then too, this is a simple example and it seems the big trick in either case is
in knowing how divide works.

In Python's case it's truly a trick question as there presently is absolutely
no way to predict what the result will be without several bits of additional
information.  (I came late to the argument, but personally I think so called
"true division" is/was a big mistake.)

Then too, the results probably vary among Lisp implementations, so you'd have
to be more specific in that case too.  IIRC, some systems were integer only,
some were real only and some had both numeric types.  I dunno what's standard
these days.

So I guess the "trick" parts of the question would far outweigh any data about
infix vs. postfix.

> > > I'd expect Lisp syntax to be
> > > rather less error-prone, if anything, since it's much simpler
> >
> > I think that's debatable.  Certainly simplicity makes it easier to
> > learn, so that's a plus.  But it's still very tedious to get all the
> > parentheses matched up just right.
>
> It seems that many (most?) people who actually try to write lisp
> without making a big issue of the parentheses, very rapidly come to a
> conclusion that disagrees with yours.

1. I think that's a fair statement.  Earlier in the week, someone suggested I'd
be more happy with Python after I was "reprogrammed" to think "Pythonically"
(which sounds like something L. Ron Hubbard would say.)  I understand that
after a period of adjustment, people can function normally while wearing
glasses that flip their entire field of vision upside down.  It takes a few
days to adjust after putting them on and another couple days to adjust after
taking them off.  So, although I can't disagree with your statement (people
adapt), I don't think it's a very strong point.  Humans are adept at overcoming
much more severe handicaps.

2. We can speculate and rationalize all day but without knowing the actual
error statistics, productivity numbers, etc., we're just airing opinions and
assertions.

3. I'm personally conflicted in the matter because when I was doing a lot of
lisp programming I DID come to appreciate that the syntax was not a huge
problem.  You get used to it and in some ways it's kind of nice.

4. The point I've been trying to make is that the syntax is enough of a
departure from most peoples' training that it constitutes a genuine barrier to
entry.  Peoples' common objections may be superstitious and unfounded in the
long run but it does require a period of adjustment and retraining (thus it
costs).   And whether they're better or worse off in the long run I still
maintain is "debatable."

5. Finally, I think the problem here is that we're all making passionate,
definitive statements about items that ultimately really boil down to matters
of personal preference.  And there's no accounting for taste.

Thanks for the clarifications and for taking the time to reply.

REgards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From gerhard at bigfoot.de  Tue Apr 30 07:01:32 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 30 Apr 2002 11:01:32 GMT
Subject: Third Draft: Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch> <MNiz8.32691$8D3.925320@news1.tin.it> <mailman.1020117911.9648.python-list@python.org>
Message-ID: <slrnacsuc9.vg.gerhard@developer2.opus-gmbh.net>

In article <mailman.1020117911.9648.python-list at python.org>, Laura Creighton wrote:
> Alex, Alex, surely you know that the main advantage of oct is that 
> dec 25 == oct 31  which you can use to prove that all Christians are
> Satanists.
 
I don't remember the proof for this obvious fact. Could you tell me how it
works again?

Gerhard


From jim at publishingresources.com  Tue Apr 16 12:27:43 2002
From: jim at publishingresources.com (Jim Abrams)
Date: Tue, 16 Apr 2002 16:27:43 -0000
Subject: KeyboardInterrupt errors in ASP pages
References: <508cb8ae.0204110045.431dadee@posting.google.com> <3CB550C9.1000902@mxm.dk> <3CB83C5C.9040704@skippinet.com.au>
Message-ID: <Xns91F27EC4E98Ajimpublishingresourc@209.249.90.101>

Mark Hammond <mhammond at skippinet.com.au> wrote in
news:3CB83C5C.9040704 at skippinet.com.au: 

> Max M wrote:
>> B wrote:
>> 
>>> I am using Python as the language for creating server side asp pages.
>>> The problem is that when using Response.Redirect(page), I seem to get
>>> the message below randomly but frequently enough to be a problem.
>>>
>>> Has anyone else found this problem and is there a way around it. Is
>>> it a bug in the Response.Redirect code?
>>>
>>> Python ActiveX Scripting Engine error '80020009'
>>> Traceback (innermost last): KeyboardInterrupt
>>> /Default.asp, line 0
>> 
>> 
>> As far as I understand it's a known bug, but without any known
>> solution. I get it too at times.
>> 
>> I have just signed over a project to a customer on .asp/iis so I am a 
>> bit worried about it. But I don't really know what to do about it.
>> 
>> Maybe I should write a test harness to pound the IIS to see if I can 
>> provoke the error. 
> 
> Yes please - that is exactly what I need.  I need some way to provoke 
> the error often enough that I can afford to "wait" for it in a
> debugger. 
>   Anything you can do to get a reliable repro would be a big help!
> 
> Mark.
> 

If I remember correctly, (from what Mark has said), this error is caused 
not by Python or win32 alone, but by extraneous system factors affecting 
Python.

I've been able to workaround it with this in global.asa:

<SCRIPT LANGUAGE=VBScript RUNAT=Server>
function noOp()
end function
</SCRIPT>

<script language=Python runat=server>
def Application_OnStart():
	try:
		import signal
		signal.signal(signal.SIGINT, lambda a,b:None)
	except AttributeError:
		pass
	except ValueError:
		pass
</SCRIPT>


Desperate measures...

-J


From Oschler at earthlink.net  Thu Apr  4 02:07:15 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 4 Apr 2002 02:07:15 -0500
Subject: IDE Wars, the prequel!
References: <mailman.1017880688.9896.python-list@python.org>
Message-ID: <6RSq8.29207$0T6.8753904@e3500-atl1.usenetserver.com>

Brian Quinlan <brian at sweetapp.com> wrote in message
news:mailman.1017880688.9896.python-list at python.org...
> ActivePython is ActiveState's Python distribution, which is available
> for several platforms (at least Windows and Linux). ActivePython is 99+%
> compatible with PythonLabs Python but it has a nicer installer and some
> extra modules. One of those extra modules is win32all, which is a Python
> extension written by Mark Hammond. This extension, which includes the
> PythonWin IDE, only works on Windows.
>
> So you can install ActivePython on Linux but it will not include the
> PythonWin IDE (though it does include IDLE).
>
Brian,

Thanks, that clears it up for me.  Seeing as I'm going to use Zope, Quanta
is starting to look very attractive since it does HTML editing, integrates
well with Zope, and has facilities for Python.

thx






From grante at visi.com  Wed Apr 17 15:39:56 2002
From: grante at visi.com (Grant Edwards)
Date: Wed, 17 Apr 2002 19:39:56 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net> <ypjv8.42546$vm6.5594931@ruti.visi.com> <a9khoc$j7h$1@nets3.rz.RWTH-Aachen.DE> <a9ki3d$jip$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <gckv8.42736$vm6.5602598@ruti.visi.com>

In article <a9ki3d$jip$1 at nets3.rz.RWTH-Aachen.DE>, Philipp Lenssen wrote:
> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
> news:a9khoc$j7h$1 at nets3.rz.RWTH-Aachen.DE...
>>..
>> The three commented lines above would not have been needed if I'd use
>> something like '... name="' + (key != 'station') ? key : 'station' + '"
> ...'
>>..
> 
> That should be '... name="' + (key != 'station') ? key : 'lastStation' + '"
> ...'
> 
> Or since I sometimes prefer testing in a positive way rather than the most
> likey, '... name="' + (key == 'station') ? 'lastStation' : key + '" ...'

There's a dictionary lookup method called get() you can use
where you can specify the value to be returned if the key
doesn't exist. Using your previous example:

>>> shorthand = {'m':'male', 'f':'femail'}
>>> key = 'm'
>>> shorthand.get(key,key)
'male'
>>> key = 'whatever'
>>> shorthand.get(key,key)
'whatever'

-- 
Grant Edwards                   grante             Yow!  Nipples, dimples,
                                  at               knuckles, NICKLES,
                               visi.com            wrinkles, pimples!!


From davidccarson at hotmail.com  Fri Apr 19 10:10:02 2002
From: davidccarson at hotmail.com (David Carson)
Date: 19 Apr 2002 07:10:02 -0700
Subject: raw_input bug only in cygwin python
References: <12257ec4.0204181907.7a79b15b@posting.google.com> <kuavbukpa505nru9g6kfivo3bfrtirda05@4ax.com>
Message-ID: <12257ec4.0204190610.44c9774@posting.google.com>

Tim Roberts <timr at probo.com> wrote in message news:<kuavbukpa505nru9g6kfivo3bfrtirda05 at 4ax.com>...
> davidccarson at hotmail.com (David Carson) wrote:
> 
> >Here is a snippet of code:
> >
> >01: #!/usr/bin/env python
> >02: 
> >03: list = []
> >04: while 1:
> >05:     try:
> >06:         stuff = raw_input("[^D to exit]: ")
> >07:     except EOFError:
> >08:         print "got EOF"
> >09:         break
> >10:     list.append(stuff)
> >11: 
> >12: raw_input("bang! ")
> >
> >
> >This works on every Python environment I've tried but cygwin (Win2000
> >-- I haven't got NT to try it):
> >  Python 2.2 (#1, Dec 31 2001, 15:21:18)
> >  [GCC 2.95.3-5 (cygwin special)] on cygwin
> >
> >In this environment, the EOFError is caught.  Then the "bang! " prompt
> >for the raw_input() outside the loop is printed, but immediately the
> >program exits with an uncaught EOFError.
> >
> >This code works with:
> >- ActiveState Python, 2.1 on Win2000
> >- python 2.2 on Mandrake
> >- python 1.5.2 on RedHat
> >and everywhere else, I'm sure.
> >
> >Is this a known problem with cygwin?
> 
> I wouldn't call it a problem with cygwin, I'd call it "relying on
> unspecified behavior".  Ctrl-D is end of file; the ability to read past EOF
> is not promised anywhere in the docs.


Ah, I see.  However, I've seen this technique recommended in several
docs/tutorials.  Also, the docs do promise raw_input() will raise
EOFError when EOF is read.  So, if it is true that I'm relying on
unspecified behavior, my question is: How do I make this code
reliable?  Is there a simple way to clear the EOF indication on stdin?
 I hope there is a good way, because I like this kind of input loop
much better than the "type 'q' to quit" ilk.

Thanks,
David


From mertz at gnosis.cx  Mon Apr  8 02:38:20 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 08 Apr 2002 02:38:20 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk>
 <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk>
Message-ID: <crTs8kKkXg5S092yn@gnosis.cx>         

ralph at inputplus.demon.co.uk (Ralph Corderoy) wrote previously:
|To calculate the major sort key you might do
|    major = can_be_moved(o) + can_be_sized(o) + can_be_stacked(o) +
|        can_be_coloured(o) + can_be_hidden(o)
|as opposed to
|    major = 0
|    if can_be_moved(o):
|        major += 1
|    if can_be_sized(o):
         [...]

Of course, I'm sure I'll be justly condmened for preferring:

    fs = (can_be_moved, can_be_sized, can_be_stacked,
          can_be_coloured, can_be_hidden)
    major = len([1 for f in fs if f(o)]))

to either :-).

Yours, Lulu...


--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From dalke at dalkescientific.com  Sat Apr 13 06:10:03 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 13 Apr 2002 04:10:03 -0600
Subject: Inspecting multiple-inheritance
References: <mailman.1018674026.19083.python-list@python.org>
Message-ID: <a98vv3$bde$1@slb6.atl.mindspring.net>

Terry Hancock:
>I'm trying to understand a portion of Zope, which involves
>highly multiply-inherited objects.  There are several
>instances of methods being overloaded, and so I'm trying
>to figure out "who wins".
  ...
>I know I can find out what methods/attributes an object
>has by using dir() -- but is it possible to figure out
>what arguments they expect, or what superclass they
>came from?

Does this help any?

>>> class Base:
...  def add(self, i):
...   pass
...
>>> class Acid:
...  def add(self, i):
...   pass
...
>>> class Salt(Base, Acid):
...  pass
...
>>> Salt.add
<unbound method Base.add>
>>> salt = Salt()
>>> salt.add
<method Base.add of Salt instance at 1061720>
>>> salt.add.im_class
<class __main__.Base at 1060c00>
>>>

                    Andrew
                    dalke at dalkescientific.com





From db3l at fitlinxx.com  Wed Apr  3 19:53:42 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Apr 2002 19:53:42 -0500
Subject: Threading problem
References: <mailman.1017868749.1550.python-list@python.org>
Message-ID: <uadskwag9.fsf@ctwd0143.fitlinxx.com>

Gustavo Cordova <gcordova at hebmex.com> writes:

> > The Python global interpreter lock requires that even with native
> > threading, only one thread is actually interpreting Python code at any
> > moment in time.  The net result of this is that while threading is
> > excellent for overlapping I/O, it won't efficiently distribute CPU
> > amongst processors.  That's because if you are CPU bound, then you are
> > stuck in Python code execution, and only one processor can be doing
> > that at any given moment in time.
> > 
> 
> This is evil :-(

It's trade-offs.  It may be less than desirable for multi-processor
utilization, but it actually carries quite a few implementation
benefits for the Python core, as well as minimizes the performance
overhead for maintaining native threading support.  In all but SMP
cases, it works quite well (barring non-cooperative extensions).

The only way to change this would be to push all the interlocking down
to the individual data structure/object levels, which would be no
small feat (at least for the current CPython implementation), but one
which I believe gets discussed from time to time.  Wouldn't hold my
breathe though :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From brian at sweetapp.com  Sat Apr  6 19:37:38 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 6 Apr 2002 16:37:38 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
In-Reply-To: <Xns91E8A12A8C2A3JamesLLugojcom@216.148.53.96>
Message-ID: <003e01c1ddcc$6bb81ea0$445d4540@Dell2>

James wrote:
> Unicode is nice. So are all the other features that are being added to
2.x.
> It is that "x" in 2.x is growing large at an alarming rate and these
> "minor" releases are not minor in the colloquial meaning.

Unicode actually appeared in 1.6. Other than that, there weren't many
forward compatibility problems. I think that the backwards compatibility
issue is easily solved: just do your development on the LCD that you
want to support. And maybe have a simple test script that runs your test
suite against all of the Python versions that you have installed (I do
this using a batch file).

> > What is the total amount of time that you have spent resolving
> > compatibility problems?
> 
> Probably a few days and some worried nights. But the context of how
these
> problems cropped up and the unease they imparted were also important.

You worry too much :-) You have to measure your few days against the
productivity gains that Python has handed you.

> As I've said elsewhere, I don't know. I'll look again in not less than
6
> months. Might look at Lua or Ruby to start.

I remember reading some complaints, along similar lines, on the Ruby
list a few months ago. Try doing a google search. I don't know anything
about Lua.

> None are, so the next best thing is to look for ones that have a
> manageable number of extant releases.

I think that you will be stuck with either new or dead languages.

Cheers,
Brian





From ssweeney at idiosync.net  Thu Apr  4 18:05:08 2002
From: ssweeney at idiosync.net (Scott Sweeney)
Date: Thu, 4 Apr 2002 15:05:08 -0800
Subject: ANN: SLiP and SLIDE
In-Reply-To: <3CAC89C7.4E637D76@juno.com>
Message-ID: <PLEFIJEECKLKCPNJFOFAMECDCCAB.ssweeney@idiosync.net>

YAML and SOX use fairly proprietary syntax... I've tried to keep SLiP's
close to Python's so that if you know Python, you pretty much know SLiP.
That way you can also leverage any Python editor as well and get all of the
auto-indenting, coloring, etc.  However, they all serve basically the same
purpose and goals (shorthand for XML), just different techniques.  I'm sure
you could find some limitations in SLiP or YAML or SOX or YAYAML (maybe in
whitespace handling, etc.), just as you would comparing Python to Perl to
Ruby, but these are mostly edge cases and ones that should be overcome
pretty easily by any of these.

I'm afraid SLiP hasn't been around long enough for me to have a more
detailed list of differences than this.  If you do find limitations in SLiP,
I'd love to know about it.

YAML and SOX do look pretty cool though.  I'm looking forward to digging in
to them a bit more.



-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Jeff Sandys
Sent: Thursday, April 04, 2002 9:14 AM
To: python-list at python.org
Subject: Re: ANN: SLiP and SLIDE


jm wrote:
>
> Scott Sweeney <ssweeney at idiosync.net> writes:
> >
> > SLiP 0.1: a "Sorta Like Python" shorthand syntax for XML
> >       http://www.scottsweeney.com/projects/slip
>
> How different is Slip from SOX ?
> http://www.langdale.com.au/SOX/

How are either of these different from YAML?
http://yaml.org
--
http://mail.python.org/mailman/listinfo/python-list





From phd at phd.pp.ru  Wed Apr 17 16:28:23 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 18 Apr 2002 00:28:23 +0400
Subject: [OT] What is Open Source?
In-Reply-To: <899f842.0204170840.29e579ae@posting.google.com>; from anthony_barker@hotmail.com on Wed, Apr 17, 2002 at 09:40:52AM -0700
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <899f842.0204170840.29e579ae@posting.google.com>
Message-ID: <20020418002823.C4068@phd.pp.ru>

On Wed, Apr 17, 2002 at 09:40:52AM -0700, Anthony_Barker wrote:
> For me open source is simply the ability to "look under the hood" of
> any program.
> Almost every other type of human endevor allows this. 

   But not software. This is where the big problem lies. If you look into
M$ "shared source" code, and use a part of the code later, M$ kills you
using copyright and patent laws.
   It is not enough to have the code open. One needs the rights to read, to
use, to copy, to modify, to distribute modified and unmodified versions (so
everyone benefits from the open code). Only Free Software gives you such
freedom and rights.

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




From tim.one at comcast.net  Fri Apr 26 21:42:33 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 21:42:33 -0400
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to
 do math?)
In-Reply-To: <slrnacjga3.2uv.huaiyu@gauss.almadan.ibm.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEGDPBAA.tim.one@comcast.net>

[Huaiyu Zhu]
> The current arrangement between repr and str has some shortcomings:
>
> - nested structures use repr for elements
> - repr is constrained by the desire to be inverse of eval for simple
>   types
> - default repr for instances are optimized for object id, not content
> - str cannot differentiate between 2 and "2"
>
> It appears that another display level is needed, somewhat between repr
> and str.

Except in the details, this is a lot like ssctsoos ("str() special-casing
the snot out of strings").  That's a hypothetical function discussed at
nauseating length several times over the years, at least on Python-Dev, but
not recently.

> Let's call it 'disp' here.  It is intended to be used for
> display in a controlled way, esp for nested structures.
>
> Here's what I think it should do:
>
> - disp(0.05)    -> "0.05", just like str(0.05).
> - disp("abc")   -> "'abc'", just like repr('abc').
> - disp([a,b,c]) -> '['+', '.join(map(disp, [a,b,c]))+']'.
> - disp(obj)     ->  obj.__disp__().

ssctsoos had several differences:

- People using non-7-bit ASCII character sets hate repr(string):
  instead of their native character glyphs, they see a bunch of
  octal or hex escape sequences instead.  So ssctsoos(string)
  *wanted* to act like str(string), but then there's the nasty
  problem you mentioned above:  "2" versus 2.  But you probably
  don't want that 2-bit tail to wag the 8-bit dog.

- ssctsoos(container) (list, tuple, dict) simply passed itself
  down to containees.  The top-level decorations (choice of bracket
  characters and containee seperators) are the same as repr(container)
  (which happens to be identical to str(container) for the builtin
  container types today).

- ssctsoos(user_defined_obj) was the same as str.

In other words, ssctsoos(x) *was* str(x), except that container types
"passed it down" instead of passing repr down, and *something* special
needed to be done for strings.

> Having and additional display method also reduces the need to override
> __repr__, which is the source of very confusing diagnostics when not
> done right (ie, when diagnostics are needed).

I think the real source of the pain is that the interactive prompt shouldn't
use repr.  That works kinda OK for builtin types (although see the "8-bit
string" discussion above), but is usually a poor thing to do for
user-defined types.  People then warp __repr__ to play nice with the
interactive prompt, instead of using it for its intended purpose, and a
world of artificial difficulties follows.

IOW, str() and repr() are both fine for their intended purposes, but the
interactive prompt specifically wants neither as defined today.  What most
people want there is str(), with containers passing str down to containees
instead of repr.  Alas, the niggling 2 vs "2" problem has killed every
attempt to date.





From cliechti at gmx.net  Sat Apr 27 17:41:31 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Apr 2002 23:41:31 +0200
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org>
Message-ID: <Xns91FDF14397530cliechtigmxnet@62.2.16.82>

[posted and mailed]

holger krekel <pyth at devel.trillke.net> wrote in
news:mailman.1019941968.7326.python-list at python.org: 

> On Sat, Apr 27, 2002 at 10:23:26PM +0200, holger krekel wrote:
>> On Sat, Apr 27, 2002 at 03:06:08PM -0500, jepler at unpythonic.net
>> wrote: 
>> > On Sat, Apr 27, 2002 at 09:57:25PM +0200, holger krekel wrote:
>> > > Hello,
>> > > 
>> > > how can you dynamically update the locals() namebindings?
>> > 
>> > You don't.  See the library documentation:
>> >     locals()
>> >      Return a dictionary representing the current local symbol
>> >      table. Warning: The contents of this dictionary should not be
>> >      modified; changes may not affect the values of local variables
>> >      used by the interpreter. 
> 
> ok. i found a nice workaround (if one dare call it that:-)
> 
> class dictlookup(type({})):

use 'dict' here:

class dictlookup(dict):

(or UserDict for older pythons than 2.2)

>     def __getattr__(self, name):
>         return self[name]
> 
> then i can do:
> 
> def func(...,somedict,...):
>     m = dictlookup(somedict)
>     ...
>        m.key1 
>     ...
> 
> which is reasonably short and additionally seems more robust.
> I think it's generally quite nice for dictionaries containing 
> 'strings' as keys which you know in advance and want to access
> without noise. 
> 
>     holger
> 
> 
> 



-- 
Chris <cliechti at gmx.net>



From shalehperry at attbi.com  Thu Apr 18 02:45:46 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 17 Apr 2002 23:45:46 -0700 (PDT)
Subject: Reading from a binary file in a portable way
In-Reply-To: <3cbe6892_11@news.newsgroups.com>
Message-ID: <XFMail.20020417234546.shalehperry@attbi.com>

On 18-Apr-2002 jb wrote:
> 
> I should like to read a (binary) file x.y in a buffer and then do things 
> like getting the 16 bit integer that is at the offset n, n+1 in the file 
> (maybe lowbyte, then higyhbyte.
> 
> Previously I did it only under MS-Windows where for example I could open the 
> file with the flag O_BINARY but this is no longer possible under Linux.
> 
> What would a solution that works under both Unix and MS-Windows look like?
> 

Why can't you open the file in binary mode under Linux/UNIX?

You may also find the 'struct' module useful.




From JamesL at Lugoj.com  Mon Apr  1 18:46:50 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 1 Apr 2002 23:46:50 GMT
Subject: multithreading and Tkinter
References: <uok7s6qqph.fsf@bashful.cdf.toronto.edu> <m3pu1yrz2w.fsf@mira.informatik.hu-berlin.de> <Xns91E2A4FE99EA8JamesLLugojcom@216.148.53.85> <m3u1qvrexs.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Xns91E3A1127FCCFJamesLLugojcom@216.148.53.85>

martin at v.loewis.de (Martin v. Loewis) wrote:
> James Logajan <JamesL at Lugoj.com> writes:
> 
>> Has anyone ever tried using after_idle in a worker thread as a way to 
>> signal a Tkinter mainloop thread? I'm wondering if this proposal is 
>> really thread safe, since technically a worker thread is making a call
>> to Tk (even if it is nothing more than registering a callback) while
>> the mainloop thread may also be in the process of making calls to Tk. 
> 
> That cannot happen. _tkinter maintains a Tcl lock, guaranteeing that
> only one Python thread at any time calls into Tcl. 
> 
> You may wonder how you then can call in to Tcl while another Python
> thread calls tk_mainloop: tk_mainloop does not call a Tcl loop, but
> calls Tcl_DoOneEvent; if there was no event, it sleeps for 20ms
> (giving up the Tcl lock during that time). As a result, other threads
> can call into Tcl while Tcl is idle. If Tcl is active, those other
> threads will block in the Tcl lock.
> 
> People have been complaining about the presence of the Tcl lock, and
> the Sleep call (which means that there might be a 20ms delay between
> an event and the time when processing the even starts). You are
> encouraged to study the source of _tkinter.c, both to get trust in my
> statements, and to suggest improvements to the status quo.

Thanks! My old Linux system still has 1.5.2 so I took a quick look at the 
source code for that version. I _think_ I understand. When compiled 
_without_ thread support, mainloop loops over Tcl_DoOneEvent(0) calls, 
which will process one event or _wait_ for one to process before returning. 
But when compiled _with_ thread support it loops over 
Tcl_DoOneEvent(TCL_DONT_WAIT) calls, around which a Tcl lock has been 
placed (but the Python interpreter lock has been released). A 20ms window 
is opened where the Tcl and Python interpreter locks are opened to allow 
other threads to run unimpeded.

No good alternatives come to mind. I'm not sure if launching a thread 
dedicated to calling into Tcl would work (e.g. all threads making Tcl calls 
are really entering requests into a queue and signalling the dedicated Tcl 
thread to pick up and process their requests). I don't know anything about 
Tcl calls, which limits what I can suggest. It would also be a bit of work 
to implement in any case. I assume others with more knowledge of these 
things have considered all reasonable alternatives.


From skip at pobox.com  Tue Apr 16 21:11:20 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 16 Apr 2002 20:11:20 -0500
Subject: optimization with the "compiler" module
In-Reply-To: <a9icab$phq$1@nntp9.atl.mindspring.net>
References: <mailman.1018990292.9573.python-list@python.org>
        <a9icab$phq$1@nntp9.atl.mindspring.net>
Message-ID: <15548.52152.546290.596813@12-248-41-177.client.attbi.com>

    Andrew> Suppose you have

    Andrew> class Strange:
    Andrew>     def __add__(self, other):
    Andrew>         return self
    Andrew> x = Strange()

    Andrew> Then x+3+7.0 = (x+3)+7.0 = x+7.0 = x

    Andrew> The compiler can't know what 'x' might be so can't fold
    Andrew> constants in this case.  You could if it was '3+7.0+x'.

This is a trivial peephole optimization.  If two constants are summed, they
will appear in the bytecode like

    LOAD_CONST 3
    LOAD_CONST 7.0
    BINARY_ADD

You recognize the pattern, perform the optimization, and move on.

In your example, it would look like

    LOAD_GLOBAL x
    LOAD_CONST 3
    BINARY_ADD
    LOAD_CONST 7.0
    BINARY_ADD

The pattern (const, const, add) never turns up in the instruction stream.

Details here:

    http://manatee.mojam.com/~skip/python/spam7/optimizer.html

Skip




From mertz at gnosis.cx  Mon Apr 22 14:50:09 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 22 Apr 2002 14:50:09 -0400
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
 <mailman.1019299357.11513.python-list@python.org>
 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
 <Xns91F67C40EB0E0lumberjack@216.148.53.83>
 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
 <3CC3A89C.BE2BF65@tundraware.com>
 <mailman.1019491732.12046.python-list@python.org>
 <3CC4536A.81C9B57D@tundraware.com>
Message-ID: <htFx8kKkXQsG092yn@gnosis.cx>         

|> Mathematics as a special branch of Psychology.  Mathematics describes
|> the laws of thought, but at a level of abstraction that applies to any
|> thinking being.

Tim Daneliuk <tundra at tundraware.com> wrote previously:
|no compelling evidence to support the idea that "...thought patterns
|will be constrained by the rules of Mathematics."
|If you take this position, you are arguing that
|*all* human intellectual activitiy like creative action, analysis,
|emotion, need, and so on can, at least in priciple, be described
|mathematically.

This goes further than I need to or intended my remarks to.  I would
certainly not claim that *ALL* human intellectual activity is
mathematical.  There are many other facets of human thought that have
nothing to do with mathematics (e.g. those examples I listed in passing:
Oedipal complex; short/middle/long-term memory; prototyping... lots of
others).

The quantifier sort of goes in the other direction.  Any *BEING* that
thinks about mathematics must think about it in certain ways (the
martians, etc).  It might be that something (or someone, say with
certain brain injuries; some of the cases Oliver Sacks describes are
interesting here) can think in many ways, but does not think about
mathematics.  OK, good enough.  But anything that *does* conceive the
concepts "two", "sum" and "four" *must* think that 2+2=4 (or at least
not conclude its contrary).

I don't think of 2+2=4 as a fact about apples that might be counted,
exchanged, stored, etc.  Of course there is a practical application of
addition to these concrete matters; and quite likely piles of apples are
useful for teaching a child basic arithmetic.  But fundamentally 2+2=4
is a (Kantian) fact about the nature of cognition.  And so is Godel's
incompleteness proof, even if most thinkers never happen to think about
it--if they *were* to think about it, they nature of thought compells
the conclusions of the proof.

Yours, Lulu...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From aleax at aleax.it  Mon Apr 29 03:22:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 07:22:42 GMT
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net> <3CCCD547.6090405@SPAMnwinternet.com>
Message-ID: <6x6z8.30735$8D3.846221@news1.tin.it>

<posted & mailed>

Jim wrote:
        ...
> to have sip and PyQt installed, but I assume you do, since PyKDE2 won't
> build and install without them. Make sure, however, that they all match
> in version (latest is sip3.1, PyQt3.1, PyKDE2-3.1 - the 3.1 is the *sip*
> version *not* the KDE version).

Having KDE 2.2.2, Qt 3.0.3, PyQt 3.1, and Python 2.2.1, I'd like to take
PyKDE for a spin, but can't figure out what to download from where, and/or
how to configure it properly.  All my attempts so far (admittedly I haven't
given it much focuses attention) end up with something like:

checking for Qt version >= 230 and < 233... configure: error: "incompatible 
Qt version"

and indeed I don't have such an old Qt (and can't/won't downgrade).  Is
there a way to do what I'm looking for -- try some PyKDE on an installation
with all the "latest stable releases" of the various other tools, as
listed -- or do I just have to wait for a stable KDE 3 or whatever...?


Thanks,

Alex



From logiplexsoftware at earthlink.net  Wed Apr  3 18:40:23 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 15:40:23 -0800
Subject: wxPython Book ?
In-Reply-To: <11e94203.0204031526.f11bfe6@posting.google.com>
References: <11e94203.0204031526.f11bfe6@posting.google.com>
Message-ID: <20020403154023.053a72a9.logiplexsoftware@earthlink.net>

On 3 Apr 2002 15:26:46 -0800
Jose Rivera wrote:

> Hi...
> Is there a wxPython book available? that gives a depth tutorial?

No.  The best resource for learning is the demo that comes with wxPython. 
After that, the wxWindows manual will make more sense.  You might also look
at the wxPython wiki:

http://wiki.wxpython.org/


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From eric at enthought.com  Mon Apr 15 09:56:21 2002
From: eric at enthought.com (eric)
Date: Mon, 15 Apr 2002 09:56:21 -0400
Subject: python on cygwin
References: <01d801c1e1b8$ac962e40$6b01a8c0@ericlaptop> <20020412121141.GA948@tishler.net>
Message-ID: <056901c1e485$53125c20$6b01a8c0@ericlaptop>

Thanks Jason,

I'll look this over and try to get it running.  Is this something that needs to
be fixed in the python disribution, or something that needs to be fixed in
cygwin?  Any prospect of this eventually working out of the box, or is this
something that will always need tweaking?

thanks,
eric


----- Original Message -----
From: "Jason Tishler" <jason at tishler.net>
To: "eric" <eric at scipy.org>
Cc: <python-list at python.org>
Sent: Friday, April 12, 2002 8:11 AM
Subject: Re: python on cygwin


> Eric,
>
> On Thu, Apr 11, 2002 at 08:26:22PM -0400, eric wrote:
> > I read a few mails about a rebase solution at:
> >
> >     http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
> >
> > However, after building and running this on the percribed files, I get:
> >
> > $ rebase -d -b 0x68000000 -o 0x10000 cygXpm-X4.dll ...
> > [snip]
> > cygXpm-X4.dll: new base = 67fe0000, new size = 20000
> > cygXpm-noX4.dll: new base = 67fc0000, new size = 20000
> > cygbz21.0.dll: new base = 67f90000, new size = 30000
> > cygcrypto.dll: new base = 67ee0000, new size = b0000
> > ReBaseImage() failed with last error = 2
> >
> > Can anyone point out the error in my ways, or point me to a fix?
>
> Replace the argument list above with the cyg*.dll files in *your*
> /usr/bin.  The "last error = 2" error message is caused by a file not
> found condition.  A later version of rebase displays better error messages
> that make this more obvious.
>
> See the following for more (possibly useful) information:
>
>     http://www.zope.org/Members/dgeorgieff/howto_zope_cvs_on_cygwin/
>
> Otherwise, just follow Michael's suggestion and build _socket statically.
>
> I will try to release a pre-built 2.2.1 ASAP.
>
> Jason
>






From jeff at ccvcorp.com  Fri Apr 19 13:57:19 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 10:57:19 -0700
Subject: PEP for an unrestricted __init__ ?!
References: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com> <mailman.1019182657.12773.python-list@python.org>
Message-ID: <MPG.1729fa43d442b9779896b6@news.nwlink.com>

In article <mailman.1019182657.12773.python-list at python.org>, 
pyth at devel.trillke.net says...

> But __new__ or __init__ still restrict you to return instances
> of their respective class? This does sound rather sane,
> but i imagined some other uses for an unrestricted __init__ (or __new__)
> such as making functions-objects a special case of classes:
> 
>    def func(args):
>       "compute some stuff
>       return result
> and
>    class func:
>        def __init__(self, args):
>           "compute some stuff"
>           return result

Well, you can already accomplish something like what you are 
asking for, by defining a __call__() method for the class.  That 
lets class instances be used as callable objects -- using 
function syntax on them.

>>> class foo:
... 	def __init__(self, food):
... 		self.food = food
... 	def __call__(self):
... 		print "Spam and %s!" % self.food
... 
>>> eggs = foo('eggs')
>>> beans = foo('baked beans')
>>> eggs()
Spam and eggs!
>>> beans()
Spam and baked beans!
>>> beans.food = 'bacon'
>>> beans()
Spam and bacon!
>>> 

Obviously, you can do any amount of processing and checking that 
you want in your __call__() method.  

So most of what you want to do, already exists.  :)  (Guido's 
time machine strikes again!)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From aleax at aleax.it  Wed Apr 10 09:54:03 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 13:54:03 GMT
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool type)
References: <mailman.1018392859.8707.python-list@python.org> <OyUs8.2398$b62.59604@news1.tin.it> <mailman.1018443200.3936.python-list@python.org>
Message-ID: <%tXs8.3035$b62.74386@news1.tin.it>

Gerhard H?ring wrote:

> * Alex Martelli <aleax at aleax.it> [2002-04-10 10:34 +0000]:
>> (let's PLEASE forget that unnamed language where you must use + only
>> to add integers and +. to add floats... have they changed it yet?-).
> 
> No. Why is this so bad, except for overloading operators in subclassed
> numeric types? Dunno if Ocaml can even do that.

It is "so bad" because it loses polymorphism.  Polymorphism being good, 
losing it wantonly is ungood (there IS no ill-effect to a polymorhpic +,
of course, as at least Standard ML amply testifies -- let's not even
get into Haskell and its wonderful typeclass-based polymorphism).


Alex



From aleax at aleax.it  Tue Apr  2 10:46:22 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 15:46:22 GMT
Subject: The Python Way
References: <5515F24370BCB753.DA7CF91464346EF6.E3A577CFA6882D82@lp.airnews.net> <jjb5-B81262.13422827032002@newsstand.cit.cornell.edu> <mailman.1017263072.31748.python-list@python.org> <a7tll9$k2v$1@panix1.panix.com>
Message-ID: <inkq8.52089$S52.1780087@news2.tin.it>

Aahz wrote:

> In article <mailman.1017263072.31748.python-list at python.org>,
> Skip Montanaro  <skip at pobox.com> wrote:
>>
>>Aahz is always admonishing us to just use Queue.Queue.  Maybe it's time
>>I paid a bit closer attention to what he's been saying...
> 
> <grin>  Thank Tim; he bashed it into my head, and all I'm doing is
> carrying the torch.

Just so all know this fire keeps burning -- I have invested a few precious
pages of the forthcoming Nutshell's chapter on threads (yes I _am_ still
writing the Nutshell -- when I'm not editing the Cookbook or working or
reading c.l.p, and I swear I'll drop the latter incredibly enticing but
timeconsuming activity again veryveryvery soon...:-) to sketching a
general threaded-program architecture very firmly based on Queues (with
a couple reusable classes wrapping up a thread object and a couple of
reference to Queues).  [If any true threading expert such as Aahz or Tim
should want to have a look, I'll happily send a sneak preview of the
chapter, of course -- just get in touch, you do know my email:-)].


Alex



From emile at fenx.com  Wed Apr 10 20:02:18 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 11 Apr 2002 00:02:18 GMT
Subject: mac-adress with python (windows)
References: <3cb493ed.40186328@news.easynews.net>
Message-ID: <eo4t8.206217$Yv2.67449@rwcrnsc54>

Marcus Stojek
> Hi,
> following some links from this and other groups I had to
> fight through pages and pages of C code and loooong
> threads. Then I found the following example in my
> win32com docu:
>
> >>>import pythoncom
> >>>print pythoncom.CreateGuid()
>
> <iid:{ECD43800-4CB9-11D6-ABD9-0050DA4545E7}>
>
> with the last six bytes being the ethernet card ID.
>
> Can anybody tell me if this works for all Window versions?
> (I can only test for Win NT 4 SP6)
>


It doesn't work here on win2k.

>>> pythoncom.CreateGuid()
<iid:{70630B60-CE9D-4405-BDC6-AADC8F89DE0E}>

C:\>ipconfig /all

Windows 2000 IP Configuration
Ethernet adapter Local Area Connection:
        Physical Address. . . . . . . . . : 00-20-78-05-4F-92

See
http://groups.google.com/groups?selm=NJ0E4.5818%24V56.23602%40news-serve
r.bigpond.net.au for Mark Hammonds comments and solution on this.

--

Emile van Sebille
emile at fenx.com

---------



From phr-n2002a at nightsong.com  Tue Apr  2 22:05:22 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 19:05:22 -0800
Subject: verifying ssl certificates
References: <3CA88D15.140053DC@ecn.purdue.edu> <7x7knrdv9k.fsf@ruckus.brouhaha.com> <3CA9C2E0.EA8A2B15@ecn.purdue.edu>
Message-ID: <7xu1qt4h3h.fsf@ruckus.brouhaha.com>

"Christopher N. Deckard" <cnd at ecn.purdue.edu> writes:
> In other words:
> 
> import socket
> 
> cert = open('my_cert')
> s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> s.connect(('host', 'port'))
> ss = ssl(s, cert, None)
> 
> That last line being the key.  Does that do the certificate
> validation?

I'm not familiar with how the socket module does ssl, but if ssl() is
supposed to do an SSL session negotiation, then yes, it should check
the cert.


From claird at starbase.neosoft.com  Tue Apr 23 13:57:23 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 23 Apr 2002 12:57:23 -0500
Subject: Python & autoftp
References: <53b7d218.0204230720.686e473f@posting.google.com> <aa42gi$6jg$1@nntp6.u.washington.edu>
Message-ID: <FB7A2771C2A3A23E.9E0965D2C89F8A9A.D356B4DD4886393B@lp.airnews.net>

In article <aa42gi$6jg$1 at nntp6.u.washington.edu>,
Donn Cave <donn at u.washington.edu> wrote:
>Quoth davies at jkevin.freeserve.co.uk (kevin):
>| I'm pretty new to Python and I would like to develop a script in
>| Python to automatically FTP. I have already used bash and expect and
>| now would like to replace bash entirely with Python.
>|
>| Is there any merit in replacing bash with Python??
>
>In general, each is superior in its own domain of application.
>More specifically, if your program works fine now, it probably
>won't work any better in Python, but you may learn something
>about Python and UNIX internals while you're at it.
>
>| Can anyone point me in the right direction.
>
>The expect part is probably the hardest.  I assume you mean to
>replace expect too, not just bash.  Look at openpty and forkpty
>in the posix ("os") module, and see if you can get them to work
>in a little test application (if you even have them.)  There
>are also more complete expect solutions, if that's interesting
>you might look in the Vaults of Parnassus exit from www.python.org.
>
>Alternatively, whatever you're doing might be done more easily
>by replacing ftp itself with ftplib module.  That will make the
>expect part unnecessary.
>
>	Donn Cave, donn at u.washington.edu

In the year 2002, what I emphasize the most is that
the Expect part is unnecessary.  It has been, in fact,
for about a decade; even most vanilla shell scripts
which do FTP work don't really need the PTY-management
of Expect <URL: http://
starbase.neosoft.com/~claird/comp.unix.programmer/ftp_automation.html >.

If you want to work with Python for automation of FTP
operations, I STRONGLY recommend the ftplib module.
That's much the best way to advance.

I agree with the rest of what Donn's written, of
course.
-- 

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


From mark at mceahern.com  Wed Apr 24 14:36:26 2002
From: mark at mceahern.com (Mark McEahern)
Date: Wed, 24 Apr 2002 13:36:26 -0500
Subject: mxCGIPython
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEHFKMAA.mark@mceahern.com>

Hi, I found this wonderful site:

  http://www.egenix.com/files/python/mxCGIPython.html

However, when I try to gunzip this file, I get this message:

  $ gunzip -f cgipython-2.2-Linux-2.2.19-i686-with-debian-2.2.gz
  gunzip: cgipython-2.2-Linux-2.2.19-i686-with-debian-2.2.gz: not in gzip
format

This happens both on Cygwin and on Linux.  Any ideas?

Thanks,

// mark





From martin at v.loewis.de  Tue Apr  9 17:34:02 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Apr 2002 23:34:02 +0200
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de>
Message-ID: <m3sn644kvp.fsf@mira.informatik.hu-berlin.de>

Thomas Guettler <pan-newsreader at thomas-guettler.de> writes:

> Up to now I was happy with emacs because I never used
> to programm it with lisp. Now I need to and find it difficult.
> 
> Is there an emacs-like editor which is scriptable with python?

You might want to try pymacs:

http://www.iro.umontreal.ca/~pinard/pymacs/

Regards,
Martin


From aahz at pythoncraft.com  Fri Apr 12 09:41:12 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Apr 2002 09:41:12 -0400
Subject: PEP 282: A Logging System
References: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru> <mailman.1018538209.18235.python-list@python.org> <a95c7k$801$1@panix1.panix.com> <2e37dc1.0204120001.11ee98b0@posting.google.com>
Message-ID: <a96o5o$j4h$1@panix1.panix.com>

In article <2e37dc1.0204120001.11ee98b0 at posting.google.com>,
Vinay Sajip <vinay_sajip at yahoo.co.uk> wrote:
>aahz at pythoncraft.com (Aahz) wrote in message news:<a95c7k$801$1 at panix1.panix.com>...
>>
>> Your main difficulty with thread safety is WRT I/O.  How do you plan to
>> handle that?
>
>I/O is the province of the handlers, and I expect to make them
>thread-safe with respect to the underlying I/O through the use of
>thread locks. Of course this applies only to the handlers I provide as
>part of the distribution; user-defined handlers will need to make
>their own provision.

Yup.  You might want to work on some metaclass hacking or something that
permits locks to be automatically applied, or perhaps checks for a
locking API on registered handlers.

>The code released up till now does not have any locking at all.
>Empirically, a simple test harness I put together (just to see how the
>current code does) seems to run fine (eyeball-wise) on Python
>2.1.1/Windows and Python 1.5.2/Windows and Linux, though with 1.5.2 I
>had to remove calls to inspect. The test was logging to a FileHandler
>and a DatagramHandler. I was a little surprised at first not to see
>any problems, might this be due to the GIL? I don't know enough in
>detail about how the GIL works, but given that these handlers all work
>through C objects I would expect the GIL to play a role.

Yup.  Generally speaking, if you're making one-and-only-one I/O call in
your handler, you're mostly safe.  But don't count on it.

Your code looks fine to an eyeball test.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From rajat at cs.utexas.edu  Thu Apr 25 16:18:39 2002
From: rajat at cs.utexas.edu (Rajat Chopra)
Date: 25 Apr 2002 13:18:39 -0700
Subject: CallDLL Package
Message-ID: <d9773b18.0204251218.3e96bf7@posting.google.com>

Hi, 

I am attempting to build a python library (pyd file) for the CallDLL
package (http://www.nightmare.com/~rushing/dynwin/index.html) - which
allows you to dynamically load and call any function in a DLL - that
will run on the newest version of Python (v2.2).

I am using MinGW32 (particularly GCC and the BINUTILS package) to
rebuild the Python library file. When I run the script to rebuild the
library file, however,  I continuously get the error of "undefined
reference to LoadLibraryA, FreeLibrary, GetModuleHandleA, and
GetProcAddress functions".

I understand that these are Windows specific functions that are
contained in Kernel32.dll. However, as this is a critical DLL, I
believe it is directly linked to in the normal C compile. In fact,
when I did a verbose run of the script, I indeed saw that kernel32
(along with several other DLLs) are linked to automatically by GCC.

I am wondering what could cause this error. It is strange because when
I link to the Python DLL (python22.dll), the Linker resolves all the
function names referenced within that DLL without a problem. I am
placing the script I am running to build my python library below:

 -------
|SCRIPT |
 -------
gcc -I"C:\Program Files\Python\include" -O2 -c calldllmodule.c -o
calldllmodule.o

dllwrap --dllname calldll.pyd --driver-name gcc --def calldll.def -o
calldll.pyd calldllmodule.o -s --entry _DllMain at 12
--target=i386-mingw32
-L"C:\WINNT\system32" -lpython22 

(DLLWrap is a utility that is included in BINUTILS. I believe it is
equivalent to DLLTOOL).

Also, the verbose script shows the following libraries being linked to
by GCC (notice both python22 & kernel32 are linked to):
-python22 -lmingw32 -lgcc -lmoldname -lmsvcrt -luser32 -lkernel32
-ladvapi32
-lshell32 

Thank you in advance.

-Rajat


From loewis at informatik.hu-berlin.de  Thu Apr 18 10:50:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 16:50:42 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <Pine.GSO.4.40.0204181304460.24683-100000@neelix>
References: <Pine.GSO.4.40.0204181304460.24683-100000@neelix>
Message-ID: <j4g01t6oxp.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> I can't yet: I don't know which properties are reversed:
>  * Modules included in the config.c file
>  * List of objects to be added to library archive
>  * linker options added to linker options
>  * Rules added to Makefile
>  * Names collected in SHAREDMODS variable.

The state of 'will the following modules be static or dynamic'. In a
Setup file, you can have

list static modules here
*shared*
list shared modules here
*static*
list static modules here
*shared*
list shared modules here
*static*
list static modules here
...

> > Use 'gcc --print-prog-name=ld' to find out which ld binary gcc uses.
> /usr/local/sparc-sun-solaris2.7/bin/ld
> /usr/local/sparc-sun-solaris2.7/bin/ld --version gives
> 
> GNU ld version 2.12
> Copyright 2002 Free Software Foundation, Inc.

Ok; that explains it: configure does

   $CC -Xlinker -V 2>&1 | grep BFD

expecting that the linker output reads like

GNU ld version 2.11.2 (with BFD 2.11.2)
  Supported emulations:
   elf32_sparc
   elf64_sparc

Apparently, in binutils 2.12, GNU ld stops printing BFD in its
output. The proper fix is to invoke --help instead, and to see whether
export-dynamic appears in the output.

> I think I will delay doing that for now, because most of the tests
> worked without using GCC.  Another post suggests that this is not common
> to Solaris7, (referring to pre-releases of Python-2.2.1).  Therefore I
> may not really be helping anyone else by exploring that.

It turns out that you are the first one to use binutils 2.12 on
Solaris, trying to build Python with that. So you are likely helping
others; Python 2.2.2 (and perhaps 2.1.4) will have this problem
corrected.

Regards,
Martin





From martin at v.loewis.de  Sat Apr 20 19:24:40 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Apr 2002 01:24:40 +0200
Subject: Parsing indention based language?
References: <a9sf96$ga6$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <m33cxqar7r.fsf@mira.informatik.hu-berlin.de>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:

> I was wondering if anybody knows any elegant approaches.

I recommend to use the same approach that Python uses: process
indentation on the lexical level. That is, the lexer ought to create
NEWLINE, INDENT and DEDENT tokens. Then you can describe the syntax
with a traditional grammar, and apply traditional parser/code
generator technology.

HTH,
Martin



From lenssen at hitnet.rwth-aachen.de  Mon Apr 15 18:09:07 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 00:09:07 +0200
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE> <m3wuv8iras.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9fj34$3mi$1@nets3.rz.RWTH-Aachen.DE>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3wuv8iras.fsf at mira.informatik.hu-berlin.de...
> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
>
> > Is it possible to use XPath in Python, if so are there any small
samples?
>
> Yes, there is an XPath library in 4Suite, which is been partially
> ported to PyXML 0.7.
>..

Yes, I've been trying to reach their site, but various links didn't return
anything.
What do I need to do? I already installed PyXML.
Thanks...




From pc_olsen at yahoo.com  Tue Apr  9 16:23:24 2002
From: pc_olsen at yahoo.com (Peter Olsen)
Date: 9 Apr 2002 13:23:24 -0700
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org>
Message-ID: <12bb92e6.0204091223.738e0798@posting.google.com>

Itamar Shtull-Trauring <twisted at itamarst.org> wrote in message news:<mailman.1017656355.5887.python-list at python.org>...

> "Software pirates not only steal from the companies that make the software,
> but with less money for research and development of new software, all users
> are hurt.  That's why all software piracy - even one copy you make for a
> friend, is illegal."
> 
> Software piracy is a serious crime, and one that the Open Source community
> has been remarkably lax in pursuing and protecting against.  This is why
> Twisted Matrix Laboratories is taking the forefront in Open Source software
> registration technology.
> 
> In order to do your part to prevent the tragedy of unregistered, unlicensed
> software, all you need to do is go to http://www.twistedmatrix.com/license
> and enter your user information to obtain a license key.  You can provide us
> with as much or as little information as you like!
> 
> For more information on how Twisted helps you run licensed, unpirated
> Open-Source software, visit:
> http://twistedmatrix.com/documents/howto/register
> 
...
> 
> For more information, visit http://www.twistedmatrix.com, or join the list
> at http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python

While the small amount of software I have written has been release 
under terms more liberal than those of the GPL (and the license
appears to have disappeared), I see absolutely no conflict between 
"Open Source" and software licensing, indeed, I don't think I see
much conflict between licensing and the GPL.

There are lots of good reasons to license open source software.

I might want to know who was using a binary distribution (perhaps in 
a commercial environment).  I might want make sure that users had 
the latest and most-bug-free version. I might want to insure that
people without the skill to review the source and compile it for
themselves weren't getting binaries munged by some middleman.

I even think that these uses might be laudable.  

I have seen several packages with license terms stating that 
users could not further distribute the program, but that it 
was freely available from the author's website.  What's wrong
that?

I have also seen cases in which a buggy early version of a program
was much more widely distributed than a later, bug-free one.  I like
to think that the code I write today is better than the code I wrote
years ago --- and I'd just as soon not have my reputation rest on 
the code I wrote back then.

What's wrong with addressing those problems with a license that 
prohibits users from further distributing either binary or
source code while making them both freely available from the
author's website?

Eric Raymond has written that one of the motivations for building
Open Source software is to gain the satisfaction of public 
accomplishment.  What better way to know that your accomplishment
is recognized than to count the number of people who took the time
to register their use of it?

I think that, for my purposes, licensing is not worth my effort or
my users' time.  But if you want to license your Open Source software,
be my guest.

Peter Olsen, P.E, Ae.E.
pc_olsen at yahoo.com


From cliechti at gmx.net  Thu Apr 11 15:36:07 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Apr 2002 21:36:07 +0200
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it>
Message-ID: <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82>

Alex Martelli <aleax at aleax.it> wrote in 
news:kTft8.6500$b62.159068 at news1.tin.it:
> Max M wrote:
>         ...
>> def newJoin(self, items):
>>      return self.join([str(i) for i in items])
>> 
>> Is there any case where it isn't applicable at all?
> 
> Could be a disaster if any item was Unicode:
> 
>>>> print ''.join(('mapp','erch',u'\u00e8')).encode('latin-1')
> mapperch?
> 
> vs:
> 
>>>> print ''.join([str(x) for x in 
> ('mapp','erch',u'\u00e8')]).encode('latin-1')
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> UnicodeError: ASCII encoding error: ordinal not in range(128)
>>>>
> 
> but presumably you could have something subtler that used
> str() or unicode() as appropriate (maybe not trivial to
> determine what's "appropriate" in each given case).

how about a formater argument?

>>> class mystr(str):
... 	def join(self, elements, formater = None):
... 		if formater is not None:
... 			elements = map(formater, elements)
... 		return str.join(self,elements)
... 
>>> mystr(',').join(range(10),str)
'0,1,2,3,4,5,6,7,8,9'

so the user can use 'unicode', 'str' or 'curry(unicode, encoding="latin1")'
etc.

chris

PS: it would be realy cool if curry was a builtin :-)


-- 
Chris <cliechti at gmx.net>



From jb at cascade-sys.com  Tue Apr 30 17:36:58 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 14:36:58 -0700
Subject: Newbie: finding the key/index of the min/max element
References: <200204301620.00255.bhan@andrew.cmu.edu>
Message-ID: <3CCF0E7A.E5C3531E@cascade-sys.com>

Benjamin Han wrote:

> In my quest of becoming a "Pythonistas" (where does this word come from BTW?),

The term derrives from the Sandinistas in Nicaragua, the Communist rebels who
fought against the US-installed conservative Somoza dictatorship (and others).
The suffix is sometimes used to refere to any contrarian group, and in cases
simply any group.

> here comes my 2nd trivial question: how do you find the key/index of the
> max/min element in a sequence/map? I know you can call max(l) but this only
> gives you the element itself, not the key. I also know it's not that hard to
> code it up the hard way (for k in d.keys(): ... # do comparison), but is
> there a more elegant way of doing this?

I have trouble thinking of a clearer or more efficient way than simply iterating
on the list.

There's a lot of other ways to do it if you goal is to get rid of the for loop,
but I think they're all less efficient and I'm not sure you could say any are more
elegant.

E.g., you could say

    x = min( list )
    minkey = list.index( x )

and do it in "two steps".  But this actually requires on average 1.5 comparisons
on the list, one to find the min and however many checks it takes to find that
value again.

Then too, for small lists performance need not be a consideration.  However, I
personally would avoid obfuscating the matter with auxilliary lists or data
structures, even though it's insanely easy to do.

Sometimes a cigar is simply a cigar.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From jeff at ccvcorp.com  Wed Apr 10 18:52:51 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 10 Apr 2002 15:52:51 -0700
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu> <Xns91EC96C0C8E9CRASXnewsDFE1@130.133.1.4>
Message-ID: <MPG.171e61f8eb76328a9896aa@news.nwlink.com>

In article <Xns91EC96C0C8E9CRASXnewsDFE1 at 130.133.1.4>, 
starx at pacbell.net says...
> Joel Bender || Wed 10 Apr 2002 01:35:10p:
> 
> > I would like to process the elements of a list that match a condition, 
> > then remove them.  I'll use integers in my example, but my list items 
> > are actually objects.
> 
> Two approaches come to mind, both using while instead (Because we need 
> to use the number instead of just the item):
> 
> i = 0
> while i < len(list)
>     	if list[i] < 5:
>     	    	print list[i]
>     	    	del list[i]
>     	i += 1

This will *not* work as expected.

When you del list[I], then list[I+1] becomes list[I].  You then 
increment I, and get the next item... except that the item you 
get is the one that *used* to be list[I+2].  You skip an item 
each time you delete an item.

>>> mylist = range(5)
>>> mylist
[0, 1, 2, 3, 4]
>>> I = 0
>>> while I < len(mylist):
... 	if mylist[I] < 3:
... 		print mylist[I]
... 		del mylist[I]
... 	I += 1
... 
0
2
>>> mylist
[1, 3, 4]
>>> # note that 1 got skipped!

This is why it's always dangerous to modify the list you're 
iterating over.  It can be done safely (see Daniel's back-to-
front method), but it takes some work.

(As a side note, don't use list as an identifier name, since 
you're masking the builtin function list() ... )

-- 

Jeff Shannon
Technician/Programmer
Credit International


From bokr at oz.net  Wed Apr 10 16:40:15 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 20:40:15 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <1018450381.103820@cswreg.cos.agilent.com> <yu991ydnmtzc.fsf@europa.research.att.com>
Message-ID: <a927vf$euh$0@216.39.172.122>

On Wed, 10 Apr 2002 15:52:23 GMT, Andrew Koenig <ark at research.att.com> wrote:

>Greg> Andrew Koenig (ark at research.att.com) wrote:
>
>Greg> : My motivation is to be able to write something that is analogous
>Greg> : to call-by-reference in C++.
>
>Greg> *Not* having call-by-reference means that you can read code without
>Greg> wondering whether a particular function call is or is not by reference.  I
>Greg> consider this a blessing, and I'd like to think that the developers of
>Greg> Lisp, C, Java, and Python feel the same way.
>
>Of course, C arrays behave as if they are pass-by-reference
>(only!), even though we both know that's not strictly true
>under the skin.
>
>Anyway, I would be quite content even if the references were
>explicit at the point of call, which would meet your objection.
>And there is even a straightforward way to achieve that state
>now:  Pass a pair of get/set functions instead of the variable
>itself.
>
>So in fact there is a way to do what I want; it's just that it's
>more verbose than I'd like.
>
What about x() as the spelling of get-x's-value and x(v) as the
spelling of set-x's-value-to-v-and-return-v?

That's what I suggested in my other post. I.e., the get/set pair
is implemented by the __call__ method, for conciseness.

Regards,
Bengt Richter


From mhammond at skippinet.com.au  Wed Apr 17 22:46:25 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 18 Apr 2002 02:46:25 GMT
Subject: ANN: Python sidebar for Mozilla
Message-ID: <3CBE33D1.4070902@skippinet.com.au>

Hi all,
   I have written a sidebar for the Mozilla browser.  If you are not 
using Mozilla or Netscape 6, then please ignore this message.

The sidebar sits with the other Mozilla sidebars.  You can jump to bugs 
or PEPs, and can perform some very nice searching.  You can search the 
Python mailing list archive, Python web resources, or the Python 
documentation index.

For further information, including installation instructions, see 
http://starship.python.net/crew/mhammond/mozilla

Mark.



From ak at silmarill.org  Mon Apr 22 20:25:44 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 23 Apr 2002 00:25:44 GMT
Subject: Twisted ircd is written in python
Message-ID: <slrnac9ak3.fho.ak@ak.silmarill.org>

google for twisted matrix


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From s328940 at student.uq.edu.au  Sun Apr 28 16:27:27 2002
From: s328940 at student.uq.edu.au (Alejandro Dubrovsky)
Date: Mon, 29 Apr 2002 06:27:27 +1000
Subject: how to pickle a lambda?
Message-ID: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au>

hi,
i've got a class that defines something like the following:

def t(self,a,b):
	...

def h(self):
	self.thelambdafunctiontobepickled = lambda x: self.t(x,4)

is there any simple way of pickling that lambda function above so that
when i unpickle it will be set to the right function?

Thanks,
Alejandro


From justin at iago.org  Tue Apr 16 13:35:51 2002
From: justin at iago.org (Justin Sheehy)
Date: Tue, 16 Apr 2002 13:35:51 -0400
Subject: python threading
In-Reply-To: <8MVu8.26112$vF6.828988@news2.tin.it> (Alex Martelli's message
 of "Tue, 16 Apr 2002 13:35:00 GMT")
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>
	<KcUu8.22993$Ji7.1840@atlpnn01.usenetserver.com>
	<3cbc16f7$0$71662$e4fe514c@dreader4.news.xs4all.nl>
	<8MVu8.26112$vF6.828988@news2.tin.it>
Message-ID: <x7elhfpmvc.fsf@ra.iago.org>

Alex Martelli <aleax at aleax.it> writes:

> Right!  Queue is THE way to go.  Now, life would be perfect if threads
> generally devoted to such tasks as GUI's or networking didn't have to
> _periodically poll_ to check if there's something new on the Queue for
> them (with a timeout in select, a Tkinter .after callback, a QTimer
> instance in Qt, etc etc).  I.e., if Queue had some way to "jig" the
> waiting thread out of its waiting state, when new stuff is posted to
> the Queue, in such a way that the thread knows it must check its incoming 
> Queue. 

Unless I'm misunderstanding the question, threading.Event is the answer.

I've wrapped Queues with Events before, and it makes things a little
nicer.  It is quite straightforward to make a Queue that exposes a
get method that will simply block until an Event is set by another
thread or an optional timeout has elapsed.

It doesn't solve the whole problem, but waiting on an Event is much
nicer than a get/sleep loop.

-Justin

 





From bokr at oz.net  Mon Apr 15 13:59:45 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 17:59:45 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <m3d6x1mqbu.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9f4eh$pjp$0@216.39.172.122>

On 15 Apr 2002 08:24:53 +0200, martin at v.loewis.de (Martin v. Loewis) wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> If hex long representation were interpreted to represent actual twos
>> complement bits, with the top hex digit 8-bit being extended indefinitely
>> to the left, you could have hex(-1L) be represented as 0xFL (which would
>> require that hex(15L) be represented as 0x0FL).
>
>So you would you represent +7 and -7 in hex?
   ^--I assume missing 'how' ;-)

newhex(+7) => 0x7 # implied extension of msb => 0x000...007
newhex(-7) => 0x9 # implied extension of msb => 0xfff...ff9

I realize there would be code breakage using 0x, since that, e.g.,
changes the current meaning of 0x9. So I thought to introduce
an alternate '0x' hex prefix in the form of '0h'.

Regards,
Bengt Richter


From cliechti at gmx.net  Mon Apr  8 16:13:45 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 8 Apr 2002 21:13:45 +0100
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com> <Xns91E9B2CEE1835cliechtigmxnet@62.2.16.82> <d7fcf8b4.0204071740.7edb1acc@posting.google.com>
Message-ID: <Xns91EAD820EFAB5cliechtigmxnet@62.2.16.82>

samschul at pacbell.net (Samuel Schulenburg) wrote in
news:d7fcf8b4.0204071740.7edb1acc at posting.google.com: 
> Under windows the Pythonwin ide has built in windows support for
> printing either the interactive windows, or any script file window.

yes i know, sorry for my bad english... i meant the calculated output of a 
script, not the script itself. :-)

> 
> Sam Schulenburg
> 
> Chris Liechti <cliechti at gmx.net> wrote in message
> news:<Xns91E9B2CEE1835cliechtigmxnet at 62.2.16.82>... 
>> jimw1956 at cs.comnojunk (Jim Williams) wrote in
>> news:20020407103654.19127.00002640 at mb-fm.news.cs.com: 
>> 
>> > Hi all...I'm new to Python and really enjoying it so far, but I am
>> > somewhat perplexed that as far as I have seen it has no facilities
>> > for printing to a printer?  Surely this is possible. Someone please
>> > enlighten me...I have done a good bit of research and searching and
>> > can find no reference to how to do this. 
>> 
>> one neat solution is to use Reportlab to generate a PDF and you'll
>> find an Acrobat reader on all big platforms today.
>> 
>> you can also look at some GUI systems. e.g. wxPython has some
>> printing support etc. but when i need a printed output from a script,
>> i generate a PDF.
>> 
>> chris



-- 
Chris <cliechti at gmx.net>



From andreas at andreas-jung.com  Thu Apr  4 19:03:38 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Thu, 4 Apr 2002 19:03:38 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com> <slrnaaphdc.39o.quinn@hurl.ugcs.caltech.edu> <200204042338.XAA12847@crown.off.ekorp.com>
Message-ID: <01c701c1dc35$5a2312f0$02010a0a@suxlap>

----- Original Message -----
From: "Richard Jones" <rjones at ekit-inc.com>
To: "Quinn Dunkan" <quinn at hurl.ugcs.caltech.edu>; <python-list at python.org>
Sent: Thursday, April 04, 2002 18:38
Subject: Re: PEP 287: reStructuredText Standard Docstring Format


> On Fri, 5 Apr 2002 07:29, Quinn Dunkan wrote:
> > On Wed, 03 Apr 2002 16:00:42 GMT, Terry Reedy <tejarex at yahoo.com> wrote:
> > >Just curious: are you willing to (consider, at least) switching to
> > >ReST, or are we going to end up with two similar standards used by
> > >different parts of the Python community?  IE, do Goodger's arguments
> > >about why the differences from ST are improvements cut any ice with
> > >you?
> >
> > I'm not speaking for Max, but my answer is "no."
>
> I'm all for it, and I hope the ZC guys are too. If they don't I'll just
have
> to write my own Zope product that wraps the ReST implementation. It's just
> too much of an improvement over the StructuredTexTNG implementation to
ignore.


Backward compatibilty is a major issue for inclusion into Zope. I still have
not
figured how to use it. Are there generators for HTML available yet?

-aj





From eppstein at ics.uci.edu  Tue Apr 16 02:36:19 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 23:36:19 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de> <eppstein-922653.14382815042002@news.service.uci.edu> <m3ads4maa1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <eppstein-16AAA5.23361915042002@news.service.uci.edu>

In article <m3ads4maa1.fsf at mira.informatik.hu-berlin.de>,
 martin at v.loewis.de (Martin v. Loewis) wrote:

> Unfortunately, you cannot represent such a number in hex, atleast not
> under any of the usual conventions: the convention is that you can
> strip infinitely many leading zeroes, but stripping off infinitely many
> leading ones is confusing.

I don't understand why not.

Exactly the same difficulty applies to decimal representations (you can 
strip off leading 0's but not leading 9's) but we have no trouble 
representing negative numbers in decimal: just write a minus sign 
followed by the decimal string for the absolute value of the number.

Similarly, the obvious and only way of representing a negative number in 
hex is to write a minus sign followed by the hex string for the absolute 
value of the number.

The "usual" way of representing 32-bit values in hex, ffff0123 etc, has 
nothing to do with negative numbers, it is a way of representing 
unsigned values.  Since Python doesn't have unsigned values, we 
shouldn't do things that way.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From aleax at aleax.it  Wed Apr  3 01:27:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 06:27:58 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA5777.66E7C2D5@python.org> <kupu1hple7.fsf@lasipalatsi.fi> <eppstein-CC3FF6.18410402042002@news.service.uci.edu>
Message-ID: <Ohxq8.45137$pT1.1283197@news1.tin.it>

David Eppstein wrote:
        ...
> I can't think of a language change that has made me very unhappy, but

I can only think of one (print>>bah)...

> There's plenty of change I'm neutral or ignorant of, like most of the
> object-oriented infrastructure stuff.  What is staticmethod etc anyway?

class Peep:
    def anyfunc(): print "note, no 'self'"
    anyfunc = staticmethod(anyfunc)

p = Peep()

p.anyfunc()
Peep.anyfunc()

prints "note, no 'self'" twice.  Just a peculiar way of packaging functions
so they can be attributes of a class without thereby becoming unbound
methods when accessed.  classmethod is more powerful, but according
to the tutorial the inability to upcall naturally from it means Guido knows
of no use cases for it and is thinking about taking it away again (maybe
some Smalltalkers will put together an impressive array of practical
use cases... let's hope...!).  It works similarly:

class Qeep:
    def anyfunc(cls): print "called on class", cls.__name__
    anyfunc = classmethod(anyfunc)

class Reep(Qeep): pass

q = Qeep()
r = Reep()

q.anyfunc()
Qeep.anyfunc()
r.anyfunc()
Reep.anyfunc()

prints "called on class Qeep" twice, then "called in class Reep" twice.


> So that seems like a very good ratio to me.

A great ratio, yes.


Alex



From noah at noah.org  Mon Apr 15 00:28:08 2002
From: noah at noah.org (Noah)
Date: Sun, 14 Apr 2002 21:28:08 -0700
Subject: split() off last substring
In-Reply-To: <3CBA41CB.27DCA084@uniserve.com>
Message-ID: <PAENINBOOODFDEPOKPKCKEDJCLAA.noah@noah.org>

Python does not have a reverse split, but of course
you can still do it. The simplest way would be to split 
the string into all the fields and then just take the last field.
	a = 'aaa.bbb.ccc'
	f = a.split('.') # Split does not split on '.' by default, 
                       # so you have to tell it to split on '.'.
	c = a[-1]  # first field from the right. Negative indexes
                 # look weird to a newbie, but it's cool.

You could also do something like this:
	a = 'aaa.bbb.ccc'
	reverse (a)
	c,b = a.split('.', 1)
	reverse (c)
Writing the reverse() function (because Python doesn't
have one of those either :-) is left as an exercise to the reader...
Arguably, you might just write your own split routine
for the same amount of trouble and twice the speed.

Yours,
Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of bvdpoel at uniserve.com
Sent: Sunday, April 14, 2002 7:58 PM
To: python-list at python.org
Subject: split() off last substring



Can I tell split() to split off the LAST seqment? I guess I want
something like:

	a="aaa.bbb.ccc"
	b,c=a.rsplit(1)
	print b,c
	# "aaa.bbb", "ccc"

Thx.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel






From siegfried.gonzi at kfunigraz.ac.at  Wed Apr  3 02:41:44 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 03 Apr 2002 09:41:44 +0200
Subject: A faster Python?
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53>
Message-ID: <3CAAB238.290440C5@kfunigraz.ac.at>

Emile van Sebille wrote:
> 
> Nick Arnett
> > To be specific, I do a lot of time series, correlation, regression and
> > clustering analysis.  Those can really bog down.  While I'm fine with
> > exporting to Excel, SPSS, etc., for heavy-duty analysis,
> 
> This surprised me.  I like python with Numeric exactly because it's
> faster and processes more data than Excel.  I'm now regularly
> calculating and delivering pivot-style results from a million row source
> with 240 columns.

Where it comes --otherwise to your experience-- that people believe that
the Numeric-package is slow? I did not compare Numeric to lets say
Matlab or IDL. But why is it that people believe that Numeric is slower
than for example Matlab (for me, when one mentions Numric in combination
with speed, this implies that Numeric is way slower than Matlab or IDL)?

My accompanied Yorick manual says that Yorick can gain 10% of the speed
of compiled C code. I, personaly, made the experience that Yorick is
quite as fast as Matlab or IDL. 


Please, people should realize that Matlab or IDL too is slow compared to
C!

The above writing will not help improving the speed of Python (which
itself, without compiled libraries, would be really  a pain in the
neck).

I think it is important to mention Numeric in combination with Matlab or
IDL or Yorick or...otherwise one could easily fall for it and believe
that Numeric is inherently slow; but I believe my notion is not far away
from reality that there are not any tools available  like
Matlab,IDL,Yorick,Numric  which have matured to compiled C speed in a
side by side comparison.

Or are there any hints that Numric is really way slower than Matlab?
Please no matrix-matrix multiplication; this test is useful for a
compiled language and when on does implement it himself. But there are
often quite a few differences how a specific task is implemented (e.g.
the IDL FFT does care on arrays which dimensionality is not exactly
2^n).


S. Gonzi


From dalke at dalkescientific.com  Fri Apr 19 16:11:32 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Fri, 19 Apr 2002 14:11:32 -0600
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>	<NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>	<mailman.1019068175.11920.python-list@python.org>	<99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com> <mailman.1019233076.19513.python-list@python.org> <3CC0578E.822EB961@metaslash.com>
Message-ID: <a9ptgh$2r6$1@nntp9.atl.mindspring.net>

Neal Norwitz:

and/or
>0.248443961143

create the dictionary each time
>0.696393966675

create the dictionary once, and use get
>0.322777986526

create the dictionary once, and use try/except
>0.239256024361

One other thing skews these numbers.  You are looking up a module
variable each time for the last two timings, and that's a hash table
lookup.  If it's inside a function then it's a constant time lookup.
Here's the same timings on my machine, with the second batch of
timings done by code identical to the first, except inside of a
function.

All variables in module scope
0.382031083107
1.22617602348
0.645591974258
0.358752012253

All variables in function scope
0.266326904297
1.18248593807
0.544940948486
0.250100016594

You can see there are speedups all over the board, and the fastest
is still the try/except.

The reason the try/except is faster than the .get is because while
exceptions are slow, they are never called for your test (you always
use 'f').  The overhead for the [] on dicts is lower than the overhead
to create the bound method ".get".


                    Andrew
                    dalke at dalkescientific.com






From jeff at ccvcorp.com  Wed Apr 24 21:25:59 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 24 Apr 2002 18:25:59 -0700
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org> <mailman.1019682973.29139.python-list@python.org> <871yd4odap.fsf@pvv.org>
Message-ID: <MPG.1730fb06f5bab3e99896c8@news.nwlink.com>

In article <871yd4odap.fsf at pvv.org>, oyving+news at pvv.org says...
> * Cliff Wells
> |
> | On 24 Apr 2002 22:57:20 +0200
> | Oyvind Gronnesby wrote:
> | 
> | > * Cimarron Taylor
> | > |
> | > | Will this do what you want?
> | > | 
> | > | verboseGender = ('female','male')[gender=='m']
> | > 
> | > Wouldn't this break in a PEP 285 Python world?
> | 
> | No.  That was the whole point of subclassing bool from int.
> 
> But wouldn't that just imply that True/False would turn into 1/0 if coerced
> into it by int()?

Yes, they will.  This is by design, to maintain backwards 
compatibility.  The only effective difference between 1 and True, 
and 0 and False, is how it's written, but they add, subtract, 
multiply, and hash the same.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From sholden at holdenweb.com  Tue Apr 30 11:22:46 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 11:22:46 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1> <mailman.1020172854.26653.python-list@python.org>
Message-ID: <JHyz8.16736$xy4.6769@atlpnn01.usenetserver.com>

"Roman Neuhauser" <neuhauser at mail.cz> wrote ...
> > From: Duncan Booth <duncan at NOSPAMrcp.co.uk>
> > Subject: Re: Comment on draft PEP for deprecating six builtins
> > To: python-list at python.org
> > Date: Tue, 30 Apr 2002 12:39:30 +0000 (UTC)
> >
> > Bernhard Herzog <bh at intevation.de> wrote in
> > news:6qn0vlbcpk.fsf at abnoba.intevation.de:
> >
> > >>     Ugh, that looks to me as appropriate as join() being a string
> > >>     method / module function.
> > >
> > > Exactly. I still think that join as a function makes more sense than
as
> > > a string method and for the same reasons as stated for chr above. Even
> > > if join is implemented as a string method.
> >
> > I prefer writing
> >    str.join(' ', aList)
> > rather than
> >    ' '.join(aList)
> >
> > When the string is a literal I find the former reads much better. YMMV
>
>     IMO that's because join being a string method is just awkward. How
>     does this read:
>     aList.join(' ')
>     After all, you tell the list to give you it's concatenation, using
>     the argument as the glue.
>
> > Sometimes though I take advantage of it being a method:
> >    joinlines = '\n'.join
> >    ...
> >    s = joinlines(aList)
>
>     This is not a method. You masked it as a global function.

While it might look like a function when you use it, Python seems to be
pretty clear that it isn't, as the following interpreter session shows:

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> joinlines = '\n'.join
>>> joinlines
<built-in method join of str object at 0x0079A740>
>>>

>
> > Note that there was a suggestion on the python-dev list to move the
> > remaining objects in the string module into the str type and then
deprecate
> > the string module. So str.hexdigits instead of string.hexdigits etc.
>
>     That should be done, of course.
>

There will doubtless be cries about the depracation of the string module
too.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From guido at python.org  Fri Apr  5 20:40:01 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 06 Apr 2002 01:40:01 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org> <Npor8.8281$q63.818@atlpnn01.usenetserver.com>
Message-ID: <3CAE52CA.60F56E3D@python.org>

Somebody (probably Cliff Wells) wrote:
> > I didn't necessarily agree with all of Laura's assertions regarding the
> > difficulty of understanding how booleans would fit into Python, but she
> > raised several compelling and thoughtful arguments.  I totally agree that
> > bools as return values of functions are less than useful.  Booleans have
> > the lowest information density of any type and thus would tend to increase
> > code verbosity if their use were to be encouraged.

Steve Holden replied:
> I too thought Laura made a thoughtful contribution to the discussion, that
> deserved to be considered more respectfully.

I agree.  I can assure you that I *did* read Laura's objections carefully,
but ended up disagreeing with her conclusion that accepting the PEP was a
bad idea.  I was baffled that she and Alex used so many words to explain
their position ("do I have to read and respond to all *that*?!" were my
exact
thoughts) and this made me speechless in response.  Eventually I recovered
from the blow and decided that Laura's fear that adding bool would make her
teaching job harder was ungrounded.  By then, so many words had gone back
& forth that I didn't feel like going back and responding to her piece in
detail -- others had done that and the discussion had blown up to such a
volume that I didn't think I could convince any of the PEP's detractors
anyway.  I apologize to Laura that my judgement came across as rude.

> > Frankly Guido's dismissal of Laura's arguments as FUD was not only a bit
> > rude, but IMHO a bit short-sighted as well (given the fact that someday,
> > there may be a desire for a real boolean type).  Maybe Tim Peters has
> > borrowed his time-machine and not returned it ;)
> >
> However it happened, it left me feeling a little embarrassed that I had
> originally defended Guido's inegrity and suggested that despite his
> authorship the PEP's acceptance was not a foregone conclusion. I hope I
> wasn't wrong, but that's not how it looks right now. Oh well.

I'm sorry I embarrassed you.  Despite how it looks, acceptance of the PEP
was *not* a foregone conclusion.  I have looked carefully for a killer
reason against its acceptance, and did not find it.  I found a lot of
disagreement on what people consider a "real" Boolean type.  I did find
wide agreement on something that I had myself taken for granted, but which
some proponents of "real" Booleans were challenging: the usefulness of
Python's ability to test almost *any* object for its truth value, what
Laura calls something/nothing.

Finally, some folks believe that the new bool should not behave like an
int.  I've thought about this too.  *Maybe*, if I had added a bool back
in 1990, I would have agreed and introduced a "pure" bool type.  Right now,
if I added such a type, it would not be usable as the return value of
comparisons and other operations that return a binary result, because of
backwards compatibility.  And backwards compatibility is a key requirement
for all language changes (see the division discussion).  Backwards
compatibility would forever prevent returning a pure bool from existing
built-in operations.  The only possible transition path has to use an
"impure" bool like the PEP proposes.  Then the only open item is whether
this should be considered an impurity that should be deprecated, or a
permanent wart in the language.  I believe that it's a useful enough wart
that it can be allowed to stay, and I don't believe that the language
would be easier to learn or use with pure bools, so I think there's no
point in deprecating it.  But if years from now our experience points
out that there is in fact a problem with impure bools, I'm willing to
reconsider.  However, since I see no way to get there without going
through the impure bool, I see no reason to wait in implementing the PEP
*now*, and that's what I have done.

Summary: the impure bool makes it *possible* to transition to a pure bool
*if* we find that a pure bool is needed; introducing a pure bool from day
one would be impossible without breaking reams of existing code.

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


From timr at probo.com  Sat Apr 13 01:06:36 2002
From: timr at probo.com (Tim Roberts)
Date: Fri, 12 Apr 2002 22:06:36 -0700
Subject: saving contents of wxHtmlWindow
References: <vxBt8.12401$C21.2528592@news6-win.server.ntlworld.com>
Message-ID: <mvefbu8u9g6s6ffr0tj86f7mbtb9h4440b@4ax.com>

Alan <alan.salmoni at ntlworld.com> wrote:

>apologies if this is the wrong place, but when using a wxHtmlWindow, how 
>can I asve the contents? I have checked the wxWindows docs, but found 
>nothing, and I am relatively new at this.

As far as I know, wxHtmlWindow is for display only.  If you set the
contents with SetPage, then you already have the contents.  If you used
LoadPage, then you will have to use something like htmllib or wxHTML to
fetch the contents.

The best source of information on wxPython is the wxpython mailing list.
If you are working with it at all, I STRONGLY recommend the list.
Subscribe here:
  http://lists.wxwindows.org/mailman/listinfo/wxpython-users
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From ngps at vista.netmemetic.com  Wed Apr 10 11:15:14 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 10 Apr 2002 15:15:14 GMT
Subject: SSL error:  2.1.1 works but 2.2 doesn't
References: <mailman.1018331718.5276.python-list@python.org>
Message-ID: <a91ku2$76q$1@clematis.singnet.com.sg>

According to Walter Miller  <w3miller at yahoo.com>:
> Apologies if this has been discussed/solved already. 
> I looked for prior postings but haven't found a
> solution for fixing this error I'm getting with
> _socket.pyd on 2.2.
>
> doc =
> urllib2.urlopen("https://www.medi-cal.ca.gov").read()
> print doc

ngps at vista:~$ python
Python 2.2 (#7, Mar 31 2002, 23:17:41) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> from M2Crypto import m2urllib
>>> doc = m2urllib.urlopen('https://www.medi-cal.ca.gov').read()
send: 'GET / HTTP/1.1\r\n'
send: 'Host: www.medi-cal.ca.gov:443\r\n'
send: 'Accept-Encoding: identity\r\n'
send: 'User-agent: Python-urllib/1.15\r\n'
send: '\r\n'
reply: 'HTTP/1.1 200 OK\r\n'
header: Server: Microsoft-IIS/5.0
header: Date: Wed, 10 Apr 2002 15:15:04 GMT
header: Content-Length: 36424
header: Content-Type: text/html
header: Cache-control: private
>>> print doc
[ bunch of html scrolled past ]
<style type="text/css">
<!--
.BannerStyle A:link {color:white;text-decoration:none;style:bold}
.BannerStyle A:visited {color:white;text-decoration:none;style:bold}
.BannerStyle A:active {color:white;text-decoration:none;style:bold} 
.BannerStyle A:hover {color:white;text-decoration:underline}

.replace A:link {color:white;text-decoration:none;style:bold}
.replace A:visited {color:white;text-decoration:none;style:bold}
.replace A:active {color:white;text-decoration:none;style:bold} 
.replace A:hover {color:white;text-decoration:none}

>>> 

ngps at vista:~$ uname -a
FreeBSD vista.netmemetic.com 4.5-STABLE FreeBSD 4.5-STABLE #0: Tue Feb 26 00:06:08 SGT 2002     
root at vista.netmemetic.com:/usr/local/sys-src/sys/compile/VISTA  i386


I imagine this should also work on Win2K. Grab the M2Crypto Win32 package and try it out.

Cheers.


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



From pythonista at uk2.net  Tue Apr 23 04:38:55 2002
From: pythonista at uk2.net (Cliff)
Date: 23 Apr 2002 01:38:55 -0700
Subject: Import conundrum
Message-ID: <b9039be.0204230038.3132cf96@posting.google.com>

If I put the following Python 2.1 script in a file called 'xml.py':

   from xml.dom.minidom import parseString
   dom = parseString('<?xml version="1.0" ?><test/>')
   print dom.writexml(sys.stdout)

and then try to run it on Windows NT I get the following:

   Traceback (most recent call last):
     File "C:\home\temp\xml.py", line 5, in ?
       from xml.dom.minidom import parseString
     File "C:\home\temp\xml.py", line 5, in ?
       from xml.dom.minidom import parseString
   ImportError: No module named dom.minidom

If I rename the file to something else, then everything is as expected. 
What I am interested to know is whether this is 'correct' behaviour, or am 
I doing something really stupid?

I ask, because I would personally expect the Python interpreter to know the 
difference between the file called 'xml.py' and the library path 'xml.dom' 
(which I would expect to be mapped to .\xml\dom on Windows).

And there I was thinking I understood this Python stuff... ;-)

Cliff


From rburnham at cri-inc.com  Mon Apr 15 09:49:44 2002
From: rburnham at cri-inc.com (Roger Burnham)
Date: Mon, 15 Apr 2002 13:49:44 GMT
Subject: upgrading to python2.2
References: <292c8da4.0204150521.35d4a842@posting.google.com>
Message-ID: <3cbad91d.1063499@news-server.socal.rr.com>

On 15 Apr 2002 06:21:08 -0700, westernsam at hotmail.com (sam) wrote:

>Hi,
>
>I'm trying to upgrade from python 1.52 to python 2.2. Most of my old
>code works fine. However the SioModule I used
>(http://starship.python.net/crew/roger/) doesn't seem to work (even
>though I've got the 2.2 edition). When import it I get the following
>error:
>
>C:\Python22\Serial\MsTimer.py:5: RuntimeWarning: Python C API version
>mismatch f
>or module elapsedTimer: This Python has API version 1011, module
>elapsedTimer ha
>s version 1007.
>  import elapsedTimer
>
You seem to have the old pyd laying about.  Search your disks for
elapsedTimer.pyd and remove the older one.

>When trying to use the module the ReadTerminated method doesn't seem
>to recognise the 'rspTerm' value I use in the port dictionary.
>
>Has anybody else successfully ported (no pun intended) from 1.52 to
>2.2 using this module?
>

I sent the 2.2 installer to Sanjay Kapoor to try out.  Did not get any repsonse
one way or the other.

>Any clues to this gratefully received
>
>Sam

Roger Burnham
rburnham at blennylips.com


From brad-g-johnson at uiowa.edu  Mon Apr 22 16:15:01 2002
From: brad-g-johnson at uiowa.edu (Johnson, Brad)
Date: Mon, 22 Apr 2002 15:15:01 -0500
Subject: Interacting with other Windoze apps
Message-ID: <412F4D3C5614D211853600A0C99E563F0E776CD5@com83-173.medicine.uiowa.edu>

Hello,

I'm a new Python programmer and have a question about "interacting" with
other Windoze applications.  I'd like to send key strokes/characters to an
application window, but have not been able to find any documentation on
this.  Is it possible?  For example, suppose a user has Microsoft Word open.
Is it possible to send an "Alt + F" then an "O" to the Word window to
activate the *F*ile menu and select *O*pen?

Any pointers (and sample code!) would be appreciated.  Thanks.

Brad

____________________________________
Brad Johnson
Department of Pediatrics - Neurology
200 Hawkins Drive, 2505 JCP
Iowa City, IA 52242-1083
Voice: +1 319 356 1851
Fax: +1 319 356 4855
e-mail: brad-g-johnson at uiowa.edu





From db3l at fitlinxx.com  Wed Apr  3 16:05:44 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Apr 2002 16:05:44 -0500
Subject: Threading problem
References: <a8f8f6$vdu$1@news.rchland.ibm.com>
Message-ID: <uit78wl07.fsf@ctwd0143.fitlinxx.com>

"Larry Whitley" <ldw at us.ibm.com> writes:

> I have a PC with two processors at work and have decided to try to use
> threading to run some long running console applications.  My basic idea is
> to start two threads, one for each processor, and give them a list of things
> to do.  As each thread finishes it one job will go to the list and pick off
> the next thing to do.  When the list is empty, both threads are done.

Ignoring the specific coding error that has been addressed by other
responses, it should also be pointed out that if you're trying to get
effective utilization of your pair of CPUs, this probably won't do it.

The Python global interpreter lock requires that even with native
threading, only one thread is actually interpreting Python code at any
moment in time.  The net result of this is that while threading is
excellent for overlapping I/O, it won't efficiently distribute CPU
amongst processors.  That's because if you are CPU bound, then you are
stuck in Python code execution, and only one processor can be doing
that at any given moment in time.

This can be improved somewhat if you use extensions that release the
GIL, but in most cases that's done during I/O and not processing time,
particularly if the latter will require access to Python structures.
But in theory a purely CPU bound extension could release the GIL while
performing computations that it knew didn't require any interface to
Python.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From maxm at mxm.dk  Thu Apr 18 04:22:47 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 18 Apr 2002 10:22:47 +0200
Subject: My First .py
References: <6Juv8.134255$QB1.35129169@typhoon3.we.ipsvc.net> <3CBE7FB9.9080907@mxm.dk>
Message-ID: <3CBE8257.1050003@mxm.dk>

Max M wrote:

whoops ... python or was offline when I wrote the code so I changed it 
to google, but naturally the code should then read ...

from urllib import urlopen

googleCom = urlopen('http://www.google.com')
content = googleCom.read()
googleCom.close()

print
content[content.find('<title>')+len('<title>'):content.find('</title>')]

>>> Google



From quinn at hurl.ugcs.caltech.edu  Thu Apr  4 15:28:12 2002
From: quinn at hurl.ugcs.caltech.edu (Quinn Dunkan)
Date: 4 Apr 2002 20:28:12 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com> <3CAC6283.424163FB@python.org> <7xu1qr4jnq.fsf@ruckus.brouhaha.com>
Message-ID: <slrnaapdqr.39o.quinn@hurl.ugcs.caltech.edu>

On 04 Apr 2002 06:34:33 -0800, Paul Rubin <phr-n2002a at nightsong.com> wrote:
>Guido van Rossum <guido at python.org> writes:
>> > I think there will need to be a format code for truth values:
>> > 
>> >   ("%b" % True) => "True"
>> 
>> There's no need.  "%s" % True already yields 'True'.  if you want
>> coercion to bool, use "%s" % bool(b).
>
>In that case there's also no need for %d, since "%s" % 17 already
>yields '17'.  If ints have a format code and bools don't, one is
>left with the impression that bools aren't as respectable a type
>as ints.  That doesn't seem in the spirit of the PEP.

"'%03s' % 17" is not the same as "'%03d' % 17".

There are really only two types supported by %---numeric types and strings (and
C holdovers like %c (which is just a variant on strings for python)).  And I
think the string formatting is just a clever pun on the meaning of precision
and leading zero.

So % is not really about formatting types, it's about formatting numbers and
strings.  There's no useful interpretation of precision and the other % flags
for booleans, and even if there were I wouldn't want to further overload
things.

The way to format types is to give them  __str__ and use %s.  If you want
formatting with arguments, I think a perfectly reasonable approach is
"'%s ...' % obj.fmt(1, 2)".  You could do "'%(1,2)s ...' % d(obj)" with a
suitably devious definition of d(), but I'd go with obj.fmt().


From martin at v.loewis.de  Thu Apr  4 13:47:07 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Apr 2002 20:47:07 +0200
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: <3CAB8C99.232A8571@metaslash.com>
References: <200204032232.g33MWqJ28933@odiug.zope.com>
	<3CAB8C99.232A8571@metaslash.com>
Message-ID: <m3bsczfgic.fsf@mira.informatik.hu-berlin.de>

Neal Norwitz <neal at metaslash.com> writes:

> 	def test(a):
> 	    if a is True: print 'True'
> 	    if a == False: print 'False'
> 
> generates these warnings:
> 	test.py:2: Should not assign to True, it is (or will be) a builtin
> 	test.py:3: Should not assign to False, it is (or will be) a builtin

While Gustavo suggests to let pass 'a is True', I recommend to give a
warning for this - people who want to see whether something is True
should write 'True is a'.

Regards,
Martin





From aleax at aleax.it  Mon Apr 29 17:19:40 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 21:19:40 GMT
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch>
Message-ID: <MNiz8.32691$8D3.925320@news1.tin.it>

Boris Borcic wrote:
        ...
> Octal has some virtues, e.g. 365 == 0555 making it easy
> to remember the number of days in a usual (non-leap) year :)

I'm surprised you miss the main advantage -- 0666, the Number
of the Beast in octal, translates to decimal 438, the A.D. in which
Emperor Theodosius II promulgated the Theodosian Code, the
wildly antisemitic laws which, among other things, forbid Christians
to even give alms to Jews, forbid Jews to hold State office or build
synagogues, condemned to death any Jew converting a Christian.

Surely this is therefore a number which is often handy in the everyday
practice of programming, and even thinking of eventually making a
non-builtin of the precious Python function that helps us work with a
numbering system which has such wondrous properties is unholy.


Alex



From phr-n2002a at nightsong.com  Wed Apr 17 17:17:00 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 17 Apr 2002 14:17:00 -0700
Subject: Pure python crypto
References: <mailman.1019024313.11744.python-list@python.org> <7x3cxubvgb.fsf@ruckus.brouhaha.com> <mailman.1019053415.9589.python-list@python.org>
Message-ID: <7x662qqb3n.fsf@ruckus.brouhaha.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
>   The nice thing about blowfish is its simplicity. I started thinking
> about implementing DES and 3DES though and its ugly. 

Blowfish isn't so simple to implement in Python, plus the key setup
is awfully slow.

>   I was thinking that it might be nice to have an 'ssh' module or
> 'sftp' module. Specifically, I wanted sftp for an application and
> couldn't find any such modules. For portability reasons, it might
> be nice to have a pure python module, however, it would require
> some tweaking I'm sure.

I think RC4 coded in Python would run much faster than Blowfish coded
in Python.  SSH can use RC4.  I don't know what SFTP is.

FWIW, I coded RC4 in Python a while back and if I remember correctly,
got about 100k/sec throughput on a P3-750.  That's sort of marginal
for (say) sftp over ADSL.  Using the SHA-OFB hack that I described,
I got about 400-500k/sec.  


From brueckd at tbye.com  Wed Apr 17 10:55:57 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 17 Apr 2002 07:55:57 -0700 (PDT)
Subject: Counterintuitive Python behavior
In-Reply-To: <lkbscisakh.fsf@pc150.maths.bris.ac.uk>
Message-ID: <Pine.LNX.4.33.0204170751290.10117-100000@ns0.tbye.com>

On Wed, 17 Apr 2002, Michael Hudson wrote:

> It's you :) As I can't read my email at the moment[1], I have no
> better way of wasting my time to hand than drawing you some ascii art.
[snip]

This is a great explanation! The next time a co-worker asks a
assignment/binding/references question I'm going to refer them to this
message. Thanks for writing this!

[snip]
>                                                     +---+
>     ,------.       +-------+                     ,->| 1 |
>     | dict |------>|+-----+|             +-----+ |  +---+
>     `------'       || "a" |+------------>| [0]-+-'
>                    |+-----+|             | [1]-+-.
>                    |+-----+|             +-----+ |  +---+
>                    || "b" |+-----.         /\    `->| 3 |
>                    |+-----+|     |    ,----'        +---+
>                    +-------+     |    |
>                                  `----+-----.
>     ,------.       +-----+            |     \/
>     | list |------>| [0]-+------------'   +-----+   +---+
>     `------'       | [1]-+--------------->| [0]-+-->| 2 |
>                    +-----+                +-----+   +---+

(sorry for reposting this, but it was too cool to snip)

-Dave





From bokr at oz.net  Thu Apr 25 10:46:18 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Apr 2002 14:46:18 GMT
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at> <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com> <3CC80EBE.E9735284@kfunigraz.ac.at>
Message-ID: <aa94rq$4qb$0@216.39.172.122>

On Thu, 25 Apr 2002 16:12:14 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
[...]
>I just start the script from the idle editor: 'Run script'.
>
>I even tried to use del(list1), del(list2),... but it won't help. I also
>write the results to a file (and not to the output screen) in order not
>to interference the calculation.
>
>After 4 simulations it comes to a crash.
>
What happens if you run it from a plain "DOS" console window (I am assuming
from your post header that you are running on NT?), typing

    python your_script.py arg1 arg2 ...

at the prompt?

Regards,
Bengt Richter


From sjmachin at lexicon.net  Mon Apr 29 20:09:57 2002
From: sjmachin at lexicon.net (John Machin)
Date: 29 Apr 2002 17:09:57 -0700
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <mailman.1020102370.5619.python-list@python.org>
Message-ID: <c76ff6fc.0204291609.115c33d6@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1020102370.5619.python-list at python.org>...
> On Mon, Apr 29, 2002 at 01:26:25PM -0400, PoulsenL at capecon.com wrote:
> > I have a Numeric array with missing elements from which I wish to calculate
> > a histogram.  However, in order to do so I must first remove the 'None'
> > values from the array.  I have a kludgy solution, but I was wondering if
> > someone could point me in the right direction on what is the proper method
> > of collapsing a masked array to just the non-masked values.
> 
> there are too solutions. choose yourself :-)
> 
> nums = [ 1,2,3,None,2,3]
> 
> #version1
> def f(x): return x is not None
> newnums = filter(f, nums)
> 
> #version2
> newnums = [ x for x in nums if x is not None]
> 
> The first version uses filter. 
> 'filter' calls 'f' for every item in 'nums' and puts
> the result into the new list iff it has a 'true' value.
> 
> The second version uses list comprehension.
> It basically does the same if you look closely
> at it. 

*Don't* try this if zero is a valid value for your histogram:

newnums = filter(None, nums)

because it will rip out any "false" entries (including 0 or 0.0) that
you have.

Caveat lector: The OP was talking about a Numeric array. A Numeric
user (i.e. not me) may be able to comment on (a) whether filter()
works on Numeric arrays (b) what the preferred solution is for Numeric
arrays.


From mmueller at dgfz.de  Tue Apr 16 06:58:04 2002
From: mmueller at dgfz.de (=?ISO-8859-1?Q?Mike_M=FCller?=)
Date: 16 Apr 2002 03:58:04 -0700
Subject: os.utime() for directories on win32?
Message-ID: <7049ba55.0204160258.1d05d685@posting.google.com>

Changing the modification time of directories with os.utime() works on
Linux.

On Windows (here: NT4.0, Python2.1) os.utime() doesn't change the
modification time. No exception is raised.
Is this a bug or a feature?
Is there any other way to change the modification time of directories?

Thanks

Mike


From guido at python.org  Sat Apr  6 18:13:31 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 06 Apr 2002 18:13:31 -0500
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "Sat, 06 Apr 2002 21:13:38 +0200."
             <200204061913.g36JDcdj017764@ratthing-b246.strakt.com> 
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net>  
            <200204061913.g36JDcdj017764@ratthing-b246.strakt.com> 
Message-ID: <200204062313.g36NDV428946@pcp742651pcs.reston01.va.comcast.net>

[Laura]
> I think there is a time and a place for bools --  Describing things
> that have two states that can never, ever, have 3 or more states.  
> These things are vanishingly close to non-existent, unless you are
> a mathematician in which case you may be playing with these sorts
> of things all the time.  What I am trying to do is save people
> like Mark McEahern from his isValid function, which returns whether
> a username is valid or not.  On the day his boss demands a third
> value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
> he will thank me.

Laura, you've said this over and over, and I don't buy it.

Two arguments from XP apply here:

- YAGNI (You Ain't Gonna Need It), or also "do the simplest thing that
  can possibly work".  If all you need today is a bool, use a bool.

- "Refactor mercilessly".  If you find later that an int makes more
  sense than a bool for a particular function or variable, change it!
  It's not the end of the world.

In addition, it seems the same situation applies with all sorts of
other data types.  Why use an int when one day you might need a float?
Why use a float when you might need a complex number?  Why use a tuple
when you might need a list?  Why use a list when you might change your
mind and need a dict?

> That is what I mean by teaching people not to use bools -- teaching
> when not to use them. This is very hard to teach.  I know it because
> I have been doing it for years and years.   I even know _why_ it is
> hard to teach, but that doesn't make it any easier.

Have you been teaching Python all those years?  I'd like to get to the
bottom of this, because it's so against my own intuition.  I haven't
been teaching, but I've sure helped a lot of people who were
struggling with programming problems -- both in C and in Python.

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




From mwh at python.net  Wed Apr  3 12:49:36 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 3 Apr 2002 17:49:36 GMT
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <lkvgb88yfk.fsf@elios.maths.bris.ac.uk>

roy at panix.com (Roy Smith) writes:

> I've got a dictionary d, which may or may not have a value I'm looking
> for.  If it does, I want to do a few things.  If it doesn't, I want to
> do a few other things.  I could write:
> 
> try:
>     value = d[key]
>     foo = stuff (value)
> except KeyError:
>     foo = otherStuff (key)
> 
> which is fine as long as I can guarantee (hint: I can't) that stuff()
> doesn't raise a KeyError.

You're hunting for

try:
    value = d[key]
except KeyError:
    foo = otherStuff(key)
else:
    foo = stuff(value)

This construct is definitely Something Worth Knowing About.

Cheers,
M.


From eppstein at ics.uci.edu  Tue Apr  2 18:23:41 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 02 Apr 2002 15:23:41 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <3CAA03E4.A79BA593@noaa.gov> <Xns91E4EF89BD4EDcliechtigmxnet@62.2.16.82> <pTpq8.53356$S52.1844974@news2.tin.it>
Message-ID: <eppstein-E6E1AD.15234102042002@news.service.uci.edu>

In article <pTpq8.53356$S52.1844974 at news2.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> > yes. but a real bool could be introduced as module for now (for the RPC
> > and marshalling people that need one) i think its not needed otherwise
> 
> A standard library module sounds good.  Or maybe it could go into operator.

So you'd be happy with a Boolean class, as long as it isn't convertable 
to integers and none of the standard operators are changed to return it?

That doesn't seem very nice either from the aesthetic purity-of-design 
point of view (non-convertable is more pure than convertable maybe but 
it's impure to have a Boolean type but return int from a comparison) nor 
from the code-documentation point of view (have to put explicit Boolean 
conversions on results of comparisons if you want to properly express 
the fact that your function's return value should be interpreted as a 
truth value not an integer) nor from the pragmatic point of view 
(Booleans would be too cumbersome to use except in unusual situations).

In comparison the actual PEP is similarly not-pure but gives better 
expressivity and possibly adds minor functionality (ability to make code 
that behaves differently for ints and for bools, just like now you can 
behave differently for lists and for tuples even though they are 
interconvertable).

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From logiplexsoftware at earthlink.net  Mon Apr 29 19:03:14 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 16:03:14 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <oqpu0inmxa.fsf@titan.progiciels-bpi.ca>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
	<oqpu0inmxa.fsf@titan.progiciels-bpi.ca>
Message-ID: <20020429160314.55b3e6de.logiplexsoftware@earthlink.net>

On 29 Apr 2002 18:46:09 -0400
Fran?ois Pinard wrote:

> I prefer intelligent discussions between nice people.

That's why no one likes you.

;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From aahz at pythoncraft.com  Tue Apr  9 17:50:19 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 17:50:19 -0400
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org>
Message-ID: <a8vnmr$r78$1@panix1.panix.com>

In article <mailman.1018366940.13473.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>
>Boris' phrasing indicates that I was 'trying to make a point' as if
>this was some sort of ego-driven competition between myself and Guido
>with some sort of pay-off in terms of whether the PEP passes, or
>whether more big names on comp.lang.python come out in favour of me
>than Guido or some other piece of nonsense.  I did not want this.
>Boris has revealed a flaw in our process, because that he took it that
>way means that the PEP process is being perceived as an ego-game.

Unfortunately, I have to say that the stridency you exhibited in this
thread (to my startlement) made me feel a bit that you *were* trying to
score a point.  I'm not talking about your initial long post, but
subsequent ones.

>This _really_ stinks, people.  I trusted Guido to be able to take off his
>'I am an implementor and I just implemented a cool new feature' hat
>and put on his 'I am a langauge designer, and I am now looking at the
>language implications' hat.  This trust we must have in Guido or else
>he will not be able to launch his own new features into Python.  This
>is the faith we must have or else we think that we ONLY GOT LUCKY
>in that Python is great.  

See?  You're implying here that Guido did *not* put on his "language
designer" hat, and in a rather subtle way, to boot.  You're
contradicting the very point you're trying to make by your use of
language.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From philh at comuno.freeserve.co.uk  Mon Apr 22 08:20:40 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 22 Apr 2002 13:20:40 +0100
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz>
Message-ID: <slrnac800o.10k.philh@comuno.freeserve.co.uk>

On 22 Apr 2002 17:04:13 +1200, Paul Foley <see at below> wrote:
>On Sun, 21 Apr 2002 02:32:29 -0700, James J Besemer wrote:
>
>> That being said, I guess I just don't get the strong reaction to this secondary
>> point.  I'm new here but I get the impression that "lisp" is a naughty word,
>> for some unfathomable reason
>
>A lot of people think "Lisp" means "Scheme".
>Scheme sucks, so they think "Lisp" sucks.
>
>[Scheme isn't Lisp.  Scheme are Lisp are only distantly related, and
>Scheme is about as closely related to Algol as to Lisp...]

(both (of them) (look (similar)) (to me))

Of course, C++ and Java have similar syntax, and it doesn't make 
them closely related.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From justin at iago.org  Mon Apr  1 15:36:32 2002
From: justin at iago.org (Justin Sheehy)
Date: Mon, 01 Apr 2002 15:36:32 -0500
Subject: PEP 285: Adding a bool type
In-Reply-To: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> (Guido
 van Rossum's message of "Sat, 30 Mar 2002 00:39:10 -0500")
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <x7k7rrw3zj.fsf@ra.iago.org>

Guido van Rossum <guido at python.org> writes:

>     1) Should this PEP be accepted at all.

-0

I don't think it is needed, and I think that the negative aspects
(most of which were well-put by Alex Martelli) are as significant as
the positive ones.

Assuming that it does go through as exists:

>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".)

"True"

>     3) Should the constants be called 'True' and 'False'
>        (corresponding to None) or 'true' and 'false' (as in C++, Java
>        and C99).

'True' and 'False'

>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

No.  If pure boolean values had been in the language from very early
on, then it would make sense to have them break when used in
arithmetic.  Since people have written much code that relies on the
integerness of booleans, and it remains convenient to do so, it should
remain allowed.

>     5) Should operator.truth(x) return an int or a bool.  Tim Peters
>        believes it should return an int because it's been documented
>        as such.  I think it should return a bool; most other standard
>        predicates (e.g. issubtype()) have also been documented as
>        returning 0 or 1, and it's obvious that we want to change those
>        to return a bool.

int, though I don't think it matters much.

-Justin

 




From piet at cs.uu.nl  Wed Apr 24 08:05:05 2002
From: piet at cs.uu.nl (Piet van Oostrum)
Date: 24 Apr 2002 14:05:05 +0200
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com> <mailman.1018931022.12502.python-list@python.org>
Message-ID: <wzd6wp480e.fsf@cs.uu.nl>

>>>>> "Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> (CG) writes:

CG> If you mean, why don't substrings share space with parent strings,
CG> hmm, that's an interesting question.  I can see that a string object
CG> implemented something like this struct:

CG>     struct ExampleString {
CG>         int len;
CG>         char *buf;
CG>     }

CG> would allow that, and it sounds like an interesting optimization. 

But it would make garbage collection harder.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From cgale1 at cox.net  Fri Apr  5 03:51:15 2002
From: cgale1 at cox.net (Van Gale)
Date: Fri, 05 Apr 2002 08:51:15 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAD4CD8.2090700@mxm.dk>
Message-ID: <7Adr8.1447$%A.217236@news1.west.cox.net>

"Max M" wrote:

> Oh stop whining!

Was that really necessary?

> Just do a search and replace in your codebase::
>
>      s/1/True/gi
>
> That should solve all those problems in a few seconds.

Yeah... it'll solve **those** problems, but think about it for a sec.
Maybe, just maybe, this will create a slew of entirely different problems?
:)

Van





From news.ASkwar at DigitalProjects.com  Mon Apr 22 15:11:59 2002
From: news.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Mon, 22 Apr 2002 21:11:59 +0200
Subject: Checking if class is derived of some other class
Message-ID: <pan.2002.04.22.19.11.59.254871.5745@teich.garten.DigitalProjects.com>

Hello.

I'm using Python 2.2.1.

Suppose I've got the following classes/objects:

class a:
	pass
class b(a):
	pass
class c(b):
	pass

myObject = c

How can I check if myObject is either directly instantiated from class a
or is if of a class which inherited from class a?

Thanks,

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 12 hours 23 minutes


From logiplexsoftware at earthlink.net  Wed Apr 17 17:10:18 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 17 Apr 2002 14:10:18 -0700
Subject: [OT] What is Open Source?
In-Reply-To: <20020418002823.C4068@phd.pp.ru>
References: <mailman.1018662715.17645.python-list@python.org>
	<slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com>
	<mailman.1018897470.30703.python-list@python.org>
	<899f842.0204170840.29e579ae@posting.google.com>
	<20020418002823.C4068@phd.pp.ru>
Message-ID: <20020417141018.76b4a8e2.logiplexsoftware@earthlink.net>

On Thu, 18 Apr 2002 00:28:23 +0400
Oleg Broytmann wrote:

> On Wed, Apr 17, 2002 at 09:40:52AM -0700, Anthony_Barker wrote:
> > For me open source is simply the ability to "look under the hood" of
> > any program.
> > Almost every other type of human endevor allows this. 
> 
>    But not software. This is where the big problem lies. If you look into
> M$ "shared source" code, and use a part of the code later, M$ kills you
> using copyright and patent laws.

Especially when it affects your ability to even mimic the functionality of the
code with a different implementation.  You can no longer claim a "clean-room"
reimplementation since you've /seen/ their code.  In this sense, it's better
*not* to see proprietary code, even if it is available.  I suppose Microsoft's
claim that the GPL is viral could be better applied to their own source since
merely *seeing* it could affect your code whereas you have to actually *use*
GPL'd code to be affected.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From NO_SPAM_jmpurser2 at attbi.com  Fri Apr  5 17:37:55 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Fri, 05 Apr 2002 22:37:55 GMT
Subject: Beginning Tkinter problem
Message-ID: <7Hpr8.233293$q2.21629@sccrnsc01>

When working in IDLE(2.2) on Win2000 and attempting some Hello world class
programs I get stuck after entering this code:
>>> from Tkinter import Label
>>> widget = Label(None, text='Hello GUI World!')
>>> widget.pack()
>>> widget.mainloop()

The little window pops up, I click on the x in the corner and the window
disappears.  But I never get my IDLE interactive prompt back.  Even going
into the task manager and killing it there takes a while.

So what am I doing wrong?

Thanks,

John Purser




From comdog at panix.com  Thu Apr  4 18:25:17 2002
From: comdog at panix.com (brian d foy)
Date: Thu, 04 Apr 2002 17:25:17 -0600
Subject: _The Perl Review_ does Python
Message-ID: <040420021725172972%comdog@panix.com>

for our April 2002 issue (April Fool!), The Perl Review replaced its Perl content 
with Python articles, so Python folks might want to read this issue:

   http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf

You will find:

   * an interview with Guido van Rossum -- Adam Turoff
   * The Python Software Foundation: A Primer -- David Ascher
   * Python and the Golden Mean -- Paul Prescod 

thanks to the various people in the Python community who helped us pull off
this little prank :)

-- 
brian d foy <comdog at panix.com> - Perl services for hire
The Perl Review - a new magazine devoted to Perl 
<http://www.theperlreview.com>


From nospam at nospam.com  Sat Apr 13 17:46:55 2002
From: nospam at nospam.com (Jens Baader)
Date: Sat, 13 Apr 2002 23:46:55 +0200
Subject: 2.1 vs. 2.2 
Message-ID: <3cb8a755$0$12714$9b622d9e@news.freenet.de>

I'm downloading Python 2.2.1 right now and I wonder
why you still produce bugfix releases for the old 2.1 development line.
It seems that Python 2.2 is somewhat broken. If not what's the
reason that keeps the people from upgrading to 2.2?

Another question: When will we see an official ISO/ANSI standard
for Python? I dislike working with nonstandardized languages.
The fact that radical language changes are still made (the
type/class unification) seem to suggest that Python is still not
mature enough for a real standard. Or maybe Python and
the other "newschool" languages (like Ruby, Java) are not
meant to by standardized? Don't know but I really would
like to. Are there any plans to get Python "stable" or will
the language continue to change with each major release
of the interpreter?

TIA,
Jens






Jens




















From maxm at mxm.dk  Fri Apr 19 17:30:38 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 19 Apr 2002 23:30:38 +0200
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
References: <mailman.1019071474.3193.python-list@python.org> <a9p1g4$51jr8$1@ID-59885.news.dfncis.de>
Message-ID: <3CC08C7E.9020000@mxm.dk>

Thomas Heller wrote:
> "Trent Mick" <TrentM at ActiveState.com> wrote in message news:mailman.1019071474.3193.python-list at python.org...
> 
> Are the HTMLHelp files available separately? Can they be installed
> separately?

You will probably like what you finde here ...

http://www.orgmf.com.ar/condor/pytstuff.html

regards Max M



From ron.teeter at supplysolution.com  Tue Apr 30 18:11:37 2002
From: ron.teeter at supplysolution.com (Ron Teeter)
Date: Tue, 30 Apr 2002 18:11:37 -0400
Subject: Why do class methods share mutable defaults between instances?
Message-ID: <000701c1f094$00688380$8210a8c0@singularity>

Greetings,

Today I made a rather perplexing discovery.  It appears that mutable defaults 
in class methods are shared between instances.  Here is a simple example of 
the scenario.

class Demo:
    def __init__(self,name,dict={},list=[]):
        self.name = name
        self.dict = dict
        self.list = list
        self.dict[name] = name
        self.list.append(name)

def dump(name,dict={},list=[]):
    print name,dict,list

for name in ['A1','B1','C1']:
    tmp = Demo(name)
    dump(tmp.name,tmp.dict,tmp.list)

>>>
A1 {'A1': 'A1'} ['A1']
B1 {'B1': 'B1', 'A1': 'A1'} ['A1', 'B1']
C1 {'C1': 'C1', 'B1': 'B1', 'A1': 'A1'} ['A1', 'B1', 'C1']
>>>

My intent was that each time Demo was instanciated a new dictionary would be 
assigned to 'dict' if there was no corresponding argument.  However, the 
behavior is as though the {} were replaced with a class constant.

In order to avoid this behavior I had to reorganize Demo as follows:

class Demo2:
    def __init__(self,name,dict=None,list=None):
        if dict==None:
            dict = {}
        if list==None:
            list = []
        self.name = name
        self.dict = dict
        self.list = list
        self.dict[name] = name
        self.list.append(name)

I am wondering if this is a bug or a subtle usage error on my part.  I would 
expect that syntax on functions would be valid for class methods as well, but 
in this case you get drastically different behaviors.

Thanks,

Ron Teeter




From parker at gol.com  Wed Apr 17 18:29:51 2002
From: parker at gol.com (Ian Parker)
Date: Wed, 17 Apr 2002 22:29:51 GMT
Subject: py script web page problem ??
References: <61fe942.0204152214.3cbe6c0a@posting.google.com>
 <mailman.1018944513.17890.python-list@python.org>
Message-ID: <gs6x90Lncfv8EwHW@hiredata.gol.com>

In article <mailman.1018944513.17890.python-list at python.org>, Oleg 
Broytmann <phd at phd.pp.ru> writes
>On Mon, Apr 15, 2002 at 11:14:08PM -0700, linux_on_wood wrote:
>> <FORM action="friends1.py" method=get>
>>
>> my python script code is:    friends1.py  as chmod 755 on linux box
>>
>> #! /usr/bin/env python
>>
>> This script code looks very correct to me but it
>> gets called --friends1.py DOES NOT render the
>> html, it just shows the source code as text.
>
>   A script (any script) should not "rebder" HTML - it will just pass the
>HTML back to a browser. Do I understand your question right that instead of
>HTML you get the source code of the script?
>
>> ANY IDEAS, HELP , WHY  has a problem ???????????
>
>   It seems your web-server was not configured to recognize .py extension
>as a CGI extension. Reconfigure your server.
>
>Oleg.

Or that the web-server doesn't recognise scripts in this directory as 
being executable.

Regards
-- 
Ian Parker


From slime at vsnl.net  Fri Apr 12 04:29:27 2002
From: slime at vsnl.net (Prahlad Vaidyanathan)
Date: Fri, 12 Apr 2002 04:29:27 -0400
Subject: SMTP with attachments
In-Reply-To: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>
References: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>
Message-ID: <E16vwQx-0001YW-00@mail.python.org>

Hi,

On Thu, 11 Apr 2002 Ezequiel, Justin spewed into the ether:
> How do I send an e-mail with attachments?

I recently found a module to do just that, which I find very useful.

Since I don't remember where I got it from, I've put it up here :

    http://www.symonds.net/~prahladv/files/MailMsg.py

The email module had a bug until recently when handling certain types of
attachments, so unless you have a fairly recent Python, I don't advise
using it. MimeWriter et al work fine :-)

HTH,

pv.
-- 
Prahlad Vaidyanathan  <http://www.symonds.net/~prahladv/>

If God hadn't wanted you to be paranoid, He wouldn't have given you such
a vivid imagination.







From syver-en+usenet at online.no  Fri Apr 26 08:36:05 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 26 Apr 2002 12:36:05 GMT
Subject: Killing a process under Windows NT/2K
References: <aab8hn$o4s$05$1@news.t-online.com>
Message-ID: <ur8l2wsch.fsf@online.no>

"Ingo Blank" <spam at yourself.com> writes:

> Hi,
> 
> I need to kill a process under Windows (NT/2K), whose PID is known.
> Unfortunately os.kill() doesn't work under Windows.
> How can I achieve this anyway?
> 
> Thanks in advance
> 

This might help. You have to have the win32all extensions installed,
but so should *everybody* that uses python on win32 anyway.

# Utility functions that should have been implemented in the os module
import signal
signal.SIGKILL = 120 # cludge: try to mimic the standard library kill

def kill(pid, sig):
    assert sig == signal.SIGKILL # cludge: try to mimic the standard library kill
    handle = win32api.OpenProcess(win32con.PROCESS_TERMINATE, 0, pid)
    win32api.TerminateProcess(handle, 1) # exit status
    

def waitpid(pid):
    handle = win32api.OpenProcess(win32con.SYNCHRONIZE|win32con.PROCESS_QUERY_INFORMATION , 0, pid)
    win32event.WaitForSingleObject(handle, win32event.INFINITE)
    exitCode = win32process.GetExitCodeProcess(handle)
    return pid, exitCode


-- 

Vennlig hilsen 

Syver Enstad


From boud at valdyas.org  Tue Apr  9 20:17:01 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 09 Apr 2002 23:17:01 -0100
Subject: PyQt question (colors)
References: <3cb2d4e7_11@news.newsgroups.com> <3cb2dff6$0$85444$e4fe514c@dreader3.news.xs4all.nl> <3cb328e8$1_3@news.newsgroups.com>
Message-ID: <3cb35c9e$0$71646$e4fe514c@dreader4.news.xs4all.nl>

Janos Blazi wrote:

> 
> Thank you for your answer.
> I tried to read the Trolltech manuals (and your book as well by the way)
> but to no avail. Ist it really so difficult to set the two colors of a
> progressbar? Does that require 20 lines of code?

Yes. If it isn't the backgroundcolor you want to set, (and you probably 
shouldn't want to do that, because should follow the user preferences, 
just like the face color of buttons and everything else), but the color of 
the bar itself -- by default black in some styles, blue in others, then I 
now remember I have come across the same problem when I wanted to build a 
unittest dialog for my book.

<rummages through code>

Yes, you can easily alter the bar color in PyQt, simply using a function 
like:

    def setForegroundColor(self, widget, color):
        palette=widget.palette()
        palette.setColor(QColorGroup.Highlight, color)
        widget.setPalette(palette)

that you then call with something like:

    self.setForegroundColor(self.mw.progressBar, QColor("green"))

If you really need to set the backgroundcolor (the part that is normally 
medium gray), then you're on your own still: you will have to experiment
with the various colors in QColorGroup:

const QColor & button () const
const QColor & light () const
const QColor & dark () const
const QColor & mid () const
const QColor & text () const
const QColor & base () const
const QColor & background () const
const QColor & midlight () const
const QColor & brightText () const
const QColor & buttonText () const
const QColor & shadow () const
const QColor & highlight () const
const QColor & highlightedText () const
const QColor & link () const
const QColor & linkVisited () const

And see which one is the one that fits your needs. The fiddling with the 
palette remains the same.

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


From aleax at aleax.it  Thu Apr 11 04:52:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 08:52:23 GMT
Subject: old vs new iteration protocols (was Re: Booleans, &c)
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204100552.6b822307@posting.google.com> <C%Xs8.3139$b62.77265@news1.tin.it> <a925qq$ne4$1@panix1.panix.com>
Message-ID: <b9ct8.6406$vF6.199796@news2.tin.it>

Aahz wrote:

> In article <C%Xs8.3139$b62.77265 at news1.tin.it>,
> Alex Martelli  <aleax at aleax.it> wrote:
>>
>>With the previous iterator protocol, "for item in somedict:" COULD
>>NOT "work smoothly and seamlessly".  Nor, at all.  Think about it.
> 
> Oh, come on, Alex; with your twisted imagination, I'm sure you can think
> of ways in which a dictionary could be forced to work with the old
> iterator protocol -- it would just be very, very messy.

Twisted imagination, me?  I'm a plodding infantry type and I take this
as a challenge.  How does e.g, somedict[0] return some "first thingy
in somedict" AND at the same time raise KeyError if 0 ain't a key in
somedict, just to start with?  The latter is part of the fundamental
social contract of dictionaries -- the former is part of the old
iteration protocol.


Alex



From eppstein at ics.uci.edu  Thu Apr 11 22:54:52 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 11 Apr 2002 19:54:52 -0700
Subject: Tuple comprehension
References: <mailman.1018541080.13442.python-list@python.org> <slrnabchgs.c8l.mlh@vier.idi.ntnu.no>
Message-ID: <eppstein-3037AA.19545211042002@news.service.uci.edu>

In article <slrnabchgs.c8l.mlh at vier.idi.ntnu.no>,
 mlh at vier.idi.ntnu.no (Magnus Lie Hetland) wrote:

> I think the only _real_ reason to have tuples at all is that they are
> immutable (and therefore usable as dictionary keys). And a slight
> efficiency advantage over lists, maybe...

I'm guessing (without looking at the source) that tuples use less memory 
because they don't need the extra room for growth that lists do.
But my mental model of them is just "immutable list".
Which as you say is important for use in dictionaries...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From pj at engr.sgi.com  Mon Apr  8 20:06:46 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 9 Apr 2002 00:06:46 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018281413.8301.python-list@python.org>
Message-ID: <a8tbam$pm65o$2@fido.engr.sgi.com>

|> But that's not well-formed xml

Nor as readable.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From logstx at bellatlantic.net  Sun Apr 14 20:10:03 2002
From: logstx at bellatlantic.net (logistix)
Date: Mon, 15 Apr 2002 00:10:03 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <eppstein-7156A5.13521614042002@news.service.uci.edu> <3cba0068@news.mhogaming.com>
Message-ID: <vTou8.20361$8W5.17808@nwrddc02.gnilink.net>

>
> Ahhhh!!!  that's a GREAT idea!  That way, you could write
> something like:
>

It gets a little trickier though.  Here's the real reason for short-circuit
evaulation:

>>> a = 1
>>> a or c
1
>>> c or a
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
NameError: name 'c' is not defined
>>>

So you can write stuff like:

if isinstance(a, foo) and a.fooOnlyValue:
    pass

instead of the less clear

if isinstance(a, foo):
    if a.fooOnlyValue:
        pass

without throwing an error.  Returning the actual values is more of a side
effect than anything.  So then the next question is, Are non-existant values
considered False or should they throw an error?

If they're false, it looks like you need to incorporate eval's into the
function wrapped with try/except statments to work properly. Some people
don't seem to like doing things like that for some reason ;)





From ralph at inputplus.demon.co.uk  Sat Apr  6 11:11:31 2002
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 6 Apr 2002 16:11:31 GMT
Subject: [Python-Dev] RE: RELEASED: Python 2.2.1c2
References: <NEBBIGHKAHJGMPJEEAGPKENMGGAA.fp@contact.de> <mailman.1017228282.28692.python-list@python.org>
Message-ID: <a8n6nj$j88$1@inputplus.demon.co.uk>

Hi Martin,

> > > Note for AIX users: there are still problems with AIX in this
> > > release, and there has not been time to repair them without fear
> > > of breaking platforms that currently work.  Hopefully by the time
> > > of the release of 2.2.1 final we will be able to provide a recipe
> > > that allows AIX users to build and run Python.  [...]
> > 
> > Hmm, can someone shed more light on this issue?
> 
> I think this is a red herring: Some user, for some reason, does not
> have cc_r. According to a number of other AIX users, cc_r should
> always be available, so it appears to be an installation bug of the
> AIX system in question. I don't think it deserves fixing.

That `some user' is me.  Can you tell me what version of AIX those
users who say it should always be available are referring to?  As the
bug report says, AIX 3.2.5 is what I'm using and cc_r doesn't exist.
It isn't a requirement to install it at the same time as cc.

Perl has always built happily on AIX 3.2.5 in my experience and given
all Python has to do is search through the list [cc_r cc] for the first
one it finds when the OS is AIX I don't see the big deal.  Escpecially
since it already has a `when OS is AIX do a bodge' statement.

I'll add more to the bug report shortly when I've time.


Ralph.



From allexpro at myrealbox.com  Thu Apr 18 23:51:32 2002
From: allexpro at myrealbox.com (allexpro)
Date: Thu, 18 Apr 2002 23:51:32 -0400
Subject: ANN: Twisted 0.17.2 Released
Message-ID: <CvMv8.15243$ue4.582660@weber.videotron.net>

Quote of the Release:

<glyph> skreech: you think "Acquireable" is hard to spell? ;-)
[ed: dict acquireable]

What is this?

  Twisted is an event-based framework for internet applications.  It
includes a
  web server, a telnet server, a multiplayer RPG engine, a generic client
and
  server for remote object access, and APIs for creating new protocols and
  services. Twisted supports integration of the Tk, GTK+, Qt or wxPython
event
  loop with its main event loop. The Win32 event loop is also supported, as
is
  basic support for running servers on top of Jython.

  Twisted currently supports the following protocols, all implemented in
pure
  python, most of them as both servers and clients:

    - FTP
    - HTTP
    - SOCKSv4
    - SMTP
    - IRC
    - telnet
    - POP3
    - AOL's instant messaging TOC
    - OSCAR, used by AOL-IM as well as ICQ (client only)
    - DNS
    - LDAP (client only)
    - finger
    - Echo, discard, chargen and friends
    - Twisted Perspective Broker

  For more information, visit http://www.twistedmatrix.com, or join the list
at
  http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python


What's new in 0.17.2?
=====================

- Added news server and NNTP protocol support

- jelly serialization more properly return a Deferred Failure -- this will
make
hangs in PB apps (most notably distributed web) less common.


- Enhanced the web server:
  * Added .trp and .rpy support in the static web server -- Very simple
    handlers to allow you to easily add Resource objects dynamically to a
    running server, by merely changing files on the filesystem.

   * As resources are being looked up by repeated calls to getChild, this
     change creates instances of
     twisted.spread.refpath.PathReferenceAcquisitionContext and puts them in
     the request as "request.pathRef"

   * Added DOM-based templating system( to manipulate the HTML node tree)
     --Subclasses of twisted.web.domtemplate.DOMTemplate can override the
     templateFile attribute and the getTemplateMethods method

   * Added a new feature that allows requests for /foo to return
     /foo.extension, which is disabled by default. If you want a --static
     webserver that uses this feature, use 'mktap web --static <dir>
     --allow_ignore_ext'.

   * Also switched --static to --path; it doesn't make sense to call
something
     that automatically executes cgis, epys, rpys, php, etc., "static". :-)

   * Added HTTP 1.1 persistent and pipelined connection support.

- Added new coil configurator for words.tendril.








From shalehperry at attbi.com  Sat Apr  6 17:35:19 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 06 Apr 2002 14:35:19 -0800 (PST)
Subject: yet another "Here's to 1.5.2" post
In-Reply-To: <1018131203.756898@cswreg.cos.agilent.com>
Message-ID: <XFMail.20020406143519.shalehperry@attbi.com>

> 
> List comprehension is cutesy and practically Perlish, an idiosyncrasy
> reminiscent (in flavor and aroma) to Perl's ".." operator.
> 

I'll see your list comprehensions and raise you a set of string methods.

" ".join(mylist) # is this perl or python?  whoever decided this is the new
                 # style should be shot, a lot





From robin at jessikat.fsnet.co.uk  Mon Apr 29 04:24:23 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 29 Apr 2002 09:24:23 +0100
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <eUTmiIA3MQz8EwSA@jessikat.fsnet.co.uk>

map, filter & reduce et al are the fastest ways to do many things....,
pow and divmod are also extremely useful. I don't use input, but I'm
sure others do.

Ther are no fast equivalents to the functional forms, list
comprehensions are as slow as for loops. We need more fast operations
for functional style programming, not less.

This PEP smacks of political requirements not practical ones. The
builtin name space is only crowded if we want to use the names for
something else.

Again this is a backward incompatible change. It's not broken and
doesn't need fixing.

-999
-- 
Robin Becker


From jb at cascade-sys.com  Mon Apr 29 15:20:29 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 12:20:29 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net>
Message-ID: <3CCD9CFD.E1869228@cascade-sys.com>

These changes sound like a BIG improvement over the original draft PEP.

I pretty much agree with the changes, with a few exceptions noted below.

Regards

--jb

Raymond Hettinger wrote:

> >From the comments, I've learned:
>
> 5.  oct() is another candidate based on the rarity of usage

Candidate perhaps but some people still like Octal.

This still is a gratuitious change and thus should not be implemented.

If it ain't really broke, you shouldn't fix it.

> 7. the language needs some means of deprecating so that clutter
>     does accumulate.  that means needs to be well documented,
>     have a long, slow phase-in, and have a mechasism for
>     restoring old behavior if it is ever needed.

Without necessarily disagreeing, I would seriously QUESTION this point
and invite discussion on it

Making it easier to phase changes in and out makes things easier only
for the most active Python users.  Changes themselves are hard on the
user base generally.  Furthermore, having a language where things like
base functionality routinely changes significantly from release to
release seems a Bad Thing.

Picture the poor newbie sitting down with his O'Reilly book, having just
installed 2.3 or 2.4 and can't figure out why some of the examples no
longer work.  O'Reilly has how many thousands of these books in the
pipeline?  I'm sure they're not anxious to issue a revised edition.
Steve Holden barely had time to get get the 2.1 appendix in his book.
This is not helping ANYBODY.

I frankly think it's ALARMING that people are becomming comfortable with
the notion that you can't interpret the meaning of programs without (a)
being up to date on how things change from release to release and (b)
paying close attention to the declarations or (super ugh) command
switches that might override defaults.  This is a much greater burden
than simply learning any particular version of the language.  It's a
moving target.  Of course users don't have to upgrade to each new
release and generally only use one version at a time.  But in trying to
use other people's code you're likely to encounter anything from old to
new, so you have the problem whether you want it or not.  Unless you
hide in your hole and only use code you write yourself.

As it is, we can no longer look at a fragment of code containing a
division operator and know what it means without checking version and
import declarations (which may appear ANYWHERE in the source).  I
frankly find this to be BIZARRE.

The more we add to this confusion the worse service we do to the user
community at large.  Easily phasing features in and out is clearly
something that ONLY benefits the most sophisticated users, leaving
beginners and non-programmers in the dust.  Some of the seeming experts
on this list cannot keep up, so how does the non-programmer have a
chance?

I think the world would be a better place if the language experienced a
lot FEWER non-backwards compatible changes.   Absolutely none is the
norm in most languages and would be an ideal here.  Also users would be
better off if we had a lot fewer releases altogether.  One major release
per year seems a healthy goal, with debug-only releases in between.
Even then, changes should focus on backwards compatible enhancements or
new features that don't break old colde rather than trying to fine-tune
existing features, even some of the few arguably regrettable ones.
Simply write them off to Python's rich heritage.

So, I think one could argue just the opposite of #7 -- that we should
look into policies that make it HARDER to deprecate and phase out old
functionality.  I don't see that it ever hurts to leave it in but
there's ALWAYS some pain to take it out.  None of the old vestiges are
so awful that they require change.  So don't do it.

IMHO.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From jeff at ccvcorp.com  Mon Apr  8 18:38:15 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 8 Apr 2002 15:38:15 -0700
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com> <1018304015.46066@cswreg.cos.agilent.com>
Message-ID: <MPG.171bbba6ee73df8c98969f@news.nwlink.com>

In article <1018304015.46066 at cswreg.cos.agilent.com>, 
weeks at vitus.scs.agilent.com says...
> Jeremy Hylton (jeremy at alum.mit.edu) wrote:
> : > Anyway, it increased visibility enough so that it is no longer possible to
> : > call a function and pass it the scope as two dictionaries, which I had 
> : > found handy on occasion.
> 
> : I'm not sure I follow the details here, but I don't see how the change
> : in the scope rules would affect a function unless it specifically
> : exploited nested scopes.  If you wrote a function that only uses
> : locals and globals, can you not call it just the way you always did?
> 
> Yes, the function continues to work.  But the documentation changes.
> Originally, calling the function with the arguments globals() and locals()
> passed to the function all the variable bindings visible at the point where
> the function was called.  That is a tidy concept, and I'm sorry to lose it.

The thought of anyone *expecting* that behavior is shocking and 
frightening to me.  For one thing, it violates the whole point of 
scoping.  Besides which, if you read your documentation 
carefully, the dictionary returned by locals(), at least, is not 
guaranteed to allow write-access.

If you want access to variables inside of a function, you should 
pass those specific variables as arguments.  If you've got lots 
of arguments in common between lots of functions, assemble the 
lot into a class.  If neither of these is appropriate, then you 
should redesign -- you're using a very ugly approach.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From roy at panix.com  Wed Apr 17 09:09:00 2002
From: roy at panix.com (Roy Smith)
Date: 17 Apr 2002 09:09:00 -0400
Subject: shipping a python app
Message-ID: <a9js5c$2m9$1@panix2.panix.com>

I've been given a task to write a small gui-based app which needs to
run on both NT and Unix platforms.  About all I know about NT is how
to spell it, and I've heard it's got some cool flight simulator games.

The scope of the development project rules out C/C++/Java because of
extremely limited development and test time available.  In most
peoples' minds, that means doing it in TCL/Tk, since it's "available
everywhere".

I'd love to write this in python/Tkinter, but it'll be a bit of an
uphill battle to convince management.  While I do a lot of python
development for internal tools, my company doesn't currently ship any
python-based products because of the impression that it's not widely
supported.

Then it turns out that for the existing TCL/Tk apps we ship, we don't
even depend on finding a pre-installed interpreter.  It might not be
there, it might be broken, it might be the wrong version, it might not
have the right libraries, etc, etc.  Turns out we package and ship a
TCL interpreter right in our distribution package, so the "available
everywhere" argument becomes a bit of a red herring.

So, that brings us around to what, exactly, would be involved in
shipping a stand-alone python app?  All my python development so far
has assumed that the run-time environment already exists, or can be
installed by the user.  That will not be the case here.  We'll need to
assume that nothing is there and package up whatever we'll need.  I'm
sure people have done this before.  Anybody willing to share your
experiences?

Fortunately, for this particular app, there's no intellectual property
being protected so we don't care if the source code is visible.


From fperez528 at yahoo.com  Sun Apr  7 23:24:47 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sun, 07 Apr 2002 21:24:47 -0600
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com> <a8q9bi$2vka$1@news.idiom.com> <d9ea2696.0204071920.1b9d75c4@posting.google.com>
Message-ID: <a8r62k$20b$1@peabody.colorado.edu>

distrex wrote:

> Don't take this the wrong way but I am a newbie and didn't understand
> 86% of what you just said. All I got was modules are classes functions
> and variables. and a few other things but that was to complicated for
> me (the newbie)
> 

Don't despair. The above was one of the best descriptions of modules I've ever 
read (but I'm not a newbie), to the point where I'd strongly encourage its 
author to contact the doc team and push for its inclusion in the official 
docs.

Just have a bit of patience, read it over again, slowly, go through the whole 
tutorial doing the examples and it will make sense. You may be a bit green 
still but it will come quickly, I'm sure. Python is one of the easiest 
languages on earth to learn: you'll be up and running in no time at all, and 
having fun programming!

Cheers,

f.


From aleax at aleax.it  Sun Apr  7 11:55:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 07 Apr 2002 15:55:15 GMT
Subject: ignored return codes
References: <E16tGLp-0007ou-00@mail.python.org> <mailman.1018014793.1604.python-list@python.org> <axir8.52642$pT1.1485864@news1.tin.it> <a8nq2n$8s3$0@216.39.172.122>
Message-ID: <DZZr8.69904$S52.2469494@news2.tin.it>

Bengt Richter wrote:
        ...
>>Button(buttonsFrame, text='Squirk', command=lambda: self.squirkIt(23, 45))
>>
>>you can code:
>>
>>def specificSquirker(): self.squirkIt(23, 45)
>                         ^-- return
> if you really want equivalent code, though maybe only side effects matter
> in the the example context?

Yes, the return code of command= callables is ignored (thus I never
code a return in those cases) -- sorry, I had dropped into Tkinter
mindset here.

>>Button(buttonsFrame, text='Squirk', command=specificSquirker)

Another nice alternative, btw, is

Button(buttonsFrame, text='Squirk', command=curry(self.squickIt, 23, 45))

Look up curry in the Python Cookbook if it only suggests Indian food to 
you...:-).


Alex




From aahz at pythoncraft.com  Sat Apr  6 17:31:47 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 17:31:47 -0500
Subject: I need to go to Malmö tomorrow
References: <mailman.1018126154.20514.python-list@python.org>
Message-ID: <a8nt0j$411$1@panix1.panix.com>

In article <mailman.1018126154.20514.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>
>I haven't forgotten you all, but it will be a while before I get to
>replying.  Alex and I have code to write together as well.  I get a
>week of pair programming with Alex!  Way Cool.

You're just trying to make us feel jealous.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From sholden at holdenweb.com  Wed Apr 17 12:51:07 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 12:51:07 -0400
Subject: Accessing mdb(access 2000) files on linux using python
References: <508cb8ae.0204170837.6f800a24@posting.google.com>
Message-ID: <ZLhv8.1231$Ip3.1214@atlpnn01.usenetserver.com>

"B" <martin_byrne2 at yahoo.co.uk> wrote in message
news:508cb8ae.0204170837.6f800a24 at posting.google.com...
> Hi,
> Just wondering if anyone knows how to manipulate mdb files stored on a
> linux webserver using python.Using win32com on windows is fine but on
> linux can i manipulate the database using pythons odbc connectivity or
> is there another way.
>
> Any suggestions, links or sample code would be of great help.
>
I'm sorry to say I suspect you're chasing a phantasm here. I'm not aware of
any such code. Neither do I know of a Unix ODBC driver for Jet database
files. I'll be delighted if somebody tells me I'm talking through my hat.

Of course, it's quite possible to migrate the data (using Python) from
Windows+Access to Linux+MySQL or some other such combination. I presume
there are specific reasons you don't want to do this (such as use of the
database over Samba by Windows-based systems)?

regards
 Steve





From orlov at diasoft.ru  Mon Apr 15 07:49:44 2002
From: orlov at diasoft.ru (Oleg Orlov11980884)
Date: Mon, 15 Apr 2002 15:49:44 +0400
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs>
Message-ID: <a9eekb$1p90$1@gavrilo.mtu.ru>

"Max Ischenko" <max at malva.com.uaREMOVE.IT> wrote in message
news:of1e9a.4vg.ln at cvs...
>
> Hi there.
>
> I have a Python program that uses data stored in XML file to generate some
> C++ code. The problem (or, more precisely, code smell) I have is that the
> actual generation algorithm is hardwired into python and therefore is
> cryptic, difficult to understand and maintain.
>
> Small example:
> def genBody(self, c):
> name = c.cxxClassName
> new = 'new %s(%s);' % (name, c.factoryCtorArg or '')
> if c.factoryNargs == 0:
> return '\treturn %s' % new
> body = '\t%s * c = %s\n' % (name, new)
> body = body + '\tc->setCmdArgs(args);\n'
> return '%s\treturn c;' % body
>
> I'd like to have the algorithm moved out from Python source.
> I've heard about XSLT language that is used to describe transformation of
> XML data into HTML. I'm seeking for something similar. The difference is
> that target language is C++ and not HTML.
>
> Any suggestions?

Try www.reportlab.co.uk/rl-cgi/viewcvs.cgi/rlextra/preppy/preppy.py

Oleg





From phd at phd.pp.ru  Tue Apr  9 10:56:48 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 18:56:48 +0400
Subject: pointer
In-Reply-To: <a8uukg$fuu$04$1@news.t-online.com>; from pete@dexterslab.de on Tue, Apr 09, 2002 at 04:45:08PM +0200
References: <a8uukg$fuu$04$1@news.t-online.com>
Message-ID: <20020409185648.F23382@phd.pp.ru>

On Tue, Apr 09, 2002 at 04:45:08PM +0200, Henning Peters wrote:
> is there anything similar to the pointer concept from c/c++ in python?

   In Python *everything* is a pointer. Really.

> i would like to store values in two different locations. both values should
> always be equal. the one could be a pointer on the other. is this possible
> to do this in python?

a = 1 # put a pointer to an Integer objetc into variable a
b = a # copy the pointer into b; now both b and a hold pointers to the same
      # Integer object

c = [a, b] # make a list holding two pointers

etc...

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




From ai.matrix at eudoramail.com  Tue Apr 23 22:27:15 2002
From: ai.matrix at eudoramail.com (Ramil D. Tolentino)
Date: Wed, 24 Apr 2002 10:27:15 +0800
Subject: Php vs Python
Message-ID: <LJEINIKGMNIGIAAA@shared1-mail.whowhere.com>

Could someone give me a run down between python and php. Which is easier to use? which is better or more powerful? Thanks


Join 18 million Eudora users by signing up for a free Eudora Web-Mail account at http://www.eudoramail.com




From maxm at mxm.dk  Tue Apr  9 06:08:12 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 09 Apr 2002 12:08:12 +0200
Subject: Discussion of Python midi library
References: <mailman.1018277417.26553.python-list@python.org>
Message-ID: <3CB2BD8C.5000606@mxm.dk>

Ken Seehof wrote:

> I have written a midi module that wraps part of the low-level Windows
> API.  I'm planning to release it to open source after we ship our
> commercial product that uses it.
> 
> It doesn't actually do any of the items you requested though, since
> I wrapped the midi input and output devices, but not midi streams or
> files.

I think that it might be very usefull for::

 >>     4) Play midifiles (especially on Windows)

Because what I meant here was not that it should play midifiles, but 
that it should play music in "real time" as generated from the midi library.

> I believe that you will find midi file processing in the Vaults:
> 
> http://www.vex.net/parnassus/

None of them did what I need, they where not very well written, or 
written in c/c++, and poorly documented. So I am starting from scracth. :-)

regards Max M



From gtaylor at lowebroadway.com  Tue Apr  9 10:48:03 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 9 Apr 2002 07:48:03 -0700
Subject: Bizarre keypressevent and keypress release behaviour
Message-ID: <f0fd5987.0204090648.256e23f4@posting.google.com>

I just upgraded to QT 3.03 from QT 2.something, and amongst the
incompatabilities I experienced, this is the one that I cannot trace.
keypressevents generally speaking do not work, but keypressrelease
generally speaking do. However, it's really only shift,control, and
alt I am interested in but get this:

For shift, only release works.
For Ctrl, both work,
For Alt,  press works, but release does not, but press it again, and
press does not work, but release does, and turn about, so basically
press and release work every other time, turn about.

I use PyQt 3 to hook up to it, when I used QT 2 and PyQt 2, I had no
such problems, I am using Python 2.2. Does anyone know what is going
on here or had a similar problem?

TIA 

Garry


From hgs at dmu.ac.uk  Thu Apr 18 12:12:12 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 17:12:12 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j47kn56ny0.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181648060.24683-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > It is worse
> >
> > Traceback (most recent call last):
> >   File "./Lib/test/test_socket.py", line 12, in ?
> >     import socket
> >   File "/home/hgs/Python-2.2.1/Lib/socket.py", line 41, in ?
> >     from _socket import *
> > ImportError: No module named _socket
>
> Here is the procedure to analyse this in detail:

Just to be clear, this is the CC=cc CXX=CC Forte6 compiler case:

 CC: WorkShop Compilers 5.0 98/12/15 C++ 5.0
 cc: WorkShop Compilers 5.0 98/12/15 C 5.0
>
> Theory 1: during building of _socket with setup.py, setup notices that
> building of _socket went wrong, and deletes it.

OK, analysing the make output I see 2 cc cmmands relating to building
the socket module.  These are now in a script, with -# added for verbose
output:

#!/bin/ksh -vx

# Script to rebuild the socket module
cc -# -DNDEBUG -O -DUSE_SSL=1 -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/socketmodule.c -o build/temp.solaris-2.7-sun4u-2.2/socketmodule.o
cc -# -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so

Results in:


Script started on Thu 18 Apr 2002 17:04:40 BST
neelix hgs 11 %> ./do_build_socket
#!/bin/ksh -vx

# Script to rebuild the socket module
cc -# -DNDEBUG -O -DUSE_SSL=1 -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/socketmodule.c -o build/temp.solaris-2.7-sun4u-2.2/socketmodule.o
+ cc -# -DNDEBUG -O -DUSE_SSL=1 -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/socketmodule.c -o build/temp.solaris-2.7-sun4u-2.2/socketmodule.o
### Note: LM_LICENSE_FILE = /micro_stripes_4.0/license.dat:/workshop_license/licenses_combined
### Note: NLSPATH = /progs/forte6/SUNWspro/bin/../WS6U2/bin/../lib/locale/%L/LC_MESSAGES/%N.cat:/progs/forte6/SUNWspro/bin/../WS6U2/bin/../../lib/locale/%L/LC_MESSAGES/%N.cat
###     command line files and options (expanded):
### -c -DNDEBUG -DUSE_SSL=1 -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -xO2 /home/hgs/Python-2.2.1/Modules/socketmodule.c -o build/temp.solaris-2.7-sun4u-2.2/socketmodule.o
/progs/forte6/SUNWspro/bin/../WS6U2/bin/acomp -i /home/hgs/Python-2.2.1/Modules/socketmodule.c -o /tmp/acompAAAWmaiMG -r /tmp/acompBAAXmaiMG -DNDEBUG -DUSE_SSL=1 -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -O2 -Qy -D__SunOS_5_7 -D__SUNPRO_C=0x530 -D__SVR4 -D__unix -D__sun -D__sparc -D__BUILTIN_VA_ARG_INCR -Xa -D__PRAGMA_REDEFINE_EXTNAME -Dunix -Dsun -Dsparc -D__RESTRICT -I/progs/forte6/SUNWspro/WS6U2/include/cc "-g/progs/forte6/SUNWspro/bin/../WS6U2/bin/cc -DNDEBUG -O -DUSE_SSL='1' -I/usr/local/ssl/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c -o build/temp.solaris-2.7-sun4u-2.2/socketmodule.o " -D__SUN_PREFETCH
/progs/forte6/SUNWspro/bin/../WS6U2/bin/iropt -F -o /tmp/iroptCAAYmaiMG /tmp/acompBAAXmaiMG -xarch=generic -xcache=generic -xchip=generic -xprefetch=no -xvector=no -xbuiltin=%none -O2 -I
/progs/forte6/SUNWspro/bin/../WS6U2/bin/cg -xcode=abs32 -O2 -Qy -xmemalign=4s -xprefetch=no -xvector=no -xarch=generic -xcache=generic -xchip=generic -oo build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -is /tmp/acompAAAWmaiMG -ir /tmp/iroptCAAYmaiMG -xbuiltin=%none
cc -# -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so
+ cc -# -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so
### Note: LM_LICENSE_FILE = /micro_stripes_4.0/license.dat:/workshop_license/licenses_combined
### Note: NLSPATH = /progs/forte6/SUNWspro/bin/../WS6U2/bin/../lib/locale/%L/LC_MESSAGES/%N.cat:/progs/forte6/SUNWspro/bin/../WS6U2/bin/../../lib/locale/%L/LC_MESSAGES/%N.cat
###     command line files and options (expanded):
### -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so
### Note: LD_LIBRARY_PATH = /progs/forte6/SUNWspro/lib:/ansys50/SC2.0.1:/usr/openwin/lib:/opt/panorama/lib
### Note: LD_RUN_PATH = <null>
/usr/ccs/bin/ld /progs/forte6/SUNWspro/WS6U2/lib/crti.o /progs/forte6/SUNWspro/WS6U2/lib/values-xa.o -o build/lib.solaris-2.7-sun4u-2.2/_socket.so -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -Y "P,/progs/forte6/SUNWspro/WS6U2/lib:/usr/ccs/lib:/usr/lib" -Qy /progs/forte6/SUNWspro/WS6U2/lib/crtn.o
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Thu 18 Apr 2002 17:05:17 BST
End of Results.
>
>   If that is the case (analyse setup.py output): Invoke the commands
>   to build _socket, and report whether this succeeds. If it does,
>   start ./python, and 'import socket'; report any errors that this

I can't tell if that worked:

Script started on Thu 18 Apr 2002 17:06:22 BST
neelix hgs 11 %> ./python
Python 2.2.1 (#1, Apr 18 2002, 16:33:40) [C] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import socket
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/home/hgs/Python-2.2.1/Lib/socket.py", line 41, in ?
    from _socket import *
ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/build/lib.solaris-2.7-sun4u-2.2/_socket.so: symbol __eprintf: referenced symbol not found
>>>
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Thu 18 Apr 2002 17:07:01 BST
>   gives.
>
> To work around this, we need
>
> Theory 2: building the socket module fails because setup.py detects
> presence of OpenSSL, yet the OpenSSL versions is to old for the socket
> module.

How old is acceptable ( before I disclose the version number publicly!!
:-) )
>
>   If that is the case, you'll need to edit Modules/Setup, and build
>   the socket module explicitly without SSL support (either shared or
>   static).
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you,
        Hugh






From logstx at bellatlantic.net  Sat Apr 13 20:07:28 2002
From: logstx at bellatlantic.net (logistix)
Date: Sun, 14 Apr 2002 00:07:28 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com>
Message-ID: <4L3u8.703$93.228@nwrddc04.gnilink.net>

There's a bit of a semantic distinction here.  In reality, bitwise
operations hit primitives like BYTES and WORDS (or DWORDS), not ints.  The
fact that ints are equivilent to words is just a conincidence that we take
advantage of.  The whole concept of "performing bitwise operations on
integers" is totally illogical (even though I do it too ;) ) but it's alot
easier to use ints as words instead of creating a distinct type.

Now a long integer is an arbitrary complex structure (in the sense that
processors don't support long integers at the machine code level), so it's
really impractical to perform some sort of official, definitive bitwise
operation on.

So what does Python do? Mimic the behavior to achieve expected results, even
though it's not performing a genuine bit shift.  Makes sense to me.

--
-

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote in message
news:3cb88ae7 at news.mhogaming.com...
> Since Python does not have unsigned numbers, I think Python should have a
> (Java-like) '>>>' operator.  The '>>>' means 'logical-shift-right', where
a zero
> is copied into the high bit.  This would be kind of nice when doing some
logical
> bit tweaking...
>
> Here is some code to implement the '>>>' operator. (Unfortunately, this
code
> does not work for long integers):
>
> # logical-shift-right operator:
> def lsr(i,n):
>     if n: i = i >> n ^ 0x80000000 >> n - 1
>     return i
>
> Although the '<<' and '>>' operators do not now "do the expected thing"
with
> long integers, they COULD if we introduced the '>>>' operator.  In order
to
> support the "all integers are the same" model, I think that they should.
Thus,
> for long integers '<<' and '>>' could give different answers than '*2' and
'/2'.
> (I don't know how to implement this properly for long integers without
breaking
> some old code though.)
>
> If this was discussed before, then it needs to be discussed again, as we
are now
> trying to provide seamless integration of the 'int' and 'long int'
types...
>
>
>




From kzaragoza at attbi.com  Wed Apr  3 23:59:24 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: Thu, 04 Apr 2002 04:59:24 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com>
Message-ID: <slrnaannkc.6ls.kzaragoza@localhost.localdomain>

In article <7xy9g4nk9w.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
> I think there will need to be a format code for truth values:
> 
>   ("%b" % True) => "True"
> 
> etc.

Maybe, but I'm not sure this is strictly necessary, since according to
the PEP:

    ("%s" % str(True)) => "True"

It's not as pretty as having a format specifier for it, but I'd
imagine it will do for the short term.  Then, of course, you have to
wonder what exactly

    ("%d" % True) => ?

should print.  After all, bools are ints too.  Should this be "1" or
an error of some sort?  And what if you've done some arithmetic on a
bool and end up with another value?  What would your proposed %b
format specifier print for

   val = int.__new__(bool,42)
   ("%b" % val) => ?

I really should stop thinking about these things and just go to
bed... :-)

-Kris

-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From theller at python.net  Fri Apr 19 08:10:43 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 19 Apr 2002 14:10:43 +0200
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
References: <mailman.1019071474.3193.python-list@python.org>
Message-ID: <a9p1g4$51jr8$1@ID-59885.news.dfncis.de>

"Trent Mick" <TrentM at ActiveState.com> wrote in message news:mailman.1019071474.3193.python-list at python.org...
> We are pleased to announce that versions 2.1.3 and 2.2.1 of
> ActivePython are now available for download from
> http://www.ActiveState.com/ActivePython
> Both of these are bugfix releases and supercede previous ActivePython
> releases.

Are the HTMLHelp files available separately? Can they be installed
separately?

Regards,

Thomas




From ark at research.att.com  Wed Apr  3 09:57:57 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 3 Apr 2002 14:57:57 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <ualr7hmlhkemab@news.supernews.com> <J7qT5iA2Rvq8EwLD@jessikat.fsnet.co.uk>
Message-ID: <yu991ydwx216.fsf@europa.research.att.com>

Robin> C is a very successful language. I don't think of its conditional
Robin> testing as a hack, but then I'm an engineer.

C99 has a bool type.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From max at alcyone.com  Tue Apr  9 14:04:43 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 09 Apr 2002 11:04:43 -0700
Subject: ANN: Twisted 0.16.0
References: <mailman.1017787747.12194.python-list@python.org> <3CAA8CFF.B6EDDCDE@alcyone.com> <8Fgs8.27735$Sw5.8757@atlpnn01.usenetserver.com> <23891c90.0204090100.38023e58@posting.google.com>
Message-ID: <3CB32D3B.74369D36@alcyone.com>

Paul Boddie wrote:

> Given that the announcement is also archived in "the public record", I
> wonder if the joke has had any effect on the number of people willing
> to evaluate the software. It probably isn't such a great idea to joke
> about software licensing if you want people to download and deploy
> your product.

It's also traditional that April Fool's Day only be perpetuated on April
1, and when found out, the perpetrators come clean.  Arguing or
continuing to drag on the "joke" only makes people wonder.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From a.zitzewitz at generic-home.de  Mon Apr  8 03:45:16 2002
From: a.zitzewitz at generic-home.de (Alexander v. Zitzewitz)
Date: Mon, 8 Apr 2002 09:45:16 +0200
Subject: Creating a DB app with mySql and Windows GUI
Message-ID: <a8rhml$1id$03$1@news.t-online.com>

Hello Python Freaks !

What components should I use for building a GUI Database App with a MySQL
Server and a GUI on Windows platforms with Python. The following components
are needed:

- a feasable IDE with Debugger interface
- a GUI Builder
- a MySQL inding

Please let me know, if anyone already did a similar project and what
experiences he/she made with using Python

Best regards

Alexander v. Zitzewitz




From fperez528 at yahoo.com  Mon Apr 22 18:49:56 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 16:49:56 -0600
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org>
Message-ID: <aa2460$534$1@peabody.colorado.edu>

holger krekel wrote:

> On Mon, Apr 22, 2002 at 11:39:24AM -0600, Fernando P?rez wrote:
>> Jaco Schieke wrote:
>> 
>> > Looks like the regexp in rlcompleter that searches the string does not
>> > work correctly.  After changing the line
>> > 
>> >> m = re.match(r"(\w+(\.\w+)*)\.(\w*)", text)
>> > 
>> > to
>> > 
>> >> m = re.match(r"([\w\[\]]+(\.[\w\[\]]+)*)\.(\w*)", text)
> 
> or maybe even r"(\S+(\.\w+)*)\.(\w*)" ? This would principally
> allow to add
> 
> "".<tab>
> 
> to work as expected. Of course you need to remove " and '
> from the delimiter list.
> 
> But then again, i wonder if the rlcompleter should not simply
> see the *whole* string and do the string-split according to delimiters
> itself? This might be more flexible for 'general purpose' completion.
> I consider the delimiters to be a rlcompleter property and
> not a readline property.

Well, your new regexp suggestion works great for me. You may want to repost 
this to python-dev or as a patch to SF, so that the powers-that-be have a say 
on the issue. For the time being, I'm using your regexp in my patched version 
of rlcompleter and running the following code:

delims = readline.get_completer_delims()
del_chars = "-/[]\"'" 
readline.set_completer_delims(delims.translate(string._idmap,del_chars))

This seems to work beautifully in all cases, completing attributes for things 
like 'hi', mylist[0] and even nested directory structures (b/c of the '/'). 
So thanks for the ideas! I'd like to see this incorporated in the standard 
rlcompleter code, as I don't see any problems caused by this (I may well be 
wrong though :)

Cheers,

f.


From aleax at aleax.it  Fri Apr  5 04:44:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 09:44:32 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze>
Message-ID: <4mer8.51712$pT1.1468090@news1.tin.it>

Peter Hayes wrote:

> Does anyone know if there is a version of Python that runs on the new
> Linux based Sharp SL5000 PDA?

Yes, assuming, as Sharp promised, that whatever was developed/ported
for the Zaurus 'developer' (less RAM) thingy runs on the 'real' SL5000
now that it's finally out.  Get it all at:

http://www.riverbankcomputing.co.uk/download/zaurus/


Reports of pystones and other Zaurus/SL5000 Python/PyQT experiences
are VERY welcome -- it's a really sexy little thingy and I'm SORELY
tempted to have a fling, but getting it in Italy isn't easy as yet...


Alex



From mal at lemburg.com  Sat Apr 20 05:10:48 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Sat, 20 Apr 2002 11:10:48 +0200
Subject: How to make regexes faster? (Python v. OmniMark)
References: <LNBBLJKPBEHFEDALKOLCCEKOOPAA.tim.one@comcast.net>
Message-ID: <3CC13098.47EB0E18@lemburg.com>

Tim Peters wrote:
> 
> [Donn Cave]
> > By the way, I'll second Johannes Stiehler's recommendation of
> > MxTextTools.  Definitely appropriate for SGML parsing, and much
> > better than regexps for extensive parsing in my opinion - not
> > just in terms of speed, but I suspect a more powerful way to
> > describe text patterns than regexps.
> 
> Yes, it is.  "More convenient" is arguable, though -- there's a steep
> learning curve, but then people often forget how hard it was to learn regexp
> syntax and pragmatics too.

BTW, in you care, the next version will have full Unicode support,
so parsing e.g. a Unicode XML file is possible too. I've also
added better support for writing tag tables: jump targets can
now be string labels, there's a tag table compiler and some
other interesting things in the new beta.

Also, all the other functions and tools in mxTextTools
are now Unicode aware.

If you're interested in the beta, please subscribe to the
egenix-users mailing list (see the web-site). The archives
have the latest announcements.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From dietmar at schwertberger.de  Sat Apr 13 10:55:37 2002
From: dietmar at schwertberger.de (Dietmar Schwertberger)
Date: Sat, 13 Apr 2002 16:55:37 +0200
Subject: Newbie: Background processing with a GUI?
References: <3CB821A8.AA3274AA@bellsouth.net>
Message-ID: <ant131437bc82%2U@schwertberger.freenet.de>

In article <3CB821A8.AA3274AA at bellsouth.net>, Jeff Layton
<URL:mailto:laytonjb at bellsouth.net> wrote:
>    I'm not sure the subject is entirely accurate, but I really don't
> know what to call what I want to do (that's really clear).
>    What I want to do is have a GUI (right now I have it written
> in wxPython) that displays some data in a grid (spreadsheet
> kind of thing) but is updated every x seconds. I've got the classes
> written that get the data and put it into a list. I've also got the
> classes written that create the GUI (with some attempts at
> methods that update the values in the grid). However, I don't
> know how to update the grid every x seconds. Any pointers or
> insight?
>    Another question that goes along with this is, how do I do this
> but allow the user to change the time between updates x?
Have a look at the wxPython demo:
Miscellaneous -> wxTimer


Regards,

Dietmar



From matt at greenroom.com.au  Tue Apr  9 01:10:58 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Tue, 09 Apr 2002 15:10:58 +1000
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com> <slrnab4m82.glr.mattdm@jadzia.bu.edu>
Message-ID: <3CB277E2.8060505@greenroom.com.au>

Matthew Miller wrote:

> Anthony_Barker <anthony_barker at hotmail.com> wrote:
> 
>>Python 2.2x is stable why does redhat still insist on releasing 1.5x?
>>All other distributions are now on 2.x.
>>It seems from the Redhat's skipjack beta 2 that it will only ship the
>>older version of python(1.5.2).


I'm confused also, for a while there, sure there were a number of apps
that were tied to this old release, but most of them seem to be updated.

I reckon that if you haven't made it compatible with python2.x then you
should have to do a HUGE days work and _ACTUALLY_ have to download the
1.5 package.

In addition to that (i promise i'll stop soon), it is a bit annoying 
that whenever i install red hat i have to

rm -f /usr/bin/python
ln -s /usr/bin/python2 /usr/bin/python

because they believe that more of us will use 1.5 than 2.1.

Aah well, thats my 2 cents.



From jeremy at alum.mit.edu  Wed Apr 10 23:52:19 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 10 Apr 2002 20:52:19 -0700
Subject: Strange behaviour of CO_NEWLOCALS and function
References: <pan.2002.04.10.17.29.41.396116.31396@cantal.lip6.fr>
Message-ID: <b0f083db.0204101952.2b58dc9e@posting.google.com>

Boris Boutillier <boris at cantal.lip6.fr> wrote in message news:<pan.2002.04.10.17.29.41.396116.31396 at cantal.lip6.fr>...
> Here is a little example, where I create a new codeobject from one of a
> function by just removing the CO_NEWLOCALS flags :

Boris, I've got no idea what you're trying to accomplish, and I think
I'm happier that way :-).

> The strange thing is that no new dictionnary is created as expected, but
> local variable have been deleted from this dictionnary, why is this ? Is
> there some way to modify my code to avoid this deletion ?

If you want to understand what CO_NEWLOCALS does, please read the
source of frameobject.c.  It's not documented, and I don't expect it
ever will be.  I think the basic idea is: If CO_NEWLOCALS is set,
create a new locals dict (f_locals attribute) for the frame.  If not,
use f_globals for f_locals.

In your example, I don't believe there's any deletion going on.  't'
is set after the call to locals(), so you shouldn't expect to see it. 
If you call locals() aftering binding 't', I expect you will see it.

Note that regardless of whether CO_NEWLOCALS is set, the code compiled
for a function body is going to use "fast locals."  The compiler
generates LOAD_FAST and STORE_FAST instructions that take an integer
argument that specifies the array index of the PyObject * in the
frame's array of local variables.

Jeremy


From jeremy at alum.mit.edu  Mon Apr 29 17:28:37 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 29 Apr 2002 14:28:37 -0700
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org> <mailman.1020076925.25849.python-list@python.org>
Message-ID: <b0f083db.0204291328.e22664e@posting.google.com>

Christian Tismer <tismer at tismer.com> wrote in message news:<mailman.1020076925.25849.python-list at python.org>...
> What we are talking about is a way to support PyZine,
> while at the same time providing a missing structure
> for Starship.

Why are these two connected? Couldn't people support PyZine by subscribing?

Jeremy


From ldw at us.ibm.com  Thu Apr  4 11:59:44 2002
From: ldw at us.ibm.com (Larry Whitley)
Date: Thu, 4 Apr 2002 10:59:44 -0600
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com> <3ouoaukmu32afgatil65cmcb7pvb8e5pcb@4ax.com>
Message-ID: <a8i0pv$bu6$1@news.rchland.ibm.com>

I made the command look like: "start /belownormal /wait cmd /c
myprog.exe -opt 3 >out3.txt".  No joy.
Then I tried: "cmd /c start /belownormal /wait myprog.exe -opt 3 >out3.txt".
Also no joy.

Finally I went for simplicity and tried just: "myprog.exe -opt 3 >out3.txt".
And it works.  The only problem with this is that messages printed out while
the threads are running are a little confusing.  I have only one window for
both threads and can't tell one message with another.  I guess I'll put up
with that for now.

Larry


"Michael Kelly" <mkelly2002 at adelphia.net> wrote in message
news:3ouoaukmu32afgatil65cmcb7pvb8e5pcb at 4ax.com...
> On Thu, 4 Apr 2002 08:53:00 -0600, "Larry Whitley" <ldw at us.ibm.com>
> wrote:
>
> >Unfortunately, while the programs appear to run, at completion their
output
> >files are empty. The files exist but contain no data as a result of the
run.
> >When I run the same program in a non-threaded fashion from a batch file
> >(without the start statement) I get output in the output files.  In this
> >case the command looks like this "myprog.exe -optoin 3.0 >out3.txt".
>
> Hi Larry.  I would take a shot at getting cmd.exe in on the act
> since command line redirection is a function of the command
> processor shell.  Perhaps even putting your command line in
> a batch file(with start and its settings and the redirection) and
> just calling cmd.exe batchfilename from the Python launcher.
> It should open a command prompt and run your batch and
> then close when done IIRC.
>
>
> Mike
>
> --
>
> "Only choice is an oxymoron."
>
> --




From aahz at pythoncraft.com  Thu Apr 25 21:29:48 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Apr 2002 21:29:48 -0400
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org>
Message-ID: <aaaaic$3b7$1@panix1.panix.com>

In article <mailman.1019781707.18338.python-list at python.org>,
James J. Besemer <jb at cascade-sys.com> wrote:
>Aahz wrote:
>>
>> Ewww!  Where did you get the idea that awk is a precursor of Python?
>
>Look up "precursor" in the dictionary and you'll see it's an accurate
>statement.

www.dictionary.com provides definitions that agree with my reaction.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From brian at sweetapp.com  Sat Apr 27 11:46:27 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 27 Apr 2002 08:46:27 -0700
Subject: Python's Lisp heritage
In-Reply-To: <tyfd6wl33n7.fsf@pcitapi22.cern.ch>
Message-ID: <002101c1ee02$b1da3ee0$445d4540@Dell2>

Jacek wrote: 
> My example was indended to disabuse James of the notion that
> parentheses are used in lisp "around sub expressions".

How would you convert the following Python expression to Lisp?

sin(x/2 + y/2)





From andrey at app.omsk.su  Tue Apr  9 12:02:51 2002
From: andrey at app.omsk.su (Andrey Panchenko)
Date: Tue, 9 Apr 2002 23:02:51 +0700
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org>
Message-ID: <a8v36u$j3p$1@news.nsk.su>

"Oleg Broytmann" <phd at phd.pp.ru> wrote:
> On Tue, Apr 09, 2002 at 04:45:08PM +0200, Henning Peters wrote:
> > is there anything similar to the pointer concept from c/c++ in python?
>
>    In Python *everything* is a pointer. Really.
>
> > i would like to store values in two different locations. both values
should
> > always be equal. the one could be a pointer on the other. is this
possible
> > to do this in python?
>
> a = 1 # put a pointer to an Integer objetc into variable a
> b = a # copy the pointer into b; now both b and a hold pointers to the
same
>       # Integer object
>
> c = [a, b] # make a list holding two pointers
>
I think it's rather an ambiguous statement. I can think that after

a = 1
b = a
a = 3

b must hold 3. But it's not true. IMHO it's better to say that the python
names are like pointers only for mutable objects.
(Caution - newbie opinion)

Andrey mailto:andrey at app.omsk.su





From brueckd at tbye.com  Sun Apr 14 19:42:11 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Sun, 14 Apr 2002 16:42:11 -0700 (PDT)
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <Pine.LNX.4.33.0204141639320.4987-100000@ns0.tbye.com>

On Sun, 14 Apr 2002, phil hunt wrote:

> It is not new features that are the problem, it is new features that
> break existing code.
>
> IMO until Python developers adopt a policy of being very reluctant
> to do this,

That already *is* their policy. 'Very reluctant to' != 'zero tolerance
for'.

-Dave





From jblazi at hotmail.com  Tue Apr  9 13:48:30 2002
From: jblazi at hotmail.com (Janos Blazi)
Date: Tue, 9 Apr 2002 19:48:30 +0200
Subject: PyQt question (colors)
References: <3cb2d4e7_11@news.newsgroups.com> <3cb2dff6$0$85444$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <3cb328e8$1_3@news.newsgroups.com>

> > If I say
> >
> >     stat_lab1.setBackgroundColor(QColor("DarkGreen"))
> >
> > and stat_lab is a QLabel, this has the required effect. But if I say
> >
> >     pb.setBackgroundColor(QColor("DarkGreen"))
> >
> > and pb is a QProgressBar object, so this has no effect at all. Can
anybody
> > tell me the reason for this? I use PyQt with Qt 2.3 on a MS-Win2000
> > system.
> >
>
> Using colors in Qt is quite complicated, unfortunately, and it has
> become even more complicated in Qt 3, where setBackgroundColor is
> so deprecated as to have disappeared from PyQt.
>
> In this case I think you will really have to read Troll Tech's
documentation
> to get a real understanding: it's not something that can be accurately
> explained in a Usenet posting. The upshot is that what you perceive
> to be the same thing -- the background of a label and a progressbar,
> are in the Qt system of widget painting different things.

Thank you for your answer.
I tried to read the Trolltech manuals (and your book as well by the way) but
to no avail. Ist it really so difficult to set the two colors of a
progressbar? Does that require 20 lines of code?

J.B.




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From vinay_sajip at yahoo.co.uk  Thu Apr 11 11:15:14 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: Thu, 11 Apr 2002 16:15:14 +0100
Subject: PEP 282: A Logging System
References: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru>
Message-ID: <002e01c1e16b$b0cd9d60$652b6992@alpha>

> On Thu, 11 Apr 2002, Andy Gimblett wrote:
>
> AG> On Wed, Apr 10, 2002 at 06:54:48PM +0100, Vinay Sajip wrote:
> AG>
> AG> > The logging module is not designed to use threading, or
> AG> for thread safety.
> AG>
> AG> Really?  It's _not_ thread safe?  Oh.  Will it be, ever?
>
Denis> Thread safety of logging module is very important. If it won't,
Denis> every module thread safe itself will become unsafe when using
Denis> logging.

Whoa - sorry, I got a bit careless there. What I meant to say was,

"the module is not designed to use threading, nor is it currently tested for
thread-safety".

That's to say, I expect it will be thread-safe in the not-too-distant
future, but I have not yet tested it in conditions such that I can
confidently comment on its thread-safety. When writing it I aimed not to use
instance data or global data (other than read-only) except where absolutely
necessary, so thread-safe operation should be straightforward to achieve.
The main areas of work to ensure thread-safe operation are being worked on
at the moment - they are:

(a) Serializing access to the hierarchy structure and logging level tables,
which are global.
(b) Checking the code in detail to ensure that there is no thread-unsafe
code.
(c) Allowing for on-the-fly reconfiguration by a thread while other threads
are using the system for logging.
(d) Coming up with a good test harness designed to try and break the module
in terms of thread-safety.

Currently, for example, if the logging system is initialized/configured in a
main/startup thread, then other threads should be able to log without
problems. The main thread would have created the logger hierarchy, the
handlers etc. During actual calls to loggers, all of the global data is used
in read-only mode and so no contention between threads should be expected.

Once all the above points are done and the module works against the tests, I
will add the test harness to the distribution and make appropriate changes
to the documentation.

Regards

Vinay Sajip


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From ftobin at neverending.org  Thu Apr 11 01:39:59 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Thu, 11 Apr 2002 05:39:59 GMT
Subject: SocketServer.DatagramRequestHandler.setup bug?
Message-ID: <20020411010624.N16832-100000@palanthas.neverending.org>

In SocketServer.DatagramRequestHandler.setup, it calls:
        self.wfile = StringIO.StringIO(self.packet)

It doesn't make sense to initialize wfile to anything; one probably wants
to start off with a 'clean' packet.  Should this not rather be:
        self.wfile = StringIO.StringIO('')

?

-- 
Frank Tobin		http://www.neverending.org/~ftobin/



From prova at prova.it  Tue Apr 30 14:11:24 2002
From: prova at prova.it (Max)
Date: Tue, 30 Apr 2002 18:11:24 GMT
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com> <f0fd5987.0204300105.51529004@posting.google.com> <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk>
Message-ID: <g7Bz8.35106$8D3.1017931@news1.tin.it>

----- Original Message -----
From: "JohnO" <jo at johnochiltree.uklinux.net>
Newsgroups: comp.lang.python
Sent: Tuesday, April 30, 2002 11:32 AM
Subject: Re: Python vs. Perl, which is better to learn?


> Garry Taylor wrote:

[...]

> I don't know if anyone else has noticed this but I've found that operators
> and sys admins use perl and that developers tend towards python. Also sys
> admins rarely utilise indirection but developers do.
>
> John

Interesting... i had the same feeling!
I'm pretty new to Python, but it's very close to the way i think... i can
concentrate on the problem and not on how to implement my solution... very
very powerful for an analyst/programmer....




From martin at v.loewis.de  Wed Apr 17 14:40:06 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 20:40:06 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <m3n0w2yxrt.fsf@mira.informatik.hu-berlin.de>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:

> Not that it's a big deal since there's many ways to do that, but can I use
> something like
>     verboseGender = (gender == 'm') ? 'male' : 'female'
> in Python?

http://www.python.org/doc/FAQ.html#4.16

HTH,
Martin


From gbreed at cix.compulink.co.uk  Mon Apr 29 05:35:48 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Mon, 29 Apr 2002 09:35:48 +0000 (UTC)
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <aaj45k$odt$1@thorium.cix.co.uk>

Raymond Hettinger wrote:

>     Pow() and divmod() are more related to math module functions than 
> any
>     of the built-in functions.  In general, they are rarely used.  
> Because
>     the ** operator is available, the pow() function is rarely called
>     directly.

Hold on there!  The easiest way of doing integer division that works 
across "from __future__ import division" is to use divmod.  Now, as soon 
as that change becomes the default, you're proposing to deprecate it?  
That means it'll get even harder to do consistent integer division across 
2.1 and 2.5.  

The rationale's very brief about the disadvantages of a bloated 
__builtins__.  It isn't like they're reserved words.  We don't need to be 
familiar with them all, because we can always check the library reference 
if we see one we don't know.  Or print the docstring.  There already seem 
to be 114 entries in __builtins__, so removing a mere 5 functions isn't 
going to make it dramatically simpler.

I don't see any benefits to outweigh the pain this is going to cause some 
of us.


                    Graham

       <http://www.microtonal.co.uk/>


From emmanuel at onda-latina.com  Sat Apr 13 04:50:29 2002
From: emmanuel at onda-latina.com (manouche)
Date: Sat, 13 Apr 2002 10:50:29 +0200
Subject: problem with wxHtmlWindow
Message-ID: <a98rht$pok$1@wanadoo.fr>

What are the libraries I have to import in order to use wxHtmlWindow?

I tried importing the libraries of the example

from wxPython.wx import *
from wxPython.html import *
import wxPython.lib.wxpTag

but my python shell says that:

Traceback (most recent call last):
  File "D:\prog\boa1\wxApp1.py", line 22, in ?
    main()
  File "D:\prog\boa1\wxApp1.py", line 18, in main
    application = BoaApp(0)
  File "C:\PYTHON22\lib\site-packages\wxPython\wx.py", line 1587, in
__init__
    _wxStart(self.OnInit)
  File "D:\prog\boa1\wxApp1.py", line 12, in OnInit
    self.main = wxFrame1.create(None)
  File "D:\prog\boa1\wxFrame1.py", line 6, in create
  File "D:\prog\boa1\wxFrame1.py", line 23, in __init__
  File "D:\prog\boa1\wxFrame1.py", line 19, in _init_ctrls
    self.SetClientSize(wxSize(624, 642))
NameError: global name 'wxHtmlWindow' is not defined

Thanks for your help





From starx at pacbell.net  Wed Apr 17 06:27:20 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 17 Apr 2002 10:27:20 GMT
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org> <a9icab$phq$1@nntp9.atl.mindspring.net>
Message-ID: <Xns91F32355A8FA9RASXnewsDFE1@130.133.1.4>

Andrew Dalke || Tue 16 Apr 2002 04:35:54p:

> sean.bowman at acm.org:
>>I'm wondering if there's a way to optimize code generated using the
>>compiler module.  for example, if I say::
>>
>> import dis
>> from parser import compileast, expr
>>
>> code = compileast(expr('x+3+7.'))
>> print dis.dis(code)
>>
>>I notice that the constants aren't folded.  is there a way to do this, 
or
>>do I have to fiddle with the ast myself?
> 
> Suppose you have
> 
> class Strange:
>     def __add__(self, other):
>         return self
> x = Strange()
> 
> Then x+3+7.0 = (x+3)+7.0 = x+7.0 = x
> 
> The compiler can't know what 'x' might be so can't fold constants
> in this case.  You could if it was '3+7.0+x'.

In this case, wouldn't 3+7.0+x be a Typeerror or something? Int doesn't 
know how to add with a Strange, and Strange dosen't have an upgrading 
function... or do we assume addition is always commutitive and try '3+x+
7.0' (effectively), therefore breaking the optimization mentioned by the 
other... =)


-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From czaban at dgt.com.pl  Fri Apr 19 05:04:41 2002
From: czaban at dgt.com.pl (Piotr Czabañski)
Date: Fri, 19 Apr 2002 11:04:41 +0200
Subject: kinterbasdb
Message-ID: <a9omib$sv1$1@news.tpi.pl>

ProgrammingError: (-104, 'execute.isc_dsql_prepare: Dynamic SQL Error. SQL
error code = -104. token size exceeds limit. ')

How to get round it? The command SQL string has 1352 character.




From gerhard at bigfoot.de  Mon Apr 15 10:34:22 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 15 Apr 2002 14:34:22 GMT
Subject: Looking for GUI Form Builder for Python
References: <a9en77$4vf8@imsp212.netvigator.com>
Message-ID: <slrnablp7d.fa.gerhard@lilith.my-fqdn.de>

Calvin wrote in comp.lang.python:
> Dear all,
> Can I build Python TK GUI interface using a form builder. Anyone know such
> builder. I just want to build my interface in a visual way.

Pythonware offers a commercial solution for this as part of their IDE.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From jeff at ccvcorp.com  Wed Apr 10 16:06:41 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 10 Apr 2002 13:06:41 -0700
Subject: Creating objects in thread created with subclass
References: <mailman.1018401560.15646.python-list@python.org>
Message-ID: <MPG.171e3b1bd28320e19896a9@news.nwlink.com>

In article <mailman.1018401560.15646.python-list at python.org>, 
narnett at mccmedia.com says...
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Jeff Shannon
> 
> ...
> 
> > No, that is definitely *not* the right way to do it -- as I keep
> > saying, exec/eval() are very rarely the right way to do it,
> > whatever "it" may be.  ;)
> 
> Ah.  Well.  Then as long as we're on the subject... I couldn't figure out a
> way to build an insert command for MySQLdb without it.  Given a list of
> tuples to insert, I need to give the database handle a statement that looks
> like this:
> 
> self.dbh.execute("INSERT INTO Foo (fieldnames) VALUES (%s,[%s])",(data[1:]))
> 
> The first tuple in the list of tuples is the fieldnames, which can vary.
> Thus, the number of "%s" occurrences in the statement has to be calculated.
> Is there a way to create this insertion statement without using exec?

Okay, first I should warn you that I don't know SQL and haven't 
used any DB modules yet, but...  you need to dynamically *create* 
the string that you pass to execute, but that doesn't necessarily 
mean that you need to dynamically execute that string, if you get 
the difference. ;)

Try something like this:

mystring = "INSERT INTO Foo (fieldnames) VALUES (%s"
mystring +=  ",%s" * (len(data)-2) + ")"
self.dbh.execute(mystring, (data[1:]))

(Note that I'm subtracting 2 from the length of data -- one for 
the fieldnames tuple that you're excluding with your slice, and 
one for the initial '%s' which does not want a comma before it.)

There's probably plenty of ways to clean this up (and speed it 
up, too), and I won't place any bets on whether this is exactly 
the format you need the string to be in, but this general 
approach should work.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mwh at python.net  Fri Apr  5 04:28:02 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 5 Apr 2002 09:28:02 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de> <3cab55f4.61190667@news.laplaza.org> <a8fn61$s1nk7$1@ID-41653.news.dfncis.de> <lky9g4o3z7.fsf@elios.maths.bris.ac.uk> <mailman.1017914955.11127.python-list@python.org>
Message-ID: <ulmc2sd6w.fsf@python.net>

Andy Gimblett <gimbo at ftech.net> writes:

> On Wed, Apr 03, 2002 at 09:40:45PM +0000, Michael Hudson wrote:
> > Tobias Klausmann <klausman-un030402 at tuts.net> writes:
> > 
> > > So far, I've only written libraries/modules (no shebang) or console
> > > apps that are at least semi-interactive. But if I'd write some such
> > > utility that might run cronned, I'd surely stumble across the system
> > > where cron doesn't know about /usr/local/bin/python15 =)
> > 
> > The distutils can help here.
> 
> How can distutils help with the problem of what to put on the first
> line of my scripts?

Because if you go

python setup.py install_scripts

said scripts get their first lines munged to point to the python used
to install them.

Cheers,
M.

-- 
58. Fools ignore complexity. Pragmatists suffer it. Some can avoid
    it. Geniuses remove it.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From guido at python.org  Sat Apr  6 20:56:23 2002
From: guido at python.org (Guido van Rossum)
Date: Sun, 07 Apr 2002 01:56:23 GMT
Subject: Booleans, integer division, backwards compatibility; where is 
 Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <1018120666.804421@yasure>
Message-ID: <3CAFA81B.B9E4AF9@python.org>

Donn Cave wrote:
> I sure fail to see the point of this middle ground.  On one hand, you
> have a language that is manifestly very useful to programmers in a variety
> of disciplines but fails to be the perfect language.  On the other, you
> have a vision of the perfect language, which in principle ought to be
> more useful but definitely will break some significant amount of code
> written today.  In between, you have all the problems of both - breaks
> code today, doesn't prevent tomorrow's breakage, and isn't the perfect
> language anyway.  Major Python heads fight tooth and nail against changes
> that they might well accept without complaint as features of a really
> overhauled language.  Developers turn away, platforms like Redhat who
> could ship Python are put in a difficult position.  In the present release
> schedule, "don't upgrade" may be easy to say but it isn't really supported -
> neither at the core, where Tim's quick to point out that we're on our own
> for bug fixes, nor among the external contributors who have no notion that
> any such thing is going on, nor among client sites who are apt to install
> whatever's current.  Is this fun?  Doesn't sound like it.  Will it be more
> fun to artificially prolong the process?  Surely it will not.  Why not
> just do what you have to do, and concede that a few releases hence will
> be bleeding edge, not for general production use?

You have a point, although a problem is that it's not like I know exactly
where I want the language to end up -- I'm continually updating the
direction
based upon feedback.  I don't want to end up with the second system
syndrome,
where everything gets rewritten and redesigned from scratch, but the result
is so different that it takes years to get the user community to migrate.

We're discussing this in python-dev right now, I'd welcome your views there
(note that python-dev no longer has a closed-subscription policy).

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


From jasonjurkowski at yahoo.com  Mon Apr  8 19:39:51 2002
From: jasonjurkowski at yahoo.com (jason jurkowski)
Date: Mon, 08 Apr 2002 19:39:51 -0400
Subject: Lex and Yacc for Python
References: <slrnab19hf.23u.Alex@Tron.ddeus.lan>
Message-ID: <3CB22A47.2060409@yahoo.com>

Alex,
You might want to look at Beazly's PLY.  Its implemented entirely in Python.
http://systems.cs.uchicago.edu/ply/
Jason

Oleksandr Kozachuk wrote:
> Hi all!
> 
> I need something, comparable to Yacc/Lex. Or something for
> Context-Free-Grammar. Can any one recommend me anything?
> 
> 
> Thanks,
> Alex.




From sjmachin at lexicon.net  Wed Apr 24 22:41:21 2002
From: sjmachin at lexicon.net (John Machin)
Date: 24 Apr 2002 19:41:21 -0700
Subject: Quick: Format Specs in %
References: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
Message-ID: <c76ff6fc.0204241841.57f83bb5@posting.google.com>

Philip Swartzleonard <starx at pacbell.net> wrote in message news:<Xns91FA8B766EF98RASXnewsDFE1 at 130.133.1.4>...
> '%2f' % some_number pads the number with 2 blank spaces
> '%.2f' % some_number fixes the number to 2 decimal places
> 
> the problem:
> '%2.2f' % some_number only does the second of those two things. I need 
> both. I tried looking up the % operator in the index of the language 
> reference, but couldn't find it.

The language reference manual talks about % only as a numeric
remainder/modulus operator. What you are looking for is cunningly
hidden in the library reference manual, in "2.2.6.2 String Formatting
Operations".

If you have a Windows PC, grab a copy of Hernan Foffani's help kit in
Windows .chm format (or grab the Activestate distribution of Python
instead of the python.org distribution) -- the .chm allows
search/indexing across all the manuals.

For the meaning of say %8.2f, given the heritage, you could look in
any C book under "printf" -- interestingly, Python seems to have also
cloned C's documentation methodology: K&R v2 has an entry for "%
modulus operator" in its index, but not "% thingy in printf and
scanf".


From aahz at pythoncraft.com  Thu Apr 25 11:48:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Apr 2002 11:48:16 -0400
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <j4helzdfbl.fsf@informatik.hu-berlin.de>
Message-ID: <aa98g0$9i4$1@panix3.panix.com>

In article <j4helzdfbl.fsf at informatik.hu-berlin.de>,
Martin v. =?iso-8859-1?q?L=F6wis?= <loewis at informatik.hu-berlin.de> wrote:
>"Adonis Vargas" <deltapigz at telocity.com> writes:
>>
>> i am just starting my venture into XML as a means of document output
>> and soon as a form of internet protocols. and i would like to know
>> how i can parse XML in Python? now i know i have access to SAX and
>> DOM, but how do i use them? i have looked into the documentation,
>> and it seems to be written in a martian language, and i am looking
>> in doing basic stuff just regular nodes, in particular. any
>> link/sample usage just anything to get my feet wet in this subject,
>> im a decent Python programmer, so there shouldnt be too much
>> problems, just need something i can work with.
>
>Have you looked at the HOWTO? Which part did you not understand (just
>indicate the first unclear sentence)?

What HOWTO?  There isn't one at http://py-howto.sourceforge.net/
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From googlegroups at sixtyten.org  Mon Apr 22 10:28:26 2002
From: googlegroups at sixtyten.org (Allan Crooks)
Date: 22 Apr 2002 07:28:26 -0700
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com>
Message-ID: <304d20df.0204220628.6aa88bc2@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<a9ut8d$dql$1 at panix3.panix.com>...
> In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
> Allan Crooks <googlegroups at sixtyten.org> wrote:
> >
> >Another thought I've had is doing it by specifying, say for module
> >'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
> >what versions of Python it is running on, and then compiling all
> >modules it can deal with, and then importing all objects into it's
> >namespace.
> 
> Yup.  But do it with a package.

Umm... pardon my ignorance, but what exactly is a package in terms of
Python? I assume it's just a module that does some form of importing
of sub-modules (perhaps it uses "from module import *")?

Allan.


From jochen at unc.edu  Tue Apr  9 19:38:06 2002
From: jochen at unc.edu (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 09 Apr 2002 19:38:06 -0400
Subject: python 2.2.1 on Cygwin
Message-ID: <ly7kngfnoh.fsf@bock.chem.unc.edu>

I have just compiled latest cvs release22-maint on Cygwin.  I get the
following test results:
,----
| 149 tests OK.
| 7 tests failed:
|     test_mimetypes test_mmap test_pyclbr test_socket test_sundry
|     test_urllib test_urllib2
| 31 tests skipped:
|     test___all__ test_al test_asynchat test_bsddb test_cd test_cgi
|     test_cl test_curses test_dbm test_dl test_fork1 test_gl
|     test_imgfile test_largefile test_linuxaudiodev test_locale
|     test_minidom test_nis test_ntpath test_pyexpat test_sax
|     test_socket_ssl test_socketserver test_sunaudiodev test_thread
|     test_threaded_import test_threadedtempfile test_threading
|     test_unicode_file test_winreg test_winsound
`----

When I run the individual tests I get:
,----[python Lib/test/test_mimetypes.py]
| test_data_urls (__main__.MimeTypesTestCase) ... ok
| test_default_data (__main__.MimeTypesTestCase) ... ok
| test_file_parsing (__main__.MimeTypesTestCase) ... ok
| test_non_standard_types (__main__.MimeTypesTestCase) ... ok
| 
| ----------------------------------------------------------------------
| Ran 4 tests in 0.092s
| 
| OK
`----

,----[python Lib/test/test_mmap.py]
| <type 'mmap.mmap'>
|   Position of foo: 1.0 pages
|   Length of file: 2.0 pages
|   Contents of byte 0: '\x00'
|   Contents of first 3 bytes: '\x00\x00\x00'
| 
|   Modifying file's content...
|   Contents of byte 0: '3'
|   Contents of first 3 bytes: '3\x00\x00'
|   Contents of second page: '\x00foobar\x00'
|   Regex match on mmap (page start, length of match): 1.0 6
|   Seek to zeroth byte
|   Seek to 42nd byte
|   Seek to last byte
|   Try to seek to negative position...
|   Try to seek beyond end of mmap...
|   Try to seek to negative position...
|   Attempting resize()
|   Creating 10 byte test data file.
|   Opening mmap with access=ACCESS_READ
|   Ensuring that readonly mmap can't be slice assigned.
|   Ensuring that readonly mmap can't be item assigned.
|   Ensuring that readonly mmap can't be write() to.
|   Ensuring that readonly mmap can't be write_byte() to.
|   Ensuring that readonly mmap can't be resized.
|   Opening mmap with access=ACCESS_WRITE
|   Modifying write-through memory map.
|   Opening mmap with access=ACCESS_COPY
|   Modifying copy-on-write memory map.
|   Ensuring copy-on-write maps cannot be resized.
|   Ensuring invalid access parameter raises exception.
| Traceback (most recent call last):
|   File "Lib/test/test_mmap.py", line 274, in ?
|     test_both()
|   File "Lib/test/test_mmap.py", line 249, in test_both
|     f = open(TESTFN, 'w+')
| IOError: [Errno 13] Permission denied: '@test'
`----
(this is on a local filesystem, one-user on the machine, me)

,----[python Lib/test/test_pyclbr.py]
| test_easy (__main__.PyclbrTest) ... ok
| test_others (__main__.PyclbrTest) ... ok
| 
| ----------------------------------------------------------------------
| Ran 2 tests in 2.925s
| 
| OK
`----

,----[python Lib/test/test_socket.py]
| socket.error
| BOCK
| 152.2.48.182
| bock.chem.unc.edu [] ['152.2.48.182']
| ['bock.chem.unc.edu']
| 23
| parent accepting
| child connecting
| connected by ('127.0.0.1', 3547)
| fileno: 4
| peer: ('127.0.0.1', 3547)
| sockname: ('127.0.0.1', 50007)
| file obj: <open file '<socket>', mode 'r' at 0x100dd008>
| received: socket test
`----

the following gave no output:
,----
| python Lib/test/test_sundry.py
| python Lib/test/test_urllib.py
| python Lib/test/test_urllib2.py
`----

I am not sure why test reported to fail seem to work ok?

Anyway, hope it helps.  Please cc me if you have questions or need
more info.

Greetings,
Jochen
-- 
University of North Carolina                       phone: +1-919-962-4403
Department of Chemistry                            phone: +1-919-962-1579
Venable Hall CB#3290 (Kenan C148)                    fax: +1-919-843-6041
Chapel Hill, NC 27599, USA                            GnuPG key: 44BCCD8E


From aleax at aleax.it  Mon Apr  8 06:37:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 10:37:24 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8pqal$ftm$1@panix1.panix.com> <7xpu1bh481.fsf@ruckus.brouhaha.com> <a8q0g1$qu5$1@panix1.panix.com> <a8qqf7$p4h$0@216.39.172.122>
Message-ID: <Epes8.59265$pT1.1705378@news1.tin.it>

Bengt Richter wrote:

> On 7 Apr 2002 13:43:29 -0400, aahz at pythoncraft.com (Aahz) wrote:
> [...]
>>
>>Assignments in expressions will never happen.
>>
> I guess you mean they won't be spelled concisely ;-)
> 
>  >>> a='eigh'
>  >>> a
>  'eigh'
>  >>> locals().__setitem__('a','bee') and 0 or a
>  'bee'

Now try that same, identical expression inside a 
function, will you...?

You can't assume locals() is modifiable at all:
attempts to modify it may fail silently and/or
they may raise exceptions ("undefined behavior").


Alex





From martin_byrne2 at yahoo.co.uk  Thu Apr 18 04:01:22 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 18 Apr 2002 01:01:22 -0700
Subject: Accessing mdb(access 2000) files on linux using python
References: <508cb8ae.0204170837.6f800a24@posting.google.com> <ZLhv8.1231$Ip3.1214@atlpnn01.usenetserver.com>
Message-ID: <508cb8ae.0204180001.2f9abf92@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<ZLhv8.1231$Ip3.1214 at atlpnn01.usenetserver.com>...
> "B" <martin_byrne2 at yahoo.co.uk> wrote in message
> news:508cb8ae.0204170837.6f800a24 at posting.google.com...
> > Hi,
> > Just wondering if anyone knows how to manipulate mdb files stored on a
> > linux webserver using python.Using win32com on windows is fine but on
> > linux can i manipulate the database using pythons odbc connectivity or
> > is there another way.
> >
> > Any suggestions, links or sample code would be of great help.
> >
> I'm sorry to say I suspect you're chasing a phantasm here. I'm not aware of
> any such code. Neither do I know of a Unix ODBC driver for Jet database
> files. I'll be delighted if somebody tells me I'm talking through my hat.
> 
> Of course, it's quite possible to migrate the data (using Python) from
> Windows+Access to Linux+MySQL or some other such combination. I presume
> there are specific reasons you don't want to do this (such as use of the
> database over Samba by Windows-based systems)?
> 
> regards
>  Steve

Yeah I'd like to keep with access if at all possible. The reasons for
this are as follows. The databases are being used to drive websites
for certain customers of ours. The current setup is that each site has
its own mdb database which the customer can change themselves, such as
changing the products they are selling,vat rates,delivery charges,
etc. This makes things easier for both parties(less work for us and
more control for the customers).
At the moment our webserver is IIS but we are hoping to migrate to
apache in the near future and there we have the problem.
I had guessed that accessing mdb databases on linux was a no go, just
thought there might be an obscure way to do it.
Looks like I'll have to write some sort of mysql frontend before we
change to apache.

Thanks for the help
B


From jriveramerla at yahoo.com  Sun Apr  7 00:47:29 2002
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 6 Apr 2002 21:47:29 -0800
Subject: Compatibilty table between BOA, wxPython, Python?
Message-ID: <11e94203.0204062147.5d2281b9@posting.google.com>

Hi...
Does somebody knows where is a table to know wich version of wxPython
is compatible with wich version of Python, and BOA, I  have tried to
find this with no success.

When I try start BOA using wxPython 2.3.2.1 and python 2.2, I get an
error.

I want to keep using python, and use the last version available to use
Python, wxPython, and BOA.

Thanks...


From rxg218 at psu.edu  Sat Apr  6 08:35:59 2002
From: rxg218 at psu.edu (Rajarshi Guha)
Date: Sat, 06 Apr 2002 08:35:59 -0500
Subject: a problem with subclassing a dict
Message-ID: <a8o0ro$1qoa@r02n01.cac.psu.edu>

Hi,
  I was playing with some code from a tutorial by GvR 
(http://www.python.org/2.2/descrintro.html) - the relevant code is:

class mydict(dict):

    def __init__(self, default=None):
        dict.__init__(self)
        self.default = default

    def __getitem__(self, key):
        try:
            return dict.__getitem__(self, key)
        except KeyError:
            return self.default

md = mydict("Sorry  - no such key!!")
md = {1:2}

print md[1]
print md[2]
---------------------------------

I get:

2
Traceback (most recent call last):
  File "p2.py", line 22, in ?
    print md[2]
KeyError: 2


Should'nt I be seeing a message instead of the Traceback?

My python is compiled  on a RH 7.2 system, with GCC 3.0 - 

Python 2.2.1c2 (#1, Apr  6 2002, 07:51:35) 
[GCC 3.0.2 20010905 (Red Hat Linux 7.1 3.0.1-3)]linux2

Any suggestions would be appreciated.
TIA,


-------------------------------------------------------------------
Rajarshi Guha                  | email: rajarshi at presidency.com
417 Davey Laboratory           | web  : http://jijo.cjb.net
Department of Chemistry        | ICQ  : 123242928
Pennsylvania State University  | AIM  : LoverOfPanda
-------------------------------------------------------------------
GPG Fingerprint: DCCB 4D1A 5A8B 2F5A B5F6  1F9E CDC4 5574 9017 AF2A 
Public Key     : http://pgp.mit.edu/
-------------------------------------------------------------------
A meeting is an event at which the minutes are kept and the hours are lost.



From gerhard at bigfoot.de  Tue Apr 16 20:44:38 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 17 Apr 2002 02:44:38 +0200
Subject: list_a-list_b
In-Reply-To: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>
References: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>
Message-ID: <20020417004437.GA615@lilith.my-fqdn.de>

* John <nobody at nobody.com> [2002-04-16 19:30 -0500]:
> 
> Hi, I need a help in writing a function that does
> 
> list_a=[1,2,4], list_b=[2,4]
> list_minus(list_a,list_b)
> print list_a #will print out a=[1] after removing all elements ([2,4])
> appearing in list_b

Ok.

>>> l1 = [3,4,5,6]
>>> l2 = [4,5]

Do you still need the original list? If yes, you'd better make a copy of
it:

>>> l3 = l1[:]

Then remove all elements of l2 in l3 like this:

>>> for item in l2:
...     l3.remove(item)
>>> l3
[3, 6]

To fully understand what's going on here, you can read in the
documentation about lists. The distinction about mutable/immutable
objects comes into play here, too: lists are mutable => you change them
"in-place". That's also the reason why l3 = l1[:] creates a copy, but
l3 = l1 would not.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From egbert at bork.demon.nl  Tue Apr 30 06:41:42 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Tue, 30 Apr 2002 12:41:42 +0200
Subject: Tkinter-helper-gui and a main class: their communication
Message-ID: <20020430124141.A560@bork.demon.nl>

The experiment below illustrates my problem, but the idea is this:
- a class uses a helper-gui-class for collecting data
- I want to use a helper-class for building the gui in stead of
  just a method that does the job, because the helper-class
  inherits from a super-class
- each time I push a button in the gui it should send a signal 
  to the main class: new data are available, you may come and use them
- the gui is non-modal: it stays on screen eternally, waiting for data

It doesn't work: the main class turtle does not receive any signal,
or at least the method buttonwork() is never used.

Is there a simple solution (which one), or is my whole idea flawed,
and why ?

egbert
--------------------------------------------------------------------
#!/usr/bin/python
import exceptions
from Tkinter import *
root = Tk()
#
class ButtonPush(exceptions.Exception):
    def __init__(self,args=None):
        self.args = args
#
class Amphibian_gui:
    def __init__(self):
        pass
#
class Turtle_gui(Amphibian_gui):
    def __init__(self):
        Amphibian_gui.__init__(self)
        knop=Button(root,text='New Turtle',command=self.pushed)
        knop.grid(row=0,column=0)
    def pushed(event=None):
        print "The button raises the ButtonPush exception"
        raise ButtonPush, "Here comes a new turtle"
#
class Turtle:
    def __init__(self):
        try:
            gui = Turtle_gui()
            print "I am trying ..."
        except Buttonpush, e:
            buttonwork()
    def buttonwork(self):
        print "main class Turtle sees the The ButtonPush exception"
        # do something useful with this knowledge

turtle = Turtle()
root.mainloop()

-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From aleax at aleax.it  Tue Apr 30 11:58:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 15:58:53 GMT
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de> <vIxz8.81546$vF6.2424489@news2.tin.it> <aamebj$c2olc$1@ID-59885.news.dfncis.de>
Message-ID: <1bzz8.82059$vF6.2440100@news2.tin.it>

Thomas Heller wrote:

> "Alex Martelli" <aleax at aleax.it> wrote in message
> news:vIxz8.81546$vF6.2424489 at news2.tin.it...
>> Ralf Juengling wrote:
>>         ...
>> > BTW: It is not easy to add __iter__ to list's meta-type, is it?
>>
>> list's "metatype" (metaclass) is the built-in type named 'type'.
>> like for all other built-in types and new-style classes.  Adding
>> __iter__ to type would thus make almost 'everything' iterable.
>>
> Correction: adding __iter__ to type would make instances of
> type iterable (these things are normally called 'classes').

Classes (new-style) and built-in types, as I said.  Maybe I wasn't
as clear as needed to express once again the distinction between
levels -- instances of a metaclass are types (classes), so thanks
for pointing it out.

> Not too useful, it seems.

Particularly given that any class or type (or other callable) that
is callable without arguments can be used, in the same was as
Ralg, the OP, wanted to use an "iterable list", by passing it
in as the first argument to the two-argument form of iter.  The
second arg need only be any object that is sure to differ from
what calling the class/type/other callable can result in, to
ensure the resulting iterator is unbounded.


Alex



From h_schneider at marketmix.com  Thu Apr 11 05:41:04 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Thu, 11 Apr 2002 11:41:04 +0200
Subject: Include zlib.dll in McMillan Installer
Message-ID: <a93ld3$jb4$03$1@news.t-online.com>

Hi,

is there a possibility to include zlib.dll into the single file distribution
of the installer?
E.g. PIL uses zlib and complains about missing it on some systems.

Thanks!

Bye,
Harald Schneider





From robin at jessikat.fsnet.co.uk  Sat Apr 13 05:06:56 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 10:06:56 +0100
Subject: hex and long
References: <8N9+ELA0u5t8Ew2n@jessikat.fsnet.co.uk>
 <mailman.1018668146.15677.python-list@python.org>
Message-ID: <GLcQVLAwU$t8EwQ3@jessikat.fsnet.co.uk>

In article <mailman.1018668146.15677.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
>[Robin Becker]
>> In 2.2 I find the following peculiarity
>>
>> >>> 0xffffffff
>> 68719476735L
>
>That looks an error in transcription; I expect you dropped an "f" character
>in the input.

whoops to late at night you're right
 
>
>> >>> 0xffffffff
>> -1
>> >>>
>>
>> clearly the second is being treated as a 32 bit int and the first as a
>> long.
>
....
thanks for the 'explanation'. It just looks odd now that we get a smooth
transition for add and multiply etc, but not for << and hex conversions. 

Is there an easy way to get back to the low 32 bits of a long or do I
have to do bit twiddling?
-simple failure is not better than complex success-ly yrs- 
Robin Becker


From spamtrash at softhome.net  Thu Apr  4 07:28:00 2002
From: spamtrash at softhome.net (Johannes Gijsbers)
Date: 4 Apr 2002 04:28:00 -0800
Subject: Catalog of Python Patterns? [LONG!]
References: <a8g97m$10jq$1@news.idiom.com>
Message-ID: <e87ce949.0204040427.7a4df8cd@posting.google.com>

jimd at vega.starshine.org (Jim Dennis) wrote in message news:<a8g97m$10jq$1 at news.idiom.com>...
> GoF Design Patterns in Python
> 
>  [Warning LONG POST AHEAD!]
> 
>  I'd like to see a canonical catalog of OODP in Python.

Try taking a look at the upcoming book by Bruce Eckel, Thinking in
Python[1]. Although it isn't finished yet, a lot of useful information
can already be found in there, including a nice implementation(no
metaclasses) of the Factory pattern that Paul Rubin was referring to.

[1] http://www.mindview.net/Books/TIPython


From stephen at xemacs.org  Wed Apr 17 07:31:36 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 17 Apr 2002 20:31:36 +0900
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <87vgaswc1s.fsf@tleepslib.sk.tsukuba.ac.jp> <7934d084.0204161757.72ff525f@posting.google.com>
Message-ID: <87u1qatvc7.fsf@tleepslib.sk.tsukuba.ac.jp>

To begin with, let me apologize to Dr. Mertz for the unwarranted ad
hominem smear I committed.  That was evil.

>>>>> "Andrae" == Andrae Muys <amuys at shortech.com.au> writes:

    Andrae> However to attempt to suggest that the terms themselves
    Andrae> have not been well defined is dishonest.

Who said "free software" or "open source software" is ambiguous?
(Admittedly, the definition of "free software" _is_ inherently
incomplete.  The FSF explicitly says that the criteria will be revised
to rationalize their judgements about individual licenses ex post.)

Those discussing these things have long since published definitions,
agreed to accept each others' definitions, _and they have also long
since agreed that for the purposes of most discussions comparing the
broad idea of open source with that of free software, "free software
== open source software"_.  Huaiyu clearly found Tim Peters's
contribution in that vein useful.  So his _question_, and the
appropriate answer, were ambiguous.

I personally prefer the kind of answer based on abstracting from the
miniscule differences, but recognize the answer that David Mertz
prefers as a valid, but equally partial, alternative.  Dr. Mertz's
refusal to recognize this ambiguity, however, is an error.

    Andrae> To attempt to cast someone who objects to this as some
    Andrae> raving idealogue,

You're right, I was wrong to do this.

    Andrae> and thereby dismiss them with casual wave handling as a
    Andrae> self declared 'seeker of truth' (tm) is the most appalling
    Andrae> piece of intellectual hypocracy I have seen for some time.

YBT.

Actually, I have to laugh at myself, because I was in fact wrong to
assert that the sets of licenses/covered programs are the same, and
Huaiyu led me to truth.  Believe me, I didn't plan that, but there is
proof: I was seeking (despite being fairly sure I already had the
answer) and I found.<wink>

Satori-remains-out-of-reach-ly y'rs

Steve

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From bbollenbach at shaw.ca  Thu Apr 11 18:19:58 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Thu, 11 Apr 2002 22:19:58 GMT
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com>
Message-ID: <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca>

In article <12bb92e6.0204091223.738e0798 at posting.google.com>, Peter Olsen wrote:
> I think that, for my purposes, licensing is not worth my effort or
> my users' time.  But if you want to license your Open Source software,
> be my guest.

This doesn't actually make sense. To call software "Open Source" is to
acknowledge it as being distributed under a license that is defined as
compatible with what Richard Stallman calls "Free Software". So you
don't have the "choice" of licensing Open Source software. It got to be
called "Open Source" *because* of the license you've already chosen for
it.

And, FWIW, thinking that you're doing your users a favour by monitoring
who's using which version of what, and in what kind of environment
(commercial vs. non-commercial) is, at best, dependant on the types of
users, and, at worst, and absolutely disasterous idea. If they're 
"end-user" types, you might legimitately be providing some capabilities 
they want, but adopting this scheme for Open Source Hacker types will 
just reduce your user-base by an order of magnitude or more.

My $0.02,

Brad


From gerhard at bigfoot.de  Mon Apr  8 03:43:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 8 Apr 2002 09:43:34 +0200
Subject: IsList?
In-Reply-To: <slrnab2feb.atp.mlh@vier.idi.ntnu.no>
References: <9F9s8.37463$tg4.444710@vixen.cso.uiuc.edu> <slrnab2b3b.1gd.gerhard@lilith.my-fqdn.de> <slrnab2feb.atp.mlh@vier.idi.ntnu.no>
Message-ID: <20020408074334.GA1844@lilith.my-fqdn.de>

* Magnus Lie Hetland <mlh at vier.idi.ntnu.no> [2002-04-08 06:50 +0000]:
> In article <slrnab2b3b.1gd.gerhard at lilith.my-fqdn.de>, Gerhard
> =?iso-8859-15?Q?H=E4ring?= wrote:
> >John wrote in comp.lang.python:
> >> Hi, How can one know wheter var 'a' is list or not? Is there any
> >> IsList-like function?
> [snip lots of stuff]
> 
> If you want to check for sequences in general you can use the
> isSequenceType function:
> 
> >>> from operator import isSequenceType
> >>> isSequenceType([1, 2, 3])
> 1
> >>> isSequenceType((1, 2, 3))
> 1
> >>> isSequenceType("1, 2, 3")
> 1
> >>> isSequenceType(123)
> 0

I know about these, but I've read Guido wants them depreciated:

http://groups.google.com/groups?selm=199808242112.RAA20599%40eric.CNRI.Reston.Va.US

I don't know if he has changed his mind in the meantime, in that
isSequenceType could be implemented reliably.

IMO it gets even hairier with isMappingType.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From pinard at iro.umontreal.ca  Mon Apr 29 18:46:09 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 18:46:09 -0400
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCD9649.98AED86F@cascade-sys.com>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
Message-ID: <oqpu0inmxa.fsf@titan.progiciels-bpi.ca>

[James J. Besemer]

> Brian Quinlan wrote:

> > Why the hostility?

> I don't interpret this strongly worded and slightly sarcastic critique to
> be hostile.

For me, sarcasm and strong wording are not welcome.  There are ways to
disagree, and even disagree much, while staying friendly.  Even a big
accumulation of ironic statements are not going to make the shadow of a
good argument.  I prefer intelligent discussions between nice people.

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




From rjones at ekit-inc.com  Tue Apr  2 21:05:05 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 12:05:05 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <3CAA60A5.F88697D5@engcorp.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204030123.BAA15136@crown.off.ekorp.com> <3CAA60A5.F88697D5@engcorp.com>
Message-ID: <200204030205.CAA21272@crown.off.ekorp.com>

On Wed, 3 Apr 2002 11:53, Peter Hansen wrote:
> I would suggest an even shorter version of the quickref, focused on
> only those items needed in 80% of embedded documentation.  It does
> _look_ complicated at first.

I agree - I think the quickref definitely needs:

1. reorganising to filter the most common, simple markup to the top, and
2. much higher status/metioning on the ReST website.


> [snip discussion of more complex documentation and appropriate formats]
> But why would anyone want to embed such complex documentation?

As mentioned in another response, I'm actually using ReST to document Roundup 
outside of the code modules. I fully expect that one day (when platypus is up 
to it) to be able to produce PDF documentation from the ReST source. That'll 
be _neat_ :)

Of course, if you want to get really complicated, Frame/DocBook is the only 
way to go :)

I already write my slide presentations using ReST - it's converted to 
PythonPoint XML and from there to PDF. Very nice, and as simple as can be :)


     Richard



From paul at boddie.net  Sun Apr 28 19:16:31 2002
From: paul at boddie.net (Paul Boddie)
Date: 28 Apr 2002 16:16:31 -0700
Subject: zope and jboss or weblogic
References: <e786d63d.0204262051.14ebe470@posting.google.com> <mailman.1019890132.10241.python-list@python.org>
Message-ID: <23891c90.0204281516.4225c2fe@posting.google.com>

Gerhard H?ring <gerhard at bigfoot.de> wrote in message news:<mailman.1019890132.10241.python-list at python.org>...
> * Wenshan Du <python at dohao.org> [2002-04-26 21:51 -0700]: > hi,
> > I like zope, I want to know what's the difference between zope and
> > jboss or weblogic? which is better? Can python & zope replace J2EE?
> > 
> Not only do Zope and J2EE play in a different league, but it's also
> different kinds of sports.

Indeed. Zope seems to lend itself especially to "content management"
applications, whereas J2EE doesn't provide specific "application
readiness" - instead the elementary building blocks are provided,
although projects do extend upon the basic J2EE features to support
different kinds of applications.

> J2EE is really only a bunch of so-called enterprise-APIs thrown together
> and made sure the versions of the selected APIs work well together.

It would seem that the J2EE standards don't move particularly quickly,
but sometimes the most interesting features might not always be
available to you in a certain project because the application server
in use does not support them yet - container-managed persistence
"version 2.0" is a good example of this right now.

> These APIs include data persistence, distributed computing, others I
> don't remember, and last but not least, a web application framework
> (servlets) with a everybody-has-it-so-we-too code-in-HTML solution
> (JSP). OTOH, I don't know (yet), what ZOPE really is about.

JSP is a "me too" solution, in my opinion, but there are always
opportunities to choose different/better, but non-J2EE-standard,
technologies as well. In several respects, Zope acts like a J2EE
servlet container, but has tighter "object persistence" integration.

> Many of these APIs are overhyped and have a lot of potential to be used
> incorrectly, IMO especially EJBs.
> 
> If I were to compare J2EE with a Python solution, I'd first construct
> something equivalent in Python, i. e. a bunch of similar APIs thrown
> together, like "WebWare for Python" (servlets, code-in-HTML, MiddleKit
> for object-relational mapping), a distributed computing framework like
> PYRO or omniORBpy, etc.

Webware for Python acts rather like a servlet container as well, but
unlike Zope and J2EE, the motivation behind the Webware project seems
to be to keep components as close to fundamental language techniques
and practices as possible - in contrast, learning Zope and J2EE can
often seem like learning completely new technologies.

Paul

P.S. I started a comparison between J2EE and Python Web frameworks,
but it isn't available on the Web just yet.


From aahz at pythoncraft.com  Sun Apr 21 21:21:17 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 21:21:17 -0400
Subject: Mailman vs References: (was Re: How to make regexes faster? (
	Python v. OmniMark))
References: <mailman.1019430519.32535.python-list@python.org>
Message-ID: <a9void$9k6$1@panix3.panix.com>

In article <mailman.1019430519.32535.python-list at python.org>,
Delaney, Timothy <tdelaney at avaya.com> wrote:
> From: aahz at pythoncraft.com [mailto:aahz at pythoncraft.com]
>> In article <mailman.1019363618.25367.python-list at python.org>,
>> Tim Peters  <tim.one at comcast.net> wrote:
>>>[Erno Kuusela]
>>>>
>>>> maybe mailman could be hacked to fake a "references" header based
>>>> on "in-reply-to".
>>>
>>>Barry won't do that. 
>> 
>> You sure?  Last time I talked to him about it a couple of years ago, he
>> said he'd at least consider it if I gave him a patch.
>
>Make sure it also goes the other way ... "In-reply-to:" is created from a
>"References:" header.

Nope.  Many MUAs understand References: and I am *not* going to take any
chances on mucking with headers that don't absolutely need it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From rhhyatt at charter.net  Tue Apr 30 20:46:10 2002
From: rhhyatt at charter.net (Bob Hyatt)
Date: Tue, 30 Apr 2002 17:46:10 -0700
Subject: Call for kjBuckets.c [was Re: relational database?]
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <mailman.1020024596.32196. <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com> <mailman.1020138953.11638.python-list@python.org>
Message-ID: <ucuee4b9gmohd6@corp.supernews.com>

Richard,

I have, temporarily, stopped any work on Linux.  The thing I want to do has
to be done on Windows XP, therefor I am looking for a 'single' solution. In
this case, Python 'thru and thru'.

'Speed' would be nice, but I will start with 'accuracy' first. And I think
it is neat that a real database is done in a high level language. It proves
the power of Python, as far as I am concerned.

e.g., when you have little time to develop something, you get it working,
then optimize as necessary.

Thanks for your response to my query.

Regards, Bob Hyatt


"Richard Jones" <rjones at ekit-inc.com> wrote in message
news:mailman.1020138953.11638.python-list at python.org...
> On Mon, 29 Apr 2002 23:15, Steve Holden wrote:
> > "Oleg Broytmann" <phd at phd.pp.ru> wrote in message
> > news:mailman.1020024596.32196.python-list at python.org...
> >
> > > On Sun, Apr 28, 2002 at 08:11:36AM -0700, Bob Hyatt wrote:
> > > > I have had a fruitless search for Gadfly. As you noted, the website
for
> > > > Chordate is not responding.
> > >
> > >    I am keeping (temporary) a copy of kjbuckets at
> > > http://phd.pp.ru/Software/Python/misc/ and mirrors:
> > > http://phd.by.ru/Software/Python/misc/
> > > http://phd2.chat.ru/Software/Python/misc/
> > >    I am looking for a permanent maintainer.
> > >
> > >    Gadfly now lives in the Zope sources at www.zope.org.
> >
> > Oleg, hi! I also remember seeing somewhere patches that moved gadfly to
re
> > usage? gadfly.sf.net is a placeholder, it seems to me.
>
> I'm working on repackaging gadfly right now and will place the  results up
on
> gadfly.sf.net. I've started with the Zope version, so the re conversion is
> done. I've also integrated Oleg's kjbuckets efforts. I'm almost there...
just
> got a few more wrinkles  to sort out before I make the first release.
>
>
>    Richard
>
>
>




From akuchlin at ute.mems-exchange.org  Thu Apr  4 11:15:41 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 04 Apr 2002 16:15:41 GMT
Subject: CGIHTTPServer and custom headers
References: <mailman.1017932831.28556.python-list@python.org>
Message-ID: <slrnaaov1b.v6s.akuchlin@ute.mems-exchange.org>

In article <mailman.1017932831.28556.python-list at python.org>, Denis S. Otkidach wrote:
> 
> Have you tried Medusa
> (http://www.amk.ca/files/python/medusa-0.5.1.tar.gz)?

Note that I've already made a 0.5.2 release with another few bugfixes.

--amk


From philh at comuno.freeserve.co.uk  Thu Apr  4 07:59:45 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Thu, 4 Apr 2002 13:59:45 +0100
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <slrnaaoji1.pto.philh@comuno.freeserve.co.uk>

On Wed, 03 Apr 2002 17:32:52 -0500, Guido van Rossum <guido at python.org> wrote:
>
>    True and False will properly round-trip through pickling and
>    marshalling; for example pickle.loads(pickle.dumps(True)) will
>    return True, and so will marshal.loads(marshal.dumps(True)).


Not if they are written and read on machines running different 
versions of Python, they won't.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From siegfried.gonzi at kfunigraz.ac.at  Thu Apr 11 05:25:06 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 11 Apr 2002 11:25:06 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at> <a92533$4bs$1@peabody.colorado.edu> <3CB4D5DF.4912EFE4@engcorp.com>
Message-ID: <3CB55672.ED95892B@kfunigraz.ac.at>

Peter Hansen wrote:

> It's not like we contribute much to world peace or solving world
> hunger or anything.

A scientist too does not much to save the world. In front of me is
sitting a scientist which is really talented (a lot of publications in
peer reviewed journals), but if you ask him what he thinks about
science, he will tell you: "It is a great waste of time and money". In
German language we would call this dispute "Wissenschaftsgl?ubigkeit" or
not.

I would say programming is like an engineering discipline and therefore
as much important as science. But often I get the feeling that people
believe that programming alone does solve problems (that is the reason
why I often say that I am not interested in in programming, but the
reality just is: I am programming all the time).


S. Gonzi


From niemeyer at conectiva.com  Mon Apr  8 15:42:39 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Mon, 8 Apr 2002 16:42:39 -0300
Subject: problem zipping a text file with zlib
In-Reply-To: <ub3gp3lfn2rc3d@corp.supernews.com>
References: <ub3gp3lfn2rc3d@corp.supernews.com>
Message-ID: <20020408164239.A27694@ibook.distro.conectiva>

> I am using a python script and the zlib module to zip two text files into
> two separate zip files.  They are zipping fine.  However, when I use winzip
[...]

Have you tried using the zipfile module?

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From aleax at aleax.it  Tue Apr 30 04:20:39 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 08:20:39 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Xns91FF62618790Flumberjack@216.148.53.85> <b03e80d.0204291719.1f36c42b@posting.google.com> <mailman.1020130493.781.python-list@python.org>
Message-ID: <rtsz8.33227$8D3.939478@news1.tin.it>

Cliff Wells wrote:
        ...
> Especially since lumberjacks are supposed to be "okay".

Yeah, but they sleep all night and work all day, so they have no time to 
post to Usenet (even ignoring their endearing habit of putting on women's 
clothing and hanging around in bars, which must eat into their free time).


Alex



From aahz at pythoncraft.com  Fri Apr 12 09:51:31 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Apr 2002 09:51:31 -0400
Subject: Looking for doh!.. simple sequence type
References: <3CB6BED8.2000502@mxm.dk>
Message-ID: <a96op3$kla$1@panix1.panix.com>

In article <3CB6BED8.2000502 at mxm.dk>, Max M  <maxm at mxm.dk> wrote:
>
>Hmm ... I am trying to create a simple sequence type that I cant iterate 
>over like (Python 2.1)::

This is a *REALLY* good reason to upgrade to Python 2.2 if possible.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From gashton at cmedltd.com  Thu Apr 18 05:14:03 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 18 Apr 2002 10:14:03 +0100
Subject: UTC equivalent of time.mktime() ?
In-Reply-To: <a9kcnu$11v6$1@nntp6.u.washington.edu>
References: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com>
	<mailman.1019062773.24881.python-list@python.org> 
	<a9kcnu$11v6$1@nntp6.u.washington.edu>
Message-ID: <1019121243.554.14.camel@tp6>

On Wed, 2002-04-17 at 18:51, Donn Cave wrote:
> I have not been following, so pardon me if this has already been
> suggested, but on a platform where you have 2.0 or later and strptime,
> this might do that:
> 
>   lt = time.mktime(time.strptime(sys.argv[1], '%d %b %Y'))
>   utoffset = time.mktime(time.gmtime(lt)) - lt
>   print lt - utoffset

Thanks for that.

I was using strptime() and mktime() to get seconds since the epoch in
localtime. I was trying to work out if I could do anything simpler than
this (in which two of the "-" signs should be "+" signs):

>>> def local2utc(secs):
...     if time.daylight:
...         return secs - time.timezone - time.altzone
...     else:
...         return secs - time.timezone

to get the seconds since the epoch into UTC. It's looking as though the
answer (in general) is "no".

-- 
Graham Ashton





From nhodgson at bigpond.net.au  Fri Apr 12 21:49:42 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 13 Apr 2002 01:49:42 GMT
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de> <m31ydlndtd.fsf@chvatal.cbbrowne.com> <a980jr$9d07k$1@hades.rz.uni-sb.de>
Message-ID: <W8Mt8.524$o66.2527@news-server.bigpond.net.au>

Uwe Schmitt:

> oops. i forgot to tell you that I'm using windows...
> and i really want to print "raw" data and not pdf et al...

   PythonWin handles printing of source code in its CScintillaView.OnPrint
method in the Pythonwin\pywin\scintilla\view.py. This code relies on being
embedded in the MFC framework used by PythonWin. You may be able to use the
Windows API more directly through the win32all extensions including the
win32print and win32ui modules.

   Or did you want to send commands directly to the printer?

   Neil





From paul at boddie.net  Wed Apr  3 06:21:33 2002
From: paul at boddie.net (Paul Boddie)
Date: 3 Apr 2002 03:21:33 -0800
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <newscache$hzbztg$btb$1@www-neu.dzsh.de>
Message-ID: <23891c90.0204030321.e19d6df@posting.google.com>

"Ruediger Maehl" <ruediger.maehl_nospam at web.de> wrote in message news:<newscache$hzbztg$btb$1 at www-neu.dzsh.de>...
> "Paul Boddie" <paul at boddie.net> schrieb im Newsbeitrag
> news:23891c90.0204020644.3ea4bc40 at posting.google.com...
> > I'm not sure I would second that recommendation unless SAP DB has
> > changed dramatically over the past few weeks. I found that the
> > installation was quite a bit more complicated than, for example, MySQL
> > (at least on Windows) and upon issuing the arcane commands for
> 
> Yes, you need to get used to the commands. But by now there is a
> database manager (client/server, web) available, which supports
> you in daily routine.

I just found it much more "involved" to get started with in comparison
to MySQL (on Windows) or PostgreSQL, Sybase ASA (on UNIX).

> > restarting database instances, the software would complain that the
> > instances needed restarting, which was hardly very reassuring with
> > respect to either the stability or the usability of the software.
> 
> Sorry, but I don't understand this statement.

Well, there's a combination of commands that one apparently needs to
execute to restart the database server, and this appears necessary
when recreating databases. Unfortunately, the "how to" documentation
isn't particularly clear about this, beyond a terse summary of how it
might be done:

  http://www.sapdb.org/sap_db_howto.htm

Then, there are the obvious questions? What does <DOMAIN_password>
mean when one hasn't been specified in the instructions? It certainly
isn't a Windows domain password, for example, but one is left
wondering.

> > As for the Python DB-API module, I couldn't get any version of it I
> > found to work with the database system. Admittedly, my objective was
> > to get a JDBC connection up and running, which was eventually
> > successful, and I therefore cut that particular side-investigation
> > short.
> 
> I had no problem to use the native interface.

Well, perhaps I was too ambitious in expecting it to work with a
binary installation of Python on Windows. Had I been evaluating it on
UNIX, I would have been more inclined to investigate the exact
problem.

> Compared to SAPDB, MySQL and PostgreSQL are well developed
> over a long time and there are more tools available. IIRC, MySQL has had
> a 2GB limit for the size of a table and was not free and PostgreSQL
> needed some Cygwin thing to run (on W*ndows). That were the reasons
> for me to use SAPDB and it is in production for 14 months now without
> any problems. Anyway, YMMV.

Indeed. Anyway, my main point was that even for people who have
installed and evaluated a fair number of database systems, such as
myself, the documentation for SAP DB is more "opaque" than for many
other systems. For those people who don't tend to install lots of
packages, I would have a hard time recommending SAP DB for that very
reason.

Paul


From eviljewel at evilemail.com  Thu Apr 25 11:05:37 2002
From: eviljewel at evilemail.com (Eviljeweller)
Date: 25 Apr 2002 08:05:37 -0700
Subject: bitmap images
Message-ID: <287a1ca.0204250705.54df316f@posting.google.com>

If anyone can tell me how to add bitmap images to a python program in
any way shape or form i would be very grateful.

thanx


From aleax at aleax.it  Wed Apr 17 17:22:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 21:22:14 GMT
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <mailman.1019073733.27071.python-list@python.org> <Filv8.32338$vF6.998647@news2.tin.it> <a9kocn$6mu$1@news.mathworks.com>
Message-ID: <aIlv8.32491$vF6.1001531@news2.tin.it>

Joshua Marshall wrote:
        ...
>> If it were, there would be little reason for it to exist -- saving
>> a couple of keystrokes would be a silly reason to introduce a
>> bunch of new operators.
> 
> To me there's a more compelling reason for the existence of "+=": you
> don't need to use a temporary variable to prevent the lvalue from being
> computed twice. Eg:

But the "temporary variable" IS a couple of keystrokes.  Since it's
only a reference that's copied to the temp, cost difference is truly
minute anyway.

> Granted, since in Python assignment is a statement and not an expression,
> this is less of a win than the existence of += in C, where it might be
> inconvenient to introduce a temporary.

It's such a tiny 'win' in Python, that it would not justify += and friends.
_Just a couple of keystrokes_...!  Almost doubling the number of operators
in a language for THAT?!  No, it's only the NON-equivalence of a+=b to
a = a + b that made introducing += &c really sensible.


Alex



From starx at pacbell.net  Sat Apr  6 04:53:37 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 6 Apr 2002 09:53:37 GMT
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com> <3cadd841.225619243@news.laplaza.org>
Message-ID: <Xns91E81370BB5FERASXnewsDFE1@130.133.1.4>

Mats Wichmann || Fri 05 Apr 2002 09:56:05a:

> On 5 Apr 2002 02:05:29 -0800, alanmk at hotmail.com (Alan Kennedy) wrote:
> 
>:brian d foy wrote:
>:> for our April 2002 issue (April Fool!), The Perl Review replaced
>:> its Perl content with Python articles, so Python folks might want
>:> to read this issue:
>:> 
>:>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf
>:
>:Forgot to mention: Does this mean that the Golf competition is open to
>:Python entries :-)
>:
>:The golf competition is a neat idea. Python hasn't completely
>:suppressed the hacker in me :-)
> 
> 
> Ugh, it's horrid, and seems very perl-ish.  Do something in the
> smallest amount of space, readability/maintainability be damned.
> 
> We need something that emphasizes the beauty, simplicity, clarity of
> the solution (some Japanese term, perhaps?)

Well, i'm not someone to ask about the detailed meanings of japanese 
words, but i'm bored, so i did some dictionary diving on these terms. 
Here's a couple hits (kanji meanings in parens).

keishikibi (form/style - system/expression - beauty) 'beauty of form'
kotan (wither/die - faint/pale O_O) 'elegant simplicity'
tanryoubi (simple - clear - beautiful) 'i just made this word up'
-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From sholden at holdenweb.com  Wed Apr 17 08:43:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 08:43:39 -0400
Subject: client timeout!
References: <mailman.1019042568.4819.python-list@python.org>
Message-ID: <Z7ev8.12438$hl2.4867@atlpnn01.usenetserver.com>

"Stian Jenssen" <stian.jenssen at nrk.no> wrote ...
> Hi all!
>
> Have a little problem.
> Have a script that runs on a ms server.
> It parses other machines for information, but
> when it don't get an answer from the other end it halts.
>
> I want to bypass this problem with a shorter timout, in the .lib file.
>
> Have anybody had a similar problem.
> What file do i have to modify (urllib.py, httplib.py or other) and how?
>
The usual answer is to use the timeoutsocket module from Timothy O'Malley.
See

    http://www.timo-tasi.org/python/timeoutsocket.py

regards
 Steve






From jolsen at mailme.dk  Mon Apr 15 04:40:21 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 15 Apr 2002 01:40:21 -0700
Subject: used ports - socket bind
Message-ID: <cf0ad9fb.0204150040.2f70bc4@posting.google.com>

When binding a socket to a port, is there a way to test if the port
is already used?

(Apart from catching the resulting exception if it is...)

Cheers
Jesper


From j.spies at hccnet.nl  Mon Apr  8 18:35:18 2002
From: j.spies at hccnet.nl (Jaap Spies)
Date: Mon, 08 Apr 2002 23:35:18 +0100
Subject: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org>
Message-ID: <3CB21B26.8000201@hccnet.nl>

Tim Peters wrote:
> [James Logajan]
> 
>>...
>>Seriously Guido, how often will I need to check back to www.python.org
>>to keep up with the changes you are making?
> 
> 
> It's unclear why you would ever need to, if you want to stick with an older
> version, unless you wanted to pick up a pure bugfix release (a 3-digit
> release, like next week's upcoming 2.2.1 and 2.1.3, being, respectively, the
> first pure bugfix release in the 2.2 series, and the third in the 2.1
> series).
> 
> 
>>Do you have a schedule or some constraints in place on how fast you are
>>releasing major versions?

Most car companies release an new model every year.
James, I suppose you want to drive always the latest model.

Jaap



From jeffbi at jetsoft.com  Tue Apr  2 11:56:27 2002
From: jeffbi at jetsoft.com (Jeff Bienstadt)
Date: Tue, 2 Apr 2002 08:56:27 -0800
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net> <3ca8f000$1@news.nwlink.com> <JLcfwNF0zPq8EwJm@nym.alias.net>
Message-ID: <3ca9e2bd@news.nwlink.com>

"Mindbomb" <Mindbomb at localhost.net> wrote in message
news:JLcfwNF0zPq8EwJm at nym.alias.net...

 [ interesting stuff about anonymous posting snipped ]

The anonymous posting stuff sounds interesting.  I'm still a bit
concerned about the spam issue, but I don't like the idea of
dis-allowing something useful because of how it *might* be used.

If it's useful, let's do it :-)

>
>
> >As far as your Python knowledge goes, I wouldn't worry about it.
> >My Python experience is currently measured in weeks, although
> >I've been doing the coding thing (mostly C/C++) for more years
> >than I care to admit.  I'm finding it surprisingly easing to learn;
> >the hard part is giving up my hard-earned C++ habits :-)
>
> Sounds like me, except I never did get my head around OOP, so I stuck
> with C (which is one of the reasons I want to get into python - C++ is
> just too arcane for my liking, it's like reading OS/390 assembler 3
> years after you wrote the code). My python experience weighs in at hours
> at the moment, although I just installed the laptop with a fresh copy of
> MDK 8.2 so I'll have a little play this week at work.
>
> >My goals for this project (personally speaking) are, in this order:
>
> >   1) Learn Python
> >   2) build a newsreader that I'm happy with
> >   3) build a newsreader that others are happy with :-)
>
> Sounds like a go-er to me. Are you a Windows man, or do you prefer the
> delights of the Linux/Unix graphical toolkits?
>

One of my other goals is to have a newsreader that works (as much as
possible) the same on both Linux and on Windows, so I was thinking
that using Tkinter for the GUI would probably be the best way to accomplish
that.

I'm currently using Outlook Express for news, and while I can't say that I
actually like it, it has some "features" that I've gotten used to, and I
can't
seem to find a Linux-based newsreader that behaves the way I've grown
accustomed to --- and I want one :-)

---jkb





From gashton at cmedltd.com  Fri Apr 19 10:37:14 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 19 Apr 2002 15:37:14 +0100
Subject: Should exceptions have an __init__ method?
Message-ID: <1019227035.1583.5.camel@tp6>

I've got a quick query about defining exceptions. In Beazley's Essential
Reference book (2nd ed.), on page 59 it says that you should define your
exceptions with their own __init__ method that accepts an args argument,
if you want the values that you pass into them to be handled correctly
in "tracebacks and other diagnostics" (see class F below for an
example).

I'm unable to determine the difference between these two exception
classes though (where class E is essentially empty):

Python 2.2 (#2, Mar 11 2002, 13:24:00) 
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> class E(Exception):
...     pass
... 
>>> raise E, "foobar"
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
__main__.E: foobar
>>> class F(Exception):
...     def __init__(self, args=None):
...         self.args = args
... 
>>> raise F, "foobar"
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
__main__.F: foobar


As you can see, I'm on 2.2, and Beazley wrote his book for 2.1. Has 2.2
made subclasses of Exception more intelligent, or am I missing
something?

Cheers.

-- 
Graham Ashton





From ark at research.att.com  Wed Apr  3 23:16:16 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 4 Apr 2002 04:16:16 GMT
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com> <mailman.1017875711.27201.python-list@python.org>
Message-ID: <yu99lmc4drov.fsf@europa.research.att.com>

Neal> This code:
Neal> 	True = 1
Neal> 	False = 0

Neal> 	def test(a):
Neal> 	    if a is True: print 'True'
Neal> 	    if a == False: print 'False'

Neal> generates these warnings:
Neal> 	test.py:2: Should not assign to True, it is (or will be) a builtin
Neal> 	test.py:3: Should not assign to False, it is (or will be) a builtin
Neal> 	test.py:6: Comparisions with True are not necessary
Neal> 	test.py:7: Comparisions with False are not necessary

I'd like to suggest that comparisons with True are not only
unnecessary but dangerous.  Comparisons with False are much
less dangerous but still unnecessary.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From tejarex at yahoo.com  Mon Apr 22 13:00:46 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 22 Apr 2002 17:00:46 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <2lXw8.130011$3L2.11915313@bin7.nnrp.aus1.giganews.com>

"Gon?alo Rodrigues" <op73418 at mail.telepac.pt> wrote in message
news:dit7cuo1nguiha34nq80m9q39igpjlr70d at 4ax.com...
, 22 Apr 2002 06:10:02 GMT, Tim Daneliuk <tundra at tundraware.com>
...
> You should understand that this is a philosophical viewpoint. The
> mathematical field is roughly divided in the Platonist field, those
who
> believe that mathematical objects exist somewhere out there in space
(to
> quote Sun Ra) in a Wonderful World of Platonic Ideas,

This is such a crude and misleading caricature of the minority
Platonist viewpoint that I suspect you must incline toward

> the formalist
> field of those who believe mathematics is essentialy a game we play
on
> paper with symbols with no "real" meaning attached to it besides the
> convenience in describing natural phenomena.

Terry J. Reedy





From paul at boddie.net  Thu Apr 11 13:07:22 2002
From: paul at boddie.net (Paul Boddie)
Date: 11 Apr 2002 10:07:22 -0700
Subject: Indenting lines 400..457 help
References: <a929tk$ik2$1@pcls4.std.com> <Fc2t8.90644$w7.7532194@bin6.nnrp.aus1.giganews.com> <a92f90$i57$1@pcls4.std.com> <HNgt8.335984$Gf.29623810@bin2.nnrp.aus1.giganews.com>
Message-ID: <23891c90.0204110907.326eaf0e@posting.google.com>

"Brett g Porter" <BgPorter at NOartlogicSPAM.com> wrote in message news:<HNgt8.335984$Gf.29623810 at bin2.nnrp.aus1.giganews.com>...
> "Ned Harvey" <rahvee at yaahhoo.com> wrote in message
> news:a92f90$i57$1 at pcls4.std.com...
> > By using the >> or << method below, vi uses the tab-characters, at least
> > in my account.  Is there any way you're aware of, that I can specify the
> > characters to use?
> >

[expandtab]

I tend to use the following line at the end of my Python files:

  # vim: tabstop=4 expandtab shiftwidth=4

I think this makes virtually all tab instances soft and equivalent to
4 spaces in vim/gvim. Having been a "hard tab user" for years, I
finally switched to soft tabs out of compatibility with just about
everything else written in Python - oh, the shame!

;-)

Paul


From maxx at easynews.com  Tue Apr 16 12:37:18 2002
From: maxx at easynews.com (maxx at easynews.com)
Date: Tue, 16 Apr 2002 16:37:18 GMT
Subject: Why are so many shorts in a knot? Was: Do you QA your Python?
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <Xns91EFCC39E5237JamesLLugojcom@216.148.53.84>
Message-ID: <ldkobukmsrbu3uuvadsqooqil2cjsm8uh1@4ax.com>

There seems to be an enormous amount of F.U.D. lately about the regular new
versions of Python. I am curious to know what environment these upset
individuals work within that do not undergo regular and constant change?

In my career, I have NEVER had the opportunity to believe, "This tool is
perfect, and needs no further changes." Sometimes, new features are released,
and I can choose to use them or not. Sometimes, updates are released to support
new operating system versions. Sometimes, very useful and valuable tools (to
me), are simply abandoned by the developer. The only thing I have been able to
depend on, is that I will have to expend significant effort to keep pace with
the industry.

Why should Python be any different? Every commercially successful language,
database, or operating systems I can think of gets regular updates. Sometimes
they are good for our projects, sometimes not. Sometimes we can ignore the
upgrades, sometimes we have modify systems to meet the new requirements. The
interesting thing is that I am certain that some time ago, some discussions were
made about why Python was still in a 1.X version, since IT Managers do not trust
such a "young" language.

Of course I may have forgotten one law of human nature - No matter what you do
(or do not do), someone will feel the need to complain about it.



From ajs at ix.netcom.com  Mon Apr  1 19:00:26 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Mon, 1 Apr 2002 19:00:26 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <000f01c1d9d9$73e80280$0334fea9@carol>


>Its designers were very concerned about newbies 
>being confused by traditional computer science
>terms like "function", "procedure" and "variable", 
>so they invented all new terminology. Functions 
>ended up being called YIELDs, procedures were 
>HOW-TOs, statements became commands, and variables
>became locations.

>The main effect this had was to confuse the died-in-the-wool
>programmers; it didn't make any difference for the newbies because
>they had to learn the concepts anyway.

I'm confused again.

One of Alice's signficant "lessons" was that it was important
to use terms like "move" and "turn" rather than "translate" and
"rotate" - in the context of doing 3d.  I thought it was a ridiculous
conclusion for the reasons you outline (among others).  I never had any
reason to suspect before now that you shared my view. Quite, in
fact, the opposite.

Your article in the Linux journal was quite explicit in promoting
the use of a 'simplified' vocabulary in the context of the IDE
of the future - clearly rejecting the concept that adopting
settled vocabulary was a good thing.

And PEP228 which you now announce has no importance
for Python's future was specifically referred to as a basis 
for the adoption of the PEP238 - what a few months ago.

I will be accused of course of flame baiting again.

But I in fact just don't get it.

Though at least I am able to just not get it with significantly
less passion attached than in prior go rounds.

Art





From martin at v.loewis.de  Mon Apr  1 08:27:28 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 15:27:28 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <roy-502972.07485701042002@news1.panix.com>
Message-ID: <m3663b8s73.fsf@mira.informatik.hu-berlin.de>

Roy Smith <roy at panix.com> writes:

> Stefan Schwarzer <s.schwarzer at ndh.net> wrote:
> > Example: "(val1, val2)[0]" is pretty clear, but "(val1, val2)[False]"
> > looks quite strange.
> 
> Does it look any stranger than "False : val1 ? val2" ?

Indeed. I'd just spell this is "val11".

Regards,
Martin


From barry at zope.com  Thu Apr 18 00:35:06 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Thu, 18 Apr 2002 00:35:06 -0400
Subject: RELEASED Mailman 2.0.10
Message-ID: <15550.19706.501470.740935@anthem.wooz.org>

I've just released version 2.0.10 of Mailman, the GNU Mailing List
Manager.  Mailman is released under the GNU General Public License
(GPL).  Version 2.0.10 fixes a few minor, but mostly harmless, bugs in
version 2.0.9.  I recommend that you upgrade, but it is not critical.
Details of what's changed are included below.

GNU Mailman is software to help manage electronic mail discussion
lists.  Mailman gives each mailing list a unique web page and allows
users to subscribe, unsubscribe, and change their account options over
the web.  Even the list manager can administer his or her list
entirely via the web.  Mailman has most of the features that people
want in a mailing list management system, including built-in
archiving, mail-to-news gateways, spam filters, bounce detection,
digest delivery, and so on.

Mailman is compatible with most web servers, web browsers, and mail
servers.  It runs on any Unix-like operating system.  Mailman 2.0.10
requires Python 1.5.2 or newer.  To install Mailman from source, you
will need a C compiler.

For more information on Mailman, including links to file downloads,
please see the Mailman web page: http://www.gnu.org/software/mailman

And its mirrors at:

    http://mailman.sourceforge.net
    http://www.list.org

Patches and source tarballs are available at

    http://sourceforge.net/project/showfiles.php?group_id=103

There are email lists (managed by Mailman, of course!) for both
Mailman users and developers.  See the web sites above for details.

Cheers,
-Barry

2.0.10 (09-Apr-2002)

    - Closed another small race condition.

    - Add the RFC-2822 recommended Message-ID: header on internally
      generated outgoing messages.  Not all MTAs add this field if
      missing (read: Qmail).




From borcis at geneva-link.ch  Mon Apr  8 07:52:11 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 08 Apr 2002 13:52:11 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk> <3CAFAAA4.41EA7CF9@python.org> <+CcOX7A2gWs8EwPR@jessikat.fsnet.co.uk>
Message-ID: <3CB1846B.5050109@geneva-link.ch>


Robin Becker wrote:

> 
>>>
>>You just made my kill file.
>>
>>--Guido van Rossum (home page: http://www.python.org/~guido/)
>>
> wow, I thought I was already there since the division rumble.
> Interesting how the irreverent spirit of Monty Python seems singularly
> lacking these days.



I'd rather note the typical elegance and economy of the BDFL's designs;
did you notice how he wrote "made my kill file" and not "made it to my
kill file" ?

Boris
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From wurmy at earthlink.net  Tue Apr  9 20:39:00 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Wed, 10 Apr 2002 00:39:00 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no> <usn64ame1.fsf@python.net> <3CB31335.2040501@mxm.dk> <u8Fs8.462$b62.9923@news1.tin.it>
Message-ID: <3CB38A4D.8ADE6717@earthlink.net>

Alex Martelli wrote:

> Max M wrote:
> 
> > Michael Hudson wrote:
> >> Syver Enstad <syver-en+usenet at online.no> writes:
> >
> >>>Am I the only one in the entire world that thinks '\t'.join(mylist) is
> >>>elegant and natural?
> >>
> >> No.
> >
> > Well I would have found mylist.join('\t') more natural, but I can live
> > with the other.
>
[...]
> 
> joiner.join(sequence) was an EXCELLENT design choice...

I'm sure it was, and also useful and necessary. Unfortunately,
it's also seriously ugly, so I'm glad we can still use
string.join. :-)

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA==')) 
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/


From tim.one at comcast.net  Thu Apr 18 00:49:20 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 18 Apr 2002 00:49:20 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <FWvAiLA3XSv8Ewwi@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCGECOOPAA.tim.one@comcast.net>

[Tim]
> Getting rid of the "delete in-use files" gimmick does [eliminate
> the vanishing unwise.exe problem, under at least one scenario].  I
> don't know whether that's the *only* problem here, though (Robin
> Becker seems to think not, but I don't believe everything I read
> on Usenet <wink>).

[Robin Becker]
> well actually I think your hypothesis is correct.

Well, it was definitely *a* cause, and the sole cause of every vanishing act
I was able to provoke.  I reported it as a bug to Wise, and changed our
installer to stop using the "delete in-use files" gimmick for the duration.

Thanks for bringing it up, Robin!  While I've been seeing this happen for a
year, the lack of any other report caused me to file it in my "huh --
something mysterious happened on Win98 -- so what else is new" bin.

BTW, if UNWISE.EXE goes missing for someone, see whether a file named

    GLB1A2B.EXE

is sitting in your temp directory.  The Wise uninstaller creates this, and
except for the name it's identical to UNWISE.EXE.  That's how they manage to
delete UNWISE.EXE during a normal uninstall without a reboot (it can't
delete itself because Windows won't let anyone delete an open file; instead
UNWISE.EXE makes a renamed clone of itself in the temp directory, and spawns
the clone at the end of the uninstall process to get rid of itself).  You
can move GLB1A2B.EXE to your Python root, rename it to UNWISE.EXE, and all
the normal ways of uninstalling should work again.





From sholden at holdenweb.com  Mon Apr 15 13:49:03 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 13:49:03 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCMEPOONAA.tim.one@comcast.net> <mailman.1018763429.14619.python-list@python.org>
Message-ID: <pqEu8.14418$Ji7.9663@atlpnn01.usenetserver.com>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote ...
> Rather than confuse Tim further with abstract samples of where I want to
> define a "NULL" default argument, here's a particular example of where
> the "NULL pattern" I've described is used:
>
> NULL = []
>
Note that this does indeed cause the NULL variable to refer to a specific
instance of the empty list (and a tuple would be acceptable here rather than
a list).

> class BasicProperty (property):
> ...
> def __init__( self, name, documentation = "", bounds = (),
> defaultValue=NULL, defaultFunction=NULL ):
> ...
> if defaultValue is not NULL:
> self.defaultValue = defaultValue
> elif defaultFunction is not NULL:
> self.defaultFunction = defaultFunction
>
[sorry, OutOfLuck Express doesn't see your tabs]

The only situation this will not help you with is if you need to distinguish
between the following calls:

    BasicProperty("One")

    BasicProperty("One", defaultValue=NULL)
and
    BasicProperty("One", defaultValue=())

> Now, I want to store/define defaultValue only if defaultValue is
> explicitly specified (same for defaultFunction).
>
The situation where you do not provide an argument value is guiaranteed to
work as long as you don't rebind the NULL variable. I believe the first and
second cases are entirely equivalent under that assumption.

The third case is trickier, however, as there is no guarantee that the
implementation won't optimize, so it is permissible for different
implementations to use either the same or different lists (or tuples).

> I _could_ (as you suggest) use namedarguments and require that the user
> name the arguments explicitly.  However, when I do that, the
> defaultValue and defaultFunction arguments don't show up in the user's
> call-tip window using Python IDEs.  (And I dislike the aesthetics of
> using **namedarguments when all acceptable names are known anyway :) ).
>
I agree that it's clunkier, but that's the timbot for you: always suggesting
things that work, no matter how inconvenient this is for us mere mortals :-)

> As for the "designed purpose of default arguments" :) .  I see the value
> "no value/not specified" as a perfectly suitable default value.  Sure,
> it may offend your sensibilities, but when has that stopped me in the
> past ;) .
>
Apparently never, if the above is typical of your disregard for the bot's
feelings.

regards
 Steve






From huaiyu at gauss.almadan.ibm.com  Mon Apr 15 14:02:02 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Mon, 15 Apr 2002 18:02:02 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <slrnabm5cp.nkd.huaiyu@gauss.almadan.ibm.com>

Stephen J. Turnbull <stephen at xemacs.org> wrote:
>I have no issue with the GPL, because the GPL only affects people who
>chose to use GPLed code in their projects.  It's a wonderful behavior-
>modification tool; it often causes people to free software they would
>otherwise keep proprietary.  Most of my development effort is spent on
>XEmacs, which is GPL through and through.  The GPL is certainly does
>not affect my decisions about what existing projects to contribute to.

I'm glad to hear that.  It is very respectable.   I salute you.

>I do have issues with some GPL advocates, because their arguments
>amount to claiming that nobody should have property rights in his own
>code.  (rms says this explicitly, but very few are so courageous as
>he.)  And I have issues with many other GPL advocates, because _their_
>arguments amount to claiming they have property rights in other
>people's code.  But that's a different thread.

Do you realize that the way you advocate your point of view gives an
impression that might be similar to the impression left on you by those GPL
advocates whom you have issues with?

RMS has my respect in being precise, consistent and honest.  He tries hard
to make clear the specific meanings he assigns to his words, and takes pains
to point out its difference with other usages or misunderstandings.

I don't think it helpful at all to sacrifice the precision of a term that is
central to an argument in favor of closeness to its everyday usage.
Otherwise we probably should all be arguing whether physicists used the
words "energy", "power" and "work" correctly.  :-)

>    Huaiyu> How come they are not equal if they are also the same?
>
>I can only suggest that you read what the protagonists say on their
>own web sites.  

I still don't know if _you_ are using them as synonyms for the same concept.
That does not help me to understand what _you_ are trying to say.  But
thanks for the links.  They certainly provide more perspectives.

Huaiyu


From distrex at hotmail.com  Sat Apr  6 13:49:25 2002
From: distrex at hotmail.com (distrex)
Date: 6 Apr 2002 10:49:25 -0800
Subject: modules
Message-ID: <d9ea2696.0204061049.551bd53b@posting.google.com>

I understand most of python. But I still don't get modules. I know
they are just basicaly a function but i don't know how to use them
effectively and what each one does. Everything good in python you will
at some point need to use a module right. If anyone can help me with
this i'd be greatful. Info or websites would be nice.


From tim.one at comcast.net  Thu Apr 11 21:18:37 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 11 Apr 2002 21:18:37 -0400
Subject: frameobject.c bug ? No call to PyFrame_LocalsToFast.
In-Reply-To: <pan.2002.04.11.12.44.16.889815.14362@cantal.lip6.fr>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEGHONAA.tim.one@comcast.net>

[Boris Boutillier]
> I've been looking through the frameObject.c code and looked for what
> happened, and surprised, there is no call to PyFrame_LocalsToFast when a
> frame is created, so if f_locals isn't empty (which can be the case if
> CO_NEWLOCALS is set to 0 the fast access values aren't initialised ).
> Is there a reason for this ?

CO_NEWLOCALS is always 1, except for code at module scope, in which latter
case locals and globals are the same thing and the LOAD_NAME (etc) opcodes
are used instead of LOAD_FAST (etc).  IOW, it's skipping an expense that's
never needed in code generated by the Python compiler.





From chris at cmb-enterprises.com  Sun Apr  7 14:31:28 2002
From: chris at cmb-enterprises.com (Chris)
Date: Sun, 07 Apr 2002 14:31:28 -0400
Subject: [Q] Python CGI Q
References: <8Far8.36437$tg4.432342@vixen.cso.uiuc.edu> <o3kvauc4u8i6cp803df18s5924qq0s7mlm@4ax.com>
Message-ID: <chris-C9DD6D.14312707042002@corp.supernews.com>

In article <o3kvauc4u8i6cp803df18s5924qq0s7mlm at 4ax.com>,
 Tim Roberts <timr at probo.com> wrote:

> "John" <nobody at nobody.com> wrote:
> >
> >Hi, I'm wondering if I can write a CGI that loads/edits/saves files on
> >server side.. (I don't need to worry about security concerns for this
> >application) Any starting point or open-source?
> 
> EVERYTHING in a CGI script happens on the server side.  You can create a
> <TEXTAREA> and copy the contents of your file inside it, then have a second
> script wait for the response and write the new contents out.

I actually prefer to have one script cover both of these things.  The 
receiving script should handle errors, and redisplay the form with 
appropriate error information, anyway, so giving it the ability to 
display a blank form if nothing has been submitted is trivial.

form(HTML) -> processor(CGI)

It also saves bandwidth, and thus production costs by not requiring a 
page to link back to the form if the CGI is called without any user 
input.

But Im getting off-topic...


From jeff at ccvcorp.com  Fri Apr 19 13:37:10 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 10:37:10 -0700
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com> <mailman.1019068175.11920.python-list@python.org> <99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com> <a9pjl6$pj6$1@slb4.atl.mindspring.net>
Message-ID: <MPG.1729f5a35f9bed979896b5@news.nwlink.com>

In article <a9pjl6$pj6$1 at slb4.atl.mindspring.net>, 
dalke at dalkescientific.com says...
> Dale Strickland-Clark:
> >But several orders of magnitude more processing [to use a dictionary
> >compared to an and/or solution]. Certainly not for tight, efficient loops.
> 
> Perhaps your intuition comes from C?  That doesn't seem to be the
> case for Python.
> 
[...] 
> Dictionary looks faster to me, in this small, tight loop.

And would be even faster if the dictionary creation were hoisted 
out of the loop, since it's invariant anyhow.

This definitely looks better to me.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From magnus.heino at pleon.sigma.se  Wed Apr 17 09:07:17 2002
From: magnus.heino at pleon.sigma.se (Magnus Heino)
Date: Wed, 17 Apr 2002 15:07:17 +0200
Subject: minidom and =?ISO-8859-15?Q?=E5=E4=F6?= once again :-P
References: <WXdv8.2962$ZN4.5053@news2.bredband.com>
Message-ID: <9sev8.3366$ZN4.5195@news2.bredband.com>

If i replace ??? below with codecs.utf_8_encode('???')[0], then parseString 
is happy.

toxml() doesnt produce a u'' string though, so I have to do 
unicode(doc.toxml()) and then feed that into parseString.

Eh??

/Magnus

Magnus Heino wrote:

> 
> 
> Hi.
> 
> I am (still) trying to get minidom and ??? to play together. And I guess
> the main problem really is me ;-)
> 
> But why doesnt this work? Shouldnt parseString eat what toxml produce?
> 
> /Magnus
> 
> [magnus at bombardier magnus]$ python2.2
> Python 2.2 (#1, Jan 23 2002, 14:50:45)
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import xml.dom.minidom
>>>> d=xml.dom.minidom.parseString('<?xml version="1.0"
> encoding="iso-8859-15" ?><test>???</test>')
>>>> d.toxml()
> '<?xml version="1.0" ?>\n<test>\xe5\xe4\xf6</test>'
>>>> d=xml.dom.minidom.parseString(d.toxml())
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line
> 965, in parseString
>     return _doparse(pulldom.parseString, args, kwargs)
>   File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line
> 952, in _doparse
>     toktype, rootNode = events.getEvent()
>   File "/usr/lib/python2.2/site-packages/_xmlplus/dom/pulldom.py", line
> 256, in getEvent
>     self.parser.feed(buf)
>   File "/usr/lib/python2.2/site-packages/_xmlplus/sax/expatreader.py",
>   line
> 143, in feed
>     self._parser.Parse(data, isFinal)
> UnicodeError: UTF-8 decoding error: invalid data
>>>>



From just at xs4all.nl  Thu Apr 18 06:43:09 2002
From: just at xs4all.nl (Just van Rossum)
Date: Thu, 18 Apr 2002 12:43:09 +0200
Subject: ranlib problem building 2.2.1 on Mac OS X
References: <m28z7l7hc4.fsf@javamonkey.com>
Message-ID: <just-4CD581.12430918042002@news1.xs4all.nl>

In article <m28z7l7hc4.fsf at javamonkey.com>,
 Peter Seibel 
 <peter at localhost.i-did-not-set--mail-host-address--so-shoot-me> wrote:

> I'm trying to build Python 2.2.1 from the source tarball on Mac OS X.
> I ran ./configure and then make and the make dies with:
> 
> ranlib libpython2.2.a
> ranlib: warning: cannot infer operation mode from `libpython2.2.a'
> ranlib: you must specify a MODE
> Try `ranlib --help' for more information.
> make: *** [libpython2.2.a] Error 1
> 
> while trying to build this target:
> 
> # Build static library
> # avoid long command lines, same as LIBRARY_OBJS
> $(LIBRARY): $(LIBRARY_OBJS)
> 	-rm -f $@
> 	$(AR) cr $@ Modules/getbuildinfo.o
> 	$(AR) cr $@ $(PARSER_OBJS)
> 	$(AR) cr $@ $(OBJECT_OBJS)
> 	$(AR) cr $@ $(PYTHON_OBJS)
> 	$(AR) cr $@ $(MODULE_OBJS) $(SIGNAL_OBJS)
> 	$(AR) cr $@ $(MODOBJS)
> 	$(RANLIB) $@
> 
> RANLIB is defined as:
> 
> RANLIB=		ranlib
> 
> Any ideas?

I just tried it, 2.2.1, ./configure with no options, on 10.1.4 with the 
december dev tools: "Works for me."

(Maybe ask on the macpython mailing list, more mac-savvy people out 
there.)

Just


From bruthasj at yahoo.com  Thu Apr 25 20:03:07 2002
From: bruthasj at yahoo.com (Jeff Pitman)
Date: Fri, 26 Apr 2002 08:03:07 +0800
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org> <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com> <slrnacb5ko.4i6.akuchlin@ute.mems-exchange.org> <b0f083db.0204231422.785cb57c@posting.google.com> <7llx8.2088$0x3.1467@atlpnn01.usenetserver.com>
Message-ID: <aaa5fu$2u4@netnews.hinet.net>

Steve Holden wrote:

> "Jeremy Hylton" <jeremy at alum.mit.edu> wrote ...
>> "A.M. Kuchling" <akuchlin at ute.mems-exchange.org> wrote ...
>> > In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>,
>> > Steve Holden wrote:
>> > > Since urllib knows nothing of cookies, you will need to integrate
>> > > some
> sort
>> > > of a cookie jar into the library, with a new API  for the clients to
>> > > retrieve and store the cookies.

Or do it transparently.


>> > This is worthwhile, but I don't think it belongs in urllib.  It
>> > belongs in a module or package of its own that provides general
>> > Web-browser features such as cookies, remembering authentication
>> > usernames and passwords, and a cache.  This package could then be used
>> > for implementing HTML-scraping scripts, spiders, or a Web browser.

No, it belongs in urllib2 because it uses recursion to open websites that, 
for example, redirect from http://site/index.php to 
http://site/login_page.php to  https://site/login_page.php.  All I can say 
is "good luck!" trying to intervene outside of the package without 
rewriting AbstractHTTPHandler.

>> I'm not sure what the difference between an HTTP client, like urllib
>> or urllib2, and a Web-browser is.  Other than urllib's monolithic
>> design, why wouldn't you want these sorts of features in the module?

Exactly!

> Personally I imagined passing a dictionary as an optional cookie jar
> argument, keyed by (domain, path) tuples. The library code would update
> this as dictated by its interactions with web sources.

This is what I did:

http://sourceforge.net/tracker/index.php?func=detail&aid=548197&group_id=5470&atid=305470

And, right now, it is transparent to anyone using urllib2.  It uses a 
persistent Dict (within a script who imports urllib2) that keys on the 
hostname and stores a "Cookie" object.  This Cookie object is scraped every 
time Cookies are sent from the server to the client.  The Cookie object is 
then consulted and its headers sent on each request to the server.

Obviously this is v0.0.0.1, but I think it is a start of something useable.  
I'm trying to create a library that "HTML-scrapes" websites and while doing 
so I hit a brick wall with Cookies.  This library is going to be similar to 
related perl scripts, except much cleaner.

Sample screen-scrape:

    ua = HTMLAgent( "http://www.yahoo.com/" )
    ua.report()

    form = ua.getFormByIndex( 0 )
    form.fill( 'q', 'python' )
    ua.submit( form )

    ua.report()
    ua.clickByName( 'Python Language Website' )

    print "Now at", ua.location.geturl()

So far so good with the library I've written so far. Except it can be slow 
as it uses minidom to parse the HTML.  And, I'm a newbie at this stuff so, 
I don't know how to limit the parse to only <form></form> tags in the 
screen-scrape process.

I'll clean it up a little tonight and drop it somewhere if you want to look 
at it.


-jeff


From jon+usenet at unequivocal.co.uk  Sat Apr 20 11:27:10 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sat, 20 Apr 2002 15:27:10 -0000
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org> <slrnac2v9d.kv6.jon+usenet@snowy.squish.net> <mailman.1019315200.29890.python-list@python.org>
Message-ID: <slrnac326d.gba.jon+usenet@snowy.squish.net>

In article <mailman.1019315200.29890.python-list at python.org>, holger krekel wrote:
>> >> Need a raw string there, i.e.
>> >> 
>> >>   r"\{\{(.*?)\}\}"
>> > 
>> > yes, i am sometimes confused by 'raw strings' and 'strings'. 
>> > Could you help by saying why exactly it is needed here?
>> 
>> Hmm, well, actually you don't, since \{ is not a string literal escape
>> sequence and Python is weird and leaves in unknown escape sequences
>> unchanged.
> 
> Actually my intention was to prevent the re module from thinking
> i might mean 'repetition' with {...}. I thought that every special
> regexp-character like '.' or '*' needs to be escaped if you want
> it literally. I was surprised to see that you can omit the '\'
> in the above case

Um, I didn't omit the \. I expect you are right that you do need it to
escape it from the regexp parser. I was wrong that you need to use a
raw string to prevent the Python parser from thinking the \ means
something, because in the specific case of \{ the Python parser
doesn't do anything with it. It is still a good idea to use
raw strings always for regexps though, because then you're not relying
on things turning out right accidentally.

> But is it true that you 'never' use them somewhere else? 
> I would think that one would like to use 'r' whenever you don't 
> want the escape-sequence-string-interpretation which is not really
> restricted to regexp-strings. 

True, but personally I've never found any other use for them.


From gerhard at bigfoot.de  Wed Apr 24 03:53:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 24 Apr 2002 07:53:17 GMT
Subject: What webhosting
References: <mailman.1019073817.29018.python-list@python.org>
Message-ID: <slrnaccp14.s4.gerhard@developer2.opus-gmbh.net>

In article <mailman.1019073817.29018.python-list at python.org>, steve at seasoned-software.com wrote:
> It'll never happen!!!!!!!!!!!   If you send unsolicited mail, the user may
> get fed up!

But I want to spam to MAKE MONEY FAST!!!!!!11

Gerhard


From ruediger.maehl_nospam at web.de  Thu Apr 11 09:33:41 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Thu, 11 Apr 2002 15:33:41 +0200
Subject: How to access attributes of nodes in xml.minidom
Message-ID: <newscache$78oeug$5lo$1@www-neu.dzsh.de>

Hello Pythoneers,

although I have tried as much as possible, I couldn't avoid that I now
have to use XML formatted files for different purposes. I just started
using the xml.minidom and I took the example and started to change
it to my needs.

But, how can I access the attributes of a node?

The documentation in the Python Library Reference is fairly short
and I cannot get further. Can anyone please help or point me to a
more extensive documentation?

Many TIA,

R?diger

This is my script:

#################
import xml.dom.minidom

document = """\
<slideshow>
  <title attrib="new">Demo slideshow</title>
  <slide>
    <title>Slide title</title>
    <point>This is a demo</point>
    <point>Of a program for processing slides</point>
  </slide>
</slideshow>
"""

def main(dom):
    for node in dom.childNodes:
        print node.tagName
        # That's how I expected it to work, but it doesn't
        for attrib in node.Attributes:
            print attrib

dom = xml.dom.minidom.parseString(document)

main(dom)

dom.unlink()  # minidom only?
#################






From op73418 at mail.telepac.pt  Mon Apr 22 15:42:29 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 20:42:29 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <3CC45005.2E9C0F86@tundraware.com>
Message-ID: <b8p8cu8f7u9lkbmpcrksjfcqp67nkoq3om@4ax.com>

On Mon, 22 Apr 2002 18:10:03 GMT, Tim Daneliuk <tundra at tundraware.com>
wrote:

>"Gon?alo Rodrigues" wrote:
><SNIP>
>
>> 
>> You should understand that this is a philosophical viewpoint. The
>> mathematical field is roughly divided in the Platonist field, those who
>> believe that mathematical objects exist somewhere out there in space (to
>> quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
>> field of those who believe mathematics is essentialy a game we play on
>> paper with symbols with no "real" meaning attached to it besides the
>> convenience in describing natural phenomena.
>
>Yup, I'm aware of this - I didn't want to digress more into philosophy
>than absolutely necessary.  IIRC, the Enlightenment thinkers also viewed
>mathematics as a branch of natural science and they were (mostly) 
>Aristoteleans, right?
>
><SNIP>
>
>> You should be careful when you say this. There is *virtually no*
>> mathematical field that has not been applied.
>
>True enough, but that is not a sufficient condition to argue that mathematics
>is correspondent one-for-one to the natural world.  All I'm arguing here is that
>mathematics embraces a problem set *larger* than just the natural world and
>this cannot simply be a natural science.

Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
would-be applications to the natural sciences. I think it was P. Erdos
who remarked once that all natural science is conceivably finite - e.g.
there would be a time when there would be nothing more to learn about
the universe - but mathematics is literally infinite.

>
>
><SNIP>
>
>> This is plainly not true. String theory models live in 10 and 24
>> dimensions. You would not want to disparage the biggest fad in current
>> theoretical physics, would you?
>
>
>Well, OK, I'll say it more formally:  Let n denote the hightest dimension
>used by any of the natural sciences.  Mathematicians routinely operate
>in k dimensions where k >> n.   Acceptable?

In quantum mechanics the state spaces are Hilbert spaces of infinite
dimensions (wether analytically or algebraically speaking). Believe me,
theoretical physics (or more correctly, mathematical physics)
uses/invents the same arsenal as mathematics. There is really no clean
separation.

>
>> 
>> Even so let me reiterate, that almost everything that mathematicians
>> have concocted is currently used in theoretical physics. To give you
>> just an example, my own, I use homotopy and higher order category
>> theory.
>> 
>> Even more when mathematics advances by theoretical physics promptings. I
>> could give you numerous examples, but just remind you that quantum
>> mechanics contributed a LOT (the celebrated Von Neumann, for example) to
>> the growth of abstract mathematics (functional analysis, Hilbert spaces,
>> operator algebras, etc.). Some of the more active work in mathematics
>> today is also prompted by theoretical physics.
>> 
>> Somehow (I do not intend to make this more precise now) new mathematical
>> formalisms and the discover of new laws in the Universe go hand in hand.
>
>
>Again, my claim was not that mathematics cannot be/is not appliable, merely
>that it is bigger than just its applications.
>
>> >
>> >I'm not choosing sides here on the math vs. science debate - both are important
>> >artifacts of the human intellect.  I merely take umbrage with the notion that
>> >mathematics is innately wired into the universe somehow.
>> 
>> See above to understand that what you are saying is really not that
>> obvious.
>
>
>Well almost nothing about epistemology is obvious.  As I said, I am willing
>to be persuaded otherwise, but it sure looks to me like mathematics is the
>construct I claim, not innate to the natural world...

The problem is what is exactly "innate to the world". A Platonist would
argue that there is a universe out there where the mathematical objects
exist independently. In short, If the universe ended right now, these
objects would continue to exist.

All the best,
Gon?alo Rodrigues



From merman at freenet.de  Tue Apr  9 17:23:24 2002
From: merman at freenet.de (merman)
Date: Tue, 9 Apr 2002 23:23:24 +0200
Subject: Embedding Python with C++
References: <mailman.1018363820.30451.python-list@python.org>
Message-ID: <3cb359c3$0$362$9b622d9e@news.freenet.de>

"Ken Seehof" <kseehof at neuralintegrator.com> schrieb im Newsbeitrag
news:mailman.1018363820.30451.python-list at python.org...
> >
> > > merman wrote:
> > > > Hi Ken,
> > > >
> > > > can I use your wizard with bcc55 too?
> > > >
> > > > o-o
> > > >
> > > > Thomas
> >
> > Ken wrote:
> > > No, it's a plug-in for the Visual C++ IDE.
> > >
> > > However you should be able to compile the generated code
> > > created by PyExtWizard with bcc.
> > >
> > > Much of the code generation functionality is independent
> > > of Visual C++, so it would be a feasible project to make
> > > PyExtWizard a multi-platform tool.  I will eventually
> > > need to do this for my own needs.  The GUI for the tool
> > > will be in wxPython.
> > >
> > > - Ken
>
> merman wrote:
> > ... thanx Ken - but I have VS7.NET (Beta2).
> > There are too much failures if I compile - no chance.
>
> Haven't tried it, but I doubt that VS7.NET supports VC6 wizards.
> Both products are by Microsoft and are therefore incompatible :-)
> If the old wizards don't work, it's useless to try to compile
> PyExtWizard (see below).
>
> > and with swig:
> > Maybe is my Makefile not valid - yesterday I posted it to
comp.lang.python
> > (thread swig-problem from merman).
> > I have try it with swig and bcc55 under win2k.
>
> You won't get anywhere :-(  SWIG has nothing to do with anything here.
>
> > Last error-messages from bcc55 are:
> >
> > Error: Unresolved external '_Py_InitModule4' referenced from
> > C:\MYC\PY_EXT\TREE_
> > WRAP.OBJ
> > Error: Unresolved external '_PyModule_GetDict' referenced from
> > C:\MYC\PY_EXT\TRE
> > E_WRAP.OBJ
> >
> > ... and so on. Is there a linker-probem?
> > I have no solution.
> >
> > o-o
> >
> > regards
> >
> > Thomas
>
> I think you misunderstood me.  You won't be able to build
> PyExtWizard with anything other than MS Visual C++ (MSVC++ 6.0).
> It makes no sense to build it with anything else since it is a
> plug-in wizard for MSVC++.  PyExtWizard is not a stand-alone
> application.  As a plug-in, it enhances VC++ by allowing you to
> create a new VC++ projects that implement python extension modules.
>
> It makes no sense to use SWIG on PyExtWizard.  PyExtWizard is not
> an API and it does not contain any C++ code that anyone would have
> any reason to wrap.  The code generation functionality is implemented
> in python and uses the capabilities of the MSVC++ wizard system.
>
> When I said, "you should be able to compile the generated code
> created by PyExtWizard with bcc", I meant that if you were to use
> PyExtWizard with Visual C++ to generate a new python extension
> module to your specifications, you would probably be able to compile
> the resulting code with a different compiler (hopefully anywhere).
>
> In other words, PyExtWizard is not portable at all, but it should
> generate portable code (this hasn't been tested yet).  I guess you
> could call it a "cross wizard" (Harry Potter on a bad hair day :-)
>
> I alluded to my future intent to make PyExtWizard independent of
> Visual C++, and therefore portable.  This would involve the following
> steps:
>
> 1. Throw away -all- of the C++ code.  The C++ portion of PyExtWizard
> serves only to interact with the MSVC++ IDE.
>
> 2. Replace the MSVC++ wizard functionality with python code.
>
> 3. Reimplement the GUI in wxPython.
>
> - Ken Seehof

Hi,

I think I understand you Ken - I only choose two ways to solve my problem:

first way:
with swig/ bcc55/ Makefile (see copmiler-messages above)

second way:
I used your PyExtWizard + VS7.NET without success - you know that.
I have no VC6.

I thought you have experience with SWIG maybe.

Last year I used swig/ gcc/ Makefile under Suse-Linux for the same purpose
and it runs!
But not so under win2k today. And I don't know why, you know?

Thats the problem.

best regards

thomas





From brian at sweetapp.com  Tue Apr 16 14:40:26 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 16 Apr 2002 11:40:26 -0700
Subject: C-extension not threadable?
In-Reply-To: <3cbc783d.39583390@news.easynews.net>
Message-ID: <008d01c1e576$2d4039d0$445d4540@Dell2>

Marcus wrote:
> The C-extension does what it is supposed to do, the callback function
> works, but the C-function does not start as a parallel thread but is
> freezing my UI. Anything basic I miss here ?

To allow meaningful threading, C extensions must explicitly release the
global interpreter lock. What extension is it?

Cheers,
Brian





From sholden at holdenweb.com  Mon Apr 22 11:04:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 11:04:18 -0400
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net> <3cc3765f$0$15478$afc38c87@news.optusnet.com.au> <B8E8F587.21E4A%goodger@users.sourceforge.net> <3cc38091$0$15473$afc38c87@news.optusnet.com.au>
Message-ID: <sGVw8.75816$T%5.32984@atlpnn01.usenetserver.com>

"Neville Franks" <nospam-readonly at getsoft.com> wrote ...
> Thanks, one small example that works is worth a thousand
somethingoranothers
> ;-)
>
making-a-somethingoranother-a-milli-example-that-works-ly y'rs  - steve






From gmuller at worldonline.nl  Tue Apr 30 15:28:48 2002
From: gmuller at worldonline.nl (GerritM)
Date: Tue, 30 Apr 2002 21:28:48 +0200
Subject: remove read only flag
References: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>
Message-ID: <aamrae$a3i$1@reader1.tiscali.nl>

"Francis Meyvis" <francis.meyvis at sonycom.com> schreef in bericht
news:XlAz8.52$FK2.1293 at nreader2.kpnqwest.net...
> Hoi,
>
> How can one remove the read-only flag from a file
> (e.g. c:\temp\file.txt) using python in win32.
>
> Tx
> --
> ---
> Kind regards, best wishes,
> francis meyvis
>
This does not answer your request. I asked myslef the same question about
half an hour ago, becauase I copies an extensive directory structure from
CDR to my hard disk, with the nasty side effect that all files are marked as
read-only. Windows 98 explorer does not seem to provide any easy way to
switch all the files back to writeable. When thinking about a short Python
script I thought up the work around (which are needed all the time under
windows):
via the search window you can get all the files by searching for *; CTRL-A
selects all and then via properties youcan change all these files at once.
So for now I don't need the Python call to reset the read only flag :-)

regards Gerrit

--
www.extra.research.philips.com/natlab/sysarch
>




From ny_r_marquez at yahoo.com  Mon Apr 29 11:08:04 2002
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 29 Apr 2002 08:08:04 -0700
Subject: Downloading pages from a Windows Intranet
References: <mailman.1017936436.8304.python-list@python.org> <uit77cgha.fsf@ctwd0143.fitlinxx.com>
Message-ID: <8a27e309.0204290708.1061659@posting.google.com>

After taking a look at my options (and considering my humble state of
expertise)
I decided that the easiest route for me to work around the
authentication issue was to use COM to control IE.  It worked.
Using win32com.Distpatch to navigate to the pages works well for my
needs inside our intranet.

However, trying to use DispatchWithEvents (to make a more generic way
of waiting for a page to load for use outside our intranet) has given
me problems.

As suggested, I created a simple class to recieve the events. 
Something like:

class IEEvents:
    def OnBeginDownload(self):
        print "Download began"
    def OnDownloadComplete(self):
        print "Download completed"
    def ...

Then, when I try this: 

import win32com.client
ie = win32com.client.DistpatchWithEvents('InternetExplorer.Application',
IEEvents)
ie.Navigate(someurl)

I see that the OnBeginDownload event gets fired, but that is it.  IE
seems to be help up by something so that it can not complete the
Navigation.  If I kill Python at that point, IE proceeds to finish the
download of the page and displays it.  Any suggestions?

(PS: I stil hope that one day someone savy enough with the use of
calldll will be able to make a module to control the WININET.DLL and
share it with the rest of us. :-)


From timr at probo.com  Mon Apr 22 12:50:37 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 22 Apr 2002 09:50:37 -0700
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com> <1019325450.422371@cswreg.cos.agilent.com>
Message-ID: <dkf8cuo0du07qr445btmukrpg5st9patu5@4ax.com>

weeks at vitus.scs.agilent.com (Greg Weeks) wrote:

>Some comments regarding "conceptual sameness":
>
>On the other hand, suppose we define our own Complex number class.
>
>    x = Complex(3,4)
>    y = Complex(3,0) + Complex(0,4)
>
>Are x and y the same?  Of course.  There is only one 3+i4.  In this case,
>conceptual sameness corresponds to sameness of content.

I disagree.  I don't think you can say "there is only one 3+i4" any more
than you can say "there is only one string 'abc'".  We're deeply into
semantics and meaning here, but correct semantics are critically important
to programming.

I would say that "x and y are equal", but I would not say that "x and y are
the same".  For example, given this next:

    x += Complex(1,0)

If x and y are the same, I would expect this to alter y.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From d.eilers at freenet.de  Wed Apr 17 17:58:15 2002
From: d.eilers at freenet.de (daniela eilers)
Date: 17 Apr 2002 14:58:15 -0700
Subject: mod_python
Message-ID: <e1e40a36.0204171358.7af66fdf@posting.google.com>

hi, experts
currently, i'm a python beginner.
normal python script run with apache, no probs.
trying to run mod_python under apache i get the following error message:

premature end of script headers. 

there is the code of my cgi script:

from mod_python import apache 
         def handler(req): 
              req.content_type = "text/html" 
              req.send_http_header() 
              req.write("Hello World") 
              return apache.OKimport apache 

what's wrong? have somebody an idea?


thx for help
greetz
daniela


From jhalttun at pp.htv.fi  Tue Apr 23 07:14:12 2002
From: jhalttun at pp.htv.fi (janne halttunen)
Date: Tue, 23 Apr 2002 14:14:12 +0300
Subject: curses mouse-events in linux console
References: <3CC3E067.6080107@pp.htv.fi> <lk662jrjav.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3CC54204.5010100@pp.htv.fi>

Michael Hudson wrote:
> janne halttunen <jhalttun at pp.htv.fi> writes:
> 
> 
>>I'm on Mandrake Linux 8.2 with Python 2.2, and I have a problem of
>>mouse-events not emitting properly in curses when on console with
>>gpm. In X it works fine, the getch-method returning instantly on
>>mouse-clicks, but on console it doesn't return until after pressing
>>key on the keyboard.
> 
> 
> Umm, does any curses application respond to the mouse on the console?
> I thought it was an xterm-and-derivatives only thing, but I could well
> be wrong.

Well, it was working in python 2.0 or 2.1 with previous Mandrake Linuxes 
8.1 or 8.0..  And it's working now, except for the triggering of the event.

janne



From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 22:53:13 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 19:53:13 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com> <a9fcut$e9l$0@216.39.172.122> <3cbb8f83@news.mhogaming.com>
Message-ID: <3cbb926a$1@news.mhogaming.com>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote in message
news:3cbb8f83 at news.mhogaming.com...

| # when we reach Python version >= 2.3.x:
| >>> d = d >> 1  # d is NOW an 'int' on most machines
| >>> print type(d)
| <type 'int'>
|
| >>> print d
| 0x43210986

oops-- this should have been:
>>> print d
1126238598

>>> print hex(d)
0x43210986

# <<< sorry for the typo >>>





From padhia at yahoo.com  Fri Apr  5 10:05:08 2002
From: padhia at yahoo.com (P Adhia)
Date: 5 Apr 2002 07:05:08 -0800
Subject: Tutoral < ? < Language Reference
Message-ID: <3281a460.0204050705.4b9b595f@posting.google.com>

Hi,

I think I have graduated to the next level after reading the tutorial.
But I find language reference just describing the formal grammer,
probably meant for programmers who already know python well (rather
too well). Just as an example, by readking lang. reference I still
don't know what "..." is and couldn't find refences to __all__,
__slots__. Is there a "python referece" available somewhere, which
covers every aspects of python, possibly with explanation+examples
(something akin to what K&R is for C) ?

Thanks

P. Adhia


From jeff at ccvcorp.com  Wed Apr 10 15:54:29 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 10 Apr 2002 12:54:29 -0700
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8vk7h$apk$1@news.mathworks.com> <MPG.171d167b861e1d669896a4@news.nwlink.com> <a921d2$2og$1@news.mathworks.com>
Message-ID: <MPG.171e3839572dc9c99896a8@news.nwlink.com>

In article <a921d2$2og$1 at news.mathworks.com>, 
jmarshal at mathworks.com says...
> Jeff Shannon <jeff at ccvcorp.com> wrote:
> > In article <a8vk7h$apk$1 at news.mathworks.com>, 
> > jmarshal at mathworks.com says...
> >> 
> >> To emulate the semantics of pointers, I like using single-element
> >> lists, as some others have suggested.
> 
> > I prefer to *not* emulate the semantics of pointers.  If I have 
> > data that needs to be changed in such a way, I'll put it in a 
> > class and change it through a method.  If I wanted to fiddle with 
> > pointers, I'd program in C.
> 
> But then you weren't the one asking about pointers in Python :)
> 
> I do agree though.  Other non-pointery approaches are cleaner.

Well, my point was exactly that doing pointery things in Python 
is generally a bad idea, regardless of whether it was being 
asked about or not.  There are other, better ways to do those 
same things, in a non-pointery fashion, and if one is truly 
attached to pointer semantics, then one should not use Python.  

Sometimes, when someone asks "How do I do this?", the appropriate 
response really is "Don't do that!"  :)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From fdrake at acm.org  Wed Apr  3 23:32:19 2002
From: fdrake at acm.org (Fred L. Drake)
Date: Wed,  3 Apr 2002 23:32:19 -0500 (EST)
Subject: [maintenance doc updates]
Message-ID: <20020404043219.912DC28696@beowolf.fdrake.net>

The maintenance version of the documentation has been updated:

    http://python.sourceforge.net/maint21-docs/

Preliminary documentation for Python 2.1.3.




From rjones at ekit-inc.com  Thu Apr 18 01:39:52 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 18 Apr 2002 15:39:52 +1000
Subject: Pure python crypto
In-Reply-To: <20020418013346.O924@eecs.tufts.edu>
References: <20020417021703.H16954@eecs.tufts.edu> <3CBE53E8.8080707@verio.net> <20020418013346.O924@eecs.tufts.edu>
Message-ID: <200204181539.52524.rjones@ekit-inc.com>

On Thu, 18 Apr 2002 15:33, Michael Gilfix wrote:
>   Heh. So would I. I'm definitely interested in the project. Perhaps
> we could launch one? It *is* a large undertaking though. What crypto
> modules are already out there though (stuff that's been implemente
> that we could use)...?

A good place to start:

    http://www.google.com/search?q=python+crypto


      Richard





From phr-n2002a at nightsong.com  Mon Apr  1 15:22:15 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 01 Apr 2002 12:22:15 -0800
Subject: verifying ssl certificates
References: <3CA88D15.140053DC@ecn.purdue.edu>
Message-ID: <7x7knrdv9k.fsf@ruckus.brouhaha.com>

"Christopher N. Deckard" <cnd at ecn.purdue.edu> writes:
> I was wondering if anyone has some sample code, or could point me at
> some documentation, to where I can see how to verify SSL
> certificates after making an SSL connection via sockets.  Is it as
> simple as comparing the response of the server to that of something
> in a file?

Usually you configure the SSL stack to do this for you.  If the
certificate doesn't verify, the SSL connection won't even open,
so your application will never get to see an invalid cert.


From jimw1956 at cs.comnojunk  Sun Apr  7 10:36:54 2002
From: jimw1956 at cs.comnojunk (Jim Williams)
Date: 07 Apr 2002 14:36:54 GMT
Subject: Can I print from Python?
Message-ID: <20020407103654.19127.00002640@mb-fm.news.cs.com>

Hi all...I'm new to Python and really enjoying it so far, but I am somewhat
perplexed that as far as I have seen it has no facilities for printing to a
printer?  Surely this is possible. Someone please enlighten me...I have done a
good bit of research and searching and can find no reference to how to do this.




From peter.west at hotmail.com  Sun Apr 21 17:18:11 2002
From: peter.west at hotmail.com (peter.west at hotmail.com)
Date: Sun, 21 Apr 2002 21:18:11 GMT
Subject: This guy is trying to steal my mates wife.  7246
Message-ID: <n0Gw8.40272$51.503475@wards>

This guy is trying to steal my mates wife.  He dumped his last partner for having too much "baggage" so if my mates wife becomes too much baggage he will forget her too.  we can arrange this by giving him loads of hassel.  For anyone that would like to help, his details are below:-

Richard Harper
26 Ashview Gardens
Ashford
TW15 3RD
Mob: +44(0)07973 527357
Bus'Mob: +44(0)7889769261

Give the bastard hell :-)

Thanks to all those who help.

cdizgycyrtqqvxiivdqtnijdlzlngswpewbkzsvpymvthgsdmgchpgfptrlkzeormfnyoihnphfqtnkjydk



From Chris.Barker at noaa.gov  Mon Apr 29 13:33:38 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 29 Apr 2002 10:33:38 -0700
Subject: Numeric question
References: <pan.2002.04.25.21.32.03.310021.2578@hooper.net> <YM2y8.71958$HH5.5067@rwcrnsc51.ops.asp.att.net> <3CC9A844.F1BE7E7B@noaa.gov> <%O1z8.808$SY6.23031@news.itd.umich.edu>
Message-ID: <3CCD8333.F9554C82@noaa.gov>

Cameron Hooper wrote:
> Thanks for the replies. So my understanding now is that the concatenate
> function really combines two references to two separate locations in memory
> into a single object. The two locations are not contiguous. Data is not
> 'moved around' so to speak. So x = concatenate(y,z) can be viewed as
> returning a reference to two non-contiguous references.  But x =
> array(concatenate(y,z)) (or using copy()) actually copies data.

Not quite. frankly, I'm a bit confused about how concatenate works, as I
don't see how it would be possible to re-use the data in the input
arrays. Also, a test shows that contatonate does not return a reference
to the same data:

>>> from Numeric import *
>>> a = ones((2,3))
>>> b = ones((3,3)) * 4
>>> 
>>> c = concatenate((a,b))
>>> a
array([[1, 1, 1],
       [1, 1, 1]])
>>> b
array([[4, 4, 4],
       [4, 4, 4],
       [4, 4, 4]])
>>> c
array([[1, 1, 1],
       [1, 1, 1],
       [4, 4, 4],
       [4, 4, 4],
       [4, 4, 4]])
>>> c.iscontiguous()
1
# so in this case, c is contiguous!
>>> 
>>> c[0,0] = 5
>>> c
array([[5, 1, 1],
       [1, 1, 1],
       [4, 4, 4],
       [4, 4, 4],
       [4, 4, 4]])
# change a value in c
>>> a
array([[1, 1, 1],
       [1, 1, 1]])
# a does not change, so they are not sharing data.
# this is different than slicing, which does result in two arrays
sharing data:

>>> d = c[:,2]
>>> d
array([1 1 4 4 4])
>>> d[2] = 6
>>> d.iscontiguous()
0
# slicing often results in discontiguous arrays
>>> c
array([[5, 1, 1],
       [1, 1, 1],
       [4, 4, 6],
       [4, 4, 4],
       [4, 4, 4]])
# so in this case, changing d did change c, as they do share data.

I have to say that I'mm in the dark as to why concatenate does not
always result in a contiguous array, but maybe with some more thought I
could figure it out.

-CHB
 


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From chooper at umich.edu  Tue Apr  2 23:20:37 2002
From: chooper at umich.edu (CGH)
Date: Tue, 02 Apr 2002 23:20:37 -0500
Subject: Newbie: Classes and Lists
References: <8Guq8.1209$zT5.41108@news.itd.umich.edu> <3CAA7F39.94FE7EA9@engcorp.com>
Message-ID: <oqvq8.1211$zT5.40775@news.itd.umich.edu>

Thanks, that fixed my problem and taught me something I had not known about 
classes in Python.

Cameron


From maik.hertha at volkswagen.de  Wed Apr  3 01:57:59 2002
From: maik.hertha at volkswagen.de (Maik Hertha)
Date: Wed, 03 Apr 2002 08:57:59 +0200
Subject: ctrl-c and os.system problem
References: <mailman.1017777544.11640.python-list@python.org>
Message-ID: <3CAAA7F7.9000304@volkswagen.de>

Noom Knews wrote:

>I have a simple program like this
>
>os.system("ls -lR / > /tmp/ls.out")
>print "end"
>
>When I press ctrl-c, the "end" message is still print.
> I'd like to terminate the whole process when pressing
>ctrl-c instead of just "end".
>
>I tried to catch the signal already but can't catch
>the SIGINT or anything else at all during the
>os.system call.   This does not work either
>
>try:
>   os.system("ls -lR / > /tmp/ls.out")
>   print "end"
>except:
>   print "get all exception"
>
>This will keep printing "end" when ctrl-c is pressed
>too. I think this could be os.system problem. If I use
>other statement instead of os.system, the program
>works as I expected.
>
>This was tried on Redhat Linux with python 1.5.2
>
works on python 2.2

-- 

mit freundlichem Gru? /
best regards

Maik Hertha

--------------------------------------------------- h+h
EBSP Anwenderbetreuung, +49 5361 9-74950
Volkswagen AG / Brieffach 1721 / D-38436 Wolfsburg
http://ebsp.wob.vw.de              
maik.hertha at volkswagen.de
---------------------------------------------------
hartmann+hertha
it (beratung / entwicklung / support)
http://www.hartmann-hertha.de
mhertha at hartmann-hertha.de
--------------------------------------------------- h+h




From maxm at mxm.dk  Thu Apr  4 08:34:55 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 04 Apr 2002 15:34:55 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk> <slrnaaok7j.4ei.jon+usenet@snowy.squish.net>
Message-ID: <3CAC567F.6090407@mxm.dk>

Jon Ribbens wrote:
> In article <3CAC1665.8040209 at mxm.dk>, Max M wrote:

>>>relative_url("http://foo/a/b/c", "http://foo/a/d")
>>> --> return: "../d"
>>
>>I hope you mean that the result should be "../../d" ??
> 
> 
> No, he means "../d".
> 
> http://foo/a/b/c + d       == http://foo/a/b/d
> http://foo/a/b/c + ../d    == http://foo/a/d
> http://foo/a/b/c + ../../d == http://foo/d

Then what is:

http://foo/ + d       == ?

regards Max M



From magnus.heino at sigma.pleon.se  Wed Apr 17 14:48:28 2002
From: magnus.heino at sigma.pleon.se (Magnus Heino)
Date: Wed, 17 Apr 2002 20:48:28 +0200
Subject: minidom and =?ISO-8859-1?Q?=E5=E4=F6?= once again :-P
References: <WXdv8.2962$ZN4.5053@news2.bredband.com> <j47kn6bb5t.fsf@informatik.hu-berlin.de>
Message-ID: <asjv8.3473$ZN4.5422@news2.bredband.com>

Martin v. L?wis wrote:

> Magnus Heino <magnus.heino at pleon.sigma.se> writes:
> 
>> >>> d.toxml()
>> '<?xml version="1.0" ?>\n<test>\xe5\xe4\xf6</test>'
> 
> This is not what I'm getting. I get
> 
> ...
>     writer.write(data)
>   File "/usr/local/lib/python2.2/StringIO.py", line 139, in write
>     s = str(s)
> UnicodeError: ASCII encoding error: ordinal not in range(128)
> 
> Apparently, you've changed the system default encoding, so that the
> conversion from Unicode strings to byte strings silently converted the
> Unicode string.

Well, I got confused by this;

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.102.htp

But I have removed that now.
 
> In any case, your best bet is to use an explicitly-encoding stream:
> 
>>>> s=StringIO.StringIO()
>>>> d.writexml(codecs.getwriter('utf-8')(s))
>>>> s.getvalue()
> '<?xml version="1.0" ?>\n<test>e\xc3\xa4\xc3\xb6</test>'

Ah.... there we go! Thank you for fast and accurate answers!

Now I finally understand whats going on, and why :-)

Maybe someone with commit access could add some short example like this one 
to the python documentation for minidom?

-- 

  /Magnus


From aleax at aleax.it  Sat Apr 27 02:51:21 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 06:51:21 GMT
Subject: iis: comparsion of two lists
References: <20020426224924.GA1318@rehor.surfbest.net> <20020426232802.GA1927@rehor.surfbest.net>
Message-ID: <JTry8.26517$8D3.706486@news1.tin.it>

Matej Cepl wrote:

> On Fri, Apr 26, 2002 at 10:49:24PM +0000, Matej Cepl wrote:
> 
>>    erase = ()
>>    for msg in proc_old:
>>    if not(msg in loc_old):
>>    erase.append(msg)
> 
> Oh, boy,
> 
> I an answer myself. Sorry, for that. Am I correct, that mere
> erase = ()
> erase = filter (lambda m: return not(m in loc_old), proc_old)
> 
> would fly?

It would, but the first assignment is useless and the approach
needlessly complicated IMHO.

erase = [ m for m in proc_old if m not in loc_old ]


Alex



From eppstein at ics.uci.edu  Sat Apr 20 12:16:20 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 20 Apr 2002 09:16:20 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org> <eppstein-2C56E0.23265219042002@news.service.uci.edu> <mailman.1019285080.29613.python-list@python.org>
Message-ID: <eppstein-2EE7C3.09162020042002@news.service.uci.edu>

In article <mailman.1019285080.29613.python-list at python.org>,
 Gerhard H?ring <gerhard at bigfoot.de> wrote:

> > The point is that your responses don't have the appropriate References: 
> > header, so they don't get threaded with other messages on the same topic.
> > It appears as if you are using a mail agent rather than a news agent, 
> > since References: is a news header (and your other headers say you're 
> > using MS Outlook).
> 
> This might also be a bug/missing feature in Mailman's mail<->news
> gateway, but if this, my response shows up correctly for you, then it's
> another proof that Mutt rocks and Outlook sucks :-P

Did we need any confirmation of that?
Yes, it shows up correctly.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From gbreed at cix.compulink.co.uk  Thu Apr  4 10:22:02 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Thu, 4 Apr 2002 15:22:02 +0000 (UTC)
Subject: ACCEPTED: PEP 285
References: <7xu1qr4jnq.fsf@ruckus.brouhaha.com>
Message-ID: <a8hr2q$qfo$1@thorium.cix.co.uk>

Paul Rubin wrote:

> Guido van Rossum <guido at python.org> writes:
> > > I think there will need to be a format code for truth values:
> > > 
> > >   ("%b" % True) => "True"
> > 
> > There's no need.  "%s" % True already yields 'True'.  if you want
> > coercion to bool, use "%s" % bool(b).
> 
> In that case there's also no need for %d, since "%s" % 17 already
> yields '17'.  If ints have a format code and bools don't, one is
> left with the impression that bools aren't as respectable a type
> as ints.  That doesn't seem in the spirit of the PEP.

The format code for integers is %i, not %d.  There's a very good reason 
for %d being there.  How else to you propose getting

>>> '%d'%1
'1'
>>> '%d'%1.0
'1'
>>> '%d'%1.3
'1'
>>> '%d'%(10L**100)
'1000000000000000000000000000000000000000000000000000000000000000000000000
000000
0000000000000000000000'
>>> '%d'%bool.True
'1'

to all work?  (bool module cut and pasted from the PEP)  %i itself is the 
simplest way to format ints, longs, bools and the integer part of floats 
so that they all look like integers between 1.5 and 2.3 (he says, not 
having 1.5 to check).  Also a good way of getting bools looking the same 
before and after the PEP.  And a nice simple way to do

>>> '%03i'%5
'005'

without importing string.

I'll bet most of the time you want to stringify booleans it won't be as 
'true' of 'false' either.  Rather something like

('It worked!', 'It didn't work!')[success]

Still, yes, if one type can have a format code why not the others?  Let's 
see what would make sense for xrange, function, slice, frame, ...


                  Graham

     <http://www.microtonal.co.uk/>


From aleax at aleax.it  Sat Apr 13 12:00:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 16:00:24 GMT
Subject: hex and long
References: <8N9+ELA0u5t8Ew2n@jessikat.fsnet.co.uk> <mailman.1018668146.15677.python-list@python.org> <GLcQVLAwU$t8EwQ3@jessikat.fsnet.co.uk>
Message-ID: <sCYt8.16132$vF6.501773@news2.tin.it>

Robin Becker wrote:
        ...
> Is there an easy way to get back to the low 32 bits of a long or do I
> have to do bit twiddling?

Say
>>> x=23**45

Now, in what category would you range:

>>> x & 2**32-1
2112869559L

>>> x^(x>>32<<32)
2112869559L

>>> x & 0xffffffffL
2112869559L

>>> x & ~(-1L<<32)
2112869559L

To me, they all look BOTH easy AND bit-twiddling (how else than by
twiddling bits are you going to extract a certain string of bits?).


Alex



From loewis at informatik.hu-berlin.de  Thu Apr 18 07:12:00 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 13:12:00 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018875990.21271.python-list@python.org> <j43cxwm1o4.fsf@informatik.hu-berlin.de> <a9id9a$33oqt$1@ID-99513.news.dfncis.de> <m37kn6wzu4.fsf@mira.informatik.hu-berlin.de> <a9m5r9$4b8ru$1@ID-99513.news.dfncis.de>
Message-ID: <j4ads18dmn.fsf@informatik.hu-berlin.de>

"P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:

> But binary output is not 100% compatible with system libraries. If you
> use the native linker and assembler, the code will be compatible. This
> has been considered as "God given truth" in the sun-managers list for
> more than 10 years (since I first subscribed). 

That is unfortunate, since the truth has changed over the
years. Certain binutils releases don't work correctly on certain
releases of Solaris; later binutils releases have been corrected.

Likewise, certain Solaris versions contain bugs that only show up when
used with gcc; both in Solaris /usr/ccs/bin/as and /usr/ccs/bin/ld.
Furthermore, Solaris (in all versions) is missing features that make
GNU ld superior to /usr/ccs/bin/ld for certain applications.

Hence, on the gcc list, the "God given truth" for the last five years
is that gcc is best used with a recent binutils release.

> > Actually, you *can* talk the Solaris linker to incorporate non-PIC
> > text into a shared library. For that, it will make the text segment
> > writable. The -mimpure-text option of gcc triggers that feature.
> 
> And open a full can of worms. Why hack the linker flags when you can
> impose yourself some discipline in your administration tasks and
> make life easier in the future? :)

Because /usr/ccs/bin/ld fails to link the extension otherwise. It's a
work-around, but your choices are to either have the extension
available, or not to be able to use the extension.

> BTW, this advice of yours does apply to the curses modules under
> Solaris, as they are compiled as non-PIC code.

Can you elaborate? /usr/lib/libcurses.so is PIC enough for me...

> > If you link static libraries into Python, I recommend to make the
> > modules that use them also static. Then you don't need PIC code for
> > those.
> 
> And still they won't work. As is the case of the curses modules and the
> readline modules in 2.1 and 2.2...

This is not true. If you compile both curses and readline as static
modules, it is irrelevant whether the libraries themselves are
available as static or shared libraries - either setup works just fine.

To compile a module as static, you edit Modules/Setup appropriately
before making Python.

Regards,
Martin


From h_schneider at marketmix.com  Fri Apr 12 02:51:39 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 12 Apr 2002 08:51:39 +0200
Subject: Include zlib.dll in McMillan Installer
References: <a93ld3$jb4$03$1@news.t-online.com> <Xns91ED608AADBDEgmcmhypernetcom@199.171.54.213>
Message-ID: <a95vrj$4eh$02$1@news.t-online.com>

Hello Gordon,

at first my compliments for your Installer :)

> In ancient days (1.5.2), zlib.pyd loaded zlib.dll. But
> since then, zlib.pyd has changed to statically link in
> zlib, so zlib.dll should not be needed (in fact, it
> doesn't exist in the distribution).

I use the Activestate distribution of Python 2.1 It seems, that
here zlib.dll is still needed. It's located in \python\DLLs\zlib.dll

A view into the arhive says that zlib.pyd is included, but not the
dll:

[(0, 354059, 354059, 0, 'z', 'out1.pyz'),
 (354059, 20972, 20972, 0, 'm', 'iu'),
 (375031, 16264, 16264, 0, 'm', 'archive'),
 (391295, 192, 192, 2, 's', '_mountzlib'),
 (391487, 16, 16, 2, 's', 'useUnicode'),
 (391503, 1909, 1909, 2, 's', 'umsg_agent'),
 (393412, 19399, 57439, 1, 'b', 'win32api.pyd'),
 (412811, 10762, 28672, 1, 'b', '_imagingtk.pyd'
 (423573, 16074, 49236, 1, 'b', '_sre.pyd'),
 (439647, 123592, 278528, 1, 'b', '_imaging.pyd'
 (563239, 9219, 28760, 1, 'b', '_tkinter.pyd'),
 (572458, 23232, 61524, 1, 'b', 'zlib.pyd'),
 (595690, 10445, 32855, 1, 'b', '_socket.pyd'),
 (606135, 18777, 57344, 1, 'b', 'PyWinTypes21.dl
 (624912, 320031, 708696, 1, 'b', 'python21.dll'
 (944943, 388214, 950272, 1, 'b', 'tk83.dll'),
 (1333157, 244541, 483328, 1, 'b', 'tcl83.dll'),
 (1577698, 45250, 96528, 1, 'b', 'IMM32.dll')]

All the best,
Harald Schneider

>
> Installer will always include zlib[.pyd/.so], unless
> your installation doesn't have it, or you turn it off
> by editting config.dat. Windows comes with zlib.pyd,
> so I expect it's there (you can check with ArchiveViewer.py
> to see if it's in the exe).
>
> So more details are needed.
>
> -- Gordon
> http:/www.mcmillan-inc.com/
>
>




From emmanuel at onda-latina.com  Sun Apr  7 01:38:06 2002
From: emmanuel at onda-latina.com (manouche)
Date: Sun, 7 Apr 2002 08:38:06 +0200
Subject: global instance question (newbie)
References: <a8njqq$phs$1@wanadoo.fr> <Xns91E8E54158554cliechtigmxnet@62.2.16.82>
Message-ID: <a8opho$bh6$1@wanadoo.fr>

thank you

Chris Liechti <cliechti at gmx.net> a ?crit dans le message :
Xns91E8E54158554cliechtigmxnet at 62.2.16.82...
> "manouche" <emmanuel at onda-latina.com> wrote in
> news:a8njqq$phs$1 at wanadoo.fr:
>
> > hello,
> >
> > How can I create a global object that I can be used in all the files
> > of my project?
> > Using "global" I can only access to the instance in the file where it
> > was created.
>
> put it in a separate module and import it where you need it.
> of course you can add and modify the values on that module.
>
> there is only one module instance in the interpreter, no matter how much
> you import it (singelton is the keyword).
>
> chris
>
> --
> Chris <cliechti at gmx.net>
>




From logiplexsoftware at earthlink.net  Wed Apr  3 15:43:06 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 12:43:06 -0800
Subject: Threading problem
In-Reply-To: <a8f8f6$vdu$1@news.rchland.ibm.com>
References: <a8f8f6$vdu$1@news.rchland.ibm.com>
Message-ID: <20020403124306.3d8e30ba.logiplexsoftware@earthlink.net>

On Wed, 3 Apr 2002 09:52:07 -0600
Larry Whitley wrote:

> I have a PC with two processors at work and have decided to try to use
> threading to run some long running console applications.  My basic idea
is
> to start two threads, one for each processor, and give them a list of
things
> to do.  As each thread finishes it one job will go to the list and pick
off
> the next thing to do.  When the list is empty, both threads are done.

You've already got your question answered, but your post raises another
question:  I don't believe you are going to be able to take much advantage
of SMP even with threads since the Python interpreter itself runs on a
single processor.  Does anyone know different?  I've been somewhat confused
on this issue since I believe Python uses the native threads on any given
platform.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From philh at comuno.freeserve.co.uk  Tue Apr  9 11:12:46 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 9 Apr 2002 16:12:46 +0100
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at>
Message-ID: <slrnab617e.97a.philh@comuno.freeserve.co.uk>

On Tue, 09 Apr 2002 10:06:01 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
>Laura Creighton wrote:
>>  However, if your real goal is
>> to hack C, and all the above is just one misused word -- you want
>> K & R.  Everybody wants K & R.  
>
>Not really. This book is really,really boring.

I disagree. K&R has a mixture of clarity and concision that makes 
it a delight to read.

IMO, anyone for whom K&R isn't to their taste, probably won't make a 
very good programmer anyway.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From see at below  Tue Apr 23 07:37:07 2002
From: see at below (Paul Foley)
Date: 23 Apr 2002 23:37:07 +1200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org>
Message-ID: <m2g01my7bw.fsf@mycroft.actrix.gen.nz>

On Mon, 22 Apr 2002 23:35:27 -0700, James J Besemer wrote:

>> On Sun, 21 Apr 2002 02:32:29 -0700, James J Besemer wrote:
>> 
>> A lot of people think "Lisp" means "Scheme".
>> Scheme sucks, so they think "Lisp" sucks.

> That complaint seems specific to the Lisp community.  I don't sense that a lot of
> people here know either language well enough to be making that particular complaint.

People who're already knowledgeable about Lisp obviously aren't
susceptible to thinking that Lisp sucks just because they don't like
Scheme.  The problem is only with people who _don't_ know Lisp, who
keep hearing that Scheme is Lisp.

>> Well, that's not entirely true.  Lisp has always used lexical scoping
>> in compiled code.  There used to be a difference between compiled and
>> interpreted code, though, where the interpreter used (only) dynamic
>> scope.

> My statement was 100% accurate, as "original Lisp" had no notion of compiled code.

See McCarthy's 1960 paper, available from his web site.  Page 22,
immediately under the heading "The LISP Programming System":

  ...  [LISP] has been used, or will be used for:

  1) Writing a compiler to compile LISP programs into machine code.


John Aycock's paper entitled "A Brief History of Just-In-Time" says:

  Instead we suspect that the earliest published work on JIT
  compilation is McCarthy's 1960 LISP paper. He mentions compilation
  of functions into machine language, a process fast enough that the
  compiler's output needn't be saved [McCarthy 1960]. This can be seen
  as an inevitable result of having programs and data share the same
  notation [McCarthy 1981].


[Also see MIT AI Memo 39, on the LISP 1.5 compiler]

> The original semantics (as formally defined in Lisp itself [McCarthy]) called for
> dynamic binding.  Most original implementations did it that way, many later

ISTM reading somewhere that the dynamic scoping in early
implementations was actually the result of an error; compiled Lisp
has always used lexical scoping (for normal variables; e.g., those
not declared "special" or "common" in LISP 1.5)

Where do you think dynamic scope is called for?

> However, you have to admit that having to surround virtually each and every
> statement, expression, sub-expression, etc. in a pair of parentheses is at the very
> least more tedious and error prone for the programmer than, say, using operator
> precedence to organize arithmetic, semicolons or newlines to separate statements and
> braces or indentation to group statements.

No I don't have to admit that...I don't even have to admit that there
are any parentheses in Lisp :-)  [parentheses are only present in the
printed representation, to provide the necessary structure; they're
not part of the code -- if you work with a structure editor in a Lisp
system, instead of a text editor, there are no parentheses anywhere]

In any case, parentheses are not used around sub-expressions, or to
separate or group statements.  They're only used (in evaluated
contexts) for "function" calls, and most other languages, including
Python, use parentheses for that purpose, too:

 Python: foo(x)
 Lisp: (foo x)

Same number of parentheses, right?  I'd expect Lisp syntax to be
rather less error-prone, if anything, since it's much simpler
[there're no special syntaxes for specific data types, operator
precedence rules, etc., etc., to remember; just a single, clean
and obvious rule, and a small handful of "special operators" with
nonstandard evaluation rules (20 in ANSI Common Lisp, IIRC)]

> prefer controlling the semantics with an "Algol-like" syntax.

Which is also possible in Lisp, of course; in fact, Lisp was
originally intended to have an "Algol-like" syntax (McCarthy's
M-expressions) on top of S expressions...

>> [And it's worth noting that, as with other aspects of Lisp, people
>> just keep reinventing it, badly --

> Hmmm.... I was recently accused of making this claim, which is entirely untrue.

Eh?  What's untrue?  The claim or that you made it?

-- 
In modern physics, the more logical you are, the more wrong you are.
                                                           -- Lama Govinda
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From dsavitsk at e-coli.net  Thu Apr 11 02:51:35 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 11 Apr 2002 06:51:35 GMT
Subject: Sound programming?
References: <3CB50DBC.1B0E533F@alum.mit.edu>
Message-ID: <Xnat8.6513$d7.1742279@newssrv26.news.prodigy.com>

for win32 only look at the winsound module.

-d


"Will Ware" <wware at alum.mit.edu> wrote in message
news:3CB50DBC.1B0E533F at alum.mit.edu...
> Somebody was asking me this evening if there's an easy way to
> play a tone of a specified frequency and duration in Python.
> I didn't know off the top of my head. I've since poked around
> the web a bit and it appears like the Snack Sound Toolkit
> (http://www.speech.kth.se/snack/) might be a good way to go.
> One of its virtues is being cross-platform, which I suspect is
> among his unstated wishes. Are there any other options?




From achim.domma at syynx.de  Sat Apr 27 06:26:04 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Sat, 27 Apr 2002 12:26:04 +0200
Subject: problems redirecting unittest output to sys.stdout
References: <aadplo$rs7$06$1@news.t-online.com> <1quy8.68737$vF6.2063037@news2.tin.it>
Message-ID: <aadttj$ujh$04$1@news.t-online.com>

"Alex Martelli" <aleax at aleax.it> wrote:

> > I want to redirect the output of a unittest to a textfile (on Windows).
> > The default stream is sys.stderr which can not be redirected on windows.
>
> Of course it can be -- depends on what Windows version you have.  With
> NT's CMD.EXE (thus I believe also /2000 and /XP), 2>blah on the command
> line works fine.

yes, 2> works fine ! Thanks !

Achim




From mhammond at skippinet.com.au  Fri Apr 19 11:03:18 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 19 Apr 2002 15:03:18 GMT
Subject: ANN: Python For Win32 Extensions documentation in HtmlHelp format.
Message-ID: <3CC031B7.20104@skippinet.com.au>

Hi all,
   I have decided to abandon Windows Help files, and move to the much 
better HtmlHelp format (as used by ActivePython).  There is a single 
.chm for all the extensions, meaning all links should correctly work, 
and a single index.

It will appear in the next win32all, but in the meantime you can 
download the current version via 
http://starship.python.net/crew/mhammond/win32/Downloads.html

Mark.



From gbreed at cix.compulink.co.uk  Wed Apr  3 08:19:04 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Wed, 3 Apr 2002 13:19:04 +0000 (UTC)
Subject: date / time
References: <3CAAE929.8000709@beth.uniforum.org.za>
Message-ID: <a8evg8$6dd$1@thorium.cix.co.uk>

mixo wrote:

> By "valid" date I meant subtracting  a day, and  then finding that the
> new date does actually exsist.

The date returned by gmtime should always be valid.  To check an arbitrary 
date, try

>>> (year, month, day) == time.gmtime(time.mktime((year, month, day, 1, 1, 
1, 1, 1, 0)))[:3]
1

> One note/question on time.gmtime():
> +++++++++++
> time.gmtime( time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60))
> +++++++++++
> 
> returns
> +++++++++++
> (2002, 2, 27, 22, 0, 0, 2, 58, 0)
> +++++++++++
> 
> There is a day missing. If you add 1 second , i.e ((24*60*60)+1) you get
> the required dated. What's going on?

It's probably something to do with your timezone.  You can work strictly 
in GMT if you ignore strptime and do

>>> year, month, day = map(int,re.findall('(\d+)', '2002-03-01'))
>>> time.gmtime(time.mktime((year, month, day, 0, 0, 0, 0, 0, 0))-24*3600)
(2002, 2, 28, 0, 0, 0, 3, 59, 0)


                       Graham


From mkellner at rpvreport.com  Mon Apr 29 18:39:31 2002
From: mkellner at rpvreport.com (Martin Kellner)
Date: 29 Apr 2002 15:39:31 -0700
Subject: Free tool for reports development
Message-ID: <6a5f9873.0204291439.41156c92@posting.google.com>

Dear community.
 
Sorry for disturbing, but I would like to notify you that we're
offering a FREE TOOL for report development: Rpv Printing System.
 
This tool lets you develop Windows reports from any programming
language and we've been noticed that Python community could be glad to
hear about it.
 
This program is free for distribution so if you consider that it is
good, don't hesitate to develop and distribute it.
 
The application is available at http://www.rpvreport.com and it comes
in english and spanish (for now).
 
Thanks for your time.
Martin Kellner
mkellner at rpvreport.com


From phd at phd.pp.ru  Tue Apr 30 11:16:20 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 30 Apr 2002 19:16:20 +0400
Subject: How to retrive attached file from newsgroup mail
In-Reply-To: <63170f57.0204300655.3ae78971@posting.google.com>; from sanjay2kind@yahoo.com on Tue, Apr 30, 2002 at 07:55:17AM -0700
References: <63170f57.0204300045.788bc634@posting.google.com> <mailman.1020157433.29195.python-list@python.org> <63170f57.0204300655.3ae78971@posting.google.com>
Message-ID: <20020430191620.K19566@phd.pp.ru>

On Tue, Apr 30, 2002 at 07:55:17AM -0700, sanjay wrote:
> Oleg Broytmann <phd at phd.pp.ru> wrote in message news:<mailman.1020157433.29195.python-list at python.org>...
> > On Tue, Apr 30, 2002 at 01:45:42AM -0700, sanjay wrote:
> > >   Im using NNTP to retrive the newsgroups mails to display on web
> > > pages.The email is multipart i,e text/html with doc,pdf,gif file has
> > > attached. Is there any python script to get all the plain text/html
> > > message with attached file from mail.
> > 
> >    http://phd.pp.ru/Software/Python/#extract_mime,
> > http://phd.pp.ru/Software/Python/#mimedecode and faster mirrors at
> > http://phd.by.ru/Software/Python/ and http://phd2.chat.ru/Software/Python/
> > 
> >    I am working on rewriting mimedecode.py to use email package. Actually
> > I've finished rewriting it, but the code requires Python 2.3 (i.e. it
> > requires my latest patches for the email package).
> > 
>  extract_mime is helpful for 7/8 bit, not for base64(attached file).
> any python way to  solve it.

   mimedecode.py. It does not extracet attachments, but you can modify the
code any way you want.

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




From peter at engcorp.com  Tue Apr  2 23:06:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 23:06:40 -0500
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net>
Message-ID: <3CAA7FD0.56C285E0@engcorp.com>

Jim Jacobs wrote:
> 
> Is it possible to simulate "C" style static variables in functions?  In
> methods?

Maybe ...  what do you mean by simulate, and what characteristics
of the C statics would you like to have?

In C, static inside a function causes the variable to be assigned
in the heap rather than on the stack (to put it one way) and
that gives the variable persistence across invocations of
the function (but scope local to the function).

In Python, you would use either an instance variable, inside
a method of an object, or the answer would be "no, but can
you get by with using 'global'?".

-Peter


From pyth at devel.trillke.net  Mon Apr 29 13:44:25 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 19:44:25 +0200
Subject: Remove masked elements from an array
In-Reply-To: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>
Message-ID: <20020429174425.GN16217@devel.trillke>

On Mon, Apr 29, 2002 at 01:26:25PM -0400, PoulsenL at capecon.com wrote:
> I have a Numeric array with missing elements from which I wish to calculate
> a histogram.  However, in order to do so I must first remove the 'None'
> values from the array.  I have a kludgy solution, but I was wondering if
> someone could point me in the right direction on what is the proper method
> of collapsing a masked array to just the non-masked values.

there are too solutions. choose yourself :-)

nums = [ 1,2,3,None,2,3]

#version1
def f(x): return x is not None
newnums = filter(f, nums)

#version2
newnums = [ x for x in nums if x is not None]

The first version uses filter. 
'filter' calls 'f' for every item in 'nums' and puts
the result into the new list iff it has a 'true' value.

The second version uses list comprehension.
It basically does the same if you look closely
at it. 

regards,

    holger




From marklists at mceahern.com  Wed Apr 17 13:29:51 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 17 Apr 2002 12:29:51 -0500
Subject: UTC equivalent of time.mktime() ?
In-Reply-To: <1019062680.653.398.camel@tp6>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEJEKLAA.marklists@mceahern.com>

[Graham Ashton]
> I appreciate that I could (and thanks for the suggestion), but does this
> really mean that via the standard library there is no easy way to
> convert "1 Jan 1970" to 0 (i.e. string to seconds since epoch)?
>
> Perhaps I should inspect the mxDateTime code...

Anyway, mxDateTime rocks.  I don't know why it's not part of the standard
distribution.

Here's a fer instance on why you oughter use it:

	Windows doesn't have strptime

cygwin:

$ python
Python 2.2 (#1, Dec 31 2001, 15:21:18)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import time
>>> time.strptime("1 Jan 1970", "%d %b %Y")
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'module' object has no attribute 'strptime'

plain win:

C:\Python22>python
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import time
>>> time.strptime("1 Jan 1970", "%d %b %Y")
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'module' object has no attribute 'strptime'

linux:

$ python
Python 2.2.1 (#1, Apr 10 2002, 09:15:55)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import time
>>> time.strptime("1 Jan 1970", "%d %b %Y")
(1970, 1, 1, 0, 0, 0, 3, 1, 0)

whereas with mxDateTime, you don't even have to bother with the format:

>>> mx.DateTime.DateTimeFrom("1 Jan 1970")
<DateTime object for '1970-01-01 00:00:00.00' at 816d7b8>

Of course, epoch-relative values are still in local time:

>>> epoch = mx.DateTime.DateTimeFrom("1 Jan 1970")
>>> epoch.ticks()
21600.0

You want that to say zero, right, no matter what the local time zone is?

Thinking about UTC vs. localtime always starts to hurt my head...

// m





From dsavitsk at e-coli.net  Fri Apr 12 12:42:44 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 12 Apr 2002 16:42:44 GMT
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de>
Message-ID: <78Et8.6685$d7.1856092@newssrv26.news.prodigy.com>

See

http://groups.google.com/groups?hl=en&th=28cf583f00d81e60&seekm=a949oq%24it2
%241%40nyheter.chalmers.se&frame=off


"Uwe Schmitt" <uwe at rocksport.de> wrote in message
news:a96q8o$9cd5d$1 at hades.rz.uni-sb.de...
> Hi,
>
> I'd like to send "raw" data to a printer, how can I do this,
> using Python ???
>
> Greetings, Uwe.
>
> --
> Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
> Universitaet des Saarlandes                     Angewandte Mathematik
> Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
> Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468




From paul at boddie.net  Thu Apr 11 06:26:21 2002
From: paul at boddie.net (Paul Boddie)
Date: 11 Apr 2002 03:26:21 -0700
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer   division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <Xns91ECDF3AC49F7lumberjack@216.148.53.85> <3CB5221D.EE464B76@engcorp.com>
Message-ID: <23891c90.0204110226.57d6bb96@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3CB5221D.EE464B76 at engcorp.com>...
> 
> In any case what's your point?  You have every right, and 
> apparently the strong motivation, to fork off your own branch
> of the language and continue development in your own fashion.
> Surely that will let you demonstrate how to bring legitimacy
> in ways the current project leads have not been able...
> 
> So I suggest you take my advice, and fork off.

I think this is really unfair, as well as failing to address the valid
points raised by the contributor in question. I presume that
"Lumberjack" is, like I am, grateful for being able to use Python in
as many projects as they have time for. However, having made a
commitment to use the language, they probably feel that they're being
put in a position where they have to decide between...

  * Tracking the "latest and greatest" releases, but having to put up
    with compatibility issues, as well as needing to upgrade their
    "infrastructure" frequently (or at least more often than they
    might like).

  * Sticking with previous releases and missing out on fixes of
    potentially critical bugs, new and exciting packages, and so on.

This isn't implicitly meant as a critique of the core Python
development team (before people read it that way) - indeed, the team
have done a great job of managing compatibility issues - but this is a
choice that affects a reasonable number of people. Unfortunately, I
haven't seen substantial amounts of authoritative advice on how people
should resolve such issues, other than "upgrade and fix your code" -
something that the Red Hat experience suggests is not always possible
or desirable.

Finally, it just isn't feasible for some rogue user to fork Python and
suddenly have their needs satisfied (although I suspect that the
comment was made with this realisation in mind) - maintaining and
developing a system as complicated as Python requires a lot of effort
and we have to be thankful that there are a number of people who can
spend large amounts of their time doing so. Indeed, adopting a
language like Python often occurs when people have realised precisely
that a well-supported language is a better choice than a home-made
scripting language.

Paul


From hzi at uol.com.br  Tue Apr 16 00:02:22 2002
From: hzi at uol.com.br (Henry Lebowzki)
Date: Tue, 16 Apr 2002 04:02:22 GMT
Subject: Did anyone have problems installing Numeric module on Debian Woody
Message-ID: <3cbba193.2712952@news.cis.dfn.de>

Hi-

	I've been having some problems installing the Numeric module
on Debian Woody GNU/Linux.

	Whenever I use IDLE, it just can't find the correct path. I
can see the Numeric module with MC (Midnight Commander) or another
file viewer. However, IDLE's Path Browser doesn't "register" it.

	Should I be doing some "make" here? 
	
	The funny thing is that dpkg installed python-numeric without
any problem.

	Any ideas?
	TIA
	Regs,

	H



From mlh at vier.idi.ntnu.no  Sat Apr 13 06:32:50 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 13 Apr 2002 10:32:50 +0000 (UTC)
Subject: ANN: Munch 0.1 -- Embed Python in text as markup
References: <3CB75C38.757A849D@alcyone.com>
Message-ID: <slrnabg2ai.g9m.mlh@vier.idi.ntnu.no>

In article <3CB75C38.757A849D at alcyone.com>, Erik Max Francis wrote:
[snip]

If you swap '$' for '@' this is quite similar to the Cheetah template
system. Have you looked at that?

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From tejarex at yahoo.com  Wed Apr 17 11:29:50 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 17 Apr 2002 15:29:50 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu> <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk> <mailman.1019031274.26548.python-list@python.org>
Message-ID: <Oxgv8.78475$K5.6922718@bin5.nnrp.aus1.giganews.com>

>(e.g. while debugging). Instead of just commenting out the code, one
>must also add a "pass" statement, and then try to remember to take it
 >out again when re-enabling the code.

If one *does* use 'pass' instead of a docstring as a placeholder, it
is *not* necessary to remove it (except for aesthetics).

def f():
  return 1
  pass

is perfectly legal, as is

def f():
  pass
  return 1

Terry J. Reedy





From weeks at vitus.scs.agilent.com  Tue Apr 23 12:37:25 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Tue, 23 Apr 2002 16:37:25 +0000 (UTC)
Subject: file_ref = open(string_ref)
Message-ID: <1019579845.565223@cswreg.cos.agilent.com>

Is this code

C1:	file_ref = open(string_ref)

better than this?:

C2:	file = open(string)

C1 is more accurate, given the view that data in Python are references to
objects.  According to this view, the left-hand variable names a reference
to a file object.

I can't believe that C1 is superior to C2 or that the millions of lines of
Python, Lisp, and Java code similar to C2 are in any way wrong.  So, in my
opinion, our view of data is wrong.  To align our view of data to our code,
we need to replace the view

V1:	Objects are regions of memory.

with:

V2:	Objects are addresses (of regions of memory).

If we adopt V2, then C2 (and all the code like it) is correct.  file is an
object, string is an object, and even open is an object.  If we prefer V1,
on the other hand, then we must consider C1 more correct.

The problem with V1 is not simply a mismatch between our concepts and our
Lisp, Java, and Python code.  The sad fact is that the majority of C code
that I've seen resembles C1.  This code is just as ugly and unnecessary in
C as it would be in Python.

So, every now and then, I raise the "Objects are addresses" flag.


Greg


PS: Note that I don't proclaim "Objects are references".  The term
"reference" is ambiguous and heavy with misleading connotation.

PS: In "Objects are addresses", "Objects" does not include the numerical
values of Java or the numerical or structure values of C.  If this is a
problem, then V1 has the same problem.

PS: What about C++.  Ucch, the worst.  C++ extravagently supports V1 by
allowing "initialization by reference", whereby a variable is initialized
to the same region of memory as another variable.  The result is
considerably more complicated than Python, Lisp, and Java.  And it doesn't
quite hang together, since there is no assignment by reference.


From loewis at informatik.hu-berlin.de  Mon Apr 29 03:14:19 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 29 Apr 2002 09:14:19 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it>
Message-ID: <j4adrn9dtg.fsf@informatik.hu-berlin.de>

Alex Martelli <aleax at aleax.it> writes:

> I think a finer-grained mechanism might be preferable, and handled by
> site.py / siteconfigure.py, perhaps, much like today are handles such
> issues as sys.path normalization, and Unicode default-encoding.  The
> former may be a more successful example than the latter, so maybe the
> need might be best approached via an extension to the parsing and
> handling of .PTH files.
> 
> Right now, .PTH files handle importing and path-setting needs of
> installed packages.  Having them handle such packages' builtins-needs
> as well seems a minor extension.

The current .PTH design allows for that kind of operation already:

# functional_compat.py
import __builtin__
exec "from functional import map,filter,reduce" in __builtin__.__dict__

# functional_compat.pth
import functional_compat

Alternatively, you could install a single .pth file enabling several
compatibility features:

# Numeric.pth
Numeric
Numeric/more_stuff
import functional_compat
import pow_compat

My original intention was that those import lines in .pth files can
install Unicode codecs in the codecs registry, but they, of course,
all for the execution of arbitrary stuff.

HTH,
Martin


From trentm at ActiveState.com  Fri Apr 26 17:26:32 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 26 Apr 2002 14:26:32 -0700
Subject: How do I format this
In-Reply-To: <8f8ffe67.0204261312.77787e94@posting.google.com>; from sjoshi@ingr.com on Fri, Apr 26, 2002 at 02:12:16PM -0700
References: <8f8ffe67.0204261312.77787e94@posting.google.com>
Message-ID: <20020426142632.A24954@ActiveState.com>

[Sunit Joshi wrote]
> Hello
> I have a format string which I'm formatting using a dict. The format
> string is
> """
> <td width="1\%">			
>  <input type=radio value="%(fullname)s****%(isdir)s" name=radioName>
>  <IMG height=17 hspace=2 src="ezmedia/window/filephotoicon.jpg"
> width=19>
>  </td>																				
>  <td>
> <div class="menucontents"><a href="%(fullname)s">%(name)s</a></div>
>   </td>																				
>  <td width="10\%">
> <div class="menucontents">%(ext)s</div>				
>  </td>																				
>  <td width="15\%">
> <div class="menucontents">%(size)s</div>			
>  </td>																				
>  <td width="22\%">						
>  <div class="menucontents">%(modified)</div>
>  </td>
> </tr>
> """

1. You want to use %% to escape a percent sign, rather than /%.
2. You missed at least one trailing 's' on a '%(foo)s' block

Try this:

"""
<td width="1%%">			
 <input type=radio value="%(fullname)s****%(isdir)s" name=radioName>
 <IMG height=17 hspace=2 src="ezmedia/window/filephotoicon.jpg"
width=19>
 </td>																				
 <td>
<div class="menucontents"><a href="%(fullname)s">%(name)s</a></div>
  </td>																				
 <td width="10%%">
<div class="menucontents">%(ext)s</div>				
 </td>																				
 <td width="15%%">
<div class="menucontents">%(size)s</div>			
 </td>																				
 <td width="22%%">						
 <div class="menucontents">%(modified)s</div>
 </td>
</tr>
"""

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From philh at comuno.freeserve.co.uk  Wed Apr  3 18:12:31 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Thu, 4 Apr 2002 00:12:31 +0100
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org>
Message-ID: <slrnaan32v.n0n.philh@comuno.freeserve.co.uk>

On 03 Apr 2002 10:59:13 -0500, Andrew Koenig <ark at research.att.com> wrote:
>
>Specifically: If I want to write a function that answers a yes/no
>question, I have lots of possible ways of spelling yes (1, 2, "yes",
>and so on) and lots of possible ways of spelling no (0, {}, None,
>and so on).  There isn't a single preferred way.

Yes there is, 1 or 0.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From martin at v.loewis.de  Mon Apr  1 18:27:38 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 01:27:38 +0200
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com> <a8apju$kcb$0@216.39.172.122>
Message-ID: <m3pu1jvw2d.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> But it does make me think, should _all_ strings be subtypes
> of a raw octet-string type according to their encoding? Then one
> could visualize automatic inter-encoding promotions analogous to
> numeric promotions, and if i/o sources and sinks have encoding
> designators, Gon?alo's f.write("Mar?o 2002" + march.Name()) should
> "just work" if the output encoding permits.

That assumes that the output encoding is known, or can be
determined. As-is, it can't - you don't know the encoding of f, and
you don't know the encoding of "Mar?o" (furthermore, the encoding of f
won't help, since you have to perform the addition before invoking
write).

In an all-Unicode approach, f would have been obtained from
codecs.open, and the string literal would have been a Unicode literal.

Regards,
Martin



From marklists at mceahern.com  Mon Apr 15 12:08:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 15 Apr 2002 11:08:22 -0500
Subject: metaclass
In-Reply-To: <r88fbu8f86i6hmave4g2v9dnedmm84g14k@4ax.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCECGKLAA.marklists@mceahern.com>

[Greg Chapman]
> I got a bit boggled looking at the Trace metaclass.  However, if
> I understand it correctly, it may not be necessary in 2.2 to use
> a metaclass to trace all method calls.
> [...]
> class C(Traced):
>     def __init__(self, x=0): self.x = x
>     def m1(self, x): self.x = x
>     def m2(self, y): return self.x + y
>     __trace_output__ = sys.stdout
>

Greg, thanks for the reply.  This helps me be a little more explicit about
why I'd prefer a metaclass-based approach:

I want to be able to turn tracing on all at once for all classes within a
particular interpreter session, without having to wire them beforehand (say
by subclassing from Traced).  Of course, theoretically another approach
would be some sort of aspect-oriented solution, but there doesn't seem to be
any convergence on a canonical approach to that in Python.

Cheers,

// mark





From twisted at itamarst.org  Tue Apr  2 19:05:22 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Wed, 03 Apr 2002 02:05:22 +0200
Subject: ANN: Twisted 0.16.0
References: <Pine.LNX.4.44.0204021726250.30418-100000@mercury.sabren.com>
Message-ID: <3CAA4742.5070706@itamarst.org>

Michal Wallace wrote:

> But I *dont* understand how it has anything whatsoever to do
> with stopping piracy. There's just a huge mental gap.

It would help at this point to quote the intro the 5th edition of the 
Prinicipia Discordia (http://www.ology.org/principia/intro5.html):

"""
Perhaps best of all is simply uttering a mondo. That is like picking up 
the telephone when it rings and saying, "Wrong number, please!" However 
much you think about a mondo it makes no sense - even clamps and pliers 
cannot get hold of it. Yet at the same time, if it is a good mondo, the 
longer you think about it the more it seems light it ought to make sense 
- although you can never figure out why. Beyond that much, a truly great 
mondo sticks to your mind like hot pine pitch - gumming up your thought 
process for weeks on end.
"""

To read the rest of the PD, I recommend the version at 
http://www.principiadiscordia.com

> The arguments I've read so far just makes me think that
> Twisted is going closed source, and that I should therefore
> stop trying to learning it... That would be a big
> disappointment for me, because so far everything I've seen
> in twisted looks great.

Nope, Twisted is and always will be Open Source and/or Free Software, 
whichever makes you feel warm and fuzzy inside.

> Is twisted going commercial?

We will gladly take any money you send us - but you can still use it for 
free, of course - it's LGPL.







From mwh at python.net  Fri Apr 12 05:33:22 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 09:33:22 GMT
Subject: frameobject.c bug ? No call to PyFrame_LocalsToFast.
References: <mailman.1018591127.18785.python-list@python.org> <pan.2002.04.12.10.18.02.982610.14362@cantal.lip6.fr>
Message-ID: <lklmbtjj73.fsf@pc150.maths.bris.ac.uk>

Boris Boutillier <boris at cantal.lip6.fr> writes:

> This is not completely true, as we can generate some code with
> CO_NEWLOCALS removed, with new.code for example.

Yes, but you can also do things like create instructions along the
lines of "LOAD_CONST 30000" which are very likely to crash the
interpreter.

Do *you* know about my bytecodehacks?  (I seem to be asking this a lot
lately).

Cheers,
M.

-- 
  nonono,  while we're making wild  conjectures about the behavior
  of completely irrelevant tasks, we must not also make serious
  mistakes, or the data might suddenly become statistically valid.
                                        -- Erik Naggum, comp.lang.lisp


From logstx at bellatlantic.net  Sat Apr  6 15:04:01 2002
From: logstx at bellatlantic.net (logistix)
Date: Sat, 06 Apr 2002 20:04:01 GMT
Subject: beginners syntax question
References: <3caf47a5@news1.homechoice.co.uk>
Message-ID: <RwIr8.5707$Sa2.2505@nwrddc01.gnilink.net>

If you really need to do this with strings, and can't make a list like the
others suggested, use a dictionary to translate strings to objects.  Python
also provides two builtin functions global() and local() that return the
respective global and local namespace dictionaries.

>>> x,y = 1,2
>>> globals()
{'__builtins__': <module '__builtin__' (built-in)>,
 '__doc__': None,
 '__name__': '__main__',
 'pprint': <function pprint at 0x01508790>,
 'pywin': <module 'pywin' from
'C:\PYTHON22\lib\site-packages\Pythonwin\pywin\__init__.pyc'>,
 'x': 1,
 'y': 2}
--
-

"Tony" <tonyl at homechoice.co.uk> wrote in message
news:3caf47a5 at news1.homechoice.co.uk...
> Hi,
>
> I'm trying to call the update() method of several objects ie:
>
> object1.update()
> object2.update()
> etc
>
> I'd really like to do
>
> for x in range(7):
>   "object"+str(x).update
>
> but can't figure outthe syntax, is this possible?
>
> Thanks
> Tony
>
>
>




From tim.one at comcast.net  Sun Apr 28 13:34:09 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 28 Apr 2002 13:34:09 -0400
Subject: struct module - '<q' equivalent for old python?
In-Reply-To: <42079966076-BeMail@sugoi>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEKBPBAA.tim.one@comcast.net>

[jaf]
> In my application, I need to be able to read and write a 64-bit integer
> (in little-endian format) from/to a file.  For Python 2.2, the struct
> module has a nice '<q' datatype argument that will do just that, but it
> isn't implemented in older versions and throws an exception.
>
> Does anyone know of a good alternative method I could use as a fallback
> in this case?  Some of the platforms I need to support don't have
> Python2.2 readily available, so I'd like to be able support older
> versions of Python, if possible.

What does "good" mean?  If you can settle for correct, these will work fine:

def write_leq(f, i):
    for count in range(8):
        f.write(chr(i & 0xff))
        i = i >> 8

def read_leq(f):
    result = 0L
    for shift in range(0, 64, 8):
        result = result | (long(ord(f.read(1))) << shift)
    if result >> 63:
        result = result - (1L << 64)
    return result

Be sure f is opened in binary mode.  Those do "the obvious" things, viewing
a 64-bit signed int as 8 unsigned 8-bit fields.  If you need it to be
zippier, you can use the struct module to view a 64-bit signed int as 2
mixed-sign 32-bit fields, although it's harder to understand:

import struct

def write_leq(f, i):
    f.write(struct.pack('<Ii', i & 0xffffffffL, i >> 32))

def read_leq(f):
    lo, hi = struct.unpack('<Ii', f.read(8))
    return (long(hi) << 32) | lo





From db3l at fitlinxx.com  Fri Apr 26 22:39:07 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Apr 2002 22:39:07 -0400
Subject: String handling bug in Python
References: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com> <mailman.1019857431.12024.python-list@python.org>
Message-ID: <usn5hyies.fsf@ctwd0143.fitlinxx.com>

Cliff Wells <logiplexsoftware at earthlink.net> writes:

> It seems if you imposed the limitation that the z-string not contain
> the quote character used to delimit the string then you could simply
> use concatenation to get around that limitation:
> 
> z"'" z'"' z'\'

Of course, the same string concatenation can also get around the final
backslash raw string problem:

    r'some\file\path' '\\'

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tim.one at comcast.net  Sun Apr 14 00:10:16 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 00:10:16 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <Xns91EFCC39E5237JamesLLugojcom@216.148.53.84>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEABOOAA.tim.one@comcast.net>

[James Logajan]
> ...
> Another important criteria in commercial environments is what
> versions are supported commercially.  The "core competencies" of many
> businesses is in the specific applications they write, not compiler or
> interepter development.  And open source can't change that limitation.
> Business people believe in contractual obligations and monetary
> incentives as their insurance that support will be there when they
> need it.

Well, at least ActiveState will be happy to sell you all the Python support
you want, contracts and all.  I expect PythonLabs would too, if people were
willing to pay the full cost; as is, we can't even afford a single full-time
tester, or porter, or doc writer, or anything else (not even Guido is
full-time on Python).  Zope Corp's direct "monetary incentives" in
supporting part-time Python work are all in the red.

It's in the nature of open source development that volunteers scratch their
own itches, then share what they've done because it does them good too to
have their work widely adopted.  If you can ride that curve, you can get an
enormous amount of high-quality work for free.  If not, you'll have to pay
for it or go without.

A crushing amount of debate over "stability" has gone by on Python-Dev the
last week, and the one thing that strikes me over and over is that, with
just a few minor exceptions, nobody volunteers to *do* anything except tell
other people what to do.  What are you willing to do to make your desires a
reality?  If it doesn't involve contributing work, time or money, you just
want a free ride.  But nobody owes anyone a free ride, and open source can't
change that fact of life either.

open-source-helps-those-who-help-themselves-and-the-payback-is-
    phenomenal-for-those-playing-the-game-ly y'rs  - tim





From uwe at rocksport.de  Fri Apr 12 21:11:23 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 13 Apr 2002 01:11:23 GMT
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de> <m31ydlndtd.fsf@chvatal.cbbrowne.com>
Message-ID: <a980jr$9d07k$1@hades.rz.uni-sb.de>

Christopher Browne <cbbrowne at acm.org> wrote:
| Uwe Schmitt <uwe at rocksport.de> wrote:
|> I'd like to send "raw" data to a printer, how can I do this,
|> using Python ???

| It's terribly platform dependent.

oops. i forgot to tell you that I'm using windows...
and i really want to print "raw" data and not pdf et al...

greetings, uwe.

-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From peter at engcorp.com  Wed Apr 10 20:28:35 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:28:35 -0400
Subject: Easter Eggs
References: <mailman.1018446981.22409.python-list@python.org> <slrnab8kfs.n1u.cpbotha@crabtree.cpbotha.net>
Message-ID: <3CB4D8B3.20DC2E5C@engcorp.com>

"Charl P. Botha" wrote:
> 
> >> "import this", IIRC.
> >
> > It's in 2.1.2, 2.1.3, 2.2, and 2.2.1.
> 
> Are you sure?
> 
> [dutidad:/home/cpbotha]% python
> Python 2.1.2 (#1, Mar 16 2002, 00:56:55)
> [GCC 2.95.4 20011002 (Debian prerelease)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> import this
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: No module named this

Check your python library file where the other ones are stored.
It is just a this.py file like all the other .py files.  If you
don't have it, maybe you don't have a standard Python distribution...


From phr-n2002a at nightsong.com  Thu Apr  4 10:25:47 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 07:25:47 -0800
Subject: Java to Python translation
References: <mailman.1017931452.24548.python-list@python.org>
Message-ID: <7xk7rn32pw.fsf@ruckus.brouhaha.com>

Kemp Randy-W18971 <Randy.L.Kemp at motorola.com> writes:
> What the heck kind of application has 100 MB of Java code anyway? 
> 
> Boy, that is a good question.  Let's look at the size of some applications I have on my Windows 2000 machine:
> JBoss-2.4.4_Tomcat-4.0.1  -  36.1 MB -- Jboss EJB server and Tomcat servlet container
> Jonas244                  -  23.0 MB -- Jonas EJB server
> mysql                     -  26.3 MB -- Mysql Database server
> Resin-2.1.0               -  17.4 MB -- Resin servlet container    
> Total space               - 102.8 MB
> Now the sum of these products equals about 100 MB of Java + C code

I doubt if there's 100 MB of source code between those programs.


From mgilfix at eecs.tufts.edu  Mon Apr 22 23:19:35 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 22 Apr 2002 23:19:35 -0400
Subject: need encryption algorithm
In-Reply-To: <e2e9e807.0204221741.597acd60@posting.google.com>; from h.e.w.frater@cs.cf.ac.uk on Mon, Apr 22, 2002 at 06:41:08PM -0700
References: <e2e9e807.0204221741.597acd60@posting.google.com>
Message-ID: <20020422231935.B21282@eecs.tufts.edu>

On Mon, Apr 22 @ 18:41, Hugh wrote:
> Hi all,
>   I need a simple encryption algorithm that will not give me
> characters that I can't store in my oracle table. At the moment I'm
> using the rotor encryption module because I like it's simple interface
> but It's generating characters that oracle can't store which is a damn
> pain. What other modules are available that are easy to use and will
> do what I wan't. Remember that security is not paramount here, I'm
> only including it for functionality.

  Anything can be transformed into base64 for storage. But there's no
point in using a weak cipher as its akin to not really using crypto at
all. You should probably ask yourself: 1) why you're encrypting in the
first place? 2) How important security of the data is 3) Who is your
intended audience.

  If you're using a password to store information, then you should
select a symmetric key encryption algorithm. I suggest searching
google for the Mcrypto module or here's an implementation of mine of
blowfish in python: http://bofh.concordia.ca/blowfish.py.

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From jimd at vega.starshine.org  Sun Apr  7 16:14:42 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 7 Apr 2002 20:14:42 GMT
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com>
Message-ID: <a8q9bi$2vka$1@news.idiom.com>

In article <d9ea2696.0204061049.551bd53b at posting.google.com>, distrex wrote:

>I understand most of python. But I still don't get modules. I know
>they are just basicaly a function but i don't know how to use them
>effectively and what each one does. Everything good in python you will
>at some point need to use a module right. If anyone can help me with
>this i'd be greatful. Info or websites would be nice.

 Modules are things that you import into your namespace.  The
 process of import them (or import *from* them) executes any 
 code in them.  Usually (almost always) a module will only contain
 class and function definitions, perhaps some "members" (variables)
 and an "if __name__ == '__main__':" suite.

 You can think of a module as a library of classes, functions and
 variables.  As such, it would be inappropriate to create modules 
 that executed "visible" statements (such as printing stuff) or made 
 other modifications to the importer's state.  Basically modules 
 should normally just execute whatever is necessary to define their
 namespace --- which usually requires NO "execution" (in the traditional
 sense) at all.  (The confusion is that python is a dynamic language,
 so things like:
 
 	def foo(): pass
	# and
	class Bar: pass

 ... are actually executable statements in Python, where they are 
 "definitions" or "declarations" in other languages.

 The most important thing to realize about Python modules is that they
 are singletons.  They are only executed upon initial import for any given 
 interpreter session.   Additional imports are not an error, and won't 
 raise an exception, but they will not re-execute the module's code.

 That can be confusing because subsequent import statements can 
 change change namespaces around.  I can do something like:

 	import bisect as bifoo
	from bisect import insort
	import bisect 

 ... and I've only execute the contents of bisect once, but I've
 made subsequent changes to my namespace (making the bisect.insort()
 available directly available without any qualifiers (in my globals()
 dictionary). and making it accessible via the bifoo and bisect
 namespaces (which is silly, and used just for an example).

 You might be aware that there are three sorts of modules in
 Python.  There are binaries (compiled code, usually from C sources)
 which appear as .so (shared object) files under Linux and UNIX, and
 .DLL files under MS Windows (don't know for Macs).  There are Python
 files (simple .py files in python) and there are packages which 
 are directories that contain __init__.py files with assignments to
 __all__ lists.  You don't have to know or care about these differences
 among these when using the modules.  You just use import the same way 
 regardless of the underlying implementation details.

 While *writing* your own modules you can start a module as a .py
 and later refactor it into a package or rewrite it in C and compile
 as your needs require.



From hgs at dmu.ac.uk  Mon Apr 22 12:21:36 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Mon, 22 Apr 2002 17:21:36 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4y9fjsk12.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204221715340.913-100000@neelix>

On 19 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > 20 tests were skipped but that would be less if I could enable that
> > regrtest.py thing you told me about.  Whereabouts is the correct place
> > to add the options to enable those tests?
>
> Neither, nor: invoke regrtest with appropriate options.

Oh, Ok, so it is done "outside" of the normal make, make test, make
install process....

The final upshot is:

largefile crashes  -- expected as we have little space.
curses crashes -- Not really surprising, given I'm doing this from a PC,
        and using an xterm tool from there.
nis gets skipped -- again, not staggering if it assumes we are not using
        nisplus, because we are using nisplus.

I therefore think I can live with these limitations, and go ahead and
install.
Thank you to everyone, particularly Martin, for their help, patience and
encouragement.
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you,
        Hugh






From mwh at python.net  Tue Apr  9 12:28:12 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 9 Apr 2002 16:28:12 GMT
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8v36u$j3p$1@news.nsk.su>
Message-ID: <uofgsal3x.fsf@python.net>

"Andrey Panchenko" <andrey at app.omsk.su> writes:

> I think it's rather an ambiguous statement. I can think that after
> 
> a = 1
> b = a
> a = 3
> 
> b must hold 3. But it's not true. 

True.

> IMHO it's better to say that the python names are like pointers only
> for mutable objects.

No, this implies a distinction that doesn't exist, IMHO.

It's best not to mention pointers at all, people who know C just get
the wrong idea.

Cheers,
M.

-- 
  The bottom tier is what a certain class of wanker would call
  "business objects" ...                      -- Greg Ward, 9 Dec 1999


From sholden at holdenweb.com  Sat Apr 27 07:45:30 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 27 Apr 2002 07:45:30 -0400
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
References: <mailman.1019830848.8415.python-list@python.org>
Message-ID: <Ydwy8.52326$z55.42220@atlpnn01.usenetserver.com>

"Arthur Siegel" <ajs at ix.netcom.com> wrote in message
news:mailman.1019830848.8415.python-list at python.org...
> Steve had written -
>
> >Many c.l.py readers are *not* "professional programmers", but
> >chemists, mathematicians, biophysicists, and a host of others
> >for whom programming is a means to an end. Many others are,
> >like me, professional programmers who attempt to
> >increase Python's popularity by assisting those less familiar with
> >programming concepts.
>
> I think this is honestly (and revealingly) worded, represents
> increasing the most consistent thread of your posts, and
> is its own kind of problem.
>
> The attempt to "increase Python's popularity by" becomes too
> transparent. Better a side effect, than a motive. Or play it
> smart enough that it seems as such.
>
> Reprimanding a long-time Python contributor for the wording
> of a post that you fear will spread FUD was strange enough.
> But there a consistency.
>
> Steve - I bought and enjoy your book and do consider it a
> service to the non-professional community.
>
> But the cheerleading and defensiveness on cl py has
> in IMO gone a bit over the top, and is a service to
> no one.
>
> Accept, this is Usenet.  I can say dugapoopoo if I want.
>
And I can say dugapoopoo too. I'll cut out the cheerleading and
defensiveness immediately :-)

regards
 Steve






From ajs at ix.netcom.com  Sun Apr  7 14:55:23 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 7 Apr 2002 14:55:23 -0400
Subject: PEP 285: Adding a bool type
Message-ID: <000001c1de6d$55a89e40$0334fea9@carol>

Aahz -

> people like Arthur and Paul Rubin...

Irish?

Paul is one of the few people on the list (other than
some of the Powers that Be) with whom I have had a
personal exchange of some nastiness.

He had, I believe, in the course of these discussions
volunteered to do some hands on development work
on some point where he and Guido seemed to agree some
attention was necessary.

How this poor guy got himself lumped in with the 
likes of *me*, is hard to understand.

Art








From shalehperry at attbi.com  Mon Apr 29 15:34:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 29 Apr 2002 12:34:12 -0700 (PDT)
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <3CCD9CFD.E1869228@cascade-sys.com>
Message-ID: <XFMail.20020429123412.shalehperry@attbi.com>

>> >From the comments, I've learned:
>>
>> 5.  oct() is another candidate based on the rarity of usage
> 
> Candidate perhaps but some people still like Octal.
> 
> This still is a gratuitious change and thus should not be implemented.
> 
> If it ain't really broke, you shouldn't fix it.
> 

here here!

>> 7. the language needs some means of deprecating so that clutter
>>     does accumulate.  that means needs to be well documented,
>>     have a long, slow phase-in, and have a mechasism for
>>     restoring old behavior if it is ever needed.
> 
> Making it easier to phase changes in and out makes things easier only
> for the most active Python users.  Changes themselves are hard on the
> user base generally.  Furthermore, having a language where things like
> base functionality routinely changes significantly from release to
> release seems a Bad Thing.
> 
> Picture the poor newbie sitting down with his O'Reilly book, having just
> installed 2.3 or 2.4 and can't figure out why some of the examples no
> longer work.  O'Reilly has how many thousands of these books in the
> pipeline?  I'm sure they're not anxious to issue a revised edition.
> Steve Holden barely had time to get get the 2.1 appendix in his book.
> This is not helping ANYBODY.
> 
> I frankly think it's ALARMING that people are becomming comfortable with
> the notion that you can't interpret the meaning of programs without (a)
> being up to date on how things change from release to release and (b)
> paying close attention to the declarations or (super ugh) command
> switches that might override defaults.  This is a much greater burden
> than simply learning any particular version of the language.  It's a
> moving target.  Of course users don't have to upgrade to each new
> release and generally only use one version at a time.  But in trying to
> use other people's code you're likely to encounter anything from old to
> new, so you have the problem whether you want it or not.  Unless you
> hide in your hole and only use code you write yourself.
> 

The problem is compounded when you are not in control of the python on the
system.  Imaging CGIs broken when the person's ISP decides to upgrade their
machines to a new OS rev which has a new python.





From ivanlan at pauahtun.org  Fri Apr 26 23:21:13 2002
From: ivanlan at pauahtun.org (Ivan Van Laningham)
Date: Fri, 26 Apr 2002 21:21:13 -0600
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com>
Message-ID: <3CCA1929.674E5C43@pauahtun.org>

Hi All--

Christian Tismer wrote:
> 
> Dear friends,
> 
> --------------------------------------
> I would like to start the SPA:
> 
>     Starship Python Activity
> 
> as a non-profit organisation.
> --------------------------------------
> 
> Here the proposed layout:
> SPA membership costs USD 25 by the year.
> This price includes Starship membership
> and a continous subscription to PyZine.
> 
> The generated extra money will be used
> for sponsoring of projects, programming
> contests, a gift to the Starship webmaster,
> there are many opportunities.
> 
> I think this potential should not be
> wasted any longer.
> 

It's a great idea, but *please* find some other way to collect money
than PayPal.  I'd rather send a check than deal with them.  Besides,
haven't they been put on hold or something?  Acting like an interstate
bank, wasn't it?

Metta,
Ivan
----------------------------------------------
Ivan Van Laningham
God N Locomotive Works
http://www.pauahtun.org/
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours





From jrup at mail.com  Mon Apr 15 02:49:35 2002
From: jrup at mail.com (jrup)
Date: Mon, 15 Apr 2002 06:49:35 GMT
Subject: Specifying a time from previous day
References: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net> <7876a8ea.0204141521.1f6570ed@posting.google.com>
Message-ID: <3Kuu8.28739$%s3.10299386@typhoon.ne.ipsvc.net>

David,

Thanks very much for the ideas and examples! I'll study this closely.

Jim R

"David Rushby" <woodsplitter at rocketmail.com> wrote in message
news:7876a8ea.0204141521.1f6570ed at posting.google.com...
> "jrup" <jrup at mail.com> wrote:
> > I need to search some log files, but only if they were created after
4:00pm
> > the previous day.
> >
> > I've looked at the python time module and its documentation, and have
> > skimmed mxDateTime documentation. I'm using activestate python 2.1.1 and
> > would like to stay with it's native library since it's the one
sanctioned by
> > our senior nt person (also not sure how it would interact with the
> > activestate install, or I'd try it).
> >
> > Can someone please point me in the right direction? Is this doable using
the
> > time module? I haven't had much luck figuring it out, yet.
>
>   It's certainly "doable" using the time module (see below), but
> since:
>
> a) the time module is rather low-level
> b) mxDateTime is freely available, and high-level
> c) you're a self-described newbie
>
>   , why confuse yourself by using the time module?
>
>   I see no reason why installing the eGenix.com mx Base Package would
> harm an installation of Activestate Python.
>
>   An mxDateTime-based example appears below (with equivalent code
> based on the time module also present, but commented out):
> ----------------------------------------------------------------
> import glob, os
> import mx.DateTime as dt
> # import time
>
> beginningOfToday = dt.today()
> eightHours = dt.DateTimeDelta(0, 8)
>
> # beginningOfToday = time.mktime(list(time.localtime()[:3])
> #    + [-1] + [0] * 5)
> # eightHours = 60*60 * 8
>
> minTime = beginningOfToday - eightHours
> minTimeAsTicks = minTime.ticks()
> print 'Will process all log files modified at or later than', minTime
>
> # minTimeAsTicks = beginningOfToday - eightHours
> # print 'Will process all log files modified at or later than',
> # print time.asctime(time.localtime(minTimeAsTicks))
>
> # logFilenamePattern could be input as a command-line
> # argument using sys.argv
> logFilenamePattern = r'c:\temp\*.txt'
> logFilenames = glob.glob(logFilenamePattern)
>
> # Build a list of relevant log filenames, including only the names of
> # files that have been modified at or later than minTimeAsTicks.
> relevantLogFilenames = [
>     logFilename
>     for logFilename in logFilenames
>     if os.stat(logFilename)[8] >= minTimeAsTicks
>   ]
>
> for logFilename in relevantLogFilenames:
>     logFile = open(logFilename)
>
>     print 'processing %s...' % logFilename
>     # Do some processing here.
>
>     logFile.close()




From BPettersen at NAREX.com  Fri Apr 19 17:52:36 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 19 Apr 2002 15:52:36 -0600
Subject: Substitution with Hash Values
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158DF7@admin56.narex.com>

> From: Jeff Jones [mailto:jeff at jonesj.com] 
> 
> I am a long time perl user who is trying to get a grip on 
> python.  I really like what I have seen so far, but I have 
> hit a little snag regarding substitution (like perl).
> 
> I use html templates for all of my CGI programming, and 
> everything that needs to be replaced is contained within 
> double curly braces ( {{value}} ).  In perl I use a hash to 
> keep all of the variables that are going to replace the 
> placeholders in the template, and replace them the following way:
> 
> 	....
> 	s/{{(.*?)}}/$hash{$1}/g;
> 	....
> 
> That substiution is accomplished as I read in the template file.
> 
> Can anyone tell me how I might do this in python?  I have 
> tried many different ways, but none seem to work.

# create a dictionary with your substitutions
subst = { 'value1' : 123, 'value2' : 'hello world' }

# create the text with appropriate %(dict-key)s placeholders
txt = """
  <body>
     this is %(value1)d and this is %(value2)s
  </body>
"""

# perform the substitutions
print txt % subst

Hth,
-- bjorn




From mck at mpinet.net  Tue Apr 16 14:51:50 2002
From: mck at mpinet.net (Matt)
Date: Tue, 16 Apr 2002 18:51:50 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com> <3CBC5B6D.70427378@noaa.gov>
Message-ID: <gpsobu0952s65n93clkqadtq27cj1j4qul@4ax.com>

On Tue, 16 Apr 2002 10:12:14 -0700, Chris Barker <Chris.Barker at noaa.gov> wrote:

>Imagine if Microsoft decided to drop VBasic
>and start selling MSPython!

Oh boy, that is a whole different can of worms (I can see the thread now...) ;)


From offer at sgi.com  Tue Apr 23 10:28:26 2002
From: offer at sgi.com (offer at sgi.com)
Date: 23 Apr 2002 14:28:26 GMT
Subject: Python, PyQt make install problems on IRIX
References: <f0fd5987.0204220106.7673e345@posting.google.com> <3cc45c8d$0$307$e4fe514c@dreader1.news.xs4all.nl> <f0fd5987.0204230057.25fdc747@posting.google.com>
Message-ID: <aa3r2a$tiabi$1@fido.engr.sgi.com>

$ from gtaylor at lowebroadway.com -#196241 | sed "1,$s/^/> /"
>
>
>Irix certainly is'nt all that common, in a way I think that kind of
>helps people make installers, because there are not many hardware
>configurations to support or variations of the OS, like Linux, there
>are loads of variations on the same thing.

Hey, Irix is pretty commone here :-)

Anyway its my intention to ship PyQt in the next release of Irix Freeware
(Aug), so I'll keep an eye out for this problem (I have no idea what's
causing it).

>
>Thanks
>
>Garry


richard.


From aleax at aleax.it  Sun Apr 28 08:13:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 12:13:44 GMT
Subject: Scripting lanuages for laboratory tests
References: <aagkfm$6kf$1@news5.svr.pol.co.uk>
Message-ID: <YHRy8.72400$vF6.2161044@news2.tin.it>

Etta Makwetta wrote:


> Cons: GUI slightly more complicated than Tcl, More to learn, Not support

Python gives you a wide choice of GUI toolkits.  I doubt the "slightly more 
complicated" idea applies to all.  PyUi, http://pyui.sourceforge.net/ , 
seems to be _extremely_ simple, as also anygui aims to be; at the other
extreme, PyQt is truly rich and powerful, but not quite as simple as, e.g.,
Tkinter (which I suspect is the one you deem "slightly more complicated"
than Tk -- a disputable claim, IMHO, but that's another issue).

> calling an arbitrary dll
> 
> Neither Tcl nor Python supports calling an arbitrary dll.  Tcl supports

http://www.nightmare.com/~rushing/dynwin/calldll-2001-05-20.zip


Alex



From neal at metaslash.com  Fri Apr 19 13:44:46 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 19 Apr 2002 13:44:46 -0400
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
		<NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>
		<mailman.1019068175.11920.python-list@python.org>
		<99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com> <mailman.1019233076.19513.python-list@python.org>
Message-ID: <3CC0578E.822EB961@metaslash.com>

Cliff Wells wrote:
> 
> On Fri, 19 Apr 2002 10:37:23 +0100
> Dale Strickland-Clark wrote:
> 
> > >
> > >Or it might be preferable to use a dictionary:
> >
> > But several orders of magnitude more processing. Certainly not for
> > tight, efficient loops.
> >
> 
> It does take considerably longer, but longer is still relative:

Depends, here's 2 more examples, in addition to Cliff's:

import time
gender = 'f'
GENDER_DICT = {'m': 'male', 'f': 'female'}

t = time.time()
for i in range(100000):
    verboseGender = (gender == 'm') and 'male' or 'female'
print time.time() - t

t = time.time()
for i in range(100000):
    verboseGender = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
print time.time() - t

t = time.time()
for i in range(100000):
    verboseGender = GENDER_DICT.get(gender, 'unknown')
print time.time() - t

t = time.time()
for i in range(100000):
    try: verboseGender = GENDER_DICT[gender]
    except KeyError: verboseGender = 'unknown'
print time.time() - t

0.248443961143
0.696393966675
0.322777986526
0.239256024361

Neal


From Reply.To.The.Group at LeaveMeAlone.invalid  Sat Apr 27 15:25:21 2002
From: Reply.To.The.Group at LeaveMeAlone.invalid (Ken Peek)
Date: Sat, 27 Apr 2002 12:25:21 -0700
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk> <just-9E8624.12281026042002@news1.xs4all.nl> <ucjicvlrf7djd3@news.supernews.com> <q5Z7JWAa$ty8EwuL@jessikat.fsnet.co.uk>
Message-ID: <3ccafb6d@news.mhogaming.com>

So, I went to the doctor, and I told the doctor: "Hey doctor, it hurts when I do
THIS!"-- and the doctor said: "Well, don't DO that then!"





From sjmachin at lexicon.net  Wed Apr 24 20:49:49 2002
From: sjmachin at lexicon.net (John Machin)
Date: 24 Apr 2002 17:49:49 -0700
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <c76ff6fc.0204241649.2b19b4f1@posting.google.com>

[Rad (zaka07 at hotmail.com)]

I am preparing myself to work on extracting data from 4 text files
(fixed width format) which combined size is about 80GB.  Considering
deadlines, costs, and my limited programming knowledge I thought using
Python/Windows for the job would be the best option for me.  However,
I am worried about the speed in which Python (me and my hardware) will
be able to deal with these massive data sets but I am hoping that this
is still a quicker route then learning C.
I still haven't received above mentioned files so I can't test the
time needed to (for example) read a 15GB "file1", filter by few
variables, and write a resulting subset as a "sub_file1".  Things
would afterwards get more complicated cause I will have to pullout
ID's from "sub_file1", remove duplicate ID's create
"no_dup_sub_file1", match those to ID's in remaining 3 main files and
pullout data linked with those ID's.

JM-> [JM == John Machin] Here is the short answer:
Somebody who can afford to accumulate 80GB of data should be able to
afford an IT professional to work on it. IMHO (80GB of unknown data) +
Windows + SQL + mmap + grep + awk + newbie - Python produces a high
potential-disaster rating.

However in case you are stuck with it, here are some comments on the
details:

[Aahz (aahz at pythoncraft.com)]

Python *can* handle this kind of task, but you'll be much better off
if
you interface Python to a database.  The problem is that even 15GB
(not
even talking about 80GB) is simply too big to fit in RAM, so you'll
need
a way to process partial sets that do fit in RAM.

JM-> 15GB of *fixed width* records (presumably lots of trailing spaces
and leading zeroes) could shrink considerably when only the relevant
info is considered.

JM-> *** IMPORTANT ***
(1) How many rows/records in each file??? How much relevant info in
each record???
(2) This exercise is "doable" in a reasonable run-time IF a Python
dictionary of the keys of the records in "no_dup_sub_file1" can be
held in real (non-swapped) memory.
(3) What is the definition of "duplicate IDs"??? Does it involve
"fuzzy matching" as in deduplication of mailing lists?

[Aahz]
I strongly suggest that you push *VERY* *HARD* to get some small
sample
files (100MB to 1GB range).  Get those samples ASAP.

JM-> Even 10 MB might do if you get every nth record. Don't believe
any specification you are given. To quote my son on arrival back from
a database integration job that took 10 days instead of 3: "Users are
bastards. They lie to you".

[Aahz]
Finally, make sure that you have at least four times the disk space as
the total size of all the files (that's probably a conservative guess,
but you'll definitely need at least 2.5 times).

JM-> Maybe not so much if you do things sequentially. However it's
still a whole lot more disk space than is usually associated with a
non-server Windows box. Interesting question: on what medium is the
80GB arriving?

[holger krekel (pyth at devel.trillke.net)]

I disagree. Don't be so humble :-)
Using a database requires

- setting up/configuring an appropriate database
  for the task (may not be easy)

- getting the info from the files into a database
  (requires reading the file anyway!)

- reading from the database with partial result sets
  and convert again to a file.

i think this is errorprone and quite complex.

JM-> If Rad is new to programming he might just be able to do this in
Python if the job is a one-off and is as simple as it seems to be.
However Murphy's Law will operate. Learning SQL and configuring a
database is certainly not newbie territory.

[holger]
why not a more pythonic way like this:

- use module mmap [snip]

JM-> I don't see what this gives you, except complication. Remember
Rad is
a newbie.

[Holger]
it is important to know what your time constraints are.
10 minutes, an hour, a night?

JM-> I'd multiply these numbers by 10, at least, for a one-off or
first-time effort.

[David Bolen (db3l at fitlinxx.com) gives *lots* of good advice]

JM-> However installing Cygwin would be yet another bump in the
learning curve for Rad the newbie.

[Neal Norwitz (neal at metaslash.com) gives upshifts a 650MB file]

JM-> I expect Rad will have to do some more complicated
transformations than upshifting. Even superficially simple things like
converting strings to internal integer or date format so that they can
be compared will add considerably to the run time.

[Neal]
for x in ifile:
        ofile.write(x.upper())
JM->
write_out = ofile.write
for x in ifile:
        write_out(x.upper())

[Fernando P?rez (fperez528 at yahoo.com)]

May I suggest that you also pick up some basic (4 hrs worth) grep/awk
knowledge?

JM-> Yes, these are basic programmer tools. You don't need to install
Cygwin. Just get the tools that you need from the GNUWin32 site.

[Bo (bosahvremove at netscape.net)]

If this is a recurring requirement at work I would consider hiring a
contractor to create an elegant solution with defined steps for you to
operate it.

JM-> Generally agree, but I'd s/elegant/robust/

Some further comments:

Does the data include dates? Do you need to select data by date
comparison?
If you are lucky, dates will be in YYYYMMDD or YYYY-MM-DD format and
you can use a string comparison. Unlucky: MM/DD/YYYY format -- you
will need to convert to a "comparable" form. Do you need to do any
date calculations? If so, grab a copy of the mxDateTime package; it's
implemented in C. Python-based date modules will be far too slow,
protestations from their authors notwithstanding :-)

Write yourself an every-nth-record selector in Python. You will need
it to produce reduced datasets for eyeballing and testing. ***Reality
check***: If you find writing such a thing within 30 minutes to be a
struggle, activate Plan B [hire a professional] now.

HTH,
John


From gerhard at bigfoot.de  Fri Apr  5 05:40:22 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 5 Apr 2002 10:40:22 GMT
Subject: Fail to use jythonc
References: <a8jv8m$s7k$1@pegasus.csx.cam.ac.uk>
Message-ID: <slrnaar174.20s.gerhard@lilith.my-fqdn.de>

Xiao-Qin Xia wrote in comp.lang.python:
> I installed a jython recently, but cannot use jythonc.
> for example, a very simple program is:
> [...]
> [xqxia at mwise1 jython-2.1]$ ./jythonc mytest.py
> processing mytest
> 
> Required packages:
> 
> Creating adapters:
> 
> Creating .java files:
>   mytest module
> 
> Compiling .java to .class...
> Compiling with args: ['/usr/local/j2re1.3.0/bin/javac', '-classpath', 
> '/home/xqxia/jython-2.1/jython.jar::./jpywork::/home/xqxia/jython-2.1/Tools/jythonc:/home/xqxia/jython-2.1/.:/home/xqxia/jython-2.1/Lib', 
> './jpywork/mytest.java']
> 1  java.io.IOException: /usr/local/j2re1.3.0/bin/javac: not found
> 
> Consider using the -C/--compiler command line switch, or setting
> the property python.jythonc.compiler in the registry.
> ERROR DURING JAVA COMPILATION... EXITING
> [xqxia at mwise1 jython-2.1]$
> """

Looks like you have the JRE installed, but not them JDK.

JRE = Java Runtime Environment -- only for running Java programs
JDK = Java Development Kit -- for running and compiling Java programs

You need a JDK for using jythonc. Ok, you *could* also use an
alternative compiler like Jikes, that's what the last message about
"-C/--compiler" is trying to tell you :)

Gerhard


From mcherm at destiny.com  Mon Apr  8 08:33:15 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 08 Apr 2002 08:33:15 -0400
Subject: need something like __init__ in classes __classinit__
Message-ID: <3CB18E0B.3040508@destiny.com>

Kevin Jacobs writes:
>   o A metaclass that lazily instantiates slots so that you can write:
       [...]
>   o A metaclass that adds constraint checking to attributes and slots.  This
       [...]
>   o A tuple building metaclass
       [...]
>   o The above ClassInit metaclass.
       [...]

COOL!!!  These seem FUN. Where did you say you had these posted? Sounds 
like the beginning of a new section of the cookbook or something... nice 
examples!

-- Michael Chermside




From jeff at ccvcorp.com  Tue Apr 23 14:20:48 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 23 Apr 2002 11:20:48 -0700
Subject: Python, search engine and NLP...
References: <409a56e2.0204230104.16d60f52@posting.google.com>
Message-ID: <MPG.172f45cbadc8d0fb9896c5@news.nwlink.com>

In article <409a56e2.0204230104.16d60f52 at posting.google.com>, 
shagshag13 at yahoo.fr says...
> Hello, 
> 
> I'm interested in Pyhon but not sure it can fit my needs :

Well, I don't know enough to address your specific questions, but 
I *do* have a couple of thoughts...

> - As by now i don't know Python, do you think that it will not be a
> waste of time even if I had to learn it ???

Python is very easy to learn, an experienced programmer can be 
productive in it within days.  Don't let the fact that you don't 
know it yet keep you away.  :)

> - Do you know about some Python's natural language processing module
> ???
> 
> - No, as there are many Java NLP library, I should definitively do it
> with Java...

I don't know either way about any Python NLP modules, but if you 
have access to Java libraries, then you might want to use Jython 
(www.jython.org), which gives you Python's clean syntax and ease-
of-use on top of a Java virtual machine.  This allows you to 
access all of your Java libraries from Python code.  I haven't 
used it myself, but I for some people this is wonderful.  

-- 

Jeff Shannon
Technician/Programmer
Credit International


From pan-newsreader at thomas-guettler.de  Fri Apr 12 15:10:58 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Fri, 12 Apr 2002 21:10:58 +0200
Subject: Queue (Threads)
Message-ID: <pan.2002.04.12.21.10.57.987428.703@thomas-guettler.de>

I programmed a Producer Consumer problem. But unfortunately
I get "cannot join thread before it is started".

I found a different approach which uses two Queues. 
http://www.bagley.org/~doug/shootout/bench/prodcons/alt/prodcons.python2.python

But I prefere a solution with just one queue:
 
import sys
from Queue import Queue
from threading import Thread

queue=Queue(4) #Only for items are allowed in the queue

def consume():
	print "Consuming makes fun"

class Producer(Thread):
	def run(self):
		print "Producing is hard"
		self.result="stone"

class Produce100Times(Thread):
	def run(self):
		for i in range(100):
			pt=Producer()
			queue.put(pt)
			pt.start()
		queue.put(None)

if (__name__=="__main__"):
	p100t=Produce100Times()
	p100t.start()
	while 1:
		produced_thread=queue.get()
		if produced_thread==None:
			#All items produced. Time to go home
			break
		print "Got Thread"
		produced_thread.join()
		print "Result:", produced_thread.result
		

I tried to find examples for the Queue class with google, but
no look. Has someone links to exmamples or hints?

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From sholden at holdenweb.com  Tue Apr 30 08:04:01 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 08:04:01 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org>
Message-ID: <kNvz8.12234$xy4.11902@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1020108174.25482.python-list at python.org...
>
> These changes sound like a BIG improvement over the original draft PEP.
>
> I pretty much agree with the changes, with a few exceptions noted below.
>
[ ... ]
> As it is, we can no longer look at a fragment of code containing a
> division operator and know what it means without checking version and
> import declarations (which may appear ANYWHERE in the source).  I
> frankly find this to be BIZARRE.
>
If you're talking about "from __future__ ...", these can only appear before
any other source lines, and so (bugs notwithstanding) they aren't as
difficult to detect as you believe. Remember that some imports from
__future__ modify how the compiler acts, so you can't just switch them on
and off at random.

> The more we add to this confusion the worse service we do to the user
> community at large.  Easily phasing features in and out is clearly
> something that ONLY benefits the most sophisticated users, leaving
> beginners and non-programmers in the dust.  Some of the seeming experts
> on this list cannot keep up, so how does the non-programmer have a
> chance?
>
Honestly, there's has been a LOT of talk about this on python-dev, and the
developers are very aware of the pain such changes cause.

> I think the world would be a better place if the language experienced a
> lot FEWER non-backwards compatible changes.   Absolutely none is the
> norm in most languages and would be an ideal here.  Also users would be
> better off if we had a lot fewer releases altogether.  One major release
> per year seems a healthy goal, with debug-only releases in between.
> Even then, changes should focus on backwards compatible enhancements or
> new features that don't break old colde rather than trying to fine-tune
> existing features, even some of the few arguably regrettable ones.
> Simply write them off to Python's rich heritage.
>
It would also be a nicer place (for me :-) if everyone just did what I said.
Hasn't happened yet, though.

> So, I think one could argue just the opposite of #7 -- that we should
> look into policies that make it HARDER to deprecate and phase out old
> functionality.  I don't see that it ever hurts to leave it in but
> there's ALWAYS some pain to take it out.  None of the old vestiges are
> so awful that they require change.  So don't do it.
>
But don't underestimate the cruft that has to be carried forward into new
implementations either. I appreciate what you're trying to say, but under
the hood there have been some radical changes in the last two releases. It's
amazing that such a small amount of code was broken as a result.

guido-feels-your-pain-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From boud at valdyas.org  Tue Apr 16 11:10:45 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 16 Apr 2002 14:10:45 -0100
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <KcUu8.22993$Ji7.1840@atlpnn01.usenetserver.com>
Message-ID: <3cbc16f7$0$71662$e4fe514c@dreader4.news.xs4all.nl>

Steve Holden wrote:

> "sameer" <sameer_ at email.com> wrote ...
>> I am trying to run multiple threads in a wxFrame class to instantiate
>> different panels within it on Win 2k,  and well, it's a little hairy.
>> Anyone got a good web source that discusses python threading in depth?
>>
>> More so, my question is, how can I get the return value of a function
>> that I run on a seperate thread?  I don't see anything for this in the
>> threading module.
> 
> The canoncial reference was written by Aahz, whe keeps threatening to
> update it. See
> 
>     http://starship.python.net/crew/aahz/OSCON2001/index.html
> 
> The Queue module might help you with passing values between threads, as
> it's specifically designed to be thread-safe.
> 

That's funny -- only yesterday this cropped up too, and Alex Martelli sent
me a link to: http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965,
where I quickly added a PyQt version of the example code... It uses Queue. 
Actually instantiating GUI objects from different threads might be hard. I 
would place instantiation requests on the Queue, and have the GUI thread
execute them.

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


From achim.domma at syynx.de  Sat Apr 20 04:39:19 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Sat, 20 Apr 2002 10:39:19 +0200
Subject: Integrating C/C++ in to python/Zope
References: <a9q49n$tq7$1@news.datasync.com>
Message-ID: <a9r91p$1sn$03$1@news.t-online.com>

http://www.boost.org/libs/python/doc/index.html




From peter at engcorp.com  Wed Apr  3 01:29:29 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 03 Apr 2002 01:29:29 -0500
Subject: Python compilers
References: <3CAA9786.2070700@moofed.com> <3CAAA0BF.E8E4659F@engcorp.com>
Message-ID: <3CAAA149.9831CD98@engcorp.com>

Peter Hansen wrote:
> 
> > What is the performance of python like compared to other languages such
> > as C and java?
> 
> Similar to Java, ...

Paul's response reminds me of the existence of just-in-time
and native compilers for Java, which clearly run much faster
than vanilla Java running on a virtual machine.  I meant
"similar to Java" only in the sense of the original Java.

-Peter


From predictr at bellatlantic.net  Mon Apr 15 06:57:17 2002
From: predictr at bellatlantic.net (Will Dwinnell)
Date: 15 Apr 2002 03:57:17 -0700
Subject: Neural Networks
References: <3CB9E3D9.121EA05C@ntlworld.com>
Message-ID: <2b7b8021.0204150257.6c184939@posting.google.com>

tony clarke <a.clarke11 at ntlworld.com> wrote:
"Does anyone know any sites, or have any code or other info about
artificial intelligence. I am specifically interested in simple neural
network programming, being a simple soul myself!"


I would suggest reading the FAQ and asking questions over in the
comp.ai.neural-nets Usenet group.  Also, see http://www.kdnuggets.com
for a reasonable listing of commercial neural network products.  Last,
I will humbly direct the reader to the Publications section at my own
Web site, at http://will.dwinnell.com     See my articles on neural
networks and especially "Will's Technical Book List" (which includes
ISBN numbers and short descriptions of each recommended title),
featuring good introductory titles on the subject.


From pyth at devel.trillke.net  Thu Apr 18 23:05:35 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 19 Apr 2002 05:05:35 +0200
Subject: PEP for an unrestricted __init__ ?!
In-Reply-To: <7mHv8.101687$K5.8348821@bin5.nnrp.aus1.giganews.com>
References: <mailman.1019162616.18095.python-list@python.org> <7mHv8.101687$K5.8348821@bin5.nnrp.aus1.giganews.com>
Message-ID: <20020419030535.GG28515@devel.trillke>

On Thu, Apr 18, 2002 at 10:00:36PM +0000, Terry Reedy wrote:
> To expand on JR's answer and identify your misconceptions:
> 
> > In python each new instance of a class is created by __init__
> 
> No, instance is first created by internal default constructor or
> new-style user-supplied equivalent __new__().  Then __init__ is
> called, if it exists, with new instance as first arg to initialize
> that new instance.

thanks! i also realized this after reading PEP 253. i certainly
haven't understood this PEP completly, though :-)
 
> > __init__ would then
> > be allowed to return something other[1] than the implicite 'self'.
> 
> > [1] For backward compatibility we might need to treat
> >     'return None' as 'return self' because __init__ currently
> >     seems to follow this rule.
> 
> __init__ is called as a proceedure -- *any* return is ignored

actually a TypeError is raised if you don't return None. At least
in python2.2rc2 this is so.

The question for me still is if relaxing the restrictions with 
__init__  or rather __new__ (which seems to have the same restrictions) 
would bring the benefit of making constructors as flexible as a separate
factory object could be. This way the class would have the control 
over what it actually means to 'create an object with the class's behaviour'. 
Two quick examples of uses:

- wrapping the instance with a generic cache class
  (see http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52201 for
  an example).

- returning a proxy class which actually performs remote network
  calls to the 'real' instance.

the point here is that the calling side of the constructor
doesn't need to *know* about what the factory-like constructor
thinks is appropriate but can continue to rely on 'expected behaviour'. 

Of course you can put this functionality into a separate factory
object. But my whole point is that python already obsoletes
reasons for having factories (like in java or c++). And that it might
be interesting to introduce even more flexibility for constructors
so that the classical redundancies of Factories and Constructors
can be eliminated for many cases.

regards,

    holger




From erno-news at erno.iki.fi  Tue Apr  9 01:25:41 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 09 Apr 2002 08:25:41 +0300
Subject: Python server from inetd?
References: <iml4bukknqf9nqe13h7dp0qco3qj2r86rt@4ax.com>
Message-ID: <kuwuvhmoiy.fsf@lasipalatsi.fi>

In article <iml4bukknqf9nqe13h7dp0qco3qj2r86rt at 4ax.com>, Tim Roberts
<timr at probo.com> writes:

|   #! /usr/bin/python
|   import socket

|   client = socket.fromfd( 0, socket.AF_INET, socet.SOCK_STREAM )
|   while 1:
|     bfr = client.read(1024)
|     if not bfr:
|       break
|     ...

| But this does not work.  The "read" call returns 0 immediately, over and
| over, until the client exits.  I never get any data.

it shouldn't work, since socket objects don't have a read method, only
recv.  getting 0 sounds really odd, since both s.recv nor f.read only
return strings. are you sure you're interpreting this correctly?
can you post the real code?

  -- erno


From dsavitsk at e-coli.net  Mon Apr  8 00:20:07 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 08 Apr 2002 04:20:07 GMT
Subject: Newbie: Alarm at end of program
References: <es5HyN42BHA.1996@cpimsnntpa03> <YS8s8.5401$d7.1425130@newssrv26.news.prodigy.com>
Message-ID: <XT8s8.5402$d7.1425262@newssrv26.news.prodigy.com>

oops ...


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:YS8s8.5401$d7.1425130 at newssrv26.news.prodigy.com...
> below are 2 solutions that showed up.  try google groups for getting old
> postings.
>
> -d
>
>
> "dsavitsk" <dsavitsk at e-coli.net> wrote in message
> news:5aRq8.3571$d7.1106160 at newssrv26.news.prodigy.com...
> > on win32 also try
> >
> > >>> import winsound
> > >>> winsound.Beep(440,250)
> >
> > the first number is the frequency, the second is the duration.
> >
> > -d
> >
> >
> > "Trond Gaarder" <trondga at -NOSPAM-ifi.uio.no> wrote in message
> > news:a8gfr8$h4u$1 at maud.ifi.uio.no...
> > > Don't know if this is the best/correct way of doing it in Python, but
it
> > > works for me (on a PC):
> > >
> > > print "\a"
> > >
> > > This should emit a beep from the pc. What it sounds like is highly
> > > machine-dependent, however...
> > >
> > >
> > > -Trond
>
>
> "billludden" <billludden at email.msn.com> wrote in message
> news:es5HyN42BHA.1996 at cpimsnntpa03...
> > Hi,
> > I have written a script that runs from a few minutes to several hours
> before
> > it finishes.  How would I make the computer play a 'beep' to signal that
> it
> > is finished?
> > Thank you,  Bill
> >
> >
>
>




From sholden at holdenweb.com  Tue Apr 30 11:27:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 11:27:00 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net> <mailman.1020176703.26327.python-list@python.org>
Message-ID: <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com>

"Roman Neuhauser" <neuhauser at mail.cz> wrote in message
news:mailman.1020176703.26327.python-list at python.org...
> > From: "Fredrik Lundh" <fredrik at pythonware.com>
> > Subject: Re: Comment on draft PEP for deprecating six builtins
> > To: python-list at python.org
> > Date: Tue, 30 Apr 2002 14:01:42 GMT
> >
> > Roman Neuhauser wrote:
> >
> > > having spotty knowledge of python
> >
> > that's pretty obvious from your posts.
> >
> > maybe you should spend a little more time learning, and
> > a little less time telling us about your "opinions" on issues
> > you don't really understand?
> >
> > further reading:
> >
> >
http://www.google.com/search?q=cache:opG_4HjTnbIC:www.theonion.com/onion3811
/child-rearing_advice.html
>
>     how does that relate to python or OO?
>
I  think it actually relates a little more to giving advice on subjects you
confess not to know much about.

If you haven't come across the effbot before, you should know that he tends
to be rude in this way. I think it happened when the parity unit was fitted
to his main memory. Fortunately it does not stop him from being a valuable
source of information on topics Pythonic. Every now and then he jumps out
and bites a poster on the ankles due to simple crotchetiness, and then he
will subside into inactivity once more, contemplating the innards of his
software in silent satisfaction.

going-the-same-way-myself-ly y'rs  - steve





From martin at v.loewis.de  Mon Apr 15 02:34:09 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:34:09 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <3CBA64AD.AE4C563D@cosc.canterbury.ac.nz> <3cba6963@news.mhogaming.com>
Message-ID: <m3zo05lbby.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> | The only reason negative integers come into
> | the picture now is that setting bit 31 of a
> | non-long int has the side effect of making it
> | negative. Once ints and longs are unified,
> | this presumably will no longer happen.
> 
> "presumably" ???
> 
> Well, WILL it, or WON'T it?  

It WILL, in phase B.1.

> "presumably" just doesn't cut the mustard with me.  

If you want to cut the mustard, RTFP.

Regards,
Martin


From jepler at unpythonic.net  Tue Apr 30 12:50:38 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 30 Apr 2002 11:50:38 -0500
Subject: Third Draft: Pep for Deprecating Builtins
In-Reply-To: <200204301647.g3UGlVv30660@bald.unpythonic.net>
References: <mailman.1020129837.31112.python-list@python.org> <CEsz8.80366$vF6.2381303@news2.tin.it> <20020430111828.C17488@unpythonic.net> <200204301647.g3UGlVv30660@bald.unpythonic.net>
Message-ID: <20020430115038.E17488@unpythonic.net>

On Tue, Apr 30, 2002 at 06:47:25PM +0200, Alex Martelli wrote:
> No doubt the purpose of "from __past__" is indeed to enable such
> further additions, which is part of what makes me -0 on it.  But it
> IS still something of a <shrug>.

Actually, I'd say I'm -1 on it, despite being the person who wrote it,
since I don't think that the deprecation process needs to be dragged on
any longer than it already is.  In fact, I was trying to make light of
the situation.

Ah well...
Jeff




From marklists at mceahern.com  Sun Apr 14 00:40:02 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 13 Apr 2002 23:40:02 -0500
Subject: copy protection
In-Reply-To: <EFEFIHAINIINNELLHFAIMEBKCBAA.kseehof@neuralintegrator.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEPIKKAA.marklists@mceahern.com>

[Ken Seehof]
> Does anyone know of anyone who provides copy protection services
> for python?
> What I'm looking for is something based on bytecode obfuscation.

I don't know anything about it, but here's the tip of a thread:

  http://groups.google.com/groups?hl=en&selm=39de791e_4%40corp.newsfeeds.com

I found with this search:


http://groups.google.com/groups?hl=en&q=bytecode+obfuscation+python&btnG=Goo
gle+Search

Cheers,

// mark





From never at mind.info  Sun Apr 28 19:41:33 2002
From: never at mind.info (G. Willoughby)
Date: Mon, 29 Apr 2002 00:41:33 +0100
Subject: Is there a text file somewhere i could download that contains all the words in the english language that i could read into python to create a spell checker?
Message-ID: <aai16e$m3$1@newsg1.svr.pol.co.uk>

Is there a text file somewhere i could download that contains all the words
in the english language that i could read into python to create a spell
checker?

Ta,

G. Willoughby




From max at alcyone.com  Thu Apr  4 14:18:07 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 04 Apr 2002 11:18:07 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net> <3CAC149F.3F1BA77B@lemburg.com> <mailman.1017930740.20674.python-list@python.org> <eppstein-AA64FA.09075504042002@news.service.uci.edu>
Message-ID: <3CACA6EF.B0E180AB@alcyone.com>

David Eppstein wrote:

> In article <mailman.1017930740.20674.python-list at python.org>,
>  Guido van Rossum <guido at python.org> wrote:
> 
> > I don't think Boolean algebra is a big subject.  I bet that most
> > people only learn about it explicitly when they learn computer
> > programming.
> 
> There are advanced graduate math books on Boolean algebra.
> For instance Halmos's _Lectures on Boolean Algebras_ or possibly
> Johnstone's _Stone Spaces_.  So the subject may be bigger than you
> think
> it is...

Operative word here being _most_.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From srumbalski at prodigy.net  Wed Apr 10 23:29:24 2002
From: srumbalski at prodigy.net (Steven Rumbalski)
Date: Thu, 11 Apr 2002 03:29:24 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <a8tov6$484$1@panix1.panix.com> <231u8a.lc.ln@127.0.0.1> <a8vnt9$roa$1@panix1.panix.com>
Message-ID: <oq7t8.661$hi.96709670@newssvr15.news.prodigy.com>

Aahz wrote:

> In article <231u8a.lc.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
>>Aahz wrote:
>>>
>>> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
>>> free to buy another book if K&R turns out to be too heavy for you right
>>> now, but recognize that you'll need it eventually.
>>
>>Really?  I've been programming in C for ten years, and I've never even
>>seen the K&R book.  I think I learned it just fine.  Not that I have a
>>better suggestion.
> 
> I'm of the opinion that if you used K&R, you'd agree with me.

I'm with Carl on this one.  One does not *need* K&R to become a good C 
programmer.  One can pick up the zen of C (if there such a thing as a "zen 
of C") without K&R.  Of course reading K&R would probably make that path 
less painful.

Steven Rumbalski


From max at alcyone.com  Tue Apr  2 14:13:46 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 02 Apr 2002 11:13:46 -0800
Subject: Quick return question
References: <0Ujq8.18$c6.7717@typhoon.mn.ipsvc.net>
Message-ID: <3CAA02EA.A98543B6@alcyone.com>

joel8bit wrote:

> Why am I getting None as well, because there's no return value for the
> function?

In a word, yes.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From centuarii at yahoo.com  Thu Apr 25 15:45:02 2002
From: centuarii at yahoo.com (BoB W)
Date: 25 Apr 2002 12:45:02 -0700
Subject: newbie question about cygwin and wxpython
Message-ID: <d0074b71.0204251145.23d0607b@posting.google.com>

I have loaded cygwin 1.3.10, which installs Python 2.2.  I have also
loaded wxpython 2.3.2.1 hybrid installer for windows to
c:\cygwin\lib\python2.2.  wxpython does show up as evidenced by the
fact that python will attempt to import it, but when python calls the
__init__.py function, it can't find wxc to import it.  The only wxc
file I see is wxc.pyd in the wxpython directory.  From within the bash
directory, this shows up as /usr/lib/python2.2/wxpython/wxc.pyd.  Is a
*.pyd file a python dll file?  It seems to me that python does not
have the write path as the message I get is

Traceback (most recent call last):
  File "./grayul.py", line 4, in ?
    import wxPython
  File "/usr/lib/python2.2/wxPython/__init__.py", line 20, in ?
    import wxc
ImportError: No module named wxc

I have tried various things to make the wxpython directory show up in
my sys.path, but none have worked.  Is this a path issue?  If it is, I
can post what I have already tried and see if someone can guide me on
how to get the wxpython directory to show up in sys.path.  Otherwise,
am I supposed to have a wxc.py or wxc.pyc file?

BoB Wood


From greg at cosc.canterbury.ac.nz  Thu Apr 11 01:24:37 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 11 Apr 2002 17:24:37 +1200
Subject: ANN: Pyrex 0.1.1
Message-ID: <mailman.1018559647.27725.clpa-moderators@python.org>

I have fixed a few bugs in Pyrex and made it available
as version 0.1.1:

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

While you're there, pick up the latest version of
Plex too (1.1.2), which doesn't use yield as a
method name any more.

Fixes in this version:

* Bogus re-use of temp variables shouldn't happen now.
* Access to builtin names now works.
* Exceptions should work now (in simple cases, anyway).
* Conditions in if-statements now evaluated correctly.

More bug fixes are in the works. Thanks to all who
have provided feedback so far -- keep it coming!

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




From bkline at rksystems.com  Tue Apr 30 20:29:05 2002
From: bkline at rksystems.com (Bob Kline)
Date: Tue, 30 Apr 2002 20:29:05 -0400
Subject: Continuation after 100 header in httplib
Message-ID: <ucudihfo3n5v38@corp.supernews.com>

I'm looking through the docs for the httplib module, and I can't find any
way to get the next set of headers and the message body when the server
sends back a status code of 100 ("Continue").  It's not appropriate for the
client to invoke Request() again, since it has already sent the request in
the first POST.  Trying getresponse() again (the most intuitively logical
next step) gets back a ResponseNotReady exception.  What am I missing?
Looking at the code for the module it seems as if this condition was
overlooked, but that seems too implausible.




From richard at electrophobia.com  Wed Apr 24 09:24:21 2002
From: richard at electrophobia.com (Richard Parker)
Date: Wed, 24 Apr 2002 13:24:21 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com>
Message-ID: <B8EC0013.EFCE%richard@electrophobia.com>

Paul Rubin at phr-n2002a at nightsong.com wrote on 4/24/02 5:37 AM:

> How about if I just hash a second time, i.e. H(H(K || x)) or
> H(K || H(K || X)), rather than coding up the full HMAC machinery?

H(K1 || H(K2 || X)) is secure.  The key-length equivalent security of this
construction is given by the length of one key, not both.  This construction
is essentially HMAC, except that the HMAC construction simplifies the keying
and makes the key-length equivalent security more obvious by adding a simple
key-separation technique to derive both keys from a single key.

-Richard



From phr-n2002a at nightsong.com  Mon Apr  1 06:57:31 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 01 Apr 2002 03:57:31 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it>
Message-ID: <7xk7rr3a38.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> No, -1, please don't.  It seems to me that the advantages are small:
> ...
> Alex

Alex's post is excellent and I hope people read it carefully.

The point about pickle/marshal is especially good.  I can probably
think of a other dozen other serialization and remote object Python
modules that would also have to be updated.


From johnroth at ameritech.net  Sun Apr  7 20:02:51 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 7 Apr 2002 20:02:51 -0400
Subject: OT: US election
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net> <3CB08659.F3ED147@geneva-link.ch>
Message-ID: <ub1n6plde70f87@news.supernews.com>

"Boris Borcic" <borcis at geneva-link.ch> wrote in message
news:3CB08659.F3ED147 at geneva-link.ch...
> Andrew Dalke wrote:
> >
> > Just as an off-topic reminder.  The difference between Bush and
> > Gore was less than 2,000.  Given the final Florida results of
> >     Bush    2,909,176, Republican
> >     Gore    2,907,451, Democrat
> >     Nader      96,837, Green
> >     Browne     18,856, Libertarian
> >     Buchanan   17,356, Reform
> >     Phillips    4,280, Official Constitution Party
> >     Hagelin     2,287, Natural Law
>
> BTW, I've been wondering if all the problems with voting machines
> (if not Bush's election itself) should not be ascribed to the infamous
> Y2K bug : after all, if programmers had not been monopolized by the
> task of preventing the bug hitting, maybe some work force would have
> been applied to design and implement more modern voting machines.
>
> Boris Borcic

No. US elections are dominated by an attitude of how many
votes a party can steal while still claiming they are honest. There
has been a great deal of discussion of alternative voting methods
over here, and the issues are not programming related. On one side,
they come down to money (which is always in chronically short
supply,) and on the other they come down to protecting the
integrity of the voting process. The technologically snazziest
solutions (touch screen computers) are three times as expensive
as the next technology, and don't do a very good job of protecting
the integrity of the individual ballot, while also protecting the
anonymity of the individual voter.

The money issue is up front investment versus running costs.
Touch screen is much more expensive up front, but it has
much lower per-election costs: all those separate ballots
don't have to be printed.

John Roth





From fperez528 at yahoo.com  Wed Apr 10 15:50:00 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 10 Apr 2002 13:50:00 -0600
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at>
Message-ID: <a92533$4bs$1@peabody.colorado.edu>

Siegfried Gonzi wrote:

> And I have really, really not the intention to waste my life as a
> professional programmer

You might want to consider that in a programming group such as c.l.p, there's 
bound to be more than one person who will consider 'to waste my life as a 
professional programmer' to be a rather offensive way of asking for 
suggestions on any topic.

Cheers,

f.

ps. for the record, I'm _not_ a professional programmer. Nor am I a wannabe 
Miss Manners, I simply feel that a bit of civility every now and then isn't a 
bad thing, esp. considering how hot the flames have recently been around 
here.


From bbollenbach at shaw.ca  Thu Apr 11 20:43:27 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 00:43:27 GMT
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
References: <oOjt8.8815$vF6.254710@news2.tin.it>
Message-ID: <P4qt8.15576$Kq4.661039@news2.calgary.shaw.ca>

In article <oOjt8.8815$vF6.254710 at news2.tin.it>, Alex Martelli wrote:
> Studying PyQt (latest releases of everything), I started noticing 
> segmentation faults as I quit my application and eventually whittled it 
> down back to...:
> 
> #!/usr/bin/env python2.2
[snip strangeness] 

For what it's worth, and if it's of any help to either martellibot or
Phil, the code runs without error -- clicking Quit or closing via the
"X" -- on Debian woody, Python 2.1, and with the python-pyqt package
installed.

HTH,

Brad Bollenbach


From richard at stockcontrol.net  Fri Apr 19 13:14:44 2002
From: richard at stockcontrol.net (Richard Walkington)
Date: Fri, 19 Apr 2002 18:14:44 +0100
Subject: raw_input bug only in cygwin python
References: <12257ec4.0204181907.7a79b15b@posting.google.com> <kuavbukpa505nru9g6kfivo3bfrtirda05@4ax.com> <12257ec4.0204190610.44c9774@posting.google.com>
Message-ID: <3CC05084.4020104@stockcontrol.net>

David Carson wrote:

>Tim Roberts <timr at probo.com> wrote in message news:<kuavbukpa505nru9g6kfivo3bfrtirda05 at 4ax.com>...
>  
>
>>davidccarson at hotmail.com (David Carson) wrote:
>>
>>    
>>
>>>Here is a snippet of code:
>>>
>>>01: #!/usr/bin/env python
>>>02: 
>>>03: list = []
>>>04: while 1:
>>>05:     try:
>>>06:         stuff = raw_input("[^D to exit]: ")
>>>07:     except EOFError:
>>>08:         print "got EOF"
>>>09:         break
>>>10:     list.append(stuff)
>>>11: 
>>>12: raw_input("bang! ")
>>>
>>>
>>>This works on every Python environment I've tried but cygwin (Win2000
>>>-- I haven't got NT to try it):
>>> Python 2.2 (#1, Dec 31 2001, 15:21:18)
>>> [GCC 2.95.3-5 (cygwin special)] on cygwin
>>>
>>>In this environment, the EOFError is caught.  Then the "bang! " prompt
>>>for the raw_input() outside the loop is printed, but immediately the
>>>program exits with an uncaught EOFError.
>>>
>>>This code works with:
>>>- ActiveState Python, 2.1 on Win2000
>>>- python 2.2 on Mandrake
>>>- python 1.5.2 on RedHat
>>>and everywhere else, I'm sure.
>>>
>>>Is this a known problem with cygwin?
>>>      
>>>
>>I wouldn't call it a problem with cygwin, I'd call it "relying on
>>unspecified behavior".  Ctrl-D is end of file; the ability to read past EOF
>>is not promised anywhere in the docs.
>>    
>>
>
>
>Ah, I see.  However, I've seen this technique recommended in several
>docs/tutorials.  Also, the docs do promise raw_input() will raise
>EOFError when EOF is read.  So, if it is true that I'm relying on
>unspecified behavior, my question is: How do I make this code
>reliable?  Is there a simple way to clear the EOF indication on stdin?
> I hope there is a good way, because I like this kind of input loop
>much better than the "type 'q' to quit" ilk.
>
>Thanks,
>David
>  
>

I don't see anything wrong with your code. I wouldn't have throught the 
second raw_input should receive an EOF. I've also seen this technique 
used in tutorials etc.

Regards
Richard Walkington





From grante at visi.com  Wed Apr 10 17:24:20 2002
From: grante at visi.com (Grant Edwards)
Date: Wed, 10 Apr 2002 21:24:20 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at> <a92533$4bs$1@peabody.colorado.edu>
Message-ID: <842t8.23204$vm6.3654628@ruti.visi.com>

In article <a92533$4bs$1 at peabody.colorado.edu>, Fernando P?rez wrote:

>> And I have really, really not the intention to waste my life as
>> a professional programmer
> 
> You might want to consider that in a programming group such as
> c.l.p, there's bound to be more than one person who will
> consider 'to waste my life as a professional programmer' to be
> a rather offensive way of asking for suggestions on any topic.

I am one (well, my business card says "software engineer"), and
some days it does feel like a life-wasting experience. :) Most
of the time it's getting paid to do stuff I'd probably be doing
in my spare time anyway.

-- 
Grant Edwards                   grante             Yow!  I'm in DISGUISE as a
                                  at               BAGGAGE CHECKER....I can
                               visi.com            watch the house, if it's
                                                   ORANGE...


From python at rcn.com  Mon Apr 15 23:07:54 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 15 Apr 2002 23:07:54 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <a9eppf$ghv$1@panix1.panix.com> <a9f95n$jg0$1@bob.news.rcn.net> <ubmftha3r74h80@news.supernews.com>
Message-ID: <a9g4k1$bn7$1@bob.news.rcn.net>

> > Rules of the game:
> > '''A PEP that has not been discussed on python-list at python.org and/or
> > python-dev at python.org will not be accepted.  ''' -- from Pep 1
>
> Does that mean by the PEP editor or by the BDFL? I thought the
> process was: write it, send it in, discuss it, then the BDFL makes
> a Pronouncement.

You're correct.  That's the way PEP 1 says to do it.

My additions to the process are:
-- float the idea around a little bit first so I don't end-up proposing
something that will look preposterous in retrospect.
-- just before posting, don a flame-proof suit and flack jacket.  Pep
writers should have no illusions about being greeted with open arms (even
Guido took some abuse for pep 285).


Raymond Hettinger






From dalke at dalkescientific.com  Sat Apr 20 13:12:17 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 11:12:17 -0600
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at>
Message-ID: <a9s7el$sr7$1@slb2.atl.mindspring.net>

Siegfried Gonzi:
>What Python --from my pragmatic point of view-- has in common with Lisp is
>the nice feature that you can put in your return  list different objects,
>
>e.g. [[1,2,3],["334.233,"Ein bloeder Oesterreicher",777],array]

This is also present in Perl and Tcl, and also I believe in
Smalltalk, Mathematica, and even APL.

>And it is possible in Python (sorry I know only the idle) to build your
>functions in the command-line (every Common Lisp environment provides you
>this).

I don't know what it's called in Common Lisp, but in Python I think
you are refering to the interactive mode.  The command-line is what's
used to start Python from, say, the unix shell.

>I often have heard that CLOS (Common Lisp Object System) lets you
>change your class objects on the fly, I do not have any experience with
that
>and I am not sure how this relates to Python (my Python object oriented
>programming experience is zero).

You can modify Python classes and instances as well, though I don't
know how similar it is to CLOS.  I do know there are differences between
how Python does it and how Smalltalk does it.  (Eg, reload() doesn't
update old versions of newly redefined classes.)

                    Andrew
                    dalke at dalkescientific.com





From eppstein at ics.uci.edu  Thu Apr  4 20:19:05 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 04 Apr 2002 17:19:05 -0800
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org>
Message-ID: <eppstein-4311CC.17190504042002@news.service.uci.edu>

In article <mailman.1017952983.647.clpa-moderators at python.org>,
 Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

> Oh, yeah, here's the example. It finds prime
> numbers at C speed.
> 
> def primes(int kmax):
>   cdef int n, k, i
>   cdef int p[1000]
>   result = []
>   if kmax > 1000:
>     kmax = 1000
>   k = 0
>   n = 2
>   while k < kmax:
>     i = 0
>     while i < k and n % p[i] <> 0:
>       i = i + 1
>     if i == k:
>       p[k] = n
>       k = k + 1
>       result.append(n)
>     n = n + 1
>   return result

Not to criticise your project, but this is not a good example to use -- 
you've started with a very unnecessarily slow algorithm.  For finding 
prime numbers, the Sieve of Eratosthenes is much better.  See e.g.
<http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/117119> which I 
implemented using simple generators so you don't need the kmax parameter.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From sth.news at uni.de  Sat Apr  6 16:04:13 2002
From: sth.news at uni.de (Stephan Hohe)
Date: Sat, 6 Apr 2002 23:04:13 +0200
Subject: Size of Newsgrouppostings with nntplib
Message-ID: <a8nnlo$tv5d1$1@ID-109762.news.dfncis.de>

Hi,

I wrote a python skript to generate some newsgroup-statistics. I use the 
NNTP object in module nntplib to connect to the newsserver and recieve 
interesting message-headers useing the xhdr method. Now I'm searching for 
a way to recieve the size of the postings. The only thing I found is the 
xover method, witch submitts many informations, including the article 
size in bytes. Is there a way to recieve the article size without loading 
all the additional informations submitted by xover?

Thanks,
Stephan


From jolsen at mailme.dk  Tue Apr  9 02:45:25 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 8 Apr 2002 23:45:25 -0700
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <cf0ad9fb.0204082245.58ec2e@posting.google.com>

You should definately go for Kernighans "C Programming Language".

Jesper


From ny_r_marquez at yahoo.com  Thu Apr  4 10:54:19 2002
From: ny_r_marquez at yahoo.com (R.Marquez)
Date: 4 Apr 2002 07:54:19 -0800
Subject: Downloading pages from a Windows Intranet
Message-ID: <8a27e309.0204040754.5073e0b2@posting.google.com>

Trying to use urllib or urllib2 to get a page from our Intranet
(Windows) gives me an Access Denied error.  The funny thing is that on
some pages it works (?).  Any way, someone found a Perl module
(Win32::Internet) that creates an object oriented interface to the
Win32 Internet Functions (WININET.DLL).  With it we can get any page
from our intranet, even asp pages.  Is there an equivalent module or
functions that I could use from Python?  Or can anyone give me an
example of how to work around this?  Or should I just settle for a
Perl/Python solution?


From mgilfix at eecs.tufts.edu  Wed Apr 17 10:36:47 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 10:36:47 -0400
Subject: Div crypto stuff
In-Reply-To: <scbd9e21.061@Riksvasa.syd.kth.se>; from tdi01fhu@syd.kth.se on Wed, Apr 17, 2002 at 04:08:50PM +0200
References: <scbd9e21.061@Riksvasa.syd.kth.se>
Message-ID: <20020417103647.M16954@eecs.tufts.edu>

  Heh. Perhaps I'll do RSA next. Much easier to implement than DES :)

On Wed, Apr 17 @ 16:08, FREDRIK HULDTGREN wrote:
> >From:   Michael Gilfix <mgilfix at eecs.tufts.edu> 
> >Date:   Wednesday - April 17, 2002 8:17 AM 
> >Subject:   Pure python crypto 
> >   
> > Would anyone have any use for a pure python implementation of
> >Blowfish? Or any other crypto algorithms for that matter.  As a matter
> >of interest, I found myself implementing the blowfish module recently
> >but was unsure if any audience would be interested. No C compiler
> >is required, which is nice. Could use some testing though.
> >             -- Mike 
> 
> >From:   Fr?d?ric Laurent <liste_nospam_ at opikanoba.org> 
> >Subject:   amkCrypto installation on win32 
>    
> >Hello,
> >I want to use RSA encoding/decoding, so I downloaded amkCrypto 0.1.3.
> >I use python 2.2 on windows XP.
> 
> >Must I have to download and install swig in order to make amkCrypto work
> >(for the mxCrypto module) ?
> >I don't have any C++ compiler installed, and it seems to be quite complicated >:
> >amkCrypto -> mxCrypto -> swig -> c++ and so on...
> >moreover the link http://starship.skyport.net/crew/amk/python/crypto.html, >given
> >in the documentation, on the pycrypt package is broken...>>
> 
> >So, I just want to generate a pblic/private key, crypt and decrypt a text
> >buffer... but it sounds to be not so simple to do this !>
> 
> >Did I miss something ???
> 
> >thanks in advance
> 
> >--
> >Fr?d?ric Laurent
> >http://www.opikanoba.org  
> 
> 
> These type of questions are more likely to be answered on the Python-Crypto mailing list.
> 
> http://listserv.surfnet.nl/archives/python-crypto.html
> 
> /crast
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (tdi01fhu)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/73205c45/attachment.sig>

From JamesL at Lugoj.com  Sun Apr  7 21:33:38 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 8 Apr 2002 01:33:38 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84> <m4isn6712s9.fsf@macquarie.com.au>
Message-ID: <Xns91E9BCA9CF9ACJamesLLugojcom@216.148.53.85>

Timothy Docker <timd at macquarie.com.au> wrote:
> James Logajan <JamesL at Lugoj.com> writes:
>> When I see that happening (maybe in a few
>> years?) then I would consider rejoining the ranks of Python
>> programmers. 
> 
> Out of curiousity, which language are you using in the meantime?

When I have the choice of language I'm using C and occasionally C++, as 
needed. Low level design and coding per se does not consume a significantly 
large fraction of my time on current projects. Also, I've programmed in C 
for about 20 years, so it flows very quickly and I have a large repository 
of solutions. I'll reevaluate scripting languages at some point in the 
future with more of an eye this time around on what might be called 
"infrastructure" issues as well as the languages.


From cgale1 at cox.net  Fri Apr 19 03:41:19 2002
From: cgale1 at cox.net (Van Gale)
Date: Fri, 19 Apr 2002 07:41:19 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net>
Message-ID: <zSPv8.41559$%A.3708530@news1.west.cox.net>

"Frederick H. Bartlett" <fbartletFIXIT at optonline.net> wrote in message
news:3CBF71D4.5668CF1A at optonline.net...

> So I did it in Python, too. But the best time I could get from Python
> was .57 sec, while OmniMark came in at .20 sec. What's the most
> efficient technique for Pythonesque regex-based text processing?
>
> My best time came from using a single rather large regex and findall; I
> also tried smaller regexes and scan and match.

Internally OmniMark builds one big state machine for all the find rules...
pretty much the same thing as what you did with the single large regex.
This would have been the only recommendation I could make, but I don't know
any performance tricks for Python regex's.

> Has anyone else compared OmniMark and Python?

OmniMark has had years and years to optimize their engine and it's the only
thing they do.  I've never seen anything faster, and that makes a HUGE
difference when you're processing gigabytes of data, however the language is
a travesty.  Well, I should qualify that... it's still more readable and
writable than perl, and I mostly used it back in the days before they had
those new-fangled "function" thingies.

I'd recommend OmniMark if the performance really will make a difference or
if you need 100% standard SGML or XML processing. Otherwise, it'll be easier
to find Python programmers, and the code will be maintainable :)

Van





From rjgruet at yahoo.com  Thu Apr 18 20:35:19 2002
From: rjgruet at yahoo.com (Richard Gruet)
Date: Fri, 19 Apr 2002 02:35:19 +0200
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <E4Iv8.2972$8D3.78501@news1.tin.it>
Message-ID: <a9noiq$btq$1@aphrodite.grec.isp.9tel.net>

Alex

Indeed, your arguments are very good and make me think : e.g. I wouldn't
like to get rid of the operators, so why dislike functions ? And I  simply
forgot that there was already a __len__ method !

I humbly implore you, Great Guru, have mercy, don't hit me this time ! I am
sincere and naive in my stupidity. I didn't fully realize how arrogant and
foolish I was by expressing shallow feelings without having first completely
understood the design of the language. I forgot my rank. Will you forgive me
? I swear I'll try hard to be worthy of this newsgroup next time and won't
ask silly questions that waste your precious time and might spoil the spirit
of innocent readers.

Peace on earth (including newsgroups),

Richard




From henk.derudder at barconet.com  Mon Apr 29 10:29:39 2002
From: henk.derudder at barconet.com (henk.derudder at barconet.com)
Date: Mon, 29 Apr 2002 14:29:39 GMT
Subject: deploying Python
Message-ID: <3ccd5803.290199187@news.skynet.be>

Hi,

I have written some Python scripts which are loaded (via hooks) in my
C++ code. The C++ code and the scripts work fine on my machine because
Python is installed on it. However, I want my app to run smoothly on
somebody else's computer, especially one which has NOT Python
pre-installed, so I need to deploy Python. Can anybody help me out on
this one please?

Kind regards,

Henk


From hbl at st-andrews.ac.uk  Wed Apr 17 14:16:06 2002
From: hbl at st-andrews.ac.uk (Hamish Lawson)
Date: Wed, 17 Apr 2002 19:16:06 +0100
Subject: Making a C extension compatible with 2.2
Message-ID: <3CBDBBE6.77DD6FD3@st-andrews.ac.uk>

Richard Townsend wrote:

> I found a very similar problem with the ingmod package (Python 
> interface to Ingres DB) on HP-UX 11.

Aha! It so happens it's also ingmod that my problem relates to! :-)  I
recently notified Holger Meyer about it and he said he would try to
address it soon, but I thought I would see what I could do myself in the
meantime (though I'm far from expert on writing C extensions for
Python).

Hamish Lawson




From bokr at oz.net  Sat Apr 13 17:24:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 13 Apr 2002 21:24:12 GMT
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com>
Message-ID: <a9a7ls$gu4$0@216.39.172.122>

On Sat, 13 Apr 2002 12:19:35 -0700, bvdpoel at uniserve.com wrote:

>
>Yeah, I thought this would be easy...
>
>I have a list like this:
>
>	[ [0, x, [0, y, [4, n] ...]
>
Is this what it appears to be? I.e., a linked list of [offset,event,link]
rooted in a [] ? Thus:
        [ + ]
          |
          +[0, x, +]
                  |
                  +[0, y, +]
                          |
                          +[4, n *?*]

(*?* what goes here, or are you using len(element)<3 to indicate eol?)

If it is a linked list, do you need that for fast insertions on the fly, or what?
If not, why not [ [0,x], [0,y], [4,n] ...] or [ (0,x), (0,y), (4,n) ...] ?

The sorting is going to be different, depending ;-)
BTW, is there no cheap way to maintain desired order as you build/modify the list,
so you won't need a sort per se?

Regards,
Bengt Richter


From yduppen at chello.nl  Tue Apr 23 01:36:32 2002
From: yduppen at chello.nl (Yigal Duppen)
Date: Tue, 23 Apr 2002 07:36:32 +0200
Subject: Mass Mailing help needed (Paul Rubin)
References: <mailman.1019325819.29016.python-list@python.org>
Message-ID: <3cc4f2e1$0$37897$e4fe514c@dreader3.news.xs4all.nl>

Is MailMan an option? (http://www.gnu.org/software/mailman/mailman.html) 
It's a mailinglist app written in Python, so it must be easy to incorporate 
it into your program.

Not only are mailinglists good examples of mass mailing, but by using a 
mailinglist, you get the opt-out part for free. And it will save you a fair 
amount of coding/testing; after all, network related programs are a *** to 
make and test. 

YDD
-- 
.sigmentation fault


From marklists at mceahern.com  Tue Apr 30 23:05:58 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 22:05:58 -0500
Subject: Python vs. Perl,  which is better to learn?
In-Reply-To: <9UHz8.23$s55.5@newsread2.prod.itd.earthlink.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEJCKNAA.marklists@mceahern.com>

[Billy Ng]
> I am also new to python.  I think python will be my replacement of perl.
> However, if I need to do very heavy regular expression, I still
> prefer perl.
> I just don't feel comfortable with the python's re syntax.

I've heard this sentiment expressed before.  I can count on my hands the
number of times I've used Perl--all before I learned about Python--so I'm
largely ignorant of Perl.

Could you give us a concrete example where the use of regular expressions in
Perl is superior to Python?

Thanks,

// mark





From martin_byrne2 at yahoo.co.uk  Mon Apr 15 07:39:44 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 15 Apr 2002 04:39:44 -0700
Subject: KeyboardInterrupt errors in ASP pages
References: <508cb8ae.0204110045.431dadee@posting.google.com> <3CB550C9.1000902@mxm.dk> <3CB83C5C.9040704@skippinet.com.au>
Message-ID: <508cb8ae.0204150339.711d4e8c@posting.google.com>

Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<3CB83C5C.9040704 at skippinet.com.au>...
> Max M wrote:
> > B wrote:
> > 
> >> I am using Python as the language for creating server side asp pages.
> >> The problem is that when using Response.Redirect(page), I seem to get
> >> the message below randomly but frequently enough to be a problem.
> >>
> >> Has anyone else found this problem and is there a way around it. Is it
> >> a bug in the Response.Redirect code?
> >>
> >> Python ActiveX Scripting Engine error '80020009'
> >> Traceback (innermost last): KeyboardInterrupt
> >> /Default.asp, line 0
> > 
> > 
> > As far as I understand it's a known bug, but without any known solution. 
> > I get it too at times.
> > 
> > I have just signed over a project to a customer on .asp/iis so I am a 
> > bit worried about it. But I don't really know what to do about it.
> > 
> > Maybe I should write a test harness to pound the IIS to see if I can 
> > provoke the error.
> 
> Yes please - that is exactly what I need.  I need some way to provoke 
> the error often enough that I can afford to "wait" for it in a debugger. 
>   Anything you can do to get a reliable repro would be a big help!
> 
> Mark.

Woops made a bit of a mistake in my last posting:

Check out 
http://bugs.activestate.com/show_bug.cgi?id=10204

One of the entries,the third last one, is basically the same as whats
happening to me(my problem is with Response.Redirect). Considering
that the example has only
Response.Write in it, it would seem fairly certain that the
Response object is the culprit. However its not infeasible that
something deeper down is causing the problem which then manifests its
self as a keyboard interrupt during a
Response.Redirect/Response.Write.


From brian at sweetapp.com  Thu Apr 25 23:18:00 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Thu, 25 Apr 2002 20:18:00 -0700
Subject: Python and Cocoa?
In-Reply-To: <eppstein-593D22.19215525042002@news.service.uci.edu>
Message-ID: <004401c1ecd0$f8b70060$445d4540@Dell2>

> Does anyone know of anything useable to write Cocoa apps with Python?

You might want to start with:
http://www.onlamp.com/pub/a/python/2002/1/31/pythonnews.html






From aleax at aleax.it  Tue Apr 30 17:23:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 21:23:05 GMT
Subject: Newbie: finding the key/index of the min/max element
References: <200204301620.00255.bhan@andrew.cmu.edu>
Message-ID: <ZWDz8.35801$8D3.1035880@news1.tin.it>

Benjamin Han wrote:

> In my quest of becoming a "Pythonistas" (where does this word come from

I think it's plural, and the singular "Pythonista".

> BTW?), here comes my 2nd trivial question: how do you find the key/index
> of the max/min element in a sequence/map? I know you can call max(l) but
> this only gives you the element itself, not the key. I also know it's not
> that hard to code it up the hard way (for k in d.keys(): ... # do
> comparison), but is there a more elegant way of doing this?

For a list, you _might_ do L.index(max(L)) -- that's conceptually two
passes on the list, but performance should still be OK.

This doesn't work for a mapping, though.  However, for a mapping (assuming
Python 2.2, else you need to iterate on D.keys() rather than on D):

        max([ (D[key], key) for key in D ])[1]

works (a variation of the DSU pattern -- we could call it DEU, for 
Decorate, Extremal, Undecorate:-).

As to whether this is "more elegant" than a loop -- I'll pass.


Alex



From joost_jacob at hotmail.com  Sun Apr 14 11:37:42 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 14 Apr 2002 08:37:42 -0700
Subject: playing with Tkinter, expert needed ?
Message-ID: <13285ea2.0204140737.55dc49f1@posting.google.com>

Hello all,

I have been playing with Tkinter to make a widget that is portable and
usable from Python version 1.5.2 on.  The widget is for emulation of
very old-fashioned character based computer screens.  Earlier I tried
all sorts of curses modules but not one of them was really portable,
or needed recompiling things.  The source is at
http://www.cwi.nl/~jacob/textmap.py and you can see a syntax colored
version at http://www.cwi.nl/~jacob/textmap.py.html

Basically it gives you a window where you can do things like:
writexy(x, y, text)
settext(text)
myscreen = gettext()
inputString = readxy(x, y)
clear()
change(coloroptions)
scroll(x, y)
etc.

Download and just run
C:> python TextMap.py
for a demonstration.
Please let me know if it does not work on your operation system, it
should be as portable as python with Tkinter is.

My questions are:
a) How can I have a 'fat' cursor blinking instead of the small one
when waiting for input?
b) Do you see implementation flaws in the TextMap class?  I am new to
Tkinter and I am sure Tkinter has all sorts of 'hidden' conventions
about how to use it I don't know about.


From egrossde at yahoo.de  Thu Apr 18 03:28:37 2002
From: egrossde at yahoo.de (jb)
Date: Thu, 18 Apr 2002 09:28:37 +0200
Subject: Reading from a binary file in a portable way
References: <mailman.1019112514.13299.python-list@python.org>
Message-ID: <3cbe7518_5@news.newsgroups.com>

> Why can't you open the file in binary mode under Linux/UNIX?
> 
> You may also find the 'struct' module useful.

I used os.open under MS-WIN with the flag O_BINARY and this is not supported 
under Linux.
-- 
Janos Blazi


-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From aahz at pythoncraft.com  Fri Apr 12 20:04:18 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Apr 2002 20:04:18 -0400
Subject: Queue (Threads)
References: <pan.2002.04.12.21.10.57.987428.703@thomas-guettler.de>
Message-ID: <a97sm2$dqn$1@panix1.panix.com>

In article <pan.2002.04.12.21.10.57.987428.703 at thomas-guettler.de>,
Thomas Guettler  <pan-newsreader at thomas-guettler.de> wrote:
>
>I tried to find examples for the Queue class with google, but
>no look. Has someone links to exmamples or hints?

See my web page.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From cce at clarkevans.com  Wed Apr 17 13:50:54 2002
From: cce at clarkevans.com (Clark C . Evans)
Date: Wed, 17 Apr 2002 13:50:54 -0400
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
In-Reply-To: <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp>; from stephen@xemacs.org on Sat, Apr 13, 2002 at 04:32:15PM +0900
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <20020417135054.A88552@doublegemini.com>

On Sat, Apr 13, 2002 at 04:32:15PM +0900, Stephen J. Turnbull wrote:
| rms's fundamental position is radical: property rights in software are
| morally wrong.  Unfortunately, even rms himself seems incapable of
| behaving in a way consonant with this belief.  (Ie, he clearly
| believes that he has a special right to manage development of projects
| he has initiated, such as Emacs, gcc, or glibc.

There is a difference between "trademark" and "copyright".  RMS firmly
believes in trademarks; and he believes that he has a special right to
manage development of projects (named by a trademark) he has initiated.
RMS has _never_ said that you couldn't take the code and make your 
own fork... *calling it something else*, like XEmacs.  He may frown upon
it since forks in some cases waste effort; but this is a different issue
entirely.  This makes sence.  People have come to know "emacs" as having
a particular flavor and quality, and this flavor and quality is a product
of Stallman's vision for Emacs.  There is nothing wrong with this, as it
does not hurt your ability to fork the code and do-as-you-please.  Not that
you have to believe in RMS's vision, but please don't mis-represent what
he stands for... 

Best,

Clark




From eppstein at ics.uci.edu  Mon Apr  8 02:23:27 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 07 Apr 2002 23:23:27 -0700
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org> <eppstein-4311CC.17190504042002@news.service.uci.edu> <3CB13147.89F02706@cosc.canterbury.ac.nz>
Message-ID: <eppstein-F7B726.23232707042002@news.service.uci.edu>

In article <3CB13147.89F02706 at cosc.canterbury.ac.nz>,
 Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

> > Not to criticise your project, but this is not a good example to use --
> > you've started with a very unnecessarily slow algorithm.
> 
> The point of the example isn't to show how to find prime numbers 
> quickly. It's just a representative CPU-intensive task that 
> demonstrates what Pyrex can achieve. Using a more efficient algorithm 
> wouldn't make that point any more effectively.

But I think your example makes a different point even more effectively: 
if you're having speed bottlenecks that would cause you to want to use 
Pyrex or similar extension modules, you can often do a lot better by 
instead applying some thought and using a better algorithm.

If you're using algorithms that are as efficient as you can make them 
and still not getting enough speed, that may be the time to start using 
extension modules.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From brian at sweetapp.com  Mon Apr 29 02:45:46 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 28 Apr 2002 23:45:46 -0700
Subject: Update locals()
In-Reply-To: <02042908013901.25806@arthur>
Message-ID: <004a01c1ef49$7d858d00$445d4540@Dell2>

> Right -- and having such change attempts be diagnosed at once (by
> making locals() a pseudo-dict that raises on set attempts) would help
> debugging.  

It would be best if the pseudo-dict would only raise an exception if the
variable cannot be set. Debuggers and other tools might want to at least
make the attempt.

> Similar optimizations to globals() are under consideration
> (there are three PEPs out about that, I believe).

And Ping's beautiful diagrams :-)

Cheers,
Brian





From db3l at fitlinxx.com  Thu Apr 25 16:54:08 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 25 Apr 2002 16:54:08 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com> <mailman.1019747686.17849.python-list@python.org>
Message-ID: <upu0n1oun.fsf@ctwd0143.fitlinxx.com>

holger krekel <pyth at devel.trillke.net> writes:

> On Thu, Apr 25, 2002 at 07:29:27AM -0700, Rad wrote:
(...)
> > having been learning Python in the last few weeks
> > I kind of like it + I was encouraged by your posts and I'm now pretty
> > convinced that Python2.2 is the way to go. 
> 
> Be careful. Some people pointed out that in the end it might
> be a task which a database can handle more reliably. 

It's likely not an either/or scenario - even if a database is used
there will probably be some Python code driving its use.

I'm not sure I follow all of the database suggestions though.  While
true, there's some correlation needed amongst the data, the
requirements really seem to point to largely linear scans of the data
followed by some reporting.  Just loading the data into a database is
unlikely to prove that much better than just scanning the data in the
first place.

However, I suppose what might prove useful is to scan the data and
only insert into the database the small portion (perhaps keys needed
for searching/joining and an filename/offset pointer to the disk
record) so you can then use the natural joining operation of the
database for queries.  But given the 2GB of memory on the machine in
question I have a feeling that tracking the keys during processing
right in Python is still very viable.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From idbaxter at semdesigns.com  Sun Apr 14 22:13:44 2002
From: idbaxter at semdesigns.com (Ira D. Baxter)
Date: Sun, 14 Apr 2002 21:13:44 -0500
Subject: Topy news - and a request for Python pretty-printing
References: <LNBBLJKPBEHFEDALKOLCOEBPOOAA.tim.one@comcast.net> <mailman.1018828351.9865.python-list@python.org>
Message-ID: <3cba3516@giga.realtime.net>

"Fran?ois Pinard" <pinard at iro.umontreal.ca> wrote in message
news:mailman.1018828351.9865.python-list at python.org...

> > > Would someone happens to know if there is a pretty-printer for Python
> > > source _lines_, where I would find some inspiration maybe?
>
> Of course, I've been interested in the `pprint' module, at least
> for structures, but not enough to dive and extend it to statements and
> expressions.  The main point is that `pprint' works on in-memory
structures,
> while I need to beautify surface, textual Python

Our DMS Software Reengineering Toolkit can easily build prettyprinters
for arbitrary langauges; we've done it for many standard ones.
See http://www.semdesigns.com/Products/Formatters/index.html
Not Python yet, but all DMS really needs is a good grammar definition.
 (Anybody have a Python lexer/grammar definition?)

--
Ira D. Baxter, Ph.D. CTO Semantic Designs, Inc.
http://www.semdesigns.com




From aleax at aleax.it  Fri Apr 12 03:45:41 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 12 Apr 2002 07:45:41 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk>
Message-ID: <Fgwt8.8607$b62.224911@news1.tin.it>

Max M wrote:
        ...
>>>>>mystr(',').join(range(10),str)
>>>>
>> '0,1,2,3,4,5,6,7,8,9'
> 
> That sounds like a good idea!

I have my doubts that the BDFL wants to special-case the join method to 
save 5 characters over a map call:

        ','.join(map(str, range(10)))

(the saving kicking in only if the mystr(...) is not needed, of course).

Fewer and simpler "Lego Bricks" (TM) that fit together regularly, flexibly, 
and unsurprisingly, seem more Pythonic to me than "pre-fitted bricks" that
are necessarily more copious and each, individually, less flexible.  Since
pragmatism trumps purity, an idiom that's very pervasive (and that is
worth encouraging) AND awkward to express with "few simple bricks" will
sometimes get its own special-cased "pre-fitted brick".  However, it does
not appear to me that this case meets the requirements: this idiom is not
all that widespread, it's quite well expressed with existing tools, and
indeed existing tools have it all over the new proposal in terms of power
and flexibility (what if you want to call a function that takes two
parameters, for example -- map doesn't bat an eyelid, what would the
new proposed extension to the join method do?).

Personally, I'd rather see curry built-in -- it IS easy to write curry
in Python 2.2, but it's such a wonderful idiom (and would replace so
many lambda's;-) that I'd love to see it officially blessed.  Not that
I expect it to be (and nor will my heart be broken because of this:-).


Alex



From logstx at bellatlantic.net  Mon Apr 29 21:41:09 2002
From: logstx at bellatlantic.net (logistix)
Date: Tue, 30 Apr 2002 01:41:09 GMT
Subject: PythonCOM syntax question
References: <mailman.1020109433.4914.python-list@python.org>
Message-ID: <VCmz8.20506$iJ.5901@nwrddc02.gnilink.net>

The problem is that Field is a property, and Call is a method.  If you can
take a look at the IDL (and make sense of it) that shows how the native
calls get tranlated into COM.  I can't remember offhand how PythonCM deals
with properties, but if you use the MakePy utility under tools in PythonWin,
then pythonwin's autocomplete stuff should make it pretty easy to locate the
appropriate function names.

--
-

"Greg Pierce" <greg at turtleprod.com> wrote in message
news:mailman.1020109433.4914.python-list at python.org...
I'm trying to interact with the Navision Axapta ERP package via it's COM
connector...and I'm having trouble with particular types of calls.  I'm
hoping someone can help me if there's some syntax problem I'm having.

I'm able to create the Axapta COM objects and can call certain methods on
them, but not others.  I'm no COM expert, but there is a difference in the
method specs that I can see.  Those that I _cannot_ call all have a "get" in
the API documentation for Axapta -- like this:

HRESULT get_Field( [in] VARIANT field, [out, retval] VARIANT *pVal );

Those that I _can_ call don't -- like this:

HRESULT Call(  [in] BSTR bstrMethod, [in, optional] VARIANT v1,  [out,
retval] VARIANT *pResult);

Is there some different syntax I should need to use in Python to access
those "get" methods?

This code in VB works ( where "buffer" is an already created COM object) :

MyValue = buffer.Call( "openBalanceMST" )
AccountNumber = buffer.Field("AccountNum")

But in Python, the "call" line works, and then the "field" line returns a
"com_error: (-2147352573, 'Member not found.', None, None)"

Any ideas?  Did that explanation make sense?

Thanks,

g.








From phr-n2002a at nightsong.com  Sat Apr 20 06:09:57 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 20 Apr 2002 03:09:57 -0700
Subject: Mass Mailing help needed
References: <mailman.1019288501.20546.python-list@python.org>
Message-ID: <7xlmbiu1e2.fsf@ruckus.brouhaha.com>

"Jonathan Vanasco" <jvanasco at hotmail.com> writes:
> Until I got to the mass-mailing part of the gig.  Based on user
> registration opt-ins, they can receive certain mailings.  ...
> I dont know what that will do to
> postfix should i need to mail 50,000 people (will that fill up the
> disk w/tempfiles, will it skew  the mailqueue too much?) 

Why should we think those 50,000 people have really opted in?  Are you
sure you're not spamming?  What site is this for?


From alwyn at alwyn.demon.co.uk  Mon Apr 15 12:22:48 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Mon, 15 Apr 2002 16:22:48 GMT
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu>
Message-ID: <20020415172247905+0100@pc-62-30-160-65-hw.blueyonder.co.uk>

In <eppstein-7165E1.08012115042002 at news.service.uci.edu> David Eppstein
<eppstein at ics.uci.edu> wrote:
>In article 
><20020415144153325+0100 at pc-62-30-160-65-hw.blueyonder.co.uk>,
> Alwyn <alwyn at alwyn.demon.co.uk> wrote:
>
>> >MacPython does also have a facility for creating a standalone 
>> >application, but these are much bigger (several megabytes?) and I 
>> >haven't been able to get this to work.
>> 
>> But if the target platform is Mac OS X, you can run Python scripts from
>> the command line, as you would on any other Unix-type system. 
>
>If you've installed a *different* version of Python (Unix Python, not 
>MacPython), you can do this.  It doesn't come installed by default, and 
>I have no idea whether it can do GUI stuff (unless you also install 
>X-windows, or something).

I'm just building from the latest Python distribution as I write. As far
as I'm aware, it has Mac-specific functionality similar to that of MacPython. 


Alwyn


From g0riaman at cdf.toronto.edu  Tue Apr 23 11:00:42 2002
From: g0riaman at cdf.toronto.edu (so very tired)
Date: Tue, 23 Apr 2002 15:00:42 GMT
Subject: non-blocking sockets
Message-ID: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>

When trying to read from a socket, I don't want it hanging, so I set it to
non-blocking by calling

setblocking(0)

but then when I try to read from it and there is no data in the socket, I
get an error. I tried this in both windows and linux and in both I get an
error and the program stops. It says there's an exception but it doesn't
say the name of the exception so I can't even try to catch.
Does anyone know how I can check if there's data coming from a socket
connection without hanging or having the program barf on me?
Thanks.
mRiaz




From fperez528 at yahoo.com  Wed Apr 24 21:04:40 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 24 Apr 2002 19:04:40 -0600
Subject: An IMHO interesting optimization question...
References: <3cc73c81@nntp.server.uni-frankfurt.de>
Message-ID: <aa7kr0$eob$1@peabody.colorado.edu>

Michael 'Mickey' Lauer wrote:

> Hello,
> 
> I have an optimiziation question. Given the following code
> which generates a word length statistic up to words of the length
> <longWord>.
> 

[snip rest]

The simple rewrite below gives the following results:

[MySQL-3.23.47]> time ~/test/wordstats_ori.py < manual.texi
5.000u 0.010s 0:05.02 99.8%     0+0k 0+0io 484pf+0w
[MySQL-3.23.47]> time ~/test/wordstats_ori.py < manual.texi
4.960u 0.030s 0:05.01 99.6%     0+0k 0+0io 484pf+0w
[MySQL-3.23.47]> time ~/test/wordstats.py < manual.texi
2.420u 0.260s 0:02.70 99.2%     0+0k 0+0io 484pf+0w
[MySQL-3.23.47]> time ~/test/wordstats.py < manual.texi
2.560u 0.130s 0:02.70 99.6%     0+0k 0+0io 484pf+0w

This is running on a 2.3Mb file (the MySQL manual in texinfo format).

So these simple changes (using an array of integers, inlining count and 
forgoing the many unnecessary loops) give a factor of 2 improvement. Plus the 
code is more readable IMHO.

A simple check shows that the expensive part is the regexp stuff. So you could 
optimize the counting trivially in C using weave, but it wouldn't buy you 
much: about 80% of the execution time is spent in the regexp findall.

So it looks like if you need even faster execution, you'll have to go beyond 
python, but writing regexp code by hand is _exactly_ the kind of thing for 
which I'll kick and scream to avoid C. Just have a cup of coffee while it 
finishes :)

Cheers,

f.

ps. Don't do 'cat file | script', that's what 'script < file' is for ;)

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

def printout():
    "Formatted output of dictionary"
    for i in range( 1, longWord ):
        print "Length %d = #%d" % ( i, stats[i] )
    print "Length >%d = #%d" % ( longWord, stats[longWord] )

# import libraries and check for help request

import sys, re

if len( sys.argv ) != 1:
    print """
wordStatistic.py 1.0
--------------------
Reads from stdin and prints a word statistic.
Usage: wordStatistic.py < file
"""
    sys.exit( 0 )


defWord = "[A-Za-z]+"
wordExp = re.compile( defWord )

# initialize array, so we don't have to check for key-existance later
longWord = 15
stats = [0]*(longWord+1)

# main loop. iterates over standard input as long as there's
# something coming in...
for word in wordExp.findall( sys.stdin.read() ):
   try:
       stats[len(word)] += 1
   except IndexError:
       stats[-1] += 1

#printout()
#------------------------------------------------------------------


From narnett at mccmedia.com  Wed Apr 10 21:27:02 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 10 Apr 2002 18:27:02 -0700
Subject: Creating objects in thread created with subclass
In-Reply-To: <MPG.171e3b1bd28320e19896a9@news.nwlink.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJKEPGKAAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Jeff Shannon
> Sent: Wednesday, April 10, 2002 1:07 PM

...

> Okay, first I should warn you that I don't know SQL and haven't
> used any DB modules yet, but...  you need to dynamically *create*
> the string that you pass to execute, but that doesn't necessarily
> mean that you need to dynamically execute that string, if you get
> the difference. ;)
>
> Try something like this:
>
> mystring = "INSERT INTO Foo (fieldnames) VALUES (%s"
> mystring +=  ",%s" * (len(data)-2) + ")"
> self.dbh.execute(mystring, (data[1:]))

This approach does work... although I'm generating the '%s' series a bit
differently.  Thanks for the help.  I've never used this style of output
formatting and I'm still wrapping my brain around it.  And I was initially
confused by figuring out how to make that MySQLdb, not Python's built-in,
was doing the formatting.

And thanks to Jeff Epler, too, for more-or-less the same solution.

And other Jeffs care to chime in?

Nick





From wilk at flibuste.net  Sat Apr 27 17:24:34 2002
From: wilk at flibuste.net (William Dode)
Date: Sat, 27 Apr 2002 23:24:34 +0200
Subject: snippet source code
Message-ID: <20020427232434.13edc1ca.wilk@flibuste.net>

hi,

when i learn i make a lot of little examples for me. Now i want to share
my work, where can i put them ? I know activestate cookbook, sourceforge,
but is there something else ?

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From aseem at neurobs.com  Mon Apr  1 17:16:17 2002
From: aseem at neurobs.com (A M)
Date: 1 Apr 2002 14:16:17 -0800
Subject: Apache, Zope, SSL on Win32
Message-ID: <ed6d1c1a.0204011416.682f73e5@posting.google.com>

Hi all,

I was trying to get zope to work behind Apache on Win32. I was able to
accomplish that and am currently running ZServer behind Apache using
the first approach provided in
http://www.zope.org/Members/anser/apache_zserver
is using the Proxy... directives in httpd.conf. Since I am running the
server on an intranet the virtual host thingy is kinda messy so I
opted for the proxy option.

I do have mod_ssl enabled on my apache and would like to use Zserver
with SSL without having to do the virtual host setup thingy.

Is there a way to do that? Are there any resources on the web that can
help me to do that. I went through the apache and Zope docs but could
not find anything that would help...

Any help would be greatly appreciated. Thanks.
Aseem


From mpoulin at verinet.com  Sat Apr 20 23:50:50 2002
From: mpoulin at verinet.com (Marc Poulin)
Date: Sat, 20 Apr 2002 21:50:50 -0600
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <3cc0e552$0$186$75868355@news.frii.net>

In article <3CC0D892.2060006 at nyu.edu>, "Shane Hoversten" <srh232 at nyu.edu>
wrote:

> Hi -
> 
> 	This is really dumb, but I just want to print out a character without a
> newline and without a space behind it.  You can't do:
> 
> print c,
> 
> 	Because Python puts a space behind it.  I can't find anywhere in the
> 	dox
> that actually talk about "print" in any detail at all.  Surely there
> must be some way to do this?
> 
> Thanks,
> 
> Shane

Not dumb at all. Here is some documentation on "print".
Note the rules about when the space is printed.

http://www.python.org/doc/current/ref/print.html#l2h-330


Try writing to stdout directly.

>>> def print_a():
	sys.stdout.write("a")
	
>>> def print_b():
	sys.stdout.write("b")
	
>>> def print_ab():
	print_a()
	print_b()

>>> print_ab()
ab
>>>


From emile at fenx.com  Sun Apr 14 13:23:54 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 14 Apr 2002 17:23:54 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <Xns91EFCC39E5237JamesLLugojcom@216.148.53.84>
Message-ID: <KWiu8.34757$CH1.27941@sccrnsc02>

James Logajan
> Business people
> believe in contractual obligations and monetary incentives as their
> insurance that support will be there when they need it.

Any one who bets the future of a company on a contractual obligation to
provide future support is making a mistake.  A _guarantee_ of access to
source is now standard in contracts -- recognition that companies come
and go and fail in their contractual obligations.

--

Emile van Sebille
emile at fenx.com

---------



From jkraska at san.rr.com  Wed Apr  3 20:37:03 2002
From: jkraska at san.rr.com (Courageous)
Date: Thu, 04 Apr 2002 01:37:03 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <32cnausana9835ava9ar07e7b6c4isft0k@4ax.com>

>I need to translate a (rather voluminous 100MB) java program to python. After a
>few searches on google, apparently the best tools for this task are Lex-Yacc
>(Flex-Bison on linux). While I am working on the Lex-YACC HOWTO, I would be very
>interested in hearing of others' successful experiences/feedback at this. Any
>pointers of interest would similarily be appreciated.

I *highly* recommend ANTLR. http://www.antlr.org. It has built in support
for translators (i.e., it's not a pain in the ass to copy over comments
which you often _do_ want).

C//



From sholden at holdenweb.com  Mon Apr  1 07:15:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 07:15:31 -0500
Subject: Root access with python
References: <d906b2c1.0203250237.60f6ad12@posting.google.com> <7x663krjtp.fsf@ruckus.brouhaha.com> <a7ndu9$bme$1@panix1.panix.com> <20020325123855.A1716@ibook.distro.conectiva> <mailman.1017071438.28406.python-list@python.org> <Ou4o8.17920$Ii4.1702@atlpnn01.usenetserver.com> <a88h4i$1v8n$1@news.idiom.com>
Message-ID: <AcYp8.107080$Ou.70390@atlpnn01.usenetserver.com>

"Jim Dennis" <jimd at vega.starshine.org> wrote in message
news:a88h4i$1v8n$1 at news.idiom.com...
> In article <Ou4o8.17920$Ii4.1702 at atlpnn01.usenetserver.com>, Steve Holden
wrote:
[...]
>
> > As far as setuid scripts goes, just say no. There are far too many
security
> > holes in almost every Unix to think that it's safe to use setuid
scripts.
>
[...]
>   As for your overall "advice."  It's of little value to tell
>   someone: "don't do that" without offering any suggestion about
>   what he or she SHOULD try.
>
>   Whenever discussing computer system security I avoid
>   proscriptive pronouncements like: "you can't do that" or
>   "no one should ..."
>
[...]
>   When we say, "you can't do that," we're imposing requirements on the
>   OP without regard for his or her other requirements and without
>   regard for any risk assessment or cost/benefit analysis that he or
>   she may have done.
>
I'm not imposing any requirements on anybody, this is Usenet, remember.

Any advice I offer on Usenet is offered on the basis that the recipient
knows more about their situation than I do, and they are free to ignore such
advice as does not suit them.

>   Unfortunately I don't know how to make SUID python scripts.  In my
>   current ignorance I'd have to recommend sperl with "taint checking,"
>   or a sudoers entry, or a small SUID C wrapper program.
>
So, now you've told me off, what *should* the OP do?

regards
 Steve






From aahz at pythoncraft.com  Tue Apr  2 09:56:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Apr 2002 09:56:45 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <mailman.1017752105.30270.python-list@python.org>
Message-ID: <a8cgrd$q0r$1@panix1.panix.com>

In article <mailman.1017752105.30270.python-list at python.org>,
Rich Harkins  <rich at worldsinfinite.com> wrote:
>
>Hmmm... this is slightly OT for the thread but I have made a bit of a habit 
>of doing the following:
>
>if result is None:		# As opposed to result == None
>	stuff
>
>This is somewhat out of paranoia, and slightly out of interest in speed (my 
>recollection is that previous Python versions could end up calling __coerce__ 
>on comparisons with None and in many of these cases I didn't really want 
>that).  Is there any compelling reason not to do this?  Isn't None to be a 
>singleton in perpetuity?

Don't worry; you're doing the right thing.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From spammers.do.not.bother at void.bogus  Mon Apr 29 13:32:26 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Mon, 29 Apr 2002 17:32:26 GMT
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no>
Message-ID: <Ksfz8.34667$n4.7298632@newsc.telia.net>

Johan Fredrik ?hman wrote:

> Hi, there.
> 
> I previously saw a posing named "Python vs Php" and sys atht
> "... there are many packages you can download..."
> 
> However, only one package is mentioned: jonpy
> 
> What is the most popular and promising webpackages for python ?
> 
> --
> JF?


BTW: I friend talked about Twisted the other day. I haven't tried it 
personally but maybe I should?

Has anyone tried it and can share some experiences?

http://www.twistedmatrix.com/

Regards,
Magnus



From jb at cascade-sys.com  Thu Apr 25 14:13:29 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 11:13:29 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com>
Message-ID: <3CC84749.1AA2A087@cascade-sys.com>

Christian Tismer wrote:

> There is some nice introduction to Limbo at
> http://www.vitanuova.com/inferno/papers/descent.html
> written by Brian W. Kernighan, one of the inventors
> of the C language.

Just a nit:

Dennis Ritchie invented C.  http://cm.bell-labs.com/cm/cs/who/dmr/chist.html

Brian helped invent the AWK language, a precursor to Python.  AWK stands for
Aho Weinberger and Kernighan.  Brian wrote a lot of books about C but he was
not one of the inventors.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 17:23:58 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 23:23:58 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE> <Xns91F2ED00B6997cliechtigmxnet@62.2.16.82>
Message-ID: <a9i4q5$ef2$1@nets3.rz.RWTH-Aachen.DE>

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns91F2ED00B6997cliechtigmxnet at 62.2.16.82...
>..
> have a look at the rexec module
>..
> you can disable the builtin functions you want and populate the namespace
> with save functions and classes (see Bastion for that)
>..

(Thanks to you and everybody for the great help!)




From jeff at ccvcorp.com  Fri Apr 19 14:28:41 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 11:28:41 -0700
Subject: European Win32. Python weirdness
References: <3CC03738.2725CC26@atlsci.com>
Message-ID: <MPG.172a01af1e490109896b9@news.nwlink.com>

In article <3CC03738.2725CC26 at atlsci.com>, aboufer at atlsci.com 
says...
> Greetings all,
> 
> I am having problems with a module of python, mainly on Win32 platform
> using European languages. In some of the code we wrote, python does not
> seem to recognise a
> number as a number. More specificcaly, when python reads a string of
> digits, 'he' does not
> recognise it as a number but as a string. This of course causes the
> reste of the processing to fail!

It's hard to be sure without seeing your code, but this is 
probably normal/expected behavior.  The raw_input() function, 
which gets user input, will always return a string.  If a 
numerical value is expected, then it should be converted with a 
call to int() or float().  Similarly, reading and writing to 
files is normally done with strings, and can be converted to 
numbers in the same way.  There are other ways to get numerical 
input, but generally reading it as a string and then converting 
to the expected/desired number type is the safest way to go.

If you post a few segments of your code that are exhibiting this 
issue, then we can perhaps offer more specific suggestions.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From sholden at holdenweb.com  Tue Apr 23 12:37:47 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 12:37:47 -0400
Subject: Python & autoftp
References: <53b7d218.0204230720.686e473f@posting.google.com>
Message-ID: <M7gx8.96698$T%5.24889@atlpnn01.usenetserver.com>

"kevin" <davies at jkevin.freeserve.co.uk> wrote ...
> Hi there
>
> I'm pretty new to Python and I would like to develop a script in
> Python to automatically FTP. I have already used bash and expect and
> now would like to replace bash entirely with Python.
>
> Is there any merit in replacing bash with Python??
> Can anyone point me in the right direction.
>

You might take a look at

    http://www.holdenweb.com/Python/PDCode/ftpStream.py

for an example of how to use the FTP module.

regards
 Steve
--
home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/






From martin at v.loewis.de  Mon Apr 15 02:39:12 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:39:12 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <m3662uy7h8.fsf@mira.informatik.hu-berlin.de> <slrnabjgvi.741.philh@comuno.freeserve.co.uk>
Message-ID: <m3r8lhlb3j.fsf@mira.informatik.hu-berlin.de>

philh at comuno.freeserve.co.uk (phil hunt) writes:

> And what if one is bool and the other is an int?

>>> True ^ 2
3

Regards,
Martin


From JamesL at Lugoj.com  Thu Apr  4 19:55:39 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 5 Apr 2002 00:55:39 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com>
Message-ID: <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96>

Eri Ma Franci <max at alcyone.com> wrote:
> Could you find something better to complain about?

Then I'd like to complain that neither you nor Guid va Rossu addressed the 
question I asked in my previous message concerning bools as keys. If a bool 
is to be treated as an int sometimes and sometimes not, I don't recall 
seeing anything about whether it will be treated as an int as a dictionary 
key. If I do:

x = ["is", "are"]
y = {0: "", 1:"s"}

for n in range(3):
   print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])

Will the behavior be backword compatible and equivalent or not? Simple 
question I assume; waiting for an answer.


From loewis at informatik.hu-berlin.de  Mon Apr 29 07:44:50 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 29 Apr 2002 13:44:50 +0200
Subject: expat and entity references
References: <slrnacp5lb.b5i.clgonsal@keeshah.penguinpowered.com>
Message-ID: <j4g01eenkd.fsf@informatik.hu-berlin.de>

clgonsal at alumni.uwaterloo.invalid (C. Laurence Gonsalves) writes:

> I'm working on an Python application that uses xml.parsers.expat to
> parse XML files.  These XML files are generally hand-written. I'd like
> to be able support more than just amp, lt, gt, apos and quot.
> 
> What's the best way to do this without requiring that every XML file
> contain a huge header declaring all of the entities I want to support? 

If you are using expat directly, you need to implement an
ExternalEntityRef handler. In that handler, you can either put the
entity definition right away into your processing results, or you need
to call ExternalEntityParserCreate to create a "nested parser" - the
latter is necessary if the external entity may contain markup.

If you are using SAX, you need to set an EntityResolver with the
parser, and return an InputSource in response to the resolveEntity
callback.

If you are using minidom.parse, you need to first create an
appropriately configured SAX parser, and pass that to build the
minidom tree.

HTH,
Martin


From duncan at NOSPAMrcp.co.uk  Tue Apr  9 04:16:22 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 9 Apr 2002 08:16:22 +0000 (UTC)
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <Xns91EA7D4BE5CABduncanrcpcouk@127.0.0.1> <1018303417.472398@cswreg.cos.agilent.com>
Message-ID: <Xns91EB5D9648363duncanrcpcouk@127.0.0.1>

weeks at vitus.scs.agilent.com (Greg Weeks) wrote in
news:1018303417.472398 at cswreg.cos.agilent.com: 

> I don't have the documentation or a Python 2.1 handy.  But I did
> originally copy from the documentation the text:
> 
>     Names in class scope are not accessible.
> 
>     Under this proposal, it will not be possible to gain
>     dictionary-style access to all visible scopes.
> 
> My interpretation of this is that visible bindings from an
> intermediate scope will be in neither globals() nor locals().

I must admit I didn't read the documentation, I just checked it out by 
defining a function that accessed a nested scope variable and printed out 
locals() and globals(). This was under 2.2 BTW.
I don't see why locals() shouldn't include the nested scope variables, 
after all they are all known at compile time just like local vars.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From tejarex at yahoo.com  Sat Apr  6 11:34:33 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 06 Apr 2002 16:34:33 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com> <mailman.1018074974.20211.python-list@python.org>
Message-ID: <tsFr8.13642$l7.1474145@bin7.nnrp.aus1.giganews.com>

"Tim Churches" <tchur at optushome.com.au> wrote in message
news:mailman.1018074974.20211.python-list at python.org...

> I sometimes use a statistics package called Stata. It has been
around  for
> well over a decade, with major upgrades to the core language
released
> about every two years or so - hence there are three or four versions
in common
> use.

Stata charges several hundred dollars for its statistics package.
Hence, it can pay a bunch of programmers to work behind closed doors.
I suspect that there are several internal releases during the
development and test period.  Of course, this may mean that outsiders
have to wait for as much as two years to get a new feature after it is
finalized.

Would you prefer that Python developers went behind closed doors?
Should they have kept 2.1, 2.2 private?  Should they keep 2.3, etc
private until ready to release 3.0 maybe two years from now?  And not
get the benefit of free user feedback?  -- which commercial companies
pay for when they want to keep it secret.

> In order to allow users who are still running Stata V4
> to write routines which will behave correctly in Stata V7 (and vice
versa),
> Stata has a "version" statement. If I am running Version 7, I can
put something like
> "USE VERSION 4" (I don't recall the exact syntax) at the top of my
programs
> and the program will be executed as if it were running in Stata V4.

Stata users PAY to get this sort of maintenance of the past.  Perhaps
there is a business opportunity for a Python entrepreneur.  Perhaps
not.

Terry J. Reedy





From erno-news at erno.iki.fi  Mon Apr 29 14:07:42 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 29 Apr 2002 21:07:42 +0300
Subject: CGI and client's IP?
References: <1a6qcuc92l8ac0tcegqj95fqemdctda21v@4ax.com>
Message-ID: <ku4rhuz8cx.fsf@lasipalatsi.fi>

In article <1a6qcuc92l8ac0tcegqj95fqemdctda21v at 4ax.com>, Fernando
Rodr?guez <frr at easyjob.net> writes:

| Can I get the client's IP froma a cgi script? O:-)

you can find a list of environment variables defined by the cgi
protocol at <URL: http://hoohoo.ncsa.uiuc.edu/cgi/env.html>.
the client ip address is one of them.

  -- erno


From fperez528 at yahoo.com  Tue Apr  9 18:00:42 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 09 Apr 2002 16:00:42 -0600
Subject: shutil.copytree - wishlist
References: <slrnab6nf1.1du.msoulier@pmerd071.ca.nortel.com>
Message-ID: <a8vobb$rmt$1@peabody.colorado.edu>

Michael P. Soulier wrote:


> Optionally, it would be nice to have C -> B.
> 
>     Is there an official way to make such a wishlist request?

I think you can file it as a bug report on SF. Those tend to be better 
received when accompanied by code implementing the desired feature :)

Sounds like a good thing, though.

Cheers,

f.


From philh at comuno.freeserve.co.uk  Tue Apr 16 18:20:03 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:20:03 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <edAu8.10842$Ji7.6909@atlpnn01.usenetserver.com>
Message-ID: <slrnabp8sj.ipr.philh@comuno.freeserve.co.uk>

On Mon, 15 Apr 2002 09:01:56 -0400, Steve Holden <sholden at holdenweb.com> wrote:
>"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
>> On Sun, 14 Apr 2002 01:02:17 -0400, Tim Peters <tim.one at comcast.net>
>wrote:
>> >> ...
>[ talk on leaving things out left out :-)]
>>
>> It is not new features that are the problem, it is new features that
>> break existing code.
>>
>And these would be ... ?

PEP 285, for example.

>> IMO until Python developers adopt a policy of being very reluctant
>> to do this, many IT managers will be wary of using it in anything
>> other thsan toy projects, and Python's user base will not grow as
>> quickly as (IMO) it should.
>>
>And your evidence that they have not yet adopted such a policy is ... ?

Gut feeling.

>> In summary:
>>
>>    New features => good
>>
>>    New features that break existing code => bad
>>
>While this last assertion is uncontentious (though hardly surprising or
>original), you have totally failed to provide any real evidence of code
>breakage due to the recent changes in the language. On the other hand,
>Andrew Kuchling recently reported on this list (in message
><slrnabba3h.h8h.akuchlin at crystal.mems-exchange.org>) that it took *minimal
>effort* to port over 170,000 lines of Python code to 2.2.

IOW, some code did break? And how much wore will break when people 
try to serialize code on a post-PEP285 python and read it on a 
pre-PEP285?

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From salmonia at cf.ac.uk  Wed Apr 10 08:04:32 2002
From: salmonia at cf.ac.uk (Alan James Salmoni)
Date: Wed, 10 Apr 2002 13:04:32 +0100
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de>
Message-ID: <3CB42A50.8020200@cf.ac.uk>

Thomas Guettler wrote:
> Up to now I was happy with emacs because I never used
> to programm it with lisp. Now I need to and find it difficult.
> 
> Is there an emacs-like editor which is scriptable with python?
> 
>   thomas
> 

There is also CoolEdit - it is small, has a snazzy interface, and can be 
custom programmed using Python (all extensions are available). I prefer 
Emacs myself, but I haven't needed to customise it. Go to: 
http://cooledit.sourceforge.net/ for more info etc.

Al.



From gh_pythonlist at gmx.de  Wed Apr 17 19:17:26 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 18 Apr 2002 01:17:26 +0200
Subject: Mailman lists at Starship
Message-ID: <20020417231726.GA761@lilith.my-fqdn.de>

Does anybody know who is the technial support contact ;-) for these?

I'm the list admin of
http://starship.python.net/mailman/listinfo/python-de
and I'm having some problems.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 8.0 ?C      Wind: 1.4 m/s




From just at xs4all.nl  Sun Apr 14 04:23:13 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sun, 14 Apr 2002 10:23:13 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <roy-0BE2A8.19181913042002@news1.panix.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de> <Prau8.17347$vF6.558738@news2.tin.it>
Message-ID: <just-A67336.10231314042002@news1.xs4all.nl>

In article <Prau8.17347$vF6.558738 at news2.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> [ ... ]  Indeed, given the small amount of bit-twiddling in
> typical Python usage, I think & ~ | ^ are already overboard -- they
> make a lot of sense in C, a very low-level language that is used a
> LOT for bit-twiddling, but, in Python, named built-in functions might
> have been more appropriate in hindsight (fewer precedence levels to
> remember) -- if not, even, functions in a separate module.

Although I agree with the bit-twiddling argument, I find that 
overloading these operators is *very* useful to get a compact (and 
readable) spelling for set-like operations (union, section, etc.).

Just


From elian at inbrief.net  Mon Apr 22 15:11:23 2002
From: elian at inbrief.net (Edmund Lian)
Date: Mon, 22 Apr 2002 15:11:23 -0400
Subject: Overriding methods in classes that use property()
Message-ID: <OFE451B18E.B91E6997-ON85256BA3.00693F99@juscom.com>

Thomas wrote:

>You can achive what you want with a metaclass, see code below.
>Every time a new subclass of A is created, the __metaclass__
>__init__ method is called, which retrieves the correct
>getVal and setVal methods, and creates a property from them:

Hmmm... interesting. OK, I'll check it out, thanks!

...Edmund.





From aahz at pythoncraft.com  Wed Apr 17 13:28:33 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Apr 2002 13:28:33 -0400
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com> <mailman.1019061131.9382.python-list@python.org>
Message-ID: <a9kbc1$oku$1@panix1.panix.com>

In article <mailman.1019061131.9382.python-list at python.org>,
Skip Montanaro  <skip at pobox.com> wrote:
>
>    Greg> There should have been an operator in Python that says when
>    Greg> objects are conceptually the same.  For the moment, call it "eq".
>    Greg> The following would hold:
>
>    Greg>     3 eq 3
>    Greg>     not (3 eq 4)
>    Greg>     a eq a            # PROVIDED a HAS A VALUE
>    Greg>     "foo" eq "foo"
>    Greg>     not ([] eq [])
>    Greg>     not ({} eq {})
>    Greg>     () eq ()
>
>Guido's time machine strikes again:
>
>    >>> 3 is 3
>    True
>    >>> not (3 is 4)
>    True
>    >>> a = "sdfsdf"
>    >>> a is a
>    True
>    >>> "foo" is "foo"
>    True
>    >>> not ([] is [])
>    True
>    >>> not ({} is {})
>    True
>    >>> () is ()
>    True

As Greg pointed out, the first one and the last four are implementation
dependent.  Do *NOT* rely on them.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From phr-n2002a at nightsong.com  Mon Apr 15 19:10:54 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 16:10:54 -0700
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com>
Message-ID: <7xu1qcmubl.fsf@ruckus.brouhaha.com>

huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) writes:
> In any case, if I now take a poll on this list for the following question:
> 
> Do you think the set of "open source software" as defined by OSI and the set
> of "free software" as defined by FSF are 
> 1) the same set,
> 2) two different sets,
> 3) none of the above;
> 
> I'd expect a dozen different answers.  Oh, well.  :-)

The two sets of programs are basically the same (there are some minor
disjunctions)--the difference is in the philosophy expressed by the
two terms.  See the FSF article linked earlier for a discussion of the
difference.


From aahz at pythoncraft.com  Wed Apr  3 18:59:27 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 18:59:27 -0500
Subject: Threading problem
References: <mailman.1017868749.1550.python-list@python.org>
Message-ID: <a8g50v$ko8$1@panix1.panix.com>

In article <mailman.1017868749.1550.python-list at python.org>,
Gustavo Cordova  <gcordova at hebmex.com> wrote:
>
>So, a solution would be to make an extension that...
>
>1. Copies all pertinent data from python-space into it's own memory.
>2. Releases the GIL.
>3. Does all the work that needs to be done only on it's private memory.
>4. Recapture the GIL.
>5. Copy back results into python-space.

Yup, exactly.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From huaiyu at gauss.almadan.ibm.com  Mon Apr 15 13:30:13 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Mon, 15 Apr 2002 17:30:13 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1018662715.17645.python-list@python.org>
Message-ID: <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com>

Tim Peters <tim.one at comcast.net> wrote:
>
>That's not.  <http://www.opensource.org/> uses nothing else, although it's
>often spelled Open Source there.  From OSI's FAQ:
>
>    How is "open source" related to "free software"?
>
>    The Open Source Initiative is a marketing program for free software.
>    It's a pitch for "free software" on solid pragmatic grounds rather
>    than ideological tub-thumping.  The winning substance has not
>    changed, the losing attitude and symbolism have.
>
>As marketing collateral, I suppose that paragraph could stand its own
>attitude adjusmment <wink>.

Thanks Tim.  I guess I need re-education again.  By now I'm pretty much sure
that these things are not as they were several years ago when the term "open
source" first came about.  I was excited about the new term which seemed
easier to explain than "free", but was promptly educated by both camps that
they were not the same.

In any case, if I now take a poll on this list for the following question:

Do you think the set of "open source software" as defined by OSI and the set
of "free software" as defined by FSF are 
1) the same set,
2) two different sets,
3) none of the above;

I'd expect a dozen different answers.  Oh, well.  :-)

Huaiyu


From logiplexsoftware at earthlink.net  Mon Apr 29 16:01:36 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 13:01:36 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCD9649.98AED86F@cascade-sys.com>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
Message-ID: <20020429130136.5b0c34f7.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 11:51:53 -0700
James J. Besemer wrote:

> 
> Brian Quinlan wrote:
> 
> > > Huh? The "core" language wouldn't be measurably shrunk by this
> > proposal
> > > fella. Your rationale is built on air. Show us some genuine
> > _advantage_
> > > that is worth the huge _cost_ in time and effort caused by your ill-
> > > considered proposal. How about you and other PEP writers take some
> > > responsibilities for your fantasies for once? You up to helping
> > thousands
> > > rewrite their code? I thought not.
> >
> > Why the hostility?
> 
> I don't interpret this strongly worded and slightly sarcastic critique to
> be hostile.
> 
> If you call this hostile I don't think you know the meaning of the word.

It was definitely hostile.  I've seen arguments that take this route end in
physical violence.

> > I think that every PEP is a good one. Even if the PEP is ultimately
> > rejected, it promotes thought about some aspect of the language. Taking
> > responsibility sounds like stifling discussion to me.
> 
> And I don't see how his offering a strong opinion stiffles discussion.

Personal attacks are counterproductive in at least two ways:
1. They take the focus away from the actual topic being argued.  The fact that
we are discussing it now should demonstrate that. 
2. They tie the argument to the person making it.  It will be difficult for
that person to ever concede the argument as doing so would also seem to concede
the validity of the personal attacks.  It becomes a grudge match rather than a
real discussion.

> I happen to agree with just about everything he said.
> 
> This PEP is angles in the clouds stuff.  A lot of work for everybody and
> virtually zero practical benefit to anybody.

Well that's a stupid statement - *your* head is in the clouds.  

See what I mean?  Gets us nowhere. ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From s1500 at sihope.nospam.com  Sun Apr  7 19:18:50 2002
From: s1500 at sihope.nospam.com (Spitfire 1500)
Date: Sun, 7 Apr 2002 18:18:50 -0500
Subject: Any reason why PythonWin wouldn't start at all?
Message-ID: <3cb0d423$1@news.sihope.com>

I'm running Windows 2K, and each time i try to start pythonwin, absolutley
nothing happens. It shows it as running on the task manager, but I can't get
to it whatsoever.

I've reinstalled it several times, killed the reg keys, but nothing seems to
do it. It runs just fine on my laptop, which runs WIndows 98.

With the numerous problems of the freezing IDLE, I'd really like to use
Pythonwin instead. Anyone?

--
Kids love the rich taste of web content!
http://british.nerp.net




From logiplexsoftware at earthlink.net  Mon Apr  1 16:10:42 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 13:10:42 -0800
Subject: wxPython  / wxSingleChoiceDialog ...
In-Reply-To: <a8ahju$8dmek$2@hades.rz.uni-sb.de>
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de>
	<slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr>
	<a89vu0$8bm2c$1@hades.rz.uni-sb.de>
	<mailman.1017686284.32405.python-list@python.org>
	<a8ahju$8dmek$2@hades.rz.uni-sb.de>
Message-ID: <20020401131042.289b7f5a.logiplexsoftware@earthlink.net>

On 1 Apr 2002 20:57:34 GMT
Uwe Schmitt wrote:
> o.k  i used that and found a wxListBox as children. How do I now
> manipulate the font of the entries of the ListBox ???

Assuming you have a reference to the wxListBox at this point, you should be
able to simply use listbox.SetFont().  Does that not work?


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From greg at cosc.canterbury.ac.nz  Thu Apr  4 01:52:37 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 04 Apr 2002 18:52:37 +1200
Subject: ANN: Pyrex - a language for writing Python extension modules
Message-ID: <mailman.1017952983.647.clpa-moderators@python.org>

Folks here may be interested in my latest
project: A new language specially designed for
writing Python extension modules.

Pyrex is more or less Python with C data types
added. You can write functions which freely
intermix operations on Python and C data, with
all Python reference counting and error checking
completely automated.

An example Pyrex module is shown below. For more
information, see:

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

where you can also download version 0.1 and try it
out. I'm very keen to get feedback on this, both to
find out whether people think it's useful, and to
help me debug it!

Oh, yeah, here's the example. It finds prime
numbers at C speed.

def primes(int kmax):
  cdef int n, k, i
  cdef int p[1000]
  result = []
  if kmax > 1000:
    kmax = 1000
  k = 0
  n = 2
  while k < kmax:
    i = 0
    while i < k and n % p[i] <> 0:
      i = i + 1
    if i == k:
      p[k] = n
      k = k + 1
      result.append(n)
    n = n + 1
  return result

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




From m.faassen at vet.uu.nl  Mon Apr 15 15:58:43 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Apr 2002 19:58:43 GMT
Subject: Calling All PSU Members
Message-ID: <a9fbdj$5uq$1@newshost.accu.uu.nl>

Secret Secret Secret -- For Python Secret Underground Members ONLY.
Do NOT distribute outside the PSU. You will be eliminated from the
space-time continuum entirely. That is not an idle threat. 

As stated in previous dispatches, flame wars can be used to extend the
maximum age of the universe. Subsequently PSU members have been dilligently
adding to such discussions. In particular, PEP 285 (Adding a bool type),
has been a point of contention. This has resulted in a change of the
cosmological constant, protonic decay rate and the spawning quotient of
quantum child universes. The cumulative effect lengthened the future 
by approximately 534.731 billion years.

It also leads to a catastrophic war in the year 285,285 (Common
Human Reckoning).

A theory first put forward by Arthumus Rex of the Reconstituted Terran
Dawn Federation in 8125 CHR explains that abrupt changes to universal 
constants affect blue quantum spin torsion. This and other warnings
are not heeded, leading to an acute scarcity of FUD in the 285th millennium.
In 285,285 CHR this results in the Boolean Wars, the conflict that puts all
others in the shade. In it, a constantly shifting mesh of alliances between 
four parties battle each other to a standstill and beyond, risking an
ultimate continuum breach. These are the Nothing Union, the League of 
Falsehood, the Protoknights of Truth and the entity simply known as
Existence.

This is exactly what the PSU desires. Unfortunately the PSU of eras 
post 285,285 does not, as it will not exist anymore. This vile and
evil PSU is therefore threatening us with Boolean Re-Negation. We do not
know what that is, but it sounds ominous. We are therefore ordering all
members that, in the spirit of friendly cooperation and not at all due to
these threats, they moderate their discussions in the future.


From weeks at vitus.scs.agilent.com  Mon Apr  8 15:33:36 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Mon, 8 Apr 2002 19:33:36 +0000 (UTC)
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <Xns91EA7D4BE5CABduncanrcpcouk@127.0.0.1>
Message-ID: <1018303417.472398@cswreg.cos.agilent.com>

Duncan Booth (duncan at NOSPAMrcp.co.uk) wrote:
: weeks at vitus.scs.agilent.com (Greg Weeks) wrote in
: news:1018044224.815568 at cswreg.cos.agilent.com: 

: > Anyway, it increased visibility enough so that it is no longer
: > possible to call a function and pass it the scope as two dictionaries,
: > which I had found handy on occasion.

: I don't think this is correct.
: If your function accesses variables in a nested scope, then those variables 
: are present in the dictionary locals().

I don't have the documentation or a Python 2.1 handy.  But I did originally
copy from the documentation the text:

    Names in class scope are not accessible.

    Under this proposal, it will not be possible to gain
    dictionary-style access to all visible scopes.

My interpretation of this is that visible bindings from an intermediate
scope will be in neither globals() nor locals().


Greg


From aahz at pythoncraft.com  Mon Apr 22 16:08:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 16:08:16 -0400
Subject: Checking if class is derived of some other class
References: <pan.2002.04.22.19.11.59.254871.5745@teich.garten.DigitalProjects.com> <pan.2002.04.22.19.43.59.567352.5745@teich.garten.DigitalProjects.com>
Message-ID: <aa1qjg$g16$1@panix1.panix.com>

In article <pan.2002.04.22.19.43.59.567352.5745 at teich.garten.DigitalProjects.com>,
Alexander Skwar  <news.ASkwar at DigitalProjects.com> wrote:
>
>class a:
>	pass
>
>class b(a):
>	pass
>
>class c(b):
>	pass
>
>myObject = c
>
>if isinstance(myObject, c):
>	print "yes"
>else:
>	print "No"

Try

myObject = c()
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From infocap at wanadoo.fr  Mon Apr 29 14:01:11 2002
From: infocap at wanadoo.fr (infocap)
Date: Mon, 29 Apr 2002 20:01:11 +0200
Subject: Python sous UNIX
Message-ID: <aak1q3$fgl$1@wanadoo.fr>

Bonjour ? tous,

Je travaille pour une entreprise d'?dition mais voil? que depuis quelques
temps je m'int?resse beaucoup ? l'informatique et particuli?rement au
langage de programmation PYTHON  sous UNIX et j'aimerai approfondir mes
connaissances au travers de petits programmes int?ressants, alors, si
quelqu'un peut satisfaire ma demande, je le remercie d'avance.

Cordialement
David



--





From eppstein at ics.uci.edu  Sat Apr 20 02:26:52 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 19 Apr 2002 23:26:52 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org>
Message-ID: <eppstein-2C56E0.23265219042002@news.service.uci.edu>

In article <mailman.1019279977.28194.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> > I'm really delighted to hear about performance improvements in file
> > object input in recent Python versions (in a message that would
> > have been in this thread if Tim Peters used a newsreader instead
> > posting to the list.)
> 
> Hmm.  Shows up fine in the newsreader I use -- this, like hearing about the
> speedups for the first time from that msg, may be another symptom of General
> Upgrade Fear <0.9 wink>.

The point is that your responses don't have the appropriate References: 
header, so they don't get threaded with other messages on the same topic.
It appears as if you are using a mail agent rather than a news agent, 
since References: is a news header (and your other headers say you're 
using MS Outlook).

It affects other things than threading, too -- e.g. I use References: to 
pick out responses to my own messages, and yours don't get caught.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From st-newsgroups at thomas-guettler.de  Thu Apr 11 03:55:11 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Thu, 11 Apr 2002 09:55:11 +0200
Subject: Create Excel Files without COM
References: <mailman.1018449081.32584.python-list@python.org>
Message-ID: <3CB5415F.9050207@thomas-guettler.de>

James Strater wrote:

> Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::
> 
> 
>>Hi!
>>
>>Is there something like WriteExcel from perl in python?
>>I need to generate an MS-Excel file. I don't want to use
>>COM although it will run on windows.
>>
>>The data is simple. Just text in columns and rows.
>>
> 
> 
> You can format your output as HTML, and then put a .xls extension on it.
> When it is opened in excel, it will be fine.  To see what your input needs
> to look like, take your spreadsheet and 'save as' HTML.


Thank you!

At least Excel 2002 and 2000 can open the created HTML.
I pass it to the webbrowser with a content-type like this:
REQUEST.RESPONSE.setHeader('Content-Type',
    'application/vnd.ms-excel')
and Excel opens it.

  Fine!

thomas






From cliechti at gmx.net  Wed Apr  3 14:33:03 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 20:33:03 +0100
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017857237.27591.python-list@python.org>
Message-ID: <Xns91E5D13ADC09cliechtigmxnet@62.2.16.82>

Ka-Ping Yee <ping at lfw.org> wrote in news:mailman.1017857237.27591.python-
list at python.org:
> [*] By real booleans, i mean the following.  (Booleans would have
>     to behave like this for me to consider them "good enough" to
>     be better than what we have now.)
> 
>     >>> False, repr(False), str(False)
>     (False, 'False', 'False')
>     >>> True, repr(False), str(False)
>     (True, 'False', 'False')
>     >>> False + True
>     TypeError...

or better "True" (0+1==1 --> True)

>     >>> False == None
>     0
>     >>> False == 0
>     0
>     >>> True == 1
>     0

i disagree. comparing True/False with other values should give the same 
result as when it were used with an 'if':
>>> True == 1, True == 6
True, True
>>> False == 0, Flase == [], Flase == 1
True, True, False

because 0 and [] etc. are representing Flasehood they should also be equal 
to False. 'False is []' etc. will allways be False except for "False is 
False".


>     >>> {0: 0, False: False, 1: 1, True: True}
>     {0: 0, False: False, 1: 1, True: True}
> 
> ... and probably
> 
>     >>> None < False < True < 0
>     True

None has no value in a nummerical sense an the docs clearly say that you 
shouldn't use it that way...

False and True should be represented as 0 and 1 in a nummerical usage.
 
> (Hee hee -- i suppose the fact that "boolean" starts with a "b"
> gets us this for free.  But i wonder how many people are going
> to be puzzled by True < 0?)


-- 
Chris <cliechti at gmx.net>



From jkn at nicorp.f9.co.uk  Tue Apr  2 11:54:24 2002
From: jkn at nicorp.f9.co.uk (Jon Nicoll)
Date: 2 Apr 2002 08:54:24 -0800
Subject: wxPython redefines string.letters?
Message-ID: <8351bb33.0204020854.272eaa7b@posting.google.com>

I realise this should go to the wxPython list, and after I've joined
it will do ;-), but in the meantime:

>>> import string
>>> string.letters
'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
>>> from wxPython.wx import *
>>> string.letters
'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\x83\x8a\x8c\x8e\x9a\x9c\x9
e\x9f\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd
2\xd3\xd4\xd5\xd6\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe
7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf8\xf9\xfa\xfb\xf
c\xfd\xfe\xff'

?

    Jon N


From mark at mceahern.com  Mon Apr 15 15:04:41 2002
From: mark at mceahern.com (Mark McEahern)
Date: Mon, 15 Apr 2002 14:04:41 -0500
Subject: slight ot: python like string interpolation with j*script
Message-ID: <NCBBLFCOHHDIKCAFGCFBMECPKLAA.mark@mceahern.com>

Hi, I was messing around with JScript this weekend and found myself missing
Python.  One of the many things I missed was string formatting.  You know:

  s = "The expected string %s did not match the generated string %s."
  print s % (expected, generated)

That sort of thing.

I got to wondering, could you do that with JScript?  I figured maybe I
wasn't reading the docs closely enough.  Well, it turns out that there's no
native support for it in JScript, but you can get something quite close with
regular expressions.  I'd be interested in any feedback on the following
from fellow Pythonistas, especially those who sometimes dabble in JScript.

Cheers,

// mark

<%@ LANGUAGE="JScript" %>
/*
** Written to be run from a ASP web page, but it wouldn't be too hard to
extract
** this for running it, say, in a WScript environment or client-side
JavaScript.
*/

<%

function debugWrite(s)
{
  Response.Write("<p>" + s + "</p>")
}

// Who needs Scripting.Dictionary, when this is such a piece of cake?
dict = new Object()
dict.expected = "foo"
dict.generated = "bar"

// Build this functionality into String, why don't ya?
String.prototype.formatWith = function(dict)
{
  var re = /\$(\w+)/gi
  var results
  var temp = new String(this.valueOf())
  while((results = re.exec(this)) != null)
  {
    temp = temp.replace(results[0], dict[results[1]])
  }
  return temp
}

s = "The $expected string does not match the $generated string."
t = s.formatWith(dict)
debugWrite(t)

%>





From jwickard**nospam** at litriusgroup.com  Mon Apr  1 15:35:46 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Mon, 01 Apr 2002 20:35:46 GMT
Subject: [OT] Developers Living/Working in France
Message-ID: <Cw3q8.1270$M.436350@typhoon.mn.ipsvc.net>

My wife is interested in a student exchange program which would move her to
France.  I'd like to talk with a few developers living/working in France to
see what the job and realestate market are like.  Any one interested in an
information exchange, or telling me where to find more information can reply
to me off-list at: joellist at litriusgroup.com

Thank you.




From starx at pacbell.net  Mon Apr  8 00:10:41 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 04:10:41 GMT
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4> <3CB0CD74.B1696987@engcorp.com>
Message-ID: <Xns91E9D79CB4AEBRASXnewsDFE1@130.133.1.4>

Peter Hansen || Sun 07 Apr 2002 03:51:32p:

>> >>I'm using the time module to make part of the game random and
> [...]
>> >
>> > Use the random module.
>> 
>> Yep. For example:
>> 
>> ---
>> import random
>> 
>> # No vaule - uses time on it's own
>> random.seed()
> 
> Random seeds itself using time.time() so this is unnecessary 
> in simple code.

Ah, yes, I didn't realize -- i thought the standard way of using a 
randomizer with a constant seed (for debug), was to just not call the seed 
function -- i may be wrong. I guess in python you'd just call 'seed
(some_number_i_like)' to do this :).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From p.magwene at snet.net  Thu Apr 25 13:15:00 2002
From: p.magwene at snet.net (Paul Magwene)
Date: Thu, 25 Apr 2002 17:15:00 GMT
Subject: ANN: disipyl plotting library - version 0.7
Message-ID: <pan.2002.04.25.17.14.57.280429.8553@snet.net>

DESCRIPTION:
-----------
    disipyl is an object-oriented wrapper around the DISLIN plotting
library.  DISLIN is a powerful and flexible multiplatform (Win32, Unix,
Linux, etc.) library designed for displaying scientific data. DISLIN's
author, Helmut Michels, has made available a DISLIN plotting extension for
the Python programming language (see http://www.linmpi.mpg.de/dislin/ for
more details).

    disipyl provides a set of classes which represent various aspects
of DISLIN plots, as well as providing some easy to use classes for
creating commonly used plot formats (e.g. scatter plots, histograms, 3-D
surface plots).  A major goal in designing the library was to facilitate
interactive data exploration and plot creation.

Documentation, a tutorial, and a demo program are included.  The library
has been tested on Win32, Linux, and FreeBSD, but I anticipate that it
should work on any platform which can make use of Python, NumPy, and the
DISLIN python extensions.

    Feedback, comments, and critique are gladly accepted (email:
paul.magwene at yale.edu).



VERSION:
-------
    This is release 0.7 of disipyl.

CHANGES:
-------
New in this release:
    
* Include a wxPython demo based on an example submitted by Les Schaffer

* Added support for generating PNG and PDF versions of plots as string
buffers (useful for serving images to web pages)

* Fixed a bug in Text class

* Fixed bug in FunctionPlot class

* Modified the format of the INFOMODULE option to take both a module name
and a class name

* Implemented "mylabels" options for Axis, as submitted by Reggie Duggard

* updated optioninfo documentation


URL:
----
    You can find disipyl at:
http://pantheon.yale.edu/~pmm34/disipyl.html


From fperez528 at yahoo.com  Sat Apr  6 21:53:30 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 06 Apr 2002 19:53:30 -0700
Subject: a problem with subclassing a dict
References: <a8o0ro$1qoa@r02n01.cac.psu.edu>
Message-ID: <a8ocbf$dbi$1@peabody.colorado.edu>

Rajarshi Guha wrote:

> Hi,
>   I was playing with some code from a tutorial by GvR
> (http://www.python.org/2.2/descrintro.html) - the relevant code is:
> 
> class mydict(dict):
> 
>     def __init__(self, default=None):
>         dict.__init__(self)
>         self.default = default
> 
>     def __getitem__(self, key):
>         try:
>             return dict.__getitem__(self, key)
>         except KeyError:
>             return self.default
> 
> md = mydict("Sorry  - no such key!!")
> md = {1:2}
> 
> print md[1]
> print md[2]
> ---------------------------------
> 
> I get:
> 
> 2
> Traceback (most recent call last):
>   File "p2.py", line 22, in ?
>     print md[2]
> KeyError: 2
> 
> 
> Should'nt I be seeing a message instead of the Traceback?

The problem is simply that your second statement (md={1:2}) wiped out the 
assignment of md and now md is a regular dict:

In [17]: md = mydict("Sorry  - no such key!!")

In [18]: md[1]
Out[18]: 'Sorry  - no such key!!'

In [19]: type md
-------> type (md)
Out[19]: <class '__main__.mydict'>

In [20]: md = {1:2}

In [21]: type md
-------> type (md)
Out[21]: <type 'dict'>

In [22]: md[1]
Out[22]: 2

In [23]: md[2]
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)

?

KeyError: 2

So, it's not a problem with the new classes. You simply assigned md to a 
normal dict, you can't expect it then to behave like a mydict, can you?

Cheers,

f.


From gimbo at ftech.net  Thu Apr  4 05:03:24 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 4 Apr 2002 11:03:24 +0100
Subject: List files recursively: use recursive function or os.path.walk()
In-Reply-To: <4cc96c48.0204031030.53bfb69f@posting.google.com>
References: <4cc96c48.0204031030.53bfb69f@posting.google.com>
Message-ID: <20020404100324.GB1779@andy.tynant.ftech.net>

On Wed, Apr 03, 2002 at 10:30:14AM -0800, python-list-admin at python.org wrote:
> I want to list all files recursively in a a directory. I can either
> use a recursive function all or use os.path.walk(). I tested both, and
> found that os.path.walk() was actually slightly slower.

[snip]

> conclusions:
> * recursive call is slightly faster than os.path.walk() .
> * code is shorter using os.path.walk() .

To me, that time difference is practically neglible.  Use the code
that's easiest to grok when you look at it again in six months.
Whether that's os.path.walk or your version is up to you, but I'd go
with the standard library every time.

Also, check out the python style guide.  ;-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From etienne at coetzeee.freeserve.co.uk  Sun Apr 28 07:01:33 2002
From: etienne at coetzeee.freeserve.co.uk (Etta Makwetta)
Date: Sun, 28 Apr 2002 12:01:33 +0100
Subject: Scripting lanuages for laboratory tests
Message-ID: <aagkfm$6kf$1@news5.svr.pol.co.uk>

To whom it may concern

I am tring to decide on which scripting language to use in one of our
laborities during testing procedures, and I don't really know what the real
advantages and disadvantages are.  I have made the following list, and would
like to know if anybody knows of any obvoius other advantages/disadvantages
that Python has compared to Tcl.  Any other comments about Ruby, Perl etc.
are also welcome.

Tcl:

Pros: Easy to learn, Easy for simple applications, Easy GUI tool (tk),
Portable, Easy Embeddability (from c-functions), Easy Extension (to
c-functions), Free, (Can be object oriented with extension [incr Tcl]

Cons: Less exception handling, No overflow checking, Less support for
complicated data structures, Not support calling an arbitrary dll,

Python:

Pros: Object oriented, Modular (name scope etc), Portable, Overflow
checking, Exception handling, Easy Embeddability (from c-functions), Easy
Extension (to c-functions), Support for complicated data structures,
Copyrighted but free

Cons: GUI slightly more complicated than Tcl, More to learn, Not support
calling an arbitrary dll

Neither Tcl nor Python supports calling an arbitrary dll.  Tcl supports
loading of dll's, however they must contain an initialisation procedure that
translates the functions  to Tcl commands. If this is not an option, e.g. if
you want to use a Win32 dll, a wrapper must be written for each dll. Similar
conditions apply to Python. Otherwise there are no significant differences.
Which to choose is basically a matter of taste.

Kind Regards

Etta Makwetta







From starx at pacbell.net  Mon Apr  8 14:11:36 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 18:11:36 GMT
Subject: newbie question about event based programming
References: <mailman.1018257016.24202.python-list@python.org> <%des8.59197$pT1.1704440@news1.tin.it>
Message-ID: <Xns91EA720D11016RASXnewsDFE1@130.133.1.4>

Alex Martelli || Mon 08 Apr 2002 03:24:59a:

> ashimali at btinternet.com wrote:
> 
>> Hello All,
>> I'm wondering if there is a way to put an "event listener" (sorry for
>> the Java terminology), on a directory. What I'm trying to do is have
>> a python script be invoked on files whenever new file/s are 
>> copied/moved to a particular directory. Any tips gratefully recieved!
>> Cheers Adam 
> 
> That depends, first of all, on the functionality of your underlying
> operating system -- for example, Windows lets you do that (if your
> event loop is coded in a powerful enough way, but that's another
> issue).  I don't believe there is any cross-platform way -- it's not
> up to the language, alas.  The best you can achieve in most platforms
> would be to "poll", say every second, to see if the directory was
> modified.  Even just getting a function of yours called every second
> is hard to arrange in a cross-platform way.  Fine if you're using
> some given GUI toolkit, such as Tkinter (it has an .after method to
> schedule a function call, and you need to call it again in the
> scheduled function to get periodic repetitions), but there's no way
> that works across platforms and toolkits.
> 
> Event-driven programming is a great thing, but it's not supported
> as well as it could and should be across every platform and toolkit
> in the world:-(.

If you just want it to inform you when something adds files to the 
directory, and you don't need split second response... I made a program 
to monter a network in directory a while ago, and it did something like 
this:

while 1:
    	self.check_dir()
    	os.sleep(60) #I think that's hte sleep i used

I can't seem to find the flie anymore though -- I gave up on it because 
I couldn't get the systray icons to display properly =).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From brueckd at tbye.com  Tue Apr 16 12:41:19 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 16 Apr 2002 09:41:19 -0700 (PDT)
Subject: pre-PEP for optional 'pass'
In-Reply-To: <Xns91F25DDBFEB2Eduncanrcpcouk@127.0.0.1>
Message-ID: <Pine.LNX.4.33.0204160935300.9176-100000@ns0.tbye.com>

On Tue, 16 Apr 2002, Duncan Booth wrote:

> You should always head each method and class with a docstring.

No! You should head each method and class with a docstring *only if the
docstrings would add value*. Otherwise you're doing more work now AND
later (during refactoring) without producing any real benefit, cluttering
the code in the process.

> class MyClass:
>    '''A simple class'''
>    def __init__(self, x, ,y, z): '''MyClass constructor'''
>    def getTwiddle(self): '''Get a twiddle'''
>    def setTwiddle(self, twiddle): '''Set a twiddle'''

The above code snipped contains exactly 4 examples of docstrings that
should not exist - all of them either contain no useful information or no
information that isn't already painfully obvious.

A good rule of thumb is that if the docstring (or comment) could have been
auto-generated by your editor, you shouldn't include it.

-Dave





From tejarex at yahoo.com  Wed Apr 17 11:09:13 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 17 Apr 2002 15:09:13 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <tegv8.70558$GS6.7156270@bin3.nnrp.aus1.giganews.com>

<dominikush at yahoo.com> wrote in message
news:3CBD1C02.8AB2CF62 at yahoo.com...

Preliminary note: over-riding the builtin constructor function names
'list' and 'dict' with program values is a bad habit to get into.  So
I modified the quotes.

vlist = ddict.values()
...
> However, if a dictionary has lists as value entries, I get
> a counterintuitive behavior (which, just recently, broke my
> code): If you change the dict, the list you previously
> created via ddict.values() gets automagically updated

The only list created by ddict.values() is the list of references to
the values in ddict.

> >>> ddict['a']=3

This changes the reference to 2 to a new reference to 3.
ddict.values() is changed and now different from vlist.

> >>> ddict['a'].append(3)

This does not change the reference bound to a and hence ddict.values()
is unchanged and still equal to vlist.  It uses the reference bound to
'a' to access and change the list, but *this* change is invisible to
ddict.

Terry J. Reedy







From aahz at pythoncraft.com  Sun Apr  7 10:48:56 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 10:48:56 -0400
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com> <just-16031B.10170807042002@news1.xs4all.nl>
Message-ID: <a8pm8o$9hm$1@panix1.panix.com>

In article <just-16031B.10170807042002 at news1.xs4all.nl>,
Just van Rossum  <just at xs4all.nl> wrote:
>In article <a8nric$sbb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
>wrote:
>>
>> The three execution scopes are function local (hereafter referred to as
>> local), module global (hereafter referred to as global), and builtin
>> scope.  Local scope exists any time Python's instruction pointer is
>> inside a function/method.  Global scope refers to the currently
>> executing module; explicitly accessing a function in another module
>> through attributes changes the current module:
>> 
>>     from M import f
>>     f()                 # f()'s global scope is the current module
>>     import M
>>     M.f()               # f()'s global scope is now the module M
>
>No: f()'s global scope is always module M. Scoping is static after all. 
>Or are you saying something different?

Whoops!  No, you're right, you caught me being too clever without
actually testing code.  <slaps hand>  Bad Aahz!
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From reply at in_newsgroup.tks  Fri Apr  5 14:43:17 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 5 Apr 2002 19:43:17 GMT
Subject: Newbie's 1st Web App decisions!
References: <mailman.1018028292.11988.python-list@python.org>
Message-ID: <a8kuol$19l$1@hfc.pacific.net.hk>

Geoffrey Talvola <gtalvola at nameconnector.com> wrote in
news:mailman.1018028292.11988.python-list at python.org: 

> Scrumpy wrote:
>> I do like to separate style, layout, app and data and will 
>> avoid using PSP, 
>> Cheetah etc. completely or at least in ways which integrate 
>> app code within 
>> HTML pages.

[snip]

> A better argument against PSP is that it isn't web-designer-friendly. 
> That was a primary motiviation for creating Cheetah -- because its
> syntax is considered better for non-programmers than PSP's syntax.

I can't argue ;-) I'm just looking for informed suggestions on a subject I 
know nothing about.

I only pointed out a personal preference i.e. I don't like to mix app logic 
and html layout in the same page, as can be seen in many pages with PHP.

On reflection, a better way to have written the sentence would have been 
"...will avoid using PSP, Cheetah etc. in ways which integrate app logic 
within HTML pages." =)


To get back on track with my original post, here's part of what I wrote:

"At the moment I'm considering using either Webware or Slither [snip]"

"So far though, I have not found any independant info on Slither (and 
nothing at all on Slither with mod_snake) [snip]"

"What I would be most appreciative of, is your helpful suggestions on how 
to start in Python Web app development and on relevant books."

That last quote above is the most important. I haven't fixed my sights on 
any particular Python related web app server/components/frameworks/RAD 
tools and am seeking Python scholars' help on how to proceed.

Although it is nice to be told about this or that web app related project, 
I'm not in a position to know if it's better or worse than the 
alternatives. I'd really appreciate hearing (relatively) unbiased 
comparisons between projects and perhaps being given a slight push towards 
one.

Thanks for your input.

Cheers,
Scrumpy :)


From phil at river-bank.demon.co.uk  Mon Apr 15 07:56:33 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Mon, 15 Apr 2002 12:56:33 +0100
Subject: Questions on the C Thread API
Message-ID: <3CBABFF1.3220B98D@river-bank.demon.co.uk>

I'm trying to properly implement the Qt threading classes in PyQt so
that they correctly interact with Python's thread support, and I'm
having a problem getting my head around some of the issues.

In a PyQt script, execution flips between Python and C++. A script will
call a C++ method, which may call a C++ virtual method that has been
re-implemented in Python, which itself calls another C++ method, etc,
etc. The boundaries points between Python and C++ have to manage the
global lock and the current thread state to allow other (Python) threads
to run.

Currently, any C++ method that might block effectively has a call to
PyEval_SaveThread() before it and a call to PyEval_ReleaseThread() after
it. Any C++ virtual method that has been re-implemented in Python first
checks to see if it already has the global lock, and acquires it if it
doesn't. The Python method is then evaluated, and the global lock
released if it had previously been acquired. The check if the lock has
already been acquired is necessary because I only release the lock on a
small subset of C++ methods, ie. those that might block, which I do for
reasons of efficiency.

All the above works fine, so long as all the C++ calls are made from the
same thread. With the support of the Qt thread classes, that assumption
is no longer valid.

Therefore, just before calling the Python code that re-implements a C++
virtual method, the global lock has to be acquired, and the correct
thread state restored. However, to prevent a deadlock the thread state
can only be restored if it is not the current one, but there is no way
to get the value of the current thread state without having acquired the
global lock. Catch 22.

The obvious solution (I think) is to always release the global lock when
entering C++ code and always acquire it again when entering Python code
- but as only a couple of dozen out of over 5000 methods are affected
I'd like to avoid it if possible.

Is my thinking correct here? Are there other alternatives?

Thanks,
Phil




From fredp at mygale.org.nospam  Fri Apr 12 13:20:28 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 12 Apr 2002 17:20:28 GMT
Subject: CGIHTTPServer, revisited (sorry)
References: <Xns91EEBD0068607PaCmAnRDLM@161.48.128.20>
Message-ID: <Xns91EEC531063F4PaCmAnRDLM@161.48.128.20>

Fred Pacquier <fredp at mygale.org.nospam> said :

Of course, just after posting this, I found one of those Google groups
posts I'd overlooked :

http://groups.google.fr/groups?hl=fr&threadm=67abb823.0105301417.6be429ae%40posting.google.com&rnum=28&prev=/groups%3Fq%3DCGIHTTPServer%26hl%3Dfr%26scoring%3Dd%26as_drrb%3Db%26as_mind%3D12%26as_minm%3D4%26as_miny%3D2000%26as_maxd%3D12%26as_maxm%3D4%26as_maxy%3D2002%26start%3D20%26sa%3DN

This describes the exact symptoms (although they are said to be MSIE-
specific, where I also have problems with Netscape 4). There were two
related bugs open of SourceForge, one of them by Kevin Altis :

http://sourceforge.net/tracker/?func=detail&atid=105470&aid=427345&group_id=5470

All of this is almost a year old (May 2001), and still stands in 2.2.1
(the topic seems to irritate Guido a lot, although it does make a few
people like me miserable :-)

BTW, I have tried out the one-line patch to socket.py suggested in one
of the bug follow-ups, and behold, everything now works perfectly (AND
much faster) in both MSIE _and_ Netscape 4... Probably tricky to distribute
an app with a specific socket.py though ? :)
(there is also that 'reading an extra 2 bytes from the rfile before sending 
your response' suggestion, but it's not clear to me where/how to apply it...)

Anybody knows if Kevin's modified CGIHTTPServer.py is available somewhere ?

Thanks,
fp

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


From maxm at mxm.dk  Fri Apr  5 03:57:44 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 10:57:44 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAD4CD8.2090700@mxm.dk> <7Adr8.1447$%A.217236@news1.west.cox.net>
Message-ID: <3CAD6708.3080602@mxm.dk>

Van Gale wrote:


>>Just do a search and replace in your codebase::
>>
>>     s/1/True/gi
>>
>>That should solve all those problems in a few seconds.
> 
> 
> Yeah... it'll solve **those** problems, but think about it for a sec.
> Maybe, just maybe, this will create a slew of entirely different problems?

Oh yes I see.

Well Alex ... if you did the search and replace I adviced, please do a 
reverse search and replace::

     s/True/1/gi

Then those problems that the first search and replace created should be 
solved.

regards Max M



From liste_nospam_ at opikanoba.org  Fri Apr 19 04:22:18 2002
From: liste_nospam_ at opikanoba.org (=?windows-1252?Q?Fr=E9d=E9ric=20Laurent?=)
Date: Fri, 19 Apr 2002 08:22:18 GMT
Subject: amkCrypto installation on win32
Message-ID: <1103_1019204538@news.free.fr>

Sur 18 Apr 2002 15:05:30 GMT, vous avez ?crit dans comp.lang.python:
> > > Try M2Crypto:
> > 
> > and there is no precompiled dll in the win directory of the binary zip package,
> 
> Is too!
> 
>         0  03-08-02 00:30   m2crypto-0.07-snap3/win/py22/
>     19350  03-05-02 22:18   m2crypto-0.07-snap3/win/py22/_m2crypto.py
>    114688  03-05-02 22:18   m2crypto-0.07-snap3/win/py22/_m2cryptoc.pyd
> 
> 
> I just dropped this into a Python 2.1 on a W2K boxen. Works fine.
> 


As I use python 2.2, I expected to find a .dll file !
It works for me too...
sorry ;-)
thanks

--
Fr?d?ric Laurent
http://www.opikanoba.org 




From mal at lemburg.com  Fri Apr 26 14:04:33 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 26 Apr 2002 20:04:33 +0200
Subject: configuring mxCGIPython
References: <NCBBLFCOHHDIKCAFGCFBOELMKMAA.marklists@mceahern.com>
Message-ID: <3CC996B1.978C346A@lemburg.com>

Mark McEahern wrote:
> 
> [M.-A. Lemburg]
> > Hmm, interesting. Looks like sys.path is empty at this point.
> > I wonder whether this is a bug in the frozen cgipython built
> > or just a consequence of freezing the interpreter.
> >
> > Could you try setting PYTHONPATH to the dir where cgipython
> > lives ?
> 
> I tried that.  The first error related to site.py no longer occurs, but I
> still get the second error.  Output appended at the end of the message.  I
> don't really know what line 27 of cgipython.py is?

You have to call cgipython as:

	cgipython script.py [options]

You probably have passed in something else as first parameter
which is why you get the error in cgipython.py:

# run the script file passed in sys.argv[1]
import sys
try:
        script = sys.argv[1]
except IndexError:
        sys.stderr.write('Usage: cgipython <pyfile> [parameters]\n')
else:
        del sys.argv[0]
        execfile(script) # <-- this is line 27

I should probably add another try:except: here which generates
a better error messages.

> > Since the stuff currently done in site.py doesn't really
> > make all that much sense here, I think cgipython should
> > probably use an empty site.py file instead of the standard
> > one (should enhance startup times even more).
> 
> Sadly, I don't know enough to really judge this.
>
> > Good question. It probably doesn't even start looking for
> > it since it's already preregistered in the built-in
> > modules.
> 
> I'm building my own fresh copy of cgipython to see if that makes any
> difference.  I'll let you know.
> 
> Again, I appreciate your feedback and help.

That's ok. I want cgipython to be as easy to use as possible
and our little chat helps with this.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From tdelaney at avaya.com  Mon Apr 22 20:42:07 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 23 Apr 2002 10:42:07 +1000
Subject: __iadd__ and fellows missing (Python 2.2)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A287@natasha.auslabs.avaya.com>

> From: Jeff Shannon [mailto:jeff at ccvcorp.com]
> 
> While there might be some theoretical argument for considering 
> augmented assignment on immutable types to be an error, keep in 
> mind that practicality beats purity.  Making augmented assignment 
> on integers into an error would, in practice, be quite silly.

Especially as it is the place I am most likely to use it ...

There is actually nothing stopping a mutable object from implementing a = a
+ b semantics in __iadd__ and fellows either ... or having different binding
semantics in __iadd__, __imul__, etc ... (although the latter would be truly
horrendous).

Basically, if you don't want any chance of an object being mutated by
(optionally) in-place operations, make a copy of the object and operate on
that.

Tim Delaney




From phr-n2002a at nightsong.com  Mon Apr 15 19:34:15 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 16:34:15 -0700
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp> <7xelhhmq5e.fsf@ruckus.brouhaha.com> <slrnablf4d.5k9.mlh@vier.idi.ntnu.no>
Message-ID: <7xpu10mt8o.fsf@ruckus.brouhaha.com>

mlh at vier.idi.ntnu.no (Magnus Lie Hetland) writes:
> Well -- the Python license is GPL-compatible, so couldn't you just
> release a GPL'ed version of it, if that would make you happy? Of
> course, the original would still be freely available without the added
> burden of the GPL restrictions, but if you're really interested in
> freedom, that shouldn't bother you...

That would mean forking the implementation and having to do a lot of
work to maintain a separate version and probably track fixes from the
non-GPL version to merge them into the GPL version.  That's possible
in principle, but not very sensible technically, and I would see it as
unfriendly.  So I likely wouldn't do it unless the changes were very
large, or if I did a new implentation that used some pieces of the
original but was mostly written from scratch.  And if I did something
like that, it would be for technical reasons (e.g. if I wanted to try
writing a native-code compiler), rather than for ideological ones.
Life is too short to reimplement a perfectly good GPL-compatible
program just because it's not pure GPL.  (I do sometimes think about
seeing if I can implement Python on top of an existing Lisp or Scheme
compiler/interpreter).

I'm glad Python is out there and that it's free.  I don't have any
ill-will toward its license and I don't see it as a "competitor" the
way a proprietary implementation might be a competitor.  I'm happy to
contribute any small improvements that I happen to make.  I just feel
less motivated (as an unpaid volunteer) to write large amounts of code
likely to end up in someone else's proprietary product, than I'd be to
write volunteer code for a GPL'd program.  If I'm getting paid, I'm
happy to write any kind of free code.

Everyone has to decide for themselves what types of programming they
personally feel good about.  I'm not as dedicated a free software
purist as RMS--I prefer to write free code but I'm willing to write
proprietary code if I'm getting paid.  I draw the line at writing
proprietary code when I'm NOT getting paid. :)


From akuchlin at ute.mems-exchange.org  Mon Apr 15 17:04:53 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 15 Apr 2002 21:04:53 GMT
Subject: problem negative lookahead assertion
References: <4b926dbb.0204151202.6d709633@posting.google.com>
Message-ID: <slrnabmg3g.hbr.akuchlin@ute.mems-exchange.org>

In article <4b926dbb.0204151202.6d709633 at posting.google.com>, 
	sjoerd siebinga wrote:
> emph = re.compile(r'\s.\\emph\{(.*?)\}([\s,\W])(?!\\index)')

I had to change \s. to just \s to get the pattern to match at all.

> \begin{germdata} ON \emph{va\th a} \index{on~va\th a}`wade, rush, walk

The pattern is matching too far. This text was turned into the following:
\begin{germdata} ON\emph{va\th a} \index{on~va\th a}` \index{unl~va\th a}
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^ contents of 
group 1.

I can make it work the way you intend by writing ([^}]*?) instead of
(.*?); this ensures that group 1 in the regex can only match up to the
very next '}'.  Note that this will break in a different situation, if
you have {} inside the contents of \emph, as for example in
\emph{ab^{cd}}, because this problem really needs a full parser to
handle the general case.  But maybe your input data never uses {}
inside \emph{}, in which case you'd be OK.

--amk                                                             (www.amk.ca)
The more I know me, the less I like me.
    -- The Doctor, in "Time and the Rani"


From klm at zope.com  Fri Apr 26 16:26:32 2002
From: klm at zope.com (Ken Manheimer)
Date: Fri, 26 Apr 2002 16:26:32 -0400 (EDT)
Subject: [Crew] PSA revival - kind of?
In-Reply-To: <0d9001c1ed5b$29939830$e000a8c0@thomasnotebook>
Message-ID: <Pine.LNX.4.44.0204261625590.23319-100000@korak.zope.com>

On Fri, 26 Apr 2002, Thomas Heller wrote:

> > --------------------------------------
> > I would like to start the SPA:
> > 
> >     Starship Python Activity
> > 
> > as a non-profit organisation.
> > --------------------------------------
> > 
> > Here the proposed layout:
> > SPA membership costs USD 25 by the year.
> > This price includes Starship membership
> > and a continous subscription to PyZine.
> 
> Do anything you want with my money (within resonable
> bounds), but don't make me a member of another
> organization. Please ;-)

It cost only a little bit more to contribute but *not* become a member...

-- 
Ken
klm at zope.com





From borcis at geneva-link.ch  Mon Apr 15 07:33:05 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 15 Apr 2002 13:33:05 +0200
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com>
Message-ID: <3CBABA71.3040909@geneva-link.ch>


Ken Peek wrote:

> I think an 'xor' operator is needed.  The 'xor' operator means
> 'logical-exclusive-or', and returns a 1 or 0.
> 
> Here is some code that does what I want (in all the cases I can think of) for
> 'xor':
> 
> # logical-exclusive-or operator:
> def xor(a,b):
>     return (not a) ^ (not b)
> 


For the shortcut semantics, we'd like to have :
     a xor b

behave exactly like :
     (not b and a) or (not a and b)

In Lisp, we would implement that with a special form
macro.

On another hand, the shortcut semantics of test expressions
(and I guess it extends to cases like a < b < c) seems
to flourish to meaningful or at least interesting generalizations,
when terms are generators or iterators.

Indeed, shouldn't two generators g1 and g2, when combined
in an expression of the form e.g. "g1 and g2" or "g1 or g2",
combine meaningfully to become a third generator ? Respectively
one that returns a successive value of g2 for every value that
g1 returns, until one of them returns IterationError, and one
that concatenates the values of g1 with the values of g2 ?

Regards, Boris Borcic
--
"Art is a trigram"



From david.meleedy at analog.com  Wed Apr 17 02:27:51 2002
From: david.meleedy at analog.com (David Meleedy)
Date: 16 Apr 2002 23:27:51 -0700
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk> <20020412144929.B12997@phd.pp.ru> <200204121227.g3CCRSd04250@localhost.localdomain> <mailman.1018615046.3490.python-list@python.org>
Message-ID: <258694d4.0204162227.656eb6d3@posting.google.com>

I also am having problems getting python compiled on Solaris 2.5.1.  I
am
seeing the same problem (kind of).  I am using gcc 2.95.3, and I am
using
/usr/ccs/bin/ld because gcc -shared did not work for any version of
Solaris (using gnu binutils) to build the shared object modules for
python.


If I do a configure with
--with-threads=no, then I get:

building 'struct' extension
creating build
creating build/temp.solaris-2.5.1-sun4u-2.2
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I.
-I/cpd/src/misc/python/2.2.1/./Include -I/usr/local/include -IInclude/
-c /cpd/src/misc/python/2.2.1/Modules/structmodule.c -o
build/temp.solaris-2.5.1-sun4u-2.2/structmodule.o -DNDEBUG -g -O3
-Wall -Wstrict-prototypes
creating build/lib.solaris-2.5.1-sun4u-2.2
ld -G build/temp.solaris-2.5.1-sun4u-2.2/structmodule.o
-L/usr/local/lib -o build/lib.solaris-2.5.1-sun4u-2.2/struct.so
Segmentation Fault
gmake: *** [sharedmods] Error 139

If I build with threads, putting in the hack to
Python/thread_pthreads.h
#if defined(__APPLE__) || defined(sun)
#define destructor xxdestructor
#endif

or

#if defined(__APPLE__) || defined(sun)
#define destructor xxdestructor
#endif
#include <pthread.h>
#if defined(__APPLE__) || defined(sun)
#undef destructor
#endif

Then when I try to compile, I get a bit further, and then things get
fouled
up again:

building 'time' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I.
-I/cpd/src/misc/python/2.2.1/./Include -I/usr/local/include -IInclude/
-c /cpd/src/misc/python/2.2.1/Modules/timemodule.c -o
build/temp.solaris-2.5.1-sun4u-2.2/timemodule.o -DNDEBUG -g -O3 -Wall
-Wstrict-prototypes
ld -G build/temp.solaris-2.5.1-sun4u-2.2/timemodule.o -L/usr/local/lib
-lm -o build/lib.solaris-2.5.1-sun4u-2.2/time.so
Segmentation Fault
gmake: *** [sharedmods] Error 139


If I then modify setup.py to not compile the time extension, but to
move
along, then other modules get the seg fault later on too.

What I discovered doing this, is that I could get some of the later
modules to work that were previously failing, if I commented out
some of the eariler modules in the setup.py file.  So perhaps lines
like:

exts.append( Extension('xreadlines', ['xreadlinesmodule.c']) )

is causing a memory problem if too many are added?

However I never was able to get time.so to work, so there probably is
a thread
problem in that file.

I hope this helps a bit.

-Dave


From peter at engcorp.com  Mon Apr 15 22:12:48 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 15 Apr 2002 22:12:48 -0400
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <3CBB88A0.CD3ED00D@engcorp.com>

Philipp Lenssen wrote:
> 
> Is it possible to use XPath in Python, if so are there any small samples? (I
> already saw some XML DOM and SAX mentioned in the context of Python, but
> little of XPath.) I don't mean using XSLT files, but opening an XML file in
> Python, and executing something like evaluate/ match/
> selectNodes(xPathString), returning a set of nodes.

Try this out for a start:

>>> import xml.dom.minidom
>>> from xml.xpath.Context import Context
>>> import xml.xpath
>>> s = '''<?xml version="1.0"?>
... <root>
... <elem attr="1">Some text</elem>
... <elem attr="2">More text</elem>
... </root>'''
>>> d = xml.dom.minidom.parseString(s)
>>> path = xml.xpath.Compile('//elem[@attr="1"]')
>>> result = path.evaluate(Context(d))
>>> print result
[<DOM Element: elem at 18968128>]
>>> print result[0].nodeName
elem
>>> print result[0].firstChild.nodeValue
Some text

I believe there is a small step or to required to install the latest
PyXML over the standard distribution which has a previous version.
It was documented clearly somewhere, but I don't recall what the
steps were.  (I don't think xpath was in the latest standard release,
but I could remember incorrectly.)

-Peter


From aahz at pythoncraft.com  Mon Apr  8 23:59:34 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Apr 2002 23:59:34 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <a8tov6$484$1@panix1.panix.com>

In article <4a7fde99.0204081858.58baabdd at posting.google.com>,
Ryan <ryanmorillo at hotmail.com> wrote:
>
>    I have been doodling with python (nasty habit I know but I have to
>do somthing with my snake) and think I may be ready to learn a second
>language.  I am leaning towards C over C++ and Java even though I
>loose objects it will ground me firmly in functional programing.  I am
>wondering what book would be the best for a begginer?  The two I have
>seen that look interesting are
>C for Dummies (parts one and two) or 
>The C Programming Language
>by Brian W. Kernighan, Dennis M. Ritchie.

If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
free to buy another book if K&R turns out to be too heavy for you right
now, but recognize that you'll need it eventually.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From seth at jtan.com  Thu Apr 25 02:31:11 2002
From: seth at jtan.com (Seth Delackner)
Date: Thu, 25 Apr 2002 06:31:11 GMT
Subject: urllib.urlopen("file://...") fails
Message-ID: <m2k7qwjnma.fsf@jtan.com>

With python 2.2 on windows, with the file t:\tmp\package.html existing,

>>> import urllib
>>> u = urllib.urlopen("file://t:/tmp/package.html")

Fails with the error below, but "file:t:/tmp/package.html" works. Is
that even a valid url?

>>> u = urllib.urlopen("file:///t:/tmp/package.html")
    Traceback (most recent call last):
      File "<pyshell#40>", line 1, in ?
        u = urllib.urlopen("file:///t:/tmp/package.html")
      File "d:\python\lib\urllib.py", line 73, in urlopen
        return _urlopener.open(url)
      File "d:\python\lib\urllib.py", line 178, in open
        return getattr(self, name)(url)
      File "d:\python\lib\urllib.py", line 405, in open_file
        return self.open_local_file(url)
      File "d:\python\lib\urllib.py", line 412, in open_local_file
        stats = os.stat(localname)
    OSError: [Errno 2] No such file or directory: '\\t:\\tmp\\package.html'

>>> u = urllib.urlopen("file:t:/tmp/package.html")
>>> u.read()
    contents of the html file spew forth.

Just for fun I tried localhost instead of just "///":

>>> u = urllib.urlopen("file://localhost/t:/tmp/package.html")
    Traceback (most recent call last):
      File "<pyshell#41>", line 1, in ?
        u = urllib.urlopen("file://localhost/t:/tmp/package.html")
      File "d:\python\lib\urllib.py", line 73, in urlopen
        return _urlopener.open(url)
      File "d:\python\lib\urllib.py", line 178, in open
        return getattr(self, name)(url)
      File "d:\python\lib\urllib.py", line 403, in open_file
        return self.open_ftp(url)
      File "d:\python\lib\urllib.py", line 470, in open_ftp
        self.ftpcache[key] = \
      File "d:\python\lib\urllib.py", line 702, in __init__
        self.init()
      File "d:\python\lib\urllib.py", line 708, in init
        self.ftp.connect(self.host, self.port)
      File "D:\PYTHON\lib\ftplib.py", line 131, in connect
        raise socket.error, msg
    IOError: [Errno ftp error] (10061, 'Connection refused')


From mlh at vier.idi.ntnu.no  Mon Apr 15 07:35:24 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 15 Apr 2002 11:35:24 +0000 (UTC)
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <slrnablens.5k9.mlh@vier.idi.ntnu.no>

In article <a9cv63$qhu$1 at bob.news.rcn.net>, Raymond Hettinger wrote:
>The help and tutor maillists indicate that the use of input() rather than
>raw_input() is widespread. I think this function is so harmful and so
>likely to be used that it should be deprecated hasta pronto (right away).
>When really needed, eval(raw_input()) is easy enough to code and explicit
>about what it does.
>
><no wink -- this function needs to go away>
>

But (in the (very?) long run) I think that the name "input" should
then be used for what is now raw_input. I know -- breakage, and all
that... I just think it fairly odd to have *only* raw_input, since the
"raw" qualification then becomes fairly meaningless.

>Raymond Hettinger

- Magnus (who in his early Python days was very baffled by the name
  "raw_input", until he found "input"...)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From bbollenbach at shaw.ca  Fri Apr 12 17:52:10 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 21:52:10 GMT
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
Message-ID: <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca>

In article <GrAt8.11681$F43.8566 at atlpnn01.usenetserver.com>, Steve Holden wrote:
> "Brad Bollenbach" <bbollenbach at shaw.ca> wrote ...
>> Peter Olsen wrote:
>> > I think that, for my purposes, licensing is not worth my effort or
>> > my users' time.  But if you want to license your Open Source software,
>> > be my guest.
>>
>> This doesn't actually make sense. To call software "Open Source" is to
>> acknowledge it as being distributed under a license that is defined as
>> compatible with what Richard Stallman calls "Free Software". So you
>> don't have the "choice" of licensing Open Source software. It got to be
>> called "Open Source" *because* of the license you've already chosen for
>> it.
>>
> So the code I have put in the public domain isn't open source? That's
> interesting.

That's correct. 

To "put it into the public domain" the only thing that
is implicit is the fact that you hold the copyright. When you start
adding clauses like "feel free to modify and distribute it as you like",
etc. you've begun to establish it as Open Source (or "Free" depending on
whose terminology you're more interested in using). When you've attached
a license to it that is compatible with the GNU definition of "Free"
you've officially made it "Free Software" (or "Open Source",
again...depending on whose terminology you prefer to use to convey that
meaning).

For more, see:

http://www.gnu.org/philosophy/free-sw.html

Note the importance of actually attaching a Free Software license,
because, as Stallman says in the above link:

"In order for these freedoms to be real, they must be irrevocable as
long as you do nothing wrong; if the developer of the software has the
power to revoke the license, without your doing anything to give cause,
the software is not free."

So, if the "only" thing you've done is to "put in the public domain"
you've not made these rights real. When you release a piece of code with
a license though -- and one that is compatible with Open Source/Free
Software -- you *have* made these rights real, and thus, you've released
Open Source/Free Software.

>> And, FWIW, thinking that you're doing your users a favour by monitoring
>> who's using which version of what, and in what kind of environment
>> (commercial vs. non-commercial) is, at best, dependant on the types of
>> users, and, at worst, and absolutely disasterous idea. If they're
>> "end-user" types, you might legimitately be providing some capabilities
>> they want, but adopting this scheme for Open Source Hacker types will
>> just reduce your user-base by an order of magnitude or more.
>>
> I think you might have been taken in by a date-dependent (think "April 1"
> here) post that wasn't intended to be taken that seriously.

Not quite. :) 

I'm well aware that it was a joke. My comments make no mention of the 
Twisted development framework.


--
Brad Bollenbach


From boud at valdyas.org  Thu Apr 25 01:38:16 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 25 Apr 2002 07:38:16 +0200
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <c76ff6fc.0204241734.40e57406@posting.google.com>
Message-ID: <3cc7967a$0$37911$e4fe514c@dreader3.news.xs4all.nl>

John Machin wrote:
> 
> Presumably the point of having a multi-character pronunciation table
> is that it is possible that pronounce("xy") can be != pronounce("x") +
> pronounce("y"). With careful thought, you may be able to remove
> redundant entries from your more-than-one-char dicts, so that they
> contain only the necessary exception cases -- but do try the basic
> approach first.
> 

Isn't big-5 a variable length encoding? I thought that was his
problem, not translating two or more character words. 

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


From huaiyu at gauss.almadan.ibm.com  Thu Apr 18 15:09:49 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 18 Apr 2002 19:09:49 +0000 (UTC)
Subject: [OT] What is Open Source?
References: <20020418002823.C4068@phd.pp.ru> <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov> <mailman.1019085816.3890.python-list@python.org> <a9lp2o$nfv$1@watserv3.uwaterloo.ca> <MPG.1728a93cfcd0fc8e9896b1@news.nwlink.com>
Message-ID: <slrnabu6ft.2tn.huaiyu@gauss.almadan.ibm.com>

Jeff Shannon <jeff at ccvcorp.com> wrote:
> There are no limits to the usage of GPL software.  (How 
>many businesses are using Linux?)  What the GPL limits, is 
>derivative works.  You can *use* the software, but when you 
>*modify* it, then you come under the restrictions that the GPL 
>lays out.  (Particularly, when you modify it and then distribute 
>your modified version.)

I think that even if you _modify_ the software, there is still no
restriction from GPL.  A modification follows almost exactly the same rule
as original (except credits, etc).  Only distribution counts.  The GPL
mentions "modify" only to avoid claims like "I modified it, so I can
distribute my modification with my own license" and similar issues.

Think of a printed book.  If you buy one, you can use it in many ways,
including study it and add the knowledge to your own, copy it to make marks
on it in your study, lend it to friends, cite paragraphs for critical
review, cut it to pieces to make paper airplanes, etc.  You can even combine
a theorem from the book with a theorem from another book and prove a theorem
of your own, and publish it.  You can even publish another book using the
ideas you learned.

But you cannot copy it and give/sell the copy to others, modified or not.

The GPL allows you to do all the things you could with a printed book in
terms of using it.  But it gives a little bit more.  It gives you a blanket
permission to copy it and distribute the copies, modified or not, for fee or
not, provided you keep the same license on the copy and follow some
reasonable restrictions (credits, liabilities, etc).

In stark contrast, some of MS licenses (end user license agreements)
actually take away some of the user's rights he would otherwise enjoy on a
purchased printed book.  They essentially sneak in some patent rights,
trademark rights and contracts rights undercover of copyrights.

Huaiyu


From nmjoseph at attbi.com  Sat Apr 27 11:11:21 2002
From: nmjoseph at attbi.com (neal josephson)
Date: Sat, 27 Apr 2002 15:11:21 GMT
Subject: GetOpenFilename from python extension
Message-ID: <tczy8.79369$HH5.5211@rwcrnsc51.ops.asp.att.net>

I have a pretty simple python program that I want to add a windows file
dialog box to - I tried writing a python extension that called
GetOpenFilename but it just returns a DLG_INITIALIZE (or something like
that) error.  Has anyone managed to get this to work?




From ykingma at accessforall.nl  Tue Apr 23 15:41:16 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Tue, 23 Apr 2002 20:41:16 +0100
Subject: Python, search engine and NLP...
References: <409a56e2.0204230104.16d60f52@posting.google.com>
Message-ID: <3CC5B8D4.F1C12DA9@accessforall.nl>


Shagshag wrote:
> 
> Hello,
> 
> I'm interested in Pyhon but not sure it can fit my needs :
> 
> I have a huge search engine to write with many XML input/output, it
> must be able to index and process something like 3 GO of data, many
> text files, parse text and other natural language processing, not be
> too slow, and so on...
> 
> - Do you think Python can help me ???

Yes, definitely.
For starters, there's an XML parser in available in python.

> 
> - As by now i don't know Python, do you think that it will not be a
> waste of time even if I had to learn it ???

You'll easily recover your python learning time by writing the text
indexing scripts in python. These are not normally time critical and need
some experimentation for which scripting in python is a good fit.

> - Do you know about some Python's natural language processing module
> ???

There are several, I'd guess others will let you know.
 
> - No, as there are many Java NLP library, I should definitively do it
> with Java...

I have some experience using the Lucene search engine
with Jython (ie. python running in a JVM).
This combination will probably do what you need.
Looking around for other combinations never hurts, though.

> Thanks in advance,

My pleasure.
Ype

http://jakarta.apache.org/lucene
http://www.jython.org

-- 
email at xs4all.nl


From andrew at acooke.org  Thu Apr 25 10:11:29 2002
From: andrew at acooke.org (andrew cooke)
Date: 25 Apr 2002 07:11:29 -0700
Subject: XML, Python, XML-SIG, documentation, PyXml, 4
References: <aef33705.0204231902.40694af4@posting.google.com> <3CC62D76.2E0D3A54@engcorp.com>
Message-ID: <aef33705.0204250611.44622e31@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3CC62D76.2E0D3A54 at engcorp.com>...
> andrew cooke wrote:
> > 
> > Which of these packages should I use, and what documentation will help
> > me, if I want to:
> > 
> > - automatically generate a tree from the XML file (I believe this is
> > related to the DOM)
> > - supply a class to some (already existing) tree walking code that
> > does local rearrangements of the XML data
> > - automatically generate an XML file from the modified tree
> > 
> > (I've already looked at XSLT and I don't believe that is the right
> > tool for this kind of job).
> 
> Why do you say that?  What you describe above is pretty much
> what XSLT is for...

My understanding (which may be completely wrong - in which case I'd
like to know!) is that XSLT is for generating new representations from
the old data, but that the new representation is separate from the old
structure.  In other words, it's easy to generate a completely new
form based on the original information, but difficult to keep the old
data largely intact with just small modifications.

If that's wrong, please could someone point me to an example (and
apologies for being obtuse, but the examples I've seen are of that
form and, in one case, stressed what I thought was the point I'm
making above).

Thanks,
Andrew


From tanzer at swing.co.at  Wed Apr 10 05:26:54 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Wed, 10 Apr 2002 11:26:54 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "Tue, 09 Apr 2002 11:02:33 -0000."
             <slrnab5ii9.ho6.mlh@vier.idi.ntnu.no>
Message-ID: <E16vENS-0003V9-00@tswings.swing.cluster>

mlh at vier.idi.ntnu.no (Magnus Lie Hetland) wrote:

> In article <mailman.1018339899.2981.python-list at python.org>, Christian
> Tanzer wrote:
> [snip]
> >There are several dimensions to typing, including:
> >
> >- static vs. dynamic
> >
> >- strong vs. weak
> >
> [snip]
> >
> >Another important point with typing is whether typing applies to
> >variables or to objects. In Python, objects are (strongly) typed, not
> >variables. Thinking `variable` when considering typing (or objects in
> >general) will not work very well in Python.
>
> Hm. I thought this was what dynamic typing was all about, i.e. having
> typed values but untyped variables/parameters (which means that
> variable types need not be determined statically, at compile time)...
> What am I missing here?

Should I have said `In other words...` instead of `Another important
point with typing...`?


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





From bptonner at yahoo.com  Sat Apr  6 20:02:30 2002
From: bptonner at yahoo.com (bptonner at yahoo.com)
Date: Sat, 06 Apr 2002 20:02:30 -0500
Subject: calling dll's functions with python
References: <8wmq8.52601$S52.1805120@news2.tin.it>
Message-ID: <3CAF9AA6.4AE62F20@yahoo.com>

Use the calldll() function.

Giovanni Laudicina wrote:
> 
> I'm a beginner with python and i did'n find this info on the FAQ and
> tutorials:
> How can I call a function included in a dll (I have only the .dll file, not
> the source code) ?
> Thanks to everyone,
> Giovanni Laudicina


From loewis at informatik.hu-berlin.de  Sun Apr 14 08:31:35 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 14 Apr 2002 14:31:35 +0200
Subject: C API Q I : calling Python methods from C
References: <a9bl5j$2aq$1@green.tninet.se>
Message-ID: <j4vgaupil4.fsf@informatik.hu-berlin.de>

"Johan Hahn" <johahn at delta.telenordia.se> writes:

> I can't figure out how to call methods written in python from C
> using the low level API.

I could not quite figure out what your question is, but I interpret it
as "How do I do the following Python code with the C API"?

> /* b4 module */
> import code
> iii = code.InteractiveInterpreter()
> def runsource(code):
>     iii.runsource(code)


> You are supposed to read the class' dictionary for methods, convert
> a method to a function and call the function with the instance as
> the first parameter, right?

That sounds way to complicated. Here is what I would do

code = PyImport_ImportModule("code");
iii = PyObject_CallMethod(code, "InteractiveInterpreter", "");
runsource_result = PyObject_CallMethod(iii, "runsource",
                                       "i", 42);

assuming you want to pass the argument 42 to iii.runsource.

Don't forget to check for exceptions!

HTH,
Martin


From sholden at holdenweb.com  Tue Apr  9 18:53:22 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 9 Apr 2002 18:53:22 -0400
Subject: ANN: Twisted 0.16.0
References: <mailman.1017787747.12194.python-list@python.org> <3CAA8CFF.B6EDDCDE@alcyone.com> <8Fgs8.27735$Sw5.8757@atlpnn01.usenetserver.com> <23891c90.0204090100.38023e58@posting.google.com> <3CB32D3B.74369D36@alcyone.com>
Message-ID: <hjKs8.5381$GM.2@atlpnn01.usenetserver.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CB32D3B.74369D36 at alcyone.com...
> Paul Boddie wrote:
>
> > Given that the announcement is also archived in "the public record", I
> > wonder if the joke has had any effect on the number of people willing
> > to evaluate the software. It probably isn't such a great idea to joke
> > about software licensing if you want people to download and deploy
> > your product.
>
> It's also traditional that April Fool's Day only be perpetuated on April
> 1, and when found out, the perpetrators come clean.  Arguing or
> continuing to drag on the "joke" only makes people wonder.
>
Well, if I can go out on a limb here, my own impression of the twisted
software is that the whole effort is aither an attempt at obscurity or a
very poor example of documentation. It sometimes seems that those guys are
doing it just for themselves, which is perfectly OK by me.

The clarity meter's pretty low on the whole project, which is a shame when
you look at what it can achieve. I guess it's just another open source
project where documentation is undervalued.

regards
 Steve






From phd at phd.pp.ru  Tue Apr 30 09:44:22 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 30 Apr 2002 17:44:22 +0400
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <NOwz8.13638$xy4.7951@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Tue, Apr 30, 2002 at 09:13:46AM -0400
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <mailman.1020024596.32196. <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com> <200204301324.25823.rjones@ekit-inc.com> <mailman.1020153536.6149.python-list@python.org> <NOwz8.13638$xy4.7951@atlpnn01.usenetserver.com>
Message-ID: <20020430174422.D19566@phd.pp.ru>

On Tue, Apr 30, 2002 at 09:13:46AM -0400, Steve Holden wrote:
> Gadfly is great for people who don't have an existing
> relational system but want to do some Python experimentation with database.

   Other alternatives are BerkeleyDB with PyBSDDB wrapper, and MetaKit with
MkSql.

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




From grante at visi.com  Fri Apr 12 21:43:58 2002
From: grante at visi.com (Grant Edwards)
Date: Sat, 13 Apr 2002 01:43:58 GMT
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com>
Message-ID: <slrnabf3c3.2e6.grante@tuxtop.visi.com>

In article <7xy9fse6ol.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:

>> > I'd say it's hopeless or at least impractical.  Those processors are
>> > much too constrained to run Python.  It would be like trying to run an
>> > SQL database on them.  Python has many virtues but being able to run
>> > on very tiny processors isn't one of them.
>>
>>     What about something like the Zaurus SL-5500?
> 
> The SL5500 is a high powered computer with a 32-bit CPU and 64 MB of
> ram.  It runs Python just fine.  The computers the OP was asking about
> are 8-bit CPU's with a few dozen to a few hundred bytes of ram (well
> maybe a few kbytes for some 68HC11 configurations).  There's a
> tremendous difference.  While there are some C compilers that can
> produce code for those cpu's, they are normally programmed in assembler.

Maybe PICs are normally programmed in assembler, but I've seen
a lot of 68HC11 projects in the past 10-12 years, and they were
all in C -- and most of them ran on top of a multi-tasking
kernel.  None of them had a chance of running Python, though. 

-- 
Grant Edwards                   grante             Yow!  I've got a COUSIN
                                  at               who works in the GARMENT
                               visi.com            DISTRICT...


From theivo at abv.bg  Sun Apr 14 04:13:20 2002
From: theivo at abv.bg (I.C.)
Date: 14 Apr 2002 01:13:20 -0700
Subject: textarea to a python sript
References: <c73096a5.0204131612.700cd7ae@posting.google.com>
Message-ID: <6ac1ecad.0204140013.5db89bef@posting.google.com>

flyguysyl at hotmail.com (flyguy) wrote in message news:<c73096a5.0204131612.700cd7ae at posting.google.com>...
> How can I get a <TEXTAREA> to a python script?

If you mean a <textarea> like in HTML, you should first open a Tk-Window,
where you can use Buttons, Checkboxes, Entrys, Textareas etc...

If you need some information about Tkinter:
           http://www.python.org/topics/tkinter/doc.html
           www.math.utsa.edu/~gokhman/ftp/soft/man/intro-tkinter.pdf

Ivo


From jim at interet.com  Wed Apr 17 11:01:40 2002
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 17 Apr 2002 11:01:40 -0400
Subject: zipfile problems
References: <2e1b0869.0204161112.425f41d4@posting.google.com>
Message-ID: <3CBD8E54.6070805@interet.com>

Austin Luminais wrote:

> I have an archive of several hundred zip files, and I'm trying to
> automate unzipping them with the zipfile module (Python 2.2,
> Activestate).  Most of them work, but 58 of them don't.  Most that
> don't work get this error:
> 
> BadZipfile: File is not a zip file, or ends with a comment


I looked at 2force.zip, and it has an ending comment of 287 bytes
which seems to print "DC-1" in ascii art.  The zipfile module
will not handle zip files with ending comments because it
requires a search of the file.

 
> And a few get:
> 
> BadZipfile: Unsupported compression method 6 for file (filename)


The zipfile module supports compression methods 0 (no compression)
and 8 only, because the others are obsolete.

 
> WinZip and Powerdesk have no problem extracting these files.
> Would this be considered a bug in zipfile or zlib, or would they
> instead be considered corrupt files that WinZip happens to handle?


No, they just have unsupported compression methods and comments.

 
> A few of the files are available here:
> http://www.chromosome57.com/python/badfiles/


Thanks, that was very useful to see what was happening.

 

I can supply a patch so the files with comments can be read
if that would help you.  I don't have time to write a whole
decompression #6 method so at least some file would remain unreadable.

If you uncompress and recompress the problem files with WinZip
(or similar), then zipfile will have no trouble with them.

Jim Ahlstrom



From nhodgson at bigpond.net.au  Mon Apr  1 15:37:26 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Mon, 01 Apr 2002 20:37:26 GMT
Subject: Files, Seek, and Readlines. A newbie question
References: <Ts%p8.149254$ZR2.64157@rwcrnsc52.ops.asp.att.net>
Message-ID: <ay3q8.8429$hi7.32827@news-server.bigpond.net.au>

John Purser:

> I'm using a large (>120,000 lines) sorted ASCII text file as an index.  I
> started running into problems as soon as I began to use the .seek() to
> position before a read.  Even when not opened as a binary file seek moves
by
> bytes not lines.  Is there a line oriented seek version?

   If you embed record numbers within the records then you could write a
binary search that would find a particular line without reading all the file
up to that point. Or if you know the average line length then start reading
at line / averageLineLength. These days computers come with lots of
memeory - just read the whole file into memory.

   Neil





From mlh at vier.idi.ntnu.no  Wed Apr 10 23:26:08 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 03:26:08 +0000 (UTC)
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3CB4BB04.1B8DD2A6@bgb-consulting.com>  <3CB4E083.7EC2EDC@engcorp.com>
Message-ID: <slrnaba0ig.efc.mlh@vier.idi.ntnu.no>

In article <3CB4E083.7EC2EDC at engcorp.com>, Peter Hansen wrote:
>Don Garrett wrote:
>> 
>> "M. J. Mix" wrote:
>> >
>> > Is anyone using Python as a spacecraft ground system scripting
>> > language, or onboard (embedded) language?
>> >
>> > Pros and cons when compared with TCL or JAVAScript.  Just
>> > some real life, honest opinions.
>> 
>>   I'm not aware of ANY scripting languages that have been developed in a manor
>> suitable for life critical fail-safe situations. I like Python a lot, and find
>> it very reliable for day to day stuff, but I'd hate to have it embedded in my
>> pace maker. 
>
>Agree.  Unsuitable for pacemakers, or any other life-critical
>application, at this time.  Not inherently unsuitable mind you.

There was a similar discussion a while ago about Air Traffic Control.
Lots of important points came up. Perhaps take a look at it before
repeating all of them in this thread?

Just a suggestion :)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From wilk at flibuste.net  Fri Apr 26 15:50:17 2002
From: wilk at flibuste.net (William Dode)
Date: Fri, 26 Apr 2002 21:50:17 +0200
Subject: reportlab and BaseDocTemplate
References: <20020426180122.3f7ce976.wilk@flibuste.net>
	<y9Y9YKAdgZy8EwN2@jessikat.demon.co.uk>
Message-ID: <20020426215017.6b2d6581.wilk@flibuste.net>

Le Fri, 26 Apr 2002 19:10:37 +0100
Robin Becker <robin at jessikat.fsnet.co.uk> ?crivait:

> In article <20020426180122.3f7ce976.wilk at flibuste.net>, William Dode
> <wilk at flibuste.net> writes
> >hello,
> >
> >I would like to use BaseDocTemplate to make report with different
> >PageTemplate. But the documentation is not enough for me, where can i
> >find some examples ?
> >
> >thanks
> >
> If you direct your mail to the reportlab users list
> see http://two.pairlist.net/mailman/listinfo/reportlab-users 
> you'll get a more reportlab aware audience.
> 
> Anyhow the docs/userguide directory contains a very large example ie the
> userguide itself which uses the base doc template class. It might be
> easier to look first at demos/odyssey/dodyssey.py which has a title
> page, chapter page in single column and later pages in double column.
thanks for theses links, i will have a look

bye

> -- 
> Robin Becker


-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From geoff at gerrietts.net  Thu Apr 11 15:39:55 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 11 Apr 2002 12:39:55 -0700
Subject: Indenting lines 400..457 help
In-Reply-To: <23891c90.0204110907.326eaf0e@posting.google.com>
References: <a929tk$ik2$1@pcls4.std.com> <Fc2t8.90644$w7.7532194@bin6.nnrp.aus1.giganews.com> <a92f90$i57$1@pcls4.std.com> <HNgt8.335984$Gf.29623810@bin2.nnrp.aus1.giganews.com> <23891c90.0204110907.326eaf0e@posting.google.com>
Message-ID: <20020411193955.GA4247@isis.gerrietts.net>

Quoting Paul Boddie (paul at boddie.net):
> 
> [expandtab]
> 
> I tend to use the following line at the end of my Python files:
> 
>   # vim: tabstop=4 expandtab shiftwidth=4
> 
> I think this makes virtually all tab instances soft and equivalent to
> 4 spaces in vim/gvim. Having been a "hard tab user" for years, I
> finally switched to soft tabs out of compatibility with just about
> everything else written in Python - oh, the shame!


As I've said elsewhere in this thread, I actually recommend:

# vim: sts=4:ts=8:et:sw=4

The reason I recommend you leave tabstop at 8 and use softtabstop at 4
is because Python will treat your hard tabs as 8 spaces, so if one (or
more) DO creep in, it will be difficult to see what's wrong when your
eyes aren't treating tabs the same way as the interpreter.

Softtabstop (sts) is in most situations indistinguishable from a
tabstop setting, so you really get the best of both worlds. Yay vim!
:)

--G.

-- 
Geoff Gerrietts <geoff at gerrietts dot net>
-rw-rw-rw-:         permissions of the beast




From djc at object-craft.com.au  Wed Apr 17 17:42:35 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 18 Apr 2002 07:42:35 +1000
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de>
Message-ID: <m3g01um27o.fsf@ferret.object-craft.com.au>

>>>>> "Martin" == Martin v Loewis <martin at v.loewis.de> writes:

Martin> Dave Cole <djc at object-craft.com.au> writes: Yes, the Global
Martin> Interpreter Lock (GIL). When C code runs that was invoked from
Martin> Python code, it holds the GIL, and only one such C code can
Martin> run at any time. So you need to release the lock, using
Martin> Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS.

>>  If the API which is being wrapped by the extension module utilises
>> callbacks, and you want to make those callbacks available to Python
>> code, and you want to do this in a threading environment where the
>> GIL is released when calling the API, then the above will not work.

Martin> You must not call *any* Python API while not holding the GIL,
Martin> yes.

Martin> That doesn't mean that my approach won't work - it is just not
Martin> the full story.

>> This seems to be working fine so far (fingers crossed).

Martin> Sounds good; if possible, you should indeed pass the thread
Martin> state forward to the callback. If you have a lock around the
Martin> library, passing that thread state in a global variable works
Martin> fine (see _tkinter for an example).

That would be another way to do it.  The problem in the database case
is that some API's will take a long time to execute at the server so
it is not a good idea to block all access to the API across all
connections.  It would defeat one of the major benefits of threading.

Martin> If callbacks can come out of arbitrary threads, or if it is
Martin> not possible to pass the thread-state through, you can always
Martin> create a new thread state when entering the callback. In that
Martin> case, you still need to obtain the interpreter somehow.

If you create a new thread state for the callback, what will happen to
Python exceptions raised in the callback?

- Dave

-- 
http://www.object-craft.com.au


From dalke at dalkescientific.com  Mon Apr 29 18:35:20 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 29 Apr 2002 16:35:20 -0600
Subject: comparing str's with ints
References: <20020429115856.R1840-100000@palanthas.neverending.org> <83adrmrwge.fsf@panacea.canonical.org>
Message-ID: <aakhth$fme$1@slb6.atl.mindspring.net>

Kragen:
>Thus, heterogeneous collections are sortable.  This might be a wart;
>once, couldn't raise exceptions, and now we're stuck with its old
>behavior in this case for compatibility.  It's kind of nice that
>heterogeneous collections are sortable, though.

Nice perhaps, but not true.

>>> a = [1, 1j]
>>> a.sort()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: cannot compare complex numbers using <, <=, >, >=
>>>

                    Andrew
                    dalke at dalkescientific.com





From stephen at xemacs.org  Tue Apr 16 09:19:18 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 16 Apr 2002 22:19:18 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabm5cp.nkd.huaiyu@gauss.almadan.ibm.com>
Message-ID: <87ofgjwzl5.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> writes:

    Huaiyu> I don't think it helpful at all to sacrifice the precision
    Huaiyu> of a term that is central to an argument in favor of
    Huaiyu> closeness to its everyday usage.

Well, that's the free/open source software argument in a nutshell,
isn't it.  Free software advocacy is implicitly limited to developers,
and then precision is important.[1]  Open source software advocates
are targeting the whole world, most of whom will never care enough to
be willing to learn precision.  If we don't talk to _them_ in everyday
usage, how can there be communication?

    Huaiyu> I still don't know if _you_ are using them as synonyms for
    Huaiyu> the same concept.

That's because the meaning varies by context.

Basically, when applied to licenses or programs, I consider them
synonyms because it's the same set of licenses and the same set of
programs.[2]

When applied to movements, the goals, or at least the emphases, are
rather different, and therefore the concepts denoted are different.




Footnotes: 
[1]  rms disagrees, of course.  He's right, in the sense that the
freedoms he talks about should morally apply to all human beings.  In
practice, though, only the developers will ever exercise the right,
and the rest realize that, and stop listening quite early on.

[2]  Of course the definitions are worded differently.  So when
referring to the definitions, they have to be different.

And I admit a slight theoretical possibility I could be wrong.  But go
look at the license lists---they're the same.  Except for the original
Perl "Artistic License", which is OSI-certified but in the "non-free"
group at GNU.  So look close---it's a technical DQ for being _too
vague_!  Ie, it's not that it's not free---it's that it's not really a
license according to the FSF.  The exception that proves the rule.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From cengiz at digitalplatform.com  Sun Apr 14 14:54:51 2002
From: cengiz at digitalplatform.com (cengiz at digitalplatform.com)
Date: Sun, 14 Apr 2002 21:54:51 +0300
Subject: Python-list digest, Vol 1 #10481 - 15 msgs
Message-ID: <B3A69981A991B142B6F565E1FDEC5532A89098@dtyilexh1.DIGITURK.LOCAL>


-----Original Message-----
From: python-list-request at python.org [mailto:python-list-request at python.org]
Sent: Sunday, April 14, 2002 9:51 PM
To: python-list at python.org
Subject: Python-list digest, Vol 1 #10481 - 15 msgs


Send Python-list mailing list submissions to
	python-list at python.org

To subscribe or unsubscribe via the World Wide Web, visit
	http://mail.python.org/mailman/listinfo/python-list
or, via email, send a message with subject or body 'help' to
	python-list-request at python.org

You can reach the person managing the list at
	python-list-admin at python.org

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Python-list digest..."





From gh_pythonlist at gmx.de  Sun Apr  7 14:11:49 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 20:11:49 +0200
Subject: Regex question about \d
Message-ID: <20020407181149.GA28549@lilith.my-fqdn.de>

Can somebody please explain this to me:

>>> print re.search(r"\d*", "asdfsfafsw 453 dfsa").group(0)

>>> print re.search(r"\d+", "asdfsfafsw 453 dfsa").group(0)
453

Why doesn't the first regex match the digits?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.4 ?C      Wind: 2.5 m/s




From bergeston at yahoo.fr  Thu Apr 11 11:26:17 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Thu, 11 Apr 2002 17:26:17 +0200
Subject: urllib.urlopen never end up in an HTTPRequestHandler
Message-ID: <a949u7$70ot$1@ID-135695.news.dfncis.de>

I set my box as a proxy on IE (port 8001), I launch the script below from
the command line (dos box  under w2k) and I never reach the "print 'ready to
read' ..." line.
On the other hand, if I try to do "urllib.urlopen" in the interactive window
of PythonWin, it works perfectly well (with the same URL obviously).
Any explanation ?
TIA for any hint.
B.

----ProxyHTTPServer.py------------------------------------------------------
----------------------------------------------------------------------------
------------------------
import BaseHTTPServer, SimpleHTTPServer, urllib

class ProxyHTTPRequestHandler(SimpleHTTPServer.SimpleHTTPRequestHandler):

    def handle(self):
        SimpleHTTPServer.SimpleHTTPRequestHandler.handle(self)
        print '*********************\n Command: %s\n Path: %s\n Request
version: %s\n' % (self.command, self.path, self.request_version)

    def do_GET(self):
        """Serve a GET request."""
        print 'ready to open', self.path
        distant_file = urllib.urlopen(self.path)
        print 'ready to read', self.path
        content = distant_file.read()
        print 'ready to send back', self.path
        self.wfile.write(content)
        print 'everything done with', self.path

    def do_HEAD(self):
        print 'unexpected de_HEAD request'

def run(server_class=BaseHTTPServer.HTTPServer, handler_class =
ProxyHTTPRequestHandler):
    server_address = ('', 8001)
    httpd = server_class(server_address, handler_class)
    httpd.serve_forever()

if __name__ == '__main__':
    run()






From tejarex at yahoo.com  Wed Apr 10 10:03:19 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 10 Apr 2002 14:03:19 GMT
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au> <a8v0mo$j50$1@newsg1.svr.pol.co.uk> <a914t4$g70$1@rzmux02.rz.uni-augsburg.de>
Message-ID: <HCXs8.82135$l7.6504138@bin7.nnrp.aus1.giganews.com>

"Sibylle Koczian" <Sibylle.Koczian at Bibliothek.Uni-Augsburg.de> wrote
in message news:a914t4$g70$1 at rzmux02.rz.uni-augsburg.de...
> >I'm a newbie myself, so this is probably wrong, but wouldn't this
be *even*
> >better?
> >
> >x = 2
> >def wrong(num):
> >    x = x+1
> >    print x

No, this should be

def incx():
  global x
  x = x+1

x = 2
incx()
print x
# prints 3

> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "<interactive input>", line 2, in wrong
> UnboundLocalError: local variable 'x' referenced before assignment
>
> I think the exception is new (result of nested scopes?), but it's
quite
> probable you meant to write

No, message means just what it says: without global declaration, name
bound to new value is taken to be local, but its value is requested
before being set.

> for the reason already stated: call by value, which works on a copy
of the
> parameter inside the function and doesn't change the parameter
itself.
> (Would be the same in C, by the way.)

Python namespaces associate names and objects via object references.
Function calls are call by reference: a *new* name (the parameter name
in the function definition) is bound to each reference within the
function's local namespace.  The reference is copied (and the the
reference count bumped up), not the value.  This 'call-binding' is
undone by any subsequent assignment to the parameter name within the
function.

>>> x = 0
>>> def test(y):
...   print y is x
...   y = y+1
...   print y is x
...
>>> test(x)
1
0

Terry J. Reedy






From phd at phd.pp.ru  Tue Apr  9 14:26:03 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 22:26:03 +0400
Subject: big ints & elliptic curves
In-Reply-To: <a8v338$4sl$1@watserv3.uwaterloo.ca>; from none@none.com on Tue, Apr 09, 2002 at 12:00:29PM -0400
References: <a8v338$4sl$1@watserv3.uwaterloo.ca>
Message-ID: <20020409222603.F27632@phd.pp.ru>

On Tue, Apr 09, 2002 at 12:00:29PM -0400, John wrote:
> Is there any good library for multi-precision arithmetics in Python? Also

   Python alread has long integers of unlimited precision. For floats,
download and install GMPy: http://gmpy.sourceforge.net/

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




From niemeyer at conectiva.com  Fri Apr 12 10:41:18 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Fri, 12 Apr 2002 11:41:18 -0300
Subject: 2.1 stable and 2.2 dev ?
In-Reply-To: <20020412162608.22381c75.wilk@flibuste.net>
References: <20020412162608.22381c75.wilk@flibuste.net>
Message-ID: <20020412114118.C1932@ibook.distro.conectiva>

> I would like to know if the version 2.1.x is the stable version and the
> 2.2.x the dev version ? If it's like that, it's oposite of linux version
> where odd number are for dev...

Right now, there's no such concept. 2.2.X *and* 2.1.X are stable
releases, while (e.g.) 2.2.1aX (alpha), 2.2.1bX (beta), and 2.2.1cX
(release candidate) were development.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From loewis at informatik.hu-berlin.de  Wed Apr 17 13:27:15 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Apr 2002 19:27:15 +0200
Subject: Making a C extension compatible with 2.2
References: <915a998f.0204170914.66aee851@posting.google.com>
Message-ID: <j4zo028ccs.fsf@informatik.hu-berlin.de>

hamish_lawson at yahoo.co.uk (Hamish Lawson) writes:

> Must anything be done to make a C extension compatible with Python
> 2.2? 

Not that I know of. About the only thing is that the GC API has
changed; if you use the old GC API, it will still work fine, except
that your types don't participate in GC anymore.

> I have an extension (not my own) that ran fine under Python 2.1
> on Solaris 2.6. I have since recompiled it for Python 2.2, but
> sometimes I now get tuple indexing errors - and not necessarily at a
> point in my Python code where the loaded extension is actually being
> called upon. Running the same code under Python 2.1 is OK. Would I be
> right in thinking it may be related to the changes in Python 2.2 for
> the new iterator protocol?

It might be that Python 2.2 exposes a bug in the module that was
hidden so far. But I would expect that the bug was always present, and
is just now exposed.

Unfortunately, the only way to find out will be to analyse this in
detail. Using a Python debug build might help, though.

Regards,
Martin



From thomas at gatsoft.no  Tue Apr  2 01:28:02 2002
From: thomas at gatsoft.no (Thomas Weholt)
Date: Tue, 02 Apr 2002 06:28:02 GMT
Subject: CD Device Name
References: <1017720021.720264@cheroke.optimation.co.nz>
Message-ID: <Sbcq8.41$Vme.170713088@news.telia.no>

Take a look at http://home.online.no/~t-weh/releases/cdrom_info.py it might
do what you want. Messy code, but it works on Linux and WinNT/Win2k, not
Win98/ME though.

Any enhancements etc you might come up with or platform related
problems/updates are highly welcome.

Best regards,
Thomas Weholt

"Russell" <Russell.Every at MentumGroup.com> wrote in message
news:1017720021.720264 at cheroke.optimation.co.nz...
> Hi all
>
> Anyone ideas (prefereably Python code) on how to get the label of a CD
under
> Windows 2000?
>
> Thanks in advance.
>
> Russell
>
>




From stojek at part-gmbh.de  Wed Apr 17 05:24:29 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 17 Apr 2002 09:24:29 GMT
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cbd3f0d.3305390@news.easynews.net>

>Yes, the Global Interpreter Lock (GIL). When C code runs that was
>invoked from Python code, it holds the GIL, and only one such C code
>can run at any time. So you need to release the lock, using
>Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS.
>
>Regards,
>Martin

Never heard about that ;-)
But now its clear and everything works. Very effective answer.
Thanks
Marcus



From stephen at xemacs.org  Thu Apr 18 03:31:22 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 18 Apr 2002 16:31:22 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org>
	<12bb92e6.0204091223.738e0798@posting.google.com>
	<i_nt8.15754$ir6.679622@news1.calgary.shaw.ca>
	<GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
	<mailman.1018626392.16328.python-list@python.org>
	<XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca>
	<slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com>
	<mailman.1018655308.18781.python-list@python.org>
	<slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com>
	<87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp>
	<20020417135054.A88552@doublegemini.com>
Message-ID: <87k7r5qx85.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Clark" == Clark C <Clark> writes:

    Clark> On Sat, Apr 13, 2002 at 04:32:15PM +0900, Stephen
    Clark> J. Turnbull wrote:

    >> rms's fundamental position is radical: property rights in
    >> software are morally wrong.  Unfortunately, even rms himself
    >> seems incapable of behaving in a way consonant with this
    >> belief.  (Ie, he clearly believes that he has a special right
    >> to manage development of projects he has initiated, such as
    >> Emacs, gcc, or glibc.

    Clark> There is a difference between "trademark" and "copyright".
    Clark> RMS firmly believes in trademarks;

Interesting but beside the point.  The law provides one primary remedy
for trademark infringement: make the offender stop using the name.
rms does not do this AFAIK (although many people will respect his
priority voluntarily when they fork).  He certainly does try to
control the development of the code itself.  Thus "trademark" is a red
herring here.

    Clark> There is nothing wrong with this,

Nobody said anything was wrong, merely inconsistent.

    Clark> as it does not hurt your ability to fork the code and
    Clark> do-as-you-please.

Hmm.  This is a good point.  rms certainly displays an emotional
attachment to some of the code (especially Emacs)[1], which I have
confused with belief in a property-like right to control.  But what
really is happening is that he uses his control over GNU and FSF
resources (eg, for development and distribution), and moral suasion.
He's not using control over the code itself.

So this is consistent with denial of property right in software.
Rather, he's saying "don't hijack my _effort_ for purposes I don't
like; isn't the software itself enough?"

Thank you, that helped.

    Clark> Not that you have to believe in RMS's vision, but please
    Clark> don't mis-represent what he stands for...

But I do believe in RMS's vision of a free-software-only world.[2] And
to the best of my knowledge, I have not misrepresented his position on
property in software.  (URLs on request; I may have mail copies but
the archives of the mailing list are not easily searchable, and
apparently aren't googled.)


Footnotes: 
[1]  As do we all.

[2]  I just believe that GNU and the FSF are not primus inter pares
any more, and that the concentration of ownership of free software in
the FSF is no longer especially useful.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From dennis at horsethieves.com  Tue Apr  9 10:24:12 2002
From: dennis at horsethieves.com (Dennis)
Date: 9 Apr 2002 07:24:12 -0700
Subject: Beginnger Question
Message-ID: <c4a65c5f.0204090624.710228ff@posting.google.com>

If there's a better forum for asking newb questions, please point me
there!

I'm new to programming.  My only experience up to this point is some
aborted attempts to learn C and Java.  I'm finding Python much easier
to learn than those.  Currently, I'm pushing my way through
"Non-Programmers Tutorial for Python" by Josh Cogliati.

My current problem is this:

x = 2

def wrong(num):
    return num + 1

print wrong(x)
print wrong(x)

I cannot figure out why this (simplified) program won't increment x
and produce 3 and 4.  This is probably indicative of a bigger problem
with understanding functions (and their limitations).  Are there any
sources out there that explain functions more in-depth and clearly?! 
Thanks for any and all help.

Dennis
dennis at horsethieves.com


From BPettersen at NAREX.com  Thu Apr 18 17:01:48 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 18 Apr 2002 15:01:48 -0600
Subject: ActivePython vs. Python
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158DF2@admin56.narex.com>

> From: Bob H [mailto:bobx at linuxmail.org] 
> 
> What is the benefit of getting/using ActivePython over Python?

On windows it comes bundled with the win32 extensions, and it's easier
to install certain third party packages through the activestate package
manager.

-- bjorn




From jessica at sonic.net  Sat Apr 20 13:50:46 2002
From: jessica at sonic.net (jessica at sonic.net)
Date: Sat, 20 Apr 2002 17:50:46 GMT
Subject: Jessica Felix Designs - Art and all that Jazz
Message-ID: <comp.lang.python.8f1c388b@tempest.sonic.net>

If the stale frames can cover strangely, the sick smog may play more earths.

"ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
that incorporates the two passions of its owner, Jessica Felix -
art and jazz.

Jessica is an artist whose medium is jewelry. She creates what she calls
"spirited art," specializing in images from indigenous cultures around the world.

http://www.artandallthatjazz.com

119A Plaza Street
Healdsburg, CA  95448
1-707-433-7900

They are lifting through the shore now, won't climb poultices later.




Why will we waste after Lara burns the hot street's weaver?



From cliechti at gmx.net  Sat Apr  6 13:18:06 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 19:18:06 +0100
Subject: "Memory could not be Read/Written" crashes with C extension DLL... HELP PLEASE!
References: <5a4226f0.0204052223.75aa0318@posting.google.com>
Message-ID: <Xns91E8C487E7F1Acliechtigmxnet@62.2.16.82>

kevin at cazabon.com (Kevin Cazabon) wrote in 
news:5a4226f0.0204052223.75aa0318 at posting.google.com:
> I've gotten past many hurdles so far, but I can't seem to figure out
> what's causing my problems.  The DLL I've built works just fine...
> unless you call pyHasp (see below) multiple times (random.. sometimes
> it works fine for 40 calls, sometimes it works for 300 calls, then
> crashes).
> 
...
> ///////////////////////////////////////////////////////////////
> static PyObject *
> pyHasp(PyObject *self, PyObject *args)
> {
>   // declarations
>   PyObject *pReturn = NULL;
> 
>   PyObject *pLong1 = NULL;
>   PyObject *pLong2 = NULL;
>   PyObject *pLong3 = NULL;
>   PyObject *inBuffer = NULL;
>   int inBufferSize = 0;
> 
>   char *pInBufferAsReadBuffer = NULL;
>   char *pInBufferAsWriteBuffer = NULL;
> 
>   ULONG Service = 0;
>   ULONG SeedCode = 0;
>   ULONG LptNum = 0;
>   ULONG Pass1 = 0;
>   ULONG Pass2 = 0;
> 
>   ULONG p1 = 0;
>   ULONG pp1 = 0;
>   ULONG p2 = 0;
>   ULONG p3 = 0;
>   ULONG pp3 = 0;
> 
>   char *EncodeDecodeBuffer = NULL;
>   ULONG pUL_EncodeDecodeBuffer = 0;
> 
>   // parse the PyObject arguments, assign to variables accordingly
>   // inBuffer is a Python 'B' array, which supports the buffer
> interface
>   if (!PyArg_ParseTuple(args, "llllllllOl", &Service, &SeedCode,
> &LptNum, &Pass1, &Pass2, &p1, &p2, &p3, &inBuffer, &inBufferSize)) {
>     return Py_BuildValue("llls", -9999, -9999, -9999, "Could not parse
> the Python Tuple!");

PyArg_ParseTuple already set up an exception. it usual to "return NULL;" 
here.


>   }
> 
>   // Check the size of the inBuffer for validity
>   if (inBufferSize > 1000) {
>     return Py_BuildValue("llls", -9998, -9998, -9998, "Encode / Decode
> buffer must be less than 1000 characters.");
>   }
> 
>   if (inBufferSize < 8) {
>     return Py_BuildValue("llls", -9997, -9997, -9997, "Encode / Decode
> buffer must be 8 characters or more in length.");
>   }
> 
>   // allocate the memory for the EncodeDecodeBuffer
>   EncodeDecodeBuffer = malloc(inBufferSize);
>   if (!EncodeDecodeBuffer) {
>     return Py_BuildValue("llls", -9994, -9994, -9994, "Could not
> allocate memory for the EncodeDecodeBuffer!");
>   }
>   memset (EncodeDecodeBuffer, 0, inBufferSize);
> 
>   // Obtain a pointer to the Python Buffer Interface for inBuffer
> (readable)
>   if (PyObject_CheckReadBuffer(inBuffer)) {
>     if ((Service != 1) && (Service != 5)) {
>       if (PyObject_AsReadBuffer(inBuffer, &pInBufferAsReadBuffer,
> &inBufferSize) == -1) { // PyObject_AsReadBuffer is fucked up in its
> return values... -1 is error, 0 is success
>         return Py_BuildValue("llls", -9995, -9995, -9995, "Could not
> access the passed string as a Python buffer!");

don't return here - you have a memory leak...
maloc is above and free below...

there follow some "return" with the same problem below

>       }
>     }
>   } 
>   else {
>       return Py_BuildValue("llls", -9995, -9995, -9995, "inBuffer does
> not support Buffer Interface!");
>   } 
> 
>   // the Python values should NOT be modified in place, make copies to
> pass to hasp function
>   pp1 = p1;
>   pp3 = p3;
> 
>   // copy the inBuffer to the EncodeDecodeBuffer, use memcpy so that
> nulls don't cause problems
>   if  ((Service != 1) && (Service != 5)) {
>     memcpy(EncodeDecodeBuffer, pInBufferAsReadBuffer, (unsigned
> int)inBufferSize);
>   }
> 
>   // get a ULONG pointer to EncodeDecodeBuffer
>   pUL_EncodeDecodeBuffer = (ULONG) EncodeDecodeBuffer;

you say "ULONG pointer" in the comment so i would expect that the cast is
"(ULONG *)". ok i see that pUL_EncodeDecodeBuffer
is declared as ULONG and not as pointer.

you're sure that the pointer is needed in that form?
usualy all pointers have the same byte-size e.g. 4 bytes on current PCs.
and when you speak of a ULONG pointer usualy a pointer to a long is meant, 
thus "ULONG *"

>   // send the data to the hasp driver for encoding/decoding/whatever
>   // this function modifies pp1, inBufferSize, pp3, EncodeDecodeBuffer
> (through the pUL_ pointer) in place
>   CallHasp(Service, SeedCode, LptNum, Pass1, Pass2, &pp1,
> &inBufferSize, &pp3, &pUL_EncodeDecodeBuffer);
> 
>   // copy the EncodeDecodeBuffer back to the inBuffer, through a
> pointer to the Python Buffer interface (writable)
>   if  ((Service != 1) && (Service != 5)) {
>     if (PyObject_AsWriteBuffer(inBuffer, &pInBufferAsWriteBuffer,
> &inBufferSize) == -1) {
>       return Py_BuildValue("llls", -9994, -9994, -9994,
> "PyObject_AsWriteBuffer Failed!");
>     }
>     memcpy(pInBufferAsWriteBuffer, EncodeDecodeBuffer, (unsigned
> int)inBufferSize);
>   }
> 
>   free(EncodeDecodeBuffer);
> 
> 
>   // build pp1, inBufferSize, pp3 into PyObjects for returning
>   pLong1 = PyLong_FromLong(pp1);
>   pLong2 = PyLong_FromLong(inBufferSize);
>   pLong3 = PyLong_FromLong(pp3);
> 
>   // build pLong1, pLong2, pLong3, pString into a tuple to return
> 
>   pReturn = PyTuple_New(4);
>   PyTuple_SetItem(pReturn, 0, pLong1); 
>   PyTuple_SetItem(pReturn, 1, pLong2);
>   PyTuple_SetItem(pReturn, 2, pLong3);
>   PyTuple_SetItem(pReturn, 3, inBuffer);

you could use Py_BuildValue here as well.

>   return pReturn;
> }
> 

why are you not using exceptions for the error handling? i think it would 
be much cleaner for the usage of this function.

-- 
Chris <cliechti at gmx.net>



From tim.one at comcast.net  Fri Apr 12 14:46:18 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 12 Apr 2002 14:46:18 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans,
 integer   division,backwards compatibility; where is  Python  going?)
In-Reply-To: <1018508767.552240@yasure>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKHONAA.tim.one@comcast.net>

[Donn Cave]
> ...
> The hostility directed towards Mr. Lumberjack suggests that maybe he
> was smart to use a nom de plume for this message.  Tomorrow he can
> post under his real name and interact on amicable terms with the
> people who want to chase him off today for criticizing our leader.
> That's fine, I guess, as long as you all are having fun, but don't
> kid yourselves that this is any kind of a service to the community.
> It's actually kind of a sorry picture.

Donn, what does this have to do with pipes <wink>?

Do a google search on Lumberjack's c.l.py posting history if you want to
know where the hostility really comes from.  It starts with pure flamebait:

    This from a person who is constantly trying to score points on
    Usenet.  A check of google shows you have an inflated notion of
    yourself.  Maybe you should stop measuring others by your own limited
    character.

That wasn't directed at me, but I don't care who it was directed at -- it's
pure troll.  The next Lumberjack post wasn't much better, and I stopped
reading then.  If Lumberjack has something useful to say, perhaps a new
pseudonym would evade the kill files.





From gashton at cmedltd.com  Mon Apr  8 08:06:21 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 08 Apr 2002 13:06:21 +0100
Subject: PEP 285: Adding a bool type
In-Reply-To: <200204070151.g371pEdk018077@ratthing-b246.strakt.com>
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com>
	<200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net>
	<200204061913.g36JDcdj017764@ratthing-b246.strakt.com> 
	<200204062313.g36NDV428946@pcp742651pcs.reston01.va.comcast.net>  
	<200204070151.g371pEdk018077@ratthing-b246.strakt.com>
Message-ID: <1018267581.1128.7.camel@tp6>

On Sun, 2002-04-07 at 02:51, Laura Creighton wrote:
> But we agree that refactoring is not good for its own sake, it would
> be better to have not needed the refactoring because we got it correct
> the first time.

Sure, but I'm not convinced that (in general) getting things right the
first time is an achievable goal. I suspect that (in order to achieve
maintainable code) refactoring is a must.

> But if we
> are writing functions that are to be used by others in libraries someplace,
> we have to try to minimize the effect of our refactoring on others.

A valid goal, but is it (in this case) worth the cost of legibility?
Reading between the lines on the last few posts, I'm concerned that
you're advocating code like this:

  if foo.getStatus() == 2:
      # do something

instead of:

  if foo.isValid() and foo.isPaid():
      # do something

> If other people have to refactor their code, every time that we have
> to refactor ours, they will get pissed.

If you add another return value to getStatus() they are going to have to
modify their code to make use of it, right? I don't see the problem with
making changes that will enable them to write more legible code, instead
of returning multiple values.

I don't generally write libraries for use by customers, so I'm probably
wrong...

-- 
Graham Ashton





From ballabio at mac.com  Tue Apr 16 07:12:53 2002
From: ballabio at mac.com (Luigi Ballabio)
Date: Tue, 16 Apr 2002 12:12:53 +0100
Subject: newbie-question: extending python with c++
In-Reply-To: <3CBBEA69.7BA7EAAC@kfunigraz.ac.at>
References: <mailman.1018890724.31468.python-list@python.org>
 <3CBB1AE8.E2843017@kfunigraz.ac.at>
 <a9f8ru$lef$2@peabody.colorado.edu>
Message-ID: <5.1.0.14.0.20020416120622.00a53680@mail.mac.com>

At 11:10 AM 4/16/02 +0200, Siegfried Gonzi wrote:
>I think it is not inappropiate to post here the instruction (thought I
>got it due to private mail; but there is no private junk in it and hence
>I believe the original author does not mind: author: Luigi Ballabio; I
>augmented Luigi's B. stuff with the pointer stuff and some other
>explanations).

I don't mind at all. However, there's one point which can be further 
detailed : I didn't in my original post since you specified that you 
weren't using SWIG with the -shadow option.
If one does instead, the shared extension should no longer be called 
example.so but examplec.so, and a Python file example.py is generated.
In this case, the call to setup() should read:

setup(name         = "Example",
        description  = "Example Python extension",
        py_modules   = ["example"],                   # add the .py file
        ext_modules  = [Extension("examplec",         # rename the extension

["C:/Wissenschaft/SWIG-1.3.11/ex/example_wrap.c",
                                   "C:/Wissenschaft/SWIG-1.3.11/ex/example.c"],
                                  libraries = [],
                                  define_macros = [],
                                  include_dirs =
[C:/Wissenschaft/SWIG-1.3.11/ex/],
                                  library_dirs = [],
                                  extra_compile_args = [],
                                  extra_link_args = [])
                       ]
       )

Bye,
         Luigi





From sholden at holdenweb.com  Mon Apr  8 17:15:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 17:15:18 -0400
Subject: beginners syntax question
References: <3caf47a5@news1.homechoice.co.uk>
Message-ID: <nNns8.3044$vT6.1793@atlpnn01.usenetserver.com>

"Tony" <tonyl at homechoice.co.uk> wrote in message
news:3caf47a5 at news1.homechoice.co.uk...
> Hi,
>
> I'm trying to call the update() method of several objects ie:
>
> object1.update()
> object2.update()
> etc
>
So, you've bound each of the objects to a different variable. How did you
perform these bindings? If you wrote

object0 = SomeClass(0)
object1 = SomeClass(1)

etc., then you really ought to be asking yourself why you didn't write:

olist = []
for i in range(6):
    olist.append(SomeClass(i))

or some such construct. Then you can do

for i in range(6):
    olist[i].update()

> I'd really like to do
>
> for x in range(7):
>   "object"+str(x).update
>
> but can't figure outthe syntax, is this possible?
>
As you've already been shown, yes, it's possible. Unfortunately the answer
to "is this desirable?" is almost always a resounding "No".

regards
 Steve






From jimmy at retzlaff.com  Wed Apr 24 09:47:01 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Wed, 24 Apr 2002 06:47:01 -0700
Subject: variable substitution
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F2D98@MAYOR.retzlaff.com>

Steve Holden [mailto:sholden at holdenweb.com] wrote
> "Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote ...
>> How do I get the %(variable)s thing to work. I need to substitute a
>> variable into a print statement so I can send links with sessionId's
>> attached back to the browser. I'm trying this at the moment, and It
>> says invalid sysntax.
>> print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
>> Type</a>'
>> I saw this in a document written by Guido that I found here:
>> http://www.w3j.com/6/s3.vanrossum.html
>
> You need to add a "%" operator and put a dictionary on its RHS. The
> parenthesised words are then used to look up the appropriate values in
the
> dictionary (technically, any mapping type will work). e.g.:

The built-in function locals() can prove to be a very handy source for
that dictionary.

Jimmy




From gashton at cmedltd.com  Wed Apr 17 12:58:00 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 17 Apr 2002 17:58:00 +0100
Subject: UTC equivalent of time.mktime() ?
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com>
References: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com>
Message-ID: <1019062680.653.398.camel@tp6>

On Wed, 2002-04-17 at 17:11, Mark McEahern wrote:
> Why not just use the mxDateTime library and just do something like this:
> 
> 	mx.DateTime.utc().ticks()

I appreciate that I could (and thanks for the suggestion), but does this
really mean that via the standard library there is no easy way to
convert "1 Jan 1970" to 0 (i.e. string to seconds since epoch)?

Perhaps I should inspect the mxDateTime code...

-- 
Graham Ashton





From noomnews at yahoo.com  Tue Apr  2 23:55:47 2002
From: noomnews at yahoo.com (Noom News)
Date: 2 Apr 2002 20:55:47 -0800
Subject: ctrl-c and os.system problem
References: <mailman.1017777544.11640.python-list@python.org> <a8dbup$hec$1@nntp6.u.washington.edu>
Message-ID: <e27b96c5.0204022055.74c3d570@posting.google.com>

"Donn Cave" <donn at u.washington.edu> wrote in message news:<a8dbup$hec$1 at nntp6.u.washington.edu>...
> Quoth Noom Knews <noomnews at yahoo.com>:
> | I have a simple program like this
> |
> | os.system("ls -lR / > /tmp/ls.out")
> | print "end"
> |
> | When I press ctrl-c, the "end" message is still print.
> |  I'd like to terminate the whole process when pressing
> | ctrl-c instead of just "end".
> |
> | I tried to catch the signal already but can't catch
> | the SIGINT or anything else at all during the
> | os.system call.   This does not work either
>  ...
> | This was tried on Redhat Linux with python 1.5.2
> 
> NetBSD's system(3) man page explains that system ignores SIGINT
> and SIGQUIT.  I assume the same must be true on Red Hat Linux,
> and Python calls C system().
> 
> Maybe you could use spawnv()?
> 
> 	Donn Cave, donn at u.washington.edu

Thanks!  The strange thing is that SIGINT kill the os.system but was
not propated to kill the whole python process.  This behavior made
python different from perl where the whole process would be killed.

-Nick


From billludden at email.msn.com  Tue Apr 23 12:54:07 2002
From: billludden at email.msn.com (billludden)
Date: Tue, 23 Apr 2002 10:54:07 -0600
Subject: Newbie: Running a script on WindowsXP
Message-ID: <#dmZtdu6BHA.1188@cpimsnntpa03>

Hi,

I am trying to figure out how to run a Python21 script
from the "Run" box on WindowsXP Home edition.
When I run it, the Command Prompt box just briefly
flashes on the screen and closes itself.  I have tried
using the -i option like you would use in Windows 98
but it doesn't seem to work.  I also tried appending
Autoexec with:
    SET PATH=.........;C:\PROGRA~1\PYTHON21

Does anyone understand what I'm doing wrong?
(incidentally, I had the same problem with "hello world"
 when I tried it in C)

Thank you!
Bill






From donn at drizzle.com  Sat Apr  6 14:17:51 2002
From: donn at drizzle.com (Donn Cave)
Date: Sat, 06 Apr 2002 19:17:51 -0000
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org>
Message-ID: <1018120666.804421@yasure>

Quoth Guido van Rossum <guido at python.org>:
...
| You clearly fall in the most conservative camp.  I just had a long
| exchange with someone who pleaded strongly for moving forward at a
| higher pace, breaking more existing code so that the perfect language
| can be obtained sooner.  I will have nothing of that, but I hope you
| understand my predicament: for every user like you who complains that
| things change too fast, there's someone with an equally strong desire
| for faster change.  I have to pick a middle ground (and it's easier to
| say "don't upgrade" than "wait a year").

I sure fail to see the point of this middle ground.  On one hand, you
have a language that is manifestly very useful to programmers in a variety
of disciplines but fails to be the perfect language.  On the other, you
have a vision of the perfect language, which in principle ought to be
more useful but definitely will break some significant amount of code
written today.  In between, you have all the problems of both - breaks
code today, doesn't prevent tomorrow's breakage, and isn't the perfect
language anyway.  Major Python heads fight tooth and nail against changes
that they might well accept without complaint as features of a really
overhauled language.  Developers turn away, platforms like Redhat who
could ship Python are put in a difficult position.  In the present release
schedule, "don't upgrade" may be easy to say but it isn't really supported -
neither at the core, where Tim's quick to point out that we're on our own
for bug fixes, nor among the external contributors who have no notion that
any such thing is going on, nor among client sites who are apt to install
whatever's current.  Is this fun?  Doesn't sound like it.  Will it be more
fun to artificially prolong the process?  Surely it will not.  Why not
just do what you have to do, and concede that a few releases hence will
be bleeding edge, not for general production use?

	Donn Cave, donn at drizzle.com


From loewis at informatik.hu-berlin.de  Wed Apr  3 04:13:45 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Apr 2002 11:13:45 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <j44rit16wm.fsf@informatik.hu-berlin.de>

vio <vmilitaru at sympatico.ca> writes:

> I need to translate a (rather voluminous 100MB) java program to
> python. After a few searches on google, apparently the best tools
> for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> on the Lex-YACC HOWTO, I would be very interested in hearing of
> others' successful experiences/feedback at this. Any pointers of
> interest would similarily be appreciated.

Not that I can offer any inside - but I think many people would be
interested to hear your results; whether the project fails or
succeeds. I'm also sure some people might be interested in the
converter itself. Please do consider publishing it, atleast if you
have to give up, so that other might be able to continue.

Regards,
Martin



From just at xs4all.nl  Sun Apr  7 01:28:01 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sun, 07 Apr 2002 08:28:01 +0200
Subject: Argument against iterable ints...
References: <mailman.1018117586.26522.python-list@python.org>
Message-ID: <just-D808D9.08280107042002@news1.xs4all.nl>

[Magnus Lie Hetland]
> Just a situation that cropped up recently: While writing a little
> generator for flattening nested iterables, I used
>
>   try: iter(foo)
>   except: return foo
>   else:
>       ....
>
> (of course I could have used a for loop directly in the try clause --
> that doesn't change anything.)
>
> If integers became iterable, this sort of code would suddenly become
> infinitely recursive. Given the propensity for using exceptions in
> this manner rather than type checking in Python, this seems like a
> weighty argument to me...

No matter how I dislike iterable ints, this argument stinks, as the 
situation you describe already exists with strings: each element of a 
string is itself a string and...

Just


From jwickard**nospam** at litriusgroup.com  Fri Apr 12 07:14:18 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Fri, 12 Apr 2002 11:14:18 GMT
Subject: Built in file copy utility
References: <Ozrt8.4709$c6.399223@typhoon.mn.ipsvc.net> <3CB64FD4.62FA1AAA@metaslash.com>
Message-ID: <ekzt8.4743$c6.411174@typhoon.mn.ipsvc.net>

"Neal Norwitz" <neal at metaslash.com> wrote in message
news:3CB64FD4.62FA1AAA at metaslash.com...
> joel8bit wrote:
> >
> > I just wanted to check and see if there was a built in module to copy
files
> > from one directory to another.
>
> >>> import shutil
> >>> print dir(shutil)
> ['__all__', '__builtins__', '__doc__', '__file__', '__name__',
'_build_cmdtuple', 'copy', 'copy2', 'copyfile',
> 'copyfileobj', 'copymode', 'copystat', 'copytree', 'os', 'rmtree', 'stat',
'sys']
> >>> print shutil.copyfile.__doc__
> Copy data from src to dst
>
> Neal

Thank you very much sir.  I did RTFM before I posted, I was just looking for
said utility under the os, os.path module.




From com-nospam at ccraig.org  Fri Apr 19 22:29:36 2002
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 19 Apr 2002 22:29:36 -0400
Subject: why no 'length' method in sequences ?
In-Reply-To: <a9q8f4$svn$1@aphrodite.grec.isp.9tel.net>
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net>
	<zwFv8.36236$vF6.1097456@news2.tin.it>
	<a9nbuq$916$1@aphrodite.grec.isp.9tel.net>
	<uc14cjalpru3b5@news.supernews.com>
	<a9q8f4$svn$1@aphrodite.grec.isp.9tel.net>
Message-ID: <t1wbscft84v.fsf@kermit.wreck.org>

"Richard Gruet" <rjgruet at yahoo.com> writes:

> Actually, as Alex pointed out, a single len() method in a
> hypothetical 'sequence' base class would be mythical, simply because
> there is no common implementation of len() for all kinds of
> collection (I just forgot this!).

I have the utmost respect for Alex, so I don't understand why I can't
figure out what he is saying, but why couldn't len(a) be equivalent to
a.len()?

The len() builtin calls PyObject_Size(), which, in turn, tries two things: 
obj->ob_type->tp_as_sequence->sq_length()
then
obj->ob_type->tp_as_mapping->mp_length()

In the case of class instances both are defined to execute the __len__
method.

Also, the type metatype defines __len__ to be PyObject_Size for all of
it's subtypes iff either of those two C functions is defined.  I don't
see why naming it "len" instead of (or even in addition too) "__len__"
wouldn't work.

The main problem I see with this is that it wouldn't have worked
before you got the ability to subtype in Python2.2, so historically it
had to be a method before.  Plus adding it now means you either have a
stub "len" method which may not work, or you have a major backward
incompatibility.  

I personally disagree with adding it now, but not because it can't be
done.

-- 
Christopher A. Craig <com-nospam at ccraig.org>
"Tragedy is when I cut my finger.  Comedy is when you fall in an open
sewer and die."  Mel Brooks.




From loewis at informatik.hu-berlin.de  Mon Apr 15 07:26:44 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Apr 2002 13:26:44 +0200
Subject: Building Python 2.1.x on Solaris
References: <32dd230a.0204150117.3f4a5bcc@posting.google.com>
Message-ID: <j4y9fpmccr.fsf@informatik.hu-berlin.de>

oliver.erlewein at sqs.de (Oliver Erlewein) writes:

> I've also experienced similar behaviour on Linux. The thing is, that
> as soon as I use precompiled stuff (RPM's) all works fine and
> threading is no problem.

I don't think you can change the behaviour of Python through the way
in which you build it. So if you are seeing different behaviour, most
likely, you've been using different source code.

If you have a specific RPM that behaves differently, locate the source
RPM and analyse the build options, and any potential patches that
might be applied before building it.

Regards,
Martin



From ASkwar at DigitalProjects.com  Fri Apr 26 17:20:06 2002
From: ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 23:20:06 +0200
Subject: Removing dupes from a list
In-Reply-To: <XFMail.20020426135630.shalehperry@attbi.com>
References: <20020426203946.GV21205@teich.Garten.DigitalProjects.com> <XFMail.20020426135630.shalehperry@attbi.com>
Message-ID: <20020426212006.GY21205@teich.Garten.DigitalProjects.com>

?Sean 'Shaleh' Perry? sagte am 2002-04-26 um 13:56:30 -0700 :
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560

Thanks, but connections to activestate.com simply time out :(  However,
a traceroute to aspn.activestate.com doesn't show any unusual long hops.
But AS is reachable from other networks.  Strange.

Anyhow, seems like uniquer() from the 1st comment is what I'm looking
for, because unique() also looses order just like my two "functions".

Thanks.

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 2 days 16 hours 12 minutes




From Answer.via.news.please.prikryl at skil.nospam.cz  Wed Apr 10 10:39:37 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Wed, 10 Apr 2002 16:39:37 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <3CB427D3.7020404@geneva-link.ch> <3cb432ca$0$231$edfadb0f@dspool01.news.tele.dk>
Message-ID: <a91itn$2fgp$1@news.nextra.cz>

Anders J. Munch wrote:
> Boris Borcic wrote...
> > Anders J. Munch wrote:
> >
> > > Pythons truth values are not "Something or
> > > Nothing".  The phrase seems to have caught on
> > > following Laura's post, but I believe it is
> > > misleading.
> > >
> > > [] and {} are objects, they are objects with
> > > identity, and they are mutable objects.  They are
> > > empty, but they are by no means "nothing".
>
> > I could discuss the "Something" side of the
> > matter, but I feel the point you raise here very
> > much boils down to "the type of zero is a
> > manifest information that disallows viewing the
> > typed zero as a nothing".

> No, I was specifically discussing how objects
> created by the expressions []and {} are poorly
> described by "nothingness", and how for this
> reason equating pythonic falsity with nothingness
> can lead you astray, as shown in the example.

I fully agree with Anders.  Indroducing fuzzy
concepts like something vs. nothing may be nice
in philosophical discussions.  Programming
requires somehow more exact approach.  Thinking
in true/false terms is much clearer and
definitely not more difficult.

Here is the snippet from my "True/False story...
(PEP 285)".  The HTML version is temporatily available at
http://www.skil.cz/test/TrueFalseStory.html.


The implicit conversion
~~~~~~~~~~~~~~~~~~~~~~~

The key point which clearly fits with conditional
expressions of the boolean type is called **the
implicit conversion concept**.

In my opinion, explaining the concept of the
*implicit conversion* together with the *context*
where the construct is used allows one to explain
the things better than *something* vs. *nothing*
concept.  I am going to show that.

The Python 2.2.1c2 documentation (i.e. before
bool era) says in *2.2.1 Truth Value Testing*

  Any object can be tested for truth value, for
  use in an if or while condition or as operand
  of the boolean operations below. The following
  values are considered false:

  - None
  - zero of any numeric type, for example, 0, 0L, 0.0, 0j.
  - any empty sequence, for example, '', (), [].
  - any empty mapping, for example, {}.
  - instances of user-defined classes, if the
    class defines a __nonzero__() or __len__()
    method, when that method returns zero.

  All other values are considered true -- so
  objects of many types are always true.

  Operations and built-in functions that have a
  boolean result always return 0 for false and 1
  for true, unless otherwise stated. (Important
  exception: the boolean operations "or" and
  "and" always return one of their operands.)

In other words, it says that in the *boolean
context* the named cases are *converted* to
*false* or *true* values respectively.  So, there
is nothing like...

::

  > {} is a dictionary-shaped nothing. [] is a
  > list-shaped nothing. 0 is an integer-shaped
  > nothing. 0.0 is a float shaped nothing.

Again, it only obscures the things.  The students
should have asked "What do you mean by the
*shaped nothing*?".  (At least for me, it is far
more obscure than the simple notion of yes/no or
true/false.)

Compare it with what the documentation already
says (in different wording):

  When an object appears in a boolean context,
  the implicit conversion is done.  The following
  objects are implicitly converted to False:

  - None
  - zero of any numeric type, for example, 0, 0L, 0.0, 0j.
  - any empty sequence, for example, '', (), [].
  - any empty mapping, for example, {}.
  - instances of user-defined classes, if the
    class defines a __nonzero__() or __len__()
    method, when that method returns zero.

  All other objects are implicitly converted to True -- so
  objects of many types are always converted to True.

--
Petr Prikryl (prikrylp at skil dot cz)





From brueckd at tbye.com  Mon Apr  8 11:01:08 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 8 Apr 2002 08:01:08 -0700 (PDT)
Subject: __init__.py
In-Reply-To: <3CB19437.C1B2B3E9@gol.ge>
Message-ID: <Pine.LNX.4.33.0204080759210.20374-100000@ns0.tbye.com>

On Mon, 8 Apr 2002, Giorgi Lekishvili wrote:

> I wonder if an experienced developer in Python gives me a few hints on
> how to make the initfile of a package.
> I find the section in the tutorial
> http://www.python.org/doc/current/tut/node8.html#SECTION008400000000000000000
>
> not very much helpful:(

>From the command line of the package's directory:
touch __init__.py

-Dave





From emile at fenx.com  Sun Apr 21 20:05:55 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 22 Apr 2002 00:05:55 GMT
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au>
Message-ID: <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net>

Neville Franks
> I enter 'import difflib' which I assume is necessary and get no
complaints.
> Then if I type 'ndiff' I get: NameError: name 'ndiff' is not defined.
If I
> type: 'ndiff file1 file2' I get: SyntaxError: invalid systax.
>
> I've also tried:
> >>> s = SequenceMatcher( None, "abcd", "bcde" )
> and I get:
> NameError: name 'SequenceMatcher' is not defined.

Those names live in the difflib namespace.  When using them like this,
you'll need to prefix them.  Try difflib.SequenceMatcher


--

Emile van Sebille
emile at fenx.com

---------



From Russell.Every at MentumGroup.com  Mon Apr  1 23:59:36 2002
From: Russell.Every at MentumGroup.com (Russell)
Date: Tue, 2 Apr 2002 14:59:36 +1000
Subject: CD Device Name
Message-ID: <1017720021.720264@cheroke.optimation.co.nz>

Hi all

Anyone ideas (prefereably Python code) on how to get the label of a CD under
Windows 2000?

Thanks in advance.

Russell




From jb at cascade-sys.com  Mon Apr 29 17:04:25 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 14:04:25 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>
		<3CCD9649.98AED86F@cascade-sys.com> <20020429130136.5b0c34f7.logiplexsoftware@earthlink.net>
Message-ID: <3CCDB559.A1BF0E47@cascade-sys.com>

Cliff Wells wrote:

> Well that's a stupid statement - *your* head is in the clouds.

Oh yeah?  Fuck you.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From geoff at gerrietts.net  Mon Apr 15 16:50:40 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 15 Apr 2002 13:50:40 -0700
Subject: String interpolation question
In-Reply-To: <a9fcmo$npi$1@peabody.colorado.edu>
References: <a9fcmo$npi$1@peabody.colorado.edu>
Message-ID: <20020415205040.GB12476@isis.gerrietts.net>

Quoting Fernando P?rez (fperez528 at yahoo.com):
...
> y=[5,6]
> i=1
> 
> I want to print y[i]. The following fails:
> 
> print '%(y[%(i)s])s' % locals()
> 
> as does the simpler
> 
> print '%(y[%(i)s])s' % locals()
> 
> but I don't want that because in my real problem, the string is long and I 
> don't want the dangers of positionally matched %s arguments.

Given your constraints, I think I'd prefer:

>>> print "%(val1)s" % { 'val1': y[i] }

Or if you insist on using locals() (a habit I find at least as
dangerous as positional arguments):

>>> z = y[i]
>>> print '%(z)s' % locals()
6

On the other hand, if you're looking at a long y and multiple values
of i (though your example provides no such hint), I'd prefer:

d = {}
for indx in range(len(y))
  d['y_%d'%indx] = y[indx]
d.update(locals())

s = '%%(y_%d)s' % (i,)
print s % d

If none of these solutions seem to satisfy your needs, maybe you
should be a little more explicit about the problem domain. If one of
the requirements is that it be a one-liner, I'm not sure if I can
help.

On the other hand, if what you really want is to push the string
interpolation PEP, well, I'm just trying to help, not trying to argue
for or against a feature. As long as I don't hafta use the feature,
and I don't hafta work around the feature, I don't mind if the
feature's present.

Thanks,
--G.

-- 
Geoff Gerrietts             "If life were measured by accomplishments, 
<geoff at gerrietts net>     most of us would die in infancy." 
http://www.gerrietts.net/       --A.P. Gouthey




From max at alcyone.com  Thu Apr  4 00:13:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 21:13:16 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com>
Message-ID: <3CABE0EC.3EA3D4C5@alcyone.com>

Paul Rubin wrote:

> I think there will need to be a format code for truth values:
> 
>   ("%b" % True) => "True"
> 
> etc.

That sounds very reasonable to me; %b should do an implicit bool(...)
just like %s does an implicit str(...).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From tim.one at comcast.net  Fri Apr 12 21:49:57 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 12 Apr 2002 21:49:57 -0400
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
In-Reply-To: <slrnabf05v.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEMBONAA.tim.one@comcast.net>

[Huaiyu Zhu]
> So RMS says using "open source" to mean "free" is not what FSF
> prefers, even though many people use it that way.

That's so.

> It is definitely not what OSI prefers, either.

That's not.  <http://www.opensource.org/> uses nothing else, although it's
often spelled Open Source there.  From OSI's FAQ:

    How is "open source" related to "free software"?

    The Open Source Initiative is a marketing program for free software.
    It's a pitch for "free software" on solid pragmatic grounds rather
    than ideological tub-thumping.  The winning substance has not
    changed, the losing attitude and symbolism have.

As marketing collateral, I suppose that paragraph could stand its own
attitude adjusmment <wink>.





From gimbo at ftech.net  Thu Apr  4 05:08:02 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 4 Apr 2002 11:08:02 +0100
Subject: duelling pythons
In-Reply-To: <lky9g4o3z7.fsf@elios.maths.bris.ac.uk>
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de> <3cab55f4.61190667@news.laplaza.org> <a8fn61$s1nk7$1@ID-41653.news.dfncis.de> <lky9g4o3z7.fsf@elios.maths.bris.ac.uk>
Message-ID: <20020404100802.GC1779@andy.tynant.ftech.net>

On Wed, Apr 03, 2002 at 09:40:45PM +0000, Michael Hudson wrote:
> Tobias Klausmann <klausman-un030402 at tuts.net> writes:
> 
> > So far, I've only written libraries/modules (no shebang) or console
> > apps that are at least semi-interactive. But if I'd write some such
> > utility that might run cronned, I'd surely stumble across the system
> > where cron doesn't know about /usr/local/bin/python15 =)
> 
> The distutils can help here.

How can distutils help with the problem of what to put on the first
line of my scripts?

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From bokr at oz.net  Wed Apr 10 17:35:40 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 21:35:40 GMT
Subject: Baffled by CGI cross-browser issues... help sought :)
References: <Xns91ECB3C2132FBPaCmAnRDLM@161.48.128.20>
Message-ID: <a92b7c$lmb$0@216.39.172.122>

On 10 Apr 2002 15:37:38 GMT, Fred Pacquier <fredp at mygale.org.nospam> wrote:
[... cgi problems ...]

>* on Netscape 4.x things sort of work, except that each time I submit a 
>form I get an error message like "A network error occurred while receiving 
>data (connection reset by peer)". After dismissing the dialog the resulting 
>page usually displays correctly (although sometimes truncated).
>
Possibly your CGI output is set up as buffered and not getting flushed?

Regards,
Bengt Richter


From dswegen at software.plasmon.com  Wed Apr 17 09:12:04 2002
From: dswegen at software.plasmon.com (Dave Swegen)
Date: Wed, 17 Apr 2002 14:12:04 +0100
Subject: Python signal handling
Message-ID: <20020417131204.GG30292@software.plasmon>

I've recently been bitten by a nasty (undocumented) feature in Perl,
where installing a signal handler to catch SIGCHLD will eventually cause
a core dump. So before I venture into any daemon handling in python I
was wondering if Python too suffers from problems with signal handling?

I believe the reason for perl's problem is that it isn't re-entrant, and
thus two rapidly incoming signals will cause perl to fo *Pooof*.

Any info will be much appreciated.

Cheers
    Dave




From jidanni at deadspam.com  Wed Apr 24 02:05:21 2002
From: jidanni at deadspam.com (Dan Jacobson)
Date: 24 Apr 2002 14:05:21 +0800
Subject: replacing Chinese chars with their spellings
Message-ID: <m2g01lfx7i.fsf@jidanni.org>

Before I start learning python, here's what I want to do: I have a
table of Hakka Chinese words and their pronunciations.  I scan a file
and replace any Hakka ["big5" 2, 4, 6 ... byte long strings] there
with their pronunciations.  If it were just one character [two byte]
words I would use the "c2t" program.  Is there a template that munches
forth in a file and replaces the longest match in a database
before moving on?
-- 
http://jidanni.org/ Taiwan(04)25854780


From robin at jessikat.fsnet.co.uk  Mon Apr 29 18:18:08 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 29 Apr 2002 23:18:08 +0100
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>
 <just-9E8624.12281026042002@news1.xs4all.nl>
 <ucjicvlrf7djd3@news.supernews.com>
 <b0f083db.0204290751.52c145f4@posting.google.com>
Message-ID: <Io2j9JAgacz8EwST@jessikat.fsnet.co.uk>

In article <b0f083db.0204290751.52c145f4 at posting.google.com>, Jeremy
Hylton <jeremy at alum.mit.edu> writes
......
>LOAD_NAME is an old-fashioned beast that looks in the current
>namespace, the global namespace, and the builtin namespace.  It never
>raises an exception unless the name can't be found in any of those
>namespaces. Note in particular that it's always three namespaces; in
>the case of a nested block, enclosing namespaces are not searched.
.....
thanks, but my primitive investigations didn't reveal any LOAD_NAME ops.
The one op I saw was LOAD_CLOSURE, but of course I'm not really sure
what I'm seeing with this.


>This stuff isn't documented, so it's no surprise that it's confusing. 
>But I think the documentation will mostly be of the form:  The results
>of XXX are undefined.  (The CPython implements currently does this
>...)
>
>Jeremy
the one thing I think is that other languages eg pascal/modula have a
better/simpler grasp of what humans see in programs. Lexical scope ie
higher on the page and further left means precede is a 'good' rule. The
'local' idea that anything that is assigned to is 'local' is much harder
since that can be realised very distantly. I assume there are reasons
for the current 'oddness', but they seem to be hidden from us poor
lusers.
-- 
Robin Becker


From merman at freenet.de  Sun Apr 14 16:01:45 2002
From: merman at freenet.de (merman)
Date: Sun, 14 Apr 2002 22:01:45 +0200
Subject: Jython help pls
References: <3cb9c6c3.115241899@news.mweb.co.za> <Xns91F0D381895A5cliechtigmxnet@62.2.16.82>
Message-ID: <3cb9ddfb$0$150$9b622d9e@news.freenet.de>

"Chris Liechti" <cliechti at gmx.net> schrieb im Newsbeitrag
news:Xns91F0D381895A5cliechtigmxnet at 62.2.16.82...
> sqrt_2 wrote in news:3cb9c6c3.115241899 at news.mweb.co.za:
> > Can anyone help with Jython compiler error messages. I find it
> > difficult to understand what the errors are as I have only just
> > started with this product.
>
> maybe you want to work through a python tutorial? the one one python.org
> is for CPython but 90% or more will apply to Jython too.
>
> > Messages  like:
> > 'NameError: Java'
> note that python and java are case sensitive. you probly meant "from java
> import ..." (small 'J')
>
> > What does that mean?? Is there a reasonable site where I can find
> > explanations for common error messages? The Jython site FAQ is no help
> > at all.
>
> The python docs should help
>
>
>
> --
> Chris <cliechti at gmx.net>
>


... or try www.jython.org

o-o

thomas





From gmcm at hypernet.com  Fri Apr 26 09:30:26 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 26 Apr 2002 13:30:26 GMT
Subject: shipping a python app
References: <a9js5c$2m9$1@panix2.panix.com> <upfv8.14124$hl2.9868@atlpnn01.usenetserver.com> <20020425231732.GJ21205@teich.Garten.DigitalProjects.com> <mailman.1019777991.16703.python-list@python.org>
Message-ID: <Xns91FC60D052D28gmcmhypernetcom@199.171.54.213>

Alexander Skwar wrote:

>?Alexander Skwar? sagte am 2002-04-26 um 01:17:32 +0200 : ^
>> SyntaxError: invalid syntax
>> 
>> Well, why is that?
> 
> Hm, seems like the packages aren't cross platform.  When I use the zip
> package, I get the error message.  When I use the .tar.gz file, I don't
> get any errors on my Linux machine.

The tar.gz is built on Linux; the .zip on Windows. So line endings are
different. eval wants the "right" line endings.
 
> Can I use the Installer to create a cross-platform binary on Linux which
> will run on Windows for a wxPython application?  Which requirements do I
> have to meet to be able to do so?

No. Cross-compiling is sort-of conceivable, but even if I went to all
that work, you'd still have to have a "workable" Windows Python 
installation on your Linux box, with no way of verifying that it
is "workable".

-- Gordon
http://www.mcmillan-inc.com/


From tdelaney at avaya.com  Sun Apr 21 19:23:20 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 22 Apr 2002 09:23:20 +1000
Subject: nested exceptions?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A27B@natasha.auslabs.avaya.com>

> From: Roy Smith [mailto:roy at panix.com]
> 
> What I've come up with uses recursion:
> 
>     def getBaseObject (self, objects, oid):
>         try:
>            return objects[oid]
>        except KeyError:
>            return self.getBaseObject (objects, 
> self.getParentOid (oid)) 
> 
>     def getParentOid (self, oid):
>         parts = oid.split ('.')
>         parentParts = parts[:-1]
>         return '.'.join (parentParts)

Try the following (self omitted because it doesn't need to be part of a
class):

def getBaseObject (objects, oid):
    """"""

    path = string.split(oid, '.')

    while path:
        try:
            return objects[string.join(path, '.')]
        except KeyError:
            del path[-1]

You can of course change string.join to a local name as

    join = string.join

but this is more straighforward. If no part of the path is in objects, None
will be returned (instead of the infinite recursion you get with your
implementation). You may wish to raise an exception instead.

Tim Delaney




From perry at subdimension.com  Fri Apr 26 05:09:15 2002
From: perry at subdimension.com (Perry)
Date: 26 Apr 2002 02:09:15 -0700
Subject: ActiveX events in wxPython not working
References: <6ddcff55.0204250625.1992ce1b@posting.google.com> <mailman.1019763678.16478.python-list@python.org>
Message-ID: <6ddcff55.0204260109.194be8d9@posting.google.com>

Cliff Wells <logiplexsoftware at earthlink.net> wrote in message news:<mailman.1019763678.16478.python-list at python.org>...
> On 25 Apr 2002 07:25:36 -0700
> Perry wrote:
> 
> > Hi,
> > 
> > I'm using wxPython as an ActiveX (GUI) container. The OCX draws a
> > gantt
> > chart from supplied data. Everything works fine, appart events.
> > 
> > Here is the code:
> 
> [snip]
> 
> > class MyGanttPanel(wxPanel):
> >   def __init__(self, parent):
> >      wxPanel.__init__(self, parent, -1)
> > 
> >      sizer = wxBoxSizer(wxVERTICAL)
> >         
> >      GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
> >      gantt.VcGantt = GanttClass(self, -1, wxPoint(0,0), style=wxMAXIMIZE)
>        ^^^^^ Where is this instantiated?
> 
> [snip]
> 
> I don't see where the object "gantt" is being created.  You can't add an
> attribute to a non-existent object:
> 
> Python 2.2 (#1, Mar 15 2002, 11:52:42) 
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-81)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> f.a = 1 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> NameError: name 'f' is not defined
> >>> class foo: pass
>  ... 
> >>> f = foo()
> >>> f.a = 1
> >>> 
> 
> Are you running the program from a shell or by double-clicking its icon?  It
> may be that there are errors you aren't seeing if you aren't using a shell. 
> Other than that, I'm afraid my ActiveX knowledge is (happily) non-existent, so
> if this isn't your problem then I'm afraid I can't help much.


gantt is a global class created like this:

class gantt:
  VcGantt = None
  status = "somestring"
  start = startDate
  end = endDate
  ...

This class holds all the data needed by the program.
So the class gantt exists outside the class MyGanttPanel, I'm using it
in other classes to feed the data to the Gantt
(gantt.VcGantt.InsertNode).
So the object is not destroyed.
I have searched this news group for similar problems and found a
similar
post, the code was almost the same. Somebody suggested to downgrade
win32com
to the prior version and it seemed it worked, but this was at win32com
version around 120.

Thanks for all the replies.


From jeff at ccvcorp.com  Tue Apr 30 20:20:46 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 30 Apr 2002 17:20:46 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <uctooia295ne2f@news.supernews.com> <bHDz8.34855$n4.7400665@newsc.telia.net>
Message-ID: <MPG.1738d4a4f15042339896cd@news.nwlink.com>

In article <bHDz8.34855$n4.7400665 at newsc.telia.net>, 
fredrik at pythonware.com says...
> John Roth wrote:
> > To me, .join() makes much more sense as a list or tuple method than
> > either. Semantically, .join() takes a sequence and transforms it into
> > a string, therefore it should be a method on the input type (the
> > sequence)
> 
> there is no sequence type in Python.  try again.
> 
> </F>

Not only that, but Python encourages developers to create objects 
that fulfill the sequence protocol.  Your proposal would force 
__join__() to be part of that protocol, along with __getitem__() 
et. al., thus making custom sequences significantly more awkward 
to write.  (Especially considering that the most sensible 
implementation of this join method would not vary significantly, 
thus creating heaps of boilerplate.)  Oh, and let's not forget 
that these __join__() methods must account for plain-string-
vs.-unicode issues, too.

Conversely, by declaring join() to be a method on strings, it 
becomes quite simple (indeed, free) for *any* sequence to be 
joined, regardless of whether that was expected by the original 
author.  And being sure to get the right plain-vs.-unicode option 
is trivial, too -- that's handled by Python's normal 
polymorphism, instead of requiring typechecking in join().  And 
if, god forbid, a third type of string-like object were 
introduced, then (providing that it implemented join() 
reasonably) all existing sequences would *already* be able to be 
joined into that third string-like type, too -- again, for free, 
instead of requiring modification (and *more* typechecking!) in 
*each* sequence-like class.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From cliechti at gmx.net  Tue Apr 16 17:16:36 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 16 Apr 2002 23:16:36 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <Xns91F2ED00B6997cliechtigmxnet@62.2.16.82>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in
news:a9i327$cvi$1 at nets3.rz.RWTH-Aachen.DE: 
> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
> news:a9i2kg$cg8$1 at nets3.rz.RWTH-Aachen.DE...
>> Is it possible to have a safe eval() on the server-side even when the
> string
>> consists of submitted user data?
>>..

have a look at the rexec module

> I forgot to mention: all that should be allowed is arithmetics
> (boolean as well) plus a limited set of functions that I write &
> define. Say the following is OK (I already replaced internal variables
> here): 1 == 2 and false or (10 + 3 * 2) == 4 or
> internal_harmlessFunction() == true 

you can disable the builtin functions you want and populate the namespace 
with save functions and classes (see Bastion for that)

chris

-- 
Chris <cliechti at gmx.net>



From bokr at oz.net  Sun Apr  7 05:57:36 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Apr 2002 09:57:36 GMT
Subject: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> <mailman.1018118665.30693.python-list@python.org>
Message-ID: <a8p56g$r42$0@216.39.172.122>

On Sat, 06 Apr 2002 20:39:41 +0200, Laura Creighton <lac at strakt.com> wrote:

>Mark McEahern writes:
>> [Laura Creighton]
>> > [...]  I am rejecting the existence of booleans in the language
>> > because it is too hard to teach how to use them properly.
>> 
>> How do you use them properly?
>> 
>> Suppose I want to validate users.  I might write something like this:
>> 
>> 	def isValid(user, password):
>> 		if ...:
>> 			return True
>> 		else:
>> 			return False
>
>My pleasure, especially since this is such a cool example.  You have
>just shot yourself in the foot, and do not know it.  When you have
>shipped your isValid function, across the world to your 10,000 clients,
>you will discover this.
>
>Your boss comes in and says, oh, Mark, we need a change to that
>isValid function.  We need to return 3 states, Valid, Invalid, and
>Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
>
>What do you do now?  
>
>> Using isValid, that returns one of two possible answers, is as clear and
>> bright as day to me.  If later on I want a function that does something
>> different, it would most definitely be named something different.  Why on
>> earth would I want a function named isFoo to return one of more than two
>> possible values?  (Your answer is probably that you never want a function
>> named isFoo.  Hmm, I still don't see why.)
>
>Do you now? Use Booleans when you want to return ON and OFF -- the
>state of that bit in memory, of something that is constrained by
>physical reality, or mathematical reality to only have 2 values, ever,
>cross my heart and the world will end if this ever ceases to be the
>case.  Don't artificially limit yourself to 2 values because your
>vision at the time only extended to 2 values.  The one thing constant
>in this world is change.   If you don't need a third value, somebody 
>else will, later.  Integers are your friends.
>
Classes and objects are even better friends. How do you think we got from DOS errorlevels
to Python exceptions? Magic numbers are out of style ;-)

IANAT, but IMHO, the teaching problems you allude to are not about types themselves,
but about the appropriate use of types to model the problem domain. Your complaint
about isLeap IMHO should be redirected as Kent Seehof suggests. If the Swedish 1712
leap year had two extra days, you can say isLeap is still a valid concept, and can only
be true or false, however that it is an incomplete model of reality for some purposes.
You could say def isLeap(y): return numberOfExtraDays(y)>0 to make the distinctions
clear in one line. From there it's easy to point to the fallacy of 365+isLeap(y)
(and the documentation problem with making any is-someproposition(x) function return
anything but true vs false, however represented).

Frankly, I'm horrified when a teacher prohibits (other than disruptive selfishness).
I'd much rather have the teacher say, solve the problem any way you like, but I have
taught you that some ways often lead to difficulties, so I will demand (a) that you
show you know you are breaking one of my rules of thumb, and (b) that your provide
a rationale for your decision to do so. Just (a) goes a long way towards QA-ing that
they have the idea, and they get to play if they are bright and motivated. An annotated
most-horrible-example-solution may teach more than a spiffy cliche.

If you only have a few acceptable answers, why not just give 'em a multiple choice
test and be done with it? I'd guess it'd be a lot less work than real teaching.

I guarantee you will eventually have a student that is smarter than you are, and who
will surprise (and delight, unless you are ego-deficient) you. Prohibition is a symptom of
management style in the sheep-dip philosopy of teaching, IMHO. Rote patterns are ok
up to a point, but they're plastic flowers vs seeds and bulbs depending on how done.

Regards,
Bengt Richter


From taradino at softhome.net  Sat Apr  6 15:55:31 2002
From: taradino at softhome.net (Johannes Gijsbers)
Date: Sat, 6 Apr 2002 22:55:31 +0200
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net>
Message-ID: <MPG.17197f04a3f463e8989680@news.planet.nl>

logistix wrote:
> Use python's type system:
> 
> >>> import types
> >>> a, b = 1, 1.0
> >>> if type(a) == types.IntType: print "int"
> ...
> int
> >>> if type(b) == types.IntType: print "int"
> ...
> >>>

That's not really a good idea, as math.sqrt() returns a float. I'd say:

>>> import math
>>> def isInt(number):
...     return int(number) == number
...
>>> isInt(math.sqrt(289))
1
>>> isInt(math.sqrt(315))
0

Johannes Gijsbers
--
-- 
will you be my big plaything
my ninja power
my number cruncher


From imbosol at vt.edu  Sun Apr 21 13:50:44 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 21 Apr 2002 13:50:44 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com>
Message-ID: <k5uu9a.621.ln@127.0.0.1>

John Roth wrote:
> 
> "Carl Banks" <imbosol at vt.edu> wrote in message
> news:tnos9a.de.ln at 127.0.0.1...
>>
>> Computer programming is not science.
>>
>> The study of algorithms, their properties, etc., is science.
>> Implementing a quicksort algorithm is not science, but studying its
>> properties is.
> 
> I would dispute that. The study of algorithms may very well
> be a branch of mathematics, but mathematics is not science.

Algorithms is unquestionably a branch of mathematics.  It has the same
formalism, same proofs and theorems, etc.  I imagine it's even
possible to define things like topologies of algorithms.


> Science is, among other parts, the study of the natural universe
> and things in it. The study of algorithms is a study of things
> that have been created by the human mind.

Sorry, but I don't agree that algorithms, or any mathematical concept,
is an invention of the human mind.  The algorithms have existed since
the beginning of time.  The quicksort had the property of requiring ~
n^2 comparisons in the worst case before it had ever been used or
applied.  We merely discovered this algorithm; we didn't create it.

So I say mathematics, including the study of algorithms, is a natural
science.  And, for now, that's all I have to say about it.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From jwbaxter at spamcop.net  Wed Apr 24 22:24:46 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Wed, 24 Apr 2002 19:24:46 -0700
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com> <aa5f7l$uck$1@news.nextra.cz>
Message-ID: <jwbaxter-B5D4A7.19244624042002@corp.supernews.com>

In article <aa5f7l$uck$1 at news.nextra.cz>,
 "Petr Prikryl" <Answer.via.news.please.prikrylp at nospam_skil.cz> wrote:

> Well, I have meant what quotes should I use in cases
> when only single or only double are to be used (i.e.
> when I do not want to include the opposite inside
> the string.  Read it what *style* is prefered.

I use single quote (that is, apostrophe) to avoid a shift key 
manipulation.  Unless I forget.  And of course if the string begins with 
a shifted character, the manipulation is deferred, not avoided.

If for some reason I want a string to be
    Eggs!
I'll use double quote.  Unless I forget.

In other words, you probably shouldn't follow my example.  And you may 
possibly detect that I don't find it a major issue.

  --John (my opinions seem to be my own...no one has claimed them yet in 
60 years)


From mal at lemburg.com  Wed Apr  3 08:05:37 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 03 Apr 2002 15:05:37 +0200
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <3CA9EF24.2B9A0C27@lemburg.com> <20020403012223.GB562@lilith.my-fqdn.de>
Message-ID: <3CAAFE21.B8C7E84F@lemburg.com>

Gerhard H?ring wrote:
> 
> * M.-A. Lemburg <mal at lemburg.com> [2002-04-02 19:49 +0200]:
> > SAP DB is fast, full-featured, open-source and very stable.
> 
> Open-source it sure is, but that doesn't mean that mere mortals can
> compile it or even port it to not currently supported platforms like
> FreeBSD.

That's true. Their compile system is rather complicated, due to many
files being generated on-the-fly and various third-party tools
which massage the code into compilable bits -- Python is among
them, so at least this part should be comprehensible ;-)
 
> > Note that Unicode support in mxODBC is available both as emulation
> > (mxODBC does the conversion to e.g.  UTF-8) and using native support
> > (the ODBC driver has to support Unicode).
> 
> Very interesting. This Unicode stuff isn't specified in a new draft of
> the DB-API, is it?

No.
 
> I'm planning to add Unicode (and perhaps non-default 8-bit charset)
> support to pyPgSQL. In my first tries, I found that there are really two
> different problems:
> 
> - getting Unicode strings as parameters of execute into the database,
>   like:
> 
>   cursor.execute("insert into test(col) values('%s'" % \
>         u"someUnicodeString")
> 
>   This doesn't look difficult.

I think that this part is really user-level programming. What's
more interesting is:

cursor.execute('insert into test (col) values (?)', (u"JapaneseText",))

There are three possibilities to choose from:
1. the DB can handle Unicode, then pass through the data as-is
2. the DB doesn't handle Unicode, but does use a predefined code-page,
   then convert the Unicode data to that code-page
3. the DB doesn't handle Unicode, and doesn't have a predefined code-page,
   then convert the Unicode data to a connection specific encoding

1. and 3. are implemented in mxODBC 2.0. 2. is currently not 
implemented since it requires fetching the code page data from the
database which is not always possible.
 
> - getting Unicode out of the database: My first idea was to add a flag
>   to the connection constructor that indicates wether to returns strings
>   as Python strings or Unicode strings. Any better ideas?

Nope. That's how mxODBC does it too. Connections and cursors
have a .stringformat attribute which can be set to haev mxODBC
return 8-bit strings, mixed 8-bit/Unicode (depending on the
information returned by the DB) or all Unicode.
 
-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/



From max at alcyone.com  Mon Apr 29 17:40:54 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 29 Apr 2002 14:40:54 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net>
Message-ID: <3CCDBDE6.93F17F@alcyone.com>

Raymond Hettinger wrote:

> From the comments, I've learned:
> 
> 1. map, filter, divmod, and pow may be immovable and should
>     be left alone.
> 
> 2. reduce has fewer friends, but should probably also be
>     left alone
> 
> 3. input() has almost no supporters and should probably be
>     deprecated.  it is security risk, a trap for new users, easily
>     replaced with eval(raw_input), and is named in a way that
>     doesn't suggest that eval is occurring.
> 
> 4. apply() was repeatedly suggested as another candidate for
>     deprecation since the * and ** forms make it obsolete
> 
> 5.  oct() is another candidate based on the rarity of usage
> 
> 6. command line options and environment variables are
>     universally hated, so this idea is dropped
> 
> 7. the language needs some means of deprecating so that clutter
>     does accumulate.  that means needs to be well documented,
>     have a long, slow phase-in, and have a mechasism for
>     restoring old behavior if it is ever needed.

So far I haven't seen any indication that what you're talking about
moving qualifies as "clutter."  If your best attempt at cleaning up
clutter in a language is moving a _handful_ of builtlins to a submodule,
I'd say that the language doesn't have much clutter at all.

What is the supposed advantage to moving, say, map to a module named
functional?

Deprecating oct doesn't make any sense to me at all.  You already have
hex; having oct simply makes good sense and improves symmetry.  Getting
rid of (or simply moving to a module) oct but not hex makes no sense to
me whatsoever.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From gerson.kurz at t-online.de  Tue Apr  9 14:01:02 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Tue, 09 Apr 2002 18:01:02 GMT
Subject: Thread/Process-safety of functions in the Standard Library, eg the dbm classes
Message-ID: <3cb32a63.17540718@news.t-online.de>

I wonder if it is documented which functions of the Python Standard
Library are threadsafe, and which not. More explicitly, I was having a
look at the anydbm interface and am pretty sure that it is not safe to
access from multiple-processes at the same time, and I don't even know
whether or not it is safe for multiple threads.

So, is there such a documentation? 

Furthermore, do you have any suggestions for these requirements: I
want to be able to store some data (<100K), and access it from
multiple processes in a safe way. Currently I'm using mySQL with
mysqldb, which seems like heavy overkill for that amount of data.
Plus, I want users to use my program without having to first install
MySQL (or any other dbm for that matter).

[A hardcore idea would be to make a "anydbm-server" that you can
communicate with using sockets ;]


From kseehof at neuralintegrator.com  Fri Apr 12 07:59:22 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Fri, 12 Apr 2002 04:59:22 -0700
Subject: Looking for doh!.. simple sequence type
In-Reply-To: <EFEFIHAINIINNELLHFAIOEOCCAAA.kseehof@neuralintegrator.com>
Message-ID: <EFEFIHAINIINNELLHFAIMEODCAAA.kseehof@neuralintegrator.com>

Ken wrote:
> 
> Max wrote:
> > 
> > Hmm ... I am trying to create a simple sequence type that I 
> cant iterate 
> > over like (Python 2.1)::
> > 
> > lst = Lst()
> > for i in lst:
> >      print i
> > 
> > And I thought I could do it like:
> > 
> > class Lst:
> > 
> >      def __len__(self):
> >          return 4
> > 
> >      def __getitem__(self, idx):
> >          return 42
> > 
> > But I have to write it like below to avoid an endless loop..
> > 
> > lst = Lst()
> > for i in range(len(lst)):
> >      print i,
> > 
> >  >>>42 42 42 42
> > 
> > I am shure it's so simple tha I will go Doh! .. But perhaps I haven't 
> > had enough cofee yet. 'Cause I cannot seem to figure it out.
> > 
> > regards Max M
> > 
> The problem is that __getitem__ doesn't check length.  Instead, it
> terminates on an IndexError exception.
> 
> class Lst:
>     def __getitem__(self, idx):
>         if idx>=4:
>             raise IndexError
>         else:
>             return 42
>     
> lst = Lst()
> for i in lst:
>     print i
>     
> 42
> 42
> 42
> 42

What I meant to say is that the default __iter__ doesn't check length,
but does a __getitem__ until it terminates on an IndexError exception.

Best regards,
- Ken





From marklists at mceahern.com  Fri Apr 26 17:05:52 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 26 Apr 2002 16:05:52 -0500
Subject: configuring mxCGIPython
In-Reply-To: <3CC996B1.978C346A@lemburg.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOENIKMAA.marklists@mceahern.com>

[M.-A. Lemburg]
> You have to call cgipython as:
>
> 	cgipython script.py [options]
>
> You probably have passed in something else as first parameter
> which is why you get the error in cgipython.py:

This is the shell script wrapper I'm using to test it:

  #!/bin/bash
  # test2.cgi

  echo content-type: text/plain
  echo

  export LD_LIBRARY_PATH="/home/chain/cgi-bin"
  export PYTHONVERBOSE="1"
  export PYTHONPATH="/home/chain/cgi-bin"

  if [ -f cgipython ]
  then
    echo cgipython exists
    # Trying to send stderr to stdout.
    ./cgipython test.py 2>&1
  else
    echo cgipython does not exist
  fi

test.py is in the same folder as cgipython.  Are the stderr/stdout
redirectors confusing it?  If I run that from my local Linux shell (i.e.,
NOT on the ISP), I get:

	$ cgipython test.py 2>&1
	2.2.1 (#1, Apr 26 2002, 09:41:38)
	[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)]

> # run the script file passed in sys.argv[1]
> import sys
> try:
>         script = sys.argv[1]
> except IndexError:
>         sys.stderr.write('Usage: cgipython <pyfile> [parameters]\n')
> else:
>         del sys.argv[0]
>         execfile(script) # <-- this is line 27

Yes, thank you.  After I built my own version of mxCGIPython, I was able to
find cgipython.py in the temp folder and find the line.  By the way, I can
send you the cgipython I generated.  It's not working for me on my ISP--I'm
getting this error, but that may have more to do with the kernel/libraries
my ISP has than anything else:

  cgipython exists
  ./cgipython: /lib/libpthread.so.0: version `GLIBC_2.2' not found (required
by ./cgipython)
  ./cgipython: /lib/libc.so.6: version `GLIBC_2.2' not found (required by
./cgipython)

> I should probably add another try:except: here which generates
> a better error messages.

Sounds good.

Thanks,

// mark





From phr-n2002a at nightsong.com  Tue Apr 23 06:49:17 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 23 Apr 2002 03:49:17 -0700
Subject: Encryption
References: <70h8cugniao40148h76fv4o7bgci2j3ddd@4ax.com> <7x662j76iu.fsf@ruckus.brouhaha.com> <359acu8n2qode4gnnmh38rkekp204dss3n@4ax.com>
Message-ID: <7xn0vu4rma.fsf@ruckus.brouhaha.com>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> writes:
> >Why do you especially want to use mcrypt, in that case?
> 
> I don't. The site supplying the data want to use mcrypt. We just want
> to accomodate. 

If the data volume isn't too large and the security requirements
aren't too intense, you can code RC4 in Python and get the other end
to configure Arcfour encryption into Mcrypt, and do all the decryption
in pure Python.  That's probably the easiest approach.  Otherwise you
have to either get python-mcrypt working under Windows, or find some
other library that works under Windows.

I coded RC4 in Python a few months ago and it ran at about 100
kbytes/sec on my P3-750, if that's of any help.  I don't know if
that's fast enough for your application.  A C implementation would run
at tens of megabytes/sec.


From kwokng at earthlink.net  Tue Apr 30 21:53:09 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Wed, 01 May 2002 01:53:09 GMT
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com> <f0fd5987.0204300105.51529004@posting.google.com> <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk> <g7Bz8.35106$8D3.1017931@news1.tin.it>
Message-ID: <9UHz8.23$s55.5@newsread2.prod.itd.earthlink.net>

I am also new to python.  I think python will be my replacement of perl.
However, if I need to do very heavy regular expression, I still prefer perl.
I just don't feel comfortable with the python's re syntax.

Billy Ng

"Max" <prova at prova.it> wrote in message
news:g7Bz8.35106$8D3.1017931 at news1.tin.it...
> ----- Original Message -----
> From: "JohnO" <jo at johnochiltree.uklinux.net>
> Newsgroups: comp.lang.python
> Sent: Tuesday, April 30, 2002 11:32 AM
> Subject: Re: Python vs. Perl, which is better to learn?
>
>
> > Garry Taylor wrote:
>
> [...]
>
> > I don't know if anyone else has noticed this but I've found that
operators
> > and sys admins use perl and that developers tend towards python. Also
sys
> > admins rarely utilise indirection but developers do.
> >
> > John
>
> Interesting... i had the same feeling!
> I'm pretty new to Python, but it's very close to the way i think... i can
> concentrate on the problem and not on how to implement my solution... very
> very powerful for an analyst/programmer....
>
>
>




From sholden at holdenweb.com  Sun Apr 28 14:31:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 28 Apr 2002 14:31:00 -0400
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no>
Message-ID: <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnacm8a7.mru.mlh at vier.idi.ntnu.no...
> In article <mailman.1019792457.31426.python-list at python.org>, James J.
> Besemer wrote:
> [snip]
> >4. Actually, this all makes me wonder if your web source is accurate in
the first
> >place?  I've come to place a lot of trust in my well worn OAD.  Shit.
Now I'll be
> >pouring over different dictionaries all evening.
>
> While you're at it you might want to look the difference between
> "pore" and "pour", unless you want to ruin your dictionaries <wink>.
>

I only recently realised that the OAD makes an amazing omission, completely
failing to provide a definition for the word "gullible".

no-don't-bother-to-go-and-look-it-up-ly y'rs  - steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From mgilfix at eecs.tufts.edu  Wed Apr 17 13:03:19 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 13:03:19 -0400
Subject: Pure python crypto
In-Reply-To: <02041718383201.00171@toad>; from martinb@otenet.gr on Wed, Apr 17, 2002 at 06:38:32PM +0300
References: <20020417021703.H16954@eecs.tufts.edu> <02041718383201.00171@toad>
Message-ID: <20020417130319.A25898@eecs.tufts.edu>

  Ah. I'll make it available on my home page then and post you a URL
when its ready. Wasn't sure if there was any interest.

              -- Mike

On Wed, Apr 17 @ 18:38, Martin Barnard wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> On Wednesday 17 April 2002 09:17, you wrote:
> >   Would anyone have any use for a pure python implementation of
> > Blowfish? Or any other crypto algorithms for that matter.  As a matter
> > of interest, I found myself implementing the blowfish module recently
> > but was unsure if any audience would be interested. No C compiler
> > is required, which is nice. Could use some testing though.
> >
> >                -- Mike
> I'd be interested in it.
> Where can I download it from?
> 
> 
> Martin
> - -- 
> You look like a million dollars.  All green and wrinkled.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.0.6 (GNU/Linux)
> Comment: For info see http://www.gnupg.org
> 
> iEYEARECAAYFAjy9lvgACgkQpgoq4z9IM6hwEgCgiOGAbHrPrc4VO1R+2m5n/kHy
> tu4AnAmeYMVHw9XKWNQaUJ9E27sHKmvr
> =cF7Q
> -----END PGP SIGNATURE-----
`-> (martinb)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From eric at enthought.com  Wed Apr  3 17:11:04 2002
From: eric at enthought.com (eric)
Date: Wed, 3 Apr 2002 17:11:04 -0500
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at> <lbh8z84nake.fsf@xihu.inria.fr> <15531.7864.698585.485436@12-248-41-177.client.attbi.com> <lbh7knolreu.fsf@xihu.inria.fr> <15531.9709.924953.953487@12-248-41-177.client.attbi.com> <mailman.1017850275.2134.python-list@python.org>
Message-ID: <0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop>

Hey Jan,
>
> for i in xrange(1,a.shape[0]):
>         if a[i,i]>30: a[i,i]+=a[i,i+1]
>         else:          a[i,0]*=2.0
>
> Not a particularly useful example, but one which could be much
> accelerated by translating it (automatically) into C.

For this particular example, weave.inline is a perfect solution.  The C++ code
is pretty readable:

from scipy import *

def foo(a):
    code = """
           for (int i = 0; i < _Na(0);i++) {
               if (a(i,i) > 30) a(i,i) += a(i,i+1);
               else             a(i,0) *= 2.0; }
           """
    weave.inline(code,'a',compiler='gcc',
                 type_converters = weave.converters.blitz)

a = rand(3,3)*60
print 'before:', a
foo(a)
print 'after:', a

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

C:\temp>python ex.py
before: [[ 52.34835505  45.81040621  34.39133048]
 [ 59.23025608  20.20882666  42.10668683]
 [ 41.57790542  43.00394297  26.22276664]]
after: [[  98.15876126   45.81040621   34.39133048]
 [ 118.46051216   20.20882666   42.10668683]
 [  83.15581083   43.00394297   26.22276664]]


The place where weave breaks down is when you have a lot of calls within the
loop to Python functions.  This is a common problem, and the only way around it
is to have something that compiles the Python to C or assembly or something.
That is what Psycho does.  weave may also do something similar in the forseeable
future.  Pat Miller is integrating some work similar to Psycho into weave that
converts Python to C on the fly using type inferencing and then uses weave to
build the extension module.  Several of the aims are different than psycho, but
the concepts are similar.

I'm all for a faster Python, but a robust compiler for a dynamic language is
difficult.  I'm hoping Psycho pulls it off.  weave is less ambitious and can
fill at least some of the gaps until it does.

regards,

eric


>
> J.
>
> --
> -------------------------------------------------------------------------
> Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
>        or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
>                     http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/
>
>






From mvcc_cr at yahoo.com  Sun Apr 14 22:40:24 2002
From: mvcc_cr at yahoo.com (Marco V. =?ISO-8859-1?Q?Caama=F1o?= Castro)
Date: 14 Apr 2002 20:40:24 -0600
Subject: Question about httplib and HTTP response 100 Continue
Message-ID: <1018837561.612.2.camel@prowler>

I've been trying to figure out why this code[1] does not work[2] on
connecting with a IIS server, but it works with Apache on Linux. I found
that using curl to post the request works. Digging into the response I
got using curl I found that the server is first sending a HTTP 100
Continue response and then the result of the post:

-----------------------
HTTP/1.1 100 Continue
Server: Microsoft-IIS/4.0
Date: Mon, 15 Apr 2002 00:49:27 GMT

HTTP/1.1 200 OK
Server: Microsoft-IIS/4.0
Date: Mon, 15 Apr 2002 00:49:27 GMT
Connection: close
Content-Type: x-data/xact-error
-----------------------

but apache just outputs the results right away:

-----------------------
HTTP/1.1 200 OK
Date: Mon, 15 Apr 2002 00:38:17 GMT
Server: Apache/1.3.22 (Unix) Debian/GNU mod_ssl/2.8.5 OpenSSL/0.9.6a PHP/4.1.0RC2
X-Powered-By: PHP/4.1.0RC2
Transfer-Encoding: chunked
Content-Type: text/html; charset=iso-8859-1
-----------------------

So, does the httplib work when it encounters a 100-Continue response?

I would really apreciate a little help here as I am stuck, I can't
figure out why using the httplib library doesn't work with this server.


Marco C


[1]:
def testurl():
	print '\n\nUsing HTTPS request'
	body=fmsg('\x82test')
	host=("ssl.pgs.wcom.net:443")
	hs=httplib.HTTPS()
	hs.set_debuglevel(4)
	hs.connect(host)
	hs.putrequest('POST','/scripts/gateway.dll?Transact')
	hs.putheader('Content-Type','x-Visa-II/x-auth')
	hs.putheader('Content-Length', str(len(body)))
	hs.endheaders()
	hs.send(body)
	status, reason, header= hs.getreply()
	data=hs.getfile().read()
	print data
	print status
	return

[2]:
Using HTTPS request
send: 'POST /scripts/gateway.dll?Transact HTTP/1.0\r\n'
send: 'Content-Type: x-Visa-II/x-auth\r\n'
send: 'Content-Length: 5\r\n'
send: '\r\n'
send: '\x82test'
Traceback (most recent call last):
  File "test.py", line 4, in ?
    misc_func.testurl2()
  File "misc_func.py", line 180, in testurl2
    status, reason, header= hs.getreply()
  File "/usr/lib/python2.2/httplib.py", line 728, in getreply
    response = self._conn.getresponse()
  File "/usr/lib/python2.2/httplib.py", line 570, in getresponse
    response = self.response_class(self.sock, self.debuglevel)
  File "/usr/lib/python2.2/httplib.py", line 98, in __init__
    self.fp = sock.makefile('rb', 0)
  File "/usr/lib/python2.2/httplib.py", line 607, in makefile
    buf = self.__ssl.read()
socket.sslerror: (5, 'EOF occurred in violation of protocol')


-- 

Marco V. Caama?o Castro


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From siegfried.gonzi at kfunigraz.ac.at  Sat Apr 20 13:26:19 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 20 Apr 2002 18:26:19 +0100
Subject: Question about IDLE
References: <0p5w8.17527$CH.4180@rwcrnsc52.ops.asp.att.net>
Message-ID: <a9s4rr$v70$1@newsreader1.netway.at>

"Matt" <insta at no.spam> schrieb im Newsbeitrag
news:0p5w8.17527$CH.4180 at rwcrnsc52.ops.asp.att.net...
> I'm trying to learn Python, and I was wondering if the IDLE program was
> sufficient enough for basic programming, or should I find some sort of
text
> editor?
> I'm currently using a win32 box.
>
> Also, any links to tutorials or general information about Python that is
> not listed on the Python website would be much appriciated.

I find it quite useable (at least on my windows box). You should also try to
become aquainted with the command line. You can there develop your code
incrementally.

Otherwise you can run your script directly from within your program code
(okay that saving before running is a little bit awkward).

Thanks to God that the developers of Python did not assume that everybody is
an Emacs or vi freak...

S. Gonzi




From erno-news at erno.iki.fi  Tue Apr  2 21:26:56 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 03 Apr 2002 05:26:56 +0300
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
Message-ID: <kupu1hple7.fsf@lasipalatsi.fi>

In article <3CAA5777.66E7C2D5 at python.org>, Guido van Rossum
<guido at python.org> writes:

| I think it's time I learn to say "yes" even if the community doesn't
| see the light yet.

maybe some of the community have mistaken the dearth of features in
the language for simplicity by design, whereas in reality you just
hadn't gotten around to adding all the little features yet.

since 1.5, i find the following changes to the language come
immediately mind as worthy:
- nested_scopes
- string methods
- unicode (maybe, haven't really used or needed it yet)
- new-style classes (ditto)

the following language changes come immediately
to mind as feeling superfluous:
- list comprehensions (map/filter)
- f(*args, **kw) (apply)
- staticmethod/classmethod

i haven't decided if this is a good ratio or not, yet.

  -- erno


From sholden at holdenweb.com  Mon Apr  8 08:22:58 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 08:22:58 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1018207935.1323.python-list@python.org>
Message-ID: <h_fs8.26783$Sw5.5213@atlpnn01.usenetserver.com>

"Arthur Siegel" <ajs at ix.netcom.com> wrote in message
news:mailman.1018207935.1323.python-list at python.org...
> Aahz -
>
> > people like Arthur and Paul Rubin...
>
> Irish?
>
> Paul is one of the few people on the list (other than
> some of the Powers that Be) with whom I have had a
> personal exchange of some nastiness.
>
Hmm, I don't think I needed to know that.

> He had, I believe, in the course of these discussions
> volunteered to do some hands on development work
> on some point where he and Guido seemed to agree some
> attention was necessary.
>
> How this poor guy got himself lumped in with the
> likes of *me*, is hard to understand.
>
Perhaps you both belong to the set of people who don't understand why
they're lumped in together?

regards
 Steve






From hfoffani at yahoo.com  Wed Apr  3 07:06:35 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Wed, 3 Apr 2002 14:06:35 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>  <mailman.1017799668.13959.python-list@python.org> <slrnaalr5a.fa8.mlh@vier.idi.ntnu.no>
Message-ID: <20020403070637.620$SQ@news.newsreader.com>

> >or is that, in the words of the late Milton Berle:
> >
> >     You can fool some of the people some of the time, and all of the
> >people some of the time, and that's usually enough.
>
> Isn't that "all of the time" in the first sentence? Otherwise it
> doesn't make much sense (in that the first sentence would follow from
> the second).

Berle's missing "all of the time" is on pourpose. The original quote is:
      You can fool some of the people some of the time and all of
      the people some of the time, but you can't fool all of the people
      all of the time. --Abraham Lincon

Apparently that saying inspired others like M.Berle's one and:
      You can fool all the people all the time if the advertising is
      right and the budget is big enough.--Joseph E. Levine

But may be I was fooled and got it all wrong ;-)

Regards,
-Hernan





From aleax at aleax.it  Fri Apr 26 17:09:55 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 21:09:55 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com> <mailman.1019212237.13240.python-list@python.org> <NETv8.45693$eS3.10263@atlpnn01.usenetserver.com>
Message-ID: <Dmjy8.66344$vF6.2029394@news2.tin.it>

Steve Holden wrote:
        ...
> I'm sure Alex's book is going to be an excellent contribution to the
> language and its growth. While I was disappointed that it had to line up
> behind "The Python Cookbook", it will probably be better for the wait. (I

We'll see -- I _am_ ensuring it covers 2.2, of course.

> can imagine Alex gnashing his teeth whenever new features requiring
> changes are introduced, though. When you are writing, sometimes you just
> want to shout "Stop!" at the developers ;-)

Actually, as a writer, it's to my net potential financial advantage if the
technology I'm covering changes (as long as the rate of change does not
reduce the acceptance of the technology in the market) -- this gives
me the "upgrade" after-market.  There are technologies which I
needed to track for which I've bought two, three, up to four such
"upgrade" books in a few years.  For the TCP/IP fundamentals, I've
made do with one book throughout my career -- one day or another,
I'll have to buy another for IPv6, but it won't be an "upgrade" of that
dog-eared book which served me so well.

My problems with the prospective of continuing, frequent changes to
the language come from other issues -- basically that innocuous
looking little parenthesis in the previous paragraph.  A rate of change
perceived as "too fast" may well inhibit the language's uptake.  Quite
apart from book authors' financial prospects, I feel it would be a pity
for the world at large if the benefits that Python can bring were to be
less widespread because of such effects.


Alex



From emile at fenx.com  Fri Apr 19 14:56:00 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 19 Apr 2002 18:56:00 GMT
Subject: Class method ptrs and dictionaries
References: <3CC05EF6.2020101@nyu.edu>
Message-ID: <4LZv8.40399$HH5.3423@rwcrnsc51.ops.asp.att.net>

Shane Hoversten
[snip]
>  >>> a = foo.foo()
>  >>> a.bar()
> printOne: Hello, what!
> printTwo: Hello, what! I say!
> I'm in nothing.
>
> My question: how does python "know" that the functions I'm putting in
the
> dictionary are class methods that need to be handed a self pointer in
> the first two cases, and not handed one in the third case?  Are these
> issues described somewhere that somebody could point me to so I could
> read about how python handles this behind the scenes?  I did a couple
> quick searches but didn't turn anything up...
>

If you look at the def's involved, you'll see that they're recognized
differently:

>>> a = foo()
>>> a.bar
<bound method foo.bar of <__main__.foo instance at 0x0076BCF8>>
>>> nothing
<function nothing at 0x007A8B20>
>>> foo.bar
<unbound method foo.bar>
>>> foo.printOne(foo(), 'hiya!')
printOne: hiya!
>>>

Searching the docs for bound, unbound, method and function turns up:

http://python.org/doc/current/lib/typesmethods.html


HTH,

--

Emile van Sebille
emile at fenx.com

---------



From niemeyer at conectiva.com  Sun Apr  7 05:34:58 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Sun, 7 Apr 2002 06:34:58 -0300
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: <slrnaass9f.178.mlh@vier.idi.ntnu.no>
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com> <mailman.1017895453.22508.python-list@python.org> <slrnaass9f.178.mlh@vier.idi.ntnu.no>
Message-ID: <20020407063458.B1600@ibook.distro.conectiva>

> >Testing if "a is True" is ok. True and False are singletons this is
> >the best way to know if 'a' is really 'True' (and not 1, [0], "foo",
> >etc). This test will be necessary sometimes.
> 
> It is neither worse nor better than using '==', IMO. Yes, it is a
> singleton, but the danger is using 'a is True' instead of simply 'a'.

It *is* worse. "a is True" works as expected, while "a == True" doesn't.
If the programmer doesn't know what "is" means, that's another story,
and then the warning should be "read the manuals, please". When PyChecker
becomes so pedantic to warn about everything a programmer that doesn't know
the language at all could be doing wrong, I'll stop using it.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From mgilfix at eecs.tufts.edu  Wed Apr 17 00:37:14 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 00:37:14 -0400
Subject: PEP 237 and long ints
Message-ID: <20020417003714.F16954@eecs.tufts.edu>

  Hi guys. I'm kinda new to the dev list so forgive me if this is
an old thread. Just getting acquainted with the community.  This
particular PEP though touched upon something that I had been dealing
with - getting my longs back into ints. Specifically, I did bitwise
operations on longs (because I had some addition that overflowed and
then took the modulus) and then wanted to cast the result back into
int. It got tricky if my long was particularly large say 4294967295L,
which is really just 0xFFFFFFFF binary and a valid int but difficult
to convert. Instead I had to do a hack where I subtracted a constant
to make the binary equivalent but to bring the value back into integer
range. The PEP seems to suggest making -1 equivalent to the hex long
value. Would this be sufficient to perform the actual type conversion?

  I kinda like having the concept of integer overflow though and am
not sure I'd want my ints to change to longs automagically. It's sort
of a quick sanity check to make sure some calculation isn't running
away (I've caught bugs like that). The PEP suggests having some sort
of warning that would be off by default and then could be turned on
explicitly. But I'm unable to see the difference between handling an
exception in that case explicitly and disabling the warning?  More so,
I'm a little fearful of such features simpley because I would hate to
see a scenario that parallels perl's 'use strict' where any decent,
well-mainting script always turns on such features.

            Regards,

                     -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/9a2ee721/attachment.sig>

From pyth at devel.trillke.net  Sun Apr 21 10:19:17 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 16:19:17 +0200
Subject: nested exceptions?
In-Reply-To: <just-DF77C7.15571021042002@news1.xs4all.nl>
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com> <a9ufpv$6hb$1@panix1.panix.com> <just-DF77C7.15571021042002@news1.xs4all.nl>
Message-ID: <20020421141917.GD10574@devel.trillke>

On Sun, Apr 21, 2002 at 03:57:10PM +0200, Just van Rossum wrote:
> In article <a9ufpv$6hb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
> wrote:
> 
> > ...and if you're using Python 2.2, write it as
> > 
> >     while oid not in objects:
> > 
> > 
> > Hmmmm....  it occurs to me that those of us teaching Python will need to
> > be *very* careful in making sure that people avoid 'in' over lists.
> 
> I'm already happy if they don't do
> 
>   if x in d.keys(): ...
>
> :-)


but that's better than:

    for key in d.keys():
    	if key==x:
		....

:-)

anyone worse-case (realistic) suggestions?

  holger




From laytonjb at bellsouth.net  Sun Apr 28 15:03:47 2002
From: laytonjb at bellsouth.net (Jeff Layton)
Date: Sun, 28 Apr 2002 15:03:47 -0400
Subject: Newbie: Problem with QTable
Message-ID: <3CCC4793.C3B2EE17@bellsouth.net>

Hello,

   I hate to bother everyone, but I'm trying to figure out how to
use QTable. I cooked up a small code that doesn't work (natch).
I'm attaching the code, but here's the error messages I receive:

[laytonj at home2 PYQT]$ /usr/local/python-2.2.1/bin/python table1.py
Traceback (most recent call last):
  File "table1.py", line 17, in ?
    table=MyTable(numRows, numCols)
  File "table1.py", line 12, in __init__
    self.setCaption("Table1")
AttributeError: sipThis


I've built and installed python-2.2.1 in /usr/local/python-2.2.1, sip
3.1 in /usr/local/sip-3.1, qt 3.0 in /usr/local/qt, and PyQt-3.1 in
/usr/local/PyQt-3.1.
   Could anybody give a newbie some pointers?

TIA,

Jeff



-------------- next part --------------
import sys
from qt import *
import qttable

numRows=20
numCols=20


class MyTable( qttable.QTable ):
    def __init__(self, rows, cols):
        """ help me """
        self.setCaption("Table1")



app=QApplication(sys.argv)
table=MyTable(numRows, numCols)     
app.setMainWidget(table)
table.show()
app.exec_loop()



From bobx at linuxmail.org  Thu Apr 18 16:51:02 2002
From: bobx at linuxmail.org (Bob H)
Date: Thu, 18 Apr 2002 20:51:02 GMT
Subject: ActivePython vs. Python
Message-ID: <WkGv8.38$Ud2.86702@news2.news.adelphia.net>

What is the benefit of getting/using ActivePython over Python?




From starx at pacbell.net  Wed Apr 24 04:35:10 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 08:35:10 GMT
Subject: Accessing class-level variables
References: <mailman.1019621385.21532.python-list@python.org> <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4> <mailman.1019624445.12502.python-list@python.org>
Message-ID: <Xns91FA105A11462RASXnewsDFE1@130.133.1.4>

Gerhard =?iso-8859-15?Q?H=E4ring?= || Tue 23 Apr 2002 09:58:46p:

> * Philip Swartzleonard <starx at pacbell.net> [2002-04-24 04:27 +0000]:
>> Gerhard =?iso-8859-15?Q?H=E4ring?= || Tue 23 Apr 2002 09:04:04p:
>> 
>> > I'm trying to access a class variable in a derived class. I
>> > sincerly hope there's a less kludgy way than what I'm currently
>> > using: [...]
>> > print self.__class__.a
>> 
>> Well, i knew this worked for dispatching __init__'s, but... well here
>> =):
>> 
>> [...]
> 
> Erh. Here you are accessing the class variable of class A explicitely.
> I know how that's working :) But it's not what I want. I want to
> access the class variable in a derived class. And even if I don't know
> what my derived class is. Here's an example that is more to my real
> code: 
> 
> import new
> 
> class X:
>      def p(self):
>          print self.__class__.a
> 
> def make_class(val):
>     klass = new.classobj("Y", (X,), {"a": val})
>     return klass
> 
> y = make_class(5)()
> y.p()
> 
> 
> This works, but using self.__class__.a is ugly.
> 
> Heh, now that I come to think of it, here I somehow have the class
> inheritance backwards :-) Guess there's no other way, right?
> 
> Gerhard

Er, i think you're trying too hard. +) Now that i realized what you 
really want, I just tried this...:

>>> class A:
	d=5

>>> class B(A):
	pass

>>> B.d
5
>>> class B(A):
	def p(self):
		B.d += 1
		print B.d
		print A.d

>>> B().p()
6
5
>>> B().p()
7
5

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From logiplexsoftware at earthlink.net  Mon Apr 29 18:52:07 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 15:52:07 -0700
Subject: (no subject)
In-Reply-To: <AC310CEA8E38D611BE2F00D0B74A0DD6073046@sumailb01.studentuniverse.com>
References: <AC310CEA8E38D611BE2F00D0B74A0DD6073046@sumailb01.studentuniverse.com>
Message-ID: <20020429155207.7773f0fb.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 18:37:19 -0400
Russo, Tom wrote:

> 
> > This PEP smacks of political requirements not practical ones. The
> > builtin name space is only crowded if we want to use the names for
> > something else.
> 
> I think that this hits the nail right on the head.  One of the reasons I
> like python is that it is friendly to functional programmers like myself.
> The goal of python _should not_ be to weed out "non-standard" features,
> which seems to be the goal of this PEP.  What python _should_ try to do (in
> my opinion at least) is take the best of all worlds, to the extent possible.
> The functions at issue here (with the possible exception of input) represent
> the best of functional programming integrated into python.  The fact that
> certain people never learned to program in the function idiom is no basis
> for removing them.

I am neither for nor against this PEP, but I'm wondering where this idea that
the PEP author wanted to *remove* funtionals came from.   He wants to put these
things in modules, not remove them.  In fact, he even suggested *adding*
additional functionals.


[from the PEP]
> The functionals (map, filter, and reduce) had greater importance prior
> to the introduction of list comprehensions which are now the preferred (and
> more readable) approach.  Reduce is easily (and more clearly) replaced by
> an equivalent for loop.  Limits on the number of built-in functions
> prevented the addition of other functionals which may have be useful
> when that style of programming is needed.  Creating a separate module allows
> room for a large variety of functionals (e.g. fold, head, tail, take,
> drop, etc.) to be introduced without cluttering the built-in namespace.


The only thing I've seen that he suggested removing was input().

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From martin at v.loewis.de  Mon Apr 15 02:38:36 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:38:36 +0200
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org> <3cba62f8@news.mhogaming.com>
Message-ID: <m3vgatlb4j.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> You may be right that I am getting confused about the hex
> display of longs-- That's because the display of a hex
> number with a '-' sign prepended is such a ridiculous
> notion

The display of numbers in the decimal system with a sign is such a
ridiculous notion...

Please remember what a positional system is; if B is the base, then
abcde means

   e + d*B + c*B**2 + b*B**3 + a*b**4

and -abcd means

   -e - d*B - c*B**2 - b*B**3 - a*b**4

I don't find this ridiculous, whether B is 10 or 16.

> There is another _BUG_ in Python:
> 
> >>> 0x40000000 << 1
> -2147483648
> 
> This should have been automatically converted to a long, not
> a negative int.

This is by design; read PEP 237.

> If we _ARE_ going to unify the long and int types, then
> let's bite the bullet, and DO IT RIGHT.  

In phase B.1, it will be right.

Regards,
Martin


From johnroth at ameritech.net  Sat Apr  6 20:44:07 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 20:44:07 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com>
Message-ID: <uav8og9h0sdqcd@news.supernews.com>

"djw" <dwelch91 at nospam.attbi.com> wrote in message
news:3CABB156.1010401 at nospam.attbi.com...
> Pardon my ignorance (I've only been working with Python for a short
> while), but I don't understand why an idiom like:
>
> if x == True:
>
> would be/should be an error (or warning initially). I see C/C++ code
> like this all the time:
>
> bool x;
> x = Foo(Bar);
> if (x==true)
> {...

Guido explains this in the latest version of the PEP. As I understand
it, if
True compared True to anything except the standard False
things (zero, empty sequences, etc.) then you could get
relationships like

6 == True == 7.

Since equality is transitive, this leads to the inescapable
conclusion that 6 == 7.

John Roth





From andersjm at dancontrol.dk  Wed Apr 10 04:43:28 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Wed, 10 Apr 2002 10:43:28 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
Message-ID: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk>

Pythons truth values are not "Something or Nothing".  The phrase seems
to have caught on following Laura's post, but I believe it is
misleading.

[] and {} are objects, they are objects with identity, and they are
mutable objects.  They are empty, but they are by no means "nothing".

Example:

def show(foo, history=None):
    """prints foo, optionally storing what was printed in a history list."""
    print foo
    if history:
        history.append(foo)

This works as intended:

    # Omit default argument, don't remember history
    show("hello world")

This works as intended:

    history = ["in the beginning was nothing"]
    show("hello world", history) # adds "hello world" to history list

But this doesn't:

   history = []
   show("hello world", history) # strange..history doesn't change here?

It doesn't work, because show() tests for emptyness with "if
show_history:".  What show() should have done is test for nothingness
using "if show_history is None:".

--
Anders Munch. Software Engineer, Dancontrol A/S, Haderslev, Denmark
Still confused but at a higher level.



From support at 4sms.org  Mon Apr  1 21:56:22 2002
From: support at 4sms.org (Free SMS Support)
Date: 2 Apr 2002 02:56:22 GMT
Subject: FREE!!! E-mail->SMS, WWW->SMS,  SMS-Games!!! Welcome to WWW.4SMS.ORG
Message-ID: <a8b6km$m1p$2340@dipt.donbass.net>

Absolutelly FREE!!!

Multilanguage, multiSMSservice gate  www.4sms.org : English, Germany Finnish, Russian!!! 

Send SMS FREE more than 50 countries! 


From abagaric_NOSPAM at rest-art.hr  Thu Apr 18 10:20:07 2002
From: abagaric_NOSPAM at rest-art.hr (Ante Bagaric)
Date: Thu, 18 Apr 2002 16:20:07 +0200
Subject: Assignment operator?
Message-ID: <a9mksa$8dqs$1@as201.hinet.hr>

After reading that PEP about static namespaces I'm now pretty confused with
the behaviour of the += operator.

The PEP states that if there is a binding of certain name anywhere within
the body of the function, it is considered local. Then it states that
binding operations include stuff like def, for, except... AND assignment.

I guess that for some reason operator += is considered an 'assignment'
(because it is often called 'one of assignment operators') but is it really?
Is 'a = a + 2' the _same_ as 'a += 2'? I don't think it is, or should be.
The 'a = a + 2' means 'store the value of a + 2 to the variable a' and 'a +=
2'
means 'increment a by 2'.

The fact is that Python already behaves exactly like that. If you say

>>> li = [1,2,3]
>>> li2 = li
>>> li += [4,5]
>>> li is li2
1

but if you say

>>> li = [1,2,3]
>>> li2 = li
>>> li = li + [4,5]
>>> li is li2
0

According to this, += operator can't be though of as an 'assignment
operator'.
And conversely, it shouldn't cause the compiler to consider 'foo' to be
local
if there is 'foo += stuff' somewhere within the body of the function.
After all, 'foo.append(stuff)' or 'foo.extend(stuff)' don't do it.

Or there is something I didn't quite understand?

Ante.





From jcmason at student.math.uwaterloo.ca  Thu Apr 18 02:28:40 2002
From: jcmason at student.math.uwaterloo.ca (Joe Mason)
Date: 18 Apr 2002 06:28:40 GMT
Subject: [OT] What is Open Source?
References: <20020418002823.C4068@phd.pp.ru> <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov> <mailman.1019085816.3890.python-list@python.org>
Message-ID: <a9lp2o$nfv$1@watserv3.uwaterloo.ca>

In article <mailman.1019085816.3890.python-list at python.org>,
holger krekel  <pyth at devel.trillke.net> wrote:
>> As I understand it, the GPL does not give you right to use the software
>> as you see fit.  You can only use it by linking with other GPL'd
>> software.
>
>cant be true: gpled software under windows wouldnt be possible...

And yet it happened.

The full text is available on a ton of web sites.  Go read it, and many of
your questions will be answered.  (You'd think that would be the first thing
to do *before* getting involved in a debate on the GPL.)

Joe


From garys at ihug.com.au  Mon Apr  1 16:29:56 2002
From: garys at ihug.com.au (Gary Stephenson)
Date: Tue, 2 Apr 2002 07:29:56 +1000
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <mailman.1017504362.25937.python-list@python.org> <slrnaacavs.2gs.philh@comuno.freeserve.co.uk> <a85od6$9i$1@panix1.panix.com> <DlTp8.48032$S52.1614804@news2.tin.it>
Message-ID: <a8ajpi$i36$1@lust.ihug.co.nz>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:DlTp8.48032$S52.1614804 at news2.tin.it...

> So, it's mutability that complicates things in these cases.  The
> exponents of Functional Programming have built rather powerful
> languages around immutable objects -- immutability may not be
> as easy and comfortable to work with in practice, and it does not
> model our computer systems all that well, but it sure does remove
> certain mathematical anomalies!

Agreed.  However, even with immutability, a client of the Rectangle class
could still create instances where height == width,  and imho it would be
both useful and cool for the system to automagically recognise and promote
such instances to having a most-specific-type of square.

Whilst the geometric arguments (rectangle/square, ellipse/circle) do have an
air of artificiality about them, there are numerous instances in the real
world where things _do_ mutate over time, and belong to varying sets of
categories as a consequence of such mutation.   Given that our objects _are_
allowed to mutate their state, wouldn't it be preferable for them to
do so in more intelligent ways?  Whether such a notion is actually
implementable in the real world is an entirely different question  of course
;-)

Simple immutability is also insufficient to enable the subclass == subtype
identity [1], which seems to me to be one of the fundamental issues
underlying all this.

cheers,

gary

[1] - http://okmij.org/ftp/Computation/Subtyping/




From sjmachin at lexicon.net  Mon Apr 29 20:47:03 2002
From: sjmachin at lexicon.net (John Machin)
Date: Tue, 30 Apr 2002 10:47:03 +1000
Subject: Remove masked elements from an array
In-Reply-To: <20020430002740.GT16217@devel.trillke>
Message-ID: <TNECXS1Y4WQCBD0VU52GB4ZT93OMV.3ccde987@Egil>

30/04/2002 10:27:40 AM, holger krekel <pyth at devel.trillke.net> wrote:

>On Mon, Apr 29, 2002 at 05:09:57PM -0700, John Machin wrote:
>> holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1020102370.5619.python-list at python.org>...
>> > On Mon, Apr 29, 2002 at 01:26:25PM -0400, PoulsenL at capecon.com wrote:
>> > > I have a Numeric array with missing elements from which I wish to calculate
>> > > a histogram.  However, in order to do so I must first remove the 'None'
>> > > values from the array.  I have a kludgy solution, but I was wondering if
>> > > someone could point me in the right direction on what is the proper method
>> > > of collapsing a masked array to just the non-masked values.
>> > 
>> > there are too solutions. choose yourself :-)
>> > 
>> > nums = [ 1,2,3,None,2,3]
>> > 
>> > #version1
>> > def f(x): return x is not None
>> > newnums = filter(f, nums)
>> > 
>> > #version2
>> > newnums = [ x for x in nums if x is not None]
>> > 
>> > The first version uses filter. 
>> > 'filter' calls 'f' for every item in 'nums' and puts
>> > the result into the new list iff it has a 'true' value.
>> > 
>> > The second version uses list comprehension.
>> > It basically does the same if you look closely
>> > at it. 
>> 
>> *Don't* try this if zero is a valid value for your histogram:
>
>i think you are on the wrong track.

Re-read my posting, especially the bit that you left out.
What you posted is correct as far as it goes.
I was pointing out to the OP that he shouldn't try

newnums = f(None, nums)
[this is the line that you mentally and physically snipped]

Grammar:
'Don't try this' <blah blah> <colon>
   <referent of 'this'>

HTH,
John






From opm at hrz.tu-chemnitz.de  Tue Apr 16 15:36:57 2002
From: opm at hrz.tu-chemnitz.de (Mirko Opitz)
Date: Tue, 16 Apr 2002 21:36:57 +0200
Subject: Newbie problem: global variables
Message-ID: <a9i1uk$1is$1@narses.hrz.tu-chemnitz.de>

Hi,

I've a small problem I need a variable which is present in all subfunctions
an also in the main prog but I don't want do give it in the header of all
the functions. Is there anyway to declaire it as Public or global like in
other languages?

Thx for now

Mirko




From opsys at voodooland.net  Thu Apr 25 21:29:19 2002
From: opsys at voodooland.net (Chris Watson)
Date: Thu, 25 Apr 2002 20:29:19 -0500
Subject: Python and Cocoa?
Message-ID: <20020426012304.7B1843FA5A@mars.sutv.com>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I just ordered my first mac, a dual Ghz G4 with OS X. (I know ill hate myself 
in july when Apple hopefully gets a faster bus that isnt in the stoneage and 
DDR ram.) Are there any hooks into Cocoa with Python? I have been searching 
google for a bit with not very either/or results. It seems someone was 
working on MacPython to work with Cocoa but im not sure what stage thats at. 
Does anyone know of anything useable to write Cocoa apps with Python?

Chris
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (FreeBSD)
Comment: For info see http://www.gnupg.org

iD8DBQE8yK1v7KmREXttmE0RArlWAJ9cur4oZ/3WQva7cNjxHBZGGEPqPACcC479
b9wuBnXhwhpsp1ouRgVWi9E=
=+44c
-----END PGP SIGNATURE-----




From aahz at pythoncraft.com  Thu Apr  4 16:22:30 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 16:22:30 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABEA67.FD3F75C0@alcyone.com> <a8hq6o$o9a$1@panix1.panix.com> <3CACA61D.C59BC837@alcyone.com>
Message-ID: <a8ig6m$mb$1@panix1.panix.com>

In article <3CACA61D.C59BC837 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>Aahz wrote:
>>
>> Did you read the final version?  C'mon, don't start arguments where
>> you have no clue what you're talking about.
>
>Hence why I said "alternate choice."  Sheesh.

There is no alternate choice following BDSL pronouncement.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From maxm at mxm.dk  Thu Apr 11 05:00:57 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 11 Apr 2002 11:00:57 +0200
Subject: KeyboardInterrupt errors in ASP pages
References: <508cb8ae.0204110045.431dadee@posting.google.com>
Message-ID: <3CB550C9.1000902@mxm.dk>

B wrote:

> I am using Python as the language for creating server side asp pages.
> The problem is that when using Response.Redirect(page), I seem to get
> the message below randomly but frequently enough to be a problem.
> 
> Has anyone else found this problem and is there a way around it. Is it
> a bug in the Response.Redirect code?
> 
> Python ActiveX Scripting Engine error '80020009' 
> 
> Traceback (innermost last): KeyboardInterrupt 
> 
> /Default.asp, line 0

As far as I understand it's a known bug, but without any known solution. 
I get it too at times.

I have just signed over a project to a customer on .asp/iis so I am a 
bit worried about it. But I don't really know what to do about it.

Maybe I should write a test harness to pound the IIS to see if I can 
provoke the error.

regards Max M



From peter at engcorp.com  Thu Apr 11 01:49:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 11 Apr 2002 01:49:49 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer 
 division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <mailman.1018495402.12499.python-list@python.org> <Xns91ECE5A88C0F8RASXnewsDFE1@130.133.1.4>
Message-ID: <3CB523FD.7D83C0C@engcorp.com>

Philip Swartzleonard wrote:
> 
> I must say I am impressed by a project with such a interactive crew out
> here in the public. Nothing fosters good feelings about a project's life
> as much as a good presense, i think.
> 
> Geoff again:
> > We're leeches (so far) but at least we're very grateful leeches.

Just for the record, 'twas I not Geoff who said that.  Geoff may
be a leech, he may even be a veritable wart on the bottom side of
society, but so far he hasn't admitted it here.  <wink>

'course-he-may-also-be-a-really-smashingly-nice-guy-ly yr's,
-Peter


From logiplexsoftware at earthlink.net  Fri Apr 12 14:47:24 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 12 Apr 2002 11:47:24 -0700
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer   division,backwards compatibility; where is  Python  going?)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEKHONAA.tim.one@comcast.net>
References: <1018508767.552240@yasure>
	<LNBBLJKPBEHFEDALKOLCCEKHONAA.tim.one@comcast.net>
Message-ID: <20020412114724.0191f125.logiplexsoftware@earthlink.net>

On Fri, 12 Apr 2002 14:46:18 -0400
Tim Peters wrote:

> Do a google search on Lumberjack's c.l.py posting history if you want to
> know where the hostility really comes from.  It starts with pure flamebait:
> 
>     This from a person who is constantly trying to score points on
>     Usenet.  A check of google shows you have an inflated notion of
>     yourself.  Maybe you should stop measuring others by your own limited
>     character.

Google can tell you whether people have inflated notions of themselves?  Is
that under "Advanced search"?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From eric.brunel at pragmadev.com  Mon Apr  8 11:24:30 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 8 Apr 2002 15:24:30 +0000
Subject: __init__.py
References: <3CB19437.C1B2B3E9@gol.ge>
Message-ID: <a8s5dr$1vb5$1@norfair.nerim.net>

Giorgi Lekishvili wrote:
> I wonder if an experienced developer in Python gives me a few hints on
> how to make the initfile of a package.

What do you want to do? I've done quite a few Python packages already and 
my __init__.py files have always been empty, which perfectly suited my 
needs... Maybe you have some other requirements, but if you do, please 
state them...
 - eric -



From aleax at aleax.it  Sun Apr 14 15:34:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 14 Apr 2002 19:34:45 GMT
Subject: Specifying a time from previous day
References: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net>
Message-ID: <pRku8.15554$b62.440441@news1.tin.it>

jrup wrote:

> Hi,
> 
> This is my first post to this group and I am new to python. I have some
> experience with perl and vbscript for sysadmin-type scripting, but am not
> a programmer.
> 
> I need to search some log files, but only if they were created after
> 4:00pm the previous day.

>>> import time
>>> time.time()
1018811834.620291

this gives us the time in seconds since some arbitrary starting point
("the epoch").  24 hours ago was clearly 24*60*60 seconds ago, i.e.:

>>> time.time()-24*60*60
1018725558.410575

Given this representation for time, we can obtain one in terms of
year-month-day-hour-minute-second-etc (9-items tuple) -- say
in local time (GMT, aka UTC, is the alternative):

>>> yesterday = time.localtime(time.time()-24*60*60)
>>> yesterday
(2002, 4, 13, 21, 20, 29, 5, 103, 1)
>>>

As you want a given time, you'll only get the first three items of
this (year, month, day), then your specified time (16, 0, 0), then
2 zeros, and finally the same last item as yesterday (DST flag: this will 
break by 1 hour when DST has just started or finished -- if you could have 
your specs in UTC...); time.mktime can rebuild the "time in seconds
since the epoch" from this:

>>> threshold = time.mktime(yesterday[:3]+(16,0,0)+(0,0,yesterday[-1]))
>>> threshold
1018706400.0

When you examine a file, os.stat gives you a lot of data about it:

>>> os.stat('/home/alex/.bashrc')
(33188, 32453L, 776L, 1, 501, 501, 403L, 1018648057, 1018267187, 1018267187)

The last three numbers are times in the seconds-since-the-epoch
form -- last access, modification, inode change.  It appears that what
you want is to look at files such that os.stat(thefilename)[-2] > threshold,
where threshold is computed as above.


I hope this information is enough for you to complete your program.


Alex



From paul at boddie.net  Tue Apr  9 04:14:11 2002
From: paul at boddie.net (Paul Boddie)
Date: 9 Apr 2002 01:14:11 -0700
Subject: Division change in 3.0! (was Re: Booleans, integer division, backwards compatibility; where is Python  going?)
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com> <a8st6c$jlp$1@panix1.panix.com>
Message-ID: <23891c90.0204090014.30a0b0a1@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<a8st6c$jlp$1 at panix1.panix.com>...
> 
> Once again, for the record: the mandatory change in the way division
> works is going to be in 3.0.

I forget these things rather easily. ;-)

> This is not the way it was originally presented (and I agree with you
> about the presentation), but it's a good example of Guido changing his
> mind after getting input.

Yes. Contrary to popular belief, too. :-)

Paul


From alwyn at alwyn.demon.co.uk  Mon Apr 22 06:36:35 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Mon, 22 Apr 2002 10:36:35 GMT
Subject: Interesting topics
References: <mailman.1019468028.28617.python-list@python.org>
Message-ID: <alwyn-94496B.11363122042002@news-text.blueyonder.co.uk>

In article <mailman.1019468028.28617.python-list at python.org>,
 Doru-Catalin Togea <doru-cat at ifi.uio.no> wrote:
> 
> Just curious: I have come across some pretty weird statements in emails,
> like the two spam mails supposedly from Jessica F. In the last one, one
> can read:
> 
> 	Who will we open after Bernadette kicks the pathetic star's boat?
> 
> 	We fear the polite gardner.
> 
> I wonder, are these statements computer generated? 

Yes. I believe the Hipcrime tool does this. 

> Or do they make sense
> if you have seen a certain movie, a TV-show, or read a certain book? I
> don't want to hurt anybodys feelings, but TO ME, their being sent in
> emails seems like the symptoms of a mental disorder of some kind.

The point of the random 'gibberish' sentences is to make each message 
unique, so as to evade the spam detectors. 


Alwyn


From boud at valdyas.org  Sat Apr 13 05:20:16 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 13 Apr 2002 11:20:16 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl> <m3662wypqt.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cb7f8a7$0$85437$e4fe514c@dreader3.news.xs4all.nl>

Martin v. Loewis wrote:

> Boudewijn Rempt <boud at valdyas.org> writes:
> 
>> Then I tried to write the utf-8 data to a file. I have tried to
>> construct that file with two methods:
>> 
>>     f = open("syllables", "w+")
>>     d2 = codecs.EncodedFile(f, "unicode_internal", "utf-8")
>>     f2.write(u"a")
>>     f2.close()
> 
> This can't work (even though it should not crash). The EncodedFile
> performs transparent recoding from two named encodings. In this
> context, unicode_internal is the name of a *byte* encoding, namely the
> encoding which exposes the internal memory layout of Unicode objects.
> 

Right. That wasn't clear to me.

> 
> import codecs
> f3 = codecs.open("syllables2", "w+", "utf-8")
> f3.write(u"\N{LATIN LETTER GLOTTAL STOP}")
> f3.close()
> 
> print repr(open("syllables2").read())
> 
> I get
> 
> '\xca\x94'
> 
> which indeed is the UTF-8 representation of the glottal stop. What did
> you get?
>

'\xc3\x8a\xc2\x94' ...

At least, when I do exactly what you did, I get the same result.

Oh, verdraaid, I see what I did wrong. On my previous computer, which
has melted down, I had hacked sys so my default encoding was the same
as that in my locale -- utf-8.

On my new computer, I just compiled and installed the latest Python,
and I can only enter source in ASCII and escape sequences.

Was PEP 263 implemented already? Must investigate...


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


From uwe at rocksport.de  Fri Apr 12 10:16:56 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 12 Apr 2002 14:16:56 GMT
Subject: how to print to printer ?
Message-ID: <a96q8o$9cd5d$1@hades.rz.uni-sb.de>

Hi,

I'd like to send "raw" data to a printer, how can I do this,
using Python ???

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From tim.one at comcast.net  Wed Apr 10 19:38:28 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 19:38:28 -0400
Subject: binding a reference to a variable
In-Reply-To: <yu99662zmufg.fsf@europa.research.att.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEAKONAA.tim.one@comcast.net>

[Andrew Koenig]
> Sure.  I'm thinking about how to implement Snobol-style pattern
> matching in Python.  I know that someone has done it before, but
> I want to approach it from a completely different direction.

I did that a looong time ago, but the code didn't survive my then-employer's
financial meltdown.  Unless you bought their assets at the fire sale, I'm
not sure how you could know whether your approach is completely different
<wink>.

It is, though.  The use of unevaluated expressions in SNOBOL4 is crucial for
advanced matching, and the "how do you do that with a bare name in Python?"
problem seems unsolvable in a squeaky clean way.  Like:

    ARB = NULL | LEN(1) *ARB

I ended up with something like

    import S

    arb = S.var('arb')
    arb.set(S.null | (S.len(1) + arb))

I could get *used* to the subtleties (since the Python binding for arb never
changes, cross- and self- references in the pattern-graph structures comes
for free), but fiddle it as I might, it never felt wholly natural.

> ...
> Focusing just on the  (break("@") . id)  subexpression, how would
> one implement it?  That's what led me on this path.

After I left that path <wink>, I found it was much easier to model Icon's
spelling of patterns, and with generators in 2.2 that can be a breeze.
However, Icon pattern-matching has a more procedural flavor, and SNOBOL4
pattern-matching a more declarative flavor, and each has strengths and
weaknesses following from that.

Before I left that path, it looked something like

    id = S.var('id')
    ... id.ass(S.break('@')) ...

were .ass() had an optional argument to choose between delayed (.) and
immediate ($) assignment, defaulting to delayed.  BTW, operator overloading
was much harder to predict in Python back then, so I often settled for
ordinary method-call notation.  After operator overloading because easy to
predict in Python, I learned to hate it from being forced to endure reading
code that used it <0.7 wink>.

An iteration before that I used string names as keys to an explicit
namespace dictionary (as others have suggesting), but found that modeling
"SNOBOL4 variable" as a class in its own right felt at least a little
better, despite needing explicit "declarations", i.e. lines of the form

    var1, var2, ... = S.var(name1, name2, ...)

Remembering to do that once was less painful than needing to slop in quotes
everywhere.

Have fun!  It's an interesting challenge.





From srh232 at nyu.edu  Tue Apr 16 17:35:24 2002
From: srh232 at nyu.edu (Shane Hoversten)
Date: Tue, 16 Apr 2002 17:35:24 -0400
Subject: Simulating multi-dim array problem - OR - reference confusions
Message-ID: <3CBC991C.7090202@nyu.edu>

Hi -

	I'm doing code that needs the functionality of multi-dimensional arrays. 
  I'm getting around this right now by using lists of lists, which seems 
to work right.  For instance:

 >>> a = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
 >>> b = a[0]
 >>> b[0]
1
 >>> a[0][0] = 20
 >>> b[0]
20

	This is all well and good; it seems to imply that when you say:

b = a[0]

	That b is actually a reference to the appropriate list in a, which is 
just what I need.  The problem is that I'm writing a program that needs 
to work with a chained structure.  Specifically, each element of the 
list of lists contains a "pointer" to another element in the list of 
lists.  These "pointers" can be traversed much like a linked list.  In 
other words, an "element" looks like:

[x, y, z, <PTR>]

	where <PTR> is set the way I described above, with b = a[0], or, in this 
case, a[0][3] = a[10], etc.

	Here's are my questions:

	1) Am I going to get into trouble assuming (using the example above) that 
b is basically a reference to the sublist a[0]?  I'm not clear how 
sharing works in python.

	2) Is there a way I can print out the address of a "reference"?  In 
particular I want some concise symbolic representation - like a memory 
address, or a handle id, or SOMETHING - and not a verbatim 
reconstruction of the list structure which is how python's "print" 
function does it.  If I have elements containing references to elements 
containing references to elements it's impossible to follow what's going 
on with the chain.  I'd like to have it print something like:

[[x, y, z, 0x123123], [a, b, c, 0x123958], [q, w, e, 0x358481]]

	or something to that effect, instead of interpolating and making a huge mess.

	I've tried to find references (no pun intended) to these issues but have 
come up empty.  I know the info is out there.  Can someone either tell 
me or tell me where to find it?

Thanks,

Shane



From martin at v.loewis.de  Sat Apr 13 04:19:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 13 Apr 2002 10:19:54 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl>
Message-ID: <m3662wypqt.fsf@mira.informatik.hu-berlin.de>

Boudewijn Rempt <boud at valdyas.org> writes:

> Then I tried to write the utf-8 data to a file. I have tried to
> construct that file with two methods:
> 
>     f = open("syllables", "w+")
>     d2 = codecs.EncodedFile(f, "unicode_internal", "utf-8")
>     f2.write(u"a")
>     f2.close()

This can't work (even though it should not crash). The EncodedFile
performs transparent recoding from two named encodings. In this
context, unicode_internal is the name of a *byte* encoding, namely the
encoding which exposes the internal memory layout of Unicode objects.

Unicode objects themselves have no "encoding"; they are just sequences
of Unicode characters. That they have an internal representation also
should not matter in most cases.

So what happens here is that you pass a Unicode object to the
unicode_internal decoder, which expects a byte string. u"a" will be
converted to the byte string "a" which is then interpreted as an
internal encoding of a Unicode object, which it is not.

> The second method writes garbage-encoded data to the file:
> 
>     f3 = codecs.open("syllables2", "w+", "utf-8")
>     f3.write(u"?")
>     f3.close()

That should work fine, and it works for me in all cases. 

> (Where u"?" contains any Unicode character you like -- in this
> case a glottal stop.)

When I do

import codecs
f3 = codecs.open("syllables2", "w+", "utf-8")
f3.write(u"\N{LATIN LETTER GLOTTAL STOP}")
f3.close()

print repr(open("syllables2").read())

I get

'\xca\x94'

which indeed is the UTF-8 representation of the glottal stop. What did
you get?

Regards,
Martin


From joost_jacob at hotmail.com  Tue Apr 30 15:51:12 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 30 Apr 2002 12:51:12 -0700
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com> <13285ea2.0204290619.3563152f@posting.google.com>
Message-ID: <13285ea2.0204301151.1cf35ce@posting.google.com>

Here is the GUI version.  It has menu options for a new fortune and
for copying to the system clipboard.  The whole thing is 104 lines of
source code.

Usage:
Put fw.py and your 'allfortunes' file somewhere in your sys.path.  The
format of the allfortunes file should be like the variable called
modulequotes in fw.py; modulequotes is used if it cannot find the
allfortunes file.
Now do:
python -c "import fw; fw.run()"

It should be portable and backwards compatible to python version
1.5.2, please let me know if it does not work with your python
installation.

Joost Jacob
www.liacs.nl/home/jjacob/

I like work ... I can sit and watch it for hours.


--------------------  fw.py  --------------------------------------

modulequotes = """At the end of every pot of gold there's a rainbow
%
To a three year old with a hammer, everything looks like a nail
%
"""

from Tkinter import *
import string, random, os, StringIO

quoteslist = []

def getQuote():
    """
    Return a quote <string> from file allfortunes,
    assume the file allfortunes is located in the same
    directory as this module."""
    if not quoteslist:
        try:
            datafile = os.path.join(
              os.path.dirname(__import__(__name__).__file__),
              'allfortunes' )
            quotesfile = open(datafile)
        except IOError:
            quotesfile = StringIO.StringIO(modulequotes)
        quote = ''
        for s in quotesfile.readlines():
            if s[0] == '%':
                if quote:
                    quoteslist.append(quote)
                    quote = ''
            else: quote = quote + s
        if quote: quoteslist.append(quote)
    return random.choice(quoteslist)


class TextMap(Frame):
    def __init__(self, parent=None, height=25, width=80):
        Frame.__init__(self, parent)
        self.width = width
        self.height = height
        self.pack()
        self.text = Text(self, wrap='none', relief=SUNKEN,
            font=('courier',10,'normal'),
            height=height, width=width)
        self.text.pack(side=LEFT, fill=NONE)
        self.text.config(state=DISABLED)
    def resize(self, height=None, width=None):
        if height:
            self.height = height
            self.text.config(height=height)
        if width:
            self.width = width
            self.text.config(width=width)
    def settext(self, textstr=''):
        self.text.config(state=NORMAL)
        self.text.delete('1.0', END)
        self.text.insert('1.0', textstr)
        self.text.update()
        self.text.config(state=DISABLED)
    def clear(self):
        self.text.config(state=NORMAL)
        self.text.delete('1.0', END)        # delete current text
        self.text.update()
        self.text.config(state=DISABLED)


class FortuneWindow:
    def showQuote(self): 
        self.root = Tk()
        self.root.title('Fortune')
        # add menu
        top=Menu(self.root)
        self.root.config(menu=top)
        top.add_command(label='Next', underline=0,
          command=lambda self=self: self._onNew())
        top.add_command(label='Copy', underline=0,
          command=lambda self=self: self._onCopy())
        # add fortune in textmap widget
        self.quote = getQuote()
        self._getDimensions()
        self.textmap = TextMap(height=self.height, width=self.width)
        self.textmap.settext(textstr=self.quote)
        # catch events
        self.root.mainloop()
    def _getDimensions(self):
        self.width = 0
        self.height = 0
        for s in string.split(self.quote, '\n'):
            self.height = self.height + 1
            if len(s) > self.width: self.width = len(s)
        if self.width == 0: self.width = 10
        if self.height == 0: self.height = 2
    def _onCopy(self):
        self.root.clipboard_clear()
        self.root.clipboard_append(self.quote)
    def _onNew(self):
        self.quote = getQuote()
        self._getDimensions()
        self.textmap.resize(width=self.width, height=self.height)
        self.textmap.clear()
        self.textmap.settext(textstr=self.quote)

def run():
    FortuneWindow().showQuote()


From ykingma at accessforall.nl  Sat Apr  6 19:24:31 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sun, 07 Apr 2002 01:24:31 +0100
Subject: PEP 285: Adding a bool type (partially OT)
References: <mailman.1017466974.6751.python-list@python.org> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAE24A2.7ECBB78@accessforall.nl> <onpr8.64923$S52.2260694@news2.tin.it>
Message-ID: <3CAF91A9.EE3607CC@accessforall.nl>

Alex,

Alex Martelli wrote:
> 
> Ype Kingma wrote:
>         ...
> >> Even if having to modify the __getitem__, __setitem__, __delitem__ (at
> >> the least) of a zillion user-defined types out there was feasible (and
> >> it's not), I think this would not fully exhaust, as I indicated.
> >
> > The users writing such code were knowingly taking a risk by using
> > a name that starts with a double underscore. For these cases I think
> > deprecation is quite acceptable.
> 
> You completely misunderstand the role of special names in Python.
> It doesn't mean "I'm taking a risk": it means "I'm following fully
> documented, STABLE interfaces".

The language spec does not specify the type of the index or key
argument to __getitem__.
As for the possible types of this argument you are at the mercy
of the possible types and the program doing the indexing.
So indeed, the double underscore does not matter: this is also
the case for normal methods in user code.

<snip>

> > When the object passed to your __getitem__ method does not
> > have .start and .stop attributes you could simply call int()
> > on it from within your __getitem__ implementation.
> 
> Incompatible with built-in lists, which raise an exception when
> indexed with complex or float numbers for example.  There are
> good reasons for that, and it's unwise to suggest breaking the
> mold with as little understanding of those reasons as your posts
> are consistently suggesting.

You didn't mention before that you need compatibility with built-in
lists. I did wonder what your indexable objects would be doing, but
I preferred not to speculate on that.
 
> > But you would only need to do this in certain cases, see below.
> 
> In fact, doing it in most other cases would be very unadvisable.

Surprise, surprise: a bit of agreement.

> >> > Indexing a dict with both int and bool keys might involve some
> >> > confusion when True is not 1 anymore. This case should be rare,
> >> > however.
> >>
> >> Why should it be rare?  All hashables are used uniformly as keys
> >> into dictionaries.
> >
> > It should be rare because the confusion would only occur when both
> > True and 1 are used as keys and they are intended to be the same key.
> 
> Any existing code doesn't know about any distinction between True and 1.
> There is no distinction between them up to the current release.  Therefore,
> any existing code will become utterly fragile if and when a distinction
> should suddenly appear out of the blue.  From the point of view of code
> implementing a mapping, there's no "intended to be the same" -- as there
> is no notion of any distinction between two cases (there is only one case
> today).  Similarly for client-code, code that USES a mapping (most often
> polymorphically).  What "intended to be" can it mean (or fail to mean)
> between objects that's always been identical?  For example, using a dict
> as a set has always yielded identical results for:
>     - testing membership with aset.has_key(item),
> or:
>     - indicating membership with aset[item] = 1, then
>     - testing membership with aset.get(item, 0)
> 
> Why would you think this is "rare"?  It's common currency.  How many tens
> of thousands of lines of python have you written, reviewed, or maintained,
> that you presume to teach me what is common and what is rare in python
> "as she is spoken", pray?

In how many of these cases would the 'item' key actually be confused
when it could be either a bool or an int? So far you failed to convince me
that this is not a rare case.

Even though I don't see the relevance of the second question
the answer is: a few, and I'm not teaching here, I'm explaining
my opinion.

> > That should only happen when True and 1 are used interchangeably,
> > which might happen by programmer mistake or by using
> > old (1/0) and new (True/False) style code together.
> 
> Old-style only will do perfectly well, since lots of extremely common
> idioms therein (a<b, d.has_key(k), x in y, and so on) change return
> values silently.  Neither any mistake nor any 'together' is needed,
> unless you're somehow trying to imply that anybody coding some kind
> of "advanced" (?!) construct such as "a<b" was using "new style code"
> unwittingly years ago...

By new style I meant explicit use of True/False. As for "advanced (?!):
I don't have a time machine.
 
> > The slowdown would be there for backwards compatibility only.
> 
> In yoyr proposal the slowdown would be there in every dict access.
> 
> > In the pure case an int() call would not be needed at all:
> 
> But to check whether it's needed or not, a type-check would be
> needed, and that IS a substantial slow-down.  You don't seem to

If it's not needed at all, there is also no need for the check.

<snipped some irrelevant rank pulling>

> > When adding the int() call to the __setitem__ method of the standard
> > dicts for backward compatibility the dict shown above would only a have
> > a single key 1 and the value would be the last value put there.
> 
> ...and pystones would become...?
> 
> >> "only hope" is not acceptable in this context.  Fortunately, we can
> >> do better than that (and I'm confident that the unacceptability of
> >> slowing down dictionaries and lists is quite obvious to decision
> >> makers in this case, so I don't fear it will happen).
> >
> > I'm still putting my hopes on deprecation (as above) after PEP 285 enters
> > into force. During the deprecation phase with non int bools such cases
> > could be warned for eg. by a special verion of int() that would also cause
> > a slowdown, and after the deprecation a pure version with normal speed
> > would remain.

And now for some fine gorgheggio:

> We have a saying in Italy -- "chi vive sperando muore cantando".  "He
> who lives hoping, dies singing".  I suggest you practice gorgheggio -- not
> a popular singing technique any more, alas, but should you ever find
> yourself cast in some (e.g.) Salieri or Gluck (Italian-style) it might come
> in handy.  And if you don't understand a word I'm saying, may I suggest
> remedying that with a course in opera buffa -- its 17-th century and
> earlier roots, its 18-th century apogee (you'll find debate on that timing,
> but that's the school I subscribe to), its 19-th century long, slow decline
> (ditto).  With any luck, such a course might just happen to show you the
> career you were always destined to, and distract you from subjects you
> appear to be unsuited for, such as programming language design and
> implementation.  One can always hope (recursive call to the beginning of
> this paragraph).

End of fine gorgheggio. Applause.

As it happens I have enjoyed an opera buffa from the 18th century live
towards the end of the previous century. So, I guess there is still hope.
Moreover the BDFL today called the new bool type 'impure' on c.l.p.,
giving me even more hope.

> > P.S. I havn't followed the change of behaviour of the / operator in
> > detail, but wasn't/isn't this done using a deprecation period with
> > warnings, too?
> 
> And commandline switches (see -Q, still alive and kicking) and so
> on.  It's not a matter of "following the change", you just need some
> modest amount of studying the current Python release.
> 
> Now a surprise-test: how many divisions are there in a typical piece
> of Python, versus, how many dictionary lookups...?  Fair warning: this
> IS a trick question.  Only sensible way to answer it is, instrument the
> Python compiler.  If you find this a problem, then you're not very

Instrumenting the compiler would only give you the number of times
these operations are present in the source code.
To measure the more relevant actual numbers of executed operations I'd
rather instrument the python byte code. In case these operations can be
deduced from the byte codes, it might be easier to instrument the
byte code interpreter.

> qualified to comment on language implementation strategy, are you...?

I am posting on c.l.p, not on python-dev. 
I am not talking about an actual implementation.
I'm talking about a specification, while trying to keep an eye on
implementABILITY.
Since you didn't say it cannot be done, I'm actually encouraged.


Ype

-- 
email at xs4all.nl


From starx at pacbell.net  Tue Apr  9 03:25:04 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 9 Apr 2002 07:25:04 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <Xns91EB472B5F17RASXnewsDFE1@130.133.1.4>

Ryan || Mon 08 Apr 2002 07:58:05p:

> I know this is way off topic, but I figure the heavy hackers here
> will understand my mindset better than those at comp.lang.C

I don't know if I count as a heavy hacker, but what the heck =)

> I have been doodling with python (nasty habit I know but I have to
> do somthing with my snake) and think I may be ready to learn a
> second language.

Computer languages are surprisingly like human ones, in that learning a
second teaches you a lot about your first.

> I am leaning towards C over C++ and Java even though I loose objects
> it will ground me firmly in functional programing.

Well, I think you mean procedural programming ... functional
programming is that kind of weird abstract alien lambda stuff that you
find in LISP and such (at least that's how I understand it.)

C is a nice solid language, and quite usable provided your programs
don't get exceedingly complex, personally I really can't recommend it
for large projects after what I've seen in just a few medium sized
ones. That's what C++ is for... dealing with insane amounts of
complexity without exploding... which is why it itself is so complex
and hard to learn. But if you started at python, and already
understand all of this fancy object stuff, you don't really have to
worry about all that 'learn c or c++ first' arguing anymore.

Oh, and raw C is easier to use as a python extension, I think. That
might be nice =).

> I am wondering what book would be the best for a begginer? The two I
> have seen that look interesting are

www.accu.org is the place to go for book reviews and recommendations
for c and c++. If you had asked this question in the form you
mentioned, someone would have dropped this link within ten seconds :).

> C for Dummies (parts one and two) or

I really don't know anything about this book, but I've heard that the
dummies series in general tends to be insulting and talks down to you
because, well, you're a dummy ;).

> The C Programming Language
> by Brian W. Kernighan, Dennis M. Ritchie.

I don't really use C very much anymore, but this book is where I go
when I need to remember what something or other does. If you are just
starting in the language, it is probably too dense and too brief in
explanation (unless you're good at that 'hear one, know ten' thing :).
(Also, i doubt that the first one's still in print, but make sure you
buy the second edition, that covers standardized C)

> I am looking for some input from someone who knows and knows the
> problems I will face coming from a __nice__ language like python.
> Thanks (and sorry about the off topic)

The first problem: Learning out of date or bad practice things like
'void main()' from out of date or just plain bad books and teachers.
This is what ACCU and usenet are for =).

The second, and biggest problem: Pointers. Pointers pointers pointers
and some more pointers. Pointers are to C what subject-object-verb
sentence order is to Japanese: they are The Thing That Makes The
Language Hard To Learn. Coming from a background where you already
understand programming logic and some data structures and what not
should make things _much_ easier than they could be, but make sure you
fully and throughly understand this topic before you try to do much
of anything with dynamically allocated ... anything, cause this is how
all of it must be handled in C.

Okie, that's enough ranting about C on a Python form =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From marklists at mceahern.com  Tue Apr 30 15:37:59 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 14:37:59 -0500
Subject: import from a file in a subdirectory
In-Reply-To: <644f6688.0204301127.5d3372d@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEHMKNAA.marklists@mceahern.com>

[rdack]
> i have a script running in a directory.
> i want to call a subroutine in a file in a subdirectory.
> 
> in the first file i said  from subdir/ff.py import  subrtn
> python doesn't like the slash.
> 
> how do i do it?

Short answer:  Look into something called "packages".

Demo:

  parent/
    a.py
    /subdir/
      __init__.py
      b.py

# a.py
from subdir import b
b.foo()

# __init__.py
# The mere presence of this file makes the dir a package.
pass

# b.py
def foo():
  print "lo and behold"

// mark




From sholden at holdenweb.com  Sat Apr  6 09:56:55 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 6 Apr 2002 09:56:55 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org> <slrnaan32v.n0n.philh@comuno.freeserve.co.uk> <3CACA68C.C7B44901@alcyone.com>
Message-ID: <k2Er8.26244$q63.13414@atlpnn01.usenetserver.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CACA68C.C7B44901 at alcyone.com...
> phil hunt wrote:
>
> > On 03 Apr 2002 10:59:13 -0500, Andrew Koenig <ark at research.att.com>
> > wrote:
> >
> > >Specifically: If I want to write a function that answers a yes/no
> > >question, I have lots of possible ways of spelling yes (1, 2, "yes",
> > >and so on) and lots of possible ways of spelling no (0, {}, None,
> > >and so on).  There isn't a single preferred way.
> >
> > Yes there is, 1 or 0.
>
> No, there isn't.  Some return 0 or 1, some return 0 or an unspecified
> integer, some return None or an object, etc.
>
I think the post to whom you are responding would probably take the view
that the value returned by the comparison and logical operators (==, !=,
not, etc.) is preferred. For that purpose Python currently uses 1 or 0.

In other words, for all x, not not x is in [0, 1]

regards
 Steve






From wilk at flibuste.net  Tue Apr 16 03:25:17 2002
From: wilk at flibuste.net (William Dode)
Date: Tue, 16 Apr 2002 09:25:17 +0200
Subject: google apis
Message-ID: <20020416092517.5a9e5845.wilk@flibuste.net>

hi,

I just read this :

With the Google Web APIs service, software developers can query more than
2 billion web documents directly from their own computer programs. Google
uses the SOAP and WSDL standards so a developer can program in his or her
favorite environment - such as Java, Perl, or Visual Studio .NET.

Is there a python api somewhere ?
Google is known to use python, maybe we must write to them to ask a python
wrapper for their api ?

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From aleax at aleax.it  Thu Apr  4 13:53:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 18:53:08 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com>
Message-ID: <oi1r8.50382$pT1.1425832@news1.tin.it>

A. Jones wrote:
        ...
> Furthermore, you can wrap it inside a class.

But then there's no need for the horrors of exec and you can do it MUCH
more cleanly and safely, to wit:

> class test:
> def wrongHeaded(self, name, value):
> exec "self."+name+'='+repr(value)
> print vars()

class test:
    def rightHeaded(self, name, value):
        setattr(self, name, value)

You can print vars(self) too, if you want.  But the point is that setattr
is perfectly clean and safe, well-performing, specific and restricted in
its semantics -- all the good things that exec is not.

But of course you can't use setattr to create *VARIABLES* -- only
*ATTRIBUTES* of some object (you could use it on a module to make
"global variables" for that module, but that would not be quite as safe
and advisable as using it on an instance object).

It's exactly because it's so convenient to use (e.g.) attributes of an
instance, rather than variables, for such purposes, that it's unjustifiably
gauche to abuse exec in such wise.


Alex



From torkel at sm.luth.se  Tue Apr 23 04:17:17 2002
From: torkel at sm.luth.se (Torkel Franzen)
Date: 23 Apr 2002 10:17:17 +0200
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> 	 <mailman.1019299357.11513.python-list@python.org> 	 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> 	 <Xns91F67C40EB0E0lumberjack@216.148.53.83> 	 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> 	 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> 	 <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at> <vcbk7qzezty.fsf@beta13.sm.luth.se> <3CC51589.B9D9658B@kfunigraz.ac.at>
Message-ID: <vcbhem2g776.fsf@beta13.sm.luth.se>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:

> "Wittgenstein hielt Mengenlehre und mathematische Logik f?r entbehrliche
> Teile der Mathematik."
> 
> Which means: "Wittgenstein says that we don't need any set-theory and
> logic".

  Not quite. It means "Wittgenstein regarded set theory and
mathematical logic as dispensable parts of mathematics." Now,
Wittgenstein waffled quite a lot about mathematics, and in
the Philosophical Grammar he said a lot of things that sounded
much like intuitionism, but he also at various times said that
he did not wish to change anything at all in mathematics.

 > I wrote an email to the official Austrian association for "Kurt Goedel"
 > in order to figure out whether Goedel met Wittgenstein or not.
 > Wittgenstein was in Vienna (and many other times) from 1924 to 1928 and
 > he even had contact with the Wiener Kreis (due to Schlick) but it
 > remains unclear for me whether he really met Goedel or not.

  Wittgenstein was a great hero of Schlick's and Waismann's, but was
of course in no sense a member of the Vienna Circle. Neither was
Godel, although he sat in on some of their meetings. I don't think
there is any record of Godel and Wittgenstein ever meeting.


From erno-news at erno.iki.fi  Wed Apr 10 22:23:55 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Apr 2002 05:23:55 +0300
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
Message-ID: <ku7knfynus.fsf@lasipalatsi.fi>

In article <3cb4bd45$1_1 at goliath.newsgroups.com>, "Brad Clements"
<bkc at Murkworks.com> writes:

| Could someone explain this to me.
| Lots of code does  type(x) in [InstanceType]

| One might imagine that to support new style classes, this would be changed
| to:

| type(x) in [InstanceType,ObjectType]

it is not a very useful contruct, since everything in python is an
object.

i suspect you have not fully grokked the idea of the 2.2 object
system changes.

the whole idea is to remove the distinction between "types" and
"classes".

  -- erno


From greg at cosc.canterbury.ac.nz  Wed Apr 10 02:35:23 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 18:35:23 +1200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1018314090.9984.python-list@python.org> 
		<_1vs8.1787$CA6.114998@newsread1.prod.itd.earthlink.net> <mailman.1018331915.6071.python-list@python.org>
Message-ID: <3CB3DD2B.D23647B9@cosc.canterbury.ac.nz>

Ian Bicking wrote:
> 
> But
> that doesn't mean prototyping Pyrex code in interpreted Python wouldn't
> be very nice -- especially since Pyrex runtime errors are unlikely to
> have nice tracebacks

I hope to improve the traceback situation at some
point, so you'll get entries in the traceback for
calls between functions in the Pyrex module.

> Maybe
> all that's required is a custom loader that filters out cdef and type
> declarations.

It wouldn't work, unless the Pyrex code restricted
itself to a very boring subset of C types (essentially
just numbers and strings) and was careful what it
did with them.

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


From robin at jessikat.fsnet.co.uk  Mon Apr  8 05:51:50 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 8 Apr 2002 10:51:50 +0100
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
 <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk> <3CAFAAA4.41EA7CF9@python.org>
Message-ID: <+CcOX7A2gWs8EwPR@jessikat.fsnet.co.uk>

In article <3CAFAAA4.41EA7CF9 at python.org>, Guido van Rossum
<guido at python.org> writes
>Robin Becker wrote:
>> 
>> In article <mailman.1017873312.18135.python-list at python.org>, Guido van
>> Rossum <guido at python.org> writes
>> >Here's one final posting of PEP 285.  I've read (or at least skimmed)
>> >all feedback that was on our news server by 9:30am EST on April 2nd,
>> >plus anything mailed directly to me or to python-dev until now.
>> 
>> If this was a final post why are you now retiring to the dev list after
>> another 5 days.
>> 
>> It seems pretty pointless to appeal to the masses if you wish only to
>> have a resounding clap on the back and then when that doesn't happen to
>> retreat to your friendly worshipers.
>> 
>> O Gods when Rome burns where are you?
>> --
>> Robin Becker
>
>You just made my kill file.
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)
wow, I thought I was already there since the division rumble.
Interesting how the irreverent spirit of Monty Python seems singularly
lacking these days.
-- 
Robin Becker


From claird at starbase.neosoft.com  Tue Apr 16 18:05:09 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 16 Apr 2002 17:05:09 -0500
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <3CBB8ABB.57BAA189@engcorp.com> <mailman.1018989153.2077.python-list@python.org>
Message-ID: <B348A2683937CC65.3942DE81F07F4146.ED3D7BA18F22B5A1@lp.airnews.net>

In article <mailman.1018989153.2077.python-list at python.org>,
Cliff Wells  <logiplexsoftware at earthlink.net> wrote:
			.
			.
			.
>Especially when the money saved by developing in Python versus C could easily
>cover the cost of a faster PC (or two).  It seems so bizarre that there are
>people who would spend $10,000 (or more) in developer time to optimize an
>in-house program rather than simply spend $1000 on a new PC.
			.
			.
			.
'Haven't studied abnormal organizational psychology,
have you?  This is only a blip in the collective
irrationalities rampant around us.  Those $10,000 and
$1000 might be decoupled from the decision-makers
managing software development.  Moreover, someone in
the system might be aware that the total cost of the
PC should be amortized as around $10,000, once all
management overhead is properly accounted.

That's only the beginning.
-- 

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


From gerhard at bigfoot.de  Sat Apr  6 17:47:04 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 6 Apr 2002 22:47:04 GMT
Subject: 2.2.1c2 test_fcntl hangs
References: <lysn69qrp7.fsf@bock.chem.unc.edu> <m3r8lsfq1x.fsf@mira.informatik.hu-berlin.de> <ly1ydspj0u.fsf@bock.chem.unc.edu>
Message-ID: <slrnaauun4.tl.gerhard@lilith.my-fqdn.de>

Jochen K?pper wrote in comp.lang.python:
> On 06 Apr 2002 23:57:46 +0200 Martin v Loewis wrote:
> 
> Martin> "Jochen K?pper" <jochen at unc.edu> writes:
> 
>>> | Traceback (most recent call last):
>>> |   File "Lib/test/test_fcntl.py", line 38, in ?
>>> |     rv = fcntl.fcntl(f.fileno(), fcntl.F_SETLKW, lockdata)
>>> | KeyboardInterrupt
>>> `----
>>> 
>>> This is RedHat-7.0 with all updates + gcc-3.0.4. Python is
> 
> Martin> On a local file system, or through NFS? 
> 
> Through NFS, mounted with /default/ from a RedHat-7.1 machine...
> 
> Martin> Is there any chance that anybody else was holding locks to the
> Martin> test file at the same time?
> 
> No, I don't think anybody else was holding a lock on that file, as
> this is inside my $HOME hierarchy.

AFAIK some filesystems have problems when being exported via NFS.
(Older) reiserfs come to mind.

Gerhard


From graeme.matthew at unite.com.au  Tue Apr 16 07:01:00 2002
From: graeme.matthew at unite.com.au (Graeme Matthew)
Date: Tue, 16 Apr 2002 21:01:00 +1000
Subject: Classes, why ? so many problems ?
Message-ID: <3cbc0473@news.comindico.com.au>

hi all im really having problems with Python can anyone please tell me why I
cannot run this code, I keep on getting:
SyntaxError: invalid syntax
Traceback (most recent call last):
  File "C:\www\cgi-bin\test.py", line 9, in ?
    template.PrintOut()
AttributeError: Template instance has no attribute 'PrintOut'

class Template:

    def __init__(self):
        self.name = 'Template'

    def PrintOut(self):
        print self.name

Run from:

#! C:\\python22\\python.exe

import sys

sys.path.append('C:\\www\\classes')
import Template

template = Template.Template()
template.PrintOut()
print "done"






From logiplexsoftware at earthlink.net  Fri Apr 26 17:03:08 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 14:03:08 -0700
Subject: Removing dupes from a list
In-Reply-To: <20020426203946.GV21205@teich.Garten.DigitalProjects.com>
References: <20020426203946.GV21205@teich.Garten.DigitalProjects.com>
Message-ID: <20020426140308.658e2ae6.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 22:39:46 +0200
Alexander Skwar wrote:

> Hi!
> 
> What's the fastest way to make sure that a list only contains unique
> entries?
> 
> Currently, I get all the elements from the list and store them as keys
> in a dictionary.  I then use dict.keys() as the new value for the list,
> like so:
> 
> def de_dupe(liste):
>         d = {}
>         for element in liste:
>                 d[element] = None
>         return list(d)
> 
> But this will obviously only work, if all the elements of list are
> immutable.

Well, there's probably a faster way, but this is fairly simple:

def de_dup(liste):
    dups = filter(lambda i: liste.count(i) > 1, liste)
    return [i for i in liste if i not in dups]

>>> l = [[1,2,3],[1,2,3],[4,5],[1,2],[1,2,3], 1, 2, 3, 4, 2]
>>> de_dup(l)
[[4, 5], [1, 2], 1, 3, 4]
>>> 

I expect the list.count() will be a performance killer on large lists.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From steve at seasoned-software.com  Wed Apr 17 16:01:45 2002
From: steve at seasoned-software.com (steve at seasoned-software.com)
Date: Wed, 17 Apr 2002 13:01:45 -0700
Subject: What webhosting
In-Reply-To: <3CBDE7B4.3348.4EB76@localhost>
Message-ID: <3CBB5E860000221F@mta07.san.yahoo.com>

It'll never happen!!!!!!!!!!!   If you send unsolicited mail, the user may
get fed up!  Bear in mind that yours may be one of HUNDREDS that they get!
 The ISP HAS to get rid of your account eventually!  WHY????

Because if they DON'T, THEIR ISP may get rid of THEIR account(Which means
they can't support you anyway), because THEIR ISP may get rid of THEIRS!

Sometimes, companies allow this, and they lose their ISP.  What happens
THEN?  You lose your account ANYWAY!

This is why so many spammers use FREE Email(Of course, the addressees may
complain to their sponsors, which is why so many ads state that, or just
don't work.), and many are now having OTHERS do their dirty work. 

Steve


>-- Original Message --
>From: "A" <printers at sendme.cz>
>To: python-list at python.org, tutor at python.org,
>  activepython at listserv.activestate.com, python-help at python.org
>Subject: What webhosting
>Reply-to: printers at sendme.cz
>Date: Wed, 17 Apr 2002 21:23:00 +0200
>
>
>Hi,
>Does anybody know about a webhosting provider ( with Python and 
>Telnet access) that does not remove a user's account if the user 
>sends occasionally some unsolicited emails?
>Thanks for reply
>Ladislav
>
>
>_______________________________________________
>ActivePython mailing list
>ActivePython at listserv.ActiveState.com
>To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs






From paul at boddie.net  Wed Apr  3 06:31:29 2002
From: paul at boddie.net (Paul Boddie)
Date: 3 Apr 2002 03:31:29 -0800
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <mailman.1017769807.20362.python-list@python.org> <yivq8.14048$hi7.49995@news-server.bigpond.net.au>
Message-ID: <23891c90.0204030331.71d8da2c@posting.google.com>

"Trung Hoang" <thoa0025 at mail.usyd.edu.au> wrote in message news:<yivq8.14048$hi7.49995 at news-server.bigpond.net.au>...
> So would GadFly be the easiest to setup out of all the databases?

Provided you can find a version of Gadfly that works with the version
of Python you have, I would think that it would be much easier to
install and configure than almost any other relational database system
- especially since no special Python DB-API module needs to be built.
As someone else has probably mentioned, Gadfly is provided as part of
Zope, and Debian packages also exist, should you find it difficult to
locate a downloadable package.

Paul


From martin at strakt.com  Sat Apr  6 04:01:35 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Sat, 6 Apr 2002 11:01:35 +0200
Subject: Need help with distutils
Message-ID: <20020406090135.GA21020@strakt.com>

Hello. I need some help with using distutils. When I run
   python setup.py sdist
all my .cvsignore files are included, and I don't see a reason for them to
be included in the source dist, so I want to get rid of them
automagically. I tried using 'prune' in MANIFEST.in, but that didn't work,
so I tried 'recursive-exclude . .cvsignore' but that didn't seem to work
either (warning: no previously-included files matching '.cvsignore' found
under directory '.')

Am I attacking this the wrong way?

Any help appreciated.


Regards,
Martin Sj?gren

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From varial at gmx.de  Sat Apr  6 12:34:19 2002
From: varial at gmx.de (Robert Spielmann)
Date: 6 Apr 2002 17:34:19 GMT
Subject: checking if a number is int
Message-ID: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4>

Hi,

is there a way in python to check if a number is integer? My problem is 
the following: I need to know if sqrt(x) is an integer value or if there 
is a fractional part.

Thanks for help
Robert Spielmann


From mertz at gnosis.cx  Tue Apr  9 18:22:23 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 09 Apr 2002 18:22:23 -0400
Subject: Easter Eggs
Message-ID: <fm2s8kKkXAiR092yn@gnosis.cx>         

Who can tell me what's in there.  Someone on c.l.py recently let drop
the hint about finding the Zen of Python at the interactive prompt in
2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
magic command was.  Can someone remind me, please?

Also, is anything else hidden in the Python interpreter that issues
similar wisdom?  (I know...  I could read the source; but this is
probably quicker).

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From pinard at iro.umontreal.ca  Mon Apr  8 19:52:14 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 08 Apr 2002 19:52:14 -0400
Subject: Topy 0.1
Message-ID: <oq6631rbo1.fsf@titan.progiciels-bpi.ca>

Hi!  An initial snapshot of the in-the-works Topy is available as:

    http://www.iro.umontreal.ca/~pinard/topy/Topy.tar.gz

A few days ago, I told your that I would announce any result when available,
even if only dimly usable.  So, here is a little something!  I append
the beginning of the `README' file for this project.  Later announcements
might be more succinct, hoping anyway that they will convey better news! :-)


* README for Topy					allout -*- outline -*-

.. WARNING!

   This is merely a snapshot of an incomplete program.  The Scheme to Python
   converter is _NOT_ usable yet.  I'm only making it available for people
   who want to comment or help even at an early stage of this project.
   If `scmtopy' crashes on you, send me a small input showing the problem.

   Only Scheme integers get translated for now.  Nesting of `define' is
   naive, in that variables from intermediate levels may not be read nor
   modified from inner scopes; a related problem is that closures are not
   supported yet.  Tail recursion is not implemented, and this is required
   before `let' looping constructs generate correct code.  Please tell me
   if I forgot something else!

.. Introduction

. : What is Topy?

    Topy is meant to help its users at converting of programs or modules
    from a few other languages to Python (only Scheme is supported by now).
    Compromises favour legibility over strict semantics, so the resulting
    program is not likely to be equivalent to the original.

    The user is not freed from understanding and debugging the resulting
    program, and a serious clean up with a good editor is the first normal
    step after Topy conversion.  We believe that this approach yields
    more maintainable Python code in the long run, than stricter and less
    legible conversions.  Data types, statement constructs and overall
    concepts of the original language may be more or less foreign to Python,
    but wherever the source language and Python coincide, the generated
    Python might closely approximate the intent of the original sources.

    This is alpha status software: specifications are not frozen, things may
    change.  Report problems to Fran?ois Pinard at `pinard at iro.umontreal.ca'.
    For discussing specifications or making suggestions, please also copy
    the `python-list at python.org' mailing list, so to help brain-storming!

. : Installation

    Topy itself requires Python 2.0 or better to execute.  By default,
    it generates Python code that should be compatible with Python 1.5.2.

    The installation of Topy is fairly simple and standard.  Once into the
    directory of the Topy distribution, execute `python setup.py install'.
    To get an option reminder, do `python setup.py install --help'.
    See the Distutils documentation if you need more information about this.

    To check the installation, type `scmtopy -q < /dev/null > /dev/null'
    in a shell; you should get a mere "::: File `<stdin>' is empty.".

[... remainder deleted ...]

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




From borcis at geneva-link.ch  Tue Apr  9 07:05:58 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 09 Apr 2002 13:05:58 +0200
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>
Message-ID: <3CB2CB16.8080103@geneva-link.ch>

Oleg Broytmann wrote:

> 
>    I do not think the PEP *did* damage - it only revealed that our
> community has problems. I don't understand what problems.


Well, let's list PLDPs - Python-List Diagnostics Proposals - and then 
vote on them :)

A couple to get the ball rolling :

(a) Bool himself first viewed boolean algebra as the "laws of thought", 
a perception that some boolean lawyers verisimilarly share. And clearly 
a BDFL's position becomes more difficult when his work touches to what 
others percieve as their working laws of thought, whatever the realities.

(b) Laura made a lengthy point, overlooking the improvable while clear 
to the language designer's POV, that it very much boiled down to : "I've 
stacked my didactics of truth values in Python on the name of None, and 
the PEP makes me feel I've been telling lies". Which certainly amounts 
to something of a valid point, but it should be understandable to 
anybody that a designer can't be overly patient with what he can't help 
percieving as trying to imprison him in his own shadow while making it 
look at length as if he'd failed to understand... what he knows to be 
his own mind, and stateable in a few words.

(c) note that the ordinal of the PEP was a bad omen. Note how the three 
digits of

     285, actually mimic the three letter of
     RHS, a well known acronym for
     "Right Hand Side", itself not an inspiration for democrats.

Regards, Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")







From jeff at ccvcorp.com  Tue Apr  2 16:55:39 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 13:55:39 -0800
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de>
Message-ID: <MPG.1713c8b646c14e7098968e@news.nwlink.com>

In article <3ca9d7d1$1_2 at news.uni-bielefeld.de>, mario at bio58.uni-
bielefeld.de says...

> wxPython looked good, but it is so damn slow on my machine 
> (PPro 200) even on very small programs (hello - world).

Hm, wxPython seemed to perform adequately for me, on a (Win95) 
P200 with only 64MB RAM.  

(Yeah, it's slow, but so is everything *else*, too! ;) )

-- 

Jeff Shannon
Technician/Programmer
Credit International


From martin_byrne2 at yahoo.co.uk  Mon Apr 22 05:23:42 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 22 Apr 2002 02:23:42 -0700
Subject: writing a body using email module
Message-ID: <508cb8ae.0204220123.493c35be@posting.google.com>

Hi,

I've recently been looking at the relativly new email module. I've
been able to create the email with to,from,subject and attachments but
I cant for the life of me figure out how to write the body of the
mail.

Can anyone help me out with this?

Cheers,
B


From rjones at ekit-inc.com  Tue Apr 30 01:37:48 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 30 Apr 2002 15:37:48 +1000
Subject: Python vs. Perl,  which is better to learn?
In-Reply-To: <df30afd4.0204292111.31e5914a@posting.google.com>
References: <df30afd4.0204292111.31e5914a@posting.google.com>
Message-ID: <200204301537.48871.rjones@ekit-inc.com>

On Tue, 30 Apr 2002 15:11, David wrote:
> I have a lot of experience in ANSI C/C++, I want to learn a new
> language and which it can help me in the work.
> I just want to clear which(python or perl) is more comfort for me?

Perl will have more familiar things:
 . braces for block delimitation
 . distinction between pointers and values
 . semicolon statement endings

Python, on the other hand, throws these annoying language burdens out the 
window (in python, indentation delimits blocks, the parser is smart enough to 
know when a statement finishes and all variables are references to values). 
It does have a large amount of familarity for the C programmer - some of the 
syntax is similar, a large amount of the OS libraries look like their C 
equivalents, but with all the housekeeping done for you.

I understand that Python's easier to extend in C (having done some myself, I 
know it's really simple) and C++.

Python is also much, much easier to learn than Perl. Nuff said, really.


    Richard





From gcordova at hebmex.com  Tue Apr  9 15:01:15 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Tue, 9 Apr 2002 14:01:15 -0500 
Subject: What has PEP 285 done to us?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB94@mail.hebmex.com>

>
> Is there a way to peek into the latest python-dev messages? I could
> not find a link at http://www.python.org/psa/MailingLists.html
> Maybe if you subscribe to that mailing list (have to find their
> address first) you get too much email to handle so being able to just
> have a look at the lastest postings would be nice.
>

The address is posted at the front desk in the PSU's off




From NO_SPAM_jmpurser2 at attbi.com  Fri Apr  5 17:56:07 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Fri, 05 Apr 2002 22:56:07 GMT
Subject: Beginning Tkinter problem
References: <7Hpr8.233293$q2.21629@sccrnsc01>
Message-ID: <bYpr8.233323$q2.21339@sccrnsc01>

Okay, I think I got it.  I didn't notice the example wasn't typed in at an
interactive prompt, it was run from a script.  Still I'd like to know what
step to skip so I can run them from the interactive prompt.  Watching a GUI
interface get built up one line at a time from code is very cool!

And again answering my own question with dubious authority it appears that
if you leave the widget.mainloop() out it works fine.  Possibly because IDLE
is a Tkinter application itself and is already running a mainloop() and my
little app inherited it?  Now I'm really guessing so somebody clue me in!

Thanks,

John Purser

"John Purser" <NO_SPAM_jmpurser2 at attbi.com> wrote in message
news:7Hpr8.233293$q2.21629 at sccrnsc01...
> When working in IDLE(2.2) on Win2000 and attempting some Hello world class
> programs I get stuck after entering this code:
> >>> from Tkinter import Label
> >>> widget = Label(None, text='Hello GUI World!')
> >>> widget.pack()
> >>> widget.mainloop()
>
> The little window pops up, I click on the x in the corner and the window
> disappears.  But I never get my IDLE interactive prompt back.  Even going
> into the task manager and killing it there takes a while.
>
> So what am I doing wrong?
>
> Thanks,
>
> John Purser
>
>




From mlh at vier.idi.ntnu.no  Sun Apr 28 17:06:05 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 28 Apr 2002 21:06:05 +0000 (UTC)
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org>  <PNFy8.28407$8D3.758650@news1.tin.it>
Message-ID: <slrnacop1t.dcb.mlh@vier.idi.ntnu.no>

In article <PNFy8.28407$8D3.758650 at news1.tin.it>, Alex Martelli wrote:
>holger krekel wrote:
>        ...
>> ok. i found a nice workaround (if one dare call it that:-)
>> 
>> class dictlookup(type({})):

Why use type({}) when you can use dict?

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From donn at drizzle.com  Sat Apr 20 00:57:21 2002
From: donn at drizzle.com (Donn Cave)
Date: Sat, 20 Apr 2002 04:57:21 -0000
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <c76ff6fc.0204191806.406ec53b@posting.google.com>
Message-ID: <1019278636.737636@yasure>

Quoth sjmachin at lexicon.net (John Machin):
...
| Another practical perspective:
|
| When people mention excessive run-time to me, I ask them to rank the
| severity of their problem on this scale:
| 1. Job doesn't finish before they get back to their desk with a fresh
| mug of coffee.
| 2. Job has (exclusive access to database or excessive impact on
| on-line response time) and doesn't finish by sun-up.
| 3. Job doesn't finish before the weekly cold back-up is due.

How would you score a CGI program on a busy server in that scale?
Or for that matter any program that runs on a busy server.  And
I drink enough coffee, without refilling the cup every time I run
some software that makes me wait.

I'm really delighted to hear about performance improvements in file
object input in recent Python versions (in a message that would
have been in this thread if Tim Peters used a newsreader instead
posting to the list.)  Mostly because it's a good thing in itself,
assuming the numbers are there, but also because it shows that
someone must have felt that performance was important enough to
work pretty hard on it - that's a healthy sign.

By the way, I'll second Johannes Stiehler's recommendation of
MxTextTools.  Definitely appropriate for SGML parsing, and much
better than regexps for extensive parsing in my opinion - not
just in terms of speed, but I suspect a more powerful way to
describe text patterns than regexps.

	Donn Cave, donn at drizzle.com


From baas at ira.uka.de  Fri Apr  5 15:21:30 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Fri, 05 Apr 2002 22:21:30 +0200
Subject: ANN: Python Computer Graphics Kit v1.0beta1
Message-ID: <mailman.1018271583.5842.clpa-moderators@python.org>

I have just released the first beta version of the Python Computer
Graphics Kit, a collection of Python modules that contain the basic
types and functions required for creating 3D computer graphics images.

http://cgkit.sourceforge.net

The kit includes several new types such as vectors, matrices and
quaternions. It contains a binding for Pixar's RenderMan interface
which is a renderer independent API to communicate 3D data to
renderers which will finally produce a 2D image. There are already
several RenderMan compliant renderers freely available (they are
not part of the kit). The kit also includes some of the functionality
from the RenderMan Shading Language which enables you to create
procedural models or textures from within Python.

Even though the kit focuses on RenderMan, the new types or the
Shading Language functionality can also be used for other rendering
mechanism such as OpenGL or other renderers such as POV-Ray.

The kit should run on any platform where Python (and a C/C++ compiler)
is available. Windows users can download a binary version for Python
2.1 or Python2.2.

For more information, visit:

http://cgkit.sourceforge.net


- Matthias -





From aahz at pythoncraft.com  Mon Apr 29 11:03:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Apr 2002 11:03:59 -0400
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org> <mailman.1020076925.25849.python-list@python.org>
Message-ID: <aajncv$gfj$1@panix2.panix.com>

In article <mailman.1020076925.25849.python-list at python.org>,
Christian Tismer  <tismer at tismer.com> wrote:
>
>What we are talking about is a way to support PyZine, while at the same
>time providing a missing structure for Starship.

*Is* that what we're talking about?  Maybe you missed it on python-dev,
but there's been some recent discussion about resurrecting an active
Starship membership policy, so I'd bet that people other than myself
assumed that was your primary thrust.

If the goal is to push PyZine, I'd prefer to make that a separate issue
from Starship.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From starx at pacbell.net  Sat Apr  6 18:34:28 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 6 Apr 2002 23:34:28 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org> <Xns91E898777C949JamesLLugojcom@216.148.53.83>
Message-ID: <Xns91E89E99C5CBFRASXnewsDFE1@130.133.1.4>

James Logajan || Sat 06 Apr 2002 02:55:36p:

> Before I set Python aside as being unacceptably non-portable for a 
> scripting language, I _had_ tried to stop at 1.5.2. But I couldn't
> develop open source scripts in such a way that they were version
> independent without either a lot of version checking or freezing
> applications with a specific version of the interpreter (which defeats
> one of the points of open source IMHO).

Freeze a simple script that imports all of the libraries you use, and then 
execfiles the startups script of your app (or a script defined in a config 
file). This has worked pretty well for me, using McMillan's installer. The 
idea is you can either download the script set if you have or want to get 
the required version of python and all of the dependencies, or dowload it 
and an executible that runs it for you. Nothing significant is hidden in 
there, on the level of my program. You can see an example of this at 
www.rubydragon.com/sproj, though i haven't tested it very much yet.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From pyth at devel.trillke.net  Thu Apr 18 16:42:45 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 18 Apr 2002 22:42:45 +0200
Subject: PEP for an unrestricted __init__ ?!
Message-ID: <20020418204245.GA28515@devel.trillke>

I'd would like to know if anybody thinks it's
worthwile to start a formal PEP for an 
"unrestricted __init__". The guiding idea is to bring full
factory-pattern power to the __init__ constructor
of a class.

The factory pattern aims at decoupling object creation 
from the current declaration (data-layout) of an object. 
In c++, java or corba (or other interface/implementation-systems)
you often need factories because objects
are more or less defined by their method/data-layouts
and if you directly call the constructor you make
your calling code dependent on the current data-layout
but you only want it to be dependend on the 'interface'
- or in python's words - on the objects  'behaviour'.

In python each new instance of a class is created by
__init__  and this doesn't tie the calling code to 
a particular data layout. You don't need factories 
to be able to work with the behaviour of an object
that's how python works anyway.

But a separate factory object has an advantage
that python's __init__ does not offer:

for example a factory object can return the same instance 
for certain parameters.  e.g.  multiple calls to

    aFactory.create('123') 

might always return the same object. (This is often
used for persistent objects to transparently deserialize
an object from a database.)

It might be very interesting to extend the __init__ semantics
to offer the same flexibility. __init__ would then
be allowed to return something other[1] than the implicite 'self'.

for example you could write:

class a:
	def __init__(self, id):
		if id in cache.keys():
			return cache[id] # currently raises TypeError
		.... # really create the object
		cache[id]=self
	
i guess that by allowing an unrestricted 'return' statement 
we would have the same flexibility that (clumsy)
factories for other languanges provide. Note that caching
like promoted is only *one* possibility what factories 
can do[2]. 

After all factories and constructors are not orthogonal
concepts. On the contrary, factories often are just
overhead both for the user and the programmer of a class.
Especially if you have several connected objects (referencing
each other) factories often become a pain and somewhat tricky to do.

comments? 

regards,

	holger krekel


--
[1] For backward compatibility we might need to treat 
    'return None' as 'return self' because __init__ currently 
    seems to follow this rule. 

[2] I know that in this special case we might also
    use the "borg-pattern": 
    	self.__dict__ == cache[id].__dict__ 
    instead of the return statement.
    But then counterintuitively 

	    a('123')!=a('123')  is true

    which of course could be fixed but
    unneccesarily complicates matters...
    	





From jon+usenet at unequivocal.co.uk  Tue Apr 23 22:15:06 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 02:15:06 -0000
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com>
Message-ID: <slrnacc59a.tg8.jon+usenet@snowy.squish.net>

In article <u3ex8.94967$T%5.10614 at atlpnn01.usenetserver.com>, Steve Holden wrote:
> This error was caused by Fernando's following a bogus suggestion of mine.
> His original SQL syntax was correct

Uh, no it wasn't, he was using '%s' instead of %s. I'm confused now.


From fredrik at pythonware.com  Wed Apr 10 12:22:53 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Apr 2002 16:22:53 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018425802.24011.python-list@python.org>
Message-ID: <xFZs8.257$p56.69811@newsb.telia.net>

Arthur Siegel wrote:

> never has and never will speak for the core development
> team.

and who exactly are you speaking for?

</F>




From marklists at mceahern.com  Wed Apr  3 10:55:48 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 09:55:48 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <02a201c1db21$5c8ddfd0$0900a8c0@spiff>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEPKKIAA.marklists@mceahern.com>

[me]
> This is Python with the PEP:
> 
> falseValues = [{}, [], 0, None, False]
> for x in falseValues:
>     if not x:
>         print "%s is 'false'." % x

[Fredrik Lundh]
> can you explain how your second example covers
> everything that's in the PEP?

It doesn't cover everything in the PEP.

Here's how I think of the PEP:

With this PEP, we get to have our cake and eat it too:

	1.  There is one obviously correct way to return a truth value.
	2.  Polymorphous perversity with respect to truth still works.

What's everybody complaining about?

Cheers,

// mark




From aahz at pythoncraft.com  Wed Apr 10 16:03:38 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 16:03:38 -0400
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <RYTs8.2219$b62.57325@news1.tin.it> <23891c90.0204100552.6b822307@posting.google.com> <C%Xs8.3139$b62.77265@news1.tin.it>
Message-ID: <a925qq$ne4$1@panix1.panix.com>

In article <C%Xs8.3139$b62.77265 at news1.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>With the previous iterator protocol, "for item in somedict:" COULD
>NOT "work smoothly and seamlessly".  Nor, at all.  Think about it.

Oh, come on, Alex; with your twisted imagination, I'm sure you can think
of ways in which a dictionary could be forced to work with the old
iterator protocol -- it would just be very, very messy.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From mwh at python.net  Sat Apr 13 05:40:20 2002
From: mwh at python.net (Michael Hudson)
Date: Sat, 13 Apr 2002 09:40:20 GMT
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018632996.10456.python-list@python.org> <m3d6x4fz3j.fsf@mira.informatik.hu-berlin.de>
Message-ID: <lkn0w8hqzm.fsf@pc150.maths.bris.ac.uk>

martin at v.loewis.de (Martin v. Loewis) writes:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
> 
> > ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/Modules/mathmodule.so: symbol PyExc_OverflowError: referenced symbol not found
> > 
> > Could someone suggest how to proceed, please?
> 
> Can you please report the command line option that was used to link
> the python binary? (delete it, and re-invoke make) PyExc_OverflowError
> ought to be exported from the Python binary, and there ought to be a
> gcc command line option to achieve that.

But this is really strange; mathmodule.so must have imported at some
point, or setup.py would have deleted it...

Cheers,
M.

-- 
  Remember - if all you have is an axe, every problem looks 
  like hours of fun.                                        -- Frossie
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From aleax at aleax.it  Fri Apr  5 09:29:26 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 14:29:26 GMT
Subject: Python-list digest, Vol 1 #10324 - 13 msgs
References: <E16tGLp-0007ou-00@mail.python.org> <mailman.1018014793.1604.python-list@python.org>
Message-ID: <axir8.52642$pT1.1485864@news1.tin.it>

Jonathan Claggett wrote:
        ...
> Actually, I really wasn't thinking in terms of functional programming (I
> can't Lisp my way out of a paper bag :-). My immediate motive was only
> skin deep in that I'd like to replace the (unique) lambda syntax in my
> code with the (standard) def syntax.

You can do that now!  def is a statement and you can put it wherever you
need.  Just choose a good name, as you would for some complicated
sub-expression in order to clarify things.  For example, instead of

Button(buttonsFrame, text='Squirk', command=lambda: self.squirkIt(23, 45))

you can code:

def specificSquirker(): self.squirkIt(23, 45)
Button(buttonsFrame, text='Squirk', command=specificSquirker)


It's generally not hard to come up with a name which helps readability.
I think un-lambdaing your code in this way is a good thing...


Alex




From teddrain at earthlink.net  Sun Apr  7 17:15:24 2002
From: teddrain at earthlink.net (Ted Drain)
Date: Sun, 07 Apr 2002 21:15:24 GMT
Subject: Lex and Yacc for Python
References: <slrnab19hf.23u.Alex@Tron.ddeus.lan>
Message-ID: <MF2s8.23647$nt1.1918243@newsread2.prod.itd.earthlink.net>

I've been using ply with good results for some time now.

http://systems.cs.uchicago.edu/ply/

Ted

"Oleksandr Kozachuk" <Alex at DDeus.de> wrote in message
news:slrnab19hf.23u.Alex at Tron.ddeus.lan...
> Hi all!
>
> I need something, comparable to Yacc/Lex. Or something for
> Context-Free-Grammar. Can any one recommend me anything?
>
>
> Thanks,
> Alex.




From frr at easyjob.net  Mon Apr 29 05:58:56 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Mon, 29 Apr 2002 11:58:56 +0200
Subject: CGI and client's IP?
Message-ID: <1a6qcuc92l8ac0tcegqj95fqemdctda21v@4ax.com>

Can I get the client's IP froma a cgi script? O:-)

TIA







-----------------------
Fernando Rodriguez


From just at xs4all.nl  Sat Apr  6 10:24:00 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sat, 06 Apr 2002 17:24:00 +0200
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <just-89BA7A.17235906042002@news1.xs4all.nl>

In article <uDDr8.66496$S52.2342405 at news2.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> Linus Thorvalds seems to have done pretty well with picking TWO middle 
> grounds -- two parallel tracks for Linux, "stable" and "experimental".

It seems 2.1.x and 2.2.x currently serve as these two tracks, 
respectively. I think it's a very good idea to think about how to 
formalize this for future releases.

Just


From DavidA at ActiveState.com  Fri Apr 26 20:10:22 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Fri, 26 Apr 2002 17:10:22 -0700
Subject: Python.NET?  Jython.NET?  Ni!
References: <RGly8.3497$Z6.36288@nnrp1.uunet.ca>
Message-ID: <3CC9EC6E.1030108@ActiveState.com>

Bruce Dodson wrote:

>Is ActiveState's Python for .NET dead?  I can find no reference to it on the
>ActiveState page, other than an old research build for .NET Beta 2.
>
>Seems to me the easiest place to start for a .NET version of Python would be
>with Jython.  I haven't looked at the Jython source code, but can someone
>who has done so comment?
>
"Dead" sounds so harsh, in this age of Buffy.  The truth is that no one 
is working on it from ActiveState, and as far as I know Mark isn't 
either.  There are a few people who are interested, however, so there's 
always hope.

--david ascher






From brian at sweetapp.com  Sat Apr  6 12:50:54 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 6 Apr 2002 09:50:54 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
In-Reply-To: <Xns91E7DF1A19529JamesLLugojcom@216.148.53.84>
Message-ID: <002c01c1dd93$99bac610$445d4540@Dell2>

James wrote:

> _I_ didn't have any problem with Unicode. Python did. Python caused
> _forward_ incompatibilities, not just backward ones.

Unicode was too important not to add to the language, even if it did
cause a few headaches. But the addition was done in a manner that
favored compatibility over optimal semantics.
 
> There are other compatibility issues that gave me some problems.

What is the total amount of time that you have spent resolving
compatibility problems? What language are you going to use instead of
Python?
 
> I condemn the growing number of versions that are proliferating. More
> versions invite a large probability of forward and backward
compability
> problems.

Please list some languages that are very forward and backwards
compatible between releases.

I'm just trying to get a handle on the exact rational for your decision
not to use Python anymore.

Cheers,
Brian





From ark at research.att.com  Tue Apr  9 15:23:18 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 9 Apr 2002 19:23:18 GMT
Subject: binding a reference to a variable
Message-ID: <yu994rikfzh5.fsf@europa.research.att.com>

I would like to be able to write an expression that yields an
object that I can subsequently use to rebind a name that I mention
only in that expression.  How do I do it?

My motivation is to be able to write something that is analogous
to call-by-reference in C++.

This question, as phrased, is probably a little hard to understand;
here are some examples that might help.

Here's a function definition:

        def f(v):
                global x
                x = v

Now, f is an object that I can use to rebind the name x.  For
example, if I execute f(42), that sets x to 42.

However, this example doesn't do what I want because the
``expression'' in this example would have to be f, not the definition
of f (which isn't an expression).  The name x is not mentioned
anywhere in the expression.  Therefore, each time I want to affect a
different variable this way, I have to define a new function.

What I would really like is to be able to define a function called,
say, `set' such that set(x, 42) has the same effect as `x = 42'.

I can't use a lambda expression, because they're not allowed to
contain assignments.

So far, the closest I've been able to come is this:

        def set(var, val):
                var[:] = [val]
                
It requires that the variable in question be primed by giving it
a list as its value, and detects failure to do so:

        set(x, 42)              # error -- x is not a list
        x = []
        set(x, 42)              # x is now [42]

Is it possible to do better (i.e., to solve this problem more succinctly)?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From lumber at jack.com  Mon Apr 29 12:39:55 2002
From: lumber at jack.com (Lumberjack)
Date: 29 Apr 2002 16:39:55 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <Xns91FF62618790Flumberjack@216.148.53.85>

"Raymond Hettinger" <python at rcn.com> wrote:
> Abstract

Deleted. "Abstract" is right. Reality doesn't enter into it.

>     There are separate motivations for each function.  One motivation
>     in common is that the number of total built-ins should be as small
>     as possible to limit the amount of "core language" one needs to
>     know in order to read code that is not module specific.

Huh? The "core" language wouldn't be measurably shrunk by this proposal 
fella. Your rationale is built on air. Show us some genuine _advantage_ 
that is worth the huge _cost_ in time and effort caused by your ill-
considered proposal. How about you and other PEP writers take some 
responsibilities for your fantasies for once? You up to helping thousands 
rewrite their code? I thought not.

>     Creating a separate module
>     allows room for a large variety of functionals (e.g. fold, head,
>     tail, take, drop, etc.) to be introduced without cluttering the
>     built-in namespace. 

Run away! Run away! You know, you can _still_ create your precious 
functional style programming module _without_ removing or deprecating the 
existing so-called functionals. But no, you'd rather break code and make 
existing instructional material obsolete and require existing Python 
programmers to _unlearn_ some existing patterns. But it is for a good cause 
- if only you could articulate it!

>     Pow() and divmod() are more related to math module functions than
>     any of the built-in functions.  In general, they are rarely used. 
>     Because the ** operator is available, the pow() function is rarely
>     called directly.

Proof by assertion. Wonderful! Is the evidence hiding in the period at the 
end of your last sentence? Cause I didn't see it anywhere else.

>     Input() is easily replaced by eval(raw_input()).  It has few
>     legitimate uses and presents a security risk in most programs where
>     it is used. Though the risks are well documented, experience on
>     python-tutor and help at python.org indicates that new users
>     frequently use input() when raw_input() was intended and that they
>     are unaware of the embedded eval().

Why not just shoot the beggars who use the function? The would be in line 
with the contempt you seem to have for others.

> Key Issue
> 
>     All of these functions have been in-place since the beginning, are
>     used in mountains of code, and appear in every Python book.  The
>     situation would appear to be immovable.

Then why the hell are you proposing it? You think the world revolves around 
your notion of "rightness"?

> Implementation and Deprecation Timetable
> 
>     The module additions would go into Python 2.3.  All six functions
>     would be left intact and raise deprecation warnings in Python 2.3
>     and 2.4. The six functions would be removed in Python 2.5 (having
>     allowed two versions and a full year for code updates).

A FULL year! The mind boggles! Civilizations can grow, live, and die in 
that time! Maybe you should make it three weeks! Of course, a pregnancy 
lasts 9 months and girls begin menustruating around 12 years, and people 
typically aren't considered duly educated till their early 20's, still, we 
shouldn't let human time scales dictate how fast we can roll out new 
computer language changes!

I believe the organization you belong to has this chant:
REMEMBER: Change is progress! Progress is change! Either you are for 
progress or for congress! All right thinking embrace change! Evil people 
hate change!

> Copyright
> 
>     This document has been placed in the public domain.

Take it back! I want no part of it.


From joost_jacob at hotmail.com  Mon Apr 29 10:19:33 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 29 Apr 2002 07:19:33 -0700
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com>
Message-ID: <13285ea2.0204290619.3563152f@posting.google.com>

Thank you for your reply [Steve Holden] and the weblinks !
Also thanks to the other people who reacted.

A nicely optimized fortune-webserver will have to wait until i have
more solid ground under my feet when doing web programming.  I
noticed you wrote a book about webprogramming and i am considering
buying it $-) 

I did like to see the ideas for improvements.  The idea by [Michael
Gilfix] about reserving the same amount of space for each quote and
then using seek() is elegant.  At the moment i have a 3.2MB file
with 19930 quotes and i am thinking about an easy way to add/remove
quotes and maybe using the system's clipboard to be able to add
quotes from my own collection to usenet postings.

Here is a slightly improved version of fortune.py.  If there is no
allfortunes file it will use the modulequotes variable, it is easy
to add your favorite quotes to the modulequotes variable.  This
version keeps the quotes in memory when calling quote() or
getQuote() repeatedly from a program that loaded this module:


-----------------------------  fortune.py  ------------------------
import random, os, StringIO

modulequotes = """At the end of every pot of gold there's a rainbow
%
To a three year old with a hammer, everything looks like a nail
%
"""

quoteslist = []

def getQuote():
    """
    Return a quote <string> from file allfortunes,
    assume the file allfortunes is located in the same
    directory as this module."""
    try:
        datafile = os.path.join(
          os.path.dirname(__import__(__name__).__file__),
          'allfortunes' )
        quotesfile = open(datafile)
    except IOError:
        quotesfile = StringIO.StringIO(modulequotes)
    quote = ''
    if not quoteslist:
        for s in quotesfile.readlines():
            if s[0] == '%':
                if quote:
                    quoteslist.append(quote)
                    quote = ''
            else: quote = quote + s
        if quote: quoteslist.append(quote)
    return random.choice(quoteslist)

def count(): return len(quoteslist)

def quote(): print getQuote(),
-------------------------------------------------------------------


Joost Jacob
www.liacs.nl/~jjacob/

YAY i can add a quote now even when using Windows:
The outraged husband discovered his wife in bed with another man.
        "What is the meaning of this?" he demanded.  "Who is this fellow?"
        "That seems like a fair question," said the wife, rolling over.
"What IS your name?"

<omg-where-is-the-option-to-prevent-offensive-quotes-'ly-yrs>


From jwbaxter at spamcop.net  Fri Apr 26 11:05:51 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Fri, 26 Apr 2002 08:05:51 -0700
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com> <gb4hcu4b55gq99blg08eudh14pm8ivju0k@4ax.com>
Message-ID: <jwbaxter-DA711D.08055126042002@corp.supernews.com>

In article <gb4hcu4b55gq99blg08eudh14pm8ivju0k at 4ax.com>,
 Chris Spencer <clspence at one.net> wrote:

> 	I want to thank everyone who responded to this.  I read the floating
> point tutorial and now understand the issues.

And now...just a wee reminder that it is possible for the chip makers to 
build errors into their floating point processors.  Even large, 
well-known chip makers whose names start with "Intel."  

And it's possible for software writers to write bugs into their floating 
point software routines.  Even large companies deeply involved in the 
IEEE standards process and producing an excellent implementation of that 
standard...including one whose name starts with "Apple" (whose code in 
rare cases messed up the low bit of the mantissa).

So...while the issue you raised isn't the result of an error, that 
doesn't mean that no strange result is ever an error.  Keep questioning 
oddities!  Understanding flows therefrom.

  --John


From hellen at claggetts.net  Thu Apr  4 11:38:12 2002
From: hellen at claggetts.net (Hellen Claggett)
Date: Thu, 04 Apr 2002 11:38:12 -0500
Subject: Anonymous functions (and classes)
Message-ID: <1017938292.1490.81.camel@localhost.localdomain>

Hello,

Here is a relativly simple feature that I would like to use while
programming in Python. First, I did make a (brief) effort to determine
if this specific request had been made and found nothing obvious in the
PEPs or in this year's python-list archives.

I would like to be able to define anonymous functions and classes by
omitting the function or class name normally found after the def or
class keyword. For example, the following two function definitions would
be equivalent:

def comfy(color='red'): print color, "comfy chair"
comfy = def(color='red'): print color, "comfy chair"

similarly, the following two class definitions would be equivalent

class Comfy(object): pass
Comfy = class(object): pass


My two Pythonic* reasons for wanting this are:

- Greater control over your namespace allowing the programmer to define
a new function/class that is _not_ in the current namespace (Design
Principle 19 :-)

- A lambda replacement which is more consistent with all other function
definitions. (Design Principles 9 & 13)


Any comments? I'm particually curious to know how painful this change
would be to python's innards.

Thanks for your time,

Jonathan

*See: http://python.org/dev/culture.html











From gerhard at bigfoot.de  Wed Apr 10 01:00:25 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 10 Apr 2002 05:00:25 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83>
Message-ID: <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de>

Lumberjack wrote in comp.lang.python:
> Tim Peters <tim.one at comcast.net> wrote:
>> PEP 285 taught us that people don't agree, and that some people
>> disagree in disagreeable ways.  Frankly, that wasn't news,
> 
> [drivel attacking Tim and Guido]

Why would anybody care if you aren't man/woman enough to post with
your real name?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From donn at u.washington.edu  Tue Apr 23 12:47:03 2002
From: donn at u.washington.edu (Donn Cave)
Date: 23 Apr 2002 16:47:03 GMT
Subject: non-blocking sockets
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf> <mailman.1019576266.25727.python-list@python.org>
Message-ID: <aa4367$6jg$2@nntp6.u.washington.edu>

Quoth Michael Gilfix <mgilfix at eecs.tufts.edu>:
|   You need to use the select module to poll the socket. The select
| module allows you to specify how long you want your timeout (in this
| case, you probably want a timeout of 0, which is simply a poll) and
| the module tells you which sockets are waiting for reading/writing.
|
|   import select
|   read_sock = <my readable socket>
|   w_ready, r_ready, e_ready = select.select ([ ], [ read_sock ], [ ], 0)
|
|   if read_sock in r_ready:
|     # Process socket
|
|   ... will probably do what you want.

Yes - and then, we might add, you don't need to make the
socket non-blocking.

	Donn Cave, donn at u.washington.edu


From tim.one at comcast.net  Fri Apr 26 18:30:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 18:30:11 -0400
Subject: [Crew] PSA revival - kind of?
In-Reply-To: <3CC9CBFE.527E8C28@python.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEFLPBAA.tim.one@comcast.net>

[Guido]
> You can send your donation to the PSF.

No, you cannot.  Not yet.  The PSF needs a favorable advance ruling from the
Internal Revenue Service before it can accept any donations, and we haven't
yet completed the mountains of paperwork needed for that.  Before then, and
even despite appearances <wink>, nothing Guido says is to be read as a
solicitation for donations.  Guido is not soliciting there!  I believe he
wrote that line from a few months in the future, after the IRS ruled in the
PSF's favor.  Of course, if you want to travel to the future and donate,
that would be way cool.  But I'm not soliciting, and neither is Guido.  I
would like a trip in your time machine, though.

> Soon this will be tax-deductible (for US citizens).

Well, so long as Guido is writing from the future and just reporting facts,
I have no objection.

>  http://www.python.org/psf/  Membership is free but only on invitation;

Note that invitations aren't offered by Guido, so bribing him won't help.
Invitations are offered only upon majority vote of the existing membership;
and the BDFL only has one vote (so long as the membership doesn't vote him
out).

> but anyone can donate, much in the PSA spirit.

Another statement from the future.  Unlike the PSA, the PSF actually has a
mission in mind:

    http://www.python.org/psf/mission.html

More details will be forthcoming in the PSF's IRS Form 1023 filing.

> We can talk about adding a starship account in the deal; Py
> subscriptions are not in the cards though.

Dates with Uncle Timmy are, although somehow that didn't make it into the
mission statement.





From elimli at yahoo.com  Wed Apr 24 05:52:26 2002
From: elimli at yahoo.com (eli li)
Date: Wed, 24 Apr 2002 11:52:26 +0200
Subject: urllib + headers
Message-ID: <aa5rnk$68k$1@news.iucc.ac.il>

I open url with urllib.urlopen, open method is POST. How is it possible in
request to specify headers like 'Referer', 'Accept-Language', etc?




From grante at visi.com  Sun Apr 21 21:30:47 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 22 Apr 2002 01:30:47 GMT
Subject: newbie dictionary question
References: <TiHw8.99936$CH1.86776@sccrnsc02> <3CC34519.E6C1063A@gmx.de> <g1Jw8.89176$G72.64386@sccrnsc01>
Message-ID: <slrnac6pvp.9c3.grante@tuxtop.visi.com>

In article <g1Jw8.89176$G72.64386 at sccrnsc01>, James Chapman wrote:
> Hi,
> 
> I'd preferably like to use something like this but my problem
> is precisely this. I have for example to dictionaries as
> follows:
> 
> a= {4.95:100,4.97:90,4.99:95}
> b={4.95:100,4.96:99}
> 
> and I would wonder if there is an elegent way to do something like:
> 
> c = a+b
> c={4.95:200,4.96:99,4.97:90,4.99:95}

c={}
c.update(a)
c.update(b)

Though I can pretty much guarantee that using floats as
dictionary keys isn't going to do what you want at some point.

-- 
Grant Edwards                   grante             Yow!  I'd like some JUNK
                                  at               FOOD... and then I want to
                               visi.com            be ALONE --


From bh at intevation.de  Tue Apr 23 05:32:14 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 23 Apr 2002 11:32:14 +0200
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <a9l0bo$h1j$1@panix1.panix.com> <6qd6wx72gh.fsf@abnoba.intevation.de> <3CC4BBF5.4CBD0276@cosc.canterbury.ac.nz>
Message-ID: <6qvgaipxpd.fsf@abnoba.intevation.de>

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

> Bernhard Herzog wrote:
> > 
> > Why would you want to implement __iadd__ for immutable objects?
> 
> So that, e.g. i += 1 works when i is an integer.

You don't need __iadd__ for that and indeed the int implementation in
2.1.3 does not have the __iadd__ method (more precisely, its C-level
equivalent). The interpreter automatically falls back to using __add__
so that i += 1 becomes equivalent to i = i + 1. The only difference is
that in e.g. x[i] += 1, i is only be evaluated once, as I understand it
anyway.

An immutable object implementing __iadd__ to behave different than
__add__ is probably asking for trouble, although in some rare
circumstances it might offer some optimizations not possible with
__add__ but I can't think of one at the moment.


   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From jb at cascade-sys.com  Tue Apr 30 13:29:45 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 10:29:45 -0700
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com>
Message-ID: <3CCED489.8F530C87@cascade-sys.com>

David wrote:

> I have a lot of experience in ANSI C/C++, I want to learn a new
> language and which it can help me in the work.
> I just want to clear which(python or perl) is more comfort for me?
> --
> http://mail.python.org/mailman/listinfo/python-list

Even without knowing your background I think one can confidently predict
that Python will be lots more "comfortable" for you, at least if you are
at all normal.

Both languages are of comparable power and capability.  Perl has been
around longer so it has attracted a wider following, though Python is
gradually catching up.

Python is a carefully designed language with a few consistent rules.
It's a first-class programming language.

Perl is a hodge-podge of random features and syntax thrown together.
It's more like a super-fancy C-Shell scripting language.

I think it's a known fact that the amount of goofy punctuation in Perl
will lower your productivity by 20-30% and shorten your life by several
years.  ;o)

I mean really, a variable is one of the most common "objects" in any
programming language.  What the frig is with having to prefix every one
with a dollar sign (or some other particle of punctuation)?  What the
frig is with NOT being able to define the arguments to your
functions/subroutines?

Perl appeals to gnomes who like programming tricks, secret handshakes
and language puns.  Python tends to attract people who value clarity of
expression   You will continually be perplexed, surprised, frustrated
and amazed (in a bad way) with Perl.  You will be continually pleased,
encouraged, rewarded and amazed (in a good way) with Python.

Of course if you asked a Perl group they'd say you'd be better off with
Perl.  But they'd be lying.  ;o)

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From david at dataovation.com  Sat Apr  6 03:15:33 2002
From: david at dataovation.com (David McInnis)
Date: Sat, 6 Apr 2002 00:15:33 -0800
Subject: Modify Word Doc with Python
Message-ID: <009601c1dd43$398e9f00$9865fea9@bigdaddy>

I need to alter the document properties of Word DOCS that our users
upload to our Web site.  Does anyone know how to do this?  (I am python
installed on a Linux box).

David





From nospam at unknown.com  Thu Apr 18 21:11:22 2002
From: nospam at unknown.com (Gregg)
Date: Fri, 19 Apr 2002 01:11:22 GMT
Subject: Python n00b here...
Message-ID: <3cbf6d6d.11416265@news.telus.net>

Hi,

I just downloaded the latest version of Python and Snack toolkit
widget. I chose Python over many others, including VB, Java and C++
because it seem to fit what I was looking for in power/user
friendlyness.

I'm into high-end analog engineering and also a musician, so I learn
quick ;-)

What I would like to do is create a full featured freeware DSP for
recording musicians that can use most any DirectX or VST processing
plugin.

What I would like to know is:

1) Has this been done before in Python and if so, where can I get it?

2) There are no such projects in the works at Sourceforge. If there is
an active group doing such, where can I find them?

3) This one's probably the most important - HOW THE HECK DO I DO
THIS???

Much thanks in advance :-)

--
Gregg


From aahz at pythoncraft.com  Mon Apr 22 12:20:15 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 12:20:15 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <304d20df.0204220628.6aa88bc2@posting.google.com>
Message-ID: <aa1d7v$a2s$1@panix2.panix.com>

In article <304d20df.0204220628.6aa88bc2 at posting.google.com>,
Allan Crooks <googlegroups at sixtyten.org> wrote:
>aahz at pythoncraft.com (Aahz) wrote in message news:<a9ut8d$dql$1 at panix3.panix.com>...
>> In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
>> Allan Crooks <googlegroups at sixtyten.org> wrote:
>>>
>>>Another thought I've had is doing it by specifying, say for module
>>>'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
>>>what versions of Python it is running on, and then compiling all
>>>modules it can deal with, and then importing all objects into it's
>>>namespace.
>> 
>> Yup.  But do it with a package.
>
>Umm... pardon my ignorance, but what exactly is a package in terms of
>Python? I assume it's just a module that does some form of importing
>of sub-modules (perhaps it uses "from module import *")?

See section 6.4 of the Python tutorial.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From bokr at oz.net  Tue Apr 16 19:58:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 23:58:57 GMT
Subject: Python 2.2 unwise.exe
References: <mailman.1018989634.5599.python-list@python.org>
Message-ID: <a9ids1$j54$0@216.39.172.122>

On Tue, 16 Apr 2002 16:38:52 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[Tim]
>> On a clean Win2K Pro box, I'm finally able to provoke this (or at
>> least one of its causes) reproducibly, via:
>>
>> 1. Boot the box.
>>
>> 2. Install 2.2.1.
>>
>> 3. Uninstall 2.2.1.
>>
>> 4. Install 2.2.1 again.  Note that \Python22\UNWISE.EXE is there.
>>
>> 5. Reboot the box (just logging off and on again isn't enough).
>>
>> 6. Now \Python22\UNWISE.EXE no longer exists.
>
>For anyone playing along, this particular way of provoking the problem
>appears stoppable via inserting step 2a:
>
>2a: Delete this line from near the end of \Python22\INSTALL.LOG:
>
>    Delete in-use files: On
>
That's good info, but you're not suggesting that as a solution, are you?
ISTM that would be interfering with a normal and legitimate function of
installers/uninstallers.

IOW, I'm guessing uninstall at 3 uses that line to decide whether to
talk to windows about files it can't delete because they're locked as in use.

Uninstallers talk to windows and say "delete this and that when you get chance
while they're not locked." That "while" occurs by design at a point in the boot
process, before those files get a chance to be locked again for some reason.
Likewise, an installer may call for deferred replacement of a file when the
file can't be deleted.

I'm guessing it takes a reboot or some special program operation to flush pending
deletes. Thus the uninstall at 3 would leave a normal pending delete, but apparently
there is no special programming in the install at 4 to deal with it. So windows remembers
until the next boot. However the install at 4 seems to have replaced unwise.exe (or
it scheduled a boot time replace, and windows got the now two pending operations out of order?)

In any case, after 4 there seems to be either a single pending delete or an out of order
pair of delete unwise.exe and delete/replace unwise.exe at 5, and the boot triggers it,
hence 6.

At least that's my theory.
Why do you think windows always wants you to reboot after installing, after all?
Seems like better file transaction stuff could have been designed into he system ;-/

Regards,
Bengt Richter


From eppstein at ics.uci.edu  Wed Apr  3 21:11:54 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 03 Apr 2002 18:11:54 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com>
Message-ID: <eppstein-AB22BE.18115403042002@news.service.uci.edu>

In article <3CABB156.1010401 at nospam.attbi.com>,
 djw <dwelch91 at nospam.attbi.com> wrote:

> Pardon my ignorance (I've only been working with Python for a short 
> while), but I don't understand why an idiom like:
> 
> if x == True:
> 
> would be/should be an error (or warning initially).

Because it will give the wrong result when x is a list, or other type of 
object that is not a bool but can be used as a truth value.

> I see C/C++ code 
> like this all the time:
> 
> bool x;
> x = Foo(Bar);
> if (x==true)
> {...

C/C++ forces its variables to have declared types, so you can be sure 
that x is a bool -- if it is not true, it must be false.
In Python, you have no guarantees about the type of x, and should be 
prepared to handle any type of value that might make sense in the 
context.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From dale at riverhall.NOTHANKS.co.uk  Mon Apr 22 12:05:19 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Mon, 22 Apr 2002 17:05:19 +0100
Subject: ANN: PyUK (The UK Python User Group) 
Message-ID: <h4d8cu84qjdhhkk4vv9meahtk1hi4vo731@4ax.com>

PyUK (The UK Python User Group) 
ACCU Python SIG

INAUGURAL MEETING

2002 April 5

Approximately 30 Python users met during the ACCU Spring Conference to
launch the Python UK user group.

It was agreed that the temporary name for the group would be "PyUK".

THANKS

There was a show of thanks to the ACCU for offering the Python stream
on their conference.

INFRASTRUCTURE

It was further agreed that PyUK would operate as a SIG (special
interest group) within the ACCU (http://www.accu.org). The ACCU
already has the organisational infrastructure that the Python groups
needs and the expansion of the ACCU's activities to include Python was
seen as an all round good thing.

Python users would have to join the ACCU but the costs are modest and
no one saw this as a barrier.
(http://www.accu.org/membership/public/index.htm) 

Information on the constitution of the ACCU can be found here:
http://accu.org/membership/public/constit.htm 

ACCU COMMITTEE REPRESENTATIVE

Dr. Tim Couper volunteered to be a co-opted member on the ACCU
committee.

PUBLICATIONS

Python users would contribute to and benefit from the ACCU's two
publications: "C-Vu" and "Overload". Paul Brian volunteered to
coordinate submissions for publication.

EVENTS

Dr. Tim Couper (TAC Software) volunteered to organise PyUK events.

WEB SITE

While there may be the possibility of some Web space on the ACCU site,
there was also seen to be a need for some dedicated space. Chris
Withers (New Information Paradigms Ltd) offered some Wiki/Zope
hosting.

TRAINING

There was some group discussion about the possibility of offering
Python training. Paul Brian volunteered to coordinate trainers and
Andy Robinson suggested a possible venue. Dale Strickland-Clark
(Riverhall Systems Ltd), John Pinner (Clockwork Software Systems),
Duncan Booth (RCP Consultants Ltd) and others could prepare courses,
if sufficient interest was generated. The courses could be advertised
on the Web site, in ACCU publications and elsewhere.




--
Dale Strickland-Clark
Riverhall Systems Ltd



From tegbert at abl.med.utah.edu  Tue Apr  2 14:46:47 2002
From: tegbert at abl.med.utah.edu (tegbert)
Date: Tue, 2 Apr 2002 12:46:47 -0700
Subject: Problem Calling Python callbacks from C
In-Reply-To: <lk7knpx64t.fsf@elios.maths.bris.ac.uk>
References: <8351bb33.0204020854.272eaa7b@posting.google.com> <m3d6xigg6a.fsf@mira.informatik.hu-berlin.de> <Pine.LNX.4.10.10204021028590.32630-100000@abl.med.utah.edu> <lk7knpx64t.fsf@elios.maths.bris.ac.uk>
Message-ID: <Pine.LNX.4.10.10204021246180.2860-100000@abl.med.utah.edu>

Oops.  Looks like it works fine and my error was elsewhere. --Tim

> On Tue, 2 Apr 2002, Michael Hudson wrote:
>
> Tim Egbert <tegbert at abl.med.utah.edu> writes:
> 
> > I'm able to call a Python callback _function_ from C code by following the      
> > information in "Extending and Embedding the Python Interpreter," section
> > 1.6, but I can't get the callback to work if I'm calling a Python object
> > _method_.  Does anyone know how to do that?
> > 
> > In other words, it works if I'm using a Python function such as:
> >    
> >     def callback_func(i):
> >         print i
> > 
> > and I set "callback_func" as the callback but not if the callback is an
> > object method such as:
> > 
> >     class callbackClass:
> >         def callback_method(self,i):
> >             print i
> >     cbc=callbackClass()
> > 
> > and I set the callback as "cbc.callback_method".
> > 
> > I'm using the "PyEval_CallObject(callback,arglist)" function in my C code
> > to make the callback.
> 
> That ought to work.
> 
> > I assume the problem has to do with not having the "self" argument
> > available in the arglist for the callback method, but I haven't been
> > able to find any documentation on how to provide "self" or what
> > Python function to use for calling an object method.
> 
> Well, I'm afraid I haven't bought my telepathy upgrade for 2002, so
> you'd better
> 
> a) post some more code
> b) tell us exactly what goes wrong
> 
> Cheers,
> M.



From greg at cosc.canterbury.ac.nz  Mon Apr 22 22:30:27 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 23 Apr 2002 14:30:27 +1200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com> <1019325450.422371@cswreg.cos.agilent.com> <mOTw8.73436$T%5.33234@atlpnn01.usenetserver.com>
Message-ID: <3CC4C743.5A3D89AF@cosc.canterbury.ac.nz>

Steve Holden wrote:
> 
> I don't feel you have yet even adequately *defined* "conceptual sameness",
> so it's surely premature to assert that it's missing from the language.

The definition he seems to have in mind is something like

  def eq(a, b):
    if is_immutable(a) and is_immutable(b):
      return a == b
    else:
      return a is b

for some suitable definition of is_immutable.

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


From phr-n2002a at nightsong.com  Wed Apr 17 18:33:53 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 17 Apr 2002 15:33:53 -0700
Subject: What webhosting
References: <3CBDE7B4.3348.4EB76@localhost> <mailman.1019078977.17448.python-list@python.org>
Message-ID: <7xads2vtta.fsf@ruckus.brouhaha.com>

William Park <opengeometry at yahoo.ca> writes:
> > Hi,
> > Does anybody know about a webhosting provider ( with Python and 
> > Telnet access) that does not remove a user's account if the user 
> > sends occasionally some unsolicited emails?
> > Thanks for reply
> > Ladislav
> 
> Get a static IP and do the hosting yourself.  Then, you don't have to
> answer to anyone.

They'll generally remove accounts for spamming even if the spammer has
a static IP.  Die, spammers, die, die, die.


From jason at tishler.net  Mon Apr 15 10:34:36 2002
From: jason at tishler.net (Jason Tishler)
Date: Mon, 15 Apr 2002 10:34:36 -0400
Subject: python 2.2.1 on Cygwin
In-Reply-To: <ly7kngfnoh.fsf@bock.chem.unc.edu>
References: <ly7kngfnoh.fsf@bock.chem.unc.edu>
Message-ID: <20020415143436.GA1324@tishler.net>

Jochen,

[Sorry for the sluggish response time...]

On Tue, Apr 09, 2002 at 07:38:06PM -0400, Jochen K?pper wrote:
> The following message is a courtesy copy of an article
> that has been posted to comp.lang.python as well.

Why not just cross post?  Although the use of cross posting should be
minimized, IMO, it is appropriate is this case.

> I have just compiled latest cvs release22-maint on Cygwin.  I get the
> following test results:
> ,----
> | 149 tests OK.
> | 7 tests failed:
> |     test_mimetypes test_mmap test_pyclbr test_socket test_sundry
> |     test_urllib test_urllib2
> | 31 tests skipped:
> |     test___all__ test_al test_asynchat test_bsddb test_cd test_cgi
> |     test_cl test_curses test_dbm test_dl test_fork1 test_gl
> |     test_imgfile test_largefile test_linuxaudiodev test_locale
> |     test_minidom test_nis test_ntpath test_pyexpat test_sax
> |     test_socket_ssl test_socketserver test_sunaudiodev test_thread
> |     test_threaded_import test_threadedtempfile test_threading
> |     test_unicode_file test_winreg test_winsound
> `----

On my system:

    Cygwin CVS from 2002/4/5
    Python 2.2.1 with a static_socket module
    Windows 2000 SP2
    NTFS

I get the following results:

    156 tests OK.
    2 tests failed:
        test_commands test_mmap
    29 tests skipped:
        test_al test_asynchat test_bsddb test_cd test_cl test_curses
        test_dbm test_dl test_fork1 test_gl test_imgfile test_largefile
        test_linuxaudiodev test_locale test_minidom test_nis test_ntpath
        test_pyexpat test_sax test_socket_ssl test_socketserver
        test_sunaudiodev test_thread test_threaded_import
        test_threadedtempfile test_threading test_unicode_file test_winreg
        test_winsound


I diff-ed test_commands and test_mmap between 2.2 and 2.2.1 and there
are differences that explain the above failures:

    test_commands fails because the regular expression is pickier in
    2.2.1 and gets confused by spaces in the user and/or group name
    (e.g. Domain Users) fields in the output of "ls -ld /.".

    test_mmap fails because a test file is left open (i.e., only
    unlink-ed) which causes a subsequent (and completely independent)
    mmap test to fail due to the standard open file "problem" under
    Windows.

I'm not sure how to best fix the test_commands problem, but following
did the trick on my system :,)

    $ chgrp Administrators /

The attached patch fixes the test_mmap problem.  I will submit it to
Python CVS for consideration.

After making the above changes, all 2.2.1 regression tests pass on
my system.

> I am not sure why test reported to fail seem to work ok?

Me neither.

> Anyway, hope it helps.  Please cc me if you have questions or need
> more info.

What is your Windows and Cygwin version?

Jason

-------------- next part --------------
--- test_mmap.py.orig	Mon Apr 15 08:24:54 2002
+++ test_mmap.py	Mon Apr 15 09:31:29 2002
@@ -238,6 +238,7 @@ def test_both():
                 pass
             else:
                 verify(0, "Incompatible parameters should raise ValueError.")
+            f.close()
     finally:
         try:
             os.unlink(TESTFN)
@@ -252,6 +253,7 @@ def test_both():
         data = 'aabaac\x00deef\x00\x00aa\x00'
         n = len(data)
         f.write(data)
+        f.flush()
         m = mmap.mmap(f.fileno(), n)
         f.close()
 


From geoff at gerrietts.net  Thu Apr 11 17:14:22 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 11 Apr 2002 14:14:22 -0700
Subject: Persistence
In-Reply-To: <a94qk5$22mg@r02n01.cac.psu.edu>
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu>
Message-ID: <20020411211422.GC4247@isis.gerrietts.net>

Quoting Rajarshi Guha (rxg218 at psu.edu):
> On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de Maria wrote:
> 
> > Is there a way to save in a file the status, objects, functions of
> > interpreter
> > in order to restore it after the interpreter has been closed?
> > 
> > If it is possible, one can work on a project stop and then continue
> > later adding functions, objects and so on.
> 
> I have a related question - does the OP refer to object persistence or 
> object serialization? Is there a difference between the two?
> 
> I looked at the shelve and pickle modules - the former uses a dbm database 
> whereas the latter just dumps an ASCII representation of the object. The 
> fact that shelve uses a dbm indicates that it should be more efficient. Is 
> this true? When would I choose one over the other?

Shelve uses pickle under the hood -- you probably knew that but if you
didn't, now you do.

Shelve is effectively the equivalent of pickling a dictionary of
objects, one at a time, with the exception (I believe?) that you must
use strings as keys.

Shelve consequently allows you named access to your objects, and
allows you to store your objects in a disk cache, retrieving and
writing one at a time.

Pickle-to-a-file will dump a single object or multiple objects one
after the other. When un-pickling, they come back out one at a time. I
don't remember whether the pickle is a FILO or a FIFO, but it's EITHER
a queue or a stack.

That means no random access to your objects, but also no hash-table
overhead, and none of the limitations that go along with it (if you're
not using GDBM, your objects have a size limit, etc).

My preference for small graphs of objects is to dump the whole graph
and retrieve the whole graph and do my operations in-memory (for
definitions of small < 1MB). It's simpler, has fewer limitations, and
fewer dependencies. If I were in a situation where I were doing more
sophisticated things with larger volumes, I would consider shelve, but
probably also look very hard at ZODB or even just going to a full
relational DBMS.

Good luck,
--G.

-- 
Geoff Gerrietts             "That's it! I've had it with your sassy mouth!
<geoff at gerrietts net>     I didn't want to do this! (Well, actually, 
http://www.gerrietts.net/    I did....)"  -- Mojo Jojo, "Bubblevicious"




From bogus@does.not.exist.com  Mon Apr 22 00:27:21 2002
From: bogus@does.not.exist.com (James Chapman)
Date: Mon, 22 Apr 2002 04:27:21 GMT
Subject: newbie dictionary question
References: <TiHw8.99936$CH1.86776@sccrnsc02> <3CC34519.E6C1063A@gmx.de> <g1Jw8.89176$G72.64386@sccrnsc01> <a9vn1g$md7$1@bob.news.rcn.net>
Message-ID: <JiMw8.51044$HH5.4376@rwcrnsc51.ops.asp.att.net>

Hi,

Thanks for the solution. I think it's better then what I wrote.

I think I'll also change the keys to intergers. While the bids come in as
floats they are constrained too be no more then three decimal places.

Thanks again,


"Raymond Hettinger" <python at rcn.com> wrote in message
news:a9vn1g$md7$1 at bob.news.rcn.net...
> "James Chapman ude.awoiu at nampahc-semaj>" <<reverse> > Hi,
> >
> I'm not sure if this counts as elegant, but it works:
> >>> [a.__setitem__(k, a.get(k,0)+v) for k,v in b.iteritems()]
> [None, None]
> >>> a
> {4.9699999999999998: 90, 4.9500000000000002: 200, 4.9900000000000002: 95,
> 4.96: 99}
>
> FWIW, I think using floats as a key is perilous.
>
>
> Raymond Hettinger
>
>




From ajs at ix.netcom.com  Sat Apr  6 09:10:03 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sat, 6 Apr 2002 09:10:03 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <001501c1dd74$bf5be4e0$0334fea9@carol>

Christian asks me to get a life.

But I think he mis-intuits.

Certainly *I* don't think, for example,
that there was anything unreasonable about
Guido's remark that

"I don't think Boolean algebra is a big subject."

Yet David's - a PEP285 supporter -  remark
indicating there might be more interest
in the subject than Guido thought - was purely
conversational and friendly.

That Mark felt compelled to intercede in  Guido's
defense I intuited as strange - a example of the
Phenom.

After  my post, but presuming I will be attacked
I take look in at Mark's weblog:

"my instant reaction to pep 285 (yes, i'm still thinking about this) was,
"rock on, guido!" it just seemed obviously right. then came the fud (labeled
such by guido-i wish i were that clever)."

For some, Guido's cleverness goes well beyond proven
talent at computer language design.

I am agnostic on the subject of Guido's complicity in
creating what I consider to be a "willies" atmosphere.

And if my interest was *not* in the area of education
I'd have a *much* better sense of humor about it all.

Art








From twisted at itamarst.org  Tue Apr  2 15:00:44 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Tue, 02 Apr 2002 22:00:44 +0200
Subject: ANN: Twisted 0.16.0
References: <Pine.LNX.4.44.0204021043340.31094-100000@mercury.sabren.com>
Message-ID: <3CAA0DEC.5060302@itamarst.org>

I should probably mention, in case it isn't clear - Twisted is licensed 
under the LGPL.




From tejarex at yahoo.com  Sat Apr 20 17:08:10 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 20 Apr 2002 21:08:10 GMT
Subject: Disappearing global variables
References: <slrnac3lc2.3hm.ak@ak.silmarill.org>
Message-ID: <_Mkw8.5895$Dm.409622@bin2.nnrp.aus1.giganews.com>

"Andrei Kulakov" <ak at silmarill.org> wrote in message
news:slrnac3lc2.3hm.ak at ak.silmarill.org...
> You can look at full program at silmarill.org/pybook/pybook.py -
it's
> a little over 500 lines.

can you produce a considerable shorter (minimal) version that exhibits
the same problem?





From kragen at pobox.com  Mon Apr 29 18:17:58 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 29 Apr 2002 18:17:58 -0400
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <3CCC730A.FC60D13F@alcyone.com> <3ccd1ed9$0$15477$afc38c87@news.optusnet.com.au>
Message-ID: <837kmqrvxl.fsf@panacea.canonical.org>

Alejandro Dubrovsky <s328940 at student.uq.edu.au> writes:
> yes, but that means saving the definition as a string everytime that the
> lambda is changed.  i was looking for something "neater". eg for function
> variables, i can just save and restore the variables'.__name__ attribute.
> i was wondering if there was something similar for lambdas.

Unfortunately not.  You can't marshal lambdas either.

Generally, you can refactor code that uses lambdas into code that uses
object methods; your closure data goes into object attributes, and
each lambda becomes a class.  Or you can just use named functions,
especially if you don't have closure data.



From djc at object-craft.com.au  Wed Apr 17 20:57:08 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 18 Apr 2002 10:57:08 +1000
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de> <m3g01um27o.fsf@ferret.object-craft.com.au> <m3ads2dkpy.fsf@mira.informatik.hu-berlin.de>
Message-ID: <m38z7lrfh7.fsf@ferret.object-craft.com.au>

>>>>> "Martin" == Martin v Loewis <martin at v.loewis.de> writes:

Martin> Dave Cole <djc at object-craft.com.au> writes:
>> If you create a new thread state for the callback, what will happen
>> to Python exceptions raised in the callback?

Martin> They propagate back to the bottom of the callback. If you
Martin> don't treat them special, deleting the thread state will also
Martin> release the exception object and traceback object. Assuming
Martin> the thread callback came "out of nowhere", there is nothing
Martin> reasonable to do, anyway: you might want to print the
Martin> exception to stderr or something like that.

I think that it is preferable to keep the exception and propagate it
back from the callback through the API to the thread which called the
API.  I am not sure why discarding the exception would be a good idea.

IMO, from the point of view of the extension module user it should not
be visible that there was some non-Python code between the Python code
where the exception was raised and the Python code which called the
module API.

- Dave

-- 
http://www.object-craft.com.au


From skip at pobox.com  Wed Apr 17 13:39:13 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 17 Apr 2002 12:39:13 -0500
Subject: Readline/command history w/ interactive prompt
In-Reply-To: <20020417163818.GA19289@andy.tynant.ftech.net>
References: <20020417163818.GA19289@andy.tynant.ftech.net>
Message-ID: <15549.45889.881737.540933@12-248-41-177.client.attbi.com>


    Andy> One of the things I find "warty" about the interactive prompt is
    Andy> that the command-line history retains multiple copies of identical
    Andy> successive lines, rather than trimming them down to a single
    Andy> instance.

See if the attached patch works for you.  If so, I'll submit it to SF.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: readline.diff
Type: test/plain
Size: 1359 bytes
Desc: patch for readline to suppress duplicate history entries
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/f06952bd/attachment.bin>

From reply at in_newsgroup.tks  Fri Apr  5 07:40:02 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 5 Apr 2002 12:40:02 GMT
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk> <23891c90.0204050156.1840d913@posting.google.com>
Message-ID: <a8k5v2$8kp$1@hfc.pacific.net.hk>

Thanks for your feedback.

paul at boddie.net (Paul Boddie) wrote in
news:23891c90.0204050156.1840d913 at posting.google.com: 

> I hadn't seen this one before, but it looks rather interesting. A
> number of people have already asked about the Web modules overview
> which I had, and in the replacement document(s) I should attempt to
> cover Slither as well as the other major frameworks.

Alias, I also wanted to see your previous Web modules overwiew but am 
looking forward to seeing your replacement doc(s) ;-) Do you have a website 
I can visit so that I'll know when you publish those docs?

[snip]

> Cheetah is a templating system, but you should not let the other
> options listed confuse you too much - whether one uses mod_python,
> mod_snake or PyApache should really only be a deployment and/or
> performance consideration.

>From what I'd read, it seemed like mod_snake was ahead of mod_python in 
implementing Apache 2x support.

[snip]

> There are live sites running Webware (such as http://www.vorbis.com)
> and whilst, in my opinion, there are various issues that arise about
> support for certain things which one might have believed to have been
> covered by now, the development community is very active and very
> eager to improve the framework. Reading the list shows how hard it is
> to write a general Web framework, and I would recommend anyone
> thinking of writing their own framework to skim the archives (and view
> the Wiki) at the very least.

Personally I find it confusing at this stage of development of Python based 
web app servers/frameworks. Although many projects overlap (some more than 
others), I guess it just shows that no one has yet found a 'silver bullet' 
solution. Perhaps with the development requirements of various types of 
Python web apps, it won't be possible for one Python web app server/RAD 
tool to even come close to covering all the bases.

As I'm a Python newbie, I'm looking for a general, rather than specialized, 
web app platform and dev tools. I'm not a purist (dev would take too long 
:) ) as I don't want to create everyting and the kitchen sink myself.

> There are actually many persistence modules in Python, as it turns
> out, partly because persistence is a bit like Web page templating -
> everyone has their own idea about how it should be done - but it's
> also a subject which lends itself to different approaches. (In the
> article you referenced, I still agree with those comments on
> templating systems.)

I do like to separate style, layout, app and data and will avoid using PSP, 
Cheetah etc. completely or at least in ways which integrate app code within 
HTML pages.

Just the sound of templates/rendering tools on the other hand, does sound 
interesting. Provided that they are sufficiently flexible i.e. all sites 
don't look like clones, then it would be worth looking at projects which 
include these.

> Having experimented with Zope, before becoming disappointed or
> concerned about development, deployment and management issues, and
> finally looking into Zope again after seeing interesting applications
> like Squishdot in widespread production use, I think Zope is very
> suitable for a number of application types. Moreover, through various
> features, Zope specially addresses areas that the other frameworks are
> less likely to address "out of the box", such as the way applications
> can be configured by creating objects at appropriate places in the
> object database - this is very intuitive (and quite remarkable) once
> the concepts are understood.

It'll be good to hear more about the application types to which Zope is 
more suited.

I think that Zope is one of the first projects that Python newbies 
'consider' using, when looking for web app servers/RAD tools to help in 
developing their first web app. There is quite a lot of user 'noise' though 
, which makes me concerned whether I should begin by devoting time to learn 
Zope. That's why I am looking at other, (less complex?) projects.

Thanks again,
Scrumpy :)
has left the building


From martin at v.loewis.de  Wed Apr 17 14:40:30 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 20:40:30 +0200
Subject: Curses for windows.
References: <f8e5025.0204170953.570cfb43@posting.google.com>
Message-ID: <m3it6qyxr5.fsf@mira.informatik.hu-berlin.de>

alx78 at hotmail.com (alex) writes:

> Where is "curses" module for python 2.2 in windows NT.

Curses is not supported on Windows NT.

Regards,
Martin


From sholden at holdenweb.com  Thu Apr 25 20:14:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 25 Apr 2002 20:14:54 -0400
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <mailman.1019778048.17911.python-list@python.org>
Message-ID: <t01y8.22574$z55.20598@atlpnn01.usenetserver.com>

"Alexander Skwar" <lists.ASkwar at DigitalProjects.com> wrote in message
news:mailman.1019778048.17911.python-list at python.org...
?Chris Spencer? sagte am 2002-04-25 um 17:14:12 -0500 :
> Okay, I might not be a math genius, but this is most definately WRONG.

Yep, you're not a math genius.  The result is completely correct.  Read the
tutorial for a good explanation!


s/correct/explicable/

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From seb_haase at yahoo.com  Fri Apr  5 14:05:22 2002
From: seb_haase at yahoo.com (Sebastian Haase)
Date: 5 Apr 2002 11:05:22 -0800
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de> <f445a2fe.0204041708.6fd6fd0d@posting.google.com> <m37knm395e.fsf@mira.informatik.hu-berlin.de>
Message-ID: <f445a2fe.0204051105.6f67b848@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m37knm395e.fsf at mira.informatik.hu-berlin.de>...
> seb_haase at yahoo.com (Sebastian Haase) writes:
> 
> > I think that wouldn't work - since I'm still interested in debugging
> > _my_ stuff - so that needs me to #define _DEBUG
> 
> Can you elaborate? If you want code that is only active under
> debugging, why can't you wrap it in #ifdef HAASE_DEBUG or some such?
> 
> If you think that you would not get symbolic debugger support if
> _DEBUG is not defined, you are wrong.

I'm just saying: Why do I need to touch _DEBUG at all? Normally if I
compile my code in debug-mode _DEBUG is defined -- thats just how
people are doing that. (As I understand...)
The story just goes the other way around: Why wants the python.h force
us to include a lib (python22_d.lib)  that it doesn't ship. (And I'm
_not_ interested in debugging the python lib anyway?)

What was this comment though about the "linking the MSVC debug C libs"
 -- might this issue cause me trouble ?? I'm not experiencing any
problems so far...

Regards,
Sebastian


From netzapper at magicstar.net  Tue Apr  2 11:43:12 2002
From: netzapper at magicstar.net (A. Jones)
Date: Tue, 02 Apr 2002 16:43:12 GMT
Subject: file.write[lines]() weirdness
References: <3ca96e30.52835380@mammoth.usenet-access.com> <j4adsm2l57.fsf@informatik.hu-berlin.de>
Message-ID: <3ca9df6b.81826899@mammoth.usenet-access.com>

On 02 Apr 2002 17:08:36 +0200, loewis at informatik.hu-berlin.de (Martin
v. =?iso-8859-1?q?L=F6wis?=) wrote:

>
>I recommend to open the file in binary.
>
>Regards,
>Martin
>

Will this result in text in the database that is unreadable by humans?
(not that this particularly matters)


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From phr-n2002a at nightsong.com  Wed Apr  3 02:48:33 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 23:48:33 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <mailman.1017817926.31983.python-list@python.org>
Message-ID: <7xhemtz0ha.fsf@ruckus.brouhaha.com>

"Delaney, Timothy" <tdelaney at avaya.com> writes:
> > That looks pretty similar to an Info file (texinfo output).  I think
> > I'd rather type "@section{Overview}" than have to supply those special
> > underscores directly.
> 
> The whole point is that this is the *input*, optionally to produce another
> format, and is also human-readable in the source code (or any other text
> file). Your statement above simply supports the use of ReST, as its native
> format is "pretty similar" to texinfo output, which you have stated
> previously that you prefer to the texinfo source.

I should be more clear--normally you read Info files through an Info
browser rather than reading the files directly.  They have funny markup
characters in them like ReST does.  My comment can be read as "ReST looks
to me more like object code than source code".

> Huh? If you'd read the accompanying docs, you would know about the ReST
> directives. I wouldn't call the .. contents:: directive "explicit markup"
> considering that it inserts an entire table of contents with one short line

Of course it's explicit markup, regardless of what it does.

> > I looked briefly at the ReST documentation linked from the PEP, but
> > if I'm supposed to read it in detail to be able to use it, I might
> > as well use a more traditional markup language.
> 
> I looked briefly at the Python documentation linked from the web site, but
> if I'm supposed to read it in detail to be able to use it, I might as well
> use a more traditional programming language.

Python is a traditional programming language.


From djc at object-craft.com.au  Mon Apr 15 19:51:16 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 16 Apr 2002 09:51:16 +1000
Subject: Questions on the C Thread API
References: <mailman.1018871909.16619.python-list@python.org> <Xns91F18C58A3AAAduncanrcpcouk@127.0.0.1> <mailman.1018886099.26322.python-list@python.org>
Message-ID: <m3d6x08qrv.fsf@ferret.object-craft.com.au>

>>>>> "Phil" == Phil Thompson <phil at river-bank.demon.co.uk> writes:

Phil> This is the bit I don't understand. How can I get the saved
Phil> thread state for the current thread? Do you mean I can use the
Phil> Python API to get it (in which case I get back to Catch 22)? Or
Phil> do you mean I have to write code that determines by other means
Phil> what the current thread is and then get the corresponding Python
Phil> thread state?

I am about to try doing almost exactly the same thing with my Sybase
module.  The way I understand it is that your code is responsible for
associating the Python thread state with the actual thread.

The way I am going to do it is to store the Python thread state inside
one of my structures which I associate with one of the Sybase API
callback parameters.  When Sybase invokes a callback I then locate the
correct threadstate by searching through non-Python structures and
recquire the GIL.

- Dave

-- 
http://www.object-craft.com.au


From fredrik at pythonware.com  Tue Apr 30 17:06:15 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 21:06:15 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <uctooia295ne2f@news.supernews.com>
Message-ID: <bHDz8.34855$n4.7400665@newsc.telia.net>

John Roth wrote:
> To me, .join() makes much more sense as a list or tuple method than
> either. Semantically, .join() takes a sequence and transforms it into
> a string, therefore it should be a method on the input type (the
> sequence)

there is no sequence type in Python.  try again.

</F>




From johnroth at ameritech.net  Fri Apr 19 17:52:38 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 19 Apr 2002 17:52:38 -0400
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net>
Message-ID: <uc14cjalpru3b5@news.supernews.com>

"Richard Gruet" <rjgruet at yahoo.com> wrote in message
news:a9nbuq$916$1 at aphrodite.grec.isp.9tel.net...
> Alex
>
> And BTW, people would not to have implement any __len__ method if all
the
> sequence types would derive from a common ancestor (sequence), in
which the
> (in)famous method would be implemented. Would it be completely
nonsensical ?
> I have been told that starting from 2.2, new style classes (including
most
> built-in types) all derive from a common base class 'object'.

As far as I know, that's correct. Inserting a .len() method in the
object class would automatically make it available to all new style
classes.

As far as I can see, there is one major issue with this: if the
class in question didn't have a __len__() method, it would
still have the (inherited) len() method. This could be rigged to
raise an exception, but it would still be a bit of a trap for the
unwary.

There's also the question of the inheritance hierarchy, which
I don't pretend to understand.

It seems to be a case of simple implementation, small benefit,
small downside, no discernable side effects. Write a PEP.

John Roth




From brueckd at tbye.com  Fri Apr  5 10:26:08 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 5 Apr 2002 07:26:08 -0800 (PST)
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
In-Reply-To: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
Message-ID: <Pine.LNX.4.33.0204050721550.9382-100000@ns0.tbye.com>

On Thu, 4 Apr 2002, phil hunt wrote:

> Recently on this newsgroup there's been a discussion about adding a
> bool datatype to Python; and Guido has decided to do this.
[snip]

> In both cases, there are good arguments in favour of the change.
[snip]

> Are the language developers constantly going to tinker with the
> language, out of a sense of fun?

I doubt they're doing it for the fun factor. See your own previous
comment, above.

-Dave





From rjones at ekit-inc.com  Thu Apr 18 00:44:08 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 18 Apr 2002 14:44:08 +1000
Subject: Hostname Resolution not working in Explorer
In-Reply-To: <200204181437.26855.rjones@ekit-inc.com>
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net> <3CBE4DAA.E710D618@engcorp.com> <200204181437.26855.rjones@ekit-inc.com>
Message-ID: <200204181444.08605.rjones@ekit-inc.com>

On Thu, 18 Apr 2002 14:37, Richard Jones wrote:
> This is a common problem with Windows - it doesn't know how to resolve
> "localhost" in some situations. Usually it coincides with you changing the
> network configuration (unplugging a network cable or something).

Oh, I forgot to add - I believe in _some_ versions of Windows you can 
hard-code an entry in a hosts file you place in the Windows System directory 
(IIRC - it's been a while - there'll be a sample file in there I think). The 
entry would read

127.0.0.1 localhost

just the one line in a file called "hosts" in the Windows System dir, and it 
should fix things. Well, it did once for me :)


    Richard





From jacek.generowicz at cern.ch  Fri Apr 26 14:09:24 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 26 Apr 2002 20:09:24 +0200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <tyf1yd25zgs.fsf@pcitapi22.cern.ch> <mailman.1019840809.5035.python-list@python.org>
Message-ID: <tyfvgae2uy3.fsf@pcitapi22.cern.ch>

"James J. Besemer" <jb at cascade-sys.com> writes:

> Jacek Generowicz wrote:
> 
> > No, as Paul said, they are used for "function calls". Admittedly,
> > in mathematical expressions function call often coincide with
> > sub-expressions, but it does depend on the sub-expression in
> > question.
> 
> Thanks for the clarification.  Now I get it.
> 
> > Compare
> >
> >    (+ 1 2 3 4 5 6 7 8 9 10)
> >
> > with
> >
> >    1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10
> >
> > Gosh, what a lot of typing I have to do in Python, compared to lisp.
> 
> So that begs the question: is it more common to have lengthy "sub
> expressions" like your example or is my example with a mixture of
> operations more typical?

My example was only half serious. (Well, three-quarters serious <0.75 wink>)

> > But, if you are dealing with mathematical expressions, then,
> > obviously, you will use some sort of infix notation macro:
> 
> I am not familiar with Common Lisp and did not know that was common.

I'm not sure what this has to do with _Common_ Lisp in particular.

I see it a bit like like criticizing python, for example, for not
allowing you to build a usefully fast matrix class, while completely
ignoring the existence of Numpy.

> Though I suppose I should have guessed that anything's possible.

Now you're talking :-)

Actually, if you are not familiar with lisp macros, then you may well
be very surprised by just how far "anything" goes in this case.

> > Incidentally, it would be interesting to see how many people would
> > correctly work out in their heads (on first attempt) what
> > 1+2*3/4-5 evaluates to (leaving aside the fact that in Python it
> > evaluates to -3[*]) . . . and then compare it to the success rate
> > when considering the equivalent (+ 1 (/ (* 2 3) 4) -5). Or maybe
> > how long it takes them to get the right answer in each case.
> 
> Hard to say.  Do the survey and lemme know.
> 
> Then too, this is a simple example and it seems the big trick in
> either case is in knowing how divide works.
> 
> In Python's case it's truly a trick question as there presently is
> absolutely no way to predict what the result will be without several
> bits of additional information.  (I came late to the argument, but
> personally I think so called "true division" is/was a big mistake.)

I specifically mentioned that I wanted to leave aside consideration of
python's answer to this, because the point I was trying to make was
that it seems to me that _I_ can work out (+ 1 (/ (* 2 3) 4) -5) more
easily (at first blush) than 1+2*3/4-5, which surprised me (just a
little), given that I had just argued (in agreement with you) that (in
general) the latter is probably easier for mathematical expressions.

> Then too, the results probably vary among Lisp implementations, so
> you'd have to be more specific in that case too.  IIRC, some systems
> were integer only, some were real only and some had both numeric
> types.  I dunno what's standard these days.

I'm not sure what you mean. I'm no expert on the features of various
lisp implementations in history. ANSI Common Lisp (which was
standardized around the time python was born IIRC) is what I would
guess most lisp users use these days (elisp has a very specific
domain), has a plethora of numeric types including unlimited size
integers, rationals and complex numbers.

All ANSI CL compliant implementations will agree.

> So I guess the "trick" parts of the question would far outweigh any
> data about infix vs. postfix.

Which is exactly why I was trying to steer the conversation clear of
the trick parts. What I had in mind was:

1) Ensure that the guinea pig (gp) understands the meaning of the
   mathematical operators.

2) Ensure that gp understands both prefix and infix notation.

3) Show gp one of the expressios written down on a piece of paper.

4) Note the time taken for gp to produce the correct answer.

By "correct answer" I mean any of -5/2, -2.5, "minus two and a half",
"the negative square root of six and a quarter", etc., but most
certainly _not_ -3 or -2.

5) Repeat for many different gps with different expressions.

> > > > I'd expect Lisp syntax to be rather less error-prone, if
> > > > anything, since it's much simpler
> > >
> > > I think that's debatable.  Certainly simplicity makes it easier
> > > to learn, so that's a plus.  But it's still very tedious to get
> > > all the parentheses matched up just right.
> >
> > It seems that many (most?) people who actually try to write lisp
> > without making a big issue of the parentheses, very rapidly come
> > to a conclusion that disagrees with yours.
> 
> 1. I think that's a fair statement.  Earlier in the week, someone
> suggested I'd be more happy with Python after I was "reprogrammed"
> to think "Pythonically"

I think it is generally true that embracing the features of a
language, rather that struggling against them and trying to make the
language fit the mindset one has picked up in other languages, is a
good thing. At least in the initial stages :-)

Cheers,


From neuhauser at mail.cz  Sat Apr 27 13:10:24 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Sat, 27 Apr 2002 19:10:24 +0200
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
In-Reply-To: <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com>
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org> <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com>
Message-ID: <20020427171024.GA82912@roman.mobil.cz>

> From: "Steve Holden" <sholden at holdenweb.com>
> Subject: Re: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
> To: python-list at python.org
> Date: Sat, 27 Apr 2002 06:47:50 -0400
 
> > "James J. Besemer" wrote:
> > > Then too, I'm crazy enough to count assignment operators as another
> > > one of Python's few serious omissions.  I wish to hell I could simply
> > > write:
> > >
> > >     while data = socket.recv( 1024 ):
> > >         process( data )
> > >
> Aha! So you are a closet "Python should be C" adherent? You should (since
> you are thoroughly familiar with the above FAQ entry) know from FAQ 6.30
> this is a definite no-no. Indeed, once you allow assignment operators you
> bring in as allowable the very case you say above that Python would be able
> to detect easily. The FAQ unambiguously states """The reason for not
> allowing assignment in Python expressions is a common, hard-to-find bug in
> those other languages, caused by this construct:
>     if (x = 0) {
>         ...error handling...
>     }
>     else {
>         ...code that only works for nonzero x...
>     }
> 
> """

    Anyone who writes "if (x == 0)" instead of "if (0 == x)" deserves
    what he begs for.

-- 
FreeBSD 4.4-STABLE
7:04PM up 12 days, 8:17, 11 users, load averages: 0.00, 0.02, 0.00




From maxm at mxm.dk  Fri Apr  5 06:44:15 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 13:44:15 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAD4CD8.2090700@mxm.dk> <7Adr8.1447$%A.217236@news1.west.cox.net> <Ster8.62071$S52.2165674@news2.tin.it>
Message-ID: <3CAD8E0F.3080404@mxm.dk>

Alex Martelli wrote:

 > Actually, the VIM-only case-insensitivity flag is crucial here,
 > since it lets you substitute both uppercase 1 and lowercase 1
 > in one sweep (pity that, as written, it works for one line only,
 > but, if one's Python code IS just one line in all, that's OK).

Maybe I should brush up on my reg-ex's

But though I love the Python syntax I really HATE the indentation issue. 
So to keep my code more readable I usually put it all in one line, as my 
editor wraps it automatically at #80.

If I need more code than can fit into a single line I just write it in 
another file.

All those files also makes the project almost self-documenting.

Maybe lesser programmers choose to do othervise, but see how simple my 
regex for replacing is as an added bonus.

;-) <- smiley added this time around

regards Max M



From JamesL at Lugoj.com  Sat Apr  6 18:47:01 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 23:47:01 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <mailman.1018115475.21810.python-list@python.org>
Message-ID: <Xns91E8A12A8C2A3JamesLLugojcom@216.148.53.96>

"Brian Quinlan" <brian at sweetapp.com> wrote:
> Unicode was too important not to add to the language, even if it did
> cause a few headaches. But the addition was done in a manner that
> favored compatibility over optimal semantics.

Unicode is nice. So are all the other features that are being added to 2.x. 
It is that "x" in 2.x is growing large at an alarming rate and these 
"minor" releases are not minor in the colloquial meaning.

> What is the total amount of time that you have spent resolving
> compatibility problems?

Probably a few days and some worried nights. But the context of how these 
problems cropped up and the unease they imparted were also important.

> What language are you going to use instead of
> Python?

As I've said elsewhere, I don't know. I'll look again in not less than 6 
months. Might look at Lua or Ruby to start.

> Please list some languages that are very forward and backwards
> compatible between releases.

None are, so the next best thing is to look for ones that have a manageable 
number of extant releases.


From geoff at gerrietts.net  Mon Apr 22 20:50:58 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 22 Apr 2002 17:50:58 -0700
Subject: What is an import and where is it on the hard disk
In-Reply-To: <3CC4A51A.A7D770ED@sympatico.ca>
References: <3CC4A51A.A7D770ED@sympatico.ca>
Message-ID: <20020423005058.GB13077@isis.gerrietts.net>

Quoting francis049 at sympatico.ca (francis049 at sympatico.ca):
> I receive an error message like :
> 
>     Traceback (innermost last) :
>     File : /usr/sbin/up2date, line 9, in ?
>       import rpm
>     Import error : No module named rpm
> 
> 
> Anyone can tell me why is this error happening and give me some details
> ... please ?

Your posting is amazingly spare with its details, but I'll venture a
guess that you've got a broken python installation on a RedHat system.

I'm not sure exactly how your python installation is broken. The most
common problems with RedHat are either a python version mismatch, or a
broken python install. It looks like you've got some version of python
installed. So, try this to figure out which version:

python -c "import sys; print sys.version"

If this says anything BUT 1.5.2, then you're likely to have problems
under RedHats 6.something forward.

If that's right, check:

python -c "import sys; print sys.path"

If you don't see /usr/lib/python1.5/site-packages and
/usr/lib/python1.5 in those places, then it's likely that your
installation is "broken", in that you're missing some files.

Anyway, that's my guess -- provide some more details on your system,
your python installation, etc. if you're still having problems.

Thanks,
--G.

-- 
Geoff Gerrietts             "People talk fundamentals and superlatives 
<geoff at gerrietts net>     and then make some changes of detail." 
http://www.gerrietts.net                  --Oliver Wendell Holmes Jr




From vinay_sajip at yahoo.co.uk  Fri Apr 12 04:01:57 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 12 Apr 2002 01:01:57 -0700
Subject: PEP 282: A Logging System
References: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru> <mailman.1018538209.18235.python-list@python.org> <a95c7k$801$1@panix1.panix.com>
Message-ID: <2e37dc1.0204120001.11ee98b0@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<a95c7k$801$1 at panix1.panix.com>...
> Your main difficulty with thread safety is WRT I/O.  How do you plan to
> handle that?

I/O is the province of the handlers, and I expect to make them
thread-safe with respect to the underlying I/O through the use of
thread locks. Of course this applies only to the handlers I provide as
part of the distribution; user-defined handlers will need to make
their own provision.

The code released up till now does not have any locking at all.
Empirically, a simple test harness I put together (just to see how the
current code does) seems to run fine (eyeball-wise) on Python
2.1.1/Windows and Python 1.5.2/Windows and Linux, though with 1.5.2 I
had to remove calls to inspect. The test was logging to a FileHandler
and a DatagramHandler. I was a little surprised at first not to see
any problems, might this be due to the GIL? I don't know enough in
detail about how the GIL works, but given that these handlers all work
through C objects I would expect the GIL to play a role.

I think that only the emit() method of handlers needs to be
synchronized - that's where the I/O happens. I expect to do something
like this:

class Handler(Filterer):
  def __init__(self, ...):
    ...
    self.getlock()
		
  def getlock(self):
    if thread:
      self.lock = thread.allocate_lock()
    else:
      self.lock = None
	
  def acquire(self):
    if self.lock:
      self.lock.acquire()
			
  def release(self):
    if self.lock:
      self.lock.release()
	
Then, in individual handlers...

class MyHandler(Handler):
  ...
  def emit(self, record):
    self.acquire()
    try:
      try:
        ... here, do the actual emission
      except:
        ... silently deal with exceptions during emission
    finally:
      self.release()
			
If anybody reading this spots any flaws in the above, please do let me
know :-)

Regards


Vinay Sajip


From krists at ifi.uio.no  Fri Apr  5 04:13:22 2002
From: krists at ifi.uio.no (Kristin Skoglund)
Date: Fri, 05 Apr 2002 11:13:22 +0200
Subject: Diff in lists
Message-ID: <3CAD6AB2.1D4E9846@ifi.uio.no>

Hi

Is it possible to run a diff-function on lists in Python?

Kristin




From tdelaney at avaya.com  Thu Apr  4 19:13:06 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 5 Apr 2002 10:13:06 +1000 
Subject: ACCEPTED: PEP 285
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A251@natasha.auslabs.avaya.com>

> From: Terry Reedy [mailto:tejarex at yahoo.com]
>
> > Clarification
> >
> >     This PEP does *not* change the fact that almost all object types
> >     can be used as truth values.  For example, when used in an if
> >     statement, an empty list is false and a non-empty one is true;
> >     this does not change and there is no plan to ever change this.
> 
> almost? all?  I am under the impression that 'if x:' is universally
> valid, and consider that one of the good features of Python's prime
> dichotomy system.

Something of the form:

class Klass:

    def __nonzero__(self):
        raise NotImplementedError

will fail an if x: test ...

Tim Delaney




From eav at ifrance.com  Wed Apr 24 06:02:13 2002
From: eav at ifrance.com (eav)
Date: Wed, 24 Apr 2002 11:02:13 +0100
Subject: linux console graphics
Message-ID: <3CC682A5.3050905@ifrance.com>

I am currently working on a project needing a graphical interface, 
supposed to run quickly on quite modest PC (first pentium generation) 
with linux or a *bsd.

What can Python do form me ? I've been considering projects such as 
pygame - but it seems that SDL can't be run from the console. Am I wrong?

I've read about a GGI wrapper, but it doesn't seem to be maintained...

What do you think ? Do I rather use C and svgalib or GGI ?


Ddom



From brueckd at tbye.com  Wed Apr 17 10:11:28 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 17 Apr 2002 07:11:28 -0700 (PDT)
Subject: pre-PEP for optional 'pass'
In-Reply-To: <Xns91F37CE5BFE3Cduncanrcpcouk@127.0.0.1>
Message-ID: <Pine.LNX.4.33.0204170706260.10117-100000@ns0.tbye.com>

On Wed, 17 Apr 2002, Duncan Booth wrote:

> > We're veering slightly from the original reason for this point, which
> > was how to write stub functions/methods which will be implemented
> > later.  The above certainly doesn't indicate any intent to actually
> > implement the methods (in fact, on the contrary, it suggests they
> > could be removed).
>
> I think you trimmed a bit too much when you followed up. I would have
> described what the functions were intended to do if I had known that,
> but as I said I have no idea what these particular functions do, so all
> I could document was the lack of implementation. Since I was being
> chastised for failing to produce meaningful docstrings for someone
> else's imaginary functions I made them as meaningful as I could in the
> circumstances.

No, you were being chastised for promoting a horrible standard - that
*all* classes and *all* methods should have a docstring. It's just plain
wrong.





From juenglin at informatik.uni-freiburg.de  Wed Apr  3 06:22:06 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 03 Apr 2002 13:22:06 +0200
Subject: PyInt_AsLong()
References: <xdu7knpaz87.fsf@leto.informatik.uni-freiburg.de> <EnBq8.29186$n4.5840811@newsc.telia.net>
Message-ID: <xduzo0l9gdd.fsf@leto.informatik.uni-freiburg.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Ralf Juengling wrote:
> 
> > PyInt_AsLong() returns '-1? to indicate an error, this meets the convetion
> > for PyXXX()-functions which return integers. But how do you distinguish an
> > '-1' indicating an error from a the legal long-value '-1'?
> 
> hint: when in doubt, grep the Modules directory in the
> Python source distribution:
> 
>     v = PyInt_AsLong(obj);
>     if (v == -1 && PyErr_Occurred())
>         return NULL; /* error */
>     /* v is valid */
> 
> hope this helps!
> 
> </F>

Okay. So the reason to indicate an error by '-1' is efficiency, 
I guess (i.e. calling PyErr_Occured() each time is more 
expensive than the prior check 'v == -1') ?

Thanks,
Ralf


From ods at fep.ru  Mon Apr 22 11:24:23 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Mon, 22 Apr 2002 19:24:23 +0400 (MSD)
Subject: Overriding methods in classes that use property()
In-Reply-To: <OF0B3F6398.7B97C194-ON85256BA3.00520742@juscom.com>
Message-ID: <Pine.LNX.4.44L.0204221923330.2654-100000@ods.pp.ru>

On Mon, 22 Apr 2002, Edmund Lian wrote:

EL> When I run the code appended below, class B inherits all the
EL> methods of
EL> class A, including the property() definition as I would
EL> expect.
EL> Unfortunately, class C also does the same; I cannot override
EL> class A's
EL> setVal method. Is this a bug, or do I have to do something
EL> different (and
EL> unexpected) with Python 2.2's new style classes?
EL>
EL> ...Edmund.
EL>
EL> class A(object):
EL>     def __init__(self):
EL>         self._val = None
EL>
EL>     def setVal(self, value):
EL>         self._val = value
EL>
EL>     def getVal (self):
EL>         return self._val
EL>
EL>     val = property(getVal, setVal)
EL>
EL> class B(A):
EL>     pass
EL>
EL> class C(A):
EL>     def setVal(self, value):
EL>         self._val = 'Huh?'

Simply add one line here:
        val = property(getVal, setVal)





From aleax at aleax.it  Thu Apr 18 18:50:12 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 22:50:12 GMT
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net>
Message-ID: <E4Iv8.2972$8D3.78501@news1.tin.it>

Richard Gruet wrote:

> Alex
> 
> Well, this is a very impressive and learned argumentation and I don't
> doubt that using operators and built-in functions offer a better
> polymorphism. I just meant that to my human eyes, it is easier to memorize
> when everything is consistently object oriented and there are only methods
> (like in Java - and I prefer Python to Java :-), or the contrary, only
> functions. Then I have not to remember which functions are methods and
> which ones are built-in. I think this is because a mixed model is more
> complicated than a non-mixed one.

"Entities must not be multiplied without necessity", no -- but in this case
the necessity, or rather the opportunity, IS there, as I showed.

Everything IS "consistently object-oriented" -- polymorphism is the
principal magic of object-orientedness.  There would be substantial
loss of power without measurable conceptual simplification if Python
had no methods (how would I store a bound-method just as any
other callable then?) or if it had no functions (serious polymorphism
loss).

Besides, I doubt your objection is whole-hearted.  What language
except some Lisp-like ones has ONLY one way to invoke code?  I
don't hear you complaining that Python has operators too -- yet
they differ in nothing but syntax-sugar from the builtin functions
that you appear to dislike so.  How is x**y so different from pow(x,y)
that you accept one and not the other -- syntax-sugar strictly --
and fail to see that x.pow(y) as the sole alternative would be UTTER
disaster?  pow(x, y) offers extras too -- if it can it calls x.__pow__(y),
else y.__rpow__(x).  Just like cmp(x, y) can call various different
special methods on x OR y depending on what works.

We can't have operators for _every_ functionality that could benefit
from such treatment, as I hope is obvious.  Defining everything as
method would be a HORRIBLE loss.


> And BTW, people would not to have implement any __len__ method if all the
> sequence types would derive from a common ancestor (sequence), in which
> the (in)famous method would be implemented. Would it be completely
> nonsensical ? 

Sorry, but to me it sure seems that way.  How would this mythical 'sequence'
implement your desired method in a way that's correct for all sequences?!
I see no sensible alternative to letting each sequence class determine what 
"length" means to it.

> I have been told that starting from 2.2, new style classes
> (including most built-in types) all derive from a common base class
> 'object'.

Yes, and?  What do you think type object implements, exactly?

> NB: I am NOT setting a dispute, but just expressing my feeling from a user
> perspective.

And I'm expressing my own feelings in return -- and motivating them, as
I see you're trying to motivate yours.  My feelings are that you (like most
people, it seems at times) "can't judge a book by its cover": instead of
trying to understand WHY a designer has set things up in a certain way,
and only forming judgment when you do have a decent grasp of that,
you seem to be snapping to "feelings" based on superficial impressions.

You may ASK "why", but you don't really WANT to know -- as evidenced
by your reaction when the question is answered, it appears to me.

If you want to program productively and understand things, Python and
this newsgroup can help.  If you prefer to "express feelings" based on
superficial syntax-sugar issues, I suspect they can help much less.


Alex



From spam at melkor.dnp.fmph.uniba.sk  Wed Apr 24 07:12:14 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Wed, 24 Apr 2002 13:12:14 +0200
Subject: linux console graphics
References: <3CC682A5.3050905@ifrance.com>
Message-ID: <eu36aa.2gm.ln@127.0.0.1>

eav <eav at ifrance.com> wrote:
 : I am currently working on a project needing a graphical interface, 
 : supposed to run quickly on quite modest PC (first pentium generation) 
 : with linux or a *bsd.

 : What can Python do form me ? I've been considering projects such as 
 : pygame - but it seems that SDL can't be run from the console. Am I wrong?

it can. In svga or framebuffer mode, or even aalib :-)
The question is if pygame supports it - but I think it should.


-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From duncan at NOSPAMrcp.co.uk  Tue Apr 30 08:37:47 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 12:37:47 +0000 (UTC)
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com>
Message-ID: <Xns9200893E7F263duncanrcpcouk@127.0.0.1>

Roy Smith <roy at panix.com> wrote in
news:roy-E96293.07453430042002 at news1.panix.com: 

> Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:
>> The problem with int.chr() is that a lot of people are going to get 
>> confused when they find that "42.chr()" doesn't work (although "42
>> .chr()" does).
> 
> What about "(42).chr"?  Wouldn't that work?  Seems a lot more natural
> than putting a space in there (which I didn't even realize was legal,
> and looks horrible to me).
> 
There are lots of things that would work. '*' is probably the cleanest.
It is just that methods on int have this slight wart. As is clear from the 
followups to my posting, most people find it surprising that the parser 
works that way and (IMHO) it is good to avoid surprising people.

[I'm not sure that came out quite right: I'm sure lots of the people on 
this newsgroup can be surprising, and I don't feel I have to avoid you.]

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From les_ander at yahoo.com  Sat Apr 27 14:58:34 2002
From: les_ander at yahoo.com (les ander)
Date: 27 Apr 2002 11:58:34 -0700
Subject: Removing dupes from a list
References: <mailman.1019853649.15936.python-list@python.org> <Jvjy8.66361$vF6.2030499@news2.tin.it>
Message-ID: <a2972632.0204271058.72ac51cf@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<Jvjy8.66361$vF6.2030499 at news2.tin.it>...
> Alexander Skwar wrote:
> 
> > What's the fastest way to make sure that a list only contains unique
> > entries?
> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560
> 
> 
> Alex


def unique(List):
  return list(dict(zip(List,[None]*len(List))))

les


From hamish_lawson at yahoo.co.uk  Wed Apr 17 13:14:43 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 17 Apr 2002 10:14:43 -0700
Subject: Making a C extension compatible with 2.2
Message-ID: <915a998f.0204170914.66aee851@posting.google.com>

Must anything be done to make a C extension compatible with Python
2.2? I have an extension (not my own) that ran fine under Python 2.1
on Solaris 2.6. I have since recompiled it for Python 2.2, but
sometimes I now get tuple indexing errors - and not necessarily at a
point in my Python code where the loaded extension is actually being
called upon. Running the same code under Python 2.1 is OK. Would I be
right in thinking it may be related to the changes in Python 2.2 for
the new iterator protocol?

Hamish Lawson


From taradino at softhome.net  Sun Apr  7 05:29:45 2002
From: taradino at softhome.net (Johannes Gijsbers)
Date: Sun, 7 Apr 2002 11:29:45 +0200
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net> <MPG.17197f04a3f463e8989680@news.planet.nl> <7xvgb4to7h.fsf@ruckus.brouhaha.com> <ad052e5c.0204062127.68335604@posting.google.com>
Message-ID: <MPG.171a2ff0813857e7989683@news.planet.nl>

Dan Bishop wrote:
> As it shouldn't.  In CPython 2.2, math.sqrt(1e100) is 7 629 769 841
> 091 887 003 294 964 970 946 560 higher than its actual value.  With an
> absolute error this high, who cares about whether there's a fractional
> part?
> 

I'm not sure if anyone would care, but it's quite easy to change the 
function to also accept large numbers:

>>> import math
>>> def isWhole(number):
...     return long(number) == number
...
>>> isWhole(math.sqrt(1e100))
1

Johannes Gijsbers

-- 
talk to me..talk to my machine
talk to me
talk to me


From amuys at shortech.com.au  Mon Apr 15 20:08:36 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 15 Apr 2002 17:08:36 -0700
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <3CBABA71.3040909@geneva-link.ch>
Message-ID: <7934d084.0204151608.271b1737@posting.google.com>

Boris Borcic <borcis at geneva-link.ch> wrote in message news:<3CBABA71.3040909 at geneva-link.ch>...
> Indeed, shouldn't two generators g1 and g2, when combined
> in an expression of the form e.g. "g1 and g2" or "g1 or g2",
> combine meaningfully to become a third generator ? Respectively
> one that returns a successive value of g2 for every value that
> g1 returns, until one of them returns IterationError, 

Ahh you would mean

[yield elem for tupl in xzip(g1, g2) for elem in tupl]

> and one
> that concatenates the values of g1 with the values of g2 ?

[yield elem for tupl in (g1, g2) for elem in tupl]

I personally don't see these as being natural applications of and/or
to lists.
As far as concatenation is concerned, I remain a little surprised that
'+' isn't defined for generators, after all it is for lists and has a
well defined semantic that is obvious.  I have always found the
absense of a multiply method on lists strange, I personally would have
expected it to perform a cartesian product.  Still it's easy enough to
do long hand, so it doesn't really bother me, especially given the
non-intuitive ordering required if you want to handle infintate
generators at all gracefully.

Andrae Muys


From phr-n2002a at nightsong.com  Fri Apr  5 20:49:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 17:49:24 -0800
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>
Message-ID: <7xwuvliokb.fsf@ruckus.brouhaha.com>

Alastair Nicol <calaban.madness at blueyonder.co.uk> writes:
> 1) Can only hosts which can see the interface your program is bound to 
> communicate with your program. In the "normal" case if you bind to 
> 127.0.0.1 for example ONLY your machine can see that interface. Other 
> machines cannot connect to it. Is this a safe / secure assumtion? (not just 
> for localhost)

Normally yes; however, if your app is strictly local and you're
running a *nix system, you may prefer using AF_UNIX (unix domain
socket) instead of an IP socket.  That avoids the whole TCP stack etc.



From martin at v.loewis.de  Tue Apr  2 12:32:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 19:32:13 +0200
Subject: wxPython redefines string.letters?
References: <8351bb33.0204020854.272eaa7b@posting.google.com>
Message-ID: <m3d6xigg6a.fsf@mira.informatik.hu-berlin.de>

jkn at nicorp.f9.co.uk (Jon Nicoll) writes:

> I realise this should go to the wxPython list, and after I've joined
> it will do ;-), but in the meantime:

I guess wxPython calls locale.setlocale, and that changes
string.letters.

Regards,
Martin



From eppstein at ics.uci.edu  Tue Apr  2 21:41:04 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 02 Apr 2002 18:41:04 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <kupu1hple7.fsf@lasipalatsi.fi>
Message-ID: <eppstein-CC3FF6.18410402042002@news.service.uci.edu>

In article <kupu1hple7.fsf at lasipalatsi.fi>,
 Erno Kuusela <erno-news at erno.iki.fi> wrote:

> since 1.5, i find the following changes to the language come
> immediately mind as worthy:
> - nested_scopes
> - string methods
> - unicode (maybe, haven't really used or needed it yet)
> - new-style classes (ditto)
> 
> the following language changes come immediately
> to mind as feeling superfluous:
> - list comprehensions (map/filter)
> - f(*args, **kw) (apply)
> - staticmethod/classmethod
> 
> i haven't decided if this is a good ratio or not, yet.

I came to the language after list comprehensions were already there, so 
maybe that changes the picture for me, but they're one of the things I 
really like about the language -- they don't do anything you couldn't do 
with a loop, but when they can replace a loop I think they can lead to 
much clearer code.

Nested scopes are also good, although it confuses me that they are 
read-only.  The other recent addition that I really like is simple 
generators.

I can't think of a language change that has made me very unhappy, but 
maybe I haven't seen enough of them to judge.  I was strongly opposed to 
the integer division thing when I first heard about it, but Guido 
convinced me that it was the right thing to do.

There's plenty of change I'm neutral or ignorant of, like most of the 
object-oriented infrastructure stuff.  What is staticmethod etc anyway?

So that seems like a very good ratio to me.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From fredrik at pythonware.com  Wed Apr  3 12:04:05 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 17:04:05 GMT
Subject: PyInt_AsLong()
References: <xdu7knpaz87.fsf@leto.informatik.uni-freiburg.de> <EnBq8.29186$n4.5840811@newsc.telia.net> <xduzo0l9gdd.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <9CGq8.36868$l93.7327576@newsb.telia.net>

Ralf Juengling wrote:
> Okay. So the reason to indicate an error by '-1' is efficiency,
> I guess

the function has to return something. it might as well
return a known value...

> (i.e. calling PyErr_Occured() each time is more
> expensive than the prior check 'v == -1') ?

on most computers, yes.

</F>




From sophiec at scripps.edu  Wed Apr 17 22:23:19 2002
From: sophiec at scripps.edu (Sophie Coon)
Date: Wed, 17 Apr 2002 19:23:19 -0700
Subject: New release: PMV, ADT, ViPEr
References: <mailman.1019073817.29018.python-list@python.org>
Message-ID: <3CBE2E17.2939E801@scripps.edu>

Hello,

We are happy to announce the release of the Python-based Molecule
Viewing environment (Pmv), the AutoDockTools (ADT) and the Visualization
Programming Environment (ViPEr) version 1.1alpha. on April 17th 2002
available to download on our web site
:http://www.scripps.edu/~sanner/python/packager.html

An academic distribution and a commercial distribution are available for
download. The version labeled 'commercial' does not contain packages
requiring commercial license. For commercial licenses please contact
Michel Sanner (sanner at scripps.edu).

You can either download the complete release which contains:
 - a python interpreter version 1.5.2
 - python packages developped in the MGL lab (platform dependent and
independent)
 - Third party python packages (platform dependent and independent)
 - Applications : PMV, ADT and ViPEr.
Or only platform dependent packages which contains:
 - python packages developped in the MGL lab
 - Third party python packages (platform dependent and independent)
 This tar files is useful to add support for another platform after
installing a full release

The download "A la carte" is of course still available. This lets you
download individual packages (with the danger to mismatch them !)
You can find more information on our web site
:http://www.scripps.edu/~sanner/python/index.html


Two mailing lists have been created pmv and pmvdev (see README.html file
once you downloaded and installed our packages).

For any question concerning :
      - the installation or PMV please contact Sophie COON
(sophiec at scripps.edu)
      - ADT please contact Ruth Huey (rhuey at scripps.edu)
      - ViPEr please contact Michel Sanner (sanner at scripps.edu)
 Have Fun !
Michel Sanner, Ruth Huey and Sophie Coon


From pj at engr.sgi.com  Thu Apr  4 22:02:27 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 5 Apr 2002 03:02:27 GMT
Subject: "Don't Kill Opera" documentation release
References: <mailman.1017709803.3793.python-list@python.org>
Message-ID: <a8j443$omncm$1@fido.engr.sgi.com>

Could you explain easily what you had to change to the Docs
in order to satisfy Opera 6.01?  As an Opera user, I am
curious what was the bug in Opera that you were working
around (or at least I assume that's the situation).

Oh - Opera support informed me in an email today that the
next Beta should be out any day.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From henrykhecht at my-deja.com  Mon Apr  1 19:16:59 2002
From: henrykhecht at my-deja.com (HH)
Date: Tue, 02 Apr 2002 00:16:59 GMT
Subject: (Help) Embedding in Python 1.5
Message-ID: <slrnaahu3v.21k.id@c1299689-a.saltlk1.ut.home.com>

This seems to be almost a FAQ, but I still can't find a comprehensive
enough answer to get it to work....

I'm trying to embed some Python in some C.  Python version 1.5, as
packaged by Red Hat (not that that should make a difference), and I'm
having a little trouble getting the right linker options.  Basically,
what I've got so far is:

gcc -L/usr/lib/python1.5/config -lpython1.5 -lieee -ldl -lpthread \
		-lm -o testpy testpy.o

which spews a few screenfuls about things from </python1.5/Python.h>
being undefined.  Looking at section 5.6 of "Extending and Embedding
the Python Interpreter", I queried python with:

>>> import distutils.sysconfig
>>> distutils.sysconfig.get_config_var('LINKFORSHARED')
'-rdynamic'

'-rdynamic' means -Wl,--export-dynamic to gcc I believe, though you
will find this nowhere in the gcc documentation.  Unfortuantely,
adding -rdynamic (or -Wl,--export-dynamic or -Xlinker --export-dynamic)
to the above command line seems to do...nothing.

Unfortunately the Python docs don't get any more verbose about this,
and the copy of "Programming Python" I have lying around includes a
very helpful example makefile for Python 1.3, referencing four
libraries that I don't have at all (Modules, Python, Objects, and
Parser).

What am I missing here?  Could someone supply me with a modern (but
not too modern, i.e. 2.x) example makefile, or at least some advice?

-- 
HH


From tim.one at comcast.net  Sat Apr  6 23:27:08 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 06 Apr 2002 23:27:08 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <Xns91E898777C949JamesLLugojcom@216.148.53.83>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKBOLAA.tim.one@comcast.net>

[Tim, suggests James not pick up new releases more often than he
 feels like upgrading]

[James Logajan]
> Your advice may seem reasonable but wasn't quite applicable. I developed
> some Usenet moderation programs on 1.5.2 because I knew the co-moderators
> would be running on Linux (1.5.2 and 2.0), Solaris (1.5.2), and Windows
> (2.1). I didn't have Solaris handy, so by distributing this freeware
> program as a simple zipped file of *.py files, I wouldn't have to
> contend with the near impossible task of freezing my application on all
> the target platforms.  And with luck it could be used by users running
> on Macs.

That's a tough one.  I didn't recommend it in this particular thread, but I
do usually recommend that people with extreme compatibility desires ship the
version of Python they're using along with their .py files.  (I know you
don't agree you're "extreme" here, but in the plain statistical meaning you
are.)  Of course I recommend the same to people shipping Perl, Scheme, Icon
or C apps too (note that Python on Windows ships with the specific release
of the Microsoft C libraries it was tested with -- that's one way we try to
insulate Python's users from Microsoft's x-release C incompatibilities,
although it's only partially effective).

> By choosing Tkinter and 1.5.2 I hoped to minimize version
> incompatibility.

I expect that you did!  Good choice.  Your complaint seems to be that
"minimize" turned out to be strictly larger than 0 <wink>.

> But in the end I had to contend with unexpected forward
> incompatibilities in a  few places, including in sockets (this one
> was well known, still some old code tripped me up), cStringIO's close
> semantics changed, and of course Tkinter problems. The Tkinter issue
> was more vexing than the other issues,

It still is, FWIW:  the Unicode versus Tkinter saga is still unfolding.
Ditto the Unicode versus XYZ story across many languages and many values of
XYZ -- Unicode support is a Huge Change, and especially for projects
predominantly staffed by Americans.  If Python were a commercial project and
could pay users to test in advance of release, I would support doing so.
Instead we did what we could, shipping alpha and beta and release candidates
and begging people to *try* them.  If you didn't pick one of those up to
test your specific cross-release needs, then you missed a great chance to
help yourself.

> since it took a while to surface and when it did it took a while to track
> down and come up with a good solution. I guess you could say I got
> "burned" by it.

Sure, and I'm sorry you did.  At the same time, I know of no way to stop all
cross-version cross-platform surprises short of stopping Python development
cold.

> Sorry, I meant to use "major" here as in "there are new features or
> semantic changes in this release that if you use in your code, it
> wont run at all or as you expect on previous releases".

We do our best to meet that in the three-digit "bugfix releases", but
there's still no absolute guarantee.  Any change, even a pure bugfix, *can*
change some program's visible behavior, and sometimes does.  That's a fact,
and can't be wished away.

> Doesn't matter what you call it, if a 2.1 script can't run at all or as
> expected on 2.0, that is a major change in the _colloquial_ sense.

I expect you phrased that backwards, yes?  If you use a new-in-2.1 feature,
then of course it's not going to run under 2.0.  The other direction is
almost always tractable, although in some cases it does take foresight, or
the active application of hindsight, to pull it off.

> I consider the nature of the changes between so-called "minor" releases
> and their frequency a major problem to _me_.  I don't think you can
> realistically "argue" those problems away. I am hoping that you will try
> to accommodate my experiences into your Python world view before
> brushing them off.

I think you're determined to take offense, James.  I simply know of no
realistic way to get the extreme level of x-release, x-platform invariance
you want, and regardless of language.  If you think you'll be better off in
this respect using C or Lua or Ruby or whatever, I can can only wish you
luck, and assure you we'll be still be here when you give up <wink>.

I don't brush off your experience, but I do dismiss your implications that
this is something we could easily fix for you if only we didn't hold
Python's users in contempt (or whatever it is you really think).  You could
try using Java instead, which has often *claimed* "write once, run
everywhere (and forever)", but in practice I believe Python gets closer to
meeting that than Java has.  I don't brush off your desire to improve that
situation, but it's true that I don't take it as an absolute good -- it's
one of many competing desires, and one of the hardest to achieve.

> ...
> Before I set Python aside as being unacceptably non-portable for a
> scripting language, I _had_ tried to stop at 1.5.2. But I
> couldn't develop open source scripts in such a way that they were
> version independent without either a lot of version checking or
> freezing applications with a specific version of the interpreter
> (which defeats one of the points of open source IMHO).

If you factor out the Tkinter part of your experience, is that still true?
You hit a really tough knot there.

>> When the release rate slows to what you'd like to see, I'd take it as a
>> strong indicator that Python may be entering its death phase.

> The Bourne shell (bsh) and C shell (csh) haven't changed in a
> _long_ time.  Do you consider these scripting languages dead in any
> meaningful sense of the word?

I doubt that it's relevant unless you view them as realistic alternatives to
Python for x-platform programming.  If so, go right ahead and *try* to write
a real app in csh <wink>.  sh and csh get to stay stable because nobody
wants anything new from them anymore -- they do a small job, and at least
Bourne shell does it well.  Languages like awk, Perl and Python took the
"new feature request" burden off their backs, and gave a large measure of
OS-independence to boot.  I class the Unix shells more with ls and expr
(etc) than with programming languages -- they're more like little tools, and
little tools don't have to grow to survive (GNU's prolific intoduction of
dubious new options notwithstanding).





From huaiyu at gauss.almadan.ibm.com  Tue Apr  9 13:47:08 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 9 Apr 2002 17:47:08 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com> <a8slgr$pin$1@nyheter.chalmers.se> <3CB1E745.EDA08705@alcyone.com> <a8sr75$rug$1@nntp6.u.washington.edu>
Message-ID: <slrnab6a8r.dvl.huaiyu@gauss.almadan.ibm.com>

Donn Cave <donn at u.washington.edu> wrote:
>Quoth Erik Max Francis <max at alcyone.com>:
>|
>| Weak typing usually refers to the idea that operations, not the objects
>| themselves, decide as what types to manipulate objects.  Dynamic typing
>| indicates that variables (or bindings) do not have an intrinsic type.
>
>At best that seems confusing to me.  It would seem to say that
>languages that use type inference would be weakly typed - they
>sure let the operations decide object type.  In Objective Caml:
>
>  # let oneup x = x + 1;;
>  val oneup : int -> int = <fun>
>
>  (i.e., because parameter x appears in the expression x + 1, x must
>   have type int, and the function result must also be of type int.)
>
>But Objective Caml's typing is hardly "weak" in any useful sense of
>the word, rather it's quite rigorous.

The types are uniquely associated with objects.  They are just not specified
explicitly.  The system infers, but not decides, what they are, using clues
from the operations.

The basis of type inference is that each object has exactly one type.

In contrast, in a weakly typed system, the _same_ object can have multiple
types.   In a hyperthetical example

		4 add 2  ===> 6
		4 concat 2  ===> 42

In such systems type inference in the style of functional languages is
impossible, since the type inferred in one context is not guaranteed to be
valid in another.

So I'd venture a claim that any language that is capable of type inference
is necessarily strongly typed.

Huaiyu


From ako at nd.edu  Fri Apr 19 11:16:12 2002
From: ako at nd.edu (Alice NgarKit Ko)
Date: Fri, 19 Apr 2002 10:16:12 -0500
Subject: os.system() with redirection
Message-ID: <3CC034BC.FAF56EBB@nd.edu>

Hi,
I want to call a program with os.system().  The program redirects output
to file only through >& outputfile.   Here is my script (temp.py).

line = 'protomol temp.conf.PME.Ewald >& run.temp.PME.Ewald'
os.system(line)

>python temp.py
>sh:  run.temp.PME.Ewald:  bad number

Any idea on how to fix it?  Thanks





From martin at v.loewis.de  Thu Apr  4 14:50:48 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Apr 2002 21:50:48 +0200
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com>
Message-ID: <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de>

seb_haase at yahoo.com (Sebastian Haase) writes:

> Solution 1)  -->> BAD!! <<-
[Solution 2)]
> =====
> The REAL 
> Solution !  --> "we" are not really interested in debugging the python
> lib !
> ========    ==> we only want to debug _our_ own C/C++ code!!!

Solution 3) Don't define _DEBUG in your debug builds.

Regards,
Martin


From aleax at aleax.it  Tue Apr  2 17:01:57 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 22:01:57 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <3CAA03E4.A79BA593@noaa.gov> <Xns91E4EF89BD4EDcliechtigmxnet@62.2.16.82>
Message-ID: <pTpq8.53356$S52.1844974@news2.tin.it>

Chris Liechti wrote:
        ...
>> Thinking in terms of something vs. nothing is an excellent way to
>> explain how Python treats these things, but I still have a problem with
>> it. Who is to say that an empty list is nothing?
> 
> you can guide that in your classes with __nonzero__ (or __len__). if you

You can do that (subclass list and add __nonzero__ -- don't muck
with __len__, of course) but it's hard to see what you will use to decide
that your list "is something" or "is nothing", except the issue of whether
it has or lack contents.

> don't think of 'find' as a boolean operation. its result is an index, not
> a truth value. maybe it would be clearer if it returns None or raises an
> exception if nothing was found but its as it is now (the strstr C function

Guido's time machine has beaten you to this one...:

>>> 'carciofo'.index('cio')
3
>>> 'carciofo'.index('ciao')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ValueError: substring not found in string.index

If you want an exception on failure, somestring.index already does it, see.
somestring.find is just like somestring.index BUT doesn't raise an 
exception on failure -- rather, it returns -1 in that case.

However, Python doesn't use "success" (no exception) or "failure"
(exception) to steer its if -- I've already suggested looking into Icon,
which DOES, but you'll never make Python into Icon, of course.

> yes. but a real bool could be introduced as module for now (for the RPC
> and marshalling people that need one) i think its not needed otherwise

A standard library module sounds good.  Or maybe it could go into operator.


Alex



From James_Althoff at i2.com  Wed Apr 24 15:17:45 2002
From: James_Althoff at i2.com (James_Althoff at i2.com)
Date: Wed, 24 Apr 2002 12:17:45 -0700
Subject: PEP 276 -- What else could iter(5) mean?
Message-ID: <OF56F34B33.93B51A68-ON88256BA5.00694648@i2.com>

[John Roth]
>...
> I'm against the proposal, simply on the basis that I don't find
> 5.iter to have a blindingly obvious meaning. It's ok if you explain
>...

[Charles Hixson]
> And to me the meaning *is* blindingly obvious ... and different!
>
> to me 5.iter should return an unlimitedly long string of 5's.

Perhaps one might even consider

    for i in <5,5,5,...>:

to be "blindingly" *useful* compared to

    for i in <0,1,2,3,4>:

<wink>

Jim





From gerhard at bigfoot.de  Sat Apr 13 15:13:16 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 13 Apr 2002 19:13:16 GMT
Subject: Post-installation with distutils windows installer
References: <1104_1018717651@news.netcom.no>
Message-ID: <slrnabh0qc.5jr.gerhard@lilith.my-fqdn.de>

Ole Fostad wrote in comp.lang.python:
> Hi everyone!
> 
> I wondered if there is a way to run a post installation script when using the distutils 
> Windows installer?

AFAIK not yet, but Thomas Heller was working on a possibility to do
so.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From goodger at users.sourceforge.net  Sun Apr 21 21:23:47 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Mon, 22 Apr 2002 01:23:47 GMT
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au>
Message-ID: <B8E8DEB0.21E3D%goodger@users.sourceforge.net>

Neville Franks wrote:
> I still can't get ndiff to work. If I enter: 'difflib.ndiff' I get
> 'AttributeError: 'difflib' module has no attribute 'ndiff'

Tim Peters' ndiff.py is a script in the tools/scripts directory.
Originally, it was self-contained, but at some point Tim migrated the
SequenceMatcher class into the difflib module.  I extracted the rest of the
generic code and turned it into the difflib.Differ class and difflib.ndiff()
convenience function, Tim converted it to use generators, and it was
released with Python 2.2.  You're running Python 2.1.2, so difflib.ndiff
doesn't exist yet.  You'll have to upgrade to get the full functionality.
You can't use the Python 2.2 difflib module in Python 2.1, because it uses
the generator feature, new to Python 2.2.  However, you can get a Python
2.0/2.1-compatible version from
http://docutils.sourceforge.net/test/difflib.py; this is the module after I
added Differ and ndiff(), but before Tim converted it to use generators.

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
  - The Go Tools Project: http://gotools.sourceforge.net/



From zaka07 at hotmail.com  Thu Apr 25 10:29:27 2002
From: zaka07 at hotmail.com (Rad)
Date: 25 Apr 2002 07:29:27 -0700
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com>
Message-ID: <ad381f5b.0204250629.5144d196@posting.google.com>

Thanks for all the suggestions and ideas, I'll try to answer to all
your questions in this one post.

It is a longish term project but after a few initial data extractions
I'll have some time to re-implement the procedures if there is a need
to do so.
At this time I don't know for sure how many unique ID's are going to
be, not less than 15 million I guess.
Don't have any options on the way data is delivered, it's going to
come on tape/s and all relevant files will be in fixed width format. 
I am OK with that, shouldn't have any problems reading it in and
storing it on the HD unless DDS starts clicking :(.  I have four 73 GB
SCSI disks on the computer that will run win2k, 2 GB of RAM, dual xeon
processor and three power units.

I agree that the actual work is not that complex even though there are
quite a few other bits involved in the project that I thought were not
relevant to mention in my earlier post, I also agree that it could be
done in a relational database but if the files were smaller, for
example Access (which I know quite well), I also know some SQL, but I
never used MySQL, which is perhaps a possible route to take.  However,
as I just mentioned I don't know anything about MySQL (except that one
has to pay for it) + having been learning Python in the last few weeks
I kind of like it + I was encouraged by your posts and I'm now pretty
convinced that Python2.2 is the way to go. I won't have to access data
randomly, just pull out the data fitting certain criteria from
original 4 files and create resulting files.  I might decide
(depending on the size) to put those into a database though.
I don't understand how I can interface Python to a database?  

All four files have dates in them but they do come in YYYYMMDD format
and I was planning to use string comparisons, same for the rest of
data,I was thinking to treat it all as strings.

My initial plan is to either split files into smaller files that could
fit into the memory or read chunks of data from original big file
using file_object.redlines(some_size_that_ can_ fit_ in_ the_ memory).

I'm stuck with it (the project) and I'm looking forward to the
challenge.
Hope I don't regret saying this publicly!?

I believe that initial filtering will reduce file size for at least
50% so I'll do some research on awk (for windows).  I guess there is
info about it on the Internet.

Thanks again for all your suggestions and ideas.
Probably "talk" to you soon.


From gh_pythonlist at gmx.de  Wed Apr 10 07:43:40 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 10 Apr 2002 13:43:40 +0200
Subject: Help: Python for a big commercial project?
In-Reply-To: <F194tIpqqJgSt51NDa200002509@hotmail.com>
References: <F194tIpqqJgSt51NDa200002509@hotmail.com>
Message-ID: <20020410114339.GB562@lilith.my-fqdn.de>

* Cereal Muesli <cereal_con_fibra at hotmail.com> [2002-04-10 13:24 +0200]:
> - First of all; do you think that Python is suitable for this
>  kind of "serious, big" project?

Yes.

> - The core DW system and ETL logic runs under a HP-UX 11 system.

Dunno much about HP-UX, but I did once manage to build Python 2.0 for a
HP-UX 10 (or was it 9) system. It was certainly more pain than on other
platforms.

> - Installation: How easy or how hard is the installation of
>  Python for HP-UX 11? I don't want to disturb sys admins more
>  than necessary; do they have to compile everything? Do they
>  need to install GCC to do that? Are there any reliable
>  full-featured pre-built binaries for this platform available to
>  "download-and-install-in-a-minute"?

I don't think there are any prebuilt binaries for most Unixen, but a
competent sysadmin will compile once and move the resulting binary
tarball to the other machines. Or even use rdist or install Python on a
NFS mounted directory.

> - Finally: Which version of Python should we use?

2.2.1, definitely.

> it seems that 1.5.x still rules under *NIX, doesn't it?

No. 1.5.2 doesn't rule at all, it's only that one (!) certain Linux
distribution still uses it by default for very stupid reasons.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 8.3 ?C      Wind: 8.1 m/s




From aleax at aleax.it  Sat Apr 27 16:15:07 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 20:15:07 GMT
Subject: slices: foo[i:j:k] => TypeError
References: <mailman.1019856470.541.python-list@python.org> <aCky8.66523$vF6.2038703@news2.tin.it> <mailman.1019924208.19241.python-list@python.org>
Message-ID: <fFDy8.70854$vF6.2116208@news2.tin.it>

Roman Neuhauser wrote:
        ...
>     Ok. What I don't understand then is why is extended slicing
>     described among primaries in the language reference. And I'm not

Because the parser needs to know about it (to support Numeric).  Indeed,
it's not a syntax error that you're getting.

>     talking about a development version of the manual, I'm talking about
>     the 2.1 version.

I think it was introduced in 1.5.2.

>     What's the expected date for the builtin types to accept slice
>     objects?

No idea.  The original intro was before PEPs, so there's not even a
PEP to refer to, I believe.  You may want to ask on Python-Dev --
if the BDFL OK's it for 2.3, I can write the code, if you or some other
volunteer can handle the rest of the patch submissal process (sorry,
can't volunteer more time than what it takes to write & debug the code).


Alex



From aahz at pythoncraft.com  Mon Apr  1 14:22:28 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 14:22:28 -0500
Subject: Access to main namespace from imported module?
References: <re8haugrbcchbephckl9t1p1s27tvg18hr@4ax.com>
Message-ID: <a8ac1k$cl5$1@panix1.panix.com>

In article <re8haugrbcchbephckl9t1p1s27tvg18hr at 4ax.com>,
Dale Strickland-Clark  <dale at riverhall.NOTHANKS.co.uk> wrote:
>
>Just curious, like, but is it possible for an included module to
>reference the namespace of the outer-most module, "__main__".

Okay, now that you know how to do it, tell us why you think you need it
so we can tell you a better way to do it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From dale at riverhall.NOTHANKS.co.uk  Sat Apr 13 11:18:08 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Sat, 13 Apr 2002 16:18:08 +0100
Subject: wich odbc is stable ?
References: <20020413141754.6d14e50d.wilk@flibuste.net>
Message-ID: <rrigbugftrojvmcoop68pkpid08mdb4f2t@4ax.com>

William Dod? <wilk at flibuste.net> wrote:

>hi,
>
>I need to connect to an MSAccess database for simple query (no lock, no
>transaction), so i would like to know if the standard odbc provided by
>win32extension is enough stable even if it's written to be no supported...
>Or must i use callDll or ado ? 
>
>thanks to read my so bad english writing

The ODBC module in win32all is OK for most jobs. I have heard that it
is buggy but my limited use hasn't found any.

A more robust ODBC module, called mx.odbc, can be found here:
http://www.egenix.com/files/python/
There is a licence fee for comercial use.

ADO also works well but is somewhat slower than ODBC for most common
tasks.

--
Dale Strickland-Clark
Riverhall Systems Ltd



From philh at comuno.freeserve.co.uk  Sun Apr 28 16:04:11 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 28 Apr 2002 21:04:11 +0100
Subject: [OT] What is Open Source?
References: <mailman.1019858688.3528.python-list@python.org> <7x1yd2qc4q.fsf@ruckus.brouhaha.com> <slrnacm45g.jk3.jajvirta@sirppi.helsinki.fi>
Message-ID: <slrnacoldr.nvo.philh@comuno.freeserve.co.uk>

On 27 Apr 2002 20:57:20 GMT, Jarno J Virtanen <jajvirta at cc.helsinki.fi> wrote:
>26 Apr 2002 16:21:57 -0700 Paul Rubin wrote:
>> So we'l have to start issuing PEP's with the disclaimer "I found this
>> PEP in my grandma's old trunk in the attic, dated April 1, 1920, so
>> it is out of copyright now."
>
>And if "they" start questioning how come your grandma cared about
>booleans in Python, you could refer to the time machine and come up with
>a reasonable story.  And if even 10% of the stuff they show on TV about
>court practice in United States is true, the jury would swallow the
>story whole.

Hah, just tell 'em you've been eating Twinkies, they'll let you off.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From brian at sweetapp.com  Tue Apr 16 18:09:02 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 16 Apr 2002 15:09:02 -0700
Subject: Unicode name questions
In-Reply-To: <15548.40527.385680.204020@12-248-41-177.client.attbi.com>
Message-ID: <009d01c1e593$52d74f40$445d4540@Dell2>

Skip,

>     Brian> Your spelling of the first two isn't quite right (at least
>     Brian> according to the Unicode guys). Remove the "B" from
"LAMBDA".
> 
> "Lambda" has been spelled with a "b" as long as I can remember.  I see
> both "LAMBDA" and "LAMDA" in the comments in the encodings modules
though.

You aren't disagreeing with me; you are disagreeing with the Unicode
committee :-) Did you look at the Unicode charts?

> I suspect perhaps Python's codecs just haven't caught up yet.

If we are talking about the same characters, then Python seems fully
inline with the published Unicode names. 

"GREEK SMALL LETTER THETA SYMBOL" could be (and all work with Python):

GREEK SMALL LETTER THETA
GREEK THETA SYMBOL
GREEK CAPITAL THETA SYMBOL

For the Greek names, see:
http://www.unicode.org/charts/PDF/U0370.pdf

Cheers,
Brian





From pan-newsreader at thomas-guettler.de  Sat Apr  6 04:23:11 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Sat, 06 Apr 2002 11:23:11 +0200
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org>
Message-ID: <pan.2002.04.06.11.23.11.491490.3709@thomas-guettler.de>

On Sat, 06 Apr 2002 10:15:33 +0200, David McInnis wrote:

> I need to alter the document properties of Word DOCS that our users
> upload to our Web site.  Does anyone know how to do this?  (I am python
> installed on a Linux box).

You could do this with python and COM if you would use windows with
MS-Office installed.

If you have a Windows server around you could use some RPC method
to do this form Linux.

Wvware can change your MSWord files to HTML or an other format on unix.
But it can't change the MSWord file as far as I know.

BTW: I think it is called "I have python installed". But I am not a
native speaker

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From jkraska at san.rr.com  Wed Apr  3 22:56:23 2002
From: jkraska at san.rr.com (Courageous)
Date: Thu, 04 Apr 2002 03:56:23 GMT
Subject: OK
References: <e8d5abee.0204031816.2bfec035@posting.google.com>
Message-ID: <c7knauo4dqpsipk3rdhg5jhcaqvkna4pat@4ax.com>

On 3 Apr 2002 18:16:14 -0800, huangcg at 163.net (huangcg) wrote:

>Is python out of fashion?

To the contrary, it's becoming trendy. I can't really say that
it ever _was_ "in fashion," except in the present. I therefore
find your question rather strange.

C//



From phil at river-bank.demon.co.uk  Sun Apr 14 14:04:36 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sun, 14 Apr 2002 19:04:36 +0100
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
References: <oOjt8.8815$vF6.254710@news2.tin.it>
Message-ID: <3CB9C4B4.DC454AA8@river-bank.demon.co.uk>

Alex Martelli wrote:
> 
> Studying PyQt (latest releases of everything), I started noticing
> segmentation faults as I quit my application and eventually whittled it
> down back to...:
> 
> #!/usr/bin/env python2.2
> 
> import sys
> import qt
> # from qt import *

The fix has now been applied to CVS. I will probably release 3.2 in
about a week.

The problem was related to not being able to control the order in which
objects in a Python dictionary are deleted when the whole dictionary is
cleared.

Phil




From martin at v.loewis.de  Wed Apr  3 02:09:22 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Apr 2002 09:09:22 +0200
Subject: C Extension
References: <mailman.1017806174.1666.python-list@python.org>
Message-ID: <m3wuvpntr1.fsf@mira.informatik.hu-berlin.de>

Eric Merritt <cyberlync at yahoo.com> writes:

> I hope you guys dont mind another stupid question. I
> have created a new C extension to python (Its my first
> one so I expect issues). This extension contains two
> types. In general these types seem to work just fine.
> However when I enter a value of one of these types on
> a line by itself python segfaults. 

I recommend to run Python in a (C) debugger. This will give you much
clearer information as to why it crashes than guessing.

Notice that the crash does not necessarily come from Python printing
something. It could also come from Python deallocating your object.
Even more involved, it could come from Python doing something
completely different, you your extension managed to corrupt central
data structures (such as the malloc heap) somehow.

> At first I thought it was the repr function, or perhaps there string
> function. But I stuck a couple of print strings there to verify that
> and those dont seem to be getting called.

Then something is clearly wrong - either tp_repr or tp_str will be
called, unless you've implemented tp_print (which you don't need to).
If those *aren't* called, this is an indication that it either crashes
before getting that far, or that you managed to get out of synch in
your type object - double-check that every function is in the right
slot (for this, it really helps if you don't need to cast the function
pointers).

Regards,
Martin


From harisri at bigpond.com  Mon Apr 22 23:08:29 2002
From: harisri at bigpond.com (Srihari Vijayaraghavan)
Date: 22 Apr 2002 20:08:29 -0700
Subject: stdin - seek() issue on Linux.
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com> <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com> <aa1bqb$lek$1@nntp6.u.washington.edu>
Message-ID: <284b44c2.0204221908.1280cfc3@posting.google.com>

Hello,

> Quoth "Terry Reedy" <tejarex at yahoo.com>:
> | "Srihari Vijayaraghavan" <no at spam.please> wrote in message
> | news:WMTw8.46875$uR5.106037 at newsfeeds.bigpond.com...

> |> I am a new-bie python user. I have Python 2.2.1 installed on a Linux
> |> computer. I get the following error message when I try to seek() in the
> |> stdin, like:
> |> #cat /var/log/messages | python test.py
> |> (where /var/log/messages is over 1MB in size)
> |>
> |> Traceback (most recent call last):
> |>   File "test.py", line 3, in ?
> |>     sys.stdin.seek(-512, 2)
> |> IOError: [Errno 29] Illegal seek
> |
> | By 'cat'ting the random access disk file, you turned it into a
> | forward-access-only stream of bytes.  You definitely cannot seek
> | backwards (and maybe not forward - don't know).  Instead, open() the
> | file and use the resulting file object:
> |
> | msgs = open("/var/log/messages", 'r')
> | msgs.seek(-512, 2)
> | ...
> 
> Right, though it can be fixed without changing test.py at all -
> "python test.py < /var/log/messages" will work fine.  This idiom
> ("cat file | ...") used to regularly receive a UUOC award, "Useless
> Use Of Cat", from notorious Perl hacker Randall Schwartz when I
> followed comp.unix.shell a few years back.
> 
> Of course if the real program uses some filter, instead of "cat",
> then there's no solution - you can't seek in a pipe, so the answer
> has to be "don't do that".
> 

Thanks for both of your responses. Very helpful.

My mistake, I haven't given the complete context. The above mentioned
method doesn't throw exceptions in Windows using the same Python
version (neither does it print anything, which is the ultimate problem
I would like to be solved).

Since I have already coded it in 'UUOC' :-) style, I thought it should
work even in Linux. But it looks like a wrong practice.

Are there any better method of redirecting the stdout of a program to
stdin of my program in both windows and linux? (specifically I should
be able to seek() on it)

Here is a sample program, where I am trying to use a simillar (bad)
idea:

### start
#!/usr/bin/python

# A python program to print tail portion of a file/stdin(not quite
yet)

import sys, getopt, xreadlines

def usage():
    print '''
    Usage: python tail.py -n<lines> myfile.txt yourfile.txt...
    (Where -n option is to specify the number of lines to be printed,
    and it defaults to 10 lines if not specified)

    If you do not specify any files, script will read from the
standard input.
    '''

def tail(stuff, lines):
    '''prints the specifed number of lines (if possible) from a given
file, stuff'''

    #print stuff.tell()
    stuff.seek(0, 2)
    #print stuff.tell()
    number_of_lines_read = 0

    while number_of_lines_read < lines:
        size = stuff.tell()
        if size >= 512:
            stuff.seek(-512, 1)
            position = stuff.tell()
            data = stuff.read(512)
            stuff.seek(position)
            number_of_lines_read = number_of_lines_read +
data.count('\n')
        elif size > 0 and size <= 511:
            stuff.seek(-1, 1)
            position = stuff.tell()
            data = stuff.read(1)
            stuff.seek(position)
            if data == '\n':
                number_of_lines_read = number_of_lines_read + 1
        elif size == 0:
            break

    if number_of_lines_read > lines:
        difference = number_of_lines_read - lines
        while difference:
            stuff.readline()
            difference = difference - 1

    #print stuff.tell()
    
    for everyline in xreadlines.xreadlines(stuff):
        print everyline,


def main():
    
    try:
        opts, args = getopt.getopt(sys.argv[1:], 'n:h')
    except:
        usage()
        sys.exit()

    # let's default to 10 lines like GNU 'tail' utility
    lines = 10

    for opt, arg in opts:
        if opt in ('-n',): lines = int(arg)
        if opt in ('-h',):
            usage()
            sys.exit()

    if len(args) == 0:
        stuff = sys.stdin
        tail(stuff, lines)
        stuff.close()
    else:
        for everyfile in args:
            one_file = open(everyfile, 'r')
            tail(one_file, lines)
            one_file.close()


if __name__ == '__main__': main()
### End

If you enable the commented 'print' statements, you will know that
Python seek() successfully to various portions of the given file, for
eg, stdin or a file.

Thanks for your time.

Hari,
harisri at bigpond.com


From tim at vegeta.ath.cx  Tue Apr 30 17:38:50 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 30 Apr 2002 21:38:50 GMT
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net> <mailman.1020118433.15459.python-list@python.org> <Xns91FFB1BDEE1BCRASXnewsDFE1@130.133.1.4>
Message-ID: <slrnacu417.pma.tim@vegeta.ath.cx>

Philip Swartzleonard graced us by uttering:
> James J. Besemer || Mon 29 Apr 2002 03:11:25p:
> 
>> 
>> Vincent Foley wrote:
>> 
>>> What's the Python equivalient of Perl's "a"++; ?
>> 
>> You can't get there from here.
>> 
>> I.e. there is no practical equivalant.
>> 
>> Including it would allow deadly side-effects within an expression and
>> thus would percipitate the end of the world, or at least mankind.
>> 
>> Regards
>> 
>> --jb
> 
> But if you just want to increment something, use 'foo+=1'. Exactly the
> same, except one more character and only works on a line of it's own
> (avoiding above said doom... man pointer-walking funcitons in C are
> nuts with that... =)

Well, yes and no. The type being incremented is a string, not an
int. For example:

    # perl
    $a = 10;    $a++;
    $b = "m";   $b++;
    $c = "az";  $c++;

    print $a, "\n";     ->  11
    print $b, "\n";     ->  "n"
    print $c, "\n";     ->  "ba"

Many question the usefulness of this perl feature, but it's come in
handy many times when dealing with alphanumerics with a limited
fieldwidth (Win32 FAT16/32 filenames, to name one).

I don't believe there's a built-in equivalent in python, but it's always
possible to write it.

HTH
Tim Hammerquist
-- 
Wow, I'm being shot at from both sides.  That means I *must* be right.  :-)
    -- Larry Wall in <199710211959.MAA18990 at wall.org>


From jbublitzNO at SPAMnwinternet.com  Mon Apr 29 01:08:17 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Mon, 29 Apr 2002 05:08:17 GMT
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net>
Message-ID: <3CCCD547.6090405@SPAMnwinternet.com>

Tom Chance wrote:
> I've just started out with Python, having been programming Perl for a 
> few years. I've been trying to get an app I downloaded a while ago 
> working without success until I found it couldn't import gnome.ui...

> I duly downloaded every RPM from the SuSE server I could find to do with 
> python, and tried again... no luck. I can't quite figure out what's 
> wrong. If I try "import gtk" or "import gnome" or "import kde" or 
> "import qt" I just get errors telling me I don't have modules by those 
> names.

> An rpm query for python gives the following results:

> tom at tigger:~> rpm -qa | grep py
> gnome-python-1.4.1-127
> pygnome-libglade-1.4.2-2
> python-tkinter-2.1.1-78
> pygtk-libglade-0.6.9-2
> pygnome-1.4.2-2
> python-tk-2.1.1-78
> python-gtk-1.4.1-83
> python-2.1.1-45
> pyxml-0.6.6-68
> python-nothreads-2.1.1-78

> So from that I would have thought it would work fine, but it doesn't. 
> I've also got the KDE bindings for KDE3.0 installed, and I can't seem to 
> get them either.

> I'm running SuSE 7.3 btw, if anyone else uses the same bindings on the 
> same distro...

> Any ideas? I'm at a dead end here...

First, the currently released PyKDE2 bindings only support KDE3.0beta2, 
or KDE2.x.x, not 3.0 final (that's about a week or two away)  If you've 
installed KDE3.0 final release on SuSE 7.3, you'll have to wait for a 
compatible set of PyKDE2 bindings.

I do PyKDE2 (the bindings for KDE), and I do them on SuSE7.1/7.3 
(depending on the version). Find your python directory (/usr/lib/python 
is normal for SuSE, although that should link to something like 
/usr/lib/python2.1). Next look for the site-packages subdirectory - if 
you have PyKDE2 installed, there should be a number of libk*cmodule.so 
entries there (somewhere between 9 and 11 different names), as well as 
some corresponding .py files.

Next, at the command line, open the Python interpreter and do:

 >>> import sys
 >>> print sys.path

You should see 'site-packages' included in the list that prints out.
(Ctrl-D to exit)

If that's all OK, go to the PyKDE2 directory, find the examples 
subdirectory, and try the example programs. They all have a .py 
extension, so to run them, enter 'python pykless.py' for example.

Things should be similar for PyGTK and some of the other programs you 
listed.

PS: The import statement(s) for KDE depends on which module(s) you want 
to use, for example:

from kdeui import KEdit
from kfile import KFileDialog
from kdecore import KAccel, KStdAccel, KCmdLineArgs

etc. The module names correspond to the KDE library names.You also need 
to have sip and PyQt installed, but I assume you do, since PyKDE2 won't 
build and install without them. Make sure, however, that they all match 
in version (latest is sip3.1, PyQt3.1, PyKDE2-3.1 - the 3.1 is the *sip* 
version *not* the KDE version).

Jim



From emile at fenx.com  Sat Apr 13 15:37:30 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 13 Apr 2002 19:37:30 GMT
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com>
Message-ID: <_N%t8.25483$CH1.22269@sccrnsc02>

<bvdpoel at uniserve.com> wrote in message
news:3CB884C7.EE7E04D at uniserve.com...
>
> Yeah, I thought this would be easy...
>
> I have a list like this:
>
> [ [0, x, [0, y, [4, n] ...]
>

Could you provide a better example of the data?  The mismatch of four
open brackets with two close brackets doesn't give me enough info to
interpret the pattern.


--

Emile van Sebille
emile at fenx.com

---------



From aries at bitstream.net  Tue Apr 16 02:14:08 2002
From: aries at bitstream.net (linux_on_wood)
Date: 15 Apr 2002 23:14:08 -0700
Subject: py script web page problem ??
Message-ID: <61fe942.0204152214.3cbe6c0a@posting.google.com>

Hey all,

ANY IDEAS, HELP , WHY  has a problem ???????????

I have tried this on (non-web server pc's)
my linux box and my win98 box both have python installed.
tried it with netscape 4.  and ms-IE 5.0
the book shows it should work

this html form renders properly and I dont have a 
problem with it...............but my python script
that gets called --friends1.py DOES NOT render the
html, it just shows the source code as text.
-------------------------------------

html and body tags and etc here........
<!-- form start -->
<a href="friends1.py">activate python script</a>

<FORM action="friends1.py" method=get>

<CENTER>

<p><p><br>
<!-- the radio's -->
<b>Enter your Name:</b>
<input type=text name=person size=20>
<p><br>
<input type=radio name=howmany value="10">10
&nbsp;&nbsp;&nbsp;&nbsp;
<input type=radio name=howmany value="20" checked>20
&nbsp;&nbsp;&nbsp;&nbsp;
<input type=radio name=howmany value="30">30
&nbsp;&nbsp;&nbsp;&nbsp;
<!--submitbelow-->
<BR><p><BR><FONT color=red>Look over the checkboxes, if everything looks
ok then click the Query button.</FONT>
<INPUT type=submit value="Go HTML Python script">

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



my python script code is:    friends1.py  as chmod 755 on linux box

#! /usr/bin/env python
# program to process html form
# by Joseph Franzmeier Apr 14 2002

import cgi

reshtml = '''Content-Type: text/html\n
<html><head><title>
The Friends Python CGI Demo (dynamic screen) !!
</title></head>
<body><h3>Friends list ofyour name is: <i>%s</i></h3>
Your name is: <b>%s </b><p>
The number of friends you have is: <b>%s</b> total
</body></html>'''

form = cgi.FieldStorage()
who = form['person'].value
howmany = form['howmany'].value
print reshtml % (who, who, howmany)

#print_environ_usage()

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

This script code looks very correct to me but it
gets called --friends1.py DOES NOT render the
html, it just shows the source code as text.

ANY IDEAS, HELP , WHY  has a problem ???????????

Thanks,
Joe F.


From starx at pacbell.net  Mon Apr  8 00:15:02 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 04:15:02 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018222875.17263.python-list@python.org> <3CB0ECEB.E9A3C179@alcyone.com>
Message-ID: <Xns91E9D8599C9D0RASXnewsDFE1@130.133.1.4>

Erik Max Francis || Sun 07 Apr 2002 06:05:47p:

> Arthur Siegel wrote:
> 
>> But the whole thing risks coming off as a loyalty
>> test of some sort.
>> 
>> Does it bother you that Paul Rubin is being
>> "classified" by someone as close in to the
>> inner circle as Aahz.
> 
> No, I haven't.  As a matter of fact, I have no idea what you're talking
> about.
> 

FWIW, neither do I, and i've been following the PEP 285 aftermath postings 
as well as anyone, i think...

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From martin at v.loewis.de  Wed Apr 24 03:17:21 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Apr 2002 09:17:21 +0200
Subject: XML, Python, XML-SIG, documentation, PyXml, 4
References: <aef33705.0204231902.40694af4@posting.google.com>
Message-ID: <m3znzth8fy.fsf@mira.informatik.hu-berlin.de>

andrew at acooke.org (andrew cooke) writes:

> The web pages at http://pyxml.sourceforge.net/topics/ seems to be out
> of date.  In particular, the download talks about 0.6.6 when at
> http://sourceforge.net/projects/pyxml there's a version of PyXml at
> version 7.0 from December last year.  

No, that version is 0.7.0. Using 0.6.6 is still reasonable. I don't
think the Web pages are out of date.

> Which of these packages should I use, and what documentation will help
> me, if I want to:
> 
> - automatically generate a tree from the XML file (I believe this is
> related to the DOM)

Yes, you can use the DOM for that. There are a number of alternative
DOM implementations: Python 2 comes with a package called "minidom";
PyXML has an implementation called "4DOM".

> - supply a class to some (already existing) tree walking code that
> does local rearrangements of the XML data

That is not really supported; you can try the traversal interface of
4DOM, but few people use it, so expect to run into problems.

> - automatically generate an XML file from the modified tree

All DOM implementations support this in some form.

> Also, I know that it's boring updating web sites, and that this is all
> a volunteer effort, but the current XML pages are not very useful.  It
> would be less confusing to simply point people to the mailing list
> archives, rather than to incomplete HOWTO documents that were last
> changed years ago.  

On http://pyxml.sourceforge.net/topics/, there *is* a link to the
mailing list archives. The HOWTO document, to my knowledge, is still
correct (even if admittedly incomplete).

> Perhaps the links to the sourceforge pages would be useful too.

Not sure what pages you are talking about here: The pages you were
complaining about *are* on SourceForge.

Regards,
Martin



From mwh at python.net  Wed Apr 17 10:41:09 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 14:41:09 GMT
Subject: 2.0*4 ==> Bus error on IRIX 6.5
References: <mailman.1019052034.30806.python-list@python.org>
Message-ID: <lkk7r6qt86.fsf@pc150.maths.bris.ac.uk>

Steve ROBBINS <stever at bic.mni.mcgill.ca> writes:

> I just built python 2.1.3 on an IRIX 6.5 machine (Origin 200) using
> MIPSpro compilers (Version 7.30) and the test suite fails.

Compile Objects/floatobject.c without optimization.

> It turns out that multiplying a float by an integer causes a
> bus error:
> 
>   $ ./python
>   Python 2.1.3 (#2, Apr 17 2002, 09:37:30) [C] on irix646
>   Type "copyright", "credits" or "license" for more information.
>   >>> 2.0*4 
>   Bus error (core dumped)
> 
> 
> [Oddly, it works if you multiply two floats or two integers]
> 
> From the core dump, I see that the fault happens here:
> 
>   #0  0x1004fd7c in float_mul () at floatobject.c:382
>   382             CONVERT_TO_DOUBLE(v, a);
> 
> 
> I looked on the python.org web pages, FAQs, etc and didn't
> find any similar problem reported.

I find this slightly hard to believe; it's certainly come up many,
many times before.  If true, something should be added to the FAQ.

> I couldn't find the way to search the python bug tracker,

Go to http://sf.net/projects/python, click on "bugs".  Then there
should be a text area somewhere you can type in together with a button
labelled "search".  You can tell I didn't design this website...

> so I had a quick look at the summary lines.  There was one about
> alignment that caught my eye (#472568) but the workaround suggested
> there does not ameliorate the situation.

This will (probably) be because previous reports of this problem have
been closed as "3rd Party".

> Help?

Hope this did.

Cheers,
M.

-- 
40. There are two ways to write error-free programs; only the third
    one works.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From peter at engcorp.com  Wed Apr  3 01:19:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 03 Apr 2002 01:19:38 -0500
Subject: Zope internal db or MySQL?
References: <Gswq8.20256$0T6.6662133@e3500-atl1.usenetserver.com>
Message-ID: <3CAA9EFA.77961D4C@engcorp.com>

Robert Oschler wrote:
> 
> I'm a Zope newbie.  Would it be better for me to initially set up my site
> using Zope's internal database, and switch to MySQL later?  Or should I just
> bite the bullet and setup MySQL initially.  My database needs won't be very
> heavy, maybe a hundred transactions a day.

(There is a Zope mailing list that may be better for such questions.)

ZODB works well.  Why would you want to make extra work for yourself?
Set it up with the standard stuff, work with it and see what you 
get.  If it isn't "good enough" (you don't give any criteria), you
can always change later once you know more.

Better to learn things one at a time, than to try to do everything at
once.

-Peter


From mertz at gnosis.cx  Mon Apr 15 14:53:33 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Mon, 15 Apr 2002 14:53:33 -0400
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org>
 <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com>
Message-ID: <tGyu8kKkXc0J092yn@gnosis.cx>                

|Do you think the set of "open source software" as defined by OSI and the
|set of "free software" as defined by FSF are
|1) the same set,
|2) two different sets,
|3) none of the above;

Well, both the OSI and the FSF are quite unambiguous in stating that
these are two different sets.

And both groups are equally unambiguous in stating that there is a lot
of overlap between the sets.  And moreover, the FSF is awfully clear in
stating exactly what they don't like about the set of things that are
Open Source but not Free Software (i.e. exactly the same the same thing
they don't like about all non-Free software).

It is somewhat up-in-the-air whether the set of Open Source Software is
a strict superset of Free Software.  At the least, Open Source includes
generally more software, and -very nearly- all software that is Free
Software is also Open Source Software.

There is nothing very difficult (and nothing -at all- ambiguous) in
here.  However, it *IS* true--as with any area of human cognition--that
there are people who fail to understand the categories.  Big deal.  Some
people are wrong, and some things are so.

Yours, David...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From greg at cosc.canterbury.ac.nz  Mon Apr 15 01:19:35 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 15 Apr 2002 17:19:35 +1200
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
References: <mailman.1018682966.2837.python-list@python.org>
Message-ID: <3CBA62E7.F54CCD08@cosc.canterbury.ac.nz>

Noah wrote:
> 
> So now I have non-blocking
> pipes wrapped with a select

I'm pretty sure you shouldn't be making anything
non-blocking. What exactly are you making non-
blocking, and why?

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


From pearu at cens.ioc.ee  Mon Apr  1 11:53:18 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Mon, 1 Apr 2002 19:53:18 +0300 (EEST)
Subject: PEP 285: Adding a bool type
In-Reply-To: <EOYp8.108214$Ou.56262@atlpnn01.usenetserver.com>
Message-ID: <Pine.LNX.4.21.0204011821200.1428-100000@cens.ioc.ee>

On 1 Apr 2002, Steve Holden wrote:

> If you seriously think that a proposal from the language's principal
> designer would cause such massive code breakage, not to mention
> contradicting almost every published Python book, then you aren't reading
> the same newsgroup, or using the same language, as I am!

Don't know about the books (never read a Python one), but isn't some
major breakage expected in Python 3? So, I'll try to keep open mind to any
changes, good or bad, less surprises later...  I want my current code
running with Python oo too.

> > -1 --- I have succesfully programmed in Python without needing the bool
> > function and I share Ka-Ping's concerns about Python future. To me the
> > only reasonable point in rationale seems to be that "other languages have
> > or will have boolean". That also sounds quite weak to me --- other
> > languages have many useful concepts that would be more worth for
> > introducing to Python than this boolean concept.
> >
> The most reasonable point in the rationale to *me* was the "experience from
> teaching Python...".

Don't have one. So I missed that point. Python is more than just a first
language, right? For me Python is something like 10th language and I am
used to all kinds of weird (read: useful) stuff. Boolean in Python would 
not be weird enough that I could see it useful in the proposed form. If
True and False would be something unique, like None is currently, then I
could see the point introducing them. But with its integer nature, they
are just 1 and 0 to me (in the == sense), just exposed a bit differently.

So far I have learned to use almost all new features in Pyhton to make my
programs better (shorter, easier to maintain, readable, efficient). I find
very little to learn from PEP 285, just an additional concept to
remember without any apparent advantage.

> > 0  --- this PEP will be hardly rejected, it seems to me.
> >
> Again, I think you're being unfair to Guido. If the PEP is accepted ist will
> be because you are in a minority who think it will not result in improvement
> to the language (or at least will not degrade it at all).

I think that the story of this PEP would be much shorter if it would have
come from somebody else, especially, when considering such diverse
opinions, from -1 to +1, even among highly regarded Python
users/developers/educators. From what I have learned last years about
Python development (other people comments being the main source), then it
is not exactly a democratic process where majority rules.

And I didn't mean to be unfair to anyone. Without having a privilege to
know  Guido personally, I draw my conclusions assuming basic human
nature. Guido must be a human, I have seen his picture on Internet;-). I
can only trust that he is a super one.

> Did it catch you on a bad day?

Well, it didn't make my day any better...

Regards,
	Pearu




From wadler at avaya.com  Mon Apr  1 11:07:48 2002
From: wadler at avaya.com (Philip Wadler)
Date: 1 Apr 2002 08:07:48 -0800
Subject: Programming Language Technologies for XML
Message-ID: <1901ed66.0204010807.41d4cdee@posting.google.com>

[Please contact me if you have any questions.  -- P]


            PLAN-X: PROGRAMMING LANGUAGE TECHNOLOGIES FOR XML

                     Oct 3, 2002   Pittsburgh, PA
                        (Co-located with PLI)

                            CALL FOR PAPERS
                    Submission deadline: May 1, 2002

XML has emerged as the de facto standard for data interchange on the web.
The use of XML as a common format for representation, interchange, and
transformation of data poses new challenges to programming languages,
applications, and database systems.  During the last few years, the
database research community has devoted a lot of attention to XML's data
representation challenges, as evidenced by the number of XML-related
publications in premier database conferences and journals.

In contrast, the attention devoted to XML by the programming language
research community has been minimal.  This is unfortunate, since the
robustness of current and future programming standards and tools for XML
will depend on the strength of their foundations in core programming
technologies e.g., XML parsing (parsing theory and incremental parsing),
XML schemas (type systems), XPATH expressions and XSLT programs
(pattern-matching languages and their optimization), XSLT debuggers
(dynamic program analysis and slicing).  Since XML is a new domain, core
programming technologies developed in past research cannot be used
unchanged; instead, novel research is required to address the unique
challenges posed by XML and its use in web applications and standalone
applications.

This workshop aims to bring together researchers from the programming
languages and XML communities,

  a) to foster novel research to address unique challenges being posed by
     XML on current and future programming technologies;

  b) to exchange information on early research experiences with XML-related
     programming systems, tools, and languages; 

and

  c) to expose the PLI community to XML technologies and the potential
     impact of these technologies on future software.

SUBMISSION PROCEDURE

We solicit submissions on original research not previously published or
currently submitted for publication elsewhere, in the form of extended
abstracts. These extended abstracts should not exceed 5000 words
(approximately 10 pages). Detailed submission instructions will be posted
at http://www.research.avayalabs.com/user/wadler/planx by early April.

PROCEEDINGS

There will be no formal proceedings.  An informal proceedings will be
distributed at the workshop.

IMPORTANT DATES
  Paper submission deadline                     May 1
  Notification of acceptance                    June 21
  Final papers due for informal proceedings     Sep 4

WEB PAGE:
    http://www.research.avayalabs.com/user/wadler/planx/

GENERAL CHAIR: 
    Vivek Sarkar, IBM

PROGRAM COMMITTEE:  
    Allen Brown (Microsoft) 
    Peter Buneman (Edinburgh)
    Sophie Cluet (Xyleme / INRIA)
    Mary Fernandez (AT&T Labs)
    Shriram Krishnamurthi (Brown)
    Makoto Murata (IBM Japan)
    Benjamin Pierce (University of Pennsylvania), co-chair
    Michael Schwartzbach (Aarhus)
    Dan Suciu (University of Washington)
    Philip Wadler (Avaya Labs), co-chair

INVITED SPEAKER:
    James Clark

-----------------------------------------------------------------------
Philip Wadler                                          wadler at avaya.com
Avaya Labs               http://www.research.avayalabs.com/user/wadler/
233 Mount Airy Road, room 2C05                  office: +1 908 696 5137
Basking Ridge, NJ 07920
USA
-----------------------------------------------------------------------
"When a Mathematical Reasoning can be had it's as great a folly to make
use of any other, as to grope for a thing in the dark,  when you have a
Candle standing by you."                        -- John Arbuthnot, 1692
-----------------------------------------------------------------------


From ajs at ix.netcom.com  Thu Apr 11 09:00:40 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Thu, 11 Apr 2002 09:00:40 -0400
Subject: What has PEP 285 done to us?
Message-ID: <002301c1e158$e230b8a0$0334fea9@carol>

Tim writes - 

>You're a sweetheart in person, Art.

Cigarettes just went up another buck a pack in
NY.  

Consider the possibility that I *am* a sweetheart.
And that my indignation is righteous enough, and
is expressed in the way it is because it is the
only possibility I can find, and in spite of,
among other things, a true crush on the wit of 
a Mr.Peters.

But at some point, after having satisfied myself
I have expressed what I need to express, I can
settle down a bit again, one hopes.

Art
  





From jimmy at retzlaff.com  Fri Apr 19 09:07:16 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Fri, 19 Apr 2002 06:07:16 -0700
Subject: use of calldll
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F564B@MAYOR.retzlaff.com>

> Piero [mailto:picodello at yahoo.it] writes:
>  I use the very good calldll module to interface to Windows DLL.

I agree that it is very good. I find it even better when using the
windll module which is part of the dynwin package from the same author.
windll is a Python wrapper around calldll that simplifies looking up and
calling dll functions. You can use windll separately from the rest of
dynwin (which has many examples of using windll).

> Something can help on using VARIANT type in Python and CALLDLL

The article here describes VARIANTs:
http://msdn.microsoft.com/archive/en-us/dnarextvb/html/msdn_article4.asp

The most important part is the C struct definition right near the top.
>From a Python calldll / struct module perspective, you want to read out
the first 2 bytes (as a struct module type 'H'). This tells you how to
interpret the contained value, i.e. its data type. The possible values
are documented here:
http://www.marin.clara.net/COM/variant_type_definitions.htm

You need to translate those data types into struct module format
strings. A dictionary can be very nice here. For example:

variantTypeMappings = {
    3  : 'l', # VT_I4
    17 : 'B', # VT_UI1
    2  : 'h', # VT_I2
    # etc.
}

Then, bytes [8:8+struct.calcsize(variantTypeMappings[vt])] from the
memory buffer should be passed to struct.unpack with
variantTypeMappings[vt] as the format string.

Win32all does automatic translation to/from variants for working with
COM Automation. I don't know if that functionality is exposed for direct
use, but if it is that would be a lot simpler than rolling your own.

Jimmy




From just at xs4all.nl  Sun Apr 21 09:57:10 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sun, 21 Apr 2002 15:57:10 +0200
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com> <a9ufpv$6hb$1@panix1.panix.com>
Message-ID: <just-DF77C7.15571021042002@news1.xs4all.nl>

In article <a9ufpv$6hb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> ...and if you're using Python 2.2, write it as
> 
>     while oid not in objects:
> 
> 
> Hmmmm....  it occurs to me that those of us teaching Python will need to
> be *very* careful in making sure that people avoid 'in' over lists.

I'm already happy if they don't do

  if x in d.keys(): ...

:-)

Just


From bokr at oz.net  Tue Apr 16 22:00:03 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 02:00:03 GMT
Subject: list_a-list_b
References: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu> <mailman.1019005834.26063.python-list@python.org>
Message-ID: <a9ikv3$p0u$0@216.39.172.122>

On Wed, 17 Apr 2002 02:44:38 +0200, Gerhard =?unknown-8bit?Q?H=E4ring?= <gerhard at bigfoot.de> wrote:

>* John <nobody at nobody.com> [2002-04-16 19:30 -0500]:
>> 
>> Hi, I need a help in writing a function that does
>> 
>> list_a=[1,2,4], list_b=[2,4]
>> list_minus(list_a,list_b)
>> print list_a #will print out a=[1] after removing all elements ([2,4])
>> appearing in list_b
>
>Ok.
>
>>>> l1 = [3,4,5,6]
>>>> l2 = [4,5]
>
>Do you still need the original list? If yes, you'd better make a copy of
>it:
>
>>>> l3 = l1[:]
>
>Then remove all elements of l2 in l3 like this:
>
>>>> for item in l2:
>...     l3.remove(item)
>>>> l3
>[3, 6]
>
>To fully understand what's going on here, you can read in the
>documentation about lists. The distinction about mutable/immutable
>objects comes into play here, too: lists are mutable => you change them
>"in-place". That's also the reason why l3 = l1[:] creates a copy, but
>l3 = l1 would not.
>
If you have duplicate elements in your list, you may also want to
think about what you really want list_minus to mean, and whether
you want to preserve duplicates in the result:

 >>> a=[1,2,3,3,4,5]
 >>> a
 [1, 2, 3, 3, 4, 5]
 >>> a.remove(3)
 >>> a
 [1, 2, 3, 4, 5]

For large lists with no duplicates it may be faster to create a dictionary
from the first list and delete entries with keys from the second list.

Regards,
Bengt Richter


From twisted at itamarst.org  Sat Apr 20 10:09:47 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Sat, 20 Apr 2002 16:09:47 +0200
Subject: ANN: Twisted 0.17.4
Message-ID: <3CC176AB.9000103@itamarst.org>

Twisted is an event-based framework for internet applications, written
in Python and licensed under the LGPL. For more information, visit:

     http://www.twistedmatrix.com

or join the list at:

     http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python


What's New since 0.16.0
=======================

- NNTP support and a framework for news servers.

- HTTP/1.1 now supports persistent connections and pipelining, for
improved performance.

- Many feature enhancements for the web framework.

- Added component architecture inspired by Zope3.

- Many bug fixes and feature enhancements.


What is Twisted?
================

Twisted is an event-based framework for internet applications.  It 
includes a web server, a telnet server, a chat server, a news server, a
generic client and server for remote object access, and APIs for
creating new protocols and services. Twisted supports integration of the
Tk, GTK+, Qt or wxPython event loop with its main event loop. The Win32
event loop is also supported, as is basic support for running servers on 
top of Jython.

Twisted currently supports the following protocols, all implemented in 
pure python, most of them as both servers and clients:

      - NNTP
      - FTP
      - HTTP
      - SOCKSv4
      - SMTP
      - IRC
      - telnet
      - POP3
      - AOL's instant messaging TOC
      - OSCAR, used by AOL-IM as well as ICQ (client only)
      - DNS
      - LDAP (client only)
      - finger
      - Echo, discard, chargen and friends
      - Twisted Perspective Broker

For more information, visit http://www.twistedmatrix.com, or join the 
list at http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python





From terrybyrne1963 at yahoo.com  Tue Apr 30 10:14:26 2002
From: terrybyrne1963 at yahoo.com (Terry Byrne)
Date: 30 Apr 2002 07:14:26 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <aaevge$emr$1@peabody.colorado.edu> <93d52e82.0204290618.349a0e5f@posting.google.com> <mailman.1020104212.2013.python-list@python.org>
Message-ID: <c9534282.0204300614.4edba80@posting.google.com>

Thanks, Geoff!

That's the clear sort of instruction I have a tough time finding elsewhere. 

Terry

Geoff Gerrietts <geoff at gerrietts.net> wrote in message news:<mailman.1020104212.2013.python-list at python.org>...
> The use of "pass" in the example doesn't actually do anything. I think
> what you're looking for is "continue". I can't explain the dramatic
> speedup you got out of adding them. It defies logic.
> 
> The "pass" keyword is a no-op. It's a placeholder. The idea is that
> you might want to do something like this:
> 
> class Spam:
>   pass
> 
> In this example, "pass" is a way to indicate that you don't want
> anything special -- your Spam class can use a default constructor, and
> needs no special methods. But you can't just write:
> 
> class Spam:
> 
> That's a syntax error. So "pass" allows you to say "do nothing
> special". This is also a way you can insert a virtual method:
> 
> class Spam:
>   def subclassed_operation(self):
>     pass
> 
> But "pass" isn't always the preferred approach. Lots of people have
> commented that better form might be:
> 
> class Spam:
>   def subclassed_operation(self):
>     """ this operation must be overridden by the subclass if it is to
>         do anything.
>     """
> 
> The doc string counts as a "statement" for the purposes of the
> compiler, making this valid syntax. Meanwhile, it provides explicit
> explanation of your intent.
> 
> Still others will prefer both: have a doc string AND a pass.
> 
> The "continue" keyword says "let's go back to the top of the loop and
> do the next item". The "break" keyword says "let's exit the loop right
> now, even if there's more work pending".
> 
> Those are your loop-control statements.
> 
> Luck,
> --G.


From cgale1 at cox.net  Fri Apr 26 20:17:01 2002
From: cgale1 at cox.net (Van Gale)
Date: Sat, 27 Apr 2002 00:17:01 GMT
Subject: comparsion of two lists
References: <20020426224924.GA1318@rehor.surfbest.net>
Message-ID: <16my8.10553$Oq2.1509222@news1.west.cox.net>

"Matej Cepl" wrote:
> I would like to compare two
> list of messages (in the local and remote folders) and create
> list of messages, to be deleted/downloaded. In plain Python, the
> part of deletion would go like this:

I'm not able to test your snippet at the moment, but it sounds like your
doing something similar to me when I wrote:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81932

Van





From mark at freelance-developer.com  Mon Apr  8 19:14:26 2002
From: mark at freelance-developer.com (Mark Nenadov)
Date: Mon, 8 Apr 2002 19:14:26 -0400
Subject: Creating a DB app with mySql and Windows GUI
References: <a8rhml$1id$03$1@news.t-online.com>
Message-ID: <wrps8.20370$Yt.763733@read1.cgocable.net>

Alexander v. Zitzewitz wrote:

> Hello Python Freaks !
> 
> What components should I use for building a GUI Database App with a MySQL
> Server and a GUI on Windows platforms with Python. The following
> components are needed:
> 
> - a feasable IDE with Debugger interface
> - a GUI Builder
> - a MySQL inding
> 
> Please let me know, if anyone already did a similar project and what
> experiences he/she made with using Python
> 

Alexander,

In the past, I wrote a few Python GUI apps that also used MySQL. You can 
download the files required for MySQL support in Python here 
(http://sourceforge.net/projects/mysql-python).

As for the GUI builder and IDE, I don't really know if I can help you. I 
used a wide variety of editors such as EditPlus, Vim, Ultraedit, Glimmer, 
etc. I rarely use IDE's that have debugging built in. As for building 
GUI's, I use wxPython. I hear that Boa Constructor is great for easily 
building wxPython GUI's, but I have never used it.

-- 
- Mark Nenadov (http://www.freelance-developer.com)


From rjones at ekit-inc.com  Thu Apr  4 18:38:27 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 5 Apr 2002 09:38:27 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <slrnaaphdc.39o.quinn@hurl.ugcs.caltech.edu>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com> <slrnaaphdc.39o.quinn@hurl.ugcs.caltech.edu>
Message-ID: <200204042338.XAA12847@crown.off.ekorp.com>

On Fri, 5 Apr 2002 07:29, Quinn Dunkan wrote:
> On Wed, 03 Apr 2002 16:00:42 GMT, Terry Reedy <tejarex at yahoo.com> wrote:
> >Just curious: are you willing to (consider, at least) switching to
> >ReST, or are we going to end up with two similar standards used by
> >different parts of the Python community?  IE, do Goodger's arguments
> >about why the differences from ST are improvements cut any ice with
> >you?
>
> I'm not speaking for Max, but my answer is "no."

I'm all for it, and I hope the ZC guys are too. If they don't I'll just have 
to write my own Zope product that wraps the ReST implementation. It's just 
too much of an improvement over the StructuredTexTNG implementation to ignore.


> I agree with Goodger that a convention for docstrings would be nice, and
> that structured text is the most reasonable candidate, but I don't agree
> that we need a complicated extensible format for it.  I use paragraphs,
> lists, verbatim code, and internal links (in the format I use, verbatim
> code goes in ''s and if it happens to be the name of a documented symbol in
> the current module it gets replaced with a link to the docs, so actually
> the format doesn't support links at all, but the translator can insert
> them), and anything with many more features, even if I personally don't use
> them, is likely to go over my "too complicated" threshold.

I've written a "primer" for ReST, which covers the exact simple components 
you describe above. At the moment, it's only available on my website:

   http://mechanicalcat.net/tech/quickstart.html

I hope that the primer clears up some of this confusion about the format 
being too complicated.


> A quick look at the python stdlib docs reveals that they could do just fine
> with the above features, except the use of footnotes, which I'm happy
> without.

I agree - I personally will be using ReST only for module-level docstrings, 
where they'll excell (since most module-level docstrings do tend to be 
verbose enough to warrant it :)


> Why not take Bobo's StructuredText and simplify it to the essentials rather
> than generalizing it to "the ultimate extensible structured text format"?

The problem with "Bobo's StructuredText" (wow, someone else who remembers 
Bobo :) is that it was never fully desgined, just implemented and hacked on 
over time. The beauty of ReST is that it's been designed from the start, and 
that means that if you want to ignore half of its features, you can. There's 
almost no chance of there being an accidental use of a feature you had no 
knowledge of, or inclination to use.




      Richard





From ngo at cartan.de  Thu Apr  4 18:53:04 2002
From: ngo at cartan.de (Nils Goesche)
Date: 05 Apr 2002 01:53:04 +0200
Subject: Eval only in homeopathic dosage (was: turning a string into an object name)
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com> <MPG.17165ec2e3c10c6c98969a@news.nwlink.com> <3D605F0E18400355.00C9CA937F5F0F26.B72EE95D42998046@lp.airnews.net>
Message-ID: <87hemr9g2n.fsf@darkstar.cartan>

claird at starbase.neosoft.com (Cameron Laird) writes:

> In article <MPG.17165ec2e3c10c6c98969a at news.nwlink.com>,
> Jeff Shannon  <jeff at ccvcorp.com> wrote:

> >If you have numerous objects that are conceptually similar, so 
> >that it makes sense to name them sequentially, then you *really* 
> >want to put them in a list or dictionary instead.
> >
> >The same thing is true of almost *every* use of exec or eval() -- 
> >it seems handy, but if you look for it, there's a better solution 
> >to your real problem.

> Of course.
> 
> The question I have is this:  do the (Common, let's say)
> Lispers agree yet?  Their usual response to lots of situ-
> ations is that Lisp has hygienic macros, and that makes

Lisp is not Scheme.  There is DEFMACRO, though...

> all the difference in the world.  Well, yeah, Lisp's 
> great for definition of Little Languages, which are
> always inherently desirable.  On the other hand, intro-
> spective fiddling is always, *always* tricky, and never
> helps maintenance of (end-user) applications.  We pro-
> pellorheads lust for it, and of course it's crucial in
> small doses, but it's simply a hazard beyond one or 
> two molecules per program.
> 
> I'll make this concrete:  what's a modern example of 
> use of a hygienic macro in an end-user application
> that wouldn't truly be better off as a dictionary or
> related data structure?

I am not sure I get what you mean.  How would you replace, say,
the WITH-OPEN-FILE or DEFCLASS macro by a dictionary?

Regards,
-- 
Nils Goesche
Ask not for whom the <CONTROL-G> tolls.

PGP key ID #xC66D6E6F


From stephen at xemacs.org  Sat Apr 13 02:20:12 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 13 Apr 2002 15:20:12 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <aAKt8.22045$ir6.979928@news1.calgary.shaw.ca> <slrnabf05v.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <874rig5dcz.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> writes:

    Huaiyu> So RMS says using "open source" to mean "free" is not what
    Huaiyu> FSF prefers, even though many people use it that way.  It
    Huaiyu> is definitely not what OSI prefers, either.  Why use it
    Huaiyu> that way?

    Huaiyu> Using two terms to mean the same concept when we do have
    Huaiyu> two distinct concepts to name leaves us one term short.

But we only have one concept.  What we have two of is _movements_.

Which term to use depends on which movement you want to imply
membership in.  When I advocate Linux, I talk about _free software_,
because software should be free (for many reasons), and because it
encourages community.  When I advocate bazaar development styles, I
usually talk about _open source software_, because I'm usually talking
to people who have a fiduciary responsibility to owners, and have to
justify "giving away the store" on the basis of return on equity.

    Huaiyu> What name do you propose for the broader concept of "open
    Huaiyu> source" as defined by OSI?

Either "open source" or "free" will do, because the concepts are the
same.

When talking about semi-free approaches, in which _some_ rights are
granted, I use "proprietary" when I dislike the approach or the
copyright holder (qmail), and "source available" or "semi-free" when I
like the approach or copyright holder (Aladdin Ghostscript).<wink>
Sometimes I use "published source" for licenses like AFPL.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From martin at v.loewis.de  Fri Apr 26 01:56:40 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 26 Apr 2002 07:56:40 +0200
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <j4helzdfbl.fsf@informatik.hu-berlin.de> <aa98g0$9i4$1@panix3.panix.com> <slrnacgag6.98m.akuchlin@ute.mems-exchange.org> <aaaknh$3uk$1@panix2.panix.com>
Message-ID: <m3r8l3roiv.fsf@mira.informatik.hu-berlin.de>

aahz at pythoncraft.com (Aahz) writes:

> >http://pyxml.sourceforge.net/topics/docs.html links to
> >http://py-howto.sourceforge.net/xml-howto/xml-howto.html, but the text
> >should really live on pyxml.sourceforge.net, because the master copy
> >of the text lives in the PyXML CVS tree.
> 
> Ah.  And conversely, py-howto should link to pyxml.  Well, I don't have
> access to either of those, so someone else needs to fix it.  Martin?

I've now removed all traces of the XML howto from the py-howto
project. It is not clear to my why the py-howto pages should have a
link, so I have not added one (in any case, I could not locate the
corresponding index.ht, anyway).

Regards,
Martin


From starx at pacbell.net  Tue Apr 30 20:01:57 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 1 May 2002 00:01:57 GMT
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net> <mailman.1020118433.15459.python-list@python.org> <Xns91FFB1BDEE1BCRASXnewsDFE1@130.133.1.4> <slrnacu417.pma.tim@vegeta.ath.cx>
Message-ID: <Xns9200AD794AD56RASXnewsDFE1@130.133.1.4>

Tim Hammerquist || Tue 30 Apr 2002 02:38:50p:

>> But if you just want to increment something, use 'foo+=1'. Exactly the
>> same, except one more character and only works on a line of it's own
>> (avoiding above said doom... man pointer-walking funcitons in C are
>> nuts with that... =)
> 
> Well, yes and no. The type being incremented is a string, not an
> int. For example: [..]

Oh... =). I figured it was just a wierd way of stating a variable name. 
But then, i have <2 minutes of total perl experience (if anything, 
changing shebangs and 'global settings' at the top of a few. maybe.)



-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From donn at u.washington.edu  Tue Apr 16 13:38:54 2002
From: donn at u.washington.edu (Donn Cave)
Date: 16 Apr 2002 17:38:54 GMT
Subject: File locking using fcntl.flock()
References: <mailman.1018927830.29158.python-list@python.org>
Message-ID: <a9hnje$a2k$1@nntp6.u.washington.edu>

Quoth Richard Jones <rjones at ekit-inc.com>:
| I can't get any locking to work on Solaris 5.7 using python 2.1.3. On Linux 
| 2.4.9, I get an exception when I try to re-lock something that's locked. On 
| Solaris, I don't get the exception (and without LOCK_NB, I don't block 
| either). I've also tried the lockf and posixfile.lock methods to no avail.
|
|>>> import fcntl
|>>> f=open('foo', 'w')
|>>> fcntl.flock(f.fileno(), fcntl.LOCK_EX)
|>>> f2=open('foo', 'w')
|>>> fcntl.flock(f2.fileno(), fcntl.LOCK_EX|fcntl.LOCK_NB)

I don't have Solaris, but I bet it doesn't have flock.  The real
flock() is more of a Berkeley UNIX thing.  Today you'll get it
on (Net/Free/Open)BSD & Darwin, Linux, Digital UNIX and probably
others.

The other locking function is the F_SETLK etc. option to fcntl,
which is specified in POSIX 1003.1 and supported by all modern
UNIX platforms.  And it's supported over NFS, where flock(2)
is not.  On the other hand, its semantics are bizarre.  Probably
the worst part is that a lock is removed when the process closes
_any_ file descriptor open on that file.  I have personally seen
this cause some havoc even when the programmer (not me) knew all
about the problem.  Another quirk is that, as you can see, it
doesn't support interlocking within a process, while flock() does.
This fcntl option is also invoked as lockf().

I think HP/UX 10.20 just doesn't have flock, period.  AIX makes
the worst of it, as usual, by providing an "flock" that just
calls fcntl.  I don't know which way Solaris goes, but it really
doesn't matter - because Python flock likewise calls fcntl if
true flock isn't supported.

	Donn Cave, donn at u.washington.edu


From nas at python.ca  Tue Apr 16 10:16:10 2002
From: nas at python.ca (Neil Schemenauer)
Date: Tue, 16 Apr 2002 07:16:10 -0700
Subject: Extension class problem
In-Reply-To: <3CBB14C9.A934DA18@anansispaceworks.com>; from hancock@anansispaceworks.com on Mon, Apr 15, 2002 at 10:58:33AM -0700
References: <3CBB14C9.A934DA18@anansispaceworks.com>
Message-ID: <20020416071610.A27829@glacier.arctrix.com>

Terry Hancock wrote:
> class GalleryArt(NaryaContrib, OFS.Image.Image):

Try putting the extension class first.

  Neil




From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 03:18:43 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 09:18:43 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE><7xhemb744m.fsf@ruckus.brouhaha.com>  <a9i41d$dqo$1@nets3.rz.RWTH-Aachen.DE> <mailman.1018994014.10229.python-list@python.org>
Message-ID: <a9j7o6$4tg$1@nets3.rz.RWTH-Aachen.DE>

"Ian Bicking" <ianb at colorstudy.com> wrote in message
news:mailman.1018994014.10229.python-list at python.org...
>..
> What you should do is sign your hidden variables.  That is, have a
> secret server-side key, and hash the key concatenated with the
> expression (using the md5 library, for instance).  When they resubmit
> the data you can confirm that the data originated from your server by
> comparing the hash they passed you with the hash you reconstruct.
>..

That sounds very interesting. Maybe I will go that route, additional to
other security (since the original XML data stored on the server should be
relatively safe from malicious authors, too). I will have to think about how
to store the secret server-side key, since the Python interpreter is open
source and everybody can download it. It would probably be sufficient to
have a text file with a key, which cannot be downloaded (achieving that
easily on every server might become another problem), and the key will have
to be changed by everyone once they put it up on their servers (it should
not have to be changed regularly since I'd lose the feauture of the client
being able to save the page and continue the game any time later). Thanks!




From gimbo at ftech.net  Fri Apr 12 06:16:11 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 12 Apr 2002 11:16:11 +0100
Subject: ftplib and FTP Objects...
In-Reply-To: <a94i7u$6id$1@news.wrc.xerox.com>
References: <a94i7u$6id$1@news.wrc.xerox.com>
Message-ID: <20020412101611.GB15466@andy.tynant.ftech.net>

On Thu, Apr 11, 2002 at 05:47:42PM +0000, jic wrote:

>       ftp.storbinary('STOR ' + fname,f.read )

The second parameter's wrong: you want to call f.read() to get the
file's contents, but in fact you're just passing a reference to the
file object's read method.

Try:

ftp.storbinary('STOR ' + fname, f.read())

Hope this helps,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From dkuhlman at rexx.com  Mon Apr 22 12:55:19 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: 22 Apr 2002 16:55:19 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <aa1f9m$6p9sv$1@ID-139865.news.dfncis.de>

Others have given better explanations than I can. But, you might
also look at the 'is' operator and the built-in function id().

    a is b if and only if id(a) == id(b)

The id() function will tell you whether you are dealing with a
reference to the same object or with a copy of the object.

dominikush at yahoo.com wrote:
> 
> Hello,
> 
> one thing I like very much about Python is that statements
> work like you would expect them to work. Take for example
> the use of dict.values() for dictionaries: If you store the
> result of dict.values(), and change the dictionary after-
> wards, the previously stored result remains untouched.


[snip]

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman


From walterm at parque.homelinux.net  Mon Apr 15 14:42:12 2002
From: walterm at parque.homelinux.net (Walter Moreira)
Date: Mon, 15 Apr 2002 15:42:12 -0300
Subject: ANN: RPy (R from Python) 0.2
Message-ID: <20020415184212.GA5149@parque.homelinux.net>

Hello.

RPy version 0.2 is released.  You can download it from the Sourceforge site:

    http://rpy.sourceforge.net

WHAT IS RPy?

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

CHANGES IN RPy 0.2

Fixed bugs

o There were several bugs related to memory and R and Python garbage
  collectors, which caused Python segfaults.
o There was a problem with signals traped by R, which made Python crash when
  <Ctl-C> was pressed.
o R objects of class 'factor' or 'data.frame' were incorrect converted to
  Python.
o Instances of any class converted to R were treated as sequences and made
  Python segfault.

Conversion

o RPy has now a new system for conversion of types from R to Python (see the
  docs), which includes user defined conversion functions.
o The conversion to R types is also customizable.
o Some utility functions are added.
o When Numeric (NumPy) is available, it is used for the standard conversion
  between R arrays and Python objects.

Documentation

o The documentation is converted to texinfo format, updated and enlarged.  In
  particular, better documentation for the compilation procedure and for the
  location of libraries is included.

Distribution

o There are now some examples included in the distribution.
o Regression tests are also included.




From jb at cascade-sys.com  Fri Apr 19 23:38:33 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 19 Apr 2002 20:38:33 -0700
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <20020419205101.17293.2839.Mailman@mail.python.org>
Message-ID: <3CC0E2B9.F12761D6@cascade-sys.com>

>
> Quoting Just van Rossum (just at xs4all.nl):
> ? I can totally see why people find it strange that to
> ? get the length of an object you write
> ?
> ?     len(o)
> ?
> ? and to get (eg.) the keys from a dictionary you do
> ?
> ?     o.keys()
>
> To which Geoff Gerrietts ?geoff at gerrietts.net> added:
> Maybe that distinction is artificial, or maybe it's less about Python
> and more about how I think about functions vs. methods. It's certainly
> a more OO-centric way of thinking about the two.

It seems that Python started out (as a Lisp derivitave) with much less of an OO emphasis.  I expect len(o) has precedent dating back to the original language.

However, over time Python has increasingly placed much greater emphasis on Objects.  Furthermore, OO programming generally has evolved from an experimental following to mainstream.  In this process, the universe has changed.  For one to say he expects o.len() to work is eminently reasonable and it's increasingly a wart on the language
that you have to revert to an alternative notation.

After all, when it is increasingly common for us to use OO notation for things like

    "a b c".split()

it's increasingly grating for "a b c".len() to be an ERROR.

I think o.len() should be recognized as a synonym for len(o).  Certainly the language has sustained a number of more dramatic changes in recent years.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From ralph at penguin.com  Wed Apr  3 08:00:03 2002
From: ralph at penguin.com (Ralph)
Date: Wed, 03 Apr 2002 13:00:03 GMT
Subject: How to wait only so long for a server to respond?
References: <mailman.1017800895.19820.python-list@python.org>
Message-ID: <n1Dq8.12893$ml2.1014180@newsread1.prod.itd.earthlink.net>

Sean 'Shaleh' Perry wrote:

> 
> On 03-Apr-2002 Ralph wrote:
>> Pythoners:
>> 
>> I have some code that requests a response from a server on the Internet,
>> but I only want to wait so long for the server to respond.  How can I
>> tell python to wait x seconds/minutes, then move on to the next section
>> of code if the server has not responded?
>> 
>> Thanks as always.
>> 
> 
> a select loop is a common way to wait for data on a socket and specify a
> timeout.  Look for documentation on select().
Sean:

Thanks for the hint.  I'll take a look.

Ralph


From python at rcn.com  Mon Apr 15 15:19:27 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 15 Apr 2002 15:19:27 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <a9eppf$ghv$1@panix1.panix.com>
Message-ID: <a9f95n$jg0$1@bob.news.rcn.net>

[Raymond Hettinger]
> >The help and tutor maillists indicate that the use of input() rather than
> >raw_input() is widespread.  I think this function is so harmful and so
> >likely to be used that it should be deprecated hasta pronto (right away).
> >When really needed, eval(raw_input()) is easy enough to code and explicit
> >about what it does.

[Aahz]
> Write a PEP.

Rules of the game:
'''A PEP that has not been discussed on python-list at python.org and/or
python-dev at python.org will not be accepted.  ''' -- from Pep 1




From aleax at aleax.it  Tue Apr 30 04:51:07 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 08:51:07 GMT
Subject: How to retrive attached file from newsgroup mail
References: <63170f57.0204300045.788bc634@posting.google.com>
Message-ID: <%Vsz8.33296$8D3.940548@news1.tin.it>

sanjay wrote:

>   Im using NNTP to retrive the newsgroups mails to display on web
> pages.The email is multipart i,e text/html with doc,pdf,gif file has
> attached. Is there any python script to get all the plain text/html
> message with attached file from mail.

Yes, there are Python modules in the standard library to let you
perform such tasks easily.  Most general and powerful is package
'email', that comes with Python 2.2 (but can also be downloaded and
installed separately on 2.1, I believe).  See, e.g.:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/86676

for a simple example.


Alex



From marklists at mceahern.com  Fri Apr 12 09:57:50 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 08:57:50 -0500
Subject: Variables common to all objects of a class
In-Reply-To: <3CB6CD99.70409@thomas-guettler.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEELIKKAA.marklists@mceahern.com>

[Thomas Guettler]
> How do you store variables which should be common to all
> objects of a class?

Short answer:

	class Foo:
		classVar = "whatever"

Example:

#!/usr/bin/env python

class Dog:

    species = "canine"

    def __init__(self, name=None):
        if not name:
            name = "mystery dog"
        self.name = name

    def bark(self):
        return "Hi, I'm a %s and my name is %s." % (Dog.species, self.name)

rover = Dog("rover")
fido = Dog("fido")

print rover.bark()
print fido.bark()

print Dog.species
print rover.species
print fido.species






From tanzer at swing.co.at  Tue Apr 16 02:28:21 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Tue, 16 Apr 2002 08:28:21 +0200
Subject: Python and spacecraft onboard or ground control scripting
In-Reply-To: Your message of "15 Apr 2002 13:08:47 CDT."
             <C04BD5ACF621A707.9626610E9084F7B7.9644F5C8B424796F@lp.airnews.net>
Message-ID: <E16xMRx-0007PC-00@tswings.swing.cluster>

claird at starbase.neosoft.com (Cameron Laird) wrote:

> Yes, for example, Python objects are dynamically created, and there is
> the potential for "out of memory". I have a personal belief, however,
> that the memory analysis of general C programs is no easier to compute
> than that of Python programs.

IIRC, the OP talked about mission critical systems. In many such
systems, you are not allowed to allocate memory dynamically. While I
agree that memory errors are much more likely in a C program than in a
Python program, this doesn't hold when there isn't any runtime memory
allocation going on in the corresponding C program.

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





From tonyl at homechoice.co.uk  Sat Apr  6 14:59:38 2002
From: tonyl at homechoice.co.uk (Tony)
Date: Sat, 06 Apr 2002 20:59:38 +0100
Subject: beginners syntax question Thanks
References: <3caf47a5@news1.homechoice.co.uk> <3CAF4E4B.ECE28871@ntlworld.com> <Xns91E8DD351888Acliechtigmxnet@62.2.16.82>
Message-ID: <3caf5499@news1.homechoice.co.uk>

Thanks Chris,

problem solved :)

Tony

Chris Liechti wrote:

> tony clarke <a.clarke11 at ntlworld.com> wrote in
> news:3CAF4E4B.ECE28871 at ntlworld.com:
> 
>> Hi
>> 
>> Try listing the objects  l=["object1','object2'..]
>> then
>> for i in l:
>>  i.update()
> 
> almost, leave away the quotes:
> l=[object1, object2]
>  
>>> I'm trying to call the update() method of several objects ie:
>>>
>>> object1.update()
>>> object2.update()
>>> etc
>>>
>>> I'd really like to do
>>>
>>> for x in range(7):
>>>   "object"+str(x).update
>>>
>>> but can't figure outthe syntax, is this possible?
> 
> yes it is, but there are better ways.
> why are you not storing the objects in a list
> instead of naming them with numbers?
> 
> calling methods on objects in a list is shown above.
> 
> chris
> 



From jhalttun at pp.htv.fi  Mon Apr 22 06:05:27 2002
From: jhalttun at pp.htv.fi (janne halttunen)
Date: Mon, 22 Apr 2002 13:05:27 +0300
Subject: curses mouse-events in linux console
Message-ID: <3CC3E067.6080107@pp.htv.fi>

Hello,

I'm on Mandrake Linux 8.2 with Python 2.2, and I have a problem of 
mouse-events not emitting properly in curses when on console with gpm. 
In X it works fine, the getch-method returning instantly on 
mouse-clicks, but on console it doesn't return until after pressing key 
on the keyboard.

The following script can be run to demonstrate the problem:



# curses mouse test: ESC to exit

import curses

def main(scr):
     curses.mousemask(curses.ALL_MOUSE_EVENTS)
     key=0
     while key!=27:
         key=scr.getch()
         scr.erase()
         if key==curses.KEY_MOUSE:
             scr.addstr(0,0,str(curses.getmouse()))
         else:
             scr.addstr(0,0,str(key))
         scr.refresh()

if __name__=='__main__':
     curses.wrapper(main)

# END


Hope there is a solution to this.


janne



From siegfried.gonzi at kfunigraz.ac.at  Fri Apr  5 08:11:18 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Fri, 05 Apr 2002 15:11:18 +0200
Subject: Theoretical threshold of improving Python's performance
Message-ID: <3CADA276.D91B7544@kfunigraz.ac.at>

I got a private email where I can read that Python can be never made as
fast as Common Lisp due to Python's dynamic nature. May I ask  without
any ulterior-motive: "Has the beformentioned proposition any value?".

I know there were discussion concerning the topic; but I would be really
interested in, from a theoretical point of few, whether it is a *fact*
that Python can *never* become as fast as Common Lisp* or not.

I think it is documented that I am not really (maybe until yet) striving
for a faster Python (clearly, would not say no) but I am really
interested in and a little bit clueless whether Python is really much
more dynamic than Common Lisp and this is an upper threshold.


S. Gonzi
*[I got some complaints that I sometimes stress the 10 times Common Lisp
performance hit; if some experienced profi programmer can tell me that
he is using Common Lisp in combination with declaring variables on a
daily basis I am going to quit my rants immediately]


From cstoy at nc.rr.com  Wed Apr 10 00:20:34 2002
From: cstoy at nc.rr.com (Chris Stoy)
Date: Wed, 10 Apr 2002 04:20:34 GMT
Subject: Embedding Python with C++
References: <mailman.1018363820.30451.python-list@python.org> <3cb359c3$0$362$9b622d9e@news.freenet.de>
Message-ID: <m4Ps8.33916$XP2.12331779@typhoon.southeast.rr.com>

Thanks for the responses.  However, I guess I wasn't totally clear in what I
was asking.

I can't use the Python.dll.  I HAVE to make a static link version so that I
can port it to another platform (lets just say its another Microsoft
product...)

It is also not possible for me to use Python as the main program.  The idea
is to be able to hook in not just Python, but perhaps other scripting
languages later on.

I've read the python.org docs and haven't found info on how to recompile
into a lib.

If I'm missing something, please let me know.

Thanks.

Chris.





From sholden at holdenweb.com  Wed Apr 10 11:30:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 10 Apr 2002 11:30:02 -0400
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk> <vnys8.12$p56.2197@newsb.telia.net>
Message-ID: <2WYs8.20054$GM.11784@atlpnn01.usenetserver.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:vnys8.12$p56.2197 at newsb.telia.net...
> > So perhaps Guido need a little work with the people skills ;-) and the
> > rest of us need to act properly, which includes not getting upset when
> > others don't.
>
> "It's all about doing what's best for the little ones."
>
> http://www.theonion.com/onion3811/child-rearing_advice.html
>
Apparentlyt the Onion doesn't like this link, and redirects to the front
page of the current issue. Oh, wll.

regards
 Steve






From martin at v.loewis.de  Wed Apr 17 14:34:20 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 20:34:20 +0200
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au>
Message-ID: <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de>

Dave Cole <djc at object-craft.com.au> writes:

> Martin> Yes, the Global Interpreter Lock (GIL). When C code runs that
> Martin> was invoked from Python code, it holds the GIL, and only one
> Martin> such C code can run at any time. So you need to release the
> Martin> lock, using Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS.
> 
> If the API which is being wrapped by the extension module utilises
> callbacks, and you want to make those callbacks available to Python
> code, and you want to do this in a threading environment where the GIL
> is released when calling the API, then the above will not work.

You must not call *any* Python API while not holding the GIL, yes.

That doesn't mean that my approach won't work - it is just not the
full story.

> This seems to be working fine so far (fingers crossed).

Sounds good; if possible, you should indeed pass the thread state
forward to the callback. If you have a lock around the library,
passing that thread state in a global variable works fine (see
_tkinter for an example).

If callbacks can come out of arbitrary threads, or if it is not
possible to pass the thread-state through, you can always create a new
thread state when entering the callback. In that case, you still need
to obtain the interpreter somehow.

Regards,
Martin



From gerhard at bigfoot.de  Tue Apr  9 13:20:18 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 9 Apr 2002 17:20:18 GMT
Subject: Beginner's performance problem
References: <memo.20020409174302.1584B@a.a>
Message-ID: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de>

Mark Charsley wrote in comp.lang.python:
> For reasons best not gone into, I needed to correct the case of a whole 
> bunch of SQL code. As such I created the little script below...
> 
> It reads in a source file, then for each line it does a case-insensitive 
> search for each TableName and ColumnName (as contained in the "names" 
> collection), checks that it's not just matching a substring, and then 
> corrects the case of the word if necessary.
>
> [snip code]
>
> 1) why is it so slow? I wouldn't have thought that processing 13,000 lines 
> * 200 names shouldn't take more than a few seconds on a modern PC.

String manipulation is costly in Python because strings are immutable.
Every string manipulation creates a new string object in effect. A
good approach is to tokenize your string into a list (which is
mutable) and operate on the list only, then join the list into a
string again.

I haven't tried but an approach using the re module should be really
fast, and shorter. Untested code follows:

import re
names = ["ALGORITHM_CLASS", ...]

text = open("ThirteenThousandLines.sql").read()
for name in names:
    regex = re.compile(name, re.I) # case insensitive search for name
    text = regex.sub(name, text)   # replace all occurences
print text

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From hfoffani at yahoo.com  Wed Apr  3 07:56:33 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Wed, 3 Apr 2002 14:56:33 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <ualr7hmlhkemab@news.supernews.com>
Message-ID: <20020403075637.591$NC@news.newsreader.com>

"John Roth"
> Consequently, I think of it (0 and  1 as false and true) as a
> hack carried over from C and similar languages.

The hacks are zero false and not zero true. Both are JPZ, JPNZ
assembler instructions surviving in high level languages. :-)

Regards,
-Hernan






From fperez528 at yahoo.com  Fri Apr 12 16:32:13 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 12 Apr 2002 14:32:13 -0600
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>
Message-ID: <a97gab$8n5$1@peabody.colorado.edu>

Bryce C. Mason wrote:

> Hi folks.
> 
> I am doing my development on a PC (activeware's distribution) and the final
> project will be running on a macintosh.  Umm...I am a little worried that
> there is not a command line interpreter in the Mac, at least one I have not
> seen.  Help.  How do I port my project to the mac?
> 
> Bryce Mason

If it's mac OSX you're fine, it's a full  posix environment. I don't know 
anything about Mac Classic.

f.


From phr-n2002a at nightsong.com  Sun Apr 14 00:35:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 13 Apr 2002 21:35:24 -0700
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org>
Message-ID: <7xr8lic2yb.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> It's in the nature of open source development that volunteers scratch their
> own itches, then share what they've done because it does them good too to
> have their work widely adopted.  If you can ride that curve, you can get an
> enormous amount of high-quality work for free.  If not, you'll have to pay
> for it or go without.

A lot of stuff gets done because it needs to get done, rather than
scratch an itch.  

> A crushing amount of debate over "stability" has gone by on Python-Dev the
> last week, and the one thing that strikes me over and over is that, with
> just a few minor exceptions, nobody volunteers to *do* anything except tell
> other people what to do.  What are you willing to do to make your desires a
> reality?  If it doesn't involve contributing work, time or money, you just
> want a free ride.  But nobody owes anyone a free ride, and open source can't
> change that fact of life either.

In that particular debate, a lot of people were asking for certain new
features to be LEFT OUT of Python.  I don't understand how someone can
volunteer to contribute work, time, or money toward leaving something out.


From tim.one at comcast.net  Mon Apr 29 15:38:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 29 Apr 2002 15:38:32 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <20020429184935.GO16217@devel.trillke>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEOGPBAA.tim.one@comcast.net>

Note that Guido's comment on Python-Dev was:

    I strongly suggest sticking to the status quo.  There are more
    important fish to fry.

Relax.

or-don't-ly y'rs  - tim




From claird at starbase.neosoft.com  Wed Apr 10 16:33:05 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 10 Apr 2002 15:33:05 -0500
Subject: Help: Python for a big commercial project?
References: <mailman.1018437980.14078.python-list@python.org> <3b091a1c.0204100926.4ba2f13b@posting.google.com>
Message-ID: <D61EEEC80E329F9C.107BAF1C9197FB14.82A079054299F67D@lp.airnews.net>

In article <3b091a1c.0204100926.4ba2f13b at posting.google.com>,
cmkl <cmkleffner at gmx.de> wrote:
>I sucessfully use Python2.2 on HPUX 11 as well as I used Python2.1 on HPUX 10.2
>with InformixDB v1.3 and other useful extensions for development of medium sized
>inhouse test tools and database service tools.
>I agree: IT IS PAINFUL learn to build Python on HPUX but nevertheless it is 
>worth doing. DONT'T use the prebuilt buggy python versions of the 
>HPUX porting Center !! Build python yourself, it is then much easier to get 
>working C-extensions like database modules, numeric python ...
>
>threads is an issue on HPUX platforms. With HPUX11 I was sucessful to build
>thread enabled python (tests sucessful) but I was not able to build a working 
>multithreaded database application with a multithreaded enabled Informixdb
>module. But it may be possible. It is a matter of time for trying and testing
>different settings at compile time. 
			.
			.
			.
... independently of Python.  People working with Java,
Perl, ksh, ... regularly stumble into problems with
HP-UX.  From what I can tell, Python's no worse than
other languages in its support of all of HP-UX's
idiosyncrasies, and better than some.
-- 

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


From sanjay2kind at yahoo.com  Tue Apr 30 04:45:42 2002
From: sanjay2kind at yahoo.com (sanjay)
Date: 30 Apr 2002 01:45:42 -0700
Subject: How to retrive attached file from newsgroup mail
Message-ID: <63170f57.0204300045.788bc634@posting.google.com>

Hi,

  Im using NNTP to retrive the newsgroups mails to display on web
pages.The email is multipart i,e text/html with doc,pdf,gif file has
attached. Is there any python script to get all the plain text/html
message with attached file from mail.

Thanks,
Sanjay


From aahz at pythoncraft.com  Mon Apr 22 22:12:18 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 22:12:18 -0400
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019510980.7097.python-list@python.org> <3CC48EDD.2BEE5215@tundraware.com> <qX0x8.82827$T%5.64632@atlpnn01.usenetserver.com>
Message-ID: <aa2fu2$m33$1@panix1.panix.com>

In article <qX0x8.82827$T%5.64632 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>
>While interesting, however, this is all a rather long way from Python
>["spoilsport!!", shouted the crowd]. Next: "How many angels can dance on the
>head of a pin?"

Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam
Spam spam spam spam spam spam spam spam spam spam spam spam spam spam


From mwh at python.org  Wed Apr 10 06:47:24 2002
From: mwh at python.org (Michael Hudson)
Date: Wed, 10 Apr 2002 11:47:24 +0100 (BST)
Subject: RELEASED: Python 2.2.1
Message-ID: <Pine.LNX.4.44.0204101144550.29300-100000@starship.python.net>

We've released a bugfix release of Python, 2.2.1!

This release should be fully compatible with Python 2.2.  Everyone
using 2.2 is recommended to upgrade to 2.2.1.

On the 2.2.1 pages

    http://www.python.org/2.2.1/

you can find more information, binaries for Windows, Redhat 7.2 & 6.2,
Mandrake 8.2 and MacOS 8, 9 & X, as well as the usual tarball.  Thanks
are due to those who provided the binaries and to all who tested the
prerelases during the 2.2.1 release process!

One nasty bug (that affected Zope in obscure circumstances) and a few
less serious ones have been fixed since the second release candidate.
Two fixes may be of special interest if you use MacOS X:

 - This release finally works on Mac OS X multiprocessors
 - The IDE is now fully functional on Mac OS X.

This being a bugfix release, there are no exciting new features -- we just
fixed a lot of bugs.  For a moderately complete list, please see:

    http://www.python.org/2.2.1/NEWS.txt

Cheers,
The Python development team.





From aleax at aleax.it  Thu Apr  4 04:13:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 09:13:58 GMT
Subject: C++ etc (was Re: ACCEPTED: PEP 285)
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com>
Message-ID: <qPUq8.48457$pT1.1378522@news1.tin.it>

djw wrote:

> Pardon my ignorance (I've only been working with Python for a short
> while), but I don't understand why an idiom like:
> 
> if x == True:
> 
> would be/should be an error (or warning initially)

Ignorance is no crime!  Let's see if we can help you, and not just
in a Python context either.


> I see C/C++ code
> like this all the time:

So did I use to, as long as a very substantial part of my job was
teaching / tutoring / mentoring in C++ (i.e., up to 2 months ago).

Which IS a good part of why I quit that job (otherwise quite good
and fulfilling) and switched to one where I'm basically dealing
with Python full-time.  I'm still a Brainbench MVP for C++, and if
some C++ code needs dealing with I'm not going to pull back, but I
dearly hope it WON'T be "like this" (Boost Python's excellent use
of C++ and other really masterful C++ is more what I have in mind).

OK, back into "teach / tutor / mentor in C++" mindset, then...:

> bool x;
> x = Foo(Bar);

This is very, *VERY* bad C++ style.  Get into the habit of initializing
EVERY variable contextually with its declaration.  This is really the
ONLY way to ensure you won't be faced with uninitialized variables.

When you're a real master at C++, then and only then can you declare
without initializing in those extremely rare cases where this is
truly what you want to do.  But wait for a year or three of intense
C++ practice before you self-certify yourself as ready for that.  (Or
take the C++ test with Brainbench, or other similar self-assessment
procedure, and convince yourself that you're in the top percentile of
all C++ users).


> if (x==true)

This is almost as bad, for a different reason.  C++ forces you to
code a lot of boilerplate already.  Do *NOT* code boilerplate that
is NOT necessary: EVERY "dot of ink" on the "page" should be either
meaningful or mandated by the language.  (If you follow a style guide
that forces you to put meaningless dots of ink on the page beyond
those that the language itself mandates, change the guide).

Boilerplate hides bugs and can even cause them.  Some day,
somewhere, somebody will accidentally skip one of those two
'=' characters and will be working with a compiler that does
not scream to high heavens when you do (e.g. MS VC++ 6).  Then
instead of a comparison you will be testing an assignment,
ignoring the previous (actual!) value of 'x', and take the
'true' branch in what is actually an unconditional manner.

You should take every precaution to ward against such problems.
If your current compiler is able to warn you against them, turn
those warnings on, but don't rely EXCLUSIVELY on them, because
one day you WILL be using another compiler (or your code will
be maintained by somebody else who does some edits and then
uses a different compiler, etc, etc).  I would even suggest (I
have had scarce luck convincing people of this!-) to code
comparisons habitually with the constant on the LEFT, when
the comparison is in fact necessary, as in "if(23==z)" -- THIS
habit would basically make the skip-an-equals bug non-existent
.
But even if you can't or won't do that (not all people realize
that == defines an equivalence relationship, so you CAN count
on it being commutative, as well as reflexive and transitive!-),
there are other good habits you SHOULD definitely pick up, and
will help avoid the bug.  One of them is, NEVER compare any
boolean with the constants true and false.  Never.  If you never
do that, the bug will never bite in this specific context.


But, there is more.  You should always maximize polymorphism
in your code even when you're not exploiting it actively.  One
day you WILL want to make this code into a template.  The
fewer gratuitous non-polymorphic constructs your code embodies,
the less exertion will it take to templatize it, and the fewer
bugs will be left in it when you're done.  I.e., one day you
will want to have this code inside a

template <class T>
   ...
T x = Foo(Bar);


Now, if the next statement is "if(x)", your template will be
perfectly working with many diverse T, such as int, X*, and
so on.  If it's "if(x==true)" instead, you may forget to
edit it, the compiler will not complain if T is int, and
the behavior will NOT be equivalent.  Don't do it.


Besides, where do you stop?  x and x==true (when x is bool)
are absolutely equivalent -- and so are x==true==true and
(x==true)==true and x==(true==true) and ... so where do you
stop?  The only sensible answer: stop at once and just
write x.  The fewer extraneous, redundant and useless
'==' and 'true' tokens you have, the better.



> Of course, this could also be written in the more abbreviated form as
> well. But, why force that on the programmer? I really prefer the more
> explicit  comparison of x with true rather than the if(x) idiom. Myself,

It's a misguided preference, and it WILL cause you problems down
the road.

> being a newbie Python programmer (but not a newbie programmer in C, C++,
> etc), I would find it somewhat confusing that I could not compare a
> variable to a valid value for the variable using what appears to be a
> valid operator.

Making this comparison an ERROR would be a bit overboard, perhaps.

The language cannot forbid every absurd usage (C++ even happily
allows "if(x=true)", although finding a sensible use case for this
is impossible for me -- forbidding it would require complicating
the already-complicated language rules yet further, and there's a
limit to the amount of complication even C++ can tolerate!-).  A
warning, based on the heuristic consideration that this usage is
VERY error-prone and has absolutely NO good reason to recommend
it, is just about right, IMHO.


> Am I missing something more subtle here?

I think you are, as I tried to indicate -- even in a C++ context.


In Python, it's even more blatant, for several reasons.  First
of all, you're "always writing templates" in a sense: when a very
experienced C++ coder meets Python, he or she often gets this
eerie sense that, yes, it FEELS sort of like coding templates,
and for good reason -- Python is intrinsically "generic" just
like C++ is only within templates (the syntax is utterly different,
checks are runtime rather than compile-time, etc, but it's still
quite close).  Second, Python strives to MINIMIZE boilerplate, so
any boilerplate you gratuitously add stands out like a sort thumb.
Etc, etc.  But you should really consider adopting a consistent
habit of eschewing comparisons with boolean constants in just
about every language you may ever use...


Alex



From sholden at holdenweb.com  Wed Apr 10 12:36:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 10 Apr 2002 12:36:59 -0400
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <a90rv2$fvh$1@slb4.atl.mindspring.net>
Message-ID: <QUZs8.20974$GM.2449@atlpnn01.usenetserver.com>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:a90rv2$fvh$1 at slb4.atl.mindspring.net...
> Emile van Sebille:
> >Oooh... that's almost as much fun as saying 'xyzzy' the first time...
>
> I just said it, but nothing happened here.
>
Sorry, that's an old, worn-out magic word

rogue-ish-ly y'rs  - steve






From maxm at mxm.dk  Tue Apr  2 02:30:18 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 02 Apr 2002 07:30:18 +0000
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>
Message-ID: <3CA95E0A.3070507@mxm.dk>

Steve Holden wrote:


> Arguably, most of that code should have read
> 
>     if ret:
>         ...
> 
> rather than
> 
>     if ret == None:
>         ...
> 
> since the latter has the flavor of "if expr==True".
> 
> So, is breaking bad code more excusable than breaking good code ;-)?


Oh I don't think so. This is a common idiom:

def f(names=None):
     if names == None:
         names = []
     ...


regards Max M



From Oschler at earthlink.net  Mon Apr 15 07:59:14 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 15 Apr 2002 07:59:14 -0400
Subject: I'm lookin' 4 cute forum scripts
References: <mt4lbucdcmfpf9ln5ma5n4kfdm1vr4qf0a@4ax.com>
Message-ID: <Yfzu8.93570$gA5.7429614@e3500-atl2.usenetserver.com>

Johann <programisci at NOSPAM.murator.com.pl> wrote in message
news:mt4lbucdcmfpf9ln5ma5n4kfdm1vr4qf0a at 4ax.com...
> I am looking for cute (functional and visual) discussion forum for
> www. I would like to display the tree with threads, sort by author,
> date and search.
>

Have you looked at Zope?

www.zope.org






From mtaylor at lowebroadway.com  Mon Apr  8 11:19:30 2002
From: mtaylor at lowebroadway.com (Moray Taylor)
Date: 8 Apr 2002 08:19:30 -0700
Subject: import and function question
References: <d906b2c1.0204080155.77b87631@posting.google.com> <0aes8.59173$pT1.1704009@news1.tin.it>
Message-ID: <d906b2c1.0204080719.3212a039@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<0aes8.59173$pT1.1704009 at news1.tin.it>...
> Moray Taylor wrote:
> 
> > Hi,
> > 
> > I have two questions..
> > 
> > a) Can I 'import' a module with the full path, i.e. I don't want to
> > add it to sys.path, just something like 'import /home/me/mod.py', is
> > this possible.
> > I have tried adding it to the path, importing, and immediatly removing
> > it from the path, but the next time I do an 'import' with a different
> > module of the same name, it picks up the old one, even though it isn't
> > in the sys.path.
> 
> The import statement will always pick up sys.modules[name] when you
> import name.  If that's not what you need (and for plugins you may
> not want that, indeed) see the imp module in the standard library.
> imp.find_module lets you find a module on a directory-path of your
> choice, then imp.load_module lets you load the module you've found.
> 
> > b) I need to overwrite a function with a new one of the same name, but
> > 'import'ed from a module. Is this possible, I really just need to
> > destroy this function before I import the new one, can this be done?
> 
> del thefunctionname
> 
> will unbind the name from the function object ("destroy" only if there
> are no other references -- you cannot destroy an object to which there
> are outstanding references, Python saves you from dangling reference
> problems; if you have subtle needs e.g. for caching, see the weak
> references module).  Most often, though, just:
> 
> thefunctionname = themodulename.thefunctionname
> 
> (after importing the module and binding the module object to name
> themodulename -- whether by an import statement or by more devious
> means) will suffice.
> 
> 
> Alex

Thanks a lot, that was exactly what I was after, works great now!

Thanks again

Moray


From skip at pobox.com  Fri Apr 19 23:26:14 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 19 Apr 2002 22:26:14 -0500
Subject: An extraordinarily stupid question
In-Reply-To: <3CC0D892.2060006@nyu.edu>
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <15552.57302.644790.775510@12-248-41-177.client.attbi.com>

    Shane> ... I just want to print out a character without a newline and
    Shane> without a space behind it.

    Try this:

       import sys
       sys.stdout.write(c)
       sys.stdout.flush()

You can omit the flush if it's not critical that it display immediately.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From tim.one at comcast.net  Fri Apr 26 16:31:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 16:31:49 -0400
Subject: String handling bug in Python
In-Reply-To: <b16e4ef7.0204260944.193432b9@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEENPBAA.tim.one@comcast.net>

[Stephen Ferg]
> ...
> I would describe the situation this way... Python's "raw" strings
> aren't truly raw at all.  Python does not have any capability for
> TRULY raw strings -- that is, for string literals that are not escaped
> in any way.

That's true of raw strings now -- there are no "escape sequences" in
Python's raw strings.

> Escaping is still performed even with "raw" strings.

This isn't so.  r'\\' == "\\\\" and r'\'' == "\\'" and r"\"" == "\\\"" and
so on.  What you see is exactly what you get, no exceptions.  In return,
some strings can't be spelled *at all* using Python's raw-string notation,
and you appear to have the same problem with your scheme (read on).

> So the term "raw" is misleading.
>
> I would like to see Python support TRULY raw string literals -- string
> literals that are not escaped in any way.  Perhaps they could be added
> with a prefix of z (for zero escaping?), so that
>
>      z"\" == "\\"
>      z"\'" == "\\\'"

The first one can't be spelled at all with a raw string today.  The second
one can be, and indeed the same way you spelled it, as r"\'".  What do you
do in your scheme if you want a z-string to contain a single double-quote?
Don't brush that off, because it's the heart of the problem:  you need
*some* way to detect where the "TRULY raw string literal" ends.  Whatever
way you pick to spell that then becomes "a problem" if you want to include
the very same spelling *in* the raw string literal.

fortran-hollerith-strings-didn't-have-this-problem-but-people-got-
    tired-of-counting-how-many-characters-they-needed-ly y'rs  - tim





From hughett at mercur.uphs.upenn.edu  Mon Apr 29 17:09:07 2002
From: hughett at mercur.uphs.upenn.edu (Paul Hughett)
Date: 29 Apr 2002 21:09:07 GMT
Subject: Slice objects with negative increment
Message-ID: <aakcpj$5f$1@netnews.upenn.edu>

I am writing a Python extension in which it has become useful to
define slice objects that run backwards (i.e. that have a negative
increment) and want to use the existing convention, if any, for
such objects.  But I don't recall ever seeing such a convention.

Hence my question: Is there an established meaning within Python for
slice objects with a negative increment?  E.g. slice(1, 10, -2) ?

If not, then I am inclined to define slice(i,j,-n) to mean the same as
reverse(slice(i,j,n)); then slice(1,10,-2) would mean (9, 7, 5, 3, 1).
This seems simpler for users than the alternatives that I've considered.

The primary use case that I have in mind is a multi-dimensional array
type in which it may be necessary to access a given dimension in
either the forward or backward direction.  Under the proposed rule,
elements i up to j in the forward direction would be i:j:1 or just
i:j, while the same elements in the reverse direction would be i:j:-1.
Even-numbered elements up to j in the forward direction would be 0:j:2
and in the reverse direction would be 0:j:-2.

Any comments?


Paul Hughett


From aleax at aleax.it  Sat Apr 27 05:44:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 09:44:29 GMT
Subject: problems redirecting unittest output to sys.stdout
References: <aadplo$rs7$06$1@news.t-online.com>
Message-ID: <1quy8.68737$vF6.2063037@news2.tin.it>

Achim Domma wrote:

> Hi,
> 
> I want to redirect the output of a unittest to a textfile (on Windows).
> The default stream is sys.stderr which can not be redirected on windows.

Of course it can be -- depends on what Windows version you have.  With
NT's CMD.EXE (thus I believe also /2000 and /XP), 2>blah on the command
line works fine.  The 4NT shell works for Win/95 and up and supports
stderr redirection.  
http://www.developersdomain.com/vb/articles/redirectpipe.htm shows
how to to it in VB and C and includes links.
ftp://ftp.externet.hu/pub/mirror/sac/utilmisc/errout.zip is one of
a zillion utilities doing the same.

But of course it's even easier in Python -- just ensure you execute

import sys
sys.stderr = open('wop', 'w')

or

sys.stderr = sys.stdout

or whatever, before you import unittest.



Alex



From gerhard at bigfoot.de  Mon Apr 29 03:13:37 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 29 Apr 2002 09:13:37 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <d06z8.12562$iU4.1091750@newsread2.prod.itd.earthlink.net>
References: <aaim48$7fd$1@bob.news.rcn.net> <d06z8.12562$iU4.1091750@newsread2.prod.itd.earthlink.net>
Message-ID: <20020429071337.GA22512@lilith.my-fqdn.de>

* Chris Tavares <christophertavares at earthlink.net> [2002-04-29 06:47 +0000]:
> 
> "Raymond Hettinger" <python at rcn.com> wrote in message
> news:aaim48$7fd$1 at bob.news.rcn.net...
> [ ... snip ... ]
> >
> >     I propose that a simple means be provided to re-enable those functions
> > at
> >     will when they are needed to run old code.  Add a command line
> option -b
> >     and an equivalent environment variable, PYTHONBUILTINRESTORE which
> would
> >     start-up Python by running:
> >
> >         from functionals import map, filter, reduce
> >         from math import pow, divmod
> >         from deprecated import input
> >
> 
> No, no, no, no, no, no, no, no, no. This would place the burden of calling
> this code correctly on the end user. How's the average programmer going to
> know when to use -b? What if somebody set PYTHONBUILTINRESTORE once, then
> trashes his .bashrc file, and now none of his previously working Python code
> works? These kinds of command line switches are evil.
> 
> Also, what about environments without command lines? Python runs on a lot
> more than just Unix.
> 
> The "directive" needs to be in the code. And you've already provided the
> form of this directive: the import statements above.
> 
> I would think the reasonable migration would be to add the modules listed
> above (or add the functions as mentioned), but leave the builtins for a
> version or two. Then remove the builtins.

100 % agree with Chris. Also, Python grows new command line switches
like weed. By the current speed, it won't take long until we have -a to
-z in use. I don't like to control language behaviour by command line
switches - I want to see *in the code* how it behaves, and no, the
shebang line doesn't count, as it is a unixoid feature only.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From phd at phd.pp.ru  Tue Apr 30 03:57:06 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 30 Apr 2002 11:57:06 +0400
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <200204301324.25823.rjones@ekit-inc.com>; from rjones@ekit-inc.com on Tue, Apr 30, 2002 at 01:25:41PM +1000
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <mailman.1020024596.32196. <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com> <200204301324.25823.rjones@ekit-inc.com>
Message-ID: <20020430115706.B15421@phd.pp.ru>

On Tue, Apr 30, 2002 at 01:25:41PM +1000, Richard Jones wrote:
> I'm working on repackaging gadfly right now and will place the  results up on 
> gadfly.sf.net. I've started with the Zope version, so the re conversion is 
> done. I've also integrated Oleg's kjbuckets efforts. I'm almost there... just 
> got a few more wrinkles  to sort out before I make the first release.

   Thank you! We are waiting...

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




From martin at v.loewis.de  Sat Apr  6 16:57:46 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Apr 2002 23:57:46 +0200
Subject: 2.2.1c2 test_fcntl hangs
References: <lysn69qrp7.fsf@bock.chem.unc.edu>
Message-ID: <m3r8lsfq1x.fsf@mira.informatik.hu-berlin.de>

"Jochen K?pper" <jochen at unc.edu> writes:

> | Traceback (most recent call last):
> |   File "Lib/test/test_fcntl.py", line 38, in ?
> |     rv = fcntl.fcntl(f.fileno(), fcntl.F_SETLKW, lockdata)
> | KeyboardInterrupt
> `----
> 
> This is RedHat-7.0 with all updates + gcc-3.0.4. Python is

On a local file system, or through NFS? Is there any chance that
anybody else was holding locks to the test file at the same time?

Regards,
Martin



From aleax at aleax.it  Mon Apr 29 03:48:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 07:48:47 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <j4adrn9dtg.fsf@informatik.hu-berlin.de>
Message-ID: <zV6z8.30796$8D3.848020@news1.tin.it>

Martin v. L?wis wrote:
        ...
>> Right now, .PTH files handle importing and path-setting needs of
>> installed packages.  Having them handle such packages' builtins-needs
>> as well seems a minor extension.
> 
> The current .PTH design allows for that kind of operation already:
> 
> # functional_compat.py
> import __builtin__
> exec "from functional import map,filter,reduce" in __builtin__.__dict__
> 
> # functional_compat.pth
> import functional_compat
        ...
> My original intention was that those import lines in .pth files can
> install Unicode codecs in the codecs registry, but they, of course,
> all for the execution of arbitrary stuff.

Right -- "import" can indeed "execute arbitrary stuff".  So, one way
to do the "minor extension" I was musing about would be to provide
suitable modules (perhaps a package) for the typical compatibility
needs related to builtins being deprecated or moved away.  Although
it appears to me that having the PTH file accept "from x import y,z"
and implicitly execute that in __builtin__.__dict__ might be more
direct and immediate for people who write .PTH files.


Alex



From fbartletFIXIT at optonline.net  Fri Apr 19 16:35:50 2002
From: fbartletFIXIT at optonline.net (Frederick H. Bartlett)
Date: Fri, 19 Apr 2002 20:35:50 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <a9pj29$pdm$1@nntp6.u.washington.edu>
Message-ID: <3CC07FA1.1653067B@optonline.net>

Donn Cave wrote:
> 
> Indeed, it could be partly I/O.

Probably not, since I slurped the file into a string before I began
processing. I'd be happy to do that for up to several tens of MBs (and
I've never had to work on any single file that large), since most of my
machines have > 100 MB ram.

Am I wrong?

Fred


From elian at inbrief.net  Mon Apr 22 11:19:39 2002
From: elian at inbrief.net (Edmund Lian)
Date: Mon, 22 Apr 2002 11:19:39 -0400
Subject: Overriding methods in classes that use property()
Message-ID: <OF0B3F6398.7B97C194-ON85256BA3.00520742@juscom.com>

Hi,

When I run the code appended below, class B inherits all the methods of
class A, including the property() definition as I would expect.
Unfortunately, class C also does the same; I cannot override class A's
setVal method. Is this a bug, or do I have to do something different (and
unexpected) with Python 2.2's new style classes?

...Edmund.

class A(object):
    def __init__(self):
        self._val = None

    def setVal(self, value):
        self._val = value

    def getVal (self):
        return self._val

    val = property(getVal, setVal)

class B(A):
    pass

class C(A):
    def setVal(self, value):
        self._val = 'Huh?'





From mathieu543 at hotmail.com  Wed Apr 10 04:27:09 2002
From: mathieu543 at hotmail.com (Imer)
Date: 10 Apr 2002 01:27:09 -0700
Subject: class callable with any method name
References: <9d095c5a.0204091256.4d88a246@posting.google.com> <a9054j$iki$0@216.39.172.122>
Message-ID: <9d095c5a.0204100027.584ca47d@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<a9054j$iki$0 at 216.39.172.122>...
> On 9 Apr 2002 13:56:09 -0700, mathieu543 at hotmail.com (Imer) wrote:
> 
> >Hi, I'm trying to find a way to implement a class that would allow any
> >method name to be called.
> >When that happens, a special method would be called and the name of
> >the method that was originally called would be passed as an argument.
> >
> >For instance:
> >########
> >class A:
> >  def __method_called__ (self, method_name, method_argument):
> >    print "The method", method_name, "was called with the argument",
> >method_argument
> >########
> >
> >Then the following code:
> >########
> >a=A()
> >a.this_is_a_random_method(arg=1)
> >a.yet_another_random_method(param=3)
> >########
> >
> >Would print out:
> >########
> >The method this_is_a_random_method was called with the argument
> >{'arg':1}
> >The method yet_another_random_method was called with the argument
> >{'param':3}
> >########
> >
> >Is there a way to do that ?
> >
> >Thanks.
> 
> No guarantees, but try this:
> 
>  >>> class A:
>  ...     def __getattr__(self, attname):
>  ...         if attname.startswith('__'): return self.__class__.__getattr__(attname)
>  ...         self._attname = attname
>  ...         return self._special
>  ...     def _special(self, *args, **kwds):
>  ...         print 'The method "%s" was called with the arguments\n%s, %s' % (
>  ...             self._attname, args, kwds
>  ...         )
>  ...
>  >>> a=A()
>  >>> a.this_is_a_random_method(arg=1)
>  The method "this_is_a_random_method" was called with the arguments
>  (), {'arg': 1}
>  >>> a.yet_another_random_method(param=3)
>  The method "yet_another_random_method" was called with the arguments
>  (), {'param': 3}
>  >>> a.x()
>  The method "x" was called with the arguments
>  (), {}
>  >>> a.y
>  <bound method A._special of <__main__.A instance at 0x007D39E0>>
>  >>> a._attname
>  'y'
>  >>> a.z
>  <bound method A._special of <__main__.A instance at 0x007D39E0>>
>  >>> a.z(k=3)
>  The method "z" was called with the arguments
>  (), {'k': 3}
>  >>> a.z(1,2,3)
>  The method "z" was called with the arguments
>  (1, 2, 3), {}
>  >>> a.z(1,2,3, i=4, j=5)
>  The method "z" was called with the arguments
>  (1, 2, 3), {'i': 4, 'j': 5}
> 
> Regards,
> Bengt Richter

Thanks to all of you guys. I didn't know about __getattr__, but I knew
there had to be a way to do it ... Python is so great !

Mathieu.


From pyth at devel.trillke.net  Mon Apr 29 23:18:10 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 05:18:10 +0200
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <Epnz8.831$SY6.24713@news.itd.umich.edu>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <Epnz8.831$SY6.24713@news.itd.umich.edu>
Message-ID: <20020430031810.GY16217@devel.trillke>

On Tue, Apr 30, 2002 at 02:35:16AM +0000, Michael George Lerner wrote:
> Are you sure you're putting appropriate things into the list?

i guess he is. He just didn't know that python 
wants expliciteness rather than c++ish impliciteness
(type-conversions). 

    holger




From zopestoller at thomas-guettler.de  Fri Apr 26 11:10:23 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 26 Apr 2002 17:10:23 +0200
Subject: Thesaurus with Python
Message-ID: <3CC96DDF.6010405@thomas-guettler.de>

I want to design a thesaurus in python.

If I search for thesaurus and pyhton with google I get
no reasonable result.

The design of the thesaurus should be simple:

There are words with attributes.

An attribute is one of the following:

1. a link to a different word
   -BT BroadTerm for ...
   -NT NarrowTerm for ...
   -SA See also ...
   -USE use ... instead
   -LANG same as ... in language [English|German|French|...]
   (Some abbreviations are from z39.19.pdf Page xii)

2. A description
    Text, possible HTML

3. A URL to a webpage

4. An tag:
   -TT (TopTerm)
   -Language: [English|German|French]

The application should be front-end indepandent (GUI or
web). This could be done with xml-rpc or soap.
I want it to be independant because I think the best way
is to use a treeview. Treeviews can be done with HTML, too,
but they are not as good as a real GUI.

For searching you could use Zope's (Z)Catalog. Or has some a
better solution for searching?

I looked at wordnet [http://www.cogsci.princeton.edu/~wn/],
but it is not what I am looking for. Its intend is
to analyse the english language. I want to store knowledge and terms.

Freshmeat had nothing which looked good.

I found a description for RDF for thesauri
[http://ceres.ca.gov/thesaurus/RDF.html] but I didn't find it
very usefull. I wonder if RDF is interesting for this, because
I prefere to think in hashtables and lists and not in XML.

What do you think?

Have you hints or links?
which interface: xml-rpc, soap or other (should be over http)?
which search engine: ZCatalog or other?

  Have a nice weekend

   thomas

PS: Sorry if I don't reply immidiatley. I am not at home this weekend.
Maybe at saturday, but at least at monday.



From mwh at python.net  Sat Apr 13 07:53:01 2002
From: mwh at python.net (Michael Hudson)
Date: Sat, 13 Apr 2002 11:53:01 GMT
Subject: libtk.so
References: <3CB7F5B6.12C8C5AA@club-internet.fr>
Message-ID: <lkd6x3izey.fsf@pc150.maths.bris.ac.uk>

Fabien H?non <ffjhenon at club-internet.fr> writes:

> I have downloaded the latest version of Pyhon (2.2) for Linux ( Python
> 2.2 + tools + Tkinter 2.2 +devel)

Wot, the RPMs?

> But once I start an application that uses Tkinter, it asks me for
> libtk.so.0 and libtcl.so.0.

Works for me.

> I searched on my hard drive and could not find them. Where or how can I
> get them.

Installing tk?  What does 

$ rpm -q tk

say?  Though I don't see how you could have installed python2-tkinter
without having tk there... Apologies if you don't use redhat.

Oh hang on; the latest version of Python is 2.2.1.  Maybe the
dependencies are different in the 2.2 RPM.

You probably want 2.2.1, anyway.

Cheers,
M.

-- 
  You can lead an idiot to knowledge but you cannot make him 
  think.  You can, however, rectally insert the information, 
  printed on stone tablets, using a sharpened poker.        -- Nicolai
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From zeitlin at seth.lpthe.jussieu.fr  Tue Apr  9 09:08:43 2002
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Tue, 9 Apr 2002 13:08:43 +0000 (UTC)
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>
Message-ID: <slrnab5pur.ngm.zeitlin@seth.lpthe.jussieu.fr>

On 8 Apr 2002 14:43:17 -0500, Chris Spencer <clspence at one.net> wrote:
> Cons:
> 1.  BAD documentation.  They've taken the wxWindows library documentation, added
> a few notes, and called it the wxPython documentation.  And the thing is, they
> haven't kept it updated for more recent releases of wxWindows/wxPython, so
> there's a lot of undocumentated features that just don't show up.

 The problem with having wxPython-only documentation would be an even
increased risk of inconsistencies between wxWindows and wxPython docs. For
example, now when I (C++ developer) add a new wxClass I document it for
wxWindows and the wxPython users get at least some docs as well. If wxPython
docs were separate, do you think I (or anyone else but Robin for that matter)
would update wxPython docs at all? As much as I want to believe in this, I
don't - especially considering that although wxPython is the most successful
set of wxWindows bindings, it's not the only one. There is also wxPerl,
wxBasic, wxLUA and wxEiffel (and probably more) and so having separate
documentation packages for each of them is simply impossible.

 So even though I understand that there is a problem here (especially for
someone who doesn't know C++ at all), I simply don't see any satisfactory
solution to it. All I can say is that we're trying to improve the
documentation constantly and that usually the most effort is spent on this
just before a release so there is some hope that it will get better for 2.4.0
wxWindows release.

> 4.  Did I mention that the documentation is both incomplete and odious?  I
> should not have to constantly do a "find in files" to find out what parameters a
> damned method call takes because the way wxPython implemented a method is
> different than the wxWindows way and someone forgot to make a note of it in the
> rubbish passed off as a "manual".

 :-( Whatever people may think of it you should still realize that it requires
a lot of efforts to produce it. The manual is something that wxPython
programmers could help to improve without problems and I'm very grateful to
all people who send us the patches to it - if each time someone found an error
in the manual he'd sent us a patch, I'm sure the manual would be much better!

> 5.  I've had a REAL problem with conservative "window refreshes", where unless I
> specifically tell each window to repaint itself, I get a lot of scroll bars
> missing, a lot of windows being "grey" (shown without controls), Tab windows
> without the tab headers, etc.

 I don't know which version of wxPython did you use but this looks a lot like
the bug which was there in wxPython 2.3.2.1 and which has been fixed in wxMSW
recently. If it's not this, I'd like to know about the problem in more details
(on wx-users at lists.wxwindows.org a.k.a. comp.soft-sys.wxwindows if possible
please).

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From eppstein at ics.uci.edu  Sun Apr 14 18:50:39 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 15:50:39 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <eppstein-7156A5.13521614042002@news.service.uci.edu> <3cba0068@news.mhogaming.com>
Message-ID: <eppstein-CE1F85.15503914042002@news.service.uci.edu>

In article <3cba0068 at news.mhogaming.com>,
 "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:

> # give me either a or b, but only if one
> # of them is true, else give me false:
> c = a xor b
> 
> # the above would be equivalent to:
> def xor(a,b):
>     if   a and not b: return a
>     elif b and not a: return b
>     else:             return false

Or shorter

    def xor(a,b):
        return (not b and a) or (not a and b) or False

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From gashton at cmedltd.com  Wed Apr 17 10:40:51 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 17 Apr 2002 15:40:51 +0100
Subject: UTC equivalent of time.mktime() ?
Message-ID: <1019054451.562.319.camel@tp6>

Sorry to perpetuate the time theme, but I'm hoping somebody will cast an
eye over the following for me.

I'm trying to work out how to convert dates specified as strings (e.g.
"1 Jan 1970") into seconds since the epoch in UTC (e.g. 0). In a fit of
stupidity I've got stuck on a UTC vs localtime issue. If I explain what
I've been doing you'll see what I mean.

My approach is as follows:

  - use time.strptime() to convert the string into a time tuple

  - use time.mktime() to convert the tuple into seconds since the epoch
    (which is in localtime, for some God awful reason that continues to
     elude me - why?)

  - modify the epoch value according to the local time zone and DST
    (which seems horridly hackish).

In code, it looks like this:

Python 2.2 (#2, Mar 11 2002, 13:24:00) 
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import time
>>> s = "1 Jan 1970"
>>> tup = time.strptime(s, "%d %b %Y")
>>> tup
(1970, 1, 1, 0, 0, 0, 3, 1, 0)
>>> secs = time.mktime(tup)
>>> secs  # (I'm in the UK)
-3600.0
>>> def local2utc(secs):
...     if time.daylight:
...         return secs - time.timezone - time.altzone
...     else:
...         return secs - time.timezone
... 
>>> local2utc(secs)
0.0

I'm not convinced that a sensible implementation of time keeping
libraries would force me to go to the effort of writing local2utc() in
order to convert "1 Jan 1970" to 0, so suspect that I've missed
something rather large. ;)

Thanks.

P.S. I've tried getting strptime() to understand %Z in it's format
string, and specify the timezone in the initial date string, but it
can't do it (even though the docs state that it understands everything
that stftime() does). Is this a bug, or a platform dependent thing? (I
suspect the latter -- I'm on Linux).

-- 
Graham Ashton





From weeks at vitus.scs.agilent.com  Fri Apr  5 13:04:15 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Fri, 5 Apr 2002 18:04:15 +0000 (UTC)
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
Message-ID: <1018038856.507460@cswreg.cos.agilent.com>

phil hunt (philh at comuno.freeserve.co.uk) wrote:
: In both cases, there are good arguments in favour of the change. 

I'd like to note that if every change for which there were good arguments
were added to Python, the result would be terrible.

: But is this changing going to be continual? Are the language 
: developers constantly going to tinker with the language, out of a 
: sense of fun?

Well ... If I had created something fabulous and then found myself with
nothing further fabulous to do, I'd have a hard time letting go of the
original fabulous something.


Greg


From dalke at dalkescientific.com  Wed Apr 17 16:00:47 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 17 Apr 2002 14:00:47 -0600
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net> <Xns91F37887977AFRASXnewsDFE1@130.133.1.4>
Message-ID: <a9kk2i$ls8$1@slb1.atl.mindspring.net>

Philip Swartzleonard added to my table:
>>   verboseGenderTable = {
>>      "m": "male",
>>      "f": "female", <          # Most large eukaryotes
>    "a": "asexual",           # Bacteria, etc
>    "h": "hermaphrodite",     # Earthworms, etc
>    "n": "neuter",            # Robots, AIs
>    "r": "refuse-to-stateist",# The paranoid
>    "u": "unknown"            # We lack information /
>                              # last seen in dark ally /
>                                 # raised by wolves, dosen't know
>>   }
>>   verboseGender = verboseGenderTable[gender]

Well, it gets even worse.  For example, there are lizards which
are parthenogenetic (reproduce asexually) but are all female.  Here's
a nice list of terms if you really want to populate that table  :)

  http://www.soton.ac.uk/~cpd/repmod.html

                    Andrew
                    dalke at dalkescientific.com





From wideopen at shinbiro.com  Wed Apr 17 03:29:11 2002
From: wideopen at shinbiro.com (Haegi Kim)
Date: 17 Apr 2002 00:29:11 -0700
Subject: Type Conversion to COM Variant
Message-ID: <a4488154.0204162329.1f7ec7f@posting.google.com>

Hi, experts,
 
I have a COM client which gets a reference to a COM server.
The COM client has a exec expression in order to execute a script in
which the reference to the COM server is used.
It works find with the COM server in InProcServer mode, but doesn't in
LocalServer and raises the following error.
-----------------------------------------------------------------------------
exceptions.TypeError: Objects of type 'instance' can not be converted
to a COM VARIANT"
-----------------------------------------------------------------------------
In what case the type conversion causes the error? And what should I
do?
Thanks in advance.


From JamesL at Lugoj.com  Mon Apr  8 23:43:09 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 9 Apr 2002 03:43:09 GMT
Subject: Flying Cars? Was: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org> <3CB21B26.8000201@hccnet.nl> <6mos8.3235$vT6.3113@atlpnn01.usenetserver.com>
Message-ID: <Xns91EAD2826E9E3JamesLLugojcom@216.148.53.83>

"Steve Holden" <sholden at holdenweb.com> wrote:
> "Jaap Spies" <j.spies at hccnet.nl> wrote:
>> Most car companies release an new model every year.
>> James, I suppose you want to drive always the latest model.
>>
> Not at all. James only wants to drive the latest rocket-powered car if
> all the old cars also run off the new rocket fuel. Otherwise he feel
> the car manufacturers have broken backwards compatibilty -- even if the
> new rocket-powered cars will also run on gasoline.

Hey, it is 2002, where the %#&#@ is my flying car!? I'm not too particular 
what it runs on or what it is compatable with. And if it is rocket powered 
I expect it to reach orbital speed. :-)

P.S. When my '88 Acura Integra's transmission finally failed after 11 
years, I went out and bought a '99 Acura Integra. But don't read too much 
into that.

P.P.S. I have no idea what cars, rocket, flying, or otherwise have to do 
with a programming language. But this IS Usenet after all.


From sholden at holdenweb.com  Tue Apr  2 07:06:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 2 Apr 2002 07:06:08 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <3CA95E0A.3070507@mxm.dk>
Message-ID: <01d401c1da3e$c6c1ed30$d200000a@COMPUTER>

----- Original Message -----
From: "Max M" <maxm at mxm.dk>
Newsgroups: comp.lang.python
To: "Steve Holden" <sholden at holdenweb.com>
Sent: Tuesday, April 02, 2002 2:30 AM
Subject: Re: PEP 285: Adding a bool type


> Steve Holden wrote:
>
>
> > Arguably, most of that code should have read
> >
> >     if ret:
> >         ...
> >
> > rather than
> >
> >     if ret == None:
> >         ...
> >
> > since the latter has the flavor of "if expr==True".
> >
> > So, is breaking bad code more excusable than breaking good code ;-)?
>
>
> Oh I don't think so. This is a common idiom:
>
> def f(names=None):
>      if names == None:
>          names = []
>      ...
>
Should anyway have been

    if not ret:
        ...

but certainly when you are chossing between two false values you have to be
explicit about which is the sentinel.

regards
 Steve





From paul.osman at sympatico.ca  Fri Apr 26 12:59:52 2002
From: paul.osman at sympatico.ca (Paul Osman)
Date: Fri, 26 Apr 2002 12:59:52 -0400
Subject: Sending mail to an SMTP server that requires POP authentication ?
References: <aabrbv$2vg$1@norfair.nerim.net>
Message-ID: <3CC98788.9020100@sympatico.ca>

    I too found this problem... I have an SMTP server at my work that 
requires authentication.. now the version of smtplib I was using had a 
login method, which took a username and password as its arguments, but I 
kept getting an SMTPException: No suitable authentication method was 
found. I spent a good half hour talking back and forth with the server, 
and I recall that I tried a few things (can't remember at the moment) 
but no matter what i did, i couldn't get it to authenticate... I ended 
up doing the damn thing in perl <shutter> which had an auth method that 
worked. Has anybody found a solution for this? Is there a patch 
available? If so, please let me know because I'd love to remove all the 
perl scripts I had to write from action and replace them with python..

-Paul

Gillou wrote:

>With this kind of server, only holders of a pop account can send mail
>through SMTP using their POP credentials.
>
>I did not find how to do this with smtplib.
>
>Did someone succeed ? How ?
>
>Thanks in advance.
>
>--Gilles
>glenfant AT bigfoot DOT com
>
>






From aleax at aleax.it  Thu Apr  4 04:24:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 09:24:52 GMT
Subject: Catalog of Python Patterns? [LONG!]
References: <a8g97m$10jq$1@news.idiom.com>
Message-ID: <EZUq8.48488$pT1.1378962@news1.tin.it>

Jim Dennis wrote:

>  GoF Design Patterns in Python
> 
>  [Warning LONG POST AHEAD!]
> 
>  I'd like to see a canonical catalog of OODP in Python.

Then get Bruce Eckel's "Thinking in Python" release 0.1 at
http://www.mindview.net/Books/TIPython .

"a somewhat straightforward partial translation of 
Thinking in Patterns", as Bruce puts it.  It can be
enhanced a lot, sure, but that's why Bruce puts his
books up for download, commentary and feedback.


Alex



From ark at research.att.com  Wed Apr  3 11:19:33 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 3 Apr 2002 11:19:33 -0500 (EST)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <00af01c1db2a$e7316d10$ced241d5@hagrid> (fredrik@pythonware.com)
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff> <yu99wuvovkmm.fsf@europa.research.att.com> <00af01c1db2a$e7316d10$ced241d5@hagrid>
Message-ID: <200204031619.g33GJXm02577@europa.research.att.com>

>> Specifically: If I want to write a function that answers a yes/no
>> question, I have lots of possible ways of spelling yes (1, 2, "yes",
>> and so on) and lots of possible ways of spelling no (0, {}, None,
>> and so on).  There isn't a single preferred way.

Fredrik> if you study real python code, you'll find that there is.

I mean that the language doesn't prefer a single way,
as opposed to programmers' conventions.







From alwyn at alwyn.demon.co.uk  Mon Apr 22 12:48:16 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Mon, 22 Apr 2002 16:48:16 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com>
Message-ID: <alwyn-D10DB2.17481022042002@news-text.blueyonder.co.uk>

In article <CmUw8.85199$XV5.7694212 at bin4.nnrp.aus1.giganews.com>,
 "Brett g Porter" <BgPorter at NOartlogicSPAM.com> wrote:
> 
> In what ways are Scheme and Lisp only distantly related?

As a Schemer, I'd be interested in an answer to that question as well. 


Alwyn


From tim.one at comcast.net  Tue Apr 16 14:52:36 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 14:52:36 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <GSYATKAqUGv8EwRX@jessikat.demon.co.uk>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMENPCPAA.tim.one@comcast.net>

[Robin Becker]
> On the office machine well I installed Python 2.2.1 last Friday evening.
> My unwise.exe is now gone as of Tuesday 10:00 am.
>
> I had a hard crash Monday morning.

Did you have a previous version of 2.2 on your machine, and uninstall it
before installing 2.2.1?

On a clean Win2K Pro box, I'm finally able to provoke this (or at least one
of its causes) reproducibly, via:

1. Boot the box.

2. Install 2.2.1.

3. Uninstall 2.2.1.

4. Install 2.2.1 again.  Note that \Python22\UNWISE.EXE is there.

5. Reboot the box (just logging off and on again isn't enough).

6. Now \Python22\UNWISE.EXE no longer exists.





From Oschler at earthlink.net  Thu Apr  4 18:19:41 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Thu, 4 Apr 2002 18:19:41 -0500
Subject: Rule Engine in Python
References: <eb612768.0203101940.5565b282@posting.google.com> <m3it82d32l.fsf@wilma.localdomain>
Message-ID: <I45r8.34886$0T6.10550048@e3500-atl1.usenetserver.com>

Harry George <hgg9140 at seanet.com> wrote in message
news:m3it82d32l.fsf at wilma.localdomain...
>
> If you are willing to use prolog as the rules language, there are
> python/prolog bindings (e.g., SWI prolog).  We use python and prolog
> jointly in a C++ CADenvironment.
>

Harry,

I'm a long term user of LPA Prolog, how does SWI Prolog compare and do you
know what the cost would be if I used SWI Prolog on a single Linux box
acting as a web server (that was consulting the Prolog Rule engine, accessed
via Python scripts running under Zope).

thx






From rasmussn at lanl.gov  Wed Apr 17 18:44:50 2002
From: rasmussn at lanl.gov (rasmussn at lanl.gov)
Date: Wed, 17 Apr 2002 16:44:50 -0600
Subject: [OT] What is Open Source?
In-Reply-To: <20020418002823.C4068@phd.pp.ru>
Message-ID: <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov>

On Wednesday, April 17, 2002, at 02:28 PM, Oleg Broytmann wrote:

>    It is not enough to have the code open. One needs the rights to read, 
> to
> use, to copy, to modify, to distribute modified and unmodified versions 
> (so
> everyone benefits from the open code). Only Free Software gives you such
> freedom and rights.

As I understand it, the GPL does not give you right to use the software
as you see fit.  You can only use it by linking with other GPL'd
software.

Craig





From maxm at mxm.dk  Fri Apr 12 04:24:43 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 12 Apr 2002 10:24:43 +0200
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it>
Message-ID: <3CB699CB.5050107@mxm.dk>

Alex Martelli wrote:

> I have my doubts that the BDFL wants to special-case the join method to 
> save 5 characters over a map call:
> 
>         ','.join(map(str, range(10))


Given the three choises discussed:

','.join(range(10), str)                # new proposal
','.join(map(str, range(10)))           # map
','.join([str(i) for i in range(10)]))  # list comprehension

I must agree. There isn't much saved and it's less flexible.

Damn I hate discussing with you!

     regards Max M



From francesco.contini at tiscalinet.it  Fri Apr  5 21:42:30 2002
From: francesco.contini at tiscalinet.it (FrancescO)
Date: Sat, 6 Apr 2002 04:42:30 +0200
Subject: BaseHTTPServer
Message-ID: <a8ljkn$6ht$1@lacerta.tiscalinet.it>

Hi!

i'm trying to write a sort of thing similar to an application server that
works
reading the uri requested like parameters (more or less like a querystring
for a cgi)
and send an xml structure for result.
My starting point is the BaseHTTPServer module, and following the example in
the documentation, i wrote all the code needed to work but...
if i try to stop and restart the server i have an exception:
port already in use

it seems that the connections with the clients aren't closed when i stop the
server...
after a minutes, or more,  the port is free... and i can restart the
program, but, obviously,
i cannot leave the server off for too much seconds.

I don't run a particular code to stop.. (kill -9 ;P ) and i like to use the
same way used
by the /etc/init.d/... scripts to start and stop.

there's something that i forget? maybe i must write a destructor for the
RequestHandler derived class?
there's a way to stop the server in a "clean way"?

Thanks a lot!

Francesco Contini
Software Developer
Excite Italia B.V.
www.excite.it




From borcis at geneva-link.ch  Wed Apr 10 18:08:41 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 11 Apr 2002 00:08:41 +0200
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <3CB49ADD.89B242A4@sympatico.ca>
Message-ID: <3CB4B7E9.D58F5E4B@geneva-link.ch>

Colin J. Williams wrote:
> 
[...]
> > Note this would introduce three-valued logic into Python, which would be
> > helluva confusing to beginners, who tend to understand the law of the
> > excluded middle implicitly.
> 
> I wonder about the confusion.  It would perhaps add clarity to an aspect of
> Python.
> 
> Let's look at some examples:

[snip over the instance traces]

> Logical operations
> 0 or  None, 1 or  None: None 1
> 0 and None, 1 and None: 0 None
>               not None: 1
> 
> With a three value scheme, the results would
> be the same for the first two, whereas the
> result of `not None` would itself be None.
> 
> I would appreciate comments.

Beautiful, and so what ? :) And there's indeed my own
use of None over Laura.

I'd say indeed that we should value Laura's intervention
not exactly for what she appears to have meant to mean,
but as an true alarm bell... I mean that if I take to the
psychic gesture to believe her, to make it plausible I have to
evoke an older idea of mine.

Tadah, OIOM :

[b] Python is unlike [a] GatesDesigns(tm) because :

Grammar is valued like a hidden beauty,
holding promises of evolution, not yet
to be fathomed, and this in the service
(choice where they differ) [a] of growing
GatesDesign(tm)'s mass of consumership
[b] of not wasting grammar pockets-of-untapped
-dynamics to silly-purposes-according-to
-better-common-interest.

IOW GvR is serious where BG isn't and LW too clever.

OK, now my current theory of Laura is that she's there
like an Except: clause that fired over the thread of PEPS
because at PEP 285 exactly Sharon invaded Palestine while
"the border of the inner square" of "aesthetic manna" Laura
"was guarding" underwent a "boundaries violation" because of
"quasifascist boolean pseudowisdom taking over python family
jewels in eg falsety gathering."

Should Python grow restartable interrupts to fully
solve the case ?

Should PEPS become mythified ? This needs order. Someone
proposed "chinese torture", I prefer (while I did not look)
to positivise the work, saying "in the ideal" the PEP process
should look like "a linear run of mutually orthogonal bytecodes
of what we'd need to call a 'crazy diagonal VM'".

Sorry, I see I'll have to write a couple PEPS, with no chance
to be taken seriously at having them see pass.

Mmmhhh. This gives me an idea. If we could efficiently market
Python as "God's very own choice of programming language" to
credulous masses, then a market geometry for adaptation of
Python patterns, metapatterns and minipatterns would spring
to existence :)

And this in turn would make it become natural for Python
to grow to the niche of being the only programming language
endowed with a so-called "boolean style sheets" with 
standard styles of the distribution being Style152 or
Style22.285 or StyleCyclotomicF or StyleIconIdol or
StyleDonJim :)

"Boolean style sheets" can certainly be sold as "refactoring
from __future__ statements"

Maybe something as simple as

"from __future__ StyleDonJim import booleans"

Mmmhh. This seems to lead us to somewhere. Three digits releases
tend to be bugfix releases closing their generation. Naked code
runs on them with at least an implied

from __future__ Style3DVersion import defaults

Where 3DVersion stands for the 3 digits of the version :)

> 
> Colin W.
> 

Cheers, Boris "30215" B.
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")


From pinard at iro.umontreal.ca  Mon Apr 29 14:01:42 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 14:01:42 -0400
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCD7A03.8A17CE41@cascade-sys.com>
References: <aaim48$7fd$1@bob.news.rcn.net>
	<eUTmiIA3MQz8EwSA@jessikat.fsnet.co.uk>
	<3CCD7A03.8A17CE41@cascade-sys.com>
Message-ID: <oqd6wipent.fsf@titan.progiciels-bpi.ca>

[James J. Besemer]

> I think we should be very reluctant about changes that break old code.
> Further we should focus on the important, substantial ones instead of
> dinking around with trivial crap like this.

Granted, Python should reasonably avoid breaking old code.  The proposed
PEP has the virtue of at least trying to propose an acceptable cleanup.
I guess we agree on the fact it fails to do so.

Surely, if we keep adding substance to Python without ever cleaning behind,
Python is going to progressively loose its simplicity (which probably
attracted us all at the beginning) and consequently, part of its legibility.
I'm not at all saying that the PEP should be accepted.  But yet, there is
a real problem in the long run, that we should accept to foresee.

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




From mark at mceahern.com  Wed Apr 24 15:43:22 2002
From: mark at mceahern.com (Mark McEahern)
Date: Wed, 24 Apr 2002 14:43:22 -0500
Subject: mxCGIPython
In-Reply-To: <20020424142536.A572@mlug.missouri.edu>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEHNKMAA.mark@mceahern.com>

Someone suggested trying again and that did the trick!

Thanks!

// mark




From jon+usenet at unequivocal.co.uk  Tue Apr  9 17:05:58 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 09 Apr 2002 21:05:58 -0000
Subject: CGI multiple select?
References: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>
Message-ID: <slrnab6ltm.hvu.jon+usenet@snowy.squish.net>

In article <56ns8.37728$tg4.448055 at vixen.cso.uiuc.edu>, John wrote:
> When only one entry is selected:
> form["spred"].value (will be the object selected)
> 
> When multiple entries were selected:
> for o in form["spred"]:
>   o.value (--> each of these will be the object selected)

If you use jonpy, you would name the form field "spred*" and it would
always be passed to you as a list ;-)

(Allowing the evil hacker to dictate the type of the input data seems
like bad design to me.)

http://jonpy.sourceforge.net/


From grey at despair.dmiyu.org  Fri Apr 12 22:26:38 2002
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Sat, 13 Apr 2002 02:26:38 -0000
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com>
Message-ID: <slrnabf5r2.s6b.grey@teleute.dmiyu.org>

On 12 Apr 2002 18:19:38 -0700, Paul Rubin <phr-n2002a at nightsong.com> wrote:
> Steve Lamb <grey at despair.dmiyu.org> writes:
>>     What about something like the Zaurus SL-5500?
 
> The SL5500 is a high powered computer with a 32-bit CPU and 64 MB of
> ram.  It runs Python just fine.  

    *Eyes Sharp's list of retailers for the SL-5500 a little closer.*
Re-e-eaaaalllly.  Hmmmm..

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------


From aleax at aleax.it  Thu Apr  4 01:17:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 06:17:23 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it> <3CAB87ED.C5567DBE@accessforall.nl>
Message-ID: <TdSq8.48142$pT1.1370343@news1.tin.it>

Ype Kingma wrote:
        ...
>> > The problem is that a lot of code has been written assuming 1/0
>> > behaviour of boolean values.
>> 
>> The "problem" is that behavior is VERY useful.  So, taking it out would
        ...
> I wasn't talking about removing this behaviour, although I don't like it.

When I call a language feature "VERY useful" it sort of implies I like it
(well, I AM an engineer -- "useful" and "beautiful" are almost synonyms
for our ilk:-) and therefore I object to it being called a _problem_:-).  I
put "problem" in quotes for the same purpose, to indicate a "so to speak".


> Thank you for pointing this out. I only read the contents of both parts
> (the language ref and the standard library ref) and based my post on that.
> I should have read the manual completely.

I agree that you should have.


>> This is the Python LANGUAGE we know and love -- the LANGUAGE
>> Reference says so, and us Language Lawyers know that.  You want to
>> change Python, that's one thing.  But don't tell us that this part of the
>> Language Reference "is not part of the python language".
> 
> I do not want to change that part of the language. I tried to
> say that the backward compatible 1/0 behaviour and
> the Something/Nothing behaviour can be implemented in the special methods
> of the standard types, even when a non int-a-new-hat bool would be added
> to the language.

Even if having to modify the __getitem__, __setitem__, __delitem__ (at
the least) of a zillion user-defined types out there was feasible (and it's
not), I think this would not fully exhaust, as I indicated.

> but they are already explicitly available. (How often do you
> actually need slices? I'd expect you to use bool more often
> than that if you could.)

I use slices in every sequence-like type I code -- a slice object
is what gets passed to my __getitem__ (etc) methods when
client code uses myobject[3:] or whatever (special methods
__getslice__ etc are deprecated for new objects, although for
backwards compatibility they're still used if present, and quite
unfortunately list &c still use them).  My __getitem__ &c need
to know whether they're called with an index that has .start
and .stop attributes (I feature-test for those, obviously, NOT
typetest for slice!) or with an index that is a number.  I feel
absolutely no need to know whether an argument passed to
my object's method comes from -- and on the other side of
things, to know whether my object is being tested for truth,
why, I already have the (unfortunately named) __nonzero__.

When this PEP passes and 2.3 becomes the only Python
release that my code must support I may have to switch to
"return True" vs "return 1" -- we're quite keen on the concept
of *TEAM* programming, against "code ownership", and part
of that is that everybody follows strict coding-style standards.
If we agree on coding-style standards mandating this, then I
will follow them -- not happily, but I will.  And that's the extent
of "use" (not "usefulness") I expect from booleans.


> Indexing a dict with both int and bool keys might involve some confusion
> when True is not 1 anymore. This case should be rare, however.

Why should it be rare?  All hashables are used uniformly as keys
into dictionaries.

> For standard types, ie. normal dictionaries there is no problem since
> one can simply add the int() method call in the implementation of
> __getitem__ of the standard dictionary for backwards compatibility.

You propose a substantial slow-down of standard dictionaries "for
backwards compatibility" against an incompatibility that need never
be introduced in the first place -- you can't "simply add the int()
method call", you first have to typetest against boolean on EVERY
dictionary access to know if the "int()" kludge is needed at all (you
can't just call int() on EVERY key without breaking a megazillion
lines of code -- i hope you realize that? -- and if you could that
would be a slow-down too).  I don't think you realize at all how
CRUCIAL dictionary look-up performance is to Python.  The number
of dictionary accesses is huge in every real Python program.  If there
is one thing that MUST be always optimized, that's the one.

The incompatibility is needless: apart from un-engineeringly yearnings
for "purity" (which, in Python, is trumped by pragmatics) there's no
reason for hash(True) to differ from 1, and True==1 is similarly not
a problem at all.  Easiest would be not to distinguish booleans from
ints at all, second easiest, as the PEP proposes, to make booleans
a subclas of ints, but even without the latter pragmatic fix a sensible
strategy would still be feasible.  Slowing down all dictionary access
is NOT a sensible strategy, nor acceptable at all.

Slowing down all list accesses as you want to do is not quite as
horrible, but still a pretty bad move.  What Python most needs right
now is a bit more attention to performance issues, NOT slow-downs
in its core to satisfy some people's desires for "purity".


> User code trying to implement a dictionary will need a similar
> fix as I pointed out above. I can only hope there is no more breakage in
> user code.

"only hope" is not acceptable in this context.  Fortunately, we can
do better than that (and I'm confident that the unacceptability of
slowing down dictionaries and lists is quite obvious to decision
makers in this case, so I don't fear it will happen).


Alex



From sholden at holdenweb.com  Mon Apr  1 16:49:49 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 16:49:49 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org>
Message-ID: <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>

"Gustavo Niemeyer" <niemeyer at conectiva.com> wrote in message
news:mailman.1017673984.1773.python-list at python.org...
> Hi Martin!
>
> > The standard library has 110 occurrences of "return 0", 70 occurrences
> > of "return 1", and 136 occurrences of "return None". Optimistically,
> > this makes 300 functions. It would help users of these 300 functions
> > if they would know whether those functions return a boolean or an
> > integral value (if integral, what range?); for the "return None"
> > cases, the boolean type would help to clarify whether this is "return
> > no value", "return False", or "return nil object";
>
> I hope nobody is planning to change "return None" to "return False".
> This would break a lot of code doing "ret == None".
>
Arguably, most of that code should have read

    if ret:
        ...

rather than

    if ret == None:
        ...

since the latter has the flavor of "if expr==True".

So, is breaking bad code more excusable than breaking good code ;-)?

regards
 Steve






From p.magwene at snet.net  Sat Apr 13 20:23:07 2002
From: p.magwene at snet.net (Paul Magwene)
Date: Sun, 14 Apr 2002 00:23:07 GMT
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <3cb8c690$0$19942$9b622d9e@news.freenet.de>
Message-ID: <pan.2002.04.13.20.23.07.90315.93808@snet.net>

On Sat, 13 Apr 2002 21:00:33 -0400, Jens Baader wrote:


> "Terry Reedy" <tejarex at yahoo.com> schrieb im Newsbeitrag
> news:WL2u8.26958$GS6.2775752 at bin3.nnrp.aus1.giganews.com...
> 
> I see. Thank you for your answers.
> 
> BTW, try this in interactive mode:
> 
> "The Python interpreter is very stable"*30000
> 
> At least at my system the interpreter promptly hangs itself. I guess I
> just found my first bug.. (not bad I just read the first few pages of
> the tutorial)
> 
> 
No, what you just encountered was the behavior of IDLE (a Tkinter based
interface to Python) when presented with large strings .  Eventually IDLE
will spit out the string - it took several minutes on one of my machines.
That's a "fault" of IDLE not of the language.

>From a purely text based command-prompt (bash on Unix or the Dos box on
Windows) there's no problem at all, either on Unix or Windows (though my
Unix boxes spit it out faster!).

If you try this:

>>> s = "I'm a snotty newcomer with an attitude" * 30000

You'll see that it returns almost immediately.

HTH,
Paul


From aleax at aleax.it  Thu Apr 18 16:10:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 20:10:52 GMT
Subject: PyQt QTable
References: <a9mksg$275d$1@news6.isdnet.net>
Message-ID: <gLFv8.2534$8D3.60744@news1.tin.it>

Michael Melchert wrote:

> self.connect(qtb,SIGNAL('clicked(....)'),self.the_method)] my problem
> arises from the parameters which are defined for
> QTable.clicked(int,int,int,QPoint&). what would be the correct way to
> write the connect() call, especially regarding the QPoint& reference? And
> how would the connected python method be defined?

import sys
import qt
import qttable
 
a = qt.QApplication(sys.argv)
 
table = qttable.QTable()
def tableClicked(a,b,c,d):
    print 'clicked',a,b,c,d
a.connect(table, qt.SIGNAL("clicked(int,int,int,const QPoint&)"), 
tableClicked)
 
a.setMainWidget(table)
table.show()
a.exec_loop()


Alex



From mwh at python.net  Tue Apr  2 03:44:36 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 2 Apr 2002 08:44:36 GMT
Subject: PyArg_BuildValue, how to return a PyString object?
References: <5a4226f0.0204012004.24bb7b00@posting.google.com>
Message-ID: <lkelhya3rg.fsf@elios.maths.bris.ac.uk>

kevin at cazabon.com (Kevin Cazabon) writes:

> I'm writing a C extension dll for Python, and am having a bit of
> trouble figuring out how to return a PyString_FromStringAndSize object
> (along with some normal stuff).
> 
> (I need to use PyString_FromStringAndSize because the string may
> contain null bytes, am I correct in this, or is there an easier way?)

Doubt it.

> Can I use something like this (it doesn't seem to work like this
> though):
> 
> PyObject *myPyStringObject = PyString_FromStringAndSize(cstringarray,
> size);
> PyObject *myReturnObject =  Py_BuildValue("llo", long1, long2,
> myPyStringObject);

I think you have to mae that "(llO)", don't you?

There might be a s# format code, I can't rememeber -- Py_BuildValue
*is* documented, though so you can check for yourself:

> Or do I have to build a tuple manually?

No.  Rememebr to DECREF "O" args after calling Py_BuildValue --
something I'm always forgetting.

Cheers,
M.


From martin at v.loewis.de  Sat Apr 13 18:57:37 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 00:57:37 +0200
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <m3k7rb1a1q.fsf@mira.informatik.hu-berlin.de>

"Jens Baader" <nospam at nospam.com> writes:

> I'm downloading Python 2.2.1 right now and I wonder
> why you still produce bugfix releases for the old 2.1 development line.

The main reason is that a really bad bug was found in 2.1.2 which
needed to get fixed.

> It seems that Python 2.2 is somewhat broken. If not what's the
> reason that keeps the people from upgrading to 2.2?

I guess people continue use Python 2.1 for two reasons:
- Zope (or some other package) is certified to work only for Python
  2.1.  If the software author tells you to use Python 2.1, you use
  that because that should cause least problems.
- Python 2.1 is what comes with several systems these days. For example,
  the next Debian release will include Python 2.1 as the default Python
  package; all packages that depend on Python will use 2.1. The rationale
  was that Python 2.2 was not complete at the time when the selection was
  made.

There are some backwards compatibility issues that may cause programs
to break when changing the Python version from 2.1 to 2.2, Andrew
Kuchling recently posted a list of changes he needed to make for his
software. I'd be curious to see similar lists for other packages.

> Another question: When will we see an official ISO/ANSI standard
> for Python? 

Probably never, atleast not in any foreseeable future. Even if people
started to work on an ISO standard today (which nobody currently plans
to), the standard would not be completed before, say, 2004. It takes a
lot of energy to produce such a standard, and nobody is willing to
invest this energy.

> I dislike working with nonstandardized languages.

Why is that?

> The fact that radical language changes are still made (the
> type/class unification) seem to suggest that Python is still not
> mature enough for a real standard.

I'm not sure what a "radical change" is in your view. The new-style
classes of Python 2.3 are a pure extension, not a radical change.

>From a standards point of view, conforming applications would have
continued to work - old style classes have seen no change, and the
only way to tell the difference is to look at the type of some
builtins.

If there was a real standard, it probably would deliberately leave out
implementation details (such as the type of the int builtin), and
merely mandate that it is callable - precisely to give implementations
the liberty to expose proprietary extensions (which newstyle classes
then would have been).

> Or maybe Python and
> the other "newschool" languages (like Ruby, Java) are not
> meant to by standardized? 

Nobody "means a language to be standardized". It is rarely the case
that a language springs into existance by somebody saying: I will
develop a new standard today.

> Don't know but I really would like to.

You would like to what? Develop a Python standard? Just go
ahead. Write it draft standard down, and contact your national
standardization body.

> Are there any plans to get Python "stable" or will the language
> continue to change with each major release of the interpreter?

Python will continue to evolve. Notice that almost every live language
evolves - even if there is a standard for the language.

"Stable" means different things to different people:
- does not break unexpectedly: Both Python 2.1.3 and Python 2.2.1
  are stable in this sense.
- permits to write code that works over a wide variety of
  implementations (ie. versions): Python always did that.
- does not break old applications "suddenly": PEP 5 defines
  the guidelines for language evolution to avoid sudden breakage.
- guarantees that all old applications continue to work the same way
  with all future releases: Python will never be stable in this
  sense, nor do I know any other language that has this property,
  yet provides new features in new versions.

Regards,
Martin


From max at alcyone.com  Tue Apr 16 14:30:52 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 11:30:52 -0700
Subject: Where should 'import' statement be put?
References: <4620daca.0204160749.1f68a62a@posting.google.com>
Message-ID: <3CBC6DDC.CCDA2553@alcyone.com>

gyromagnetic wrote:

> I was wondering what the 'best practice' is for locating 'import'
> statements. If I have a module comprising several classes, each of
> which has several functions, should the import statements be grouped
> - at the top of the module?
> - at the top of the classes in which they are relevant?
> - at the top of the functions in which they are relevant?
> 
> I believe imports are done only once regardless of the number of times
> the statement appears, but are there any other considerations here? I
> realize there may be scoping issues as well.

You're correct; importing modules is idempotent (only done once no
matter how many times an import statement is encountered), so this is
really a matter of style.

I nearly always put import statements at the top of the file, except in
the case where only one function uses them which may never get called. 
If, for instance, sys isn't used in the module proper, but _is_ used in
the main wrapper to scan for arguments, one might do something like:

	import StringIO
	import socket
	... etc. ...

	def main():
	    import sys
	    if len(sys.argv) ...

	if __name__ == '__main__': main()

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From martin at v.loewis.de  Sat Apr 13 05:58:59 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 13 Apr 2002 11:58:59 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl> <m3662wypqt.fsf@mira.informatik.hu-berlin.de> <3cb7f8a7$0$85437$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <m37knb9axo.fsf@mira.informatik.hu-berlin.de>

Boudewijn Rempt <boud at valdyas.org> writes:

> Oh, verdraaid, I see what I did wrong. On my previous computer, which
> has melted down, I had hacked sys so my default encoding was the same
> as that in my locale -- utf-8.
> 
> On my new computer, I just compiled and installed the latest Python,
> and I can only enter source in ASCII and escape sequences.
> 
> Was PEP 263 implemented already? Must investigate...

No, not in any released version. Meanwhile, I recommend to stay away
from Unicode literals if you need "funny" characters, and write

  unicode("<string containing funny characters>", "utf-8")

intead.

Regards,
Martin



From sholden at holdenweb.com  Mon Apr  1 08:05:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 08:05:24 -0500
Subject: Debugging flags [was: Re: PEP 285: Adding a bool type]
References: <mailman.1017466974.6751.python-list@python.org> <7xhemy1jo4.fsf@ruckus.brouhaha.com> <slrnaac6kf.2gs.philh@comuno.freeserve.co.uk> <slrnaacj3a.2gs.philh@comuno.freeserve.co.uk>
Message-ID: <lXYp8.108382$Ou.84158@atlpnn01.usenetserver.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
[...]
> I've gone through some of my herbivore code, looking to see where
> I've used 0 or 1 as a boolean. Example 1:
>
> debug = 0  # debugging this module?
>
> Frankly, if that is unclear to anyone, they are not cut out to be a
> programmer. Bear in mind that most non-programmers are aware of the
> 0/1 paradigm for boolean values: it appears on much electrical
> equipment.
>
I find it much more helpful to use

    debug = (1, 2, 7, 93) # debugging parsing, tree build, code generation
and error messages

with corresponding tests like

    if 7 in debug:
        ...

Given that when I'm not debugging the list is empty, these tests are not
unduly time-consuming. But then I've often gone over the top to allow
selective testing in complex developments.

regards
 Steve






From fredrik at pythonware.com  Wed Apr 10 11:43:12 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Apr 2002 15:43:12 GMT
Subject: PIL + McMillan Installer
References: <a91f3r$4o7$06$1@news.t-online.com>
Message-ID: <k4Zs8.244$p56.67402@newsb.telia.net>

Harald Schneider wrote:

> I use Activestate Python 2.1 and McMillan Installer 5.
> A script that uses PIL runs fine. But the one-file-distribution seems to
> have a problem with Image.open()
>
> ----
> img = Image.open(File)
>   File "C:\usr\python\PIL\Image.py", line 902, in open
>     raise IOError, "cannot identify image file"
> IOError: cannot identify image file
> ----

PIL's plugin scanner won't find the plugins when they're
embedded in the EXE file.  there's probably some installer
recipe somewhere, but if not, the explicit-import approach
described in this article might help:

http://www.pythonware.com/products/pil/articles/speeding-up-initialization.htm

</F>

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




From gh_pythonlist at gmx.de  Wed Apr 10 07:20:58 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 10 Apr 2002 13:20:58 +0200
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool type)
In-Reply-To: <OyUs8.2398$b62.59604@news1.tin.it>
References: <mailman.1018392859.8707.python-list@python.org> <OyUs8.2398$b62.59604@news1.tin.it>
Message-ID: <20020410112057.GA562@lilith.my-fqdn.de>

* Alex Martelli <aleax at aleax.it> [2002-04-10 10:34 +0000]:
> (let's PLEASE forget that unnamed language where you must use + only
> to add integers and +. to add floats... have they changed it yet?-).

No. Why is this so bad, except for overloading operators in subclassed
numeric types? Dunno if Ocaml can even do that.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 8.0 ?C      Wind: 6.6 m/s




From gcordova at hebmex.com  Mon Apr 22 12:38:06 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 22 Apr 2002 11:38:06 -0500
Subject: constructors in python
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062338E651A@mail.hebmex.com>

> 
> But what does happened when I forget (I'm just a man) to call 
> father's c-or?
> Semi-initialized object?
>

Well, you get buggy code; after all, you're just a man.

>
>C++ doesn't permit not to call the base c-or.
> 

And thanks to all the Gods and Goddesses for that.

-gustavo




From dalke at dalkescientific.com  Wed Apr 10 05:46:14 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 10 Apr 2002 03:46:14 -0600
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <a90rv2$fvh$1@slb4.atl.mindspring.net> <Xns91EC17861D8C7RASXnewsDFE1@130.133.1.4>
Message-ID: <a911vf$cp1$1@slb0.atl.mindspring.net>

Philip Swartzleonard:
>Wouldn't a more likly WoP around here be something like 'NI!' =)

Nah.  That went out of popularity in 1.4 when the 'ni' module
was replaced by 'knee'.  But it wasn't meant to be used and I
just found out that knee.py was removed from Lib a couple days
ago and moved to Demo/imputil.

So the proper WoP in that lineage is "IMPUTIL!"  Or maybe "IMP".
Or perhaps "BUILT-IN".  It depends, really.

Regardless, Python is now no longer the language that says "NI!".

                    Andrew
                    dalke at dalkescientific.com





From fperez528 at yahoo.com  Tue Apr  9 18:07:08 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 09 Apr 2002 16:07:08 -0600
Subject: property() type that only protects set?
References: <3cb34fb8$1_4@goliath.newsgroups.com> <3CB357E3.95EFC8E1@metaslash.com> <3cb358cb$1_2@goliath.newsgroups.com>
Message-ID: <a8vone$rmt$3@peabody.colorado.edu>

Brad Clements wrote:

> If you leave out fget (that is, let it default to None) then you have an
> attribute that cannot be read.
> 
> This isn't what I want.. I want an attr that can be read without a special
> getter function
> 
> (something like
> 
> myprop = property(self.__getattr,my_set)

Your request sounds reasonable enough to almost (IMHO) qualify as a bug report 
on mis-functionality of property(). It sounds to me reasonable to expect that 
un-specified property helpers would default to the normal behavior, instead 
of this all-or-none deal which seems to be the current behavior.

But I may be off-base on this one, I've only looked superficially at the 
issue.

Cheers,

f.


From martin at v.loewis.de  Wed Apr 17 15:38:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 21:38:50 +0200
Subject: minidom and =?iso-8859-1?q?=E5=E4=F6?= once again :-P
References: <WXdv8.2962$ZN4.5053@news2.bredband.com> <j47kn6bb5t.fsf@informatik.hu-berlin.de> <asjv8.3473$ZN4.5422@news2.bredband.com>
Message-ID: <m3662qyv1x.fsf@mira.informatik.hu-berlin.de>

Magnus Heino <magnus.heino at sigma.pleon.se> writes:

> Well, I got confused by this;
> 
> http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.102.htp
> 
> But I have removed that now.

Yes, that's a confusing FAQ item - it suggests as a solution what
really is just a work-around.

> Maybe someone with commit access could add some short example like this one 
> to the python documentation for minidom?

Feel free to add a patch to SF; somebody with commit access will then
happily integrate it :-)

Actually, I think .toxml should optionally accept the requested output
encoding, and default to 'utf-8'. If you can come up with a patch for
that as well, that would be even better. Extra points if you still
manage to make it return Unicode objects upon request, to allow
porting of applications that currently expect Unicode objects in
return.

Regards,
Martin


From rpm1deletethis at frontiernet.net  Mon Apr 29 21:17:55 2002
From: rpm1deletethis at frontiernet.net (Patrick)
Date: Mon, 29 Apr 2002 21:17:55 -0400
Subject: Access violation trying to run Python over Win NT network.
Message-ID: <ucrs88tkj2ttc7@corp.supernews.com>

I'm trying to run Python from a network drive.  On most machines it works,
but on some machines it generates an access violation about 60% of the time,
(just by typing 'python' with no arguments).  If I install Python locally it
runs fine on all the machines.

I would really appreciate any help.  I've been trying to get the people at
work to use Python and they have finally agreed to give it a try, but they
don't want to install it on everyones machines, so we have to run from a
network drive.

Thanks,

Patrick





From baas at ira.uka.de  Thu Apr 18 08:00:37 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Thu, 18 Apr 2002 14:00:37 +0200
Subject: Finalizing C extension modules
Message-ID: <7kctbustl4vsrp54fk7i7s8iu3ecf90oko@4ax.com>

Hi,

I have a question concerning extension modules written in C/C++.
Suppose I have a C/C++ module that allocates some resources in its
init function, where would be the proper place to release them again
when the Python interpreter is shutting down? Is there an equivalent
to Py_Finalize() for extension modules? 

There's this part in the docs of Py_Finalize() which makes me a bit
uneasy: 

"Dynamically loaded extension modules loaded by Python are not
unloaded. [...] Some memory allocated by extension modules may not be
freed."

Why is it that extension modules aren't unloaded?

So to be a bit more specific about my problem. I have a C extension
module containing some functions that are supposed to return a
specific pure Python object from another pure Python module. So I
import this Python module in the init function, get the class object
and keep it so that my own functions can create instances of this
Python class and return them to the caller. However, I will never call
Py_DECREF() for that class object I was retrieving from the pure
Python module. Is this ok? Or will this result in a memory leak? Where
am I supposed to call Py_DECREF()?

- Matthias -



From robin at jessikat.fsnet.co.uk  Wed Apr 24 08:00:27 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 24 Apr 2002 13:00:27 +0100
Subject: Python call tree tool?
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>
 <lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk>
 <7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>
 <mailman.1019641424.6210.python-list@python.org>
Message-ID: <ita58uAb5px8Ew2A@jessikat.demon.co.uk>

In article <mailman.1019641424.6210.python-list at python.org>, Graham
Ashton <gashton at cmedltd.com> writes
>On Tue, 2002-04-23 at 18:37, Robin Becker wrote:
>
>> trace.py seems to trace the execution of code. What I want is the call
>> graph/inverted call graph a la CFlow so that I can determine which
>> functions/classes are not called or referred to in some set of modules.
>
>It sounds like you're after a code coverage tool. This might have some
>ideas in it:
>
>  http://www.garethrees.org/2001/12/04/python-coverage/
>
that looks useful.
-- 
Robin Becker


From whisper at oz.net  Fri Apr  5 22:54:30 2002
From: whisper at oz.net (David LeBlanc)
Date: 6 Apr 2002 03:54:30 GMT
Subject: [Python-Dev] ACCEPTED: PEP 285 (truth pickles)
References: <mailman.1017974474.30402.python-list@python.org>
Message-ID: <a8lrhm$e8t$0@216.39.170.247>

Tim Peters wrote:
> [Roy Smith]
> 
>>If I pickle a bool, and then restore it, do I get the same singleton
>>back?
> 
> 
> Yes.  However, booling a pickle is a different question entirely.
> 
> don't-even-think-about-it-ly y'rs  - tim
> 
> 

Then there is the matter of Tim Peter's pickle, but i'm not going there LOL.



From maxm at mxm.dk  Thu Apr 18 04:11:37 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 18 Apr 2002 10:11:37 +0200
Subject: My First .py
References: <6Juv8.134255$QB1.35129169@typhoon3.we.ipsvc.net>
Message-ID: <3CBE7FB9.9080907@mxm.dk>

Bryce C. Mason wrote:

> I think I'll go get a book on how to let python retrieve information over
> the net.  This would remove another step of him having to capture screen
> output.  Anyone have a good suggestion for some examples of how python can
> manipulate network resources?

ahh you hardly need a book for just that :-)

But I would recommend "Learning Python" and the "Python Essential 
Reference" Or the one By Frederik Lundh.

regards Max M


###########################

from urllib import urlopen

pythonOrg = urlopen('http://www.google.com')
content = pythonOrg.read()
pythonOrg.close()

print 
content[content.find('<title>')+len('<title>'):content.find('</title>')]

 >>> Google




From phr-n2002a at nightsong.com  Wed Apr  3 20:58:42 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 17:58:42 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <7x3cyc5inh.fsf@ruckus.brouhaha.com>

vio <vmilitaru at sympatico.ca> writes:
> I need to translate a (rather voluminous 100MB) java program to
> python. After a few searches on google, apparently the best tools
> for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> on the Lex-YACC HOWTO, I would be very interested in hearing of
> others' successful experiences/feedback at this. Any pointers of
> interest would similarily be appreciated.

Why do you want to do this?  I think if you do it, you'll wish you hadn't.
An automatic translation really isn't what you want.  The languages
are just too different.  Parsing is the least of your problems.

What is the application, if you can say?


From eppstein at ics.uci.edu  Wed Apr  3 17:11:14 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 03 Apr 2002 14:11:14 -0800
Subject: PEP 284, Integer for-loops
References: <eppstein-BB3186.08413106032002@news.service.uci.edu> <slrna8d4bj.utc.grey@teleute.dmiyu.org> <eppstein-07CB42.14230206032002@news.service.uci.edu> <3c87d103.26130283@news.laplaza.org> <eppstein-B61874.13474907032002@news.service.uci.edu> <3c87f022.34099342@news.laplaza.org>
Message-ID: <eppstein-B55E0C.14111403042002@news.service.uci.edu>

In article <3c87f022.34099342 at news.laplaza.org>,
 mats at laplaza.org (Mats Wichmann) wrote:

> :I don't want to teach Python.  I want to teach other stuff.  
> 
> Right.  That's why I separated that out:  the intesting part is
> looping over multiple values.  range() happens to be a way to generate
> those values, but not the only way by any means, and that you don't
> like the syntax of range should have no effect on the teaching
> objective of showing how the concept of loops works....
> Mats Wichmann

Well, that's not what I want to teach either -- I want to show them how 
more advanced algorithms work, they should already have learned how 
loops work in their freshman classes or before.

Python's method of looping over general sequences is powerful and 
useful, no question.  But sometimes ranges of integers are exactly what 
you want to loop over, and for some types of integer range the Python 
range() syntax is very awkward.  Especially if you want to spend your 
attention on what you're trying to accomplish with the integer range 
loop, and not on the mechanics of coding it.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From nospam at nospam.com  Sat Apr 13 20:00:33 2002
From: nospam at nospam.com (Jens Baader)
Date: Sun, 14 Apr 2002 02:00:33 +0200
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com>
Message-ID: <3cb8c690$0$19942$9b622d9e@news.freenet.de>

"Terry Reedy" <tejarex at yahoo.com> schrieb im Newsbeitrag
news:WL2u8.26958$GS6.2775752 at bin3.nnrp.aus1.giganews.com...

I see. Thank you for your answers.

BTW, try this in interactive mode:

"The Python interpreter is very stable"*30000

At least at my system the interpreter promptly
hangs itself. I guess I just found my first bug..
(not bad I just read the first few pages of the
tutorial)

Jens






From starx at pacbell.net  Mon Apr  8 07:17:03 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 11:17:03 GMT
Subject: Win32 .pyc icon suggestion
References: <439969A9369AE4439637588ADB9B434767861B@weblink73.weblink.com.au> <mailman.1018238716.32275.python-list@python.org>
Message-ID: <Xns91EA2BC457492RASXnewsDFE1@130.133.1.4>

Gerhard =?iso-8859-15?Q?H=E4ring?= || Sun 07 Apr 2002 09:04:41p:

> * Peter Aylett <petera at weblink.com.au> [2002-04-08 13:49 +1000]:
>> Hi all,
>> 
>> Just a small idea:
>> I find that there is very little distinction between the icon for the
>> .py files and the .pyc files, which slows down the process of opening
>> and managing my program files.
>> 
>> Here's a replacement icon I suggest for the .pyc files that makes life
>> a bit easier, if anyone's interested.  (basically coloured yellow
>> instead of green).
> 
> Would you mind uploading this as a patch to Sourceforge, so it doesn't
> get forgotten? If you don't have an idea (I don't either), where in the
> Python distribution this fits, just open up a new patch and only upload
> the icon file instead of a full-blown patch.
> 
> Gerhard

*figures out how to download attachment with Xnews for maybe the thrid 
time ever*

That's nice. Some time back i had the same problem, and mad a red varient 
of the closed-eyes snake-head icon for .pycs, a blue version of the full 
snake for .pyws, and something I don't even remember for .pyos (cause I 
never bothered seperating the file type info for .pyc and .pyo into two 
seperate entries).  I'll toss em up there too incase anyone wants em 
(tommrow or sometime when i shouln't be sleeping instead =).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From ben at wblogan.net  Tue Apr  9 02:32:08 2002
From: ben at wblogan.net (Ben Logan)
Date: Tue, 9 Apr 2002 02:32:08 -0400
Subject: Fwd: Mapscript and Python: no mapscriptc module.
Message-ID: <20020409023208.A27398@wblogan.net>

I originally sent this to the MapServer mailing list, but no-one has
has responded.  The references to documentation, etc. are to MapServer
docs.  Perhaps someone here can help?

Thanks,
Ben

----- Forwarded message from Ben Logan <ben at wblogan.net> -----

From: Ben Logan <ben at wblogan.net>
To: mapserver-users at lists.gis.umn.edu
Subject: Mapscript and Python: no mapscriptc module.

I've been searching the docs and list archives, but haven't found much
on getting the Python Mapscript wrapper to work.  I changed to the
mapscript/python directory, ran python, and tried to import
mapscript.py, but got the following error:

Python 1.5.2 (#1, Mar  3 2001, 01:35:43)  [GCC 2.96 20000731 (Red Hat
Linux 7.1 2 on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import mapscript
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "mapscript.py", line 2, in ?
    import mapscriptc
ImportError: No module named mapscriptc

Here's the contents of the directory:

CVS
examples
mapscript.i
mapscript.py
mapscript.pyc
mapscript_wrap.c

As you can see, there is no mapscriptc.* .  From the messages in the
list archives is appears that I need to compile mapscript_wrap.c into
mapscriptcmodule.so (since I have a mapscript_wrap.c, I'm assuming
that I don't need to run SWIG--if so I'll have to install it first).

I ran

gcc -c mapscript_wrap.c -I/usr/include/python1.5 -o mapscriptcmodule.o

without any errors.  I then ran

ld -shared mapscriptcmodule.o -o mapscriptcmodule.so

also successful.  However, now I get the same error as the OP:

Python 1.5.2 (#1, Mar  3 2001, 01:35:43)  [GCC 2.96 20000731 (Red Hat
Linux 7.1 2 on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import mapscript
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "mapscript.py", line 2, in ?
    import mapscriptc
ImportError: ./mapscriptcmodule.so: undefined symbol: ms_error

Norman Vine responded to his post and attached a setup script to link
the module with the necessary mapscript libraries, but the list
archive software won't let me download it.

I tried

ld -o mapscriptcmodule.so mapscriptcmodule.o -shared ../../libmap.a

and it seemed to work, but now I get this error:

Python 1.5.2 (#1, Mar  3 2001, 01:35:43)  [GCC 2.96 20000731 (Red Hat
Linux 7.1 2 on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import mapscript
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "mapscript.py", line 2, in ?
    import mapscriptc
ImportError: ./mapscriptcmodule.so: undefined symbol: gdImageCreate

I have a feeling this could go on for a while.  As you can probably
tell, I don't know too much about building libraries. :) Can anyone
help me out?  BTW, I'm using MapServer 3.5 and you can see my system
setup from the errors above.

Thanks,
Ben

-- 
Ben Logan: ben at wblogan dot net
OpenPGP Key KeyID: A1ADD1F0

((lambda (foo) (bar foo)) (baz))

----- End forwarded message -----

-- 
Ben Logan: ben at wblogan dot net
OpenPGP Key KeyID: A1ADD1F0

I'm pretending I'm pulling in a TROUT!  Am I doing it correctly??




From ak at silmarill.org  Sun Apr  7 10:47:56 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sun, 07 Apr 2002 14:47:56 GMT
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com>
Message-ID: <slrnab0n45.rrn.ak@ak.silmarill.org>

In article <20020407103654.19127.00002640 at mb-fm.news.cs.com>, Jim
Williams wrote:
> Hi all...I'm new to Python and really enjoying it so far, but I am
> somewhat
> perplexed that as far as I have seen it has no facilities for printing to a
> printer?  Surely this is possible. Someone please enlighten me...I have done a
> good bit of research and searching and can find no reference to how to do this.
> 
It's very platform-dependant; on unix, if your lp command is set up
correctly, you can save to a text file and then do os.system("lp
file.txt"). I don't know about windows.

> 


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From philh at comuno.freeserve.co.uk  Tue Apr 16 18:20:59 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:20:59 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <3CBB83B1.8EC2E2F6@python.org>
Message-ID: <slrnabp8ub.ipr.philh@comuno.freeserve.co.uk>

On Tue, 16 Apr 2002 01:48:56 GMT, Guido van Rossum <guido at python.org> wrote:
>phil hunt wrote:
>
>> IMO until Python developers adopt a policy of being very reluctant
>> to do this, many IT managers will be wary of using it in anything
>> other thsan toy projects, and Python's user base will not grow as
>> quickly as (IMO) it should.
>
>If in your opinion Python should grow faster, perhaps you could
>help out by contributing rather than by telling us what to do...

I would contribute, but you have just said that you won't accept my 
proposed changes.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From uwe at rocksport.de  Mon Apr 29 17:26:09 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 29 Apr 2002 21:26:09 GMT
Subject: wxHtml - newbie
References: <aajv7m$3hl$1@wanadoo.fr>
Message-ID: <aakdph$8om5c$1@hades.rz.uni-sb.de>

Emmanuelgagniere <emmanuelgagniere at wanadoo.fr> wrote:
| Hello,

| Do you know where can I find an example of a simple application using
| wxHtmlWindow?

| Thanks.

have a look at the examples in the wxWindows-distribution. The
translation from wxWin to wxPy is quite easy...

greetings, uwe.



-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From peter at engcorp.com  Sun Apr 21 12:47:51 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 21 Apr 2002 12:47:51 -0400
Subject: constructors in python
References: <a9uoa0$t0r$1@news.iucc.ac.il>
Message-ID: <3CC2ED37.3D9225B3@engcorp.com>

eli li wrote:
> 
> Consider, we have two classes: one derived from other. In derived class c-or
> how is it possible to call the base c-or? Simple call base.__init__ (...)?

Assuming c-or means constructor:

[untested code]

class Base:
   def __init__(self):
       return

class Derived(Base):
   def __init__(self):
       Base.__init__(self)
       return


-Peter


From JamesL at Lugoj.com  Sat Apr  6 13:52:26 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 18:52:26 GMT
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <m3bscwhgkb.fsf@localhost.localdomain>
Message-ID: <Xns91E86F533194JamesLLugojcom@216.148.53.83>

Steingrim Dovland <steingrd at student.matnat.uio.no> wrote:
> Robert Spielmann <varial at gmx.de> writes:
> 
>> is there a way in python to check if a number is integer? My problem
>> is the following: I need to know if sqrt(x) is an integer value or if
>> there is a fractional part. 
> 
> You can do it using the modulo-operator %.  An integer value % 1 
> will return 0 and a non-integer value % 1 will return something
> greater than 0.

Another way is to try something like:

import math

def IsWhole(x):
   return x == math.floor(x)

if IsWhole(math.sqrt(4.0)):
   print "Whole number"


From goodger at users.sourceforge.net  Tue Apr 16 21:12:06 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Wed, 17 Apr 2002 01:12:06 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com> <a9fu9g$mj3$1@watserv3.uwaterloo.ca> <3CBBAB60.2CC717D7@alcyone.com> <slrnabp974.ipr.philh@comuno.freeserve.co.uk> <3CBCB42B.CB515B66@alcyone.com>
Message-ID: <B8E24468.21A38%goodger@users.sourceforge.net>

Erik Max Francis wrote:
>> I'm not suggesting removing it, merely making it optional. People
>> can still use it if they want to.
> 
> Well, this is all academic now anyway, since Guido Has Spoken.

Phil, I encourage you to write up a PEP anyhow.  There's a slim chance you
may be able to persuade Guido.  At the very least a rejected PEP would do
the service of documenting the proposal for posterity.  From PEP 1:

    A PEP can also be `Rejected'.  Perhaps after all is said and done
    it was not a good idea.  It is still important to have a record of
    this fact.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From fgeiger at datec.at  Tue Apr 23 12:56:36 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Tue, 23 Apr 2002 18:56:36 +0200
Subject: Class Methods with Python 2.1
References: <3CC57C5F.7040804@thomas-guettler.de>
Message-ID: <aa43p3$123$1@newsreaderg1.core.theplanet.net>

"Thomas Guettler" <zopestoller at thomas-guettler.de> schrieb im Newsbeitrag
news:3CC57C5F.7040804 at thomas-guettler.de...
> Is it possible to have methods in classes
> which don't take a self argument, because
> they only change attributes common to all
> instances of a class?
>
> class car:
>      available=["porsche", "volvo", "trabbi"]
>      def __init__(self, name):
>         if name not in available:
>             raise "Name not in available cars"
>         ....
>      def add_available_car(name):
>          availalble.append(name)
>
> One solutionn is to define this method
> at module level. But that's not nice.
>
>
>
> Unfortunately I have to stick to Python2.1
>
>    thomas
>




From usenet at minus1.de  Sat Apr 20 12:03:10 2002
From: usenet at minus1.de (Konrad Anton)
Date: Sat, 20 Apr 2002 18:03:10 +0200
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com>
Message-ID: <slrnac349u.ea.usenet@fohnhab.minus1.de>

Peter Saffrey <theoryboy at my-deja.com> schrieb:
>When I place some buttons on a frame I don't seem to be allowed to use
>arguments in the function call attached to the button. If I do use
>arguments, or even just empty brackets, the function is called when
>the button is placed, not when the button is clicked on.
>(...)
>Button(win, text='print 3', command=printNumber(3)).pack()
>Button(win, text='print 5', command=printNumber(5)).pack()

As long as there is no native support for currying in the Python
language, you can still use 

  Button(win, text='print 3', command=lambda : printNumber(3)).pack()

which will create a zero-argument anonymous function that will call
printNumber(3) when it is called. 


HTH 
Konrad


-- 
 Konrad Anton <konrad at minus1.de> ; Web: www.minus1.de ; gpg: 0x22954D8A
 Tel. +49-761-881-2122, Fax +49-721-151318943.  Alice:
Yes! 


From huaiyu at gauss.almadan.ibm.com  Tue Apr 30 20:52:29 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 1 May 2002 00:52:29 +0000 (UTC)
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
References: <aan5hv$aor$1@bob.news.rcn.net> <ucuagjg8it3v4b@news.supernews.com>
Message-ID: <slrnacuf2d.9l1.huaiyu@gauss.almadan.ibm.com>

John Roth <johnroth at ameritech.net> wrote:
>I strongly disagree about apply(). The only use I've found for it
>varies the function which it calls, not the parameter lists! Having
>variable parameter lists doesn't hack it - I need the variable
>function.

>>> from math import *
>>> a = (2.3,)
>>> for f in [sin, cos, sqrt]: print apply(f, a)
... 
0.745705212177
-0.66627602128
1.51657508881
>>> for f in [sin, cos, sqrt]: print f(*a)
... 
0.745705212177
-0.66627602128
1.51657508881

Huaiyu


From sholden at holdenweb.com  Wed Apr 17 09:56:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 09:56:25 -0400
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <87vgaswc1s.fsf@tleepslib.sk.tsukuba.ac.jp> <7934d084.0204161757.72ff525f@posting.google.com> <87u1qatvc7.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <9cfv8.13747$hl2.2413@atlpnn01.usenetserver.com>

"Stephen J. Turnbull" <stephen at xemacs.org> wrote in message
news:87u1qatvc7.fsf at tleepslib.sk.tsukuba.ac.jp...
> To begin with, let me apologize to Dr. Mertz for the unwarranted ad
> hominem smear I committed.  That was evil.
>
[ ... ]
>
>     Andrae> To attempt to cast someone who objects to this as some
>     Andrae> raving idealogue,
>
> You're right, I was wrong to do this.
>
>     Andrae> and thereby dismiss them with casual wave handling as a
>     Andrae> self declared 'seeker of truth' (tm) is the most appalling
>     Andrae> piece of intellectual hypocracy I have seen for some time.
>
> YBT.
>
> Actually, I have to laugh at myself, because I was in fact wrong to
> assert that the sets of licenses/covered programs are the same, and
> Huaiyu led me to truth.  Believe me, I didn't plan that, but there is
> proof: I was seeking (despite being fairly sure I already had the
> answer) and I found.<wink>
>
> Satori-remains-out-of-reach-ly y'rs
>
Good God! Honest, open-minded acceptance of criticism? Do you realise you
are flying in the face of Usenet tradition? Am I going to have to migrate to
the other c.l.p... to see real blood and guts?

Seriously: it's good to see such open-mindedness accompanied by a sensoie of
humor. I think more of you for it.

not-that-you-need-care-about-that-ly y'rs  - steve






From tundra at tundraware.com  Mon Apr 22 02:10:02 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 06:10:02 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
Message-ID: <3CC3A89C.BE2BF65@tundraware.com>

Carl Banks wrote:
> 
<SNIP>

> 
> Sorry, but I don't agree that algorithms, or any mathematical concept,
> is an invention of the human mind.  The algorithms have existed since
> the beginning of time.  The quicksort had the property of requiring ~
> n^2 comparisons in the worst case before it had ever been used or
> applied.  We merely discovered this algorithm; we didn't create it.
> 
> So I say mathematics, including the study of algorithms, is a natural
> science.  And, for now, that's all I have to say about it.

I disagree strongly (but am willing to be convinced otherwise).  All of
mathematics is a formal construct of the human mind created with the
intent of removing ambiguity and enhancing our ability to describe what
we *think* and what we *observe*.  We can (and do) create all kinds of new
calculi that are problem- or domain-specific.  That is, we *invent*
mathematics to suit our needs - mathematics is not one-for-one
correspondent to the natural universe.

e.g., Calling the upper-bound of Quicksort N^2 is merely one way of looking at 
the time complexity of this "idea".  There is not some hardwired constant
in the universe that declares heapsort to be O*(N^2) - this is merely an
artifact of the calculus used for Big-O analysis.

The reason I take this position is pretty simple.  We are able to construct
mathematical systems which describe intellectual abstractions for which
there is no natural analog.   That is, mathematics can embrace way more than
just the natural world and its workings.  (Theoretical mathematicians are famous
for their contempt of "mere" Applied mathematicians whom they view as operating in a
very limited, mechanical intellectual domain.)  For instance, even the theoretical
physicists only need about 5 dimensions to describe Life, The Universe, and
Everything, but mathematicians routinely work in arbitrary "n-space" for which
there is not physical/natural correspondent.  Similarly, a mathematician
can introduce ideas like an "Incompleteness Theorem" which is a meta-mathematical
commentary on mathematics itself - something far beyond the domain and range of
mere natural science.

I'm not choosing sides here on the math vs. science debate - both are important
artifacts of the human intellect.  I merely take umbrage with the notion that
mathematics is innately wired into the universe somehow.  


------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From geoff at gerrietts.net  Tue Apr 16 16:57:06 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 16 Apr 2002 13:57:06 -0700
Subject: Safe eval?
In-Reply-To: <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE>
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <20020416205706.GD20804@isis.gerrietts.net>

Quoting Philipp Lenssen (lenssen at hitnet.rwth-aachen.de):
> I just realized, as opposed to PHP, there's much stricter namespaces in
> Python and maybe not using "import" on anything harmful could be the
> solution?

You can hack on the rexec module to make it do what you want to do;
that's what I do. It may or may not provide ample security for your
purposes; it's certainly better than a raw eval().

The code I use is attached; under 2.something, this breaks unless you
override make_re to pass.

Thanks,
--G.

-- 
Geoff Gerrietts             "There is no fate that cannot be 
<geoff at gerrietts net>     surmounted by scorn." --Albert Camus

import rexec

class SEval(rexec.RExec):
    nok_builtin_names = tuple(__builtins__.keys())
    ok_builtin_modules = ()
    ok_path = ()
    ok_posix_names = ()
    ok_sys_names = ()

    def __crunch(*args):
        raise SystemError, "Untrusted code did bad things."

    def make_re(*args):
        pass

    r_execfile = s_eval = s_exec = r_exec = __crunch
    s_execfile = r_open = r_reload = r_import = __crunch
    r_unload = s_import = s_reload = s_unload = __crunch

se = SEval()

def secure_eval(string):
    try:
        res = se.r_eval(string)
        return res
    except:
        log("Untrusted code evaluated to an error: %s" % string)
        raise




From slzatz at hotmail.com  Sun Apr  7 22:58:46 2002
From: slzatz at hotmail.com (Steve Zatz)
Date: Mon, 08 Apr 2002 02:58:46 GMT
Subject: radiobutton menu item issue on Windows
Message-ID: <GH7s8.206591$u77.42778741@news02.optonline.net>

On W2K and XP, when I run Tkinter programs with radiobutton menu items, the
check mark is not visible until the mouse is run over the selected menu
item.  I can see a small fraction of the checkmark before I run the cursor
over the item but the full check mark isn't visible until the cursor runs
over the item.  To be clear, the menu item was selected previously, however
when you return to that menu, the selected menu item only has a partially
visible check until you move the cursor over the item.  And when you do
bring the check mark 'to life' by rolling the cursor over it, the check mark
overwrites most of the first letter of the menu text.

Has anyone seen this?  I realize the problem is more aesthetic than fatal,
but would like to fix if possible.  I am using whatever version of Tkinter
and Tcl/Tk came with the standard Python 2.2 Windows release.






From timr at probo.com  Mon Apr 22 14:10:16 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 22 Apr 2002 11:10:16 -0700
Subject: (Win32) Mapping WebDAV to Drive Letter
References: <3CC00F34.8000807@thomas-guettler.de>
Message-ID: <jck8cukmp84qeae0qq3j0vnvgdou9jdrj7@4ax.com>

Thomas Guettler <zopestoller at thomas-guettler.de> wrote:
>
>Does someone know how you could use the win32 api
>with python to map a webdav server to a drive letter?
>
>There is a C library for the webdav part. But I have
>no clue how you could map it to a drive letter.

This requires an "installable file system driver", which is a kernel-mode
component.  Unless you are a driver writer, believe me when I say you do
NOT want to go there.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From phr-n2002a at nightsong.com  Tue Apr 23 00:03:02 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 22 Apr 2002 21:03:02 -0700
Subject: need encryption algorithm
References: <mailman.1019528622.26907.python-list@python.org>
Message-ID: <7x1yd7yscp.fsf@ruckus.brouhaha.com>

Anthony Baxter <anthony at interlink.com.au> writes:
> base64 encode whatever you're planning to use. easy peasy.

You have to be careful with base64, since it uses some non-alphabetic
characters that confuse some systems (maybe not Oracle).  Hexadecimal
(base 16) is the safest method of dealing with this problem.



From DeepBleu at DeepBleu.org  Thu Apr 18 13:58:53 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Thu, 18 Apr 2002 12:58:53 -0500
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com>
Message-ID: <ubu2a0dbso0u66@corp.supernews.com>

How are you importing the DCOracle2 module, and what is *the exact* error
that you are getting?
DeepBleu

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in message
news:e2e9e807.0204171619.46de6db1 at posting.google.com...
> Hi All, I can import the DCOracle2 module fine from the interactive
> interpreter but my CGI script causes an internal server error. The
> interactive interpreter is running on the same machine as my scripts.
> A tail of the apache error log reveals that my script can't import the
> DCOracle2 module.
> Any ideas anyone?
>
> Hugh




From peter at engcorp.com  Sun Apr  7 18:51:32 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 18:51:32 -0400
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4>
Message-ID: <3CB0CD74.B1696987@engcorp.com>

> >>I'm using the time module to make part of the game random and
[...]
> >
> > Use the random module.
> 
> Yep. For example:
> 
> ---
> import random
> 
> # No vaule - uses time on it's own
> random.seed()

Random seeds itself using time.time() so this is unnecessary 
in simple code.

-Peter


From mlerner at umich.edu  Mon Apr 29 23:28:45 2002
From: mlerner at umich.edu (Michael George Lerner)
Date: Mon, 29 Apr 2002 23:28:45 -0400 (EDT)
Subject: Newbie question: how to join a list of elements of user-defined
 types?
In-Reply-To: <20020430031810.GY16217@devel.trillke>
Message-ID: <Pine.SOL.4.44.0204292321410.25646-100000@xexex.gpcc.itd.umich.edu>

On Tue, 30 Apr 2002, holger krekel wrote:

> On Tue, Apr 30, 2002 at 02:35:16AM +0000, Michael George Lerner wrote:
> > Are you sure you're putting appropriate things into the list?
>
> i guess he is. He just didn't know that python
> wants expliciteness rather than c++ish impliciteness
> (type-conversions).

d'oh.  i noticed that after i responded.  and the

' '.join([str(i) for i in l])

response summed it up pretty well, so i decided to keep my mouth shut :).

-michael






From tim.one at comcast.net  Sun Apr 14 19:30:05 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 19:30:05 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F5640@MAYOR.retzlaff.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIECHOOAA.tim.one@comcast.net>

[Jimmy Retzlaff]
> ...
> Any chance some package is also using Wise and treating unwise.exe as
> its own?

It's unlikely, unless it *overwrites* the existing Python unwise.exe as part
of its install.  Wise doesn't special-case "unwise.exe" on an uninstall,
it's exactly like any other file then.  The only reason it gets deleted
during an uninstall is because it's explicitly listed (along with all other
files the install created) in install.log.  For example, here's its line
from the Python 2.2.1 install.log:

File Copy: C:\Python22\UNWISE.EXE | 05-24-2001 | 12:59:30 | ...

The uninstaller deletes the files named in "File Copy" lines, using the full
path recorded there.

It's possible that a poorly written installer is leaving behind a "File
Tree:" line in its install.log, deleting *all* .exe files.  For example, the
PythonLabs Windows installer injects these lines in install.log:

File Tree: C:\PYTHON22\*.pyc
File Tree: C:\PYTHON22\*.pyo

to get rid of compiled Python files when you uninstall Python.  But if
someone were doing that with *.exe, then python.exe and pythonw.exe would
get deleted too -- but every time unwise.exe has gone missing on me,
python.exe is still there.

BTW, I never saw this happen before we upgraded to a new version of the Wise
installer builder, and one thing did change in *my* treatment of the
uninstaller:  the uninstaller file used to be named UNWISE.EXE, and we
installed it under that name.  In the newer Wise system, the uninstaller
file is named UNWISE32.EXE instead, but in order to foster the illusion
<wink> that nothing had changed, I *rename* that to UNWISE.EXE as part of
the install process.  This has been nagging me for a year, but as I've never
been able to *provoke* the problem at will, I don't have a conclusive way to
test whether the renaming somehow confuses Windows.

I have noticed that unwise.exe is at least much more likely to vanish if
Win98 crashes between the time I do an install and get around to doing a
normal reboot.  That suggests Wise is doing something special with it --
although I've no idea what (or, really, even "if").





From seant at iname.com  Fri Apr  5 14:05:12 2002
From: seant at iname.com (Sean True)
Date: 5 Apr 2002 11:05:12 -0800
Subject: [ANN] JpGraph 1.5.3 Graphs, Charts, Plots ported to Python
Message-ID: <mailman.1018271582.5841.clpa-moderators@python.org>

A Python version of JpGraph, a substantial presentation graphics
module written in PHP by Johan Persson.

"JpGraph is a fully OO graph library which makes it easy to both draw
a "quick and dirty" graph with a minimum of code and quite complex
graphs which requires a very fine grain of control. The library tries
to assign sensible default values for most parameters hence making the
learning curve quite flat since for most of the time very few commands
is required to draw graphs with a pleasing aesthetic look." -- Johan
Persson

Our port of JpGraph depends on gdmodule, and of gd itself. Both are
included in the tar file. The port also expects Freetype and
appropriate TTF fonts, which are not included. Gantt charts and image
maps still need work, but the package in general pretty well. Sample
output is available from the download
site.

Thanks to Johan Persson, Chris Gonnerman, and Thomas Bouttel.

Download from http://metagram.webreply.com

Sean True
Webreply, Inc.




From jon+usenet at unequivocal.co.uk  Mon Apr 29 12:05:21 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 16:05:21 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org>
Message-ID: <slrnacqrps.sdd.jon+usenet@snowy.squish.net>

In article <mailman.1020090053.19284.python-list at python.org>, Fran?ois Pinard wrote:
>> Can you think of an example where the "extra power" provided by a more
>> complicated templating system actually makes something easier to do than
>> with jonpy? ;-)
> 
> To really reply, one would have to study `jonpy'.

Yes. But I know that Paul, who I was replying to, is somewhat familiar
with jonpy and hence would most likely in a position to be able to
answer the question.

> Yet, I got from an earlier message that `jonpy' does not allow
> arbitrary Python in the HTML pages.

It does not allow code of any sort in the HTML pages. This is because
I do not think that code of any sort should be in the HTML pages. Code
and HTML are separate things that belong in separate files rather than
muddled together.

> Like many others, I made my own templating system, but did not overly
> publish it as there are so many already.  Surely in my case, I would loose
> big if I was not allowed to link Web pages with installed Python modules,
> these take care of database access, specialised HTML formatting (that
> would only encumber the templating language if it was done there), etc.

Linking web pages with installed Python modules is exactly what jonpy
does.

> In fact, before writing my own, I glanced around to see what was written,
> and my overall feeling was that systems were either weakish or over-bloated.

Yes, that's why I made my system too.

> If you nicely marry Python with your templating system, you get an extremely
> powerful templating system that is also very simple.

Yes, that's what jonpy does. It doesn't involve putting the Python
code in the same file as the HTML though.

> The power _is_ needed.

Yes, it is my belief that jonpy, while being very simple, is also very
powerful. Hence my question to Paul about what features of other more
complicated templating systems make him feel that they have "extra
power" in comparison, because personally I can't think of any.

Cheers


Jon


From chris.gonnerman at newcenturycomputers.net  Wed Apr 17 23:10:35 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 17 Apr 2002 22:10:35 -0500
Subject: Problem Writing Files
References: <DCF398FA38882C418E2F4F41F5146D870B28DD@exchange3.slac.stanford.edu>
Message-ID: <000d01c1e686$9f6cd340$0101010a@local>

----- Original Message -----
From: "Jones, Roger M." <rmj at SLAC.Stanford.EDU>


> Dear Chris,
>
> Thank you for your reply.  It is helpful.
> I tried rewind and seek but they
> do not work properly.

First rule of asking for help via email or USENET:  post
lots of detail.  What exactly did you do?  Copy the actual
screen with the error message and paste it into your email.

For instance, this is the "right" way to write and then
re-read a file, copied from an actual session (you may need
to widen your window to read this):

----------------------------------------------------------------------------
-------
PythonWin 2.1.2 (#31, Jan 15 2002, 17:28:11) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (MarkH at ActiveState.com) - see
'Help/About PythonWin' for further copyright information.
>>> afile = open("myfile.txt", "w+")
>>> afile.write("a line of text.\n")
>>> afile.seek(0)
>>> afile.read()
'a line of text.\n'
>>> afile.seek(0)
>>> afile.readline()
'a line of text.\n'
>>> afile.write("another line.\n")
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
IOError: (0, 'Error')
>>> afile.seek(0, 1)
>>> afile.write("another line\n")
>>> afile.seek(0, 0)
>>> afile.readline()
'a line of text.\n'
>>> afile.readline()
'another line\n'
>>> afile.readline()
''
>>>
----------------------------------------------------------------------------
-------

I open the file in "w+" mode (read the documentation for the
builtin open function), then write a line of text, then seek
the start of the file (read the file object docs in section
2.1 of the library documentation; 2.1.7.9 in version 2.1.2
manuals).  The seek is required on most platforms when
switching between write and read or read and write on a
stream.  Next I read the entire file (with the read method).

I attempt to write a line immediately (the file pointer is at
the end of file, so I expect I am appending a new line) but
I get an exception of type IOError because I neglected to
seek before changing modes.  Sheepishly I do a seek to 0
bytes from the current position (that's what the second
argument to seek does; read thos docs in the library manual
section 2.1!) and write the line.  The next seek is to the
start of file (I don't have to include the second argument
but I did anyway).

Now the readline method returns the first line, because that's
where the file pointer is right now.  Another readline gets
another line from the file; the null string after the third
readline is the end of file indicator.

> Is there a module
> that needs to be loaded before using rewind or seek?

Nope.  READ THOSE LIBRARY DOCS, especially the file object
documentation.

> You are right, I failed to type "open"
> which I of course used in the actual python session.

Yup.

Everything you need for basic understanding of file I/O
is in the example above.  Absorb it.

OK, I left one thing out:

>>> afile.close()

(Note that the >>> is the prompt from Pythonwin, where I
ran this test.)  Closing the file is not required, as
Python closes all open files automatically on exiting the
interpreter, but as each open file represents a resource
in use, and you don't want to run out, you should close
file handles (file objects) explicitly when you don't need
them anymore.

Hope this helps.

Oh, yeah, READ THE MANUALS.

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





From alet at unice.fr  Fri Apr 26 03:45:57 2002
From: alet at unice.fr (Jerome Alet)
Date: Fri, 26 Apr 2002 09:45:57 +0200
Subject: urllib.urlopen("file://...") fails
References: <m2k7qwjnma.fsf@jtan.com>
Message-ID: <3CC905B5.209F86B6@unice.fr>

Seth Delackner wrote:
> 
> With python 2.2 on windows, with the file t:\tmp\package.html existing,
> 
> >>> import urllib
> >>> u = urllib.urlopen("file://t:/tmp/package.html")
> 
> Fails with the error below, but "file:t:/tmp/package.html" works. Is
> that even a valid url?

IMHO this is a bug in the urllib module when you want to open a local
file
and you specify the disk on which to find it under Windows.

if t: is your current working disk then don't put the t: in the path and 
it should work fine.

under *nix systems this problem doesn't exist (no disk letters)

hth.

Jerome Alet


From niemeyer at conectiva.com  Fri Apr 12 10:34:14 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Fri, 12 Apr 2002 11:34:14 -0300
Subject: A simpler super.
In-Reply-To: <u662x3u1j.fsf@online.no>
References: <u662x3u1j.fsf@online.no>
Message-ID: <20020412113414.B1932@ibook.distro.conectiva>

> def Super(yourself):
>     if isinstance(yourself, type): # for classes
>         return super(yourself, yourself)
>     return super(yourself.__class__, yourself)

I had the same idea a while ago. This can't be done. Think about what
happens here:

class A:
	def __init__(self):
		pass
class B(A):
	def __init__(self):
		Super(self).__init__(self) # <-- problem
class C(B):
	def __init__(self):
		Super(self).__init__(self)

In this case, you'll get a loop, because self.__class__ is C, no matter
where it is called.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From sjmachin at lexicon.net  Thu Apr 25 06:12:21 2002
From: sjmachin at lexicon.net (John Machin)
Date: 25 Apr 2002 03:12:21 -0700
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <c76ff6fc.0204241734.40e57406@posting.google.com> <3cc7967a$0$37911$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <c76ff6fc.0204250212.af07541@posting.google.com>

Boudewijn Rempt <boud at valdyas.org> wrote in message news:<3cc7967a$0$37911$e4fe514c at dreader3.news.xs4all.nl>...
> John Machin wrote:
> > 
> > Presumably the point of having a multi-character pronunciation table
> > is that it is possible that pronounce("xy") can be != pronounce("x") +
> > pronounce("y"). With careful thought, you may be able to remove
> > redundant entries from your more-than-one-char dicts, so that they
> > contain only the necessary exception cases -- but do try the basic
> > approach first.
> > 
> 
> Isn't big-5 a variable length encoding? I thought that was his
> problem, not translating two or more character words.

big5 is a 1-2 byte encoding. A byte 0-127 is more-or-less ASCII; a
byte 128-255 (or less) is the first byte of a two-byte Chinese
character. So it's variable-length only to that extent.

Contrary to popular mythology, Chinese words can have more than one
syllable. As the OP said:

> ["big5" 2, 4, 6 ... byte long strings] there
> with their pronunciations.  If it were just one character [two byte]
> words I would use the "c2t" program.


From hgs at dmu.ac.uk  Thu Apr 18 13:00:45 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 18:00:45 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4u1q955fd.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181736420.16636-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> >         [*shared*, *static* intereaved]
> >
> > But to change the wording one needs to know which of the above
> > properties are reversed, and how commenting out modules affects them in
> > each case.
>
> I'm not sure what your question is. The build process of Python works
> that way:
        [Clear explanation trimmed]
>
OK, how about this for a suggested patch:
<patch>

--- Python-2.2.1/Modules/Setup.dist.orig	Wed Oct 17 14:46:28 2001
+++ Python-2.2.1/Modules/Setup.dist	Thu Apr 18 17:51:35 2002
@@ -36,9 +36,20 @@
 #
 # Finally, if a line contains just the word "*shared*" (without the
 # quotes but with the stars), then the following modules will not be
+# built statically.  The build process works like this:
+#
+# 1. Build all modules that are declared as static in Modules/Setup,
+#    combine them into libpythonxy.a, combine that into python.
+# 2. Build all modules that are listed as shared in Modules/Setup.
+# 3. Invoke setup.py. That builds all modules that
+#    a) are not builtin, and
+#    b) are not listed in Modules/Setup
+#    c) can be build on the target
+#
+# Therefore, modules declared to be shared will not be
 # included in the config.c file, nor in the list of objects to be
 # added to the library archive, and their linker options won't be
-# added to the linker options, but rules to create their .o files and
+# added to the linker options. Rules to create their .o files and
 # their shared libraries will still be added to the Makefile, and
 # their names will be collected in the Make variable SHAREDMODS.  This
 # is used to build modules as shared libraries.  (They can be
@@ -46,8 +57,9 @@
 # toplevel "make install" target.)  (For compatibility,
 # *noconfig* has the same effect as *shared*.)
 #
-# In addition, *static* reverses this effect (negating a previous
-# *shared* line).
+# In addition, *static* explicitly declares the following modules to
+# be static.  Lines containing "*static*" and "*shared*" may thus
+# alternate thoughout this file.

 # NOTE: As a standard policy, as many modules as can be supported by a
 # platform should be present.  The distribution comes with all modules
</patch>

This provides context for the linking process, breaks up the long
sentence, and describes the interleaving of *static* and *shared*.
>
> > So the fastest fix is for me to doengrade binutils?
>
> No. The fastest fix is to apply the patch below, and rerun configure.

Excellent. I will give that a spin.  Thank you.
>
> Regards,
> Martin
>
        [configure patch trimmed]

        Hugh





From maxm at mxm.dk  Wed Apr  3 03:48:59 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 03 Apr 2002 08:48:59 +0000
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
Message-ID: <3CAAC1FB.2050105@mxm.dk>

David Goodger wrote:
> Here's a serious proposal, safe to post now that April Fool's is over.
> Please read & comment.

Through Zope I have used Structured text extensively. Also I have good 
experience teaching stuctured text to my customers. After a very short 
while they start to use it insted of html when generating content for 
their sites due to ease of use.

A very nice bonus that isn't mentioned in the PEP is that it is very 
easy to store the text in a database and then query in free text. No 
need to parse it beforehand to remove markup tags.

Structured text is easy to write and easy to read, and flexible enough 
to write documentation that contains code. So in theory it could even be 
used in a *readable* literate programming tool.

Something that I would like to use for writing better documentation.

regards Max M



From starx at pacbell.net  Mon Apr  8 14:01:49 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 18:01:49 GMT
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com> <a8q9bi$2vka$1@news.idiom.com> <d9ea2696.0204071920.1b9d75c4@posting.google.com> <3CB112E4.B5884860@engcorp.com>
Message-ID: <Xns91EA7063DEC28RASXnewsDFE1@130.133.1.4>

Peter Hansen || Sun 07 Apr 2002 08:47:48p:

> Date: Sun, 07 Apr 2002 23:47:48 -0400
> Subject: Re: modules
> From: Peter Hansen <peter at engcorp.com>
> Newsgroups: comp.lang.python
> 
> distrex wrote:
>> Jim Dennis wrote:
>> > distrex wrote:
>> > >I understand most of python. But I still don't get modules. I know
>> > >they are just basicaly a function but i don't know how to use them
>> > >effectively and what each one does.  [...]
>> >
>> >  Modules are things that you import into your namespace.  
> [...snip rest of Jim's reply]
>> 
>> Don't take this the wrong way but I am a newbie and didn't understand
>> 86% of what you just said. All I got was modules are classes
>> functions and variables. and a few other things but that was to
>> complicated for me (the newbie)
> 
> I can see how it might have been above the level you are at just yet,
> so don't take this the wrong way: what are you going to do about it?
> [ ... ] 
>
> Your question about "what each one does" is also strange to us.  
> It almost sounds like you know there are dozens of standard modules
> and were asking what each one does.  There is full documentation
> on each module already, so that must not have been what you meant.
> But in that case the question is bizarre. "Each one"?  What "each"
> are you talking about?
> 
> And in case it helps any, here's a shorter answer than Jim's:
> 
> Modules are files which contain collections of functions, variables,
> and classes.  They help organize your code so it doesn't all 
> have to be in one large file, and they let you put code in 
> libraries to make it easier to re-use, so you don't have to
> copy-and-paste all the time to re-use other people's code.
> (This is a somewhat simplified description but may be more 
> useful to a newbie.)

It may help to add this:

Some nice people have already set up modules that contain useful functions 
and classes organized into coherent sets. There are quite a few of these, 
and a great many make up the 'standard library' that comes with Python. 
The master guide to all this usefulness is the Python Library Referance*. 
Read over the table of contents here... each of these names that you see 
in typewriter font is a diferent module you can import into your program: 
sys, urllib, string, random, pickle, threading, socket, os, Tkinter, 
email.. each of these is a module, some .py file sitting somewhere in your 
Python folder just waiting to be evoked for some useful or sinister 
purpose. Sort of like magic :).

If these are the names you are thinking of when you ask 'what each one 
does', then this reference should be your first stop for information. It 
might not make sense, but if you can at least try to understand it you 
will be that much more prepared to ask a question here, and people will be 
more able (and more willing) to help you if you can pick out what you 
don't understand and ask more specific questions.

* http://www.python.org/doc/current/lib/lib.html - the contents, alhtough 
it's more like a subject index

http://www.python.org/doc/current/lib/modindex.html - index of module 
names, often easier to find something here if you only know it's name 
(though i usually just use find in the browser on the other page :)

> Help us a little more, and we can help you a little more. :-)

Learining... moving from ignorance to understanding of a subject... is 
rarely easy. However, with work and ye shall be rewarded.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com
"There's more than one answer to these questions-- pointing me in a 
crooked line" - (Closer to fine)


From roy at panix.com  Wed Apr 10 21:26:13 2002
From: roy at panix.com (Roy Smith)
Date: Wed, 10 Apr 2002 21:26:13 -0400
Subject: Easter Eggs
References: <mailman.1018485864.13992.python-list@python.org>
Message-ID: <roy-DBB765.21261310042002@news1.panix.com>

Tim Peters <tim.one at comcast.net> wrote:
> they're-all-backward-compatible-too-ly y'rs  - tim

Actually, backward compatibility can be harder to achieve than you seem to 
think.  By adding the "this" module, you broke existing code which relied 
on:

try:
   import this
except ImportError:
   do stuff

doing stuff.  Stick that in your PEP and smoke it :-)


From fortepianissimo at yahoo.com.tw  Mon Apr 29 21:54:12 2002
From: fortepianissimo at yahoo.com.tw (Fortepianissimo)
Date: 29 Apr 2002 18:54:12 -0700
Subject: Newbie question: how to join a list of elements of user-defined types?
Message-ID: <ef08387c.0204291754.1224fd7b@posting.google.com>

This question is so fundamental that I strongly suspect it's already
answered - ig that's the case, please forgive this poor newbie, for he
has done some work searching over Python FAQ and Google.

Basically I want to have a subclass of list, like

class MyList (list):
    ...

then I want to

l=MyList()
... do stuff to fill l
s=" ".join(l)

with the hope that s will be a nicely joined string with " " as the
delimiter. Given my C++ background, I tried to define __str__/__repr__
for both MyList and the user-defined types of the list members,
assuming the type conversion would silently be invoked. But I was
wrong!

Even trying this failed:

l=[1,2,3]
s=" ".join(l)

and I thought built-in types all have default __str__ and __repr__?

What am I missing here? Many thanks to someone who would take time to
rescue a poor soul.


Ben


From mgilfix at eecs.tufts.edu  Wed Apr 17 02:17:03 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 02:17:03 -0400
Subject: Pure python crypto
Message-ID: <20020417021703.H16954@eecs.tufts.edu>

  Would anyone have any use for a pure python implementation of
Blowfish? Or any other crypto algorithms for that matter.  As a matter
of interest, I found myself implementing the blowfish module recently
but was unsure if any audience would be interested. No C compiler
is required, which is nice. Could use some testing though.

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/d35209bc/attachment.sig>

From gbreed at cix.compulink.co.uk  Tue Apr  9 06:24:26 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Tue, 9 Apr 2002 10:24:26 +0000 (UTC)
Subject: Yeah whatever... (was RE: PEP 285: Adding a bool type)
References: <mailman.1018296200.32231.python-list@python.org>
Message-ID: <a8ufgq$n1v$1@thorium.cix.co.uk>

Ken Seehof wrote:

> Sorry, I just don't get why this is relevant.  Let's try to keep this
> discussion in the real world, ok?

The real world's no fun.  A peek at the types module shows that Python 
already has 33 fundamental types, which means there are already 
33!=8683317618811886495518194401280000000 ways of combining them by this 
strange factorial rule.  It's trivial to show that Python is already 
impossible to use as a programming language.  Unless perhaps we can all 
give ourselves super intelligence as well.


                      Graham


From timr at probo.com  Fri Apr 19 01:34:43 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 18 Apr 2002 22:34:43 -0700
Subject: raw_input bug only in cygwin python
References: <12257ec4.0204181907.7a79b15b@posting.google.com>
Message-ID: <kuavbukpa505nru9g6kfivo3bfrtirda05@4ax.com>

davidccarson at hotmail.com (David Carson) wrote:

>Here is a snippet of code:
>
>01: #!/usr/bin/env python
>02: 
>03: list = []
>04: while 1:
>05:     try:
>06:         stuff = raw_input("[^D to exit]: ")
>07:     except EOFError:
>08:         print "got EOF"
>09:         break
>10:     list.append(stuff)
>11: 
>12: raw_input("bang! ")
>
>
>This works on every Python environment I've tried but cygwin (Win2000
>-- I haven't got NT to try it):
>  Python 2.2 (#1, Dec 31 2001, 15:21:18)
>  [GCC 2.95.3-5 (cygwin special)] on cygwin
>
>In this environment, the EOFError is caught.  Then the "bang! " prompt
>for the raw_input() outside the loop is printed, but immediately the
>program exits with an uncaught EOFError.
>
>This code works with:
>- ActiveState Python, 2.1 on Win2000
>- python 2.2 on Mandrake
>- python 1.5.2 on RedHat
>and everywhere else, I'm sure.
>
>Is this a known problem with cygwin?

I wouldn't call it a problem with cygwin, I'd call it "relying on
unspecified behavior".  Ctrl-D is end of file; the ability to read past EOF
is not promised anywhere in the docs.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From printers at sendme.cz  Wed Apr 17 15:23:00 2002
From: printers at sendme.cz (A)
Date: Wed, 17 Apr 2002 21:23:00 +0200
Subject: What webhosting
Message-ID: <3CBDE7B4.3348.4EB76@localhost>

Hi,
Does anybody know about a webhosting provider ( with Python and 
Telnet access) that does not remove a user's account if the user 
sends occasionally some unsolicited emails?
Thanks for reply
Ladislav






From wesc at deirdre.org  Tue Apr  9 14:56:10 2002
From: wesc at deirdre.org (Wesley J. Chun)
Date: 9 Apr 2002 11:56:10 -0700
Subject: ANN: BayPIGgies this WED nite 7:30pm
Message-ID: <mailman.1018559645.27708.clpa-moderators@python.org>

What:    BayPIGgies meeting
When:    Wed, 10 Apr 2002, 7:30-9 PM
Where:   Stanford University, Palo Alto, CA
Agenda:  Newbies Night
Speaker: everyone...

Invite everyone you know who may be interested in Python but have
questions, would like to learn more about it, or need some advice on
an application! These meetings have been very popular in the past,
with a good mix of newbies as well as old hands. They are even more
fun when one or more Perl experts come around wondering what the big
deal is about Python. :-) Come join us for this interactive session!

Next Meeting (5/8): Eating Out with Python (our meeting room will not
be available, so we will meet and chat over dinner at a local
restaurant!)

--------------------------------------------------------------------------------
Call For Talks: We are actively seeking speakers for BayPIGgies! If
you would like to give a talk at one of our meetings (any Python
related topic), contact us to coordinate!

for more information including driving and public transit directions:

http://deirdre.org/baypiggies

-wesley

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://deirdre.org/baypiggies

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/




From max at alcyone.com  Mon Apr  8 14:23:08 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 08 Apr 2002 11:23:08 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com> <eppstein-CB6351.08534907042002@news.service.uci.edu> <ub315bi2fkdube@news.supernews.com>
Message-ID: <3CB1E00C.1DBA7DE4@alcyone.com>

John Roth wrote:

> Ints are neither cardinal nor ordinal. They just are. Cardinal and
> ordinal are ways of analyzing the situation that may, or may not,
> apply.

Indeed, even in mathematics, the cardinal number that is represented
with "1" (that is, the cardinality of of a set with a single element) is
_not_ the same thing as the integer 1.  Same goes for ordinal numbers.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From aleax at aleax.it  Sat Apr  6 04:43:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 06 Apr 2002 09:43:16 GMT
Subject: abstraction, names, and lack thereof
References: <E16tGLp-0007ou-00@mail.python.org> <mailman.1018014793.1604.python-list@python.org> <axir8.52642$pT1.1485864@news1.tin.it> <slrnaat3fn.3cd.jajvirta@sirppi.helsinki.fi>
Message-ID: <Uqzr8.54598$pT1.1547186@news1.tin.it>

Jarno J Virtanen wrote:
        ...
>> def specificSquirker(): self.squirkIt(23, 45)
>> Button(buttonsFrame, text='Squirk', command=specificSquirker)
>> 
>> It's generally not hard to come up with a name which helps readability.
>> I think un-lambdaing your code in this way is a good thing...
> 
> I agree.  Furthermore, I don't even think that un-lambdaing that way is
> about the maintainability issue per se (ie., 'the "functional" features
> make python code unreadable' [1]); 
        ...
> [1] Which I might or might not agree.

I wouldn't.  map, filter, list comprehensions (from Haskell), lazy
evaluation (aka iterators), are all examples of typical functional
programming features that make Python powerful and readable.

lambda, specifically, isn't, IMHO.


> it's just that abstracting everywhere
> is a good thing.  

I'd nuance that to ALMOST everywhere.  Sometimes you DO want to think
about the plumbing, so naked tuples or dicts can be more useful than
the various abstractions they could stand for.  Guido makes this point
in his essay on graphs, and a speaker at IPC10 made it about his use
of Python to teach a university course in algorithms and data structures.


But such cases ARE the exception, not the rule.  So,

try: abstraction is good
except WhenYouAreStudyingThePlumbingItself:
    (a, rare, case)

would seem to be the Pythonic approach to this.  A slightly expanded
version of your:

> [2] Unless writing a programming language compiler or interpreter.


> The highest level of logic at that point is not about
> lambdas, it's about "squirking" and that's what you should see in the
> source code.  Making routines and giving them names that represent the
> highest level of logic, which is not about the actual programming
> language code statements [2], is a very powerful tool (available in
> other languages too, of course :-).

The ABILITY to name things is almost always good -- if the "things"
are well-identified (and a function almost invariably is).  Insert
diatribe pro/con macros, hygienic or otherwise, to taste.

The REQUIREMENT to name things you'd rather leave unnamed could be
argued to be bad.  Naming gives identity and conceptual dignity to
something you might perhaps prefer to sweep more under the carpet.

When extracting a few things from a tuple I might code:

    fleep, gorp, __, __, swant, plosk, __ = subberTuplest(lark)

where the __ plays the role of an "anonymous name" (somewhat of
an oxymoron perhaps...:-).  Two underscores, not one, to avoid any
issues wrt interactive interpreter sessions or i18n via gettext,
is what I like to use here -- it also stands out just the right
amount (IMHO) as a place-holder.

Consider the alternatives when you do need to give names to items
0, 1, 4 and 5 of a 7-item tuple returned by a function call.  You
could bind the function call's result to a name -- but that would
"give identity and conceptual dignity" to what's probably a mere
artefact, a way for the function to "return several results" --
and would the resulting indexing over several statements be as
readable and clear as this unpacking?  Or you could build some
more elaborate special-purpose infrastructure, e.g.:

def indexedSelector(sequence, *indices):
    return [sequence[index] for index in indices]

fleep, gorp, swant, plosk = indexedSelector(
    subberTuplest(lark), 0, 1, 4, 5)

Worth it to avoid the "anonymous names"?  Maybe.  But it seems
to me that my preferred alternative remains more readable, in a
typical situation where those index numbers mean nothing special
to the human reader.  The visual "pattern matching" of the
original unpacking assignment seems to work pretty well.


I'd like to make "__ as anonymous name" a widespread convention
in Python, for such purposes as this one.  Should it ever catch
on, then, for those rare cases where you just can't abide naming
(giving identity and conceptual dignity) a small "throwaway"
function that you syntactically need, you COULD sensibly code:

def __(): self.squirkIt(23, 45)
Button(buttonsFrame, text='Squirk', command=__)

I wouldn't do it today, because readers would puzzle over the
unfamiliar usage (strangely enough, use of __ in unpacking
assignments doesn't seem to provoke similar puzzlement, even
today).  But language's pragmatics are molded by usage and
familiarity, so it's conceivable that such an idiom may become
acceptable one day.

It IS peculiar to go to such lengths to avoid naming a _function_.
Most often it's OK to give it identity and conceptual dignity.
Still, when exceptions to this DO arise...


> (There are other issues, but this abstraction thingy has been
> haunting me last couple of days and I had to let out the steam. ;-)

Just a couple of days?  This abstraction thingy (how best to
achieve it, where it gets in the way, and so on) haunts _me_
since at least a couple of decades...:-).


Alex



From ianb at colorstudy.com  Mon Apr 15 20:53:09 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 15 Apr 2002 19:53:09 -0500
Subject: Python webmail packages?
In-Reply-To: <7876a8ea.0204151614.138853be@posting.google.com>
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com> 
	<7876a8ea.0204151614.138853be@posting.google.com>
Message-ID: <1018918390.1149.721.camel@localhost>

On Mon, 2002-04-15 at 19:14, David Rushby wrote:
> BoboMail:
> http://bobomail.sourceforge.net/

BoboMail works okay.  It uses some portions of an old version of Zope
(hence it's named after that old version -- actually, it might be new
versions of those pieces, I don't know) -- but it's actually CGI-based. 
It only works with POP.  It doesn't always deal gracefully with
misformatted mail, giving Zope errors which are hard to read for the
non-programmer.  But it does work, and I was happy with the code when I
was making modifications.  It's been a sleepy project for some time now.

  Ian






From maxx at easynews.com  Thu Apr 11 15:08:16 2002
From: maxx at easynews.com (Max)
Date: Thu, 11 Apr 2002 19:08:16 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83>
Message-ID: <m1nbbus765pra82gmeof51tfn32ooqqtr3@4ax.com>

On 10 Apr 2002 04:22:55 GMT, Lumberjack <lumber at jack.com> wrote:

>People aren't going to change, so Guido needs to adjust to the reality of 
>the political consequences of his current strategy.

Sounds like someone is having a hard time keeping up.

If you have any concept of computing history, then you should be able to observe
that people MUST change, or become left behind by the changes. Ever heard the
phrase "the one thing that remains constant is change" ?

There are many cases where technologies have undergone significant changes, and
_paying_ customers who have not been overly happy about the changes continue to
pay licensing fees anyway. The most currently successful software company in the
world practices this way. Heard the VB developers screaming lately? 

In other cases, I have witnessed many a developer or technician, who solidly
refuses to adapt to changes in their preferred environment be left behind, while
the rest of the community moves along with the group. Sometime we call them
"dinosaurs" - middle aged developers, whose supervisors were born a generation
later. 

"Political consequences" sound like the typical hippie-hacker nonsense from
those whose identities are defined by their preferred computing envrionment. Who
can name a single language that remains absolutely static until everyone (or
even a significant majority) agrees on the changes? Sooner or later, design
decisions are made by an individual or small group, and we all are affected.

This is the reality that needs adjusting to. 



From emile at fenx.com  Sun Apr 14 14:45:01 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 14 Apr 2002 18:45:01 GMT
Subject: Specifying a time from previous day
References: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net>
Message-ID: <N6ku8.22340$G72.14938@sccrnsc01>

jrup
> I need to search some log files, but only if they were created after
4:00pm
> the previous day.

Look into os.stat

>>> print os.stat.__doc__
stat(path) -> (mode,ino,dev,nlink,uid,gid,size,atime,mtime,ctime)
Perform a stat system call on the given path.


--

Emile van Sebille
emile at fenx.com

---------



From dalke at dalkescientific.com  Tue Apr 16 01:30:27 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 15 Apr 2002 23:30:27 -0600
Subject: Does Python need a '>>>' operator?
References: <mailman.1018925938.16233.python-list@python.org> <a9g8ok$9a3$1@panix1.panix.com>
Message-ID: <a9gcmq$bmb$1@nntp9.atl.mindspring.net>

Aahz:
>What's a co-finite set"?  (Yeah, I could Google for it, but I usually
>understand *your* explanations -- eventually. ;-)

I'm not my Uncle Timmy, but I can give it a go.  Let me dust off
my math degree, as soon as I can find the sucker...

It's a set whose complement (that is, the set containing everything
which isn't in the original set) is finite.

For example, in integers the set

  ...-3, -2, -1, 0, 1, 3, 4, 5, 6, 7, ...
                     ^^^ 2 is missing

is cofinite because its complement, which is {2}, is finite.

The set of even integers is not a cofinite subset of the integers
because its complements, the odd integers, is also infinite.

They are relevant here because bits can only be constructed either
explicitly (you need to say them) or though operators.  Operators
like 'and' and 'or' don't make infinite sets, but 'not' does.
This value cannot be represented directly as an infinite set of
bits, but can be represented as a finite set of bits and a flag
that says "use the complement."  There is no way to construct
a bitlist using these operators (in finite time) which is not
either finite or cofinite, which means there is a relatively
easy way to store these calculations on a computer.

                    Andrew
                    dalke at dalkescientific.com





From rjones at ekit-inc.com  Thu Apr  4 19:32:34 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 5 Apr 2002 10:32:34 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <01c701c1dc35$5a2312f0$02010a0a@suxlap>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204042338.XAA12847@crown.off.ekorp.com> <01c701c1dc35$5a2312f0$02010a0a@suxlap>
Message-ID: <200204050032.AAA20911@crown.off.ekorp.com>

On Fri, 5 Apr 2002 10:03, Andreas Jung wrote:
> [on switching Zope from STNG to ReST]
> Backward compatibilty is a major issue for inclusion into Zope.

This is the major problem facing such a switch, yes. Automatic conversion 
_should_ be possible,  but I have a strong feeling this issue will be a 
blocker. Converting from STNG to ReST is usually just a process of dedenting.


> I still have not figured how to use it.

This is a big issue at the moment - there's a lot of detailed spec and 
related documentation, but nothing solid on the user side of things. I'm 
hoping to help that along with documentation like the primer and a simple 
command-line tool exposing the most commonly needed functionality. At the 
moment, the restructuredtext/tools/html.py script is it (pass it a ReST file 
name and it spits out HTML).


> Are there generators for HTML available yet?

Yep. The primer I just posted uses it. Roundup uses it for its non-docstring 
documentation (only because there's no simple docstring extractor yet). Here 
at ekit we've used it for internal documentation too.

The generation is slow at present. The emphasis has been on implementation, 
and optimisation is yet to come.


    Richard




From lac at strakt.com  Sat Apr  6 10:26:23 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 17:26:23 +0200
Subject: Selecting cipher suites with socket.ssl 
In-Reply-To: Message from loewis@informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=) 
   of "06 Apr 2002 12:40:51 +0200." <j4k7rl2jpo.fsf@informatik.hu-berlin.de> 
References: <84cff98e.0204041830.7c0d0171@posting.google.com> <mailman.1017978613.9042.python-list@python.org> <m3y9g21uea.fsf@mira.informatik.hu-berlin.de> <200204051346.g35Dkerq000532@ratthing-b246.strakt.com> <m3it769ee9.fsf@mira.informatik.hu-berlin.de> <mailman.1018087333.20495.python-list@python.org>  <j4k7rl2jpo.fsf@informatik.hu-berlin.de> 
Message-ID: <200204061526.g36FQNdj017294@ratthing-b246.strakt.com>

> Martin:
> I was mainly objecting the assertion "It's broken." I don't think that
> is true - it is mainly incomplete.

I think I should have said that instead, yes.  Apologies.

> > Besides, whatever he comes up with may be rejected on the grounds
> > that the code isn't good enough, or it adds the sort of complexity
> > you don't want in the standard library.  Modifying an existing
> > library shouldn't be by definition easier to do than adding a new
> > one -- foul code is foul code
> 
> I disagree that the current SSL support is foul code.

This is my day for misunderstandings.  I'm not saying that the SSL
is foul code.  I'm saying that if you write foul code, you shouldn't
think that you will have an easier time fitting it into an existing
library than submitting it as a new library for inclusion.

Laura Creighton




From h.e.w.frater at cs.cf.ac.uk  Mon Apr 22 21:41:08 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 22 Apr 2002 18:41:08 -0700
Subject: need encryption algorithm
Message-ID: <e2e9e807.0204221741.597acd60@posting.google.com>

Hi all,
  I need a simple encryption algorithm that will not give me
characters that I can't store in my oracle table. At the moment I'm
using the rotor encryption module because I like it's simple interface
but It's generating characters that oracle can't store which is a damn
pain. What other modules are available that are easy to use and will
do what I wan't. Remember that security is not paramount here, I'm
only including it for functionality.

Thanks,
hugh


From hgs at dmu.ac.uk  Fri Apr 19 10:41:44 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Fri, 19 Apr 2002 15:41:44 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <3CC02703.DB2E7D8D@metaslash.com>
Message-ID: <Pine.GSO.4.40.0204191540120.24244-100000@neelix>

On Fri, 19 Apr 2002, Neal Norwitz wrote:

> Hugh Sasse Staff Elec Eng wrote:
> >
> > Now the build gives:
>
> 	[compiles & links]
>
> > case $MAKEFLAGS in \
        [...]
> > ld.so.1: ./python: fatal: libtk8.3.so: open failed: No such file or directory
> >
> > But
>
> 	[tcl/tk 8.3 is in /usr/local/lib]
>
> Does LD_LIBRARY_PATH have /usr/local/lib in it?

Grr! I should have spotted that, thanks.  That was the cuase of this.
>
>
> Neal
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Hugh





From mwh at python.net  Tue Apr  2 11:53:08 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 2 Apr 2002 16:53:08 GMT
Subject: __slots__ query
References: <KvTTsIAQYbq8EwrM@jessikat.fsnet.co.uk>
Message-ID: <lk663am498.fsf@elios.maths.bris.ac.uk>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> I notice that in 2.2 it's possible to do strange things to slots via
> instances.

There were lengthy threads on python-dev about this not so long ago (I
recommend the thread on "Meta-reflections" from February this year).

I think the best summary of the answer would be "yes, you can" with
overtones that you probably shouldn't be using __slots__ without a
thorough understanding of the consequences (but, as I seem to be
saying a lot of late, I didn't really concentrate on the thread).

Cheers,
M.


From seth at jtan.com  Thu Apr 25 11:27:55 2002
From: seth at jtan.com (Seth Delackner)
Date: Thu, 25 Apr 2002 15:27:55 GMT
Subject: urllib.urlopen("file://...") fails
References: <m2k7qwjnma.fsf@jtan.com> <aa89g0$pv8$1@peabody.colorado.edu>
Message-ID: <m2d6wn7q85.fsf@jtan.com>

Fernando P?rez <fperez528 at yahoo.com> writes:

> Seth Delackner wrote:
>>>>> u = urllib.urlopen("file:///t:/tmp/package.html")
>                                 ^
>
> One / too many? You're directing it to \t:\.... which it obviously
> can't find.
>
> I don't run windows so I can't test so I may be wrong. It just looks
> like that is your problem.

You would think, right?  But I just tried 'file://t:/tmp/package.html'
and that bombed:

>>> urllib.urlopen('file://t:/tmp/package.html')
    Traceback (most recent call last):
      File "<pyshell#114>", line 1, in ?
        urllib.urlopen('file://t:/tmp/package.html')
      File "d:\python\lib\urllib.py", line 73, in urlopen
        return _urlopener.open(url)
      File "d:\python\lib\urllib.py", line 178, in open
        return getattr(self, name)(url)
      File "d:\python\lib\urllib.py", line 403, in open_file
        return self.open_ftp(url)
      File "d:\python\lib\urllib.py", line 447, in open_ftp
        host = socket.gethostbyname(host)
      IOError: [Errno socket error] (7, 'getaddrinfo failed')

The only thing that actually works for me is:
>>> urllib.urlopen('file:t:/tmp/package.html')

Which is not even a valid url!


From sholden at holdenweb.com  Fri Apr 19 21:48:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 19 Apr 2002 21:48:00 -0400
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com> <e2e9e807.0204191705.6569c572@posting.google.com>
Message-ID: <pP3w8.6885$T%5.6696@atlpnn01.usenetserver.com>

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote ...
> I'm importing it in the interactive interpreter using:
> >>> import cgi #runs fine
>
> here's the cgi script:
>
> import DCOracle2
> import md5
> import time
> import base64
> import cgi
> import rotor
>
> connectString = "'scm1hewf/carthage at cs2000'"
>
> def makeSessionId(st):
>   string = ""
>   m = md5.new()
>   m.update('this is a test of the emergency broadcasting system')
>   m.update(str(time.time()))
>   m.update(str(st))
>   return string.replace(base64.encodestring(m.digest())[:-3], '/', '$')
>
> def enc(password):
>   rt = rotor.newrotor('this is a test of the emergency broadcast system',
20)
>   return rt.encrypt(password)
>
> def checkid(id, password):
>   db = DCOracle2.connect(connectString)
>   c = db.cursor()
>   query = "'select password from tblborrower where id = " + formid + "'"
>   c.execute(query)
>   return c.fetchone()
>
> form = cgi.FieldStorage()
> id = form.getvalue('id')
> password = form.getvalue('password')
>
> if (form.getvalue('id')):
>   encrytedpassword = enc(password)
>   sessionId = makeSessionId(id)
>   result = checkid(id, encryptedpassword)

I note you still haven't shown us the error message that you find in the
error log. That would really help.

Python's import mechanism makes use of sys.path, a list of directories to be
scanned for modules. It's entirely possible that the list has different
contents when scritps are executed under your web server.

Note also that you should provide a conformant HTTP data stream when you are
ruinning under the server (though its absence probably isn't causing your
error). Something like

print """Content-Type: text/html

<HTML>
<HEAD><TITLE>Password Changed</TITLE></HEAD>
<BODY>
    <H2>OK, Done!</H2>
</BODY>
</HTML>
"""

Putting this statement in also has the advantage that you can move it higher
and higher in the script until you actualy see some output - then you know
nothing before it is failing. You may need to amke it line 1 in the present
case 8<(

The most recent release of Python (2.2) also has a cgitb module, which is
very helpful in reporting tracebacks. Just put

    import cgitb; cgitb.enable()

at the head of your script and any Python error messages autmagically appear
in the browser.

Hope all this helps. You'll probably find it was something fairly trivial in
the end.

regards
 Steve


--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From jbperez808 at yahoo.com  Tue Apr 16 02:15:31 2002
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 15 Apr 2002 23:15:31 -0700
Subject: Python for .NET vs. Jython
Message-ID: <f57664b9.0204152215.57312efa@posting.google.com>

Jython is undeniably a very successful implementation of a language on
the JVM.

Considering Python for .NET is still very far from the level of
maturity that Jython enjoys and the fact that the CLR is supposed to
be designed from the ground up to support multiple languages (MS even
commissioned and asked feedback from Mark Hammond and the others
working with him on Python for .NET regarding the CLR), could it be
the CLR is (ironically) less of a fit for Python than the JVM?

If you read Jim Hugunin's original paper on JPython, it seems that
from the beginning, Python on the JVM was a good idea and that,
despite Sun's not designing the JVM for different languages, it
actually turned out to be a really good fit for our favorite language
after all.


From jcmason at student.math.uwaterloo.ca  Mon Apr 15 21:20:48 2002
From: jcmason at student.math.uwaterloo.ca (Joe Mason)
Date: 16 Apr 2002 01:20:48 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com>
Message-ID: <a9fu9g$mj3$1@watserv3.uwaterloo.ca>

In article <3CBB76CA.A423AAD8 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>phil hunt wrote:
>
>> OK, I've decided to write a PEP. Preliminary to this, I'd like
>> people's comments.
>> 
>> I wish to make the pas startement optional. That is, in any Python
>> program where there is a pass it can be left out and the meaning of
>> the program is unchanged.
>
>The purpose of pass to explicitly disclaim what you're intending to do;
>it reduces programmer error.  In cases in code where you see something
>like:
>
>	if condition:
>	doSomething
>
>It is far more likely that this is an indentation mistake rather than a
>deliberate null statement.  The pass keyword removes all doubt:
>
>	if condition:
>	    pass
>	doSomething
>
>The pass keyword helps solidify this kind of behavior.  Convenience and
>beauty can't really hold a candle to utility (even if it is marginal
>utility) and precendent.

A warning might suffice.  If the problem is mainly writing code that will be
filled in later, ignoring the warning should be fine during development,
and when you're ready to release it will remind you of any you haven't
implemented yet.  (Then again, I hate seeing warnings for problems I know
about, since they distract me from new warnings, so I'd probably prefer to
fiddle with pass than to see a warning every time.  But at least there'd be
the option.)

Joe


From tejarex at yahoo.com  Sat Apr 13 10:39:49 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 13 Apr 2002 14:39:49 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com>

"Ralf Juengling" <juenglin at informatik.uni-freiburg.de> wrote in
message news:xdulmbr51x3.fsf at leto.informatik.uni-freiburg.de...
> I understand now, that 'i+=1' is no real in-place operation

Nor is it a false in-place operation :<)

> but  just a shortcut for 'i=i+1' since int is immutable.

Correct.  This is true for all immutable types.

> For the sake of clarity, an 'in-place' operation should really
> work in-place, don't you think?

Yes, but +=, etc, are not in-place operations, and are not advertised
as such.

Python is not C, or any other language except itself.

Terry J. Reedy





From aleax at aleax.it  Mon Apr  1 13:39:28 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 18:39:28 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017680049.17235.python-list@python.org>
Message-ID: <AP1q8.41430$pT1.1144020@news1.tin.it>

Pearu Peterson wrote:
        ...
> users/developers/educators. From what I have learned last years about
> Python development (other people comments being the main source), then it
> is not exactly a democratic process where majority rules.

Absolutely not, *thanks be*.  I strongly suspect I wouldn't _want_ to use
a language developed by "a democratic process where majority rules" -- cfr
e.g. Larry Wall's quip "Perl is worse than Python because users WANTED
it worse"... I think there's truth in it.  A single language architect's 
intuition, supported by advice and reasoned opinions of whoever he cares
to listen to (as opposed to votes which are just counted, no matter who
expresses them and what arguments, if any, they're bolstered by) stands
a stronger chance of producing a gem than a committee's productions.

I don't agree with every decision by the BDFL, but I'm still pretty happy
that HE decides, not a committee -- and not even me (if I somehow felt
sure that my decisions would be better, I could after all go elsewhere... I
doubt a better language would result that way, though).


Alex



From stephen at xemacs.org  Tue Apr 30 08:30:13 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 30 Apr 2002 21:30:13 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> <mailman.1020071811.14821.python-list@python.org> <87adrm38fu.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnacrvhu.6u4.huaiyu@gauss.almadan.ibm.com>
Message-ID: <87adrl1i96.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> writes:

    Huaiyu> If you stick to RMS's intent, you'll see that it is indeed
    Huaiyu> about consumer protection from the very beginning.  There
    Huaiyu> are plenty of evidence all over the net, from GNU, FSF and
    Huaiyu> RMS writings, and also summarized by several posters here.

I've read the writings.  Let's agree to disagree on interpretation; it
won't advance the technical discussion.

So, how about a tentative definition, then:

    _Consumer protection_ is a framework of regulation which places on
    the vendor, and possibly a third party, the responsibility for
    preventing certain kinds of damages to the consumer.

What I have in mind by "certain kinds of damages" are damages which
the vendor (or regulator) has a lot more power to prevent than the
consumer, due to information about the product, or access to
complementary technologies, or experience, or market power, for
example.  The reason for the "possible" third party is that in some
cases it makes sense to have a specialized regulator inspect the
products or transactions to be sure they're fair, eg in the case of
food or drug purity.  However, in some regimes (like the US with
medical malpractice), the burden of enforcement may be placed on the
consumer via a civil suit.  Laura's already cautioned me not to be
biased toward U.S. institutions, but I don't want to rule them out,
either.

This definition is tentative because it's biased against "free
software as consumer protection".  With this definition, it seems
natural to me to say "Free software exempts the vendor from the burden
of responsibility by shifting it back on the consumer," in some sense.
That doesn't sound like "protection" to me.  Can we fix that?  (I can
see and acknowledge my bias, but it's hard for me to work around it.)


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN


From Christopher.Saunter at durham.ac.uk  Sun Apr 14 10:48:30 2002
From: Christopher.Saunter at durham.ac.uk (Christopher Saunter)
Date: Sun, 14 Apr 2002 14:48:30 +0000 (UTC)
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <3cb8c690$0$19942$9b622d9e@news.freenet.de>
Message-ID: <a9c4ru$7dq$1@sirius.dur.ac.uk>

Jens Baader (nospam at nospam.com) wrote:

: "The Python interpreter is very stable"*30000

: At least at my system the interpreter promptly
: hangs itself. I guess I just found my first bug..
: (not bad I just read the first few pages of the
: tutorial)

Probably not a bug...  If I do that from Python
under a win32 console, things get silly as Python
spends the next five minuites printing stuff out.

If I run that from IDLE, it apears to hang, but 
it's actually just spending a few minuites
drawing rather a lot of text, before responsivity
returns.  I know this through having accidentally 
printed the odd 10^4 by 10^4 array in IDLE by
mistake...

Cheers,
	Chris Saunter





From martin at v.loewis.de  Mon Apr  1 08:24:39 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 15:24:39 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it>
Message-ID: <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de>

Alex Martelli <aleax at aleax.it> writes:

> b. The hoped-for uniformity of notation won't happen -- we've already 
> seen on these threads people who want every constant to be uppercased
> (to make it 'stand out') and so will use TRUE=True etc anyway.

That won't happen. People don't dare rename builtins, normally, and
will appreciate the increased readability if the builtins keep their
names.

> f. the claimed conceptual benefits in easing the lives of programmers and
> newbies in particular I consider as illusions.  Programmers want booleans
> iff they're used to them from previous languages -- I've never seen people
> used to (e.g.) APL, good old (uncommon?-) Lisp, or (dare I say it) Perl,
> feel sad and deprived by the lack of booleans in their respective languages.

That does not prove the claimed conceptual benefits as
illusions. Programmers that are used to booleans from previous
languages might well find their lifes easier. Even people that you
have not seen sad might enjoy this enhancement - they just did not
know what they were missing :-)

> Which leads me right onto disadvantages of the PEP -- because that's
> the main one I see: extra conceptual load in teaching the language.  "0 can
> be spelled either 0 or False" with subtly (but generally irrelevantly:-) 
> different consequences... 

0 cannot be spelled False. 0 is spelled 0 (or perhaps 00, or 0x0). 0
is a integral number, False is a boolean. These are different things.

In the past, untruth could be spelled in many ways (0, None, (), ...);
with the PEP, you can say "untruth is spelled as False". You still
need to explain that other things "count as false" also, but the
proper spelling becomes "False" (whereas 0 served this position only
by convention).

> I've wasted too much of my life trying to explain to people why they must
> NOT code "if some_condition==True:" or the equivalent thereof in various
> languages, but just "if some_condition:".  The "== True" appendix, useless
> at best, often damaging, seems one of the most natural errors by newbies.

That is a problem. Perhaps comparing booleans and non-booleans for
equality should cause a warning (for that matter, perhaps even
comparing booleans and booleans should cause that warning...)

More realistically, the *parser* should emit a warning if it sees the
name True used in such a context.

> In Python I was blessedly spared that waste of teaching time.  If the PEP
> is approved, I'll be back to wasting that time and effort (perhaps more,
> given how seductive "is True" will appear to many newbies...!).

Fortunately, "is True" won't be much worse than "== True" (except when
you have 1 on the other side).

> And what happens to other forms of output, such as marshaling and
> pickling -- does the bool or int nature of a 0 or 1 get preserved or
> lost in a round trip?

That's a good point: booleans should round-trip when pickling. I guess
it is a bug in the PEP that this hasn't been spelled out.

Regards,
Martin


From bogus@does.not.exist.com  Sun Apr 21 18:46:11 2002
From: bogus@does.not.exist.com (James Chapman)
Date: Sun, 21 Apr 2002 22:46:11 GMT
Subject: newbie dictionary question
Message-ID: <TiHw8.99936$CH1.86776@sccrnsc02>

Hi,

I have a question about dictionaries. I've been coding up a system of
objects that represents the reuslts of different auctions. One thing I've
been doing constantly though is iterating over dictionaries where the keys
are the prices of the bids and the values are the bids at that amount.

I was wondering if there is an easy way to consolidate to dictionaries into
one, where if they both have the same key the resulting value can be
concatenated.

Also is there any similar syntax for dictionaries like the list syntax
"[some_function(x) for x in other_list]"?

Thanks,

James Chapman




From jajvirta at cc.helsinki.fi  Sat Apr 27 16:57:20 2002
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 27 Apr 2002 20:57:20 GMT
Subject: [OT] What is Open Source?
References: <mailman.1019858688.3528.python-list@python.org> <7x1yd2qc4q.fsf@ruckus.brouhaha.com>
Message-ID: <slrnacm45g.jk3.jajvirta@sirppi.helsinki.fi>

26 Apr 2002 16:21:57 -0700 Paul Rubin wrote:
> So we'l have to start issuing PEP's with the disclaimer "I found this
> PEP in my grandma's old trunk in the attic, dated April 1, 1920, so
> it is out of copyright now."

And if "they" start questioning how come your grandma cared about
booleans in Python, you could refer to the time machine and come up with
a reasonable story.  And if even 10% of the stuff they show on TV about
court practice in United States is true, the jury would swallow the
story whole. Now, PSU has prepared a emergency plan that


From DavidA at ActiveState.com  Tue Apr  2 21:04:12 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Tue, 02 Apr 2002 18:04:12 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
Message-ID: <3CAA631C.7674F116@activestate.com>

Guido van Rossum wrote:
> 
> Alex Martelli wrote:
> > I _have_ seen old-time Fortranites rant quite convincingly against that
> > useless, newfangled LOGICAL stuff.
> 
> And that is exactly what many of the responses in this thread
> (including Laura's over-long FUD piece) sound like.
> 
> I'm reading (at least skimming) everything, but the more I read the
> reactions, the more I believe that this is just the response I have to
> expect for *every* change to the language I propose, and the more I'm
> tempted to follow my gut instincts.  

IIRC, there wasn't nearly this much heat in the discussion over the
long/int unification or the type/class unification proposals, even
though the latter especially was much more complicated, ambitious, and
compatibility-challenging.

This reminds me of the old saying about academic politics: "the smaller
the stakes, the more vicious the fights".

IMO, you shouldn't expect unanimity on any language change. 
Furthermore, there's going to be a natural bias for people who are
against an idea proposed by GvR to want to speak up more than those who
agree with him, since they feel they're on the winning side already.

My own take-away from this thread is that Laura identified a compelling
description of what truth testing in Python currently corresponds to. 
Unlike Laura, I don't think that the PEP invalidates the core of the
"something vs. nothing" distinction.  False would just be another form
of nothing and True something.  Which leads to next year's Python
slogan:

     Something is True, and Nothing is False!

or is that, in the words of the late Milton Berle:

     You can fool some of the people some of the time, and all of the
people some of the time, and that's usually enough.

--david

PS: I wish that this much heat got generated discussing changes and
additions to the standard library...



From fabrice.lorrain at chello.fr  Sun Apr 28 06:21:10 2002
From: fabrice.lorrain at chello.fr (Fabrice Lorrain (home))
Date: Sun, 28 Apr 2002 12:21:10 +0200
Subject: win32api : getfileversion?
Message-ID: <3CCBCD16.8010604@chello.fr>

Hi all,

I'm new to python and I try to get informations (Name, version)  on .exe 
files under win*.

Under the MS API & perl win32API, there seems to have a GetFileVersion 
wich does exactly what I need.
I couldn't find the equivalent in win32all API.

Any idees or suggestions ?

Thanks,

	Fab






From loewis at informatik.hu-berlin.de  Wed Apr  3 04:18:47 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Apr 2002 11:18:47 +0200
Subject: problem with http post
References: <cf43c01d.0204022342.73e5c8fa@posting.google.com>
Message-ID: <j4zo0lywaw.fsf@informatik.hu-berlin.de>

mongiaamit at yahoo.com (Amit Mongia) writes:

> Can anyone please tell me what the possible error can be. 

You need to send a Content-Length as well.

Regards,
Martin


From pyth at devel.trillke.net  Fri Apr 19 12:29:05 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 19 Apr 2002 18:29:05 +0200
Subject: minor generator question
Message-ID: <20020419162905.GH28515@devel.trillke>

i have a recursive generator to flatten
nested containers (lists and tuples):

def flatten(*args):
    for arg in args:
            try: 
                for i in arg:
-->                 for l in flatten(i):
-->                     yield l

            except TypeError,e: yield arg


Is there a better way to say that i want all results
generated from a (sub-)generator be "forwarded" ? 
(see the two marked lines). 

thanks, holger




From ryanmorillo at hotmail.com  Mon Apr  8 22:58:05 2002
From: ryanmorillo at hotmail.com (Ryan)
Date: 8 Apr 2002 19:58:05 -0700
Subject: Off topic C question
Message-ID: <4a7fde99.0204081858.58baabdd@posting.google.com>

I know this is way off topic, but I figure the heavy hackers here will
understand my mindset better than those at comp.lang.C
    I have been doodling with python (nasty habit I know but I have to
do somthing with my snake) and think I may be ready to learn a second
language.  I am leaning towards C over C++ and Java even though I
loose objects it will ground me firmly in functional programing.  I am
wondering what book would be the best for a begginer?  The two I have
seen that look interesting are
C for Dummies (parts one and two) or 
The C Programming Language
by Brian W. Kernighan, Dennis M. Ritchie.
I am looking for some input from someone who knows and knows the
problems I will face coming from a __nice__ language like python.
Thanks (and sorry about the off topic)


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 18:58:19 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 15:58:19 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cb8b86a@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3ofgn1avt.fsf at mira.informatik.hu-berlin.de...
| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
|
| > I think an 'xor' operator is needed.  The 'xor'
| > operator means 'logical-exclusive-or', and
| > returns a 1 or 0.
|
| In Python 2.3, you will be able to write
|
|    bool(a) ^ bool(b)
|
| Seems pretty clear to me.
|
| Regards,
| Martin

You didn't read my post well enough-- we can already _DO_
that with Python v2.2.x:

(not a) ^ (not b)

IMHO, neither the above, nor the future v2.3.x idiom:

bool(a) ^ bool(b)

is as clear as:

a xor b

...

Everyone_is_entitled_to_have_my_opinion'ly yours,

--Ken Peek





From ballabio at mac.com  Fri Apr 12 16:14:55 2002
From: ballabio at mac.com (Luigi Ballabio)
Date: Fri, 12 Apr 2002 22:14:55 +0200
Subject: Macintosh Development
In-Reply-To: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>
Message-ID: <a05100305b8dceff6b668@[1.83.220.98]>

At 4:58 PM +0000 4/12/02, Bryce C. Mason wrote:
>I am doing my development on a PC (activeware's distribution) and the final
>project will be running on a macintosh.  Umm...I am a little worried that
>there is not a command line interpreter in the Mac, at least one I have not
>seen.  Help.  How do I port my project to the mac?

It depends on your project. If it's pure Python, it should run on the 
Mac distribution available at http://www.cwi.nl/~jack/macpython.html 
which includes a command-line Python interpreter.

I might be answering the wrong question, though. Did you mean a 
command-line *Python* interpreter, or a terminal shell? If the 
latter, how is it used in your project?

Bye,
	Luigi

-- 




From donald.braman at yale.edu  Thu Apr  4 13:13:44 2002
From: donald.braman at yale.edu (don)
Date: Thu, 4 Apr 2002 10:13:44 -0800
Subject: dbi odbc error
References: <a8i4hh$htk$1@news.ycc.yale.edu>
Message-ID: <a8i4vj$ieq$1@news.ycc.yale.edu>

Nevermind. :-) For any who encounter the same problem, swap double and
single quotes in the sql command.

"don" <donald.braman at yale.edu> wrote in message
news:a8i4hh$htk$1 at news.ycc.yale.edu...
> I'm new to python. I'm trying to update tables in an MS Access database. I
> can connect and retrieve data just fine. The following *does* work well:
>
>                 import dbi, odbc
>                 conn = odbc.odbc(mybase)
>                 cur = conn.cursor()
>                 cur.execute('SELECT * FROM mytable WHERE myfield < 1830')
>                 print cur.description
>                 for tup in cur.description:
>                         print tup[0],
>                 print
>                 while 1:
>                         rec = cur.fetchmany(10)
>                         if not rec: break
>                         print rec
>
> But when I try to update a field the in table, I get an error. So, the
> following does *not* work:
>
>                 import dbi, odbc
>                 conn = odbc.odbc(mybase)
>                 cur = conn.cursor()
>                 cur.execute('UPDATE mytable SET mytable.myfield= "test";')
>                 conn.commit()
>                 cur.close()
>                 conn.close()
>
> The error I get is:
>
>   File "F:\simple update.py", line 4, in ?
>     cur.execute('UPDATE mytable SET mytable.myfield = "test";')
> dbi.program-error: [Microsoft][ODBC Microsoft Access Driver] Too few
> parameters. Expected 1. in EXEC
>
>
>
>
>




From donn at u.washington.edu  Wed Apr 17 22:54:06 2002
From: donn at u.washington.edu (Donn Cave)
Date: Thu, 18 Apr 2002 02:54:06 -0000
Subject: Python webmail packages?
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com>
Message-ID: <1019098444.973175@yasure>

Quoth Paul Rubin <phr-n2002a at nightsong.com>:
| Does anyone have a favorite webmail system written in Python?
| Something like Squirrelmail (www.squirrelmail.org) but written in
| Python and hopefully with fewer bugs ;-).  It would be good if it
| talks to an IMAP server like Squirrelmail does, but that's not mandatory.

I don't have any favorites, but here's another that seems to be
more or less inactive (latest release here is a 1-year old alpha),
http://sourceforge.net/projects/worldpilot/.  Uses Zope.  Does
speak IMAP.

	Donn Cave, donn at drizzle.com


From aleax at aleax.it  Mon Apr  1 15:01:43 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 20:01:43 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de> <KM_p8.41075$pT1.1129125@news1.tin.it> <m3g02f76o6.fsf@mira.informatik.hu-berlin.de>
Message-ID: <H03q8.41590$pT1.1149573@news1.tin.it>

Martin v. Loewis wrote:
        ...
>> > Programmers that are used to booleans from previous
>> > languages might well find their lifes easier. Even people that you
>> 
>> In the same sense as they might well find their lifes easier if suddenly
>> they had to start decorating their Python programs with deuced braces
>> or BEGIN/END to indicate block structure -- they might find the
>> familiarity comforting.
> 
> Nah, that is hardly the same thing. With the PEP, nobody *has* to do
> anything; existing programs will continue to work just fine (with very

"if suddenly they _were allowed_ to start decorating", etc, then -- optional
BEGIN/END rather than mandatory.  "more than one way"...?

> very few exceptions). If people want to return 0 when they mean False
> - fine.

_Mostly_ fine... "with very few exceptions" -- sporadic breakage.


>> I think teaching this as alternate ways of *spelling* equal (but
>> distinct: ==, and essentially interchangeable in use, but NOT 'is')
>> values will be the least of evils.
> 
> I disagree. Somebody could use the language (with the change) and
> believe booleans are completely unrelated to numbers, and would rarely
> find this view incorrect - at which time you'll need to explain that
> booleans interact with numbers in a certain way.

Somebody could use the language under all sort of weird misconceptions,
as shown by reasonably frequent questions both here and on python-help
by people who've been using the language for quite a while and still
display some fundamental misunderstandings.  My preference when teaching
is not to encourage such misconceptions "for simplicity": experience shows
that such oversimplifications, meant to be "stepping stones" towards some
better future understanding, all too often end up as millstones 
metaphorically around students' necks.

It's best for what one's teaching to BE simple, rather than to try to make 
it APPEAR simple by hiding complications.  If complications *are* there (be 
it for good and sufficient reasons, or not), students had better be warned
(rather than waste hours "debugging" an isinstance(x, int) before being told
that, oh btw, bools ARE int...).


>> Exactly the same set of objects (plus the new object False, that's ==0
>> anyway and can be used just about wherever 0 can be)
> 
> While it is true that False can be used everywhere, it doesn't mean it
> should be. Use booleans if you want to express the value of a
> predicate; use integers if you want to count things.

"Do that because that's what you're supposed to do, even if it makes
no difference" is another approach that's totally alien to my style of
teaching.  It's not rare in the evolution of a system for an assumed
"boolean" to become a count -- "do I have a printer or not?" becomes
"how many printers do I have?", a set (each item being or not being in
the set) becomes a bag (each item being in the bag 0, 1, ... N times),
and so on.  Fortunately, these issues are not as sharp in Python as they
would be in languages that can publish type-rigid interfaces, at least
as long as booleans remain confusable-enough with integers.  The best
case, of course, would be _total_ confusability... no separate boolean
type at all, just like today.  But given the high 'confusability' of the PEP
proposed booleans, the cost, or damage, is basically conceptual, more
than pragmatical -- "introducing a distinction without a difference" being
almost as naughty as "filling a badly-needed gap":-).


>> The "by convention" status of the canonical spelling for untrue
>> still holds, except that you want to change the convention to the
>> new object which exists solely for the value of changing that
>> canonical spelling, more or less.
> 
> No. If the goal had been to change the canonical spelling, two new
> builtins would have been sufficient. The goal was to create a new type
> for those, in addition to the new spelling.

Introducing two builtins would have a subset of the claimed benefits 
(attempt to standardize spelling to other than 0/1, but only in source,
not in program output) and a subset of the costs I see (e.g. the
"==True" issue would probably arise even then).  But note that the
new type was probably the simplest way to affect spelling in output
(by defining __str__ or __repr__), and the major issues are all
connected with spelling (in source and/or output).  Spelling apart,
you _could_ have an almost-transparent subtype of int and only
have very few issues indeed (breaking on type()==type() horrors
that probably do deserve to break, pickle roundtrip) -- maybe even
to the level that newbies _might_ be left ignorant of the change.
(I'd still rather not have that almost-transparent subclass of int, but
it would be such a minor issue -- almost invisible to most users,
blissfully -- that I'd probably not bother posting about it, I think).

Naah, "the spelling's the thing" -- in both output and source (two
issues which are connected by desiring eval(repr(x))==x).


>> 'if x == y:' should cause no warning, of course -- it's a general,
>> perfectly
>> good polymorphic usage.  Why should the user have to silence warnings
>> or break the smoothness of polymorphism to ensure against x, or y, or
>> both being of this newfangled bool type?
> 
> Indeed. I wonder how often the warning would trigger, though. I'd
> expect that people very rarely compare the result the got from a
> predicate to something else.

Indeed, I've seen people code "(x&&y) || ((!x)&&(!y))" (including the
overflow of parentheses) in cases where each of x and y were known
to be surely true or false, so the much-smoother "x==y" would have
sufficed -- they just hadn't _thought_ of that... some sort of
psychological block, I guess.  Still, that's no reason to penalize
those programmers who _are_ more facile with choosing simple
and smooth expressions for their logical needs.


>> > More realistically, the *parser* should emit a warning if it sees the
>> > name True used in such a context.
>> 
>> ...assuming the parser can determine at parse time that True stands
>> for the built-in value at that point -- or do you want to give True an
>> exhalted semi-keyword status, beyond that of, say, None, a dignity that
>> will produce frequent warnings if you rebind that identifier e.g. in a
>> local namespace?
> 
> I would indeed have the warning produced for all occurrences of True -
> perhaps a different warning if the parser can tell that True is bound
> at module scope. This is essentially what pychecker does in many
> similar cases, and Guido is in favour of moving some of the PyChecker
> warnings into the interpreter.
> 
> There would be an option to turn off parser warnings, of course.

OK.  On reflection, the 'slippery slope' of adding lint-like functionality
to the interpreter is one on which I probably won't mind sliding down -- as
long as there isn't only 'one' option (e.g. on the command-line) but also
a reasonably easy way for the user who knows what he or she is doing
to assert that to the checker (the warnings module seems a bit clunky
for that, but maybe I just haven't understood it well enough yet) for a
specific statement.


>> I know 'slippery slope' arguments are disliked around here, but both
>> this exchange and the various voices raised in clamor towards
>> 'stronger, purer booleans' suggest to me that PEP 285's own slope
>> may be slippery indeed.
> 
> These are different issues, though: those voices want to ban
> interactions between booleans and numbers; I want to ban binding of
> names that are also builtins (starting with None, True, and False).

Not just bindings -- also comparisons, you said.  A warning when any
builtin name is rebound (except when specifically turned off) sounds
good -- no reason to start from None, True, and False, when the
built-in names most often accidentally rebound are probably open (from
an errant 'from os import *' and the like), and type names (it's _so_ 
tempting to name an integer 'int', a list 'list', a dictionary 'dict', etc).
Warnings on comparisons and/or other operations involving booleans
are on the slope with very different slant that leads to just where I'd
_detest_ to see Python go;-).


>> No, but it's one more warning-case you have to propose (and how else
>> will one check for the wonderful new objects' identities if 'is True'
>> warns?! by twists and turns of isinstance checking for bool then int()
>> and == checking for 1...?!).  And 'is True' is very newbie-alluring: a
>> true 'attractive nuisance' (in more than one meaning of the phrase).
> 
> Ok, so the problem is more involved. I would suggest to produce the
> warning for 'is True', but not for 'True is' ... (I'm not sure whether
> I mean this seriously :-)

Given today's date, I wouldn't know.  I did use to campaign (not to much
use) for '0==x' in C (to avoid the typical bug of dropping an '=' from the
more usual 'x==0'), so playing around with operators that _are_
commutative but more error-prone depending on operand order is
hardly an entirely new concept:-).


>> > That's a good point: booleans should round-trip when pickling. I guess
>> > it is a bug in the PEP that this hasn't been spelled out.
>> 
>> If they do round-trip, then, it seems to me, we get a 'gratuitous'
>> incompatibility of pickles produced by 2.3 wrt 2.2 programs trying to
>> read them -- one more annoyance (and not a very attractive one either,
>> this time)
> 
> I'm not sure how big this problem is: it should be easy to add
> bool-loading function to Python installations for use with Python
> before 2.3. Packages that need to support cross-version pickling could
> incorporate such a feature.

Probably 'should', at least -- yet more work ('transitional', to be sure),
and so yet another cost of the PEP's adoption.  I don't consider it to
be necessarily all that 'easy' to encode metadata for all objects to
indicate which fields have switched to booleans, by the way -- quite
possible, sure, but 'easy' may be an overbid.  At the very least, it
seems to me that qualifying it as an 'annoyance' is hardly overdoing it.


>> And of course old pickles will give 0 and 1 so that much work will
>> be needed to reconstruct them into False and True when appropriate
>> *IF* it matters at all if something is True or 1.
> 
> This is a transitional issue. It may matter, but it may not matter
> enough to take action - depending on the application. An application
> that worries about cross-versioning pickling (presumable across
> versions of its own software as well) will have to take efforts
> already; avoiding to dump booleans will be a minor complication only.

An annoyance -- another little nugget on the 'costs' plate of the
scales, to add to the others.


>> So _why_ pay the price -- not a huge, world-racking one, but still
>> one nevertheless?
> 
> I could repeat the rationale of the PEP here, but I guess the question
> was rhetoric.

No, but neither would it be appropriate to repeat arguments against many
of which I've been arguing (either to point out why a claimed benefit would
not be one, or to indicate it would be truly minimal); rather, a summary of
what true, concrete benefits are left to offset against the costs might be
appropriate.  But as clearly we don't agree on most of what we've been
discussing on this thread, it may be that such a summary would serve no
real purpose.


Alex



From jb at cascade-sys.com  Tue Apr 30 14:38:06 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 11:38:06 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>	<3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca> <mailman.1020126468.7915.python-list@python.org> <G8xz8.14199$xy4.5322@atlpnn01.usenetserver.com>
Message-ID: <3CCEE48E.3B8BFDB3@cascade-sys.com>

Steve Holden wrote:

> Well, speaking as a US-based Englishman (from Yorkshire, where sarcasm comes
> as naturally as calling your best friend a stupid bastard) I have to say my
> experience is that *some* Americans will *say* they value straight talk
> until they actually receive some, Then one is being "aggressive".

Sad but true.  I trust you don't lump me in that category.

> Seriously, little mild-mannered me had to undergo self-administered behavior
> therapy to stop me from saying what I really thought, just in the interests
> of "getting along with my colleagues". The fact that they were building crap
> web sites in a piss-poor language with no architecture and not even the
> slightest acknowledgement that security might be important [...].

So tell us what you REALLY thought.

> We got on even better after I left, and now I haven't spoken to any of them
> for three years we are the best of pals.

Yeah, I gotta a few friends like that.  ;o)

> They can, but they can also alienate someone who is genuinely trying to
> reach a meeting of the minds over some topic. Particularly if there's any
> ambiguity about the intent of the sarcasm, which is after all a double-edged
> sword.

I reluctantly admit it can be unproductive with people you don't know well.
Even well intentioned humor can backfire, as humor generally advances at
someone's expense.

> What's a rhetorical question?

Would you really be asking if you didn't already know the answer?

> So would this be a good time to say "Shut up, you stupid bastard"?

Oh yeah?  I fahart in your general direction.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From tejarex at yahoo.com  Mon Apr 15 13:35:56 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 15 Apr 2002 17:35:56 GMT
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <0cEu8.47362$iw.4776263@bin4.nnrp.aus1.giganews.com>

"Stephen J. Turnbull" <stephen at xemacs.org> wrote in message
news:87wuv9xdr8.fsf at tleepslib.sk.tsukuba.ac.jp...
> I have no issue with the GPL, because the GPL only affects people
who
> chose to use GPLed code in their projects.
...
> I do have issues with some GPL advocates, because their arguments
> amount to claiming that nobody should have property rights in his
own
> code.  (rms says this explicitly, but very few are so courageous as
> he.)

Interersting.  If so, then RMS is 'guilty' of an almost humorous
contradiction.  The GPL *is* an assertion of property rights in code,
backed up by FSF lawyers.  It explicity extracts a price from those
who would use GPL software.  It is not directly monetary, but it is a
price nonetheless, as evidenced by the fact that many do not choose to
pay, even though they might be willing to pay a reasonable amount of
money to get the same software without the non-monetary price
attached.

Terry J. Reedy





From frithiof.jensen at removethis.ted.ericsson.dk  Thu Apr 11 04:04:28 2002
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 11 Apr 2002 10:04:28 +0200
Subject: Java to Python translation
References: <3CB01A9E.4DFB9D9D@sympatico.ca>
Message-ID: <1018512289.446501@robert>

"vio" <vmilitaru at sympatico.ca> wrote in message
news:3CB01A9E.4DFB9D9D at sympatico.ca...

> First, I had some problem running the demo in my current window manager
(icewm),
....

That is what I do not like about Java: The platform independence is broken
in subtle and often surprising ways and there is version dependancies
introduced by the JVM. Reminds one of Windows, in fact - Bad!

> Second, it did reverse engineer (java to UML) simple java sample code, but
I
> couldn't make it to generate python code from the UML model (nor CORBA IDL
code,
> another supported language).

Thats appalingly bad - just as well you did not shell money for this, eh?

I *would* give it a spin on Windows thouogh, just in case of it being "yet
another JVM bug" - Linux is not well supported in the Java arena and some
distributions even comes preloaded with the Kaffee JVM, which is a nice
enough effort, but mostly does not work with 3rd-party Java applications.
With SUN's JVM, on a RedHat distro, there is somewhat better odds.

> Third, and the big "show stopper" though, is that during 'java to UML'
> translation, all the code inside my java methods are simply lost.

That's what is *supposed* to happen: UML tools captures *design*; you have
to provide the implementation yourself - they do not translate code, *you*
have to do that (Whatever you do IMNSHOO :-).





From johnroth at ameritech.net  Tue Apr 16 07:40:49 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 16 Apr 2002 07:40:49 -0400
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5 <3cbb7e33@news.mhogaming.com> <a9fvvl$nr4$1@watserv3.uwaterloo.ca>
Message-ID: <ubo3dnl1k93f93@news.supernews.com>

"Joe Mason" <jcmason at student.math.uwaterloo.ca> wrote in message
news:a9fvvl$nr4$1 at watserv3.uwaterloo.ca...
> In article <3cbb7e33 at news.mhogaming.com>,
> Ken Peek <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
> >Quite frankly, I don't give a damn HOW the long (or int) is
represented
> >internally to the machine-- I simply ALWAYS want the same to
interface to an
> >object that behaves as if it is a twos-complement integer.  When I
left shift
> >it, I don't want the damn SIGN to change-- when I right shift it, I
want to be
> >able to CONTROL the sign bit that is shifted in, when I display the
integer in
> >decimal, I want a minus sign prepended to the string if the number is
negative,
> >and when I print out the number in hexadecimal format, I DON'T want
"sign and
> >magnitude"-- I simply want a string of hex digits with a '0x'
prepended to them,
> >representing the twos-complement form of the number, and I want 2*N
hex digits
> >EVERY TIME, where N is the current size of the integer (in bytes.)
>
> As I see it, the problem here is an interface mismatch.  The fact that
> integers are represented as twos-complement numbers is, most of the
time,
> an implementation detail.  As you point out, you don't care if it
actually
> *is* a bitfield, you just want the interface of one.  This is an
interface
> that should definitely be available, but not necessarily for every
integer:
> most of the time, it's not important at all.

I think you've hit the root of the issue: integers are not the same as
bit strings. We tend to get confused because most hardware architectures
treat them the same when it gets down to machine representation.

> So I'd argue that the ability to behave as a twos-complement bitfield
is
> not something a standard integer should support, and that what you
need is
> a bitfield object (which a constructor that takes an integer).  This
could
> be a new language construct or completely encapsulated in a module.
(I'd
> lean towards a module, but I think that would mean that every bitwise
logical
> operator would become a method or function, and I can see a case being
made
> that keeping the operators available is important enough to be part of
the
> language.)

I don't know whether there is really any need for bitfields to behave
as integers, or vice versa. If so, I presume that int() on a bitfield
instance, and bitfield() on an integer would do the job.
>
> Also, to be completely clean, any operators that currently have no use
> except to treat integers as bitfields should be made to work only on
> bitfield objects instead.

That would break backwards compatiblity, so it won't happen.

>   x = bitfield(i)
>   y = bitfield(j)
>   z = x << 8 & y
>
> or
>
>   x = bitfield(i)
>   y = bitfield(j)
>   z = x.shiftleft(8).bitand(y)
>
> All the arguments about whether doing things with sign extension is
> violating the transparency of integers (in an 'it should just work'
sense)
> now become invalid (or at least are confined to the constructor and
any
> methods to convert back to an int).  The bitfield contracts to work
just
> like a string of bits.
>
> Obviously this is an unrealistic proposal: removing bitwise logical
> operators isn't going to happen.  But I'm still interested, in case I
> ever design a Python-like language (heh): from your point of view,
would
> this be acceptable or desirable?  Or would it be too much of a pain in
the
> ass?  (I don't work with bitfields much, but I can imagine that if you
> commonly have to keep swapping back and forth between the two ways of
> conceiving of a number, this would be a bad idea.)

Actually, it's a good proposal, except for removing the operations
from integers. It's got all of the hallmarks of something that could
happen: it can be implemented as a new type, doesn't involve
any new keywords, it can be prototyped independently of changes
to the python core, etc.

John Roth
>
> Joe




From mertz at gnosis.cx  Fri Apr 12 22:39:52 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Fri, 12 Apr 2002 22:39:52 -0400
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
In-Reply-To: <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca>
References: <mailman.1017656355.5887.python-list@python.org>
 <12bb92e6.0204091223.738e0798@posting.google.com>
 <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca>
 <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
 <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca>
 <MPG.172101b4b7f13f639896ae@news.nwlink.com>
 <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca>
Message-ID: <4p5t8kKkXI6d092yn@gnosis.cx>                

|> (See http://www.fsf.org/philosophy/categories.html)

|I overlooked the specific legal meaning of "Public Domain"...
|However, I was still correct in saying that just because you release in
|the public domain doesn't mean you're releasing "free" or even more
|generally "open source" software

Huh?!

Did you look at the URL given above?  At the top of that page is a very
nice Venn that shows many of the categories of software, for a cost and
freedom POV.  Among other features, this diagram (correctly) shows
Public Domain software as entirely contained in BOTH of the overlapping
categories Free Software and Open Source.

Study this chart a bit, it is extremely good.  (The one quibble I might
have with it is the fact that "Shareware" is entirely encompassed in the
"Free Download" category.  I used to get shareware programs on floppy
distributions before the internet was widespread or bandwidths were
fast...  I'm pretty sure an occassional shareware program was
distributed *only* on certain limited media.  It really is a tiny
quibble).

The quote also, btw., is quite consistent with the Venn diagram.

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From BPettersen at NAREX.com  Thu Apr 18 17:07:42 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 18 Apr 2002 15:07:42 -0600
Subject: Changing the repr() of a class?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158DF3@admin56.narex.com>

> From: Lulu of the Lotus-Eaters [mailto:mertz at gnosis.cx] 
> 
> Alex Martelli <aleax at aleax.it> wrote previously:
> |class FunkyReprMeta(type):
> |    def __repr__(cls):
> |        return "My wonderful class %s" % cls.__name__
> |class FunkyReprObject:
> |    __metaclass__ = FunkyReprMeta
> |class C(FunkyReprObject):
> |    "whatever you want here"
> 
> I don't really understand what a metaclass buys you here.  I 
> think I might be missing something important, since Alex proposes it.
> 
> However, it appears that one could simply define the __repr__ 
> method in the regular class FunkyReprObject, ignore 
> metaclasses altogether, and wind up with the same results.  
> What is better about Alex' approach?

I'm assuming you mean something like:

>>> class C(object):
...   def __repr__(klass): return 'class %s' % klass.__class__
...   __repr__ = classmethod(__repr__)
...

Unfortunately, that doesn't work, so I'd stick with Alex' way <wink>

-- bjorn




From quinn at pfennig.ugcs.caltech.edu  Tue Apr 30 18:09:35 2002
From: quinn at pfennig.ugcs.caltech.edu (Quinn Dunkan)
Date: 30 Apr 2002 22:09:35 GMT
Subject: Dictionary that adds items to itself
References: <4620daca.0204250809.563a8c0b@posting.google.com>
Message-ID: <slrnacu5gr.eft.quinn@pfennig.ugcs.caltech.edu>

On 25 Apr 2002 09:09:27 -0700, gyromagnetic <gyromagnetic at excite.com> wrote:
>Hi,
>I would appreciate some advice on the following. I would like to
>create a dictionary-like object with the same properties as a normal
>dictionary except with regard to adding items. When items are added, I
>would like them to be processed by a function that may generate
>additional items. I would like these additional items to be added to
>the dictionary, have the function applied to them, have the resulting
>items added to the dictionary, etc.

Off the top of my head:

import UserDict
class Apply_dict(UserDict.UserDict):
    def __init__(self, mapf):
        self.mapf = mapf
        UserDict.UserDict.__init__(self)
    def __setitem__(self, k, v):
        v, extras = self.mapf(k, v)
        self.data[k] = v
        for k, v in extras:
            self[k] = v

Just make sure mapf() doesn't generate extras that eventually generate the
original input or it'll loop.


From sak at essc.psu.edu  Tue Apr 23 11:10:00 2002
From: sak at essc.psu.edu (Sridhar Anandakrishnan)
Date: Tue, 23 Apr 2002 11:10:00 -0400
Subject: BLT/Python PMW problem
Message-ID: <3CC57948.14D0ABD6@essc.psu.edu>

Hi, 

When I try and load BLT from within Tkinter, I get this error:

Python 2.2.1 (#1, Apr  9 2002, 13:10:27) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from Tkinter import *
>>> m=Tk() 
>>> m.tk.call('package', 'require', 'BLT')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TclError: couldn't load file "/usr/lib/libBLT24.so":
/usr/lib/libBLT24.so: undefined symbol: TclpAlloc
>>> 

Any idea what's up?  I don't get this problem if I do this directly from
Tcl, so it is some sort of path problem?  I don't know Tcl well enough
to track this down - any help appreciated.

(python 2.2.1, tcl/tk 8.3.4, blt 2.4x)

Sincerely,
Sridhar.


From kseehof at neuralintegrator.com  Mon Apr  8 10:53:04 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Mon, 08 Apr 2002 07:53:04 -0700
Subject: Discussion of Python midi library
In-Reply-To: <3CB19732.8060304@mxm.dk>
Message-ID: <EFEFIHAINIINNELLHFAIGECCCAAA.kseehof@neuralintegrator.com>

I have written a midi module that wraps part of the low-level Windows
API.  I'm planning to release it to open source after we ship our
commercial product that uses it.

It doesn't actually do any of the items you requested though, since
I wrapped the midi input and output devices, but not midi streams or
files.

I believe that you will find midi file processing in the Vaults:

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

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Max M
> Sent: Monday, April 08, 2002 6:12 AM
> To: python-list at python.org
> Subject: Discussion of Python midi library
> 
> 
> I am writing a Midi Module for Python.
> 
> I basically want it make it possible to:
> 
>      1) Read midi from a file
> 
>      2) transform the midi data
> 
>      3) Write midi to a file
> 
>      4) Play midifiles (especially on Windows)
> 
> The biggest challenge here is getting the internal representation of the 
> midi data right.
> 
> The midi data format in it's raw format is too specialized to make it 
> possible to load some midi-data and then do any kind meningfull 
> transformations on it.
> 
> So generally I want to be able to load in the data into a Python 
> specific objects representation that can be easily modified and then 
> saved back out as a new midi file.
> 
> I am doing algorithmical composition hence the need.
> 
> Has anybody got an interrest in participating in a discussion about how 
> such a module should be designed please contact me.
> 
> I allready have some code and some ideas.
> 
> regards Max M
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From rjones at ekit-inc.com  Tue Apr 30 19:40:31 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 1 May 2002 09:40:31 +1000
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
In-Reply-To: <ucuagjg8it3v4b@news.supernews.com>
References: <aan5hv$aor$1@bob.news.rcn.net> <ucuagjg8it3v4b@news.supernews.com>
Message-ID: <200205010940.31124.rjones@ekit-inc.com>

On Wed, 1 May 2002 09:35, John Roth wrote:
> "Raymond Hettinger" <python at rcn.com> wrote in message
> > I no longer think map(), filter(), and reduce() should even be
> > psuedo-deprecated,
> > but apply(), oct() and hex() are prime candidates.  The input()
>
> function
>
> > should
> > still be considered for genuine deprecation and deletion, but that
>
> reflects
>
> > my
> > view of input() as a bug.
>
> I strongly disagree about apply(). The only use I've found for it
> varies the function which it calls, not the parameter lists! Having
> variable parameter lists doesn't hack it - I need the variable
> function.

Eh?

>>> def f():
...   print 'hi'
...
>>> a=f
>>> a()
hi
>>>





From mgilfix at eecs.tufts.edu  Wed Apr 17 00:13:22 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 00:13:22 -0400
Subject: list_a-list_b
In-Reply-To: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>; from nobody@nobody.com on Tue, Apr 16, 2002 at 07:30:19PM -0500
References: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>
Message-ID: <20020417001321.E16954@eecs.tufts.edu>

  Here's a one line example that doesn't touch either of the lists
that you might like and I think it pretty clear.

  new_list = filter (lambda x, list=list_b: x not in b, list_a)

                 -- Mike

On Tue, Apr 16 @ 19:30, John wrote:
> 
> Hi, I need a help in writing a function that does
> 
> list_a=[1,2,4], list_b=[2,4]
> list_minus(list_a,list_b)
> print list_a #will print out a=[1] after removing all elements ([2,4])
> appearing in list_b

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/bc6fb53d/attachment.sig>

From joost_jacob at hotmail.com  Sat Apr 27 12:57:34 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 27 Apr 2002 09:57:34 -0700
Subject: portable fortune program
Message-ID: <13285ea2.0204270857.2f34f8a4@posting.google.com>

I have always liked the unix fortune program.
Unfortunately nowadays some *n.x's come without fortune,
and sometimes I have to work with Windows.  I wanted to be
able to do:
    C:\mydirectory> python -c "import fortune; fortune.quote()"
printing a nice quote to stdout.
The fortune module has to be able to locate a file with
the name 'allfortunes'.  And you want everything to be
portable.  And working with python 1.5.2 too.

The allfortunes file contains quotes separated by lines
with a '%' as its first character.  I downloaded a couple
of BSD fortune files and combined them into allfortunes.
If you put the following fortune.py file with an
allfortunes file into one of the directories in your
sys.path, or in you current working directory, it will work.

This is a first hack, it is slow, the allfortunes file is
not compressed, so comments are welcome, maybe you can
suggest improvements?  I was trying to put the allfortunes
file into a compressed string at the end of the fortune.py
file, so you do not need a separate file anymore, but I
did not see a best way to do it.  Please let me know if it
does not work with your python distribution.

joost_jacob at hotmail.com



# file fortune.py
import random, os

def randlistindex(yourlist):
    "Return a random index number into yourlist"
    return random.randint(0, len(yourlist)-1)

def quote():
    """
    Return a quote <string> from file allfortunes,
    assume the file allfortunes is located in the same
    directory as this module."""
    datafile = os.path.join(
      os.path.dirname(__import__(__name__).__file__),
      'allfortunes' )
    quotesfile = open(datafile)
    quoteslist, quote = [], ''
    for s in quotesfile.readlines():
        if s[0] == '%':
            if quote:
                quoteslist.append(quote)
                quote = ''
        else: quote = quote + s
    if quote: quoteslist.append(quote)
    print quoteslist[randlistindex(quoteslist)],


From SBrunning at trisystems.co.uk  Wed Apr 10 04:48:16 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 10 Apr 2002 09:48:16 +0100
Subject: Win32 and Excel question
Message-ID: <31575A892FF6D1118F5800600846864DCBCF68@intrepid>

> From:	Matthew [SMTP:matthewm at ihug.co.nz]
> I'm playing with win32COM with a little success. I have a large XL table
> with 8 columns and about 3500 rows. What I want to do is search only
> column B for a value and return the row number where this value is
> found. So, I have 2 questions:-
> 1) Howto construct a Win32COM statement to do this? So far I have this:-
> 
> >>>
> found=myxl.ActiveSheet.Range("B1:B3500").Find(<value_i'm_looking_for>
> However, this search finds the first match in any column between B1 and
> B3500. Yep, their are similar values in other cells of the s/s
> >>> found.Row
> 25
> >>> found.Column
> 1 (ie: Column 'A')
> Help please.
> 2) Howto return a tuple of all values in only column 'B' and then use
> python to search the tuple for the value I'm after. It strikes me that
> this is probably a faster approach?
 
I reccomend that you get a copy of "Python Programming on Win32" if you
don't already have it. It contains a very useful helper class for accessing
Excel. I have a version with a couple of additional methods - I'll send it
to you directly.
 
> BTW: Thanks Mark Hammond for a v.cool and useful tool.
 
Seconded.

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




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




From stephen at xemacs.org  Thu Apr  4 01:12:32 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 04 Apr 2002 15:12:32 +0900
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com> <3CAB0015.6563C468@python.org> <3CAB5757.6EFD1303@alcyone.com>
Message-ID: <87n0wkvvov.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Erik" == Erik Max Francis <max at alcyone.com> writes:

    Erik> Guido van Rossum wrote:

    >> That's an interesting response, because it highlights what I
    >> think is the main misunderstanding about the PEP.  You should
    >> think of it a very *small* change!

I do now ...

    Erik> The something-vs.-nothing objection seems to be missing the
    Erik> point.  The addition of bool does not affect what truth and
    Erik> falsity mean in Python, it just adds another explicit set of
    Erik> values for use when that is all you mean to talk about.

... and of course Guido has said that all along, but ...

    Erik> I'm really surprised at the volume of objections to a change
    Erik> that can only help make code clearer and will have very few
    Erik> compatibility ramifications.

... (1) if you are not a language implementer (and even if one's done
a bit of it oneself), adding a Whole New Builtin Type looks like a
BFD, not a "very *small* change", which makes (2) the many attempts to
hijack the PEP into the opening wedge of a full boolixing up<wink> of
Python look very plausible.  If there were a serious possibility of
boolixation, then "something-vs.-nothing" _would_ be the central issue.
(It wasn't just the "education faction" that missed the point.)

FWIW, I think that's why I didn't "get" it until a couple hours ago.
I don't know if it accounts for anyone else's thinking, of course.

I still think both Alex's and Laura's posts were terrific, and
important contributions to my Pythonic understanding.  And they'll
provide lots of ammunition in the never-ending battle against the
demon "if x is True: ..."!

Can-introspecting-confusion-ever-help-clarify-ing-ly y'rs,

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From ale_xp at yahoo.com  Sun Apr  7 05:14:06 2002
From: ale_xp at yahoo.com (alex)
Date: 7 Apr 2002 01:14:06 -0800
Subject: help:problem using the RawPen class/turtle module.
References: <8374ee10.0204060303.3333260a@posting.google.com> <slrnaauvbv.3v.missive@whave.frontiernet.net>
Message-ID: <8374ee10.0204070114.56260848@posting.google.com>

missive at frontiernet.net (Lee Harr) wrote in message news:<slrnaauvbv.3v.missive at whave.frontiernet.net>...
> On 6 Apr 2002 03:03:14 -0800, alex <ale_xp at yahoo.com> wrote:
> > Hi all, I'm a Python beginner, and I'm trying to use
> > the module "turtle" to do some turtle graphics.
> > But I have a problem when I try to use the turtle
> > class RawPen in order to use it with a Canvas. 
> > In particular when I try to run the following code: ...
> > 
> > ...
> > root = Tkinter.Tk()
> > c       =   Tkinter.Canvas(root,heigh=200,width=300)
>                                   ^^^^^
>             just a transcription typo, right?

yes a bad copy&paste it was 'height' sorry.

> 
> > c.pack()
> > mypen   =   turtle.RawPen(c)
> > ... 
> > 
> 
> I tried this at the interactive python prompt:
> 
> import turtle
> import Tk
> root=Tkinter.Tk()
> c=Tkinter.Canvas(root)
> c.pack(expand=1)
> mypen=turtle.RawPen(c)
> 
> 
> and got no errors.
> 
> This was with python2.1 configured with Tkinter
> 
> So it may be a configuration (rather than a
> usage or syntax) problem.

Hi, 
thank you for your reply,

I tryied to type your code on my python command line
and unfortunatly I have the same message.
So well yes I think you are right, something's wrong
in my configuration.

It seems that it comes from the turtle.py module, I'll
try to have a look at it(even if I'm not an expert of
python modules), 
it seems strange to me that it doesn't work because 
it is in the standard distribution and of course I 
haven't changed it.

I made a copy & paste of my little test.


Python 2.2 (#1, Mar  9 2002, 18:21:48) 
[GCC 2.95.2 19991024 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import turtle
>>> import Tkinter
>>> root=Tkinter.Tk()
>>> c=Tkinter.Canvas(root)
>>> c.pack(expand=1)
>>> mypen=turtle.RawPen(c)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 16, in __init__
    self.reset()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 42, in reset
    self.clear()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 53, in clear
    self._draw_turtle()
  File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 262, in _draw_turtle
    self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
AttributeError: 'NoneType' object has no attribute 'create_line'


alex.


From weeks at vitus.scs.agilent.com  Mon Apr  8 15:43:33 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Mon, 8 Apr 2002 19:43:33 +0000 (UTC)
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com>
Message-ID: <1018304015.46066@cswreg.cos.agilent.com>

Jeremy Hylton (jeremy at alum.mit.edu) wrote:
: > Anyway, it increased visibility enough so that it is no longer possible to
: > call a function and pass it the scope as two dictionaries, which I had 
: > found handy on occasion.

: I'm not sure I follow the details here, but I don't see how the change
: in the scope rules would affect a function unless it specifically
: exploited nested scopes.  If you wrote a function that only uses
: locals and globals, can you not call it just the way you always did?

Yes, the function continues to work.  But the documentation changes.
Originally, calling the function with the arguments globals() and locals()
passed to the function all the variable bindings visible at the point where
the function was called.  That is a tidy concept, and I'm sorry to lose it.

Regards,
Greg


From roy at panix.com  Wed Apr 10 20:47:07 2002
From: roy at panix.com (Roy Smith)
Date: Wed, 10 Apr 2002 20:47:07 -0400
Subject: Help: Python for a big commercial project?
References: <F194tIpqqJgSt51NDa200002509@hotmail.com> <mailman.1018464082.15807.python-list@python.org>
Message-ID: <roy-EB08BB.20470710042002@news1.panix.com>

Cliff Wells <logiplexsoftware at earthlink.net> wrote:
> Absolutely.  I would not hesitate to use Python for *any* project.

I think one of the keys to a successful large-scale python project is 
understanding its strengths and weaknesses.  Take advantage of the former, 
and don't try to brute-force your way past the latter.

For example, python is not very good at low-level character manipulation.  
If you're parsing text, for example, and are used to typical C-style 
tokenizer loop:

   while (c = getchar()) {
      switch (c):
         blah, blah

and try to translate that directly into python, you'll probably end up with 
something which is slow as a dog (DAMHIKT).  The key is to figure out which 
little bits of your code would benefit most from re-writing them as a C 
module, and doing that.  You may discover that writing 2% of your code in C 
will give you something that's 10 times faster than the pure python 
implementation.  Without having looked at the code, I'm guessing that the 
pickle/cPickle modules fall into this category.

The nice thing is that you can start out by writing everything in python, 
then, once you've got it all working, profile the thing and figure out 
what's the bottleneck.  You can then do a selective C rewite as time 
permits and performance demands.


From emile at fenx.com  Wed Apr  3 07:40:27 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 03 Apr 2002 12:40:27 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com>
Message-ID: <%KCq8.184591$uA5.150637@rwcrnsc51.ops.asp.att.net>

John Roth
> Maybe my post in response didn't propagate? As I said in that
> post, the root of her problem is a poorly chosen name for the
> function. If she paid attention to constructing the name properly,
> it would come out as something like "numberOfLeapDays"
> rather than "isLeapYear", and there would be **no**
> instructional issue.
>

Your post propagated here -- I just assumed you referred to Guido's
original reference to continuing to allow things like isLeapYear and not
Laura's reuse of it.  From the PEP:

Minor additional issues:

    4) Should we strive to eliminate non-Boolean operations on bools
       in the future, through suitable warnings, so that for example
       True+1 would eventually (in Python 3000) be illegal.
       Personally, I think we shouldn't; 28+isleap(y) seems totally
       reasonable to me.



--

Emile van Sebille
emile at fenx.com

---------



From davies at jkevin.freeserve.co.uk  Tue Apr 23 11:20:09 2002
From: davies at jkevin.freeserve.co.uk (kevin)
Date: 23 Apr 2002 08:20:09 -0700
Subject: Python & autoftp
Message-ID: <53b7d218.0204230720.686e473f@posting.google.com>

Hi there

I'm pretty new to Python and I would like to develop a script in
Python to automatically FTP. I have already used bash and expect and
now would like to replace bash entirely with Python.

Is there any merit in replacing bash with Python??
Can anyone point me in the right direction.


Thanks


Kevin


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 02:55:32 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 23:55:32 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cba79ab@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3lmbpmqij.fsf at mira.informatik.hu-berlin.de...

| So why should hexadecimal be different? It is just the
| base-16 representation of the number, nothing more.

1) Actually, if the long an int types are properly unified,
the long will be converted to an int as soon as it can
be properly represented by an int.

2) Secondly, when we are printing integers in the hex format,
we are not interested in the sign-- we are looking at the
bit pattern.

3) With unification, the long and the int types should not
print differently when calling the 'hex()' method.  (The
'int' version of this method NEVER has, and shouldn't
print a minus sign, and _I_ don't think the long version
of 'hex()' should either.  This _is_ what needs to happen
to have proper unification.

4) We are not using 1.5.2 anymore.  We are on 2.2.x now.
You need to imprint on your brain that there now should be
_NO_ visible difference (to the programmer) between the long
and the int type.  In fact, the int and long representations
are now just an implementation detail, and should be hidden,
and transparent to the programmer.

5) These two types (int and long) _CAN_ be unified for
_ALL_ operators and methods-- I know they can, because the
language known as "Ruby" does a very good job of it.  In
Ruby, there is simply no difference (at least visibly) to
the programmer _at_all_.  The 2 types are automatically
converted back and forth without the knowledge or consent
of the programmer.  If Ruby does it, so can Python.


Martin, why do you resist the unification of these two
types for _ALL_ operations and methods?  Don't you think
Python would be better if there simply was no visible,
operational or programmatical difference between an int
and a long?





From matthewm at ihug.co.nz  Thu Apr 18 03:22:45 2002
From: matthewm at ihug.co.nz (Matthew)
Date: Thu, 18 Apr 2002 19:22:45 +1200
Subject: Hostname Resolution not working in Explorer
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net> <w9sv8.17352$L1.1367388@newsread2.prod.itd.earthlink.net>
Message-ID: <3CBE7444.8D2D96BC@ihug.co.nz>

Try 127.0.0.1 in lmhosts file rather than hosts file. iirc the lmhosts file maps
netbios computer names to ip addresses which i think is what you want. matthew.

Gary Spivey wrote:

> Looks like I screwed up my first post ...
>
> I have windows XP Home, and IE6. I have a zope server running on my box. The
> server is running fine, and I can access if it I access
>
> http://127.0.0.1:8080/manage
>
> However, if I use
>
> http://localhost:8080/manage
>
> it doesn't work. Same holds true if I use xx.xx.xx.xx:8080/manage and
> hostname:8080/manage.
> The names don't work, the numbers do. And they all work just fine when
> pinging from a command window. So, something between me, internet explorer,
> and the name resolution isn't working.
>
> "Gary Spivey" <spivey at ieee.org> wrote in message
> news:edrv8.17246$L1.1349660 at newsread2.prod.itd.earthlink.net...
> >
> > I have windows XP Home, and IE6. I have a zope server running on my box.
> > The server is running fine, and I can access if it I access
> >
> > http://127.0.0.1:8080/mange
> >
> > However, if I use
> >
> > http://localhost/mange
> >
> > the page isn't found. The same holds true if I use my hostname rather than
> > localhost - also, the actual IP address works.
> >
> > I have made sure that the name entries exist in the
> > window/system32/drivers/etc/HOSTS file, and I can ping both localhost and
> my
> > hostname from a command window (with the correct address resolution).
> >
> > So - why can I not get the resolution to work from within Internet
> Explorer?
> >
> > Cheers
> > Gary
> >
> >
> >



From mwh at python.net  Thu Apr 18 06:04:18 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 10:04:18 GMT
Subject: Readline/command history w/ interactive prompt
References: <mailman.1019061574.15837.python-list@python.org> <a9kcml$69k$1@news.wrc.xerox.com> <mailman.1019117616.20456.python-list@python.org>
Message-ID: <lk662pmi8v.fsf@pc150.maths.bris.ac.uk>

Andy Gimblett <gimbo at ftech.net> writes:

> On Wed, Apr 17, 2002 at 05:51:17PM +0000, Mark Jackson wrote:
> 
> > Try hitting control-r and typing a short string that last appeared
> > in the line desired ("my" or even "m" would work in this case).
> 
> _Nice_!  Thanks for that.  One question though: where's this feature
> documented?  :-)

"man bash", "man readline".  Not too hard...

Cheers,
M.

-- 
  A.D. 1517: Martin Luther nails his 95 Theses to the church door and
             is promptly moderated down to (-1, Flamebait).
        -- http://slashdot.org/comments.pl?sid=01/02/09/1815221&cid=52
                                        (although I've seen it before)


From ufechner at csi.com  Sun Apr  7 06:50:02 2002
From: ufechner at csi.com (Uwe Fechner)
Date: Sun, 07 Apr 2002 12:50:02 +0200
Subject: Compatibilty table between BOA, wxPython, Python?
References: <11e94203.0204062147.5d2281b9@posting.google.com>
Message-ID: <3CB0245A.3010808@csi.com>

Jose Rivera wrote:


> I want to keep using python, and use the last version available to use
> Python, wxPython, and BOA.
> 
> Thanks...

Hello,

if you use python2.2, you should try the newest version of boa (1.1a) 
from the sourceforge cvs.

It works fine for me (on win2k).

Regards:

Uwe Fechner



From boud at valdyas.org  Thu Apr  4 03:38:41 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 04 Apr 2002 07:38:41 -0100
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl> <slrnaak6oe.s4j.mlh@vier.idi.ntnu.no> <3caabd6d$0$45613$e4fe514c@dreader3.news.xs4all.nl> <slrnaanebn.1jcm.kamikaze@kuoi.asui.uidaho.edu>
Message-ID: <3cabe8e6$0$32306$e4fe514c@dreader4.news.xs4all.nl>

Mark 'Kamikaze' Hughes wrote:

> Wed, 03 Apr 2002 10:24:59 -0100, Boudewijn Rempt <boud at valdyas.org> spake:
>>>   http://www.opendocspublishing.com/pyqt/
>> I've forwarded this to Joshua -- I haven't tried buying one, of course
>> :-).
> 
>   It would also be very nice if the opendoc version had a nice single
> download archive - I had to fiddle with wget and do a lot of sed work to
> fix the URLs so I could read it on my local HD and not hit the site
> every time I turn the page.
> 

Well, I don't think Joshua will do this -- but you could ask him, he's
the publisher.

>   I'd really MUCH rather be able to buy just a CDROM with the book in
> clean HTML (go ahead and take a look at the generated HTML...  Evil!)
> with all the illustrations (which aren't in the opendoc version), but
> failing that, I'll probably pick up the dead-tree edition.
> 

The html is just what docbook produces, so it can't be _that_ evil. The
lack of illustrations isn't apparently intentional, as the captions are
present :-).

>   So far it seems to be an excellent book - just last week I started
> teaching myself PyQt, so it came out at the right time.  I've skipped
> over all the BlackAdder stuff, but I may try that out next.
> 

I'd say, give Designer or the Designer module of BlackAdder a try anyway.
It's amazingly comfortable, even for designing complex widgets that should
be embedded in your main windows.

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


From shalehperry at attbi.com  Tue Apr 23 15:42:02 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 23 Apr 2002 12:42:02 -0700 (PDT)
Subject: dictionary method question
In-Reply-To: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com>
Message-ID: <XFMail.20020423124202.shalehperry@attbi.com>

On 23-Apr-2002 dsavitsk wrote:
> for some dictionary d is there any relationship between the returns from
> d.keys() and d.values()?  that is, to get them in some arbitrary but
> matching order, do i need to loop over d.items() or can i count on a
> similarity of the above? in small tests this seems to be the case.
> 

I would only trust d.items() but it makes sense that keys and values should be
in the same order.




From kybic at ieee.org  Wed Apr  3 11:10:26 2002
From: kybic at ieee.org (Jan Kybic)
Date: 03 Apr 2002 18:10:26 +0200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at> <lbh8z84nake.fsf@xihu.inria.fr> <15531.7864.698585.485436@12-248-41-177.client.attbi.com> <lbh7knolreu.fsf@xihu.inria.fr> <15531.9709.924953.953487@12-248-41-177.client.attbi.com>
Message-ID: <lbhn0wkkbkd.fsf@xihu.inria.fr>

The following message is a courtesy copy of an article
that has been posted to comp.lang.python as well.

> Not to belabor the point, but you can simply use Python's control
> structures:
> 
>     if today == "Monday":
>         expr = "a[1:-1,1:-1] =  (b[1:-1,1:-1] + b[2:,1:-1] + b[:-2,1:-1]" \
>                               "+ b[1:-1,2:] + b[1:-1,:-2]) / 5."
>         weave.blitz(expr)

Yes, but you what if you want the structures 'inside' the loop? Example:

for i in xrange(1,a.shape[0]):
        if a[i,i]>30: a[i,i]+=a[i,i+1]
        else:          a[i,0]*=2.0

Not a particularly useful example, but one which could be much
accelerated by translating it (automatically) into C.
        
J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/




From cppdan at dansmart.com  Sat Apr  6 17:25:40 2002
From: cppdan at dansmart.com (Dan Smart)
Date: Sat, 06 Apr 2002 22:25:40 GMT
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <mailman.1018120753.6552.python-list@python.org>
Message-ID: <Xns91E8B14D34FD8Me@167.206.112.134>

Laura Creighton <lac at strakt.com> wrote in 
news:mailman.1018120753.6552.python-list at python.org:

> Thank you very much for the reply.
> I just want to correct something you ascribed to me.
> 
>> I guess we'll have to agree to disagree here: I don't think that bools
>> are bad, either Python's variety or any other language's kind.  I've
>> seen your arguments, and they just don't convince me.  So be it. 
> 
> I think there is a time and a place for bools --  Describing things
> that have two states that can never, ever, have 3 or more states.  
> These things are vanishingly close to non-existent, unless you are
> a mathematician in which case you may be playing with these sorts
> of things all the time.  What I am trying to do is save people
> like Mark McEahern from his isValid function, which returns whether
> a username is valid or not.  On the day his boss demands a third
> value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
> he will thank me.
And on the day that a maintenance programmer comes to try and understand 
the obscene mess you have recommended, he will curse you. There are two 
correct solutions to the hypothetical problem you posed, neither of which 
involve isValid() (an intergatory function that obviously has a YES/NO 
answer) returning anything other than YES/NO.

The first is to add a second function isAllowedToLogin (another 
interogatory function that obviously has a YES/NO answer).

The second is to replace the isValid() function with a function like 
validateUser() which either returns OK or a failure reason. In fact I'd 
probably code it to return OK, or raise an exception, YMMV.

Both these alternative will involve refactoring the code to account for 
this change, but that is a GOOD THING, and was anyway required by your 
change. The difference being that both the above solutions accurately 
describe the requirements, where your solution obfuscates those 
requirements.

Another example you gave, that of isinstance, shows equally flawed 
reasoning. Either an object is an instance of a given class, or it isn't. 
It is a fact that instances of a class derived from the given class are 
also instances of the given class. I understand that there are times when 
it might be useful to differentiate between objects that are both instances 
of a given class and instances of a class that is derived from that class, 
and objects that are only instances of the given class, but this is a 
different question. Overriding the former question so that it also answers 
the second question *decreases* the clarity of code that has to deal with 
these issues; by (a) obscuring the intent of the question, readers now have 
to understand that a single question serves two purposes, and intuit which 
question was really being asked; and (b) forcing code that is only 
interested in the answer to the former question (which is the most likely 
case) to account for the fact that they also getting an answer to the 
second. As an aside there is another case that you failed to think of, you 
can modify the methods and variables of an instance of a class. Note that 
in this case isinstance is still a valid question which still only has a 
YES/NO answer, but it may be useful to have a third query to differentiate 
this case.

Your suggestion that very few questions only have YES/NO (or TRUE/FALSE) 
answers surpises me, I can think of several immediately: isValid(), 
isNonZero(), isModified(), isInstance(), contains(). In general if you 
discover that such a question turns out to be Yes/No/Maybe, it indicates 
that you either originally provided the wrong question, or that 
circumstances have changed such that a new question (or set thereof) would 
be more appropiate. In either case refactoring the code to reflect this 
change in reality/understanding is the correct thing to do. Changing the 
function to return extra state is hacking, and while there's a time and a 
place for hacking, suggesting that this is good design is almost as 
outrageous as suggesting that the function should have originally been 
written to make such hacking easier.

> 
> That is what I mean by teaching people not to use bools -- teaching
> when not to use them. This is very hard to teach.  I know it because
> I have been doing it for years and years.   I even know _why_ it is
> hard to teach, but that doesn't make it any easier.
> 
> Laura
> 
Perhaps you would be better off teaching them sound design, that code 
should accurately and concisely express both the problem it is trying to 
solve, and that solution; and that when the problem, or the authors 
understanding of the problem, changes, the code should be re-written to 
reflect that change.

Just in case you think I've missed something, I completely agree that:
    	Using bools to represent other bi-valued states (like color/side in 
chess) is terrible design.
    	Using the integer value of a bool (as in the isLeapYear example) is 
terrible design (although not because the number of days may not 0 or 1, 
but because such code doesn't say what it means).
    	Using bools as the return value of functions that are not directly 
enquiring as to whether a given assertion holds is almost certainly a 
mistake. Note that functions that are enquiring as to whether a given 
assertion holds should have an interogatory name that reflects that.
    	In particular, using a boolean return to indicate success/failure is 
almost certainly a mistake, as success/failure is not the same as 
true/false, and knowing what failed is usually important. This is probably 
the most difficult case, as there are rare cases where you are only 
interested in the two states, and there is a temptation to believe that in 
such cases there is an implicit question "wasThisSuccessful" which does 
have a Yes/No answer, this temptation should be resisted. It may be 
appropiate in such cases to provide a function like "succeeded()" which 
takes the return value of the function in question, and returns true/false.

Dan "It doesn't seem that hard to say when not to use bools, although of 
course if you include the cases where bools should be used, it gets much 
harder" Smart
-- 
Dan Smart. C++ Programming and Mentoring.
cppdan at dansmart.com
ADDvantaged


From martin at v.loewis.de  Mon Apr 15 02:29:10 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:29:10 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu>
Message-ID: <m38z7pmq4p.fsf@mira.informatik.hu-berlin.de>

David Eppstein <eppstein at ics.uci.edu> writes:

> >>> 0x40000000<<1
> -2147483648
> 
> This is inconsistent with int-long unification.  In fact, I think it is 
> a bug.  The answer should, of course, be 2147483648L, just as you would 
> get by multiplying by 2 instead of shifting.

Indeed, if you write

>>> 0x40000000L<<1
2147483648L

it works the way you expect it to work. Please see

http://www.python.org/peps/pep-0237.html

Python has currently completed phase A of the PEP, meaning that
operations that previously raised an OverflowError now return a long;
nothing else has changed.

Notice that your example is the historic meaning of shift; in phase
B.1, it will change to the arithmetic meaning (multiplication with
2). Since this is a semantic change, you'll get a warning if you make
use of a short int shift that loses a bit.

Regards,
Martin



From ponderor at lycos.com  Tue Apr  2 18:13:49 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 2 Apr 2002 17:13:49 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr  2)
Message-ID: <195407FD2A84CA3D.DC4247817634A604.1B161784FA0BDD0D@lp.airnews.net>

QOTW:  "Is this one of those deals where you tickle a file with that
'mouse' thing, and then a dialog box from an irrelevant program pops
up and steals the focus before you can finish what you started?  I
love stuff like that." -Tim Peters

"[T]he distinctive difference between PCs and other consumer electronics
is that only PCs are allowed to fail in routine use." -Peter Coffee


    Does Python deserve a boolean type?  Powerful minds analyze PEP 285.
	http://groups.google.com/groups?th=3d8f05b445cdbf9b

    Aahz echoes timbot's entreaties to esteem the benefits of Queue.Queue.
	http://groups.google.com/groups?th=1e465bd53977862f

    April Fool's roundup:
    ZWiki Author Disillusioned...Rubymortis sets in. - see "Big news"
        http://zwiki.org/GeneralDiscussion
        http://zwiki.org/GeneralDiscussion200204 (future archive)
    ZopeNewbies: Microsoft takes over Zope
        http://www.zopenewbies.net/1017662225/index_html
    Sestina deemed best format for switch statement implementation
	http://www.acmenet.net/~jilhan/sesform.html
    Docstrings requirement feature in 2.2.1 avoidable by use of "<wink>"  
    SOAP to be transported over IRC; finger/whois replaces WSDL.

    The *real* Tkinter documentation URL:
	http://groups.google.com/groups?th=83aa5100ac5e4d76&seekm=slrnaa7ic5.ehu.philh%40comuno.freeserve.co.uk#link20

    Bob McMillan elicits "Python's Creator Guido van Rossum's Take on Open
    Source's Favorite Serpent"
        http://www.linux-mag.com/2001-12/rossum_01.html

    Plotting Pointers:
        http://mail.python.org/pipermail/python-list/2002-March/094177.html

    Nicolas Chauvat is looking for interested people to form a Logic-SIG 
    or CLP-SIG.
        http://mail.python.org/pipermail/python-list/2002-March/094019.html

    Reading binary as text is not the easiest way..
        http://mail.python.org/pipermail/python-list/2002-March/094424.html
    
    Coding can be pure poetry...
        http://groups.google.com/groups?th=83aa5100ac5e4d76&seekm=FC718F7B3AF20F62.09ECC4E1E82F69E7.7E08A6AB55BAC1B9%40lp.airnews.net#link6

    Measuring memory allocation
        http://mail.python.org/pipermail/python-list/2002-March/094191.html

    Messaging or Threading..and safely
        http://mail.python.org/pipermail/python-list/2002-March/094235.html

    PEP 262: Database of Installed Python Packages Moving into Interface
    Development and on to Integration

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

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

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

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

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

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

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

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


-- The Python-URL! Team--

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

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


From gregLANDRUM at mindspring.com  Mon Apr 29 17:47:37 2002
From: gregLANDRUM at mindspring.com (Greg Landrum)
Date: Mon, 29 Apr 2002 14:47:37 -0700
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <aakf1c$acb$1@slb2.atl.mindspring.net>

On the off chance that this proposal might actually be considered by the
powers-that-be:

-999

Changes that:
a) break lots and lots of code
b) render most (all?) Python books inaccurate
c) eliminate analogies to other common languages (I immediately knew what
"map" and "filter" would do when I was learning python).
d) break lots and lots of code
are *bad*.

I'm just a dumb-ass programmer, but I don't even begin to understand the
argument behind doing this.

-greg





From maxm at mxm.dk  Fri Apr 12 08:00:36 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 12 Apr 2002 14:00:36 +0200
Subject: Looking for doh!.. Answer found
References: <3CB6BED8.2000502@mxm.dk>
Message-ID: <3CB6CC64.8080809@mxm.dk>

Answer found:

The problem is that __getitem__ doesn't check length.  Instead, it
terminates on an IndexError exception.

class Lst:
     def __getitem__(self, idx):
         if idx>=4:
             raise IndexError
         else:
             return 42

lst = Lst()
for i in lst:
     print i

42
42
42
42


Max M wrote:
> Hmm ... I am trying to create a simple sequence type that I cant iterate 
> over like (Python 2.1)::
> 
> lst = Lst()
> for i in lst:
>     print i
> 
> And I thought I could do it like:
> 
> class Lst:
> 
>     def __len__(self):
>         return 4
> 
>     def __getitem__(self, idx):
>         return 42
> 
> But I have to write it like below to avoid an endless loop..
> 
> lst = Lst()
> for i in range(len(lst)):
>     print i,
> 
>  >>>42 42 42 42
> 
> I am shure it's so simple tha I will go Doh! .. But perhaps I haven't 
> had enough cofee yet. 'Cause I cannot seem to figure it out.
> 
> regards Max M
> 



From pyth at devel.trillke.net  Mon Apr 22 11:58:58 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 22 Apr 2002 17:58:58 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <304d20df.0204220633.50bf23bf@posting.google.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org> <304d20df.0204220633.50bf23bf@posting.google.com>
Message-ID: <20020422155858.GD27395@devel.trillke>

On Mon, Apr 22, 2002 at 07:33:15AM -0700, Allan Crooks wrote:
> > But if you are truly set on using incompatible constructs  
> > *everywhere* for the fun of it, you are going to have a problem
> > with code duplication :-)
> 
> I wouldn't actually mind having code which would do the following:
> 
> --------
> 
> try:
>     import warning
> except ImportError:
>     warning = None
> 
> ....
> 
> if warning is not None:
>     <do something with warning module>
> 
> --------
> 
> The main problem is when I want to use 'yield', there's no real way of
> compensating for a language change... I suppose it is my fault for
> trying too much. :) The more I think about it, the more maintaining
> concurrent versions seems appealing... at least there is no 'hacking'
> involved...

You probably refer to my posting :-)
If so, let me note that "execing a string" is what python does itself. 
e.g. when you import a module the interpreter is execing a module-string or 
code-object. It's some weird hack but the regular python way.

As long as you use your pre2.2/post2.2 "function" with
"for i in function()"  or "filter(foo,function())"
you can easily do the following:

minversion("2.2",'''

from __future__ import generators

def function():   # full generator version
	...
	yield xyz

''','''

def function():   # fully-constructed list version
	l=[]
	for item in soandso:
		...
		l.append(item)
	return l

''')

This is precisely what i want to express.

Your mileage may vary, of course.

  holger




From Matthias.Hanitzsch at dynamic.de  Wed Apr 10 02:22:42 2002
From: Matthias.Hanitzsch at dynamic.de (Matthias Hanitzsch)
Date: Wed, 10 Apr 2002 08:22:42 +0200
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <3CB2FDF3.BBBD99C0@dynamic.de> <xm5vgb06g7f.fsf@decius.daimi.au.dk>
Message-ID: <3CB3DA32.75529CF1@dynamic.de>


Torben Lauritzen wrote:

>
> As I see it, num is not incremented, before it is returned. So, even
> if it had been a 'call by reference', x would not increase.

Correct.



From h_schneider at marketmix.com  Fri Apr 12 06:28:32 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 12 Apr 2002 12:28:32 +0200
Subject: Problem with PIL and compression format
References: <mailman.1018602385.24023.python-list@python.org>
Message-ID: <a96cib$bkf$04$1@news.t-online.com>

You can ship around this problem, by doing all imaging e.g. with .bmp and
then calling an external bmp to sff (structured fax file) converter.

All the best,
Harald
"Maric MICHAUD" <maric.michaud at cirec.com> schrieb im Newsbeitrag
news:mailman.1018602385.24023.python-list at python.org...
I post my request again, if someone can help.

-----Message d'origine-----
De : Maric MICHAUD
Envoy? : mercredi 10 avril 2002 15:45
? : Python-List (E-mail)
Objet : Problem with PIL


I tried ro include tif image in a pdf document with reportlab and Python
2.0, PIL 1.1.3.
This work pretty well with uncompressed and 'Pack Bits" compressed TIFF
files, but not with "CCITT group 4" that I need.
Anyone has a suggestion for this ?


--
http://mail.python.org/mailman/listinfo/python-list






From jb at cascade-sys.com  Tue Apr 30 14:06:59 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 11:06:59 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2><3CCD9649.98AED86F@cascade-sys.com><20020429130136.5b0c34f7.logiplexsoftware@earthlink.net><3CCDB559.A1BF0E47@cascade-sys.com> <mailman.1020116996.21109.python-list@python.org> <fZwz8.13890$xy4.10493@atlpnn01.usenetserver.com>
Message-ID: <3CCEDD43.A56AFEBF@cascade-sys.com>

Steve Holden wrote:

> "Cliff Wells" ?logiplexsoftware at earthlink.net? wrote in message
> news:mailman.1020116996.21109.python-list at python.org...
> ? On Mon, 29 Apr 2002 14:04:25 -0700
> ? James J. Besemer wrote:
> ?
> ? ?
> ? ? Cliff Wells wrote:
> ? ?
> ? ? ? Well that's a stupid statement - *your* head is in the clouds.
> ? ?
> ? ? Oh yeah?  Fuck you.
> ?
> ? Heh.
> ?
>
> This little exchange is about as perfect a demonstration of Cliff's point as
> could be made. Now, if anyone but me is still reading this thread, nobody
> knows who was joking and who wasn't in the first place. The additional loss
> of context certainly doesn't help, either.

So long as my new good buddy Cliff and I aren't complaining, what the f-
business is it of yours?  ;o)

Sincerely,

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From nobody at nobody.com  Fri Apr  5 00:35:35 2002
From: nobody at nobody.com (John)
Date: Thu, 4 Apr 2002 23:35:35 -0600
Subject: [Q] Python CGI Q
Message-ID: <8Far8.36437$tg4.432342@vixen.cso.uiuc.edu>

Hi, I'm wondering if I can write a CGI that loads/edits/saves files on
server side.. (I don't need to worry about security concerns for this
application) Any starting point or open-source?




From robin at jessikat.fsnet.co.uk  Wed Apr  3 07:31:59 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 3 Apr 2002 13:31:59 +0100
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org>
 <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
 <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
Message-ID: <DrIRduA$Yvq8Ewq+@jessikat.fsnet.co.uk>

In article <3CAAEF17.32333986 at kfunigraz.ac.at>, Siegfried Gonzi
<siegfried.gonzi at kfunigraz.ac.at> writes
.....
>Good to see that my assumption that Numeric is not slower than Matlab or
>IDL can be confirmed.
>
.....
a while ago I translated a matlab generational accounting model into
Python+Numeric. The result was faster than the DLLs generated for matlab
by a proprietary compiler. It was certainly a bit faster than the
original matlab. Incidentally I think much of the speed improvement over
the compiler form was due to the non-copying slices in Numeric (ie more
work and much more memory was used in the DLL form). I understand some
propose to change that.
-- 
Robin Becker


From bogus@does.not.exist.com  Tue Apr 30 19:32:31 2002
From: bogus@does.not.exist.com ()
Date: Tue, 30 Apr 2002 23:32:31 GMT
Subject: Newbie - Making Windows Add-in with Python?
Message-ID: <3ccf27e2.18898518@news.dial.pipex.com>

Help please!
I'm trying to build an add-in module to work with a shrink-wrap
commercial application. The API documentation supplied  with the app
shows how to build an add-in with VB or C++/ATL, in the VB example it
requires you to choose "add-in" and select the application for the
add-in. from deciphering the later documentation (I'm not familiar
with VB6) it implies that this causes the DLL Vb6 produces to be
registered in relation to the app, so that when the app starts it
autoloads the DLL.

I've had a play with the WIN32 COM extensions in PYTHON but can't see
how to duplicate the above behaviour.

Any clues please?


From shalehperry at attbi.com  Sat Apr 20 11:28:10 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 20 Apr 2002 08:28:10 -0700 (PDT)
Subject: Substitution with Hash Values
In-Reply-To: <20020420150508.GB10574@devel.trillke>
Message-ID: <XFMail.20020420082810.shalehperry@attbi.com>

> 
> But is it true that you 'never' use them somewhere else? 
> I would think that one would like to use 'r' whenever you don't 
> want the escape-sequence-string-interpretation which is not really
> restricted to regexp-strings. 
> 
> I have noticed that e.g. the docstring of the module os.py uses
> r'...'
> 

'never' as in 'so rare as to be an exception and not the rule'.




From johnroth at ameritech.net  Mon Apr 15 17:00:45 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 15 Apr 2002 17:00:45 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <a9eppf$ghv$1@panix1.panix.com> <a9f95n$jg0$1@bob.news.rcn.net>
Message-ID: <ubmftha3r74h80@news.supernews.com>

"Raymond Hettinger" <python at rcn.com> wrote in message
news:a9f95n$jg0$1 at bob.news.rcn.net...
> [Raymond Hettinger]
> > >The help and tutor maillists indicate that the use of input()
rather than
> > >raw_input() is widespread.  I think this function is so harmful and
so
> > >likely to be used that it should be deprecated hasta pronto (right
away).
> > >When really needed, eval(raw_input()) is easy enough to code and
explicit
> > >about what it does.
>
> [Aahz]
> > Write a PEP.
>
> Rules of the game:
> '''A PEP that has not been discussed on python-list at python.org and/or
> python-dev at python.org will not be accepted.  ''' -- from Pep 1

Does that mean by the PEP editor or by the BDFL? I thought the
process was: write it, send it in, discuss it, then the BDFL makes
a Pronouncement.

John Roth
>
>




From skip at pobox.com  Wed Apr 10 13:41:40 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 10 Apr 2002 12:41:40 -0500
Subject: Transferring files using sockets
In-Reply-To: <Pine.LNX.4.33.0204101034250.24093-100000@ns0.tbye.com>
References: <15540.27738.975194.101310@beluga.mojam.com>
        <Pine.LNX.4.33.0204101034250.24093-100000@ns0.tbye.com>
Message-ID: <15540.31060.637356.483031@beluga.mojam.com>

    Dave> Also, 'send' doesn't necessarily send the whole chunk of data; as
    Dave> a return value it gives you the number of bytes actually
    Dave> sent. 

Oh yeah, I forgot about that...

Skip




From neuhauser at mail.cz  Tue Apr 30 08:21:30 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 14:21:30 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <roy-E96293.07453430042002@news1.panix.com>
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com>
Message-ID: <20020430122130.GJ325@roman.mobil.cz>

> From: Roy Smith <roy at panix.com>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Tue, 30 Apr 2002 07:45:34 -0400
> 
> Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:
> > The problem with int.chr() is that a lot of people are going to get 
> > confused when they find that "42.chr()" doesn't work (although "42 .chr()" 
> > does).
> 
> What about "(42).chr"?  Wouldn't that work?  Seems a lot more natural than 
> putting a space in there (which I didn't even realize was legal, and looks 
> horrible to me).

    (42).chr does look better, and is IMO more readable.
    I didn't originally suggest this as I (having spotty knowledge of
    python) thought it would make it a tuple. But you're right, no
    comma, no tuple.

-- 
FreeBSD 4.4-STABLE
2:19PM up 1 day, 4:44, 11 users, load averages: 0.01, 0.05, 0.01




From aleax at aleax.it  Tue Apr 30 04:58:07 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 08:58:07 GMT
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <z0tz8.33311$8D3.941294@news1.tin.it>

Ralf Juengling wrote:

> when I want to make sure within a function that it can deal with
> the arguments passed, the only choice is to check it's type via
> 'isinstance()'. However, in general the type of an argument is
> not the crux but whether the argument supports a certain
> 'interface' or 'protocol'. Thus, a function 'hasinterface()'
> would appear more natural (more 'Pythonic') to me.
> 
> So, what is the proper, 'pythonic' way of argument checking?

PEP 246, but I haven't yet been able to communicate the how's
and why's of it to the BDFL.

Meanwhile, http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52291
may help.


Alex



From sfconsulting at attbi.com  Thu Apr  4 11:06:02 2002
From: sfconsulting at attbi.com (Daniel Popowich)
Date: 4 Apr 2002 08:06:02 -0800
Subject: 2.2 descriptor API
Message-ID: <410890c4.0204040806.363f5728@posting.google.com>

I have been working with the new 2.2 descriptor API and have a
question about the __get__ method.

If I have the following:

    class Foo(object):
       x = mydescriptor()

where mydescriptor is a class that defines  __get__, __set__ and
__delete__, then the following code:

    f = Foo()
    f.x = 1
    print f.x
    del f.x

behaves as you would expect and calls the __set__, __get__ and
__delete__ methods in turn.  No problems here.

Now to the meat of my question which relates to descriptor usage when
referenced as class attributes...

If I write:

   Foo.x = 1

The class attribute x is overwritten with the integer 1.  The __set__
method is not involved.  If I write:

   del Foo.x

The x attribute is removed from the class namespace.  The __delete__
method is not involved.

HOWEVER, if I write:

   Foo.x

The __get__ method is called!!!  WHY??  Why isn't the instance of my
descriptor returned as it would be with any other object?  In fact, the
canonical implementation of a __get__ method that I have seen thus far
in examples...

   def __get__(self, inst, type=None):
       if inst is None: return self
       ...

implements exactly what one would expect to happen: return the
instance of the descriptor.  But why do I have to code this?

While I'm on this subject, could someone explain the second argument?
I have not seen any code that makes use of it and I'm not sure what
it's about.

In summary:

1.  Why is the __get__ method called on a descriptor when it is
    referenced as a class attribute?

2.  What is the purpose behind the second argument; does someone have
    an example of how or when I would make use of it?


Thanks,

Daniel Popowich


From db3l at fitlinxx.com  Fri Apr 26 15:08:18 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Apr 2002 15:08:18 -0400
Subject: File sizes from MSWin...
References: <mailman.1019843557.24893.python-list@python.org>
Message-ID: <uvgaexopp.fsf@ctwd0143.fitlinxx.com>

Christopher Encapera <ChrisE at lantech.com> writes:

> When I try:
> 
> >>> os.path.getsize('x:\\exchsrvr\\mdbdata\\PRIV.EDB')
> 1475354624
> >>> 
> 
> (1,475,354,624 bytes)
> The actual size is 14,360,256,512 bytes.  When I try:

I expect you're seeing truncation due to the use of 32-bit values for
the size.  Until Python 2.2, the Windows version didn't include large
file support (not sure if you could manually build it that way or not,
but it didn't come pre-built with it).

> >>> import win32file
> >>> a = win32file.GetFileSize('X:\\exchsrvr\\mdbdata\\PRIV.EDB')
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: The object is not a PyHANDLE object

The GetFileSize() function can only be used on an open file handle
(not on a filename).  So you could do:

    source = win32file.CreateFile('X:\\exchsrvr\\mdbdata\\PRIV.EDB'),
                                  win32file.GENERIC_READ,
                                  win32file.FILE_SHARE_READ,None,
                                  win32file.OPEN_EXISTING,0,None)
    size = win32file.GetFileSize(source)
    source.Close()

The other option (closer to what os.path.getsize is doing) would be to
do a directory listing with Win32 calls (FindFirstFile - implemented
within win32api.FindFiles) and use the size from the directory entry
which can be 64-bit.

> This is the MS Exchange private information store, and ours is large (13.3
> G).  Any suggestions on how to get an accurate size?  I have ActiveState
> ActivePython 2.1 installed, and X: is a mapped drive to a WinNT machine from
> a Win2k workstation...
> Thanks in advance!

If you upgrade to 2.2 you should be fine.  That includes large file
(64-bit) support under Windows by default, and the os.path.getsize()
call will return a long integer rather than an integer.  The bonus is
that it's then also portable across systems rather than limited to
Windows.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From simon at piranhaflt.freeserve.co.uk  Tue Apr  9 11:20:32 2002
From: simon at piranhaflt.freeserve.co.uk (¨)
Date: Tue, 9 Apr 2002 16:20:32 +0100
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au>
Message-ID: <a8v0mo$j50$1@newsg1.svr.pol.co.uk>

Matthew Baulch <matt at greenroom.com.au> wrote in message
news:3CB3039C.3040108 at greenroom.com.au...
> I think the code you are looking for could be:
>
> x = 2
>
> def wrong(num):
>    return num + 1
>
> x = wrong(x)
> print x
>
> x = wrong(x)
> print x
>

I'm a newbie myself, so this is probably wrong, but wouldn't this be *even*
better?

x = 2
def wrong(num):
    x = x+1
    print x

wrong(x)
wrong(x)




From sheershion at mailexpire.com  Tue Apr 30 16:18:22 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Tue, 30 Apr 2002 20:18:22 -0000
Subject: wxHtml - newbie
References: <aajv7m$3hl$1@wanadoo.fr>
Message-ID: <Xns9200E1EF9C8D3rcamesz@amesz.demon.nl>

Emmanuelgagniere wrote:

> Do you know where can I find an example of a simple application
> using wxHtmlWindow?


A wxHtmlWindow can be used like any other window or control. Use the
SetPage() member function to set or change its html-text, that's all
there is to it in its most basic form. (The way you worded your
question, that's about the best I can do.) 

For more elaborate examples: the demo has loads of very
cut-and-pastable tidbits. The wxWindows manual will get you more
specific information. (When I'm doing some wxPython programming, I
have that manual opened somewhere in the corner of my screen more
often than not.)


Robert Amesz


From jon+usenet at unequivocal.co.uk  Fri Apr  5 08:45:16 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Fri, 05 Apr 2002 13:45:16 -0000
Subject: cgi.py performance
References: <a8ght4$o8o$1@web1.cup.hp.com> <slrnaaok15.4ei.jon+usenet@snowy.squish.net> <a8i60e$csg$1@web1.cup.hp.com>
Message-ID: <slrnaaraja.7u2.jon+usenet@snowy.squish.net>

In article <a8i60e$csg$1 at web1.cup.hp.com>, jimh wrote:
>> You could try jon.cgi from jonpy, it might be more efficient.
>> http://jonpy.sourceforge.net/
> 
> Hey, that looks pretty cool.  I do have one question before I try it out: in
> mime.py, Entity.__init()__, which splits up the data on the boundaries,
> loads the entire file at once, if I am not mistaken ("msg = fp.read()").
> What will happen if I am transferring a very large file, 50-100MB?

In theory: your computer uses as much RAM as it has, then starts using
swap space instead. So it's like using a temporary file except it uses
as much RAM as possible first, which makes it more efficient.

That's the idea anyway. I haven't tested it with huge files because
I've never had the need to use even reasonably large files. I'd be
very interested in your experiences if you try it.


From tim.one at comcast.net  Fri Apr  5 17:59:07 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 05 Apr 2002 17:59:07 -0500
Subject: compare function in sort function of lists
In-Reply-To: <0C7CA8D8DF75494EB09AB6016990107F011B1982@NOXMAIL.noetixad.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEFJOLAA.tim.one@comcast.net>

[Gabe Newcomb]
> 	I'm trying to sort a list numerically, e.g.,  so that 11 comes
> AFTER 2.

I suspect you're sorting a list of strings, not numbers (in Python before
2.3, these will display 0 and 1 respectively):

>>> '11' > '2'
False
>>> 11 > 2
True
>>>

> The documentation says I need to pass sort() a compare
> function...but I can't find an example of this.

You don't *need* to pass a compare function, and usually shouldn't.

>>> x = ['2', '11']
>>> x.sort()
>>> x
['11', '2']
>>> y = map(int, x)  # convert the strings to integers
>>> y
[11, 2]
>>> y.sort()
>>> y
[2, 11]
>>>





From sheershion at mailexpire.com  Wed Apr 24 13:57:44 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Wed, 24 Apr 2002 17:57:44 -0000
Subject: atoi
References: <mailman.1019649779.5458.python-list@python.org>
Message-ID: <Xns91FACA96A5899rcamesz@amesz.demon.nl>

Mariappan, MaharajanX wrote:

> is there any python buildin method whihc will convert strings into
> integer, like atoi in c lib.?
> 
> "234" should be converted into interger value 234

>>> int("234")
234


Robert Amesz


From boris at cantal.lip6.fr  Wed Apr 10 11:29:41 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Wed, 10 Apr 2002 17:29:41 +0200
Subject: Strange behaviour of CO_NEWLOCALS and function
Message-ID: <pan.2002.04.10.17.29.41.396116.31396@cantal.lip6.fr>

Here is a little example, where I create a new codeobject from one of a
function by just removing the CO_NEWLOCALS flags :

>>> import new
>>> CO_NEWLOCALS = 0x2
>>> 
>>> def g():
...     print 'locals in g :',locals()
...     if x: t=2
... 
>>> gcode = g.func_code
>>> 
>>> newcode = new.code(
...     gcode.co_argcount
... ,   gcode.co_nlocals
... ,   gcode.co_stacksize
... ,   gcode.co_flags & ~CO_NEWLOCALS
... ,   gcode.co_code
... ,   gcode.co_consts
... ,   gcode.co_names
... ,   gcode.co_varnames
... ,   gcode.co_filename
... ,   gcode.co_name
... ,   gcode.co_firstlineno
... ,   gcode.co_lnotab
... )
>>> 
>>> l = {'test':3,'toto':4,'t':2}
>>> g = {'x':1}
>>> 
>>> print l
{'test': 3, 'toto': 4, 't': 2}
>>> eval(newcode,g,l)
locals in g : {'test': 3, 'toto': 4}  # There is no more 't' here !!!
>>> print l
{'test': 3, 'toto': 4}                # 't' has been deleted from
dictionnary


The strange thing is that no new dictionnary is created as expected, but
local variable have been deleted from this dictionnary, why is this ? Is
there some way to modify my code to avoid this deletion ?

Boris

executio


From BPettersen at NAREX.com  Fri Apr 26 18:35:18 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 26 Apr 2002 16:35:18 -0600
Subject: String handling bug in Python
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192215166@admin56.narex.com>

> From: Tim Peters [mailto:tim.one at comcast.net] 
> > I would think that would occur less frequently than \ at 
> the end of a 
> > string though (especially with windows path names).  I feel 
> I must be 
> > missing something?
> 
> The good news is that I also feel that you are <wink>.  

Well, I'm glad we're in agreement <wink>

> r-strings were primarily intended to ease writing regular 
> expressions (which can never end with an odd number of 
> backslashes), and people using \ in Windows pathnames are 
> usually missing an obvious alternative (like the forward 
> slash -- Windows API functions couldn't care less which kind 
> of slash you use; and portable code should avoid any literal 
> slashes via os.path.join()).

It seems a shame to limit such a potentially useful feature to only one
domain.

Believe me, I get your windows comment, but when you work in a purely
windows shop what "should" be done to make code portable doesn't fly
unless it's also shorter, easier, and makes the code faster <wink>.

At-least-they-provide-us-with-free-beer'ly y'rs
-- bjorn




From kseehof at neuralintegrator.com  Sat Apr  6 23:34:01 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 6 Apr 2002 20:34:01 -0800
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204070151.g371pEdk018077@ratthing-b246.strakt.com>
Message-ID: <MEENLMFCHGAOOICLMCHHMEDNDLAA.kseehof@neuralintegrator.com>

> Guido:
> > Laura, you've said this over and over, and I don't buy it.
> > 
> > Two arguments from XP apply here:
> > 
> > - YAGNI (You Ain't Gonna Need It), or also "do the simplest thing that
> >   can possibly work".  If all you need today is a bool, use a bool.
> 
> Aha, now I understand a large part of why we disagree.  A bool is not
> simpler than an int.

Yes it is.  For example, a boolean can be fully interpreted and
processed by a single if statement.

> 'Something that can change state' is simpler than
> 'something that can change state and must only have one of two values'.

Is "Is it?" simpler than "What is it?" or isn't it?

> Do you go around making functions that return bool all the time, only
> then to change them to return int every time you have the idea that 
> there actually are 3 states here you want to model?  Then you will
> be littered with one obsolete 2-valued function every time this happens.
> They pile up, and since your wretched customers have all started using
> the 2-value functions, you are stuck having to maintain these things,
> even though you now hate them, and want everybody to use the 3-value
> versions.

Hmm.  My code isn't littered with obsolete "2-valued functions", and I've
been writing boolean functions since the first Borland C++ compiler was
released.  Actually I don't recall -ever- changing a boolean function to
return anything else.  I've simply never had the need.  You're still
missing the point: the concept of "2-value" functions is completely
separate from the concept of "boolean" functions.  Please separate these
two utterly distinct concepts in your mind.  A boolean function answers a
Yes/No question.  If you still don't understand this, I will be forced to
concede your earlier point, that the concept of "boolean" is difficult to
teach :-)

All of your examples fit in one of two categories:

1. Things that very clearly should not have been boolean in the first
place.  This includes anything that is not a Yes/No concept.

2. Things that should remain boolean, and would break everything if
changed to return more values.  These include anything that the user
expects a Yes/No from (based on documentation and naming).

You really shouldn't encourage your students to change interrogatory
functions to return 3 or more values.  It's a really bad idea.  Actually
go ahead, I don't mind.  When I'm 60, I might need a job fixing their
bugs :-)  The problem is that existing code expects a boolean value
(even if expressed as 0,1) and will break.

Note that if I wrote a function to classify your examples into the two
categories listed above, I would not use a boolean function.  Get it?

- Ken Seehof





From l0819m0v0smfm001 at sneakemail.com  Thu Apr 11 22:26:04 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Fri, 12 Apr 2002 02:26:04 GMT
Subject: Beginner's performance problem
References: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de> <memo.20020411143211.1400B@a.a> <a95d1v$avb$1@panix1.panix.com>
Message-ID: <3CB60C73.1010101@sneakemail.com>

Aahz wrote:

> In article <memo.20020411143211.1400B at a.a>,
> Mark Charsley <mark.charsley at REMOVE_THIS.radioscape.com> wrote:
> 
>>TVM to everyone who helped.
>>
> 
> I'm familiar with TVP, but what's TVM?
> 

   TVM is obvious Ta Ever So

   ... or something like that. <wink>

   Joshua



From vze3myxh at verizon.net  Mon Apr 15 19:15:07 2002
From: vze3myxh at verizon.net (d506)
Date: Mon, 15 Apr 2002 23:15:07 GMT
Subject: Newbie: Win32 gui examples?
Message-ID: <3CBB5EC6.4050602@verizon.net>

Hi,

Does anyone know some place where I can get some code samples for making 
simple GUIs using Win32all extensions?

My users are not entirely DOS prompt savvy, so all I want to do is to 
make a small GUI that has File->Open, File->Exit, Reports->1,2,3..., 
Options, About menus that call different actions (no parameters other 
than file name is required). At the moment, I compile things using 
Py2exe so that the whole thing fits in two floppies. I'd like to stick 
to Win32 extensions to keep the file size small - as opposed to PyQT or 
wxPython.

Thanks in advance
Babu



From doru-cat at ifi.uio.no  Mon Apr 22 05:32:32 2002
From: doru-cat at ifi.uio.no (Doru-Catalin Togea)
Date: Mon, 22 Apr 2002 11:32:32 +0200 (MET DST)
Subject: Interesting topics
Message-ID: <Pine.GSO.4.21.0204221109040.6191-100000@fjorir.ifi.uio.no>

Hi all,

I have subscribed to several other lists besides this one, (not python
lists) and I noticed that there are being posted quite interesting mails
here. On other lists, even though unmoderated, people do not tolerate that
many off-the-topic postings. 

Personally, I don't really care, YET. I have to delete over 150
postings everyday anyways, so I've become really proficient at hitting the
'D' key which marks mails for deletion in PINE. But if the ratio of noise
to signal increases too much on this list, it can become boring to remain
on it.

On the "Jessica Felix Designs - Art and all that Jazz" site, 
(http://www.artandallthatjazz.com), they claim that somebody is spamming
in her name, so maybe not everybody should be fried.

Just curious: I have come across some pretty weird statements in emails,
like the two spam mails supposedly from Jessica F. In the last one, one
can read:

	Who will we open after Bernadette kicks the pathetic star's boat?

	We fear the polite gardner.

I wonder, are these statements computer generated? Or do they make sense
if you have seen a certain movie, a TV-show, or read a certain book? I
don't want to hurt anybodys feelings, but TO ME, their being sent in
emails seems like the symptoms of a mental disorder of some kind.

Regards,
Catalin

	<<<< ================================== >>>>
	<<     We are what we repeatedly do.      >>
	<<  Excellence, therefore, is not an act  >>
	<<             but a habit.               >>
	<<<< ================================== >>>>







From unendliche at hanmail.net  Fri Apr 12 05:47:52 2002
From: unendliche at hanmail.net (Seo Sanghyeon)
Date: 12 Apr 2002 02:47:52 -0700
Subject: Magic words
Message-ID: <45e6545c.0204120147.3707d410@posting.google.com>

Hello. I am new to Python. I have a question. I got confused
by all those 'magic words' -- __init__, __class__, __doc__,
__name__ and so on. Where can I find a documentation about
these words? Or is there a list of 'magic words'?

Happy Python!
Seo Sanghyeon


From phlip_cpp at yahoo.com  Tue Apr 23 22:20:22 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 24 Apr 2002 02:20:22 GMT
Subject: PyGtk TestRunner for PyUnit?
Message-ID: <aa54p6$g2h@dispatch.concentric.net>

Hypo Nt:

I found this textage on the web:

"A Tkinter-based GUI test runner is in the works. It will look suspiciously 
like JUnit, and will be made available as soon as I have learnt enough 
about Tkinter to complete it. Any volunteers for this task would be 
welcome. How about integrating such a GUI with the IDLE IDE? (There is 
probably little point in writing a GUI front end for only one target 
platform, e.g. by using PyGTK."

Well [flame war bait] PyGTK is not just "one" target platform.

But more important, if I want to seamlessly bond PyUnit with Glimmer (using 
scripts I'l be publishing on their list server soon), then I need to 
interact healthily with Gtk.

Each system's mainloop occludes the other.

I don't want to 'popen' to another shell because I want the interaction 
between Glimmer and PyUnit to be bidirectional.

Has anyone written a PyGtk TestRunner for PyUnit?

-- 
Phlip

http://www.greencheese.org/HatTrick


From speeves at unt.edu  Mon Apr 29 15:05:29 2002
From: speeves at unt.edu (speeves)
Date: 29 Apr 2002 12:05:29 -0700
Subject: changes to win server registry with python script not appeari ng in gui tools
References: <mailman.1019573801.4240.python-list@python.org>
Message-ID: <d6351bca.0204291105.6c3474c6@posting.google.com>

I mean the odbc tool in Admin tools in the control panels.  I was
thinking about this over the weekend, and it doesn't make sense to me.
 I said python, but I am only using python to make calls to the win32
api.  (and creating .reg files and using regedit commandline to create
dsns)  If these are not working, then this has to be a bug right?
Calls using Microsoft's own tools should be working on the os, I
think...  Also, I need the ability to create mysql/ms sql server
dsn's, but now have to physically go to the machine to create them,
because the coldfusion admin, is looking for dsns in the same place as
the odbc gui tool, meaning that they don't see the dsns, unless I
create them in the odbc gui.  I need this resolved asap, cause I am in
the middle of many migrations of database. :(

Unhappy Server Admin,
speeves
UNT

Christopher Encapera <ChrisE at lantech.com> wrote in message news:<mailman.1019573801.4240.python-list at python.org>...
> You do mean regedit and regedt32 - or does gui tools refer to something
> else...?
> 
> 
> 
> Christopher Encapera
> Lantech.Com
> chrise at lantech.com
> 502-267-4200
> 
> 
> -----Original Message-----
> From: speeves at unt.edu [mailto:speeves at unt.edu]
> Sent: Monday, April 22, 2002 8:56 PM
> To: python-list at python.org
> Subject: changes to win server registry with python script not appearing
> in gui tools
> 
> 
> Hi!
> 
> The title says it all...  Why don't the changes that I am making to
> the registry appear in the gui tools?  I have completely automated
> many of the gui-functions with a simple python script, but any query
> to the registry must now come from the python scripts, and not the gui
> tools, as they are never current anymore :(
> (Even after rebooting)
> 
> Just wondering...  (Seems like a fluke in the win os to me, but...)
> 
> speeves
> central Web Support
> UNT


From pyth at devel.trillke.net  Sat Apr 20 07:49:06 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 20 Apr 2002 13:49:06 +0200
Subject: Substitution with Hash Values
In-Reply-To: <slrnac1lbh.nma.jon+usenet@snowy.squish.net>
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net>
Message-ID: <20020420114906.GA10574@devel.trillke>

On Sat, Apr 20, 2002 at 02:41:53AM -0000, Jon Ribbens wrote:
> In article <mailman.1019254077.7316.python-list at python.org>, holger krekel wrote:
> > 	 re.sub('\{\{(.*?)\}\}',lambda x: dict[x.group(1)], str)
> 
> Need a raw string there, i.e.
> 
>   r"\{\{(.*?)\}\}"

yes, i am sometimes confused by 'raw strings' and 'strings'. 
Could you help by saying why exactly it is needed here?

> Plus 'str' is perhaps not the best choice of variable name ;-)

right. 

	holger




From martin at v.loewis.de  Wed Apr  3 02:02:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Apr 2002 09:02:43 +0200
Subject: A faster Python?
References: <a8dcqr$p29$1@laurel.tc.umn.edu>
Message-ID: <m31ydxp8mk.fsf@mira.informatik.hu-berlin.de>

Scott Hennesy <henne021 at tc.umn.edu> writes:

> What's the consensus out there? Do Psyco, NumPy, and SciPy fulfill
> the need for speed or is there a desire for a faster Python?

Among the Python contributors, I think the consensus is that there is
a desire for Python working faster, not for a faster Python.

If you have something that runs too slow, analyse why that is, then
think about how to improve it. If you find that changes to the
interpreter are needed, and if you think the changes are generally
useful, don't break existing applications (or break them in a way
compatible with PEP5), and don't involve "dirty hacks", please
contribute a patch to sf.net/projects/python.

IOW, a rewrite of the complete interpreter is nowhere in sight -
better aim for gradual evolution than revolution.

Regards,
Martin



From narnett at mccmedia.com  Wed Apr 24 11:59:33 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 24 Apr 2002 08:59:33 -0700
Subject: Probably I'm just too stupid, but...
In-Reply-To: <34e7f752.0204240746.7d2c41b0@posting.google.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJCENPKDAA.narnett@mccmedia.com>

What were you expecting?

If you wanted to call the function, you'd use c().  When you entered just
'c', you were asking about the object, not running it.

I'm sure that many of us have made the same mistake sometimes... I have.

Nick

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of redflag
> Sent: Wednesday, April 24, 2002 8:46 AM
> To: python-list at python.org
> Subject: Probably I'm just too stupid, but...
>
>
> >>> def c():
> 	for i in range(10):
> 		print i





From fperez528 at yahoo.com  Mon Apr 15 17:55:41 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 15:55:41 -0600
Subject: String interpolation question
References: <a9fcmo$npi$1@peabody.colorado.edu> <uuHu8.23410$vF6.768198@news2.tin.it>
Message-ID: <a9fial$r0g$1@peabody.colorado.edu>

Alex Martelli wrote:

> If you're a string-interpolation junky:
> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66018
> 
> and particularly Skip Montanaro's variant (acknowledgedly inspired
> by a post by Steve Majewski) may be to your taste.  With that:
> 
>         print "%$(y[i])s" % Eval()
> 
> works basically as you wish (I could quibble with a few details, but
> it would be quibbling).

Thanks! Very nice, and now that you mention it, I _had_ that bookmarked but 
had forgotten it. I'm now using Skip's variant. I still feel that here perl's 
syntax is cleaner, but that argument is getting old, so I'll shut up ;)

Regards,

Fernando.


From torkel at sm.luth.se  Tue Apr 23 01:43:26 2002
From: torkel at sm.luth.se (Torkel Franzen)
Date: 23 Apr 2002 07:43:26 +0200
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <alwyn-0BB274.17465022042002@news-text.blueyonder.co.uk>
Message-ID: <vcbit6jezr4.fsf@beta13.sm.luth.se>

Alwyn <alwyn at alwyn.demon.co.uk> writes:

> I recall that Russell reports having had discussions with Goedel, 
> Einstein and other distinguished emigres in America. He found their 
> ideas impossibly 'Platonist', and he could not abide them. 

  Russell does indeed comment on Godel's platonism, but Einstein
doesn't enter into it. Russell was himself very much a platonist
in his early days.




From cmkleffner at gmx.de  Wed Apr 10 13:26:35 2002
From: cmkleffner at gmx.de (cmkl)
Date: 10 Apr 2002 10:26:35 -0700
Subject: Help: Python for a big commercial project?
References: <mailman.1018437980.14078.python-list@python.org>
Message-ID: <3b091a1c.0204100926.4ba2f13b@posting.google.com>

I sucessfully use Python2.2 on HPUX 11 as well as I used Python2.1 on HPUX 10.2
with InformixDB v1.3 and other useful extensions for development of medium sized
inhouse test tools and database service tools.
I agree: IT IS PAINFUL learn to build Python on HPUX but nevertheless it is 
worth doing. DONT'T use the prebuilt buggy python versions of the 
HPUX porting Center !! Build python yourself, it is then much easier to get 
working C-extensions like database modules, numeric python ...

threads is an issue on HPUX platforms. With HPUX11 I was sucessful to build
thread enabled python (tests sucessful) but I was not able to build a working 
multithreaded database application with a multithreaded enabled Informixdb
module. But it may be possible. It is a matter of time for trying and testing
different settings at compile time. 

On the other hand I was able to build a single-file python interpreter with
several frozen modules - perfect for painless distribution to other HPUX's.
Try this i.e. with perl or the JAVA JRE !!!

Development time is dratically reduced compared to the classical approach
with mixed C / ksh applications. 

Sincerly

Carl KLeffner

"Cereal Muesli" <cereal_con_fibra at hotmail.com> wrote in message news:<mailman.1018437980.14078.python-list at python.org>...
> Hi,
> 
> 	I need a piece of advice from you, Oh wise Pythoneers! 8-)
> 
>   I'm involved in the development of a big DataWarehouse for an important 
> TelCo, and I want to propose Python as
>   the language of choice for internal maintenance tasks, high level control 
> tasks for the loading procedures, etc...
>   the people here is not likely to use any free stuff for "serious" 
> development (you know...no money = no quality, sigh!),
>   but i want to change this, since I believe that Python is really a very 
> good product. I have played with Python for
>   a year or so under Windows, and before trying to convince people here that 
> it is a good choice, I need to know
>   some things:
> 
> - First of all; do you think that Python is suitable for this
>   kind of "serious, big" project?
> 
> - The core DW system and ETL logic runs under a HP-UX 11
>   system. It seems to me that there are too many complains
>   floating around about problems compiling/using Python under
>   this platform; am I right? Does anybody have any good/bad
>   experiences to share about using Python in this kind of OS?
> 
> - Installation: How easy or how hard is the installation of
>   Python for HP-UX 11? I don't want to disturb sys admins more
>   than necessary; do they have to compile everything? Do they
>   need to install GCC to do that? Are there any reliable
>   full-featured pre-built binaries for this platform available to
>   "download-and-install-in-a-minute"?
> 
> - At least two of the candidate modules to be developed in
>   Python require a GUI under X Windows. Does the distribution of
>   Python for HP-UX include TKinter and TK? Do we need a separate
>   download/installation of TCL/TK? In this case, is
>   TCL/TK/TKinter hard to build/install/configure for Python?
> 
> - Database access would be required for some data analysys
>   modules, specifically ORACLE 8 access would definitely be a
>   requirement; how well is this supported under HP-UX? I have
>   seen there are some 3rd party contributions for DB access
>   under Python, but don't know if they work fine for this
>   arquitecture/OS. And again, how hard is to make them run in HP-UX
>   11 (compilation needed, additional libraries, etc...)?
> 
> - Finally: Which version of Python should we use? it seems that
>   1.5.x still rules under *NIX, doesn't it?
> 
> 
> BIG thanks to everybody, help is *really* appreciated.
> 
> 
> Have a nice day!
> 
> _________________________________________________________________
> MSN Photos es la manera m?s sencilla de compartir e imprimir sus fotos: 
> http://photos.latam.msn.com/Support/WorldWide.aspx


From shalehperry at attbi.com  Wed Apr 17 09:50:00 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 17 Apr 2002 06:50:00 -0700 (PDT)
Subject: Classes Again !
In-Reply-To: <3cbd7527$1@news.comindico.com.au>
Message-ID: <XFMail.20020417065000.shalehperry@attbi.com>

On 17-Apr-2002 Graeme Matthew wrote:
> Hi all  I am going mad again, but making progress, been on a problem for 1
> hour already and cannot see the cause ! please could someone help
> 
> The getSubDocument method in my template class is failing suddenly , i dont
> have a clue why ! thanks a mil
> 

$ python
>>> class Foo:
...   def one(self): pass
...   def two(self): pass
... 
>>> dir(Foo)
['__doc__', '__module__', 'one', 'two']

this shows me everything in the Foo class.  Try the same with your class and
see what happens.  Perhaps you have a subtle indentation problem so
getSubDocument is being defined in the previous function's namespace.




From ajs at ix.netcom.com  Sun Apr  7 19:38:00 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 7 Apr 2002 19:38:00 -0400
Subject: ACCEPTED: PEP 285
Message-ID: <000101c1de8f$93645400$0334fea9@carol>

Erik -

>And many people responded that it was a good idea.

But the whole thing risks coming off as a loyalty
test of some sort.

Does it bother you that Paul Rubin is being 
"classified" by someone as close in to the
inner circle as Aahz.

Do you understand what set that off beyond
Paul expressing what he felt is right for Python,
after having been *asked* to do so.

Has Paul (with whom, to repeat, I only had
one exchange with - nasty as hell) somehow 
exhibited less devotion to the "cause"?

I am in some sense a true dissident.

There is a something here I find unpleasant 
and I have chosen to be vocal about it.

Stuff coming at *me* I sort of understand.

Paul was *asked* to express an opinion
and did so. 

I truly do try to hold open the possibility I
am missing something. And say so. Because 
sometimes I most certainly am, and I need the 
escape hatch.  Survival instinct.

But the fact that Aahz has Paul on the
hotseat seems at the moment as more of the
same kind of stuff that has be off the bus.   

Can Aahz explain where Paul sinned, other 
than questioning 'policy' - after in effect his
(and others) opinions were actively sort?

Does he just have the wrong opinion. 

Technically wrong?  Wrongly intentioned. 
An IQ issue?

What the hell was that?

Can someone explain it for Aahz?

Art 








From emile at fenx.com  Wed Apr 10 19:40:43 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 10 Apr 2002 23:40:43 GMT
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu>
Message-ID: <%34t8.36244$CU.41218@rwcrnsc51.ops.asp.att.net>

Rajarshi Guha
>   I have a list like this:  l = [1,'s', 2,'t', 3,'d']
> 
[snip]
> whereas I want
> 
> 1 s
> 2 t
> 3 d
> 

>>> for x,y in [(x,y) for x,y,z in zip(l,l[1:],[1,0]*len(l)) if z]:
...     print x,y


Have fun...  ;-)

--

Emile van Sebille
emile at fenx.com



From mlh at vier.idi.ntnu.no  Fri Apr 12 08:01:44 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 12:01:44 +0000 (UTC)
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com> <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no>  <mailman.1018592665.14227.python-list@python.org>
Message-ID: <slrnabdj58.mnq.mlh@vier.idi.ntnu.no>

After a few draft replies: Yes, type checking is needed for some
applications, such as pickling. However, I don't think there is a
general (intensive) description of the set of "instances" in the new
paradigm; you would probably be better of by a specific (extensive)
description. I.e. special-case what you need (e.g. strings and
numbers) and then check whether something is one of those types or
not.

Maybe... :)

On the other hand -- you don't _have_ to special-case anything. I
believe it's possible to construct any Python value through
function/method calls. But then your XML probably wouldn't be very
pretty ;)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From logiplexsoftware at earthlink.net  Tue Apr 16 18:32:12 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 16 Apr 2002 15:32:12 -0700
Subject: A little advice please? (Convert my boss to Python)
In-Reply-To: <B348A2683937CC65.3942DE81F07F4146.ED3D7BA18F22B5A1@lp.airnews.net>
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
	<3CBB8ABB.57BAA189@engcorp.com>
	<mailman.1018989153.2077.python-list@python.org>
	<B348A2683937CC65.3942DE81F07F4146.ED3D7BA18F22B5A1@lp.airnews.net>
Message-ID: <20020416153212.0bde8cde.logiplexsoftware@earthlink.net>

On 16 Apr 2002 17:05:09 -0500
Cameron Laird wrote:

> In article <mailman.1018989153.2077.python-list at python.org>,
> Cliff Wells  <logiplexsoftware at earthlink.net> wrote:
> >Especially when the money saved by developing in Python versus C could
easily
> >cover the cost of a faster PC (or two).  It seems so bizarre that there are
> >people who would spend $10,000 (or more) in developer time to optimize an
> >in-house program rather than simply spend $1000 on a new PC.
>
> 'Haven't studied abnormal organizational psychology,
> have you?  This is only a blip in the collective
> irrationalities rampant around us.  Those $10,000 and
> $1000 might be decoupled from the decision-makers
> managing software development.  Moreover, someone in
> the system might be aware that the total cost of the
> PC should be amortized as around $10,000, once all
> management overhead is properly accounted.

- Cost of paperwork/meetings to get funds approved for $1000 PC: $500
- Consulting fees to decide if perhaps the PC should instead be an AS/400 and
  should we upgrade everyone to Windows XP: $5000
- Cost of additional paperwork/meetings to convince management otherwise: $500
- $1000 PC through purchase-agreement: $4000
- Flock of migrating developers: Priceless.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 20:50:07 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 13 Apr 2002 00:50:07 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <aAKt8.22045$ir6.979928@news1.calgary.shaw.ca>
Message-ID: <slrnabf05v.mfc.huaiyu@gauss.almadan.ibm.com>

Brad Bollenbach <bbollenbach at shaw.ca> wrote:
>In article <slrnabeqkg.mfc.huaiyu at gauss.almadan.ibm.com>, Huaiyu Zhu wrote:
>> Brad Bollenbach <bbollenbach at shaw.ca> wrote:
>>>
>>>This is a semantic debate. I know that Stallman chose the term "Free
>>>Software" because he felt it was the best (and most concise) way to
>>>express what he was thinking of.
>> 
>> That's right.  A large portion of OSS is FS.  But another large 
>> portion is  not.  FS is a proper subset of OSS.
>
>Fine, I guess I'll have to cite Stallman directly to back up what I
>said. :) From http://www.fsf.org/philosophy/categories.html#OpenSource:
>
><begin gnu>
>Open Source software
>====================
>
> The term ``open source'' software is used by some people to mean more
> or less the same thing as free software. We prefer the term ``free
> software''; follow that link to see the reasons.
><end gnu>

So RMS says using "open source" to mean "free" is not what FSF prefers, even
though many people use it that way.  It is definitely not what OSI prefers,
either.  Why use it that way?

Using two terms to mean the same concept when we do have two distinct
concepts to name leaves us one term short.  What name do you propose for the
broader concept of "open source" as defined by OSI?

Names are precious commodity in today's world.

Huaiyu


From stephen at xemacs.org  Mon Apr 15 10:00:59 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 15 Apr 2002 23:00:59 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain>
Message-ID: <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <hzhu at mars.localdomain> writes:

    Huaiyu> Well, what you wrote does not help in any way to clarify
    Huaiyu> the situation, except to make it clear that you have a big
    Huaiyu> issue with GPL.

I have no issue with the GPL, because the GPL only affects people who
chose to use GPLed code in their projects.  It's a wonderful behavior-
modification tool; it often causes people to free software they would
otherwise keep proprietary.  Most of my development effort is spent on
XEmacs, which is GPL through and through.  The GPL is certainly does
not affect my decisions about what existing projects to contribute to.

I do have issues with some GPL advocates, because their arguments
amount to claiming that nobody should have property rights in his own
code.  (rms says this explicitly, but very few are so courageous as
he.)  And I have issues with many other GPL advocates, because _their_
arguments amount to claiming they have property rights in other
people's code.  But that's a different thread.

    Huaiyu> How come they are not equal if they are also the same?

I can only suggest that you read what the protagonists say on their
own web sites.  FWIW, I find the OSI pages to be more accurate, but
unnecessarily insulting (and quite sparse, perhaps a veiled insult?)
The GNU site often substitutes rhetorical tricks for substance, but
does its best to give the OSI position a fair reading.

http://www.opensource.org/
http://www.gnu.org/philosophy/

Perhaps what you're looking for is Joe Barr's think piece "Live and
let license" (http://www.itworld.com/AppDev/350/LWD010523vcontrol4/).
This seems to mirror many of your concerns, and the differences you
have with people who have responded in this thread.  However, I doubt
you'll find any "open source" activist who agrees with him,
particularly his conclusion #1.

The main problem is that Barr writes about "open source": "Think of
source code that is widely available. Source code that can be both
viewed and changed by just about anyone who wants to bother." and then
continues as if this has no relation to licensing.  How can it not?
Barr doesn't even try to answer that, thus sowing more seeds of
confusion than he resolves.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From hgs at dmu.ac.uk  Wed Apr 17 10:59:09 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Wed, 17 Apr 2002 15:59:09 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <Pine.GSO.4.40.0204171007080.18149-100000@neelix>
Message-ID: <Pine.GSO.4.40.0204171545270.18149-100000@neelix>

I have made another attempt to build Python-2.2.1, this time only adding
sunaudiodev to the list of modules.


Script started on Wed Apr 17 15:33:33 2002
neelix hgs 11 %> gmake test
        [...]
WARNING: removing "struct" since importing it failed
        [...]
WARNING: removing "regex" since importing it failed
        [...]
WARNING: removing "pcre" since importing it failed
        [...]
WARNING: removing "_hotshot" since importing it failed
        [...]
WARNING: removing "_weakref" since importing it failed
        [...]
WARNING: removing "xreadlines" since importing it failed
        [...]
WARNING: removing "array" since importing it failed
        [...]
WARNING: removing "cmath" since importing it failed
        [...]
WARNING: removing "math" since importing it failed
        [...]
WARNING: removing "strop" since importing it failed
        [...]
WARNING: removing "time" since importing it failed
        [...]
WARNING: removing "operator" since importing it failed
        [...]
WARNING: removing "_codecs" since importing it failed
        [...]
WARNING: removing "_testcapi" since importing it failed
        [...]
WARNING: removing "unicodedata" since importing it failed
        [...]
WARNING: removing "_locale" since importing it failed
        [...]
WARNING: removing "fcntl" since importing it failed
        [...]
WARNING: removing "pwd" since importing it failed
        [...]
WARNING: removing "grp" since importing it failed
        [...]
WARNING: removing "errno" since importing it failed
        [...]
WARNING: removing "select" since importing it failed
        [...]
WARNING: removing "md5" since importing it failed
        [...]
WARNING: removing "sha" since importing it failed
        [...]
WARNING: removing "binascii" since importing it failed
        [...]
WARNING: removing "parser" since importing it failed
        [...]
WARNING: removing "cStringIO" since importing it failed
        [...]
WARNING: removing "cPickle" since importing it failed
        [...]
WARNING: removing "mmap" since importing it failed
        [...]
WARNING: removing "rotor" since importing it failed
        [...]
WARNING: removing "syslog" since importing it failed
        [...]
WARNING: removing "timing" since importing it failed
        [...]
WARNING: removing "audioop" since importing it failed
        [...]
WARNING: removing "imageop" since importing it failed
        [...]
WARNING: removing "rgbimg" since importing it failed
        [...]
WARNING: removing "crypt" since importing it failed
        [...]
WARNING: removing "_socket" since importing it failed
        [...]
WARNING: removing "dbm" since importing it failed
        [...]
WARNING: removing "gdbm" since importing it failed
        [...]
WARNING: removing "termios" since importing it failed
        [...]
WARNING: removing "resource" since importing it failed
        [...]
WARNING: removing "nis" since importing it failed
        [...]
WARNING: removing "_curses" since importing it failed
        [...]
WARNING: removing "_curses_panel" since importing it failed
        [...]
WARNING: removing "fpectl" since importing it failed
        [...]
WARNING: removing "zlib" since importing it failed
        [...]
WARNING: removing "_tkinter" since importing it failed
running build_scripts
not copying /home/hgs/Python-2.2.1/Tools/scripts/pydoc (up-to-date)
./python -E -c 'import sys ; from distutils.util import get_platform ; print get_platform()+"-"+sys.version[0:3]' >platform
find ./Lib -name '*.py[co]' -print | xargs rm -f
./python -E -tt ./Lib/test/regrtest.py -l
Traceback (most recent call last):
  File "./Lib/test/regrtest.py", line 54, in ?
    import random
  File "/home/hgs/Python-2.2.1/Lib/random.py", line 76, in ?
    from math import log as _log, exp as _exp, pi as _pi, e as _e
ImportError: No module named math
gmake: [test] Error 1 (ignored)
./python -E -tt ./Lib/test/regrtest.py -l
Traceback (most recent call last):
  File "./Lib/test/regrtest.py", line 54, in ?
    import random
  File "/home/hgs/Python-2.2.1/Lib/random.py", line 76, in ?
    from math import log as _log, exp as _exp, pi as _pi, e as _e
ImportError: No module named math
gmake: *** [test] Error 1
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Wed Apr 17 15:44:07 2002


So even building statically doesn't work.

        Hugh





From aahz at pythoncraft.com  Mon Apr 15 21:36:17 2002
From: aahz at pythoncraft.com (Aahz)
Date: 15 Apr 2002 21:36:17 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <a9eppf$ghv$1@panix1.panix.com> <a9f95n$jg0$1@bob.news.rcn.net>
Message-ID: <a9fv6h$f9c$1@panix3.panix.com>

In article <a9f95n$jg0$1 at bob.news.rcn.net>,
Raymond Hettinger <python at rcn.com> wrote:
>[Aahz]
>>[Raymond Hettinger]
>>>
>>>The help and tutor maillists indicate that the use of input() rather than
>>>raw_input() is widespread.  I think this function is so harmful and so
>>>likely to be used that it should be deprecated hasta pronto (right away).
>>>When really needed, eval(raw_input()) is easy enough to code and explicit
>>>about what it does.
>>
>> Write a PEP.
>
>Rules of the game:
>'''A PEP that has not been discussed on python-list at python.org and/or
>python-dev at python.org will not be accepted.  ''' -- from Pep 1

First write the PEP, then submit it, then post it to c.l.py for
discussion.  Incorporate the results of the discussion into the revised
PEP (making sure to include dissenting opinions), then request BDFL
pronouncement ("BDFL pronouncement" == "accepted" or "rejected" or
"needs more work").

Make sure to reread *all* of PEP 1, plus PEP 9.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From phd at phd.pp.ru  Fri Apr 26 04:59:45 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 26 Apr 2002 12:59:45 +0400
Subject: ftplib and attributes
In-Reply-To: <aab41f$5hk$1@news1.wdf.sap-ag.de>; from daniel.dittmar@sap.com on Fri, Apr 26, 2002 at 10:44:31AM +0200
References: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it> <aab41f$5hk$1@news1.wdf.sap-ag.de>
Message-ID: <20020426125945.F24848@phd.pp.ru>

On Fri, Apr 26, 2002 at 10:44:31AM +0200, Daniel Dittmar wrote:
> > hi, im new bie and im trying to use ftplib, but my problem is to get
> > the attributes of the files, how can i do ?
> 
> Although you have to use .dir () as the other posters have written, it is
> important to know that different ftp servers will format the information in
> different ways.

   On the http://py.vaults.ca/parnassus/apyllo.py/126307487 there are few
high-level modules and .dir() parsers.

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




From marklists at mceahern.com  Wed Apr  3 16:06:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 15:06:17 -0600
Subject: date / time
In-Reply-To: <MPG.171506a638d9ca2c989694@news.nwlink.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEAOKJAA.marklists@mceahern.com>

I haven't followed this post.  For all I know, the OP has their solution.
For what it's worth, mx.DateTime works like a charm every time.  Consider
this:

import mx.DateTime
years = range(1999, 2010, 1)
dates = [mx.DateTime.Parser.DateTimeFromString("3/1/%s" % y) for y in years]
for d in dates:
  print "One day before %s is %s." % (d, d -
mx.DateTime.RelativeDateTime(days=1))

Cheers,

// m





From garys at ihug.com.au  Tue Apr  2 20:34:26 2002
From: garys at ihug.com.au (Gary Stephenson)
Date: Wed, 3 Apr 2002 11:34:26 +1000
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <a85od6$9i$1@panix1.panix.com> <DlTp8.48032$S52.1614804@news2.tin.it> <a8ajpi$i36$1@lust.ihug.co.nz> <k2dq8.50144$S52.1717644@news2.tin.it>
Message-ID: <a8dmgf$3ms$1@lust.ihug.co.nz>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:k2dq8.50144$S52.1717644 at news2.tin.it...

> Gary Stephenson wrote:
>         ...
> > Given that our objects
> > _are_ allowed to mutate their state, wouldn't it be preferable for them
to
> > do so in more intelligent ways?
>
> No.  "Change" is such a troublesome notion already, that to have it
> train behind itself a "keeping track of uncountably infinite sets" is
> quite a conceptual mountain.

I'm not sure we are actually talking about uncountably infinite sets here -
are we?.  Isn't it rather the finite and (probably) small set of heir
classes that need to be kept track of?  (I do agree that even this is
probably way too much book-keeping overhead to make it practicable in the
real world.)  And modelling "Change" well is surely something that we all
need to continually improve upon.

> Objects that do not change are perfectly well implementable and quite
> usable.  Anybody looking for a closer match to maths should IMHO turn
> right to lazy FP (Haskell...) rather than tilt agains the windmill of
> 'proper' modeling of change.

I have yet to really come to grips with Haskell - but look forward to doing
so.  However, I think what we are looking for here is not so much a closer
match to maths, but a closer match to reality - which has always been one of
the O-O's great marketing points - that it (supposedly) models reality more
closely.  Besides, just because nobody has yet come up with a really good
imperative system for subtyping doesn't necessarily mean that it's all
"tilting at windmills" (though it may very well prove to be so!).

> > Simple immutability is also insufficient to enable the subclass ==
subtype
> > identity [1], which seems to me to be one of the fundamental issues
> > underlying all this.
>
> Yes.  If you realize that 'behavior' is never intrinsic to a set of
values,
> and systematically _superimpose_ behavior on value sets (Haskell's
> typeclasses being of course the right way to do that:-), then you complete
> the conceptual trip to systematic purity.

Err, as I said, I'm not terribly Haskell enlightened (yet).  Does this
perhaps relate in some way to Dylan's approach of eschewing methods in
favour of generic functions (multi-methods)?

> Then you can decide whether
> the pragmatics of Python are preferable to the purity of Haskell, but if
> you do then it's a peculiar waste of time to keep yearning for purity
after
> choosing the compromise route of pragmatism.

Probably so.  But one man's "peculiar waste of time" might be another man's
life obsession.   Personally I am rather reluctant to label another person's
obsessions a "waste of time" - you just never know when they might come up
with something insanely great!  For example, some _very_ smart folk are
striving to
achieve a synergistic synthesis of the functional and imperative paradigms
[1]

> The set/bag example under immutability depends on functions such as
> put and remove returning new sets/bags -- a conceptual approach that
> simplifies things a lot, even though implementation-wise it may be too
> heavy.  But the insights you gain from there are still valuable and to
some
> extent applicable to impure worlds with mutating objects.  Among bag's
> axioms (w has/put/remove) should be something like the material
implication:
>     (has item bag) => (has item (remove item (put item bag)))
> and among set's axioms w has/put/remove should be:
>     () => (not (has item (remove item set)))
> so the impossibility of any one entity X satisfying both sets of axioms
> is trivial to prove, no?

Yes. Extremely lucid and succinct!  I wonder how often such axiomatic
analysis
would identify pathology in our inheritance relationships.  Or does
"pragmatism beats purity" mean we shouldn't worry overly much about such
things?  "If it gets the job done .." and all that. <g>

Many thanks for your thoughts - they are, as always, most illuminating.

respectfully,

gary

[1] - http://www.uwm.edu/~whopkins/seqcd/jolt.txt





From marklists at mceahern.com  Tue Apr  2 11:13:04 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 2 Apr 2002 10:13:04 -0600
Subject: httpget
In-Reply-To: <2eea6430.0204020737.67749830@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEMDKIAA.marklists@mceahern.com>

[Max]
> I'm looking for a script like 'httpget' with a option to check a url
> (exists or not).

Have you looked at urllib?  (I've never used it, but it seems to provide
what you're looking for.)

// mark




From shalehperry at attbi.com  Thu Apr 18 12:33:21 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 18 Apr 2002 09:33:21 -0700 (PDT)
Subject: help splitting strings
In-Reply-To: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com>
Message-ID: <XFMail.20020418093321.shalehperry@attbi.com>

On 18-Apr-2002 Tony Case wrote:
> Hi all
>       I'm pretty new to Python which probably explains my problems,
> I'm trying to read data from a number of comma separated variable
> files.  These are fairly large, 5000 lines plus.
>       I'm reading the files and using split to separate the
> individual values, some numeric some string.  The problem is that a
> number of the lines have commas and double quotes embeded in the
> strings so there are more elements in the list returned from split
> than there should be, in one file there are 400 + lines like this so
> it's not easy to edit the original file.
>       Is there a programatic way of dealing with this?
> The input files are in the form :-
> int,int,int,"string","string",int,"string"
> etc.
> I tried using the re.findall and re.split but didn't get very far.
>       TIA     T Case
> -- 
> http://mail.python.org/mailman/listinfo/python-list

my first thought is something like this:

re.compile(r'^(\d+),(\d+),(\d+),((?:"[^"]+")|(?:[^,]+)),((?:"[^"]+")|(?:[^,]+)),
(\d+),((?:"[^"]+")|(?:[^,]+))$')

\d+ means any number.  The ugly part is the string matching.  "[^"]+" means
match a double quote, then some number of non-double quote chars followed by a
double quote.  This is wrapped in (?:) to make it not put that chunk in a
group.  Same goes for the alternative which is simply a string of anything but
commas.




From dradul at yahoo.com  Wed Apr 17 22:59:02 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Wed, 17 Apr 2002 21:59:02 -0500
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018875990.21271.python-list@python.org> <j43cxwm1o4.fsf@informatik.hu-berlin.de> <a9id9a$33oqt$1@ID-99513.news.dfncis.de> <m37kn6wzu4.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9m5r9$4b8ru$1@ID-99513.news.dfncis.de>

"Martin v. Loewis" <martin at v.loewis.de> escribi? en el mensaje
news:m37kn6wzu4.fsf at mira.informatik.hu-berlin.de...
> "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
>
> > And while you are at it... You have not compiled python itself as
PIC
> > code. No, gcc -shared *doesn't* work in Solaris unless you are using
GNU
> > Binutils instead of the native linker and assembler, something *not
> > recommended at all*.
>
> Can you elaborate? gcc -shared works fine on Solaris, and binutils
> also works without problems.

But binary output is not 100% compatible with system libraries. If you
use the native linker and assembler, the code will be compatible. This
has been considered as "God given truth" in the sun-managers list for
more than 10 years (since I first subscribed). The sun-managers list,
btw, is where all admins, hackers and users with root access discuss
Solaris issues.

> > Solaris' linker can't build a dynamic loading library/module using
> > object code compiled in position dependent mode.
>
> And rightly so. That should never be a problem, though.
>
> Actually, you *can* talk the Solaris linker to incorporate non-PIC
> text into a shared library. For that, it will make the text segment
> writable. The -mimpure-text option of gcc triggers that feature.

And open a full can of worms. Why hack the linker flags when you can
impose yourself some discipline in your administration tasks and
make life easier in the future? :)

BTW, this advice of yours does apply to the curses modules under
Solaris, as they are compiled as non-PIC code.

> > This means, that you need to have all external libraries compiled
with
> > the -fPIC flag, even if they are static.
>
> If you link static libraries into Python, I recommend to make the
> modules that use them also static. Then you don't need PIC code for
> those.

And still they won't work. As is the case of the curses modules and the
readline modules in 2.1 and 2.2...

> > Ahh! And make sure that you give the linker the actual library paths
> > you use with an -R flag, else the dynamic loader won't find them and
> > you'll have to use an explicit LD_LIBRARY_PATH environment variable
> > for each invocation of the interpreter.
>
> That is the prime reason why I dislike shared libraries in the first
> place - statically linked binaries are much easier to manage; use of
> shared libraries should be restricted to system libraries, IMO.

It depends on your mindset. Shared libraries do save disk space, and
Solaris default installs tend to create too small partitions for
application space. IMO, few people have the guts to do a hot repartition
(actually it should be called "partition stretching" :).



From cereal_con_fibra at hotmail.com  Wed Apr 10 07:24:59 2002
From: cereal_con_fibra at hotmail.com (Cereal Muesli)
Date: Wed, 10 Apr 2002 13:24:59 +0200
Subject: Help: Python for a big commercial project?
Message-ID: <F194tIpqqJgSt51NDa200002509@hotmail.com>

Hi,

	I need a piece of advice from you, Oh wise Pythoneers! 8-)

  I'm involved in the development of a big DataWarehouse for an important 
TelCo, and I want to propose Python as
  the language of choice for internal maintenance tasks, high level control 
tasks for the loading procedures, etc...
  the people here is not likely to use any free stuff for "serious" 
development (you know...no money = no quality, sigh!),
  but i want to change this, since I believe that Python is really a very 
good product. I have played with Python for
  a year or so under Windows, and before trying to convince people here that 
it is a good choice, I need to know
  some things:

- First of all; do you think that Python is suitable for this
  kind of "serious, big" project?

- The core DW system and ETL logic runs under a HP-UX 11
  system. It seems to me that there are too many complains
  floating around about problems compiling/using Python under
  this platform; am I right? Does anybody have any good/bad
  experiences to share about using Python in this kind of OS?

- Installation: How easy or how hard is the installation of
  Python for HP-UX 11? I don't want to disturb sys admins more
  than necessary; do they have to compile everything? Do they
  need to install GCC to do that? Are there any reliable
  full-featured pre-built binaries for this platform available to
  "download-and-install-in-a-minute"?

- At least two of the candidate modules to be developed in
  Python require a GUI under X Windows. Does the distribution of
  Python for HP-UX include TKinter and TK? Do we need a separate
  download/installation of TCL/TK? In this case, is
  TCL/TK/TKinter hard to build/install/configure for Python?

- Database access would be required for some data analysys
  modules, specifically ORACLE 8 access would definitely be a
  requirement; how well is this supported under HP-UX? I have
  seen there are some 3rd party contributions for DB access
  under Python, but don't know if they work fine for this
  arquitecture/OS. And again, how hard is to make them run in HP-UX
  11 (compilation needed, additional libraries, etc...)?

- Finally: Which version of Python should we use? it seems that
  1.5.x still rules under *NIX, doesn't it?


BIG thanks to everybody, help is *really* appreciated.


Have a nice day!

_________________________________________________________________
MSN Photos es la manera m?s sencilla de compartir e imprimir sus fotos: 
http://photos.latam.msn.com/Support/WorldWide.aspx





From eric.brunel at pragmadev.com  Wed Apr 17 05:57:52 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 17 Apr 2002 09:57:52 +0000
Subject: Problems building Python tkinter module on Solaris
References: <mailman.1018982613.15346.python-list@python.org>
Message-ID: <a9j9l9$221d$1@norfair.nerim.net>

Alex Jacques wrote:
[snip]
> I read the Python FAQ 3.18 and other mentions of trying another version of
> tcl/tk. Since I cannot remove the tcl/tk in my /usr/local (insufficent
> privilege, and other folks use it), I built tcl/tk 8.3.4 in my
> /usr/freeware tree (same place where I build/install Python).
> 
> My Python build "procedure" is
> 
> ./configure --prefix=/usr/freeware
> make
> 
> No luck: the make file still grabs the tcl/tk stuff from /usr/local.

To get Tcl/Tk from another location, you have to edit your Setup file in 
the Modules sub-directory of your Python installation. Search for 
"_tkinter" and edit the appropriate lines. Refer to the comments in the 
file: they're quite clear...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From greg at cosc.canterbury.ac.nz  Mon Apr 15 01:02:19 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 15 Apr 2002 17:02:19 +1200
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>
Message-ID: <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz>

"Bryce C. Mason" wrote:
> 
> I am doing my development on a PC (activeware's distribution) and the final
> project will be running on a macintosh.  Umm...I am a little worried that
> there is not a command line interpreter in the Mac, at least one I have not
> seen.  Help.  How do I port my project to the mac?

MacPython has a facility for creating an "applet" from
a Python script. This is a small Macintosh application
that can be launched from the Finder. You can drop
files on the applet, and their pathnames will appear in
sys.argv.

Standard input and output is done through a text
window put up by the Python interpreter.

You will have to avoid any need for redirecting
stdin/out/err from outside the program, because there
is no way to do this on the Mac. There is also no
way of specifying dash-option command line
arguments.

(Actually, I believe that if you hold down the
Option key while launching the applet, there is an
option for entering a Unix-style command line
through a dialog box. It's an awfully clunky
way to use a Mac application, though.)

Hope that helps,

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


From tim.one at comcast.net  Sun Apr 14 23:56:27 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 23:56:27 -0400
Subject: Topy news - and a request for Python pretty-printing
In-Reply-To: <3cba3516@giga.realtime.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEDAOOAA.tim.one@comcast.net>

[Ira D. Baxter]
> Our DMS Software Reengineering Toolkit can easily build prettyprinters
> for arbitrary langauges; we've done it for many standard ones.
> See http://www.semdesigns.com/Products/Formatters/index.html
> Not Python yet, but all DMS really needs is a good grammar definition.
>  (Anybody have a Python lexer/grammar definition?)

The input for Python's parser generator is in the Python source
distribution, file Grammar/Grammar.

Note that block structure is denoted *by* indentation in Python source, and
there are no "curly braces" (or other visible "begin block"/"end block"
tokens).  In the grammar, INDENT and DEDENT are the tokens generated by the
lexer when a change in indentation level (from the indentation of the
previous statement) is seen.





From egrossde at yahoo.de  Thu Apr 18 02:35:12 2002
From: egrossde at yahoo.de (jb)
Date: Thu, 18 Apr 2002 08:35:12 +0200
Subject: Reading from a binary file in a portable way
Message-ID: <3cbe6892_11@news.newsgroups.com>

I should like to read a (binary) file x.y in a buffer and then do things 
like getting the 16 bit integer that is at the offset n, n+1 in the file 
(maybe lowbyte, then higyhbyte.

Previously I did it only under MS-Windows where for example I could open the 
file with the flag O_BINARY but this is no longer possible under Linux.

What would a solution that works under both Unix and MS-Windows look like?

-- 
Janos Blazi


-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From sholden at holdenweb.com  Mon Apr 15 09:34:26 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 09:34:26 -0400
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122>
Message-ID: <LHAu8.11094$Ji7.10494@atlpnn01.usenetserver.com>

"Bengt Richter" <bokr at oz.net> wrote ...
[ ... ]
> Well, to tie in with existing spelled-out 'and' and 'or' operations,
> it should probably do something along those lines (i.e., evaluate to
> one or other of the objects, not one of the boolean values used in the
logic),
> but I can't think of something sensible.
>
Since in Boolean terms a xor b is equivalent to ((a and not b) or (b and not
a)), maybe a xor b should return whichever of a or b is not false.

>>> a = 42
>>> b = 0
>>> a ^ b
42
>>> b ^ a
42
>>>

Since that's exactly what it appears to do, stop fiddling with what already
works !

regards
 Steve






From mlh at vier.idi.ntnu.no  Fri Apr  5 23:44:45 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 04:44:45 +0000 (UTC)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org> <mailman.1017893479.16809.python-list@python.org>  <7xu1qsnk5j.fsf@ruckus.brouhaha.com>
Message-ID: <slrnaasv9t.178.mlh@vier.idi.ntnu.no>

In article <7xu1qsnk5j.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
[snip]
>Python already has a bunch of eq but not equal numeric types:
>
>  1 == 1L      1 is not 1L
>  1 == 1.0     1 is not 1.0
>  1 == 1+0j    1 is not 1+0j

  10**2 == 10**2    10**2 is not 10**2

This doesn't necessarily have to do with types... Object identity
isn't that important here, I think.

>I felt better about the PEP after Guido pointed this out.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From ture at lysator.liu.se  Sat Apr 27 18:58:11 2002
From: ture at lysator.liu.se (Ture =?iso-8859-1?q?P=E5lsson?=)
Date: 28 Apr 2002 00:58:11 +0200
Subject: String handling bug in Python
References: <mailman.1019853229.10992.python-list@python.org>
Message-ID: <upsn5gsq9o.fsf@svartsmurf.lysator.liu.se>

Tim Peters <tim.one at comcast.net> writes:

> Don't brush that off, because it's the heart of the problem:  you need
> *some* way to detect where the "TRULY raw string literal" ends.  Whatever
> way you pick to spell that then becomes "a problem" if you want to include
> the very same spelling *in* the raw string literal.

Obviously, the string containing just a single quote is written as

  1H'

just as the string containing three backslashes is written as

  3H\\\

:-)

  -- 4HTure



From mitchchapman at earthlink.net  Fri Apr  5 09:41:50 2002
From: mitchchapman at earthlink.net (Mitch Chapman)
Date: Fri, 05 Apr 2002 14:41:50 GMT
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com> <050420020836327939%comdog@panix.com>
Message-ID: <3CAD9B5F.8B64400@earthlink.net>

brian d foy wrote:
> In article <f72dac0d.0204050205.4cb3e8ac at posting.google.com>, Alan Kennedy <alanmk at hotmail.com> wrote:
> > Forgot to mention: Does this mean that the Golf competition is open to
> > Python entries :-)
> 
> i think Python code would be at a disadvantage since the golf umpires
> count whitespace in the number of strokes :(.


?  That's what gzip is for.

--
Mitch
(_What_ SyntaxError?)


From tejarex at yahoo.com  Thu Apr  4 03:53:14 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 04 Apr 2002 08:53:14 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <_vUq8.552234$pN4.39754796@bin8.nnrp.aus1.giganews.com>

"Guido van Rossum" <guido at python.org> wrote in message
news:mailman.1017873312.18135.python-list at python.org...
>     7) Should the name 'bool' be changed?
>     => No.
>        One reviewer has argued to make the name 'truth'.  I find
this
>        an unattractive name, and would actually prefer to reserve
this
>        term (in documentation) for the more abstract concept of
truth
>        values that already exists in Python.  For example: "when a
>        container is interpreted as a truth value, an empty container
>        is considered false and a non-empty one is considered true."

Especially since reading Laura's essay, I no longer consider that to
be the proper term.  'Truth' is only one model for dichotomy, which is
the proper generic descriptive term, but which is too long to use as
is and not gracefully abbreviated.  I thought of suggesting 'bit', but
that falsely implies a length of 1, even though a significant bit
difference of one is true.  I can see +s and -s to every name I have
considered, including bool.

> Clarification
>
>     This PEP does *not* change the fact that almost all object types
>     can be used as truth values.  For example, when used in an if
>     statement, an empty list is false and a non-empty one is true;
>     this does not change and there is no plan to ever change this.

almost? all?  I am under the impression that 'if x:' is universally
valid, and consider that one of the good features of Python's prime
dichotomy system.

>     The only thing that changes is the preferred values to represent
>     truth values when returned or assigned explicitly.  Previously,
>     these preferred truth values were 0 and 1; the PEP changes the
>     preferred values to False and True, and changes built-in
>     operations to return these preferred values.

I consider the change to be a change of hat (type) rather than a
change of value, just as I and Python consider 1 and 1L to be two
different guises for the same value.

>     - It has been suggested that, in order to satisfy user
>       expectations, for every x that is considered true in a Boolean
>       context, the expression x == True should be true, and likewise
>       if x is considered false, x == False should be true.

Python's nothing/something dichotomy divides all(?) Python objects
into two equivalence classes.  As is typical in mathematics, we choose
one member of each, 0/False and 1/True, as (canonical)
representatives.  It is not sensible to expect members of a class to
equal the chosen representative (except in relation to the
class-defining attribute).  People should really learn this much of
class logic.  If explained this way, I would hope that such
expectations would vanish.

Terry J. Reedy





From shagshag13 at yahoo.fr  Wed Apr 24 03:12:32 2002
From: shagshag13 at yahoo.fr (Shagshag)
Date: 24 Apr 2002 00:12:32 -0700
Subject: Python, search engine and NLP...
References: <409a56e2.0204230104.16d60f52@posting.google.com> <3CC60496.4070509@earthlink.net>
Message-ID: <409a56e2.0204232312.74e0cf71@posting.google.com>

Thank a lot for all yours answers !!! That's really cheering!!!

Here are some lib which might be useful for nlp :

pyxml : http://sourceforge.net/project/showfiles.php?group_id=6473&release_id=66587
a stemmer http://sourceforge.net/projects/pystemmer/
Natural Language Understanding program http://pytalk.sourceforge.net/
Natural Language Processing Toolkit http://nltk.sourceforge.net/
Data mining http://ai.fri.uni-lj.si/~aleks/orng/

If you know about others which might help let me know !

S13.


From marklists at mceahern.com  Wed Apr 17 12:11:50 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 17 Apr 2002 11:11:50 -0500
Subject: UTC equivalent of time.mktime() ?
In-Reply-To: <1019054451.562.319.camel@tp6>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com>

Why not just use the mxDateTime library and just do something like this:

	mx.DateTime.utc().ticks()

// mark




From phr-n2002a at nightsong.com  Fri Apr 26 02:10:47 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 23:10:47 -0700
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz>
Message-ID: <7xsn5jat20.fsf@ruckus.brouhaha.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
> But if RMS would be happy with a world in which there was simply no
> copyright, he wouldn't bother with the GPL -- he could get the same
> effect just by putting all the stuff he writes in the public domain.
> 
> The fact that he hasn't done that means he *does* want *some* kind
> of intellectual property laws.

That's illogical.  I'd like to live in a world where there's no guns.
That doesn't mean I advocate disarming the cops in a world where the
criminals still carry guns.  

As a free software author, I'll consider stopping using copyright if
all the other software publishers out there are willing to do the
same.  Until then, I'll stay with the GPL.


From seb_haase at yahoo.com  Mon Apr  1 20:18:12 2002
From: seb_haase at yahoo.com (Sebastian Haase)
Date: 1 Apr 2002 17:18:12 -0800
Subject: embedded  (numeric -) python misses multiarray
Message-ID: <f445a2fe.0204011718.419c57d0@posting.google.com>

Hi,
I have an embedded python interpreter running just fine.
On windows using VC6++.

But when I'm trying  to 
import Gnuplot
(which uses Numeric) 
I get
Traceback (most recent call last):
  File "<pyshell#0>", line 1, in ?
    import Gnuplot
  File "C:\Python-2.2\Lib\SITE-P~1\GNUPLO~1.4\__init__.py", line 190, in ?
  File "C:\Python-2.2\Lib\site-packages\Numeric\Numeric.py", line 91, in ?
    import multiarray
ImportError: No module named multiarray

BUT THIS HAPPENS ACTUALLY ONLY if I compile my app in DEBUG mode
on release mode its running fine.

So -- where do I get some the file  multiarray.pyd
so that it runs with my app compiled in DEBUG mode (linked against  python2.2_d.lib)
???

Thanks,
Sebastian
haase at msg.ucsf.edu


From mlh at vier.idi.ntnu.no  Wed Apr  3 06:51:06 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Wed, 3 Apr 2002 11:51:06 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>  <mailman.1017799668.13959.python-list@python.org>
Message-ID: <slrnaalr5a.fa8.mlh@vier.idi.ntnu.no>

In article <mailman.1017799668.13959.python-list at python.org>, David
Ascher wrote:
[snip]
>
>     Something is True, and Nothing is False!

:)

>or is that, in the words of the late Milton Berle:
>
>     You can fool some of the people some of the time, and all of the
>people some of the time, and that's usually enough.

Isn't that "all of the time" in the first sentence? Otherwise it
doesn't make much sense (in that the first sentence would follow from
the second).

>--david

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From ajs at ix.netcom.com  Thu Apr  4 20:40:38 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Thu, 4 Apr 2002 20:40:38 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285
Message-ID: <001101c1dc42$e4427f40$0334fea9@carol>

"Strong and bitter words indicate a weak
    cause."  It reminded me of some of the posts against this
    PEP... :-)

Strong and bitter words:-}


Chinese proverb say they who dish, must also
take.

Art






From dalke at dalkescientific.com  Tue Apr 23 15:56:36 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 23 Apr 2002 13:56:36 -0600
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <aa4ei3$pir$1@nntp9.atl.mindspring.net>

Petr Prikryl:
>Being quite new to Python -- is there any preference of
>single or double quotes when writing strings?

In addition to the comments made by others, I tend to use
double quotes because they are more visible -- easier to spot.
Single quotes are harder to make out.

The counter example is for small strings.  In that case, 'a'
seems esthetically nicer than "a" because they don't dominate
the string.  It's also conviently similar to C's usage.

                    Andrew
                    dalke at dalkescientific.com





From aleax at aleax.it  Thu Apr 18 16:23:00 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 20:23:00 GMT
Subject: Counting the number of elements in a list - help.
References: <a9hu8e$2fg$1@lumberjack.rand.org> <MPG.1728bf07ecc579969896b3@news.nwlink.com> <20020418215424.28033abb.christophe.delord@free.fr>
Message-ID: <EWFv8.36351$vF6.1101881@news2.tin.it>

Christophe Delord wrote:
        ...
> It could be shorter and smarter like this :
> 
> def InRange(sequence, index):
>     return 0 <= index < len(sequence)

responding to:

>> > 2) a function which takes a list and an index and returns a zero when
>> > that list index is out of range.

No!  You're wrong.  -1 is in range for any non-empty list (it means the
last element), so why is your inRange function going to deny it?

FAR better Jeff Shannon's approach:

>> def InRange(sequence, index):
>>     try:
>>         sequence[index]
>>     except IndexError:
>>         return 0
>>     else:
>>         return 1

This lets *the sequence object* decide what indices are in or not-in
range.  By far the best approach.

If you insist that you know better than the object which indices are
in-range for it, at least use a body such as:

        thelen = len(sequence)
        return -thelen <= index < thelen

but no "checking" approach is going to be as complete and reliable
as try/except!


Alex
        


From fperez528 at yahoo.com  Mon Apr 22 20:38:32 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 18:38:32 -0600
Subject: Proper syntax for adding methods to a class instance dynamically?
Message-ID: <aa2ahm$8pr$1@peabody.colorado.edu>

I'd like to know what the proper way to do this is. I've been using the new 
module with new.instancemethod(), but the docstring for new isn't very 
confidence inducing:

Docstring:
    Functions to create new objects used by the interpreter.

    You need to know a great deal about the interpreter to use this!


Is there a simpler, perhaps 'safer' way to achieve the same result? What I 
want is to have code which runs after an instance of a certain class has been 
created and which can add new methods to the class instance. I'd also like to 
know how to add them to the whole class. I have some solutions in store but I 
feel they're kind of kludgy.

Thanks for any help,

f.


From chris at psychofx.com  Fri Apr 12 14:19:46 2002
From: chris at psychofx.com (Chris Miles)
Date: Sat, 13 Apr 2002 04:19:46 +1000
Subject: SMS Server implemented in Python
In-Reply-To: <2D4B57FE40B5FE2A.08DB31F492EEFF0B.F76D2BD5D7AAECF5@lp.airnews.net>; from claird@starbase.neosoft.com on Thu, Apr 11, 2002 at 09:42:19AM -0500
References: <k1ft8.1694$ws6.37396@news2.nokia.com> <2D4B57FE40B5FE2A.08DB31F492EEFF0B.F76D2BD5D7AAECF5@lp.airnews.net>
Message-ID: <20020413041946.B28038@psychofx.com>

You could look at Pagey, which is a simple SMS server written entirely in
Python.  See if that suits your needs.

http://psychofx.com/pagey/

Regards,
Chris.


On Thu, Apr 11, 2002 at 09:42:19AM -0500, Cameron Laird wrote:
> He's likely to tell you that AT&T hasn't
> and won't make the server public, but that
> he knows another one that is.




From siegfried.gonzi at kfunigraz.ac.at  Fri Apr 26 04:27:22 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Fri, 26 Apr 2002 10:27:22 +0200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org>
Message-ID: <3CC90F6A.A2416DCB@kfunigraz.ac.at>

"James J. Besemer" wrote:

> So now compare
> 
>         if x > y or a = 0:
>             a1 = m * x + b
>         else:
>             a2 = m / x - b
> 
> with
> 
>         (if (or (> x y ) (zerop a))
>             (setq a1 (+ (* m x ) b )
>             (setq a2 (- (/ m x ) b )).
> 
> In this case, Lisp requires about 33% more typing, almost all parentheses.
> 
> And, oh yeah, I forgot, aside from the parentheses, the pervasive prefix notation can be a
> drag.  And I'm the type that still finds HP's stack notation less confusing in a
> calculator than infix.  But having to mentally convert everything back and forth between
> infix and prefix notation also can make you nuts.

I have to beg differ. Though my stance actually is that when one does
not like any style or paradigm then he must not excuse his behavior,
which becomes often required by some advocates; e.g., I hate vim or
Emacs but I do not have to defend my position to anyone. Programming
languages are like Coca-Cola/Pepsi.

I used Lisp a few month (before I encountered Python) and I had never
any problems with infix/prefix or parentheses. I was even very suprised
that parentheses did not come in the way which some of us want to make
believe us. Okay, I have to admit that my programs seldom surface the
1000LOC threshold.

In the meantime I have got a rather pragmatic view of programming
languages: libraries. I am the first who would vote for Clean or even
Lisp when it comes to some "elegance", but they all completely fail when
it comes to useful libraries. Neither Clean nor Common Lisp delivers me:
DISLIN and the Numeric library. Sure, it would be a tiny deal for an
expert to implement it in Clean or Common Lisp, but this is only
hypothetical and does not happen.


Call me ignoarant, but here Python or even Perl are really light years
ahead of Common Lisp.


S. Gonzi


From aaron at reportlab.com  Mon Apr 22 18:05:16 2002
From: aaron at reportlab.com (Aaron Watters)
Date: 22 Apr 2002 15:05:16 -0700
Subject: calendar.py doesn't use lists???
Message-ID: <9a6d7d9d.0204221405.7d394207@posting.google.com>

Hi guys.  I got an irksome problem.  It's possible that
it's fixed in python 2.2.1, but even if it is, it irks me.

I tried the following script:

"""
import sys
print sys.version
if 1:
    from calendar import month_abbr, month_name
    # hey python guys, i want lists!!
    for x in (month_abbr, month_name):
        print type(x)
    month_abbr = list(month_abbr)
    month_name = list(month_name)
    print month_abbr
    print month_name
"""

Using python 2.1 I get this output

"""
2.1.1 (#20, Jul 26 2001, 11:38:51) [MSC 32 bit (Intel)]
<type 'list'>
<type 'list'>
['   ', 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'N
ov', 'Dec']
['', 'January', 'February', 'March', 'April', 'May', 'June', 'July', 'August', '
September', 'October', 'November', 'December']
"""

(completing normally)

Using Python 2.2 I get this

"""
2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)]
<type 'instance'>
<type 'instance'>
[CRASH AND BURN... WINDOWS XP SUGGESTS I SEND A COMPLAINT TO MICROSOFT...]
"""

Of course it may be that this is fixed in python 2.2.1 but...

WHO'S GOOFY IDEA WAS THIS?

Seriously, why did you need to change these sequences from
lists where they worked with my code, to somethingsimilartolists
that broke my code?

Irked.  In general I've noticed a tendancy with recent python
releases towards changing stuff for the fun of changing stuff,
I'm sorry to say.  Please remember that when you fix something
that isn't broken, you might just be breaking somebody else's code.

Sorry to whine.

   -- Aaron Watters

===
If there is a god, he or she sure liked to design
beetles.
  -- attribution forgotten


From DeepBleu at DeepBleu.org  Wed Apr 24 10:59:53 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Wed, 24 Apr 2002 09:59:53 -0500
Subject: (no subject)
References: <mailman.1019632185.27320.python-list@python.org>
Message-ID: <ucdi2nd48euif8@corp.supernews.com>

An example of Zen Posting.  Silent and to the point :o)
DeepBleu

"Yaser Ansari" <yaser at lambenttek.com> wrote in message
news:mailman.1019632185.27320.python-list at python.org...





From bellman at lysator.liu.se  Wed Apr 10 15:23:44 2002
From: bellman at lysator.liu.se (Thomas Bellman)
Date: Wed, 10 Apr 2002 19:23:44 +0000 (UTC)
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204100109.4a44932b@posting.google.com> <RYTs8.2219$b62.57325@news1.tin.it> <23891c90.0204100552.6b822307@posting.google.com> <C%Xs8.3139$b62.77265@news1.tin.it>
Message-ID: <a923g0$ks2$1@news.island.liu.se>

Alex Martelli <aleax at aleax.it> wrote:

> Once you're using (e.g.) iterators in your package, or subclassing
> a standard class, there's no extra cost in having the modules of
> your package use (e.g.) generators or true-division as well, so
> why shouldn't a package author freely do so?

For the specific case of iterators, it is fairly easy to add a
backwards compatible interface to your classes, so they work even
in an older interpreter:

    try:
	StopIteration
    except NameError:
	class StopIteration(Exception):
	    pass

    class IteratorEmulator:
	def __getitem__(self, i):
	    if i == 0:
		self.__iterator_index = 0
		self.__iterator = self.__iter__()
	    if self.__iterator_index != i:
		raise IndexError("not indexed sequentially")
	    self.__iterator_index = self.__iterator_index + 1
	    try:
		return self.__iterator.next()
	    except StopIteration:
		raise IndexError("index out of range")

Then just let your classes inherit from IteratorEmulator.  This,
of course, assumes that you don't need __getitem__() for other
purposes, or that you are returning infinite sequences.

Generators, inheriting from builtin types, and true division, are
not as easy to get running in an older interpreter, so if you are
using one of *them*, you might as well use iterators too.


-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"When C++ is your hammer, everything         !  bellman @ lysator.liu.se
 looks like a thumb."                        !  Make Love -- Nicht Wahr!


From joost_jacob at hotmail.com  Tue Apr 30 07:03:12 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 30 Apr 2002 04:03:12 -0700
Subject: Tkinter Text widget reading input (how to do overwrite mode?)
References: <13285ea2.0204290624.4fbe8cba@posting.google.com> <20020429100034.A12526@unpythonic.net> <mailman.1020104511.3714.python-list@python.org>
Message-ID: <13285ea2.0204300303.1343dc44@posting.google.com>

Thank you !
Frankly i do not understand all of the code, maybe you could put
comments in for Tk newbies? :-), but a quick test here did show it
worked.  With this i think i can build a readxy() function as i
described in the earlier post with 'overwrite-mode' as i intended.


From tdelaney at avaya.com  Mon Apr 22 20:32:41 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 23 Apr 2002 10:32:41 +1000
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A286@natasha.auslabs.avaya.com>

> From: Michael Gilfix [mailto:mgilfix at eecs.tufts.edu]
> 
>   void both_ways (int val, int direction)
>   {
>     another_func ( (direction) ? NULL : &val,
>                    (direction) ? &val : NULL,
>                     ....);
>   }

Urgh. Yuck. Ugly. I have *never* done this, in any language. I will always
compose the arguments correctly first, then call the function.

>   That way you can easily switch arguments to another function without
> more typing. But most people use the ternary operator as an if
> statement, which is ugly.

As is the above. The only use I have ever had for the ternary operator is of
the form:

    void both_ways (int val, int direction)
    {
        int *arg1 = direction ? NULL : &val;
        int *arg2 = direction ? &val : NULL;

        another_func(arg1, arg2, ...);
    }

Isn't that much cleaner and understandable than the above?

>   While I'm not commenting on whether this is really useful in python
> (because I get along fine without it), the python equiv would be:
> 
>   def both_ways (arg, dir):
> 
>     if dir:
>       arg1 = blah
>       arg2 = some_blah
>     else:
>       arg1 = something_else
>       arg2 = some_other_thing
> 
>     another_func (arg1, arg2)

I would rewrite this as:

def both_ways (val, direction):

    if dir:
        arg1, arg2 = None, val
    else:
        arg1, arg2 = val, None

    another_func(arg1, arg2, ...)

The thing is, the ternary operator is only more readable when dealing with a
single condition and single results. Adding even a single expression in
there starts making it unreadable. Any time things get more complicated, you
need to move to the if () {} else {} format anyway.

Tim Delaney




From aleax at aleax.it  Tue Apr  2 02:26:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 07:26:08 GMT
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <a85od6$9i$1@panix1.panix.com> <DlTp8.48032$S52.1614804@news2.tin.it> <a8ajpi$i36$1@lust.ihug.co.nz>
Message-ID: <k2dq8.50144$S52.1717644@news2.tin.it>

Gary Stephenson wrote:
        ...
> an air of artificiality about them, there are numerous instances in the
> real world where things _do_ mutate over time, and belong to varying sets
> of
> categories as a consequence of such mutation.   

Sure.  It's easy to see that any change removes membership of an
uncountably infinite number of sets and adds membership ditto.  We
can't model that perfectly (infinities are hard to model, uncountable
ones in particular) but we can come decently close (for most sets of
interest) by modeling each set by its characteristic predicate.

> Given that our objects
> _are_ allowed to mutate their state, wouldn't it be preferable for them to
> do so in more intelligent ways?  

No.  "Change" is such a troublesome notion already, that to have it
train behind itself a "keeping track of uncountably infinite sets" is
quite a conceptual mountain.

> Whether such a notion is actually
> implementable in the real world is an entirely different question  of
> course ;-)

Objects that do not change are perfectly well implementable and quite
usable.  Anybody looking for a closer match to maths should IMHO turn
right to lazy FP (Haskell...) rather than tilt agains the windmill of 
'proper' modeling of change.

> Simple immutability is also insufficient to enable the subclass == subtype
> identity [1], which seems to me to be one of the fundamental issues
> underlying all this.

Yes.  If you realize that 'behavior' is never intrinsic to a set of values, 
and systematically _superimpose_ behavior on value sets (Haskell's
typeclasses being of course the right way to do that:-), then you complete
the conceptual trip to systematic purity.  Then you can decide whether
the pragmatics of Python are preferable to the purity of Haskell, but if
you do then it's a peculiar waste of time to keep yearning for purity after
choosing the compromise route of pragmatism.

The set/bag example under immutability depends on functions such as
put and remove returning new sets/bags -- a conceptual approach that
simplifies things a lot, even though implementation-wise it may be too
heavy.  But the insights you gain from there are still valuable and to some
extent applicable to impure worlds with mutating objects.  Among bag's
axioms (w has/put/remove) should be something like the material implication:
    (has item bag) => (has item (remove item (put item bag)))
and among set's axioms w has/put/remove should be:
    () => (not (has item (remove item set)))
so the impossibility of any one entity X satisfying both sets of axioms
is trivial to prove, no?


Alex



From aleax at aleax.it  Sat Apr 27 18:35:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 22:35:59 GMT
Subject: What is NotImplemented ?
References: <slrnacinls.j5.alf@virgo.logilab.fr> <lkn0vq34fi.fsf@pc150.maths.bris.ac.uk> <slrnacm7li.mru.mlh@vier.idi.ntnu.no>
Message-ID: <jJFy8.28401$8D3.758727@news1.tin.it>

Magnus Lie Hetland wrote:
        ...
> But how is this different from raising NotImplementedError? (If that
> doesn't work in this case, why doesn't it?)

Because raising NotImplementedError means "this method is abstract
in this class, it MUST be overridden by concrete subclasses".  That's
quite different from "try alternate ways" such as:

>>Returning NotImplemented from the __add__ makes Python go off and call
>>the __radd__ method.


Alex



From deltapigz at telocity.com  Thu Apr 25 10:12:15 2002
From: deltapigz at telocity.com (Adonis Vargas)
Date: Thu, 25 Apr 2002 10:12:15 -0400
Subject: XML and Python
Message-ID: <3cc812a3$1_1@nopics.sjc>

i am just starting my venture into XML as a means of document output and
soon as a form of internet protocols. and i would like to know how i can
parse XML in Python? now i know i have access to SAX and DOM, but how do i
use them? i have looked into the documentation, and it seems to be written
in a martian language, and i am looking in doing basic stuff just regular
nodes, in particular. any link/sample usage just anything to get my feet wet
in this subject, im a decent Python programmer, so there shouldnt be too
much problems, just need something i can work with.

any help would greatly be appreciated.

Adonis




From aleax at aleax.it  Mon Apr 29 03:37:51 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 07:37:51 GMT
Subject: inline assignments in conditionals
References: <mailman.1020042769.15036.python-list@python.org>
Message-ID: <jL6z8.75063$vF6.2243835@news2.tin.it>

Delaney, Timothy wrote:
        ...
>> >    "If Sunday is today, I'm going to go shopping"
>> 
>> I don't think that anybody has had any trouble understanding me when I
>> say 'If today is Saturday, then we are going to the concert house,
> 
> Not at all - this is exactly how I say things.
> 
>> otherwise we are going to hack.'  But then I have been writing
>> if 0 == x for a long time as well.
> 
> The one doesn't follow the other. "If today is Sunday" corresponds to "if
> x is 0" - "today" is the variable, "Sunday" is a constant.

Take another typical English usage then.  "If Bush is the President of the
US ...".  Which one is the "variable" now, and which one the "constant"?
Clearly, who is President varies over time, just as what is 'today', while
'Bush' in this context refers to a specific individual (it may be ambiguous
which one, of course:-).

The distinction in connotation is not variable vs constant -- it is about
what term is supposed to be in doubt:

    "If Bush is the President of the US ... " (rather than being, e.g.,
        the President of General Motors) 

    "if the President of the US is Bush ... " (rather than being, e.g.,
        Gore)

Actually my feeling is that the first form would be quite acceptable
in the second connotation as well, while the latter form would be
totally unacceptable in the former connotation (i.e., the latter form
unambiguously indicates we're musing about different individuals
covering the "President of the US" role; the former form might
indicate that, too, although it's quite OK to indicate the different
case of musing about different roles for one given individual).  But
I doubt i can "prove" it.  Natural language just doesn't fit well into
hard-and-fast formulas, despite the unending efforts of grammarians,
pundits, and purists -- it IS, after all, part of the biology and
sociology of human beings, notoriously funny and fuzzy subjects.
(Which explains why it's so unappropriate as a basis for designing
programming languages, but that's an issue deserving longer discourse).


Yours for descriptive-rather-than-prescriptive linguistics,

Alex



From Patrick.Bussi at space.alcatel.fr  Mon Apr 22 04:58:01 2002
From: Patrick.Bussi at space.alcatel.fr (Patrick.Bussi at space.alcatel.fr)
Date: Mon, 22 Apr 2002 10:58:01 +0200
Subject: HTTP error on receiving a form
Message-ID: <C1256BA3.00316CEA.00@vzmta01.netfr.alcatel.fr>

Hello

I am implementing a simple tool to answer post requests for automatic testing of
a Web APACHE Server (implemented by other people). It is rather easy with use of
something like:

dataOfTheForm= {"data1":"val1","data2":"val2"[...etc...]}
encodedData=urllib.urlencode(dataOfTheForm)
f=urllib.urlopen(webSite,encodedData)

However, the web server find errors in the data received from my tool, while all
is OK with a web browser. I have TCPdumped answers both from my test tool and
from an ordinary web browser (say Mozilla). As far as I could understand, the
only difference is the order in which data are encoded.

My questions:

1 - is it possible that a web Server be sensitive to the order in which data are
encoded (as they are all sent in one single IP packet)

2 - is there a quick way to encode data from a dictionnary in a specific order.


Thank you for any clue.


---
Patrick Bussi
patrick.bussi at space.alcatel.fr


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






From aahz at pythoncraft.com  Sun Apr 21 14:23:26 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 14:23:26 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org>
Message-ID: <a9v02u$lal$1@panix3.panix.com>

In article <mailman.1019412399.4283.python-list at python.org>,
holger krekel  <pyth at devel.trillke.net> wrote:
>On Sun, Apr 21, 2002 at 01:35:09PM -0400, Aahz wrote:
>> In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
>> Allan Crooks <googlegroups at sixtyten.org> wrote:
>>>
>>>Another thought I've had is doing it by specifying, say for module
>>>'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
>>>what versions of Python it is running on, and then compiling all
>>>modules it can deal with, and then importing all objects into it's
>>>namespace.
>> 
>> Yup.  But do it with a package.
>
>But doesn't this lead to lots of duplicated -hard to maintain-  code?

Depends.  You can certainly construct your package to use common code.
But if you're truly set on using the idioms of each Python version to
the full extent while still providing backwards compatibility, you're
going to have a problem with code duplication.  Period.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From joonas at olen.to  Tue Apr  9 12:53:58 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Tue, 09 Apr 2002 16:53:58 GMT
Subject: Sending a signal to a process started by another user
Message-ID: <3CB31C2C.9010503@olen.to>

How can I send a signal to a process started by another user. In this 
case my intention is to send a signal 14 to a process started by user jp 
from a cgi script. The script raises an "Operation not permitted"- 
OSError perhaps because the script is run by user "apache" and the 
process by user "jp". How can In make this work properly?

- Joonas



From lac at strakt.com  Fri Apr 26 18:56:55 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 27 Apr 2002 00:56:55 +0200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...) 
In-Reply-To: Message from Paul Rubin <phr-n2002a@nightsong.com> 
   of "25 Apr 2002 23:10:47 PDT." <7xsn5jat20.fsf@ruckus.brouhaha.com> 
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz>  <7xsn5jat20.fsf@ruckus.brouhaha.com> 
Message-ID: <200204262256.g3QMuukG009495@ratthing-b246.strakt.com>

People here seem to be thinking that the GPL is there to protect
creators rights to their software.  As a thought experiment, consider
that the GPL is instead intended only to benefit the purchasers of
software.  When the damn thing breaks, the purchaser can pay somebody
to fix it.  This is a consumer-protection issue - akin to to the
requirement that the ingredients in food be listed on the outside.
You can still go buy the can whose first ingredient is sugar and
whose next 4 are artificial flavours if you like, just as you can
buy buggy software written by idiots if you like as well.  The thing
is you will be able to tell (or hire somebody to tell you) before
you buy -- and if the damn thing breaks there will be something that
you can do about it besides hope that the seller a) is competant enough
to fix things and b) gives a damn.

Laura Creighton




From jimmy at retzlaff.com  Tue Apr 30 16:54:25 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 30 Apr 2002 13:54:25 -0700
Subject: Newbie: finding the key/index of the min/max element
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F565B@MAYOR.retzlaff.com>

Benjamin Han [mailto:bhan at andrew.cmu.edu] asked:
> how do you find the key/index of the max/min element in a
sequence/map?

In Python tuples are compared element-wise from first to last. So, e.g.:

>>> min((1, 2, 3), (1, 3, 2))
(1, 2, 3)

So what you need is to represent your dictionary as a list of tuples
with the values first and keys second. Here is one way to do that in
recent versions of Python:

>>> d = {'one' : 1, 'two' : 2, 'three' : 3, 'four' : 4, 'five' : 5}
>>> invertedList = zip(d.values(), d.keys())
>>> invertedList
[(4, 'four'), (3, 'three'), (5, 'five'), (2, 'two'), (1, 'one')]

Then min/max can operate on that list:

>>> min(invertedList)
(1, 'one')
>>> max(invertedList)[1]
'five'


Jimmy




From skip at pobox.com  Mon Apr 15 09:25:32 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 15 Apr 2002 08:25:32 -0500
Subject: Question about httplib and HTTP response 100 Continue
In-Reply-To: <1018837561.612.2.camel@prowler>
References: <1018837561.612.2.camel@prowler>
Message-ID: <15546.54476.895761.105912@12-248-41-177.client.attbi.com>

    Marco> So, does the httplib work when it encounters a 100-Continue
    Marco> response?

I suggest you try recoding your example using the HTTPConnection class.  I
believe the HTTP and HTTPS are now included only for backward compatibility
(though I think they've yet to be deprecated).

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From s328940 at student.uq.edu.au  Mon Apr 29 06:32:13 2002
From: s328940 at student.uq.edu.au (Alejandro Dubrovsky)
Date: Mon, 29 Apr 2002 20:32:13 +1000
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <3CCC730A.FC60D13F@alcyone.com>
Message-ID: <3ccd1ed9$0$15477$afc38c87@news.optusnet.com.au>

On Mon, 29 Apr 2002 08:09:14 +1000, Erik Max Francis wrote:

> Alejandro Dubrovsky wrote:
> 
>> i've got a class that defines something like the following:
>> 
>> def t(self,a,b):
>>         ...
>> 
>> def h(self):
>>         self.thelambdafunctiontobepickled = lambda x: self.t(x,4)
>> 
>> is there any simple way of pickling that lambda function above so that
>> when i unpickle it will be set to the right function?
> 
> You can't pickle a lambda function.  However, you could for instance
> pickle a string representation of it that when eval'd gives the desired
> lambda.
 
yes, but that means saving the definition as a string everytime that the
lambda is changed.  i was looking for something "neater". eg for function
variables, i can just save and restore the variables'.__name__ attribute.
i was wondering if there was something similar for lambdas.
Thanks,
Alejandro


From marklists at mceahern.com  Wed Apr  3 23:07:48 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 22:07:48 -0600
Subject: OK
In-Reply-To: <c7knauo4dqpsipk3rdhg5jhcaqvkna4pat@4ax.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKECCKJAA.marklists@mceahern.com>

> On 3 Apr 2002 18:16:14 -0800, huangcg at 163.net (huangcg) wrote:
> 
> Is python out of fashion?

The latest Cosmo had something about snake charmers--does that help?

// m




From m_mariappanX at trillium.com  Tue Apr 23 03:05:59 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Tue, 23 Apr 2002 00:05:59 -0700
Subject: A Q on using one of the Pmw widget Dialog
Message-ID: <53A7943A5BD8D411B6930002A5073155022D746E@bgsmsx90.iind.intel.com>

Hi all,

A Q on using one of the Pmw widget Dialog



From greg at cosc.canterbury.ac.nz  Mon Apr 15 01:31:01 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 15 Apr 2002 17:31:01 +1200
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
References: <mailman.1018755873.29647.python-list@python.org>
Message-ID: <3CBA6595.104ED599@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> As
> originally posted, they were the "20 Pythonic Theses".  They never need
> revision, because I only gave 19 -- the 20th (or perhaps the first -- they
> weren't numbered) was left open for Guido to fill in.

Perhaps it should be "19 Pythonic theses is sufficient."

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


From aleax at aleax.it  Sat Apr 13 18:24:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 22:24:45 GMT
Subject: cStringIO write
References: <QHxVGIALCDu8EwLs@jessikat.fsnet.co.uk> <krYt8.13007$b62.354393@news1.tin.it> <fTIWTKAuNHu8Ewq$@jessikat.fsnet.co.uk>
Message-ID: <Ne2u8.13850$b62.382109@news1.tin.it>

Robin Becker wrote:
        ...
> strange in my docs I see no mention of these special types. I see for

You may not be looking at the latest version of the docs, at

http://www.python.org/dev/doc/devel/lib/module-cStringIO.html


Alex



From aleax at aleax.it  Mon Apr  8 07:36:10 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 11:36:10 GMT
Subject: Bitwise/Binary operations in Python < 2.1
References: <slrnab2sb5.j99.petro@lists.bounty.org>
Message-ID: <Kgfs8.59392$pT1.1709038@news1.tin.it>

petro wrote:
        ...
> * priorities/facilities are encoded into a single 32-bit quantity, where
> the * bottom 3 bits are the priority (0-7) and the top 28 bits are the
> facility
> * (0-big number).  Both the priorities and the facilities map roughly
> * one to strings in the syslogd(8) source code.  This mapping is
> * included in this file.
> 
> The problem is, well, basically I'm pretty stupid, and can for the life
> of me figure out how to do this in such a way that it's portable from
> Python 1.5.2 to 2.x

what about:

def extractPriorityAndFacility(single32bitQuantity):
    priority = single32bitQuantity & 7
    facility = (single32bitQuantity >> 3) & 0xFFFFFFF
    return priority, facility



Alex




From djc at object-craft.com.au  Thu Apr 18 04:50:44 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 18 Apr 2002 18:50:44 +1000
Subject: chopping csv-files (newbie)
References: <3CBE768C.844C90FF@itea.ntnu.no>
Message-ID: <m3it6p5r17.fsf@ferret.object-craft.com.au>

> I have a csv-file.. delimited by ;
> example-line from the csv-file
> "90383";"ooooxWOW   90383;";"Some discription";
> "99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";
> 
> In this file there is 3 fields extracted from a database.
> The second field, can vary som 1 to 4 fields separated by ; but not with
> "
> 
> What I need to do is to build an array after opening a given file
> with the information from the second and third fields.
> 
> E.g:
> 
> 90383	Some discription
> 99875	New line-descriptor, Some discription
> 
> You see where I'm going at?
> 
> the second record in this file has a reference to the first record in
> its second field.

You could use my CSV parser:

>>> import csv
>>> p = csv.parser()
>>> p.field_sep = ';'
>>> p.parse('"90383";"ooooxWOW   90383;";"Some discription";')
['90383', 'ooooxWOW   90383;', 'Some discription', '']
>>> p.parse('"99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";')
['99875', 'ooooxSomede90383;0000xNewli99875;', 'New line-descriptor', '']

The module is available at:

        http://www.object-craft.com.au/projects/csv/

- Dave

-- 
http://www.object-craft.com.au


From gerhard at bigfoot.de  Mon Apr  8 10:56:01 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 8 Apr 2002 14:56:01 GMT
Subject: Adjuntar ficheros en python.
References: <mailman.1018265955.20370.python-list@python.org>
Message-ID: <slrnab3bs2.f1.gerhard@lilith.my-fqdn.de>

I won't pretend to be able to *write* Spanish, so I'll answer in
English.

jmanzano at antena3tv.es wrote in comp.lang.python:
> Hola
> 
> Alguien me puede decir como puedo adjuntar ficheros en python a la
> hora de enviar emails.

How do I attach files in Python when sending emails?

If you're using Python >= 2.2, simply use the email module and look
this up in its documentation, which has examples for exactly that.

For Python 2.1, this module is available seperately in the mimelib
project at Sourceforge.

For even older Pythons, you'll have to play some tricks with the
mimetools package yourself. I've seen answers on comp.lang.python for
that, so searching groups.google.com for "email mime attachment"
should give some options.

HTH,

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From sholden at holdenweb.com  Mon Apr 15 13:22:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 13:22:10 -0400
Subject: Python 2.2 unwise.exe
References: <mailman.1018734147.26303.python-list@python.org>
Message-ID: <c1Eu8.13834$Ji7.6048@atlpnn01.usenetserver.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1018734147.26303.python-list at python.org...
> [Robin Becker, on unwise.exe vanishing]
> > Installed under C:\Python; system is Win2k. I installed win32all-146
> > after Python-2.2, but didn't save backups.
> >
> > I have a couple of other things installed eg Numeric-20.3 and PIL, but
> > nothing  that came with an installer. I don't remember anything untoward
> > about the Python-2.2 install.
> >
> > The unwise.exe was just not there for some reason.
>
> Crud, I was hoping it was unique to Win9x.  "Just not there for some
reason"
> is the same thing I've seen.
[ ... ]
Oh, crap. Same thing's happened on my Win2k installation. I habitually run
Windows Explorer with "show hidden files", so I know it's really disappeared
and isn't just in hiding.

Perhaps we're being unwise relying on this software? I wonder whether some
other uninstall of a Python related component might be causing this?

regards
 Steve






From alanmk at hotmail.com  Fri Apr  5 04:12:46 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 5 Apr 2002 01:12:46 -0800
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com>
Message-ID: <f72dac0d.0204050112.11d4ef89@posting.google.com>

brian d foy <comdog at panix.com> wrote:

> for our April 2002 issue (April Fool!), The Perl Review replaced its
> Perl content with Python articles, so Python folks might want to read
> this issue:
> 
>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf

Brian,

Thanks for that! I learnt a load about python I never knew before.

What a refreshing take on April 1st: Look at the world from another point of view.

Hopefully it will catch on everywhere :-)

Cheers again,

Alan Kennedy.


From pmullh at yahoo.com  Tue Apr 30 13:31:36 2002
From: pmullh at yahoo.com (PM)
Date: 30 Apr 2002 10:31:36 -0700
Subject: Access violation trying to run Python over Win NT network.
References: <ucrs88tkj2ttc7@corp.supernews.com> <Xns920058DCEAB82gmcmhypernetcom@199.171.54.215>
Message-ID: <4be6f063.0204300931.530a0cda@posting.google.com>

Gordon McMillan <gmcm at hypernet.com> wrote in message news:<Xns920058DCEAB82gmcmhypernetcom at 199.171.54.215>...
> Patrick wrote:
> 
> > I'm trying to run Python from a network drive.  On most machines it
> > works, but on some machines it generates an access violation about 60%
> > of the time, (just by typing 'python' with no arguments).  If I install
> > Python locally it runs fine on all the machines.
> 
> One possibility is that the install is updating some MS redistributable
> dll. This goes into the server's Windows\System. The client is still
> using an out of date copy from the client machine's Windows\System.
> 
> That would explain why it doesn't fail on Win2K boxes (as you mention
> in a later post).
> 

I am currently using pythonware's python 2.1 on the network drive.  I
believe it does not change anything in the registry or add/modify any
system dlls.

I was originally using the www.python.org 2.2.1 distribution with
py2exe.  When I started getting the access violation, I switched to
your installer, but the error was the same.  So then I switched to
just trying to run python itself over the network.  That's when I
started using pythonware's distribtution for the reason that it
doesn't modify your system.  Still get the error.

This morning I tried running pythonware's python on the network using:

Win 2000 drive on machine 'A': worked fine
Win NT drive on machine 'A': access violation
Same Win 2000 drive on machine 'B': worked fine
Same Win NT drive on machine 'B': worked fine

Python works when run locally on all machines.

Thanks,
Patrick


From sixtysix at inwind.it  Thu Apr  4 10:52:11 2002
From: sixtysix at inwind.it (Massimo Valentini)
Date: Thu, 4 Apr 2002 17:52:11 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk>
Message-ID: <eR_q8.117876$%g7.3768326@twister2.libero.it>

"Max M" wrote
: A function that generates a relative path given two absolute urls?
:
: Not that I am aware of, but this problem was more fun than what I really
: should be working on today :-/, so here is an attempt:
:
: regards Max M
:
: -----------------------------------------------------
:
: from urlparse import urlparse, urljoin
:
:
: def relativePath(url1,url2):
:      "Returns relative path from url1 to url1"
...
:
:      diffStart = 0 # from where do the paths differ?
:      for i in range(min(u1Len, u2Len)):
:          if u1Parts[i] == u2Parts[i]:
:              diffStart += 1
...
: url1 = "http://foo/a/b/c"
: url2 = "http://foo/a/b/c"
: print "Excpect: ''"
: print repr(relativePath(url1,url2))
: print ""
...

I would suggest to consider even a case like this:

url1 = "http://foo/a/b/images"
url2 = "http://foo/a/b/c/d/images"
print "Expect: '../../../images'"
print repr(relativePath(url1,url2))
print ""


Massimo




From hgs at dmu.ac.uk  Thu Apr 18 08:03:30 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 13:03:30 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4662p8d5u.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181249260.24683-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
>
>
> > > Actually, you *can* talk the Solaris linker to incorporate non-PIC
> > > text into a shared library. For that, it will make the text segment
> > > writable. The -mimpure-text option of gcc triggers that feature.
> >
> > OK, but shouldn't this sort of information be in the configure script?

I wrote that, actually.
>
> No. Using -mimpure-text has undesirable side effects, such as making
> the text segment writable (which in turn means that the shared library
> won't be shared across address spaces).

OK, for that case I agree. I was thinking that the more arcane aspects
of this are just those that the human installer will not want to track
down themselves...
>
> There is only so much that configure can find out automatically. If
> you have a slightly broken system, or one that lacks certain features,
> you need to give explicit instructions to your tools.

Yes.  Does this mean that my Solaris2.7 system is broken, or is this a
problem with Solaris2.7 generally?   If the latter, then I believe that
configure should either fail in a more informative manner, or do
something about the problem, a belief based on suns being widely used.
If I could do the more constructive thing of providing a fix then I would.
>
> > But seriously, Solaris is quirky, but no one contributes the appropriate
> > patches to CVS... (I am not going to, I have other things to worry about
> > presently).
>
> Patches that propose to replace -shared with -G when using gcc will be
> rejected, as they are plain wrong. Patches that always add
> -mimpure-text will be rejected as the cause performance degradation
> for extension modules that don't need relocations in their text
> segment.

Agreed, patches that do the wrong thing should be avoided.
>
> > > If you link static libraries into Python, I recommend to make the
> > > modules that use them also static. Then you don't need PIC code for
> > > those.
> >
> > Wouldn't this make the memory image of all python programs much larger?

I wrote that line too...
>
> What are "all python programs"? There is only one Python
> executable.

All programs written in Python, which use the said executable,
regardless of whether they use all the modules or just some.

>
> If you are concerned about the memory consumption of python processes:
> Since Solaris uses demand-paged memory, code of extensions won't be
> loaded into memory until somebody touches the page that has the code,
> which usually won't happen until the module is imported.

OK, that is reassuring.
        [...]
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you,
        Hugh





From ianb at colorstudy.com  Mon Apr 15 16:46:10 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 15 Apr 2002 15:46:10 -0500
Subject: String interpolation question
In-Reply-To: <a9fcmo$npi$1@peabody.colorado.edu>
References: <a9fcmo$npi$1@peabody.colorado.edu>
Message-ID: <1018903571.1147.395.camel@localhost>

On Mon, 2002-04-15 at 15:19, Fernando P?rez wrote:
> but I don't want that because in my real problem, the string is long and I 
> don't want the dangers of positionally matched %s arguments.

Then do 
  print somestring.replace("%", "%%") % y[i]

  Ian






From sholden at holdenweb.com  Mon Apr 15 15:00:52 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 15:00:52 -0400
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <a9apoc$k52$1@nntp9.atl.mindspring.net>
Message-ID: <LtFu8.15762$Ji7.1848@atlpnn01.usenetserver.com>

"Dennis Lee Bieber" <wlfraed at ix.netcom.com> wrote ...
> On Tuesday 09 April 2002 14:45, a penguin (Henning Peters) squawked:
>
> > is there anything similar to the pointer concept from c/c++ in python?
> >
Indeed, but it's mostly transparent, and therefore hardly visible.

> > i would like to store values in two different locations. both values
> > should always be equal. the one could be a pointer on the other. is
> > this possible to do this in python?
> >
>         You first have to understand the differences between variables in
> traditional languages and variables in Python. I'm going to go back to
> "Intro to Computing 101", circa mid-70s for this.
>
>         Have you ever encounted the "mailboxes" example for
variables/memory?
> IE, the old sorting grid of boxes (you need a fixed width font here):
>
>         +--+--+--+--+--+
>         |  |  |  |  |  |
>         +--+--+--+--+--+
>         |  |  |  |  |  |
>         +--+--+--+--+--+
>
>         In the traditional language, each of those boxes has an address
> (symbolically, say, A, B, C, D, E, ...).
>
Let me say right from the start that I've always thought this analogy was
very weak for languages like Python and Icon, where assignment/rebinding
effectively implements homonyms.

It's much more effective for assembly languages, for example.

>         A statement in those languages of:      A = B
> in effect says: take what ever is in box B, make a copy of it (leaving
> the original in box B) and put the copy in box A.
>

>         In Python, the boxes do not have fixed addresses. Instead they
have
> Post-It notes stuck to the front (one note with A on it, one with B,
> etc.). The statement:   A = B
> in Python is the equivalent of taking the Post-It note with "A" on it,
> and sticking it onto the front of whatever box currently has the Post-It
> note with "B". If the box that used to have the "A" label now has no
> Post-It's attached, the contents can be garbage-collected and disposed
> of.
>
It seems to make much more sense (to me) to think of a name in a Python
program being associated with a value, and the associations being made by
binding/rebinding. I've explained this in the past in terms of namespaces
vs. object spaces.

>         Later, if you do something like:        B = 3047302
> Python creates a piece of paper with "3047302", then looks for an empty
> box (one with no Post-It attached). It places the paper inside that box,
> and moves the Post-It "B" from where ever it currently is to the newly
> filled box.
>
When a name is rebound in a Python namespace, the location of the name (and
therefore the technique used to reference the associated value) does not
change. What changes is the object reference associated with the name (in
other words, the content associated with the name). For this reason I think
this analogy is a little stretched, and does not reflect the underlying
mechanisms in a way which is helpful to the Python newbie.

>         So in effect, normal Python variables are always "pointers" --
> assignment consists of changing where the variable points, not of
> changing the contents pointed to by a variable.
>
This, if course, I have no problem with at all. That is exactly the case.
And a list and a tuple are ordered collections of pointers, and dictionaries
are collections of pointers that can be retrieved by arbitrarily
subscripting [a reference to] the object containing them.
>
>         At this point I fall dry -- some data types may perform "in place"
> modifications, leaving the "label" alone and changing the contents of
> the box, but I don't believe there is a direct procedure for doing a
> "dereferencing" operation.
>
Well, Python dereferences names automatically. When you deal with objects
such as instances and lists, it helps to remember that each list element has
properties very similar to those discussed above. Thus, changing the value
of a list element does not require any memory shuffling, since a list is
(until you start adding or removing elements from it) simply an ordered
collection of references, whose names are generated by indexing the list
object.

References to a name in an expression are automatically dereferenced. When a
name is used as the left-hand side of an assignment, however, no
dereferencing is performed.

Tuples, being immutable, do not allow any changes to their contents. Lists
and dictionaries, being mutable, do. A name which is bound to an immutable
object can, of course, easily be rebound to some other object. It can also
be completely unbound with "del".

consider the statements

    a = 3.1
    a += 7.5

The first statement binds the name "a" to a floating point value. The second
statement generates a new floating point value by adding the value
referenced by the variable's current binding, adding it to the (unnamed)
floating-point constant 7.5, and rebinding the name "a" to the resulting
floating-point value (which should be a close approximation to 10.6.

There is very little difference between this and the statements

    b = [3.1]
    b[0] += 7.5

except that what's being changed is a list element rather than an explicit
name. In this case the name "b" is bound to a list, and b[1] is bound to a
floationg-point value. Similarly

    c = {0: 3.1}
    c[0] += 7.5

will change an element of the dictionary to which the name "c" is bound,
rebinding the element to a new value. In these last two cases no name
rebinding is involved -- the name is bound to an object in the first
statement, and an element of the referenced object is rebound by the second.

Attempts to use a tuple instead of a list will raise an exception due to the
immutability of tuples.

Augmented assignments are just a little more complicated, since the
implementors of mutable types are free to *either* modify an existing value
in-place or overwrite a reference to the old value with a reference to the
new value.

In the case where the same object has been bound to multiple variables,
changes to the bindings stored inside the object are naturally seen
whichever name is used to reference the object.

What Henning seems to want is parallel binding of multiple names, so that
rebinding "a" will automatically rebind "b" as well. There are indirect ways
to achieve this end, but nothing that will work directly.

regards
 Steve







From jb at cascade-sys.com  Mon Apr 29 12:54:30 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 09:54:30 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <aaj45k$odt$1@thorium.cix.co.uk>
Message-ID: <3CCD7AC6.FC7871C9@cascade-sys.com>

gbreed at cix.compulink.co.uk wrote:

> The rationale's very brief about the disadvantages of a bloated
> __builtins__.  It isn't like they're reserved words.  We don't need to be
> familiar with them all, because we can always check the library reference
> if we see one we don't know.

There is ZERO benefit to moving these several operators from builtins to a
module.  People unfamiliar with them will be equally confused in either case,
perhaps more so.

> I don't see any benefits to outweigh the pain this is going to cause some
> of us.

Ditto.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From tim at frontier.net.nospam  Wed Apr 10 12:23:26 2002
From: tim at frontier.net.nospam (Tim H)
Date: Wed, 10 Apr 2002 10:23:26 -0600
Subject: Python 2.2.1 on Win prob?
Message-ID: <3cb466fe$0$16416$4c41069e@reader1.ash.ops.us.uu.net>

On a Windows 2000 SP2 machine with the newly installed Python 2.2.1:

python -c"from test import testall"

the last two lines read:

1 skip unexpected on win32:
    test_threaded_import

Something to worry about?

Tim

PS Fairly inexperienced Python user.




From johanfo at essay.org  Sun Apr 28 15:33:22 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Sun, 28 Apr 2002 21:33:22 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net>
Message-ID: <aahiq2$6t3$1@maud.ifi.uio.no>

I have just had a look at Webware myself,  and it seems like one of the more
promesing projects.  At least it is not abandoned.
--
JF?

> I tested Webware the other week and it looks very promising. It is very
easy
> to e.g. develop a template as a servlet and let other servlets inherit the
> layout and functionality which makes Webware extremly powerful and easy to
> use.
>
> There is also support for PSP which I haven't tried (embedded Python in
> HTML).
>
> Another powerful (not yet final though) module is Middlekit which e.g.
makes
> it easy to deploy a servlet that acts with some glueware to e.g. MySQL.
>
> I am going to follow the work of Webware since I like the freedom it
offers.
> It seems more convenient to use comapred to Zope.





From Scott.Daniels at Acm.Org  Fri Apr 26 04:23:38 2002
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Fri, 26 Apr 2002 08:23:38 GMT
Subject: An IMHO interesting optimization question...
References: <3cc73c81@nntp.server.uni-frankfurt.de> <aa7nv9$aju$0@216.39.172.122>
Message-ID: <1103_1019809418@news.dsl-only.net>

The best I found (I read from a file to try several strategies 
per run) did reasonably well as "python -OO race.py"
..
longWord=15
filename = 'folio.txt'

def lcounts( words ):
   "Increments the wordcounter -- locals and seq arg"
   lstats = stats
   for word in words:
       length = len( word )
       if length > longWord:
           length = longWord
       lstats[length] += 1

def laruns():
    global stats
    stats = [0] * (longWord+1)
    start = time.time()
    findall = wordExp.findall
    for line in file(filename):
        lcounts( findall( line ))
    return time.time() - start
..

-Scott.Daniels at Acm.Org





From tim.one at comcast.net  Tue Apr  2 00:48:58 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 02 Apr 2002 00:48:58 -0500
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
In-Reply-To: <qzbq8.3238$tW2.2164@nwrddc04.gnilink.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEOIOJAA.tim.one@comcast.net>

[logistix]
> So this PEP will break the ternary hack (a and b or c) and other crazy
> exploitation of short-circuiting tricks in lambdas?

Nope!  All of that will continue to work fine.  You could try reading the
PEP if you want an explanation <wink>.

> Not that I mind. ;)

Someone would.  Someone always does.  Some people even fight to keep gross
bugs intact, although not yet to preserve segfaults.

picky-picky-ly y'rs  - tim




From emmanuelgagniere at wanadoo.fr  Mon Apr 29 13:17:41 2002
From: emmanuelgagniere at wanadoo.fr (Emmanuelgagniere)
Date: Mon, 29 Apr 2002 10:17:41 -0700
Subject: wxHtml - newbie
Message-ID: <aajv7m$3hl$1@wanadoo.fr>

Hello,

Do you know where can I find an example of a simple application using
wxHtmlWindow?

Thanks.




From rdsteph at earthlink.net  Tue Apr 23 21:03:54 2002
From: rdsteph at earthlink.net (rdsteph)
Date: Wed, 24 Apr 2002 01:03:54 GMT
Subject: Python, search engine and NLP...
References: <409a56e2.0204230104.16d60f52@posting.google.com>
Message-ID: <3CC60496.4070509@earthlink.net>

http://sourceforge.net/projects/nltk is the url for the sourceforge page 
for the Python Natural Language Toolkit. Also you might try 
http://pywordnet.sourceforge.net/ which is PyWordNet, a Python interface 
to WordNet.

I do not know if these will be of any help, but maybe????

Good Luck,

Ron Stephens



From mats at laplaza.org  Wed Apr  3 14:43:33 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Wed, 03 Apr 2002 19:43:33 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <7xelhx3zz6.fsf@ruckus.brouhaha.com>
Message-ID: <3cab597a.62091883@news.laplaza.org>

On 03 Apr 2002 01:15:09 -0800, Paul Rubin <phr-n2002a at nightsong.com>
wrote:

:vio <vmilitaru at sympatico.ca> writes:
:> I need to translate a (rather voluminous 100MB) java program to
:> python. After a few searches on google, apparently the best tools
:> for this task are Lex-Yacc (Flex-Bison on linux). While I am working
:> on the Lex-YACC HOWTO, I would be very interested in hearing of
:> others' successful experiences/feedback at this. Any pointers of
:> interest would similarily be appreciated.
:
:That's crazy.  Don't even think about it.  If you want Python-Java
:integration, use Jython.

It's doubful if any possible benefits of switching from Java to Python
can be realized by converting the code, which will give you not
Python, but "Java written in Python".

Instead, suggest leveraging the existing work by using Jython as a way
to begin recoding bits, probably starting with those that are shown to
need upgrading, maintenance, etc. 


Mats Wichmann



From aahz at pythoncraft.com  Sun Apr  7 14:55:27 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 14:55:27 -0400
Subject: I need to go to =?iso-8859-1?Q?Malm=F6?= tomorrow
References: <mailman.1018126154.20514.python-list@python.org> <a8nt0j$411$1@panix1.panix.com> <slrnab1477.cde.mlh@vier.idi.ntnu.no>
Message-ID: <a8q4mv$7u3$1@panix1.panix.com>

In article <slrnab1477.cde.mlh at vier.idi.ntnu.no>,
Magnus Lie Hetland <mlh at vier.idi.ntnu.no> wrote:
>In article <a8nt0j$411$1 at panix1.panix.com>, Aahz wrote:
>>In article <mailman.1018126154.20514.python-list at python.org>,
>>Laura Creighton  <lac at strakt.com> wrote:
>>>
>>>I haven't forgotten you all, but it will be a while before I get to
>>>replying.  Alex and I have code to write together as well.  I get a
>>>week of pair programming with Alex!  Way Cool.
>>
>>You're just trying to make us feel jealous.
>
>On which one of them?

Yes.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From jepler at unpythonic.net  Sun Apr 14 11:17:57 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 14 Apr 2002 10:17:57 -0500
Subject: Does Python need an 'xor' operator?
In-Reply-To: <ubirifj8o6fu48@news.supernews.com>
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <ubirifj8o6fu48@news.supernews.com>
Message-ID: <20020414101753.A11353@unpythonic.net>

On Sun, Apr 14, 2002 at 07:56:06AM -0400, John Roth wrote:
> So it's not possible to return one of the operands, we have
> to return True or False. However, now the operation
> is not consistent with 'and' and 'or,' which imposes an
> issue with teaching the language.

You could come up with something more perverse if you want to better keep
the spirit of the 'and' and 'or' operators for 'a xor b'.

If exactly one operand is true, return that operand.  Otherwise, if both
operands are false return 'b'.  If both operands are true, return 'not b'.

This follows (kinda) from the identity
    a xor b == (a or b) and not (a and b)

I'm sure this is in keeping with the spirit of Python.  It doesn't
short circuit, but it returns one of its arguments when possible,
and returns the 'not' of one of its arguments in the fourth case.
There's no short-circuit evaluation, but that's just too bad.

Jeff




From paul at prescod.net  Wed Apr 10 00:01:31 2002
From: paul at prescod.net (Paul Prescod)
Date: Tue, 09 Apr 2002 21:01:31 -0700
Subject: [Python-Dev] Re: ANN: Pyrex - a language for writing Python 
 extension modules
References: <200204100229.OAA00739@s454.cosc.canterbury.ac.nz>
Message-ID: <3CB3B91B.3E20FE02@prescod.net>

Greg Ewing wrote:
> 
>...
> 
> You're not meant to compile all your Python code with Pyrex.
> You only put Python code in a Pyrex module if it's needed to
> interface with non-Python code. Pyrex isn't designed to speed
> up Python code, but to bridge the gap between Python and
> non-Python. Speed, if any, comes from the non-Python code
> (Pyrex code operating on C data, or external C code).

It would be clearer if the canonical[1] PyRex example was a bridge
example instead of a "pure pyrex" example.

 Paul Prescod

[1]http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/Doc/About.html




From tim.one at comcast.net  Tue Apr 16 22:59:39 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 22:59:39 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <a9im4e$rf4$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCMENOOOAA.tim.one@comcast.net>

[Bengt Richter]
> ...
> If so, you could just specify reboot after uninstall before the
> next install as a fix, I would think.

Rebooting Windows is a PITA for users, and if someone is confused enough to
try to uninstall Python while they're using Python, I'd rather they get a
complaint from Wise.  That is, silent deletion of in-use files is a dubious
idea for Python anyway, and that it tickles a Wise bug (Wise uses a
different mechanism to get rid of UNWISE.EXE -- they shouldn't be trying to
do a redundant at-reboot delete of this specific file) nails it for me.





From tim.one at comcast.net  Wed Apr 10 20:42:40 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 20:42:40 -0400
Subject: Easter Eggs
In-Reply-To: <roy-E50C96.20260810042002@news1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEBBONAA.tim.one@comcast.net>

[Roy Smith]
> Software is my job (both writing it and useing it).  If I can't have a
> little fun doing my job, it's time for a new career.

Actually, we were trying to give people a reason to keep their t-shirts:
somehow or other "import this" got picked as the slogan for the last Python
conference, and since nobody was quite sure what it meant, we decided to
*make* it mean something in time for the conference.  So, it was no fun at
all, just another hard-headed business decision to ensure the continuing
value of users' Python investement.  We take t-shirts very seriously at
PythonLabs.

they're-all-backward-compatible-too-ly y'rs  - tim





From gusraan at terra.com.br  Wed Apr 10 15:10:53 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: Wed, 10 Apr 2002 16:10:53 -0300
Subject: Python FREE IDE for windows
Message-ID: <005b01c1e0c3$707dc620$01c8a8c0@raphaelofizs8j>

does anyone know?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020410/159b20d6/attachment.html>

From aleax at aleax.it  Fri Apr 26 17:19:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 21:19:37 GMT
Subject: Removing dupes from a list
References: <mailman.1019853649.15936.python-list@python.org>
Message-ID: <Jvjy8.66361$vF6.2030499@news2.tin.it>

Alexander Skwar wrote:

> What's the fastest way to make sure that a list only contains unique
> entries?

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560


Alex



From brueckd at tbye.com  Mon Apr 22 12:30:38 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 22 Apr 2002 09:30:38 -0700 (PDT)
Subject: HTTP error on receiving a form
In-Reply-To: <q_Vw8.76070$T%5.17278@atlpnn01.usenetserver.com>
Message-ID: <Pine.LNX.4.33.0204220926310.25157-100000@ns0.tbye.com>

On Mon, 22 Apr 2002, Steve Holden wrote:

> > 2 - is there a quick way to encode data from a dictionary in a specific
> order.
> >
> >
> Not using urllib.urlencode().

(sorry to respond to Steve's post, but I missed the start of this thread)

Not from a dictionary, but urlencode can keep the order from a sequence:

>>> import urllib
>>> help(urllib.urlencode)
Help on function urlencode in module urllib:

urlencode(query, doseq=0)
[snip]
    If the query arg is a sequence of two-element tuples, the order of the
    parameters in the output will match the order of parameters in the
    input.

But like Steve said, the chances that it's an ordering problem are really,
really low (IOW, if it *is* an ordering problem then the server/CGI is
broken :)  ).

-Dave





From DeepBleu at DeepBleu.org  Tue Apr  9 09:31:20 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Tue, 9 Apr 2002 08:31:20 -0500
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org>
Message-ID: <ub5s4paqolpo0a@corp.supernews.com>

"Gustavo Niemeyer" <niemeyer at conectiva.com> wrote in message
news:mailman.1018294997.29344.python-list at python.org...
> > I am using a python script and the zlib module to zip two text files
into
> > two separate zip files.  They are zipping fine.  However, when I use
winzip
> [...]
>
> Have you tried using the zipfile module?
>
Yes, I am using zipfile (from zipfile import ZipFile).  Sorry I stated zlib.
Still with one text file in particular, WinZip reports "missing 2 bytes"
DeepBleu

> --
> Gustavo Niemeyer
>
> [ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]
>
>




From eppstein at ics.uci.edu  Mon Apr  1 22:21:16 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 01 Apr 2002 19:21:16 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com> <a8an7b$eme$0@216.39.172.122> <eppstein-250858.15583201042002@news.service.uci.edu> <a8b4kk$je2$0@216.39.172.122>
Message-ID: <eppstein-8A375A.19211601042002@news.service.uci.edu>

In article <a8b4kk$je2$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
wrote:

> >If you called it "binary" instead of "bool" would it help you think of 
> >it as being more like a number?
> >
> No. Because numbers are ordered, and bools are not. bools are selected
> elements from an unordered set of two possibilities. The bool is one chosen
> element, but the element can be anything that can be distinguished from the
> other element of the set. They have names str(b) associated with them, but
> they don't have an order. One is chosen to mean logical true and the other
> logical false, but they are just tokens with no necessary order.

But there are systems of numbers that are just as much "selected 
elements from an unordered set" as the boolean truth values: the members 
of the various finite fields.  In fact the field of order two is a 
pretty good match for the booleans.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From mal at lemburg.com  Tue Apr  2 13:49:41 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 02 Apr 2002 20:49:41 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>
Message-ID: <3CA9FD45.E6E33129@lemburg.com>

I haven't followed this thread for quite some time, but since the
PEP still seems alive, let me add some experience I've had with
using the already existing singletons (Py_True and Py_False)
in Python for recognizing truth values.

Py_True and Py_False can be exposed in Python via

True = (1==1)
False = (1!=1)

and most comparison operations and quite a few other APIs
returning truth values make use of these singletons.

I thought it would be a good idea to use those two
singletons for protocols which use booleans such as
XML-RPC. My experiences with this approach are, well,
not so good :-/

The reason is that True and False are really integers
and not of a special new type. Now they are singletons,
which is good, since they represent the only states
a boolean can have and in many cases work reasonably
well as boolean representative, but not always 
(e.g. arithmetic operations such as True - True == False). 
Also, when trying to recognize the two singletons you have 
to use the "is" comparison -- "==" will fail to 
differentiate between 1 and True...

def isboolean(x):
    return x in (True, False)

...doesn't work...

def isboolean(x):
    return (x is True) or (x is False)

..does.

As a conclusion, I think it would be better to make bool() a 
new type which does *not* inherit from integers, but which 
does know how deal with other types which are commonly
used together with booleans such as integers. However, the
type should implement boolean algebra and not try to
mimic integer arithemtic, i.e. True - True raises an
exception.

Py_True and Py_False should then be made singletons
of this new type (possibly after a transition phase which
replaces the singletons with a version that doesn't raise
exceptions but instead issues warnings).

This may sound painful at first, but in the long run,
I believe, it'll result in the same benefits as other 
painful changes have or will (e.g. the change from integer 
division to floating point division).

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/



From liste_nospam_ at opikanoba.org  Wed Apr 17 12:36:14 2002
From: liste_nospam_ at opikanoba.org (=?windows-1252?Q?Fr=E9d=E9ric=20Laurent?=)
Date: Wed, 17 Apr 2002 16:36:14 GMT
Subject: amkCrypto installation on win32
Message-ID: <1103_1019061374@news.free.fr>

Sur 17 Apr 2002 14:38:15 GMT, vous avez ?crit dans comp.lang.python:
> According to Fr?d?ric Laurent  <liste_nospam_ at opikanoba.org>:
> > I want to use RSA encoding/decoding, so I downloaded amkCrypto 0.1.3.
> > I use python 2.2 on windows XP.
> > 
> > So, I just want to generate a pblic/private key, crypt and decrypt a text
> > buffer... but it sounds to be not so simple to do this !
> 
> Try M2Crypto:

I can't ! why ? because neither MS VC++ 6.0 nor Borland C++ is installed on my 
computer !
So I can't compile the module ...
and there is no precompiled dll in the win directory of the binary zip package,
even if the documentation says the opposite ...




> 
> ngps at vista:~$ python
> Python 2.2 (#7, Mar 31 2002, 23:17:41) 
> [GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
> Type "help", "copyright", "credits" or "license" for more information.
> >>> from M2Crypto import RSA
> >>> r = RSA.gen_key(1024, 3) 
> ............++++++
> ....................++++++
> >>> r
> <M2Crypto.RSA.RSA instance at 0x8197d2c>
> >>> ptxt = 'plain text'
> >>> import sha              
> >>> x = sha.sha(ptxt).digest()
> >>> ctxt = r.public_encrypt(x, RSA.pkcs1_oaep_padding)
> >>> ptxt2 = r.private_decrypt(ctxt, RSA.pkcs1_oaep_padding)
> >>> ptxt2 == x


--
Fr?d?ric Laurent
http://www.opikanoba.org 




From p.magwene at snet.net  Thu Apr 11 09:07:15 2002
From: p.magwene at snet.net (Paul Magwene)
Date: Thu, 11 Apr 2002 13:07:15 GMT
Subject: ANN: disipyl - a 2D and 3D plotting library for Python/Jython
Message-ID: <pan.2002.04.11.09.07.15.342883.78423@snet.net>

DESCRIPTION:
-----------
    disipyl is an object-oriented wrapper around the DISLIN plotting
library.  DISLIN is a powerful and flexible multiplatform (Win32, Unix,
Linux, etc.) library designed for displaying scientific data. DISLIN's
author, Helmut Michels, has made available a DISLIN plotting extension for
the Python programming language (see http://www.linmpi.mpg.de/dislin/ for
more details).

    disipyl provides a set of classes which represent various aspects
of DISLIN plots, as well as providing some easy to use classes for
creating commonly used plot formats (e.g. scatter plots, histograms, 3-D
surface plots).  A major goal in designing the library was to facilitate
interactive data exploration and plot creation.

    Documentation, a tutorial, and a demo program are included.  The
library has been tested on Win32, Linux, and FreeBSD, but I anticipate
that it should work on any platform which can make use of Python, NumPy,
and the DISLIN python extensions.

    Feedback, comments, and critique are gladly accepted (email:
paul.magwene at yale.edu).



VERSION:
-------
    This is release 0.6 of disipyl.

CHANGES:
-------
New in this release:

* support for plot interaction (record plot coordinates via mouse clicks)

* new mechanism for quering disipyl objects about the options they support

* cleanup of data/option setting mechanisms for disiypl objects

* Jython support (when DISLIN java library available - most but not all
classes)

* updated documentation and tutorial


URL:
----
    You can find disipyl at:
http://pantheon.yale.edu/~pmm34/disipyl.html


From Oschler at earthlink.net  Mon Apr  1 00:14:49 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 00:14:49 -0500
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <a886hv$st6$1@panix1.panix.com> <U5Op8.61601$Q7.48626359@e3500-atl1.usenetserver.com> <mailman.1017629043.9252.python-list@python.org>
Message-ID: <oVRp8.63158$Q7.50001123@e3500-atl1.usenetserver.com>

Gerhard,

Just ordered "The Zope Construction Kit" book from Amazon today.

thx

Gerhard H?ring <gh_pythonlist at gmx.de> wrote in message
news:mailman.1017629043.9252.python-list at python.org...
> I'd recommend to start with the ZOPE tutorial and examples on a live ZOPE
> installation if you want to get an impression of it.
>






From phd at phd.pp.ru  Sat Apr 20 02:59:37 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Sat, 20 Apr 2002 10:59:37 +0400
Subject: Integrating C/C++ in to python/Zope
In-Reply-To: <a9q49n$tq7$1@news.datasync.com>; from kalphonso@diamonddata.com on Fri, Apr 19, 2002 at 05:04:11PM -0500
References: <a9q49n$tq7$1@news.datasync.com>
Message-ID: <20020420105937.C4819@phd.pp.ru>

On Fri, Apr 19, 2002 at 05:04:11PM -0500, Keith Alphonso wrote:
> Can anyone point me to documentation explaining how to integrate my C or C++
> with Python, specifically for Zope?  I can't seem to find anything
> difinitive anywhere.
> 
> Do I basically have to modify the Python interpreter source code to put my
> code with it?  Is there any mechanisms for linking in dynamic libraries?

   http://www.python.org/doc/current/ext/ext.html

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




From fgeiger at datec.at  Wed Apr 17 07:22:27 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Wed, 17 Apr 2002 13:22:27 +0200
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>
Message-ID: <a9jluf$kn8$1@newsreaderm1.core.theplanet.net>

The wxPython demo program shows the use of threads within wxPython GUIs.
Threads, that want to display something in say a notebook page, send
messages to the main frame, which actually does the display.

There's also a Wiki (see wxPython.org), where advanced topics are discussed.

Cheers
Franz


"sameer" <sameer_ at email.com> schrieb im Newsbeitrag
news:6bd9f01b.0204152132.6a6ad09e at posting.google.com...
> I am trying to run multiple threads in a wxFrame class to instantiate
> different panels within it on Win 2k,  and well, it's a little hairy.
> Anyone got a good web source that discusses python threading in depth?
>
> More so, my question is, how can I get the return value of a function
> that I run on a seperate thread?  I don't see anything for this in the
> threading module.
>
> Thanks in advance for your help!




From duncan at NOSPAMrcp.co.uk  Tue Apr 16 04:18:55 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 16 Apr 2002 08:18:55 +0000 (UTC)
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <Xns91F25DDBFEB2Eduncanrcpcouk@127.0.0.1>

philh at comuno.freeserve.co.uk (phil hunt) wrote in 
news:slrnabmoci.dhk.philh at comuno.freeserve.co.uk:

> (1) Convenience
> 
> I would like to be able to write as code classes and functions 
> without implementing the body to start with, e.g.:
> 
> 
> class MyClass:
>    def __init__(self, x, ,y, z):
>    def getTwiddle(self):
>    def setTwiddle(self, twiddle):
> 
> 
> At the moment if I want to do this I have to put pass in and then 
> take them out again. A minor inconvenience, to be sure, but one that 
> niggles slightly due to the highly exploratory style of coding I 
> often use.
You should always head each method and class with a docstring.
Apart from being incredibly useful when you come back to the methods some 
months later and cannot remember what they did, this has exactly the effect 
you wanted of making 'pass' optional:

class MyClass:
   '''A simple class'''
   def __init__(self, x, ,y, z): '''MyClass constructor'''
   def getTwiddle(self): '''Get a twiddle'''
   def setTwiddle(self, twiddle): '''Set a twiddle'''

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From boud at valdyas.org  Sat Apr 13 06:36:18 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 13 Apr 2002 12:36:18 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl> <m3662wypqt.fsf@mira.informatik.hu-berlin.de> <3cb7f8a7$0$85437$e4fe514c@dreader3.news.xs4all.nl> <m37knb9axo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cb80a78$0$43852$e4fe514c@dreader1.news.xs4all.nl>

Martin v. Loewis wrote:
> No, not in any released version. Meanwhile, I recommend to stay away
> from Unicode literals if you need "funny" characters, and write
> 
>   unicode("<string containing funny characters>", "utf-8")
> 
> instead.

Yes. Although I might be going to try the relevant patch.
-- 
Boudewijn Rempt | http://www.valdyas.org


From geoff at gerrietts.net  Wed Apr 10 23:42:31 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 10 Apr 2002 20:42:31 -0700
Subject: Easter Eggs
In-Reply-To: <b0f083db.0204101941.1f37632b@posting.google.com>
References: <mailman.1018391602.4247.python-list@python.org> <b0f083db.0204101941.1f37632b@posting.google.com>
Message-ID: <20020411034231.GD30655@isis.gerrietts.net>

Quoting Jeremy Hylton (jeremy at alum.mit.edu):
> Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote in message news:<mailman.1018391602.4247.python-list at python.org>...
> > Also, is anything else hidden in the Python interpreter that issues
> > similar wisdom?  (I know...  I could read the source; but this is
> > probably quicker).
> 
> I can think of one other small feature of this sort, but I don't think
> I should tell you.

Now that's just plain cruel. ;)

Is this your one-man crusade to get a lot of eyes on the source? ;)

Thanks,
--G.

-- 
Geoff Gerrietts             "Don't get suckered in by the comments-- 
<geoff at gerrietts net>     they can be terribly misleading.
www.gerrietts.net/geoff/     Debug only code." --Dave Storer




From fredrik at pythonware.com  Tue Apr  9 05:20:25 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 09 Apr 2002 09:20:25 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org>
Message-ID: <tnys8.11$p56.2333@newsb.telia.net>

Sean 'Shaleh' Perry wrote:
>
> > List comprehension is cutesy and practically Perlish, an idiosyncrasy
> > reminiscent (in flavor and aroma) to Perl's ".." operator.
> >
>
> I'll see your list comprehensions and raise you a set of string methods.
>
> " ".join(mylist) # is this perl or python?

*why* string objects have join method has been explained
over and over and over again, and should also be obvious for
anyone with some knowledge how strings and sequences work
in Python.

> whoever decided this is the new style should be shot, a lot

I suggest you start by punishing yourself, for not knowing
what you're talking about, and not paying attention.

</F>




From aahz at pythoncraft.com  Tue Apr 23 18:45:41 2002
From: aahz at pythoncraft.com (Aahz)
Date: 23 Apr 2002 18:45:41 -0400
Subject: Passing arguments to threading objects
References: <pan.2002.04.23.22.39.14.487844.11225@teich.garten.DigitalProjects.com>
Message-ID: <aa4o6l$1ce$1@panix1.panix.com>

In article <pan.2002.04.23.22.39.14.487844.11225 at teich.garten.DigitalProjects.com>,
Alexander Skwar  <news.ASkwar at DigitalProjects.com> wrote:
>
>When I subclass threading.Thread, how can I access the args parameter
>passed to the constructor?

Override __init__().  See my home page for more info.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From mcfletch at rogers.com  Sun Apr 14 01:48:40 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sun, 14 Apr 2002 01:48:40 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCMEPOONAA.tim.one@comcast.net>
Message-ID: <3CB91838.6000902@rogers.com>

Rather than confuse Tim further with abstract samples of where I want to 
define a "NULL" default argument, here's a particular example of where 
the "NULL pattern" I've described is used:

NULL = []

class BasicProperty (property):
...
	def __init__( self, name, documentation = "", bounds = (), 
defaultValue=NULL, defaultFunction=NULL ):
...
		if defaultValue is not NULL:
			self.defaultValue = defaultValue
		elif defaultFunction is not NULL:
			self.defaultFunction = defaultFunction

Now, I want to store/define defaultValue only if defaultValue is 
explicitly specified (same for defaultFunction).

I _could_ (as you suggest) use namedarguments and require that the user 
name the arguments explicitly.  However, when I do that, the 
defaultValue and defaultFunction arguments don't show up in the user's 
call-tip window using Python IDEs.  (And I dislike the aesthetics of 
using **namedarguments when all acceptable names are known anyway :) ).

As for the "designed purpose of default arguments" :) .  I see the value 
"no value/not specified" as a perfectly suitable default value.  Sure, 
it may offend your sensibilities, but when has that stopped me in the 
past ;) .

As for the brittle-ness.  Yes, it's a problem, especially if someone 
decides to implement copy-on-write semantics for lists some day (so that 
different lists could have the same id).  Of course, when you do that, 
all hell will break loose with code everywhere crashing and burning, we 
all know how much you like to party ;) :) .  For now, I'll switch the 
pattern over to using Singleton objects I suppose.  I don't think even 
you're planning to optimise away instance identity any time soon.

Okay, enough, I'm getting back to work :) .  Enjoy,
Mike



Tim Peters wrote:
> [Mike C. Fletcher]
> 
>>I was unaware that there was no guarantee for 0-length tuples always
>>being unique objects.  There are lots of places in my code where I use
>>0-length tuples in this pattern:
>>
>>NULL = ()
>>
>>class A:
>>	def b( self, object=NULL ):
>>		if object is NULL:
>>			#was not specified
>>		elif object is None:
>>			#specified to be object None
> 
> 
> Mike, I'm pretty lost.  Are you saying that if a user explicitly calls
> 
>     A().b(())
> 
> or
> 
>     A().b((1, 2, 3)[2:2])
> 
> you *want* to hit the "was not specified" block?  That's all a guarantee
> about 0-length tuples would buy you.  But if it's not what you want, such a
> guarantee is irrelevant:  the method is full of holes regardless of whether
> the empty tuple is unique.
> 
> Note that default arguments aren't designed to tell you whether a specific
> argument was explicitly passed.  They're designed to supply reasonable
> defaults that a user could just as well pass explicitly if they feel like
> it.  Rather than invent more brittle tricks, I suggest pondering whether the
> intended use of this gimmick isn't good enough for your purposes (and noting
> that it's always been good enough for mine).
> 
> If you really need to know whether a thing was passed explicitly, it makes
> more sense to insist that specifying it be done via keyword, and write the
> function to examine the keys in the keyword dict passed to it.  It's all
> obvious then.
> 
>     def f(required, **kw):
>         if 'abc' in kw:
>             print "'abc' was explicitly passed"
>             abc = kw['abc']
>         else:
>             print "'abc' was not specified"
>             abc = 42
>         print "'abc' is", abc
> 
>     f(3)
>     f(3, abc=666)
> 
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From jeff at ccvcorp.com  Thu Apr  4 16:00:26 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 4 Apr 2002 13:00:26 -0800
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com>
Message-ID: <MPG.17165ec2e3c10c6c98969a@news.nwlink.com>

In article <3cac943f.132176618 at mammoth.usenet-access.com>, 
netzapper at magicstar.net says...
> On Thu, 04 Apr 2002 11:56:24 GMT, Alex Martelli <aleax at aleax.it>
> wrote:
> 
> >You cannot use ANY identifier safely after that exec
> >statement, as you have no idea any more what the
> >identifier can refer to.  As a bonus, your code crawls,
> >since the Python compiler knows it does not know, and
> >therefore executes a full identifier look-up at runtime
> >rather than recognizing local variables at compile time.
> >
> >What kind of useful code can you write without being
> >able to use ANY identifier safely?  WHY would you ever
> >WANT to trample all over your namespace to ensure every
> >identifier becomes an utter and total mystery?  Beats me.
> 
> I agree with you if it's an arbitrary string being passed to your exec
> call .  However, if you have some idea of what the form of the string
> is going to be, it doesn't seem so dangerous.  For instance:
> 
> for x in range(0, 100):
> 	exec foo + `x` + " = " + "myClass()"
> 
> allows you to easily make a large list of sequentially named
> variables, with no chance of stepping on a system call.

And *why* on earth would you want to do that, instead of:

foo = [myClass() for x in range(100)]

which creates a *true* list of 100 objects, which are not just 
sequentially named, but are an actual sequence?  Surely it's not 
so much easier to write foo23 than to write foo[23] ... and of 
course, your method will require even *more* exec statements when 
you want to modify all 100 of these objects:

for x in range(100):
    exec foo + `x` + " = " + foo + `x` + " % 5"

compared to:

foo = [x % 5 for x in foo]

Creating sequentially-named variables is the *wrong* approach.  
If you have numerous objects that are conceptually similar, so 
that it makes sense to name them sequentially, then you *really* 
want to put them in a list or dictionary instead.

The same thing is true of almost *every* use of exec or eval() -- 
it seems handy, but if you look for it, there's a better solution 
to your real problem.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From eppstein at ics.uci.edu  Thu Apr  4 15:46:49 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 04 Apr 2002 12:46:49 -0800
Subject: Response to Guido's essay on graph representation
Message-ID: <eppstein-87C133.12464904042002@news.service.uci.edu>

I've posted an implementation of Dijkstra's algorithm, which is in some 
sense a response to Guido's essay 
<http://www.python.org/doc/essays/graphs.html> on the proper Pythonic 
representation of graphs, to
<http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466>.

In essence, I agree with Guido's point that a full-blown object oriented 
representation is too cumbersome, and that it is better to use Python's 
built-in data structures to form the overall graph structure, and to let 
the graph vertices be anything that can index a dictionary.
However his suggestion of a dictionary mapping vertices to lists of 
neighbors is a little too inflexible: instead I propose a dictionary 
mapping vertices to dictionaries of neighbors.  As a bonus, the 
neighbor-dictionary entries can be used to store additional graph 
information, which I use for the edge lengths needed for Dijkstra.

Comments welcome, here or (preferably) via the comment form on the 
recipe page.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From brueckd at tbye.com  Fri Apr 12 11:20:36 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 12 Apr 2002 08:20:36 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <eppstein-3037AA.19545211042002@news.service.uci.edu>
Message-ID: <Pine.LNX.4.33.0204120800530.27483-100000@ns0.tbye.com>

On Thu, 11 Apr 2002, David Eppstein wrote:

> In article <slrnabchgs.c8l.mlh at vier.idi.ntnu.no>,
>  mlh at vier.idi.ntnu.no (Magnus Lie Hetland) wrote:
>
> > I think the only _real_ reason to have tuples at all is that they are
> > immutable (and therefore usable as dictionary keys). And a slight
> > efficiency advantage over lists, maybe...
>
> I'm guessing (without looking at the source) that tuples use less memory
> because they don't need the extra room for growth that lists do.
> But my mental model of them is just "immutable list".

Yes, this is where the difference in opinion lies: as immutable lists,
tuple comprehensions can make sense. I like the mental model of tuples
put forth by the Python tutorial:

[5.3] "Tuples have many uses, e.g. (x, y) coordinate pairs, employee
records from a database, etc."

The implication is that tuples group bits of often dissimilar but related
information that, grouped together, represent a unique entity - sort of a
poor man's object instance.

In this sense tuple comprehensions are less valuable because they'd really
only be useful in cases where all the members are homogeneous, where it'd
make sense to apply the same sort of operation to each item. There's a few
obvious cases where this happens (coordinates, open/high/low/close prices
for a stock), but they don't seem all that common.

-Dave





From DeepBleu at DeepBleu.org  Fri Apr 19 11:16:23 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Fri, 19 Apr 2002 10:16:23 -0500
Subject: The Python Manual Index
References: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de>
Message-ID: <uc0d5d68i0vhf1@corp.supernews.com>

WOW!
Great Job!
DeepBleu

"Thomas Heller" <theller at python.net> wrote in message
news:a9pc0t$54o6t$1 at ID-59885.news.dfncis.de...
> Hi, folks.
>
> A new Python-powered search engine is online at
> http://starship.python.net/crew/theller/pyhelp.cgi
>
> It searches for keywords in the Python manuals at
http://www.python.org/doc/.
>
> The script can also be downloaded, and can be configured to search HTML
manual
> pages installed locally. The keyword index is build on the fly.
> It can also (probably) be easily integrated in your editor/ide to provide
> context sensitive help.
>
> Mark Hammond has integrated an interface to this keyword search in his
cool
> Mozilla/Netscape sidebar he recently announced:
http://starship.python.net/crew/mhammond/mozilla
>
> Regards,
>
> Thomas
>
>




From jaf at lcsaudio.com  Sat Apr 27 15:33:47 2002
From: jaf at lcsaudio.com (jaf)
Date: Sat, 27 Apr 2002 12:33:47 -0700 (-0700)
Subject: inline assignments in conditionals
Message-ID: <2143388891-BeMail@sugoi>

Roman Neuhauser wrote:

>Anyone who writes "if (x == 0)" instead of "if (0 == x)" deserves what

> he begs for.

Writing things that way is really unnatural, at least for English
speakers.  It makes the intent of the code harder to understand (at
least for me).  For example, if you were talking to a friend, you
woudn't say:

   "If today is Sunday, I'm going to go shopping"

but rather:

   "If Sunday is today, I'm going to go shopping"

and in any case, writing the constant first only protects in the case
where one of the compared values is a constant.  If you are doing

   if (a == b):

then there is no way to protect yourself against actually writing '='.

Given that, I think not supportig inline assignment is a reasonable
choice.  Perhaps another option would be to coin a synonymous operator
that works in comparisons, but is distinct enough from '==' so as not
to be used by accident?

   if (data gets socket.recv(1024)):
      process(data)

-Jeremy





From fgeiger at datec.at  Sat Apr 13 12:08:45 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Sat, 13 Apr 2002 18:08:45 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com><vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com><3CB274D2.A7554C79@engcorp.com><a8u6tj$quj$02$1@news.t-online.com> <20020409125632.40965c79.wilk@flibuste.net>
Message-ID: <a99l6j$rck$1@rex.ip-plus.net>

Yes, at ActiveState. Go for Cookbooks there.

Cheers
Franz

"William Dod?" <wilk at flibuste.net> schrieb im Newsbeitrag
news:20020409125632.40965c79.wilk at flibuste.net...
> Le Tue, 9 Apr 2002 10:03:04 +0200
> "Harald Schneider" <h_schneider at marketmix.com> ?crivait:
>
> > Thats the fact! If the community gives back the experience, people like
> > me would stop thinking about using
> > Delphi or VB ... we soon would have an excellent and up-to-date
> > documentation.
> Me to !
>
> >
> > What about posting them to
> > http://wiki.wxpython.org/index.cgi/wxPython_20Cookbook :)
> Is-there an equivalent for python (langage and library) ?
>
> bye
>
> --
> William Dod? - Informaticien Ind?pendant
> http://www.flibuste.net




From tejarex at yahoo.com  Sat Apr 13 19:00:06 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 13 Apr 2002 23:00:06 GMT
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com>

"Jens Baader" <nospam at nospam.com> wrote in message
news:3cb8a755$0$12714$9b622d9e at news.freenet.de...
> I'm downloading Python 2.2.1 right now and I wonder
> why you still produce bugfix releases for the old 2.1 development
line.

Because someone wants it done badly enough (perhaps with encouragement
from others) to volunteer to do it.  Sone of the 2.2.1 bugfixes were
backported from the nascent 2.3 deveopement version.

> It seems that Python 2.2 is somewhat broken.

Hogwash.  Bugfixes were ported from 2.2 back to 2.1 *because* 2.2
fixed a few things that were previous broken.

> If not what's the
> reason that keeps the people from upgrading to 2.2?

Perhaps you should ask before jumping to foolish conclusions.  Some
(overlapping) answers based on newsgroup posts as I remember them:

* Laziness; fatigue; lack of time.

* Lack of perceived positive (>1) benefit/cost ratio.

* Fear of rocking a boat that is sailing fine.

* Non-programmer end users do not care as long as their programs run
ok, so why bother unless forced to by the supplier of the programs
they run.

* Some programmers/businesspeople do not want to force an upgrade on
customers that will not benefit the customers.

* Some programmers do not currently intend or see that they would use
the new features.

* Others might be afraid that they might use the new features if
available, but cannot because of their customer base.

* ???

* ???

> When will we see an official ISO/ANSI standard for Python?

I hope never.  Committee standards are needed when multiple vendors
produce mutually imcompatible (in part) dialects.

> I dislike working with nonstandardized languages.

Then don't.

But notice that there is considerable difference between 'officially'
standardized and standardized in practice.  The cross-platform
portability of single-source-standardized Python is, I believe, as
great or better than some multisource officially-standardized
languages.  Official standards sometimes define features as
implementation-defined (which is to say, not standardized!).  Some
details of Python behavior are 'implementation-defined' precisely
because the same behaviors in C are not standardized and are not (yet)
hidden by current CPython.  There are many other platform and C
variations that have been hidden by Python acting as a standardization
layer.

Terry J. Reedy





From joonas at olen.to  Fri Apr  5 10:03:05 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Fri, 05 Apr 2002 15:03:05 GMT
Subject: location of a module's directory
References: <a8k1cb$s9v$1@news6.svr.pol.co.uk>
Message-ID: <3CADBDB4.7090007@olen.to>

Dave Berkeley wrote:
> Is there a way to discover where a module was loaded from?
> 
> I want to include a text configuration file with a .dll, and place them in
> the same directory. If the module knows where it was loaded from it will be
> able to find the config file.
> 
> If you do (Python 2.1.1)
> 
> 
>>import sys
>>print sys.modules
> 
> 
> you can get a dictionary of loaded modules, and a description, which
> includes the path the module was loaded from. However, I can find no way of
> accessing this information programmatically.

Are you looking for this?

Python 2.2a3 (#23, Sep  7 2001, 01:43:22) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE Fork 0.8 -- press F1 for help
 >>> import cgi
 >>> cgi.__file__
'C:\\Python22\\lib\\cgi.py'



From czaban at dgt.com.pl  Mon Apr 22 05:36:46 2002
From: czaban at dgt.com.pl (Piotr Czabañski)
Date: Mon, 22 Apr 2002 11:36:46 +0200
Subject: DB-API 2.0
Message-ID: <aa0lig$asv$1@news.tpi.pl>

Hi,

Two problems:

1.
    Fetch all (remaining) rows of a query result, returning them as a
sequence of sequences (e.g. a list of tuples).

    J think the proper result would be a list of lists - specilally for
people working with wxPython.

2.
    ProgrammingError: (-104, 'execute.isc_dsql_prepare: Dynamic SQL Error.
SQL
    error code = -104. token size exceeds limit. ')
    How to get round it? My command SQL string has 1352 character (it is
"insert" command).

Best regards
Piotr




From erno-news at erno.iki.fi  Tue Apr  2 20:57:50 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 03 Apr 2002 04:57:50 +0300
Subject: wxPython redefines string.letters?
References: <8351bb33.0204020854.272eaa7b@posting.google.com>
Message-ID: <kuu1qtpmqp.fsf@lasipalatsi.fi>

In article <8351bb33.0204020854.272eaa7b at posting.google.com>,
jkn at nicorp.f9.co.uk (Jon Nicoll) writes:

|>>> from wxPython.wx import *
|>>> string.letters
| 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\x83\x8a\x8c\x8e[...]

not to defend wxPython setting the locale behind your back, but
generally you should be using string.ascii_letters (& friends)
if you aren't prepared to handle locales being in use.

  -- erno


From gh_pythonlist at gmx.de  Tue Apr  2 00:08:48 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 2 Apr 2002 07:08:48 +0200
Subject: Extending Python on Windows
In-Reply-To: <3CA90BE1.9010703@cfl.rr.com>
References: <3CA90BE1.9010703@cfl.rr.com>
Message-ID: <20020402050848.GD1247@lilith.hqd-internal>

* Gregory Propf <gpropf at cfl.rr.com> [2002-04-01 20:39 -0500]:
> [...] the GUI will be done with wxWindows in python.  I will need to
> build python extensions for this.

> I have tried to build this extension using the version of gcc and ld
> that comes with cygwin.

You don't really want to use Cygwin as target on Windows, do you? It's
perfectly possible to use Mingw's or Cygwin's gcc (or even the free
Borland compiler) to compile Python extensions for *native* win32
Python. For native win32 Python, there's a wxPython installer. I just
wanted to warn you that building wxPython for Cygwin could be quite
tricky. And running wxPython over wxGTK over GTK over the Cygwin X
server over the Windows sounds pretty pointless ;-)

I already recommended distutils. If you want to use the Cygwin or Mingw
compiler to compile extensions for native win32 Python, the build
command using disutils is

    python setup.py --compiler=mingw32

Just for kicks, you can build an Windows installer for your module with:

    python setup.py bdist_wininst

or a RPM installer with:

    python setup.py bdist_rpm

Enough advertising distutils for today :-)

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.2 ?C      Wind: 0.9 m/s



From Matthias.Hanitzsch at dynamic.de  Tue Apr  9 10:42:59 2002
From: Matthias.Hanitzsch at dynamic.de (Matthias Hanitzsch)
Date: Tue, 09 Apr 2002 16:42:59 +0200
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com>
Message-ID: <3CB2FDF3.BBBD99C0@dynamic.de>


Dennis wrote:

> x = 2
>
> def wrong(num):
>     return num + 1
>
> print wrong(x)
> print wrong(x)
>
> I cannot figure out why this (simplified) program won't increment x
> and produce 3 and 4.

Hi,
this is "call by value", that means that the parameter num of your
function will get the VALUE of the variable x.
Then num is incremented and returned.
The variable x is not changed, as you never write to it. You just
take its value.

HTH,
Matthias



From ark at research.att.com  Wed Apr 10 11:42:43 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 15:42:43 GMT
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <mailman.1018385900.7252.python-list@python.org> <u8z7valyh.fsf@python.net>
Message-ID: <yu99662zmufg.fsf@europa.research.att.com>

>> I want to be able to stuff an object into a larger data
>> structure and have that object reach out into a variable and leave a
>> trail of breadcrumbs.

Michael> Can you come up with a more fleshed out example?  There
Michael> probably are some ways of doing what you want but they will
Michael> be hackish and fragile (grovelling around in frame objects,
Michael> etc).  OTOH, if you explain what you are trying to acheive,
Michael> perhaps we can come up with a more elegant way (it strikes me
Michael> that 2.2's properties may help here, for instance).

Sure.  I'm thinking about how to implement Snobol-style pattern
matching in Python.  I know that someone has done it before, but
I want to approach it from a completely different direction.

Snobol lets you write stuff like this:

        name ? pos(0) (break('@') . id) '@' (rpos(0) . domain)

(Experienced Snobol programmers will note that I have deliberately
introduced some extra parentheses and an '?' operator for clarity)

Here, "pos" denotes an offset from the beginning of a string,
"rpos" denotes an offset from the end, and break(s) matches the
string from the present position up to and not including the first
character that occurs anywhere in s.

The point of this example is the . operator, which takes a pattern
(by value) as its left argument and a variable (by reference) as its
right argument.  The result is a pattern that matches the same strings
as the left argument, but if the match is successful, it assigns
the substring matched to the variable that is the right argument.

As it happens, I can implement much the same idea in C++, though
the exact details of operator overloading make it look somewhat
different:

    string id, domain;
    match(name, pos(0) + (break("@") >> id) + "@" + (rpos(0) >> domain));

and I can also implement a similar idea in ML, though I don't
think it's worth going into details here.

So I was thinking about how I would do something similar in Python.

Focusing just on the  (break("@") . id)  subexpression, how would
one implement it?  That's what led me on this path.

Michael> Having random variables change their values at random times
Michael> strikes me as an idea that might come from - ooh - C++?
Michael> <wink>

Well, in this case, the idea came from Snobol, vintage 1968 or so.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From shalehperry at attbi.com  Fri Apr  5 17:22:00 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 05 Apr 2002 14:22:00 -0800 (PST)
Subject: icotl and Pointers
In-Reply-To: <200204052206.g35M6X304288@catfish.lcs.mit.edu>
Message-ID: <XFMail.20020405142200.shalehperry@attbi.com>

On 05-Apr-2002 Umar Saif wrote:
> Hi,
>       I am trying to write a piece of python code to control a sound device
(with 
> an oss driver).
>  The ioctls for the sound driver take in a pointer to an integer for the 
> argument -- I was wondering whether it is possible to do this in python ? Or 
> writing a native library is the only way around this ?
> 

see fcntl.ioctl().




From nobody at nobody.com  Sun Apr  7 01:36:59 2002
From: nobody at nobody.com (John)
Date: Sun, 7 Apr 2002 00:36:59 -0600
Subject: _Pg module and CGI
Message-ID: <vKRr8.36988$tg4.439776@vixen.cso.uiuc.edu>

Hi, assuming to build an application getting SQL statement from html form
and processing the query in postgresql, how can I add this "web user" to
access the database?

Error I get:
Traceback (most recent call last): File "c:\inetpub\scripts\sd.cgi", line
15, in ? db=_pg.connect('template1','localhost') _pg.error: FATAL 1: user
"IUSR_VALUED-VYL4TJPA" does not exist

Remedy I tried:
adduser IUSR_VALUED-VYL4TJPA in UNIX prompt
(but there wasn't such command)

I'd appreciate your help very much!





From eric at enthought.com  Mon Apr  8 15:33:08 2002
From: eric at enthought.com (eric)
Date: Mon, 8 Apr 2002 15:33:08 -0400
Subject: Pyrex and weave.accelerate
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1
 <00e101c1dd14$5620a930$6b01a8c0@ericlaptop> <3CB1FA63.6ADACEDE@llnl.gov>
Message-ID: <034701c1df34$36d24d50$777ba8c0@ericlaptop>

> 
> Where Eric alludes to scientific needs, he is
> referring to array/field acceleration.  Now
> the initial target of weave is Numpy arrays...
> 
> 
> def div(u):
>   d = zeros(len(u),'d')
>   for i in range(len(u)-1):
>      d[i] = u[i+1]-u[i]
>   return d
> 
> So, weave.accelerate tries to replace function calls with
> equivalent C/C++ code.

I'm also alluding to creating C/C++ versions of Python functions
that have a standard C calling convention (instead of the self, args)
stuff that has to be parsed on function calls.  If a pure C version
is possible, scientific codes that call the function a zillion 
times can do so at the speed of C instead of having to callback
into python.  This is important for a lot of scientific codes
that have been wrapped for use in Python.

eric







From cliechti at gmx.net  Wed Apr 10 18:05:03 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Apr 2002 00:05:03 +0200
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com>
Message-ID: <Xns91ED10E9E5E7cliechtigmxnet@62.2.16.82>

marilyn.mix at omitron.com (M. J. Mix) wrote in 
news:7226483f.0204101216.2c8fcda4 at posting.google.com:
> Is anyone using Python as a spacecraft ground system scripting
> language, or onboard (embedded) language?

i once runned it on a embedded linux PPC 60Mhz worked basicaly fine. there 
is a Palm port (pippy). don't expect it to run on a embedded chip with 64k 
RAM or so.

in what time frame do you think? don't expect from python to react within 
microseconds but its well suited for a timeframe above, say, 100ms (stronly 
dependant on your hardware and sw design). i have an app with hardrealtime 
constaints and 180ms reaction time. it works fine on a pentium class PC, 
never tried slower ones.
 
> Pros and cons when compared with TCL or JAVAScript.  Just 
> some real life, honest opinions.

i don't realy know those. form the small amount of JS that i did for web 
programming, i can only say that python is much more powerful and simpler.
when it come to GUI and more you have powerful libraries for python but not 
for JS.
 
> The usual issues: reliability and maintenance effort in applications. 
> Also, use as a scripting language by non-programmers. (I saw the post
> on indenting, so I'm concerned about that.)

for maintanance and non-programmers is python very well suited!

its also very reliable. i run a pure python webserver since weeks 24/7 
without problems. the exception mechanism makes it easy to write fault 
tolerant SW.

don't care about the indentation - thats easy to learn.
some people don't like it and others like to fight problems by mixing tabs 
and spaces in an editor with Tab=4 spaces. if you stick with spaces the 
indentation-is-syntax in python is much more saver than counting braces.

chris

-- 
Chris <cliechti at gmx.net>



From carole at autosim.no  Wed Apr  3 08:10:26 2002
From: carole at autosim.no (Carole Valentin)
Date: 3 Apr 2002 05:10:26 -0800
Subject: Swig/Python: data type problem
References: <9d8d2e93.0204020314.362621a4@posting.google.com> <uajfhvcudhk5cb@corp.supernews.com>
Message-ID: <9d8d2e93.0204030510.6bd7dffd@posting.google.com>

"Lyle Johnson" <ljohnson at resgen.com> wrote in message news:<uajfhvcudhk5cb at corp.supernews.com>...
> > But in C++, we expect a data type:
> >
> > void MyObject::read(String* namefile)
> >
> > I tryed with 3 data types: char / string / string*
> > Each time, I have the same error:
> >
> > File "c:\python21\Module1.py", line 172, in read
> >     val = apply(Interfc.RoadNetwork_lecture,args)
> > TypeError: Type error. Expected _p_String
> >
> > Do you have any suggestion?
> 
> Well, the short answer is that you need to specify a typemap for String
> pointers in your SWIG interface file. Since I'm not familiar with your
> library's String class, let's just assume that the String class has a
> constructor that takes a const char * for the string's contents. In that
> case you could write an "in" typemap:
> 
> %typemap(in) String * {
>     $1 = new String(PyString_AsString($input));
> }
> 
> For more details, please read the SWIG user manual (especially the stuff
> about typemaps).
> 
> Hope this helps,
> 
> Lyle

Hello,

Thank you for your answer.
I have read the swig documentation, but I have still some more questions:

- where I have to put the code:
%typemap(in) String * {
    $1 = new String(PyString_AsString($input));
}
in the .i file?
- what is $1 and $input ? They are not recognized...
- is it the only piece of code I must write to solve my data type problem?
for instance, do I have to write something in the read function?

Thanks again,
Carole.


From fperez528 at yahoo.com  Thu Apr 18 13:15:18 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 18 Apr 2002 11:15:18 -0600
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com>
Message-ID: <a9mv1o$7jh$1@peabody.colorado.edu>

Jaco Schieke wrote:

> Being new to interactive python, I have the problem that rlcompleter
> does not seem to be able to resolve the items in a list:
> 
> This works as expected:
> ======================================
>>>> import readline, atexit, rlcompleter
>>>> readline.parse_and_bind('tab: complete')
>>>> a=[[],"",1]
>>>> a.<tab>
> a.append   a.count    a.extend   a.index    a.insert   a.pop
> a.remove   a.reverse  a.sort
> 
> 
> Now is this also supposed to work?
> ==================================
>>>> a[0].<tab>
> .   ..         <---- this is the output I receive
>                      instead of the properties of an []
> 
> 
> 'cause this works:
> ==================================
>>>> b=a[0]
>>>> b.<tab>
> b.append   b.count    b.extend   b.index    b.insert   b.pop
> b.remove   b.reverse  b.sort

I was going to ask the same thing as I'm inclined to think this is a bug: 
readline has '[' and ']' as part of its internal list of word delimiters, so 
when you do 
a[0].<tab>
it breaks the word at ']' and sees a bare '.', hence the completions you get. 
So far so good.

BUT: if you change the list of delimiters (with readline.set_delimiters()) to 
remove '[' and ']', the problem persists. That's the part I think to be a bug 
(I could be wrong, but I'd like to understand why if that's the case). At 
this point it shouldn't break at ']' anymore, and the completion should be 
done on the full 'a[0]'. I put print statements in the rlcompleter code to 
see this and indeed, readline _is_ breaking on ']' even after removing the 
character from the list of delimiters.

Unfortunately the completion code is in C and I don't have the sources around, 
so I haven't had the chance to take a look at this.

I'd love to hear from someone who knows the code as to this behavior really is 
a bug or not before shooting off to SF and looking like an idiot :)

Cheers,

f.


From peter at seahaze.demon.co.uk  Fri Apr  5 14:18:19 2002
From: peter at seahaze.demon.co.uk (Peter Hayes)
Date: Fri, 5 Apr 2002 19:18:19 +0000 (UTC)
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it> <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20>
Message-ID: <slrnaaru18.2vl.peter@vlad.seahaze>

In article <Xns91E77C8D2BE34PaCmAnRDLM at 161.48.128.20>, Fred Pacquier wrote:
> Alex Martelli <aleax at aleax.it> said :
> 
>> Reports of pystones and other Zaurus/SL5000 Python/PyQT experiences
>> are VERY welcome -- it's a really sexy little thingy and I'm SORELY
>> tempted to have a fling, but getting it in Italy isn't easy as yet...
> 
> So am I, since my preferred platform of past years (Psion) is now dead in 
> the water. But just yesterday I read a review of the just-released (in the 
> US) Zaurus, and the last line mentions that Sharp is "thinking about 
> introducing the new Zaurus line in Europe"... next year. So if you want one 
> soon you'll probably have to take your chances and order one overseas, and 
> make do with the Qwerty keyboard (probably less of a problem for you than 
> for us French though :-)

They're being advertised as being "available shortly" by some UK
dealers, which is why I was wondering about running Python on them.

-- 

Peter


From insta at no.spam  Fri Apr 19 23:38:37 2002
From: insta at no.spam (Matt)
Date: Sat, 20 Apr 2002 03:38:37 GMT
Subject: Question about IDLE
Message-ID: <0p5w8.17527$CH.4180@rwcrnsc52.ops.asp.att.net>

I'm trying to learn Python, and I was wondering if the IDLE program was 
sufficient enough for basic programming, or should I find some sort of text 
editor?
I'm currently using a win32 box.

Also, any links to tutorials or general information about Python that is 
not listed on the Python website would be much appriciated.

-Matt


From jajvirta at cc.helsinki.fi  Sat Apr  6 00:56:08 2002
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 6 Apr 2002 05:56:08 GMT
Subject: Python-list digest, Vol 1 #10324 - 13 msgs
References: <E16tGLp-0007ou-00@mail.python.org> <mailman.1018014793.1604.python-list@python.org> <axir8.52642$pT1.1485864@news1.tin.it>
Message-ID: <slrnaat3fn.3cd.jajvirta@sirppi.helsinki.fi>

Fri, 05 Apr 2002 14:29:26 GMT Alex Martelli wrote:

> For example, instead of
> 
> Button(buttonsFrame, text='Squirk', command=lambda: self.squirkIt(23, 45))
> 
> you can code:
> 
> def specificSquirker(): self.squirkIt(23, 45)
> Button(buttonsFrame, text='Squirk', command=specificSquirker)
> 
> It's generally not hard to come up with a name which helps readability.
> I think un-lambdaing your code in this way is a good thing...

I agree.  Furthermore, I don't even think that un-lambdaing that way is
about the maintainability issue per se (ie., 'the "functional" features
make python code unreadable' [1]); it's just that abstracting everywhere
is a good thing.  The highest level of logic at that point is not about
lambdas, it's about "squirking" and that's what you should see in the
source code.  Making routines and giving them names that represent the
highest level of logic, which is not about the actual programming
language code statements [2], is a very powerful tool (available in
other languages too, of course :-). 

(There are other issues, but this abstraction thingy has been
haunting me last couple of days and I had to let out the steam. ;-)

[1] Which I might or might not agree.

[2] Unless writing a programming language compiler or interpreter.



From max at alcyone.com  Tue Apr 30 05:08:58 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 30 Apr 2002 02:08:58 -0700
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <3CCE5F2A.1103E744@alcyone.com>

Ralf Juengling wrote:

> when I want to make sure within a function that it can deal with
> the arguments passed, the only choice is to check it's type via
> 'isinstance()'. However, in general the type of an argument is
> not the crux but whether the argument supports a certain
> 'interface' or 'protocol'. Thus, a function 'hasinterface()'
> would appear more natural (more 'Pythonic') to me.
> 
> So, what is the proper, 'pythonic' way of argument checking?

The "Pythonic" way of argument checking is to simply use it and catch
the exceptions that will be thrown if it doesn't behave as planned.

Depending on what you want, some of the is... functions in the operator
module might be to your liking.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From jb at cascade-sys.com  Mon Apr 29 21:10:47 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 18:10:47 -0700
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <20020429174425.GN16217@devel.trillke> <mailman.1020115270.9666.python-list@python.org> <aaknh8$1c4$1@peabody.colorado.edu>
Message-ID: <3CCDEF17.E4D1DC88@cascade-sys.com>

Fernando P?rez wrote:

> James J. Besemer wrote:
>
> > THREE if you include LAMBDA.
>
> No, please heed other advice and use the MA approach. Please note that I'm not
> dissing this post on its own merits, it's a nice discussion, it just is
> ill-adapted to numeric processing. Numeric provides its own _fast_ way to
> manipulate arrays with invalid data, by all means use it. After all, speed is
> one of the big reasons why you use Numeric to begin with (you could do most
> things in it with regular lists, if you had the age of the universe to wait).

I would agree with you that it might be worth the extra trouble and effort to
download, install and learn to use Numerics IF the user in question has a big,
production application to run.

Failing that, if it's just a little application, say, filtering a small array of
numbers in a histogram, then I humbly submit that one of the several "stock"
solutions would do just nicely if not better.

Afterall, even on a crummy old sub-200 MHz PC, IIRC, the list solution was able to
so process on average about 100K elements per second (25% empty).  Probably could
do 10X as fast on a new PC.

Too, I think the original question was about Language technique, as the poster
already had in hand a solution that worked.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From christophe.delord at free.fr  Wed Apr 17 15:20:44 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Wed, 17 Apr 2002 21:20:44 +0200
Subject: Trinary operator?
References: <mailman.1019066614.29792.python-list@python.org>
Message-ID: <20020417212044.1df4d4bd.christophe.delord@free.fr>

You have to be very carefull with and/or. If you evaluate A and B or C, if A is true and B is false, the result is C !!!

For example, "number=='zero' and 0 or 1" returns 1 whatever the value of number

>>> number='one'
>>> number=='zero' and 0 or 1
1
>>> number='zero'
>>> number=='zero' and 0 or 1
1
>>> 

Christophe.

On Wed, 17 Apr 2002 13:02:56 -0500
"Mark McEahern" <marklists at mceahern.com> wrote:

> > Not that it's a big deal since there's many ways to do that, but can I use
> > something like
> >     verboseGender = (gender == 'm') ? 'male' : 'female'
> > in Python?
> 
> People seem to use and/or for that:
> 
> >>> gender = 'm'
> >>> verbose = (gender == 'm' and 'male' or 'female')
> >>> verbose
> 'male'
> >>> gender = 'f'
> >>> verbose = (gender == 'm' and 'male' or 'female')
> >>> verbose
> 'female'
> 
> // mark
> 
> 


-- 
Christophe Delord
http://christophe.delord.free.fr/


From Alex at DDeus.de  Sun Apr  7 16:04:31 2002
From: Alex at DDeus.de (Oleksandr Kozachuk)
Date: 7 Apr 2002 20:04:31 GMT
Subject: Lex and Yacc for Python
Message-ID: <slrnab19hf.23u.Alex@Tron.ddeus.lan>

Hi all!

I need something, comparable to Yacc/Lex. Or something for
Context-Free-Grammar. Can any one recommend me anything?


Thanks,
Alex.


From ajs at ix.netcom.com  Tue Apr  2 09:07:18 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 2 Apr 2002 09:07:18 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <001301c1da4f$b56d5f40$0334fea9@carol>

Steve writes -

>I think the word you're looking for is "pilloried".

Thanks Steve, you got right to my point.

Steve also writes -

>If the PEP is accepted ist will be because you are in a >minority who think
it will not result in improvement
>to the language (or at least will not degrade it at all).

Shamefully dishonest. You know quite well those are
not the rules of the game. But you are on the right side,
which is the important thing.

I by the way am not criticizing the rules of the game.

Have done nothing to contribute to Python, and don't
consider myself to have standing to do so. Nor do I have
an alternative to suggest.  Vote on language design -
silly.

But I can add my commentary as the game plays itself
out.

Art




From aleax at aleax.it  Thu Apr 11 04:48:28 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 08:48:28 GMT
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <oAfs8.26107$Sw5.2561@atlpnn01.usenetserver.com> <3CB35101.AF4FA8F7@noaa.gov> <slrnab9811.hm3.huaiyu@gauss.almadan.ibm.com>
Message-ID: <w5ct8.6353$vF6.199608@news2.tin.it>

Huaiyu Zhu wrote:
        ...
> I don't see any situation that you need "if bool(some_expression)".

Right, it's just as redundant as:

    for item in iter(some_iterable):

or

    print str(some_expression)


I.e., if or while intrinsically use bool, just like for uses iter and
print uses str.  Now, what we need is a release 1.0.1 of this.py
that, right below "explicit is better than implicit", adds a
"but-clause" of (I suggest) "but superfluous redundancy and
repetitive boilerplate aren't" (self-referential, of course:-).
Note this also helps ward against "if bool(x)==True:" &c...


Alex



From frr at easyjob.net  Tue Apr 23 09:40:19 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 15:40:19 +0200
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com>
Message-ID: <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com>

On Tue, 23 Apr 2002 08:21:30 -0400, "Steve Holden" <sholden at holdenweb.com>
wrote:


>I don't think you need the string quotes around the parameter markers - the
>module determines the required type, and in many cases will perform
>conversions for you as necessary. Since you are using a list of tuples, you
>should really be calling executemany() rather than execute, though most DB
>API modules will accept lists of tuples. Try this:
>
>curs.execute(
>    """insert into customer_info (email, url, referrer, first_visit)
>    values ( %s, %s, %s, %s)""",
>    ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07
>12:32:00')
>)

It gives an error:
>>> curs.execute("""insert into customer_info (email, url, referrer, first_visit) values ( %s, %s, %s, %s)""", ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
Traceback (most recent call last):
  File "<pyshell#53>", line 1, in ?
    curs.execute("""insert into customer_info (email, url, referrer,
first_visit) values ( %s, %s, %s, %s)""", ('billg at microsoft.com',
'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
  File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line 230, in
execute
    raise error, msg
mysqldb.error: (1064, "You have an error in your SQL syntax near
'@microsoft.com, www.easyjob.net, www.google.com, 2001-03-07 12:32:00)' at
line 1")
>>> 

If I add the string quotes again, I get the same AttributeError exception I
was getting before:

>>> curs.execute("insert into customer_info (email, url, referrer, first_visit) values ( '%s', '%s', '%s', '%s')", ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
Traceback (most recent call last):
  File "<pyshell#56>", line 1, in ?
    curs.execute("insert into customer_info (email, url, referrer,
first_visit) values ( '%s', '%s', '%s', '%s')", ('billg at microsoft.com',
'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
  File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line 227, in
execute
    af =af+self.__res.affectedrows()
AttributeError: 'NoneType' object has no attribute 'affectedrows'
>>> 



>It may or may not make a difference. What's the difference between
>CompatMySQLdb and MySQLdb?

I'm not sure. I haven't been able to find MySQLdb. I found this Compat thing
in MySQL-python-0.9.1.win32-py2.2.exe at
http://sourceforge.net/project/showfiles.php?group_id=22307&release_id=51760

Where can I find MySQLdb for win32 _and_ Linux? O:-)




-----------------------
Fernando Rodriguez


From timr at probo.com  Wed Apr 24 02:05:59 2002
From: timr at probo.com (Tim Roberts)
Date: Tue, 23 Apr 2002 23:05:59 -0700
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net>
Message-ID: <0kiccu46a2580bbohtpavbla241q95etj0@4ax.com>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote:

>In article <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v at 4ax.com>, Fernando Rodr?guez wrote:
>> No. Step (b) processes the order and does give feedback. Step (a) saves in a
>> db some info conatined in a cookie (date of first visit and first referrer). I
>> want step (a) to be seamless.
>
>Must admit I don't have a clue what you mean. Why aren't scripts (a)
>and (b) in the same script? If they really must be separate scripts,
>why doesn't script (a) just output nothing at all to stdout?

My guess is that the (b) option is a URL at a third-party order processing
site that is not under his control.  If my guess is correct, that means the
original poster is actually hoping to have a single <FORM> fire to two
different URLs.  As far as I know, that is not possible.

Theoretically, it would be possible to have the <FORM> fire a URL on the
home system with the order processing URL, and have that script generate
its own HTTPS connection to the order processing site, but that means the
web site user will not see that his connection is secure.  I would never
enter a credit card number in that case.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From ark at research.att.com  Fri Apr  5 10:23:26 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 5 Apr 2002 15:23:26 GMT
Subject: newbie: declare function before definition?
References: <Pine.LNX.4.44.0204050953040.17812-100000@kramer.thekramers.net>
Message-ID: <yu993cyanp8x.fsf@europa.research.att.com>

David> It appears to me that in Python, functions need to be defined
David> before they are called (physically in the file, I mean).  This
David> means any piddly little utility functions have to be before my
David> main routine, and the reader has to wade through them before
David> getting to the main routine.

Not really.

If you call a function, it must be defined at the point the call
is executed.  However, if that call is in another function, it
is not executed until you call that other function.

So, for example:

        def f():
                g()

        def g():
                print "I am in g"

There is no problem with f calling g, as long as you don't attempt
to execute f until after you have defined g.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From geoff at gerrietts.net  Fri Apr 12 15:16:39 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 12 Apr 2002 12:16:39 -0700
Subject: Queue (Threads)
In-Reply-To: <pan.2002.04.12.21.10.57.987428.703@thomas-guettler.de>
References: <pan.2002.04.12.21.10.57.987428.703@thomas-guettler.de>
Message-ID: <20020412191639.GA12191@isis.gerrietts.net>

Quoting Thomas Guettler (pan-newsreader at thomas-guettler.de):
> I tried to find examples for the Queue class with google, but
> no look. Has someone links to exmamples or hints?

The code below allows a main thread to delegate a Command to a
subthread. The producer (main thread) calls 

  threads.queue(func, (arg,arg,arg,matey), {'kw': 'kval'})

The consumer threads (StoutpamThread instances) pluck functions off
the queue and run them all day long.

Enjoy,
--G.



""" threads.py
    Implements threading for Stoutpamphlet.

    To delegate a process to a thread, call threads.queue(func, args, kw)

    To check whether you're inside a thread or not, (for example to decide
    whether you want to lock the gui or not) call threads.in_subthread()
"""

has_threads = 0
try:
    import threading, Queue
    has_threads = 1
except:
    pass

if has_threads:
    
    ActivityQueue = Queue.Queue(5)

    def queue(func, args=(), kw={}):
        ActivityQueue.put((func,args,kw))

    class StoutpamThread(threading.Thread):
        def run():
            while (1):
                func, args, kw = ActivityQueue.get()
                apply(func,args,kw)
            # end loop
        # end def

    threads = []
    for i in range(3):
        t = StoutpamThread(name="Thread %d" % (i,))
        threads.append(t)
    # end loop

    def in_subthread():
        t = threading.currentThread()
        return ("MainThread" == t.getName())
else:
    
    def queue(func, args, kw):
        apply(func, args, kw)

    def in_subthread():
        return 0
    



-- 
Geoff Gerrietts             "I am always doing that which I can not do, 
<geoff at gerrietts net>     in order that I may learn how to do it." 
http://www.gerrietts.net                    --Pablo Picasso




From sehamran at frisurf.no  Wed Apr 24 04:37:47 2002
From: sehamran at frisurf.no (seh)
Date: 24 Apr 2002 01:37:47 -0700
Subject: Data from Snack
Message-ID: <7557b51f.0204240037.e773443@posting.google.com>

I would like to record some data from the sound card using the Snack
module. Then get the recorded data from the Snack sound class into a
Python variable (array?). Then process the data. Anybody know how to
do this?

Any other ways to do this in Python?

Svein-Erik


From m_mariappanX at trillium.com  Thu Apr 25 03:34:30 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Thu, 25 Apr 2002 00:34:30 -0700
Subject: is there any dedicated mailing list fo python PMW project?
Message-ID: <53A7943A5BD8D411B6930002A5073155022D7498@bgsmsx90.iind.intel.com>

Python user  comunity,

is there any dedicated mailing list fo python Pmw project?

Maharajan




From cliechti at gmx.net  Thu Apr 11 18:20:32 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 12 Apr 2002 00:20:32 +0200
Subject: ASCII to HEX conversion
References: <34e98ec2.0204111415.504d246f@posting.google.com>
Message-ID: <Xns91EE3AF2DC0cliechtigmxnet@62.2.16.82>

leisinmr at purdue.edu (Matt) wrote in 
news:34e98ec2.0204111415.504d246f at posting.google.com:

> I need to write a program that converts ASCII test into the
> hexidecimal equivalent.  It would help if it could convert to binary,
> too.
> 
> Example:
> 
> ASCII = "A"
> HEX = 41
> 
> I tried using the hex() function, but it does not work.  If anyone
> could tell me how to do this without making a lookup table of all the
> ASCII values.

use the ord function to get the ascii code of a character. you can convert 
that number to any format you want. 

>>> [hex(ord(x)) for x in "ABCDE"]
>>> ["%02X" % (ord(x)) for x in "ABCDE"]

chris

-- 
Chris <cliechti at gmx.net>



From aleax at aleax.it  Mon Apr 15 15:07:02 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 19:07:02 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de> <Prau8.17347$vF6.558738@news2.tin.it> <eppstein-15B457.11430415042002@news.service.uci.edu>
Message-ID: <qxFu8.18112$b62.514495@news1.tin.it>

David Eppstein wrote:

>?Is it really appropriate to design a language so that only "typical"
>?usages are simple, and everything else is made more difficult?

Yes, if the amount of "more difficulty" is reasonable, as it would be
here -- using built-in functions such as bitand, bitnot, and so on.

Languages must be small -- THAT is crucial. ?Every aspect that is
hardwired into a language must fully pull its weight. ?Python does
pretty well at this -- so the few imperfections stand out more.
`x` (backquotes) is another example, a silly duplication of repr.


>?My most recent Python program (a testbed for some signal encoding
>?schemes) made extensive use of both bit twiddling and simple generators.
>?I'm very happy that Python was available for that purpose.


And me, I've done lot of combinatorics with Python -- so why couldn't
I have a postfix ! to indicate factorial? ?Why did I have to write out (not
even a builtin, o heavens) fact(x) rather than the neat
???????? x!
??? ?Etc, etc. ?Everyone and his cousin has their own peculiar uses, and
if the language tried to wire them all in it would be as unusable as your
worst nightmares. ?The language should be lean and mean, with wired
support for _typical_ uses only, and libraries taking up the slack -- that
is *basically* Python today, btw, I'd only like it more so (in my dreams:-).

If we hadn't "wasted" `` as a duplication of repr, we could maybe use
them Haskellishly to make any function syntactically into an operator
and vice versa -- a `flep` b as a synonym for flep(a, b) -- but maybe
we don't need that syntax sugar all that badly:-).


Alex



Alex



From pinard at iro.umontreal.ca  Tue Apr 23 14:52:23 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 23 Apr 2002 14:52:23 -0400
Subject: Python call tree tool?
In-Reply-To: <7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>
	<lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk>
	<7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>
Message-ID: <oqg01mqmc8.fsf@titan.progiciels-bpi.ca>

[Robin Becker]

> I guess I can use inspect to figure out the functions/classes that are
> declared in an arbitrary module, but it seems harder to determine if
> these things are actually referenced in other modules. For that I assume
> I will have to do some code analysis.

It may be difficult to do in general.  Suppose you have a `maker' module
containing the definition of class `Class' (having a rich set of methods), a
`processor' module with a `do_it' function, and a `driver' module containing:

    def main():
        import maker, processor
        processor.do_it(maker.Class())

All references to `Class' methods will occur in `processor' module, but
you have no clues in `processor' about the type of the argument of `do_it'.
Moreover, in `main' above, `make.Class()' may well have been stored into an
intermediate variable or more structured container, before being actually
used as an argument to `processor.do_it'.

Such usages do not appear uncommon at all to me, and it surely requires an
extremely clever and sophisticated code analyser to find out what really
happens at run-time.  This is only a small dark side to a great virtue of
Python, by which a variable, with no fixed type in advance, gives simple
access to all methods of the class of the object it currently refers to.

Other languages might give you good tools for tracking dependencies,
but there is a high price to pay in flexibility, for making these possible.

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




From kzaragoza at attbi.com  Mon Apr  1 13:33:14 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: Mon, 01 Apr 2002 18:33:14 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <3CA62CD8.6ABD44D0@accessforall.nl> <m3adsp8zie.fsf@mira.informatik.hu-berlin.de> <3CA6F012.E2631003@accessforall.nl> <m3adspm1kl.fsf@mira.informatik.hu-berlin.de> <3CA79401.CFA41B65@python.org>
Message-ID: <slrnaaha69.nn5.kzaragoza@localhost.localdomain>

In article <3CA79401.CFA41B65 at python.org>, Guido van Rossum wrote:
> But see also my comments on these two in
> http://www.python.org/doc/essays/pepparade.html; neither PEP is likely to
> have any effect on Python's future.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)

Guido,

I have to ask one question that I haven't seen brought up yet: What
does PEP 285 imply for the future of Python?  Do you intend to
eventually change the concept of what is true?

Specifically, I refer to the Python Language Reference
(http://www.python.org/doc/2.2/ref/lambda.html):

"""In the context of Boolean operations, and also when expressions are
used by control flow statements, the following values are interpreted
as false: None, numeric zero of all types, empty sequences (strings,
tuples and lists), and empty mappings (dictionaries). All other values
are interpreted as true."""

Do you intend to eventually (NOTE: I say eventually, I know it's not
immediate) modify this to read that only the value True is true and
the value False is false?  If this is the case, I believe this adds
needless complexity to code whereas (in a simple case)

if foo:
    ...do whatever...

would now have to read

if bool(foo):
    ...do whatever...

in order to convert None, empty sequences/mappings, and non-integer
zeros to boolean values.  It may seem like a small change, but it
would add unnecessary verbiage to what is already fairly obvious in
control constructs and other boolean expressions.

If you don't intend to modify this concept, then what's the real gain
for the new boolean type?  The concept is already fairly clear, and
would be needlessly muddied by this extra type that holds one of two
values.  (Try explaining to a newbie that it's actually one of two
pre-allocated instances!)

The only real advantage I can see in the latter case is to make some
code a bit more self-documenting.  If you assign a value of False or
True to a variable, it's pretty clear what that variable is intended
to hold.  However, given the firestorm of discussion that has erupted
around the details of this PEP, is this small gain really worth the
effort?  Is there some other advantage I'm missing?

Kris

--
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org



From tdelaney at avaya.com  Wed Apr  3 18:48:25 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 4 Apr 2002 09:48:25 +1000 
Subject: [Python-Dev] ACCEPTED: PEP 285
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A24C@natasha.auslabs.avaya.com>

> From: Neal Norwitz [mailto:neal at metaslash.com]
>
> This code:
> 	True = 1
> 	False = 0
> 
> 	def test(a):
> 	    if a is True: print 'True'
> 	    if a == False: print 'False'
> 
> generates these warnings:
> 	test.py:2: Should not assign to True, it is (or will 
> be) a builtin
> 	test.py:3: Should not assign to False, it is (or will 
> be) a builtin
> 	test.py:6: Comparisions with True are not necessary
> 	test.py:7: Comparisions with False are not necessary

I would change that to ...

# only generated in Python 2.3 or higher

test.py:2: Should not assign to True, it is a builtin
test.py:3: Should not assign to False, it is a builtin

# only generated in Python 2.3 or higher,
# or if there are constants in scope with the names
# 'TRUE', 'FALSE', 'True', 'False', 'true', 'false'
# and the appropriate name should be included in the
# warning.

test.py:6: Comparisons with [TRUE|True|true] may give incorrect results
                   ^ extra i ;)
test.py:7: Comparisons with [FALSE|False|false] may give incorrect results
                   ^ extra i ;)

and would also add a warning for code in 2.3+ for code like

TRUE = <something>
FALSE = <something>
true = <something>
false = <something>

test.py:2: TRUE appears to be a boolean constant - builtin True should be
used.
test.py:3: FALSE appears to be a boolean constant - builtin False should be
used.
test.py:4: true appears to be a boolean constant - builtin True should be
used.
test.py:5: false appears to be a boolean constant - builtin False should be
used.

Tim Delaney




From SBrunning at trisystems.co.uk  Fri Apr 26 06:11:16 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 26 Apr 2002 11:11:16 +0100
Subject: Python doesn't know how to do math?
Message-ID: <31575A892FF6D1118F5800600846864DCBD08E@intrepid>

> From:	Chris Spencer [SMTP:clspence at one.net]
> 	I want to thank everyone who responded to this.  I read the floating
> point tutorial and now understand the issues.  I wasn't really looking for
> that
> much precision anyways (2 decimal places was all I really need).  So now
> that I
> know, I shant ask again.
 
Hmmm. It sounds to me like you don't want to use floats at all. Have a look
at <http://starship.python.net/crew/aahz/FixedPoint.py> - you might find it
more useful for holding stuff like currency values.

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




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




From phil at river-bank.demon.co.uk  Thu Apr 18 19:14:35 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Fri, 19 Apr 2002 00:14:35 +0100
Subject: PyQT, how to compile SIP on MAC? (long)
References: <3C9A62C2.ABA9AE5D@gmx.de> <mailman.1016790154.12480.python-list@python.org> <de404184.0204181134.6bd4bc2c@posting.google.com>
Message-ID: <3CBF535B.BCF1B67A@river-bank.demon.co.uk>

Barron wrote:
> 
> Phil Thompson <phil at river-bank.demon.co.uk> wrote in message news:<mailman.1016790154.12480.python-list at python.org>...
> > Ingo Linkweiler wrote:
> > >
> > > Hi,
> > >
> > > I am willing to use PyQt on Mac OS X
> > >
> > > But I was not able to compile SIP --
> > > http://www.river-bank.demon.co.uk/software/sip-3.0.tar.gz
> > >
> > > Are there any hints you can give to compile SIP with success?
> > >
> > > With windows and linux it works fine.
> >
> > I've had reports of SIP building fine, but PyQt having problems with
> > pyuic.
> >
> > I can't test any of this myself because I don't have the tools (cue
> > another begging letter for somebody to donate me a Mac and a development
> > environment).
> >
> > Phil
> 
> I, too, would love to get PyQt running on OS X! Any help is
> appreciated.
> 
> Phil, I wish I could ship you a machine to do the work, but until I
> win the lottery, my resources are quite limited <g>. However, I would
> be willing to contribute a few dollars toward the effort....
> 
> It seems like the first step is compiling SIP.
> So, I ran the .configure script. I received a few "no's" in the
> output. Is this a problem? (full output at the end of this message) It
> didn't seem to be, so I ran the "make check" to run the checks. (full
> output at the end of this message) The only errors seemed to come at
> the end:

I suggest you join the PyKDE mailing list.

I have produced v3.2rc1 for MacOS in
http://www.riverbankcomputing.co.uk/download/snapshots/ which gets
further. I will release v3.2rc2 this weekend with further changes.
Several people are helping me with this at the moment - more helpis
welcome.

Phil




From ldw at us.ibm.com  Thu Apr  4 09:53:00 2002
From: ldw at us.ibm.com (Larry Whitley)
Date: Thu, 4 Apr 2002 08:53:00 -0600
Subject: No output from threads
Message-ID: <a8hq64$114s$1@news.rchland.ibm.com>

On a windows 2k system I have a single master thread, spawning two worker
threads and ending. The master thread is started from the command prompt
(run.py) and the worker threads use "start /belownormal /wait ..." to run a
sequence of C++ programs that I have writen that take a long time to execute
(they're simulation performance models). The start commands have this form:
"start /belownormal /wait myprog.exe -option 3.0 >out3.txt". The two
threads, one for each processor, will work their way through a queue or work
provided by the master thread when the worker threads started.

Unfortunately, while the programs appear to run, at completion their output
files are empty. The files exist but contain no data as a result of the run.
When I run the same program in a non-threaded fashion from a batch file
(without the start statement) I get output in the output files.  In this
case the command looks like this "myprog.exe -optoin 3.0 >out3.txt".

My workstation (windows 2k pro) has two processors so I'd like to use
python's threading to let me run two of my simulations at the same time.
But, I would also like output from the runs.  :-)   Here's the code that
starts the C++ jobs.  It's called by the run() method in the
MyThread(threading.Thread) class.  self.prt() is just adds a thread safe
wrapper around the normal python print command.

    def runJob(self, job):
        command = "start /belownormal /wait "
        command += job[0] + " " # program
        command += job[1] + " " # options
        command += ">" + job[2] # outfile
        self.prt( self.getName() + " is starting " + command )
        os.system( command ) # submit the command to the system



Larry

PS: My reason for using the start command is to let me set the priority of
the jobs to something less than normal.  That way I can continue to use the
workstation without undue delays while the simulations run.





From max at alcyone.com  Tue Apr 16 03:30:37 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 00:30:37 -0700
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
Message-ID: <3CBBD31D.AA2E653C@alcyone.com>

William Dode wrote:

> Is there a python api somewhere ?
> Google is known to use python, maybe we must write to them to ask a
> python
> wrapper for their api ?

	http://diveintomark.org/projects/#pygoogle

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From logiplexsoftware at earthlink.net  Fri Apr  5 17:41:26 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 5 Apr 2002 14:41:26 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
In-Reply-To: <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
	<Xns91E75A62ACE63JamesLLugojcom@216.148.53.96>
Message-ID: <20020405144126.5cad70e9.logiplexsoftware@earthlink.net>

On 5 Apr 2002 16:48:57 GMT
James Logajan wrote:

> The answer to this isn't as interesting to me as _when_ will this
tinkering 
> end? Something that isn't up for debate: because of the problems I have 
> actually encountered with making code work with the many extant versions
of 
> Python, I will no longer be doing any new projects in the language and am

> not recommending Python to anyone. The language tinkering is one of
several 
> reasons I left Python and this newsgroup many months ago. (I have 
> resubscribed recently to find out if there was a way to rework an old 
> Tkinter app of mine that people have asked for some changes to, and will
be 
> exiting here in a few days.)

Ah, you must have chose APL or Forth, since they're pretty much in the
feature-lock stage.  It can't be C since they've also added the damn
boolean. 

Seriously though, if this is a big concern for you (and it might well be),
target the 1.5.2 language spec and the chances of your code not working
even in 2.2 are slim, so I'm not sure what would be compelling enough to
make you abandon Python.  I myself don't like every change in Python (or
even some things that have been in place since 1.5.2) but it's still an
excellent language.  I suspect you are masking some other complaint with
this one.

Technology that ceases to change is already dead.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From SBrunning at trisystems.co.uk  Wed Apr 24 05:04:02 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 24 Apr 2002 10:04:02 +0100
Subject: Python call tree tool?
Message-ID: <31575A892FF6D1118F5800600846864DCBD042@intrepid>

> From:	Robin Becker [SMTP:robin at jessikat.fsnet.co.uk]
> I searched in vain for a tool that can create a call tree or  inverted
> call tree for python modules/packages. It seems inconceivable that the
> Python community hasn't created such a tool so I must be looking in the
> wrong places. 
> 
> Any one know where such a beast is?
 
Might it be worth looking at py2exe and/or Installer? I would have thought
that these tools must create a call tree for internal use.

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




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




From phr-n2002a at nightsong.com  Sun Apr 14 01:44:35 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 13 Apr 2002 22:44:35 -0700
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org>
Message-ID: <7xn0w696m4.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> > A lot of stuff gets done because it needs to get done, rather than
> > scratch an itch.
> 
> Do you have a concrete example?  I can't think of one.  In a world of
> volunteers, "need" is in the judgment of each volunteer, and counts as
> scratching their own itches to me.  In contrast, Guido said 1.5.2 "needed" a
> bugfix release about 6 months after it was released, and nobody volunteered,
> and 1.5.2 never got a bugfix release.  Volunteers can't be forced to do
> anything; they do what they want to do, and for people playing the game
> that's great for everyone.

I was thinking about various parts of the GNU project, not
specifically about Python.  For example, the whole GNOME system was
written because the FSF found the Trolltech Qt license unsatisfactory
(the Qt license later was fixed).  Or the Gutenberg project--people
scanning and manually correcting millions of pages of public domain
texts that they can easily get in printed form in bookstores, but they
feel the texts should be downloadable and somebody needs to do it, so
they do.

> > In that particular debate, a lot of people were asking for certain new
> > features to be LEFT OUT of Python. I don't understand how someone can
> > volunteer to contribute work, time, or money toward leaving something
> > out.
> 
> That was barely mentioned in the "crushing amount of debate over 'stability'
> [that] has gone by on Python-Dev the last week", so I don't know what "that
> particular debate" means to you, but it surely doesn't refer to anything in
> the paragraph you were responding to.  The debate on Python-Dev has about
> backward compatibility, and especially about producing bugfix releases for
> older releases for a much longer time. 

I was thinking of the PEP 285 discussion and the complaints about
too-frequent releases messing up compatibility because of new stuff
constantly being added--the remedy being to not add so much stuff.  If
you were referring to a different debate, I misunderstood.


From max at alcyone.com  Tue Apr 16 14:27:30 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 11:27:30 -0700
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
		<3CBBD31D.AA2E653C@alcyone.com>
		<20020416115225.1d9fa2d2.wilk@flibuste.net>
		<3CBC0078.C10C5F4@alcyone.com> <20020416165356.3aeb88eb.wilk@flibuste.net>
Message-ID: <3CBC6D12.6BE3A139@alcyone.com>

William Dode wrote:

> I mean that on the page of google they speak about api for java,
> visual.net and perl, but not for python. I would like to see python
> everywhere !

Well, that's because the Python interface was not done by them.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From cliechti at gmx.net  Sun Apr 28 15:47:20 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Apr 2002 21:47:20 +0200
Subject: email images
References: <a2972632.0204281142.30eb1e8@posting.google.com>
Message-ID: <Xns91FEDDE883787cliechtigmxnet@62.2.16.82>

les_ander at yahoo.com (les ander) wrote in 
news:a2972632.0204281142.30eb1e8 at posting.google.com:

> Hi,
> to avoid being a victim of spam bots, i would like to have a image 
> representation of my email to post on my web site.
> 
> is there a way to convert string to images using python?
> 

i think PIL can do it.
http://www.pythonware.com/products/pil/

chris

-- 
Chris <cliechti at gmx.net>



From mertz at gnosis.cx  Wed Apr  3 15:18:31 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Wed, 03 Apr 2002 15:18:31 -0500
Subject: ANN: Update to Gnosis package (xml_pickle, etc.)
Message-ID: <XO2q8kKkXEQA092yn@gnosis.cx>                

Subject: [Announce] Update to gnosis package
Date: Wed 04-03-2002

Building on the numerous significant improvements Frank McIngvale and
David Mertz made to xml_pickle for the 0.9 release of the Gnosis XML
Utilities package, this release adds some new features and some revised
packaging.

Obtain the current release at:

   http://gnosis.cx/download/Gnosis_Utils-0.95.tar.gz

Or check its parent directory for any releases that supercede 0.95.

 * For the 0.95 release adds "mutators" and type families to xml_pickle.
   The documentation at gnosis.xml.pickle.doc provides information on
   these concepts.  (Mutators replace helpers; if anyone has written
   custom helpers, they should be easy to convert to mutators.  Email
   frankm at hiwaay.net with any problems.)

 * Better handling of Numeric.arrays and array.arrays.

 * Added a SAX parser (4-5x the speed of the DOM parser).  The DOM
   parser is still the default.  Please try the SAX parser and report
   any issues.

 * Added "binary" flag to dump()/dumps() to create compressed streams.

 * Self-referencing objects correctly handled now.

 * Python builtins can be pickled as toplevel objects.  All builtins
   should be handled now.

 * Additional software modules have been added to the package, under
   gnosis.util and gnosis.util.convert (mostly under-documented, for
   now).

 * Documentation organization improved.  Each level of subpackages
   should contain a doc/ subdirectory, containing text files that can be
   viewed in any text viewer easily.  The content of these files is
   imported into *.doc subpackage names, allowing inspection such as:

      >>> from gnosis.xml.pickle.doc import AUTHOR
      >>> AUTHOR
      ['David Mertz <mertz at gnosis.cx>', 'Frank McIngvale <frankm at hiwaay.net>']

    Or also:

      % pydoc.py gnosis.xml.pickle.doc.COPYRIGHT

Summary:
<P><A HREF="http://gnosis.cx/download/Gnosis_XML_Util.ANNOUNCE">
Gnosis_Util 0.95</A>
xml_pickle, xml_objectify and indexer/xml_indexer; also xml2sql,
sql2dtd, txt2html, etc.  (Wed 04-03-2002)


The prior announcement is still generally informative:
------------------------------------------------------------------------

Subject: [Announce] Major update to xml_pickle package

Hi, I just wanted to send a note to announce a rather large
update to the xml_pickle package. I started with xml_pickle-0.51
and (together with David Mertz) added a bunch of things to make
xml_pickle a functional superset of the standard pickle module.

[Background blurb: xml_pickle is a package that looks, internally
and externally, like the standard pickle module, but produces XML
output using a minimalistic encoding.]

New features, compared to xml_pickle-0.51:

  o Can now fully restore class functionality, just like the
    standard pickle module.

  o Full internal compatibility with the pickle protocol
    (objects with __getstate__,__setstate__, and __getinitargs__
     are pickled/unpickled exactly as pickle does).

  o A pickle-compatible API layer
    (xml_pickle.dump()/dumps()/load()/loads() are drop-in compatible
     with the standard pickle module.)

  o A type-extension "plugin" system allows you to create "pretty
    printed" XML representations of objects if so desired.

  o Virtually all objects that can be pickled can be XML-pickled.
    (The only exception is that builtin types cannot be pickled as
     *toplevel* objects, since they are not true InstanceTypes.
     This restriction may be removed in a future version.)

  o Python 2.2 compatibility (Remains compatible with Python 2.0+)

  o Fine-grained security model added. I consider this an improvement
    over regular pickle, since it allows you to control how classes
    are created during unpickling. This means you can load pickles
    from "untrusted" sources and not worry about what you're getting
    (ie. no one can trick you into loading an EvilList where you
     expected a UserList). See "pydoc gnosis.xml.pickle.doc" for more
     information about the security model.

  o DEEPCOPY option added, so you can choose whether or not to have
    references inserted in the XML stream. References are great for
    saving space, but deepcopying allows third-party/non-Python parsers
    to get the entire data set without having to understand references.

  o References now keep the "pointed-to" type in the XML tag.
    This allows third-party/non-Python parsers to easily tell what the
    referenced object contains without reconstructing the object tree.

  o A few other bugfixes/improvements:
      - Corrected refchecking of temporary objects and key/val/items.
      - Numerical.arrays are now refchecked
      - SREs are now pickled

Given that there are alternative XML encoding schemes (such as XML-RPC
and WDDX), a few words to explain the particular purpose and goals of
xml_pickle seem to be in order:

   o It looks & feels just like the standard pickle module, both
     internally and externally, so all extension classes that
     speak the pickle protocol will "just work".

   o Has a minimalistic encoding scheme. All objects are represented
     using only five tags (<attr>,<entry>,<item>,<key>,and <val>) to
     simplify parsing by third-party/non-Python code.

   o The XML format is rich enough to enable xml_pickle to restore
     objects "fully alive", yet simple enough for "dumb" parsers
     to grab the complete data structures without needing to understand
     anything about Python classes, pickling, etc.

Despite the numerous changes, xml_pickle remains backward compatible
with the xml_pickle-0.51 API (and with XML documents created by older
versions of xml_pickle). The only source-level change you should have
to make would be to change the "import" line, since xml_pickle is now
part of a larger package.

For example:
    import xml_pickle
Could be rewritten:
    import gnosis.xml.pickle as xml_pickle

The complete package (Gnosis_XML_Utils), containing xml_pickle and
other useful components like xml_objectify and xml_indexer, is
released as public domain software, and is available from:

   http://gnosis.cx/download

Specifically (as of this writing):
   http://gnosis.cx/download/Gnosis_XML_Utils-0.9.tar.gz

Given the numerous changes, it'd be great to have as many people as
possible test it. The new package includes a good set of test cases
that also demonstrate the many new features of xml_pickle. Unpack the
sources, and look in gnosis/xml/pickle/test. Run "test_all.py" for a
complete set of tests.

Please send feedback, bug reports, etc., to either myself
(frankm at hiwaay.net) or David Mertz (mertz at gnosis.cx).

(Development was done on Linux and OS/2, so feedback about successes
or failures on other platforms would be greatly appreciated!)

Thanks!
frank

Summary:
<P><A HREF="http://gnosis.cx/download/Gnosis_XML_Util.ANNOUNCE">
Gnosis_XML_Util 0.9</A>
xml_pickle (pickle to XML), xml_objectify and xml_indexer (15-Feb-02)


--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From duncan at NOSPAMrcp.co.uk  Wed Apr 10 05:25:49 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 10 Apr 2002 09:25:49 +0000 (UTC)
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no>
Message-ID: <Xns91EC685318301duncanrcpcouk@127.0.0.1>

Syver Enstad <syver-en+usenet at online.no> wrote in 
news:upu18x4h3.fsf at online.no:

> Am I the only one in the entire world that thinks '\t'.join(mylist) is
> elegant and natural?
> 
Probably not, but I think that if the string is a literal I prefer:

    str.join('\t', mylist)

although I have also been known to write code like this:
    joinlines = '\n'.join
    ...
    joinlines(mylist)

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From sholden at holdenweb.com  Mon Apr 29 07:56:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 07:56:08 -0400
Subject: inline assignments in conditionals
References: <mailman.1019941549.3704.python-list@python.org>
Message-ID: <wAaz8.7312$M63.1038@atlpnn01.usenetserver.com>

"jaf" <jaf at lcsaudio.com> wrote in message
news:mailman.1019941549.3704.python-list at python.org...
> Aggh!  Of course I meant it the other way around.  :^P
> It would have been a nice example too, if I hadn't completely
> stuffed it up.  So for the record, "If today is Sunday" is my
> preferred form (i.e. if day == 0), and I should never again
> be allowed to post without proofreading first.
>

Don't see why I should be the only one to make a fool of myself (dammit).

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From sholden at holdenweb.com  Tue Apr 23 18:33:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 18:33:11 -0400
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org> <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com> <slrnacb5ko.4i6.akuchlin@ute.mems-exchange.org> <b0f083db.0204231422.785cb57c@posting.google.com>
Message-ID: <7llx8.2088$0x3.1467@atlpnn01.usenetserver.com>

"Jeremy Hylton" <jeremy at alum.mit.edu> wrote ...
> "A.M. Kuchling" <akuchlin at ute.mems-exchange.org> wrote ...
> > In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>,
> > Steve Holden wrote:
> > > Since urllib knows nothing of cookies, you will need to integrate some
sort
> > > of a cookie jar into the library, with a new API  for the clients to
> > > retrieve and store the cookies.
> >
> > This is worthwhile, but I don't think it belongs in urllib.  It
> > belongs in a module or package of its own that provides general
> > Web-browser features such as cookies, remembering authentication
> > usernames and passwords, and a cache.  This package could then be used
> > for implementing HTML-scraping scripts, spiders, or a Web browser.
>
> urllib2 provides a more flexible framework for implementing
> URL-loading programs, like a spider.  I think it would be helpful to
> have the features you mention integrated into urllib2.
>
Given that urllib2 is much newer than urllib it does seem to make sense to
augment the more recently developed code.

> I'm not sure what the difference between an HTTP client, like urllib
> or urllib2, and a Web-browser is.  Other than urllib's monolithic
> design, why wouldn't you want these sorts of features in the module?
>
I imagine Andrew thought code breakage would be a problem, but I wouldn't
presume to channel him ;-).

Personally I imagined passing a dictionary as an optional cookie jar
argument, keyed by (domain, path) tuples. The library code would update this
as dictated by its interactions with web sources.

Of course there's a problem if multiple connections need to update the
cookie jar concurrently, so this might well turn out to be too simplistic
anyway.

regards
 Steve






From marklists at mceahern.com  Wed Apr  3 08:35:20 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 07:35:20 -0600
Subject: PEP 285: Adding a bool type
In-Reply-To: <J7qT5iA2Rvq8EwLD@jessikat.fsnet.co.uk>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEPCKIAA.marklists@mceahern.com>

[Robin Becker]
> C is a very successful language. I don't think of its conditional
> testing as a hack, but then I'm an engineer. What works is good and
> Python's current conditional testing seems to work very well and
> moreover efficiently in many cases of interest.
>
> Forcing all these into the (true, false) set using a Boole (if we're
> going to honour the man let's spell him right) variable seems to put too
> much emphasis on formalism. It also ignores the eminent sense of not
> fixing unbroken things and not adding yet another current to the Python
> type cake.

How is the PEP "forcing" anything?

As far as I can tell, the PEP simply means that I can do something like
this:

	def isPaid(account):
		...
		return True

without having to define my own silly truth/bool/boole (whatever the hell
you want to call it) constants.

Now, whether I should instead design this function like so:

	def paidThroughDate(account):
		...
		# someDate might actually be None
		return someDate

is a different matter.  And even if I did the latter, nothing about the PEP
means I can't still do this:

	if not paidThroughDate(account):
		...
	else:
		...

So where's the problem?

I'm still totally +1 on this PEP.  Of course, maybe all the subtle
objections raised are simply way over my head.

Cheers,

// mark




From garrett at bgb-consulting.com  Mon Apr 15 15:26:50 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Mon, 15 Apr 2002 19:26:50 GMT
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu>
Message-ID: <3CBB29B9.95DED100@bgb-consulting.com>

  I have a couple of Pentium 450s that can help. RedHat 7.x with Python 2.2
but no additional modules yet. I can install them, but probably take me a day
or two to get to it. So... 4-6 problems, I guess?

"D. Michael McFarland" wrote:
> 
> Dear Fellow Pythonistas,
> 
>     A friend of mine is in the throes of finishing his
> Ph.D. dissertation, comparing experimental and numerical results for
> material damage caused by grinding.  For the numerical part, he's
> using a finite element library I wrote in Python.  This library made
> programming the tensor operations much easier than they might have
> been in, say, Fortran, but the end result takes rather longer to run
> than one could hope--about a day per problem case on a PII-450.  He
> has two machines of this class available, and I have two, and that
> might just be enough to let him make his deadline, but it occurred to
> me where to find more machines: this newsgroup.
> 
>     In short, we want your flops.
> 
>     I'm writing in the hope that a few people with CPU time to spare
> over the next couple of weeks will agree to run a case or two each.
> Volunteers would need to be running a fairly recent Python and the
> Numerical module, and install Neil Schemenauer's sparsemodule from
> 
> http://python.ca/nas/python/sparsemodule/
> 
> The output files are large, so some bandwidth is needed.  Unix or
> something like it is the only OS we're prepared to run on, unless
> someone wants to try to make our code run under MS Windows (I wouldn't
> know where to begin, myself, but it might not be too difficult).
> 
>     I'm open to suggestions, but the simplest strategy seems to be to
> mail source code, problem input files and a Makefile to control the
> whole mess to each participant, with results to be FTPed back.
> 
>     Incidentally, we have always meant to release this code once it
> was polished, and we still plan to do so.  But first this guy needs to
> graduate, for which he needs to crunch a _lot_ of numbers.  He's
> swamped with related work, and I feel like I got him into this, so I'm
> appealing to the Python community on his behalf.  Any offers of
> assistance or other advice will be gratefully received.
> 
>                               Best regards,
>                               Michael
> 
> --
>     D. Michael McFarland                      dmmcf at uiuc.edu
>     Visiting Senior Research Scientist & Visiting Lecturer
>     Department of Aeronautical and Astronautical Engineering
>     University of Illinois at Urbana-Champaign

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From aahz at pythoncraft.com  Wed Apr  3 16:53:57 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 16:53:57 -0500
Subject: Py3K (was Re: PEP 285: Adding a bool type)
References: <mailman.1017466974.6751.python-list@python.org> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <ualr7hmlhkemab@news.supernews.com>
Message-ID: <a8ftll$a4$1@panix1.panix.com>

In article <ualr7hmlhkemab at news.supernews.com>,
John Roth <johnroth at ameritech.net> wrote:
>
>In fact, at this point I'd like everything substantive to be defered
>to Python 3000, and get started on that right away. I've got a
>few suggestions...

I don't remember when Guido said it, and I don't feel like searching for
it, but there will not be a Python 3000.  It was half a joke and half a
gedanken when originally proposed.  Python 3.0 will introduce some
incompatibilities and do some cleanup, but it will *NOT* be a complete
rewrite.  The only certain thing about Python 3.0 at this point (that's
in my memory banks, at least) is that classic classes will be dead and
all classes will inherit from object.  Complete removal of old-style
division will probably be in Python 3.0, too.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From BgPorter at NOartlogicSPAM.com  Wed Apr 10 17:33:25 2002
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Wed, 10 Apr 2002 21:33:25 GMT
Subject: Indenting lines 400..457 help
References: <a929tk$ik2$1@pcls4.std.com>
Message-ID: <Fc2t8.90644$w7.7532194@bin6.nnrp.aus1.giganews.com>

"Ned Harvey" <rahvee at yaahhoo.com> wrote in message
news:a929tk$ik2$1 at pcls4.std.com...
> For the life of me, I can't find any easy way to indent (or unindent)
lines
> numbered 400..457 or whatever.  Does anyone have any good suggestions?
>
> I'm usually a vi user, but I'll happily learn a command of some other
> format.  I've tried
>     :400,457s/(.*)/   \1/g
> and a ton of variations of the same idea, but it seems that the
> search/replace doesn't do what I expect it to.
>
> Thanks for your help.
In vi, place the cursor on the first line that you want {in|de}dented. Then
type the number of lines to affect and >> to indent, or << to dedent.


--
--
//  Today's Oblique Strategy (? Brian Eno/Peter Schmidt):
//  Don't be afraid of things because they're easy to do
//
//  Brett g Porter * BgPorter at acm.org
//  http://mywebpages.comcast.net/bgporter/





From mgilfix at eecs.tufts.edu  Mon Apr 22 11:13:55 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 22 Apr 2002 11:13:55 -0400
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
In-Reply-To: <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Mon, Apr 22, 2002 at 09:51:38AM -0400
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com>
Message-ID: <20020422111355.A2066@eecs.tufts.edu>

On Mon, Apr 22 @ 09:51, Steve Holden wrote:
> I agree with this. But don't try to suggest that there aren't problems with
> the ternary construct in C, most notably the omission of an "=" in the test,
> turning a test into an assignment.

  That's the whole point of the ternary operator isn't it? To make
binary decisions before an assignment. A useful example is a funtion
with mutli-directions:

  void both_ways (int val, int direction)
  {

    another_func ( (direction) ? NULL : &val,
                   (direction) ? &val : NULL,
                    ....);
  }

  That way you can easily switch arguments to another function without
more typing. But most people use the ternary operator as an if
statement, which is ugly.

  While I'm not commenting on whether this is really useful in python
(because I get along fine without it), the python equiv would be:

  def both_ways (arg, dir):

    if dir:
      arg1 = blah
      arg2 = some_blah
    else:
      arg1 = something_else
      arg2 = some_other_thing

    another_func (arg1, arg2)

  Yeah, I'm *sure* you can get fancier. I think I saw a PEP where
someone suggesting having if return a value. There have been
times when I think my syntax would have looked nicer/simpler/more
elegant/pick one if that were the case. I know it can be done
another way. But a bit of syntactic sugar isn't always a bad
thing.

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From aleax at aleax.it  Tue Apr 30 17:08:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 21:08:59 GMT
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <jCzz8.17994$xy4.10082@atlpnn01.usenetserver.com> <sfAz8.82271$vF6.2450688@news2.tin.it> <6OAz8.1251$Au5.843@atlpnn01.usenetserver.com>
Message-ID: <LJDz8.35771$8D3.1034180@news1.tin.it>

Steve Holden wrote:
        ...
> sloppiness. And, of course, to confound the confusion int() is *not* a
> call to the integer class/type creator, but a no-argument call of the
> built-in int() function. What happens to "int(x[, radix]) " when int is a
> built-in type, by the way, or isn't that going to happen?

Steve, what are you _taking_ about...?

>>> type(int)
<type 'type'>
>>> int('23',6)
15

int *is* a built-in type, a call to it is a call to the int type, and one of
the ways you can call the int type to create an int istance is with a
string argument and optionally a radix integer argument.

Are you stuck in the far, dark ages of Python 2.1 ... ?-)


> so-they-don't-call-it-a-callable-iterator-because-
>         -it's-an-iterator-that's-callable-they-call-it-a-
>         -callable-iterator-because-it-iterates-callables-
>         -ly y'rs  - steve

Bingo.


Alex



From TBelich at skcc.org  Mon Apr 15 11:47:33 2002
From: TBelich at skcc.org (Tom Belich)
Date: Mon, 15 Apr 2002 08:47:33 -0700
Subject: Neural Networks
Message-ID: <EC29AFD5E8A1D311A86C0008C79F87C5ABD384@sidney.skcc.org>

Check out Con-X
http://dangermouse.brynmawr.edu/cx/

If you download the Pyro package:
http://emergent.brynmawr.edu/~dblank/pyro/html/

Con-X is included.

If you like Java as well, check out Joone:
http://joone.sourceforge.net/index.html

Lots of luck,
Tom
-----Original Message-----
From: tony clarke [mailto:a.clarke11 at ntlworld.com]
Sent: Sunday, April 14, 2002 1:17 PM
To: python-list at python.org
Subject: Neural Networks


Hi

Does anyone know any sites, or have any code or other info about
artificial intelligence. I am specifically interested in simple neural
network programming, being a simple soul myself!

Thanks in advance

Tony

-- 
http://mail.python.org/mailman/listinfo/python-list




From marklists at mceahern.com  Wed Apr  3 22:46:37 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 21:46:37 -0600
Subject: turning a string into an object name
In-Reply-To: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOECAKJAA.marklists@mceahern.com>

[Marco Herrn]
> [...]what I want to achieve is turning a string into an object name.

Short answer:

Use a dictionary to store a mapping between strings and object names, like
this:

  class foo:
    def __init__(self):
      print "Hi, I'm an instance of foo."

  dict = {'foo': foo}
  className = 'foo'
  # Avoid using class, since that's reserved.
  cls = dict.get(className, None)
  if cls:
    x = cls()

Longer answer:

Checkout the inspect module.

Cheers,

// mark





From christophertavares at earthlink.net  Mon Apr 29 02:47:37 2002
From: christophertavares at earthlink.net (Chris Tavares)
Date: Mon, 29 Apr 2002 06:47:37 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <d06z8.12562$iU4.1091750@newsread2.prod.itd.earthlink.net>

"Raymond Hettinger" <python at rcn.com> wrote in message
news:aaim48$7fd$1 at bob.news.rcn.net...
[ ... snip ... ]
>
>     I propose that a simple means be provided to re-enable those functions
> at
>     will when they are needed to run old code.  Add a command line
option -b
>     and an equivalent environment variable, PYTHONBUILTINRESTORE which
would
>     start-up Python by running:
>
>         from functionals import map, filter, reduce
>         from math import pow, divmod
>         from deprecated import input
>

No, no, no, no, no, no, no, no, no. This would place the burden of calling
this code correctly on the end user. How's the average programmer going to
know when to use -b? What if somebody set PYTHONBUILTINRESTORE once, then
trashes his .bashrc file, and now none of his previously working Python code
works? These kinds of command line switches are evil.

Also, what about environments without command lines? Python runs on a lot
more than just Unix.

The "directive" needs to be in the code. And you've already provided the
form of this directive: the import statements above.

I would think the reasonable migration would be to add the modules listed
above (or add the functions as mentioned), but leave the builtins for a
version or two. Then remove the builtins.

-Chris





From programisci at NOSPAM.murator.com.pl  Mon Apr 15 04:48:58 2002
From: programisci at NOSPAM.murator.com.pl (Johann)
Date: Mon, 15 Apr 2002 10:48:58 +0200
Subject: I'm lookin' 4 cute forum scripts
Message-ID: <mt4lbucdcmfpf9ln5ma5n4kfdm1vr4qf0a@4ax.com>

I am looking for cute (functional and visual) discussion forum for
www. I would like to display the tree with threads, sort by author,
date and search.

I am interested in Perl, PHP or Python scripts only.  I do not know
what to choose. I found e.g.
http://www.w-agora.net/current/index.php?site=demos but it cannot sort
by date, author etc. (I didn't check it carefully, but all examples
cannot)

Mabe someone coud recommend me something?

-- 
Johann


From robin at jessikat.fsnet.co.uk  Sat Apr  6 18:04:41 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 7 Apr 2002 00:04:41 +0100
Subject: 2.2 schizo win32 linkage
References: <naBIBIAc1ur8EwVD@jessikat.fsnet.co.uk>
 <slrnaauuj6.tl.gerhard@lilith.my-fqdn.de>
Message-ID: <uTUaTJAJ83r8EwkA@jessikat.fsnet.co.uk>

In article <slrnaauuj6.tl.gerhard at lilith.my-fqdn.de>, Gerhard H?ring
<gerhard at bigfoot.de> writes
...
>
>The best solution I can think of is for dependencies of an extension
>module to place them in the same directory as the extension module:
>this is where Windows searches first when trying to satisfy the DLL
>dependency.
so where should Python's tcl/tk dlls live. Other extensions also depend
on them.
-- 
Robin Becker


From warren-postma at NOSPAM.rogers.com  Tue Apr 23 12:18:23 2002
From: warren-postma at NOSPAM.rogers.com (Warren Postma)
Date: Tue, 23 Apr 2002 16:18:23 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020422124126.GD23706@andy.tynant.ftech.net> <Pine.LNX.4.33.0204220729500.25157-100000@ns0.tbye.com> <mailman.1019485603.16764.python-list@python.org>
Message-ID: <jPfx8.47701$Gpn.14899@news02.bloor.is.net.cable.rogers.com>

> Discovery: Mathematics is intrinsically "there" ...
> Invention: We're making mathematics up as we go along ...

What about a third alternative?  Mathematics is Occult.  It only makes sense
to people who are posessed by [sic] "daemons".   We should have burned all
those Mathematicians at the stake and thus saved highschool students from
the horrors of Algebra and Calculus, forever and ever.   Amen.

<grin>

Warren





From gjlap at yahoo.com  Mon Apr  1 00:31:41 2002
From: gjlap at yahoo.com (Gil)
Date: 31 Mar 2002 21:31:41 -0800
Subject: duelling pythons
Message-ID: <1eed326b.0203312131.2ef13b62@posting.google.com>

I think my original install of Python1.5 with RedHat 7.1 is not
playing nice with a later install of Python2.2. How do I troubleshoot
this?

TIA


From aahz at pythoncraft.com  Sun Apr 21 13:36:11 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 13:36:11 -0400
Subject: c++ vs python!
References: <3CC2D8D7.20601@student.kun.nl>
Message-ID: <a9utab$e18$1@panix3.panix.com>

In article <3CC2D8D7.20601 at student.kun.nl>,
husam  <h.jehadalwan at student.kun.nl> wrote:
>
>If we compare c++ with python, which one is faster when it comes to 2D, 
>3D graphics and math. And which one is faster when it comes to design 
>and implementation of "heavy" applications?

Use Python to call C/C++ code for compute-intensive tasks.  See the
Python Game Library and NumPy for examples.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From vomjom at vomjom.org  Sat Apr 20 04:02:45 2002
From: vomjom at vomjom.org (Jonathan Hseu)
Date: Sat, 20 Apr 2002 03:02:45 -0500
Subject: Fundamental problem with urllib...
Message-ID: <20020420080245.GA2245@vomjom>

Urllib will always have problems with some sites.  Notably, sites with
cookies that also redirect.

Here's what happens:
Say, perhaps after POSTing, you want to grab a cookie from the headers.
The website puts the cookie in a 302 message (redirection for those of
you who don't know HTTP responses), and then redirects you elsewhere.
Now, to grab the cookie, you need to be able to get the headers of that
message before getting redirected.

This is certainly not possible without overriding some functions in
FancyURLopener or declaring some in URLopener (which creates exceptions
for all HTTP responses != 200).  This isn't a big deal (I did it myself
for my program), but this can be simply fixed by adding support for
cookies within urllib.

Shall I take it upon myself to make such a patch?  Comments/Flames?

I personally think easy cookie handling within urllib is a good thing.
-- 
Jonathan Hseu <vomjom at vomjom.org, vomjom at debian.org, jh4 at cec.wustl.edu>
GPG ID:          5228D713
GPG fingerprint: 220B A4EF 70FE B884 CB38  F93F EA8A 1024 5228 D713




From tanzer at swing.co.at  Thu Apr 18 03:15:20 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 18 Apr 2002 09:15:20 +0200
Subject: pre-PEP for optional 'pass'
In-Reply-To: Your message of "Wed, 17 Apr 2002 10:21:15 BST."
             <20020417092115.GD2063@andy.tynant.ftech.net>
Message-ID: <E16y68W-0003HE-00@tswings.swing.cluster>

Andy Gimblett <gimbo at ftech.net> wrote:

> How about this?
>
> class MyClass:
>     """Class encapsulating a Twiddle.  Unfinished.  Document me.  XXX"""
>     def __init__(self, x, ,y, z):
>         """Not yet implemented. XXX"""
>     def getTwiddle(self):
>         """Not yet implemented. XXX"""
>     def setTwiddle(self, twiddle):
>         """Not yet implemented. XXX"""
>
> where you can grep for XXX to find unfinished business - very handy.

If you want to go that road, I'd prefer

        raise NotImplementedError

over the doc-string. Add XXX according to taste.

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





From siegfried.gonzi at kfunigraz.ac.at  Fri Apr 26 04:07:30 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Fri, 26 Apr 2002 10:07:30 +0200
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at> <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com> <3CC80EBE.E9735284@kfunigraz.ac.at> <aa94rq$4qb$0@216.39.172.122>
Message-ID: <3CC90AC2.BD87A807@kfunigraz.ac.at>

Bengt Richter wrote:

> What happens if you run it from a plain "DOS" console window (I am assuming
> from your post header that you are running on NT?), typing
> 
>     python your_script.py arg1 arg2 ...
> 
> at the prompt?

Oh Christ, thank you very much! This actually works without any
problems. It really frees the memory (that is observable in the Task
Manager). You could also work on other projects when the simulation is
running in the DOS-window as opposed to the idle: which would occupy the
complete system without any pardon.

This means: for calculation which heavy memory consumption you should
run it from the DOS-window and not from the idle itself.


I am now happy again with Python.

Thank you,
S. Gonzi


From phd at phd.pp.ru  Wed Apr 24 03:58:15 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 24 Apr 2002 11:58:15 +0400
Subject: unsuscribe
In-Reply-To: <20020423222038.20dc4118.ngreco@softhome.net>; from ngreco@softhome.net on Tue, Apr 23, 2002 at 10:20:38PM -0300
References: <20020423222038.20dc4118.ngreco@softhome.net>
Message-ID: <20020424115815.B22328@phd.pp.ru>

On Tue, Apr 23, 2002 at 10:20:38PM -0300, Nahuel Greco wrote:
> unsuscribe

   Not "unsubscribe" but "unsubscribe, please". Joke. Wrong address,
actually. Please don't send control messages to the list, send them to the
appropriate control address.

> - Nahuel Greco                   Web Development - Open Source - 
> - http://nahuel.codelarvs.com.ar Game Programming - Research   -
> - Freelance coding / sysadmin    Networking. The answer is 42. -
> -- 
> http://mail.python.org/mailman/listinfo/python-list

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




From gerhard at bigfoot.de  Fri Apr 12 19:35:32 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 13 Apr 2002 01:35:32 +0200
Subject: porting python to embeded devices
In-Reply-To: <slrnabeqci.s6b.grey@teleute.dmiyu.org>
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org>
Message-ID: <20020412233532.GB15144@lilith.my-fqdn.de>

* Steve Lamb <grey at despair.dmiyu.org> [2002-04-12 23:11 -0000]:
> On 12 Apr 2002 16:04:50 -0700, Paul Rubin <phr-n2002a at nightsong.com> wrote:
> > I'd say it's hopeless or at least impractical.  Those processors are
> > much too constrained to run Python.  It would be like trying to run an
> > SQL database on them.  Python has many virtues but being able to run
> > on very tiny processors isn't one of them.
> 
>     What about something like the Zaurus SL-5500?
 
It runs fine on it, and I'm currently using Phil Thompson's Python 2.2
and PyQt packages that he built for the Zaurus on my iPAQ running opie.

These devices both have a StrongARM processor running at 206 MHz (only
software floating point, though) and between 16 and 64 MB of RAM. So
they're quite well equipped for a handheld ;-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From mgilfix at eecs.tufts.edu  Wed Apr 17 10:21:46 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 10:21:46 -0400
Subject: Pure python crypto
In-Reply-To: <7x3cxubvgb.fsf@ruckus.brouhaha.com>; from phr-n2002a@nightsong.com on Wed, Apr 17, 2002 at 01:06:28AM -0700
References: <mailman.1019024313.11744.python-list@python.org> <7x3cxubvgb.fsf@ruckus.brouhaha.com>
Message-ID: <20020417102146.J16954@eecs.tufts.edu>

  The nice thing about blowfish is its simplicity. I started thinking
about implementing DES and 3DES though and its ugly. The perl
module version seemed to go at great lengths to do some interesting
optimizations which are difficult to follow but increase speed
considerably.

  I was thinking that it might be nice to have an 'ssh' module or
'sftp' module. Specifically, I wanted sftp for an application and
couldn't find any such modules. For portability reasons, it might
be nice to have a pure python module, however, it would require
some tweaking I'm sure.

                 -- Mike

On Wed, Apr 17 @ 01:06, Paul Rubin wrote:
> Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
> >   Would anyone have any use for a pure python implementation of
> > Blowfish? Or any other crypto algorithms for that matter.  As a matter
> > of interest, I found myself implementing the blowfish module recently
> > but was unsure if any audience would be interested. No C compiler
> > is required, which is nice. Could use some testing though.
> 
> I wouldn't mind having it, but it wouldn't be THAT useful, as it
> would be awfully slow, and I don't know of any important programs
> to interoperate with that need Blowfish.
> 
> DES/3DES seems more important for interoperability purposes.  I coded
> them in pure Python but made a mistake somewhere, so the test vectors
> don't work correctly, and I never got around to debugging it.  I should
> take a look at it again.
> 
> If you just want to encrypt in Python without custom C modules, after
> a fair amount of experimentation the fastest way I found is to use the
> SHA module in output feedback mode to generate a keystream from the
> key and a random initialization string, and use the array module to
> combine the keystream with the plaintext.  This works pretty well on
> Un*x/Linux because the /dev/urandom device makes it easy to get the
> initialization strings.  However, there's no good source of such
> strings under Windows using the standard modules at the moment.
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (phr-n2002a)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/191b7321/attachment.sig>

From roy at panix.com  Mon Apr 29 08:35:24 2002
From: roy at panix.com (Roy Smith)
Date: Mon, 29 Apr 2002 08:35:24 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <roy-4E6B7A.08352429042002@news1.panix.com>

"Raymond Hettinger" <python at rcn.com> wrote:
> The six functions would be removed in Python 2.5 (having allowed
> two versions and a full year for code updates).

I don't use map(), reduce(), filter(), or for that matter lambdas.  Some 
part of me agrees with you that they don't belong in the core langugae.

But, given that they're there, and some people use them and depend on them 
for their code to work, removing them just for the sake of cleaning up the 
language is a poor idea.

As for giving people "a full year", before their code is broken, that's 
absurd.  Lots of folks run software a lot older than a year.  To break 
stuff that's a year old is just plain stupid.


From bmason at rand.org  Tue Apr 30 18:59:10 2002
From: bmason at rand.org (Bryce C. Mason)
Date: Tue, 30 Apr 2002 15:59:10 -0700
Subject: Macintosh Question
Message-ID: <aan7ih$jve$1@lumberjack.rand.org>

I developed a script in windows which opens a file, does "stuff" to it, and
writes a few files back to the disk.

How do I do the "drag and drop" Macintosh version?  I want the user to have
my python script on his desktop and all he has to do is drag the file over
the script, let go, and out pop the files the script generates right on his
desktop.

Any ideas?

Bryce





From ianb at colorstudy.com  Tue Apr  9 02:28:27 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 09 Apr 2002 01:28:27 -0500
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <1018331838.472.796.camel@localhost>
References: <mailman.1018314090.9984.python-list@python.org> 
	<_1vs8.1787$CA6.114998@newsread1.prod.itd.earthlink.net> 
	<1018331838.472.796.camel@localhost>
Message-ID: <1018333707.539.899.camel@localhost>

On Tue, 2002-04-09 at 00:57, Ian Bicking wrote:
> One of the big advantages they had of this system was that they could
> run the VM inside Squeak and use Squeak's excellent debugging tools
> (well, I think they were actually running it in another version of
> Smalltalk when they were bootstrapping).  The language was a subset of
> Smalltalk, which made this possible.  Pyrex isn't a subset, since it
> introduces new keywords and static types -- the static types could
> potentially be ignored, with the expectation that you might encounter
> new type-related bugs when you moved to C.  This would have been an
> issue for Squeak as well -- it's certainly no closer to C than Python --
> but I don't know how they solved this problem.

I looked at this a bit, and I think I realize how they did it: Slang
(Squeak's C-like subset) is really, really boring.  It can't interact
with Squeak -- it's used to implement the VM and potentially plugins to
the VM, but the layers remain very separate.  It doesn't seem to have
type definitions, so I don't know what's going on with that... maybe
it's making everything an int... they speak of it as a portable
assembler, and maybe they mean that more than you'd think.

This is all from spending 15 minutes looking at the VM code in Squeak,
so I could be entirely wrong.  There doesn't seem to be any
documentation for Slang.

  Ian






From aleax at aleax.it  Tue Apr  9 17:09:21 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 21:09:21 GMT
Subject: class callable with any method name
References: <9d095c5a.0204091256.4d88a246@posting.google.com>
Message-ID: <5MIs8.1212$vF6.49294@news2.tin.it>

Imer wrote:

> Hi, I'm trying to find a way to implement a class that would allow any
> method name to be called.
> When that happens, a special method would be called and the name of
> the method that was originally called would be passed as an argument.
        ...
> Is there a way to do that ?

Sure: __getattr__ is called for methods as well as attributes.  You
just need to return a callable that wraps/curries whatever you want.


Alex



From eppstein at ics.uci.edu  Mon Apr  1 18:58:32 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 01 Apr 2002 15:58:32 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com> <a8an7b$eme$0@216.39.172.122>
Message-ID: <eppstein-250858.15583201042002@news.service.uci.edu>

In article <a8an7b$eme$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
wrote:

> Well, for types that are related in certain ways ... ;-)
> 
>  >>> 0 == 0.0
>  1
>  >>> 1 == 1.0
>  1
>  >>> 1 == 2.0
>  0
> 
> That gets into what relational operators really should mean.
> I think I exclude bool from the number tree, so I don't
> like the effect of subtyping bool from int, even though
> I support context-dependent implicit conversions to integer
> (though that context should be defined by the client expression ops).

If you called it "binary" instead of "bool" would it help you think of 
it as being more like a number?

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From magnus.heino at pleon.sigma.se  Thu Apr 11 08:27:11 2002
From: magnus.heino at pleon.sigma.se (Magnus Heino)
Date: Thu, 11 Apr 2002 14:27:11 +0200
Subject: distutils
References: <GTet8.4573$Dw4.8174@news1.bredband.com> <j4sn62xx1y.fsf@informatik.hu-berlin.de>
Message-ID: <Jift8.8$WJ4.85@news1.bredband.com>

Martin v. L?wis wrote:

> Magnus Heino <magnus.heino at pleon.sigma.se> writes:
> 
>> Im using distutils to create an rpm for me,
>> 
>> python2.2 setup.py bdist_rpm
>> 
>> Now, this installs the module in /usr/lib/python2.2/site-packages/
>> 
>> How can I install it somewhere else instead?
> 
> See
> 
> http://www.python.org/doc/current/inst/alt-install-windows.html

Well, thats valid for 

python2.2 setup.py install

while im doing 

python2.2 setup.py bdist_rpm

There isnt anything you can do in setup.py?

/Magnus


From aleax at aleax.it  Wed Apr  3 05:16:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 10:16:08 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com>
Message-ID: <IDAq8.45703$pT1.1295831@news1.tin.it>

Paul Rubin wrote:

> Jeff Shannon <jeff at ccvcorp.com> writes:
>> > I am opposed to the addition of the new proposed type on the grounds
>> > that it will make Python harder to teach both to people who have never
>> > programmed before and to people who _have_.
>> 
>> At first glance, I was marginally in favor of this PEP.  After
>> reading Alex Martelli's arguments, I was marginally opposed to
>> it.  I think that Laura's comments are *very* convincing, and I'm
>> now fairly solidly opposed to it.  Well said, Laura.
> 
> I agree, I missed Laura's post earlier but it's terrific.

Full agreement.  But not from Guido, who dubs it an "over-long FUD
piece" -- all others who commented on Laura's post, whether they
agreed with it or not, seemed to appreciate its worth.  I guess that
tells us something about how worthwhile it is to spend hours studying
and reflecting in depth upon issues, and composing beautiful posts 
on the subject, when arguments are requested: the message is quite
clear, "if you don't agree with what he's already decided to do,
don't bother, you'll only peeve him off".  Nice PR exercise...;-).

Oh well.


Alex



From sholden at holdenweb.com  Mon Apr 15 15:04:32 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 15:04:32 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <a8tov6$484$1@panix1.panix.com>
Message-ID: <9xFu8.15792$Ji7.11388@atlpnn01.usenetserver.com>

"Aahz" <aahz at pythoncraft.com> wrote in message
news:a8tov6$484$1 at panix1.panix.com...
> In article <4a7fde99.0204081858.58baabdd at posting.google.com>,
> Ryan <ryanmorillo at hotmail.com> wrote:
> >
> >    I have been doodling with python (nasty habit I know but I have to
> >do somthing with my snake) and think I may be ready to learn a second
> >language.  I am leaning towards C over C++ and Java even though I
> >loose objects it will ground me firmly in functional programing.  I am
> >wondering what book would be the best for a begginer?  The two I have
> >seen that look interesting are
> >C for Dummies (parts one and two) or
> >The C Programming Language
> >by Brian W. Kernighan, Dennis M. Ritchie.
>
> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
> free to buy another book if K&R turns out to be too heavy for you right
> now, but recognize that you'll need it eventually.

Harbison and Steele is IMHO more beginner-friendly, and will help you until
you graduate to Kernigan and Ritchie.

regards
 Steve






From gerhard at bigfoot.de  Mon Apr 15 10:31:15 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 15 Apr 2002 14:31:15 GMT
Subject: Python installation on a Windows shared disk
References: <a9e5u2$1j0u$1@norfair.nerim.net>
Message-ID: <slrnablp1j.fa.gerhard@lilith.my-fqdn.de>

Eric Brunel wrote in comp.lang.python:
> Hi all,
> 
> I'm trying to do an installation of Python 2.1 on a shared Windows 2K disk.
> People should then be able to use this installation from all PCs seeing this
> disk. [...]

This should work:

- Install Python in a directory on your share
- Copy the Python$VERSION.DLL from $SYSTEMROOT/System32/ into the
  directory you installed Python in (in the same one the python.exe
  went)
- Put this share in the PATH on your client PCs

Nothing more should be needed.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From sholden at holdenweb.com  Mon Apr 22 11:25:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 11:25:35 -0400
Subject: HTTP error on receiving a form
References: <mailman.1019466221.11245.python-list@python.org>
Message-ID: <q_Vw8.76070$T%5.17278@atlpnn01.usenetserver.com>

<Patrick.Bussi at space.alcatel.fr> wrote in message
news:mailman.1019466221.11245.python-list at python.org...
>
> Hello
>
> I am implementing a simple tool to answer post requests for automatic
testing of
> a Web APACHE Server (implemented by other people). It is rather easy with
use of
> something like:
>
> dataOfTheForm= {"data1":"val1","data2":"val2"[...etc...]}
> encodedData=urllib.urlencode(dataOfTheForm)
> f=urllib.urlopen(webSite,encodedData)
>
> However, the web server find errors in the data received from my tool,
while all
> is OK with a web browser. I have TCPdumped answers both from my test tool
and
> from an ordinary web browser (say Mozilla). As far as I could understand,
the
> only difference is the order in which data are encoded.
>
> My questions:
>
> 1 - is it possible that a web Server be sensitive to the order in which
data are
> encoded (as they are all sent in one single IP packet)
>
This should not be a factor. It might affect things if the same name
appeared twice in the encoded data, but encoding from a dictionary more or
less explicitly ensure this won't happen.

> 2 - is there a quick way to encode data from a dictionary in a specific
order.
>
>
Not using urllib.urlencode().

I think you should be looking elsehwere for the problem, it seems rather
unlikely that sequencing is causing the error.

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From brian at sweetapp.com  Fri Apr  5 13:03:07 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 5 Apr 2002 10:03:07 -0800
Subject: Filename Too Long
In-Reply-To: <be230158.0204050948.7fa6a62f@posting.google.com>
Message-ID: <002501c1dccc$241a5870$445d4540@Dell2>

Peter wrote:
> I'm using windows and trying to call os.path.getmtime() after using
> os.path.walk... However, I'm choking with "[Errno 38] Filename too
> long" on paths with len > ~260
> 
> Is there any way around this error?

I assume that you are using absolute paths? If that is the case then you
can work around this problem by changing directories (os.chdir) and
using relative paths e.g.

file = r'c:\200chardirectoryname\200charfilename'
os.chdir(r'c:\200chardirectoryname')
os.path.getmttime('200charfilename')

Cheers,
Brian





From jeff at ccvcorp.com  Mon Apr 22 14:28:10 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 11:28:10 -0700
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org> <slrnac2v9d.kv6.jon+usenet@snowy.squish.net> <mailman.1019315200.29890.python-list@python.org> <slrnac326d.gba.jon+usenet@snowy.squish.net>
Message-ID: <MPG.172df6156ea0ac7c9896bd@news.nwlink.com>

In article <slrnac326d.gba.jon+usenet at snowy.squish.net>, 
jon+usenet at unequivocal.co.uk says...
> In article <mailman.1019315200.29890.python-list at python.org>,
>   holger krekel wrote:

> > But is it true that you 'never' use them somewhere else? 
> > I would think that one would like to use 'r' whenever you don't 
> > want the escape-sequence-string-interpretation which is not really
> > restricted to regexp-strings. 
> 
> True, but personally I've never found any other use for them.

Then you probably use a unixoid OS.  ;)  Since Windows uses \ as 
a path separator, any Windows path strings tend to have lots of 
\'s in them, and are thus good candidates for using raw strings.

Holger is correct -- they are useful anywhere you don't want the 
escape-sequence interpretation done.  But \'s aren't terribly 
common except in regexes and Windows paths, so those are the main 
places that raw strings get used.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From opengeometry at yahoo.ca  Wed Apr 17 17:26:52 2002
From: opengeometry at yahoo.ca (William Park)
Date: Wed, 17 Apr 2002 17:26:52 -0400
Subject: What webhosting
In-Reply-To: <3CBDE7B4.3348.4EB76@localhost>; from printers@sendme.cz on Wed, Apr 17, 2002 at 09:23:00PM +0200
References: <3CBDE7B4.3348.4EB76@localhost>
Message-ID: <20020417172652.B10816@node0.opengeometry.ca>

On Wed, Apr 17, 2002 at 09:23:00PM +0200, A wrote:
> Hi,
> Does anybody know about a webhosting provider ( with Python and 
> Telnet access) that does not remove a user's account if the user 
> sends occasionally some unsolicited emails?
> Thanks for reply
> Ladislav

Get a static IP and do the hosting yourself.  Then, you don't have to
answer to anyone.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8 CPU cluster, NAS, (Slackware) Linux, Python, LaTeX, Vim, Mutt, Tin




From rxg218 at psu.edu  Wed Apr 10 18:42:34 2002
From: rxg218 at psu.edu (Rajarshi Guha)
Date: Wed, 10 Apr 2002 18:42:34 -0400
Subject: looping throuhg a list 2 items at a time
Message-ID: <a92f5j$250g@r02n01.cac.psu.edu>

Hi,
  I have a list like this:  l = [1,'s', 2,'t', 3,'d']

I'd like to loop through the list but within the loop I want 
the current element and the element just after it:
I tried this

liter = iter(l)
for i in l:
  print i, liter.next()

But I get:

1 1
s s
2 2
t t
3 3
d d

whereas I want

1 s
2 t
3 d

I can see I'm wrong - any ideas on how to fix it?

TIA,



-- 
Rajarshi Guha
rajarshi at presidency.com


From torn at coit13m.europe.nokia.com  Thu Apr 11 08:08:48 2002
From: torn at coit13m.europe.nokia.com (Peter Torngaard)
Date: Thu, 11 Apr 2002 12:08:48 GMT
Subject: SMS Server implemented in Python
Message-ID: <k1ft8.1694$ws6.37396@news2.nokia.com>

Hello,

just happend to read Frank Stajano's paper 
'Implementing the SMS server, or why I ....' and found that 
the server sounded cool. Now I'm looking for it but it seems 
like it do not want to be found! It have been on the net, 
alot of search engines are pointing to 404 pages!

My question is do anyone hold a copy? or know where it can be 
found.
-- 
Peter Torngaard (email: Peter.torngaard at nokia.com)
  "There are two kinds of fools, one that says 'This is old, and 
  therefore good'. And one that says 'This is new, and therefore 
  better'", John Brunner "The Shockwave Rider"


From joonas at olen.to  Sun Apr  7 10:54:03 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Sun, 07 Apr 2002 17:54:03 +0300
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com>
Message-ID: <3CB05D8B.F9FD9318@olen.to>

Jim Williams wrote:
> 
> Hi all...I'm new to Python and really enjoying it so far, but I am somewhat
> perplexed that as far as I have seen it has no facilities for printing to a
> printer?  Surely this is possible. Someone please enlighten me...I have done a
> good bit of research and searching and can find no reference to how to do this.

There is some information on printing on windows at
http://www.faqts.com/knowledge_base/view.phtml/aid/4463
Also a Google Groups search will bring up some interesting articles.


From borcis at geneva-link.ch  Mon Apr  8 04:57:48 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 08 Apr 2002 10:57:48 +0200
Subject: Icon  (Re: PEP 285: Adding a bool type)
References: <mailman.1017466974.6751.python-list@python.org> <3CA8E37A.E89CA002@noaa.gov> <7x4rivj5ya.fsf@ruckus.brouhaha.com> <3CA9E7CF.52D6CD37@noaa.gov> <QPoq8.44596$pT1.1256392@news1.tin.it>
Message-ID: <3CB15B8C.6070106@geneva-link.ch>

Alex Martelli wrote:

> Tut, tut -- in Icon you do write "if expression" (well, syntax's a bit
> different and my memory rusty, but, basically).  There's no concept
> of true or false or a boolean type: rather, any expression succeds
> or fails. Only if it succeeds does it return anything (and there's no
> such thing as 'returning a boolean', of course).


Guido wrote somewhere early on that Icon participated to Python's 
inspiration sources, and I've always tended to locate that contribution 
with the treatment of "logical" values and expressions. The name of None 
I believe is revealing. This would be BTW my own only aesthetical 
objection to the PEP, to have "None is not False" be True; or 
equivalently "None is False" be False. (Unless I have the PEP wrong, 
mind you).

> 
> There's more, of course, such as the ability to backtrack (trying to
> make the expression succeed by pumping sub-expressions for
> multiple results),

Together with generators and adequate special methods invocation logic, 
a  subclassable bool type looks like an early first step to evolve
python towards allowing similar idioms.


> However, it has had nothing like Python's success.

The reason for this is imho to be found in the thoroughness of Python's 
standard libs.

> I think it's worth studying if only to understand what DID
> Python get right, or luckier, than Icon.  (And anyway, Icon's
> paradigms are different enough from most other language's
> to make its study worthwhile as a mind-widening experience).

The backtracking paradigm is actually very close to Prolog's (and even 
more to Lisp+Screamer).


> If you want "booleans" given the respect you seem to think they
> deserve, you may actually be happier with Icon than with Python.

Given (what I saw of) Chris Barker's expressed position, I find this 
suggestion surprising. To me, Icon's treatment of "boolean values" is 
"idiosyncratic" in a manner that "exagerates" Python's way. Prolog and 
its logical variables otoh go full sweep to a mathematically more 
litterate understanding of logic.

Regards, Boris
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From sholden at holdenweb.com  Mon Apr  8 10:51:17 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 10:51:17 -0400
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <k9is8.832$eJ6.421@atlpnn01.usenetserver.com>

"Paul Rubin" <phr-n2002a at nightsong.com> wrote in message
news:7xadshtrj1.fsf at ruckus.brouhaha.com...
> Peter Hansen <peter at engcorp.com> writes:
> > No, that's "stagnant"...
> >
> > Stable is also a good adjective for it, since most of the older
> > code still runs unchanged, and Python from ten years ago is
> > still very recognizable to new coders today.
>
> Stable doesn't just mean old scripts run without change in new
> interpreters.  It also means new scripts usually run without change in
> old interpreters.  That's what keeps breaking in Python.
>
Please stop promoting your misunderstanding about backwards compatibility.
(OK, here you call it "stability", but I'm not aware of any major language
which hasn't introduced new features).

If all old interpreters must run all new scripts then the only possible
solution is indeed the time machine reputed to be in Guido's possession.

regards
 Steve






From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 17:39:32 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 23:39:32 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i4uv$k3h$1@slb4.atl.mindspring.net>
Message-ID: <a9i5ne$f9r$1@nets3.rz.RWTH-Aachen.DE>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:a9i4uv$k3h$1 at slb4.atl.mindspring.net...
> Philipp Lenssen:
> >Is it possible to have a safe eval() on the server-side even when the
> string
> >consists of submitted user data?
>
> What do you call 'safe'?  Ahh, your next post
>
> > all that should be allowed is arithmetics (boolean as
> > well) plus a limited set of functions that I write & define.
>
> Will you allow "blahblahblahblah" * 2 ** 31 ? (which is a possible denial
> of service attack if you don't have a few GB of RAM handy, since it makes
> an 8GB string.)
>

Oh yes, and I forgot that complicated part -- not only do I want to have my
limited set of allowed functions, I want to definitely allow string
comparison. This makes it a bit harder, since technically e.g. the following
should be allowed:
'john doe' == 'delete_all_files()' or 1

Now to tell the truth I don't care 100% if a once-in-a-million user input
like "delete_all_files()" is actually part of a harmless, well intended
string value, and I accidentally remove it (if I'd use a straightforward
replacement algorithm as last resort) because of my security... problems
would arise if the function name sounds more natural (like, I removed
"touch" in PHP).

But, if this is about any kind of buffer overflow or denial of service
attacks, I didn't even think about that. Well one problem at a time...

> If so, take a look at 'eval' in the documentation, under
>   http://www.python.org/doc/current/lib/built-in-funcs.html
> Call it with an empty locals and an almost empty globals
>
>   eval("2+3", {"__builtins__": {}}, {})
>..

I saw the __builtins__ being used by Geoff Gerrietts in the class he
provided, so that should be OK.

Thanks.






From gerhard at bigfoot.de  Sat Apr 27 02:47:14 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 27 Apr 2002 08:47:14 +0200
Subject: zope and jboss or weblogic
In-Reply-To: <e786d63d.0204262051.14ebe470@posting.google.com>
References: <e786d63d.0204262051.14ebe470@posting.google.com>
Message-ID: <20020427064714.GA2032@lilith.my-fqdn.de>

* Wenshan Du <python at dohao.org> [2002-04-26 21:51 -0700]: > hi,
> I like zope, I want to know what's the difference between zope and
> jboss or weblogic? which is better? Can python & zope replace J2EE?
> 
Not only do Zope and J2EE play in a different league, but it's also
different kinds of sports.

J2EE is really only a bunch of so-called enterprise-APIs thrown together
and made sure the versions of the selected APIs work well together.
These APIs include data persistence, distributed computing, others I
don't remember, and last but not least, a web application framework
(servlets) with a everybody-has-it-so-we-too code-in-HTML solution
(JSP). OTOH, I don't know (yet), what ZOPE really is about.

Many of these APIs are overhyped and have a lot of potential to be used
incorrectly, IMO especially EJBs.

If I were to compare J2EE with a Python solution, I'd first construct
something equivalent in Python, i. e. a bunch of similar APIs thrown
together, like "WebWare for Python" (servlets, code-in-HTML, MiddleKit
for object-relational mapping), a distributed computing framework like
PYRO or omniORBpy, etc.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From starx at pacbell.net  Mon Apr  8 07:06:48 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 11:06:48 GMT
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4> <chris-CFFF94.21350207042002@corp.supernews.com> <3CB106AA.1AA3A5BF@engcorp.com> <chris-B48101.01250808042002@corp.supernews.com>
Message-ID: <Xns91EA2A0707D9ARASXnewsDFE1@130.133.1.4>

Chris || Sun 07 Apr 2002 10:25:08p:

> In article <3CB106AA.1AA3A5BF at engcorp.com>,
>  Peter Hansen <peter at engcorp.com> wrote:
> 
>> Chris wrote:
>> > 
>> > In article <Xns91E989929996CRASXnewsDFE1 at 130.133.1.4>,
>> >  Philip Swartzleonard <starx at pacbell.net> wrote:
>> > 
>> > > def percent_chance( percentage ):
>> > >       "Assumes percentage as number 0-100"
>> > >       if random.random() > (percentage / 100.0):
>> > >               return 1
>> > >       else:
>> > >               return 0
>> > 
>> > Might more concisely be written as:
>> > 
>> > def percent_chance( percentage ):
>> >      return random.random() > (percentage / 100.0)
>> 
>> But this is Python, and a newbie.  I think Philip's
>> effort was more suited to the audience, as it is
>> more readable. 
> 
> It's redundant, and "random.random() > (percentage / 100.0)" is no
> more readable with a "if" in front of it than it is with a "return".
> 
> I'm not trying to be rude, it just seems silly to do the work twice.

When defining small utility functions (such as these, and things like 
'cap number between two others' and 'calcualte slope', and that usually 
have only a few lines of code no matter how you stretch it ;), i like to 
keep things as open and as mentally simple as possible. And it seems 
like it takes less mental effort to look at my version and see the 
general nature of it than yours; where I have to parse the whole line, 
think about it, and realise that the return value is the result of the 
boolish-returning > operator, and not any of the other stuff on the 
line. I also like the pattern of 'compute somehting, and then act on 
it'... it seems to act in the same way we would if we eg. rolled a 
percentile die, then looked at the chart to see if we passed.

Not tryint to be rude, just a matter of style really. Also, i should go 
to bed before I completly stop functioning. (I should queue a message 
like this and look at it in the morning...... but i never do :\)

O-ya-su-mi
-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From gtaylor at lowebroadway.com  Tue Apr 23 04:57:17 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 23 Apr 2002 01:57:17 -0700
Subject: Python, PyQt make install problems on IRIX
References: <f0fd5987.0204220106.7673e345@posting.google.com> <3cc45c8d$0$307$e4fe514c@dreader1.news.xs4all.nl>
Message-ID: <f0fd5987.0204230057.25fdc747@posting.google.com>

Boudewijn Rempt <boud at valdyas.org> wrote in message news:<3cc45c8d$0$307$e4fe514c at dreader1.news.xs4all.nl>...
> Garry Taylor wrote:
> 
> > 
> > Has anyone else had this problem, or have any clues as to why it is
> > happening?
> > 
> 
> Well, Irix isn't exactly a common thing (I think I've
> seen only one machine running it in my life), but could
> it be possible that the installation process updates the
> dynamic library cache or something, which subsequently
> gets un-updated upon reboot? If you've installed PyQt
> from one xterm, can you run PyQt apps from another?

Yes, they will run from another terminal, I hav'nt heard of a Dynamic
Library cache, but this would seem an odd thing for a make install to
do, I was under the impression all they were supposed to do was to
copy all binaries and libraries to /usr/local or whereever. But if it
was this, how would I correct the problem?

Irix certainly is'nt all that common, in a way I think that kind of
helps people make installers, because there are not many hardware
configurations to support or variations of the OS, like Linux, there
are loads of variations on the same thing.

Thanks

Garry


From bhan at andrew.cmu.edu  Tue Apr 30 16:13:44 2002
From: bhan at andrew.cmu.edu (Benjamin Han)
Date: Tue, 30 Apr 2002 16:13:44 -0400
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <nlalaa.ko2.ln@127.0.0.1>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <200204292311.57363.bhan@andrew.cmu.edu>
    <nlalaa.ko2.ln@127.0.0.1>
Message-ID: <200204301613.44775.bhan@andrew.cmu.edu>

On Tuesday 30 April 2002 01:39 am, you wrote:
> Benjamin Han wrote:
>
> > How much is an effort in
> > converting Python into C++? Anyone would like to comment on some of
> > these questions, or point to some article, FAQ (esp. helpful from a
> > C++ programmer's point of view)?
>
> Well, let me give you a little heads up on the differences you might
> see.  Many C++ programmers are brought up to believe that C++ is
> synonymous with OOP.  You seem to tend that way, with your idea that
> implicit typecasting is an OO feature.  (If not, I apologize I seem
> facetious.)
>

I guess you're right - I did learn a lot from so many immensely useful 
replies. Time to shed my C++ skin and learn something new. Thank you guys!

(wouldn't it be nice if someone writes up a comparison list between Python and 
the other major OO languages, e.g., SmallTalk, C++, Java, etc.?)

Ben


From emile at fenx.com  Mon Apr 22 22:36:53 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 23 Apr 2002 02:36:53 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019510980.7097.python-list@python.org> <3CC48EDD.2BEE5215@tundraware.com> <qX0x8.82827$T%5.64632@atlpnn01.usenetserver.com> <aa2fu2$m33$1@panix1.panix.com>
Message-ID: <9N3x8.268614$Yv2.68820@rwcrnsc54>

Aahz, making a point I'm sure ;-),
>
> Spam spam spam spam spam spam spam spam spam spam spam spam spam spam

[Balance of Monty Python theme song snipped -- but I'd always wondered
just *exactly* how many times they said, and still stayed on key!]


Emile



From pinard at iro.umontreal.ca  Fri Apr  5 17:45:53 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 05 Apr 2002 17:45:53 -0500
Subject: The Topy project
In-Reply-To: <Xns91E7EF76BA86Acliechtigmxnet@62.2.16.82>
References: <mailman.1018039992.15896.python-list@python.org>
	<Xns91E7EF76BA86Acliechtigmxnet@62.2.16.82>
Message-ID: <oqadshpxwe.fsf@carouge.sram.qc.ca>

[Chris Liechti]

> I think the ideal solution would be to translate the source language to
> a meta language that describes the algorithm and its properties.

This is already what Topy does, yet the meta-language is produced and left
in tree form.  There was no need to make a surface representation for it.
Well, I did make a bastardised representation, but for debugging only!

> [...] then this cookbook could be used to generate code in any other
> language - however I don't think that such an approach is really possible
> with today's technologies...

I would agree, if you speak of "any" language.  Python has great lengths of
expressive power, it beats to this many other languages I learned before.
We may inject many small languages into Python, and still have plenty of
unused room on the Python side.  A few things appear to be missing (like
`goto's!), but we prefer to live without such things anyway, don't we?

The current meta-representation is surprisingly simple, at least so far.
It seems sufficient to convey most of Scheme (I have nothing to represent
lazy evaluation, nor continuations).  Some reasonable trickery is needed to
represent nested scopes and `let' loops: I found nothing intractable yet,
and no real problem will pop up.  Plans are still opened about `letrec';
hopefully, I'll find something not too ugly. :-)

It is notable to me that the meta-representation is much more simpler than
what would be needed to represent Python itself.  It prosaically means,
in particular, that it would not be thinkable to use Topy for translating
Python to something else, because that middle representation is just
unable to convey generic Python code.  This limitation is very expectable;
we cannot succeed into making Python compilers because it is difficult
to automatically guess the intent of a random source fragment, it looks
equally difficult to translate Python legibly to most other languages.
Or else, we would need a target language _significantly_ more expressive
than Python itself; and waiting for this, we should not hold our breath...

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





From me at mikerobin.com  Mon Apr 15 21:31:46 2002
From: me at mikerobin.com (Michael Robin)
Date: Tue, 16 Apr 2002 01:31:46 GMT
Subject: Diagramming library?
Message-ID: <6aLu8.48797$CH1.40700@sccrnsc02>

Does anyone know of a real-time diagramming
library that will work with Python? Cross platform
would be nice (wxWindows/wxPython, or a lib I
can wrap w/SWIG, etc),
but I'm willing to use a Win32 COM component as well.
Solution can be open or closed source.

Desired features :
- API-level, not batch.
- Dynamic layout of (a)cyclic graphs
- Several layout managers
- Ability to change/add/remove graph elements and labels in real-time,
  during graph evolution, preferably without a complete re-draw
- Callbacks to client code for graph info - shouldn't have
to build "shadow" data-structure. (May impose too much of a
performance hit?)
- Topological "Focus" areas to optimize display and
  updates in particular subsections
- Ability to handle about 10k nodes
- Hit-testing support for nodes and acrs

There are some good packages out there, but I haven't
seen much that supports the real-time nature and hit-testing, etc.

thanks,
mike
me at mikerobin.com




From jb at cascade-sys.com  Sat Apr 20 06:41:01 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sat, 20 Apr 2002 03:41:01 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
Message-ID: <3CC145BD.4660B5E9@cascade-sys.com>

Andrew <dalke at dalkescientific.com> replies:

> As I understand it, Python's Lisp heritage is pretty indirect.
> The most direct one I know of is docstrings, based on a similar
> feature in elisp.

Although they may be found in clisp, I don't think docstrings are regarded a general feature of "Lisp" or Scheme.

Seems to me the more prominent genetic similarities are the overall semantics: late binding, garbage collection, lists and sequences as a way of life.  Also, the fact that functions are values that may be free rebound to other names.  The semantics of and/or are rarely
encountered outside of lisp and derivatives.  MAP, APPLY, and especially LAMBDA also are pretty unique to Lisp.  This is VERY like Lisp and very UNLIKE "Algol like" languages.

I suppose Moshe Zadka is more accurate about the actual history without contradicting the point I was trying to make (that there are some strong similarities):

    Python and Scheme are very similar languages,
        that came from very different backgrounds.

Perhaps some of the Lisp heritage slipped in via ABC or via proposals from others without Guido's knowledge.

> If there was lisp heritage, why isn't there anything like car/cdr?

Easy.  Some things got left out.  They're not the SAME language.

List indexing and and slices are better and more powerful, an important step forward.

Note that CAR and CDR stand for Contents Address Register and Contents Data Register.  The original Lisp implementation (before some of you were born) was very specific to the IBM 709X architecture, in which it was easy to pack two addresses into a single word (with a
few bits to spare).  The more important LANGUAGE abstraction is that dotted pairs can be formed into lists and lists may be manipulated at a higher levels, and this same abstraction exists in Python.

> To back up my statement, here's some quotes from Guido I found in
> the Google usenet archives (no URLs since the Google links are too
> long)
> ] Also, the Python docs (to the extent that I wrote them)
> ] *never* attribute *anything* to Scheme or Lisp.  Python comes from the
> ] Algol family of languages.

In syntax yes.  In semantics it seems quite the opposite.

> ] (I'm a rabid Lisp-hater -- syntax-wise, not semantics-wise! -- and
> ] excessive parentheses in syntax annoy me.  Don't ever write return(i)
> ] or if(x==y): in your Python code! :-)

I hate Perl but I am well acquainted with how it works.  I agree excessive parentheses can be a bad thing.

> ] Common Lisp and Scheme.
> ]
> ] These languages are close to Python in their dynamic semantics,

Which was my point.

> but so
> ] different in their approach to syntax that a comparison becomes almost
> ] a religious argument: is Lisp's lack of syntax an advantage or a
> ] disadvantage?

I don't agree.  I think syntax and semantics are almost orthogonal, with syntax being the less complicated, less important of the two.  The fact that Python and Scheme have similar semantics make them very similar despite dramatic differences in their outer appearance.

Other than that, I think it's widely acknowledged that Lisp's lack of syntax is a disadvantage for most applications.  The exception might be the AI sub-domain of "program understanding," a tiny niche indeed.

> ] My conjecture: writing lambdas in Python is a bad habit, brought on
> ] from exposure to Lisp or Scheme.  Most Python code looks better after
> ] rewriting without lambdas.

How'd it get in there in the first place?  MAP?  APPLY? EVAL?  They're not in ABC or any other "Algol language" that I'm aware of.  There's nothing you can do with them in Python that you can't accomplish some other way (albiet with having to perhaps introduce
auxilliary control structures and temporary variables).

> There's also a relevant quote from Tim Peters:
> ] The problem Python set out to solve is that ABC was useless for
> ] real-life programming.  If Guido knows anything substantial about
> ] Lisp or Scheme, he's done an excellent job of hiding it for more
> ] than a decade.  I've seen him struggle with simple elisp, for God's
> ] sake <wink>.
>

I thought I read an essay of Guido's once long ago where he acknowledged a lisp heritage but I now see that I was mistaken.

And since I was mistaken about the historical asides, my suggestions about inline choice and len() are blithely ignored.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From aahz at pythoncraft.com  Sun Apr  7 11:58:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 11:58:13 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com>
Message-ID: <a8pqal$ftm$1@panix1.panix.com>

In article <ZMZr8.1903$Sw5.1235 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>
>It's becoming clear to me that Guido's biggest mistake was proposing
>this as a PEP. Since none of the arguments expressed as a result have
>swayed his opinion that the change is a good thing, he might as well
>have simply introduced the change by fiat.
>
>That way he could have ignored the resulting howls of protest
>rather than being forced to respond to them. Not much else would be
>different...

Well, no.  How could Guido have known there were no good arguments
against, save by proposing the PEP?  Do recall that the division
argument turned up some points that resulted in modification of the PEP,
including the critical issue that the default operation of / would *not*
change until Python 3.0.

I think it might be possible to make changes in the way Guido-introduced
PEPs get presented to reduce the amount of shouting from people like
Laura.  I hold no hopes that anything would get better results from
people like Arthur and Paul Rubin.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From robin at jessikat.fsnet.co.uk  Tue Apr 23 12:09:25 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 23 Apr 2002 17:09:25 +0100
Subject: Python call tree tool?
Message-ID: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>

I searched in vain for a tool that can create a call tree or  inverted
call tree for python modules/packages. It seems inconceivable that the
Python community hasn't created such a tool so I must be looking in the
wrong places. 

Any one know where such a beast is?
-- 
Robin Becker


From jb at cascade-sys.com  Sat Apr 27 02:02:08 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 26 Apr 2002 23:02:08 -0700
Subject: Python's Lisp heritage
References: <001201c1ed54$99d56760$445d4540@Dell2>
Message-ID: <3CCA3EDF.B7873D8D@cascade-sys.com>


Brian Quinlan wrote:

> William Park wrote:
> > > Lisp:   (+ 1 2 3 4 5 6 7 8 9 10)
> > > Python: 1+2+3+4+5+6+7+8+9+10
> >
> > Behold...
> >         seq -s+ 1 10
> > :-)
>
> Behold....
>         55
>
> :-)

LOL!

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From abf at fly.srk.fer.hr  Thu Apr 11 06:33:56 2002
From: abf at fly.srk.fer.hr (Ante Bagaric)
Date: 11 Apr 2002 10:33:56 GMT
Subject: A trivial question about print
Message-ID: <a93oqk$jf6$1@bagan.srce.hr>

Is there a way to prevent the print statement to automatically add a space
between two objects?

>>>print "a=",a
a= 1

how to make that cursed blank character dissappear?
Other than print "a=%d" % a because it wouldnt work for iteration of prints..

for i in range(5):
    print i,
	
gives 0, 1, 2, 3, 4 and I want 0,1,2,3,4

Yes, I know it's possible to circumvent this behaviour, but I'm curiuous
whether it can be done within the print statement alone.





From mixo at beth.uniforum.org.za  Wed Apr  3 04:52:17 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Wed, 03 Apr 2002 11:52:17 +0200
Subject: date / time
Message-ID: <3CAAD0D1.2040309@beth.uniforum.org.za>

Suppose there is a given date, 2002-03-01 (YYYY-MM-DD).
How can you get the day (2002-02-28 ) before the given date and
check that it is valid?



From aahz at pythoncraft.com  Tue Apr 16 00:30:42 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 00:30:42 -0400
Subject: Need help
References: <mailman.1018920450.17397.python-list@python.org>
Message-ID: <a9g9di$a77$1@panix1.panix.com>

In article <mailman.1018920450.17397.python-list at python.org>,
Faizal  <faizal at umland.com.my> wrote:
>
>I am still new(no exprience in programming)...&  need help from you guys...
>Can u teach me how to learn programming (with example....)

http://www.python.org/doc/Newbies.html
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From tim.one at comcast.net  Thu Apr 25 21:06:51 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 25 Apr 2002 21:06:51 -0400
Subject: Dynamic Binding
In-Reply-To: <3CC8A126.595FDDC0@cascade-sys.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEAPPBAA.tim.one@comcast.net>

[Tim]
> It seems this would have been a surprise to McCarthy <wink>:
>
>     http://www-formal.stanford.edu/jmc/history/lisp/node4.html
>
> See paragraph "d. Free variables".  When a user bumped into a function
> where the difference between dynamic and lexical scoping mattered, and
> lexical scoping was what they needed, "I must confess that I regarded
> this difficulty as just a bug".  The understanding of these issues was
> much muddier at the time.

[James J. Besemer]
> I disagree with your interpretation.

I take it you didn't follow the link and read the referenced paragraph.

> Rather he is saying that the implementation DID use dynamic "scoping,"
> that it worked just fine in most circumstances and that when people
> encountered a situation where "lexical scoping" would help McCarthy
> regarded their "difficulty" simply to be a bug.

I'll just finish the sentence I partially quoted:

     I must confess that I regarded this difficulty as just a bug and
     expressed confidence that Steve Russell would soon fix it.

Russell wrote the interpreter.  James Slagle wrote the code that was failing
due to dynamic scoping.  See the paper for more.  The differences between
lexical and dynamic scoping weren't clearly understood at the time.  Yes, it
used dynamic scoping, but the full implications of that implementation
choice weren't understood in advance, and some came as unwelcome surprises.

> I presume he means that the bug was in their code.  "It hurts when arg
> names conflict with your globals?  Don't DO that."

Reading the paper remains a recommended alternative to presumption <wink>.





From logstx at bellatlantic.net  Sat Apr  6 14:50:56 2002
From: logstx at bellatlantic.net (logistix)
Date: Sat, 06 Apr 2002 19:50:56 GMT
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4>
Message-ID: <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net>

Use python's type system:

>>> import types
>>> a, b = 1, 1.0
>>> if type(a) == types.IntType: print "int"
...
int
>>> if type(b) == types.IntType: print "int"
...
>>>



--
-

"Robert Spielmann" <varial at gmx.de> wrote in message
news:Xns91E8C6EF779FEvarialgmxde at 130.133.1.4...
> Hi,
>
> is there a way in python to check if a number is integer? My problem is
> the following: I need to know if sqrt(x) is an integer value or if there
> is a fractional part.
>
> Thanks for help
> Robert Spielmann




From gtaylor at lowebroadway.com  Tue Apr 30 05:10:31 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 30 Apr 2002 02:10:31 -0700
Subject: Destroying instances with PyQT
References: <f0fd5987.0204230240.56804d3f@posting.google.com> <3cc59504$0$41709$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <f0fd5987.0204300110.24a9539e@posting.google.com>

Boudewijn Rempt <boud at valdyas.org> wrote in message news:<3cc59504$0$41709$e4fe514c at dreader4.news.xs4all.nl>...
> Garry Taylor wrote:
> 
> > I have attached a QTimer to one of my objects, but I cannot seem to
> > get rid of it when I'm finished, which cause the whole object not to
> > be destroyed, normally my instances are erased simply by erasing the
> > Dictionary in which they reside, which has always worked before the
> > QTimer was added. In the QT manual it says you can destruct an
> > instance like this:
> > 
> > QTimer::~QTimer ()
> > 
> > How can I translate this into Python?
> > 
> 
> Simple remove the reference to the QObject that is the parent
> of the QTimer. The following code removes the owning object
> after ten ticks:
> 
> 
> from qt import *
> import sys
> 
> class Q(QObject):
> 
>         def __init__(self):
>                 QObject.__init__(self)
>                 self.o=O(self)
>                 
>         def delete(self):
>                 self.o = None
> 
> class O(QObject):
> 
>         def __init__(self, parent):
>                 QObject.__init__(self)
>                 self.parent = parent
>                 self.i = 0
>                 self.t = QTimer(self)
>                 QObject.connect(self.t, SIGNAL("timeout()"), self.p)
>                 self.t.start(0, 0)
> 
>         def p(self):
>                 self.i += 1
>                 print "bla", self.i
>                 if self.i == 10:
>                         self.parent.delete()
> 
> q = Q()
> app=QApplication(sys.argv)
> app.exec_loop()

Thanks, that works fine, but it's not any different to what I was
doing, except that I had sub-classed the QTimer. Anyway, problem
solved.

Thanks

Garry


From alain at onesite.org  Sat Apr 27 19:57:54 2002
From: alain at onesite.org (Alain Tesio)
Date: Sun, 28 Apr 2002 01:57:54 +0200
Subject: local variable not found
References: <pan.2002.04.27.23.01.32.51309.30401@onesite.org> <HzEy8.28287$8D3.754996@news1.tin.it>
Message-ID: <pan.2002.04.28.01.57.54.335756.30401@onesite.org>

On Sat, 27 Apr 2002 23:17:27 +0200, Alex Martelli wrote:

> Change your eval(...) to an exec ... and it should work.

Thanks, it works fine with exec indeed.

> It's close to impossible that such an approach is
> anywhere like halfway-decent, by the way.  eval and
> exec make sense a tiny percentage of the time, but
> (in my experience) only when used with explicit
> dictionaries, NOT by letting them default to locals()
> and globals().

My problem is to make html templates where the user
can insert python code.
I'm working on a dynamic interface to mailing list
archives : http://onesite.org/exarch/

Alain


From hopfgartner at rolmail.net  Mon Apr 15 04:04:27 2002
From: hopfgartner at rolmail.net (Peter Hopfgartner)
Date: Mon, 15 Apr 2002 08:04:27 GMT
Subject: SAP DCOM Connector, return type of DimAs
Message-ID: <pan.2002.04.15.09.02.11.851376.559@rolmail.net>

Dear members of the Python community,

is there anybody, who used Python successfully to connect to SAP through a
DCOM Connector?

My current problem is:

I get the COM project through:


        # Rufe die COM-Komponente auf
        self.sap_com = win32com.client.Dispatch('SAP.BAPI.1')
(... connect successfully to SAP ... )

        self.porders = self.sap_com.GetSAPObject('PlannedOrder')
        self.header =  self.sap_com.DimAs(self.porders, 'Create',
                                                                   'HeaderData')

        self.header =  self.porders.DimAs('Create','HeaderData')

Then, I would like to insert some data through self.header.Value, but this
is seen only as an integer from Python.

In Visual Basic the data is inserted as

header.Value("Material") = 123456

In Visual Basic everything works fine.

All COM objects are early bounded, as described in Hammonds (excellent)
Python on Win32 book.

I'm using ActiveState Python 2.2 on Windows 2000 Server.

If anybody could help me he will earn my eternal gratitude.

Peter


From noah at noah.org  Sat Apr 13 14:10:06 2002
From: noah at noah.org (Noah)
Date: Sat, 13 Apr 2002 11:10:06 -0700
Subject: Python and spacecraft onboard or ground control scripting
In-Reply-To: <3CB853CA.8050507@cutter.rexx.com>
Message-ID: <PAENINBOOODFDEPOKPKCOEAPCLAA.noah@noah.org>

Seg faults are the least of your worries for
mission critical applications. Deterministic threading
becomes the real headache. Deterministic garbage collection and 
deterministic thread scheduling; validated compilers; 
strong, static types are all the hallmarks of 
mission critical application languages.
I can't imagine a dynamically typed language like Python 
(in its current form) could ever be suitable for 
real-time embedded systems or life-critical controls.

I think there is an initiative to make a Java VM that
will meet the requirements for real-time programming.
RTJ or JRT or something like that...
Maybe some version of Jython could be created that
will compile class files for that JVM. You would probably
have to give up most of Python's dynamic features to make
it work.

I wonder what those stackless Python people are up to...

Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Dave Kuhlman
Sent: Saturday, April 13, 2002 8:51 AM
To: python-list at python.org
Subject: Re: Python and spacecraft onboard or ground control scripting


Cameron Laird wrote:

> Folklore has it that Tcl *did* travel to Mars, and its
> interpretive nature was crucial in responding to some
> mission contingency; the control team needed to upload
> new code during flight.  We ought to document this ...
> 
> While I understand the observations that Python, Tcl, 
> and so on surely aren't safe enough for medical devices,
> nuclear plants, air traffic control, military devices,
> and so on, I have reasonably certain knowledge of instances
> of each of these.
> 


The idea that Python, Tcl, and so on aren't safe enough for medical 
devices sounds like conventional wisdom to me.  Someday a scripting 
language will come along which is so good that it proves this 
conventional wisdom wrong, and does this so conclusively that we are 
forced to give up that idea.  I believe that Python was/is/will be that 
language.

Are there examples and evidence that Python fails when applied to 
long-running, mission critical tasks?

By the way, I write C++ code.  And, the idea that C++ code is safer than 
Python code seems screwy to me.  Can you say "seg fault"?  I thought so.

   - Dave










From fredp at mygale.org.nospam  Mon Apr  8 12:48:52 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 8 Apr 2002 16:48:52 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <Xns91E77A4404E25PaCmAnRDLM@161.48.128.20> <3cb1c295.11651323@news.laplaza.org>
Message-ID: <Xns91EABFD35CF37PaCmAnRDLM@161.48.128.20>

mats at laplaza.org (Mats Wichmann) said :

>:Yes, it was ported during the beta phase last year. I remember seeing
>the :announcement on the developer's list when it was on Yahoo Groups,
>but it :has moved and I don't know where it is hosted now.
> SourceForge.

Thanks, I'll have a look over there...
 
> And as to the availability, Sharp has replied to this question with
> some pablum about how each individual sales region (usually ==
> country) has to commit to enough volume from the factory in order to
> make the creature available.  That's happened in the US, and
> UK/Ireland for the early-access model (5000) which is now no longer
> available (I have no info on whether there's going to continue to be a
> Developer box available), and for the released model, the 5500,
> Germany has added.  And yes the whole business of local configuration
> and packaging rules lies behind this.  You can't even buy one in
> Canada yet.
 
Not to mention localization and keyboards :)

> I've been lurking on a lot of the forums recently and there are some
> real problems with the Zuarus as a consumer-grade device.

Such as, in a nutshell ?...

> Programmers,
> who don't mind rough edges generally, love it.  I have some worries
> about its' survival in the marketplace.

All the more reason to grab one while it lasts... (I love lost causes, I 
still use a Psion Series 5mx :-)

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


From mwh at python.net  Fri Apr 26 10:40:59 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 26 Apr 2002 14:40:59 GMT
Subject: What is NotImplemented ?
References: <slrnacinls.j5.alf@virgo.logilab.fr>
Message-ID: <lkn0vq34fi.fsf@pc150.maths.bris.ac.uk>

Alexandre Fayolle <alf at logilab.fr> writes:

> A quick question, here. What is the built in NotImplemented object? 
> And what can you do with it? I thought I could raise it as an 
> exception, but this is not the case (I use NotImplementedError instead).

I think you can return it from implementations of, say, __add__ to
mean roughly "I don't know what to do here".

Eg:

>>> class Num: 
...  def __add__(self, other):
...   return NotImplemented
...  def __radd__(self, other):
...   print "radd!"
...   return 1
... 
>>> Num() + Num()
radd!
1

Returning NotImplemented from the __add__ makes Python go off and call
the __radd__ method.

HTH,
M.

-- 
  ... but I guess there are some things that are so gross you just have
  to forget, or it'll destroy something within you.  perl is the first
  such thing I have known.              -- Erik Naggum, comp.lang.lisp


From mgilfix at eecs.tufts.edu  Wed Apr 17 23:55:30 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 23:55:30 -0400
Subject: Problems with copies (and references) to lists
In-Reply-To: <3CBE0F6C.17CA6751@uniserve.com>; from bvdpoel@uniserve.com on Wed, Apr 17, 2002 at 05:12:28PM -0700
References: <3CBE0F6C.17CA6751@uniserve.com>
Message-ID: <20020417235528.G924@eecs.tufts.edu>

  I found this a little misleading. What's 's' have to do with
anything? Here's a couple of comments:

  * Check out 'import copy' and use that to copy lists.
  * Try using map to apply functions to various elements
    without any iteration. That might help some.

               -- Mike

On Wed, Apr 17 @ 17:12, bvdpoel at uniserve.com wrote:
> 
> Took awhile, but I do have the following bit of code working as it
> should. The doc string is misleading, but I'm not sure of how better to
> describe it out of context... The important thing is, it is giving the
> right result, and it is NOT changing the original list.
> 
> -------
> def doubleUp(a, fact):
>         """ Makes fact/2 copies of list [ [s, l, v,...] ].
>                 New copy has all values of l divided by fact/2 and
>                 copies of s set to s/2 and 50+(s/2).
> 		fact will always be a power of 2.
>         """
> 
>         new = a
>         while fact >1:
>                 b1 = []
>                 b2 = []
>                 for t in new[:]:
>                         n=t[:]          # this prevents munging our
> original
>                         n[0] /= 2
>                         n[1] /= 2
>                         b1+=[n[:]]      # [:] forces new copy of n!
>                         n[0] += 50
>                         b2+=[n]
>                 fact /=2
>                 new = b1 + b2
>         return new
> 
> a = [[0.0, 20.0, 66, 44],[50.0, 20.0, 55, 33]]
> print doubleUp(a, 2)
> print doubleUp(a,4)
> print a
> 
> --------
> 
> But, I'm wondering if it is at all efficient. Or even good python. Seems
> that the number of [:] needed just makes for ugly code.
> 
> -- 
> Bob van der Poel ** Wynndel, British Columbia, CANADA **
> EMAIL: bvdpoel at uniserve.com
> WWW:   http://users.uniserve.com/~bvdpoel
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (bvdpoel)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From bh at intevation.de  Thu Apr 18 05:58:38 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 18 Apr 2002 11:58:38 +0200
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <a9l0bo$h1j$1@panix1.panix.com>
Message-ID: <6qd6wx72gh.fsf@abnoba.intevation.de>

aahz at pythoncraft.com (Aahz) writes:

> Heh. That's because of the way __iadd__ works. Immutable objects must
> return a new object with __iadd__;

Why would you want to implement __iadd__ for immutable objects? Just
__add__ (and maybe __radd__) should be enough.

> mutable objects *MAY* return self (and usually do).

Wouldn't it have been enough to assume that __iadd__ always, if defined,
modifies the object in place?

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From JamesL at Lugoj.com  Thu Apr  4 21:46:09 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 5 Apr 2002 02:46:09 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com>
Message-ID: <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83>

David Bolen <db3l at fitlinxx.com> wrote:
> Guido did cover this explicitly in one of his responses to the bool
> threads [ ... ]

Thank you. I only just resubscribed to comp.lang.python after many months 
of absence and was only reading a handful of articles of immediate interest 
and was not following the PEP 285 thread until Guid posted the ACCEPTED 
message. When I looked at it I saw a "class bool(int)" definition, so I 
wasn't sure how instances of such a (pseudo?) class would be handled in a 
dictionary. The statement "The values False and True will be singletons, 
like None," further added to my confusion, since None is a valid dictionary 
key, distinct from all other values, which when combined with the class 
definition pretty much sealed the confusion. I assumed the "ACCEPTED: PEP 
285" posting would have subsumed previous discussions, including any 
concerning the class instances as dictionary keys. I couldn't find a clear 
answer there, hence my question.

My apologies to those who have sent me angry e-mails for not realizing that 
PEPs do not stand alone from the lengthy discussion threads on Usenet. The 
fault is entirely mine.


From martin at v.loewis.de  Fri Apr  5 02:19:09 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Apr 2002 09:19:09 +0200
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de> <f445a2fe.0204041708.6fd6fd0d@posting.google.com>
Message-ID: <m37knm395e.fsf@mira.informatik.hu-berlin.de>

seb_haase at yahoo.com (Sebastian Haase) writes:

> I think that wouldn't work - since I'm still interested in debugging
> _my_ stuff - so that needs me to #define _DEBUG

Can you elaborate? If you want code that is only active under
debugging, why can't you wrap it in #ifdef HAASE_DEBUG or some such?

If you think that you would not get symbolic debugger support if
_DEBUG is not defined, you are wrong.

Regards,
Martin



From vinay_sajip at yahoo.co.uk  Wed Apr  3 18:23:07 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 3 Apr 2002 15:23:07 -0800
Subject: ANN: Logging Module v0.4.1 released
Message-ID: <2e37dc1.0204031523.29ab3d6@posting.google.com>

A new version of the proposed Python standard logging module (as per
PEP 282) has been released. You can get all the information from

http://www.red-dove.com/python_logging.html

There are "download" and "recent changes" links at the top of that
page, and there is quite a lot of new stuff including support for XML,
HTTP and SOAP, user-defined logger classes, a ConfigParser-based
configuration file format, a Tkinter GUI-based configurator and more!

As always, your feedback is most welcome (especially bug reports,
patches and suggestions for improvement).

Cheers

Vinay Sajip
Red Dove Consultants Ltd.

Changes since the last version:
=================================
Bug fix SMTPHandler - extra \r\n needed (Oleg Orlov)
Added BufferingHandler, BufferingFormatter
Renamed getChainedPriority to getEffectiveLevel
Removed Logger.getRoot as it is redundant
Added log_test9.py to test Buffering classes and to show an
XMLFormatter example.
Added setLoggerClass.
Added log_test10.py to test setLoggerClass, using an example
Logger-derived class which outputs exception info even for DEBUG level
logging calls
Added log_test11.py to test a buffering implementation of SMTPHandler
Changed logging call implementation to allow keyword arguments (Kevin
Butler and others)
Changed default SysLogHandler implementation.
Renamed "additive" to "propagate" as it better describes the
attribute.
Added HTTPHandler.
Modified logrecv.py to remove "both" option and to add "HTTP" and
"SOAP" options (SOAP option needs you to have PyXML-0.6.6 and ZSI
installed - for logrecv.py only, and not for the core logging module
itself).
Added log_test12.py to test HTTPHandler.
Added log_test13.py to test SOAPHandler.
Formatted to Python source guidelines (spaces, indent of 4, within 80
columns).
More method renamings (result of feedback) - _handle() renamed to
emit(), _logRecord() renamed to handle(). Renamed FATAL to CRITICAL
(David Goodger), but left fatal() and FATAL in (until PEP is changed)
Changed configuration file format to ConfigParser format.
Factored filter application functionality out to a new Filterer class.
The isLoggable() method is renamed to filter() in both Filter and
Filterer classes.
Altered SMTPHandler __init__ to accept (host, port) for the mail
internet address.
Added GUI configurator which uses Tkinter and the new configuration
file format. (See logconf.py and an example configuration file in
logconf.ini)
Altered log_test3.py to test with the new file format.


From djc at object-craft.com.au  Thu Apr 18 08:08:28 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 18 Apr 2002 22:08:28 +1000
Subject: chopping csv-files (newbie)
References: <3CBE768C.844C90FF@itea.ntnu.no> <m3it6p5r17.fsf@ferret.object-craft.com.au> <3CBEB009.8B4B7A56@stop.spam.volvo.com>
Message-ID: <m3bschb45f.fsf@ferret.object-craft.com.au>

>>>>> "Mikael" == Mikael Lex?n <mikael.lexen at stop.spam.volvo.com> writes:

Mikael> ImportError: DLL load failed with error code 193

Mikael> What's wrong? I'am on NT4.

I probably need to recompile the module with a more recent Python.
You can do it if you have access to MSVC.

- Dave

-- 
http://www.object-craft.com.au


From mwh at python.net  Thu Apr 18 06:08:01 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 10:08:01 GMT
Subject: ranlib problem building 2.2.1 on Mac OS X
References: <m28z7l7hc4.fsf@javamonkey.com>
Message-ID: <lk1yddmi2m.fsf@pc150.maths.bris.ac.uk>

Peter Seibel <peter at localhost.i-did-not-set--mail-host-address--so-shoot-me> writes:

> I'm trying to build Python 2.2.1 from the source tarball on Mac OS X.

Wierd.  Built fine for me.  What version of Mac OS X is it?  I think
you need at least 10.1.

You could also trying asking on the macpython-sig.

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme


From googlegroups at sixtyten.org  Sun Apr 21 10:29:08 2002
From: googlegroups at sixtyten.org (Allan Crooks)
Date: 21 Apr 2002 07:29:08 -0700
Subject: Modules that provide differing functionality amongst different Python versions...
Message-ID: <304d20df.0204210629.3f2b4d58@posting.google.com>

Hi,

I was wondering what the best approach is on providing libraries for
varying versions of Python?

My aim is to distribute a module that will provide a set of core
functions, but provide more functions to later versions of Python.

For example, I make a module that works on 2.0, but if used on 2.1,
will provide more functions to say, do with the warning framework. If
used on 2.2, certain functions that return lists in earlier versions
of Python, will use generators instead.

Rather than keeping several versions of the same module (but for use
with different versions of Python), I was wondering if there was a
more elegant solution?

I had thought initially of doing something like this:

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

<2.0 code>
if sys.version[:2] < (2, 1): return

<2.1 code>
if sys.version[:2] < (2, 2): return

<2.2 code>

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

But that wouldn't work, because earlier versions of Python would still
report SyntaxErrors, as it would have a problem with iterator
statements.

Another thought I've had is doing it by specifying, say for module
'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
what versions of Python it is running on, and then compiling all
modules it can deal with, and then importing all objects into it's
namespace.

That idea sounds a bit 'hacky' to me, so can anyone suggest a nicer
way of doing this? There's also the problem of trying to specify class
definitions which provide more methods in later versions of Python. Do
later versions subclass earlier ones, or do they just interfere with
the namespace?

Thanks,
Allan.


From brian at sweetapp.com  Sun Apr 28 19:05:49 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 28 Apr 2002 16:05:49 -0700
Subject: Update locals()
In-Reply-To: <20020428205757.GH16217@devel.trillke>
Message-ID: <003d01c1ef09$3d4c28b0$445d4540@Dell2>

>   locals()
>   Return a dictionary representing the current local symbol table.
>   Warning: The contents of this dictionary should not be modified;
changes
>   may not affect the values of local variables used by the
interpreter.

The reason that changes to the locals dictionary might not be visible is
that LOAD_FAST instructions, which are used for local variable lookup,
use an indexed list and not the locals dictionary.

Cheers,
Brian





From logiplexsoftware at earthlink.net  Thu Apr  4 14:07:07 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 4 Apr 2002 11:07:07 -0800
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: <m3bsczfgic.fsf@mira.informatik.hu-berlin.de>
References: <200204032232.g33MWqJ28933@odiug.zope.com>
	<3CAB8C99.232A8571@metaslash.com>
	<m3bsczfgic.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020404110707.5077d9dc.logiplexsoftware@earthlink.net>

On 04 Apr 2002 20:47:07 +0200
Martin v. Loewis wrote:

> While Gustavo suggests to let pass 'a is True', I recommend to give a
> warning for this - people who want to see whether something is True
> should write 'True is a'.

But only if they do it in a gravelly Yoda voice ;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mwh at python.net  Mon Apr  8 08:08:11 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 8 Apr 2002 12:08:11 GMT
Subject: [Python-Dev] RE: RELEASED: Python 2.2.1c2
References: <NEBBIGHKAHJGMPJEEAGPKENMGGAA.fp@contact.de> <mailman.1017228282.28692.python-list@python.org> <a8n6nj$j88$1@inputplus.demon.co.uk>
Message-ID: <uhemmbd8p.fsf@python.net>

ralph at inputplus.demon.co.uk (Ralph Corderoy) writes:

> Hi Martin,
> 
> > > > Note for AIX users: there are still problems with AIX in this
> > > > release, and there has not been time to repair them without fear
> > > > of breaking platforms that currently work.  Hopefully by the time
> > > > of the release of 2.2.1 final we will be able to provide a recipe
> > > > that allows AIX users to build and run Python.  [...]
> > > 
> > > Hmm, can someone shed more light on this issue?
> > 
> > I think this is a red herring: Some user, for some reason, does not
> > have cc_r. According to a number of other AIX users, cc_r should
> > always be available, so it appears to be an installation bug of the
> > AIX system in question. I don't think it deserves fixing.
> 
> That `some user' is me.  Can you tell me what version of AIX those
> users who say it should always be available are referring to?  As the
> bug report says, AIX 3.2.5 is what I'm using and cc_r doesn't exist.
> It isn't a requirement to install it at the same time as cc.
> 
> Perl has always built happily on AIX 3.2.5 in my experience and given
> all Python has to do is search through the list [cc_r cc] for the first
> one it finds when the OS is AIX I don't see the big deal.  Escpecially
> since it already has a `when OS is AIX do a bodge' statement.

But you can make it compile with ./configure --without-gcc, right?

> I'll add more to the bug report shortly when I've time.

You don't have very long!

Cheers,
M.

-- 
  And then the character-only displays went away (leading to
  increasingly silly graphical effects and finally to ads on
  web pages).                      -- John W. Baxter, comp.lang.python


From gordon.scott at peregrine.com  Thu Apr  4 11:15:24 2002
From: gordon.scott at peregrine.com (Gordon Scott)
Date: Thu, 4 Apr 2002 08:15:24 -0800 
Subject: Catalog of Python Patterns? [LONG!]
Message-ID: <D54BF26DB5A7D511B36B00508B9596A402472A24@sdgcaexch2.peregrine.com>

There IS a separate Smalltalk companion book to DP (but different authors).
I'm currently using in it a class I'm taking.  It describes implementations
of the 23 GoF patterns in Smalltalk, utilizing it's dynamic aspects that 
are sometimes much different than DP's C++ implementations and probably
more applicable to python.

If you're interested in patterns for Python it's probably at least worth
a trip to the bookstore to flip through...if you can find it. Might be a
special order item..
 

-----Original Message-----
From: jimd at vega.starshine.org [mailto:jimd at vega.starshine.org]
Sent: Thursday, April 04, 2002 1:03 AM
To: python-list at python.org
Subject: Re: Catalog of Python Patterns? [LONG!]


In article <mailman.1017886389.26908.python-list at python.org>, Delaney, 
Timothy wrote:

>> From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
>> jimd at vega.starshine.org (Jim Dennis) writes:

>>>  At a minimum it should list each of the 23 GoF patterns, and
>>>  at least one minimal Python implementation of that pattern.

>> Yucch, I've never been a fan of that book, it seems too cultish to me.

> More importantly, many of the patterns are not applicable to Python.
_Design
> Patterns_ comes from a C/C++ point of view ... there is also _Design
> Patterns for Smalltalk_ (or something along those lines) which is
apparently
> much more applicable.

 Actually the GoF book has some C++ and some Smalltalk examples.  They
 aren't separate.

> It is useful to *know* the GoF patterns, but it is important not to be a
> slave to them.

>Tim Delaney

 I agree that the book has somewhat of a cult following, and I'm
 naturally wary of cults.  I also agree that one should not be a 
 slave to fashion, and that some of the hype of "patterns" (and
 XP, and ...) is "fashion" more than practical.

 However, I'm trying to understand them well enough to distinguish
 which parts of useful versus cosmetic.

 I disagree (with one followup that's not quoted herein) that
 patterns are not applicable to Python.  The author of that comment
 claimed that GoF was a C++ specific book (which is untrue, as I 
 point out above).  Python is much more like Smalltalk than C++,
 and the GoF specific chose it because it is a dynamic language
 --- for its contrast to C++ and its ilk.

 As I pointed out in my post, Python employs several patterns in
 its core and standard libraries.  Every class is a Factory, the
 Bastion module returns Proxy objects, we see Iterators all over
 the place, etc.  So I still want to see more.


-- 
http://mail.python.org/mailman/listinfo/python-list




From sholden at holdenweb.com  Mon Apr 29 09:15:57 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 09:15:57 -0400
Subject: Call for kjBuckets.c [was Re: relational database?]
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <5yEp8.5051$hi7.18667@news-server.bigpond.net.au> <3CA7192C.E1C8974A@engcorp.com> <mailman.1017612966.7043.python-list@python.org> <uco40vl11ipj04@corp.supernews.com> <mailman.1020024596.32196.python-list@python.org>
Message-ID: <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com>

"Oleg Broytmann" <phd at phd.pp.ru> wrote in message
news:mailman.1020024596.32196.python-list at python.org...
> On Sun, Apr 28, 2002 at 08:11:36AM -0700, Bob Hyatt wrote:
> > I have had a fruitless search for Gadfly. As you noted, the website for
> > Chordate is not responding.
>
>    I am keeping (temporary) a copy of kjbuckets at
> http://phd.pp.ru/Software/Python/misc/ and mirrors:
> http://phd.by.ru/Software/Python/misc/
> http://phd2.chat.ru/Software/Python/misc/
>    I am looking for a permanent maintainer.
>
>    Gadfly now lives in the Zope sources at www.zope.org.
>

Oleg, hi! I also remember seeing somewhere patches that moved gadfly to re
usage? gadfly.sf.net is a placeholder, it seems to me.

regards
 Steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From emile at fenx.com  Tue Apr  2 16:16:30 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 02 Apr 2002 21:16:30 GMT
Subject: Zope Q: no restart button & shutdown problem
References: <ed6d1c1a.0204021258.53585416@posting.google.com>
Message-ID: <Ocpq8.160483$af7.77579@rwcrnsc53>

"A M" <aseem at neurobs.com> wrote in message
news:ed6d1c1a.0204021258.53585416 at posting.google.com...
> Hi all,
> I am running  (Zope 2.5.0 (binary release, python 2.1, win32-x86),
> python 2.1.2, win32) on Win2K. Everythin seems to be running fine. My
> only problem is that in the Control Panel I only get the Shutdown
> button and there is no restart button. I just installed the binary and
> did as was told in the docs, but I cant seem to find an answer or
> explantion of this behaviour. Any help would be greatly appreciated.
>

I'd suspect you're running from the command line instead of as a
service.

--

Emile van Sebille
emile at fenx.com

---------



From lists.ASkwar at DigitalProjects.com  Fri Apr 26 16:39:46 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 22:39:46 +0200
Subject: Removing dupes from a list
Message-ID: <20020426203946.GV21205@teich.Garten.DigitalProjects.com>

Hi!

What's the fastest way to make sure that a list only contains unique
entries?

Currently, I get all the elements from the list and store them as keys
in a dictionary.  I then use dict.keys() as the new value for the list,
like so:

def de_dupe(liste):
        d = {}
        for element in liste:
                d[element] = None
        return list(d)

But this will obviously only work, if all the elements of list are
immutable.

Next, I came up with this class:

class de_duper:
        def __init__(self, liste):
                # Store the last compared element
                self.__last = None
                check_list = liste[:]
                check_list.sort()
                self.clean = filter(self.__check, check_list)
        def __check(self, current_element):
                check = self.__last
                self.__last = current_element
                return cmp(check, current_element)
                
Here, I use filter and cmp.  The reason that I use a class is, that cmp
requires two arguments and I thus need to store the last checked value.
Futher, this approach requires that the list first needs to be sorted.
Because of the sort, the order of the elements is all mixed up, compared
to the original list.

What I'm now looking for, is an easy and fast way to remove duplicates
from a list without destroying the list order.  I'm looking for
something like PHP's array_unique().  Is there something like this in
Python?

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 2 days 13 hours 6 minutes




From jim at jim.jim  Thu Apr  4 13:28:27 2002
From: jim at jim.jim (jimh)
Date: Thu, 4 Apr 2002 10:28:27 -0800
Subject: cgi.py performance
References: <a8ght4$o8o$1@web1.cup.hp.com> <slrnaaok15.4ei.jon+usenet@snowy.squish.net>
Message-ID: <a8i60e$csg$1@web1.cup.hp.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnaaok15.4ei.jon+usenet at snowy.squish.net...
> In article <a8ght4$o8o$1 at web1.cup.hp.com>, jimh wrote:
> > I am using cgi.py to handle the transfer of large incoming files.
Because
> > of the way that cgi.py handles multipart/form-data, the performance is
VERY
> > poor.
>
> You could try jon.cgi from jonpy, it might be more efficient.
> http://jonpy.sourceforge.net/

Hey, that looks pretty cool.  I do have one question before I try it out: in
mime.py, Entity.__init()__, which splits up the data on the boundaries,
loads the entire file at once, if I am not mistaken ("msg = fp.read()").
What will happen if I am transferring a very large file, 50-100MB?

Thanks,
Jim




From gimbo at ftech.net  Mon Apr 22 12:51:35 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Mon, 22 Apr 2002 17:51:35 +0100
Subject: Interesting topics
In-Reply-To: <Xns91F85897F9A6DRASXnewsDFE1@130.133.1.4>
References: <mailman.1019468028.28617.python-list@python.org> <Xns91F85897F9A6DRASXnewsDFE1@130.133.1.4>
Message-ID: <20020422165134.GD25778@andy.tynant.ftech.net>

On Mon, Apr 22, 2002 at 03:41:18PM +0000, Philip Swartzleonard wrote:
> Doru-Catalin Togea || Mon 22 Apr 2002 02:32:32a:
> > Personally, I don't really care, YET. I have to delete over 150
> > postings everyday anyways, so I've become really proficient at hitting
> > the 'D' key which marks mails for deletion in PINE. But if the ratio
> > of noise to signal increases too much on this list, it can become
> > boring to remain on it.
> 
> Remember, the mailing list you're on is tied directly into a usenet 
> newsgroup called comp.lang.python. Newsgroups and the related software 
> all scale a lot better to this kind of volume, with the ability to just 
> read what you want and ignore the rest, and i'm not sure in general, but 
> i'm pretty sure it's a lot faster than it would be to get all this 
> through the little pipe from our crappy email servers here =).

OTOH, if you're determined to use your favourite MUA to read the
list/group (as I am, 'cos mutt rocks), it helps to get to know that
MUA's features as well as you can.  I don't know if pine has an
equivalent, but in mutt I find Ctrl-D (delete entire thread) and
Meta-D (delete subthread) _indispensable_ around here.  :-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From donn at u.washington.edu  Thu Apr 11 16:32:26 2002
From: donn at u.washington.edu (Donn Cave)
Date: 11 Apr 2002 20:32:26 GMT
Subject: Communication between a Python process and a Python cgi script
References: <3CB5E224.4E3DC5BA@olen.to>
Message-ID: <a94rsq$29e8$1@nntp6.u.washington.edu>

Quoth Joonas Paalasmaa <joonas at olen.to>:
| I have a Python program started by root that does some magic stuff
| every 10 minutes. So the program in an infinite loop of sleep, do stuff,
| sleep, do stuff... How can I make a cgi-script to end the process's
| sleep phase and make it to "do stuff" immediately. I tried signals, but
| it is propably too risky because the process is run as root. Any
| suggestions?

You could try a named pipe -
$ mkfifo /tmp/fifo
$ cat /tmp/ptest
pipe = open('/tmp/fifo', 'r')
while 1:
    ping = pipe.read()
    print 'ping', repr(ping)
$ python /tmp/ptest &
$ echo junk > /tmp/fifo
ping 'junk\n'
$ echo junk > /tmp/fifo
ping 'junk\n'

Find a better name and place for the fifo, check that it will be
writable by the CGI process.  Consider security issues, who will
be able to do this - might be no compromise, yet you could have
a "denial of service" kind of attack.

	Donn Cave, donn at u.washington.edu


From johnroth at ameritech.net  Mon Apr 29 07:38:53 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 29 Apr 2002 07:38:53 -0400
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <ucqc5464pt5b95@news.supernews.com>

"Raymond Hettinger" <python at rcn.com> wrote in message
news:aaiu6e$pfr$1 at bob.news.rcn.net...
> Please send me comments on a draft PEP deprecating
> and relocating six built-in functions.  The text file is at:
> http://users.rcn.com/python/download/deppep.txt
>
> This is version two and incorporates the comments received
> from David, Alex, Brian, Martin, and Chris.
>
> It is also listed below.
>
>
> Raymond Hettinger

-1.

I don't see any benefit to eliminating five of the
builtins from the global name space. Making a
separate module for functional programming
style functions seems reasonable, and there is
no particular reason why the three functions in
that category couldn't appear in both places,
at least until 3.0.

If we're going to keep the functions in the
global name space until 3.0, there is no need
for a depreciation warning, and no need for
any transition mechanism.

On the other hand, eliminating input() seems
to be a good thing, as it is a snare for the
unwary.

On a more global comment, I think we should
get to work on 3.0 now. There are a lot of good
ideas floating around that are not forward
compatible, and that need some time to think
about them.

John Roth







From johnroth at ameritech.net  Sat Apr  6 10:24:19 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 10:24:19 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org>
Message-ID: <uau4ebrvk4e563@news.supernews.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1017865089.20676.python-list at python.org...
> On Tue, 2 Apr 2002 21:21:27 -0500
> John Roth wrote:
>
> >
> > "Laura Creighton" <lac at strakt.com> wrote in message
> > news:mailman.1017755045.4308.python-list at python.org...
> > > Dear Guido:
> > >
> > > I would first like to thank you for giving us an opportunity to
> > respond.
> > > I have spent most of the weekend thinking and writing a reply to
this,
> > > and I think that this has made me a better teacher.  For this I am
> > > grateful.  I realise that this is rather long, but I have
condensed
> > > it substantially from my efforts earlier this weekend.   Thank you
> > > for taking the time to read this.
> > >
> > > Laura Creighton
> > >
> > > -----------------------------
> > >
> > I'm not going to quote this, as it's already too long. I read up to
> > the diatribe on the Swedish leap year, and I went AAAARRRRGGGH!
> >
> > I really don't care if you want the silly function to return
> > a Boolean, the number of leap days in the year, or the name
> > of the idiot parliamentarian who decided to fix the calendar
> > so that 1712 was different from the rest of Europe.
> >
> > HOWEVER, if you want true or false as the return
> > value, the name should be something like 'isLeapYear()'.
> > If you want the number of leap days in the year, then
> > it should be named something like 'numberOfLeapDays().'
>
> > Sorry to make this sound like a flame, but I try to
> > practice XP. The first time I ever come across a
> > function named "isSomething" that returns a conceptual
> > integer, it's going to get refactored, right then and
> > there to a name that actually expresses what it does.
>
> Laura was using the example given by Guido, since she was responding
to
> Guido's statements.
> I totally agree that a function name that implies a boolean should
return a
> boolean.  Guido was stating otherwise, that a boolean could also be
used as
> an int, which Laura felt (and apparently you and I agree) is wrong.

Agreed.

> If you
> had read the entire post you would have seen that that was (partially)
her
> point:
>
> [GvR]
> >     4) Should we strive to eliminate non-Boolean operations on bools
> >        in the future, through suitable warnings, so that e.g. True+1
> >        would eventually (e.g. in Python 3000 be illegal).
Personally,
> >        I think we shouldn't; 28+isleap(y) seems totally reasonable
to
> >        me.
>
> [Laura]
> > This is not an argument for allowing non-Boolean operations on
bools();
> > this is an argument for not writing functions that return Booleans.
Make
> > them return numbers instead, so that you can use them as you did.
Last
> > month we discussed why in 1712 February had 30 days in Sweden.
(See:
>
> So what you are saying fits perfectly with what Laura asserted.  If
you
> have a function numberOfLeapDays() why on earth would you need
> isLeapYear()?  The solution is simple: don't return a boolean when
another
> type can not only express true/false but also convey other information
as
> well.

You need both because the fact that the number of leap days
is either 0 or 1 is not exactly obvious unless you know the algorithm.
Using a number where you need to use a boolean is just as bad
as using a boolean where you need to use a number.

> I didn't necessarily agree with all of Laura's assertions regarding
the
> difficulty of understanding how booleans would fit into Python, but
she
> raised several compelling and thoughtful arguments.  I totally agree
that
> bools as return values of functions are less than useful.  Booleans
have
> the lowest information density of any type and thus would tend to
increase
> code verbosity if their use were to be encouraged.

Disagree. Clarity is more important than code density.

> Frankly Guido's dismissal of Laura's arguments as FUD was not only a
bit
> rude, but IMHO a bit short-sighted as well (given the fact that
someday,
> there may be a desire for a real boolean type).  Maybe Tim Peters has
> borrowed his time-machine and not returned it ;)

I agree, and he's already commented on that elsewhere. I would
not have classified it as FUD - I disagree with Laura's analysis, but
what she said wasn't FUD in any reasonable definition of the term,
it was simply wrong.

John Roth
>
>
> --
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
>
>




From pete at dexterslab.de  Tue Apr  9 10:45:08 2002
From: pete at dexterslab.de (Henning Peters)
Date: Tue, 9 Apr 2002 16:45:08 +0200
Subject: pointer
Message-ID: <a8uukg$fuu$04$1@news.t-online.com>

hi NG,

is there anything similar to the pointer concept from c/c++ in python?

i would like to store values in two different locations. both values should
always be equal. the one could be a pointer on the other. is this possible
to do this in python?

bye,
henning





From vfiack at webmails.com  Tue Apr  9 08:53:19 2002
From: vfiack at webmails.com (Vincent Fiack)
Date: Tue, 9 Apr 2002 13:53:19 +0100
Subject: Simple question! urgent
Message-ID: <20020409115319.4160.qmail@webmails.com>

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020409/776aa523/attachment.ksh>

From mmelchert at cgg.com  Thu Apr 18 11:16:58 2002
From: mmelchert at cgg.com (Michael Melchert)
Date: Thu, 18 Apr 2002 15:16:58 +0000
Subject: PyQt QTable
Message-ID: <a9mksg$275d$1@news6.isdnet.net>

Hello,

I am having some difficulties using QTable (PyQt3.0-2.3.1) 
in a python QDialog:
basically all I want to do is to connect the QTable.clicked() signal to
a python method [ self.connect(qtb,SIGNAL('clicked(....)'),self.the_method)]
my problem arises from the parameters which are defined for 
QTable.clicked(int,int,int,QPoint&).
what would be the correct way to write the connect() call, especially 
regarding the QPoint& reference? And how would the connected python method 
be defined?

cheers
Michael

-- 
--------------------------------------
Michael Melchert

E-Mail: mmelchert at cgg.com
--------------------------------------



From loewis at informatik.hu-berlin.de  Thu Apr 18 12:46:05 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 18:46:05 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <Pine.GSO.4.40.0204181648060.24683-100000@neelix>
References: <Pine.GSO.4.40.0204181648060.24683-100000@neelix>
Message-ID: <j4pu0x550y.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> cc -# -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so

Alright, so it does link SSL (and SSL is likely in a useful version, also).

> I can't tell if that worked:
> 
> >>> import socket
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/home/hgs/Python-2.2.1/Lib/socket.py", line 41, in ?
>     from _socket import *
> ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/build/lib.solaris-2.7-sun4u-2.2/_socket.so: symbol __eprintf: referenced symbol not found

That is the problem: The SSL libraries refer to a symbol __eprintf,
which originates from compiling the SSL libraries with GCC. GCC
defines this symbol in libgcc.a, which is implicitly linked always.

Now, when you link _socket.so with SunPRO, libgcc.a is not implicitly
linked, hence the error.


> How old is acceptable ( before I disclose the version number publicly!!
> :-) )

It turns out that this theory was only partially true; the SSL version
is fine, but the compiler the wrong one. You have the following
options (all alternative):

1. Build Python with GCC.
2. Rebuild OpenSSL with SunPRO.
3. Explicitly link libgcc.a to _socket.so, by editing Modules/Setup
4. Build without SSL support, by editing Modules/Setup.

HTH,
Martin




From paustin at eos.ubc.ca  Thu Apr 11 12:34:58 2002
From: paustin at eos.ubc.ca (phil)
Date: 11 Apr 2002 09:34:58 -0700
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it>
Message-ID: <m3k7rechxp.fsf@thrush.eos.ubc.ca>

Alex Martelli <aleax at aleax.it> writes:
> 
> I'll explain to you why lots of code does typetesting if you explain to me 
> lemmings' alleged propensity for hurling themselves off a cliff.  You go 
> first -- I suspect I can then just copy and paste your explanation right 
> back to you:-).
> 

It seems we have Walt Disney to blame for this:

http://www.snopes2.com/disney/films/lemmings.htm

Unlikely that typetesting has a similar origin, although there is at
least one python in the Disney version of the Jungle Book.

Regards, Phil




From guido at python.org  Thu Apr  4 09:37:14 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:37:14 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org> <mailman.1017893479.16809.python-list@python.org> <7xu1qsnk5j.fsf@ruckus.brouhaha.com>
Message-ID: <3CAC65FB.4175721A@python.org>

Paul Rubin wrote:
> Python already has a bunch of eq but not equal numeric types:
> 
>   1 == 1L      1 is not 1L
>   1 == 1.0     1 is not 1.0
>   1 == 1+0j    1 is not 1+0j

In addition, Python has always had

    (1==1) == 1  (1==1) is not 1

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


From Kevin.Smith at sas.com  Mon Apr  8 10:07:01 2002
From: Kevin.Smith at sas.com (Kevin Smith)
Date: 8 Apr 2002 14:07:01 GMT
Subject: Runtime mix-ins with new-style classes
Message-ID: <20020408100657387-0400@braeburn.themorgue.org>

Before I started using the new-style classes (i.e. 'object' based), I was
able to add mix-in base classes at runtime by modifying the __bases__
attribute.  However, I recently tried to do the same thing with a new
-style class and got an error telling me that __bases__ was a readonly
attribute.  Is there a was to do runtime mix-ins with new-style classes?
  
-- 
Kevin Smith
Kevin.Smith at sas.com


From dalke at dalkescientific.com  Wed Apr 17 12:37:05 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 17 Apr 2002 10:37:05 -0600
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org> <a9icab$phq$1@nntp9.atl.mindspring.net> <Xns91F32355A8FA9RASXnewsDFE1@130.133.1.4>
Message-ID: <a9k945$57k$1@slb5.atl.mindspring.net>

Philip Swartzleonard:
>In this case, wouldn't 3+7.0+x be a Typeerror or something? Int doesn't
>know how to add with a Strange, and Strange dosen't have an upgrading
>function...

Yeah, it would.  First though, 3+7.0 is a float.  To get the behaviour
I mentioned, add

    __radd__ = __add__

to the class definition.

> or do we assume addition is always commutitive and try '3+x+
>7.0' (effectively), therefore breaking the optimization mentioned by the
>other... =)

Python nor most other languages can assume addition of unknown types
is communitive.  After all, "a"+"b"+"c" != "a"+"c"+"b".

                    Andrew






From aleax at aleax.it  Tue Apr 30 13:13:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 17:13:36 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020176703.26327.python-list@python.org> <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com> <mailman.1020182455.6633.python-list@python.org>
Message-ID: <4hAz8.82281$vF6.2450915@news2.tin.it>

Roman Neuhauser wrote:
        ...
>     Given the high tongue-in-cheek ration of this list, I'm now really
>     confused as to whether Fredrik Lundh is a bot, or if you're "rocking
>     me", as the Czech saying has it.

Of course /F is a bot.  So am I, and so is Tim.  So far, we're the only
contributors to c.l.py to have been _found out_ to be bots.


Alex



From mkc+dated+1022285475.ae3858 at mathdogs.com  Wed Apr 24 20:12:58 2002
From: mkc+dated+1022285475.ae3858 at mathdogs.com (Mike Coleman)
Date: 24 Apr 2002 19:12:58 -0500
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6o7l$nbi$1@panix3.panix.com> <mailman.1019669144.3345.python-list@python.org>
Message-ID: <87662g7i0l.fsf@mathdogs.com>

holger krekel <pyth at devel.trillke.net> writes:
> - use module mmap to map source and destination files 
>   into addressspace.

At least for Python2.2 on Unix, I think mmap's are currently limited to
32-bits or so.  (Of course, if you're on a 32-bit box, this doesn't really
matter.)

So, for 80GB files, you must read.



From phd at phd.pp.ru  Tue Apr  9 14:29:52 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 22:29:52 +0400
Subject: pointer
In-Reply-To: <a8v36u$j3p$1@news.nsk.su>; from andrey@app.omsk.su on Tue, Apr 09, 2002 at 11:02:51PM +0700
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8v36u$j3p$1@news.nsk.su>
Message-ID: <20020409222952.G27632@phd.pp.ru>

On Tue, Apr 09, 2002 at 11:02:51PM +0700, Andrey Panchenko wrote:
> I think it's rather an ambiguous statement. I can think that after
> 
> a = 1
> b = a
> a = 3
> 
> b must hold 3.

   Why? You've rebinded a, but b still points to Integer(1).

> Andrey mailto:andrey at app.omsk.su

   Already subscribed to Russian-speaking Python and Zope User Group?

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




From mwh at python.net  Thu Apr 18 06:08:21 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 10:08:21 GMT
Subject: Readline/command history w/ interactive prompt
References: <mailman.1019061574.15837.python-list@python.org> <a9kcml$69k$1@news.wrc.xerox.com> <mailman.1019117616.20456.python-list@python.org>
Message-ID: <lky9fll3hn.fsf@pc150.maths.bris.ac.uk>

Andy Gimblett <gimbo at ftech.net> writes:

> On Wed, Apr 17, 2002 at 05:51:17PM +0000, Mark Jackson wrote:
> 
> > Try hitting control-r and typing a short string that last appeared
> > in the line desired ("my" or even "m" would work in this case).
> 
> _Nice_!  Thanks for that.  One question though: where's this feature
> documented?  :-)

"man bash", "man readline".  Not too hard...

Cheers,
M.

-- 
  A.D. 1517: Martin Luther nails his 95 Theses to the church door and
             is promptly moderated down to (-1, Flamebait).
        -- http://slashdot.org/comments.pl?sid=01/02/09/1815221&cid=52
                                        (although I've seen it before)


From jacek.generowicz at cern.ch  Sat Apr 27 14:33:16 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 27 Apr 2002 20:33:16 +0200
Subject: Python's Lisp heritage
References: <mailman.1019929248.17784.python-list@python.org>
Message-ID: <tyf662d0z6b.fsf@pcitapi22.cern.ch>

"Brian Quinlan" <brian at sweetapp.com> writes:

> Jacek wrote:
> > What? you object to my obtuse avoidance of Python's nifty infix
> > synonyms for these operators which cunningly call the functions
> > without recourse to parentheses?  In that case you surely allow me to
> > object to obtuse avoidance of such possibilities in Lisp.
> 
> The Lisp version is still more obtuse. Ask 100 people to evaluate the
> expressions "sin(0/2 + pi/2)" and "#i(sin(0/2 + pi/2))". I bet 90% of
> people won't know what sin is, 10% will be able to evaluate the Python
> version and 1% will be able to evaluate the Lisp version.

OK, I understand: I am participating in this thread to clear up some
misconceptions about Lisp that I've seen raised; you are participating
in order to perpetuate your own prejudices against Lisp.

Clearly we have nothing constructive to say to eachother on this
matter.

Bye.


From samschul at pacbell.net  Sun Apr  7 21:40:55 2002
From: samschul at pacbell.net (Samuel Schulenburg)
Date: 7 Apr 2002 18:40:55 -0700
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com> <Xns91E9B2CEE1835cliechtigmxnet@62.2.16.82>
Message-ID: <d7fcf8b4.0204071740.7edb1acc@posting.google.com>

Under windows the Pythonwin ide has built in windows support for
printing either the interactive windows, or any script file window.

Sam Schulenburg

Chris Liechti <cliechti at gmx.net> wrote in message news:<Xns91E9B2CEE1835cliechtigmxnet at 62.2.16.82>...
> jimw1956 at cs.comnojunk (Jim Williams) wrote in
> news:20020407103654.19127.00002640 at mb-fm.news.cs.com: 
> 
> > Hi all...I'm new to Python and really enjoying it so far, but I am
> > somewhat perplexed that as far as I have seen it has no facilities for
> > printing to a printer?  Surely this is possible. Someone please
> > enlighten me...I have done a good bit of research and searching and
> > can find no reference to how to do this. 
> 
> one neat solution is to use Reportlab to generate a PDF and you'll find an 
> Acrobat reader on all big platforms today.
> 
> you can also look at some GUI systems. e.g. wxPython has some printing 
> support etc. but when i need a printed output from a script, i generate a 
> PDF.
> 
> chris


From andreas at mtg.co.at  Mon Apr  8 16:35:56 2002
From: andreas at mtg.co.at (Andreas Kostyrka)
Date: Mon, 8 Apr 2002 22:35:56 +0200
Subject: Tkinter + import question
In-Reply-To: <3ca754d4$1_1@nopics.sjc>
References: <3ca754d4$1_1@nopics.sjc>
Message-ID: <20020408223556.6882603f.andreas@mtg.co.at>

On Sun, 31 Mar 2002 13:09:58 -0500
"Adonis Vargas" <deltapigz at telocity.com> wrote:

> i am writting a program utilizing Tk as my gui engine, and i have broken
> down sections of the form into modules and loaded them through a main
> script. my only question is in every module i am forced to import the
> Tkinter module, and does this pose consume too much memory? or does python
Nope.
Importing in Python consists of two "steps":
1.) load the .pyc or .so (.dll) file.
2.) add it to the namespace of the modules that does the import.

The first step is done only once, so later imports are quite fast and do not consume memory (well they do for the dictionary entry in the second module, but this is negliable.)

Actually one can get the effect you want to have like this: (DO NOT USE THIS UNLESS YOU KNOW WHAT YOU DO)
import __builtins__
import Tkinter
__builtins__.Tkinter=Tkinter

This makes Tkinter a "builtin" symbol that is visible in all modules.

Andreas




From skip at pobox.com  Thu Apr  4 16:50:55 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 4 Apr 2002 15:50:55 -0600
Subject: ANN: Pyrex - a language for writing Python extension modules
In-Reply-To: <mailman.1017952983.647.clpa-moderators@python.org>
References: <mailman.1017952983.647.clpa-moderators@python.org>
Message-ID: <15532.51903.323557.147303@12-248-41-177.client.attbi.com>

    Greg> Pyrex is more or less Python with C data types added. You can
    Greg> write functions which freely intermix operations on Python and C
    Greg> data, with all Python reference counting and error checking
    Greg> completely automated.

Interesting, but why not simply attach the type information to function
attributes so the Pyrex code really is just Python?  The Pyrex compiler can
fish the type information out of function attributes as readily as from the
program source code.  

The advantages?  I can think of a couple right off the bat:

    You can defer Pyrex compilation until you realize you need it.  Naively
    compiling all your code to C makes for a lot of (object) code bloat and
    doesn't necessarily speed anything up (if, for example, the code is I/O
    bound or not executed frequently enough).  This was one of the problems
    with Python2C.

    You could train PyChecker about the data type information, which it
    could use to produce more (and more useful) warnings about suspect
    Python usage.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From python at rcn.com  Tue Apr  2 09:16:01 2002
From: python at rcn.com (Raymond Hettinger)
Date: Tue, 2 Apr 2002 09:16:01 -0500
Subject: simulated annealing
References: <3CAA1FDC.7FEF351D@gol.ge>
Message-ID: <a8cega$410$1@bob.news.rcn.net>

> Has somebody written a free module for simulated annealing?
> I have an experiense of making this, just I'd like to avoid inventing a
> weel:(

A google search on 'simulated annealing python' turns up a few links:
http://www.google.com/search?sourceid=navclient&q=simulated+annealing+python


Raymond Hettinger




From JamesL at Lugoj.com  Sat Apr  6 14:33:31 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 19:33:31 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com>
Message-ID: <Xns91E87646CA7FCJamesLLugojcom@216.148.53.83>

"John Roth" <johnroth at ameritech.net> wrote:
> The original example was something like
> 
> number of days in year = 365 + isLeap(year).
> 
> To put it bluntly, that's a very confusing hack that depends
> on conceptual True being 1 and conceptual False being
> zero.

Just FYI, that example is not quite theoretical since you can find almost 
equivalent usage in the standard Lib/calendar.py module:

# Return weekday (0-6 ~ Mon-Sun) and number of days (28-31) for year, month
def monthrange(year, month):
    if not 1 <= month <= 12: raise ValueError, 'bad month number'
    day1 = weekday(year, month, 1)
    ndays = mdays[month] + (month == February and isleap(year))
    return day1, ndays

Of course in this case it is the result of the "and" that is being treated 
as an integer.


From vinay_sajip at yahoo.co.uk  Fri Apr 26 12:30:33 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 26 Apr 2002 09:30:33 -0700
Subject: AS/400
References: <1e140f34.0204250934.2fee1f26@posting.google.com>
Message-ID: <2e37dc1.0204260830.546ae83@posting.google.com>

I agree with all the other posts - and if screen scraping is what you
need to do, fair enough. ODBC has already been mentioned, but it's
sometimes a bit of a sledgehammer to crack a nut. I've had good
experience with an alternative approach (similar to JTOpen) - I've
used a product called DataGate/400, from ASNA - http://www.asna.com/ -
which is low cost and provides fine-grained read/write access to
AS/400 resources. It does require software to be installed on the
AS/400, but it's low cost and quite effective. I wrote my own COM
wrapper around it for use under Windows, but I believe their product
range now includes a COM component.

Best of luck,


Vinay Sajip


From joonas at olen.to  Sat Apr 13 16:57:57 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Sat, 13 Apr 2002 23:57:57 +0300
Subject: FileUpload via CGI by Python
References: <slrnabgabq.gjo.jarvin24@tutor.cc.tut.fi>
Message-ID: <3CB89BD5.573A8B3B@olen.to>

J?rvinen Petri wrote:
> 
> Hey, i'm trying to do form that uploads file to server.
> 
> Is there any ready solutions?

Use cgi module to do the parsing on server side. Remember to use correct
enctype in your html code, like below.

<form action="upload.py" method="POST" enctype="multipart/form-data">


From emile at fenx.com  Sat Apr 27 17:22:03 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 27 Apr 2002 21:22:03 GMT
Subject: inline assignments in conditionals
References: <mailman.1019941549.3704.python-list@python.org>
Message-ID: <%DEy8.66450$CH.10901@rwcrnsc52.ops.asp.att.net>

"jaf" 
> Aggh!  Of course I meant it the other way around.  :^P
> It would have been a nice example too, if I hadn't completely
> stuffed it up.  So for the record, "If today is Sunday" is my
> preferred form (i.e. if day == 0), and I should never again
> be allowed to post without proofreading first.
> 

No.  That'd take a lot of the fun out of it!  ;-)


--

Emile van Sebille
emile at fenx.com

---------



From tejarex at yahoo.com  Wed Apr  3 11:00:42 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 03 Apr 2002 16:00:42 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <3CAAC1FB.2050105@mxm.dk>
Message-ID: <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com>

"Max M" <maxm at mxm.dk> wrote in message news:3CAAC1FB.2050105 at mxm.dk...
> Through Zope I have used Structured text extensively. Also I have
good
> experience teaching stuctured text to my customers. After a very
short
> while they start to use it insted of html when generating content
for
> their sites due to ease of use.

Just curious: are you willing to (consider, at least) switching to
ReST, or are we going to end up with two similar standards used by
different parts of the Python community?  IE, do Goodger's arguments
about why the differences from ST are improvements cut any ice with
you?

Terry






From logiplexsoftware at earthlink.net  Tue Apr  2 17:12:19 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 2 Apr 2002 14:12:19 -0800
Subject: Why do Windows sockets break after exactly 4 minutes?
In-Reply-To: <3CAA28FF.9E94F7F5@sage.att.com>
References: <mailman.1016754947.17945.python-list@python.org>
	<3CAA28FF.9E94F7F5@sage.att.com>
Message-ID: <20020402141219.19a121f7.logiplexsoftware@earthlink.net>

On Tue, 2 Apr 2002 21:56:15 GMT
Garry Hodgson wrote:

> Geoffrey Talvola wrote:
> > 
> > I've noticed that Windows sockets seem to close all by themselves if
there
> > is no activity for 4 minutes.
> 
> your windows machine stays up that long?  wow!

You obviously have your Windows machine misconfigured.  Mine stays up at
least twice that long (although the first 4 minutes is taken while Windows
optimizes memory usage by filling all available memory with registry
entries).

A good Windows optimization trick is to insert this line at the beginning
of autoexec.bat:

c:\loadlin c:\vmlinuz root=/dev/hda3 ro 

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From jepler at unpythonic.net  Tue Apr 30 12:18:28 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 30 Apr 2002 11:18:28 -0500
Subject: Third Draft: Pep for Deprecating Builtins
In-Reply-To: <CEsz8.80366$vF6.2381303@news2.tin.it>
References: <mailman.1020129837.31112.python-list@python.org> <CEsz8.80366$vF6.2381303@news2.tin.it>
Message-ID: <20020430111828.C17488@unpythonic.net>

On Tue, Apr 30, 2002 at 08:32:34AM +0000, Alex Martelli wrote:
> So, types.ClassType COULD stay around "forever" from my POV, even long after
> the default metaclass has become type -- at worst, the way to enable the
> alternative metaclass as the default one would be 2 lines at module scope:
> 
> import types
> __metaclass__ = types.ClassType

Why not
    from __past__ import old_style_classes
this would extend well to
    from __past__ import integer_division
    from __past__ import no_yield_token
> 
We could have one version warn about old code, and permit new behavior
by importing from the __future__.  Then, in the next version, code could
still get the old behavior by 'from __past__ import'.  Finally, in
Python 3.0, the old behavior could be removed altogether, like the
vestigal appendix or a tomcat's testicles.

Jeff




From ods at fep.ru  Tue Apr  9 10:08:12 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Tue, 9 Apr 2002 18:08:12 +0400 (MSD)
Subject: Problem with packages and modules
In-Reply-To: <117880a1.0204090542.258f0f51@posting.google.com>
Message-ID: <Pine.LNX.4.44L.0204091805020.13708-100000@ods.pp.ru>

On 9 Apr 2002, [ISO-8859-1] J?rgen Hansen wrote:

JH> Hello,
JH>
JH> I have been working for some time now with Python and really
JH> enjoy it.
JH> Right now I am having serious difficulties crasping several
JH> issues
JH> when using packages.
JH>
JH>
JH> This is an example of directory hierachy:
JH>
JH> company/                   Top-level package
JH>     __init__.py            Initialize the company package
JH>     plot/                  A plot package
JH>         __init__.py
JH>         axes.py            Modules in the plot package
JH>         axis.py
JH>         bar.py
JH>         figure.py
JH>         line.py
JH>         plotobject.py
JH>         points.py
JH>         tick.py
JH>         utils.py
JH>     log.py                 Modules in the company package
JH>     utils.py
JH>
JH>
JH> If I were to use the plot package I would type in
JH>    >>> from company import plot
JH>
JH> and assume that I could access modules as:
JH>    >>> plot.figure.Figure()    # where Figure is a class
JH> name.
JH>
JH> but this is not the case. Neither of the modules in the plot
JH> package
JH> are available, and it doesn't matter if I use the __all__
JH> attribute in
JH> __init__.py and type in all the module names.

Certanly, module figure is not imported. You should write "import
figure" in __init__.py or use "from company.plot import figure"
(or "from company.plot import *" if you know what you are doing).

JH> Further if I typed in:
JH>    >>> import company
JH>
JH> I would assume that I could access all subpackages and their
JH> modules
JH> with a dotted name (company.plot.figure.Figure() for
JH> instance), but
JH> this is obviously not the case.

Similar, write "import plot" in company/__init__.py or import
modules directly.





From gcordova at hebmex.com  Wed Apr  3 16:08:14 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 3 Apr 2002 15:08:14 -0600 
Subject: Threading problem
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB54@mail.hebmex.com>

> 
> The Python global interpreter lock requires that even with native
> threading, only one thread is actually interpreting Python code at any
> moment in time.  The net result of this is that while threading is
> excellent for overlapping I/O, it won't efficiently distribute CPU
> amongst processors.  That's because if you are CPU bound, then you are
> stuck in Python code execution, and only one processor can be doing
> that at any given moment in time.
> 

This is evil :-(

>
> This can be improved somewhat if you use extensions that release the
> GIL, but in most cases that's done during I/O and not processing time,
> particularly if the latter will require access to Python structures.
> But in theory a purely CPU bound extension could release the GIL while
> performing computations that it knew didn't require any interface to
> Python.
> 

I suspected this.

So, a solution would be to make an extension that...

1. Copies all pertinent data from python-space into it's own memory.

2. Releases the GIL.

3. Does all the work that needs to be done only on it's private memory.

4. Recapture the GIL.

5. Copy back results into python-space.

Something like that?

-gustavo






From martin at v.loewis.de  Mon Apr  1 10:57:45 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 17:57:45 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de> <KM_p8.41075$pT1.1129125@news1.tin.it>
Message-ID: <m3g02f76o6.fsf@mira.informatik.hu-berlin.de>

Alex Martelli <aleax at aleax.it> writes:

> > Programmers that are used to booleans from previous
> > languages might well find their lifes easier. Even people that you
> 
> In the same sense as they might well find their lifes easier if suddenly
> they had to start decorating their Python programs with deuced braces
> or BEGIN/END to indicate block structure -- they might find the
> familiarity comforting.  

Nah, that is hardly the same thing. With the PEP, nobody *has* to do
anything; existing programs will continue to work just fine (with very
very few exceptions). If people want to return 0 when they mean False
- fine.

> I think teaching this as alternate ways of *spelling* equal (but distinct: 
> ==, and essentially interchangeable in use, but NOT 'is') values will be
> the least of evils.  

I disagree. Somebody could use the language (with the change) and
believe booleans are completely unrelated to numbers, and would rarely
find this view incorrect - at which time you'll need to explain that
booleans interact with numbers in a certain way.

> Exactly the same set of objects (plus the new object False, that's ==0
> anyway and can be used just about wherever 0 can be) 

While it is true that False can be used everywhere, it doesn't mean it
should be. Use booleans if you want to express the value of a
predicate; use integers if you want to count things.

> The "by convention" status of the canonical spelling for untrue
> still holds, except that you want to change the convention to the
> new object which exists solely for the value of changing that
> canonical spelling, more or less.

No. If the goal had been to change the canonical spelling, two new
builtins would have been sufficient. The goal was to create a new type
for those, in addition to the new spelling.

> 'if x == y:' should cause no warning, of course -- it's a general, perfectly
> good polymorphic usage.  Why should the user have to silence warnings
> or break the smoothness of polymorphism to ensure against x, or y, or
> both being of this newfangled bool type?

Indeed. I wonder how often the warning would trigger, though. I'd
expect that people very rarely compare the result the got from a
predicate to something else.

> > More realistically, the *parser* should emit a warning if it sees the
> > name True used in such a context.
> 
> ...assuming the parser can determine at parse time that True stands
> for the built-in value at that point -- or do you want to give True an
> exhalted semi-keyword status, beyond that of, say, None, a dignity that 
> will produce frequent warnings if you rebind that identifier e.g. in a local
> namespace?  

I would indeed have the warning produced for all occurrences of True -
perhaps a different warning if the parser can tell that True is bound
at module scope. This is essentially what pychecker does in many
similar cases, and Guido is in favour of moving some of the PyChecker
warnings into the interpreter.

There would be an option to turn off parser warnings, of course.

> I know 'slippery slope' arguments are disliked around here, but both
> this exchange and the various voices raised in clamor towards
> 'stronger, purer booleans' suggest to me that PEP 285's own slope
> may be slippery indeed.

These are different issues, though: those voices want to ban
interactions between booleans and numbers; I want to ban binding of
names that are also builtins (starting with None, True, and False).

> No, but it's one more warning-case you have to propose (and how else
> will one check for the wonderful new objects' identities if 'is True' 
> warns?! by twists and turns of isinstance checking for bool then int()
> and == checking for 1...?!).  And 'is True' is very newbie-alluring: a
> true 'attractive nuisance' (in more than one meaning of the phrase).

Ok, so the problem is more involved. I would suggest to produce the
warning for 'is True', but not for 'True is' ... (I'm not sure whether
I mean this seriously :-)

> > That's a good point: booleans should round-trip when pickling. I guess
> > it is a bug in the PEP that this hasn't been spelled out.
> 
> If they do round-trip, then, it seems to me, we get a 'gratuitous'
> incompatibility of pickles produced by 2.3 wrt 2.2 programs trying to
> read them -- one more annoyance (and not a very attractive one either,
> this time)

I'm not sure how big this problem is: it should be easy to add
bool-loading function to Python installations for use with Python
before 2.3. Packages that need to support cross-version pickling could
incorporate such a feature.

> And of course old pickles will give 0 and 1 so that much work will
> be needed to reconstruct them into False and True when appropriate
> *IF* it matters at all if something is True or 1.

This is a transitional issue. It may matter, but it may not matter
enough to take action - depending on the application. An application
that worries about cross-versioning pickling (presumable across
versions of its own software as well) will have to take efforts
already; avoiding to dump booleans will be a minor complication only.

> So _why_ pay the price -- not a huge, world-racking one, but still
> one nevertheless?

I could repeat the rationale of the PEP here, but I guess the question
was rhetoric.

Regards,
Martin



From python at rcn.com  Mon Apr 29 03:52:40 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 03:52:40 -0400
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
Message-ID: <aaiu6e$pfr$1@bob.news.rcn.net>

Please send me comments on a draft PEP deprecating
and relocating six built-in functions.  The text file is at:
http://users.rcn.com/python/download/deppep.txt

This is version two and incorporates the comments received
from David, Alex, Brian, Martin, and Chris.

It is also listed below.


Raymond Hettinger

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

PEP: XXX
Title: Deprecate Six Built-ins
Version: $Revision: 0.2 $
Last-Modified: $Date: 2002/04/29 19:42:56 $
Author: python at rcn.com (Raymond D. Hettinger)
Status: Draft
Type: Standards Track
Created: 28-Apr-2002
Python-Version: 2.3
Post-History:


Abstract

    This PEP proposes to compact the list of built-in functions through
    relocation and deprecation.  map(), reduce() and filter() would move to
a
    separate module for functionals.  pow() and divmod() would move to into
    the operator module.  input() would be eliminated altogether.



Rationale

    There are separate motivations for each function.  One motivation in
    common is that the number of total built-ins should be as small as
    possible to limit the amount of "core language" one needs to know in
order
    to read code that is not module specific.

    The functionals (map, filter, and reduce) had greater importance prior
to
    the introduction of list comprehensions which are now the preferred (and
    more readable) approach.  Reduce is easily (and more clearly) replaced
by
    an equivalent for loop.  Limits on the number of built-in functions
    prevented the addition of other functionals which may have be useful
when
    that style of programming is needed.  Creating a separate module allows
    room for a large variety of functionals (e.g. fold, head, tail, take,
    drop, etc.) to be introduced without cluttering the built-in namespace.

    Pow() and divmod() are more related to operator module functions than
any
    of the built-in functions.  In general, they are rarely used.  Because
    the ** operator is available, the pow() function is rarely called
    directly (unless the optional third argument is needed).  Another reason
    to move pow() is that is conflicts with the identically named, but
    semantically different function in the math module.  As a builtin, pow()
    is in a vulnerable position for getting clobberred by a
    "from math import *".

    Input() is easily replaced by eval(raw_input()).  It has few legitimate
    uses and presents a security risk in most programs where it is used.
    Though the risks are well documented, experience on python-tutor and
    help at python.org indicates that new users frequently use input() when
    raw_input() was intended and that they are unaware of the embedded
eval().



Key Issue

    All of these functions have been in-place since the beginning, are used
    in mountains of code, and appear in every Python book.  The situation
    would appear to be immovable.

    I propose that a simple means be provided to re-enable those functions
at
    will when they are needed to run old code -- Something equivalent to:

        from functionals import map, filter, reduce
        from math import pow, divmod
        from deprecated import input



Implementation and Deprecation Timetable

    The module additions would go into Python 2.3.  All six functions would
be
    left intact and raise deprecation warnings in Python 2.4 and 2.5.  The
six
    functions would be removed in Python 2.6 (having allowed three versions
    and over a year for code updates).  The restore builtins mechanism would
    go into Python 2.4 and remain available in perpetuity for other
    deprecations, allowing the language to evolve without accumulating
    clutter.


Discussion

    Q.  Why wasn't zip() included.
    A.  It doesn't take and apply a function argument.  Also, it is used in
        non-functional programming to neatly code lockstep iteration.

    Q.  Why wasn't apply() or oct() included?
    A.  I picked six candidates for relocation or elimination based on
        whether I could make a case for them.  No doubt there will be
others.



Copyright

    This document has been placed in the public domain.



Local Variables:
mode: indented-text
indent-tabs-mode: nil
fill-column: 70
End:








From kseehof at neuralintegrator.com  Sat Apr  6 15:32:18 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 6 Apr 2002 12:32:18 -0800
Subject: PEP 285: Adding a bool type
In-Reply-To: <a8nhfi$gbt$1@nyheter.chalmers.se>
Message-ID: <MEENLMFCHGAOOICLMCHHOECPDLAA.kseehof@neuralintegrator.com>

> 1) Should this PEP be accepted at all.
>
> No.
>
> I come from a background where I have taught students hands-on programming
> in the saltmines of my local university, taught engineers in the industry
> how to program and where I have implemented actual programs for
> various uses,
> from the humble MCU that controls a valve, to systems that manage the
> collected literature of all academia in Sweden. I have taught using about
> a dozen different programming languages and last I counted, I had written
> programs in about 30 languages, including some assembler dialects.
>
> My humble opinion is that new features in a language should make it easier
> to accomplish what you set out to do - create a working program.
>
> Python is a language that is weakly typed, essentially allowing a
> statement
> like
>
> x = f(y)
>
> or
>
> x = a + b
>
> to assign a value of any type to x. This value may be an int, a float,
> a string, a list or a dictionary (and an imaginary).
>
> If we ignore the imaginary, which has a very specialized scope, we have 5
> base types. These can interact in approximately 5! (120)
> different ways in a
> statement. If we add one more base type we get 6! (720) different ways
> that our types can interact.

In python the number of numeric interactions is essentially 3!, not 6!.
Most of these interactions are trivial and predictable, and are not
increased
by bool.  My brain hurts trying to figure out how to enumerate the 6! that
you suggest :-)  Perhaps in other languages your argument has merit.

One thing that I agree with is that one would not be likely to add a bool to
a complex number.  :-)

> For a student as well as an active implementor, it is necessary to learn
> what the results of all (or at least most of) these interactions are.
> Add one more base type and you have increased the complexity of the type
> system by a large factor.
>
> As a contrast, teaching how the current type system works is a matter of
> a 15 minute lecture, that no student will misunderstand. The hard thing to
> teach is how to interpret a "Type mismatch error" and why unexpected
> automatic conversions happen. Automatic conversions are a pitfall even for
> very seasoned programmers.

Properly used bools will tend not to interact with other numeric types in
practice.  Anyway, a TypeError won't occur.

> Closely related types also make things more difficult than highly
> ortogonal
> ones. Booleans, like ints and floats are scalar, so relations between
> these 3 will be more important than say, the relationship between
> an int and
> a dictionary.
>
> If we take the example of C++, one of the reasons it is really
> hard to teach
> and to use is the very close relationship between references and pointers.
> Of course it is very nice to be able to take a reference, cast it into a
> pointer, subtract it from another pointer and get an int, or a long, or an
> unsigned long, or whatever. (Yes, multiple length, closely related types
> add a lot of complexity. (And I didn't even mention unsigned ;-)))

Sure, but we are talking about python here.  I agree that C++ suffers
greatly from the proliferation of numeric types.

> You can learn to master complexity as you become proficient in a
> language and
> you will select a subset of its features that you know and understand.
> People who write C code and have never used a function pointer or
> an enum or
> a const declaration are in vast majority, even though these are all great
> tools in the right context. C is only moderately more complex than Python.
>
>
> So, is a boolean always a waste of time?
>
> No, I don't think so. In a statically typed language, they actually catch
> some errors at compile time. Python is dynamically typed.
>
> In a language for embedded applications, you can save execution time and
> memory space by using a single bit for your boolean variable. Python
> is good for many things, but it is not optimised for embedded
> applications.
>
> It does add readability to some code to use the literals 'true'
> and 'false'.
> Python contains ample constructs for the user to do this. You can even add
> a module in the standard library called "bool", exporting "true"
> and "false".

The net effect of PEP 285 is very similar to that suggestion.  Except that
the PEP has several advantages, such as having a __repr__ that returns
"true" and "false".

> Guido thinks this is a small change to the language. I strongly disagree.
> My conclusion is that the addition of a boolean type to Python would
> hit it where it hurts most. Python is a simple language, because it has
> a very small and simple type system. Making it harder to learn by
> increasing
> the complexity is going to drive Pythons core supporters - the biologists,
> physicists, astronomers, sociologists and other scientists - to look for
> other alternatives, that require less work to understand.
>
> Jacob Hall?n
> AB Strakt
>
> (Yes, I know that booleans are implemented in a cheat way
> according to the PEP,
>  making them almost aliases for int 0 and int 1. I does not ease
> the burden
>  of knowing the type interaction information.)

Yes it does.  There isn't any new type interaction information associated
with adding bool (except the information that there isn't any new type
interaction information associated with adding bool :-)

- Ken Seehof





From gulliver at atr.co.jp  Wed Apr  3 16:59:04 2002
From: gulliver at atr.co.jp (Roberto Lopez-Gulliver)
Date: 3 Apr 2002 13:59:04 -0800
Subject: How to wait only so long for a server to respond?
References: <mailman.1017800895.19820.python-list@python.org> <n1Dq8.12893$ml2.1014180@newsread1.prod.itd.earthlink.net>
Message-ID: <26b7aa41.0204031359.2171591d@posting.google.com>

go 'google' for "timeoutsocket.py".

-r


From fperez528 at yahoo.com  Wed Apr 17 01:50:40 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 16 Apr 2002 23:50:40 -0600
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu> <861ydg3ff3.fsf@haggis.aae.uiuc.edu> <3CBC5686.BCBCF663@noaa.gov> <a9holn$5t1$2@peabody.colorado.edu> <86g01vrxcr.fsf@haggis.aae.uiuc.edu>
Message-ID: <a9j2hd$gr$1@peabody.colorado.edu>

D. Michael McFarland wrote:

> Fernando P?rez <fperez528 at yahoo.com> writes:
> 
>> This is another one for which weave can be a godsend. The blitz
>> module in weave can for many numeric expressions generate equivalent
>> C++ code which avoids all the intermediate copies. I don't want to
>> sound like a cheerleader (I'm not a developer of weave, only a
>> user), but weave is really amazing in how easily it allows you to
>> integrate Numeric pyhton code with C snippets.
> 
> I played with weave 0.2.3 shortly after that version was released, but
> never got weave.test() to run cleanly.  I said to myself, "Something
> here has great promise," and went back to my old ways of doing
> things.  Should I try the CVS version?

Yes. In fact, the code I'm referring you to here requires the CVS code to work 
(though it could be made to work with the old code). Scipy is shaping up very 
nicely, and the core developers are very close to an updated release in the 
next few weeks.

> 
> Re cheerleading, well, if the student in question hadn't listened to
> my proselytizing about Python, he wouldn't have quite his current
> problem.  (He'd have different and worse problems, but that's another
> post.)
> 
> I've snipped away your mention of example files, but by all means,
> please send me a few.

I've posted it at http://www-hep.colorado.edu/~fperez/python_misc/

Grab the weave_examples.py, and you may also look at the python-c.pdf file 
which has another python/NumPy/weave comparison of the same problem (with a 
factor of 25 speedup between weave and python!).

Let me know if this helps.

Best,

f.


From andy47 at halfcooked.com  Mon Apr  1 23:03:12 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Tue, 2 Apr 2002 04:03:12 +0000 (UTC)
Subject: selecting files from MySQL
References: <a87o30$qabnh$1@fu-berlin.de>
Message-ID: <Xns91E48EC445537andy47halfcookedcom@203.109.252.31>

"amy" <amylang at zedat.fu-berlin.de> wrote in news:a87o30$qabnh$1 at fu-
berlin.de:

> hello python-knowing,
> can somebody help me to solve a little problem in a script, sc selecting
> files from a mysql database?
> so far, it doesn't want to work:
> 
> 
>     if fileid['translation'] == 'none':
>        actie = suchen fileid == 'none'
>     elif fileid['translation'] == 'de-orig':
>        actie = suchen fileid == 'de-orig'
>     elif fileid['translation'] == 'nl-orig':
>        actie = suchen fileid == 'nl-orig'
>     elif fileid['translation'] == 'no-orig':
>        actie = suchen fileid == 'no-orig'
>     else:
>        search = opt['such']
> 
> 
> thanks a lot, if it's possible
> amy
> 
> 

Amy,

This isn't C, you can either do a comparison OR an assigment on each line, 
not both.

Thus, all of your lines like;
    actie = suchen fileid == 'no-orig'

are invalid syntax. Are you trying to assign a value to 'actie' and to 
'fileid', or are you trying to assign a value to 'actie' if 'fileid' has 
another specific value in it? In Python '=' is assignment (well, reference 
but lets not go there just yet) and '==' is a comparison operator. 

Perhaps it would be best if you said what you want to do and then ask for 
examples of how to do it. 

Also, this code fragment doesn't appear to have anything to do with MySQL 
but your subject says you are trying to select files from MySQL? Is that 
important here or do you just need to get this particular section of your 
code working first?

Finally, a good tip for posts to the newsgroup is to try out your code in 
the Python interpreter (or idle, or Pythonwin) and include in your post any 
and all error messages or exceptions you get when you try to run it.

HTH,
Andy
-- 
Contents free posts a speciality


From noah at noah.org  Tue Apr 16 12:47:25 2002
From: noah at noah.org (Noah)
Date: Tue, 16 Apr 2002 09:47:25 -0700
Subject: Backing up mail in Outlook Express using Python
In-Reply-To: <jeWu8.104$1kc.179321856@news.telia.no>
Message-ID: <PAENINBOOODFDEPOKPKCOEHJCLAA.noah@noah.org>

What?! My version of Outlook Express stores them in binary files.
What version are you using?

Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Thomas Weholt
Sent: Tuesday, April 16, 2002 7:07 AM
To: python-list at python.org
Subject: Re: Backing up mail in Outlook Express using Python


Well, by searching the net a bit more I've found some pointers and as soon
as I found out where the mail was actually stored and that it's stored in
plain files, backing them up was easy.

Thanks anyway :-)

Best regards,
Thomas Weholt


"Steve Holden" <sholden at holdenweb.com> wrote in message
news:JCUu8.23257$Ji7.8640 at atlpnn01.usenetserver.com...
> "Thomas Weholt" <thomas at gatsoft.no> wrote in message
> news:cGSu8.101$1kc.179320832 at news.telia.no...
> > I need to backup all my mail stored in Outlook Express. Can I use Python
> to
> > do this?
> >
> No reason why not, though I'm not sure why you'd want to, since the
message
> are just stored in files.
>
> You can find out where they are from the Maintenance tab on Tools |
Options
> (click the "Store Folder" button).
>
> I did adapt some code from Delphi to read the mailbox files, and can let
you
> have that if it would help, but sounds to me like a file copy is all
you're
> looking for.
>
> regards
>  Steve
>
>
>
>







From tejarex at yahoo.com  Sun Apr 14 04:18:05 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sun, 14 Apr 2002 08:18:05 GMT
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <3cb8c690$0$19942$9b622d9e@news.freenet.de>
Message-ID: <1Xau8.32080$K5.3134945@bin5.nnrp.aus1.giganews.com>

"Jens Baader" <nospam at nospam.com> wrote in message
news:3cb8c690$0$19942$9b622d9e at news.freenet.de...
>
> "Terry Reedy" <tejarex at yahoo.com> schrieb im Newsbeitrag
> news:WL2u8.26958$GS6.2775752 at bin3.nnrp.aus1.giganews.com...
>
> I see. Thank you for your answers.
>
> BTW, try this in interactive mode:
>
> "The Python interpreter is very stable"*30000
>
> At least at my system the interpreter promptly
> hangs itself. I guess I just found my first bug..
> (not bad I just read the first few pages of the
> tutorial)

Perhaps you should write an iterative fibonacci function in standard C
or C++ and try fib(100) or whatever n is needed to overflow ints on
that machine.  Or write a simple recursive function and try
recurse(30000).  If it works with one compiler on one OS and system,
try another.  Then write to the compiler maker and C standards
committee about the bugs you are finding after barely learning C.  I
am sure Microsoft will be glad to hear from you.

Terry J. Reedy





From stian.jenssen at nrk.no  Wed Apr 24 09:00:22 2002
From: stian.jenssen at nrk.no (Stian Jenssen)
Date: Wed, 24 Apr 2002 15:00:22 +0200
Subject: Window problem in windows!
Message-ID: <D186E3C450FED2119AFA00508B08736F040D9974@maexch05.nrk.no>

Hi all!

When I use 

	---code---

    command = "pause \r\n"
    command += "pause"
    print os.system(command)

	----code--


It pops up 2 window instead of 1, in a windows enviroment. The code works
greate on *nix box, what seems to be the problem.



Stian






From tim.one at comcast.net  Mon Apr  1 18:42:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 01 Apr 2002 18:42:49 -0500
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
In-Reply-To: <slrnaahnu8.nsu.kzaragoza@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCEENAOJAA.tim.one@comcast.net>

[Kris J. Zaragoza, resisting digital catharsis]
> Gotta love those digital seances. :) Thanks, Tim.  This does make me
> feel a bit better about what direction this is taking.

If you want to feel a lot better, read the most recent revision of the PEP:
I see Guido added a "Clarification" section near the end, in which he says
"Kris Zaragoza's concerns are paranoid foolishness, not worth my precious
time to expose for the baseless hysteria they most obviously are".  Or
something like that -- I'm not actually sure he named you specifically.

> ...
> I have read the PEP, but based on that alone, I don't see any reasons
> that are terribly compelling.  Perhaps it's just that I'm an
> experienced developer or that I have a CS degree.  I just don't find
> the basic concepts as they are that hard to grasp.  That's why I keep
> wondering how much aid this will really be to a new or inexperienced
> programmer.

That's why I asked whether you had read the PEP.  The Rationale section
contains 6 distinct paragraphs, only one of which mentions (indirectly)
new/inexperienced programmers.  Will they, as claimed, require a millisecond
less thought each time they see

>>> 2-1  # doesn't change
1
>>> 2<1  # changes to False
0
>>>

?  Probably, but I don't really care -- nobody tried to make computers easy
for me, and suffering builds character <wink>.  I've been at this for
decades now, and I'll need a millisecond less thought.

To the extent I like the PEP, I'm mostly in the second paragraph:  there's
always some uncertainty when reading a function whether a "return 1" is
intended as a true/false result, or as an integer result.  I don't care that
the PEP leaves them much the same under the covers, I care that it allows
clearer expression of *intent*.  It's the same reason hundreds of C projects
#define their way into the same almost-inconsequential (under the covers)
distinction.  In the heat of battle, I'm not going to bother writing

    return 1  # true

but I will bother writing

    return True

because my editor has extensive word-completion facilities <wink>.




From alexandrecourbot at linuxgames.com  Sun Apr 14 05:33:48 2002
From: alexandrecourbot at linuxgames.com (Alexandre Courbot)
Date: 14 Apr 2002 11:33:48 +0200
Subject: Getting a Python library for cross-compiling with mingw32
In-Reply-To: <20020413192547.GA5477@lilith.my-fqdn.de>
References: <1018706814.23939.26.camel@progeny> 
	<20020413192547.GA5477@lilith.my-fqdn.de>
Message-ID: <1018776833.807.1.camel@progeny>

> > [...] I then attempted to cross-compile Python itself. It looks like
> > it's not intented for that :) After disabling configure tests that
> > didn't like cross compilers, I had to twiddle even more to get a
> > binary.  Finally, I managed to get a binary and a library, that even
> > worked under Wine!
> 
> Wow! Cool. But you know that the resulting binary isn't necessarily
> binary compatible with the Standard win32 Python distribution, do you?

Mmm, not sure what you mean. You mean my binary might not run with the
official release modules and so on? Actually my goal is just to embed
Python in my app (statically if possible) that would come with the
mandatory Python modules and its own wrappers, and set the import path
appropriately so the user doesn't have to install the whole Python stuff
just to play. So it'd just need to be compatible with itself, plus maybe
the official Python DLL if I can't link statically (don't know why, but
despite of my linker options and the fact I had the .a the binaries
always required the DLL of the libs I used so far (SDL and z))

> > [...] Please note that I largely prefer having a version >= 2.2.
> > Anyone knows of a place which provides suitable binaries, or any
> > Python hacker could help me properly compiling my binary?
> 
> Next time I reboot to Windows, I can send you my gcc import library for
> the standard win32 Python. It's easy to create with Cygwin, btw. I'd
> have to check if this is suitable for cross-compiling from a *nix,
> though.

Would certainly help - please don't send it by mail though (I suppose
it's huge) as I only have RTC here ;) I don't think there would be any
compatibility problems. As far as I've seen, mingw32 cross compiler for
Linux just works as well as the Windows native one, so anything working
with gcc should make it.

> > It looks like the source distribution wouldn't need big changes to be
> > cross-compileable, which would be nice.
> 
> No, but since distutils is part of the Python build process, this got
> messier. In fact, in my cross-compiling effort, I had to replace the
> StrongARM ./python binary temporarily with a x86 ./python binary of the
> same version to have distutils build its extension modules. For
> cross-compiling, I'd suggest to add a switch --with-native-python= to
> the configure script.

I'll give it a second try following your suggestions. But I found no
trace of --with-native-python in the configure script.

> > I'd be glad to help making the source distribution cross-compileable,
> > provided I can solve this issue and you guys think the feature would
> > be usefull.
> 
> It sure would be very useful. One point: doesn't the configure script
> use "uname" at some point? This would need to be worked around, too. And
> one of these days I should resume my efforts with adding mingw32 as a
> supported platform for Python's autoconf.

Haven't cared about uname, but it's worth checking. And I'll probably
try harder to get a "out-of-the-box" compilation with mingw32. Your help
would certainly be precious, as you seem to already have attempted the
thing. Not to mention I still have no clue on how to include the builtin
module required for Windows systems! :)

Thanks for the reply anyway!
Alex.
-- 
http://www.gnurou.org





From phr-n2002a at nightsong.com  Fri Apr  5 07:30:25 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 04:30:25 -0800
Subject: PEP 285 and pickle compatibility
Message-ID: <7xsn6auy3i.fsf_-_@ruckus.brouhaha.com>

Here's a kludge that might steer around PEP 285 causing pickle
incompatibility between new and old python versions.  The issue
is if you pickle the value of 1==1, in new versions it will get
pickled as something that comes back as a bool rather than an int,
and the obvious method (adding a new pickle code for bool objects)
will cause breakage if you try to read the pickle with an old
version of python that doesn't know about bools.

Proposed fix: in new pickles, dump boolean True as "I01" and boolean
False as "I00".  The current unpickler should load these as the integer
values 1 and 0 and everything should work fine.  But these strings
will never be generated by the current pickler, which wouldn't emit
the leading 0's.  

In the new unpickler, put a special hack into the load_int routines
in cPickle.c and pickle.py, that notice the leading 0 and push the
appropriate bool value.

I think the ugliness of this kludge is outweighed by escaping making
an incompatible change to the pickle format.  Sooner or later, though,
a real incompatible change will be needed, so I'd like to propose
a further change now, that shouldn't cause compatibility probs:

The current pickle dump/dumps functions currently take a "bin" arg
specifying whether to dump in text or binary format.  Unfortunately,
this arg is treated as a bool and the dumps routine only cares whether
it's true or false.  So I think a further optional arg should be
added, giving a "format code" saying what unpickler versions have
to be able to read the pickle being created:

  dump(object, format=0)  # same as dump(object), text format pickle
  dump(object, format=1)  # same as dump(object,bin=1), binary format
  dump(object, format=2)  # new, incompatible format

Pickles created with format=2 might for example actually contain a
bool code instead of the I00/I01 hack, so old unpicklers couldn't read them.

If a future pickler/unpickler introduces an incompatible format, one
feature I'd hope to see added is efficient binary pickling of long
ints.  Currently they're pickled as ascii decimal strings (even in
binary pickles), which are very slow to convert to and from (all that
arithmetic).

Paul


From bokr at oz.net  Mon Apr  1 21:58:49 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Apr 2002 02:58:49 GMT
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com> <a8apju$kcb$0@216.39.172.122> <m3pu1jvw2d.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a8b6p9$q40$0@216.39.172.122>

On 02 Apr 2002 01:27:38 +0200, martin at v.loewis.de (Martin v. Loewis) wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> But it does make me think, should _all_ strings be subtypes
>> of a raw octet-string type according to their encoding? Then one
>> could visualize automatic inter-encoding promotions analogous to
>> numeric promotions, and if i/o sources and sinks have encoding
>> designators, Gon?alo's f.write("Mar?o 2002" + march.Name()) should
>> "just work" if the output encoding permits.
>
>That assumes that the output encoding is known, or can be
>determined. As-is, it can't - you don't know the encoding of f, and
Well, I was wondering more where we're heading than 'As-is' ;-)

IOW, assume encoding was an optional keyword parameter to open/file.
Then you'd know what output encoding was desired.

>you don't know the encoding of "Mar?o" (furthermore, the encoding of f
>won't help, since you have to perform the addition before invoking
>write).
Whoa, f is the last step. First would come the addition, and the premise
was that strings would have encoding attributes one way or another (different
subtype names?), so "Mar?o 2002" would have a known encoding and what comes
from march.Name() would have a known encoding. The encoding of the sum would
involve identifying an encoding that contained codes for both character sets,
and converting to that as necessary. Then comes the comparison with the f encoding.
Logically speaking. It might be optimized in case f is unicode, since then there's no
point in merging to an in-between encoding, (if that's even a possibility).

But no string would exist without at least an assumption as to its encoding.
I guess you could do unix file-type magic to infer encoding if you had to,
but it wouldn't seem reliable or cheap except utf & co.

It's an interesting exercise in keeping distinctions between codes and
representations straight ;-)

>In an all-Unicode approach, f would have been obtained from
>codecs.open, and the string literal would have been a Unicode literal.
>
One day maybe I'll need to think about this ;-)

Regards,
Bengt Richter


From gerhard at bigfoot.de  Tue Apr  9 12:48:40 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 9 Apr 2002 16:48:40 GMT
Subject: Beginner's performance problem
References: <memo.20020409174302.1584B@a.a>
Message-ID: <slrnab66r7.15p.gerhard@lilith.my-fqdn.de>

Mark Charsley wrote in comp.lang.python:
> 2) what's the best way of finding inefficiencies in python code (other 
> than asking here?). Are there any profilers out there for python?

Yes, in the profile and hotshot modules.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From just at xs4all.nl  Sat Apr 13 14:00:54 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sat, 13 Apr 2002 20:00:54 +0200
Subject: binding a reference to a variable
References: <mailman.1018682547.1577.python-list@python.org> <a99q7f$ss6$1@panix1.panix.com>
Message-ID: <just-44DB3A.20005413042002@news1.xs4all.nl>

In article <a99q7f$ss6$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> But it *is* silly.  That's why we have iterators.  One thing I noticed
> recently about how fragile the while 1 idiom is: it depends on
> readline() returning linebreaks.  Writing a file-like class for which
> this *isn't* true breaks the idiom.

I'd say it breaks the file protocol, part of which is: if read() or 
readline() returns the empty string you're at the end of the file. 
Nitpicking, nitpicking...

Just


From quitelikely at yahoo.com.au  Tue Apr 30 23:05:54 2002
From: quitelikely at yahoo.com.au (Patrick W)
Date: 01 May 2002 13:05:54 +1000
Subject: Python vs. Perl, which is better to learn?
References: <mailman.1020147892.13578.python-list@python.org> <aamjat$qku$1@panix1.panix.com> <aamm5k$kv3$1@bubba.NMSU.Edu>
Message-ID: <826628efe5.fsf@acropolis.localdomain>

rmelson at gilia.nmsu.edu ("Bob Melson") writes:

> If you're asked to build a house, you certainly don't confine
> yourself to a single tool -- you use the tools appropriate to the
> job you have to do; why should it be different with scripting
> languages?

Because the "right tool for the right job" approach is wrong in this
context. It should apply to the choice of language _categories_, not
languages _within_ a category. There are better things to do with time
and mental energy than learn a dozen different ways of doing the same
thing. 

If it took a minimum of six months to learn how to push two different
brands of wheelbarrow, can you imagine an experienced builder saying
to his apprentice: learn both?


From mertz at gnosis.cx  Fri Apr 26 13:54:09 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Fri, 26 Apr 2002 13:54:09 -0400
Subject: [OT] What is Open Source?
References: <mailman.1019526833.9592.python-list@python.org>
 <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift>
 <3CC8D767.599720@cosc.canterbury.ac.nz>
Message-ID: <BRZy8kKkXIyO092yn@gnosis.cx>         

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote previously:
|But if RMS would be happy with a world in which there
|was simply no copyright, he wouldn't bother with the
|GPL -- he could get the same effect just by putting
|all the stuff he writes in the public domain.

Surely, Greg, you see the error in this comment!  Being in a *world*
with no copyright is a very different thing than just putting your own
work in the public domain (no copyright).

Let's try analogy.  I would be happy in a *world* with no private
automobiles.  In such a world, almost per force, there would be good
public transportation most everywhere for people to use in getting
around.  In my real world, a bus comes to my entire town thrice a day,
and to get most anywhere, I need to spend hours waiting for transfers.
So in this actual world, I own an automobile.  My ideal would not be
obtained by disposing myself of my privately owned car.  Right?!

Yours, Lulu...

FWIW.  Even in this world, I still put my work in the public domain
(whenever I have control over the matter)... because I prefer to live in
a fantasy world of ethical purity.  But I realize that RMS is right
about the implications of the licenses in the real world.

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From tejarex at yahoo.com  Thu Apr 18 18:00:36 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 18 Apr 2002 22:00:36 GMT
Subject: PEP for an unrestricted __init__ ?!
References: <mailman.1019162616.18095.python-list@python.org>
Message-ID: <7mHv8.101687$K5.8348821@bin5.nnrp.aus1.giganews.com>

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1019162616.18095.python-list at python.org...

To expand on JR's answer and identify your misconceptions:

> In python each new instance of a class is created by __init__

No, instance is first created by internal default constructor or
new-style user-supplied equivalent __new__().  Then __init__ is
called, if it exists, with new instance as first arg to initialize
that new instance.

> __init__ would then
> be allowed to return something other[1] than the implicite 'self'.

> [1] For backward compatibility we might need to treat
>     'return None' as 'return self' because __init__ currently
>     seems to follow this rule.

__init__ is called as a proceedure -- *any* return is ignored

Terry J. Reedy





From astavale at yahoo.co.uk  Tue Apr  2 13:35:14 2002
From: astavale at yahoo.co.uk (Alistair Thomas)
Date: Tue, 02 Apr 2002 19:35:14 +0100
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de>
Message-ID: <uajueoqkg3f044@corp.supernews.com>

In article <3ca9d7d1$1_2 at news.uni-bielefeld.de>, "Mario Wehbrink"
<mario at bio58.uni-bielefeld.de> wrote:

> i wonder wether someone is using pygtk and if it is nice toolkit to
> learn. 

Take a look at this tutorial which I found very useful:

http://www.icon.co.za/%7Ezapr/Project1.html

Glade is a Gnome application for designing graphical interfaces. You will
then need libglade to parse the file and use pygtk to attach actions to
the events generated by the user when they do something.


From sholden at holdenweb.com  Wed Apr 24 08:21:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 08:21:56 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <slrnacag2g.1u3.gerhard@lilith.my-fqdn.de> <aa63ju$nva$1@news.ilk.de>
Message-ID: <7uxx8.44918$0x3.9466@atlpnn01.usenetserver.com>

"Steffen Kirschke" <steffen at molli15.org> wrote ...
> Gerhard H?ring wrote:
>
> > Fernando Rodr?guez wrote in comp.lang.python:
> >>
> >> I'm using CompatMysqldb from
http://sourceforge.net/projects/mysql-python
> >> with Python 2.2
> >>
> >> Whenever I try to insert some data into a mysql db, I get this error:
> >>
> >>>>> curs.execute("insert into customer_info (email, url, referrer,
> >>>>> first_visit) values ( '%s', '%s', '%s', '%s')",
> >>>>> [('billg at microsoft.com', 'www.easyjob.net', 'www.google.com',
> >>>>> '2001-03-07 12:32:00')])
> > AFAIC, you should be using a tuple as second parameter to execute, not
> > a list containing a tuple.
>
> and also use a '%' instead of the ','.
>
> curs.execute("""insert into table (col1, col2,...) values ('%s',
>         '%s',...)""" % (val1, val2, ...)
>

Please don't do that unless you have to (which you sometimes do if it's a
table that needs to vary between queries, for example).

The original example uses a parameterized SQL query. This means that the SLQ
engien has a chance to optimize the query independent of the actual data
values used, which in turn can speed up your application if the same query
is repeatedly used with different data.

Your version (which I used to suggest myself, so it's not like I'm accusing
you of a crime <wink>) generates a unique SQL statement which is guranteed
to be different from the next, and the one after that, depriving the SQL
implementation of the chance to optimize.

For completeness, I will repeat that *most* DBAPI-compliant modules '
execute() method will accept a list of tuples as data, repeating the query
with each tuple. Nowadays one should really explicitly use the executemany()
cursor method for such purposes, but befioer it was available execute() was
often written to accept a list of tuples.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From tim.one at comcast.net  Fri Apr 12 15:28:56 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 12 Apr 2002 15:28:56 -0400
Subject: New classes, __slots__, __dict__
In-Reply-To: <nilt8kKkX8EC092yn@gnosis.cx>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKKONAA.tim.one@comcast.net>

[David Mertz, Ph.D., wonders about __slots__ versus pickling]

[Michael Hudson, points to
    http://mail.python.org/pipermail/python-dev/2002-February/020072.html
    http://mail.python.org/pipermail/python-dev/2002-February/020129.html
]

Another link:

    http://www.python.org/sf/520644

That's bug report "[520644] __slots__ are not pickled".  The short course is
that someclass.__slots__ is unreliable because mutable, and so how to find
all the actual slots in all the classes (direct and base) by magic is
something of a puzzle and/or PITA.  Unless/until that does get done by
magic, you need to define a __getstate__ to spell out which slots do and
don't want to be pickled.





From look at sig.invalid  Thu Apr 18 07:29:11 2002
From: look at sig.invalid (Simo Salminen)
Date: Thu, 18 Apr 2002 11:29:11 +0000 (UTC)
Subject: Curses for windows.
References: <f8e5025.0204170953.570cfb43@posting.google.com> <m3it6qyxr5.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnabtbg7.8h5.look@pikkukorppi.cs.tut.fi>

* Martin v. Loewis [17 Apr 2002 20:40:30 +0200]
> alx78 at hotmail.com (alex) writes:
> 
>> Where is "curses" module for python 2.2 in windows NT.
> 
> Curses is not supported on Windows NT.

but you can use curses on NT if you use cygwin-python.

-- 
simo <dot> salminen <at> iki <dot> fi


From ak at silmarill.org  Sat Apr 20 18:51:17 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 20 Apr 2002 22:51:17 GMT
Subject: Disappearing global variables
References: <slrnac3lc2.3hm.ak@ak.silmarill.org> <_Mkw8.5895$Dm.409622@bin2.nnrp.aus1.giganews.com>
Message-ID: <slrnac3sau.4fs.ak@ak.silmarill.org>

In article <_Mkw8.5895$Dm.409622 at bin2.nnrp.aus1.giganews.com>, Terry
Reedy wrote:
> 
> "Andrei Kulakov" <ak at silmarill.org> wrote in message
> news:slrnac3lc2.3hm.ak at ak.silmarill.org...
>> You can look at full program at silmarill.org/pybook/pybook.py -
> it's
>> a little over 500 lines.
> 
> can you produce a considerable shorter (minimal) version that exhibits
> the same problem?
>
I figured this out.. I had a datafile that held a list of Book()
instances; then I changed Book() class a little, so here I was trying to
use old instances in the file where class was changed. I don't quite
understand why it deleted other instances from global namespace. I guess
if you do something like this, outcome is unpredictable :P.

 - Andrei


> 
> 
> 


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From robin at jessikat.fsnet.co.uk  Fri Apr  5 03:34:09 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 5 Apr 2002 09:34:09 +0100
Subject: PEP 285: Adding a bool type
References: <zpL9tKA1hNr8EwkQ@jessikat.fsnet.co.uk>
 <mailman.1017990311.8899.python-list@python.org>
Message-ID: <Uy2E7JABGWr8Eww9@jessikat.fsnet.co.uk>

In article <mailman.1017990311.8899.python-list at python.org>, Christian
Tanzer <tanzer at swing.co.at> writes
>
>Robin Becker <robin at jessikat.fsnet.co.uk> wrote:
>
>> In article <mailman.1017946214.8160.python-list at python.org>,
>> arthur.siegel at rsmi.com writes
>(snip)
>> >Guido will always be right to a following that will work to find
>> >the meaning that makes him right.
>>
>> In any large subject a 'Guido' will always claim to be bigger than 'God'
>> thus leading to Sergeant Pepper phenomena.
>
>Robin & Arthur,
>
>Why don't the two of you go and design and use a language of your own?
>
>Your repeated ad-hominem attacks on Guido make me sick.
>
>Or-go-get-a-life,
>
The table was sold as the best ever; why the need constantly to repair
it? I suspect because that's what the carpenter knows. 
-- 
Robin Becker


From marklists at mceahern.com  Wed Apr  3 09:46:02 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 08:46:02 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <016701c1db1a$569543e0$0900a8c0@spiff>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEPHKIAA.marklists@mceahern.com>

[Fredrik Lundh]
> more than one way to do it.  more than one concept to
> explain to newcomers.  more than one "obviously correct"
> way to do things.  less is more.  etc.

Hmm, I'm not sure I follow.

This is Python without the PEP:

	falseValues = [{}, [], 0, None]
	for x in falseValues:
		if not x:
			print "%s is 'false'." % x

This is Python with the PEP:

	falseValues = [{}, [], 0, None, False]
	for x in falseValues:
		if not x:
			print "%s is 'false'." % x

I don't see this vast conceptual chasm between pre and post PEP that you
appear to see.  Please help me understand.

Would the above be clearer if I called the collection of false values
'nothingValues' instead?  I, personally, don't think so.

// mark




From bokr at oz.net  Mon Apr  1 18:14:06 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 23:14:06 GMT
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com>
Message-ID: <a8apju$kcb$0@216.39.172.122>

On Mon, 01 Apr 2002 23:41:31 +0100, Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote:

>My problem is the following:
>
>Let f be a file object. If I do
>
>f.write("Mar?o 2002")
>
>all goes well. But if I try
>
>f.write("Mar?o 2002" + march.Name())
>
>where march.Name() returns a unicode string I get a unicode error. I
>tried converting both unicodes to strings via str but obviously I got an
>error (in the first string the culprit is the "?" character).
>
>Can someone help me out here and show me the way to write these strings
>to the file?
>
Someone else will be able to say OTTOTH, so I'll leave it to them (Martin?)

But it does make me think, should _all_ strings be subtypes
of a raw octet-string type according to their encoding? Then one
could visualize automatic inter-encoding promotions analogous to
numeric promotions, and if i/o sources and sinks have encoding
designators, Gon?alo's f.write("Mar?o 2002" + march.Name()) should
"just work" if the output encoding permits. Similarly for other
string character stream destinations such as exec, eval,
and compile, of course execfile, etc.

Is this where we're heading? Just a thought ;-)

Regards,
Bengt Richter


From brueckd at tbye.com  Mon Apr 22 12:10:20 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 22 Apr 2002 09:10:20 -0700 (PDT)
Subject: Science And Math  Was: Python's Lisp heritage
In-Reply-To: <MqVw8.47908$vm6.7506303@ruti.visi.com>
Message-ID: <Pine.LNX.4.33.0204220854000.25157-100000@ns0.tbye.com>

On Mon, 22 Apr 2002, Grant Edwards wrote:

> In article <mailman.1019483442.32389.python-list at python.org>, brueckd at tbye.com wrote:
> >
> >> > You should understand that this is a philosophical viewpoint. The
> > [snip]
> >> To put this another way, ask yourself where you stand on this
> >> question: is mathematics discovered or invented?
>
> Very well put.
>
> > Is there a difference?
>
> Yes.  Science verifies it's theories by comparing them to the
> physical world which they are attempting to describe.  It
> doesn't matter how elegant, or internally consistent Science
> is.  If it doesn't successfully describe the physical world,
> it's wrong.
>
> Science has an external reference point. Mathematics does not.
> Internal consistency is the only thing which mathematics can
> attempt to verify.  Mathematics is not an attempt to describe
> the physical world.

First, please back up, I was asking if there's a difference between
discovery and invention, which you don't mention anywhere in your
response.

Second, math *is* an attempt to describe the physical world - that's what
makes it useful. Case in point: calculus, which was invented/discovered
specifically to deal with and describe the motion of physical bodies. Had
it not been useful in describing such motion, it would have been tossed
out, or at least not so widely accepted. Math is useful because of its
relevance to the world around us.

Finally, in the more general sense, formal mathematical proofs and whatnot
might not require an external reference point, but the foundation upon
which they are built certainly does. Indeed, the fact that we've gotten as
far as something like abstract algebra is largely due to the fact that the
underlying building blocks *are* verifiable and applicable to our
experiences outside of math and many of the "advances" in mathematics have
been due to intuition or hunches provoked by real world phenomena.

-Dave





From pearu at cens.ioc.ee  Sun Apr  7 16:34:20 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sun, 7 Apr 2002 23:34:20 +0300 (EEST)
Subject: Lex and Yacc for Python
In-Reply-To: <slrnab19hf.23u.Alex@Tron.ddeus.lan>
Message-ID: <Pine.LNX.4.21.0204072332180.13987-100000@cens.ioc.ee>

On 7 Apr 2002, Oleksandr Kozachuk wrote:

> I need something, comparable to Yacc/Lex. Or something for
> Context-Free-Grammar. Can any one recommend me anything?

I recommend PLY:

  http://systems.cs.uchicago.edu/ply/

Pearu





From amuys at shortech.com.au  Thu Apr 11 02:51:31 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 10 Apr 2002 23:51:31 -0700
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <%34t8.36244$CU.41218@rwcrnsc51.ops.asp.att.net> <C25t8.5982$d7.1714559@newssrv26.news.prodigy.com>
Message-ID: <7934d084.0204102251.1cd87702@posting.google.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message news:<C25t8.5982$d7.1714559 at newssrv26.news.prodigy.com>...
> I guess this proves TMTOWTDI.

Which is a Tautology, but is it also obvious? :)

Andrae

> "Emile van Sebille" <emile at fenx.com> wrote in message
> news:%34t8.36244$CU.41218 at rwcrnsc51.ops.asp.att.net...
> > >>> for x,y in [(x,y) for x,y,z in zip(l,l[1:],[1,0]*len(l)) if z]:
> > ...     print x,y


From logiplexsoftware at earthlink.net  Fri Apr 26 13:25:29 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 10:25:29 -0700
Subject: python and very large data sets???
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F062339B30E6@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F062339B30E6@mail.hebmex.com>
Message-ID: <20020426102529.09777fa0.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 09:55:24 -0500
Gustavo Cordova wrote:

> MySQL holds it's own managing *huge* datasets, and it's *VERY*
> (yes, in capital letters and surrounded with asterisks) fast
> about doing it's work.
> 
> Just don't expect transactions, nor references. Strictly flat
> table stuff.
> 
> Bloody-evilly-fast, though.

Very:

http://www.eweek.com/article/0,3658,s=708&a=23115,00.asp

I've never tried it, but you _can_ get full transaction support if you use
MySQL-Max and Berkley DB or InnoDB tables rather than the standard ISAM tables.
 I don't know how large an effect this would have on performance.  The PC Labs
testers used a mixed-table approach for testing (InnoDB and ISAM).

http://www.mysql.com/downloads/mysql-max-3.23.html

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From bkc at Murkworks.com  Mon Apr 15 18:11:04 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 15 Apr 2002 18:11:04 -0400
Subject: Calling All PSU Members
References: <a9fbdj$5uq$1@newshost.accu.uu.nl>
Message-ID: <3cbb4e21$1_6@goliath.newsgroups.com>

I wish I had enough free time to write something as detailed as this..

;-)

--
Novell DeveloperNet Sysop #5





-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From tim.one at comcast.net  Sat Apr 13 23:23:30 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 23:23:30 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
In-Reply-To: <3CB87B52.9010702@rogers.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEPOONAA.tim.one@comcast.net>

[Mike C. Fletcher]
> I was unaware that there was no guarantee for 0-length tuples always
> being unique objects.  There are lots of places in my code where I use
> 0-length tuples in this pattern:
>
> NULL = ()
>
> class A:
> 	def b( self, object=NULL ):
> 		if object is NULL:
> 			#was not specified
> 		elif object is None:
> 			#specified to be object None

Mike, I'm pretty lost.  Are you saying that if a user explicitly calls

    A().b(())

or

    A().b((1, 2, 3)[2:2])

you *want* to hit the "was not specified" block?  That's all a guarantee
about 0-length tuples would buy you.  But if it's not what you want, such a
guarantee is irrelevant:  the method is full of holes regardless of whether
the empty tuple is unique.

Note that default arguments aren't designed to tell you whether a specific
argument was explicitly passed.  They're designed to supply reasonable
defaults that a user could just as well pass explicitly if they feel like
it.  Rather than invent more brittle tricks, I suggest pondering whether the
intended use of this gimmick isn't good enough for your purposes (and noting
that it's always been good enough for mine).

If you really need to know whether a thing was passed explicitly, it makes
more sense to insist that specifying it be done via keyword, and write the
function to examine the keys in the keyword dict passed to it.  It's all
obvious then.

    def f(required, **kw):
        if 'abc' in kw:
            print "'abc' was explicitly passed"
            abc = kw['abc']
        else:
            print "'abc' was not specified"
            abc = 42
        print "'abc' is", abc

    f(3)
    f(3, abc=666)





From op73418 at mail.telepac.pt  Mon Apr  1 15:21:02 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 01 Apr 2002 21:21:02 +0100
Subject: None not None in Excel COM?
References: <s4dhaukcq7v1v7gdgi9vcb9a46tiu6kv20@4ax.com>
Message-ID: <l0ghau4cdo1son3kejk0v7bel7fiutiqq2@4ax.com>

On Mon, 01 Apr 2002 20:30:33 +0100, Gon?alo Rodrigues
<op73418 at mail.telepac.pt> wrote:

>Maybe someone can help me out here.
>
>Im using Excel via Python COM and have already run makepy on its type
>library.
>
>The problem is the following:
>
>Let sheet be a reference to an Excel worksheet and suppose
>sheet.Cells(row, 1) returns None for a given row > 1. The problem is
>that if I do
>
>sheet.Cells(row, 1) == None
>sheet.Cells(row, 1) is None
>
>both return 0. What am I doing wrong?
>
>Thanx for all the help,
>Gon?alo Rodrigues

Am I a STUPID klut or what? I forgot to add .Value to the above, That
was why it was not working.

My apologies, 
Gon?alo Rodrigues


From phd at phd.pp.ru  Fri Apr 12 06:49:29 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 12 Apr 2002 14:49:29 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <lk4rihjgek.fsf@pc150.maths.bris.ac.uk>; from mwh@python.net on Fri, Apr 12, 2002 at 10:33:40AM +0000
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020412144929.B12997@phd.pp.ru>

On Fri, Apr 12, 2002 at 10:33:40AM +0000, Michael Hudson wrote:
> > case $MAKEFLAGS in \
> > *-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py -q build;; \
> > *) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py build;; \
> > esac
> > running build
> > running build_ext
> > building 'struct' extension
> > Segmentation Fault - core dumped
> 
> Yikes!  Can you dig into this?  I can't, not having access to a
> relavent machine... (you did run make clean, didn't you?)

   I'll try. What is worse, Python 2.2 failed the same way :( It seems
upgrading gcc from 2.8.0 to 2.95.3 did more harm than good :(((

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




From borcis at geneva-link.ch  Thu Apr 25 05:20:01 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 25 Apr 2002 11:20:01 +0200
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com>
Message-ID: <3CC7CA41.8030208@geneva-link.ch>

Tim Daneliuk wrote

> 
> I disagree strongly (but am willing to be convinced otherwise).  All of
> mathematics is a formal construct of the human mind created with the
> intent of removing ambiguity and enhancing our ability to describe what
> we *think* and what we *observe*.


(1) one may well deny you the adequacy of speaking of "the human mind"
in that manner, esp. if you further deny its existence as a bona-fide 
component of nature. Same remark for your use of "we".

(2) the nature/position/interpretation of ambiguity would deserve 
volumes by itself. Most people's understanding of the distribution of 
ambiguity is biassed by having it confused with cases of visible impact.

(3) I believe it's you who use in another post of this thread, the 
classical question "Is mathematics discovered or invented ? It's clearly 
invented". To me, such rhetorics display some lack of sophistication as 
regards the ambiguity of the value of single words/concepts.

My reply to it : consider *misunderstandings* as objects; clearly they 
have scientific status, and simple misunderstandings recur 
isomorphically. Clearly they are creations of "the human mind", as they 
don't exist and can't be described without rooting parts of any 
description of a misunderstanding to human minds. But are they 
"discovered" or "invented" ? Neither word applies; this kind of 
disproves your implied notion about innovation, that it should be either 
"discovered" or "invented".

My belief is that all this debate is very much complicated by "bootstrap 
notions" about the nature of understanding and knowledge, that need to 
be sorted out before any serious discussion is possible. Mathematics is 
indeed very much concerned with building the means to describe solutions 
to misunderstandings (if not the misunderstandings themselves). That's 
what imho you allude to when referring to "ambiguity", but this way of 
putting it kind of implies the confusion of misunderstandings with 
"misunderstandings traceable to the effects of spot lexical 
ambiguities". My understanding of Godel's theorem is precisely that 
there's much more to ambiguities than spot ambiguities of a single 
lexeme unit or word.

Nothing forbids to describe natural science in exactly the same terms, 
to say that it is concerned with destroying ambiguities, and solving 
eventual misunderstandings : calling repeatable experiments in as 
witnesses. Is being witness to a classical experiment of physics that 
much different from being witness to a proof of some classical math 
theorem ? Not to me.

> We can (and do) create all kinds of new
> calculi that are problem- or domain-specific.  That is, we *invent*
> mathematics to suit our needs - mathematics is not one-for-one
> correspondent to the natural universe.


To me what you are really saying is that mathematics can't claim to have 
a special adequacy to the natural universe, any different in essence 
than natural language. I mean that's how I need to translate your 
opinion to make it more or less my own.


> The reason I take this position is pretty simple.  We are able to construct
> mathematical systems which describe intellectual abstractions for which
> there is no natural analog.


This line of reasoning is imho biassed by a closed-world assumption that 
doesn't apply. Your sentence should really end with "...no natural 
analog I can spot as such".

> That is, mathematics can embrace way more than just the natural world

> and its workings.


I'd rather say that mathematics supplements our natural senses to the 
point of allowing us to touch matters that our (other) natural senses 
and languages do not allow us to touch or share (for various reasons).


> Similarly, a mathematician
> can introduce ideas like an "Incompleteness Theorem" which is a meta-mathematical
> commentary on mathematics itself - something far beyond the domain and range of
> mere natural science.


As I've already alluded above, my belief is that the Goedel Theorems 
have an interpretation in non-mathematical terms, as the idea that 
trying to kill lexical ambiguity by mechanically hardening the 
definitory relationship between lexemes is doomed to fail (in some 
situations) because the whole system of lexemes then becomes ambiguous, 
allowing coordinated reinterpretation of many lexemes in such a manner 
that the mechanical definitions stay invariant.


> 
> I'm not choosing sides here on the math vs. science debate - both are important
> artifacts of the human intellect.  I merely take umbrage with the notion that
> mathematics is innately wired into the universe somehow.  


I see what you mean, and if I am to choose corresponding umbrage with 
something that seems to base your notion, it's the idea that mankind 
could be ascribed the role of a single cognizer that can be contrasted 
to the universe as a whole.

Regards, Boris Borcic
--
"God is Truth too complicated for children, and that Mankind failed
  to teach itself ever since"



From pobrien at orbtech.com  Wed Apr  3 14:34:37 2002
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Wed, 3 Apr 2002 13:34:37 -0600
Subject: [Python-Dev] PEP 285: Adding a bool type
In-Reply-To: <NBBBIOJPGKJEKIECEMCBKEFDMLAA.pobrien@orbtech.com>
Message-ID: <NBBBIOJPGKJEKIECEMCBMEFEMLAA.pobrien@orbtech.com>

> The issue of dictionary keys has been raised, and is a bit perplexing, but
> not overly so, IMO. However, I'd like confirmation that this is indeed the
> intended behavior:
>
> >>> d = {}
> >>> d[0] = 'First'
> >>> d
> {0: 'First'}
> >>> d[False] = 'Second'
> >>> d
> {0: 'Second'}
> >>> d[0.0] = 'Third'
> >>> d
> {0: 'Third'}
> >>> d[()] = 'Fourth'
> >>> d
> {0: 'Third', (): 'Fourth'}
> >>> d[None] = 'Fifth'
> >>> d
> {0: 'Third', (): 'Fourth', None: 'Fifth'}
>
> Or are there any situations like this where you would want True
> and False to
> be a bit more like None and a bit less like ints? Meaning True and False
> could be dictionary keys separate from 0 and 1.

I see that you just addressed this and confirmed that True==1 and False==0
in all contexts. That's fine by me. I just wanted confirmation.

---
Patrick K. O'Brien
Orbtech





From cliechti at gmx.net  Wed Apr 24 20:15:21 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Apr 2002 02:15:21 +0200
Subject: sql statement generation problem
References: <aa7g4o$djf$1@mannews.swan.ac.uk>
Message-ID: <Xns91FB1732029BAcliechtigmxnet@62.2.16.82>

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in
news:aa7g4o$djf$1 at mannews.swan.ac.uk: 
> when using the following line to create an sql statement:  query2 =   
> "select * from tblitem where itemtypeid = %(itemTypeID)s and
> %(searchIn)s like '%(query)s'" % sqlValues    
> 
> it works, but it need to have this  like '%%(query)s%' so that oracle
> will 
>  perform the search properly. However, the extra %'s seem to mess
>  things  

whe you use the % operator on a string you must use double '%%' to get one 
in the resulting string:

.. '%%%(query)s%%'" % sqlValues

> up. Is there any way to get round this. Also, DCOracle2 is returning
> 1.0 for a value of 1 in the database. This is totally messing things
> up, and I need to find a way round this.

"int(x)" ? when you know that there are no floats, why not?

> Any ideas,
> Hugh
> 
> 



-- 
Chris <cliechti at gmx.net>



From eric.brunel at pragmadev.com  Mon Apr 15 06:08:46 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 15 Apr 2002 10:08:46 +0000
Subject: Python-tkinter and canvas tags
References: <ca58272d.0204131249.3f7e8663@posting.google.com>
Message-ID: <a9e1hq$1ede$1@norfair.nerim.net>

JB wrote:
[snip]
> def handleEnter(self, event):
> self.canvasRef.itemConfig(CURRENT,
> fill=self.HLColor)
> 
> When I run it this way, the events are definitely being handled, but I get
> an error stating that "NameError: Global name 'CURRENT' is not defined".

Errrr, maybe you forgot to do "from Tkinter import *" in your module? 
CURRENT is just a constant defined in this module with the string value 
'current', so you just have to do a "from Tkinter import *" to use it. You 
may also just do an "import Tkinter" and use Tkinter.CURRENT instead of 
just CURRENT. In other words, don't look for magic here: CURRENT is just a 
constant variable like any other one, and getting a NameError using it just 
means that you didn't do the right things to get it properly defined...

BTW, AFAIK, the method itemConfig on canvases does not exist; it's named 
itemconfigure.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From steve at ferg.org  Fri Apr 26 13:44:34 2002
From: steve at ferg.org (Stephen Ferg)
Date: 26 Apr 2002 10:44:34 -0700
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net>
Message-ID: <b16e4ef7.0204260944.193432b9@posting.google.com>

The other replies in this thread have answered your question.  Now I'd
like to express sympathy with it.

As a person coming from a non-Unix background (IBM and Windows), I
experienced the same confusion and problems with "raw" strings that
you did.  And I'm sure others will, in the future, also.

I would describe the situation this way... Python's "raw" strings
aren't truly raw at all.  Python does not have any capability for
TRULY raw strings -- that is, for string literals that are not escaped
in any way.  Escaping is still performed even with "raw" strings.  So
the term "raw" is misleading.

I would like to see Python support TRULY raw string literals -- string
literals that are not escaped in any way.  Perhaps they could be added
with a prefix of z (for zero escaping?), so that

     z"\" == "\\"
     z"\'" == "\\\'"

This might help reduce the probability of others experiencing the same
confusions in the future.


From donn at u.washington.edu  Tue Apr  9 13:10:38 2002
From: donn at u.washington.edu (Donn Cave)
Date: 9 Apr 2002 17:10:38 GMT
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8v36u$j3p$1@news.nsk.su>
Message-ID: <a8v7ae$a10$1@nntp6.u.washington.edu>

Quoth "Andrey Panchenko" <andrey at app.omsk.su>:
| "Oleg Broytmann" <phd at phd.pp.ru> wrote:
|> On Tue, Apr 09, 2002 at 04:45:08PM +0200, Henning Peters wrote:
|> > is there anything similar to the pointer concept from c/c++ in python?
|>
|>    In Python *everything* is a pointer. Really.
|>
|> > i would like to store values in two different locations. both values should
|> > always be equal. the one could be a pointer on the other. is this possible
|> > to do this in python?
|>
|> a = 1 # put a pointer to an Integer objetc into variable a
|> b = a # copy the pointer into b; now both b and a hold pointers to the same
|>       # Integer object
|>
|> c = [a, b] # make a list holding two pointers
|
| I think it's rather an ambiguous statement. I can think that after
|
| a = 1
| b = a
| a = 3
|
| b must hold 3. But it's not true. IMHO it's better to say that the python
| names are like pointers only for mutable objects.
| (Caution - newbie opinion)

Well, it's hard to use these words without some risk of confusion.
There is indeed some similarity between a Python reference and a
"pointer" in C, and all you need for your example is to bear in mind
that "b = a" means "let b point to the same object as a" - in other
words, the assignment does make a pointer of "b", but the "right hand
side" automatically dereferences pointer "a".

Meanwhile, the answer to the original question ("how can I be two places
at once") does require a mutable object.  If your two places both refer
to a single object, then when attributes of that object change, both will
see the same change.

But that's really changing the question so that it can have a positive
answer.  Whatever the relationship between Python names and pointers,
it really applies to every single object, whether mutable or immutable.

   class T:
       def __init__(self):
           self.x = 'original'
   a = T()
   b = a
   b.x = 'new'
   print a.x   (-> 'new' ... this is the mutable object strategy.)

but
   b = T()
   print a.x   (-> 'new' ... whether a and b were mutable or not, 
                             they now point to different objects.)

Mutable objects' only special property outside of their particular kinds
of mutability, is that they can be dictionary keys.

	Donn Cave, donn at u.washington.edu


From tim.one at comcast.net  Sun Apr  7 14:59:47 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 07 Apr 2002 14:59:47 -0400
Subject: Regex question about \d
In-Reply-To: <20020407181149.GA28549@lilith.my-fqdn.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCAELMOLAA.tim.one@comcast.net>

[Gerhard H?ring]
> Can somebody please explain this to me:
>
> >>> print re.search(r"\d*", "asdfsfafsw 453 dfsa").group(0)
>
> >>> print re.search(r"\d+", "asdfsfafsw 453 dfsa").group(0)
> 453
>
> Why doesn't the first regex match the digits?

A Python/Perl regexp search matches at the leftmost position it can match.
Any regexp of the form (whatever)* can match an empty string, so searching
for any such regexp matches at the start of the string.  If the only thing
it can match at the start of the string is the empty string, good enough:
that is a match.






From aleax at aleax.it  Sun Apr 28 13:16:45 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 19:16:45 +0200
Subject: Update locals()
In-Reply-To: <20020428115221.GF16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke> <0204280830290C.16995@arthur> <20020428115221.GF16217@devel.trillke>
Message-ID: <E171sIX-0007um-00@mail.python.org>

On Sunday 28 April 2002 01:52 pm, holger krekel wrote:
	...
> > Please give ONE example which, in your opinion, is best handled
> > with 'exec' without explicit dictionaries rather than in any other way.
>
> i try to make it short and answer questions (if any of you gurus
> ever have any :-) as needed. Obviously you can always pass explicit
> dictionaries to exec (globals, locals). That can't be the point.

Nope.  locals() and globals() are allowed by the language definition to be 
mappings rather than fully general dictionaries, and in particular any
alteration performed on locals() or globals() is allowed (this could change
at any time, even in a micro-version or a bug-fix one -- the docs have been 
saying so for a LONG time) to [a] alter the namespace, [b] silently
ignore the alteration attempt, [c] raise an appropriate exception.

In the current implementation, locals() alterations when the namespace is in 
a function follow rule [b] (they're silently ignored), other locals() and 
globals() alterations rule [a], but no program that relies on either behavior 
is correct Python (might be nice for PyChecker to warn, but it's probably too 
hard to track).  I _think_ the best solution might be to have them as 
mappings that give a warning when altered, on the way to making them 
read-only per rule [c] in a release or two. 

> I cite some of my code.
>
> class filename:
>     """ Abstraction for filename ... """
>     # ...
>
>     for name in filters._stateful_all:
>         doc=filters.__dict__.get(name,undoc).__doc__.strip()+'\n'
>         exec indentedcode("""
>
>     def %(name)s (self,*args,**kargs):
>         '''%(doc)s'''
>         return filters.%(name)s (*args,**kargs) (self)
>     """ % locals())

Basically you want to add to class filename a bunch of methods, each
delegating as you show to the homonym callable attribute of global object 
filters.  This is most easily done after the filename class object is
complete and available - i.e. right after the class statement, or in a
metaclass.  The "right after" approach is probably easier (using Python
2.2 for simplicity, doesn't make much difference anyway):

def addMethods(class_, filters, statefulList):
    for name in statefulList:
        delegatee = getattr(filters, name)
        def method(self, *a, **k): return delegatee(*a, **k)(self)
        method.__doc__ = getattr(filters, name, undoc).__doc__.strip() + '\n'
        setattr(class_, name, method)

You may need to tweak this (I'm not sure why the filter.name access is
unconditional when looking for the delegatee, but falls back to undoc when
looking for the __doc__, for example).  But the point is that this way the
code is reusable (you may apply it to as many separate classes as you
wish, several sequences of methodnames for the same class, etc), clean,
efficient -- all claims that don't hold for the exec-based approach.  Most
often, setattr and getattr do offer superior alternatives for such kinds of
metaprogramming.

If having to explicitly call

addMethods(filename, filters, filters._stateful_all)

right after the 'class filename:' statement body's end nags you, you can
use a custom metaclass for the same purpose, e.g.:

class filename(methods_adder):
    _methods_source = filter
    _methods_names = filters._stateful_all
    # etc etc

with:

class MC_methods_adder(type):
    def __init__(self, *args):
        class_ = type.__init__(self, *args)
        methods_source = getattr(class_, '_methods_source', None)
        methods_names = getattr(class_, '_methods_names', ())
        addMethods(class_, methods_source, methods_names)
        return class_

class methods_adder:
    __metaclass__ = MC_methods_adder

or the like (untested code, but I hope the general idea is clear).
Custom metaclasses (in 2.2., at least) can be handy for such
metaprogramming jobs.


> - i am a maniac at avoiding redundancies.

...and yet the infrastructure embodied in class filename as you coded
it is not reusable save by copy-and-paste, the worst kind of redundancy.
addMethods and the MC_methods_adder metaclass avoid redundancy
by placing such metacoding-infrastructure in ONE place.


Alex




From ianb at colorstudy.com  Tue Apr 16 15:54:26 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 16 Apr 2002 14:54:26 -0500
Subject: Python webmail packages?
In-Reply-To: <7xsn5ve9hj.fsf@ruckus.brouhaha.com>
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com>
	<7876a8ea.0204151614.138853be@posting.google.com>
	<mailman.1018918474.3622.python-list@python.org>
	<7xsn5wjk85.fsf@ruckus.brouhaha.com>
	<mailman.1018941693.2296.python-list@python.org> 
	<7xsn5ve9hj.fsf@ruckus.brouhaha.com>
Message-ID: <1018986866.16813.187.camel@localhost>

On Tue, 2002-04-16 at 14:20, Paul Rubin wrote:
> Ian Bicking <ianb at colorstudy.com> writes:
> > > Hmm, how much code is involved?  I didn't want to deal with installing
> > > a huge package like Zope.
> > 
> > Installation isn't hard, like I said it's CGI.  It happens to reuse some
> > code from Zope, but all that code is contained in the package itself.
> 
> Thanks.  Any impressions of how it compares to squirrelmail?  It
> sounds like it isn't being actively maintained, which is not so
> attractive.  Squirrelmail on the other hand has bugs that I can't seem
> to convince the developers are really bugs.  I can patch my local copy
> with fixes, but I'd rather do that with something written in Python.

It's in maintenance mode -- the developer isn't putting much work into
it, or adding features, but he hasn't disappeared either.

  Ian






From fgeiger at datec.at  Sat Apr 13 12:07:16 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Sat, 13 Apr 2002 18:07:16 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>
Message-ID: <a99l3r$r8g$1@rex.ip-plus.net>

Hi Chris,

> null pointers.  Yet the grid control kept GPF-ing on me for weeks until I
> figured out that Python was "blowing away pointers" behind the scenes for
> garbage collection.

How did you find, *where* this happend and how did you solve this problem?
Did you deactivate gc (don't know if it's possible)?.

Best regards
Franz


"Chris Spencer" <clspence at one.net> schrieb im Newsbeitrag
news:vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v at 4ax.com...
> Okay, I guess I should chime in here.
> I am a professional developer for a small software company.  My company
> has been using Python for its projects for the past 18 or so months.  We
have a
> whole product suite written in Python.  Until recently, this product was
95%
> console driven, even on our Windows port.  Customers have been asking for
a GUI
> to tie all the parts of our product together.
> For the past month, I have been developing a GUI Administrator program
> for our product using wxPython.
>
> Some statistics:
> Code is hovering around 6000+ lines in 12 modules
> Approx. 60% of that code is directly related to wxPython
> Development time in man-hours: 280
>
> Pros:
> 1.  It has incredible flexibility.  I've done some pretty weird stuff that
I'd
> plain just be afraid to try in MSVC++.
> 2.  Rapid development.  This project would've taken me nearly a year to do
in
> C++/MSVCRT.  At a ROUGH guess, it would've hovered around 30,000-40,000
lines.
> 3.  Robust library.  I used to dive into www.codeguru.com to grab source
code
> for custom controls so I could modify them to my needs.  I've felt no real
> reason to "build" new controls to suit my needs with wxPython.
> 4.  Maintainability.  Those 6000 lines of code I've done are all pretty
much in
> the category of "self-documenting code".
>
> Cons:
> 1.  BAD documentation.  They've taken the wxWindows library documentation,
added
> a few notes, and called it the wxPython documentation.  And the thing is,
they
> haven't kept it updated for more recent releases of wxWindows/wxPython, so
> there's a lot of undocumentated features that just don't show up.  I had
to
> start using the wxWindows docs from the wxwindows site, and interpret (and
> source-code dive) what a reasonable wxPython equivalent would be.  LOTS of
> trial-and-error.
>
> 2.  This project would've taken half the time it eventually took if
wxPython was
> either documented better, or took greater pains to protect the pointers
that
> wxWindows sometimes uses.  I'm in Python fer chrissakes, I shouldn't care
about
> null pointers.  Yet the grid control kept GPF-ing on me for weeks until I
> figured out that Python was "blowing away pointers" behind the scenes for
> garbage collection.
>
> 3.  Bloat.  The support files alone for my program (the wxWindows .dll,
the
> wxPython .pyd files, etc) amount to about 6 megs of space.  The program's
BASE
> memory usage upon startup is 14 megs.  This might not seem so bad for a
6000+
> line program until you realize that a *100* line program would STILL take
up 14.
> megs of memory.
>
> 4.  Did I mention that the documentation is both incomplete and odious?  I
> should not have to constantly do a "find in files" to find out what
parameters a
> damned method call takes because the way wxPython implemented a method is
> different than the wxWindows way and someone forgot to make a note of it
in the
> rubbish passed off as a "manual".
>
> 5.  I've had a REAL problem with conservative "window refreshes", where
unless I
> specifically tell each window to repaint itself, I get a lot of scroll
bars
> missing, a lot of windows being "grey" (shown without controls), Tab
windows
> without the tab headers, etc.  wxWindows/wxPython should be more
aggressive
> about repainting windows.  Even MCVCRT controls are better about
repainting
> themselves without having to be told that their render clipping planes are
> dirty.
>
> The pros outweight the cons by a WIDE margin once you've climbed the
> learning curve.  While on that slope, though, the lack of robust and
complete
> documentation will HURT you.
>
> Chris.
>
> On Mon, 8 Apr 2002 11:55:24 +0200, "Harald Schneider"
> <h_schneider at marketmix.com> wrote:
>
> >Hi,
> >
> >what are your experiences, using wxPython/wxWindows in large projects?
> >
> >Large means in my case a sales system, based on MySQL.
> >I'm thinking about implementing it in wxPython, but I'm not sure about
the
> >stability of
> >wxWindows and the wxPython wrapper ...
> >
> >The alternative way would be a project in Borland Delphi, which has
proofed
> >its
> >stability over years ...
> >
> >So what do you think about it!
> >
> >Thanks a lot,
> >Harald
> >
>




From richard at electrophobia.com  Wed Apr 24 22:13:36 2002
From: richard at electrophobia.com (Richard Parker)
Date: Thu, 25 Apr 2002 02:13:36 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com>
Message-ID: <B8ECB435.F916%richard@electrophobia.com>

Paul Rubin at phr-n2002a at nightsong.com wrote on 4/24/02 4:42 PM:

> Do the keys really need to be independent?

As I recall, it was essential to the HMAC proof of security that the keys be
independent.  I don't know if is possible to prove HMAC secure with
dependent keys, but it would almost certainly require stronger assumptions
about the underlying hash function.

> How is H(K || (H || 'a' || K || X))?

I am concerned about it.  You're basically using HMAC with a weaker
pseudorandom key derivation.  If you absolutely can not use the xor of ipad
and opad, make sure you use a key derivation for both K1 and K2.  Perhaps
the following:

  H(c1 || K || H(c2 || K || x))

I strongly recommend you just use HMAC, rather than inventing your own MAC
construction.

-Richard



From mixo at beth.uniforum.org.za  Thu Apr 11 03:37:25 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Thu, 11 Apr 2002 09:37:25 +0200
Subject: a list from lists
Message-ID: <3CB53D35.5030308@beth.uniforum.org.za>

I have a list of lists (which are not of the same size but can be).
How can I join lists to for a new list with elements from
the list?

i.e. from,
   listOfList = [ [1,2,3],[4,5],[6,7,8,9]]

I want to get,
   NewList= [1,2,3,4,5,6,7,8,9]

The new list does not need to be in any particular order.



From noah at noah.org  Tue Apr 23 02:06:04 2002
From: noah at noah.org (Noah Spurrier)
Date: Mon, 22 Apr 2002 23:06:04 -0700
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com>
Message-ID: <0e2e01c1ea8c$f6319490$0200a8c0@saysyou>

Most crypto libraries output 8-bit binary. You are probably trying to store
that binary in a VARCHAR2 field in Oracle. Maybe it thinks its Unicode.

What you want to do is also use a binary to ASCII encoder.
Python includes a number of different modules to encode binary data
as plain ASCII text. Take a look at the base64 module:
    http://www.python.org/doc/current/lib/module-base64.html#l2h-3019
or for more detailed control try the binascii module:
    http://www.python.org/doc/current/lib/module-binascii.html
So just take the result of your rotor encryption and pass it through base64
then store that result in your database.

You could also look into what BLOB fields Oracle supports. I've never used
them.
I have always had painful experiences using BLOBs in databases.

The Enigma rotor encryption is really bad. I wish Python came with at least
one
half-way decent encryption algorithm like AES or DES.

Noah

----- Original Message -----
From: "Hugh" <h.e.w.frater at cs.cf.ac.uk>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, April 22, 2002 6:41 PM
Subject: need encryption algorithm


> Hi all,
>   I need a simple encryption algorithm that will not give me
> characters that I can't store in my oracle table. At the moment I'm
> using the rotor encryption module because I like it's simple interface
> but It's generating characters that oracle can't store which is a damn
> pain. What other modules are available that are easy to use and will
> do what I wan't. Remember that security is not paramount here, I'm
> only including it for functionality.
>
> Thanks,
> hugh
>





From loewis at informatik.hu-berlin.de  Mon Apr 29 07:38:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 29 Apr 2002 13:38:03 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <aaj45k$odt$1@thorium.cix.co.uk>
Message-ID: <j4k7qqenvo.fsf@informatik.hu-berlin.de>

gbreed at cix.compulink.co.uk writes:

> The rationale's very brief about the disadvantages of a bloated 
> __builtins__.  It isn't like they're reserved words.  

Actually, some of them will become reserved words: None, True, False.

Regards,
Martin


From pyth at devel.trillke.net  Mon Apr 22 23:39:16 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 23 Apr 2002 05:39:16 +0200
Subject: rlcompleter not calling __getattr__ on [ ]
In-Reply-To: <aa2460$534$1@peabody.colorado.edu>
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org> <aa2460$534$1@peabody.colorado.edu>
Message-ID: <20020423033916.GJ27395@devel.trillke>

On Mon, Apr 22, 2002 at 04:49:56PM -0600, Fernando P?rez wrote:
> holger krekel wrote:
> 
> > or maybe even r"(\S+(\.\w+)*)\.(\w*)" ? This would principally
> > allow to add
> > 
> > "".<tab>
> > 
> > to work as expected. Of course you need to remove " and '
> > from the delimiter list.
> > 
> > But then again, i wonder if the rlcompleter should not simply
> > see the *whole* string and do the string-split according to delimiters
> > itself? This might be more flexible for 'general purpose' completion.
> > I consider the delimiters to be a rlcompleter property and
> > not a readline property.
> 
> Well, your new regexp suggestion works great for me. You may want to repost 
> this to python-dev or as a patch to SF, so that the powers-that-be have a say 
> on the issue. For the time being, I'm using your regexp in my patched version 
> of rlcompleter ...

great. but i guess we should try to come up with a solution that
allows 

  a=[[].<tab>

and similar cases to work. Also

  a=().<tab> 

should be allowed, too. I'll check that tommorow. Additionally
i want to see if it is possible to 

a) (optionally) just return method names which don't start with '__'

b) return to 'readline' strings of the form:
   
   COMPLETION "docstring"

   as this would show you brief documentation for each method.

i am probably looking into that tommorow.
Right now i am writing some long awaited dynamic 
'hall of fame' webpages for nethack :-)

	holger




From eppstein at ics.uci.edu  Tue Apr 30 15:40:15 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 30 Apr 2002 12:40:15 -0700
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org> <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no> <mailman.1020193732.29753.python-list@python.org>
Message-ID: <eppstein-346BE0.12401530042002@news.service.uci.edu>

In article <mailman.1020193732.29753.python-list at python.org>,
 holger krekel <pyth at devel.trillke.net> wrote:

> > Thes names aren't reserved in any way...
> 
> right. i do it all the time for file,str,dict,string,list...
> they are simply to short and descriptive to be left bound
> to something i don't access directly. and using type('') or
> friends is usually a very precise way of expressing that
> you want the type object for that particular thing.

So what do you do when you have overridden type?

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From neuhauser at mail.cz  Tue Apr 30 06:42:12 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 12:42:12 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <6qvga9bizr.fsf@abnoba.intevation.de>
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de>
Message-ID: <20020430104212.GI325@roman.mobil.cz>

> From: Bernhard Herzog <bh at intevation.de>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: 30 Apr 2002 12:04:08 +0200
> 
> 
> Roman Neuhauser <neuhauser at mail.cz> writes:
> 
> > > >> The problem with int.chr() is that a lot of people are going to get
> > > >> confused when they find that "42.chr()" doesn't work (although "42
> > > >> .chr()" does).
> [Alex explains that 42. is a float]
> >     Right, I didn't realize 42. was a valid float. The space would be a
> >     reasonable (beit tedious) solution. As long as this is highlighted
> >     in the docs, I don't see a problem.
> 
> Why should chr be an int method in the first place? It's not tied to
> ints all that much[1]. IMO it sits somewhere between ints (including
> int-like objects !) and strings so it shouldn't be a method of either
> class. Leaving it a function is perfectly fine, therefore, although
> putting it into the string module might be a good idea. The same would
> apply to ord, IMO.

    Ugh, that looks to me as appropriate as join() being a string
    method / module function.

> [1] In fact, chr today accepts any object that can be converted
> successfully to an int, so that chr(42.5) does work, but that's not
> always desirable.

    Any object that can be converted to a character should have a chr()
    method.

-- 
FreeBSD 4.4-STABLE
12:40PM up 1 day, 3:05, 11 users, load averages: 0.00, 0.04, 0.06




From brian at sweetapp.com  Mon Apr 29 13:54:33 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 29 Apr 2002 10:54:33 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <Xns91FF62618790Flumberjack@216.148.53.85>
Message-ID: <007001c1efa7$04ede140$445d4540@Dell2>

> Huh? The "core" language wouldn't be measurably shrunk by this
proposal
> fella. Your rationale is built on air. Show us some genuine
_advantage_
> that is worth the huge _cost_ in time and effort caused by your ill-
> considered proposal. How about you and other PEP writers take some
> responsibilities for your fantasies for once? You up to helping
thousands
> rewrite their code? I thought not.

Why the hostility? 

I think that every PEP is a good one. Even if the PEP is ultimately
rejected, it promotes thought about some aspect of the language. Taking
responsibility sounds like stifling discussion to me.

Cheers,
Brian





From wilk at flibuste.net  Sat Apr 13 08:17:54 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Sat, 13 Apr 2002 14:17:54 +0200
Subject: wich odbc is stable ?
Message-ID: <20020413141754.6d14e50d.wilk@flibuste.net>

hi,

I need to connect to an MSAccess database for simple query (no lock, no
transaction), so i would like to know if the standard odbc provided by
win32extension is enough stable even if it's written to be no supported...
Or must i use callDll or ado ? 

thanks to read my so bad english writing

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From mikeb at mitre.org  Mon Apr  8 13:13:25 2002
From: mikeb at mitre.org (Mike Brenner)
Date: Mon, 08 Apr 2002 13:13:25 -0400
Subject: Discussion of Python midi library
Message-ID: <3CB1CFB5.AF4A221F@mitre.org>

Here is a suggestion as to what to encapsulate (each one is harder than the next.

1. Midi Instruments (preferably the hundred general midi instruments from grand-piano up through gunshot and the hundreds of xg midi instruments available on the more advanced synthesizers and sequencers). The attributes of the instruments would include:

	- a handle with which to grab the instrument -- for example, if my application wishes to grade the quality of low pitch bells (gongs) on a particular sequencer, giving me a set of poor gongs (clavinet, celesta, cow bell, and fx_5_brightness), a set of fair gongs (tubular bells, glockenspiel), and a set of good gongs (lower range pan flute, charang, tinkle bell, and silver bell), I would need unique names for all these instruments
	- the name of the instrument
	- the midi code to invoke that instrument on the currently selected hardware
	- over what ranges of pitch they are valid (don't use a piccolo 3 octaves below middle C)
	- over what ranges of rhythm they are valid (don't use legato strings for 64th notes)
	- instruments they are similar to, if the original choice is not available


2. Midi Selection Sequences. Each midi hardware will have certain selection sequences, bank selection commands, patch selection commands, controllers, etc. There must be a way of doing the machine independent stuff in a machine independent manner and yet there must be a way to insert the device-specific stuff. The currently selected hardware must be part of the model. There must be ways to insert, delete, modify, move, and copy notes and other commands into channels, including setting their velocity (loudness), frequency (pitch), duration, etc. A single command will enter a chord, a triple, a comment such as key, a meta event, etc.


3. Image. A string image function must represent every object, possibly in a standard format like xml, musicml, etc.)


4. Time. There must be a consistent time model used throughout, which easily represents notes, rests, start, finish, etc.


5. Graphical representations in staff format, say, in jpeg or png format.


6. Reading in a bit-mapped graphical representation (say a jpeg or png file) of a staff into midi format.


7. Pattern recognition of a wav file into a midi format, choosing the right notes, durations, and instruments.





From loewis at informatik.hu-berlin.de  Thu Apr 18 07:32:10 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 13:32:10 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019123256.19883.python-list@python.org>
Message-ID: <j41ydd8cp1.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> So:
> 
> #
> # In addition, *static* reverses this effect (negating a previous
> # *shared* line).
> 
> is ambiguous, because it does not reverse all the effects of *shared*?

Maybe. This is meant in the context of the Setup file: all modules
listed after shared are compiled statically. If no modules are listed,
this has no effect.

If you can propose a rewording, please submit a patch to
sf.net/projects/python.

> I have tried to configure this in different ways:
> configure
> configure -- with-gnu-ld

Python's configure does not support --with-gnu-ld, so this option has
no effect.

> What else does it look for?  

Use 'gcc --print-prog-name=ld' to find out which ld binary gcc uses.

> If I symling /usr/local/bin/ld to gld will configure find it?

No. If you create an ld link in
/usr/local/lib/gcc-lib/<system>/<version>; that will make gcc use a
different ld binary.

*However*, there is no need to do that; the system linker should work
 as well. Please invoke the python linker line line manually, adding
 -v:

gcc -v -o python Modules/python.o libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm

That will tell precisely which linker is invoked, and what options are
given to it (most likely, the linker will be collect2 which invokes
the real linker, so you have to consider the output where the linker
identifies itself as well).

Regards,
Martin


From ruediger.maehl_nospam at web.de  Fri Apr 12 01:37:38 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Fri, 12 Apr 2002 07:37:38 +0200
Subject: How to access attributes of nodes in xml.minidom
References: <mailman.1018534534.16214.python-list@python.org>
Message-ID: <newscache$buwfug$8y3$1@www-neu.dzsh.de>

Thanks, Mark.
That's helping.

R?diger




From op73418 at mail.telepac.pt  Mon Apr 15 19:58:46 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 16 Apr 2002 00:58:46 +0100
Subject: Calling All PSU Members
References: <a9fbdj$5uq$1@newshost.accu.uu.nl>
Message-ID: <71qmbugpvlrcjfk6e7feq1on5046e49atu@4ax.com>

On 15 Apr 2002 19:58:43 GMT, m.faassen at vet.uu.nl (Martijn Faassen)
wrote:

>Secret Secret Secret -- For Python Secret Underground Members ONLY.
>Do NOT distribute outside the PSU. You will be eliminated from the
>space-time continuum entirely. That is not an idle threat. 
>
>As stated in previous dispatches, flame wars can be used to extend the
>maximum age of the universe. Subsequently PSU members have been dilligently
>adding to such discussions. In particular, PEP 285 (Adding a bool type),
>has been a point of contention. This has resulted in a change of the
>cosmological constant, protonic decay rate and the spawning quotient of
>quantum child universes. The cumulative effect lengthened the future 
>by approximately 534.731 billion years.

You've got your physics wrong: The change in the cosmological constant
has the wrong sign! This means that the B-fields entering in the vertice
calculations for the spin foam interactions contrtibute much less, which
induces a contraction of the universe at an exponential rate. It will
cease to exist in a mere few million year at which point the laws of
physics (now and in 1023794 years ahhhhhhhrrrrrrrgggggggggg




From marklists at mceahern.com  Wed Apr 10 10:27:04 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 10 Apr 2002 09:27:04 -0500
Subject: Working with untrusted coders
In-Reply-To: <9739136773arocas@alumni.uv.es>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEEIKKAA.marklists@mceahern.com>

[Arnau Rossells]
> [...]Python security
> about private elements and others is merely a "don't step on the grass"
> sign, everybody can change private functions and data in classes and even
> modules. I'd like to know if there is some way of really securing python,
> so parts of the code can't be overriden by the coders. Has anybody tried
> it?

I haven't used this but it may be what you're looking for:

  http://www.lemburg.com/files/python/mxProxy.html

By the way, here's the Google search that resulted in the above as the first
hit:

  http://www.google.com/search?hl=en&q=python+proxy+access

It's always a good idea to search for things first and Google is your
friend.

// mark





From greg at cosc.canterbury.ac.nz  Sun Apr 14 23:18:08 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 15 Apr 2002 15:18:08 +1200
Subject: Looking for doh!.. simple sequence type
References: <3CB6BED8.2000502@mxm.dk> <a99k6c$g9c$1@news.ilk.de>
Message-ID: <3CBA4670.3189B2A8@cosc.canterbury.ac.nz>

Steffen Kirschke wrote:
> 
> or, instead of __getitem__
> 
>         def __iter__(self):
>                 self.iter = 0
>                 return self
> 
>         def next(self):
>                 if self.iter == self.len:
>                         raise StopIteration
>                 else:
>                         self.iter += 1
>                         return 42

The __iter__ method should really create and return
a new object, otherwise two for-loops iterating
over the same object will get in each other's way.
For example,

  class LstIter:

    def __init__(self, seq):
      self.seq = seq
      self.count = 0

    def next(self):
      if self.count == self.seq.len:
        raise StopIteration
      else:
        return 42

  class Lst:
    ...
    def __iter__(self):
      return LstIter(self)

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


From jm.hoeltzli at coelis.com  Wed Apr 24 06:12:57 2002
From: jm.hoeltzli at coelis.com (Jean-Marc Hoeltzli)
Date: Wed, 24 Apr 2002 12:12:57 +0200
Subject: urllib2, cookies and redirect.
Message-ID: <006301c1eb78$9c3b0100$0201a8c0@fmcf>

Hi,

I would like to get an http documment whith the ulrlib2, but
-the remote server send me a cookie whith a documment I retrieved bofore.
-i do
    myreq=myurllib2.Request(the_url)
    myreq.add_header("Cookie",Cook)
    urlh=myurllib2.urlopen(myreq)
    data=urlh.read()
to get my documment.
-the webserver answer is an HTTP 302, which is handeled by the urllib2, but
the second get doesn't send the cookie.

any idea ??

Jean-Marc









From westernsam at hotmail.com  Mon Apr 15 09:21:08 2002
From: westernsam at hotmail.com (sam)
Date: 15 Apr 2002 06:21:08 -0700
Subject: upgrading to python2.2
Message-ID: <292c8da4.0204150521.35d4a842@posting.google.com>

Hi,

I'm trying to upgrade from python 1.52 to python 2.2. Most of my old
code works fine. However the SioModule I used
(http://starship.python.net/crew/roger/) doesn't seem to work (even
though I've got the 2.2 edition). When import it I get the following
error:

C:\Python22\Serial\MsTimer.py:5: RuntimeWarning: Python C API version
mismatch f
or module elapsedTimer: This Python has API version 1011, module
elapsedTimer ha
s version 1007.
  import elapsedTimer

When trying to use the module the ReadTerminated method doesn't seem
to recognise the 'rspTerm' value I use in the port dictionary.

Has anybody else successfully ported (no pun intended) from 1.52 to
2.2 using this module?

Any clues to this gratefully received

Sam


From mwh at python.net  Thu Apr 11 09:43:11 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 11 Apr 2002 13:43:11 GMT
Subject: Easter Eggs
References: <mailman.1018391602.4247.python-list@python.org> <b0f083db.0204101941.1f37632b@posting.google.com> <mailman.1018497202.25727.python-list@python.org>
Message-ID: <lkwuvewau5.fsf@pc150.maths.bris.ac.uk>

Geoff Gerrietts <geoff at gerrietts.net> writes:

> Quoting Jeremy Hylton (jeremy at alum.mit.edu):
> > Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote in message news:<mailman.1018391602.4247.python-list at python.org>...
> > > Also, is anything else hidden in the Python interpreter that issues
> > > similar wisdom?  (I know...  I could read the source; but this is
> > > probably quicker).
> > 
> > I can think of one other small feature of this sort, but I don't think
> > I should tell you.
> 
> Now that's just plain cruel. ;)
> 
> Is this your one-man crusade to get a lot of eyes on the source? ;)


I'll give you a clue:

f

Cheers,
M.

-- 
  The bottom tier is what a certain class of wanker would call
  "business objects" ...                      -- Greg Ward, 9 Dec 1999


From sholden at holdenweb.com  Tue Apr 23 17:59:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 17:59:42 -0400
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net>
Message-ID: <KRkx8.1147$0x3.1000@atlpnn01.usenetserver.com>

"Colin Brown" <cbrown at metservice.com> wrote in message
news:3cc5d5c8$1 at news.nz.asiaonline.net...
> I have entered the following at the command line in Python 2.1.1 on Linux
> and in Python 2.2 on NT and get the same results:
>
> >>> r'\'
>   File "<string>", line 1
>      r'\'
>         ^
> SyntaxError: invalid token
> >>>
>
> It appears as if the r option for strings does not handle a backslash as
the
> last string character properly.
>

That's correct, and the documentation explains why in Section 2.4.1 of the
Python Reference Manual:

"""When an `r' or `R' prefix is present, a character following a backslash
is included in the string without change, and all backslashes are left in
the string. For example, the string literal r"\n" consists of two
characters: a backslash and a lowercase `n'. String quotes can be escaped
with a backslash, but the backslash remains in the string; for example,
r"\"" is a valid string literal consisting of two characters: a backslash
and a double quote; r"\" is not a valid string literal (even a raw string
cannot end in an odd number of backslashes). Specifically, a raw string
cannot end in a single backslash (since the backslash would escape the
following quote character).""" (italics omitted to avoid posting HTML)

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp





From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 13:48:30 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 19:48:30 +0200
Subject: Trinary operator?
Message-ID: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>

Not that it's a big deal since there's many ways to do that, but can I use
something like
    verboseGender = (gender == 'm') ? 'male' : 'female'
in Python?




From sugarking666 at hotmail.com  Mon Apr 15 21:41:07 2002
From: sugarking666 at hotmail.com (Gus Tabares)
Date: Mon, 15 Apr 2002 21:41:07 -0400
Subject: [Tutor] Need help
Message-ID: <F305ZGUTwSIQK9VPRfL00009b57@hotmail.com>

Faizal,

  Although we cannot teach you to learn to program, something that is 
complex and time consumming, we can point you in the right direction. Try 
www.python.org for Guido's tutorial on Python or try a few links that have 
been mentioned here the past few days.

http://www.freenetpages.co.uk/hp/alan.gauld/
http://livewires.org.uk/python/index.html

Good luck...

Gus Tabares


>From: Faizal <faizal at umland.com.my>
>To: "'python'" <tutor at python.org>, "'python'" <python-list at python.org>
>Subject: [Tutor] Need help
>Date: Tue, 16 Apr 2002 09:19:59 +0800
>
>Hi there...
>
>I am still new(no exprience in programming)...&  need help from you guys...
>Can u teach me how to learn programming (with example....)
>
>
>Thank you
>
>Faizal
>
>---
>Outgoing mail is certified Virus Free.
>Checked by AVG anti-virus system (http://www.grisoft.com).
>Version: 6.0.325 / Virus Database: 182 - Release Date: 19/02/02
>
>
>
>
>
>_______________________________________________
>Tutor maillist  -  Tutor at python.org
>http://mail.python.org/mailman/listinfo/tutor




_________________________________________________________________
Join the world?s largest e-mail service with MSN Hotmail. 
http://www.hotmail.com





From xx758 at cam.ac.uk  Fri Apr  5 06:48:37 2002
From: xx758 at cam.ac.uk (Xiao-Qin Xia)
Date: Fri, 05 Apr 2002 12:48:37 +0100
Subject: Fail to use jythonc
Message-ID: <a8jv8m$s7k$1@pegasus.csx.cam.ac.uk>

Dear everybody,

I installed a jython recently, but cannot use jythonc.
for example, a very simple program is:

# mytest.py
import time
t1=time.time()
k=0
while k<10000000:
    k=k+1
print time.time()-t1


jython can run it, but when use jythonc, error ocurred:

"""
[xqxia at mwise1 jython-2.1]$ ./jythonc mytest.py
processing mytest

Required packages:

Creating adapters:

Creating .java files:
  mytest module

Compiling .java to .class...
Compiling with args: ['/usr/local/j2re1.3.0/bin/javac', '-classpath', 
'/home/xqxia/jython-2.1/jython.jar::./jpywork::/home/xqxia/jython-2.1/Tools/jythonc:/home/xqxia/jython-2.1/.:/home/xqxia/jython-2.1/Lib', 
'./jpywork/mytest.java']
1  java.io.IOException: /usr/local/j2re1.3.0/bin/javac: not found

Consider using the -C/--compiler command line switch, or setting
the property python.jythonc.compiler in the registry.
ERROR DURING JAVA COMPILATION... EXITING
[xqxia at mwise1 jython-2.1]$
"""

Anybody know how to solve this problem?

Many Thanks from 

Xiao-Qin Xia


From mlh at vier.idi.ntnu.no  Fri Apr  5 22:58:11 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 03:58:11 +0000 (UTC)
Subject: [Python-Dev] ACCEPTED: PEP 285 (truth pickles)
References: <mailman.1017974474.30402.python-list@python.org>
Message-ID: <slrnaassij.178.mlh@vier.idi.ntnu.no>

In article <mailman.1017974474.30402.python-list at python.org>, Tim Peters wrote:
>[Roy Smith]
>> If I pickle a bool, and then restore it, do I get the same singleton
>> back?
>
>Yes.  However, booling a pickle is a different question entirely.
>
>don't-even-think-about-it-ly y'rs  - tim

According to the "something vs. nothing" principle, a pickle should
always be true... While no pickle should be false.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From jeff at ccvcorp.com  Wed Apr  3 22:09:40 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 3 Apr 2002 19:09:40 -0800
Subject: NTLM authorization - compile into EXE
References: <9b309481.0204031300.70ce3c63@posting.google.com>
Message-ID: <MPG.171563d334a3f2cf989695@news.nwlink.com>

In article <9b309481.0204031300.70ce3c63 at posting.google.com>, 
mscheuner at hfx.eastlink.ca says...
>
> But under Windows, having an extra command line window open just for
> this NTLM proxy thingie is a bit "annoying" - so I thought I'd compile
> it into an EXE, using Py2EXE. 

Actually, while creating an EXE might have other advantages, it's 
not the right way to go about solving your problem.

The exact way to solve it depends somewhat on how your proxy 
script is being triggered, but the general solution is that you 
need the script to be run by pythonw.exe (a windows executable 
with no visible window) instead of by python.exe (a console 
application).  If you are launching the script directly (through 
a "python NTLMproxy.py" command) then you can simply change that 
command to "pythonw NTLMproxy.py".  If the script is being 
launched indirectly, then you can probably achieve the same 
results by changing your script's extension to .pyw -- I.e., 
"NTLMproxy.pyw".  The registry will recognize that .pyw files are 
used with pythonw.exe and .py files are used with python.exe.

Hope that helps...

-- 

Jeff Shannon
Technician/Programmer
Credit International


From bokr at oz.net  Sat Apr 13 22:16:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 02:16:44 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com>
Message-ID: <a9aoqc$j94$0@216.39.172.122>

On Sat, 13 Apr 2002 15:58:19 -0700, "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:

>"Martin v. Loewis" <martin at v.loewis.de> wrote in message
>news:m3ofgn1avt.fsf at mira.informatik.hu-berlin.de...
>| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
>|
>| > I think an 'xor' operator is needed.  The 'xor'
>| > operator means 'logical-exclusive-or', and
>| > returns a 1 or 0.
>|
>| In Python 2.3, you will be able to write
>|
>|    bool(a) ^ bool(b)
>|
>| Seems pretty clear to me.
>|
>| Regards,
>| Martin
>
>You didn't read my post well enough-- we can already _DO_
>that with Python v2.2.x:
>
>(not a) ^ (not b)
>
>IMHO, neither the above, nor the future v2.3.x idiom:
>
>bool(a) ^ bool(b)
>
Hm, ^ should probably be != to return bools instead of integers?

>is as clear as:
>
>a xor b
>
>...
>
>Everyone_is_entitled_to_have_my_opinion'ly yours,
>
Well, to tie in with existing spelled-out 'and' and 'or' operations,
it should probably do something along those lines (i.e., evaluate to
one or other of the objects, not one of the boolean values used in the logic),
but I can't think of something sensible.

Regards,
Bengt Richter


From Oschler at earthlink.net  Fri Apr  5 22:58:13 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 5 Apr 2002 22:58:13 -0500
Subject: Rule Engine in Python
References: <eb612768.0203101940.5565b282@posting.google.com> <m3it82d32l.fsf@wilma.localdomain> <I45r8.34886$0T6.10550048@e3500-atl1.usenetserver.com> <m3u1qq3ma8.fsf@wilma.localdomain> <B4nr8.36171$ku5.3121156@e3500-atl2.usenetserver.com> <m3zo0hiqaa.fsf@wilma.localdomain>
Message-ID: <Pfur8.49644$0T6.13163838@e3500-atl1.usenetserver.com>

Harry George <hgg9140 at seanet.com> wrote in message
news:m3zo0hiqaa.fsf at wilma.localdomain...
>
> Yes, at least for our context.  Out of perhaps 100 prolog modules,
> only 1 or 2 really have declarative rules which take advantage of
> backtracking.  The others are procedural (usually enforced with cuts).
> For that python is a better choice.  Either way, the heavy lifting for
> 3D graphics is in the underlying C/C++/OpenGL libraries.  Prolog/Pyton
> are there to add intelligence.
>

Ok, got it now.  Having once made a largely procedural email application
many years ago in Prolog, I know that kind of shoehorning to be very
painful.  What I couldn't understand was the possibility that Python is
better at rule creation/maintenance/processing than Prolog. IMHO, that is
precisely what Prolog excels at compared to other languages, along with
self-manipulation of code as data, expression parsing, and limited
constraint solving.

thx





From aahz at pythoncraft.com  Wed Apr 24 18:28:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Apr 2002 18:28:02 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa731c$q53$0@216.39.172.122>
Message-ID: <aa7bhi$iam$1@panix1.panix.com>

In article <aa731c$q53$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>
>All this stuff will be re-inventing data base stuff, so you might
>want to consider going that route, as others have suggested. A flat
>fixed-width source sounds like an easy importing/loading problem. I
>haven't tried MySQL on windows, but on Linux is was not hard to get
>going, and it might be faster than a fully transactional DB like
>postgresql (I don't know any recent benchmarks). I would guess that you
>could beat the disk space requirements by writing a special-purpose
>program, though.

A recent PC Magazine article said that MySQL is *very* fast,
particularly if you don't need transactional processing.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From aahz at pythoncraft.com  Tue Apr 16 01:06:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 01:06:59 -0400
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com>
Message-ID: <a9gbhj$dhm$1@panix1.panix.com>

In article <3CBBAB98.91FA1B17 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>jepler at unpythonic.net wrote:
>>
>>     [2, 3, 5, 7, 9, ...]
>> would generate successive primes.
>
>Pretty lousy code, since 9 is not prime.  :-)

That just proves he's an engineer (or is it a physicist? I can never
remember that joke properly).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From JamesL at Lugoj.com  Fri Apr  5 23:26:16 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 04:26:16 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018040171.16545.python-list@python.org> <3CAE1D58.4050500@mxm.dk>
Message-ID: <Xns91E7D088415CFJamesLLugojcom@216.148.53.84>

Max M <maxm at mxm.dk> wrote:
> brueckd at tbye.com wrote:
> 
>>>because of the problems I have
>>>actually encountered with making code work with the many extant
>>>versions of Python, I will no longer be doing any new projects in the
>>>language and am not recommending Python to anyone. 
> 
>> So, what language did you move on to?
> 
> 386 assembler probably. It's a little less portable, but the specs 
> haven't changed in over a decade.

I'm at a loss to understand all the snide remarks. Some I understand, some 
I don't. I've been using Python for at least 4 years and think it is a nice 
scripting language. It is very useful and allows quick development, but it 
is not indispensable. Besides, you weren't entirely wrong: I need something 
that runs very fast in a small amount of ram and assembler or cross-
compiled C is what I'll probably be using. But it isn't an x86. And PC 
support code will be C++ most likely. Ask me again in 6 months or so and 
I'll let you know what I've decided on (okay, it wont be Perl for sure).


From tim.one at comcast.net  Wed Apr  3 15:27:13 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 03 Apr 2002 15:27:13 -0500
Subject: [Python-Dev] PEP 285: Adding a bool type
In-Reply-To: <200204031946.g33JktP25857@odiug.zope.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEOECOAA.tim.one@comcast.net>

[pobrien at orbtech.com]
> How about operator.truth()? Has Tim changed his mind at all? Has Guido?

Guido change his mind?  That's about as likely as Tim changing his <wink>.
Guido Pronounced on this and that's good enough for me.  I'd still prefer
operator.truth() return 0/1, but this always was (as the PEP said from the
start) a "minor additional issue".  People can learn to use int(bool(x))
when they need 0/1, and operator.truth() becomes dead weight.





From daniel at dittmar.net  Sat Apr 27 15:01:10 2002
From: daniel at dittmar.net (Daniel Dittmar)
Date: Sat, 27 Apr 2002 21:01:10 +0200
Subject: Python as scripting plug-in for large C++ project
References: <mailman.1019880409.19217.python-list@python.org>
Message-ID: <3CCAF576.3060304@dittmar.net>

Bjorn Pettersen wrote:

>  - We extensively (exclusively) use our own basic datatypes,
>    e.g. NInteger which in addition to keeping a value also
>    stores whether it is NULL (think databases). How much work
>    is it going to be to expose these to embedded Python?


When mapping from/to Python, translate NULL values to Py_None. Creating 
multiple full fledged numeric types is probably more trouble than it is 
worth.

 
>  - Our object/relational mapping tool (written in Python :-)
>    outputs ~100KLocs of C++ that defines an object tree that 
>    ends in timeseries data ("vectors") or records. The code
>    is highly optimized and makes sure we keep data integrity. 
>    Would it be simpler to use SWIG/SIP/boost to expose this, 
>    or would it be just as easy to have the mapping tool 
>    generate the Python glue directly.


Just a shot in the dark: having both C++ and Python representation of 
the same data in memory will probably make data integrity quite 
difficult, so expose your current data structures via SWIG etc. for 
production use. You might also think about generating the Python glue 
directly. This isn't really that difficult. It might actually be easier 
to generate this with your knowledge of the data structures rather than 
trying to sqeeze it into the SWIG etc. input.

Generating Python bindings directly will give you a great tool to play 
with new ideas, so I would probably (time permitting) add this one as well.


Daniel




From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 16:54:06 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 22:54:06 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9i2kg$cg8$1 at nets3.rz.RWTH-Aachen.DE...
> Is it possible to have a safe eval() on the server-side even when the
string
> consists of submitted user data?
>..

I forgot to mention: all that should be allowed is arithmetics (boolean as
well) plus a limited set of functions that I write & define.
Say the following is OK (I already replaced internal variables here):
1 == 2 and false or (10 + 3 * 2) == 4 or internal_harmlessFunction() == true

Whereas this clearly isn't:
1 == 2 and false or (10 + 3 * 2) == 4 or delete_all_files('c:\\') == true


I just realized, as opposed to PHP, there's much stricter namespaces in
Python and maybe not using "import" on anything harmful could be the
solution?




From jon+usenet at unequivocal.co.uk  Thu Apr  4 11:07:50 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Thu, 04 Apr 2002 16:07:50 -0000
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk> <slrnaaok7j.4ei.jon+usenet@snowy.squish.net> <3CAC567F.6090407@mxm.dk>
Message-ID: <slrnaaouim.h07.jon+usenet@snowy.squish.net>

In article <3CAC567F.6090407 at mxm.dk>, Max M wrote:
>> http://foo/a/b/c + d       == http://foo/a/b/d
>> http://foo/a/b/c + ../d    == http://foo/a/d
>> http://foo/a/b/c + ../../d == http://foo/d
> 
> Then what is:
> 
> http://foo/ + d       == ?

http://foo/d of course. I don't see your point.                                 
See RFC 2396 for the rules on relative URL resolution.                          


From jeff at ccvcorp.com  Thu Apr 18 15:05:01 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 18 Apr 2002 12:05:01 -0700
Subject: Win32 Open Source or free .NET integration?
References: <RSJu8.3093$MM2.1226076@news1.news.adelphia.net> <uhemc4ano.fsf@ctwd0143.fitlinxx.com> <e00pbugs69lk497ga1u2j5oh8bllbcf7a8@4ax.com>
Message-ID: <MPG.1728b89c75b3d3bf9896b2@news.nwlink.com>

In article <e00pbugs69lk497ga1u2j5oh8bllbcf7a8 at 4ax.com>, 
mkelly2002 at adelphia.net says...
> On 15 Apr 2002 22:52:59 -0400, David Bolen <db3l at fitlinxx.com> wrote:
> 
> >That actually seems reasonably unlikely at this point.  The "Visual
> >Python" project at ActiveState is to tie into Visual Studio for .NET
> >(e.g., the IDE) and not tying Python itself into the CLR (Common
> >Language Runtime) of .NET.
> 
> That's a bummer.  Reading the stuff on their site
> "full integration" leads one to believe if you plug
> Visual Python in it's plugged in just as good as
> C# and that's obviously not the case. 

No, that *is* the case, you're just not reading the whole site.  
It's integrated into Visual Studio, *not* into the .NET 
framework.  The confusion results from the fact that Microsoft 
has decided to call the new version of VS by their latest 
buzzword -- Visual Studio .NET -- when it is not directly tied to 
the framework or the CLR at all.  They're two entirely separate 
issues, although MS would probably be happy to have you never 
realize that.  (Admittedly, the ActiveState site didn't seem to 
explicitly point out the difference, last time I looked, but 
their meaning *is* clear if you know the difference -- it's MS's 
confusion, not ActiveState's.)

But Visual Python integrates Python into the Visual Studio IDE as 
well as C# is integrated into VS.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mwh at python.net  Wed Apr 10 06:26:03 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 10 Apr 2002 10:26:03 GMT
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54>
Message-ID: <u4rijals0.fsf@python.net>

"Emile van Sebille" <emile at fenx.com> writes:

> Kalle Svensson
> >
> > "import this", IIRC.
> >
> 
> 
> Oooh... that's almost as much fun as saying 'xyzzy' the first time...
> ;-)

That's not as much fun as "syzygy" (which is a real word,
astonishingly enough).

Cheers,
M.

-- 
  Reading Slashdot can [...] often be worse than useless, especially
  to young and budding programmers: it can give you exactly the wrong
  idea about the technical issues it raises.
 -- http://www.cs.washington.edu/homes/klee/misc/slashdot.html#reasons


From torkel at sm.luth.se  Tue Apr 23 01:41:45 2002
From: torkel at sm.luth.se (Torkel Franzen)
Date: 23 Apr 2002 07:41:45 +0200
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> 	 <mailman.1019299357.11513.python-list@python.org> 	 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> 	 <Xns91F67C40EB0E0lumberjack@216.148.53.83> 	 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> 	 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> 	 <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at>
Message-ID: <vcbk7qzezty.fsf@beta13.sm.luth.se>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:

> Wittgenstein dissmised the pure behavior of mathematicians: proof n, do
> not rest and proof n+1, do not rest and proof n+2,...

  What comments of Wittgenstein's do you have in mind here?

> Wittgenstein and Goedel met in the so called "Wiener Kreis" (an exquiste
> circle --on a regular basis-- for intellectuals in Vienna/Austria).

  Wittgenstein and Godel never met.



From pricee at hhbrown.com  Tue Apr  2 16:29:01 2002
From: pricee at hhbrown.com (Erik Price)
Date: Tue, 2 Apr 2002 16:29:01 -0500
Subject: compiling Python with readline
In-Reply-To: <20020402192502.GA13827@isis.gerrietts.net>
Message-ID: <A6BB000A-4680-11D6-A42D-0050E4857868@hhbrown.com>

On Tuesday, April 2, 2002, at 02:25  PM, Geoff Gerrietts wrote:

> Quoting Erik Price (pricee at hhbrown.com):
>> Hello,
>>
>> Sorry to trouble if this is not the right list -- I would like to
>> re-compile my Python 2.2 source (on Mac OS X 10.1.3) to support 
>> readline
>> (or whatever it takes to get emacs-style key bindings to work in the
>> interpreter, like I get in bash).

> The Modules/Setup file is automatically created from (I believe)
> Modules/Setup.dist, but will not be overwritten if it exists.
>
> In other words, if you need to use a nondefault Setup file, cd into
> Modules, cp Setup.dist to Setup, and edit Setup.
>
> When I built 2.2, the configure process automatically discovered
> the existence of readline and built the interpreter properly. If you
> haven't tried that, you might consider trying it. If it doesn't work,
> you may have some configuration problems with the readline library (or
> you may not, but it's possible).
>
> Best of luck (if you need more help, feel free to ask).

Thanks for the reply, Geoff!  I found the Modules/Setup.dist file and 
copied it to Setup as per your instructions (and the instructions in the 
file).  I also found the following text in the file:

<quote source="Modules/Setup">
# GNU readline.  Unlike previous Python incarnations, GNU readline is
# now incorporated in an optional module, configured in the Setup file
# instead of by a configure script switch.  You may have to insert a
# -L option pointing to the directory where libreadline.* lives,
# and you may have to change -ltermcap to -ltermlib or perhaps remove
# it, depending on your system -- see the GNU readline instructions.
# It's okay for this to be a shared library, too.
</quote>

So I uncommented the line, but my 'make' command failed with the 
following error:

<quote source="make output">
cc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -no-cpp-precomp -I. 
-I./Include -DHAVE_CONFIG_H   -c ./Modules/readline.c -o 
Modules/readline.o
./Modules/readline.c:21: readline/readline.h: No such file or directory
./Modules/readline.c:22: readline/history.h: No such file or directory
make: *** [Modules/readline.o] Error 1
</quote>

I then rm'd the source tree and re-extracted (to avoid messing with 
config.cache and make clean), and tried adding the -L option described 
in the Modules/Setup script.  My syntax looks like this:

<quote source="Modules/Setup">
readline -L /sw/lib/ readline.c -lreadline -ltermcap
</quote>

But this time my ./configure failed.

My libreadline.* files are located in /sw/lib, they are 
/sw/lib/libreadline.4.dylib, /sw/lib/libreadline.a, and 
/sw/lib/libreadline.dylib.  I am guessing that I am giving the wrong 
syntax for the -L option.  Can anyone see what I'm doing wrong here?

Thank you for your time,


Erik



----

Erik Price
Web Developer Temp
Media Lab, H.H. Brown
pricee at hhbrown.com




From fperez528 at yahoo.com  Mon Apr 22 13:03:52 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 11:03:52 -0600
Subject: Numeric Python, COM and Eigenvalues
References: <9e1f166b.0204220437.57a087de@posting.google.com>
Message-ID: <aa1ft2$mn7$2@peabody.colorado.edu>

Stefano wrote:

> I am working with Numpy and trying to create a COM server.
> I managed to call this server by Excel and get results like Inverse
> matrix or Cholesky decomposition but the server doesn't work as I ask
> for Eigenvalues/vectors (Excel does not respond anymore).
> Has anyone heard of a problem of this kind?

Well, with zero example code and details provided that's going to be a hard 
one to get help on ;)

Do you know for a fact you're feeding it a diagonalizable matrix? Does numpy 
compute the eigenvals/vects nicely if you test it with the same matrix on its 
own (not via excel)? Start testing with a symmetric 2x2 real matrix for which 
you can hand-compute the eigenvalues trivially, that might tell you where the 
problem is.

Good luck,

f.


From gashton at cmedltd.com  Fri Apr 12 11:42:35 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 12 Apr 2002 16:42:35 +0100
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
In-Reply-To: <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
References: <mailman.1017656355.5887.python-list@python.org>
	<12bb92e6.0204091223.738e0798@posting.google.com>
	<i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> 
	<GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
Message-ID: <1018626156.27398.16.camel@tp6>

On Fri, 2002-04-12 at 13:28, Steve Holden wrote:
> "Brad Bollenbach" <bbollenbach at shaw.ca> wrote ...
> >
> > This doesn't actually make sense. To call software "Open Source" is to
> > acknowledge it as being distributed under a license that is defined as
> > compatible with what Richard Stallman calls "Free Software". 

This isn't really the right place for this discussion, but the above is
incorrect. Stallman didn't coin the term "Open Source", the OSI
(www.opensource.org) did. IIRC, Stallman doesn't agree with the term
Open Source, as it can dilute the "free as in speech" aspect of GPL
compatible licenses, which the FSF stands for. It's all explained
here....

  http://www.fsf.org/philosophy/free-software-for-freedom.html

> > So you don't have the "choice" of licensing Open Source software. It got 
> > to be called "Open Source" *because* of the license you've already
> > chosen for it.

So you're saying that you can't license *your own* software to different
people under different agreements, if you so choose?

> So the code I have put in the public domain isn't open source? That's
> interesting.

Yes, it is (but I suspect you knew that). See the venn diagram on:

  http://www.fsf.org/philosophy/categories.html

-- 
Graham Ashton





From jkring at calbay.com  Tue Apr  2 11:30:36 2002
From: jkring at calbay.com (Jim)
Date: 2 Apr 2002 08:30:36 -0800
Subject: shelve files
Message-ID: <89a9b6ce.0204020830.7934adf0@posting.google.com>

I moved a shelve file from one machine to another and I cannot open it
on the new machine.  Is there something that I should be aware of?  I
moved it from a Redhat 6.X machine to a 7.X machine.

-Jim

>>> db = shelve.open('shelveFile')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python1.5/shelve.py", line 152, in open
    return DbfilenameShelf(filename, flag)
  File "/usr/lib/python1.5/shelve.py", line 142, in __init__
    Shelf.__init__(self, anydbm.open(filename, flag))
  File "/usr/lib/python1.5/anydbm.py", line 86, in open
    return mod.open(file, flag, mode)
  File "/usr/lib/python1.5/dbhash.py", line 8, in open
    return bsddb.hashopen(file, flag, mode)
bsddb.error: (22, 'Invalid argument')


From daniels at mail.dsl-only.net  Sat Apr 13 02:35:33 2002
From: daniels at mail.dsl-only.net (Scott David Daniels)
Date: Sat, 13 Apr 2002 06:35:33 GMT
Subject: Forcing quote characters in repr(STRING), how?
References: <mailman.988841104.7613.python-list@python.org> <Xns90968DA40B8A3duncanrcpcouk@127.0.0.1> <mailman.1018662566.16282.python-list@python.org>
Message-ID: <1103_1018679733@news.dsl-only.net>

On 12 Apr 2002 21:48:25 -0400, pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard) wrote:
..
> Despite not recommended, I nevertheless followed the trick above in Pymacs.
> Preceded by a long comment, the little piece of code looks almost clean :-).
> 
>     [...]
>     elif type(value) == types.StringType:
>         # Python delimits a string it by single quotes preferably, unless
>         # single quotes appear within the string while double quotes do
>         # not, in which case it uses double quotes for string delimiters.
>         # Checking the string contents, the C code stops at the first NUL.
>         # We prefix the string with a single quote and a NUL, this forces
>         # double quotes as delimiters for the whole prefixed string.  Then,
>         # we get rid of the representation of the single quote and the NUL.
>         write('"' + repr("'\0" + value)[6:])
>     [...]

I'd just add a slight tweak tbe a nickel faster:
..
    # Python delimits a string it by single quotes preferably, unless
    # single quotes appear within the string while double quotes do
    # not, in which case it uses double quotes for string delimiters.
    # Checking the string contents, the C code stops at the first NUL.
    # We prefix the string with a single quote, a NUL,, and a double 
    # quote -- this not only forces double quotes as delimiters for the
    #  whole prefixed string, it then follows that with a backslash and 
    # double quote, forcing the initial cruft to end in a double quote.
    # The resulting string will always begin "'\0\"...  by dropping the
    " "'\0\ part of the repr, we are left with a perfeectly fine repr.
    write(repr("'\0\"" + value)[6:])
 
This saves one string creation.
Isn't fiddling code fun?

-Scott David Daniels
Scott.Daniels at Acm.Org






From mwh at python.net  Tue Apr  2 14:17:09 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 2 Apr 2002 19:17:09 GMT
Subject: Problem Calling Python callbacks from C
References: <8351bb33.0204020854.272eaa7b@posting.google.com> <m3d6xigg6a.fsf@mira.informatik.hu-berlin.de> <Pine.LNX.4.10.10204021028590.32630-100000@abl.med.utah.edu>
Message-ID: <lk7knpx64t.fsf@elios.maths.bris.ac.uk>

Tim Egbert <tegbert at abl.med.utah.edu> writes:

> I'm able to call a Python callback _function_ from C code by following the      
> information in "Extending and Embedding the Python Interpreter," section
> 1.6, but I can't get the callback to work if I'm calling a Python object
> _method_.  Does anyone know how to do that?
> 
> In other words, it works if I'm using a Python function such as:
>    
>     def callback_func(i):
>         print i
> 
> and I set "callback_func" as the callback but not if the callback is an
> object method such as:
> 
>     class callbackClass:
>         def callback_method(self,i):
>             print i
>     cbc=callbackClass()
> 
> and I set the callback as "cbc.callback_method".
> 
> I'm using the "PyEval_CallObject(callback,arglist)" function in my C code
> to make the callback.

That ought to work.

> I assume the problem has to do with not having the "self" argument
> available in the arglist for the callback method, but I haven't been
> able to find any documentation on how to provide "self" or what
> Python function to use for calling an object method.

Well, I'm afraid I haven't bought my telepathy upgrade for 2002, so
you'd better

a) post some more code
b) tell us exactly what goes wrong

Cheers,
M.


From juenglin at informatik.uni-freiburg.de  Thu Apr  4 06:35:46 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 04 Apr 2002 13:35:46 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <_vUq8.552234$pN4.39754796@bin8.nnrp.aus1.giganews.com>
Message-ID: <xdupu1fae7h.fsf@leto.informatik.uni-freiburg.de>

"Terry Reedy" <tejarex at yahoo.com> writes:

> > Clarification
> >
> >     This PEP does *not* change the fact that almost all object types
> >     can be used as truth values.  For example, when used in an if
> >     statement, an empty list is false and a non-empty one is true;
> >     this does not change and there is no plan to ever change this.
> 
> almost? all?  I am under the impression that 'if x:' is universally
> valid, and consider that one of the good features of Python's prime
> dichotomy system.

Hi,

this might have been brought up already (I didn't follow the whole
discussion): Currently, when an object x is tested as in 
'if x:' or 'if not x:'

the interpretation of x as a boolean is done by PyObject_IsTrue()
at the very end. PyObject_IsTrue() itself tries to interpret the
given object as a number, then (if this fails) as a mapping, 
finally as a sequence. Here is where the something/nothing semantic
roots in.

But what about user-defined types? 

I think it should be specified by the type itself how to interpret
its instances in a boolean context. Thus, there should be a 
further slot 'tp_as_bool', a special method '__asbool__', respectively,
and this method should be invoked by PyObject_IsTrue(). If it were
not defined, an exception was raised.

Regards,
Ralf



From tdi01fhu at syd.kth.se  Mon Apr 15 19:31:20 2002
From: tdi01fhu at syd.kth.se (FREDRIK HULDTGREN)
Date: Tue, 16 Apr 2002 01:31:20 +0200
Subject: Going through folders...
Message-ID: <scbb7efb.022@Riksvasa.syd.kth.se>

I want to go through a catalogue structure and make a textfile that lists the files in each catalogue/undercatalogue. I know that there is a function for doing this somewhere, I just can't remember what it is right now...
all help is appreciated...

/fredrik




From mgilfix at eecs.tufts.edu  Wed Apr 17 23:24:21 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 23:24:21 -0400
Subject: Pure python crypto
In-Reply-To: <7x662qqb3n.fsf@ruckus.brouhaha.com>; from phr-n2002a@nightsong.com on Wed, Apr 17, 2002 at 02:17:00PM -0700
References: <mailman.1019024313.11744.python-list@python.org> <7x3cxubvgb.fsf@ruckus.brouhaha.com> <mailman.1019053415.9589.python-list@python.org> <7x662qqb3n.fsf@ruckus.brouhaha.com>
Message-ID: <20020417232421.D924@eecs.tufts.edu>

On Wed, Apr 17 @ 14:17, Paul Rubin wrote:
> Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
> >   The nice thing about blowfish is its simplicity. I started thinking
> > about implementing DES and 3DES though and its ugly. 
> 
> Blowfish isn't so simple to implement in Python, plus the key setup
> is awfully slow.

  Well, I'll have to post it when I return home to Canada. Not a good
idea to distribute it in the US :) I can send you a copy if you're in
the US if you email me directly.

> >   I was thinking that it might be nice to have an 'ssh' module or
> > 'sftp' module. Specifically, I wanted sftp for an application and
> > couldn't find any such modules. For portability reasons, it might
> > be nice to have a pure python module, however, it would require
> > some tweaking I'm sure.
> 
> I think RC4 coded in Python would run much faster than Blowfish coded
> in Python.  SSH can use RC4.  I don't know what SFTP is.
> 
> FWIW, I coded RC4 in Python a while back and if I remember correctly,
> got about 100k/sec throughput on a P3-750.  That's sort of marginal
> for (say) sftp over ADSL.  Using the SHA-OFB hack that I described,
> I got about 400-500k/sec.  

 Do you still have a copy of this? I would be nice to assemble a suite
that can be used to create a pure python implementation of SSH (which
could also use wrappers to C implementations that exist as well).  I
wish I had a better idea of what was out there module-wise though.

 SFTP is essentially the ftp protocol written on top of SSH. Very
handy as you have all the functionality of FTP except encrypted.

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From h_schneider at marketmix.com  Mon Apr  8 06:47:59 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Mon, 8 Apr 2002 12:47:59 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com><20020408120242.41456ace.wilk@flibuste.net><a8rqh2$9h5$03$1@news.t-online.com> <20020408123251.48eab8d6.wilk@flibuste.net>
Message-ID: <a8rs5h$7na$01$1@news.t-online.com>

> When i see boa constructor i'm a litle bit afraid... but it doesn't mean
> that it's a problem of wxPython. Also when i make some tests of wxPython
> with idle or pythonwin i have often big scratch (whitout python log). But
> i have never theses scratch if i run my script from the command line.

What do you mean exactly with scratch? Crashes ?

Boa is the most complex app, I know. That was the reason for this
investigation ;)

> The opinion of an experienced user of wxPython would be welcome !

So please (you - out there) join our thread! :-)

Bye,
Harald





From magnus at hetland.org  Tue Apr 30 15:09:38 2002
From: magnus at hetland.org (Magnus Lie Hetland)
Date: Tue, 30 Apr 2002 21:09:38 +0200
Subject: Update locals() / resolved
In-Reply-To: <20020430190553.GE16217@devel.trillke>; from pyth@devel.trillke.net on Tue, Apr 30, 2002 at 09:05:53PM +0200
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org> <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no> <20020430190553.GE16217@devel.trillke>
Message-ID: <20020430210938.B3173@idi.ntnu.no>

holger krekel <pyth at devel.trillke.net>:
[snip]
> > Thes names aren't reserved in any way...
> 
> right. i do it all the time for file,str,dict,string,list...
> they are simply to short and descriptive to be left bound
> to something i don't access directly.

Sure. I wasn't arguing with you here (believe it or not ;)

> and using type('') or friends is usually a very precise way of
> expressing that you want the type object for that particular thing.

Sure. I just like using their names when I need them, I guess.

> regards,
> 
>     holger

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org




From martin at v.loewis.de  Wed Apr 17 04:30:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 10:30:50 +0200
Subject: bug in build? configuration --with-cxx ends up using PY_CFLAGS instead of CPPFLAGS
References: <ad496f8.0204170006.596b6818@posting.google.com>
Message-ID: <m3sn5uvi9x.fsf@mira.informatik.hu-berlin.de>

vvainio at tp.spt.fi (Ville Vainio) writes:

> I need to create a c++ - friendly version of python 2.2.1 (long story
> short: hp-ux11, omniORB[py]). When I do ./configure --with-cxx=aCC,
> the Makefile that gets generated has the following rule:
> 
> Modules/ccpython.o: $(srcdir)/Modules/ccpython.cc
> 	$(CXX) -c $(PY_CFLAGS) -o $@ $(srcdir)/Modules/ccpython.cc
> 
> So it uses PY_CFLAGS, while it should IMHO use CPPFLAGS (it appears to
> work properly with CPPFLAGS).
> 
> So, the Makefile.pre.in should apparently be altered? Why has this not
> been done? Would it break something on other platforms?

CPPFLAGS are just the flags for the preprocessor, so they are clearly
the wrong set of flags. Can you elaborate which flags is being passed
that should not? In any case, you need both OPT and CFLAGSFORSHARED
when compiling ccpython.

Regards,
Martin



From claird at starbase.neosoft.com  Thu Apr 11 10:42:19 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 11 Apr 2002 09:42:19 -0500
Subject: SMS Server implemented in Python
References: <k1ft8.1694$ws6.37396@news2.nokia.com>
Message-ID: <2D4B57FE40B5FE2A.08DB31F492EEFF0B.F76D2BD5D7AAECF5@lp.airnews.net>

In article <k1ft8.1694$ws6.37396 at news2.nokia.com>,
Peter Torngaard  <peter.torngaard at nokia.com> wrote:
>Hello,
>
>just happend to read Frank Stajano's paper 
>'Implementing the SMS server, or why I ....' and found that 
>the server sounded cool. Now I'm looking for it but it seems 
>like it do not want to be found! It have been on the net, 
>alot of search engines are pointing to 404 pages!
>
>My question is do anyone hold a copy? or know where it can be 
>found.
			.
			.
			.
Write directly to Frank.  He's a nice guy.
He's likely to tell you that AT&T hasn't
and won't make the server public, but that
he knows another one that is.

Even though a competitor employs you.
-- 

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


From rdsteph at earthlink.net  Sun Apr 28 21:51:05 2002
From: rdsteph at earthlink.net (Ron Stephens)
Date: Mon, 29 Apr 2002 01:51:05 GMT
Subject: 300 Latest Pythonic web articles
Message-ID: <3CCCA704.4060000@earthlink.net>

Now updated daily at http://www.awaretek.com/plf.html



From max at alcyone.com  Thu Apr  4 04:38:01 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 04 Apr 2002 01:38:01 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85>
Message-ID: <3CAC1EF9.C4A97594@alcyone.com>

James Logajan wrote:

> Too bad. I believe his name was spelled Boole, not Bool or Boolean.
> You had
> a chance to properly honor him and didn't for some reason. Why drop a
> single letter of a dead man's name? This seems to be a pettiness that
> has
> gained universal appeal, alas. Shades of "creat" versus "create".
> Maybe
> there is some conspiracy against the letter "e" on the end of words?

Could you find something better to complain about?

"Bool" is a standard abbreviation, like it or not.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From sdeibel at wingide.com  Tue Apr  9 20:05:16 2002
From: sdeibel at wingide.com (Stephan R.A. Deibel)
Date: Tue, 9 Apr 2002 20:05:16 -0400 (EDT)
Subject: Help with Python advocacy and marketing
Message-ID: <Pine.LNX.4.30.0204091918230.896-100000@pangolin.wingide.com>

Hi,

Some time ago on c.l.p.a., I announced a new mailing list for those
interested in discussing marketing Python to a wider audience:

http://pythonology.org/mailman/listinfo/marketing-python

Several projects have grown out of that and I would like to invite
the wider community of Python users to participate in these:

* We are collecting user-side advocacy quotes for use in press
  releases announcing future versions of Python.  If you have
  a compelling statement about how Python has helped your
  project or business, this is a great way to help Python and
  get some media exposure in the process.  We can't guarantee
  your quote will be used in a press release, but with your
  permission it will be included on a Python advocacy website
  that is under development.

  For more information, or to submit a quote, please email to
  python-quotes at pythonology.org

* We are working on a collection of Python Success Stories,
  which showcase Python in a specific project or setting.
  If you've used Python for a large, complex, or interesting
  project, and have compelling things to say about how Python
  has helped you, please consider being interviewed for this
  effort.  The resulting stories will be made available on
  the web, will help build credibility for Python, and can be
  a great way to get exposure for your research, product,
  or company.

  For more information, please email success-story at pythonology.org

* We are working on a collection of Python Quick Start Guides
  that are include condensed instructions and download urls that
  help new users of Python get started on solving specific
  kinds of problems.  Examples might include "Writing a GUI
  with wxPython", "Writing Python CGIs", "Using Jython to
  Test Java Libraries", and so forth.

  For more information about contributing to this effort, please
  email quick-start at pythonology.org

* Future versions of Python will be announced in the technical
  media with formal press releases.  If you have media contacts
  that might be interested in receiving and/or redistributing
  press releases, please send email to python-press at pythonology.org

All of these efforts are strictly on a volunteer basis.  We do
not have funds to support any of them at this time.  But these
are great ways to give back to Python if you don't already
contribute to the language's development in some way (or even
if you do!).

Please direct discussion of this topic to the marketing-python
mailing list, which can be joined with the URL given above.

Thanks,

- Stephan

------------------------------------------------------------------------
Wing IDE for Python                          Archaeopteryx Software, Inc
www.wingide.com                              Take Flight!












From thomas at gatsoft.no  Tue Apr 16 10:07:11 2002
From: thomas at gatsoft.no (Thomas Weholt)
Date: Tue, 16 Apr 2002 14:07:11 GMT
Subject: Backing up mail in Outlook Express using Python
References: <cGSu8.101$1kc.179320832@news.telia.no> <JCUu8.23257$Ji7.8640@atlpnn01.usenetserver.com>
Message-ID: <jeWu8.104$1kc.179321856@news.telia.no>

Well, by searching the net a bit more I've found some pointers and as soon
as I found out where the mail was actually stored and that it's stored in
plain files, backing them up was easy.

Thanks anyway :-)

Best regards,
Thomas Weholt


"Steve Holden" <sholden at holdenweb.com> wrote in message
news:JCUu8.23257$Ji7.8640 at atlpnn01.usenetserver.com...
> "Thomas Weholt" <thomas at gatsoft.no> wrote in message
> news:cGSu8.101$1kc.179320832 at news.telia.no...
> > I need to backup all my mail stored in Outlook Express. Can I use Python
> to
> > do this?
> >
> No reason why not, though I'm not sure why you'd want to, since the
message
> are just stored in files.
>
> You can find out where they are from the Maintenance tab on Tools |
Options
> (click the "Store Folder" button).
>
> I did adapt some code from Delphi to read the mailbox files, and can let
you
> have that if it would help, but sounds to me like a file copy is all
you're
> looking for.
>
> regards
>  Steve
>
>
>
>




From martin at v.loewis.de  Sun Apr 14 05:22:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 11:22:52 +0200
Subject: Weak references and tracebacks - NEVER MIND
References: <a9albr$3ha$1@peabody.colorado.edu> <a9altu$3rp$1@peabody.colorado.edu>
Message-ID: <m3pu12ws5v.fsf@mira.informatik.hu-berlin.de>

Fernando P?rez <fperez528 at yahoo.com> writes:

> > Is it ok to use weak references to hold tracebacks without generating
> > circular references? From the documentation:
> 
> Never mind, you simply can't make those weak refs. Sorry (tested with other 
> types. dumb me).

Even if it was possible, it would be less useful than you think: Most
likely, your weak reference would be the *only* reference to the
traceback object, which in turns would cause the traceback to die, and
your weak reference go stale immediately.

Regards,
Martin



From fperez528 at yahoo.com  Wed Apr 10 20:01:09 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 10 Apr 2002 18:01:09 -0600
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu>
Message-ID: <a92jp9$cto$1@peabody.colorado.edu>

Rajarshi Guha wrote:

> Hi,
>   I have a list like this:  l = [1,'s', 2,'t', 3,'d']
> 
> I'd like to loop through the list but within the loop I want
> the current element and the element just after it:
> I tried this
> 
> liter = iter(l)
> for i in l:
>   print i, liter.next()

This works (notice you need to iterate over the _iterator_ for this to work):

In [9]: liter = iter(l)

In [10]: for a in liter:
   ....:   print a,liter.next()
   ....:
1 s
2 t
3 d

cheers,

f.


From hove at phys.ntnu.no  Mon Apr 29 03:23:14 2002
From: hove at phys.ntnu.no (Joakim Hove)
Date: 29 Apr 2002 09:23:14 +0200
Subject: Converting a Python List to a C double *
References: <k0n4rhy3gtn.fsf@metropolis.phys.ntnu.no> <jPfy8.65668$vF6.2006478@news2.tin.it>
Message-ID: <k0nadrn3r4t.fsf@metropolis.phys.ntnu.no>

Aleax,

thank you very much for a detailed suggestion. It works just as I
intended :-)

Joakim


-- 
==== Joakim Hove      www.phys.ntnu.no/~hove/   =======================
|| Institutt for fysikk  (735) 93637 / E3-166  |  Sk?yensgate 10D    ||
|| N - 7491 Trondheim    hove at phys.ntnu.no     |  N - 7030 Trondheim ||
================================================= 73 93 31 68 =========


From op73418 at mail.telepac.pt  Tue Apr  2 19:58:00 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 03 Apr 2002 01:58:00 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <l9ikau855v1gghv7vl4djvmh7aks0jad2n@4ax.com>

On Sat, 30 Mar 2002 00:39:10 -0500, Guido van Rossum <guido at python.org>
wrote:

>I offer the following PEP for review by the community.  If it receives
>a favorable response, it will be implemented in Python 2.3.
>
>A long discussion has already been held in python-dev about this PEP;
>most things you could bring up have already been brought up there.
>The head of the thread there is:
>
>    http://mail.python.org/pipermail/python-dev/2002-March/020750.html
>
>I believe that the review questions listed near the beginning of the
>PEP are the main unresolved issues from that discussion.
>
>This PEP is also on the web, of course, at:
>
>    http://python.org/peps/pep-0285.html
>
>If you prefer to look at code, here's a reasonably complete
>implementation (in C; it may be slightly out of date relative to the
>current CVS):
>
>    http://python.org/sf/528022
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)
>
>PEP: 285
>Title: Adding a bool type
>Version: $Revision: 1.12 $
>Last-Modified: $Date: 2002/03/30 05:37:02 $
>Author: guido at python.org (Guido van Rossum)
>Status: Draft
>Type: Standards Track
>Created: 8-Mar-2002
>Python-Version: 2.3
>Post-History: 8-Mar-2002, 30-Mar-2002
>
>
>Abstract
>
>    This PEP proposes the introduction of a new built-in type, bool,
>    with two constants, False and True.  The bool type would be a
>    straightforward subtype (in C) of the int type, and the values
>    False and True would behave like 0 and 1 in most respects (for
>    example, False==0 and True==1 would be true) except repr() and
>    str().  All built-in operations that conceptually return a Boolean
>    result will be changed to return False or True instead of 0 or 1;
>    for example, comparisons, the "not" operator, and predicates like
>    isinstance().
>
>
>Review
>
>    Dear reviewers:
>
>    I'm particularly interested in hearing your opinion about the
>    following three issues:
>
>    1) Should this PEP be accepted at all.

-0 on this. Generally, I think having a bool type is a Good Thing, but
the problem is that Python is already in version 2.2 and lots of code
has been written. So, in order to preserve backwards-compatibility some
choices have to be made. The real damage lies in

A) True == 1

The integer 1 and the boolean value True are distinct things. They
belong to different universes and as such they should have different
types => Asking if True is equal to 1 is just nonsense. A side-effect of
this is that Boolean arithmetic (that is, the +, *, etc. operations when
viewing the boolean algebra {0, 1} as the order 2 field) cannot be
implemented, e.g. True + True should be False yet because of A) this
cannot be, or else we'd get False == 2.

In fact, now that I think of it, in no circumstance we can have Boolean
arithmetic + the requisite that

B) True + 1 = 2

and such similar things (e.g. via __add__ and __radd__ checking for the
presence of integers in their arguments and making an implicit cast of
bool into int) Because in this case the expression

True + True + 1

is ambiguous (it depends on the parenthesation). And since B) is the
more important, boolean arithmetic is gone with the wind. This leaves me
uneasy.

Anyway, what we end up with is half-a-boolean, an integer in disguise.
Is it really worth adding extra baggage - and all the ensuing confusion
- for the few perceivable gains? I think not.

What I think it might be useful is to make a full blown bool type in a
module for those that need it.

My 2 cents on the subject, anyway.

Best regards,
Gon?alo Rodrigues


From w3miller at yahoo.com  Tue Apr  9 02:01:54 2002
From: w3miller at yahoo.com (Walter Miller)
Date: Mon, 8 Apr 2002 23:01:54 -0700 (PDT)
Subject: SSL error: 2.1.1 works but 2.2 doesn't 
In-Reply-To: <200204090557.g395vqE02023@mbuna.arbhome.com.au>
Message-ID: <20020409060154.59291.qmail@web21410.mail.yahoo.com>

I tried 2.2.1rc2 with the same results.  I haven't
tried 2.1.3.

--- Anthony Baxter <anthony at interlink.com.au> wrote:
> 
> >>> Walter Miller wrote
> > Apologies if this has been discussed/solved
> already. 
> > I looked for prior postings but haven't found a
> > solution for fixing this error I'm getting with
> > _socket.pyd on 2.2.
> 
> Does it work with 2.1.3? 2.2.1rc2? 
> 
> 
> -- 
> Anthony Baxter     <anthony at interlink.com.au>   
> It's never too late to have a happy childhood.
> 


__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/




From skip at pobox.com  Mon Apr 15 15:32:47 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 15 Apr 2002 14:32:47 -0500
Subject: input() vs raw_input()
In-Reply-To: <slrnabm6im.a8q.ak@ak.silmarill.org>
References: <slrnabm6im.a8q.ak@ak.silmarill.org>
Message-ID: <15547.10975.286697.85984@12-248-41-177.client.attbi.com>

    Andrei> input() must not be mentioned in any tutorials for new users. It
    Andrei> belongs somewhere in the depths of the thickest reference
    Andrei> manual. Where do new users *learn* about it? I don't remember
    Andrei> seeing it in any tutorial and yet it pops up all the time.

Maybe it's mentioned in some books.  It's right there in black-and-white in
the documentation of the builtin functions as well

    http://www.python.org/doc/current/lib/built-in-funcs.html

though it does warn people against using it.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From mwh at python.net  Sat Apr 13 08:04:18 2002
From: mwh at python.net (Michael Hudson)
Date: Sat, 13 Apr 2002 12:04:18 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <Xns91EF2A8A1CF47RASXnewsDFE1@130.133.1.4>
Message-ID: <lk7knbiyw4.fsf@pc150.maths.bris.ac.uk>

Philip Swartzleonard <starx at pacbell.net> writes:

> It appears that python only kees one copy of any given int around, 

Only for ints in the range 0...100, I think (might be -10...100 or
something now).

> and any name that happens to gain the status of being that integer
> really just gets another copy of it's pointer.

Yep.  This happens for the empty tuple, the empty string and one
character strings too (and maybe some other I forget now).

> It seems like it'd be less effort, but...

This is a performance & memory usage issue.

Cheers,
M.

-- 
  C is not clean -- the language has _many_ gotchas and traps, and
  although its semantics are _simple_ in some sense, it is not any
  cleaner than the assembly-language design it is based on.
                                        -- Erik Naggum, comp.lang.lisp


From tejarex at yahoo.com  Tue Apr  9 10:44:17 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 09 Apr 2002 14:44:17 GMT
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com>
Message-ID: <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com>

"Dennis" <dennis at horsethieves.com> wrote in message
news:c4a65c5f.0204090624.710228ff at posting.google.com...
> If there's a better forum for asking newb questions, please point me
> there!

There is (or at least has been) a beginner tutorial email list: see
python.org

> I'm new to programming.  My only experience up to this point is some
> aborted attempts to learn C and Java.  I'm finding Python much
easier
> to learn than those.

Welcome to the club.

>  Currently, I'm pushing my way through
> "Non-Programmers Tutorial for Python" by Josh Cogliati.
> My current problem is this:
>
> x = 2
>
> def wrong(num):
>     return num + 1
>
> print wrong(x)
> print wrong(x)
>
> I cannot figure out why this (simplified) program won't increment x
> and produce 3 and 4.

Ask yourself why you expect it to and then tell yourself, "There is
something wrong with that reasoning."  (Hint: num+1 generates a new,
unnamed object.)

>  This is probably indicative of a bigger problem
> with understanding functions (and their limitations).

Keep experimenting until you understand better.  Put print statements
inside functions.

Terry J. Reedy





From tim.one at comcast.net  Sat Apr 13 22:53:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 22:53:11 -0400
Subject: 2.1 vs. 2.2
In-Reply-To: <3cb8c8d5$0$12708$9b622d9e@news.freenet.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPNONAA.tim.one@comcast.net>

[Jens Baader]
> BTW, try this in interactive mode:
>
> "The Python interpreter is very stable"*30000
>
> At least at my system the interpreter promptly
> hangs itself.


> ...
> It works well if you use the command
> line interface so the bug seem to be related to
> IDLE (I'm using Python 2.2.1 with WinXP)

Ah.  It's not a bug, but if you want to call it one you'll have to talk to
the Tcl/Tk people:  the Tk text widget has severe performance problems with
very long lines (and this line is about a megabyte).  Note that it's not
hung:  if you wait long enough, it will eventually display the whole thing,
chewing up close to 100% of your CPU until that happens.  I'm trying it
right now under IDLE, and so far it's consumed about 18 CPU minutes.

looks-like-i'll-be-staying-up-late<wink>-ly y'rs  - tim

PS:  Oooh!  It finished!  It took 20 CPU minutes and 32 CPU seconds on
     an 866 MHz Win98SE box.





From fperez528 at yahoo.com  Sat Apr  6 15:18:17 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 06 Apr 2002 13:18:17 -0700
Subject: Reading large multidimensional arrays off file
References: <87k7rksjeg.fsf@sheffield.ac.uk>
Message-ID: <a8nl6d$lb$1@peabody.colorado.edu>

 
>         The previous process takes ages to read the file in (of the
> order of an hour for one 13 Mb file), so I was wondering why I don't
> just read the whole array in, after all, I know that it is NxMx10
> bytes, and I know the order. My question is fairly basic: How do I
> specify the order of the data? I would like to read it in the order
> shown above, into an 3-dimensional array size (10,N,M), but I am not
> sure what sizes come first or second in Python.

See if this helps as a sample start:

    def read(self,fname):
        """Read in a topological density file.

        Does NOT do byte-reversal.
        """

        # MILC code:
        TOPO_VERSION = 66051
        # of bytes to skip in file *from the beginning* before data starts
        offset = 108

        topo = open(fname)
        topo_version,nx,ny,nz,nt = struct.unpack('i'*5,topo.read(5*4))
        if topo_version != TOPO_VERSION:
            warn('Wrong topological density version number (or 
byte-reversal)!',3)

        topo.seek(offset)
        data_size = nx*ny*nz*nt*4  # these files use C float, not double
        topo_flat = fromstring(topo.read(data_size),Float32)
        topo_flat.shape = (nt,nz,ny,nx)
        with(self,nx=nx,ny=ny,nz=nz,nt=nt,data=topo_flat)
        print 'Read',data_size/4,'data points. Shape:',(nx,ny,nz,nt)


It reads binary files which represent 4-dimensional C float arrays. The key 
calls are fromstring() (there's a from Numeric import * somewhere in the 
code) and the .shape assignment later. You read the data 'flat' first, then 
make the shape assignments.

My files aren't as big as yours (under a megabyte) but the read time is 
negligible for me, so I'm sure it would be far less than one hour for a 13mb 
file.

Here's another similar reader but for big text files (~16 MB) where I need to 
scan for a tag to extract the data. The trick there is to have awk 
(super-fast) do the first pass:

    def read(self,fname,mode_num):
        """Read in a text density file.

        Assumes nx=ny=nz (but not necessarily =nt)."""

          # read nx,nt first so we know the size of the data to load
        for line in file(fname):
            if line.startswith('nx'):
                nx = ny = nz = int(line.strip().split()[1])
            if line.startswith('nt'):
                nt = int(line.strip().split()[1])
                break

        cmd = "awk '/%sPG5P/ { print $6}' %s" % (mode_num,fname)
        tfile = os.popen(cmd)
        data_a_flat = array(map(float,tfile))
        data_a_flat.shape = (nt,nz,ny,nx)

        with(self,nx=nx,ny=ny,nz=nz,nt=nt,data_a=data_a_flat)
        data_size = nx*ny*nz*nt
        print 'Read',data_size,'data points. Shape:',(nx,ny,nz,nt)

This takes 2-3 seconds to scan a 16 mb text file and extract about 1 mb of 
data from it. Let awk do fast the main scanning, then map float() onto the 
resulting in-memory string and presto, you have your array.

Hope this helps to get you started.

Cheers,

f.


From dkuhlman at rexx.com  Mon Apr 29 15:45:04 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Mon, 29 Apr 2002 12:45:04 -0700
Subject: python2.2
References: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net> <slrnacr3io.nil.gerhard@lilith.my-fqdn.de> <A7hz8.22$Uw4.4633@newsread1.prod.itd.earthlink.net>
Message-ID: <aak7qv$bf46e$1@ID-139865.news.dfncis.de>

The testing version is woody.  So change "stable" to "woody" in 
your sources.list.  (Perhaps you could instead change "stable" to 
"testing", but I'm not sure about that one.)

Not to distract you, but ... If you have gcc on your system, 
downloading, building, and installing the source is very straight 
forward.  Just:

  1. Down load the source from www.python.org.
  2. Unroll it with something like "tar xvzf ..."
  3. Then do:

    ./configure
    make
    make install     # as root

This is what I do on my Debian system.  Takes a few minutes, but 
it's painless and works great.  The README file will lead you 
through it.

And, now you can get the latest Python sources (2.2.1).

  - Dave

Billy Ng wrote:

> This is my source.list, would you please tell me how to change
> this file to allow me to install the testing packages?
> 
> deb ftp://ftp.debian.org/debian/ stable main
> deb-src ftp://ftp.debian.org/debian/ stable main
> deb http://non-us.debian.org/debian-non-US stable/non-US main
> deb-src http://non-us.debian.org/debian-non-US stable/non-US main
> 
> deb ftp://ftp.debian.org/debian/ stable main
> deb-src ftp://ftp.debian.org/debian/ stable main
> deb http://non-us.debian.org/debian-non-US stable/non-US main
> deb-src http://non-us.debian.org/debian-non-US stable/non-US main
> 
> Thanks!
> 
> Billy Ng

[snip]

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From aleax at aleax.it  Mon Apr 15 17:22:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 21:22:50 GMT
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org> <3cbb355a$0$166$9b622d9e@news.freenet.de>
Message-ID: <KwHu8.23421$vF6.768479@news2.tin.it>

merman wrote:
        ...
> Ok Brian, and how can I write a disutils setup script? I don't know that!

Try http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66509 --
the last part of the recipe gives a setup.py script that should be easy to
adapt to your needs, I hope.


Alex



From gbreed at cix.compulink.co.uk  Tue Apr  9 06:52:40 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Tue, 9 Apr 2002 10:52:40 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk>
Message-ID: <a8uh5p$nu7$1@thorium.cix.co.uk>

Robin Becker wrote:

> This article in salon 
>         http://www.salon.com/tech/feature/2002/04/08/lehman/print.html
> 
> about Manny Lehman, my old department head, gives an insight into why
> people like myself are often afraid of software change.

And why Python is an exception

"""
Michael Godfrey, a University of Waterloo scientist, is equally hesitant 
but still finds the Lehman approach useful. In 2000, Godfrey and a fellow 
Waterloo researcher, Qiang Tu, released a study showing that several 
open-source software programs, including the Linux kernel and fetchmail, 
were growing at geometric rates, breaking the inverse squared barrier 
constraining most traditionally built programs. 
"""


                                 Graham


From deathfox at moochercrew.org  Sat Apr 27 01:00:01 2002
From: deathfox at moochercrew.org (Chris Emery)
Date: Fri, 26 Apr 2002 22:00:01 -0700
Subject: reload problem
Message-ID: <20020426220001.A13373@arisu.moochercrew.org>

I am atempting to do something like the autoreloader, that will work for
some threaded objects I'm dealing with, which in 2.2.1 autoreloader
doesn't seem to like.  I figure if I update each instance of the class,
it will work properly.  Unfortunatly, I seem to be running into errors.
Any advice would be greatly welcomed, especially if I'm doing this the
wrong way, attached are the two files, and below is the output from my
session.

>>> import auto
>>> a = auto.auto(10)
>>> a.reload()
auto.pyc
0
stat <module 'stat' from '/usr/lib/python2.2/stat.pyc'>
__future__ <module '__future__' from '/usr/lib/python2.2/__future__.pyc'>
copy_reg <module 'copy_reg' from '/usr/lib/python2.2/copy_reg.pyc'>
posixpath <module 'posixpath' from '/usr/lib/python2.2/posixpath.pyc'>
auto <module 'auto' from 'auto.py'>
UserDict <module 'UserDict' from '/usr/lib/python2.2/UserDict.pyc'>
signal <module 'signal' (built-in)>
time <module 'time' from '/usr/lib/python2.2/lib-dynload/time.so'>
site <module 'site' from '/usr/lib/python2.2/site.pyc'>
__builtin__ <module '__builtin__' (built-in)>
sys <module 'sys' (built-in)>
atexit <module 'atexit' from '/usr/lib/python2.2/atexit.pyc'>
reloadtmp <module 'reloadtmp' from 'reloadtmp.py'>
posix <module 'posix' (built-in)>
types <module 'types' from '/usr/lib/python2.2/types.pyc'>
__main__ <module '__main__' (built-in)>
exceptions <module 'exceptions' (built-in)>
readline <module 'readline' from '/usr/lib/python2.2/lib-dynload/readline.so'>
os <module 'os' from '/usr/lib/python2.2/os.pyc'>
rlcompleter <module 'rlcompleter' from '/usr/lib/python2.2/rlcompleter.pyc'>
os.path <module 'posixpath' from '/usr/lib/python2.2/posixpath.pyc'>
<reloadtmp.reload object at 0x81541ec> auto

It appears that reload, is reloading it's own module, even if I pass in the fact that it's suposed to reload the auto module, any idea of why this is happening?

Chris
-------------- next part --------------
import time
import sys
import reloadtmp

class auto(object, reloadtmp.reload):
	"Thisis another doc string"
	def __init__(self, x):
		"""Yet another doc string"""
		self.x = x
		self.d = 0
		self.rinit()

	def count(self):
		"""This is a doc string"""
		self.d = 0
		for a in range(1, self.x + 1):
			self.d += a
			#time.sleep(1)
			print 'test'

	def reload(self):
		reloadtmp.reload.reload(self, "auto")
-------------- next part --------------
import os
import sys

class reload(object):
	def rinit(self):
		self.file = str(self.__module__ + ".pyc")
		lclass = str(self.__class__)
		lclass = lclass.split()
		lclass = lclass[1]
		lclass = lclass[1:-2]
		self.classname = lclass.split(".")

	def reload(self, modname):
		print self.file
		print os.stat(self.file[:-1])[8] >= os.stat(self.file)[8]
		mod = reload(sys.modules[modname])
		for a,b in sys.modules.items():
			print a, b
		print mod, self.__module__
		if os.stat(self.file[:-1])[8] >= os.stat(self.file)[8]:
			print mod
			for a in mod.__dict__:
				print a
			sys.stdout.flush()
			#for name, object in mod.__dict__[self.classname[1]].__dict__.items():
			#	self.__dict__[name] =  object
			#	print name, object
		

From dalke at dalkescientific.com  Sat Apr 20 05:16:51 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 03:16:51 -0600
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org>
Message-ID: <a9rbi7$fr2$1@slb4.atl.mindspring.net>

[Donn Cave]
> By the way, I'll second Johannes Stiehler's recommendation of
> MxTextTools.  Definitely appropriate for SGML parsing, and much
> better than regexps for extensive parsing in my opinion - not
> just in terms of speed, but I suspect a more powerful way to
> describe text patterns than regexps.

Uncle Tim:
> Yes, it is.  "More convenient" is arguable, though -- there's a steep
> learning curve, but then people often forget how hard it was to learn
regexp
> syntax and pragmatics too.

Let's see.  Regexps.  I started learning them because of archie
searches.  I treated them as modified versions of file globs, and
slowly picked up a few other things over a couple months.

I then took a theory of computer languages course, which went into
DFAs, NFAs, and several other As.  I then had to learn how Perl did
it, which was different from grep, which was different from ....

So, figure three to four months, on and off.

Now, mxTextTools.  Three attempts.  First two failures took a week,
thereabouts.  The final success took another couple weeks to get a
working regexp parser (yes, on top of mxTextTools) going.  Let's
call it three weeks.  And this was after knowing a lot about how
those As work.

But I gotta say, mxTextTools is amazingly fast and flexible.  On
one test it was only 50% slower than fgrep for a simple text search.
(Not quite a direct comparison because the Python code was doing
other things as well.)

                    Andrew
                    dalke at dalkescientific.com





From donn at u.washington.edu  Mon Apr  8 15:31:49 2002
From: donn at u.washington.edu (Donn Cave)
Date: 8 Apr 2002 19:31:49 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com> <a8slgr$pin$1@nyheter.chalmers.se> <3CB1E745.EDA08705@alcyone.com>
Message-ID: <a8sr75$rug$1@nntp6.u.washington.edu>

Quoth Erik Max Francis <max at alcyone.com>:
| Jacob Hallen wrote:
|
| > I have been brought up in an environment which called anything
| > allowing mixed types in a list weakly typed.
| > 
| > If you prefer to call it dynamically typed, I don't mind.
|
| Weak typing usually refers to the idea that operations, not the objects
| themselves, decide as what types to manipulate objects.  Dynamic typing
| indicates that variables (or bindings) do not have an intrinsic type.

At best that seems confusing to me.  It would seem to say that
languages that use type inference would be weakly typed - they
sure let the operations decide object type.  In Objective Caml:

  # let oneup x = x + 1;;
  val oneup : int -> int = <fun>

  (i.e., because parameter x appears in the expression x + 1, x must
   have type int, and the function result must also be of type int.)

But Objective Caml's typing is hardly "weak" in any useful sense of
the word, rather it's quite rigorous.

| Python is strongly typed, since objects have intrinsic type, and those
| types are what determine how operations behave, rather than the reverse.

Maybe it would be best to just concede that these terms aren't useful.
I don't think I've ever seen anyone use them to shed any light on
Python programming, they're just occasions for academic quibbles that
have little to do with the substance of what people are saying.  Even
worse than "pass by reference" and "pass by value."

	Donn Cave, donn at u.washington.edu


From mickey at tm.informatik.uni-frankfurt.de  Sat Apr  6 08:05:18 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 6 Apr 2002 15:05:18 +0200
Subject: wxPython Book ?
References: <11e94203.0204031526.f11bfe6@posting.google.com>
Message-ID: <3caef28e@nntp.server.uni-frankfurt.de>

Jose Rivera <jriveramerla at yahoo.com> wrote:
> Hi...
> Is there a wxPython book available? that gives a depth tutorial?

<shameless self-promotion>

There is a german book "Python und GUI-Toolkits", which implements
a contact database with four GUI-Toolkits (namely Tk, Qt, PyGTK 2 and
wxWindows) which might be of interest - but it's targetted at the
intermediate Python Programmer.

It will be out in May or early June.

</shameless self-promotion>

Yours,

Mickey.

:M:
--
|----------------------------------------------------------------------------|
| Dipl.-Inf. Michael 'Mickey' Lauer    mickey at tm.informatik.uni-frankfurt.de |
|   Raum 10b - ++49 69 798 28358        Fachbereich Informatik und Biologie  |
|----------------------------------------------------------------------------|


From donald.braman at yale.edu  Thu Apr  4 20:37:57 2002
From: donald.braman at yale.edu (don)
Date: Thu, 4 Apr 2002 17:37:57 -0800
Subject: rowcount in dbi w/ odbc
Message-ID: <a8iv0i$56r$1@news.ycc.yale.edu>

I'm new to Python programming. I am trying to gain access to specific
columns in specific rows in an Access database, but am not having much luck.
I thought I'd first try to access specific rows and columns. However, when I
try to use "rowcount" to see how many rows there are, I get "AttributeError:
rowcount". I assume there is some easy way to do this, but I can't seem to
figure it out. Here's my attempt that fails:

import dbi, odbc
try:
        con = odbc.odbc('mydatabase')
        cur = con.cursor()
        cur.execute('SELECT * FROM mytable')
        con.close()

        for row in range(0, cur.rowcount()-1):
                for col in range(0, cur.colCount()-1):
                        print cur.getField(row, col), ',',
                print cur.getField(row, col)

except NameError,e:
        print 'error ', e, 'undefined'


I suppose I should note that I can access the data using the following, but
that it doesn't give me access to a specific rows:

import dbi, odbc
try:
    s = odbc.odbc('mydatabase')
    cur = s.cursor()
    cur.execute('SELECT * FROM mytable)
    print cur.description
    for tup in cur.description:
        print tup[0],
    print
    while 1:
        rec = cur.fetchmany(1)
        if not rec: break
        print rec

except NameError,e:
        print 'error ', e, 'undefined'

-don




From laytonjb at bellsouth.net  Sat Apr 13 08:16:41 2002
From: laytonjb at bellsouth.net (Jeff Layton)
Date: Sat, 13 Apr 2002 08:16:41 -0400
Subject: Newbie: Background processing with a GUI?
Message-ID: <3CB821A8.AA3274AA@bellsouth.net>

Good morning,

   I'm not sure the subject is entirely accurate, but I really don't
know what to call what I want to do (that's really clear).
   What I want to do is have a GUI (right now I have it written
in wxPython) that displays some data in a grid (spreadsheet
kind of thing) but is updated every x seconds. I've got the classes
written that get the data and put it into a list. I've also got the
classes written that create the GUI (with some attempts at
methods that update the values in the grid). However, I don't
know how to update the grid every x seconds. Any pointers or
insight?
   Another question that goes along with this is, how do I do this
but allow the user to change the time between updates x?

TIA,

Jeff





From andymac at bullseye.apana.org.au  Sat Apr  6 18:26:47 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sun, 7 Apr 2002 10:26:47 +1100 (edt)
Subject: 2.2.1c2 test_fcntl hangs
In-Reply-To: <lysn69qrp7.fsf@bock.chem.unc.edu>
Message-ID: <Pine.OS2.4.32.0204071013560.6070-100000@tenring.andymac.org>

On 6 Apr 2002, Jochen [iso-8859-1] K?pper wrote:

> With latest cvs release22-maint (April 6, 1am EST) test_fcntl hangs.
> (I couldn't find any postings about this.  I think I have seen this
> before myself, though...)

I don't believe there have been any recent changes that might have
provoked this - at least I don't remember any checkins on the 2.2
maintenance branch (or the head branch for that matter) that would seem
plausible as causes (not that I'd rely too much on my memory either).

If you have succeeded in building from the 2.2 maintenance branch before,
what was the checkout time of the last successfully built tree?

Have you changed anything else on your system in similar timeframe - such
as glibc?

Please report this to the SF bug tracker.  I'd suggest assigning it to
Michael Hudson, as 2.2.1 patch czar.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From martin at v.loewis.de  Fri Apr 26 12:53:16 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 26 Apr 2002 18:53:16 +0200
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <slrnacgag6.98m.akuchlin@ute.mems-exchange.org> <aaaknh$3uk$1@panix2.panix.com> <m3r8l3roiv.fsf@mira.informatik.hu-berlin.de> <aabubi$1le$1@panix3.panix.com>
Message-ID: <m3hely4d1f.fsf@mira.informatik.hu-berlin.de>

aahz at pythoncraft.com (Aahz) writes:

> http://py-howto.sourceforge.net/ says
> 
>     Python HOWTOs are documents that cover a single, specific topic, and
>     attempt to cover it fairly completely. Modelled on the Linux
>     Documentation Project's HOWTO collection, this collection is an
>     effort to foster documentation that's more detailed than the Python
>     Library Reference.

That is what the page says. I recall Andrew saying that he'd prefer to
get as many of these documents off the project, since he'll shut it
down over time.

Regards,
Martin



From donn at u.washington.edu  Mon Apr 22 11:55:55 2002
From: donn at u.washington.edu (Donn Cave)
Date: 22 Apr 2002 15:55:55 GMT
Subject: stdin - seek() issue on Linux.
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com> <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com>
Message-ID: <aa1bqb$lek$1@nntp6.u.washington.edu>

Quoth "Terry Reedy" <tejarex at yahoo.com>:
| "Srihari Vijayaraghavan" <no at spam.please> wrote in message
| news:WMTw8.46875$uR5.106037 at newsfeeds.bigpond.com...
|> I am a new-bie python user. I have Python 2.2.1 installed on a Linux
|> computer. I get the following error message when I try to seek() in the
|> stdin, like:
|> #cat /var/log/messages | python test.py
|> (where /var/log/messages is over 1MB in size)
|>
|> Traceback (most recent call last):
|>   File "test.py", line 3, in ?
|>     sys.stdin.seek(-512, 2)
|> IOError: [Errno 29] Illegal seek
|
| By 'cat'ting the random access disk file, you turned it into a
| forward-access-only stream of bytes.  You definitely cannot seek
| backwards (and maybe not forward - don't know).  Instead, open() the
| file and use the resulting file object:
|
| msgs = open("/var/log/messages", 'r')
| msgs.seek(-512, 2)
| ...

Right, though it can be fixed without changing test.py at all -
"python test.py < /var/log/messages" will work fine.  This idiom
("cat file | ...") used to regularly receive a UUOC award, "Useless
Use Of Cat", from notorious Perl hacker Randall Schwartz when I
followed comp.unix.shell a few years back.

Of course if the real program uses some filter, instead of "cat",
then there's no solution - you can't seek in a pipe, so the answer
has to be "don't do that".

	Donn Cave, donn at u.washington.edu


From jeff at ccvcorp.com  Tue Apr 30 15:37:34 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 30 Apr 2002 12:37:34 -0700
Subject: wxHtml - newbie
References: <aajv7m$3hl$1@wanadoo.fr>
Message-ID: <MPG.1738925df3af14349896cc@news.nwlink.com>

In article <aajv7m$3hl$1 at wanadoo.fr>, emmanuelgagniere at wanadoo.fr 
says...
> Hello,
> 
> Do you know where can I find an example of a simple application using
> wxHtmlWindow?

Have you looked at the demo provided with wxPython?  It is a 
marvelous source of example code for every aspect of wxPython, 
and is frequently clearer than reading the annotated wxWindows 
docs.  Consulting both of those (docs + demo) will answer about 
90% of wxPython questions.  For the others, consider joining the 
wxpython-users mailing list (from wxpython.org), where all sorts 
of questions are quickly answered.  :)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From roy at panix.com  Wed Apr 17 12:06:48 2002
From: roy at panix.com (Roy Smith)
Date: 17 Apr 2002 12:06:48 -0400
Subject: slow gdbm performance (doesn't honor "f" flag?)
Message-ID: <a9k6io$704$1@panix2.panix.com>

I'm building a bunch of gdbm files with a fairly large amount of data.
The original text file is about 17 Mbytes which ends up being
about 50,000 gdbm records.  I'm using RedHat 6.2, and Python 2.1.2.

I open the gdbm files with:

db = gdbm.open (fileName, 'nf')

My understanding of the 'f' flag is that it would do everything in
memory and only flush data out to the disk when I did a close() or
sync(), but it doesn't seem to work that way.  My application takes
about 5 minutes to run, during which time the gdbm files keep getting
updated, and vmstat shows mostly idle CPU and lots of disk writes.

Is the gdbm module ignoring the 'f' flag, or does it just not work
quite the way I envisioned, or am I simply doing something wrong? :-)


From jepler at unpythonic.net  Mon Apr 29 14:20:06 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 29 Apr 2002 13:20:06 -0500
Subject: Tkinter Text widget reading input (how to do overwrite mode?)
In-Reply-To: <20020429100034.A12526@unpythonic.net>
References: <13285ea2.0204290624.4fbe8cba@posting.google.com> <20020429100034.A12526@unpythonic.net>
Message-ID: <20020429132005.B12526@unpythonic.net>

On Mon, Apr 29, 2002 at 10:00:39AM -0500, Jeff Epler wrote:
> I have implemented an 'overwrite mode' in the Text widget in tcl.
> Conversion to Python is left as an exercise to the reader.

Okay, I can't stand an unfinished exercise.  In my "testing", I decided
that it could stand to be a little smarter, in particular when the
insertion point is at the end of a line and I'm in overwrite mode, I don't
expect the newline to be removed to make room for the new character.

Since arbitrary new attributes can be hung on a Python class, there's no
fooling around with a fake tag on the text widget .. yay Python!

Things still behave slightly oddly when the widget contains the
selection.  If the overlined characters are selected and | shows the
insertion point, here's what happens when you overwrite "x":
        ____
    abc|defghijk
becomes
    abcx|ijk
This suggests that a test should be made for whether "nextchar" lies
inside the selection -- if so, the step inside Text's KeyPress binding
which deletes the selected text will delete it for us.

# --- overwritetext.py
import Tkinter

def comparable_index(index):
    return map(int, index.split("."))

class OverwriteText(Tkinter.Text):
    def __init__(self, *args, **kw):
	Tkinter.Text.__init__(self, *args, **kw)
	tags = list(self.bindtags())
	idx = tags.index("Text")
	tags[idx:idx] = ["OverwriteText"]
	self.bindtags(tuple(tags))
	self.is_overwrite = 0

	if not self.bind_class("OverwriteText"):
	    self._init_bindings()

    def _init_bindings(self):
	self.bind_class("OverwriteText", "<KeyPress>",
	    lambda evt: OverwriteText.KeyPress(evt.widget, evt))
	self.bind_class("OverwriteText", "<Insert>",
	    lambda evt: OverwriteText.Insert(evt.widget, evt))

    def toggle_overwrite(self):
	self.is_overwrite = not self.is_overwrite

    def KeyPress(self, evt):
	if self.is_overwrite:
	    eol = self.index("insert lineend")
	    nextchar = self.index("insert + %dc" % len(evt.char))
	    eol_num = comparable_index(eol)
	    nextchar_num = comparable_index(nextchar)
	    if eol_num < nextchar_num:
		index = eol
	    else:
		index = nextchar
	    self.delete("insert", index)

    def Insert(self, evt):
	self.toggle_overwrite()
	return "break" # Don't insert clipboard text

def _test():
    import sys
    t = OverwriteText()
    t.pack()
    t.insert(0.0, open(sys.argv[0]).read())
    t.mainloop()

if __name__ == '__main__': _test()




From idbaxter at semdesigns.com  Fri Apr 19 18:12:56 2002
From: idbaxter at semdesigns.com (Ira D. Baxter)
Date: Fri, 19 Apr 2002 17:12:56 -0500
Subject: source code cross reference tool
References: <mailman.1019208819.9434.python-list@python.org>
Message-ID: <3cc09423@giga.realtime.net>

"Tripp Scott" <tripps81 at yahoo.com> wrote in message
news:mailman.1019208819.9434.python-list at python.org...
> ... it would be cool to have the source
> code in a set of html pages that allows a user to click around
> and jump to function/class definitions, etc.

Don't have that for Python, but could do it if there was enough
interest.  See the COBOL equivalent (ZIP file download), if you like,
at http://www.semdesigns.com/Products/Formatters/index.html

--
Ira D. Baxter, Ph.D. CTO Semantic Designs, Inc.
http://www.semdesigns.com




From johnroth at ameritech.net  Fri Apr 26 17:42:28 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 26 Apr 2002 17:42:28 -0400
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk> <just-9E8624.12281026042002@news1.xs4all.nl>
Message-ID: <ucjicvlrf7djd3@news.supernews.com>

"Just van Rossum" <just at xs4all.nl> wrote in message
news:just-9E8624.12281026042002 at news1.xs4all.nl...
> In article <C2cBaXANXSy8Ewbb at jessikat.demon.co.uk>,
>  Robin Becker <robin at jessikat.fsnet.co.uk> wrote:
>
> > With 2.2.1 I'm puzzled why the first returns the local x and the
second
> > the global I suppose x=x rather than z=x is the problem, but don't
> > understand why. I opposed local scopes originally as I liked the two
> > scope rule, but this seems more complex than I imagined.
> >
> > >>> x=2
> > >>> def bingo(x):
> > ...     class A:
> > ...             z=x
> > ...             def get(self):
> > ...                     return self.z
> > ...     return A().get()
> > ...
> > >>> bingo(4)
> > 4
> > >>> def bango(x):
> > ...     class A:
> > ...             x=x
> > ...             def get(self):
> > ...                     return self.x
> > ...     return A().get()
> > ...
> > >>> bango(4)
> > 2
>
> What a coincidence:
>    http://www.python.org/sf/532860
>    http://mail.python.org/pipermail/python-dev/2002-April/023427.html
>
> Just

This doesn't appear to be the same bug. Robin's example
has two functions which are identical, other than the
substitution of 'x' for 'z' in the definition of class A.

The discussion in the bug report indicates that the 'x'
following the '=' should be compiled using the 'LOAD_NAME'
op code, which implements the old rules. If it in fact did this,
I would expect the first example to return 4, not 2,
because 'LOAD_NAME' is supposed to ignore the
parameter.

Clearly, it isn't. At a guess, I'd think the problem is that
the compiler is trying to make 'z' a local variable of the
def, so it's using a different op code in the first case than
in the second.

Of course, I could be totally confused, as usual.

John Roth




From op73418 at mail.telepac.pt  Mon Apr 22 13:19:35 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 18:19:35 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <2lXw8.130011$3L2.11915313@bin7.nnrp.aus1.giganews.com>
Message-ID: <o4h8cuo0j9b2kmc787fmk6f5bjt4b5n88n@4ax.com>

On Mon, 22 Apr 2002 17:00:46 GMT, "Terry Reedy" <tejarex at yahoo.com>
wrote:

>
>"Gon?alo Rodrigues" <op73418 at mail.telepac.pt> wrote in message
>news:dit7cuo1nguiha34nq80m9q39igpjlr70d at 4ax.com...
>, 22 Apr 2002 06:10:02 GMT, Tim Daneliuk <tundra at tundraware.com>
>...
>> You should understand that this is a philosophical viewpoint. The
>> mathematical field is roughly divided in the Platonist field, those
>who
>> believe that mathematical objects exist somewhere out there in space
>(to
>> quote Sun Ra) in a Wonderful World of Platonic Ideas,
>
>This is such a crude and misleading caricature of the minority
>Platonist viewpoint that I suspect you must incline toward
>
>> the formalist
>> field of those who believe mathematics is essentialy a game we play
>on
>> paper with symbols with no "real" meaning attached to it besides the
>> convenience in describing natural phenomena.
>
>Terry J. Reedy
>
>

I confess: I did not try to make an exact and rigorous picture of any of
the fields. And if I was misleading in any way do you care to correct
me? 

Also, I am not so sure that the Platonists are in the minority.

best regards,
Gon?alo Rodrigues



From jon+usenet at unequivocal.co.uk  Wed Apr 24 11:28:32 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 15:28:32 -0000
Subject: Php vs Python
References: <mailman.1019615325.15799.python-list@python.org> <3cc6a741$0$15474$afc38c87@news.optusnet.com.au>
Message-ID: <slrnacdjov.ja1.jon+usenet@snowy.squish.net>

In article <3cc6a741$0$15474$afc38c87 at news.optusnet.com.au>, nobody wrote:
> However, PHP was made for HTML, is well supported and is easily inlined in
> HTML whereas, AFAIK, Python pages have to be entirely Python.

This is not true, there are many packages you can download, some
simple, some complicated, in order to create web applications using
Python. Some of these emulate almost exactly the PHP-style
code-embedded-in-HTML model, some are more sophisticated.

One of the simpler (but still powerful) ones happens to be at
http://jonpy.sourceforge.net/ . I just mention this in passing <wink>


From borcis at geneva-link.ch  Tue Apr 30 04:33:03 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 30 Apr 2002 10:33:03 +0200
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch> <MNiz8.32691$8D3.925320@news1.tin.it>
Message-ID: <3CCE56BF.9000109@geneva-link.ch>

Alex Martelli wrote:

> 
> I'm surprised you miss the main advantage -- 0666, the Number
> of the Beast in octal, translates to decimal 438, the A.D. in which
> Emperor Theodosius II promulgated the Theodosian Code,


Very neat, but...

Surely you are aware that 666 refers to president Bush' notion that
"whoever isn't with us is against us" - since the excluded middle is 
written excluded "third" in most reasonable languages, including latin, 
e.g. "tertium non datur", and 1 - 1/3 = 0.666... ?

OTOH, the aversion for the excluded middle implied by this (quite 
natural) interpretation of the 666 should as well apply to mutual 
exclusion *of interpretations in general*, and of interpretations of the 
666 in particular (ambiguity is abusively maligned, if one believes in 
Goedel's theorem).

Now the chapter of the Rev. that ends with 666 is inviting the naive 
reader to assume the number can't refer to himself, while calling for 
wisdom and intelligence... so I found the following way to associate it 
with my initials.

Let B = 666
     K = 4.66920... Feigenbaum's delta constant

Thus K**K = B + 1 + B

Cheers, Boris J. Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From claird at starbase.neosoft.com  Thu Apr 25 13:47:56 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Apr 2002 12:47:56 -0500
Subject: Python v. Euphoria
References: <394fcus92g8rt0e2ektsd36s9qu7cq2cij@4ax.com> <aa8531$fho$1@panix3.panix.com>
Message-ID: <DBD012232B537A40.AC23647CB5D9162D.E46A489263D39B3C@lp.airnews.net>

In article <aa8531$fho$1 at panix3.panix.com>, Aahz <aahz at pythoncraft.com> wrote:
>In article <394fcus92g8rt0e2ektsd36s9qu7cq2cij at 4ax.com>,
>Stom Fairecroire  <stom at hotpop.com> wrote:
>>
>>Many, many years ago I did a lot of programming in HPL (Hewlet Packard
>>Language) then did lots using PICK basic. Now I want to start some
>>very low key programming again - just for my own amusement. Anyone
>>have any views on the pros and cons  of Python v. Euphoria ? Which
>>would be the better option?
>
>Based on a quick look at the Euphoria web site, Python would be a much
>better option.  Euphoria is a proprietary product that is only available
>on a few platforms; it looks like Euphoria programs tend to not be
>cross-platform even for the available platforms.  Euphoria also looks a
>bit uglier than Python, and it lacks some of Python's features, such as
>threads.
			.
			.
			.
I'd put the case stronger than that:  Python has 
proven to be applicable over an exceptionally wide
scope of programming requirements, and Python has
a record of satisfying many beginners and near-
beginners in programming.
-- 

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


From logiplexsoftware at earthlink.net  Wed Apr 10 17:54:30 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 14:54:30 -0700
Subject: Extracting from a list
In-Reply-To: <3CB4B1D6.1030504@dittmar.net>
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
	<3CB4B1D6.1030504@dittmar.net>
Message-ID: <20020410145430.4432d2be.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 23:42:46 +0200
Daniel Dittmar wrote:

> How about:
> 
> for i in xrange (len (lst) - 1, -1, -1):
>      item = lst [i]
>      if item < 5:
>          print item
>          del lst [i]
> 
> This makes it possible to delete items from the list as the access is 
> back to front.

Good answer.  Why didn't I think of that...

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From lac at strakt.com  Fri Apr 19 06:22:29 2002
From: lac at strakt.com (Laura Creighton)
Date: Fri, 19 Apr 2002 12:22:29 +0200
Subject: Changing the repr() of a class? 
In-Reply-To: Message from "Donald McCarthy" <paddy3118@tiscali.co.uk> 
   of "Fri, 19 Apr 2002 08:00:21 BST." <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com> 
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <I6Fv8.36158$vF6.1093396@news2.tin.it> <mailman.1019162916.19664.python-list@python.org> <7cIv8.36928$vF6.1122050@news2.tin.it> <mailman.1019184878.1733.python-list@python.org>  <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com> 
Message-ID: <200204191022.g3JAMUdj014958@ratthing-b246.strakt.com>

> Thanks Alex and thanks to all who discussed my problem. I am honoured to
> participate in such a group.
> 
> The 'Alex solution' worked but it does lead on to another problem -
> Reference material!
> 
> I bought 'Learning Python' by Mark Lutz, First Edition. Read it when I
> bought it a couple of years back, but I usually get by with the online
> documentation that comes with python.
> 
> With all the changes in Python since then I need a new reference work. Would
> anyone recommend a  book that includes info on metaclasses?
> 
> Thanks again,
> 
>  Donald 'Paddy' McCarthy.

If you are willing to wait, Alex's own Python in a Nutshell will be
published by O'Reilly soon.  In this particular case, I think the
Nut is more of a coconut than a pistachio nut :-).  Its not PP2E sized, and
no doubt Alex will complain about all the wonderful things he had
to _cut_, but nonetheless ...

Laura Creighton




From tdelaney at avaya.com  Thu Apr  4 20:34:39 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 5 Apr 2002 11:34:39 +1000 
Subject: ACCEPTED: PEP 285
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A252@natasha.auslabs.avaya.com>

> From: James Logajan [mailto:JamesL at Lugoj.com]
> 
> Eri Ma Franci <max at alcyone.com> wrote:
> > Could you find something better to complain about?
> 
> Then I'd like to complain that neither you nor Guid va Rossu 
> addressed the 
> question I asked in my previous message concerning bools as 
> keys. If a bool 
> is to be treated as an int sometimes and sometimes not, I 
> don't recall 
> seeing anything about whether it will be treated as an int as 
> a dictionary 
> key. If I do:
> 
> x = ["is", "are"]
> y = {0: "", 1:"s"}
> 
> for n in range(3):
>    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])

This has been addressed *many* times in these threads. If you have failed to
read them, that is your problem.

One last time.

True compares equal to 1. False compares equal to 0. Therefore, in a
dictionary, True and 1 are synonymous as keys, and False and 0 are
synonymous as keys. So, the following code:

d = {}
d[0] = 'Hello'
d[1] = 'World'
d[True] = 'T'
d[False] = 'F'

print len(d)
print d[1]
print d[True]
print d[False]
print d[0]

will print:

2
T
T
F
F

If that doesn't answer your question, I have no idea what will. There is no
ambiguity.

A bool *is an* int according to PEP 285, which has been accepted.

Tim Delaney




From greg at cosc.canterbury.ac.nz  Wed Apr 10 01:39:39 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 17:39:39 +1200
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com>
Message-ID: <3CB3D01B.8F11B041@cosc.canterbury.ac.nz>

Peter Olsen wrote:
> 
> What's wrong with addressing those problems with a license that
> prohibits users from further distributing either binary or
> source code while making them both freely available from the
> author's website?

I think I'd prefer to address things like that by *asking*
people not to redistribute, etc., rather than bludgeoning
them with legalities.

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


From ajs at ix.netcom.com  Mon Apr  1 15:00:10 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Mon, 1 Apr 2002 15:00:10 -0500
Subject: re PEP 285: Adding a bool type
Message-ID: <000001c1d9b9$13198020$0334fea9@carol>

Erik writes -

>I'm not really sure how beneficial this would be. In the real 
>world, not everything about a programming language -- even
>a low overhead language like Python -- can be instantly 
>intuited by non-programmers.

Isn't this exactly the argument that seems to be consistently 
rejected when put forth by those of us that feel that much of
this concern about 'non-programmers' is, at best, much ado 
about nothing and at worst directly counterproductive to
Python status as a language particularly useful as an 
introduction to real world programming.

There seems to be two forms of status quo seekers in 
the Python community - one, that sees that status quo 
in the semantics of Python as of its development at a point
in time (let's say 1.5.2) - and feel there should be some serious
burden of proof on those seeking semantic changes.
And even expect Guido to at least seriously try to carry
that burden. 

I happen to be of that camp.

The other seems to take the position that Guido got us to
1.5.2,  the status quo is in accepting his intuitions about
moving forward from there, the burden of proof being squarely
on those who seek to question or resist those intuitions.

Which I as a member of the other camp realize to be a
reasonable position.

But one to which I (and others) are constitutionally 
unable to adapt.

What does seem clearest is that the rest
is pretty much just a debating game.

Art





From alun at chromosome57.com  Tue Apr 16 15:12:10 2002
From: alun at chromosome57.com (Austin Luminais)
Date: 16 Apr 2002 12:12:10 -0700
Subject: zipfile problems
Message-ID: <2e1b0869.0204161112.425f41d4@posting.google.com>

I have an archive of several hundred zip files, and I'm trying to
automate unzipping them with the zipfile module (Python 2.2,
Activestate).  Most of them work, but 58 of them don't.  Most that
don't work get this error:

BadZipfile: File is not a zip file, or ends with a comment

And a few get:

BadZipfile: Unsupported compression method 6 for file (filename)

WinZip and Powerdesk have no problem extracting these files.
Would this be considered a bug in zipfile or zlib, or would they
instead be considered corrupt files that WinZip happens to handle?

A few of the files are available here:
http://www.chromosome57.com/python/badfiles/

alonerbm.zip gives the unsupported compression method error, and the
others give the "not a zip file" message.


From fperez528 at yahoo.com  Mon Apr  1 12:40:43 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 01 Apr 2002 10:40:43 -0700
Subject: [Newbie] embed c++ in Python
References: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <a8a62r$57u$1@peabody.colorado.edu>

Matthias Stern wrote:

> Hi!
> 
> I'm new to Python and I'm looking for a good tutorial how to embed a C/C++
> function into Python. I found some tutorials but no one describes what I
> need:
> 
> I want to know how to write a function which receives some variables as
> argument changes them and sends the results back to Python. E.g.:
> 
> int doSomething(int *a, int *b) {
>    :
>    *a = 10;
>    *b = 123;
>    :
> }
> 
> Can I do this in Python? I'd like to call the function doSomething() from
> Python and receive the changed values a and b in Python again.
> 
> Is there a tutorial? Has someone got a simple example for this?
> 
> Thx, Mac

The easiest of the many possible ways to do it:

http://www.scipy.org/site_content/weave

f.


From logiplexsoftware at earthlink.net  Mon Apr  8 14:33:46 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 8 Apr 2002 11:33:46 -0700
Subject: wxPython / Your Experiences in large project ???
In-Reply-To: <a8rp2u$ht8$02$1@news.t-online.com>
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <20020408113346.74889690.logiplexsoftware@earthlink.net>

On Mon, 8 Apr 2002 11:55:24 +0200
Harald Schneider wrote:

> Hi,
> 
> what are your experiences, using wxPython/wxWindows in large projects?
> 
> Large means in my case a sales system, based on MySQL.
> I'm thinking about implementing it in wxPython, but I'm not sure about the
> stability of
> wxWindows and the wxPython wrapper ...

While working as a developer at a call center (telephone sales and surveys) I
developed a system using Python, wxPython and MySQL.  I did encounter the
occasional problem when upgrading wxPython versions, but nothing that wasn't
readily fixed.  It was a moderately complex system, having a custom markup
language for prompting the telemarketer (so different sales regions could use a
different "script" without code changes) that was parsed and passed through as
HTML to the wxPython HTML widget.  Besides the database access, the software
also had to control an autodialer via telnet and have a fallback dialing system
in case the autodialer failed.  The system ran on 90 seats, 16 hours a day, 7
days a week for over two years. It was fast, stable, and any of the problems we
had could be usually be directly attributed to me =)

Also, if you have doubts about whether the lack of a standard/stable/* IDE for
wxPython is a problem, consider that this system went live less than 30 days
after the initial spec. There was only one developer (me) using only XEmacs
(although the first couple of days in production were a special sort of hell ;)

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mlh at vier.idi.ntnu.no  Tue Apr 30 14:33:35 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 30 Apr 2002 18:33:35 +0000 (UTC)
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no> <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>  <mailman.1020033011.19748.python-list@python.org>
Message-ID: <slrnactoru.pv.mlh@vier.idi.ntnu.no>

In article <mailman.1020033011.19748.python-list at python.org>, James J.
Besemer wrote:
>
>Steve Holden wrote:
>
>> I only recently realised that the OAD makes an amazing omission, completely
>> failing to provide a definition for the word "gullible".
>
>Made me look.

Gullible, anyone? <wink>

>--jb

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From phr-n2002a at nightsong.com  Mon Apr 15 06:58:11 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 03:58:11 -0700
Subject: Python webmail packages?
Message-ID: <7xr8lhb54s.fsf@ruckus.brouhaha.com>

Does anyone have a favorite webmail system written in Python?
Something like Squirrelmail (www.squirrelmail.org) but written in
Python and hopefully with fewer bugs ;-).  It would be good if it
talks to an IMAP server like Squirrelmail does, but that's not mandatory.


From gashton at cmedltd.com  Mon Apr  8 06:25:43 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 08 Apr 2002 11:25:43 +0100
Subject: Nested scopes hitch
In-Reply-To: <MPG.17179f847bf41a6998969b@news.nwlink.com>
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> 
	<MPG.17179f847bf41a6998969b@news.nwlink.com>
Message-ID: <1018261543.690.5.camel@tp6>

On Fri, 2002-04-05 at 20:49, Jeff Shannon wrote:

> In class A, the attribute a is a class attribute, not a name in an 
> enclosing scope, so the nested scope rules don't apply.  You can 
> access it by referring to self.a, however -- that will find both 
> instance and class attributes.

I remember being slightly surprised by the fact that self finds class
variables. Is there any particular reason why, considering that the you
can reference them as ClassName.var anyway?

I suspect that my surpise stemmed form an assumption that readibility
would be improved if you can tell just by looking at it whether a
variable is a class or instance variable.

I'm just curious...

Python 2.2 (#2, Mar 11 2002, 13:24:00) 
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> class Spam:
...   a = 1
...   def test(self):
...     if Spam.a == self.a:
...       print "equal"
...     if Spam.a is self.a:
...       print "identical"
... 
>>> s = Spam()
>>> s.test()
equal
identical


-- 
Graham Ashton





From mats at laplaza.org  Thu Apr 18 11:08:07 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Thu, 18 Apr 2002 15:08:07 GMT
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <899f842.0204170840.29e579ae@posting.google.com> <20020418002823.C4068@phd.pp.ru> <mailman.1019077897.8110.python-list@python.org> <87r8ldqygo.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <3cbee101.152536606@news.laplaza.org>

On 18 Apr 2002 16:04:39 +0900, "Stephen J. Turnbull"
<stephen at xemacs.org> wrote:

:>>>>> "Cliff" == Cliff Wells <logiplexsoftware at earthlink.net> writes:
:
:    Cliff> I suppose Microsoft's claim that the GPL is viral could be
:    Cliff> better applied to their own source since merely *seeing* it
:    Cliff> could affect your code whereas you have to actually *use*
:    Cliff> GPL'd code to be affected.
:
:I've not seen this before.  Please send it to rms.

It's not really different then when most of the early GNU Project
utilities were developed, at that time seeing the UNIX source code had
a chance of being "polluting" in the same way.


Mats Wichmann



From trentm at ActiveState.com  Thu Apr 18 12:26:17 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Thu, 18 Apr 2002 09:26:17 -0700
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
In-Reply-To: <a9kuj8$bsa$1@nsnmrro2-gest.nuria.telefonica-data.net>; from gusraan@terra.com.br on Wed, Apr 17, 2002 at 07:56:59PM -0300
References: <mailman.1019071474.3193.python-list@python.org> <Xns91F4680F3221cliechtigmxnet@62.2.16.82> <a9kuj8$bsa$1@nsnmrro2-gest.nuria.telefonica-data.net>
Message-ID: <20020418092617.L28973@ActiveState.com>

[Raphael Ribeiro wrote]
> Do i have to remove the last version of active python i have?

That depends. In general though, yes, you'll be happier if you uninstall
other ActivePython's first. ActivePython 2.2.x cannot co-exist with
other Python 2.2.x installations and ActivePython 2.1.x cannot co-exist
with other Python 2.1.x installations. You *can* have both a Python
2.1.x and a Python 2.2.x installed at the same time.

For ActivePython's Install Notes:
http://aspn.activestate.com/ASPN/Downloads/ActivePython/install_22

- Directory Contents: If you are installing ActivePython to an existing
  directory, ActivePython's setup may overwrite files in that directory.
  We recommend that you install ActivePython to a new or empty
  directory. 

- Other Python 2.2 Installations: On Windows, ActivePython 2.2 cannot
  coexist with other Python 2.2 installations (for example, the
  reference implementation of Python 2.2). Uninstall any other Python
  2.2 installations before installing ActivePython 2.2. 

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From claird at starbase.neosoft.com  Tue Apr  9 16:42:31 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 9 Apr 2002 15:42:31 -0500
Subject: Beginner's performance problem
References: <memo.20020409174302.1584B@a.a> <slrnab68mh.15p.gerhard@lilith.my-fqdn.de>
Message-ID: <D948EE567EE6A22E.5C4CE0A478940342.35FB5D603AE15D83@lp.airnews.net>

In article <slrnab68mh.15p.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>Mark Charsley wrote in comp.lang.python:
>> For reasons best not gone into, I needed to correct the case of a whole 
>> bunch of SQL code. As such I created the little script below...
>> 
>> It reads in a source file, then for each line it does a case-insensitive 
>> search for each TableName and ColumnName (as contained in the "names" 
>> collection), checks that it's not just matching a substring, and then 
>> corrects the case of the word if necessary.
>>
>> [snip code]
>>
>> 1) why is it so slow? I wouldn't have thought that processing 13,000 lines 
>> * 200 names shouldn't take more than a few seconds on a modern PC.
>
>String manipulation is costly in Python because strings are immutable.
>Every string manipulation creates a new string object in effect. A
>good approach is to tokenize your string into a list (which is
>mutable) and operate on the list only, then join the list into a
>string again.
>
>I haven't tried but an approach using the re module should be really
>fast, and shorter. Untested code follows:
>
>import re
>names = ["ALGORITHM_CLASS", ...]
>
>text = open("ThirteenThousandLines.sql").read()
>for name in names:
>    regex = re.compile(name, re.I) # case insensitive search for name
>    text = regex.sub(name, text)   # replace all occurences
>print text
				.
				.
				.
It's part of the Python Way.  Write a seven-line solution,
and it works.  Write a hand-tuned hundred-line alternative,
and not only does it take a day longer to get right, but it
runs two orders of magnitude slower.  Guido's Master Plan
emphasizes this element of behavorial conditioning.
-- 

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


From max at alcyone.com  Tue Apr 16 19:30:51 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 16:30:51 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com> <a9fu9g$mj3$1@watserv3.uwaterloo.ca> <3CBBAB60.2CC717D7@alcyone.com> <slrnabp974.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <3CBCB42B.CB515B66@alcyone.com>

phil hunt wrote:

> I'm not suggesting removing it, merely making it optional. People
> can still use it if they want to.

Well, this is all academic now anyway, since Guido Has Spoken.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From guido at python.org  Wed Apr  3 14:46:55 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 14:46:55 -0500
Subject: [Python-Dev] PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 13:25:05 CST."
             <NBBBIOJPGKJEKIECEMCBKEFDMLAA.pobrien@orbtech.com> 
References: <NBBBIOJPGKJEKIECEMCBKEFDMLAA.pobrien@orbtech.com> 
Message-ID: <200204031946.g33JktP25857@odiug.zope.com>

> It feels like there are too many combinations available in the
> PEP. (Or maybe too many questions and options which could result in
> weird combinations.) I'd like to see a new version with Guido's
> latest thoughts.

My thoughts haven't changed at all. :-)

The first round of discussion (on python-dev) brought up some
contentious issues where people disagreed strongly with particular
details proposed in the PEP, and I put those contentious issues at the
top of the review list.  If you go to the latest version of the PEP on
the web (http://www.python.org/peps/pep-0285.html) you'll see that
I've added some other issues that were brought up in the more recent
discussion.

> For example, are we closer to a resolution on the
> str() and repr() issues?  The only behavior I like is that exhibited
> by the sample code:
> 
> >>> str(True)
> 'True'
> >>> repr(True)
> 'True'

So far I believe only Marc-Andre has expressed himself against this,
so I'll make it so.

> How about the case of the constants: true, false vs. True, False? I
> like the latter, which match None. They would also stand out in the
> new docstrings mentioned by Fred.

Ditto.

> How about operator.truth()? Has Tim changed his mind at all? Has Guido?

I don't know about Tim.  I've made an executive decision that
operator.truth() will return a bool (making it an alternate spelling
for bool()).

> If this PEP were implemented and "all built-in operations that
> conceptually return a Boolean" were updated, what verifiable
> examples of code breakage would we have?

I've changed the list of operations explicitly mentioned in the PEP
(comparisons, some builtins, some string and file methods, a few other
things).  It breaks about 12 tests in the test suite.  I haven't
looked at all of these in detail, but I believe they are all shallow,
of the following form: a test does "print x == y" and then the output
is matched to "expected" output.  The expected output strings must be
changed from 0/1 to False/True.

> The issue of dictionary keys has been raised, and is a bit
> perplexing, but not overly so, IMO. However, I'd like confirmation
> that this is indeed the intended behavior:
> 
> >>> d = {}
> >>> d[0] = 'First'
> >>> d
> {0: 'First'}
> >>> d[False] = 'Second'
> >>> d
> {0: 'Second'}
> >>> d[0.0] = 'Third'
> >>> d
> {0: 'Third'}
> >>> d[()] = 'Fourth'
> >>> d
> {0: 'Third', (): 'Fourth'}
> >>> d[None] = 'Fifth'
> >>> d
> {0: 'Third', (): 'Fourth', None: 'Fifth'}

Yes.

> Or are there any situations like this where you would want True and
> False to be a bit more like None and a bit less like ints? Meaning
> True and False could be dictionary keys separate from 0 and 1.

No.  That would break backwards compatibility.  False==0, and True==1;
everything else follows from that.  (But False is not 0, and True is
not 1!)

> Other than that, are there any concrete issues that have been raised
> that need to be addressed by the PEP? I understand the uneasy
> feeling felt by many as I felt it myself. However, I've read all the
> newsgroup postings and reread the PEP and I don't feel very queasy
> any more. I think this PEP is a step in the right direction. I would
> hope that clarifying some of the genuine issues might help to bring
> about consensus and reduce some of the rhetoric.

There's an educational issue that I'm just getting my head around.  It
appears that for various psychological and linguistic reasons, newbies
naturally write

    if b == True: ...

when they first encounter a bool variable and have to test whether
it's true.  I believe that this is most of Laura's issue.  The only
reasonable solution for this is that the Zen master slaps them over
the head with a stick, and they learn to write

    if b: ...

I'll update the PEP.

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




From aahz at pythoncraft.com  Thu Apr  4 10:07:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 10:07:04 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com> <slrnaannkc.6ls.kzaragoza@localhost.localdomain> <3CABEA67.FD3F75C0@alcyone.com>
Message-ID: <a8hq6o$o9a$1@panix1.panix.com>

In article <3CABEA67.FD3F75C0 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>"Kris J. Zaragoza" wrote:
>>
>> Maybe, but I'm not sure this is strictly necessary, since according to
>> the PEP:
>> 
>>     ("%s" % str(True)) => "True"
>
>Unless the alternate choice is accepted where str(True) == '1'.

Did you read the final version?  C'mon, don't start arguments where you
have no clue what you're talking about.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From nospam-readonly at getsoft.com  Sun Apr 21 23:16:56 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 13:16:56 +1000
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net> <3cc3765f$0$15478$afc38c87@news.optusnet.com.au> <B8E8F587.21E4A%goodger@users.sourceforge.net>
Message-ID: <3cc38091$0$15473$afc38c87@news.optusnet.com.au>

Thanks, one small example that works is worth a thousand somethingoranothers
;-)

I'm a C++ developer and as I get older I find learning new languages doesn't
necessarily get any easier.

I'll go look at the source.


"David Goodger" <goodger at users.sourceforge.net> wrote in message
news:B8E8F587.21E4A%goodger at users.sourceforge.net...
> Neville Franks wrote:
> > I thought that Tim's original ndiff compared files, but the one in
difflib
> > compares lists of strings. Does this mean I somehow need to read the
files
> > into lists of strings and then call ndiff.
>
> Yes.  It's easy to turn a file into a list of strings.  Read with
> afile.readlines(), or split a multi-line string with astring.splitlines(1)
> (the 1 keeps the line endings).
>
> > A short example of how to use
> > ndiff to compare two files would be greatly appreciated.
>
> Use the source, Luke!  There are painfully detailed examples in there.
>
> Here's a quickie (untested, but oughta work)::
>
>     import difflib
>     list1 = open('file1').readlines()
>     list2 = open('file2').readlines()
>     diff = difflib.ndiff(list1, list2)
>     print ''.join(diff)
>
> --
> David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
>   - Python Docutils: http://docutils.sourceforge.net/
>   - The Go Tools Project: http://gotools.sourceforge.net/
>




From phd at phd.pp.ru  Tue Apr  2 15:10:05 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 3 Apr 2002 00:10:05 +0400
Subject: Shelve Help (file portablility question)
In-Reply-To: <89a9b6ce.0204021113.7c4bf243@posting.google.com>; from jkring@calbay.com on Tue, Apr 02, 2002 at 11:13:30AM -0800
References: <89a9b6ce.0204021113.7c4bf243@posting.google.com>
Message-ID: <20020403001005.H17464@phd.pp.ru>

On Tue, Apr 02, 2002 at 11:13:30AM -0800, Jim wrote:
> I cannot read a shelve file that I created in Python 1.5.2 on Redhat
> 6.X using Python 1.5.2 on Redhat 7.X.  Is there any way that I can do
> this?
> 
> my error message is shown below:
> 
> >>> db = shelve.open('shelveFile')
> Traceback (innermost last):
>     File "<stdin>", line 1, in ?
>     File "/usr/lib/python1.5/shelve.py", line 152, in open return
> DbfilenameShelf(filename, flag)
>     File "/usr/lib/python1.5/shelve.py", line 142, in __init__
> Shelf.__init__(self, anydbm.open(filename, flag))
>     File "/usr/lib/python1.5/anydbm.py", line 86, in open return
> mod.open(file, flag, mode)
>     File "/usr/lib/python1.5/dbhash.py", line 8, in open return
> bsddb.hashopen(file, flag, mode) bsddb.error: (22, 'Invalid argument')

   http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.099.htp

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



From alun at chromosome57.com  Wed Apr 17 16:20:21 2002
From: alun at chromosome57.com (Austin Luminais)
Date: 17 Apr 2002 13:20:21 -0700
Subject: shipping a python app
References: <a9js5c$2m9$1@panix2.panix.com>
Message-ID: <2e1b0869.0204171220.29455f34@posting.google.com>

roy at panix.com (Roy Smith) wrote in message news:<a9js5c$2m9$1 at panix2.panix.com>...
> I'd love to write this in python/Tkinter, but it'll be a bit of an
> uphill battle to convince management.  While I do a lot of python
> development for internal tools, my company doesn't currently ship any
> python-based products because of the impression that it's not widely
> supported.

I tried Tkinter on Windows a long time ago, and had a persistent
problem with the windows not being closeable.  I believe I also had
some crashes.  I checked Sourceforge recently, and it looked like the
non-closing windows was still an open bug!

I'm using wxPython now and haven't had any problems.  A lot of people
prefer it, but it does add at least two megs to your distribution
(about two megs compressed, about four megs uncompressed).

It does sound like Tkinter would be better in your situation, but I'd
test it a bit on Windows first.


From peter at engcorp.com  Thu Apr 18 00:38:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 18 Apr 2002 00:38:02 -0400
Subject: Hostname Resolution not working in Explorer
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net>
Message-ID: <3CBE4DAA.E710D618@engcorp.com>

Gary Spivey wrote:
> 
> I have windows XP Home, and IE6. I have a zope server running on my box.
> The server is running fine, and I can access if it I access
> 
> http://127.0.0.1:8080/mange
> 
> However, if I use
> 
> http://localhost/mange
> 
> the page isn't found. The same holds true if I use my hostname rather than
> localhost - also, the actual IP address works.

Well, at least two mistakes in the above: you didn't use port 8080 in
the second example, and you spelled 'manage' as 'mange'.  With those
typos, maybe you are consistently making similar mistakes when trying
this?

Also, what is "page isn't found"?  You should actually be getting
a "The page cannot be displayed" message from IE6.

By the way, this is not Python related.  Just a basic networking 
or configuration problem or something.


From jb at cascade-sys.com  Sun Apr 21 08:51:06 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 21 Apr 2002 05:51:06 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <uc5aajf0m2dmc5@news.supernews.com>
Message-ID: <3CC2B5BA.777D7B1A@cascade-sys.com>

John Roth wrote:

> I think what people are trying to tell you is that many of the
> features that Lisp pioneered are now features of so many
> languages that it simply isn't necessary to quote Lisp as a
> direct ancestor of anything, unless you're doing historical
> research.

MESSAGE RECEIVED LOUD AND CLEAR.

> One of the things that is most unusual about Python is
> it's rejection of bracketing where it isn't needed:
> using indentation for block structure is one example.
> This is diametrically opposed to Lisp.

Right.  Along with most of the rest of Python's syntax.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From goodger at users.sourceforge.net  Sun Apr 21 21:50:02 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Mon, 22 Apr 2002 01:50:02 GMT
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au>
Message-ID: <B8E8E4D6.21E42%goodger@users.sourceforge.net>

Neville Franks wrote:
> Is there some one to run this in 2.1.2 or should I just install 2.2?

Unless you have a reason to stay with 2.1, you might as well install 2.2.1
(the latest).  If you don't want to, then as I said:

>> you can get a Python 2.0/2.1-compatible version from
>> http://docutils.sourceforge.net/test/difflib.py

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
  - The Go Tools Project: http://gotools.sourceforge.net/



From gimbo at ftech.net  Fri Apr 12 09:16:33 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 12 Apr 2002 14:16:33 +0100
Subject: ftplib and FTP Objects...
In-Reply-To: <20020412101611.GB15466@andy.tynant.ftech.net>
References: <a94i7u$6id$1@news.wrc.xerox.com> <20020412101611.GB15466@andy.tynant.ftech.net>
Message-ID: <20020412131633.GA17830@andy.tynant.ftech.net>

On Fri, Apr 12, 2002 at 11:16:11AM +0100, Andy Gimblett wrote:

> ftp.storbinary('STOR ' + fname, f.read())

Sorry, I just realised that should be:

ftp.storbinary('STOR ' + fname, f)

(ie ftplib takes care of the read() for you).

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From gdiribarne at udcast.com  Tue Apr 23 08:41:12 2002
From: gdiribarne at udcast.com (Gilles Diribarne)
Date: 23 Apr 2002 14:41:12 +0200
Subject: Problem with static library
Message-ID: <1019565672.1280.11.camel@meteor.udcast.com>

Hi,

I use python 2.2 with libpython2.2.a statically linked.
I need to have a C program totally self-contained (i.e. statically
linked). But, I use xml.parsers.expat library which needs libexpat.so.

The program generate a SIGSEV, when trying to load libexpat.so.
Can you help me?

Gilles







From skip at pobox.com  Wed Apr 17 11:57:11 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 17 Apr 2002 10:57:11 -0500
Subject: mapping HTML entities to Unicode (was: Re: Unicode name questions)
In-Reply-To: <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de>
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com>
        <009901c1e58b$28a5bd90$445d4540@Dell2>
        <mailman.1018994314.14044.python-list@python.org>
        <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15549.39767.650448.746011@12-248-41-177.client.attbi.com>

    Martin> Skip Montanaro <skip at pobox.com> writes:
    >> "Lambda" has been spelled with a "b" as long as I can remember.  

    Martin> I think the Unicode charts use an ASCII transcription of the
    Martin> native pronouncation of the letters, if possible....

Thanks for the explanation, Martin.  I now have a dictionary mapping 391
specific entity strings to their plain string or Unicode string equivalents.
I know there are more.  For example, my dictionary doesn't describe any
numeric entities encoded in hex (easy enough to preprocess back to decimal,
but I've never encountered any).  I have a map_entities function that does
the obvious chr(int(ent[2:-1])) thing for decimal numeric entities that map
into the ASCII range.

Is this of use to anyone besides me?  I tossed a copy of the dictionary into

    http://manatee.mojam.com/~skip/python/entities.py

I'm willing to maintain and improve it (and expose a function that uses the
dictionary) if others feel it would be of use.  If not, I'll just muddle
along with my own minor idiosyncracies.  I'd also be happy to dump this in
favor of a preexisting solution.

Skip




From abele  Tue Apr  9 02:52:00 2002
From: abele (abele)
Date: 9 Apr 2002 08:52:00 +0200
Subject: zope: Zahlen-Index
Message-ID: <3cb28f90$1@netnews.web.de>

Liebe Experten,


ich weis nun nicht, ob ich hier in der Liste richtig bin. 
Trotzdem m?chte ich folgendes Zope-Problem posten:


Folgendes Problem stellt sich mir:


Ich habe einen ZCatalog erstellt und habe die relevanten Objekte 
katalogisiert.
Nun gibt es jedoch dabei Objekte, die im Titel nur Zahlen (z.B. 
Jahreszahlen)aufweisen.
Offenbar werden nur Zahlen nicht in den Index ?bernommen und nach diesen 
Objekten kann ich also auch nicht suchen!
Als Indextyp habe ich Textindex verwendet.


Frage: 
Wie kann man erzwingen, dass auch Zahlen indiziert werden?
Wird da ein neues Vacabular ben?tigt? 
Muss die Splitter.py ge?ndert werden?


Kann mir da jemand helfen, an welcher Schraube zu drehen ist, da ich noch 
ziemlich neu in der Zope-Thematik bin.


Schon mal vielen Dank im Voraus


Stephan Abele
--
Stephan Abele
W?rttembergische Landesbibliothek
EDV-Abteilung
Konrad-Adenauer Str. 8
70173 Stuttgart
Tel.: 0711 / 212-4455   Fax: 0711 / 212-4422
E-Mail: abele at wlb-stuttgart.de



-- 
__________________________________________________________
News suchen, lesen, schreiben mit http://newsgroups.web.de


From tim.one at comcast.net  Tue Apr  9 23:06:29 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 09 Apr 2002 23:06:29 -0400
Subject: What has PEP 285 done to us?
In-Reply-To: <Xns91EBA33E75BDlumberjack@216.148.53.96>
Message-ID: <LNBBLJKPBEHFEDALKOLCKELEOMAA.tim.one@comcast.net>

PEP 285 taught us that people don't agree, and that some people disagree in
disagreeable ways.  Frankly, that wasn't news, so let's do a Group Hug and
get on with coding.  Hey!  That guy from dev at null hugs too hard.  And you
over there:  no hugging yourself.

or-back-to-group-backstabbing-your-choice-ly y'rs  - tim






From aleax at aleax.it  Mon Apr 29 11:52:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 15:52:37 GMT
Subject: Lambda Scoping
References: <3ccd5f99.5559046@news.t-online.de>
Message-ID: <9%dz8.31899$8D3.892983@news1.tin.it>

Gerson Kurz wrote:

> The following code
> ...
> funcs = []
> for instance in range(10):
>     funcs.append( lambda: instance )
> print map(apply,funcs)
> 
> funcs = []
> for instance in range(10):
>     funcs.append( lambda instance=instance: instance )
> print map(apply,funcs)
> ...
> prints
> 
> [9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> which I sort-of understand (I figure, in "lambda: instance" instance
> is a lazy expression; "lambda instance=instance: instance" the
> expression itself - "instance" - is a lazy expression, too, but the
> argument list is not.) However, can anybody please explain this more
> detailed?

It's not really connected to lambdas -- you'll get the same effects with
any nested function.  Anyway, if you prefer to discuss it with lambda...:

In the case:
lambda: instance
you have an argumentless callable which, when called, returns the
value of a global or free variable named 'instance' at the time of
the call.  In this case, the calls all happen when instance is
worth 10.

In the case (which you didn't try but is handy for explaining:-):
lambda flippo=instance: flippo
you have an 'argumentless' callable (actually has one optional
argument, but let's forget that:-) with a local variable 'flippo'
that is initialized at definition-time with the value that the
global or free variable named 'instance' happens to have at the
time of the definition.  When called, this callable returns the
value of flippo -- so overall, it returns the value that global
or free variable 'instance' had at the time the callable was
defined.

In other words, an argument with a default value 'snapshots' the
expression at the time of definition.

In the case
lambda instance=instance: instance
you have named the local variable 'instance' rather than 'flippo'
but the name of the local makes no difference, so the treatment
of the second ('missing') case, given above, fully applies.


With real functions, the first case, with which lambda was:
> funcs = []
> for instance in range(10):
>     funcs.append( lambda: instance )
> print map(apply,funcs)
you could translate almost-indifferently to either a more direct
equivalent:

funcs = []
for instance in range(10):
    def patakun(): return instance
    funcs.append( patakun )
print map(apply,funcs)

which like your first case creates and appends ten different
callables, but all behaving identically; or:

funcs = []
def patakun(): return instance
for instance in range(10):
    funcs.append( patakun )
print map(apply,funcs)

which appends 10 references to the same callable instead.


The second case, which in your code was (except for the
name of the callable's local variable):

> funcs = []
> for instance in range(10):
>     funcs.append( lambda instance=instance: instance )
> print map(apply,funcs)

you HAVE to translate into the building of 10 separate
callables, because each carries a different 'snapshot'
in its default-argument:

funcs = []
for instance in range(10):
    def palok(flippo=instance): return flippo
    funcs.append( palok )
print map(apply,funcs)

Here, if you try to move the def out of the loop, the
def fails, because it needs to evaluate 'instance' in
global or outer scope when 'instance' isn't bound yet
(assuming this snippet is all of the program -- else,
the name may have a stray binding remaining from some
previous part of the program).



Hope this has helped rather than confused...

Alex



From tjreedy at udel.edu  Thu Apr 25 12:23:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 25 Apr 2002 16:23:03 GMT
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com>
Message-ID: <H3Wx8.125676$XV5.11204992@bin4.nnrp.aus1.giganews.com>

"Rad" <zaka07 at hotmail.com> wrote in message
news:ad381f5b.0204250629.5144d196 at posting.google.com...
> Thanks for all the suggestions and ideas, I'll try to answer to all
> your questions in this one post.
...
> At this time I don't know for sure how many unique ID's are going to
> be, not less than 15 million I guess.

With 2GB, you might then be able to fit an id/disk address index in
memory.

> storing it on the HD unless DDS starts clicking

DDS?

If this is a one time project, I am not sure of the rational for using
AWK also.  The ideas underlying its operation may be useful.

Defragmenting disk before loading big files will optimize reading
time.

Advice based on similar projects with much smaller files:

1. Proceed methodically in reasonable size steps and KEEP A LOG of
what you do.  For each step, list purpose, input file, test or
tranformation script (.py or whatever), and output files (results or
data).

2. If you can write as well as read tapes, store intermediate versions
of data files so you do not have to start over from original data if
you discover a goof several steps along the way.

3. As for testing: I would check that files really are in the format
specified before doing anything: line length correct, ids and dates
valid, etc.  I would start with a small subset and get each test
seemingly right before combining to test multi-gigabyte files.

Terry J. Reedy





From sholden at holdenweb.com  Mon Apr  1 17:13:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 17:13:08 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017680049.17235.python-list@python.org> <AP1q8.41430$pT1.1144020@news1.tin.it>
Message-ID: <TY4q8.118105$Ou.17770@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:AP1q8.41430$pT1.1144020 at news1.tin.it...
> Pearu Peterson wrote:
>         ...
> > users/developers/educators. From what I have learned last years about
> > Python development (other people comments being the main source), then
it
> > is not exactly a democratic process where majority rules.
>
> Absolutely not, *thanks be*.  I strongly suspect I wouldn't _want_ to use
> a language developed by "a democratic process where majority rules" -- cfr
> e.g. Larry Wall's quip "Perl is worse than Python because users WANTED
> it worse"... I think there's truth in it.  A single language architect's
> intuition, supported by advice and reasoned opinions of whoever he cares
> to listen to (as opposed to votes which are just counted, no matter who
> expresses them and what arguments, if any, they're bolstered by) stands
> a stronger chance of producing a gem than a committee's productions.
>
> I don't agree with every decision by the BDFL, but I'm still pretty happy
> that HE decides, not a committee -- and not even me (if I somehow felt
> sure that my decisions would be better, I could after all go elsewhere...
I
> doubt a better language would result that way, though).
>
Even if I thought my decisions would be "better" than Guido's (which they
would be only to me) I would nevertheless be glad that Guido was taking the
flak and not me. By my current estimate, this thread is accounting for
arounf a third of current c.l.py posting volume.

The bicycle-shed syndrome rules again!

regards
 Steve






From amuys at shortech.com.au  Thu Apr 11 02:59:06 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 10 Apr 2002 23:59:06 -0700
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at> <a92533$4bs$1@peabody.colorado.edu> <3CB4D5DF.4912EFE4@engcorp.com>
Message-ID: <7934d084.0204102259.715b6d4d@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3CB4D5DF.4912EFE4 at engcorp.com>...
> I chuckled to read this, mainly because as I read Siegfried's comment
> I was not in any way offended, and I am a professional programmer.
> It's not like we contribute much to world peace or solving world
> hunger or anything.
> 

Heh?  Speak for youself.... (as I continue to prepare a requirements
spec' for a werehouse inventory system for a local food charity) ;)

yes-I-will-probably-use-python-ly yours

Andrae Muys


From johnroth at ameritech.net  Sun Apr 21 07:52:39 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 21 Apr 2002 07:52:39 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
Message-ID: <uc59vi7etg1j2c@news.supernews.com>

"Carl Banks" <imbosol at vt.edu> wrote in message
news:tnos9a.de.ln at 127.0.0.1...
>
> Computer programming is not science.
>
> The study of algorithms, their properties, etc., is science.
> Implementing a quicksort algorithm is not science, but studying its
> properties is.

I would dispute that. The study of algorithms may very well
be a branch of mathematics, but mathematics is not science.

Science is, among other parts, the study of the natural universe
and things in it. The study of algorithms is a study of things
that have been created by the human mind.


> Don't even get me started on whether is computer engineering is
> engineering.

Legally, in the United States, it isn't. Anything with the word
"engineering" in its name is state regulated, by law.

John Roth





From kperacles at yahoo.com  Tue Apr  9 17:45:03 2002
From: kperacles at yahoo.com (Karl Putland)
Date: Tue, 9 Apr 2002 14:45:03 -0700 (PDT)
Subject: [Skunkweb-list] [ANNOUNCE] SkunkWeb 3.2 Released!
In-Reply-To: <20020409120627.A1201@localhost.localdomain>
Message-ID: <20020409214503.49367.qmail@web14208.mail.yahoo.com>

--- Drew Csillag <drew_csillag at yahoo.com> wrote:
> SkunkWeb 3.2 has just been released.
> 
> The release is available from the SkunkWeb home page at
> 	http://skunkweb.sourceforge.net
[...]
> What's new? (from 3.1.3)
> ------------------------------
[...]
> * Many PyDO improvements, including a SAP driver

Are there any changes to PyDO since the separate release of 1.1b1? 
I didn't see anything new in CVS view on SF.

--Karl

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/




From sheershion at mailexpire.com  Tue Apr  2 22:13:41 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Wed, 03 Apr 2002 03:13:41 -0000
Subject: wxpython: problem with events and checklistbox
References: <3caa2be5.48482250@news.easynews.net>
Message-ID: <Xns91E534FF3A655rcamesz@amesz.demon.nl>

Marcus Stojek wrote:

> in my GUI I have a wxCheckListBox (named CheckList) and an fuction
> IfChecked(self,event) that process the checkevents (EVT_CHECLISTBOX).
> 
> Now, when unpickling all data at the start of my program,
> I want to check some items with CheckList.check(index,true)
> My problem is, that after each checking the function IfChecked is
> called. In the docs I found something about PopEventHandler, but this
> doesn't seem to solve my problem. Again, this should be an all day
> problem, but I have no idea where to look.

Well the first hint would be to ask yourself if you really need that 
event handler in the first place. If it can't cope with the program 
setting or removing the checkmarks in the control, chances are it won't 
handle user input any better or more gracefully. You might want to 
reconsider your design.

But apart from that, there are two very simple solutions:

1. Don't set the event handler in the __init__ of your class, but do it 
after you've set all the checks you want.

2. Use a flag somewhere in your class, and check (in another sense of 
the word, no pun intended) that flag at the start of your event 
handler. Simply do a return if the flag indicates the event handler 
should be temporarily disabled. (You could even do event.Skip() prior 
to that return if you want that event to propagate to the parent 
window, but somehow I don't think that's needed here.)


HTH, Robert Amesz


From cliechti at gmx.net  Fri Apr 12 21:52:42 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Apr 2002 03:52:42 +0200
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de> <m31ydlndtd.fsf@chvatal.cbbrowne.com> <a980jr$9d07k$1@hades.rz.uni-sb.de>
Message-ID: <Xns91EF27A99F8C2cliechtigmxnet@62.2.16.82>

Uwe Schmitt <uwe at rocksport.de> wrote in news:a980jr$9d07k$1 at hades.rz.uni-
sb.de:
> Christopher Browne <cbbrowne at acm.org> wrote:
>| Uwe Schmitt <uwe at rocksport.de> wrote:
>|> I'd like to send "raw" data to a printer, how can I do this,
>|> using Python ???
> 
>| It's terribly platform dependent.
> 
> oops. i forgot to tell you that I'm using windows...
> and i really want to print "raw" data and not pdf et al...

never used that module, but it could fit for you problem :-)
 
"""
win32print.WritePrinter

int = WritePrinter( hprinter, buf )
Copies the specified bytes to the specified printer. Suitable for copying 
raw Postscript or HPGL files to a printer. StartDocPrinter and 
EndDocPrinter should be called before and after. Returns number of bytes 
written to printer.
"""

that module is part of "win32all". the module has of course all the 
functions to open a printer etc.

chris

-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Mon Apr 22 08:56:12 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 08:56:12 -0400
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com> <1019325450.422371@cswreg.cos.agilent.com>
Message-ID: <mOTw8.73436$T%5.33234@atlpnn01.usenetserver.com>

"Greg Weeks" <weeks at vitus.scs.agilent.com> rode in on his hobbyhorse to
assert...
> Some comments regarding "conceptual sameness":
>
>
> From: "Petr Prikryl" <Answer.via.news.please.prikryl at skil.nospam.cz>
>
> > There are two distinct things when comparing the objects: the object
> > content and the object identity.
>
> Neither of these concepts corresponds to conceptual sameness.  I haven't
> defined "conceptual sameness", but it is a natural concept:
>
If it's a natural concept then a definition shouldn't be too hard. Come on,
you can do it.

[spurious examples]
>
> Are these the same string?  Conceptually, yes.  When you are comparing
> strings in your code, it will be conceptual sameness that you are
> interested in.  But neither the identity nor the content of s1 and s2 are
> the same.
>
> So, conceptual sameness corresponds sometimes to sameness of identity,
> sometimes to sameness of content, and sometimes to neither.  And (I
> maintain), in ordinary user code, conceptual sameness is the concept of
> *identity* that is most useful.  (I stress "identity" because there are
> weaker forms of comparison that are sometimes useful.  But not *as*
> useful.)
>
Well, if you wave your hands long enough you might persuade some readers
that this all *means* something. But I'm not convinced you are yet at the
stage where you could provide a patch to add a "conceptual sameness"
operator to the language. Clearly object identity and object equality
*might* be equivalent for some types/classes.

At present Python, with its typically pragmatic approach to these matters,
does not insist on single-copy instances for immutable types. Consequently
it is possible for a == b to be true even though a is not b.

> So Python is missing THE single most useful comparision operator.  But
> there are workarounds.  "==" means conceptual sameness for all objects
> other than lists and dictionaries, provided that __cmp__ is appropriately
> defined.  Furthermore, if __hash__ is appropriately defined, any objec
> other than a list or dictionary can be a hash key.  So, except for lists
> and dictionaries, everything's fine.
>
I don't feel you have yet even adequately *defined* "conceptual sameness",
so it's surely premature to assert that it's missing from the language.

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From mcfletch at rogers.com  Mon Apr 15 15:59:19 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 15 Apr 2002 15:59:19 -0400
Subject: Tuple/List identity guarantees Was(Re: __iadd__ and fellows missing
 (Python 2.2))
References: <LNBBLJKPBEHFEDALKOLCMEPOONAA.tim.one@comcast.net> <mailman.1018763429.14619.python-list@python.org> <pqEu8.14418$Ji7.9663@atlpnn01.usenetserver.com>
Message-ID: <3CBB3117.7090309@rogers.com>

This little mental exercise all started when Tim pointed out the tuples 
_won't_ work for the pattern (well, he pointed out that there's no 
guarantee that imutable values (particularly empty tuples) will always 
be seperate objects).  So using an empty tuple for the pattern is no 
longer suitable.

To the best of my knowledge, there _is_ an implicit guarantee (at least, 
such a guarantee is assumed in a lot of code) that two different 
_mutable_ objects (such as lists or instances) will _always_ return 
different id values, and thus always compare unequal with "is".  Sure, 
there are optimisation tricks (copy-on-write) that could invalidate such 
a guarantee, but somewhere under the covers the implementation needs to 
be tracking the "real id" of the list to implement those semantics, so 
the implementation can still return that id should someone ever 
introduce those optimisations.

Note: the "second case" in your list is something that's explicitly 
out-of-scope for this pattern, not impossible, but not something I'm 
interested in making easier.  If you want to specify, do so, if not, 
just leave the argument out.

Oh, I care deeply for the Timbot's feelings.  Six years of fighting a 
bloody guerilla war against him ;) has made me come to love the little 
floating ball of light :) (though the strange looking guy that always 
hangs out around him kinda freaks me out ;) ).

Incidentally, trying to figure out how long we've been trading volleys, 
I found an article from just about a year ago:

http://groups.google.com/groups?selm=mailman.987583046.24994.python-list%40python.org
	
At the time I was so burried in work I guess I didn't file that lack of 
guarantee away in active storage.  Though, checking my code to change 
all the NULL = () to NULL = [], I only found a single case (written 
_before_ that message) where I had used NULL = () rather than NULL = [], 
so I suppose at the code-writing level it must have lodged somewhere. 
Hmm, there's hope you might change me yet, Tim :) .

Reminds me, where have all the optimisation challenges gone these days? 
  Tim, Christian and I used to spend whole minutes on those babies, and 
now all we can do is bat around recycled debates on object identity :) .

Have fun,
Mike


Steve Holden wrote:
> "Mike C. Fletcher" <mcfletch at rogers.com> wrote ...
...
>>NULL = []
>>
> 
> Note that this does indeed cause the NULL variable to refer to a specific
> instance of the empty list (and a tuple would be acceptable here rather than
> a list).
> 
...
> The only situation this will not help you with is if you need to distinguish
> between the following calls:
> 
>     BasicProperty("One")
> 
>     BasicProperty("One", defaultValue=NULL)
> and
>     BasicProperty("One", defaultValue=())
...
> The third case is trickier, however, as there is no guarantee that the
> implementation won't optimize, so it is permissible for different
> implementations to use either the same or different lists (or tuples).
...
> As for the "designed purpose of default arguments" :) .  I see the value
>> "no value/not specified" as a perfectly suitable default value.  Sure,
>> it may offend your sensibilities, but when has that stopped me in the
>> past ;) .
>>
> 
> Apparently never, if the above is typical of your disregard for the bot's
> feelings.
> 
> regards
>  Steve
...





From mcfletch at rogers.com  Sat Apr 13 18:11:48 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 13 Apr 2002 18:11:48 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCOEOBONAA.tim.one@comcast.net> <mailman.1018723227.15055.python-list@python.org> <V01u8.27065$043.2467943@bin6.nnrp.aus1.giganews.com>
Message-ID: <3CB8AD24.707@rogers.com>

Using an unlikely-but-not-guaranteed static value such as "OpTioNaL ArG 
NoT GiVeN" for NULL makes the software's reaction dependent on the data 
it's processing (and the version of the interpreter it's running on (if 
you're  using is for comparisons), string-(or, apparently, 
tuple)-identity check is essentially a random/upredictable event).


Consider, for instance:

	an instance of a class that compares == to anything (will compare == to 
the string), so you need to use is for identity checking.

	objects that blow up when you access any attribute of them, or which take 
hours to access a given attribute (e.g. __cmp__)

	some evil user decides to put "OpTioNaL ArG NoT GiVeN" into the system, 
since the identity of that argument is entirely dependent on the 
interpreter (it may or may not be the same object as the default arg), 
you have a "random" reaction to the data.


Using a true singleton instance or even a list means that the system's 
operation is not dependent on the data at all (don't need to use ==, 
doesn't have a particular value that causes unexpected results).

Enjoy,
Mike

Terry Reedy wrote:
> "Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
...
>>  Suppose I'll wind up using singleton instances instead.
> 
> 
> Given that None is a possible input, so it cannot be used as the
> no-input flag, you could also try something like NULL="OpTioNaL ArG
> NoT GiVeN"
> 
> Terry





From peter at engcorp.com  Sun Apr 21 22:05:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 21 Apr 2002 22:05:06 -0400
Subject: Pyxie, anyone?
References: <Xns91F7F276A63DAgustafl@195.100.94.182>
Message-ID: <3CC36FD2.E98469F0@engcorp.com>

Gustaf Liljegren wrote:
> 
> I have problems using the pyxie library for processing XML. This page
> 
>   http://pyxie.sourceforge.net/
> 
> says that "To run Pyxie, you need two things:
> 
> - A Python distribution
> - The XML package for Python [...]"

And the Pyxie library itself.

Three.  There are _three_ things needed to run Pyxie....

-Peter


From mparaz at yahoo.com  Thu Apr 11 09:08:20 2002
From: mparaz at yahoo.com (Miguel A Paraz)
Date: 11 Apr 2002 06:08:20 -0700
Subject: Linking against python library causes core dumps
References: <c9f472a0.0204090306.7fb516ea@posting.google.com> <mailman.1018351106.14249.python-list@python.org>
Message-ID: <c9f472a0.0204110508.38581b2a@posting.google.com>

Oleg Broytmann <phd at phd.pp.ru> wrote in message news:<mailman.1018351106.14249.python-list at python.org>...
>    Looks like a problem with threaded Python with non-threaded freeradius.
> Or vice versa.
> 
> Oleg.


Python is threaded:
$ ldd /usr/bin/python | fgrep pthread
	libpthread.so.0 => /lib/i686/libpthread.so.0 (0x40028000)

freeradius is threaded by default.

freeradius works OK if I statically link only my python module, and
let the rest be dynamically loaded.

I tried rebuilding freeradius without threads (configure
--with-threads=no) and it's the same.

So, the problem is no big deal... I'm just curious since it sure is
odd.


From eppstein at ics.uci.edu  Mon Apr 29 02:53:31 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 28 Apr 2002 23:53:31 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <eppstein-DF7988.23533128042002@news.service.uci.edu>

In article <aaim48$7fd$1 at bob.news.rcn.net>,
 "Raymond Hettinger" <python at rcn.com> wrote:

> The functionals (map, filter, and reduce) had greater importance 
> prior to the introduction of list comprehensions which are now the 
> preferred (and more readable) approach.

Map and filter, I can see substituting with list comprehensions.
But can you tell me how to use a list comprehension in place of reduce?

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From Answer.via.news.please.prikryl at skil.nospam.cz  Thu Apr 11 03:46:00 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Thu, 11 Apr 2002 09:46:00 +0200
Subject: True/False story... (PEP 285)
References: <a8u99o$2lcu$1@news.nextra.cz> <vHys8.62333$pT1.1792330@news1.tin.it> <3CB2E617.8050600@geneva-link.ch> <a8vaba$cnv$1@slb3.atl.mindspring.net>
Message-ID: <a93f3p$rcg$1@news.nextra.cz>

Boris wrote...
>
> Alex Martelli wrote:
> >
> > there IS a natural mapping of booleans onto a subset
> > of integers, just as there is a natural mapping of
> > integers onto a subset of reals (or rationals, etc).

But does it help somehow?

I do not know from what thread the Alex's notice was
cut, and what was the point.  I agree that there is
natural mapping between booleans and integers.  There
even is natural mapping between natural numbers and any
data that is used by computers (the RAM can be
viewed as a single, very big natural number, the same
way the disk space can be interpretes, but it does not
makes the things easier).

The same way, you probably would not be happy if someone
decided to remove strings and to replace them by
sequences of integers.

To summarize, bool type is related to a different
abstraction than the integer type.  Bool is the most
natural type (conceptually) for capturing
conditional-expression results.

Moreover, strict boolean does not require ordering of
False, True.  In other words, the only comparison
operators are '==' and '!='.  Then, the mapping of
boolean into integer adds some implementation details
that should not be abused if one want to preserve the
original meaning of the booleans -- this is the problem
that was unfortunately introduced into C/C++/Python
implementation of conditional expressions at the
beginning.

Boris...
>
> Could be argued that there is no single natural mapping,
> for an alternate to the current one we could have
>
> True     == 1
> False    == -1
> not x    == -x
> x and y  == min(x,y)
> x or y   == max(x,y)
> x xor y  == -x*y
>
> A nice feature of that one (to my aesthetics), is that
>
> 1j == sqrt(False)

This alternate mapping breaks everything that exists now
and does not introduce anything better than the existing
state.

Andrew wrote...
>
> Quoting Hilbert,
>
>    "It must always be possible to substitute 'table',
>     chair' and 'beer mug' for 'point', line' and 'plane'
>     in a system of geometrical axioms."

;-)

> I personally started with a BASIC which had 0 for false
> and -1 (all bits on) for true.

In comparison with Python (C/C++), this is only a minor
*flavour* of the same.  Well, I were used to 0 and !=0
in C/C++, but I prefer to think in booleans.  And I also
always prefer new bool in C++ since it was introduced
into the language.

Petr

P.S. The HTML version of the original
     "True/False story... (PEP 285)" is temporatily available
     at http://www.skil.cz/test/TrueFalseStory.html
--
Petr Prikryl (prikrylp at skil dot cz)




From aahz at pythoncraft.com  Thu Apr 11 21:18:48 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 21:18:48 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <Xns91ECE5A88C0F8RASXnewsDFE1@130.133.1.4> <3CB523FD.7D83C0C@engcorp.com> <mailman.1018555883.18567.python-list@python.org>
Message-ID: <a95clo$a4g$1@panix1.panix.com>

In article <mailman.1018555883.18567.python-list at python.org>,
Geoff Gerrietts  <geoff at gerrietts.net> wrote:
>
>If you need references, I can provide an ex-wife and several people I
>used to play MUSHes with who will be MORE than happy to take potshots
>at my character. ;)
>
>I like to think they're wrong; please don't disabuse me of my
>delusions. ;)

Heh.  I've got plenty of people to take potshots at my character, and I
*know* they're correct.  But I seem to have enough positive attributes
that enough people like me anyway.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From aleax at aleax.it  Fri Apr 26 17:11:20 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 21:11:20 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <mailman.1019184878.1733.python-list@python.org> <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com> <mailman.1019212237.13240.python-list@python.org>
Message-ID: <Ynjy8.66347$vF6.2029617@news2.tin.it>

Laura Creighton wrote:
        ...
> If you are willing to wait, Alex's own Python in a Nutshell will be
> published by O'Reilly soon.  In this particular case, I think the
> Nut is more of a coconut than a pistachio nut :-).  Its not PP2E sized,

Nothing a swallow can't easily carry...


Alex



From gbreed at cix.compulink.co.uk  Mon Apr  8 10:18:38 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Mon, 8 Apr 2002 14:18:38 +0000 (UTC)
Subject: Discussion of Python midi library
References: <3CB19732.8060304@mxm.dk>
Message-ID: <a8s8ru$qej$1@thorium.cix.co.uk>

Max M wrote:

> I am writing a Midi Module for Python.
> 
> I basically want it make it possible to:
> 
>      1) Read midi from a file
> 
>      2) transform the midi data
> 
>      3) Write midi to a file
> 
>      4) Play midifiles (especially on Windows)

Hey, that's good, I'd like to do those things as well.

> The biggest challenge here is getting the internal representation of 
> the midi data right.

You mean getting it to match the standard, or encapsulating it in the most 
convenient way?

> The midi data format in it's raw format is too specialized to make it 
> possible to load some midi-data and then do any kind meningfull 
> transformations on it.
> 
> So generally I want to be able to load in the data into a Python 
> specific objects representation that can be easily modified and then 
> saved back out as a new midi file.

I'd like it to be generic enough that the code could be re-used to 
transform live MIDI streams as well.

> I am doing algorithmical composition hence the need.
> 
> Has anybody got an interrest in participating in a discussion about how 
> such a module should be designed please contact me.

I think a few of us are interested, but we may not have the time to 
contribute much.  I did get half way through converting my old C++ code, 
but I don't think it's really good enough for anybody else to work with.  
Unless you want an example of something that can read and write the files 
successfully.

> I allready have some code and some ideas.

Where?


                      Graham

         <http://www.microtonal.co.uk/>


From python at rcn.com  Wed Apr 17 10:10:49 2002
From: python at rcn.com (Raymond Hettinger)
Date: Wed, 17 Apr 2002 10:10:49 -0400
Subject: list_a-list_b
References: <mailman.1019022993.8455.python-list@python.org>
Message-ID: <a9jvr2$io5$1@bob.news.rcn.net>

[David Eppstein]
> > Better is to convert b to a dictionary
> > (where are dict comprehensions when you need them):
> >
> >     dict_b = {}
> >     for x in list_b: dict_b[x] = 1

[Tim Peters]
> Looks clear enough to me, although
>
>     for x in list_b:
>         dict_b[x] = 1
>
> is clearer.  If you like, you can do this:
>
>     dict([(x, 1) for x in list_b])
>
> or this:
>
>     dict(zip(list_b, [1]*len(list_b)))
>
> or even:
>
>     dict(zip(list_b, list_b))

My favorites:

dict(map(None, list_b, []))   # doesn't loop over list_b twice

or

dict_b = {}
map(dict_b.__setitem__, b, [])   # fastest for Py2.2


Raymond Hettinger




From loewis at informatik.hu-berlin.de  Fri Apr 12 11:33:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 12 Apr 2002 17:33:27 +0200
Subject: 2.1 stable and 2.2 dev ?
References: <20020412162608.22381c75.wilk@flibuste.net> <mailman.1018622545.19707.python-list@python.org>
Message-ID: <j41ydlszi0.fsf@informatik.hu-berlin.de>

Gustavo Niemeyer <niemeyer at conectiva.com> writes:

> > I would like to know if the version 2.1.x is the stable version and the
> > 2.2.x the dev version ? If it's like that, it's oposite of linux version
> > where odd number are for dev...
> 
> Right now, there's no such concept. 2.2.X *and* 2.1.X are stable
> releases, while (e.g.) 2.2.1aX (alpha), 2.2.1bX (beta), and 2.2.1cX
> (release candidate) were development.

More importantly, the "current" development source code is in CVS, and
carries no number (at the moment, it identifies itself as 2.3a0, but
that does not name a specific release). For Linux, Linus *releases*
development versions; a similar thing does not exist for Python
(releases are ongoing, through CVS).

Regards,
Martin



From coyote at step.polymtl.ca  Fri Apr 12 06:28:32 2002
From: coyote at step.polymtl.ca (Vincent Cojot)
Date: Fri, 12 Apr 2002 10:28:32 GMT
Subject: Problem in 'Doc' building in 2.2.1 (OK in 2.2.1c2)
Message-ID: <Pine.GSO.4.44.0204120620180.13652-100000@step.polymtl.ca>

Hello everyone,

	Python 2.2.1 (release) appears to be missing something in its
documentation, thereby preventing a sucessfull build of the Documentation:

[root at toffee Doc]# gmake -n
gmake: *** No rule to make target `whatsnew/whatsnew22.tex', needed by
`html/whatsnew/whatsnew22.html'.  Stop.

This problem isn't present in 2.2.1c2 because the Make dependency for
the whatsnew stuff isn't there..

Additionnally, is there -anything- that helps latex2html -not- to complain
about "dots" in the command name:

[...snip...]
Loading /usr/share/latex2html/versions/latin1.pl
 *** Fatal Error --- but easy to fix ***
Cannot have '.' in file-name prefix, else dvips fails on images
Change the name from  /root/Python-2.2.1/Doc/doc/doc.tex  and try again.
[...snip...]

The workaround is of course to renamed /usr/local/src/Python-2.2.1 into
/usr/local/src/Python-221 but it's not likeable..

Thanks for reading... (Credits to Cedric Karm for complaining about
this.. :) )

,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'`^`'~*-,
Vincent S. Cojot, Computer Engineering. STEP project. _.,-*~'`^`'~*-,._.,-*~
Ecole Polytechnique de Montreal, Comite Micro-Informatique. _.,-*~'`^`'~*-,.
Linux Xview/OpenLook resources page _.,-*~'`^`'~*-,._.,-*~'`^`'~*-,._.,-*~'
http://step.polymtl.ca/~coyote  _.,-*~'`^`'~*-,._ coyote at step.polymtl.ca

They cannot scare me with their empty spaces
Between stars - on stars where no human race is
I have it in me so much nearer home
To scare myself with my own desert places.       - Robert Frost





From skip at pobox.com  Wed Apr 10 12:46:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 10 Apr 2002 11:46:18 -0500
Subject: Transferring files using sockets
In-Reply-To: <6ac1ecad.0204100826.25217b0a@posting.google.com>
References: <6ac1ecad.0204100826.25217b0a@posting.google.com>
Message-ID: <15540.27738.975194.101310@beluga.mojam.com>

    IC> I am trying to transfer a file using the socket module, but I'm not
    IC> sure how to do this. Should I split the file into little(1KB)
    IC> packages and then transfer them, or is there a possibility to send
    IC> the whole file?

It should be possible to just execute

    sock.send(open("somefile").read())

if sock is a connected stream socket.  You might want to break it down in
chunks though, either to conserve local memory or provide transit feedback:

    f = open("somefile")
    while 1:
        data = f.read(8192)
        if not data:
            break
        sock.send(data)
        sys.stderr.write(".")
        sys.stderr.flush()
    sys.stderr.write("\n")

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From cyberlync at yahoo.com  Wed Apr  3 08:06:43 2002
From: cyberlync at yahoo.com (Eric Merritt)
Date: Wed, 3 Apr 2002 05:06:43 -0800 (PST)
Subject: C Extension
In-Reply-To: <m3wuvpntr1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020403130643.14790.qmail@web13003.mail.yahoo.com>

> Then something is clearly wrong - either tp_repr or
> tp_str will be
> called, unless you've implemented tp_print (which
> you don't need to).
> If those *aren't* called, this is an indication that
> it either crashes
> before getting that far, or that you managed to get
> out of synch in
> your type object - double-check that every function
> is in the right
> slot (for this, it really helps if you don't need to
> cast the function
> pointers).
After sleeping on it I am thinking that python
actually deallocs the object before I get to the point
printing it. I need to reread the reference counting
docs and make sure. I will check the function
references as well.

Thanks,
Eric

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/



From borcis at geneva-link.ch  Mon Apr  8 13:42:09 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 08 Apr 2002 19:42:09 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com> <eppstein-CB6351.08534907042002@news.service.uci.edu> <ub315bi2fkdube@news.supernews.com> <3CB1E00C.1DBA7DE4@alcyone.com>
Message-ID: <3CB1D671.F5012473@geneva-link.ch>

Erik Max Francis wrote:
> 
> Indeed, even in mathematics, the cardinal number that is represented
> with "1" (that is, the cardinality of of a set with a single element) is
> _not_ the same thing as the integer 1.

Well, since his name is mentionned in your sig, I don't think
Albert Einstein would have agreed with the x of it as a child.

Boris
--
"12 ? - the least integer symbolizing N by itself. Successor: 123"


> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.

Cosmetic, cosmic, comic, cmc, mcc, mc^2, E = Albert (tm)

> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.


From akuchlin at ute.mems-exchange.org  Tue Apr 30 13:54:02 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 30 Apr 2002 17:54:02 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net> <mailman.1020176703.26327.python-list@python.org> <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com>
Message-ID: <slrnactmhq.thl.akuchlin@ute.mems-exchange.org>

In article <ELyz8.16820$xy4.289 at atlpnn01.usenetserver.com>, Steve Holden wrote:
> If you haven't come across the effbot before, you should know that he tends
> to be rude in this way. I think it happened when the parity unit was fitted

He's not rude, just telegraphic in his posts.

--amk


From merman at freenet.de  Mon Apr  8 16:30:27 2002
From: merman at freenet.de (merman)
Date: Mon, 8 Apr 2002 22:30:27 +0200
Subject: swig-problem
References: <3cb1ae14$0$12336$9b622d9e@news.freenet.de> <ub3cbpdmtl65f2@corp.supernews.com>
Message-ID: <3cb1fbe2$0$146$9b622d9e@news.freenet.de>

"Lyle Johnson" <ljohnson2 at knology.net> schrieb im Newsbeitrag
news:ub3cbpdmtl65f2 at corp.supernews.com...
> Use the "-o" flag for SWIG to change the output file name:
>
>     swig -c++ -python -o Tree_wrap.cpp Tree.i
>
> Hope this helps,
>
> Lyle

Hi Lyle,

thanx for your effort - very polite!

It was a little bit better with your hint but there is maybe an additional
linker-problem (lots of failures) - I don't know why - puh.
I continue my work tomorrow - I fear I go crazy ;-).

best regards

thomas





From imbosol at vt.edu  Sun Apr 21 14:12:16 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 21 Apr 2002 14:12:16 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com>
Message-ID: <0evu9a.621.ln@127.0.0.1>

Allan Crooks wrote:
> Hi,
> 
> I was wondering what the best approach is on providing libraries for
> varying versions of Python?
> 
> My aim is to distribute a module that will provide a set of core
> functions, but provide more functions to later versions of Python.
> 
> For example, I make a module that works on 2.0, but if used on 2.1,
> will provide more functions to say, do with the warning framework. If
> used on 2.2, certain functions that return lists in earlier versions
> of Python, will use generators instead.
> 
> Rather than keeping several versions of the same module (but for use
> with different versions of Python), I was wondering if there was a
> more elegant solution?


# mymodule.py
"Does different stuff depending on what version is used."

# Stuff common to all versions goes here

# Import stuff for version 2.1
if sys.version[:2] >= (2,1):
    from mymodule2_1 import *

# Import for version 2.2
if sys.version[:2] >= (2,2):
    from mymodule2_2 import *

You still have to use several modules, but you don't have to maintain
multiple copies of code common to all the modules, and the same files
work with any version.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From akuchlin at crystal.mems-exchange.org  Thu Apr 11 13:35:41 2002
From: akuchlin at crystal.mems-exchange.org (A.M. Kuchling)
Date: 11 Apr 2002 17:35:41 GMT
Subject: Notes on 2.2's changes
References: <slrnabba3h.h8h.akuchlin@crystal.mems-exchange.org> <Xns91ED5ED8EF7A0lumberjack@216.148.53.96>
Message-ID: <slrnabbibb.iac.akuchlin@crystal.mems-exchange.org>

In article <Xns91ED5ED8EF7A0lumberjack at 216.148.53.96>, Lumberjack wrote:
> hardship. It is impossible to tell from your list how much _time_ (and 
> money) was consumed in locating and fixing those problems. Basically, you 
> haven't provided data on the more important metric. 

The multi-argument .append() and .connect() calls take a few minutes
to fix once the test suite catches them.  After asking him, Neil said
fixing ptl_compile.py takes around 20 minutes.  Grouch is likely the
messiest case, as Greg may have to spend lots of time crawling up and
down in his code, but I don't know how much time those fixes took.

> And it isn't clear why 
> you felt it was necessary to be compatible with 2.2. Wouldn't it have been 
> simpler to just say "run these on version 2.0.x of Python"? 

This software runs on our Web server, and on servers we control, so we
can run any version we like, if we decide the porting effort is worth
the additional features and bugfixes.  New features sometimes let us
write cleaner code; generators are used in a few places,
unsurprisingly, as are the new email package.  My holy grail is to
someday use a version of the ZODB that doesn't use ExtensionClass,
which seems possible with 2.2.  (ExtensionClass is far, *far* more
annoying than anything that 2.1 or 2.2 has changed; it means we can't
run PyChecker, and can't use __r*__ overloading or and had to work
around isinstance() being broken for ExtensionClass in 2.1.)

The large code base is also a decent test case for Python itself.  We
ran 2.2.1c1 and c2 on our developer machines, though not on our
production servers, and it exposed at least one bug that I could then
fix for 2.2.1final.

> The alleged 
> underlying philosophy seems to be that you can pick whatever version 
> that suits you and stick with it; you are under no compunction to track 
> changes to Python. 
> So why track it if there is no compunction? I'd like to 
> see you address that.

What's to address?  "Because we wanted to" is surely reason enough, if
we're willing to pay the corresponding cost in time.

--amk                                                             (www.amk.ca)
It seemed like the night sucked them up, took them into its dark heart. It
seemed like the darkness swallowed them... Perhaps it did.
    -- The convention breaks up, in SANDMAN #14: "Collectors"


From tim.one at comcast.net  Sat Apr 27 01:18:06 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 27 Apr 2002 01:18:06 -0400
Subject: String handling bug in Python
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192215166@admin56.narex.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEGMPBAA.tim.one@comcast.net>

[Tim]
>> r-strings were primarily intended to ease writing regular
>> expressions (which can never end with an odd number of
>> backslashes), and people using \ in Windows pathnames are
>> usually missing an obvious alternative (like the forward
>> slash -- Windows API functions couldn't care less which kind
>> of slash you use; and portable code should avoid any literal
>> slashes via os.path.join()).

[Bjorn Pettersen]
> It seems a shame to limit such a potentially useful feature to only one
> domain.

It isn't.  Ask, and people will tell you it's useful in several other
domains too.  The only people who ever whine about this are Windows
programmers who should know better.

> Believe me, I get your windows comment, but when you work in a purely
> windows shop what "should" be done to make code portable doesn't fly
> unless it's also shorter, easier, and makes the code faster <wink>.

Forward slashes are easier for almost everyone to type than backward
slashes, and work fine on Windows.  Paths using forward slashes also run
13-17 times faster in Python code, due to Python's Unix heritage <wink>.

> At-least-they-provide-us-with-free-beer'ly y'rs

while-i-provide-self-serving-lies-that's-windows-for-you-ly y'rs  - tim





From Oschler at earthlink.net  Mon Apr  1 00:14:07 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 00:14:07 -0500
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <mailman.1017628505.7732.python-list@python.org>
Message-ID: <LURp8.63151$Q7.49997488@e3500-atl1.usenetserver.com>

Gerhard,

True, beauty is at the behest of the be-renderer (ouch, that was bad).  But
I'm still a junkie for drag and drop form drawing.  Regarding your tag line
(this sig powered by Python).  Might want to add "so is Google and Inktomi".
Frick, I was astounded when I read that in "Web Programming in Python".

I'm going to look at Open Office's and Mozilla's HTML editor.  If that
doesn't work out then probably WebSphere.  I'd like to try Quanta or
Bluefish, but if you really scour the newsgroups, its hard to find strong
opinions on either and they don't seem to be getting a groundswell of
followers (I could be _very_ wrong on this, just going by my surfing and NG
posting results).

thx

>
> That, and Mozilla seem to be your only options for Linux-native WYSIWYMG.
>
> Gerhard
> --
> This sig powered by Python!
> Au?entemperatur in M?nchen: 6.3 ?C      Wind: 0.5 m/s
>





From mertz at gnosis.cx  Tue Apr 16 12:19:19 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Tue, 16 Apr 2002 12:19:19 -0400
Subject: xml2dict - dict2xml
References: <a9gudi$964$04$1@news.t-online.com>
 <3CBC3E59.5090908@thomas-guettler.de>
Message-ID: <H8Ev8kKkXQUV092yn@gnosis.cx>                

|Henning Peters wrote:
|> i am searching for the best solution to convert a small xml fragment in a
|> python dictionary and vice versa.

Thomas Guettler <zopestoller at thomas-guettler.de> wrote previously:
|http://www-106.ibm.com/developerworks/xml/library/x-matters11.html

A somewhat more up to date source than my abovementioned article is:

    http://gnosis.cx/download/Gnosis_XML_Util.ANNOUNCE

Our xml_pickle utility converts (very nearly) any Python object to XML,
and can restore it back to a Python object (including a dictionary).
What xml_pickle handles should be a strict superset of what pickle.py
handles, since at very worst, we can fall back to inserting a pickle
fragment inside an element (which isn't very XML-ish, but also doesn't
happen except in very strange cases).  The XML format is not quite the
same as the prior example, however.  It wasn't clear to me whether the
main issue was the Python dictionary or the specific XML format.  What
you get looks like:

    >>> from gnosis.xml.pickle import dumps
    >>> dct = {1:2, 'this':'that'}
    >>> print dumps(dct)
    <?xml version="1.0"?>
    <!DOCTYPE PyObject SYSTEM "PyObjects.dtd">
    <PyObject family="obj" type="builtin_wrapper"  class="_EmptyClass">
    <attr name="__toplevel__" type="dict" id="799116" >
      <entry>
        <key type="string" value="this" />
        <val type="string" value="that" />
      </entry>
      <entry>
        <key type="numeric" value="1" />
        <val type="numeric" value="2" />
      </entry>
    </attr>
    </PyObject>

Usage is simple.  If you need specific XML tags and structures, however,
you'll probably need to write your own code (or customize ours).

xml_pickle now optionally uses SAX instead of DOM for underlying
parsing, which is pretty fast.  Henning Peters' simultaneous statements
that speed was essential, and that small dictionaries were at issue,
seems slightly strange.  Certainly for any "small" dictionary,
xml_pickle is pretty much too fast to accurately time.  But the point of
xml_pickle is not speed... a custom application, probably especially
using PyRXP, would be faster (but might not handle new cases one
discovered with usage).  I suspect the speed issue is actually a red
herring though, and most any approach will be "fast enough".  FWIW:  if
you want to transmit or store an xml_pickle file, we have a compression
option, which produces output that is smaller than binary pickles!

Yours, David...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From paul at boddie.net  Tue Apr 30 04:33:04 2002
From: paul at boddie.net (Paul Boddie)
Date: 30 Apr 2002 01:33:04 -0700
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net>
Message-ID: <23891c90.0204300033.16549c52@posting.google.com>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote in message news:<slrnacrltr.6mq.jon+usenet at snowy.squish.net>...
> In article <mailman.1020119995.30995.python-list at python.org>, Fran?ois Pinard wrote:
> > 
> > Strange.  For more than one year now, we never had the shadow of a
> > problem with HTML designers.  Everybody is competent in his field and
> > respectful for the work of others.  There is no "we are good they are
> > rotten" speak here.
> 
> That's not what I meant. I'm not saying that designers are
> incompetent, I'm saying that having to work around code is not part of
> their job and the *tools* used will usually make it impossible.

Some tools do preserve not-strictly-HTML sections of files, but that
isn't to say that the designers will necessarily use such tools.
Moreover, they could just decide to produce something brand new rather
than work on existing files, or instead capture a "screenshot" (save
the Web page source in their browser) and work from that, where none
of the "structural" (template code) parts will be present.

I've had experiences where non-developers have just presented a new
concept and wanted it implemented from scratch. While I've seen talk
of the "webmaster bottleneck" on some page or other (I was working on
the Web Modules Overview last night, so I looked at a lot of template
system descriptions), I can imagine that quite often it is the
development team who is overloaded with writing code, maintaining the
application *and* writing other people's HTML for them.

> I'll just leave you with one further point. Several times on projects
> I have worked on, the customer decided upon a near-total redesign of
> an existing site. This involved us receiving completely new HTML files
> designed from scratch. With *any* system involving embedded code, this
> necessities many days of manual re-integration of the code and HTML.
> With something like jonpy, it's the work of a few hours at most.

If you can offload the HTML editing job on someone else (who may well
be less busy) and make the task of integrating their work back into
the application easier, then it's a distinct advantage. Moreover,
changes to the data model are likely to be easier to manage, requiring
less head scratching and staring at embedded code to work out what has
to change, but that's just my opinion.

Paul


From loewis at informatik.hu-berlin.de  Fri Apr 19 05:48:06 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Apr 2002 11:48:06 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <Pine.GSO.4.40.0204181736420.16636-100000@neelix>
References: <Pine.GSO.4.40.0204181736420.16636-100000@neelix>
Message-ID: <j4g01suii1.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> OK, how about this for a suggested patch:

Thanks, applied as Setup.dist 1.27.

Regards,
Martin




From aleax at aleax.it  Sun Apr 28 18:25:56 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 00:25:56 +0200
Subject: Update locals()
In-Reply-To: <20020428205757.GH16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke> <200204281716.TAA10331@prim.merlinux.de> <20020428205757.GH16217@devel.trillke>
Message-ID: <02042900255602.25256@arthur>

On Sunday 28 April 2002 22:57, holger krekel wrote:

	[trying to compact this down again]

[about locals() and globals()]

> Are you sure that this holds?

You're right in your quotes from the library manual and I can't easily
find out the various spots from where, collectively, I formed a much
stronger opinion (could be python-dev messages, comments in code,
etc).  If this matters to you I suggest you ask for official clarification
(e.g. on sourceforge) -- e.g., is it just a current implementation accident 
that locals()['x']='y' either changes the local namespace or doesn't but 
never raises exceptions, or is it a guarantee for future language versions. 
I believe the former holds.

> scope. But anyway we certainly agree that "exec" without optional
> arguments *does* execute in current scopes (no matter what behaviour the
> objects returned by locals()/globals() have). We are only arguing whether
> there is some good use for execing in current scopes.

Yes.  You brought up the irrelevant aside that locals and globals could be
passed anyway, even if one did require always passing a dictionary, and
I wanted to point out that this wasn't the case (under semantic 
equivalence).


> Hmmm. certainly a possible update-path.
> I really like about python that you can refer to classes,
> object's attributes/methods, the global/local bindings
> as objects and work with it. This is (one of) the goal

I agree _except_ for the local and global namespaces -- because [a] you 
can't (see above), [b] it's best that you can't.

> that many other current languages are missing. Nobody says that you don't
> have to care about any side effects this may and will have.

Sorry, but the last sentences is totally mysterious to me.


> > def addMethods(class_, filters, statefulList):
> >     for name in statefulList:
> >         delegatee = getattr(filters, name)
> >         def method(self, *a, **k): return delegatee(*a, **k)(self)
> >         method.__doc__ = getattr(filters, name, undoc).__doc__.strip()
> > + '\n' setattr(class_, name, method)
>
> Thanks for this transformation. But of course a 'but' follows :-)
> As soon as
>
> - you want to adapt the definition to a filter's method signature
>
> - have templated or parametrized code in the function body
>
> - want to see what actually gets defined
>   (with my version i can easily write or print the
>    actual function's definition and look at a real definition)
>
> it might get difficult or impossible to do it your way.

It might, but I asked you to provide an example, you did, and I recoded
it in a way that I consider preferable (without any exec or eval).  If you
want to continue this exercise, despite its proving rather sterile, please
keep providing examples for me to recode, rather than handwaving
generalizations as in those first two points.

Point three is murky to me.  "The actual function definition" you can't
either see or print with your method -- it's hidden somewhere in
filters, and all your method is doing is delegating to it.  (Maybe you
mean something other than 'method' when you say 'method', which
is after all a specific technical term in Python, but then you'll have to
rephrase, because even mentally translating e.g. 'method' into
'approach' this is still quite unclear).

Source code can be seen or printed at runtime without needing it to
come from a string exec'd with default dictionaries, as is pretty
obvious.  So, if this point has any importance, you should restate it.


> Writing a reusable helper function i could say something like
>
>     exec make_adapters(filters, statefulList, '''
>     def %name (self, %args):
>         """%__doc__"""
>         return %sourcescope.%name (%args) (self)
>     ''')
>
> or even better make the templated adapter definition a defaulted
> argument. This provides more flexibility for the *caller* of the helper
> function without ever having to modify the helper function.

I _thought_ I knew what the term "unmaintainable" meant.  This, however, 
takes this important term to levels I had never dreamed of.  Any bug at
all in this complicated, tightly-coupled nest of function and data, any 
little discrepancy, will be hell to find out and fix.

Even if I did need to have a function take a string template and make
code from it, I would never have it return strings to be exec'd in such
a way.  Rather, I'd return a dictionary (or sequence of pairs) with
names and function objects, and install them as needed caller-side
(again in a metaclass or after the class statement), either inline or in
an auxiliary function -- e.g. supposing inline, and a defaulted string
template:

    for name, function in make_adapters(filters, statefulList):
        setattr(filename, name, function)

THIS gives flexibility to the caller (which may easily install the 
functions in different objects and/or with changed names as need be)
AND to function make_adapters (which may generate the needed
functions in any of several different ways, rather than being constrained 
to returning a string -- strings of sourcecode are not the best ways to
sling code around in a language where functions and classes are first-class 
objects).  I suspect this might also be slightly more maintainable (although
not GOOD in this way, maybe not quite as horrible as the version based
on exec-in-local-scope).


> Also i use the template 'exec in current scope' method to allow
>
> resultfilters = "isfile and nolink and fnmatch('data*')"
>
> to be compiled at run time into a function and then be
> applied to - say - 15000 files. This technique is similar
> to what the regular expressions module does with re.compile.

*NO*!  re.compile NEVER mucks with your local namespace.
Any "similarity" may be with functions such as the built-in compile
(see below), never with exec-in-local-scope, which *does*
have uncontrollable alterations of namespace as its only reason to exist.

> This may be a better example than the original one?!
> At least i don't see how you can apply your techniques
> to it without using slow 'evals' or a slow interpretation

You seem to be unaware that built-in function 'compile' is able to make 
code objects, that eval is not slow at all when passed a code object, 
and/or that the functions of module new also let you build functions.
These are not "my" technique, btw, but Python's.


> None.strip() in
>
> ...__dict__.get(name).__doc__.strip()
>
> would fail at run time if a filters's doc has not been defined.
> Same with your code.

Yes, but later you access filter.name unconditionally, and filter
is what is at the point of the ellipsis.  The 'get', with or without a
default 2nd argument, has nothing to do with whether a specific
named filter has or lacks a docstring, but only with whether a
filter of that name exists in the filter module.  And if no filter of
that name exists, the method is going to fail at runtime anyway.
So I still don't see why you're using this peculiar construct.


> I appreciate your explanations and code very much.
>
> The only better thing i can imagine is if you tried a tiny bit
> harder to see some *valid* points rather than only pin down invalid
> points. Otherwise you give me the feeling that i am completly on the
> wrong path. (or is this intended?!)

I do believe you are "completely on the wrong path" in wanting to
use exec in local scope.  I am certainly open to being convinced,
should some example be brought that I don't find a way to recode
which seems preferable to me; it just appears to me to be highly
unlikely for such an example to emerge.

> I just try to communicate a use for "exec" where it is preferrable
> to not use explicit dictionaries.

I think "exec in local scope" is invariably overkill, and the more highly
disciplined tools that Python provides in abundance afford better,
faster and more maintainable ways to perform even tasks so dynamic
that their very wisdom is dubious.

'GOTO' (in languages that have it) may no doubt let you construct
your own data structures, finely tuned to your tastes and what you
see as the task at hand.  However, the more highly disciplined
control-flow tools that Python provides in abundance afford better,
faster, and more maintainable ways to control your program's flow.
I think the analogy is pretty strict, and long for a Python whose
exec was a function taking mandatory dict arguments.  Oh well.


I think one good language, quite close to Python in most respects, that may
support better your specific quest of runtime compilation of source in
order to alter namespaces, is Ruby.  In Python, you're fighting against
the grain of the language when you do this.  In Ruby, you are closer
to the mainstream of the language -- you may also alter in such ways
built-in types akin to Python's object, string, dict, etc, while Python does
not let you do that.  I think Python's approach is wiser, and Ruby's leads
to a programming style damaging the ability to understand and maintain
large programs (but I haven't tackled 'big' projects in Ruby, so this is
only a working hypothesis for me so far).  But I suspect you might disagree,
and perhaps find yourself more at home in Ruby.  Do give it a look.


Alex




From peter at engcorp.com  Tue Apr  2 20:53:41 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 20:53:41 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com> <3CAA562E.C5D592A2@engcorp.com> <200204030123.BAA15136@crown.off.ekorp.com>
Message-ID: <3CAA60A5.F88697D5@engcorp.com>

Richard Jones wrote:
> 
> On Wed, 3 Apr 2002 11:09, Peter Hansen wrote:
> > By the time we'd
> > scrolled through the ReStructured Text documentation I had to say
> > to him, "Hmm... that's so bloody complicated it might be easier
> > writing XML tags directly.  Sort of defeats the purpose of structured
> > text, doesn't it?"
> >
> > When structured text becomes so complicated to type that one has to
> > actually work at it, and it needs a large manual to use, it has
> > become too unwieldy to be useful.
> 
> Are you sure you aren't confusing comprehensive (and, unlike most STX-alike
> documentation formats out there, well-designed) with complicated?

Yes, you're right, I was.  My apologies.  On rereading with a more
thoughtful eye, and ignoring the bulk of the links on the site and
focusing on the quickref only, I conclude after brief practice 
that I would have little trouble using it effectively.  

I would suggest an even shorter version of the quickref, focused on 
only those items needed in 80% of embedded documentation.  It does
_look_ complicated at first.

> > My suggestion would be go with a simpler Structured Text for the 80%
> > of documentation needs it, and allow an escape to DocBook or
> > something for more complicated docs...
> 
> So we'd have to learn DocBook, or some other [SG|X|HT]ML layout? I'd
> guarantee that the STX version is going to be much easier to edit and
> maintain (without purchasing Frame, that is ;)

Actually, that would be entirely unnecessary.  You would only 
need to do that if you wanted complex documentation.  Since
we don't want Structured Text to be so complex that it 
could really handle that kind of documentation, an escape 
mechanism would still be needed for something that ugly.
But why would anyone want to embed such complex documentation?
(Yes, I know all about Web and Weave.)

-Peter



From db3l at fitlinxx.com  Wed Apr  3 16:14:48 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 03 Apr 2002 16:14:48 -0500
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <xIGq8.25783$VQ2.8306517@twister.socal.rr.com> <mailman.1017854591.19430.python-list@python.org>
Message-ID: <ug02cwkl3.fsf@ctwd0143.fitlinxx.com>

Rich Harkins <rich at worldsinfinite.com> writes:

> Because of the is comparison even if "MYBLANKVALUE" is in d then this code 
> will still work.  Just don't intern whatever "MYBLANKVALUE" is (I frequently 
> use undefined="UNDEFINED").

Of course, that's getting interned automatically anyway:

    >>> undefined="UNDEFINED"
    >>> foo="UNDEFINED"
    >>> id(undefined),id(foo)
    (8357472, 8357472)
    >>> foo is undefined
    1

You could play games (adding a space, or using a non-identifier
character) but then you're coding to a specific implementation and
optimization which could change at any time.

If you need to go this approach (wouldn't be my first choice), you
could at least use a mutable object (say []) which would be safest and
known to be unique.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aleax at aleax.it  Mon Apr 29 04:05:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 08:05:32 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <eppstein-DF7988.23533128042002@news.service.uci.edu>
Message-ID: <g97z8.75142$vF6.2246544@news2.tin.it>

David Eppstein wrote:

> In article <aaim48$7fd$1 at bob.news.rcn.net>,
>  "Raymond Hettinger" <python at rcn.com> wrote:
> 
>> The functionals (map, filter, and reduce) had greater importance
>> prior to the introduction of list comprehensions which are now the
>> preferred (and more readable) approach.
> 
> Map and filter, I can see substituting with list comprehensions.
> But can you tell me how to use a list comprehension in place of reduce?

Only kludgily, today (skipping the extra test needed to handle
a missing 'starter'):

def myreduce(function, sequence, starter):
    return [starter for item in sequence
        for starter in (function(starter,item),)][-1]

>>> from operator import add
>>> myreduce(add, range(10), 0)
45
>>> 

You can replace any use of built-in reduce with a list comprehension
based on these ideas -- if starter is missing you need a further
'for starter in sequence[:1]' as the first for clause of the LC
and of course 'for item in sequence[1:]' as the second for clause.

But I do think this IS kludgy, much like saying that Python DOES
have assignment-within-expression because where in C you might
code

while((next=makenext())) process(next);

in Python you might code

while [next for next in (makenext(),)][0]: process(next)


"[x for x in (y,)][0]" DOES have just the same semantics as a
hypothetical "x=y acceptable as part of an expression".  But it
does appear to me to be a kludgy exploit of something of an
"implementation accident" of list comprehensions.


Alex




From seb_haase at yahoo.com  Thu Apr  4 14:11:36 2002
From: seb_haase at yahoo.com (Sebastian Haase)
Date: 4 Apr 2002 11:11:36 -0800
Subject: Solution for   python22_d.lib
Message-ID: <f445a2fe.0204041111.66e3cced@posting.google.com>

Hi everybody!

Here is the issue:
Compiling a C/C++ program (that includes "python.h") with 
VC++  in "debug-mode"
wants  to link python22_d.lib.  But this isn't part of the binary
python distribution.

Solution 1)  -->> BAD!! <<-
    downloading the python source and compile python (itself) in
debug-mode.
   This was suggested to me in and others here in this group!
   But the "stuff" you get/end-up-with  doesn't work as well. 
   a) e.g numPy doesn't like it (see my [unanswered ;-( ] post ]
      about "multiarray" last week - I tried to recompile it from
source - but
      "python setup.py install" did't find the python libs
   b) wxPython/"and lot's of other nice modules" don't install
properly because
      they don't find the "standard file/dir structure" --> e.g. libs
are
      somewere in "C:\python-2.2\PCbuild" !!!

=====
The REAL 
Solution !  --> "we" are not really interested in debugging the python
lib !
========    ==> we only want to debug _our_ own C/C++ code!!!

       just go into  c:/Python22/include/pyconfig.h:
                     ------------------------------
     and change these lines:  (starting line 331)
@code@

#if !defined(USE_DL_EXPORT) && defined(_MSC_VER)
/* So nobody using MSVC needs to specify the .lib in their Makefile
any
   more (other compilers will still need to do so, but that's taken
care
   of by the Distutils, so it's not a problem). */
#ifdef _DEBUG
// seb  WHO has python22_d.lib ?? WE DON'T WANT TO DEBUG _PYTHON_
// seb #pragma comment(lib,"python22_d.lib")
#pragma comment(lib,"python22.lib")
#else
#pragma comment(lib,"python22.lib")
#endif

@/code@

the " // seb" - part refers to my changes!
----------------------------------------------------------------
--> so both debug and release can link against  python22.lib.
----------------------------------------------------------------

This has caused me a lot of trouble  -- and now I feel to be back on
track - since I can just use ALL the standard python-installations !!!

Regards,
Sebastian

PS: Maybe the above code should be taken out entirely !!! It causes
trouble ;-))


From rahvee at yaahhoo.com  Wed Apr 10 18:44:35 2002
From: rahvee at yaahhoo.com (Ned Harvey)
Date: Wed, 10 Apr 2002 18:44:35 -0400
Subject: Indenting lines 400..457 help
References: <a929tk$ik2$1@pcls4.std.com> <Fc2t8.90644$w7.7532194@bin6.nnrp.aus1.giganews.com>
Message-ID: <a92f90$i57$1@pcls4.std.com>

That's not bad.  Not bad at all.
Now of course, the reason I want to {in|de}dent a large section of text, is
because I'm copying a large section of python code.  So my *real* goal is to
insert (or delete) a given number of spaces at the beginning of a whole
bunch of lines, to make it match the indentation of the surrounding code.

By using the >> or << method below, vi uses the tab-characters, at least in
my account.  Is there any way you're aware of, that I can specify the
characters to use?

My vi-manual doesn't seem to say anything about it.

Thanks again.

"Brett g Porter" <BgPorter at NOartlogicSPAM.com> wrote in message
news:Fc2t8.90644$w7.7532194 at bin6.nnrp.aus1.giganews.com...
>
> "Ned Harvey" <rahvee at yaahhoo.com> wrote in message
> news:a929tk$ik2$1 at pcls4.std.com...
> > For the life of me, I can't find any easy way to indent (or unindent)
> lines
> > numbered 400..457 or whatever.  Does anyone have any good suggestions?
> >
> > I'm usually a vi user, but I'll happily learn a command of some other
> > format.  I've tried
> >     :400,457s/(.*)/   \1/g
> > and a ton of variations of the same idea, but it seems that the
> > search/replace doesn't do what I expect it to.
> >
> > Thanks for your help.
> In vi, place the cursor on the first line that you want {in|de}dented.
Then
> type the number of lines to affect and >> to indent, or << to dedent.
>
>
> --
> --
> //  Today's Oblique Strategy (? Brian Eno/Peter Schmidt):
> //  Don't be afraid of things because they're easy to do
> //
> //  Brett g Porter * BgPorter at acm.org
> //  http://mywebpages.comcast.net/bgporter/
>
>
>




From sholden at holdenweb.com  Tue Apr 16 15:44:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 15:44:11 -0400
Subject: pre-PEP for optional 'pass'
References: <mailman.1018972676.30534.python-list@python.org>
Message-ID: <bc%u8.421$hl2.358@atlpnn01.usenetserver.com>

<brueckd at tbye.com> wrote in message
news:mailman.1018972676.30534.python-list at python.org...
> On Tue, 16 Apr 2002, Duncan Booth wrote:
>
> > You should always head each method and class with a docstring.
>
> No! You should head each method and class with a docstring *only if the
> docstrings would add value*. Otherwise you're doing more work now AND
> later (during refactoring) without producing any real benefit, cluttering
> the code in the process.
>
> > class MyClass:
> >    '''A simple class'''
> >    def __init__(self, x, ,y, z): '''MyClass constructor'''
> >    def getTwiddle(self): '''Get a twiddle'''
> >    def setTwiddle(self, twiddle): '''Set a twiddle'''
>
> The above code snipped contains exactly 4 examples of docstrings that
> should not exist - all of them either contain no useful information or no
> information that isn't already painfully obvious.
>
> A good rule of thumb is that if the docstring (or comment) could have been
> auto-generated by your editor, you shouldn't include it.
>
Unless, of course, you are a Microsoft employee documenting your latest
creation, in which case it is mandatory to include a section for every menu
option such as:

Edit | Frob | Unfrobnicate

The Edit | Frob | Unfrobnicate command unfrobnicates a frob that has
previously been frobnicated.

About the only thing you tend *not* to find in Microsoft documentation is a
sensible discussion of frobs and what they might do for you.

WHile I'm at it, I REALLY HATE those messages that tell me to contact my
system adminstrator when a problem has arisen. Makes me feel like shouting
"But I *am* the bloody system administrator" at the dialog box or error
message.

i-feel-better-now-ly y'rs  - steve
\





From kseehof at neuralintegrator.com  Tue Apr  9 19:19:16 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 09 Apr 2002 16:19:16 -0700
Subject: True/False story... (PEP 285)
In-Reply-To: <3CB2E617.8050600@geneva-link.ch>
Message-ID: <EFEFIHAINIINNELLHFAIAEGOCAAA.kseehof@neuralintegrator.com>

> Alex Martelli wrote:
> 
> > there IS a natural mapping of booleans onto a subset
> > of integers, just as there is a natural mapping of
> > integers onto a subset of reals (or rationals, etc).
> 
> Could be argued that there is no single natural mapping, for an 
> alternate to the current one we could have
> 
> True     == 1
> False    == -1
> not x    == -x
> x and y  == min(x,y)
> x or y   == max(x,y)
> x xor y  == -x*y
> 
> A nice feature of that one (to my aesthetics), is that
> 
> 1j == sqrt(False)
> 
> It also meshes well with the cyclotomic fields.
> 
> Boris
> --
> "Hope achieves the square root of the impossible"
 
Cool.  And it maps to our linguistic understanding of positive
and negative.  Hey, let's write a PEP :-)

(I hope Guido doesn't ruin it for us by pointing out that there
is only one mapping of bool to a subset of int that is backward
compatible.)

Also, I think we should change the mapping of integers to floats.
Using an logarithmic scale would have many advantages.  Now some
of the integers would no longer be permitted (i<1).  This is
consistent with the observation that 0 is a kind of None, and
therefore shouldn't exist.  People focus way too much on the
negative.  This PEP will help people be more positive.

Combined with your PEP, this leaves us with only Truth!

- Ken
"Most of the complexity of computer science would be removed by
the abolition of zero."





From sumner-nntp5 at forceovermass.com  Wed Apr  3 10:41:09 2002
From: sumner-nntp5 at forceovermass.com (G. Sumner Hayes)
Date: Wed, 03 Apr 2002 15:41:09 GMT
Subject: A faster Python?
References: <a8dcqr$p29$1@laurel.tc.umn.edu> <7xn0wl3bg9.fsf@ruckus.brouhaha.com>
Message-ID: <poFq8.101209$VJ1.8293815@bin3.nnrp.aus1.giganews.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote:
> Psyco is a research project.  It would be great if it were developed
> further, but in its current form it's not really useful.

It is being developed further, last I heard it passed all the Python
2.2 regression tests and they were aiming for a public alpha some time
soon.  Still some Zope issues.  

I'm not sure if any of the above is still current, but it was a couple
weeks ago.  http://psyco.sourceforge.net has mailing list archives.

  Sumner

-- 
rage, rage against the dying of the light

Usenet email address changes periodically.  sumner-nntp at forceovermass
period com is more permanent.


From spahievi at vega.bg  Sat Apr 13 06:02:08 2002
From: spahievi at vega.bg (Niki Spahiev)
Date: Sat, 13 Apr 2002 13:02:08 +0300
Subject: Notes on 2.2's changes
In-Reply-To: <slrnabbibb.iac.akuchlin@crystal.mems-exchange.org>
References: <slrnabba3h.h8h.akuchlin@crystal.mems-exchange.org>
 <Xns91ED5ED8EF7A0lumberjack@216.148.53.96>
 <slrnabbibb.iac.akuchlin@crystal.mems-exchange.org>
Message-ID: <243208523.20020413130208@vega.bg>

2002-04-11, 20:35:41, A.M. Kuchling wrote:

AMK> My holy grail is to
AMK> someday use a version of the ZODB that doesn't use ExtensionClass,
AMK> which seems possible with 2.2.  (ExtensionClass is far, *far* more
AMK> annoying than anything that 2.1 or 2.2 has changed; it means we can't
AMK> run PyChecker, and can't use __r*__ overloading or and had to work
AMK> around isinstance() being broken for ExtensionClass in 2.1.)

FYI there is ZODB without ExtensionClass in Zope3 CVS

Niki Spahiev


______________________________________
VEGA(TM) Internet Service Provider
      <http://www.vega.bg>
Scanned and protected by Inflex
Tova pismo e provereno i zashtiteno
ot programite "Inflex/Antivir".






From tim.one at comcast.net  Thu Apr 11 11:49:03 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 11 Apr 2002 11:49:03 -0400
Subject: Tuple comprehension
In-Reply-To: <Pine.LNX.4.33.0204110859270.25928-100000@ns0.tbye.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHOEJFCPAA.tim.one@comcast.net>

[brueckd at tbye.com]
> ...
> I'd *love* to have dictionary comprehensions!

Try this on for size:

>>> dict([(i, i**2) for i in range(10)])
{0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25, 6: 36, 7: 49, 8: 64, 9: 81}
>>>

That is, 2.2 dict() accepts an iterable object producing iterable objects
each producing two values.  A list of 2-tuples is one instance of that, and
one particularly easy to create with a listcomp.  More possibilities open if
you ponder other kinds of iterable objects; e.g.,

>>> def squarepairs(n):
...     for i in range(n):
...         yield i, i**2
...
>>> dict(squarepairs(10))
{0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25, 6: 36, 7: 49, 8: 64, 9: 81}
>>>

is one way to get the same end result without building a list in one gulp
first.





From tim_one at email.msn.com  Wed Apr 10 02:58:21 2002
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 10 Apr 2002 02:58:21 -0400
Subject: What has PEP 285 done to us?
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A263@natasha.auslabs.avaya.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEMGOMAA.tim_one@email.msn.com>

[Delaney, Timothy]
> Actually, as a lumberjack, it really doesn't matter what sex he/she is ;)
>
> I'm hoping the post was just a joke, although it didn't carry any of the
> characteristic humour markers ...

Smelled more like a troll.  It could have credited Paul Rubin for the "water
torture" analogy, s/drop/drip/, but raw provocation seems the only goal.

just-another-day-on-usenet-ly y'rs  - tim





From peter at norvig.com  Wed Apr  3 04:43:01 2002
From: peter at norvig.com (Peter Norvig)
Date: 3 Apr 2002 01:43:01 -0800
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net> <3CAA7FD0.56C285E0@engcorp.com>
Message-ID: <da69ff6a.0204030143.679e2563@posting.google.com>

You can also use optional arguments to functions and methods.  These
share with C static variables the property of being assigned once:

def f(x, s=initialization_code()):
    # s behaves like a static variable
    ...

Peter Hansen <peter at engcorp.com> wrote in message news:<3CAA7FD0.56C285E0 at engcorp.com>...
> Jim Jacobs wrote:
> > 
> > Is it possible to simulate "C" style static variables in functions?  In
> > methods?
> 
> Maybe ...  what do you mean by simulate, and what characteristics
> of the C statics would you like to have?
> 
> In C, static inside a function causes the variable to be assigned
> in the heap rather than on the stack (to put it one way) and
> that gives the variable persistence across invocations of
> the function (but scope local to the function).
> 
> In Python, you would use either an instance variable, inside
> a method of an object, or the answer would be "no, but can
> you get by with using 'global'?".
> 
> -Peter


From starx at pacbell.net  Wed Apr 10 05:11:38 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 10 Apr 2002 09:11:38 GMT
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk>
Message-ID: <Xns91EC16845A7E9RASXnewsDFE1@130.133.1.4>

Anders J. Munch || Wed 10 Apr 2002 01:43:28a:

> Pythons truth values are not "Something or Nothing".  The phrase seems
> to have caught on following Laura's post, but I believe it is
> misleading.
> 
> [] and {} are objects, they are objects with identity, and they are
> mutable objects.  They are empty, but they are by no means "nothing".
> 
> Example:
> 
> def show(foo, history=None):
>     """prints foo, optionally storing what was printed in a history
>     list.""" print foo
>     if history:
>         history.append(foo)
> 
> This works as intended:
> 
>     # Omit default argument, don't remember history
>     show("hello world")
> 
> This works as intended:
> 
>     history = ["in the beginning was nothing"]
>     show("hello world", history) # adds "hello world" to history list
> 
> But this doesn't:
> 
>    history = []
>    show("hello world", history) # strange..history doesn't change
>    here? 
> 
> It doesn't work, because show() tests for emptyness with "if
> show_history:".  What show() should have done is test for nothingness
> using "if show_history is None:".

Hm, interesting. 

Right at this molment (and i'm a little tired), i'm thinking that None 
is only nothing by itself. Think of this, if we say:

x = None

x still exists. It isn't nothing. It's more like it's not here right 
now.*  We know what it is, we have a name for it, but we don't have any 
of it. We can test to see if it is there; if it dosen't exist, we can't 
test for it at all.

*This is sort of like how adventure games tend to handle nouns. If you 
say 'look at history', and history is in scope (in los/current area, not 
hidden, etc.), it will print something along the lines of 'history is a 
list. it contains a string.' (assuming we are in a text adventure that 
models the python interperter space as physical space... i'll have to 
try something like this sometime =). If there is a history somewhere, 
that is, that noun exists in the parser's dictonary_of_names, you'll get 
a message along the lines of "I don't see any history here.". However, 
if there is no such thing in the game world, and the parser dosen't know 
that word, it says "I don't know what a history is." (i.e. don't ever 
bother using the word again, it dosen't matter =).

ramble.Ramble.stop()

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From siegfried.gonzi at kfunigraz.ac.at  Mon Apr 22 13:09:30 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 22 Apr 2002 19:09:30 +0200
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org> <aa1fe9$mn7$1@peabody.colorado.edu>
Message-ID: <3CC443CA.A7CE9E6A@kfunigraz.ac.at>

Fernando P?rez wrote:

> Sorry but no. The fact that certain physical problems may have served as
> inspiration for some mathematical developments can be considered to be purely
> accidental. The foundational structure of mathematics is entirely one of
> logical self consistency, and massive amounts of work was done particularly
> in the 20th century to build that framework. While it is true that up to the
> 19th century most (if not all) significant advances in math were prompted by
> challenges from the physical sciences, the work of the formalists in the 20th
> century showed that mathematics could be built entirely on an abstract,
> logical foundation.

This is true, fo example even the tensor analysis ("Ricci") has been
formalized 1850 long before Hilbert (though, who did let go Einstein
first) made the conncetion to Einstein's theory of relativity.

Or, Maxwell's equations first came without any tensor analysis
formulation. This has been added/reformulated later.

This remembers me on a professor who asked me (for a certificate): "What
do you know about Maxwell's equations". I started to write  it
mathematically down, but he then stopped me and said: "We are doing
physics and not math now". I just had it to explain verbally.

S. Gonzi


From roy at panix.com  Wed Apr  3 18:32:37 2002
From: roy at panix.com (Roy Smith)
Date: Wed, 03 Apr 2002 18:32:37 -0500
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <lkvgb88yfk.fsf@elios.maths.bris.ac.uk>
Message-ID: <roy-7309E8.18323703042002@news1.panix.com>

Michael Hudson <mwh at python.net> wrote:
> try:
>     value = d[key]
> except KeyError:
>     foo = otherStuff(key)
> else:
>     foo = stuff(value)

Duh.  Been writing 4 Python for 4 years now, and while I'm sure I once knew 
that try blocks could have else clauses, I had just plain forgotten I ever 
knew it.  It's exactly what I was looking for.

Well, almost.  It's slightly annoying that the main flow of control is 
broken into two pieces; one in the try clause, the other in the else 
clause.  But it's a lot better than the alternatives I had come up with.


From r.demaria at tiscalinet.it  Sun Apr 14 04:21:13 2002
From: r.demaria at tiscalinet.it (Riccardo de Maria)
Date: 14 Apr 2002 01:21:13 -0700
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu> <mailman.1018589458.11530.python-list@python.org>
Message-ID: <90cc01ee.0204140021.5a310ceb@posting.google.com>

> My preference for small graphs of objects is to dump the whole graph
> and retrieve the whole graph and do my operations in-memory (for
> definitions of small < 1MB). It's simpler, has fewer limitations, and
> fewer dependencies. 

I want to work with graph, too. I'm a newbie, could you make an
example of implementing a graph, store and retrieve it?

Thanks
Riccardo


From mwh at python.net  Wed Apr  3 06:21:47 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 3 Apr 2002 11:21:47 GMT
Subject: PyInt_AsLong()
References: <xdu7knpaz87.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <lkit79uiwl.fsf@elios.maths.bris.ac.uk>

Ralf Juengling <juenglin at informatik.uni-freiburg.de> writes:

> Hi,
> 
> PyInt_AsLong() returns '-1? to indicate an error, this meets the convetion 
> for PyXXX()-functions which return integers. But how do you distinguish an
> '-1' indicating an error from a the legal long-value '-1'?

if ((retval == -1) && PyErr_Occurred()) {
    waagh!
}

Cheers,
M.


From gerhard at bigfoot.de  Sat Apr 13 18:49:39 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 13 Apr 2002 22:49:39 GMT
Subject: Python newbie
References: <ab96a890.0204131438.5cce8f66@posting.google.com>
Message-ID: <slrnabhdg4.6cf.gerhard@lilith.my-fqdn.de>

slack wrote in comp.lang.python:
> Hello,
> I'm interested in Python programming language since one around one
> year. I think it is an interesting language because it's very simple
> to use this language to program windows... We don't have to think
> about event handling instead of the purpose of the program itself...
> well... I don't want to explain why I find it interesting... I've
> installed a version of Python on Windows computers but I find it very
> slow compared to Linux version of Python.

What exactly was slow? In fact, the Python interpreter on Windows is
normally slightly faster then the one on Linux, because of better
optimizations of the Microsoft C compiler.

> I've noticed that the first time I used Python on Windows. Well.
> This is not really the problem. I want to use Python on Windows but
> I don't know if I can convert .py files into .exe files? Is it
> possible thanks to some utility?

py2exe and Gordon' McMillan's installer can do that.
 
Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Apr 30 20:06:47 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Wed, 1 May 2002 02:06:47 +0200
Subject: Importing dynamically loaded modules (in packages)?
Message-ID: <aanbio$pk3$1@news1.xs4all.nl>

Hello
I want to import a dynamically loaded module that comes from a package.

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82234
more or less confirms the code that I've created myself (in Pyro).
But this code only works if the module does not come from a package.

i.e. loading the .py file of "TestModule" and then "import TestModule"
works fine, but loading "Package.TestModule.py" and then
"import Package.TestModule" doesn't work.
I've been toying around with creating package/modules structures
with new.module, mimicing the package structure on disk, but
even if I have this:

sys.modules['Package']= module 'Package'
    module 'Package' . TestModule = the loaded module

"import Package.TestModule" still doesn't work: ""no module named TestModule""

How can I correct this?

Thank you!
Irmen de Jong



From marklists at mceahern.com  Wed Apr 17 14:02:56 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 17 Apr 2002 13:02:56 -0500
Subject: Trinary operator?
In-Reply-To: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>

> Not that it's a big deal since there's many ways to do that, but can I use
> something like
>     verboseGender = (gender == 'm') ? 'male' : 'female'
> in Python?

People seem to use and/or for that:

>>> gender = 'm'
>>> verbose = (gender == 'm' and 'male' or 'female')
>>> verbose
'male'
>>> gender = 'f'
>>> verbose = (gender == 'm' and 'male' or 'female')
>>> verbose
'female'

// mark




From tdelaney at avaya.com  Sun Apr 21 19:07:40 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 22 Apr 2002 09:07:40 +1000
Subject: Mailman vs References: (was Re: How to make regexes faster? (
	Python v. OmniMark))
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A27A@natasha.auslabs.avaya.com>

> From: aahz at pythoncraft.com [mailto:aahz at pythoncraft.com]
> 
> In article <mailman.1019363618.25367.python-list at python.org>,
> Tim Peters  <tim.one at comcast.net> wrote:
> >[Erno Kuusela]
> >>
> >> maybe mailman could be hacked to fake a "references" header based
> >> on "in-reply-to".
> >
> >Barry won't do that. 
> 
> You sure?  Last time I talked to him about it a couple of 
> years ago, he
> said he'd at least consider it if I gave him a patch.

Make sure it also goes the other way ... "In-reply-to:" is created from a
"References:" header.

Tim Delaney




From merman at freenet.de  Sat Apr 13 05:13:06 2002
From: merman at freenet.de (merman)
Date: Sat, 13 Apr 2002 11:13:06 +0200
Subject: newbie-question: interfaces
Message-ID: <3cb7f47e$0$360$9b622d9e@news.freenet.de>

Hi,

is there a way to simulate interfaces (like in java) in python?
... or abstract classes? (c++/ java)

o-o

thomas






From tdelaney at avaya.com  Tue Apr  2 02:39:03 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 2 Apr 2002 17:39:03 +1000 
Subject: PEP 285: Adding a bool type
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A240@natasha.auslabs.avaya.com>

> From: Max M [mailto:maxm at mxm.dk]
> 
> Steve Holden wrote:
> > 
> > So, is breaking bad code more excusable than breaking good code ;-)?
> 
> 
> Oh I don't think so. This is a common idiom:
> 
> def f(names=None):
>      if names == None:
>          names = []
>      ...

That should read:

def f (names=None):
    if names is None:
        names = []

:)

Tim Delaney



From achim.domma at syynx.de  Mon Apr 29 10:54:07 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Mon, 29 Apr 2002 16:54:07 +0200
Subject: os.path like library for urls ?
References: <aajkrn$o0a$07$1@news.t-online.com>
Message-ID: <aajmc4$lvg$05$1@news.t-online.com>

sorry, found it by myself : urlparse !




From noah at noah.org  Tue Apr 23 16:59:57 2002
From: noah at noah.org (Noah)
Date: Tue, 23 Apr 2002 13:59:57 -0700
Subject: reading unix files from NT using SSH2
In-Reply-To: <F666B26E8EF5B84C88E3DF34E5A1E49CF081E1@venus.exolink.com>
Message-ID: <PAENINBOOODFDEPOKPKCOEFACMAA.noah@noah.org>

Take a look at PuTTY. This is a free SSH client for Windows.
	http://www.chiark.greenend.org.uk/~sgtatham/putty/
Ignore the fact that it's Beta. It's been out for a long time and
it is stable. PuTTY has a few different tools.
You might want to check out Plink:
	http://the.earth.li/~sgtatham/putty/0.52/htmldoc/Chapter7.html#7

Another possibility is to try Cygwin. It comes with SSH and SCP/
	http://www.cygwin.com/

The other trick you will need to master is using Public Key authentication.
See the man pages for ssh-keygen. You need to use public keys instead of 
passwords in order to use SSH in a script. You can't easily automate giving
SSH or SCP a password from inside a script. Public Keys takes care of this
while preserving security. It only takes a few steps to setup, but 
it sometimes confuses people the first time. Email me if you need help.
Once you get the keys setup then it just works.

Then from Python you could do something like this.
This will copy a file onto the desktop (for all users to see).

	os.system(r'C:\cygwin\bin\scp.exe remotehost:/tmp/mydata 
		"//c/Documents and Settings/All Users/Desktop/mydata.txt"')

Note the raw string
	r'C:\cygwin ... '
This is because we need to give os.system a Windows path name 
that includes backslashes. Without the r'raw string' Python would get confused.
Also note that for scp.exe you need to use Cygwin's style for 
referring to Windows paths. It is also "quoted" because of the 
spaces in the path name.
	"//c/Documents and Settings/All Users/Desktop/mydata.txt"

Yours,
Noah


-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of James Strater
Sent: Tuesday, April 23, 2002 7:32 AM
To: python-list at python.org
Subject: reading unix files from NT using SSH2


I have a python script running on an NT box.  I need to connect to a  unix
box using SSH2 and read some files sitting on that box.  Can anyone point me
in the right direction on how to go about doing this?   The script needs to
stay on the NT box so that it can do some things with MS Excel using COM.  

Thanks,

James







From sholden at holdenweb.com  Tue Apr 30 13:08:27 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 13:08:27 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net> <mailman.1020176703.26327.python-list@python.org> <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com> <mailman.1020182455.6633.python-list@python.org>
Message-ID: <LeAz8.606$Au5.499@atlpnn01.usenetserver.com>

"Roman Neuhauser" <neuhauser at mail.cz> wrote ...
> > From: "Steve Holden" <sholden at holdenweb.com>
[...]
> >
> > I  think it actually relates a little more to giving advice on subjects
you
> > confess not to know much about.
>
>     Not advice. I was expressing my opinion as a python user.
>
Well you are certainly entitled to express your opinion.

> > If you haven't come across the effbot before, you should know that he
tends
> > to be rude in this way. I think it happened when the parity unit was
fitted
> > to his main memory. Fortunately it does not stop him from being a
valuable
> > source of information on topics Pythonic. Every now and then he jumps
out
> > and bites a poster on the ankles due to simple crotchetiness, and then
he
> > will subside into inactivity once more, contemplating the innards of his
> > software in silent satisfaction.
>
>     Given the high tongue-in-cheek ration of this list, I'm now really
>     confused as to whether Fredrik Lundh is a bot, or if you're "rocking
>     me", as the Czech saying has it.
>

Well if he's really a bot then the joke's on me, because I exchange email
with him occasionally. I seem to remember you can see a photograph
purporting to be him somewhere on the web

        http://www.python.org/~guido/confpix/flundh-2.jpg

has-anyone-seen-the-timbot-around-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From DeepBleu at DeepBleu.org  Mon Apr 22 10:49:54 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Mon, 22 Apr 2002 09:49:54 -0500
Subject: dco2.so problem
References: <e2e9e807.0204201749.62f3b8c6@posting.google.com>
Message-ID: <uc88ovs4pjum17@corp.supernews.com>

Hugh
Before going on these curves, and before moving all these files into the CGI
directory - what error were you getting while trying to connect using
DCOracle2?
Say you have DCO2 installed into /usr/python/DCO2; did you do the following:
>>import sys
>>sys.path.insert(1,"/usr/python/DCO2")
>>import DCOracle2
where dco2.so is in /usr/python/DCO2/DCOracle2

But you see, you've got to give me the original error message before you
moved all these files in the CGI directory.
Also, do you have Oracle configured correctly and do you have libclntsh.so
in the path?
DeepBleu

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in message
news:e2e9e807.0204201749.62f3b8c6 at posting.google.com...
> I can import DCOracle2 now that I've put dco2.so, DCOracle2.py and
> __init__.py in my cgi-bin directory (probably not the right thing to
> do, but it imports OK).
> I now have another problem. When trying to do dbh =
> DCOracle2.connect(connectstring) I get the following error in the
> apache error log:
>
> Traceback (most recent call last):
>   File
"/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/addItemType.py",
> l
> ine 12, in ?
>     dbh = DCOracle2.connect('scm1hewf/*******@cs2000')
>   File
"/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/DCOracle2.py",
> lin
> e 93, in connect
>     db = dco2.connect(u, p, dbname)
> dco2.DatabaseError: (12154, 'Error while trying to retrieve text for
> error ORA-1
> 2154')
>
> Does anyone know what this means and how to fix it.
>
> Hugh






From martin at v.loewis.de  Tue Apr  9 03:26:48 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Apr 2002 09:26:48 +0200
Subject: SSL error: 2.1.1 works but 2.2 doesn't
References: <mailman.1018332682.8578.python-list@python.org>
Message-ID: <m3k7rhl4cn.fsf@mira.informatik.hu-berlin.de>

Anthony Baxter <anthony at interlink.com.au> writes:

> >>> Walter Miller wrote
> > I tried 2.2.1rc2 with the same results.  I haven't
> > tried 2.1.3.
> 
> Log a bug on sourceforge.net - if you don't, it's unlikely to get
> addressed. The chances of it being addressed for 2.2.1 are so close
> to zero as to be ignored, but there will be a 2.2.2, so it'd be nice
> to fix it for that.

This is probably the same problem as http://python.org/sf/494762,
for which http://python.org/sf/500311 is a suggested patch.

It wouldn't hurt if more people tried to understand the problem,
though, and comment on the proposed solution.

Regards,
Martin



From mrroach at uhg.net  Fri Apr 19 17:22:02 2002
From: mrroach at uhg.net (Roach, Mark R.)
Date: 19 Apr 2002 16:22:02 -0500
Subject: PIL, printing on win32
Message-ID: <1019251323.20014.46.camel@tncorpmrr001.uhg>

Hello, all I am trying to print some images using PIL in win32. Through
much trial and error I have come up with the following

    ... create printer dc etc ...
    dc.SetMapMode(win32con.MM_ISOTROPIC)
    dc.SetWindowOrg((0, 0))
    dc.SetWindowExt((10, 10))
    dc.SetViewportOrg((0, 0))
    dc.SetViewportExt((mag, mag))  # set the image magnification
    pilDisplay.expose(dc.GetHandleOutput())  # pilDisplay is an ImageWin.Dib
    dc.EndPage()
    dc.EndDoc()


now, the problem is that I get 10 meg spool files. This is better than I
had been getting when I scaled the PIL image before expose()ing it to
the printerdc, but still seems like too much for a 68k image. Does
anyone know of a way to improve this?

Thanks for your help,

Mark Roach





From sholden at holdenweb.com  Thu Apr 25 10:04:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 25 Apr 2002 10:04:11 -0400
Subject: sql but... without server
References: <7fvfcu0vkc610fma8545v0nhdm7c6hem7q@4ax.com>
Message-ID: <14Ux8.13604$z55.3648@atlpnn01.usenetserver.com>

"Johann" <programisci at NOSPAM.murator.com.pl> wrote ...
> I  would like to use sql database with python but I do not want to
> install any sql server.
>
> I would like to use sql engine as a standalone file (e.g. dll).
>
> It is because I wanted to prepare a standalone application which run
> from CD without any servers or network connection.
>
> Do anone have any experience with similar task?
>

If you are prepared to accept a restricted version of SQL, Gadfly
(www.chordate.com) might be just the thing for you. No DLLs, it's a SQL
database written only in Python.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From amckay at merlinsoftech.com  Tue Apr  2 12:28:29 2002
From: amckay at merlinsoftech.com (Andy McKay)
Date: Tue, 2 Apr 2002 09:28:29 -0800
Subject: Zope query: importing packages from External Methods
In-Reply-To: <934d5a7e.0204012346.262d975b@posting.google.com>
References: <934d5a7e.0204012346.262d975b@posting.google.com>
Message-ID: <200204021726.g32HQEq08780@mail.merlinsoftech.com>

> I have tried ZOPE\lib\python\Shared, PYTHON\site-packages, and even
> ZOPE\Extensions.

- If your module was named "Foo" you can put it in zope\lib\python and 
import Foo. 
- If you want it in zope\lib\python\Shared\DCK it would be from Shared.DCK 
import Foo. But you will need to add a blank __init__.py to the directory

Lastly you could just add it to your python installation, look for the lib 
folder containing all the other python modules
--
  Andy McKay



From jon+usenet at unequivocal.co.uk  Tue Apr 23 22:17:01 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 02:17:01 -0000
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com>
Message-ID: <slrnacc5cs.tg8.jon+usenet@snowy.squish.net>

In article <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v at 4ax.com>, Fernando Rodr?guez wrote:
> No. Step (b) processes the order and does give feedback. Step (a) saves in a
> db some info conatined in a cookie (date of first visit and first referrer). I
> want step (a) to be seamless.

Must admit I don't have a clue what you mean. Why aren't scripts (a)
and (b) in the same script? If they really must be separate scripts,
why doesn't script (a) just output nothing at all to stdout?


From grante at visi.com  Sun Apr 14 13:13:39 2002
From: grante at visi.com (Grant Edwards)
Date: Sun, 14 Apr 2002 17:13:39 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <ubirifj8o6fu48@news.supernews.com> <mailman.1018797388.28608.python-list@python.org>
Message-ID: <slrnabje7j.12c.grante@tuxtop.visi.com>

In article <mailman.1018797388.28608.python-list at python.org>, jepler at unpythonic.net wrote:

> This follows (kinda) from the identity
>     a xor b == (a or b) and not (a and b)
> 
> I'm sure this is in keeping with the spirit of Python.  It doesn't
> short circuit, but it returns one of its arguments when possible,
> and returns the 'not' of one of its arguments in the fourth case.
> There's no short-circuit evaluation, but that's just too bad.

I wouldn't say it's "too bad" that there's no short-circuit
evaluation any more than it's just too bad that the addition
operator doesn't do short-circuit evaluation of it's operands.

The definition of the operation requires that both operands be
evaluated.  Arguing that xor shouldn't be implimented because
it doesn't do short-circuit evaluation is a non-sequitor.  If
that were a valid argument then we'd have to throw out most of
the existing binary operators.

-- 
Grant Edwards                   grante             Yow!  .. Should I get
                                  at               locked in the PRINCIPAL'S
                               visi.com            OFFICE today -- or have
                                                   a VASECTOMY??


From donn at u.washington.edu  Fri Apr 19 13:10:33 2002
From: donn at u.washington.edu (Donn Cave)
Date: 19 Apr 2002 17:10:33 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net>
Message-ID: <a9pj29$pdm$1@nntp6.u.washington.edu>

Quoth claird at starbase.neosoft.com (Cameron Laird):
...
| Next, I'd determine whether my test examples are indeed
| regex-bound (it might well be I/O which constrains your
| performance).  After that ... well, part of the charm of
| regex-s for some people is that they're so flexible that
| different techniques are superior in different circumstances.

Indeed, it could be partly I/O.

I recently went to a meeting and heard someone mention that he had
written a program in Python, his first, but was thinking of rewriting
it in Perl because he had determined that Perl was 10 times faster
at I/O and regular expressions.  At a site that employs hundreds of
at least occasional programmers, this is maybe the fourth I've seen
show this much interest in Python, so I was kind of chagrined to hear
this announcement and went back to check it out.  I was even more
chagrined to find that it was not an unreasonable claim.

Part of the problem is that when you write something like a "grep"
in Python and in Perl, the Perl program will naturally be written
like while ($line = <STDIN>) {...}, and the Python program will
naturally be written like while 1: line = sys.stdin.readline() ...
That pits a lot of function calls against what must be an inline
operation.  I think I decided that "I/O", in this practical sense
of getting a line of data, might have been about half the problem.

The xreadlines function available in later versions of Python did
reduce the disparity a little.  This optimization might help a lot
in the present case, if there's a lot of line-by-line I/O and if
Python is 2.1 or later.

	Donn Cave, donn at u.washington.edu


From sholden at holdenweb.com  Mon Apr 15 08:27:23 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 08:27:23 -0400
Subject: Question about httplib and HTTP response 100 Continue
References: <mailman.1018838549.5496.python-list@python.org>
Message-ID: <PIzu8.10159$Ji7.62@atlpnn01.usenetserver.com>

"Marco V. Caama?o Castro" <mvcc_cr at yahoo.com> wrote ...
"""
I've been trying to figure out why this code[1] does not work[2] on
connecting with a IIS server, but it works with Apache on Linux. I found
that using curl to post the request works. Digging into the response I
got using curl I found that the server is first sending a HTTP 100
Continue response and then the result of the post:

-----------------------
HTTP/1.1 100 Continue
Server: Microsoft-IIS/4.0
Date: Mon, 15 Apr 2002 00:49:27 GMT

HTTP/1.1 200 OK
Server: Microsoft-IIS/4.0
Date: Mon, 15 Apr 2002 00:49:27 GMT
Connection: close
Content-Type: x-data/xact-error
-----------------------

but apache just outputs the results right away:

-----------------------
HTTP/1.1 200 OK
Date: Mon, 15 Apr 2002 00:38:17 GMT
Server: Apache/1.3.22 (Unix) Debian/GNU mod_ssl/2.8.5 OpenSSL/0.9.6a
PHP/4.1.0RC2
X-Powered-By: PHP/4.1.0RC2
Transfer-Encoding: chunked
Content-Type: text/html; charset=iso-8859-1
-----------------------

So, does the httplib work when it encounters a 100-Continue response?

I would really apreciate a little help here as I am stuck, I can't
figure out why using the httplib library doesn't work with this server.
"""

Sorry abt the formatting here: OutOfLuck Express just seems to not quote
certain messages, and I don't have time to fight with it. There's a bug
registered in the SourceForge database that speaks to this problem:


http://sourceforge.net/tracker/?group_id=5470&atid=105470&func=detail&aid=22
7361

This also says that a fix has been generated, though maybe not yet
integrated into the distributions.

regards
 Steve






From cgale1 at cox.net  Sat Apr 20 09:45:43 2002
From: cgale1 at cox.net (Van Gale)
Date: Sat, 20 Apr 2002 13:45:43 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <c76ff6fc.0204191806.406ec53b@posting.google.com> <1019278636.737636@yasure>
Message-ID: <biew8.45681$%A.3954555@news1.west.cox.net>

"Donn Cave" wrote:
> By the way, I'll second Johannes Stiehler's recommendation of
> MxTextTools.  Definitely appropriate for SGML parsing, and much
> better than regexps for extensive parsing in my opinion - not
> just in terms of speed, but I suspect a more powerful way to
> describe text patterns than regexps.

I agree that mxTextTools are excellent but sometimes, like now, I can't keep
my yapper shut when people talk about parsing SGML with regexps :)  If SGML
could be parsed that way then XML never would have been created.  You cannot
parse SGML without completely parsing and supporting all the features in the
DTD that accompany an SGML document (a DTD is mandatory).  For example, in
SGML tags don't have to be enclosed in '<' and '>'.  You can redefine those
in the DTD.  If you are sending or receiving SGML documents from systems
outside your control you better use a big bucks tool like OmniMark,
otherwise you'll end up with egg on your face when those documents don't
validate.  If all the systems are under your control and running your own
programs, then by all means save some money and use regular expressions
tailored to your documents, just don't call them valid SGML ... call them
XML :)

Van





From theller at python.net  Sat Apr 20 07:20:44 2002
From: theller at python.net (Thomas Heller)
Date: Sat, 20 Apr 2002 13:20:44 +0200
Subject: The Python Manual Index
References: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de> <uc0d5d68i0vhf1@corp.supernews.com> <Xns91F5997F18E9ARASXnewsDFE1@130.133.1.4> <nb2w8.4665$T%5.1737@atlpnn01.usenetserver.com>
Message-ID: <a9rj1l$5gq1n$1@ID-59885.news.dfncis.de>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:nb2w8.4665$T%5.1737 at atlpnn01.usenetserver.com...
> "Philip Swartzleonard" <starx at pacbell.net> wrote in message
> news:Xns91F5997F18E9ARASXnewsDFE1 at 130.133.1.4...
> >
> [Thomas Heller's new search interface]
> > >>
> > DeepBleu || Fri 19 Apr 2002 08:16:23a:
> >
> > > WOW!
> > > Great Job!
> >
> > Indeed. Something like this needs to be added to the python.org
> > interface... i've been wondering for a long time why they don't have a
> > document search function =).
> >
> Well, all they really needed to do was link to
>
>     http://web.pydoc.org/
>
> though there's no denying that Thomas' code is effective and useful.
>
One difference at least is that my index search not only searches the module
index, but also the C-api index.
Try http://starship.python.net/crew/theller/pyhelp.cgi?keyword=Py_INCREF
for example.

Thomas




From mwh at python.net  Tue Apr 16 05:44:19 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 16 Apr 2002 09:44:19 GMT
Subject: installation probs on SGI: _socket package removed.
References: <6a3d3529.0204151211.7f38f354@posting.google.com>
Message-ID: <lkofgkrn2b.fsf@pc150.maths.bris.ac.uk>

ophirb at hotmail.com (ophir) writes:

> When building python on SGI IRIX646 I've notice that although the
> _socket library compiled and linked ok, the 'make' process failed to
> import _socket and removed it.
> Following the advice I found here at the newsgroup I attempted to
> manually compile and link _socket.so, then tried importing it from
> python.
> The results: An exception was thrown saying the symbol __eprintf was
> not defined in the shared object.
> Has anyone encountered this problem and have any help on this?
> What library is the symbol __eprintf defined in?

This usually means that you're linker and compiler don't match
(i.e. compiling with gcc but linking with the system ld).  Or maybe
openssl was compile with gcc? (quite likely if _socket is the only one
that fails) __eprintf is defined in libgcc.a, at any rate, so you
could try explicitly linking that in.

Cheers,
M.

-- 
  Well, yes.  I don't think I'd put something like "penchant for anal
  play" and "able to wield a buttplug" in a CV unless it was relevant
  to the gig being applied for...
                                 -- Matt McLeod, alt.sysadmin.recovery


From jb at cascade-sys.com  Fri Apr 26 01:26:30 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 22:26:30 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz>
Message-ID: <3CC8E506.82067BD8@cascade-sys.com>

Paul Foley wrote:

> No I don't have to admit that...I don't even have to admit that there
> are any parentheses in Lisp :-)

What color is the sky in your world?  ;o)

Seriously, I agree, the underlying abstraction -- the semantics -- does not require
parentheses.  But that's not really the language, per se.  I'm of the school that a
particular language, curly-L, consists of a tuple: ( Syntax x Semantics ).  Of course you
may change one component one and not the other but each change nevertheless results in a
distinct language.

I have seen "editors" that allow you to edit control structures in a purely graphical
form.  There was a 1:1 correspondence with a subset of the target language.  But to me the
graphical form really is is a different language or at least a different syntax for a
language, not the original language, like "RatFor" is to Fortran.

I agree, a proper editor can be a big help.  Ordinary VI's lisp mode is not bad, and
although in this case the parens could not be omitted, it allows you largely ignore them.
Still, you "have" to agree that most people's exposure is through the ascii
representation, in which case I think it's hard to argue parentheses not an intrinsic part
of the language (if "only" syntax).

That's why I like Python, as it provides access to a lot of the same functionality as Lisp
in a more familiar syntax, without all the tedious punctuation.

> In any case, parentheses are not used around sub-expressions, or to
> separate or group statements.  They're only used (in evaluated
> contexts) for "function" calls, and most other languages, including
> Python, use parentheses for that purpose, too:
>
>  Python: foo(x)
>  Lisp: (foo x)
>
> Same number of parentheses, right?

Besides, this is an entirely contrived, intellectually dishonest example.

First off, parens ARE used around sub expressions last I checked, unless I don't get how
you define sub expression.  E.g.,

    (foo (+ x (/ 1 n)))

vs.

    foo( x + 1 / n ).

So now compare

        if x > y or a = 0:
            a1 = m * x + b
        else:
            a2 = m / x - b

with

        (if (or (> x y ) (zerop a))
            (setq a1 (+ (* m x ) b )
            (setq a2 (- (/ m x ) b )).

In this case, Lisp requires about 33% more typing, almost all parentheses.

And, oh yeah, I forgot, aside from the parentheses, the pervasive prefix notation can be a
drag.  And I'm the type that still finds HP's stack notation less confusing in a
calculator than infix.  But having to mentally convert everything back and forth between
infix and prefix notation also can make you nuts.

That being said, I acknowledge that arithmetic is a small part of some large systems and
once you rise above the lower levels of abstraction, "everything" is function (or member)
calls, and your original example begins to be more valid.  Still, arithmetic crops up and
awful lot even then and can confound even the most enthusiastic programmer.

> I'd expect Lisp syntax to be
> rather less error-prone, if anything, since it's much simpler

I think that's debatable.  Certainly simplicity makes it easier to learn, so that's a
plus.  But it's still very tedious to get all the parentheses matched up just right.

> ? prefer controlling the semantics with an "Algol-like" syntax.
>
> Which is also possible in Lisp, of course; in fact, Lisp was
> originally intended to have an "Algol-like" syntax (McCarthy's
> M-expressions) on top of S expressions...

I'm not sure they're the same thing.

> ?? [And it's worth noting that, as with other aspects of Lisp, people
> ?? just keep reinventing it, badly --
>
> ? Hmmm.... I was recently accused of making this claim, which is entirely untrue.
>
> Eh?  What's untrue?  The claim or that you made it?

I only deny that I ever made the claim.

Interesting discussion.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From sholden at holdenweb.com  Tue Apr 30 08:12:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 08:12:08 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
References: <mailman.1020108957.30689.python-list@python.org>
Message-ID: <YUvz8.12366$xy4.11240@atlpnn01.usenetserver.com>

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in message
news:mailman.1020108957.30689.python-list at python.org...
[ ... ]
>
> The problem is compounded when you are not in control of the python on the
> system.  Imaging CGIs broken when the person's ISP decides to upgrade
their
> machines to a new OS rev which has a new python.
>

Thankfully, for the most part that's precisely what you have to do ...
imagine them.

reports-of-python-CGI-deaths-have-been-much-exaggerated-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From mgilfix at eecs.tufts.edu  Wed Apr 17 10:27:21 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 10:27:21 -0400
Subject: python threading
In-Reply-To: <a9jluf$kn8$1@newsreaderm1.core.theplanet.net>; from fgeiger@datec.at on Wed, Apr 17, 2002 at 01:22:27PM +0200
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <a9jluf$kn8$1@newsreaderm1.core.theplanet.net>
Message-ID: <20020417102721.K16954@eecs.tufts.edu>

  Don't have the original thread so I'll use this one. It's a bit
off topic but... With wxPython, while you want to know how python
threading and its primitives work, you always need to remember to
acquire the gui lock in threads that want to made changes to any
widgets from outside the main thread.  Note this does not apply to
callbacks and other stuff that is really executed by the main GUI
thread. Use wxMutexGuiEnter () and wxMutexGuiLeave () to wrap such
calls but remember to make calculations short because otherwise your
app will seem unresponsive.

                    -- Mike

On Wed, Apr 17 @ 13:22, F. GEIGER wrote:
> The wxPython demo program shows the use of threads within wxPython GUIs.
> Threads, that want to display something in say a notebook page, send
> messages to the main frame, which actually does the display.
> 
> There's also a Wiki (see wxPython.org), where advanced topics are discussed.
> 
> Cheers
> Franz
> 
> 
> "sameer" <sameer_ at email.com> schrieb im Newsbeitrag
> news:6bd9f01b.0204152132.6a6ad09e at posting.google.com...
> > I am trying to run multiple threads in a wxFrame class to instantiate
> > different panels within it on Win 2k,  and well, it's a little hairy.
> > Anyone got a good web source that discusses python threading in depth?
> >
> > More so, my question is, how can I get the return value of a function
> > that I run on a seperate thread?  I don't see anything for this in the
> > threading module.
> >
> > Thanks in advance for your help!
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (fgeiger)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/e5a66c57/attachment.sig>

From mixo at beth.uniforum.org.za  Wed Apr  3 06:36:09 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Wed, 03 Apr 2002 13:36:09 +0200
Subject: date / time
References: <3CAAD0D1.2040309@beth.uniforum.org.za> <a8elrm$r7du9$3@ID-41653.news.dfncis.de>
Message-ID: <3CAAE929.8000709@beth.uniforum.org.za>

Tobias Klausmann wrote:

> mixo <mixo at beth.uniforum.org.za> wrote:
> 
>>Suppose there is a given date, 2002-03-01 (YYYY-MM-DD).
>>How can you get the day (2002-02-28 ) before the given date and
>>check that it is valid?
>>
> 
> (I replace the interpreters >>> prompt with ### for better
>  legibility in USENET posts)
> 
> ### import time
> ### date="2002-03-01"
> ### time.strptime(date,"%Y-%m-%d")
> (2002, 3, 1, 0, 0, 0, 4, 60, 0)
> ### time.mktime(time.strptime(date,"%Y-%m-%d"))
> 1014937200.0
> ### time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60)
> 1014850800.0
> ### time.ctime(time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60))
> 'Thu Feb 28 00:00:00 2002'
> ###
> 
> Just remember that time.ctime() will give you local time
> while time.gmtime() will give you UTC. The input date is
> assumed to be the local time. If you drift between TZs, make
> use of the time.tzname variable.
> 
> By the way: what do you mean by "valid"?
> 
> Greets/HTH,
> Tobias
> 
> 

Thanks.

By "valid" date I meant subtracting  a day, and  then finding that the
new date does actually exsist.


One note/question on time.gmtime():
+++++++++++
time.gmtime( time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60))
+++++++++++

returns
+++++++++++
(2002, 2, 27, 22, 0, 0, 2, 58, 0)
+++++++++++

There is a day missing. If you add 1 second , i.e ((24*60*60)+1) you get
the required dated. What's going on?



From weeks at vitus.scs.agilent.com  Fri Apr  5 14:33:43 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Fri, 5 Apr 2002 19:33:43 +0000 (UTC)
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl>
Message-ID: <1018044224.815568@cswreg.cos.agilent.com>

Marek Augustyn (maug at poczta.onet.pl) wrote:
: Would somebody explain the following behavior (Python 2.2)?

: Function nested in class:
: (why nested scoping doesn't work?)

One way to put it is this: Python 2.2 did not create nested scopes (they've
always been there) or make nested scopes visible in function/method code.
It simply *increased* the visibility of nested scopes beyond the original
visibility of local+global.  How much is it increased?  Well, gee, I
forget.  (I guess the language is getting too big for me.)

Anyway, it increased visibility enough so that it is no longer possible to
call a function and pass it the scope as two dictionaries, which I had 
found handy on occasion.


Greg


From borcis at geneva-link.ch  Wed Apr 10 06:05:35 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Wed, 10 Apr 2002 12:05:35 +0200
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2CB16.8080103@geneva-link.ch>  <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org>
Message-ID: <3CB40E6F.90509@geneva-link.ch>

Laura Creighton wrote:

>>In article <3CB2CB16.8080103 at geneva-link.ch>, Boris Borcic wrote:
>>[snip]
>>
>>>(b) Laura made a lengthy point, overlooking the improvable while clear 
>>>to the language designer's POV, that it very much boiled down to : "I've 
>>>stacked my didactics of truth values in Python on the name of None, and 
>>>the PEP makes me feel I've been telling lies".
>>>
>>That's not how I saw it. The main point I extracted was: Until now,
>>the Pythonic "truth values" have been "something" and "nothing" (i.e.
>>the empty list, zero, None, etc.) With the bool type this is still
>>true, but we get a type which will give newbies the impression that it
>>isn't.


Well, cut "very much boiled down to" and paste "canonically equivalent 
over the right background".


> Magnus understood me very well.  I only want to add is that the 
> something-nothing is the real empirical truth of Nature, and all rest
> is formalism.


Don't take me too seriously above. You know what you have written, and 
so does everybody who read you.

I disagree with the letter of your addition, while I understand what you 
mean in context. The nothing/something matter has lots of aesthetics in 
it, and it is (was if you want) one of the subtle small beauties of 
Python. Python is like that. This is not an effect of God's will, but of 
Guido's instincts.

"Something-nothing is the real empirical truth of Nature" : with this 
I'd say again that you have something of a point but litteraly speaking 
saying what you say implies that we'd have a (adequate, consensual) 
understanding of the position of the *linguistic* zero. Which IMO - and 
globally speaking - is very far to be the case.

My point here is a bit diluted : an elegant way to put it, is that I 
think you are wrong if you think of language and minds as something 
outside Nature. To be exact, I don't think you do, but I do think your 
statement is given to be read to imply "Nature is outside language" and 
(by way of consequence) "there's no language in Nature".


> and to reject empty formalism (learning when formalism is empty and when 
> it is worthwhile is exactly what I am teaching and this is hard, hard,
> hard, hard, hard) then your job just got harder. 


Hmm. A bit flippantly, I'd be tempted to say that *of course* teaching 
something impossible to know is difficult to do. But the bit of truth in 
it wouldn't shine.

The real distinction is not between "knowing when formalism is empty" 
and "when it is worthwhile" but between "knowing why" you express 
yourself in such and such a precise way, and "not knowing why". Then is 
the question of "why might one fail to see the difference(s)".

> 
> In contrast --
> 
> Boris' phrasing indicates that I was 'trying to make a point' as if
> this was some sort of ego-driven competition between myself and Guido


Not at all my intention.


> with some sort of pay-off in terms of whether the PEP passes, or
> whether more big names on comp.lang.python come out in favour of me


Neither.


> than Guido or some other piece of nonsense.  I did not want this.


Acknowledged.


> Boris has revealed a flaw in our process, because that he took it that
> way means that the PEP process is being perceived as an ego-game.


Sorry, but I can't second your diagnostic here.


> 
> This _really_ stinks, people.

> I trusted Guido to be able to take off his
> 'I am an implementor and I just implemented a cool new feature' hat
> and put on his 'I am a langauge designer, and I am now looking at the
> language implications' hat.


In my view, you belittled the parallax between your POV as a teacher and 
Guido's POV as both large-scale coder and designer. To me it's more like 
"but given how much in the past I've demonstrated my attention to 
long-term language implications, and general conservativeness, can't you 
people give me some credit that I've pondered them to decisive details,
or do I have to face the perspective to remain bound not only to 
technical decisions but also to aesthetic decisions with little 
technical impact that I've made in the past and taken years to convince 
myself were OK to withdraw ?".

Guido is an conscientious artist, and it is a painful irony for a 
conscientious artist, to see a good art critic launch in passionate 
exegetics of the author's former aesthetic creativity at the instant he 
straws from a path of the type he'd up to then felt was his private 
business to pilot : the type of aesthetic choices.

> This trust we must have in Guido or else
> he will not be able to launch his own new features into Python.

> This
> is the faith we must have or else we think that we ONLY GOT LUCKY
> in that Python is great.


Right. 'xcept as a private matter and since you emphasize, LUCK
is not part of my primitive vocabulary.


> 
> Stop worshipping or demonising the man.  Only in bad American movies
> do the leaders of the military want blind discipline and worship.


Hmm. Should I read this as a comment on my

filter(lambda W : W not in "ILLITERATE","BULLSHIT") ?


> Leadership is hard, and way more fascinating and creative than that.


Au royaume des aveugles les borgnes sont rois.


> But this list is not the place to discuss this, either.


Who knows ?


> 
> Laura Creighton


You should read that I feel our particular attachment to Python (as 
compared to other people's) has similarities, despite my not aligning on 
your neatly motivated plea on PEP 285.

Regards, Boris Borcic
--
"Hope achieves the square root of the impossible"



From ruediger.maehl_nospam at web.de  Tue Apr  2 04:14:00 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Tue, 2 Apr 2002 11:14:00 +0200
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au>
Message-ID: <newscache$d8oxtg$iv4$1@www-neu.dzsh.de>

"Trung Hoang" <thoa0025 at mail.usyd.edu.au> schrieb im Newsbeitrag
news:pRzp8.4373$hi7.15765 at news-server.bigpond.net.au...
> What relational database is easy to setup and use with python? besides
msql
> (im having some problems)
>
> Cheeeers..
> trungie

SAPDB, a full featured database for free at www.sapdb.org
with Python interface.

R?diger




From logiplexsoftware at earthlink.net  Mon Apr 29 21:17:21 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 18:17:21 -0700
Subject: Remove masked elements from an array
In-Reply-To: <3CCDEF17.E4D1DC88@cascade-sys.com>
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>
	<20020429174425.GN16217@devel.trillke>
	<mailman.1020115270.9666.python-list@python.org>
	<aaknh8$1c4$1@peabody.colorado.edu>
	<3CCDEF17.E4D1DC88@cascade-sys.com>
Message-ID: <20020429181721.0da66800.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 18:10:47 -0700
James J. Besemer wrote:

> 
> Fernando P?rez wrote:
> 
> > James J. Besemer wrote:
> >
> > > THREE if you include LAMBDA.
> >
> > No, please heed other advice and use the MA approach. Please note that I'm
not
> > dissing this post on its own merits, it's a nice discussion, it just is
> > ill-adapted to numeric processing. Numeric provides its own _fast_ way to
> > manipulate arrays with invalid data, by all means use it. After all, speed
is
> > one of the big reasons why you use Numeric to begin with (you could do most
> > things in it with regular lists, if you had the age of the universe to
wait).
> 
> I would agree with you that it might be worth the extra trouble and effort to
> download, install and learn to use Numerics IF the user in question has a
big,
> production application to run.

Yes, but the OP clearly stated that he had a Numeric array that he wanted to
filter the values from.  Other than that, no argument.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From Answer.via.news.please.prikryl at skil.nospam.cz  Tue Apr  9 04:36:55 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Tue, 9 Apr 2002 10:36:55 +0200
Subject: True/False story... (PEP 285)
Message-ID: <a8u99o$2lcu$1@news.nextra.cz>

(The HTML version temporatily available at
http://www.skil.cz/test/TrueFalseStory.html;
converted by the reStructuredText tool -- see
http://structuredtext.sourceforge.net/)

Created: 9th April, 2002

Warning: It is rather long.

True/False story... (PEP 285)
=============================

This article should express my view of acceptance
of "PEP 285: Adding a bool type".

I want to present some points of view that (I
think) were not presented here this way.  Based
on my observations, I am going to suggest how
bool and even the strict implementation of bool
could be introduced into Python while preserving
the old code.  The ideas are rather *conceptual*
than *the lines of code*.

I am very new to Python, (say two weeks old
knowledge, started from zero) -- one reason why I
did not want to spoil the existing threads.  On
the other hand, I have working knowledge of other
languages, including C/C++ (the bool was
introduced in C/C++ with probably similar reasons
and similar problems like it is going to be done
in Python with presumably similar problems
related to backward compatibility).

I have also worked as a teacher earlier.  So I hope
I can compare my view with the view of other
teachers who participated in the discussion.

I have read threads related to PEP 285, including
the ones at Python-dev mailing list.  I have read
also the original proposal by Guido van Rossum
and also one of the longest saying *No* by Laura
Creighton.  I have noticed both positive and
negative reactions to both messages.

Firstly, I do not think that acceptance of PEP
285 was done without listening the opposite
arguments.  The discussion on Python-dev mailing
list (the URL to the archive was presented in the
"ACCEPTED" message) clearly shows that Guido did
NOT decided on his own, ignoring the voices of
others.  Frankly speaking, I have also found some
of the Laura's arguments for not accepting the
PEP slightly confusing (for me) in her own
context. Still, she definitely has the right to
express her opinion.  It was a good article --
wishing my was at least that good ;)

Here is the content of the rest of the article:

- `The booleans as the very basic concept`_
- `Learning the language vs. learning programming`_
- `What makes the good programming languages good`_
- `Implementation of concepts`_
- `Evolution vs. revolution (backward compatibility)`_


The booleans as the very basic concept
--------------------------------------

::

  > I am opposed to the addition of the new proposed
  > type on the grounds that it will make Python
  > harder to teach both to people who have never
  > programmed before and to people who _have_.  If
  > they have no preconceived idea of booleans, then
  > I do not propose to need to teach it to them in
  > an introductory lesson.  There is a time to
  > learn symbolic logic but while trying to learn
  > how to program for the first time is not it.
  > If, on the other hand, they _do_ have some
  > preconceived idea of booleans then Python will
  > not have what they want.  They will want
  > stricter Booleans that 'behave properly'.

I simply do not believe that introduction of
booleans makes teaching harder.  I think that the
opposite is true.

The *if-then-else* construct is very natural for
the procedural languages and it always
*conceptually* presents the expression that
decides the branching as the *conditional
expression*.  The same holds for conditions of
loop constructs.  The *conceptual result* of
such expression is always boolean -- true/false.
(The implementation of booleans is another story
-- see the `Implementation of concepts`_ below.)

For beginners (students), the notion of yes/no or
true/false plus some logical operations are quite
enough for introduction to booleans.  There is no
need to explain the details of the boolean
algebra.  Your are not explaining them the
details of how integers are defined
mathematically either, do you?

Even the small children are used to questions
like *Are you hungry?*  It's very unlikely that
they answer *Zero!*  Yet, they have no formal
knowledge of booleans. ;)

See `What makes the good programming languages
good`_ for the further discussion on booleans.


Learning the language vs. learning programming
----------------------------------------------

I have been teaching introductory/intermediate
subjects on programming for eight years
(University level, Dept. of CS).  So I hope I
have the right xsto talk about it.

I always tried to teach the concept first and the
syntax alongside as the second.  From my own
experience, the students were sometimes taught
programming (on a secondary school) the way that
I call *syntax-way of programming* -- which I
consider the really bad way of programming.

When converted to the working experience, I do
not like the *trial and failure* approach of
learning programming based on syntax knowledge.
To illustrate the approach in the Pascal
language...

   ... Let's type ``begin end.``, add something,
   compile it, and try what happens.

Well, it is an attractive way for many students,
because they immediately see some results (i.e
something is printed, something moves, etc.), and
they may feel powerful...

I prefer what I would call *semantics-based
programming*, or even better **thinking in
abstractions**.  Teaching to think in
abstractions is more difficult, but also more
important, in my opinion.  Well, there always are
some tradeoffs.  One should show students working
examples to attract their attention.  It is also
a good start to explain the syntax.

Think about the repeated "Hello World!" examples
in many languages.  The example should be
presented first as the one that displays (in some
form) the string of letters.  The syntax of how
it is done is important, but it is not the key
point.

Well, when presenting new language or system, the
key point may be different -- e.g. to show how
briefly it may be done using new language or to
compare what must be done to show it inside a
message window.  In such cases, one (often
silently) assumes that the abstract goal is well
understood; moreover, one usually assumes that
the students know how the "Hello World!" is done
in the other languages.


What makes the good programming languages good
----------------------------------------------

This title can be used as the question that is not
easy to answer. (If it were easy, only the
excelent programming languages would exist ;)

Say, a programmer thinks in abstractions that
nicely model the problem being solved.  Then the
programming language is good when it allows to
transform the mental picture of the solution into
the working application.  This means that the
language have to offer the implementation of the
abstractions (via language constructs or via
libraries) and it must be easy to rewrite your
ideas into the source text.  The language is even
better if it offers some efficient abstractions
that can be used for building the mental picture
of the solution (having for example The C++
Standard Library, or STL on my mind).  Say, the
concepts supported by the language *guide* the
mental process of searching for the mental
solution.

If the Object Oriented approach is to be taken,
then the language should support the Object
Oriented Programming.

In its domain the *language must be as clear as
possible*.  (Putting together this and the OOP,
the Perl vs. Python clearly shows the difference
here.)

The clarity of the language means that the
natural things to the *concepts* must be
naturally expressed by the *language constructs*.
If brevity does not spoil the things, then also
brevity of the language is welcome.

Now, back to booleans.

::

  > Python does not distinguish between True and
  > False -- Python makes the distinction between
  > something and nothing. And I think that this is
  > one of the reasons why well-written Python
  > programs are so elegant.  And this is what I am
  > trying to teach people.

Being very new to Python, I do not know if it is
the usual way of thinking of those who call
themselves *Pythonistas* ;-)  But I did not notice
that Python explicitly says anything about the
concept of *something* vs. *nothing*.  I have
noticed only the existence of ``None``, but I do
not think that the None fits well to your
*something/nothing* concept.

::

  > So I out-and-out tell people this.  {} is a
  > dictionary-shaped nothing. [] is a list-shaped
  > nothing. 0 is an integer-shaped nothing. 0.0 is a
  > float shaped nothing.

In my opinion, this is rather the source of
*concept confusion* than the way that makes
everything clear.  If I understand it well, you
say *something* vs. *nothing* concept brings
elegance into Python, and having *true* vs.
*false* would break it.

If you agree with the previous citation, then I
still feel that you are thinking in booleans
(*something* equal to *true* and *nothing* equal
to *false*), but you obscure it giving it other
names.  I guess I know the reason -- see `The
implicit conversion`_ in the following text.

::

  > I want to save them from the error of writing
  >
  > if (bool(myDict) == True):

In typical OO languages, the ``if myDict.empty():
some code`` or  ``if myDict.isempty(): some code``
is much more natural and understandable (closing
both eyes not to see ``b == True`` ;).  It's
because it is easy to think in yes/no (boolean)
answers.  It's because thinking in predicates is
natural not only to programmers.  When thinking
and writing code this way, your student would
probably not be tempted to write::

  if myDict.isempty() == True:
      some code

It looks simply stranger than::

  if myDict.isempty():
      some code

>From that point of view, the following code...

::

  if myDict
      some code

is brief but it is not as clear to beginners as
the previous one.  But I have to admit that it is
very well understandable for non-beginners, and
it may be prefered way for its brevity.


The implicit conversion
~~~~~~~~~~~~~~~~~~~~~~~

The key point which clearly fits with conditional
expressions of the boolean type is called **the
implicit conversion concept**.

In my opinion, explaining the concept of the
*implicit conversion* together with the *context*
where the construct is used allows one to explain
the things better than *something* vs. *nothing*
concept.  I am going to show that.

The Python 2.2.1c2 documentation (i.e. before
bool era) says in *2.2.1 Truth Value Testing*

  Any object can be tested for truth value, for
  use in an if or while condition or as operand
  of the boolean operations below. The following
  values are considered false:

  - None
  - zero of any numeric type, for example, 0, 0L, 0.0, 0j.
  - any empty sequence, for example, '', (), [].
  - any empty mapping, for example, {}.
  - instances of user-defined classes, if the
    class defines a __nonzero__() or __len__()
    method, when that method returns zero.

  All other values are considered true -- so
  objects of many types are always true.

  Operations and built-in functions that have a
  boolean result always return 0 for false and 1
  for true, unless otherwise stated. (Important
  exception: the boolean operations "or" and
  "and" always return one of their operands.)

In other words, it says that in the *boolean
context* the named cases are *converted* to
*false* or *true* values respectively.  So, there
is nothing like...

::

  > {} is a dictionary-shaped nothing. [] is a
  > list-shaped nothing. 0 is an integer-shaped
  > nothing. 0.0 is a float shaped nothing.

Again, it only obscures the things.  The students
should have asked "What do you mean by the
*shaped nothing*?".  (At least for me, it is far
more obscure than the simple notion of yes/no or
true/false.)

Compare it with what the documentation already
says (in different wording):

  When an object appears in a boolean context,
  the implicit conversion is done.  The following
  objects are implicitly converted to False:

  - None
  - zero of any numeric type, for example, 0, 0L, 0.0, 0j.
  - any empty sequence, for example, '', (), [].
  - any empty mapping, for example, {}.
  - instances of user-defined classes, if the
    class defines a __nonzero__() or __len__()
    method, when that method returns zero.

  All other objects are implicitly converted to True -- so
  objects of many types are always converted to True.


Implementation of concepts
--------------------------

Let's start with the following comment::

  > The problem is that _everybody_ has some
  > conceptual understanding of True and False.  And
  > whatever that understanding is, it is unlikely to
  > be the one used by Python.

When a programming language offers the direct
implementation of an abstraction, then it is
difficult to imagine how this could be done
better.  Then we can only wish for more efficient
(in the terms of performance) implementation.

I simply cannot imagine a different understanding
of the *true/false concept*.  I think that it is
so easy to understand, that there is almost no
space for confusion.  I am not talking here about
details of boolean algebra, I am talking about
how would you explain the concept of true/false
to anyone who is capable of the basic abstract
thinking.

To put it together, introduction of *bool* with
*True* and *False* values into Python cannot be
wrong -- conceptually.  On the other hand, it may
be difficult to do it smoothly considering the
backward compatibility.


Booleans implemented as int in C/C++
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The example of the unfortunate mixture of the
concept of booleans and integers, that is
identical to the situation in Python, can be
observed in the C/C++ languages.

Clearly---as someone in PEP 285 threads already
said---the roots are in the memory representation
of booleans and in available machine instructions
for testing them.  We may think about the *if*
command int the C language as about somehow very
low-level or simple-to-implement command,
because there always was the *jump-if-zero*
instruction in the machine code.

The smallest possible integer with well chosen
values seems to be natural for implementing
booleans.  Because the FALSE abstraction was
implemented as zero, the code generated from
the *if* statement ended with testing the
condition against zero.  This way there was no
*implementation* difference between FALSE and
zero and it was natural to represent TRUE as
*anything else than zero* --- because the jump
instruction lead to the second branch of the
*if-then-else* construct.

In my opinion, this was the reason why the
boolean type was not supported in C by the
language from the very beginning. (Pascal, on the
other hand, was written as the language for
teaching. That probably was the reason for
introducing boolean as one of the basic types.)

Unfortunately, that easy way of interpretation of
integers as booleans lead to the programming
style where returning integer value is perceived
as very natural even in situations where only two
possible cases may happen.


When zero means O.K...
~~~~~~~~~~~~~~~~~~~~~~

If you learned Unix or the like OS, you may
remember how *exit code* of utilities is defined:

  zero
    Everything went O.K.  No problem detected.

  positive
    Something wrong happened.

In shell scripts, you may conditionally execute
utilities in a sequence where the execution of
the second is done only if the previous utility
exited with zero exit code (or the reversed).
The syntax resembles using of the boolean
expressions, but one have to think twice to use
it correctly (zero means O.K. termination for the
utility but false for the boolean expression).


Evolution vs. revolution (backward compatibility)
-------------------------------------------------

No doubt, working code should not be broken by
new features -- if possible.

I may be wrong, but it seems to me that
introduction of the *strict bool* together with
`The implicit conversion`_ of bools in the *int
context* would solve many problems discussed in
the threads. The possible warning about implicit
conversion of bool into int could be the future
mechanism to purify the existing code later
without breaking it.

I do not know the details of the Python
implementation, so I may easily be wrong.  But I
also do not say that the things must be
implemented by the context recognition and the
implicit conversion.  I only suggest that this
concept may help to think about special cases
when the implementation details should be
decided, for example::

  >> comparing True/False with other values should give the same
  >> result as when it were used with an 'if':
  >> >>> True == 1, True == 6
  >> True, True
  >
  > So 6 == True == 1, but 6 != 1?

The problem can be converted to the question
"What is the context of the used bool?". Without
discussing some details, let's assume that the
context is *boolean*.  Then the other arguments
should be *IMPLICITLY* converted to bool, like
here::

  >>> True == bool(1), True == bool(6)
  True, True

  bool(6) == True == bool(1)

which apparently holds.

If the decision was the opposite (for some
reasons unknown to me), then the bools should be
*IMPLICITLY* converted into int::

  6 == int(True) == 1

which apparently never holds, regardless the old
application, the new bool implemented over int,
or the strict implementation of bool.

Then even the following example is easy to
explain::

  When  6 == True  and  7 == True  and  True == True,
  then why 6 != 7?

Simply, bool cannot be *conceptually* directly
compared with int.  Some implicit conversion must
happen, depending on the context (the boolean or
the integer one).  This implicit conversion is
invisible to the reader of the source, but it
changes the meaning of the bool value or of the
int value -- depending on the context.  Because
of that one cannot derive the fact ``6 == 7``
from the previous equations.  In other words, the
equations make sense *only* when the implicit
conversion is taken into account.

To summarize: Thinkink in *context* and
*implicit conversion* terms makes things
unambiguous (at least, I hope so ;)

>From PEP 285 proposal and direct reactions::

    1) Should this PEP be accepted at all.

**YES** (i.e. big yes), I definitely would be for
accepting the PEP if it was not accepted already
;) I hope I did explain my reasons in the
previous text.

::

    2) Should str(True) return "True" or "1": "1" might reduce
       backwards compatibility problems, but looks strange to me.
       (repr(True) would always return "True".)

I guess that it requires the "proof by fire". If
some test utilities rely on 0/1 results, they
should be updated to bool anyway.  I prefer
str(True) returning "True".

::

  > There's also another issue here: "True" and "False"
  > are English words, "0" and "1" are language neutral.

As someone else said, I do not think that it is
the problem.  If some application is tailored to
be more user friendly, then it probably convert
the 0/1 into some meaningful text.  If it is only
for an administrator or a programmer (expected to
understand English), then "True" is quite
descriptive.

Similarly, if some log file (for example) relies
on 0/1, then it should be easy to fix the script
that produces the content to produce 0/1
explicitly.  There should not be too much well
written code that relies on str() in such cases.

::

    3) Should the constants be called 'True' and 'False'
       (corresponding to None) or 'true' and 'false' (as in C++, Java
       and C99).

In my opinion, it is rather minor issue.
Mentally, the "true", "True", "TRUE", and "tRUe"
is the same (the last one looking too weird to be
accepted ;-)

::

    4) Should we strive to eliminate non-Boolean operations on bools
       in the future, through suitable warnings, so that e.g. True+1
       would eventually (e.g. in Python 3000 be illegal).  Personally,
       I think we shouldn't; 28+isleap(y) seems totally reasonable to
       me.

I am not going to comment the ``isleap``
identifier. I understand clearly what the
question wants to ask ;)

I would recommend to think in the *implicit
conversion* category in such cases in the *given
context*.  Here, the isleap() is used probably in
numerical context, so there sould be implicit
conversion of possible boolean result of isleap()
into int (think about bbb() if the isleap() makes
you crazy ;) In other words, if the expression
``28+bbb()`` is used in the numeric context
(determined by the plus operator) I would prefer
it to be implicitly converted ``28+int(bbb())``
in the case when the bbb() does not return int.

If the implicit conversion could easily be
implemented into new version of Python, then it
would make no difference if the booleans were
implemented on the top of integers or if they
were implemented in the most strict way.  Then,
the new (even strict) booleans could be smoothly
mixed with integers.  The enforcement of using
strict booleans can then be solved by deciding
whether warning is to be issued when the implicit
conversion is done (in specific cases) or not.

Again, I do not say that there *must* be
introduced the term of implicit conversion into
the Python implementation.  Often, it may be
enough to think about it this way in cases when
the things are not clear.


Thanks for your attention,
  Petr

--

Petr Prikryl (prikrylp at skil dot cz)





From sid at claudius87.plusserver.de  Sat Apr 13 09:46:54 2002
From: sid at claudius87.plusserver.de (Juergen Gather)
Date: Sat, 13 Apr 2002 15:46:54 +0200
Subject: porting python to embeded devices
In-Reply-To: <3CB83354.90CB406@alum.mit.edu>
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net>
 <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org>
 <7xy9fse6ol.fsf@ruckus.brouhaha.com> <slrnabf5r2.s6b.grey@teleute.dmiyu.org>
 <7xr8lk46nl.fsf@ruckus.brouhaha.com> <3CB83354.90CB406@alum.mit.edu>
Message-ID: <6765943231.20020413154654@claudius87.plusserver.de>

WW> Paul Rubin wrote:
>> The SL-5500 hardware is similar to an Ipaq 36/37xx plus it has a cute
>> little slide-out keyboard that's not all that useful.  I got the
>> developer version for $199.00 which was a good deal at the price.  I
>> wouldn't recommend it at $500 though.  I'll post a review to my website soon.

WW> This thread got me curious about it. There are SL-5000D's (the developer
WW> version) on eBay trading in the ballpark of $300. Is there any significant
WW> difference between the 5000D and the 5500?

The 5000 is 32MB and the 5500 is 64MB. And due to reports on several
sites a RAM upgrade is impossible at the moment.





From tejarex at yahoo.com  Fri Apr  5 12:03:02 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Fri, 05 Apr 2002 17:03:02 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <3CAAC1FB.2050105@mxm.dk> <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com> <3CAD53EC.1010703@mxm.dk>
Message-ID: <aNkr8.13682$w7.998355@bin6.nnrp.aus1.giganews.com>

"Max M" <maxm at mxm.dk> wrote in message news:3CAD53EC.1010703 at mxm.dk...
> What I don't understand here is that it seems that people considers
> structured text to be difficult. It really isn't. Any programmer can
> learn the basics in 15 minutes or less.
>
> And it really is very very easy to write documents in it. I do
belive
> that if I where ever to write a book (again) i would write it in
ReST.

I have had the same thought myself.

> ########################################
> # 'old' structured text
> ########################################
>
> This is an example of a headline in structured text
>
>      Here is a paragraph.
...

vs.

> ########################################
> # ReST
> ########################################
>
> This is an example of a headline in structured text
> ===================================================
>
> Here is a paragraph.
...

I consider marking subheads and NOT indenting paragraghs to be a
reason to prefer ReST.  Thanks for the feature-by-feature comparison.

Terry J. Reedy





From sholden at holdenweb.com  Mon Apr  8 11:09:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 11:09:35 -0400
Subject: rowcount in dbi w/ odbc
References: <a8iv0i$56r$1@news.ycc.yale.edu>
Message-ID: <uqis8.1061$eJ6.328@atlpnn01.usenetserver.com>

"don" <donald.braman at yale.edu> wrote in message
news:a8iv0i$56r$1 at news.ycc.yale.edu...
> I'm new to Python programming. I am trying to gain access to specific
> columns in specific rows in an Access database, but am not having much
luck.
> I thought I'd first try to access specific rows and columns. However, when
I
> try to use "rowcount" to see how many rows there are, I get
"AttributeError:
> rowcount". I assume there is some easy way to do this, but I can't seem to
> figure it out. Here's my attempt that fails:
>
> import dbi, odbc
> try:
>         con = odbc.odbc('mydatabase')
>         cur = con.cursor()
>         cur.execute('SELECT * FROM mytable')
>         con.close()
>
>         for row in range(0, cur.rowcount()-1):
>                 for col in range(0, cur.colCount()-1):
>                         print cur.getField(row, col), ',',
>                 print cur.getField(row, col)
>
> except NameError,e:
>         print 'error ', e, 'undefined'
>
Note that the odbc module is not fully conformant to the Python DB API, but
that even conformant modules do not have to implement rowcount. It's
explicitly stated that the value rowcount provides may change as access to
the retrieved data takes place.

>
> I suppose I should note that I can access the data using the following,
but
> that it doesn't give me access to a specific rows:
>
> import dbi, odbc
> try:
>     s = odbc.odbc('mydatabase')
>     cur = s.cursor()
>     cur.execute('SELECT * FROM mytable)
>     print cur.description
>     for tup in cur.description:
>         print tup[0],
>     print
>     while 1:
>         rec = cur.fetchmany(1)
>         if not rec: break
>         print rec
>
> except NameError,e:
>         print 'error ', e, 'undefined'
>
The easiest way is to use the returned data to answer your questions, with
code like what follows.

    cur.execute("SELECT * FROM mytable")
    data = cur.fetchall()
    print "Row count is", len(data)

Then handle the rows with something like

    for row in data:
        print "Column zero is", row[0]

If you need the database to explain the structure of the data, again you
should be aware that not all modules and database make a complete
description of hte returned data available. Take a look at

    http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/81189

for one way to get some insight into the structure of a query's output.

regards
 Steve






From peter at engcorp.com  Tue Apr  9 01:12:41 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 01:12:41 -0400
Subject: Need help with distutils
References: <mailman.1018083739.10368.python-list@python.org> <Xns91EA988AA8260andy47halfcookedcom@203.109.252.31> <slrnab3gei.1i0.akuchlin@crystal.mems-exchange.org>
Message-ID: <3CB27849.104E33E3@engcorp.com>

"A.M. Kuchling" wrote:
> 
> In article <Xns91EA988AA8260andy47halfcookedcom at 203.109.252.31>,
>    Andy Todd wrote:
> > notice that the pattern you pass to global-exclude has to be a Unix-style
> > "glob" pattern. In that notation '.' is a special character so using
> > '.cvsignore' doesn't actually match your files ;-)
> 
> No, '.' is not special in glob notation.

I believe it is special in glob.py, which explicitly excludes files
with a leading '.' unless the pattern explicitly includes a leading '.'
This may not be the issue involved, however.

-Peter


From paulo.pinto at altitude.com  Fri Apr 12 09:10:05 2002
From: paulo.pinto at altitude.com (Paulo Pinto)
Date: Fri, 12 Apr 2002 13:10:05 GMT
Subject: 4Suite for Python 2.2
References: <zxft8.37$Mh5.1049@nreader1.kpnqwest.net> <3CB64FFD.6D0D6AD7@engcorp.com> <rNut8.1$xV3.411@news.ecrc.de>
Message-ID: <N0Bt8.143$2L2.1265@nreader2.kpnqwest.net>

Yes ! It does work.

I was trying to use the binary distribution in a Windoze machine
but the latest version looks for Python 2.1. Since I have version
2.2 it didn't find it.

Now I have installed from the source and it works.

Thanks.

Paulo Pinto


"Markus Jais" <mjais at web.de> wrote in message
news:rNut8.1$xV3.411 at news.ecrc.de...
> Peter Hansen wrote:
>
> > Paulo Pinto wrote:
> >>
> >> Hello,
> >>
> >> does anyone know when 4Suite for Python 2.2
> >> would be available ?
> >>
> >> I can only find support up to version 2.1 :(
> >
> > Doesn't it run under 2.2?  (I haven't tried and don't know --
> > this is a serious question.)
> >
> > I thought 4Suite was simply a very large collection of XML
> > processing tools in native Python (which I would have
> > expected would still run under 2.2).  Or are there binary
> > libraries which are keyed to python21.dll or something?
> >
> > -Peter
>
> hello
> I have it running under python 2.2 and it seems to work fine.
> have just used the 4xslt command line tool until now
> but this works good.
>
> markus




From eppstein at ics.uci.edu  Mon Apr 15 00:12:27 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 21:12:27 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com>
Message-ID: <eppstein-9E09EB.21122714042002@news.service.uci.edu>

In article <3cba4cf7 at news.mhogaming.com>,
 "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:

> The other thing I was trying to say, is that if this _BUG_
> (gee- that wasn't so hard!) in Python is fixed, then we need
> a way to consistently do all of this bit shifting.  Now that
> ints and longs are 'unified', you are supposed to be able to
> write code that doesn't care what the internal
> representation is-- that should also apply to the '<<' and
> '>>' operators.  BUT, since we NOW don't know where the high
> bit is (but the class that is managing the long object
> DOES), we need a '>>>' operator to consistently handle a
> logical right shift-- (i.e.: there is no way to code this
> behavior in Python.)

There is no "highest nonzero bit" -- if it's a negative number, the bits 
in all sufficiently high positions are nonzero.  The long class code 
doesn't know where the highest meaningful bit is in an value that is 
being used as a bitstring, it only knows where the highest 
explicitly-represented bit is, and that's not the same thing.

If you do somehow know where the highest meaningful bit is, you can 
always do something like this:

def shiftLeftMasked(x, shiftAmount, highBitPosition):
    return (x & ((1L << highBitPosition) - 1)) >> shiftAmount

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From trondga at -NOSPAM-ifi.uio.no  Wed Apr  3 22:07:45 2002
From: trondga at -NOSPAM-ifi.uio.no (Trond Gaarder)
Date: Thu, 4 Apr 2002 05:07:45 +0200
Subject: Newbie: Alarm at end of program
References: <es5HyN42BHA.1996@cpimsnntpa03>
Message-ID: <a8gfr8$h4u$1@maud.ifi.uio.no>

Don't know if this is the best/correct way of doing it in Python, but it
works for me (on a PC):

print "\a"

This should emit a beep from the pc. What it sounds like is highly
machine-dependent, however...


-Trond


"billludden" <billludden at email.msn.com> wrote in message
news:es5HyN42BHA.1996 at cpimsnntpa03...
> Hi,
> I have written a script that runs from a few minutes to several hours
before
> it finishes.  How would I make the computer play a 'beep' to signal that
it
> is finished?
> Thank you,  Bill
>
>




From ralph at inputplus.demon.co.uk  Sun Apr  7 11:14:37 2002
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 7 Apr 2002 15:14:37 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk> <3CB0495F.7010507@mxm.dk>
Message-ID: <a8pnot$eck$1@inputplus.demon.co.uk>

Hi Max,

> > I think we all perfectly understand your opinion too and no matter how
> > many times you repeat it there will still be some of us that like to do
> > 
> >     a = b + is_movable(c
> 
> I can see that it is clever, but I don't get why it is smart. Isn't it 
> just another way of writing?:
> 
>      if is_movable(c):
>          a = b + 1

No.  BTW, you snipped a closing parenthesis there.

It's equivalent to

    a = b
    if is_movable(c):
        a += 1

or

    if is_movable(c):
        a = b + 1
    else:
        a = b

or

    a = b + is_movable(c) ? 1 : 0

but that last one isn't valid Python but does show a strong C
background is likely to make you familiar with this approach.

> If it is, I find that it is more obfuscation than expression. It's
> akind to using a magic value. My paraphrase of our code should really
> read something like::
> 
>      motionLength = 1
>      if is_movable(c):
>          a = b + motionLength

That doesn't cover `not is_movable(c)' again.

> Or your code should possibly have read::
> 
>      a = b + motionLength(c)
> 
> But probably I misunderstand you intentions.

The quantity being added to b isn't a length, it's 1 indicating that c
is movable.  Trying to name it anything other than what it is is
contortion for some ideal that doesn't work in this case.

Say you had a list of items and wanted to sort them to place those with
the most abilities first, with all abilities considered equal, and then
sorting on the item's name where two have the same number of abilities.

To calculate the major sort key you might do

    major = can_be_moved(o) + can_be_sized(o) + can_be_stacked(o) +
        can_be_coloured(o) + can_be_hidden(o)

as opposed to

    major = 0
    if can_be_moved(o):
        major += 1
    if can_be_sized(o):
        major += 1
    if can_be_stacked(o):
        major += 1
    if can_be_coloured(o):
        major += 1
    if can_be_hidden(o):
        major += 1

I think the former is way clearer, and not because it uses less
key-strokes, although brevity can aid clarity sometimes.

Cheers,


Ralph.



From jblazi at hotmail.com  Tue Apr  9 08:06:39 2002
From: jblazi at hotmail.com (Janos Blazi)
Date: Tue, 9 Apr 2002 14:06:39 +0200
Subject: Simple question! urgent
References: <62l5buo5do5eh0h93b4r9ua8skp61b7302@4ax.com>
Message-ID: <3cb2d8cf_6@news.newsgroups.com>

> How to get NUMBER OF ITEMS in array? I need it for checkin nr of items
> in argv[]

simply len(x)
J.B.




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From dale at riverhall.NOTHANKS.co.uk  Tue Apr 16 15:21:05 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 16 Apr 2002 20:21:05 +0100
Subject: Brackets in brackets, braces in braces
References: <2e1a250.0204160730.637139f1@posting.google.com>
Message-ID: <q8uobu8vs2eiq7olfbsg2tmv3uri37b26g@4ax.com>

ahorvath at szgti.bmf.hu (Horvath, Arpad) wrote:

>How can I find strings like this with regular expressions:\something
>[something1] {something2} {something3}Where in the braces can be also
>another braces. For example:\frac{1\times10^{-16}}{4}   can be at the
>place, where something2 is.(braces in pairs)\interval [J]
>{\frac{1\times10^{-16}}{4}} {1..5}My half solution is at
>www.szgti.bmf.hu/~ahorvath/regexp.txt.

You can't do it with re but, if I recall correctly, someone posted a
link to a class based on SPITBOL/SNOBOL's pattern matching engine a
little while ago.

That would have no trouble with nested structures.

--
Dale Strickland-Clark
Riverhall Systems Ltd



From Oschler at earthlink.net  Sat Apr  6 05:31:20 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Sat, 06 Apr 2002 10:31:20 GMT
Subject: Destructors
Message-ID: <Y7Ar8.5367$Tf4.2431098@news2.news.adelphia.net>

My favorite object construct/destruct paradigm is the Boost Library's
shared_ptr template class for C++ objects.  When the reference count of the
shared pointer reaches zero, the destructor of the object it wraps is
called.  Delphi has "interfaces" which do the same, although they behave
differently than classes do when it comes to inheritance and other subtle
programming aspects so it can be a little awkward at times.

Python seems to follow the Java finalize/gc() model where destructors (or
finalize() in Java) aren't called until the garbage collector frees the
object.  As in Java, the Python book I'm reading stresses that you have no
idea when the destructor will be called, or if it will happen at all.  Just
like the Java books I have, the Python book recommends creating a method,
for example close(), to be called explicitly, probably in a finally portion
of a try/finally block.  So of course you have to remember, for every class
that needs to immediately free a resource when the class object refcount
reaches 0,  to pair each such class instance creation with a call to the
custom "close()" method and wrap the code in a try/finally block.

Is there a way to achieve the same functionality as the shared_ptr template
class (stack-based refcounted object) in Python?  If not, is there a place
to post a language feature suggestion for discussion and possible review by
the Python Illuminati?

thx
















From brian at sweetapp.com  Mon Apr 29 02:15:02 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 28 Apr 2002 23:15:02 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <004901c1ef45$328df2f0$445d4540@Dell2>

> The functionals (map, filter, and reduce) had greater importance prior
> to the introduction of list comprehensions which are now the preferred
> (and more readable) approach.  

reduce() does not have a list comprehension equivalent. What about
zip()?

> Pow() and divmod() are more related to math module functions than any
> of the built-in functions.  In general, they are rarely used.  Because
> the ** operator is available, the pow() function is rarely called
> directly.

pow() has an optional third argument, so the ** operator cannot act as a
direct replacement. Also, a pow function, with different semantics,
exists in the math module.

I've seen divmod() used in lots of code. 

And, if you are going this far already, why shouldn't abs() and round()
be moved too?

In any case, math seems like an odd home for these functions as all of
the routines currently in math work with doubles.

> The -b command line option and related environment variable would go 
> into Python 2.3 and remain available in perpetuity.

1. Using a command-line switch seems evil. The first 3rd party module
that someone uses will provoke them to add the -b switch, which will not
encourage them to fix their own code. 
2. Leaving the switch in forever doesn't make sense. 

Cheers,
Brian 





From JamesL at Lugoj.com  Sat Apr  6 15:13:49 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 20:13:49 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84> <a8lvod$589$1@panix1.panix.com>
Message-ID: <Xns91E87D184B4BCJamesLLugojcom@216.148.53.83>

aahz at pythoncraft.com (Aahz) wrote:
> Python 2.1 (particularly 2.1.2) is a fine release.  There's no
> compelling reason to upgrade unless you care about the type/class
> unification.  But it's foolish at best to stick your nose into a thread
> about language changes without finding what the current state of the
> language is.
> 
> You seem to make a habit of irresponsible posts.  (Remember your
> actions in the comp.distributed brouhaha?)

Ad hominem attacks, such as calling someone foolish and irresponsible are 
irrelevant to the issue being discussed and needlessly inflammatory. At 46, 
I do try to limit my foolishness and try to act responsibly. But when I 
fail (and I do sometimes, despite my best efforts) and I manage to 
recognize such failures I try to find the courage to apologize for any 
damage done and do what I can to rectify the situation. But I don't yet see 
why _I_ have to apologize for being a victim of my published 1.5.2 scripts 
breaking on other releases, despite claims of forward compatability.

> The current standard is approximately one release every six months.

If that rate is maintained, then I suspect that in a few years I will have 
a lot more company.


From merkosh at hadiko.de  Thu Apr 18 13:22:04 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Thu, 18 Apr 2002 19:22:04 +0200
Subject: problems with Perl; changing to Python
Message-ID: <MPG.17291f2bfcb85301989684@news.rz.uni-karlsruhe.de>

[This followup was posted to comp.lang.python and a copy was sent to the cited author.]

Hallo,

I'm a troubled Perl programmer who's got enough of Perl and needs 
another programming language. Perhaps you can help me here.
I've got a view questions I myself cannot answer. Perhaps those of you 
who got more experience can:

Perl is great for scripts. Perl is great for programs with less than 
perhaps 200 lines. A program should be aware of its memory usage; for a 
script I don't care.

I need a language that can be used to write programs with. GUI, and 
commandline tools. When the program run in background they shouldn't 
take up more memory than necessary.
F.e. I wrote a Perl script checking for changes on homepages, expiry 
date, etc in regular intervals. This "program" is 600 lines long and 
uses 7mb of ram... too much in my oppinion.
Can I do that in Python?

I'll need to write programs which have to be extensible. The program 
may, during calculation require up to 500 mb ram. Fine with me. On the 
other hand I'll neet to write something working in background and then 
(depending on its task and size) 1-2 MB should be enough. 
What about Python here?

Installing Perl modules was a torture. I am sitting behind a firewall 
and in order to install the modules (compiling them did never work on my 
Windows 2k mashine) I had to run my own DNS server and tunnel the 
connection. Will that work better with Python? (f.e. Proxy support would 
be sufficient!)

And for the GUI programming: I used Perl/Tk previously, so Tkinter would 
be most fammiliar, I guess. What other GUI packages can Python offer me 
that will work (perhaps even in compiled .exe format) on windows 
mashines? What about features like "minimize to tray" or "always on 
top" that I wasn't able to accomplish in Perl/Tk?

I need a language where I can rely on OOP if I need to. A language where 
there's a difference between class- and instance- varriables. 
How about multiple inheritance (perl, c++) or interfaces (jave) in 
python?

Are there any applications written in Python?
Am I right here with Python or will I have to stick to a hybid C++/Java 
solution?

Thanks in advance - 
Yours
Uwe
--
C makes it easy to shoot into your foot.
C++ makes that harder, but when you do - it blows away your whole leg.


From achim.domma at syynx.de  Sun Apr 28 06:09:40 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Sun, 28 Apr 2002 12:09:40 +0200
Subject: Help again: py2exe and win98
References: <LxPy8.72046$vF6.2149827@news2.tin.it>
Message-ID: <aaghao$d9u$04$1@news.t-online.com>

"Utente" <prova at qui.com> wrote:
[...]
> >    add_resource(exe_name, str(vr), RT_VERSION, 1, delete)
> >RuntimeError: BeginUpdateResource: La funzione ? valida soltanto in
> modalit? Win
> >32.
> >
> >C:\dati\Python\PrimaNota
>
> Where "La funzione ? valida soltanto in modalit? Win32" is the italian for
> something like "The function is valid only in Win32 mode".
> Windows 98 is not a win 32 system???

I don't know anything about py2exe, but BeginUpdateResource is a function
which is not supported on Win95/98. Look at http://msdn.microsoft.com if you
are interested in details.

greetings
Achim




From aahz at pythoncraft.com  Mon Apr  1 11:07:18 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 11:07:18 -0500
Subject: Files, Seek, and Readlines. A newbie question
References: <Ts%p8.149254$ZR2.64157@rwcrnsc52.ops.asp.att.net>
Message-ID: <a8a0jm$s2a$1@panix1.panix.com>

In article <Ts%p8.149254$ZR2.64157 at rwcrnsc52.ops.asp.att.net>,
John Purser <NO_SPAM_jmpurser2 at attbi.com> wrote:
>
>I'm using a large (>120,000 lines) sorted ASCII text file as an index.  I
>started running into problems as soon as I began to use the .seek() to
>position before a read.  Even when not opened as a binary file seek moves by
>bytes not lines.  Is there a line oriented seek version?  Is there a better
>way to get to line 50 than:
>myFile.seek(0)
>myFile.readlines(49)
>Record = myFile.readline()

Not unless each line has a fixed length.  Better yet, use a database.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From mats at laplaza.org  Wed Apr  3 14:43:28 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Wed, 03 Apr 2002 19:43:28 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de>
Message-ID: <3cab55f4.61190667@news.laplaza.org>

On 3 Apr 2002 08:08:18 GMT, Tobias Klausmann
<klausman-un030402 at tuts.net> wrote:

:Cliff Wells <logiplexsoftware at earthlink.net> wrote:
:> On 01 Apr 2002 21:56:44 -0500
:> nbecker at fred.net wrote:
:> 
:>> Keep lobbying Redhat.  The correct solution is for them to put
:>> 
:>> #! /usr/bin/python1.5
:>> 
:>> If they need a specific version.  I argued this with them already.
:>> Keep up the pressure.  Maybe they will get it.
:> 
:> Absolutely agree.  Or better #!/usr/bin/python.rh with a link python.rh ->
:> python1.5 so they can eventually upgrade without changing every shebang.
:
:In a related matter: how common is /usr/bin/env ?
:I really like that tool to run python wherever it is: 
:
:#!/usr/bin/env python
:
:As it is part of sh-utils, I guess it's pretty widely used.
:
:Anybody know of specific environments where it is not commonly
:available?

Windows :-)

It's a standard utility from the UNIX world so it's pretty widespread.

The thing to remember with env is that it uses the current environment
to lookup the thing named python.  Which is great if you can precisely
control this in all cases that matter.  The environment /is/ going to
be different between your interactive shell and some other program
started in a different context, such as "cron" (which could be asked
to run a Python script for you), or a webserver (as in using Python
for cgi scripts).   That's fine as long as you remember to check what
"env" will find in non-shell contexts, and make any necessary
adjustments.
Mats Wichmann



From Scott.Daniels at Acm.Org  Sun Apr 14 17:43:40 2002
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Sun, 14 Apr 2002 21:43:40 GMT
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4>
Message-ID: <1104_1018820620@news.dsl-only.net>

On 14 Apr 2002 02:24:24 GMT, Philip Swartzleonard <starx at pacbell.net> wrote:
> Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:
> > Philip Swartzleonard <starx at pacbell.net> wrote
> >> The first section of the index[1] for the language reference manual
> >> lists all of these things (the underscore section), with links.
> > 
> > Thank you very much for pointers, but I can think of something not
> > listed there. For example, __all__ for "from module import *", pickle
> > protocols, __new__ for subclassing from a type and so on. Is there a
> > complete (or near-complete) list?
> ...

You could also take this lack of a list as a charter for your collecting _all_ 
(not __all__) such names, and creating a document for others who, like you
want to know the list of such names.  This could include package references 
if you so desire (others will complain if you leave something out, providing 
you with more fodder for your document).

-Scott David Daniels
Scott.Daniels at Acm.Org





From aleax at aleax.it  Sun Apr 14 03:44:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 14 Apr 2002 07:44:47 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <roy-0BE2A8.19181913042002@news1.panix.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Prau8.17347$vF6.558738@news2.tin.it>

Martin v. Loewis wrote:

> Roy Smith <roy at panix.com> writes:
> 
>> Any operator can get replaced by a function call.  It's just a matter of
>> what's more convenient to write.
> 
> Indeed, and so it is desirable to have infix shorthand notations for
> the frequent operations. However, as the number of infix operators
> grows, the convenience loses: you now have to start remembering line
> noise.

Very well put.  Indeed, given the small amount of bit-twiddling in
typical Python usage, I think & ~ | ^ are already overboard -- they
make a lot of sense in C, a very low-level language that is used a
LOT for bit-twiddling, but, in Python, named built-in functions might
have been more appropriate in hindsight (fewer precedence levels to
remember) -- if not, even, functions in a separate module.

A simple histogram of non-alphameric tokens in Lib/*.py only partially
conforts this thesis:

     0  18495 )
     1  18495 (
     2  16820 .
     3  15467 :
     4  15327 ,
     5  12989 =
     7   4487 ]
     8   4487 [
     9   2426 +
    10   1434 ==
    11   1044 -
    12    699 %
    13    488 `
    14    481 *
    15    342 !=
    17    305 <
    18    280 }
    19    280 {
    20    256 >
    21    150 >=
    22    128 <=
    23    117 &
    24    100 /
    25     80 +=
    26     73 ;
    27     67 |
    28     34 >>
    29     29 **
    30     27 <<
    31     21 //
    35      9 ~
    [ cut here ]

I'll have to get a better / wider / more representative corpus of
Python sources before my dimly-remembered background in computational
linguistics can be used here, I guess -- these numbers are just too
small (and from a probably non-representative sample).  Hmmm, maybe
Parnassus...  It's all academic, of course, anyway.  But even though
we're looking at code so bit-oriented than it bitmasks (&) more than
it divides (or just a bit less, if you count both / and //), bitwise
xor is hardly popular in it -- _that_ much we can already state:-).


Alex



From phr-n2002a at nightsong.com  Fri Apr 26 02:00:36 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 23:00:36 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com> <3CC7889D.2080002@nowhere.org> <7x8z7cvusy.fsf@ruckus.brouhaha.com> <3CC8BCD1.7090804@nowhere.org>
Message-ID: <7xwuuvatiz.fsf@ruckus.brouhaha.com>

Bryan Olson <fakeaddress at nowhere.org> writes:
>  > I thought of adding a function that would allow encrypting in segments
>  > but don't currently have applications that need that.  Remember that
>  > the main goal of this code (and I didn't expect it to become such a
>  > complicated project) was to just have something in portable Python
>  > that could be used instead of the silly rotor module.
> 
> I'm surprised such a trivial goal interests you.  If this isn't worth
> doing well, then I think it's just another case of "hey look - I
> invented a cipher", and personally I respond to those with "O.K. throw
> it on the 'unused' pile."

Well, the original reason I wrote it was as part of a fancier crypto
module that would provide a bunch of additional functions:
  
  http://www.nightsong.com/phr/crypto.txt

so I figured anything fancier than simple string encryption/signing
would go there (that's just the docs-- your comments on them are
welcome too--I guess I can also upload the code, though I want to
reorganize it a little first).

However, I'm planning to write an AES module in C that would be less
kludgy (plus a lot faster) than this SHA thing.  So the fancy module
should be able to use AES in most situations.

> The reason I'm enthusiastic about this project is that it actually meets
> a need.  Due to silly politics, a variety of programming environments
> ship with SHA-1 and without a decent cipher.  The value here is in
> creating a strong and efficient cipher without stepping outside the
> native programming languages.

Yes, that was my motivation too, both for writing it and for putting
some effort into making it run fast.

I think the politics though, at least in the US, have settled down to
the point where exporting AES shouldn't be a big problem, so a
standard AES module may be in Python's future.  That will obviate the
need for this thing if it happens.

> The definition of the cryptosystem is much more valuable than any
> particular implementation.  If you don't have a need to encrypt in
> segments while preserving state in-between, then fine don't implement
> it.  The definition of the system should still have that potential.
> 
> The reference implementation I offered doesn't offer a piece-by-piece
> facility either, but by separating the IV and MAC from the ciphertext,
> it defines what such a system should produce.

I think most users just want to pass in a string and get back a
string, rather than a tuple.  If the user wants to get the pieces
back separately, there could be a separate entry point for that.

>  > It's slow
>  > enough that it would be painful to encrypt large files with it.
> 
> Sure, but stand-alone (en/de)cryption a large file at one time is
> usually a mistake.  Even my reference implementation, on my 'budget'
> Celeron 1200, is fast enough to saturate my DSL line many times over.

Your "budget" Celeron 1200 is about 10x faster than the "high end" 200
mhz StrongARM PDA that I plan to run this thing on ;-).  The point
about large files is that's the situation where you need to encrypt in
segments.  For smaller files, it's ok to encrypt all at once, which
the function already does.  FWIW, you could probably saturate your DSL
line (1.4 mbit/sec) with RC4 written in Python, though just barely.

Anyway, I'm ok with adding more entry points if it doesn't confuse
non-cryptographer users too much or if there's user demand.  If I want
to confuse people though, I have the whole fancier library for that ;-).

> A Python-only cipher is of little, maybe negative, value.  A cipher that
> is very fast in C (or ASM) and moderately fast in a variety of scripting
> languages could be a serious player.  We should aim for the big-time.

I'm not sure how this SHA-OFB in C would compare to AES in C or asm.
FWIW, I'd tried using SHA-CTR but I found that in Python, SHA-OFB is
faster, so I switched.  So this design is very much steered by details
of the implementation, usually not a good sign in protocols that are
supposed to be long-lived.  I think of this scheme as more of a
short-term hack (famous last words...).

>  > And in an SSL-like protocol it still seems ok to encrypt-and-sign
>  > each message.  (You couldn't implement real SSL with this module
>  > because of the nonstandard cipher).
> 
> Well, I kind of disagree with the principle.  If this cipher is to be of
> real value, we should definitely not assume it will always be
> "nonstandard".  It's unlikely to be a FIPS, but it could be at least as
> standard as RC4.

I think RC4 has been popular because 1) it's already a standard (part
of SSL because of marketing stunts from a past era that wouldn't be so
easy to accomplish now); 2) it's extremely simple to implement even in
scripting languages; 3) it's extremely fast when coded in C or asm
(faster than AES).  SHA-OFB is none of the above.

> SSL/TLS is designed to support multiple ciphers.  Certainly support for
> AES is coming.  There's no reason this cipher can't be included.  I see
> little use for an "SSL-like" protocol.  Protocols are valuable if and
> only if they are widely implemented.

I guess SHA-OFB support could be integrated into something like
OpenSSL, but would anyone really want to write an interoperating SSL
peer completely in Python using SHA-OFB?  The same question would
arise for SSH, IPSEC, and so forth.  Do you really want to hair up
those implementations with more silly ciphers?


From gerhard at bigfoot.de  Wed Apr 10 19:42:36 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 10 Apr 2002 23:42:36 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83> <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de> <a91mdd$ck$1@panix1.panix.com>
Message-ID: <slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>

Aahz wrote in comp.lang.python:
> In article <slrnab7hn9.4tj.gerhard at lilith.my-fqdn.de>,
> Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>>
>>Why would anybody care if you aren't man/woman enough to post with
>>your real name?
> 
> Because many people, respectable people, *do* post to Usenet without
> using their "real name".

Yes, there are such people, but they are rare. For one of those, there
come a hundred trolls, 1337 hax0rz and their like.

> What the heck is a real name, anyway?

In western contries, firstname + lastname.

IMO using your real name is polite and whatever you say is not worth
being said if you can't put your name under it.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From pyth at devel.trillke.net  Mon Apr 29 20:27:40 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 02:27:40 +0200
Subject: Remove masked elements from an array
In-Reply-To: <c76ff6fc.0204291609.115c33d6@posting.google.com>
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <mailman.1020102370.5619.python-list@python.org> <c76ff6fc.0204291609.115c33d6@posting.google.com>
Message-ID: <20020430002740.GT16217@devel.trillke>

On Mon, Apr 29, 2002 at 05:09:57PM -0700, John Machin wrote:
> holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1020102370.5619.python-list at python.org>...
> > On Mon, Apr 29, 2002 at 01:26:25PM -0400, PoulsenL at capecon.com wrote:
> > > I have a Numeric array with missing elements from which I wish to calculate
> > > a histogram.  However, in order to do so I must first remove the 'None'
> > > values from the array.  I have a kludgy solution, but I was wondering if
> > > someone could point me in the right direction on what is the proper method
> > > of collapsing a masked array to just the non-masked values.
> > 
> > there are too solutions. choose yourself :-)
> > 
> > nums = [ 1,2,3,None,2,3]
> > 
> > #version1
> > def f(x): return x is not None
> > newnums = filter(f, nums)
> > 
> > #version2
> > newnums = [ x for x in nums if x is not None]
> > 
> > The first version uses filter. 
> > 'filter' calls 'f' for every item in 'nums' and puts
> > the result into the new list iff it has a 'true' value.
> > 
> > The second version uses list comprehension.
> > It basically does the same if you look closely
> > at it. 
> 
> *Don't* try this if zero is a valid value for your histogram:

i think you are on the wrong track.

the check in question is 'x is not None'. look at it.

> because it will rip out any "false" entries (including 0 or 0.0) that
> you have.

Numbers are always 'not None'. Even 0.000 or 0 or 1-1 or whatever.

>>> nums = [ 1,2,0,3,None,2,3]
>>> print filter(f,nums)
[1, 2, 0, 3, 2, 3]

regards,

    holger




From DavidA at ActiveState.com  Tue Apr  2 17:38:34 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Tue, 02 Apr 2002 14:38:34 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <gIpq8.44728$pT1.1266378@news1.tin.it>
Message-ID: <3CAA32EA.62C105B5@activestate.com>

Alex Martelli wrote:

> Seconded.  _I_ am going to steal it (if I get Laura's permission,
> and if the PEP is rejected) for the Nutshell, for sure...

I like it too.  Explaining Pythonic truth always seemed a little
convoluted, but Laura's description is right on the ball.

I'm not sure how it translates to other languages.  "Quelque chose" vs.
"Rien" in French, I suppose.  I'm not sure what "None" would be in
French.  FWIW, I'm not sure what it means in English.  That word always
struck me as the most non-native aspect of Python, but being a
non-native speaker myself, I don't know if I'm the weird one here.

--da



From fredrik at pythonware.com  Thu Apr 11 16:05:31 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 11 Apr 2002 20:05:31 GMT
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
References: <mailman.1018549044.28363.python-list@python.org>
Message-ID: <f0mt8.538$p56.132821@newsb.telia.net>

"Noah" wrote:
> I could have sworn that I have seen interactive programs
> such as telnet and ftp scripted from Python.
>
> I looked at the pty stuff, but I was going to implement that
> AFTER I got pipes working. Pipes seem to be the most portable
> and intuitive way to solve this problem.

I'd say the most portable way is to forget about running external
commands, and use ftplib and telnetlib instead.

</F>




From aahz at pythoncraft.com  Mon Apr 22 16:02:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 16:02:45 -0400
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <3CC45005.2E9C0F86@tundraware.com> <b8p8cu8f7u9lkbmpcrksjfcqp67nkoq3om@4ax.com>
Message-ID: <aa1q95$eb5$1@panix1.panix.com>

In article <b8p8cu8f7u9lkbmpcrksjfcqp67nkoq3om at 4ax.com>,
Gon?alo Rodrigues  <op73418 at mail.telepac.pt> wrote:
>
>Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
>would-be applications to the natural sciences. I think it was P. Erdos
>who remarked once that all natural science is conceivably finite - e.g.
>there would be a time when there would be nothing more to learn about
>the universe - but mathematics is literally infinite.

I'm not so sure about the latter assertion, and I'm *definitely* not
certain about the assertion that natural science is finite.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From mparaz at yahoo.com  Tue Apr  9 07:06:22 2002
From: mparaz at yahoo.com (Miguel A Paraz)
Date: 9 Apr 2002 04:06:22 -0700
Subject: Linking against python library causes core dumps
Message-ID: <c9f472a0.0204090306.7fb516ea@posting.google.com>

Hi,
I'm working on the Python interface to the FreeRADIUS AAA server
<http://www.freeradius.org>.

When linked against the Python library libpython2.2.a, and my code, it
fails like this:

#0  ptmalloc_init () at malloc.c:1713
#1  0x4026de64 in malloc_hook_ini (sz=21, caller=0x40270a34) at
malloc.c:1856
#2  0x40269f8d in __libc_malloc (bytes=21) at malloc.c:2798
#3  0x40270a34 in __strdup (s=0x810ab28 "/usr/local/etc/raddb") at
strdup.c:43
#4  0x08058b23 in main (argc=1, argv=0xbffffa0c) at radiusd.c:550
#5  0x40206507 in __libc_start_main (main=0x8058a9c <main>, argc=1, 
    ubp_av=0xbffffa0c, init=0x8056d20 <_init>, fini=0x81098e0 <_fini>,
    rtld_fini=0x4000dc14 <_dl_fini>, stack_end=0xbffff9fc)
    at ../sysdeps/generic/libc-start.c:129

Where radiusd.c:550 is simply:

    radius_dir = strdup(RADIUS_DIR);


At this point, no Python functions are even called yet.
If I remove libpython from linking, it starts up fine.
I'm stumped - I see no reason why simply linking a lib would mess up
the malloc() system.

Thanks in advance!


From jo at johnochiltree.uklinux.net  Tue Apr 30 05:32:47 2002
From: jo at johnochiltree.uklinux.net (JohnO)
Date: Tue, 30 Apr 2002 09:32:47 +0000
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com> <f0fd5987.0204300105.51529004@posting.google.com>
Message-ID: <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk>

Garry Taylor wrote:

> Chris <chris at cmb-enterprises.com> wrote in message
> news:<chris-4E0B1A.01281230042002 at corp.supernews.com>...
>> In article <df30afd4.0204292111.31e5914a at posting.google.com>,
>>  xiaotsing_79 at hotmail.com (David) wrote:
>> 
>> > I have a lot of experience in ANSI C/C++, I want to learn a new
>> > language and which it can help me in the work.
>> > I just want to clear which(python or perl) is more comfort for me?
>> 
>> In terms of how it looks, Perl might look more comfortable, but Python
>> is likely to function more similarly to C++, while C will have more in
>> common with Perl.
>> 
>> Learn both.
> 
> I use Perl at work day in day out, and use Python for my own personal
> stuff, I would say that Python is better for pretty much everything,
> particularly larger programs, where it's more structured syntax makes
> code easier to read. I quite like Perl for CGI scripting, and little
> 1-page scripts, but would never contemplate using it for a big
> project.
> 
> Learning both is an ideal solution, but personally I wish I started on
> Python far earlier than I did.
> 
> Garry


I don't know if anyone else has noticed this but I've found that operators 
and sys admins use perl and that developers tend towards python. Also sys 
admins rarely utilise indirection but developers do.

John


From marklists at mceahern.com  Mon Apr 15 19:52:14 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 15 Apr 2002 18:52:14 -0500
Subject: pre-PEP for optional 'pass'
In-Reply-To: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEDOKLAA.marklists@mceahern.com>

[phil hunt]
> I wish to make the pass statement optional. That is, in any Python
> program where there is a pass it can be left out and the meaning of
> the program is unchanged.
[...]
> Thoughts?

I imagine you'll need to provide working code that demonstrates what's
required to accomplish this proposed PEP--or at least some analysis of the
scope of the work and the potential for breakage.  It would be really
interesting to see you post something about that.

Having never really delved into the C source code for Python itself, I have
no real sense of the scope of the change you're proposing.  Would existing
code break?  Is there code out there that depends on the presence of pass?
PyChecker, for example?

This may be one of those things that is tremendously easier said than done.

For what it's worth, I personally don't mind having to specify pass.

Cheers,

// mark





From martin at v.loewis.de  Sat Apr 13 18:38:36 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 00:38:36 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com>
Message-ID: <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> Since Python does not have unsigned numbers, I think Python should
> have a (Java-like) '>>>' operator.  The '>>>' means
> 'logical-shift-right', where a zero is copied into the high bit.

Since the Python int and long int types are going to be merged, I
doubt that this will be added to the language.

One might question whether the bitwise operators should have been
there in the first place - functions can do the same things just as
well.

Regards,
Martin



From pyth at devel.trillke.net  Thu Apr 18 18:20:07 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 19 Apr 2002 00:20:07 +0200
Subject: PEP for an unrestricted __init__ ?!
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com>
Message-ID: <20020418222007.GC28515@devel.trillke>

On Thu, Apr 18, 2002 at 01:59:19PM -0700, Jimmy Retzlaff wrote:
> > holger krekel [mailto:pyth at devel.trillke.net] writes:
> > I'd would like to know if anybody thinks it's
> > worthwile to start a formal PEP
> ...
> > to bring full factory-pattern power to the __init__
> > constructor of a class.
>
> I believe that's exactly what __new__ is for in new-style classes. See
> PEP 285 for an example of __new__ usage for exactly this purpose (in
> the
> Specification section).

good pointer. It seems like PEP 253 is the original PEP
for the use described in PEP 285. Guido's time machine :-)

But __new__ or __init__ still restrict you to return instances
of their respective class? This does sound rather sane,
but i imagined some other uses for an unrestricted __init__ (or __new__)
such as making functions-objects a special case of classes:

   def func(args):
      "compute some stuff
      return result
and
   class func:
       def __init__(self, args):
          "compute some stuff"
          return result

start to look very similar. This opens up some interesting uses,
like letting the functions "constructor" decide if it has enough
arguments and
returning a function instance (with some prebound arguments)
otherwise. Currently functions are always singletons...

but i might be terribly on the wrong path ...

        holger




From lac at strakt.com  Fri Apr  5 09:18:12 2002
From: lac at strakt.com (Laura Creighton)
Date: Fri, 05 Apr 2002 16:18:12 +0200
Subject: Python-list digest, Vol 1 #10324 - 13 msgs 
In-Reply-To: Message from Jonathan Claggett <jonathan@claggetts.net> 
   of "." <1018015095.1486.141.camel@localhost.localdomain> 
References: <E16tGLp-0007ou-00@mail.python.org>  <1018015095.1486.141.camel@localhost.localdomain> 
Message-ID: <200204051418.g35EICrq000708@ratthing-b246.strakt.com>

> On Thu, 2002-04-04 at 18:08, python-list-request at python.org digested:
> 
> > >I would like to be able to define anonymous functions and classes by
> > >omitting the function or class name normally found after the def or
> > >class keyword. 
> > 
> > Feel free to write a PEP so that this abomination can be rejected once
> > and for all.  ;-)  (I don't think there have been any large threads on
> > this in 2002, but there certainly were at least two or three in 2001.)
> 
> I might create the PEP for posterity's sake. Should I dub it the
> 'sacrificial PEP'?

not _The_ -- I have beaten you to it.  See PEP 
http://www.python.org/peps/pep-0666.html  (ps, it's funny).

> 
> > (Yes, I can be quite certain it will be rejected, given Guido's regret
> > over including lambda in the first place.  Python has rightly included
> > many useful bits from functional programming languages, but Python is
> > also emphatically *NOT* a functional language itself.)
> 
> This is interesting news to me since I've been picking Python habits up
> from the latest edition of Programming Python and which uses lambda a
> fair amount. Specifically, I've been using them in my Tkinter programs
> to pass extra parameters to my event handlers. Is it more proper
> (Pythonic) to def the helper function first and pass the helper function
> name second?

Lots of Tkinter programmers, including me, would say no.   There is also
a third approach -- Gustavo Cordoba's : see  
http://groups.google.com/groups?q=comp.lang.python+Tkinter+callback&hl=en&scoring=d&selm=mailman.1017245242.12269.python-list%40python.org&rnum=4

(last month).

Laura Creighton





From bkc at Murkworks.com  Wed Apr 10 18:38:47 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Wed, 10 Apr 2002 18:38:47 -0400
Subject: type() new style class instance says "class", not "ObjectType"
Message-ID: <3cb4bd45$1_1@goliath.newsgroups.com>

Could someone explain this to me.

Lots of code does  type(x) in [InstanceType]

One might imagine that to support new style classes, this would be changed
to:

type(x) in [InstanceType,ObjectType]


But that doesn't work because type(new style instance) returns "Class"

>>> class x(object):
...  pass
...
>>> y = x()
>>> type(y)
<class '__main__.x'>

>>> type(x)
<type 'type'>

This seems the best way:

>>> isinstance(y,ObjectType)
1
>>>

I'm hoping someone can illucidate this for me.

--
Novell DeveloperNet Sysop #5

_




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From pdm at zamazal.org  Sat Apr 13 11:56:23 2002
From: pdm at zamazal.org (Milan Zamazal)
Date: 13 Apr 2002 17:56:23 +0200
Subject: emacs python mode completion
References: <mailman.1018619185.28393.python-list@python.org> <a97gda$8n5$2@peabody.colorado.edu>
Message-ID: <87elhj3848.fsf@blackbird.zamazal.org>

>>>>> "FP" == Fernando P?rez <fperez528 at yahoo.com> writes:

    FP> I wish still that there were true namespace-based completion,
    FP> but that would require some serious python-lisp hacking, as
    FP> emacs would have to import your objects and parse your python
    FP> code to know which attributes to complete.

    FP> Sounds like a nice project for a serious lisp hacker. Any takers
    FP> :)

Emacs needn't parse Python code, it can ask the running Python
interpreter about its objects.  The only problem why this feature is
difficult to implement right now is how to communicate with the Python
interactive interface within multi-line inputs.

If I only had the support in IPython ;-), I might try to do something
about the completion in Emacs.

Milan Zamazal

-- 
http://www.zamazal.org


From pinard at iro.umontreal.ca  Tue Apr 23 13:13:56 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 23 Apr 2002 13:13:56 -0400
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
In-Reply-To: <aa3dti$n0$1@news.nextra.cz>
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <oqy9feqqwb.fsf@titan.progiciels-bpi.ca>

[Petr Prikryl]

> [...] is there any preference of single or double quotes when writing
> strings?  [...]  Thanks for your opinions,

When I came to Python, I asked myself the same question, and discussed the
matter with this list as you do.  I hesitated for a while before adopting
a consistent style.  Here is the rule I finally gave myself.

If the string is text, or even sentence fragments meant to be read by
a human, I use double quotes.  If the string is anything else, a file
name, a character, the name of an identifier, etc., I use single quotes.
When in doubt, I ask myself: would this string be translated or not if the
program was later internationalised?  I agree with you that single quotes
are less noisy, yet the fact is that single quotes often appear in human
text (as apostrophes).

Some people tend to merely transport in Python habits coming from other
languages.  We need a very good reason for doing that, in my opinion, as
reasons which are good in some languages are pretty meaningless in Python.
For example, in Perl and shell, I use double quotes when I want to allow
interpolation, and single quotes when I want to inhibit it: so I could
translate this usage in using double quotes for formats and single quotes
for the rest, but it looks to me that this is poor usage of the available
difference.  C mandates double quotes for strings and single quotes for
characters, this distinction is not really useful nor worth in Python.
Best is to develop a style which much more fit to Python idiosyncrasies.

Other details.  For Python, I tend to use raw strings (the `r' prefix to
strings) consistently for regular expressions (especially with SPARK and
PLY in these days :-), even if there is no embedded backslashes.  I use
triple single quoted strings for grammar rules, and only in SPARK so far.
I limit triple double quoted strings for long strings, meant to be read
by a human, having embedded newlines; the first line is always written
with an escaped newline immediately after the initial triple double quote,
so the first line of text is flushed left, aligned with the others.

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




From feldspar at ix.netcom.com  Tue Apr 23 19:03:44 2002
From: feldspar at ix.netcom.com (Antaeus Feldspar)
Date: Tue, 23 Apr 2002 20:03:44 -0300
Subject: Art and all that Jazz:  are you ugly, I mean, attacking through outer powders
References: <comp.lang.python.d42583a5@agozgi.sc.us>
Message-ID: <3CC5E850.3080903@ix.netcom.com>

This Jessica Felix really is a moron, isn't she?

I mean, she steals newsgroup space for her commercial spam (the fact 
that she's using HipCrime tools to spam *shows* that she knows she's in 
the wrong) and expects that we'll *want* to do business with her after 
she's shown herself to be contemptuous of courtesy and law?

Jessica Felix wrote:

> Some durable cups are bizarre and other brave dusts are clean, but will Abu dream that?
> 
> "ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
> that incorporates the two passions of its owner, Jessica Felix -
> art and jazz.
> 
> Jessica is an artist whose medium is jewelry. She creates what she calls
> "spirited art," specializing in images from indigenous cultures around the world.
> 
> http://www.artandallthatjazz.com/index-2.shtml
> 
> 119A Plaza Street
> Healdsburg, CA, USA 95448
> 1-707-433-7900
> 
> Occasionally, Kenny never attempts until Frank smells the think cloud tamely.
> 
> They are sowing near sweet, under clean, with sick jugs.
> 
> 
> 
> 



From niemeyer at conectiva.com  Sun Apr  7 06:12:26 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Sun, 7 Apr 2002 07:12:26 -0300
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
In-Reply-To: <7xadshtrj1.fsf@ruckus.brouhaha.com>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <20020407071225.D1600@ibook.distro.conectiva>

[...]
> Most recent changes in Python are good (list comprehensions, x in dict,
> etc) but I'd rather that they appear only in major releases (2.0, 3.0, etc)
> which happen every 2 years or so, rather than in the minor releases
> that come out every few months.
> 
> A mature language is like a grown adult who needs to buy new clothes
> every few years.  Python is more like a rapidly growing child who
> needs bigger clothes every few months.  There's a tension between
[...]

You're not using any of those features (list comp, rich comparisions, x
in dict, weak references, function attributes, distutils, nested scopes,
new style classes, iterators, generators, etc), right!?  If so, you'd be
acting against your own principles, obligating users to use newer Python
versions, and then you'd be like an adult buying new clothes every few
months. ;-)

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From Henrik.Weber at sys.aok.de  Mon Apr  8 08:48:57 2002
From: Henrik.Weber at sys.aok.de (Henrik Weber)
Date: 8 Apr 2002 05:48:57 -0700
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>	<20020408120242.41456ace.wilk@flibuste.net>	<a8rqh2$9h5$03$1@news.t-online.com> <20020408123251.48eab8d6.wilk@flibuste.net>
Message-ID: <7e964d1d.0204080448.1c099671@posting.google.com>

> When i see boa constructor i'm a litle bit afraid... but it doesn't mean
> that it's a problem of wxPython. Also when i make some tests of wxPython
> with idle or pythonwin i have often big scratch (whitout python log). But
> i have never theses scratch if i run my script from the command line.
> 
> The opinion of an experienced user of wxPython would be welcome !

The problem with Pythonwin (and Idle?) has something to do with
conflicting event loops. Pythonwin has it's own event loop and a
wxPython application has one as well. They get into the way of each
other when you run the wxPython app from Pythonwin.

I'm not an experienced user of wxPython but you may find some on the
wxWindows mailing list.


From kseehof at neuralintegrator.com  Mon Apr  8 16:06:48 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Mon, 08 Apr 2002 13:06:48 -0700
Subject: Yeah whatever... (was RE: PEP 285: Adding a bool type)
In-Reply-To: <a8smam$png$1@nyheter.chalmers.se>
Message-ID: <EFEFIHAINIINNELLHFAIAEDACAAA.kseehof@neuralintegrator.com>

Jacob Hall?n wrote:
> In article <mailman.1018124969.17602.python-list at python.org>,
> Ken Seehof <kseehof at neuralintegrator.com> wrote:
> >> If we ignore the imaginary, which has a very specialized
> scope, we have 5
> >> base types. These can interact in approximately 5! (120)
> >> different ways in a
> >> statement. If we add one more base type we get 6! (720) different ways
> >> that our types can interact.
> >
> >In python the number of numeric interactions is essentially 3!, not 6!.
> >Most of these interactions are trivial and predictable, and are not
> >increased
> >by bool.  My brain hurts trying to figure out how to enumerate
> the 6! that
> >you suggest :-)  Perhaps in other languages your argument has merit.
> >
> >One thing that I agree with is that one would not be likely to
> add a bool to
> >a complex number.  :-)
>
> The difficulty does not come from the legitimate combinations. It
> comes from
> the erroneous ones. Understanding what a correct program does is trivial,
> compared to understanding why an incorrect one does not work.
> The possible number of ways of combining X types is the factorial of X.

It seems like were playing number games to me.  This factorial stuff is
mathematically correct, but rediculously abstract and meaningless.  To
have a complete understanding of booleans I simply do not have to learn
the difference between:

  foo_dict[foo_list[2.0+(x==5)+2L]] and
  foo_dict[foo_list[2.0+int(x==5)+2L]]

And then learn the diffence between:

  foo_dict[foo_list[2+2.0+(x==5)+2L]] and
  foo_dict[foo_list[2+2.0+int(x==5)+2L]]

And then ....

Can you list the rest of the 5*5! new real world interactions that result
from adding bool to python?

Sorry, I just don't get why this is relevant.  Let's try to keep this
discussion in the real world, ok?

There are two concepts that I need in order to understand booleans as
added in PEP 285:

0. My understand of the interaction between numbers in python extends
to include all interactions with boolean.

1. The values of True and False behave like 1 and 0 respectively.

That's two (count em), not 5*5! new things to learn.

> Each wrong combination requires the understanding of the error message and
> the reason why it is wrong. A seasoned programmer can classify the errors
> and handle a fair number of these classes. Newbies have a seriously
> difficult time. I would even go as far as to say that it is the most
> important factor controlling how much time a novice programmer needs
> in order to be comfortable with the language.

I have no idea what you are talking about.  There are no new kinds of
errors.
False works wherever 0 works and True works wherever 1 works.  This is even
true of interactions with lists and dictionaries.

I didn't even need to read to PEP to be fairly certain that my two
assertions are correct (though I read it anyway just to be sure).  The
only explanation for any other behaviour would have been that Guido bumped
his head a little too hard before implementing the PEP :-)

You keep talking about error messages.  Read the PEP, then give me one
example of an error message caused by using bool instead of int.  See?
There isn't one.

<fantasyland>
Hmm... perhaps you have a point... let's see...

>>> x = 1
>>> x[5]
Traceback (most recent call last):
  File "<input>", line 1, in ?
TypeError: unsubscriptable object

Ok, so now we add booleans and get an <sarcasm> entirely brand new
</sarcasm> error condition to learn:

>>> x = True
>>> x[5]
Traceback (most recent call last):
  File "<input>", line 1, in ?
TypeError: unsubscriptable object
</fantasyland>

Can we get back to the real world now?

- Ken Seehof






From gbreed at cix.compulink.co.uk  Fri Apr  5 06:33:46 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Fri, 5 Apr 2002 11:33:46 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <gzxZLgAd4Xr8Ewmq@jessikat.demon.co.uk>
Message-ID: <a8k22q$213$1@thorium.cix.co.uk>

Robin Becker wrote:

> heh heh, at last someone with a sense of humour. Of course now we can
> argue whether it's 1 or True that 'Guido' > 'God'
> 
> and at least no one mentioned Hitler. duh!

It's fine so long as you properly define the comparison.

>>> 'van Rossum' > 'Hitler'
1

and

>>> 'Guido' > 'Adolf'
1

so it doesn't really matter that

>>> 'Guido' > 'Hitler'
0


                       Graham


From aahz at pythoncraft.com  Sun Apr  7 10:52:48 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 10:52:48 -0400
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com> <just-16031B.10170807042002@news1.xs4all.nl>
Message-ID: <a8pmg0$9q2$1@panix1.panix.com>

In article <just-16031B.10170807042002 at news1.xs4all.nl>,
Just van Rossum  <just at xs4all.nl> wrote:
>In article <a8nric$sbb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
>wrote:
>>
>> The three execution scopes are function local (hereafter referred to as
>> local), module global (hereafter referred to as global), and builtin
>> scope.  Local scope exists any time Python's instruction pointer is
>> inside a function/method.  Global scope refers to the currently
>> executing module; explicitly accessing a function in another module
>> through attributes changes the current module:
>> 
>>     from M import f
>>     f()                 # f()'s global scope is the current module
>>     import M
>>     M.f()               # f()'s global scope is now the module M
>
>No: f()'s global scope is always module M. Scoping is static after all. 
>Or are you saying something different?

Okay, here's the new version:

The three execution scopes are function local (hereafter referred to as
local), module global (hereafter referred to as global), and builtin
scope.  Local scope exists any time Python's instruction pointer is
inside a function/method.  Global scope refers to the currently
executing module; functions defined in another module still consider
that module their global scope even if imported into the current module's
namespace:

    import M
    M.f()               # f()'s global scope is the module M
    from M import f
    f()                 # f()'s global scope is still M
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From greg at cosc.canterbury.ac.nz  Mon Apr 15 22:03:12 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 16 Apr 2002 14:03:12 +1200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <m3d6x1mqbu.fsf@mira.informatik.hu-berlin.de> <a9f4eh$pjp$0@216.39.172.122>
Message-ID: <3CBB8660.A4859BAB@cosc.canterbury.ac.nz>

Bengt Richter wrote:
> 
> newhex(+7) => 0x7 # implied extension of msb => 0x000...007
> newhex(-7) => 0x9 # implied extension of msb => 0xfff...ff9

I think that would be very confusing. It's pretty
subtle that the MSB of the binary equivalent of the
first hex digit tells you whether the original number
was positive or negative.

If we were to have something like this, I would rather
see

  newhex(+7) => 0x07
  newhex(-7) => 0xF9

i.e. the leading hex digit is always either 0 or F.

Then I would also want an extra arg to hex() for
specifying how many digits I want.

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


From ods at fep.ru  Fri Apr 19 13:09:49 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Fri, 19 Apr 2002 21:09:49 +0400 (MSD)
Subject: minor generator question
In-Reply-To: <20020419162905.GH28515@devel.trillke>
Message-ID: <Pine.LNX.4.44L.0204192105240.14030-100000@ods.pp.ru>

On Fri, 19 Apr 2002, holger krekel wrote:

hk> i have a recursive generator to flatten
hk> nested containers (lists and tuples):
hk>
hk> def flatten(*args):
hk>     for arg in args:
hk>             try:
hk>                 for i in arg:
hk> -->                 for l in flatten(i):
hk> -->                     yield l
hk>
hk>             except TypeError,e: yield arg
hk>
hk>
hk> Is there a better way to say that i want all results
hk> generated from a (sub-)generator be "forwarded" ?
hk> (see the two marked lines).

Related question: why we can write

[j for i in seq for j in func(i)]

but cannot

for i in seq for j in func(i):
    ...


If you don't like deep stair there is solution:

for j in [j for i in seq for j in func(i)]:
    ...

It looks a bit strange...





From davidccarson at hotmail.com  Fri Apr 26 16:22:51 2002
From: davidccarson at hotmail.com (David Carson)
Date: 26 Apr 2002 13:22:51 -0700
Subject: raw_input bug only in cygwin python
References: <12257ec4.0204181907.7a79b15b@posting.google.com> <kuavbukpa505nru9g6kfivo3bfrtirda05@4ax.com> <12257ec4.0204190610.44c9774@posting.google.com> <mailman.1019236547.15671.python-list@python.org>
Message-ID: <12257ec4.0204261222.70c81506@posting.google.com>

Which is it?  How do I carry on using raw_input()?

Richard Walkington <richard at stockcontrol.net> wrote in message news:<mailman.1019236547.15671.python-list at python.org>...
> >>I wouldn't call it a problem with cygwin, I'd call it "relying on
> >>unspecified behavior".  Ctrl-D is end of file; the ability to read past EOF
> >>is not promised anywhere in the docs.
> >>    
> >>
> >
> >
> >Ah, I see.  However, I've seen this technique recommended in several
> >docs/tutorials.  Also, the docs do promise raw_input() will raise
> >EOFError when EOF is read.  So, if it is true that I'm relying on
> >unspecified behavior, my question is: How do I make this code
> >reliable?  Is there a simple way to clear the EOF indication on stdin?
> > I hope there is a good way, because I like this kind of input loop
> >much better than the "type 'q' to quit" ilk.
> >
> 
> I don't see anything wrong with your code. I wouldn't have throught the 
> second raw_input should receive an EOF. I've also seen this technique 
> used in tutorials etc.
> 
>


From daniel.dittmar at sap.com  Mon Apr 22 11:51:04 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Mon, 22 Apr 2002 17:51:04 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org> <304d20df.0204220633.50bf23bf@posting.google.com>
Message-ID: <aa1bh8$frm$1@news1.wdf.sap-ag.de>

Allan Crooks wrote:
> The main problem is when I want to use 'yield', there's no real way of
> compensating for a language change... I suppose it is my fault for
> trying too much. :) The more I think about it, the more maintaining
> concurrent versions seems appealing... at least there is no 'hacking'
> involved...

You could do 'yield (expr, ...)', which looks like a yield to 2.2 and like a
routine call to anyone else.

There is still the problem how to declare
from __future__ import generators
It has to be the very first statement, so it can't be inside a try-except.
You might add a module __future__.py with a variable 'generators', but this
works only in 1.5 and 2.0, not in 2.1 (SyntaxError: future feature
generators is not defined).

Another possibility would be to have a custom importer who knows how to
parse top level if:
if PythonVersion >= 2.2:
    def generator ():
        ... yield (expr, ...) ...
else:
    class generator:
        ...

One difficulty is that you would have to cache the byte code for each Python
version:
mod.15.pyc
mod.20.pyc
...

Daniel





From gerhard at bigfoot.de  Mon Apr  1 12:56:50 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Apr 2002 17:56:50 GMT
Subject: Most important extensions?
References: <a8614l$dfa$1@panix1.panix.com> <a87gnj$e75$1@panix1.panix.com> <slrnaafsqi.8mf.gerhard@lilith.my-fqdn.de> <a89qsk$b65$1@panix1.panix.com>
Message-ID: <slrnaai7gn.bo.gerhard@lilith.my-fqdn.de>

In article <a89qsk$b65$1 at panix1.panix.com>, Aahz wrote:
> In article <slrnaafsqi.8mf.gerhard at lilith.my-fqdn.de>,
> Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>>
>>This post is the first one coming thru my own nntp-proxy :-)
> 
> Cool!  Written in Python?

Of course.

I've just hacked an additional backend for the Papercut nntp server recently
announced here. Most of a ForwardingProxy/HeaderRewringProxy is implemented
now. But my real aim is to write a (Header-)CachingProxy to replace the buggy
NNTPCache thing that I run currently.

Gerhard


From tim.one at comcast.net  Sat Apr 13 22:03:21 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 22:03:21 -0400
Subject: 2.1 vs. 2.2
In-Reply-To: <3cb8c690$0$19942$9b622d9e@news.freenet.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPJONAA.tim.one@comcast.net>

[Jens Baader]
> BTW, try this in interactive mode:
>
> "The Python interpreter is very stable"*30000
>
> At least at my system the interpreter promptly
> hangs itself.

Which version of Python, which OS, and exactly what do you mean by
"interactive mode"?  The code you showed isn't a challenge for any version
of Python I'm aware of (and Python & I go back more than a decade <wink>),
but if you're running Python under a GUI or windowing system of some kind,
the megabyte string Python tries to display may be giving *it* fits.  Try
this instead:

len("The Python interpreter is very stable"*30000)

If that doesn't display 1110000 in an eyeblink, then you may have a real
problem.

> I guess I just found my first bug..

Possibly, but the odds that it's a Python bug are small.





From phr-n2002a at nightsong.com  Mon Apr 29 19:31:10 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 29 Apr 2002 16:31:10 -0700
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <3CCC730A.FC60D13F@alcyone.com> <3ccd1ed9$0$15477$afc38c87@news.optusnet.com.au> <837kmqrvxl.fsf@panacea.canonical.org>
Message-ID: <7xit6aqdz5.fsf@ruckus.brouhaha.com>

Kragen Sitaker <kragen at pobox.com> writes:
> Unfortunately not.  You can't marshal lambdas either.

Oh well.  I'm surprised by this.  Maybe marshal should be extended.

> Or you can just use named functions, especially if you don't have
> closure data.

I thought a named function was just a variable whose value was a lambda:

  def square(x): return x*x

is the same as

  square = lambda x: x*x

Of course the syntax is different (you can't have control statements in
a lambda) but I'd have thought if you can't marshal a lambda, you also
can't marshal a named function.


From sholden at holdenweb.com  Mon Apr  1 17:44:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 17:44:50 -0500
Subject: Most important extensions?
References: <a8614l$dfa$1@panix1.panix.com>
Message-ID: <Bq5q8.118385$Ou.64260@atlpnn01.usenetserver.com>

"Aahz" <aahz at pythoncraft.com> wrote in message
news:a8614l$dfa$1 at panix1.panix.com...
> I'm teaching Python for [Perl] Programmers at OSCON this July.  I've
> already decided to mention mxODBC, NumPy, and PIL, plus PythonPoint (aka
> ReportLab) gets a mention on my title page.  I'd like this group to vote
> on the most important extensions for me to add to this list.  Please vote
> for no more than three.

gadfly - SQL in pure Python

sping - a very useful add-on for graphics

win32all - first encouraged me to use the Windows-specific interface.

Good luck at OSCON.

regards
 Steve






From max at alcyone.com  Tue Apr 23 15:52:20 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 23 Apr 2002 12:52:20 -0700
Subject: dictionary method question
References: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com>
Message-ID: <3CC5BB74.9070009@alcyone.com>

dsavitsk wrote:

> for some dictionary d is there any relationship between the returns from
> d.keys() and d.values()?  that is, to get them in some arbitrary but
> matching order, do i need to loop over d.items() or can i count on a
> similarity of the above? in small tests this seems to be the case.

I presume they would return values in the same object order, but I 
wouldn't count on it.  The .items method is designed expressly for that 
purpose anyway; you're better off using .items since it's potentially 
faster than zipping the two lists together yourself, and it's clearer 
anyway.
-- 
  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
\__ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I like young girls.  Their stories are shorter.
\__/ Thomas McGuane
/   Kepler's laws / http://www.alcyone.com/max/physics/kepler/
  A proof of Kepler's laws.



From tim.one at comcast.net  Sat Apr 13 23:42:47 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 23:42:47 -0400
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
In-Reply-To: <w5ct8.6353$vF6.199608@news2.tin.it>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEAAOOAA.tim.one@comcast.net>

[Alex Martelli]
> ...
> Now, what we need is a release 1.0.1 of this.py that, right below
> "explicit is better than implicit", adds a "but-clause" of (I suggest)
> "but superfluous redundancy and repetitive boilerplate aren't" (self-
> referential, of course:-).  Note this also helps ward against
> "if bool(x)==True:" &c...

Well, "The Zen of Python" is somebody else's name for this thing.  As
originally posted, they were the "20 Pythonic Theses".  They never need
revision, because I only gave 19 -- the 20th (or perhaps the first -- they
weren't numbered) was left open for Guido to fill in.  The admonition to
explicitness is indeed taken to extremes by those still struggling to
uncover their Pythonic Nature.

Raymond Smullyan wrote a great little story about a world where only a few
knew how to laugh.  Others strove to imitate the body movements and
breathing patterns of the laughers, and somberly committed entire Marx
Brothers movies to memory.  Having to be explicit about the limitations of
explicitness amounts to the same thing <wink>.





From root at [127.0.0.1]  Thu Apr 18 06:57:57 2002
From: root at [127.0.0.1] (nobody)
Date: Thu, 18 Apr 2002 20:57:57 +1000
Subject: MySQLdb: UPDATE borked?
Message-ID: <3cbea6b6$0$15472$afc38c87@news.optusnet.com.au>

When trying to UPDATE a blob field with more than 64k, I get this:

Traceback (most recent call last):
  File "e:\code\projects\bbs\tix\work\test.py", line 18, in ?
    main()
  File "e:\code\projects\bbs\tix\work\test.py", line 14, in main
    cursor.execute("UPDATE cache SET filedata=%s WHERE id=1", buff)
  File "C:\PROGRA~1\Python22\Lib\site-packages\MySQLdb\cursors.py", line 66,
in
execute
    r = self._query(query % self.__conn.literal(args))
  File "C:\PROGRA~1\Python22\Lib\site-packages\MySQLdb\cursors.py", line
168, in
 _query
    rowcount = self._BaseCursor__do_query(q)
  File "C:\PROGRA~1\Python22\Lib\site-packages\MySQLdb\cursors.py", line
118, in
 __do_query
    self._check_for_warnings()
  File "C:\PROGRA~1\Python22\Lib\site-packages\MySQLdb\cursors.py", line
150, in
 _check_for_warnings
    raise Warning, self._info
_mysql_exceptions.Warning: Rows matched: 1  Changed: 1  Warnings: 1

Updates under 64k, and any sized INSERT works just fine though (yes, I set
the max packet size to 16M for MySQL).

The same update also works just fine using the mysql client (dumped it to
file from the python script, and piped it straight in to mysql, no problem).






From gerhard at bigfoot.de  Wed Apr  3 09:25:24 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 3 Apr 2002 14:25:24 GMT
Subject: audio stream server by python?
References: <mailman.1017839888.2448.python-list@python.org>
Message-ID: <slrnaam45r.o0.gerhard@developer2.opus-gmbh.net>

In article <mailman.1017839888.2448.python-list at python.org>, Brian Lee wrote:
> Hi, is there any simple code to transfer audio stream data 
> to remote host by Python?

Doesn't a normal HTTP server with the correct MIME type declared do?

Btw. http://edna.sf.net/ streams mp3 files.

If the streaming server needs QoS, I'm pretty sure that there's little besides
the commercial servers, like the Real one.

Gerhard


From kevin at cazabon.com  Mon Apr  1 23:04:20 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 1 Apr 2002 20:04:20 -0800
Subject: PyArg_BuildValue, how to return a PyString object?
Message-ID: <5a4226f0.0204012004.24bb7b00@posting.google.com>

I'm writing a C extension dll for Python, and am having a bit of
trouble figuring out how to return a PyString_FromStringAndSize object
(along with some normal stuff).

(I need to use PyString_FromStringAndSize because the string may
contain null bytes, am I correct in this, or is there an easier way?)

Can I use something like this (it doesn't seem to work like this
though):

PyObject *myPyStringObject = PyString_FromStringAndSize(cstringarray,
size);
PyObject *myReturnObject =  Py_BuildValue("llo", long1, long2,
myPyStringObject);

Or do I have to build a tuple manually?

Or...?


Thanks!

Kevin Cazabon.


From syver-en+usenet at online.no  Tue Apr  9 11:44:20 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 09 Apr 2002 15:44:20 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net>
Message-ID: <upu18x4h3.fsf@online.no>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Sean 'Shaleh' Perry wrote:
> > " ".join(mylist) # is this perl or python?
> 
> *why* string objects have join method has been explained
> over and over and over again, and should also be obvious for
> anyone with some knowledge how strings and sequences work
> in Python.
> 
> > whoever decided this is the new style should be shot, a lot

Am I the only one in the entire world that thinks '\t'.join(mylist) is
elegant and natural?

-- 

Vennlig hilsen 

Syver Enstad


From pyth at devel.trillke.net  Tue Apr 30 15:38:02 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 21:38:02 +0200
Subject: import from a file in a subdirectory
In-Reply-To: <644f6688.0204301127.5d3372d@posting.google.com>
References: <644f6688.0204301127.5d3372d@posting.google.com>
Message-ID: <20020430193802.GF16217@devel.trillke>

On Tue, Apr 30, 2002 at 12:27:03PM -0700, rdack wrote:
> i have a script running in a directory.
> i want to call a subroutine in a file in a subdirectory.
> 
> in the first file i said  from subdir/ff.py import  subrtn
> python doesn't like the slash.
> 
> how do i do it?

import sys
sys.path.append('subdir')

import ff

--holger




From mongiaamit at yahoo.com  Mon Apr  1 01:56:47 2002
From: mongiaamit at yahoo.com (Amit Mongia)
Date: 31 Mar 2002 22:56:47 -0800
Subject: PyQt installation problems on windows
References: <20020322164756.46454.qmail@web11302.mail.yahoo.com> <mailman.1016821723.15034.python-list@python.org>
Message-ID: <cf43c01d.0203312256.405146a4@posting.google.com>

Thanks a ton Phil. I sent you a thank you email to your account also. 
Regards,
Amit Mongia.

Phil Thompson <phil at river-bank.demon.co.uk> wrote in message news:<mailman.1016821723.15034.python-list at python.org>...
> Amit Mongia wrote:
> > 
> > Hi,
> > I have installed Python 2.2 , on my computer. I
> > created a gui using Qt Designer and used pyuic to make
> > a python file out of it.
> > Now i want to run that file using the python
> > interpreter. Whenever i try doing that it gives an
> > error:
> > 
> > Traceback (most recent call last):
> >   File "WrapperInterface.py", line 10, in ?
> >     from qt import *
> >   File "F:\PYTHON22\lib\qt.py", line 44, in ?
> >     import libqtc
> > ImportError: DLL load failed: One of the library files
> > needed to run this applic
> > ation cannot be found.
> > 
> > I even tried all the py and pyc file from PyQt/lib
> > folder to Python22/lib folder and the same with the
> > Dll files. but it still gives the same error. Can
> > anyone offer a solution for this on windows.
> 
> What version of PyQt? What version of Qt?
> 
> Is libqtcmodule.pyd in F:\python22\dlls?
> 
> Is sip.dll (assuming you are not running a PyQt v3.1 pre-release
> version) installed in a directory that is on your path?
> 
> Phil


From phd at phd.pp.ru  Wed Apr 10 11:44:31 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 10 Apr 2002 19:44:31 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <j4sn63zht2.fsf@informatik.hu-berlin.de>; from loewis@informatik.hu-berlin.de on Wed, Apr 10, 2002 at 05:36:57PM +0200
References: <mailman.1018447222.24371.python-list@python.org> <j4wuvfzlbw.fsf@informatik.hu-berlin.de> <mailman.1018450821.8190.python-list@python.org> <j4sn63zht2.fsf@informatik.hu-berlin.de>
Message-ID: <20020410194431.I19899@phd.pp.ru>

On Wed, Apr 10, 2002 at 05:36:57PM +0200, Martin v. L?wis wrote:
> That is a bug: All API in a header must start with a Py prefix to
> avoid the kind of conflict you got.
> 
> Unfortunately, this probably cannot be fixed for 2.2; you'll have to
> work around it locally. It is unfortunate that this wasn't changed in
> the Grand Renaming - it sounds like rename3.h will be needed. (*)
> 
> You could also report this as a bug to Sun: 'destructor' is not an
> identifier reserved for the implementation, so they should not use it
> in their header. Indeed, Solaris 9 (and probably earlier) has
> 
> extern int pthread_key_create(pthread_key_t *, void (*)(void *));
> 
> Finally, you could report this as a GCC bug: they should fix the
> system header via fixincludes. This is also the work-around which I'd
> offer: copy pthread.h to .../gcc-lib/system/2.8.1/include/pthread.h,
> and adjust the copy accordingly.

   Ouh! The small Python bugfix release revealed THREE bugs in the OS,
copiler and the Python itself :)
   Thank you!

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




From paddy3118 at tiscali.co.uk  Fri Apr 19 03:00:21 2002
From: paddy3118 at tiscali.co.uk (Donald McCarthy)
Date: Fri, 19 Apr 2002 08:00:21 +0100
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <I6Fv8.36158$vF6.1093396@news2.tin.it> <mailman.1019162916.19664.python-list@python.org> <7cIv8.36928$vF6.1122050@news2.tin.it> <mailman.1019184878.1733.python-list@python.org>
Message-ID: <3cbfc027_2@mk-nntp-1.news.uk.worldonline.com>

Thanks Alex and thanks to all who discussed my problem. I am honoured to
participate in such a group.

The 'Alex solution' worked but it does lead on to another problem -
Reference material!

I bought 'Learning Python' by Mark Lutz, First Edition. Read it when I
bought it a couple of years back, but I usually get by with the online
documentation that comes with python.

With all the changes in Python since then I need a new reference work. Would
anyone recommend a  book that includes info on metaclasses?

Thanks again,

 Donald 'Paddy' McCarthy.





From aahz at pythoncraft.com  Thu Apr 11 21:21:14 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 21:21:14 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <wTit8.1381$F43.347@atlpnn01.usenetserver.com>
Message-ID: <a95cqa$adr$1@panix1.panix.com>

In article <wTit8.1381$F43.347 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>
>Personally I regard your frequent posts on this group as a significant
>contribution to "the community" (maybe because I'd like my own to be
>regarded in the same way), so I think you're being entirely too
>self-denigrating.

Besides, you never know when a bunch of Usenet posts might turn into a
book contract.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From loewis at informatik.hu-berlin.de  Thu Apr 18 07:40:36 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 13:40:36 +0200
Subject: HTTPS: memory leakage
References: <1918e8fd.0204180152.26cb929b@posting.google.com>
Message-ID: <j4sn5t6xqj.fsf@informatik.hu-berlin.de>

paolo at prosa.it (Paolo Andreetto) writes:

> Does it depend on the code or is there any problem in the system?

I recommend to run gc.collect() at each iteration. If you want to know
more details, you should activate gc.DEBUG_STATS before the loop.
After each collect invocation, make sure gc.garbage is empty.

If you find that gc does indeed collect objects, you can use
gc.DEBUG_SAVEALL to put the objects into gc.garbage instead of
collecting them.

If you find that memory consumption still goes up even though GC runs
often, and even though gc.garbage does not grow, you'll need more
powerful analysis tools.

Regards,
Martin



From BPettersen at NAREX.com  Mon Apr 22 12:52:31 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 22 Apr 2002 10:52:31 -0600
Subject: constructors in python
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158DF8@admin56.narex.com>

> From: Gustavo Cordova [mailto:gcordova at hebmex.com] 
[snip]
> 
> >
> >C++ doesn't permit not to call the base c-or.
> > 
> 
> And thanks to all the Gods and Goddesses for that.

But it does allow you to call any of the Base's ctors, which is an
excellent way to introduce bugs <wink>

-- bjorn




From sholden at holdenweb.com  Wed Apr 17 09:12:57 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 09:12:57 -0400
Subject: Simulating multi-dim array problem - OR - reference confusions
References: <3CBC991C.7090202@nyu.edu> <mailman.1018999655.9328.python-list@python.org>
Message-ID: <rzev8.12773$hl2.1966@atlpnn01.usenetserver.com>

"holger krekel" <pyth at devel.trillke.net> wrote ...
[In reply to "Shane Hoversten" <srh232 at nyu.edu> ] ...
> > I'm doing code that needs the functionality of multi-dimensional
> > arrays. I'm getting around this right now by using lists of lists, which
> >  seems to work right.  For instance:
> >
> > >>> a = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
> > ...
> > b = a[0]
> >
> > That b is actually a reference to the appropriate list in a, which
> > is just what I need.
>
> actually speaking in python semantics
>
> b=a[0]
>
> means that the "name 'b' is bound to 'a[0]'". Internally
> the reference count of the object a[0] is incremented.
>
Well, newbies might avoid mis-perceptions by reading the Python as "name 'b'
is bound to the object currently bound to 'a[0]'", to avoid any belief that
anything in a changes as a result of the assignment. In strict terms, a[0]
doesn't *have* a reference count, as far as I can tell without reading the
interpreter code, though both a and the object bound to a[0] will have a
reference count.

> the difference is that after you change b with e.g.
>
>   b = ['bert']
>
> a[0] will not have changed. You just changed the binding
> of the name 'b' again -- not touching 'a' in any way
> (except that the reference count of a[0] is decremented)!
>
Again, it's important to realise that *nothing* in 'a' changes as a result
of this rebinding, though the reference count of the object referenced by
a[0] will have been reduced by one.

> > Here's are my questions:
> >
> > 1) Am I going to get into trouble assuming (using the example above)
> > that b is basically a reference to the sublist a[0]?  I'm not clear how
> > sharing works in python.
[ ... ]
Well, since that's exactly what b *is*, you'll only get into the trouble you
make for yourself. You seem to have a pretty clear understanding of Python's
binding model.
>
> > 2) Is there a way I can print out the address of a "reference"?  In
> > particular I want some concise symbolic representation - like a memory
> > address, or a handle id, or SOMETHING - and not a verbatim
> > reconstruction of the list structure which is how python's "print"
> > function does it.  If I have elements containing references to elements
> > containing references to elements it's impossible to follow what's going
> > on with the chain.  I'd like to have it print something like:
> >
> > [[x, y, z, 0x123123], [a, b, c, 0x123958], [q, w, e, 0x358481]]
>
[ ... ]
The id() finction gives a unique value for each distinct object in Python.
Rumour has it that in the current C implementation of Python the value id()
returns is not unrelated to the object's address in memory, although I don't
believe this is guaranteed by the documentation.

I understand that the "is" test actually tests for equality of id() values
(since two references to the same object must give the same id() value).

However, you may have to get involved in type-checking to determine whether
to print out the value of a reference or its id() value. This may oir may
not ultimately be a good thing.

Finally, I'm not sure how you would get Python to print a list of lists in
the form you originally mentioned, since it would normally dereference the
names a, b, c, etc., and simply print the values they referenced. You might
therefore get either

[[1, 2, 3, 0x123123], [7, 8, 9, 0x123958], [12, 13, 14, 0x358481]]

or possibly

[['x', 'y', 'z', 0x123123], ['a', 'b', 'c', 0x123958], ['q', 'w', 'e',
0x358481]]

but not, without some programming effort on your part,

[[x, y, z, 0x123123], [a, b, c, 0x123958], [q, w, e, 0x358481]]

regards
 Steve







From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 22:45:00 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 19:45:00 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5 <3cbb7e33@news.mhogaming.com> <a9fvvl$nr4$1@watserv3.uwaterloo.ca>
Message-ID: <3cbb907d@news.mhogaming.com>

"Joe Mason" <jcmason at student.math.uwaterloo.ca> wrote in message
news:a9fvvl$nr4$1 at watserv3.uwaterloo.ca...
| In article <3cbb7e33 at news.mhogaming.com>,
| Ken Peek <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
| >Quite frankly, I don't give a damn HOW the long (or int) is represented
| >internally to the machine-- I simply ALWAYS want the same to interface to an
| >object that behaves as if it is a twos-complement integer.  When I left shift
| >it, I don't want the damn SIGN to change-- when I right shift it, I want to
be
| >able to CONTROL the sign bit that is shifted in, when I display the integer
in
| >decimal, I want a minus sign prepended to the string if the number is
negative,
| >and when I print out the number in hexadecimal format, I DON'T want "sign and
| >magnitude"-- I simply want a string of hex digits with a '0x' prepended to
them,
| >representing the twos-complement form of the number, and I want 2*N hex
digits
| >EVERY TIME, where N is the current size of the integer (in bytes.)
|
| As I see it, the problem here is an interface mismatch.  The fact that
| integers are represented as twos-complement numbers is, most of the time,
| an implementation detail.  As you point out, you don't care if it actually
| *is* a bitfield, you just want the interface of one.  This is an interface
| that should definitely be available, but not necessarily for every integer:
| most of the time, it's not important at all.
|
| So I'd argue that the ability to behave as a twos-complement bitfield is
| not something a standard integer should support, and that what you need is
| a bitfield object (which a constructor that takes an integer).  This could
| be a new language construct or completely encapsulated in a module.  (I'd
| lean towards a module, but I think that would mean that every bitwise logical
| operator would become a method or function, and I can see a case being made
| that keeping the operators available is important enough to be part of the
| language.)
|
| Also, to be completely clean, any operators that currently have no use
| except to treat integers as bitfields should be made to work only on
| bitfield objects instead.
|
|   x = bitfield(i)
|   y = bitfield(j)
|   z = x << 8 & y
|
| or
|
|   x = bitfield(i)
|   y = bitfield(j)
|   z = x.shiftleft(8).bitand(y)
|
| All the arguments about whether doing things with sign extension is
| violating the transparency of integers (in an 'it should just work' sense)
| now become invalid (or at least are confined to the constructor and any
| methods to convert back to an int).  The bitfield contracts to work just
| like a string of bits.
|
| Obviously this is an unrealistic proposal: removing bitwise logical
| operators isn't going to happen.  But I'm still interested, in case I
| ever design a Python-like language (heh): from your point of view, would
| this be acceptable or desirable?  Or would it be too much of a pain in the
| ass?  (I don't work with bitfields much, but I can imagine that if you
| commonly have to keep swapping back and forth between the two ways of
| conceiving of a number, this would be a bad idea.)
|
| Joe

This would be too much of a pain in the ass (for me at least.)  See my later
post, where I give examples.





From pyth at devel.trillke.net  Mon Apr 29 08:15:12 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 14:15:12 +0200
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
In-Reply-To: <aaiu6e$pfr$1@bob.news.rcn.net>
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <20020429121512.GK16217@devel.trillke>

> Rationale
> 
>     There are separate motivations for each function.  One motivation in
>     common is that the number of total built-ins should be as small as
>     possible to limit the amount of "core language" one needs to know in
> order
>     to read code that is not module specific.

This won't change the fact that some people really like to work with 
filter, lambda and friends. They will probably include the import
statement for their examples. 

>     The functionals (map, filter, and reduce) had greater importance prior
> to
>     the introduction of list comprehensions which are now the preferred (and
>     more readable) approach.  

Is 

    map(func, list)

really less readable than

    [ func(x) for x in list]

? If we voted on this are you sure 50% would say that the
list-comprehension is more readable and more easily understandable?
My take:

- map expresses without noise that func is applied to the list elements
- map uses no special syntax (like the 'for' loop without ':')
- map is shorter

Btw, what is the best/shortest way to express 'filter' and 'reduce' with 
list comprehensions?

regards,

    holger




From fredrik at pythonware.com  Tue Apr 30 10:27:52 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 14:27:52 GMT
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org> <200204292311.57363.bhan@andrew.cmu.edu>
Message-ID: <IRxz8.34779$n4.7373208@newsc.telia.net>

Benjamin Han wrote:

> This surely looks nicer. I'm just wondering how far Python has fared in terms
> of this concept of OO? This example seems to show a very different character
> of OO in Python, in that __str__ is not automatically called, which seems
> rather counter-intuitive.

join expects a homogenous sequence of string objects.

if you have something else, it's up to you to make it match
the expectations.  in this case, map(str, object) is probably
the best way to do that.

> How much is an effort in converting Python into C++?

none of it, I hope ;-)

to write pythonic python, think templates (interfaces), not
subclasses.

</F>




From dale at riverhall.NOTHANKS.co.uk  Tue Apr  2 04:26:08 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 02 Apr 2002 10:26:08 +0100
Subject: CD Device Name
References: <1017720021.720264@cheroke.optimation.co.nz>
Message-ID: <07uiaucl52krro7h3hr9p1q4pbmpedo4ap@4ax.com>

"Russell" <Russell.Every at MentumGroup.com> wrote:

>Anyone ideas (prefereably Python code) on how to get the label of a CD under
>Windows 2000?
>

You need Mark Hammond's Win32 extensions. Then:

from win32api import GetVolumeInformation

label = GetVolumeInformation("E:\\")[0]


--
Dale Strickland-Clark
Riverhall Systems Ltd


From sholden at holdenweb.com  Tue Apr 16 15:52:12 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 15:52:12 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com>
Message-ID: <Ij%u8.517$hl2.136@atlpnn01.usenetserver.com>

<maxx at easynerws.com> wrote ...
> On Sun, 14 Apr 2002 19:36:27 +0100, philh at comuno.freeserve.co.uk (phil
hunt)
> wrote:
>
> >IMO until Python developers adopt a policy of being very reluctant
> >to do this, many IT managers will be wary of using it in anything
> >other thsan toy projects, and Python's user base will not grow as
> >quickly as (IMO) it should.
>
> Why?
>
> At my company we are stuck in the realm of Windows, and the IT staff must
> constantly patch the webservers to keep up with security and bugfixes. A
quick
> check of the IT media will show management's displeasure over this
continual
> process, but it does not appear to have hampered the deployment of Windows
> servers.
>
However, this is merely a demonstration of what I think of as Erlichman's
Principle: "When you've got them by the balls, their hearts and minds will
surely follow".

> As a matter of fact, what seems to be more reliable is that IT managers
will
> deploy more of what they already have, what their developers are most
> productive, or what their required component requires. If change rates
were a
> critical factor, no one would be deploying Microsoft's IIS.
>
I think it's more a matter of justifying their decision to back Microsoft
technologies in the first place, or simply knowing no better. Microsoft are
dominating because they have the same mindshare that IBM did in the 1980s,
and if they aren't careful they will go the same way.

Although I can't disagree that such market dominance does tend to ensure
that productivity aids are plentiful for the Microsoft environment. There
are signs that this won't last, though, as the open source/free software
movement(s) climbs up from the infrastructure towards the applications.

> As mentioned in elsewhere this thread, Java and VB, have and are
undergoing
> complete architecture restructuring, and they still are expanding their
> territories.
>
> Keep in mind, IT managers often do not think and act like developers.

Too true. But then, if developers had had the mandatory lobotomy they'd be
no good as developers anyway, would they?

i'd-rather-have-a-bottle-in-front-of-me-than-a-frontal-lobotomy-ly y'rs  -
steve






From greg at cosc.canterbury.ac.nz  Fri Apr 26 00:31:40 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 26 Apr 2002 16:31:40 +1200
Subject: ANN: Pyrex 0.1.2
Message-ID: <mailman.1020103159.18247.clpa-moderators@python.org>

I've fixed a few more bugs and made a new version
of Pyrex available:

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

Synopsis:

Pyrex is a new language for writing Python extension
modules. It lets you write code that freely mixes
operations on C and Python data, with all Python
reference counting and error checking handled
automatically.

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




From jb at cascade-sys.com  Mon Apr 29 18:11:25 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 15:11:25 -0700
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net>
Message-ID: <3CCDC50D.61DC661@cascade-sys.com>

Vincent Foley wrote:

> What's the Python equivalient of Perl's "a"++; ?

You can't get there from here.

I.e. there is no practical equivalant.

Including it would allow deadly side-effects within an expression and
thus would percipitate the end of the world, or at least mankind.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From steffen at molli15.org  Wed Apr 24 07:06:37 2002
From: steffen at molli15.org (Steffen Kirschke)
Date: Wed, 24 Apr 2002 13:06:37 +0200
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <slrnacag2g.1u3.gerhard@lilith.my-fqdn.de>
Message-ID: <aa63ju$nva$1@news.ilk.de>

Gerhard H?ring wrote:

> Fernando Rodr?guez wrote in comp.lang.python:
>> 
>> I'm using CompatMysqldb from http://sourceforge.net/projects/mysql-python
>> with Python 2.2
>> 
>> Whenever I try to insert some data into a mysql db, I get this error:
>> 
>>>>> curs.execute("insert into customer_info (email, url, referrer,
>>>>> first_visit) values ( '%s', '%s', '%s', '%s')",
>>>>> [('billg at microsoft.com', 'www.easyjob.net', 'www.google.com',
>>>>> '2001-03-07 12:32:00')])
> AFAIC, you should be using a tuple as second parameter to execute, not
> a list containing a tuple.

and also use a '%' instead of the ','.

curs.execute("""insert into table (col1, col2,...) values ('%s', 
        '%s',...)""" % (val1, val2, ...)

Steffen
> Gerhard



From guido at python.org  Fri Apr 26 22:05:20 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 27 Apr 2002 02:05:20 GMT
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <3CCA07D3.7A767866@python.org>

Philipp Lenssen wrote:
> Is there a simple, open source text-editor project available in Python?

How about IDLE?  www.python.org/idle/ -- also comes with every Python
distribution.

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


From ale_xp at yahoo.com  Sun Apr  7 09:23:00 2002
From: ale_xp at yahoo.com (alex)
Date: 7 Apr 2002 06:23:00 -0700
Subject: help:problem using the RawPen class/turtle module.
References: <8374ee10.0204060303.3333260a@posting.google.com> <slrnaauvbv.3v.missive@whave.frontiernet.net> <8374ee10.0204070114.56260848@posting.google.com>
Message-ID: <8374ee10.0204070523.3bbf801b@posting.google.com>

> 
> Hi, 
> thank you for your reply,
> 
> I tryied to type your code on my python command line
> and unfortunatly I have the same message.
> So well yes I think you are right, something's wrong
> in my configuration.
> 
> It seems that it comes from the turtle.py module, I'll
> try to have a look at it(even if I'm not an expert of
> python modules), 
> it seems strange to me that it doesn't work because 
> it is in the standard distribution and of course I 
> haven't changed it.
> 
> I made a copy & paste of my little test.
> 
> 
> Python 2.2 (#1, Mar  9 2002, 18:21:48) 
> [GCC 2.95.2 19991024 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import turtle
> >>> import Tkinter
> >>> root=Tkinter.Tk()
> >>> c=Tkinter.Canvas(root)
> >>> c.pack(expand=1)
> >>> mypen=turtle.RawPen(c)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 16, in __init__
>     self.reset()
>   File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 42, in reset
>     self.clear()
>   File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 53, in clear
>     self._draw_turtle()
>   File "/usr/local/lib/python2.2/lib-tk/turtle.py", line 262, in _draw_turtle
>     self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
> AttributeError: 'NoneType' object has no attribute 'create_line'
> 
> 
> alex.

I think I've found the problem,
it seems that the call to 

...
self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
                                  width=self._width,
                                  arrow="last",
                                  capstyle="round",
                                  fill=self._color)

...

in the turtle.py ( _draw_turtle() method)
doesn't work because the _canvas to refer to is
the one in the 'self' instance,  
so with the following change:

self._arrow = self._canvas.create_line(x-dx,y+dy,x,y,....
              ^^^^^

everything is ok.

Without it the _canvas refers to a global that it
isn't been initialized to a Canvas when there is a
call to the RawPen class.

ale><


From philh at comuno.freeserve.co.uk  Sun Apr 14 14:25:56 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 19:25:56 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org> <7xr8lic2yb.fsf@ruckus.brouhaha.com> <a9b9a9$iam$0@216.39.172.122>
Message-ID: <slrnabjidk.741.philh@comuno.freeserve.co.uk>

On 14 Apr 2002 06:58:17 GMT, Bengt Richter <bokr at oz.net> wrote:
>>
>>In that particular debate, a lot of people were asking for certain new
>>features to be LEFT OUT of Python.  I don't understand how someone can
>>volunteer to contribute work, time, or money toward leaving something out.
>
>Maintaining older baseline releases, where the features are already left out?

The problem here is that then there multiple "current" versions of 
Python.

IMO there is no need for multiple versions. Just 2 would do for me, 
a stable branch and an unstable branch, like Linux uses.

Alkso, I have no problem with changes that DO NOT BREAK EXISTING 
CODE.

>I don't think that work is zero, so should the BPWIs[1] be doing that, and who should say?
>
>The trick is what happens when various versions are adopted out in the real world
>without asking those whose scripts are expected to continue to run.
>Can one distribute a script with first line #!/usr/bin/python21 and expect the
>link to be there wherever python21 (or fully compatible other version) is available?
>
>Otherwise new versions will potentially make extra work for me, which I'd rather
>volunteer in other ways ;-)
>
>Apparently 1.5.2 and 2.1.x and 2.2.x are semi-official baseline versions. Is it/could it
>be official policy that #!/usr/bin/python(15|21|22) links should exist if a site
>has the identified version? 

IMO this is a good idea.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From johnroth at ameritech.net  Wed Apr  3 06:57:52 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:57:52 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org>
Message-ID: <ualr7hmlhkemab@news.supernews.com>

"David Ascher" <DavidA at ActiveState.com> wrote in message
news:mailman.1017799668.13959.python-list at python.org...
> Guido van Rossum wrote:
> >
> > Alex Martelli wrote:
> > > I _have_ seen old-time Fortranites rant quite convincingly against
that
> > > useless, newfangled LOGICAL stuff.
> >
> > And that is exactly what many of the responses in this thread
> > (including Laura's over-long FUD piece) sound like.
> >
> > I'm reading (at least skimming) everything, but the more I read the
> > reactions, the more I believe that this is just the response I have
to
> > expect for *every* change to the language I propose, and the more
I'm
> > tempted to follow my gut instincts.
>
> IIRC, there wasn't nearly this much heat in the discussion over the
> long/int unification or the type/class unification proposals, even
> though the latter especially was much more complicated, ambitious, and
> compatibility-challenging.

Well, long / int unification was pretty obviously healing a wart, and I
doubt if very many people understood the stakes in the type/class
unification proposal. I know I didn't, or I'd have very vocally
asked for it to be postponed to Python 3000!

In fact, at this point I'd like everything substantive to be defered
to Python 3000, and get started on that right away. I've got a
few suggestions...

> This reminds me of the old saying about academic politics: "the
smaller
> the stakes, the more vicious the fights".

I don't think that's the case here. This proposal opens up a whole
can of worms involving the concept of 'truth', that is, what the
conditional and logical operators see as 'true' or 'false'.

In current Python, truth is whatever an object says it is, using either
the __nonzero__ or the __len__ magic methods. Defining 0 to be
false runs into problems anywhere you have a function that indexes
from zero, and also can return 'not found.' It also makes no sense
for the return from the cmp() function, and I suspect that I could
find other examples if I looked.

Defining empty sequence objects to be false is sometimes what
you want, and sometimes not.

Consequently, I think of it (0 and  1 as false and true) as a
hack carried over from C and similar languages.

John Roth




From carole at autosim.no  Tue Apr  2 06:14:46 2002
From: carole at autosim.no (Carole Valentin)
Date: 2 Apr 2002 03:14:46 -0800
Subject: Swig/Python: data type problem
Message-ID: <9d8d2e93.0204020314.362621a4@posting.google.com>

Hello, 

I have 2 modules: Module 1 written in C++ (Visual C++ 5.0) and Module
2 in Python (Python 2.1).
I use Swig to make accessible the data from the Module 1 to the Module
2.
In the Module 1, I have a function which read a file: 

void read(String* namefile); 

In the Module 2, I build a window (with wxPython) with a popup list
which contains the names of the different files. When I choose a file
name:

# import the Module 1 project: 
from Module1 import * 
# create a new object:  
object = MyObject() 
# read the database file with the 
# parameter "namefile" = the name choosen just before: 
object.read(namefile) 

In Python, we don't declare the data so "namefile" doesn't have any
data type.
But in C++, we expect a data type: 

void MyObject::read(String* namefile) 

I tryed with 3 data types: char / string / string* 
Each time, I have the same error: 

File "c:\python21\Module1.py", line 172, in read 
    val = apply(Interfc.RoadNetwork_lecture,args) 
TypeError: Type error. Expected _p_String 

Do you have any suggestion? 
Thanks a lot, 
Carole.


From aleax at aleax.it  Wed Apr 10 10:29:54 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 14:29:54 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204100109.4a44932b@posting.google.com> <RYTs8.2219$b62.57325@news1.tin.it> <23891c90.0204100552.6b822307@posting.google.com>
Message-ID: <C%Xs8.3139$b62.77265@news1.tin.it>

Paul Boddie wrote:

> Alex Martelli <aleax at aleax.it> wrote in message
> news:<RYTs8.2219$b62.57325 at news1.tin.it>...
>> 
> 
> [Iterators]
> 
>> I've noticed they immensely help Python beginners that come to them
>> without preconcepts.  Such beginners are surprised to learn that
>> these are extremely recent additions to Python, so well they fit in.
> 
> Well, I was excited a few years ago when iterators appeared in C++, so
> I suppose I should be even more excited now. :-) However, it's
> arguably a bigger deal for the implementer of the iterator (or
> generator, even) than the "users" or "clients" of them. Many beginners
> just expect things to work in the way that iterators themselves
> promote, and it was previously up to the implementer of a framework to
> provide nice interfaces to "users"/"clients" - now, it's just easier
> for the implementer.

A beginner won't go around calling iter(), whatever.next(), and
try: ... / except StopIteration: ... , of course -- he or she will
just expect "for item in container:" to work smoothly and
seamlessly.

With the previous iterator protocol, "for item in somedict:" COULD
NOT "work smoothly and seamlessly".  Nor, at all.  Think about it.

Therefore, the previous iterator protocol did NOT allow "the
implementor" of such a fundamental container as a dictionary (or
many other a mapping type) to offer to beginners (and other users
of the container) this crucial tidbit of functionality in the
seamless manner provided by "for item in container:".

Now, it's not *JUST* "easier" -- it's finally become *POSSIBLE*
at all, where it earlier wasn't.  I guess somebody with sufficient
Jesuitic education COULD try arguing that if something that was
not possible and now is CAN be fairly said to be "easier" now,
but, fair warning: I may have had more years with the Jesuits than
you, I think that "just" gives me enough purchase to tear such
an attempted, hypothetical argument into reasonably tiny pieces,
and while that might even be fun (in the usual perverted way
semantics flamewars can be fun) I'm busy enough that I don't truly
look forwards to it (whence the fair warning:-).

In many, indeed most, cases, showing a nice, seamless iterable face to 
beginners and other users is now "just" easier; in many others, it's 
become possible where it previousty was impossible (you had at least
to give up the "seamless" part of it).

>> They're great progress over the previously murky "iteration protocol".
> 
> What protocol was that again? ;-)

for item in container:
    body(item)

translated, roughly, into:

_index = 0
while 1:
    try: item = container[_index]
    except IndexError: break
    _index += 1

    body(item)

Perhaps "murky" is not an ideal characterization of this, but it
does have the problem of requiring the container to implement
__getitem__ yielding all items progressively when called with
indices of 0, 1, ..., and raising IndexError when done -- fine
for sequences, not so fine at all for mappings (not to mention
a few auxiliary issues with sequenceoids that ONLY wanted to
allow iterations and not "random-access" indexing).

In contrast, the same statement now translates, roughly, into:

_iter = iter(container)
while 1:
    try: item = _iter.next()
    except StopIteration: break

    body(item)

The key advantage is not overloading indexing-on-container to
serve for iteration-on-container as well.  Built-in function
iter knows how to build an iterator-on-sequence object if need
be (also an iterator-to-sentinel when called with 2 args,
but that's another, albeit sort of nice, thing).


> [Applications which subclass built-in types]
> 
>> Or maybe not being used to them you just don't think of cases where
>> they'd be handy (if you need to keep compatibility with 2.1, then
>> of course you'd better NOT start thinking of such cases:-).  I.e.,
>> Sapir-Whorf hypothesis somewhere between its weakest and strongest
>> forms (though I still prefer the way Wittgenstein independently
>> put it earlier -- "The limits of my language are the limits of
>> my world").
> 
> It's probably best not to bring semi-controversial linguistic themes
> into the discussion - Whorf allegedly had less of a grasp of the
> languages he used to illustrate his hypothesis than I do of the most
> recent Python features. ;-)

The ones you're nevertheless busy putting down?-)  Maybe, but that
is part of why I prefer to use the Wittgenstein quote, independently
penned in 1918 -- Wittgenstein's grasp of German and English was
excellent, his Italian and his formal-mathematics not bad, and his
thesis is in any case not based on empirical issues but on a survey
of the analytical nature of language (Tractatus Logico-Philosophicus).

> Seriously, I can't think of cases off the top of my head where
> inheriting from an integer, floating point, complex number, string,
> tuple, list or dictionary are so compelling. An "ordered dictionary"
> can be useful, but we already had UserDict to define the interface for
> that kind of thing.

Inheriting from immutables is unusual -- but far from unheard of,
as shown by the existence of UserString too.  Inheriting from
mutables is more usual, as UserList and UserDict's existence and
popularity show -- and now we don't need them any longer, as we
can subclass list and dict directly, an obvious win.  Besides
directness (no need to teach and learn two separate types for
every interface, an inheritable and a fast one) and performance,
you stand a little bit more chance to survive type-testing (not
if the adversary type-tests with type(a)==type(b), but at least
if he uses isinstance instead).  You also get many more types
you can inherit from -- you didn't have a UserFile earlier for
example, so no direct and immediate tweaking of file objects.

Sure, to some extent you could kludge it up with delegation, but
only to some extent (there are several examples in the Cookbook,
you may want to have a look).  Delegation is often more powerful
and flexible than inheritance (and nobody is taking it away),
but inheritance, when what it offers is all you need, is handier,
faster and more direct -- often a win via "do the simplest thing
that can possibly work".  Sometimes (via the Template Method
design pattern, for example) it even does offer more power than
delegation (AFAIK no existing built-in type uses the TM DP, but
nobody stops you doing it in YOUR built-in types, now that you
know how to write them in order to be inheritable-from).


> [New features in the tutorial]
> 
>> Yes, just not in the tutorial (at least not until they're in the
>> "current" language rather than needing import from the future).
> 
> That's part of my original argument, though. I get the impression that
> some packages not only require Python 2.2 because they use new,
> standard features, but also so that the author can drag users further
> into the future by making use of not-yet-standard features, if only
> for their own entertainment.

Once you're using (e.g.) iterators in your package, or subclassing
a standard class, there's no extra cost in having the modules of
your package use (e.g.) generators or true-division as well, so
why shouldn't a package author freely do so?

That something MAY not be quite ready to explain to beginners is
one thing.  A package's author is presumably no beginner.


>> AMK's mission is not to tell you WHY you can be interested in
>> major stuff such as subclassing builtins, iterators, generators,
>> though -- just to ensure you DO know easily what IS new.
> 
> It wouldn't be a bad thing to have such a source of brief rationales
> for new features, however.

Since each new feature has its own PEP and each PEP its rationale,
I suspect such a "source" could be automated (more easily if PEPs
were in easily parsable XML, of course, but, even as is...).  Of
course, as rationales would come from a variety of authors, etc,
there is no guarantee such a compilation would be appealing or
very nice to read (also, some PEP authors are better at other
tasks than at writing appealing rationales).

Is it SO hard to go read the Rationale of each PEP yourself, btw?


Alex



From mtaylor at lowebroadway.com  Fri Apr 19 07:50:15 2002
From: mtaylor at lowebroadway.com (Moray Taylor)
Date: 19 Apr 2002 04:50:15 -0700
Subject: Is there a PyQT 'send'?
Message-ID: <d906b2c1.0204190350.2d5ae684@posting.google.com>

Hi,

Is there an equivalent command to Tk's 'send' in Python/QT?

Please reply to mtaylor at lowebroadway.com as well as the NG.

Thanks in advance for any help.

Moray Taylor


From aahz at pythoncraft.com  Sun Apr  7 13:43:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 13:43:29 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <a8pqal$ftm$1@panix1.panix.com> <7xpu1bh481.fsf@ruckus.brouhaha.com>
Message-ID: <a8q0g1$qu5$1@panix1.panix.com>

In article <7xpu1bh481.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2002a at nightsong.com> wrote:
>aahz at pythoncraft.com (Aahz) writes:
>>
>> I think it might be possible to make changes in the way Guido-introduced
>> PEPs get presented to reduce the amount of shouting from people like
>> Laura.  I hold no hopes that anything would get better results from
>> people like Arthur and Paul Rubin.
>
>I'm mildly opposed to the bool PEP but it's not that big a deal for
>me.  I'm more unhappy with the process that results in adding
>something like that to a minor version (2.3).  If it was saved for
>3.0, I think that would make more sense, for reasons I've explained.
>There are other features that I would LIKE to see added, like
>assignments in expressions; but those, too, shouldn't first appear
>in a minor release.

Assignments in expressions will never happen.

That said, seems to me that booleans are the stuff of a minor release,
unless you claim that minor releases should never contain language
changes.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From rjones at ekit-inc.com  Tue Apr 16 21:05:53 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 17 Apr 2002 11:05:53 +1000
Subject: File locking using fcntl.flock()
In-Reply-To: <a9hnje$a2k$1@nntp6.u.washington.edu>
References: <mailman.1018927830.29158.python-list@python.org> <a9hnje$a2k$1@nntp6.u.washington.edu>
Message-ID: <200204171105.53895.rjones@ekit-inc.com>

On Wed, 17 Apr 2002 03:38, Donn Cave wrote:
> Quoth Richard Jones <rjones at ekit-inc.com>:
> | I can't get any locking to work on Solaris 5.7 using python 2.1.3. On
> | Linux 2.4.9, I get an exception when I try to re-lock something that's
> | locked. On Solaris, I don't get the exception (and without LOCK_NB, I
> | don't block either). I've also tried the lockf and posixfile.lock methods
> | to no avail.
> |
> |>>> import fcntl
> |>>> f=open('foo', 'w')
> |>>> fcntl.flock(f.fileno(), fcntl.LOCK_EX)
> |>>> f2=open('foo', 'w')
> |>>> fcntl.flock(f2.fileno(), fcntl.LOCK_EX|fcntl.LOCK_NB)
>
> I don't have Solaris, but I bet it doesn't have flock.

It does, under the heading of "BSD compatibility".

That's almost a moot point though, because I also mentioned that I tried both 
fcntl.lockf() ("This is essentially a wrapper around the fcntl() locking 
calls.") and posixfile.lock() and neither of those worked either. The _real_ 
problem here is that Solaris appears to allow a process to have two locks on 
a single file in the same process. Things are fine if the lock is requested 
from separate processes.

Thus in separate processes fcntl.lockf() and friends work fine. This is 
troublesome because threads can't use file locks. Just gotta avoid threading 
:(


    Richard





From dsavitsk at e-coli.net  Mon Apr  8 00:19:04 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 08 Apr 2002 04:19:04 GMT
Subject: Newbie: Alarm at end of program
References: <es5HyN42BHA.1996@cpimsnntpa03>
Message-ID: <YS8s8.5401$d7.1425130@newssrv26.news.prodigy.com>

below are 2 solutions that showed up.  try google groups for getting old
postings.

-d


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:5aRq8.3571$d7.1106160 at newssrv26.news.prodigy.com...
> on win32 also try
>
> >>> import winsound
> >>> winsound.Beep(440,250)
>
> the first number is the frequency, the second is the duration.
>
> -d
>
>
> "Trond Gaarder" <trondga at -NOSPAM-ifi.uio.no> wrote in message
> news:a8gfr8$h4u$1 at maud.ifi.uio.no...
> > Don't know if this is the best/correct way of doing it in Python, but it
> > works for me (on a PC):
> >
> > print "\a"
> >
> > This should emit a beep from the pc. What it sounds like is highly
> > machine-dependent, however...
> >
> >
> > -Trond


"billludden" <billludden at email.msn.com> wrote in message
news:es5HyN42BHA.1996 at cpimsnntpa03...
> Hi,
> I have written a script that runs from a few minutes to several hours
before
> it finishes.  How would I make the computer play a 'beep' to signal that
it
> is finished?
> Thank you,  Bill
>
>




From tejarex at yahoo.com  Sat Apr 13 17:01:41 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 13 Apr 2002 21:01:41 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCOEOBONAA.tim.one@comcast.net> <mailman.1018723227.15055.python-list@python.org>
Message-ID: <V01u8.27065$043.2467943@bin6.nnrp.aus1.giganews.com>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
news:mailman.1018723227.15055.python-list at python.org...
> I was unaware that there was no guarantee for 0-length tuples always
> being unique objects.  There are lots of places in my code where I
use
> 0-length tuples in this pattern:
>
> NULL = ()
>
> class A:
> def b( self, object=NULL ):
> if object is NULL:
> #was not specified
> elif object is None:
> #specified to be object None
>
> Of course, I can use:
>
> NULL = []
>
> But I'd always avoided that because of bad associations with having
> mutable objects as default parameters (yes, it doesn't actually
matter
> in this pattern, but it's a thirty-times-burned-thirty-first-shy
thing).
>   Suppose I'll wind up using singleton instances instead.

Given that None is a possible input, so it cannot be used as the
no-input flag, you could also try something like NULL="OpTioNaL ArG
NoT GiVeN"

Terry





From mtvpro at usavideopro.com  Tue Apr 23 08:58:08 2002
From: mtvpro at usavideopro.com (MTVPRO)
Date: Tue, 23 Apr 2002 08:58:08 -0400
Subject: =?iso-8859-1?Q?30.000_fatalidades_em_acidentes_de_tr=E2nsito!?=
Message-ID: <200204231258.g3NCvH2Q029499@smtp-server6.tampabay.rr.com>

Prezados,
Conhecedor dos empreendimentos de sua empresa no Brasil, com o compromisso de 
influenciar de maneira positiva seus funcion?rios e a sociedade, desenvolvemos 
um programas em video para treinamento sobre Educa??o e Seguran?a no Tr?nsito 
com a assessoria t?cnica da ABPA - Associa??o Brasileira para Preven??o de Acidentes 
www.abpa.com.br

Com o seu apoio, poderemos inserir em nosso programa at? 1 minuto de imagens 
dos esfor?os da sua empresa sobre o assunto.
Por favor, caso n?o seja o seu departamento, encaminhe ao departamento que possa 
ser interessado em apoiar a nossa iniciativa.
Quantas vezes ao ano a sua empresa promove Educa??o no Tr?nsito?
? a preven??o ajudando a salvar vidas.
Grato por sua aten??o.
Manoel Torres
www.usavideopro.com





From fperez528 at yahoo.com  Tue Apr 16 13:43:09 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 16 Apr 2002 11:43:09 -0600
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu>
Message-ID: <a9hnt6$5t1$1@peabody.colorado.edu>

Russell E. Owen wrote:

> In article <slrnabmoci.dhk.philh at comuno.freeserve.co.uk>,
>  philh at comuno.freeserve.co.uk (phil hunt) wrote:
> 
>>OK, I've decided to write a PEP. Preliminary to this, I'd like
>>people's comments.
>>
>>I wish to make the pas startement optional. That is, in any Python
>>program where there is a pass it can be left out and the meaning of
>>the program is unchanged...
> 
> I'd go along with it and offer a third rationale:
> 
> The need for pass is a pain when commenting out entire blocks of code
> (e.g. while debugging). Instead of just commenting out the code, one
> must also add a "pass" statement, and then try to remember to take it
> out again when re-enabling the code.

??? You can simply wrap the block you want to comment out in """...""" and be 
done with it. It becomes a dummy string, no need for pass and it's out of the 
way.

> In my opinion an optional pass is a very good thing -- sometimes you
> want to be able to assert "this block intentionally left blank", but
> sometimes it makes more sense to just accept a truly empty block.

And then you have the following:
if this_is_true:
do_that()

Was that an indentation mistake? Or is the if... still unimplemented with an 
implied pass? At that point python is back to the joys of C's

if (this_is_true)
        do_that();
        and_that();  // For non-C-ers, this is NOT part of the if block :)

Remember? :) No thanks.


> Thanks for proposing this. I'm sure you'll get lots of horrified
> reactions (remember the fuss whenever someone suggests removing the
> required colon at the start of blocks? Forgetting that damn colon is
> still the most common mistake I make when coding Python, so I'd love to
> see it go, but it'll be a cold day in hell...)

Besides, you already saw Guido's unambiguous response, so this discussion is 
strictly of academic interest.

Best,

f.


From gimbo at ftech.net  Thu Apr 11 11:19:51 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 11 Apr 2002 16:19:51 +0100
Subject: PEP 282: A Logging System
In-Reply-To: <002e01c1e16b$b0cd9d60$652b6992@alpha>
References: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru> <002e01c1e16b$b0cd9d60$652b6992@alpha>
Message-ID: <20020411151951.GA7181@andy.tynant.ftech.net>

On Thu, Apr 11, 2002 at 04:15:14PM +0100, Vinay Sajip wrote:

> That's to say, I expect it will be thread-safe in the not-too-distant
> future, but I have not yet tested it in conditions such that I can

[snip]

Thanks for the clarification and reassurance Vinay - it sounds like
you're very much on the case!  In particular your comment that if the
hierarchy is set up in the main thread, children should be able to log
with no problems _now_ tells me I can go ahead and use it.  Nice.

I'd just like to say big thanks for the work you've done here: the
module's cool and I'm enjoying using it.  Wish I had time to help
develop it... :-/

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From gerhard at bigfoot.de  Mon Apr  1 00:35:27 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Apr 2002 05:35:27 GMT
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <mailman.1017628505.7732.python-list@python.org> <LURp8.63151$Q7.49997488@e3500-atl1.usenetserver.com>
Message-ID: <slrnaafotk.8j8.gerhard@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> True, beauty is at the behest of the be-renderer (ouch, that was bad).  But
> I'm still a junkie for drag and drop form drawing.  Regarding your tag line
> (this sig powered by Python).  Might want to add "so is Google and Inktomi".
> Frick, I was astounded when I read that in "Web Programming in Python".
> 
> I'm going to look at Open Office's and Mozilla's HTML editor.  If that
> doesn't work out then probably WebSphere.

If neither of these meets your requirements, and you have a decently
powered computer, you could try emulating Windows software like
Dreamweaver under Linux. Unfortunately, vmWare has gotten quite
expensive, but there's a much cheaper alternative, of which the name
currently escapes me, but I think it's in the EUR 100 price range.

Or just get a Mac with OS X instead, then you'll have a Unix and can
still run Dreamweaver and Photoshop ;-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From stephen at xemacs.org  Thu Apr 25 22:55:44 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 26 Apr 2002 11:55:44 +0900
Subject: Slight irritation with try/finally indentation
References: <mailman.1019657853.14965.python-list@python.org>
Message-ID: <87bsc7b233.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Fran?ois" == Fran?ois Pinard <pinard at iro.umontreal.ca> writes:

    Fran?ois> What I find irritating is that `cleanup()' is not
    Fran?ois> aligned anymore with `setup()', as it was originally, so
    Fran?ois> we loose on the legibility of the parenthetical idiom we
    Fran?ois> wanted to stress.  Surely, I may _not_ write:

    [...]
    try:
        setup()
        process()
    finally:
        cleanup()
    [...]

    Fran?ois> despite it would look nicer, because if `setup()' fails,
    Fran?ois> `cleanup()' gets called when it should not have been.

Of course you may not write that.  You must write

    try:
        setup()
        process()
    except SetupError:
        pass
    else:
        cleanup()

I think that you're trying to bind setup() more closely to process()
and cleanup() than your assumptions about failure modes allow.

Alternatively, to be truly perverse:

    try:
        setup()
        process()
    except ProcessSuccess:
        cleanup()

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


From fperez528 at yahoo.com  Tue Apr  9 14:33:25 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 09 Apr 2002 12:33:25 -0600
Subject: big ints & elliptic curves
References: <a8v338$4sl$1@watserv3.uwaterloo.ca>
Message-ID: <a8vc6u$kim$1@peabody.colorado.edu>

John wrote:

> Is there any good library for multi-precision arithmetics in Python? Also

google(python arbitrary precision)




From bergeston at yahoo.fr  Thu Apr  4 09:42:49 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Thu, 4 Apr 2002 16:42:49 +0200
Subject: Java to Python translation
References: <mailman.1017831968.9815.python-list@python.org>
Message-ID: <a8hop5$rq6ba$1@ID-135695.news.dfncis.de>

"Andrew McNamara" <andrewm at object-craft.com.au> wrote in message
news:mailman.1017831968.9815.python-list at python.org...
> >> > I need to translate a (rather voluminous 100MB) java program to
> >> > python. After a few searches on google, apparently the best tools
> >> > for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> >> > on the Lex-YACC HOWTO, I would be very interested in hearing of
> >> > others' successful experiences/feedback at this. Any pointers of
> >> > interest would similarily be appreciated.
> [...]
> >I think for a project like that, even a success would be disastrous.
> [...]
>
> Indeed.
>
> But I bet a complete re-write in python would less than 10MB, and take
> less than a tenth of the time of the original project... 8-)
>
> The suggestion of using Jython would be good one - subsystems can be
> rewritten and tested as time and resources allow.
>

To translate Java to Python (no idea if it is worthy in this case, but in a
more general perpective, this is a problem that makes sense IMHO) , I wonder
about this intermediate solution:
1. use a parser/generator to translate the Java code to Jython code (with
eventually some generated 'util' Java classes for the calls to the JDK
packages and other libraries)
2. translate manually the Jython code to Python code if is is really worthy.

My guess is that step 2 could be a lot simpler to achieve than the whole
translation and that step 1 can be done automatically (even if the
development of the generator requires some investment that could be shared
in an open source project) because :
- Java and Python are not so far in their concepts,
- Python is more dynamic than Java and I suppose that this makes it simpler
than the opposite (Jython)
- part of the effort could help anyway (classes to manage the java method
overridde in Python, ...)

I have no experience in such area, any comments are welcome.



> --
> Andrew McNamara, Senior Developer, Object Craft
> http://www.object-craft.com.au/
>




From ben at wblogan.net  Mon Apr  8 17:28:37 2002
From: ben at wblogan.net (Ben Logan)
Date: Mon, 8 Apr 2002 17:28:37 -0400
Subject: Reading/writing binary data.
In-Reply-To: <uQks8.20420$vm6.3149576@ruti.visi.com>; from grante@visi.com on Mon, Apr 08, 2002 at 05:55:38PM +0000
References: <mailman.1018285938.26500.python-list@python.org> <uQks8.20420$vm6.3149576@ruti.visi.com>
Message-ID: <20020408172837.A22281@wblogan.net>

Thanks, Grant and Jeff, for your help.  Both of those modules look
very promising and I never would have found them. :)

Regards,
Ben

-- 
Ben Logan: ben at wblogan dot net
OpenPGP Key KeyID: A1ADD1F0

A company is known by the men it keeps.




From uwe at rocksport.de  Mon Apr  1 10:55:44 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 1 Apr 2002 15:55:44 GMT
Subject: wxPython  / wxSingleChoiceDialog ...
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de> <slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <a89vu0$8bm2c$1@hades.rz.uni-sb.de>

Vadim Zeitlin <zeitlin at seth.lpthe.jussieu.fr> wrote:
| On 1 Apr 2002 13:54:42 GMT, Uwe Schmitt <uwe at rocksport.de> wrote:
|> I tried to manipulate wxSinglChoiceDialog so that the presented 
|> choices are printed with a fixed width font. I tried SetFont()
|> but nothing changed...

|  You should change the font of the listbox used by this dialog, not the dialog
| itself. Although there is no way to access this listbox directly (and unlikely
| to ever be), you can iterate over the dialog children until you find it.

I inspectetd the __dict__ attribute of my singlechoicedialog object and
tried dir() but found no hints how to iterate over the dialog children...

Greetings, Uwe.

|  Regards,
| VZ

-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From ianb at colorstudy.com  Wed Apr 17 16:51:50 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Apr 2002 15:51:50 -0500
Subject: preg_replace?
In-Reply-To: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE>
References: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <1019076711.1306.226.camel@localhost>

On Wed, 2002-04-17 at 15:09, Philipp Lenssen wrote:
> I am porting a program that I already wrote in VBScript and PHP.
> My current task: I need to replace everything inbetween square brackets with
> "0". In VBScript I did this "by hand" (looping, inString and so on). In PHP
> somebody helped me with the following solution:
> 
>     $s = preg_replace("%\[[^\]]*\]%", "0", $s);
> 
> Even though I've been reading through regex-tutorials in the meantime, I'm
> still not familiar with it. But is there a simple rule how to convert the
> above into Python code? Not only would it solve my current problem but I
> already got some helpful regex's for PHP and could reuse them in the future.
> 
> Anyway, the best (or worst) I did so far is:
>     pattern = re.compile(r'(\[[^\]]*\])')
>     pattern.sub('0', s)
> 
> I'm not convinced this remotely resembles how it's done in Python; it
> doesn't cause an error, doesn't do what I want either.
> 
> Here's my hoped before after:
>     s = "Hello [test] world [bla bla]..." # before regular expression
>     s = "Hello 0 world 0..." # after

That's what it does for me.  What's the problem?  

You don't need the parenthesis in your regex, but otherwise it looks
perfectly fine.  You could also do it as r'\[.*?\]', where the ? makes
the .* match as short a string as possible, as opposed to as long a
string as possible.  But these are stylistic -- your code is fine.

Maybe you wanted to do:

pattern = re.compile(r'(?<=\[)[^\]]+(?=\])')

The (?<=...) and (?=...) do look-behind and look-ahead matching.  You
can also do:

result = pattern.sub('[0]', s)

By using functions instead of strings for your substitution argument,
you can do far more advanced things, like substituting something into
the braces based on the text inside the braces.

  Ian






From jeremy at alum.mit.edu  Mon Apr  8 12:37:55 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 8 Apr 2002 09:37:55 -0700
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com>
Message-ID: <b0f083db.0204080837.625d270f@posting.google.com>

weeks at vitus.scs.agilent.com (Greg Weeks) wrote in message news:<1018044224.815568 at cswreg.cos.agilent.com>...
> One way to put it is this: Python 2.2 did not create nested scopes (they've
> always been there) or make nested scopes visible in function/method code.
> It simply *increased* the visibility of nested scopes beyond the original
> visibility of local+global.  How much is it increased?  Well, gee, I
> forget.  (I guess the language is getting too big for me.)

It enlarged the environment in which free variables were resolved from
the global scope to the scope of all enclosing function blocks.  I
hope that isn't too much to remember :-).
 
> Anyway, it increased visibility enough so that it is no longer possible to
> call a function and pass it the scope as two dictionaries, which I had 
> found handy on occasion.

I'm not sure I follow the details here, but I don't see how the change
in the scope rules would affect a function unless it specifically
exploited nested scopes.  If you wrote a function that only uses
locals and globals, can you not call it just the way you always did?

Jeremy


From mstern at aon.at  Thu Apr  4 03:37:19 2002
From: mstern at aon.at (Matthias Stern)
Date: Thu, 4 Apr 2002 10:37:19 +0200
Subject: Py_BuildValue() and C++ arrays
Message-ID: <3cac101d$0$25808$6e365a64@newsreader02.highway.telekom.at>

I have a C++ function which has to give back to PYTHON an array:

---
static PyObject * ex_readData(PyObject *self, PyObject *args)
{
     // define my data array
     float data[10000];
     :
     // call another C++ function which calculates some data
     err =  TeastAPP::readData(file, elConnName, channel, tmin, tmax, unit,
data, numElem);

     return Py_BuildValue( ??? );
}
---

I want to return the values "numElem" (int), "unit" (char[250]) and the
array "data" (float,10000).
What is the correct syntax for "Py_BuildValue( ??? )" ?

The first two parameters work e.g. this way:
---
     :
     Py_BuildValue("(si)", unit, numElem);
     :
---

But how can i send back the data array?!? I cant't write it like this:
---
     Py_BuildValue("(iiiiiiiii... (10.000 time)... iiiii)",  data);
---

Thx, Mac




From tripps81 at yahoo.com  Sat Apr 13 03:29:36 2002
From: tripps81 at yahoo.com (Tripp Scott)
Date: Sat, 13 Apr 2002 14:29:36 +0700
Subject: inspect.getargspec() for built-in functions
Message-ID: <5.1.0.14.2.20020413142745.00aca140@202.95.131.154>

is there any?

t





From tjreedy at udel.edu  Sun Apr 28 20:36:50 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 29 Apr 2002 00:36:50 GMT
Subject: Is there a text file somewhere i could download that contains all the words in the english language that i could read into python to create a spell checker?
References: <aai16e$m3$1@newsg1.svr.pol.co.uk>
Message-ID: <CA0z8.54606$fg.5260393@bin2.nnrp.aus1.giganews.com>

"G. Willoughby" <never at mind.info> wrote in message
news:aai16e$m3$1 at newsg1.svr.pol.co.uk...
> Is there a text file somewhere i could download that contains all
the words
> in the english language that i could read into python to create a
spell
> checker?

Try google or other search site if you have not yet.





From phr-n2002a at nightsong.com  Mon Apr 22 22:21:33 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 22 Apr 2002 19:21:33 -0700
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com>
Message-ID: <7xk7qz9mtu.fsf@ruckus.brouhaha.com>

h.e.w.frater at cs.cf.ac.uk (Hugh) writes:
>   I need a simple encryption algorithm that will not give me
> characters that I can't store in my oracle table. At the moment I'm
> using the rotor encryption module because I like it's simple interface
> but It's generating characters that oracle can't store which is a damn
> pain. What other modules are available that are easy to use and will
> do what I wan't. 

The simplest thing is take run the ciphertext through
binascii.hexlify.  That turn it into hex characters, which should
store ok in just about any database, but of course it doubles the
length of the text.

Note that the rotor module is dreadfully insecure and you should use
something better.

> Remember that security is not paramount here, I'm only including it
> for functionality.

That's not a valid reason to use bad cryptography.  If someone wants
encryption functionality, it's because they want to secure their data
against unauthorized inspection and they feel that the access controls
already provided by Oracle and the system it's running on are
insufficient.  Adding a layer of tissue paper encryption may create a
false sense of security for the user, while failing to stop a serious
attacker.  It's a bad idea.


From johnroth at ameritech.net  Mon Apr 22 14:17:42 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 22 Apr 2002 14:17:42 -0400
Subject: OT Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org> <pUWw8.48099$vm6.7526151@ruti.visi.com>
Message-ID: <uc8ktj1fv8gccc@news.supernews.com>

"Grant Edwards" <grante at visi.com> wrote in message
news:pUWw8.48099$vm6.7526151 at ruti.visi.com...
> In article <mailman.1019489191.17192.python-list at python.org>,
brueckd at tbye.com wrote:
>
>
> > Second, math *is* an attempt to describe the physical world
>
> In my experience, that's a rather unique opinion, and one not
> held by any of the math type people I've met.

That wasn't true in the past. One of the major problems
with zero was all the people going around saying: show
me zero apples! Or show me -1 oranges. The notion of
mathematics as a conceptual construct that "could" be
useful, but didn't need to be is fairly recent.

Another case in point is Euclid's parallel postulate.
Mathematicians worked on that for two millenia before
Riemann, and several actually had it solved (or at least
the spherical geometry version) and rejected their
solution because "it didn't make sense," that is, they
couldn't wrap their minds around the notion that a
straight line was a closed loop.

John Roth

>





From mal at lemburg.com  Fri Apr  5 07:35:57 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 05 Apr 2002 14:35:57 +0200
Subject: ANN: eGenix.com mxCGIPython Version 0.5.0
Message-ID: <3CAD9A2D.1BB4F7A7@lemburg.com>

________________________________________________________________________

ANNOUNCING:
                 eGenix.com mxCGIPython Version 0.5.0

            An easy-to-install one-file Python Interpreter

________________________________________________________________________

WHAT IS IT ?:

mxCGIPython is a project to provide a complete ready-to-run
Python interpreter in one single executable file. The main
intent for supplying such a single file executable is to 
ease deployment of Python for CGI scripting even when the ISP
is not willing or able to install Python for you.

The mxCGIPython archive contains everything needed to build
the one file interpreter. The web page provides a set of links
to contributed binaries for direct use in addition to setup
instructions to get the binaries up and running at your ISP.

See the documentation page for a list of available binary
packages. Binaries for new platforms are always welcome :-)

________________________________________________________________________

WHAT'S NEW ?

Version 0.5.0 now supports importing shared modules using
the pre-compiled cgipython binary.

Since all the already existing binaries don't support this
yet, I'm looking for volunteers to send in compiled versions
for all the popular platforms (e.g. different Linux distros,
FreeBSD, NetBSD, AIX, Sun Cobalt, etc.).

________________________________________________________________________

WHERE CAN I GET IT ?

The download archives and instructions for installing the packages can
be found at:

     http://www.egenix.com/files/python/

________________________________________________________________________

WHAT DOES IT COST ?

The packages comes with a Python 2.0 style license, which means
that you can use them in both commercial and non-commercial settings
without fee or charge.

The binaries are covered by the usual Python licenses.

________________________________________________________________________

WHERE CAN I GET SUPPORT ?

Free support is available through the eGenix.com User Mailing List
which provides a forum for eGenix.com Software Users:

http://lists.egenix.com/mailman/listinfo/egenix-users/

Commercial quality support for these packages is available from
eGenix.com Software GmbH. Please see

http://www.egenix.com/files/python/eGenix-mx-Extensions.html#Support 

for details about the eGenix support offerings.

________________________________________________________________________

Enjoy,
-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/





From woodsplitter at rocketmail.com  Sun Apr 14 19:21:56 2002
From: woodsplitter at rocketmail.com (David Rushby)
Date: 14 Apr 2002 16:21:56 -0700
Subject: Specifying a time from previous day
References: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net>
Message-ID: <7876a8ea.0204141521.1f6570ed@posting.google.com>

"jrup" <jrup at mail.com> wrote:
> I need to search some log files, but only if they were created after 4:00pm
> the previous day.
> 
> I've looked at the python time module and its documentation, and have
> skimmed mxDateTime documentation. I'm using activestate python 2.1.1 and
> would like to stay with it's native library since it's the one sanctioned by
> our senior nt person (also not sure how it would interact with the
> activestate install, or I'd try it).
> 
> Can someone please point me in the right direction? Is this doable using the
> time module? I haven't had much luck figuring it out, yet.

  It's certainly "doable" using the time module (see below), but
since:

a) the time module is rather low-level
b) mxDateTime is freely available, and high-level
c) you're a self-described newbie

  , why confuse yourself by using the time module?

  I see no reason why installing the eGenix.com mx Base Package would
harm an installation of Activestate Python.

  An mxDateTime-based example appears below (with equivalent code
based on the time module also present, but commented out):
----------------------------------------------------------------
import glob, os
import mx.DateTime as dt
# import time

beginningOfToday = dt.today()
eightHours = dt.DateTimeDelta(0, 8)

# beginningOfToday = time.mktime(list(time.localtime()[:3])
#    + [-1] + [0] * 5)
# eightHours = 60*60 * 8

minTime = beginningOfToday - eightHours
minTimeAsTicks = minTime.ticks()
print 'Will process all log files modified at or later than', minTime

# minTimeAsTicks = beginningOfToday - eightHours
# print 'Will process all log files modified at or later than',
# print time.asctime(time.localtime(minTimeAsTicks))

# logFilenamePattern could be input as a command-line
# argument using sys.argv
logFilenamePattern = r'c:\temp\*.txt'
logFilenames = glob.glob(logFilenamePattern)

# Build a list of relevant log filenames, including only the names of
# files that have been modified at or later than minTimeAsTicks.
relevantLogFilenames = [
    logFilename 
    for logFilename in logFilenames
    if os.stat(logFilename)[8] >= minTimeAsTicks
  ]

for logFilename in relevantLogFilenames:
    logFile = open(logFilename)

    print 'processing %s...' % logFilename
    # Do some processing here.

    logFile.close()


From andymac at bullseye.apana.org.au  Wed Apr 17 05:48:29 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Wed, 17 Apr 2002 20:48:29 +1100 (edt)
Subject: pre-PEP for optional 'pass'
In-Reply-To: <3CBB80CC.9C57B055@python.org>
Message-ID: <Pine.OS2.4.32.0204172037220.16906-100000@tenring.andymac.org>

On Tue, 16 Apr 2002, Guido van Rossum wrote:

> phil hunt wrote:
>
> > I wish to make the pas startement optional. That is, in any Python
> > program where there is a pass it can be left out and the meaning of
> > the program is unchanged.
>
> I expect I may be too late here, but I'd like to point out that I
> will never accept such a proposal, so you can stop wasting your time.
> I've seen your arguments, and they don't convince me.  Sorry. :-)

That is your prerogative.  However I understood one of the roles of PEPs
was that when rejected, they recorded for posterity your reasons for
rejecting them.  Such records would help quench the recurring bouts of
foaming-at-the-mouth featuritis that clogs c.l.py/python-list from time to
time, as the afflicted could be directed to the PEP - it may not cure
them, but it would tell them why they were beyond help...

Alas several rejected PEPs (don't recall which off the top of my head)
have just been recorded as "rejected" with no rationale for rejection.

Which _is_ a waste of effort, IMO.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From hughett at mercur.uphs.upenn.edu  Tue Apr 16 10:18:57 2002
From: hughett at mercur.uphs.upenn.edu (Paul Hughett)
Date: 16 Apr 2002 14:18:57 GMT
Subject: ideas for the best way to do some math, please??
References: <693d5681.0204160516.541c0bfe@posting.google.com>
Message-ID: <a9hbsh$pmn$1@netnews.upenn.edu>

Ben Carrington <bjc32 at hermes.cam.ac.uk> wrote:

: i am trying to evaluate a reatively simple equation but i keep getting
: this error:

: overflowError: math range error

: ...pointing to math.log(p)
:         ...
:         #part3
: 	part3 = 1 - math.e**part2
: 	p = abs(part3)

: 	#part4
: 	part4i = math.log(p)
: 	part4ii = k*T*part4i
:         ...

I would suspect that p is zero, in which case the log is undefined (or
minus infinity, depending on your conventions).  If that's not the case,
print out the values of the various variables and show them to us.

Paul Hughett


From sholden at holdenweb.com  Tue Apr 30 08:18:15 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 08:18:15 -0400
Subject: Asynchronous Socket with the ServerSocket module (??)
References: <8Xgz8.730$5y4.17760@castor.casema.net>
Message-ID: <J_vz8.12468$xy4.11273@atlpnn01.usenetserver.com>

"guyon moree" <cheezus at wanadoo.nl> wrote in message
news:8Xgz8.730$5y4.17760 at castor.casema.net...
> hi all,
>
> i'm trying to write some sort of server program on linux. i stumbled into
> the ServerSocket module and it does pretty much what i want. i create a
> requesthandler class and pass it to the module and call server_forever()
By
> using ForkingTCPServer i am able to make it handle multiple connecties.
>
>
> all fine...
>
> but...
>
> how can i keep a list of connections and pass for example messages from
one
> client to the other through the server?
>
> thanx in advance, i am a newbie though.
>

Note that the SocketServer library was principally designed for services in
which a TCP connection is used for each request, rather than being
maintained continuously over a bunch of interactions. The server is
therefore structured for "requesthandler" to make sanes (as a name) in this
context.

To write (for example) a telnet server using the SocketServer you would have
to have your request handler establish a session and go through all the
interactions with the remote client, terminating only when the remote client
sent end-of-file.

So RequestHandler is something opf a misnomer: if you think of it as
"SessionHandler" you might get more sense out of the code.

lots-of-work-in-the-details-though-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From starx at pacbell.net  Tue Apr  9 23:40:36 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 10 Apr 2002 03:40:36 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018379540.8525.python-list@python.org>
Message-ID: <Xns91EBD286FB57ARASXnewsDFE1@130.133.1.4>

Gustavo Cordova || Tue 09 Apr 2002 12:01:15p:

 
> The address is posted at the front desk in the PSU's off

Hm. p's u's off? =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From Chris.Barker at noaa.gov  Tue Apr 16 13:14:45 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 16 Apr 2002 10:14:45 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz>
Message-ID: <3CBC5C04.8F2C3D5A@noaa.gov>


Greg Ewing wrote:

> In theory, it would seem that you could install Unix python
> on Mac OS X together with the Aqua version of Tcl/Tk, and
> run Tkinter stuff.
> 
> When I tried this recently, however, it didn't work. I
> got a Tkinter window, but it was "dead" -- it didn't
> respond to mouse clicks. 

Check out the MacPython SIG mailing list, I think this is now working.
Also, wxPthon on MacOS-X is getting very close.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From robin at jessikat.fsnet.co.uk  Sat Apr  6 18:08:34 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 7 Apr 2002 00:08:34 +0100
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk>

In article <mailman.1017873312.18135.python-list at python.org>, Guido van
Rossum <guido at python.org> writes
>Here's one final posting of PEP 285.  I've read (or at least skimmed)
>all feedback that was on our news server by 9:30am EST on April 2nd,
>plus anything mailed directly to me or to python-dev until now.

If this was a final post why are you now retiring to the dev list after
another 5 days.

It seems pretty pointless to appeal to the masses if you wish only to
have a resounding clap on the back and then when that doesn't happen to
retreat to your friendly worshipers.

O Gods when Rome burns where are you?
-- 
Robin Becker


From tim.one at comcast.net  Tue Apr 16 20:34:42 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 20:34:42 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <a9i114$hq6$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCOENGOOAA.tim.one@comcast.net>

> 1. Boot the box.
>
> 2. Install 2.2.1.
>
> 3. Uninstall 2.2.1.

[Bengt Richter]
> What happens if you shut down and boot here?

Nothing visible.  The entire \Python22 directory is gone (including
WISE.EXE) immediately after completing step 3.  After a reboot, it's still
gone <wink>.

> (I'm wondering if there's a deferred delete waiting for a reboot
> which doesn't get cleared by the re-install at 4.)

Of course.  But since the only relevant file in use at the time UNWISE.EXE
is running *is* UNWISE.EXE, and Wise already got rid of it, it appears to be
a Wise bug that it arranges to delete, at the next reboot, any file named
\Python22\WISE.EXE the user happens to create after Wise already got rid of
the original.





From trentm at ActiveState.com  Tue Apr  9 15:22:08 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Tue, 9 Apr 2002 12:22:08 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F062334CFB94@mail.hebmex.com>; from gcordova@hebmex.com on Tue, Apr 09, 2002 at 02:01:15PM -0500
References: <F7DB8D13DB61D511B6FF00B0D0F062334CFB94@mail.hebmex.com>
Message-ID: <20020409122208.A6471@ActiveState.com>

[Gustavo Cordova wrote]
> >
> > Is there a way to peek into the latest python-dev messages? I could
> > not find a link at http://www.python.org/psa/MailingLists.html
> > Maybe if you subscribe to that mailing list (have to find their
> > address first) you get too much email to handle so being able to just
> > have a look at the lastest postings would be nice.


http://aspn.activestate.com/ASPN/Mail/Browse/Threaded/python-dev


Trent

-- 
Trent Mick
TrentM at ActiveState.com




From cameron at hooper.net  Sun Apr 28 22:00:19 2002
From: cameron at hooper.net (Cameron Hooper)
Date: Sun, 28 Apr 2002 22:00:19 -0400
Subject: Numeric question
References: <pan.2002.04.25.21.32.03.310021.2578@hooper.net> <YM2y8.71958$HH5.5067@rwcrnsc51.ops.asp.att.net> <3CC9A844.F1BE7E7B@noaa.gov>
Message-ID: <%O1z8.808$SY6.23031@news.itd.umich.edu>

Chris Barker wrote:

> 
> 
> Emile van Sebille wrote:
> 
>> OK.  It's x.  So let's build a contiguous x and try it again:
> 
>> >>> x = array(inputs[0][:])
> 
> or
> 
> x = inputs[0][:].copy()
> 
> The copy method isn't in the version of the docs I have, but it's very
> handy and quite clear as to what it is doing.
> 
> or
> 
> x = inputs[0,:].copy()
> 
> Numpy arrays are N-dimensional, you might as well index them that way.
> 
> come to think of it, why not:
> 
> inputs[0].copy() ?
> 
> 
> -Chris
> 

Thanks for the replies. So my understanding now is that the concatenate 
function really combines two references to two separate locations in memory 
into a single object. The two locations are not contiguous. Data is not 
'moved around' so to speak. So x = concatenate(y,z) can be viewed as 
returning a reference to two non-contiguous references.  But x = 
array(concatenate(y,z)) (or using copy()) actually copies data.

Cameron


From eds at yahoo.com  Mon Apr 15 21:31:15 2002
From: eds at yahoo.com (Ed Simmon)
Date: Tue, 16 Apr 2002 01:31:15 GMT
Subject: ActiveState Licenses - 21 day trial doesn't work
Message-ID: <D9Lu8.875$oj5.60546@typhoon.austin.rr.com>

I have tried to evaluate the ActiveState Visual Python and Visual Perl tools
but when I run their license EXE and then try to open a project in VS.NET it
tells me the license expired 9 days ago.  I just downloaded everything last
night, so my 21 day evaluation wasn't even close to being expired.  One of
my developers sent an email to ActiveState's licensing yesterday but it has
gone unanswered, I guess they don't need new customers.  I was hoping to
evaluate these tools before shelling out $295 per developer.

I even went went and re-registered and got another EXE license program to no
avail.  This is not a good system they have.

Anyone else had trouble with getting these eval copies of ActiveState's
tools working?





From rdacker at pacbell.net  Tue Apr 30 15:27:03 2002
From: rdacker at pacbell.net (rdack)
Date: 30 Apr 2002 12:27:03 -0700
Subject: import from a file in a subdirectory
Message-ID: <644f6688.0204301127.5d3372d@posting.google.com>

i have a script running in a directory.
i want to call a subroutine in a file in a subdirectory.

in the first file i said  from subdir/ff.py import  subrtn
python doesn't like the slash.

how do i do it?


From 374882.554635406 at 124721.179250628  Tue Apr 23 14:22:03 2002
From: 374882.554635406 at 124721.179250628 (Help Please)
Date: Tue, 23 Apr 2002 18:22:03 +0000 (UTC)
Subject: Help Please
Message-ID: <aa48oa$f9e$357@rex.ip-plus.net>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020423/fa691099/attachment.html>

From jimd at vega.starshine.org  Wed Apr  3 20:11:18 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 4 Apr 2002 01:11:18 GMT
Subject: Catalog of Python Patterns? [LONG!]
Message-ID: <a8g97m$10jq$1@news.idiom.com>

 GoF Design Patterns in Python

 [Warning LONG POST AHEAD!]

 I'd like to see a canonical catalog of OODP in Python.

 At a minimum it should list each of the 23 GoF patterns, and
 at least one minimal Python implementation of that pattern.
 It should also point out the "native" use of each pattern in
 the Python core, standard modules or among popular publicly 
 available third-party modules, packages and applications.  
 Naturally it should also discuss alternatives (Borg vs. Singleton)
 and issues (particularly cases where Pythonic idioms make 
 patterns trivial, for example).

 I realize that there used to be a "Python-Patterns" SIG (now
 "retired") and a Google search will reveal a number of articles
 on the topic (such as the one by Vespe Savikko).  Unfortunately 
 none of these sources is comprehensive.  Not one of these articles
 or web sites shows a list of all 23 GoF patterns with a minimal
 Python implementation.  Also many of them are somewhat dated.  
 I'd like to focus on Python 2.2 features (nods and caveats regarding
 older versions can be added later).

 I'll be the first to admit that I'm not a "Patterns" expert.
 That's why I'd like to see *simple* (as it brain-dead, dirt,
 simple) examples.  I'm also not an experienced professional 
 OO programmer.  I consider myself to be a student in this field.
 So all of my comments should be take with a large lump of salt,
 perhaps with a whole 5 lb. salt lick (for you equestrians).

 Here's a start.  (All I've done is copy the GoF 23 from the 
 inside cover of their book, and inserted my questions, comments 
 and suggestions.  Note that many of my comments are laced with
 the phrase "I imagine" or qualifiers like "maybe" or "could" ---
 like I said, I'm just studying this stuff and I have little
 practical experience with it).

 Who wants to contribute "text book" sample implementations?

 Creational::
   Abstract Factory:
    I've never understood the difference between an "abstract factory"
    and a "factory method."  Since Python doesn't really have 
    "abstract" versus "concrete" classes, perhaps the distinction is
    meaningless.

    Every Python class acts like a factory.  anObj = Foo() returns
    a reference to a new "Foo" object.  Additionally Python allows
    us to define normal functions which return object references.
    Indeed there is no way to tell from this assignment whether Foo
    is a class or a function.  It could easily be a function that 
    returns an object from a pool or one that returns a reference to
    a singleton. (Of course having a function masquerading as a 
    "class" in this way would inelegant and might violate the 
    principle of least surprise for some of the programmers using
    or extending this code).

    In Python2.2 it's possible to create/over-ride a class method
    with the reserved __new__ name so that we have greater control
    over the new instance creation (instantiation).  Guido shows
    how __new__ can be used to implement a Singleton. It should be
    easy to use __new__ to implement object pools (is that a 
    non-GoF pattern?), or any of creational patterns. (see below)

	Clearly, for a number of patterns, one object will be acting
	as some form of dynamic interface for one or more objects 
	"behind it."  (Decorator, Adapter, Bridge, Proxy, Facade, 
	etc).  The __init__ method (constructor) may have to instantiate
	these back-end objects, storing references to them as instance
	members.  The resulting "front-end" object must then dispatch
	method invocations (and attempted member accesses) to the 
	appropriate back-end object(s) according to the required semantics.

	Indeed this model seems to be a unifying theme among many of the
	GoF patterns.  A Decorator enhances a back-end object by adding
	it's to members or methods to it (like a wrapper) and dispatching
	"base" methods back into it's instance of the underlying object.
	Bridges and Adapters reformulate method calls as the pass them
	back.  A Proxy selectively rejects some method invocations, or
	over-rides them, to "protect" the underlying object in some way.
	A Facade allows a group of objects to appear as a single object;
	it dispatches methods to the appropriate back-end instances.
	(In any of these cases the "front-end" might also wrapper or
	over-ride the return values). 

	Quite a bit of the motivation and inspiration for the GoF 
	patterns seems to be in recognition of the limitations of 
	inheritance and class hierarchies.  They talk quit a bit about
	how these patterns allow OO programmers to avoid common
	pitfalls that result from attempts to use inheritance for 
	"has a" rather than "is a" relationships, for cases where 
	objects need to interact with other objects or in ways OTHER
	than inheritance.
	
	[Perhaps I'm wrong about this, but this is my take on the GoF
	patterns. BTW: the term "delegation" might be useful here.]

   Builder:
     I guess that "Builders" would often be used for "Composite" 
	 objects.  

   Factory Method:
     Are there any cases in Python 2.2 where this would be implemented
	 through something other than some combination of code in __new__
	 and/or __init__?

   Prototype:
     Are there any examples of this in the Python distribution?
	 I guess the implementation would usually be to create a 
	 _clone() or __clone() method which would instantiate a blank
	 object (clone = self.__class__()?) passing it appropriate 
	 constructor arguments (which it must have cached in its own
	 __init__) and calling on whatever methods are necessary to 
	 put the clone into the desired (mostly identical) state as
	 the prototypical instance.  Finally the clone() function 
	 would return a reference to the newly cloned instance.

	 I suppose one could have a Prototype (possibly some
	 selected instances of a class) define the __call__ method to
	 perform the cloning.  This would make the prototype look like
	 a factory or a normal class object throughout the rest of the
	 program.  I could envision a "Prototype" function that would
	 convert (promote?) a "normal" instance variable into a 
	 Prototype by causing it to bind its __call__ method to the
	 necessary cloning operations.  This might look a little like
	 the way that one defines a class or static method and "promotes" 
	 using syntax of the form:  mymethod = staticmethod(mymethod).
	 (Obviously the same function could be used to create a 
	 new Prototype from any extent instance: fooProto = Prototype(inst)
	 instead of myproto = Prototype(myproto); the only difference
	 is whether you bind the result to the old name or a new one).

	 One question that naturally comes up with regard to Prototypes
	 is this: How does Python's introspection features facilitate
	 the implementation of Prototypes.  How much of the cloning 
	 operation can be standardized and encapsulated in our 
	 (hypothetical) "Prototype()" function (or mix-in class?)?
	 [I can imagine all sorts of clever fiddling with self.__dict__
	 et al --- possibly using the deepcopy module/functions; but
	 the actual fiddling is beyond me at this point, so it's 
	 best left to the imagination].
	 
   Singleton:
     Alex Martelli has long been a proponent of the Borg pattern
     as an alternative to the Singleton.  While his point is well
	 taken (the focus should be on behavior, not object identity)
	 I think of the Borg as a way to work around limitation in the
	 earlier versions of Python.  With the introduction of new 
	 classes and the ability to create static methods and over-ride
	 __new__ there is no need to use the Borg. 

	 Incidentally the Borg is yet another example of the delegation
	 theme that I discussed before.  It redirects all access to 
	 any member of the collective to the "queen" (back-end) instance.

 Structural::
   Adapter:
   Bridge: 
     Frankly I don't see much difference between these two patterns.
	 I guess anygui might use Adapters to provide one uniform
	 interfacee to GUI widgets while adapting method calls into the
	 various semantics of back-end GUIs and widget collections.
	 A Bridge does the same thing.  GoF makes the distinction on 
	 "intent" (Adapters being used to access legacy or external 
	 classes through a common API, Bridges being a design decision
	 to decouple one's own classes from their APIs; it sounds like
	 hair splitting to me -- the same implementation techniques
	 should work in either case.

	 I can imagine cases where an Adapter or Bridge might have to
	 also be a bit of a Decorator or a Facade. If the semantics of 
	 one API differ from another, for example, one might have to 
	 instantiatiate additional backend objects to supply enough
	 functionality to support the desired common interface.  Otherwise
	 your interface is constrained to a least common denominator
	 among the possible back-end classes.

	 In these (and all of the other "delegation" patterns) I can
	 imagine lots of tricky, clever, subtle (and "too" extremes
	 of each: as in "too tricky," "too clever," and "too subtle")
	 ways of using __getattr__ and __setattr__ methods and the 
	 hasattr(), to avoid explicit enumeration of the back-end
	 methods for those cases where the backend does have the 
	 same interface as the Adapter/Bridge.  In other words it
	 may be possible through __getattr__ and __setattr__ or
	 similar special Python methods to allow any of these delegate
	 Patterns to only over-ride the necessary functions while
	 transparently passing along any other methods (and transparently
	 passing back results or raising exceptions?).

	 What are some pointers and caveats about this sort of "dynamic 
	 delegation" (sort of "inheritance from outside of the family 
	 tree")?
	 
   Composite:
     The GoF refers to Composites as "tree structures."  This strikes
	 me as odd since I can imagine many composites of objects that
	 aren't trees.  I can imagine chains, directed graphs, networks, 
	 and pools of objects that are interconnected through instance
	 members.  I can also see where one might have class/static 
	 members which track all of the extant instances (__new__ might
	 append to "__class__.__instancelist" and __del__ might perform
	 the remove.  Could __del__ "divert" the destruction, blank the
	 instance and add it to a pool?  Might one define a wrapper around
	 the del() built-in that would do something like: 

		def del(arg):
			if hasattr(arg,'__pool__'): arg.__pool__()
			else:                     _del(arg)

	But I gather (from reading the GoF) that the object here
	(to excuse the pun) is to allow for a method to propagate
	transparently to any subset of the composite regardless
	of whether that subset effects the whole tree (invoke it
	on the root node) or an individual leaf (the degenerate
	case).  So I guess the core of a Python Composite with
	support for a given method's propagation might be:

		Class CompositeFoo:
			def __init__(self):
				self.children = [] 
			def propagated_method(self, *args, **kwargs):
				# do stuff to self
				if supposed_to_propagate:
					for i in self.children:
						i.propagated_method(args, kwargs)

	[Does that make sense?  BTW: I'm envisioning some cases
	where the propagation is intended to be limited or conditional]

	So, what are the caveats?  Are there any examples in Python
	itself?  Are there any coding conventions that might be
    encapsulated into an abstract base class or a mix-in? 

   Decorator:
   Facade:
     As I've already said, Decorators and Facades, Adapters, Proxies
	 and Bridges (and Mediators and possibly even State objects all 
	 seem to be forms of delegation (or collaboration?).  In all 
	 cases there is a delegate (front end) object which is "used" 
	 in our code and it instantiates one or more of the underlying 
	 (back end) objects as its own member(s).  The delegate then 
	 dispatches or relays method invocations back to the underlying 
	 object(s) according the the delegate's semantics.  

	 Am I missing something here?  Is this a gross 
	 oversimplification?  Are there any differences among these
	 patterns that require special implementation (for Python)
	 or to the same language features and techniques apply to 
	 all of these "delegations" (my term).

   Flyweight:
     I've read innumerable times that the Python interpreter 
	 implements a set of static immutable objects for a small 
	 range of integers (-1 or -2 through 100 or so?).  I guess
	 this is an example of Python's native use of the Flyweight
	 pattern.  

	 Are there any others?  Would it be possible to have a 
	 sort of "copy on write" (COW) Flyweight?  These would behave
	 like Singletons for read-only purposes, but then be promoted
	 to separate/new objects on any setattr?  What would be a 
	 use for such a beast.

   Proxy:
     See my comments for Decorators, Facades, Adapters, and Bridges.


 Behavioral::
   Chain of Responsibility:
     A cascaded of Decorators or a Chain of responsibility.  
	 What's the difference.

   Command:
     Some of the GoF's comments on this sound like it would
	 be ideal for supporting AOP (aspect oriented programming).

   Interpreter:
     I just don't get this pattern.  I'll have to mull on it
	 for a bit.  Other the trite observation that Python *is*
	 an interpreter, are there any examples of the Interpreter
	 pattern in Python?

   Iterator:
     Iterators seem to be evolving into one of Pyton's most 
	 visible and commonly used features.  In Python2.2 we 
	 see automatic iteration over files, and dictionaries as
	 well as the traditional iteration over built-in sequences
	 (lists, tuples, strings).  We also see support for defining
	 iterators for our own classes.

	 It seems like iterators and generators are intertwined in
	 someway, but I don't understand it.  [So maybe I'm just
	 wrong and they only seem to be related because they're 
	 described in the same docs, or something].

   Mediator:
     This seems like a sort of bi-directional Facade to me.
	
   Memento:
     Here's where Python's "encapsulation by courtesy" and lack of
	 enforced data hiding might make a GoF pattern particularly 
	 easy; almost unecessary.  Alternatively I can imagine that 
	 some implementation of the "Memento" pattern might involve a
	 coding convention where each class that supports it implements
	 an __memoize() method.  Might that return a pickle of self?

   Observer:
	 I can imagine where some objects, including composite
	 objects might "register" with some sort of "registry" (sorry
	 for the redundancy).  I'm just not sure what sort of applications
	 one my find for this sort of "registry" even though I'm sure
	 it would be easy (almost trivial) to implement in Python.

	 I suppose such a "registry" makes quite a bit of sense in the 
	 Observer pattern.  Although I gather that this would usually be
	 a instance registry in each of the "observer/dispatchers"

   State:
     I see this as similar to an Adapter except that it may 
	 change which back end object(s) it dispatches to (or calls
	 upon or delegates for) might change dynamically during the
	 lifetime of a given instance.  For example I can imagine an
	 object that implements a simple list for some of its members
	 until the list gets beyond a given size, then it might 
	 change that to a dictionary, or some sort of AVL tree or
	 more complex data structure at the object needs to scale 
	 upwards.

   Strategy:
   Template Method:
     Would some of the mix-ins from SocketServer count as 
	 examples of the Strategy or Template Method patterns?
     
   Visitor:
     os.path.walk is an obvious example of a standard Python
	 module/function that uses Visitor.  (Obvious because it's
	 documented as the "visitor" function).

     However, this is also one of the most confusing functions
	 in the standard Python library.  I seem to see  befuddled 
	 posts about it in the c.l.py newsgroup at least once a month.
	 As the GoF point out, Visitors are often used on Composites.

	 It seems like generators might often be alternatives to 
	 Visitors.  Rather than passing a function to each object
	 and saying "do this to yourself" it seems natural to 
	 iterator over a set of object references and invoke the 
	 methods (or even, horrors! play with its attributes) directly.
	 
	 However, Python's ability to pass functions, objects and 
	 closures around as first order data types seems to make 
	 Visitors easy to implement --- if you can just understand
	 the scoping well enough for the Visitor to know what to
	 do (and who to do it to) when it gets there.



From lozinski at openstepnews.com  Wed Apr  3 12:22:59 2002
From: lozinski at openstepnews.com (Christopher Lozinski)
Date: Wed, 03 Apr 2002 09:22:59 -0800
Subject: Python Banner Exchange
In-Reply-To: <E16skiA-0000tF-00@mail.python.org>
Message-ID: <5.1.0.14.0.20020403092218.00a08bc0@imap.imeme.net>

Is growing rapidly at
www.pythonandzope.com/BannerInfo

You are invited to join.

Regards
Christopher Lozinski





From bokr at oz.net  Tue Apr  9 04:40:31 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Apr 2002 08:40:31 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8sf0k$2a5$0@216.39.172.122> <ahjs8.73608$S52.2597704@news2.tin.it> <a8sn8u$sj7$0@216.39.172.122> <Eeos8.74753$S52.2651788@news2.tin.it>
Message-ID: <a8u9dv$lfo$0@216.39.172.122>

On Mon, 08 Apr 2002 21:48:20 GMT, Alex Martelli <aleax at aleax.it> wrote:

>Bengt Richter wrote:
>        ...
>> Well, I guess that depends on how you define "assignment" in Python.
>
>I don't have to, as the Language Reference sees to that:
>http://www.python.org/dev/doc/devel/ref/assignment.html
>
>> ISTM there really is no such thing, in the sense of updating a named
>> storage place with new content, 
"...storage place with new content, so what goes by the name "assignment"
needs some explanation. If the statement a=b is an assignment,
then binding the name 'a' to object b is the meaning of Python
assignment as I see it."

>"Assignment statements?are used to (re)bind names to values and to modify 
>attributes or items of mutable objects" -- I guess one COULD at the
>limit call an attribute "a named storage place", but it would not be a very
>fruitful thing in Python.
ISTM I did not contradict or distort that.

>
>> expression ;-) I'm not sure what that side effect should be called
>> if not assignment, since it caused the identical binding that
>> a='bee' would in that context.
>
>It's a serious logical fallacy to reason "assignment statements may
>(among other things) rebind names, therefore what rebinds a name
>is an assignment". Many bindings occur other than by assignment.
Sure, but that was not my reasoning. I was saying that assignment (not
assignment statements) in Python is effectively an alias for rebinding.
I can thus call bindings other than those caused by assignment _statements_
assignments (though I try to avoid the word, because I think its misleading
in Python). I'll even include anonymous temporary bindings in the concept,
and call those temporary "assignments" (though I know they're bindings).
            
>Would you call a for statement, a def statement, a class statement,
>or an import statement "assignments", because they also rebind
>names?!
>
To quote myself, "My criterion for judging whether it was an assignment was whether
it accomplished the same name re/binding as a normal assignment,..."
I would say they do implement Python assignment functionality as part of what they do.
I certainly would not call them assignment _statements_.

>> If it binds like a duck... ;-)
>
>Should I try to display my imitatory abilities by quacking like a duck,
>that does not justify your shooting me even if you have a hunting
>license and ducks are in season.
>
Point. I left that in against my better judgement. That should teach
me to borrow ill-fitting politicians' cliches ;-)

>> My criterion for judging whether it was an assignment was whether
>> it accomplished the same name re/binding as a normal assignment,
>> and I think it did, in the demonstrated context. So IMO Aahz's
>> quote doesn't quite remain perfectly right ;-)
>
>In many restricted contexts it's long been perfectly feasible to
>perform a binding (and nothing but that) within an expression.
>Your example was horrid -- it works "by accident" in a given
Horridness is a side issue w.r.t whether "Assignments in expressions
will never happen." But we can go there... ;-)

>implementation, in a certain restricted context, because a
>undefined behavior happens to be that way in said specific
>implementation.  One can of course do much better than that:
>
>        [x for x in (y,)][0]
>
Yes, I thought of that one too, while I was out ;-)

I think that one is actually _more_ "horrid" ;-) Why? Because it depends
on what IMO should _not_ work and my example depends on what reasonably IMO
_should_ work (if optimization issues did not take practical precedence).
IMO letting a binding side effect escape the scope of a list comprehension
is a flaw (and I'm guessing that it will be fixed). In contrast, locals()
returns a purported dictionary of local bindings, and IMO it's reasonable
to expect to be able to modify them through dictionary operations,
unless warned off conspicously.

>is an expression perfectly valid in all contexts which binds or
>rebinds name x to the same value as name y is bound to and
>also has that value as its value-as-expression.  It doesn't
>take all that much Python knowledge to build it, either.
>
Nope, I thought of it too ;-)

>So, whatever you could do in a language that allowed assignments
>in expressions you could do in Python via that hard-to-read way, e.g.:
>
>while [x for x in (getnext(23),)][0]:
>    frobnicate(x)
>
>is the Python (but NOT Pythonic) equivalent of C's
>
>while(x=getnext(23))
>    frobnicate(x);
>
>That doesn't affect the "perfectly right" status of Aahz's quote, either.
>
Nope. It's a "horrid" example confirming to me what that status is ;-)


Or perhaps I should say was: I saw another post from Aahz, which said,
"""
    I define "assignment" in Python as being precisely equal to "=".

"""
So that changes the ball game and clinches it. He's perfectly right, unless Guido
does something to change the situation. ;-)

Regards,
Bengt Richter


From claird at starbase.neosoft.com  Wed Apr 24 11:02:07 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2002 10:02:07 -0500
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org> <aa1fe9$mn7$1@peabody.colorado.edu>
Message-ID: <4B8917DC906C2F52.202DFBF1D991CC04.F8F981F11AE8D2D1@lp.airnews.net>

In article <aa1fe9$mn7$1 at peabody.colorado.edu>,
Fernando =?ISO-8859-1?Q?P=E9rez?=  <fperez528 at yahoo.com> wrote:
			.
			.
			.
>What I do find most fascinating is how many seemingly super-abstract results 
>of mathematics turn up (sometimes decades later after they are found) having 
>physical implications which noone ever suspected. That 'mysterious 
>mathematical structure of the physical world' is one of the most beautiful, 
>in my opinion unanswerable questions of the foundation of science.
			.
			.
			.
It's surprised me a bit that no one has yet followed up with
the obligatory original references.  I do so: <URL: http://
www.dartmouth.edu/~matc/MathDrama/reading/Wigner.html > and
<URL: http://www.lecb.ncifcrf.gov/~toms/Hamming.unreasonable.html >.
-- 

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


From dwelch91 at nospam.attbi.com  Wed Apr  3 20:50:13 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Thu, 04 Apr 2002 01:50:13 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <3CABB156.1010401@nospam.attbi.com>

Pardon my ignorance (I've only been working with Python for a short 
while), but I don't understand why an idiom like:

if x == True:

would be/should be an error (or warning initially). I see C/C++ code 
like this all the time:

bool x;
x = Foo(Bar);
if (x==true)
{...

Of course, this could also be written in the more abbreviated form as 
well. But, why force that on the programmer? I really prefer the more 
explicit  comparison of x with true rather than the if(x) idiom. Myself, 
being a newbie Python programmer (but not a newbie programmer in C, C++, 
etc), I would find it somewhat confusing that I could not compare a 
variable to a valid value for the variable using what appears to be a 
valid operator.

Am I missing something more subtle here?

Just curious, sorry for the interruption.

Don

Guido van Rossum wrote:
> Here's one final posting of PEP 285.  I've read (or at least skimmed)
> all feedback that was on our news server by 9:30am EST on April 2nd,
> plus anything mailed directly to me or to python-dev until now.
> 
> There was a lot of negative feedback, but there was also a lot of
> support for the PEP.  The feedback made it easy to decide on most of
> the smaller issues in the review section of the PEP as previously
> posted: even reviewers who were against the PEP mostly supported my
> preferences for the various details.
> 
> Despite the negative feedback, I've decided to accept the PEP.  The
> most important perceived problem is that newbies tend to write
> 
>     if x == True: ...
> 
> where they should write
> 
>     if x: ...
> 
> I believe this problem should be solved by education (the Zen master
> hits the student on the head with a stick, and the student gets
> enlightenment) rather than by holding back what I believe will be a
> useful feature.  You can expect the bool type in Python 2.3; it will
> be in CVS tonight.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)
> 
> PEP: 285
> Title: Adding a bool type
> Version: $Revision: 1.20 $
> Last-Modified: $Date: 2002/04/03 22:11:05 $
> Author: guido at python.org (Guido van Rossum)
> Status: Accepted
> Type: Standards Track
> Created: 8-Mar-2002
> Python-Version: 2.3
> Post-History: 8-Mar-2002, 30-Mar-2002, 3-Apr-2002
> 
> 
> Abstract
> 
>     This PEP proposes the introduction of a new built-in type, bool,
>     with two constants, False and True.  The bool type would be a
>     straightforward subtype (in C) of the int type, and the values
>     False and True would behave like 0 and 1 in most respects (for
>     example, False==0 and True==1 would be true) except repr() and
>     str().  All built-in operations that conceptually return a Boolean
>     result will be changed to return False or True instead of 0 or 1;
>     for example, comparisons, the "not" operator, and predicates like
>     isinstance().
> 
> 
> Review
> 
>     I've collected enough feedback to last me a lifetime, so I declare
>     the review period officially OVER.  I had Chinese food today; my
>     fortune cookie said "Strong and bitter words indicate a weak
>     cause."  It reminded me of some of the posts against this
>     PEP... :-)
> 
>     Anyway, here are my BDFL pronouncements.  (Executive summary: I'm
>     not changing a thing; all variants are rejected.)
> 
>     1) Should this PEP be accepted?
> 
>     => Yes.
> 
>        There have been many arguments against the PEP.  Many of them
>        were based on misunderstandings.  I've tried to clarify some of
>        the most common misunderstandings below in the main text of the
>        PEP.  The only issue that weighs at all for me is the tendency
>        of newbies to write "if x == True" where "if x" would suffice.
>        More about that below too.  I think this is not a sufficient
>        reason to reject the PEP.
> 
>     2) Should str(True) return "True" or "1"?  "1" might reduce
>        backwards compatibility problems, but looks strange.
>        (repr(True) would always return "True".)
> 
>     => "True".
> 
>        Almost all reviewers agree with this.
> 
>     3) Should the constants be called 'True' and 'False' (similar to
>        None) or 'true' and 'false' (as in C++, Java and C99)?
> 
>     => True and False.
> 
>        Most reviewers agree that consistency within Python is more
>        important than consistency with other languages.
> 
>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that for example
>        True+1 would eventually (in Python 3000) be illegal?
> 
>     => No.
> 
>        There's a small but vocal minority that would prefer to see
>        "textbook" bools that don't support arithmetic operations at
>        all, but most reviewers agree with me that bools should always
>        allow arithmetic operations.
> 
>     5) Should operator.truth(x) return an int or a bool?
> 
>     => bool.
> 
>        Tim Peters believes it should return an int, but almost all
>        other reviewers agree that it should return a bool.  My
>        rationale: operator.truth() exists to force a Boolean context
>        on its argument (it calls the C API PyObject_IsTrue()).
>        Whether the outcome is reported as int or bool is secondary; if
>        bool exists there's no reason not to use it.  (Under the PEP,
>        operator.truth() now becomes an alias for bool(); that's fine.)
> 
>     6) Should bool inherit from int?
> 
>     => Yes.
> 
>        In an ideal world, bool might be better implemented as a
>        separate integer type that knows how to perform mixed-mode
>        arithmetic.  However, inheriting bool from int eases the
>        implementation enormously (in part since all C code that calls
>        PyInt_Check() will continue to work -- this returns true for
>        subclasses of int).  Also, I believe this is right in terms of
>        substitutability: code that requires an int can be fed a bool
>        and it will behave the same as 0 or 1.  Code that requires a
>        bool may not work when it is given an int; for example, 3 & 4
>        is 0, but both 3 and 4 are true when considered as truth
>        values.
> 
>     7) Should the name 'bool' be changed?
> 
>     => No.
> 
>        Some reviewers have argued for boolean instead of bool, because
>        this would be easier to understand (novices may have heard of
>        Boolean algebra but may not make the connection with bool) or
>        because they hate abbreviations.  My take: Python uses
>        abbreviations judiciously (like 'def', 'int', 'dict') and I
>        don't think these are a burden to understanding.  To a newbie,
>        it doesn't matter whether it's called a waffle or a bool; it's
>        a new word, and they learn quickly what it means.
> 
>        One reviewer has argued to make the name 'truth'.  I find this
>        an unattractive name, and would actually prefer to reserve this
>        term (in documentation) for the more abstract concept of truth
>        values that already exists in Python.  For example: "when a
>        container is interpreted as a truth value, an empty container
>        is considered false and a non-empty one is considered true."
> 
>     8) Should we strive to require that Boolean operations (like "if",
>        "and", "not") have a bool as an argument in the future, so that
>        for example "if []:" would become illegal and would have to be
>        writen as "if bool([]):" ???
> 
>     => No!!!
> 
>        Some people believe that this is how a language with a textbook
>        Boolean type should behave.  Because it was brought up, others
>        have worried that I might agree with this position.  Let me
>        make my position on this quite clear.  This is not part of the
>        PEP's motivation and I don't intend to make this change.  (See
>        also the section "Clarification" below.)
> 
> 
> Rationale
> 
>     Most languages eventually grow a Boolean type; even C99 (the new
>     and improved C standard, not yet widely adopted) has one.
> 
>     Many programmers apparently feel the need for a Boolean type; most
>     Python documentation contains a bit of an apology for the absence
>     of a Boolean type.  I've seen lots of modules that defined
>     constants "False=0" and "True=1" (or similar) at the top and used
>     those.  The problem with this is that everybody does it
>     differently.  For example, should you use "FALSE", "false",
>     "False", "F" or even "f"?  And should false be the value zero or
>     None, or perhaps a truth value of a different type that will print
>     as "true" or "false"?  Adding a standard bool type to the language
>     resolves those issues.
> 
>     Some external libraries (like databases and RPC packages) need to
>     be able to distinguish between Boolean and integral values, and
>     while it's usually possible to craft a solution, it would be
>     easier if the language offered a standard Boolean type.  This also
>     applies to Jython: some Java classes have separately overloaded
>     methods or constructors for int and boolean arguments.  The bool
>     type can be used to select the boolean variant.  (The same is
>     apparently the case for some COM interfaces.)
> 
>     The standard bool type can also serve as a way to force a value to
>     be interpreted as a Boolean, which can be used to normalize
>     Boolean values.  When a Boolean value needs to be normalized to
>     one of two values, bool(x) is much clearer than "not not x" and
>     much more concise than
> 
>         if x:
>             return 1
>         else:
>             return 0
> 
>     Here are some arguments derived from teaching Python.  When
>     showing people comparison operators etc. in the interactive shell,
>     I think this is a bit ugly:
> 
>         >>> a = 13
>         >>> b = 12
>         >>> a > b
>         1
>         >>>
> 
>     If this was:
> 
>         >>> a > b
>         True
>         >>>
> 
>     it would require a millisecond less thinking each time a 0 or 1
>     was printed.
> 
>     There's also the issue (which I've seen baffling even experienced
>     Pythonistas who had been away from the language for a while) that
>     if you see:
> 
>         >>> cmp(a, b)
>         1
>         >>> cmp(a, a)
>         0
>         >>> 
> 
>     you might be tempted to believe that cmp() also returned a truth
>     value, whereas in reality it can return three different values
>     (-1, 0, 1).  If ints were not (normally) used to represent
>     Booleans results, this would stand out much more clearly as
>     something completely different.
> 
> 
> Specification
> 
>     The following Python code specifies most of the properties of the
>     new type:
> 
>         class bool(int):
> 
>             def __new__(cls, val=0):
>                 # This constructor always returns an existing instance
>                 if val:
>                     return True
>                 else:
>                     return False
> 
>             def __repr__(self):
>                 if self:
>                     return "True"
>                 else:
>                     return "False"
> 
>             __str__ = __repr__
> 
>             def __and__(self, other):
>                 if isinstance(other, bool):
>                     return bool(int(self) & int(other))
>                 else:
>                     return int.__and__(self, other)
> 
>             __rand__ = __and__
> 
>             def __or__(self, other):
>                 if isinstance(other, bool):
>                     return bool(int(self) | int(other))
>                 else:
>                     return int.__or__(self, other)
> 
>             __ror__ = __or__
> 
>             def __xor__(self, other):
>                 if isinstance(other, bool):
>                     return bool(int(self) ^ int(other))
>                 else:
>                     return int.__xor__(self, other)
> 
>             __rxor__ = __xor__
> 
>         # Bootstrap truth values through sheer willpower
>         False = int.__new__(bool, 0)
>         True = int.__new__(bool, 1)
> 
>     The values False and True will be singletons, like None.  Because
>     the type has two values, perhaps these should be called
>     "doubletons"?  The real implementation will not allow other
>     instances of bool to be created.
> 
>     True and False will properly round-trip through pickling and
>     marshalling; for example pickle.loads(pickle.dumps(True)) will
>     return True, and so will marshal.loads(marshal.dumps(True)).
> 
>     All built-in operations that are defined to return a Boolean
>     result will be changed to return False or True instead of 0 or 1.
>     In particular, this affects comparisons (<, <=, ==, !=, >, >=, is,
>     is not, in, not in), the unary operator 'not', the built-in
>     functions callable(), hasattr(), isinstance() and issubclass(),
>     the dict method has_key(), the string and unicode methods
>     endswith(), isalnum(), isalpha(), isdigit(), islower(), isspace(),
>     istitle(), isupper(), and startswith(), the unicode methods
>     isdecimal() and isnumeric(), and the 'closed' attribute of file
>     objects.  The predicates in the operator module are also changed
>     to return a bool, including operator.truth().
> 
>     Because bool inherits from int, True+1 is valid and equals 2, and
>     so on.  This is important for backwards compatibility: because
>     comparisons and so on currently return integer values, there's no
>     way of telling what uses existing applications make of these
>     values.
> 
>     It is expected that over time, the standard library will be
>     updated to use False and True when appropriate (but not to require
>     a bool argument type where previous an int was allowed).  This
>     change should not pose additional problems and is not specified in
>     detail by this PEP.
> 
> 
> C API
> 
>     The header file "boolobject.h" defines the C API for the bool
>     type.  It is included by "Python.h" so there is no need to include
>     it directly.
> 
>     The existing names Py_False and Py_True reference the unique bool
>     objects False and True (previously these referenced static int
>     objects with values 0 and 1, which were not unique amongst int
>     values).
> 
>     A new API, PyObject *PyBool_FromLong(long), takes a C long int
>     argument and returns a new reference to either Py_False (when the
>     argument is zero) or Py_True (when it is nonzero).
> 
>     To check whether an object is a bool, the macro PyBool_Check() can
>     be used.
> 
>     The type of bool instances is PyBoolObject *.
> 
>     The bool type object is available as PyBool_Type.
> 
> 
> Clarification
> 
>     This PEP does *not* change the fact that almost all object types
>     can be used as truth values.  For example, when used in an if
>     statement, an empty list is false and a non-empty one is true;
>     this does not change and there is no plan to ever change this.
> 
>     The only thing that changes is the preferred values to represent
>     truth values when returned or assigned explicitly.  Previously,
>     these preferred truth values were 0 and 1; the PEP changes the
>     preferred values to False and True, and changes built-in
>     operations to return these preferred values.
> 
> 
> Compatibility
> 
>     Because of backwards compatibility, the bool type lacks many
>     properties that some would like to see.  For example, arithmetic
>     operations with one or two bool arguments is allowed, treating
>     False as 0 and True as 1.  Also, a bool may be used as a sequence
>     index.
> 
>     I don't see this as a problem, and I don't want evolve the
>     language in this direction either.  I don't believe that a
>     stricter interpretation of "Booleanness" makes the language any
>     clearer.
> 
>     Another consequence of the compatibility requirement is that the
>     expression "True and 6" has the value 6, and similarly the
>     expression "False or None" has the value None.  The "and" and "or"
>     operators are usefully defined to return the first argument that
>     determines the outcome, and this won't change; in particular, they
>     don't force the outcome to be a bool.  Of course, if both
>     arguments are bools, the outcome is always a bool.  It can also
>     easily be coerced into being a bool by writing for example "bool(x
>     and y)".
> 
> 
> Resolved Issues
> 
>     (See also the Review section above.)
> 
>     - Because the repr() or str() of a bool value is different from an
>       int value, some code (for example doctest-based unit tests, and
>       possibly database code that relies on things like "%s" % truth)
>       may fail.  It is easy to work around this (without explicitly
>       referencing the bool type), and it is expected that this only
>       affects a very small amount of code that can easily be fixed.
> 
>     - Other languages (C99, C++, Java) name the constants "false" and
>       "true", in all lowercase.  For Python, I prefer to stick with
>       the example set by the existing built-in constants, which all
>       use CapitalizedWords: None, Ellipsis, NotImplemented (as well as
>       all built-in exceptions).  Python's built-in namespace uses all
>       lowercase for functions and types only.
> 
>     - It has been suggested that, in order to satisfy user
>       expectations, for every x that is considered true in a Boolean
>       context, the expression x == True should be true, and likewise
>       if x is considered false, x == False should be true.  In
>       particular newbies who have only just learned about Boolean
>       variables are likely to write
> 
>           if x == True: ...
> 
>       instead of the correct form,
> 
>           if x: ...
> 
>       There seem to be strong psychological and linguistic reasons why
>       many people are at first uncomfortable with the latter form, but
>       I believe that the solution should be in education rather than
>       in crippling the language.  After all, == is general seen as a
>       transitive operator, meaning that from a==b and b==c we can
>       deduce a==c.  But if any comparison to True were to report
>       equality when the other operand was a true value of any type,
>       atrocities like 6==True==7 would hold true, from which one could
>       infer the falsehood 6==7.  That's unacceptable.  (In addition,
>       it would break backwards compatibility.  But even if it didn't,
>       I'd still be against this, for the stated reasons.)
> 
>       Newbies should also be reminded that there's never a reason to
>       write
> 
>           if bool(x): ...
> 
>       since the bool is implicit in the "if".  Explicit is *not*
>       better than implicit here, since the added verbiage impairs
>       redability and there's no other interpretation possible.  There
>       is, however, sometimes a reason to write
> 
>           b = bool(x)
> 
>       This is useful when it is unattractive to keep a reference to an
>       arbitrary object x, or when normalization is required for some
>       other reason.  It is also sometimes appropriate to write
> 
>           i = int(bool(x))
> 
>       which converts the bool to an int with the value 0 or 1.  This
>       conveys the intention to henceforth use the value as an int.
> 
> 
> Implementation
> 
>     A complete implementation in C has been uploaded to the
>     SourceForge patch manager:
> 
>         http://python.org/sf/528022
> 
>     This will soon be checked into CVS for python 2.3a0.
> 
> 
> Copyright
> 
>     This document has been placed in the public domain.
> 
> 
> 
> Local Variables:
> mode: indented-text
> indent-tabs-mode: nil
> fill-column: 70
> End:
> 
> 




From phr-n2002a at nightsong.com  Wed Apr  3 01:08:36 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 22:08:36 -0800
Subject: Python compilers
References: <3CAA9786.2070700@moofed.com>
Message-ID: <7xwuvpxqjf.fsf@ruckus.brouhaha.com>

Rob Hazlewood <rob at moofed.com> writes:
> Would I be right in saying that Python is an interpreted language?

Yes. 

> I know it can be run as an interpreted language, however I seem to
> remember people talking about python compilers.
> 
> Do Python compilers exist? are they efficient?

The current Python implmentations compile Python into either a
special-purpose byte code or into Java bytecode.  The byte code is
then interpreted by an interpreter in C or by a JVM.  There's an
experimental compiler called Psyco that makes machine code directly
and it runs a lot faster than interpreted Python bytecode, but it
doesn't implement the whole language.  It's a promising direction
for development but not generally useable in its current form.

> If there is a python compiler, surely it would be able to generate
> code that is as efficient as a C compiler, resulting in what would be
> the best available OO language? (having simplicity, speed)

It can probably come pretty close, like Lisp compilers do, if you're
willing to accept some limitations on style.

> What is the performance of python like compared to other languages
> such as C and java?

Currently Python programs run a lot slower than C or Java, but that's
usually ok, since you can write them a lot faster than their C or Java
equivalents, and your time is more valuable than the computer's.
Faster Python implementations are always a desirable thing though.


From andymac at bullseye.apana.org.au  Sat Apr  6 20:17:55 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sun, 7 Apr 2002 12:17:55 +1100 (edt)
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
In-Reply-To: <7x1ydt8ve7.fsf@ruckus.brouhaha.com>
Message-ID: <Pine.OS2.4.32.0204071040020.6070-100000@tenring.andymac.org>

On 5 Apr 2002, Paul Rubin wrote:

> Stable, mature software should be able to run for that long without
> needing to be upgraded.  I find just can't expect that from Python.

You appear to be saying that Python incorporates expiry logic, which
causes a particular version to cease functioning correctly after some
specified time.

This of course is complete BS.  Likewise your whole argument is complete
BS in the face of moves by a certain company towards subscription software
with mandated upgrade schedules.

There are OS vendors who choose to imposing grossly out-of-date releases
of 3rd party software on customers for their own selfish reasons.  I
suppose you laud them too, in the face of serious security weaknesses and
the absence of real functionality gains.

At least with OSS, you can preserve your personal "blessed" release longer
than anyone else would consider sane.  Not only that, you can distribute
it with any software you wish to supply someone else, so that you can
avoid having to deal with the compatibility issues.

You've said your piece to the point of being a broken record, and been
informed that your POV is not accepted.  Please swallow future such
utterances so that the positive contributions you've made to this
community don't get lumped into irrelevence along with your whinging.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From logiplexsoftware at earthlink.net  Tue Apr 30 13:33:20 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 30 Apr 2002 10:33:20 -0700
Subject: remove read only flag
In-Reply-To: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>
References: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>
Message-ID: <20020430103320.49a5e140.logiplexsoftware@earthlink.net>

On Tue, 30 Apr 2002 17:18:47 GMT
Francis Meyvis wrote:

> Hoi,
> 
> How can one remove the read-only flag from a file
> (e.g. c:\temp\file.txt) using python in win32.

see os.chmod()

Hm.  I wonder how this maps to Windows permissions...

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From tomchance at gmx.net  Sun Apr 28 19:11:08 2002
From: tomchance at gmx.net (Tom Chance)
Date: Sun, 28 Apr 2002 23:11:08 +0000
Subject: KDE+GNOME Bindings in SuSE 7.3
Message-ID: <3CCC818C.2070505@gmx.net>

I've just started out with Python, having been programming Perl for a 
few years. I've been trying to get an app I downloaded a while ago 
working without success until I found it couldn't import gnome.ui...

I duly downloaded every RPM from the SuSE server I could find to do with 
python, and tried again... no luck. I can't quite figure out what's 
wrong. If I try "import gtk" or "import gnome" or "import kde" or 
"import qt" I just get errors telling me I don't have modules by those 
names.

An rpm query for python gives the following results:

tom at tigger:~> rpm -qa | grep py
gnome-python-1.4.1-127
pygnome-libglade-1.4.2-2
python-tkinter-2.1.1-78
pygtk-libglade-0.6.9-2
pygnome-1.4.2-2
python-tk-2.1.1-78
python-gtk-1.4.1-83
python-2.1.1-45
pyxml-0.6.6-68
python-nothreads-2.1.1-78

So from that I would have thought it would work fine, but it doesn't. 
I've also got the KDE bindings for KDE3.0 installed, and I can't seem to 
get them either.

I'm running SuSE 7.3 btw, if anyone else uses the same bindings on the 
same distro...

Any ideas? I'm at a dead end here...

Tom



From h_schneider at marketmix.com  Mon Apr  8 05:55:24 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Mon, 8 Apr 2002 11:55:24 +0200
Subject: wxPython / Your Experiences in large project ???
Message-ID: <a8rp2u$ht8$02$1@news.t-online.com>

Hi,

what are your experiences, using wxPython/wxWindows in large projects?

Large means in my case a sales system, based on MySQL.
I'm thinking about implementing it in wxPython, but I'm not sure about the
stability of
wxWindows and the wxPython wrapper ...

The alternative way would be a project in Borland Delphi, which has proofed
its
stability over years ...

So what do you think about it!

Thanks a lot,
Harald




From bac at OCF.Berkeley.EDU  Tue Apr  2 18:08:58 2002
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Tue, 2 Apr 2002 15:08:58 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
Message-ID: <Pine.SOL.4.44.0204021501270.5861-100000@death.OCF.Berkeley.EDU>

On Tue, 2 Apr 2002, David Goodger wrote:

> PEP: 287
> Title: reStructuredText Standard Docstring Format
> Version: $Revision: 1.3 $
> Last-Modified: $Date: 2002/04/02 03:50:38 $
> Author: goodger at users.sourceforge.net (David Goodger)
> Discussions-To: doc-sig at python.org
> Status: Draft
> Type: Informational
> Created: 25-Mar-2002
> Post-History: 02-Apr-2002
> Replaces: 216

+1 vote from me!  Seems to have a good balance of implicit formatting
(back-slashing only when ambiguous, indenting code, doctest using >>>) and
verbose (the :method:doStuff() style).

As for the PEP part, it might as well be adopted.  Since anyone who writes
a PEP will most likely have quite a bit of experience writing docstrings
(and would most likely adopt this), they will already be experienced with
the syntax and thus using it for PEP writing would not be a big hurdle.

Now-if-we-could-only-get-Fred-to-put-all-docs-in-the-modules-using-this-that-would-be-great-ly
y'rs,
Brett C.



From mgilfix at eecs.tufts.edu  Tue Apr 23 11:36:03 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 23 Apr 2002 11:36:03 -0400
Subject: non-blocking sockets
In-Reply-To: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>; from g0riaman@cdf.toronto.edu on Tue, Apr 23, 2002 at 03:00:42PM +0000
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>
Message-ID: <20020423113603.B10411@eecs.tufts.edu>

  You need to use the select module to poll the socket. The select
module allows you to specify how long you want your timeout (in this
case, you probably want a timeout of 0, which is simply a poll) and
the module tells you which sockets are waiting for reading/writing.

  import select
  read_sock = <my readable socket>
  w_ready, r_ready, e_ready = select.select ([ ], [ read_sock ], [ ], 0)

  if read_sock in r_ready:
    # Process socket

  ... will probably do what you want.

  Check out the library ref docs for select at:
http://www.python.org/doc/current/lib/module-select.html

                 -- Mike

On Tue, Apr 23 @ 15:00, so very tired wrote:
> When trying to read from a socket, I don't want it hanging, so I set it to
> non-blocking by calling
> 
> setblocking(0)
> 
> but then when I try to read from it and there is no data in the socket, I
> get an error. I tried this in both windows and linux and in both I get an
> error and the program stops. It says there's an exception but it doesn't
> say the name of the exception so I can't even try to catch.
> Does anyone know how I can check if there's data coming from a socket
> connection without hanging or having the program barf on me?
> Thanks.
> mRiaz
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (g0riaman)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From tdelaney at avaya.com  Mon Apr 29 20:40:37 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 30 Apr 2002 10:40:37 +1000
Subject: String handling bug in Python
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2AA@natasha.auslabs.avaya.com>

> From: David Bolen [mailto:db3l at fitlinxx.com]
> 
> Tim Peters <tim.one at comcast.net> writes:
> 
> > Forward slashes are easier for almost everyone to type than backward
> > slashes, and work fine on Windows.  Paths using forward 
> slashes also run
> > 13-17 times faster in Python code, due to Python's Unix 
> heritage <wink>.
> 
> Just as an FYI, forward slashes don't actually work *everywhere* on
> Windows - there are notable UNC names that must use \\ as a prefix and
> won't accept // in the API.  It's not that common but I've been bitten
> by it in the past.
> 
> But they definitely work in any local file API.

import os.path

path = '//computer/path/to/file'
path = os.path.normpath(path)
print path

Tim Delaney




From sholden at holdenweb.com  Thu Apr 11 17:10:21 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 11 Apr 2002 17:10:21 -0400
Subject: Easter Eggs
References: <mailman.1018485864.13992.python-list@python.org> <on6t8.206591$Yv2.67543@rwcrnsc54>
Message-ID: <l%mt8.4625$F43.3338@atlpnn01.usenetserver.com>

"Emile van Sebille" <emile at fenx.com> wrote ...
> Tim Peters
> > We take t-shirts very seriously at
> > PythonLabs.
> >
> > they're-all-backward-compatible-too-ly y'rs  - tim
> >
>
> I don't know about that... mine seems too big, backward compatibly
> speaking...

You're lucky. My particular backward compatibilty problem tends to make old
T-shirts too *small*.

definitely-not-wasting-away-ly y'rs  - steve






From bokr at oz.net  Sun Apr  7 07:00:34 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Apr 2002 11:00:34 GMT
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com> <just-16031B.10170807042002@news1.xs4all.nl>
Message-ID: <a8p8si$5ef$0@216.39.172.122>

On Sun, 07 Apr 2002 10:17:08 +0200, Just van Rossum <just at xs4all.nl> wrote:

>In article <a8nric$sbb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
>wrote:
>
>> The three execution scopes are function local (hereafter referred to as
>> local), module global (hereafter referred to as global), and builtin
>> scope.  Local scope exists any time Python's instruction pointer is
>> inside a function/method.  Global scope refers to the currently
>> executing module; explicitly accessing a function in another module
>> through attributes changes the current module:
>> 
>>     from M import f
>>     f()                 # f()'s global scope is the current module
>>     import M
>>     M.f()               # f()'s global scope is now the module M
>
>No: f()'s global scope is always module M. Scoping is static after all. 
>Or are you saying something different?
>
>Just
 +---< M.py >---
 |a='a bound in M.py'
 |def f(): return a
 +--------------
 >>> import M
 >>> M.__file__
 'M.py'
 >>> M.f()
 'a bound in M.py'
 >>> a='a bound interactively'
 >>> M.f()
 'a bound in M.py'
 >>> from M import f
 >>> f()
 'a bound in M.py'
 >>> f2=M.f
 >>> f2()
 'a bound in M.py'
 >>> a
 'a bound interactively'

Regards,
Bengt Richter


From paul.osman at sympatico.ca  Tue Apr 16 14:36:17 2002
From: paul.osman at sympatico.ca (Paul Osman)
Date: Tue, 16 Apr 2002 14:36:17 -0400
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca> <a9hjvc$cdq$1@as201.hinet.hr> <mailman.1018979135.19771.python-list@python.org> <m2wuv7ijsa.fsf@asfast.com>
Message-ID: <3CBC6F21.3050305@sympatico.ca>

Lloyd Zusman wrote:

>>[ ... ]
>>
>><code>
>>num = 5
>>def dummy():
>>    newnum = num + 5
>>    return num
>></code>
>>
>>[ ... ]
>>
>
>I think that you meant to type this:
>
>    return newnum
>
I did indeed... thanks for catching that, the example would be rather 
useless otherwise. :)

Cheers,
Paul Osman





From bokr at oz.net  Wed Apr  3 23:16:06 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Apr 2002 04:16:06 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <3CA62CD8.6ABD44D0@accessforall.nl> <m3adsp8zie.fsf@mira.informatik.hu-berlin.de> <3CA6F012.E2631003@accessforall.nl> <m3adspm1kl.fsf@mira.informatik.hu-berlin.de> <3CA79401.CFA41B65@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain> <a8akqh$8bk$0@216.39.172.122>
Message-ID: <a8gk26$b16$0@216.39.172.122>

On 1 Apr 2002 21:52:17 GMT, bokr at oz.net (Bengt Richter) wrote:
[...a post to which Kris replied...]

Unfortunately my newsfeed cycles too damn fast, so I have to reply to Kris
by replying to myself, after reading from the archive and pasting as quoted
from there:

>Kris J. Zaragoza kzaragoza at attbi.com 
>Mon, 01 Apr 2002 22:38:39 GMT 

>In article <a8akqh$8bk$0 at 216.39.172.122>, Bengt Richter wrote:
>> Of course the 'hidden property' is only rhetoric here, but why not
>> make this a clear and controllable unified mechanism instead of a
>> collection of special effects? I would like to see an overridable
>> __bool__ method (not actually a programmatic property) for the base
>> object class, perhaps making deprecation of __nonzero__ possible.

>Now this is an idea with some merit.  I can see some real advantages
>to being able to define the truth value of a given object, or at least
>control its conversion to a bool.  But then, Guido's PEP already shows
>this as possible when creating a bool.  If you look at the __new__()
>constructor, it checks the truth value of the object when deciding
>which global value to return.  If this is how things are implemented
>in the C code (and I haven't checked), then __nonzero__() (or
>__len__()) is called if defined.  This gives us the control we're
>looking for.

Well, I wouldn't want to mess with __len__ if I e.g., were subtyping str
for something where I wanted to be able to write

    if s: print '"%s" (length %d) is special.' % (s, len(s))

I.e., in a logical context( if x..., x and ..., etc.), I think x.__bool__()
should have priority over x.__nonzero__() or x.__len__().

>> Well, what do you think of the above?

>It certainly clarifies matters a bit, but I still have to wonder
>whether it's worth the effort.  The way I see it, the mechanism for
>the customizations you describe is already present, albeit badly
>named.
I'm suggesting there may be cases where e.g., __len__() should still
function to control the range of an iteration, but __bool__() might
exist at the same time to signify some other condition.

>I certainly don't object to implementing the bool type.  I guess I
>just don't see it as a significant advantage.

I'd like the semantic clarity. And I'd like to define a class or subtype whose
boolean logical value doesn't depend on anyones concept of Nothing or Empty
if those don't fit. How about has-converged as a boolean meaning associated
with some object in a worker thread? Do I have to say the status is
empty of convergedness? Or that is has non-zero convergence?

Say I don't want a measure, I just want to know if the data is ready to use
when I poll. Or take a whimsical example: What if I want to subclass int and
be able to say if n: print n,'is a prime', expecting that n.__bool__()
will be called so I can calculate primeness and cache the result in a
boolean attribute for faster return next time, etc.

The point is to have freedom to use a useful general mechanism without
semantic contortions, or giving up the ability if __nonzero__ and/or __len__
already have uses for the class.

Regards,
Bengt Richter


From jeremy at alum.mit.edu  Tue Apr  9 12:19:52 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 9 Apr 2002 09:19:52 -0700
Subject: changing local dictionnary for functions
References: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr> <pan.2002.04.08.18.45.23.620630.19669@cantal.lip6.fr>
Message-ID: <b0f083db.0204090819.66be16bd@posting.google.com>

Boris Boutillier <boris at cantal.lip6.fr> wrote in message news:<pan.2002.04.08.18.45.23.620630.19669 at cantal.lip6.fr>...
> To be more precise, I'm looking for this kind of function :
> execfunction(func,*args,globals,locals)
> 
> func is a function
> *args : is a list of arguments to pass to the function
> globals is a dictionnary (or subclass of it) being used as global dict
> locals is a dictionnary (or subclass of it) being used as local dict
> 
> I think this kind of function can be written by playing with the
> func.func_code bytecode, but before trying to do it this way I'd like to
> be sure there is no simple way to do it.

I think the table in the docs is mistaken.  I can't think of any way
to call a function and cause it to use a different namespace for its
locals.  The implementation of local namespaces is fairly fundamental;
the namespace isn't even implemented as a dictionary.

You can exec the bytecode of the func (func.func_code) using a custom
locals dictionary, but there are two significant limitations:  The
function can't take any arguments and its return value is swallowed by
the exec statement.

Also, be careful if you pass a subclass of dict to exec.  The
interpreter uses the PyDict API which operates on the concrete
representation of the dict rather than the abstract API.  For example,
if your subclass overrides __getitem__() or __setitem__(), those
methods won't be called.

I'm having trouble imagining what you are really trying to do, so I
can't be more helpful.  In general, when I do very dynamic things with
namespaces, I use classes and instances because the language provides
more user control over attribute lookup than name lookup.

Jeremy


From sjmachin at lexicon.net  Sat Apr 27 07:05:13 2002
From: sjmachin at lexicon.net (John Machin)
Date: 27 Apr 2002 04:05:13 -0700
Subject: Argh, I give up.
References: <mailman.1019863308.13671.python-list@python.org>
Message-ID: <c76ff6fc.0204270305.7be129dc@posting.google.com>

Gustavo Cordova <gcordova at hebmex.com> wrote in message news:<mailman.1019863308.13671.python-list at python.org>...
> What can I do to make MingW32
> properly understand python's *.lib files, which are needed for
> linking?
> 
> 

If you aren't using distutils, you should be. Read the manual:

http://www.python.org/doc/current/inst/non-ms-compilers.html

and follow the pointers at the bottom of the page. You will need to
create a file called libpython22.a (that MinGW understands) and put it
in the libs directory of your Python binary distribution (the one
where python22.lib resides).

Here is another link which may now be only of historical interest; I
note that Robert hasn't updated it since 1999.

http://starship.python.net/crew/kernr/mingw32/Notes.html


From aahz at pythoncraft.com  Thu Apr  4 16:24:20 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 16:24:20 -0500
Subject: Threading problem
References: <mailman.1017931757.25464.python-list@python.org>
Message-ID: <a8iga4$18b$1@panix1.panix.com>

In article <mailman.1017931757.25464.python-list at python.org>,
Gustavo Cordova  <gcordova at hebmex.com> wrote:
>
>Yeah, sounds like what's been happening to the Linux kernel
>in the last iterations; locking has become steadily finer-grained,
>so that multiprocessor setups will scale better.
>
>I also believe that Python, eventually, will have to follow
>that same path.

Why?  What's wrong with just making multiple Python processes work
better together?  Or with writing extensions that release the GIL?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From achim.domma at syynx.de  Mon Apr 15 04:27:41 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Mon, 15 Apr 2002 10:27:41 +0200
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs>
Message-ID: <a9e2g8$nhh$00$1@news.t-online.com>

Hi,

"Max Ischenko" <max at malva.com.uaREMOVE.IT> wrote in message
news:of1e9a.4vg.ln at cvs...
> I'd like to have the algorithm moved out from Python source.
> I've heard about XSLT language that is used to describe transformation of
> XML data into HTML. I'm seeking for something similar. The difference is
> that target language is C++ and not HTML.
>
> Any suggestions?

XSLT can transform XML into any kind of text, so you are not restricted to
HTML. In fact, XSLT seems to be the solution you are looking for.

greetings
Achim




From maxm at mxm.dk  Sun Apr  7 09:27:59 2002
From: maxm at mxm.dk (Max M)
Date: Sun, 07 Apr 2002 15:27:59 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk>
Message-ID: <3CB0495F.7010507@mxm.dk>

Ralph Corderoy wrote:

> I think we all perfectly understand your opinion too and no matter how
> many times you repeat it there will still be some of us that like to do
> 
>     a = b + is_movable(c


I can see that it is clever, but I don't get why it is smart. Isn't it 
just another way of writing?:

     if is_movable(c):
         a = b + 1

If it is, I find that it is more obfuscation than expression. It's akind 
to using a magic value. My paraphrase of our code should really read 
something like::

     motionLength = 1
     if is_movable(c):
         a = b + motionLength

Or your code should possibly have read::

     a = b + motionLength(c)

But probably I misunderstand you intentions.

regards Max M



From sqrt_2  Sun Apr 14 14:19:56 2002
From: sqrt_2 (sqrt_2)
Date: Sun, 14 Apr 2002 18:19:56 GMT
Subject: Awesome wxPython IDE
Message-ID: <3cb9c75d.115396601@news.mweb.co.za>

For those of you who develop apps using wxPython there is this great
product I know of.
It is called Boa-Constructor and also incorporates Zope.
Have a look at:
http://boa-constructor.sourceforge.net

Enjoy!!




From aleax at aleax.it  Thu Apr 18 15:27:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 19:27:36 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com>
Message-ID: <I6Fv8.36158$vF6.1093396@news2.tin.it>

Donald McCarthy wrote:

> I would like to alter the way python displays what a class is. I can do
> this for an instance of a class using __repr__ but I don't know how to do
> this for a class.

You do it by defining a custom metaclass.  Python 2.2 of course:

class FunkyReprMeta(type):
    def __repr__(cls):
        return "My wonderful class %s" % cls.__name__

class FunkyReprObject:
    __metaclass__ = FunkyReprMeta


class C(FunkyReprObject):
    "whatever you want here"


print C


this emits "My wonderful class C" and has otherwise just the semantics
of newstyle classes.  A class is an instance of its metaclass, after all, so
that's how you define how/what happens when a class object gets
printed, added to something, etc -- define special methods in its
metaclass.  Instead of deriving from FunkyReprObject, by the way,
you may prefer to define __metaclass__ in the class-body of C directly,
or as a module-level global variable.


Alex



From lenssen at hitnet.rwth-aachen.de  Fri Apr 26 16:30:51 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Fri, 26 Apr 2002 22:30:51 +0200
Subject: Simple text-editor in Python?
Message-ID: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>

Is there a simple, open source text-editor project available in Python?




From hgs at dmu.ac.uk  Thu Apr 18 11:11:03 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 16:11:03 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4g01t6oxp.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181557541.24683-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > I can't yet: I don't know which properties are reversed:
> >  * Modules included in the config.c file
> >  * List of objects to be added to library archive
> >  * linker options added to linker options
> >  * Rules added to Makefile
> >  * Names collected in SHAREDMODS variable.
>
> The state of 'will the following modules be static or dynamic'. In a
> Setup file, you can have
>
        [*shared*, *static* intereaved]

But to change the wording one needs to know which of the above
properties are reversed, and how commenting out modules affects them in
each case.
> ...
>
> > > Use 'gcc --print-prog-name=ld' to find out which ld binary gcc uses.
> > /usr/local/sparc-sun-solaris2.7/bin/ld
> > /usr/local/sparc-sun-solaris2.7/bin/ld --version gives
> >
> > GNU ld version 2.12
> > Copyright 2002 Free Software Foundation, Inc.
>
> Ok; that explains it: configure does
>
>    $CC -Xlinker -V 2>&1 | grep BFD

gcc -Xlinker -V on my system gives:
Script started on Thu 18 Apr 2002 16:07:12 BST
neelix hgs 11 %> gcc -Xlinker -V
/usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.3/crt1.o: In function `_start':
/usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.3/crt1.o(.text+0x5c): undefined reference to `main'
GNU ld version 2.12
  Supported emulations:
   elf32_sparc
   elf64_sparc
collect2: ld returned 1 exit status
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Thu 18 Apr 2002 16:07:52 BST
>
> expecting that the linker output reads like
>
> GNU ld version 2.11.2 (with BFD 2.11.2)
>   Supported emulations:
>    elf32_sparc
>    elf64_sparc
>
> Apparently, in binutils 2.12, GNU ld stops printing BFD in its
> output. The proper fix is to invoke --help instead, and to see whether
> export-dynamic appears in the output.
output from :  /usr/local/sparc-sun-solaris2.7/bin/ld --help  is [rather
long!]:

Usage: /usr/local/sparc-sun-solaris2.7/bin/ld [options] file...
Options:
  -a KEYWORD                  Shared library control for HP/UX compatibility
  -A ARCH, --architecture ARCH
                              Set architecture
  -b TARGET, --format TARGET  Specify target for following input files
  -c FILE, --mri-script FILE  Read MRI format linker script
  -d, -dc, -dp                Force common symbols to be defined
  -e ADDRESS, --entry ADDRESS Set start address
  -E, --export-dynamic        Export all dynamic symbols
  -EB                         Link big-endian objects
  -EL                         Link little-endian objects
  -f SHLIB, --auxiliary SHLIB Auxiliary filter for shared object symbol table
  -F SHLIB, --filter SHLIB    Filter for shared object symbol table
  -g                          Ignored
  -G SIZE, --gpsize SIZE      Small data size (if no size, same as --shared)
  -h FILENAME, -soname FILENAME
                              Set internal name of shared library
  -I PROGRAM, --dynamic-linker PROGRAM
                              Set PROGRAM as the dynamic linker to use
  -l LIBNAME, --library LIBNAME
                              Search for library LIBNAME
  -L DIRECTORY, --library-path DIRECTORY
                              Add DIRECTORY to library search path
  -m EMULATION                Set emulation
  -M, --print-map             Print map file on standard output
  -n, --nmagic                Do not page align data
  -N, --omagic                Do not page align data, do not make text readonly
  -o FILE, --output FILE      Set output file name
  -O                          Optimize output file
  -Qy                         Ignored for SVR4 compatibility
  -q, --emit-relocs           Generate relocations in final output
  -r, -i, --relocateable      Generate relocateable output
  -R FILE, --just-symbols FILE
                              Just link symbols (if directory, same as --rpath)
  -s, --strip-all             Strip all symbols
  -S, --strip-debug           Strip debugging symbols
  -t, --trace                 Trace file opens
  -T FILE, --script FILE      Read linker script
  -u SYMBOL, --undefined SYMBOL
                              Start with undefined reference to SYMBOL
  --unique [=SECTION]         Don't merge input [SECTION | orphan] sections
  -Ur                         Build global constructor/destructor tables
  -v, --version               Print version information
  -V                          Print version and emulation information
  -x, --discard-all           Discard all local symbols
  -X, --discard-locals        Discard temporary local symbols (default)
  --discard-none              Don't discard any local symbols
  -y SYMBOL, --trace-symbol SYMBOL
                              Trace mentions of SYMBOL
  -Y PATH                     Default search path for Solaris compatibility
  -(, --start-group           Start a group
  -), --end-group             End a group
  -assert KEYWORD             Ignored for SunOS compatibility
  -Bdynamic, -dy, -call_shared
                              Link against shared libraries
  -Bstatic, -dn, -non_shared, -static
                              Do not link against shared libraries
  -Bsymbolic                  Bind global references locally
  --check-sections            Check section addresses for overlaps (default)
  --no-check-sections         Do not check section addresses for overlaps
  --cref                      Output cross reference table
  --defsym SYMBOL=EXPRESSION  Define a symbol
  --demangle [=STYLE]         Demangle symbol names [using STYLE]
  --embedded-relocs           Generate embedded relocs
  -fini SYMBOL                Call SYMBOL at unload-time
  --force-exe-suffix          Force generation of file with .exe suffix
  --gc-sections               Remove unused sections (on some targets)
  --no-gc-sections            Don't remove unused sections (default)
  --help                      Print option help
  -init SYMBOL                Call SYMBOL at load-time
  -Map FILE                   Write a map file
  --no-define-common          Do not define Common storage
  --no-demangle               Do not demangle symbol names
  --no-keep-memory            Use less memory and more disk I/O
  --no-undefined              Allow no undefined symbols
  --allow-shlib-undefined     Allow undefined symbols in shared objects
  --no-warn-mismatch          Don't warn about mismatched input files
  --no-whole-archive          Turn off --whole-archive
  --noinhibit-exec            Create an output file even if errors occur
  -nostdlib                   Only use library directories specified on
				the command line
  --oformat TARGET            Specify target of output file
  -qmagic                     Ignored for Linux compatibility
  --relax                     Relax branches on certain targets
  --retain-symbols-file FILE  Keep only symbols listed in FILE
  -rpath PATH                 Set runtime shared library search path
  -rpath-link PATH            Set link time shared library search path
  -shared, -Bshareable        Create a shared library
  --sort-common               Sort common symbols by size
  --spare-dynamic-tags COUNT  How many tags to reserve in .dynamic section
  --split-by-file [=SIZE]     Split output sections every SIZE octets
  --split-by-reloc [=COUNT]   Split output sections every COUNT relocs
  --stats                     Print memory usage statistics
  --target-help               Display target specific options
  --task-link SYMBOL          Do task level linking
  --traditional-format        Use same format as native linker
  --section-start SECTION=ADDRESS
                              Set address of named section
  -Tbss ADDRESS               Set address of .bss section
  -Tdata ADDRESS              Set address of .data section
  -Ttext ADDRESS              Set address of .text section
  --verbose                   Output lots of information during link
  --version-script FILE       Read version information script
  --version-exports-section SYMBOL
                              Take export symbols list from .exports, using
				SYMBOL as the version.
  --warn-common               Warn about duplicate common symbols
  --warn-constructors         Warn if global constructors/destructors are seen
  --warn-multiple-gp          Warn if the multiple GP values are used
  --warn-once                 Warn only once per undefined symbol
  --warn-section-align        Warn if start of section changes due to alignment
  --fatal-warnings            Treat warnings as errors
  --whole-archive             Include all objects from following archives
  --wrap SYMBOL               Use wrapper functions for SYMBOL
  --mpc860c0 [=WORDS]         Modify problematic branches in last WORDS (1-10,
				default 5) words of a page
/usr/local/sparc-sun-solaris2.7/bin/ld: supported targets: elf32-sparc elf64-sparc a.out-sunos-big elf64-little elf64-big elf32-little elf32-big srec symbolsrec tekhex binary ihex
/usr/local/sparc-sun-solaris2.7/bin/ld: supported emulations: elf32_sparc elf64_sparc
/usr/local/sparc-sun-solaris2.7/bin/ld: emulation specific options:
elf32_sparc:
  -Bgroup		Selects group name lookup rules for DSO
  --disable-new-dtags	Disable new dynamic tags
  --enable-new-dtags	Enable new dynamic tags
  --eh-frame-hdr	Create .eh_frame_hdr section
  -z combreloc		Merge dynamic relocs into one section and sort
  -z defs		Disallows undefined symbols
  -z initfirst		Mark DSO to be initialized first at runtime
  -z interpose		Mark object to interpose all DSOs but executable
  -z loadfltr		Mark object requiring immediate process
  -z nocombreloc	Don't merge dynamic relocs into one section
  -z nocopyreloc	Don't create copy relocs
  -z nodefaultlib	Mark object not to use default search paths
  -z nodelete		Mark DSO non-deletable at runtime
  -z nodlopen		Mark DSO not available to dlopen
  -z nodump		Mark DSO not available to dldump
  -z now		Mark object non-lazy runtime binding
  -z origin		Mark object requiring immediate $ORIGIN processing
			  at runtime
  -z KEYWORD		Ignored for Solaris compatibility
elf64_sparc:
  -Bgroup		Selects group name lookup rules for DSO
  --disable-new-dtags	Disable new dynamic tags
  --enable-new-dtags	Enable new dynamic tags
  --eh-frame-hdr	Create .eh_frame_hdr section
  -z combreloc		Merge dynamic relocs into one section and sort
  -z defs		Disallows undefined symbols
  -z initfirst		Mark DSO to be initialized first at runtime
  -z interpose		Mark object to interpose all DSOs but executable
  -z loadfltr		Mark object requiring immediate process
  -z nocombreloc	Don't merge dynamic relocs into one section
  -z nocopyreloc	Don't create copy relocs
  -z nodefaultlib	Mark object not to use default search paths
  -z nodelete		Mark DSO non-deletable at runtime
  -z nodlopen		Mark DSO not available to dlopen
  -z nodump		Mark DSO not available to dldump
  -z now		Mark object non-lazy runtime binding
  -z origin		Mark object requiring immediate $ORIGIN processing
			  at runtime
  -z KEYWORD		Ignored for Solaris compatibility

Report bugs to bug-binutils at gnu.org
>
> > I think I will delay doing that for now, because most of the tests
> > worked without using GCC.  Another post suggests that this is not common
> > to Solaris7, (referring to pre-releases of Python-2.2.1).  Therefore I
> > may not really be helping anyone else by exploring that.
>
> It turns out that you are the first one to use binutils 2.12 on
> Solaris, trying to build Python with that. So you are likely helping
> others; Python 2.2.2 (and perhaps 2.1.4) will have this problem
> corrected.

So the fastest fix is for me to doengrade binutils?
Can I provide any further diags at this point which might be helpful?
>
> Regards,
> Martin
>
        Thank you,
        Hugh
>






From siegfried.gonzi at kfunigraz.ac.at  Thu Apr 11 04:56:17 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 11 Apr 2002 10:56:17 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at> <slrnab8qb5.ec2.philh@comuno.freeserve.co.uk>
Message-ID: <3CB54FB1.5E3A6C90@kfunigraz.ac.at>

phil hunt wrote:
 
> That's a fair enough attitude if all you are going to do is write
> the odd 200-line script.
> 
> If you are writing larger programs, it rapidly becomes
> self-defeating.

There is an upper limit of code amount which a scientist (normally)
can/would handle. 3000 lines of code is really huge for me. A
professional would laugh.

But I am not sure whether it is a good idea to go with C for a large
project (up to 10000 and more lines of code). Yesterday I was confronted
with a Mie-calculation code which is implemented in C; and you then know
the story with post-fence errors and the like; and that you can access,
in C, every array object even it was never intented for. Surely, nobody
would ever deliberately to this in C, but it happens (and this all the
time).

Though it is possible to write very,very robust code C (for example
Yorick).


S. Gonzi


From lumber at jack.com  Wed Apr 10 00:22:55 2002
From: lumber at jack.com (Lumberjack)
Date: 10 Apr 2002 04:22:55 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org>
Message-ID: <Xns91EBD90A7449Dlumberjack@216.148.53.83>

Tim Peters <tim.one at comcast.net> wrote:
> PEP 285 taught us that people don't agree, and that some people
> disagree in disagreeable ways.  Frankly, that wasn't news,

Amazing to see so much slow learning taking place. So why don't you all do 
it again some time. Like the next PEP and the next one and the next one, 
and so on ad infinitum. Just like previous PEPs. The old Chinese water 
torture would be more humane than continuing these unending so-called 
"enhancements" to the language. Drip drip drip... Neither you nor Guido 
have a clue; unless this is his secret plan to destroy the appeal of Python 
by setting Python programmers at each other's throats, or they begin to 
associate emotional discord with the language and decide to keep their 
distance. If so, the plan is well executed.

People aren't going to change, so Guido needs to adjust to the reality of 
the political consequences of his current strategy.


From bokr at oz.net  Tue Apr  9 20:07:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 00:07:38 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <a8vvoa$3g8$0@216.39.172.122>

On Tue, 9 Apr 2002 19:23:18 GMT, Andrew Koenig <ark at research.att.com> wrote:

>I would like to be able to write an expression that yields an
>object that I can subsequently use to rebind a name that I mention
>only in that expression.  How do I do it?
>
>My motivation is to be able to write something that is analogous
>to call-by-reference in C++.
>
>This question, as phrased, is probably a little hard to understand;
>here are some examples that might help.
>
>Here's a function definition:
>
>        def f(v):
>                global x
>                x = v
>
>Now, f is an object that I can use to rebind the name x.  For
>example, if I execute f(42), that sets x to 42.
>
>However, this example doesn't do what I want because the
>``expression'' in this example would have to be f, not the definition
>of f (which isn't an expression).  The name x is not mentioned
>anywhere in the expression.  Therefore, each time I want to affect a
>different variable this way, I have to define a new function.
>
>What I would really like is to be able to define a function called,
>say, `set' such that set(x, 42) has the same effect as `x = 42'.
>
>I can't use a lambda expression, because they're not allowed to
>contain assignments.
>
>So far, the closest I've been able to come is this:
>
>        def set(var, val):
>                var[:] = [val]
>                
>It requires that the variable in question be primed by giving it
>a list as its value, and detects failure to do so:
>
>        set(x, 42)              # error -- x is not a list
>        x = []
>        set(x, 42)              # x is now [42]
>
>Is it possible to do better (i.e., to solve this problem more succinctly)?
>
You've probably thought of this (using a class instance instead of []),
but maybe commenting on it will help us see where you're going:

 >>> class Ref:
 ...     def __init__(self, thing): self.thing = thing
 ...     def __call__(self, *newthing):
 ...         if newthing: self.thing = newthing[0]
 ...         return self.thing
 ...     def __repr__(self): return repr(self.thing)
 ...     __str__ = __repr__
 ...
 >>> x = Ref(42)
 >>> x
 42
 >>> x, x(23), x
 (23, 23, 23)

(Function eval got priority, unlike order in print args):

 >>> print x, x(42), x
 23 42 42
 >>> def foo(y):
 ...     y('set by foo')
 ...     return 'returned by foo'
 ...
 >>> x
 42
 >>> foo(x)
 'returned by foo'
 >>> x
 'set by foo'
 >>> x(100)
 100
 >>> x()+5
 105
 >>> x
 100
 >>> x+5
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: unsupported operand types for +: 'instance' and 'int'

I.e,. it's the str/repr methods of the Ref instance x that are showing the
printed results, not a automatic retrieval of the thing values followed by
str conversion, so you have to write x() to get at the actual value or
x(newvalue) to set and also get it.

Of course you can also write methods for x to participate in various operator
expressions besides being called with ().

Regards,
Bengt Richter


From pyth at devel.trillke.net  Tue Apr 16 19:21:42 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 17 Apr 2002 01:21:42 +0200
Subject: Simulating multi-dim array problem - OR - reference confusions
In-Reply-To: <3CBC991C.7090202@nyu.edu>
References: <3CBC991C.7090202@nyu.edu>
Message-ID: <20020416232142.GC29479@devel.trillke>

> 	I'm doing code that needs the functionality of multi-dimensional 
> 	arrays. I'm getting around this right now by using lists of lists, which 
>  seems to work right.  For instance:
> 
> >>> a = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
> ...
> b = a[0]
>
> 	That b is actually a reference to the appropriate list in a, which 
> 	is just what I need.  

actually speaking in python semantics 

	b=a[0]

means that the "name 'b' is bound to 'a[0]'". Internally 
the reference count of the object a[0] is incremented.

the difference is that after you change b with e.g.

  b = ['bert'] 

a[0] will not have changed. You just changed the binding
of the name 'b' again -- not touching 'a' in any way
(except that the reference count of a[0] is decremented)! 

> [x, y, z, <PTR>]
> 	where <PTR> is set the way I described above, with b = a[0], or, in 
> 	this case, a[0][3] = a[10], etc.
> 
> 	Here's are my questions:
> 
> 	1) Am I going to get into trouble assuming (using the example above) 
> 	that b is basically a reference to the sublist a[0]?  I'm not clear how 
> sharing works in python.

see above. if you only work with the "binding" (not really sharing!) like

  b[i]=...  

then you are fine. 
 
> 	2) Is there a way I can print out the address of a "reference"?  In 
> particular I want some concise symbolic representation - like a memory 
> address, or a handle id, or SOMETHING - and not a verbatim 
> reconstruction of the list structure which is how python's "print" 
> function does it.  If I have elements containing references to elements 
> containing references to elements it's impossible to follow what's going 
> on with the chain.  I'd like to have it print something like:
> 
> [[x, y, z, 0x123123], [a, b, c, 0x123958], [q, w, e, 0x358481]]

you might want to work with a wrapper around rows to solve the two
issues:

class mrow:
    "read/write-wrapper for a row of a list"
    def __init__(self, mlist, row):
        self.mlist = mlist
        self.row   = row

    def get(self):
    	" get a reference 
        return self.mlist[self.row]

    def set(self, arg):
        self.mlist[self.row]=arg
        return self.get()

    def __repr__(self):
	" what you see when you print it """
        return 'row %d' % (self.row)

you can use it like this:

>>> li=[ [1,2,3], [4,5,6], [7,8,9]]  # as usual
>>>
>>> b = mrow(li,2)  
>>> print b
row 2
>>>
>>> for row in li:
...     row.append(b) # append the row reference
...
>>> print li
[[1, 2, 3, row 2], [4, 5, 6, row 2], [7, 8, 9, row 2]]
>>>
>>> print li[0][3].get()
[7, 8, 9, row 2]
>>>
>>> li[0][3].set( [42,23,5,mrow(li,0)])
[42, 23, 5, row 0]
>>> print li
[[1, 2, 3, row 2], [4, 5, 6, row 2], [42, 23, 5, row 0]]
>>>

hope that's understandable and what you want. if not just say so...

	holger




From wolfson at uchicago.edu  Sat Apr 27 18:15:44 2002
From: wolfson at uchicago.edu (Ben Wolfson)
Date: Sat, 27 Apr 2002 17:15:44 -0500
Subject: Python's Lisp heritage
References: <mailman.1019922475.8361.python-list@python.org> <tyfsn5h13mh.fsf@pcitapi22.cern.ch> <ZADy8.70845$vF6.2115921@news2.tin.it>
Message-ID: <pan.2002.04.27.17.15.43.188185.9182@uchicago.edu>

On Sat, 27 Apr 2002 15:10:33 -0500, Alex Martelli wrote:

> acek Generowicz wrote:
>         ...
>> The purpose of the parentheses (in this context) is to call functions
>> and associate parameters with those function calls. Python has a not
>> too dissimilar use of parentheses:
> 
> Anyone for Haskell's f a b c ...?  Now THAT is clean and elegant... the
> concept of "a function with > 1 parameter" is unnecessary (as Haskell
> Curry pointed out), so there's no need to associate a function with any
> more than its first (and only) parameter: f a returns an unnamed
> function which takes argument b, returning another unnamed function
> which takes argument c.  Look, ma, no parentheses -- and no tricks
> needed to achieve that, just some deep mathematical reality:-).

No parentheses in the simple case, maybe, but at least in SML, which is
similar, the parentheses can multiply pretty quickly if one of the
parameters is the result of a function, or a function returns a function:
(f a (g b)) (h c) isn't all that pretty, unless you like Lisp.  Haskell
may be smarter about that kind of thing, though.

-- 
BTR    
BEN WOLFSON HAS RUINED ROCK MUSIC FOR A GENERATION
 -- Crgre Jvyyneq


From Chris.Barker at noaa.gov  Thu Apr 25 16:15:30 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 25 Apr 2002 13:15:30 -0700
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at> <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com> <3CC80EBE.E9735284@kfunigraz.ac.at>
Message-ID: <3CC863A6.C7857C25@noaa.gov>

Siegfried Gonzi wrote:
> it even includes an external mie
> calculation program (written in C)

Is this being called as a Python extension? My gues is your memory leak
is right there.

Other's suggestions that you run each instance with a separate startup
of the interpreter will probalby make your program usable, but you still
seem to have a substantial memory leak, which really is very rare in
Python. 

If it's not too long, post you code and we'll see what we can find. Or,
try to shrink it down to a little code as possible, and you'll probably
find the problem, and if not post what code is left here.

-Chris


From TBrixi at CSAS.cz  Wed Apr  3 05:46:49 2002
From: TBrixi at CSAS.cz (=?iso-8859-2?Q?Brixi_Tom=E1=B9?=)
Date: Wed, 3 Apr 2002 12:46:49 +0200 
Subject: Python compilers
Message-ID: <414401922FF9D011BEC600805FEA07E301A865E9@XI0SXJ01>


-----Original Message-----
From: Peter Hansen [mailto:peter at engcorp.com]
Sent: Wednesday, April 03, 2002 8:29 AM
To: python-list at python.org
Subject: Re: Python compilers


Peter Hansen wrote:
> 
> > What is the performance of python like compared to other languages such
> > as C and java?
> 
> Similar to Java, ...

Paul's response reminds me of the existence of just-in-time
and native compilers for Java, which clearly run much faster
than vanilla Java running on a virtual machine.  I meant
"similar to Java" only in the sense of the original Java.

-Peter


BTW

Sun ships Hotspot JIT with their jdk/jre as a default VM
sice Java 2 platform was introduced I think

Tom




From boud at valdyas.org  Tue Apr  2 03:16:24 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 02 Apr 2002 07:16:24 -0100
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl> <1d6q8.118888$Ou.77410@atlpnn01.usenetserver.com>
Message-ID: <3ca940d3$0$32314$e4fe514c@dreader4.news.xs4all.nl>

Steve Holden wrote:

> Congratulations! I remember your original posts asking for good examples
> and choices for the graphics subnsystem. You must feel a weight has been
> lifted ...
> 

I do. It has certainly been an interesting experience. I was wrapping up the
last of the text last year, when there was this thread about writing books
on comp.lang.python. It felt very to the point.

Anyway, I'm quite satisfied with the results :-).

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


From frithiof.jensen at removethis.ted.ericsson.dk  Thu Apr  4 03:57:41 2002
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 4 Apr 2002 10:57:41 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org>
Message-ID: <1017910682.648791@robert>

"Jeremy Bowers" <newsfroups at jerf.org> wrote in message
news:3CABDF6B.5020808 at jerf.org...

I agree totally -

The only glimmer of hope would be to capture the design with an UML tool,
that understands both Java and Python and then reimplement -
http://www.objectdomain.com/index.html is one - but....but...but... it is
probably easier to just capture the requirements and start afresh :)




From cpbotha at crabtree.cpbotha.net  Tue Apr  2 05:37:45 2002
From: cpbotha at crabtree.cpbotha.net (Charl P. Botha)
Date: Tue, 02 Apr 2002 10:37:45 GMT
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <a84561$p5g$1@wanadoo.fr> <slrnaabbct.npb.cpbotha@crabtree.cpbotha.net> <4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com> <20020401020912.GE333@lilith.hqd-internal> <mailman.1017681605.21669.python-list@python.org> <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net> <slrnaaivjc.565.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <slrnaaj2o9.a62.cpbotha@crabtree.cpbotha.net>

In article <slrnaaivjc.565.zeitlin at seth.lpthe.jussieu.fr>, 
Vadim Zeitlin wrote:
> On Mon, 01 Apr 2002 19:04:16 GMT, Charl P. Botha <cpbotha at i_triple_e.org> wrote:
>> Try getting a running wxPython installation on SGI Irix or SUN Solaris for
>> instance.
> 
>  And just what exactly is the problem with it? I don't have access to SGI but
> it did build without any problems on a Solaris box here. If you have
> encountered any problems, have you reported them to us (wxWindows developers)
> as bugs? I'm sorry to say this, but you have to consider spending some of your
> time on writing (useful) bug reports - without the users help there is
> absolutely no chance that wxPython or wxGTK builds out of the box on the
> platforms we don't have access to. It's amazing by how much the number of
> complains about the difficulties of building wxGTK exceeds the number of the
> bug reports (of which there are maybe 1 or 2 right now).

Please read the whole thread before replying.  Also try to reply with less
accusatory language.  I did NOT say that wxWindows doesn't build on these
architectures.  To clarify, and I've said this previously in this thread, I
have the highest regard for wxWindows and wxPython.

What I DID say was that getting a working wxPython installation on one of
those boxes can take quite a while.  One has to get GTK and all the packages
it depends on and configure and build all of these.  Then one has to
configure and build wxGTK and wxPython... all of this can take some time.

I'm not complaining about this.  I was mentioning it to illustrate that on
average, Tkinter is still the fastest to get up and running on all platforms
and that that was part of the reason that it was the default GUI of Python.

I hope this was clear enough,

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From aleax at aleax.it  Mon Apr 29 08:36:38 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 12:36:38 GMT
Subject: Update locals()
References: <20020427195725.GA16217@devel.trillke> <20020428205757.GH16217@devel.trillke> <02042900255602.25256@arthur> <mailman.1020042230.9019.python-list@python.org>
Message-ID: <q7bz8.76797$vF6.2277789@news2.tin.it>

<posted & mailed>

holger krekel wrote:
        ...
> OF COURSE i didn't mean that re.compile mucks with the local namespace.

Right, and THAT is the issue: exec-without-explicit-dicts DOES muck
with the local namespace in ways that cannot be controlled and thus
lead to unmaintainable code, hard-to-find bugs, and so on.


> re.compile compiles a pattern into a pattern object which
> can be used for searches. In the above case the helper
> uses exec to dynamically construct a function object from
> a pattern. That's why i said 'similar' technique. Why are
> you often assuming the worst of all interpretations?

Because the ONLY important issue here is whether having your local
namespace uncontrollably altered depending on a string's contents
can ever be *preferable* to keeping your namespace under control.

The ONLY specific functionality that exec-without-explicit-dicts
(EWOED for short) "gives" you is exactly the ability to enable such
uncontrolled namespace alteration.  And that is exactly what makes
it an abomination.

You can "dynamically construct a function object from a pattern"
in many OTHER ways than with EWOED.  These are "designed", or maybe
better, "architected" ways, controlled and controllable, while
EWOED is a "doomsday device" able to wipe out anything in its path.


> If you are coming to EuroPython maybe we find half an hour where we
> can discuss it on the screen? Doing this in mail makes it hard to not
> misunderstand each other.

I think I'll be at EuroPython, yes, although I don't understand what
you can possibly mean by "on the screen" in this context (it seems to
me that the inability to understand what you can possibly mean is
turning out to be frequent on my part -- just to double check, is
everybody else perfectly clear about what this "on the screen" is
meant to convey, or is it just as murky to everyone?).  Let's see if
that can help.  But the specific issue is "what are the alleged
advantages of losing control of my local namespace" -- dynamic
generation of code through the many OTHER, CONTROLLED means that
Python offers has its (SMALL) place, it's the uncontrolled 'power'
of namespace-destruction that (like the uncontrolled 'power' of
GOTO, vs the controlled means of architected control-structures)
I believe should be 'considered harmful' and to have NO place in
production code.


Alex



From pyth at devel.trillke.net  Tue Apr 30 09:04:11 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 15:04:11 +0200
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
In-Reply-To: <Xns92005AC7EB928duncanrcpcouk@127.0.0.1>
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1> <mailman.1020100431.16386.python-list@python.org> <Xns92005AC7EB928duncanrcpcouk@127.0.0.1>
Message-ID: <20020430130411.GB16217@devel.trillke>

On Tue, Apr 30, 2002 at 07:59:27AM +0000, Duncan Booth wrote:
> holger krekel <pyth at devel.trillke.net> wrote in
> news:mailman.1020100431.16386.python-list at python.org: 
> 
> > You really feel that these free floating strings with
> > no clear separations and lots of micro-management where
> > you dont't even know what does what and where the parts
> > start to float around and you ask yourself 
> > when the sentence will ever end 
> > 
> I didn't say that. The question was what was the best/shortest way to 
> replace the functions with list comprehensions. I answered the question. I 
> did not say anywhere that they were always better than the functional 
> alternative.

Citing from your initial posting:

> All of these replacements are big winners if they remove a use of the
> 'l' word. I don't think any of them lose in clarity even if there are no
> 'l's around.

regards,

    holger




From ajs at ix.netcom.com  Thu Apr  4 21:29:09 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Thu, 4 Apr 2002 21:29:09 -0500
Subject: Case insensitivity
Message-ID: <000e01c1dc49$ad1d81c0$0334fea9@carol>

I know I have this wrong.

We have a Boolean "False" that's not
really a Boolean as a synonym that's
not really a synonym for 0.

A clearly identifiable benefit being
avoiding the situation where I
define FALSE = 0 and you
define false = 0,  which leaves
open the possibility that FALSE
and false (in 2 different modules)
could mean the same thing.

Another blow to case insensitivity.

I'm a +1.

Art





From landy at despiertapr.com  Tue Apr 30 18:32:17 2002
From: landy at despiertapr.com (RR)
Date: 30 Apr 2002 18:32:17 -0400
Subject: [Zaurus-general] ANN: Python 2.2.1 for Sharp Zaurus
In-Reply-To: <3CCD1D46.B519567@river-bank.demon.co.uk>
References: <3CCD1D46.B519567@river-bank.demon.co.uk>
Message-ID: <1020205937.8210.15.camel@pepino>

On Mon, 2002-04-29 at 06:15, Phil Thompson wrote:
> Riverbank Computing is pleased to annouce the release of the port to
> Python v2.2.1 to the Sharp Zaurus. We believe the port will also run
> unchanged on the Compaq iPAQ (running Linux).
> 
> The download page is
> http://www.riverbankcomputing.co.uk/zaurus/download.php
> 


stll not able to install

storage i have 630kb  currently  tried cf no luck





From andymac at bullseye.apana.org.au  Mon Apr  8 06:31:59 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Mon, 8 Apr 2002 21:31:59 +1100 (edt)
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
In-Reply-To: <7xit74vzr5.fsf@ruckus.brouhaha.com>
Message-ID: <Pine.OS2.4.32.0204082124010.6077-100000@tenring.andymac.org>

On 6 Apr 2002, Paul Rubin wrote:

> Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:

{...}

> > There are OS vendors who choose to imposing grossly out-of-date releases
> > of 3rd party software on customers for their own selfish reasons.
>
> Are you talking about Red Hat?  If a Python release as recent as 1.5.2
> is already "grossly out-of-date", that confirms what I'm saying about
> releases happening too fast.  The idea of stability and maturity is that
> releases don't become grossly out of date anywhere near that quickly.

In actual fact Sendmail was the example I had principally in mind.

RH could have managed the python issue much more gracefully than they
chose to, as Python offers the facility to support multiple releases
in parallel.  That they still consider (and apparently have no plans to
reconsider for some time) a nearly 3 year old release as the "standard"
goes a bit further than I thought even you were arguing for.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From news at tylerdave.com  Wed Apr 10 23:21:31 2002
From: news at tylerdave.com (Dave Forgac)
Date: Thu, 11 Apr 2002 03:21:31 GMT
Subject: FTP via HTTP proxy
Message-ID: <%i7t8.723$7f4.443149@news1.news.adelphia.net>

I'm trying to write a component of a program that will use FTP to send and
recieve files.  The problem is that in certain cases, this needs to happen
over a proxy (WinGate, WinProxy, Proxy+, etc.).  Does anyone have any idea
how to do this?  I would really like to be able to finish this project in
python and not have to find an external solution.  Any help is appreciated.

-Dave Forgac




From borcis at geneva-link.ch  Thu Apr 11 04:30:06 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 11 Apr 2002 10:30:06 +0200
Subject: True/False story... (PEP 285)
References: <a8u99o$2lcu$1@news.nextra.cz> <vHys8.62333$pT1.1792330@news1.tin.it> <3CB2E617.8050600@geneva-link.ch> <a8vaba$cnv$1@slb3.atl.mindspring.net> <a93f3p$rcg$1@news.nextra.cz>
Message-ID: <3CB5498E.4030404@geneva-link.ch>

Petr Prikryl wrote:

> Boris wrote...
> 
>>Could be argued that there is no single natural mapping,
>>for an alternate to the current one we could have
>>
>>True     == 1
>>False    == -1
>>not x    == -x
>>x and y  == min(x,y)
>>x or y   == max(x,y)
>>x xor y  == -x*y
>>
>>A nice feature of that one (to my aesthetics), is that
>>
>>1j == sqrt(False)
>>
> 
> This alternate mapping breaks everything that exists now


It breaks nothing since it's not a PEP; and clearly it is natural enough 
not to break *everything* if it were and was adopted.


> and does not introduce anything better than the existing state.


A very boolean declaration, that :) In any case, look at context : 
nobody proposed it should take over the state.

Blind spots tax any vision. Many argued that enums were better than 
booleans; and cyclotomic fields are algebra's version of enums. Some 
mathematician wrote that a good representation is equivalent to a good 
teacher. Means it has to be inspiring.

Regards, Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From phr-n2002a at nightsong.com  Wed Apr  3 23:46:19 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 20:46:19 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <7xy9g4nk9w.fsf@ruckus.brouhaha.com>

I think there will need to be a format code for truth values:

  ("%b" % True) => "True"

etc.


From tdelaney at avaya.com  Tue Apr  2 19:35:08 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 3 Apr 2002 10:35:08 +1000 
Subject: PEP 287: reStructuredText Standard Docstring Format
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A244@natasha.auslabs.avaya.com>

> From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
> 
> David Goodger <goodger at users.sourceforge.net> writes:
> > Here's a serious proposal, safe to post now that April 
> Fool's is over.
> > Please read & comment.
> 
> Man, I thought it was an April fools joke when I saw it last night.
> All those `different` :kinds: 'of' "quoting" are much more confusing
> than explicit markup.  So I think it's better to use explicit markup.
> The ultimate explicit system is tangle/weave but that's a bit extreme.
> Javadoc has been very successful in practice and you could do a lot
> worse than just copying it lock stock and barrel.

I use javadoc. I don't like it. What I like least is the explicit markup -
it makes the source code documentation very difficult to read. In javadoc
you have lots of ugly constructions, mixed in with *different*,
inconsistent, ugly constructions if you want anything unusual (for example,
a list). The documentation for my unit tests looks like:

/**
 * Test:
 * <ul>
 *    <li>Class.method1();</li>
 *    <li>Class.method2();</li>
 *    <li>Class.method3().</li>
 * </ul>
 * Fails if:
 * <ul>
 *    <li><code>Class.method1</code> fails.</li>
 *    <li><code>Class.method2</code> fails.</li>
 *    <li><code>Class.method3</code> fails.</li>
 * </ul>
 **/

This is not very easy to read, although it produces nice HTML docs (the
<ul><li></li></ul> is ignored in the summary, but displayed in the
description, the semicolons split up the methods in the summary, the
fullstop terminates the summary).

Compare that to:

"""
Test:

- Class.method1()
- Class.method2()
- Class.method3()

Fails if:

- `Class.method1` fails.
- `Class.method2` fails.
- `Class.method3` fails.
"""

I think I got the above right - I've never read anything about
reStructuredText before. I have implemented a form of structured text for a
specific domain (in VBScript no less ... that was painful).

I read through all the reStructuredText docs, comparing the text versions to
the html versions. I found the text versions to be *very* easy to read,
whilst making it obvious in most cases when something was "special".

I particularly like the system of doing hyperlinks, and much prefer the more
verbose

.. _hyperlink: url

form than the short form.

> The thing I like least about the PEP is that it's written like
> marketing material--it sets out to push a particular solution and
> looks for things to bash about the alternatives rather than treating
> them evenhandedly and being willing to accept the idea that
> reStructuredText may not be the answer.

It does feel somewhat like marketing material in some documents, however the
"alternatives" document dispells that notion. There has obviously been a lot
of thought and discussion put into the system, and the author is shameless
about admitting that a lot of ideas have been stolen.

> widely-deployed, practical systems; a lot of people use them and like
> them and there are REASONS for that.  The PEP should acknowledge their
> success and see if there's something to be learned from it, rather
> than dismissing them so airily.

I think the PEP (and supporting documentation) did a good job of
acknowledging these systems, but quickly concluded that the systems were not
appropriate for the purpose.

Definitely +1 from me ... I would really like a standard, clean docstring
format. Might make it easier to get my next project done in Python (maybe
Jython).

Tim Delaney



From narnett at mccmedia.com  Fri Apr 26 16:06:55 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Fri, 26 Apr 2002 13:06:55 -0700
Subject: MySQLdb and DateTime objects
Message-ID: <NDBBLAEMOMEHKALAMKKJIEGLKEAA.narnett@mccmedia.com>

I'm have a bit of trouble figuring out how to use a date field retrieved
from MySQL when updating a record in another table.  Rather than a date
string, MySQLdb returns a DateTime object.  At first, imagining that since
it came out of a field, the date object would also then match the field in a
subsequent operation.  But no, it doesn't -- it produces an error.

E.g., "SELECT DISTINCT(Date) FROM Foo" returns a list of tuples whose first
member is a DateTime object, not a date string.  If the dates are in a list
called myDates, and I do something like, "UPDATE Foo SET X=%s WHERE Date=%s"
I get an error, as MySQLdb tries to insert a reference to the DateTime
object, rather than a date string.

What seemed to be the obvious solution was to use
MySQLdb.times.format_TIMESTAMP() or MySQLdb.times.format_DATE()-- but that
doesn't work, it returns an error:

  File "C:\Python22\Lib\site-packages\MySQLdb\times.py", lin
STAMP
    return d.strftime("%Y-%m-%d %H:%M:%S")
AttributeError: 'tuple' object has no attribute 'strftime'

Nor does str(), but that seemed like a long shot, anyway.

How the heck do I get this object back into a form that will work in a
query?

And is it just me, or is it rather inelegant that MySQLdb converts
timestamps into a DateTime object on the way out, but doesn't do the
reverse?

Thanks in advance.

Nick

--
narnett at mccmedia.com
(408) 904-7198





From aleax at aleax.it  Tue Apr 30 16:23:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 20:23:59 GMT
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu> <3kjz8.32779$8D3.928578@news1.tin.it> <aam88o$oec$1@netnews.upenn.edu> <g8yz8.81675$vF6.2428847@news2.tin.it> <aamt6p$dfa$1@netnews.upenn.edu>
Message-ID: <z3Dz8.83132$vF6.2475027@news2.tin.it>

Paul Hughett wrote:
        ...
> Hmm.  Is the value of r[-1:-1:-1] equal to array([12, 11, 10, 9, 8, 7,
> 6, 5, 4, 3, 2, 1, 0]) or to an empty array?  It could plausibly be

Any slice x[a:a] is empty.  Why ever should x[a:a:-1] be any different
from x[a:a:1]?!  I suggest you download and install Numeric so you
can easily check on such doubts.

Slicing rules are simple.  All indices, positives or negatives, indicate
specific items.  Too-small indices point "before the start"; too-large 
ones, "after the end", as does a missing stop -- there's no IndexError as 
there could be in indexing.  If start is missing it's the first if stride>0,
the last if stride<0; if stop it's missing it's after-the-end if stride>0,
before-the-start if stride<0.

OK, so look at the start and stop items after this normalization.  The
slice is empty if they're the same, if start is after stop and stride>0,
if start is before stop and stride<0.

If the slice is not empty, the start element is included.  Now add
(algebraically, i.e. including sign) the stride to the index of the start
element to get "another start element" and go back to the previous
paragraph to see if the overall result slice is finished or if you need
to keep repeating this.

Apart from the normalization of negative and out-of-range indices,
these are the same rules that range and xrange use, except they
return "the indices", so to speak, rather than any items.


> In any case, I will be adopting the Numeric convention.  As you point
> out, having multiple conventions is definitely a bad idea.

Even having just one convention might be a problem if the convention
was bad -- complicated, irregular, inconsistent.  I think Python's
arrangement here is quite good.


Alex



From sholden at holdenweb.com  Sat Apr  6 10:02:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 6 Apr 2002 10:02:51 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff> <yu99wuvovkmm.fsf@europa.research.att.com> <00af01c1db2a$e7316d10$ced241d5@hagrid> <200204031619.g33GJXm02577@europa.research.att.com> <mailman.1017853522.14491.python-list@python.org>
Message-ID: <V7Er8.26349$q63.14266@atlpnn01.usenetserver.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:mailman.1017853522.14491.python-list at python.org...
> andrew wrote:
> >
> > Fredrik> if you study real python code, you'll find that there is.
> >
> > I mean that the language doesn't prefer a single way,
> > as opposed to programmers' conventions.
>
> it's more than just a convention -- built-in functions and operators
> always return 0 for false and 1 for true whenever they need to in-
> vent a boolean value.  this is documented in the language reference.
>
> (and if you look inside, there's something called Py_False that's an
> integer zero, and something called Py_True that is an integer one).
>
Which the Win32 extensions, for example, actually wrap up in an entirely
separate COM type, so that they don't compare equal to integer values.

> this is also true for the standard library, and lots of major extensions.
>
> also note that the PEP says "The bool type would be a straightforward
> subtype (in C) of the int type, and the values False and True would
> behave like 0 and 1 in most respects".  Guido didn't pick those values
> out of thin air...
>
Since the decision not to move forward with Python 3000 it appear my
favorite language is dying the death of 1,000 cuts!

regards
 Steve






From maxm at mxm.dk  Tue Apr  2 06:27:57 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 02 Apr 2002 11:27:57 +0000
Subject: ASP / IIS problems
References: <3ca99102.602353265@news.mch.sni.de>
Message-ID: <3CA995BD.5090103@mxm.dk>

Nikolai Kirsebom wrote:

> Does anyone have any idea where to look.
> I've compared the settings for the IIS server (running Python 2.0)
> which works with the one which does not work, and all are identical.


Problem and solution here (one line url):

http://groups.google.com/groups?hl=da&threadm=vtk68.20338%248d1.7162721%40news1.rdc1.md.home.com&rnum=1&prev=/groups%3Fq%3Dpython%2Biis%2Bwin32all-146%26hl%3Dda%26selm%3Dvtk68.20338%25248d1.7162721%2540news1.rdc1.md.home.com%26rnum%3D1

regards Max M



From ark at research.att.com  Wed Apr 10 23:02:31 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 11 Apr 2002 03:02:31 GMT
Subject: binding a reference to a variable
References: <mailman.1018481963.1924.python-list@python.org>
Message-ID: <yu99hemj7xa0.fsf@europa.research.att.com>

Tim> I did that a looong time ago, but the code didn't survive my
Tim> then-employer's financial meltdown.  Unless you bought their
Tim> assets at the fire sale, I'm not sure how you could know whether
Tim> your approach is completely different <wink>.

The implementation I was thinking of was actually done in Ada, and
gatewayed to Python...

Tim> It is, though.  The use of unevaluated expressions in SNOBOL4 is
Tim> crucial for advanced matching, and the "how do you do that with a
Tim> bare name in Python?"  problem seems unsolvable in a squeaky
Tim> clean way.  Like:

Tim>     ARB = NULL | LEN(1) *ARB

Yup.  But why not

        arb = null | (len(1) + indirect(lambda: arg))

?

>> Focusing just on the  (break("@") . id)  subexpression, how would
>> one implement it?  That's what led me on this path.

Tim> After I left that path <wink>, I found it was much easier to
Tim> model Icon's spelling of patterns, and with generators in 2.2
Tim> that can be a breeze.

Indeed.  But I want more Snobolish than Iconish patterns.

Tim> However, Icon pattern-matching has a more procedural flavor, and
Tim> SNOBOL4 pattern-matching a more declarative flavor, and each has
Tim> strengths and weaknesses following from that.

Agreed completely.

Tim> Before I left that path, it looked something like

Tim>     id = S.var('id')
Tim>     ... id.ass(S.break('@')) ...

Tim> were .ass() had an optional argument to choose between delayed
Tim> (.) and immediate ($) assignment, defaulting to delayed.  BTW,
Tim> operator overloading was much harder to predict in Python back
Tim> then, so I often settled for ordinary method-call notation.
Tim> After operator overloading because easy to predict in Python, I
Tim> learned to hate it from being forced to endure reading code that
Tim> used it <0.7 wink>.

Maybe there's another way.  If I knew, I'd say what it is :-)

Tim> An iteration before that I used string names as keys to an
Tim> explicit namespace dictionary (as others have suggesting), but
Tim> found that modeling "SNOBOL4 variable" as a class in its own
Tim> right felt at least a little better, despite needing explicit
Tim> "declarations", i.e. lines of the form

Tim>     var1, var2, ... = S.var(name1, name2, ...)

Tim> Remembering to do that once was less painful than needing to slop
Tim> in quotes everywhere.

Might well be.

Tim> Have fun!  It's an interesting challenge.

Indeed.  And a good learning vehicle.


-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From wilk at flibuste.net  Fri Apr 26 12:01:22 2002
From: wilk at flibuste.net (William Dode)
Date: Fri, 26 Apr 2002 18:01:22 +0200
Subject: reportlab and BaseDocTemplate
Message-ID: <20020426180122.3f7ce976.wilk@flibuste.net>

hello,

I would like to use BaseDocTemplate to make report with different
PageTemplate. But the documentation is not enough for me, where can i find
some examples ?

thanks

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From tejarex at yahoo.com  Sat Apr  6 12:08:26 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 06 Apr 2002 17:08:26 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <%_Br8.20276$nt1.1661286@newsread2.prod.itd.earthlink.net>
Message-ID: <eYFr8.255468$Gf.23514820@bin2.nnrp.aus1.giganews.com>

"robin and jim" <robinjim at earthlink.net> wrote in message
news:%_Br8.20276$nt1.1661286 at newsread2.prod.itd.earthlink.net...
>
> Gee, I thought it would be obvious why I was asking such a question
<wink>!

It *is* obvious.  Your question is based on a misunderstanding of
Python which you have so far resisted giving up <wink>!

> Well, yours truly thought he would be clever and add some additional
> information to this docstring as it became available at run time via
the %
> placeholder mechanism.  Alas, the result is a single-quoted string

No, no, no!  There are single-quote string *literals* (what you type)
and triple-quote string *literals*.  Both get translated into internal
string *objects* after *stripping off* the quotes and interpreting (or
not) backslash escapes.  There are also string *expressions* that
yield a string object and other expressions (with variables and/or
function calls) that *might* yield a string object .  A string object
carries with it *no* history of how it was generated.

A string *literal* (regardedless of how quoted!) in the proper place
get stashed with the module/class/function/method object as a __doc__
attribute when that object is *created* (rather than later when
*called*, if callable).  When you got clever and turned the literal
into an *expression*, it was no longer a literal with a value known at
object creation time but an expression with value not known until
later, if ever.  Expressions are not and cannot be used as
automatically-assigned docstrings.  This has NOTHING to do with the
quotedness (or rawness) of the literal.

As Emile van Sebille pointed out in his posting, you must explicit do
any run time manipulation of the docstring yourself.

Terry J. Reedy





From guido at python.org  Wed Apr  3 08:10:08 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 13:10:08 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com>
Message-ID: <3CAB0015.6563C468@python.org>

Someone wrote:

> It's a cost/benefit thing.  Type/class unification may have cost more
> in complexity than bools cost, but it has high benefits.  The bool
> thing would a complicated change with limited benefit, so it's provoked
> some sharp reactions.

That's an interesting response, because it highlights what I think is
the main misunderstanding about the PEP.  You should think of it a
very *small* change!

Some operations that conceptually return a truth value (like a==b)
have to pick something to represent that result.  This is only needed
when no other object among the operation's arguments is suitable to
represent the result, for example in "not x".  While almost any value
is usable to express a truth value, traditionally the Python core and
standard library have consistently picked 0 to represent falsehood
and 1 for truth.

A problem with this is that the intention isn't always clear.  Does a
particular 1 represents a specific count or is it simply a
representative of truth?  While ultimately this is a matter of
documentation, having a notation helps clarify the matter (in cases
where it is used).

To solve the problem of clarifying intent, the PEP introduces two new
values that the core will use in preference over 0 and 1, and standard
names for them.  It makes them the only values of a new type, so that
they can be recognized as the canonical representatives when
necessary.  For backwards compatibility, these values act just like
the integers 0 and 1 in non-Boolean contexts (except when printed or
converted to a string).  For simplicity of implementation, and to
guarantee backwards compatibility at the C level, the new type is
implemented as a subtype of the built-in int type.

No user code needs to be changed unless it is sensitive to the string
representation of Python objects.  There is no prejudice against user
code that continues to use 0 and 1 (or something else) to represent
truth values.

In order to give people coming from (or going to) other languages an
easy time getting used to the new type, it is called bool, and its
values are called False and True.  But like so many things in
Python[*], its bool type acts slightly different than the Boolean type
in other languages.  That's fine -- other languages don't all agree on
the details of Boolean types anyway, varying from very strict (Pascal,
Java) to very loose (C, Lisp).

This is long enough of a response.  To those people who complain that
I don't take their long thoughts seriously, I can only respond that I
thought long and hard about this PEP too, and am similarly disappointed
that it faces so much opposition based upon what *might* happen.

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

[*] Take variables for example.  You can write "x = 1" in Python just
as in C or Perl, and superficially it does the same thing.  But the
meaning is profoundly different!  1 is an object, and x is a name
binding for that object.  This goes on in Python all the time: other
examples are x+y, for loops, even function argument default values.
Booleans have their own twists and always had ("if x", "x and y").
Python is not unique here either -- many language have different
semantics underneath familiar notations, and this can represent a real
breakthrough (like Smalltalk, which redefined x+y as sending the
message "+y" to the object x).


From mhammond at skippinet.com.au  Wed Apr  3 20:08:34 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 04 Apr 2002 01:08:34 GMT
Subject: os.remove() permissions problems
References: <VCJq8.208$c6.57043@typhoon.mn.ipsvc.net>
Message-ID: <3CABA7C0.1090307@skippinet.com.au>

joel8bit wrote:
> I'm trying to process some files on my system.  when I run a script with
> os.remove() in it I'm denied permission to delete the file.  I've created
> these files as the same user that I'm running the script as.  This is on a
> win2k box, when the python interpreter runs, does it run as it's own
> unprivleged user?

The files are almost certainly in use.

Mark.



From mwh at python.net  Wed Apr 10 10:56:12 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 10 Apr 2002 14:56:12 GMT
Subject: RELEASED: Python 2.2.1
References: <mailman.1018435709.31373.python-list@python.org> <LWXs8.208801$VJ1.16864889@bin3.nnrp.aus1.giganews.com>
Message-ID: <uelhnvbro.fsf@python.net>

"Terry Reedy" <tejarex at yahoo.com> writes:

> "Michael Hudson" <mwh at python.org> wrote in message
> news:mailman.1018435709.31373.python-list at python.org...
> > We've released a bugfix release of Python, 2.2.1!
> 
> Hooray!  Just as soon as I send off my income tax forms....
> 
> > This release should be fully compatible with Python 2.2.  Everyone
> > using 2.2 is recommended to upgrade to 2.2.1.
> >
> > On the 2.2.1 pages
> >     http://www.python.org/2.2.1/
> 
> > This being a bugfix release, there are no exciting new features --
> we just
> > fixed a lot of bugs.  For a moderately complete list, please see:
> >
> >     http://www.python.org/2.2.1/NEWS.txt
> 
> Ah, but you did introduce one (unexciting?) new feature that users
          ^^^
It was Guido, not me...

> should know about:
> 
> - Added new builtin function bool() and new builtin constants True and
>   False to ease backporting of code developed for Python 2.3.  In 2.2,
>   bool() returns 1 or 0, True == 1, and False == 0.
> 
> This will also make it easier to write code that looks forward to 2.3+
> as long as one does not test type(True), `True`, and str(True) (and
> ditto for False).

...and I'm not overwhelmingly convinced that it's a good idea, so I
didn't shout about it...

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme


From dperedo at dow.com  Thu Apr  4 12:02:37 2002
From: dperedo at dow.com (Dickerson, Dee)
Date: Thu, 4 Apr 2002 12:02:37 -0500 
Subject: Python Numeric Installation problems
Message-ID: <911A53117F96D31184500000F8BDCA870E40838B@TXNTE21>

ALL:

I am trying to install the Numeric Python on a linux Redhat 7.1 machine.  I
have installed the python 2.0 rpm and devel rpm.  My envirnment variables
are set to:
PYTHONHOME=/usr/lib/python2.0
PYTHONPATH=/usr/lib/python2.0

when trying to the the build I get:
[u370471 at n16 Numeric-21.0]$ python2 setup.py build
running build
running build_py
not copying Lib/ArrayPrinter.py (output up-to-date)
not copying Lib/Numeric.py (output up-to-date)
not copying Lib/Precision.py (output up-to-date)
not copying Lib/UserArray.py (output up-to-date)
not copying Lib/numeric_version.py (output up-to-date)
not copying Lib/LinearAlgebra.py (output up-to-date)
not copying Lib/MLab.py (output up-to-date)
not copying Lib/Matrix.py (output up-to-date)
not copying Lib/RandomArray.py (output up-to-date)
not copying Lib/site.py (output up-to-date)
not copying Packages/FFT/Lib/FFT.py (output up-to-date)
not copying Packages/FFT/Lib/__init__.py (output up-to-date)
not copying Packages/MA/Lib/MA_version.py (output up-to-date)
not copying Packages/MA/Lib/__init__.py (output up-to-date)
not copying Packages/MA/Lib/MA.py (output up-to-date)
not copying Packages/Properties/Lib/Properties.py (output up-to-date)
not copying Packages/Properties/Lib/__init__.py (output up-to-date)
not copying Packages/Properties/Lib/Properties_version.py (output
up-to-date)
not copying Packages/RNG/Lib/__init__.py (output up-to-date)
not copying Packages/RNG/Lib/Statistics.py (output up-to-date)
not copying Packages/kinds/Lib/__init__.py (output up-to-date)
not copying Packages/kinds/Lib/kinds.py (output up-to-date)
not copying Packages/kinds/Lib/kinds_version.py (output up-to-date)
running build_ext
Traceback (most recent call last):
  File "setup.py", line 195, in ?
    ext_modules = ext_modules
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/core.py", line
138, in setup
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/dist.py", line
829, in run_commands
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/dist.py", line
849, in run_command
  File
"/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/command/build.py",
line 106, in run
  File "/usr/lib/python2.0/cmd.py", line 328, in run_command
    print "\n"
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/dist.py", line
849, in run_command
  File
"/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/command/build_ext.py"
, line 200, in run
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/sysconfig.py",
line 106, in customize_compiler
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/sysconfig.py",
line 365, in get_config_vars
  File "/usr/lib/python2.0/pprint.py", line 42, in ?
    from StringIO import StringIO
  File "/usr/lib/python2.0/StringIO.py", line 32, in ?
    import errno
ImportError: No module named errno

Trying to run install I get

[u370471 at n16 Numeric-21.0]$ python2 setup.py install
running install
Traceback (most recent call last):
  File "setup.py", line 195, in ?
    ext_modules = ext_modules
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/core.py", line
138, in setup
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/dist.py", line
829, in run_commands
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/dist.py", line
848, in run_command
  File "/usr/lib/python2.0/cmd.py", line 112, in ensure_finalized
    return self.default(line)
  File
"/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/command/install.py",
line 255, in finalize_options
  File "/var/tmp/python2-2.0-root/usr/lib/python2.0/distutils/sysconfig.py",
line 365, in get_config_vars
  File "/usr/lib/python2.0/pprint.py", line 42, in ?
    from StringIO import StringIO
  File "/usr/lib/python2.0/StringIO.py", line 32, in ?
    import errno
ImportError: No module named errno

Any suggestions?

Please respond to me as I am not a member of the list yet.  Thank you in
advance

 ____________________________________________
Dee Dickerson
Unix System Administrator
The Dow Chemical Company
2301 N. Brazosport Blvd., B-1217
Freeport, TX 77541
Ph:  (979) 238-4449
Fax: (979) 238-0244 
____________________________________________








From h.e.w.frater at cs.cf.ac.uk  Sat Apr 20 23:59:05 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 20 Apr 2002 20:59:05 -0700
Subject: more oracle problems!
Message-ID: <e2e9e807.0204201959.7232ae91@posting.google.com>

hi all, I'm sending this again because It hasn't appeared on the group
so I thought it might have got lost somewhere. Apologies If it has got
there and I haven't realised.
I can now import the module but only because I've domped the files
from my DCOracle directory into my cgi-bin directory. However, when ir
run my script I now get this error:
Traceback (most recent call last):
  File "/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/addItemType.py",
l
ine 12, in ?
    dbh = DCOracle2.connect('scm1hewf/********@cs2000')
  File "/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/DCOracle2.py",
lin
e 93, in connect
    db = dco2.connect(u, p, dbname)
dco2.DatabaseError: (12154, 'Error while trying to retrieve text for
error ORA-1
2154')

What does this mean, and how do I fix it?

Hugh


From peter at engcorp.com  Mon Apr 15 22:21:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 15 Apr 2002 22:21:47 -0400
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
Message-ID: <3CBB8ABB.57BAA189@engcorp.com>

Duncan Smith wrote:
> 
> My boss is considering moving to Python (from Poplog), so I've coded up
> something to read in records from a text file (as below) and return a
> statistic based on the numbers of unique and paired records (as he
> requested).  Easy enough (code below), but now he wants to compare it for
> speed against some existing code (Poplog?).  

As another "boss", I want to make the strong point that speed 
is not what Python is about (although you may well make this
fast enough to suit him... it's not necessarily slow).

Python provides exceptional benefits in the areas of development
time, readability and therefore code maintenance, flexibility,
reliability, and ease of learning.

If he concludes that it's not "fast enough", I suspect he
is focusing too much on a relatively unimportant issue 
and likely to miss out on a some major competitive advantages.

Most of the time when someone concludes Python is not fast
enough, they have a situation where something is expected
to run once or twice a day, and it might take ten minutes
with Python and one minute with C, and they conclude that
is too slow...  meanwhile the C version is buggy, took 
four times longer to develop, and nobody likes to maintain
it because the code is five times longer than the Python
one would be.  

So make sure he takes a good hard look at just how much
speed he really _needs_ (not wants), and at the places
where Python really shines, before closing the door on
your suggestion.

(For what it's worth, we've more or less outlawed 
any internal tools not written in Python as being a 
waste of the developer's time, and NONE of our utilities
are "too slow".  (No, we're not quite as anal about it
as that sounds, but you get the point.))

-Peter


From sandysj at juno.com  Thu Apr  4 12:13:43 2002
From: sandysj at juno.com (Jeff Sandys)
Date: Thu, 4 Apr 2002 17:13:43 GMT
Subject: ANN: SLiP and SLIDE
References: <mailman.1017879911.8562.python-list@python.org> <3cac2b07$0$21495$626a54ce@news.free.fr>
Message-ID: <3CAC89C7.4E637D76@juno.com>

jm wrote:
> 
> Scott Sweeney <ssweeney at idiosync.net> writes:
> >
> > SLiP 0.1: a "Sorta Like Python" shorthand syntax for XML
> >       http://www.scottsweeney.com/projects/slip
> 
> How different is Slip from SOX ?
> http://www.langdale.com.au/SOX/

How are either of these different from YAML?
http://yaml.org


From joonas at olen.to  Fri Apr  5 10:16:02 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Fri, 05 Apr 2002 15:16:02 GMT
Subject: how to convert an integer to a string?
References: <a8ho5l$ubt$00$2@news.t-online.com>
Message-ID: <3CADC0BD.4000008@olen.to>

Sascha Ernst wrote:
> thanx, sascha
> 
> 
> 

Q: how to convert an integer to a string?

A:
integer = 1
integerstring = str(integer)



From guido at python.org  Sun Apr  7 10:23:45 2002
From: guido at python.org (Guido van Rossum)
Date: Sun, 07 Apr 2002 14:23:45 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018161735.24484.python-list@python.org> <20020407134159.3680634c.wilk@flibuste.net>
Message-ID: <3CB05744.4D7D049E@python.org>

William Dod? wrote:
> > If we had more active Python developers, that might be feasible.
> how many their are ?

(a) Not enough.

(b) The SourceForge project has just shy of 40 developers,
    but over half of them never contribute anything.

(c) The python-dev list has over 200 members.

(d) The Misc/ACKS file in the distribution has under 500 names.

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


From calaban.madness at blueyonder.co.uk  Fri Apr  5 20:16:48 2002
From: calaban.madness at blueyonder.co.uk (Alastair Nicol)
Date: Sat, 06 Apr 2002 01:16:48 GMT
Subject: socket security
Message-ID: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>

Hi,
  I've been enjoying the wonders of socket programming with Python. I have 
read a lot of information on the subject, but still have a few unanswered 
questions which is why, out of desperation, i'm asking here. 

1) Can only hosts which can see the interface your program is bound to 
communicate with your program. In the "normal" case if you bind to 
127.0.0.1 for example ONLY your machine can see that interface. Other 
machines cannot connect to it. Is this a safe / secure assumtion? (not just 
for localhost)

2) Also when a client connects to a socket the server gets the IP address 
of the client. Is this the "real" ip number? Ive heard you can spoof IP 
numbers so what hoops are needed to validate that the client is really that 
client.

openSSL is an option i would like to use, but for reasons to numerious to 
list here, i cant.


Thanks, Alastair Nicol



From aleax at aleax.it  Mon Apr 29 02:01:39 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 08:01:39 +0200
Subject: Update locals()
In-Reply-To: <003d01c1ef09$3d4c28b0$445d4540@Dell2>
References: <003d01c1ef09$3d4c28b0$445d4540@Dell2>
Message-ID: <02042908013901.25806@arthur>

On Monday 29 April 2002 01:05, Brian Quinlan wrote:
	...
> >   may not affect the values of local variables used by the
> interpreter.
>
> The reason that changes to the locals dictionary might not be visible is
> that LOAD_FAST instructions, which are used for local variable lookup,
> use an indexed list and not the locals dictionary.

Right -- and having such change attempts be diagnosed at once (by
making locals() a pseudo-dict that raises on set attempts) would help
debugging.  Similar optimizations to globals() are under consideration
(there are three PEPs out about that, I believe).


Alex




From mlh at vier.idi.ntnu.no  Fri Apr  5 22:54:59 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 03:54:59 +0000 (UTC)
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com> <m3bsczfgic.fsf@mira.informatik.hu-berlin.de>  <mailman.1017947953.14025.python-list@python.org>
Message-ID: <slrnaasscj.178.mlh@vier.idi.ntnu.no>

In article <mailman.1017947953.14025.python-list at python.org>, Cliff
Wells wrote:
>On 04 Apr 2002 20:47:07 +0200
>Martin v. Loewis wrote:
>
>> While Gustavo suggests to let pass 'a is True', I recommend to give a
>> warning for this - people who want to see whether something is True
>> should write 'True is a'.
>
>But only if they do it in a gravelly Yoda voice ;)

ROTFLMAO!

Or ... (in a gravelly Yoda voice):

OTFRMAOL

:))

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From skip at pobox.com  Thu Apr 18 11:32:40 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 18 Apr 2002 10:32:40 -0500
Subject: Readline/command history w/ interactive prompt
In-Reply-To: <20020418081246.GB3689@andy.tynant.ftech.net>
References: <mailman.1019061574.15837.python-list@python.org>
        <a9kcml$69k$1@news.wrc.xerox.com>
        <20020418081246.GB3689@andy.tynant.ftech.net>
Message-ID: <15550.59160.67740.852428@12-248-41-177.client.attbi.com>

    >> Try hitting control-r and typing a short string that last appeared in
    >> the line desired ("my" or even "m" would work in this case).

    Andy> _Nice_!  Thanks for that.  One question though: where's this
    Andy> feature documented?  :-)

Well, assuming you use Emacs, your fingers already know about it without any
interference from your brain. :-)

More seriously, try executing "info readline".  There is a subsection
entitled "Readline Interaction" under the heading "Command Line Editing".
If you're not an Emacs person you can choose select vi-style key bindings.

Of course, if you're a Word person the arrow keys also work...

Skip





From juenglin at informatik.uni-freiburg.de  Fri Apr 12 09:23:27 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 12 Apr 2002 15:23:27 +0200
Subject: __iadd__ and fellows missing (Python 2.2)
Message-ID: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de>

Shouldn't I find the special methods '__iadd__' et al, when
asking for the attributes of 'int'?

>>> i=1
>>> i+=1
>>> dir(i)
['__abs__', '__add__', '__and__', '__class__', '__cmp__', '__coerce__', '__delattr__', '__div__', '__divmod__', '__doc__', '__float__', '__floordiv__', '__getattribute__', '__hash__', '__hex__', '__init__', '__int__', '__invert__', '__long__', '__lshift__', '__mod__', '__mul__', '__neg__', '__new__', '__nonzero__', '__oct__', '__or__', '__pos__', '__pow__', '__radd__', '__rand__', '__rdiv__', '__rdivmod__', '__reduce__', '__repr__', '__rfloordiv__', '__rlshift__', '__rmod__', '__rmul__', '__ror__', '__rpow__', '__rrshift__', '__rshift__', '__rsub__', '__rtruediv__', '__rxor__', '__setattr__', '__str__', '__sub__', '__truediv__', '__xor__']
>>> dir(int)
['__abs__', '__add__', '__and__', '__class__', '__cmp__', '__coerce__', '__delattr__', '__div__', '__divmod__', '__doc__', '__float__', '__floordiv__', '__getattribute__', '__hash__', '__hex__', '__init__', '__int__', '__invert__', '__long__', '__lshift__', '__mod__', '__mul__', '__neg__', '__new__', '__nonzero__', '__oct__', '__or__', '__pos__', '__pow__', '__radd__', '__rand__', '__rdiv__', '__rdivmod__', '__reduce__', '__repr__', '__rfloordiv__', '__rlshift__', '__rmod__', '__rmul__', '__ror__', '__rpow__', '__rrshift__', '__rshift__', '__rsub__', '__rtruediv__', '__rxor__', '__setattr__', '__str__', '__sub__', '__truediv__', '__xor__']
>>>

I couldn't find them either in ther operator-module:

>>> import operator 
>>> dir(operator)
['__abs__', '__add__', '__and__', '__concat__', '__contains__', '__delitem__', '__delslice__', '__div__', '__doc__', '__eq__', '__file__', '__floordiv__', '__ge__', '__getitem__', '__getslice__', '__gt__', '__inv__', '__invert__', '__le__', '__lshift__', '__lt__', '__mod__', '__mul__', '__name__', '__ne__', '__neg__', '__not__', '__or__', '__pos__', '__repeat__', '__rshift__', '__setitem__', '__setslice__', '__sub__', '__truediv__', '__xor__', 'abs', 'add', 'and_', 'concat', 'contains', 'countOf', 'delitem', 'delslice', 'div', 'eq', 'floordiv', 'ge', 'getitem', 'getslice', 'gt', 'indexOf', 'inv', 'invert', 'isCallable', 'isMappingType', 'isNumberType', 'isSequenceType', 'le', 'lshift', 'lt', 'mod', 'mul', 'ne', 'neg', 'not_', 'or_', 'pos', 'repeat', 'rshift', 'sequenceIncludes', 'setitem', 'setslice', 'sub', 'truediv', 'truth', 'xor']


Why?

Cheers,
Ralf






From tismer at tismer.com  Thu Apr 25 07:43:53 2002
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 25 Apr 2002 13:43:53 +0200
Subject: Stackless goes Limbo
References: <mailman.1019662503.1913.python-list@python.org> <h%Ex8.184144$K5.15424154@bin5.nnrp.aus1.giganews.com>
Message-ID: <3CC7EBF9.9020905@tismer.com>

Terry Reedy wrote:
> "Christian Tismer" <tismer at tismer.com> wrote in message
...

>>Besides the tasklets, there are "channel" objects,
>>which allow for communication between tasklets.
>>This approach is simple and powerful. Blocking of
>>tasklets takes place on writes when no tasklet
>>is listening, or on reads, when no data is available.
> 
> 
> Whatever else you do, please consider adding an iterator interface for
> reading channels:
> 
> for item in channel: whatever()

Sure. Pipes are nice.

> Also nice, but probably much harder, would be to hijack 'yield' to
> write to a channel (for tasklets that write to only one channel, or
> which have a default output channel).

Yes. I already hijacked yield in my first attempt.
Before I became familiar with channels, I was always
thinking around generators and coroutines. For yield,
I had a "client" field as the implicit destination.
Replacing this with an implicit channel would be
quite easy, and the overhead would be neglectible.

Where I still have problems to get things clear is
how to handle exceptions.
With tasklets, waiting on some channel, this is
no question. The tasklets are on their own, and
they handle their exceptions alone.
But what about generators/coroutines? They look
tighter coupled, and I'm not so sure who has
to handle exceptions. On a yield or coroutine
transfer, it looks like a caller/callee situation,
where exceptions might be propagated.

Any ideas?

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From boud at valdyas.org  Wed Apr  3 06:24:59 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Wed, 03 Apr 2002 10:24:59 -0100
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl> <1d6q8.118888$Ou.77410@atlpnn01.usenetserver.com> <3ca940d3$0$32314$e4fe514c@dreader4.news.xs4all.nl> <slrnaak6oe.s4j.mlh@vier.idi.ntnu.no>
Message-ID: <3caabd6d$0$45613$e4fe514c@dreader3.news.xs4all.nl>

Magnus Lie Hetland wrote:

> In article <3ca940d3$0$32314$e4fe514c at dreader4.news.xs4all.nl>,
> Boudewijn Rempt wrote:
>>Steve Holden wrote:
>>
>>> Congratulations! I remember your original posts asking for good
>>> examples and choices for the graphics subnsystem. You must feel a
>>> weight has been lifted ...
>>> 
>>
>>I do. It has certainly been an interesting experience. I was wrapping
>>up the last of the text last year, when there was this thread about
>>writing books on comp.lang.python. It felt very to the point.
>>
>>Anyway, I'm quite satisfied with the results :-).
> 
> I'm getting a "Not found" error for a javascript file:
> 
>   http://www.opendocspublishing.com/pyqt/validate.js
> 
> (A bit irritating when MSIE users don't notice missing js and css
> files, because MSIE ignores them... I can't even access the page in
> Netscape :I)
> 
> Anyway, congrats! :)
> 

I've forwarded this to Joshua -- I haven't tried buying one, of course :-).

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


From gimbo at ftech.net  Mon Apr 15 05:06:39 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Mon, 15 Apr 2002 10:06:39 +0100
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
In-Reply-To: <3CBA6595.104ED599@cosc.canterbury.ac.nz>
References: <mailman.1018755873.29647.python-list@python.org> <3CBA6595.104ED599@cosc.canterbury.ac.nz>
Message-ID: <20020415090639.GA23784@andy.tynant.ftech.net>

On Mon, Apr 15, 2002 at 05:31:01PM +1200, Greg Ewing wrote:
> Tim Peters wrote:
> > 
> > As originally posted, they were the "20 Pythonic Theses".  They
> > never need revision, because I only gave 19 -- the 20th (or
> > perhaps the first -- they weren't numbered) was left open for
> > Guido to fill in.
> 
> Perhaps it should be "19 Pythonic theses is sufficient."

Surely the gap is just another Spanish Inquisition reference?

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From rjones at ekit-inc.com  Tue Apr  2 00:57:02 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 2 Apr 2002 15:57:02 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
Message-ID: <200204020557.FAA22283@crown.off.ekorp.com>

On Tue, 2 Apr 2002 15:28, David Goodger wrote:
> Here's a serious proposal, safe to post now that April Fool's is over.
> Please read & comment.

Gets my vote :)

Disclaimer: I've been making quite good use of the ReST format for a month or 
two now. It's used in the Roundup documentation [1]_ and is definitely easy 
to use.


   Richard

.. [1] http://roundup.sf.net/doc/



From tanzer at swing.co.at  Sun Apr  7 05:17:48 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Sun, 07 Apr 2002 11:17:48 +0200
Subject: ACCEPTED: PEP 285
In-Reply-To: Your message of "06 Apr 2002 22:15:56 -0800."
             <7xelhsvxsz.fsf@ruckus.brouhaha.com>
Message-ID: <E16u8o0-0005XS-00@tswings.swing.cluster>

Paul Rubin <phr-n2002a at nightsong.com> wrote:

> Tim Peters <tim.one at comcast.net> writes:
> > > Doesn't matter what you call it, if a 2.1 script can't run at all or as
> > > expected on 2.0, that is a major change in the _colloquial_ sense.
> >
> > I expect you phrased that backwards, yes?  If you use a new-in-2.1
> > feature, then of course it's not going to run under 2.0.  The other
> > direction is almost always tractable, although in some cases it does
> > take foresight, or the active application of hindsight, to pull it off.
>
> I don't think he phrased it backwards.  "If you use a new-in-2.1
> feature" presupposes that 2.1 has features that aren't in 2.0, and I
> think James is arguing that it should not.  The issue is just how
> often you're willing to tell people they have to upgrade.
>
> If I'm running 2.0, and 2.1 comes out, it's pretty similar, so I don't
> upgrade.  Then someone sends me a script that using some new-in-2.1
> feature and I have to upgrade.  Then 2.2 comes out with new features
> of its own, someone sends me a 2.2 script and I have to upgrade again.
> Next, 2.3 comes out and I want to run a script that uses bools.
> Whoops, another upgrade.

FUD.

> It gets tedious!

Indeed.

If you need a newer version of Python to run someone's cool script,
just `make altinstall` of the newer version and use the shiny new
Python for that one cool script only.

Upgrading Python is not a boolean operation <wink>. Various Python
versions can live happily side by side. FWIW, I currently have
installed three different versions on one machine and four on another.

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





From marklists at mceahern.com  Fri Apr 12 12:25:30 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 11:25:30 -0500
Subject: separate software from data
Message-ID: <NCBBLFCOHHDIKCAFGCFBMELNKKAA.marklists@mceahern.com>

Here's the short form of my question:

How do people generally keep their Python modules and any configuration data
for those Python modules separate?

I'm looking for both conventions (e.g., config goes in /etc/foo) and
mechanics (e.g., use distutils with options x, y, and z).

The longer form of the question is a bit more convoluted, so for those with
patience, read on.  ;-)

I'm transitioning from Windows to Linux.  So part of my question is about
the conventions for separating software from data on Linux.  I assume it's
self-explanatory why you'd want to separate software from data.  In case
it's not, here's why I think it's important to separate:  You don't really
have to backup software (although you can if you want), but you do have to
backup data (it's not like you can re-download it).  So, if push comes to
shove, and you have limited resources/time, it's nice to be able to focus on
the important stuff and only back that up.

There's always a gray line between software and data that I label
configuration data.  In the Linux world, I guess that'd constitute anything
you do to install any piece of software beyond the equivalent of configure;
make; make install.  If you specify any configuration options, create daemon
accounts, etc.--ideally you'd do that with a script and you'd store that
script somewhere along with your data, so that you not only knew what you
did, but you could easily re-do it.

I've looked at distutils and it's not immediately obvious that it was
designed to solve the distribution problem in the particular what I want to
try to solve it, although I suppose I may be able to extend distutils with
special commands to do something like this?

I want a way to say something like this:

Install my module here:

  /usr/local/lib/pythonX.Y/site-packages/

By default, place configuration data here:

  /etc/my-package/

But if the installer specifies different configuration options when running
setup.py, then use those values when generating mypackage.conf from
mypackage.conf.template.

Of course, the person installing the package should be able to override
where configuration data is stored:

  $ python setup.py --configdir=/var/foo

If such an option existed, its effect might be to generate a configuration
file in the root of the package's module distribution folder:

  /site-packages/mymodule/mymodule.conf

containing something readable by ConfigParser:

[locations]
prefix=/etc/mymodule/
connections=%prefix/connections.xml

So that at runtime, the module first looks to the configuration information
in its package distribution to determine where it needs to look for the
distribution-specific configuration information?  Shwew.

Has anybody tackled this problem or do most folks just punt on this issue
and assume they can plop configuration information alongside their
distributed modules without providing any easy way for the person who
installs it to override that?

Thanks,

// mark





From roy at panix.com  Wed Apr  3 11:40:40 2002
From: roy at panix.com (Roy Smith)
Date: 3 Apr 2002 11:40:40 -0500
Subject: How best to write this try/except block?
Message-ID: <a8fba8$ln8$1@panix2.panix.com>

I've got a dictionary d, which may or may not have a value I'm looking
for.  If it does, I want to do a few things.  If it doesn't, I want to
do a few other things.  I could write:

try:
    value = d[key]
    foo = stuff (value)
except KeyError:
    foo = otherStuff (key)

which is fine as long as I can guarantee (hint: I can't) that stuff()
doesn't raise a KeyError.  The alternative would be something like:

foundIt = 0
try:
    value = d[key]
    foundIt = 1
except KeyError:
       foo = otherStuff (key)

if (foundIt):
   foo = stuff (value)

but that's exceptionally ugly.  Any suggestions for how better to
write this?



From mjackson at wrc.xerox.com  Wed Apr 17 13:51:17 2002
From: mjackson at wrc.xerox.com (Mark Jackson)
Date: 17 Apr 2002 17:51:17 GMT
Subject: Readline/command history w/ interactive prompt
References: <mailman.1019061574.15837.python-list@python.org>
Message-ID: <a9kcml$69k$1@news.wrc.xerox.com>

Andy Gimblett <gimbo at ftech.net> writes:
> Just a thought...
> 
> One of the things I find "warty" about the interactive prompt is that
> the command-line history retains multiple copies of identical
> successive lines, rather than trimming them down to a single instance.
> 
> Slightly fictional example:
> 
> >>> import my_mysql_wrapper
> >>> d = my_mysql_wrapper.connection(parms)
> >>> d.select('select count(*) from some_table')
> >>> reload(my_mysql_wrapper)
> <module 'my_mysql_wrapper' from 'my_mysql_wrapper.py'>
> >>> d.select('select count(*) from some_table')
> >>> d.select('select count(*) from some_table')
> >>> d.select('select count(*) from some_table')
> >>> d.select('select count(*) from some_table')
> >>> d.select('select count(*) from some_table')
> .... another 15 instances of that same .select() call
> 
> At this point, I go and fix a bug in my_mysql_wrapper, so I want to
> reload it.  Alas, I have to hit the up arrow 20 times to make that
> happen, and I have to watch carefully to see when I get the line I
> want.  If only it could be done with a mere 2 hits instead...

Try hitting control-r and typing a short string that last appeared in
the line desired ("my" or even "m" would work in this case).

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	The greatest masterpiece in literature is only a dictionary
	out of order.		- Jean Cocteau




From uwe at rocksport.de  Mon Apr  1 15:57:34 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 1 Apr 2002 20:57:34 GMT
Subject: wxPython  / wxSingleChoiceDialog ...
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de> <slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr> <a89vu0$8bm2c$1@hades.rz.uni-sb.de> <mailman.1017686284.32405.python-list@python.org>
Message-ID: <a8ahju$8dmek$2@hades.rz.uni-sb.de>

Cliff Wells <logiplexsoftware at earthlink.net> wrote:
| On 1 Apr 2002 15:55:44 GMT
| Uwe Schmitt wrote:

|> Vadim Zeitlin <zeitlin at seth.lpthe.jussieu.fr> wrote:
|> | On 1 Apr 2002 13:54:42 GMT, Uwe Schmitt <uwe at rocksport.de> wrote:
|> |> I tried to manipulate wxSinglChoiceDialog so that the presented 
|> |> choices are printed with a fixed width font. I tried SetFont()
|> |> but nothing changed...
|> 
|> |  You should change the font of the listbox used by this dialog, not the
| dialog
|> | itself. Although there is no way to access this listbox directly (and
| unlikely
|> | to ever be), you can iterate over the dialog children until you find
| it.
|> 
|> I inspectetd the __dict__ attribute of my singlechoicedialog object and
|> tried dir() but found no hints how to iterate over the dialog children...


| dialog.GetChildren().  Look at the docs for wxWindow.

o.k  i used that and found a wxListBox as children. How do I now
manipulate the font of the entries of the ListBox ???

Greetings,Uwe


-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From martin at v.loewis.de  Mon Apr 15 17:06:43 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 23:06:43 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122>
Message-ID: <m3662sk6xo.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> IMO 'internal representation' is a red herring. We have an abstract
> concept of number representation (twos complement) that corresponds
> to many hardware devices' usage, but it is the abstraction that is
> important IMO, not the hardware.

Yes, but this notion is inherently bound with the notion of "size" of
an integer, atleast in the signed variant (for unsigned integers,
there is no need for twos complement). If you create the complement,
you have to know how many bits to invert.

I absolutely agree with your (earlier) conclusion that bit arithmetic
is best done with unsigned numbers.

> When looking at integers cast as arrays of booleans, hex provides a
> representation, but w.r.t. boolean values it is not very useful to
> print it as base-16 integer with sign+absolute value.

I also agree.

> If I infer correctly, Ken is arguing that hex is more often used to
> view the boolean aspect of integers than their numeric
> aspect. That's true for me too.

And so it is for me. Still, I disagree that a display of a hex number
with a sign is illogical or incorrect. In fact, when talking about
"unbounded" numbers, it is illogical to assume twos complement.

> >Yes-- I read the PEP too-- when this happens, I will consider it a
> >Python "_WART_", and will set about inventing ways around it for my
> >own use.  I have forgotten more computer languages than most people
> >will ever learn-- an I have NEVER found printing a hexadecimal
> >number in sign and magnitude format to be useful...
>
> I'll second that.

At a minimum, you need to propose an alternative behaviour then. I
must say that I don't find the proposed behaviour of mandatory leading
zeroes (the 0h proposal) very intuitive, or following tradition.

Instead, there is a simple solution: use unsigned numbers if you want
to represent bit arrays.

> I would guess so. I do. But for language purposes my preference is
> to refer to an abstract bit pattern, not hardware (even though
> hardware mappings will exist and be useful).

That is only meaningful for positive numbers - it just doesn't work
for twos complement.

Regards,
Martin


From tim.one at comcast.net  Sat Apr  6 00:59:31 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 06 Apr 2002 00:59:31 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEHAOLAA.tim.one@comcast.net>

[James Logajan]
> ...
> Seriously Guido, how often will I need to check back to www.python.org
> to keep up with the changes you are making?

It's unclear why you would ever need to, if you want to stick with an older
version, unless you wanted to pick up a pure bugfix release (a 3-digit
release, like next week's upcoming 2.2.1 and 2.1.3, being, respectively, the
first pure bugfix release in the 2.2 series, and the third in the 2.1
series).

> Do you have a schedule or some constraints in place on how fast you are
> releasing major versions?

Major releases change the first digit; there have been only three such in
all of Python history, counting 0.9.6 and 1.0 and two major releases.
Python 3.0 is at least years away.  The goal between minor releases (bumping
the 2nd digit) is 6 months.  This is what Guido strived for before and after
1.5.2.  1.5.2 was an anomaly.  More about the release philosophy can be
found in PEP 6:

    http://www.python.org/peps/pep-0006.html

> If you could please please please slow your major release rate down,
> and  make that an "official" policy

Why do you feel you can't slow down your adoption of new releases?  We'll
(PythonLabs) package and distribute bugfix releases in a series for as long
as people in the community care enough about a Python version to produce
bugfixes releases for it (speaking of which, you can think Michael Hudson
for 2.2.1 next week, and Anthony Baxter for 2.1.3 -- although I'll be
suprised if you do <wink>).

> I would reconsider my departure from Pythondom. It would be nice to see
> major releases at a _minimum_ of 18 to 24 months apart.

That's much faster than major releases happen now, as Python uses the word.
It's much slower than minor releases happen now.  If you only care to
upgrade every 2 or 3 years (whatever), you're not required to pick up
releases any faster than that.  It's not like you paid us for a subscription
and are going to get cheated out of your money's worth if you refuse to
accept delivery <wink>.

> And frankly I would think the law of diminishing returns has to kick in
> at some point; even if you disregard my request I would think it would
> become progressively more difficult for you to justify new releases all
> that often.

It's actually progressively easier:  Python is in a high growth phase, and
countless people want countless more things from it.  There is strong
sentiment for scaling way back on changes to the language proper, but that
leaves person-centuries of desirable work in improving the libraries and
implementation.

> When I see that happening (maybe in a few years?) then I would consider
> rejoining the ranks of Python programmers.

When the release rate slows to what you'd like to see, I'd take it as a
strong indicator that Python may be entering its death phase.  Indeed, I
took the long quiet after 1.5.2 as a sign that Python was in trouble, and it
turned out it was (note that Guido left his then-employer to get on with
2.0; apply some imagination).  It's not in trouble now, and frequent
releases are a sign of health for an open source project of this nature.





From bokr at oz.net  Tue Apr  9 21:39:31 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 01:39:31 GMT
Subject: class callable with any method name
References: <9d095c5a.0204091256.4d88a246@posting.google.com>
Message-ID: <a9054j$iki$0@216.39.172.122>

On 9 Apr 2002 13:56:09 -0700, mathieu543 at hotmail.com (Imer) wrote:

>Hi, I'm trying to find a way to implement a class that would allow any
>method name to be called.
>When that happens, a special method would be called and the name of
>the method that was originally called would be passed as an argument.
>
>For instance:
>########
>class A:
>  def __method_called__ (self, method_name, method_argument):
>    print "The method", method_name, "was called with the argument",
>method_argument
>########
>
>Then the following code:
>########
>a=A()
>a.this_is_a_random_method(arg=1)
>a.yet_another_random_method(param=3)
>########
>
>Would print out:
>########
>The method this_is_a_random_method was called with the argument
>{'arg':1}
>The method yet_another_random_method was called with the argument
>{'param':3}
>########
>
>Is there a way to do that ?
>
>Thanks.

No guarantees, but try this:

 >>> class A:
 ...     def __getattr__(self, attname):
 ...         if attname.startswith('__'): return self.__class__.__getattr__(attname)
 ...         self._attname = attname
 ...         return self._special
 ...     def _special(self, *args, **kwds):
 ...         print 'The method "%s" was called with the arguments\n%s, %s' % (
 ...             self._attname, args, kwds
 ...         )
 ...
 >>> a=A()
 >>> a.this_is_a_random_method(arg=1)
 The method "this_is_a_random_method" was called with the arguments
 (), {'arg': 1}
 >>> a.yet_another_random_method(param=3)
 The method "yet_another_random_method" was called with the arguments
 (), {'param': 3}
 >>> a.x()
 The method "x" was called with the arguments
 (), {}
 >>> a.y
 <bound method A._special of <__main__.A instance at 0x007D39E0>>
 >>> a._attname
 'y'
 >>> a.z
 <bound method A._special of <__main__.A instance at 0x007D39E0>>
 >>> a.z(k=3)
 The method "z" was called with the arguments
 (), {'k': 3}
 >>> a.z(1,2,3)
 The method "z" was called with the arguments
 (1, 2, 3), {}
 >>> a.z(1,2,3, i=4, j=5)
 The method "z" was called with the arguments
 (1, 2, 3), {'i': 4, 'j': 5}

Regards,
Bengt Richter


From huaiyu at gauss.almadan.ibm.com  Tue Apr  2 20:34:30 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 3 Apr 2002 01:34:30 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com>
Message-ID: <slrnaakn16.17k.huaiyu@gauss.almadan.ibm.com>

I don't think Guido follows this thread. I hope someone at least refer to
him Laura's and Alex' excellent articles.

Following is my $0.02 from algebraic point of view.

Several aspects need to be considered for a boolean type in a language:

- boolean objects (True, False)
- boolean algebra (and, or, not, xor, ...)
- boolean functionality (if, while, ...)
- IO representation

Python currently does not have boolean objects, but it has boolean-like
operators operating on all the objects.  The algebra is homomorphic to a
boolean algebra by the mapping operator.truth.  In other words,

	truth(a and b) == truth(a) and truth(b).

Similarly for 'or' and 'not'.  

The language structures that expects booleans accept all objects, and are
consistent with the truth mapping.  For example, if a <==> if truth(a).

It also happens that the range of operator.truth and 'not' consists of
integers 0, 1, but this is not necessary.

The operator truth divides all objects into two sets, those representing
nothing (mapped to 0) and those representing something (mapped to 1).  Under
this partition, we could regard the 'object' type as the boolean type,
except that equality based propositions do not work, such as

	not (a==b)  ==>  (a == not b).

This is one reason to introduce a boolean type (normalization). Another is
IO representation, which we will come to later.  There can be several
different ways to introduce a boolean type:

1. as a completely new and separate type, with strict and correct algebra.
2. as a subtype of int which is self contained with correct algebra.
3. as a new type with alternative printing methods, but automatically
   coerceable to integers, as normalization of boolean values.

The first two alternatives merely add new objects that could be classified
as "something" or "nothing".  They still need truth (or bool) to "normalize"
to booleans values.  They are both viable.  The first might be better if
booleans are equivalent to the partition of objects, so that we could say

	a in True
	b not in False

However, most of their benefits might be obtained by just having a few
builtins, such as truth, True and False.  The equality and IO problems are
solved using

	truth(a) == truth(b)
	print truth(a)
	a = False

Unfortunately, the proposal on the table is alternative 3, not just as
additional representations of booleans, but as normalization of booleans.
It is not consistent in the sense that it tries to normalize boolean values
without dropping the polymorphism of the boolean-like algebra and controls
structures.  It cannot solve the printing problem either because all the
true objects besides integer 1 would still be printed as themselves.

As Laura pointed out, making the distinction between something and nothing
(as opposed to true and false) is one of the reasons why well-written Python
programs are so elegant.  If we want to keep the polymorphism, ie. if we
want 'true' and 'false' to be representable by a great many objects, it is
futile to come up with an implicit printing solution for booleans.  Saying
print truth(a) is much more reliable and explicit and not much more
cumbersome.

Huaiyu


From prova at prova.it  Sun Apr 28 12:52:44 2002
From: prova at prova.it (Max)
Date: Sun, 28 Apr 2002 16:52:44 GMT
Subject: Still on py2exe - solved!
References: <bWSy8.29406$8D3.782924@news1.tin.it>
Message-ID: <wNVy8.29716$8D3.797588@news1.tin.it>

"Utente" <prova at qui.com> ha scritto nel messaggio
news:bWSy8.29406$8D3.782924 at news1.tin.it...
> Excuse me for bothering...

[...]

Very simple solution, once known...
I used the shelve module, that relies on the dbhash module (under windows).
Simply, i forced the import of the dbhash module with "--force-imports
dbhash" and it worked even "compiled"...





From amuys at shortech.com.au  Tue Apr  9 21:30:45 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 9 Apr 2002 18:30:45 -0700
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <7934d084.0204091730.f9df19d@posting.google.com>

ryanmorillo at hotmail.com (Ryan) wrote in message news:<4a7fde99.0204081858.58baabdd at posting.google.com>...
>     I have been doodling with python (nasty habit I know but I have to
> do somthing with my snake) and think I may be ready to learn a second
> language.  
It's always time to learn another language, I haven't found a language
yet that hasn't taught me to program better in the others.  Even VB
(yuck!) taught me the power of graphical interface builders when
writing graphical interfaces (of course the lanugage itself just
taught me how to run away *fast* ;).

> I am leaning towards C over C++ and Java even though I
> loose objects it will ground me firmly in functional programing.  I am
> wondering what book would be the best for a begginer? 

C maybe, Java maybe, I would recommend you wait on C++ until you have
a firm grasp of either or both of the others (Java probably more then
C).  (btw, as others have already pointed out, you probably mean
procedual, rather than functional.  At some point you should
definately learn a functional language as well, Haskell appears to be
the functional language of choice 'round here).

> The two I have
> seen that look interesting are
> C for Dummies (parts one and two) or 

Simple answer... No.

> The C Programming Language
> by Brian W. Kernighan, Dennis M. Ritchie.

If you decide to learn C, Yes.

Should you decide to consider Java I personally found
 The Java Programming Language
 by Ken Arnold and James Gosling
a good intro.

> I am looking for some input from someone who knows and knows the
> problems I will face coming from a __nice__ language like python.

Pointers.  Pretty much that's the number one issue most people have a
problem with coming to C from anywhere.  OTOH don't be afraid to
tackle them head-on, I haven't seen clean C code that dosn't use them
extensively, and I've seen some appalling code resulting from attempts
to avoid using them.  Of course this is why most other languages (such
as python) provide other, safer, ways of solving the problems pointers
solve in C.

Andrae Muys


From theivo at abv.bg  Sat Apr 13 08:35:07 2002
From: theivo at abv.bg (I.C.)
Date: 13 Apr 2002 05:35:07 -0700
Subject: dos promt simulation
Message-ID: <6ac1ecad.0204130435.93d0ad3@posting.google.com>

Hello,
I would like to simulate a simple dos prompt using a Tk window. I use
os.popen(command) to run some commands like 'echo', 'dir' etc... Then
I read from the pipe and it works, but if I try to execute a command
like 'time', which expects input, I can't read from the pipe. Is it
possible to simulate such commands, without passing some
parameters('time 22:10')?

Thank you for the help !!!

Ivo


From vvainio at tp.spt.fi  Fri Apr 19 09:43:25 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 19 Apr 2002 06:43:25 -0700
Subject: ActivePython vs. Python
References: <mailman.1019201079.9672.python-list@python.org>
Message-ID: <ad496f8.0204190543.21f046a4@posting.google.com>

"Brian Quinlan" <brian at sweetapp.com> wrote in message news:<mailman.1019201079.9672.python-list at python.org>...

> > There was a long period when nothing happened with ActivePython, which
> > no doubt alarmed many people.
> 
> It's not like ActivePython locks you in; it's just a nicer packaging of
> PythonLabs Python. 

That's true. In fact I'm running ActivePython on windows platforms myself.


From loewis at informatik.hu-berlin.de  Sun Apr 28 13:39:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 28 Apr 2002 19:39:27 +0200
Subject: struct module - '<q' equivalent for old python?
References: <mailman.1020011569.12050.python-list@python.org>
Message-ID: <j4znznafjk.fsf@informatik.hu-berlin.de>

jaf <jaf at lcsaudio.com> writes:

> Does anyone know of a good alternative method I could use as a fallback
> in this case?  Some of the platforms I need to support don't have
> Python2.2 readily available, so I'd like to be able support older
> versions of Python, if possible.

Implement it yourself; it is just a few lines of bit twiddling.

Regards,
Martin


From aleax at aleax.it  Mon Apr 15 18:27:56 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 22:27:56 GMT
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
Message-ID: <MtIu8.18713$b62.536574@news1.tin.it>

Duncan Smith wrote:
        ...
> Identification        Var1    Var2    Var3    Var4    Var5 ...
> 
> 0000000000001    N        0           2           3        0
> 0000000000002    N        0           2           2        0
> 0000000000003    N        1           3           3        1
> 0000000000004    Y        0           2           2        2
> 0000000000005    N        0           2           1        0
> 
> def do_stuff(filename, S):  #(S is floating point)
>     f = open(filename, 'r')
>     lines = f.readlines()
>     f.close()
>     lines = [line.split()[1:] for line in lines]
>     lines = [line for line in lines if line != []]
>     lines = lines[1:]

You can collapse this as much as you wish -- hard to say what
will have best performance, but perhaps something like (untested
but I hope the intention is clear)

    lines = [ tuple(x) for line in lines for x in (line.split()[1:],) if x]

If the file is very large, in Python 2.2, you might be better off by
iterating on the file rather than slurping it in in one go:

    lines = [ tuple(x) for line in open(filename)
        for x in (line.split()[1:],) if x]

If you need to select from each line's tuple of variables only some,
as you indicate as one possibility, then maybe:

    lines = [ tuple([x[i] for i in takethese])
        for line in open(filename)
            for x in (line.split()[1:],) if x]

where takethese is the sequence of indices you need.

Each of these ideas ends up with lines as a list of tuples in
memory.  Alternatively, you could produce each tuple on the
go as you iterate below, rather than 'for line in lines:'; again,
either approach could be faster, depending on filesize vs
memory issues.

>     dict = {}
>     for line in lines:
>         my_key = tuple(line)
>         if dict.has_key(my_key):
>             dict[my_key] += 1
>         else: dict[my_key] = 1

Having made of lines a sequence of tuples, I'd code this as:

adict = {}
for tup in lines:
    adict[tup] = 1 + adict.get(tup,0)

particularly avoiding shadowing the name of builtin type dict
(in 2.2) with a local variable (one of my pet little hates:-), but
for speed purposes the win would be to avoid the if/else test.

>     U = P = 0
>     for value in dict.values():
>         if value == 1:
>             U += 1
>         elif value == 2:
>             P += 1

Again I think it might be faster to avoid the if/else with yet
another dictionary:

counts = {}
for count in adict.values():
    counts[count] = 1 + counts.get(count, 0)

U = counts.get(1, 0)
P = counts.get(2, 0)

>     return U*S/(U*S+P*(1-S))

This one can stay:-).


Alex



From tejarex at yahoo.com  Mon Apr  8 21:07:44 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 09 Apr 2002 01:07:44 GMT
Subject: Norton problem
References: <mailman.1018285042.24105.python-list@python.org>
Message-ID: <A9rs8.186857$VJ1.15080324@bin3.nnrp.aus1.giganews.com>

"Angus Duncan", apparently a victim of Norton FUD, wrote

>My computer keeps crashing when I download new Norton Anti Virus
updates. I >undersstand from them that this is due to Python 2.0.

I *really* doubt this.

>I am not familiar with the program or what I use it for. where would
I find it and how
> can I Take it off when doing new updates.

>I attach the message fom Norton.

I assume you have Win98SE?  Even if so, you only have Python2.0b2 on
your machine if you or someone intentionally downloaded this
experimental version from the site and installed it (about 6-8 months
ago, with default location = C:\Program Files\Python.. .  Is it
there?).  Even if it is loaded, it only has an effect if it is running
a Python program and the program uses an os.popenx() function -- which
is pretty rarely used.  According to their message, you also need to
have NAV installed and running in Auto-Protect mode.  I suspect that
you need to reinstall NAV.

Terry J. Reedy

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

>Error: "An exception 0E has occurred at 0028:C02AD9D0... It may be
possible to >continue normally" when using Python 2.0

Is this (or the longer version below) the error message you saw?  Were
you using Python at the time?  (Do you see it with Cntl-Alt-Del?)
===========================================================
# Norton quote not quoted because of posting being html instead of
plain text.

Situation:
You are using Norton AntiVirus 2000/2001(NAV) under Windows 98 SE and
have Auto-Protect enabled. You have Python 2.0 beta 2 from PythonLabs
installed. While using certain features in Python, you see a blue
screen with the error message "An exception 0E has occurred at
0028:C02AD9D0 in VxD IFSMGR(04) + 000050E4. This was called from
0028:C000B511 in VxD VMM(01) + 0000A511. It may be possible to
continue normally."

Solution:
The solution depends on your version of NAV:

NAV 2001
An update is available that fixes this problem. Please run LiveUpdate
to download the program update.

NAV 2000
PythonLabs is aware of this problem. The following information is from
the PythonLabs Web page:

  "Incompatibility warning: Norton Antivirus 2000 can cause blue
screen crashes on Windows 98 when a function in the os.popen*() family
is invoked. To prevent this problem, disable Norton Antivirus when
using Python. (Confirmed on Windows 98 Second Edition with Norton
Antivirus version 6.10.20. The same Norton Antivirus version doesn't
have this problem on Windows 2000. Norton Antivirus version 5 on
Windows 98SE doesn't have this problem either.)"

For additional information, go to the following Internet address:

http://www.pythonlabs.com/products/python2.0/download_python2.0b2.html

To use the Python program, you can disable Auto-Protect while using
Python.

NOTE: If you disable Auto-Protect, then you are no longer protected
against viruses when you open or save files.

To disable Auto-Protect, right-click the Norton AntiVirus icon in the
System Tray, and then click Disable Auto-Protect. Be sure to enable it
again as soon as you finish using Python.

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

Thank you for your help





From phr-n2002a at nightsong.com  Fri Apr  5 22:49:54 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 19:49:54 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com>
Message-ID: <7xadshtrj1.fsf@ruckus.brouhaha.com>

Peter Hansen <peter at engcorp.com> writes:
> No, that's "stagnant"...
> 
> Stable is also a good adjective for it, since most of the older
> code still runs unchanged, and Python from ten years ago is
> still very recognizable to new coders today.

Stable doesn't just mean old scripts run without change in new
interpreters.  It also means new scripts usually run without change in
old interpreters.  That's what keeps breaking in Python.  

> Evolving is another word we need to use, though, and that's what 
> leads some to feel uncomfortable or downright scared of the situation.
> 
> Embrace change!  The changes make Python that much more effective
> and strong.  

That is true, but the marketing is written to make people think the
evolution has ALREADY HAPPENED, and Python is sufficiently effective
and strong to not need further rapid change.  C is not stagnant (C
implementations keep improving) but it is mature and stable.  If I
download an interesting new C program today, I can probably build and
run it on my 5 year old Debian 1.x system without upgrading my C compiler.
I generally can't do that with Python, because of all the new features
that keep getting introduced.

> On the other hand, beware change!  At least for the sake of change.  
> I think those driving the changes need to be ever more cautious in 
> allowing changes to make it through, and have an ever-rising threshold 
> of added value before something is accepted.

Most recent changes in Python are good (list comprehensions, x in dict,
etc) but I'd rather that they appear only in major releases (2.0, 3.0, etc)
which happen every 2 years or so, rather than in the minor releases
that come out every few months.

A mature language is like a grown adult who needs to buy new clothes
every few years.  Python is more like a rapidly growing child who
needs bigger clothes every few months.  There's a tension between
language buffs looking for excitement and growth, and production users
who want a full-grown, mature language that can be left alone for long
periods.  There's nothing wrong with youthful flexibility but it
shouldn't be marketed as something that it's different from.


From tdelaney at avaya.com  Wed Apr 10 01:15:32 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 10 Apr 2002 15:15:32 +1000
Subject: What has PEP 285 done to us?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A263@natasha.auslabs.avaya.com>

> From: Gerhard H?ring [mailto:gerhard at bigfoot.de]
> 
> Lumberjack wrote in comp.lang.python:
> > Tim Peters <tim.one at comcast.net> wrote:
> >> PEP 285 taught us that people don't agree, and that some people
> >> disagree in disagreeable ways.  Frankly, that wasn't news,
> > 
> > [drivel attacking Tim and Guido]
> 
> Why would anybody care if you aren't man/woman enough to post with
> your real name?

Actually, as a lumberjack, it really doesn't matter what sex he/she is ;)

I'm hoping the post was just a joke, although it didn't carry any of the
characteristic humour markers ...

Tim Delaney




From chris.gonnerman at newcenturycomputers.net  Tue Apr 16 00:21:59 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 15 Apr 2002 23:21:59 -0500
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com>
Message-ID: <00b401c1e4fe$41b82da0$0101010a@local>

----- Original Message -----
From: "Mike Coleman" <mkc+dated+1021521407.f909ec at mathdogs.com>


> Does anyone know why strings (i.e., those of length >1) don't share their
data
> in Python?  Since their immutable, it seems like this would be the obvious
> thing to do.  So, for example, the space behavior of this code could be
linear
> rather than quadratic/horrific:
>
> d = {}
> for i in xrange(100000000):
>     d[mybigstring[i:]] = mybigstring[i:]

You, the human, have an implicit semantic understanding that the RHS
of the assignment above is the same as the key in the LHS expression.
The compiler knows no such thing; so it computes the substring twice
(for each loop pass).  You are, in effect, asking for it.

If you mean, why don't substrings share space with parent strings,
hmm, that's an interesting question.  I can see that a string object
implemented something like this struct:

    struct ExampleString {
        int len;
        char *buf;
    }

would allow that, and it sounds like an interesting optimization.  If
the string object is like this:

    struct AnotherExampleString {
        int len;
        char buf[1];
    }

(where the instances are overallocated to contain the data) would not
allow it.

Are you certain Python doesn't do it?  Can you submit a patch to
implement it?  (PEP + Patch == Implemented Change in many cases).

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





From geoff at gerrietts.net  Mon Apr  1 16:58:38 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 1 Apr 2002 13:58:38 -0800
Subject: runtime location discovery
In-Reply-To: <20020401084421.3aa73878.logiplexsoftware@earthlink.net>
References: <20020330014832.GF7602@isis.gerrietts.net> <20020401084421.3aa73878.logiplexsoftware@earthlink.net>
Message-ID: <20020401215838.GE6604@isis.gerrietts.net>

Quoting Cliff Wells (logiplexsoftware at earthlink.net):
> On Fri, 29 Mar 2002 17:48:32 -0800
> Geoff Gerrietts wrote:
> >     Image.open(os.path.join(root_location, "image/naked_chicks.png"))
> 
> Maybe if you post this image we could come up with some ideas ;)

Hee. Well, the image is *actually* "image/error.png" but I was hoping
to get my message scored higher with the other. ;)

--G.

-- 
Geoff Gerrietts             <geoff at gerrietts dot net>
  I AM YOUR KING! BOW BEFORE ME, PEASANT!   -- Dogbert



From brian at mcconnell.net  Mon Apr 15 16:39:16 2002
From: brian at mcconnell.net (Brian McConnell)
Date: 15 Apr 2002 13:39:16 -0700
Subject: Worldwide Lexicon project, call for volunteers
References: <e2f36e18.0204111716.78d6f0a6@posting.google.com> <7xsn60h5at.fsf@ruckus.brouhaha.com>
Message-ID: <e2f36e18.0204151239.306b1819@posting.google.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote in message news:<7xsn60h5at.fsf at ruckus.brouhaha.com>...
> Why yet another protocol instead of using DICT?

Three reasons....

1. DICT is a TCP protocol, and is harder to implement. WWL is built on
top of SOAP, which is supported in most scripting and programming
languages. In most environments, invoking a SOAP method on a remote
server is about the same as invoking a local subroutine. So, for most
developers, it is much easier to deal with. This isn't a criticism of
DICT, it was developed about five years ago, before XML, SOAP, etc
came along.

So imagine, for example, that you run a Spanish-French dictionary
server. You know basic web scripting and MySql, but aren't a transport
protocol expert. Implementing DICT on your server would be a chore.
Implementing WWL is much simpler as you simply update your existing
CGI script to reply by returning its results via a SOAP method (same
procedure as returning results from a function call).

2. WWLP provides a mechanism for automatic server discovery. The
client application can therefore talk to WWL servers as needed. This
also allows for WWL servers to join and leave the network as needed.
DICT doesn't really address this. Clients have to know where to find
servers. With WWL this can be automated. The client talks to one or
more supernodes, and can then switch from WWL server to WWL server for
different types of queries.

3. Real-time feedback. WWLP includes a couple nifty features that
allow real-time interaction between dictionaries and their users. One
method allows users running a client program (think seti at home for
dictionaries) to poll WWL servers that allow public contributions.

The last point I would make is that WWLP is not being proposed as a
standard. I like to use GNUtella as a model. If it turns into a widely
used service, that's wonderful. However, I am not interested in going
through the whole IETF process. I've done that before, and spent a lot
of time getting bogged down in politics. I'd rather build an
interesting service, if people use it great, if not, I'll still have a
good time working on it.

You can find a draft copy of the spec at www.worldwidelexicon.org

Thanks,

Brian McConnell


From aleax at aleax.it  Sun Apr  7 05:32:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 07 Apr 2002 09:32:32 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018153694.532.python-list@python.org> <7xelhsvxsz.fsf@ruckus.brouhaha.com>
Message-ID: <QmUr8.56786$pT1.1621156@news1.tin.it>

Paul Rubin wrote:
        ...
> The stability-valuing approach would be DON'T PUT NEW FEATURES into
> releases like 2.1 without good reason to get the particular features
        ...
> The proposal of having separate "development" and "stable" release
> tracks like Linux has sounds like a promising direction for keeping
> both camps (language development enthusiasts and would-be production
> users) happier than they are in the current situation.

Not really, because "new features" *DO* get into Linux "stable" releases,
as long as they don't break anything in terms of backwards compatibility.
(That's the theory -- defects in execution often mean that the "stable"
releases aren't all that stable, but that's another issue).

And that is exactly what you and Mr. Logajan seem dead set against -- and
it's utterly different from the concern which I was trying to address with 
my dual-track proposal (now being discussed in python-dev).  That concern
being the reluctance of some substantial software development shops to 
choose a platform for which the only choices are [a] feature-frozen, 
nothing-new, bugfix-only releases or [b] a frequent stream of releases that 
break previously-correct code.  [b] is perceived as bad because of the
need to rework existing stable code developed in said shop (or purchased
from third parties).  [a] is perceived as bad because, as Eric Raymond so
well puts it in his "Magic Cauldron" essay, the perceived value of software
lies in very good part, not in its bits, but in the expected future stream 
of maintenance, enhancements and fixes.  A "frozen" release, in which no
new feature will ever go, lowers expectations on that score too drastically
to let the managers of those shops feel good about it.

Your concerns on the other hand seem all tied to what OTHER people
may be doing out there.  You WANT case [a] software -- but you're
dead set against OTHER people getting to choose [b], because they
may then use [b], the bleeding-edge experimental version, to write some
neat script that you'd like to run (but can't while staying stuck at [a]).

I don't think that any reasonable variation on my proposal will satisfy
your concern -- even assuming there IS a large constituency for "not
letting OTHERS have new features [so they won't write software that
uses them -- so you won't receive that software and need upgrades]".
I'm not sure it's all that large and substantial a constituency.  And even
if it were, I don't see how to meet that concern without driving away
an important group of neophile/hackers to more welcoming languages
(or a fork): the "experimental" branch of releases, if existing at all, 
would still produce the scripts you don't want others to write -- if it
did not exist, neophiles would go elsewhere.


Alex



From dkuhlman at cutter.rexx.com  Sat Apr 13 11:50:34 2002
From: dkuhlman at cutter.rexx.com (Dave Kuhlman)
Date: Sat, 13 Apr 2002 08:50:34 -0700
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <7E283A6EA4F0DBD6.B68D98BE8C574F5D.CE8B287F1A68C959@lp.airnews.net> <LFGz2JAYZxt8EwR6@jessikat.demon.co.uk> <459A680B92F06092.DD5B3EAA28701812.3A8CCB250DB66074@lp.airnews.net>
Message-ID: <3CB853CA.8050507@cutter.rexx.com>

Cameron Laird wrote:

> Folklore has it that Tcl *did* travel to Mars, and its
> interpretive nature was crucial in responding to some
> mission contingency; the control team needed to upload
> new code during flight.  We ought to document this ...
> 
> While I understand the observations that Python, Tcl, 
> and so on surely aren't safe enough for medical devices,
> nuclear plants, air traffic control, military devices,
> and so on, I have reasonably certain knowledge of instances
> of each of these.
> 


The idea that Python, Tcl, and so on aren't safe enough for medical 
devices sounds like conventional wisdom to me.  Someday a scripting 
language will come along which is so good that it proves this 
conventional wisdom wrong, and does this so conclusively that we are 
forced to give up that idea.  I believe that Python was/is/will be that 
language.

Are there examples and evidence that Python fails when applied to 
long-running, mission critical tasks?

By the way, I write C++ code.  And, the idea that C++ code is safer than 
Python code seems screwy to me.  Can you say "seg fault"?  I thought so.

   - Dave





From aleax at aleax.it  Mon Apr 29 17:05:41 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 21:05:41 GMT
Subject: cPickle problem
References: <Ssiz8.63$bK3.1993791@newnews.cc.stevens-tech.edu>
Message-ID: <FAiz8.78905$vF6.2347612@news2.tin.it>

Vojin Jovanovic wrote:
        ...
>  def __getattr__(self,arg):
>   try:
>    exec 'result =' + self.e[arg]
>   except:
>    return '#U'
>   return result

Note that this __getattr__ answers "yes, here it is!" WHATEVER
attribute the object is being accessed for.

Now suppose that Python itself is accessing an instance of this
class to check if it has a certain special-method.  What happens?

Python tries accessing attribute (e.g.) '__getstate__', ready for
an AttributeError exception to mean "no such attribute", or else
to get and use (call) the result.

And the object answers "Sure, I DO have a __getstate__, and
its value is '#U'".

So of course Python tries calling '#U' when NOT ready for any
exception, since by spec, the attribute __getstate__, if it exists,
must be a callable (typically a method) with certain semantics.
But, in this case:


>>>> cPickle.dump(A,file)
> Traceback (most recent call last):
>   File "<pyshell#62>", line 1, in ?
>     cPickle.dump(A,file)
> TypeError: object of type 'string' is not callable

Indeed, object '#U', of type 'string', is NOT callable.  And yet
that is what the object CLAIMS it has as attribute __getstate__.


Any __getattr__ special method should ALWAYS raise
AttributeError when asked for any name starting and ending
with two underscores, except if it does know the name and
is specifically prepared to return some right, meaningful value
for it.  All identifiers that start and end with two underscores
are "reserved to Python" in that any Python version can add
some of them, so it's totally unsafe to *pretend* you have ones
that you don't actually know about.


Alex



From ChrisE at lantech.com  Fri Apr 26 09:42:32 2002
From: ChrisE at lantech.com (Christopher Encapera)
Date: Fri, 26 Apr 2002 09:42:32 -0400
Subject: Killing a process under Windows NT/2K
Message-ID: <BC74F8F7E87DD511990C000103E2FEFA2FFC@XCHG>



"Ingo Blank" <spam at yourself.com> writes:

> Hi,
> 
> I need to kill a process under Windows (NT/2K), whose PID is known.
> Unfortunately os.kill() doesn't work under Windows.
> How can I achieve this anyway?
> 
> Thanks in advance
> 
>
>This might help. You have to have the win32all extensions installed,
>but so should *everybody* that uses python on win32 anyway.
>

You could also call kill.exe from the WinNT resource kit.  I went looking
for this handy app in the win2k kit the other day and found a gui version
(rkill.exe if I remember) but not the plain cmd line kill.exe, so I just
borrowed it from my WinNT partition and it works fine on 2k...

Christopher Encapera
Lantech.Com
chrise at lantech.com
502-267-4200





From flognat at flognat.myip.org  Tue Apr 16 04:23:21 2002
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Tue, 16 Apr 2002 08:23:21 GMT
Subject: Win32com usage of IE raises exception.. (how do I interpret win32com errs?)
Message-ID: <m33cxwxdae.fsf@flognat.myip.org>

I am trying to reuse some win32com demo that was posted ages ago in
this group, firing up ActiveState Python (i tried the python.org
python+win32com, same 'problem), win32makepy, and the code:

>>> import win32com.client
>>> ie = win32com.client.Dispatch('InternetExplorer.Application')
>>> ie.Navigate("http://www.microsoft.com")
Traceback (most recent call last):  
   File "<interactive input>", line 1, in ?  
   File "C:\Python22\Lib\site-packages\win32com\gen_py\EAB22AC0-30C1-11CF-A7EB-0000C05BAE0Bx0x1x1.py",
        line 1100, in Navigate    return self._oleobj_.InvokeTypes(0x68,
        LCID, 1, (24, 0), ((8, 1), (16396, 17), (16396, 17), (16396, 17),
        (16396, 17)),URL, Flags, TargetFrameName, PostData,
        Headers)com_error: (-2147352567, 'Exception occurred.', (0, None,
        None, None, 0, -2147467259), None)

What have I done wrong? Anyone who has a good class/functions for
using IE as browser, fetching stuff and so on.. 

How do I get into win32com 'programming'/usage?? Read the docs?

        /Andy

p.s. The 'tutorial' I fetched inspiration from is:
        http://mail.python.org/pipermail/edu-sig/2000-September/000636.html

about 1.5 years old.. 

-- 
 The eye of the beholder rests on the beauty!


From sabren at manifestation.com  Tue Apr  2 17:48:08 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Tue, 2 Apr 2002 17:48:08 -0500 (EST)
Subject: ANN: Twisted 0.16.0
In-Reply-To: <3CAA0A6F.1040600@itamarst.org>
Message-ID: <Pine.LNX.4.44.0204021726250.30418-100000@mercury.sabren.com>

Hi Itamar,

Thanks for trying to answer my question. I'm afraid I still
don't follow the logic of the registrations though.

(I'm BCC'ing glyph because I really want to understand this.)

On Tue, 2 Apr 2002, Itamar Shtull-Trauring wrote:

> bin/twistd can load license files, yes. If you don't have
> one it prints a nice long error message (which I removed
> in CVS, so it won't be in 0.16.1).

So registration will *not* be required in the future?

> Glyph has some insane reason he wants it to be in (rumor
> has it a rogue Pavlovian has linked the registration form
> to a Certain Person's cage, giving him a food pellet every
> time someone registers. Try not to overfeed him.)

:) I can see that registration is a great way to track the
growth of the user base, and I understand the extra benefits
mentioned on the how to register page:

  http://www.twistedmatrix.com/documents/howto/register

But I *dont* understand how it has anything whatsoever to do
with stopping piracy. There's just a huge mental gap.

If you want to track the popularity, great! Then say that,
but why tag it as a "help fight piracy" issue when you don't
present an actual argument for how it will lead to reduced
piracy. :)


>  > If so, why? How can piracy be a problem for an open source
>  > project?  How does this licensing scheme prevent it?
> 
> Well, let's quote one the BSA's latest press release:
> 
> """
> Washington, DC (March 21) ?In its ongoing effort to protect consumers
> from Internet software pirates and support a safe, secure and legal
> online community, the Business Software Alliance (BSA) announced that
> last month it executed three unannounced court-ordered seizures on
> individuals whom BSA investigated for selling unlicensed software out of
> their homes in Kentucky, Missouri and Pennsylvania. Under the
> supervision of the U.S. Marshals, BSA carried away computers, equipment
> and CDs used in the mail-order operations.


Um, okay, but I read this to mean that they were taking
closed-source software and selling it, thus stealing money
from the original creators.

Twisted is a free and open source product, isn't it? The
version I have comes with a copy of the GNU LGPL, which
explicity grants me permission to distribute copies. The
first clause even says I can charge for the distribution:

-- You may charge a fee for the physical act of transferring
-- a copy, and you may at your option offer warranty
-- protection in exchange for a fee.

How can I possibly pirate it?


> Now, you wouldn't want any Sikrit PSU Agents showing up at
> your house, asking to see *your* Twisted license, and not
> being able to show one, now would you? And do you really
> trust a bunch of people giving away software for FREE?

:) But I can show them the license. It's in a file called
"LICENSE", and it contains the GNU LGPL.

I'm not opposed to asking people to register, if its
beneficial... I just don't understand your logic, or see how
it relates to the closed-source concept of piracy.

The arguments I've read so far just makes me think that
Twisted is going closed source, and that I should therefore
stop trying to learning it... That would be a big
disappointment for me, because so far everything I've seen
in twisted looks great.

Is twisted going commercial?

Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Give your ideas the perfect home: http://www.cornerhost.com/
 cvs - weblogs - php - linux shell - perl/python/cgi - java
------------------------------------------------------------




From mertz at gnosis.cx  Thu Apr 11 14:13:32 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 11 Apr 2002 14:13:32 -0400
Subject: Help: Python for a big commercial project?
References: <F194tIpqqJgSt51NDa200002509@hotmail.com>
 <mailman.1018464082.15807.python-list@python.org>
 <roy-EB08BB.20470710042002@news1.panix.com>
 <60F7BB039BFDD68F.7FE6889DFED7EA98.93812C99068FD653@lp.airnews.net>
Message-ID: <MJdt8kKkXAtS092yn@gnosis.cx>         

claird at starbase.neosoft.com (Cameron Laird) wrote previously:
|Python style result in code which is more maintainable, only
|a fraction as large in lines-of-source-code, and, remarkably,
|at least as fast as the hand-coded C.  Why and how this latter
|is possible have been explained here often in the past; I'll
|abbreviate the details by just saying that Python's run-time
|library is itself written in C

The other important aspect is the motto/mantra that all Pythonistas[*]
should repeat to them upon waking in the morning:  "Tim Peters is
smarter than I am."

Yours, Lulu...

[*] Nope! There is not an off-by-one error here.

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From aleax at aleax.it  Mon Apr  1 15:23:07 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 20:23:07 GMT
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <a85od6$9i$1@panix1.panix.com> <DlTp8.48032$S52.1614804@news2.tin.it> <a89qjc$ah1$1@panix1.panix.com>
Message-ID: <Lk3q8.49428$S52.1679119@news2.tin.it>

Aahz wrote:

> In article <DlTp8.48032$S52.1614804 at news2.tin.it>,
> Alex Martelli  <aleax at aleax.it> wrote:
>>
>>Actually, this specific problem is there, like several others, only
>>*for MUTABLE objects*.  Take an "immutable" rectangle:
        ...
>>class ImmutableRectangle(object):
>>    def __init__(self, width, heigth):
>>        object.__setattr__(self, 'width', width)
        ...
[I think I forgot to override __delattr__ and raise a
TypeError to shore up the immutability, forgot to
override __eq__ as a hashable should, etc, but anyway,
I hope the concept of immutability _is_ clear, even
though I messed some details up:-)]
        ...
>>Now, subclassing this is perfectly fine:
>>
>>class ImmutableSquare(ImmutableRectangle):
>>    def __init__(self, side):
>>        ImmutableRectangle.__init__(self, side, side)
> 
> Where is the side attribute?

I haven't added it.  If I needed to, it would be:

        def getSide(self): return self.width)
        side = property(getSide)

in the body of ImmutableSquare, right after the def
for __init__.  This seems smoothest.

The alternative approach is perhaps more interesting:

class ImmutableSquare(ImmutableRectangle):
    def __init__(self, side):
        object.__setattr__(self, 'side', side)
    def getSide(self): return self.side
    width = property(getSide)
    heigth = property(getSide)

This 'transparently, retroactively and noninvasively' turns
all of ImmutableRectangle's methods that access self.width
and self.heigth into 'Template Methods' (in the Gof4 DP
sense).  I think it's cool and of course pre-emptively answers
objections from misguided souls about 'but ImmutableSquare
should hold only one reference to its side, not two'.  But it
will make no measurable difference (if ImmutableRectangle
has done serious memory optimization by defining __slots__,
there are two per-instance references willy nilly, and all that
ImmutableSquare should do should be to define its own empy
__slots__ to avoid wasting yet more memory:-), and I think
the 'coolness factor' plays both way -- the first solution is
trivially obvious, the second one is 'clever', so the first one is
more immediately understandable and therefore, I think, it IS
the preferable one in this simple case.

The 'trick' of the "noninvasive TemplateMethod'ing" (turning
'actual' attributes that a base class references on self, into
properties in a derived class, so one can compute them on
the fly) may be worth keeping in mind for possible future needs
where it might offer the best approach, though.  It's not all
THAT tricky, after all -- just enough to make a plainer idea
preferable when one offers itself as evident:-).


Alex



From sholden at holdenweb.com  Tue Apr 30 09:20:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 09:20:39 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <3CCD9649.98AED86F@cascade-sys.com><007401c1efb0$91d9d7e0$445d4540@Dell2> <mailman.1020109852.21265.python-list@python.org>
Message-ID: <eVwz8.13791$xy4.677@atlpnn01.usenetserver.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote ...
> On Mon, 29 Apr 2002 12:03:35 -0700
> Brian Quinlan wrote:
>
> > > > Why the hostility?
[...]
> I have to agree with Brian.  While Lumberjack's response did address
issues
> with the PEP, it was laced with personal insinuations about the PEP's
author
> (and PEP writers in general):
>
> [Lumberjack]
> "How about you and other PEP writers take some responsibilities for your
> fantasies for once?"
> "But no, you'd rather break code and make existing instructional material
> obsolete"
> "The [sic] would be in line with the contempt you seem to have for others.
> "You think the world revolves around your notion of "rightness"?"
> "I believe the organization you belong to has this chant: ..."
>
> The overall tone was anger not reason.  Arguments stated in this fashion
> deserve the same consideration they give to their opponent: none.
>
Good point. Perhaps he should change his nom-de-guerre from "lumberjack" to
"gumby".

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From aleax at aleax.it  Sat Apr 13 18:22:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 22:22:44 GMT
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it> <a9a7sa$qd3$1@reader05.wxs.nl>
Message-ID: <Uc2u8.13848$b62.382031@news1.tin.it>

Jozef wrote:
        ...
>> def relatively_prime(a, b): return gcd(a, b) == 1
> 
> Thats funny, I wanted to do that but I don't know how to use filter with a
> function that takes 2 variables. E.g., to get a list if all numbers less
> than 10 relatively prime to 3, I would write
> 
> filter(relatively_prime, range(1, 10)) # aaah, where does the 3 go???

filter wants a unary function, and, as you saw in Haskell, you do that by 
_currying_ (Haskell Curry is the name of the mathematician who thought
that up...):

> So I learned the lambda way. I also looked at Haskell recently, and  there
> (relatively_prime 3) is a valid function that takes only 1 argument, so no

Python doesn't have built-in curry, so you hack it up in any of several 
ways (of which lambda one), such as a closure:

def relativelyPrimeToN(N):
    def relatively_prime(x, N=N): return gcd(x, N) == 1
    return relatively_prime

and then

filter(relativelyPrimeToN(3), range(1, 10))

curry itself isn't hard to write (see the Python Cookbook for various 
approaches) -- e.g., with nested scopes, and only caring about
positional arguments, again as a closure:

def curry(f, *args):
    def curried(*more_args): return f(*(args+more_args))
    return curried

and then

filter(curry(relatively_prime, 3), range(1, 10))

> problem there. Just a thought: a nice way to get rid of lambda's and save
> a little typing would be if for example
> 
> filter(is_even, alist)
> and
> filter(is_even(_), alist)
> 
> meant the same thing, and in
> 
> filter(relatively_prime(3, _), alist)

Not a chance, IMHO.  _ is a valid Python identifier, used in the 
interactive interpreter (result of the latest expression statement)
and elsewhere (e.g. the gettext module).

If "currying with different patterns of supplied/nonsupplied args"
was crucial (but Haskell doesn't have it... you have to define a
named function for that), you could, I guess, use some other
place-holder object -- but not, I think, with function-call syntax,
or you could never pass that place-holder to any function.
Something like curry(relatively_prime, 3, curry_arg) -- though
here it would have no added value wrt the simpler curry I suggest.


Alex



From Chris.Barker at noaa.gov  Fri Apr 26 15:29:19 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 26 Apr 2002 12:29:19 -0700
Subject: Python and Cocoa?
References: <mailman.1019791185.27819.python-list@python.org>
Message-ID: <3CC9AA8D.99ECC911@noaa.gov>

Brian Quinlan wrote:
> > Does anyone know of anything useable to write Cocoa apps with Python?
> You might want to start with:
> http://www.onlamp.com/pub/a/python/2002/1/31/pythonnews.html

Also, it's no Cocoa, but wxMacPython is coming along nicely, and gives
you a way to write the same code for Mac, Windows, and *nix.

See:
http://radio.weblogs.com/0102677/2002/04/24.html

for a recent update on progress.

-Chris


From peter at engcorp.com  Mon Apr 15 21:39:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 15 Apr 2002 21:39:31 -0400
Subject: Python installation on a Windows shared disk
References: <a9e5u2$1j0u$1@norfair.nerim.net> <slrnablp1j.fa.gerhard@lilith.my-fqdn.de>
Message-ID: <3CBB80D3.8F5FF3EB@engcorp.com>

Gerhard H?ring wrote:
> 
> Eric Brunel wrote in comp.lang.python:
> > Hi all,
> >
> > I'm trying to do an installation of Python 2.1 on a shared Windows 2K disk.
> > People should then be able to use this installation from all PCs seeing this
> > disk. [...]
> 
> This should work:
> 
> - Install Python in a directory on your share
> - Copy the Python$VERSION.DLL from $SYSTEMROOT/System32/ into the
>   directory you installed Python in (in the same one the python.exe
>   went)
> - Put this share in the PATH on your client PCs
> 
> Nothing more should be needed.

I've found it very helpful to have PYTHONHOME set on each 
machine that uses it, but then again the directory with
the executable is not in our PATHs, but instead we use
a batch file in a different folder which is in the PATH
to launch python.exe.  Not sure whether that's why we
thought we needed PYTHONHOME set.

The advantage is possibly that merely by changing 
PYTHONHOME and running a different batch file, we
can support older versions also on the network.

-Peter


From laytonjb at bellsouth.net  Mon Apr 29 19:38:04 2002
From: laytonjb at bellsouth.net (Jeff Layton)
Date: Mon, 29 Apr 2002 19:38:04 -0400
Subject: Newbie: QTable Questions - Part II
Message-ID: <3CCDD95C.280F7DC6@bellsouth.net>

Hello,

   I'm in the midst of learning PyQT (and python more in-depth).
I haven't found much if any documentation about QTable for
PyQT. Andrew Dalke helped me get started (and pointing out
that in my panic I forgot to sit down and think :). Now I want to
modify the appearance of my table created with QTable. In
particular, I want to change the :

(1) column headers (via QHeader I think)
(2) change the column width and row height
(3) Make the cells read-only
(4) Change the color of text in the cells (and maybe the background
    of the cells as well).

Could a kind soul give me a kick in the pants to get me started?

TIA,

Jeff





From rs96057 at hotmail.com  Tue Apr 23 01:47:18 2002
From: rs96057 at hotmail.com (Dimitris Garanatsios)
Date: Tue, 23 Apr 2002 08:47:18 +0300
Subject: Classes: Old or New and How?
References: <Xns91F8BDF3CF983RASXnewsDFE1@130.133.1.4>
Message-ID: <aa2sh6$2o7v$1@ulysses.noc.ntua.gr>

"Philip Swartzleonard" <starx at pacbell.net> wrote in message
news:Xns91F8BDF3CF983RASXnewsDFE1 at 130.133.1.4...
>     I've been using py2.2 for a while now, although i haven't really
> used many of the new features aside from the odd list compt to ease
> initing 2-3 dimensional lists, and some class hierarchy i made off list
> some months ago. (No generators or anything like that).
>     What i'm wondering is this: when i define some random class,
> weather it's something small (initially) like a Link_Node (datastructures
> class =) or a GL_Color, or something way comlicated like ARCS_FOV or a
> game world controller, in anycase they don't require any special
> functionality -- just data and regular methods. When i start a class like
> this, should I bother explicetly deriving from Object, or should I set
> that flag that makes everything like that, or should I just not bother?
>
> Thanks.
>
> --
> Philip Sw "Starweaver" [rasx] :: www.rubydragon.com

It doesn't matter whether you use "object " as a base class/type or not in
Python 2.2. Old-style classes still work as well and (for the moment) you
are free to choose between styles any time you create a class. Using the new
style should not make any difference from the class beeing used the
old-style way, though it offers some interesting options...

I think that, new style is some kind of an extension of the old one and that
in later versions of Python the default style will change to the new (maybe
a unified one).

Dimitris







From no-news at darkeldar.fsnet.co.uk  Fri Apr 12 17:44:51 2002
From: no-news at darkeldar.fsnet.co.uk (Grim Archon)
Date: Fri, 12 Apr 2002 22:44:51 +0100
Subject: Tkinter Win32 key handling problem?
Message-ID: <a97kip$93m$1@newsg1.svr.pol.co.uk>

I hope you can help me!

I have been developing a Windows based program with a TKinter GUI that runs
batched test scripts on devices connected to serial ports. In order to debug
the scripts and check the syntax of  'AT' commands I have added a simple
terminal emulation module. This is currently working well except for one
very annoying feature -

Using the Shift key to access characters such as '&' or '+' has a big delay
(seconds) before the user sees the program respond. I am using a scrolled
text widget and have bound the <key> event to an input handling method. I
suspect it is something to do with multiple key presses but I don't know.
Has anyone seen this before and is there a way a round it? If not what other
methods do people use?

Grim






From sholden at holdenweb.com  Wed Apr 24 08:32:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 08:32:25 -0400
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com> <aa5f7l$uck$1@news.nextra.cz>
Message-ID: <nExx8.45080$0x3.1309@atlpnn01.usenetserver.com>

"Petr Prikryl" <Answer.via.news.please.prikrylp at nospam_skil.cz> wrote in
message news:aa5f7l$uck$1 at news.nextra.cz...
> Well, I have meant what quotes should I use in cases
> when only single or only double are to be used (i.e.
> when I do not want to include the opposite inside
> the string.  Read it what *style* is prefered.
>
There are no quote police in the Python world, only the PSU, and you are
unlikely to get trouble from them as long as you





From fredrik at pythonware.com  Wed Apr 10 12:14:15 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Apr 2002 16:14:15 GMT
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <mailman.1018385900.7252.python-list@python.org> <u8z7valyh.fsf@python.net> <yu99662zmufg.fsf@europa.research.att.com>
Message-ID: <rxZs8.254$p56.69633@newsb.telia.net>

Andrew Koenig wrote:

> The point of this example is the . operator, which takes a pattern
> (by value) as its left argument and a variable (by reference) as its
> right argument.  The result is a pattern that matches the same strings
> as the left argument, but if the match is successful, it assigns
> the substring matched to the variable that is the right argument.

you could of course use a RE-like approach:

    m = match(name, pos(0) + break("@", "id") + "@" + rpos(0, "domain"))
    if m:
        print m.group("id", "domain")
        # or perhaps: print m.id, m.domain

</F>

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




From philh at comuno.freeserve.co.uk  Wed Apr 10 13:54:00 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 10 Apr 2002 18:54:00 +0100
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>
Message-ID: <slrnab8v1o.emj.philh@comuno.freeserve.co.uk>

What have the PEPs ever done for us?

(Apart from the roads. The roads go without saying.)

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From timd at macquarie.com.au  Sun Apr  7 19:56:22 2002
From: timd at macquarie.com.au (Timothy Docker)
Date: 08 Apr 2002 09:56:22 +1000
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84>
Message-ID: <m4isn6712s9.fsf@macquarie.com.au>

James Logajan <JamesL at Lugoj.com> writes:

> And frankly I would think the law of diminishing returns 
> has to kick in at some point; even if you disregard my request I would 
> think it would become progressively more difficult for you to justify new 
> releases all that often. When I see that happening (maybe in a few years?) 
> then I would consider rejoining the ranks of Python programmers.

Out of curiousity, which language are you using in the meantime? I
presume that you find it suits you better in the tradeoff between a
useful feature set and rate of change.

Tim


From nbecker at fred.net  Mon Apr  1 21:56:44 2002
From: nbecker at fred.net (nbecker at fred.net)
Date: 01 Apr 2002 21:56:44 -0500
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org>
Message-ID: <m3d6xibyfn.fsf@nbecker.fred.net>

Keep lobbying Redhat.  The correct solution is for them to put

#! /usr/bin/python1.5

If they need a specific version.  I argued this with them already.
Keep up the pressure.  Maybe they will get it.


From marklists at mceahern.com  Wed Apr  3 23:50:18 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 22:50:18 -0600
Subject: turning a string into an object name
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBOECAKJAA.marklists@mceahern.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAECFKJAA.marklists@mceahern.com>

Here's some sample code using inspect.  There are two files:

	factory.py
	classes.py

Run factory.py and watch the results.

Bad names, I know.  Oh well.  Here they are:

#! /usr/bin/env python

import inspect
import classes

def getClassByName(module, className):
    all = [x[1] for x in inspect.getmembers(module) if inspect.isclass(x[1])
and x[0] == className]
    if all:
        return all[0]
    else:
        return all

x = getClassByName(classes, "foo")
if x:
    x()

y = getClassByName(classes, "bar")
if y:
    y()

# classes.py
class foo:
  def __init__(self):
    print "Hi, I'm an instance of foo."

class bar:
  def __init__(self):
    print "Hi, I'm an instance of bar."






From s328940 at student.uq.edu.au  Tue Apr 30 00:43:48 2002
From: s328940 at student.uq.edu.au (Alejandro Dubrovsky)
Date: Tue, 30 Apr 2002 14:43:48 +1000
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <3CCC730A.FC60D13F@alcyone.com> <3ccd1ed9$0$15477$afc38c87@news.optusnet.com.au> <837kmqrvxl.fsf@panacea.canonical.org>
Message-ID: <3cce1eaf$0$15477$afc38c87@news.optusnet.com.au>

On Tue, 30 Apr 2002 08:17:58 +1000, Kragen Sitaker wrote:


> Unfortunately not.  You can't marshal lambdas either.
> 
> Generally, you can refactor code that uses lambdas into code that uses
> object methods; your closure data goes into object attributes, and each
> lambda becomes a class.  Or you can just use named functions, especially
> if you don't have closure data.
 
i do have closure data, but making it a class seems like a good solution
i hadn't thought about.  Thanks.
Alejandro


From donn at drizzle.com  Wed Apr 24 01:19:55 2002
From: donn at drizzle.com (Donn Cave)
Date: Wed, 24 Apr 2002 05:19:55 -0000
Subject: stdin - seek() issue on Linux.
References: None <mailman.1019617785.428.python-list@python.org>
Message-ID: <1019625593.731162@yasure>

Quoth "Noah" <noah at noah.org>:
| On the other hand, perhaps the File object could be more consistent.
| In some cases seek() will throw an IOError exception, but in
| other cases it will do nothing (no-op). For example if you
| call seek on disk files opened for writing in append mode (mode 'a') 
| then seek will be do nothing (no-op)... Hmmm... 
| I wonder why it throws an exception in some cases, but does nothing in others.

The rules, and their application, come from the C library stdio
implemention.  Python doesn't second-guess it - if it says the seek
was good, it was good.  And in fact, it was good, inasmuch as the
file pointer did move as specified.  But when a file was opened in
append mode, that means that every write goes to the end of the
file, period, like there's an implicit seek.  So you're right that
the seek() wouldn't have any practical effect.

	Donn Cave, donn at drizzle.com


From guido at python.org  Fri Apr 26 17:50:03 2002
From: guido at python.org (Guido van Rossum)
Date: Fri, 26 Apr 2002 21:50:03 GMT
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org>
Message-ID: <3CC9CBFE.527E8C28@python.org>

Thomas Heller wrote:
> Do anything you want with my money (within resonable
> bounds), but don't make me a member of another
> organization. Please ;-)

You can send your donation to the PSF.  Soon this will be tax-deductible
(for US citizens).  http://www.python.org/psf/  Membership is free
but only on invitation; but anyone can donate, much in the PSA spirit.
We can talk about adding a starship account in the deal; Py subscriptions
are not in the cards though.

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


From johnroth at ameritech.net  Sun Apr  7 10:04:15 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 7 Apr 2002 10:04:15 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu>
Message-ID: <ub0k49taktp243@news.supernews.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-8A20F8.18322106042002 at news.service.uci.edu...
> In article <uavbgcabvl642b at news.supernews.com>,
>  "John Roth" <johnroth at ameritech.net> wrote:
>
> > Here I thought that teaching about bools was quite easy.
> > Bools answer a specific question: is it x? True or False.
> > They do not answer the question: which X is it?
> >
> > In other words, bools select one possible case out of
> > however many there are; they are not a substitute for a
> > universe which has only two cases.
>
> By the way, this is conceptually very similar to ints (at least,
> nonnegative ints): ints answer a specific question: how many things?
> They do not answer the question: which things are they.

Not exactly. Consider the common case of 'take a number'
played at various burocracies. The manager can tell how many
there are by inspecting the next number on the roll, but each person
waiting has his own number, and they can be called out of order.

In other words, ints do allow you to tag each case with a
specific identifier. However, you can do that with any other
type that has enough distinguishable values: I prefer strings for
debugging purposes. If I was really an OO purist, I might prefer
using the object itself, and dispensing with having to identify
it!

To take it to an extreme, I consider using an int as a proxy
for cases to be a code smell! Unless, of course, you need
them for some other reason, like indexing an array.

We might notice that Unix, for example, defines a list of
errors, indexed by small integers. Python returns an error
object, which is much more useful.

John Roth






From paul at boddie.net  Mon Apr 29 04:56:28 2002
From: paul at boddie.net (Paul Boddie)
Date: 29 Apr 2002 01:56:28 -0700
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <mailman.1020024831.2366.python-list@python.org>
Message-ID: <23891c90.0204290056.76b8375e@posting.google.com>

Gerhard H?ring <gerhard at bigfoot.de> wrote in message news:<mailman.1020024831.2366.python-list at python.org>...
>  
> The more I think about it, the less I get the point of object-relational
> mappers like MiddleKit. Especially the ones that require you to rebuild
> the database schema every time you change the object schema. AFAIC
> MiddleKit falls into this category.

That's the way I've perceived it as well, thus explaining my
indifference to MiddleKit, but then I must confess that I don't do any
real work with Webware at the moment.

> I think I'd prefer going either straight SQL with a dictionary mapper
> (pyPgSQL's PgResultSet has such a thing built in [1]) or using an object
> database like ZODB.

I can see substantial advantages in using an object-relational mapper
because it avoids the need to write and maintain near-trivial (and
thus highly tedious) database access code.

> [1] in pyPgSQL, you can say, for example:
> cursor.execute("select id, name from test")
> print cursor.id, cursor.name
> print cursor['id'], cursor['name']
> print cursor[0], cursor[1]
> 
> All forms are equivalent, but only the last one is DB-API2 compliant.

The second one is present in JDBC, at least, but I'm currently unsure
as to how it is guaranteed to behave when there are table qualifiers
or aliases in the query.

Paul


From gerhard at bigfoot.de  Sat Apr 13 01:26:11 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Sat, 13 Apr 2002 07:26:11 +0200
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
In-Reply-To: <005001c1e2aa$3d554280$0101010a@local>
References: <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <005001c1e2aa$3d554280$0101010a@local>
Message-ID: <20020413052610.GA5561@lilith.my-fqdn.de>

* Chris Gonnerman <chris.gonnerman at newcenturycomputers.net> [2002-04-13 00:15 -0500]:
> Another reason often given (by the misinformed) is that GPL code can
> somehow "infect" non-GPL free code (BSD let's say); it goes something
> like this: Bob writes an excellent library and releases it under the
> BSD license.  George has his previously mentioned GPL'd library, which
> is also most cool.  Fred comes along, and writes a program which uses
> both Bob's and George's libraries, and for some reason Fred actually
> incorporates the library source in his project rather than expecting
> you to separately install it.  Fred must release his program under the
> GPL (he has no choice in this matter).

This is wrong. Fed must release his program under a Free Software
license (see the FSF page I mentioned in a previous posting for a
listing of those). So Fred can release his program under a
BSD/Python-like license, too.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From pyth at devel.trillke.net  Sat Apr 27 15:57:25 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 27 Apr 2002 21:57:25 +0200
Subject: Update locals()
Message-ID: <20020427195725.GA16217@devel.trillke>

Hello,

how can you dynamically update the locals() namebindings?
Example what does *not* work:

def func(somedict):
    locals().update(somedict)
    print i

func({'i': None})

This fails on 'print i' with

    NameError: global name 'i' is not defined

dissassembled bytecode with 'import dis ; dis.dis(func)':

          0 SET_LINENO               1

          3 SET_LINENO               2
          6 LOAD_GLOBAL              0 (locals)
          9 CALL_FUNCTION            0
         12 LOAD_ATTR                1 (update)
         15 LOAD_FAST                0 (somedict)
         18 CALL_FUNCTION            1
         21 POP_TOP

         22 SET_LINENO               3
         25 LOAD_GLOBAL              3 (i)
         28 PRINT_ITEM
         29 PRINT_NEWLINE
         30 LOAD_CONST               0 (None)
         33 RETURN_VALUE

Did the compiler decide [25] that 'i' must mean a global
binding? Is there an elegant way to get around this?

thanks,

    holger




From JamesL at Lugoj.com  Sat Apr  6 17:55:36 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 22:55:36 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018072887.13645.python-list@python.org>
Message-ID: <Xns91E898777C949JamesLLugojcom@216.148.53.83>

Tim Peters <tim.one at comcast.net> wrote:
> It's unclear why you would ever need to, if you want to stick with an
> older version, unless you wanted to pick up a pure bugfix release (a
> 3-digit release, like next week's upcoming 2.2.1 and 2.1.3, being,
> respectively, the first pure bugfix release in the 2.2 series, and the
> third in the 2.1 series).

Your advice may seem reasonable but wasn't quite applicable. I developed 
some Usenet moderation programs on 1.5.2 because I knew the co-moderators 
would be running on Linux (1.5.2 and 2.0), Solaris (1.5.2), and Windows 
(2.1). I didn't have Solaris handy, so by distributing this freeware 
program as a simple zipped file of *.py files, I wouldn't have to contend 
with the near impossible task of freezing my application on all the target 
platforms. And with luck it could be used by users running on Macs. By 
choosing Tkinter and 1.5.2 I hoped to minimize version incompatibility. But 
in the end I had to contend with unexpected forward incompatibilities in a 
few places, including in sockets (this one was well known, still some old 
code tripped me up), cStringIO's close semantics changed, and of course 
Tkinter problems. The Tkinter issue was more vexing than the other issues, 
since it took a while to surface and when it did it took a while to track 
down and come up with a good solution. I guess you could say I got "burned" 
by it.

> Major releases change the first digit; there have been only three such
> in all of Python history, counting 0.9.6 and 1.0 and two major
> releases. Python 3.0 is at least years away.  The goal between minor
> releases (bumping the 2nd digit) is 6 months.

Sorry, I meant to use "major" here as in "there are new features or 
semantic changes in this release that if you use in your code, it wont run 
at all or as you expect on previous releases". Doesn't matter what you call 
it, if a 2.1 script can't run at all or as expected on 2.0, that is a major 
change in the _colloquial_ sense. I consider the nature of the changes 
between so-called "minor" releases and their frequency a major problem to 
_me_. I don't think you can realistically "argue" those problems away. I am 
hoping that you will try to accommodate my experiences into your Python 
world view before brushing them off.

> Why do you feel you can't slow down your adoption of new releases? 

Before I set Python aside as being unacceptably non-portable for a 
scripting language, I _had_ tried to stop at 1.5.2. But I couldn't develop 
open source scripts in such a way that they were version independent 
without either a lot of version checking or freezing applications with a 
specific version of the interpreter (which defeats one of the points of 
open source IMHO).

> If you only
> care to upgrade every 2 or 3 years (whatever), you're not required to
> pick up releases any faster than that.  It's not like you paid us for a
> subscription and are going to get cheated out of your money's worth if
> you refuse to accept delivery <wink>.

My last Python project was a freeware moderation tool: my attempt to 
contribute some small thing back to the maintenance of Usenet. It was the 
difficulties encountered in that project that made me think about the 
future of Python and my use of it.

> When the release rate slows to what you'd like to see, I'd take it as a
> strong indicator that Python may be entering its death phase.

The Bourne shell (bsh) and C shell (csh) haven't changed in a _long_ time. 
Do you consider these scripting languages dead in any meaningful sense of 
the word?


From barry at zope.com  Mon Apr  8 17:43:36 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Mon, 8 Apr 2002 17:43:36 -0400
Subject: RELEASED email 2.0 (standalone)
Message-ID: <15538.3848.65410.460971@anthem.wooz.org>

I've just released version 2.0 of the standalone email package.  This
will likely be the last official standalone release, as I intend to
complete the merge of the package into the Python 2.3 cvs tree.

The email package is a library for managing email messages, including
MIME and other RFC 2822-based message documents.  It subsumes most of
the functionality in several older standard modules such as rfc822,
mimetools, multifile, and other non-standard packages such as
mimecntl.

The email package is compatible with both Python 2.1 and Python 2.2.
Online documentation is available at

    http://www.python.org/doc/current/lib/module-email.html

although this does not yet cover the new features in email 2.0.

You can download the lastest version from

    http://sourceforge.net/project/showfiles.php?group_id=25568

Enjoy,
-Barry




From JamesL at Lugoj.com  Sun Apr  7 20:19:01 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 8 Apr 2002 00:19:01 GMT
Subject: Get Button Text
References: <3CB0D763.3090901@acrossthebigpond.com>
Message-ID: <Xns91E9B00BFBC38JamesLLugojcom@216.148.53.96>

Kevin Doyle <doyle at acrossthebigpond.com> wrote:
> I have code that creates x amount of Buttons on the fly.  Each buttons 
> will have a unique Text displayed on it.
> 
> I want to be able to get the text vale of the button when it is pressed.
> 
> I also want to know is it possible to create just one function and bind 
> this one function to all the buttons, and call the function with a 
> reference to the button clicked as an argument.

The secret is the use of lambda to define the command (there is another 
way, but this has worked for me). Try this; hopefully it will become clear.

#-----------------------
from Tkinter import *

buttonWidgets = []
buttonLabels = [ "OK", "Cancel", "Whatever" ]

def YourFunctionCall(button_text):
    print "You pressed ", button_text
    # Just for the heck of it, delete the button pressed.
    buttonLabels.remove(button_text)
    UpdateButtons()

def UpdateButtons():
    global buttonWidgets
    global root
    for button in buttonWidgets:
        button.destroy()
    buttonWidgets = []
    for buttonLabel in buttonLabels:
        button = Button(root, text = buttonLabel, command =
                    lambda txt = buttonLabel: YourFunctionCall(txt))
        button.pack(side = LEFT)
        buttonWidgets.append(button)

root = Frame()
root.pack()
UpdateButtons()
root.mainloop()


From paul at boddie.net  Mon Apr  8 05:56:04 2002
From: paul at boddie.net (Paul Boddie)
Date: 8 Apr 2002 02:56:04 -0700
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org>
Message-ID: <23891c90.0204080156.27f36d22@posting.google.com>

Guido van Rossum <guido at python.org> wrote in message news:<3CAEFC4E.2778AC55 at python.org>...
> 
> About the change rate in general: it's hard to find the proper pace.
> Most new features are requested by users (even the division change!)

...which I eventually came to see as a good idea, although I
personally believe that the way it was announced caused unnecessary
panic. Postponing mandatory changes in semantics even to the release
after next, when the frequency of releases appears to be doubling over
time, might not have been enough to satisfy the most conservative
users.

[...]

> You clearly fall in the most conservative camp.  I just had a long
> exchange with someone who pleaded strongly for moving forward at a
> higher pace, breaking more existing code so that the perfect language
> can be obtained sooner.

Do such people actually do any real work with the language, though?
Wouldn't they be better served by a "big bang, super upgrade" some
time when all aspects of perfection are ready?

> I will have nothing of that, but I hope you understand my predicament: for
> every user like you who complains that things change too fast, there's
> someone with an equally strong desire for faster change.

For your amusement, perhaps you should insist that anyone voting (*)
for particular changes must demonstrate competence with the features
introduced by such changes (and prove that they're a real user of
them) before being allowed to ask for any more.

Let's say that Joe User threw his not inconsiderable charisma behind a
suggested and adopted new feature ("I fully support dictionary
generator comprehension microthreads and encourage you all to do so,
too!"); when subsequently asking for "transaction support in dynamic
class modification", one would ask him to prove that he actually uses
"dictionary generator comprehension microthreads", or whether he just
fancied having them in the language because they seemed like a good
idea at the time. Only after seeing that Joe User was sincere would
his latest support/advocacy/proposal be considered seriously.

(*) Of course, nothing gets decided by voting, but there's always a
lot of noise about new features. Perhaps this would reduce the noise
substantially, albeit at the cost of increasing the administration
ten-fold. ;-)

Anyway, I'm still using 2.0 for most things because Unicode was an
important feature that 2.0 provided. As for the other features
introduced in 2.0 and later, I can't say that I use any of them apart
from the += operator - I hardly ever even use the -= operator! I would
also guess that a reasonable number of people who use the language
still regard these newer features as advanced or magical, especially
those people who learned the language in the pre-1.6 days.

As far as the 2.2 features are concerned, they seem to have a
reputation in certain circles as being badly-defined, immature and
"best steered clear of until later". That's a shame, since some of
them are really very interesting (although the intersection of
"interesting" and "essential" is obviously measured in different ways
by different people), but I can understand (and support) the intention
of focusing on stability for 2.3.

Paul


From Solveig.Wiesmayr at t-online.de  Sat Apr 27 10:46:56 2002
From: Solveig.Wiesmayr at t-online.de (Solveig Wiesmayr)
Date: Sat, 27 Apr 2002 16:46:56 +0200
Subject: name of the current function ?
References: <mailman.1019914669.31757.python-list@python.org>
Message-ID: <aahriu$nkg$01$1@news.t-online.com>

Thanks for your quick and instructive help :-)






From tismer at tismer.com  Fri Apr 26 15:28:09 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 26 Apr 2002 21:28:09 +0200
Subject: PSA revival - kind of?
Message-ID: <3CC9AA49.7050907@tismer.com>

Dear friends,

since the PSA has vanished, Starship has been a
bit stumbling, because PSA membership was the
single requirement for membership.

Starship was quite a good supporter for PSA,
it raised the number of members considerably.

Couple of days ago, I got the idea to use the
new Python magazine subscription as a replacement,
which would give us something similar, and it
would help PyZine very much to grow.

Today, I think we can do even better:

--------------------------------------
I would like to start the SPA:

    Starship Python Activity

as a non-profit organisation.
--------------------------------------

Here the proposed layout:
SPA membership costs USD 25 by the year.
This price includes Starship membership
and a continous subscription to PyZine.

The generated extra money will be used
for sponsoring of projects, programming
contests, a gift to the Starship webmaster,
there are many opportunities.

I think this potential should not be
wasted any longer.

Please send me your encouragements and objections,
I'd like to do it :-)

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From chris at cmb-enterprises.com  Sun Apr  7 21:35:02 2002
From: chris at cmb-enterprises.com (Chris)
Date: Sun, 07 Apr 2002 21:35:02 -0400
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4>
Message-ID: <chris-CFFF94.21350207042002@corp.supernews.com>

In article <Xns91E989929996CRASXnewsDFE1 at 130.133.1.4>,
 Philip Swartzleonard <starx at pacbell.net> wrote:

> def percent_chance( percentage ):
>     	"Assumes percentage as number 0-100"
>     	if random.random() > (percentage / 100.0):
>     	    	return 1
>     	else:
>     	    	return 0

Might more concisely be written as:

def percent_chance( percentage ):
     return random.random() > (percentage / 100.0)


From bokr at oz.net  Wed Apr 10 21:51:08 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Apr 2002 01:51:08 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <1018450381.103820@cswreg.cos.agilent.com> <yu991ydnmtzc.fsf@europa.research.att.com> <a927vf$euh$0@216.39.172.122> <yu99u1qjfdi7.fsf@europa.research.att.com>
Message-ID: <a92q6c$npp$0@216.39.172.122>

On Wed, 10 Apr 2002 21:30:08 GMT, Andrew Koenig <ark at research.att.com> wrote:

>Bengt> What about x() as the spelling of get-x's-value and x(v) as the
>Bengt> spelling of set-x's-value-to-v-and-return-v?
>
>That still assumes that the value of x was set previously.
>
x as container, yes

>If I make that assumption, there are lots of ways of solving the
>problem -- the first one I suggested was simply to set x to [].
>
yes, I was in effect just suggesting eliminating the need for a set
function in the expressions if a named container was going to exist anyway.

If you need to introduce new names into a namespace as a side effect
of an expression evaluation, does it have to be the local namespace,
or could you live with a name space identifier as part of your expression?

The local namespace is going to be a problem in general, but if you are
experimenting, perhaps the restrictions on locals() will make it ok for
you to use

 >>> def set(name, val, ns=locals()):
 ...     ns[name]=val
 ...     return val
 ...
 >>> z
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 NameError: name 'z' is not defined
 >>> set('z',123)
 123
 >>> z
 123

Note that z wasn't preset, and set was executed as an expression.

But if you need to depend on this, maybe Guido will make a pronouncement
as to what functionality will be guaranteed for locals() and globals().

(I've been told that locals() worked for me in similar usage "... "by accident"
in a given implementation, in a certain restricted context, because a undefined
behavior happens to be that way in said specific implementation." So beware ;-)

Of course, def set ... must be executed where it can capture ns=locals()
for use inside the function, where locals() is another space (and read-only).

If you could live with a name space identifier, you obviously have all
sorts of options. Does your problem preclude that way?

Other nasty tricks to create a new bare visible symbol from an expression could be
based on globals() in place of locals(), or even setattr(__builtins__, name, val),
though the latter particularly seems not very advisable ;-)

Regards,
Bengt Richter


From aleax at aleax.it  Mon Apr  1 15:38:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 20:38:16 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8a0gt$roc$1@panix1.panix.com> <l_1q8.49099$S52.1668112@news2.tin.it> <Xns91E3DE242622Fcliechtigmxnet@62.2.16.82>
Message-ID: <Yy3q8.41656$pT1.1153178@news1.tin.it>

Chris Liechti wrote:
        ...
> the above argument about speed seems to be stronger one against the
> change. it would mean that the python core must be adapted so that

I think the speed issue is tiny -- I raised it as a 'by the by', not because
I think the Python language should ever be determined by such issues
as a 10% speed difference in a rather trivial case (how many empty
functions returning a boolean constant do you write in a typical day?-).

> i'm reading in the thread pros and cons, regarding the problems for such a
> small benefit of having True and False i vote for inluding it later in
> Python 3000, if ever.

I do agree that the several small problems (together with the one I
don't consider all that small -- making Python a bit harder to teach)
add up to more than the few small benefits (not just having the True
and False constants -- I include the ability for low-level code to
distinguish between e.g. 0 and False, to ease a bit the task of
interfacing to databases, RPC systems, etc, which do distinguish,
among the small benefits).  And therefore, I'd much rather not see
this change today, even if that meant keeping the Damocles' sword
of the fear of seeing it appear in Py3000:-).


Alex



From fredrik at pythonware.com  Wed Apr  3 13:18:50 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 3 Apr 2002 20:18:50 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff><yu99wuvovkmm.fsf@europa.research.att.com><00af01c1db2a$e7316d10$ced241d5@hagrid><200204031619.g33GJXm02577@europa.research.att.com><016b01c1db31$8da51010$ced241d5@hagrid> <yu99ofh0vg4x.fsf@europa.research.att.com>
Message-ID: <031001c1db3c$0460ad90$ced241d5@hagrid>

Andrew Koenig wrote:

> On the other hand, the library isn't nearly so consistent.  Just for
> fun, I started browsing the library documentation at random.
> 
> The very first page I visited was the description of gc, the
> garbage collector interface.  The first function that returns
> a truth value there is described as
> 
>         isenabled()
>             Returns true if automatic collection is enabled.

>>> import gc
>>> gc.isenabled()
1
>>> type(gc.isenabled())
<type 'int'>

>         endswith(suffix[,start[,end]])
>             Return true if the string ends with the specified suffix,
>             otherwise return false.

>>> "spam".endswith("m")
1
>>> type("spam".endswith("m"))
<type 'int'>

>         isatty()
>             Return true if the file is connected to a tty(-like) device,
>             else false.

>>> sys.stdout.isatty()
1
>>> type(sys.stdout.isatty())
<type 'int'>

>         exists(path)
>             Return true if path refers to an existing path.

>>> os.path.exists("/etc/passwd")
1
>>> type(os.path.exists("/etc/passwd"))
<type 'int'>

looks pretty consistent to me ;-)

(Python docs tend to use "true" to mean anything which isn't "false", and
where operator.truth returns the right thing.  this won't change post-PEP
-- if you start interpreting "return true" as "return True", you'll be asking
for trouble)

</F>





From missive at frontiernet.net  Thu Apr 11 19:46:52 2002
From: missive at frontiernet.net (Lee Harr)
Date: Thu, 11 Apr 2002 23:46:52 -0000
Subject: sockets
References: <uVnt8.8550$jD4.264772@weber.videotron.net>
Message-ID: <slrnabc847.32.missive@whave.frontiernet.net>

On Thu, 11 Apr 2002 18:13:48 -0400, allexpro <allexpro at myrealbox.com> wrote:
>  I'm trying to make a python MUD for the moment, but i don't know what to
> use
> for sockets. i've heard alot of things about socket, serverSoket, medusa,
> simpleHTTPServer,
> asyncore, etc. but i just don't know which to use. Could anyone help me
> please?
> (I'm using python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on
> win32)
> 
> 

You might look at Twisted also.



From guido at python.org  Fri Apr  5 14:09:00 2002
From: guido at python.org (Guido van Rossum)
Date: Fri, 05 Apr 2002 19:09:00 GMT
Subject: PEP 285 and pickle compatibility
References: <7xsn6auy3i.fsf_-_@ruckus.brouhaha.com>
Message-ID: <3CADF727.8A875581@python.org>

Paul Rubin wrote:
> Proposed fix: in new pickles, dump boolean True as "I01" and boolean
> False as "I00".  The current unpickler should load these as the integer
> values 1 and 0 and everything should work fine.  But these strings
> will never be generated by the current pickler, which wouldn't emit
> the leading 0's.
> 
> In the new unpickler, put a special hack into the load_int routines
> in cPickle.c and pickle.py, that notice the leading 0 and push the
> appropriate bool value.

I like this.  I'll see if I can implement it.  (Though it's not clear what
to do in binary mode.)

> I think the ugliness of this kludge is outweighed by escaping making
> an incompatible change to the pickle format.  Sooner or later, though,
> a real incompatible change will be needed, so I'd like to propose
> a further change now, that shouldn't cause compatibility probs:
> 
> The current pickle dump/dumps functions currently take a "bin" arg
> specifying whether to dump in text or binary format.  Unfortunately,
> this arg is treated as a bool and the dumps routine only cares whether
> it's true or false.  So I think a further optional arg should be
> added, giving a "format code" saying what unpickler versions have
> to be able to read the pickle being created:
> 
>   dump(object, format=0)  # same as dump(object), text format pickle
>   dump(object, format=1)  # same as dump(object,bin=1), binary format
>   dump(object, format=2)  # new, incompatible format
> 
> Pickles created with format=2 might for example actually contain a
> bool code instead of the I00/I01 hack, so old unpicklers couldn't read them.
> 
> If a future pickler/unpickler introduces an incompatible format, one
> feature I'd hope to see added is efficient binary pickling of long
> ints.  Currently they're pickled as ascii decimal strings (even in
> binary pickles), which are very slow to convert to and from (all that
> arithmetic).

I like this recommendation, but I'm not motivated enough to implement
it myself.  Maybe you can get some people to work with you on a PEP
and a sample implementation?

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


From borcis at geneva-link.ch  Sun Apr  7 13:48:10 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Sun, 07 Apr 2002 19:48:10 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net>
Message-ID: <3CB08659.F3ED147@geneva-link.ch>

Andrew Dalke wrote:
> 
> Just as an off-topic reminder.  The difference between Bush and
> Gore was less than 2,000.  Given the final Florida results of
>     Bush    2,909,176, Republican
>     Gore    2,907,451, Democrat
>     Nader      96,837, Green
>     Browne     18,856, Libertarian
>     Buchanan   17,356, Reform
>     Phillips    4,280, Official Constitution Party
>     Hagelin     2,287, Natural Law

BTW, I've been wondering if all the problems with voting machines
(if not Bush's election itself) should not be ascribed to the infamous
Y2K bug : after all, if programmers had not been monopolized by the
task of preventing the bug hitting, maybe some work force would have
been applied to design and implement more modern voting machines.

Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")


From tdelaney at avaya.com  Wed Apr  3 21:12:07 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 4 Apr 2002 12:12:07 +1000 
Subject: Catalog of Python Patterns? [LONG!]
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A24F@natasha.auslabs.avaya.com>

> From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
> jimd at vega.starshine.org (Jim Dennis) writes:
> > 
> >  At a minimum it should list each of the 23 GoF patterns, and
> >  at least one minimal Python implementation of that pattern.
> 
> Yucch, I've never been a fan of that book, it seems too cultish to me.

More importantly, many of the patterns are not applicable to Python. _Design
Patterns_ comes from a C/C++ point of view ... there is also _Design
Patterns for Smalltalk_ (or something along those lines) which is apparently
much more applicable.

It is useful to *know* the GoF patterns, but it is important not to be a
slave to them.

Tim Delaney




From mkelly2002 at adelphia.net  Thu Apr  4 11:13:12 2002
From: mkelly2002 at adelphia.net (Michael Kelly)
Date: Thu, 04 Apr 2002 16:13:12 GMT
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com>
Message-ID: <3ouoaukmu32afgatil65cmcb7pvb8e5pcb@4ax.com>

On Thu, 4 Apr 2002 08:53:00 -0600, "Larry Whitley" <ldw at us.ibm.com>
wrote:

>Unfortunately, while the programs appear to run, at completion their output
>files are empty. The files exist but contain no data as a result of the run.
>When I run the same program in a non-threaded fashion from a batch file
>(without the start statement) I get output in the output files.  In this
>case the command looks like this "myprog.exe -optoin 3.0 >out3.txt".

Hi Larry.  I would take a shot at getting cmd.exe in on the act
since command line redirection is a function of the command
processor shell.  Perhaps even putting your command line in
a batch file(with start and its settings and the redirection) and
just calling cmd.exe batchfilename from the Python launcher.
It should open a command prompt and run your batch and
then close when done IIRC.


Mike

--

"Only choice is an oxymoron."

--


From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 14:41:09 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 18:41:09 +0000 (UTC)
Subject: How to avoid mutual imports
Message-ID: <slrnabeai4.lrg.huaiyu@gauss.almadan.ibm.com>

I need advice on how to handle the following situation:

utils1.py:
    a
	b
    c
    if __name__ == '__main__':
	   from interface import data_initialization  # circular
	   test_a()
	   test_b()
	   test_c()

utils2.py:
    x
	y
    z
    if __name__ == '__main__':
	   from interface import data_initialization  # circular
	   test_x()
	   test_y()
	   test_z()

interface.py:
	from utils1 import a, b, c
	from utils2 import x, y, z
	data_initialization

application.py:
	from interface import data_initialization
	do_something()


Note that the interface module defines the data initialization that depends
on both utils1 and utils2.  And the testing of utils1 and utils2 both depend
on data being initialized for both.

Of course I could put everything into a single module, but that would make
it very large.  The definitions of a,b,c and x,y,z really are independent.

Another idea is to write separate tests for each module.  Yes I have done
that, but many things that I want to test rely on data for both modules.

Huaiyu


From duncan at NOSPAMrcp.co.uk  Tue Apr 30 04:20:04 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 08:20:04 +0000 (UTC)
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1> <Bhgz8.58769$vm6.9975333@ruti.visi.com>
Message-ID: <Xns92005E2F4AE3Cduncanrcpcouk@127.0.0.1>

[posted and mailed]

grante at visi.com (Grant Edwards) wrote in
news:Bhgz8.58769$vm6.9975333 at ruti.visi.com: 

> In article <Xns91FF9515A57Eduncanrcpcouk at 127.0.0.1>, Duncan Booth
> wrote: 
>> holger krekel <pyth at devel.trillke.net> wrote in 
>> news:mailman.1020082610.7467.python-list at python.org:
>> 
>>> Btw, what is the best/shortest way to express 'filter' and 'reduce'
>>> with list comprehensions?
>> 
> I find the "reduce" spelling simpler and easier to both read
> and write.
> 
>> All of these replacements are big winners if they remove a use of the
>> 'l' word.
> 
> If getting rid of "lambda" is the goal, then why not state that
> openly and propose elimination of "lambda".

No, getting rid of lambda is not a goal, getting rid of inappropriate use 
of lambda might be. Most lambdas are best replaced by a named function or 
by writing the code in a non-functional manner. Sometimes one is 
appropriate, sometimes the other, sometimes the lambda is best. I do 
believe that lambda is dangerous (because it is often overused), but I 
don't think that it right to get rid of something just because it is 
dangerous.


>> I don't think any of them lose in clarity even if there are no 'l's 
>> around.
> 
> You've got to be kidding -- especially looking at the "reduce"
> and "filter a string" examples.
The filter a string was a bit messy, but in real life I have never used 
filter on a string. I wasn't kidding about the reduce though.

> If you want to move map/filter/reduce into a module, fine.  I
> think it's too much breakage for too little gain, but whatever...
You could try addressing these comments to the people who made them.

I think there is plenty of scope for tidying up the builtins, but not until 
Python 3 and I don't think that map/filter/reduce would be first on my 
hitlist.

> 
> However, claiming that they aren't useful because the above
> "replacements" are just as clear is just silly.
Who claimed that?

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jeff at ccvcorp.com  Tue Apr 30 21:55:00 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 30 Apr 2002 18:55:00 -0700
Subject: OT: Happy Birthday Larry Niven
References: <3cce9f3a$1_6@goliath.newsgroups.com> <3CCEBD17.5040903@jerf.org> <3ccf161d$1_7@goliath.newsgroups.com>
Message-ID: <MPG.1738eace4acc3cd9896ce@news.nwlink.com>

In article <3ccf161d$1_7 at goliath.newsgroups.com>, 
bkc at Murkworks.com says...
> Guess I better go back and re-read the book.
> 
> As I recall.. Moties are essentially genetically designed for their
> particular task. .As in the case of the vehicle drivers who swerved around
> each other and looked like they were going to crash, but didn't.

Motie engineers were capable of adapting bits and pieces of 
whatever in ways that would make a human's brain explode, but 
turned out to function remarkably well.

This sounds exactly like Perl to me -- it *can* work remarkably 
well, if you're genetically designed to grok it, but for a normal 
human, it's more likely to make your brain explode.   :)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From dalcolmo at vh-s.de  Thu Apr 11 08:30:33 2002
From: dalcolmo at vh-s.de (Josef Dalcolmo)
Date: 11 Apr 2002 14:30:33 +0200
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de> <3CB42A50.8020200@cf.ac.uk>
Message-ID: <wkg022l8nq.fsf@vh-s.de>

there is a file called 
pymacs.tar.gz
somewhere. This is supposed to make emacs extensible in Python. 
I haven't had the time to try it yet.

- Josef


From altis at semi-retired.com  Fri Apr 26 14:02:41 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 26 Apr 2002 11:02:41 -0700
Subject: ANN: wxPython Mac on OS X preview
Message-ID: <SBgy8.46$EE5.53683@news.uswest.net>

There are disk images (binary installers) of MachoPython 2.2.1 and the
wxMacPython 2.3.3 preview build available for OS X at:
http://sourceforge.net/project/showfiles.php?group_id=10718&release_id=56880

This software should be considered beta quality. It is not the final
wxPython Mac 2.3.3 release. MachoPython 2.2.1 is required to use wxPython
Mac.

I made an announcement on my blog, but forgot to post to c.l.py.
http://radio.weblogs.com/0102677/2002/04/24.html#a49

The most recent information is available on the wxPython-mac mailing list
http://lists.wxwindows.org/mailman/listinfo/wxpython-mac

If you want to help with the project or just join in the fun please
subscribe.

Bug fixes and new binaries should be appearing frequently as a lot of
testing and work is being done on wxMac (the C++ wxWindows sources) and
wxPython Mac right now. Frank Vercruesse has provided some instructions for
building from cvs.

http://lists.wxwindows.org/pipermail/wxpython-mac/2002-April/000227.html

Finally, as might be expected, I'm pretty excited about this because it
means PythonCard works on OS X. More info on that topic is available on the
pythoncard-users mailing list
http://lists.sourceforge.net/lists/listinfo/pythoncard-users

and the wiki page
http://wiki.wxpython.org/index.cgi/PythonCardOnMac

ka
---
Kevin Altis
altis at semi-retired.com




From gbreed at cix.compulink.co.uk  Wed Apr 17 06:39:01 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Wed, 17 Apr 2002 10:39:01 +0000 (UTC)
Subject: Safe eval?
References: <Xns91F2ED00B6997cliechtigmxnet@62.2.16.82>
Message-ID: <a9jjc5$9g6$1@thorium.cix.co.uk>

Chris Liechti wrote:

> you can disable the builtin functions you want and populate the 
> namespace with save functions and classes (see Bastion for that)

I've never worked out how to do the second bit.  Bastion you say?  The 
documentation for that's all about restricting access to classes.  How do 
I get such a class into the namespace for an reval?  Every time this comes 
up, somebody says "use reval" but nothing about how you supply locals to 
it.

I'm currently using normal eval with a local dictionary, and checking it 
doesn't have file, eval or __ in the supplied string.  That covers 
everything I can think of for cracking my website.  I didn't notice you 
could supply a cleaned up globals dictionary as well, but classes and 
modules being leaky I think vetoing all __ is safest anyway.

My ISP does kill long running scripts (standard feature with Apache) but 
how would I defend against DoS in Python?


                    Graham
           
       <http://www.microtonal.co.uk/>


From chasm at galileo.rift  Thu Apr 25 18:02:55 2002
From: chasm at galileo.rift (Julian Tibble)
Date: Thu, 25 Apr 2002 22:02:55 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz>
Message-ID: <slrnacgvik.irb.chasm@galileo.rift>

In article <3CC61DCA.DAD9DD8E at cosc.canterbury.ac.nz>, Greg Ewing wrote:
>Andrew McNamara wrote:
>> 
>> >Interersting.  If so, then RMS is 'guilty' of an almost humorous
>> >contradiction.
>> 
>> Speculation - maybe he's aware of this and simply sees it as playing
>> what he sees as a fatally flawed system against itself? 8-)
>
>I wonder what he would do if there were no intellectual
>property laws. Would he be campaigning for the introduction
>of a law that enabled him to enforce the price that the
>GNU copyleft exacts?
>
>Greg Ewing, Computer Science Dept, University of Canterbury

erm... If there is no copyright then there is no *need* for copyleft.

BTW, don't say "ahhh, but the GPL also stops you from commercialising
something unless you give away the source," because without copyright
you would be free to copy that commercial software (and reverse engineer
it - once you've copied it, it is yours to do with as you like.)

Julian


From ajs at ix.netcom.com  Thu Apr 18 10:11:06 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Thu, 18 Apr 2002 10:11:06 -0400
Subject: pre-PEP for optional 'pass'
Message-ID: <000701c1e6e2$e282b9e0$0334fea9@carol>

Terry writes: 
>>I am not horrified, however, I do have two objections:

>>1. We will have two styles of Python to read: with and 
>>without pass. I expect this would be confusing, especially 
>>since one >>would not know
>>for sure until one sees a pass.

Phil responds:

>I don't see that as a serious problem. We already have as many 
>styles of python as there are python programmers. For example, I 
>have never used lambda. As another example, some people prefer short
>variable names and others long ones, t versus totalCount. Etc.

This, under my interpretation, was exactly the PEP285 problem.
Can't understand why we would add alternative ways of accomplishing
things that are already fully accomplishable, without a showing of
true enhancement - performance, etc. 

Because there *is* a clear downside, IMO, to creating too many
ways to accomplish the same thing - especially to those new
to the language, and more particularly to those new to programming.

And on the other hand, I can't see any reason to make changes to
the syntax, for the *benefit* of the newbie.  There is little, let's
say, of 1.5.2 that cannot be grasped  by anyone undertaking a serious
enough effort to absorb available docs, reference materials, tutorials 
and sample code. 

Unfortunately I don't think the same can be said of 2.2, for example.

But presumably things will stabilize at some point again, and 
that statement will again be true.


Art





From SBrunning at trisystems.co.uk  Thu Apr 18 08:00:23 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 18 Apr 2002 13:00:23 +0100
Subject: chopping csv-files (newbie)
Message-ID: <31575A892FF6D1118F5800600846864DCBCFE9@intrepid>

> From:	Mikael Lex?n [SMTP:mikael.lexen at stop.spam.volvo.com]
> I get the following when I try to import your module
> 
> PythonWin 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on
> win32.
> Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) -
> see
> 'Help/About PythonWin' for further copyright information.
> >>> import csv
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> ImportError: DLL load failed with error code 193
> 
> What's wrong? I'am on NT4.
 
Did you download the pyd file as well as the package? If not, then that's
the problem.

If you did, did you download the right on for your version of Python?

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




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




From icottee at bluefountain.com  Sat Apr 20 23:30:41 2002
From: icottee at bluefountain.com (Ian J Cottee)
Date: 21 Apr 2002 12:30:41 +0900
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
Message-ID: <m3pu0tpw2m.fsf@localhost.localdomain>

Carl Banks <imbosol at vt.edu> writes:

> Don't even get me started on whether is computer engineering is
> engineering.

Damn, I was looking forward to that! :-)

Ian


From roy at panix.com  Fri Apr  5 21:51:57 2002
From: roy at panix.com (Roy Smith)
Date: Fri, 05 Apr 2002 21:51:57 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018031177.21308.python-list@python.org> <c76ff6fc.0204051701.68c63ecb@posting.google.com>
Message-ID: <roy-A2A363.21515705042002@news1.panix.com>

sjmachin at lexicon.net (John Machin) wrote:
> My crates are in the museum but AFAIR the mag tapes were 7-track or
> 9-track and the paper ones were 5-track [or was there a (6th) parity
> track?]; what's this 8-track story?

8-tracks were audio tapes.  Endless loop, 4 programs (stereo) per tape, 
nice satisfying <ka-Thunk> noise as it switched tracks automatically at the 
end of the loop.  Popular in the 60's and 70's.  Lost in the head-to-head 
competition with cassettes.


From pinard at iro.umontreal.ca  Mon Apr 29 18:38:01 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 18:38:01 -0400
Subject: Python Webpages
In-Reply-To: <slrnacra60.v3g.jon+usenet@snowy.squish.net>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net>
	<aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no>
	<slrnacpc38.obc.jon+usenet@snowy.squish.net>
	<23891c90.0204290051.3edcc62d@posting.google.com>
	<slrnacq7jk.gav.jon+usenet@snowy.squish.net>
	<mailman.1020090053.19284.python-list@python.org>
	<slrnacqrps.sdd.jon+usenet@snowy.squish.net>
	<mailman.1020102500.7281.python-list@python.org>
	<slrnacra60.v3g.jon+usenet@snowy.squish.net>
Message-ID: <oqu1punnau.fsf@titan.progiciels-bpi.ca>

[Jon Ribbens]

> > Yet, seeing the whole thing, both the designer and the programmer
> > can leave the page in a consistent state, even if not fluent in the
> > _other_ field.

> Woah, *big* disagreement here.  Not sure what kind of designers you are
> used to, but in my experience they, and/or the tools they use are
> incapabale of editing the HTML without destroying the code.

Strange.  For more than one year now, we never had the shadow of a
problem with HTML designers.  Everybody is competent in his field and
respectful for the work of others.  There is no "we are good they are
rotten" speak here.  Maybe you are not working with the proper people,
and this had some influence on `jonpy' design? :-)

> I must admit I have never yet seen a system which integrates HTML and
> code in the same file without the file ending up an unreadable mess.

I must admit that HTML is pretty messy to start with, especially the HTML
as output by some Web editor tools.  Included Javascript is not always
nice either.  When I had to massage many HTML files, a few years ago, I
wrote something named `xxml.el' to re-indent and re-fill HTML more nicely
(`xxml.el' extends on Lennart Staflin's PSGML), and surprisingly to me,
from a friend to another, `xxml.el' spreaded and acquired some popularity
in the community over the years.

Once the HTML cleaned up, turning it into active pages, even with embedded
Python, has been rather clean in my eyes.  This is my experience, at least.

> If you take a look at the following URL you can see examples of jonpy
> template use:
>   http://jonpy.sourceforge.net/wt-examples.html#templatecode

OK, thanks, I'll take a look when I'll find a few bits of free time.

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




From marklists at mceahern.com  Sat Apr  6 12:57:07 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 6 Apr 2002 11:57:07 -0600
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204061633.g36GXWdj017461@ratthing-b246.strakt.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com>

[Laura Creighton]
> [...]  I am rejecting the existence of booleans in the language
> because it is too hard to teach how to use them properly.

How do you use them properly?

Suppose I want to validate users.  I might write something like this:

	def isValid(user, password):
		if ...:
			return True
		else:
			return False

I can see why it might be better to write:

	def validateUser(user, password):
		# Yes, of course, I'd really hash it.
		try:
			realUser = getUser(user)
		except UserNotFoundError:
			raise AuthenticationError, ...
		realPassword = getRealPassword(realUser)
		if realPassword != password:
			raise AuthenticationError, ...

I don't see how the latter is somehow avoiding the sin of using booleans,
though.  (Of course, I'm sure there's an even better way to write it.)  I
mean, the comparison between realPassword and password--is that not boolean?
Or are you saying I really care here that Python actually allows me to know
whether realPassword is less than or greater than password?

If so, what's wrong with encapsulating that intention in a function?

Using isValid, that returns one of two possible answers, is as clear and
bright as day to me.  If later on I want a function that does something
different, it would most definitely be named something different.  Why on
earth would I want a function named isFoo to return one of more than two
possible values?  (Your answer is probably that you never want a function
named isFoo.  Hmm, I still don't see why.)

Worrying about the fact that I already have something named isValid when I
go to scratch this newfound itch seems to me like worrying that I'm going to
run out of numbers.

That you want a function whose return values are broader than isinstance
doesn't mean there's something wrong with isinstance.

Cheers,

// mark





From sholden at holdenweb.com  Tue Apr 23 12:58:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 12:58:39 -0400
Subject: Mass Mailing help needed (Paul Rubin)
References: <mailman.1019325819.29016.python-list@python.org><MPG.172e025d803fc2009896bf@news.nwlink.com><mailman.1019553582.18769.python-list@python.org> <slrnacarct.3s6.akuchlin@ute.mems-exchange.org> <mailman.1019575369.19253.python-list@python.org>
Message-ID: <jrgx8.119$_i3.97@atlpnn01.usenetserver.com>

"Graham Ashton" <gashton at cmedltd.com> wrote ...
> On Tue, 2002-04-23 at 15:20, A.M. Kuchling wrote:
>
> > It's certainly possible to send multiple messages over a single SMTP
> > connection,
>
> I must confess to not being that well up on the Python library with
> regard to email messages. Are there any python modules that support
> this?
>

The smtplib library is explicitly provided to allow your applications to
make direct connections to SMTP servers.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From jon+usenet at unequivocal.co.uk  Wed Apr  3 08:18:15 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 03 Apr 2002 13:18:15 -0000
Subject: Python embedded like PHP
References: <mailman.1015988122.18668.python-list@python.org> <mailman.1016619871.27287.python-list@python.org> <7xk7s7mpiy.fsf@ruckus.brouhaha.com> <mailman.1016626151.9950.python-list@python.org> <slrna9haud.e6b.akuchlin@ute.mems-exchange.org> <slrnaa3bi8.lam.jon+usenet@snowy.squish.net> <23891c90.0203290655.25bbcab9@posting.google.com> <slrnaacob1.bu2.jon+usenet@snowy.squish.net> <23891c90.0203310820.60696fb1@posting.google.com> <slrnaahnai.i1t.jon+usenet@snowy.squish.net> <23891c90.0204020047.4d4ae046@posting.google.com>
Message-ID: <slrnaam08l.8al.jon+usenet@snowy.squish.net>

In article <23891c90.0204020047.4d4ae046 at posting.google.com>, Paul Boddie wrote:
> Some common validation cases are numbers, dates and monetary amounts,
> and I've also seen systems which validated measurements. To an extent,
> one can break down complicated fields into collections of simpler
> ones, although some people dislike such interfaces. Nevertheless,
> being able to declaratively associate form fields with the means of
> validation, without the explicit need to code it, could well be
> beneficial, in my opinion.

Well, it could be done as follows:

  class EmailInput(GenericInput):
    def __init__(self, val):
      if not re.match(r"[^@]+@[A-Za-z0-9-]+\.[A-Za-z0-9.-]+$", val):
        raise ValueError, "%s is not a valid email address" % `val`
      self.val = val
    
    def __str__(self):
      return val

The GenericInput class hierarchy could obviously also have any other
class variables or methods which you might find helpful to define
input types.

In jonpy using multiform you would then simply do:

  ...
  class stage0(wt.multiform.Stage):
    ...
    def update_email(self, val):
      try:
        self.container["email"] = EmailInput(val)
      except ValueError:
        self.errors.append("Sorry, the email address you entered is not valid.")

... and there you have it. I don't see how you could get much more
automated than that, really.

Cheers


Jon


From matt at greenroom.com.au  Tue Apr  9 11:07:08 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Wed, 10 Apr 2002 01:07:08 +1000
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com>
Message-ID: <3CB3039C.3040108@greenroom.com.au>

I think the code you are looking for could be:

x = 2

def wrong(num):
   return num + 1

x = wrong(x)
print x

x = wrong(x)
print x



From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 15:16:57 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 21:16:57 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net> <ypjv8.42546$vm6.5594931@ruti.visi.com>
Message-ID: <a9khoc$j7h$1@nets3.rz.RWTH-Aachen.DE>

"Grant Edwards" <grante at visi.com> wrote in message
news:ypjv8.42546$vm6.5594931 at ruti.visi.com...
>..
> Unless we _are_ talking about bioligical stuff, then the term
> you want is probably "sex" rather than "gender"
>..

No, that was just an example.

To explain: I read all form-variables into a dictionary in the beginning of
the script -- I don't know but the form-fields-to-dictionary was a bit of a
hassle -- and at the end I output them, but what was "station" now becomes
"lastStation", and what will become the next "station" depends on the
submit-button the user clicks.
So my actual function where this popped into my head is more boring less
obvious:

    def getHiddenStates(self):
        xhtml = ''
        for key in self.states.keys():
            if key != 'lastStation':
                thisName = key # OK for most cases
                if key == 'station': # happens once only
                    thisName = 'lastStation' # then needs this value
                xhtml += '<input type="hidden" name="' + thisName + '" '
                xhtml += 'value="' + str(self.states[key]) + '" />\n'
        return xhtml

The three commented lines above would not have been needed if I'd use
something like '... name="' + (key != 'station') ? key : 'station' + '" ...'

(There's always a number of occassions where I like to use this construct*,
but this happened to be the first since I only started Python monday.)


* Yes, I could write my own returnIf(state, valueIfTrue, valueIfFalse)
function.
Or use (gender == 'm' and 'male' or 'female') as Mark McEahern pointed out,
but I'm afraid I don't quite get that.




From maxm at mxm.dk  Thu Apr 11 17:25:10 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 11 Apr 2002 23:25:10 +0200
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82>
Message-ID: <3CB5FF36.6060300@mxm.dk>

Chris Liechti wrote:


> how about a formater argument?
> 
>>>>class mystr(str):
>>>
> ... 	def join(self, elements, formater = None):
> ... 		if formater is not None:
> ... 			elements = map(formater, elements)
> ... 		return str.join(self,elements)
> ... 
> 
>>>>mystr(',').join(range(10),str)
>>>
> '0,1,2,3,4,5,6,7,8,9'

That sounds like a good idea!

Max M



From Solveig.Wiesmayr at t-online.de  Sat Apr 27 09:04:48 2002
From: Solveig.Wiesmayr at t-online.de (Solveig Wiesmayr)
Date: Sat, 27 Apr 2002 15:04:48 +0200
Subject: name of the current function ?
Message-ID: <aae7kf$apv$06$1@news.t-online.com>

A beginners question :

I would like to get the name 'my_name' of the current function to pass it
to some other (debugging-) function.

def aaa():
    my_name = 'aaa'
    current_f(my_name)

def current_f(name):
    print name

aaa()

Thanks for hints.





From tim.one at comcast.net  Sun Apr  7 01:41:12 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 07 Apr 2002 01:41:12 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <7xelhsvxsz.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEKIOLAA.tim.one@comcast.net>

[Paul Rubin]
> I don't think he phrased it backwards.  "If you use a new-in-2.1
> feature" presupposes that 2.1 has features that aren't in 2.0, and I
> think James is arguing that it should not.  The issue is just how
> often you're willing to tell people they have to upgrade.

I never tell anyone they have to upgrade:  all choices are tradeoffs, and
you make the ones best for your situation.

> ...
> If I'm running 2.0, and 2.1 comes out, it's pretty similar, so I don't
> upgrade.  Then someone sends me a script that using some new-in-2.1
> feature and I have to upgrade.

This is a tradeoff among your desire to run their script, your reluctance to
upgrade, and your ability to convince that "someone" to code to an older
version of the language.  I don't see "have to" anywhere there.

> ...
> I'd have to do a line by line code review of every one of those damn
> releases before installing it,

Have you done a line-by-line code review of any release?  If not,
multiplication by 0 returns 0 in all versions of Python (so far <wink>).

> ...
> The stability-valuing approach would be DON'T PUT NEW FEATURES into
> releases like 2.1 without good reason to get the particular features
> out fast.

Which competes with other peoples' desires for new features.  I don't buy
that stability is an absolute good; neither new features; overall I've been
very happy with Guido's balancing of these tradeoffs in its effects on my
own code.  I understand that you're not.

> ...
> The proposal of having separate "development" and "stable" release
> tracks like Linux has sounds like a promising direction for keeping
> both camps (language development enthusiasts and would-be production
> users) happier than they are in the current situation.

If we had more active Python developers, that might be feasible.





From ct at gras.de  Fri Apr 26 03:11:01 2002
From: ct at gras.de (Christian Tismer)
Date: Fri, 26 Apr 2002 09:11:01 +0200
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <3CC84749.1AA2A087@cascade-sys.com>
Message-ID: <3CC8FD85.D28F6D4B@gras.de>


"James J. Besemer" schrieb:
> 
> Christian Tismer wrote:
> 
> > There is some nice introduction to Limbo at
> > http://www.vitanuova.com/inferno/papers/descent.html
> > written by Brian W. Kernighan, one of the inventors
> > of the C language.
> 
> Just a nit:
> 
> Dennis Ritchie invented C.  http://cm.bell-labs.com/cm/cs/who/dmr/chist.html
> 
> Brian helped invent the AWK language, a precursor to Python.  AWK stands for
> Aho Weinberger and Kernighan.  Brian wrote a lot of books about C but he was
> not one of the inventors.

Oops!

Sorry about that. I knew it of course, but the
association with the well-known "Kernighan/Ritchie"
Book made me write this mess.

BTW, AWK is more related to Perl than to Python.
I really loved that little language.

ciao - chris




From lac at strakt.com  Sat Apr  6 12:06:30 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 19:06:30 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from Guido van Rossum <guido@python.org> 
   of "Sat, 06 Apr 2002 01:40:01 GMT."
Message-ID: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com>

Apology accepted.  It must be hard to work hard on something and find 
support luke-warm, at best.  Explanation: The reason that the article 
was so long was in part, because the people I asked for help in cutting 
it said that I should not cut it -- I realise now that asking Alex Martelli
for help in cutting an article is like asking the Atlantic Ocean
for help in 'Dryness' -- and in part because to understand my point
of view I need a response from the part of the brain that is not
concerned with symbolic manipulation and language.  I need a _pop_
from the visual cortex of your brain.  This was my research topic
before I abandoned efforts to get a phd and I can talk to you until 
we are all dead of old age about how to measure such reponses in people's 
brains, and what it means about thinking and the like, but you are 
already sick of words from me, so I won't.  But maybe some day over a 
few beers ...

However, you don't know me well enough to know that finding me 
afraid is cause for calling the newspapers ;-).  So I am not
afraid of your bools and my students, I just hate that it is
now going to be harder to teach students not to use them.  I
have a longer response outlining this as a reply to John Roth,
who also did not understand me, in case you want a rehash of the 
argument about why using bools is bad, in general, but I am damn 
close to positive you know all of that.  

I am just pleased that I did not make your 'people not worth
talking to list'.  That was the only bit that worried me.

Laura Creighton




From l0819m0v0smfm001 at sneakemail.com  Sat Apr  6 15:05:45 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Sat, 06 Apr 2002 20:05:45 GMT
Subject: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> <mailman.1018118665.30693.python-list@python.org>
Message-ID: <3CAF1B8F.3050306@sneakemail.com>

Laura Creighton wrote:

> My pleasure, especially since this is such a cool example.  You have
> just shot yourself in the foot, and do not know it.  When you have
> shipped your isValid function, across the world to your 10,000 clients,
> you will discover this.
> 
> Your boss comes in and says, oh, Mark, we need a change to that
> isValid function.  We need to return 3 states, Valid, Invalid, and
> Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
> 
> What do you do now?  

   Write another function, accountPaid perhaps, that clearly expresses 
the intent.  To me something like:

    valid = isValid(user, password)
    if not valid:
       warn("Invalid password")
    elif valid == 2:
       warn("Account not paid up")
    else:
       grantAccess()

is clearly worse (uses magic values, is less clear and harder to extend) 
than something like:

   if not isValid(user, password):
      warn("Invalid Password")
   else:
      if not accountPaid(user):
         warn("Account not paid up")
      else:
         grantAccess()

I think coding a bunch of magic integer return values is far more likely 
to shoot yourself in the foot than a bunch of properly-named boolean 
functions.  To me that kind of enum-think is a legacy of non-object 
oriented code, with vast switch statements to handle program state 
encoded into integers.


   Joshua



From gh_pythonlist at gmx.de  Sun Apr  7 13:32:03 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 19:32:03 +0200
Subject: socket security
In-Reply-To: <slrnaave1r.13h.jon+usenet@snowy.squish.net>
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk> <slrnaave1r.13h.jon+usenet@snowy.squish.net>
Message-ID: <20020407173203.GA1165@lilith.my-fqdn.de>

* Jon Ribbens <jon+usenet at unequivocal.co.uk> [2002-04-07 03:08 -0000]:
> In article <40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk>,
>   Alastair Nicol wrote:
> > 1) Can only hosts which can see the interface your program is bound to 
> > communicate with your program. In the "normal" case if you bind to 
> > 127.0.0.1 for example ONLY your machine can see that interface. Other 
> > machines cannot connect to it. Is this a safe / secure assumtion? (not just 
> > for localhost)
> 
> No. Other machines on your local net can talk to your localhost by
> sending ethernet packets with your machine's MAC address and 127.0.0.1
> as the IP address.

Which you can fight by activating route verification on your OS.

If your OS doesn't have builtin route verification, you should set up
packet filtering rules to deny packets with obviously spoofed addresses,
like non-local packets coming in from elsewhere than the loopback
interface. Or packets that pretend to come from the internal network,
but come in thru the device that's connected to the Evil Internet [tm].

On Linux, enabling CONFIG_IP_ADVANCED_ROUTER and exuting this on bootup
does the job already:

# Route verification (fight IP spoofing)
for f in /proc/sys/net/ipv4/conf/*/rp_filter; do
	echo 1 > $f
done

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 7.2 ?C      Wind: 4.2 m/s




From h_schneider at marketmix.com  Fri Apr 12 05:25:06 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Fri, 12 Apr 2002 11:25:06 +0200
Subject: Include zlib.dll in McMillan Installer
References: <a93ld3$jb4$03$1@news.t-online.com> <Xns91ED608AADBDEgmcmhypernetcom@199.171.54.213>
Message-ID: <a968re$qp9$07$1@news.t-online.com>

Hello Gordon,

removing zlib.dll from excludes[] in bindepend.py adds zlib.dll to the
archive.
Can I do this by a command in the .spec file?

All the best,
Harald


"Gordon McMillan" <gmcm at hypernet.com> schrieb im Newsbeitrag
news:Xns91ED608AADBDEgmcmhypernetcom at 199.171.54.213...
> Harald Schneider wrote:
>
> > is there a possibility to include zlib.dll into the single file
> > distribution of the installer?
> > E.g. PIL uses zlib and complains about missing it on some systems.
>
> In ancient days (1.5.2), zlib.pyd loaded zlib.dll. But
> since then, zlib.pyd has changed to statically link in
> zlib, so zlib.dll should not be needed (in fact, it
> doesn't exist in the distribution).
>
> Installer will always include zlib[.pyd/.so], unless
> your installation doesn't have it, or you turn it off
> by editting config.dat. Windows comes with zlib.pyd,
> so I expect it's there (you can check with ArchiveViewer.py
> to see if it's in the exe).
>
> So more details are needed.
>
> -- Gordon
> http:/www.mcmillan-inc.com/
>
>




From cliechti at gmx.net  Wed Apr  3 17:21:53 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 23:21:53 +0100
Subject: NTLM authorization - compile into EXE
References: <9b309481.0204031300.70ce3c63@posting.google.com>
Message-ID: <Xns91E5EDDAF2FD3cliechtigmxnet@62.2.16.82>

mscheuner at hfx.eastlink.ca (Marc Scheuner) wrote in 
news:9b309481.0204031300.70ce3c63 at posting.google.com:
...
> and I also downloaded Python 1.5.2 itself - everything works okay.
...
> File "C:\Program Files\Python\py2exe\resources\VersionInfo.py", line
> 56, in ?
>   raise ImportError, "Could not import StringTables, no unicode
> available"

i thinnk unicode support was added in python 2.0
i suggest anyway using at least 2.1.2 (i use 2.2)

chris

-- 
Chris <cliechti at gmx.net>



From vmilitaru at sympatico.ca  Thu Apr  4 09:50:52 2002
From: vmilitaru at sympatico.ca (vio)
Date: Thu, 04 Apr 2002 09:50:52 -0500
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org> <3CAC19D8.142A918@sympatico.ca> <7xhemru1pg.fsf@ruckus.brouhaha.com> <3CAC5182.F6CB3F07@sympatico.ca> <7xzo0j4jzz.fsf@ruckus.brouhaha.com>
Message-ID: <3CAC684C.E0C1B789@sympatico.ca>

Paul Rubin wrote:

> What the heck kind of application has 100 MB of Java code anyway?
> It sounds like it may already be the output of some automatic process.

Well, it's an ERP app. 100MB is a very rough figure. Actually not not all java
source code, but a good part of it is. Obviously it grew out of control. Like
some onion which keeps growing up layers upon layers. Though I'm sure half of it
is all brackets and puctuation :-)
Vio


From jacek.generowicz at cern.ch  Mon Apr 29 05:36:48 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 29 Apr 2002 11:36:48 +0200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <m2znzpwc7a.fsf@mycroft.actrix.gen.nz> <mailman.1020032874.18514.python-list@python.org>
Message-ID: <tyf1ycyhmmn.fsf@pcitapi22.cern.ch>

"James J. Besemer" <jb at cascade-sys.com> writes:

> Paul Foley wrote:
> 
> > I think you're missing the point: Lisp is not defined in terms of
> > text; instead, it's the objects that result from calling READ on
> > that text.  There are no parentheses in that result; there are
> > only lists, symbols, strings, integers, etc., etc.  That's why you
> > can define things like the infix syntax reader Jacek Generowicz
> > mentioned.
> >
> > READ can call arbitrary user-defined functions to parse the
> > textual representation into internal objects, so it can read any
> > surface syntax you want.
> 
> I appreciate that Lisp can be customized to process any alternative
> input that you like.  And that the default S-expression notation may
> be abstracted away from the underlying language semantics.
> 
> However, seems to me that out of the box there is this default
> S-expression notation that is rather pervasive to programming in
> Lisp.  If you pick up a book on a particular dialect of Lisp or a
> manual on a particular implementation they all begin with this same
> default syntax.  This default notation, it seems to me and many
> others, requires more parentheses than you normally have occasion to
> use in Algol-like languages such as Pascal, C or Python.
> 
> Although I personally did not find these differences to be a problem
> when I used to program in Lisp I sympathize with those who find them
> strange and unfamiliar.  I feel the default S-expression syntax
> represent a legitimate barrier to entry, however slight, to new
> people approaching and learning the language for the first time.

To re-state publicly the sentiment I expressed in our private exchange
...

As your experience suggests, it is a very slight barrier. The
multitude of people who have never made a sincere effort to overcome
this barrier, but are nevertheless all too eager to suggest publically
that the issue is a serious one, are guilty of mindless
vandalism. They do _nobody_ any favours.

I have no problems with people who do not wish to learn or use lisp,
but I do so much wish that such people would not prejudice others
against what (in the case of Common Lisp) is, above all, an extremely
_practical_, _useful_ and _productive_ language.


From logiplexsoftware at earthlink.net  Mon Apr 29 22:02:05 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 19:02:05 -0700
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <ef08387c.0204291754.1224fd7b@posting.google.com>
References: <ef08387c.0204291754.1224fd7b@posting.google.com>
Message-ID: <20020429190205.1cd7d1bc.logiplexsoftware@earthlink.net>

On 29 Apr 2002 18:54:12 -0700
Fortepianissimo wrote:

> This question is so fundamental that I strongly suspect it's already
> answered - ig that's the case, please forgive this poor newbie, for he
> has done some work searching over Python FAQ and Google.
> 
> Basically I want to have a subclass of list, like
> 
> class MyList (list):
>     ...
> 
> then I want to
> 
> l=MyList()
> ... do stuff to fill l
> s=" ".join(l)

s = " ".join([str(i) for i in l])


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From spammers.do.not.bother at void.bogus  Mon Apr 29 16:13:18 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Mon, 29 Apr 2002 20:13:18 GMT
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <Ksfz8.34667$n4.7298632@newsc.telia.net> <slrnacr583.nil.gerhard@lilith.my-fqdn.de>
Message-ID: <yPhz8.34708$n4.7315779@newsc.telia.net>

Gerhard H?ring wrote:

> Magnus wrote in comp.lang.python:
>> BTW: I friend talked about Twisted the other day. I haven't tried it
>> personally but maybe I should?
> 
> You can join the #twisted and #python channels at irc.openprojects.net
> to get your share of indoctrination ;-)
> 
> I haven't yet fully groked twisted, but IMO it's a great toolkit for
> writing all kinds of network applications. A lot of protocols are
> already implemented for you (http, ftp, telnet, irc, nntp, ...), which
> you can use both for clients and servers.
> 
> But it isn't primarily a solution for writing *web* applications or a
> templating system.
> 
> Gerhard

Thanks! I'll do that later this week.

/Magnus



From paul at fxtech.com  Sat Apr 27 15:09:45 2002
From: paul at fxtech.com (Paul Miller)
Date: Sat, 27 Apr 2002 19:09:45 GMT
Subject: handling different arguments depending on first argument
Message-ID: <3CCAF77B.3BEBF259@fxtech.com>

I've written a C function (callable from Python) that can take a
variable number of arguments, and the arguments after the first are
dependent on what the first was. For example, the signatures for 2
"modes" might look like this:

	my_func(MODE1, var1, var2, var3=None)
	my_func(MODE2, var1, var2=None)

	etc.

Is it possible to handle this kind of different calling convention in
pure Python? I'd like to write the equivalent of an "overloaded" Python
object constructor, which takes various arguments depending on the
first.


From ralph at penguin.com  Tue Apr  2 21:10:27 2002
From: ralph at penguin.com (Ralph)
Date: Wed, 03 Apr 2002 02:10:27 GMT
Subject: How to wait only so long for a server to respond?
Message-ID: <nwtq8.11707$ml2.924462@newsread1.prod.itd.earthlink.net>

Pythoners:

I have some code that requests a response from a server on the Internet, 
but I only want to wait so long for the server to respond.  How can I tell 
python to wait x seconds/minutes, then move on to the next section of code 
if the server has not responded?

Thanks as always.

Raplh


From geoff at gerrietts.net  Mon Apr  1 17:19:22 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 1 Apr 2002 14:19:22 -0800
Subject: runtime location discovery
In-Reply-To: <20020330014832.GF7602@isis.gerrietts.net>
References: <20020330014832.GF7602@isis.gerrietts.net>
Message-ID: <20020401221922.GF6604@isis.gerrietts.net>

Quoting Geoff Gerrietts (geoff at gerrietts.net):
> Now I can go look for symlinks and resolve them, that's not a problem.
> But a hard link would create the same problem all over again. So I'm
> curious -- this can't be an uncommon hurdle to hop. How have other
> people solved the problem?

I've spent some time thinking about this and working at it, and I've
come to the conclusion that no solution is a sure-fire winner. I've
got two ideas toward achieving the goal that I'm pretty happy with.

First: start the program with a shell-script wrapper that is
configured to know where the program is installed. The shell script
can be linked or symlinked to or whatever else, and it won't cause a
problem. I'm not entirely pleased with this solution, since all it
does is move the deterministic configuration step out to the shell
script, but it's acceptable.

Second: put an __init__.py in the images directory and import it --
then use images.__path__[0] as the prefix for my resource loads. That
solves the sys.argv[0] problem without having to special-case for
symlinks, but it still breaks on a hard link.

Of course, I can't even rely on os.path[''] when there's a hard link
involved, which is really the whole reason for the "First" part of the
solution.

I've heard it said that the best way to get a response on the internet
is to post the wrong information, so I'll make a pleasantly audacious
claim in closing: this is the best possible way to address the problem
described. I almost believe that's true, so argue with me. Please. :)

Thanks,
--G.

-- 
Geoff Gerrietts             "If life were measured by accomplishments, 
<geoff at gerrietts net>     most of us would die in infancy." 
http://www.gerrietts.net/       --A.P. Gouthey



From roy at panix.com  Sat Apr  6 20:46:13 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 06 Apr 2002 20:46:13 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com> <uav8og9h0sdqcd@news.supernews.com>
Message-ID: <roy-CAAACC.20461306042002@news1.panix.com>

"John Roth" <johnroth at ameritech.net> wrote:
> Since equality is transitive, this leads to the inescapable
> conclusion that 6 == 7.

Only for very large values of 6.


From logiplexsoftware at earthlink.net  Fri Apr 26 17:58:51 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 14:58:51 -0700
Subject: Removing dupes from a list
In-Reply-To: <20020426140308.658e2ae6.logiplexsoftware@earthlink.net>
References: <20020426203946.GV21205@teich.Garten.DigitalProjects.com>
	<20020426140308.658e2ae6.logiplexsoftware@earthlink.net>
Message-ID: <20020426145851.4e92baae.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 14:03:08 -0700
Cliff Wells wrote:

 
> def de_dup(liste):
>     dups = filter(lambda i: liste.count(i) > 1, liste)
>     return [i for i in liste if i not in dups]
> 
> >>> l = [[1,2,3],[1,2,3],[4,5],[1,2],[1,2,3], 1, 2, 3, 4, 2]
> >>> de_dup(l)
> [[4, 5], [1, 2], 1, 3, 4]
> >>> 
> 
> I expect the list.count() will be a performance killer on large lists.

And I expect this will also be the wrong answer as it removes *all* of the
duplicates :P 

Er, is there any chance that no one saw this? 


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From robin at jessikat.fsnet.co.uk  Tue Apr  2 08:45:52 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 2 Apr 2002 14:45:52 +0100
Subject: __slots__ query
Message-ID: <KvTTsIAQYbq8EwrM@jessikat.fsnet.co.uk>

I notice that in 2.2 it's possible to do strange things to slots via
instances.

Thus 
>>> class A(object):
...     __slots__ = ['a','b']
... 
>>> a=A()
>>> a.c=2
===> error
>>> a.__slots__.append('c')
>>> a.c=2
===> error
>>> a.__slots__
['a', 'b', 'c']
>>> A.__slots__
['a', 'b', 'c']
>>> b=A()
>>> b.c=2
===> error

shouldn't __slots__ be a read only attribute?
-- 
Robin Becker


From dalke at dalkescientific.com  Sat Apr 20 13:06:02 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 11:06:02 -0600
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org>
Message-ID: <a9s72u$gjj$1@slb3.atl.mindspring.net>

James J. Besemer:
>Seems to me the more prominent genetic similarities are the overall
> semantics: late binding, garbage collection, lists and sequences as
> a way of life.

Those are also present in Perl, Tcl, Smalltalk, APL, Mathematica,
and many other languages.  Yes, Lisp was there first, but it's only an
indirect influence on Python.

>  Also, the fact that functions are values that may be free rebound
> to other names.

And Perl and Tcl and ...

>The semantics of and/or are rarely encountered outside of lisp and
> derivatives.

As far as I am aware, that may be true.  I know of no language besides
Python which does that.

>  MAP, APPLY, and especially LAMBDA also are pretty unique to Lisp.
>  This is VERY like Lisp and very UNLIKE "Algol like" languages.

I first learned about map through Perl.  That's also where I learned
about anonymous functions, although Perl doesn't use the word 'lambda.'

They were added to Python just before the 1.0 release, as a contribution
by someone other than Guido.  He's noted for saying way back in 1994
] To be honest, I wish I hadn't introduced lambda, map, filter and
] reduce -- they support a style that is inconsistent with the rest of
] Python. ...

One long-time complaint about them is that they weren't lisp-like
enough, for example, in how lambda handles scope.  Python used to
be quite different than Lisp in how it did static scope.  But as I
recall, others have pointed out that some lisp handle scoping
differently than others, and closer to how Python does it.

BTW, the first two features are best done in modern Python using
list comprehensions and the f(*a, **kw) syntax.

> Perhaps some of the Lisp heritage slipped in via ABC or via proposals
> from others without Guido's knowledge.

Oh, I believe there's a lot of indirect influence on Lisp on Python,
just as there is on most every other language.  Even the first language
I wrote (a BASIC written in BASIC) likely had some influence from Lisp,
and that was before I knew the language existed.

I just argue that there are few direct influences.

>> If there was lisp heritage, why isn't there anything like car/cdr?
>
>Easy.  Some things got left out.  They're not the SAME language.
>
>List indexing and and slices are better and more powerful, an important
> step forward.
  ...
> The more important LANGUAGE abstraction is that dotted pairs can be
> formed into lists and lists may be manipulated at a higher levels, and
> this same abstraction exists in Python.

I think you misunderstood my statement, but then again, I wasn't
very clear.  Lisp's idea of lists is as a composition of pairs.  As
such, certain operations like list indexing are (usually) slow, while
list mutation is fast.  On the other hand, Python uses a contiguous
block of memory, so indexing is fast but mutation (other than append)
is usually slow.

So algorithms which might run fast under Lisp do not under Python,
and vice versa.  The decision for this difference was made after
observations in ABC that most list use was through indexing, and not
through a direct decision to be different from Lisp.

>> ] Python comes from the Algol family of languages.
>
>In syntax yes.  In semantics it seems quite the opposite.

I've heard the statement from Lisp people that all other languages
are an attempt to reproduce Lisp, badly.

>> ] Common Lisp and Scheme.
>> ]
>> ] These languages are close to Python in their dynamic semantics,
>
>Which was my point.

You said Python was a "Lisp derivitave" with the use of len(obj)
based in "precedent dating back to the original language."  You
also said "Python claims a Lisp heritage."

I am not attempt to say that Python and Lisp don't have similar
semantics.  What I am saying is that these specific points you
made are not correct.  Python isn't a Lisp derivative, except in
the sense that all modern languages have Lisp influence.

And that's not the sense you mean since you said "So from a linguistic
heritage standpoint, it's actually rather odd that Python lacks
some equivalent of Scheme's IF."

> I don't agree.  I think syntax and semantics are almost orthogonal,
> with syntax being the less complicated, less important of the two.
> The fact that Python and Scheme have similar semantics make them very
> similar despite dramatic differences in their outer appearance.

Spoken, by the way, like a Lisp programmer.  I found that language
almost impenetrable because of the syntax and my several attempts to
learn the language have failed.  I enjoy Python's semantics, which
implies at least to me that syntax is very important.  Perl and Python
classes also have similar semantics -- Perl got them from Python --
but I couldn't figure out how to do OO programming in Perl because
the syntax was so cumbersome.

But I think you and I agree here, by your saying
>Other than that, I think it's widely acknowledged that Lisp's lack of
> syntax is a disadvantage for most applications.

> How'd [map, apply, lambda] get in there in the first place?  MAP?
> APPLY? EVAL?  They're not in ABC or any other "Algol language" that
> I'm aware of.

As mentioned above, 3rd party contributions which were never part of
the original plan for Python.  Perl has a 'map.'  Tcl has an 'eval'.
Others in this thread point out that functional languages like Clean
and Haskell have map, apply, and lambda.

C doesn't need an apply because it comes for free as part of the syntax.

> And since I was mistaken about the historical asides, my suggestions
> about inline choice and len() are blithely ignored.

I didn't reply on that because I thought other posters covered that
topic reasonably well, and I have very little to contribute in that regard.

The only this I can add is that after watching many non-programmers
(biologists, chemists, and physicists) learn Python, *none* have
complained to me that list.len() doesn't work.

Instead, since your commentary is at least partially based on a
view of history which is incorrect, I decided to comment on that
instead.

I could also have pointed out that except for metaobjects (which in Python
comes more from a Smalltalk heritage) the way Python has done classes
has effectively not changed in the last 10 years, so another statement
you made
> over time Python has increasingly placed much greater emphasis on Objects
is suspect.  But I don't have a firm understanding of the history of
that aspect of Python's history to back up my belief.  Neither, I
believe, do you.

                    Andrew
                    dalke at dalkescientific.com





From bergeston at yahoo.fr  Thu Apr  4 04:58:21 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Thu, 4 Apr 2002 11:58:21 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <mailman.1017869530.3706.python-list@python.org> <b5cnaucrm55rpm8kgt0nibnd4lr43neljs@4ax.com>
Message-ID: <a8h83p$rqigt$1@ID-135695.news.dfncis.de>

"Courageous" <jkraska at san.rr.com> wrote in message
news:b5cnaucrm55rpm8kgt0nibnd4lr43neljs at 4ax.com...
>
> >But... I would aslo suggest that if you do decide to
> >automate part of the transition, you might want to take a
> >look at ANTLR...
> >
> >   http://www.antlr.org/
>
> I recently built a whole Python lexer and parser with ANTLR.
> I played around with about a half dozen compiler-compilers
> while trying to find one that would support my needs. In my
> opinion, ANTLR is the _top_ of the pack of all the freely
> available compiler/translator generators.
>
Did you compare it with JavaCC or SableCC ? I really would like to read your
comments about that. Any link to an article comparing compiler-compilers is
welcome as well.

Many thanks in advance.

> For one thing, the lexer, parser, tree walker, and ast-
> generator are all unified into one meta language. This is
> a very, very clean implementation. It's a full suite all
> in one spot, and fully deserves all the accolades that it
> gets.
>
> C//
>





From logiplexsoftware at earthlink.net  Wed Apr 10 19:23:41 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 16:23:41 -0700
Subject: Easter Eggs
In-Reply-To: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
References: <fm2s8kKkXAiR092yn@gnosis.cx>
	<mailman.1018391839.6535.python-list@python.org>
	<14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <20020410162341.61a42ab6.logiplexsoftware@earthlink.net>

On Thu, 11 Apr 2002 00:17:40 +0100
Robin Becker wrote:
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers. 

Maybe not, but probably the only one who got up on the wrong side of the bed
today.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From brueckd at tbye.com  Mon Apr 22 18:15:31 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 22 Apr 2002 15:15:31 -0700 (PDT)
Subject: Interacting with other Windoze apps
In-Reply-To: <412F4D3C5614D211853600A0C99E563F0E776CD5@com83-173.medicine.uiowa.edu>
Message-ID: <Pine.LNX.4.33.0204221506590.25587-100000@ns0.tbye.com>

On Mon, 22 Apr 2002, Johnson, Brad wrote:

> I'm a new Python programmer and have a question about "interacting" with
> other Windoze applications.  I'd like to send key strokes/characters to an
> application window, but have not been able to find any documentation on
> this.  Is it possible?  For example, suppose a user has Microsoft Word open.
> Is it possible to send an "Alt + F" then an "O" to the Word window to
> activate the *F*ile menu and select *O*pen?
>
> Any pointers (and sample code!) would be appreciated.  Thanks.

Hi Brad, Here's two ways to do this, and both should be possible from
Windows, but I've only tried one of them. In both cases you need the win32
Python extensions by Mark Hammond, which you already have if you are using
the ActiveState distribution of Python.

(1) (the more hacky way) If you're familiar with Win32 programming, you
can probably solve this problem the same way you would from C (use the
win32 API to find the correct window handle and then send it the correct
message. I don't have any sample code to do this, but your Python code
will follow the C equivalent almost call-for-call on the API level (the
Python extensions are wrappers for the equivalent win32 calls so the names
are usually very close, if not identical, to the originals).

(2) If the application you want to interact with exposes a COM interface
(Word falls into this category), then you can use the win32com package of
Python extensions (part of the same win32 package mentioned above) to
automate it (so you could then tell Word to open a new document, write
stuff stuff, save it, etc). The Python section of www.faqts.com has an
example, you can also search on Google groups for
'win32com.client.Dispatch'.

-Dave





From jeremy at alum.mit.edu  Fri Apr 19 13:08:12 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 19 Apr 2002 10:08:12 -0700
Subject: Assignment operator?
References: <a9mksa$8dqs$1@as201.hinet.hr> <WiBv8.15852$eS3.5828@atlpnn01.usenetserver.com> <a9mpq4$940c$1@as201.hinet.hr>
Message-ID: <b0f083db.0204190908.fefa55d@posting.google.com>

"Ante Bagaric" <abagaric_NOSPAM at rest-art.hr> wrote in message news:<a9mpq4$940c$1 at as201.hinet.hr>...
> Hmm, and the compiler doesn't know in advance what object a name references,
> so it has no choice but to flag '+=' as rebinding operation even if it is
> not
> the case. It doesn't sound nice, but I admit there's no obvious alternative
> that
> comes to mind. This whole namespace and local/global issue has an aura of
> 'unfinishness' around it :)

Well, I'm not really sure what 'unfinishness' means.  You didn't mean
un-Finnishnes, right?

The namespace rules in Python are unusual, perhaps unique, because
there are no declarations (except the global statement).  As a
consequence, assignment and other name-binsding operations play an
important role in determining scope.

Jeremy


From esloane at bigpond.net.au  Thu Apr 18 08:03:04 2002
From: esloane at bigpond.net.au (Eric Sloane)
Date: Thu, 18 Apr 2002 22:03:04 +1000 (AUS Eastern Standard Time)
Subject: What webhosting
References: <3CBB5E860000221F@mta07.san.yahoo.com>
Message-ID: <3CBEB5F8.000001.01364@mercury>

Hmmm - Good point Steve - hope you feel better now :~)



-------Original Message-------



From: steve at seasoned-software.com

Date: Thursday, April 18, 2002 2:26:10 PM

To: printers at sendme.cz; python-list at python.org; tutor at python.org;
activepython at listserv.activestate.com; python-help at python.org

Subject: RE: What webhosting



It'll never happen!!!!!!!!!!! If you send unsolicited mail, the user may

get fed up! Bear in mind that yours may be one of HUNDREDS that they get!

The ISP HAS to get rid of your account eventually! WHY????



Because if they DON'T, THEIR ISP may get rid of THEIR account(Which means

they can't support you anyway), because THEIR ISP may get rid of THEIRS!



Sometimes, companies allow this, and they lose their ISP. What happens

THEN? You lose your account ANYWAY!



This is why so many spammers use FREE Email(Of course, the addressees may

complain to their sponsors, which is why so many ads state that, or just

don't work.), and many are now having OTHERS do their dirty work.



Steve





>-- Original Message --

>From: "A" <printers at sendme.cz>

>To: python-list at python.org, tutor at python.org,

> activepython at listserv.activestate.com, python-help at python.org

>Subject: What webhosting

>Reply-to: printers at sendme.cz

>Date: Wed, 17 Apr 2002 21:23:00 +0200

>

>

>Hi,

>Does anybody know about a webhosting provider ( with Python and

>Telnet access) that does not remove a user's account if the user

>sends occasionally some unsolicited emails?

>Thanks for reply

>Ladislav

>

>

>_______________________________________________

>ActivePython mailing list

>ActivePython at listserv.ActiveState.com

>To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs





_______________________________________________

ActivePython mailing list

ActivePython at listserv.ActiveState.com

To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs

.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020418/81b9aea7/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 494 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020418/81b9aea7/attachment.gif>

From maric.michaud at cirec.com  Fri Apr 12 05:05:00 2002
From: maric.michaud at cirec.com (Maric MICHAUD)
Date: Fri, 12 Apr 2002 11:05:00 +0200
Subject: Problem with PIL and compression format
Message-ID: <27E327AB8201B54495D54105FBBBB1485D130E@SRVMSG.GERLAND2.local>

I post my request again, if someone can help.

-----Message d'origine-----
De : Maric MICHAUD 
Envoy? : mercredi 10 avril 2002 15:45
? : Python-List (E-mail)
Objet : Problem with PIL


I tried ro include tif image in a pdf document with reportlab and Python
2.0, PIL 1.1.3.
This work pretty well with uncompressed and 'Pack Bits" compressed TIFF
files, but not with "CCITT group 4" that I need.
Anyone has a suggestion for this ?


-- 
http://mail.python.org/mailman/listinfo/python-list




From francis049 at sympatico.ca  Mon Apr 22 20:04:42 2002
From: francis049 at sympatico.ca (francis049 at sympatico.ca)
Date: Mon, 22 Apr 2002 20:04:42 -0400
Subject: What is an import and where is it on the hard disk
Message-ID: <3CC4A51A.A7D770ED@sympatico.ca>

I receive an error message like :

    Traceback (innermost last) :
    File : /usr/sbin/up2date, line 9, in ?
      import rpm
    Import error : No module named rpm


Anyone can tell me why is this error happening and give me some details
... please ?



From mal at lemburg.com  Thu Apr 25 16:28:57 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 25 Apr 2002 22:28:57 +0200
Subject: configuring mxCGIPython
References: <NCBBLFCOHHDIKCAFGCFBEEJEKMAA.marklists@mceahern.com>
Message-ID: <3CC86709.30433BD3@lemburg.com>

Mark McEahern wrote:
> 
> [Paul Boddie]
> > I found that it isn't always possible to deploy Python programs in a
> > cgi-bin directory and to use the above notation to persuade the
> > operating system to locate the Python executable. In the end, I used a
> > wrapper shell script for each of my Python programs.
> 
> Here's the wrapper shell script I'm using.  I'd be very interested in any
> feedback, suggestions you might have:
> 
>   #!/bin/bash
> 
>   echo content-type: text/plain
>   echo
> 
>   echo Trying to capture the stderr of cgipython
> 
>   if [ -f cgipython ]
>   then
>     echo cgipython exists
>     # Trying to send stderr to stdout.
>     ./cgipython test.py 2>&1
>   else
>     echo cgipython does not exist
>   fi
> 
> test.py, fwiw, is simply:
> 
>   import sys
>   print sys.version
> 
> When I run this, I get:
> 
>   Trying to capture the stderr of cgipython
>   cgipython exists
>   ./cgipython: error in loading shared libraries: libxmlparse.so.1: cannot
> open shared object file: No such file or directory
> 
> I'm not sure yet what that means yet, but I think it means I may be able to
> set LD_LIBRARY_PATH, if: 1) that file resides on the server; and 2) I know
> where it is.

Right.
 
> Is there a way I can tell cgipython not to bother loading that particular
> shared object?

No. Could be that libxmlparse is referenced by pyexpat which is part
of cgipython. On most systems the expat libs should be part
of the system, so setting LD_LIBRARY_PATH to e.g. /usr/lib
should help. If not, you can try getting that file from some
other source and install it in the same dir as cgipython.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From kevin at cazabon.com  Sat Apr  6 13:49:24 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 6 Apr 2002 10:49:24 -0800
Subject: "Memory could not be Read/Written" crashes with C extension DLL... HELP PLEASE!
References: <5a4226f0.0204052223.75aa0318@posting.google.com>
Message-ID: <5a4226f0.0204061049.c2b1026@posting.google.com>

Through some help, I've found that this section is causing the
problem... If I simply use "return Py_BuildValue("lllO", pp1,
inBufferSize, pp3, inBuffer);", it works fine with no crashes.

What is the difference though, shouldn't the code below work too?  Is
there maybe something wrong in the PyTuple code?  (As mentioned
before, it only crashed when called multiple times (40+ usually), but
worked fine otherwise.

Kevin.

>   // build pp1, inBufferSize, pp3 into PyObjects for returning
>   pLong1 = PyLong_FromLong(pp1);
>   pLong2 = PyLong_FromLong(inBufferSize);
>   pLong3 = PyLong_FromLong(pp3);
> 
>   // build pLong1, pLong2, pLong3, pString into a tuple to return
> 
>   pReturn = PyTuple_New(4);
>   PyTuple_SetItem(pReturn, 0, pLong1); 
>   PyTuple_SetItem(pReturn, 1, pLong2);
>   PyTuple_SetItem(pReturn, 2, pLong3);
>   PyTuple_SetItem(pReturn, 3, inBuffer);
> 
>   return pReturn;
> }


From ods at fep.ru  Thu Apr 11 12:16:23 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 11 Apr 2002 20:16:23 +0400 (MSD)
Subject: binding a reference to a variable
In-Reply-To: <3CB5B230.CDECEBFB@llnl.gov>
Message-ID: <Pine.LNX.4.44L.0204112012020.13708-100000@ods.pp.ru>

On Thu, 11 Apr 2002, Patrick Miller wrote:

PM> The following does not work as intended....
PM>
PM> def f():
PM>    a = 7
PM>    L = locals()
PM>    print L.keys() # Prints ['a','L'] or some such
PM>    L['a'] = 99  # Update the value in L
PM>    print a      # But wait!  This prints out '7'

But this work and AFAIK is not implementation dependent:

>>> def set(name, val):
...     exec '%s=__val__' % name in {'__val__': val}, \
...                           sys._getframe().f_back.f_locals
...
>>> a = 7
>>> set('a', 99)
>>> a
99





From max at alcyone.com  Mon Apr 15 20:56:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 15 Apr 2002 17:56:42 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <3CBB76CA.A423AAD8@alcyone.com>

phil hunt wrote:

> OK, I've decided to write a PEP. Preliminary to this, I'd like
> people's comments.
> 
> I wish to make the pas startement optional. That is, in any Python
> program where there is a pass it can be left out and the meaning of
> the program is unchanged.

The purpose of pass to explicitly disclaim what you're intending to do;
it reduces programmer error.  In cases in code where you see something
like:

	if condition:
	doSomething

It is far more likely that this is an indentation mistake rather than a
deliberate null statement.  The pass keyword removes all doubt:

	if condition:
	    pass
	doSomething

The pass keyword helps solidify this kind of behavior.  Convenience and
beauty can't really hold a candle to utility (even if it is marginal
utility) and precendent.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From phr-n2002a at nightsong.com  Sat Apr 13 02:55:14 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 23:55:14 -0700
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <005001c1e2aa$3d554280$0101010a@local> <mailman.1018675649.29978.python-list@python.org>
Message-ID: <7xn0w8cckt.fsf@ruckus.brouhaha.com>

Gerhard H?ring <gerhard at bigfoot.de> writes:
> > Another reason often given (by the misinformed) is that GPL code can
> > somehow "infect" non-GPL free code (BSD let's say); it goes something
> > like this: Bob writes an excellent library and releases it under the
> > BSD license.  George has his previously mentioned GPL'd library, which
> > is also most cool.  Fred comes along, and writes a program which uses
> > both Bob's and George's libraries, and for some reason Fred actually
> > incorporates the library source in his project rather than expecting
> > you to separately install it.  Fred must release his program under the
> > GPL (he has no choice in this matter).
> 
> This is wrong. Fed must release his program under a Free Software
> license (see the FSF page I mentioned in a previous posting for a
> listing of those). So Fred can release his program under a
> BSD/Python-like license, too.

That's incorrect; from section 2 of the GPL:

        b) You must cause any work that you distribute or publish, that in
        whole or in part contains or is derived from the Program or any
        part thereof, to be licensed as a whole at no charge to all third
        parties under the terms of this License...

    These requirements apply to the modified work as a whole.  If
    identifiable sections of that work are not derived from the Program,
    and can be reasonably considered independent and separate works in
    themselves, then this License, and its terms, do not apply to those
    sections when you distribute them as separate works.  But when you
    distribute the same sections as part of a whole which is a work based
    on the Program, the distribution of the whole must be on the terms of
    this License, whose permissions for other licensees extend to the
    entire whole, and thus to each and every part regardless of who wrote it.

The "this License" above refers to the GPL.  So if Fred's program
incorporates George's GPL'd library, Fred's program must itself be
GPL'd.  If Fred doesn't want to follow George's conditions, he is of
course free to decide to not use George's library.  

There's a separate "Lesser GPL" that you might be thinking of, used
for some libraries, which have more relaxed conditions the full GPL.
The decision of which license to use on a given library is up to the
library author--some choose the GPL and some choose the LGPL.

Hope-this-helps-'ly yours,
phr


From dpeter at designtheory.org  Tue Apr 30 06:44:12 2002
From: dpeter at designtheory.org (Peter Dobcsanyi)
Date: 30 Apr 2002 10:44:12 GMT
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no>
Message-ID: <aalshs$3p5$1@beta.qmul.ac.uk>

While we are at it, I would be interested in opinions on
Albatross:

    http://object-craft.com.au/projects/albatross/

Thanks,
    Peter


From philh at comuno.freeserve.co.uk  Sun Apr 14 14:03:30 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 19:03:30 +0100
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <ubhlb81tfqo23e@news.supernews.com>
Message-ID: <slrnabjh3i.741.philh@comuno.freeserve.co.uk>

On Sat, 13 Apr 2002 21:03:25 -0400, John Roth <johnroth at ameritech.net> wrote:
>
>"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote in message
>news:3cb88af1 at news.mhogaming.com...
>> I think an 'xor' operator is needed.  The 'xor' operator means
>> 'logical-exclusive-or', and returns a 1 or 0.
>>
>> Here is some code that does what I want (in all the cases I can think
>of) for
>> 'xor':
>>
>> # logical-exclusive-or operator:
>> def xor(a,b):
>>     return (not a) ^ (not b)
>>
>> This will work, but I think it is less clear (in an expression) than:
>'a xor b'.
>
>Write a PEP. I've got no objection, but you will get
>the usual slew of objections from people that think
>that it is too minor to warrent breaking code by
>introducing a new operator.

How will it break code to introduce a new operator? Any existing 
program that uses xor will be a syntax error, and broken already.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From mwh at python.net  Mon Apr  8 09:34:25 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 8 Apr 2002 13:34:25 GMT
Subject: yet another "Here's to 1.5.2" post
References: <1018131203.756898@cswreg.cos.agilent.com>
Message-ID: <ud6xab990.fsf@python.net>

weeks at vitus.scs.agilent.com (Greg Weeks) writes:

> Actually, Python 1.5.2 was already a bit too large.  The functional
> programming constructs (lambda, map, reduce, filter) are fluff.

Yes, but that doesn't mean all later constructs are fluff too.

scheme is over there ---------------------------------->

> List comprehension is cutesy and practically Perlish, an idiosyncrasy
> reminiscent (in flavor and aroma) to Perl's ".." operator.

Strange, seeing as they were inspired my Haskell...

> "print" has always been something of a blemish.

But an important one:

    I mean, just take a look at Joe Strout's brilliant little "python
    for beginners" page. Replace all print-statements with
    sys.stdout.write( string.join(map(str, args)) + "\n") and you'll
    surely won't get any new beginners. And That Would Be A Very Bad
    Thing.
                    Fredrik Lundh, 27 Aug 1996 

(from http://www.amk.ca/quotations/python-quotes/page-2.html)

[...]
> The so-called "nested scopes" added in 2.1 are a mistake. 

Says you.

[...]
> With somewhat less confidence -- and really, what's the point of being sure
> about this (like it matters!) -- I think the redefinition "/" and the
> addition of a boolean type were bad choices.

Oh well.

> Now, if only the energy that was put into adding the above blemishes had
> been placed into making Python more robustly installable.  On my system
> (HP-UX 10.2) with my reasonable efforts at building a system:
> 
>     in 1.5.2, there is no FCNTL module
> 
>     in 2.2, "configure" fails due to erroneous C pre-processor code
> 
>     in 2.2, "make test" fails for reasons I haven't had the time to pin down
> 
> I'll keep using Python, so long as Python fills the Perl niche better than
> Perl.  But creeping TIMTOWTDI and installation problems make me worry,
> worry, worry.

Hey, if you want HP-UX 10.20 support, you are probably going to have
to do it.  I wouldn't mind spending a while on it, but I can't.  FWIW,
it seems Python 2.2.1(c1) on 11.00 actually works.

Anybody want to post a HP-UX box to me?

Cheers,
M.

-- 
  I also feel it essential to note, [...], that Description Logics,
  non-Monotonic Logics, Default Logics and Circumscription Logics 
  can all collectively go suck a cow. Thank you.
              -- http://advogato.org/person/Johnath/diary.html?start=4


From neuhauser at mail.cz  Mon Apr 29 10:30:28 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Mon, 29 Apr 2002 16:30:28 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1>
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1>
Message-ID: <20020429143028.GE325@roman.mobil.cz>

> From: Duncan Booth <duncan at NOSPAMrcp.co.uk>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Mon, 29 Apr 2002 13:33:31 +0000 (UTC)
> 
> Roman Neuhauser <neuhauser at mail.cz> wrote in 
> news:mailman.1020083690.16055.python-list at python.org:
> 
> >     __builtin__.chr(int) -> int.chr()
> >     __builtin__.abs(int) -> int.abs()
> >     __builtin__.ord(str) -> str.ord()
> >     etc
> > 
> 
> The problem with int.chr() is that a lot of people are going to get 
> confused when they find that "42.chr()" doesn't work (although "42 .chr()" 
> does).

    why? letters aren't valid decimal digits, are they?
 
-- 
FreeBSD 4.4-STABLE
4:28PM up 6:53, 10 users, load averages: 0.02, 0.09, 0.08




From Gabe.Newcomb at noetix.com  Tue Apr 23 16:18:09 2002
From: Gabe.Newcomb at noetix.com (Gabe Newcomb)
Date: Tue, 23 Apr 2002 13:18:09 -0700
Subject: Fundamental problem with urllib...
Message-ID: <0C7CA8D8DF75494EB09AB6016990107F011B1A97@NOXMAIL.noetixad.com>

I think you mis-spelled wgah'nagl...

-----Original Message-----
From: A.M. Kuchling [mailto:akuchlin at ute.mems-exchange.org]
Sent: Tuesday, April 23, 2002 10:15 AM
To: python-list at python.org
Subject: Re: Fundamental problem with urllib...


In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>, 
	Steve Holden wrote:
> Since urllib knows nothing of cookies, you will need to integrate some
sort
> of a cookie jar into the library, with a new API  for the clients to
> retrieve and store the cookies.

This is worthwhile, but I don't think it belongs in urllib.  It
belongs in a module or package of its own that provides general
Web-browser features such as cookies, remembering authentication
usernames and passwords, and a cache.  This package could then be used
for implementing HTML-scraping scripts, spiders, or a Web browser.

--amk
(www.amk.ca)
Ph-nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn. 
    -- H.P. Lovecraft, "The Call of Cthulhu"

-- 
http://mail.python.org/mailman/listinfo/python-list




From aleax at aleax.it  Thu Apr 18 03:55:51 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 07:55:51 GMT
Subject: unhandled exceptions without a stack-trace
References: <mailman.1019111684.8904.python-list@python.org> <just-209240.09395718042002@news1.xs4all.nl>
Message-ID: <b_uv8.33276$vF6.1023831@news2.tin.it>

Just van Rossum wrote:

> In article <mailman.1019111684.8904.python-list at python.org>,
>  Paul Rodger <paul at paulrodger.com> wrote:
> 
>> Is there any way I can:
>> 
>> 1) raise an exception ...
>> 2) not catch that exception (I want python to abort) ...
>> 
>> without displaying the "Traceback (most recent call last):" stack trace?
> 
> Catch the exception anyway, then do sys.exit() or raise SystemExit.

sys.excepthook should be another viable alternative too.


Alex



From skip at pobox.com  Wed Apr  3 23:10:48 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 3 Apr 2002 22:10:48 -0600
Subject: cgi.py performance
In-Reply-To: <a8ght4$o8o$1@web1.cup.hp.com>
References: <a8ght4$o8o$1@web1.cup.hp.com>
Message-ID: <15531.53832.385265.70888@12-248-41-177.client.attbi.com>

    jim> ....  I have rewritten read_lines_to_outerboundary() to process the
    jim> incoming data in large (binary) chunks and have gotten the
    jim> performance of a 40MB file from 1:41 down to just 15 seconds(!).
    ...
    jim> I guess my first question should be: has something already been
    jim> done to address the performance of multipart/form-data?  And if
    jim> not, could I get what I have written reviewed?

Search for "cgi.py" at SF in the bugs and patches section.  If you see
something already submitted, comment on it and attach your patch if
appropriate.  If you don't see something already, submit a patch.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 15:44:16 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 12:44:16 -0700
Subject: Does Python need a '>>>' operator?
Message-ID: <3cb88ae7@news.mhogaming.com>

Since Python does not have unsigned numbers, I think Python should have a
(Java-like) '>>>' operator.  The '>>>' means 'logical-shift-right', where a zero
is copied into the high bit.  This would be kind of nice when doing some logical
bit tweaking...

Here is some code to implement the '>>>' operator. (Unfortunately, this code
does not work for long integers):

# logical-shift-right operator:
def lsr(i,n):
    if n: i = i >> n ^ 0x80000000 >> n - 1
    return i

Although the '<<' and '>>' operators do not now "do the expected thing" with
long integers, they COULD if we introduced the '>>>' operator.  In order to
support the "all integers are the same" model, I think that they should.  Thus,
for long integers '<<' and '>>' could give different answers than '*2' and '/2'.
(I don't know how to implement this properly for long integers without breaking
some old code though.)

If this was discussed before, then it needs to be discussed again, as we are now
trying to provide seamless integration of the 'int' and 'long int' types...





From JamesL at Lugoj.com  Fri Apr  5 23:08:06 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 04:08:06 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org>
Message-ID: <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84>

Guido van Rossum <guido at python.org> wrote:
> You could have typed the code in to Python 2.2 and gotten
> your answer. 

Therein lies the problem: I typed it into Python 2.1 and naturally it 
didn't work. I loaded 2.1 onto my laptop mid-year last year to be "up to 
date" and I see that less than a year later my version is out of date. (By 
comparison, my Linux machine still has 1.5.2 so I can do version checks.)

Seriously Guido, how often will I need to check back to www.python.org to 
keep up with the changes you are making? Do you have a schedule or some 
constraints in place on how fast you are releasing major versions?

If you could please please please slow your major release rate down, and 
make that an "official" policy I would reconsider my departure from 
Pythondom. It would be nice to see major releases at a _minimum_ of 18 to 
24 months apart. And frankly I would think the law of diminishing returns 
has to kick in at some point; even if you disregard my request I would 
think it would become progressively more difficult for you to justify new 
releases all that often. When I see that happening (maybe in a few years?) 
then I would consider rejoining the ranks of Python programmers.


From martin at v.loewis.de  Fri Apr  5 16:21:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Apr 2002 23:21:49 +0200
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de> <f445a2fe.0204041708.6fd6fd0d@posting.google.com> <m37knm395e.fsf@mira.informatik.hu-berlin.de> <f445a2fe.0204051105.6f67b848@posting.google.com>
Message-ID: <m3r8lt96z6.fsf@mira.informatik.hu-berlin.de>

seb_haase at yahoo.com (Sebastian Haase) writes:

> I'm just saying: Why do I need to touch _DEBUG at all? Normally if I
> compile my code in debug-mode _DEBUG is defined -- thats just how
> people are doing that. (As I understand...)

Yes. And normally, on Windows, you also need to use the debug
libraries of everything - whether you want to debug them or not.

> The story just goes the other way around: Why wants the python.h force
> us to include a lib (python22_d.lib)  that it doesn't ship. (And I'm
> _not_ interested in debugging the python lib anyway?)

This is not because we want you to debug Python, but because
otherwise, things may go wrong.

> What was this comment though about the "linking the MSVC debug C libs"
>  -- might this issue cause me trouble ?? I'm not experiencing any
> problems so far...

No, they are unlikely, and if they happen, they are obscure and hard
to find.

Just visit the modules list in the VC++ debugger when you have the
debug version of your extension loaded, but you are linking to the
nodebug version of Python.

You will notice that you have two libraries loaded: msvcrt.dll (or
msvcrt40.dll), which is the no-debug C library, that was loaded
because python22.dll requires that DLL. You will also notice that
msvcrt40d.dll is loaded, which is the debug version of the C library.

Now, both copies of the C library have the same global data structures
- but you get them twice now. This involves in particular the heap: if
you invoke malloc with one of the functions, and free with the other,
bad things may happen - at a minimum, the memory will leak. Microsoft
advises against such scenarios, but I can't find the relevant MSKB
article right now.

Regards,
Martin


From op73418 at mail.telepac.pt  Mon Apr 22 15:51:00 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 20:51:00 +0100
Subject: Science And Math
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <mailman.1019498507.4893.python-list@python.org>
Message-ID: <hup8cu8lrglr4e5tn1sn3c171bua1fsan2@4ax.com>

On Mon, 22 Apr 2002 13:49:00 -0400, Lulu of the Lotus-Eaters
<mertz at gnosis.cx> wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote previously:
>|mathematical field is roughly divided in the Platonist field, those who
>|believe that mathematical objects exist somewhere out there in space
>|...and the formalist field of those who believe mathematics is
>|essentialy a game we play on paper with symbols with no "real" meaning
>|attached to it besides the convenience in describing natural phenomena.
>
>There is at least one philosophical position missing from Rodrigues'
>enumeration.  The intuitionist approach to mathematics sees mathematical
>reasoning as exposing the "rules of thought."  These rules are not
>independently existing structures in the world (Platonism), but neither
>is mathematics merely symbolic (Formalism).  Instead, there is a
>substantial but procedural fact about the way mathematicians think about
>their subject matter.

Yes, you are right. But I wouldn't be surprised if they really are a
minority.

>
>It is obvious from my other little note in the thread that I feel
>sympathy with the Intuitionist idea about mathematics.  Of course, this
>sympathy is somewhat forclosed by my attachment to the Axiom of Choice.
>I'm not sure I'm able to quite go as far as Brouwer (even though I only
>guiltily fail to).
>
>Yours, Lulu...

If it were just the axiom of choice... The biggest problem for the
constructivist/intuitionist (and my apologies if my lumping them
together is just plainly wrong) is that no non-constructive proofs (for
example, existence proofs relying in double negation, e.g. object x does
not exist ==> a contradiction, therefore object x must exist) is
accepted.

These came about because these non-constructive proofs can yield some
really hair-raising examples (an orange can be split and reassembled in
such a way to yield two oranges, etc.)

But rest assured, there is *intuitinistic/constructivist* mathematics
being done (at least that is what my Topos theory friends tell me, if I
understand them right)

Best regards,
Gon?alo Rodrigues



From nhodgson at bigpond.net.au  Sat Apr 20 19:20:27 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 20 Apr 2002 23:20:27 GMT
Subject: Python for .NET vs. Jython ... Python.Net thoughts?
References: <f57664b9.0204152215.57312efa@posting.google.com> <mailman.1018976552.29487.python-list@python.org> <e81be8b2.0204201335.77865efd@posting.google.com>
Message-ID: <%Imw8.30126$o66.93750@news-server.bigpond.net.au>

Dean Goodmanson:

> What's the most practical way to get Python into the a .NET
> development environment?
>
> - CLR Python compiler : True "Python.net"
>
> - C# implementation , simpilar to Jython : "Shython" ??
>
> - Others???

   You can interoperate using COM.

   One option ActiveState implemented for Perl called PerlNET runs the Perl
interpreter as unmanaged code along with .NET managed code in one process
and provides interoperability between the two worlds. A similar facility
could be produced for Python.

   Neil





From hughett at mercur.uphs.upenn.edu  Tue Apr 30 10:04:08 2002
From: hughett at mercur.uphs.upenn.edu (Paul Hughett)
Date: 30 Apr 2002 14:04:08 GMT
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu> <3kjz8.32779$8D3.928578@news1.tin.it>
Message-ID: <aam88o$oec$1@netnews.upenn.edu>

Alex Martelli <aleax at aleax.it> wrote:
: Paul Hughett wrote:
:         ...
:> Hence my question: Is there an established meaning within Python for
:> slice objects with a negative increment?  E.g. slice(1, 10, -2) ?

: Sure!  [Use the same convention as Numeric]

Okay, I've looked up the convention for Numeric, but the description
isn't entirely clear in the case of negative increments.  It states that
the default values for start and stop when the increment is negative are
"the beginning" and "the end" of the axis.  Suppose that the axis has
n elements.  Does ::-k default to 

   n:0:-k        which refers to element n, which doesn't exist
   (n-1):0:-k    which skips element 0
   n:-1:-k       which refers to element n, and apparently to -1 => n-1, or
   (n-1):-1:-k   which is not generally the reversal of 0:n:k for k > 1


By the way, I don't see anything in "Numerical Python" about the
meaning of a negative start or stop.  Do they follow the usual Python
convention of meaning the nth element from the end?


Paul Hughett


From fgeiger at datec.at  Tue Apr 23 12:59:12 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Tue, 23 Apr 2002 18:59:12 +0200
Subject: Class Methods with Python 2.1
References: <3CC57C5F.7040804@thomas-guettler.de>
Message-ID: <aa43tu$13k$1@newsreaderg1.core.theplanet.net>

I did this with PythonWin IDE:


from PyFL_0_1_1.static import static


class Car:

   __Availables = ["porsche", "volvo", "trabbi"]

   def __init__(self, name):
      if name not in Car.__Availables:
         raise Exception("Name not in available cars")
      return

   def Add_available_car(name):
      Car.__Availables.append(name)
      return
   Add_available_car = static(Add_available_car)

   def Available_cars():
      return Car.__Availables
   Available_cars = static(Available_cars)

   def available_cars(self):
      return Car.__Availables


def main():
   try:
      print 'Try to create my car: '
      myCar = Car('Benz')
   except Exception, e:
      print e
      print 'Add my car.'
      Car.Add_available_car('Benz')
      myCar = Car('Benz')

   print 'Cars available now: '
   print Car.Available_cars()
   print
   print myCar.available_cars()


if __name__ == '__main__':
   main()


To be able to run this, you need 'static':

class static:
   """
   SYNOPSIS:
      'Static-methods' (aka 'class-methods') in Python. Copied from Alex
Martelli's
      contribution in the ActiveState Python Cookbook.

   SAMPLE:
      class MyClass:
         def __init__(self):
            pass

         def myStaticMethod(self):
            pass

         myStaticMethod = static(myStaticMethod)

      def test():
         MyClass.myStaticMethod()

      test()
   """
   def __init__(self, anyCallable):
      self.__theCallable = anyCallable

   def __call__(self, *args, **kwds):
      return self.__theCallable(*args, **kwds)



This prints:

Try to create my car:
Name not in available cars
Add my car.
Cars available now:
['porsche', 'volvo', 'trabbi', 'Benz']

['porsche', 'volvo', 'trabbi', 'Benz']


Hope this helps

Cheers
Franz

"Thomas Guettler" <zopestoller at thomas-guettler.de> schrieb im Newsbeitrag
news:3CC57C5F.7040804 at thomas-guettler.de...
> Is it possible to have methods in classes
> which don't take a self argument, because
> they only change attributes common to all
> instances of a class?
>
> class car:
>      available=["porsche", "volvo", "trabbi"]
>      def __init__(self, name):
>         if name not in available:
>             raise "Name not in available cars"
>         ....
>      def add_available_car(name):
>          availalble.append(name)
>
> One solutionn is to define this method
> at module level. But that's not nice.
>
>
>
> Unfortunately I have to stick to Python2.1
>
>    thomas
>




From brueckd at tbye.com  Wed Apr 17 10:28:55 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 17 Apr 2002 07:28:55 -0700 (PDT)
Subject: pre-PEP for optional 'pass'
In-Reply-To: <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1>
Message-ID: <Pine.LNX.4.33.0204170711450.10117-100000@ns0.tbye.com>

On Wed, 17 Apr 2002, Duncan Booth wrote:

> <brueckd at tbye.com> wrote in
> news:mailman.1018972676.30534.python-list at python.org:
>
> >> You should always head each method and class with a docstring.
> >
> > No! You should head each method and class with a docstring *only if
> > the docstrings would add value*. Otherwise you're doing more work now
> > AND later (during refactoring) without producing any real benefit,
> > cluttering the code in the process.
>
> I kind of agree with you while standing by my original statement.

(IOW: "I must get in the last word" :-) )

> In the context of the original message, where the poster was proposing
> writing a function that only contained 'pass', they could always use a
> good docstring instead.

Perhaps, with heavy emphasis on 'good'. This may seem like a nitpick, but
it does play an important part in writing clean, readable code. The
absolute best docstring is none at all because something is so darn
obvious that it doesn't need one.  Many (but not all) get/set methods fall
into this category.

> Even if the docstring simply says that the function is
> unimplemented this provides useful information to anyone who does not have
> the original source in front of them.

Quite an esoteric case, really: someone has the .pyc of a half-baked .py
with no explanation of what is unfinished. Most people won't read the
docstrings anyway, so in the case it'd be better to raise a not
implemented exception.

Docstrings and comments put in your code just to satisfy some sort of
documentation requirement ("all classes and all methods should have a
docstring") are distracting and detrimental to code readability and make
refactoring more of a chore. I believe _Code Complete_ has a chapter or
two worth reading on this subject, and the author does a much better job
explaining it that I have.

Moving on....

-Dave





From mrstephenson at dof.se  Sun Apr 14 00:22:51 2002
From: mrstephenson at dof.se (Rod Stephenson)
Date: 14 Apr 2002 06:22:51 +0200
Subject: Unposting tearoff menus under win32...
Message-ID: <wkbscmj4dg.fsf@dof.se>

I'm using Tkinter and the tearoff menu feature. However in certain
situations I'd like the torn-off menu to go away without user
interaction. 

Using Tk I can say 
	winfo children .kMain
where .kMain is the toplevel window, and get something like

.kMain.menubar .kMain.#kMain#menubar .... .kMain.tearoff1 .kMain.tearoff2

with .kMain.tearoff1  and .kMain.tearoff2 being a couple of tornoff
menus; I can then destroy these directly. 

Trying this under Tkinter I get an exception: 
root=Tk()
...
print root.winfo_children()
.....
  File "C:\PYTHON22\lib\lib-tk\Tkinter.py", line 609, in winfo_children
    self.tk.splitlist(self.tk.call(
  File "C:\PYTHON22\lib\lib-tk\Tkinter.py", line 979, in nametowidget
    w = w.children[name]
KeyError: #11512128

which I assume has something to do with the Tkinter naming convention
for menus. If I use root.children instead I get a list of the
children, but the same list whether or not the tearoff menus are
posted. I assume this means that in Tkinter they are implemented as
children of some other (non toplevel) widget, but I'm still trying to
track this down.

Back to the main task (just getting rid of the tearoff), is there
another way to accomplish this? unpost() apparently doesnt work under
win32. Any insights appreciated...

Thanks

Rod Stephenson




From Oschler at earthlink.net  Mon Apr  1 11:58:23 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 11:58:23 -0500
Subject: IDLE for SUSE Linux?
Message-ID: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com>

I'm looking for a good of binaries, or a good URL concerning setting up IDLE
under SUSE Linux 7.3 Pro.  Got URL?

Also, if anyone has any experience working with Python under KDeveloper
(KDE) I'd love to hear about it.  As soon as I get Glimpse/GlimseIndex
installed I'm going to give it a whirl.

thx






From rapskat at hotmail.com  Thu Apr 25 18:37:19 2002
From: rapskat at hotmail.com (rapskat)
Date: Thu, 25 Apr 2002 22:37:19 +0000
Subject: PyGTK libs for Zaurus?
Message-ID: <aaa0ev$963r5$1@ID-139866.news.dfncis.de>

Anyone know anything about installing the PyGTK ligs on the Sharp Zaurus
5500?

I'm using Qutopia for the unit's  WM/desktop, would GNOME be required?  If
so, I guess this is just a pipe dream then.

Thanks!

-- 
rapskat  - 10:35pm  up 1 day,  4:18,  3 users,  load average: 0.12, 0.11, 0.09

You Know You Need To Upgrade When...
The only web you know of  is the one you found on your hard drive.


From vvainio at tp.spt.fi  Tue Apr 23 05:13:10 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 23 Apr 2002 02:13:10 -0700
Subject: Portage (of Gentoo)
Message-ID: <ad496f8.0204230113.15f9731e@posting.google.com>

So, how usable is Portage when taken out of Gentoo context? I'm
thinking standalonePortage here... has anyone used it for any "own"
stuff, like building commercial projects etc.?

-- Ville


From gerhard at bigfoot.de  Sat Apr 13 23:01:50 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 14 Apr 2002 05:01:50 +0200
Subject: 2.1 vs. 2.2
In-Reply-To: <a9appm$l47$0@216.39.172.122>
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <mailman.1018738411.20464.python-list@python.org> <a9appm$l47$0@216.39.172.122>
Message-ID: <20020414030150.GA8965@lilith.my-fqdn.de>

* Bengt Richter <bokr at oz.net> [2002-04-14 02:33 +0000]:
> On Sun, 14 Apr 2002 00:41:37 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> wrote:
> 
> >* Jens Baader <nospam at nospam.com> [2002-04-13 23:46 +0200]:
> >                      ^^^^^^^^^^
> >
> >If you don't have permission of the nospam.com domain owner to use their
> >domain, you should stop abusing their services *now*. If you don't want
> >to get spam, I'd recommend you get a throwaway email address at a
> >freemailer.
> >
> I agree, but OTOH you'd think the IETF would come up with a standard
> global /dev/null sort of address for this purpose. Is there such a thing?
> I.e., why burden freemailers with garbage storage, and the net with the traffic?

That's already been done:

http://www.rfc-editor.org/rfc/rfc2606.txt

<quote>
    To safely satisfy these needs, four domain names are reserved as
    listed and described below.

    .test
    .example
    .invalid
    .localhost

    ".test" is recommended for use in testing of current or new DNS
    related code.

    ".example" is recommended for use in documentation or as examples.

    ".invalid" is intended for use in online construction of domain
    names that are sure to be invalid and which it is obvious at a
    glance are invalid.
</quote>

So if you really *never* want to receive email replies from usenet
posts, you should construct a fake email address and add .invalid at the
end.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From tanzer at swing.co.at  Tue Apr  9 03:05:30 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Tue, 09 Apr 2002 09:05:30 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "08 Apr 2002 17:54:35 GMT."
             <a8slgr$pin$1@nyheter.chalmers.se>
Message-ID: <E16uph4-0001JQ-00@tswings.swing.cluster>

jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen) wrote:

> In article <3CAF524F.63E46C81 at alcyone.com>,
> Erik Max Francis  <max at alcyone.com> wrote:
> >Jacob Hallen wrote:
> >
> >> Python is a language that is weakly typed, ...
> >
> >You mean dynamically typed, not weakly typed.
>
> I have been brought up in an environment which called anything
> allowing mixed types in a list weakly typed.

By this reasoning all object-oriented languages would be called weakly
typed. Possible but not very helpful (and decidedly non-standard).

> If you prefer to call it dynamically typed, I don't mind.

There are several dimensions to typing, including:

- static vs. dynamic

- strong vs. weak

Python's typing is dynamic and strong. Calling it weakly typed
lumps it together with languages like Perl. I would argue that
Python's typing is stronger than for instance C++'s.

Another important point with typing is whether typing applies to
variables or to objects. In Python, objects are (strongly) typed, not
variables. Thinking `variable` when considering typing (or objects in
general) will not work very well in Python.

Names-don't-matter-except-when-they-do-ly yr's

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





From sjmachin at lexicon.net  Fri Apr 26 19:30:07 2002
From: sjmachin at lexicon.net (John Machin)
Date: 26 Apr 2002 16:30:07 -0700
Subject: Converting a Python List to a C double *
References: <k0n4rhy3gtn.fsf@metropolis.phys.ntnu.no> <jPfy8.65668$vF6.2006478@news2.tin.it>
Message-ID: <c76ff6fc.0204261530.3b9fa701@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<jPfy8.65668$vF6.2006478 at news2.tin.it>...

>     xdata = malloc(listlength * sizeof(double));

Joakim's platform may have not the Cornucopia version of malloc() but
the Mother Hubbard version -- a test of the returned value may be a
good idea.

>     for(i=0; i<listlength; i++) {
>         temp = PySequence_GetItem(thelistofdoubles, i);
>         if(!temp) {
>             free(xdata);
>             return 0;
>         }
>         if(!PyFloat_Check(temp)) {
>             Py_DECREF(temp);
>             free(xdata);
>             PyErr_SetString(PyExc_TypeError, "all items must be float");
>             return 0;
>         }
>         xdata[i] = PyFloat_AS_DOUBLE(temp);
>         Py_DECREF(temp);
>     }


From davebutlerREMOVE at hotmail.com  Wed Apr  3 12:44:36 2002
From: davebutlerREMOVE at hotmail.com (Dave Butler)
Date: Wed, 3 Apr 2002 11:44:36 -0600
Subject: Jython Books
Message-ID: <ddHq8.11630$1a.2753554@twister.kc.rr.com>

A search of Amazon.com reveals two books dedicated to Jython (Pedroni &
Rappin and Robert W. Bill). If anyone has read them both (or at least
browsed), please share your opinions.

Thanks,

Dave Butler




From bokr at oz.net  Mon Apr  1 17:33:15 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 22:33:15 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com>
Message-ID: <a8an7b$eme$0@216.39.172.122>

On Mon, 01 Apr 2002 11:28:12 -0800, Erik Max Francis <max at alcyone.com> wrote:
[...]
>> 2 Or, as the PEP suggests, introduce a bool type which behaves quite
>>   unintuitive, but keeps existing programs running.
>
>I don't see why you think the behavior (I presume you're referring to
>syntaxes like True == 2) is counterintuitive.  Why _would_ you expect
>this to do what you want?  It doesn't for any other Python types that
>I'm aware of.  == tests for equality, it doesn't do automatic type
>coercion.
>
Well, for types that are related in certain ways ... ;-)

 >>> 0 == 0.0
 1
 >>> 1 == 1.0
 1
 >>> 1 == 2.0
 0

That gets into what relational operators really should mean.
I think I exclude bool from the number tree, so I don't
like the effect of subtyping bool from int, even though
I support context-dependent implicit conversions to integer
(though that context should be defined by the client expression ops).

Regards,
Bengt Richter


From pobrien at orbtech.com  Wed Apr  3 18:22:25 2002
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Wed, 3 Apr 2002 17:22:25 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org>
Message-ID: <NBBBIOJPGKJEKIECEMCBKEGEMLAA.pobrien@orbtech.com>

> Guido van Rossum wrote:
> > False and True are numbers [...]
>
> That's exactly where this falls down.
>
> Patrick O'Brien wrote:
> > [...] True and False could be dictionary keys separate from 0 and 1.
>
> Guido van Rossum wrote:
> > No.  That would break backwards compatibility.  False==0, and True==1;
> > everything else follows from that.  (But False is not 0, and True is
> > not 1!)
>
> Ka-Ping Yee wrote:
> This is a strange distinction to make -- one that is not made in any
> other programming language i have ever encountered.  Of course *i* get
> it and *you* get it -- but it will be so hard to teach this particular
> weirdness that i consider it fatal to the proposal.

This strange distinction seems to be the crux of the controversy surrounding
this PEP. I thought it a bit strange at first too. But the more I think
about it the less it troubles me. In fact, it is starting to feel downright
Pythonic.

True and False are singleton instances of a Boolean subclass of the Integer
class. As a subclass, much of, but not all of, their behavior is inherited
from Integer. So False often acts like 0 and True often acts like 1. But not
always. Because they are instances of a subclass. That's what subclassing
allows. This is Object Orientation 101.

I think the concepts truly are no less difficult to understand or explain
than plenty of other aspects of the Python language. The way Boolean is
being implemented may not be what one would expect, and it may not be what
other languages do, and it may take a little getting used to, and it may
upset some people's preconceptions of truth values, but I think if everyone
steps back and looks at this objectively they have to admit that this
"strange distinction" really isn't too terribly strange after all.

There are plenty of Python features that still make my head hurt. This new
Boolean isn't one of them.

---
Patrick K. O'Brien
Orbtech





From jacek.generowicz at cern.ch  Sat Apr 27 12:57:10 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 27 Apr 2002 18:57:10 +0200
Subject: Python's Lisp heritage
References: <mailman.1019922475.8361.python-list@python.org>
Message-ID: <tyfsn5h13mh.fsf@pcitapi22.cern.ch>

"Brian Quinlan" <brian at sweetapp.com> writes:

> Jacek wrote: 
> > My example was indended to disabuse James of the notion that
> > parentheses are used in lisp "around sub expressions".
> 
> How would you convert the following Python expression to Lisp?
> 
> sin(x/2 + y/2)

#i(sin(x/2 + y/2))

I'm guessing that you want me to answer (sin (/ (+ x y) 2)) so that
you can then triumphantly point out that this is putting parentheses
around sub expressions. But then you would be confusing the appearance
with the purpose. (I already acknowledged in my original followup to
James, that in mathematical expressions function calls and
sub-expressions coincide rather frequently.)

The purpose of the parentheses (in this context) is to call functions
and associate parameters with those function calls. Python has a not
too dissimilar use of parentheses:

  sin(operator.div(operator.add(x,y),2))

Oh look, exactly the same number of parentheses!

What? you object to my obtuse avoidance of Python's nifty infix
synonyms for these operators which cunningly call the functions
without recourse to parentheses?  In that case you surely allow me to
object to obtuse avoidance of such possibilities in Lisp.

In summary, a pair of parentheses is necessary to call functions in
both Python and Lisp, though there are ways of getting around this
requirement in both languages: Python provides a set of infix
operators which call a subset of the mathematical functions; there are
Lisp macros which provide very similar functionality.


Rhetorical question:

  Imagine you use the splark function so often in your expressions
  that you would really like to extend the set of infix operators,
  with say @, to allow you to call splark without having to type those
  infernal parentheses. Is it going to be easier in Python or in Lisp?

  Another one liner, right?


From claird at starbase.neosoft.com  Mon Apr  8 12:26:55 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 8 Apr 2002 11:26:55 -0500
Subject: Adjuntar ficheros en python.
References: <mailman.1018265955.20370.python-list@python.org> <slrnab3bs2.f1.gerhard@lilith.my-fqdn.de>
Message-ID: <3776A7C82CC557D4.05DFDD66985981E9.309AF811BFDFE3AD@lp.airnews.net>

In article <slrnab3bs2.f1.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>I won't pretend to be able to *write* Spanish, so I'll answer in
>English.
>
>jmanzano at antena3tv.es wrote in comp.lang.python:
>> Hola
>> 
>> Alguien me puede decir como puedo adjuntar ficheros en python a la
>> hora de enviar emails.
			.
			.
			.
>For even older Pythons, you'll have to play some tricks with the
>mimetools package yourself. I've seen answers on comp.lang.python for
>that, so searching groups.google.com for "email mime attachment"
>should give some options.
			.
			.
			.
I collect alternative implementations that might be valuable in <URL:
http://starbase.neosoft.com/~claird/comp.unix.programmer/attachments.html >.
-- 

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


From ping at lfw.org  Wed Apr  3 11:44:01 2002
From: ping at lfw.org (Ka-Ping Yee)
Date: Wed, 3 Apr 2002 10:44:01 -0600 (CST)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <200204031619.g33GJXm02577@europa.research.att.com>
Message-ID: <Pine.LNX.4.33.0204031031440.16190-100000@server1.lfw.org>

Andrew Koenig wrote:
> Specifically: If I want to write a function that answers a yes/no
> question, I have lots of possible ways of spelling yes (1, 2, "yes",
> and so on) and lots of possible ways of spelling no (0, {}, None,
> and so on).  There isn't a single preferred way.

Fredrik Lundh wrote:
> if you study real python code, you'll find that there is.

Andrew Koenig wrote:
> I mean that the language doesn't prefer a single way,
> as opposed to programmers' conventions.

This doesn't mean anything in practice.  Every yes/no function
i've encountered returns 1 for yes and 0 for no.  (If it returns
something else, then it really has more to say than yes or no.)

In fact the language actually does prefer a particular way,
since x == x returns 1.

The addition of another type really does change the set of preferred
truth values from {0, 1} to {0, False, 1, True}.  The typical practicing
Python programmer who wants to say "true" would have to decide -- for
each instance -- whether to say "1" or "True".

And the programmer cannot happily ignore the issue and stick to using
0 and 1, because if comparisons start to return True and False, dealing
with a mix of four truth values is unavoidable.  This PEP makes me uneasy
because it causes me to see lots of extra casts in Python's future (not
to mention confused beginners).


-- ?!ng





From aahz at pythoncraft.com  Thu Apr 11 21:11:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 21:11:16 -0400
Subject: PEP 282: A Logging System
References: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru> <mailman.1018538209.18235.python-list@python.org>
Message-ID: <a95c7k$801$1@panix1.panix.com>

In article <mailman.1018538209.18235.python-list at python.org>,
Vinay Sajip <vinay_sajip at yahoo.co.uk> wrote:
>
>That's to say, I expect it will be thread-safe in the not-too-distant
>future, but I have not yet tested it in conditions such that I can
>confidently comment on its thread-safety. When writing it I aimed not to use
>instance data or global data (other than read-only) except where absolutely
>necessary, so thread-safe operation should be straightforward to achieve.
>The main areas of work to ensure thread-safe operation are being worked on
>at the moment - they are:
>
>(a) Serializing access to the hierarchy structure and logging level tables,
>which are global.
>(b) Checking the code in detail to ensure that there is no thread-unsafe
>code.
>(c) Allowing for on-the-fly reconfiguration by a thread while other threads
>are using the system for logging.
>(d) Coming up with a good test harness designed to try and break the module
>in terms of thread-safety.

Your main difficulty with thread safety is WRT I/O.  How do you plan to
handle that?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From philh at comuno.freeserve.co.uk  Mon Apr 15 10:18:14 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 15 Apr 2002 15:18:14 +0100
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs>
Message-ID: <slrnablo96.bcr.philh@comuno.freeserve.co.uk>

On Mon, 15 Apr 2002 11:03:04 +0300, Max Ischenko <max at malva.com.uaREMOVE.IT> wrote:
>
>Hi there.
>
>I have a Python program that uses data stored in XML file to generate some
>C++ code. The problem (or, more precisely, code smell) I have is that the
>actual generation algorithm is hardwired into python and therefore is
>cryptic, difficult to understand and maintain.
>
>Small example:
>	def genBody(self, c):
>		name = c.cxxClassName
>		new = 'new %s(%s);' % (name, c.factoryCtorArg or '')
>		if c.factoryNargs == 0:
>			return '\treturn %s' % new
>		body = '\t%s * c = %s\n' % (name, new)
>		body = body + '\tc->setCmdArgs(args);\n'
>		return '%s\treturn c;' % body

How about:

template1 = """\
   return new $name($ctorArg);
"""

template2 = """\
   $name * c = new $name($ctorArg);
   c->setCmdArgs(args);
   return c;
"""
   
   def genBody(self, c):
      substitutions = { 
         "name": c.cxxClassName,
         "ctorArg": c.factoryCtorArg or ''}
      if c.factoryNargs == 0:
         return subst(template1, substitutions)
      else 
         return subst(template2, substitutions)


   def subst(formatString, args):
      f = ... in formatString convert all $xxx to %(xxx)s ...
      return f % args

The implementation of subst() is left as an exercise.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From aahz at pythoncraft.com  Thu Apr  4 23:13:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 23:13:45 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <7x7knnvsc6.fsf@ruckus.brouhaha.com>
Message-ID: <a8j89p$cmn$1@panix1.panix.com>

In article <7x7knnvsc6.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2002a at nightsong.com> wrote:
>
>   print "Calm down, it's only %s's and %s's." % (1==1, 1==0)

"Calm down, it's only Banned From Argo."
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From matt at greenroom.com.au  Thu Apr 11 02:14:56 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Thu, 11 Apr 2002 16:14:56 +1000
Subject: Sound programming?
References: <3CB50DBC.1B0E533F@alum.mit.edu>
Message-ID: <3CB529E0.3010807@greenroom.com.au>

Will Ware wrote:

> Somebody was asking me this evening if there's an easy way to
> play a tone of a specified frequency and duration in Python.
> I didn't know off the top of my head. I've since poked around
> the web a bit and it appears like the Snack Sound Toolkit
> (http://www.speech.kth.se/snack/) might be a good way to go.
> One of its virtues is being cross-platform, which I suspect is
> among his unstated wishes. Are there any other options?
> 

There aren't any functions (that i'm aware of) of that nature in the 
standard python libs (except for manipulation of *.wav, *.au) .etc. Of 
the options like gump and pyXMMS (not cross platform), i'd consider 
snack to be one of the best although being wrapped around tcl like tk grrr!.



From sjmachin at lexicon.net  Sun Apr 21 22:45:08 2002
From: sjmachin at lexicon.net (John Machin)
Date: 21 Apr 2002 19:45:08 -0700
Subject: Pyxie, anyone?
References: <Xns91F7F276A63DAgustafl@195.100.94.182>
Message-ID: <c76ff6fc.0204211845.6d17014f@posting.google.com>

Gustaf Liljegren <gustafl at algonet.se> wrote in message news:<Xns91F7F276A63DAgustafl at 195.100.94.182>...
> I have problems using the pyxie library for processing XML. This page
> 
>   http://pyxie.sourceforge.net/
> 
> says that "To run Pyxie, you need two things:
> 
> - A Python distribution
> - The XML package for Python [...]"
> 
> I have the latest versions of both and still I get the error "ImportError: 
> No module named pyxie" when I try to import the module. Has Pyxie been 
> phased out from the XML package?
> 
> Gustaf

Is it possible that "To run Pyxie, you need two things" (which appears
under a heading "dependencies") should be interpreted as "To run
Pyxie, you need, not only Pyxie itself, but also two *other*
things"??? IOW, have you downloaded Pyxie from
http://pyxie.sourceforge.net/ ???


From tanzer at swing.co.at  Tue Apr  9 03:13:48 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Tue, 09 Apr 2002 09:13:48 +0200
Subject: Nested scopes hitch
In-Reply-To: Your message of "Mon, 08 Apr 2002 17:43:54 PDT."
             <MPG.171bd911b760bfa19896a0@news.nwlink.com>
Message-ID: <E16upp6-0001Js-00@tswings.swing.cluster>

Jeff Shannon <jeff at ccvcorp.com> wrote:

> In article <mailman.1018310800.31349.python-list at python.org>,
> pinard at iro.umontreal.ca says...
> > [Jeff Shannon]
> >
> > > If you want access to variables inside of a function, you should
> > > pass those specific variables as arguments.  If you've got lots
> > > of arguments in common between lots of functions, assemble the
> > > lot into a class.  If neither of these is appropriate, then you
> > > should redesign -- you're using a very ugly approach.
> >
> > I beg to disagree, a bit strongly.
> >
> > Something is not ugly just because you prefer something else.  I have very,
> > very neat applications which use globals() and locals(), and they will
> > loose much if they use arguments all over, or get assembled into classes.
>
> Very well, I suppose we will have to disagree, then.  To my mind,
> using globals() and locals() in this way is a gross violation of
> encapsulation.

Sometimes it is, sometimes it isn't.

Without knowing the problem it's hard to tell.

> I obviously can't speak to your specific designs, not having seen
> them, but I can't imagine how globals()/locals() would be
> necessary unless you have a large amount of interdependencies in
> your code

Try to imagine interactive access to an application, or a function
displaying all local variables in case of an exception, or ...

Dogmatism-doesn't-improve-SW-design-ly yr's

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





From vvainio at tp.spt.fi  Thu Apr 18 02:44:38 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 17 Apr 2002 23:44:38 -0700
Subject: bug in build? configuration --with-cxx ends up using PY_CFLAGS instead of CPPFLAGS
References: <ad496f8.0204170006.596b6818@posting.google.com> <m3sn5uvi9x.fsf@mira.informatik.hu-berlin.de>
Message-ID: <ad496f8.0204172244.13246ac7@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3sn5uvi9x.fsf at mira.informatik.hu-berlin.de>...

> CPPFLAGS are just the flags for the preprocessor, so they are clearly
> the wrong set of flags. Can you elaborate which flags is being passed
> that should not? In any case, you need both OPT and CFLAGSFORSHARED
> when compiling ccpython.

I used gcc as my c compiler, and aCC as a C++ compiler to 1) compile
ccpython and 2) link. aCC can't digest stuff that gcc can, e.g. -O3.
Standard build does try to feed -O3 to aCC. aCC did seem to be able to
compile ccpython with just CPPFLAGS. However, not everything seems to
work without a hitch at the moment, but I assume these are other,
unrelated problems (that I have trying to make omniORBpy work with
hp-ux).

Or should I just use hp-ux c compiler (cc) instead of gcc? 

-- Ville


From borcis at geneva-link.ch  Wed Apr 10 07:53:55 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Wed, 10 Apr 2002 13:53:55 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk>
Message-ID: <3CB427D3.7020404@geneva-link.ch>


Anders J. Munch wrote:

> Pythons truth values are not "Something or Nothing".  The phrase seems
> to have caught on following Laura's post, but I believe it is
> misleading.
> 
> [] and {} are objects, they are objects with identity, and they are
> mutable objects.  They are empty, but they are by no means "nothing".


Yeah but test expressions provide a context of abstract interpretation,
were their coalescence as eg shadows of zero makes beautiful sense. One 
needs to note that in a "deepest" linguistic sense, zero is everywhere; 
more precisely, everywhere grammar would allow to splice in a supplement 
to the code...guess I need to expose a bit of my linguistic views to 
make myself clear.

Language - and natural language I have first in mind here - is 
structured by paradigms. A simple manner to define a paradigm, is : the 
set of variants that I could fit in place of some segment of (text, 
code), in the latter's context. Then comes a first classification of 
segments, segments that have the linguistic zero in their paradigm, and 
segments that don't. One could probably grow that division to the 
separation between grammar and lexemes, grammar relating to the segments 
equivalent to the linguistic zero, and lexical information relating to 
atomic units not by themselves repleaceable with a linguistic zero 
(=litteral nothing).

I feel the multiplicity of "false" objects, e.g. 0, [], (,), {}, None - 
and now False itself, in Python, first deserves notice as one of 
Python's most brilliant idiosyncrasies, and relates to what I just said 
above, by viewing the primary data types/structures as embedded 
mini-languages over which python code rules as a meta-language. As a 
meta-language, it's necessary that Python code can distinguish between 
the (meta-representations of) zeroes of the embedded mini-languages by 
type. At the same time, the uniform treatment of them in test 
expressions contexts, has a feeling of Big Truth to it and (I believe) I 
fully understand how Laura launched in defending the elaborate 
interpretation of that feeling that she grew while teaching Python.

I could discuss the "Something" side of the matter, but I feel the point 
you raise here very much boils down to "the type of zero is a manifest 
information that disallows viewing the typed zero as a nothing".

To sum up, I certainly agree that Python is a better and richer language 
than it would be if all other things being equal, it had a bool type and 
every test expression should first evaluate to a value of that type. But 
this is not the change that the PEP introduces, AFAIK. Fortran and 
Pascal have this, and while I understand some people feel they are 
better morals to rigidly typed booleans, to me Python's current state of 
matters is both, more interesting, simpler to use, and mysteriously 
linked to subtle "metaphysical" (meta-mathematical actually imho) 
justifications that while (a) fail having been made explicit (b) have no 
real reasons to be less compelling than "the known morals" of booleans.

Regards, Boris Borcic
--
" 12 ? The least integer symbolizing the natural numbers by itself.
  Successors : 123, 1234..."



From aleax at aleax.it  Thu Apr 11 04:35:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 08:35:24 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de> <a91mdd$ck$1@panix1.panix.com> <slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>
Message-ID: <gVbt8.5591$b62.137508@news1.tin.it>

Gerhard H?ring wrote:
        ...
> IMO using your real name is polite and whatever you say is not worth
> being said if you can't put your name under it.

Yet "Primary Colors" was lots of fun to read -- which to me means it
WAS worth writing, even though the author couldn't safely sign it.

The "ad hominem" fallacy is to judge some argument on the basis of
who said what, rather than on what is being said.  We're all prone
to it, of course -- it's part of human biology that the same words,
spoken from a Leader, a Loser, or a Contrarian, receive different
consideration.  But to assert that it MUST be made possible for you
to apply the "Pecking Order" criterion to an argument, or else the
argument was "not worth saying", is to elevate "ad hominem", from a
fallacy excusable on biological grounds, to a guiding principle of
discourse.  Try considering arguments on their merit, instead.

[Of course, I _can_ turn around and argue the other side, a little:
cfr. the solutions to iterated Prisoner's Dilemma problems that
are only enabled by the ability to identify your counterpart, for
example - that can be subsumed in there being sound pro-survival
reasons behind our biological leanings, as should be obvious...
if those leanings weren't adaptive in long-prevailing circumstances,
they'd long have been bred out of the collective genome.  But how
does our current environment compare with those circumstances?-)]

Me, I sign because I'm an egotist and like seeing my name written.

Of course, should I be flaming, I might well pounce on some
interlocutor's pseudonimity as an excuse to insult him or her,
if I couldn't sink my teeth into something more substantial.


Alex



From lal at pro.onet.pl  Wed Apr 24 08:55:19 2002
From: lal at pro.onet.pl (Artur Lipowski)
Date: Wed, 24 Apr 2002 12:55:19 +0000 (UTC)
Subject: atoi
References: <mailman.1019649779.5458.python-list@python.org>
Message-ID: <slrnacdaqp.16g.lal@artur2.krakow.tpg.pl>

In article <mailman.1019649779.5458.python-list at python.org>, 
Mariappan, MaharajanX wrote:
...
> is there any python buildin method whihc will convert strings into integer,
> like atoi in c lib.?
> "234" should be converted into interger value 234
...
int() and company.
Look at "Python Library Reference" chapter 2.1 "Built-in Functions". Also 
using searchable Python manual at 
http://starship.python.net/crew/theller/pyhelp.cgi 
with "conversion" keyword will find out what you need.
In other words RTFM 8-)

Regards,
--
Artur Lipowski


From eric.brunel at pragmadev.com  Mon Apr 15 12:39:18 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 15 Apr 2002 16:39:18 +0000
Subject: Looking for GUI Form Builder for Python
References: <a9en77$4vf8@imsp212.netvigator.com>
Message-ID: <a9eodu$2c5p$1@norfair.nerim.net>

Calvin wrote:
> Dear all,
>  Can I build Python TK GUI interface using a form builder. Anyone know
>  such
> builder. I just want to build my interface in a visual way.
> -Calvin

Try Mike Clarkson's SpecTix:
http://starship.python.net/crew/mike/src/Spectix/Spectix.html
It's multi-platform and can generate Python code.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From starx at pacbell.net  Wed Apr 17 06:41:05 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 17 Apr 2002 10:41:05 GMT
Subject: Unicode name questions
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com> <009901c1e58b$28a5bd90$445d4540@Dell2> <mailman.1018994314.14044.python-list@python.org> <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Xns91F325AA52316RASXnewsDFE1@130.133.1.4>

Martin v. Loewis || Wed 17 Apr 2002 01:25:23a:

> Skip Montanaro <skip at pobox.com> writes:
> 
>> "Lambda" has been spelled with a "b" as long as I can remember.  
> 
> I think the Unicode charts use an ASCII transcription of the native
> pronouncation of the letters, if possible. In any case, the Unicode
> names for the characters are "official" in the context of Unicode.
> The Unicode consortium refrains from renaming them. Between Unicode
> 2.0 and Unicode 3.2, no character was renamed.
> 
> [Etc Etc Greekness]

Hm. I was browing through some of their code charts for the heck of it, 
and I noticed that they listed certain kata/hiragana characters's primary 
names as SI, TI, TU, etc. I've listened to a considerable amount of 
japanese media, and AFAICT the correct romanization of those sounds would 
be SHI, CHI, TSU, etc.

But then they also have some other wierdness, e.g., listing VA, VI, VE, VO 
as WA, WI, WE, WO + diacrit, where WI and WE are old, obsolite* characters 
as mentioned in my dictonary, and to the best of my knowledge the methods 
for doing V sounds are eihter old-style use of B sounds, or new-style use 
of A,I,E,O + diacrit. And there's this 'katakana iteration symbol' (and 
hira, and voiced of both) that i've never so much as heard of... :\

(*I can't for the life of me find a spelling of this word that i'm sure 
of, and I don't have checker or dictionary or time on hand... bed. now.)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From gerhard at bigfoot.de  Tue Apr 16 00:43:34 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Tue, 16 Apr 2002 06:43:34 +0200
Subject: Why don't strings share data in Python?
In-Reply-To: <00b401c1e4fe$41b82da0$0101010a@local>
References: <87d6x09trz.fsf@mathdogs.com> <00b401c1e4fe$41b82da0$0101010a@local>
Message-ID: <20020416044334.GB10507@lilith.my-fqdn.de>

* Chris Gonnerman <chris.gonnerman at newcenturycomputers.net> [2002-04-15 23:21 -0500]:
> From: "Mike Coleman" <mkc+dated+1021521407.f909ec at mathdogs.com>
> 
> > Does anyone know why strings (i.e., those of length >1) don't share
> > their data in Python?  Since their immutable, it seems like this
> > would be the obvious thing to do.  So, for example, the space
> > behavior of this code could be linear rather than
> > quadratic/horrific:
> >
> > d = {}
> > for i in xrange(100000000):
> >     d[mybigstring[i:]] = mybigstring[i:]
> 
> [...]
> Are you certain Python doesn't do it?

It doesn't. There's only one optimization in
Objects/stringobject:string_slice, and it's about slicing the whole
string:

>>> x = "hello"
>>> id(x)
134867304
>>> y=x[:]
>>> id(y)
134867304

> Can you submit a patch to implement it? (PEP + Patch == Implemented
> Change in many cases).

Why a PEP? This isn't a language change, only an implementation detail.
No PEP needed IMO.

The problem with implementing this is that it wouldn't exactly simplify
garbage collection of strings ;-) If it can be efficiently and safely
implemented, then I'd like to see this change, though. I'm not sure if
this is possible, and if it's worth the effort.
 
Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From sholden at holdenweb.com  Tue Apr 30 09:13:46 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 09:13:46 -0400
Subject: Call for kjBuckets.c [was Re: relational database?]
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <mailman.1020024596.32196. <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com> <200204301324.25823.rjones@ekit-inc.com> <mailman.1020153536.6149.python-list@python.org>
Message-ID: <NOwz8.13638$xy4.7951@atlpnn01.usenetserver.com>

"Oleg Broytmann" <phd at phd.pp.ru> wrote ...
> On Tue, Apr 30, 2002 at 01:25:41PM +1000, Richard Jones wrote:
> > I'm working on repackaging gadfly right now and will place the  results
up on
> > gadfly.sf.net. I've started with the Zope version, so the re conversion
is
> > done. I've also integrated Oleg's kjbuckets efforts. I'm almost there...
just
> > got a few more wrinkles  to sort out before I make the first release.
>
>    Thank you! We are waiting...
>

I echo Oleg's thanks. Gadfly is great for people who don't have an existing
relational system but want to do some Python experimentation with database.
Let me know when you need testers.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From kfw at woeltje.org  Sat Apr 27 15:42:55 2002
From: kfw at woeltje.org (Keith Woeltje)
Date: Sat, 27 Apr 2002 15:42:55 -0400
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <3CCAFF3F.9020302@woeltje.org>

Although the question has  been amply answered with references to the 
tutorial et al., I came across this reference today, and thought of this 
thread.

http://docs.sun.com/htmlcoll/coll.648.2/iso-8859-1/NUMCOMPGD/ncg_goldberg.html

It is an edited reprint of the paper What Every Computer Scientist 
Should Know About Floating-Point Arithmetic, by David Goldberg, 
published in the March, 1991 issue of Computing Surveys




From philh at comuno.freeserve.co.uk  Mon Apr 22 09:36:36 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 22 Apr 2002 14:36:36 +0100
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83> <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>
Message-ID: <slrnac84f4.1c6.philh@comuno.freeserve.co.uk>

On 22 Apr 2002 10:50:37 GMT, Jarno J Virtanen <jajvirta at cc.helsinki.fi> wrote:
>
>I find every other metaphor lacking something essential too.  One
>popular metaphor is building construction or engineering activity in
>general.  Personally I just don't get the feeling that there's any of
>those so called components that the final product is made of in computer
>programming.  These components are supposed to be realiable (and in a
>way simple) yet provide something significant to the product.  I can't
>figure any non-trivial and substantial programming component which can
>be viewed having the same reliability (or reliability properties, such
>as "this bolt holds the pressure of N") as the compontents that
>buildings are made of. 

Data structures such as lists and dictionaries.

Built-in libraries such as those for windowing, TCP/IP 
communication, etc.

Automatic storage management.

Imagine having to write a mailer or newsreader without any of them.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From starx at pacbell.net  Tue Apr  9 02:52:28 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 9 Apr 2002 06:52:28 GMT
Subject: multiple ranges to for statement?
References: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu>
Message-ID: <Xns91EAF30F4955CRASXnewsDFE1@130.133.1.4>

Damian Menscher || Mon 08 Apr 2002 11:33:04p:

> I'm hoping to do something along the lines of:
> 
> for i in range(a,b),range(c,d):
>     stuff goes here
> 
> But of course that sets i to the *two* values: range(a,b) and
> range(c,d).  Is there a simple way to get it to take on every value
> from a to b and then every value from c to d?  I would think there
> should be, but I'm a newbie and can't figure it out.
> 
> So far, the best I can think of would be:
> 
> values = range(a,b)
> for i in range(c,d):
>     values.append(i)
> for i in values:
>     stuff goes here
> 
> It just doesn't seem very clean, so I'm wondering if I'm missing
> something obvious.

There may be some gotchas here that i'm not seeing... Anyway, just
remeber that the range command returns a sequence. Sequences can be
concatanated. Here's an exapmle: 

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> range(1,5)
[1, 2, 3, 4]
>>> range(10,15)
[10, 11, 12, 13, 14]
>>> range(1,5) + range(10,15)
[1, 2, 3, 4, 10, 11, 12, 13, 14]
>>> for x in range(1,5) + range(10,15):
	print x,

1 2 3 4 10 11 12 13 14
>>> 

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From bokr at oz.net  Tue Apr 16 03:21:01 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 07:21:01 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <m3d6x1mqbu.fsf@mira.informatik.hu-berlin.de> <a9f4eh$pjp$0@216.39.172.122> <3CBB8660.A4859BAB@cosc.canterbury.ac.nz>
Message-ID: <a9gjct$odr$0@216.39.172.122>

On Tue, 16 Apr 2002 14:03:12 +1200, Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

>Bengt Richter wrote:
>> 
>> newhex(+7) => 0x7 # implied extension of msb => 0x000...007
>> newhex(-7) => 0x9 # implied extension of msb => 0xfff...ff9
>
>I think that would be very confusing. It's pretty
>subtle that the MSB of the binary equivalent of the
>first hex digit tells you whether the original number
>was positive or negative.

Familiarity breeds private illusions of clarity, I guess ;-)
I accept your point.

>
>If we were to have something like this, I would rather
>see
>
>  newhex(+7) => 0x07
>  newhex(-7) => 0xF9
>
>i.e. the leading hex digit is always either 0 or F.
>
An excellent suggestion!

Unfortunately, we have to worry about backwards compatibility, though,
so it might have to be introduced with 0h prefix instead of 0x.

>Then I would also want an extra arg to hex() for
>specifying how many digits I want.
>
Fine with me, too, if default is normalized to minimal prefixing.

Regards,
Bengt Richter


From starx at pacbell.net  Tue Apr 23 01:19:07 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 23 Apr 2002 05:19:07 GMT
Subject: Developing and ircd with python
References: <aa278j$8c8$1@nsnmrro2-gest.nuria.telefonica-data.net> <3CC4A648.3070100@jerf.org>
Message-ID: <Xns91F8E339D9A04RASXnewsDFE1@130.133.1.4>

Jeremy Bowers || Mon 22 Apr 2002 05:07:52p:

> Raphael Ribeiro wrote:
>> Is it possible?
> 
> Yes.
> 

There's something entirely perfect about this conversation...

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From imbosol at vt.edu  Thu Apr  4 15:52:26 2002
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 4 Apr 2002 15:52:26 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <aeei8a.i11.ln@127.0.0.1>

Guido van Rossum wrote:
> Despite the negative feedback, I've decided to accept the PEP. 


What a pity.  I've grown fond of referring to == as the Kronicker
Delta operator; no more.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From bokr at oz.net  Sun Apr  7 20:09:48 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 00:09:48 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk> <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk>
Message-ID: <a8qn4c$fmc$0@216.39.172.122>

On 7 Apr 2002 15:14:37 GMT, ralph at inputplus.demon.co.uk (Ralph Corderoy) wrote:

>Hi Max,
>
>> > I think we all perfectly understand your opinion too and no matter how
>> > many times you repeat it there will still be some of us that like to do
>> > 
>> >     a = b + is_movable(c
>> 
>> I can see that it is clever, but I don't get why it is smart. Isn't it 
>> just another way of writing?:
>> 
>>      if is_movable(c):
>>          a = b + 1
>
>No.  BTW, you snipped a closing parenthesis there.
>
>It's equivalent to
>
>    a = b
>    if is_movable(c):
>        a += 1
>
>or
>
>    if is_movable(c):
>        a = b + 1
>    else:
>        a = b
>
>or
>
>    a = b + is_movable(c) ? 1 : 0
>
>but that last one isn't valid Python but does show a strong C
>background is likely to make you familiar with this approach.
You could write that in Python as

    a = b + (is_movable(c) and 1 or 0)
>
>> If it is, I find that it is more obfuscation than expression. It's
>> akind to using a magic value. My paraphrase of our code should really
>> read something like::
>> 
>>      motionLength = 1
>>      if is_movable(c):
>>          a = b + motionLength
>
>That doesn't cover `not is_movable(c)' again.
>
>> Or your code should possibly have read::
>> 
>>      a = b + motionLength(c)
>> 
>> But probably I misunderstand you intentions.
>
>The quantity being added to b isn't a length, it's 1 indicating that c
>is movable.  Trying to name it anything other than what it is is
>contortion for some ideal that doesn't work in this case.
You could write

      a = b + oneIfMovableZeroIfNot(c)

if you wanted to be conspicuously clear about your function's intent.

>
>Say you had a list of items and wanted to sort them to place those with
>the most abilities first, with all abilities considered equal, and then
>sorting on the item's name where two have the same number of abilities.
>
>To calculate the major sort key you might do
>
>    major = can_be_moved(o) + can_be_sized(o) + can_be_stacked(o) +
>        can_be_coloured(o) + can_be_hidden(o)
>
It would express the semantics better if these were true boolean functions
returning True or False, and you wrote it as

    major = [can_be_moved(o), can_be_sized(o), can_be_stacked(o),
             can_be_coloured(o), can_be_hidden(o) ].count(True)

There is not much ambiguity in this form, whereas doing arithmetic leaves
a lingering doubt as to whether someone has 'enhanced' one of the can_be_'s
to give extra weight to one capability without renaming the function appropriately.

Using the new bools as return values does resolve the ambiguity and makes the
arithmetic usage an acceptable convention, though it requires that mental cast
from the true/false semantics of typical function names to 1/0 return values.
In some cases I suppose the better (I assume) speed of the arithmetic may be important.

>as opposed to
>
>    major = 0
>    if can_be_moved(o):
>        major += 1
>    if can_be_sized(o):
>        major += 1
>    if can_be_stacked(o):
>        major += 1
>    if can_be_coloured(o):
>        major += 1
>    if can_be_hidden(o):
>        major += 1
>
>I think the former is way clearer, and not because it uses less
>key-strokes, although brevity can aid clarity sometimes.
I think it is less cluttered, but it is more ambiguous (unless plainly
using the new bools), and that is not really clearer.

When you have your own usage conventions and have written the code
yourself, it is easy to take meanings for granted, and have no doubts,
but others may not have such faith in your adherence to a convention
unless you document it one way or another. One way is unambiguous code.
Another is a docstring comment like

"""Note: all is_xxx and can_be_xxx functions defined in this module
are guaranteed to return 1 if xxx is true and 0 if xxx is false."""

Personally, I prefer to read unambigous code, and not have to refer
to supporting documentation that I may have to double check anyway
by reading function source, if I'm doing a serious review, e.g., chasing
a bug. The new bools will help, since the arithmetic values can only be 0 or 1,
and that is unambigous where integer function return values are not.

Regards,
Bengt Richter


From thomas_judge at hotmail.com  Tue Apr 30 23:31:44 2002
From: thomas_judge at hotmail.com (Thomas)
Date: 30 Apr 2002 20:31:44 -0700
Subject: Python, Perl, &  BBSes
Message-ID: <35180ff3.0204301931.512b7f2e@posting.google.com>

There is a perl program in "Matt's Script Archive" called wwwboard.pl
which, as you might guess, creates a simple but powerful threaded
message board for websites. Has anyone out there either: translated
this to python (I need to implement something like this in python in a
short period of time and don't have time to learn perl); come up with
an automated perl-python translation tool; or found/written a good
python based bbs which does not rely on an external database?

Many, many thanks
Thomas


From gimbo at ftech.net  Wed Apr 17 05:21:15 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Wed, 17 Apr 2002 10:21:15 +0100
Subject: pre-PEP for optional 'pass'
In-Reply-To: <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1>
References: <mailman.1018972676.30534.python-list@python.org> <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1>
Message-ID: <20020417092115.GD2063@andy.tynant.ftech.net>

On Wed, Apr 17, 2002 at 08:33:44AM +0000, Duncan Booth wrote:

> Or you should edit it so it contains some useful content.
> Here is another go at the class above
> 
> class MyClass:
>    '''A completely useless class with no obvious purpose in life'''
>    def __init__(self, x, ,y, z):
>       '''Parameters x, y, and z are all dropped in the bit bucket.
>          Perhaps we are passed them for compatibility with something
>          else in which case we could say what...'''
>    def getTwiddle(self):
>     	 '''Despite the implications of its name, this method actually
>          returns None. It also has no side effects and should therefore
>          be refactored out.'''
>    def setTwiddle(self, twiddle):
>       '''Whatever a twiddle may be, this method actually drops the
>          argument and does nothing useful. Returns None. Another
>          refactoring candidate.'''
> 
> I would say that all of these comments could be automatically generated, 
> but if I look at the output of (say) pydoc, they add useful information as 
> I might otherwise be misled into thinking that the class and methods did 
> something.

We're veering slightly from the original reason for this point, which
was how to write stub functions/methods which will be implemented
later.  The above certainly doesn't indicate any intent to actually
implement the methods (in fact, on the contrary, it suggests they
could be removed). 

How about this?

class MyClass:
    """Class encapsulating a Twiddle.  Unfinished.  Document me.  XXX"""
    def __init__(self, x, ,y, z):
        """Not yet implemented. XXX"""     
    def getTwiddle(self):
        """Not yet implemented. XXX"""     
    def setTwiddle(self, twiddle):
        """Not yet implemented. XXX"""     

where you can grep for XXX to find unfinished business - very handy.

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From reply at in_newsgroup.tks  Fri Apr  5 11:35:58 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 5 Apr 2002 16:35:58 GMT
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk> <23891c90.0204050156.1840d913@posting.google.com> <a8k5v2$8kp$1@hfc.pacific.net.hk> <slrnaaravf.7u2.jon+usenet@snowy.squish.net>
Message-ID: <a8kjpe$kns$1@hfc.pacific.net.hk>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote in
news:slrnaaravf.7u2.jon+usenet at snowy.squish.net: 

> Try http://jonpy.sourceforge.net/

Congrats on your new project. My choices have been expanded even further!

Cheers,
Scrumpy :)


From marklists at mceahern.com  Fri Apr  5 10:23:13 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 5 Apr 2002 09:23:13 -0600
Subject: how to convert an integer to a string?
In-Reply-To: <a8ho5l$ubt$00$2@news.t-online.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEFFKJAA.marklists@mceahern.com>

s = str(10)

// m




From chris.gonnerman at newcenturycomputers.net  Wed Apr 24 21:41:22 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 24 Apr 2002 20:41:22 -0500
Subject: Is  fcntl.ioctl broken?
References: <PAENINBOOODFDEPOKPKCIEIACMAA.noah@noah.org>
Message-ID: <007b01c1ebfa$554ac780$0101010a@local>

----- Original Message ----- 
From: "Noah" <noah at noah.org>


> The documentation for ioctl in the fcntl module says:
> "ioctl(fd, op, arg) 
> This function is identical to the fcntl() function, 
> except that the operations are typically defined in 
> the library module IOCTL."
> 
> The only problem I can see is that there is no IOCTL module.
> I'm looking to get and set the winsize structure for the
> controlling terminal.

What version of Python?  What platform?  How did you install it?

You are leaving a lot out, I'm afraid.

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






From kseehof at neuralintegrator.com  Tue Apr  9 18:57:01 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 09 Apr 2002 15:57:01 -0700
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool type)
In-Reply-To: <roy-7658EF.08144409042002@news1.panix.com>
Message-ID: <EFEFIHAINIINNELLHFAIMEGMCAAA.kseehof@neuralintegrator.com>

Roy wrote:
>  tanzer at swing.co.at (Christian Tanzer) wrote:
> > There are several dimensions to typing, including:
> > - static vs. dynamic
> > - strong vs. weak
> 
> Putting it another way...
> 
> In Perl, if I write "$x = 1;", x can act as either an integer or 
> a string, 
> depending on how I use it.
> 
> In Python, if I write "x = 1", x can only act as an integer.  In that 
> sense, it's strongly typed.  On the other hand, it's also true that 
> sometimes later, I can say "x = '1'", and now x can only act as a 
> string.  
> At any particular point in time, its type is strongly defined, 
> but the type 
> can also (dynamically) change from time to time.
> 
> In Python, the variable (x) can change type, but the value itself (1) 
> cannot.  In Perl, the value itself can change type as coerced by the 
> context.

Python is mostly strongly typed.  Within the context of numbers, I
would classify python as weakly typed in the sense that x = 1 can act
as a float, int, longint, or bool, depending on context.  All numbers
of equal value are pretty much interchangable (this becomes more true
as we add more numeric unification features).  In c/c++ numbers are
more strongly typed, in that there are many ways to get error messages
for using the wrong type of number without explicit conversion.

The fact that an object's type doesn't actually change seems like an
implementation detail to me.  Numbers in python -feel- as if they are
weakly typed.

On the other hand, the ability to typecast pointers makes c/c++ rather
weakly typed compared to python on the whole.

- Ken Seehof





From sholden at holdenweb.com  Thu Apr 18 07:06:46 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 18 Apr 2002 07:06:46 -0400
Subject: Accessing mdb(access 2000) files on linux using python
References: <508cb8ae.0204170837.6f800a24@posting.google.com> <ZLhv8.1231$Ip3.1214@atlpnn01.usenetserver.com> <508cb8ae.0204180001.2f9abf92@posting.google.com>
Message-ID: <bPxv8.10890$eS3.57@atlpnn01.usenetserver.com>

"B" <martin_byrne2 at yahoo.co.uk> wrote in message
news:508cb8ae.0204180001.2f9abf92 at posting.google.com...
> "Steve Holden" <sholden at holdenweb.com> wrote in message
news:<ZLhv8.1231$Ip3.1214 at atlpnn01.usenetserver.com>...
> > "B" <martin_byrne2 at yahoo.co.uk> wrote in message
> > news:508cb8ae.0204170837.6f800a24 at posting.google.com...
> > > Hi,
> > > Just wondering if anyone knows how to manipulate mdb files stored on a
> > > linux webserver using python.Using win32com on windows is fine but on
> > > linux can i manipulate the database using pythons odbc connectivity or
> > > is there another way.
> > >
> > > Any suggestions, links or sample code would be of great help.
> > >
[Steve: No]
>
> Yeah I'd like to keep with access if at all possible. The reasons for
> this are as follows. The databases are being used to drive websites
> for certain customers of ours. The current setup is that each site has
> its own mdb database which the customer can change themselves, such as
> changing the products they are selling,vat rates,delivery charges,
> etc. This makes things easier for both parties(less work for us and
> more control for the customers).
> At the moment our webserver is IIS but we are hoping to migrate to
> apache in the near future and there we have the problem.
> I had guessed that accessing mdb databases on linux was a no go, just
> thought there might be an obscure way to do it.
> Looks like I'll have to write some sort of mysql frontend before we
> change to apache.
>
> Thanks for the help

A pleasure. One thing you *might* consider: use remote object technology
like Pyro to have the Linux web processes communicate with a database client
that runs on the Windows machine and uses either ODBC or ADO to access the
Jet database.

regards
 Steve






From martin at v.loewis.de  Fri Apr  5 02:21:12 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Apr 2002 09:21:12 +0200
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de> <d7fcf8b4.0204041558.1ecf5f27@posting.google.com>
Message-ID: <m33cya391z.fsf@mira.informatik.hu-berlin.de>

samschul at pacbell.net (Samuel Schulenburg) writes:

> I have used the folloing technique to get around the need to have a
> debug library for python. The following technique has been posted
> previously by Tim Peters.

I believe this approach has the flaw of linking the MSVC debug C
libraries to your module. This is undesirable, as it means that you
are simultaneously using the debug and the non-debug C
libraries. This, in turn, is undesirable since you get duplicate
copies of global data of the C libraries, which may interact in funny
ways.

Regards,
Martin



From mjackson at wrc.xerox.com  Wed Apr 17 21:47:50 2002
From: mjackson at wrc.xerox.com (Mark Jackson)
Date: 18 Apr 2002 01:47:50 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com> <a9gbhj$dhm$1@panix1.panix.com> <slrnabs3mm.1kfm.Gareth.McCaughan@g.local> <3CBE215D.F146ED46@alcyone.com>
Message-ID: <a9l8k6$f0t$1@news.wrc.xerox.com>

Erik Max Francis <max at alcyone.com> writes:
> Gareth McCaughan wrote:
> 
> > (These days there ought to be a computer scientist in the
> > mix too. Suggestions?)
> 
> "I'll just write a program to exhaustively find them all.  What's a
> prime number again?"

"My program halted with no output, so there must not be any."

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	The greatest masterpiece in literature is only a dictionary
	out of order.		- Jean Cocteau




From gcordova at hebmex.com  Fri Apr 26 10:19:09 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 26 Apr 2002 09:19:09 -0500
Subject: AS/400
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339B2FF1@mail.hebmex.com>

> 
> Another thing to look at - Per Gummedal's port of Python to 
> the iSeries at <http://home.no.net/pgummeda/>.
> 
> Cheers,
> Simon Brunning
> TriSystems Ltd.
> sbrunning at trisystems.co.uk
> 

Wheee, this would be most cool.

Using Pyro, you could instantiate Excel objects at the win32 side,
and just create a few methods to read data from certain position
or whatever you would need.

This would be quick, easiest (I believe), and also a most cool
example of python application.

Good luck!

-gustavo




From danb_83 at yahoo.com  Sun Apr  7 00:27:40 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 6 Apr 2002 21:27:40 -0800
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net> <MPG.17197f04a3f463e8989680@news.planet.nl> <7xvgb4to7h.fsf@ruckus.brouhaha.com>
Message-ID: <ad052e5c.0204062127.68335604@posting.google.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote in message news:<7xvgb4to7h.fsf at ruckus.brouhaha.com>...
> Johannes Gijsbers <taradino at softhome.net> writes:
> > That's not really a good idea, as math.sqrt() returns a float. I'd say:
> > 
> > >>> import math
> > >>> def isInt(number):
> > ...     return int(number) == number
> > ...
> > >>> isInt(math.sqrt(289))
>  1
> > >>> isInt(math.sqrt(315))
> > 0
> 
> isInt(math.sqrt(1e100)) doesn't work so well.

As it shouldn't.  In CPython 2.2, math.sqrt(1e100) is 7 629 769 841
091 887 003 294 964 970 946 560 higher than its actual value.  With an
absolute error this high, who cares about whether there's a fractional
part?


From mkelly2002 at adelphia.net  Mon Apr 15 20:02:57 2002
From: mkelly2002 at adelphia.net (Michael Kelly)
Date: Tue, 16 Apr 2002 00:02:57 GMT
Subject: Win32 Open Source or free .NET integration?
Message-ID: <RSJu8.3093$MM2.1226076@news1.news.adelphia.net>

I haven't used .NET myself yet but another programmer
using it raves about the back end integration among
the languages.  Also I noticed ActiveState is doing
a Visual Python that's supposed to integrate fully
into .NET.  I know there are some open source project
for .NET for Linux/unix but I was wondering if anyone
has heard of a free Win32 implementation in the works?

It would be way cool to plug Python into this system
from what I'm hearing(but $295 I don't have.) :)


From boris at cantal.lip6.fr  Thu Apr 11 06:44:19 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Thu, 11 Apr 2002 12:44:19 +0200
Subject: frameobject.c bug ? No call to PyFrame_LocalsToFast.
Message-ID: <pan.2002.04.11.12.44.16.889815.14362@cantal.lip6.fr>

I've been looking through the frameObject.c code and looked for what
happened, and surprised, there is no call to PyFrame_LocalsToFast when a
frame is created, so if f_locals isn't empty (which can be the case if
CO_NEWLOCALS is set to 0 the fast access values aren't initialised ).
Is there a reason for this ?

For the moment I've added a line:
> PyFrame_LocalsToFast(f,0);
just before line 261 : 
261> _PyObject_GC_TRACK(f);

This change allow to create codeobject without CO_NEWLOCALS that really
can be initialised with a local dictionnary. Before that as f_localsplus
wasn't initialised, CO_NEWLOCALS was a useless flag, this was as if it
was always set to 1.


Boris


From mlh at vier.idi.ntnu.no  Wed Apr 10 22:10:40 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 02:10:40 +0000 (UTC)
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk>
Message-ID: <slrnab9s50.6g8.mlh@vier.idi.ntnu.no>

In article <3cb3fbc0$0$78808$edfadb0f at dspool01.news.tele.dk>, Anders
J. Munch wrote:
>Pythons truth values are not "Something or Nothing".  The phrase seems
>to have caught on following Laura's post, but I believe it is
>misleading.

I don't.

>[] and {} are objects,

Yes, as are None and 0.

> they are objects with identity, and they are
>mutable objects.  They are empty, but they are by no means "nothing".

Would you call them false? Would you call 0 empty? It's an intuitive
way of explaining the concept, not a precise one. Nothing is nothing,
and as long as you try to represent it with a value (which must, by
necessity, be something, you will fail).

>Example:
>
>def show(foo, history=None):
>    """prints foo, optionally storing what was printed in a history list."""
>    print foo
>    if history:
>        history.append(foo)

Very odd function. "If there is a history" (i.e. the history list
contains something) "append something else"...

I don't think this is less clear if one considers the empty list as
"nothing". If your history is represented as a history, I'd say you
had no history if the list were empty.

The fact that you're using None as a placeholder means that you
shouldn't use a plain boolean check for the history here -- you should
use the standard idiom:

def show(foo, history=None):
    print foo
    if history is not None:
        history.append(foo)

or perhaps:

def show(foo, history=None):
    print foo
    try: history.append(foo)
    except TypeError: pass

(might be more efficient but will hide the error of passing a
non-appendable object as history.)

>This works as intended:
[snip]
>But this doesn't:
>
>   history = []
>   show("hello world", history) # strange..history doesn't change here?

IMO that should be obvious from your function definition. I don't see
what it is you feel this example proves. That seeing the empty list as
"a list-shaped nothing" (in Laura's words) would make people think
that only None qualifies as false (as opposed to an empty list)? I
really don't understand your interpretation here.

>It doesn't work, because show() tests for emptyness with "if
>show_history:".

Of course.

> What show() should have done is test for nothingness
>using "if show_history is None:".

No -- it should have tested whether history was the default value you
provided. This is not a test of nothingness -- it's not even a test of
None-ness (using the "ness" ending to indicate structural polymorphic
equivalence, which is normal in Python); it's a test of whether or not
the object is identical to None.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From marklists at mceahern.com  Wed Apr  3 12:12:58 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 11:12:58 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.33.0204031031440.16190-100000@server1.lfw.org>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEAAKJAA.marklists@mceahern.com>

[Ka-Ping Yee]
> The addition of another type really does change the set of preferred
> truth values from {0, 1} to {0, False, 1, True}.  The typical practicing
> Python programmer who wants to say "true" would have to decide -- for
> each instance -- whether to say "1" or "True".

Nope, the set of preferred truth values would be:

	True, False

You could, of course, continue to use 1, 0 if you like.  In the long run,
why would you?

> And the programmer cannot happily ignore the issue and stick to using
> 0 and 1 [...]

Of course she can.  In the long run, why would she?

> [...] because if comparisons start to return True and False, dealing
> with a mix of four truth values is unavoidable.  This PEP makes me uneasy
> because it causes me to see lots of extra casts in Python's future (not
> to mention confused beginners).

Can you give me just one example of these supposedly necessary 'extra casts'
that is not gratuitous and intentionally sloppy?  I'm too dense to come up
with one on my own.

Hmm, this PEP makes me delighted because I see lots of clean code and happy,
productive beginners.

But, to each his own.

Cheers,

// mark





From bokr at oz.net  Tue Apr 16 16:19:48 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 20:19:48 GMT
Subject: Python 2.2 unwise.exe
References: <mailman.1018983273.20776.python-list@python.org>
Message-ID: <a9i114$hq6$0@216.39.172.122>

On Tue, 16 Apr 2002 14:52:36 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[Robin Becker]
>> On the office machine well I installed Python 2.2.1 last Friday evening.
>> My unwise.exe is now gone as of Tuesday 10:00 am.
>>
>> I had a hard crash Monday morning.
>
>Did you have a previous version of 2.2 on your machine, and uninstall it
>before installing 2.2.1?
>
>On a clean Win2K Pro box, I'm finally able to provoke this (or at least one
>of its causes) reproducibly, via:
>
>1. Boot the box.
>
>2. Install 2.2.1.
>
>3. Uninstall 2.2.1.
>
What happens if you shut down and boot here?
(I'm wondering if there's a deferred delete waiting for a reboot
which doesn't get cleared by the re-install at 4.)

>4. Install 2.2.1 again.  Note that \Python22\UNWISE.EXE is there.
>
>5. Reboot the box (just logging off and on again isn't enough).
>
Deferred delete happens during reboot??

>6. Now \Python22\UNWISE.EXE no longer exists.
>
>
>

Regards,
Bengt Richter


From aleax at aleax.it  Tue Apr 16 07:23:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 11:23:15 GMT
Subject: Python Classes and IDLE
References: <3cbc06b0@news.comindico.com.au>
Message-ID: <DQTu8.25388$vF6.812770@news2.tin.it>

<posted & mailed>

Graeme Matthew wrote:

> Sorry to bug u all, but im at the point of giving up and going back to
> Perl (much to my disgust as python is so clean !). I have been trying to
> get classes going, even with two bloody books, one by Wesley Chun and its
> just a nighmare, sometimes I have to close down IDLE and now I find out
> that it must be reading the .pyc file thats why when I make changes they
> are not taken into account. I eitehr have to delete the .pyc file or close
> down IDLE, if this is the truth then this will take ages to do a project,
> anyone got some advice ? thanks a mil

If you modify a module, called e.g. floop.py, which you have already
imported in some running program (such as any interactive interpreter,
including the one in IDLE), then in order to re-load module floop (so
getting the newer version as you just modified it) you must call the
built-in function reload.

reload(floop)

will go to the disk and load floop.py up again -- as a side effect it
will recompile it into floop.pyc if necessary (checking modification
times of the source and compiled files).

If you find you must remove floop.pyc in order for floop.py to be
recompiled when you reload it, there must be something weird going
on with the file dates and times in your machine (but in that case,
closing and restarting IDLE would do nothing for you -- while if
your problem was related to not having called reload, then closing
and restarting IDLE would work, BUT deleting the .pyc file would
do nothing for you -- so if you're reporting everything accurately,
there must be something yet stranger going on).


Assuming you're not doing reload, rather than having weird file
time problems on your machine...:

To expand on the reload issue: when you
        import floop
Python goes to disk for module floop only ONCE in a given program's
run -- then, it caches the module into sys.modules['floop'], so that
any further imports become extremely fast.  This is an absolutely
crucial optimization, since modules are often imported from many
other modules, and they _don't_ usually change -- for those cases
where they do change (typically during interactive development),
the built-in function reload was introduced.


I'm somewhat perplexed about what you say about "going back to Perl".
What interactive environment are you used to, in Perl, which
implicitly reloaded modules as you changed them, without needing
you to ask for a reload explicitly?  I thought Perl development
didn't normally use interactive development environments, but
rather an edit/run/edit/... cycle.  If that's what you are most
comfortable with, you can use it with Python, too -- write your
code into whatever.py and run it from a Dos command box with
c:\python22\python whatever.py or the like.  Each new run of the
program will pick up all then-current versions of modules, with
no need to reload anything.  In other words, while Python LETS
you use an interactive environment (and many people find that
quite productive), it doesn't FORCE you to -- the usual edit/run
cycle at the Dos or Unix commandline is still available, too.


Alex



From TerryByrne1963 at yahoo.com  Sat Apr 27 13:01:21 2002
From: TerryByrne1963 at yahoo.com (Terry Byrne)
Date: 27 Apr 2002 10:01:21 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
Message-ID: <93d52e82.0204270901.3cf1ecd@posting.google.com>

All,

In a recent project I needed to read in a text file (a log file
generated by a utility program) and test each line of the file,
looking for certain error codes. I used file.readlines() to load the
log file into a list, then I used for aLine in aList: to test each
line.

One of my log files was ~60,000 lines long. At first I was using
re.search to find the relevant error codes. But it took ~5 hours to
process this log file -- too slow. Knowing that the string module is
much faster than re, I switched to string.find(), and only if that
found a certain error code would I run the re operation. (re is needed
to obtain more specific error information.) That cut my time down to
~1 hr and 45 minutes, a big improvement but still too slow.

Then I used the keyword "pass" whenever I ran an re since each line
can contain only one error; after running an re it's not necessary to
look for other error codes at all. Now that 60,000-line log file takes
less than a minute for Python to handle. Good enough.

I had no idea just how powerful and fast Python's re engine can be.
Having come from a Perl background, Python's way of doing regexes
annoyed me at first. But no more.

Terry


From max.wenzel at web.de  Tue Apr  2 10:37:14 2002
From: max.wenzel at web.de (Max)
Date: 2 Apr 2002 07:37:14 -0800
Subject: httpget
Message-ID: <2eea6430.0204020737.67749830@posting.google.com>

Hi,
I'm looking for a script like 'httpget' with a option to check a url
(exists or not).

Thanks,
Max


From paul at JUNKCATCHER.ciphergoth.org  Thu Apr 25 03:42:30 2002
From: paul at JUNKCATCHER.ciphergoth.org (Paul Crowley)
Date: 25 Apr 2002 08:42:30 +0100
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com>
Message-ID: <87ofg88brt.fsf@saltationism.subnet.hedonism.cluefactory.org.uk>

Paul Rubin <phr-n2002a at nightsong.com> writes:

> Currently I'm using
> 
>   K1 = H('auth1' + K)
>   K2 = H('auth2' + K)
>   MAC = H(K1 + H(K2 + ciphertext))

This looks good.  You can optimise this a little by padding K1 and K2
out to 64 bytes.  That means you can store them as internal SHA
states.

> Note that there's a trivial O(2**64) attack on my authentication since
> I'm truncating the MAC to 8 bytes.  So if a security difference
> between HMAC and what I'm doing needs more than O(2**64) work to
> exploit, it's not really useful to an attacker. 

Not so - it is far easier for the attacker to do 2^64 offline work
than it is to present 2^64 forged messages to the remote end.
-- 
  __  Paul Crowley
\/ o\ sig at paul.ciphergoth.org http://www.ciphergoth.org/
/\__/ BiCon 2002, discount before 10th May: http://www.2002.bicon.org.uk/


From fperez528 at yahoo.com  Wed Apr 24 17:13:15 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 24 Apr 2002 15:13:15 -0600
Subject: Quick: Format Specs in %
References: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
Message-ID: <aa7792$71a$1@peabody.colorado.edu>

Philip Swartzleonard wrote:

> 
> '%2f' % some_number pads the number with 2 blank spaces
> '%.2f' % some_number fixes the number to 2 decimal places
> 
> the problem:
> '%2.2f' % some_number only does the second of those two things. I need
> both. I tried looking up the % operator in the index of the language
> reference, but couldn't find it.
> 

For those quick questions, there's nothing like the PQR:

http://www.brunningonline.net/simon/python/PQR2.1.html

I actually keep a local copy permanently open while developing (it's a longish 
document to download every time). Combined with opera's incremental search in 
page (a la emacs), I typically find everything I need for such details in a 
few seconds.

I hope the author has time to update it to 2.2 sometime.

Cheers,

f.


From aleax at aleax.it  Mon Apr 29 03:01:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 07:01:59 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <Hd6z8.74979$vF6.2240539@news2.tin.it>

<posted & mailed>

Raymond Hettinger wrote:
        ...
>     There are separate motivations for each function.  One motivation in
>     common is that the number of total built-ins should be as small as
>     possible to limit the amount of "core language" one needs to know in
> order
>     to read code that is not module specific.

OK.  This may well be worth the substantial amount of work involved.

>     Pow() and divmod() are more related to math module functions than any
>     of the built-in functions.  In general, they are rarely used.  Because
>     the ** operator is available, the pow() function is rarely called
>     directly.

Built-in pow has a 3-argument form pow(a,b,c) which computes (a**b)%c far
more efficiently than separate operators can (for large integers).  There
are other obvious incompatibilities between math.pow and builtin pow:

>>> math.pow(2,3)
8.0
>>> pow(2,3)
8

Even though I use 3-arguments builtin pow reasonably often myself, that's
because I _am_ a combinatorics &c junkie.  I agree it's too rarely used to
be worth having it as a built-in -- importing it from a suitable module
would be fine.  But I double the suitable module is math, as this would
introduce further incompatibilities one way or another.  Module operator
might be most appropriate for both pow and divmod.


> Key Issue
> 
>     All of these functions have been in-place since the beginning, are
>     used
>     in mountains of code, and appear in every Python book.  The situation
>     would appear to be immovable.

Right.  Or at least, it should be moved slowly and with care.

 
>     I propose that a simple means be provided to re-enable those functions
> at
>     will when they are needed to run old code.  Add a command line option
>     -b and an equivalent environment variable, PYTHONBUILTINRESTORE which
>     would start-up Python by running:
> 
>         from functionals import map, filter, reduce
>         from math import pow, divmod
>         from deprecated import input

This would only make these 6 names available in __main__.  Do you propose
to inject them in every single module...?  Now THAT would be invasive
indeed.  No, they really need to go into __builtin__ if they're installed
for backwards compatibility.

I think a finer-grained mechanism might be preferable, and handled by
site.py / siteconfigure.py, perhaps, much like today are handles such
issues as sys.path normalization, and Unicode default-encoding.  The
former may be a more successful example than the latter, so maybe the
need might be best approached via an extension to the parsing and
handling of .PTH files.

Right now, .PTH files handle importing and path-setting needs of
installed packages.  Having them handle such packages' builtins-needs
as well seems a minor extension.  Should this slow start-up too much, maybe
some modest acceleration could be provided e.g. by sys.whatever helper
functions.  No objections to a PYTHONBUILTINS equivalent of PYTHONPATH,
in addition, for off-the-cuff needs.  Dunno 'bout the proposed flag: a
binary on/off switch seems too coarse-grained to be useful.


>     in perpetuity.  Like "from __future__" this mechanism can be used for
>     other deprecations, allowing the language to evolve without
>     accumulating clutter.

Exactly because of such "ongoing needs", the mechanism deserves careful
design.  No doubt more builtin candidates for deprecation can be proposed.

Take, for example, 'apply'.  What need does it serve, ever since the
forms *args and **kwds started applying to actual arguments (i.e., at
function-call time) as well as to formal arguments (i.e., at function
definition time)?  I.e., since 2.0 (or 1.6, I don't recall).  Indeed
that would have been my first candidate for built-in deprecation (also
for alphabetical reasons:-).

abs, buffer, chr, cmp, coerce, id, intern, oct, ord, are a few others
that might be worth thinking about in terms of frequency of need --
they're all needed some of the time, of course, but how often?  Often
enough to be warranted as built-ins?  I surely use, e.g., map, more
often than, e.g., cmp, coerce, or id.  Not to mention intern.  So why
should map go away and the others remain?  This is not meant to be a
rhetorical question, but to prompt active reflection/discussion/study.


Alex



From stephen at xemacs.org  Tue Apr 16 14:31:43 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 17 Apr 2002 03:31:43 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabm5cp.nkd.huaiyu@gauss.almadan.ibm.com> <87ofgjwzl5.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabommq.q79.huaiyu@gauss.almadan.ibm.com>
Message-ID: <87k7r7v6k0.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> writes:

    Huaiyu> The Apple Public Source License (APSL) version 1.2 is
    Huaiyu> listed as "Non-Free Software License" on
    Huaiyu> http://www.gnu.org/licenses/license-list.html, mainly
    Huaiyu> because of "Possibility of revocation at any time".

Ah, I missed that one.  But note it's not the revocation problem,
that's listed as fixed as of 1.2.

It's the "must publish deployed software" clause that's objectionable.

    Huaiyu> To me, one counterexample is enough, and one with a clear
    Huaiyu> explanation is very satisfying.

It's actually not a clear counterexample for the definition.  What I
mean is that I think that clause violates section 6 of the OSD.  It
clearly discriminates against internal business uses vis-a-vis
personal and R&D uses.

But that does make the covered portions of MacOS/X open-source but not
free.  I wonder what the OSI was thinking when they made that
decision.  The FSF's objection to the "deployment" clause was
well-known from earlier versions of the APSL.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From theller at python.net  Fri Apr  5 10:29:11 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 5 Apr 2002 17:29:11 +0200
Subject: CGI GET vs. POST
References: <a8k613$suseq$1@ID-59885.news.dfncis.de>
Message-ID: <a8kfse$teqtj$1@ID-59885.news.dfncis.de>

[replying to my own post]
"Thomas Heller" <theller at python.net> wrote in message news:a8k613$suseq$1 at ID-59885.news.dfncis.de...
> I'm trying to use the cgi module in a CGI script,
> but encountered a problem.
>
[snip}
>
> If I use method=POST in the CGI-script, the
> "keyword" is found as a field in FieldStorage,
> but NOT if I use method=GET.
> Is FieldStorage only usable with POST?
>

Problem solved, although I don't understand why.
Maybe someone can enlighten me?

The script was in a file pyhelp.cgi on the server,
and I called it with 'http://...../pyhelp' when it did
not work.
Calling it with 'http://..../pyhelp.cgi' made it work.

Thomas




From david.abrahams at rcn.com  Tue Apr  2 02:53:07 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Tue, 2 Apr 2002 02:53:07 -0500
Subject: [Newbie] embed c++ in Python
References: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at> <a8a62r$57u$1@peabody.colorado.edu> <3ca95375$0$3698$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <a8bo3v$1cs$1@bob.news.rcn.net>

Boost.Python uses just C++ and Python.
www.boost.org/libs/python/doc/index.html

"Matthias Stern" <mstern at aon.at> wrote in message
news:3ca95375$0$3698$6e365a64 at newsreader02.highway.telekom.at...
> > > I want to know how to write a function which receives some variables
as
> > > argument changes them and sends the results back to Python.
>
> Thank you for the links I got (http://www.scipy.org/site_content/weave
>  and http://www.acl.lanl.gov/siloon/ ). I've visited the sites and
> discovered a new problem: Both solutions require some additional
> stuff to be installed. Maybe it's much simpler to use them but it
> is quite hard to add some additional stuff like this to our project. It's
> a cross-platform project with many operating systems and we want as
> little additional programs as possible.
>
> So I have to ask again: is it possible to do this without any additional
> installations ("pure" Python 2.2) ? (L ike the "spam" example on the
Python
> Homepage) Has anyone a simple example code or link for this?
>
> Thx, Mac
>
>




From gashton at cmedltd.com  Wed Apr 24 05:42:15 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 24 Apr 2002 10:42:15 +0100
Subject: Python call tree tool?
In-Reply-To: <7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>
	<lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk> 
	<7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>
Message-ID: <1019641335.540.15.camel@tp6>

On Tue, 2002-04-23 at 18:37, Robin Becker wrote:

> trace.py seems to trace the execution of code. What I want is the call
> graph/inverted call graph a la CFlow so that I can determine which
> functions/classes are not called or referred to in some set of modules.

It sounds like you're after a code coverage tool. This might have some
ideas in it:

  http://www.garethrees.org/2001/12/04/python-coverage/

-- 
Graham Ashton





From gusraan at terra.com.br  Tue Apr 23 11:38:44 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: Tue, 23 Apr 2002 12:38:44 -0300
Subject: Developing and ircd with python
References: <aa278j$8c8$1@nsnmrro2-gest.nuria.telefonica-data.net> <3CC4A648.3070100@jerf.org> <Xns91F8E339D9A04RASXnewsDFE1@130.133.1.4>
Message-ID: <aa3v5n$8jg$1@nsnmrro2-gest.nuria.telefonica-data.net>

Do you know any place where i can find info about that?

"Philip Swartzleonard" <starx at pacbell.net> wrote in message
news:Xns91F8E339D9A04RASXnewsDFE1 at 130.133.1.4...
> Jeremy Bowers || Mon 22 Apr 2002 05:07:52p:
>
> > Raphael Ribeiro wrote:
> >> Is it possible?
> >
> > Yes.
> >
>
> There's something entirely perfect about this conversation...
>
> --
> Philip Sw "Starweaver" [rasx] :: www.rubydragon.com




From pyth at devel.trillke.net  Sun Apr 28 16:57:57 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 28 Apr 2002 22:57:57 +0200
Subject: Update locals()
In-Reply-To: <200204281716.TAA10331@prim.merlinux.de>
References: <20020427195725.GA16217@devel.trillke> <0204280830290C.16995@arthur> <20020428115221.GF16217@devel.trillke> <200204281716.TAA10331@prim.merlinux.de>
Message-ID: <20020428205757.GH16217@devel.trillke>

On Sun, Apr 28, 2002 at 07:16:45PM +0200, Alex Martelli wrote:
> On Sunday 28 April 2002 01:52 pm, holger krekel wrote:
> 	...
> > > Please give ONE example which, in your opinion, is best handled
> > > with 'exec' without explicit dictionaries rather than in any other way.
> >
> > i try to make it short and answer questions (if any of you gurus
> > ever have any :-) as needed. Obviously you can always pass explicit
> > dictionaries to exec (globals, locals). That can't be the point.
> 
> Nope.  locals() and globals() are allowed by the language definition to be 
> mappings rather than fully general dictionaries, and in particular any
> alteration performed on locals() or globals() is allowed (this could change
> at any time, even in a micro-version or a bug-fix one -- the docs have been 
> saying so for a LONG time) to [a] alter the namespace, [b] silently
> ignore the alteration attempt, [c] raise an appropriate exception.

Are you sure that this holds?

the documentation from python.org says
"""
  globals() 
  Return a dictionary representing the current global symbol table. This
  is always the dictionary of the current module (inside a function or
  method, this is the module where it is defined, not the module from
  which it is called). 
"""

"""
  locals() 
  Return a dictionary representing the current local symbol table.
  Warning: The contents of this dictionary should not be modified; changes
  may not affect the values of local variables used by the interpreter.
"""

So globals() has no such warning (which you stated) and even the warning
in locals() doesn't sound like python is going to *systematically*
prohibit such alterations. It indicates to me that there are cases (which 
we also had yesterday) where the alteration may not affect the calling 
scope. But anyway we certainly agree that "exec" without optional arguments 
*does* execute in current scopes (no matter what behaviour the objects
returned by locals()/globals() have). We are only arguing whether 
there is some good use for execing in current scopes.

> In the current implementation, locals() alterations when the namespace is in 
> a function follow rule [b] (they're silently ignored), other locals() and 
> globals() alterations rule [a], but no program that relies on either behavior 
> is correct Python (might be nice for PyChecker to warn, but it's probably too 
> hard to track).  I _think_ the best solution might be to have them as 
> mappings that give a warning when altered, on the way to making them 
> read-only per rule [c] in a release or two. 

Hmmm. certainly a possible update-path. 
I really like about python that you can refer to classes, 
object's attributes/methods, the global/local bindings 
as objects and work with it. This is (one of) the goal
of e.g. "reflective middleware" or generally software which can
introspect and modify its behaviour. It's a very powerful idiom
that many other current languages are missing. Nobody says that you don't
have to care about any side effects this may and will have.

> Basically you want to add to class filename a bunch of methods, each
> delegating as you show to the homonym callable attribute of global object 
> filters.  This is most easily done after the filename class object is
> complete and available - i.e. right after the class statement, or in a
> metaclass.  The "right after" approach is probably easier (using Python
> 2.2 for simplicity, doesn't make much difference anyway):
> 
> def addMethods(class_, filters, statefulList):
>     for name in statefulList:
>         delegatee = getattr(filters, name)
>         def method(self, *a, **k): return delegatee(*a, **k)(self)
>         method.__doc__ = getattr(filters, name, undoc).__doc__.strip() + '\n'
>         setattr(class_, name, method)

Thanks for this transformation. But of course a 'but' follows :-)
As soon as 

- you want to adapt the definition to a filter's method signature 

- have templated or parametrized code in the function body

- want to see what actually gets defined 
  (with my version i can easily write or print the
   actual function's definition and look at a real definition)

it might get difficult or impossible to do it your way. 
Writing a reusable helper function i could say something like

    exec make_adapters(filters, statefulList, '''
    def %name (self, %args):
        """%__doc__"""
        return %sourcescope.%name (%args) (self)
    ''')

or even better make the templated adapter definition a defaulted 
argument. This provides more flexibility for the *caller* of the helper
function without ever having to modify the helper function.

Also i use the template 'exec in current scope' method to allow 

resultfilters = "isfile and nolink and fnmatch('data*')"

to be compiled at run time into a function and then be
applied to - say - 15000 files. This technique is similar 
to what the regular expressions module does with re.compile. 
This may be a better example than the original one?!
At least i don't see how you can apply your techniques 
to it without using slow 'evals' or a slow interpretation
layer.

> You may need to tweak this (I'm not sure why the filter.name access is
> unconditional when looking for the delegatee, but falls back to undoc when
> looking for the __doc__, for example)

None.strip() in 

...__dict__.get(name).__doc__.strip() 

would fail at run time if a filters's doc has not been defined.
Same with your code.

> If having to explicitly call
> 
> addMethods(filename, filters, filters._stateful_all)
> 
> right after the 'class filename:' statement body's end nags you, you can
> use a custom metaclass for the same purpose...

That's very interesting and i will investigate this further.
But my above points can certainly be raised here, too.

> > - i am a maniac at avoiding redundancies.
> 
> ...and yet the infrastructure embodied in class filename as you coded
> it is not reusable save by copy-and-paste, the worst kind of redundancy.
> addMethods and the MC_methods_adder metaclass avoid redundancy
> by placing such metacoding-infrastructure in ONE place.

OK. 

1. My quest against redundancy is certainly not over :-)

2. It's obvious that i could turn the given example into
   something reusable as well (and it would honour the actual
   arguments via inspect.getargspec among other nicities).
   I just didn't want to blow the example to inifity
   and provide some direct context.

I appreciate your explanations and code very much. 

The only better thing i can imagine is if you tried a tiny bit 
harder to see some *valid* points rather than only pin down invalid points.
Otherwise you give me the feeling that i am completly on the wrong path.
(or is this intended?!)

I just try to communicate a use for "exec" where it is preferrable
to not use explicit dictionaries. 

regards,

    holger




From jwickard**nospam** at litriusgroup.com  Fri Apr  5 09:02:34 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Fri, 05 Apr 2002 14:02:34 GMT
Subject: file rewind?
Message-ID: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>

I have this code:

def countLines(file):
    numLines = 0
    for line in file.readlines():
        print line
        numLines = numLines + 1

    print numLines

def countChars(file):
    numChars = 0
    for line in file.readlines():
        print line
        for char in line:
            print char
            numChars += 1
    print numChars

def test(name):
    file = open(name, "r")
    countLines(file)
    countChars(file)

if __name__ == "__main__":
    test("C:\\test.txt")

when I run the module from the comman line by itself, it counts the lines
alright, but does not count the characters.  do I need to "rewind" the file
reference after the first function call to countLines()?




From starx at pacbell.net  Wed Apr 24 14:23:07 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 18:23:07 GMT
Subject: Window problem in windows!
References: <mailman.1019653396.7431.python-list@python.org> <Ozzx8.64548$HH5.4797@rwcrnsc51.ops.asp.att.net>
Message-ID: <Xns91FA740A8E146RASXnewsDFE1@130.133.1.4>

Emile van Sebille || Wed 24 Apr 2002 07:47:10a: 

> Stian Jenssen 
>> ---code--- 
>>
>>     command = "pause \r\n" command += "pause" print
>>     os.system(command) 
>>
>> ----code-- 
>>
>>
>> It pops up 2 window instead of 1, in a windows enviroment. The code 
>> works greate on *nix box, what seems to be the problem. 
> 
> 
> Well, then it averages out properly.  Here it opens no window! ;-) 
> 
> Which really just means to say that we'll need more info on how it is 
> you're trying to do this, or a complete example the exhibits the 
> problem. 

I'm fairly sure he means that double-clicking the script in windows causes 
two command-promptish output windows to come up. Sounds wierd.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com 


From brian at sweetapp.com  Sat Apr  6 13:14:52 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 6 Apr 2002 10:14:52 -0800
Subject: ANN: SilverCity 0.6.0
Message-ID: <002d01c1dd96$f23d6fb0$445d4540@Dell2>

SilverCity is a lexing package, based on Scintilla, for over 20
programming and markup languages. 
 
The SilverCity API is similar to that of the Python tokenize module but:
o it works with languages other than Python
o it works with syntactically invalid code
o it includes an (optional) event-driven API (think SAX)
o it includes modules to generate syntax-styled HTML documents
o it includes a few useful scripts (see PySilverCity/Scripts in the
source distribution)
 
For more information, see:
http://silvercity.sourceforge.net <http://silvercity.sourceforge.net/> 
 
The documentation is available at:
http://silvercity.sourceforge.net/docs.html
 
Cheers,
Brian
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020406/18c4d9c4/attachment.html>

From loewis at informatik.hu-berlin.de  Thu Apr 18 11:12:07 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 17:12:07 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019136395.28275.python-list@python.org>
Message-ID: <j47kn56ny0.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> It is worse
> 
> Traceback (most recent call last):
>   File "./Lib/test/test_socket.py", line 12, in ?
>     import socket
>   File "/home/hgs/Python-2.2.1/Lib/socket.py", line 41, in ?
>     from _socket import *
> ImportError: No module named _socket

Here is the procedure to analyse this in detail:

Theory 1: during building of _socket with setup.py, setup notices that
building of _socket went wrong, and deletes it.

  If that is the case (analyse setup.py output): Invoke the commands
  to build _socket, and report whether this succeeds. If it does,
  start ./python, and 'import socket'; report any errors that this
  gives.

To work around this, we need

Theory 2: building the socket module fails because setup.py detects
presence of OpenSSL, yet the OpenSSL versions is to old for the socket
module.

  If that is the case, you'll need to edit Modules/Setup, and build
  the socket module explicitly without SSL support (either shared or
  static).

Regards,
Martin


From sholden at holdenweb.com  Fri Apr 12 08:28:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 12 Apr 2002 08:28:38 -0400
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca>
Message-ID: <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>

"Brad Bollenbach" <bbollenbach at shaw.ca> wrote ...
> Peter Olsen wrote:
> > I think that, for my purposes, licensing is not worth my effort or
> > my users' time.  But if you want to license your Open Source software,
> > be my guest.
>
> This doesn't actually make sense. To call software "Open Source" is to
> acknowledge it as being distributed under a license that is defined as
> compatible with what Richard Stallman calls "Free Software". So you
> don't have the "choice" of licensing Open Source software. It got to be
> called "Open Source" *because* of the license you've already chosen for
> it.
>
So the code I have put in the public domain isn't open source? That's
interesting.

> And, FWIW, thinking that you're doing your users a favour by monitoring
> who's using which version of what, and in what kind of environment
> (commercial vs. non-commercial) is, at best, dependant on the types of
> users, and, at worst, and absolutely disasterous idea. If they're
> "end-user" types, you might legimitately be providing some capabilities
> they want, but adopting this scheme for Open Source Hacker types will
> just reduce your user-base by an order of magnitude or more.
>
I think you might have been taken in by a date-dependent (think "April 1"
here) post that wasn't intended to be taken that seriously.

regards
 Steve






From aahz at pythoncraft.com  Sat Apr  6 16:11:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 16:11:04 -0500
Subject: New classes in 2.2
References: <un0wglgqf.fsf@online.no>
Message-ID: <a8no98$leu$1@panix1.panix.com>

In article <un0wglgqf.fsf at online.no>,
Syver Enstad  <syver-en+usenet at online.no> wrote:
>
>I have to admit that I am excited about the revisions to pythons
>object model. I hope that the current state of affairs where there are
>two kinds of classes will eventually fade away, as it seems quite ugly
>at the time being (wouldn't want to be a new to Python just now). It
>would be sad if Python we're to develop along the lines as C++, where
>there seems to be no obvious way to do even the simplest of things (in
>my view, even though I think it's a fascinating language, my main
>language in fact until I discovered Python).

Python 3.0 will get rid of "classic" classes entirely.  Right now, we're
in a transitional phase where we're experimenting with the precise
semantics for new-style classes.  Guido hasn't said, but I'll bet that
2.4 or 2.5 will be the last of the 2.x line.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From Oschler at earthlink.net  Mon Apr  1 23:33:52 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 23:33:52 -0500
Subject: IDLE for SUSE Linux?
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com> <3CA8D724.336A22BA@3captus.com>
Message-ID: <5waq8.5401$ku5.476012@e3500-atl2.usenetserver.com>

Bernard,

Then I have something strange.  There is no /Tools directory under
/usr/share/doc/packages/python.  I tried doing a search on Python in Yast2
and installing a bunch more of the Python related packates (and online
updated them), but still no luck.  A full disk search for Idle.Py turned up
empty.

thx

Bernard Yue <bernie at 3captus.com> wrote in message
news:3CA8D724.336A22BA at 3captus.com...
>
> SuSE Linux 7.3 Pro comes with Python 2.1.1, so IDLE comes with it.  IDLE
> is located at /usr/share/doc/packages/python/Tools/idle/idle.py.
>






From sam at sambayer.com  Mon Apr 22 23:12:07 2002
From: sam at sambayer.com (Sam Bayer)
Date: Tue, 23 Apr 2002 03:12:07 GMT
Subject: subprocess monitoring in Windows?
Message-ID: <3CC4D1B7.7070303@sambayer.com>

Hi all -

Lately, I've been forced to take a hard look at getting one of my 
Python/Tk apps running under Windows. This app monitors and controls 
subprocesses, essentially. I've been tripping over a lot of things, 
starting with the rather considerable differences in the Windows 
programming model, but I think I'm going to be able to handle the 
Windows equivalents of nonblocking reads and writes with pipes to 
subprocesses. However, there are a couple of odd omissions which I'm 
curious about, and I'm wondering whether they're just things that no one 
ever had the time to implement, or whether there's some deeper reason 
they're not around.

(1) The Unix popen2 module provides an object which gives the programmer 
simultaneous access to stdin/stdout for the subprocess as well as the 
process ID of the subprocess. For some reason, Mark Hammond's Windows 
extensions for Python don't do this. The C code which implements the 
popen2/3/4 functions hangs onto a process ID internally, but doesn't 
return it to the user; the only way to duplicate the Unix functionality 
is to translate the C code into Python (fortunately, someone posted 
something akin to such a translation a couple years ago). Has no one 
else ever found this to be useful?

(2) Apparently, Tcl/Tk replaced createfilehandler on Windows with the 
more general fileevent/channel abstraction starting with 8.0. However, 
this abstraction hasn't made it into Tkinter, and I can't find any hints 
on the newsgroup about how to get around this. A couple people reference 
the asyncore module, but I'm dealing with pipes, not sockets. What I'll 
probably do is write my own formatters to construct the appropriate 
fileevent calls, but I'm wondering what people have been doing in the 
meantime. I know I could use threads, but having had extensive annoying 
experience with mutexes in C, I'm not looking forward to having to 
duplicate the experience in Python if I don't have to.

Don't get me wrong; I love Python with all my heart, and I know that 
much of what makes it into the language is a generous gift of people's 
time. But I'm kind of puzzled that I'm finding it so hard to find 
references to workarounds for these omissions, and I'm wondering whether 
I'm missing something.

Thanks in advance -
Sam Bayer
The MITRE Corporation
sam at mitre.org



From bmason at rand.org  Tue Apr 16 16:05:44 2002
From: bmason at rand.org (Bryce C. Mason)
Date: Tue, 16 Apr 2002 13:05:44 -0700
Subject: Counting the number of elements in a list - help.
References: <a9hu8e$2fg$1@lumberjack.rand.org> <m3n0w3fmhf.fsf@localhost.localdomain>
Message-ID: <a9i05f$2pf$1@lumberjack.rand.org>

Sure was!  Thanks


"Steingrim Dovland" <steingrd at ifi.uio.no> wrote in message
news:m3n0w3fmhf.fsf at localhost.localdomain...
> "Bryce C. Mason" <bmason at rand.org> writes:
>
> > 1) a function which returns the number of elements in a list
> > 2) a function which takes a list and an index and returns a zero when
that
> > list index is out of range.
>
> I think you're looking for len(), a built in function.
>
> >>> x = [1,2,3]
> >>> len(x)
> 3
>
> --
> Here the sea spray give
> I was with her
> We're under the ship so get me over




From starx at pacbell.net  Sat Apr  6 18:15:27 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 6 Apr 2002 23:15:27 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com> <Xns91E6D45CAAFEcliechtigmxnet@62.2.16.82>
Message-ID: <Xns91E89B600B746RASXnewsDFE1@130.133.1.4>

Chris Liechti || Thu 04 Apr 2002 11:51:30a:

> netzapper at magicstar.net (A. Jones) wrote in
> news:3cac943f.132176618 at mammoth.usenet-access.com: 
> 
>> On Thu, 04 Apr 2002 11:56:24 GMT, Alex Martelli <aleax at aleax.it>
>> wrote:
>> 
>> Furthermore, you can wrap it inside a class.
>> 
>> class test:
>>      def wrongHeaded(self, name, value):
>>          exec "self."+name+'='+repr(value)
>>          print vars()
>> 
>> this produces no exceptions, and doesn't cause any problems, since
>> vars is never defined... only self.vars is.
> 
> until someone has the idea to pass some fancier names like
> t.wrongHeaded('a; del None', 57)
> 
> and then have fun watching your entire program fail...

Hm, i think you want something like
t.wrongHeaded('a=5; del None;b', 57)

which will evaluate to 
self.a=5; del None; b=57

instead of
self.a; del None=57

which is two syntax errors in one (no a defined, and i think 'del none=
57' is invalid?)



-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From mlh at vier.idi.ntnu.no  Tue Apr  2 15:56:46 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 2 Apr 2002 20:56:46 +0000 (UTC)
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl> <1d6q8.118888$Ou.77410@atlpnn01.usenetserver.com>  <3ca940d3$0$32314$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <slrnaak6oe.s4j.mlh@vier.idi.ntnu.no>

In article <3ca940d3$0$32314$e4fe514c at dreader4.news.xs4all.nl>,
Boudewijn Rempt wrote:
>Steve Holden wrote:
>
>> Congratulations! I remember your original posts asking for good
>> examples and choices for the graphics subnsystem. You must feel a
>> weight has been lifted ...
>> 
>
>I do. It has certainly been an interesting experience. I was wrapping
>up the last of the text last year, when there was this thread about
>writing books on comp.lang.python. It felt very to the point.
>
>Anyway, I'm quite satisfied with the results :-).

I'm getting a "Not found" error for a javascript file:

  http://www.opendocspublishing.com/pyqt/validate.js

(A bit irritating when MSIE users don't notice missing js and css
files, because MSIE ignores them... I can't even access the page in
Netscape :I)

Anyway, congrats! :)

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

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From tim.one at comcast.net  Fri Apr 26 15:35:53 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 15:35:53 -0400
Subject: File sizes from MSWin...
In-Reply-To: <uvgaexopp.fsf@ctwd0143.fitlinxx.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEEIPBAA.tim.one@comcast.net>

[David Bolen]
> ...
> I expect you're seeing truncation due to the use of 32-bit values for
> the size.  Until Python 2.2, the Windows version didn't include large
> file support (not sure if you could manually build it that way or not,
> but it didn't come pre-built with it).

You could not; enabling large file support on Windows required structural
changes to the Python implementation.  The limitations before 2.2 were
inherited from that Microsoft's implementations of "regular old portable C
library functions", like stat(), were restricted to 32-bit results.  So
getting sensible results for large files on Windows required calling a
variety of Windows-specific functions under the covers.





From ole.fostad at netcom.no  Sat Apr 13 13:00:45 2002
From: ole.fostad at netcom.no (Ole Fostad)
Date: Sat, 13 Apr 2002 17:00:45 GMT
Subject: Post-installation with distutils windows installer
Message-ID: <1103_1018717245@news.netcom.no>

Hi!

I just started using Python and wondered if it is possible to run a post installation script 
when using the distutils windows installer. 
I cannot find anything in the manuals.

Ole Fostad






From mwh at python.net  Wed Apr 17 08:06:29 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 12:06:29 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com> <3CBC5B6D.70427378@noaa.gov> <gpsobu0952s65n93clkqadtq27cj1j4qul@4ax.com>
Message-ID: <lkg01useye.fsf@pc150.maths.bris.ac.uk>

Matt <mck at mpinet.net> writes:

> On Tue, 16 Apr 2002 10:12:14 -0700, Chris Barker <Chris.Barker at noaa.gov> wrote:
> 
> >Imagine if Microsoft decided to drop VBasic
> >and start selling MSPython!
> 
> Oh boy, that is a whole different can of worms (I can see the thread
> now...) ;)

I think that one's already happened (see google if you really care).

Cheers,
M.

-- 
  Sufficiently advanced political correctness is indistinguishable
  from irony.                                           -- Erik Naggum


From mats at laplaza.org  Wed Apr  3 14:42:23 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Wed, 03 Apr 2002 19:42:23 GMT
Subject: PEP 284, Integer for-loops
References: <eppstein-BB3186.08413106032002@news.service.uci.edu> <slrna8d4bj.utc.grey@teleute.dmiyu.org> <eppstein-07CB42.14230206032002@news.service.uci.edu> <3c87d103.26130283@news.laplaza.org> <eppstein-B61874.13474907032002@news.service.uci.edu>
Message-ID: <3c87f022.34099342@news.laplaza.org>

On Thu, 07 Mar 2002 13:47:49 -0800, David Eppstein
<eppstein at ics.uci.edu> wrote:

:In article <3c87d103.26130283 at news.laplaza.org>,
: mats at laplaza.org (Mats Wichmann) wrote:
:
:> If you'e going on to teach Python, you'd tell 'em those can of course
:> be combined.
:
:I don't want to teach Python.  I want to teach other stuff.  

Right.  That's why I separated that out:  the intesting part is
looping over multiple values.  range() happens to be a way to generate
those values, but not the only way by any means, and that you don't
like the syntax of range should have no effect on the teaching
objective of showing how the concept of loops works....
Mats Wichmann



From taw at redhat.com  Thu Apr 11 19:38:02 2002
From: taw at redhat.com (Todd Warner)
Date: Thu, 11 Apr 2002 19:38:02 -0400 (EDT)
Subject: Communication between a Python process and a Python cgi script
In-Reply-To: <3CB5E224.4E3DC5BA@olen.to>
Message-ID: <Pine.LNX.4.44.0204111601010.8053-100000@devserv.devel.redhat.com>

On Thu, 11 Apr 2002, Joonas Paalasmaa wrote:
> I have a Python program started by root that does some magic stuff
> every 10 minutes. So the program in an infinite loop of sleep, do stuff,
> sleep, do stuff... How can I make a cgi-script to end the process's
> sleep phase and make it to "do stuff" immediately. I tried signals, but
> it is propably too risky because the process is run as root. Any
> suggestions?

One way:
Use a cron job to execute it every ten minutes would be very simple.
But it can't be nicely ended via a CGI script that I can think of off
the top of my head. And it won't perform an action immediately.

Another:
Alternatively, and probably ideally. Make it a service. Start playing
with os.fork() to toss it to the background, and have it listen through
a socket for an exit or action string. Make a cgi script that sends that
enabling string through the socket. Look in /etc/init.d/ for examples of
how to make something a service.

Another option:
cgi script sends a kill and SIGALRM. *shrug* Depends on what the risk is.

-- 
____________
 /odd "traceback" Warner                        <taw@{redhat,pobox}.com>
Tank Commander - NC Army National Guard   Bit Twiddler - Red Hat Network
---------------------gpg info in the message headers--------------------
"Eat dessert now; life is uncertain" -- unknown








From peter at engcorp.com  Thu Apr 11 23:09:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 11 Apr 2002 23:09:49 -0400
Subject: 4Suite for Python 2.2
References: <zxft8.37$Mh5.1049@nreader1.kpnqwest.net>
Message-ID: <3CB64FFD.6D0D6AD7@engcorp.com>

Paulo Pinto wrote:
> 
> Hello,
> 
> does anyone know when 4Suite for Python 2.2
> would be available ?
> 
> I can only find support up to version 2.1 :(

Doesn't it run under 2.2?  (I haven't tried and don't know --
this is a serious question.)

I thought 4Suite was simply a very large collection of XML
processing tools in native Python (which I would have
expected would still run under 2.2).  Or are there binary
libraries which are keyed to python21.dll or something?

-Peter


From j.l.gomez-dans at sheffield.ac.uk  Sat Apr  6 14:43:03 2002
From: j.l.gomez-dans at sheffield.ac.uk (=?iso-8859-1?q?Jos=E9?= Luis)
Date: 06 Apr 2002 20:43:03 +0100
Subject: Reading large multidimensional arrays off file
Message-ID: <87k7rksjeg.fsf@sheffield.ac.uk>

        Hi all,
        Sometime ago, I posted a query about reading data from a
binary file, with its header and all that. Chris Barker very kindly
answered my questions, and I was a happy person. I now have a new set
of data files, and I was trying to read them in with Python. the files
are simple: there are N "lines", each with M "samples", and each
sample represents 10 signed bytes. I need to access each of the
unsigned bytes individually, and at the moment, I am doing this
through a loop, as follows (I am using the SciPy library):



from scipy import io
from scipy import plt
from scipy import *
import gui_thread

#Define the filename
fname = "/cdrom/ddata.dat"

#Get a file pointer to the file
fp = io.fopen(fname,'r',format='b')


#Skip header (30720 bytes in this case)
fp.fread(30240,'b','f')
#The image array
img=array(zeros((1280,1024),Complex)) 

for line in range(1279):
    for sample in range(1024):
        matrix=fp.fread(10,'1','f')
        #Random processing of bytes :-)
        M = StokesMatrix(matrix,scale)
        C = Stokes2Covariance(M)
        img[line,sample] = abs(C[0,0])
        
#Save the image to a matlab file to play around with later...
io.savemat('flev116-1a-c',{'img':img})


        The previous process takes ages to read the file in (of the
order of an hour for one 13 Mb file), so I was wondering why I don't
just read the whole array in, after all, I know that it is NxMx10
bytes, and I know the order. My question is fairly basic: How do I
specify the order of the data? I would like to read it in the order
shown above, into an 3-dimensional array size (10,N,M), but I am not
sure what sizes come first or second in Python.

        Many thanks in advance,
Jos?
-- 
Jos? L G?mez Dans			PhD student
Tel: +44 114 222 5582			Radar & Communications Group
FAX; +44 870 132 2990			Department of Electronic Engineering
					University of Sheffield UK


From alanmk at hotmail.com  Sat Apr  6 08:09:51 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 6 Apr 2002 05:09:51 -0800
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com> <3cadd841.225619243@news.laplaza.org>
Message-ID: <f72dac0d.0204060509.22c7833e@posting.google.com>

mats at laplaza.org (Mats Wichmann) wrote:

> Alan Kennedy> 
> :The golf competition is a neat idea. Python hasn't completely
> :suppressed the hacker in me :-)

> Mats Wichmann
> 
> Ugh, it's horrid, and seems very perl-ish.  Do something in the
> smallest amount of space, readability/maintainability be damned.

Point taken. But I don't think the Golf competition is about the
aesthetics of readability and maintainability, it's about the
aesthetics of minimisation.

And the golf competition problems are hardly everyday problems that
need maintainable solutions.

> We need something that emphasizes the beauty, simplicity, clarity of
> the solution (some Japanese term, perhaps?)

I find this idea very attractive as well. But how could such a
competition be judged objectively? Beauty is in the eye of the
beholder (as are simplicity and clarity).

Alan Kennedy.


From clgonsal at alumni.uwaterloo.invalid  Sun Apr 28 20:58:15 2002
From: clgonsal at alumni.uwaterloo.invalid (C. Laurence Gonsalves)
Date: 28 Apr 2002 17:58:15 -0700
Subject: expat and entity references
Message-ID: <slrnacp5lb.b5i.clgonsal@keeshah.penguinpowered.com>

I'm working on an Python application that uses xml.parsers.expat to
parse XML files.  These XML files are generally hand-written. I'd like
to be able support more than just amp, lt, gt, apos and quot.

What's the best way to do this without requiring that every XML file
contain a huge header declaring all of the entities I want to support? 

-- 
  C. Laurence Gonsalves
  clgonsal at alumni.uwaterloo.invalid
  -1 s/invalid$/ca


From sholden at holdenweb.com  Mon Apr 15 13:03:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 13:03:50 -0400
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <LHAu8.11094$Ji7.10494@atlpnn01.usenetserver.com> <3cbaebd1@news.mhogaming.com>
Message-ID: <%LDu8.13590$Ji7.6884@atlpnn01.usenetserver.com>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote in message
news:3cbaebd1 at news.mhogaming.com...
> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:LHAu8.11094$Ji7.10494 at atlpnn01.usenetserver.com...
>
> | "Bengt Richter" <bokr at oz.net> wrote ...
> |
> | Since in Boolean terms a xor b is equivalent to ((a and not b) or (b and
not
> | a)), maybe a xor b should return whichever of a or b is not false.
> |
> | >>> a = 42
> | >>> b = 0
> | >>> a ^ b
> | 42
> | >>> b ^ a
> | 42
> | >>>
> |
> | Since that's exactly what it appears to do, stop fiddling with what
already
> | works !
>
> NO it DOESN'T!! ::
>
I know, but I suppose I'll have to come clean: I chose the data to support
my argument. As your example shows clearly, the ^ operator is bitwise rather
than what the documentation currently calls a Boolean operator (not, and).

> >>> a = "123"
> >>> b = 0
> >>> a ^ b
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: unsupported operand type(s) for ^: 'str' and 'int'
> >>>
>
> 'xor' should return the string '123' in this case-- not an error.  In
other
> words, '^' deals with BITS (as it now does), and 'xor' deals with objects.
>
> I need (in my work) the ability to diddle bits-- these operators need to
"do the
> good thing" when operating on ints (and longs).  I don't want these
operators
> overloaded to recognize other objects-- that is the purpose of 'not',
'or',
> 'and', (and would be operator 'xor').
>
> The operator '!=' performs a similar operation, but always returns a
boolean.
> 'xor' should return the non-false object if the other operand is false,
and
> false if both operands are either true or both operands are false.
>
> The 3 different behaviors of '^', '!=', and 'xor' each have their place in
> writing code...
>
Perhaps so, but since the addition of xor might break code by creating a new
keyword I suspect you would have a difficult time persuading anyone who
counts to do it. So you're stuck with

    (a and not b) or (b and not a)

for the present. Maybe in Python 3?

regards
 Steve






From patmiller at llnl.gov  Thu Apr 11 13:33:24 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Thu, 11 Apr 2002 10:33:24 -0700
Subject: Python-list digest, Vol 1 #10442 - 15 msgs
References: <20020411164626.5925.85110.Mailman@mail.python.org>
Message-ID: <3CB5C8E4.34043420@llnl.gov>

Denis writes:
> But this work and AFAIK is not implementation dependent:
> 
> >>> def set(name, val):
> ...     exec '%s=__val__' % name in {'__val__': val}, \
> ...                           sys._getframe().f_back.f_locals
> ...
> >>> a = 7
> >>> set('a', 99)
> >>> a
> 99

Alas, it doesn't really work for the reasons outlined regarding changing
local variables in one of my previous messages.  In your test
case, you have the illusion of it working because the
outermost frame's local and global dictionaries are coincident
(i.e. a is treated as a member of the global variable namespace).
The example below shows that if you enter a new frame, then
locals() does NOT allow you to update values.

Another victim of the locals() globals() namespace wars...
Pat


% python
Python 2.1 (#1, Aug 23 2001, 10:42:50) [C] on osf1V4
Type "copyright", "credits" or "license" for more information.
>>> def set(name,val):
...   import sys
...   exec '%s=__val__' % name in {'__val__': val}, sys._getframe().f_back.f_locals
...
>>> a = 7
>>> set('a',99)
>>> print a
99
>>> # So far so good, but......
...
>>> def test():
...   b = 7
...   set('b',99)
...   print b
...
>>> test()
7
>>>

-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

Access to power must be confined to those who are not in love with it. 
 -- Plato




From roy at panix.com  Sun Apr 21 09:34:12 2002
From: roy at panix.com (Roy Smith)
Date: Sun, 21 Apr 2002 09:34:12 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <uc5aajf0m2dmc5@news.supernews.com>
Message-ID: <roy-F25CD7.09321421042002@oldnews.panix.com>

"John Roth" <johnroth at ameritech.net> wrote:
> One of the things that is most unusual about Python is
> it's rejection of bracketing where it isn't needed:
> using indentation for block structure is one example.
> This is diametrically opposed to Lisp.

The one nice thing about lisp syntax is that it's absurdly easy to parse.  
Given that the machines it originally ran on had less memory and processing 
power than your typical kitchen appliance today, not wasting anything on 
parsing was important.

What's truly mind boggling is that those same machines parsed fortran.


From mgilfix at eecs.tufts.edu  Thu Apr 25 11:14:27 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 25 Apr 2002 11:14:27 -0400
Subject: bitmap images
In-Reply-To: <287a1ca.0204250705.54df316f@posting.google.com>; from eviljewel@evilemail.com on Thu, Apr 25, 2002 at 08:05:37AM -0700
References: <287a1ca.0204250705.54df316f@posting.google.com>
Message-ID: <20020425111427.A25670@eecs.tufts.edu>

  Can you be more specific? What do you want to use them for? I assume
you probably want to use them in a GUI program, in which case, which
GUI toolkit are you using?

                 -- Mike

On Thu, Apr 25 @ 08:05, Eviljeweller wrote:
> If anyone can tell me how to add bitmap images to a python program in
> any way shape or form i would be very grateful.
> 
> thanx
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (eviljewel)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From vasvir at iit.demokritos.gr  Tue Apr  2 06:19:46 2002
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Tue, 02 Apr 2002 13:19:46 +0200
Subject: new classes with __slots__ and weakref
Message-ID: <3CA993D2.7030500@iit.demokritos.gr>

Halo,

The following fragment of code is not working as expected

import weakref
class newclass(object):
   __slots__ = []
   pass

inst = newclass()
prx = weakref.proxy(inst)
print inst
print prx


If you comment out the __slots__ statement it works as expected.

I have found this problem acknowledged in goggle. However I would like 
to be informed if there are any recent advances or workarounds.

Thanks for you time.

    .Bill



From martin at v.loewis.de  Wed Apr 10 13:30:37 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Apr 2002 19:30:37 +0200
Subject: Transferring files using sockets
References: <6ac1ecad.0204100826.25217b0a@posting.google.com>
Message-ID: <m3n0wb31he.fsf@mira.informatik.hu-berlin.de>

theivo at abv.bg (I.C.) writes:

> I am trying to transfer a file using the socket module, but I'm not
> sure how to do this. Should I split the file into little(1KB) packages
> and then transfer them, or is there a possibility to send the whole
> file?

You can transfer the entire file contents in a single .write or .send
call; the TCP implementation will itself split it into smaller
packets. You still need to have a (byte) string object as an argument
for .write/.send; as a result, you would need to read the entire file
first. If the file is very large, reading it all at once may exhaust
the memory - that's why it might be desirable to split it into chunks.

If you don't want to write the copy loop yourself, you could consider
shutil.copyfileobj.

HTH,
Martin



From phd at phd.pp.ru  Fri Apr 12 05:16:41 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 12 Apr 2002 13:16:41 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <20020410192232.G19899@phd.pp.ru>; from phd@phd.pp.ru on Wed, Apr 10, 2002 at 07:22:32PM +0400
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru>
Message-ID: <20020412131641.L11075@phd.pp.ru>

On Wed, Apr 10, 2002 at 07:22:32PM +0400, Oleg Broytmann wrote:
> On Wed, Apr 10, 2002 at 02:54:20PM +0000, Michael Hudson wrote:
> > >    Environment: SUN Sparc Ultra-1, Solaris 2.5.1, gcc 2.8.1.
> >                                                         ^^^^^
> > Is this negotiable?  I.e. could you try upgrading?  This is pretty old now.

   I upgarded my software, including gcc (to 2.95.3). Not much help:

gcc -s  -Xlinker --export-dynamic -o python \
                Modules/python.o \
                libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm
case $MAKEFLAGS in \
*-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py build;; \
esac
running build
running build_ext
building 'struct' extension
Segmentation Fault - core dumped
make: *** [sharedmods] Error 139

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




From ajs at ix.netcom.com  Tue Apr  2 22:10:53 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 2 Apr 2002 22:10:53 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <001601c1dabd$2adbb340$0334fea9@carol>

Tim writes -

>The PEP doesn't mention simplicity, except perhaps >indirectly and
partially via its claim that the existence of a >standard, distinguishable
true/false type could make life >easier for a particular class of external
library
>(which has been confirmed by people maintaining >such .libraries).

Probably the best way to keep me out of future PEP debates is
to make sure I don't understand *any* of it.

Unfortunately I understand the significance of the millisecond
recognition time, which is the significance of a millisecond
of recognition time, at the potential loss of the ability
to read and understand lots of current and probably future
Python code using the idiom false = 0.

I will otherwise defer to Alex's comments.

His take on  the non-technical issues here seem to
express with much better authority my own take.

Of course on the technical stuff I find him way off base.

But I'm not in the mood to tackle him on it.

Art





From boud at valdyas.org  Mon Apr 22 14:53:17 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 22 Apr 2002 17:53:17 -0100
Subject: Python, PyQt make install problems on IRIX
References: <f0fd5987.0204220106.7673e345@posting.google.com>
Message-ID: <3cc45c8d$0$307$e4fe514c@dreader1.news.xs4all.nl>

Garry Taylor wrote:

> 
> Has anyone else had this problem, or have any clues as to why it is
> happening?
> 

Well, Irix isn't exactly a common thing (I think I've
seen only one machine running it in my life), but could
it be possible that the installation process updates the
dynamic library cache or something, which subsequently
gets un-updated upon reboot? If you've installed PyQt
from one xterm, can you run PyQt apps from another?

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


From neal at metaslash.com  Thu Apr  4 21:00:18 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 04 Apr 2002 21:00:18 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <7x7knnvsc6.fsf@ruckus.brouhaha.com>
Message-ID: <3CAD0532.C9B76EE4@metaslash.com>

Paul Rubin wrote:
> 
> James Logajan <JamesL at Lugoj.com> writes:
> > x = ["is", "are"]
> > y = {0: "", 1:"s"}
> >
> > for n in range(3):
> >    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
> >
> > Will the behavior be backword compatible and equivalent or not? Simple
> > question I assume; waiting for an answer.
> 
> Yes, if I understand the PEP, that should still work.
> 
> This, however, won't work compatibly:
> 
>    print "Calm down, it's only %s's and %s's." % (1==1, 1==0)

Correct, 2.1:
	>>> print "Calm down, it's only %s's and %s's." % (1==1, 1==0)
	Calm down, it's only 1's and 0's.

2.3:
	>>> print "Calm down, it's only %s's and %s's." % (1==1, 1==0)
	Calm down, it's only True's and False's.

However, if you really want integers, then ask for them:

2.1:
	>>> print "Calm down, it's only %d's and %d's." % (1==1, 1==0)
	Calm down, it's only 1's and 0's.

2.3:
	>>> print "Calm down, it's only %d's and %d's." % (1==1, 1==0)
	Calm down, it's only 1's and 0's.

That is backwards compatible.  And explicit.


From zopestoller at thomas-guettler.de  Fri Apr 12 04:50:43 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 12 Apr 2002 10:50:43 +0200
Subject: Worldwide Lexicon project, call for volunteers
References: <e2f36e18.0204111716.78d6f0a6@posting.google.com>
Message-ID: <3CB69FE3.6000108@thomas-guettler.de>

Brian McConnell wrote:

> Worldwide lexicon... call for volunteers...


Sound interesting. How are you planing the user management?

What is your document format? (HTML, Docbook, ...)?

Can you make links between lexicon servers? How do you keep them
from getting broken?

  thomas





From pyth at devel.trillke.net  Fri Apr 19 18:05:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 20 Apr 2002 00:05:14 +0200
Subject: Substitution with Hash Values
In-Reply-To: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com>
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com>
Message-ID: <20020419220514.GL28515@devel.trillke>

On Fri, Apr 19, 2002 at 04:48:12PM -0500, Jeff Jones wrote:
> I am a long time perl user who is trying to get a grip on python.  I
> really like what I have seen so far, but I have hit a little snag
> regarding substitution (like perl).
> 
> I use html templates for all of my CGI programming, and everything that needs
> to be replaced is contained within double curly braces ( {{value}} ).  In
> perl I use a hash to keep all of the variables that are going to replace
> the placeholders in the template, and replace them the following way:
> 
> 	....
> 	s/{{(.*?)}}/$hash{$1}/g;
> 	....
> 

my try (although my perl knowledge is at best dusty):

	 def value(match):
	    return dict[match.group(1)]
	
	 regex_key=re.compile('\{\{(.*?)\}\}')

	 regex_key.subn(value, str)

you might like to write that in one line

	 re.sub('\{\{(.*?)\}\}',lambda x: dict[x.group(1)], str)

where dict is a dictionary. As you see you can pass functions
as a 'substitution pattern'. This function operates on the
match object. 

is this what you want?

regards,

	holger





From imbosol at vt.edu  Sun Apr 21 16:19:28 2002
From: imbosol at vt.edu (Carl Banks)
Date: Sun, 21 Apr 2002 16:19:28 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <0evu9a.621.ln@127.0.0.1> <mailman.1019413960.21990.python-list@python.org>
Message-ID: <gs6v9a.vb1.ln@127.0.0.1>

holger krekel wrote:
> On Sun, Apr 21, 2002 at 02:12:16PM -0400, Carl Banks wrote:
>> Allan Crooks wrote:
>> > Hi,
>> > 
>> > I was wondering what the best approach is on providing libraries for
>> > varying versions of Python?
>> > (...)
> 
>> # Import stuff for version 2.1
>> if sys.version[:2] >= (2,1):
>>     from mymodule2_1 import *
>> 
>> # Import for version 2.2
>> if sys.version[:2] >= (2,2):
>>     from mymodule2_2 import *
> 
> this solves part of the 'code duplication' problem. Might get a bit tricky
> if you just want to define a few additional methods for some classes
> (for example 'generator versions' of methods otherwise returning lists).

I wouldn't say so.  It is straightforward enough to add a method to a
class any time:

>>> class a:
...    pass
...
>>> def p(self,x):
...    print "hello, %s" % x
...
>>> a.p = p
>>> b = a()
>>> b.p("world")
hello, world



> Ah, and by the way how is
> 
>        sys.version[:2] >= (x,y) 
> 
> supposed to work? On my python 2.2 
> 
>        sys.version[:2] == '2.'   is True
>
> am i missing something here?

I just did it how the OP did it, without bothering to check.  It
should be sys.version_info[:2] >= (x,y)


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From cliechti at gmx.net  Sat Apr 13 12:38:10 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Apr 2002 18:38:10 +0200
Subject: dos promt simulation
References: <6ac1ecad.0204130435.93d0ad3@posting.google.com>
Message-ID: <Xns91EFBDC8F2FCcliechtigmxnet@62.2.16.82>

theivo at abv.bg (I.C.) wrote in 
news:6ac1ecad.0204130435.93d0ad3 at posting.google.com:
> Hello,
> I would like to simulate a simple dos prompt using a Tk window. I use
> os.popen(command) to run some commands like 'echo', 'dir' etc... Then
> I read from the pipe and it works, but if I try to execute a command
> like 'time', which expects input, I can't read from the pipe. Is it
> possible to simulate such commands, without passing some
> parameters('time 22:10')?

there is also popen2, popen3 etc. they give you a file for stdin where you 
can write to.
(don't know if it works for _all_ dos programs. some might well read the 
keyboard through hacky ways.)

chris

-- 
Chris <cliechti at gmx.net>



From anandpillai6 at yahoo.com  Mon Apr  1 06:23:36 2002
From: anandpillai6 at yahoo.com (Anand)
Date: 1 Apr 2002 03:23:36 -0800
Subject: Problem with PIL
Message-ID: <41c203e.0204010323.12f9350b@posting.google.com>

Hi

 I installed the latest beta version of PIL on Windows NT. When I try
to
run one of the sample programs I am getting the error as follows:


$python player.py D:/users/anand/docs/docs/*.gif
loading...
Traceback (most recent call last):
  File "player.py", line 120, in ?
    UI(root, im).pack()
  File "player.py", line 51, in __init__
    self.image = ImageTk.PhotoImage(im)
  File "d:\python21\lib\PIL\ImageTk.py", line 68, in __init__
    image.load()
  File "d:\python21\lib\PIL\ImageFile.py", line 126, in load
    self.load_prepare()
  File "d:\python21\lib\PIL\ImageFile.py", line 180, in load_prepare
    self.im = Image.core.new(self.mode, self.size)
  File "d:\python21\lib\PIL\Image.py", line 37, in __getattr__
    raise ImportError, "The _imaging C module is not installed"
ImportError: The _imaging C module is not installed
Exception exceptions.AttributeError: "PhotoImage instance has no
attribute '_Pho
toImage__photo'" in <method PhotoImage.__del__ of PhotoImage instance
at 00D9301
C> ignored

I can find the installation file _imaging.pyd under "DLLs" directory
in the installation directory of PIL. What else am I missing here...

Thanks for any help in advance,

Anand


From reply at in_newsgroup.tks  Tue Apr  9 07:01:48 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 9 Apr 2002 11:01:48 GMT
Subject: Newbie - web app deployment issues
References: <a8skcd$mr5$1@hfc.pacific.net.hk> <23891c90.0204090043.287c317f@posting.google.com>
Message-ID: <a8uhms$b5m$1@hfc.pacific.net.hk>

paul at boddie.net (Paul Boddie) wrote in
news:23891c90.0204090043.287c317f at posting.google.com: 

> It probably looks like I'm tracking down all your posts to this
> newsgroup, but then it's a solid topic you're bringing up and I can't
> resist responding. ;-)

No problem, I appreciate your feedback. It is an interesting area 
(developing and deploying web apps based on Python) and I hope any replies 
to my post can also assist other Python newbies.

IMO it would be great if the use of Python in producing web apps was to 
increase. Better promotion and visibility (marketing) of Python to new 
developers and to higher educational establishments is essential.

I'm sure there are skilled people who want to contribute to Open Source 
projects but cannot find ways to do so. In many projects, this is also a 
failing of the project's coordinators. Of course the 'bread and butter' of 
projects is the actual coding and bug testing but establishing a website 
and posting in a few forums and newsgroups is the extent of most projects' 
'marketing' efforts.

PR and marketing is an essential area and one in which people with those 
skills, who want to assist Open Source projects, could make a major 
contribution (cue some appropriately stirring music, hehe!).

BTW, I'm not in PR or marketing, just in case anyone thought I was doing a 
hard sell :-p

>> As a Python newbie who wants to try developing web apps, I want to
>> avoid as much as possible, the need to deploy to specialized hosting
>> environments. For example, out of the many hosting companys in the
>> world, only a very small percentage provide Zope hosting. I haven't
>> confirmed this but there's probably a premium to pay for Zope hosting
>> (which may be warranted but it's not something I want to pay for at
>> this point in time). 
> 
> I've seen http://www.freezope.org, but that seems really to be for
> experimentation only. However, you could always choose a very-low-cost
> Web host which only provides static Web pages and then link to dynamic
> content on FreeZope or other such sites.

Thanks for the pointer.

[snip]

> [1] mxCGIPython is a good start for those wanting a binary, but I then
> discovered the FreeBSD "ports" system (see http://www.freebsd.org) and
> it's possible to find lots of binaries for download in .tar.gz format,
> including various Python extension modules (MySQLdb, for example) -
> it's very convenient indeed!

I had a quick look at the "FreeBSD Ports" system and it looks very 
convenient.

On the other hand, I definitely prefer to test my apps on my home PCs 
before uploading them to a host and that would mean installing and 
familiarizing myself with FreeBSD. 

I don't feel I would be sufficiently disadvantaged by not using FreeBSD and 
will therefore not add it to my current list of things to learn. Something 
for the future perhaps ;-)

FYI, regarding my previous post about web app development, I will begin by 
learning to develop simple Python CGI apps and progress from there.

Cheers,
Scrumpy :)


From ajs at ix.netcom.com  Sun Apr 14 21:55:39 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 14 Apr 2002 21:55:39 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
Message-ID: <000b01c1e420$a77a96c0$0334fea9@carol>

[phil hunt]
> It is not new features that are the problem, it is new features that
> break existing code.

[tim]
>According to some people.  Other people have said new >features are a
problem even when they're 100% backward >compatible.

I'm more of that camp.  And in my opinion, its more an issue of pace and
timing than of change itself.

But it has to be clear who I consider my constituency to be. And I will
describe that simply as a constituency of one - myself.  But posit that I
represent an important subgroup of Python's potential constituency - but one
that would logically
be least heard from and least heard in a forum such as c.l. py.

Described briefly as a non-professional programmer, using Python as a first
language, to explore and problem solve.

My only comment on PEP284 had been that any perceived gain on behalf of
novices justified on the basis of a syntax that:

"closely matches standard mathematical notation, so is  likely to be more
familiar to Python novices than the current
range() syntax."

should be considered in light of the fact the pace of change of Python
syntax has become such that - even if the PEP284 assertion is true, its
adoption *at this time* could arguably be considered a net loss for novices.

There were follow-up posts that indicated I was not alone in having a sense
of frustration in chasing a moving target, and at having fewer and fewer
available resources - in terms of reference materials and code samples and
the like - which spoke Python in its current dialect and could be used for
reference and learning purposes.

But in truth I think the issue goes deeper than that.

Maybe I can phrase it as the Culture of Python vs. Python as Culture. The
culture of Python being highly technical and mostly among computer
scientists and professional programmers, Python as culture, a liberal art -
Python significance as a  programming language very much tied to its ability
to act as a common language for the highly technical and the less technical.

One of many examples being one of the IPC10 presentations on a system to
archive antique musical scores, with Python becoming the important glue
between the technologists and the archivists.

And I hope I am correct that it is really only a matter of pace - which I
would still argue has been unnecessarily brisk - rather than substance.
Will Python remain a common language, a point of contact, between folks with
different expertise, or evolve into  a language spoken in dialects so
divergent that it eventually fails as a touchstone between such groups.

Art






From maxm at mxm.dk  Tue Apr  9 04:48:31 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 09 Apr 2002 10:48:31 +0200
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>
Message-ID: <3CB2AADF.9050908@mxm.dk>

Oleg Broytmann wrote:

>    I do not think the PEP *did* damage - it only revealed that our
> community has problems. I don't understand what problems.

The whole bool thing as seen from my point of view happened something 
like this:

Guide posted a PEP. Some people agreed, some disagreed. Personally I 
didn't really care for or against so I just kept quiet.

Then Laura made the (very long) argument against the PEP. When I read it 
I could see that she had some points, but at the same time i thought 
that a boolean type really isn't that hard to fathom as she tried to 
point out. Anyway most of us define our own true=1 false=0 so what was 
wrong with having it in the language?

Due to the very verbose, but rather precise explanation a lot of people 
agreed with her. I still had the feeling that it was a rather small 
problem and though I agreed in her point, I didn't agree with the 
importance of her point.

I once heard a saying "TV is a lot like a microscope. If you enlarge it 
enough, even a breadcrumb seems important" This was how I felt about 
Lauras post.

So when Guido wrote that the post was an overlong piece of FUD I more or 
less agreed. But I was rather surprised by his choice of words, and 
found it to be a rather rude brush-of. And I do believe that if I had 
made a lot of work making an argument like Lauras, that I would feel a 
bit bad about it being turned down that way.

It was hinted that discussions had taken place on python-dev, but I 
don't subscribe to it so all I, and I guess most others, got was the 
short reply from Guido.

So even though I agree that Guido didn't have to reply seperately to 
each of Lauras points, I believe that a longer softer and more humane 
reply would not have hurt in this case.

Online you often need to state what you find obvious so others won't 
miss the crucial.


So perhaps Guido need a little work with the people skills ;-) and the 
rest of us need to act properly, which includes not getting upset when 
others don't.


regards Max M



From aleax at aleax.it  Mon Apr 29 03:27:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 07:27:47 GMT
Subject: Super-newbie syntax error question
References: <mailman.1020045350.28826.python-list@python.org>
Message-ID: <TB6z8.30750$8D3.846736@news1.tin.it>

Sean 'Shaleh' Perry wrote:
        ...
>> Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>>   File "C:\PYTHON22\lib\rss2html.py", line 46
>>     f = file("/article.txt", "w")
>>     ^
>> SyntaxError: invalid syntax
>> 
> 
> Should be:
> 
> f = open() not file().

No matters, they're interchangeable here in 2.2, and
anyway would never cause a SYNTAX error.  Diagnostics
such as that shown by the OP are invariably caused by
problems on PREVIOUS lines (e.g., unclosed parentheses)
or indentation issues with the current one.


Alex



From phr-n2002a at nightsong.com  Thu Apr 25 07:33:25 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 04:33:25 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com> <B8ECCE7D.F941%richard@electrophobia.com>
Message-ID: <7xu1q0vwqi.fsf@ruckus.brouhaha.com>

Richard Parker <richard at electrophobia.com> writes:
> > Currently I'm using
> > 
> > K1 = H('auth1' + K)
> > K2 = H('auth2' + K)
> > MAC = H(K1 + H(K2 + ciphertext))
> > 
> > which I hope is at least as good as HMAC (the keys are more
> > independent) and it uses fewer interpreter operations (though more SHA
> > calls).  It's speed isn't too bad, but I always like to look for
> > improvements.
> 
> I like that idea.  I do, however, recommend that as a precaution you choose
> values for the two constants that differ by more than one bit.

Hmm, that's my instinct too, but I just finished making myself get
over it in order to believe that counter mode AES encryption is ok,
and now I'm supposed to change back?  ;-)

>  While more expensive than HMAC in the theoretical sense, this HMAC
> variant looks secure and I'm not surprised to hear that it is
> actually faster in Python.  I can't see any attack that wouldn't
> also imply the ability to crack the hash function, which is the same
> level of security offered by HMAC.  If you are using the MAC more
> than once with same key you could improve performance by only
> computing the values of K1 and K2 once (this optimization would also
> apply to HMAC).

K in this system is derived from the password and nonce, which in turn
depends on the password, the plaintext, and a random value.

It bugs me that HMAC goes out of its way to make sure the compression
function runs separately for the key and message for both of its
hashes, and my variant doesn't.  There must be a reason for that.  So
maybe I should read the paper and/or bite the bullet and use HMAC.

Anyway, I begin to miss my original scheme of hashing under the
encryption (MAC-then-encrypt):

  ciphertext = encrypt(plaintext + H(plaintext))

The Bellare/Namprempre article list this scheme as insecure but I
don't understand the reasons yet.  I'll have to read the article more
carefully (it's a good article--thanks).  I don't see what problems it
causes, and as we've seen it avoids some problems and may be a little
faster.

Btw, here's my current hmac implementation, which should be faster
than the 2.2 library version on short strings, but I still have to do
some comparative timing.

# sha-1 hmac, RFC 2104

from array import array
import sha
from string import translate

def _setup():
    global _ipad, _opad, _itrans, _otrans
    _itrans = array('B',[0]*256)
    _otrans = array('B',[0]*256)    
    for i in xrange(256):
        _itrans[i] = i ^ 0x36
        _otrans[i] = i ^ 0x5c
    _itrans = _itrans.tostring()
    _otrans = _otrans.tostring()

    _ipad = '\x36'*64
    _opad = '\x5c'*64

_setup()  # initialize constants

def hmac_str(key, msg, digestmod=sha):
    if len(key) > 64:
        key = digestmod.new(key).digest()
    ki = (translate(key,_itrans)+_ipad)[:64] # inner
    ko = (translate(key,_otrans)+_opad)[:64] # outer
    return digestmod.new(ko+digestmod.new(ki+msg).digest()).digest()


From rjones at ekit-inc.com  Wed Apr  3 02:01:59 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 17:01:59 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <7xzo0lth3r.fsf@ruckus.brouhaha.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017795008.700.python-list@python.org> <7xzo0lth3r.fsf@ruckus.brouhaha.com>
Message-ID: <200204030702.HAA29062@crown.off.ekorp.com>

On Wed, 3 Apr 2002 16:46, Paul Rubin wrote:
> Richard Jones <rjones at ekit-inc.com> writes:
> > For a _practical_ example of the "markup", please see the Roundup
> > documentation mentioned in my response. I think you'll find that the
> > original text is perfectly readable and significantly easier to edit
> > and maintain that any "explicit" markup language. Apologies for the
> > long
> >
> > http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/roundup/roundup/doc/instal
> >lation.txt?rev=HEAD
>
> That looks pretty similar to an Info file (texinfo output).  I think
> I'd rather type "@section{Overview}" than have to supply those special
> underscores directly.

Yeah, I guess at this point it's down to personal preference. I think we can 
summarise by saying that you're happy to have your documentation source in 
the texinfo format [#]_, whereas the Structure Text camp would rather see it 
more looking like actual documentation. Is that fair?

. [#] or something else that's got "bearable" readability issues, where
      "bearable" is determined on an entirely personal scale :)


> > Note that my manual Table of Contents at the top is now possible using a
> > directive::
> >
> >    .. contents:: **Table of Contents**
> >
> > The output is at http://roundup.sf.net/doc/installation.html
>
> So the ink isn't even dry on this PEP and already it's sliding down
> the slippery slope into explicit markup <wink>.

I think it's justifiable in this case - it saves a lot of typing :)

Also, the other commonly used explicit markup is images::

  .. image:: graphic.png

... that's pretty hard to do without an explicit markup (short of 
ASCII-art'ing the PNG :)



> > Please, read the justification text, and the anlysis of existing formats.
> > POD, while not mentioned, shares a lot of the shortcomings of other STX
> > formats in that it was grown over time.
>
> I did read the justification text but found it scanty and
> unpersuasive.  It read too much like a sales pitch and not enough like
> an objective analysis.

Fair enough, I thought it was pretty balanced. 


> > Again, have you _read_ the ReST documentation, specifically the
> > discussion of alternatives? It's extensive, and very well
> > written. It discusses many existing STX systems and their strengths
> > and weaknesses.
>
> I looked briefly at the ReST documentation linked from the PEP, but
> if I'm supposed to read it in detail to be able to use it, I might
> as well use a more traditional markup language.

I've started writing a "primer" that should be the first port of call. It's a 
somewhat more verbose version of selected parts of the quick reference. Just 
the basics, explained in a little more detailed and human way than the formal 
documentation. I'd hope that that helps a little...



     Richard



From anthony_barker at hotmail.com  Mon Apr  8 21:14:05 2002
From: anthony_barker at hotmail.com (Anthony_Barker)
Date: 8 Apr 2002 18:14:05 -0700
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
Message-ID: <899f842.0204081714.25eab798@posting.google.com>

Python 2.2x is stable why does redhat still insist on releasing 1.5x?
All other distributions are now on 2.x.

It seems from the Redhat's skipjack beta 2 that it will only ship the
older version of python(1.5.2).

I find this annoying because it means I have to upgrade each redhat
machine individually as it breaks a couple of my scripts.


Anthony
http://xminc.com/anthony/


From sholden at holdenweb.com  Mon Apr 15 07:32:43 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 07:32:43 -0400
Subject: Regex question about \d
References: <mailman.1018203796.20857.python-list@python.org>
Message-ID: <BVyu8.9724$Ji7.271@atlpnn01.usenetserver.com>

"Gerhard H?ring" <gh_pythonlist at gmx.de> wrote in message
news:mailman.1018203796.20857.python-list at python.org...
> Can somebody please explain this to me:
>
> >>> print re.search(r"\d*", "asdfsfafsw 453 dfsa").group(0)
>
> >>> print re.search(r"\d+", "asdfsfafsw 453 dfsa").group(0)
> 453
>
> Why doesn't the first regex match the digits?
>
Because it matches the empty string at the beginning -- the "*" implies zero
or more occurrences, so the pattern matches right at the start, and need
look no further for digits.

regards
 Steve






From pereira at cis.upenn.edu  Mon Apr 22 20:50:45 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Mon, 22 Apr 2002 20:50:45 -0400
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019510980.7097.python-list@python.org> <3CC48EDD.2BEE5215@tundraware.com> <qX0x8.82827$T%5.64632@atlpnn01.usenetserver.com>
Message-ID: <B8EA2825.83F6%pereira@cis.upenn.edu>

On 4/22/02 7:20 PM, in article
qX0x8.82827$T%5.64632 at atlpnn01.usenetserver.com, "Steve Holden"
<sholden at holdenweb.com> wrote:
> Maybe Cambridge had it right in
> including both mathematics and science under "Natural Philosophy".
Pretty appropriate if "natural philosophy" is allowed to include psychology,
as it did in the Enlightenment (see Roy Porter's "The Creation of the Modern
World", esp. chapter 7). Since, as everyone knows without realizing,
mathematics is just a very elaborate kind of perceptual technology <0.5
wink>. 

-- F



From richard at electrophobia.com  Thu Apr 25 05:30:10 2002
From: richard at electrophobia.com (Richard Parker)
Date: Thu, 25 Apr 2002 09:30:10 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com> <87ofg88brt.fsf@saltationism.subnet.hedonism.cluefactory.org.uk>
Message-ID: <B8ED1AB1.F992%richard@electrophobia.com>

Paul Crowley wrote on 4/25/02 12:42 AM:

> Not so - it is far easier for the attacker to do 2^64 offline work
> than it is to present 2^64 forged messages to the remote end.

Paul Rubin's application is general purpose library.  It is entirely
plausible that Paul Rubin's library could be used in an application that
essentially provides the attacker with an oracle.  I agree that it is not a
good idea to replace an operation that to break requires 2^64 queries for
one that requires 2^64 work.  However, while queries are, in practice,
usually more difficult, I think it is a mistake in a general purpose library
to assume that making 2^64 queries is necessarily any "harder" than 2^64
work.

-Richard



From BPettersen at NAREX.com  Thu Apr 25 18:36:08 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 25 Apr 2002 16:36:08 -0600
Subject: Python doesn't know how to do math?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192B982E2@admin56.narex.com>

> From: Chris Spencer [mailto:clspence at one.net] 
> 
> 	I tested this using the Python command interpreter on a 
> Windows XP box
> running:
> Python 2.1 (#15, Apr 16 2001, 18:25:49) [MSC 32 bit (Intel)] 
> on win32 Type "copyright", "credits" or "license" for more 
> information.
> >>> float(5)/float(100)
> 0.050000000000000003
> >>> 5.0/100.0
> 0.050000000000000003
> 
> 	Okay, I might not be a math genius, but this is most 
> definately WRONG. What gives?  Anyone have insight into this 
> little "feature" of Python floating point operations?

http://www.python.org/doc/current/tut/node14.html

Wrong-is-in-the-eye-of-the-representation'ly y'rs
-- bjorn




From sholden at holdenweb.com  Tue Apr 30 18:14:15 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 18:14:15 -0400
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <jCzz8.17994$xy4.10082@atlpnn01.usenetserver.com> <sfAz8.82271$vF6.2450688@news2.tin.it> <6OAz8.1251$Au5.843@atlpnn01.usenetserver.com> <LJDz8.35771$8D3.1034180@news1.tin.it>
Message-ID: <sJEz8.4912$Au5.1585@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:LJDz8.35771$8D3.1034180 at news1.tin.it...
> Steve Holden wrote:
>         ...
> > sloppiness. And, of course, to confound the confusion int() is *not* a
> > call to the integer class/type creator, but a no-argument call of the
> > built-in int() function. What happens to "int(x[, radix]) " when int is
a
> > built-in type, by the way, or isn't that going to happen?
>
> Steve, what are you _taking_ about...?
>
> >>> type(int)
> <type 'type'>
> >>> int('23',6)
> 15
>
> int *is* a built-in type, a call to it is a call to the int type, and one
of
> the ways you can call the int type to create an int istance is with a
> string argument and optionally a radix integer argument.
>
> Are you stuck in the far, dark ages of Python 2.1 ... ?-)

[...]

Well, I was only quoting the 2.2 Python Library Reference (section 2.1,
ironically enough) [he said, mildly]:

"""
int(x[, radix])
Convert a string or number to a plain integer. If the argument is a string,
it must contain a possibly signed decimal number representable as a Python
integer, possibly embedded in whitespace; this behaves identical to
string.atoi(x[, radix]). The radix parameter gives the base for the
conversion and may be any integer in the range [2, 36], or zero. If radix is
zero, the proper radix is guessed based on the contents of string; the
interpretation is the same as for integer literals. If radix is specified
and x is not a string, TypeError is raised. Otherwise, the argument may be a
plain or long integer or a floating point number. Conversion of floating
point numbers to integers truncates (towards zero).
"""

So, while I thought the change had been made (because type(int) ==
type(type)) the documentation seemed to say otherwise. This is the kind of
situation that tends to confuse. Even though the type/class unification is
going ahead, the documentation isn't explicit about it, and int() is still
represented as a built-in function when really it's a type. Unless, that is,
you know different.

Types, of course, are callables, so it's reasonable to represent them as
functions, but it's interesting that the new types' constructors are so
explicitly polymorphic across the "sensible" arguments. The documentation is
unchanged in the CVS tree.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/








From aahz at pythoncraft.com  Mon Apr  1 18:30:28 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 18:30:28 -0500
Subject: The fastest XML parser around
References: <mailman.1017654423.12312.clpa-moderators@python.org> <a89pnc$7j9$1@panix1.panix.com> <UD5q8.874$aO4.4453379@news-text.cableinet.net>
Message-ID: <a8aqik$kfe$1@panix1.panix.com>

In article <UD5q8.874$aO4.4453379 at news-text.cableinet.net>,
Andy Robinson <andy at reportlab.com> wrote:
>
>I plan to rewrite pythonpoint using our XML toolkit - indeed, use
>it as the exercise in the manual - but we need to finalize two
>more components first: a 'lazy accessor' giving friendly syntax
>for navigating the tree, and something to convert attributes to
>numbers/lengths/colors en masse.  After that the Pythonpoint code could
>get a lot shorter.  Some broad plans are in the back of the pyRXP
>documentation, referenced from the web page my announcement pointed at.

Okay, then, I probably shouldn't submit patches (not that I can, anyway,
until I figure out how to make SF work with Lynx).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From st-newsgroups at thomas-guettler.de  Thu Apr  4 02:06:20 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Thu, 04 Apr 2002 09:06:20 +0200
Subject: Absolute to relative URL?
Message-ID: <3CABFB6C.7050308@thomas-guettler.de>

Hi!

Is there already a function in the standard python modules
which returns the a relative url given two absolute urls?

Example:

relative_url("http://foo/a/b/c", "http://foo/a/d")
  --> return: "../d"

This method are only some line of codes. But I prefere
to take tested functions.

  thomas



From tanzer at swing.co.at  Mon Apr  1 02:14:59 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Mon, 01 Apr 2002 09:14:59 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "Sat, 30 Mar 2002 04:35:32 EST."
             <LNBBLJKPBEHFEDALKOLCMEBLOJAA.tim.one@comcast.net>
Message-ID: <E16rw1r-00016o-00@tswings.swing.cluster>

Tim Peters <tim.one at comcast.net> wrote:

> [Christian Tanzer]
> > ...
> > If str and repr were to behave differently from each other, I'd expect
> > repr(True) to return "1" and str(True) to return "True". Interchanging
> > that seems strange.
>
> We try to ensure that eval(repr(x)) reproduce x whenever reasonably
> possible.  That best way to do that here is for repr(True) to return
> 'True'.

I understood the motivation for `repr(bool(1)) == "True"`.

What I wanted to say is that having repr return a readable result and
str return a result looking like an internal representation runs
counter to the usual conventions used for repr and str.

> > OTOH, I'm worried about backwards compatibility.
>
> That's the only reason for str(True) to return '1'.

In this case, I'd value consistency between repr/str higher than
backwards compatibility.

> How about str(True) return '1' but str(False) return 'False'? That's
> what a committee would compromise on <wink>.

Any self respecting committee would try harder and probably come up
with a really symmetric compromise:

    value    str      repr
    True     '1'      'True'
    False    'False'  '0'

<1.4. wink>

> > How will bool influence '__nonzero__'? Currently, Python insists that
> > '__nonzero__ should return an int'. Will this be changed to 'should
> > return a bool'?
>
> I hope so.
>
> > As that would break existing code, I hope not.
>
> No, "should" != "must".

Fine display of sophisticated hair splitting, but:

    >>> class Foo:
    ...   def __nonzero__(self): return None
    ...
    >>> f = Foo ()
    >>> not f
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    TypeError: __nonzero__ should return an int
    >>>

I don't care about `should` vs. `must` as long as the new and improved
Python won't spit an exception at __nonzero__ code returning an int
instead of a bool.

> Guido won't call code that returns an int here broken, although he
> might call it deprecated, and leave you wondering when the hammer will
> drop <wink>.

Nice tease. But Guido's hammer with regard to future strictification
of bool is exactly what makes me uneasy.

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




From pinard at iro.umontreal.ca  Wed Apr 24 10:14:45 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 24 Apr 2002 10:14:45 -0400
Subject: Slight irritation with try/finally indentation
Message-ID: <oqy9fdp4iy.fsf@titan.progiciels-bpi.ca>

Hello, people.

There is something which I find slightly irritating with the try/finally
construct, and wonder if someone would not have a good advice to offer.
I guess it is very usual that one has, within a suite, something like:

    [...]
    setup()
    process()
    cleanup()
    [...]

with `cleanup' undoing the effect of `setup', more or less, really seeing
`setup()' and `cleanup()' as a kind of big parentheses around `process()'.
In a context where `process()' may raise exceptions, `cleanup()' might be
mandatory once `setup()' has completed, so one rewrites the above as:

    [...]
    setup()
    try:
        process()
    finally:
        cleanup()
    [...]

What I find irritating is that `cleanup()' is not aligned anymore with
`setup()', as it was originally, so we loose on the legibility of the
parenthetical idiom we wanted to stress.  Surely, I may _not_ write:

    [...]
    try:
        setup()
        process()
    finally:
        cleanup()
    [...]

despite it would look nicer, because if `setup()' fails, `cleanup()'
gets called when it should not have been.  Resorting to tricks like:

    [...]
    if True:
        setup()
    try:
        process()
    finally:
        cleanup()
    [...]

is the only solution I see, but this is not much appealing either, maybe
the cure is not worse than the illness.  I am not sure.  Are others on
this list bothered with this issue?  And if yes, what do you choose to do?

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




From huaiyu at gauss.almadan.ibm.com  Tue Apr  9 15:09:56 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 9 Apr 2002 19:09:56 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com> <a8slgr$pin$1@nyheter.chalmers.se> <3CB1E745.EDA08705@alcyone.com> <a8sr75$rug$1@nntp6.u.washington.edu> <slrnab6a8r.dvl.huaiyu@gauss.almadan.ibm.com>
Message-ID: <slrnab6f44.dvl.huaiyu@gauss.almadan.ibm.com>

I wrote:
>The basis of type inference is that each object has exactly one type.
 ^^^^^^^^^

I meant  "one basic requirement".

Huaiyu



From rick at bioinformatics.org  Sun Apr 21 20:16:15 2002
From: rick at bioinformatics.org (Rick Ree)
Date: 21 Apr 2002 17:16:15 -0700
Subject: correct Numeric behavior?
Message-ID: <m3hem4si40.fsf@loco.ucdavis.edu>

The following seems counterintuitive:

>>> import Numeric
>>> a = Numeric.array([0])
>>> a
array([0])
>>> not a
1
>>> b = Numeric.array([1])
>>> not b
0

That is, a rank-1 array if length 1 tests false if it's value is 0,
but true otherwise.  This is not the case with the array module:

>>> import array
>>> x = array.array("i", [0])
>>> x
array('i', [0])
>>> not x
0

Is this the intended behavior of Numeric, or is this a bug?


From phil at mrinfo.de  Thu Apr 25 04:33:56 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Thu, 25 Apr 2002 10:33:56 +0200
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com> <aa5f7l$uck$1@news.nextra.cz> <jwbaxter-B5D4A7.19244624042002@corp.supernews.com>
Message-ID: <aa8f00$50o$1@swifty.westend.com>

"John Baxter" <jwbaxter at spamcop.net> wrote in message
news:jwbaxter-B5D4A7.19244624042002 at corp.supernews.com...
>..
> I use single quote (that is, apostrophe) to avoid a shift key
> manipulation.  Unless I forget.  And of course if the string begins with
> a shifted character, the manipulation is deferred, not avoided.
>..

I need to press Shift to get a single quote. Or double-quote, greater-than,
and $.
I'm not sure what is exclusive german keyboard layout here.




From newsfroups at jerf.org  Sun Apr 14 23:40:19 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Mon, 15 Apr 2002 03:40:19 GMT
Subject: currying (was: Re: A trivial question about print)
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com> <3CB756E4.5030404@jerf.org> <slrnabesm3.mfc.huaiyu@gauss.almadan.ibm.com> <1103_1018819947@news.dsl-only.net>
Message-ID: <3CBA4B83.9060604@jerf.org>

Scott David Daniels wrote:
> I had kind of thought that the keyword args would take care of this.  
> Sure there are cases where this wouldn't be true, but  I'd hate to have 
> to read code that included ... curry([2, 3, 5], f, a1, a4, a6) ..., I'd have 
> the devil's own time determining which args were specified.

In the case I'm using it, it would still be a net gain. Basically, I 
have a whole list of function calls (to set up bindings), and I'm trying 
to make the calls as simple as possible.

(Why not use a data file? Well, I'm basically using the Python code 
itself as my data file. Python code understands the Python runtime data 
structures better then any data file interpreter I could possibly cook up.)

I was mostly curious if anybody had come up with a good way of 
expressing it; obviously, with your curry example in hand, creating the 
code to do what you want is the easy part. It's the specs that are hard.

> As to Alex's comment on the recipe -- of course he's right; I just hadn't seen 
> the possibility of curry by closure until nested scopes showed up after I
> had written curry.

BTW, thanks for that recipe. Not only have I found it useful code, but 
it taught me what currying *is*. I'd seen the pure functional kind, in 
lambda calculus style, but I found that hard to follow, so I tried 
http://www.google.com/search?hl=en&q=python+curry on a whim, and lo, 
there it was. (I think that's also how I found the whole recipes database.)



From jon+usenet at unequivocal.co.uk  Mon Apr 29 19:31:09 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 23:31:09 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org>
Message-ID: <slrnacrltr.6mq.jon+usenet@snowy.squish.net>

In article <mailman.1020119995.30995.python-list at python.org>, Fran?ois Pinard wrote:
>> Woah, *big* disagreement here.  Not sure what kind of designers you are
>> used to, but in my experience they, and/or the tools they use are
>> incapabale of editing the HTML without destroying the code.
> 
> Strange.  For more than one year now, we never had the shadow of a
> problem with HTML designers.  Everybody is competent in his field and
> respectful for the work of others.  There is no "we are good they are
> rotten" speak here.

That's not what I meant. I'm not saying that designers are
incompetent, I'm saying that having to work around code is not part of
their job and the *tools* used will usually make it impossible.

I'll just leave you with one further point. Several times on projects
I have worked on, the customer decided upon a near-total redesign of
an existing site. This involved us receiving completely new HTML files
designed from scratch. With *any* system involving embedded code, this
necessities many days of manual re-integration of the code and HTML.
With something like jonpy, it's the work of a few hours at most.

Cheers


Jon


From shalehperry at attbi.com  Tue Apr 16 16:00:04 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 16 Apr 2002 13:00:04 -0700 (PDT)
Subject: Counting the number of elements in a list - help.
In-Reply-To: <a9hu8e$2fg$1@lumberjack.rand.org>
Message-ID: <XFMail.20020416130004.shalehperry@attbi.com>

> 
> anyone know of a "list size" function?  It has to be out there..hehe, i'm
> just really green at python.  If I knew #2, I could build the function
> myself, but #1 is preferable.
> 

len() returns the size of any python container -- string, dictionary, list,
tuple.




From shalehperry at attbi.com  Tue Apr  2 22:41:10 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 02 Apr 2002 19:41:10 -0800 (PST)
Subject: Newbie: Classes and Lists
In-Reply-To: <8Guq8.1209$zT5.41108@news.itd.umich.edu>
Message-ID: <XFMail.20020402194110.shalehperry@attbi.com>

> 
> import random
> 
> class Chromosome:
> 
>     data = []
>     

this makes data global to ALL members of Chromosome.  You want to move it into
the constructor.

>     def __init__(self, length):

ie. self.data = []

>         for bit in range(length):
>             self.data.append(random.randrange(0,2,1))
> 
>     def display(self):
>         print self.data
> 
> 
> population = []
> 
> for i in range(100):
>     x = Chromosome(20)
>     population.append(x)
> 
> population[50].display()
> 



From jkring at calbay.com  Tue Apr  2 14:13:30 2002
From: jkring at calbay.com (Jim)
Date: 2 Apr 2002 11:13:30 -0800
Subject: Shelve Help (file portablility question)
Message-ID: <89a9b6ce.0204021113.7c4bf243@posting.google.com>

I cannot read a shelve file that I created in Python 1.5.2 on Redhat
6.X using Python 1.5.2 on Redhat 7.X.  Is there any way that I can do
this?

my error message is shown below:

>>> db = shelve.open('shelveFile')
Traceback (innermost last):
    File "<stdin>", line 1, in ?
    File "/usr/lib/python1.5/shelve.py", line 152, in open return
DbfilenameShelf(filename, flag)
    File "/usr/lib/python1.5/shelve.py", line 142, in __init__
Shelf.__init__(self, anydbm.open(filename, flag))
    File "/usr/lib/python1.5/anydbm.py", line 86, in open return
mod.open(file, flag, mode)
    File "/usr/lib/python1.5/dbhash.py", line 8, in open return
bsddb.hashopen(file, flag, mode) bsddb.error: (22, 'Invalid argument')

-Jim


From phr-n2002a at nightsong.com  Fri Apr 26 22:13:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 26 Apr 2002 19:13:24 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com> <3CC7889D.2080002@nowhere.org> <7x8z7cvusy.fsf@ruckus.brouhaha.com> <3CC8BCD1.7090804@nowhere.org> <7xwuuvatiz.fsf@ruckus.brouhaha.com> <aact9l$9odgq$3@ID-99513.news.dfncis.de>
Message-ID: <7xr8l199dn.fsf@ruckus.brouhaha.com>

"P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
> See the project at http://pyaes.sourceforge.net/[1] for a working
> implementation of AES and SHA as C extensions that are unfortunately
> abandoned. Don't need to reinvent the wheel, make it better.
> 
> [1] I may be wrong about the project name, but the code is there
> somewhere.

Yes, I know about that module (the author is on the python-crypto list
where the new module is being discussed).  The new module is intended
as a successor.


From gimbo at ftech.net  Fri Apr  5 12:04:00 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 5 Apr 2002 18:04:00 +0100
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? [LONG!])
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F062334CFB74@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F062334CFB74@mail.hebmex.com>
Message-ID: <20020405170400.GG3249@andy.tynant.ftech.net>

On Fri, Apr 05, 2002 at 10:44:00AM -0600, Gustavo Cordova wrote:

> Andy hadn't caught on that a module is loaded only once,
> the first time it's imported. Any subsequent imports are

To be fair to myself, that's not strictly true.  I did know that, I
just hadn't made the connection between that fact and the implication
that the module was, thus, actually, in fact, a singleton.  :-)

> only accesses to the original imported version, so that
> makes them effectively singletons.

> So if you keep lotsa utility functions, variables and
> constants in a certain module, and import it from a lot
> of diferent places in a single application, then you're
> really referencing the originally imported version, all
> others are only references.
> 
> Neat, eh?

It really is.  My biggest problem with my python programming is
definitely a tendency to overcomplicate things.  :-/

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From lists.ASkwar at DigitalProjects.com  Thu Apr 25 19:39:02 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 01:39:02 +0200
Subject: Python doesn't know how to do math?
In-Reply-To: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <20020425233902.GL21205@teich.Garten.DigitalProjects.com>

?Chris Spencer? sagte am 2002-04-25 um 17:14:12 -0500 :
> 	Okay, I might not be a math genius, but this is most definately WRONG.

Yep, you're not a math genius.  The result is completely correct.  Read the
tutorial for a good explanation!

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 18 hours 38 minutes




From bokr at oz.net  Thu Apr  4 01:01:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Apr 2002 06:01:12 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org>
Message-ID: <a8gq78$hg$0@216.39.172.122>

On 03 Apr 2002 10:59:13 -0500, Andrew Koenig <ark at research.att.com> wrote:

>>> So what's different?
>
>Fredrik> more than one way to do it.  more than one concept to
>Fredrik> explain to newcomers.  more than one "obviously correct"
>Fredrik> way to do things.  less is more.  etc.
>
>I'm a little confused, because it seems to me that Python already
>has more than one way to do this particular thing.
>
>Specifically: If I want to write a function that answers a yes/no
>question, I have lots of possible ways of spelling yes (1, 2, "yes",
>and so on) and lots of possible ways of spelling no (0, {}, None,
>and so on).  There isn't a single preferred way.
>
Everyone seems focused in bools as "variable values" or "function return values"
(incidentally usually ignoring the concept of binding symbols as opposed to
assigning values in the midst of pedagogical concerns, but never mind that).

I've been trying to point to objects, and get some attention to how bools
will/should work with them. if x : foo(x) will test x and not do foo(x)
if x names one of the magic objects 0,{},[],'', -- or: an object whose
__nonzero__() (or __len__() if no __nonzero__) method returns zero.

This is very useful, but what if you would like to subtype int and have its default
bool value be of your design (e.g., true if prime) rather than the usual non-zero test?

I would like a PEP for bool that does not preclude this possibility, or
other custom bool values for objects that may already have other uses for __len__
etc. What if I wanted to subtype str and have a bool signifying URL-escaped or not,
yet still have normally functioning __len__() etc.?

I think a logical (;-) method for this would be __bool__(), and have it override
__nonzero__ just as __nonzero__ apparently overrides __len__ for the determination
of logical value. Then bool(x) can call x.__bool__(), and the designer of x.__class__
can decide whether to return True or False or even a subtype of bool (on which
str(b) could give On/Off, Converged/Computing or whatever suits the design.
My goal freedom for the designer.

Bottom line, I think __bool__ is a better general concept to design classes with than
forcing abuse of someone's Nothing concept.

BTW, I would suggest renaming operator.truth to operator.oitzif and let it return
One If True Zero If False. This doesn't pervert semantics, and does something useful.
Operator.truth(x) can just be defined as bool(x) if someone then still wants it.

An integer-based bool might better be named ibool, so that the name bool can be reserved
for the real thing.

All IMHO ;-)

Regards,
Bengt Richter


From siegfried.gonzi at kfunigraz.ac.at  Thu Apr  4 03:52:38 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 04 Apr 2002 10:52:38 +0200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org>
		<r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
		<lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
		<lbh8z84nake.fsf@xihu.inria.fr>
		<15531.7864.698585.485436@12-248-41-177.client.attbi.com>
		<lbh7knolreu.fsf@xihu.inria.fr>
		<15531.9709.924953.953487@12-248-41-177.client.attbi.com>
		<mailman.1017850275.2134.python-list@python.org>
		<0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <mailman.1017906916.22194.python-list@python.org>
Message-ID: <3CAC1456.C8A51DC9@kfunigraz.ac.at>

Jan Kybic wrote:

> Great, I am becoming even more impressed. However, the main criticism
> remains: You have to code your function twice, first in Python for
> prototyping, then in C. But I agree that weave saves you a lot of
> work.

This can become a burden; especially if one has to code every day some
small functions in Python first and then in C or C++. But isn't it
sometimes a cost/benefit question? If you have a problem where you know
in advance that it will involve heavy duty array access in loops (and
your experience gained before tell you that Python slows down
remarkably), wouldn't it be wise to implement the function in C or C++
from the start on?

I can imagine that in complicated programs it is not always clear what
parts should become exposed to some tuning and what parts not. And the
often strained panacea: "Code in Python and your profiler will tell you
where the weakness lies", can become an obstacle in the long run. 

You are not the first who complained the speed of environments like
Matlab/IDL... I once red a commentar that one was tired of waiting more
then one day in Matlab of an execution result. He then rewrote the
simulation in Fortran 90 and everything went right in about 1 hour or
so.

But you should take into account that Python handles C interfaces very
well. I am not sure how you just use your Python; whether you use Python
only in combination with Numeric and not with a
plotting/image-processing library. I say this in the light of your
possible contemplation to give OCaml a try:

- Is it easy in OCaml to link to a plotting library (e.g. DISLIN)?
- Are there any numerical libraries available (e.g. Numeric in Python)?
- Can you read and handle binary files?
- No joke: Can you write a numer in a specific format to a file (e.g. in
Python "fix()")?; try it in Clean

If your simulation takes always hours then it is not a tradegy to save
the results to a file and consult a plotting program; but if you always
work in the range of seconds and minutes and want to plot your results
in an interactive mode then it can be awkward to work simultaneously
with two or three different programs.


S. Gonzi


From jeff at ccvcorp.com  Mon Apr  8 20:43:54 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 8 Apr 2002 17:43:54 -0700
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com> <1018304015.46066@cswreg.cos.agilent.com> <MPG.171bbba6ee73df8c98969f@news.nwlink.com> <mailman.1018310800.31349.python-list@python.org>
Message-ID: <MPG.171bd911b760bfa19896a0@news.nwlink.com>

In article <mailman.1018310800.31349.python-list at python.org>, 
pinard at iro.umontreal.ca says...
> [Jeff Shannon]
> 
> > If you want access to variables inside of a function, you should 
> > pass those specific variables as arguments.  If you've got lots 
> > of arguments in common between lots of functions, assemble the 
> > lot into a class.  If neither of these is appropriate, then you 
> > should redesign -- you're using a very ugly approach.
> 
> I beg to disagree, a bit strongly.
> 
> Something is not ugly just because you prefer something else.  I have very,
> very neat applications which use globals() and locals(), and they will
> loose much if they use arguments all over, or get assembled into classes.

Very well, I suppose we will have to disagree, then.  To my mind, 
using globals() and locals() in this way is a gross violation of 
encapsulation.  As I see it, if there is data that truly needs to 
be accessed throughout an application (or subsystem), then it 
should be put in an explicit configuration dictionary or object 
(which could be either passed around, or set as global).  

I obviously can't speak to your specific designs, not having seen 
them, but I can't imagine how globals()/locals() would be 
necessary unless you have a large amount of interdependencies in 
your code, and it is generally accepted that interdependent code 
is not a good thing.  If you can explain to me how you can 
require access to large portions of another scope's namespace 
without having large interdependencies, then perhaps I'll change 
my mind about this method.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From donn at u.washington.edu  Thu Apr 11 12:21:57 2002
From: donn at u.washington.edu (Donn Cave)
Date: 11 Apr 2002 16:21:57 GMT
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
References: <mailman.1018515924.12250.python-list@python.org>
Message-ID: <a94d75$hde$1@nntp6.u.washington.edu>

Quoth "Noah" <noah at noah.org>:
| I'm having headaches with nonblocking IO on a pipe.
| I opened a UNIX command through os.popen4.
| Then I set O_NONBLOCK on the output stream of the pipe.
| Finally I did a select() on the output pipe to wait for data.
| If select returns I would try to read the data.
| The problem is that select() is blocking. Select should block 
| until some data is available on the pipe, so I assume that
| select is never even seeing any data on the pipe.
| I added a timeout to select() and then tried to read the
| pipe, but it was empty (read() returned '').
|
| One program that I am attempting to talk to via a pipe is ftp.
| Any interactive program I open will block. 
| If I try open a program that terminates like "ls" or "uname -a" 
| then my script works as expected.
|
| What am I doing wrong that is causing interactive programs to block?

It isn't you.  I'd argue that your O_NONBLOCK is unnecessary, as long
as you're using select anyway, but it won't make any difference if
the command on the other end doesn't send the data.  It doesn't send
the data because its output goes through C library stdio buffers, which
treat a pipe like a disk file and buffer up whole blocks.  There could
be other problems, like some of the data is actually going to stderr,
but when you say the only programs that work are the ones that terminate,
that's a hallmark of this problem (buffers are flushed on exit.)

The only solution is to get away from popen2 and pipes, and use a
pseudotty instead.  2.1 and up have a usable posix.openpty() on some
UNIX platforms, and there's an older pty.py module that may or may
not be useful.  Ptys are sort of tricky and cross-platform support is
poor, but they'll work better in a couple of ways for applications
that are intended for interactive use.

	Donn Cave, donn at u.washington.edu


From sholden at holdenweb.com  Mon Apr 29 07:53:55 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 07:53:55 -0400
Subject: Dammit [was: something else, dammit]
Message-ID: <syaz8.7310$M63.5178@atlpnn01.usenetserver.com>

----- Original Message -----
From: "James J. Besemer" <jb at cascade-sys.com>
To: "Steve Holden" <sholden at holdenweb.com>
Cc: <python-list at python.org>
Sent: Sunday, April 28, 2002 5:29 PM
Subject: Re: Python's Lisp heritage


> Steve Holden wrote:
>
> > Don't quote me out of context, dammit. Just because you didn't detect
the
> > humour didn't mean it wasn't there. And it waas "reeducated", not
> > "reprogrammed to think Pythonically".
>
> (a) Ironically, seriously and truly, you weren't the only person to make a
> remark like that to me last week.  So I wasn't intentionally quoting you
or
> anybody specific.  I was merely paraphrasing my sense of the group
consciousness
> to the best of my recollection. Damn it.
>
Stop saying damn it, dammit.

> (b) "Just because you didn't detect the humor didn't mean it wasn't
there."
> Frankly, I think the L. Ron Hubbard/reprogramming angle is way funnier
than
> being reeducated by the Khmer Rogue.  Damn it.
>
Well, each to their own. Dammit.

> (c) You need to lighten up, damn it.  There's a whole vast world out there
> beyond the kingdom of Python.  Damn it.
>
You've got me laughing at myself, now, dammit.

> > People can also learn to ride bicycles where turning the handlebars to
the
> > left turns the wheel to the right.
> >
> > People can even learn to program in C, dammit.
> >
> > Your problem is you just won't take those damned glasses off again now
> > you've got used to them.
>
> What glasses?  Oh, you mean the example *I* used earlier about how people
can
> adapt to anything?  And how just because people can adapt (e.g., to
Python)
> doesn't mean it is flawless?
>
What, you mean I don't get to reeducate anyone? Dammit!

> What's your point again?  Damn it.  ;o)
>
Dammit, I've forgotten.

> I've programmed in many different languages over the decades and even
maintained
> and implemented a few myself.  I think my perspective is at least as broad
and
> objective as anybody here.  Way more so than some I've seen in my short
time
> here.
>
> You seem to be the one who is locked into just one way of viewing the
world.
>
Perhaps I'm just a leetle bit tired of seeing the same discussions play out
on c.l.py again and again and again. We are all locked in to one way of
viewing the world, we just each have our own way. Dammit.

> > True. But passion is a relative judgement. Programming languages and
> > operating systems stopped being a religious issue for me many years ago,
and
> > so I tend to reserve passion for more appropriate activities.
>
> Geeze you sure fooled me.   ;o)
>
Well, I guess I should lighten up then, dammit.

> I gotta say, though, if one more ankle biter sycophant quotes the
scriptures
> according to Guido to me, I think I'm gonna puke.
>
> Elsewhere, Steve Holden adds:
>
> > And I can say dugapoopoo too. I'll cut out the cheerleading and
> > defensiveness immediately :-)
>
> I'm not holding my breath.  ;o)
>
Well OK then, "rah, rah for our side". I never said that. Happy now? Dammit.

refusing-to-take-myself-too-seriously-and-thanks-for-that-dammit-ly y'rs  -
steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/








From sholden at holdenweb.com  Mon Apr 15 09:11:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 09:11:50 -0400
Subject: explicit is better, BUT... (was Re: PEP 285: Adding a bool type)
References: <mailman.1018755873.29647.python-list@python.org>  <3CBA6595.104ED599@cosc.canterbury.ac.nz> <slrnablepu.5k9.mlh@vier.idi.ntnu.no>
Message-ID: <wmAu8.10927$Ji7.6097@atlpnn01.usenetserver.com>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnablepu.5k9.mlh at vier.idi.ntnu.no...
> In article <3CBA6595.104ED599 at cosc.canterbury.ac.nz>, Greg Ewing wrote:
> >Tim Peters wrote:
> >>
> >> As
> >> originally posted, they were the "20 Pythonic Theses".  They never need
> >> revision, because I only gave 19 -- the 20th (or perhaps the first --
they
> >> weren't numbered) was left open for Guido to fill in.
> >
> >Perhaps it should be "19 Pythonic theses is sufficient."
>
> Or: "Number 20: There is *no*..... Number 20."
>
Or possibly insert as Thesis 12: "The thesis after Thesis 12 shall be
numbered 12a".

regards
 Steve






From narnett at mccmedia.com  Fri Apr 26 18:00:02 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Fri, 26 Apr 2002 15:00:02 -0700
Subject: MySQLdb and DateTime objects
In-Reply-To: <NDBBLAEMOMEHKALAMKKJIEGLKEAA.narnett@mccmedia.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJGEHDKEAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Nick Arnett

[snip]

> What seemed to be the obvious solution was to use
> MySQLdb.times.format_TIMESTAMP() or MySQLdb.times.format_DATE()-- but that
> doesn't work, it returns an error:
>
>   File "C:\Python22\Lib\site-packages\MySQLdb\times.py", lin
> STAMP
>     return d.strftime("%Y-%m-%d %H:%M:%S")
> AttributeError: 'tuple' object has no attribute 'strftime'

I was nailed by a typo -- the above actually *does* work, if one is careful
to use the function on the appropriate list member, rather than the entire
list.

But it still seems inelegant...

Nick





From logiplexsoftware at earthlink.net  Fri Apr 12 14:57:10 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 12 Apr 2002 11:57:10 -0700
Subject: wxImage problem
In-Reply-To: <d5534716.0204121048.3a666b3c@posting.google.com>
References: <d5534716.0204121048.3a666b3c@posting.google.com>
Message-ID: <20020412115710.71c97615.logiplexsoftware@earthlink.net>

On 12 Apr 2002 11:48:50 -0700
Ed wrote:

> how can i put an image into a wxPanel?
> i read on the net that the code is ...
> 
> wxImage_AddHandler(wxGIFHandler())
> gif = wxImage('bitmaps/image.gif', wxBITMAP_TYPE_GIF).ConvertToBitmap()
> wxStaticBitmap(panel, -1, bmp, wxPoint(10, pos),wxSize(bmp.GetWidth
> (),bmp.GetHeight()))

Try something like this (not looking at the docs, so if I make a dumb mistake,
feel free to shoot me):

wxImage_AddHandler(wxGIFHandler())
gif = wxBitmapFromImage(wxImage('bitmaps/image.gif', wxBITMAP_TYPE_GIF))
bmp = wxStaticBitmap(panel, -1, gif, wxPoint(10, pos))

You should probably post questions like this to the wxPython user list (see
www.wxpython.org).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From max at alcyone.com  Tue Apr 23 14:20:48 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 23 Apr 2002 11:20:48 -0700
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <3CC5A600.4080100@alcyone.com>

Petr Prikryl wrote:

> Being quite new to Python -- is there any preference of
> single or double quotes when writing strings?  I am thinking
> about cases when both may be used, not when I need to
> include the alternative inside the string.
> 
> In other words, should I prefer writing:
>     s1 = 'first string'
> or
>     s2 = "second string"
> 
> In yet other words, should I prefer the Pascal look of
> the strings or the C/C++ look of the strings?

It's a style issue, and I'd be highly surprised if there were a 
consensus.  I personally use single quotes for trivial strings (such as 
around invidividual characters, keywords, or identifiers), and double 
quotes for strings that are intended to be English text.  If the text 
spans multiple lines, I use triple double quotes.

> On the other hand, when the Python code is occasionally
> to be converted to C/C++ code, then double quotes
> may be appreciated.

In my opinion, this is a pretty weak reason.  You should never be 
writing code in one language with the hope/intent that someday you're 
going to convert (rather than port) that very code to some _other_ 
language.  Code against the language you're using, and nothing else.

-- 
  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
\__ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I like young girls.  Their stories are shorter.
\__/ Thomas McGuane
/   Kepler's laws / http://www.alcyone.com/max/physics/kepler/
  A proof of Kepler's laws.



From db3l at fitlinxx.com  Thu Apr  4 20:44:23 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 04 Apr 2002 20:44:23 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96>
Message-ID: <u7knmdimg.fsf@ctwd0143.fitlinxx.com>

James Logajan <JamesL at Lugoj.com> writes:

> Then I'd like to complain that neither you nor Guid va Rossu addressed the 
> question I asked in my previous message concerning bools as keys. If a bool 
> is to be treated as an int sometimes and sometimes not, I don't recall 
> seeing anything about whether it will be treated as an int as a dictionary 
> key. If I do:

Guido did cover this explicitly in one of his responses to the bool
threads - see:

http://groups.google.com/groups?selm=mailman.1017860357.6448.python-list%40python.org

which includes among other comments:

    "False and True are numbers, and they are equal (==) to 0 and 1;
     everything else follows from there."

and

    "No you're not.  d[0] and d[False] retrieve the same value, as do
     d[0L], d[0.0], and d[0j]."


--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From billludden at email.msn.com  Wed Apr  3 21:51:18 2002
From: billludden at email.msn.com (billludden)
Date: Wed, 3 Apr 2002 19:51:18 -0700
Subject: Newbie: Alarm at end of program
Message-ID: <es5HyN42BHA.1996@cpimsnntpa03>

Hi,
I have written a script that runs from a few minutes to several hours before
it finishes.  How would I make the computer play a 'beep' to signal that it
is finished?
Thank you,  Bill




From phr-n2002a at nightsong.com  Thu Apr 25 07:45:40 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 04:45:40 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com> <mailman.1019709538.1087.python-list@python.org>
Message-ID: <7xd6wovw63.fsf@ruckus.brouhaha.com>

Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> writes:
> > I'll experiment with some ways of coding HMAC to run fast in Python.
> > The implementation that comes with the 2.2 library is pretty slow.
> 
> Hmm, as the one who submitted this code, I don't see much room for
> improvement:
> 
> - the hexdigest method should really use binascii.hexlify(), but I
>   didn't know about the existence of it at the time
> 
> - I don't see how to optimize _strxor
> 
> But then, I never benchmarked the module.

_strxor should be a lot faster if it uses string.translate (or xor's
using the array module) rather than doing what it's doing.  I appended
my version to another message in this thread.


From phd at phd.pp.ru  Wed Apr  3 04:17:22 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 3 Apr 2002 13:17:22 +0400
Subject: How to check whether email /domain exists
In-Reply-To: <3CAADA7B.25443.509AE1@localhost>; from printers@sendme.cz on Wed, Apr 03, 2002 at 10:33:31AM +0200
References: <3CAADA7B.25443.509AE1@localhost>
Message-ID: <20020403131722.I23871@phd.pp.ru>

On Wed, Apr 03, 2002 at 10:33:31AM +0200, A wrote:
> What is the simplest way how to check 
>  whether a given email address( or domain for this email) exists?

   Send a mail and ask the user to reply or click an URL.

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



From kybic at ieee.org  Wed Apr  3 10:42:49 2002
From: kybic at ieee.org (Jan Kybic)
Date: 03 Apr 2002 17:42:49 +0200
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <15531.7864.698585.485436@12-248-41-177.client.attbi.com>
References: <mailman.1017795845.3464.python-list@python.org>
	<r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
	<lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
	<lbh8z84nake.fsf@xihu.inria.fr>
	<15531.7864.698585.485436@12-248-41-177.client.attbi.com>
Message-ID: <lbh7knolreu.fsf@xihu.inria.fr>

The following message is a courtesy copy of an article
that has been posted to comp.lang.python as well.

> You're thinking of weave.inline().  I was talking about weave.blitz().
> Reread the last statement in the text I quoted, then take a look here:
> 
>     http://www.scipy.org/site_content/weave/tutorial.html#Blitz

OK, I agree this is a nice piece of work. However, this is still
limited to relatively simple, one line expressions. No way of adding
more complex calculations (with ifs, loops,...). 

J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/




From chris_burn at NO-SPAM.bigfoot.com  Thu Apr  4 21:08:25 2002
From: chris_burn at NO-SPAM.bigfoot.com (Chris Burn)
Date: Fri, 05 Apr 2002 12:08:25 +1000
Subject: Can someone comment on my code?
Message-ID: <3CAD0719.6050707@NO-SPAM.bigfoot.com>

Hi, I've been writing perl, shell, c programs for a while but I thought 
I should move on to python.

So this is my first real venture in python. Basically I work on an app 
that auto saves to your home dir every 10 mins, but I keep my working 
files in a dir elsewhere. So this script is intended to be run after a 
crash, it searches for the last autosave and the last normal save in the 
current dir. It then copies the autosave to the current dir with _rec 
stuck on the end.

Anyway, it works, but I'm new to error handling in python and it would 
be good to know if I'm handling errors correctly and also if there are 
easier ways to do things or conventions I should be keeping to? It seems 
to be a bit long to me.

Oh it needs to run on python 1.5.2

Thanks
Chris

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: autoShakeRecover.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020405/f044e9eb/attachment.ksh>

From mgilfix at eecs.tufts.edu  Sun Apr 28 15:28:00 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sun, 28 Apr 2002 15:28:00 -0400
Subject: portable fortune program
In-Reply-To: <aahhsg$a1k$2@nyheter.chalmers.se>; from jacob@boris.cd.chalmers.se.cd.chalmers.se on Sun, Apr 28, 2002 at 07:17:36PM +0000
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com> <aahhsg$a1k$2@nyheter.chalmers.se>
Message-ID: <20020428152800.A22480@eecs.tufts.edu>

  Or other simple suggestion:

    1) Have 1 index file that just contains the offsets of the fortunes
       in the other index file and has each entry taking up a fixed amount
       of space.
    2) Get size of the file
    3) Get a random index from file_size / index_size
    4) Seek to fortune in actual file

  That will be really fast as will have no skewing.

                   -- Mike

On Sun, Apr 28 @ 19:17, Jacob Hallen wrote:
> Simple suggestion by me added:
> 
> 1. Get the size of the file.
> 2. Get a random number 0 - size-1.
> 3. Index to this point in the file.
> 4. Iterate back to the nearest previous '%'.
> 5. Print fortune from there.
> 
> This will of course skew the probability of getting printed towards longer
> cookies, but nobody will notice that.

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From sholden at holdenweb.com  Tue Apr 23 10:14:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 10:14:16 -0400
Subject: check, if an url exists
References: <2eea6430.0204230603.420cf7ee@posting.google.com>
Message-ID: <c1ex8.94935$T%5.80990@atlpnn01.usenetserver.com>

If you have a Tools subdirectory in your Python installation, there's a
webchecker subdirectory that wil give you a good handle on this. If not, you
can download the source and extract it quite easily.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/


"Max" <max.wenzel at web.de> wrote in message
news:2eea6430.0204230603.420cf7ee at posting.google.com...
> Halli-Hallo
>
> I'm looking for a script, that checks an url, if exists or not.
> Can anyone help me ??
>
> Thanks a lot and many greetings,
> Max





From buzzard at urubu.freeserve.co.uk  Mon Apr 15 19:55:27 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 16 Apr 2002 00:55:27 +0100
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <MtIu8.18713$b62.536574@news1.tin.it>
Message-ID: <a9fpff$gfl$1@newsg1.svr.pol.co.uk>

Thanks Alex (and Christophe).  Will try these things tomorrow.  Sorry about
using 'dict' for a dictionary; and after having been bitten by calling a
list 'list' before now.:-)  Cheers.

Duncan




From max at malva.com.uaREMOVE.IT  Mon Apr 15 06:57:49 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Mon, 15 Apr 2002 13:57:49 +0300
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com>
Message-ID: <dnbe9a.jfl.ln@cvs>

 Ken Peek wrote:

> 5) These two types (int and long) _CAN_ be unified for
> _ALL_ operators and methods-- I know they can, because the
> language known as "Ruby" does a very good job of it.  In
> Ruby, there is simply no difference (at least visibly) to
> the programmer _at_all_.  The 2 types are automatically
> converted back and forth without the knowledge or consent
> of the programmer.  If Ruby does it, so can Python.

Not quite:
irb(main):001:0> 1222.class
Fixnum
irb(main):002:0> 12232333333333332.class
Bignum


For me, Common Lisp's numeric types look like a good source of
inspiration.


-- 
BOFH excuse #71:
The file system is full of it


From grante at visi.com  Sat Apr  6 10:08:50 2002
From: grante at visi.com (Grant Edwards)
Date: Sat, 06 Apr 2002 15:08:50 GMT
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org>
Message-ID: <slrnaau3t2.1si.grante@tuxtop.visi.com>

In article <mailman.1018081037.4211.python-list at python.org>, David McInnis wrote:
> I need to alter the document properties of Word DOCS that our users
> upload to our Web site.

os.unlink(filename)

> Does anyone know how to do this?

On windows, you can use COM interface to MS-Office/Word. Under
Unix, you're pretty much screwed.  You could ask your users to
upload something less evil:  SGML, HTML, RTF, ???.

> (I am python installed on a Linux box).

AI has come a long way!  Is this some sort of Turing test thread?

-- 
Grant Edwards                   grante             Yow!  HELLO KITTY gang
                                  at               terrorizes town, family
                               visi.com            STICKERED to death!


From phd at phd.pp.ru  Mon Apr 15 07:18:39 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 15 Apr 2002 15:18:39 +0400
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
In-Reply-To: <7xelhhmq5e.fsf@ruckus.brouhaha.com>; from phr-n2002a@nightsong.com on Sun, Apr 14, 2002 at 11:28:45PM -0700
References: <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp> <7xelhhmq5e.fsf@ruckus.brouhaha.com>
Message-ID: <20020415151839.A23775@phd.pp.ru>

On Sun, Apr 14, 2002 at 11:28:45PM -0700, Paul Rubin wrote:
> The non-GPL-ness of Python is a main reason I don't have much interest
> in doing significant development work on the Python implementation

   It is not like this with me - I sent, is sending and will send those
patches I can. But I certainly would like to see Python GPL'd.

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




From brueckd at tbye.com  Thu Apr 11 12:13:42 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 11 Apr 2002 09:13:42 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <slrnaba052.efc.mlh@vier.idi.ntnu.no>
Message-ID: <Pine.LNX.4.33.0204110859270.25928-100000@ns0.tbye.com>

On Thu, 11 Apr 2002, Magnus Lie Hetland wrote:

> It seems this has been brought up before, but I could only find two
> posts about it (about a year old). The question is: Could we have
> tuple comprehensions?
>
> The argument I saw used earlier was based on the author's
> preconceptions about tuples being primarily "records" and not
> "containers". The only defining quality I see to tuples as opposed to
> lists is that they are immutable. (Yes, I see the "recordness", but I
> don't think it's an argument against tuple comprehension.)

Maybe not directly, but it might argue against them simply because tuple
comprehensions wouldn't be used as much. They'd be useful when you're
using tuples as containers and not records, or when all the record members
are the same "type" to your application, e.g.:

coord = (x,y,z)
scaled = (f/2 for f in coord)

I use list comprehensions all the time, but I have trouble coming up with
too many cases where tuple comprehensions would be useful. Maybe I'm
missing a large set of potential uses?

> As far as I can see, the PEP on dict comprehension is still open... I
> think both dictionary comprehension and tuple comprehension seem like
> natural parallels to list comprehension. But then I would, wouldn't I?

I'd *love* to have dictionary comprehensions!

I guess the reason list and dictionary comprehensions make sense to me and
tuple comprehensions do not is because for lists and dictionaries they are
direct replacements for the initialize-to-empty-then-loop-to-populate
task, for which there is no tuple equivalent:

d = empty list or dict
for item in someInputSet:
    do something to item and add it to d

 -Dave





From gcordova at hebmex.com  Wed Apr 24 12:52:40 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 24 Apr 2002 11:52:40 -0500
Subject: python and very large data sets???
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339535CF@mail.hebmex.com>

Whoooeeee, sounds interesting.

Hmmmm.. actually, it seems to me that the information is
vaguely relational, at least. If that's the case,
wouldn't it be better to load the flat files into
some database tables, manipulate it using normal
SQL statements, and then generate any resulting
files from there?

-gustavo




From matt at mondoinfo.com  Mon Apr  8 23:07:37 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Tue, 09 Apr 2002 03:07:37 GMT
Subject: Python server from inetd?
References: <iml4bukknqf9nqe13h7dp0qco3qj2r86rt@4ax.com>
Message-ID: <ZVss8.21014$vm6.3263812@ruti.visi.com>

On Mon, 08 Apr 2002 19:57:37 -0700, Tim Roberts <timr at probo.com> wrote:

>I have a quick little socket server that works OK as a continuously running
>app, listening for and handling connections.

[. . .]

>Thus, I tried to port it to inetd.

[. . .]

>I thought all I needed to do was basically:

[. . .]

Tim,
You have to do even less than that. inetd (I generally use xinetd but
I'm quite sure they're the same in this respect) does all the socket
work for you. Your server only has to read from stdin and write to
stdout.

Regards,
Matt


From fperez528 at yahoo.com  Sun Apr 14 14:52:11 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sun, 14 Apr 2002 12:52:11 -0600
Subject: Weak references and tracebacks - NEVER MIND
References: <a9albr$3ha$1@peabody.colorado.edu> <a9altu$3rp$1@peabody.colorado.edu> <m3pu12ws5v.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9cj6h$6jd$1@peabody.colorado.edu>

Martin v. Loewis wrote:

> Fernando P?rez <fperez528 at yahoo.com> writes:
> 
>> > Is it ok to use weak references to hold tracebacks without generating
>> > circular references? From the documentation:
>> 
>> Never mind, you simply can't make those weak refs. Sorry (tested with other
>> types. dumb me).
> 
> Even if it was possible, it would be less useful than you think: Most
> likely, your weak reference would be the *only* reference to the
> traceback object, which in turns would cause the traceback to die, and
> your weak reference go stale immediately.

You're 100% right. And as usual in such cases, a few minutes thought (which I 
should have done before hand :) showed me what the proper solution was.

Thanks,

Fernando.


From flz at cfdrc.com  Tue Apr  2 10:55:34 2002
From: flz at cfdrc.com (gother)
Date: 2 Apr 2002 07:55:34 -0800
Subject: Initialization problem of shared library in HP-UX11
References: <7aafbf19.0204010929.62d17d6a@posting.google.com> <m37knr71pl.fsf@mira.informatik.hu-berlin.de> <7aafbf19.0204011414.25acdd0@posting.google.com> <m3d6xjxc7h.fsf@mira.informatik.hu-berlin.de>
Message-ID: <7aafbf19.0204020755.7b8bd91f@posting.google.com>

Thanks again for your reply.

I used gcc compiler. I tried once using ld to build the shared lib and
found that even the executable main linking to this shared lib doesn't
do the initialization properly.

Yes, I definely need a HP developer expert.

Regards,

Gother

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3d6xjxc7h.fsf at mira.informatik.hu-berlin.de>...
> flz at cfdrc.com (gother) writes:
> 
> > I know windows do the initialization when it load the shared lib. In
> > HP, when the shared lib links to an main function, it do the same
> > thing. But somehow HP-UX11 failed to do it inside the python which
> > called shl_load function. So I guess inside the executable the shared
> > lib is loaded with different system function call instead of shl_load
> > or different arguments of shl_load. If anybody be familar with the
> > internal loading process of HP, he must have the answer.
> 
> Perhaps the answer is: It just can't work; it is a known deficiency of
> HP-UX. Then you are back to my suggestion. I don't know whether this
> would be a correct answer.
> 
> Now that you seem to want understanding this, we should debug it it
> step by step. 
> 
> First question: What is the command line that you use to link the
> extension module?
> 
> If it starts with 'ld', that could explain it. Many systems require
> special options to the linker (in particular addition of object
> files) to support initialization of shared libraries.
> 
> If it starts with 'gcc', that still could explain it. Try the system
> compiler; it might be a bug in gcc.
> 
> If it is the bundled C compiler, this definitely explains it.
> 
> If it is the unbundled compiler, next question:
> 
> What is the documentation for all the command line option you are
> giving? What other options related to shared libraries are mentioned
> in the documentation of ld(1) and c89(1)?
> 
> And so on. It is unlikely that you find a lot of HP-UX expertise in
> this group - I recommend to contact HP.
> 
> Regards,
> Martin


From marklists at mceahern.com  Mon Apr  8 11:54:39 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 8 Apr 2002 10:54:39 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <roy-995FAF.10470208042002@news1.panix.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEMHKJAA.marklists@mceahern.com>

[Roy Smith]
> If he wanted to be truly elegant and economical, he'd have simply written
> "<ploink>".

But that's not well-formed xml.  ;-)

// m





From gerhard at bigfoot.de  Sat Apr 27 02:49:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 27 Apr 2002 08:49:34 +0200
Subject: Compiled socket.pyd available for Win32 Python 2.2.1?
In-Reply-To: <3CCA3D6F.7020506@xs4all_removethisalso.nl>
References: <3CCA3D6F.7020506@xs4all_removethisalso.nl>
Message-ID: <20020427064934.GB2032@lilith.my-fqdn.de>

* iwk <iwk_removethis at xs4all_removethisalso.nl> [2002-04-27 07:55 +0200]:
> Hi there,
> 
> Does any benevolent soul have socket.pyd compiled for Win32 Python 2.2.1 
> available for download? I need SSL support and the default Win install 
> doesn't have ssl support compiled (why? US export regulations?) in. I 
> don't have access to Visual C and lower versions of Python or Cygwin are 
> unfortunaltely not an option (stalinist system management :-)
 
You can use the one from my homepage (see sig). It's from Python 2.2,
but I haven't heard of any SSL fixes for 2.2.1. Python's with different
micro release numbers (x.y.z1 <-> x.y.z2) are binary compatible, even on
Windows.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From ods at fep.ru  Thu Apr  4 10:06:11 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 4 Apr 2002 19:06:11 +0400 (MSD)
Subject: CGIHTTPServer and custom headers
In-Reply-To: <1017930044.10710.7.camel@tp6>
Message-ID: <Pine.LNX.4.44L.0204041905150.13708-100000@ods.pp.ru>

Have you tried Medusa
(http://www.amk.ca/files/python/medusa-0.5.1.tar.gz)?

On 4 Apr 2002, Graham Ashton wrote:

GA> I tried to use the CGIHTTPServer module to implement a
GA> simple server for
GA> running a simple CGI that I'm developing. I thought "Great,
GA> this can be
GA> a nice stand alone little test environment, without the need
GA> for
GA> apache".
[...]
GA> Can anybody recommend the most straightforward approach? My
GA> app is dead
GA> simple and doesn't need speed or scalability. I'm currently
GA> browsing the
GA> twisted docs, and the asyncore docs. Is there an easier way?





From aleax at aleax.it  Wed Apr 10 10:32:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 14:32:05 GMT
Subject: Working with untrusted coders
References: <9739136773arocas@alumni.uv.es>
Message-ID: <F1Ys8.3143$b62.77608@news1.tin.it>

Arnau Rossell? Castell? wrote:

> Hi,
> I want to develop a MUD, and i wat thinking that python would be a good
> language to make it, but such a MUD would have many coders whom i don't
> trust completely, and i'd like to restrict his access. Python security
> about private elements and others is merely a "don't step on the grass"
> sign, everybody can change private functions and data in classes and even
> modules. I'd like to know if there is some way of really securing python,
> so parts of the code can't be overriden by the coders. Has anybody tried

Have a look at modules rexec and Bastion, and more generally at
"Restricted Execution" in Python -- that's what it's for.


Alex



From starx at pacbell.net  Wed Apr 17 14:18:45 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 17 Apr 2002 18:18:45 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com>
Message-ID: <Xns91F3733E2F78BRASXnewsDFE1@130.133.1.4>

Greg Weeks || Wed 17 Apr 2002 09:15:58a:

> There should have been an operator in Python that says when objects
> are conceptually the same.  For the moment, call it "eq".  The
> following would hold:
> 
>     3 eq 3
>     not (3 eq 4)
>     a eq a          # PROVIDED a HAS A VALUE
> [etc, etc]

Well, if a is defined, a has a value. So this is either true or a 
NameError (?). Or do you mean if a is None, a eq a should be false?


-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From chene at cs.queensu.ca  Mon Apr  8 21:15:59 2002
From: chene at cs.queensu.ca (Elvis Chen)
Date: Mon, 8 Apr 2002 21:15:59 -0400
Subject: Newbie tkinter question:  Scale release event?
Message-ID: <Pine.SOL.4.33.0204082107020.21145-100000@radius>

Greetings,

I'm a relative newbit to Python/tkinter, so please go easy with my
question.  I have search through google/python.org, but havn't found the
answer to my question yet:

I'm trying to do some OpenGL programming with pythong/tkinter.  I would
like to use Scale widget of tkinter to control the movement/rotation of a
camera.  That I would like to do is, upon the release of the mouse button
off a Scale widget, the value of the Scale widget is used via a callback
mechanism to update camera orientation.  After the update, the scale is
reset back to zero.

For example, I currently have:

def rotx(a):
    global rotx_scale  # a Scale widget
    angle_in = rotx_scale.get()
    ren.GetActiveCamera().Elevation( angle_in )
    rotx_scale.set( 0 )

and in the main:

    # Rotation about X-axis
    global rotx_scale
    rotx_scale = Scale( root, from_=-180, to=180, resolution=5,
                        command=rotx )

The problem of my code is, the callback function is called whenever the
mouse is "paused" on the screen.  I would like to change it so that the
callback is only called when the mouse button is released.

any help is very much appreciated,

ECC



From brobbins333 at shaw.ca  Thu Apr 11 21:43:37 2002
From: brobbins333 at shaw.ca (brobbins333 at shaw.ca)
Date: Fri, 12 Apr 2002 01:43:37 GMT
Subject: File Size Mystery??
References: <3cb6358d.33229187@news> <a95dri$mc1$0@216.39.172.122>
Message-ID: <3cb63b9c.34780328@news>

Five minutes after I posted the message I figured it out, but thanks
anyway. Old VB habits are hard to break.

BR 


On 12 Apr 2002 01:38:58 GMT, bokr at oz.net (Bengt Richter) wrote:

>On Fri, 12 Apr 2002 01:18:56 GMT, brobbins333 at shaw.ca wrote:
>
>>this is foofile.txt (created in Windows Notepad):
>>ur75ht8495u7475ht89302oskv476830rj
> 01234567890123456789012345678901234
>           1         2         3
>>
>>and here is what Python tells me about it:
>>
>>
>>>>> print os.path.getsize('foofile.txt')
>>34
>>>>> print len('foofile.txt')
>                012345678901
>                          1
>>11
>>
>>
>>os.path.getsize() returns the correct size of the file, 
>>but len() does not. Why is there a discrepancy?
>>
>;-)
>
>Regards,
>Bengt Richter



From gerhard at bigfoot.de  Thu Apr 25 21:10:13 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 26 Apr 2002 01:10:13 GMT
Subject: newbie question about cygwin and wxpython
References: <d0074b71.0204251145.23d0607b@posting.google.com>
Message-ID: <slrnacha7k.en.gerhard@lilith.my-fqdn.de>

BoB W wrote in comp.lang.python:
> I have loaded cygwin 1.3.10, which installs Python 2.2.  I have also
> loaded wxpython 2.3.2.1 hybrid installer for windows to
> c:\cygwin\lib\python2.2.

Don't mix Python extensions for native Python with Cygwin extensions.
They're not compatible. Why do you want to use Cygwin's Python instead
of the normal win32 Python, anyway?  It's main purpose is, as with all
Cygwin programs, to be able to run Unixoid Python extensions on
Windows. Do you really need to do this?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From aahz at pythoncraft.com  Mon Apr  1 13:44:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 13:44:02 -0500
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <m3adspm1kl.fsf@mira.informatik.hu-berlin.de> <3CA79401.CFA41B65@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain>
Message-ID: <a8a9pi$4ha$1@panix1.panix.com>

In article <slrnaaha69.nn5.kzaragoza at localhost.localdomain>,
Kris J. Zaragoza <kzaragoza at attbi.com> wrote:
>
>Do you intend to eventually (NOTE: I say eventually, I know it's not
>immediate) modify this to read that only the value True is true and
>the value False is false?  If this is the case, I believe this adds
>needless complexity to code whereas (in a simple case)
>
>if foo:
>    ...do whatever...

I can't find a reference off-hand, but I think it's pretty clear that
this will remain a standard Python idiom forever.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From DeepBleu at DeepBleu.org  Thu Apr 11 09:50:35 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Thu, 11 Apr 2002 08:50:35 -0500
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com> <3CB2FE25.86A37EA1@imk.fraunhofer.de> <ub6dji5plutpc5@corp.supernews.com> <SKYs8.19881$GM.6702@atlpnn01.usenetserver.com>
Message-ID: <ubb56d6fvmpjf0@corp.supernews.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:SKYs8.19881$GM.6702 at atlpnn01.usenetserver.com...
> "DeepBleu" <DeepBleu at DeepBleu.org> wrote in message
> news:ub6dji5plutpc5 at corp.supernews.com...
> >
> > "Joachim Kaeber" <joachim.kaeber at imk.fraunhofer.de> wrote in message
> > news:3CB2FE25.86A37EA1 at imk.fraunhofer.de...
> > > DeepBleu wrote:
> > >
> > > > Yes, I am using zipfile (from zipfile import ZipFile).  Sorry I
stated
> > zlib.
> > > > Still with one text file in particular, WinZip reports "missing 2
> bytes"
> > > > DeepBleu
> > >
> > > is the script run on Windows? If yes, is the zipfile opened
> > > in binary mode?
> > >
> > The script is running on Sun Solaris with Python 2.1
> >
> > >>>from zipfile import ZipFile
> > >>>zip_archive = ZipFile("text1.zip", "w", 8)
> > >>>zip_archive.write(PathToTextFile, os.path.split(PathToTextFile)[1])
> > >>>zip_archive.close()
> >
> > Before starting this process, I am running Unix2DOS on the text file:
> > Unix2Dos TextFile TextFile
> >
> > Reason is that the zip file will be opened later in windows (using
WinZip
> > for example).
> > Hope this answers your question,
>
> Do you close the zipfile after you've finished writing it?
>
Do you mean text1.zip (from the code above) as the zipfile to be closed?
No, I am not!  I thought it was enough to close the ZipFile instance
(zip_archive).  Since text1.zip makes its appearance (first used) in:
>>>zip_archive = ZipFile("text1.zip", "w", 8)
How do I close the text1.zip (if that is the one you are refering to)?
DeepBleu





From phr-n2002a at nightsong.com  Wed Apr 24 18:41:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 15:41:24 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com>
Message-ID: <7x4ri0oh2j.fsf@ruckus.brouhaha.com>

bryanjugglercryptographer at yahoo.com (Bryan Olson) writes:
> Not so good.  The attacker only has to guess one block to break
> all subsequent blocks.  For a demo, put the code below into 
> the p2.py file

ARGGH!  I knew about that before too, but forgot about it while writing
the code.  I will replace SHA(block) with SHA(key + block) and release
a new version tonight, with the MAC improved too.  Sound ok?

Thanks.


From gashton at cmedltd.com  Wed Apr 17 09:18:23 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 17 Apr 2002 14:18:23 +0100
Subject: strftime weirdness
Message-ID: <1019049503.653.166.camel@tp6>

I clearly don't understand timezones. Can somebody please explain why I
get +0000 for both these time tuples, with the first two calls to
strftime() ?

>>> p
(1970, 1, 1, 0, 0, 0, 3, 1, 0)
>>> q
(1970, 1, 1, 0, 0, 0, 3, 1, 1)
>>> time.strftime("%d %b %Y %H:%M:%S %z", p)
'01 Jan 1970 00:00:00 +0000'
>>> time.strftime("%d %b %Y %H:%M:%S %z", q)
'01 Jan 1970 00:00:00 +0000'
>>> time.strftime("%d %b %Y %H:%M:%S %Z", q)
'01 Jan 1970 00:00:00 BST'
>>> time.strftime("%d %b %Y %H:%M:%S %Z", p)
'01 Jan 1970 00:00:00 GMT'

-- 
Graham Ashton





From boud at valdyas.org  Thu Apr 25 06:24:45 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 25 Apr 2002 12:24:45 +0200
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <c76ff6fc.0204241734.40e57406@posting.google.com> <3cc7967a$0$37911$e4fe514c@dreader3.news.xs4all.nl> <c76ff6fc.0204250212.af07541@posting.google.com>
Message-ID: <3cc7d9dc$0$37919$e4fe514c@dreader3.news.xs4all.nl>

John Machin wrote:


> Contrary to popular mythology, Chinese words can have more than one
> syllable. As the OP said:
> 
I know that! I just didn't know how big-5 encoded characters -- I
use Unicode exclusively for my work with Chinese. I just didn't read
the original post closely enough...

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


From jb at cascade-sys.com  Mon Apr 29 17:15:21 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 14:15:21 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <LNBBLJKPBEHFEDALKOLCAEOGPBAA.tim.one@comcast.net>
Message-ID: <3CCDB7E9.2FCC06BB@cascade-sys.com>


Tim Peters wrote:

> Note that Guido's comment on Python-Dev was:
>
>     I strongly suggest sticking to the status quo.  There are more
>     important fish to fry.

Smart man.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From mgilfix at eecs.tufts.edu  Thu Apr 25 16:00:23 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 25 Apr 2002 16:00:23 -0400
Subject: ftplib and attributes
In-Reply-To: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>; from giannivolontosky@TOGLIyahoo.it on Thu, Apr 25, 2002 at 07:39:32PM +0000
References: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>
Message-ID: <20020425160022.A6555@eecs.tufts.edu>

  You get the attribute of files when you do a listing. For example:

    import ftplib
    ftp = ftplib.FTP ('ftp.somewhere.org')
    ftp.login ()
    listing = ftp.retrlines ('LIST')

    # Print your listing to look at it
    for i in listing: print i

  To get your attribute, first split the fields, for example:
  import string
  entry = listing[0]
  fields = string.splitfields ()

  You attributes will be in field[0]. You can then do any other
mining  you want. Hope that helps.

                       -- Mike

On Thu, Apr 25 @ 19:39, LoGan wrote:
> hi, im new bie and im trying to use ftplib, but my problem is to get the
> attributes of the files, how can i do ?
> LoGan
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (giannivolontosky)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From gimbo at ftech.net  Fri Apr  5 05:01:07 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 5 Apr 2002 11:01:07 +0100
Subject: duelling pythons
In-Reply-To: <ulmc2sd6w.fsf@python.net>
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de> <3cab55f4.61190667@news.laplaza.org> <a8fn61$s1nk7$1@ID-41653.news.dfncis.de> <lky9g4o3z7.fsf@elios.maths.bris.ac.uk> <mailman.1017914955.11127.python-list@python.org> <ulmc2sd6w.fsf@python.net>
Message-ID: <20020405100107.GC16724@andy.tynant.ftech.net>

On Fri, Apr 05, 2002 at 09:28:02AM +0000, Michael Hudson wrote:
> Andy Gimblett <gimbo at ftech.net> writes:
> 
> > How can distutils help with the problem of what to put on the first
> > line of my scripts?
> 
> Because if you go
> 
> python setup.py install_scripts
> 
> said scripts get their first lines munged to point to the python used
> to install them.

Oooh!  NICE!

Thanks for that!  :-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From max at alcyone.com  Wed Apr 24 16:24:10 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 24 Apr 2002 13:24:10 -0700
Subject: variable substitution
References: <e2e9e807.0204240519.688cf3cb@posting.google.com>
Message-ID: <3CC7146A.B7C00010@alcyone.com>

Hugh wrote:

> How do I get the %(variable)s thing to work. I need to substitute a
> variable into a print statement so I can send links with sessionId's
> attached back to the browser. I'm trying this at the moment, and It
> says invalid sysntax.
> print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
> Type</a>'

	print 'blah blah %(sessionId)s" % {'sessionId': 'ghlahglahg'}

If you really want everything in your locals stuffed in there, then call

	print 'blah blah %(sessionId)s" % locals()

although you should beware of security issues if the formatting string
is coming from somewhere else.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From andrew.henshaw at gtri.gatech.edu  Thu Apr 25 22:30:01 2002
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Thu, 25 Apr 2002 22:30:01 -0400
Subject: Stackless goes Limbo
References: <mailman.1019662503.1913.python-list@python.org>
Message-ID: <uchetjbtqegmb8@corp.supernews.com>

Christian Tismer wrote:

> Hi all,
> 
> after some time of consideration,
> I have decided to move the interface to Stackless
> Python's multitasking towards the ideas behind
> the Limbo Language.
> 
> Besides the tasklets, there are "channel" objects,
> which allow for communication between tasklets.
> This approach is simple and powerful. Blocking of
> tasklets takes place on writes when no tasklet
> is listening, or on reads, when no data is available.
> 
> There is some nice introduction to Limbo at
> http://www.vitanuova.com/inferno/papers/descent.html
> written by Brian W. Kernighan, one of the inventors
> of the C language.
> 
> I hope to have a prototype ready by end of April.
> 
> See the stackless homepage for more news,
> including the new sponsorship honor page at:
> 
> <http://www.stackless.com/>
> 
While you're at it, you might want to look at Occam.  In my opinion, no 
language handles channels as nicely as Occam.  [Sigh] PAR and ALT would 
look so nice as Python keywords...

Good luck!

Andrew Henshaw



From krissepu at vip.fi  Mon Apr  1 12:51:12 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Mon, 01 Apr 2002 20:51:12 +0300
Subject: How to close wxGrid located inside wxFrame
Message-ID: <3CA89E10.AF2C93B3@vip.fi>

Hi,

    how should I define the OnExit -method in MainFrame -class below
    so that wxWindows would not leak memory, when "Exit" is selected
    from the menu. I should somehow close the grid before the frame, but
how ?

------------------xxxx-------------------

from wxPython.wx import *
from wxPython.grid import *
from wxPython.lib.mixins.grid import wxGridAutoEditMixin
import sys, pprint

ID_EXIT  = 102

class PnGrid(wxGrid):
    def __init__(self, parent):
        wxGrid.__init__(self, parent, -1)
        self.moveTo = None
        EVT_IDLE(self, self.OnIdle)
        self.CreateGrid(25, 25)

        .... lots of code...copied from wxWindows demofile
SimpleGrid.py...

class MainFrame(wxFrame):
    """ Creates and initializes the main window"""

    def __init__(self, parent, id, title, style):
        # First, call the base class' __init__ method to create the
frame
        wxFrame.__init__(self, parent, id, title, wxPoint(100, 100),
wxSize(640, 480), style)

        self.CreateStatusBar()
        self.SetStatusText("This is the statusbar")

        # Make menu with separate MakeMenus -method
        menu = self.MakeMenus()
        self.SetMenuBar(menu)

        # Associate some events with methods of this class
        EVT_MENU(self, ID_EXIT,  self.OnExit)

        # Create grid
        pn_grid = PnGrid(self)

    def MakeMenus(self):
        """Creates the menus of the main window"""

        # Filemenu contents
        filemenu = wxMenu()
        filemenu.Append(ID_EXIT, "Exit", "Terminate the program")

        # Collect all menus
        main = wxMenuBar()
        main.Append(filemenu, "File")
        return main

    def OnExit(self, event):
        # Close program
        self.Close(true)


class MainApp(wxApp):
    # wxWindows calls this method to initialize the application
    def OnInit(self):

    # Create an instance of our customized Frame class
        frame = MainFrame(NULL, -1, " PnGrid", wxSTAY_ON_TOP |
wxCAPTION)
        frame.Show(true)

    # Tell wxWindows that this is our main window
        self.SetTopWindow(frame)
        return true

if __name__ == "__main__":
    app = MainApp(0)
    app.MainLoop()

---------------------------xxxx----------------

-pekka-




From robin at jessikat.fsnet.co.uk  Sat Apr 27 13:28:58 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 27 Apr 2002 18:28:58 +0100
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>
 <just-9E8624.12281026042002@news1.xs4all.nl>
 <ucjicvlrf7djd3@news.supernews.com>
Message-ID: <q5Z7JWAa$ty8EwuL@jessikat.fsnet.co.uk>

In article <ucjicvlrf7djd3 at news.supernews.com>, John Roth
<johnroth at ameritech.net> writes
>
>"Just van Rossum" <just at xs4all.nl> wrote in message
>news:just-9E8624.12281026042002 at news1.xs4all.nl...
>> In article <C2cBaXANXSy8Ewbb at jessikat.demon.co.uk>,
......
>
>This doesn't appear to be the same bug. Robin's example
>has two functions which are identical, other than the
>substitution of 'x' for 'z' in the definition of class A.
>
>The discussion in the bug report indicates that the 'x'
>following the '=' should be compiled using the 'LOAD_NAME'
>op code, which implements the old rules. If it in fact did this,
>I would expect the first example to return 4, not 2,
>because 'LOAD_NAME' is supposed to ignore the
>parameter.
>
>Clearly, it isn't. At a guess, I'd think the problem is that
>the compiler is trying to make 'z' a local variable of the
>def, so it's using a different op code in the first case than
>in the second.
>

well I see this from dis the x=x case seems to ignore that line totally,
the z=x case uses LOAD_CLOSURE. I guess someone is guessing that x=x is
a NOOP and removing it as an optimisation, but perhaps I'm totally
confused as well. Probably I'm misunderstanding the scope somehow, but
in this case  x = x is supposed to be equivalent to A.x = x. 

if it's treated as A.x = A.x then I should be warned that A.x is
referred to before assignment. 

>>> dis.dis(aaa.bingo)
          0 SET_LINENO               2

          3 SET_LINENO               3
          6 LOAD_CONST               1 ('A')
          9 BUILD_TUPLE              0
         12 LOAD_CLOSURE             0 (x)
         15 LOAD_CONST               2 (<code object A at 010EE328, file
"aaa.py", line 3>)
         18 MAKE_CLOSURE             0
         21 CALL_FUNCTION            0
         24 BUILD_CLASS         
         25 STORE_FAST               1 (A)

         28 SET_LINENO               7
         31 LOAD_FAST                1 (A)
         34 CALL_FUNCTION            0
         37 LOAD_ATTR                1 (get)
         40 CALL_FUNCTION            0
         43 RETURN_VALUE        
         44 LOAD_CONST               0 (None)
         47 RETURN_VALUE        
>>> dis.dis(aaa.bango)
          0 SET_LINENO              10

          3 SET_LINENO              11
          6 LOAD_CONST               1 ('A')
          9 BUILD_TUPLE              0
         12 LOAD_CONST               2 (<code object A at 010F2398, file
"aaa.py", line 11>)
         15 MAKE_FUNCTION            0
         18 CALL_FUNCTION            0
         21 BUILD_CLASS         
         22 STORE_FAST               1 (A)

         25 SET_LINENO              15
         28 LOAD_FAST                1 (A)
         31 CALL_FUNCTION            0
         34 LOAD_ATTR                1 (get)
         37 CALL_FUNCTION            0
         40 RETURN_VALUE        
         41 LOAD_CONST               0 (None)
         44 RETURN_VALUE        
>>> 
>Of course, I could be totally confused, as usual.
>
>John Roth
>
>

-- 
Robin Becker


From aahz at pythoncraft.com  Wed Apr 10 00:55:14 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 00:55:14 -0400
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54>
Message-ID: <a90gji$9mq$1@panix1.panix.com>

In article <zHOs8.201693$Yv2.65602 at rwcrnsc54>,
Emile van Sebille <emile at fenx.com> wrote:
>Kalle Svensson
>>
>> "import this", IIRC.
>
>Oooh... that's almost as much fun as saying 'xyzzy' the first time...
>;-)
>
>Plugh!-ly y'rs

kill dragon
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From ykingma at accessforall.nl  Mon Apr  8 18:23:49 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Mon, 08 Apr 2002 23:23:49 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAE24A2.7ECBB78@accessforall.nl> <onpr8.64923$S52.2260694@news2.tin.it> <3CAF91A9.EE3607CC@accessforall.nl> <JOZr8.69884$S52.2468332@news2.tin.it>
Message-ID: <3CB2185B.2AFD549C@accessforall.nl>

Alex,

> > The language spec does not specify the type of the index or key
> > argument to __getitem__.
> 
> It does for sequences (int and slice only) and for dictionaries (any
> hashable) but not for other mappings (each mapping define its
> own rules: something like "only a string is accepted" is quite
> common in various mappings implemented in the standard
> library, for example).

I concluded that the type of the key argument is not specified from:
http://www.python.org/doc/current/ref/sequence-types.html :

__getitem__(self, key) 

 Called to implement evaluation of self[key]. For sequence types, the accepted keys
 should be integers and slice objects. 
 Note that the special interpretation of negative indexes (if the class wishes to
 emulate a sequence type) is up to the __getitem__() method. If key is of an
 inappropriate type, TypeError may be raised; if of a value outside the set of
 indexes for the sequence (after any special interpretation of negative values),
 IndexError should be raised. Note: for loops
 expect that an IndexError will be raised for illegal indexes to allow proper
 detection of the end of the sequence. 

End of quote.

 
> >>     - testing membership with aset.has_key(item),
> >> or:
> >>     - indicating membership with aset[item] = 1, then
> >>     - testing membership with aset.get(item, 0)
> >>
> >> Why would you think this is "rare"?  It's common currency.  How many tens
> >> of thousands of lines of python have you written, reviewed, or
> >> maintained, that you presume to teach me what is common and what is rare
> >> in python "as she is spoken", pray?
> >
> > In how many of these cases would the 'item' key actually be confused
> > when it could be either a bool or an int? So far you failed to convince me
> > that this is not a rare case.
> 
> What does it matter what type the item is?  has_key must now return

Because that is where the confusion would lead to problems when
True and 1 are considered different keys.

> a Boolean, while the get idiom that always was equivalent to it must
> keep returning int.  Any existing code that has tests coded both ways,
> for ANY type of item, will return "mixed" Booleans and int's.  Modern
> equivalent idiom "item in aset" also returns Boolean, but THAT is quite
> likely rare since it only became possible in Python 2.2.  Still, where 2.2
> was accepted as the used language version, whenever the get idiom
> was used I refactored it into the 'in' operator on code I was revising
> anyway (but not on parts of code I wasn't revising) -- clearer, more
> concise, AND faster... irresistible (the "one obvious way to do it", once
> it became acceptable).

Such mixing of bool and int would be handled in the special methods of
the bool and int classes, eg. the __add__ method of the bool class
would use the appropriate integer value when both operands are bools.
 
> > Even though I don't see the relevance of the second question
> > the answer is: a few, and I'm not teaching here, I'm explaining
> > my opinion.
> 
> Based on WHAT experience of Python, that's what I'm asking --
> maintenance responsibility for how many tens of thousands of lines,

A few.

> etc.  How can you have a sound opinion on what is rare or frequent if
> not on the basis of some reasonably representative sample?

The opinion comes from my intuition that most of the work on mixed bool
and other types would be done in a backwards compatible way by the special
methods of the standard types.

For example, assuming pure bool True and False:

True + True   --> 2   # done by the __add__ method bool
5 * False     --> 0   # by __mul__ of int, or by __rmul__ of bool
True - 1      --> 0   # by __sub__ of bool, or by __rsub__ of int.

And even:

'01'[True]    --> '1'   # done by __getitem__ of string.

As I said earlier, I don't like this behaviour, but I am
in favour of PEP 285, and I accept this.

However, dictionary indexing can be problematic as I indicated above.

For example when indicating membership with aset[item] = 1, 
this would break:

m1 = aset.has_key(item)
someDict[m1] = value1
m2 = aset.get(item, 0)
someDict[m2] = value2

To break, it not only needs _both_ styles of membership testing
to occur in the same program, but also the results of both 
styles should be used to index a dictionary.

How often are the direct results of both styles of tests used to
index a dictionary?

> >> > That should only happen when True and 1 are used interchangeably,
> >> > which might happen by programmer mistake or by using
> >> > old (1/0) and new (True/False) style code together.
> >>
> >> Old-style only will do perfectly well, since lots of extremely common
> >> idioms therein (a<b, d.has_key(k), x in y, and so on) change return
> >> values silently.  Neither any mistake nor any 'together' is needed,
> >> unless you're somehow trying to imply that anybody coding some kind
> >> of "advanced" (?!) construct such as "a<b" was using "new style code"
> >> unwittingly years ago...
> >
> > By new style I meant explicit use of True/False.
> 
> So do I: and there is absolutely no "explicit use of True/False" in any
> of the several expressions I enumerted, nor the many more whose
> return type change silently from int to Boolean.  "True and 1 are used
> interchangeably" in any existing function that can return any of these
> expressions OR an explicit 0 or 1, as in:

I meant this to be limited to the context of indexing dictionary
like objects. Somehow that context was lost :(

> def isValueAtIndex(alist, anindex, avalue):
>     if -len(alist)<=anindex<len(alist):
>         return alist[ainindex] == avalue
>     else:
>         return 0
> 
> or equivalently and more Pythonically:
> 
> def isValueAtIndex(alist, anindex, avalue):
>     try: return alist[ainindex] == avalue
>     except IndexError: return 0
> 
> There is absolutely no programmer mistake here, nor any use of old
> style and new style code together.  Yet such existing code may very
> well return int or Boolean interchangeably _because the return type
> of very common operations has changed silently in 2.3_ ... !!!

Which I expect to be handled well by other special methods, see above.

> > As for "advanced (?!): I don't have a time machine.
> 
> So you'll have to admit that claiming "interchangeable return of int
> or Boolean" can only be caused by mistakes or "mixing old and
> new style" was a wildly unfounded assertion, yes?

Again the context was lost.

> >> > In the pure case an int() call would not be needed at all:
> >>
> >> But to check whether it's needed or not, a type-check would be
> >> needed, and that IS a substantial slow-down.  You don't seem to
> >
> > If it's not needed at all, there is also no need for the check.
> 
> How do you *KNOW* whether to check or not, without a check?

Since bool would be normally hashable and there would be no
need to 'cast' True/False to 1/0, the __getitem__ method of dict
would not need to treat bool as a special case.

> > I am not talking about an actual implementation.
> > I'm talking about a specification, while trying to keep an eye on
> > implementABILITY.
> > Since you didn't say it cannot be done, I'm actually encouraged.
> 
> It IS quite possible to implement a language where the semantics
> of, say, + and - operators, are switched in any source line which has a
> prime number of characters.  It IS totally absurd, but clearly feasible.
>
> Why such technical feasibility would encourage anybody to consider
> such absurdities is beyond me, but please feel free to go ahead and
> implement such a beauty.  It will still be less absurd and abstruse
> than your cherished "implementable" specification, anyway.
> 

Most of the work to handle mixed bool and other types I meant to be done
in a backwards compatible way.

Regards,
Ype


P.S.

Deprecation of mixed type dictionary indexing could be done in phases:
a. deprecate indexing of standard dictionaries using both PEP 285 bool
   and int keys.
b. introduce pure bool and emulate the 1/0 behaviour of pure
   True and False. This takes the performance hit for the typecheck
   on bool and the cast to 1/0 for __getitem__ (and friends) of
   standard dicts.
c. end the deprecation by removing the this typecheck.

Phase b. would off course be as short as possible.

Some other interesting code I remember reading about would also
be involved in deprecation:

type(b) == type(1) # breaks in 2.3?, ie. already deprecated?

# breaks when removing the subtype relation between bool and int:
isinstance(b, type(1))


-- 
email at xs4all.nl


From roy at panix.com  Sat Apr 27 08:37:31 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 27 Apr 2002 08:37:31 -0400
Subject: lint for python?
Message-ID: <roy-6B1ABC.08373127042002@news1.panix.com>

I committed an interesting error just now.  I've got a list, foo, and meant 
to type:

foo.sort()

but left out the parens and ended up with:

foo.sort

in my program by accident.  From python's point of view, it's not an error, 
so no error message was produced.  Even more interesting, both forms leave 
foo as a list, with the same elements, so no exceptions got raised as I 
continued to process the list further down in my program.  My program 
simply didn't do what wanted it to.

So, my question is, is there such a thing as lint for python?  Something 
like "foo.sort" could have generated a warning, "non-call of function", or 
"rvalue with no assignment", or some such.  It's certainly not something I 
want the regular interpreter to do, but as a quality control tool, having 
something that warns of questionable constructs is often useful.


From fredp at mygale.org.nospam  Fri Apr  5 04:58:34 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 5 Apr 2002 09:58:34 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze>
Message-ID: <Xns91E77A4404E25PaCmAnRDLM@161.48.128.20>

Peter Hayes <peter at seahaze.demon.co.uk> said :

> Does anyone know if there is a version of Python that runs on the new
> Linux based Sharp SL5000 PDA?

Yes, it was ported during the beta phase last year. I remember seeing the 
announcement on the developer's list when it was on Yahoo Groups, but it 
has moved and I don't know where it is hosted now.

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


From starx at pacbell.net  Sat Apr 20 05:58:45 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 20 Apr 2002 09:58:45 GMT
Subject: Mail/news + References? <- RE: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019286100.3207.python-list@python.org>
Message-ID: <Xns91F61E77DFA42RASXnewsDFE1@130.133.1.4>

Tim Peters || Sat 20 Apr 2002 12:00:05a:

> [David Eppstein]
>> The point is that your responses don't have the appropriate
>> References: header, so they don't get threaded with other messages on
>> the same topic.
> 
> I know.  The news reader I use (when I use one) uses criteria in
> addition to "References:" for threading, so I never notice it myself.
> 
>> It appears as if you are using a mail agent rather than a news agent,
>> since References: is a news header (and your other headers say you're
>> using MS Outlook).
> 
> I do use the news<->mail gateway most of the time.
> 
>> It affects other things than threading, too -- e.g. I use References:
>> to pick out responses to my own messages, and yours don't get caught.
> 
> Then you'll be spared this <wink>.

Hm, even though my newsreader can keep things in their appropriate threads 
most of the time, but anyone posting from the mailing list gets put out of 
the indentation scheme. That is, instead of appering in a hierarchy under 
the message you decided to reply to, you show up on the same level as the 
first message in the thread, breaking the normal flow of the discussion. I 
don't really think it's possible for a newsreader that works like this to 
funciton properly without the references headers; i have a felling yours 
just lists all the messages in a thread in chrono order instead of trying 
to physically show the who-replies-to-who shceme?

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From jeff at ccvcorp.com  Fri Apr 12 13:36:40 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 12 Apr 2002 10:36:40 -0700
Subject: Tuple comprehension
References: <mailman.1018539023.28081.python-list@python.org> <slrnabchd7.c8l.mlh@vier.idi.ntnu.no>
Message-ID: <MPG.1720baf1c72dfb4a9896ab@news.nwlink.com>

In article <slrnabchd7.c8l.mlh at vier.idi.ntnu.no>, 
mlh at vier.idi.ntnu.no says...

> [...]  It just seems natural to do
> 
>   geometry = (state[name] for name in names)
> 
> (I'm assuming that the added parentheses would signal the tuple
> comprehension, just like the added brackets will signal generator
> comprehension.)

This wouldn't work -- the parser already recognizes parentheses 
as expression grouping.  Even if we *could* make the parser smart 
enough to recognize a tuple-comp, it would be confusing to read, 
at least IMO.

I also agree with Dave Brueck, that the cases in which tuples 
hold uniform values (so that tuple comprehensions might make 
sense) are relatively rare, compared to cases in which tuples 
hold heterogeneous values.  (This is exactly the opposite case of 
lists, which may be heterogeneous but are usually homogenous.)  
So ISTM that tuple comprehensions would be of limited utility.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jbperez808 at yahoo.com  Sun Apr 28 03:57:35 2002
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 28 Apr 2002 00:57:35 -0700
Subject: compiled Jython vs. CPython
References: <f57664b9.0204211807.2235f3bf@posting.google.com> <3CC4790C.B7495A5E@accessforall.nl>
Message-ID: <f57664b9.0204272357.769887d4@posting.google.com>

> > Could jythonc be *the* Python compiler we wish
> > we had?
> 
> In both CPython and Jython all python code is compiled
> on the fly to byte code before being executed. For imported
> modules this byte code is also saved.

I was hoping that the one difference that might go in
favor of Jython is that while CPython bytecodes are
interpreted while the program is being run, Java bytecodes 
generated by Jython will be compiled into native code 
first by a JIT compiling JVM.


From chris.gonnerman at newcenturycomputers.net  Mon Apr 15 08:06:48 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 15 Apr 2002 07:06:48 -0500
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp>  <7xelhhmq5e.fsf@ruckus.brouhaha.com> <slrnablf4d.5k9.mlh@vier.idi.ntnu.no>
Message-ID: <004901c1e476$0c171e00$0101010a@local>

----- Original Message ----- 
From: "Magnus Lie Hetland" <mlh at vier.idi.ntnu.no>

> Well -- the Python license is GPL-compatible, so couldn't you just
> release a GPL'ed version of it, if that would make you happy? Of
> course, the original would still be freely available without the added
> burden of the GPL restrictions, but if you're really interested in
> freedom, that shouldn't bother you...

"GPL Compatible" with respect to Python's license does NOT mean that 
anyone can change the license at will.  If someone incorporates GPL'd
code into the Python core, the *aggregate* must be distributed under
the GPL but the *core* is still under the normal Python license.

If you aren't the author / copyright owner, you have no right to change
the license.
 
> <duck and cover>

Good idea.  :-)

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





From aahz at pythoncraft.com  Sun Apr 21 09:45:35 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 09:45:35 -0400
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com>
Message-ID: <a9ufpv$6hb$1@panix1.panix.com>

In article <a9t177$pi5$1 at panix1.panix.com>, Aahz <aahz at pythoncraft.com> wrote:
>In article <roy-AAA3A1.19193820042002 at news1.panix.com>,
>Roy Smith  <roy at panix.com> wrote:
>>
>>I want to look an oid up in my dictionary, and if it's not there, I want to 
>>keep working my way up the tree until I find it.  Thus, if 
>>'1.3.6.1.4.3.5.66.3.4' isn't a key, I want to try '1.3.6.1.4.3.5.66.3', and 
>>then '1.3.6.1.4.3.5.66', and then '1.3.6.1.4.3.5', and so on until I find 
>>it (if nothing else, '1' is guaranteed to be a key).
>>
>>What I've come up with uses recursion:
>>
>>    def getBaseObject (self, objects, oid):
>>        try:
>>           return objects[oid]
>>       except KeyError:
>>           return self.getBaseObject (objects, self.getParentOid (oid)) 
>
>I'd write this as
>
>    while not objects.has_key(oid):
>        oid = self.getParentOid(oid)
>    return objects[oid]

...and if you're using Python 2.2, write it as

    while oid not in objects:


Hmmmm....  it occurs to me that those of us teaching Python will need to
be *very* careful in making sure that people avoid 'in' over lists.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From wtri at wtresearch.de  Mon Apr 22 10:23:08 2002
From: wtri at wtresearch.de (wtr)
Date: Mon, 22 Apr 2002 16:23:08 +0200
Subject: yopy
Message-ID: <1019481209.166566@lix-049dz-008.inside.impress.com>

Hi,
has anybody done a test with the YOPY (also StrongArm based with
"linupy"/yopy-linux)?
Thanks,
wtr (wolfgang)




From paul at boddie.net  Tue Apr  9 05:00:13 2002
From: paul at boddie.net (Paul Boddie)
Date: 9 Apr 2002 02:00:13 -0700
Subject: ANN: Twisted 0.16.0
References: <mailman.1017787747.12194.python-list@python.org> <3CAA8CFF.B6EDDCDE@alcyone.com> <8Fgs8.27735$Sw5.8757@atlpnn01.usenetserver.com>
Message-ID: <23891c90.0204090100.38023e58@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<8Fgs8.27735$Sw5.8757 at atlpnn01.usenetserver.com>...
>
> Well, a good April Fool's joke should obviously *be* an April Fool's joke,
> so ultimately the joke's on the Twisted team.

Given that the announcement is also archived in "the public record", I
wonder if the joke has had any effect on the number of people willing
to evaluate the software. It probably isn't such a great idea to joke
about software licensing if you want people to download and deploy
your product.

It's better to claim that it runs on the Vectrex instead... ;-)

Paul


From jb at cascade-sys.com  Tue Apr 23 13:56:36 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 23 Apr 2002 10:56:36 -0700
Subject: file_ref = open(string_ref)
References: <1019579845.565223@cswreg.cos.agilent.com>
Message-ID: <3CC5A054.82E6C887@cascade-sys.com>

Greg Weeks wrote:

> Is this code
>
> C1:     file_ref = open(string_ref)
>
> better than this?:
>
> C2:     file = open(string)

I would vote for C2 for the simple reason that C1 stinks of Microsoft's dreaded
"Hungarian" notation.  It may have use in assembly code and maybe even in C but
I think encoding object types into object names is not only unnecessary it
creates a maintenance hazard.  If you wanna know the type, look at the
definition.  Better, choose good class, instance and method names so that your
code is clear without having to be reminded of the class name.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From aleax at aleax.it  Wed Apr 10 17:31:34 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 23:31:34 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <20020410230535.674eef0a.andreas@mtg.co.at>
References: <mailman.1017466974.6751.python-list@python.org> <Ohxq8.45137$pT1.1283197@news1.tin.it> <20020410230535.674eef0a.andreas@mtg.co.at>
Message-ID: <02041023313403.17153@arthur>

On Wednesday 10 April 2002 23:05, Andreas Kostyrka wrote:
	...
> Ok, some very useful case, actually I've done this in pre-2.0 Python with
> some kludges and some magic :)
>
> class MagicDataBaseTable:
> 	...
> 	def newRow(klass,**params):
> 		...
> 	newRow=classmethod(newRow)
> 	...
>
> class MyTable(MagicDataBaseTable):
> 	...
>
> Now the newRow class method knows that it is being called upon MyTable
> and can be parametrized by the table and column names of that table.

Super, but, I am not sure why is it a class method rather than a normal 
instance method, accepting self and taking the table and column names
from self (which of course could in turn implicitly delegate them to its
class).  Wouldn't I normally be adding a row to a specific table, an
instance of one of these classes, after all?  Maybe you can show us some
more details because, so far, the superiority of a class method wrt a 
normal instance method is still murky to me.  Thanks!


Alex




From tim.one at comcast.net  Fri Apr 26 16:37:23 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 16:37:23 -0400
Subject: [OT] What is Open Source?
In-Reply-To: <BRZy8kKkXIyO092yn@gnosis.cx>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEEPPBAA.tim.one@comcast.net>

[Lulu of the Lotus-Eaters]
> ...
> FWIW.  Even in this world, I still put my work in the public domain
> (whenever I have control over the matter)... because I prefer to live in
> a fantasy world of ethical purity.

Me too, but at least one lawyer has told me it is a fantasy world, i.e. that
an individual in the United States cannot put their work in the public
domain:  you can't disclaim your copyright even if you want to, unless it's
work done for the government and "they" put it in the public domain.

like-i-care<wink>-ly y'rs  - tim





From pyth at devel.trillke.net  Wed Apr 17 19:22:38 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 18 Apr 2002 01:22:38 +0200
Subject: [OT] What is Open Source?
In-Reply-To: <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov>
References: <20020418002823.C4068@phd.pp.ru> <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov>
Message-ID: <20020417232238.GC17361@devel.trillke>

On Wed, Apr 17, 2002 at 04:44:50PM -0600, rasmussn at lanl.gov wrote:
> 
> On Wednesday, April 17, 2002, at 02:28 PM, Oleg Broytmann wrote:
> 
> >   It is not enough to have the code open. One needs the rights to read, 
> >to
> >use, to copy, to modify, to distribute modified and unmodified versions 
> >(so
> >everyone benefits from the open code). Only Free Software gives you such
> >freedom and rights.
> 
> As I understand it, the GPL does not give you right to use the software
> as you see fit.  You can only use it by linking with other GPL'd
> software.

cant be true: gpled software under windows wouldnt be possible...

	holger




From aahz at pythoncraft.com  Tue Apr 16 00:19:32 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 00:19:32 -0400
Subject: Does Python need a '>>>' operator?
References: <mailman.1018925938.16233.python-list@python.org>
Message-ID: <a9g8ok$9a3$1@panix1.panix.com>

In article <mailman.1018925938.16233.python-list at python.org>,
Tim Peters  <tim.one at comcast.net> wrote:
>
>a co-finite set 

What's a "co-finite set"?  (Yeah, I could Google for it, but I usually
understand *your* explanations -- eventually. ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From starx at pacbell.net  Fri Apr 12 03:16:39 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 12 Apr 2002 07:16:39 GMT
Subject: Beginner's performance problem
References: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de> <memo.20020411143211.1400B@a.a> <a95d1v$avb$1@panix1.panix.com> <3CB60C73.1010101@sneakemail.com>
Message-ID: <Xns91EE3036F42FRASXnewsDFE1@130.133.1.4>

Joshua Macy || Thu 11 Apr 2002 07:26:04p:

> Aahz wrote:
> 
>> In article <memo.20020411143211.1400B at a.a>,
>> Mark Charsley <mark.charsley at REMOVE_THIS.radioscape.com> wrote:
>> 
>>>TVM to everyone who helped.
>>>
>> 
>> I'm familiar with TVP, but what's TVM?
>> 
> 
>    TVM is obvious Ta Ever So
> 
>    ... or something like that. <wink>
> 
>    Joshua
> 
> 

You lost me after the Ta there... but i'd think it'd mean 'thanks very 
much', at least in that context.

not-that-i-know-what-TVM-is-ly outta here

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From shalehperry at attbi.com  Thu Apr 11 01:23:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 10 Apr 2002 22:23:12 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <%r7t8.37204$CU.41841@rwcrnsc51.ops.asp.att.net>
Message-ID: <XFMail.20020410222312.shalehperry@attbi.com>

On 11-Apr-2002 Emile van Sebille wrote:
> Magnus Lie Hetland
>> As far as I can see, the PEP on dict comprehension is still open... I
>> think both dictionary comprehension and tuple comprehension seem like
>> natural parallels to list comprehension. But then I would, wouldn't I?
>> <wink>
>>
> 
> Why not just tuple([my list comp]) ?
> 

because it feels wrong to create a (possibly large) list and then duplicate it
into a tuple.  My mind sees that and thinks 'overhead'.




From mwh at python.net  Thu Apr 18 04:58:51 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 08:58:51 GMT
Subject: Mailman lists at Starship
References: <mailman.1019085516.2498.python-list@python.org> <3CBE6B4F.3030506@thomas-guettler.de>
Message-ID: <lkelhdml9w.fsf@pc150.maths.bris.ac.uk>

Thomas Guettler <zopestoller at thomas-guettler.de> writes:

> Gerhard H?ring wrote:
> 
> > Does anybody know who is the technial support contact ;-) for these?
> > I'm the list admin of
> > http://starship.python.net/mailman/listinfo/python-de
> > and I'm having some problems.
> > Gerhard
> >
> 
> Some days ago there was this message in python-announce:
> 
> """
> We plan on upgrading the mail server mail.python.org on Wednesday

But that's .org, not .net.  The starship's back up again, if that was
the problem...

Cheers,
M.

-- 
  : Giant screaming pieces of excrement, they are.
  I have a feeling that some of the people in here have a 
  MUCH more exciting time relieving themselves than I do.
                                       -- Mike Sphar & Dave Brown, asr


From grante at visi.com  Fri Apr  5 20:31:40 2002
From: grante at visi.com (Grant Edwards)
Date: Sat, 06 Apr 2002 01:31:40 GMT
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>
Message-ID: <slrnaask0t.1dm.grante@tuxtop.visi.com>

In article <40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk>, Alastair Nicol wrote:

> I've been enjoying the wonders of socket programming with Python. 

Way more fun than C, eh?

> 1) Can only hosts which can see the interface your program is
>    bound to communicate with your program? In the "normal"
>    case if you bind to 127.0.0.1 for example ONLY your machine
>    can see that interface. Other machines cannot connect to
>    it. Is this a safe / secure assumtion? (not just for
>    localhost)

Well, yes.  It's sort of a tautology.  There is no way to "see"
an IP interface other than to communicate with it.

> 2) Also when a client connects to a socket the server gets the
>    IP address of the client. Is this the "real" ip number?

Sometimes.  For some values of "real".

>    Ive heard you can spoof IP numbers so what hoops are needed
>    to validate that the client is really that client.

If the other "end" of the connection is behind a
router/firewall that does NAT (network address translation),
then the IP address you see is that of the firewall, not of the
machine to which you are ultimately talking.  Often there are
more than one NAT translation going on (e.g. when I log into
my home machine from my office there's one NAT translation in
the firewall at the office, and one in my DSL modem/router).

In general the IP address is "real" in that it's the IP address
to which you send packets.  It may or may not be useful in
identifying the machine at the other end.

If somebody really does "spoof" an IP address (uses an IP
address which doesn't really route frames back to them), then
the connection can never be completed.  Spoofed IP addresses
are often used for low-level attacks.

> openSSL is an option i would like to use, but for reasons to
> numerious to list here, i cant.

OK.

-- 
Grant Edwards                   grante             Yow!  over in west
                                  at               Philadelphia a puppy is
                               visi.com            vomiting...


From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 17:22:11 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 23:22:11 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE> <mailman.1018991313.16792.python-list@python.org>
Message-ID: <a9i4ms$ece$1@nets3.rz.RWTH-Aachen.DE>

"Geoff Gerrietts" <geoff at gerrietts.net> wrote in message
news:mailman.1018991313.16792.python-list at python.org...
> Quoting Philipp Lenssen (lenssen at hitnet.rwth-aachen.de):
> > I just realized, as opposed to PHP, there's much stricter namespaces in
> > Python and maybe not using "import" on anything harmful could be the
> > solution?
>
> You can hack on the rexec module to make it do what you want to do;
> that's what I do. It may or may not provide ample security for your
> purposes; it's certainly better than a raw eval().
>
> The code I use is attached; under 2.something,

Thanks! I will try out your class and simply use it if it does what I want.
Please email me if you a) don't want me to use it xor b) don't want to get
credited in the source file or the http://questml.com webpage.

> this breaks unless you
> override make_re to pass.
>..

Could you explain that? I didn't understand...




From marklists at mceahern.com  Wed Apr  3 13:55:26 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 12:55:26 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Xns91E5D13ADC09cliechtigmxnet@62.2.16.82>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEAHKJAA.marklists@mceahern.com>

[Chris Liechti]
> i disagree. comparing True/False with other values should give the same 
> result as when it were used with an 'if':
> >>> True == 1, True == 6
> True, True

So 6 == True == 1, but 6 != 1?

// mark




From NO_SPAM_jmpurser2 at attbi.com  Thu Apr 18 12:03:31 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Thu, 18 Apr 2002 16:03:31 GMT
Subject: Reading from a binary file in a portable way
References: <3cbe6892_11@news.newsgroups.com> <p9yv8.11123$eS3.9630@atlpnn01.usenetserver.com>
Message-ID: <n7Cv8.244313$Yv2.68657@rwcrnsc54>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:p9yv8.11123$eS3.9630 at atlpnn01.usenetserver.com...
> "jb" <egrossde at yahoo.de> wrote ...
> >
> > I should like to read a (binary) file x.y in a buffer and then do things
> > like getting the 16 bit integer that is at the offset n, n+1 in the file
> > (maybe lowbyte, then higyhbyte.
> >
> > Previously I did it only under MS-Windows where for example I could open
> the
> > file with the flag O_BINARY but this is no longer possible under Linux.
> >
> > What would a solution that works under both Unix and MS-Windows look
like?
> >
> Such a solution would probably use the mmap module. Although there are
minor
> differences on the two platforms it should be possible to write something
> that works well on both.
>
> regards
>  Steve


I just did this.  An outline of my code would look like:

import struct

fInput = file('im1.d', 'r+b')

fInput.seek(512)

record = fInput.read(180)

item_number = struct.unpack('> i', record[1:5])

print item_number

fInput.close()

It worked nicely on Windows 2000 (Python 2.2) and Debian Linux (Python 2.1).
I was reading and writing to the input file and needed data from the whole
record so there's some extra lines.  I'd suggest writing a fuction that
positioned the file pointer in the correct position based on record number,
record length, data slice, header, etc.

Good Luck,

John Purser







From robin at jessikat.fsnet.co.uk  Thu Apr  4 17:49:25 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 4 Apr 2002 23:49:25 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017946214.8160.python-list@python.org>
Message-ID: <zpL9tKA1hNr8EwkQ@jessikat.fsnet.co.uk>

In article <mailman.1017946214.8160.python-list at python.org>,
arthur.siegel at rsmi.com writes
>
>>[Guido]
>>> I don't think Boolean algebra is a big subject.
>
>>[David Eppstein]
>>> There are advanced graduate math books on Boolean algebra.
>> >[...]
>> >But these books' content is almost completely irrelevant for day-to-day
>> >computer programming.
>
>>[Mark]
>>>Don't you figure that's what Guido meant by saying it's not a 'big' subject?
>>>I mean, there are advanced philosophy books on Heidegger...
>
>Damn willies I tell you.
>
>Guido will always be right to a following that will work to find
>the meaning that makes him right.

In any large subject a 'Guido' will always claim to be bigger than 'God'
thus leading to Sergeant Pepper phenomena.
-- 
Robin Becker


From paul at boddie.net  Wed Apr 10 05:09:05 2002
From: paul at boddie.net (Paul Boddie)
Date: 10 Apr 2002 02:09:05 -0700
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204090012.7fc80c1f@posting.google.com> <Xns91EB245A0BCD9RASXnewsDFE1@130.133.1.4> <23891c90.0204090526.18675d02@posting.google.com> <0NCs8.63131$pT1.1810689@news1.tin.it>
Message-ID: <23891c90.0204100109.4a44932b@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<0NCs8.63131$pT1.1810689 at news1.tin.it>...
> Paul Boddie wrote:
>         ...
> > sentiments.) However, I do feel that the more recent features seem
> > increasingly remote to the beginner and even the more
> 
> This can only come from a very selective view of "recent features".

Perhaps I've just been following the wrong threads on
comp.lang.python.

> Nothing is more natural to a beginner than "if x in somedict:" and
> "for x in thefile:" -- having had the opportunity to teach some
> beginners with 2.2, I've seen them slurp these up instantly (just
> try to contrast with "if somedict.has_key(x):" and
> "for x in thefile.xreadlines():" ...!).

Well, both examples you gave seemed to be controversial when up for
discussion in the past, but if you say they help people, then I'll
take your word for it.

> And the new features are the right way to implement such obvious progress --
> iterators, for example, let you loop on (e.g.) dictionaries or files, without
> requiring such types to define unnatural (or unfeasible) "indexing"
> then go out of their way to stop misuse.

Iterators are indeed nice to use and a very welcome feature.

> > advanced/experienced Python developer.
> 
> You MUST be joking -- inheriting from builtin types is SUCH a win,
> and so obviously so, for advanced and experienced developers!

Perhaps I'm not writing the kinds of applications where this is the
case, then.

> Generators need a "from __future__ import", so I applaud the tutorial's
> choice not to burden beginners with them YET.

It would be nice for some coverage of that particular feature, though.
I can't say that I've ever written one myself, although they do seem
useful for certain kinds of applications.

> > other things to do during working hours) it becomes increasingly hard
> > to keep track of what's great in the more recent Python releases.
> 
> I think AMK's "what's new" articles are great for that purpose.

I must find the time to review them at my leisure, then.

Paul


From eppstein at ics.uci.edu  Mon Apr 29 12:07:21 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 29 Apr 2002 09:07:21 -0700
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <eppstein-7EA364.09072129042002@news.service.uci.edu>

In article <aaiu6e$pfr$1 at bob.news.rcn.net>,
 "Raymond Hettinger" <python at rcn.com> wrote:

> Reduce is easily (and more clearly) replaced
> by an equivalent for loop.

More clear to you, maybe.  I'd rather write programs that say what 
result I want, and let the interpreter figure out how to get that 
result; writing an explicit for-loop in place of a nice functional 
reduce is too much micro-management for my taste.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From tanzer at swing.co.at  Thu Apr  4 02:04:09 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 04 Apr 2002 09:04:09 +0200
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: Your message of "Wed, 03 Apr 2002 14:47:08 -0800."
             <3CAB866C.4DF51F72@activestate.com>
Message-ID: <E16t1I1-0007gL-00@tswings.swing.cluster>

David Ascher <DavidA at ActiveState.com> wrote:

> > Despite the negative feedback, I've decided to accept the PEP.  The
> > most important perceived problem is that newbies tend to write
> >
> >     if x == True: ...
> >
> > where they should write
> >
> >     if x: ...
> >
> > I believe this problem should be solved by education (the Zen master
> > hits the student on the head with a stick, and the student gets
> > enlightenment) rather than by holding back what I believe will be a
> > useful feature.
>
> Would it not be "relatively" easy to add a compiler-time warning for
> most uses of this dangerous idiom?
>
> At the very least, the boolean type could do the equivalent of:
>
>   def __eq__(self, other):
>      if __debug__ and other not in (0,1,True,False):
>          warnings.warn("Testing of equality is best done by asking
> objects whether they're true, not comparing with a boolean",
> RunTimeWarning)
>      return int.__eq__(self, other)
>
> or something similar?

I warning about `x == True` wouldn't hurt. But your proposed `__eq__`
certainly would -- don't do that please.

Equality testing is useful if one wants to check whether two
predicates have the same boolean value.

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





From jon+usenet at unequivocal.co.uk  Wed Apr 24 23:06:45 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Thu, 25 Apr 2002 03:06:45 -0000
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com> <slrnacc59a.tg8.jon+usenet@snowy.squish.net> <MCxx8.45058$0x3.38185@atlpnn01.usenetserver.com> <aa7c52$ehu$0@216.39.172.122> <slrnaceetc.3k8.jon+usenet@snowy.squish.net> <aa7meg$84u$0@216.39.172.122>
Message-ID: <slrnacesm5.bcl.jon+usenet@snowy.squish.net>

In article <aa7meg$84u$0 at 216.39.172.122>, Bengt Richter wrote:
> I just wanted to make it clear that leaving "'" out of the format string
> was not directly for the sake of SQL itself, but more for some intermediating
> software's purposes.

It's because the apostrophes are part of the syntax of the value, and
Python's DB API says that it automatically puts in all the synatx of
the value for you. This is a very good thing, because if it worked the
other way around and you were were expecting, say, a number, so you
didn't put the apostrophes in, but it turns out the value in the
sequence is a string, you suddenly have a security hole.

> I guess I could RTFM, but how do you do a single INSERT with many
> value tuples?

executemany(sql, [params1, params2, params3, ...])

> Next time I want to do something with a DB, I'll have to try it with Python.
> (I used the mySQL C API before).

Hrm, yes, it is much easier building SQL strings in a language where
you don't have to allocate the memory for string operations by hand ;-)


From phr-n2002a at nightsong.com  Mon Apr 15 02:28:45 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 14 Apr 2002 23:28:45 -0700
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <7xelhhmq5e.fsf@ruckus.brouhaha.com>

"Stephen J. Turnbull" <stephen at xemacs.org> writes:
> And the (rms-deprecated!) LGPL is surely good enough for anybody who
> only cares about freedom of their own code.  So the GPL is an advocacy
> tool, pure and simple.  It's all about affecting licensing decisions
> for code that others write, it has nothing to do with protecting
> freedom of yours.

Um, I value the freedom to use code that I wrote, including derivative
versions.  If you think I shouldn't have any interest in derivative
versions, try selling some derivative versions of Microsoft code and
see how fast Bill slams you.

>     Chris> There are those (uniformly programmers) who hate the GPL
>     Chris> because it restricts their ability to "hide" the source if
>     Chris> they incorporate GPL'd code in their non-GPL programs.
> 
> Consumers also dislike the GPL because it removes their freedom to
> sell their privilege of inspection of enhanced derivative software, a
> privilege most will never wish to exercise.  They don't hate it the
> way Bill Gates does, of course; they just would rather not pay for it
> in higher prices and reduced functionality. 

Or lower prices and enhanced functionality?  Look at all the GCC ports
that have been released for free because of the GPL, for example.  The
people who did those ports would have loved making them proprietary,
but that option wasn't available if they wanted to start with GCC, and
users everywhere benefitted.  (Of course if a developer really wants
to release a proprietary C compiler, he's perfectly free to write his
own instead of porting GCC, so the GPL imposes nothing on him).  

If I write a program that does XYZ and you want a program that does
XYZ+W, the GPL is my way of saying "hey, I'll do 90% of the work of
that XYZ+W implementation for you if you make the final result free".
It promotes cooperation by providing an incentive to release free
software.  Whereas if I release XYZ under a BSD license so your XYZ+W
ends up proprietary, I've simply become your unpaid employee.  That
arrangement might be nice for YOU, but I can do perfectly well without
it, thank you.

The non-GPL-ness of Python is a main reason I don't have much interest
in doing significant development work on the Python implementation
(though I'm willing to make smaller enhancements to fill my own
requirements, and send in the patches).  On the other hand, Python's
non-GPL may have attracted some corporate development money, so it's a
trade-off.  And of course some volunteers do contribute heavily.  It's
a decision everyone has to make for themselves based on their own
priorities.


From eppstein at ics.uci.edu  Mon Apr 15 02:52:57 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 23:52:57 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <a9dqkk$5kf$0@216.39.172.122>
Message-ID: <eppstein-A8E7D2.23525714042002@news.service.uci.edu>

In article <a9dqkk$5kf$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
wrote:

> >The reason I was getting into trouble anyway was because of the way 
> >right-shifting can change sign:
>  ^^^^^ ITYM left?
> >
> >>>> 0x40000000<<1
> >-2147483648

Yes, of course, thanks for the correction.

The only way I can remember which direction is left and which is right 
is to look at which hand has a ring on it...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From logiplexsoftware at earthlink.net  Wed Apr 10 14:28:22 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 11:28:22 -0700
Subject: Help: Python for a big commercial project?
In-Reply-To: <F194tIpqqJgSt51NDa200002509@hotmail.com>
References: <F194tIpqqJgSt51NDa200002509@hotmail.com>
Message-ID: <20020410112822.4ddfc783.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 13:24:59 +0200
Cereal Muesli wrote:

>   the people here is not likely to use any free stuff for "serious" 
> development (you know...no money = no quality, sigh!),

And they're using HPUX?  Hm.  I wonder how they'll defend that position once
GNOME is the default desktop on that platform ;)

>   but i want to change this, since I believe that Python is really a very 
> good product. I have played with Python for
>   a year or so under Windows, and before trying to convince people here that 
> it is a good choice, I need to know
>   some things:
> 
> - First of all; do you think that Python is suitable for this
>   kind of "serious, big" project?

Absolutely.  I would not hesitate to use Python for *any* project.  At a prior
job, I used Python develop the primary application for a contract worth over $3
million a year.  In my current position, I am moving our applications from C to
Python and these applications will be responsible for monitoring and
controlling security systems in some high-risk/top-secret facilities (overseas
military installations, government research facilities, etc - if I told you
more, well, you know the line ;)

In fact, I would recommend Python /over/ other languages since its inherent
readability helps prevent subtle programming mistakes.  Even minor flaws in
software can be fatal under the right circumstances and Python helps avoid
these by 1) reducing LOC and 2) encouraging readable code.  While it's clearly
possible to write bad code in any language, Python's nice data types (lists,
dicts, objects in general) and useful libraries help alleviate the need for
clever/complex code (read "unreadable") or often the need to write code at all
(much programming in other languages is spent developing/debugging complex data
structures rather than the application itself).  I have seen (and written)
plenty of C that had more to do with working around limitations imposed by the
compiler than actually solving the problem at hand (worse, I congratulated
myself on these feats ;).  This code would have been totally unnecessary in
Python.

[snip]
Afraid I can't comment on HPUX since I've never used it.

> - Finally: Which version of Python should we use? it seems that
>   1.5.x still rules under *NIX, doesn't it?

I suppose that depends on which Unix.  I use 2.2 on Linux and 1.5.2 on SCO
Unix.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From v.wehren at home.nl  Sun Apr 28 03:15:56 2002
From: v.wehren at home.nl (vincent wehren)
Date: Sun, 28 Apr 2002 09:15:56 +0200
Subject: AW: does readlines stop at 255?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192B9836E@admin56.narex.com>
Message-ID: <NFBBJHJACLIMCPDAKGKIMEBHCBAA.v.wehren@home.nl>

Didn't know this one slipped through - I thought I killed the message before
sending it...
Anyway, since I couldn't believe it either, I didn't stop till I found the
bug..
Somebody (who shall go nameless - since it might have been my great self)
sliced the text at some point
in the unfathomable code!

Thanks all the same

Regards
Vincent





-----Urspr?ngliche Nachricht-----
Von: Bjorn Pettersen [mailto:BPettersen at NAREX.com]
Gesendet: Samstag, 27. April 2002 20:11
An: v.wehren; python-list at python.org
Betreff: RE: does readlines stop at 255?


> From: v.wehren [mailto:v.wehren at home.nl]
>
> It appears that readlines() stops at 255...when reading a line..
>
> How can I work around this if my file has lines longer than
> 255 characters (and making them shorter is not an option...)??

readline() can handle any length lines (up to available memory). There
must be something else wrong. Post code.

-- bjorn





From martin at v.loewis.de  Tue Apr  2 01:05:34 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 08:05:34 +0200
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com> <a8apju$kcb$0@216.39.172.122> <m3pu1jvw2d.fsf@mira.informatik.hu-berlin.de> <a8b6p9$q40$0@216.39.172.122>
Message-ID: <m3d6xiocsx.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> >That assumes that the output encoding is known, or can be
> >determined. As-is, it can't - you don't know the encoding of f, and
> Well, I was wondering more where we're heading than 'As-is' ;-)
> 
> IOW, assume encoding was an optional keyword parameter to open/file.
> Then you'd know what output encoding was desired.

Provided that parameter is provided. You get this today, via
codecs.open.

> But no string would exist without at least an assumption as to its
> encoding.

This alternative (associate each string with an encoding) has been
considered and rejected: it was considered to be better to only
support Unicode, and explicit conversion. Otherwise, two strings with
different incompatible encodings might produce interesting results.

Notice that the current approach gives the same result as the "each
string has an encoding" would give if it worked: Somehow, the
application has to tell, for each string, what the encoding is. *If*
the application can tell, it should currently convert that string to
Unicode - then the information what byte means what character is
preserved in the Unicode string.

Also notice that there *is* an assumption to the encoding of each
string; sys.getdefaultencoding().

> I guess you could do unix file-type magic to infer encoding if you had to,
> but it wouldn't seem reliable or cheap except utf & co.

Indeed. That can't really work.

Regards,
Martin


From juenglin at informatik.uni-freiburg.de  Sat Apr 13 06:27:20 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 13 Apr 2002 12:27:20 +0200
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk>
Message-ID: <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de>

Michael Hudson <mwh at python.net> writes:

> Ralf Juengling <juenglin at informatik.uni-freiburg.de> writes:
> 
> > Shouldn't I find the special methods '__iadd__' et al, when
> > asking for the attributes of 'int'?
> 
> ints are immutable, so they don't implement __iadd__.  lists do:
> 

Hm. But then an expression 'i+=1' should cause an error if 'i'
is an int...

>>> i = 1
>>> type(i) 
<type 'int'>
>>> id(i)
134526912
>>> i+=1
>>> id(i)
134526852
>>> 

I understand now, that 'i+=1' is no real in-place operation but
just a shortcut for 'i=i+1' since int is immutable. The language
reference states that this is okay, anyhow, I find it confusing.

For the sake of clarity, an 'in-place' operation should really 
work in-place, don't you think?


Cheers,
Ralf



> > I couldn't find them either in ther operator-module:
> 
> Hmm, I think the issue here is that 
> 
> x += y
> 
> translates (in the presence of an __iadd__ method) roughly to 
> 
> x = x.__iadd__(x, y)
> 
> and you can't do that in the form of 
> 
> operator.blah(x,y)
> 
> .  Or it might just be oversight.
> 
> Cheers,
> M.
> 
> -- 
>   Remember - if all you have is an axe, every problem looks 
>   like hours of fun.                                        -- Frossie
>                -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Wed Apr 17 00:32:19 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Tue, 16 Apr 2002 21:32:19 -0700
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org> <3cba62f8@news.mhogaming.com> <slrnablsp5.f08.jon+usenet@snowy.squish.net> <3cbb996d@news.mhogaming.com> <slrnabobcu.2k3.jon+usenet@snowy.squish.net>
Message-ID: <3cbcfb2a@news.mhogaming.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnabobcu.2k3.jon+usenet at snowy.squish.net...
| In article <3cbb996d at news.mhogaming.com>, Ken Peek wrote:
| >| So you think that:
| >|
| >|   long(hex(-1l), 16)
| >|
| >| should not equal -1 but, err, some random choice of 15, 255, 4095,
| >| 65535, etc?
| >
| > No, since we are "unifying" the int and long type, it is none
| > of your business HOW they are stored-- so we do away with the
| > 'int()' and 'long()' functions...
|
| So you think that:
|
|   number(hex(-1), 16)
|
| should not equal -1 but, err, some random choice of 15, 255, 4095,
| 65535, etc?

This is a non-sequitur.  If you want to convert an integer that is holding a
number to a string and back (for whatever reason)-- use decimal representation.
Hexadecimal is for examining bit patterns, which are not necessarily numbers,
although integers can also be used to hold them both.







From gcordova at hebmex.com  Wed Apr  3 16:10:36 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 3 Apr 2002 15:10:36 -0600 
Subject: Deeply embedded Python still alive?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB55@mail.hebmex.com>

Go ahead and post the answers you get,
remember that they stay archived in google's
newsgroup database, so anybody asking the
same or related things can simple google
for an answer.

(neat, how a noun "google" becomes a verb "google"
with time and usage)

-gustavo




From sholden at holdenweb.com  Wed Apr 17 09:36:52 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 09:36:52 -0400
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com> <slrnabp95i.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <QVev8.13228$hl2.7933@atlpnn01.usenetserver.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
> On Mon, 15 Apr 2002 17:56:42 -0700, Erik Max Francis <max at alcyone.com>
wrote:
> >phil hunt wrote:
> >
[phil wants empty nested suites]
> >
> >The purpose of pass to explicitly disclaim what you're intending to do;
> >it reduces programmer error.  In cases in code where you see something
> >like:
> >
> > if condition:
> > doSomething
> >
> >It is far more likely that this is an indentation mistake rather than a
> >deliberate null statement.
>
> I don't recall ever having made that sort of mistake. In any case,
> that sort of thing (i.e. legal but possibly wrong code) is what lint
> tools are for. The classic example being if (a = b) in C.

Well, of course, if you never make that mistake then there's no need to
ensure that the other 231,763 Python programmers don't have any problem with
it, right <wink>?

I rather suspect Max's point was that it would take more time to examine
PyChecker output to determine there really was no error than it would to put
a pass statement in to start with. This somehow loses a lot of the point of
the exercise, wouldn't you agree? [This last is a rhetorical question, since
I'm pretty sure you wouldn't agree].

However, since you are clearly bent on finishing the PEP it will stand as a
permanent record of the decision to reject it, and will therefore be a Good
Thing.

regards
 Steve






From fredrik at pythonware.com  Tue Apr 30 17:06:52 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 21:06:52 GMT
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com> <xdur8kx1i1i.fsf@leto.informatik.uni-freiburg.de> <3CCEDA38.2559BE98@alcyone.com>
Message-ID: <MHDz8.34856$n4.7400679@newsc.telia.net>

Erik Max Francis wrote:

> *BUT*, these (eg. isSequenceType) work only for C extension types,
> > right?
>
> No, they work with any object:
>
> >>> import operator
> >>> operator.isSequenceType([1, 2, 3])
> 1
> >>> class MyList(list): pass
> ...
> >>> operator.isSequenceType(MyList())
> 1

define "work" :

>>> class NotAList: pass
...
>>> operator.isSequenceType(NotAList())
1
>>> for item in NotAList(): print item
Traceback (most recent call last):
AttributeError: NotAList instance has no attribute '__getitem__'

(the isSequenceType predicate checks if the object's type
implements the C-level __getitem__ slot.  that's not always
what you want...)

</F>




From dalke at dalkescientific.com  Tue Apr 16 19:35:54 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 16 Apr 2002 17:35:54 -0600
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org>
Message-ID: <a9icab$phq$1@nntp9.atl.mindspring.net>

sean.bowman at acm.org:
>I'm wondering if there's a way to optimize code generated using the
>compiler module.  for example, if I say::
>
> import dis
> from parser import compileast, expr
>
> code = compileast(expr('x+3+7.'))
> print dis.dis(code)
>
>I notice that the constants aren't folded.  is there a way to do this, or
>do I have to fiddle with the ast myself?

Suppose you have

class Strange:
    def __add__(self, other):
        return self
x = Strange()

Then x+3+7.0 = (x+3)+7.0 = x+7.0 = x

The compiler can't know what 'x' might be so can't fold constants
in this case.  You could if it was '3+7.0+x'.

One advantage to Python is that incorrect constant folding has
never caused a bug in user code.  Compare this to C compilers where
I've had plenty of troubles with the optimizer.  The Python
implementation leans more towards simple than squeeze out performance
 -- the latter can be done either with hand optimization or switching
to C.
                    Andrew
                    dalke at dalkescientific.com





From ianb at colorstudy.com  Wed Apr 17 14:29:02 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Apr 2002 13:29:02 -0500
Subject: preferred syntax for list extend?
In-Reply-To: <a9jaah$493b$1@as201.hinet.hr>
References: <3CBCC196.466476CE@uniserve.com>  <a9jaah$493b$1@as201.hinet.hr>
Message-ID: <1019068142.1260.55.camel@localhost>

On Wed, 2002-04-17 at 03:01, Ante wrote:
> 
> <bvdpoel at uniserve.com> wrote in message
> news:3CBCC196.466476CE at uniserve.com...
> >
> > Just doing a bit of fooling around and I've found that if:
> >
> > q1=[1,2,3,4]
> > q2=[5,6,7,8]
> >
> > then:
> >
> > q1 += q2
> >
> > is about 15% faster than
> >
> > q1.extend(q2)
> >
> > Mind you, my test was with a relatively short list, don't know if longer
> > lists would act differently.
> >
> > I would have thought that using extend() would be preferred and faster?
> 
> Even if it was faster, I would still prefer +=.
> foo.extend(yada) doesn't look beautiful to me at all.
> After all, why do we have, e.g. == operator so we can say if a == b and
> not a.equals(b)? Because the former is simply better looking, and to me the
> same applies to += versus .extend() or even .append()

+= is different from .extend()... for instance:

q1 = [1, 2, 3]
q2 = q1
q3 = [4, 5, 6]

q1 += q3   # then...
q2 == [1, 2, 3]

But...

q1 = [1, 2, 3]
q2 = q1
q3 = [4, 5, 6]

q1.extend(q3)
q2 == [1, 2, 3, 4, 5, 6]

  Ian






From ak at silmarill.org  Fri Apr 19 15:11:30 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Fri, 19 Apr 2002 19:11:30 GMT
Subject: urllib2: pwd protected urls
Message-ID: <slrnac0r2p.ol1.ak@ak.silmarill.org>

Hello,

I can't figure out how to get pwd-protected urls using urllib2. Here's
what I'm doing:

>>> from urllib2 import *
>>> a = HTTPBasicAuthHandler()
>>> a.add_password("default", "host.net", "login", "password")
>>> o = build_opener(a)
>>> install_opener(o)
>>> f = urlopen("http://host.net/file.zip")

I get Error 401: Auth required

What's 'realm'? I set it to 'default' and I also tried None.


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From rdieter at math.unl.edu  Mon Apr  8 22:49:27 2002
From: rdieter at math.unl.edu (Rex Dieter)
Date: Tue, 09 Apr 2002 02:49:27 GMT
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com>
Message-ID: <XEss8.16251$GG1.1081902@news2.east.cox.net>

"Anthony_Barker" <anthony_barker at hotmail.com> wrote in message
news:899f842.0204081714.25eab798 at posting.google.com...
> Python 2.2x is stable why does redhat still insist on releasing 1.5x?
> All other distributions are now on 2.x.

This is for binary compatibility.  Be aware, that will include a separate
python2 package, which may interest you.

--
Rex Dieter
Computer System Administrator
Department of Mathematics and Statistics
University of Nebraska Lincoln




From hgs at dmu.ac.uk  Thu Apr 18 13:06:32 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 18:06:32 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4pu0x550y.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181803330.16636-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > cc -# -G build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so
>
> Alright, so it does link SSL (and SSL is likely in a useful version, also).
>
> > I can't tell if that worked:
> >
> > >>> import socket
        [...]
> >     from _socket import *
> > ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/build/lib.solaris-2.7-sun4u-2.2/_socket.so: symbol __eprintf: referenced symbol not found
>
> That is the problem: The SSL libraries refer to a symbol __eprintf,
> which originates from compiling the SSL libraries with GCC. GCC
> defines this symbol in libgcc.a, which is implicitly linked always.

OK,
>
> Now, when you link _socket.so with SunPRO, libgcc.a is not implicitly
> linked, hence the error.

Yes, that follows.
>
>
> > How old is acceptable ( before I disclose the version number publicly!!
> > :-) )
>
> It turns out that this theory was only partially true; the SSL version
> is fine, but the compiler the wrong one. You have the following

That (version) is encouraging.
> options (all alternative):
>
> 1. Build Python with GCC.
> 2. Rebuild OpenSSL with SunPRO.
> 3. Explicitly link libgcc.a to _socket.so, by editing Modules/Setup
> 4. Build without SSL support, by editing Modules/Setup.

I'll try out the configure patch, and go with 1.
>
> HTH,

Yes, it is a lot clearer now.
> Martin
>
        Thank you.
        Hugh





From tim.one at comcast.net  Wed Apr  3 21:26:40 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 03 Apr 2002 21:26:40 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <3CABABCA.EB446FEE@python.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEJPOKAA.tim.one@comcast.net>

[Guido van Rossum]
> ...
> Maybe you and a few others can get together and start working on the
> code?  Or do I have to do everything myself?

+1.  If you do everything yourself, that's a lot less for the rest of us to
do.  Don't be so selfish.

supportively y'rs  - tim





From cliechti at gmx.net  Tue Apr  2 17:31:49 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Apr 2002 23:31:49 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <3CAA03E4.A79BA593@noaa.gov>
Message-ID: <Xns91E4EF89BD4EDcliechtigmxnet@62.2.16.82>

Chris Barker <Chris.Barker at noaa.gov> wrote in 
news:3CAA03E4.A79BA593 at noaa.gov:
> WOW! fabulous post, Laura. As a result of mainly this post, I have
> changed my thinking somewhat.
> 
> Laura Creighton wrote:
> 
>> Python does not distinguish between True and
>> False -- Python makes the distinction between something and nothing.
>>
>> So I out-and-out tell people this.  {} is a dictionary-shaped nothing.
>> [] is a list-shaped nothing. 0 is an integer-shaped nothing.
>> 0.0 is a float shaped nothing.
>
> Thinking in terms of something vs. nothing is an excellent way to
> explain how Python treats these things, but I still have a problem with
> it. Who is to say that an empty list is nothing?

you can guide that in your classes with __nonzero__ (or __len__). if you 
don't care if the list is empty then make other tests like
>>> if var is None: ...

> or 0 for that matter. 0
> particularly: it is an integer, and a very useful one at that, I find it
> very inconvenient that 0 is always interpreted as nothing (see the
> string.find() example).

don't think of 'find' as a boolean operation. its result is an index, not a 
truth value. maybe it would be clearer if it returns None or raises an 
exception if nothing was found but its as it is now (the strstr C function 
return that values so it will come from there).

> In Python the only object I want consistently
> interpreted as nothing is None.

i find it especialy useful that other empty {list,dict,string} count as 
false. its a clear concept and it even saves you typing...
 
> That being said, the proposed "int-in-a-hat" (nice coinage, Laura), does
> not solve my problem at all. Everything that is currently interpreted as
> nothing will continue to be so interpreted, which is must be so as not
> to break probably every module over ten or so lines of code long.
> 
> The Python language is very much set in its ways: it relies on the
> concept of something vs. nothing, and it has pre-defined what nothing is
> for all the standard types in a very useful way. I'd be happier if that
> legacy didn't exist, but like many people posting, I feel this halfway
> measure is worse than what we have now.
> 
> Perhaps for Py3k, if we accept code breakage galor, a real Bool can be
> introduced, and we can get rid of de-facto definitions of particular
> values of an object being nothing. For now, let's keep it the way it is.

yes. but a real bool could be introduced as module for now (for the RPC and 
marshalling people that need one) i think its not needed otherwise becuse 
of the clean something/nothing logic wich only gets spoiled by a "backwards 
compatible" implementation.
we could watch how it evolves in the future. if its a clever thing it can 
become a builtin in py3k.

> -Chris

i can only agree with the rest :-) 

chris



-- 
Chris <cliechti at gmx.net>



From tim.one at comcast.net  Sun Apr  7 17:01:15 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 07 Apr 2002 17:01:15 -0400
Subject: ACCEPTED: PEP 285
In-Reply-To: <20020407134159.3680634c.wilk@flibuste.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEMBOLAA.tim.one@comcast.net>

[William Dod?]
> how many [active Python developers] their are ?

Guido controls two "full-time equivalents", which includes his own time on
Python, Barry's work on Mailman, and Fred's work on docs.  All other work is
volunteer (including, I'd wager, most PythonLabs work on Python).  You can
determine who was active at any given time, and what they were doing then,
by studying the checkin history:

    http://mail.python.org/pipermail/python-checkins/

The number of people who actually check something in in any given month
amounts to a handful, and some are "one issue" developers (like Jack Jansen
for Macintosh support); others are "almost one issue" developers (like
Marc-Andre Lemburg for Unicode support).  Martin von Loewis and Michael
Hudson are currently Python's most prolific "all issue" frequent
contributors outside of PythonLabs.  Michael is currently over his eyeballs
getting 2.2.1 ready, and only God knows how Martin manages to do all that he
does.

Hundreds and hundreds of people have contributed to Python over the years,
and it all adds up.  When the community wants an *ongoing* commitment to a
new scheme, though, they should realize that the pool of ongoing Python
contributors-- consistently contributing Python work week after week --is
tiny.  Even PythonLabs has weeks when it can't do Python work at all.






From phr-n2002a at nightsong.com  Thu Apr  4 10:27:28 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 07:27:28 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org> <3CAC19D8.142A918@sympatico.ca> <7xhemru1pg.fsf@ruckus.brouhaha.com> <3CAC5182.F6CB3F07@sympatico.ca> <7xzo0j4jzz.fsf@ruckus.brouhaha.com> <3CAC684C.E0C1B789@sympatico.ca>
Message-ID: <7xlmc3trfj.fsf@ruckus.brouhaha.com>

vio <vmilitaru at sympatico.ca> writes:
> Well, it's an ERP app. 100MB is a very rough figure. Actually not
> not all java source code, but a good part of it is. Obviously it
> grew out of control. Like some onion which keeps growing up layers
> upon layers. Though I'm sure half of it is all brackets and
> puctuation :-) Vio

There's a saying "inside every big program there's a small program
screaming to get out" that probably applies here.

You don't need a translation.  You need a rewrite.


From jb at cascade-sys.com  Sun Apr 28 18:28:00 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 28 Apr 2002 15:28:00 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no> <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>
Message-ID: <3CCC7770.336AA7CE@cascade-sys.com>

Steve Holden wrote:

> I only recently realised that the OAD makes an amazing omission, completely
> failing to provide a definition for the word "gullible".

Made me look.

Odd.  It's in my copy, page 290.

Whew.  Faith restored.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From joe.woodward at padtinc.com  Tue Apr 30 19:18:16 2002
From: joe.woodward at padtinc.com (Joe Woodward)
Date: 30 Apr 2002 16:18:16 -0700
Subject: Bulding arrays from text file data.
Message-ID: <28bbdc67.0204301518.5e7c41c8@posting.google.com>

Does anyone know of a faster way to build an array out of text file
with formatted data. What I have been doing works well, but now my
files and 200MB and up. The following way builds a Numeric.array
without knowing the size to start with. I can then resize it after the
fact.


yldpos=Numeric.array(map(float,re.split('\s+',string.strip(open('datafile.txt').read()))))

yldpos.shape=(len(yldpos)/5,5)


Thank you,

Joe Woodward 
PHoenix Analysis & Design Technologies


From ralph at inputplus.demon.co.uk  Mon Apr  8 06:57:05 2002
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 8 Apr 2002 10:57:05 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk> <mailman.1018248742.29568.python-list@python.org>
Message-ID: <a8rt21$8ec$2@inputplus.demon.co.uk>

Hi Lulu,

> > To calculate the major sort key you might do
> >     major = can_be_moved(o) + can_be_sized(o) + can_be_stacked(o) +
> >         can_be_coloured(o) + can_be_hidden(o)
> 
> Of course, I'm sure I'll be justly condmened for preferring:
> 
>     fs = (can_be_moved, can_be_sized, can_be_stacked,
>           can_be_coloured, can_be_hidden)
>     major = len([1 for f in fs if f(o)]))
> 
> to either :-).

I think you would be if all of fs are known to return 0 or 1 since the
list comprehension is less clear IMHO.  There's a lot more going on for
the reader to take in.


Ralph.



From jon+usenet at unequivocal.co.uk  Sun Apr 28 22:31:05 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 02:31:05 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no>
Message-ID: <slrnacpc38.obc.jon+usenet@snowy.squish.net>

In article <aai5oo$99l$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> Just realized that these Servlet based development models really limits the
> usage of software like Dreamweaver.  Since all code is in the program these
> utilities are almost worthless.  Is there a good way to combine Webware with
> templates.  I think of templates as pages where certain %words% are
> substituted etc...

This is exactly why I created jonpy (http://jonpy.sourceforge.net/).
It puts *no* Python code inside your HTML templates, you simply mark
sections using comments (e.g. <!--wt:newsitem-->) and replacements
using dollar signs (e.g. $$title$$). The idea is that even if you are
suddenly given an entirely new design with new HTML files with none of
the modifications you made to make the pages work, it will only take a
few minutes to fix them up again.


From aleax at aleax.it  Tue Apr  9 05:41:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 09:41:47 GMT
Subject: True/False story... (PEP 285)
References: <a8u99o$2lcu$1@news.nextra.cz>
Message-ID: <vHys8.62333$pT1.1792330@news1.tin.it>

Petr Prikryl wrote:
        ...
> you crazy ;) In other words, if the expression
> ``28+bbb()`` is used in the numeric context
> (determined by the plus operator) I would prefer

Good attempt, but there's a fundamental misunderstanding
of Python's worldview here (and in the rest of the post)
as evidenced by "numeric context determined by the plus
operator".

Python does not work by operators determining contexts:
that's *Perl*'s way, and Python is poles apart (and we'd
like it to stay that way).  + can mean numeric addition,
or concatenation of strings, or of list, or tuples --
that depends on the OBJECTS the operator + is applied to.

Python is object-oriented, not context-oriented.

Control structures (including the two that masquerade
as operators, 'and' and 'or') and certain statements
(such as print) do in some measure define a context
which may imply asking the object for type conversion:
iter(x) is implied by "for y in x:", bool(x) [when
bool is introduced] by 'x and y' -- note however that
in the latter case the bool() is only used in an
auxiliary way, since 'x and y' MUST return x itself,
NOT bool(x), if x is taken as true... so, viewing it
in terms of type conversion is nowere as useful as it
is for print (where str() is implicitly called AND
its result is indeed what gets used) and the for loop
(where iter() is implicitly called AND its result is
indeed what gets used).

I think the PEP's conceptualization as booleans as
a subclass of int is a better approach than one
hinging exclusively on context.  You say that booleans
cannot be conceptually compared to integers -- but,
there IS a natural mapping of booleans onto a subset
of integers, just as there is a natural mapping of
integers onto a subset of reals (or rationals, etc).

Not that it makes much practical difference at this
point, with PEP285 accepted anyway, but I'm now going
to switch to defending the pep's overall stance (even
though I'd have, a little bit, preferred it had never
happened:-) given that the alternatives that might
be still open are IMHO worse.


Alex



From lists.ASkwar at DigitalProjects.com  Thu Apr 25 18:35:36 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 00:35:36 +0200
Subject: python and very large data sets???
In-Reply-To: <20020424172403.GE20209@devel.trillke>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6o7l$nbi$1@panix3.panix.com> <20020424172403.GE20209@devel.trillke>
Message-ID: <20020425223536.GG21205@teich.Garten.DigitalProjects.com>

?holger krekel? sagte am 2002-04-24 um 19:24:03 +0200 :
> - setting up/configuring an appropriate database
>   for the task (may not be easy)

Hm, I don't think so.  Setting up a basic MySQL server without any
tunings is easy as pie, even on Windows.  But he'll of course need to
know SQL, at least in some simple forms - but SQL knowledge is often
needed, isn't it?

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 17 hours 33 minutes




From mal at lemburg.com  Fri Apr 26 04:07:25 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 26 Apr 2002 10:07:25 +0200
Subject: configuring mxCGIPython
References: <NCBBLFCOHHDIKCAFGCFBEEKHKMAA.marklists@mceahern.com>
Message-ID: <3CC90ABD.73350628@lemburg.com>

Mark McEahern wrote:
> 
> [Marc-Andre Lemburg]
> > Because Python always does on startup if not run with the
> > -S switch (this is not available in cgipython though).
> > Could you send the output of running cgipython
> > with the environment variable PYTHONVERBOSE defined to 1 ?
> 
> I've appended the output below.  It looks like it's stuck here:
> 
> # site.py
> 95: if os.name == "posix" and os.path.basename(sys.path[-1]) == "Modules":
> 
> evaluating sys.path.  Hmm?  It'll take me a while to figure out what's going
> on here.

Hmm, interesting. Looks like sys.path is empty at this point.
I wonder whether this is a bug in the frozen cgipython built
or just a consequence of freezing the interpreter.

Could you try setting PYTHONPATH to the dir where cgipython
lives ?

Since the stuff currently done in site.py doesn't really
make all that much sense here, I think cgipython should
probably use an empty site.py file instead of the standard
one (should enhance startup times even more).
 
> > A simple work-around would be adding the site.py from the
> > standard Python lib -- this should really be included in
> > cgipython, though, and I think it is, so there must
> > be something in site.py which fails.
> 
> If I put a site.py in the same path as cgipython, will that prevent it from
> reading its built-in one?  I need to experiment with that.

Good question. It probably doesn't even start looking for
it since it's already preregistered in the built-in 
modules.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From tejarex at yahoo.com  Mon Apr  1 02:18:44 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 01 Apr 2002 07:18:44 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017632403.21519.python-list@python.org>
Message-ID: <oRTp8.364099$uv5.30909820@bin6.nnrp.aus1.giganews.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1017632403.21519.python-list at python.org...
> [Terry Reedy]
> > ...
> > My proposal has two parts: do not use 'bool'; do use 'truth' (or
> > possibly something else such as 'binary' or even 'tval').  I am
> > curious what you think of the four reasons I gave.  I am
specifically
> > curious whether you will volunteer to anwer a multiyear stream
> > (already started) of questions about why Python truth values,
being
> > called bool, do not act 'properly' as bools 'should' and
suggestion
> > that they 'ought' to be restricted (crippled) so that they do.
>
> I think this is easy:  I point them at the C99 standard, whose
boolean is
> very much like Guido's conception, and end with "yup, same thing in
Python".

This is helpful information.  Since switching to Python, I haven't
tracked the new standard.

> Or just point them at the PEP.  The idea that the same people are
*not*
> going to complain about True+1 if the type is called "binary" (or
anything
> else) instead isn't credible to me.  They'll insist that True+1 is
evil just
> because part of it is spelled "True".  If the new tval type had
singleton
> constants Freeble and Goroblof, people would still complain about
> Goroblof+1, but not just because of the addition <wink> ...
Perhaps you are more realistically pessimistic than me and therefore
more relaxed about what can and cannot be done ;<).

>I expect they're triggered more by True and False than by "bool". ...

At this point, I will let those who are so triggered speak for
themselves.

Terry J. Reedy





From max at alcyone.com  Thu Apr  4 14:16:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 04 Apr 2002 11:16:28 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org> <slrnaan32v.n0n.philh@comuno.freeserve.co.uk>
Message-ID: <3CACA68C.C7B44901@alcyone.com>

phil hunt wrote:

> On 03 Apr 2002 10:59:13 -0500, Andrew Koenig <ark at research.att.com>
> wrote:
>
> >Specifically: If I want to write a function that answers a yes/no
> >question, I have lots of possible ways of spelling yes (1, 2, "yes",
> >and so on) and lots of possible ways of spelling no (0, {}, None,
> >and so on).  There isn't a single preferred way.
> 
> Yes there is, 1 or 0.

No, there isn't.  Some return 0 or 1, some return 0 or an unspecified
integer, some return None or an object, etc.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From mwh at python.net  Tue Apr 23 12:30:04 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 23 Apr 2002 16:30:04 GMT
Subject: Python call tree tool?
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>
Message-ID: <lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> I searched in vain for a tool that can create a call tree or  inverted
> call tree for python modules/packages. It seems inconceivable that the
> Python community hasn't created such a tool so I must be looking in the
> wrong places. 
> 
> Any one know where such a beast is?

There's Tools/script/trace.py which might or might not do what you
want.  Apparently it's in a bit of a sorry state, but I don't really
know how/why.

Cheers,
M.

-- 
  ... so the notion that it is meaningful to pass pointers to memory
  objects into which any random function may write random values
  without having a clue where they point, has _not_ been debunked as
  the sheer idiocy it really is.        -- Erik Naggum, comp.lang.lisp


From rjones at ekit-inc.com  Wed Apr  3 03:17:38 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 18:17:38 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <7xd6xhyzvx.fsf@ruckus.brouhaha.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017817388.30724.python-list@python.org> <7xd6xhyzvx.fsf@ruckus.brouhaha.com>
Message-ID: <200204030817.IAA06170@crown.off.ekorp.com>

On Wed, 3 Apr 2002 18:01, Paul Rubin wrote:
> Richard Jones <rjones at ekit-inc.com> writes:
> > > That looks pretty similar to an Info file (texinfo output).  I think
> > > I'd rather type "@section{Overview}" than have to supply those special
> > > underscores directly.
> >
> > Yeah, I guess at this point it's down to personal preference. I think we
> > can summarise by saying that you're happy to have your documentation
> > source in the texinfo format [#]_, whereas the Structure Text camp would
> > rather see it more looking like actual documentation. Is that fair?
>
> I'm not sure.  The PEP examines a bunch of different markup-based
> alternatives rather than comparing markup vs. non-markup.  The problem
> it finds with Javadoc source isn't that it's merely non-beautiful, but
> that it's actually quite hard to read without practice and HTML
> familiarity.  So I don't see it as a matter of a markup vs. a
> non-markup camp.  If there were a markup system that was easier to
> read directly, that might be a better answer than either Javadoc or
> ReST.  I'm pretty comfortable with Texinfo, for example.

How I see it is that ReST is a middle ground between markup and non-. It has 
markup, and you can use it to the extreme. Or you can follow some simple 
conventions (the most basic form of markup) and not worry about all the finer 
detail stuff. The difference between::

   @section{The Section Title}

and::

   The Section Title
   -----------------

Is pretty clearly to me that the second doesn't _look_ like markup, even 
though it is. It looks more like a convention (and a fairly common one, when 
you let the '-------' be any non-alphanum character).


> > > >    .. contents:: **Table of Contents**
> >
> > I think it's justifiable in this case - it saves a lot of typing :)
> >
> > Also, the other commonly used explicit markup is images::
> >
> >   .. image:: graphic.png
> >
> > ... that's pretty hard to do without an explicit markup (short of
> > ASCII-art'ing the PNG :)
>
> You might also want to add a directive for enumerated lists, so that
> if you insert a new item in the numbered list in your Roundup docs,
> you don't have to manually renumber the rest of them.
> [and some more extreme examples]

I don't believe that's what directives are _for_. They're used to perform 
tasks that can't reasonably be performed by simple markup. Embedding images, 
automatically scraping the document for a table of contents, defining a 
reference to an external document... But not for creating tables - there's a 
reasonable way to do that with simple markup. Well, that's how I see it, 
anyway :)


> That will help, but I thought the idea of a PEP was to present a complete
> specification of the proposed feature, which is why I didn't follow the
> external references too carefully.

I think the problem is that the ReST and greater DPS specifications are 
simply far too big to include in the PEP, so David's summarised what he 
thought were the pertinent points out of that large volume. I thought he'd 
done a pretty good job :)


> Btw, could those responding to me please just respond here in the
> newsgroup-- I don't need the email carbon copies since I'm reading
> here.  Thanks.

My apologies - I'll stop :)

Actually, that's it for me tonight - I gotta go. I'll read the thread 
tomorrow with interest :)


    Richard



From ark at research.att.com  Wed Apr  3 12:36:14 2002
From: ark at research.att.com (Andrew Koenig)
Date: 03 Apr 2002 12:36:14 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <016b01c1db31$8da51010$ced241d5@hagrid>
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
	<yu99wuvovkmm.fsf@europa.research.att.com>
	<00af01c1db2a$e7316d10$ced241d5@hagrid>
	<200204031619.g33GJXm02577@europa.research.att.com>
	<016b01c1db31$8da51010$ced241d5@hagrid>
Message-ID: <yu99ofh0vg4x.fsf@europa.research.att.com>

>> I mean that the language doesn't prefer a single way,
>> as opposed to programmers' conventions.

Fredrik> it's more than just a convention -- built-in functions and
Fredrik> operators always return 0 for false and 1 for true whenever
Fredrik> they need to invent a boolean value.  this is documented in
Fredrik> the language reference.

On the other hand, the library isn't nearly so consistent.  Just for
fun, I started browsing the library documentation at random.

The very first page I visited was the description of gc, the
garbage collector interface.  The first function that returns
a truth value there is described as

        isenabled()
            Returns true if automatic collection is enabled.

Some more examples:

        2.2.6.1 String Methods

        endswith(suffix[,start[,end]])
            Return true if the string ends with the specified suffix,
            otherwise return false.

        Similarly for isalnum, isalpha, isdigit, islower, isspace.

        2.2.8 File Objects:

        isatty()
            Return true if the file is connected to a tty(-like) device,
            else false.

        6.2 os.path -- Common pathname manipulations

        exists(path)
            Return true if path refers to an existing path.

I'm sure there are others.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark




From cliechti at gmx.net  Wed Apr  3 17:35:23 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 23:35:23 +0100
Subject: File opening
References: <a8frho$qqu$1@netnews.upenn.edu>
Message-ID: <Xns91E5F025911BFcliechtigmxnet@62.2.16.82>

zayats at blue.seas.upenn.edu (Salim Zayat) wrote in 
news:a8frho$qqu$1 at netnews.upenn.edu:

> Hey all.  I was wondering about opening files.  Now I know you need to 
> include the path to the file in the open statement.  Now the problems 
> arises here, though.  I am writing a module that reads in a config file 
> and creates buttons from it.  So if my module is run directly, the only 
> thing you need is 
> 
>> open('./buttons.config', 'r')
> 
> But if you wanna import the module, then the path to the open statement 
> needs to be full path to the file.  Is there a fix to this?

google brought up this:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52224 
scroll down to the comments box. others have had that problem too...



-- 
Chris <cliechti at gmx.net>



From jepler at unpythonic.net  Tue Apr 30 16:20:22 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 30 Apr 2002 15:20:22 -0500
Subject: os.environ.copy()['COPY_HUH'] = "not really"
Message-ID: <20020430152022.I17488@unpythonic.net>

Under Python2.1 and earlier (back to 1.5.2):
>>> import os
>>> os.environ.copy()['COPY_HUH'] = "not really"
>>> print os.environ['COPY_HUH']
not really

Under even 2.3a0 (CVS):
>>> import os
>>> os.environ.copy()['COPY_HUH'] = "not really"
>>> os.system("echo $COPY_HUH")
not really
0

My expectation is that modifying a copy of os.environ would not actually
modify the environment either in the current session or for children
created with os.system ...

dict(os.environ) seems to be the trick on versions of Python that have
it, os.environ.data on versions that don't.

Jeff




From philh at comuno.freeserve.co.uk  Mon Apr 15 19:26:10 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 00:26:10 +0100
Subject: pre-PEP for optional 'pass'
Message-ID: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>

OK, I've decided to write a PEP. Preliminary to this, I'd like 
people's comments.

I wish to make the pas startement optional. That is, in any Python 
program where there is a pass it can be left out and the meaning of 
the program is unchanged.

There are two rationales for this: convenience and beauty.


(1) Convenience

I would like to be able to write as code classes and functions 
without implementing the body to start with, e.g.:


class MyClass:
   def __init__(self, x, ,y, z):
   def getTwiddle(self):
   def setTwiddle(self, twiddle):


At the moment if I want to do this I have to put pass in and then 
take them out again. A minor inconvenience, to be sure, but one that 
niggles slightly due to the highly exploratory style of coding I 
often use.


(2) Beauty

In my opinion this change is Pythonic.

Consider the following code:

if condition:
   statement1()
   statement2()                                                                             
   statement3()                                                                             
else:
   statement4()                                                                             
   statement5()                                                                             
   statement6()                                                                             


If I remove the statement1() line, the program will still execute, 
and will behave as before except that staement1() is not executed. 
This is true if I then remeove the statement2() line. So the rule 
appears to be: if you have a block of n statements, you can remove 
one of them and the program behaves as before except not executing 
the one removed.

But if I then remove statement3() it is suddenly a syntax error.  
So blocks 1 statement long currently behave differently to blocks >1
statement long. I propose that this irregularity be removed.

In python, you can say:

   x = [ ]

specifying an empty list in the obvious way. You don't have to say

   x = [ empty ]

or:

   x = [ pass ]


or anything like that. I propose to bring groups of staements in 
line with that principle.


Thoughts?


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if




From mal at lemburg.com  Wed Apr  3 09:03:45 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 03 Apr 2002 16:03:45 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>  
	            <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com>
Message-ID: <3CAB0BC1.20B90FED@lemburg.com>

Guido van Rossum wrote:
> 
> > I haven't followed this thread for quite some time, but since the
> > PEP still seems alive,
> 
> :-)
> 
> > let me add some experience I've had with
> > using the already existing singletons (Py_True and Py_False)
> > in Python for recognizing truth values.
> >
> > Py_True and Py_False can be exposed in Python via
> >
> > True = (1==1)
> > False = (1!=1)
> 
> Yuck.  What a horrible implementation detail.

No really: comparisons should return booleans.
 
> > and most comparison operations and quite a few other APIs
> > returning truth values make use of these singletons.
> 
> This works purely by accident.  Given that at least Py_False is also
> known as &_Py_ZeroStruct, I'm surprised that the code that optimizes
> ints doesn't initialize small_ints[NSMALLNEGINTS] with Py_False.

Py_True and Py_False are statically allocated objects
which don't live on the heap. All the ints in small_ints
do, however.

_Py_ZeroStruct is probably just a naming inconsistency
(Py_True maps to _Py_TrueStruct).
 
> > I thought it would be a good idea to use those two
> > singletons for protocols which use booleans such as
> > XML-RPC. My experiences with this approach are, well,
> > not so good :-/
> >
> > The reason is that True and False are really integers
> > and not of a special new type. Now they are singletons,
> > which is good, since they represent the only states
> > a boolean can have and in many cases work reasonably
> > well as boolean representative, but not always
> > (e.g. arithmetic operations such as True - True == False).
> > Also, when trying to recognize the two singletons you have
> > to use the "is" comparison -- "==" will fail to
> > differentiate between 1 and True...
> >
> > def isboolean(x):
> >     return x in (True, False)
> >
> > ...doesn't work...
> >
> > def isboolean(x):
> >     return (x is True) or (x is False)
> >
> > ..does.
> 
> The correct way if PEP 285 is accepted would of course be
> isinstance(x, bool).

Let's say, that's the goal. PEP 285 is not necessarily
the complete answer to the problem ;-)
 
> > As a conclusion, I think it would be better to make bool() a
> > new type which does *not* inherit from integers, but which
> > does know how deal with other types which are commonly
> > used together with booleans such as integers. However, the
> > type should implement boolean algebra and not try to
> > mimic integer arithemtic, i.e. True - True raises an
> > exception.
> 
> That would break more code.  E.g. (a!=0) + (b!=0) + (c!=0) counts how
> many of (a, b, c) are nonzero; this would break.

Right and this is goodness, since your example is plain 
wrong if seen as boolean operation. Boolean algebra doesn't 
have +, -, *, / operations. 

The right way to write this would be:

	int(a!=0) + int(b!=0) + int(c!=0)

since you are counting values rather than finding truth
and then only if we define int(True) := 1 and 
int(False) := 0.

> > Py_True and Py_False should then be made singletons
> > of this new type (possibly after a transition phase which
> > replaces the singletons with a version that doesn't raise
> > exceptions but instead issues warnings).
> 
> That's what the PEP does (but with a different type).

Right, but my point is that if you want to make booleans
first class objects in Python, you should consider making 
them behave as defined in the text books rather than 
trying to make them feel like integers but look like
truth values.

> > This may sound painful at first, but in the long run,
> > I believe, it'll result in the same benefits as other
> > painful changes have or will (e.g. the change from integer
> > division to floating point division).
> 
> I don't see the introduction of the bool type as painful at all.
> 
> I find getting it accepted painful though. :-(

How about adding a new pure bool type to the language
*without* changing any of the old Py_False/Py_True
code (at first) ?!

Then leave the code in for one release to see how
people feel about it in real life and enable a warning
for incorrect usage of comparison arithmetic in a second 
release. The third release would then enable the new
logic per default.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/



From barry at zope.com  Mon Apr  8 10:57:38 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Mon, 8 Apr 2002 10:57:38 -0400
Subject: Scheduled maintenance for all mail at python.org & zope.og
Message-ID: <15537.45026.621346.249177@anthem.wooz.org>

We plan on doing another upgrade to the mail server mail.python.org,
a.k.a. lists.zope.org on Wednesday 10-Apr-2002, at 11am -0400 (EDT,
US/Eastern).  We will be upgrading to Exim 4, and this will affect all
mail and mailing lists on the python.org and zope.org domains.

Downtime should be minimal; if all goes well, just 15-30 minutes.  You
shouldn't even notice it.  If you still cannot get email through by
noon our time, check the www.python.org home page for status, or email
Barry (barry at zope.com), Martijn (mj at zope.com), and Greg
(gward at mems-exchange.org).

-Postmaster @ {python,zope}.org




From news at 92tr.freeserve.co.uk  Wed Apr 10 19:40:52 2002
From: news at 92tr.freeserve.co.uk (ddoc)
Date: Thu, 11 Apr 2002 00:40:52 +0100
Subject: Can I print from Python?
References: <slrnab0n45.rrn.ak@ak.silmarill.org> <20020407110524.09532.00002024@mb-cs.news.cs.com>
Message-ID: <a92io4$331$1@news7.svr.pol.co.uk>

Jim Williams wrote:
> Ahhh..that sheds a little light then.  probably could do something like
> os.system("copy file.txt lpt1") maybe.  Thanks will experiment with it.

Remember the basic Windows applications accept a comman line switch to 
print a file so you can make a rich text file and then shell Wordpad with 
/p file.rtf    to print it.

-- 
A


From aahz at pythoncraft.com  Tue Apr  9 18:27:53 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 18:27:53 -0400
Subject: Thread/Process-safety of functions in the Standard Library, eg the dbm classes
References: <3cb32a63.17540718@news.t-online.de>
Message-ID: <a8vpt9$6os$1@panix1.panix.com>

In article <3cb32a63.17540718 at news.t-online.de>,
Gerson Kurz <gerson.kurz at t-online.de> wrote:
>
>I wonder if it is documented which functions of the Python Standard
>Library are threadsafe, and which not. More explicitly, I was having a
>look at the anydbm interface and am pretty sure that it is not safe to
>access from multiple-processes at the same time, and I don't even know
>whether or not it is safe for multiple threads.
>
>So, is there such a documentation? 

Use an extra thread to handle I/O for each dbm you open, then use Queue
to handle the inter-thread communication.  See my web page for some
examples.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From ngps at vista.netmemetic.com  Mon Apr 15 10:05:04 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 15 Apr 2002 14:05:04 GMT
Subject: httplib drops ssl connection to MS IIS Server
References: <mailman.1018741166.19994.python-list@python.org>
Message-ID: <a9emmg$ijm$1@dahlia.singnet.com.sg>

According to Marco Caamano  <mvcc_cr at yahoo.com>:
> I've been experiencing some problems with ssl connections using httplib.
> When executed it connects to the server and then drops on the following 
> error:
> socket.sslerror: (5, 'EOF occurred in violation of protocol')
> ----------------
>
> I'll really apreciate a pointer here, what I am doing wrong? what should
> I try next?

M2Crypto:

ngps at vista:~$ python
Python 2.2 (#7, Mar 31 2002, 23:17:41) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> from M2Crypto import httpslib
>>> body = '\202test' 
>>> host = ('209.154.200.218:443')
>>> hs = httpslib.HTTPS()
>>> hs.set_debuglevel(4)
>>> hs.connect(host)
>>> hs.putrequest('POST', '/scripts/gateway.dll?Transact')
send: 'POST /scripts/gateway.dll?Transact HTTP/1.0\r\n'
>>> hs.putheader('Content-Type','x-Visa-II/x-auth')
send: 'Content-Type: x-Visa-II/x-auth\r\n'
>>> hs.putheader('Content-Length', str(len(body)))
send: 'Content-Length: 5\r\n'
>>> hs.endheaders()
send: '\r\n'
>>> hs.send(body)
send: '\x82test'
>>> status, reason, header= hs.getreply()
reply: 'HTTP/1.1 200 OK\r\n'
header: Server: Microsoft-IIS/4.0
header: Date: Mon, 15 Apr 2002 14:18:46 GMT
header: Content-Type: x-data/xact-error
>>> data=hs.getfile().read()
>>> print data
510 - Authorization request must be an Interleave (application type 4).

>>> print status
200
>>> 

Get M2Crypto here:

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


I'll soon write a bot to auto-answer such questions. ;-)


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



From henrik.motakef at web.de  Mon Apr  8 15:57:37 2002
From: henrik.motakef at web.de (Henrik Motakef)
Date: 08 Apr 2002 21:57:37 +0200
Subject: PyQt question (connecting signals)
References: <3cb1ed1f$1_3@news.newsgroups.com>
Message-ID: <8766327yku.fsf@apo.hmotakef.int>

"Janos Blazi" <jblazi at hotmail.com> writes:

> I have a very simple application. Can I connect a signal from a QPushButton
> to a normal python function without using classes at all?

Well, you will at least have to use the QPushButton class ;-)

However, try something like:

QObject.connect(myButton, SIGNAL("clicked()"), myFunction)

(This is documented, btw. See 
<http://www.riverbankcomputing.co.uk/pyqt/docs/x156.html>)

hth
Henrik


From jwickard**nospam** at litriusgroup.com  Wed Apr  3 15:29:41 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Wed, 03 Apr 2002 20:29:41 GMT
Subject: os.remove() permissions problems
Message-ID: <VCJq8.208$c6.57043@typhoon.mn.ipsvc.net>

I'm trying to process some files on my system.  when I run a script with
os.remove() in it I'm denied permission to delete the file.  I've created
these files as the same user that I'm running the script as.  This is on a
win2k box, when the python interpreter runs, does it run as it's own
unprivleged user?




From cliechti at gmx.net  Fri Apr  5 17:31:20 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 5 Apr 2002 23:31:20 +0100
Subject: The Topy project
References: <mailman.1018039992.15896.python-list@python.org>
Message-ID: <Xns91E7EF76BA86Acliechtigmxnet@62.2.16.82>

pinard at iro.umontreal.ca (Fran?ois Pinard) wrote in
news:mailman.1018039992.15896.python-list at python.org: 

> Hi, people.  I merely feel like chatting a bit. :-)
[long talk about per2py and scheme2py]

this reminds be to GCC and RTL. the GNU-C compiler translates the
source to an intermediate language called RTL. this intermediate code can 
then be optimized with arithmetical rules and finaly its output to 
assembler/bytecodes (many diffrent dialects for a dozen processors).
there are diffrent fontends to genereate RTL for diffrent languages as C, 
C++, Java and some other.

thats in some ways the same problem as yours...

i think the ideal soloution would be to translate the source language to a 
meta language that describes the algorithm and its properties. then this 
cookbook could be used to generate code in any other language - however i 
don't think that such an aproach is realy possible with todays 
technologies...

chris

-- 
Chris <cliechti at gmx.net>



From dkuhlman at rexx.com  Sat Apr 27 17:35:01 2002
From: dkuhlman at rexx.com (kuhlman)
Date: Sat, 27 Apr 2002 14:35:01 -0700
Subject: Python as scripting plug-in for large C++ project
References: <mailman.1019880409.19217.python-list@python.org> <3CCAF576.3060304@dittmar.net>
Message-ID: <aaf5hb$a6vgj$1@ID-139865.news.dfncis.de>

SWIG is great.  But, if you decide to write your glue code by hand, 
you may want to implement your own Python data types.  It's really 
not that hard. If you decide to do this, be sure to look at 
Objects/xxobject.c in the Python source code distribution.  It 
gives a template for implementing new data types.  And for hints 
and clues, also look at the implementation of some other the other 
Python data types (lists, strings, etc) in the Objects subdirectory.

Also see "2. Defining New Types" in "Extending and Embedding the 
Python Interpreter" in the Python standard documentation.

And, if you have *lots* of Python type extensions that you want to 
implement, you might even want to look at my dtGenerator.py, which 
will generate the implementation of a Python type from a 
specification.  Although, you will probably want to implement one 
type extension in C by hand, and then adjust the templates in 
dtGenerator.py for your own use.  dtGenerator.py assumes that there 
is an underlying C data type (and a struct) that you want to 
"wrapper".  You can find dtGenerator.py at 
http://www.rexx.com/~dkuhlman .

  - Dave


Daniel Dittmar wrote:

> Bjorn Pettersen wrote:
> 
>>  - We extensively (exclusively) use our own basic datatypes,
>>    e.g. NInteger which in addition to keeping a value also
>>    stores whether it is NULL (think databases). How much work
>>    is it going to be to expose these to embedded Python?
> 

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From roy at panix.com  Sun Apr 14 14:11:42 2002
From: roy at panix.com (Roy Smith)
Date: Sun, 14 Apr 2002 14:11:42 -0400
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <roy-33A2C0.14114214042002@news1.panix.com>

martin at v.loewis.de (Martin v. Loewis) wrote:
> As Bengt points out, it is tricky to define this analogous to 'and'
> and 'or'. Both have a shortcut meaning, evaluating the second one only
> if the first does not determine the outcome.

Well, you could look at it as "you only evaluate as many operands, from 
left to right, as you need to determine the value of the expression".  This 
describes the current behavior of "and" and "or", and with only a small 
stretch, could be said to describe "xor" as well.  This may be a little far 
fetched, but, how about:

x = {'foo': 0}
y = x['bar'] xor x(0)

would be guaranteed to raise a KeyError and not a TypeError, since the 
left-to-right evaluation would ensure that x(0) was never evaluated.


From tim.one at comcast.net  Sun Apr 14 17:24:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 17:24:52 -0400
Subject: Does Python need an 'xor' operator?
In-Reply-To: <eppstein-7156A5.13521614042002@news.service.uci.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEBPOOAA.tim.one@comcast.net>

I doubt that a new "xor" operator will ever get in (compelling use cases are
conspicuous by absence), but anyone pushing for it had better push for a new
"eqv" operator too (that is, if you're going to claim that it's ugly that
x^y may not return one of its argument, surely it's just as ugly that the
same holds for x==y).





From aahz at pythoncraft.com  Mon Apr  1 11:05:49 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 11:05:49 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de> <KM_p8.41075$pT1.1129125@news1.tin.it>
Message-ID: <a8a0gt$roc$1@panix1.panix.com>

In article <KM_p8.41075$pT1.1129125 at news1.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>0 is (2.2 is still present, not yet past) the canonical way to spell 
>untruth: it's what 'not', 'is', 'in', comparisons, and all standard library 
>iswhatever's return -- including operator.truth, of course.

Just to throw in a datapoint, I generally use None to indicate falsehood.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From sholden at holdenweb.com  Mon Apr 29 08:54:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 08:54:06 -0400
Subject: Update locals()
References: <20020427195725.GA16217@devel.trillke> <20020428205757.GH16217@devel.trillke> <02042900255602.25256@arthur> <mailman.1020042230.9019.python-list@python.org> <q7bz8.76797$vF6.2277789@news2.tin.it>
Message-ID: <Sqbz8.7645$M63.2673@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:q7bz8.76797$vF6.2277789 at news2.tin.it...
> <posted & mailed>
>
> holger krekel wrote:
>         ...
> > OF COURSE i didn't mean that re.compile mucks with the local namespace.
>
[ ... ]
>
> > If you are coming to EuroPython maybe we find half an hour where we
> > can discuss it on the screen? Doing this in mail makes it hard to not
> > misunderstand each other.
>
> [...] (it seems to
> me that the inability to understand what you can possibly mean is
> turning out to be frequent on my part -- just to double check, is
> everybody else perfectly clear about what this "on the screen" is
> meant to convey, or is it just as murky to everyone?).  [...]

A simple thinko for "off-screen" would be my vote, if this were a vote,
which it isn't. Done it many a time myself. Sheesh, and people are telling
*me* to lighten up! (Hi, Alex, sorry the notes are late again).

I really envy you guys your trip to EuroPython, it's enough to make a chap
want to live in Europe again. I wish the INS processes went a bit faster,
but about now isn't the time to start complaining.

and-don't-even-*ask*-me-to-do-it-in-a-foreign-language-ly y'rs  - steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From max at alcyone.com  Sat Apr 13 05:59:46 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 13 Apr 2002 02:59:46 -0700
Subject: ANN: Munch 0.1 -- Embed Python in text as markup
References: <3CB75C38.757A849D@alcyone.com> <slrnabg2ai.g9m.mlh@vier.idi.ntnu.no>
Message-ID: <3CB80192.C4C4AC1B@alcyone.com>

Magnus Lie Hetland wrote:

> In article <3CB75C38.757A849D at alcyone.com>, Erik Max Francis wrote:
> [snip]
> 
> If you swap '$' for '@' this is quite similar to the Cheetah template
> system. Have you looked at that?

No, although I did come across PTML in my initial searches.  I see that
Cheetah uses $ for embedding variable names as markup, but uses a
combination of # and $ to set off code, e.g.

	#for $client in $clients

Munch is simpler than this; it isn't designed to be embedded halfway.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From shredwheat at attbi.com  Wed Apr 24 11:20:57 2002
From: shredwheat at attbi.com (Pete Shinners)
Date: Wed, 24 Apr 2002 15:20:57 GMT
Subject: linux console graphics
References: <3CC682A5.3050905@ifrance.com> <eu36aa.2gm.ln@127.0.0.1>
Message-ID: <3CC6DBDB.1030707@attbi.com>

Radovan Garabik wrote:
> eav <eav at ifrance.com> wrote:
>  : I am currently working on a project needing a graphical interface, 
>  : supposed to run quickly on quite modest PC (first pentium generation) 
>  : with linux or a *bsd.
> 
>  : What can Python do form me ? I've been considering projects such as 
>  : pygame - but it seems that SDL can't be run from the console. Am I wrong?
> 
> it can. In svga or framebuffer mode, or even aalib :-)
> The question is if pygame supports it - but I think it should.

pygame supports it as well as SDL will support it. just make sure the 
environment variable SDL_DISPLAYDEVICE is set before initializing 
anything in pygame.

this may answer some questions, from the sdl linux faq
http://www.libsdl.org/faq/FAQ-Linux.html#LINUX_5



From tim.one at comcast.net  Mon Apr 22 19:41:22 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 22 Apr 2002 19:41:22 -0400
Subject: calendar.py doesn't use lists???
In-Reply-To: <fO0x8.82707$T%5.833@atlpnn01.usenetserver.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDNPAAA.tim.one@comcast.net>

[Steve Holden]
> ...
> Well, reading the source gives co clue to the guilty party. We could
> find out in SourceForge if we really wanted. It wasn't me, honest.

It started as a user-contribued patch, and everyone should be relieved to
hear that we don't nominate users for Usenet abuse just because they're
trying to make everyone's lives better by contributing to Python <wink>.
IOW, I'm not telling.

> ...
> Well, let's take a look at the documentation, shall we. In "5.13
> calendar -- General calendar-related functions" for Python 2.2 there is
> no mention of these attributes. Nothing in 2.1 or 1.5.2 either. So
> offhand I'd say the author was free to replace those sequences with
> somethingbearingnodamnedresemblanceatalltolists without anyone having
> any right to complain.

Guido wouldn't say that, though, and didn't.  He was adamant that we do
whatever it took to make these undocumented thingies "act like" the old
static lists so far as was possible.  That couldn't be a complete emulation,
as the contents can change dynamically on every access (depending on locale
setting).

> ...
> There's been a lot of complaining about language change recently, but
> precious little hard evidence of real code breakage.

This was real code breakage, despite that Aaron was mucking with
undocumented interals.  We don't *want* to break stuff like that either.
But even pure bugfixes can break somebody's code, and sometimes they do.

> ...
> I have to say in this case that the code breakage you report is not
> Python's fault (though the crash and burn behavior should be fixed:
> segfaults are a bit OTT).

It turned out there's more than one cause for the crash-and-burn.  On some
platforms (not most), the crash happens in the bowels of the platform's
strftime() implementation.  So the locale change tickled platform-specific C
library bugs too.  2.2.1 also worms around those, but you really can't
change anything without breaking something for someone.

you'd-*almost*-think-programmers-would-have-learned-this-from-
    their-own-code<wink>-ly y'rs  - tim





From johnroth at ameritech.net  Sat Apr 20 07:03:50 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 20 Apr 2002 07:03:50 -0400
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <uc14cjalpru3b5@news.supernews.com> <a9q8f4$svn$1@aphrodite.grec.isp.9tel.net>
Message-ID: <uc2io69bm63ge9@news.supernews.com>

"Richard Gruet" <rjgruet at yahoo.com> wrote in message
news:a9q8f4$svn$1 at aphrodite.grec.isp.9tel.net...
>
> "John Roth" <johnroth at ameritech.net> wrote in message
> news:uc14cjalpru3b5 at news.supernews.com...
> >
> > "Richard Gruet" <rjgruet at yahoo.com> wrote in message
> > news:a9nbuq$916$1 at aphrodite.grec.isp.9tel.net...
> > > Alex
> > >
> > > And BTW, people would not to have implement any __len__ method if
all
> > the
> > > sequence types would derive from a common ancestor (sequence), in
> > which the
> > > (in)famous method would be implemented. Would it be completely
> > nonsensical ?
> > > I have been told that starting from 2.2, new style classes
(including
> > most
> > > built-in types) all derive from a common base class 'object'.
> >
> > As far as I know, that's correct. Inserting a .len() method in the
> > object class would automatically make it available to all new style
> > classes.
> >
> > As far as I can see, there is one major issue with this: if the
> > class in question didn't have a __len__() method, it would
> > still have the (inherited) len() method. This could be rigged to
> > raise an exception, but it would still be a bit of a trap for the
> > unwary.
> >
> > There's also the question of the inheritance hierarchy, which I
don't
> pretend to understand.
> > It seems to be a case of simple implementation, small benefit, small
> downside, no discernable side effects. Write a PEP.
> >
> > John Roth
>
> Actually, as Alex pointed out, a single len() method in a hypothetical
> 'sequence' base class would be mythical, simply because there is no
common
> implementation of len() for all kinds of collection (I just forgot
this!).

It doesn't matter. As I clearly implied, the default .len() method
would call the class's .__len__() method. Therefore, there is no
need for a common implementation.

> However we could still define a sort of abstract len() method  in a
> 'sequence' (or 'collection') base class, that has to be implemented in
> derived classes (or will raise an exception otherwise).

Why bother?

AFIC, it's still a yawner. Small benefit, small cost.

John Roth




From tim.one at comcast.net  Mon Apr  1 15:00:51 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 01 Apr 2002 15:00:51 -0500
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
In-Reply-To: <slrnaaha69.nn5.kzaragoza@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCMELGOJAA.tim.one@comcast.net>

[Kris J. Zaragoza]
> I have to ask one question that I haven't seen brought up yet: What
> does PEP 285 imply for the future of Python?  Do you intend to
> eventually change the concept of what is true?
>
> Specifically, I refer to the Python Language Reference
> (http://www.python.org/doc/2.2/ref/lambda.html):
>
> """In the context of Boolean operations, and also when expressions are
> used by control flow statements, the following values are interpreted
> as false: None, numeric zero of all types, empty sequences (strings,
> tuples and lists), and empty mappings (dictionaries). All other values
> are interpreted as true."""

I'll channel him:  he intends to add False to that list.  That's all.

> ...
> If you don't intend to modify this concept, then what's the real gain
> for the new boolean type?

Have you read the PEP?  I can assure you he's not hiding a killer good
reason he hasn't already mentioned <wink>.

> ...
> (Try explaining to a newbie that it's actually one of two
> pre-allocated instances!)

Why would a newbie need to know that?  They seem to get along fine without
realizing that None is the unique instance of NoneType, and, unlike as for
None, it should be rare to see code relying on unique object identity for
True and False.




From jajvirta at cc.helsinki.fi  Mon Apr 22 11:01:32 2002
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 22 Apr 2002 15:01:32 GMT
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83> <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi> <slrnac84f4.1c6.philh@comuno.freeserve.co.uk>
Message-ID: <slrnac89ec.r20.jajvirta@sirppi.helsinki.fi>

Mon, 22 Apr 2002 14:36:36 +0100 phil hunt wrote:
> On 22 Apr 2002 10:50:37 GMT, Jarno J Virtanen <jajvirta at cc.helsinki.fi> wrote:
>> [on reliable software compontents]

> Data structures such as lists and dictionaries.
> 
> Built-in libraries such as those for windowing, TCP/IP 
> communication, etc.
> 
> Automatic storage management.

Yes, I agree. These components (or whatever one would like to call them)
are a significant contrubution and this is where Python rocks and rules.
Especially the standard library part that *as included to the language
distribution itself* is a major advantage. But still, there's a limit
where a language and its modules can take you. Major part of the
complexity is still in the so called business logic which implements
the connections and dependencies of the actual domain (or even the
unique domain of the customer organization or whatever it happens to
be). 

It seems to me that the "batteries included" part of language
relates strongly to some technological aspect of the programming, ie.
handling storages, handling UIs, speaking with specific protocols
and so on. It definitely makes software development easier but 
not by far makes it redundant. 

> Imagine having to write a mailer or newsreader without any of them.

I prefer not to imagine, but yes, I got the point. ;-)



From jkraska at san.rr.com  Wed Apr  3 20:39:41 2002
From: jkraska at san.rr.com (Courageous)
Date: Thu, 04 Apr 2002 01:39:41 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <mailman.1017869530.3706.python-list@python.org>
Message-ID: <b5cnaucrm55rpm8kgt0nibnd4lr43neljs@4ax.com>

>But... I would aslo suggest that if you do decide to
>automate part of the transition, you might want to take a
>look at ANTLR...
>
>   http://www.antlr.org/ 

I recently built a whole Python lexer and parser with ANTLR.
I played around with about a half dozen compiler-compilers
while trying to find one that would support my needs. In my
opinion, ANTLR is the _top_ of the pack of all the freely
available compiler/translator generators.

For one thing, the lexer, parser, tree walker, and ast-
generator are all unified into one meta language. This is
a very, very clean implementation. It's a full suite all
in one spot, and fully deserves all the accolades that it
gets.

C//



From mlh at vier.idi.ntnu.no  Wed Apr 10 22:18:12 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 02:18:12 +0000 (UTC)
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
Message-ID: <slrnab9sj4.6g8.mlh@vier.idi.ntnu.no>

In article <3cb4bd45$1_1 at goliath.newsgroups.com>, Brad Clements wrote:
>Could someone explain this to me.
>
>Lots of code does  type(x) in [InstanceType]

First solution: Don't do this. Type checking is bad.

Why do you have to do this? Could you give an example?

>One might imagine that to support new style classes, this would be changed
>to:
>
>type(x) in [InstanceType,ObjectType]
>
>
>But that doesn't work because type(new style instance) returns "Class"

>This seems the best way:
>
>>>> isinstance(y,ObjectType)
>1
>>>>

Well...

>>> isinstance(1, ObjectType)
1

Sorry :)

And in case you hadn't noticed:

>>> ObjectType is object
1

>I'm hoping someone can illucidate this for me.

What is it you want to detect with your type checking? Whether
something is an "object" or not is pretty meaningless with new-style
classes... (Everything is.)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From netzapper at magicstar.net  Tue Apr  2 16:51:56 2002
From: netzapper at magicstar.net (A. Jones)
Date: Tue, 02 Apr 2002 21:51:56 GMT
Subject: Can't find Python2c
Message-ID: <3caa27bc.5645718@mammoth.usenet-access.com>

I've looked all over for an active site that has a copy of Python2c,
however, I'm unable to find it.

The official site seems to be down, and the Vaults of Parnassus simply
links to that official site.

Searching revealed nothing...

Any idea where I might find a mirror of it?


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From phr-n2002a at nightsong.com  Wed Apr  3 02:45:51 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 23:45:51 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <mailman.1017794173.28885.python-list@python.org> <B8D00218.20D57%goodger@users.sourceforge.net> <7x4rituw82.fsf@ruckus.brouhaha.com> <mailman.1017816546.28826.python-list@python.org>
Message-ID: <7xlmc5z0ls.fsf@ruckus.brouhaha.com>

Richard Jones <rjones at ekit-inc.com> writes:
> > Since the PEP is being presented to the whole community and not just
> > to the Doc-SIG (I'd never heard of the Doc-SIG til just now), it
> > should make its case in detail rather than just giving the conclusion.
> 
> The PEP is already very long. Did you want David to include _all_ the 
> documentation from the ReST website in the PEP to back it up?

If the proposal is good, the PEP can be written more persuasively
without lengthening it excessively.  

> > That includes the case for having structured doc strings at all--do
> > you have formatting tools that read these doc strings and produce
> > printed output like POD?
> 
> I believe the docstring extraction is performed by pydoc and is a little 
> manual at the moment, but the upshot is yes, he does (printing of HTML is 
> valid at this early stage of development).

OK, that's a start.

> > I've written some pretty big Texinfo documents and find Texinfo to be
> > readable enough in source form, though it looks better after
> > formatting.  reStructuredText looks pretty simliar to the output of a
> > Texinfo formatter--maybe Texinfo could be adapted to produce
> > reStructuredText instead of Info files.
> 
> But the whole point of this is that the _source_ of the documentation is the 
> reStructuredText-formatted text, not some other more arcane format that must 
> be processed to produce reStructuredText files.

I guess I don't see plain english markup commands as being more arcane
than remembering what all those funny characters are supposed to mean.


From prova at prova.it  Sun Apr 28 11:24:52 2002
From: prova at prova.it (Utente)
Date: Sun, 28 Apr 2002 15:24:52 GMT
Subject: Help: py2exe and os.path
References: <JlPy8.72016$vF6.2149385@news2.tin.it> <Xns91FE6B9995466gmcmhypernetcom@199.171.54.213>
Message-ID: <8vUy8.29577$8D3.789907@news1.tin.it>

----- Original Message -----
From: "Gordon McMillan" <gmcm at hypernet.com>
Newsgroups: comp.lang.python
Sent: Sunday, April 28, 2002 4:34 PM
Subject: Re: Help: py2exe and os.path


[...]
> From import's viewpoint, there is no module named os.path, so
> don't import it.
>
> When you import os, it locates the appropriate module for doing
> path manipulations on your platform, and makes this available
> as "path" (could be posixpath, macpath...). Since it's a global
> of os, it's available as os.path.
>
> -- Gordon
> http://www.mcmillan-inc.com/

Thanks! You're absolutely right!!!
In fact, i disseminated my source with message boxes, (how else could i
debug the exe???) and i discovered that it behaves correctly. If the file
exists, it does some things, if the file doesn't exists it does other
things. In the beginning i thought the problem was the os.path.isfile()
function because i always had the same results (data not loaded and the
error message there).
Now, i discovered that my program stops when i try to load a dictionary via
the shelve module (i posted another help request for that... :) ). The
strange thing is that (again) it works perfectly with the interpreter, but
not with py2exe...

But, at least, now i know where is the real problem.
Thanks again,
Max





From flz at cfdrc.com  Mon Apr  1 12:29:48 2002
From: flz at cfdrc.com (gother)
Date: 1 Apr 2002 09:29:48 -0800
Subject: Initialization problem of shared library in HP-UX11
Message-ID: <7aafbf19.0204010929.62d17d6a@posting.google.com>

I got into the following situation in the HP machine with gcc 3.0.3
compiler

myStr.h
-----------------
class MyStr
{

public: 
  char * str;
  Mystr();
  MyStr(const char *);
  MyStr& operator=(const char *);
}
=====================================
Var.h
----------
extern MyStr mystr;
extern int myint;
=====================================
var.cpp
---------------
MyStr mystr = "xyz";
int myint = 1;
====================================
myPrint.h
---------------
extern void f();
===================================
myPrint.cpp
---------------
...
void f()
{
  printf("%x\n",mystr.str);
  printf("%d\n",myint);
}
==================================


Made those available to python by Swig

MyTest.i
----------------
%module MyTest
%{
#include "Var.h"
#include "myStr.h"
#include myPrint.h
%}
%include myPrint.h


The problem is the global variable (user defined class) "mystr" never
initialized when I import the shared library into python, but int
global variable is properly initialized.

When I linked the shared library to a main function, user defined
class variable "mystr" got properly intialized.

Anybody know how to resolve this problem?

Thanks

Gother


From cepl.m at neu.edu  Sun Apr 28 07:52:27 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Sun, 28 Apr 2002 11:52:27 +0000 (UTC)
Subject: IMAP2IMAP Synchronization -- iis [Was: Re: iis: comparsion of two lists]
In-Reply-To: <yqNy8.71678$vF6.2141429@news2.tin.it>
References: <20020426224924.GA1318@rehor.surfbest.net> <20020426232802.GA1927@rehor.surfbest.net> <JTry8.26517$8D3.706486@news1.tin.it> <20020428034636.GB1279@rehor.surfbest.net> <yqNy8.71678$vF6.2141429@news2.tin.it>
Message-ID: <20020428115225.GA999@rehor.surfbest.net>

On Sun, Apr 28, 2002 at 07:22:06AM +0000, Alex Martelli wrote:
> the first assignment is useless.  Not an issue of performance, but of
> Occam's razor: entities must not be multiplied needlessly, and a
> statement that is totally useless is is a good candidate for pruning.

OK, I recant, these needless assignements will go. Thanks.
 
> My favorite would be to forego the hard-to-read lambda for a named
> function (the name injection is inevitable in 1.5.2 or 2.0):
> 
> def not_in_loc_old(m, loc_old=loc_old): return m not in loc_old
> erase = filter(not_in_loc_old, proc_old)

OK, that's much more sane, you are correct.

> It's silly to use two operators in expressions such as
>         not (x in y)
> when a single operator called not in suffices:
>         x not in y

Thanks, I did not know it.

> likely to produce the speed bottleneck unless the lists involved
> are very, VERY small.

Probably up to 100 elements each?

> For membership checking, prepare, not lists, but dictionaries,
> and (in 1.5.2 and up to 2.1) use has_key.

Great idea.

> By modestly increasing elegance, we should also get substantial
> increases in performance unless your sets are all tiny (in which
> case performance doesn't matter all that much).  Try -- performance
> issues are always to be checked by measurement.

Thanks a lot, that's what I was exactly interested in.

Matej

-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
How many Bavarian Illuminati does it take to screw in
a lightbulb?
Three: one to screw it in, and one to confuse the issue.



From max at malva.com.uaREMOVE.IT  Tue Apr 16 02:56:54 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Tue, 16 Apr 2002 09:56:54 +0300
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs> <slrnablo96.bcr.philh@comuno.freeserve.co.uk>
Message-ID: <mvhg9a.mgs.ln@cvs>

 phil hunt wrote:

>>I have a Python program that uses data stored in XML file to generate some
>>C++ code. The problem (or, more precisely, code smell) I have is that the
>>actual generation algorithm is hardwired into python and therefore is
>>cryptic, difficult to understand and maintain.

> How about:

>   def genBody(self, c):
>      substitutions = { 
>         "name": c.cxxClassName,
>         "ctorArg": c.factoryCtorArg or ''}
>      if c.factoryNargs == 0:
>         return subst(template1, substitutions)
>      else 
>         return subst(template2, substitutions)
Hmm. Have to think about it.
It still embeds generation algorithm inside the python script, but at
least does this much more elegantly that my version.
 From the DTSTTCPW point I think I could give it a chance, thanks!

>   def subst(formatString, args):
>      f = ... in formatString convert all $xxx to %(xxx)s ...
>      return f % args

> The implementation of subst() is left as an exercise.



-- 
If something is ugly, it can't be the best solution.
             -- Kelly Johnson.


From andymac at bullseye.apana.org.au  Tue Apr  9 05:19:26 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Tue, 9 Apr 2002 20:19:26 +1100 (edt)
Subject: 2.2.1c2 test_fcntl hangs
In-Reply-To: <ulmbybe3n.fsf@python.net>
Message-ID: <Pine.OS2.4.32.0204092015130.6083-100000@tenring.andymac.org>

On Mon, 8 Apr 2002, Michael Hudson wrote:

> Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:
>
> > I don't believe there have been any recent changes that might have
> > provoked this - at least I don't remember any checkins on the 2.2
> > maintenance branch (or the head branch for that matter) that would seem
> > plausible as causes (not that I'd rely too much on my memory either).
>
> Nothing substantive has changed in Modules/fcntlmodule.c in *years* (a
> large pile of constants have been added, the functions now take file
> objects as well as file descriptors and something changed for os/2-emx
> support -- but that's not on release22-maint).

I know about that, of course - the EMX changes are only to disable lockf
support (which EMX doesn't have).

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From johnroth at ameritech.net  Fri Apr 19 07:03:12 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 19 Apr 2002 07:03:12 -0400
Subject: Trinary operator?
References: <mailman.1019066614.29792.python-list@python.org> <just-4227D7.21004717042002@news1.xs4all.nl>
Message-ID: <ubvub71mto0d04@news.supernews.com>

"Just van Rossum" <just at xs4all.nl> wrote in message
news:just-4227D7.21004717042002 at news1.xs4all.nl...
> In article <mailman.1019066614.29792.python-list at python.org>,
>  "Mark McEahern" <marklists at mceahern.com> wrote:
>
> > > Not that it's a big deal since there's many ways to do that, but
can I use
> > > something like
> > >     verboseGender = (gender == 'm') ? 'male' : 'female'
> > > in Python?
> >
> > People seem to use and/or for that:
>
> But they shouldn't.

Why not? There's no other reason for the behavior
of 'and' and 'or' returning the object that caused the
final result.

John Roth





From hoiwai930 at hotmail.com  Sun Apr 14 08:55:03 2002
From: hoiwai930 at hotmail.com (Calvin)
Date: Sun, 14 Apr 2002 20:55:03 +0800
Subject: Looking for GUI Form Builder for Python
Message-ID: <a9en77$4vf8@imsp212.netvigator.com>

Dear all,
 Can I build Python TK GUI interface using a form builder. Anyone know such
builder. I just want to build my interface in a visual way.
-Calvin




From gerhard at bigfoot.de  Sat Apr  6 17:35:47 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 00:35:47 +0200
Subject: Size of Newsgrouppostings with nntplib
In-Reply-To: <a8nnlo$tv5d1$1@ID-109762.news.dfncis.de>
References: <a8nnlo$tv5d1$1@ID-109762.news.dfncis.de>
Message-ID: <20020406223547.GA1336@lilith.my-fqdn.de>

* Stephan Hohe <sth.news at uni.de> [2002-04-06 23:04 +0200]:
> Hi,
> 
> I wrote a python skript to generate some newsgroup-statistics. I use
> the NNTP object in module nntplib to connect to the newsserver and
> recieve interesting message-headers useing the xhdr method. Now I'm
> searching for a way to recieve the size of the postings. The only
> thing I found is the xover method, witch submitts many informations,
> including the article size in bytes.  Is there a way to recieve the
> article size without loading all the additional informations submitted
> by xover?

I don't know of any way. You could fetch the "Lines" header with the
XHDR command, but AFAIK you can only access the article size in bytes
with the help of XOVER.

Gerhard




From mats at laplaza.org  Wed Apr  3 14:42:26 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Wed, 03 Apr 2002 19:42:26 GMT
Subject: Python Version Strategy?
References: <xIYj8.16613$_L.7649@news1.bloor.is> <mailman.1016094984.20818.python-list@python.org>
Message-ID: <3c94aaea.161405128@news.laplaza.org>

On Thu, 14 Mar 2002 00:35:27 -0800, Trent Mick
<trentm at ActiveState.com> wrote:


:> Redhat, and most (all?) of the linux distros seem to be slow to fully
:> upgrade and I feel like this is raining in my cornflakes.
:
:Yup. As I understand it the core Python folks don't really have any pull for
:what default Python version gets installed for the various Linux distros.

Red Hat is trying to at least /appear/ to be stabilising things for
their customers, and they won't roll key software except at major
release time.  Python is used "internally" so it qualifies on this
front. That means that the entire 7.x series continues on Python
1.5.2, and the sad news is, the next Red Hat release will be 7.3 so
figure that Red Hat will be stuck on Python 1.5.2 through the end of
this year.  You can now have a more modern version as long as you're
willing to call it python2.

The sad thing is while Red Hat is sticking to their guns on this
issue, they keep breaking other stuff for folks, making one wonder why
they bother standing on the Python issue so hard.

Most other distros do seem to have moved forward to Python 2.x. That
doesn't mean the users of those distributions have upgraded to newer
versions, however.

Mats Wichmann



From nobody at nobody.com  Mon Apr  8 01:16:42 2002
From: nobody at nobody.com (John)
Date: Mon, 8 Apr 2002 00:16:42 -0500
Subject: IsList?
Message-ID: <9F9s8.37463$tg4.444710@vixen.cso.uiuc.edu>

Hi, How can one know wheter var 'a' is list or not? Is there any IsList-like
function?





From mkc+dated+1021692907.9e7f28 at mathdogs.com  Wed Apr 17 23:39:33 2002
From: mkc+dated+1021692907.9e7f28 at mathdogs.com (Mike Coleman)
Date: 17 Apr 2002 22:39:33 -0500
Subject: Why don't strings share data in Python?
References: <mailman.1018937610.11880.python-list@python.org> <87n0w37z0n.fsf@mathdogs.com> <a9j08p$nj6$0@216.39.172.122>
Message-ID: <87y9fl7k0a.fsf@mathdogs.com>

bokr at oz.net (Bengt Richter) writes:
> >I almost thought I might be able to use NumPy arrays, as it looks like they do
> >share data.  They're not hashable, though, unfortunately, which kills the
> >usage I had in mind.
> >
> So what did you have in mind?

Various string and sequence algorithms on really large data sets (gigabytes).
Bioinformatics.

Python is ideal for prototyping this stuff, but a bit profligate with memory
sometimes.  Some things feel slower (e.g., quadratic?) than I'd expect, at
times.

Mike



From juenglin at informatik.uni-freiburg.de  Wed Apr  3 09:25:19 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 03 Apr 2002 16:25:19 +0200
Subject: PEP 235: role of PyType_GenericNew()
Message-ID: <xduwuvoamgg.fsf@leto.informatik.uni-freiburg.de>

Hi,

I'm wondering whether the current implementation of PyType_GenericNew
does, what it's supposed to do.
In typeobject.c it reads:

PyObject *
PyType_GenericNew(PyTypeObject *type, PyObject *args, PyObject *kwds)
{
	return type->tp_alloc(type, 0);
}


Thus, only the allocator of the passed type is called.
Shouldn't the base classes allocators be called as well?

Cheers,
Ralf



From geoff at gerrietts.net  Thu Apr 11 16:00:33 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 11 Apr 2002 13:00:33 -0700
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer division,backwards compatibility; where is  Python  going?)
In-Reply-To: <3CB523FD.7D83C0C@engcorp.com>
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <mailman.1018495402.12499.python-list@python.org> <Xns91ECE5A88C0F8RASXnewsDFE1@130.133.1.4> <3CB523FD.7D83C0C@engcorp.com>
Message-ID: <20020411200033.GB4247@isis.gerrietts.net>

Quoting Peter Hansen (peter at engcorp.com):
> Just for the record, 'twas I not Geoff who said that.  Geoff may
> be a leech, he may even be a veritable wart on the bottom side of
> society, but so far he hasn't admitted it here.  <wink>

If you need references, I can provide an ex-wife and several people I
used to play MUSHes with who will be MORE than happy to take potshots
at my character. ;)

I like to think they're wrong; please don't disabuse me of my
delusions. ;)

Thanks,
--G.

-- 
Geoff Gerrietts                                 <geoff at gerrietts net> 
"A man can't be too careful in the choice of his enemies." --Oscar Wilde




From rjgruet at yahoo.com  Fri Apr 19 19:18:42 2002
From: rjgruet at yahoo.com (Richard Gruet)
Date: Sat, 20 Apr 2002 01:18:42 +0200
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <uc14cjalpru3b5@news.supernews.com>
Message-ID: <a9q8f4$svn$1@aphrodite.grec.isp.9tel.net>

"John Roth" <johnroth at ameritech.net> wrote in message
news:uc14cjalpru3b5 at news.supernews.com...
>
> "Richard Gruet" <rjgruet at yahoo.com> wrote in message
> news:a9nbuq$916$1 at aphrodite.grec.isp.9tel.net...
> > Alex
> >
> > And BTW, people would not to have implement any __len__ method if all
> the
> > sequence types would derive from a common ancestor (sequence), in
> which the
> > (in)famous method would be implemented. Would it be completely
> nonsensical ?
> > I have been told that starting from 2.2, new style classes (including
> most
> > built-in types) all derive from a common base class 'object'.
>
> As far as I know, that's correct. Inserting a .len() method in the
> object class would automatically make it available to all new style
> classes.
>
> As far as I can see, there is one major issue with this: if the
> class in question didn't have a __len__() method, it would
> still have the (inherited) len() method. This could be rigged to
> raise an exception, but it would still be a bit of a trap for the
> unwary.
>
> There's also the question of the inheritance hierarchy, which I don't
pretend to understand.
> It seems to be a case of simple implementation, small benefit, small
downside, no discernable side effects. Write a PEP.
>
> John Roth

Actually, as Alex pointed out, a single len() method in a hypothetical
'sequence' base class would be mythical, simply because there is no common
implementation of len() for all kinds of collection (I just forgot this!).
However we could still define a sort of abstract len() method  in a
'sequence' (or 'collection') base class, that has to be implemented in
derived classes (or will raise an exception otherwise).
This is in fact equivalent to implementing the existing __len__ method,
except that the names are different and the purpose of __len__ is mainly to
be called by the built-in function len(), but nothing forbids to call
obj.__len__() directly AFAIK. So ultimately the issue resides solely  in the
name, as an esthetical preference (I'm not personally addicted to
underscores). And since the __xxx__ notation is a Python convention, I'm
afraid there will be no other solution but using the obj.__len__() syntax if
one doesn't want to use len(obj).

Richard




From jmanzano at antena3tv.es  Mon Apr  8 07:39:45 2002
From: jmanzano at antena3tv.es (jmanzano at antena3tv.es)
Date: Mon, 8 Apr 2002 13:39:45 +0200
Subject: Adjuntar ficheros en python.
Message-ID: <OF9C403694.D36227EA-ONC1256B95.003FCDDF@antena3tv.es>

Hola

     Alguien me puede decir como puedo adjuntar ficheros en python a la
hora de enviar emails.

     Ficheros de tipo texto.

     Salu2.

     Actualmente, utilizo algo como lo siguiente,  pero necesito enviar los
ficheros como adjuntos.

#!/usr/bin/env python
import smtplib

pax=open('/disco2/errores/datos/personas.dat','r')

server = smtplib.SMTP('192.168.14.2')
server.set_debuglevel(1)
persona=pax.readline()
while persona!="":
        fic=open('/disco2/errores/datos/salida.txt','r')
        server.sendmail('admin at prueba.com',persona,fic.read())
        persona=pax.readline()
        fic.close()
pax.close()
server.quit()






From gerhard at bigfoot.de  Mon Apr  1 01:42:04 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Apr 2002 06:42:04 GMT
Subject: Most important extensions?
References: <a8614l$dfa$1@panix1.panix.com> <a87gnj$e75$1@panix1.panix.com>
Message-ID: <slrnaafsqi.8mf.gerhard@lilith.my-fqdn.de>

Aahz wrote in comp.lang.python:
> In article <3CA732D6.91163F1E at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
>>
>>I do hope you will have a chance to mention PyUnit anyway, even
>>if not as an extension.  Learning how to use it will definitely
>>improve a programmer's productivity.
>
> Oh, of course I'll mention doctest and unittest.  My focus will be on
> "batteries included"; the question is which of the *non*-included
> extensions I should cover.

Among my favourites are:
- win32 extensions (make Windows actually useful)
- wxPython (also for its non-GUI goodies)
- 4Suite (for the stuff that's not in PyXML, yet)

This post is the first one coming thru my own nntp-proxy :-)

Gerhard
--
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From dmmcf at uiuc.edu  Tue Apr 16 20:03:41 2002
From: dmmcf at uiuc.edu (D. Michael McFarland)
Date: Tue, 16 Apr 2002 19:03:41 -0500
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu>
 <861ydg3ff3.fsf@haggis.aae.uiuc.edu> <3CBC5686.BCBCF663@noaa.gov>
Message-ID: <86k7r7ry1u.fsf@haggis.aae.uiuc.edu>

Chris Barker <Chris.Barker at noaa.gov> writes:

> "D. Michael McFarland" wrote:
>> The bottlenecks now seem to be in
>> memory management, which I'm sure Python does better than I would in
>> C,
>
> better than me too, I'm sure, but NumPy code often ends up making a lot
> more copies of data that similar C code would, so that might very well
> get better in C too.

According to the profiler output, the single biggest time sink seems
to be the creation of new tensor objects, which are essentially NumPy
arrays with some bells and whistles.  I've always assumed (and
continue to) that NumPy was doing a good job of allocating memory and
copying array elements, and that I had simply chosen algorithms that
require the creation of too many new tensors.

>> and numerical linear algebra, which I'm sure Numeric does better
>> than I would.
>
> Are you using a ggod native BLAS? that can make a differnce too.

A good point, and something I've not looked at closely yet.  The
second bottleneck in the code is the inversion of rank 4 tensors, in
the course of which we do a singular value decomposition of a 9 x 9
NumPy array.  Speeding this up would certainly help overall.

One reason I'm reluctant to start rewriting my tensor lib (the good
parts of which are due to Konrad Hinsen, by the way) is that
eventually I'd want to call an external library for this and for a
linear equation solver.  Because that interface exists in NumPy, I've
felt I'd be reinventing the wheel, but maybe I should take another
look.  (In the long run, SciPy is probably how I want to do this, no?)

>> For now, though, the safest
>> route seems to be to throw CPU time at the existing code.
>
> I understand, and I may be able to offer a few cycles. Send me a
> private email.

Will do.  Thanks.

>> But first this guy needs to
>> graduate, for which he needs to crunch a _lot_ of numbers.  He's
>> swamped with related work, and I feel like I got him into this...
>
> Where would he be if he had tried to write all his code in Fortran
> or C??

You'd have to ask him for the full answer, but I remain convinced that
the savings in programming time and the clarity possible in Python are
well worth the run time they cost.  It's just that, dissertation
writing being what it is, things have come down to a crunch at the
end.  Our running joke is that in ten years' time we'll be running
these problems on our Palm XVII PDAs and laughing about it.

Thanks for your suggestions.  I'll take another look at how I might
tighten up some parts of the existing code.

Cheers,
Michael


From bokr at oz.net  Mon Apr 15 16:39:37 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 20:39:37 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <LHAu8.11094$Ji7.10494@atlpnn01.usenetserver.com>
Message-ID: <a9fdq9$h08$0@216.39.172.122>

On Mon, 15 Apr 2002 09:34:26 -0400, "Steve Holden" <sholden at holdenweb.com> wrote:

>"Bengt Richter" <bokr at oz.net> wrote ...
>[ ... ]
>> Well, to tie in with existing spelled-out 'and' and 'or' operations,
>> it should probably do something along those lines (i.e., evaluate to
>> one or other of the objects, not one of the boolean values used in the
>logic),
>> but I can't think of something sensible.
>>
>Since in Boolean terms a xor b is equivalent to ((a and not b) or (b and not
>a)), maybe a xor b should return whichever of a or b is not false.
>
>>>> a = 42
>>>> b = 0
>>>> a ^ b
>42
>>>> b ^ a
>42
>>>>
>
>Since that's exactly what it appears to do, stop fiddling with what already
>works !
 >>> a = 42
 >>> b = None
 >>> a ^ b
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: unsupported operand type(s) for ^: 'int' and 'NoneType'
??

I'm not sure who you're talking to, but my problem was thinking of xor in terms
of short cut evaluation analogous to 'and' and 'or'. Nothing seemed intuitive,
since there is no way to short cut (i.e., have a case sensibly avoiding evaluation
of both terms) an xor.

Regards,
Bengt Richter


From p.magwene at snet.net  Thu Apr  4 23:31:17 2002
From: p.magwene at snet.net (Paul M)
Date: Fri, 05 Apr 2002 04:31:17 GMT
Subject: Response to Guido's essay on graph representation
References: <eppstein-87C133.12464904042002@news.service.uci.edu>
Message-ID: <pM9r8.11156$FU6.2998356006@newssvr10.news.prodigy.com>

For me David's code emphasizes the importance of well designed data
structures!  In particular his implementation of a priority queue (his
priorityDictionary class).  I love it, and here's why...


<long rant on optimization in Python>

I've recently been fooling around with various graph algorithms, and had
implemented a variety of "naive" implementations of classics like Prim's and
Kruskal's MST algorithms, Djikstra's shortest path, Floyd's all shortest
paths, etc.

My implimentations worked ok, but for even moderately large data sets they
were SLOW.  I figured that I was just bumping up against Python's limits,
and so decided to reimpliment the Kruskal and Prim algorithms using
structures from C extension modules.

I first reimplemented the Kruskal algorithm using the kjGraph data structure
from Aaron Watters module kjbuckets, which resulted in a considerable speed
up, but was still somewhat pokey.  After going back and reading Cormen et
al.'s Introduction to Algorithms (highly recommended!) I realized that  part
of my problem was that I wasn't using a disjoint-set data structure.  I
coded up the disjoint-set representation in pure Python and lo and behold,
the Python implementation using the "right" data structure was several times
faster than the C extension module using the "wrong" data structure!
Ka-ching!

With my renewed appreciation for the importance of data structures in hand I
turned to Prim's algorithm.  The standard approach for this algorithm is to
use a priority queue, however my initial naive implementation had been
abysmally slow.  Again I turned to somebody elses C extension for priority
queues which used Fibonacci heaps.  My implementation of Prim's based on
this was pretty darn fast, and decided I'd found the proper solution for my
needs, despite the fact that I had trouble making heads or tails out of the
fibonacci heap C code (I hate taking a black-box approach!).

Today I stumbled across David's implementation of priority queues.  I
decided to give it a shot as the underlying data structure for my
implementation of Prim's.  Dropping it into my implementation was trivial
(only a few lines of code needed to be changed) and I threw some test data
sets at it.  How did it do?  Amazingly well!

Check out these stats from runs with the profiler module for a 1000 vertex,
(1000 choose 2) edge graph.  Remember, that using the profiler adds a good
chunk of time to the runs but is good for getting a sense of relative
speed...

First the Fib. heap based C extension module:

>>> profile.run('mst.mst_prim(Avert, Adist)')
         3003 function calls in 15.172 CPU seconds
<< details snipped >>


Now, the priorityDictionary results:

>>> profile.run('pd.mst_prim(Avert, Adist)')
         62658 function calls in 20.383 CPU seconds
[ details snipped, but about 3.5 seconds is just calls to __setitem__ as I
filled up the queue ]


Obviously the C-based data structure is still faster, but I can understand
what David's implementation is doing, while the C implementation would take
quite some slogging through to understand.  That for me is worth the
tradeoff!

</long rant>

Of course, anecdotes about Python optimization are not uncommon - either in
this newsgroup or on the web (see Guido's classic here:
http://www.python.org/doc/essays/list2str.html ).  However, running into
real world examples in my own day-to-day work puts a real smile on my face.

Go Python!

Ciao,
Paul



From sholden at holdenweb.com  Thu Apr 25 10:02:36 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 25 Apr 2002 10:02:36 -0400
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at>
Message-ID: <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com>

"Siegfried Gonzi" <siegfried.gonzi at kfunigraz.ac.at> wrote in message
news:3CC7F1F7.8B987146 at kfunigraz.ac.at...
> The topic has emerged quite often in the past but never completely
> answered.
>
> I have an application which consumes at every run 60MB RAM. Every new
> start of the script ('Run scipt') augments the memory up to old_memory +
> 60MB. After a few runs I am at 400MB.
>
> I always just thought Python has a garbage collector and will dispose
> the memory after the last reference (for me this means: end of my
> script).
>
> In Clean or Common Lisp or R-language or Yorick I never had any dire
> need to
> consult the Windows Task-Manager. The explanation that Python really
> gets rid of the memory but the Windows OS does not show it sounds a
> little bit strange to me.
>
> I am now a little bit more accustomed to Python, but I cannot imagine to
> use it further for larger projects. My machine has got 256MB RAM and I
> just use Windows XP (80MB RAM) and Python2.2; there is absolutely no
> excuse
> for Python not to find any free successive memory block.
>
> There is the chance that  memory becomes free; but I am not willing to
> wait about 10 minutes (after a 10 minutes calculation) in order to see a
> possible disposing of memory. I want to start the second simulation
> right after the first, but after a few simulations I end up at 400MB and
> Python chrashes.
>
>
> Any suggestions?
>

It sounds as though you are running the successive simulations in a single
interactive editor session. Are you possibly doing something like (simulated
interpreter session follows):

>>> import simulator
>>> a = simulator.run("file1.txt")
>>> b = simulator.run("file2.txt")
>>>

or the like? If so then the references in variables might account for the
memory usage. Would it be possible to rewrite your code so you could do
simulation runs from the command line, along the lines of

python simulate file1.txt

In this case there is no chance that the interpreter would continue to hold
on to memory used by previous simulation runs -- when it terminates, the OS
gets the memory back, period.

Python is usually a reliable consumer of memory, and a reliable collector of
garbage (although a few pathological cases have been observed). You are
seeing the behavior you describe because the interpreter is somehow
retaining pointers to the data structures created during your simulations.
Without a bit more information about the program structure and your usage
it's difficult to say what might really be causing your problem.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From logiplexsoftware at earthlink.net  Tue Apr  9 14:55:50 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 9 Apr 2002 11:55:50 -0700
Subject: PEP 285: Adding a bool type
In-Reply-To: <zpF9N5ATtts8Ewc0@jessikat.demon.co.uk>
References: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk>
	<a8uh5p$nu7$1@thorium.cix.co.uk>
	<zpF9N5ATtts8Ewc0@jessikat.demon.co.uk>
Message-ID: <20020409115550.5f3f3d49.logiplexsoftware@earthlink.net>

On Tue, 9 Apr 2002 13:15:15 +0100
Robin Becker wrote:

> Scripts in sed or awk that are 20 years old still run because their
> languages are stable. Python scripts that are 5 years old don't run. The
> rate of change is increasing precisely because product developers think
> continually changing them is advantageous. New versions always create
> update fever further down the food chain which is why intermediary users
> get so fed up with continual minor adjustments that cause software
> uncertainty.

Comparisons to sed and awk don't really seem appropriate.  Python is a
general-purpose programming language whereas sed and awk are special-purpose
languages with a rather limited scope.  I think your concerns are
understandable, but I don't see how you are going to escape this.  Rapid change
is the reality of our field, whether in operating systems, programming
languages, GUI's, whatever.  This is good, but of course makes everyone's job
more difficult.  However, hoping it will change doesn't seem realistic at this
point, in fact I expect that the rate of change will steadily increase (as
technology begets more technology).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From tejarex at yahoo.com  Thu Apr 18 18:37:20 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 18 Apr 2002 22:37:20 GMT
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net>
Message-ID: <AUHv8.80647$3L2.7180902@bin7.nnrp.aus1.giganews.com>

"Richard Gruet" <rjgruet at yahoo.com> wrote in message
news:a9nbuq$916$1 at aphrodite.grec.isp.9tel.net...
> I just meant that to my human eyes, it is easier to memorize when
everything
> is consistently object oriented and there are only methods

Would you prefer that a+b be written as a.add(b) (which I believe
would be the Python equivalent of what at least one Object-Oriented
language does)?

> (like in Java - and I prefer Python to Java :-), or the contrary,
only functions.

or as add(a,b) (which is the Python version of what some functional
languages do)?

> Then I
> have not to remember which functions are methods and which ones are
> built-in. I think this is because a mixed model is more complicated
than a
> non-mixed one.

An argument can be made for syntactic uniformity, and successful
languages have been designed around that principle.  However, Python
was not.  As I remember, Guido has expressed a belief that a mixture
of function calls, operators, and method calls is *easier* to read.
And the observation that programmers read code as much or more than
they write it is one of the guiding forces of Python design.

Terry J. Reedy





From sqrt_2  Mon Apr 15 12:37:56 2002
From: sqrt_2 (sqrt_2)
Date: Mon, 15 Apr 2002 16:37:56 GMT
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za> <3CBA43B9.22850DF1@engcorp.com> <3CBA6FDB.9030706@mxm.dk>
Message-ID: <3cbaffcf.195381884@news.mweb.co.za>

Okay I have been not entirely open as I did not want to be seen as a
'promoter'. My friend and fellow employee Riaan Booysen,  wrote
Boa-constructor and in his words:
A RAD GUI-building IDE for wxPython wrapping wxWindows and so on.

What I meant with Zope is that I can edit my Zope sites through Boa as
well as use the help-files etc. This means that I do not necessarily
have to open a site(local) to edit it's files. I can just use boa.
So basically Boa is always running on my machine and is used to write
any Python apps as well as edit Zope files.
I will be working with extending Boa myself for possible products such
as Pygame or Pyui and whatever seems appropriate. Of course only in a
month or two.
I just posted the note to help some people that seemed to have lots of
wx questions. (Saw from post headers in the group)

And of course a little promotion didn't hurt I hope. <grin>

Please do not bombard me personally with mails about Boa but posted
suggestions I will look at happily.
For my company stuff please visit http://www.tbs.co.za
I will post new Boa developments if significant.

sqrt_2

On Mon, 15 Apr 2002 08:14:51 +0200, Max M <maxm at mxm.dk> wrote:

>Peter Hansen wrote:
>
>> Of course, those products are well known around here.
>> 
>> What's less well known is the relationship between Boa-Constructor 
>> and Zope.  I strongly suspect there isn't one.
>> 
>> Can you elaborate on this supposed "incorporation of Zope
>> into Boa-Constructor"?
>
>
>It can, eventually, be used as an editor for Zope. I beleive either 
>through ftp or webdav.
>
>regards Max M
>



From jon+usenet at unequivocal.co.uk  Wed Apr 24 05:12:09 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 09:12:09 -0000
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net> <3umccu09j6dkui67hnppulv8ngk71g55s4@4ax.com>
Message-ID: <slrnacctn8.s1r.jon+usenet@snowy.squish.net>

In article <3umccu09j6dkui67hnppulv8ngk71g55s4 at 4ax.com>, Fernando Rodr?guez wrote:
>> If they really must be separate scripts, why doesn't script (a)
>> just output nothing at all to stdout?
> 
> I didn't try it, but I think the browser wouldn't like a cgi url that doesn't
> return any html content, no? O:-)

I thought they were running on the same machine, in which case the
entire sequence of [run script a + run script b] would produce the
correct CGI output.

You are going to have to do as someone else suggested, use the Python
standard library routines to issue another https request yourself from
the server to run script (b).


From dradul at yahoo.com  Fri Apr 26 15:56:34 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Fri, 26 Apr 2002 14:56:34 -0500
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org>
Message-ID: <aact9h$9odgq$2@ID-99513.news.dfncis.de>

"Dan Jacobson" <jidanni at deadspam.com> escribi? en el mensaje
news:m2g01lfx7i.fsf at jidanni.org...
> Before I start learning python, here's what I want to do: I have a
> table of Hakka Chinese words and their pronunciations.  I scan a file
> and replace any Hakka ["big5" 2, 4, 6 ... byte long strings] there
> with their pronunciations.  If it were just one character [two byte]
> words I would use the "c2t" program.  Is there a template that munches
> forth in a file and replaces the longest match in a database
> before moving on?
> --
> http://jidanni.org/ Taiwan(04)25854780

This will help you get started:

        http://pythonzh.sourceforge.net/



From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 01:34:26 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 22:34:26 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu>
Message-ID: <3cba66aa@news.mhogaming.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-9E09EB.21122714042002 at news.service.uci.edu...
| In article <3cba4cf7 at news.mhogaming.com>,
|  "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
|
| > The other thing I was trying to say, is that if this
_BUG_
| > (gee- that wasn't so hard!) in Python is fixed, then we
need
| > a way to consistently do all of this bit shifting.  Now
that
| > ints and longs are 'unified', you are supposed to be
able to
| > write code that doesn't care what the internal
| > representation is-- that should also apply to the '<<'
and
| > '>>' operators.  BUT, since we NOW don't know where the
high
| > bit is (but the class that is managing the long object
| > DOES), we need a '>>>' operator to consistently handle a
| > logical right shift-- (i.e.: there is no way to code
this
| > behavior in Python.)
|
| There is no "highest nonzero bit" -- if it's a negative
number, the bits
| in all sufficiently high positions are nonzero.  The long
class code
| doesn't know where the highest meaningful bit is in an
value that is
| being used as a bitstring, it only knows where the highest
| explicitly-represented bit is, and that's not the same
thing.
|
| If you do somehow know where the highest meaningful bit
is, you can
| always do something like this:
|
| def shiftLeftMasked(x, shiftAmount, highBitPosition):
|     return (x & ((1L << highBitPosition) - 1)) >>
shiftAmount
|
| --
| David Eppstein       UC Irvine Dept. of Information &
Computer Science
| eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/

The point that I was trying to make, is that the method that
is creating the hexadecimal string (the __str__ method?)
inside
the long class is the ONLY thing that "knows" how many bytes
are being used to represent a long.  When "hex(x)" creates a
string from a long, it should not have a '-' sign if the
long
is negative, but rather should print out the hexadecimal
representation of the number in using the number of bytes
that
the object currently occupies.  If we are printing a long as
a
decimal number, then we DO add the '-' sign.

The method that "owns" the object, and shifts the long
('<<',
'>>', '>>>') can most certainly determine how many bytes are
being used to represent the object, and can most certainly
provide the correct answer for these operations.  The USER
of
the object cannot do this from outside code however, so the
method inside the long class will have to be fixed to repair
this errant behavior.

I think the unification of longs and ints is a wonderful
thing,
but the work it not finished yet-- some more careful thought
needs to be put into this-- in fact, an int is promoted to a
long (as you pointed out) by the multiply operation, but not
the
shift operation.  This is indeed a BUG.  What is ALSO a BUG,
is
that the long is then not DEMOTED back to an int, when it is
divided by either a right shift or a divide operation, and
the
number can _NOW_ be properly represented by an int.  This
should all happen automatically, "under the hood" (so to
speak)
without the programmer even having to think about it.  That
was
the whole purpose of the int-long unification, wasn't it?





From wurmy at earthlink.net  Sun Apr  7 14:53:10 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 07 Apr 2002 18:53:10 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1018169894.19586.python-list@python.org>
Message-ID: <3CB09633.12096C85@earthlink.net>

Arthur Siegel wrote:

> Guido asks "should the PEP be accepted"?
> 
> Many folks reacted that - no, it is unnecessary and
> poorly timed and took the time to elaborate as intelligently
> and thoroughly as they could the basis for that conclusion.
> 
> Only to find that the question Guido was *really* asking was
> whether anyone saw a *killer* reason not to adopt.
> 
> Unnecessary and poorly timed may well be the right answer
> but is not a  *killer* reason. You lose.
> 
> The fact as to who's football we happening to be playing
> with now being a little shoved in our face.  Guido seemed
> willing to accept the sense of the community, along as the
> community told him what he wanted to hear. When
> it didn't, he simply retroactively rephrased his question.

I don't see it that way.

- Guido wants to add a boolean to Python. 
- Now, he could have just shoved it down our collective
throat, without asking anyone's opinion. But instead,
he decides to write a PEP, and asks what people think
of it. (First in python-dev, later in c.l.py.)
- Apparently, none of the arguments against the boolean
were convincing to him, at least not enough to reject
the PEP. (People may argue, sure he won't reject it,
it's his own PEP... but if I'm not mistaken, he has
rejected his own PEP before.)
- There is a lot of talk about the PEP in the newsgroup,
but seeing that this most likely won't yield any
new reasons not to accept the PEP, Guido decides to
tie the knot and use his status as BDFL. PEP accepted.

Doesn't seem so unreasonable to me.

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA==')) 
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/


From juenglin at informatik.uni-freiburg.de  Thu Apr  4 03:39:52 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 04 Apr 2002 10:39:52 +0200
Subject: PEP 235: role of PyType_GenericNew()
References: <xduwuvoamgg.fsf@leto.informatik.uni-freiburg.de> <mailman.1017854949.20733.python-list@python.org>
Message-ID: <xdusn6bamcn.fsf@leto.informatik.uni-freiburg.de>

Gustavo Niemeyer <niemeyer at conectiva.com> writes:

> Hi Ralf!
> 
> > I'm wondering whether the current implementation of PyType_GenericNew
> > does, what it's supposed to do.
> [...]
> > Thus, only the allocator of the passed type is called.
> > Shouldn't the base classes allocators be called as well?
> 
> If you're inheriting from another object, and you don't want to
> do anything special at object initialization time, why would you
> use this function in the first place? Just leave the slot empty
> and allow the super class to initialize itself. If you implement
> your special function, it's your business to call the super class.
> 

That's the point, why I consider PyType_GenericNew's current behaviour
a bug!

What I want to do:

Write a base-class _A in C, which fills the type's tp_alloc-slot with
its special allocator, for the tp_new slot I simply use PyType_GenericNew.
At the Python level I write a class A which is a subclass of _A. Since
A does not introduce new data members, there is no need to overwrite
__new__. This way, however, the superclass-allocator (_A.alloc) never
gets called when I create instances of A...

Regards,
Ralf

 



From peter at engcorp.com  Tue Apr  2 21:58:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 21:58:55 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204030123.BAA15136@crown.off.ekorp.com> <3CAA60A5.F88697D5@engcorp.com> <200204030205.CAA21272@crown.off.ekorp.com>
Message-ID: <3CAA6FEF.9C589C1B@engcorp.com>

Richard Jones wrote:
> 
> I already write my slide presentations using ReST - it's converted to
> PythonPoint XML and from there to PDF. Very nice, and as simple as can be :)

We had a brief discussion today, inspired perhaps by the presence
of the rlzope.py file in PythonPoint, about how few steps would
be required to integrate:

 1. ZWiki 

 2. ReST

 3. Zope

 4. PythonPoint

to allow entry of text into wiki pages and, with the click of
a button, auto-generate PDF presentations.  Hello ease-of-use
and efficiency, bye-bye PowerPoint.

(The answer, by the way, was "very very few"! :-)

-Peter



From martin at v.loewis.de  Mon Apr  1 08:26:29 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 15:26:29 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <7xk7rr3a38.fsf@ruckus.brouhaha.com>
Message-ID: <m3adsn8s8q.fsf@mira.informatik.hu-berlin.de>

Paul Rubin <phr-n2002a at nightsong.com> writes:

> The point about pickle/marshal is especially good.  I can probably
> think of a other dozen other serialization and remote object Python
> modules that would also have to be updated.

Indeed. Interestingly enough, one of the driving forces of this PEP is
the ability to change them.

Regards,
Martin



From ole.fostad at netcom.no  Sun Apr 14 06:06:23 2002
From: ole.fostad at netcom.no (Ole Fostad)
Date: Sun, 14 Apr 2002 12:06:23 +0200
Subject: Editable field length in wxGrid
Message-ID: <a9bkb1$98o$1@lise.netcom.no>

Hi everyone!

I tried to figure out how to limit the available field length in a wxGrid
text field (using a wxGridCellTextEditor), but I am stuck.

Does anyone know how to limit the editable field length (to say 8
characters) ?

Thanks!

Ole Fostad




From fperez528 at yahoo.com  Mon Apr 22 18:17:24 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 16:17:24 -0600
Subject: calendar.py doesn't use lists???
References: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <aa2290$431$1@peabody.colorado.edu>

Aaron Watters wrote:

> 
> Using Python 2.2 I get this
> 
> """
> 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)]
> <type 'instance'>
> <type 'instance'>
> [CRASH AND BURN... WINDOWS XP SUGGESTS I SEND A COMPLAINT TO MICROSOFT...]
> """
> 
> Of course it may be that this is fixed in python 2.2.1 but...
> 
> WHO'S GOOFY IDEA WAS THIS?
> 
> Seriously, why did you need to change these sequences from
> lists where they worked with my code, to somethingsimilartolists
> that broke my code?

I get the feeling that this isn't so much a change which broke your code but 
rather a good'ole BUG. Under Mandrake Linux 8.2:

[~]> ipy
Python 2.2 (#1, Feb 24 2002, 16:21:58)
Type "copyright", "credits" or "license" for more information.

IPython 0.2.10pre9 -- An enhanced Interactive Python.
?       -> Introduction to IPython's features.
?object -> Details about 'object'; object? also works, ?? prints more.
help    -> Python's own help system.
@magic  -> Information about IPython's 'magic' @ functions.

In [1]: from calendar import month_abbr, month_name

In [2]: month_abbr ?
Type:           instance
Base Class:     calendar._localized_name
String Form:    <calendar._localized_name instance at 0x81a8cb4>
Namespace:      Interactive

In [3]: for m in month_abbr: print m
   ...:
Segmentation fault

Python shouldn't segfault ever, period (that is, unless it's calling your own 
extension modules and those are causing the segfault).

Cheers,

f.


From logiplexsoftware at earthlink.net  Fri Apr 19 14:25:53 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 19 Apr 2002 11:25:53 -0700
Subject: Trinary operator?
In-Reply-To: <3CC0578E.822EB961@metaslash.com>
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
	<NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>
	<mailman.1019068175.11920.python-list@python.org>
	<99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com>
	<mailman.1019233076.19513.python-list@python.org>
	<3CC0578E.822EB961@metaslash.com>
Message-ID: <20020419112553.200f2716.logiplexsoftware@earthlink.net>

On Fri, 19 Apr 2002 13:44:46 -0400
Neal Norwitz wrote:

> import time
> gender = 'f'
> GENDER_DICT = {'m': 'male', 'f': 'female'}
> 
> t = time.time()
> for i in range(100000):
>     verboseGender = (gender == 'm') and 'male' or 'female'
> print time.time() - t
> 
> t = time.time()
> for i in range(100000):
>     verboseGender = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
> print time.time() - t
> 
> t = time.time()
> for i in range(100000):
>     verboseGender = GENDER_DICT.get(gender, 'unknown')
> print time.time() - t
> 
> t = time.time()
> for i in range(100000):
>     try: verboseGender = GENDER_DICT[gender]
>     except KeyError: verboseGender = 'unknown'
> print time.time() - t
> 
> 0.248443961143
> 0.696393966675
> 0.322777986526
> 0.239256024361

Interesting.  It hadn't occurred to me that exceptions would be any faster than
.get().  I didn't use a predefined dictionary since the OP was looking for an
equivalent to C's ternary operator (which I took to mean "one line"), but
that's clearly the most important optimization if speed is a concern (athough
it's arguably more readable as well).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From bokr at oz.net  Sat Apr 13 22:33:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 02:33:26 GMT
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <mailman.1018738411.20464.python-list@python.org>
Message-ID: <a9appm$l47$0@216.39.172.122>

On Sun, 14 Apr 2002 00:41:37 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> wrote:

>* Jens Baader <nospam at nospam.com> [2002-04-13 23:46 +0200]:
>                      ^^^^^^^^^^
>
>If you don't have permission of the nospam.com domain owner to use their
>domain, you should stop abusing their services *now*. If you don't want
>to get spam, I'd recommend you get a throwaway email address at a
>freemailer.
>
I agree, but OTOH you'd think the IETF would come up with a standard
global /dev/null sort of address for this purpose. Is there such a thing?
I.e., why burden freemailers with garbage storage, and the net with the traffic?

Regards,
Bengt Richter


From shippy at nmt.edu  Fri Apr 26 16:45:51 2002
From: shippy at nmt.edu (Jeff Shipman)
Date: Fri, 26 Apr 2002 14:45:51 -0600
Subject: cgi module question
References: <3CC9A3E2.9060803@nmt.edu> <kud6wmb6wt.fsf@lasipalatsi.fi>
Message-ID: <3CC9BC7F.5020305@nmt.edu>

Uh, no, not really because it's consumed by
FieldStorage(). If I consume it before that, then
there's FieldStorage() can't grab it. I still
need to use the cgi module, but I also need
to get the POST query string.

Erno Kuusela wrote:
> In article <3CC9A3E2.9060803 at nmt.edu>, Jeff Shipman <shippy at nmt.edu> writes:
> 
> | I'm using Python 1.5.2 and the cgi module. I'm
> | wondering if there's any way to recover the string
> | that was passed to the cgi module during a POST
> | operation (normally comes in through stdin)? 
> 
> yes. read sys.stdin. 
> 
>   -- erno


-- 
Jeff "Shippy" Shipman     E-Mail: shippy at nmt.edu
Computer Science Major    ICQ: 1786493
New Mexico Institute of Mining and Technology
Homepage: http://www.nmt.edu/~shippy



From starx at pacbell.net  Tue Apr  2 19:58:09 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 3 Apr 2002 00:58:09 GMT
Subject: Quick return question
References: <0Ujq8.18$c6.7717@typhoon.mn.ipsvc.net> <YWjq8.19$c6.7993@typhoon.mn.ipsvc.net>
Message-ID: <Xns91E4ACCD9E5C4RASXnewsDFE1@130.133.1.4>

joel8bit || Tue 02 Apr 2002 07:16:08a:

> 
> "joel8bit" <jwickard**nospam**@litriusgroup.com> wrote in message
> news:0Ujq8.18$c6.7717 at typhoon.mn.ipsvc.net...
>> Hello all:
>> given the function:
>>
>> def adder(**args):
>>     for key in args.keys():
>>         print args[key]
>>
>> print adder(fName="sam", lName="scott", mName="wayne")
>>
>> I get:
>>
>> sam
>> scott
>> wayne
>> None
>>
>> Why am I getting None as well, because there's no return value for
>> the function? 
>>
>>
> 
> Disregard I understand now,
> I told it to print the function call and there was no return value.

Heh, i did the same thing a few weeks ago: had a function that did 
something or other and returned the results, and another part that did 
'print function(foo)' a bunch of times. Changed the functions to print 
instead of return, and started getting 'None' all over. Very hard to 
catch, the first time =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From ebrake2002 at yahoo.com  Fri Apr 19 21:47:38 2002
From: ebrake2002 at yahoo.com (Eric Brake)
Date: 19 Apr 2002 18:47:38 -0700
Subject: using get() with Entry()
Message-ID: <12f756b9.0204191747.7acb2b74@posting.google.com>

I've looked through all of the documentation I can find. I can't make
get() work with Entry().

Here's my code:
###
guess = StringVar()
guessbox = 0
guessbox= Entry(root, width=20, textvariable=guess).pack()
guess.set(guessbox.get(guessbox))
###

It keeps giving me a error message saying:
Traceback (most recent call last):
  File "D:\My Documents\mymodules\gui\guessthenumber.py", line 43, in
?
    guess.set(guessbox.get(guessbox))
AttributeError: 'NoneType' object has no attribute 'get'

What am I not doing? I just now learning this so break it down,
Please.
Thank you for your help


From max at alcyone.com  Sun Apr  7 15:46:00 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 07 Apr 2002 12:46:00 -0700
Subject: ACCEPTED: PEP 285
References: <mailman.1018169894.19586.python-list@python.org>
Message-ID: <3CB0A1F8.BB443044@alcyone.com>

Arthur Siegel wrote:

> Guido asks "should the PEP be accepted"?
> 
> Many folks reacted that - no, it is unnecessary and
> poorly timed and took the time to elaborate as intelligently
> and thoroughly as they could the basis for that conclusion.

And many people responded that it was a good idea.

> Only to find that the question Guido was *really* asking was
> whether anyone saw a *killer* reason not to adopt.
> 
> Unnecessary and poorly timed may well be the right answer
> but is not a  *killer* reason. You lose.

If he has a prediliction to accept it, what would you really expect? 
He's looking for a good reason _not_ to accept it, and none was
presented.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From tejarex at yahoo.com  Fri Apr 19 11:43:08 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Fri, 19 Apr 2002 15:43:08 GMT
Subject: Python scripting to dynamically generate 3D human models
References: <mailman.1019218239.16981.python-list@python.org>
Message-ID: <gWWv8.109596$K5.9156921@bin5.nnrp.aus1.giganews.com>

"Woon Shan Chin" <woonshan at time.net.my> wrote in message
news:mailman.1019218239.16981.python-list at python.org...
> Hi,
>
> I heard about 'Python scripting' from a member experts-exchange.com;
an
> IT knowledge sharing community. I'm working on a virtual fitting
room
> program, that needs to dynamically generate 3D human models during
run-
> time (based on users' input of their body measurements for eg.). So,
I
> heard from a member in that community that I could import Poser's 3D
> human models into my program and use Python script to read users'
input
> of their measurements at runtime and make changes in the parameters
of
> the basic model such as wasit, height, etc., so as to regenerate the
3D
> human model that matches the users' input.
>
> Eventually, Python scripting is something new to me and I would like
to
> know how I can use Python script to perform the functionality
mentioned
> above. For instance, the program I am developing is a MFC
application
> written in Visual C++. How could I link the Python scripts to my
> program then?

There are two ways to interface the standard version of Python, which
is written in portable C with other C or C++ functions.  One is to
have C++ use the Python interpreter as a subroutine library.  This is
called 'embedding' (of Python into C++).  This requires you to learn
the Python C API (Application Program Interface).  This is NOT what
the 'member' was suggesting.

The other is to have Python use your C++ functions as a fast-running
library.  This is called 'extending (of Python with C++)..  This
requires you to learn Python.  It also requires that you 'wrap' your
C++ functions with interface functions (also in C/C++).  Much of the
wrapping can be done automatically with a program called Swig.  This
IS what the 'member' was suggesting.

I suggest that, when ready, you go to www.python.org and read the
tutorial.  Since you are already programming in another language, you
should be able to try out simple Python programs in a few hours.  If
you decide you like the style of Python programming (and some people
just do not), then you can look for the documentation on 'extending'.
You will also have to choose a GUI package to use.  If you have
trouble finding something on the Python site, try searching web sites
and/or the newsgroup archives with google.

Terry J. Reedy




From gh_pythonlist at gmx.de  Tue Apr  9 20:05:00 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 10 Apr 2002 02:05:00 +0200
Subject: port scan
In-Reply-To: <d9ea2696.0204091410.7755db60@posting.google.com>
References: <d9ea2696.0204091410.7755db60@posting.google.com>
Message-ID: <20020410000500.GA3243@lilith.my-fqdn.de>

* distrex <distrex at hotmail.com> [2002-04-09 15:10 -0700]:
> I don't know if this is possible but is there a way to create a port
> scanner in python?

If by a port scanner you only want to scan for open ports, this is
surely possible, and not difficult, either. Simply try to connect to the
port, if this succeeds, you've found an open port. This is a minimal
example for a Python portscanner (only checks for TCP so far):

#!/usr/bin/env python
import sys, socket

def checkPortTCP(host, port):
    sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    try:
        sock.connect((host, port))
        return 1
    except socket.error, reason:
        return 0

def main():
    hostname = sys.argv[1]

    port_range = range(1024)
    open_ports = []
    for port in port_range:
        if checkPortTCP(hostname, port):
            open_ports.append(port)
    print open_ports

if __name__ == "__main__":
    main()

I recommend to look into the documentation of the socket module, and if
this looks like chinese to you, as it was to me, the Socket Programming
HOWTO at http://py-howto.sf.net/

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 3.0 ?C      Wind: 2.4 m/s




From starx at pacbell.net  Sat Apr 20 05:52:01 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 20 Apr 2002 09:52:01 GMT
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <Xns91F61D53BDF9BRASXnewsDFE1@130.133.1.4>

Shane Hoversten || Fri 19 Apr 2002 07:55:14p: 

> Hi - 
> 
> This is really dumb, but I just want to print out a character without
> a newline and without a space behind it.  You can't do: 
> 
> print c, 
> 
> Because Python puts a space behind it.  I can't find anywhere in the
> dox that actually talk about "print" in any detail at all.  Surely
> there must be some way to do this? 

Three people have already solved your problem. I would just like to add 
that this is far from being a dumb question. You have already experimented 
with the tools at hand, and looked for further information. You have 
defined your problem percisely, and are in effect saying "I want to do X, 
there must be a way to do X, I just don't know where to look." Such a 
specific query is easily answerable, and to be able to state your need so 
clearly cannot be called 'etraordinarily stupid', as far as i can see. =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com
"tenshi wa iranai"


From ark at research.att.com  Tue Apr  9 19:26:13 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 9 Apr 2002 23:26:13 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <a8vpp5$67d$1@panix1.panix.com>
Message-ID: <yu99pu18lai2.fsf@europa.research.att.com>

>> Now, f is an object that I can use to rebind the name x.  For
>> example, if I execute f(42), that sets x to 42.

Aahz> You really don't want to do this in Python.

One thing that fascinates me is how the nature of a language affects
the preferred idioms therein.  To help me further my understanding,
would you mind explaining to me why this particular idea is undesirable
in Python?  I can think of several possible reasons, but I don't know
which reason is most plausible to the Python community.

Aahz>  Your best bet is to pass
Aahz> around mutable objects (namely class instances and modules) and use
Aahz> setattr().  You could be grotesque and also permit lists and
Aahz> dictionaries:

Aahz>     def updateObject(obj, key, value):
Aahz>         try:
Aahz>             obj.setattr(key, value)
Aahz>         except AttributeError:
Aahz>             obj.setitem(key, value)

Ick.  I wanna keep it simple.

I'm thinking back to Snobol, which lets you do this

        foo(.x)

This would pass the name of x to foo.  That name isn't necessarily
the same as "x", because if a is an array, you can also call

        foo(.a[i])

which passes the name of the i'th element of a to foo.

Within foo, you say

        $arg = 42

(assuming that the formal parameter is named arg), which sets
x (or a[i], depending on the call) to 42.

Incidentally, Snobol variables are just like Python variables in
that assignment is really rebinding.


-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From gerhard at bigfoot.de  Sat Apr 27 09:02:33 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 27 Apr 2002 15:02:33 +0200
Subject: lint for python?
In-Reply-To: <roy-6B1ABC.08373127042002@news1.panix.com>
References: <roy-6B1ABC.08373127042002@news1.panix.com>
Message-ID: <20020427130232.GA3245@lilith.my-fqdn.de>

re subject: pyChecker

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From boud at valdyas.org  Wed Apr 24 16:06:07 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Wed, 24 Apr 2002 22:06:07 +0200
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org>
Message-ID: <3cc712c2$0$41712$e4fe514c@dreader4.news.xs4all.nl>

Dan Jacobson wrote:

> Before I start learning python, here's what I want to do: I have a
> table of Hakka Chinese words and their pronunciations.  I scan a file
> and replace any Hakka ["big5" 2, 4, 6 ... byte long strings] there
> with their pronunciations.  If it were just one character [two byte]
> words I would use the "c2t" program.  Is there a template that munches
> forth in a file and replaces the longest match in a database
> before moving on?

Is the whole file in Big-5? If so, it's easiest to use iconv to
convert the file to unicode, and open it using Python. If you have
also converted your table to unicode, you can match on unicode char.
-- 
Boudewijn Rempt | http://www.valdyas.org


From eppstein at ics.uci.edu  Wed Apr  3 03:21:06 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 03 Apr 2002 00:21:06 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <Xns91E4EF89BD4EDcliechtigmxnet@62.2.16.82> <pTpq8.53356$S52.1844974@news2.tin.it> <eppstein-E6E1AD.15234102042002@news.service.uci.edu> <Lkyq8.45231$pT1.1286029@news1.tin.it>
Message-ID: <eppstein-F70646.00210603042002@news.service.uci.edu>

In article <Lkyq8.45231$pT1.1286029 at news1.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> > expressivity and possibly adds minor functionality (ability to make code
> > that behaves differently for ints and for bools, just like now you can
> 
> It would be quite possible to distinguish between int and Booleans if
> you needed to, even if Booleans came from a specific module rather
> than from builtins, of course.  Such a need is a very rare issue, thus
> it should not be allowed to get in the way for everyday usage.

Sure, but it's only *useful* to distinguish between them if you're 
likely to actually see Booleans in your input, which will only happen if 
they aren't hidden in some back corner where only the 
external-language-interfacers can find them.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From s.schwarzer at ndh.net  Sun Apr  7 16:17:04 2002
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Sun, 07 Apr 2002 22:17:04 +0200
Subject: I'm going crazy using cgi module. HELP PLEASE!!!!
References: <9c5a942b.0204050836.58595d84@posting.google.com>
Message-ID: <3CB0A940.63F504E1@ndh.net>

Hello Enrico

Enrico Morelli wrote:
> I'm going crazy using cgi module.
> Follow a main routine of a program that receives for each day of the
> month, six fields from a form.
> 
> The first  'print ar'  inside of  'for' cycle, works fine, display the
> correct ar contents.
> The second one (outside of 'for' cycle) doesn't display nothing.

You probably mean "doesn't display anything" or "does display nothing"?

> In fact,  if I try to write the array on a file, the file will be
> empty.
> 
> I use python2.2 on Linux RH7.2  system.
> 
> Where I wrong???
>
> I use cgi modules for more and more other forms without problems.

I didn't investigate your code in detail but it may be that there's an
defect in it which causes an exception. You may use

import sys
sys.stderr = sys.stdout

at the top of your module to receive a traceback in the browser window
or catch the exception with

try:
    ...
except:
    # do something with sys.exc_info()
 
> Thanks
> 
> #----------------------------------------#
> # Main Program
> def main():
> 
>         months={1 : 'Gen',
>                 2 : 'Feb',
>                 3 : 'Mar',
>                 4 : 'Apr',
>                 5 : 'Mag',
>                 6 : 'Giu',
>                 7 : 'Lug',
>                 8 : 'Ago',
>                 9 : 'Set',
>                 10 : 'Ott',
>                 11 : 'Nov',
>                 12 : 'Dic'}
> 
>         MONTH_NUMBER="%m"
>         YEAR_FORMAT="%Y"
>         currenttime = time.time()
>         nmonth=time.strftime(MONTH_NUMBER,
> time.localtime(currenttime))
>         year=time.strftime(YEAR_FORMAT, time.localtime(currenttime))
> 
>         print "Content-Type: text/plain\n\n"
> 
>         form = cgi.FieldStorage()
> 
>         im=int(form['month'].value)
>         month=months[int(nmonth)+im]
>         nmonth=int(nmonth)+im
> 
>         planningsavefile=planningsave + '-' + month
>         dmonth=calendar.monthcalendar(int(year), nmonth)
> 
>         for dd in range(len(dmonth)):
>                 for mm in range(len(dmonth[dd])):
>                         if dmonth[dd][mm] == 0:
>                                 continue
>                         ar.append([dmonth[dd][mm],'','','','','',''])
> 
>         print ar # works fine
>         for name in form.keys():
> 
>                 if name=='submit':
>                         continue
>                  # name has the values like: 2-7 or 2-8 or 1-7b or
> 4-4, etc.
>                 d=string.split(name, '-')[0]
>                 s=string.split(name, '-')[1]

Maybe use

d, s = string.split(name, '-')[:2]

> 
>                 d=int(d)-1
>                 print name, d,s # Works fine

>                 if s=='7b':
>                         ar[d][2]=form[name].value
>                 elif s=='8':
>                         ar[d][1]=form[name].value
>                 elif s=='7':
>                         ar[d][3]=form[name].value
>                 elif s=='6':
>                         ar[d][4]=form[name].value
>                 elif s=='5':
>                         ar[d][5]=form[name].value
>                 elif s=='4':
>                         ar[d][6]=form[name].value

What about setting up a dictionary at module level ...

# I would use a better name if I knew what's this for :-)
s_to_index = {'7b': 2, '8': 1, '7': 3, ...}

and use

if s_to_index.has_key(s):
    ar[d][ s_to_index[s] ] = form[name].value

instead of if ... elif ... elif ...

>                 print ar # works fine
> 
>         print ar # doesn't display nothing

Stefan


From marklists at mceahern.com  Mon Apr 29 00:52:21 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 28 Apr 2002 23:52:21 -0500
Subject: configuring mxCGIPython
In-Reply-To: <3CC996B1.978C346A@lemburg.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEBLKNAA.marklists@mceahern.com>

I finally got cgipython to work.  Since my ISP didn't have libxmlparse.so.1
and libxmltok.so.1, I had to do this:

1.  Use a bash shell script as a wrapper to cgipython.

2.  Upload these files to /bin in my account along with cgipython:

  libxmlparse.so.1
  libxmltok.so.1

This is also where I put my Python modules.

3.  Put a .htaccess file in /bin to prevent web access:

  <Limit GET>
  order deny,allow
  deny from all
  </Limit>

2.  In the shell script, set LD_LIBRARY_PATH and PYTHONPATH to /bin.  This
is what the shell script looks like:

  #!/bin/bash
  bin=/home/<username>/bin
  export LD_LIBRARY_PATH=$bin
  export PYTHONPATH=$bin
  PATH=$PATH:.
  $bin/cgipython $bin/my.cgi

I really appreciate being able to use Python instead of having to resort to
Perl for this application.

Mark, thanks for all your help!  I'd be happy to contribute to a
troubleshooting section on deploying cgipython--if you think that would be
useful.

Best,

// mark





From rudi_sass at yahoo.com  Fri Apr  5 12:48:13 2002
From: rudi_sass at yahoo.com (Peter D)
Date: 5 Apr 2002 09:48:13 -0800
Subject: Filename Too Long
Message-ID: <be230158.0204050948.7fa6a62f@posting.google.com>

I'm using windows and trying to call os.path.getmtime() after using
os.path.walk... However, I'm choking with "[Errno 38] Filename too
long" on paths with len > ~260

Is there any way around this error?

Thanks-
Peter


From netzapper at magicstar.net  Thu Apr  4 01:53:54 2002
From: netzapper at magicstar.net (A. Jones)
Date: Thu, 04 Apr 2002 06:53:54 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de>
Message-ID: <3cabf705.91920985@mammoth.usenet-access.com>

On 4 Apr 2002 00:58:10 GMT, Marco Herrn <herrn at gmx.net> wrote:

>Hello. I know that it is possible to turn a string into a funtion call by
>using getattr(). But what I want to achieve is turning a string into an
>object name. 
>I would have a method like
>
>  foo(self, objname, objtype)
>
>Then I would call this method like
>  
>  foo(myObj, bar)
>
>Then the method foo should create an object with the name "myObj" from the
>type bar.
>
>Is such an thing possible? And if, how?

I'll probably get blasted for this by Python gurus... but, I've used
it with some success.  You can do

exec(myObjc + " = " + value)

or, if it's a class instance you want to create

exec(myObjc + " = " + bar + "()").

Toss whatever variables you want sent on to the instance inside "("
and ")" and you're homefree.

Unless there's some problem with that that an old-timer spots that I
haven't yet.

Netzapper/Aubrey Jones


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From johnroth at ameritech.net  Wed Apr  3 06:21:57 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:21:57 -0500
Subject: date / time
References: <3CAAD0D1.2040309@beth.uniforum.org.za>
Message-ID: <ualp46j3qnn21d@news.supernews.com>

"mixo" <mixo at beth.uniforum.org.za> wrote in message
news:3CAAD0D1.2040309 at beth.uniforum.org.za...
>
> Suppose there is a given date, 2002-03-01 (YYYY-MM-DD).
> How can you get the day (2002-02-28 ) before the given date and
> check that it is valid?
>

Convert it into some form that is representable in days,
subtract one day, and then convert it back. You don't need to
check validity then.

There's lots of ways of doing that, from the builtin date
functions to mxDateTime.

John Roth




From jeff at ccvcorp.com  Tue Apr  2 15:46:58 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 12:46:58 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org>
Message-ID: <MPG.1713b8a1bedfe05098968b@news.nwlink.com>

In article <mailman.1017755045.4308.python-list at python.org>, 
lac at strakt.com says...

> I am opposed to the addition of the new proposed type on the grounds
> that it will make Python harder to teach both to people who have never
> programmed before and to people who _have_.  

At first glance, I was marginally in favor of this PEP.  After 
reading Alex Martelli's arguments, I was marginally opposed to 
it.  I think that Laura's comments are *very* convincing, and I'm 
now fairly solidly opposed to it.  Well said, Laura.

(FWIW, this was the first time I've seen the Something vs. 
Nothing distinction pointed out explicitly.  I think that this 
explanation of how Python determines truth is exceptionally 
clear, and I think that the tutorial should be modified to 
include this terminology.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From johnroth at ameritech.net  Tue Apr 30 19:35:21 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 30 Apr 2002 19:35:21 -0400
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
References: <aan5hv$aor$1@bob.news.rcn.net>
Message-ID: <ucuagjg8it3v4b@news.supernews.com>

"Raymond Hettinger" <python at rcn.com> wrote in message
news:aan5hv$aor$1 at bob.news.rcn.net...
> Reading all of the posts on the deprecation pep shows that there is a
> tension between ability to remove cruft versus breakage of old
programs.
> My idea for getting the best of both is psuedo-deprecation.
>
> Let's take anything that shouldn't be there anymore (read as anything
that
> wouldn't be put in again if we were doing this from scratch) in a
separate
> section of the library reference called Obsolete Constructs.  We'll
pull it
> out of the main body of the docs and the tutorial; remove uses of it
from
> the Python library; and ask authors not to even mention the
> psuedo-deprecates
> (except in a separate section devoted to being able to read old code).

Agreed.

> I no longer think map(), filter(), and reduce() should even be
> psuedo-deprecated,
> but apply(), oct() and hex() are prime candidates.  The input()
function
> should
> still be considered for genuine deprecation and deletion, but that
reflects
> my
> view of input() as a bug.

I strongly disagree about apply(). The only use I've found for it
varies the function which it calls, not the parameter lists! Having
variable parameter lists doesn't hack it - I need the variable
function.

John Roth




From s328940 at student.uq.edu.au  Mon Apr 29 06:29:04 2002
From: s328940 at student.uq.edu.au (Alejandro Dubrovsky)
Date: Mon, 29 Apr 2002 20:29:04 +1000
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au> <7xpu0jnyyh.fsf@ruckus.brouhaha.com>
Message-ID: <3ccd1e1c$0$15474$afc38c87@news.optusnet.com.au>

On Mon, 29 Apr 2002 10:13:58 +1000, Paul Rubin wrote:

> Alejandro Dubrovsky <s328940 at student.uq.edu.au> writes:
>> def h(self):
>> 	self.thelambdafunctiontobepickled = lambda x: self.t(x,4)
>> 
>> is there any simple way of pickling that lambda function above so that
>> when i unpickle it will be set to the right function?
> 
> Lambdas are code objects which aren't portable between python versions,
> so you can't pickle them.
> 
> If you want to serialize a code object so you can export and re-import
> it to the same python version, use marshal rather than pickle.

marshaling the lambda does not seem to work, marshaling the lambda's
func_code attribute does, but when this object is unmarshaled, how do i
turn it back to a lambda function?

Alejandro


From andreas at mtg.co.at  Thu Apr 11 03:03:45 2002
From: andreas at mtg.co.at (Andreas Kostyrka)
Date: Thu, 11 Apr 2002 09:03:45 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <010420021349434531%alexis.layton@post.harvard.edu>
References: <mailman.1017466974.6751.python-list@python.org>
	<010420021349434531%alexis.layton@post.harvard.edu>
Message-ID: <20020411090345.352360de.andreas@mtg.co.at>

On Mon, 01 Apr 2002 18:46:02 GMT
Alexis Layton <alexis.layton at post.harvard.edu> wrote:

> As for int-compatibility, were we designing Python from the ground up
> it might be the case that it should be designed as a pure truth type. 
> But we are constrained by Pythons 1-2.2.x so True has to be
> truly-comparable to 1.
Nope. It must be truly-comparable to any non-zero object. As it happens this is what Python uses now as definition of Truth :)
False OTOH must truly-comparable to ["",u"",None,{},[]] (and whatever I've just forgot).

I personally have often written code that uses one variable to double as "boolean indicator" and "value storage":

founditem=None
while not founditem:
	....
	if ....:
		founditem="VALUE"

print "Found item with value:",founditem

Andreas




From tejarex at yahoo.com  Thu Apr 18 12:33:23 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 18 Apr 2002 16:33:23 GMT
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net>
Message-ID: <nzCv8.102885$%8.8427374@bin2.nnrp.aus1.giganews.com>

"Richard Gruet" <rjgruet at yahoo.com> wrote in message
news:a9mqn3$5ao$1 at aphrodite.grec.isp.9tel.net...
> Hi all
>
> I wonder why there is no 'length' (or 'len') method - or attribute -
in the
> sequences types. Having to use the built-in function len() doesn't
sound
> very Object-Oriented to me. I understand that the built-in function
len()
> must be kept for compatibility reasons, but why not to add also the
method
> len() to the sequence types ? Is there a reason ?

Python is object-based and increasingly object-oriented but not
Object-Oriented.  Originally, Guido decided that functions/methods
common to many types should be functions rather than methods.  Besides
which, only some types had/have methods.  Tuples still do not have
methods (that I know of).  Strings did not until fairly recently.
With respect to class instances, len(instance) is easily redirected to
__len__, which is consonat with Python's current policy for naming
common, cross-class methods.

Terry J. Reedy




From aleax at aleax.it  Tue Apr 16 09:35:00 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 13:35:00 GMT
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <KcUu8.22993$Ji7.1840@atlpnn01.usenetserver.com> <3cbc16f7$0$71662$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <8MVu8.26112$vF6.828988@news2.tin.it>

Boudewijn Rempt wrote:
        ...
>>     http://starship.python.net/crew/aahz/OSCON2001/index.html
>> 
>> The Queue module might help you with passing values between threads, as
>> it's specifically designed to be thread-safe.
>> 
> 
> That's funny -- only yesterday this cropped up too, and Alex Martelli sent
> me a link to:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965, where I
> quickly added a PyQt version of the example code... It uses Queue.
> Actually instantiating GUI objects from different threads might be hard. I
> would place instantiation requests on the Queue, and have the GUI thread
> execute them.

Right!  Queue is THE way to go.  Now, life would be perfect if threads
generally devoted to such tasks as GUI's or networking didn't have to
_periodically poll_ to check if there's something new on the Queue for
them (with a timeout in select, a Tkinter .after callback, a QTimer
instance in Qt, etc etc).  I.e., if Queue had some way to "jig" the
waiting thread out of its waiting state, when new stuff is posted to
the Queue, in such a way that the thread knows it must check its incoming 
Queue.  With select, I guess you could do it with an auxiliary socket
pair (but in Windows...?).  Not sure how to achieve that in the various GUI 
toolkits -- particularly, how to do it without having a call on the GUI 
from a thread that shouldn't be calling the GUI directly.  Polling is
quite a general concept, of course, but unsatisfactory in several
other ways (too rarely -> unresponsive, too often -> wasteful of
resources... what's a poor thread to DO?!-).


Alex



From mhammond at skippinet.com.au  Wed Apr 24 04:23:37 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 24 Apr 2002 08:23:37 GMT
Subject: Numeric Python, COM and Eigenvalues
References: <9e1f166b.0204220437.57a087de@posting.google.com> <aa1ft2$mn7$2@peabody.colorado.edu> <9e1f166b.0204222329.6f5685a6@posting.google.com>
Message-ID: <3CC66B94.3000101@skippinet.com.au>

Stefano wrote:
> Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aa1ft2$mn7$2 at peabody.colorado.edu>...
> 
>>Stefano wrote:
>>
>>
>>>I am working with Numpy and trying to create a COM server.
>>>I managed to call this server by Excel and get results like Inverse
>>>matrix or Cholesky decomposition but the server doesn't work as I ask
>>>for Eigenvalues/vectors (Excel does not respond anymore).
>>>Has anyone heard of a problem of this kind?
>>
>>Well, with zero example code and details provided that's going to be a hard 
>>one to get help on ;)
>>
>>Do you know for a fact you're feeding it a diagonalizable matrix? Does numpy 
>>compute the eigenvals/vects nicely if you test it with the same matrix on its 
>>own (not via excel)? Start testing with a symmetric 2x2 real matrix for which 
>>you can hand-compute the eigenvalues trivially, that might tell you where the 
>>problem is.
>>
>>Good luck,
>>
>>f.
> 
> 
> I have already tried with a trivial matrix via Shell.
> The same matrix makes Excel freeze.

Can you send the code for a trivial demo of the problem.  If you can get 
a 10 line script that fills a 2x2 matrix and hangs Excel it would help.

Mark.



From logiplexsoftware at earthlink.net  Fri Apr 26 13:43:54 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 10:43:54 -0700
Subject: Sending mail to an SMTP server that requires POP authentication ?
In-Reply-To: <aabrbv$2vg$1@norfair.nerim.net>
References: <aabrbv$2vg$1@norfair.nerim.net>
Message-ID: <20020426104354.6ed98204.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 17:30:40 +0200
Gillou wrote:

> I did not find how to do this with smtplib.
> 
> Did someone succeed ? How ?
> 
> Thanks in advance.

Hm.  I think you'll have to log in with POP first, and then for some predefined
period of time you'll be granted SMTP access.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Sun Apr 28 15:07:38 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 28 Apr 2002 19:07:38 GMT
Subject: Scripting lanuages for laboratory tests
References: <aagkfm$6kf$1@news5.svr.pol.co.uk>
Message-ID: <aahh9q$a1k$1@nyheter.chalmers.se>

In article <aagkfm$6kf$1 at news5.svr.pol.co.uk>,
Etta Makwetta <etienne at coetzeee.freeserve.co.uk> wrote:
>To whom it may concern
>
>I am tring to decide on which scripting language to use in one of our
>laborities during testing procedures, and I don't really know what the real
>advantages and disadvantages are.  I have made the following list, and would
>like to know if anybody knows of any obvoius other advantages/disadvantages
>that Python has compared to Tcl.  Any other comments about Ruby, Perl etc.
>are also welcome.
>
>Tcl:
>
>Pros: Easy to learn, Easy for simple applications, Easy GUI tool (tk),
>Portable, Easy Embeddability (from c-functions), Easy Extension (to
>c-functions), Free, (Can be object oriented with extension [incr Tcl]

Having learned both Python and Tcl, I found Python much easier to grasp.
As soon as you do anything slightly complicated with parameter passing in
Tcl, you'd better think very carefully or you will get things wrong.

For extensibility and embeddability, it is about as easy as with Python.

>
>Cons: Less exception handling, No overflow checking, Less support for
>complicated data structures, Not support calling an arbitrary dll,

Object support not built in from the start. Smaller standard library.
"Everything is a string"-paradigm painful to use.
>
>Python:
>
>Pros: Object oriented, Modular (name scope etc), Portable, Overflow
>checking, Exception handling, Easy Embeddability (from c-functions), Easy
>Extension (to c-functions), Support for complicated data structures,
>Copyrighted but free
>
>Cons: GUI slightly more complicated than Tcl, More to learn, Not support
>calling an arbitrary dll

Tkinter is much easier to use than plain tk. I have built fairly substantial
applications in both. You pay a small price in the form of having to have
both Python and tcl installed on your machine, and there are a few very
small spots where the tcl base of tkinter shows through.

Jacob Hall?n

-- 


From skip at pobox.com  Wed Apr 24 20:08:53 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 24 Apr 2002 19:08:53 -0500
Subject: An IMHO interesting optimization question...
In-Reply-To: <3cc73c81@nntp.server.uni-frankfurt.de>
References: <3cc73c81@nntp.server.uni-frankfurt.de>
Message-ID: <15559.18709.614568.220225@12-248-41-177.client.attbi.com>

    Mickey> =-------------------------------------------------=
    Mickey> # main loop. iterates over standard input as long as there's
    Mickey> # something coming in...
    Mickey> lines = sys.stdin.readlines()
    Mickey> for line in lines:
    Mickey>     for word in wordExp.findall( line ):
    Mickey>         count( word )
    Mickey> =-------------------------------------------------=

    Mickey> This takes 14 seconds on my machine which seems to be quite an
    Mickey> achievement.

    Mickey> Now on to the next. I try to eliminate the for loop by
    Mickey> substituting with a map statement, which is supposed to be
    Mickey> faster (because it's looping in C). The 3rd version looks like:

    Mickey> =-------------------------------------------------=
    Mickey> # main loop. iterates over standard input as long as there's
    Mickey> # something coming in...
    Mickey> lines = sys.stdin.readlines()
    Mickey> for line in lines:
    Mickey>     map( count, wordExp.findall( line ) )
    Mickey> =-------------------------------------------------=

    Mickey> Now this takes 17 seconds on my machine. Wow? Why that? Why does
    Mickey> it take longer?

The cost to run a loop is the startup/tear down cost plus the cost of
running all the iterations.  I believe the startup/tear down cost of map()
is fairly high, while it's per iteration cost is low (high and low being
relative to the cost of the same work using a for loop).  You have to
amortize that startup cost over a large number of iterations for it to pay
off.  My guess is that your wordExp.findall() call only returns a list of at
most a few elements.

You might try just using map to replace the outer loop.  That would
necessitate an interface change to count().  It would then expect a line and
run the inner loop itself.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From mwh at python.net  Tue Apr 16 06:01:43 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 16 Apr 2002 10:01:43 GMT
Subject: String interpolation question
References: <a9fcmo$npi$1@peabody.colorado.edu> <mailman.1018904544.23487.python-list@python.org> <a9fhad$qbt$1@peabody.colorado.edu>
Message-ID: <lkk7r8rm9a.fsf@pc150.maths.bris.ac.uk>

Fernando P?rez <fperez528 at yahoo.com> writes:

> Thanks for your suggestions. Basically imagine having many arrays
> and needing to generate output of the form (this is perl syntax,
> which for once is a million times cleaner than python for _this_
> problem):
> 
> print "$i arr1[$i]=$arr1[$i] arr2[$i]=$arr2[$i] arr3[$i]=$arr3[$i] ....\n";

When I found myself doing this, I wrote something like this:

import re

template = """\
def copy_array(self, arr1, arr2):
    {{ arr1[!] = arr2[!]; }}
"""

prog = re.compile("\\{\\{.*?\\}\\}")

def expand_template(template, n):
    def replacer(match):
        s = match.group(0)[2:-2]
        r = []
        for i in range(n):
            r.append(s.replace('!', `i`))
        return ''.join(r)
    return prog.sub(replacer, template)

then:

>>> print expand_template(template, 3)
def copy_array(self, arr1, arr2):
     arr1[0] = arr2[0];  arr1[1] = arr2[1];  arr1[2] = arr2[2]; 

>>> 

I say "something like" because my code was a bit less trivial, came
before nested scopes, and I can't quite remember the markup I used.

Ah yes, more here:

http://groups.google.com/groups?selm=m31ziyfy28.fsf_-_%40atrus.jesus.cam.ac.uk

> To me, the above line is clear, unambiguous and requires zero contortions to 
> either write or read (by the coder). 

To me it looks scarily long...

HTH,
M.

-- 
  incidentally, asking why things are "left out of the language" is
  a good sign that the asker is fairly clueless.
                                        -- Erik Naggum, comp.lang.lisp


From max at alcyone.com  Tue Apr 16 00:41:04 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 15 Apr 2002 21:41:04 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com> <a9fu9g$mj3$1@watserv3.uwaterloo.ca>
Message-ID: <3CBBAB60.2CC717D7@alcyone.com>

Joe Mason wrote:

> A warning might suffice.  If the problem is mainly writing code that
> will be
> filled in later, ignoring the warning should be fine during
> development,
> and when you're ready to release it will remind you of any you haven't
> implemented yet.

But then we've come full circle.  If it generates a warning, then there
must be a way to suppress the warning (i.e., tell the interpreter that
that's what you really meant and it's not a mistake).  The obvious
mechanism would be a keyword that means "don't do anything," and an
obvious choice for that keyword is pass.

Bam, we're right back where we started.

Removing pass seems like a terribly unuseful idea.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From TrentM at ActiveState.com  Wed Apr 17 15:22:27 2002
From: TrentM at ActiveState.com (Trent Mick)
Date: Wed, 17 Apr 2002 12:22:27 -0700
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
Message-ID: <20020417122227.A31200@ActiveState.com>

We are pleased to announce that versions 2.1.3 and 2.2.1 of
ActivePython are now available for download from
http://www.ActiveState.com/ActivePython
Both of these are bugfix releases and supercede previous ActivePython
releases.

ActivePython is ActiveState's quality-assured binary build of Python.
Builds are currently available for Windows and Linux (Solaris builds
are coming soon).

In addition to the core, ActivePython features:
    * expat for XML processing;
    * zlib for data compression;
    * Tkinter for Tk development;
    * a large set of useful Python documentation; and, 
    * on Windows, Mark Hammond's PyWin32 extensions, including the
      Pythonwin IDE, support for Python ASP, the PythonCOM system.

In addition, PyPPM (the Python Package Manager) enables you to easily
install a number of other popular Python packages. For example, try
this:
    > pyppm install Numeric
to install Numeric Python. Use this command to list all available
packages:
    > pyppm search

Please submit email feedback to ActivePython-feedback at ActiveState.com,
and file bugs against ActivePython at
http://bugs.ActiveState.com/ActivePython.

Thanks, and enjoy!

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From derek at wedgetail.com  Thu Apr  4 21:30:10 2002
From: derek at wedgetail.com (Derek Thomson)
Date: 4 Apr 2002 18:30:10 -0800
Subject: Selecting cipher suites with socket.ssl
Message-ID: <84cff98e.0204041830.7c0d0171@posting.google.com>

Hi,

I've been using the socket.ssl method to create client-side SSL
sockets. It works very nicely!

However, I'd like to know how I can specify which SSL cipher suites I
allow i.e. the cipher suites that will be sent to the server in the
cipher_suites field of the SSL ClientHello message.

I can't seem to find any documentation on this.

Thanks in advance,
Derek.


From kybic at ieee.org  Thu Apr  4 04:43:06 2002
From: kybic at ieee.org (Jan Kybic)
Date: 04 Apr 2002 11:43:06 +0200
Subject: Prototype in Python, Was: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at> <lbh8z84nake.fsf@xihu.inria.fr> <15531.7864.698585.485436@12-248-41-177.client.attbi.com> <lbh7knolreu.fsf@xihu.inria.fr> <15531.9709.924953.953487@12-248-41-177.client.attbi.com> <mailman.1017850275.2134.python-list@python.org> <0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <mailman.1017906916.22194.python-list@python.org> <3CAC1456.C8A51DC9@kfunigraz.ac.at>
Message-ID: <lbhlmc3iytx.fsf_-_@xihu.inria.fr>

> your experience gained before tell you that Python slows down
> remarkably), wouldn't it be wise to implement the function in C or C++
> from the start on?

If you knew from the beginning what is to be implemented, of course.
But in real word (especially in the research world), you do not know
what is going to work. You try something based on your preliminary
analysis and experience and it does not quite give you the result you
wanted. So you keep tweaking the algorithm (and implementation) little
by little, until you are satisfied. For this exploratory programming,
Python is perfect. You can of course write it directly in C/C++ but if
the edit/compile/run cycle takes 30s or even more for templated C++
libraries, you are much less productive. Moreover, you write much more
in C than in Python.  Once you are satisfied, you want your algorithm
to run fast. So you rewrite everything into C/C++, having the Python
implementation as a reference. Not difficult, just takes time.

> But you should take into account that Python handles C interfaces very
> well. I am not sure how you just use your Python; whether you use Python
> only in combination with Numeric and not with a
> plotting/image-processing library. I say this in the light of your

I am using many Python extensions - PIL for images, Gnuplot for
plotting, VTK for 3D graphics, ... and I am very satisfied. That's why
I would prefer to stick to Python if I can, it makes a perfect glue
language.

J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/


From jon+usenet at unequivocal.co.uk  Sat Apr 20 10:37:34 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sat, 20 Apr 2002 14:37:34 -0000
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org>
Message-ID: <slrnac2v9d.kv6.jon+usenet@snowy.squish.net>

In article <mailman.1019303438.1339.python-list at python.org>, holger krekel wrote:
>> Need a raw string there, i.e.
>> 
>>   r"\{\{(.*?)\}\}"
> 
> yes, i am sometimes confused by 'raw strings' and 'strings'. 
> Could you help by saying why exactly it is needed here?

Hmm, well, actually you don't, since \{ is not a string literal escape
sequence and Python is weird and leaves in unknown escape sequences
unchanged.

Basically, you always use raw strings when using regexps, and never
use them anywhere else. You need them because both string literals and
regexps treat \ as an escape character, so to get the \ through to the
regexp unchanged you use a raw string.


From tejarex at yahoo.com  Fri Apr 19 03:03:12 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Fri, 19 Apr 2002 07:03:12 GMT
Subject: Twisted 0.17.2 Released
References: <CvMv8.15243$ue4.582660@weber.videotron.net>
Message-ID: <QiPv8.107052$%8.9166874@bin2.nnrp.aus1.giganews.com>

"allexpro" <allexpro at myrealbox.com> wrote in message
news:CvMv8.15243$ue4.582660 at weber.videotron.net...
>   Twisted is an event-based framework for internet applications.  It
> includes a web server, a telnet server, a multiplayer RPG engine,

After reading the 'license' page in Apr 2, I looked at every page I
could find for something about the RPG engine but could not find much
at all.  Is there a page I missed or must I download and look for doc
or source which will tell me whether or not I wanted to download?

Terry





From lac at strakt.com  Sat Apr  6 20:51:14 2002
From: lac at strakt.com (Laura Creighton)
Date: Sun, 07 Apr 2002 03:51:14 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from Guido van Rossum <guido@python.org> 
   of "Sat, 06 Apr 2002 18:13:31 CDT." <200204062313.g36NDV428946@pcp742651pcs.reston01.va.comcast.net> 
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <200204061913.g36JDcdj017764@ratthing-b246.strakt.com>  <200204062313.g36NDV428946@pcp742651pcs.reston01.va.comcast.net> 
Message-ID: <200204070151.g371pEdk018077@ratthing-b246.strakt.com>

Guido:
> Laura, you've said this over and over, and I don't buy it.
> 
> Two arguments from XP apply here:
> 
> - YAGNI (You Ain't Gonna Need It), or also "do the simplest thing that
>   can possibly work".  If all you need today is a bool, use a bool.

Aha, now I understand a large part of why we disagree.  A bool is not
simpler than an int.  'Something that can change state' is simpler than
'something that can change state and must only have one of two values'.
Do you go around making functions that return bool all the time, only
then to change them to return int every time you have the idea that 
there actually are 3 states here you want to model?  Then you will
be littered with one obsolete 2-valued function every time this happens.
They pile up, and since your wretched customers have all started using
the 2-value functions, you are stuck having to maintain these things,
even though you now hate them, and want everybody to use the 3-value
versions.

> - "Refactor mercilessly".  If you find later that an int makes more
>   sense than a bool for a particular function or variable, change it!
>   It's not the end of the world.

But we agree that refactoring is not good for its own sake, it would
be better to have not needed the refactoring because we got it correct
the first time.  Refactoring is necessary because, we make mistakes,
and we learn and grow, and because our situation changes.  But if we
are writing functions that are to be used by others in libraries someplace,
we have to try to minimize the effect of our refactoring on others.
If other people have to refactor their code, every time that we have
to refactor ours, they will get pissed.  Rightly or wrongly, they will
conclude that either a) we are fools who never get it right the
first time b) this program is in a state of flux and it would be better
to wait for a more stable version or c) we are so taken with the joys
of refactoring for its own sake that we wish this wonderful experience
on everybody, whether they want to do so or not.

If enough people decide not to buy your product (or postpone
it until the later stable release) then your company goes bankrupt and
this _is_ the end of the world.  This is why deciding when to give your
poor customers the hell of refactoring is such a difficult skill. Agile
Programming was independently invented several places, in part as a
reaction against the idea that you must _never_ make the customer refactor.
You must always stick to the published design specs, even when they are
grossly, and horribly wrong because there is nothing more important than
stability.  This stinks, but not because instability is a virtue.

One reason that XP works so much better when you have the customer on board,
and can only be approximated and guessed at when you don't have frequent
meetings with your customer, is that when the customer drives the change
process it is easier to make them refactor.  They are willing to make the
changes because they are demanding the new features that require the 
changes.  In addition, when you get to know your customer, you can find
little places to put slack in because you know the particular habits they
have of how they change their minds.  If on 2 prior occasions he has
asked for a feature that required a list, and then changed his requirement
to somehting that needs a dictionary, you can have a fruitful discussion
with him.  'Any chance you are going to soon want this, that, and some
other feature?'  When your customer says 'oh yes, we were saving that for
next month, because we don't need the feature now, but will in September'
you will be glad that you have such a close relationship with the customer.

Building for a mass audience is different.  Then end-user customers have
to refactor in order that _some_ _other_ _person_  gets the new features 
he wants even though that end-user has no need or desire for the new 
feature. This is a significant barrier to sales. 

> Have you been teaching Python all those years?  I'd like to get to the
> bottom of this, because it's so against my own intuition.  I haven't
> been teaching, but I've sure helped a lot of people who were
> struggling with programming problems -- both in C and in Python.

I only discovered Python nearly 2 years ago, alas.  I wish it could have
been longer.

Laura Creighton




From robin at jessikat.fsnet.co.uk  Tue Apr 16 19:07:56 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 17 Apr 2002 00:07:56 +0100
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
 <3CBB8ABB.57BAA189@engcorp.com>
 <mailman.1018989153.2077.python-list@python.org>
 <B348A2683937CC65.3942DE81F07F4146.ED3D7BA18F22B5A1@lp.airnews.net>
Message-ID: <p+hLIIAM7Kv8Ew3l@jessikat.fsnet.co.uk>

In article <B348A2683937CC65.3942DE81F07F4146.ED3D7BA18F22B5A1 at lp.airnew
s.net>, Cameron Laird <claird at starbase.neosoft.com> writes
......
>the system might be aware that the total cost of the
>PC should be amortized as around $10,000, once all
>management overhead is properly accounted.
>
>That's only the beginning.
A serious administrator in 1995 once told me that it cost ?20000 to put
a PC on a desk. I disbelieved her at the time as the average graduate
trainee was getting slightly less than that. That organisation was
trading in billions every day and some of the salaries were huge so
perhaps it was true.
-- 
Robin Becker


From tismer at tismer.com  Tue Apr 16 07:46:58 2002
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 16 Apr 2002 13:46:58 +0200
Subject: Classes, why ? so many problems ?
References: <3cbc0473@news.comindico.com.au>
Message-ID: <3CBC0F32.1090209@tismer.com>

Graeme Matthew wrote:
> hi all im really having problems with Python can anyone please tell me why I
> cannot run this code, I keep on getting:
> SyntaxError: invalid syntax
> Traceback (most recent call last):
>   File "C:\www\cgi-bin\test.py", line 9, in ?
>     template.PrintOut()
> AttributeError: Template instance has no attribute 'PrintOut'
> 
> class Template:
> 
>     def __init__(self):
>         self.name = 'Template'
> 
>     def PrintOut(self):
>         print self.name

The above code works. You have probably some other
problems. Maybe your sys.path si wrong, and an older
version from an other directory comes into play?

Instead of
 > sys.path.append('C:\\www\\classes')
try
 > sys.path.insert(0, 'C:\\www\\classes')

If that works, you have probably different
versions somewhere.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From stephen at xemacs.org  Thu Apr 11 01:51:05 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 11 Apr 2002 14:51:05 +0900
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org> <a8vnmr$r78$1@panix1.panix.com>
Message-ID: <87hemiaily.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "aahz" == aahz  <aahz at pythoncraft.com> writes:

    aahz> Unfortunately, I have to say that the stridency you
    aahz> exhibited in this thread (to my startlement)

Trust your startlement, not your logic.

    aahz> See?  You're implying here that Guido did *not* put on his
    aahz> "language designer" hat, and in a rather subtle way, to
    aahz> boot.  You're contradicting the very point you're trying to
    aahz> make by your use of language.

It could be taken that way, if you gloss the past tense of "trusted"
with "and now I don't", implying "but he didn't".  Or you could take
it literally, applying to the case under discussion, and gloss it as
"and he did and now it's done and _I_ accept that (even though it was
479 lines important to me), so why don't y'all do the same and stop
making such a deal out of it?"

Neither gloss is very reliable just based on the words, although I
think the latter fits much better with the theme of Laura's post.  And
with Laura.

Trust the people, far more than the words.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From aleax at aleax.it  Wed Apr 17 16:59:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 20:59:27 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <ypjv8.42546$vm6.5594931@ruti.visi.com> <a9khoc$j7h$1@nets3.rz.RWTH-Aachen.DE> <a9ki3d$jip$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <Pmlv8.32367$vF6.999025@news2.tin.it>

Philipp Lenssen wrote:
        ...
> likey, '... name="' + (key == 'station') ? 'lastStation' : key + '" ...'

name = '"' + (key, 'lastStation')[key == 'station'] + '" ...'

or even better:

name = '"%s" ...' % (key, 'lastStation')[key == 'station']


Alex



From bokr at oz.net  Mon Apr 15 15:44:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 19:44:57 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com>
Message-ID: <a9fajp$76e$0@216.39.172.122>

On Mon, 15 Apr 2002 08:19:01 -0700, "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
[...]
>"Martin v. Loewis" <martin at v.loewis.de> wrote in message
>news:m3lmbpoyai.fsf at mira.informatik.hu-berlin.de...
>
>| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
>|
>| > 2) Secondly, when we are printing integers in the hex format,
>| > we are not interested in the sign-- we are looking at the
>| > bit pattern.
>|
>| Who is this "we" in this statement? When I print numbers in hex, I'm
>| interested in the hexadecimal representation if the number. For
>| *printing* it, I usually don't care what the internal representation
>| is.
IMO 'internal representation' is a red herring. We have an abstract
concept of number representation (twos complement) that corresponds to
many hardware devices' usage, but it is the abstraction that is important IMO,
not the hardware.

The abstraction comes into play because we have a legacy of using integers
as e.g., arrays of booleans with effective reinterpret-casts back and forth. But
there are valid and useful abstractions behind that. When looking at integers
cast as arrays of booleans, hex provides a representation, but w.r.t. boolean
values it is not very useful to print it as base-16 integer with sign+absolute value.

If I infer correctly, Ken is arguing that hex is more often used to view the boolean
aspect of integers than their numeric aspect. That's true for me too.

>
>That would be most people that need to get inside and take a look at the bit
>representation of a number-- when doing things like developing new random nubmer
>generators, and encryption techniques.  There are other examples of course, but
>you should get the idea.  When _I_ want to se a number in hex, it is because I
>am interested in the bit pattern.  I'm sure many others feel the same way about
>it.  I have only been programming for 30 years, (professionally for 25)-- so
>maybe I don't have enough experience to give you the right answer to this.
Calm down, sonny ;-) (I've been at it since '59 ;-)

>
>| > 3) With unification, the long and the int types should not
>| > print differently when calling the 'hex()' method.
>|
>| And indeed, they won't.
>|
>| > (The 'int' version of this method NEVER has, and shouldn't print a
>| > minus sign, and _I_ don't think the long version of 'hex()' should
>| > either.  This _is_ what needs to happen to have proper unification.
>|
>| The PEP currently says differently:
>|
>| # In most cases the long int semantics will prevail; however, the
>| # trailing 'L' from long int representations will be dropped.
>|
>| This clearly says that the long int semantics will prevail.
>
>Yes-- I read the PEP too--  when this happens, I will consider it a Python
>"_WART_", and will set about inventing ways around it for my own use.  I have
>forgotten more computer languages than most people will ever learn-- an I have
>NEVER found printing a hexadecimal number in sign and magnitude format to be
>useful...
>
I'll second that.

>I will say it again: when I am printing a number in hexadecimal, I am interested
>in the _BIT_PATTERN_ -- _NOT_ the numeric _VALUE_.  I'm sure many others feel
>this way as well.
I would guess so. I do. But for language purposes my preference is to refer
to an abstract bit pattern, not hardware (even though hardware mappings will exist
and be useful).

>
>| > 4) We are not using 1.5.2 anymore.  We are on 2.2.x now.
>| > You need to imprint on your brain that there now should be
>| > _NO_ visible difference (to the programmer) between the long
>| > and the int type.  In fact, the int and long representations
>| > are now just an implementation detail, and should be hidden,
>| > and transparent to the programmer.
>|
>| I completely agree.
>|
>| > Martin, why do you resist the unification of these two
>| > types for _ALL_ operations and methods?
>|
>| Why did you conclude that I resist? I'm just saying that the long
>| semantics will prevail, not the int semantics.
>
>Again-- when this change is made, I think it will be a Python "_WART_".
                                           ^^
Which 'it' do you mean? Personally, I agree on the hex print issue (though
not with your reference to what the implementation 'knows' about how many
bits happen to be used in reresenting longs -- the abstract values have
indefinitely extended sign bits).

>Numbers should only print the sign and magnitude when printing decimal numbers.
>When printing hexadecimal numbers, the bit pattern is what is needed-- not the
>sign and magnitude.
I agree, but there needs to be a convention for representing indefinitely extended
sign bits, and that is hard to do with backwards compatibility. So I suggest that
0x be used as planned and an alternate hex representation be introduced using 0h
as the prefix and the msb of the top hex digit is considered indefinitely extended.
This means that positive numbers with msb in position 4*b-1 (numbering from 0 at lsb)
will need a prefixed zero hex digit, E.g, 0h09 == +9, but 0h9 == -7.
>
>wart- Wart- WART- WWAAARRTT!!!!!
>

Regards,
Bengt Richter


From hgs at dmu.ac.uk  Wed Apr 17 05:32:17 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Wed, 17 Apr 2002 10:32:17 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <a9id9a$33oqt$1@ID-99513.news.dfncis.de>
Message-ID: <Pine.GSO.4.40.0204171007080.18149-100000@neelix>

On Tue, 16 Apr 2002, P. Alejandro Lopez-Valencia wrote:

>
> "Martin v. L?wis" <loewis at informatik.hu-berlin.de> escribi? en el
> mensaje news:j43cxwm1o4.fsf at informatik.hu-berlin.de...
> > Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
> >
> > > > But this is really strange; mathmodule.so must have imported at
> some
> > > > point, or setup.py would have deleted it...
> > >
> > > Hmm.  I don't know.  Anyway, here is the result of re-making the
> > > Python, but not the make test, after removing that file...
> > >
> > > Script started on Mon Apr 15 13:57:01 2002
> > > neelix hgs 1 %> gmake
> > > gcc -shared  Modules/mathmodule.o   -o Modules/mathmodule.so
> >
> > That explains it: he has enabled mathmodule.c in the *shared* section
> > of Setup.

<quote from my Modules/Setup file>
posix posixmodule.c		# posix (UNIX) system calls
_sre _sre.c			# Fredrik Lundh's new regular expressions
new newmodule.c			# Tommy Burnette's 'new' module

# The rest of the modules listed in this file are all commented out by
# default.  Usually they can be detected and built as dynamically
# loaded modules by the new setup.py script added in Python 2.1.  If

    "Usually".  I wanted to be sure they could be, so I enabled them

# you're on a platform that doesn't support dynamic loading, want to

    Given the comments below, this is still the right thing to do, isn't it?

# compile modules statically into the Python binary, or need to
# specify some odd set of compiler switches, you can uncomment the
# appropriate lines below.

    I suggest that this paragraph be re-written to reflect what is
    recommeneded, and the consequences of other choices.

# ======================================================================
        [...]

# Uncommenting the following line tells makesetup that all following
# modules are to be built as shared libraries (see above for more
# detail; also note that *static* reverses this effect):

*shared*
        [...]

# Modules that should always be present (non UNIX dependent):

array arraymodule.c	# array objects
cmath cmathmodule.c # -lm # complex math library functions
math mathmodule.c # -lm # math library functions, e.g. sin()
        [...]
# SunOS specific modules -- off by default:

sunaudiodev sunaudiodev.c
</qoute>

> >
> > I recommend to build Python from original sources, in a fresh
> > directory.

Lots of times, already.  The problem is the correct way to do it.
>
> And while you are at it... You have not compiled python itself as PIC
> code. No, gcc -shared *doesn't* work in Solaris unless you are using GNU
> Binutils instead of the native linker and assembler, something *not
> recommended at all*. Solaris' linker can't bild a dynamic loading
> library/module using object code compiled in position dependent mode.

I ran configure.  I have tried running configure with --enable-gnu-ld
and this didn't help, even with the latest binutils, so I have used the
sun ld.

>
> This means, that you need to have all external libraries compiled with
> the -fPIC flag, even if they are static. Ahh! And make sure that you
> give the linker the actual library paths you use with an -R flag, else
> the dynamic loader won't find them and you'll have to use an explicit
> LD_LIBRARY_PATH environment variable for each invocation invocation of
> the internpreter.
>
/progs/workshop5/SUNWspro/lib:/ansys50/SC2.0.1:/usr/openwin/lib:/opt/panorama/lib

That could be out of date.  I'll check that...
> --
> http://mail.python.org/mailman/listinfo/python-list
>

On 17 Apr 2002, Martin v. Loewis wrote:

> "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
>
> > And while you are at it... You have not compiled python itself as PIC
> > code. No, gcc -shared *doesn't* work in Solaris unless you are using GNU
        [...]
> Can you elaborate? gcc -shared works fine on Solaris, and binutils
> also works without problems.

AFAIK this is true.  I've not had problems like this with other
packages.
>
> > Solaris' linker can't bild a dynamic loading library/module using
> > object code compiled in position dependent mode.
>
> And rightly so. That should never be a problem, though.
>
> Actually, you *can* talk the Solaris linker to incorporate non-PIC
> text into a shared library. For that, it will make the text segment
> writable. The -mimpure-text option of gcc triggers that feature.

OK, but shouldn't this sort of information be in the configure script?
>
> > This means, that you need to have all external libraries compiled with
> > the -fPIC flag, even if they are static.
>
> If you link static libraries into Python, I recommend to make the
> modules that use them also static. Then you don't need PIC code for
> those.

Wouldn't this make the memory image of all python programs much larger?
        [...]
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Hugh





From eric at enthought.com  Thu Apr  4 09:37:32 2002
From: eric at enthought.com (eric)
Date: Thu, 4 Apr 2002 09:37:32 -0500
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>
Message-ID: <101f01c1dbe6$41a555c0$6b01a8c0@ericlaptop>

> > For this particular example, weave.inline is a perfect solution.  The C++
code
> > is pretty readable:
> >
> > from scipy import *
> >
> > def foo(a):
> >     code = """
> >            for (int i = 0; i < _Na(0);i++) {
> >                if (a(i,i) > 30) a(i,i) += a(i,i+1);
> >                else             a(i,0) *= 2.0; }
> >            """
> >     weave.inline(code,'a',compiler='gcc',
> >                  type_converters = weave.converters.blitz)
> >
> > a = rand(3,3)*60
> > print 'before:', a
> > foo(a)
> > print 'after:', a
> >
>
> Great, I am becoming even more impressed. However, the main criticism
> remains: You have to code your function twice, first in Python for
> prototyping, then in C. But I agree that weave saves you a lot of
> work.

If this is the only criticism, than we're doing quite well. :-)  The alternative
to Python right now if you need speed is to write your entire program (or
extension) in C/C++.  Of course, then the complaint becomes "Man, it sure does
run fast, but I sure wish C had dynamic typing and didn't have all this memory
management stuff." ;-) When I need to write things in C/C++, I generally
prototype them in Python first anyway.

That said, I just saw another project similar to weave come across the line on
python-dev called Pyrex.  very cool.  It allows C typing in Python code and then
compiles the code to an extension module.  It may be exactly what you are
looking for.  Lets hope it moves forward quickly.


With all these people working on the problem, I'm betting some sort of
accelerated Python is a reality in 1-2 years.  It may not be a full
implementation, but it'll likely accelerate scientific codes. So, take a look at
Pysco, Pyrex, and weave, choose the one you like, and pitch in!  We all know
that a faster Python is desired.  It even looks very possible.  The problem is
making it a reality.

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
http://homepages.ulb.ac.be/~arigo/psyco/
http://www.scipy.org/site_content/weave

see ya,

eric









From huaiyu at gauss.almadan.ibm.com  Mon Apr 29 18:45:30 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Mon, 29 Apr 2002 22:45:30 +0000 (UTC)
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to do math?)
References: <mailman.1019871888.29291.python-list@python.org>
Message-ID: <slrnacrj8a.6u4.huaiyu@gauss.almadan.ibm.com>

Tim Peters <tim.one at comcast.net> wrote:
>Except in the details, this is a lot like ssctsoos ("str() special-casing
>the snot out of strings").  That's a hypothetical function discussed at
>nauseating length several times over the years, at least on Python-Dev, but
>not recently.

[ snip background details about 'ssctsoos' (what a strange name!)]

>In other words, ssctsoos(x) *was* str(x), except that container types
>"passed it down" instead of passing repr down, and *something* special
>needed to be done for strings.

[snip]

>IOW, str() and repr() are both fine for their intended purposes, but the
>interactive prompt specifically wants neither as defined today.  What most
>people want there is str(), with containers passing str down to containees
>instead of repr.  Alas, the niggling 2 vs "2" problem has killed every
>attempt to date.

So we have identified at least the following desirable properties:

- passing down structures (lists, dicts, tuples).  Call it self-recursive.
- differentiating between 2 and "2".  Call it diff-num-str.
- showing floating numbers without extra precision.  Call it simple-float.
- showing national characters as in str().  Call it smart-char.

It appears that simply modifying str() or repr() or the interactive prompt
cannot achieve all these without losing some other nice properties they have
today. 

If that observation is correct, then introducing disp() __disp__() to take
care of all these issues in one stroke would appear to be worth it.  (Of
course the name can be chosen later.)

Huaiyu


From bokr at oz.net  Sat Apr  6 22:18:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Apr 2002 03:18:58 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <0rsr8.1752$CU.3728@rwcrnsc51.ops.asp.att.net>
Message-ID: <a8odr2$rdh$0@216.39.172.122>

On Sat, 06 Apr 2002 01:45:32 GMT, "Emile van Sebille" <emile at fenx.com> wrote:

>"robin and jim" <robinjim at earthlink.net>
>> Is there a way to convert a single quoted string to a triple quoted
>string?
>>
>> For example, given:
>>
>> ''' abc %d xyz '''
>>
>> the result of ''' abc %d xyz''' % 10 is:
>>
>> ' abc 10 xyz '
>>
>> which is a single quoted string.
>>
>> I would like to convert the result to a triple-quoted string.
>>
>>
>
>What do you want to do with it?  Once a string is created, it doesn't
>matter which style quoting was used -- it's simply a string.
>
True, but notice:

 >>> """ ' """
 " ' "
 >>> """ " """
 ' " '
 >>> """ ' " """
 ' \' " '
 >>>

perhaps he would have liked """ ' " """ back instead of ' \' " '
Seems like just as valid as switching the other quotes, e.g., a quick hack
to simulate extending the quote-choosing of repr:

 >>> def repr3(x):
 ...     rep = repr(x)
 ...     if rep.find('\\n')<0 and rep[1:-1].find("\\'")<0: return rep
 ...     rep=rep[1:-1].replace('\\n','\n')
 ...     rep=rep.replace("\\'","'")
 ...     return '"""' + rep + '"""'
 ...
 >>> tests = [""" ' """, """ " """, """ ' " """, """Line 1
 ... Line 2"""]
 >>> for t in tests: print 'plain: %s\n repr: %s\nrepr3: %s\n'% (t, repr(t), repr3(t))
 ...
 plain:  '
  repr: " ' "
 repr3: " ' "

 plain:  "
  repr: ' " '
 repr3: ' " '

 plain:  ' "
  repr: ' \' " '
 repr3: """ ' " """

 plain: Line 1
 Line 2
  repr: 'Line 1\nLine 2'
 repr3: """Line 1
 Line 2"""

Regards,
Bengt Richter


From cbbrowne at acm.org  Sun Apr 28 16:06:48 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: Sun, 28 Apr 2002 16:06:48 -0400
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com> <aahhsg$a1k$2@nyheter.chalmers.se>
Message-ID: <m3d6wj1tbb.fsf@chvatal.cbbrowne.com>

In the last exciting episode, jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen) wrote::
> In article <z4Xy8.72324$z55.71416 at atlpnn01.usenetserver.com>,
> Steve Holden <sholden at holdenweb.com> wrote:
>>"J.Jacob" <joost_jacob at hotmail.com> wrote in message
>>news:13285ea2.0204270857.2f34f8a4 at posting.google.com...
> ... Complicated suggestion by Steve deleted.
>>>
>>> # file fortune.py
>>> import random, os
>>>
>>> def randlistindex(yourlist):
>>>     "Return a random index number into yourlist"
>>>     return random.randint(0, len(yourlist)-1)
>>>
>>> def quote():
>>>     """
>>>     Return a quote <string> from file allfortunes,
>>>     assume the file allfortunes is located in the same
>>>     directory as this module."""
>>>     datafile = os.path.join(
>>>       os.path.dirname(__import__(__name__).__file__),
>>>       'allfortunes' )
>>>     quotesfile = open(datafile)
>>>     quoteslist, quote = [], ''
>>>     for s in quotesfile.readlines():
>>>         if s[0] == '%':
>>>             if quote:
>>>                 quoteslist.append(quote)
>>>                 quote = ''
>>>         else: quote = quote + s
>>>     if quote: quoteslist.append(quote)
>>>     print quoteslist[randlistindex(quoteslist)],
>
> Simple suggestion by me added:
>
> 1. Get the size of the file.
> 2. Get a random number 0 - size-1.
> 3. Index to this point in the file.
> 4. Iterate back to the nearest previous '%'.
> 5. Print fortune from there.
>
> This will of course skew the probability of getting printed towards longer
> cookies, but nobody will notice that.

The approach I take is to read through the cookies, one by one, and at
each one, randomly choose to keep the current cookie, thus:

(defun end-of-cookie ()
  (incf *total-cookies*) ;;; Increment count
  (if (= (random *total-cookies*) 0)
      (setq *keptcookie* *currentcookie*))
  (setq *currentcookie* '()))  

This gives an equal probability for each cookie in the file.

That's not as fast as your approach, but has similar memory
consumption characteristics, and is unbiased.

If efficiency were truly important, it would make sense to set up a
separate array, associate 1 with the start point of message 1, 2 with
the start point of message 2, and so forth, and you'd run through the
cookies to rebuild the hash table every time the cookies are changed.
-- 
(reverse (concatenate 'string "moc.enworbbc@" "sirhc"))
http://www.cbbrowne.com/info/multiplexor.html
When there's a will, I want to be in it. 


From mgilfix at eecs.tufts.edu  Tue Apr 23 15:06:07 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 23 Apr 2002 15:06:07 -0400
Subject: non-blocking sockets
In-Reply-To: <aa4367$6jg$2@nntp6.u.washington.edu>; from donn@u.washington.edu on Tue, Apr 23, 2002 at 04:47:03PM +0000
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf> <mailman.1019576266.25727.python-list@python.org> <aa4367$6jg$2@nntp6.u.washington.edu>
Message-ID: <20020423150607.B28052@eecs.tufts.edu>

On Tue, Apr 23 @ 16:47, Donn Cave wrote:
> Quoth Michael Gilfix <mgilfix at eecs.tufts.edu>:
> |   You need to use the select module to poll the socket. The select
> | module allows you to specify how long you want your timeout (in this
> | case, you probably want a timeout of 0, which is simply a poll) and
> | the module tells you which sockets are waiting for reading/writing.
> |
> |   import select
> |   read_sock = <my readable socket>
> |   w_ready, r_ready, e_ready = select.select ([ ], [ read_sock ], [ ], 0)
> |
> |   if read_sock in r_ready:
> |     # Process socket
> |
> |   ... will probably do what you want.
> 
> Yes - and then, we might add, you don't need to make the
> socket non-blocking.

  Not necessarily. Depends on his application and how much trust he
places in his network operations..

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From donn at drizzle.com  Wed Apr 24 23:08:10 2002
From: donn at drizzle.com (Donn Cave)
Date: Thu, 25 Apr 2002 03:08:10 -0000
Subject: stdin - seek() issue on Linux.
References: <mailman.1019617785.428.python-list@python.org> <1019625593.731162@yasure> <aa71nq$tukai$1@fido.engr.sgi.com>
Message-ID: <1019704085.667496@yasure>

Quoth pj at engr.sgi.com (Paul Jackson):
| The seek semantics are lower than the stdio package.  They are
| found in the semantics of the kernel.  Look at the open(2)
| and lseek(2) man pages on most any Unix or Linux or Posix
| emulation thereof.

True, and it's probably not a coincidence, since C and UNIX
are historically related.  But C is much more widely supported
than just UNIX or POSIX 1003.  Whatever seek means in Python,
it inherits that meaning from C fseek (or whatever 64-bit
alternative the platform supports, fseeko etc.)

	Donn Cave, donn at drizzle.com


From anthony at interlink.com.au  Fri Apr 26 05:08:27 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Fri, 26 Apr 2002 19:08:27 +1000
Subject: ftplib and attributes 
In-Reply-To: Message from "Daniel Dittmar" <daniel.dittmar@sap.com> 
   of "Fri, 26 Apr 2002 10:44:31 +0200." <aab41f$5hk$1@news1.wdf.sap-ag.de> 
Message-ID: <200204260908.g3Q98R214091@localhost.localdomain>

>>> "Daniel Dittmar" wrote
> One very annoying feature of many UNIX based ftp servers is the following:
> drwxrwsr-x   5 python   www          8192 Mar  1 16:54 ..
> drwxrwsr-x   2 python   www          8192 Aug  7  2001 art
> 
> Note how a recent file/directory has a time, where as an older file will
> have a year instead.

That's actually UNIX's "ls" command.

eg.

-rw-rw-r--   1 root     sys           928 Dec 13 02:12 /etc/vfstab
-rw-rw-r--   1 root     root          768 May 31  2000 /etc/vfstab.before_auto_mirror
-r-xr-xr-x   1 bin      bin          9968 Oct  6  1998 /etc/volcopy
-r--r--r--   1 root     bin           781 May 31  2000 /etc/vold.conf

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From jeff at ccvcorp.com  Tue Apr 30 14:31:03 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 30 Apr 2002 11:31:03 -0700
Subject: snippet source code
References: <20020427232434.13edc1ca.wilk@flibuste.net>
Message-ID: <MPG.173882c6a70163eb9896cb@news.nwlink.com>

In article <20020427232434.13edc1ca.wilk at flibuste.net>, 
wilk at flibuste.net says...
> hi,
> 
> when i learn i make a lot of little examples for me. Now i want to share
> my work, where can i put them ? I know activestate cookbook, sourceforge,
> but is there something else ?

For small little snippets and minor projects, try Useless Python 
(www.uselesspython.com).  It's intended for beginners to share 
ideas, code, challenges, etc...


-- 

Jeff Shannon
Technician/Programmer
Credit International


From phr-n2002a at nightsong.com  Tue Apr  2 22:17:21 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 19:17:21 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <kupu1hple7.fsf@lasipalatsi.fi>
Message-ID: <7xlmc54gji.fsf@ruckus.brouhaha.com>

Erno Kuusela <erno-news at erno.iki.fi> writes:
> since 1.5, i find the following changes to the language come
> immediately mind as worthy:
> - nested_scopes
> - string methods
> - unicode (maybe, haven't really used or needed it yet)
> - new-style classes (ditto)
> 
> the following language changes come immediately
> to mind as feeling superfluous:
> - list comprehensions (map/filter)
> - f(*args, **kw) (apply)
> - staticmethod/classmethod
> 
> i haven't decided if this is a good ratio or not, yet.

I see it as being a deeper matter than a "ratio".  I think it's wise
to resist making deep changes to the languages without deep benefits.

Nested scopes and new-style classes were deep changes with deep benefits,
so they were justified.

List comprehensions and f(*args) were shallow changes, so they can be
justified even if they don't provide deep benefits.  I think both were
worth adding, though their benefits weren't all that deep.

I don't understand staticmethod/classmethod so I won't remark on it.

I see Unicode as correcting an implementation deficiency-- it may have
been distasteful to fix, but it probably had to be done if the
language is supposed to stay current.  There just wasn't much
discretion available.

Bools seem to me to be an example of a deep change with shallow
benefits, so it's harder to justify than any of the above.


From tdelaney at avaya.com  Tue Apr 30 00:37:56 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 30 Apr 2002 14:37:56 +1000
Subject: String handling bug in Python
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2B3@natasha.auslabs.avaya.com>

> From: David Bolen [mailto:db3l at fitlinxx.com]
> "Delaney, Timothy" <tdelaney at avaya.com> writes:
> 
> > import os.path
> > 
> > path = '//computer/path/to/file'
> > path = os.path.normpath(path)
> > print path
> 
> Um, yeah, but why bother?  I suppose if I were going to present it to
> a user but wanted the source to use forward slashes I might do this,
> but it's already a Windows specific reference, so I'd probably just go
> with backslashes in the code or just leave the forward slashes
> unconverted.

1. You may need to deal with user-specified paths (e.g. from a properties
file).

2. Who knows what platform your code may be used on. In particular, someone
may try to use it on the (Classic) Mac OS, which uses colons as path
delimiters (and doubled colons for the parent folder). This of course is
most useful for relative paths, but see (1) as well.

While checking this, I noted a bug in the windows version of the docs - in
the "Index" tab, selecting "os.path (standard module)" then selecting
"os.path (standard module)" in the dialog takes you to the macpath module. I
don't know who should be informed of this ...

Tim Delaney




From greg at cosc.canterbury.ac.nz  Mon Apr 22 21:42:13 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 23 Apr 2002 13:42:13 +1200
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <a9l0bo$h1j$1@panix1.panix.com> <6qd6wx72gh.fsf@abnoba.intevation.de>
Message-ID: <3CC4BBF5.4CBD0276@cosc.canterbury.ac.nz>

Bernhard Herzog wrote:
> 
> Why would you want to implement __iadd__ for immutable objects?

So that, e.g. i += 1 works when i is an integer.

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


From juenglin at informatik.uni-freiburg.de  Tue Apr 30 03:04:13 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 30 Apr 2002 09:04:13 +0200
Subject: Is 'isinstance()' the right thing?
Message-ID: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de>

Hi,

when I want to make sure within a function that it can deal with
the arguments passed, the only choice is to check it's type via
'isinstance()'. However, in general the type of an argument is 
not the crux but whether the argument supports a certain 
'interface' or 'protocol'. Thus, a function 'hasinterface()'
would appear more natural (more 'Pythonic') to me.

So, what is the proper, 'pythonic' way of argument checking?

Regards,
Ralf


From kwokng at earthlink.net  Tue Apr 30 21:40:37 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Wed, 01 May 2002 01:40:37 GMT
Subject: telnet
References: <5cFz8.127$ke3.35@newsread2.prod.itd.earthlink.net> <Xns9201FD27582Dcliechtigmxnet@62.2.16.82>
Message-ID: <pIHz8.340$%i4.324@newsread2.prod.itd.earthlink.net>

I'm not sure if I know what you mean.  Basically, I want to know how to
write that depends on the read's response.  For example,

import getpass
import sys
import telnetlib

HOST = "localhost"
user = raw_input("Enter your remote account: ")
password = getpass.getpass()

tn = telnetlib.Telnet(HOST)

tn.read_until("login: ")
tn.write(user + "\n")
if password:
    tn.read_until("Password: ")
    tn.write(password + "\n")

tn.write("ls\n")
tn.write("exit\n")

print tn.read_all()

This example only assumes tn.read_until() will read "Password".  What if the
tn returns something else?  Therefore, I want to check the "Password"
returns before I write password.

Billy Ng

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns9201FD27582Dcliechtigmxnet at 62.2.16.82...
> "Billy Ng" <kwokng at earthlink.net> wrote in
> news:5cFz8.127$ke3.35 at newsread2.prod.itd.earthlink.net:
> > tn = telnetlib.Telnet(host)
> > tn.read_until("Please enter your id")
> > tn.write(userid+"\r\n")
> >
> > From here, tn will return either "Please enter your password" or
> > "Wrong id, please enter your id again".  My question is how can I get
> > the returning string before I do the next tn.write()
>
> can't you go for a newline? often such messages print on a separate line,
> so you can just "readline" and use string methods to find out if it was
> positive.
>
> chris
>
> --
> Chris <cliechti at gmx.net>
>
>




From aleax at aleax.it  Sat Apr 27 18:45:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 22:45:50 GMT
Subject: Update locals()
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <mailman.1019939112.20825.python-list@python.org>
Message-ID: <ySFy8.28416$8D3.759092@news1.tin.it>

holger krekel wrote:
        ...
> Uff. Thanks for the pointer although it is dissappointing. I always
> liked to assume that python does this decision at run time. Doing
> it at compile time may have performance benefits but it prevents
> some nice uses.

It only prevents some terrible hacks that Python is MUCH better
off forbidding.  Python doesn't go out of its way to stop terrible
hacks, but it's sure nice when a crucial optimization has such nice
side effects.  As you've noticed, it's easy to dress up dictionary
lookup by constant keys into attribute-access syntax -- so easy,
that you and I came up with two totally different 3-liner classes
for the same purpose.  And THAT is clean, nice, and useful.  So,
why ever should the dirty, terrible hacks be allowed -- even
though realizing there IS a compile-time is in some measure a
conceptual hurdle when one's first learning Python?


Alex



From aleax at aleax.it  Sun Apr 28 04:57:54 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 08:57:54 GMT
Subject: Indexing cells in a table
References: <3ccbac7c_3@news.newsgroups.com>
Message-ID: <mQOy8.71919$vF6.2146684@news2.tin.it>

jb wrote:

> I have a base class x with attributes a,b,c,..., that is x.a=0 is o.k.
> 
> Now T is a rectangular table of elements of the type x, that is I can say
> T[4][3].c=1.
> 
> Is there a generic way to address an area of the table and assign values
> to one of the attributes in the area?
> That is, can I write a generic function setValue that works like
> 
> T.setValue( (0,0) ,(m,n) , c , 1 )

Almost -- it has to be something like:

T.setValues( (0,0), (m,n), 'c', 1)

Note the quotes around the 'c' -- you pass the attribute name as a
string.  Later I'll show a different syntax to avoid the quoting needs...

> and assigns the value 1 to the x.c-s in the specified area.
> 
> Example:
> 
>   setValue( (1,1), (2,3), c, 1)
> 
> is the same as
> 
>   T[1][1].c=1
>   T[1][2].c=1
>   ...
>   T[2][3].c=1

That would really be peculiar semantics in Python, where intervals
always follow the rule lower-bound-included, upper-bound-excluded
which helps so much in avoiding off-by-one errors.  I STRONGLY
suggest you to consider using (3,4) as the upperbounds to mean "go
to 3 excluded, and 4 excluded".

Anyway, using your semantics instead.  You could define the function
with method-syntax if and only if T is an instance of a class or type
that you control and whose sources you can modify.  Since the
function itself is clearly feasible by using only T's public interface 
(indeed that's how you define it!) I suggest you consider making the
function a normal, free-standing one instead.  Like this:

def setValues(T, (minI, minJ), (maxI, maxJ), attName, attValue):
    for i in range(minI, maxI+1):
        for j in range(minJ, maxJ+1):
            setattr(T[i][j], attName, attValue)

Note the awkware "+1" in the range calls -- this tells you there is
something peculiar about your semantics (to wit, you want upper
range included intervals).

This needs to be called as: setValues(T, (1,1), (2,3), 'c', 1).

Anyway, if T is e.g. an instance of a class you control, you can
just use this same def inside the class body and rename T to
self in the function to use your required syntax.

I would in fact suggest yet another small syntax variation -- call
this as: setValues(T, (1,1), (2,3), c=1)...:

def setValues(T, (minI, minJ), (maxI, maxJ), **attributes):
    for i in range(minI, maxI+1):
        for j in range(minJ, maxJ+1):
            for attName in attributes:
                attValue = attributes[attName]
                setattr(T[i][j], attName, attValue)

The "for attName in attributes:" is 2.2 syntax -- use instead
"in attributes.keys():" if you need portability to old Python versions.

Note that this latter syntax easily allows setting multiple attributes:
just call, e.g., setValues(T, (1,1), (2,3), a=12, b=34, c=56) should
you ever need that.  You pay little for the generality, and get a
rather pretty calling-syntax too.


Alex



From peter at engcorp.com  Wed Apr 10 20:13:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:13:27 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <231u8a.lc.ln@127.0.0.1> <a8vnt9$roa$1@panix1.panix.com> <dotv8a.un.ln@127.0.0.1> <a925gf$luk$1@panix1.panix.com>
Message-ID: <3CB4D527.E9B76876@engcorp.com>

Aahz wrote:
> 
> In article <dotv8a.un.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
> >>>Aahz wrote:
> >>>>
> >>>> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
> >>>> free to buy another book if K&R turns out to be too heavy for you right
> >>>> now, but recognize that you'll need it eventually.
[...]
> >So if I agreed with you, I'd be wrong.
> 
> Precisely.  Of course, there's no way to prove that without your trying
> to read K&R, but note that several people in this thread have agreed
> with me (even if they haven't phrased their opinions as vociferously).

Oh, I'll back you up vociferously.  I learned C before I had
access to a C compiler or computer that ran other than BASIC
by reading K&R front to back, back to front, and eleven ways
from Sunday.  Never read any other book about C since either
and I don't seem to have missed them.  I'm still amazed that
it was so readable that I didn't feel it was a labour to get
through it that way.

-Peter


From jb at cascade-sys.com  Mon Apr 29 20:24:35 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 17:24:35 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>
		<3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca>
Message-ID: <3CCDE443.CA2525FF@cascade-sys.com>


> [James J. Besemer]
>
> > I don't interpret this strongly worded and slightly sarcastic critique to
> > be hostile.

Fran?ois Pinard wrote:

> For me, sarcasm and strong wording are not welcome.  There are ways to
> disagree, and even disagree much, while staying friendly.

Americans tend to value straight talk.  I put it all under the category of
effective communication.  Being French Canadian, this may be alien to you.

Although I agree with you that sarcasm and rhetorical questions are no
substitute for a logical argument, they can put an effective emphasis on an
otherwise factual and logical post such as Logjam's (or whatever his name is).

Nor is Sarcasm or frank talk necessarily unfriendly.  My close friends tease
each other mercilessly.  After an exchange that grew a bit tense at points,
Steve Holden and I emerged better friends and better acquainted with each other
than I think we would have otherwise.  Many people rely on close friends to
utter the bare truths that mere acquaintences would avoid.

I think you need to... [all together now]... Lighten Up.

> I prefer intelligent discussions between nice people.

Yeah, "Mean People Suck."

Incidentally, by my count, *I* am the only victim of an unprovoked personal
attack on this list today.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From aahz at pythoncraft.com  Sat Apr 13 18:54:46 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Apr 2002 18:54:46 -0400
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <a9acvm$24d$1@panix1.panix.com>

In article <3cb8a755$0$12714$9b622d9e at news.freenet.de>,
Jens Baader <nospam at nospam.com> wrote:
>
>I'm downloading Python 2.2.1 right now and I wonder
>why you still produce bugfix releases for the old 2.1 development line.
>It seems that Python 2.2 is somewhat broken. If not what's the
>reason that keeps the people from upgrading to 2.2?

You've already received a couple of good answers; here's another: some
people need to develop for multiple versions of Python.  While it's
certainly possible (and not all that difficult, actually) to support
releases from 1.5.2 through 2.2.1, each additional release does take
more effort, so I can understand why someone might choose to stick with
1.5.2 through 2.1.3.

>Another question: When will we see an official ISO/ANSI standard
>for Python? I dislike working with nonstandardized languages.  The
>fact that radical language changes are still made (the type/class
>unification) seem to suggest that Python is still not mature enough for
>a real standard. Or maybe Python and the other "newschool" languages
>(like Ruby, Java) are not meant to by standardized? Don't know but I
>really would like to. Are there any plans to get Python "stable" or
>will the language continue to change with each major release of the
>interpreter?

Never.  There's a reason Guido is called Benevolent Dictator For Life:
most of us recognize that there's no way a committee of any sort could
substitute for Guido's design intuition and expertise.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From austus at adelphia.net  Sat Apr  6 12:50:57 2002
From: austus at adelphia.net (Kevin)
Date: Sat, 06 Apr 2002 09:50:57 -0800
Subject: Python interest group for the Los Angeles area.
Message-ID: <3CAF3581.8000302@adelphia.net>

I'm interested in coordinating the formation of Python interest group in 
the Los Angeles area.  Anyone else interested?  Email me.





From fdrake at acm.org  Wed Apr  3 13:29:55 2002
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 3 Apr 2002 13:29:55 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <031001c1db3c$0460ad90$ced241d5@hagrid>
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
	<yu99wuvovkmm.fsf@europa.research.att.com>
	<00af01c1db2a$e7316d10$ced241d5@hagrid>
	<200204031619.g33GJXm02577@europa.research.att.com>
	<016b01c1db31$8da51010$ced241d5@hagrid>
	<yu99ofh0vg4x.fsf@europa.research.att.com>
	<031001c1db3c$0460ad90$ced241d5@hagrid>
Message-ID: <15531.18979.357453.21774@grendel.zope.com>

Fredrik Lundh writes:
 > (Python docs tend to use "true" to mean anything which isn't
 > "false", and where operator.truth returns the right thing.  this
 > won't change post-PEP -- if you start interpreting "return true" as
 > "return True", you'll be asking for trouble)

That's right, but see my other note:  Guido would like things like
this to say "return True", which is different.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation




From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 19:15:28 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 23:15:28 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca>
Message-ID: <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com>

Brad Bollenbach <bbollenbach at shaw.ca> wrote:
>
>This is a semantic debate. I know that Stallman chose the term "Free
>Software" because he felt it was the best (and most concise) way to
>express what he was thinking of.

There is real difference between Open Source and Free Software.  Microsoft
would be happy to use some OSS such as BSD (as they have done in their
network codes), but it would recoil at any mention of Free (as in speech).

>As it turns out, *lots* of hackers talk about "Open Source" when they
>often mean exactly the same thing as what Stallman calls "Free
>Software". 

That's right.  A large portion of OSS is FS.  But another large portion is
not.  FS is a proper subset of OSS.

In concrete terms, OSS is anything with an OSI approved licence.  FS is
anything with an FSF approved licence.  My impression is all FSF licences
are OSI approved, but many OSI approved licences are opposed by FSF.

BTW, Python is OSS but not FS.

Huaiyu


From markpasc at markpasc.org  Wed Apr 17 10:03:00 2002
From: markpasc at markpasc.org (Mark Paschal)
Date: 17 Apr 2002 07:03:00 -0700
Subject: tkFileDialog "multiple" option?
Message-ID: <b41a4ff6.0204170603.19492a23@posting.google.com>

I came across this question at python.faqts.com:

http://www.faqts.com/knowledge_base/view.phtml/aid/14332

According to the Tk documentation I linked in, tkFileDialog should support
the "multiple" option. However, when I try to pass that, I get a TclError
thus:

   >>> import tkFileDialog
   >>> f = tkFileDialog.askopenfile(multiple=1)
   Traceback (most recent call last):
     File "<stdin>", line 1, in ?
     File "d:\programs\python22\lib\lib-tk\tkFileDialog.py", line 106, in
     askopenfile
       filename = Open(**options).show()
     File "d:\programs\python22\lib\lib-tk\tkCommonDialog.py", line 53, in
     show
       s = apply(w.tk.call, (self.command,) + w._options(self.options))
   TclError: bad option "-multiple": must be -defaultextension,
   -filetypes, -initialdir, -initialfile, -parent, or -title
   >>>

Am I misreading something somewhere, or is it a case of not understanding
Tkinter (I'm very new to it)?


-- 
Mark Paschal
http://markpasc.org/blog/
markpasc at markpasc.org


From bgrotanLOVELYSPAM at itea.ntnu.no  Mon Apr 22 05:32:14 2002
From: bgrotanLOVELYSPAM at itea.ntnu.no (=?iso-8859-1?Q?Bj=F8rn?= Ove =?iso-8859-1?Q?Gr=F8tan?=)
Date: Mon, 22 Apr 2002 11:32:14 +0200
Subject: strip/trim a string
Message-ID: <3CC3D89E.9115E1A5@itea.ntnu.no>

yes, I'm a newbie at python.

and I need to strip/trim a string... e.g. I have a string of unknown
length, and only care 
for the last five(5) characters.

An example string:
xxxNOGBJ  94873

where 94873 is the only part of the string I want to keep

best regards

Bj?rn Ove


From joseph.bell at bankofamerica.com  Thu Apr 25 14:02:25 2002
From: joseph.bell at bankofamerica.com (Bell, Joseph)
Date: Thu, 25 Apr 2002 13:02:25 -0500
Subject: Errors with locale during make test
Message-ID: <F931488C3634D11180D700A02461E24507986B9F@chitmd03.nt.il.nbgfn.com>

I'm trying to compile and install Python 2.2.1 on Solaris 2.6 and I keep
getting the following
error during the make test stage:

test_locale
test test_locale produced unexpected output:
**********************************************************************
*** lines 2-4 of actual output doesn't appear in expected output after line
1:
+ '%f' % 1024 == '1024.000000' != '1,024.000000'
+ '%+10.f' % -4200 == '    -4200' != '    -4,200'
+ '%-10.f' % 4200 == '4200     ' != '4,200     '
**********************************************************************

I'm using the SUNWspro/cc and have enabled the options for 
SUNWspro in the Modules/Setup file.

I think I could resolve this issue if I knew what the error message meant,
but 
I don't know why it is producing this output or what the meaning of the
message
is so I have no idea how to approach this.

Can someone shed light on this?

Joseph Bell
joseph.bell at bankofamerica.com
Unix Admin
------------------------------------------------------>>BSF
"Remember...No matter where you go,
there you are."
-Buckaroo Banzai
------------------------------------------------------>>EOF





From jeff at ccvcorp.com  Tue Apr 23 14:38:37 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 23 Apr 2002 11:38:37 -0700
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com>
Message-ID: <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com>

In article <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv at 4ax.com>, 
frr at easyjob.net says...
> 
> Whenever a customer clicks on the order button, I'd like to:
> 
> a) Save some info into a database
> b) Send the order to the payment processing company
> 
> I was planning to use a Python CGI for a. This cgi shouldn't return anything
> and shouldn't change the contents of the browser. Is this possible? O:-)  

So that the person making the order has no idea whether it 
actually went through or not?  I think that this is bad UI 
design.  When I am submitting information to a website, 
*especially* if there's a monetary transaction involved, I want 
to have some sort of confirmation that it succeeded.  Even if you 
define "success" as your script receiving the information and 
saving it for later deposit into the database.  HTTP connections 
get lost or delayed often enough to make any submission 
uncertain.  

(If I got no confirmation that my order was received, I'd be 
likely to assume that the connection was bad, and might submit 
the order again... and again... and then, when I got charged for 
a dozen orders when I only wanted one, I might consider a lawsuit 
against your company.)

> Besides, I'd like this script to return immediatelly, either if it can save or
> not the info, so if the db server is down I don't lose the sale. How can I do
> this? O:-)  I was thinking of using 2 scripts:
> 
> scripta: Forks scriptb and returns immediatelly
> scriptb: saves the info to the db

This plan should work, but again, I *really* think that you 
should be providing some sort of confirmation to the user.  Doing 
otherwise is just *begging* for all sorts of complaints.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From phr-n2002a at nightsong.com  Sun Apr 28 20:13:58 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 28 Apr 2002 17:13:58 -0700
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au>
Message-ID: <7xpu0jnyyh.fsf@ruckus.brouhaha.com>

Alejandro Dubrovsky <s328940 at student.uq.edu.au> writes:
> def h(self):
> 	self.thelambdafunctiontobepickled = lambda x: self.t(x,4)
> 
> is there any simple way of pickling that lambda function above so that
> when i unpickle it will be set to the right function?

Lambdas are code objects which aren't portable between python
versions, so you can't pickle them.

If you want to serialize a code object so you can export and re-import
it to the same python version, use marshal rather than pickle.



From mongiaamit at yahoo.com  Wed Apr  3 02:42:02 2002
From: mongiaamit at yahoo.com (Amit Mongia)
Date: 2 Apr 2002 23:42:02 -0800
Subject: problem with http post
Message-ID: <cf43c01d.0204022342.73e5c8fa@posting.google.com>

See i have a small problem. I am posting some parameters to a
particular url. Say i am posting to
www.karlulbrich.com/address/accept/. The parameter to be passed is say
myaddress. If i post the address then the link will work. If i dont
then it will return "Invalid number of arguements". So the following
code works for me.

encdata = 'myaddress=addressvalue1'
f =  urllib.urlopen("http://www.karlulbrich.com/address/accept/",
encdata)
print f.info()
print f.read()

But the following code gives me an "Invalid number of arguements"
error.

encdata = 'myaddress=addressvalue1'
conn = httplib.HTTP("karlulbrich.com")
conn.putrequest("POST", "/address/accept/")
conn.putheader('Content-type', 'application/x-www-form-urlencoded')
conn.putheader('Accept', 'text/plain') 
conn.endheaders()
conn.send(encdata)
reply, msg, hdrs = conn.getreply()
print(reply)
print(conn.getfile().read())

The reply variable tells me that the call is successfull and i am
reaching the server but the return value indicates invalid number of
arguements. This means that the server is not receiving myaddress.

Can anyone please tell me what the possible error can be. I am using
Python. 2.1.2. The error is not with urlencoding for sure as i have
tried that too.

Finally i also need to know how to upload files over https. Thats the
final aim.

Regards,
Amit Mongia.


From maxm at mxm.dk  Fri Apr  5 01:25:49 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 08:25:49 +0200
Subject: [Q] Python CGI Q
References: <8Far8.36437$tg4.432342@vixen.cso.uiuc.edu>
Message-ID: <3CAD436D.6090807@mxm.dk>

There is a module called cgi in the standard distribution, and it is 
fairly well documented.

regards Max M

John wrote:
> Hi, I'm wondering if I can write a CGI that loads/edits/saves files on
> server side.. (I don't need to worry about security concerns for this
> application) Any starting point or open-source?
> 
> 



From brian at mcconnell.net  Thu Apr 11 21:16:24 2002
From: brian at mcconnell.net (Brian McConnell)
Date: 11 Apr 2002 18:16:24 -0700
Subject: Worldwide Lexicon project, call for volunteers
Message-ID: <e2f36e18.0204111716.78d6f0a6@posting.google.com>

Worldwide lexicon... call for volunteers...

Hello. I am working on an open source project called the worldwide
lexicon (www.worldwidelexicon.org). The worldwide lexicon is built
around a simple set of SOAP methods that allow applications to
discover and communicate with dictionary servers throughout the web.
Think of this as GNUtella for dictionaries.

WWL aims to do two things: 1) provide a simple, easy to implement
protocol for locating and communicating with web dictionary and cyc
servers (there are many, but they all sit behind proprietary front
ends), 2) create a service similar to SETI at home that instead of
tapping PCs to crunch numbers, taps idle web surfers to contribute
data to dictionaries that accept public submissions.

The WWL protocol, because it is built on SOAP, will allow developers
to embed dictionary functions in a wide variety of applications. One
obvious example, a browser or text editor plug in that uses a few
lines of code to find an english-->arabic dictionary on the fly then
perform word and phrase translation queries (embedded translation
dictionary).

Another example, chat software that guides users in composing foreign
language messages. For example, I can read spanish, but have a
marginal vocabulary. A Jabber IM client that assisted me in
translating or picking spanish words on the fly would be very useful.
NOTE: WWL is not a translation system, its main purpose is to create a
simple interface that allows apps to find and communicate with
dictionaries, encyclopediae, and semantic networks.

It sounds complicated, but the individual elements of the system are
simple, and can be developed and tested independently of other
components.

The project will be announced at O'Reilly & Associates Emerging
Technologies conference where I'll be giving a talk, also watch for an
article on their site in early May. There is some introductory
material on the website today, full specs, documentation and tutorials
to be published in early May, shortly before the O'Reilly conference.

In the meantime, I am putting out a call to Python developers who
might be interested in contributing to this project. I am interested
in using Python for most of our code libraries. Of course, some
developers will prefer other languages, but I want to use Python for
as many tools as possible because it is an easy language to work with,
and will probably appeal to the type of people who will be interested
in WWL.

Thank you for your time. For more information, visit the WWL site
(www.worldwidelexicon.org), be sure to subscribe to the topica list to
receive updates and/or drop me a line.

Brian McConnell


From tim.one at comcast.net  Fri Apr 12 23:31:03 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 12 Apr 2002 23:31:03 -0400
Subject: Python 2.2.1 on Win prob?
In-Reply-To: <3cb466fe$0$16416$4c41069e@reader1.ash.ops.us.uu.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEMJONAA.tim.one@comcast.net>

[Tim H]
> On a Windows 2000 SP2 machine with the newly installed Python 2.2.1:
>
> python -c"from test import testall"
>
> the last two lines read:
>
> 1 skip unexpected on win32:
>     test_threaded_import
>
> Something to worry about?

No.  A few of the tests have to be skipped if you try to run them as a side
effect of importing something else, and test_threaded_import is one of
those.  As the comment at the top of test/testall.py says,

# Backward compatibility -- you should use regrtest instead of this module.

Here are two other ways of running test_threaded_import, and these should
work for you:

C:\Python22>python lib/test/test_threaded_import.py
Trying 20 threads ... OK.
Trying 50 threads ... OK.
Trying 20 threads ... OK.
Trying 50 threads ... OK.
Trying 20 threads ... OK.
Trying 50 threads ... OK.

C:\Python22>python lib/test/regrtest.py test_threaded_import
test_threaded_import
1 test OK.

C:\Python22>

The *intended* way to run the full test suite is via

python lib/test/regrtest.py





From dalke at dalkescientific.com  Tue Apr  9 12:48:10 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 9 Apr 2002 10:48:10 -0600
Subject: PEP 285: Adding a bool type
References: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk><a8uh5p$nu7$1@thorium.cix.co.uk> <zpF9N5ATtts8Ewc0@jessikat.demon.co.uk>
Message-ID: <a8v6bp$nc3$1@slb1.atl.mindspring.net>

Robin Becker:
>Scripts in sed or awk that are 20 years old still run because their
>languages are stable.

Just this morning I read a message from one mailing list which was
trying to get their awk script working.  The problem was a difference
between nawk and awk.  I myself recall prefering nawk over awk because
it let me access multiple files at the same time.

Since there are several different implementation of awk (awk, nawk,
and gawk) each with different features, I wouldn't suggest using
them as good examples.  Judging from the ChangeLog, gawk added some new
features a year ago.  Eg,

Tue Feb  6 11:20:21 2001  Arnold D. Robbins  <arnold at skeeve.com>

        Have `for (iggy in foo)' save the elements and loop over them.


> Python scripts that are 5 years old don't run.

Strangely enough, I do have 5 year old Python scripts which run.
So far the upgrade cycle from 1.4 to 2.2 has been for me very
smooth, and I've handled some 50-100,000 lines of Python with
a couple dozen authors.

Take Gadfly as another example.  It was last released by Aaron
in 1998, which is about 5 years old.  It still runs, except
for deprecation warnings that the 'regex' module will eventually
be removed.  That removal has been expected for only 4 years,
to give people time to migrate.  And someone is working on the
Gadfly regex->re rewrite.

>The
>rate of change is increasing precisely because product developers think
>continually changing them is advantageous.

They might think so.  Bear in mind that they might be right.

They also change things because there were bugs and other errors
in the code that need fixing.  For examples:
  - there was a thread bug in 1.5.2 that we needed to hand patch
      for every new install until 2.0 came out
  - the type/class dichotomy took a few years until people realized
      it was a problem, and it took about 7 years before it finally
      happened

These *are* advantageous.

Python, btw, has been a more stable development platform for me
than C++ (I did C++ during the early template days, and when
for(int i=0;;) changed how it handled scopes), and more stable
than Perl (I started with 4.036, and my scripts broke going to
5) and even C (I started with Turbo C for the PC, then went to a
K&R C on IRIX, then a multitude of vendor semi-ANSI C unix
compilers).  The only development language I found more stable
was Tcl, where very little changed on me between 6 and 8.

                    Andrew
                    dalke at dalkesientific.com





From sasoft at gmx.de  Mon Apr  1 05:08:37 2002
From: sasoft at gmx.de (cruciatuz)
Date: Mon, 1 Apr 2002 10:08:37 +0000
Subject: PyOS -- was: Re: ANN: cage 1.0 -- Cellular automata engine in Python
In-Reply-To: <20020401021435.GF333@lilith.hqd-internal>
References: <3CA5194E.D5AB00E1@alcyone.com> <20020401021435.GF333@lilith.hqd-internal>
Message-ID: <20020401100837.GA1536@blackscarab>

On Mon, Apr 01, 2002 at 04:14:35AM +0200, Gerhard H?ring wrote:
> * Erik Max Francis <max at alcyone.com> [2002-03-29 17:47 -0800]:
> > Here's the latest PyOS installment:
> > [...]
> 
> For one second I thought: "Hey cool - somebody wrote an operating system in
> Python." :-)
>  
> This would be a nice exercise, IMO: use an existing kernel and C library, like
> Linux and glibc, then implement everything starting, including /sbin/init and
> the standard network servers (echo, finger, ...) in Python. Add a little Python
> shell and voila - the PyOS :-)
there was (the link is dead): "spyLinux" at
http://dilbert.physast.uga.edu/~andy/minilinux.html

it was a boot-disk which booted linux and came with a small
python-interpreter. it was quite nice.



From graham at microtonal.co.uk  Sun Apr 28 15:58:00 2002
From: graham at microtonal.co.uk (Graham Breed)
Date: Sun, 28 Apr 2002 20:58 +0100 (BST)
Subject: cookies ???
References: <aah926$k1a$1@news.iucc.ac.il>
Message-ID: <memo.20020428205808.32107A@gbreed.compulink.co.uk>

eli li wrote:

> I "open" http sites using urllib2 library. Method read() retunes content
> part only, in other words html code is returned. How can I retrieve 
> cookies?

Check the headers.



From ullrich at math.okstate.edu  Tue Apr 16 11:43:22 2002
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 16 Apr 2002 15:43:22 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com>
Message-ID: <3cbc4603.79232280@news>

On Mon, 15 Apr 2002 21:42:00 -0700, Erik Max Francis <max at alcyone.com>
wrote:

>jepler at unpythonic.net wrote:
>
>>     [2, 3, 5, 7, 9, ...]
>> would generate successive primes.
>
>Pretty lousy code, since 9 is not prime.  :-)

No, that's the beauty of it. Any moron could design
something to generate primes from a hint like 
"[2, 3, 5, 7, 11, ...]"; it takes a truly brilliant 
compiler to recognize that "[2, 3, 5, 7, 9, ...]" was 
an attempt to write down the first five primes and 
respond accordingly.

>-- 
> Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
> __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
>/  \ Nationalism is an infantile sickness.
>\__/ Albert Einstein
>    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
> A new, virtual planet, every day.


David C. Ullrich


From dalke at dalkescientific.com  Thu Apr  4 03:24:31 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Thu, 4 Apr 2002 01:24:31 -0700
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <a8h2n6$ff2$1@nntp9.atl.mindspring.net>

Guido:
>Despite the negative feedback, I've decided to accept the PEP.  The
>most important perceived problem is that newbies tend to write
>
>    if x == True: ...
>
>where they should write
>
>    if x: ...
>
>I believe this problem should be solved by education (the Zen master
>hits the student on the head with a stick, and the student gets
>enlightenment)

Personally, my problems with the PEP were mere quibbles, so I didn't
comment.  I think most code that breaks with this change wasn't very
good code in the first place.  But I've reading through Boudewijn
Rempt's  QT book at
 http://www.opendocspublishing.com/pyqt/index.lxp?lxpwrap=c1267%2ehtm
and noticed


    def add(self, occasion, function):
        if self.connections.has_key(occasion) == 0:
            self.connections[occasion]=[function]
        else:
            self.connections[occasion].append(function)

when it should have said
        if not self.connection.has_key(occasion):

I know this will still work correctly since False == 0,
but I suspect this means there's many people who are doing
explicit checks for 0/1 (/whatever?) for truth tests.

Treat this post as a stick.  :)

                    Andrew
                    dalke at dalkescientific.com





From hoel at germanlloyd.org  Thu Apr 11 05:32:52 2002
From: hoel at germanlloyd.org (Berthold =?iso-8859-15?q?H=F6llmann?=)
Date: 11 Apr 2002 11:32:52 +0200
Subject: Strange segfault in extension
Message-ID: <sewuved1h7.fsf@pc970145.GermanLloyd.de>

I have an extension that, where calling a function, generates an
segmentatin fault. This is nothing new for me, but I stumbled over the
backtrace in gdb. Why is "args" getting the address 0x1? Is this
possible?

Here is the top of my Backtrace

#0  0xfe4069f0 in LexSrt_method (self=0x0, args=0x1) at module/LexSrt.c:73
#1  0xaea38 in PyCFunction_Call (func=0x1c43f0, arg=0x274c30, kw=0x0)
    at Objects/methodobject.c:80
#2  0x4abb4 in eval_frame (f=0x30d3f8) at Python/ceval.c:2004

Thanks for your help

Berthold
-- 
Dipl.-Ing. Berthold H?llmann   __   Address:       
hoel at germanlloyd.org        G /  \ L Germanischer Lloyd
phone: +49-40-36149-7374    -+----+- Vorsetzen 32/35    P.O.Box 111606
fax  : +49-40-36149-7320      \__/   D-20459 Hamburg    D-20416 Hamburg


From jorgenhansen at hotmail.com  Wed Apr 10 02:07:37 2002
From: jorgenhansen at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Hansen?=)
Date: 9 Apr 2002 23:07:37 -0700
Subject: Problem with packages and modules
References: <117880a1.0204090542.258f0f51@posting.google.com> <3CB2FA14.1050708@mxm.dk>
Message-ID: <117880a1.0204092207.7e51ff80@posting.google.com>

Max M <maxm at mxm.dk> wrote in message news:<3CB2FA14.1050708 at mxm.dk>...
> J?rgen Hansen wrote:
> 
> > This is an example of directory hierachy:
<snip example hierachy>
> > 
> > 
> > If I were to use the plot package I would type in
> >    >>> from company import plot
> > 
> > and assume that I could access modules as:
> >    >>> plot.figure.Figure()    # where Figure is a class name.
> 
> 
> You will need to add the full path to a module.
> 
> from company.plot.figure import Figure
> fig = Figure()
> 
> or
> 
> import company.plot.figure
> fig = company.plot.figure.Figure()
> 
> or
> 
> from company.plot import figure
> fig = figure.Figure()
> 
> 
> > Further if I typed in:
> >    >>> import company
> > 
> > I would assume that I could access all subpackages and their modules
> > with a dotted name (company.plot.figure.Figure() for instance), but
> > this is obviously not the case.
> >
> 
> You would need to import the modul in company/__init__.py:
> import plot
> 
> to do:
> 
> import company
> fig = company.plot.figure.Figure()
> 
> 
> Hmm... I hope this helps a bit.

It does. Thanks a lot. After I wrote, I got thinking, that my approach
would have meant that I would import the entire hierachy. Not very
smart if the hierachy, where larger.

Well thanks again, for the prompt answer.

regards
Jorgen


From buzzard at urubu.freeserve.co.uk  Tue Apr 16 02:57:41 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 16 Apr 2002 07:57:41 +0100
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <3CBB8ABB.57BAA189@engcorp.com>
Message-ID: <a9gi77$nvj$1@newsg3.svr.pol.co.uk>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CBB8ABB.57BAA189 at engcorp.com...
> Duncan Smith wrote:
> >
> > My boss is considering moving to Python (from Poplog), so I've coded up
> > something to read in records from a text file (as below) and return a
> > statistic based on the numbers of unique and paired records (as he
> > requested).  Easy enough (code below), but now he wants to compare it
for
> > speed against some existing code (Poplog?).
>
> As another "boss", I want to make the strong point that speed
> is not what Python is about (although you may well make this
> fast enough to suit him... it's not necessarily slow).
>
> Python provides exceptional benefits in the areas of development
> time, readability and therefore code maintenance, flexibility,
> reliability, and ease of learning.
>
> If he concludes that it's not "fast enough", I suspect he
> is focusing too much on a relatively unimportant issue
> and likely to miss out on a some major competitive advantages.
>
> Most of the time when someone concludes Python is not fast
> enough, they have a situation where something is expected
> to run once or twice a day, and it might take ten minutes
> with Python and one minute with C, and they conclude that
> is too slow...  meanwhile the C version is buggy, took
> four times longer to develop, and nobody likes to maintain
> it because the code is five times longer than the Python
> one would be.
>
> So make sure he takes a good hard look at just how much
> speed he really _needs_ (not wants), and at the places
> where Python really shines, before closing the door on
> your suggestion.
>
> (For what it's worth, we've more or less outlawed
> any internal tools not written in Python as being a
> waste of the developer's time, and NONE of our utilities
> are "too slow".  (No, we're not quite as anal about it
> as that sounds, but you get the point.))
>
> -Peter

I agree completely Peter.  Python is 'fast enough' for everything I've
attempted with it - even graph algorithms, MCMC etc.  But I've never had to
deal with large files before, and I didn't want to present him with
something that was unnecessarily slow (due to my code / and not Python). :-)

Duncan




From ark at research.att.com  Wed Apr 10 16:41:57 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 20:41:57 GMT
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <yu99662zmufg.fsf@europa.research.att.com> <rxZs8.254$p56.69633@newsb.telia.net> <yu99vgazifhp.fsf@europa.research.att.com> <a925du$lj6$1@panix1.panix.com>
Message-ID: <yu99y9fvffqi.fsf@europa.research.att.com>

>> Yes, that's one possibility.

Aahz> You say that as though you don't really like the idea.  But that
Aahz> *is* the Pythonic approach: return an object that you can probe
Aahz> to get the appropriate information.

I neither like nor dislike it at present -- what I'd like is to
understand as many alternatives as possible before making up my mind.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From martin at v.loewis.de  Wed Apr 17 14:46:20 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 20:46:20 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019066315.25303.python-list@python.org>
Message-ID: <m3elheyxhf.fsf@mira.informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> > You did not build statically. Instead, all extension modules are build
> > by setup.py now.
> 
> I left *shared* commented out.  So that was not enough, and neither was
> *static* apparently.

To build all modules statically, you need to to add *static* (or just
not add *shared*), and remove the comments for all modules that you
want to build statically.

> gcc   -o python \
> 		Modules/python.o \
> 		libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm

Now we are getting somewhere. This line is wrong; it ought to have the
option -Wl,-export-dynamic *if* you are using GNU ld. Can you tell
whether you are using GNU ld?

Regards,
Martin


From gimbo at ftech.net  Thu Apr 18 06:46:29 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 18 Apr 2002 11:46:29 +0100
Subject: Readline/command history w/ interactive prompt
In-Reply-To: <lk662pmi8v.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1019061574.15837.python-list@python.org> <a9kcml$69k$1@news.wrc.xerox.com> <mailman.1019117616.20456.python-list@python.org> <lk662pmi8v.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020418104629.GJ3689@andy.tynant.ftech.net>

On Thu, Apr 18, 2002 at 10:04:18AM +0000, Michael Hudson wrote:
> Andy Gimblett <gimbo at ftech.net> writes:
> 
> > On Wed, Apr 17, 2002 at 05:51:17PM +0000, Mark Jackson wrote:
> > 
> > > Try hitting control-r and typing a short string that last appeared
> > > in the line desired ("my" or even "m" would work in this case).
> > 
> > _Nice_!  Thanks for that.  One question though: where's this feature
> > documented?  :-)
> 
> "man bash", "man readline".  Not too hard...

Hmmm...  Personally, I can't imagine making the conceptual leap from
"where's the python interpreter documentation" to "I know, I'll look
at the bash man page" without prompting.

man readline, I admit, was more obvious.

Thanks for the pointers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From hamish_lawson at yahoo.co.uk  Mon Apr  1 08:23:01 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 1 Apr 2002 05:23:01 -0800
Subject: Cannot find what I'm sure is a simple command... please help.
References: <3ca7978b.251291804@mammoth.usenet-access.com> <a886fp$sq1$1@panix1.panix.com> <uafaoipgvj7c96@news.supernews.com> <3ca7b638.259145474@mammoth.usenet-access.com>
Message-ID: <915a998f.0204010523.4063b6fc@posting.google.com>

> Right, I knew I could do that.  However, when I did that, it causes a
> meltdown when I index from that list, owing to the large size of the
> file (it's a database, with each entry on a separate line, but with an
> arbitrary line size).

The third-party 'textlines' module trades speed for memory efficency.
by providing a read-on-demand substitute for file.readlines(),
returning a randomly accessible sequence that reads lines as they're
requested.

    http://www.cs.brandeis.edu/~steele/sources/python.html

Hamish Lawson


From starx at pacbell.net  Mon Apr 29 05:33:35 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 29 Apr 2002 09:33:35 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <Xns91FF1A41961E8RASXnewsDFE1@130.133.1.4>

[mail -only-]

Raymond Hettinger || Sun 28 Apr 2002 10:34:58p:

> Please send me comments on a draft PEP deprecating
> and relocating six built-in functions.  The text file is at:
> http://users.rcn.com/python/download/deppep.txt
> [pep]

Just a techinical note: if you're going to go through the effort of 
typeing something up and sending it on the group, might want to make 
sure your newsreader dosen't break your line-formatting when you send 
it. Stuff like this:

------
Abstract

    This PEP proposes to compact the list of built-in functions through
    relocation and deprecation.  map(), reduce() and filter() would ... 
a
    separate module for functionals.  pow() and divmod() would move ... 
    the math module.  input() would be eliminated altogether.
------

(i'm just trying to show how it looks, my reader wants to wrap it -
worse- and I don't want to reformat it, so i trimmed long lines). 


-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From loewis at informatik.hu-berlin.de  Wed Apr 17 13:22:13 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Apr 2002 19:22:13 +0200
Subject: mapping HTML entities to Unicode (was: Re: Unicode name questions)
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com> <009901c1e58b$28a5bd90$445d4540@Dell2> <mailman.1018994314.14044.python-list@python.org> <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de> <mailman.1019059114.19248.python-list@python.org>
Message-ID: <j44ria9r5m.fsf@informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> Is this of use to anyone besides me?  ...
> I'd also be happy to dump this in favor of a preexisting solution.

I always thought that htmlentitydefs exists for that purpose. If that
is missing entities that are defined in any HTML version, that would
be a bug.

I think it is also unfortunate that htmlentitydefs maps from entity
names to either single-byte strings (if it fits to Latin-1) or decimal
character references (otherwise). Instead, it should map to unicode
strings.

So I emagine something like

unicode = {
  'AElig': u'\u0132',
  ...
}

entitydefs = {}
reverse = {}
for k,v in unicode.items():
  v = ord(v)
  if (v < 256):
    entitydefs[k] = chr(v)
  else:
    entitydefs[k] = '&#%d;' % v
  reverse[v] = k

I think this would be quite useful, except that it might be debatable
whether the key to htmlentitydefs.reverse should be unicode characters
or ordinals.

Regards,
Martin



From max at alcyone.com  Mon Apr  8 14:53:57 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 08 Apr 2002 11:53:57 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com> <a8slgr$pin$1@nyheter.chalmers.se>
Message-ID: <3CB1E745.EDA08705@alcyone.com>

Jacob Hallen wrote:

> I have been brought up in an environment which called anything
> allowing mixed types in a list weakly typed.
> 
> If you prefer to call it dynamically typed, I don't mind.

Weak typing usually refers to the idea that operations, not the objects
themselves, decide as what types to manipulate objects.  Dynamic typing
indicates that variables (or bindings) do not have an intrinsic type.

Python is strongly typed, since objects have intrinsic type, and those
types are what determine how operations behave, rather than the reverse.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From brian at sweetapp.com  Fri Apr 26 14:43:37 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 26 Apr 2002 11:43:37 -0700
Subject: Python's Lisp heritage
In-Reply-To: <tyf1yd25zgs.fsf@pcitapi22.cern.ch>
Message-ID: <001101c1ed52$47156450$445d4540@Dell2>

Jacek wrote:
> No, as Paul said, they are used for "function calls". Admittedly, in
> mathematical expressions function call often coincide with
> sub-expressions, but it does depend on the sub-expression in
> question. Compare
> 
>    (+ 1 2 3 4 5 6 7 8 9 10)
> 
> with
> 
>    1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10
> 
> Gosh, what a lot of typing I have to do in Python, compared to lisp.

Actually, in your own example, the lisp example requires more typing.
You disingenuously inserted unnecessary spaces in the Python example.
Allow me to fix that:

Lisp:   (+ 1 2 3 4 5 6 7 8 9 10)
Python: 1+2+3+4+5+6+7+8+9+10

You could also write it like:

>>> reduce(operator.add, (1,2,3,4,5,6,7,8,9,10))
55

Your example is not really interesting, except to show that lisp has a
multi-argument add function and Python doesn't. Implementing one, as I
demonstrated above, is a one liner e.g.

def add(*args): return reduce(operator.add, args)
add(1,2,3,4,5,6,7,8,9,10)

Cheers,
Brian





From gerhard at bigfoot.de  Thu Apr 11 20:41:22 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 12 Apr 2002 02:41:22 +0200
Subject: BUG: Python 2.1.1 nttplib
In-Reply-To: <20020412112959.73a8da53.andreas@mtg.co.at>
References: <20020412112959.73a8da53.andreas@mtg.co.at>
Message-ID: <20020412004122.GB436@lilith.my-fqdn.de>

* Andreas Kostyrka <andreas at mtg.co.at> [2002-04-12 11:29 +0200]:
> Hi!
> 
> In my Python 2.1.1 installation nntplib.py:NNTP.body
> misses the second optional argument file as described in the
> documentation.

For Python 2.1.1, you have to look into the Python 2.1.1 docs. ;-)

(The file argument was introduced with Python 2.2).

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From ruediger.maehl_nospam at web.de  Wed Apr  3 08:00:33 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Wed, 3 Apr 2002 15:00:33 +0200
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <newscache$hzbztg$btb$1@www-neu.dzsh.de> <23891c90.0204030321.e19d6df@posting.google.com>
Message-ID: <newscache$2ftztg$dee$1@www-neu.dzsh.de>

"Paul Boddie" <paul at boddie.net> schrieb im Newsbeitrag
> [snip]
> Well, there's a combination of commands that one apparently needs to
> execute to restart the database server, and this appears necessary
> when recreating databases. Unfortunately, the "how to" documentation
> isn't particularly clear about this, beyond a terse summary of how it
> might be done:
>
>   http://www.sapdb.org/sap_db_howto.htm
>
> Then, there are the obvious questions? What does <DOMAIN_password>
> mean when one hasn't been specified in the instructions? It certainly
> isn't a Windows domain password, for example, but one is left
> wondering.

That's true.

> [snip]
> Indeed. Anyway, my main point was that even for people who have
> installed and evaluated a fair number of database systems, such as
> myself, the documentation for SAP DB is more "opaque" than for many
> other systems. For those people who don't tend to install lots of
> packages, I would have a hard time recommending SAP DB for that very
> reason.

You are right, I perhaps underestimated the 'easy to setup' part of the
original posting.

> Paul

R?diger




From claird at starbase.neosoft.com  Tue Apr 30 15:18:42 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 30 Apr 2002 14:18:42 -0500
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <f0fd5987.0204300105.51529004@posting.google.com> <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk> <slrnactndm.b2u.ak@ak.silmarill.org>
Message-ID: <4EDE60B998E3F80A.6590A84D8293B23B.391672C1C5A66D5F@lp.airnews.net>

In article <slrnactndm.b2u.ak at ak.silmarill.org>,
Andrei Kulakov  <ak at silmarill.org> wrote:
			.
			.
			.
>> I don't know if anyone else has noticed this but I've found that operators 
>> and sys admins use perl and that developers tend towards python. Also sys 
>> admins rarely utilise indirection but developers do.
>>
>What do you mean by indirection?
			.
			.
			.
That one more application of which is alleged to solve all 
algorithmic problems (look for "David Wheeler" in <URL:
http://www1.ics.uci.edu/~rohit/IEEE-L7-namespaces.html >).
Examples:  in Python we idomatically parametrize by putting
stuff in a dictionary, and computing with the dictionary.
The previous correspondent was saying something like this:
"sysads write
  log('/tmp/mylogfile');
where developers are likely to put
  logfile = '/tmp/mylogfile'
   ...
  log(logfile)
"

"Indirection" means something different in psychology,
journalism, and so on.
-- 

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


From jamespp35 at yahoo.com  Wed Apr 17 19:33:23 2002
From: jamespp35 at yahoo.com (Jim Henry)
Date: Wed, 17 Apr 2002 16:33:23 -0700
Subject: listobject.c:101: bad argument to internal function
Message-ID: <a9l0o9$hdg$1@web1.cup.hp.com>

I have c code which constructs a python class and calls some methods in that
class (just two!).  That python class also constructs other classes.  There
are various lists in these 'second level' classes, and when traversing one
of these lists "for line in lines:", this error pops up, always on the last
item in the list.  I tried changing the code around in some of these
classes, and now the error happens on ANOTHER list (also being traversed
with "for ... in ...:" and also on the last item in the list).

I have gone over the C code, and it is VERY simple.  It passes one parameter
(a string) to the first python method it calls (and a "print" in that method
shows the string coming through just fine), and it passes no parameters to
the second python method (which is the one that fails).

All of this stuff works just fine when run directly in python.

Am using Python 2.2, btw.

Any ideas? (please!)

Jim Henry




From phr-n2002a at nightsong.com  Thu Apr  4 06:45:31 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 03:45:31 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org> <3CAC19D8.142A918@sympatico.ca>
Message-ID: <7xhemru1pg.fsf@ruckus.brouhaha.com>

vio <vmilitaru at sympatico.ca> writes:
> Thank you all for your suggestions and for pointing out obvious
> obstacles. Yes, 100MB of java source sounds impressive, and is.  But
> there seem to be some alternatives to re-writing that project in
> python from scrap, to not turn all that java code into total
> waste.

This makes no sense.  WHY do you want to migrate all that Java to Python?
Why not leave it in Java and keep using it?  That's what none of us
are understanding.  What do you GAIN from making an unmaintainable
machine translation?


From mertz at gnosis.cx  Mon Apr 22 13:49:00 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 22 Apr 2002 13:49:00 -0400
Subject: Science And Math
References: <20020420092101.19095.84453.Mailman@mail.python.org>
 <mailman.1019299357.11513.python-list@python.org>
 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
 <Xns91F67C40EB0E0lumberjack@216.148.53.83>
 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
 <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <M0Ex8kKkXsxb092yn@gnosis.cx>         

Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote previously:
|mathematical field is roughly divided in the Platonist field, those who
|believe that mathematical objects exist somewhere out there in space
|...and the formalist field of those who believe mathematics is
|essentialy a game we play on paper with symbols with no "real" meaning
|attached to it besides the convenience in describing natural phenomena.

There is at least one philosophical position missing from Rodrigues'
enumeration.  The intuitionist approach to mathematics sees mathematical
reasoning as exposing the "rules of thought."  These rules are not
independently existing structures in the world (Platonism), but neither
is mathematics merely symbolic (Formalism).  Instead, there is a
substantial but procedural fact about the way mathematicians think about
their subject matter.

It is obvious from my other little note in the thread that I feel
sympathy with the Intuitionist idea about mathematics.  Of course, this
sympathy is somewhat forclosed by my attachment to the Axiom of Choice.
I'm not sure I'm able to quite go as far as Brouwer (even though I only
guiltily fail to).

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From bbollenbach at shaw.ca  Sat Apr 13 18:47:29 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Sat, 13 Apr 2002 22:47:29 GMT
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <Xns91EFB465B7DAbbollenbachshawca@24.66.94.159>

"Jens Baader" <nospam at nospam.com> wrote in 
news:3cb8a755$0$12714$9b622d9e at news.freenet.de:

> I'm downloading Python 2.2.1 right now and I wonder
> why you still produce bugfix releases for the old 2.1 development line.
> It seems that Python 2.2 is somewhat broken. If not what's the
> reason that keeps the people from upgrading to 2.2?

Simple; there are environments that are running versions older than the 
latest and greatest versions. If upgrading to the newest Python isn't in 
the cards for these individuals/companies/whatever, but it's known that 
there are bugs in older versions, these should be fixed.

> Another question: When will we see an official ISO/ANSI standard
> for Python? I dislike working with nonstandardized languages.
> The fact that radical language changes are still made (the
> type/class unification) seem to suggest that Python is still not
> mature enough for a real standard. Or maybe Python and
> the other "newschool" languages (like Ruby, Java) are not
> meant to by standardized? Don't know but I really would
> like to. Are there any plans to get Python "stable" or will
> the language continue to change with each major release
> of the interpreter?

Well...if you're going to flame a language for lack of "standardization" do 
it to Perl or something. :) The only thing they have is a "reference 
implementation". At least Python has something roughly resembling a 
"language spec" in its documentation, hence the fact that we have multiple 
implementations (Jython and Stackless, along with the "official" CPython).

I can't say it's held me back from using Python to pursue laziness (and 
have a little bit of fun in the process).

Just like martellibot would say: "pragmatism trumps correctness".


From martin at v.loewis.de  Tue Apr 16 14:57:04 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 20:57:04 +0200
Subject: How does xml.dom.minidom compare?
References: <a9hidh$j94$1@swifty.westend.com>
Message-ID: <m3y9fneakf.fsf@mira.informatik.hu-berlin.de>

"Philipp Lenssen" <phil at mrinfo.de> writes:

> I saw some differences listed regarding xml.dom and xml.dom.minidom at
> http://www.python.org/dev/doc/maint22/lib/dom-objects.html . I will probably
> start to collect experiences soon but does anybody already have some
> performance experiences? And are they practically interchangeable, so that
> one could replace the one with the other at a later point without much
> hassle?

Yes, the goal is that DOM implementations can be easily replaced with
each other. For minidom, 4DOM, and 4Suite's cDomlette, this works
pretty well.

Regards,
Martin



From gerhard at bigfoot.de  Thu Apr  4 19:27:02 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Fri, 5 Apr 2002 02:27:02 +0200
Subject: 1 .__nonzero__() => memory fault (2.2 win32)
In-Reply-To: <a8ipmo$qq0$1@216.39.172.122>
References: <a8ipmo$qq0$1@216.39.172.122>
Message-ID: <20020405002702.GA2145@lilith.my-fqdn.de>

* Bengt Richter <bokr at oz.net> [2002-04-05 00:04 +0000]:
>  Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>  Type "help", "copyright", "credits" or "license" for more information.
>  >>> hasattr(1, '__nonzero__')
>  1
>  >>> 1 .__nonzero__()
> 
> (You get a memory fault).

Bugs should be reported with the Sourceforge bug tracker.

This particular one is already know and even fixed in CVS:

http://sf.net/tracker/?group_id=5470&atid=105470&func=detail&aid=527816

To find it, I entered "__nonzero__" in the bug search box.

Gerhard




From owen at nospam.invalid  Wed Apr 17 13:14:02 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Wed, 17 Apr 2002 10:14:02 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz>
Message-ID: <a9kags$50q$1@nntp6.u.washington.edu>

In article <3CBB9217.8A54E272 at cosc.canterbury.ac.nz>,
 Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

>David Eppstein wrote:
>> 
>> If you've installed a *different* version of Python (Unix Python, not
>> MacPython), you can do this.  It doesn't come installed by default, and
>> I have no idea whether it can do GUI stuff (unless you also install
>> X-windows, or something).
>
>In theory, it would seem that you could install Unix python
>on Mac OS X together with the Aqua version of Tcl/Tk, and
>run Tkinter stuff.
>
>When I tried this recently, however, it didn't work. I
>got a Tkinter window, but it was "dead" -- it didn't
>respond to mouse clicks. Tcl applications run using
>wish worked, but it wouldn't work from Python for
>some reason.
>
>I'd be interested if anyone happens to know how to
>get this to work!

I got it to work as follows:
- install fink (a Debian software installer for MacOS X)
- use fink to install:
  - Python 2.2 binary (be sure to install the full version;
    in other words, avoid the "no X" version)
  - xfree-86-base binary
  - xfree-86-root binary (approx; the rootless version whatever it's 
called)
- At this point you'll have something that won't quite work (unless fink 
has been updated), since it has installed a BROKEN version of twm window 
manager. To fix this...
-Go to the OroborOSX web site and get the installer. Install that.

Fink includes some other window managers that I have not tried 
(including the vanilla unix oroborus, of which OroborOSX is a 
MacOS-X-specific variant); some of them probably work fine, and it'd 
have the advantage of letting fink manage all the code for you.

Also, note that some of the fink binaries are (or were) rather out of 
date, including xfree-86. If you want to be more up to day, check the 
source code installer and the "unstable" area.

Regards,

-- Russell
-- 
Return      owen
address     astro
garbled     washington
in header   edu


From martin at v.loewis.de  Tue Apr 16 02:55:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:55:50 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de> <eppstein-922653.14382815042002@news.service.uci.edu> <m3ads4maa1.fsf@mira.informatik.hu-berlin.de> <eppstein-16AAA5.23361915042002@news.service.uci.edu>
Message-ID: <m3k7r8ku89.fsf@mira.informatik.hu-berlin.de>

David Eppstein <eppstein at ics.uci.edu> writes:

> > Unfortunately, you cannot represent such a number in hex, atleast not
> > under any of the usual conventions
[...]
> I don't understand why not.
[...]
> 
> Similarly, the obvious and only way of representing a negative number in 
> hex is to write a minus sign followed by the hex string for the absolute 
> value of the number.

Right; this representation is natural for me as well. However, people
were objecting that hex() should not produce a sign: *that* is not
possible following the usual conventions.

> The "usual" way of representing 32-bit values in hex, ffff0123 etc, has 
> nothing to do with negative numbers, it is a way of representing 
> unsigned values.  Since Python doesn't have unsigned values, we 
> shouldn't do things that way.

I completely agree.

Regards,
Martin


From richard at stockcontrol.net  Tue Apr  2 12:14:03 2002
From: richard at stockcontrol.net (Richard Walkington)
Date: Tue, 02 Apr 2002 18:14:03 +0100
Subject: Python message board application
Message-ID: <3CA9E6DB.4060208@stockcontrol.net>

Hello,

Does anyone know of a cgi based message board app written in Python. 
I've had a look on Freshmeat, Google etc but haven't come up with much.

I know about things like Squishdot but because of my hosting situation I 
can't use Zope.

There's a huge number of such apps written PHP and Perl but I'd much 
rather use a Python based solution.

Regards
Richard Walkington




From h.e.w.frater at cs.cf.ac.uk  Wed Apr 24 09:19:55 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 24 Apr 2002 06:19:55 -0700
Subject: variable substitution
Message-ID: <e2e9e807.0204240519.688cf3cb@posting.google.com>

How do I get the %(variable)s thing to work. I need to substitute a
variable into a print statement so I can send links with sessionId's
attached back to the browser. I'm trying this at the moment, and It
says invalid sysntax.
print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
Type</a>'
I saw this in a document written by Guido that I found here:
http://www.w3j.com/6/s3.vanrossum.html


hugh


From jeff at ccvcorp.com  Wed Apr 10 14:59:07 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 10 Apr 2002 11:59:07 -0700
Subject: how to read hardware info on Win NT
References: <3cb2e686.21131515@news.easynews.net> <3cb45107.23060437@news.easynews.net>
Message-ID: <MPG.171e2b58f71bb5039896a6@news.nwlink.com>

In article <3cb45107.23060437 at news.easynews.net>, stojek at part-
gmbh.de says...
 
> I need the id of my ethernet card for a license documentation tool we
> want to write in Python. As many commercial softwares are using this
> id for generating the license keys etc. 

ISTM that this method has potential problems.  What happens if 
the software is installed on a machine with no ethernet card?  
What happens if the software is installed, and then the ethernet 
card is upgraded?  What happens if the entire machine is 
replaced, and the licensee wishes to install the software on 
their new machine?

I have always been leery about the idea of tying software 
licenses to hardware IDs, because in an upgradable system, there 
is no guarantee that those IDs will remain constant.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jwickard**nospam** at litriusgroup.com  Thu Apr 11 22:24:46 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Fri, 12 Apr 2002 02:24:46 GMT
Subject: Built in file copy utility
Message-ID: <Ozrt8.4709$c6.399223@typhoon.mn.ipsvc.net>

I just wanted to check and see if there was a built in module to copy files
from one directory to another.




From just at xs4all.nl  Fri Apr 26 10:36:49 2002
From: just at xs4all.nl (Just van Rossum)
Date: Fri, 26 Apr 2002 16:36:49 +0200
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk> <just-9E8624.12281026042002@news1.xs4all.nl> <QnyV+gA0ITy8EwpR@jessikat.demon.co.uk>
Message-ID: <just-497E77.16364926042002@news1.xs4all.nl>

In article <QnyV+gA0ITy8EwpR at jessikat.demon.co.uk>,
 Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

> eerrrhh I'm going away from this. I have no idea what LOAD_NAME is.

In that case, the answer to the subject line is "it's you" ;-)

Just


From eppstein at ics.uci.edu  Wed Apr 17 01:29:07 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 16 Apr 2002 22:29:07 -0700
Subject: list_a-list_b
References: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu> <mailman.1019017052.23597.python-list@python.org>
Message-ID: <eppstein-6EBA52.22290716042002@news.service.uci.edu>

In article <mailman.1019017052.23597.python-list at python.org>,
 Michael Gilfix <mgilfix at eecs.tufts.edu> wrote:

>   Here's a one line example that doesn't touch either of the lists
> that you might like and I think it pretty clear.
> 
>   new list = filter (lambda x, list=list b: x not in b, list a)

It's not completely clear to me (I think some of your spaces should be 
underscores, and how come the lambda function ignores its second 
argument).  I also don't see why this should be better than

    new_list = [x for x in list_a if x not in list_b]

In any case both of these will take time proportional to the product of 
the two lists' lengths, not good.

Better is to convert b to a dictionary
(where are dict comprehensions when you need them):

    dict_b = {}
    for x in list_b: dict_b[x] = 1
    new_list = [x for x in list_a if x not in dict_b]

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From matt at greenroom.com.au  Wed Apr 10 21:54:14 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Thu, 11 Apr 2002 11:54:14 +1000
Subject: port scan
References: <d9ea2696.0204091410.7755db60@posting.google.com>
Message-ID: <3CB4ECC6.1030706@greenroom.com.au>

distrex wrote:

> I don't know if this is possible but is there a way to create a port
> scanner in python? If there is is there anyone who has successfuly
> done it that can point me in the right direction. I know there is a
> way to do it with VB and C++ but I'd like to stick with making
> programs with python for now untill I understand it.
> 

Anything you can write in VB you can write in python * 10



From eppstein at ics.uci.edu  Mon Apr 29 16:04:52 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 29 Apr 2002 13:04:52 -0700
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1> <Bhgz8.58769$vm6.9975333@ruti.visi.com>
Message-ID: <eppstein-E02885.13045229042002@news.service.uci.edu>

In article <Bhgz8.58769$vm6.9975333 at ruti.visi.com>,
 grante at visi.com (Grant Edwards) wrote:

> > reduce doesn't return a list, so a list comprehension is not an
> > appropriate replacement.
> > 
> >  reduce(fn, seq, initial)
> > becomes:
> >  res = initial
> >  for v in seq: res = fn(res, v)
> 
> I find the "reduce" spelling simpler and easier to both read
> and write.

Not to mention that the two parameter reduce (without the "initial" 
argument) becomes a lot uglier...e.g. instead of

    reduce(fn, seq)

we need something like

    sseq = iter(seq)
    res = sseq.next()
    for v in sseq:
        res = fn(res, v)

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From lac at strakt.com  Wed Apr  3 13:15:50 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 03 Apr 2002 20:15:50 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from "Mark McEahern" <marklists@mceahern.com> 
   of "Wed, 03 Apr 2002 11:03:53 MDT." <NCBBLFCOHHDIKCAFGCFBEEPPKIAA.marklists@mceahern.com> 
References: <NCBBLFCOHHDIKCAFGCFBEEPPKIAA.marklists@mceahern.com> 
Message-ID: <200204031815.g33IFprq029301@ratthing-b246.strakt.com>

[Mark McEahern]
> [Laura Creighton]
> > That is more strongly worded then I would make it; I only think that
> > deciding that a function should be compelled to return only one of
> > two values requires long and careful deliberation.  This is because it
> > is likely that some day you will wish that you could provide more
> > information than this allows.  isinstance is a good example
> > of this. I would be pleased if isinstance returned more useful info
> > than just yes/no, such as : 0, no inheritance; 1: direct instance of the
> > requested class; 2: instance of a subclass of the requested class.
> > 90% of uses would still be yes/no, but this would be fine, and you would
> > get the extra information when you needed it.
> 
> That's an eye-opening and thoughtful response. 

Thank you very much. 

>                                                All I'm saying is that this
> design philosophy is completely orthogonal to the value of PEP 285.

Certainly, but if I understand you correctly, all you want is a 
consistent thing called True and False which you can use.  I already
proposed a different way to get that, which doesn't involve inventing
a new type, which looks like a bool and behaves like an int. That is
what adding True and False to __builtins__ would do.  And I
explicitly did not discuss the value of adding a new type which looks
like a bool and behaves like a bool, which would be another way to
get this benefit.

> 
> Further, if I wanted such a function, I wouldn't call it isinstance().

Yes.  This is what the 'some day' is about.  As consider extending your
functions you frequently are less than perfectly happy with the name that
you (or somebody else) chose, sometimes years ago,  before you had your 
current insight, or new need.  Then you get to decide if it is better to 
write a new function or extend the one you have got and live with the
less than perfect name.  If you write a new one, you get to decide if
you should deprecate the old one - or remove it entirely.

> 
> Cheers,
> 
> // mark

Laura Creighton




From kevin at cazabon.com  Sat Apr  6 01:23:55 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 5 Apr 2002 22:23:55 -0800
Subject: "Memory could not be Read/Written" crashes with C extension DLL... HELP PLEASE!
Message-ID: <5a4226f0.0204052223.75aa0318@posting.google.com>

I've gotten past many hurdles so far, but I can't seem to figure out
what's causing my problems.  The DLL I've built works just fine...
unless you call pyHasp (see below) multiple times (random.. sometimes
it works fine for 40 calls, sometimes it works for 300 calls, then
crashes).

The error is one of two:  'Instruction at xxxxxx referenced memory at
xxxxx.  The memory could not be "written"' or could not be "read"...
random which one it gives.

I've tried debugging, but it seems the actual violation is in
NTDLL.DLL, I'm assuming I'm causing it somehow...

Can ANYONE look at the funciton below and let me know if I am doing
something wrong, probably in the area of retaining a reference to a
PyObject, or using the wrong pointer somewhere...  ANY suggestions are
welcome, I've tried almost everything.  I've also tried commenting out
sections one at a time (the memcpy sections, the CallHasp section, and
the AsReadBuffer/AsWriteBuffer sections) to see if it gets rid of the
problem, but it hasn't helped...???

Thanks!

Kevin.

///////////////////////////////////////////////////////////////
static PyObject *
pyHasp(PyObject *self, PyObject *args)
{
  // declarations
  PyObject *pReturn = NULL;

  PyObject *pLong1 = NULL;
  PyObject *pLong2 = NULL;
  PyObject *pLong3 = NULL;
  PyObject *inBuffer = NULL;
  int inBufferSize = 0;

  char *pInBufferAsReadBuffer = NULL;
  char *pInBufferAsWriteBuffer = NULL;

  ULONG Service = 0;
  ULONG SeedCode = 0;
  ULONG LptNum = 0;
  ULONG Pass1 = 0;
  ULONG Pass2 = 0;

  ULONG p1 = 0;
  ULONG pp1 = 0;
  ULONG p2 = 0;
  ULONG p3 = 0;
  ULONG pp3 = 0;

  char *EncodeDecodeBuffer = NULL;
  ULONG pUL_EncodeDecodeBuffer = 0;

  // parse the PyObject arguments, assign to variables accordingly
  // inBuffer is a Python 'B' array, which supports the buffer
interface
  if (!PyArg_ParseTuple(args, "llllllllOl", &Service, &SeedCode,
&LptNum, &Pass1, &Pass2, &p1, &p2, &p3, &inBuffer, &inBufferSize)) {
    return Py_BuildValue("llls", -9999, -9999, -9999, "Could not parse
the Python Tuple!");
  }

  // Check the size of the inBuffer for validity
  if (inBufferSize > 1000) {
    return Py_BuildValue("llls", -9998, -9998, -9998, "Encode / Decode
buffer must be less than 1000 characters.");
  }

  if (inBufferSize < 8) {
    return Py_BuildValue("llls", -9997, -9997, -9997, "Encode / Decode
buffer must be 8 characters or more in length.");
  }

  // allocate the memory for the EncodeDecodeBuffer
  EncodeDecodeBuffer = malloc(inBufferSize);
  if (!EncodeDecodeBuffer) {
    return Py_BuildValue("llls", -9994, -9994, -9994, "Could not
allocate memory for the EncodeDecodeBuffer!");
  }
  memset (EncodeDecodeBuffer, 0, inBufferSize);

  // Obtain a pointer to the Python Buffer Interface for inBuffer
(readable)
  if (PyObject_CheckReadBuffer(inBuffer)) {
    if ((Service != 1) && (Service != 5)) {
      if (PyObject_AsReadBuffer(inBuffer, &pInBufferAsReadBuffer,
&inBufferSize) == -1) { // PyObject_AsReadBuffer is fucked up in its
return values... -1 is error, 0 is success
        return Py_BuildValue("llls", -9995, -9995, -9995, "Could not
access the passed string as a Python buffer!");
      }
    }
  } 
  else {
      return Py_BuildValue("llls", -9995, -9995, -9995, "inBuffer does
not support Buffer Interface!");
  } 

  // the Python values should NOT be modified in place, make copies to
pass to hasp function
  pp1 = p1;
  pp3 = p3;

  // copy the inBuffer to the EncodeDecodeBuffer, use memcpy so that
nulls don't cause problems
  if  ((Service != 1) && (Service != 5)) {
    memcpy(EncodeDecodeBuffer, pInBufferAsReadBuffer, (unsigned
int)inBufferSize);
  }

  // get a ULONG pointer to EncodeDecodeBuffer
  pUL_EncodeDecodeBuffer = (ULONG) EncodeDecodeBuffer;

  // send the data to the hasp driver for encoding/decoding/whatever
  // this function modifies pp1, inBufferSize, pp3, EncodeDecodeBuffer
(through the pUL_ pointer) in place
  CallHasp(Service, SeedCode, LptNum, Pass1, Pass2, &pp1,
&inBufferSize, &pp3, &pUL_EncodeDecodeBuffer);

  // copy the EncodeDecodeBuffer back to the inBuffer, through a
pointer to the Python Buffer interface (writable)
  if  ((Service != 1) && (Service != 5)) {
    if (PyObject_AsWriteBuffer(inBuffer, &pInBufferAsWriteBuffer,
&inBufferSize) == -1) {
      return Py_BuildValue("llls", -9994, -9994, -9994,
"PyObject_AsWriteBuffer Failed!");
    }
    memcpy(pInBufferAsWriteBuffer, EncodeDecodeBuffer, (unsigned
int)inBufferSize);
  }

  free(EncodeDecodeBuffer);


  // build pp1, inBufferSize, pp3 into PyObjects for returning
  pLong1 = PyLong_FromLong(pp1);
  pLong2 = PyLong_FromLong(inBufferSize);
  pLong3 = PyLong_FromLong(pp3);

  // build pLong1, pLong2, pLong3, pString into a tuple to return

  pReturn = PyTuple_New(4);
  PyTuple_SetItem(pReturn, 0, pLong1); 
  PyTuple_SetItem(pReturn, 1, pLong2);
  PyTuple_SetItem(pReturn, 2, pLong3);
  PyTuple_SetItem(pReturn, 3, inBuffer);

  return pReturn;
}


From ponderor at lycos.com  Sat Apr 20 11:34:15 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 20 Apr 2002 08:34:15 -0700
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu> <Xns91F61D53BDF9BRASXnewsDFE1@130.133.1.4>
Message-ID: <e81be8b2.0204200734.5e87edfc@posting.google.com>

Philip Swartzleonard wrote: 
>Such a 
> specific query is easily answerable, and to be able to state your need so 
> clearly cannot be called 'etraordinarily stupid', as far as i can see. =)


Thank you for this affirmation.

In addition, these posts are important for book & documentation
publishers to recognize new user needs.


From kingprad at mail.com  Sun Apr  7 16:43:37 2002
From: kingprad at mail.com (Carl)
Date: Sun, 7 Apr 2002 15:43:37 -0500
Subject: Tkiner send command
Message-ID: <3cb0ae7d$1@news.cc.umr.edu>

I'm having trouble going from the formal specs in the Tcl/Tk Reference to
writing an actual Tkinter send command. I just want to start a text editor,
either the Windows default or just notepad.exe. Can someone please give me
an example line of how to use it? Thanks very much!

Carl
kingprad at mail.com




From w3miller at yahoo.com  Tue Apr  9 10:44:41 2002
From: w3miller at yahoo.com (Walter Miller)
Date: Tue, 9 Apr 2002 07:44:41 -0700 (PDT)
Subject: SSL error: 2.1.1 works but 2.2 doesn't 
In-Reply-To: <200204090610.g396AF802261@mbuna.arbhome.com.au>
Message-ID: <20020409144441.40134.qmail@web21411.mail.yahoo.com>

Yes - looks very familiar.  Was this patch (500311)
not accepted for 2.2.1 because it's considered a
work-around?

> This is probably the same problem as
http://python.org/sf/494762,
> for which http://python.org/sf/500311 is a suggested
patch.

> It wouldn't hurt if more people tried to understand
the problem,
> though, and comment on the proposed solution.

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/




From paul at boddie.net  Tue Apr  9 09:26:24 2002
From: paul at boddie.net (Paul Boddie)
Date: 9 Apr 2002 06:26:24 -0700
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com> <3CB1CEC0.F324F480@python.org> <23891c90.0204090012.7fc80c1f@posting.google.com> <Xns91EB245A0BCD9RASXnewsDFE1@130.133.1.4>
Message-ID: <23891c90.0204090526.18675d02@posting.google.com>

Philip Swartzleonard <starx at pacbell.net> wrote in message news:<Xns91EB245A0BCD9RASXnewsDFE1 at 130.133.1.4>...
> 
> Huh? Your argument is all well and good, but what does supporing 
> releases earlier than 2.2 have to do with it's new features being badly 
> defined and immature themselves?

I don't know, since I myself never claimed that they were. (I could,
however, dig up some of the postings that expressed those, or similar,
sentiments.) However, I do feel that the more recent features seem
increasingly remote to the beginner and even the more
advanced/experienced Python developer. List comprehensions (introduced
in 2.0) do appear in the standard tutorial, but I can't easily find
generators, even though they are documented in the language reference.

Without slavishly following the development list archives (and having
other things to do during working hours) it becomes increasingly hard
to keep track of what's great in the more recent Python releases.
Consequently, people are going to feel less informed, and will
therefore be more reluctant to accept changes to the language on top
of those which they already know very little about.

Perhaps there should be a community site which focuses on such news
from the busy package maintainer's perspective, rather than the
language architect's perspective. (Not to undermine the presumably
excellent work already done in summarising the goings-on on the
development list.)

Paul


From theller at python.net  Fri Apr 19 11:10:19 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 19 Apr 2002 17:10:19 +0200
Subject: ANN: The Python Manual Index
Message-ID: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de>

Hi, folks.

A new Python-powered search engine is online at
http://starship.python.net/crew/theller/pyhelp.cgi

It searches for keywords in the Python manuals at http://www.python.org/doc/.

The script can also be downloaded, and can be configured to search HTML manual
pages installed locally. The keyword index is build on the fly.
It can also (probably) be easily integrated in your editor/ide to provide
context sensitive help.

Mark Hammond has integrated an interface to this keyword search in his cool
Mozilla/Netscape sidebar he recently announced: http://starship.python.net/crew/mhammond/mozilla

Regards,

Thomas




From skip at pobox.com  Wed Apr  3 10:24:40 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 3 Apr 2002 09:24:40 -0600
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <lbh8z84nake.fsf@xihu.inria.fr>
References: <mailman.1017795845.3464.python-list@python.org>
        <r1tq8.161131$af7.78459@rwcrnsc53>
        <3CAAB238.290440C5@kfunigraz.ac.at>
        <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
        <3CAAEF17.32333986@kfunigraz.ac.at>
        <lbh8z84nake.fsf@xihu.inria.fr>
Message-ID: <15531.7864.698585.485436@12-248-41-177.client.attbi.com>

    SM> Have you tried weave?  It's part of SciPy.  From the SciPy front page:
    >> Weave 0.2.1 Now Available: weave is a new package that provides tools
    >> for including C/C++ code within Python code.  The weave.inline()
    >> function executes C code directly within Python, weave.blitz()
    >> translates Python Numeric expressions to C++ for fast execution...

    Jan> Basically weave permits you to put both the C code and the Python
    Jan> code into one file. Not bad, but you still have to write the code
    Jan> twice.  Useful for short C fragments, not long ones.

You're thinking of weave.inline().  I was talking about weave.blitz().
Reread the last statement in the text I quoted, then take a look here:

    http://www.scipy.org/site_content/weave/tutorial.html#Blitz

The example shows creation of a couple of Numeric arrays, followed by the
embedding of a Numeric *Python* expression in a string and a call to
weave.blitz().  There's not a hint of C or C++ code at all.  blitz() does
all the code generation for you.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)



From nixx at telia.com  Sun Apr 28 14:59:29 2002
From: nixx at telia.com (Magnus)
Date: Sun, 28 Apr 2002 18:59:29 GMT
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no>
Message-ID: <lEXy8.34518$n4.7236390@newsc.telia.net>

Johan Fredrik ?hman wrote:

> Hi, there.
> 
> I previously saw a posing named "Python vs Php" and sys atht
> "... there are many packages you can download..."
> 
> However, only one package is mentioned: jonpy
> 
> What is the most popular and promising webpackages for python ?
> 
> --
> JF?

Hi,

I tested Webware the other week and it looks very promising. It is very easy 
to e.g. develop a template as a servlet and let other servlets inherit the 
layout and functionality which makes Webware extremly powerful and easy to 
use.

There is also support for PSP which I haven't tried (embedded Python in 
HTML).

Another powerful (not yet final though) module is Middlekit which e.g. makes 
it easy to deploy a servlet that acts with some glueware to e.g. MySQL.

I am going to follow the work of Webware since I like the freedom it offers. 
It seems more convenient to use comapred to Zope.

Regards,
Magnus



From cbrown at metservice.com  Tue Apr 23 20:57:02 2002
From: cbrown at metservice.com (Colin Brown)
Date: Wed, 24 Apr 2002 12:57:02 +1200
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net>
Message-ID: <3cc6036f$1@news.nz.asiaonline.net>

Thanks guys

I did not think to look up this  "documented feature", sorry.

It still looks like a cop-out to me and can join my "ugly list" along with
leading-zero octal numbers.

Colin Brown
PyNZ


"Colin Brown" <cbrown at metservice.com> wrote in message
news:3cc5d5c8$1 at news.nz.asiaonline.net...
> I have entered the following at the command line in Python 2.1.1 on Linux
> and in Python 2.2 on NT and get the same results:
>
> >>> r'\'
>   File "<string>", line 1
>      r'\'
>         ^
> SyntaxError: invalid token
> >>>
>
> It appears as if the r option for strings does not handle a backslash as
the
> last string character properly.
>
> Colin Brown
> PyNZ
>
>
>




From angus.duncan at btclick.com  Mon Apr  8 12:57:47 2002
From: angus.duncan at btclick.com (Angus Duncan)
Date: Mon, 8 Apr 2002 17:57:47 +0100
Subject: Norton problem
Message-ID: <017a01c1df1e$9ff793a0$ecb77bd5@fkvp60j>

My computer keeps crashing when I download new Norton Anti Virus updates. I undersstand from them that this is due to Python 2.0.
I am not familiar with the program or what I use it for. where would I find it and how can I Take it off when doing new updates.

I attach the message fom Norton.

Error: "An exception 0E has occurred at 0028:C02AD9D0... It may be possible to continue normally" when using Python 2.0

Situation:
You are using Norton AntiVirus 2000/2001(NAV) under Windows 98 SE and have Auto-Protect enabled. You have Python 2.0 beta 2 from PythonLabs installed. While using certain features in Python, you see a blue screen with the error message "An exception 0E has occurred at 0028:C02AD9D0 in VxD IFSMGR(04) + 000050E4. This was called from 0028:C000B511 in VxD VMM(01) + 0000A511. It may be possible to continue normally."

Solution:
The solution depends on your version of NAV:

NAV 2001
An update is available that fixes this problem. Please run LiveUpdate to download the program update.

NAV 2000
PythonLabs is aware of this problem. The following information is from the PythonLabs Web page: 

  "Incompatibility warning: Norton Antivirus 2000 can cause blue screen crashes on Windows 98 when a function in the os.popen*() family is invoked. To prevent this problem, disable Norton Antivirus when using Python. (Confirmed on Windows 98 Second Edition with Norton Antivirus version 6.10.20. The same Norton Antivirus version doesn't have this problem on Windows 2000. Norton Antivirus version 5 on Windows 98SE doesn't have this problem either.)"

For additional information, go to the following Internet address:

http://www.pythonlabs.com/products/python2.0/download_python2.0b2.html

To use the Python program, you can disable Auto-Protect while using Python.

NOTE: If you disable Auto-Protect, then you are no longer protected against viruses when you open or save files.

To disable Auto-Protect, right-click the Norton AntiVirus icon in the System Tray, and then click Disable Auto-Protect. Be sure to enable it again as soon as you finish using Python.

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

Thank you for your help
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020408/e79066be/attachment.html>

From kirschh at lionbioscience.com  Mon Apr 22 04:59:35 2002
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 22 Apr 2002 10:59:35 +0200
Subject: handling exception of python callback called from C function
Message-ID: <yv2pu0s6rd4.fsf@lionsp093.lion-ag.de>

I have a C library which runs callbacks into user-supplied code. I
made a python extension to interface this library in a way that such a
callback ends up again in the extension which then executes python
scripts.

The following can happen:

1) the script code called via the callback throws an exception,

2) the extension catches the exception and translates it into an error
code for the calling C function,

3) the C function recognizes the error code and returns with its own
error code to the calling interfacing python extension,

4) the interfacing python extension recognizes the error code. It can
even identify the fact that the cause of the error was a
python-exception.

QUESTION:

How can step (4) initialize/augment the error traceback with the
traceback information generated in step (1)?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh bei lionbioscience punkt com
       *** Please do not send me copies of your posts. ***


From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 13:41:49 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 17:41:49 +0000 (UTC)
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it>
Message-ID: <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com>

Alex Martelli <aleax at aleax.it> wrote:

>Personally, I'd rather see curry built-in -- it IS easy to write curry
>in Python 2.2, but it's such a wonderful idiom (and would replace so
>many lambda's;-) that I'd love to see it officially blessed.  Not that
>I expect it to be (and nor will my heart be broken because of this:-).

Do you have in mind a special syntax for curry?  

The implicit syntax (omitting arguments) will interfere with default
arguments and error checking for argument numbers, etc.

What do you think about the following explicit syntax?

def f(a, b, c): return a+b+c
f1 = curry(f, a, b)
f1(c)
map(curry(f, x, y), a)  # map(lambda z:f(x, y, z), a)

Huaiyu


From jeff at ccvcorp.com  Thu Apr 25 13:59:59 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 25 Apr 2002 10:59:59 -0700
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org> <mailman.1019682973.29139.python-list@python.org> <871yd4odap.fsf@pvv.org> <MPG.1730fb06f5bab3e99896c8@news.nwlink.com> <sepu0okxo8.fsf@pc970145.GermanLloyd.de>
Message-ID: <MPG.1731e3fd71e8350b9896c9@news.nwlink.com>

In article <sepu0okxo8.fsf at pc970145.GermanLloyd.de>, 
hoel at germanlloyd.org says...
> Jeff Shannon <jeff at ccvcorp.com> writes:
> 
> > ...The only effective difference between 1 and True,=20
> > and 0 and False, is how it's written, but they add, subtract,=20
> > multiply, and hash the same.
> 
> But there is a difference:
> 
> >python
> Python 2.2.1 (#1, Apr 10 2002, 17:11:02)=20
> [GCC 3.0.4] on sunos5
> Type "help", "copyright", "credits" or "license" for more information.
> >>> a =3D 1
> >>> b =3D 1
> >>> a is b
> 1
> >>> a =3D True
> >>> a is b
> 0
> >>>=20

Yes, True and False are singleton objects that have a different 
identity than 1 and 0.  The 'is' operator tests object identity.  

>>> a = []
>>> b = []
>>> a is b
0
>>> 

This doesn't change the fact that a and b are equivalent, and I 
would still say that there is no effective difference between a 
and b, here.  Just different object IDs.

--

Jeff Shannon
Technician/Programmer
Credit International


From aleax at aleax.it  Wed Apr 17 11:20:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 15:20:37 GMT
Subject: Counterintuitive Python behavior
References: <mailman.1019052695.4161.python-list@python.org> <lksn5uqtsv.fsf@pc150.maths.bris.ac.uk> <lkofgiqteo.fsf@pc150.maths.bris.ac.uk>
Message-ID: <9pgv8.30868$vF6.957894@news2.tin.it>

Michael Hudson wrote:

> Michael Hudson <mwh at python.net> writes:
> 
>> > The next time a co-worker asks a assignment/binding/references
>> > question I'm going to refer them to this message.
>> 
>> I was hoping more people than the OP would find it useful.  I'll
>> probably knock up a version for my web pages (when the starship
>> reappears...).
> 
> I should say that Alex's answer is well worth referring to too.  I
> wonder if he would mind me incorporating it into aforementioned
> webpage...

Sure!  Go right ahead.


Alex



From gh_pythonlist at gmx.de  Tue Apr  2 20:22:24 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 3 Apr 2002 03:22:24 +0200
Subject: relational database?
In-Reply-To: <3CA9EF24.2B9A0C27@lemburg.com>
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <3CA9EF24.2B9A0C27@lemburg.com>
Message-ID: <20020403012223.GB562@lilith.my-fqdn.de>

* M.-A. Lemburg <mal at lemburg.com> [2002-04-02 19:49 +0200]:
> SAP DB is fast, full-featured, open-source and very stable.

Open-source it sure is, but that doesn't mean that mere mortals can
compile it or even port it to not currently supported platforms like
FreeBSD.

> Note that Unicode support in mxODBC is available both as emulation
> (mxODBC does the conversion to e.g.  UTF-8) and using native support
> (the ODBC driver has to support Unicode).

Very interesting. This Unicode stuff isn't specified in a new draft of
the DB-API, is it?

I'm planning to add Unicode (and perhaps non-default 8-bit charset)
support to pyPgSQL. In my first tries, I found that there are really two
different problems:

- getting Unicode strings as parameters of execute into the database,
  like:

  cursor.execute("insert into test(col) values('%s'" % \
  	u"someUnicodeString")

  This doesn't look difficult.

- getting Unicode out of the database: My first idea was to add a flag
  to the connection constructor that indicates wether to returns strings
  as Python strings or Unicode strings. Any better ideas?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 7.2 ?C      Wind: 2.8 m/s



From tchur at optushome.com.au  Thu Apr 25 16:46:02 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Fri, 26 Apr 2002 06:46:02 +1000
Subject: Python v. Euphoria
References: <394fcus92g8rt0e2ektsd36s9qu7cq2cij@4ax.com>
Message-ID: <3CC86B0A.3A6E97C9@optushome.com.au>

Stom Fairecroire wrote:
> 
> Hi, All
> Many, many years ago I did a lot of programming in HPL (Hewlet Packard
> Language) then did lots using PICK basic. Now I want to start some
> very low key programming again - just for my own amusement. Anyone
> have any views on the pros and cons  of Python v. Euphoria ? Which
> would be the better option?

Stom,

Definitely Python, for all the reasons given by others, but also because
the native data structures which Python provides will seem very familiar
to anyone who has used PICK. In fact, if you nest some Python lists
inside
a Python dictionary and store all that in a ZODB (Zope Object DataBase),
you have a pretty good facsimile of the PICK filesystem, surrounded by a 
much more powerful language than PICK Basic ever was, and even easier to 
learn. In fact, you could write a class around the above to make it
behave
exactly like the PICK filesystem if you felt so inclined. 

Tim C




From jriveramerla at yahoo.com  Wed Apr  3 18:26:46 2002
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 3 Apr 2002 15:26:46 -0800
Subject: wxPython Book ?
Message-ID: <11e94203.0204031526.f11bfe6@posting.google.com>

Hi...
Is there a wxPython book available? that gives a depth tutorial?

thanks...


From lac at strakt.com  Wed Apr  3 11:45:50 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 03 Apr 2002 18:45:50 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from "Mark McEahern" <marklists@mceahern.com> 
   of "Wed, 03 Apr 2002 07:49:17 MDT." <NCBBLFCOHHDIKCAFGCFBMEPDKIAA.marklists@mceahern.com> 
References: <NCBBLFCOHHDIKCAFGCFBMEPDKIAA.marklists@mceahern.com> 
Message-ID: <200204031645.g33Gjorq028972@ratthing-b246.strakt.com>

> ["Mark McEahern]
> 
> I thought she was trying to make the stronger point that even without PEP
> 285, any function that is designed to return true/false (1/0) is badly
> designed.  Moreover, PEP 285 would only encourage people to write such
> functions.
> 
> Does that apply to, say, isinstance()?
> 
> // mark

That is more strongly worded then I would make it; I only think that
deciding that a function should be compelled to return only one of
two values requires long and careful deliberation.  This is because it
is likely that some day you will wish that you could provide more
information than this allows.  isinstance is a good example
of this. I would be pleased if isinstance returned more useful info
than just yes/no, such as : 0, no inheritance; 1: direct instance of the 
requested class; 2: instance of a subclass of the requested class.
90% of uses would still be yes/no, but this would be fine, and you would
get the extra information when you needed it.

Laura Creighton




From sholden at holdenweb.com  Tue Apr 30 09:24:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 09:24:56 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2><3CCD9649.98AED86F@cascade-sys.com><20020429130136.5b0c34f7.logiplexsoftware@earthlink.net><3CCDB559.A1BF0E47@cascade-sys.com> <mailman.1020116996.21109.python-list@python.org>
Message-ID: <fZwz8.13890$xy4.10493@atlpnn01.usenetserver.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1020116996.21109.python-list at python.org...
> On Mon, 29 Apr 2002 14:04:25 -0700
> James J. Besemer wrote:
>
> >
> > Cliff Wells wrote:
> >
> > > Well that's a stupid statement - *your* head is in the clouds.
> >
> > Oh yeah?  Fuck you.
>
> Heh.
>

This little exchange is about as perfect a demonstration of Cliff's point as
could be made. Now, if anyone but me is still reading this thread, nobody
knows who was joking and who wasn't in the first place. The additional loss
of context certainly doesn't help, either.

sense-of-humor-failure-is-embarrassing-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From peter at engcorp.com  Mon Apr  1 01:02:58 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 01 Apr 2002 01:02:58 -0500
Subject: OT: Daylight savings time (was Re: Subclass sadness (was Re:  
 [Python-Dev] PEP 285: Adding a bool type))
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <a85od6$9i$1@pani <yu99hemx1e8r.fsf@europa.research.att.com> <mailman.1017555459.10956.python-list@python.org> <3CA73486.72B4059F@engcorp.com> <c76ff6fc.0203311421.2a67402b@posting.google.com>
Message-ID: <3CA7F812.49EEB7CD@engcorp.com>

John Machin wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote in message news:<3CA73486.72B4059F at engcorp.com>...
> > Richard Jones wrote:
> > >
> > > Hrm, just following this thought here, possibly for no good reason at all
> > > (hey, it's Sunday, and daylight savings just ended so I've got a whole _hour_
> > > extra to waste today :)...
> >
> > In which country?  Good thing it's not a work day, since I took your
> > word for it last night...  or was that another advance April Fools joke?
> >
> 
> "daylight savings just *ended*" (my emphasis) -> try thinking "other
> side of the equator"

Well, that only works if one understands daylight savings time
and knows when it starts and when it ends. :-(  This is one of 
those things that I've just never grokked.

My sister (on whom I count for complicated technical information
like this :-)  informs me that here (Canada) we'll be switching
over (don't know if that means starting or ending) _next_
weekend, and we _lose_ an hour.  Bummer...

-Peter


From johnroth at ameritech.net  Tue Apr 16 17:17:56 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 16 Apr 2002 17:17:56 -0400
Subject: Backing up mail in Outlook Express using Python
References: <mailman.1018975781.22375.python-list@python.org>
Message-ID: <ubp58ek7j07d77@news.supernews.com>

"Noah" <noah at noah.org> wrote in message
news:mailman.1018975781.22375.python-list at python.org...
>
> What?! My version of Outlook Express stores them in binary files.
> What version are you using?
>
> Noah

I think he's talking about a different problem. I suspect he only
wants to back them up so he can restore them. That might not
be too difficult since there is one file per mailbox.

On the other hand, you (and I) want to break them out into
separate mail messages so we can do something reasonable
with them. I'm going to take a look at the utility you (or someone)
suggested in one of the other posts.

John Roth
>
> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Thomas Weholt
> Sent: Tuesday, April 16, 2002 7:07 AM
> To: python-list at python.org
> Subject: Re: Backing up mail in Outlook Express using Python
>
>
> Well, by searching the net a bit more I've found some pointers and as
soon
> as I found out where the mail was actually stored and that it's stored
in
> plain files, backing them up was easy.
>
> Thanks anyway :-)
>
> Best regards,
> Thomas Weholt
>
>
> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:JCUu8.23257$Ji7.8640 at atlpnn01.usenetserver.com...
> > "Thomas Weholt" <thomas at gatsoft.no> wrote in message
> > news:cGSu8.101$1kc.179320832 at news.telia.no...
> > > I need to backup all my mail stored in Outlook Express. Can I use
Python
> > to
> > > do this?
> > >
> > No reason why not, though I'm not sure why you'd want to, since the
> message
> > are just stored in files.
> >
> > You can find out where they are from the Maintenance tab on Tools |
> Options
> > (click the "Store Folder" button).
> >
> > I did adapt some code from Delphi to read the mailbox files, and can
let
> you
> > have that if it would help, but sounds to me like a file copy is all
> you're
> > looking for.
> >
> > regards
> >  Steve
> >
> >
> >
> >
>
>
>
>
>




From roy at panix.com  Sat Apr  6 18:21:58 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 06 Apr 2002 18:21:58 -0500
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org>
Message-ID: <roy-AF1B7C.18215806042002@news1.panix.com>

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote: 
> I'll see your list comprehensions and raise you a set of string methods.

To be perfectly honest, I just don't comprehend list comprehensions, but if 
there is a single post-1.5.2 feature that I wouldn't willingly give up for 
anything, it's string methods.


From cliechti at gmx.net  Sun Apr 14 13:04:28 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 14 Apr 2002 19:04:28 +0200
Subject: Lambda  added to release 2.2 ?
References: <3CB9ABAB.E37E685B@club-internet.fr>
Message-ID: <Xns91F0C23FC37DBcliechtigmxnet@62.2.16.82>

Fabien H?non <ffjhenon at club-internet.fr> wrote in 
news:3CB9ABAB.E37E685B at club-internet.fr:
> When I start my python script, here is what I get on the console from
> Linux (Python 2.1) :
> a26.py:51: SyntaxWarning: local name 'self' in '__init__' shadows use of
> 'self' as global in nested scope 'lambda'
>   def __init__(self, master):
...
> I don't have any problem under W2000 ( Python 2.2)
> Is lambda function a new add-on to 2.2 ? 

no but, nested scopes are now enabled by default wheras in
2.1 you need to write
>>> from __future__ imort nested_scope
and the warning will go away.

chris

-- 
Chris <cliechti at gmx.net>



From mark.charsley at REMOVE_THIS.radioscape.com  Mon Apr 15 06:30:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Mon, 15 Apr 2002 11:30 +0100 (BST)
Subject: Beginner's performance problem
References: <a95154$ans$0@216.39.172.122>
Message-ID: <memo.20020415113054.1532B@a.a>

In article <a95154$ans$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
wrote:

> >Aha! I'd put off reading up on the regular expression stuff until 
> later as 
> >I thought it just did searches, as opposed to also doing 
> in-place 
> >replaces. 

> Best reread Alex's post carefully, until you understand the difference 
> from the above ;-)

Sorry s/in-place replaces/replaces within the whole string at once/

Of course if I misunderstood something else, let me know...

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From tim.one at comcast.net  Wed Apr 17 01:54:56 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 17 Apr 2002 01:54:56 -0400
Subject: list_a-list_b
In-Reply-To: <eppstein-6EBA52.22290716042002@news.service.uci.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOHOOAA.tim.one@comcast.net>

[David Eppstein]
> ...
> Better is to convert b to a dictionary
> (where are dict comprehensions when you need them):
>
>     dict_b = {}
>     for x in list_b: dict_b[x] = 1

Looks clear enough to me, although

    for x in list_b:
        dict_b[x] = 1

is clearer.  If you like, you can do this:

    dict([(x, 1) for x in list_b])

or this:

    dict(zip(list_b, [1]*len(list_b)))

or even:

    dict(zip(list_b, list_b))

since you don't really care what the dict's values are, just the keys.  I
draw the line before:

    dict(zip(*[list_b]*2))

though.  Contrary to popular belief, there's no Fabulous Prize awaiting
those who minimize either vertical or horizontal space <wink>.





From paul at boddie.net  Tue Apr 30 11:57:41 2002
From: paul at boddie.net (Paul Boddie)
Date: 30 Apr 2002 08:57:41 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de>
Message-ID: <23891c90.0204300757.17985ba3@posting.google.com>

Bernhard Herzog <bh at intevation.de> wrote in message news:<6qvga9bizr.fsf at abnoba.intevation.de>...
> 
> Why should chr be an int method in the first place? It's not tied to
> ints all that much[1]. IMO it sits somewhere between ints (including
> int-like objects !) and strings so it shouldn't be a method of either
> class. Leaving it a function is perfectly fine, therefore, although
> putting it into the string module might be a good idea. The same would
> apply to ord, IMO.

Indeed. The mistake the Java API designers made with their "standard"
classes was to invent shedloads of different methods to convert one
type/class to another, giving lots of different possibilities and
bloating the API to new, undiscovered heights/depths. For anyone in
any doubt about this, take a look at java.lang.String/Long/Integer
etc. to see what I mean.

But then I still have problems remembering if one should use...

  "/".split("a/b/c")

or

  "a/b/c".split("/")

...so perhaps I'm not sufficiently "into" the practice of making
methods out of perfectly decent enough functions to be able to
comment.

Paul


From seth at jtan.com  Thu Apr 25 12:54:03 2002
From: seth at jtan.com (Seth Delackner)
Date: Thu, 25 Apr 2002 12:54:03 -0400
Subject: urllib.urlopen("file://...") fails
In-Reply-To: <Pine.LNX.4.44L.0204252001100.2654-100000@ods.pp.ru> ("Denis S.
 Otkidach"'s message of "Thu, 25 Apr 2002 20:02:07 +0400 (MSD)")
References: <Pine.LNX.4.44L.0204252001100.2654-100000@ods.pp.ru>
Message-ID: <m2elh3ius4.fsf@jtan.com>

"Denis S. Otkidach" <ods at fep.ru> writes:

> On Thu, 25 Apr 2002, Seth Delackner wrote:
>
> SD> Fernando P?rez <fperez528 at yahoo.com> writes:
> SD>
> SD> > Seth Delackner wrote:
> SD> >>>>> u = urllib.urlopen("file:///t:/tmp/package.html")
> SD> >                                 ^
> SD> >
> SD> > One / too many? You're directing it to \t:\.... which it
> SD> obviously
> SD> > can't find.
> SD> >
> SD> > I don't run windows so I can't test so I may be wrong. It
> SD> just looks
> SD> > like that is your problem.
> SD>
> SD> You would think, right?  But I just tried
> SD> 'file://t:/tmp/package.html'
>
> I guess it should be something like 'file://t|/tmp/package.html'.

Sorry, that also dies (although IE also accepts that format).  It is
erroneously identifying this as an ftp url.





From tismer at tismer.com  Fri Apr 26 06:25:05 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 26 Apr 2002 12:25:05 +0200
Subject: Stackless goes Limbo
References: <mailman.1019662503.1913.python-list@python.org> <uchetjbtqegmb8@corp.supernews.com>
Message-ID: <3CC92B01.50808@tismer.com>

Andrew Henshaw wrote:
...

> While you're at it, you might want to look at Occam.  In my opinion, no 
> language handles channels as nicely as Occam.  [Sigh] PAR and ALT would 
> look so nice as Python keywords...

I took Limbo since it seemed to be one of the newest
parallel languages, and it was quite much simpler than
Alef. Sure, I will also take Occam into
account. At the moment, I'm trying to find the simplest
possible but yet most powerful set of primitives,
and I use the other languages to learn from and to lean
on, when I have to find names of functions.

Things like PAR and ALT could in fact convince me
to add new keywords to Stackless Python...

> Good luck!

thanks a lot - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From boris.boutillier at lip6.fr  Wed Apr 24 12:26:30 2002
From: boris.boutillier at lip6.fr (Boris Boutillier)
Date: Wed, 24 Apr 2002 18:26:30 +0200
Subject: Subclass of int and initialisation with not integer values
References: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr>
Message-ID: <pan.2002.04.24.18.26.26.813701.17049@lip6.fr>

On Wed, 24 Apr 2002 17:09:53 +0200, Boris Boutillier wrote:

> Hi all,
> 
> I've got a little problem with subclass of int.
> 
> Here is a sample of what I want to do :
> 
> class toto(int):
> 	true  = 'O'
> 	false = 'N'
> 	def __init__(self,str):
> 		self.str = str
> 		if str=='O':
> 			int.__init__(self,1)
> 		elif str == 'N':
> 			int.__init__(self,0)
> 		else:
> 			raise ValueError, 'Not valid value %s'%(str)
> 	def __str__(self):
> 		return self.str
>>>> x = toto('N')
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ValueError: invalid literal for int(): N
>>>> x = toto(1)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "<stdin>", line 11, in __init__
> ValueError: Not valid value 1
> 
> In some way it seems that the int initialisation is called before
> __init__ I want toto to be a subclass of int because I need toto objects
> to be used as index for lists.
> 
> Do you know some way to solve the problem ? Here true and false are 'O'
> and 'N' for the example but in my example I use a metaclass to create
> numerous classes of this kind, with different true letter and false
> letter.
> 
> Boris
> .

In fact I found the answer, it is that __new__ is called before __init__
and as a true int is return by new, __init__ is called too late.
So i solved my problem by redefining __new__ instead of __init__.

Sorry for this useless post.

Boris Boutillier


From sholden at holdenweb.com  Thu Apr 18 15:01:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 18 Apr 2002 15:01:18 -0400
Subject: ODBC and Excel
References: <3cbed86d@news.barak.net.il>
Message-ID: <5MEv8.19497$eS3.5069@atlpnn01.usenetserver.com>

"Clive Bluston" <clive at emultek.co.il> wrote in message
news:3cbed86d at news.barak.net.il...
> I am new to Python. I am trying to read from an Excel file and cannot even
> get connected. The following code gives me:
> dbi.internal-error: [Microsoft][ODBC Driver Manager] Invalid string or
> buffer length in LOGIN
> Has anyone any idea what I may be doing wrong?
>
> import dbi, odbc;
>
>         s = odbc.odbc('ODBC;Driver={Microsoft Excel Driver
> (*.xls)};DBQ=d:\\clive\\python\\DEFAULT.XLS;DriverID=790')
>  Timeout=5;')
>
Have you tried establishing a System DSN and just using that as the argument
to odbc.odbc?

regards
 Steve






From rjones at ekit-inc.com  Wed Apr  3 01:47:47 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 16:47:47 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <7x4rituw82.fsf@ruckus.brouhaha.com>
References: <mailman.1017794173.28885.python-list@python.org> <B8D00218.20D57%goodger@users.sourceforge.net> <7x4rituw82.fsf@ruckus.brouhaha.com>
Message-ID: <200204030647.GAA27785@crown.off.ekorp.com>

On Wed, 3 Apr 2002 16:34, Paul Rubin wrote:
> I hope PEP's are not marketing tools.  Marketing tools are material
> written by vendors to get customers to buy products because they are
> the vendors products, rather than because they are what the customer
> needs.

Wow, that's a very narrow definition of marketing. I know people who are in 
martketing, and they'd go green hearing such a definition :)

I believe you are confusing marketing with advertising.


>  If a PEP advocates a particular approach over other
> approaches, it should come from having carefully compared the
> approaches and finding that the PEP's approach is actually the best.
> This PEP doesn't give the impression of careful comparison.

It points to the documents that David and others have written that do. Did 
you read those? Ooh. Dejavu. I've said that before, I swear...

> David Goodger wrote:
> > The PEP briefly examines alternatives proposed in the past, on the
> > Doc-SIG, in terms of "the generally accepted goals for a docstring
> > format".  I think you're reading too much into the prose as it stands.
>
> Since the PEP is being presented to the whole community and not just
> to the Doc-SIG (I'd never heard of the Doc-SIG til just now), it
> should make its case in detail rather than just giving the conclusion.

The PEP is already very long. Did you want David to include _all_ the 
documentation from the ReST website in the PEP to back it up?


> That includes the case for having structured doc strings at all--do
> you have formatting tools that read these doc strings and produce
> printed output like POD?

I believe the docstring extraction is performed by pydoc and is a little 
manual at the moment, but the upshot is yes, he does (printing of HTML is 
valid at this early stage of development).


> I've written some pretty big Texinfo documents and find Texinfo to be
> readable enough in source form, though it looks better after
> formatting.  reStructuredText looks pretty simliar to the output of a
> Texinfo formatter--maybe Texinfo could be adapted to produce
> reStructuredText instead of Info files.

But the whole point of this is that the _source_ of the documentation is the 
reStructuredText-formatted text, not some other more arcane format that must 
be processed to produce reStructuredText files.


    Richard



From geoff at gerrietts.net  Mon Apr  1 15:56:04 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 1 Apr 2002 12:56:04 -0800
Subject: duelling pythons
In-Reply-To: <1eed326b.0203312131.2ef13b62@posting.google.com>
References: <1eed326b.0203312131.2ef13b62@posting.google.com>
Message-ID: <20020401205604.GB6604@isis.gerrietts.net>

Quoting Gil (gjlap at yahoo.com):
> I think my original install of Python1.5 with RedHat 7.1 is not
> playing nice with a later install of Python2.2. How do I troubleshoot
> this?

Having done this myself a few times, I can give you some pieces of
advice.

By default, if you're compiling python, you're going to install
/usr/local/bin/python. You can take special measures to do otherwise,
but by default, that's where it ends up.

This puts you in the circumstance of having /usr/local/bin/python and
/usr/bin/python, with /usr/local/lib/python2.2 and /usr/lib/python1.5
supporting these two installs.

You may or may not also have /usr/lib/python2.1 and /usr/bin/python2.

Confirm this. Fire them up, one at a time:

/usr/bin/python
/usr/local/bin/python

Verify that the versions are stated correctly.

Have a look at what else you have installed:

/usr/bin/python*
/usr/local/bin/python*

Most of these are just links to one of the core executables, but some
might be different. My RedHat 7.2 system has Python 1.5, Python 2.1,
and Python 2.2 simultaneously installed. I generally don't use
python2.1, but it does exist (which makes cross-version testing
easier).

Lots of RedHat's software relies on the contents of
/usr/lib/python1.5, which is why it sucks to use RedHat for coding
python these days. (On the other hand, they've got lots of extension
modules for 1.5 pre-installed, which is why it's kinda cool to use
RedHat for coding python these days.)

Because RedHat's software relies on the 1.5.2 binary, there's gonna be
some pain if you want to use 2.2. How you take that pain is up to you.
There are two basic options:

DO REDHAT'S DEPENDENCY MANAGEMENT FOR THEM

    The first solution is to just use "python2" whenever you want to
    use the new features or write forward-portable code. To accomplish
    this:
    - Make sure that /usr/bin/python1.5 points to the 1.5.2 binary.
    - Make sure that /usr/bin/python points to the version you want to
      use.
    - Make sure that all RedHat's scripts are fixed to point to
      python1.5.

    It's the last step that's pretty troublesome. You'll need to go
    look at /usr/share. There's a lot in this subdirectory, and a
    striking percentage of it uses python. You might try "grep -rl
    python ." to extract a list of all the files who look at python.
    Not all of these need to be changed, but it's hard to pick out
    which do and which don't without knowing how the individual
    systems involed operate.

    Then check /usr/bin and /usr/sbin the same way -- I think there
    are some shell scripts in those directories that invoke python.

    This is the way RedHat should have shipped the system, if they
    were going to be depending on an old version of the python
    interpreter, but they didn't.

SUCK IT UP

    The second solution is to make all your scripts and all your
    installs of recent tools use python2 instead of python. For
    anything that uses distutils, I've found this option works pretty
    well -- you call "python2 setup.py" and the package configures
    itself and builds itself appropriately.

    This can be a little more troublesome with packages or scripts
    that assume /usr/bin/python will be the one you want, or scripts
    that assume /usr/bin/env python will work properly.

    This is a hassle, but it's an ongoing hassle that is unlikely to
    consume more than 30 mins of work in a sitting, whereas the
    previous task is likely to take many, many hours in up-front time.

There's also one other option, but I'm not sure how feasible it is. It
might be the preferred solution, if you can manage it. It might not be
possible, though.


DO REDHAT'S SYSTEM UPDATING FOR THEM

    This path, entirely hypothetical at this point, would involve
    downloading and building all the extension modules presently
    included in /usr/lib/python1.5 for your new version of the
    interpreter.

    It may also involve debugging some of the old scripts for the new
    environment -- a lot has changed between 1.5.2 and 2.2 that could
    have backwards-compatibility ramifications for the scripts. But it
    may not.

    In some ways, this is the ideal approach, but I don't know how
    feasible it is.


It's funny that RedHat is such a big believer in Python that they've
written huge swaths of their distribution-management system in python,
yet at the same time they've made it very challenging for Python
programmers to use their distribution. There's something of an
inconsistency there. :)

Good luck,
--G.

-- 
Geoff Gerrietts <geoff at gerrietts dot net>     http://www.gerrietts.net/
 "Politics, as a practice, whatever its professions, has always been the 
          systematic organization of hatreds." --Henry Adams



From aleax at aleax.it  Wed Apr 17 16:55:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 20:55:01 GMT
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <mailman.1019073733.27071.python-list@python.org>
Message-ID: <Filv8.32338$vF6.998647@news2.tin.it>

Ian Bicking wrote:
        ...
> But that's weird...
> 
> q1 += q3
> is not equivalent to:
> q1 = q1 + q3

If it were, there would be little reason for it to exist -- saving
a couple of keystrokes would be a silly reason to introduce a
bunch of new operators.  q1 = (whatever) *MUST* rebind q1,
this is fundamental Python semantics.  q1 += (whatever) MAY
rebind q1, when q1 was referring to an immutable object (no
inplace operations then), but it was introduced mostly to ease
*inplace* operations (on, obviously, *mutable* objects only) -- it
gains extra power by working polymorphically on both mutable
and immutable objects, but of course it cannot have _exactly_
identical semantics in the two cases (it would be useless if it
didn't take the opportunity to mutate a mutable, impossible if
it DID mutate immutables:-).


Alex



From tdelaney at avaya.com  Wed Apr 17 22:02:31 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 18 Apr 2002 12:02:31 +1000
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A271@natasha.auslabs.avaya.com>

> From: philh at comuno.freeserve.co.uk 
> 
> IOW, some code did break? And how much wore will break when people 
> try to serialize code on a post-PEP285 python and read it on a 
> pre-PEP285?

None at all. Guido has already stated that he intends to take Paul Rubin's
suggestion for pickling boolean values as I00 and I01, which are valid
integral values, but never generated by pickle. These will be treated
specially to produce False and True on 2.3+, but will produce 0 and 1 on
2.2-. However, I'm not sure if there was a determination on binary pickles
...

Tim Delaney




From claird at starbase.neosoft.com  Thu Apr 11 17:03:37 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 11 Apr 2002 16:03:37 -0500
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com>
Message-ID: <7E283A6EA4F0DBD6.B68D98BE8C574F5D.CE8B287F1A68C959@lp.airnews.net>

In article <7226483f.0204101216.2c8fcda4 at posting.google.com>,
M. J. Mix <marilyn.mix at omitron.com> wrote:
>Is anyone using Python as a spacecraft ground system scripting
>language, or onboard (embedded) language?
>
>Pros and cons when compared with TCL or JAVAScript.  Just 
>some real life, honest opinions.
>
>The usual issues: reliability and maintenance effort in applications. 
>Also, use as a scripting language by non-programmers. (I saw the post
>on indenting, so I'm concerned about that.)

Others have already answered most of this more than
adequately.  I'll reinforce a few points:

Bluntly, JavaScript's a bad idea.  Don't do it.
Details available, if necessary.

I believe that Python is in use "as a spacecraft
ground system scripting language".  I'm less certain
whether that's public information.

Indentation is a non-issue.  While your question is
a reasonable one, the best research available shows
that non-programmers do *better* with Python's
significant white space than with comparable {} sorts
of grouping.  It's an advantage, not a disadvantage.
-- 

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


From tony at cases.demon.co.uk  Thu Apr 18 12:35:41 2002
From: tony at cases.demon.co.uk (Tony Case)
Date: Thu, 18 Apr 2002 16:35:41 +0000
Subject: help splitting strings
Message-ID: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com>

Hi all
	I'm pretty new to Python which probably explains my problems,
I'm trying to read data from a number of comma separated variable
files.  These are fairly large, 5000 lines plus.
	I'm reading the files and using split to separate the
individual values, some numeric some string.  The problem is that a
number of the lines have commas and double quotes embeded in the
strings so there are more elements in the list returned from split
than there should be, in one file there are 400 + lines like this so
it's not easy to edit the original file.
	Is there a programatic way of dealing with this?
The input files are in the form :-
int,int,int,"string","string",int,"string"
etc.
I tried using the re.findall and re.split but didn't get very far.
	TIA	T Case


From paddy3118 at tiscali.co.uk  Thu Apr 18 15:11:26 2002
From: paddy3118 at tiscali.co.uk (Donald McCarthy)
Date: Thu, 18 Apr 2002 20:11:26 +0100
Subject: Changing the repr() of a class?
Message-ID: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com>

I would like to alter the way python displays what a class is. I can do this
for an instance of a class using __repr__ but I don't know how to do this
for a class.

Example:
>>> class C(object):
...  def __repr__(self): return 'repr C'
...  def __str__(self): return 'str C'
...
>>> C
<class '__main__.C'>
>>> repr(C)
"<class '__main__.C'>"
>>> str(C)
"<class '__main__.C'>"
>>> i=C()
>>> i
repr C
>>> repr(i)
'repr C'
>>> str(i)
'str C'
>>>

What I want to do is change what is printed when I just type C so it returns
something more meaningful to me whilst creating my program (and learning
python).

Oh - and I'm using python version 2.2

Thanks, Paddy.





From calves at coelce.com.br  Thu Apr 11 07:43:15 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Thu, 11 Apr 2002 08:43:15 -0300
Subject: RES: initializing list with number of elements given by a variabl
	e
Message-ID: <29A97D00F387D411AC7900902770E14804C1BC3B@lcoeexc01.coelce.net>

It depends on how you should/want set the values of elements in the list.
For example, if you already know what/how elements to place in the list,you
could do:

lis=[None]*6 #Initialize the list with 6 None elements.
for i in range(6):
	lis[i]=i*i #or do something else

Now, lis is one list with 6 elements given by i*i.
If you need to enter values in run time, you could make it like this:

lis=[None]*6
for i in range(6):
	lis[i]=raw_input('Elemento %d:'%i)


> -----Mensagem original-----
> De: Oliver Pust [mailto:oliver.pust at web.de]
> Enviada em: quinta-feira, 11 de abril de 2002 06:23
> Para: python-list at python.org
> Assunto: initializing list with number of elements given by a variable
> 
> 
> Hi,
> 
> I need to create a list which number of elements is given by 
> an integer 
> variable.
> 
> So something like this:
> 
>  >>> length_of_list = 6
>  >>> list = ????????? # what to place here, is the question
>  >>> print list
>  >>> [0,0,0,0,0,0]    # the resulting list should look like this
> 
> Can anybody help me here?
> 
> Thanks,
> 	Oliver
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020411/a9fc753d/attachment.html>

From giovanni.laudicina at libero.it  Tue Apr  2 13:12:20 2002
From: giovanni.laudicina at libero.it (Giovanni Laudicina)
Date: Tue, 02 Apr 2002 18:12:20 GMT
Subject: calling dll's functions with python
Message-ID: <8wmq8.52601$S52.1805120@news2.tin.it>

I'm a beginner with python and i did'n find this info on the FAQ and
tutorials:
How can I call a function included in a dll (I have only the .dll file, not
the source code) ?
Thanks to everyone,
Giovanni Laudicina




From boud at valdyas.org  Mon Apr 15 12:35:05 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 15 Apr 2002 18:35:05 +0200
Subject: Looking for GUI Form Builder for Python
References: <mailman.1018882348.11577.python-list@python.org>
Message-ID: <3cbb03c6$0$71662$e4fe514c@dreader4.news.xs4all.nl>

> 
> -----Original Message-----
> From: Calvin [mailto:hoiwai930 at hotmail.com]
> Sent: Sunday, April 14, 2002 7:55 AM
> To: python-list at python.org
> Subject: Looking for GUI Form Builder for Python
> 
> Dear all,
>  Can I build Python TK GUI interface using a form builder. Anyone know
> such
> builder. I just want to build my interface in a visual way.
> -Calvin
> 
> 

Paul Sage wrote:

> Have you investigated QT?  I believe it might be able to do what you are
> looking for.  However, maybe someone on this list who has worked with it
> will be able to confirm that for you.  I have not worked with it, but I
> believe it has a Python wrapper available.


Yes, PyQt can use the output of Qt Designer. It doesn't design
Tk forms, of course, but Qt forms. I think there's only one Tk
form builder available: Secret Labs' Python IDE, Pythonworks.
(www.secretlabs.com)

In my biased opionion, Qt is superior, but Pythonworks has really
grown up.

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


From anthony at interlink.com.au  Mon Apr 22 22:22:30 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Tue, 23 Apr 2002 12:22:30 +1000
Subject: need encryption algorithm 
In-Reply-To: Message from h.e.w.frater@cs.cf.ac.uk (Hugh) 
   of "22 Apr 2002 18:41:08 MST." <e2e9e807.0204221741.597acd60@posting.google.com> 
Message-ID: <200204230222.g3N2MVV03621@localhost.localdomain>

>>> Hugh wrote
>   I need a simple encryption algorithm that will not give me
> characters that I can't store in my oracle table. At the moment I'm
> using the rotor encryption module because I like it's simple interface
> but It's generating characters that oracle can't store which is a damn
> pain. What other modules are available that are easy to use and will
> do what I wan't. Remember that security is not paramount here, I'm
> only including it for functionality.
> 

base64 encode whatever you're planning to use. easy peasy.

Anthony

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From rxg218 at psu.edu  Thu Apr 11 11:35:40 2002
From: rxg218 at psu.edu (Rajarshi Guha)
Date: Thu, 11 Apr 2002 11:35:40 -0400
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu> <eppstein-40A0D7.17104110042002@news.service.uci.edu> <3CB4E282.AAD468E4@engcorp.com>
Message-ID: <a94ah7$22ja@r02n01.cac.psu.edu>

On Wednesday 10 April 2002 21:10 in comp.lang.python Peter Hansen wrote:

> David Eppstein wrote:
>> 
>> In article <a92hl1$1v4c at r02n01.cac.psu.edu>,
>>  Rajarshi Guha <rxg218 at psu.edu> wrote:
>> 
>> > I got this code to work:
>> >
>> > liter = iter(l)
>> > while(1):
>> >   try:
>> >      print liter.next(), liter.next()
>> >   except StopIteration:
>> >      break
>> >
>> > Works fine - is it an efficient way to handle this situation?
>> > (I could arrange the data into tuples if necessary)
>> 
>> Does it work fine?  What if the list has an odd number of items -- the
>> last item won't be printed.
> 
> It works perfectly, given the requirements specified.
> 
> If the customer wanted to handle lists with odd numbers of items
> he would have specified the desired behaviour.  Or he will now. :-)

Exactly :)
My list will always have an even number of elements.

But I think the best thing to do is to simply restructure my data structures
and use tuples. Smart data structures are the way to go!

-- 
Rajarshi Guha
rajarshi at presidency.com


From syver-en+usenet at online.no  Tue Apr  9 20:46:23 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Wed, 10 Apr 2002 00:46:23 GMT
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de> <uvgb0eefo.fsf@turangalila.harmonixmusic.com>
Message-ID: <u8z7wwfdo.fsf@online.no>

Dan Schmidt <dfan at harmonixmusic.com> writes:

> Thomas Guettler <pan-newsreader at thomas-guettler.de> writes:
> 
> | Up to now I was happy with emacs because I never used
> | to programm it with lisp. Now I need to and find it difficult.
> | 
> | Is there an emacs-like editor which is scriptable with python?
> 
> Pymacs allows you to script Emacs with python.  I haven't used it,
> though.
> 

Pymacs is great. I use it everyday. Very neat. You do have to learn
the emacs api anyway though, even if you can write to it in Python
instead of lisp.
 

-- 

Vennlig hilsen 

Syver Enstad


From jb at cascade-sys.com  Tue Apr 23 12:53:11 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 23 Apr 2002 09:53:11 -0700
Subject: non-blocking sockets
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>
Message-ID: <3CC59176.9944B750@cascade-sys.com>

so very tired wrote:

> When trying to read from a socket, I don't want it hanging, so I set it to
> non-blocking by calling
>
> setblocking(0)
>
> but then when I try to read from it and there is no data in the socket, I
> get an error. I tried this in both windows and linux and in both I get an
> error and the program stops. It says there's an exception but it doesn't
> say the name of the exception so I can't even try to catch.
> Does anyone know how I can check if there's data coming from a socket
> connection without hanging or having the program barf on me?

I suspect you need to use "select" to query whether a given socket is ready
for reading or writing.  It has its own section in the library doc.  In
general it waits on multiple sockets and/or file descriptors and tells you
which ones are "ready".  (However, it does not work with anything EXCEPT
sockets on Windoz.)  Select also has a delay parameter so it will return after
an interval even if nothing is happening, so you program can pretend it's not
dead.

A newer alternative is "Polling Objects," in the section immediately following
"select".  I never used it but it appears marginally better than "select".

However, I personally find it MUCH easier to use blocking I/O in conjunction
with threads, rather than the older, historical techniques.  And since Python
threads are available in Windows and most Unix dialects (Linux, anyway),
there's little reason not to use them.  Furthermore, I find coding it out long
hand with threads to be lots easier to figuring out how to adapt library code,
such as "asyncore" or "SocketServer".  The result also is much clearer to
understand.

E.g., below is something I was working on this weekend (undefined variables
are left to the reader's imagination):

    import os, sys
    import socket, thread

    # first thread listens for incoming connects
    # and then creates a new thread to handle each one

    def Listener():
        s = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
        s.bind( HOST, PORT )
        while 1:
            s.listen(3)
            conn, addr = s.accept()
            thread.start_new_thread( Reader, ( conn, addr ))

    # independent reader thread, processes incoming data as it arrives

    def Reader( conn, addr ):
        while 1:
            data = conn.recv( 1024 )    # blocking!
            if not data:
                break
            process( conn, data )
        conn.close()

    # main program launches threads and
    # loops, waiting to handle incoming SIGNALS

    def main():
        thread.start_new_thread( Listener, ())
        thread.start_new_thread( Reader, ())

        # signal handling code omitted

        while 1:
            time.sleep( 5 )

If your protocol is simple enough, 'process()' may result in sends on the
socket.  If the protocol is complex then you may need separate threads for
reading and for writing.

The one hitch is where you have different threads accessing common data then
you need to serialize access via LOCKS.  This is a problem you don't have with
a single, blocking socket, or with multiple sockets tested via "select".

    aLock = thread.allocate_lock()
    ...
    aLock.acquire()
    access_shared_resource()
    aLock.release()

Also, the Queue data structure is handy for queuing data between multiple
threads and if you poke around, there are other tools for concurrent
programming.

Lemme know if this helps.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From starx at pacbell.net  Fri Apr 12 22:48:47 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 13 Apr 2002 02:48:47 GMT
Subject: Initilizing MD lists w/ objects.
Message-ID: <Xns91EEC9BB14699RASXnewsDFE1@130.133.1.4>

I'm looking for a reasonable way to initilize a multidimensional list with 
a bunch of blank objects of some type (already been burned by using [obj
()] * 400 or similar =). This is the best i've come up with, but it seems 
icky:

        self.data = [None]*40

        for x in range(40):
            self.data[x] = []
            for y in range(40):
                self.data[x].append(Vis_cell())

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From James.Strater at exolink.com  Tue Apr 23 10:32:10 2002
From: James.Strater at exolink.com (James Strater)
Date: Tue, 23 Apr 2002 09:32:10 -0500
Subject: reading unix files from NT using SSH2
Message-ID: <F666B26E8EF5B84C88E3DF34E5A1E49CF081E1@venus.exolink.com>

I have a python script running on an NT box.  I need to connect to a  unix
box using SSH2 and read some files sitting on that box.  Can anyone point me
in the right direction on how to go about doing this?   The script needs to
stay on the NT box so that it can do some things with MS Excel using COM.  

Thanks,

James




From mcfletch at rogers.com  Tue Apr 16 00:25:36 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 16 Apr 2002 00:25:36 -0400
Subject: [2.2] Weird behaviour of file.write() (no data written)?
Message-ID: <3CBBA7C0.6020405@rogers.com>

I'm trying to figure out what's going wrong with my index-storing 
routines.  These are dirt-simple pickle-dumps to disk.  I must be doing 
something terribly obviously wrong, but darned if I can figure out what 
it is.


Here's the relevant code:

fileHandle = open( self.filename,'wb',0)
values = {}
for key, value in self.tables.items():
	if isinstance( key, Persistent ):
		if not hasattr( key, '_p_oid'):
			raise ValueError( """Attempting to store a collection index for a 
collection not currently in the database %s"""%( key))
		key = key._p_oid
	values[key] = value
self.tables = values
print 'dumping indices to', self.filename
data = dumps( values)
fileHandle.write( data)
fileHandle.write( '')
fileHandle.flush()
fileHandle.close()
print 'fileHandle closed'


What happens with that code is that a 0-length file is created on disk, 
despite the pickle being 7 or 8 kb in length.  Now, this code works (no 
clue why it's even different):

values = {}
for key, value in self.tables.items():
	if isinstance( key, Persistent ):
		if not hasattr( key, '_p_oid'):
			raise ValueError( """Attempting to store a collection index for a 
collection not currently in the database %s"""%( key))
		key = key._p_oid
	values[key] = value
self.tables = values
print 'dumping indices to', self.filename
data = dumps( values)
fileHandle = open( self.filename,'wb',0)
fileHandle.write( data)
fileHandle.write( '')
fileHandle.flush()
fileHandle.close()
print 'fileHandle closed'

(The only difference being _when_ the fileHandle object is created). 
I've tried buffered and unbuffered files, using pickle.dump instead of 
dumps, and just about everything else I could think of.

The code is getting called in the OnClose handler of a wxPython wxFrame 
object.  Zope 3's ZODB is being used (it's what's pulling in 
Persistent), but I'm not using it for the indices (the only things in 
the indices are readily pickle-able objects).


There's lots of alpha or beta-quality C software running around this 
application (Zope3 and wxPython 2.3.3pre2, respectively), but I'd be 
surprised if it could affect something as stable as file.write without 
causing something a little more obvious to occur.

Ideas welcome,
Mike
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From marklists at mceahern.com  Wed Apr  3 12:05:44 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 11:05:44 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.44.0204031119240.29628-100000@penguin.theopalgroup.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEPPKIAA.marklists@mceahern.com>

[Kevin Jacobs]
> The moral of the story is that I will be extremely happy once there is a
> globally blessed way of spelling true and false other than 1 and 0.  Of
> course, I and my team aren't exactly typical Python developers.  Though
> wouldn't it be nice if we were?

I'd like to consciously go out of my way and violate all rules of netiquette
and say simply this:

Me too.

Cheers,

// mark





From grey at despair.dmiyu.org  Fri Apr 12 19:11:10 2002
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Fri, 12 Apr 2002 23:11:10 -0000
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com>
Message-ID: <slrnabeqci.s6b.grey@teleute.dmiyu.org>

On 12 Apr 2002 16:04:50 -0700, Paul Rubin <phr-n2002a at nightsong.com> wrote:
> I'd say it's hopeless or at least impractical.  Those processors are
> much too constrained to run Python.  It would be like trying to run an
> SQL database on them.  Python has many virtues but being able to run
> on very tiny processors isn't one of them.

    What about something like the Zaurus SL-5500?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------


From erict at millfilm.co.uk  Tue Apr 30 06:03:57 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Tue, 30 Apr 2002 11:03:57 +0100
Subject: Tix Vs Pmw which is better to use?
Message-ID: <3CCE6C0D.FA96F79A@millfilm.co.uk>

My application needs a very dynamic widget package, It needs to add and
remove  widget constently.
I am looking also for something that will let me create more organic
widget like tree but on a free canvas.

- Do I have to be aware of some major difference before I should choose
one or the other?
- Is there an other package to be considered ?

Thanks
Eric Texier



From aahz at pythoncraft.com  Mon Apr  8 19:49:53 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Apr 2002 19:49:53 -0400
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <MPG.17179f847bf41a6998969b@news.nwlink.com> <mailman.1018261577.4898.python-list@python.org>
Message-ID: <a8tab1$imf$1@panix1.panix.com>

In article <mailman.1018261577.4898.python-list at python.org>,
Graham Ashton  <gashton at cmedltd.com> wrote:
>
>I remember being slightly surprised by the fact that self finds class
>variables. Is there any particular reason why, considering that the you
>can reference them as ClassName.var anyway?

It's part of the whole superclass lookup mechanism:

    class A:
        def foo(self):
            print "!"

    class B(A):
        pass

    B().foo()
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From erno-news at erno.iki.fi  Thu Apr 11 02:00:50 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Apr 2002 09:00:50 +0300
Subject: Python server from inetd?
References: <iml4bukknqf9nqe13h7dp0qco3qj2r86rt@4ax.com> <kuwuvhmoiy.fsf@lasipalatsi.fi> <mb1abucnutggsi6v22mbo03qisn1l8agm3@4ax.com>
Message-ID: <kuy9fuydt9.fsf@lasipalatsi.fi>

In article <mb1abucnutggsi6v22mbo03qisn1l8agm3 at 4ax.com>, Tim Roberts
<timr at probo.com> writes:

| You're right.  That's what I get for posting from memory instead of from a
| file.  I did figure out that I needed to use stdin and skip the
| translate-to-socket model.

using socket.fromfd should work fine. you can't do things like
getpeername() if you don't make a socket object out of the fd.

  -- erno


From fperez528 at yahoo.com  Mon Apr 22 12:55:58 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 10:55:58 -0600
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org>
Message-ID: <aa1fe9$mn7$1@peabody.colorado.edu>

> Second, math *is* an attempt to describe the physical world - that's what
> makes it useful. Case in point: calculus, which was invented/discovered
> specifically to deal with and describe the motion of physical bodies. Had
> it not been useful in describing such motion, it would have been tossed
> out, or at least not so widely accepted. Math is useful because of its
> relevance to the world around us.

Sorry but no. The fact that certain physical problems may have served as 
inspiration for some mathematical developments can be considered to be purely 
accidental. The foundational structure of mathematics is entirely one of 
logical self consistency, and massive amounts of work was done particularly 
in the 20th century to build that framework. While it is true that up to the 
19th century most (if not all) significant advances in math were prompted by 
challenges from the physical sciences, the work of the formalists in the 20th 
century showed that mathematics could be built entirely on an abstract, 
logical foundation.

That is not to say that it's not _useful_ to have a reference to the physical 
world both for inspiration and for applications, but there's a crucial 
difference between that and saying that 'math *is* an attempt to describe the 
physical world'.

What I do find most fascinating is how many seemingly super-abstract results 
of mathematics turn up (sometimes decades later after they are found) having 
physical implications which noone ever suspected. That 'mysterious 
mathematical structure of the physical world' is one of the most beautiful, 
in my opinion unanswerable questions of the foundation of science.

Cheers,

f.


From nospam at bigfoot.com  Tue Apr 16 09:55:53 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 16 Apr 2002 15:55:53 +0200
Subject: xml2dict - dict2xml
References: <a9gudi$964$04$1@news.t-online.com>
Message-ID: <a9ha1u$2iri$1@norfair.nerim.net>

The PyRXP XML parser that does it (almost) directly, and very fast.
Anyway, the data structure provided by pyrxp should be easily stuffed to the
format you want.
In addition, it's said to be the fastest of the world and has the best
memory ratio.

http://www.reportlab.com/xml/pyrxp.html

HTH

--Gilles

"Henning Peters" <pete at dexterslab.de> a ?crit dans le message news:
a9gudi$964$04$1 at news.t-online.com...
[...]




From sholden at holdenweb.com  Sat Apr 20 12:17:03 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 20 Apr 2002 12:17:03 -0400
Subject: Newbie: creating an <li> list
References: <c73096a5.0204200357.f7b9169@posting.google.com>
Message-ID: <oygw8.26465$T%5.15023@atlpnn01.usenetserver.com>

"flyguy" <flyguysyl at hotmail.com> wrote in message
news:c73096a5.0204200357.f7b9169 at posting.google.com...
> Hy. I'd like to know how you can send a new html page with a group of
> radio button. Here's my attempt:
>
>   r1 is the value of a radio button that a user checks so that he can
> be offered a format of response (in this case radio buttons) to a
> question he composed while constructing a survey:
>
> if not (form.has_key("question") and form.has_key("r1") and
> form.has_key("r2")):
>     MaPage=MaPage +  "<H1>Error</H1>"
>     MaPage=MaPage + "<h3>Veuillez construire votre question et choisir
> un format de reponse</h3>"
>
>
> else:
>  if (form.has_key("question") and form.has_key("r1") and
> form.has_key("r2")):
>         Q=form["question"].value
>         MaPage=MaPage + "<center><h2>Voici votre question ainsi que
> votre choix de reponse</h2></center>"
>         MaPage=MaPage + "<br><br><table width=80%><tr><td><h2>" + Q +
> "</h2></td></tr></table>"
>
>         if(form["r1"].value==1):
>              MaPage=MaPage + "<ol><li><input type=radio name=rep1
> value=1>"
>              MaPage=MaPage + "<li><input type=radio name=rep1
> value=2>"
>              MaPage=MaPage + "<li><input type=radio name=rep1
> value=3>"
>              MaPage=MaPage + "<li>input type=radio name=rep1
> value=4></ol>"
>
> MaPage=MaPage + "</body></HTML>"
> print MaPage

You will probably find your radio buttons behave better if you actually
enclose them in "<form> ... </form>" tags!

If you don't want the user to be able to submit the form, don't put an
ACTION attriobute on it, and don't put an <input type="submit"> button
either.

You will also find it hlpful to use the triple-quoted type of string
literal, whioch would (for example) allow you to have written

        if(form["r1"].value==1):
             MaPage=MaPage + """
            <ol>
            <li><input type=radio name="rep1" value="1">
             <li><input type=radio name="rep1" value="2">
             <li><input type=radio name='rep1" value="3">
             <li><input type=radio name="rep1" value="4">
            </ol>
            </body></HTML>"""

This makes it much easier to produce conformant [X]HTML, which we should all
try to do.

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From stephen at xemacs.org  Mon Apr 15 23:35:27 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 16 Apr 2002 12:35:27 +0900
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org>
Message-ID: <87vgaswc1s.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "David" == David Mertz <mertz at gnosis.cx> writes:

    David> There is nothing very difficult (and nothing -at all-
    David> ambiguous) in here.

Nothing is difficult or ambiguous for any True Believer.  It is true
that if you define the issues narrowly as you have, there is nothing
difficult or ambiguous.

Those of us seeking truth, on the other hand, often must accept that
multiple definitions and opinions are current, and try to adapt to
others' points of view to further communication.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From phr-n2002a at nightsong.com  Fri Apr  5 18:01:32 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 15:01:32 -0800
Subject: PEP 285 and pickle compatibility
References: <7xsn6auy3i.fsf_-_@ruckus.brouhaha.com> <3CADF727.8A875581@python.org>
Message-ID: <7xwuvlagxf.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
> I like this.  I'll see if I can implement it.  (Though it's not clear what
> to do in binary mode.)

Just always write bools in 'I' format, even in binary mode.  The unpickler
shouldn't mind, unless I missed something.

> I like this recommendation, but I'm not motivated enough to implement
> it myself.  Maybe you can get some people to work with you on a PEP
> and a sample implementation?

OK.  I'll see if I can code up something, and worry about a PEP afterwards.
I'd like to see it implemented so I can stop using marshal and
its lack of cross-version compatibility to pass long ints around.


From distrex at hotmail.com  Wed Apr 10 17:29:37 2002
From: distrex at hotmail.com (distrex)
Date: 10 Apr 2002 14:29:37 -0700
Subject: port scan
References: <d9ea2696.0204091410.7755db60@posting.google.com> <mailman.1018397300.27960.python-list@python.org>
Message-ID: <d9ea2696.0204101329.5cacb9c2@posting.google.com>

> 
> If by a port scanner you only want to scan for open ports, this is
> surely possible, and not difficult, either. Simply try to connect to the
> port, if this succeeds, you've found an open port. This is a minimal
> example for a Python portscanner (only checks for TCP so far):
> 
> #!/usr/bin/env python
> import sys, socket
> 
> def checkPortTCP(host, port):
>     sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
>     try:
>         sock.connect((host, port))
>         return 1
>     except socket.error, reason:
>         return 0
> 
> def main():
>     hostname = sys.argv[1]
> 
>     port_range = range(1024)
>     open_ports = []
>     for port in port_range:
>         if checkPortTCP(hostname, port):
>             open_ports.append(port)
>     print open_ports
> 
> if __name__ == "__main__":
>     main()
> 
> I recommend to look into the documentation of the socket module, and if
> this looks like chinese to you, as it was to me, the Socket Programming
> HOWTO at http://py-howto.sf.net/
> 
> Gerhard

Thanx for that it helped. That is what I wanted so I'll check out that
URL. Thanx again


From paul at prescod.net  Tue Apr 16 17:16:32 2002
From: paul at prescod.net (Paul Prescod)
Date: Tue, 16 Apr 2002 14:16:32 -0700
Subject: O_O <- Re: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net> <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1> <Xns91F2905ED8DAARASXnewsDFE1@130.133.1.4>
Message-ID: <3CBC94B0.58505F48@prescod.net>

Yes, I will probably change the name of that. I would have put that in
long before minidom was in the Python standard library!

Philip Swartzleonard wrote:
> 
> Duncan Booth || Mon 15 Apr 2002 04:35:13a:
> > [...]
> > __paul_prescod_hack_constant__
> > [...]
> 
> Now there's an interesting variable.
> 
> Also interesting that _..._undefined_..._ has so many underscores...
> 
> --
> Philip Sw "Starweaver" [rasx] :: www.rubydragon.com
> --
> http://mail.python.org/mailman/listinfo/python-list




From cgale1 at cox.net  Sat Apr 13 21:54:54 2002
From: cgale1 at cox.net (Van Gale)
Date: Sun, 14 Apr 2002 01:54:54 GMT
Subject: Checking whether an async* dispatcher is "alive"
References: <slrnabggo4.o80.mlh@vier.idi.ntnu.no>
Message-ID: <Oj5u8.34280$%A.2510727@news1.west.cox.net>

"Magnus Lie Hetland" wrote:
> In a server I'm working on, I need to check whether an asynchat
> dispatcher is still connected, i.e. whether I can send it data. I've
> tried to use try/except (with socket.error) but since async* catches
> that itself, It's no good.
>
> Are there any methods/attributes that indicate whether I can send data
> on a dispatcher? (Possibly some method/attribute on its socket
> object?)

Hmm, this is a bit tricky and depends some on how your protocol works.

If your protocol is request/response oriented (primarily what
asyncore/asynchat were designed for), check out http_server.py and
status_handler.py in the medusa distribution (www.nightmare.com).  These
take a fairly simple approach: after every X number of connections (default
X = 500) close any sockets that have been open for longer than 30 minutes.
This works fine for request/response like HTTP where 30 minutes is probably
a good indicator it's a zombie.

I presume however, that your protocol is asynchronous over a persistent
connection since you are implying you want to send data to a client without
having just recieved a request.  In this case you really only have 2
choices: either just trust TCP's built-in reliability, tr-try, and
keep-alive features or you can implement some kind of heartbeat message.
See pg 581 of [1] for a discussion and sample C code for using TCP's ugent
mode to implement heartbeats.  His implementation has the client send the
heartbeats and the server raises a signal if it hasn't recieved any
heartbeats after a specified interval.  I don't know how portable this is,
or even how easy it would be to implement in Python if you're using async*
classes.

My personal preference (atm) is just trusting TCP and handling errors when
the connection is broken because you're going to get a send error on the
heartbeat anyway :)  If you're implementing something like a server for an
IM protocol (e.g. Jabber) which needs up to date "presence" information,
then it should have a heartbeat built-in to the protocol (as does Jabber).

Btw, the Medusa http_server, status_handler, and the monitor_server, are
probably worth checking out even if your protocol is fully asynchronous.  I
really like having a port I can connect to with a browser that gives me a
nice picture of the internal status of the running server, and
monitor_server takes the ultimate step: it gives you an interactive Python
session inside your running server.

Van
[1] W. Richard Stevens "Unix Network Programming: Networking APIs: Sockets
and XTI Volume 1"





From aahz at pythoncraft.com  Thu Apr 11 21:25:19 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 21:25:19 -0400
Subject: Beginner's performance problem
References: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de> <memo.20020411143211.1400B@a.a>
Message-ID: <a95d1v$avb$1@panix1.panix.com>

In article <memo.20020411143211.1400B at a.a>,
Mark Charsley <mark.charsley at REMOVE_THIS.radioscape.com> wrote:
>
>TVM to everyone who helped.

I'm familiar with TVP, but what's TVM?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From gimbo at ftech.net  Thu Apr 18 04:12:46 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 18 Apr 2002 09:12:46 +0100
Subject: Readline/command history w/ interactive prompt
In-Reply-To: <a9kcml$69k$1@news.wrc.xerox.com>
References: <mailman.1019061574.15837.python-list@python.org> <a9kcml$69k$1@news.wrc.xerox.com>
Message-ID: <20020418081246.GB3689@andy.tynant.ftech.net>

On Wed, Apr 17, 2002 at 05:51:17PM +0000, Mark Jackson wrote:

> Try hitting control-r and typing a short string that last appeared
> in the line desired ("my" or even "m" would work in this case).

_Nice_!  Thanks for that.  One question though: where's this feature
documented?  :-)

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From ods at fep.ru  Wed Apr  3 03:49:58 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Wed, 3 Apr 2002 12:49:58 +0400 (MSD)
Subject: How to check whether email /domain exists
In-Reply-To: <3CAADA7B.25443.509AE1@localhost>
Message-ID: <Pine.LNX.4.44L.0204031248550.13708-100000@ods.pp.ru>

On Wed, 3 Apr 2002, A wrote:

A> What is the simplest way how to check
A>  whether a given email address( or domain for this email)
A> exists?
A> Thanks for help

Check for MX record, e.g. with
http://www.interlink.com.au/anthony/python/dns.tar.gz




From emile at fenx.com  Tue Apr  9 23:48:48 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 10 Apr 2002 03:48:48 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <ACOs8.29598$CU.39610@rwcrnsc51.ops.asp.att.net>

Andrew Koenig
> I would like to be able to write an expression that yields an
> object that I can subsequently use to rebind a name that I mention
> only in that expression.  How do I do it?


If you need to, you could try:

class rebind:
    def __init__(self, var, context):
        for ky, val in context.items():
            if val is var:
                self.var = ky
                break
        self.context = context
    def __call__(self, newval):
        self.context[self.var] = newval

x = 3
tx = rebind(x, globals())
tx(4)
x

although I'd be real careful where and how you deploy.  This breaks in
lots of fun-to-find ways.  After finding the first three or four, an
alternate solution or design will start looking good.

Did-Schr?dinger's-cat-have-a-name?-ly y'rs

--

Emile van Sebille
emile at fenx.com

---------



From ark at research.att.com  Thu Apr 25 18:43:01 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 25 Apr 2002 22:43:01 GMT
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <yu994rhzmmbu.fsf@europa.research.att.com>

>>>> float(5)/float(100)
Chris> 0.050000000000000003
>>>> 5.0/100.0
Chris> 0.050000000000000003

Chris> 	Okay, I might not be a math genius, but this is most definately WRONG.

Why do you think so?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From altis at semi-retired.com  Fri Apr 19 17:28:00 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 19 Apr 2002 14:28:00 -0700
Subject: How to save bitmap from MS Windows clipboard to xxx.png?
References: <a9p1e3$efb$1@news.nextra.cz> <a9pb8b$2nlg$1@norfair.nerim.net>
Message-ID: <yY%v8.82$4K.78915@news.uswest.net>

There are two issues here. One is getting the contents of the clipboard. The
second is taking a bitmap in memory and then saving it as a PNG file. I just
tried this snippet in wxPython which will work in Windows, Linux, etc.

from wxPython import wx

...

        try:
            if wx.wxTheClipboard.Open():
                if not
wx.wxTheClipboard.IsSupported(wx.wxDataFormat(wx.wxDF_BITMAP)):
                    return
                do = wx.wxBitmapDataObject()
                wx.wxTheClipboard.GetData(do)
                bmp = do.GetBitmap()
                bmp.SaveFile('clipboard.png', wx.wxBITMAP_TYPE_PNG)
                wx.wxTheClipboard.Close()
        except:
            pass

This is not a complete app and could be better about handling the possible
exceptions, but that's the basic idea. There is something like this included
in the doodle sample in PythonCard, except that it pastes the image into the
on screen window instead of saving the bitmap to disk.

wxPython home page
http://www.wxpython.org/

PythonCard home page
http://pythoncard.sourceforge.net/

ka

"Gillou" <nospam at bigfoot.com> wrote in message
news:a9pb8b$2nlg$1 at norfair.nerim.net...
> Hi,
>
> Have a look to the Python Imaging Library (AKA PIL). This should be a good
> start for converting graphic formats
>
> http://www.pythonware.com/products/pil/index.htm
>
> And have a look to the Python Win32 extensions that may (?) provide
> scripting access to the clipboard.
>
> http://starship.python.net/crew/mhammond/
>
> HTH
>
> --Gilles
>
> "Petr Prikryl" <Answer.via.news.please.prikryl at skil.nospam.cz> a ?crit
dans
> le message news: a9p1e3$efb$1 at news.nextra.cz...
> > Hi,
> >
> > I am quite new to Python to now what was done already
> > and what was not.
> >
> > What I need is to save the window snapshot (in MS
> > Windows) into PNG file.  It just came to my mind that
> > some Python utility may already exist for that task
> > or that some PNG support is implemented.
> >
> [...]
>
>




From sholden at holdenweb.com  Mon Apr  1 08:28:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 08:28:54 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017637982.13565.python-list@python.org>
Message-ID: <nhZp8.108797$Ou.105859@atlpnn01.usenetserver.com>

"Arthur Siegel" <ajs at ix.netcom.com> wrote in message
news:mailman.1017637982.13565.python-list at python.org...
> >>Inventing new, non-standard terminology in order to be >>"simpler"  is
> rarely helpful; it doesn't substantially help  >>newbies, and it only
serves
> to confuse people actually >>involved.
>
> >This is born out by experimental psychologists, e.g.
>
> Also interesting is group psychology.
>
> I was pillared on Edu-Sig for making this point.
>
I think the word you're looking for is "pilloried".

> Of course I was referencing Alice and Guido's Linux
> journal article - not chiming in support of his insight.
>
all-in-the-interests-of-english-education-ly y'rs  - steve






From shippy at nmt.edu  Tue Apr 16 16:57:11 2002
From: shippy at nmt.edu (Jeff Shipman)
Date: Tue, 16 Apr 2002 14:57:11 -0600
Subject: using httplib
Message-ID: <3CBC9027.3010603@nmt.edu>

I'm trying to use httplib under Python 1.5.2 to
submit a POST request to a remote server. I would
use Python 2.x if I could, but I do not control
the webserver I'm using and they do not plan to
upgrade for a long while. However, it's not working
for me. I have tried it with Python 2.x and it
works fine. Here's an example:

Python 2.x
headers = { "Content-type": "application/x-www-form-urlencoded" }
conn = httplib.HTTPConnection("www.paypal.com")
conn.request('POST', "/cgi-bin/webscr", "cmd=_notify-validate", headers)
res = conn.getresponse()
print res.read()
INVALID

INVALID is the correct response.

Python 1.5.2

conn = httplib.HTTP("www.paypal.com")
conn.putrequest('POST', "/cgi-bin/webscr")
conn.putheader('Content-type', 'application/x-www-form-urlencoded')
conn.endheaders()
conn.send('cmd=_notify-validate')
conn.getreply()
(200, 'OK', <mimetools.Message instance at 808e128>)
conn.getfile().read()
... Bunch of HTML because form wasn't posted properly ...

I've also tried it this way:

conn = httplib.HTTP("www.paypal.com")
conn.putrequest('POST', "/cgi-bin/webscr?cmd=_notify-validate")
conn.endheaders()
conn.getreply()
(200, 'OK', <mimetools.Message instance at 808c330>)
conn.getfile().read()
... Bunch of HTML because form wasn't posted properly ...

Any idea of how I can get this to work with Python 1.5.2?
Could I possibly just snag the 2.x httplib and it would
work under 1.5.2?

Thanks in advance,

-- 
Jeff "Shippy" Shipman     E-Mail: shippy at nmt.edu
Computer Science Major    ICQ: 1786493
New Mexico Institute of Mining and Technology
Homepage: http://www.nmt.edu/~shippy



From aleax at aleax.it  Mon Apr 29 13:33:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 17:33:36 GMT
Subject: caching of modules?
References: <aajsp7$b84$1@netnews.upenn.edu>
Message-ID: <Qtfz8.32183$8D3.903638@news1.tin.it>

Mike Kayser wrote:
        ...
> It looks to me like it's caching the module somewhere, but then when it

Yes, in sys.modules.

> spits out error information it goes back to the file itself, even if it's

Indeed, it 'caches' the compiled form, not the sources.

> Does anyone know if my hypothesis is right? If so, how can I force the
> "module cache" to reload the updated file?

Built-in function reload lets you reload any given module object.


Alex



From opengeometry at NOSPAM.yahoo.ca  Sat Apr 27 18:47:46 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 27 Apr 2002 22:47:46 GMT
Subject: Simple bar graphs
References: <Xns91FCD43535458Latzikatz@139.174.2.56> <aacas8$9d997$2@ID-99293.news.dfncis.de> <Xns91FCE04392F6ELatzikatz@139.174.2.56>
Message-ID: <aaf9qh$a5v8a$1@ID-99293.news.dfncis.de>

Lutz Schroeer <Lutz.Schroeer at kybernetik-manufaktur.de> wrote:
> William Park <opengeometry at NOSPAM.yahoo.ca> wrote in
> news:aacas8$9d997$2 at ID-99293.news.dfncis.de: 
> 
>> I used 'gdchart' long ago.  It generates GIF directly.  But, if your
>> requirement is simple enough, you can use HTML rules, say in table
>> layout. 
> 
> Looked at that, too, good stuff, but unfortunately it's just a Python 
> frontend to the GDChart package (IMHO), which is a C lib and needs 
> compiling, which isn't permitted on my webspace for security reasons (says 
> my provider).

In that case, your option is pretty limited.  If you have access to TeX and
ImageMagick, then you can get EPS using PiCTeX, then convert it to GIF
using ImageMagick.  If not, use HTML codes (ie. <hr ...>) directly.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-cpu Cluster, Hosting, NAS, Linux, LaTeX, Python, vim, mutt, tin


From dradul at yahoo.com  Tue Apr 30 13:34:19 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Tue, 30 Apr 2002 12:34:19 -0500
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <aact9h$9odgq$2@ID-99513.news.dfncis.de> <c3d20384.0204292330.ea14313@posting.google.com>
Message-ID: <aamlqj$bu823$1@ID-99513.news.dfncis.de>

"Dan Jacobson" <jidanni at hotmail.com> escribi? en el mensaje
news:c3d20384.0204292330.ea14313 at posting.google.com...
> > This will help you get started:
> >
> >         http://pythonzh.sourceforge.net/
>
> I see nothing there.


cvs -z9 -d:pserver:anonymous at cvs.pythonzh.sourceforge.net:/cvsroot/pytho
nzh co pythonzh

from the top of my head. If it doesn't work take a look at the project
pages.



From martin at v.loewis.de  Tue Apr 16 02:23:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:23:50 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de> <eppstein-922653.14382815042002@news.service.uci.edu>
Message-ID: <m3ads4maa1.fsf@mira.informatik.hu-berlin.de>

David Eppstein <eppstein at ics.uci.edu> writes:

> > Yes, but this notion is inherently bound with the notion of "size" of
> > an integer, atleast in the signed variant (for unsigned integers,
> > there is no need for twos complement). If you create the complement,
> > you have to know how many bits to invert.
> 
> Not if you have a representation which (when interpreted as a bitvector) 
> allows bitvectors with infinitely many one-bits.  Which in fact is what 
> we have with Python's long integer type.

Unfortunately, you cannot represent such a number in hex, atleast not
under any of the usual conventions: the convention is that you can
strip infinitely many leading zeroes, but stripping off infinitely many
leading ones is confusing.

Regards,
Martin



From sholden at holdenweb.com  Wed Apr 17 09:42:40 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 09:42:40 -0400
Subject: pre-PEP for optional 'pass'
References: <mailman.1018972676.30534.python-list@python.org> <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1> <mailman.1019035355.21475.python-list@python.org>
Message-ID: <h%ev8.13380$hl2.4876@atlpnn01.usenetserver.com>

"Andy Gimblett" <gimbo at ftech.net> wrote ...
> On Wed, Apr 17, 2002 at 08:33:44AM +0000, Duncan Booth wrote:
>
[example code]
> >
> > I would say that all of these comments could be automatically generated,
> > but if I look at the output of (say) pydoc, they add useful information
as
> > I might otherwise be misled into thinking that the class and methods did
> > something.
>
> We're veering slightly from the original reason for this point, which
> was how to write stub functions/methods which will be implemented
> later.  The above certainly doesn't indicate any intent to actually
> implement the methods (in fact, on the contrary, it suggests they
> could be removed).
>
> How about this?
>
> class MyClass:
>     """Class encapsulating a Twiddle.  Unfinished.  Document me.  XXX"""
>     def __init__(self, x, ,y, z):
>         """Not yet implemented. XXX"""
>     def getTwiddle(self):
>         """Not yet implemented. XXX"""
>     def setTwiddle(self, twiddle):
>         """Not yet implemented. XXX"""
>
> where you can grep for XXX to find unfinished business - very handy.
>
But surely no less handy than, say, the following:

class MyClass:
    """Class encapsulating a Twiddle.  Unfinished.  Document me.  XXX"""
    def __init__(self, x, ,y, z):
        raise NotImplementedError    # XXX
    def getTwiddle(self):
        raise NotImplementedError    # XXX
    def setTwiddle(self, twiddle):
        raise NotImplementedError    # XXX

This also has the merit of not simply (and silently) doing nothing, and can
still be grepped for the necessary change areas. But this thread has now
gone round in circles twice without anyone changing their minds, so at this
point I'll leave it to vanish up its own orifice.

regards
 Steve






From pedroni at inf.ethz.ch  Tue Apr  2 14:12:08 2002
From: pedroni at inf.ethz.ch (Samuele Pedroni)
Date: Tue, 2 Apr 2002 21:12:08 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com>
Message-ID: <001101c1da7a$605da660$6d94fea9@newmexico>

From: M.-A. Lemburg <mal at lemburg.com>
> As a conclusion, I think it would be better to make bool() a 
> new type which does *not* inherit from integers, but which 
> does know how deal with other types which are commonly
> used together with booleans such as integers. However, the
> type should implement boolean algebra and not try to
> mimic integer arithemtic, i.e. True - True raises an
> exception.
> 

"but which 
 does know how deal with other types which are commonly
 used together with booleans such as integers" ??

with deal you mean throwing exceptions?

regards.









From observer at NOSPAM.space.pl  Tue Apr  9 07:46:27 2002
From: observer at NOSPAM.space.pl (RobW)
Date: Tue, 09 Apr 2002 13:46:27 +0200
Subject: Simple question! urgent
Message-ID: <62l5buo5do5eh0h93b4r9ua8skp61b7302@4ax.com>

x = [1,2,3,4]

How to get NUMBER OF ITEMS in array? I need it for checkin nr of items
in argv[]

--
RW


From bokr at oz.net  Mon Apr 15 16:25:01 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 20:25:01 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com>
Message-ID: <a9fcut$e9l$0@216.39.172.122>

On Mon, 15 Apr 2002 00:06:34 -0700, "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:

>
>"Martin v. Loewis" <martin at v.loewis.de> wrote in message
>news:m34ridmq00.fsf at mira.informatik.hu-berlin.de...
>
>| That's what I'm trying to tell you all the time: the
>| '>>>' operator is meaningless if it is defined as
>| "fill in zeroes". How does it know where to start
>| inserting zeroes?
>
>No, it isn't meaningless:
>
>The 'long' class has an internal representation for the
>long number.  The number of bytes that are currently
>being used to contain the number are known (internally)
>to the object.  The zeroes get shifted into the high bit
>of the number, no matter how many bytes are used to
>contain the number.
>
Actually the longs may not be byte-based at all. They could
be based on an array of shorts with 15-bit values. You don't
want to see the actual bit pattern of that, do you? At least,
not as a general purpose thing.

>An int type simply shifts a zero into bit 31, so that
>this works the same as it will on a long.
>
>The '>>' operator should also work the same for a long
>as it does for an int.
It can't, because the long does not represent a fixed-width
bit pattern, it represents an infinite bit pattern. The infinite
comes from extending the sign bit. Thus shifting in zeroes at
the 'top' of a negative number implies shifting them in at bit
position infinity. You can say that we don't actually use an
infinite number of bits in the internal representations, but
the/a point of numeric unification is to hide that representation.
All you can legitimately refer to is the implied abstraction, which
has indefinitely extended sign bits.

If you do choose some top bit in the indefinite extension as 'the'
sign bit, it should be done consistently, and what you are doing
should be clear. For example, if you scanned down from the left until
you found two adjacent differing bits (or a single bitif you reach the
bottom) you could say the ms bit of those was 'the' sign bit. Now you have
a width, and this can be computed for any integer, positive or negative.

On the basis of that width, you could shift in zeroes. I'm not sure how
useful that is, but I would say it would be more consistent that picking
an arbitraryly occuring k*4 or k*8 or k*15 bit width value that you might
get from the current state of an internal representation.

In any case, you can define a function to whatever you actually want.
(I suspect you'll wind up specifying width explicitly). But if
you want a >>> operation, what width should it use? Remember, different
platforms may use different internal long representations.

>
>The '<<' operator should also work the same for a long
>as it does for an int.
>
That one's easy ;-)

>The 'hex()' method should work the same for a long as
>it does for and int.
>
I think that's agreed too, just not the details. I agree with you that
sign/magnitude is not very useful if you're interested in bits. (I might
be referring to different bits than you though).

>Why is this so confusing to you?  Java programmers have
>been putting the difference between the '>>' and the
>'>>>' operators to good use for years!
>
I doubt whether Martin is confused, but he may have a different perspective.

Regards,
Bengt Richter


From lac at strakt.com  Wed Apr 24 00:39:04 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 24 Apr 2002 06:39:04 +0200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...) 
In-Reply-To: Message from Greg Ewing <greg@cosc.canterbury.ac.nz> 
   of "Wed, 24 Apr 2002 14:51:54 +1200." <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> 
References: <mailman.1019526833.9592.python-list@python.org>  <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> 
Message-ID: <200204240439.g3O4d4dj030129@ratthing-b246.strakt.com>

> Andrew McNamara wrote:
> > 
> > >Interersting.  If so, then RMS is 'guilty' of an almost humorous
> > >contradiction.
> > 
> > Speculation - maybe he's aware of this and simply sees it as playing
> > what he sees as a fatally flawed system against itself? 8-)

This is most definitely the case.  Plus he is trying to reform it.

> I wonder what he would do if there were no intellectual
> property laws. Would he be campaigning for the introduction
> of a law that enabled him to enforce the price that the
> GNU copyleft exacts?

RMS does not see 'property' in the same way that most people do.  In
particular, he sees intellectual property laws, at least as
exemplified by the copyright acts, as exploiting the creators on
behalf of the publishers, all the while cleverly tricking the creators
into foolishly believing that they benefit.  So. If there were no
copyrights, then there still is contract law -- and RMS would be
writing the GNU contract, not the copyleft.  (Or something.  Speculating 
what RMS would do in some hypothetical world is dangerous -- I've 
been wrong enough times on what he would do in this world, in
the next ten minutes .-))

Laura Creighton




From robin at jessikat.fsnet.co.uk  Fri Apr 26 06:55:48 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 26 Apr 2002 11:55:48 +0100
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>
 <just-9E8624.12281026042002@news1.xs4all.nl>
Message-ID: <QnyV+gA0ITy8EwpR@jessikat.demon.co.uk>

In article <just-9E8624.12281026042002 at news1.xs4all.nl>, Just van Rossum
<just at xs4all.nl> writes
....
>What a coincidence:
>   http://www.python.org/sf/532860
>   http://mail.python.org/pipermail/python-dev/2002-April/023427.html
>
>Just
eerrrhh I'm going away from this. I have no idea what LOAD_NAME is.
-- 
Robin Becker


From jeff at ccvcorp.com  Fri Apr 19 16:29:17 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 13:29:17 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <a9pj29$pdm$1@nntp6.u.washington.edu> <a9pt8a$3nh$1@peabody.colorado.edu>
Message-ID: <MPG.172a1df699b5a9ed9896ba@news.nwlink.com>

In article <a9pt8a$3nh$1 at peabody.colorado.edu>, 
fperez528 at yahoo.com says...
> 
> I don't know how far back (version wise) this works, but I'd rather use:
> 
> for line in sys.stdin.xreadlines():
> 
> which should be pretty good. And I guess if the name lookups aren't optimized 
> away (which I don't know) you can always do
> 
> xread = sys.stdin.xreadlines
> for line in xread():

I don't think this would be a significant savings.  As I 
understand it, the name lookup and function call only happens 
once anyhow, and returns an xreadlines object, which the for-loop 
then iterates over.  Trying to stash the bound method doesn't 
really gain anything if it's only used once.

I'd definitely prefer xreadlines() to readlines(), though.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jorgenhansen at hotmail.com  Tue Apr  9 09:42:11 2002
From: jorgenhansen at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Hansen?=)
Date: 9 Apr 2002 06:42:11 -0700
Subject: Problem with packages and modules
Message-ID: <117880a1.0204090542.258f0f51@posting.google.com>

Hello,

I have been working for some time now with Python and really enjoy it.
Right now I am having serious difficulties crasping several issues
when using packages.


This is an example of directory hierachy:

company/                   Top-level package
    __init__.py            Initialize the company package
    plot/                  A plot package
        __init__.py
        axes.py            Modules in the plot package
        axis.py
        bar.py
        figure.py
        line.py
        plotobject.py
        points.py
        tick.py
        utils.py    
    log.py                 Modules in the company package
    utils.py


If I were to use the plot package I would type in
   >>> from company import plot

and assume that I could access modules as:
   >>> plot.figure.Figure()    # where Figure is a class name.

but this is not the case. Neither of the modules in the plot package
are available, and it doesn't matter if I use the __all__ attribute in
__init__.py and type in all the module names.

Further if I typed in:
   >>> import company

I would assume that I could access all subpackages and their modules
with a dotted name (company.plot.figure.Figure() for instance), but
this is obviously not the case.

Can someone please help me. I don't now what missing out on. I've read
the documentation, Googled quite a lot, but haven't found any answers.

Regards
Jorgen


From brian at sweetapp.com  Fri Apr 12 14:59:56 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 12 Apr 2002 11:59:56 -0700
Subject: How to avoid mutual imports
In-Reply-To: <slrnabeai4.lrg.huaiyu@gauss.almadan.ibm.com>
Message-ID: <004301c1e254$3cf69dd0$445d4540@Dell2>

Huaiyu wrote:
> I need advice on how to handle the following situation:

[long description snipped]

If interface.py only defines data_initialization, then you could move
your imports into it i.e.

def data_initialization(...):
 	from utils1 import a, b, c
 	from utils2 import x, y, z

	# do real work

You could also move data_initialization into its own module (might be a
good idea anyway).

Cheers,
Brian





From pinard at iro.umontreal.ca  Sun Apr 14 19:51:30 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 14 Apr 2002 19:51:30 -0400
Subject: Topy news - and a request for Python pretty-printing
In-Reply-To: <LNBBLJKPBEHFEDALKOLCOEBPOOAA.tim.one@comcast.net>
References: <LNBBLJKPBEHFEDALKOLCOEBPOOAA.tim.one@comcast.net>
Message-ID: <oqn0w5vny5.fsf@carouge.sram.qc.ca>

[Tim Peters]

> [Fran?ois Pinard]
> > ...
> > Would someone happens to know if there is a pretty-printer for Python
> > source _lines_, where I would find some inspiration maybe?

> The closest I've seen is the "block comment" reflowing facilities in
> various Python editors.

Comments are not so bad.  When Topy moves comments over from the
original source to the generated Python code, it does not make them longer.
The comments coming from the `-C' option, which asks for non-comment source
lines to be re-inserted as comments near the corresponding Python code, have
`#LINE: ' added at their beginning, this causes overflow only for lines
which were long already, I consider this bearable.  Diagnostic warnings,
when inserted as comments, currently have a full FILE:LINE:COLUMN reference,
and when FILE is long, overflow may occur.  But I also take this as bearable.

> For very long lines with clear structure (lists, tuples and dict
> constructors; function calls), I suppose you could do a lot worse than to
> feed them to Emacs, injecting a newline after every comma.  Then they'll
> line up the way Guido would type them by hand <wink>.

I am surely not going to call Emacs from Topy :-).  Does Guido use Emacs?
Just curious!  In Emacs Python mode, there are many good ideas which inspired
my own writing style, thanks to those who wrote it.  On the other hand,
some choices are not satisfactory, and I saw cases when Emacs does not
properly reflect structure imbrication -- I do not have examples in head
-- but when this occurs, I prefer rewriting the little affected parts, so
Emacs Python mode behaves more nicely.  It's a give and take, and that's OK!

Of course, I've been interested in the `pprint' module, at least
for structures, but not enough to dive and extend it to statements and
expressions.  The main point is that `pprint' works on in-memory structures,
while I need to beautify surface, textual Python.  Or almost: at the time
Topy needs beautifying, the output surface line has already been generated
as a list of textual fragments, which only need to be conceptually joined
and written out.  For now, I merely kludged something quick to decide
how to split these fragments between continued lines, driven by simple
heuristics looking at the contents of each fragment (is it a string?
how many characters?  any parentheses?  any initial or trailing space?).

To do real beauty work, I should change the Topy back-end and many parts in
front-ends to generate Python structures instead of textual fragments, add
a lot more specs to the tree language between front-ends and the back-end,
and probably add one more pass in the back-end, or nearly.  I'll postpone all
this, as I do not want to be overly distracted by pretty-printing now, and
there are much more pressing things to address before Topy exists for real.

Speaking of, I started to retrofit my previous uncompleted `perl2py'
as a new Topy front-end.  Currently, Topy uses SPARK, which I find easy
and satisfying.  But for Perl, I had two problems with SPARK: it was
inordinately slow (I might not be writing good grammars), and I did not
find how to express the assymetrical precedence for Perl named operators
(they bind lousely to their left, and rather aggressively to their right).
I still have more retrofitting work to do before being really ready for
revisiting these two areas, though.

For Scheme and Milou, the two languages which Topy is likely to handle
first, I tried something which might have been fruitful speedwise, I guess,
I'm not sure.  Once the scanner has produced tokens for the whole source,
I have a quick and simple pass which segments tokens into capsules, a
capsule merely represents a single definition or expression at the outer
level of the source.  Then, I call the parser on each capsule in turn.
I have the prejudice that I help the parser by giving it smaller chunks
than the whole source.  Moreover, since the number of parse calls is known
before we begin them all, Topy may even entertain the user with progress
counters :-).  Topy collect all the parse results together, as this is
useful in subsequent passes.  Those are speedy enough over the whole thing.

The Topy grammar for Perl is significantly bigger, so I'm not sure what
will happen, even if I try to implement special parser help like above.
Another possibility would be that I consider PLY instead of SPARK, or
maybe use both at once depending on the front-end, yet I would expect a
few intricacies at the nitty-gritty level.  I did not really look at PLY
yet, but I've been told that if I like SPARK as a programmer (and I do!),
PLY might please me as well, as PLY reuses many of SPARK good ideas from
the programmers's viewpoint.  PLY grammars are weaker than SPARK's, but
they might be much faster on Perl.  Sigh!  One sure thing is that I like
SPARK generality, elegance, and unobtrusiveness.

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





From wilk at flibuste.net  Sun Apr  7 07:41:59 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Sun, 7 Apr 2002 13:41:59 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1018161735.24484.python-list@python.org>
Message-ID: <20020407134159.3680634c.wilk@flibuste.net>

Le Sun, 07 Apr 2002 01:41:12 -0500
Tim Peters <tim.one at comcast.net> ?crivait:

> > ...
> > The proposal of having separate "development" and "stable" release
> > tracks like Linux has sounds like a promising direction for keeping
> > both camps (language development enthusiasts and would-be production
> > users) happier than they are in the current situation.
> 
> If we had more active Python developers, that might be feasible.
how many their are ?

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From jb at cascade-sys.com  Mon Apr 29 12:51:15 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 09:51:15 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <eUTmiIA3MQz8EwSA@jessikat.fsnet.co.uk>
Message-ID: <3CCD7A03.8A17CE41@cascade-sys.com>

Robin Becker wrote:

> map, filter & reduce et al are the fastest ways to do many things....,
> pow and divmod are also extremely useful. I don't use input, but I'm
> sure others do.
>
> Ther are no fast equivalents to the functional forms, list
> comprehensions are as slow as for loops. We need more fast operations
> for functional style programming, not less.
>
> This PEP smacks of political requirements not practical ones. The
> builtin name space is only crowded if we want to use the names for
> something else.
>
> Again this is a backward incompatible change. It's not broken and
> doesn't need fixing.

Hear! Hear!

I don't see a problem that this change solves and it actually detracts
from the language, as Robin so eloquently puts.

Furthermore, divmod() and pow() are natural and common enough and have a
direct and familiar relation with inline operators that they reasonably
deserves to be builtins.

Furthermore the change invalidates a lot of the books published to date
for Python, for no good reason.

I gotta say, for a group who ostensibly is looking out for newbies, I
submit that the number of language changes in the last 12 months may
constitute the largest single barrier to understanding the language.  It's
a moving target.  A number of people HERE have said they can't keep up
with the latest.

I think we should be very reluctant about changes that break old code.
Further we should focus on the important, substantial ones instead of
dinking around with trivial crap like this.

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From gimbo at ftech.net  Fri Apr  5 10:34:58 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 5 Apr 2002 16:34:58 +0100
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? [LONG!])
In-Reply-To: <a8kdde$12t$1@panix1.panix.com>
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com> <20020405014732.GB17563@isis.gerrietts.net> <mailman.1017996733.25069.python-list@python.org> <a8kdde$12t$1@panix1.panix.com>
Message-ID: <20020405153458.GC3249@andy.tynant.ftech.net>

On Fri, Apr 05, 2002 at 09:47:10AM -0500, Aahz wrote:
> In article <mailman.1017996733.25069.python-list at python.org>,
> Andy Gimblett  <gimbo at ftech.net> wrote:
> >
> >Really?  I've used Singleton a fair bit (probably too much in fact)
> >for stuff like configuration information.  I'm curious as to how else
> >I could/should do this.
> 
> Use a module.  Don't use a class unless you need a class.

/me closes his eyes tight, opens his mouth, thinks about it for a
while, shaking his head around slightly, feels something click, says
"wow", opens his eyes, says "wow" again, and hits reply.

Wow.  Man, I love python.

Cheers for the epiphany

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From marklists at mceahern.com  Tue Apr 30 21:56:54 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 20:56:54 -0500
Subject: Newbie: Working with large database -- help!
In-Reply-To: <WEHz8.1942$z65.211132@news.uswest.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEIPKNAA.marklists@mceahern.com>

[Owen Marshall]
> I am trying to work with a file -- for lack of better terms, a
> database. The data is like this:
>
> data|data|data|data

Short answer:  split() -- a method of the string object -- is your friend.

Demo:

>>> s = "a|b|c|d|e"
>>> delim = "|"
>>> a = s.split(delim)
>>> a
['a', 'b', 'c', 'd', 'e']

What if one of the fields itself has a pipe in it?  Presumably, the writer
escapes those into something other than a pipe.

Cheers,

// m





From phr-n2002a at nightsong.com  Wed Apr  3 19:11:28 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 16:11:28 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017795008.700.python-list@python.org> <hyAq8.29178$n4.5840455@newsc.telia.net> <mailman.1017873324.18169.python-list@python.org>
Message-ID: <7xhemsqq4v.fsf@ruckus.brouhaha.com>

Richard Jones <rjones at ekit-inc.com> writes:
> What's the value in this statement? Nowhere does it state that it is 
> mandatory that python programmers write their docstrings in ReST. Only that 
> there will be tools available for those that do.

Well, nothing is mandatory without legislation by congress, but I
thought the idea of structured doc strings was to have cool tools that
use them, and for those to be useful, the doc string format has to be
widely used and accepted.  If you just want a few
especially-interested people writing doc strings that way, you don't
need a PEP--just start using whatever format you want.


From h_schneider at marketmix.com  Wed Apr 10 12:39:31 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Wed, 10 Apr 2002 18:39:31 +0200
Subject: SOLVED: PIL + McMillan Installer
References: <a91f3r$4o7$06$1@news.t-online.com> <k4Zs8.244$p56.67402@newsb.telia.net>
Message-ID: <a91pi2$m66$01$1@news.t-online.com>

Thanks! It works!

"Fredrik Lundh" <fredrik at pythonware.com> schrieb im Newsbeitrag
news:k4Zs8.244$p56.67402 at newsb.telia.net...
> Harald Schneider wrote:
>
> > I use Activestate Python 2.1 and McMillan Installer 5.
> > A script that uses PIL runs fine. But the one-file-distribution seems to
> > have a problem with Image.open()
> >
> > ----
> > img = Image.open(File)
> >   File "C:\usr\python\PIL\Image.py", line 902, in open
> >     raise IOError, "cannot identify image file"
> > IOError: cannot identify image file
> > ----
>
> PIL's plugin scanner won't find the plugins when they're
> embedded in the EXE file.  there's probably some installer
> recipe somewhere, but if not, the explicit-import approach
> described in this article might help:
>
>
http://www.pythonware.com/products/pil/articles/speeding-up-initialization.h
tm
>
> </F>
>
> <!-- (the eff-bot guide to) the python standard library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> -->
>
>




From tim.one at comcast.net  Mon Apr 22 00:47:22 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 22 Apr 2002 00:47:22 -0400
Subject: A simply newbie question about ndiff
In-Reply-To: <3cc38db1$0$15477$afc38c87@news.optusnet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEPDOPAA.tim.one@comcast.net>

[Neville Franks]
> One thing I've noticed now is for larger files (73,000 lines)
> ndiff can take several minutes on my 1G PC, but that is another issue
> again. I assume the fact that it is being interpreted has quite a bit
> to do with this.

Not so.  I've been reimplementing ndiff since the early 80's in a variety of
languages, and the Python version is by far the fastest, due to using fancy
algorithms that were just too painful to code in lower-level languages.
ndiff is expensive because it tries very hard to produce an "intuitive"
diff.  If you don't need that kind of diff, there are any number of diff
variants that run much faster.

ndiff can be incredibly expensive when it's faced with large blocks of
mismatching lines (meaning the blocks have no lines at all in common).  Then
it does a full cross-product search, for each line in one block computing a
similarity score between it and every line in the other block.  The
best-matching pair is used as a synch point, and then the whole business is
done twice again for the mismatcking sub-blocks on each side of the synch
pair.  To keep memory use finite, it doesn't even try to remember the
similarity computations across iterations of this, and does a ton of
redundant work instead.  In the end, that can take time cubic in the number
of mismatching lines.

So, like most tools, it works best when it isn't really needed <wink>.





From kseehof at neuralintegrator.com  Tue Apr  9 03:25:24 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 09 Apr 2002 00:25:24 -0700
Subject: Embedding Python with C++
In-Reply-To: <GYts8.26982$2J2.6830478@typhoon.southeast.rr.com>
Message-ID: <EFEFIHAINIINNELLHFAIIEEECAAA.kseehof@neuralintegrator.com>

Chris Stoy wrote:
> Hey all,
>
> I'm looking to embed Python as a scripting language in a game I am working
> on.  I want to be able to call C++ methods from Python, and call
> Python from
> C++ (for the usual code quick in Python, convert to C++ for speed later.)
>
> I've looked into using the Boost.Python code and I'm fine with that.
>
> My problem is, I have no clue how to compile Python into a static linked
> library for Windows.  Actually, I need it to be cross-platform, which is
> another issue.
>
> I've never embedded Python before, and I'm fairly new to Python, so can
> someone please give me a pointer to how to do this?  What are the files I
> need to compile to build the static lib (using MS Visual Studio
> 6.0?) Also,
> I don't need all of the standard Python components (like networking.)  Can
> those be removed from the lib?

Oh yeah, I forgot to mention PyGame.  Check it out:
http://www.pygame.org/

- Ken Seehof





From sholden at holdenweb.com  Mon Apr  8 15:21:36 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 15:21:36 -0400
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk> <slrnaave1r.13h.jon+usenet@snowy.squish.net>
Message-ID: <N6ms8.1279$vT6.636@atlpnn01.usenetserver.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnaave1r.13h.jon+usenet at snowy.squish.net...
> In article <40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk>,
>   Alastair Nicol wrote:
> > 1) Can only hosts which can see the interface your program is bound to
> > communicate with your program. In the "normal" case if you bind to
> > 127.0.0.1 for example ONLY your machine can see that interface. Other
> > machines cannot connect to it. Is this a safe / secure assumtion? (not
just
> > for localhost)
>
> No. Other machines on your local net can talk to your localhost by
> sending ethernet packets with your machine's MAC address and 127.0.0.1
> as the IP address.

Good point.

My understanding is that this "feature" depends on the particualr stack and
OS you are using, and is not guaranteed to work. It is, of course, a
perversion of the intended use of the loopback network, although one we do
well to remember.

regards
 Steve





From guido at python.org  Wed Apr  3 13:56:52 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 13:56:52 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 12:06:31 CST."
             <Pine.LNX.4.33.0204031118110.16190-100000@server1.lfw.org> 
References: <Pine.LNX.4.33.0204031118110.16190-100000@server1.lfw.org> 
Message-ID: <200204031856.g33Iuqr22751@odiug.zope.com>

> It's the *speaking*.  When i want to say true or false, then there's
> the dilemma.
> 
> I know, your answer is "you should always just say True or False",
> and Mark McEahern said the same thing.  But this cannot be so in
> practice: *everything* already returns 0 or 1.  (It may be possible to
> get around this if we commit to changing the entire standard library
> before releasing a version of Python with True and False, but alas,
> this is not the only issue... read on.)
> 
> As long as True and False are somewhere represented as 0 and 1,
> the values 0 and 1 will never lose legitimacy as booleans.  This
> business with str() and/or repr() producing "0" or "1" for backwards
> compatibility prevents us from considering 0 and 1 relegated to a
> truly non-boolean status.

But that's a variant of the PEP that no-one except Marc-Andre has
spoken in favor of.  The PEP proposes str(True) == "True".

> Consider this:
> 
>     >>> a = [0, False, 1, True]
> 
>     >>> print a
>     [0, 0, 1, 1]
> 
>     >>> for x in a: print x
>     0
>     0
>     1
>     1
> 
> Good heavens!

It will print

    0
    False
    1
    True

> What about this:
> 
>     >>> d = {}
>     >>> d[0] = 'zero'
>     >>> d[False] = 'false'
>     >>> len(d)
>     1 or 2?

What about this:

    >>> d = {}
    >>> d[0] = 'int'
    >>> d[0.0] = 'float'
    >>> d[0j] = 'complex'
    >>> print len(d)
    1 or 3?

False and True are numbers, and they are equal (==) to 0 and 1;
everything else follows from there.

> Basically, the concept of having a permanently schizophrenic type
> in the language scares me.  The above shows, i believe, that a
> reasonable implementation must print True as True and False as False,
> and never mention 1 or 0.

And this is what the PEP proposes (despite its brief mention of an
alternative).

> Moreover, as soon as you start sorting a
> bag of objects, or keying dictionaries on objects, you are forced to
> run into the distinction between 0 and False, and between 1 and True.

No you're not.  d[0] and d[False] retrieve the same value, as do
d[0L], d[0.0], and d[0j].

> I'm not against the idea of booleans, of course -- but i do think
> that halfway booleans are worse than what we have now.  And getting
> to real booleans [*] involves real pain; it's just a question of
> whether that pain is worth it.  Even if we get all the way there --
> as in we manage to convert enough code and convince everyone to use
> the new style -- i will never ever want "and" and "or" to return
> booleans (i just hope that doesn't confuse anyone).
> 
> 
> -- ?!ng
> 
> [*] By real booleans, i mean the following.  (Booleans would have
>     to behave like this for me to consider them "good enough" to
>     be better than what we have now.)
> 
>     >>> False, repr(False), str(False)
>     (False, 'False', 'False')
>     >>> True, repr(False), str(False)
>     (True, 'False', 'False')
>     >>> False + True
>     TypeError...

That's just one textbook idea of what a Boolean "should" be.

>     >>> False == None
>     0
>     >>> False == 0
>     0
>     >>> True == 1
>     0
>     >>> {0: 0, False: False, 1: 1, True: True}
>     {0: 0, False: False, 1: 1, True: True}
> 
> ... and probably
> 
>     >>> None < False < True < 0
>     True
> 
> (Hee hee -- i suppose the fact that "boolean" starts with a "b"
> gets us this for free.  But i wonder how many people are going
> to be puzzled by True < 0?)

Yuck.

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




From mhammond at skippinet.com.au  Sun Apr 14 02:21:16 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 14 Apr 2002 06:21:16 GMT
Subject: Some troubles with COMObject of the ClearCaseAutomation Library
 4.1 (ccauto.dll)
References: <3CB85F52.40000@snafu.de>
Message-ID: <3CB92021.1030208@skippinet.com.au>

Andreas Maurer wrote:
> Hi all,
> 
> I've written a little script using the ClearCase Automation Library 4.1
> with python 2.1.2
> 
> When I'm trying to get the collection of HistoryRecords of one
> CC-element this error message appears:
> ------------------
> 
>  >>> hist = Elem.HistoryRecords(0)
> 
> Traceback (most recent call last):
>   File 
> "C:\win32app\Python21\lib\site-packages\win32com\client\dynamic.py", 
> line 279, in _make_method_
> codeObject = compile(methodCode, "<COMObject %s>" % self._username_,"exec")
>   File "<string>", line 1
>      def HistoryRecords(self, pICCBranchType=pythoncom.Missing, 
> Since=<time object at 013BB240>, User='', Minor=0, 
> ExcludeCheckOutEvents=0, Recurse=0, DirectoryOnly=0):

This is a bug in the Python COM package.  If a COM method takes a DATE 
param with a default value, the makepy generated .py file is invalid.

I have fixed this locally.  Mail me for the new .py files.  Otherwise, 
just remove the invalid default value from the generated .py file and 
always the param.

Mark.



From dreed at capital.edu  Sat Apr 13 18:21:52 2002
From: dreed at capital.edu (Dave Reed)
Date: Sat, 13 Apr 2002 18:21:52 -0400
Subject: porting python to embeded devices
In-Reply-To: <7xr8lk46nl.fsf@ruckus.brouhaha.com> (message from Paul Rubin on
	12 Apr 2002 20:30:22 -0700)
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com> <slrnabf5r2.s6b.grey@teleute.dmiyu.org> <7xr8lk46nl.fsf@ruckus.brouhaha.com>
Message-ID: <200204132221.g3DMLqT07469@localhost.localdomain>

> From: Paul Rubin <phr-n2002a at nightsong.com>
> 
> Steve Lamb <grey at despair.dmiyu.org> writes:
> > > The SL5500 is a high powered computer with a 32-bit CPU and 64 MB of
> > > ram.  It runs Python just fine.  
> > 
> >     *Eyes Sharp's list of retailers for the SL-5500 a little closer.*
> > Re-e-eaaaalllly.  Hmmmm..
> 
> The SL-5500 hardware is similar to an Ipaq 36/37xx plus it has a cute
> little slide-out keyboard that's not all that useful.  I got the
> developer version for $199.00 which was a good deal at the price.  I
> wouldn't recommend it at $500 though.  I'll post a review to my website soon.


I stopped at Best Buy today and looked at one. The thing that
interests me is the ability to program it in an environment I'm used
to (Linux, Python, C/C++). I can't justify $500 right now (plus I'd
want to hook it into my LAN so there's another $120 and the extra
memory), but I do have a use for a Zaurus at one of the places I
consult. I'd like to develop the program (using Python) I want to run
on the Zaurus before I buy one (figure by the time I get done they'll
probably be cheaper :-).

I downloaded qtopia-free-1.5.0-1.i386.rpm and have been playing with
it a little on my Linux machine. I see there are python packages for
the Strong Arm processor, but I'm interested in running Python inside
the qtopia environment so I need python packages for it that are for the
i386. Can anyone tell me how to do create those or if they exist
already?

Or could I just write the program using Python and PyQt with a 240x320
window and then when I'm done, get a Zaurus, the python Strong Arm
packages, and copy my program to the Zaurus and it will work?

Thanks,
Dave







From martin at v.loewis.de  Sun Apr 14 05:30:14 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 11:30:14 +0200
Subject: Failure when profiling XML processing
References: <pan.2002.04.13.20.51.16.512445.4898@mindspringdot.com>
Message-ID: <m3lmbqwrtl.fsf@mira.informatik.hu-berlin.de>

Jyrinx <jyrinxat at mindspringdot.com> writes:

> (Is there a more specific place where I should be asking about libraries?)

The XML-SIG would be the appropriate place. From a glance at your
report, I'd agree that there is a bug somewhere, but it is not obvious
to me what the bug is, and it looks like anybody approaching this bug
would need a lot of time to investigate it. So it is questionable
whether posting a formal bug report somewhere would advance getting it
fixed.

Regards,
Martin


From jimmy at retzlaff.com  Sun Apr 14 08:34:45 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Sun, 14 Apr 2002 05:34:45 -0700
Subject: Python 2.2 unwise.exe
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F5640@MAYOR.retzlaff.com>

Tim Peters agonized:
> Crud, I was hoping it was unique to Win9x.

It's gone on my development workstation (WinXP Pro) but still on my
server (Win2K Server). The most obvious difference in usage between the
two is that I'm often installing/uninstalling/reinstalling various
extensions (Win32all, Numeric, wxPython, PIL, mx, etc.) on my WinXP
machine. Any chance some package is also using Wise and treating
unwise.exe as its own? It appears Win32all 146 uses Wise but it has its
own unwisepw32.exe in site-packages. Has it always been so careful? That
suggests protection if this is the problem (e.g., timscoolunwise.exe).

Jimmy




From fperez528 at yahoo.com  Mon Apr 22 13:39:24 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 22 Apr 2002 11:39:24 -0600
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com>
Message-ID: <aa1hvn$o9i$1@peabody.colorado.edu>

Jaco Schieke wrote:

>> I was going to ask the same thing as I'm inclined to think this is a bug:
>> readline has '[' and ']' as part of its internal list of word delimiters,
>> so when you do
>> a[0].<tab>
>> it breaks the word at ']' and sees a bare '.', hence the completions you
>> get. So far so good.
>> 
>> BUT: if you change the list of delimiters (with readline.set_delimiters())
>> to remove '[' and ']', the problem persists. That's the part I think to be
>> a bug (I could be wrong, but I'd like to understand why if that's the
>> case). At this point it shouldn't break at ']' anymore, and the completion
>> should be done on the full 'a[0]'. I put print statements in the
>> rlcompleter code to see this and indeed, readline _is_ breaking on ']' even
>> after removing the character from the list of delimiters.
>> 
>> Unfortunately the completion code is in C and I don't have the sources
>> around, so I haven't had the chance to take a look at this.
>> 
>> I'd love to hear from someone who knows the code as to this behavior really
>> is a bug or not before shooting off to SF and looking like an idiot :)
>> 
>> Cheers,
>> 
>> f.
> 
> 
> Looks like the regexp in rlcompleter that searches the string does not
> work correctly.  After changing the line
> 
>> m = re.match(r"(\w+(\.\w+)*)\.(\w*)", text)
> 
> to
> 
>> m = re.match(r"([\w\[\]]+(\.[\w\[\]]+)*)\.(\w*)", text)
> 
> it matched on the brackets too and could get to the objects.  I also
> had to change the delimiters to exclude [].

Nice! I didn't think to look there as I assumed thiss part was happening in 
the C part of readline. Do you care to submit a bug report with this one-line 
patch to SourceForge? I'd really like to have rlcompleter in 2.3 have this 
fixed (it already has a patch I sent in earlier for another problem) That way 
I could stop shipping a separate patched version with my own project.

If you can't/won't do it let me know and I can send the patch in.

Cheers,

f.


From shalehperry at attbi.com  Tue Apr  2 11:25:44 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 02 Apr 2002 08:25:44 -0800 (PST)
Subject: User-Agent, version and other mysteries
In-Reply-To: <fce95053.0204020622.3c1f2f27@posting.google.com>
Message-ID: <XFMail.20020402082544.shalehperry@attbi.com>

On 02-Apr-2002 Stefan Faltz wrote:
> I try to change the "User-Agent" header, i tried it first by something
> like
> 
> urllib.FancyURLopener.version = "Whatever Agent you like"
> 
> that didn't work, what a surprise
> 
> then i ripped some code, and i think it is reasonable
> 
> class AppURLopener(urllib.FancyURLopener):
>     def __init__(self, *args):
>         apply(urllib.FancyURLopener.__init__, (self,) + args)
>         self.version = "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT
> 5.0; T312461; Q312461)"
> 
> urllib._urlopener = AppURLopener
> 

from the library reference:

For example, applications may want to specify a different `User-Agent' header
than `URLopener' defines.  This can be accomplished with the following code:

     class AppURLopener(urllib.FancyURLopener):
         def __init__(self, *args):
             self.version = "App/1.7"
             urllib.FancyURLopener.__init__(self, *args)

     urllib._urlopener = AppURLopener()

The key here is urllib._urlopener takes a constructed class instance, not the
name of a class.  Also notice how the parent's constructor is called.



From oyving+news at pvv.org  Wed Apr 24 16:57:20 2002
From: oyving+news at pvv.org (Oyvind Gronnesby)
Date: 24 Apr 2002 22:57:20 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com>
Message-ID: <87k7qwrf0v.fsf@pvv.org>

* Cimarron Taylor
|
| Will this do what you want?
| 
| verboseGender = ('female','male')[gender=='m']

Wouldn't this break in a PEP 285 Python world?

-- 
?yvind Gr?nnesby
have castle, need dragon


From cliechti at gmx.net  Tue Apr  9 15:59:53 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 9 Apr 2002 20:59:53 +0100
Subject: Sending a signal to a process started by another user
References: <3CB31C2C.9010503@olen.to>
Message-ID: <Xns91EBD5CCB3489cliechtigmxnet@62.2.16.82>

Joonas Paalasmaa <joonas at olen.to> wrote in news:3CB31C2C.9010503 at olen.to:

> How can I send a signal to a process started by another user. In this 
> case my intention is to send a signal 14 to a process started by user jp 
> from a cgi script. The script raises an "Operation not permitted"- 

i hope so - for security's sake...

> OSError perhaps because the script is run by user "apache" and the 
> process by user "jp". How can In make this work properly?

if you can change the other script so that it communicates through say 
sockets or so or run it under the same UID as the webserver.
or write another script that uses sockets or whatever to talk to the CGI 
and let this run as 'jp' so that it can send a signal to the other script.

chris


-- 
Chris <cliechti at gmx.net>



From dale at riverhall.NOTHANKS.co.uk  Tue Apr 23 05:12:48 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 23 Apr 2002 10:12:48 +0100
Subject: Encryption
References: <70h8cugniao40148h76fv4o7bgci2j3ddd@4ax.com> <7x662j76iu.fsf@ruckus.brouhaha.com>
Message-ID: <359acu8n2qode4gnnmh38rkekp204dss3n@4ax.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote:

>Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> writes:
>> Mcrypt can use a number of encryption techniques and I'm not too
>> bothered which we use. Just as long as I can decypher it at our end.
>> 
>> Any suggestions?
>
>Why do you especially want to use mcrypt, in that case?

I don't. The site supplying the data want to use mcrypt. We just want
to accomodate. 
--
Dale Strickland-Clark
Riverhall Systems Ltd



From altis at semi-retired.com  Thu Apr 18 22:37:20 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Thu, 18 Apr 2002 19:37:20 -0700
Subject: [ANN] PythonCard 0.6.5
Message-ID: <zoLv8.1597$Xz.253924@news.uswest.net>

This release includes installation instructions and walkthroughs to get you
started with PythonCard, distutils support, and the layout editor
(resourceEditor sample) has been enhanced to edit dialogs. New components in
0.6.5: Gauge, StaticBox, ComboBox, Spinner, RightTextField, and Calendar

All the information you need about PythonCard can be found on the project
web page at:
http://pythoncard.sourceforge.net/

The installation instructions and walkthroughs are available on the main
documentation page:
http://pythoncard.sourceforge.net/documentation.html

You can download the latest release at:
http://sourceforge.net/project/showfiles.php?group_id=19015

For a list of some of the samples that have been built with PythonCard and
screenshots of them in action go to:
http://pythoncard.sourceforge.net/samples.html
A description of each sample is included in the readme.txt file in each
sample directory.

The kind people at SourceForge host the project:
http://sourceforge.net/projects/pythoncard/

If you want to get involved the main contact point is the Mailing list:
http://lists.sourceforge.net/lists/listinfo/pythoncard-users

PythonCard requires Python 2.1.x or later and wxPython 2.3.2.1 or later and
runs on every platform those are available for. wxPython can be downloaded
at http://www.wxpython.org/

ka
---
Kevin Altis
altis at semi-retired.com




From hak at iki.fi  Fri Apr 12 14:49:12 2002
From: hak at iki.fi (hankoifi)
Date: 12 Apr 2002 11:49:12 -0700
Subject: python raw socket / ping
Message-ID: <c9b1b5f8.0204121049.3bc6e949@posting.google.com>

Hi
I have been searching for raw socket examples for python.
Main idea is to build something like pathchar or clink with python
(activepython 2.1 , win2k and/or linux). For these i mainly need ICMP
but i also have to incrementally decrease/increase TTL-field in IP
header (like in traceroute). Found these:

1)http://www.python.org/~jeremy/python.html
  Very nice packet, includes ping, traceroute etc.
  However:  made with version 1.4 and doesn't work straight way.

2)ftp://ftp.visi.com/users/mdc/ping.py
  Only ping, but works straight with v2.1

3) couple of other ping variants using os.popen()
 ping = os.popen('ping -c 3 -i 1 ' + host + ' 2>&1')
 data = ping.read()

Number 3) is probably the easiest way to do it, but still
I wonder if somebody has already made revisions for jeremy's packet
for activepython v2.1 or otherwise got it working wit win2k

- hankoifi


From martin at v.loewis.de  Mon Apr  1 03:38:55 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 10:38:55 +0200
Subject: multithreading and Tkinter
References: <uok7s6qqph.fsf@bashful.cdf.toronto.edu> <m3pu1yrz2w.fsf@mira.informatik.hu-berlin.de> <Xns91E2A4FE99EA8JamesLLugojcom@216.148.53.85>
Message-ID: <m3u1qvrexs.fsf@mira.informatik.hu-berlin.de>

James Logajan <JamesL at Lugoj.com> writes:

> Has anyone ever tried using after_idle in a worker thread as a way to 
> signal a Tkinter mainloop thread? I'm wondering if this proposal is 
> really thread safe, since technically a worker thread is making a call to 
> Tk (even if it is nothing more than registering a callback) while the 
> mainloop thread may also be in the process of making calls to Tk.

That cannot happen. _tkinter maintains a Tcl lock, guaranteeing that
only one Python thread at any time calls into Tcl. 

You may wonder how you then can call in to Tcl while another Python
thread calls tk_mainloop: tk_mainloop does not call a Tcl loop, but
calls Tcl_DoOneEvent; if there was no event, it sleeps for 20ms
(giving up the Tcl lock during that time). As a result, other threads
can call into Tcl while Tcl is idle. If Tcl is active, those other
threads will block in the Tcl lock.

People have been complaining about the presence of the Tcl lock, and
the Sleep call (which means that there might be a 20ms delay between
an event and the time when processing the even starts). You are
encouraged to study the source of _tkinter.c, both to get trust in my
statements, and to suggest improvements to the status quo.

HTH,
Martin


From joost_jacob at hotmail.com  Mon Apr 29 10:32:42 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 29 Apr 2002 07:32:42 -0700
Subject: module for printing all the methods in a class, with inheritance
Message-ID: <13285ea2.0204290632.47eb212e@posting.google.com>

Sometimes when using a class from a 3rd party library I wonder what
all the methods are.  Sometimes you need list of all the methods for
a class, not just the ones the class defines but also the ones it
inherits.  Module showclasses.py below tries to do just that, and
shows where the methods are inherited from.  

I have a Windows batchfile in c:\WINNT containing the line:
python -c "import showclasses; showclasses.dumpmodule('%1')"
so i can do
C:\mydirectory\anywhere> showclasses SocketServer
when i am interested in the methods in all the classes available
after you import the 'SocketServer' module.
A *n.x batchfile would contain the same line, but with '$1' instead
of '%1'.
Of course you can also just dump all the methods in a class, for
example when you are unsure about names when coding, use dump(ClassName).

The showclasses module is not perfect but sometimes it helps me.  It
could be easily extended to also print attributes but i learned that
often manipulating attributes is evil.


Questions:

* If you do dumpmodule('copy'), copy is a standard python module, you
will see that the Error class declared in the copy module cannot be
dumped.  Also you will see a class with the name C that is declared
locally in the _test function, it should not be listed.  How to
improve this?

* Now with the new 2.2 Python version classes is this still a good
approach?  Unfortunately i do not have the 2.2 version myself.  Any
solution should be backwards compatible to python 1.5.2.  But i do
like the new developments in Python, it's just that i have to use
1.5.2 too often. 

* Please let me know if you see improvements or if it does not work
with your Python installation.


Sourcecode:

------------------  showclasses.py  ---------------------------------
import sys, pyclbr, string

def dumpmethods(c, classname=None):
    methods = c.methods.items()
    methods.sort(lambda a, b: cmp(a[1], b[1]))
    for method, lineno in methods:
        print "  def " + method,
        if classname:
            print " "*(20-len(method)),
            print "(" + classname + ")",
        print
    if classname and c.super:
        for s in c.super: dumpmethods(s, classname=s.name)


def dump(c):    # adapted from Fredrik Lundh: Python Standard Library
    try:
        s = "class " + c.name   # class header
        if c.super:
            s = s + "(" + string.join(
              map(lambda v: v.name, c.super), ", ") + ")"
        print s + ":"
        # print method names, sorted by line number
        dumpmethods(c)
        # print inherited method names, sorted by line number
        for s in c.super: dumpmethods(s, classname=s.name)
    except AttributeError:
        print 'Cannot dump the class', c
    print

def dumpmodule(m):
    try:
        mod = pyclbr.readmodule(m)
    except ImportError:
        print 'No module named', m, 'in your Python installation',
        sys.exit(1)
    for k, v in mod.items():
        dump(v)

if __name__ == '__main__':
    if len(sys.argv) < 2:
        print "Usage: $ python "+sys.argv[0]+" module_name"
        sys.exit(1)
    dumpmodule(sys.argv[1])
---------------------------------------------------------------------

Joost Jacob
www.liacs.nl/home/jjacob/

I was gratified to be able to answer promptly, and I did.  I said I
didn't know.
                -- Mark Twain

<groundcontrol-to-redhat--please-upgrade-2-2-2-'ly-yrs>


From lac at strakt.com  Mon Apr 29 05:13:21 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 29 Apr 2002 11:13:21 +0200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...) 
In-Reply-To: Message from "Stephen J. Turnbull" <stephen@xemacs.org> 
   of "29 Apr 2002 11:40:05 +0900." <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> 
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org>  <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> 
Message-ID: <200204290913.g3T9DMkG026104@ratthing-b246.strakt.com>

> >>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
> 
>     Laura> People here seem to be thinking that the GPL is there to
>     Laura> protect creators rights to their software.
> 
> I'm pretty sure that that is what rms has in mind, as long as both
> "creators" and "software" are inclusive collective nouns.
> 
> The consumer protection argument is the OSI line (a special case
> thereof), and rms don't have much truck with anything so economical
> and relativist.

1. Why did you remove my prior line, 'as a thought experiment?'  I put
   it in in order to avoid having this conversation right now?  This
   is very rude.

2. I learned this from RMS himself, when he was living in my house
   about 15 years ago.  Look up 'hoarder' in his writings.  Many of
   those writings were getting composed at about that time, and in
   part come from arguments John Gilmore and I were having with
   him.  RMS called me a software hoarder, because I was writing
   closed-software for pay.  I took great offense because he was
   eating food purchased with this pay.  Much argument ensued.
   I thought that everybody should have the right to do whatever
   they wanted with the property they created, open source it,
   close source it, whatever -- they made it, it is _theirs_.  I
   wanted people to open source whenever possible out of decency
   and kindness, but I did not want to compell them to open source
   their software.  I thought compelling people was a worse sin
   than software hoarding.

   RMS does not see things this way at all.  He thinks that software
   hoarding is a worse sin than compelling, indeed he wants a world
   where the idea of 'having the right to close source your program'
   is greeted with the same sort of horror as 'having the right to
   sell your children into slavery'.  He doesn't want software to be
   property in the sense that it is now.

   I want the same world.  RMS and I are only fighting about the
   implementation details, and the philosophical underpinnings.  (But
   that is _pleanty_.  It takes about 45 seconds for RMS and I to be
   in the same room for an argument to start.  But since we rarely see
   each other, this is not a problem.)

   And I am now with RMS -- I think software hoarding is so vile that
   I want laws preventing it.  But until we have such laws, everybody
   has to do whatever their own consciences dictate.

3. I'm glad you recognised the OSI line, and even the rhetoric.  Eric
   Raymond showed up at my house the week after.  We discussed what
   Eric is now calling 'sustainable public rhetoric' a _lot_.  I
   thought that if businesses were too stupid to already know that
   'the tools you use to run your business -- being able to fix them
   when they break -- good idea' then educating them was a waste of
   time.  But I was perfectly willing to help Eric come up with
   rhetoric -- I didn't think that RMS's vision was going to happen
   until the hypothetical Socialist World Government nationalised the
   entire labour force, something which I most devoutly wanted to
   prevent.  I thought that it would take something like that for it
   to happen -- but then I expected the GPL to be tossed out of an
   American court on a legal challenge more than a decade ago, as
   well.

   My judgement about American courts and American businesses has an
   inperfect track record.  Remember this when I discuss what I know of
   them.

4. I am not in the channelling RMS business, even the channelling RMS
   of a decade ago business.  If anybody out there wants to find out
   what RMS thinks of something -- go volunteer for something and then
   ask him.  I guarantee you will learn something.

Laura
   




From nospam at nospam.nospam  Wed Apr  3 07:32:10 2002
From: nospam at nospam.nospam (neil)
Date: Wed, 3 Apr 2002 13:32:10 +0100
Subject: Newbie: issuing shell commands with mod_python
Message-ID: <lFCq8.906$d45.95879@monolith.news.easynet.net>

Hi. I have mod_python installed on a linux, apache server and have got some
of the basic examples from www.modpython.org working ok.

I have built a web form that takes some input and sends it to a python
script. The python script is receiving the posted variables correctly as I
can send them back to the browser no problems.

What I need to do now is issue a command to the shell with the python
script. I have the following code

import os
def main(req):
    child = os.popen("/usr/bin/htpasswd -cb /usr/local/apache/users/myfile
name1 passwd1")
    data = child.read()
    err = child.close()
    return data

When I hit this script through a browser I get an internal server error
returned to the browser. There is nothing in my error logs and the following
in the access log
[03/Apr/2002:12:28:23 +0000] "GET /cgi/publisher/makepassword.py/main
HTTP/1.0" 500 596

I am a bit stuck at present as I dont really know which way to go forward.
Could it be something to do with my sys.path settings? If so how do I go
about changing them? Can monitor the shell in the same way as you can with
tail -f on logfiles?

Is this even the right way to issues commands against the shell? Should I bu
running a cgi instead of using mod_python?

Any help/direction/tutorials will be gratefull accepted.

neil






From kevin at cazabon.com  Wed Apr  3 22:59:18 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 3 Apr 2002 19:59:18 -0800
Subject: Buffer Interface, can't get PyObject_AsReadBuffer to work?
References: <5a4226f0.0204022145.49010b2e@posting.google.com> <m3sn6dntiy.fsf@mira.informatik.hu-berlin.de>
Message-ID: <5a4226f0.0204031959.5cd0febe@posting.google.com>

Thanks Martin, you're right of course.  The documentation is
incorrect... I've submitted a bug report to SourceForge.  Annoying,
heh?

Back-on-the-right-track-ly yours,
Kevin.

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3sn6dntiy.fsf at mira.informatik.hu-berlin.de>...
> kevin at cazabon.com (Kevin Cazabon) writes:
> 
> >   // Convert the pInBuffer from PyObject to a PyBuffer (readable),
> > THIS DOESN'T WORK (the Python object type that is passed in as
> > inBuffer is a 'B' array)
> >   if (PyObject_CheckReadBuffer(inBuffer)) {
> >     //this ALWAYS returns 0, meaning unsuccessful... what's wrong???
> >     ok = PyObject_AsReadBuffer(inBuffer, &pInBuffer, &p4size);
> 
> 0 means success, -1 means error. Use the source, Luke.
> 
> HTH,
> Martin


From starx at pacbell.net  Wed Apr 17 14:49:55 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 17 Apr 2002 18:49:55 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net>
Message-ID: <Xns91F37887977AFRASXnewsDFE1@130.133.1.4>

Andrew Dalke || Wed 17 Apr 2002 11:20:15a:

> Philipp Lenssen:
>>can I use something like
>>    verboseGender = (gender == 'm') ? 'male' : 'female'
>>in Python?
> 
> No.  Search the archives for "ternary operator" for plenty of
> past discussion.  Google lists 65 posts for comp.lang.python.
> 
> Besides, for this specific case you should have some error checking,
> because there are genders other than 'male' and 'female' (eg,
> asexual bacteria and hermaphradite earthworms).
> 
>   verboseGenderTable = {
>      "m": "male",
>      "f": "female", <          # Most large eukaryotes
    	"a": "asexual",           # Bacteria, etc
    	"h": "hermaphrodite",     # Earthworms, etc
    	"n": "neuter",            # Robots, AIs
    	"r": "refuse-to-stateist",# The paranoid
    	"u": "unknown"            # We lack information /
    	                          # last seen in dark ally /
                                 # raised by wolves, dosen't know
>   }
>   verboseGender = verboseGenderTable[gender]



-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From starx at pacbell.net  Sun Apr 21 14:25:42 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 21 Apr 2002 18:25:42 GMT
Subject: c++ vs python!
References: <3CC2D8D7.20601@student.kun.nl> <a9utab$e18$1@panix3.panix.com>
Message-ID: <Xns91F7747347BC5RASXnewsDFE1@130.133.1.4>

Aahz || Sun 21 Apr 2002 10:36:11a:

> In article <3CC2D8D7.20601 at student.kun.nl>,
> husam  <h.jehadalwan at student.kun.nl> wrote:
>>
>>If we compare c++ with python, which one is faster when it comes to
>>2D, 3D graphics and math. And which one is faster when it comes to
>>design and implementation of "heavy" applications?
> 
> Use Python to call C/C++ code for compute-intensive tasks.  See the
> Python Game Library and NumPy for examples.

Better yet, just use Numpy and Pygame and you'll probably have enough
speed without putting any of your code into a lower language than
python. PyOpenGL too. Python is great for the 'stand on the shoulders of
giants' / 'let other people do the hard stuff for you' (depending on
perspective =) way of working. 

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From xoltar at starship.python.net  Thu Apr 18 19:59:29 2002
From: xoltar at starship.python.net (Bryn Keller)
Date: 18 Apr 2002 16:59:29 -0700
Subject: Memory problem (possibly involves interpreter)
Message-ID: <6df8badb.0204181559.46a1c841@posting.google.com>

Hi folks,

I have an application which uses unexpectedly large amounts of memory,
and then doesn't free as much of it as it should. It's a GUI
application, and when I show a new (instance of a) window, the memory
usage as reported by Task Manager (this is NT4/sp6a) climbs by a
certain amount, say 340K. Close the window and the memory drops by
108K. Now, I'd understand that this is just garbage collection at
work, or a memory leak in my code, and wouldn't bother everyone on the
list with it, if it weren't for these facts:

1. These numbers stay roughly the same no matter how many times you
open and close the window. So opening and closing the same (class of)
window 100 times took me from 24356K to 67456K. This would seem to
indicate there is memory that really isn't being reclaimed.

2. I've verified (print statements in __del__ methods) that all the
windows and all their controls are getting garbage collected properly,
and the underlying C++ objects seem to be getting deleted also.
There's not a whole lot else involved with a window besides the
controls, certainly not enough to account for 2/3 of the memory usage.

3. In the simplest case (a window with just two controls), it takes
44K to show the window, and closing it reclaims nothing at all. Every
time.

4. Most importantly, the numbers change depending on the version of
Python. In 2.1.0, more memory was expended than described in #2, but
about the same amount was reclaimed in large windows. For the small
window, 2.1.0 showed the behavior in #3 above, while in 2.1.3, the
small window takes 36K to show. Closing it still doesn't release any
memory, *but* the second time the window is shown it will take *zero*
memory, which is what I'd hope for with GC. For larger windows there's
still a big problem, as #1 (figures for 2.1.3) indicates.

The interpreter would normally be the last thing I'd suspect, but
since the behavior changes significantly from version to version of
Python, I'm beginning to wonder if that's the culprit. Can anybody
give suggestions? 2.2.1 has the same results as 2.1.3. Is there any
way to determine whether the leak is in the interpreter's GC or my
Python code or a 3rd-party C module? Any tools available to analyze
the contents of memory?

Thanks,

Bryn


From ods at fep.ru  Thu Apr 25 12:02:07 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 25 Apr 2002 20:02:07 +0400 (MSD)
Subject: urllib.urlopen("file://...") fails
In-Reply-To: <m2d6wn7q85.fsf@jtan.com>
Message-ID: <Pine.LNX.4.44L.0204252001100.2654-100000@ods.pp.ru>

On Thu, 25 Apr 2002, Seth Delackner wrote:

SD> Fernando P?rez <fperez528 at yahoo.com> writes:
SD>
SD> > Seth Delackner wrote:
SD> >>>>> u = urllib.urlopen("file:///t:/tmp/package.html")
SD> >                                 ^
SD> >
SD> > One / too many? You're directing it to \t:\.... which it
SD> obviously
SD> > can't find.
SD> >
SD> > I don't run windows so I can't test so I may be wrong. It
SD> just looks
SD> > like that is your problem.
SD>
SD> You would think, right?  But I just tried
SD> 'file://t:/tmp/package.html'

I guess it should be something like 'file://t|/tmp/package.html'.





From richard at NOstarfighterSPAM.freeuk.com  Wed Apr 17 13:50:54 2002
From: richard at NOstarfighterSPAM.freeuk.com (Richard Townsend)
Date: Wed, 17 Apr 2002 18:50:54 +0100
Subject: Making a C extension compatible with 2.2
References: <915a998f.0204170914.66aee851@posting.google.com>
Message-ID: <1019065831.5459.0@eos.uk.clara.net>

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> wrote in message
news:915a998f.0204170914.66aee851 at posting.google.com...
> Must anything be done to make a C extension compatible with Python
> 2.2? I have an extension (not my own) that ran fine under Python 2.1
> on Solaris 2.6. I have since recompiled it for Python 2.2, but
> sometimes I now get tuple indexing errors - and not necessarily at a
> point in my Python code where the loaded extension is actually being
> called upon. Running the same code under Python 2.1 is OK. Would I be
> right in thinking it may be related to the changes in Python 2.2 for
> the new iterator protocol?
>

I found a very similar problem with the ingmod package (Python interface to
Ingres DB) on HP-UX 11.

It works fine with Python 2.1 but in 2.2 it breaks the list class's ability
to be used in a for loop (giving the strange error message 'IndexError:
tuple index out of range').

I've not been able to identify where it's failing though :-(










From paul at boddie.net  Mon Apr 29 04:51:26 2002
From: paul at boddie.net (Paul Boddie)
Date: 29 Apr 2002 01:51:26 -0700
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net>
Message-ID: <23891c90.0204290051.3edcc62d@posting.google.com>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote in message news:<slrnacpc38.obc.jon+usenet at snowy.squish.net>...
> In article <aai5oo$99l$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> > Just realized that these Servlet based development models really limits the
> > usage of software like Dreamweaver.  Since all code is in the program these
> > utilities are almost worthless.  Is there a good way to combine Webware with
> > templates.  I think of templates as pages where certain %words% are
> > substituted etc...

In the above, I think you're really referring to *SP (ASP, JSP, PSP)
or "program-in-markup" templating, rather than servlets. Certainly,
Webware doesn't restrict you to using PSP just because it has the
notion of servlets.

> This is exactly why I created jonpy (http://jonpy.sourceforge.net/).
> It puts *no* Python code inside your HTML templates, you simply mark
> sections using comments (e.g. <!--wt:newsitem-->) and replacements
> using dollar signs (e.g. $$title$$). The idea is that even if you are
> suddenly given an entirely new design with new HTML files with none of
> the modifications you made to make the pages work, it will only take a
> few minutes to fix them up again.

As I noted before, DTML started out a bit like this, and there are
other template frameworks which stick quite close to the more
"conservative" forms of templating. Personally, I believe simple
"model-based" (or "placeholder-based") templating offers the most
reward for the least additional complexity (and maintenance),
especially for certain kinds of applications, and that the extra power
offered by less "conservative" template frameworks only gives you a
certain amount of extra flexibility that not all applications really
need.

Paul


From jeff at ccvcorp.com  Fri Apr 12 20:38:26 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 12 Apr 2002 17:38:26 -0700
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca>
Message-ID: <MPG.17211dcc4bd939939896af@news.nwlink.com>

In article <1vKt8.21374$Kq4.896208 at news2.calgary.shaw.ca>, 
bbollenbach at shaw.ca says...
> > This doesn't change the fact, though, that Public Domain 
> > software is Free Software in the FSF  sense of the word.  
> > 
> > (See http://www.fsf.org/philosophy/categories.html)
> 
> Interesting that in the page you cite, they state specifically the
> opposite of what you're saying:

Uh, no they don't.

> 
> <begin gnu>
> Public domain software is [...] a
> special case of non-copylefted free software, [...]

Which *does*, in fact, mean that it is free software.  Unless 
perhaps free software means something other than free software?


> However, I was still correct in saying that just because you release in 
> the public domain doesn't mean you're releasing "free" or even more 
> generally "open source" software, which was mainly the point I was
> trying to convey.

It is indeed open source, in the sense that everyone has a right 
to access and use the source code.  It does *NOT* guarantee the 
right of access to derivative works (which the GPL does), but 
that doesn't mean that it's not open source.

(Where, specifically, are you getting whatever definition of 
"open source" that you're using?  You seem to be *far* more 
restrictive in your use of "open source" and "free software" than 
even the FSF is...)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From vmilitaru at sympatico.ca  Wed Apr  3 03:55:14 2002
From: vmilitaru at sympatico.ca (vio)
Date: Wed, 03 Apr 2002 03:55:14 -0500
Subject: Java to Python translation
Message-ID: <3CAAC372.792A0F9A@sympatico.ca>

Greetings,

I need to translate a (rather voluminous 100MB) java program to python. After a
few searches on google, apparently the best tools for this task are Lex-Yacc
(Flex-Bison on linux). While I am working on the Lex-YACC HOWTO, I would be very
interested in hearing of others' successful experiences/feedback at this. Any
pointers of interest would similarily be appreciated.

Many thanks,
Vio


From chema at imdea.com  Tue Apr 16 15:13:39 2002
From: chema at imdea.com (Jose Manuel Esteve)
Date: 16 Apr 2002 12:13:39 -0700
Subject: pyRXP Memory Likage???
Message-ID: <c5f48de3.0204161113.14919d25@posting.google.com>

I have been playing with pyRXP and is incredibly fast as compared with
minidom, also the memory footprint is smaller for one file, however if
you loop for parsing a lot of xml strings the memory usage gets higher
and higher as compared to minidom

import pyRXP
xml='<start><tag1>text here</tag1></start>'
P=pyRXP.Parser()
for i in range(0,100000):
    tup=P.parse(xml)

will show the behaviour


From Oschler at earthlink.net  Mon Apr  1 01:38:55 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 01:38:55 -0500
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <mailman.1017628505.7732.python-list@python.org> <LURp8.63151$Q7.49997488@e3500-atl1.usenetserver.com> <slrnaafotk.8j8.gerhard@lilith.my-fqdn.de>
Message-ID: <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com>

Gerhard,

Nah, I installed Linux to be my vacation from redmondland, thank you kindly
:)  Richard Jones (other poster on this thread) seems to like Quanta Gold
and says it works well with Zope so I'll check that out.  Mozilla?  SUSE 7.3
Pro installs Netscape 4.67 by default.  What's so special about Mozilla
(blue oyster cult).

What's with all the stuff in your sig?  Especially the last one that looks
like a LISP statement.  I'm feeling a bit of "sig envy".

thx

Gerhard H?ring <gerhard at bigfoot.de> wrote in message
news:slrnaafotk.8j8.gerhard at lilith.my-fqdn.de...
>
> If neither of these meets your requirements, and you have a decently
> powered computer, you could try emulating Windows software like
> Dreamweaver under Linux. Unfortunately, vmWare has gotten quite
> expensive, but there's a much cheaper alternative, of which the name
> currently escapes me, but I think it's in the EUR 100 price range.
>
> Or just get a Mac with OS X instead, then you'll have a Unix and can
> still run Dreamweaver and Photoshop ;-)
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
> public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))





From ljohnson at resgen.com  Wed Apr  3 10:05:57 2002
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 3 Apr 2002 09:05:57 -0600
Subject: Swig/Python: data type problem
References: <9d8d2e93.0204020314.362621a4@posting.google.com> <uajfhvcudhk5cb@corp.supernews.com> <9d8d2e93.0204030510.6bd7dffd@posting.google.com>
Message-ID: <uam664ftg3jj0e@corp.supernews.com>

> Thank you for your answer.
> I have read the swig documentation, but I have still some more questions:
>
> - where I have to put the code:
> %typemap(in) String * {
>     $1 = new String(PyString_AsString($input));
> }
> in the .i file?

This typemap needs to appear somewhere in your SWIG interface file before
the declaration of your "void read(String *namefile)", so that by the time
SWIG processes the function declaration it knows about the typemap.

> - what is $1 and $input ? They are not recognized...

Which version of SWIG are you using? I was writing it using SWIG 1.3
conventions. If you are instead using SWIG 1.1, try this typemap instead:

    %typemap(python, in) String * {
        $target = new String(PyString_AsString($source));
    }

> - is it the only piece of code I must write to solve my data type problem?
> for instance, do I have to write something in the read function?

Adding this typemap should solve the problem of converting a Python string
input argument to one of your library's String objects before calling
read(). Since I have absolutely no other information about your library's
code I can't speculate on what other problems might come up.




From erno-news at erno.iki.fi  Wed Apr 17 06:31:42 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 17 Apr 2002 13:31:42 +0300
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <7xhemb744m.fsf@ruckus.brouhaha.com> <Ms9v8.22165$b62.626694@news1.tin.it>
Message-ID: <kupu0yfwfl.fsf@lasipalatsi.fi>

In article <Ms9v8.22165$b62.626694 at news1.tin.it>, Alex Martelli
<aleax at aleax.it> writes:

| Paul Rubin wrote:
|| "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
||| Is it possible to have a safe eval() on the server-side even when the
||| string consists of submitted user data?
|| 
|| Well, look at the docs for the rexec (restricted execution) and Bastion
|| classes, but basically no, there's no safe way.  Rexec doesn't stop an
|| infinite loop from hosing your cpu.

| If you have to watch for that, forking and running the DoS risk in the
| child process can help -- the parent process needs to keep an eye
| over the resources (mostly, time) being used by the child process and
| be ready to kill it if need be.  Simple, it ain't.

you can use resource limits - then it's pretty simple. both cpu time
memory usage can be restricted using the resource module (remember to
use the hard limit).

it used to be the case that on linux only the address space
and stack limits were respected, i don't know if that's the
case any more. but address space works, unless you want
to mmap big files or somesuch.

  -- erno


From cepl.m at neu.edu  Sat Apr 27 23:46:39 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Sun, 28 Apr 2002 03:46:39 +0000 (UTC)
Subject: IMAP2IMAP Synchronization -- iis [Was: Re: iis: comparsion of two lists]
In-Reply-To: <JTry8.26517$8D3.706486@news1.tin.it>
References: <20020426224924.GA1318@rehor.surfbest.net> <20020426232802.GA1927@rehor.surfbest.net> <JTry8.26517$8D3.706486@news1.tin.it>
Message-ID: <20020428034636.GB1279@rehor.surfbest.net>

On Sat, Apr 27, 2002 at 06:51:21AM +0000, Alex Martelli wrote:
> It would, but the first assignment is useless and the approach
> needlessly complicated IMHO.
> 
> erase = [ m for m in proc_old if m not in loc_old ]

Concerning the first assignement, rather be (slightly) slower 
than sorry and your solution seems to work only with Python 2.*. 
I am still on 1.5.2. Do I have to _really_ download multimegabyte 
beast just to get above command?

Concerning the rest of the script, I have yesterday finalized it
and tried to send it to this NG. Unfortunately, the message
somewhere got screwed up (at least on DejaNews, it is not on my
newserver at all; I have no idea, what's up). So, sorry for
reposting (now the script is in the body of message below the
sig), but I would really like to ask people here on the group for 
review (it is my first non-trivial script in Python and I am 
afraid to trust him with my real email traffic).

Therefore, could I ask somebody for review of the below script, 
especially concerning the safety of messages and speed, please? 


Thank you very much.

Matej


-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
There's a long-standing bug relating to the x86 architecture that
allows you to install Windows.
	-- Matthew D. Fuller

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

#!/usr/bin/env python
# $Id: iis.py,v 0.7 2002/04/27 01:32:52 matej Exp $
"""
IMAP-IMAP Synchronization - iis

(always use lower case to avoid terrible misunderstanding :-)

Copyright (c) 2002 Matej Cepl <matej at ceplovi.cz>
   
Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following
conditions:

The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT.  IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS IN THE SOFTWARE.

"""
import ConfigParser
import imaplib
from sys import stderr, exit
from string import split, join
from os.path import expanduser
from time import time, gmtime

# constants
# return errorlevels
DOWNERR = 1
UPERR = 2
SELECTERR = 3
LISTERR = 4
LOGINERR = 5
CONFIGERR = 10

# other constants
IMAPPORT = "143"
LOCALSTR = 'local'

def chatch_error(errmsg,stat_num):
	stderr.write("IMAP Error: %s !\n" % errmsg)
	exit(stat_num)

class Conifg(ConfigParser.ConfigParser):
	"""
	Object containing configuration.
	
	It reads configuration file (in WIN .ini syntax) like this
	
	[localuser1]
	local = password,last_sync
	server1.com = username,password,last_sync (in IMAP shape)
	
	[localuser2]
	local = password,last_sync
	server2.com = username,password,last_sync (in IMAP shape)
	"""
	def __init__(self,fname=".iisrc"):
		ConfigParser.ConfigParser.__ini__(self)
		self.filename = expanduser("~/"+fname)
		self.options = {}
		try:
			self.read(self.filename)
		except ParsingError,e:
			chatch_error(e,CONFIGERR)
		for sect in self.sections():
			tmp_dict = {}
			for opt in self.options(sect)[:-1]:
				tmp_dict[opt] = split(self.get(sect,opt),',')
			self.options[sect] = tmp_dict

	def close(self)
		try:
			f = open(self.filename,'w')
			for sect in self.options.keys():
				f.write('[' + sect + ']\n')
				for opt in self.options[sect].keys():
					f.write(opt + '=')
					f.write(self.optionsp[sect][opt] + '\n')
			f.close()

def comp_lists(processed,local):
	"""
	Compare two lists of messages and select for download/delete.

	INPUT: two objects Server -- one to be processed and other
	OUTPUT: erase, download -- list of messages to delete from
			  the processed server, and list of messages to be
			  downloaded on other server
	What is newer than self.last and it isn't in other, should be
	downloaded from processed, and what is older and isn't on
	other, should be deleted. The same other way around.

	OK, I admitt, that code below is ugly, but it should be as
	fast as possible -- which is what we need.
	"""
	proc_old = map(lambda x: processed.old_msg[x][0],processed.old_msg.keys())
	proc_new = map(lambda x: processed.new_msg[x][0],processed.new_msg.keys())
	loc_old = map(lambda x: local.old_msg[x][0],local.old_msg.keys())
	loc_new = map(lambda x: local.new_msg[x][0],local.new_msg.keys())
	
	erase = e = download = d = ()
	e = filter(lambda m: return not(m in loc_old), proc_old)
	d = filter(lambda m: return not(m in loc_new), proc_new)

	erase = filter(lambda x: processed.old_msg[x][0] in e, \
	processed.old_msg.keys())
	download = filter(lambda x: processed.new_msg[x][0] in d, \
	processed.new_msg.keys())

	return erase,download

def sync(loc,rem):
	"""
	Synchronize two folders on servers.

	## INPUT:
	## OUTPUT:
	## description
	"""
	# delete old on the remote and download new messages
	delete,down = comp_lists(rem,loc)
	try:
		ok, answer = rem.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Remote messages were not deleted!\n")
		# rather let it be -- it is better not to have
		# something deleted, than to have deleted more than
		# we want
	for msg in down:
		try:
			ok, stuff = rem.fetch(msg,'(RFC822)')
			tmp_msg = rem.new_msg[msg]
			ok, answer = loc.append(loc.curr_folder,tmp_msg[2],tmp_msg[1],stuff)
		except Exception,e:
			catch_error(e,DOWNERR)
	
	# delete old on local and upload new ones
	delete,up = comp_lists(loc,rem)
	try:
		ok, answer = loc.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Local messages were not deleted!\n")
	for msg in up:
		try:
			stuff = loc.fetch(msg,'(RFC822)')
			tmp_msg = loc.old_msg[msg]
			ok, answer = rem.append(rem.curr_folder,tmp_msg[2],tmp_msg[1],stuff)
		except Exception,e:
			catch_error(e,UPERR)

class Server(imaplib.IMAP4):
	"""
	Main class for managing IMAP servers and folders.
	
	## some description
	"""
	def __init__(self,server='',port=IMAPPORT,user,password,last_sync):
		"""
		Connect to the server and login.
		"""
		self.host=server
		self.port=port
		self.user=user
		self.passwd=password
		self.last=last_sync
		self.new_msg={}
		self.old_msg={}
		self.curr_folder=''
		try:
			imaplib.IMAP4.__init__(self,self.host,self.port)
			self.login(self.user,self.passwd)
		except Exception,e:
			catch_error(e,UPERR)

	def message_lists(self,folder="")
		"""
		Fetch list of messages, that are older than self.last and
		newer than that.
		INPUT: correctly opened IMAP server, name of the folder,
		correctly set self.last (in IMAP representation).
		OUTPUT: sets self.new_msg and self.old_msg.
		"""
		if folder=="":
			folder="INBOX"
		try:
			self.select(folder)
		except Exception,e:
			catch_error(e,SELECTERR)
		self.curr_folder=folder
		try:
			ok, stuff = self.search(None,'( UNDELETED SINCE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					else
						raise ValueError, 'The field UID is not fetched!',
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					else
						raise ValueError, 'The field INTERNALDATE is not fetched!',
					ok, header = self.fetch(msg,'(FLAGS)')
					if ok == 'OK':
						flags = split(header[0])[2]
					else
						raise ValueError, 'The FLAGS are not fetched!',
					self.new_msg[msg]=uid,date,flags
			ok, stuff = self.search(None,'( UNDELETED BEFORE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					self.old_msg[msg]=uid,date
		except Exception,e:
			catch_error(e,LISTERR)
		self.close()
	def complete(self):
		"""
		leave server
		"""
		self.logout()

if __name__ == '__main__':
	## General outline
	conf = Config()
	for usr in conf.options.keys():
		opts = conf.options[usr][LOCALSTR]
		local = Server(user=usr,password=opts[0],last_sync=opts[1])
		local.message_lists()
		for svr in conf.options[usr].keys():
			if svr != LOCALSTR:
				ropts = conf.options[usr][svr]
				remote = Server(user=ropts[0],password=ropts[1],last_sync=ropts[2])
				remote.message_lists()
				sync(local,remote)
				sync_time = remote.Time2Internaldate(gmtime(time())
				conf.options[usr][svr][2] = sync_time
		conf.options[usr][LOCALSTR][1] = sync_time
	conf.close()

# vim: set ts=3:


From pinard at iro.umontreal.ca  Thu Apr 11 16:44:13 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 11 Apr 2002 16:44:13 -0400
Subject: Topy news - and a request for Python pretty-printing
Message-ID: <oqadsam0de.fsf@carouge.sram.qc.ca>

Hi, gang!  Just a few news and words about the Topy project.  (For those
who do not know, it is a Scheme to Python conversion tool.)

I gave Topy more hours since I last wrote to you about it.  Besides
correcting a few bugs here and there, I rewrote the scoping mechanics so
to ease later implementation of `let'/`let*'/`letrec' nuances.  The new
mechanics is oriented towards the quick switching between evaluation
contexts.  Nothing difficult theoretically, only tedious in details.

For strange reasons, I've been side-tracked into writing another Topy
front-end for Milou (a language which is 10 years old, but fully unknown
outside the company where I work, here).  Taking this opportunity, I
identified more common code and widened the back-end (yielding shortened
front-ends for both Scheme and Milou).  I also cleaning around a bit more.

Something that bothers me a bit is that the Python generated out of Scheme
input, at times, goes over 79 columns.  But for Milou, this happens very
frequently, to the point I cannot escape anymore doing something about it.
Consequently, I quickly kludged a line wrapper in the back-end which,
without being so bad, is still far from being a real pretty-printer.

Would someone happens to know if there is a pretty-printer for Python
source _lines_, where I would find some inspiration maybe?  I do not need
an indenter, there is no problem there.  The problem is to split a long
Python line into many lines which would nicely represent the original,
that is, for example, for which indentation stresses operator precedence.
Cleverly breaking long string constants is also an interesting problem!

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





From johnroth at ameritech.net  Tue Apr 16 21:32:56 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 16 Apr 2002 21:32:56 -0400
Subject: Newbie problem: global variables
References: <a9i1uk$1is$1@narses.hrz.tu-chemnitz.de>
Message-ID: <ubpk7eidausdb0@news.supernews.com>

"Mirko Opitz" <opm at hrz.tu-chemnitz.de> wrote in message
news:a9i1uk$1is$1 at narses.hrz.tu-chemnitz.de...
> Hi,
>
> I've a small problem I need a variable which is present in all
subfunctions
> an also in the main prog but I don't want do give it in the header of
all
> the functions. Is there anyway to declaire it as Public or global like
in
> other languages?
>
> Thx for now
>
> Mirko

In general, no. If you simply want to use it, then assigning
to it once at the module level is sufficient. If you want to
assign to it in various functions, you have to tell Python
that it isn't local, by using a global statement in the
function.

However, you can sidestep the issue by defining
an accessor function. That is, you can define
a function whose sole purpose is to set the variable.
That can be accessed globally, so you don't have to
declare the variable in each function where you want
to modify it.

John Roth
>
>




From starx at pacbell.net  Sun Apr  7 16:30:26 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 7 Apr 2002 20:30:26 GMT
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com>
Message-ID: <Xns91E989929996CRASXnewsDFE1@130.133.1.4>

Aahz || Sun 07 Apr 2002 01:02:01p:

> In article <d9ea2696.0204071156.4284b1f1 at posting.google.com>,
> distrex <distrex at hotmail.com> wrote:
>>
>>I'm trying my first script with a module and I can't get it to work
>>right.  I'm using the time module to make part of the game random and
>>i just want the number to be at a whole number (no decimals) and I
>>can't figure out how to make it do that. If someone needs me to post
>>the script I'd be happy to but if you can help me just from memory
>>that'd be great too. Thanx in advance.
> 
> Use the random module.

Yep. For example:

---
import random

# No vaule - uses time on it's own
random.seed()

orcs_to_place = random.randrange(12,30)

def percent_chance( percentage ):
    	"Assumes percentage as number 0-100"
    	if random.random() > (percentage / 100.0):
    	    	return 1
    	else:
    	    	return 0

def die_roll( number, sides ):
    	total = 0
    	for not_used in number:
    	    	total += random.rand_range(1, sides)
    	return total
---

This is all just off the top of my head, but i've made functions like 
this before.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From nhodgson at bigpond.net.au  Sun Apr 14 18:59:44 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sun, 14 Apr 2002 22:59:44 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <ubhlb81tfqo23e@news.supernews.com> <slrnabjh3i.741.philh@comuno.freeserve.co.uk>
Message-ID: <ARnu8.7539$o66.24404@news-server.bigpond.net.au>

phil hunt:

> How will it break code to introduce a new operator? Any existing
> program that uses xor will be a syntax error, and broken already.

   Current code:

xor = 1

   will break. Operators are keywords, so

or = 1

   produces:

Traceback (  File "<interactive input>", line 1
    or = 1
     ^
SyntaxError: invalid syntax

   Neil





From jeff at ccvcorp.com  Fri Apr  5 14:49:16 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 5 Apr 2002 11:49:16 -0800
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl>
Message-ID: <MPG.17179f847bf41a6998969b@news.nwlink.com>

In article <a8kn23$ikq$1 at zeus.polsl.gliwice.pl>, 
maug at poczta.onet.pl says...
> 
> class A:
>     a = 1
>     def f():
>         print a
>     f()
> Traceback (most recent call last):
[...]
> NameError: global name 'a' is not defined
> 
> def f1():
>     a = 1
>     def f2():
>         print a
>     f2()
> f1()
[...]
> 1

The difference here is because of the class statement, it's not 
using nested scopes in the sense that you're thinking.  In class 
A, the attribute a is a class attribute, not a name in an 
enclosing scope, so the nested scope rules don't apply.  You can 
access it by referring to self.a, however -- that will find both 
instance and class attributes.  

>>> class A:
... 	a = 1
... 	def f(self):
... 		print self.a
... 	
>>> obj = A()
>>> obj.f()
1
>>>

However, for your desired purpose -- a singleton "class" with no 
instances -- you're probably better off putting your code in a 
separate module, and importing the module wherever needed-- as 
noted in another thread today, a module is already a singleton.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jvanasco at hotmail.com  Sat Apr 20 03:40:12 2002
From: jvanasco at hotmail.com (Jonathan Vanasco)
Date: Sat, 20 Apr 2002 07:40:12 +0000
Subject: Mass Mailing help needed
Message-ID: <F5VvYGog71qPQNE3r0Y00023017@hotmail.com>

Hi.. this is my first post, so be kind...

I'm working on a freelance project right now, and there's a lot of db data 
analysis that wasn't working too well in perl.  I heard about python, made a 
couple of benchmarks, and found it working about 3-8x faster than perl.  
That made me happy.

Until I got to the mass-mailing part of the gig.  Based on user registration 
opt-ins, they can receive certain mailings.  This all works fine when it was 
done in perl.  I make a flat text file for each mailinglist with an address 
per line.  Then i used Mail::Bulkmail to handle sending it

But I migrated the analysis files over to python -- and users will 
essentially be mailed results of the analysis files.  Rather than call a 
python script to generate the messages and a perl script to mail them, i'd 
rather do it all at once.  In python.

Which brings me to my problem: for the life of me, i can't figure out how to 
mass email in python.  (note: i'm a designer, not really a programmer).  I 
can't find any python equivalent to perl's Mail::Bulkmail (which, among 
other things does some pimp stuff with the envelopes).  I waded through all 
of the files in Mailman,  and it seemed to rely on using Sendmail for every 
message.  That seems fine if users stay under 5,000 -- but I dont know what 
that will do to postfix should i need to mail 50,000 people (will that fill 
up the disk w/tempfiles, will it skew  the mailqueue too much?)  I thought  
about sending groups of 100 to sendmail as a ,  delimted list in the bcc 
field.  Then i read in a dozen places 'never use bcc'.  Which brings me 
right back to square one -- clueless.

So I'm hoping that someone here can help me out, or point me in a direction 
so I can get this all done in python, and  not  have to run 2 scripts.



_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com





From vinay_sajip at yahoo.co.uk  Thu Apr 11 14:04:12 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: Thu, 11 Apr 2002 19:04:12 +0100
Subject: ANN: Logging Module v0.4.2 released
References: <Pine.LNX.4.44L.0204111947391.13708-100000@ods.pp.ru>
Message-ID: <003101c1e183$4b0579e0$652b6992@alpha>

Denis> I've tried to use logging in a module used by CGI script and ...
Denis> it now dies with misterious message: unexpected end of header.
Denis> The reason is the message 'No handlers could be found for logger
Denis> "qq"' that goes to stdout.  I guess logging should not lo
Denis> messages by default and certanly it must not send warnings to
Denis> stdout.

Yes, this is the one-off message I mentioned earlier. It's only printed
once. It's a bug that it's sent to sys.stdout, it should be sys.stderr. Will
be fixed in 0.4.3.

This message is only there because otherwise there's no other way of
indicating that there's a problem with the logging configuration. No other
messages are printed, even if exceptions occur during logging.

Regards

Vinay


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From logiplexsoftware at earthlink.net  Thu Apr 25 15:26:37 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 25 Apr 2002 12:26:37 -0700
Subject: ActiveX events in wxPython not working
In-Reply-To: <6ddcff55.0204250625.1992ce1b@posting.google.com>
References: <6ddcff55.0204250625.1992ce1b@posting.google.com>
Message-ID: <20020425122637.014976cb.logiplexsoftware@earthlink.net>

On 25 Apr 2002 07:25:36 -0700
Perry wrote:

> Hi,
> 
> I'm using wxPython as an ActiveX (GUI) container. The OCX draws a
> gantt
> chart from supplied data. Everything works fine, appart events.
> 
> Here is the code:

[snip]

> class MyGanttPanel(wxPanel):
>   def __init__(self, parent):
>      wxPanel.__init__(self, parent, -1)
> 
>      sizer = wxBoxSizer(wxVERTICAL)
>         
>      GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
>      gantt.VcGantt = GanttClass(self, -1, wxPoint(0,0), style=wxMAXIMIZE)
       ^^^^^ Where is this instantiated?

[snip]

I don't see where the object "gantt" is being created.  You can't add an
attribute to a non-existent object:

Python 2.2 (#1, Mar 15 2002, 11:52:42) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-81)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> f.a = 1 
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: name 'f' is not defined
>>> class foo: pass
... 
>>> f = foo()
>>> f.a = 1
>>> 

Are you running the program from a shell or by double-clicking its icon?  It
may be that there are errors you aren't seeing if you aren't using a shell. 
Other than that, I'm afraid my ActiveX knowledge is (happily) non-existent, so
if this isn't your problem then I'm afraid I can't help much.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From syver-en+usenet at online.no  Sat Apr  6 14:42:21 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sat, 06 Apr 2002 19:42:21 GMT
Subject: New classes in 2.2
Message-ID: <uvgb4lip0.fsf@online.no>

I have to admit that I am excited about the revisions to pythons
object model. I hope that the current state of affairs where there are
two kinds of classes will eventually fade away, as it seems quite ugly
at the time being (wouldn't want to be a new to Python just now). It
would be sad if Python we're to develop along the lines as C++, where
there seems to be no obvious way to do even the simplest of things (in
my view, even though I think it's a fascinating language, my main
language in fact until I discovered Python).

I've started looking more closely at the new style classes (object
derived). And it struck me as strange that in Guido's Unifying types
and classes in Python 2.2, the new super function and proper
classmethods are discussed. I soon found out that super would (in
addition to ordinary methods) only work for class data members and
static functions, and not for class methods were they in my view would
be most elegant.

I am sure someone has got a better version of super for class methods
than I can present here, but I insert it below anyway as a motivation
to improve or post a link to a better implementation. 

(Shamelessly stolen from guido's Super class in his article, and
blindly hacked till it did what I wanted it to do then and there :-):

class ClassMethodSuper(cls):
    def __init__(self, cls):
        self.__cls__ = cls
    def __getattr__(self, attr):
        starttype = self.__cls__
        mro = iter(starttype.__mro__)
        for cls in mro:
            if cls is self.__cls__:
                break
        for cls in mro:
            if attr in cls.__dict__:
                x = Command(x.im_func, self.__cls__)
                return x
        raise AttributeError, attr


class Command(object):
    def __init__(self, func, *args, **kwargs):
        self.func = func
        self.args = args
        self.kw = kwargs

    def __call__(self, *args, **kwargs):
        args = self.args + args
        kwargs.update(self.kw)
        self.func(*args, **kwargs)


-- 

Vennlig hilsen 

Syver Enstad


From gerhard at bigfoot.de  Wed Apr 24 05:56:22 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 24 Apr 2002 09:56:22 GMT
Subject: linux console graphics
References: <3CC682A5.3050905@ifrance.com>
Message-ID: <slrnacd07s.eo.gerhard@developer2.opus-gmbh.net>

In article <3CC682A5.3050905 at ifrance.com>, eav wrote:
> I am currently working on a project needing a graphical interface, 
> supposed to run quickly on quite modest PC (first pentium generation) 
> with linux or a *bsd.
> 
> What can Python do form me ? I've been considering projects such as 
> pygame - but it seems that SDL can't be run from the console. Am I wrong?
> 
> I've read about a GGI wrapper, but it doesn't seem to be maintained...
> 
> What do you think ? Do I rather use C and svgalib or GGI ?
 
My recommendation: Use a light X configuration with a very lightweight Window
Manager (icewm, blackbox, fluxbox, etc.) There are tons of them here:
http://freshmeat.net/browse/56/?topic_id=56 

Then use whatever GUI library you want on X.  This will save you a lot of
trouble with poor console graphics libraries.

It's a myth that X is heavyweight - it can be trimmed down a lot. Hey, I can
even run it on my handheld :-)

Gerhard


From gashton at cmedltd.com  Tue Apr 23 11:21:14 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 23 Apr 2002 16:21:14 +0100
Subject: Mass Mailing help needed (Paul Rubin)
In-Reply-To: <slrnacarct.3s6.akuchlin@ute.mems-exchange.org>
References: <mailman.1019325819.29016.python-list@python.org>
	<MPG.172e025d803fc2009896bf@news.nwlink.com>
	<mailman.1019553582.18769.python-list@python.org> 
	<slrnacarct.3s6.akuchlin@ute.mems-exchange.org>
Message-ID: <1019575274.4103.0.camel@tp6>

On Tue, 2002-04-23 at 15:20, A.M. Kuchling wrote:

> It's certainly possible to send multiple messages over a single SMTP
> connection, 

I must confess to not being that well up on the Python library with
regard to email messages. Are there any python modules that support
this?

-- 
Graham Ashton





From starx at pacbell.net  Mon Apr 22 21:39:04 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 23 Apr 2002 01:39:04 GMT
Subject: Classes: Old or New and How?
Message-ID: <Xns91F8BDF3CF983RASXnewsDFE1@130.133.1.4>

    	I've been using py2.2 for a while now, although i haven't really 
used many of the new features aside from the odd list compt to ease 
initing 2-3 dimensional lists, and some class hierarchy i made off list 
some months ago. (No generators or anything like that).
    	What i'm wondering is this: when i define some random class, 
weather it's something small (initially) like a Link_Node (datastructures 
class =) or a GL_Color, or something way comlicated like ARCS_FOV or a 
game world controller, in anycase they don't require any special 
functionality -- just data and regular methods. When i start a class like 
this, should I bother explicetly deriving from Object, or should I set 
that flag that makes everything like that, or should I just not bother?

Thanks.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From aleax at aleax.it  Tue Apr 16 13:42:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 19:42:37 +0200
Subject: python threading
In-Reply-To: <x7elhfpmvc.fsf@ra.iago.org>
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <8MVu8.26112$vF6.828988@news2.tin.it> <x7elhfpmvc.fsf@ra.iago.org>
Message-ID: <E16xWz4-0004oN-00@mail.python.org>

On Tuesday 16 April 2002 07:35 pm, Justin Sheehy wrote:
	...
> I've wrapped Queues with Events before, and it makes things a little
> nicer.  It is quite straightforward to make a Queue that exposes a
> get method that will simply block until an Event is set by another
> thread or an optional timeout has elapsed.
>
> It doesn't solve the whole problem, but waiting on an Event is much
> nicer than a get/sleep loop.

Yes, but a GUI thread can't do that either -- nor can a network one;
they must be ready for user-input events and netword-data-arriving
events too.  The problem is how to seamlessly merge waiting for
THOSE kinds of things and for Python-thread-stuff kinds...


Alex




From ajs at ix.netcom.com  Tue Apr  2 19:47:09 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 2 Apr 2002 19:47:09 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <001601c1daa9$16d572a0$0334fea9@carol>

Alex writes -

>It's best for what one's teaching to BE simple, rather than to >try to make
it APPEAR simple by hiding complications.

God bless and Amen.

Efforts to further gloss Python's air of simplicity are going
too, too far.

We've accomplished that Python can be taught while
hiding the existence of numeric types.

If the "teaching community" demands more of Python -
I do suspect Guido faces pressures - let them find something better for
their purposes.

Of course they won't.

Its called calling a bluff.

Art




From rasmussn at lanl.gov  Thu Apr  4 18:07:13 2002
From: rasmussn at lanl.gov (rasmussn at lanl.gov)
Date: Thu, 4 Apr 2002 16:07:13 -0700
Subject: Py_BuildValue() and C++ arrays
In-Reply-To: <3cac101d$0$25808$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <B2DFA0D2-4820-11D6-8148-003065D61546@lanl.gov>

On Thursday, April 4, 2002, at 01:37 AM, Matthias Stern wrote:

> I want to return the values "numElem" (int), "unit" (char[250]) and the
> array "data" (float,10000).
> What is the correct syntax for "Py_BuildValue( ??? )" ?
>
> The first two parameters work e.g. this way:
> ---
>      :
>      Py_BuildValue("(si)", unit, numElem);
>      :
> ---


I've used something like this before a tuple filled with an array:

    PyObject* tuple = PyTuple_New(count);

    for (int i = 0; i < count; i++) {
        double rVal = data[i];
        PyTuple_SetItem(tuple, i, Py_BuildValue("d", rVal));
    }

    return tuple;


Cheers,
Craig





From andy47 at halfcooked.com  Mon Apr  8 01:00:53 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Mon, 8 Apr 2002 05:00:53 +0000 (UTC)
Subject: Need help with distutils
References: <mailman.1018083739.10368.python-list@python.org>
Message-ID: <Xns91EA988AA8260andy47halfcookedcom@203.109.252.31>

Martin Sj?gren <martin at strakt.com> wrote in 
news:mailman.1018083739.10368.python-list at python.org:

> Hello. I need some help with using distutils. When I run
>    python setup.py sdist
> all my .cvsignore files are included, and I don't see a reason for them 
to
> be included in the source dist, so I want to get rid of them
> automagically. I tried using 'prune' in MANIFEST.in, but that didn't 
work,
> so I tried 'recursive-exclude . .cvsignore' but that didn't seem to work
> either (warning: no previously-included files matching '.cvsignore' found
> under directory '.')
> 
> Am I attacking this the wrong way?
> 
> Any help appreciated.
> 
> 
> Regards,
> Martin Sj?gren
> 

The way I've done this for PythonCard (http://pythoncard.sourceforge.net/) 
is to put the following line in my MANIFEST.in file;

global-exclude *cvsignore

If you take a look at the documentation 
(http://www.python.org/doc/current/dist/sdist-cmd.html#sdist-cmd) you will 
notice that the pattern you pass to global-exclude has to be a Unix-style 
"glob" pattern. In that notation '.' is a special character so using 
'.cvsignore' doesn't actually match your files ;-)

Also note that this isn't properly implemented for distutils on Windows or 
Mac. If you try and package a source distribution on Windows, for instance, 
you will get all of your CVS directories in the .zip file as well.

Regards,
Andy
-- 
Contents free posts a speciality


From mgilfix at eecs.tufts.edu  Thu Apr 18 12:34:03 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 18 Apr 2002 12:34:03 -0400
Subject: help splitting strings
In-Reply-To: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com>; from tony@cases.demon.co.uk on Thu, Apr 18, 2002 at 04:35:41PM +0000
References: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com>
Message-ID: <20020418123403.E3177@eecs.tufts.edu>

  You definitely want to use the re module and finall
here.  That way you can have tighter control over what
constitutes a field.  The difficulty here is crafting your
regular expression so it matches what you want. Check out
http://www.python.org/doc/current/lib/re-syntax.html for a description
of the regular expression syntax.

  You'll probably want to code something like this:

   text = # line to parse

   field_re = re.compile (r' <expression here> ')
   match_obj = field_re.finall (text)

   if match_obj is not None:
     my_fields = list (match_obj.groups ())

  As for the regular expression, you'll probably want a look ahead to
check if there are commas within strings:

  something like:

       expr = r'(".*,.*")?,"

  I doubt that'll work but it'll give you an idea anyway :)

            -- Mike

On Thu, Apr 18 @ 16:35, Tony Case wrote:
> Hi all
> 	I'm pretty new to Python which probably explains my problems,
> I'm trying to read data from a number of comma separated variable
> files.  These are fairly large, 5000 lines plus.
> 	I'm reading the files and using split to separate the
> individual values, some numeric some string.  The problem is that a
> number of the lines have commas and double quotes embeded in the
> strings so there are more elements in the list returned from split
> than there should be, in one file there are 400 + lines like this so
> it's not easy to edit the original file.
> 	Is there a programatic way of dealing with this?
> The input files are in the form :-
> int,int,int,"string","string",int,"string"
> etc.
> I tried using the re.findall and re.split but didn't get very far.
> 	TIA	T Case
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (tony)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From jon+usenet at unequivocal.co.uk  Fri Apr  5 08:51:44 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Fri, 05 Apr 2002 13:51:44 -0000
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk> <23891c90.0204050156.1840d913@posting.google.com> <a8k5v2$8kp$1@hfc.pacific.net.hk>
Message-ID: <slrnaaravf.7u2.jon+usenet@snowy.squish.net>

In article <a8k5v2$8kp$1 at hfc.pacific.net.hk>, Scrumpy wrote:
> I do like to separate style, layout, app and data and will avoid using PSP, 
> Cheetah etc. completely or at least in ways which integrate app code within 
> HTML pages.

Try http://jonpy.sourceforge.net/


From aleax at aleax.it  Mon Apr  1 13:50:57 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 18:50:57 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <m3elhz8sbs.fsf@mira.informatik.hu-berlin.de> <KM_p8.41075$pT1.1129125@news1.tin.it> <a8a0gt$roc$1@panix1.panix.com>
Message-ID: <l_1q8.49099$S52.1668112@news2.tin.it>

Aahz wrote:

> In article <KM_p8.41075$pT1.1129125 at news1.tin.it>,
> Alex Martelli  <aleax at aleax.it> wrote:
>>
>>0 is (2.2 is still present, not yet past) the canonical way to spell
>>untruth: it's what 'not', 'is', 'in', comparisons, and all standard
>>library iswhatever's return -- including operator.truth, of course.
> 
> Just to throw in a datapoint, I generally use None to indicate falsehood.

...and you can presumably continue to indulge in this quirk, if you so
choose, after 285 passes.  'return 0' is obviously faster (a LOAD_CONST
opcode, rather than a LOAD_GLOBAL for either 'return False' or explicit
'return None', where the latter always has to check globals() first
and then proceed to the built-ins), but just as obviously the difference
is so minute as not to matter -- a matter of 10% slowdown on my
machine for 100,000 calls to a function whose body is just return None,
versus one whose body is just return 0 (0.45 vs 0.41 secs for 100,000
calls including timing overhead, python -O, time.clock).  An _implicit_
return None (fall-off-the-end) would be just as fast as return 0, but
I would consider it seriously obfuscated (while 'return None' strikes me
as only moderately so -- just because of the difference wrt what all
of the builtin and library equivalents).


Alex





From Alex.Adrian at gmx.de  Wed Apr 24 06:23:55 2002
From: Alex.Adrian at gmx.de (Alex Adrian)
Date: Wed, 24 Apr 2002 12:23:55 +0200
Subject: xml question?
Message-ID: <aa6121$bgf$03$1@news.t-online.com>

hallo,

i would like to make this string xml conform? is there an easy
filterfunction in python to convert characters like "<>??" .... to xml
conform characters??

thanks for help,
alex


s = """?xml version="1.0"?>
<ticker>
<scan>
<STATUS>1</STATUS>
<UID>1986</UID>
<TEXT></TEXT>
<NAME>Dumpenv</NAME>
<FAMILY>CGI abuses</FAMILY>
<SUMMARY>Checks for the presence of /cgi-bin/dumpenv</SUMMARY>
<DESCRIPTION>""</DESCRIPTION>
<PORTS>80</PORTS>
<WWW>www.imfmetal.org/imf/main/main_about.cfm?lang=spanish</WWW>
<ORG>Federaci?n Internacional de Trabajadores de las Industrias Metal?rgicas
de la Am?ricaLatina y Caribe</ORG>
</scan>
</ticker>"""





From mlh at vier.idi.ntnu.no  Sun Apr  7 14:33:11 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 7 Apr 2002 18:33:11 +0000 (UTC)
Subject: I need to go to =?iso-8859-1?Q?Malm=F6?= tomorrow
References: <mailman.1018126154.20514.python-list@python.org>  <a8nt0j$411$1@panix1.panix.com>
Message-ID: <slrnab1477.cde.mlh@vier.idi.ntnu.no>

In article <a8nt0j$411$1 at panix1.panix.com>, Aahz wrote:
>In article <mailman.1018126154.20514.python-list at python.org>,
>Laura Creighton  <lac at strakt.com> wrote:
>>
>>I haven't forgotten you all, but it will be a while before I get to
>>replying.  Alex and I have code to write together as well.  I get a
>>week of pair programming with Alex!  Way Cool.
>
>You're just trying to make us feel jealous.

On which one of them?

<wink>

>-- 
>Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From tl_news at nexgo.de  Tue Apr 16 03:32:37 2002
From: tl_news at nexgo.de (Tino Lange)
Date: Tue, 16 Apr 2002 09:32:37 +0200
Subject: problem running win32com.client application from commandline
References: <3hrjbu4qv9b9afjhr2nhnef025350k4lpd@4ax.com> <3CBA8463.30703@skippinet.com.au> <a9e8si$47q$1@crusher.de.colt.net> <3CBBA67F.6050209@skippinet.com.au>
Message-ID: <a9gk2l$bj$1@crusher.de.colt.net>

Mark Hammond wrote:

Hallo!

>> Is there a list of those useful commandline-switches for PythonWin
>> available? A "PythonWin.exe /?" or "PythonWin.exe --help" does not help.
> Nope - look in pythonwin\pywin\framework\intpyapp.py
> 
>> [other question]
> Sorry, I have no idea - that is completely up to the OCX.

Doesn't matter, but thank's again, you helped a lot!
Best regards

Tino


From redflag at directbox.com  Wed Apr 24 11:46:06 2002
From: redflag at directbox.com (redflag)
Date: 24 Apr 2002 08:46:06 -0700
Subject: Probably I'm just too stupid, but...
Message-ID: <34e7f752.0204240746.7d2c41b0@posting.google.com>

>>> def c():
	for i in range(10):
		print i

		
>>> c
<function c at 0x00AA93B0>
>>> for i in range(10):
	print i

	
0
1
2
3
4
5
6
7
8
9
>>> 

why didn't "c" work as I imagined?


From peter at engcorp.com  Wed Apr  3 22:08:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 03 Apr 2002 22:08:40 -0500
Subject: Python compilers
References: <mailman.1017830887.6595.python-list@python.org>
Message-ID: <3CABC3B8.888D968C@engcorp.com>

Brixi Tom?? wrote:
> 
> Peter Hansen wrote:
> >
> > > What is the performance of python like compared to other languages such
> > > as C and java?
> >
> > Similar to Java, ...
> 
> Paul's response reminds me of the existence of just-in-time
> and native compilers for Java, which clearly run much faster
> than vanilla Java running on a virtual machine.  I meant
> "similar to Java" only in the sense of the original Java.
> 
> Sun ships Hotspot JIT with their jdk/jre as a default VM
> sice Java 2 platform was introduced I think

Fine.  Which platforms are supported?

(I don't really need the answer... that was a rhetorical question.)

-Peter


From xnexau at yahoo.com.au  Sat Apr 13 11:30:24 2002
From: xnexau at yahoo.com.au (Patrick W)
Date: 14 Apr 2002 01:30:24 +1000
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net> <3cb843d0$0$4068$9b622d9e@news.freenet.de>
Message-ID: <82bscn4nvz.fsf@acropolis.localdomain>

"merman" <merman at freenet.de> writes:

> > "merman" <merman at freenet.de> schrieb im Newsbeitrag
> > news:3cb7f47e$0$360$9b622d9e at news.freenet.de...
> > > Hi,
> > >
> > > is there a way to simulate interfaces (like in java) in python?
> > > ... or abstract classes? (c++/ java)
> > >
> > > o-o


Do you have a technical reason to actually want interfaces, or are you
simply experimenting with the concept?

As far as I can see, which is admittedly not that far, interfaces
don't really serve a useful purpose in Python. If you want
polymorphism in statically typed OO languages like C++ and Java, you
need to design your functions and methods to accept only objects
descended from a known type. But in Python, the type need not
(actually cannot) be specified at design time, so what purpose would
an interface serve? Wouldn't it be better to just inherit
_implementation_ when necessary, rather than to inherit pure
interfaces?

I ask this not to be a prick, but to find out if I'm missing something
important.


From emmanuel at onda-latina.com  Sat Apr  6 14:54:23 2002
From: emmanuel at onda-latina.com (manouche)
Date: Sat, 6 Apr 2002 21:54:23 +0200
Subject: global instance question (newbie)
Message-ID: <a8njqq$phs$1@wanadoo.fr>

hello,

How can I create a global object that I can be used in all the files of my
project?
Using "global" I can only access to the instance in the file where it was
created.

thanks




From tl_news at nexgo.de  Sun Apr 14 17:07:35 2002
From: tl_news at nexgo.de (Tino Lange)
Date: Sun, 14 Apr 2002 23:07:35 +0200
Subject: problem running win32com.client application from commandline
Message-ID: <3hrjbu4qv9b9afjhr2nhnef025350k4lpd@4ax.com>

Hi All!

I just discovered a problem with a win32com.client based application.

If I initialize the object 
acheck = win32com.client.Dispatch(com_name)
from pythonwin it's no problem at all. The application runs and does
everything I want it to do for me :-)

But if I use it from the commandline (either python.exe or
pythonw.exe) then the above initialization of the COM-Object fails. A
MessageBox appears (from the COM-Object, Method
"UserControl_Initialize") telling that there's not enough memory to
initialize.

Unfortunately I just have the OCX - I cannot look in the COM-Object
:-(

Any idea? What's different, when I run the script via PythonWin?
Is there a possibility to run commandline-scripts?

Thanks a lot for your help!

Tino



From maxm at mxm.dk  Wed Apr 10 10:09:52 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 10 Apr 2002 16:09:52 +0200
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org>  <3CB40E6F.90509@geneva-link.ch> <mailman.1018442912.3164.python-list@python.org>
Message-ID: <3CB447B0.1030800@mxm.dk>

Laura Creighton wrote:

> re: what is in nature vs what is in perception -- see David Pye
> The Nature and Aesthetics of Design -- especially Chapter 11
> (function and fiction) and  the latter part of Chapter 13 (the
> aesthetics of design) and all of chapter 14 (perception and
> looking).

You aren't by any chance a philosophy major ;-)

regards Max M



From gh_pythonlist at gmx.de  Tue Apr  9 20:44:57 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 10 Apr 2002 02:44:57 +0200
Subject: Easter Eggs
In-Reply-To: <fm2s8kKkXAiR092yn@gnosis.cx>
References: <fm2s8kKkXAiR092yn@gnosis.cx>
Message-ID: <20020410004457.GB3243@lilith.my-fqdn.de>

* Lulu of the Lotus-Eaters <mertz at gnosis.cx> [2002-04-09 18:22 -0400]:
> Who can tell me what's in there.  Someone on c.l.py recently let drop
> the hint about finding the Zen of Python at the interactive prompt in
> 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
> 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
> magic command was.  Can someone remind me, please?

In fact, you need the CVS version. Here's the patch:

diff -r python.orig/dist/src/Include/compile.h python/dist/src/Include/compile.h
36a37
> #define CO_ENLIGHTENMENT 0x0040
76a78
> #define FUTURE_ENLIGHTENMENT "enlightenment"
diff -r python.orig/dist/src/Lib/__future__.py python/dist/src/Lib/__future__.py
50a51
>     "enlightenment",
64a66
> CO_ENLIGHTENMENT     = 0x4000   # enlightenment
104a107,110
> 
> enlightenment = _Feature((2, 3, 0, "alpha", 0),
>                          (2, 3, 0, "alpha", 1),
>                          CO_ENLIGHTENMENT)
diff -r python.orig/dist/src/Python/future.c python/dist/src/Python/future.c
36a37,38
>                 } else if (strcmp(feature, FUTURE_ENLIGHTENMENT) == 0) {
>                     printf("You are enlightened.\n");

I won't publish the patch of the additional Python features that are
only available when CO_ENLIGHTENMENT is set ;-)

Now you can do:

bash-2.05$ ./python
Python 2.3a0 (#1, Apr 10 2002, 02:29:51)
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import enlightenment
You are enlightened.

Now you may wonder how it makes a difference to you that you actually
got enlightened. Guess that's a tough question for a Zen master, too.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 2.8 ?C      Wind: 3.5 m/s




From python at rcn.com  Sun Apr 21 20:54:12 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sun, 21 Apr 2002 20:54:12 -0400
Subject: newbie dictionary question
References: <TiHw8.99936$CH1.86776@sccrnsc02> <3CC34519.E6C1063A@gmx.de> <g1Jw8.89176$G72.64386@sccrnsc01>
Message-ID: <a9vn1g$md7$1@bob.news.rcn.net>

"James Chapman ude.awoiu at nampahc-semaj>" <<reverse> > Hi,
>
> I'd preferably like to use something like this but my problem is precisely
> this. I have for example to dictionaries as follows:
>
> a= {4.95:100,4.97:90,4.99:95}
> b={4.95:100,4.96:99}
>
> and I would wonder if there is an elegent way to do something like:
>
> c = a+b
> c={4.95:200,4.96:99,4.97:90,4.99:95}

I'm not sure if this counts as elegant, but it works:
>>> [a.__setitem__(k, a.get(k,0)+v) for k,v in b.iteritems()]
[None, None]
>>> a
{4.9699999999999998: 90, 4.9500000000000002: 200, 4.9900000000000002: 95,
4.96: 99}

FWIW, I think using floats as a key is perilous.


Raymond Hettinger




From loewis at informatik.hu-berlin.de  Wed Apr 10 11:36:57 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 10 Apr 2002 17:36:57 +0200
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org> <j4wuvfzlbw.fsf@informatik.hu-berlin.de> <mailman.1018450821.8190.python-list@python.org>
Message-ID: <j4sn63zht2.fsf@informatik.hu-berlin.de>

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

>    Thank you very much for stepping in.
> 
>    It revealed that "destructor" was defined
> typedef void (*destructor)(PyObject *);
>    before.
> 
>    Not sure if I should report this as a bug.

That is a bug: All API in a header must start with a Py prefix to
avoid the kind of conflict you got.

Unfortunately, this probably cannot be fixed for 2.2; you'll have to
work around it locally. It is unfortunate that this wasn't changed in
the Grand Renaming - it sounds like rename3.h will be needed. (*)

You could also report this as a bug to Sun: 'destructor' is not an
identifier reserved for the implementation, so they should not use it
in their header. Indeed, Solaris 9 (and probably earlier) has

extern int pthread_key_create(pthread_key_t *, void (*)(void *));

Finally, you could report this as a GCC bug: they should fix the
system header via fixincludes. This is also the work-around which I'd
offer: copy pthread.h to .../gcc-lib/system/2.8.1/include/pthread.h,
and adjust the copy accordingly.

HTH,
Martin

(*) The other apparent fix is also not available: one could propose to
include pthread.h before Python.h, but the change in include order was
deliberate, so that the system headers could pick up any defines in
pyconfig.h.


From gerhard at bigfoot.de  Thu Apr 25 21:06:30 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 26 Apr 2002 01:06:30 GMT
Subject: PyGTK libs for Zaurus?
References: <aaa0ev$963r5$1@ID-139866.news.dfncis.de>
Message-ID: <slrnacha0k.en.gerhard@lilith.my-fqdn.de>

rapskat wrote in comp.lang.python:
> Anyone know anything about installing the PyGTK ligs on the Sharp Zaurus
> 5500?

The Sharp Zaurus runs on Qtopia, which uses Qt Embedded as underlying
graphics library.

> I'm using Qutopia for the unit's  WM/desktop, would GNOME be required?

No, but X.

> If so, I guess this is just a pipe dream then.

What's wrong with using PyQt?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From tjreedy at udel.edu  Mon Apr 29 11:02:32 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 29 Apr 2002 15:02:32 GMT
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <cgdz8.167961$XV5.15940232@bin4.nnrp.aus1.giganews.com>

"Raymond Hettinger" <python at rcn.com> wrote in message
news:aaiu6e$pfr$1 at bob.news.rcn.net...
> Please send me comments on a draft PEP deprecating
> and relocating six built-in functions.  The text file is at:
> http://users.rcn.com/python/download/deppep.txt

-1 on proposal as stated.

I agree with Alex M. that major concerns should be separated.

First, should we open up a can of worms and posssibly set off a
firestorm by executing
builtins = filter(keep, builtins) #  or is really better to say
builtins = [func for func in builtins if keep(func)] # or even
builtins = [func for func in builtins if not remove(func)]

Second (given First is true), what should be general mechanism and
philosophy.  Should we reduce gradually or just once (at 3.0).  Should
we keep unless there is a reason to remove, or remove unless a reason
to keep.  We could put *everything* in a separate module.  I am not
suggesting that, but it seems to me that builtins *could* easily and
justifiably be reduced by about half, given that we get started and go
beyoun the most obvious and uncontentious.

Third (having answered Second) is the contentious fight over fleshing
out the body of keep() or remove().  Input() is obvious.  So,
arguably, is apply() -- if the criterion for removal is 'now obsolete'
(in this case because of a syntax addition) rather than 'seldom used
in the past'.  On the otherhand, using divmod() has been touted as a
way to insulate onself from the int division change, so it would seem
ungracious to 'deprecate' it anytime soon.

Terry J. Reedy







From aut at c2i.net  Mon Apr 22 14:43:16 2002
From: aut at c2i.net (Audun Tonnesen)
Date: Mon, 22 Apr 2002 18:43:16 GMT
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com> <20020412101611.GB15466@andy.tynant.ftech.net> <20020412131633.GA17830@andy.tynant.ftech.net> <mailman.1018636586.2076.python-list@python.org>
Message-ID: <1103_1019500994@news1.c2i.net>

I have problems with retrbinary. The call is like this:
ftpobject.retrbinary("RETR <filename>, file.write(someData),[blocksize])

My problem is the <someData> argument to file.write(...). How do I define or bind the name to the actual datachunk?

Regards Audun







From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 15:23:02 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 21:23:02 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net> <ypjv8.42546$vm6.5594931@ruti.visi.com> <a9khoc$j7h$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9ki3d$jip$1@nets3.rz.RWTH-Aachen.DE>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9khoc$j7h$1 at nets3.rz.RWTH-Aachen.DE...
>..
> The three commented lines above would not have been needed if I'd use
> something like '... name="' + (key != 'station') ? key : 'station' + '"
...'
>..

That should be '... name="' + (key != 'station') ? key : 'lastStation' + '"
...'

Or since I sometimes prefer testing in a positive way rather than the most
likey, '... name="' + (key == 'station') ? 'lastStation' : key + '" ...'




From jeff at ccvcorp.com  Mon Apr 22 15:45:33 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 12:45:33 -0700
Subject: constructors in python
References: <a9uoa0$t0r$1@news.iucc.ac.il> <3CC2ED37.3D9225B3@engcorp.com> <a9usk3$rk$1@news.iucc.ac.il>
Message-ID: <MPG.172e082bb60a58569896c0@news.nwlink.com>

In article <a9usk3$rk$1 at news.iucc.ac.il>, elimli at yahoo.com 
says...
> But what does happened when I forget (I'm just a man) to call father's c-or?
> Semi-initialized object? C++ doesn't permit not to call the base c-or.
 
The semantics are different between C++'s constructors and 
Python's __init__() methods.

In the C++ case, the constructor is called to allocate memory for 
the object, and create an initial layout in that memory.  If the 
constructor fails, then you have no object.

In Python, the object is fully created *before* __init__() is 
called.  Thus, __init__() is more of an initializer than a 
constructor.  The equivalent C++ idiom would be something like 
this:

    foo = new foo_obj();
    foo.init();

...with the __init__() method corresponding to the foo.init() 
call, and the new being done internally by the interpreter.  If 
your __init__() method fails, then you still have an object, 
even though it is an incompletely initialized one.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From sholden at holdenweb.com  Tue Apr  9 18:50:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 9 Apr 2002 18:50:35 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1018339899.2981.python-list@python.org> <roy-7658EF.08144409042002@news1.panix.com>
Message-ID: <JgKs8.5356$GM.4821@atlpnn01.usenetserver.com>

"Roy Smith" <roy at panix.com> wrote in message
news:roy-7658EF.08144409042002 at news1.panix.com...
> tanzer at swing.co.at (Christian Tanzer) wrote:
> > There are several dimensions to typing, including:
> > - static vs. dynamic
> > - strong vs. weak
>
> Putting it another way...
>
> In Perl, if I write "$x = 1;", x can act as either an integer or a string,
> depending on how I use it.
>
> In Python, if I write "x = 1", x can only act as an integer.  In that
> sense, it's strongly typed.  On the other hand, it's also true that
> sometimes later, I can say "x = '1'", and now x can only act as a string.
> At any particular point in time, its type is strongly defined, but the
type
> can also (dynamically) change from time to time.
>
> In Python, the variable (x) can change type, but the value itself (1)
> cannot.  In Perl, the value itself can change type as coerced by the
> context.

It might be preferable to emphasise that variables have no type by
expressing the same idea as "The value bound to the name x can be of any
type, but in Python there are fewer implicit coercions than in Perl".

regards
 Steve






From bokr at oz.net  Sun Apr  7 21:06:47 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 01:06:47 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <a8pqal$ftm$1@panix1.panix.com> <7xpu1bh481.fsf@ruckus.brouhaha.com> <a8q0g1$qu5$1@panix1.panix.com>
Message-ID: <a8qqf7$p4h$0@216.39.172.122>

On 7 Apr 2002 13:43:29 -0400, aahz at pythoncraft.com (Aahz) wrote:
[...]
>
>Assignments in expressions will never happen.
>
I guess you mean they won't be spelled concisely ;-)

 >>> a='eigh'
 >>> a
 'eigh'
 >>> locals().__setitem__('a','bee') and 0 or a
 'bee'

Regards,
Bengt Richter


From paul at boddie.net  Fri Apr 19 10:17:41 2002
From: paul at boddie.net (Paul Boddie)
Date: 19 Apr 2002 07:17:41 -0700
Subject: Twisted 0.17.2 Released
References: <CvMv8.15243$ue4.582660@weber.videotron.net> <QiPv8.107052$%8.9166874@bin2.nnrp.aus1.giganews.com>
Message-ID: <23891c90.0204190617.4a49aede@posting.google.com>

"Terry Reedy" <tejarex at yahoo.com> wrote in message news:<QiPv8.107052$%8.9166874 at bin2.nnrp.aus1.giganews.com>...
> 

[Twisted]

> After reading the 'license' page in Apr 2, I looked at every page I
> could find for something about the RPG engine but could not find much
> at all.  Is there a page I missed or must I download and look for doc
> or source which will tell me whether or not I wanted to download?

One thing that I believe developers should do when releasing code is
to make it easy for people to discover the licence associated with
their software. After the April Fools' joke played by the Twisted
team, I'd really expect the actual licence to be prominent on their
site.

It may be the case that a search for "Twisted license" on Google does
bring up a reference to the GNU LGPL, but the next thing for the more
creative Python developers amongst us could well be the concealment of
the licence text for their packages in a file called
"BewareOfTheLeopard.txt". ;-)

Paul


From frr at easyjob.net  Tue Apr 23 16:45:12 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 22:45:12 +0200
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com>
Message-ID: <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com>

On Tue, 23 Apr 2002 11:38:37 -0700, Jeff Shannon <jeff at ccvcorp.com> wrote:

>In article <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv at 4ax.com>, 
>frr at easyjob.net says...
>> 
>> Whenever a customer clicks on the order button, I'd like to:
>> 
>> a) Save some info into a database
>> b) Send the order to the payment processing company
>> 
>> I was planning to use a Python CGI for a. This cgi shouldn't return anything
>> and shouldn't change the contents of the browser. Is this possible? O:-)  
>
>So that the person making the order has no idea whether it 
>actually went through or not?  

No. Step (b) processes the order and does give feedback. Step (a) saves in a
db some info conatined in a cookie (date of first visit and first referrer). I
want step (a) to be seamless.





-----------------------
Fernando Rodriguez


From huaiyu at gauss.almadan.ibm.com  Thu Apr  4 14:49:01 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 4 Apr 2002 19:49:01 +0000 (UTC)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>              <3CAC149F.3F1BA77B@lemburg.com> <mailman.1017930740.20674.python-list@python.org>
Message-ID: <slrnaapbh8.4s7.huaiyu@gauss.almadan.ibm.com>

[GvR]
>The textbooks that say that a Boolean value does not support
>arithmetic operations.  In tis discussion several people have tried to
>argue for such bools and claimed that that is the textbook definition.

Textbooks on algebra usually define algebra as sets with certain operations
and certain rules.  They do not say what operations an algebra cannot have,
except in cases of theorems showing certain operations as incompatible.

Regarding Python bools as subtype of ints, the boolean algebra and the ring
of int arithmetic are not incompatible.  They are just completely
independent.  There will only be conflict if one tries to associate the
operators + - * with 'and' 'or', etc.

>
>> FWIW, I was talking about the text books students use to learn boolean 
>> algebra.
>
>I don't think Boolean algebra is a big subject.  I bet that most
>people only learn about it explicitly when they learn computer
>programming.

I think the textbook definition of algebra fits perfectly in this situation.
The set of all Python objects with the logical operators forms an algebra.
It is not boolean because the operators are not commutative, among others.
The operator.truth is a homomorphism of this algebra to a boolean subalgebra
on the set of two elements (previously 0 and 1, now False and True).  It is
a subalgebra since it is the range of 'not'.

BTW, many people mistake boolean algebra with two-element boolean algebra,
which is only its most trivial example.  The set operations (union, joint,
complement) on all subsets of a given set is a more generic example of
boolean algebra.

Huaiyu


From pyth at devel.trillke.net  Mon Apr 22 12:21:26 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 22 Apr 2002 18:21:26 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <304d20df.0204220625.6d1430ba@posting.google.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <mailman.1019403879.24573.python-list@python.org> <304d20df.0204220625.6d1430ba@posting.google.com>
Message-ID: <20020422162126.GE27395@devel.trillke>

On Mon, Apr 22, 2002 at 07:25:47AM -0700, Allan Crooks wrote:
> holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1019403879.24573.python-list at python.org>...
> 
> > exec forversion("2.2",
> 
> <snip>
> 
> I had considered using exec as a way of getting round it, but it seems
> like quite a hack to get what I want (wrapping code in a string to
> prevent Python accessing it properly).
> Mind you, considering what I want to do, I suppose hacking is really
> the only feasible solution. :)

:-)

> Doing things via exec would ultimately mean that I couldn't take
> advantage of PyChecker...

I don't know PyChecker enough but judging from their README it seems
to work on modules. So it should be conceptionally possible to 
do the following:

import pychecker # at the start of your program

# in forversion:
    involve_pychecker(string)
    exec string

def involve_pychecker(string):
    import os,sys
    if sys.modules.has_key('pychecker'):
        fn=os.tmpnam()
        f=open(fn,'w')
        f.write(string) 
        f.close()
        sys.path.append('/tmp')
        exec "import "+fn # pychecker generates warnings, if any
        sys.path.pop()  # cleanup
	del sys.modules[fn] # kill the temporary module

don't call it hack, call it "adapter to pychecker" :-)
Guess that for debugging purposes the use of os.tmpnam()
is endurable.

	holger




From m.faassen at vet.uu.nl  Wed Apr 24 15:27:24 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 24 Apr 2002 19:27:24 GMT
Subject: atoi
References: <mailman.1019649779.5458.python-list@python.org>
Message-ID: <aa70us$k06$1@newshost.accu.uu.nl>

Mariappan, MaharajanX <m_mariappanX at trillium.com> wrote:
> Hi all,

> simple......but need to know

> is there any python buildin method whihc will convert strings into integer,
> like atoi in c lib.?

> "234" should be converted into interger value 234

> but the string function ord(character) convert it into equvalent ASCII code

Try the builtin 'int()' (no need to import anything).

so:

foo = int('10')

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From krissepu at vip.fi  Wed Apr  3 14:10:04 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Wed, 3 Apr 2002 22:10:04 +0300
Subject: Memory Leaks with wxwindow and grid
Message-ID: <3cab5418@news.vip.fi>

I get the following lines in dos prompt in win2000,
when exiting the main frame:

E:\koodit\python\grid>python pngrid.py
21:39:42: There were memory leaks.
21:39:42: ----- Memory dump -----
21:39:42: wxEvtHandler at $1067DB8, size 48
21:39:42:
21:39:42:
21:39:42: ----- Memory statistics -----
21:39:42: 1 objects of class wxEvtHandler, total size 48
21:39:42:
21:39:42: Number of object items: 1
21:39:42: Number of non-object items: 0
21:39:42: Total allocated size: 48

Note that if I comment line:
            self.pn_grid = PnGrid(self)


no memory leaks will occur then "exit" is selected from menu !!!

What am I missing here ?

----------code starts here-------------------

from wxPython.wx import *
from wxPython.grid import *
from wxPython.lib.mixins.grid import wxGridAutoEditMixin
import sys, pprint
ID_EXIT  = 102

class PnGrid(wxGrid):
    def __init__(self, parent):
        wxGrid.__init__(self, parent, -1)

        self.moveTo = None

        EVT_IDLE(self, self.OnIdle)

        self.CreateGrid(25, 25)
        #self.EnableEditing(false)

        # simple cell formatting
        self.SetColSize(3, 200)
        self.SetRowSize(4, 45)
        self.SetCellValue(0, 0, "First cell")
        self.SetCellValue(1, 1, "Another cell")
        self.SetCellValue(2, 2, "Yet another cell")
        self.SetCellValue(3, 3, "This cell is read-only")
        self.SetCellFont(0, 0, wxFont(12, wxROMAN, wxITALIC, wxNORMAL))
        self.SetCellTextColour(1, 1, wxRED)
        self.SetCellBackgroundColour(2, 2, wxCYAN)
        self.SetReadOnly(3, 3, true)

        self.SetCellEditor(5, 0, wxGridCellNumberEditor())
        self.SetCellValue(5, 0, "123")
        self.SetCellEditor(6, 0, wxGridCellFloatEditor())
        self.SetCellValue(6, 0, "123.34")

        self.SetCellValue(6, 3, "You can veto editing this cell")


        # attribute objects let you keep a set of formatting values
        # in one spot, and reuse them if needed
        attr = wxGridCellAttr()
        attr.SetTextColour(wxBLACK)
        attr.SetBackgroundColour(wxRED)
        attr.SetFont(wxFont(10, wxSWISS, wxNORMAL, wxBOLD))

        # you can set cell attributes for the whole row (or column)
        self.SetRowAttr(5, attr)

        self.SetColLabelValue(0, "Custom")
        self.SetColLabelValue(1, "column")
        self.SetColLabelValue(2, "labels")

        self.SetColLabelAlignment(wxALIGN_LEFT, wxALIGN_BOTTOM)

        # test all the events
        EVT_GRID_CELL_LEFT_CLICK(self, self.OnCellLeftClick)
        EVT_GRID_CELL_RIGHT_CLICK(self, self.OnCellRightClick)
        EVT_GRID_CELL_LEFT_DCLICK(self, self.OnCellLeftDClick)
        EVT_GRID_CELL_RIGHT_DCLICK(self, self.OnCellRightDClick)

        EVT_GRID_LABEL_LEFT_CLICK(self, self.OnLabelLeftClick)
        EVT_GRID_LABEL_RIGHT_CLICK(self, self.OnLabelRightClick)
        EVT_GRID_LABEL_LEFT_DCLICK(self, self.OnLabelLeftDClick)
        EVT_GRID_LABEL_RIGHT_DCLICK(self, self.OnLabelRightDClick)

        EVT_GRID_ROW_SIZE(self, self.OnRowSize)
        EVT_GRID_COL_SIZE(self, self.OnColSize)

        EVT_GRID_RANGE_SELECT(self, self.OnRangeSelect)
        EVT_GRID_CELL_CHANGE(self, self.OnCellChange)
        EVT_GRID_SELECT_CELL(self, self.OnSelectCell)

        EVT_GRID_EDITOR_SHOWN(self, self.OnEditorShown)
        EVT_GRID_EDITOR_HIDDEN(self, self.OnEditorHidden)
        EVT_GRID_EDITOR_CREATED(self, self.OnEditorCreated)



    def OnCellLeftClick(self, evt):
        pprint.pprint("OnCellLeftClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnCellRightClick(self, evt):
        pprint.pprint("OnCellRightClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnCellLeftDClick(self, evt):
        pprint.pprint("OnCellLeftDClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnCellRightDClick(self, evt):
        pprint.pprint("OnCellRightDClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnLabelLeftClick(self, evt):
        pprint.pprint("OnLabelLeftClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnLabelRightClick(self, evt):
        pprint.pprint("OnLabelRightClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnLabelLeftDClick(self, evt):
        pprint.pprint("OnLabelLeftDClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()

    def OnLabelRightDClick(self, evt):
        pprint.pprint("OnLabelRightDClick: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()


    def OnRowSize(self, evt):
        pprint.pprint("OnRowSize: row %d, %s\n" %
                       (evt.GetRowOrCol(), evt.GetPosition()))
        evt.Skip()

    def OnColSize(self, evt):
        pprint.pprint("OnColSize: col %d, %s\n" %
                       (evt.GetRowOrCol(), evt.GetPosition()))
        evt.Skip()

    def OnRangeSelect(self, evt):
        if evt.Selecting():
            pprint.pprint("OnRangeSelect: top-left %s, bottom-right %s\n" %
                           (evt.GetTopLeftCoords(),
evt.GetBottomRightCoords()))
        evt.Skip()


    def OnCellChange(self, evt):
        pprint.pprint("OnCellChange: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))

        # Show how to stay in a cell that has bad data.  We can't just
        # call SetGridCursor here since we are nested inside one so it
        # won't have any effect.  Instead, set coordinants to move to in
        # idle time.
        value = self.GetCellValue(evt.GetRow(), evt.GetCol())
        if value == 'no good':
            self.moveTo = evt.GetRow(), evt.GetCol()


    def OnIdle(self, evt):
        if self.moveTo != None:
            self.SetGridCursor(self.moveTo[0], self.moveTo[1])
            self.moveTo = None
        evt.Skip()


    def OnSelectCell(self, evt):
        pprint.pprint("OnSelectCell: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))

        # Another way to stay in a cell that has a bad value...
        row = self.GetGridCursorRow()
        col = self.GetGridCursorCol()
        if self.IsCellEditControlEnabled():
            self.HideCellEditControl()
            self.DisableCellEditControl()
        value = self.GetCellValue(row, col)
        if value == 'no good 2':
            return  # cancels the cell selection
        evt.Skip()


    def OnEditorShown(self, evt):
        if evt.GetRow() == 6 and evt.GetCol() == 3 and \
           wxMessageBox("Are you sure you wish to edit this cell?",
                        "Checking", wxYES_NO) == wxNO:
            evt.Veto()
            return

        pprint.pprint("OnEditorShown: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()


    def OnEditorHidden(self, evt):
        if evt.GetRow() == 6 and evt.GetCol() == 3 and \
           wxMessageBox("Are you sure you wish to  finish editing this
cell?",
                        "Checking", wxYES_NO) == wxNO:
            evt.Veto()
            return

        pprint.pprint("OnEditorHidden: (%d,%d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
        evt.Skip()


    def OnEditorCreated(self, evt):
        pprint.pprint("OnEditorCreated: (%d, %d) %s\n" %
                       (evt.GetRow(), evt.GetCol(), evt.GetControl()))



class MainFrame(wxFrame):
    """ Creates and initializes the main window"""
    def __init__(self, parent, id, title, style):
        # First, call the base class' __init__ method to create the frame
        wxFrame.__init__(self, parent, id, title, wxPoint(100, 100),
wxSize(640, 480), style)

        self.CreateStatusBar()
        self.SetStatusText("This is the statusbar")

        # Make menu with separate MakeMenus -method
        menu = self.MakeMenus()
        self.SetMenuBar(menu)

        # Associate some events with methods of this class
        EVT_MENU(self, ID_EXIT,  self.OnExit)

        # Create grid
        self.pn_grid = PnGrid(self)

    def MakeMenus(self):
        """Creates the menus of the main window"""

        # Filemenu contents
        filemenu = wxMenu()
        filemenu.Append(ID_EXIT, "Exit", "Terminate the program")

        # Collect all menus
        main = wxMenuBar()
        main.Append(filemenu, "File")
        return main

    def OnExit(self, event):
        # Close program
        self.pn_grid.Destroy()
        self.Close(true)


class MainApp(wxApp):
    # wxWindows calls this method to initialize the application

    def OnInit(self):
    # Create an instance of our customized Frame class
        # Create an instance of our customized Frame class
        frame = MainFrame(NULL, -1, " PnGrid", wxSTAY_ON_TOP | wxCAPTION)
        frame.Show(true)

    # Tell wxWindows that this is our main window
        self.SetTopWindow(frame)
        return true

if __name__ == "__main__":
    app = MainApp(0)
    app.MainLoop()





From bkc at Murkworks.com  Tue Apr  9 16:38:58 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 9 Apr 2002 16:38:58 -0400
Subject: property() type that only protects set?
Message-ID: <3cb34fb8$1_4@goliath.newsgroups.com>

Anyone have an implementation of property() that doesn't require a getter,
only a setter (and optional del)

I only need to protect set operations, and want to allow get to go straight
to __dict__

(or object when I'm using slots)


--
Novell DeveloperNet Sysop #5

_




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From hgg9140 at seanet.com  Sat Apr 13 19:52:39 2002
From: hgg9140 at seanet.com (Harry George)
Date: 13 Apr 2002 16:52:39 -0700
Subject: ANN: pdx 1.4.2
Message-ID: <mailman.1018761181.319.clpa-moderators@python.org>

Pdx is a POD-style markup language, which outputs html, latex, or
docbook.  For documentation and tarball, see:
  http://www.seanet.com/~hgg9140/comp/index.html

Bugfix for nested includes with relative path redirections.

#CHANGES
2002-04-13: v1.4.2
  - Revised Base.do_include_getdata.  It was just looking
    at the last item on the pathstack, thus failing to handle
    multiple redirections.  Now concatenates the full stack.

-- 
Harry George
hgg9140 at seanet.com




From theller at python.net  Tue Apr 30 15:08:50 2002
From: theller at python.net (Thomas Heller)
Date: Tue, 30 Apr 2002 21:08:50 +0200
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org>  <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no>
Message-ID: <aamq47$c03hp$1@ID-59885.news.dfncis.de>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message news:slrnactp07.pv.mlh at vier.idi.ntnu.no...
> ... I've got plenty of
> code with stuff like
>
>   file = open('foo.bar')
>

Perhaps you should now write

    open = file('foo.bar')

;-)

Thomas




From gimbo at ftech.net  Fri Apr  5 03:51:16 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 5 Apr 2002 09:51:16 +0100
Subject: Catalog of Python Patterns? [LONG!]
In-Reply-To: <20020405014732.GB17563@isis.gerrietts.net>
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com> <20020405014732.GB17563@isis.gerrietts.net>
Message-ID: <20020405085116.GA16724@andy.tynant.ftech.net>

On Thu, Apr 04, 2002 at 05:47:32PM -0800, Geoff Gerrietts wrote:

> I think that you could make a case for the limited applicability of
> Command pattern, the Singleton pattern, the Prototype pattern, and I'm
> sure several others. You could implement them, but there's not a solid
> reason to do so, because Python doesn't impose the restrictions that
> the patterns are designed to circumvent.

Really?  I've used Singleton a fair bit (probably too much in fact)
for stuff like configuration information.  I'm curious as to how else
I could/should do this.

Here's how I'd do it now:

-- begin Config.py -- 

_configInstance = None

def Config(args=None):
    global _configInstance
    if _configInstance is None:
        _configInstance = ConfigInstance(args)
    return _configInstance

class ConfigInstance:
    def __init__(self, args=None):
	# do some stuff

-- end Config.py --

Then to use it elsewhere I just do:

import Config
config = Config.Config()
if config.send_error_emails:
    send_it()

etc.

and I know what I'm getting.

If you can suggest a better/cleaner/simpler/more pythonic way of
achieving this, I'd be honestly interested.  :-)

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From grahamd at dscpl.com.au  Wed Apr 10 23:20:36 2002
From: grahamd at dscpl.com.au (Graham Dumpleton)
Date: 10 Apr 2002 20:20:36 -0700
Subject: Help: Python for a big commercial project?
References: <mailman.1018437980.14078.python-list@python.org>
Message-ID: <dc6f5c99.0204101920.143540b6@posting.google.com>

"Cereal Muesli" <cereal_con_fibra at hotmail.com> wrote in message news:<mailman.1018437980.14078.python-list at python.org>...
> 	I need a piece of advice from you, Oh wise Pythoneers! 8-)
> 
>   I'm involved in the development of a big DataWarehouse for an important 
> TelCo, and I want to propose Python as
>   the language of choice for internal maintenance tasks, high level control 
> tasks for the loading procedures, etc...
>   the people here is not likely to use any free stuff for "serious" 
> development (you know...no money = no quality, sigh!),
>   but i want to change this, since I believe that Python is really a very 
> good product. I have played with Python for
>   a year or so under Windows, and before trying to convince people here that 
> it is a good choice, I need to know
>   some things:

Perhaps if you can point at another Telco where Python is being used. I
will not name names, but a certain almost monopolistic telco in Australia
(not hard to work out who :-)) uses Python in some OSS, NMS and TMS
applications which help to monitor and manage their international telephone
switches and associated equipment. Aspects of the TMS application amounted
to a data warehouse as it stored the last few months worth of call records
for all phone calls going in, out or through the country through that telco.

Not everything is Python as the core infrastructure is written in C++. What
has been done is that Python wrappers have been added to a distributed
messaging infrastructure to allow parts to be written in Python. One way in
which this is used is for high level control and maintenance tasks. Python is
also being used for writing some GUIs as well.

> - First of all; do you think that Python is suitable for this
>   kind of "serious, big" project?

As far as using it as the glue to help manage the system and perhaps
to build some components then yes, as to using it for practically
everything, possibily no, this being for reasons political and depending
on what you are doing perhaps also technical. You really need to have
good management support and good technical people who can show how
the technology can be used effectively to give management a good sense
they are doing the right thing.
 
> - At least two of the candidate modules to be developed in
>   Python require a GUI under X Windows.

In the telco mentioned above, for those areas where Python was used
for GUIs, the Pmw (http://pmw.sourceforge.net) module was used. In
fact, this is where the Pmw module was first created and developed
before it was released on the Internet. Ie., the Pmw module has telco
grade breeding. :-)

> - Database access would be required for some data analysys
>   modules, specifically ORACLE 8 access would definitely be a
>   requirement

Database access was done through a slightly modified version of one
of the Oracle modules for Python. This was in some applications only
though, as there were also Oracle interfaces in C++ using OCI which
were accessible over a distributed messaging system.

In respect of the C++ core of these applications, the OSE C++ class
library was used. The distributed messaging system and Python wrappers
were developed in house, however, in the past couple of years, an
equivalent distributed messaging system has been added to OSE. This
includes Python wrappers for the distributed messaging system as well
as a HTTP servlet framework and XML-RPC and SOAP gateways. Included
with OSE are also examples of how to create database query agents which
adhere to the Python database API. In short, OSE provides a framework
which is quite capable of being used in a telco setting and which has
proven itself to be robust enough for that sort of environment.

For more information on OSE, check out:

  http://ose.sourceforge.net

It may especially be worthwhile browsing the Python manual and the
examples presented on the web site.


From theller at python.net  Tue Apr 30 11:47:59 2002
From: theller at python.net (Thomas Heller)
Date: Tue, 30 Apr 2002 17:47:59 +0200
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <199334c5.0204260645.3af87dc5@posting.google.com> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de> <vIxz8.81546$vF6.2424489@news2.tin.it>
Message-ID: <aamebj$c2olc$1@ID-59885.news.dfncis.de>

"Alex Martelli" <aleax at aleax.it> wrote in message news:vIxz8.81546$vF6.2424489 at news2.tin.it...
> Ralf Juengling wrote:
>         ...
> > BTW: It is not easy to add __iter__ to list's meta-type, is it?
>
> list's "metatype" (metaclass) is the built-in type named 'type'.
> like for all other built-in types and new-style classes.  Adding
> __iter__ to type would thus make almost 'everything' iterable.
>
Correction: adding __iter__ to type would make instances of
type iterable (these things are normally called 'classes').
Not too useful, it seems.

Thomas




From phr-n2002a at nightsong.com  Wed Apr  3 21:04:39 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 18:04:39 -0800
Subject: Catalog of Python Patterns? [LONG!]
References: <a8g97m$10jq$1@news.idiom.com>
Message-ID: <7xy9g443t4.fsf@ruckus.brouhaha.com>

jimd at vega.starshine.org (Jim Dennis) writes:
>  GoF Design Patterns in Python
> 
>  [Warning LONG POST AHEAD!]
> 
>  I'd like to see a canonical catalog of OODP in Python.
> 
>  At a minimum it should list each of the 23 GoF patterns, and
>  at least one minimal Python implementation of that pattern.

Yucch, I've never been a fan of that book, it seems too cultish to me.

>  Creational::
>    Abstract Factory:
>     I've never understood the difference between an "abstract factory"
>     and a "factory method."  Since Python doesn't really have 
>     "abstract" versus "concrete" classes, perhaps the distinction is
>     meaningless.
> 
>     Every Python class acts like a factory.  anObj = Foo() returns
>     a reference to a new "Foo" object.  Additionally Python allows
>     us to define normal functions which return object references.

I think the idea here is an abstract factory returns an instance that's
some descendant of a class, rather than necessarily a direct member.
For example, you might have a general "polygon" class, and subclasses
"pentagon" and "hexagon" for 5- and 6-sided polygons.  Then

   p = PolygonFactory(5)

could return a pentagon object, while

   p = PolygonFactory(9)

would return a general polygon with 9 sides.

The only way I know of to do that in Python with a class constructor
rather than a function involves using metaclasses in a fairly hairy way.


From theoryboy at my-deja.com  Mon Apr 22 06:26:21 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 22 Apr 2002 03:26:21 -0700
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com> <slrnac349u.ea.usenet@fohnhab.minus1.de> <ced73313.0204210643.2d88ac82@posting.google.com> <3CC2DA79.26698885@metaslash.com>
Message-ID: <ced73313.0204220226.3329b66@posting.google.com>

Yes, printNumber(self, number) is what I meant.

That worked fine, thanks very much for your help. I'm amazed at how
quick Python Tkinter development is, although I expect that's a common
response.

Peter


From philh at comuno.freeserve.co.uk  Tue Apr 16 18:22:43 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:22:43 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com>
Message-ID: <slrnabp91j.ipr.philh@comuno.freeserve.co.uk>

On Tue, 16 Apr 2002 16:27:43 GMT, maxx at easynerws.com <maxx at easynerws.com> wrote:
>On Sun, 14 Apr 2002 19:36:27 +0100, philh at comuno.freeserve.co.uk (phil hunt)
>wrote:
>
>>IMO until Python developers adopt a policy of being very reluctant 
>>to do this, many IT managers will be wary of using it in anything 
>>other thsan toy projects, and Python's user base will not grow as 
>>quickly as (IMO) it should.
>
>Why? 
>
>At my company we are stuck in the realm of Windows, and the IT staff must
>constantly patch the webservers to keep up with security and bugfixes. A quick
>check of the IT media will show management's displeasure over this continual
>process, but it does not appear to have hampered the deployment of Windows
>servers.

That's because they want to play save "no-one got fired from buying 
Microsoft", even if it is often crap.

>As mentioned in elsewhere this thread, Java and VB, have and are undergoing
>complete architecture restructuring, and they still are expanding their
>territories.

This is true. I am glad I am not using either.

>Keep in mind, IT managers often do not think and act like developers.

Indeed so.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From ods at fep.ru  Tue Apr 16 08:09:52 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Tue, 16 Apr 2002 16:09:52 +0400 (MSD)
Subject: output formating
In-Reply-To: <3CBC122E.5090306@beth.uniforum.org.za>
Message-ID: <Pine.LNX.4.44L.0204161609390.13708-100000@ods.pp.ru>

On Tue, 16 Apr 2002, mixo wrote:

m> I am looking for a way to do something like ('cal' command )

http://www.python.org/doc/2.2.1/lib/module-calendar.html





From eppstein at ics.uci.edu  Tue Apr 16 01:14:47 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 22:14:47 -0700
Subject: Does Python need a '>>>' operator?
References: <mailman.1018925938.16233.python-list@python.org> <a9g8ok$9a3$1@panix1.panix.com>
Message-ID: <eppstein-6C9350.22144715042002@news.service.uci.edu>

In article <a9g8ok$9a3$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> In article <mailman.1018925938.16233.python-list at python.org>,
> Tim Peters  <tim.one at comcast.net> wrote:
> >
> >a co-finite set 
> 
> What's a "co-finite set"?  (Yeah, I could Google for it, but I usually
> understand *your* explanations -- eventually. ;-)

The complement of a finite set.

In this case, the set of nonzero bit positions of a positive long is a 
finite set of natural numbers, as is the set of zero bit positions of a 
negative long.  So the set of nonzero positions of a negative long and 
the set of zero positions of a positive long are cofinite.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From sholden at holdenweb.com  Thu Apr 18 07:30:27 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 18 Apr 2002 07:30:27 -0400
Subject: Reading from a binary file in a portable way
References: <3cbe6892_11@news.newsgroups.com>
Message-ID: <p9yv8.11123$eS3.9630@atlpnn01.usenetserver.com>

"jb" <egrossde at yahoo.de> wrote ...
>
> I should like to read a (binary) file x.y in a buffer and then do things
> like getting the 16 bit integer that is at the offset n, n+1 in the file
> (maybe lowbyte, then higyhbyte.
>
> Previously I did it only under MS-Windows where for example I could open
the
> file with the flag O_BINARY but this is no longer possible under Linux.
>
> What would a solution that works under both Unix and MS-Windows look like?
>
Such a solution would probably use the mmap module. Although there are minor
differences on the two platforms it should be possible to write something
that works well on both.

regards
 Steve






From liste_nospam_ at opikanoba.org  Wed Apr 17 03:42:57 2002
From: liste_nospam_ at opikanoba.org (=?windows-1252?Q?Fr=E9d=E9ric=20Laurent?=)
Date: Wed, 17 Apr 2002 07:42:57 GMT
Subject: Pure python crypto
Message-ID: <1104_1019029377@news.free.fr>

Sur Wed, 17 Apr 2002 02:17:03 -0400, vous avez ?crit dans comp.lang.python:
> 
>   Would anyone have any use for a pure python implementation of
> Blowfish? Or any other crypto algorithms for that matter.  As a matter
> of interest, I found myself implementing the blowfish module recently
> but was unsure if any audience would be interested. No C compiler
> is required, which is nice. Could use some testing though.
> 
>                -- Mike

Hello, as I can't make amkCrypto work, and as I need simple
functionnalities (crypt/decrypt), your proposition interest me...
I doesn't have any c/c++ compiler installed, so a pure python
solution is welcomed ;)

Fred

--
Fr?d?ric Laurent
http://www.opikanoba.org 




From TerryByrne1963 at yahoo.com  Mon Apr 29 10:37:24 2002
From: TerryByrne1963 at yahoo.com (Terry Byrne)
Date: 29 Apr 2002 07:37:24 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <m3y9f7hdc7.fsf@box.home.de>
Message-ID: <93d52e82.0204290637.60a65a9e@posting.google.com>

Jochen,

Thanks for the response. 

I'm nobody in the Python world, wish I could help you with the
documentation! I agree with you that it's a good reference, but I too
didn't find it helpful as a tutorial.

To get started I read "Learning Python" by Lutz & Ascher, published by
O'Reilly (http://www.oreilly.com). Once I realized that everything in
Python corresponds to what Perl would call either a "scalar" or a
"list", things got easier! Of course, beyond that Python is truly
object-oriented in a way that Perl isn't, but that's another
discussion ;-) ...

Basically I'm going through a list of strings in the log file,
generating an HTML summary page and conditionally e-mailing certain
production managers. pass lets me treat the long list of lines like a
C switch or a Pascal case: once a certain error code is found, pass
skips to the next line in the list without even testing for the other
codes. Much faster, of course. Please see my seperate response to
Fernando in this thread.

I learned about "pass" in Beazley's "Python Essential Reference",
published by New Riders Publishing (http://www.newriders.com). Then
grepped the modules that ship with Python and observed that the Python
pros use it a lot, and figured it must be very helpful. I haven't
calculated the percentage change in a shortening of runtime from 1 1/2
hours to < a minute, but I know it's good enough for me!

-Terry

"Jochen K?pper <jochen at jochen-kuepper.de> wrote in message news:<m3y9f7hdc7.fsf at box.home.de>...
> On 27 Apr 2002 10:01:21 -0700 Terry Byrne wrote:
> 
> Thanks for your posting, but I don't understand the following:
> 
> Terry> Then I used the keyword "pass" whenever I ran an re since each
> Terry> line can contain only one error; after running an re it's not
> Terry> necessary to look for other error codes at all.
> 
> You are passing 'pass' to compile or search, or what are you doing
> here?  Is this documented anywhere?  I could neither find it in the
> manual nor the re HOWTO; maybe it's just me, though:O
> 
> Btw: An index would serve the HOWTO well:)) It is pretty hard to find
> information in there in a non-linear way (that is, without reading all
> of it over and over).  It's a nice document, but it is a little
> unorganized.
> 
> Greetings,
> Jochen


From aahz at pythoncraft.com  Wed Apr  3 19:16:11 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 19:16:11 -0500
Subject: IDE Wars, the prequel!
References: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com> <mailman.1017867011.28827.python-list@python.org>
Message-ID: <a8g60b$o7r$1@panix1.panix.com>

In article <mailman.1017867011.28827.python-list at python.org>,
Trent Mick  <trentm at ActiveState.com> wrote:
>
>Aside: Anybody know who maintains "http://www.python.org/editors/". Its
>comments on PythonWin are misleading.

Send proposed fix to webmaster at python.org and someone will handle it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From sheershion at mailexpire.com  Mon Apr 29 10:23:14 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Mon, 29 Apr 2002 14:23:14 -0000
Subject: Super-newbie syntax error question
References: <Cg%y8.5752$9F5.405701@typhoon.austin.rr.com>
Message-ID: <Xns91FFA4EA37280rcamesz@amesz.demon.nl>

Charlie Wood wrote:

> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "C:\PYTHON22\lib\rss2html.py", line 46
>     f = file("/article.txt", "w")
>     ^
> SyntaxError: invalid syntax


It's probably an indentation error. Alternatively, check the
preceding line carefully. 

To avoid indentation errors
   1 - Use eiter tabs or spaces, and be consistent about it.
   2 - Use the SciTE-editor: it can show indentation levels very
       clearly.

SciTE is also a very nice, albeit somewhat minimalist[*] -
general code editor, and a tiny download at about 350K.
Open source, too. Check http://www.scintilla.org .


Robert Amesz
-- 
[*] A relative term: the syntax colouring is excellent, and you can 
configure it for code completion, and in most cases you can run a 
Python script 'out of the box'. Double clicking on an error in the 
output window will move the cursor instantly to the offending line, 
even if the file isn't loaded at the time. There's more, but this
should suffise. The configuration dialogs *are* minimal, though: there 
are _none_: you'll have to edit the configuration files yourself.


From lenssen at hitnet.rwth-aachen.de  Sat Apr 20 15:24:04 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Sat, 20 Apr 2002 21:24:04 +0200
Subject: Parsing indention based language?
Message-ID: <a9sf96$ga6$1@nets3.rz.RWTH-Aachen.DE>

There's QML (Quest Markup Language, XML-based) for which I just written an
early Python port. Now I created an additional format called QL (Quest
Language, text-based) which can then be converted to QML (which is the
better storing format for interpreters). The Python language which I got to
know through the port was part in inspiring me to do an indent-based scoping
in QL. Now that I've written a nicely working but relatively ugly hack to
convert QL to QML (I used VBScript/ VisualBasic, the results can be
downloaded at http://questml.com/download.htm ), I was wondering if anybody
knows any elegant approaches.

I'm asking in this group first and foremost out of general interest and
because Python itself has similar rules, not because I immediately want to
start porting the QL-to-QML to Python (maybe some day) -- but I'd be happy
to get any hints written in Python.


QL, in short, looks like this (see http://questml.com/tutorial/ql.htm for
language details):

--------------------- Start below
"Battle sample", Philipp Lenssen
// Being a smaller sample of QL

start
    ---
        There's a big monster in front of you
        blocking the path.
        *This could be dangerous!*

    ( monkey.gif )

    % player skill = 10
    % player stamina = 10

    % enemy skill = 10
    % enemy stamina = 10

    $ after win = new path

    --> battle
        Let's start the battle
        _ player started battle2

battle
    % player power = [player skill] + {random 2, 12}
    % enemy power = [enemy skill] + {random 2, 12}

    ? [player power] = [enemy power]
        ---
            Both of your swords miss the other.
            Stamina stays the same.
        --> battle
            Battle again
    ? [player power] > [enemy power]
        ---
            You hit the enemy and wound him.
        % enemy stamina = [enemy stamina] - 2
        --> battle result
            Continue
    ...
        ---
            The enemy hits you and it hurts.
        % player stamina = [player stamina] - 2
        --> battle result
            Continue
--------------------- End above

According to the rules the above has to be converted to the following QML
(see http://questml.com/tutorial/ ):

--------------------- Start below
<quest>

<about>
    <title>Battle sample</title>
    <author>Philipp Lenssen</author>
</about>

<station id="start">
    <text>There's a big monster in front of you blocking the path.
        <em>This could be dangerous.</em>
    </text>

    <number name="player skill" value="10"/>
    <number name="player stamina" value="10"/>

    <number name="enemy skill" value="10"/>
    <number name="enemy stamina" value="10"/>

    <string name="after win" value="new path"/>

    <choice station="battle">Let's start the battle</choice>
</station>

<station id="battle">
    <number name="player power" value="[player skill] + {random 2, 12}"/>
    <number name="enemy power" value="[enemy skill] + {random 2, 12}"/>

    <if check="[player power] = [enemy power]">
        <text>Both of your swords miss the other. Stamina stays the
                same.</text>
        <choice station="battle">Battle again</choice>
    </if>
    <if check="[player power] greater [enemy power]">
        <text>You hit the enemy and wound him.</text>
        <number name="enemy stamina" value="[enemy stamina] - 2"/>
        <choice station="battle result">Continue</choice>
    </if>
    <else>
        <text>The enemy hits you and it hurts.</text>
        <number name="player stamina" value="[player stamina] - 2"/>
        <choice station="battle result">Continue</choice>
    </else>
</station>

</quest>
--------------------- End above

Thanks!


On a side-note, converting the other way round which I'm scripting now seems
fairly straightforward (I can recursively iterate through child elements by
just using XML DOM/ XSL XPath interfaces).




From logiplexsoftware at earthlink.net  Tue Apr  9 15:22:18 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 9 Apr 2002 12:22:18 -0700
Subject: wxPython / Your Experiences in large project ???
In-Reply-To: <a8u5nj$alq$01$1@news.t-online.com>
References: <a8rp2u$ht8$02$1@news.t-online.com>
	<mailman.1018291584.15367.python-list@python.org>
	<a8u5nj$alq$01$1@news.t-online.com>
Message-ID: <20020409122218.6b6b2c32.logiplexsoftware@earthlink.net>

On Tue, 9 Apr 2002 09:42:47 +0200
Harald Schneider wrote:

> Hi Cliff,
> 
> thanks for your answer! On what kind of clients do you run this software?
> My customer has an mixed Win2K, Win98-environment. Especially the Win98
> machines might cause stability probs when running wxPython. Do you have any
> experiences in this direction ?

We had (I no longer work there) a fairly homogenous environment (Windows NT 4.0
SP5) on all the PC's (Linux on the MySQL server).  Win2k should be no problem,
but I've avoided 9x like the plague that it is so I can't give you much insight
there.  I haven't heard of many problems on 98 but I have heard of issues on 95
(which IIRC were resolved by updates from MS).  FWIW, I wouldn't expect
wxPython to cause any additional problems beyond those already present on a 98
PC.

You might browse the wxPython archives to get a better feel for the issues
people have had with it in the past.

http://lists.wxwindows.org/pipermail/wxpython-users/

Also, if you are going to use wxPython, subscribing to its mailing list is a
*must*.  The things that have been said about the documentation are sadly true
(I consider the wxPython demo the primary source of documentation, with the
wxWindows docs being a more complete reference manual but not usable for
learning).

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From sholden at holdenweb.com  Tue Apr 30 13:46:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 13:46:10 -0400
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <aamebj$c2olc$1@ID-59885.news.dfncis.de> <1bzz8.82059$vF6.2440100@news2.tin.it> <jCzz8.17994$xy4.10082@atlpnn01.usenetserver.com> <sfAz8.82271$vF6.2450688@news2.tin.it>
Message-ID: <6OAz8.1251$Au5.843@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:sfAz8.82271$vF6.2450688 at news2.tin.it...
> Steve Holden wrote:
>         ...
> > What you say appears to mean is that iter(type, None), for example,
should
> > return an unbounded iterator of type. I'd be interested in what you have
>
> type is not callable without arguments.  iter(XX, None) gives an unbounded
> iterator for any XX which is a type callable without arguments: each call
> to the iterator's .next() yields [aka, each item in the iterator is] a
> "default-initialized" instance of type XX (same as XX() returns).
>
> > if-you-can't-call-a-callable-iterator-why-is-it-
> >         -called-a-callable-iterator-ly y'rs  - steve
>
> "Iterator that wraps a callable" might no doubt be a better name.
However:
>
> Much like a burger flipper is not (necessarily) a burger, so is a callable
> iterator not (necessarily) a callable.  "Callable" here is used as a noun
> (if any noun can be verbed, why can't adjectives be nouned...?).
>


In fact I had kind of made the meta-version of the mistake everyone is being
warned about now, which is to use a type's name as a variable: "type" was
intended to stand as a placemarker for "int", "float", etc., and I
completely overlooked

>>> type(type)
<type 'type'>

so I should probably have typed "iter(atype, None)" or similar. My usual
sloppiness. And, of course, to confound the confusion int() is *not* a call
to the integer class/type creator, but a no-argument call of the built-in
int() function. What happens to "int(x[, radix]) " when int is a built-in
type, by the way, or isn't that going to happen?

Anyway, just to finish up, the step I went wrong on was

>>> x = iter(int, 0)

So I decided instead to create my own type:

>>> class MyObj(object):
...         pass
...
>>> type(MyObj)
<type 'type'>
>>> x = iter(MyObj, None)
>>> x
<callable-iterator object at 0x0079A378>
>>> x()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: 'callable-iterator' object is not callable # aha!
>>> x.next()
<__main__.MyObj object at 0x007A6450>

So, I have to use the iterator protocol to get something out of this beast,
and since it's unbounded a simple for statement might not be a good idea.

>>> type(x.next())
<class '__main__.MyObj'>
>>> type(MyObj)
<type 'type'>
>>> type(MyObj())
<class '__main__.MyObj'>

That makes perfect sense. So callable-iterators can be used to produce
bounded or unbounded sequences of objects. If the type/class maintains
state, these objects can differ, otherwise they will all be the same (i.e.
each will be the result of calling the type's creator (the callable that was
provided when the iterator was created) with no arguments). Erk, that looks
a bit lisplike.

I was interested to find out that

>>> type()
Traceback ...
TypeError: type() takes 1 or 3 arguments

given the documentation for 2.2 says """
type(object)
Return the type of an object. The return value is a type object. The
standard module types defines names for all built-in types. For instance:
>>> import types
>>> if type(x) == types.StringType: print "It's a string"
"""

so-they-don't-call-it-a-callable-iterator-because-
        -it's-an-iterator-that's-callable-they-call-it-a-
        -callable-iterator-because-it-iterates-callables-
        -ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/








From guido at python.org  Mon Apr 22 15:17:53 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 22 Apr 2002 19:17:53 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <3CBB83B1.8EC2E2F6@python.org> <slrnabp8ub.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <cpofgbwnfa.fsf@pcp742651pcs.reston01.va.comcast.net>

> >phil hunt wrote:
> >> IMO until Python developers adopt a policy of being very reluctant
> >> to do this, many IT managers will be wary of using it in anything
> >> other thsan toy projects, and Python's user base will not grow as
> >> quickly as (IMO) it should.

[Guido]
> >If in your opinion Python should grow faster, perhaps you could
> >help out by contributing rather than by telling us what to do...

philh at comuno.freeserve.co.uk (phil hunt) writes:

> I would contribute, but you have just said that you won't accept my 
> proposed changes.

There are other ways to contribute besides by proposing language
features.  We need help doing code cleanup, backporting patches to
bugfix releases, documentation, and so on.

Also, I only told you about one very specific (and silly) change
you've proposed that I wouldn't accept it.  But I'd rather see you
help out in other ways than proposing language changes, since the
latter (even if they don't break anything) cause complaints about the
pace of change.

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


From alanmk at hotmail.com  Fri Apr  5 04:35:02 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 5 Apr 2002 01:35:02 -0800
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com> <3ouoaukmu32afgatil65cmcb7pvb8e5pcb@4ax.com> <a8i0pv$bu6$1@news.rchland.ibm.com> <MPG.171653199552480c989699@news.nwlink.com> <ulmc3cgm7.fsf@ctwd0143.fitlinxx.com>
Message-ID: <f72dac0d.0204050135.5b6fc286@posting.google.com>

Jeff Shannon:

> > A word of warning, though -- you may not be getting much usage of 
> > your second processor.  Python itself has a Global Interpreter 
> > Lock, forcing only one Python thread to be running at any given 
> > time.

David Bolen
> Actually, threads can certainly run on any of multiple processors,
> since they should be scheduled by the OS, but they can't be
> interpreting Python code simultaneously.  But that doesn't stop them
> from executing on different processors when they do get control of the
> GIL.  

An alternative approach which will make use of multiple processors is
jython, which, AFAIK, does not have a Global Interpreter Lock. So in
Jython, multiple python threads can run simultaneously on multiple
processors, as long as they don't access the same python objects.

Of course, you're unlikely to get a speed improvement with two
processors running jython threads versus one processor running cpython
threads, because jython is a virtual machine inside a virtual machine.

But I'd imagine that if you had 4+ processors, you'd start seeing
superior performance from jython ;-)

Regards,

Alan Kennedy.


From neal at metaslash.com  Mon Apr 22 19:07:30 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Mon, 22 Apr 2002 19:07:30 -0400
Subject: calendar.py doesn't use lists???
References: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <3CC497B2.41DFB81B@metaslash.com>

Aaron Watters wrote:
> 
> Hi guys.  I got an irksome problem.  It's possible that
> it's fixed in python 2.2.1, but even if it is, it irks me.

It is fixed.

Python 2.2.1 (#6, Apr 20 2002, 18:17:02) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import calendar
>>> print list(calendar.month_abbr)
['', 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
>>> print list(calendar.month_name)
['', 'January', 'February', 'March', 'April', 'May', 'June', 'July', 'August', 'September', 'October', 'November',
'December']


From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 19:50:27 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 23:50:27 +0000 (UTC)
Subject: currying (was: Re: A trivial question about print)
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com> <3CB756E4.5030404@jerf.org>
Message-ID: <slrnabesm3.mfc.huaiyu@gauss.almadan.ibm.com>

Jeremy Bowers <newsfroups at jerf.org> wrote:
>Has anyone done any work on creating a more general curry function? One 
>that might for instance curry the second of three positional arguments? 
>Using the above definitions:
>
>f2 = curry2(f, a, c)
>f2(b)

In general you must have a way to indicate a subset.  Like

g = curry([2, 3, 5], f, a1, a4, a6)
g(a2, a3, a5)

You might even allow something like
g = curry([2, 5, 3], f, a1, a4, a6)
g(a2, a5, a3)

Is this easy to do?  I don't know.  If it adds a for-loop for each function
call, it might be too costly.

Huaiyu


From iwk_n0spam at n0spam_xs4all.nl  Fri Apr 19 03:41:56 2002
From: iwk_n0spam at n0spam_xs4all.nl (iwk)
Date: Fri, 19 Apr 2002 09:41:56 +0200
Subject: OT: ViM compiled against Python 2.2.1 available?
Message-ID: <3CBFCA44.8010200@n0spam_xs4all.nl>

Hi there,

I've installed Python 2.2.1 on a new Windows 2000 install. The "problem" 
is that ViM 6.1 is compiled against Py2.1 so I cannot use any of the 
integration features I've gotten used to. On Linux I would compile it 
myself, but I don't have access to a compiler on Windows.  Are there 
versions of ViM6.1 availabe for download which are compiled aginst Py2.2.1?

Regards,

Iwan



From gcordova at hebmex.com  Wed Apr 17 12:34:14 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 17 Apr 2002 11:34:14 -0500
Subject: need help to choose an appropriate database
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062337FD494@mail.hebmex.com>

> 
> I use zope to access to a database, but I have to choose it.
> I mustn t store a lot of data so should i choose MySQL, PostGreSQL or
> anther one ?
> Which one is the most efficient ( speed...)
> 

Is this your sole consideration? Then go by MySQL, it's the fastest.

Will you (eventually) need transactions and such? Then go with
PostgreSQL; it has good performance and also has ACIDity.

Good luck.

-gustavo




From merman at freenet.de  Sat Apr 13 10:51:33 2002
From: merman at freenet.de (merman)
Date: Sat, 13 Apr 2002 16:51:33 +0200
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net>
Message-ID: <3cb843d0$0$4068$9b622d9e@news.freenet.de>

Hi Franz,

very polite of you - thanx for the good described hints/ examples.
I will try ;-).

best regards

Thomas


"F. GEIGER" <fgeiger at datec.at> schrieb im Newsbeitrag
news:a999hl$nmd$1 at rex.ip-plus.net...
> You have to do this by yourself, i.e. there's no built-in support for this
>
>
> class Abstract: # All subclasses of this have to conform to this
>    def __init__(self):
>       return
>
>    def method(self):
>       raise NotImplementedError # Oops, you are working with an Abstract
> object or you have forgotten to override method()!
>
>
> class Concrete:
>    def __init__(self):
>       Abstract.__init__(self)
>       return
>
>    def method(self):
>       pass # Your stuff goes here
>       return
>
>
> a = Abstract()
> a.method() # Will raise an error
>
> c = Concrete()
> c.method() # Will do, what you told it to do
>
>
> Try Google to find more on this as this is frequently discussed here.
>
> Best regards
> Franz
>
> "merman" <merman at freenet.de> schrieb im Newsbeitrag
> news:3cb7f47e$0$360$9b622d9e at news.freenet.de...
> > Hi,
> >
> > is there a way to simulate interfaces (like in java) in python?
> > ... or abstract classes? (c++/ java)
> >
> > o-o
> >
> > thomas
> >
> >
> >
> >
>
>




From jeff at ccvcorp.com  Mon Apr 15 14:03:48 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 15 Apr 2002 11:03:48 -0700
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3cb8a237@news.mhogaming.com>
Message-ID: <MPG.1724b5c6afb62a419896b0@news.nwlink.com>

In article <3cb8a237 at news.mhogaming.com>, 
Ken.Peek at SpiritSongDesigns.comNOSPAM says...

> ...  If you do not intend to operate the
> spacecraft in 'interactive mode', then why use a scripting
> language at all?

For the same reasons that one uses a "scripting" language in any 
other project -- speed of development, plus clearer, easier-to-
read code (thanks to less boilerplate).  Which leads directly to 
easily maintainable code and simpler software upgrades.

Note that the term "scripting language" is a very imprecise term, 
and it's arguable that it doesn't strictly apply to Python.  
Python is an interpreted general-purpose language, which is 
useful for scripting (among other things).  Then again, there are 
some people who use C as a scripting language...

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jeff at ccvcorp.com  Wed Apr 10 15:34:01 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 10 Apr 2002 12:34:01 -0700
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <slrnab8v1o.emj.philh@comuno.freeserve.co.uk>
Message-ID: <MPG.171e33893fed6df09896a7@news.nwlink.com>

In article <slrnab8v1o.emj.philh at comuno.freeserve.co.uk>, 
philh at comuno.freeserve.co.uk says...
> What have the PEPs ever done for us?
> 
> (Apart from the roads. The roads go without saying.)

... and the Aqueduct...

-- 

Jeff Shannon
Technician/Programmer
Credit International


From frr at easyjob.net  Tue Apr 23 11:40:54 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 17:40:54 +0200
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
Message-ID: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com>

Whenever a customer clicks on the order button, I'd like to:

a) Save some info into a database
b) Send the order to the payment processing company

I was planning to use a Python CGI for a. This cgi shouldn't return anything
and shouldn't change the contents of the browser. Is this possible? O:-)  

Besides, I'd like this script to return immediatelly, either if it can save or
not the info, so if the db server is down I don't lose the sale. How can I do
this? O:-)  I was thinking of using 2 scripts:

scripta: Forks scriptb and returns immediatelly
scriptb: saves the info to the db



-----------------------
Fernando Rodriguez


From careye at spamcop.net  Sat Apr 13 19:39:07 2002
From: careye at spamcop.net (Carey Evans)
Date: Sun, 14 Apr 2002 11:39:07 +1200
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com> <xduhemf464r.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <pan.2002.04.14.11.39.06.845014.6872@spamcop.net>

On Sun, 14 Apr 2002 09:53:56 +1200, quoth Ralf Juengling,

> Can anybody explain why the special method 'i.__iadd__' does not exist,
> while 'i+=1' is a legal expression ('i' being a int)?

In the absence of __iadd__, += uses __add__ instead.  The end result is
similar to:

    try:
        iadd = x.__iadd__
    except AttributeError:
        iadd = x.__add__
    x = iadd(y)

There's a bit more logic to enable addition of different types (e.g. int
with float) which is described in the reference manual at:

    http://www.python.org/doc/current/ref/numeric-types.html

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

			     Cave canem.


From zopestoller at thomas-guettler.de  Thu Apr 11 11:08:41 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 11 Apr 2002 17:08:41 +0200
Subject: Readline on Window
Message-ID: <3CB5A6F9.2070205@thomas-guettler.de>

I use:
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32

C-a results in ^A. But I would like to go the beginning of the line

If I use cygwin. Importing Readline brings an error:
Python 2.2 (#1, Dec 31 2001, 15:21:18)
[GCC 2.95.3-5 (cygwin special)] on cygwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import readline
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
ImportError: dlopen: Win32 error 126

Has someone a hint how I can get it working?

  thomas



From mwh at python.net  Thu Apr 18 09:32:16 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 13:32:16 GMT
Subject: Counterintuitive Python behavior
References: <mailman.1019052695.4161.python-list@python.org> <lksn5uqtsv.fsf@pc150.maths.bris.ac.uk> <lkofgiqteo.fsf@pc150.maths.bris.ac.uk> <9pgv8.30868$vF6.957894@news2.tin.it>
Message-ID: <lkk7r5qgbk.fsf@pc150.maths.bris.ac.uk>

Alex Martelli <aleax at aleax.it> writes:

> Michael Hudson wrote:
> 
> > Michael Hudson <mwh at python.net> writes:
> > 
> >> I was hoping more people than the OP would find it useful.  I'll
> >> probably knock up a version for my web pages (when the starship
> >> reappears...).
> > 
> > I should say that Alex's answer is well worth referring to too.  I
> > wonder if he would mind me incorporating it into aforementioned
> > webpage...
> 
> Sure!  Go right ahead.

OK, bits of this thread are now 22k of xhtml1-strict-y goodness at:

    http://starship.python.net/crew/mwh/hacks/objectthink.html

Please read and tell me about typos, misattributions and any other
sillynesses.

Cheers,
M.

-- 
  I'm about to search Google for contract assassins to go to Iomega
  and HP's programming groups and kill everyone there with some kind
  of electrically charged rusty barbed thing.
                -- http://bofhcam.org/journal/journal.html, 2002-01-08


From donn at drizzle.com  Tue Apr  9 02:02:47 2002
From: donn at drizzle.com (Donn Cave)
Date: Tue, 09 Apr 2002 06:02:47 -0000
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com> <1018304015.46066@cswreg.cos.agilent.com> <MPG.171bbba6ee73df8c98969f@news.nwlink.com> <mailman.1018310800.31349.python-list@python.org> <MPG.171bd911b760bfa19896a0@news.nwlink.com>
Message-ID: <1018332163.664704@yasure>

Quoth Jeff Shannon <jeff at ccvcorp.com>:
...
| I obviously can't speak to your specific designs, not having seen 
| them, but I can't imagine how globals()/locals() would be 
| necessary unless you have a large amount of interdependencies in 
| your code, and it is generally accepted that interdependent code 
| is not a good thing.  If you can explain to me how you can 
| require access to large portions of another scope's namespace 
| without having large interdependencies, then perhaps I'll change 
| my mind about this method.

I use it during class definition, to add a set of shadow functions.
In a class definition, the methods etc. are in locals().  I compute
further methods by looking at the existing methods, so I must pass
locals() to the function that does that.

You don't have to love the idea, but if there's anything horrifying
about, it's not that I'm passing locals() to a function.

	Donn Cave, donn at drizzle.com


From aleax at aleax.it  Wed Apr 17 03:26:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 07:26:36 GMT
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <7xhemb744m.fsf@ruckus.brouhaha.com>
Message-ID: <Ms9v8.22165$b62.626694@news1.tin.it>

Paul Rubin wrote:

> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
>> Is it possible to have a safe eval() on the server-side even when the
>> string consists of submitted user data?
> 
> Well, look at the docs for the rexec (restricted execution) and Bastion
> classes, but basically no, there's no safe way.  Rexec doesn't stop an
> infinite loop from hosing your cpu.

If you have to watch for that, forking and running the DoS risk in the
child process can help -- the parent process needs to keep an eye
over the resources (mostly, time) being used by the child process and
be ready to kill it if need be.  Simple, it ain't.


> Why do you want to eval strings from untrusted users?

Can't answer for the OP, but applets/plugins are a nice concept for
a larg app, and it's nice, if feasible, to let end-users just download
and try plugins from 3rd parties if a way can be found to avoid having
the end-user examine and validate them.  In such a situation the risk
of having an errant or malignant plugin get into a loop might even be
acceptable, as long as no damage can result to persistent data and
the end-user can still hit some button to terminate operations.


Alex



From phr-n2002a at nightsong.com  Mon Apr 15 06:55:13 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 03:55:13 -0700
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain>
Message-ID: <7xvgatb59q.fsf@ruckus.brouhaha.com>

hzhu at mars.localdomain (Huaiyu Zhu) writes:
> I'm interested in what people mean when they say "free software" or "open
> source software".  I might also be interested in hearing the justification
> given for different preferences, but not before the speaker makes clear what
> he means by these terms.

This is explained at length in 

  http://www.gnu.org/philosophy/free-software-for-freedom.html

> If you know that there are at least two movements that do not even agree
> on the usage of terms, it is somewhat pointless in starting a sales pitch
> without identifying which movement you are pitching for and how its terms
> differ from the others.

It's pretty clear to me that Stephen is pitching from the Open Source
perspective.  The Free Software side is generally pro-GPL.

This article (linked from the gnu.org url above) is also pretty good:

  http://www.itworld.com/AppDev/350/LWD010523vcontrol4


From ajs at ix.netcom.com  Wed Apr 10 03:58:26 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Wed, 10 Apr 2002 03:58:26 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000001c1e067$76889380$0334fea9@carol>

(This post may be in duplicate. If so apologies.)

Cliff writes -

>This was a direct attack on Guido's integrity, which was >uncalled for. You
*assumed* much about what he meant
>(perhaps he didn't make himself clear, but
>that's hardly grounds for labeling him a dictator) and when he >clarified
his position he became a liar.

I didn't label him a dictator. As far I understand he has labeled himself
such.  And the fact that he has is wonderful and honest - because he is and
should say so, and there is general agreement that software of this nature
cannot be designed by committee.

And being dictator he has the unique ability to issue the LION PEP. The LIKE
IT OR NOT PEP.

If Guido is concerned  - lets say -  about any code breakage issues he might
have missed lets limit discussion to that.

People don't waste time making arguments that will have no impact, Aahz
(oops, forgot we were suppose to be being nice) doesn't get to play friend
of the court, and at the end of the day folks decide to use or not use
Python, LION PEP included.

Clean and simple - and easier on everyone's treads.

And thanks, Cliff, for your thread.

Art





From maxm at mxm.dk  Fri Apr  5 02:36:12 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 09:36:12 +0200
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <3CAAC1FB.2050105@mxm.dk> <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com>
Message-ID: <3CAD53EC.1010703@mxm.dk>

Terry Reedy wrote:

> "Max M" <maxm at mxm.dk> wrote in message news:3CAAC1FB.2050105 at mxm.dk...
>>Through Zope I have used Structured text extensively. Also I have

> Just curious: are you willing to (consider, at least) switching to
> ReST, or are we going to end up with two similar standards used by
> different parts of the Python community?  IE, do Goodger's arguments
> about why the differences from ST are improvements cut any ice with
> you?

What I don't understand here is that it seems that people considers 
structured text to be difficult. It really isn't. Any programmer can 
learn the basics in 15 minutes or less.

And it really is very very easy to write documents in it. I do belive 
that if I where ever to write a book (again) i would write it in ReST.

It find it rather ridiculous to use the JavaDoc approach with using 
markup in the docs, as it offers allmost nothing practial that ReST 
dosn't for solving this particular problem.

And as far as I can tell from the specs, ReST solves most of the 
problems I have had with structured text. A few things gets a little 
more complicated and some get simpler. All in all a good bargain.

I would certainly use it. I also hope that it gets integrated into Zope.

regards Max M


########################################
# 'old' structured text
########################################

This is an example of a headline in structured text

     Here is a paragraph.

     This is a second paragraph with a link to 
"http://www.python.org":http://www.python.org

     Here follows a piece of Python code::

         for i in range(10):
             print 'hello world'

     That is really all there is to it.

     Unless you want to use this paragraph headline

         That subcedes this paragraph




########################################
# ReST
########################################

This is an example of a headline in structured text
===================================================

Here is a paragraph.

This is a second paragraph with a link to http://www.python.org

Here follows a piece of Python code::

     for i in range(10):
         print 'hello world'

That is really all there is to it.

Unless you want to use this paragraph headline
----------------------------------------------

That subcedes this paragraph



########################################



From DeepBleu at DeepBleu.org  Thu Apr 25 16:04:46 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Thu, 25 Apr 2002 15:04:46 -0500
Subject: ftplib and attributes
References: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>
Message-ID: <ucgoaehao5o8c@corp.supernews.com>

"LoGan" <giannivolontosky at TOGLIyahoo.it> wrote in message
news:pan.2002.04.25.21.35.03.529909.14114 at TOGLIyahoo.it...
> hi, im new bie and im trying to use ftplib, but my problem is to get the
> attributes of the files, how can i do ?
> LoGan

What attributes of what files?  The remote files on the server you are
FTPing?  If so:
>>> from fttlib import FTP
>>> ftp = FTP(host, username, password)
>>> ftp.dir()

This last one will give you a list of files of the active directory on the
remote server and their attributes.  Do you want to save this list?

>>> flist = []
>>> ftp.dir(flist.append)
>>> for l in flist:
>>>   print l
Also, even though you are a newbie, you should learn how to check the docs
and search the web.  All this above you can find.
DeepBleu




From mlh at vier.idi.ntnu.no  Tue Apr 16 09:58:03 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 16 Apr 2002 13:58:03 +0000 (UTC)
Subject: input() vs raw_input()
References: <slrnabm6im.a8q.ak@ak.silmarill.org>
Message-ID: <slrnabobfb.8ho.mlh@vier.idi.ntnu.no>

In article <slrnabm6im.a8q.ak at ak.silmarill.org>, Andrei Kulakov wrote:
>input() must not be mentioned in any tutorials for new users. It belongs
>somewhere in the depths of the thickest reference manual. Where do new
>users *learn* about it? I don't remember seeing it in any tutorial and
>yet it pops up all the time.

I use it in my tutorials. I don't see what the fuss is about. You have
to explain how it works (at least in a limited way) but so what?

> - Andrei

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From python at rcn.com  Mon Apr 29 01:34:58 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 01:34:58 -0400
Subject: Comment on draft PEP for deprecating six builtins
Message-ID: <aaim48$7fd$1@bob.news.rcn.net>

Please send me comments on a draft PEP deprecating
and relocating six built-in functions.  The text file is at:
http://users.rcn.com/python/download/deppep.txt

It is also listed below.


Raymond Hettinger

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

PEP: XXX
Title: Deprecate Six Built-ins
Version: $Revision: 1.1 $
Last-Modified: $Date: 2002/04/28 19:42:56 $
Author: python at rcn.com (Raymond D. Hettinger)
Status: Draft
Type: Standards Track
Created: 28-Apr-2002
Python-Version: 2.3
Post-History:


Abstract

    This PEP proposes to compact the list of built-in functions through
    relocation and deprecation.  map(), reduce() and filter() would move to
a
    separate module for functionals.  pow() and divmod() would move to into
    the math module.  input() would be eliminated altogether.



Rationale

    There are separate motivations for each function.  One motivation in
    common is that the number of total built-ins should be as small as
    possible to limit the amount of "core language" one needs to know in
order
    to read code that is not module specific.

    The functionals (map, filter, and reduce) had greater importance prior
to
    the introduction of list comprehensions which are now the preferred (and
    more readable) approach.  Limits on the number of built-in functions
    prevented the addition of other functionals which may have be useful
when
    that style of programming is needed.  Creating a separate module allows
    room for a large variety of functionals (e.g. fold, head, tail, take,
    drop, etc.) to be introduced without cluttering the built-in namespace.

    Pow() and divmod() are more related to math module functions than any
    of the built-in functions.  In general, they are rarely used.  Because
    the ** operator is available, the pow() function is rarely called
    directly.

    Input() is easily replaced by eval(raw_input()).  It has few legitimate
    uses and presents a security risk in most programs where it is used.
    Though the risks are well documented, experience on python-tutor and
    help at python.org indicates that new users frequently use input() when
    raw_input() was intended and that they are unaware of the embedded
eval().



Key Issue

    All of these functions have been in-place since the beginning, are used
    in mountains of code, and appear in every Python book.  The situation
    would appear to be immovable.

    I propose that a simple means be provided to re-enable those functions
at
    will when they are needed to run old code.  Add a command line option -b
    and an equivalent environment variable, PYTHONBUILTINRESTORE which would
    start-up Python by running:

        from functionals import map, filter, reduce
        from math import pow, divmod
        from deprecated import input



Implementation and Deprecation Timetable

    The module additions would go into Python 2.3.  All six functions would
be
    left intact and raise deprecation warnings in Python 2.3 and 2.4.  The
six
    functions would be removed in Python 2.5 (having allowed two versions
and
    a full year for code updates).  The -b command line option and related
    environment variable would go into Python 2.3 and remain available
    in perpetuity.  Like "from __future__" this mechanism can be used for
    other deprecations, allowing the language to evolve without accumulating
    clutter.



Copyright

    This document has been placed in the public domain.



Local Variables:
mode: indented-text
indent-tabs-mode: nil
fill-column: 70
End:







From narnett at mccmedia.com  Tue Apr  2 21:24:06 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Tue, 2 Apr 2002 18:24:06 -0800
Subject: A faster Python?
In-Reply-To: <r1tq8.161131$af7.78459@rwcrnsc53>
Message-ID: <NDBBLAEMOMEHKALAMKKJKEOPJOAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Emile van Sebille
> Sent: Tuesday, April 02, 2002 5:37 PM
> To: python-list at python.org
> Subject: Re: A faster Python?

> This surprised me.  I like python with Numeric exactly because it's
> faster and processes more data than Excel.  I'm now regularly
> calculating and delivering pivot-style results from a million row source
> with 240 columns.

Sorry, I should have made it clear that I haven't made direct comparisons.
I'm still working on data structures and such; I haven't yet seen where the
bottlenecks might be.  In other words, my comments were entirely
theoretical -- I was assuming that there are things that Excel, SPSS and
other commercial tools can do that will be significantly faster.  My
assumption is that I'll see this in in some highly iterative operations.

I was suggesting that if anyone is working on speeding up Python, those are
the areas where I never have enough speed.

And I'm quite encouraged by the experience you've shared.

Nick




From frr at easyjob.net  Wed Apr 10 04:29:02 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Wed, 10 Apr 2002 10:29:02 +0200
Subject: Parsing apache log files
Message-ID: <4st7bu4j2483nqe1fm8vb7utfm48rqeqd4@4ax.com>

Is there any module for parsing apache log files? O:-)  

I guess it could be done with regular expressions, but I'm not very gifted
with them... :-)




-----------------------
Fernando Rodriguez


From bernie at 3captus.com  Tue Apr  2 04:56:28 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Tue, 02 Apr 2002 09:56:28 GMT
Subject: IDLE for SUSE Linux?
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com> <3CA8D724.336A22BA@3captus.com> <5waq8.5401$ku5.476012@e3500-atl2.usenetserver.com>
Message-ID: <3CA98091.4B8C83A2@3captus.com>


Robert Oschler wrote:
> 
> Bernard,
> 
> Then I have something strange.  There is no /Tools directory under
> /usr/share/doc/packages/python.  I tried doing a search on Python in Yast2
> and installing a bunch more of the Python related packates (and online
> updated them), but still no luck.  A full disk search for Idle.Py turned up
> empty.
> 
> thx
> 

Have you installed python-tk?  That maybe the reason.  IDLE needs
Tkinter.


Bernie

-- 
There are three schools of magic.  One:  State a tautology, then ring
the changes on its corollaries; that's philosophy.  Two:  Record many
facts.  Try to find a pattern.  Then make a wrong guess at the next
fact; that's science.  Three:  Be aware that you live in a malevolent
Universe controlled by Murphy's Law, sometimes offset by Brewster's
Factor; that's engineering.
                -- Robert A. Heinlein


From simon at incutio.com  Fri Apr 12 05:43:32 2002
From: simon at incutio.com (Simon Willison)
Date: Fri, 12 Apr 2002 10:43:32 +0100
Subject: Problems accessing the new Google Web APIs from Python
References: <3CB6A5EA.9010804@incutio.com>
Message-ID: <3CB6AC44.6050806@incutio.com>

Simon Willison wrote:

> Google have just launched a new service - Google Web APIs - which 
> allows developers to query the google search engine via SOAP for 
> non-commercial use (you're allowed up to 1000 queries a day). To use 
> the service you have to sign up for a license key (all they ask for is 
> an e-mail address and password) and download the API package:
>
> http://www.google.com/apis/
>
> Anyway, I've been trying to get this working with Python (having never 
> used SOAP before) and I'm having problems, but I'm not sure if it's 
> their service going down or a problem with my code.

Problem solved - I found this page which has full instructions on how to 
get it working:

http://diveintomark.org/archives/2002/04/11.html








From rhhyatt at charter.net  Sun Apr 28 11:11:36 2002
From: rhhyatt at charter.net (Bob Hyatt)
Date: Sun, 28 Apr 2002 08:11:36 -0700
Subject: Call for kjBuckets.c [was Re: relational database?]
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <5yEp8.5051$hi7.18667@news-server.bigpond.net.au> <3CA7192C.E1C8974A@engcorp.com> <mailman.1017612966.7043.python-list@python.org>
Message-ID: <uco40vl11ipj04@corp.supernews.com>

Richard,

I have had a fruitless search for Gadfly. As you noted, the website for
Chordate is not responding.

I would prefer to use all python tools, even though the db processing might
be slower.

I would appreciate notification.

I will also monitor comp.lang.python.announce

Good luck,

Bob Hyatt


"Richard Jones" <rjones at ekit-inc.com> wrote in message
news:mailman.1017612966.7043.python-list at python.org...
> On Mon, 1 Apr 2002 00:11, you wrote:
> > Trung Hoang wrote:
> > > "Martin v. Loewis" <martin at v.loewis.de> wrote in message
> > > news:m3eli1m2dt.fsf at mira.informatik.hu-berlin.de...
> > >
> > > > "Trung Hoang" <thoa0025 at mail.usyd.edu.au> writes:
> > > > > What relational database is easy to setup and use with python?
> > > >
> > > > Gadfly.  It is a pure-Python database, and requires little
> > > > administrative interaction.  It does not provide good performance,
> > > > though...
> > >
> > > where can i get this ?
> >
> > Try Google?
>
> Actually, that's not going to help much - last I looked www.chordate.com
> wasn't answering its calls. I'm currently in the process of repackaging
> gadfly, including the updated code from the Zope version (regex->re
> conversion mostly).
>
> There'll be the usual sort of announcement when I'm done.
>
> Does anyone have a copy of kjbuckets.c that they could send me?
>
>
>     Richard
>




From paul at boddie.net  Fri Apr 19 10:08:35 2002
From: paul at boddie.net (Paul Boddie)
Date: 19 Apr 2002 07:08:35 -0700
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <E4Iv8.2972$8D3.78501@news1.tin.it> <just-2DF829.09350419042002@news1.xs4all.nl>
Message-ID: <23891c90.0204190608.11c70ee4@posting.google.com>

Just van Rossum <just at xs4all.nl> wrote in message news:<just-2DF829.09350419042002 at news1.xs4all.nl>...
> 
>     len(o)

[vs]

>     o.keys()
> 
> This seems quite arbitrary, even to me, after maybe 6 years of working 
> with Python. If o.__len__() were spelled o.len(), we could easily do 
> without the len() builtin. (Note that I don't argue we could do without 
> functions: I wouldn't want to miss them for the world...)

I suppose that len is considered near-universal, although len(123)
reveals a notable exception, whereas very few objects bear the notion
of having "keys". Therefore, near-universal "protocols" (as I believe
someone once called them) are supported through built-in functions,
and other "protocols" are supported more or less exclusively through
methods.

Of course, providing len as a method on all appropriate objects would
be more consistent, but as far as Python is concerned right now, it
seems redundant and potentially confusing: "But I thought len was a
function!" exclaimed the beginner.

Paul


From trentm at ActiveState.com  Fri Apr 26 20:33:05 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 26 Apr 2002 17:33:05 -0700
Subject: Python.NET?  Jython.NET?  Ni!
In-Reply-To: <RGly8.3497$Z6.36288@nnrp1.uunet.ca>; from bruce_dodson@bigfoot.com on Fri, Apr 26, 2002 at 08:47:58PM -0300
References: <RGly8.3497$Z6.36288@nnrp1.uunet.ca>
Message-ID: <20020426173305.A13797@ActiveState.com>

[Bruce Dodson wrote]
> Is ActiveState's Python for .NET dead?  I can find no reference to it on the
> ActiveState page, other than an old research build for .NET Beta 2.

Look to the PyWin32 CVS tree. A large checkin under the "dotnet"
directory recently went in.

http://mailman.pythonpros.com/mailman/listinfo/win32-checkins
http://starship.python.net/crew/mhammond/dotnet/
http://www.pythonpros.com/cgi-bin/cvsweb.cgi/PyWin32/dotnet/

I don't know if that means anyone is working on it, though.

Cheers,
Trent


-- 
Trent Mick
TrentM at ActiveState.com




From pinard at iro.umontreal.ca  Thu Apr 25 11:41:50 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 25 Apr 2002 11:41:50 -0400
Subject: Topy news - still about pretty-printing
In-Reply-To: <oqadsam0de.fsf@carouge.sram.qc.ca>
References: <oqadsam0de.fsf@carouge.sram.qc.ca>
Message-ID: <oqlmbbpyyp.fsf@titan.progiciels-bpi.ca>

Hi, gang!  Here are a few more Topy news, for those who are interested.

Quick reminder: Topy is a tool meant to help converting from other languages
to Python.  This is a project under development, it is not usable yet.

Two weeks ago, I asked this list about pretty printers for Python.  Finding
none, I finally made a good stab on this problem, or at least, I think so.
The line wrapper which I temporarily put in the back-end is now gone.

Generating pretty printed Python sources has been an interesting problem
in Topy.  There were two main difficulties.  The first is that, for doing
a good job, Topy needed a more extensive representation of expressions,
right in the tree structure allowing front-ends to communicate with the
back-end, so I extended those structures as needed.  The second is that, to
stay speedy enough, generation should be interrupted as soon as we go over
line width, and when one wanders in trees while maintaining stacked state,
properly and efficiently backtracking everything required some thought :-).

Other news are that the Perl front-end now uses PLY instead of SPARK,
(the other front-ends still use SPARK).  I made small stunts so that PLY
diagnostics are precise to the column, to be compatible with the rest.
I also merged `miltopy', `pltopy' and `scmtopy' into a single `topy'.

Speaking of, I thought for a moment that if Topy had a Python front-end
(the back-end is already for Python), then Topy would offer a Python
pretty printer for free.  This would requires two things.  First, the
structure of trees for communicating between front-ends and the back-end
has be boosted, once more.  Second, I need a scanner and parser for Python.
I thought merely using the tokenizer/parser modules provided with Python,
and even experimented a bit with them.  The resolution is rough to lines,
Not columns, so some hackery wound be needed here as well (this is not
only for diagnostics, but for proper re-insertion of comments).  Also,
ASTs produced by the Python parser are surprisingly noisy, Topy would need
more code and overhead for cleaning these.  This all means that having a
Python reformatter in Topy would require more work than I expected.  Sigh!

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




From mal at lemburg.com  Tue Apr  9 11:29:35 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 09 Apr 2002 17:29:35 +0200
Subject: Newbie - web app deployment issues
References: <a8skcd$mr5$1@hfc.pacific.net.hk> <23891c90.0204090043.287c317f@posting.google.com>
Message-ID: <3CB308DF.478DCD8E@lemburg.com>

Paul Boddie wrote:
> 
> [1] mxCGIPython is a good start for those wanting a binary, but I then
> discovered the FreeBSD "ports" system (see http://www.freebsd.org) and
> it's possible to find lots of binaries for download in .tar.gz format,
> including various Python extension modules (MySQLdb, for example) -
> it's very convenient indeed!

BTW, these extensions should now work with updated mxCGIPython
as well: I've included support for importing shared modules,
which was not previously possible.

Oleg Broytmann has provided binaries for FreeBSD, so getting
it installed on an ISP machine with FTP-only access should
be possible now.

-- 
Marc-Andre Lemburg
________________________________________________________________________
Business:                                        http://www.lemburg.com/
Python Pages:                             http://www.lemburg.com/python/




From jsmullyan at wnyc.org  Thu Apr 11 12:16:22 2002
From: jsmullyan at wnyc.org (Jacob Smullyan)
Date: Thu, 11 Apr 2002 12:16:22 -0400
Subject: [Skunkweb-list] [ANNOUNCE] SkunkWeb 3.2 Released!
References: <mailman.1018388802.21863.python-list@python.org>
Message-ID: <pan.2002.04.11.12.16.21.588655.31121@wnyc.org>

On Tue, 09 Apr 2002 17:45:03 -0400, Karl Putland wrote:

> Are there any changes to PyDO since the separate release of 1.1b1? I
> didn't see anything new in CVS view on SF.

Right.  We haven't changed PyDO since 1.1b1, which SkunkWeb 3.2 includes.
The beta label doesn't mean anything at this point; since we haven't made
any changes since the beta, there wasn't much point in making a new
release.

js


From kalphonso at diamonddata.com  Fri Apr 19 18:04:11 2002
From: kalphonso at diamonddata.com (Keith Alphonso)
Date: Fri, 19 Apr 2002 17:04:11 -0500
Subject: Integrating C/C++ in to python/Zope
Message-ID: <a9q49n$tq7$1@news.datasync.com>

Can anyone point me to documentation explaining how to integrate my C or C++
with Python, specifically for Zope?  I can't seem to find anything
difinitive anywhere.

Do I basically have to modify the Python interpreter source code to put my
code with it?  Is there any mechanisms for linking in dynamic libraries?

Are there some pre-defined C++ classes that work as Python classes that I
can extend?

thanks!





From guido at python.org  Wed Apr  3 12:12:36 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 12:12:36 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 10:44:01 CST."
             <Pine.LNX.4.33.0204031031440.16190-100000@server1.lfw.org> 
References: <Pine.LNX.4.33.0204031031440.16190-100000@server1.lfw.org> 
Message-ID: <200204031712.g33HCaP22228@odiug.zope.com>

> The addition of another type really does change the set of preferred
> truth values from {0, 1} to {0, False, 1, True}.  The typical
> practicing Python programmer who wants to say "true" would have to
> decide -- for each instance -- whether to say "1" or "True".

No.  If they want to use the *recommended* values they should always
use False or True.  0 and 1 are relegated to the domain of
non-representative values of the set of false and true objects, just
as None and "true" (the string literal).

> And the programmer cannot happily ignore the issue and stick to
> using 0 and 1, because if comparisons start to return True and
> False, dealing with a mix of four truth values is unavoidable.  This
> PEP makes me uneasy because it causes me to see lots of extra casts
> in Python's future (not to mention confused beginners).

You misunderstand it.  Strings, lists, numbers and so on are still
acceptable as truth values, and when you want to know whether x is
true or false, you still have to say "if x:" -- never "if x == True:".

In addition, for backward compatibility 0 == False and 1 == True, so
mixing them shouldn't be a problem if you know your values are
chosen from the set {0, False, 1, True}.

If you find yourself comparing an unknown value x to a known truth
value to determine whether x is true or false, stop right there:
you've *already* got a problem.  If you don't have a problem because
it works in practice (meaning x is known to be 0 or 1) the backwards
compatibility of the PEP kicks in and saves your butt.  This won't
be deprecated: the PEP promises that False==0 and True==1 will always
hold.  (There's a question for reviewers about this, but consider it
decided -- bools will always mix freely with ints.)

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




From eppstein at ics.uci.edu  Sun Apr 14 16:52:16 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 13:52:16 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <eppstein-7156A5.13521614042002@news.service.uci.edu>

In article <m31ydiy7b9.fsf at mira.informatik.hu-berlin.de>,
 martin at v.loewis.de (Martin v. Loewis) wrote:

> As Bengt points out, it is tricky to define this analogous to 'and'
> and 'or'. Both have a shortcut meaning, evaluating the second one only
> if the first does not determine the outcome. Furthermore, they both
> return the last expression evaluated. This is not possible for 'xor':
> you have to evaluate both, and there is no clear choice as to which
> one is better as a result - in fact, the OP wants it to return bool.
> 
> So xor would be very much *unlike* 'and' and 'or'.

If (a xor b) is true, it must be because exactly one of a and b is true.
So, in that case, it seems reasonable to return the one that is true.
I'm not sure what value would make most sense to return when (a xor b) 
is false, though.

I.e. (a xor b) could expand to something like
    (not b and a) or (not a and b)
but without all the repetition...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From logstx at bellatlantic.net  Sat Apr 20 15:04:16 2002
From: logstx at bellatlantic.net (logistix)
Date: Sat, 20 Apr 2002 19:04:16 GMT
Subject: Python For Win32 Extensions documentation in HtmlHelp format.
References: <3CC031B7.20104@skippinet.com.au>
Message-ID: <QYiw8.3824$uV.196@nwrddc01.gnilink.net>

Great job!

The .chm files are the only reason I uninstalled 2.2 and reinstalled the
ActiveState version when it came out.

Thanks for taking the time to do something that doesn't fit on the
fun/cool/cutting-edge list of things to do.

--
-

"Mark Hammond" <mhammond at skippinet.com.au> wrote in message
news:3CC031B7.20104 at skippinet.com.au...
> Hi all,
>    I have decided to abandon Windows Help files, and move to the much
> better HtmlHelp format (as used by ActivePython).  There is a single
> .chm for all the extensions, meaning all links should correctly work,
> and a single index.
>
> It will appear in the next win32all, but in the meantime you can
> download the current version via
> http://starship.python.net/crew/mhammond/win32/Downloads.html
>
> Mark.
>




From pyth at devel.trillke.net  Thu Apr 25 20:34:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 26 Apr 2002 02:34:28 +0200
Subject: python and very large data sets???
In-Reply-To: <20020425223536.GG21205@teich.Garten.DigitalProjects.com>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6o7l$nbi$1@panix3.panix.com> <20020424172403.GE20209@devel.trillke> <20020425223536.GG21205@teich.Garten.DigitalProjects.com>
Message-ID: <20020426003428.GD22555@devel.trillke>

On Fri, Apr 26, 2002 at 12:35:36AM +0200, Alexander Skwar wrote:
> ?holger krekel? sagte am 2002-04-24 um 19:24:03 +0200 :
> > - setting up/configuring an appropriate database
> >   for the task (may not be easy)
> 
> Hm, I don't think so.  Setting up a basic MySQL server without any
> tunings is easy as pie, even on Windows.  But he'll of course need to
> know SQL, at least in some simple forms - but SQL knowledge is often
> needed, isn't it?

it is quite easy to set up a database e.g. for a website but 
is the same true if you have 100 GB of data being processed?
Easy as pie? Maybe i am underestimating mysql...

	holger




From phil at river-bank.demon.co.uk  Mon Apr 29 06:15:34 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Mon, 29 Apr 2002 11:15:34 +0100
Subject: ANN: Python 2.2.1 for Sharp Zaurus
Message-ID: <3CCD1D46.B519567@river-bank.demon.co.uk>

Riverbank Computing is pleased to annouce the release of the port to
Python v2.2.1 to the Sharp Zaurus. We believe the port will also run
unchanged on the Compaq iPAQ (running Linux).

The download page is
http://www.riverbankcomputing.co.uk/zaurus/download.php

This Python port has been tailored to reduce the storage requirement for
PDAs:

- removal of modules orientated to development and servers

- re-use of the zlib code included in Qtopia

- reduction in module inter-dependencies (eg. the core does not require
distutils)

- modules split into separately installable "collections".

Phil




From news at 92tr.freeserve.co.uk  Sun Apr  7 18:32:56 2002
From: news at 92tr.freeserve.co.uk (ddoc)
Date: Sun, 7 Apr 2002 23:32:56 +0100
Subject: purging file buffers - to make new data be written to file
Message-ID: <a8qhjp$7go$1@news6.svr.pol.co.uk>

Can I do that from within a python script, or do I have to give a command 
line argument to make Python start without buffers?

On Linux.  The program is to save a memo in a slightly organised format, 
but I find the file is not written until Python exits.

-- 
A


From barry at zope.com  Wed Apr 10 17:05:44 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Wed, 10 Apr 2002 17:05:44 -0400
Subject: [mimelib-devel] RELEASED email 2.0 (standalone)
References: <15538.3848.65410.460971@anthem.wooz.org>
Message-ID: <15540.43304.411729.293307@anthem.wooz.org>

>>>>> "BAW" == Barry A Warsaw <barry at zope.com> writes:

    BAW> I've just released version 2.0 of the standalone email
    BAW> package.  This will likely be the last official standalone
    BAW> release, as I intend to complete the merge of the package
    BAW> into the Python 2.3 cvs tree.

Quick followup: I just released version 2.0.1 which fixes a few
problems in the test suite.  There's also one minor API change that
syncs this version with the version that's now in the Python 2.3 cvs
tree.

It is recommended that you grab the 2.0.1 version, but it's not
urgent.

-Barry




From jimd at vega.starshine.org  Fri Apr  5 14:05:09 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 5 Apr 2002 19:05:09 GMT
Subject: Tutoral < ? < Language Reference
References: <3281a460.0204050705.4b9b595f@posting.google.com>
Message-ID: <a8ksh5$2eqh$1@news.idiom.com>

In article <3281a460.0204050705.4b9b595f at posting.google.com>, P Adhia wrote:

> Hi,

> I think I have graduated to the next level after reading the tutorial.
> But I find language reference just describing the formal grammer,
> probably meant for programmers who already know python well (rather
> too well). Just as an example, by readking lang. reference I still
> don't know what "..." is and couldn't find refences to __all__,
> __slots__. Is there a "python referece" available somewhere, which
> covers every aspects of python, possibly with explanation+examples
> (something akin to what K&R is for C) ?

>Thanks
>P. Adhia

 ... is an Ellipsis; I gather that there is such a thing as an
 Ellipsis object which I guess is some way of representing 
 recursive references.  For example:

 	a = []; a.append(a)

 ... If you do this from an interactive interpreter session
 and then just type: 

 >>> a

 ... you should get:

 [[...]]

 (a list containing a list containing an Ellipsis object)?

 I won't pretend to understand this; and maybe I'm completely 
 mistaken.  However, this is the impression I've gotten.  As you 
 say, the Tutorials (and other books and docs I've see) don't 
 explain it.

 __all__ is a member of a package which lists "all" of the 
 classes that are intended to be part of the package (exposed via
 the dir() built-in).  It's mostly used in the sorts of modules
 that are implemented as a directory tree, so it would be set 
 in an __init__.py file.  
 
 __slots__ is a special member for 2.2 "New" classes (as opposed
 to "classic classes").  It provides support for constraining the 
 list of attributes for an object.  Basically it means that efforts
 to add new attributes in an object will raise an AttributeError
 exception.  (Unless they are in __slots__ or if the __setattr__
 method of an object handles it for us?).

 """
 Finally, it's possible to constrain the list of attributes that can be
 referenced on an object using the new __slots__ class
 attribute. Python objects are usually very dynamic; at any time it's
 possible to define a new attribute on an instance by just
 doing obj.new_attr=1. This is flexible and convenient, but this
 flexibility can also lead to bugs, as when you meant to write
 obj.template = 'a' but made a typo and wrote obj.templtae by accident. 

 A new-style class can define a class attribute named __slots__ to
 constrain the list of legal attribute names.
 """

 You can read much more about these and other 2.2 features at:

 	http://www.amk.ca/python/2.2/




From aahz at pythoncraft.com  Wed Apr  3 18:54:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 18:54:01 -0500
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net>
Message-ID: <a8g4mp$k71$1@panix1.panix.com>

In article <a8dt35$a2i$1 at slb2.atl.mindspring.net>,
Jim Jacobs <jim.jacobs at jacobshome.com> wrote:
>
>Is it possible to simulate "C" style static variables in functions?  In
>methods?

I see that nobody has mentioned the most Pythonic way of all: create a
module called static.py.  (That's assuming you want a global static
variable, of course.)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From jkraska at san.rr.com  Sat Apr 13 00:03:18 2002
From: jkraska at san.rr.com (Courageous)
Date: Sat, 13 Apr 2002 04:03:18 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at>
Message-ID: <vrbfbukcme1ltp2tr1u30bkk28a3vdqkvl@4ax.com>

>The last thing which I am  bothered with: whether to become a good
>programmer or not. For example, I did never use a struct in C; that
>means my C code is more or less a big spaghetti one. I do not care as
>long as I get the correct result.

Fair enough. Truth be told, a great deal of the rather impassioned
arguments that programmers have involve implicit disagreement over
what _constitutes_ "the correct result". Consider:

For you, you write some code, execute it, couldn't give a hairbrained
rat's ass whether or not you yourself could read it next year.

For others, they care whether or not they can. They'd like to be able
to read their own code in a month, two, a year, and still be able to
figure out what they did.

For other still, they care whether or not their colleagues can.

For even others, they care whether some dummies in a microsoft shop
can figure it out. :)

These are a linear scale taken from just _one_ dimension of one
measure of "correctness".

There are other dimensions.

And I'll leave it at that.

C//



From erno-news at erno.iki.fi  Sat Apr 20 10:30:37 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 20 Apr 2002 17:30:37 +0300
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org>
Message-ID: <kulmbimohe.fsf@lasipalatsi.fi>

In article <mailman.1019279977.28194.python-list at python.org>, Tim
Peters <tim.one at comcast.net> writes:

| Hmm.  Shows up fine in the newsreader I use -- this, like hearing about the
| speedups for the first time from that msg, may be another symptom of General
| Upgrade Fear <0.9 wink>.

it doesn't show up in many newsreaders, because your message lacks a
"references" header. that's the usual mechanism that newsreaders use
to do threading. your message does have a "in-reply-to" header, which
your newreader could be using. it's for mail and not news, so not
standard.

maybe mailman could be hacked to fake a "references" header based
on "in-reply-to".

  -- erno


From ak at silmarill.org  Tue Apr  2 15:21:50 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 02 Apr 2002 20:21:50 GMT
Subject: ctrl-c and os.system problem
References: <mailman.1017777544.11640.python-list@python.org>
Message-ID: <slrnaak4q3.u14.ak@ak.silmarill.org>

In article <mailman.1017777544.11640.python-list at python.org>, Noom Knews wrote:
> I have a simple program like this
> 
> os.system("ls -lR / > /tmp/ls.out")
> print "end"
> 
> When I press ctrl-c, the "end" message is still print.
>  I'd like to terminate the whole process when pressing
> ctrl-c instead of just "end".
> 
> I tried to catch the signal already but can't catch
> the SIGINT or anything else at all during the
> os.system call.   This does not work either
> 
> try:
>    os.system("ls -lR / > /tmp/ls.out")
>    print "end"
> except:
>    print "get all exception"
> 
> This will keep printing "end" when ctrl-c is pressed
> too. I think this could be os.system problem. If I use
> other statement instead of os.system, the program
> works as I expected.
>
So os.exec works fine? Use that then.

> 
> This was tried on Redhat Linux with python 1.5.2
> 
> Thanks,
> Nick
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Tax Center - online filing with TurboTax
> http://http://taxes.yahoo.com/
> 


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From phd at phd.pp.ru  Mon Apr  1 21:14:12 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 2 Apr 2002 06:14:12 +0400
Subject: PEP 262 (database of installed Python packages)
In-Reply-To: <3CA902AF.EF9C710E@bgb-consulting.com>; from garrett@bgb-consulting.com on Tue, Apr 02, 2002 at 12:59:43AM +0000
References: <3wgg8.685$Qo2.70120@news02.tsnz.net> <3CA902AF.EF9C710E@bgb-consulting.com>
Message-ID: <20020402061412.A22529@phd.pp.ru>

On Tue, Apr 02, 2002 at 12:59:43AM +0000, Don Garrett wrote:
>   If this is an approach worth taking, then broaden it further. Take advantage
> of the native install DBs that already exist. RPMs on Linux

   Unfortunately, this is not as simple as it sounds - RPMs are not for
Linux, they are only for RedHat and derivtives. Debian use DEBs, and other
Linux distributions use other different incompatible formats.

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



From mwh at python.net  Tue Apr 23 07:54:13 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 23 Apr 2002 11:54:13 GMT
Subject: curses mouse-events in linux console
References: <3CC3E067.6080107@pp.htv.fi> <lk662jrjav.fsf@pc150.maths.bris.ac.uk> <3CC54204.5010100@pp.htv.fi>
Message-ID: <lkofgapqyd.fsf@pc150.maths.bris.ac.uk>

janne halttunen <jhalttun at pp.htv.fi> writes:

> Michael Hudson wrote:
> > janne halttunen <jhalttun at pp.htv.fi> writes:
> >
> >>I'm on Mandrake Linux 8.2 with Python 2.2, and I have a problem of
> >>mouse-events not emitting properly in curses when on console with
> >>gpm. In X it works fine, the getch-method returning instantly on
> >>mouse-clicks, but on console it doesn't return until after pressing
> >>key on the keyboard.
> > Umm, does any curses application respond to the mouse on the console?
> > I thought it was an xterm-and-derivatives only thing, but I could well
> > be wrong.
> 
> Well, it was working in python 2.0 or 2.1 with previous Mandrake
> Linuxes 8.1 or 8.0..  And it's working now, except for the triggering
> of the event.

Oh.  Can't really help you I'm afraid...

Cheers,
M.

-- 
     ARTHUR:  Why are there three of you?
  LINTILLAS:  Why is there only one of you?
     ARTHUR:  Er... Could I have notice of that question?
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11


From tim.one at comcast.net  Sat Apr 20 03:00:05 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 20 Apr 2002 03:00:05 -0400
Subject: How to make regexes faster? (Python v. OmniMark)
In-Reply-To: <eppstein-2C56E0.23265219042002@news.service.uci.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEKPOPAA.tim.one@comcast.net>

[David Eppstein]
> The point is that your responses don't have the appropriate References:
> header, so they don't get threaded with other messages on the same
> topic.

I know.  The news reader I use (when I use one) uses criteria in addition to
"References:" for threading, so I never notice it myself.

> It appears as if you are using a mail agent rather than a news agent,
> since References: is a news header (and your other headers say you're
> using MS Outlook).

I do use the news<->mail gateway most of the time.

> It affects other things than threading, too -- e.g. I use References:
> to pick out responses to my own messages, and yours don't get caught.

Then you'll be spared this <wink>.





From ark at research.att.com  Wed Apr  3 10:59:13 2002
From: ark at research.att.com (Andrew Koenig)
Date: 03 Apr 2002 10:59:13 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <016701c1db1a$569543e0$0900a8c0@spiff>
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
Message-ID: <yu99wuvovkmm.fsf@europa.research.att.com>

>> So what's different?

Fredrik> more than one way to do it.  more than one concept to
Fredrik> explain to newcomers.  more than one "obviously correct"
Fredrik> way to do things.  less is more.  etc.

I'm a little confused, because it seems to me that Python already
has more than one way to do this particular thing.

Specifically: If I want to write a function that answers a yes/no
question, I have lots of possible ways of spelling yes (1, 2, "yes",
and so on) and lots of possible ways of spelling no (0, {}, None,
and so on).  There isn't a single preferred way.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark




From aleax at aleax.it  Mon Apr 15 14:39:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 18:39:08 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de> <Prau8.17347$vF6.558738@news2.tin.it> <slrnabm63e.nkd.huaiyu@gauss.almadan.ibm.com>
Message-ID: <g7Fu8.22687$vF6.744151@news2.tin.it>

Huaiyu Zhu wrote:
        ...
>>A simple histogram of non-alphameric tokens in Lib/*.py only partially
>>conforts this thesis:
> [snip statistics show small usage of & ~ | ^]

Well, & showed up better than I thought it would, actually.

> Do you plan to revive the bitwise module you wrote sometime ago?

Guess it should be PEPped.  But I don't think the powers that be
will stand for any eventual deprecation of the bitwise ops, so it
looks like a futile exercise.


Alex



From ajs at ix.netcom.com  Wed Apr 10 04:08:08 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Wed, 10 Apr 2002 04:08:08 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000101c1e067$76d35820$0334fea9@carol>

Time writes -

>It could have credited Paul Rubin for the "water
>torture" analogy

While we're on the subject of Aahz...

It was and is a piece of a advice that  - in the 
interests of better group dynamics - a message be
sent that Aahz never has and never will speak for
the core development team. 

A peek in at dev indicates things moving in the 
direct opposite direction.

Some-people-don't-know-who-their-friends-arely-yours,

Art 













From aleax at aleax.it  Sat Apr 13 18:39:28 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 22:39:28 GMT
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <As2u8.17016$vF6.537198@news2.tin.it>

Jens Baader wrote:

> I'm downloading Python 2.2.1 right now and I wonder
> why you still produce bugfix releases for the old 2.1 development line.

I think the main motivation is that the current Zope (since it relies on
the ExtensionClass extension) relies on Python 2.1.

> It seems that Python 2.2 is somewhat broken. If not what's the
> reason that keeps the people from upgrading to 2.2?

I don't see anything "broken" about 2.2.  The changes to the object
model invalidated extensions which relied on very deep characteristics
of Python 2.1 and earlier, such as ExtensionClass.  The next major
release of Zope will presumably be based on 2.2 (or later).

A more general issue, however, is that Python issues two releases
a year, and that's just too often for many commercial developers to
stop their own development and re-validate their code base.  So, such
a developer will typically choose a certain specific line (such as 2.2.*)
and stick with it for longer than six months -- anything from 12 to 24
months is more typical, and longer cycles not unheard of.  Bugfix releases
(such as 2.2.1) help such developers enormously.


> Another question: When will we see an official ISO/ANSI standard
> for Python? I dislike working with nonstandardized languages.

Funny -- for years I've worked mostly with such a standardized language 
(and I'm still a Brainbench MVP for C++), and never did I find two 
implementations that matched exactly -- I think that even today no 
commercial C++ claims 100% compliance with the standard.  Given that track 
record, I doubt anybody will ever find the funds to finance the enormous 
collective investment (mostly of people's time) required for the 
standardization process.


Alex



From peter at engcorp.com  Tue Apr 16 02:23:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 16 Apr 2002 02:23:07 -0400
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>
Message-ID: <3CBBC34B.2A325141@engcorp.com>

sameer wrote:
> 
> I am trying to run multiple threads in a wxFrame class to instantiate
> different panels within it on Win 2k,  and well, it's a little hairy.
> Anyone got a good web source that discusses python threading in depth?

Sorry, not handy.  Maybe others can help.

> More so, my question is, how can I get the return value of a function
> that I run on a seperate thread?  I don't see anything for this in the
> threading module.

Threads don't "return" in the sense of going back to the routine
which called them.  They are spun off at the time they are started,
largely independent of their creator.

You might look into the join() method of the threading.Thread class,
which you could use in a calling routine to wait for termination
of a thread (but if you do that in the wxPython mainloop, you'll
freeze the GUI).

For actually passing a value, you can just store something as
an attribute of the Thread object, and retrieve it after the
thread has terminated, if you keep a reference to the thread
in the creator.

Threads are not trivial, especially when mixed with GUI stuff.
Search the web for examples.  Try google with something like
http://www.google.com/search?q=wxpython+threading+thread+python+examples

-Peter


From kragen at pobox.com  Mon Apr 29 17:57:01 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 29 Apr 2002 17:57:01 -0400
Subject: Lambda Scoping
References: <3ccd5f99.5559046@news.t-online.de> <mailman.1020093351.17995.python-list@python.org>
Message-ID: <83d6wirwwi.fsf@panacea.canonical.org>

Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
>   Lambdas in python don't work exactly like lambdas that you would
> expect in lisp. The scoping is a little different.

True, at least up to 2.1, but completely irrelevant.  The same code in
Lisp would do the same thing.



From aleax at aleax.it  Mon Apr 29 04:31:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 08:31:44 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <aaitih$o2k$1@bob.news.rcn.net>
Message-ID: <Qx7z8.75202$vF6.2249222@news2.tin.it>

Raymond Hettinger wrote:
        ...
> Like 'open' in the os module, 'pow' is a high risk candidate from for
> being clobbered by a 'from math import *'.

It's another good example of why "from ... import *", like other
statements that play havoc with your namespace in ways you can't
fully control (e.g., exec in local/global scope), is evil.

 
>> But I double the suitable module is math, as this would
         ^^^^^^
Freudian slip, I swear!  Thinking of "math" tells my subconscious
"it's about double" -- and "doubt" becomes "double" (as for any
bridge player it might...).


Alex



From aleax at aleax.it  Mon Apr  1 01:44:51 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 06:44:51 GMT
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <mailman.1017504362.25937.python-list@python.org> <slrnaacavs.2gs.philh@comuno.freeserve.co.uk> <a85od6$9i$1@panix1.panix.com>
Message-ID: <DlTp8.48032$S52.1614804@news2.tin.it>

Aahz wrote:

> In article <slrnaacavs.2gs.philh at comuno.freeserve.co.uk>,
> phil hunt <philh at comuno.freeserve.co.uk> wrote:
>>On 30 Mar 2002 11:04:13 -0500, Andrew Koenig <ark at research.att.com> wrote:
>>>
>>>I wonder if this is the circle-ellipse problem over again?
>>
>>What is the circle-ellipse problem?
> 
> I'm not sure, but I'm going to guess for my own edification:
> 
> Instead of circle/ellipse, it's slightly easier to talk about
> square/rectangle.  In terms of standard geometry, a square is a subset
> of rectangles, such that height and width are identical.  But in OO

The set of squares is a subset of the set of rectangles, but "a square is
a subset or rectangles" is not correct (one _has_ to be nitpickish around
such delicate issues...).

> terms, it's a bit more complicated to inherit square from rectangle,
> because square has only side-length attribute where rectangle has two
> (height and width).  So to create a class square that inherits from
> rectangle, you either have to write special code that ties height and
> width together or you have to delete at least one of the attributes (in
> which case any code expecting to work with rectangle fails).

Actually, this specific problem is there, like several others, only
*for MUTABLE objects*.  Take an "immutable" rectangle:

class ImmutableRectangle(object):
    def __init__(self, width, heigth):
        object.__setattr__(self, 'width', width)
        object.__setattr__(self, 'heighth', heigth)
    def __setattr__(self, name, value):
        raise TypeError, "cannot mutate ImmutableRectangle"
    def __hash__(self):
        return hash(self.width) ^ hash(self.heigth)
    def area(self):
        return self.width*self.heigth
    def perimeter(self):
        return 2*(self.width+self.heigth)
    # etc, etc

Now, subclassing this is perfectly fine:

class ImmutableSquare(ImmutableRectangle):
    def __init__(self, side):
        ImmutableRectangle.__init__(self, side, side)

Net of tricky ways to go behind the interface and mutate the
"immutable" instances (which are neither here nor there... they
go below and behind the OO'ness, e.g. by working directly
with __dict__ etc), an instance of the (immutable) square is
fine wherever one of the similarly immutable rectangle would
be.  It's only if you're interested in mutable object state (rather
alien to the mathematical basis of geometry -- geometrical
entities are not conceived of as having separate 'identity' and
'state', so one could not even speak of the 'same' entity that
'changes'...) that things break down.

Another similar issue comes with containers -- say Set for
example.  In statically typed languages with generics, say I have
a Set<something> parameterized class ('template' in C++, e.g.).
I have a class Banana inheriting from Fruit, a function f that needs
an argument s of class Set<Fruit>, and an object b that is a
Set<Banana>.  Can I call f(b) type-safely?  Answer: if Set<...>
is immutable, yes.  But if it's mutable, no: f might insert an Apple
instance in its s parameter (type-safely, since s is a Set<Fruit>
and Apple also subclasses Fruit) and break the constraints on b's
type (all of b's items must be instances of Banana).

So, it's mutability that complicates things in these cases.  The
exponents of Functional Programming have built rather powerful
languages around immutable objects -- immutability may not be
as easy and comfortable to work with in practice, and it does not
model our computer systems all that well, but it sure does remove
certain mathematical anomalies!

Back to bool and int -- we ARE talking about immutables here,
since numbers are immutable in Python.  Therefore (net of the
str and repr issues, where IS-A breaks down either way... and
that's another problem) the "circle/ellipse anomaly" does not surface 
in any form.  Subclassing bool from int IS ok, given immutability.
    

> This kind of thing is what makes the Deitel book on Python really bad,
> because it doesn't give a clean picture of proper OOP (they make circle
> a subclass of point, and cylinder a subclass of circle, which is even a
> worse botch than trying to deal with circle/ellipse).

I agree (and I told them that, as a reviewer, but was unable to convince
them).  Inheritance does not model IS-A at all well in such cases (even
though isinstance and exception handling appear to want to believe that
Python subclassing implements IS-A), but beginners will still be confused
unless you try to give parallels between the two concepts -- and if you
try, as they do, then such examples are counterproductive.


> In general, I think the evidence has shown that subclassing is an
> exercise fraught with danger and should be de-emphasized as a primary
> technique in OOP.  I'm very glad that Python uses implicit protocols
> through magic methods for much of its OO processing; that makes it much
> easier to cut through the Gordian Knot.

Yes -- net of isinstance and exception handling.  PEP 246 would make
it that much easier again.


Alex



From loewis at informatik.hu-berlin.de  Sun Apr 14 13:03:58 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 14 Apr 2002 19:03:58 +0200
Subject: Lambda  added to release 2.2 ?
References: <3CB9ABAB.E37E685B@club-internet.fr>
Message-ID: <j4r8linrep.fsf@informatik.hu-berlin.de>

Fabien H?non <ffjhenon at club-internet.fr> writes:

> When I use the lines below :
> >    insertmenumaterial.add_command(label="interior", command = lambda :
> self.insert_example("interior.txt"))

I assume your lines look more like

  def foo(self):
    insertmenumaterial.add_command(label="interior", 
         command = lambda : self.insert_example("interior.txt"))

i.e. the insertmenumaterial is inside a method.

Notice that the lambda function itself does not take any arguments.

> >>> NameError: global name 'self' is not defined

That is the traditional way of Python: since the lambda is a function,
it has only access to its local and global variables. self is neither
local nor global, hence 'self' is not a name of any variable.

> I don't have any problem under W2000 ( Python 2.2)
> Is lambda function a new add-on to 2.2 ? I thought it was added ages
> ago.

In Python 2.2, nested function scopes where added, so that the nested
lambda function can access the variables of the outer function. This
is not directly available in Python 2.1. Add

from __future__ import nested_scopes

to the top of the file to make it work correctly with Python 2.1, and
fail immediately with Python 2.0 and earlier.

Regards,
Martin


From baas at ira.uka.de  Sat Apr  6 03:28:28 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Sat, 06 Apr 2002 10:28:28 +0200
Subject: ANN: Python Computer Graphics Kit v1.0beta1
Message-ID: <6actau4pssld5v0i1l8rs6isgpsib3af6i@4ax.com>

I've posted this earlier but it didn't seem to get through, so I post
again. My apologies if you got this twice.

I have just released the first beta version of the Python Computer
Graphics Kit, a collection of Python modules that contain the basic
types and functions required for creating 3D computer graphics images.

http://cgkit.sourceforge.net

The kit includes several new types such as vectors, matrices and
quaternions. It contains a binding for Pixar's RenderMan interface
which is a renderer independent API to communicate 3D data to
renderers which will finally produce a 2D image. There are already
several RenderMan compliant renderers freely available (they are
not part of the kit). The kit also includes some of the functionality
from the RenderMan Shading Language which enables you to create
procedural models or textures from within Python.

Even though the kit focuses on RenderMan, the new types or the
Shading Language functionality can also be used for other rendering
mechanism such as OpenGL or other renderers such as POV-Ray.

The kit should run on any platform where Python (and a C/C++ compiler)
is available. Windows users can download a binary version for Python
2.1or Python2.2.

For more information, visit:

http://cgkit.sourceforge.net


- Matthias -



From pilkevitch at gmx.net  Tue Apr 23 09:54:41 2002
From: pilkevitch at gmx.net (Konstantin Pilkevich)
Date: 23 Apr 2002 06:54:41 -0700
Subject: how to register COM interface from python COM-server
Message-ID: <2ff416fa.0204230554.40048f15@posting.google.com>

hi all

i got a problem using Python COM server. 
I have working python COM-server Dictionary Singleton (updated
dictionary, you can find 3  .py-files attached). I wrote idl file for
it. It was compliled by midl withour errors. I've got type lib file
and registered it. After that i created C++ class from type lib file
from VC++ Class Wizard. And it works normally with server.
 
The problem is that i can't use this solution, because in this case i
must use MFC, and i should not.
 
I tried to use this server in such way:
 
 CoInitialize(NULL);
 IMainDictionary *pDict;
 IDispatch *pDisp;
 _GUID guid;
 CLSIDFromProgID(L"Python.DictionarySng",&guid);
 long res = 0;
 IClassFactory *unk = NULL;
 
 res = CoCreateInstance(guid,NULL,CLSCTX_ALL,IID_IDispatch,(void**)&pDisp);
 if(res==S_OK && pDisp != NULL)
 {
  pDisp->AddRef();
  res = pDisp->QueryInterface(IID_IMainDictionary,(void**)&pDict);
 }
 else if(res==E_NOINTERFACE)
 {
  return 0;
 }

instance of object was created normally but QueryInterface failed.
i added such code to server:
 
IID_IMainDictionary =
pythoncom.MakeIID("{B5B94B7D-A76F-47b4-B755-EDCA4F5A5925}")

and overload query_interface function:
  def _query_interface_(self, iid):
      if iid==IID_IMainDictionary:
        return win32com.server.util.wrap(self)

but needed interface is still not supported.
can you help me to find the way to register my interface.


From jozef at jozef.joz  Fri Apr 12 19:58:00 2002
From: jozef at jozef.joz (Jozef)
Date: Sat, 13 Apr 2002 01:58:00 +0200
Subject: fp/lambda question
Message-ID: <a97s9h$gfg$1@reader07.wxs.nl>

I was reading on functional programming in Python and wanted to test
myself by writing a function:

functional_euler = lambda n: \
	len(filter(lambda i: gcd(i, n) == 1, range(1, n + 1)))

(gcd is the greatest common divisor function). It gives the following
error: <stdin>:1: SyntaxWarning: local name 'n' in 'lambda' shadows use of
'n' as global in nested scope 'lambda'

I am not really sure even what this means... does lambda create a new
scope or something? Is there another way to write this?

Below is the gcd function I used and two other versions of the euler
function that do work as expected.

def gcd(a, b): # Greatest Common Divisor
	if b == 0: return a
	else: return gcd(b, a % b)

def imperative_euler(n):
	xs = []
	for i in range(1, n + 1):
		if gcd(i, n) == 1: xs.append(i)
	return len(xs)

def list_euler(n):
	return len([x for x in range(1, n + 1) if gcd(x, n) == 1])


From ianb at colorstudy.com  Tue Apr  9 15:53:59 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 09 Apr 2002 14:53:59 -0500
Subject: pointer
In-Reply-To: <a8uukg$fuu$04$1@news.t-online.com>
References: <a8uukg$fuu$04$1@news.t-online.com>
Message-ID: <1018382040.2825.35.camel@localhost>

On Tue, 2002-04-09 at 09:45, Henning Peters wrote:
> i would like to store values in two different locations. both values should
> always be equal. the one could be a pointer on the other. is this possible
> to do this in python?

You can simulate explicit pointers with lists (nearly everything in
Python has implicit pointers, which is what other people are pointing
out).  Or, maybe more clearly, you can follow the pointer pattern in
Python without using anything called a pointer.  For instance:

int dosomething(int* x) {
    *x = 10; // <-- changing the slot x, not just the binding of x 
             // for this function
}

def dosomething(x):
    x = 10

The Python version will actually do nothing, since if you do:

a = 5
dosomething(a)
# a is still 5

But you could do:

def dosomething(alist)
    alist[0] = 10

a = [5]
dosomething(a)
# a is now [10]

So [0] is like the * operator, and [something] is like the & operator. 
In other words, a list creates a second pointer.  Objects slots are
similar.

  Ian






From joe at fenixasia.com  Wed Apr  3 00:07:34 2002
From: joe at fenixasia.com (joe)
Date: 2 Apr 2002 21:07:34 -0800
Subject: chdir() problem
Message-ID: <b5b450db.0204022107.37bc0a25@posting.google.com>

within python 2.2, chdir() works fine.

however, i want to be able to change the "active" directory, so that
when the script completes, the user
is left in a target directory (which is not the original directory)

reading docs and searching has provided no insight.

any suggestions?

thx, joe


From marklists at mceahern.com  Thu Apr 11 10:13:09 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 11 Apr 2002 09:13:09 -0500
Subject: How to access attributes of nodes in xml.minidom
In-Reply-To: <newscache$78oeug$5lo$1@www-neu.dzsh.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEIHKKAA.marklists@mceahern.com>

[Ruediger Maehl]
> But, how can I access the attributes of a node?

I've modified your example slightly.  I hope it gives you some ideas.  You
can also use getAttribute(name).  I know what you mean about the docs.  But
the xml-sig is a good place to ask these questions.

Cheers,

// mark

#! /usr/bin/env python
import xml.dom.minidom

document = """\
<slideshow>
  <title attrib="new">Demo slideshow</title>
  <slide>
    <title>Slide title</title>
    <point>This is a demo</point>
    <point>Of a program for processing slides</point>
  </slide>
</slideshow>
"""

def main(dom):
    for node in dom.documentElement.childNodes:
        if node.nodeType != 3:
            print node.tagName
        else:
            print node.toxml()
        if node.attributes:
            for i in range(node.attributes.length):
                a = node.attributes.item(i)
                print "%s = %s" % (a.name, a.value)

dom = xml.dom.minidom.parseString(document)

main(dom)

dom.unlink()  # minidom only?
#################





From lumber at jack.com  Sat Apr 20 15:12:58 2002
From: lumber at jack.com (Lumberjack)
Date: 20 Apr 2002 19:12:58 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
Message-ID: <Xns91F67C40EB0E0lumberjack@216.148.53.83>

aahz at pythoncraft.com (Aahz) wrote:
> What's amazing about Lisp is that it's still in many ways on the cutting
> edge of computer science.

Perhaps you should name these "many ways". Taken at face value, your 
statement pretty much claims that computer science hasn't progressed in 40 
some years. I can at least agree with that statement. Of course the 
"science" in "computer science" is pure bovine excrement. If you can't or 
wont measure it, it isn't science. "Computer religion" or "computer cults" 
would better capture the reality.


From tismer at tismer.com  Mon Apr 29 06:30:03 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 29 Apr 2002 12:30:03 +0200
Subject: Stackless goes Limbo
References: <mailman.1019911009.16159.python-list@python.org> <aaee44$p73$1@panix1.panix.com> <1019928879.371845@yasure>
Message-ID: <3CCD20AB.3040701@tismer.com>

Donn Cave wrote:
[about C, sys-> etc]

Well, this is all not relevant for me.
I'm not interested in the language, only
in the multitasking aspects.

> The important thing to me is that the system looks like it would work
> well for modular function libraries, thanks mostly to the garbage
> collection storage model, and who knows, maybe with less function call
> overhead.  The thing I miss here is exceptions - nothing like that?
> Maybe I skimmed it too fast.

I didn't find them, either. And in fact, they give me
some headaches for Stackless.

> I think I like the channels.  Not sure I like the Unicode strings.
> I wouldn't mind trying it out, if it ever escapes from Inferno.

It is the channels idea what I want to borrow. Nothing else.

thanks & cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From jimd at vega.starshine.org  Tue Apr  2 22:23:15 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 3 Apr 2002 03:23:15 GMT
Subject: Bastion Documentation Error?
Message-ID: <a8dsj3$1n4j$1@news.idiom.com>

 I'm a bit confused by the Bastion.Bastion() function.
 From the documentation I read:

 """
 Protect the object object, returning a bastion for the object. Any
 attempt to access one of the object's attributes will have to be
 approved by the filter function; if the access is denied an 
 AttributeError exception will be raised. 
 """

 ... but playing with it I find that it raises an AttributeError on 
 any attempt to access any class or instance variables.  It always 
 seems to call the filter function, then (regardless of the results) 
 it seems to then check to see if the attribute is a method and block 
 access even if the filter function returned a true value.

 If that is the intent (that only *methods* can be exposed through
 regardless of the filter function) then I think this should be made 
 more clear in the documentation.

 Here's a simple transcript that should demonstrate the issue:

Python 2.2.1c1 (#1, Mar 15 2002, 08:13:47) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import Bastion
>>> flt_args=[]
>>> def flt(attr):
...    flt_args.append(attr)
...    if attr[0] == "_": return None
...    else: return 1
... 
>>> class Kls:
...    foo = 1
...    _foo = 2
...    def __init__(self):
...       self.bar = 3
...       self._bar = 4
...    def getfoo(): return seself): return self.foo
...    def getbar(self): return self.bar
...    def _getfoo(self): return self._foo
...    def _getbar(self): return self._bar
... 
>>> a=Kls()
>>> b=Bastion.Bastion(a,flt)
>>> b.foo
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: foo
>>> flt_args
['foo']
>>> b.getfoo()
1
>>> b.getbar()
3
>>> b._getfoo()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: _getfoo
>>> b._getbar()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: _getbar
>>> flt_args 
['foo', 'getfoo', 'getbar', '_getfoo', '_getbar']
>>> b.bar
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: bar
>>> b._foo
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: _foo
>>> b._bar
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/Bastion.py", line 79, in __getattr__
    attribute = self._get_(name)
  File "/usr/lib/python2.2/Bastion.py", line 122, in get2
    return get1(name)
  File "/usr/lib/python2.2/Bastion.py", line 118, in get1
    raise AttributeError, name
AttributeError: _bar
>>> b.getbar()
3
>>> b.getbar
<bound method Kls.getbar of <__main__.Kls instance at 0x80a4a84>>
>>> 



From mertz at gnosis.cx  Thu Apr 11 23:46:47 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Thu, 11 Apr 2002 23:46:47 -0400
Subject: New classes, __slots__, __dict__
Message-ID: <nilt8kKkX8EC092yn@gnosis.cx>                

Hi Pythonistas,

This is especially to the actual Python developers, but other users
might have insight.  For background, I am co-author of the module
xml_pickle; for those unfamiliar, xml_pickle (more-or-less) does a
superset of what pickle/cPickle do, but using XML files as the data
format.  This note isn't to plug the module, but you can find info at:

    http://gnosis.cx/download/Gnosis_XML_Util.ANNOUNCE

In xml_pickle, we are trying to decide how to handle Python 2.2+ new
style classes that use the '__slots__' attribute.  Unfortunately,
pickle.py itself doesn't really do anything very sensible for
guidance... which is the nature of my question.

Let me explain for those (vast majority) of readers who have not
encountered this.  The Python 2.2(+.1c1) release notes indicate:

  - If you try to pickle an instance of a class that has __slots__ but
    doesn't define or override __getstate__, a TypeError is now raised.
    This is done by adding a bozo __getstate__ to the class that always
    raises TypeError.  (Before, this would appear to be pickled, but the
    state of the slots would be lost.)

I still have plain 2.2 on my system, so what I get is:

    >>> class New(object):
    ...     __slots__ = ('this','that')
    ...
    >>> new = New()
    >>> new.this = 'this'
    >>> from pickle import dumps, loads
    >>> new2 = loads(dumps(new))
    >>> new2.this
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    AttributeError: this

But it's easy enough to envision a TypeError being raised here instead.
In any case, neither behavior is particularly helpful.

I'd like pickle to do something nicer, but MY real concern is making
xml_pickle do something nice.  Within xml_pickle, we look at
"o.__dict__.keys()" to find the attributes for an object we are
pickling.  In an old version, xml_pickle looked at "dir(o)", but that
was fragile, and broke with Python 2.2.  The o.__dict__ approach doesn't
work with new style classes that have __slots__:

    >>> new.__dict__
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    AttributeError: 'New' object has no attribute '__dict__'

My feeling on this is that we (the xml_pickle authors) should provide a
"smart" attibute detector that looks something like:

    def dir_(o):
        if hasattr(o,'__dict__'):
            return o.__dict__.keys()
        if not hasattr(o,'__slots__'):
            raise TypeError, "Strange class with neither dict nor slots"
        attrs = []
        for attr in o.__slots__:
            if hasattr(o, attr):
                attrs.append(attr)
        return attrs

At first brush, this function returns helpful results in all cases:

    >>> class Old: pass
    ...
    >>> class Noslots(object): pass
    ...
    >>> old = Old()
    >>> old.this = 'this'
    >>> noslots = Noslots()
    >>> noslots.this = 'this'
    >>> dir_(new), dir_(old), dir_(noslots)
    (['this'], ['this'], ['this'])

However, my xml_pickle co-author, Frank McIngvale, expressed the
following reservation about substituting my new dir_() function:

| Also, my admittedly simpleminded thought is that if handling slots
| is a 2-liner, why doesn't pickle.py just do that? I have to believe that
| the pickle authors are smart people and there is some deep technical
| reason why getstate MUST exist if slots are used, and not just that the
| pickle authors were too lazy and punted.

So what do you think... smart people who wrote pickle?  What's wrong
with the (slightly-more-than-two) line solution?  What is the deep
technical mojo that argues against this approach?

Yours, David...


--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From marcominutoli at supereva.it  Fri Apr  5 04:17:58 2002
From: marcominutoli at supereva.it (Marco)
Date: 5 Apr 2002 01:17:58 -0800
Subject: ZPT : return a string HTML from a script python
Message-ID: <73f739e9.0204050117.36bde90b@posting.google.com>

I have a ZPT and I want to print an HTML string (it's arrive from a
script python) inside it.


This is a piece of ZPT code that call the script "pyPrint"

--   <span tal:content="python:here.pyPrint
--        (folder=request['folder3'])">
--   </span>      

This is a script

--  message = "Result: <br>"
--  for folder in folder:
--      folder=container.pyString2list(string=folder.id)
--      rootFolder = container.restrictedTraverse(['',]) # container 
--      folderDaCancellare = rootFolder
--      folderPadre = None
--      for item in folder[1:]:
--          folderPadre = folderDaCancellare
--          folderDaCancellare = folderDaCancellare[item]
--      message = messaggio+folderDaCancellare.id
--      if folderPadre != None and
len(folderDaCancellare.objectValues()) == 0:
--          folderPadre.manage_delObjects(folderDaCancellare.id)
--          message = message + " cancellato<br>"
--      else:
--          messaggio = message + " non cancellato<br>"
--  
--  return message


I want that the <br> tag will become a new line in ZPT page.

Sorry for my english


From geoff at gerrietts.net  Mon Apr 15 16:19:44 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 15 Apr 2002 13:19:44 -0700
Subject: problem negative lookahead assertion
In-Reply-To: <4b926dbb.0204151202.6d709633@posting.google.com>
References: <4b926dbb.0204151202.6d709633@posting.google.com>
Message-ID: <20020415201944.GA12476@isis.gerrietts.net>

Quoting sjoerd siebinga (ssiebinga at fa.knaw.nl):
> 
> When this was applied to the text below all the \emph{} phrases are
> replaced instead of only those not followed by the \index{} commando.

Okay. I put together the script I've appended below my .signature and
ran it under python 2.2, python 1.5.2, and python 2.1. The script was
basically a lot of cut and paste from your posting.

When I run it, it brings up a prompt (which is what it's supposed to
do):

>>> data == mystr
1

In other words, no change at all has been made.

In all three interpreters, that's what I get. I assume that's a
problem, but a different problem than what you're actually complaining
about.

Others may be more familiar with your problem domain or more willing
to guess at it, but I'm gonna request that you provide, in addition to
your pattern, an example of what you're trying to transform, an
example of what you want to get, and an example of what you actually
get.

If I were to supply some general advice, based on what I've seen in
your patterns, it would be to avoid using .*  in your patterns if at
all possible -- prefer a [^ ] containing any "stop characters". Even
when the .* is nongreedy, it will swallow as much as it needs to, to
make a match. That could be half your document if that's what it takes
to get to a comma, whitespace character, or non-alphanumeric character
that isn't followed by \index.

Luck,
--G.

-- 
Geoff Gerrietts             <geoff at gerrietts dot net>
  I AM YOUR KING! BOW BEFORE ME, PEASANT!   -- Dogbert

#!/usr/bin/python -i

import re

emph = re.compile(r'\s.\\emph\{(.*?)\}([\s,\W])(?!\\index)')


mystr = r"""
\begin{germdata} ON \emph{va\th a} \index{on~va\th a}`wade, rush, walk
through', OE \emph{wadan}, \index{oe~wadan} OHG, MHG \emph{watan}
\index{mhg~watan}`wade, stride', MLG \emph{w\=aden},
\index{mlg~w\=aden} MDu. \emph{waden}, \index{mdu~waden} \emph{waeyen}
`wade, go' \end{germdata}
"""


data = emph.sub('\emph{\\1}\\2 \index{unl~\\1}', mystr)





From dahicks-NOSPAM at voyager.net  Tue Apr 30 13:28:46 2002
From: dahicks-NOSPAM at voyager.net (Daniel Hicks)
Date: Tue, 30 Apr 2002 13:28:46 -0400
Subject: Python SMTP server
Message-ID: <3cced44a$0$3341$2c3e98f8@news.voyager.net>

To all,

I have been playing around with the SMTP server written in Python that came
with the distribution I downloaded.  Has anyone had any success with a
similar version, perhaps modified?  I would appreciate seeing your code

Thanks
Dan

Remove '-NOSPAM' to send me email







From paul.osman at sheridanc.on.ca  Fri Apr 26 14:01:54 2002
From: paul.osman at sheridanc.on.ca (Paul Osman)
Date: Fri, 26 Apr 2002 14:01:54 -0400
Subject: Sending mail to an SMTP server that requires POP authentication ?
References: <aabrbv$2vg$1@norfair.nerim.net> <20020426104354.6ed98204.logiplexsoftware@earthlink.net>
Message-ID: <3CC99612.7050305@sheridanc.on.ca>

Cliff Wells wrote:

>On Fri, 26 Apr 2002 17:30:40 +0200
>Gillou wrote:
>
>>I did not find how to do this with smtplib.
>>
>>Did someone succeed ? How ?
>>
>>Thanks in advance.
>>
>
>Hm.  I think you'll have to log in with POP first, and then for some predefined
>period of time you'll be granted SMTP access.
>
not true in my experience... the situation I was dealing with was a 
little different mind you, users were on IMAP and SMTP. The SMTP server 
explicitly requires authentication... it does not matter if you had 
logged into the IMAP server or not. two seperate entities in this case, 
and they do not communicate other than holding the same account information.

-paul







From emile at fenx.com  Tue Apr 30 10:15:14 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 30 Apr 2002 14:15:14 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net>
Message-ID: <SFxz8.93254$HH5.5498@rwcrnsc51.ops.asp.att.net>

Fredrik Lundh
>
>
http://www.google.com/search?q=cache:opG_4HjTnbIC:www.theonion.com/onion
3811/child-rearing_advice.html
>
>

Are you getting ready to add this link to the FAQ?

;-)

--

Emile van Sebille
emile at fenx.com

---------



From bosahvremove at netscape.net  Wed Apr 24 13:50:57 2002
From: bosahvremove at netscape.net (Bo)
Date: Wed, 24 Apr 2002 17:50:57 GMT
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <Xns91FA8BD634248bosahvnetscapenet@66.185.95.104>

I agree with the others about a SQL backend, I'd suggest something like 
Postgres. This would be much more robust and probably faster but that isnt 
the first reason I'd suggest it.

You have 4 files (80GB) now and some expectations of the combined data. 
What happens if they change those expectations? If you have established an 
SQL database you will be much more flexible and fast.

> I still haven't received above mentioned files so I can't test the
> time needed to (for example) read a 15GB "file1", filter by few
> variables, and write a resulting subset as a "sub_file1".  Things
> would afterwards get more complicated cause I will have to pullout
> ID's from "sub_file1", remove duplicate ID's create
> "no_dup_sub_file1", match those to ID's in remaining 3 main files and
> pullout data linked with those ID's.

SQL would let you remove duplicates from your results in one step or show 
where those duplicates might be significant (IE. Joe was late how many 
times?)

If this is a recurring requirement at work I would consider hiring a 
contractor to create an elegant solution with defined steps for you to 
operate it.







zaka07 at hotmail.com (Rad) wrote in 
news:ad381f5b.0204240841.52333a3e at posting.google.com:

> I am preparing myself to work on extracting data from 4 text files
> (fixed width format) which combined size is about 80GB.  Considering
> deadlines, costs, and my limited programming knowledge I thought using
> Python/Windows for the job would be the best option for me.  However,
> I am worried about the speed in which Python (me and my hardware) will
> be able to deal with these massive data sets but I am hoping that this
> is still a quicker route then learning C.
> 
> I have a few weeks to prepare myself before data arrives and my
> question is: am I going the right way about the project, is Python
> (with humanly written code) capable of doing this kind of staff
> relatively quickly?
> 
> Any help and suggestions would be greatly appreciated.
> 
> Thanks
> 
> P.S As you probably guessed I'm new to python



From ws-news at gmx.at  Thu Apr 11 10:58:19 2002
From: ws-news at gmx.at (Werner Schiendl)
Date: Thu, 11 Apr 2002 16:58:19 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <3cb5a47d$1@brateggebdc5.br-automation.co.at>

Hi Ryan,

since you are interested in both, C and C++ or C and Jave I would recommend
the freely downloadable books from Bruce Eckel (available from
http://www.bruceeckel.com/ ).

They are named 'Thinking in Java' and 'Thinking in C++' and are in my
opinion very readable. Maybe a little under-challenging in the first
chapters (since you already know how to program python) but not boring.

Both books contain a decent introduction into the world of C, that forms the
basement of Java and C++.

They do _not_ contain a full introduction into the C programming language,
though. Basically, concepts that have more convenient (or more safe or more
object-oriented) alternatives in C++ (or Java) are not discusses in detail.

In my opinion, moving from Python 'down' to Java or C++ and then 'down' to C
is probably easier for you since you are used to much higher level
constructs from Python than C provides.

Whether you learn C++ or Java first is a matter of taste, although Java
cares for some things more than C++ will do (without extensions). For
example, Java provides automatic memory allocation and reclamation, while
C++ requires you to care for that.

On the other hand, the Syntax of Java and C++ is _very_ similar. I state if
you 'speak' C++ and Python, Java is basically for free (minus some time
learning the few differences from C++ syntax). So you might also consider
Python->Java->C++->C.

Whatever direction you take, the above mentioned books provide a very good
start into C++ and Java in my opinion.

For language details, you can consider the much-mentioned K&R book for C.
The similar book for C++ is Bjarne Stroustroup's 'The C++ Programming
Language'. At least the latter is not an introductory book, in my opinion.
But nevertheless worth reading (later).

hth
Werner


"Ryan" <ryanmorillo at hotmail.com> wrote in message
news:4a7fde99.0204081858.58baabdd at posting.google.com...
> I know this is way off topic, but I figure the heavy hackers here will
> understand my mindset better than those at comp.lang.C
>     I have been doodling with python (nasty habit I know but I have to
> do somthing with my snake) and think I may be ready to learn a second
> language.  I am leaning towards C over C++ and Java even though I
> loose objects it will ground me firmly in functional programing.  I am
> wondering what book would be the best for a begginer?  The two I have
> seen that look interesting are
> C for Dummies (parts one and two) or
> The C Programming Language
> by Brian W. Kernighan, Dennis M. Ritchie.
> I am looking for some input from someone who knows and knows the
> problems I will face coming from a __nice__ language like python.
> Thanks (and sorry about the off topic)




From grante at visi.com  Mon Apr 22 10:50:20 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 22 Apr 2002 14:50:20 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019483442.32389.python-list@python.org>
Message-ID: <MqVw8.47908$vm6.7506303@ruti.visi.com>

In article <mailman.1019483442.32389.python-list at python.org>, brueckd at tbye.com wrote:
>
>> > You should understand that this is a philosophical viewpoint. The
> [snip]
>> To put this another way, ask yourself where you stand on this
>> question: is mathematics discovered or invented?

Very well put.

> Is there a difference?

Yes.  Science verifies it's theories by comparing them to the
physical world which they are attempting to describe.  It
doesn't matter how elegant, or internally consistent Science
is.  If it doesn't successfully describe the physical world,
it's wrong.

Science has an external reference point. Mathematics does not.
Internal consistency is the only thing which mathematics can
attempt to verify.  Mathematics is not an attempt to describe
the physical world.

> One could argue that the only difference between discovery and
> invention is the arrogance of the person reporting it.

One could.  But one would be missing the point.

-- 
Grant Edwards                   grante             Yow!  Of course, you
                                  at               UNDERSTAND about the PLAIDS
                               visi.com            in the SPIN CYCLE --


From baas at ira.uka.de  Tue Apr  9 02:53:09 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Tue, 09 Apr 2002 08:53:09 +0200
Subject: How do I generate 1/f (pink) noise in Python
References: <3CB17354.60609@mxm.dk>
Message-ID: <ne25bu0udkp2dmh9o0k2cv9t72cmuj9r53@4ax.com>

On Mon, 08 Apr 2002 12:39:16 +0200, Max M <maxm at mxm.dk> wrote:
>I am dabling in algorithmical composition and at times I need to 
>generate random values. Often the values should have the characteristics 
>of pink noise (1/f).

I've just released a collection of modules which contains a noise
module, see http://cgkit.sourceforge.net
Maybe this is what you want? You pass up to 3 floating point values to
the noise function (Perlin noise) and it'll return a pseudo random
value in the range from 0 to 1 (or -1 to 1). The output is dependent
on the input and will always be the same if the input stays the same
and the output is band limited.

(BTW, I've just recently noticed that you won't be able to import the
module under Linux. This is fixed for the next release, so tell me if
this module is really what you were looking for and you need it for
something other than Windows)

- Matthias -



From fperez528 at yahoo.com  Mon Apr 15 16:19:44 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 14:19:44 -0600
Subject: String interpolation question
Message-ID: <a9fcmo$npi$1@peabody.colorado.edu>

Hi group,

I have a string interpolation problem and can't seem to find a clean solution.

Suppose I have a list and an index:

y=[5,6]
i=1

I want to print y[i]. The following fails:

print '%(y[%(i)s])s' % locals()

as does the simpler

print '%(y[%(i)s])s' % locals()

Obviously this works:

print '%s' % y[i]

but I don't want that because in my real problem, the string is long and I 
don't want the dangers of positionally matched %s arguments.

By the way, Itpl interpolation here also crashes:

printpl('$y[$i]')

gives an exception.

I know that pep215 never gets anywhere everytime it gets pushed, but I 
_really_ feel that string interpolation is about the only place where python 
simply sucks (IMHO ;) I keep running into situations where the only solutions 
that work are either kludgy or require way too much work to be clean.

Thanks for any help,

f.


From hfoffani at yahoo.com  Wed Apr  3 07:22:36 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Wed, 3 Apr 2002 14:22:36 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>  <mailman.1017799668.13959.python-list@python.org> <slrnaalr5a.fa8.mlh@vier.idi.ntnu.no> <20020403070637.620$SQ@news.newsreader.com>
Message-ID: <20020403072238.382$WC@news.newsreader.com>

> > >or is that, in the words of the late Milton Berle:
> > >
> > >     You can fool some of the people some of the time, and all of the
> > >people some of the time, and that's usually enough.
> >
> > Isn't that "all of the time" in the first sentence? Otherwise it
> > doesn't make much sense (in that the first sentence would follow from
> > the second).

I wrote:
> Berle's missing "all of the time" is on pourpose. The original quote is:
>       You can fool some of the people some of the time and all of
>       the people some of the time, but you can't fool all of the people
>       all of the time. --Abraham Lincon

Ups! Yep, I was fooled. :-) Too late for a cancel.
I reread it and notice what you meant.
Sorry.

Regards,
-Hernan





From ron.teeter at supplysolution.com  Tue Apr 30 18:50:25 2002
From: ron.teeter at supplysolution.com (Ron Teeter)
Date: Tue, 30 Apr 2002 18:50:25 -0400
Subject: Why do class methods share mutable defaults between instances?
In-Reply-To: <000701c1f094$00688380$8210a8c0@singularity>
Message-ID: <001801c1f099$6b1252b0$8210a8c0@singularity>

Asked and answered... little more digging around and I found this...

Python documentation, section 4.7.1, excerpt:
"""
Important warning: The default value is evaluated only once. This makes a
difference
when the default is a mutable object such as a list or dictionary. For example,
the following function accumulates the arguments passed to it on subsequent
calls:

def f(a, L=[]):
    L.append(a)
    return L

print f(1)
print f(2)
print f(3)
This will print


[1]
[1, 2]
[1, 2, 3]

If you don't want the default to be shared between subsequent calls, you can
write
the function like this instead:

def f(a, L=None):
    if L is None:
        L = []
    L.append(a)
    return L
"""

Chalk it up to insufficient familiarity with default value initialization.

Thanks anyway,

Ron Teeter

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Ron Teeter
> Sent: Tuesday, April 30, 2002 6:12 PM
> To: python-list at python.org
> Subject: Why do class methods share mutable defaults between
> instances?
>
>
> Greetings,
>
> Today I made a rather perplexing discovery.  It appears that
> mutable defaults
> in class methods are shared between instances.  Here is a
> simple example of
> the scenario.
>
> class Demo:
>     def __init__(self,name,dict={},list=[]):
>         self.name = name
>         self.dict = dict
>         self.list = list
>         self.dict[name] = name
>         self.list.append(name)
>
> def dump(name,dict={},list=[]):
>     print name,dict,list
>
> for name in ['A1','B1','C1']:
>     tmp = Demo(name)
>     dump(tmp.name,tmp.dict,tmp.list)
>
> >>>
> A1 {'A1': 'A1'} ['A1']
> B1 {'B1': 'B1', 'A1': 'A1'} ['A1', 'B1']
> C1 {'C1': 'C1', 'B1': 'B1', 'A1': 'A1'} ['A1', 'B1', 'C1']
> >>>
>
> My intent was that each time Demo was instanciated a new
> dictionary would be
> assigned to 'dict' if there was no corresponding argument.
> However, the
> behavior is as though the {} were replaced with a class constant.
>
> In order to avoid this behavior I had to reorganize Demo as follows:
>
> class Demo2:
>     def __init__(self,name,dict=None,list=None):
>         if dict==None:
>             dict = {}
>         if list==None:
>             list = []
>         self.name = name
>         self.dict = dict
>         self.list = list
>         self.dict[name] = name
>         self.list.append(name)
>
> I am wondering if this is a bug or a subtle usage error on my
> part.  I would
> expect that syntax on functions would be valid for class
> methods as well, but
> in this case you get drastically different behaviors.
>
> Thanks,
>
> Ron Teeter
>
>
>





From boud at valdyas.org  Thu Apr 11 13:57:43 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 11 Apr 2002 16:57:43 -0100
Subject: Help: Python for a big commercial project?
References: <F194tIpqqJgSt51NDa200002509@hotmail.com> <mailman.1018464082.15807.python-list@python.org> <roy-EB08BB.20470710042002@news1.panix.com> <60F7BB039BFDD68F.7FE6889DFED7EA98.93812C99068FD653@lp.airnews.net>
Message-ID: <3cb5a4b1$0$85463$e4fe514c@dreader3.news.xs4all.nl>

Cameron Laird wrote:

> Let me reinforce this message.
> 
> Python is indeed "not very good at low-level character
> manipulation" (and other operations) *in "typical C-style"*.

I'll deinforce it :-). I recently had occasion to write a
couple of scripts to convert my novel from it's usenet-style
markup to latex or html, and I copied Knode's usenet-markup
parser to Python. (I was pretty feverish at that time and not
able to think coherently. That's why I was coding instead of
writing.)

Anyway, this is a plain silly loop over every character in the 
input files, like:

          s = f.read()
        i = 1

        while i < len(s):
            c = s[i]
            if c == "<": # Comments are ignored
                while c != ">":
                    i+=1
                    c=s[i]

            elif c == "-":
                if s[i+1] == "-" and s[i+2] != "-":
                    # mdash
          ...

But it still only takes 8 seconds to parse half a million
characters, which is pretty good going. I don't feel I will
have to do the rewrite I promised myself. After all, I'm
at 80% of the intended verbiage.

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


From prova at qui.com  Sun Apr 28 09:26:54 2002
From: prova at qui.com (Utente)
Date: Sun, 28 Apr 2002 13:26:54 GMT
Subject: Help again: py2exe and win98
References: <LxPy8.72046$vF6.2149827@news2.tin.it> <aaghao$d9u$04$1@news.t-online.com>
Message-ID: <yMSy8.29394$8D3.782164@news1.tin.it>

"Achim Domma" <achim.domma at syynx.de> ha scritto nel messaggio
news:aaghao$d9u$04$1 at news.t-online.com...
> "Utente" <prova at qui.com> wrote:

[...]

> I don't know anything about py2exe, but BeginUpdateResource is a function
> which is not supported on Win95/98. Look at http://msdn.microsoft.com if
you
> are interested in details.
>

Thanks for the info.
I found that my problem is known (and not solved) also on sourceforge as a
runtime error. I followed some links and found a dll from Ms that should
contain some extensions (unicows.dll). I tried to copy it in the windows's
system subdirectory, restart the system and re-compile but i had no success
again :(





From paul at boddie.net  Tue Apr  9 04:12:17 2002
From: paul at boddie.net (Paul Boddie)
Date: 9 Apr 2002 01:12:17 -0700
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com> <3CB1CEC0.F324F480@python.org>
Message-ID: <23891c90.0204090012.7fc80c1f@posting.google.com>

Guido van Rossum <guido at python.org> wrote in message news:<3CB1CEC0.F324F480 at python.org>...
>
> Paul wrote:
> >
> > As far as the 2.2 features are concerned, they seem to have a
> > reputation in certain circles as being badly-defined, immature and
> > "best steered clear of until later".
> 
> Huh?  I'm not sure where you heard that.  Alex Martelli himself just
> told me that he is using it and considers it stable.  Maybe the people
> who say this themselves deserve a reputation as immature and "best steered
> clear of"?

Perhaps their views were tainted by the need to support releases of
Python lower than 2.2. I'm sure that there are many nice features
which reduce the amount of code (along with the complexity of the
code) substantially in Python 2.2, but it isn't always fair to expect
everyone to upgrade Python every time someone releases a new piece of
software.

Even though I use 2.0 and only have 2.1 to hand because an increasing
number of packages required it over time (most of which I only
evaluated anyway), I couldn't see why I also had to install 2.2 when
it was released because the odd package required it, arguably on an
unnecessarily frivolous basis.

I suppose one argument against a "super upgrade" is that backwards
compatibility is readily abandoned (by package authors), but then,
regardless of the scale of the differences between releases, I would
estimate that there is still a fair amount of coding for compatibility
on a number of the larger Python projects. (Although PyXML's support
for Python 1.5.2 is possibly too heroic to be practical any more.)

Paul


From -$P-W$- at verence.demon.co.uk  Sat Apr 20 12:36:40 2002
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: 20 Apr 2002 17:36:40 +0100
Subject: source code cross reference tool
References: <mailman.1019208819.9434.python-list@python.org>
Message-ID: <a9s5eo$6hp$1@verence.demon.co.uk>

In article <mailman.1019208819.9434.python-list at python.org>,
Tripp Scott  <tripps81 at yahoo.com> wrote:
>does there exist any? or, has anyone used lxr to cross reference 
>a python source code tree? it would be cool to have the source 
>code in a set of html pages that allows a user to click around 
>and jump to function/class definitions, etc.

You could get ctags from http://ctags.sf.net/ and build one (ctags
understands Python source).

-- 
Paul Wright | http://pobox.com/~pw201 |


From vjovanov at stevens-tech.edu  Thu Apr 25 12:07:29 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Thu, 25 Apr 2002 12:07:29 -0400
Subject: PyShell environment variables question
Message-ID: <3RVx8.53$8t1.1919104@newnews.cc.stevens-tech.edu>

I noticed that if I import idle and do idle.PyShell.main() I get
a new shell which contains all of the variables contained in the
root shell.  Is it possible to do idle.PyShell.main() and specify
that the new shell contains variables defined in some dictionary?

Vin.





From ajs at ix.netcom.com  Tue Apr 30 20:09:37 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 30 Apr 2002 20:09:37 -0400
Subject: PEP 285: Adding a bool type
Message-ID: <000f01c1f0a4$9f176d60$0334fea9@carol>

These debates are difficult for some of us to follow.

Partly because all the participants reference other
language implements with which they are fluent.

I have some limited fluency in one language -
Python.

If I am concerned about the readability of my code
for other non-experts (and I usually am), I might
do a:

True = 1
False = 0

at a modules head and go from there.

I am not understanding what beyond that the
Boolean type accomplishes.

It would perhaps not detract from the debate
if the upside could be explained in terms that 
spoke to the practicalities of writing workable 
and readable code in Python, rather than in 
terms of language design theory and 
implementation internals.
 
Art




From ak at silmarill.org  Tue Apr  2 10:18:58 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 02 Apr 2002 15:18:58 GMT
Subject: creating a module
References: <3CA87B34.CF1128F7@ecn.purdue.edu>
Message-ID: <slrnaajj28.s7q.ak@ak.silmarill.org>

In article <3CA87B34.CF1128F7 at ecn.purdue.edu>, Christopher N. Deckard wrote:
> I'm having some difficulties making my python module available to
> other programs besides itself.
> 
> Say I have a module called Thinger, and a basic file layout like:
> 
> Thinger (a directory)
>        |- Thinger.py
>           from Thinger_definitions import *
>           from Commands.Command import Command
>           from Commands.Command1 improt Command1
>           class Thinger:
>             methods...
> 
>        |-  Thinger_definitions.py (global definitions for Thinger 
>                                    services.  Every other file has
>                                    import * from
> Thinger_definitions)
> 
>        |-  ThingerConnection.py (handles connections to a server)
>            class ThingerConnection:
>               methods...
> 
>        |-  Commands (a directory)
>                    |- Command.py (base command class)
>                       from Thinger_definitions import *
>                       class Command:
>                         methods...
>                    |- Command1.py
>                       from Thinger_definitions import *
>                       class Command1(Command):
>                         methods...
> 
> If I create a script in the Thinger directory, which uses the
> Thinger class, then everything works fine.  If I make a script
> outside of Thinger (parent directory) and do something like
> 
> test_script:
>   from Thinger.Thinger import Thinger
>   my_thing = Thinger()
> 
> Then this fails when the files in the Command directory try to
> import Thinger_definitions.
> 
> Really, what I'd like to do is to be able to say in Thinger.py
> 
> from Commands import *
> 
> and that would import all of the Commands that were in there and
> make them directly useable as Command(), Command1(), etc.
>
Set PYTHONPATH env variable to the list of directories you want to
import modules from.

> 
> Another question along these lines is what shouls go in the
> __init__.py script in each directory?  Right now I just have the
> keyword 'pass' and that seems to do fine for most things.
> 
> Thanks,
> -Chris
> 
> P.S.  If you could, please send mail to me as well as the group.


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From fredrik at pythonware.com  Wed Apr  3 12:04:07 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 17:04:07 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com>
Message-ID: <bCGq8.36869$l93.7327515@newsb.telia.net>

Paul Rubin wrote:
>
> It's a cost/benefit thing.  Type/class unification may have cost more
> in complexity than bools cost, but it has high benefits.  The bool
> thing would a complicated change with limited benefit, so it's provoked
> some sharp reactions.

well, if you study the PEP carefully, it isn't a complicated
change.  had GvR just checked it in, most people wouldn't
have noticed.

but the cost is still > 0, and the benefit is close to 0.  it's
hard to be enthusiastic; the time spent on this could have
been spent on more useful stuff...

> It's usually easier to justify ADDING stuff than CHANGING stuff.

I assume you mean adding stuff to the library, not adding stuff
to the language.

</F>




From aleax at aleax.it  Thu Apr 18 15:47:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 19:47:58 GMT
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <nzCv8.102885$%8.8427374@bin2.nnrp.aus1.giganews.com>
Message-ID: <OpFv8.2484$8D3.57452@news1.tin.it>

Terry Reedy wrote:
        ...
> which, only some types had/have methods.  Tuples still do not have
> methods (that I know of).  Strings did not until fairly recently.

Python 2.2.1 (#1, Apr 15 2002, 17:55:14)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> a=2,3,4
>>> dir(a)
['__add__', '__class__', '__contains__', '__delattr__', '__doc__', 
'__eq__', '__ge__', '__getattribute__', '__getitem__', '__getslice__', 
'__gt__', '__hash__',
'__init__', '__le__', '__len__', '__lt__', '__mul__', '__ne__', '__new__', 
'__reduce__', '__repr__', '__rmul__', '__setattr__', '__str__']
>>>

That's about 20 methods plus static-method __new__.  Of course,
they ARE all special-methods -- maybe that's what you meant.
Only about half are also in type object, by the way -- the others are
specifically added by type tuple.

Special methods are normally invoked via operators, built-in functions,
and/or statements -- e.g., x.__str__() is normally called indirectly via
        print x
or
        str(x)
although of course you're not forbidden from calling one directly:

>>> (5,6,7).__str__()
'(5, 6, 7)'
>>>


This subtle arrangement allows built-ins to potentially "add value" to
(special) methods.  For example, we see that tuples don't define a
method __cmp__.  Despite this, built-in cmp() can still compare two
tuples, by relying on their __eq__ and __lt__ methods.  And vice
versa, if some type defines only __cmp__, a < b can still be obtained,
by relying on that.  So, doing all of your comparisons through operators
or the cmp() built-in gives you better polymorphism than calling the
methods directly.  Similar _potential_ considerations apply in many
other cases, even though the language should not be exploiting all
of them at any given point in time.


Alex



From phd at phd.pp.ru  Wed Apr  3 03:06:56 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 3 Apr 2002 12:06:56 +0400
Subject: Can't find Python2c
In-Reply-To: <3caa27bc.5645718@mammoth.usenet-access.com>; from netzapper@magicstar.net on Tue, Apr 02, 2002 at 09:51:56PM +0000
References: <3caa27bc.5645718@mammoth.usenet-access.com>
Message-ID: <20020403120656.D23871@phd.pp.ru>

On Tue, Apr 02, 2002 at 09:51:56PM +0000, A. Jones wrote:
> I've looked all over for an active site that has a copy of Python2c,
> however, I'm unable to find it.

   The project was abandoned long ago. Use
http://homepages.ulb.ac.be/~arigo/psyco/ or
http://sourceforge.net/projects/py2cmod/ instead.

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



From sholden at holdenweb.com  Thu Apr 11 12:29:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 11 Apr 2002 12:29:08 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com>
Message-ID: <wTit8.1381$F43.347@atlpnn01.usenetserver.com>

"Peter Hansen" <peter at engcorp.com> wrote ...
[ botquote ]
>
> I'd like to take this opportunity to at least thank the developers,
> bots included, and other contributors (in the most heartfelt way I
> can without sending large cash contributions) for all the blood,
> sweat and keypresses they've put into this phenomenal project.
>
Hear hear!

> I'm classified in the pathetic subset of users known as "commercial
> developers building products using Python and to date not contributing
> a damn thing back to the community".  My accomplishments in the
> past two years using Python have been possible largely because of
> the unpaid, thankless (well, until now by me anyway) efforts of
> a large number of mostly volunteer developers.
>
Personally I regard your frequent posts on this group as a significant
contribution to "the community" (maybe because I'd like my own to be
regarded in the same way), so I think you're being entirely too
self-denigrating.

> We're leeches (so far) but at least we're very grateful leeches.
>
Gratitude don't pay the bills. Fortunately, Zope Corporation currently does,
God bless them.

> I'm dismayed by comments attacking the core developers when made
> by people who have not contributed more than I or my team have.
> Beggars can't be choosers, and nobody likes a whiner...
>
Ignorance is bliss, I suppose. Unfortunately your comments will be
disregarded by the people you are complaining about because either

a) they will have no idea you are referring to them, or
b) they will respond to your comments by subjecting you to similar attacks.

 This accounts for the low signal-to-noise ratio on many high-volume
newsgroups. But then I'm pretty sure you knew that already ...

regards
 Steve






From core.lists.python at core-sdi.com  Mon Apr 29 18:14:51 2002
From: core.lists.python at core-sdi.com (Maximiliano Caceres)
Date: Mon, 29 Apr 2002 19:14:51 -0300
Subject: Python as scripting plug-in for large C++ project
References: <mailman.1019880409.19217.python-list@python.org> <3CCAF576.3060304@dittmar.net> <aaf5hb$a6vgj$1@ID-139865.news.dfncis.de>
Message-ID: <3CCDC5DB.8040605@corest.com>

kuhlman wrote:
> SWIG is great.  But, if you decide to write your glue code by hand, 

take a look at Boost.Python
(http://www.boost.org/libs/python/doc/index.html)

max/


--- for a personal reply use: Maximiliano Caceres <maximiliano.caceres at corest.com>




From vlindberg at verio.net  Thu Apr 18 01:04:40 2002
From: vlindberg at verio.net (VanL)
Date: Wed, 17 Apr 2002 23:04:40 -0600
Subject: Pure python crypto
References: <20020417021703.H16954@eecs.tufts.edu> <02041718383201.00171@toad> <mailman.1019063074.28268.python-list@python.org>
Message-ID: <3CBE53E8.8080707@verio.net>

[Snip Various]

If you are implementing a sshlib module (for lack of a better term) I 
would be *very* interested.

VanL



From greg at cosc.canterbury.ac.nz  Mon Apr 15 22:46:51 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 16 Apr 2002 14:46:51 +1200
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu>
Message-ID: <3CBB909B.2D097D32@cosc.canterbury.ac.nz>

David Eppstein wrote:
> 
> In article <3CBA5EDB.C60BF821 at cosc.canterbury.ac.nz>,
>  Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> 
> > MacPython has a facility for creating an "applet" from
> > a Python script.
> 
> Which is great for your own use -- But it's not so good for
> distributing programs to others

It's no worse than the equivalent problem under any other
OS, though. I was just trying to give the original poster
some idea of how command-line stuff translates across to
the Mac.

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


From jdhunter at nitace.bsd.uchicago.edu  Mon Apr 22 19:06:48 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Mon, 22 Apr 2002 18:06:48 -0500
Subject: flatten tuple
Message-ID: <m2adrvnxiv.fsf@mother.paradise.lost>

I have a tuple of strings and a map from strings to either a string or
a tuple of strings

tuple of strings:
 
  x = ( 'John', 'Hunter', 'where' )


map:

  m = {
       'where': ('was', 'here'),
       'somekey': 'someval'
      }


I want to be able to create a new tuple, which is the expanded and
flattened version x, so that if an element of x is a key of m, then
the new list has that value instead of the key value .  In the example
above, the expansion would give

y = ( 'John', 'Hunter',  ('was', 'here') )

and the flattening would give

y = ( 'John', 'Hunter',  'was', 'here' )

The only hitch is that I want to be able to do this recursively, so
I could have for example

m = {
    'where': ('was', 'somekey'),
    'somekey': 'someval',
    'someval': ('asking', 'really', 'silly', 'questions')
    }

and end up with 


y = ( 'John', 'Hunter',  'was',  'asking', 'really', 'silly', 'questions')

I see that I expose myself to the dangers of infinite recursion which
I should need to guard against, but I am willing to put the burden on
the user not to define recursive tuple/map combinations if checking
against recursion is too difficult.

I would like to use tuples rather than lists because I would like
these tuples to be keys in a hash.

Suggestions?

John Hunter



From mlh at vier.idi.ntnu.no  Thu Apr 11 22:20:32 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 02:20:32 +0000 (UTC)
Subject: Tuple comprehension
References: <mailman.1018502663.25698.python-list@python.org>
Message-ID: <slrnabch3f.c8l.mlh@vier.idi.ntnu.no>

In article <mailman.1018502663.25698.python-list at python.org>, Sean
'Shaleh' Perry wrote:
>
>On 11-Apr-2002 Emile van Sebille wrote:
>> Magnus Lie Hetland
>>> As far as I can see, the PEP on dict comprehension is still open... I
>>> think both dictionary comprehension and tuple comprehension seem like
>>> natural parallels to list comprehension. But then I would, wouldn't I?
>>> <wink>
>>>
>> 
>> Why not just tuple([my list comp]) ?
>> 
>
>because it feels wrong to create a (possibly large) list and then duplicate it
>into a tuple.  My mind sees that and thinks 'overhead'.

Of course, until we get some sort of "freezing", that is what would
have to happen behind the scenes, I imagine.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From NO_SPAM_jmpurser2 at attbi.com  Mon Apr  1 14:06:59 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Mon, 01 Apr 2002 19:06:59 GMT
Subject: readlines() problem, more info
Message-ID: <nd2q8.149996$ZR2.64791@rwcrnsc52.ops.asp.att.net>

Hello,

I've got a little function:

def FindLine(Record):
     im1_Dif.seek(0)
     temp = im1_Dif.readlines(Record)
     lngCount = 0L
     for line in temp:
          print str(lngCount), '   ', line
          lngCount += 1
     test = im1_Dif.readline()
     print test

 That I'm using to trouble shoot a small script.  When I run this in the
interactive mode it returns 136 lines to "temp" no matter what number I pass
to it.  Can anyone tell me why this is happening?

I'm using Python 2.2 on Windows 2000.

Thanks,

John Purser




From morelli at cerm.unifi.it  Fri Apr  5 11:36:28 2002
From: morelli at cerm.unifi.it (Enrico Morelli)
Date: 5 Apr 2002 08:36:28 -0800
Subject: I'm going crazy using cgi module. HELP PLEASE!!!!
Message-ID: <9c5a942b.0204050836.58595d84@posting.google.com>

Dear all,

I'm going crazy using cgi module. 
Follow a main routine of a program that receives for each day of the
month, six fields from a form.

The first  'print ar'  inside of  'for' cycle, works fine, display the
correct ar contents.
The second one (outside of 'for' cycle) doesn't display nothing.
In fact,  if I try to write the array on a file, the file will be
empty.

I use python2.2 on Linux RH7.2  system.

Where I wrong??? 

I use cgi modules for more and more other forms without problems.

Thanks 

#----------------------------------------#
# Main Program
def main():

        months={1 : 'Gen',
                2 : 'Feb',
                3 : 'Mar',
                4 : 'Apr',
                5 : 'Mag',
                6 : 'Giu',
                7 : 'Lug',
                8 : 'Ago',
                9 : 'Set',
                10 : 'Ott',
                11 : 'Nov',
                12 : 'Dic'}

        MONTH_NUMBER="%m"
        YEAR_FORMAT="%Y"
        currenttime = time.time()
        nmonth=time.strftime(MONTH_NUMBER,
time.localtime(currenttime))
        year=time.strftime(YEAR_FORMAT, time.localtime(currenttime))


        print "Content-Type: text/plain\n\n"

        form = cgi.FieldStorage()


        im=int(form['month'].value)
        month=months[int(nmonth)+im]
        nmonth=int(nmonth)+im

        planningsavefile=planningsave + '-' + month
        dmonth=calendar.monthcalendar(int(year), nmonth)

        for dd in range(len(dmonth)):
                for mm in range(len(dmonth[dd])):
                        if dmonth[dd][mm] == 0:
                                continue
                        ar.append([dmonth[dd][mm],'','','','','',''])

        print ar # works fine
        for name in form.keys():

                if name=='submit':
                        continue
                 # name has the values like: 2-7 or 2-8 or 1-7b or
4-4, etc.
                d=string.split(name, '-')[0]
                s=string.split(name, '-')[1]

                d=int(d)-1
                print name, d,s # Works fine
                if s=='7b':
                        ar[d][2]=form[name].value
                elif s=='8':
                        ar[d][1]=form[name].value
                elif s=='7':
                        ar[d][3]=form[name].value
                elif s=='6':
                        ar[d][4]=form[name].value
                elif s=='5':
                        ar[d][5]=form[name].value
                elif s=='4':
                        ar[d][6]=form[name].value
                print ar # works fine

        print ar # doesn't display nothing


From xnexau at yahoo.com.au  Fri Apr 26 17:51:32 2002
From: xnexau at yahoo.com.au (Patrick W)
Date: 27 Apr 2002 07:51:32 +1000
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <tyf1yd25zgs.fsf@pcitapi22.cern.ch> <mailman.1019840809.5035.python-list@python.org>
Message-ID: <821yd22knv.fsf@acropolis.localdomain>

"James J. Besemer" <jb at cascade-sys.com> writes:

> 5. Finally, I think the problem here is that we're all making
> passionate, definitive statements about items that ultimately really
> boil down to matters of personal preference.

Sounds nice, but it's just not true. In Lisp's case, there are
technical issues related to syntax that go way beyond personal
preference or aesthetics.

Anyone who wants to follow this up and really *learn* something,
instead of listening to a bunch of turkeys arguing about it with
varying degrees of cluelessness, take a look at this book (free,
online, pdf or postscript). Find out for yourselves.  Spend 20 minutes
on it. You'd have to be brain dead beyond redemption to /not/ find
this interesting.

        http://www.paulgraham.com/onlisptext.html

The first three chapters contain a lot of ideas that many Pythonistas
will relate to. The rest of the book takes you into territory that
Python (or any other non-Lisp-related language) does not even attempt
to take you. (If you've never been exposed to these ideas, you can't
really have much idea of what you're missing because you won't find
parallels anywhere else).


From tanzer at swing.co.at  Wed Apr  3 02:01:01 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Wed, 03 Apr 2002 09:01:01 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Mon, 01 Apr 2002 16:21:39 CDT."
             <200204012121.g31LLdk00743@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <E16selR-0005Cn-00@tswings.swing.cluster>

> > Nice tease. But Guido's hammer with regard to future strictification
> > of bool is exactly what makes me uneasy.
>
> Can you elaborate that?  I thought I made it pretty clear I didn't
> want to strictify bool (though I did open up the question as to
> whether other people want that).  Is that what you call a hammer?

You made it clear enough.

There are two things making me uneasy though:

- A lot of people instantly started complaining about the new bools
  being not strict enough. Over time, all that clamour might change your
  POV.

- I don't really understand why you proposed this PEP in the first
  place.

  For me, Python's (old) way of handling boolean conditions is one of
  its great strengths. This wasn't always the case. When I started
  using Python, I was slightly put off by the lack of a boolean data
  type. But quite quickly I learned to appreciate Python's way of
  dealing with this issue.

  All of that was implicit knowledge, though. It took Laura
  Creighton's well written post to make it explicit for me (thank
  you, Laura).

After reading Alex Martelli's and Laura's arguments, I'm convinced
that Python would be better off without this PEP and change to -1.

Glad-that-I'm-not-a-language-designer-ly yr's,
Christian

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




From jon+usenet at unequivocal.co.uk  Mon Apr 29 06:20:36 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 10:20:36 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com>
Message-ID: <slrnacq7jk.gav.jon+usenet@snowy.squish.net>

In article <23891c90.0204290051.3edcc62d at posting.google.com>, Paul Boddie wrote:
> Personally, I believe simple "model-based" (or "placeholder-based")
> templating offers the most reward for the least additional
> complexity (and maintenance), especially for certain kinds of
> applications, and that the extra power offered by less
> "conservative" template frameworks only gives you a certain amount
> of extra flexibility that not all applications really need.

Can you think of an example where the "extra power" provided by a more
complicated templating system actually makes something easier to do
than with jonpy? ;-)


From phr-n2002a at nightsong.com  Fri Apr  5 07:40:15 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 04:40:15 -0800
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it>
Message-ID: <7xbscye2ts.fsf@ruckus.brouhaha.com>

Alex Martelli <aleax at aleax.it> writes:
> Reports of pystones and other Zaurus/SL5000 Python/PyQT experiences
> are VERY welcome -- it's a really sexy little thingy and I'm SORELY
> tempted to have a fling, but getting it in Italy isn't easy as yet...

If you want to benchmark Python on an ARM, there are some ARM
machines you can telnet to:

  http://handhelds.org/projects/skiffcluster.html


From ngps at vista.netmemetic.com  Thu Apr 18 11:05:30 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 18 Apr 2002 15:05:30 GMT
Subject: amkCrypto installation on win32
References: <1103_1019061374@news.free.fr>
Message-ID: <a9mnbq$dfe$1@coco.singnet.com.sg>

According to Fr?d?ric Laurent  <liste_nospam_ at opikanoba.org>:
> Sur 17 Apr 2002 14:38:15 GMT, vous avez ?crit dans comp.lang.python:
> > Try M2Crypto:
> 
> and there is no precompiled dll in the win directory of the binary zip package,

Is too!

$ fetch http://www.post1.com/home/ngps/m2/m2crypto-0.07-snap3-win32.zip 
Receiving m2crypto-0.07-snap3-win32.zip (495847 bytes): 100%
495847 bytes transferred in 25.5 seconds (18.97 kBps)

$ unzip -l m2crypto-0.07-snap3-win32.zip | egrep win
Archive:  m2crypto-0.07-snap3-win32.zip
     1438  08-27-01 20:53   m2crypto-0.07-snap3/tests/test_ssl_win.py
        0  03-08-02 00:30   m2crypto-0.07-snap3/win/
      540  06-01-01 21:58   m2crypto-0.07-snap3/win/README
        0  03-08-02 00:30   m2crypto-0.07-snap3/win/py15/
   122880  03-05-02 22:49   m2crypto-0.07-snap3/win/py15/_m2cryptoc.dll
    20089  03-05-02 22:49   m2crypto-0.07-snap3/win/py15/_m2crypto.py
        0  03-08-02 00:30   m2crypto-0.07-snap3/win/py21/
   118784  03-05-02 22:48   m2crypto-0.07-snap3/win/py21/_m2cryptoc.dll
    20089  03-05-02 22:48   m2crypto-0.07-snap3/win/py21/_m2crypto.py
        0  03-08-02 00:30   m2crypto-0.07-snap3/win/py22/
    19350  03-05-02 22:18   m2crypto-0.07-snap3/win/py22/_m2crypto.py
   114688  03-05-02 22:18   m2crypto-0.07-snap3/win/py22/_m2cryptoc.pyd


I just dropped this into a Python 2.1 on a W2K boxen. Works fine.


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



From trentm at ActiveState.com  Wed Apr 10 15:22:16 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 10 Apr 2002 12:22:16 -0700
Subject: Python FREE IDE for windows
In-Reply-To: <005b01c1e0c3$707dc620$01c8a8c0@raphaelofizs8j>; from gusraan@terra.com.br on Wed, Apr 10, 2002 at 04:10:53PM -0300
References: <005b01c1e0c3$707dc620$01c8a8c0@raphaelofizs8j>
Message-ID: <20020410122216.B23370@ActiveState.com>

[Raphael Ribeiro wrote]
> does anyone know?
> 

Check out:
http://www.python.org/editors/

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From dalke at dalkescientific.com  Fri Apr 19 13:15:32 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Fri, 19 Apr 2002 11:15:32 -0600
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com> <mailman.1019068175.11920.python-list@python.org> <99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com>
Message-ID: <a9pjl6$pj6$1@slb4.atl.mindspring.net>

Dale Strickland-Clark:
>But several orders of magnitude more processing [to use a dictionary
>compared to an and/or solution]. Certainly not for tight, efficient loops.

Perhaps your intuition comes from C?  That doesn't seem to be the
case for Python.

>>> def do_filter1(data):
...     for x in data:
...             v = x == 'm' and 'male' or 'female'
...
>>> def do_filter2(data):
...     d = {"m": "male", "f": "female"}
...     for x in data:
...             v = d[x]
...
>>> def test_size(n):
...     data = ["m", "f"] * n
...     t1 = time.clock()
...     do_filter1(data)
...     t2 = time.clock()
...     do_filter2(data)
...     t3 = time.clock()
...     print "and/or", t2-t1, "dict", t3-t2, "ratio", (t2-t1)/(t3-t2)
...
>>> import time
>>> test_size(10000)
and/or 0.04392 dict 0.02928 ratio 1.5
>>> test_size(100000)
and/or 0.437248 dict 0.298656 ratio 1.46405228758
>>> test_size(1000000)
and/or 4.379312 dict 3.004128 ratio 1.45776478233
>>>

Dictionary looks faster to me, in this small, tight loop.

                    Andrew
                    dalke at dalkescientific.com





From peter at engcorp.com  Tue Apr  9 01:05:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 01:05:55 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <3CB276B3.84B42D6E@engcorp.com>

Ryan wrote:
> 
> ... I may be ready to learn a second
> language.  I am leaning towards C over C++ and Java even though I
> loose objects it will ground me firmly in functional programing.  I am
> wondering what book would be the best for a begginer?  The two I have
> seen that look interesting are
> C for Dummies (parts one and two) or
> The C Programming Language
> by Brian W. Kernighan, Dennis M. Ritchie.
> I am looking for some input from someone who knows and knows the
> problems I will face coming from a __nice__ language like python.
> Thanks (and sorry about the off topic)

Without knowing more about your goals it is difficult to provide
a useful answer (although Aahz did exactly that).

What do you want from your second language?  Do you have
specific tasks in mind?  Specific long-term goals you think
this will move you closer towards?

-Peter


From maxm at mxm.dk  Sat Apr  6 05:02:59 2002
From: maxm at mxm.dk (Max M)
Date: Sat, 06 Apr 2002 12:02:59 +0200
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
References: <mailman.1018040171.16545.python-list@python.org> <3CAE1D58.4050500@mxm.dk> <Xns91E7D088415CFJamesLLugojcom@216.148.53.84>
Message-ID: <3CAEC7D3.6050004@mxm.dk>

James Logajan wrote:
> Max M <maxm at mxm.dk> wrote:


>>>So, what language did you move on to?
>>
>>386 assembler probably. It's a little less portable, but the specs 
>>haven't changed in over a decade.
> 
> 
> I'm at a loss to understand all the snide remarks. Some I understand, some 
> I don't.

You are right. That was a bit rude, but you didn't really give any 
solutions or pointers to alternatives. Just complaints.

Shurely Python doesn't solve all problems, but then you say that you 
have changed away from it to something else, and I really cannot see any 
modern language having the stuff that you look for, I took the liberty 
of a cheap-shot.

Well those are allways easy anyway ;-)

regards Max M



From peter at engcorp.com  Sun Apr  7 22:55:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 22:55:38 -0400
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com> <a8q8jp$fob$1@panix1.panix.com> <Xns91E989929996CRASXnewsDFE1@130.133.1.4> <chris-CFFF94.21350207042002@corp.supernews.com>
Message-ID: <3CB106AA.1AA3A5BF@engcorp.com>

Chris wrote:
> 
> In article <Xns91E989929996CRASXnewsDFE1 at 130.133.1.4>,
>  Philip Swartzleonard <starx at pacbell.net> wrote:
> 
> > def percent_chance( percentage ):
> >       "Assumes percentage as number 0-100"
> >       if random.random() > (percentage / 100.0):
> >               return 1
> >       else:
> >               return 0
> 
> Might more concisely be written as:
> 
> def percent_chance( percentage ):
>      return random.random() > (percentage / 100.0)

But this is Python, and a newbie.  I think Philip's
effort was more suited to the audience, as it is
more readable.  Conciseness is not a good goal, IMHO,
unless you choose Perl as your implementation language.

-Peter


From ngreco at softhome.net  Tue Apr 23 21:20:38 2002
From: ngreco at softhome.net (Nahuel Greco)
Date: Tue, 23 Apr 2002 22:20:38 -0300
Subject: unsuscribe
Message-ID: <20020423222038.20dc4118.ngreco@softhome.net>

unsuscribe

- Nahuel Greco                   Web Development - Open Source - 
- http://nahuel.codelarvs.com.ar Game Programming - Research   -
- Freelance coding / sysadmin    Networking. The answer is 42. -




From Jyrki.Alakuijala at iki.fi  Wed Apr  3 20:28:42 2002
From: Jyrki.Alakuijala at iki.fi (Jyrki Alakuijala)
Date: Thu, 04 Apr 2002 04:28:42 +0300
Subject: Tkinter
Message-ID: <3CABAC4A.1050106@iki.fi>

Would it be possible to offer an API access to the lines

static PyThread_type_lock tcl_lock = 0;
static PyThreadState *tcl_tstate = NULL;

currently hiding in the _tkinter.c.

Currently, I need to use my own tkinter modification to
overcome this problem.

My Tkinter-based program starts out as a python program,
The mouse events are taken as native evens for speed reasons
inside the my own tcl-widget and redirected back to the C++
computation backend. However, when the code execution
goes from there to python , which may invoke Tkinter,
it goes twice through the same Python->Tcl gate, messing
up with the locking system.

With my own tkinter, I can release the lock when I am back
in the C++ side and relock it again when returning to the tcl
system. Going to python via tcl would also be a solution, but
that is an architectural and performance nightmare.

This used to work so nicely with Python 1.4 - no thread support
there(?) - no python->tcl locking was implemented either...

best regards,

Jyrki Alakuijala



From peter at seahaze.demon.co.uk  Fri Apr  5 04:01:48 2002
From: peter at seahaze.demon.co.uk (Peter Hayes)
Date: Fri, 5 Apr 2002 09:01:48 +0000 (UTC)
Subject: sharp sl5000
Message-ID: <slrnaaqpvc.2vl.peter@vlad.seahaze>

Does anyone know if there is a version of Python that runs on the new
Linux based Sharp SL5000 PDA?

Thanks

-- 

Peter


From lac at strakt.com  Sat Apr  6 04:57:40 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 11:57:40 +0200
Subject: Selecting cipher suites with socket.ssl 
In-Reply-To: Message from martin@v.loewis.de (Martin v. Loewis) 
   of "05 Apr 2002 20:41:34 +0200." <m3it769ee9.fsf@mira.informatik.hu-berlin.de> 
References: <84cff98e.0204041830.7c0d0171@posting.google.com> <mailman.1017978613.9042.python-list@python.org> <m3y9g21uea.fsf@mira.informatik.hu-berlin.de> <200204051346.g35Dkerq000532@ratthing-b246.strakt.com>  <m3it769ee9.fsf@mira.informatik.hu-berlin.de> 
Message-ID: <200204060957.g369vedj016999@ratthing-b246.strakt.com>

Sent twice, apologies to Martin.  Some fool program is mangling things
so that Replied python-list at python.org lines come out python-list at python.
So that didn't work.  Now to find that bug.  Sorry that I did not
notice it until I sent this.

Laura
---------
> Laura Creighton <lac at strakt.com> writes:
> 
> > Why should the OP do this rather than just get our PyOpenSSL, or m2crypto
> > (which also didn't work a year ago when we tried it, either, but I believe
> > does now)?  Why not put our PyOpenSSL (or m2crypto) into Python?  This is
> > a serious question which nobody has answered for nearly a year now.
> 
> I think there are three reasons:
> - none of the authors has offered to contribute them;
> - none of the users have requested an inclusion of a specific such
>   package into Python
> - it is not clear which of these are "good" enough for the standard
>   Python library.
> 
> The third item is particularly important, IMO - apparently, the
> previous attempt at offering SSL support is considered as failed by
> those that need advanced features; those that don't need advanced
> features (like myself) are happy users of the Python SSL support.
> 
> It is IMO pointless to add another package that then will be
> considered as a failure after some time. This stuff is difficult, so
> people need to sit down and identify the requirements, and propose an
> API.
> 
> Outright exposure of the OpenSSL API is IMO the wrong thing to do:
> parts of that are historic, and parts of it are better done in
> different ways in Python than they are in C. What I like about the
> current SSL support in Python is that simple things are easy to do; I
> fear that those packages aiming at exposing OpenSSL fully don't allow
> to do easy things easily.
> 
> Regards,
> Martin

We appear to have a simple misunderstanding. 
1. We offered to contribute them.  it was refused I think due to 2.
2. You are absolutely correct about that, there appears to be almost
   no interest in them, or in certificate generation whatsoever.  About
   every 3 months one more person is found who wants one, which is a
   tiny user community.
3. Well, the only way you can find that out is to read the code.
   I cheerfully admit, given 2, there is probably a better use for your time.

But I don't like the assumption that the OP can do certificate generation
better than Martin Sj?gren did.  Maybe you are already familiar with
the OP, (or hate our code, and are being too polite to tell us, even
privately.)  Meetings for common API, though proposed, have come to
naught, basically due to lack of interest.  Those of us who have working
things wrote them for ourselves, so naturally we like them, but as
for 'giving back to the community ...' as far as we can 10 of the 27
people in the whole world who want certificates already work here ...
(I made up the 27; I have no numbers, but it is very small.)

So naturally, there is no big push here to get pyOpenSSL in the standard
library.  But I don't think that the OP should waste his time hacking
on the standard SSL module; pick something that other people want to see.
Besides, whatever he comes up with may be rejected on the grounds that the 
code isn't good enough, or it adds the sort of complexity you don't want in
the standard library.  Modifying an existing library shouldn't be
by definition easier to do than adding a new one -- foul code is foul code


From Gareth.McCaughan at pobox.com  Wed Apr 17 20:09:58 2002
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Thu, 18 Apr 2002 01:09:58 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com> <a9gbhj$dhm$1@panix1.panix.com>
Message-ID: <slrnabs3mm.1kfm.Gareth.McCaughan@g.local>

Aahz wrote:

> In article <3CBBAB98.91FA1B17 at alcyone.com>,
> Erik Max Francis  <max at alcyone.com> wrote:
> >jepler at unpythonic.net wrote:
> >>
> >>     [2, 3, 5, 7, 9, ...]
> >> would generate successive primes.
> >
> >Pretty lousy code, since 9 is not prime.  :-)
> 
> That just proves he's an engineer (or is it a physicist? I can never
> remember that joke properly).

Engineer.

A mathematician, a physicist and an engineer set out to
prove that all odd numbers are prime.

Mathematician: 3 is prime, 5 is prime, 7 is prime ...
well, it's obviously a simple proof by induction. Trivial!

Physicist: 3 is prime, 5 is prime, 7 is prime, hmm,
9 isn't prime: experimental error. 11, 13, ... yeah,
must be right.

Engineer: 3 is prime, 5 is prime, 7 is prime, 9 is prime,
11 is prime, 13 is prime, 15 is prime, ... [fades into distance]

(These days there ought to be a computer scientist in the
mix too. Suggestions?)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc


From nika at kassube.de  Fri Apr 26 16:00:19 2002
From: nika at kassube.de (Nils Kassube)
Date: 26 Apr 2002 22:00:19 +0200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz>
Message-ID: <87d6wms018.fsf@kursk.kassube.de>

Paul Foley <see at below> writes:

> In any case, parentheses are not used around sub-expressions, or to
> separate or group statements.  They're only used (in evaluated
> contexts) for "function" calls, and most other languages, including
> Python, use parentheses for that purpose, too:
> 
>  Python: foo(x)
>  Lisp: (foo x)
> 
> Same number of parentheses, right?  I'd expect Lisp syntax to be

Python: a+b/c
Lisp: (+ a (/ b c))

Please, I really do like Common Lisp, but it doesn't help to offer
wrong arguments in favor of it. The big advantage of the s-expression
syntax used by CL is that it's very easy to parse the language.  So
you're encouraged to write macros and build your own sub-language on
top of it.


From gh_pythonlist at gmx.de  Wed Apr 24 00:58:46 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 24 Apr 2002 06:58:46 +0200
Subject: Accessing class-level variables
In-Reply-To: <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4>
References: <mailman.1019621385.21532.python-list@python.org> <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4>
Message-ID: <20020424045845.GA1649@lilith.my-fqdn.de>

* Philip Swartzleonard <starx at pacbell.net> [2002-04-24 04:27 +0000]:
> Gerhard =?iso-8859-15?Q?H=E4ring?= || Tue 23 Apr 2002 09:04:04p:
> 
> > I'm trying to access a class variable in a derived class. I sincerly
> > hope there's a less kludgy way than what I'm currently using:
> > [...]
> > print self.__class__.a
> 
> Well, i knew this worked for dispatching __init__'s, but... well here 
> =):
> 
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
> >>> class A:
> 	d = 5
> 
> >>> class B:
> 	def p(self):
> 		print A.d
> 
> >>> z = B()
> >>> z.p()

Erh. Here you are accessing the class variable of class A explicitely. I
know how that's working :) But it's not what I want. I want to access
the class variable in a derived class. And even if I don't know what my
derived class is. Here's an example that is more to my real code:

import new

class X:
     def p(self):
         print self.__class__.a

def make_class(val):
    klass = new.classobj("Y", (X,), {"a": val})
    return klass

y = make_class(5)()
y.p()


This works, but using self.__class__.a is ugly.

Heh, now that I come to think of it, here I somehow have the class
inheritance backwards :-) Guess there's no other way, right?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 9.5 ?C      Wind: 3.3 m/s




From robin at jessikat.fsnet.co.uk  Fri Apr 12 22:45:08 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 03:45:08 +0100
Subject: hex and long
Message-ID: <8N9+ELA0u5t8Ew2n@jessikat.fsnet.co.uk>

In 2.2 I find the following peculiarity

>>> 0xffffffff
68719476735L
>>> 0xffffffff
-1
>>>

clearly the second is being treated as a 32 bit int and the first as a
long. Do I have to be careful with bit 31? What happens on 64 bit
machines? 
-- 
Robin Becker


From ponderor at lycos.com  Sat Apr 20 17:35:12 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 20 Apr 2002 14:35:12 -0700
Subject: Python for .NET vs. Jython ... Python.Net thoughts?
References: <f57664b9.0204152215.57312efa@posting.google.com> <mailman.1018976552.29487.python-list@python.org>
Message-ID: <e81be8b2.0204201335.77865efd@posting.google.com>

Paul,

Forgive me if this isn't too off topic, but it's scratched a dormant
curiousity:

Disclaimer: It's been months since I've read the ActiveState white
paper on Python and .NET, but I'm consistently grimacing when a .Net
developer says something like "There's even a .NET compiler for
Python."

What's the most practical way to get Python into the a .NET
development environment?

- CLR Python compiler : True "Python.net"

- C# implementation , simpilar to Jython : "Shython" ??

- Others???


Are any of these in development?

Projects such as bCodeParser 
(http://www.software4net.com/products/bcodeparser.html ) tease at
thinking a Jython -> CLR is even possible...   There's another one
that does .NET -> Java and I can't find the link. :(

Finalley, forgive me if I mis-used CLR where I should have used IDL.

Best Regards, and wishing I had more time with Jython to know if I
should even wish for a Shython.

Surrounded by C++ & C#,

Dean


From fperez528 at yahoo.com  Tue Apr 23 03:19:38 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 23 Apr 2002 01:19:38 -0600
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org> <aa2460$534$1@peabody.colorado.edu> <mailman.1019533244.22752.python-list@python.org> <aa2u72$k0a$1@peabody.colorado.edu> <mailman.1019545127.27780.python-list@python.org>
Message-ID: <aa3222$lna$1@peabody.colorado.edu>

holger krekel wrote:

>> Currently ipython seems to do these just fine, after removing the following
>> from the delimiters:
>> 
>> readline_remove_delims -/'"[]{}()
> 
> does
> 
>   b=[]
>   a=a-b.<tab>
> 
> work then? And
> 
>   [[].<tab>
>   
> 
> really works, too?

Well, almost :)

In [2]: b=[]

In [3]: a=a- b.<TAB>
b.__add__           b.__getslice__      b.__ne__            b.count
b.__class__         b.__gt__            b.__new__           b.extend
[snip]

In [3]: [ [].<TAB>
[].__add__           [].__iadd__          [].__setitem__
[].__class__         [].__imul__          [].__setslice__
[].__contains__      [].__init__          [].__str__
[snip]

The trick is the extra spaces before the b. and the []. I know it's cheating, 
but I don't see a straightforward solution to this one right now. I can get 
the a-b.<tab> case to work if I don't remove '-' from the delimiters, but I 
don't want to do that b/c then filenames with - in them don't get completed. 
I'd rather have filename completion and have to put in the occasional extra 
space in an expression. Note though that the characters removed from the 
delimiters are a user setting in ipython, so anyone can change them to their 
taste.

The second case is obviously far trickier :)

Cheers,

f.


From hjwidmaier at web.de  Mon Apr 15 10:08:34 2002
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 15 Apr 2002 07:08:34 -0700
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018825230.16052.python-list@python.org>
Message-ID: <6e990e29.0204150608.44bf1fe1@posting.google.com>

Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1018825230.16052.python-list at python.org>...

> Guido is very reluctant to break code, but it's not an absolute for him:
> among other things, it competes with the desire to fix design mistakes in
> the language, which is unusual for a language designer (not to realize that
> some ideas were mistakes, but to have enough guts to repair them).  Guido is
> also much more sympathetic to not breaking code that was in fact forbidden
> by the docs than any C or C++ vendor would be.  The __future__ + warning
> mechanism also gives explicit and controllable warnings about incompatible
> changes for at least one release cycle before they happen.  Heck, I've spent
> more time writing replies in this thread today than I spent adjusting to
> Python changes in my own code since Python 0.9.6 was released.

I think nobody likes code breaking in new releases, as well as noone
wants to introduce them. So far the incompatibilities weren't all that
great--properly written code usually ran without a hitch on new
releases. And the things that didn't were usually quite easy to spot
(threw exceptions) or gave warnings. One thing that I like about
python (apart from all the well-known things) is that the designers
actually try to fix design flaws. Even if it means some work for me,
I'd rather have it fixed *now* than wait another 2 years, when it'll
be next to impossible due to an even greater code base that yells
"don't change!" And yes, complaining about something often takes more
time than fixing it right away.

Back-to-lurking'ly,
Hans-J.


From greg at cosc.canterbury.ac.nz  Wed Apr 10 20:34:45 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 11 Apr 2002 12:34:45 +1200
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <ACOs8.29598$CU.39610@rwcrnsc51.ops.asp.att.net>
Message-ID: <3CB4DA25.1D828F7@cosc.canterbury.ac.nz>

Emile van Sebille wrote:
> 
> Did-Schr?dinger's-cat-have-a-name?-ly y'rs

Yes. And no. Until somebody asks what it was.

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


From Chris.Barker at noaa.gov  Tue Apr 16 12:51:19 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 16 Apr 2002 09:51:19 -0700
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu> <861ydg3ff3.fsf@haggis.aae.uiuc.edu>
Message-ID: <3CBC5686.BCBCF663@noaa.gov>

"D. Michael McFarland" wrote:
> The bottlenecks now seem to be in
> memory management, which I'm sure Python does better than I would in
> C,

better than me too, I'm sure, but NumPy code often ends up making a lot
more copies of data that similar C code would, so that might very well
get better in C too.

> and numerical linear algebra, which I'm sure Numeric does better
> than I would.

Are you using a ggod native BLAS? that can make a differnce too.

> For now, though, the safest
> route seems to be to throw CPU time at the existing code.

I understand, and I may be able to offer a few cycles. Send me a private email.


> But first this guy needs to
> graduate, for which he needs to crunch a _lot_ of numbers.  He's
> swamped with related work, and I feel like I got him into this...

Where would he be if he had tried to write all his code in Fortran or C??

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From brobbins333 at shaw.ca  Thu Apr 11 21:18:56 2002
From: brobbins333 at shaw.ca (brobbins333 at shaw.ca)
Date: Fri, 12 Apr 2002 01:18:56 GMT
Subject: File Size Mystery??
Message-ID: <3cb6358d.33229187@news>

this is foofile.txt (created in Windows Notepad):
ur75ht8495u7475ht89302oskv476830rj

and here is what Python tells me about it:


>>> print os.path.getsize('foofile.txt')
34
>>> print len('foofile.txt')
11


os.path.getsize() returns the correct size of the file, 
but len() does not. Why is there a discrepancy?


Bruce Robbins
Victoria, BC
Canada




From fredrik at pythonware.com  Wed Apr  3 12:03:24 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 3 Apr 2002 19:03:24 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff> <yu99wuvovkmm.fsf@europa.research.att.com> <00af01c1db2a$e7316d10$ced241d5@hagrid> <200204031619.g33GJXm02577@europa.research.att.com>
Message-ID: <016b01c1db31$8da51010$ced241d5@hagrid>

andrew wrote:
> 
> Fredrik> if you study real python code, you'll find that there is.
> 
> I mean that the language doesn't prefer a single way,
> as opposed to programmers' conventions.

it's more than just a convention -- built-in functions and operators
always return 0 for false and 1 for true whenever they need to in-
vent a boolean value.  this is documented in the language reference.

(and if you look inside, there's something called Py_False that's an
integer zero, and something called Py_True that is an integer one).

this is also true for the standard library, and lots of major extensions.

also note that the PEP says "The bool type would be a straightforward
subtype (in C) of the int type, and the values False and True would
behave like 0 and 1 in most respects".  Guido didn't pick those values
out of thin air...

</F>





From pan-newsreader at thomas-guettler.de  Sat Apr  6 08:10:32 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Sat, 06 Apr 2002 15:10:32 +0200
Subject: RPC for ZODB via https
References: <3CAD9F25.8010708@thomas-guettler.de> <a8kqtu$1d59$1@norfair.nerim.net>
Message-ID: <pan.2002.04.06.15.10.32.283632.3709@thomas-guettler.de>

On Fri, 05 Apr 2002 20:45:50 +0200, Gillou wrote:

> xmlrpclib *is* object oriented, especially when used with Zope
> 
> s = xmlrpclib.Server(http://url_to_a_zope_object) data =
> s.method_of_zope_object(param)

But you can't pass objects as arguments or returns values

The following is not possible:

s=xmlrpclib.Server("http://localhost:8080")
obj=s.getObject()
obj.method()

AFAIK Pyro can do this. But it needs TCP/IP. Would be good
if you could use it with clients which are connected to the internet
via a proxy which allows only http, https.

 thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From pan-newsreader at thomas-guettler.de  Sat Apr  6 01:56:05 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Sat, 06 Apr 2002 08:56:05 +0200
Subject: Better error message for reserved words
Message-ID: <pan.2002.04.06.08.56.05.35544.838@thomas-guettler.de>

Hi!

How difficult is it to change the parser that the following:

==guettli at sonne:~/tmp$ python ~/scripts/replace_recursive.py
  File "/home/guettli/scripts/replace_recursive.py", line 17
    in=open(temp)
     ^
SyntaxError: invalid syntax

will print "SyntaxError: invalid syntax. 'in' is an reserved word"?

I think this would make python more newbie-friendly.

  thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From rob at moofed.com  Wed Apr  3 00:47:50 2002
From: rob at moofed.com (Rob Hazlewood)
Date: Wed, 03 Apr 2002 15:47:50 +1000
Subject: Python compilers
Message-ID: <3CAA9786.2070700@moofed.com>

Hi,
I'm thinking of learning python, and am a little confused as to the 
overall structure of python programming.

Would I be right in saying that Python is an interpreted language?

I know it can be run as an interpreted language, however I seem to 
remember people talking about python compilers.

Do Python compilers exist? are they efficient?

If there is a python compiler, surely it would be able to generate code 
that is as efficient as a C compiler, resulting in what would be the 
best available OO language? (having simplicity, speed)

What is the performance of python like compared to other languages such 
as C and java?

thanks in advance,
Rob



From mickey at tm.informatik.uni-frankfurt.de  Wed Apr 24 19:15:14 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 25 Apr 2002 01:15:14 +0200
Subject: An IMHO interesting optimization question...
Message-ID: <3cc73c81@nntp.server.uni-frankfurt.de>

Hello,

I have an optimiziation question. Given the following code
which generates a word length statistic up to words of the length
<longWord>.

=-------------------------------------------------=
#!/usr/bin/python

def count( word ):
   "Increments the wordcounter"
   length = len( word )
   if length > longWord:
       length = longWord
   stats[length] += 1

def printout():
    "Formatted output of dictionary"
    for i in xrange( 1, longWord ):
        print "Length %d = #%d" % ( i, stats[i] )
    print "Length >%d = #%d" % ( longWord, stats[longWord] )

# import libraries and check for help request

import sys, re
if len( sys.argv ) != 1:
    print """
wordStatistic.py 1.0
--------------------
Reads from stdin and prints a word statistic.
Usage: cat <file> | wordStatistic.py
"""
    sys.exit( 0 )

longWord = 15
defWord = "[A-Za-z]+"

stats = {}
wordExp = re.compile( defWord )

# initialize dictionary, so we don't have to check for key-existance later
for i in xrange( 1, longWord+1 ):
    stats[i] = 0

# main loop. iterates over standard input as long as there's
# something coming in...
while 1:
    line = sys.stdin.readline()
    if line:
        for word in wordExp.findall( line ):
            count( word )
    else:
        break
=-------------------------------------------------=

For a given document, this program takes 20 seconds on my machine.
I now try to optimize it by eliminating the while loop. The 2nd version looks like:

=-------------------------------------------------=
# main loop. iterates over standard input as long as there's
# something coming in...
lines = sys.stdin.readlines()
for line in lines:
    for word in wordExp.findall( line ):
        count( word )
=-------------------------------------------------=

This takes 14 seconds on my machine which seems to be quite an achievement.

Now on to the next. I try to eliminate the for loop by substituting with a map statement,
which is supposed to be faster (because it's looping in C). The 3rd version looks like:

=-------------------------------------------------=
# main loop. iterates over standard input as long as there's
# something coming in...
lines = sys.stdin.readlines()
for line in lines:
    map( count, wordExp.findall( line ) )
=-------------------------------------------------=

Now this takes 17 seconds on my machine. Wow? Why that? Why does it take longer?

On to the final version, which transforms the outer loop into a map statement:

=-------------------------------------------------=
# main loop. iterates over standard input as long as there's
# something coming in...
map( lambda line: map( count, wordExp.findall( line ) ), sys.stdin.readlines() )
=-------------------------------------------------=

This is even worse - it takes 18 seconds on my computer. What am I missing? And do you
know a version which would be even faster than the 2nd version?

Yours,

:M:
--
|----------------------------------------------------------------------------|
| Dipl.-Inf. Michael 'Mickey' Lauer    mickey at tm.informatik.uni-frankfurt.de |
|   Raum 10b - ++49 69 798 28358        Fachbereich Informatik und Biologie  |
|----------------------------------------------------------------------------|


From mwh at python.net  Wed Apr 24 09:46:29 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 24 Apr 2002 13:46:29 GMT
Subject: dictionary method question
References: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com> <3CC5BB74.9070009@alcyone.com> <aa4gpc$7ma$1@bob.news.rcn.net>
Message-ID: <lkvgah1a09.fsf@pc150.maths.bris.ac.uk>

"Raymond Hettinger" <python at rcn.com> writes:

> A resize operation will reorder the keys.  Because resize removes
> dummy entries for previously deleted items, the order may change
> even if the contents of the dictionary remain the same.

But you can't trigger a resize without modifying the dictionary (well,
before 2.2 assigning to a key already in the dictionary could trigger
a resize but that was accidental).

Relavent?  Nope.

Cheers,
M.

-- 
  If you give someone Fortran, he has Fortran.
  If you give someone Lisp, he has any language he pleases.
    -- Guy L. Steele Jr, quoted by David Rush in comp.lang.scheme.scsh


From sth.news at uni.de  Wed Apr 10 13:13:53 2002
From: sth.news at uni.de (Stephan Hohe)
Date: Wed, 10 Apr 2002 19:13:53 +0200
Subject: Size of Newsgrouppostings with nntplib
References: <a8nnlo$tv5d1$1@ID-109762.news.dfncis.de> <mailman.1018132635.8435.python-list@python.org>
Message-ID: <a91rlr$vu766$1@ID-109762.news.dfncis.de>

Gerhard =?unknown-8bit?Q?H=E4ring?= (gerhard at bigfoot.de) schrieb...
> * Stephan Hohe <sth.news at uni.de> [2002-04-06 23:04 +0200]:
> > 
> > I wrote a python skript to generate some newsgroup-statistics. I use
> > the NNTP object in module nntplib to connect to the newsserver and
> > recieve interesting message-headers useing the xhdr method. Now I'm
> > searching for a way to recieve the size of the postings. The only
> > thing I found is the xover method, witch submitts many informations,
> > including the article size in bytes.  Is there a way to recieve the
> > article size without loading all the additional informations submitted
> > by xover?
> 
> I don't know of any way. You could fetch the "Lines" header with the
> XHDR command, but AFAIK you can only access the article size in bytes
> with the help of XOVER.

I found a way: Using XHDR and requesting the header "Bytes" works (at 
least at my newsserver).

Stephan

-- 
Alle, die pauschalisieren, sind bl?d!


From emile at fenx.com  Fri Apr  5 20:45:32 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 06 Apr 2002 01:45:32 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>
Message-ID: <0rsr8.1752$CU.3728@rwcrnsc51.ops.asp.att.net>

"robin and jim" <robinjim at earthlink.net>
> Is there a way to convert a single quoted string to a triple quoted
string?
>
> For example, given:
>
> ''' abc %d xyz '''
>
> the result of ''' abc %d xyz''' % 10 is:
>
> ' abc 10 xyz '
>
> which is a single quoted string.
>
> I would like to convert the result to a triple-quoted string.
>
>

What do you want to do with it?  Once a string is created, it doesn't
matter which style quoting was used -- it's simply a string.



--

Emile van Sebille
emile at fenx.com

---------



From phd at phd.pp.ru  Wed Apr 10 10:59:02 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 10 Apr 2002 18:59:02 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <j4wuvfzlbw.fsf@informatik.hu-berlin.de>; from loewis@informatik.hu-berlin.de on Wed, Apr 10, 2002 at 04:20:51PM +0200
References: <mailman.1018447222.24371.python-list@python.org> <j4wuvfzlbw.fsf@informatik.hu-berlin.de>
Message-ID: <20020410185902.F19899@phd.pp.ru>

On Wed, Apr 10, 2002 at 04:20:51PM +0200, Martin v. L?wis wrote:
> > gcc -c -DNDEBUG -O2 -I. -I./Include -DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
> > In file included from Python/thread_pthread.h:9,
> >                  from Python/thread.c:113:
> > /usr/include/pthread.h:238: parse error before `destructor'
> > /usr/include/pthread.h:238: parse error before `)'
> > make: *** [Python/thread.o] Error 1
> > 
> >    The line in question /usr/include/pthread.h:238 is
> > int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
> 
> I suggest to produce preprocessor output, via -E (not sure whether
> 2.8.1 already had --save-temps).

   Thank you very much for stepping in.

   It revealed that "destructor" was defined
typedef void (*destructor)(PyObject *);
   before.

   Not sure if I should report this as a bug.

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




From mwh at python.net  Wed Apr 17 10:28:44 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 14:28:44 GMT
Subject: Counterintuitive Python behavior
References: <mailman.1019052695.4161.python-list@python.org>
Message-ID: <lksn5uqtsv.fsf@pc150.maths.bris.ac.uk>

<brueckd at tbye.com> writes:

> On Wed, 17 Apr 2002, Michael Hudson wrote:
> 
> > It's you :) As I can't read my email at the moment[1], I have no
> > better way of wasting my time to hand than drawing you some ascii art.
> [snip]
> 
> This is a great explanation! 

Thanks!

> The next time a co-worker asks a
> assignment/binding/references question I'm going to refer them to this
> message.

I was hoping more people than the OP would find it useful.  I'll
probably knock up a version for my web pages (when the starship
reappears...).

Cheers,
M.

-- 
  People think I'm a nice guy, and the fact is that I'm a scheming,
  conniving bastard who doesn't care for any hurt feelings or lost
  hours of work if it just results in what I consider to be a better
  system.                                            -- Linus Torvalds


From robin at jessikat.fsnet.co.uk  Sat Apr 13 09:20:11 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 14:20:11 +0100
Subject: cStringIO write
Message-ID: <QHxVGIALCDu8EwLs@jessikat.fsnet.co.uk>

Is it deliberate that 2.2 cStringIO only has a write method when created uninitialised? StringIO
doesn't have this 'feature', but no mention is made of this in the docs.

>>> import StringIO, cStringIO
>>> y = cStringIO.StringIO('aaa')
>>> dir(y)
['close', 'flush', 'getvalue', 'isatty', 'read', 'readline', 'readlines', 'reset', 'seek',
'tell', 'truncate']
>>> y.write('aaaa')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
AttributeError: write
>>> z = cStringIO.StringIO()
>>> z.write('aaaa')
>>> dir(z)
['close', 'flush', 'getvalue', 'isatty', 'read', 'readline', 'readlines', 'reset', 'seek',
'tell', 'truncate', 'write', 'writelines']
>>> y = StringIO.StringIO('aaa')
>>> dir(y)
['__doc__', '__init__', '__iter__', '__module__', 'buf', 'buflist', 'close', 'closed', 'flush',
'getvalue', 'isatty', 'len', 'pos', 'read', 'readline', 'readlines', 'seek', 'softspace',
'tell', 'truncate', 'write', 'writelines']
>>> y.write('aaaa')
>>> z = cStringIO.StringIO()
>>> z.write('aaaa')
>>> dir(z)
['close', 'flush', 'getvalue', 'isatty', 'read', 'readline', 'readlines', 'reset', 'seek',
'tell', 'truncate', 'write', 'writelines']
>>> 
-- 
Robin Becker


From siegfried.gonzi at kfunigraz.ac.at  Thu Apr 11 05:03:33 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 11 Apr 2002 11:03:33 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <231u8a.lc.ln@127.0.0.1> <a8vnt9$roa$1@panix1.panix.com> <dotv8a.un.ln@127.0.0.1> <a925gf$luk$1@panix1.panix.com> <3CB4D527.E9B76876@engcorp.com>
Message-ID: <3CB55165.89A696A@kfunigraz.ac.at>

Peter Hansen wrote:

> Oh, I'll back you up vociferously.  I learned C before I had
> access to a C compiler or computer that ran other than BASIC
> by reading K&R front to back, back to front, and eleven ways
> from Sunday. 

That is interesting. I had a friend (meanwhile I think he a professional
programmer) who made the same learning curve: learning C without the
assistance of a computer only due to K&R.

He was a very good programmer and did things in C which we only could
complete in IDL.


S. Gonzi


From gimbo at ftech.net  Tue Apr  2 06:54:27 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Tue, 2 Apr 2002 12:54:27 +0100
Subject: The Python Way
In-Reply-To: <15522.13124.505385.871392@beluga.mojam.com>
References: <5515F24370BCB753.DA7CF91464346EF6.E3A577CFA6882D82@lp.airnews.net> <j2m1au4se21mc222877l06690vei117jc3@4ax.com> <a7qqhr$m5$1@panix1.panix.com> <jjb5-B81262.13422827032002@newsstand.cit.cornell.edu> <15522.13124.505385.871392@beluga.mojam.com>
Message-ID: <20020402115427.GC1081@andy.tynant.ftech.net>

On Wed, Mar 27, 2002 at 03:01:56PM -0600, Skip Montanaro wrote:

>     self.cache_lock.acquire()
>     try:
>         fiddle_some_cache...
>     finally:
>         self.cache_lock.release()
> 
> would look a lot cleaner if I simply dumped the threading.RLock objects
> altogether and used Queue.Queue objects holding the cache:
> 
>     some_cache = self.some_cache_queue.get()
>     fiddle_some_cache...
>     self.some_cache_queue.put(some_cache)

I looked at the second code snippet, blinked, said "wow... nice!", but
then wondered if/why it's any better than the first, apart from the
already-mentioned cleanliness.  I _do_ like it, but it smells a bit
"clever", with perhaps a hint of implicity.  ;-)  Why's it better?

> Aahz is always admonishing us to just use Queue.Queue.  Maybe it's time
> I paid a bit closer attention to what he's been saying...

So, is Aahz/Tim saying that Queue.Queue is "better" in general than
threading.RLock ?  If so, why?  I'm honestly curious (though not
curiously honest, alas).

Cheers,

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.



From martin at v.loewis.de  Wed Apr 17 04:25:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 10:25:23 +0200
Subject: Unicode name questions
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com> <009901c1e58b$28a5bd90$445d4540@Dell2> <mailman.1018994314.14044.python-list@python.org>
Message-ID: <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> "Lambda" has been spelled with a "b" as long as I can remember.  

I think the Unicode charts use an ASCII transcription of the native
pronouncation of the letters, if possible. In any case, the Unicode
names for the characters are "official" in the context of Unicode.
The Unicode consortium refrains from renaming them. Between Unicode
2.0 and Unicode 3.2, no character was renamed.

Unicode 1.0 had different names, and \u039B was indeed called GREEK
CAPITAL LETTER LAMBDA in Unicode 1.0.

> I see both "LAMBDA" and "LAMDA" in the comments in the encodings
> modules though.

Apparently, some of the modules are based on tables which use the
Unicode 1.0 names.

> Note that
> 
>     http://www.w3.org/TR/REC-html40/sgml/entities.html
> 
> spells it with a "b".

Correct. However, this is what SGML calls this letter; apparently, the
Unicode consortium found reason to call it differently. I don't know
what the rationale was - most likely, both would be considered
"correct" by a Greek speaker.

>     Brian> And change "GREEK SMALL LETTER THETA SYMBOL" to "GREEK SMALL
>     Brian> LETTER THETA"
> 
> Note that there are both '&theta;' and '&thetasym;' HTML entities.  I found
> it here:
> 
>     http://www.htmlhelp.com/reference/html40/entities/symbols.html
> 
> but also saw it here:
> 
>     http://www.w3.org/TR/REC-html40/sgml/entities.html
> 
> where it is marked "new".  I suspect perhaps Python's codecs just haven't
> caught up yet.

No. Unicode has

0398;GREEK CAPITAL LETTER THETA;Lu;0;L;;;;;N;;;;03B8;
03B8;GREEK SMALL LETTER THETA;Ll;0;L;;;;;N;;;0398;;0398
03D1;GREEK THETA SYMBOL;Ll;0;L;<compat> 03B8;;;;N;GREEK SMALL LETTER SCRIPT THETA;;0398;;0398

The second field is the Unicode name; Python suports all three of
them. The tenth field is the Unicode 1.0 name (if different); the
fifth field indicates that U+03D1 is a compatibility (deprecated)
character for U+3B8.

>From the HTML definition, you'll see that &theta; is &#952; which is
&#x3b8; which is GREEK SMALL LETTER THETA. &thetasym; is &#x3d1; i.e.
GREEK THETA SYMBOL. This was called GREEK SMALL LETTER SCRIPT THETA in
the past, but never GREEK SMALL LETTER THETA SYMBOL. One should
perhaps point out this mistake to the W3C; they ought to use the
Unicode names throughout.

Unicode 3.1 adds

03F4;GREEK CAPITAL THETA SYMBOL;Lu;0;L;<compat> 0398;;;;N;;;;03B8;

as well as various MathML thetas; these are indeed not supported in
Python, yet.

HTH,
Martin


From weeks at vitus.scs.agilent.com  Wed Apr 10 08:22:59 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Wed, 10 Apr 2002 12:22:59 +0000 (UTC)
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <1018450381.103820@cswreg.cos.agilent.com>

Andrew Koenig (ark at research.att.com) wrote:
: My motivation is to be able to write something that is analogous
: to call-by-reference in C++.

*Not* having call-by-reference means that you can read code without
wondering whether a particular function call is or is not by reference.  I
consider this a blessing, and I'd like to think that the developers of
Lisp, C, Java, and Python feel the same way.

Of course, that isn't going to make *you* feel any differently.  I just
wanted to point out the existence of the attitude where your failure to
achieve what you want is considered a success of the language.

Regards (to all to whom I've been grumpy lately),
Greg


From mgilfix at eecs.tufts.edu  Sat Apr 27 14:20:25 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 27 Apr 2002 14:20:25 -0400
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
In-Reply-To: <20020427180727.GB82912@roman.mobil.cz>; from neuhauser@mail.cz on Sat, Apr 27, 2002 at 08:07:27PM +0200
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org> <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com> <20020427171024.GA82912@roman.mobil.cz> <20020427130942.B1588@eecs.tufts.edu> <20020427180727.GB82912@roman.mobil.cz>
Message-ID: <20020427142025.D1588@eecs.tufts.edu>

On Sat, Apr 27 @ 20:07, Roman Neuhauser wrote:
>     I disagree entirely.
> 
>     When I write code @work, I think "if 'x' is equal to 0", because
>     one of my colleagues couldn't wrap his head around the safe way.
>     When I write my own code, I think "if 0 is equal to 'x'". No, wait,
>     I don't know what I think: the thought is abstract, and can be
>     written both ways. But I actually have to remember to put the
>     variable on the left, and the constant on the right side.

  For me, it works the other way around. If I'm programming in an
imperative paradigm, my thought process is sequential. I'm usually at
state A and want to get to state B and in the imperative paradigm,
my program is the list of steps to get from A to B.  Therefore,
logically, it makes sense to think of x == '0' because by deduction,
sometime earlier x must have been declared and assigned to and this is
a pre-step to 0.

  I'm sure most people find it more intuitive to put to the
left at any rate. Hence the reason it's a common bug among
newbies-intermediaries and even careless experts. However, your
thinking that you write " if 'x' is equal to 0" because your
colleagues won't understand only serves to re-enforce my point.

> >   Python is smart to avoid such assignments...
> 
>     Python thinks I'm too dumb to take care of myself.

  Big projects get complicated quickly. Especially if you're working
in an environment where you have many different developers of varying
skills and backgrounds. From a pure software engineering perspective
and even a cost perspective, it's desirable to eliminate all sources
of error - especially if they don't limit the expressiveness of the
language. Otherwise, why not just program in perl? Perl will let you
do *anything* you so desire. But then, why does everyone complain Perl
is completely unmaintainable? I've worked on maintaining a large Perl
code base and yes, it can be done, provided the people who code have
good style and produce clear code. But in a constantly changing
environment (what's the average - people switch jobs every 1 or 2
years in CS), nevermind one of differing skills, I'd rather not
leave that up to chance.

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From logiplexsoftware at earthlink.net  Thu Apr 11 14:49:28 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 11 Apr 2002 11:49:28 -0700
Subject: Help: Python for a big commercial project?
In-Reply-To: <3CB4DBAF.CC7AE350@engcorp.com>
References: <F194tIpqqJgSt51NDa200002509@hotmail.com>
	<mailman.1018464082.15807.python-list@python.org>
	<3CB4DBAF.CC7AE350@engcorp.com>
Message-ID: <20020411114928.3464d19e.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 20:41:19 -0400
Peter Hansen wrote:

> Cliff Wells wrote:
> > 
> > On Wed, 10 Apr 2002 13:24:59 +0200 Cereal Muesli wrote:
> > 
> > > - First of all; do you think that Python is suitable for this
> > >   kind of "serious, big" project?
> > 
> > Absolutely.  I would not hesitate to use Python for *any* project.  At a
prior
> > job, I used Python develop the primary application for a contract worth
over $3
> > million a year.  In my current position, I am moving our applications from
C to
> > Python and these applications will be responsible for monitoring and
> > controlling security systems in some high-risk/top-secret facilities
(overseas
> > military installations, government research facilities, etc - if I told you
> > more, well, you know the line ;)
> 
> Complete, total, and 100% redundant agreement with Cliff on all the issues
> he mentions here.  My company is writing applications from scratch in Python
> to be used to monitor and control communication systems in similar facilities

> (would be fun someday to find that our stuff went into some of the same
places) 
> and for industrial use and if I told you more I'd have to ask Cliff to get
one 
> of the people he knows in his line of business to kill you...
> 
> Hey Cliff, did Bruno call you back yet?  I'm still kicking myself for that
> accidental disclosure on the XP mailing list.  It's going to take so long for
> him to track them all down through their email addresses ...  :-(

Actually Bruno has gone missing after a um, miscalculation at the fish market. 
We've been using Crowbar and Glovebox exclusively for our public relations.  So
I suppose those XP guys are safe for the moment.  Besides, anything you might
have told them was probably refactored to the point of disinformation ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From fperez528 at yahoo.com  Fri Apr 19 14:06:30 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 19 Apr 2002 12:06:30 -0600
Subject: source code cross reference tool
References: <mailman.1019208819.9434.python-list@python.org>
Message-ID: <a9pmcu$srp$1@peabody.colorado.edu>

Tripp Scott wrote:

> does there exist any? or, has anyone used lxr to cross reference
> a python source code tree? it would be cool to have the source
> code in a set of html pages that allows a user to click around
> and jump to function/class definitions, etc.
> 
> t

Well, pydoc -g pretty does some of that. it runs on port 7464, standard http.

cheers,

f


From aleax at aleax.it  Wed Apr 17 03:07:06 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 07:07:06 GMT
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <x7elhfpmvc.fsf@ra.iago.org> <mailman.1018979071.18732.python-list@python.org> <7xofgje97v.fsf@ruckus.brouhaha.com>
Message-ID: <ua9v8.28653$vF6.909653@news2.tin.it>

Paul Rubin wrote:

> Alex Martelli <aleax at aleax.it> writes:
>> Yes, but a GUI thread can't do that either -- nor can a network one;
>> they must be ready for user-input events and netword-data-arriving
>> events too.  The problem is how to seamlessly merge waiting for
>> THOSE kinds of things and for Python-thread-stuff kinds...
> 
> What's really needed is a way to raise events asynchronously in other
> threads.  However this seems technically difficult.  There's some

Yes, and yes.  Problem is the "events" under consideration are disparate:
select sees sockets (files under Unix), each GUI kit has its own ideas.

> discussion of it in one of the sourceforge bug reports; I don't have
> the number handy but the title is something like "add a way to make
> the program completely exit" (i.e. by killing all the threads).

Tx, I'll see if I can find it.


Alex



From gerhard at bigfoot.de  Mon Apr 29 14:46:25 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 29 Apr 2002 18:46:25 GMT
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <Ksfz8.34667$n4.7298632@newsc.telia.net>
Message-ID: <slrnacr583.nil.gerhard@lilith.my-fqdn.de>

Magnus wrote in comp.lang.python:
> BTW: I friend talked about Twisted the other day. I haven't tried it 
> personally but maybe I should?

You can join the #twisted and #python channels at irc.openprojects.net
to get your share of indoctrination ;-)

I haven't yet fully groked twisted, but IMO it's a great toolkit for
writing all kinds of network applications. A lot of protocols are
already implemented for you (http, ftp, telnet, irc, nntp, ...), which
you can use both for clients and servers.

But it isn't primarily a solution for writing *web* applications or a
templating system.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From jon+usenet at unequivocal.co.uk  Tue Apr 16 09:56:47 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 16 Apr 2002 13:56:47 -0000
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org> <3cba62f8@news.mhogaming.com> <slrnablsp5.f08.jon+usenet@snowy.squish.net> <3cbb996d@news.mhogaming.com>
Message-ID: <slrnabobcu.2k3.jon+usenet@snowy.squish.net>

In article <3cbb996d at news.mhogaming.com>, Ken Peek wrote:
>| So you think that:
>|
>|   long(hex(-1l), 16)
>|
>| should not equal -1 but, err, some random choice of 15, 255, 4095,
>| 65535, etc?
> 
> No, since we are "unifying" the int and long type, it is none
> of your business HOW they are stored-- so we do away with the
> 'int()' and 'long()' functions...

So you think that:

  number(hex(-1), 16)

should not equal -1 but, err, some random choice of 15, 255, 4095,
65535, etc?


From debl2nononoSpammywhammy at bellatlantic.net  Tue Apr 23 23:44:06 2002
From: debl2nononoSpammywhammy at bellatlantic.net (David Lees)
Date: Wed, 24 Apr 2002 03:44:06 GMT
Subject: Art and all that Jazz:  are you ugly, I mean, attacking through 
 outer powders
References: <comp.lang.python.d42583a5@agozgi.sc.us> <3CC5E850.3080903@ix.netcom.com>
Message-ID: <3CC62A08.77F6AC93@bellatlantic.net>

I may have missed some of the discussion, but as a naive (with respect
to hipcrime) newsgroup reader I find an explanation at:
http://www.artandallthatjazz.com/spam.shtml which claims that the spam
was actually an an attack on the artist and her business.  Are you
claiming that those pages are a fake or did you miss them?

David Lees


Antaeus Feldspar wrote:
> 
> This Jessica Felix really is a moron, isn't she?
> 
> I mean, she steals newsgroup space for her commercial spam (the fact
> that she's using HipCrime tools to spam *shows* that she knows she's in
> the wrong) and expects that we'll *want* to do business with her after
> she's shown herself to be contemptuous of courtesy and law?
> 
> Jessica Felix wrote:
> 
> > Some durable cups are bizarre and other brave dusts are clean, but will Abu dream that?
> >
> > "ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
> > that incorporates the two passions of its owner, Jessica Felix -
> > art and jazz.
> >
> > Jessica is an artist whose medium is jewelry. She creates what she calls
> > "spirited art," specializing in images from indigenous cultures around the world.
> >
> > http://www.artandallthatjazz.com/index-2.shtml
> >
> > 119A Plaza Street
> > Healdsburg, CA, USA 95448
> > 1-707-433-7900
> >
> > Occasionally, Kenny never attempts until Frank smells the think cloud tamely.
> >
> > They are sowing near sweet, under clean, with sick jugs.
> >
> >
> >
> >


From marklists at mceahern.com  Sat Apr 27 09:37:13 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 27 Apr 2002 08:37:13 -0500
Subject: lint for python?
In-Reply-To: <roy-6B1ABC.08373127042002@news1.panix.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEOPKMAA.marklists@mceahern.com>

[Roy Smith]
> So, my question is, is there such a thing as lint for python?

Have you checked out PyChecker?

Cheers,

// mark




From fperez528 at yahoo.com  Mon Apr 29 20:09:55 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 29 Apr 2002 18:09:55 -0600
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <20020429174425.GN16217@devel.trillke> <mailman.1020115270.9666.python-list@python.org>
Message-ID: <aaknh8$1c4$1@peabody.colorado.edu>

James J. Besemer wrote:

> holger krekel wrote:
> 
>> there are too solutions. choose yourself :-)
> 
> THREE if you include LAMBDA.
> 

No, please heed other advice and use the MA approach. Please note that I'm not 
dissing this post on its own merits, it's a nice discussion, it just is 
ill-adapted to numeric processing. Numeric provides its own _fast_ way to 
manipulate arrays with invalid data, by all means use it. After all, speed is 
one of the big reasons why you use Numeric to begin with (you could do most 
things in it with regular lists, if you had the age of the universe to wait).

cheers,

f


From paiarunk at yahoo.com  Mon Apr  8 04:28:09 2002
From: paiarunk at yahoo.com (Arun KUmar)
Date: 8 Apr 2002 01:28:09 -0700
Subject: Hiding the python source from the user !
Message-ID: <5e430dc.0204080028.2443f026@posting.google.com>

hai,

i am creating an UI application in python, which i need to be given
for demo purposes to my customers,
so how can hide the source of my application?
i don't to show the python code to the customers until he buys the
product.

so how to do this?

bye
arun


From kwokng at earthlink.net  Tue Apr 30 18:49:37 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Tue, 30 Apr 2002 22:49:37 GMT
Subject: telnet
Message-ID: <5cFz8.127$ke3.35@newsread2.prod.itd.earthlink.net>

Hi folks,

I am writing a python app that uses the Telnet object to connect to the
server.  Here is the code,

tn = telnetlib.Telnet(host)
tn.read_until("Please enter your id")
tn.write(userid+"\r\n")

>From here, tn will return either "Please enter your password" or "Wrong id,
please enter your id again".  My question is how can I get the returning
string before I do the next tn.write()

Thanks!

Billy Ng






From pyth at devel.trillke.net  Tue Apr 23 16:46:48 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 23 Apr 2002 22:46:48 +0200
Subject: flatten tuple
In-Reply-To: <m2adrvnxiv.fsf@mother.paradise.lost>
References: <m2adrvnxiv.fsf@mother.paradise.lost>
Message-ID: <20020423204648.GM27395@devel.trillke>

Hello John,

because i have related interests i have taken the
time to write a very small (one screen) and robust 
more general 'SubstDict'. you can use it like this:

  subst1 = SubstDict ( {1: (2, 3), 2: (4, 5, 23), 23: ((42))} )

  subst1[ (1,1,2)] = (4, 5, 42, 3, 4, 5, 42, 3, 4, 5, 42)

Of course you can also use strings but numbers
are shorter for examples :-) See end of mail 
to see how it works for your mail-example.

It does your required (recursive) substitions. It also
detects loops and raises an exception in that
case. 

it requires python 2.2, though. (i recommend
upgrading if possible. it's even more fun than 2.1)

Additionally this module is self-testing
via the nice doctest-feature.

greetings,

	holger

-- snip

#!/usr/bin/env python2.2

from __future__ import generators

class SubstDict(type({})): 
    """ a Dictionary which allows for recursive substution
        of a given key. if key is a tuple and
        no match is found the parts of the tuple 
        are substituted.

        note: this implementation raises a KeyError 
              if a loop is detected.
    """
        
    def subst(self, key, history):
        """ 
        (generative) recursive substitute 
        - yields the finally substituted objects one by one
        - for tuple keys it descends into its parts
        - history is a list of keys visited in one recursion
        """
        if key in history:
            raise KeyError,"loop detected at key %(key)s, history=%(history)s" % locals()

        if self.has_key(key):
            history.append(key)
            for result in self.subst(self.get(key),history):
                yield result

        elif type(key)==type(()):
            for subkey in key: # recurse into iterable key
                for result in self.subst(subkey,history[:]):
                    yield result
        else: 
            yield key

    def __getitem__(self, key):
        """ calls subst to get all substitions for key 
            and returns a full-constructed tuple 
        """
        complete=[]
        for part in self.subst(key,[]):
             complete.append(part)
        return tuple(complete)

# do automatic tests to verify that
# changing the implementation doesn't 
# break anything

if __name__=='__main__':
    
    def _test():
        import doctest, substdict
        return doctest.testmod(substdict,verbose=0)

    _test()


__test__= { 'SubstDict': """
>>> key1 = ( 'John', 'Hunter', 'where' )
>>> dict1 = SubstDict ({
...        'where': ('was', 'here'),
...        'somekey': 'someval'
...        })
>>> dict1[key1]
('John', 'Hunter', 'was', 'here')
>>> dict2 = SubstDict ({
...     'where': ('was', 'somekey'),
...     'somekey': 'someval',
...     'someval': ('asking', 'really', 'silly', 'questions'),
...     'asking': (((('not asking'))))
...     })
>>> dict2[key1]
('John', 'Hunter', 'was', 'not asking', 'really', 'silly', 'questions')
>>> loopingdict = {
...     'where' : ('am','i'),
...     'i'     : ('where','what')
... }
>>> try: loopingdict[key1]
... except KeyError,e:
...      print "ok, got keyerror for loops in SubstDict"
...
ok, got keyerror for loops in SubstDict
>>> dict1[('john','john','where')]
('john', 'john', 'was', 'here')
>>>
"""}

-- snip






From phr-n2002a at nightsong.com  Wed Apr 24 04:41:34 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 01:41:34 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com>
Message-ID: <7xpu0pij41.fsf@ruckus.brouhaha.com>

Richard Parker <richard at electrophobia.com> writes:
> I just took a quick look at your Python code.  I'd encourage you to not use
> the secret prefix method to construct a MAC from a hash function, i.e.
> MAC(x) = H(K || x).  This method is generally considered to be insecure.
> Use the HMAC construction instead.

Thanks.  What's the real deal with this?  Is an insecurity known, when
K and X are the output of encryption functions, and not chosen by
attackers?  I know that HMAC tries to be robust under unknown weakness
of the underlying hash function, but since the OFB mode encryption
depends on the hash function being strong, HMAC likely doens't help
that much.  Do you think it matters in practice, given that this is
running in an interpreted language on a general purpose PC?  Practical
attacks probably involve computer viruses peeking at memory rather
than cryptanalysing SHA-1.  Also, I currently chop the MAC to 64 bits
(maybe I should make it 80 bits or so) to cut down on plaintext
expansion.  I'd rather avoid HMAC because implementing HMAC in Python
will make the function even slower than it already is.  

> The ciphertext and key passed to the decryption function could be used as
> additional sources of entropy for your "_state" variable.

I guess that's not likely to hurt, though it would probably be
deterministic in many applications.

> Perhaps you could add a function to your API that would allow the
> programmer using your API to provide additional an source of
> entropy.  This would enable the hypothetical programmer of the CGI
> using your library to fix the problem by using information from the
> HTTP requests as a source of entropy.

I guess I could do that, though what I really want is a CPRNG with
good entropy, and a web server is likely to have that.  Asking the
programmer to supply entropy is against the spirit of this supposedly
easy-to-use module.

This module is a stopgap anyway--once there's an AES module, this
thing won't be needed any more.  But I've been using it for a while
(in slightly different form) so I thought I'd release it.


From hgs at dmu.ac.uk  Thu Apr 18 07:49:06 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 12:49:06 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <a9m5rn$4b8ru$2@ID-99513.news.dfncis.de>
Message-ID: <Pine.GSO.4.40.0204181239010.24683-100000@neelix>

On Wed, 17 Apr 2002, P. Alejandro Lopez-Valencia wrote:

>
> "Hugh Sasse Staff Elec Eng" <hgs at dmu.ac.uk> escribi? en el mensaje
> news:mailman.1019036193.29349.python-list at python.org...
> On Tue, 16 Apr 2002, P. Alejandro Lopez-Valencia wrote:
>
>
> On 17 Apr 2002, Martin v. Loewis wrote:
>
> > "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
        [...]
> > Can you elaborate? gcc -shared works fine on Solaris, and binutils
> > also works without problems.
>
> >AFAIK this is true.  I've not had problems like this with other
> >packages.
>
> As said in other message, this is not the recommended procedure; the
> fact
> that you can does not mean you should. GNU as does not generate 100%
> compatible binary object code, and GNU ld tends to botch Sparc object

I don't know enough about the internals to assert anything about this
myself, so I tried a build with CC=cc and CXX=CC because we fortunately
have the Forte6 compiler set available.

Script started on Thu Apr 18 11:53:08 2002
        [...]
160 tests OK.
4 tests failed:
    test_largefile test_socket test_sundry test_urllib2
23 tests skipped:
    test___all__ test_al test_asynchat test_bsddb test_cd test_cl
    test_curses test_dl test_gdbm test_gl test_imgfile
    test_linuxaudiodev test_minidom test_nis test_ntpath test_openpty
    test_pyexpat test_sax test_socket_ssl test_socketserver
    test_sunaudiodev test_winreg test_winsound
Ask someone to teach regrtest.py about which tests are
expected to get skipped on sunos5.
*** Error code 1
make: Fatal error: Command failed for target `test'
neelix hgs 13 %> exit
neelix hgs 14 %>
script done on Thu Apr 18 12:33:51 2002

How concerned should I be about these?

        [other interesting matters trimmed, see my next reply]
>
> Alejo
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Hugh





From tdelaney at avaya.com  Sun Apr 28 20:48:21 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 10:48:21 +1000
Subject: iis -- IMAP to IMAP Synchronization [Was: Re: comparsion of t
	wo lists]
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A0@natasha.auslabs.avaya.com>

> From: Matej Cepl [mailto:cepl.m at neu.edu]
> 
> On Sat, Apr 27, 2002 at 12:17:01AM +0000, Van Gale wrote:
> > "Matej Cepl" wrote:
> > > I would like to compare two
> > > list of messages (in the local and remote folders) and create
> > > list of messages, to be deleted/downloaded. In plain Python, the
> > > part of deletion would go like this:
> 
> OK, I have created ALPHA version of my script!! Yeah (my first
> big script in Python :-)!!!
> 
> However, because the script should work with my real emails, I am 
> rather concerned, whether I have not made some stupid mistake. 
> Would you be so kind and review the attached script and comment 
> on it (especially safety of my emails and speed), please?

To test this type of thing, you should set up an IMAP server (or 2) on one
of your own boxes. Then you can create fake emails to your heart's content.
Never use real, live data for testing - there will *always* be something you
have missed which will delete everything you have.

Tim Delaney




From nas at python.ca  Tue Apr 30 22:21:36 2002
From: nas at python.ca (Neil Schemenauer)
Date: Tue, 30 Apr 2002 19:21:36 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <LeAz8.606$Au5.499@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Tue, Apr 30, 2002 at 01:08:27PM -0400
References: <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net> <mailman.1020176703.26327.python-list@python.org> <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com> <mailman.1020182455.6633.python-list@python.org> <LeAz8.606$Au5.499@atlpnn01.usenetserver.com>
Message-ID: <20020430192136.A25838@glacier.arctrix.com>

Steve Holden wrote:
> has-anyone-seen-the-timbot-around-ly y'rs  - steve

Nope and I've never seen anyone from the PSU eith




From anandpillai6 at yahoo.com  Thu Apr  4 01:51:42 2002
From: anandpillai6 at yahoo.com (Anand)
Date: 3 Apr 2002 22:51:42 -0800
Subject: Firewall settings in httplib
Message-ID: <41c203e.0204032251.76bc3c8a@posting.google.com>

How do I specify a proxy server/firewall settings in httplib .

Thanks

Anand Pillai


From francis.meyvis at sonycom.com  Fri Apr 12 06:00:40 2002
From: francis.meyvis at sonycom.com (Francis Meyvis)
Date: Fri, 12 Apr 2002 10:00:40 GMT
Subject: socket.connect problem on w2k with python 2.1.1
Message-ID: <cfyt8.42$2L2.1093@nreader2.kpnqwest.net>

Hello,

I've a socket "connect" problem on w2k SP2 with python 2.1.1
I want to connect to a program running on my local PC,
that listens as the socket server.
It works from plain cpp using socket(), connect(), recv and send()
Do you have any suggestions on what I do wrong?

import socket
a = (socket.gethostname(), 18002)
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect(a)

Returns as output:
socket.error: (10061, 'Connection refused')

In the cpp program I do basically this (code snippet):

    aiSock = socket(AF_INET, SOCK_STREAM, 0);
    if ( BAD_SOCK == aiSock ) {
        std::cerr << "socket failed: " << ERRNO_SOCKET << std::endl;
        Fail(1);
    }

    vptHost = gethostbyname(arOpt.moAddr.c_str());
    if ( !vptHost ) {
        std::cerr << "gethostbyname failed: " << ERRNO_SOCKET << std::endl;
        Fail(3);
    }

    struct sockaddr_in vtAddr;
    vtAddr.sin_family = AF_INET;
    memcpy(&vtAddr.sin_addr.s_addr, vptHost->h_addr_list[0],
sizeof(vtAddr.sin_addr.s_addr));
    vtAddr.sin_port = arOpt.miPort;

    if ( connect(aiSock, (struct sockaddr*)&vtAddr, sizeof(vtAddr)) ) {
        std::cerr << "connect failed: " << ERRNO_SOCKET << std::endl;
        Fail(2);
    }

Thanks!
--
---
Kind regards,
francis meyvis




From abagaric at rest-art.hr  Wed Apr 17 03:45:43 2002
From: abagaric at rest-art.hr (Ante)
Date: Wed, 17 Apr 2002 09:45:43 +0200
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca> <a9hjvc$cdq$1@as201.hinet.hr> <mailman.1018979135.19771.python-list@python.org>
Message-ID: <a9j9cc$4bj4$1@as201.hinet.hr>

"Paul Osman" <paul.osman at sympatico.ca> wrote in message
news:mailman.1018979135.19771.python-list at python.org...
> Okay, let me try this again... Python detects local names statically.
> Variables assigned in a function are classified as local by default.
> therefore in the following script:
>
> This is because there are actually two variables named "num", one
> global, and one local to the function dummy(). Now, if we were to do
> take out the assignment from the dummy() function, the output would be:

I am new to Python, but I'm not quite new to programming in general...
I see what you are saying, no need to use the syntax as you would with a
total programming newbie. :) I mean, I know what global and local means,
and I know 'global num' would solve the problem. But that was not the issue.
Let me try another example:

dummy.py:
--------------------------------
num = [1]

def dummy():
    num.append(2)
    print num

dummy()
------------------------------


>>> import dummy
[1, 2]

See? I was changing the list, and yet it didn't produce an error.
But, if I used
    num += [2]
instead of
    num.append(2)
it would again produce an exception.

Now. don't those two expressions mean the same? I am not changing
the reference, (as I would if I used plain =) but I am rather modifying the
list itself, without moving the reference. Here is the proof:

>>> li = [1,2,3]
>>> li2 = li
>>> li += [2]
>>> li is li2
1

So, the li reference hasn't moved after I used += operator. Wouldn't be true
if I used operator =.

So, the question was: while it made some sense that if I used the operator =
the compiler would assume it is a new, local name, it doesn't make as much
sense to assume the same thing if I used +=. Because, in order to
'increment'
something, I already need its initial value. But, if you still think any
changes to
a name should make the compiler assume it's local, then num.append() should
assume it just as well. Am I not right?


Ante





From kybic at ieee.org  Thu Apr  4 02:54:44 2002
From: kybic at ieee.org (Jan Kybic)
Date: 04 Apr 2002 09:54:44 +0200
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop>
References: <mailman.1017795845.3464.python-list@python.org>
	<r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
	<lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
	<lbh8z84nake.fsf@xihu.inria.fr>
	<15531.7864.698585.485436@12-248-41-177.client.attbi.com>
	<lbh7knolreu.fsf@xihu.inria.fr>
	<15531.9709.924953.953487@12-248-41-177.client.attbi.com>
	<mailman.1017850275.2134.python-list@python.org>
	<0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop>
Message-ID: <lbh8z83kiez.fsf@xihu.inria.fr>

> For this particular example, weave.inline is a perfect solution.  The C++ code
> is pretty readable:
> 
> from scipy import *
> 
> def foo(a):
>     code = """
>            for (int i = 0; i < _Na(0);i++) {
>                if (a(i,i) > 30) a(i,i) += a(i,i+1);
>                else             a(i,0) *= 2.0; }
>            """
>     weave.inline(code,'a',compiler='gcc',
>                  type_converters = weave.converters.blitz)
> 
> a = rand(3,3)*60
> print 'before:', a
> foo(a)
> print 'after:', a
> 

Great, I am becoming even more impressed. However, the main criticism
remains: You have to code your function twice, first in Python for
prototyping, then in C. But I agree that weave saves you a lot of
work.


> The place where weave breaks down is when you have a lot of calls within the
> loop to Python functions.  This is a common problem, and the only way around it
> is to have something that compiles the Python to C or assembly or something.

Indeed.

> That is what Psycho does.  weave may also do something similar in the forseeable
> future.  Pat Miller is integrating some work similar to Psycho into weave that
> converts Python to C on the fly using type inferencing and then uses weave to
> build the extension module.  Several of the aims are different than psycho, but
> the concepts are similar.

Great. I will stay tuned.

J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/




From NO_SPAM_jmpurser2 at attbi.com  Fri Apr 12 06:24:52 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Fri, 12 Apr 2002 10:24:52 GMT
Subject: Python and databases
Message-ID: <UByt8.1138$4D6.194@rwcrnsc52.ops.asp.att.net>

I'm working on my first real project with python and it's a client
server/database program.  I'm looking  for a good tutorial on python and
databases.  Anybody got a good link they'd care to share?

Thanks,

John Purser




From bww00 at amdahl.com  Mon Apr  8 11:52:29 2002
From: bww00 at amdahl.com (Bryan Webb)
Date: 08 Apr 2002 15:52:29 GMT
Subject: postmans sort in python
Message-ID: <a8sebt$im4@dispatch.concentric.net>

Hi,
    Has anyone done a sort in python based on the "postmans sort" at the
following link ??
http://www.rrsd.com/

Thanks
bryan Webb




From starx at pacbell.net  Sat Apr 13 22:24:24 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 14 Apr 2002 02:24:24 GMT
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com>
Message-ID: <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4>

Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:

> Philip Swartzleonard <starx at pacbell.net> wrote
> 
>> The first section of the index[1] for the language reference manual
>> lists all of these things (the underscore section), with links.
> 
> Thank you very much for pointers, but I can think of something not
> listed there. For example, __all__ for "from module import *", pickle
> protocols, __new__ for subclassing from a type and so on. Is there a
> complete (or near-complete) list?
> 

Well, pickle protocols are defined by the pickle module, right? Thus are 
not part of the language, but just __foo__ names that the module calls on 
your object. For this kind of special name, I'd think you would _have_ to 
look in the documentation for the module that provides the functionality, 
and it would not be included on a language __x__ scorecard anyway.

For the __new__, I think the problem there is that the newer type/class 
unification structures and keywords haven't all trickled down into the 
reference documentation: if you want to learn about list comprehensions 
and object-based types, you currently have to go to the 'what's new in py 
2.2' documentation. These should be added to that index list in due time.

(Never heard of __all__ though, so... :\ )

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From vmilitaru at sympatico.ca  Thu Apr  4 04:16:08 2002
From: vmilitaru at sympatico.ca (vio)
Date: Thu, 04 Apr 2002 04:16:08 -0500
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org>
Message-ID: <3CAC19D8.142A918@sympatico.ca>

Thank you all for your suggestions and for pointing out obvious obstacles. Yes,
100MB of java source sounds impressive, and is.  But there seem to be some
alternatives to  re-writing that project in python from scrap, to not turn all
that java code into total waste. Jython and Antlr, P-lex, Lex-Yacc, to name a
few suggestions, each promises to bring a little closer to a possible, though
imperfect, solution. I am very enthousiastic about all your suggestions, and
would like to give a wack at each, see what flies and what bombs.

Many Thanks,
Vio


From aahz at pythoncraft.com  Mon Apr  1 09:09:48 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 09:09:48 -0500
Subject: The fastest XML parser around
References: <mailman.1017654423.12312.clpa-moderators@python.org>
Message-ID: <a89pnc$7j9$1@panix1.panix.com>

In article <mailman.1017654423.12312.clpa-moderators at python.org>,
Andy Robinson <andy at reportlab.com> wrote:
>
>ReportLab (www.reportlab.com) are proud to announce the release of
>pyRXP, the fastest XML parsing toolkit for python, and possibly for any
>other language anywhere:
>
>                     http://www.reportlab.com/xml/pyrxp.html

Do you have information available on using pyRXP with PythonPoint?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From daniel.dittmar at sap.com  Fri Apr 26 04:44:31 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 26 Apr 2002 10:44:31 +0200
Subject: ftplib and attributes
References: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>
Message-ID: <aab41f$5hk$1@news1.wdf.sap-ag.de>

LoGan wrote:
> hi, im new bie and im trying to use ftplib, but my problem is to get
> the attributes of the files, how can i do ?

Although you have to use .dir () as the other posters have written, it is
important to know that different ftp servers will format the information in
different ways.

One very annoying feature of many UNIX based ftp servers is the following:
drwxrwsr-x   5 python   www          8192 Mar  1 16:54 ..
drwxrwsr-x   2 python   www          8192 Aug  7  2001 art

Note how a recent file/directory has a time, where as an older file will
have a year instead.

Daniel





From aleax at aleax.it  Thu Apr 11 06:25:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 10:25:42 GMT
Subject: initializing list with number of elements given by a variable
References: <3CB555E6.3080207@web.de>
Message-ID: <Gwdt8.5932$b62.144118@news1.tin.it>

Oliver Pust wrote:

> Hi,
> 
> I need to create a list which number of elements is given by an integer
> variable.
> 
> So something like this:
> 
>  >>> length_of_list = 6
>  >>> list = ????????? # what to place here, is the question

First, don't call it 'list': that's the built-in name of the list
type and it's best not to hide it with your own variable.  It is
much wiser to call it alist, somelist, whatever, NOT 'list'.

Apart from that, what you need is:

alist = length_of_list * [0]


Alex



From pete at dexterslab.de  Tue Apr  9 10:39:31 2002
From: pete at dexterslab.de (Henning Peters)
Date: Tue, 9 Apr 2002 16:39:31 +0200
Subject: exception question
References: <a8ua60$7nd$06$1@news.t-online.com> <mailman.1018344320.21056.python-list@python.org>
Message-ID: <a8uu9v$fkm$04$1@news.t-online.com>

> except:
>    e1, e2, e3 = sys.exc_info()

thanks a lot, that works...

bye,
henning




From mjbarber at ascc.artsci.wustl.edu  Mon Apr  1 08:17:20 2002
From: mjbarber at ascc.artsci.wustl.edu (Michael James Barber)
Date: 1 Apr 2002 07:17:20 -0600
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <a89ml0$4c1@ascc.artsci.wustl.edu>

True generally acts like 1 and False acts like 0.

So the following all are true (oh, this is going to be hard to talk 
about!):
True == 1
False == 0
True != 2


We can use True and False arithmetically.  I do realize that 'and' and 
'or' are the more usual spellings, but let's see how things look.

So:
True + False == True # seems reasonable
True * False == False # also nice
True - True == False # OK
True + True != True # conceptually strange

Similarly, we can test of what class True and False are instances.

So, we have true statements:
isinstance(True, bool) # or whatever spelling is preferred for bool
isinstance(False, bool)
not isinstance(True + True, bool)
not isinstance(True - True, bool)
not isinstance(True + False, bool)
not isinstance(True - False, bool)
not isinstance(True * True, bool)
not isinstance(True * False, bool)

I'm opposed.  Conceptually, it is identical to working with 1 and 0, 
except that sometimes we'll see an alternate spelling and may have to jump 
through some hoops to understand what's going on.  I don't really see what 
is gained by this, and there is increased conceptual overhead with reduced 
clarity, IMO.

A version of truth types that handled arithmetic operators more smoothly 
would better suit my tastes.  I think this version pays too big of a cost 
in clarity to maintain backwards compatibility.
-- 
					Michael J. Barber


From guido at python.org  Tue Apr  2 20:53:56 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 01:53:56 GMT
Subject: __name__ on new-style classes
References: <stcfaugkbh0u65d1euuofhs2b69ej7i34h@4ax.com> <3CA7C873.D1AC908A@python.org> <jvogausg16i1vgp93qkbb8f9sb3nicgb7u@4ax.com>
Message-ID: <3CAA619C.A38F2DC9@python.org>

"Gon?alo Rodrigues" wrote:
> 
> On Mon, 01 Apr 2002 02:35:47 GMT, Guido van Rossum <guido at python.org>
> wrote:
> 
> >It's a feature.  The reason for the change is that I've never seen an
> >example of a use for it (as opposed to changing an instance's __class__).
> >If you have one, let us know.
> 
> I bumped into this "feature" a few days ago trying to understand
> metaclasses in Python. And to understand them there is nothing better
> than actually coding one, so I'll explain my problem/solution and let
> you judge if this is a good use case for having a writable __name__.
> 
> My problem is modelling finite arithmetic in Python. Here by finite
> arithmetic I mean all the quotients of the integer ring Z. There is a
> countable number of them, one for each n > 1 (n=0 is Z and n=1 is the
> trivial ring {0}. If n is negative then it is the same as -n). If n is
> prime then we actually have a field.
> 
> The obvious way (to me, that is) to do this is to let each finite ring
> Z_n be a class, inheriting from int. Since there are an infinite number
> of these we have to have a metaclass, call it Field, that has as
> intances the Z_n rings, e.g. Z_n would be Field(n). The skeleton of this
> metaclass would be something like:
> 
> class Field(type):
>     """The Finite Field metaclass."""
> 
>     def __new__(cls, n):
>         if isinstance(n, int) and n > 1:
>             #The template class.
>             class Z(int):
>                 pass
> 
>             #Generate proper name for the instance class, e.g. "Z_n"
>             Z.__name__ = "Z_%d" % n
> 
>             #Return class.
>             return Z
>         else:
>             raise TypeError, "%s is not an integer > 1." % str(n)
> 
> If I cant write to __name__ then all the instances will have the same
> name... Not A Good Thing, IMHO.

But you're doing the meta-instantiation all wrong.  Rather than using a
class
statement, you should be using cls.__new__(cls, "Z_%d"%n, (int,), {...}).

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


From rjones at ekit-inc.com  Tue Apr 30 19:55:40 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 1 May 2002 09:55:40 +1000
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <20020430174422.D19566@phd.pp.ru>
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <NOwz8.13638$xy4.7951@atlpnn <20020430174422.D19566@phd.pp.ru>
Message-ID: <200205010955.40536.rjones@ekit-inc.com>

On Tue, 30 Apr 2002 23:44, Oleg Broytmann wrote:
> On Tue, Apr 30, 2002 at 09:13:46AM -0400, Steve Holden wrote:
> > Gadfly is great for people who don't have an existing
> > relational system but want to do some Python experimentation with
> > database.
>
>    Other alternatives are BerkeleyDB with PyBSDDB wrapper

Not really a relational system... although if roundup's hyperdb bsddb3 backed 
was working, it'd come close ;)


> and MetaKit with MkSql.

... and MetaKit isn't a pure-python relational system. Note that there's two 
python bindings for MetaKit:

   http://www.mcmillan-inc.com/mksqlintro.html
   http://www.equi4.com/metakit/python.html

so I guess it's popular ;)

Gadfly's got the really neat feature of being able to create a Python class 
that has the same interface as the gadfly SQL tables - and instances of that 
class are then useable in any SQL query you want to throw at gadfly. I know 
of at least one user who wrapped an snmp system in gadfly table classes and 
that relational queries between the snmp data and regular table data. That's 
pretty darned neat :)


    Richard





From jacob at boris.cd.chalmers.se.cd.chalmers.se  Mon Apr  8 14:08:22 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 8 Apr 2002 18:08:22 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1018124969.17602.python-list@python.org>
Message-ID: <a8smam$png$1@nyheter.chalmers.se>

In article <mailman.1018124969.17602.python-list at python.org>,
Ken Seehof <kseehof at neuralintegrator.com> wrote:
>> If we ignore the imaginary, which has a very specialized scope, we have 5
>> base types. These can interact in approximately 5! (120)
>> different ways in a
>> statement. If we add one more base type we get 6! (720) different ways
>> that our types can interact.
>
>In python the number of numeric interactions is essentially 3!, not 6!.
>Most of these interactions are trivial and predictable, and are not
>increased
>by bool.  My brain hurts trying to figure out how to enumerate the 6! that
>you suggest :-)  Perhaps in other languages your argument has merit.
>
>One thing that I agree with is that one would not be likely to add a bool to
>a complex number.  :-)

The difficulty does not come from the legitimate combinations. It comes from
the erroneous ones. Understanding what a correct program does is trivial,
compared to understanding why an incorrect one does not work.
The possible number of ways of combining X types is the factorial of X.

Each wrong combination requires the understanding of the error message and
the reason why it is wrong. A seasoned programmer can classify the errors
and handle a fair number of these classes. Newbies have a seriously
difficult time. I would even go as far as to say that it is the most
important factor controlling how much time a novice programmer needs
in order to be comfortable with the language.

Jacob Hall?n

-- 


From sholden at holdenweb.com  Mon Apr 29 08:45:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 08:45:59 -0400
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org>
Message-ID: <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com>

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1020029269.29740.python-list at python.org...
> On Sun, Apr 28, 2002 at 09:06:05PM +0000, Magnus Lie Hetland wrote:
> > In article <PNFy8.28407$8D3.758650 at news1.tin.it>, Alex Martelli wrote:
> > >holger krekel wrote:
> > >        ...
> > >> ok. i found a nice workaround (if one dare call it that:-)
> > >>
> > >> class dictlookup(type({})):
> >
> > Why use type({}) when you can use dict?
>
[ ... good reasons ...]
>
> 3. And 'dict' is something i often use in local definitions
>    when working with dictionaries.
>
Yeah, well we all have to cut that out and start using "adict" or "thisdict"
pretty sharpish. This one bites now.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From noah at noah.org  Thu Apr 11 05:03:47 2002
From: noah at noah.org (Noah)
Date: Thu, 11 Apr 2002 02:03:47 -0700
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
Message-ID: <PAENINBOOODFDEPOKPKCMEMACKAA.noah@noah.org>

I'm having headaches with nonblocking IO on a pipe.
I opened a UNIX command through os.popen4.
Then I set O_NONBLOCK on the output stream of the pipe.
Finally I did a select() on the output pipe to wait for data.
If select returns I would try to read the data.
The problem is that select() is blocking. Select should block 
until some data is available on the pipe, so I assume that
select is never even seeing any data on the pipe.
I added a timeout to select() and then tried to read the
pipe, but it was empty (read() returned '').

One program that I am attempting to talk to via a pipe is ftp.
Any interactive program I open will block. 
If I try open a program that terminates like "ls" or "uname -a" 
then my script works as expected.

What am I doing wrong that is causing interactive programs to block?
Any help?  I'm using Python 2.1 (#1, Jul  2 2001, 23:39:12).
The following sample script runs, but blocks on my system.

#!/usr/bin/env python
import fcntl, FCNTL
import select
import os
# First I open FTP using os.popen4()
(child_stdin, child_stdout_and_stderr) = os.popen4 ("/usr/bin/ftp localhost")
# This is how I'm setting nonblocking IO on child_stdout_and_stderr
original_flags = fcntl.fcntl (child_stdout_and_stderr.fileno(), FCNTL.F_GETFL, 0)
flags = original_flags | FCNTL.O_NONBLOCK
fcntl.fcntl(child_stdout_and_stderr.fileno(), FCNTL.F_SETFL, flags)
# Then I select() on child_stdout_and_stderr to wait for data
(r,w,e) = select.select([child_stdout_and_stderr], [], [], None)
if len(r) > 0:
        print r[0].read()



Yours,
Noah




From sabren at manifestation.com  Wed Apr  3 16:31:46 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Wed, 3 Apr 2002 16:31:46 -0500 (EST)
Subject: Java to Python translation
In-Reply-To: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <Pine.LNX.4.44.0204031626020.13277-100000@mercury.sabren.com>

On Wed, 3 Apr 2002, vio wrote:

> I need to translate a (rather voluminous 100MB) java
> program to python. After a few searches on google,
> apparently the best tools for this task are Lex-Yacc
> (Flex-Bison on linux). While I am working on the Lex-YACC
> HOWTO, I would be very interested in hearing of others'
> successful experiences/feedback at this. Any pointers of
> interest would similarily be appreciated.

Hi Vio,

I agree with what everyone else is saying about Jython.
That's definitely the place to start.. At least that
way you can change only certain classes at a time, and
gradually move towards python. (And then switch to plain
python if you want to after the transition is complete)

But... I would aslo suggest that if you do decide to
automate part of the transition, you might want to take a
look at ANTLR...

   http://www.antlr.org/ 

It's a java-based tool, with a somewhat different
theoretical background than lex/yacc... And it actually
comes with a grammar for the java language built in. I'm not
a parsing pro by any means, but even back when I was a
complete beginner to java, I was able to figure out ANTLR
and make my own little interpreter (for a prolog-like
language, not java).

Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Give your ideas the perfect home: http://www.cornerhost.com/
 cvs - weblogs - php - linux shell - perl/python/cgi - java
------------------------------------------------------------





From sholden at holdenweb.com  Mon Apr  1 17:34:22 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 17:34:22 -0500
Subject: web - recommended authentication?
References: <ayvp8.3121$hi7.12423@news-server.bigpond.net.au>
Message-ID: <Ng5q8.118312$Ou.94876@atlpnn01.usenetserver.com>

"Trung Hoang" <thoa0025 at mail.usyd.edu.au> wrote in message
news:ayvp8.3121$hi7.12423 at news-server.bigpond.net.au...
> Greetings,
>
> i wish to create authentication for a section of my website. what is a
good
> practiced way to do this?
>
> via cookies?
> via .htaccess? (but i do not wish to use directories is it possible to
> restrict a script & query string? suppose script.cgi?request=specialarea
via
> .htaccess??)
>
> what are your thoughts?

This topic has been fairly extensively discussed in the last couple of
months. You migfht want to take a look in the newsgroupd archives on Google.

regards
 Steve






From herrn at gmx.net  Thu Apr  4 06:29:03 2002
From: herrn at gmx.net (Marco Herrn)
Date: 4 Apr 2002 11:29:03 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com>
Message-ID: <1017919742.269270@DS9.ai-lab.fh-furtwangen.de>

A. Jones schrieb im Artikel <3cabf705.91920985 at mammoth.usenet-access.com>:
> 
> I'll probably get blasted for this by Python gurus... but, I've used
> it with some success.  You can do
> 
> exec(myObjc + " = " + value)
> 
> or, if it's a class instance you want to create
> 
> exec(myObjc + " = " + bar + "()").

Thank you. That is possibly the best solution for me.  The other answers
I got (thank you all for that) seem to assume that I know possible names
of class instances. But that is not what I want. (Maybe I am just
misunderstanding the code ;-)

Bye
Marco


-- 
Marco Herrn             herrn at gmx.net
(GnuPG/PGP-signed and crypted mail preferred)
Key ID: 0x94620736



From tim.one at comcast.net  Sat Apr 13 13:49:00 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 13:49:00 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
In-Reply-To: <lk7knbiyw4.fsf@pc150.maths.bris.ac.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEOBONAA.tim.one@comcast.net>

[Philip Swartzleonard]
> ... any name that happens to gain the status of being that integer
> really just gets another copy of it's pointer.

[Michael Hudson]
> Yep.  This happens for the empty tuple, the empty string and one
> character strings too (and maybe some other I forget now).

The implementation can choose, or choose not, to share memory blobs for any
value of an immutable type, and exactly when & which objects get shared can
(and does) vary across releases.  The only objects guaranteed to be
singletons (meaning you'll always get exactly the same object no matter how
its value is computed) are None, type objects (things the type() function
returns), and in 2.3 True and False.  So, in particular, "is" should never
be used as a test for integer equality, or equality with an empty tuple or
string.

About a year ago, aysncore.py did have a line of the form

     if s is "":

and of course it eventually failed in real life.  Most ways of computing
strings *just happen* today to return "the same" empty string when an empty
string is the result, but not all do.

>>> id("")
6602808
>>> id("abc"[1:1])
6602808
>>> id("" + "")
6602808
>>> id("abc" * 0)   # oops
6602856
>>>

That's not a bug.  It wouldn't be a bug if all of them returned different
objects.  Indeed, it wouldn't be a bug even if two distinct instances of the
literal "" returned different objects.





From niemeyer at conectiva.com  Mon Apr  8 19:18:29 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Mon, 8 Apr 2002 20:18:29 -0300
Subject: ACCEPTED: PEP 285
In-Reply-To: <Xns91E9BCA9CF9ACJamesLLugojcom@216.148.53.85>
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83> <3CADF01A.7F0974B2@python.org> <Xns91E7CD754DB67JamesLLugojcom@216.148.53.84> <m4isn6712s9.fsf@macquarie.com.au> <Xns91E9BCA9CF9ACJamesLLugojcom@216.148.53.85>
Message-ID: <20020408201829.A846@ibook.distro.conectiva>

[...]
> of solutions. I'll reevaluate scripting languages at some point in the 
> future with more of an eye this time around on what might be called 
> "infrastructure" issues as well as the languages.

James, I'm curious. Any of your python programs stopped working
recently!? Why!?

I ask that because I'm the maintainer of thousands of lines of python
code, and I'm also responsible for keeping Python up-to-date in a widely
used distribution. It means I have to update Python often, and also check
and fix anything that breaks.  I really can't find those problems you're
so worried about.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From just at xs4all.nl  Sun Apr  7 04:17:08 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sun, 07 Apr 2002 10:17:08 +0200
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com>
Message-ID: <just-16031B.10170807042002@news1.xs4all.nl>

In article <a8nric$sbb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> The three execution scopes are function local (hereafter referred to as
> local), module global (hereafter referred to as global), and builtin
> scope.  Local scope exists any time Python's instruction pointer is
> inside a function/method.  Global scope refers to the currently
> executing module; explicitly accessing a function in another module
> through attributes changes the current module:
> 
>     from M import f
>     f()                 # f()'s global scope is the current module
>     import M
>     M.f()               # f()'s global scope is now the module M

No: f()'s global scope is always module M. Scoping is static after all. 
Or are you saying something different?

Just


From JamesL at Lugoj.com  Fri Apr  5 23:17:09 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 04:17:09 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018040171.16545.python-list@python.org>
Message-ID: <Xns91E7CEFD91DC1JamesLLugojcom@216.148.53.84>

<brueckd at tbye.com> wrote:
> On 5 Apr 2002, James Logajan wrote:
>> The answer to this isn't as interesting to me as _when_ will this
>> tinkering end? Something that isn't up for debate: because of the
>> problems I have actually encountered with making code work with the
>> many extant versions of Python, I will no longer be doing any new
>> projects in the language and am not recommending Python to anyone.
> 
> So, what language did you move on to?

Since at the moment I don't have any projects in the pipeline needing a 
scripting language, I haven't decided yet. The only thing I have lined up 
right now is a project requring a PIC microprocessor (or equivalent) and it 
will require either assembler or cross compiled C. I've another 6 months at 
least to worry about scripting languages.


From jeff at ccvcorp.com  Fri Apr 12 13:58:42 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 12 Apr 2002 10:58:42 -0700
Subject: Problem with PIL and compression format
References: <mailman.1018602385.24023.python-list@python.org>
Message-ID: <MPG.1720c01d6f5d61239896ac@news.nwlink.com>

In article <mailman.1018602385.24023.python-list at python.org>, 
maric.michaud at cirec.com says...
 
> I tried ro include tif image in a pdf document with reportlab and Python
> 2.0, PIL 1.1.3.
> This work pretty well with uncompressed and 'Pack Bits" compressed TIFF
> files, but not with "CCITT group 4" that I need.
> Anyone has a suggestion for this ?

The problem here is that PIL doesn't include a codec for CCITT4 
compression.  (It's been asked for before, with no luck yet -- I 
imagine that /F has his hands full with more important things.  
And considering what we've paid for PIL, I'd hardly feel right 
about complaining... ;) )  

The only two solutions to this would be to 1) write your own 
codec (possibly submitting it as a patch to PIL ;) ) or else 2) 
find some other program/library to decode the CCITT4 images, 
possibly converting them into a different format/compression 
scheme.  (A possible third solution might be to pay for Secret 
Labs to develop and add this, themselves, but I don't expect 
that's very practical...)

I'm using wxPython to read CCITT4 TIFFs, but I don't know if it's 
practical to use it for *just* that purpose (I'm handling them as 
part of a GUI app anyhow).  There's also a number of external 
programs that you could use to batch-convert your TIFFs to some 
other format. (ImageMagick?)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mhammond at skippinet.com.au  Mon Apr  1 08:29:27 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 01 Apr 2002 13:29:27 GMT
Subject: Unicode with win32com and makepy
References: <71jaau4q7m8ljplv7r9vrog529dibnf4s9@4ax.com>
Message-ID: <3CA860DF.3040306@skippinet.com.au>

Tim Roberts wrote:

> Here's the question: what do we do in the long term?  Should makepy
> generate a __unicode__ member every time it generates a __str__ member?

Yep - and the next version will!

Mark.



From erno-news at erno.iki.fi  Fri Apr 26 15:23:46 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 26 Apr 2002 22:23:46 +0300
Subject: cgi module question
References: <3CC9A3E2.9060803@nmt.edu>
Message-ID: <kud6wmb6wt.fsf@lasipalatsi.fi>

In article <3CC9A3E2.9060803 at nmt.edu>, Jeff Shipman <shippy at nmt.edu> writes:

| I'm using Python 1.5.2 and the cgi module. I'm
| wondering if there's any way to recover the string
| that was passed to the cgi module during a POST
| operation (normally comes in through stdin)? 

yes. read sys.stdin. 

  -- erno


From niemeyer at conectiva.com  Sun Apr  7 18:11:42 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Sun, 7 Apr 2002 19:11:42 -0300
Subject: Stability and change
In-Reply-To: <a8pqec$g3r$1@panix1.panix.com>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <just-89BA7A.17235906042002@news1.xs4all.nl> <3CAF18F8.56821C0B@python.org> <mailman.1018172897.27393.python-list@python.org> <a8pqec$g3r$1@panix1.panix.com>
Message-ID: <20020407191142.A1940@ibook.distro.conectiva>

> >> Agreed.  I've moved this to python-dev; see you all there.
> >
> >Bye.
> 
> Just in case it's not clear, you're welcome to join the discussion on
> python-dev; it's not a restricted mailing list the way it used to be.

It was not clear. Thanks.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From sjmachin at lexicon.net  Thu Apr 25 22:41:28 2002
From: sjmachin at lexicon.net (John Machin)
Date: 25 Apr 2002 19:41:28 -0700
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com> <mailman.1019747686.17849.python-list@python.org>
Message-ID: <c76ff6fc.0204251841.480fac9d@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1019747686.17849.python-list at python.org>...
> On Thu, Apr 25, 2002 at 07:29:27AM -0700, Rad wrote:
 
> > All four files have dates in them but they do come in YYYYMMDD format
> > and I was planning to use string comparisons, same for the rest of
> > data,I was thinking to treat it all as strings.
> 
> Huh? YYMMDD is easily convertible to a 32bit value. see "help('time')"
> on the python-prompt.

Fortunately, Rad believes/hopes he is getting the dates in YYYYMMDD
format, not YYMMDD.

When filtering the first input file, Rad should be concentrating on
rejecting irrelevant records as cheaply as possible e.g.

   if transaction_date < '19990701': continue

Once all irrelevant records are rejected, *IF* he needs to do date
arithmetic, then he can convert his date strings to whatever other
format.

Rad is doing some _data processing_. The mxDateTime module seems to
have been designed with data processing in mind. The time module seems
to be provided by Python for Unix compatibility <wink>.

> You should really start to generate some example files (made up
> by you) and start to code *right now*.

If Rad has never seen the data before, generating some example files
made up by himself would be an exercise in utter futility. He should
be getting sample test files from the source.


From gcordova at hebmex.com  Fri Apr  5 17:55:58 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 5 Apr 2002 16:55:58 -0600 
Subject: Beginning Tkinter problem
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB7A@mail.hebmex.com>

> 
> And again answering my own question with dubious authority it 
> appears that if you leave the widget.mainloop() out it works fine.
> Possibly because IDLE is a Tkinter application itself and is already
> running a mainloop() and my little app inherited it?
> Now I'm really guessing so somebody clue me in!
> 
> Thanks,
> 
> John Purser
> 

You are correct :-)

-gustavo




From vomjom at vomjom.org  Fri Apr 19 06:24:07 2002
From: vomjom at vomjom.org (Jonathan Hseu)
Date: Fri, 19 Apr 2002 05:24:07 -0500
Subject: Bug (?) in httplib SSL
Message-ID: <20020419102407.GA4802@vomjom>

I'm not quite sure whether this is a bug, or it's just my fault.
Basically, when I try a POST over SSL on an IIS server, it hangs.

I am using Python 2.2.1, and here is the relevant code:
	params = urllib.urlencode(account)
	headers = {"Content-type": "application/x-www-form-urlencoded",
		   "Accept": "text/plain"}
	connection = httplib.HTTPSConnection(server)
	connection.set_debuglevel(1)
	connection.request("POST", "/LoginPost.asp", params, headers)
	response = connection.getresponse()

It hangs there, here's the traceback:
Traceback (most recent call last):
  File "./smote.py", line 37, in ?
    smote()
  File "./smote.py", line 30, in smote
    response = connection.getresponse()
  File "/usr/lib/python2.2/httplib.py", line 591, in getresponse
    response = self.response_class(self.sock, self.debuglevel)
  File "/usr/lib/python2.2/httplib.py", line 99, in __init__
    self.fp = sock.makefile('rb', 0)
  File "/usr/lib/python2.2/httplib.py", line 628, in makefile
    buf = self.__ssl.read()

I've been looking through old bug reports, and there was one that I was
receiving earlier (until I upgraded).  The reason people received this
bug was apparently because some web servers, such as IIS, don't complete
the SSL connection before they close it.

Perhaps that is the problem?

By the way, I am not subscribed, so please e-mail me directly.
Thank you.

-- 
Jonathan Hseu <vomjom at vomjom.org, vomjom at debian.org, jh4 at cec.wustl.edu>
GPG ID:          5228D713
GPG fingerprint: 220B A4EF 70FE B884 CB38  F93F EA8A 1024 5228 D713




From bokr at oz.net  Sun Apr 14 03:41:33 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 07:41:33 GMT
Subject: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <mailman.1018738411.20464.python-list@python.org> <a9appm$l47$0@216.39.172.122> <mailman.1018753407.10536.python-list@python.org>
Message-ID: <a9bbrd$mcg$0@216.39.172.122>

On Sun, 14 Apr 2002 05:01:50 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> wrote:

>* Bengt Richter <bokr at oz.net> [2002-04-14 02:33 +0000]:
>> On Sun, 14 Apr 2002 00:41:37 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> wrote:
>> 
>> >* Jens Baader <nospam at nospam.com> [2002-04-13 23:46 +0200]:
>> >                      ^^^^^^^^^^
>> >
>> >If you don't have permission of the nospam.com domain owner to use their
>> >domain, you should stop abusing their services *now*. If you don't want
>> >to get spam, I'd recommend you get a throwaway email address at a
>> >freemailer.
>> >
>> I agree, but OTOH you'd think the IETF would come up with a standard
>> global /dev/null sort of address for this purpose. Is there such a thing?
>> I.e., why burden freemailers with garbage storage, and the net with the traffic?
>
>That's already been done:
>
>http://www.rfc-editor.org/rfc/rfc2606.txt
>
><quote>
>    To safely satisfy these needs, four domain names are reserved as
>    listed and described below.
>
>    .test
>    .example
>    .invalid
>    .localhost
>
>    ".test" is recommended for use in testing of current or new DNS
>    related code.
>
>    ".example" is recommended for use in documentation or as examples.
>
>    ".invalid" is intended for use in online construction of domain
>    names that are sure to be invalid and which it is obvious at a
>    glance are invalid.
></quote>
>
>So if you really *never* want to receive email replies from usenet
>posts, you should construct a fake email address and add .invalid at the
>end.
>
Great. So we can recommend that Jens change his

    Jens Baader <nospam at nospam.com>
to
    Jens Baader <nospam at nospam.invalid>

I'd CC: him, but ... ;-)

Regards,
Bengt Richter


From sholden at holdenweb.com  Wed Apr 24 11:23:44 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 11:23:44 -0400
Subject: Slight irritation with try/finally indentation
References: <mailman.1019657853.14965.python-list@python.org>
Message-ID: <n8Ax8.3546$z55.2175@atlpnn01.usenetserver.com>

"Fran?ois Pinard" <pinard at iro.umontreal.ca> wrote in message
news:mailman.1019657853.14965.python-list at python.org...
> Hello, people.
>
> There is something which I find slightly irritating with the try/finally
> construct, and wonder if someone would not have a good advice to offer.

[obscure indentation topic]

Well, my advice would be to get on with your life, but I can't really hope
you'll see this as "good". Sheesh, don't you have *anything* to do? <wink>

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/






From db3l at fitlinxx.com  Thu Apr  4 23:55:28 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 04 Apr 2002 23:55:28 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83>
Message-ID: <u3cyad9rz.fsf@ctwd0143.fitlinxx.com>

James Logajan <JamesL at Lugoj.com> writes:

> Thank you. I only just resubscribed to comp.lang.python after many months 
> of absence and was only reading a handful of articles of immediate interest 
> and was not following the PEP 285 thread until Guid posted the ACCEPTED 
> message. When I looked at it I saw a "class bool(int)" definition, so I 
> wasn't sure how instances of such a (pseudo?) class would be handled in a 
> dictionary.

The class definition in the PEP does pretty much answer this because
it shows the inheritance from the built-in "int" type, but I agree you
need to work it through mentally (or test via the CVS Python) - it's
not explicitly shown in the PEP.

>   (...)                                      I assumed the "ACCEPTED: PEP 
> 285" posting would have subsumed previous discussions, including any 
> concerning the class instances as dictionary keys. I couldn't find a clear 
> answer there, hence my question.

There were several lengthy threads that cropped up in response to
Guido's original posting of the PEP (which in turn followed a lengthy
discussion on the python-dev list).  The ACCEPTED thread was just
Guido's final decision, so it's unlikely it would automatically
duplicate much of the prior discussion.

> My apologies to those who have sent me angry e-mails for not realizing that 
> PEPs do not stand alone from the lengthy discussion threads on Usenet. The 
> fault is entirely mine.

I'm not sure any angry comments were called for (hope you didn't take
mine in that vein), but I do think that in a case like this, it's a
good idea to review prior threads (e.g., with google for the
newsgroup) and perhaps python-dev archives.  At a minimum in a case
like this - and particularly if you're going to point out a lack of
response to a question of yours - searching for all posts/e-mail by
Guido first to see if perhaps the question already arose can't hurt :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From sholden at holdenweb.com  Thu Apr 18 11:05:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 18 Apr 2002 11:05:25 -0400
Subject: Assignment operator?
References: <a9mksa$8dqs$1@as201.hinet.hr>
Message-ID: <WiBv8.15852$eS3.5828@atlpnn01.usenetserver.com>

"Ante Bagaric" <abagaric_NOSPAM at rest-art.hr> wrote ...
>
> After reading that PEP about static namespaces I'm now pretty confused
with
> the behaviour of the += operator.
>
It might have been helpful to quote the PEP number. I'm gyuessing you mean
PEP 227, from

    http://www.python.org/peps/pep-0227.html

> The PEP states that if there is a binding of certain name anywhere within
> the body of the function, it is considered local. Then it states that
> binding operations include stuff like def, for, except... AND assignment.
>
Correct. The compiler makes a static analysis of each function body, and
every name which is assigned within the function body is regarded as local,
even if there are other uses, unless it's explicitly declared to be global.

This sometimes leads to oddities like:

>>> a = 3
>>> def myfunc():
...     print a
...     a = 4
...
>>> myfunc()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in myfunc
UnboundLocalError: local variable 'a' referenced before assignment

The weird thing here is that if you remove the "print" statement the
function raises no exception. It doesn't do much, but that's another story.

> I guess that for some reason operator += is considered an 'assignment'
> (because it is often called 'one of assignment operators') but is it
really?
> Is 'a = a + 2' the _same_ as 'a += 2'? I don't think it is, or should be.
> The 'a = a + 2' means 'store the value of a + 2 to the variable a' and 'a
+=
> 2'
> means 'increment a by 2'.
>
Indeed, +=. -= and the rest are referred to as "augmented assignment
operators". Whether or not you think that 'a = a + 2' should be the same as
"a += 2" is a matter of choice. The implementation, however, clearly allows
operators to implement in-place update on mutable objects *if they choose*,
but does not require it.

Obviously, immutables must implement augmented assignment by re-binding the
name to a new object, otherwise

>>> a = 3
>>> a += 2
>>> a
5

would not be possible.

> The fact is that Python already behaves exactly like that. If you say
>
> >>> li = [1,2,3]
> >>> li2 = li
> >>> li += [4,5]
> >>> li is li2
> 1
>
> but if you say
>
> >>> li = [1,2,3]
> >>> li2 = li
> >>> li = li + [4,5]
> >>> li is li2
> 0
>
Indeed. Thr PEPs aren't all proposals for the future: you might notice that
in

    http://www.python.org/peps/

PEP 227 is listed in the "finished PEPs" section. It's one of Python's
charms, or weaknesses, depending on how authoritative you expect your
documentation to be, that it's quite possible to have features in the
language whose specifications include such extracts as

"""XXX Explain the examples"""

"""XXX need some counterexamples"""

and (my favorite)

"""XXX Jeremy, is this still the case?""".

However, we should remember that the Python development team are humans, and
it's much easier to bitch about this sort of thing than it is to roll up
one's sleeves and fix them.

> According to this, += operator can't be though of as an 'assignment
> operator'.
> And conversely, it shouldn't cause the compiler to consider 'foo' to be
> local
> if there is 'foo += stuff' somewhere within the body of the function.
> After all, 'foo.append(stuff)' or 'foo.extend(stuff)' don't do it.
>
> Or there is something I didn't quite understand?
>
I think what you haven't understood is the update in-place semantics that
mutable object implementations *may* choose to adopt. Otherwise you seem to
have it down pat.

regards
 Steve






From martin at v.loewis.de  Sun Apr  7 16:14:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 07 Apr 2002 22:14:52 +0200
Subject: Lex and Yacc for Python
References: <slrnab19hf.23u.Alex@Tron.ddeus.lan>
Message-ID: <m3vgb3s1tv.fsf@mira.informatik.hu-berlin.de>

Oleksandr Kozachuk <Alex at DDeus.de> writes:

> I need something, comparable to Yacc/Lex. Or something for
> Context-Free-Grammar. Can any one recommend me anything?

I recommend YAPPS if you need a simple solutions to a simple problem,
and SPARK if you need a complex solution to a complex problem. See

http://www.python.org/sigs/parser-sig/

Regards,
Martin



From pobrien at orbtech.com  Wed Apr  3 14:25:05 2002
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Wed, 3 Apr 2002 13:25:05 -0600
Subject: [Python-Dev] PEP 285: Adding a bool type
In-Reply-To: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <NBBBIOJPGKJEKIECEMCBKEFDMLAA.pobrien@orbtech.com>

[Guido van Rossum]
>
> I offer the following PEP for review by the community.  If it receives
> a favorable response, it will be implemented in Python 2.3.

Yikes! I've read every single response to this so far. It has been quite
enlightening for me. At the same time, I hope it ends soon. ;-)

At this point I'm +1 on the proposal, but I have a favor to ask. While a lot
of the issues raised have been revealing and provocative, I think the vast
majority are tangential to the real substance of this PEP. I'm having
trouble separating what is important from what is not. Is there any way we
can focus the issue a bit?

Specifically:

It feels like there are too many combinations available in the PEP. (Or
maybe too many questions and options which could result in weird
combinations.) I'd like to see a new version with Guido's latest thoughts.
For example, are we closer to a resolution on the str() and repr() issues?
The only behavior I like is that exhibited by the sample code:

>>> str(True)
'True'
>>> repr(True)
'True'

How about the case of the constants: true, false vs. True, False? I like the
latter, which match None. They would also stand out in the new docstrings
mentioned by Fred.

How about operator.truth()? Has Tim changed his mind at all? Has Guido?

If this PEP were implemented and "all built-in operations that conceptually
return a Boolean" were updated, what verifiable examples of code breakage
would we have?

The issue of dictionary keys has been raised, and is a bit perplexing, but
not overly so, IMO. However, I'd like confirmation that this is indeed the
intended behavior:

>>> d = {}
>>> d[0] = 'First'
>>> d
{0: 'First'}
>>> d[False] = 'Second'
>>> d
{0: 'Second'}
>>> d[0.0] = 'Third'
>>> d
{0: 'Third'}
>>> d[()] = 'Fourth'
>>> d
{0: 'Third', (): 'Fourth'}
>>> d[None] = 'Fifth'
>>> d
{0: 'Third', (): 'Fourth', None: 'Fifth'}

Or are there any situations like this where you would want True and False to
be a bit more like None and a bit less like ints? Meaning True and False
could be dictionary keys separate from 0 and 1.

Other than that, are there any concrete issues that have been raised that
need to be addressed by the PEP? I understand the uneasy feeling felt by
many as I felt it myself. However, I've read all the newsgroup postings and
reread the PEP and I don't feel very queasy any more. I think this PEP is a
step in the right direction. I would hope that clarifying some of the
genuine issues might help to bring about consensus and reduce some of the
rhetoric.

---
Patrick K. O'Brien
Orbtech





From akuchlin at ute.mems-exchange.org  Wed Apr 24 08:55:10 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 24 Apr 2002 12:55:10 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com>
Message-ID: <slrnacdape.60f.akuchlin@ute.mems-exchange.org>

In article <7xsn5l6zn7.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
> Although, HMAC in Python using string multiplication instead of loops
> might not be so slow.  I'll see if I can code it that way.

Note that Python 2.2's standard library includes an hmac module.

--amk


From ianb at colorstudy.com  Tue Apr  9 01:57:17 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 09 Apr 2002 00:57:17 -0500
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <_1vs8.1787$CA6.114998@newsread1.prod.itd.earthlink.net>
References: <mailman.1018314090.9984.python-list@python.org> 
	<_1vs8.1787$CA6.114998@newsread1.prod.itd.earthlink.net>
Message-ID: <1018331838.472.796.camel@localhost>

On Tue, 2002-04-09 at 00:32, Chris Tavares wrote:
> 
> "Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> news:mailman.1018314090.9984.python-list at python.org...
> [ ... snip ... ]
> >
> > I'm not sure that Pyrex will ever compile *all* valid Python
> > code. It probably wouldn't be very effective at speeding up
> > pure Python code anyway. The reason it deals with Python code
> > at all is so that you can write code to convert between
> > Python and C data structures -- the fact that you can
> > compile pure Python with it is sort of a side effect.
> >
> [ ... snip ... ]
> 
> It could be interesting to see how far you could get with just the subset
> you do compile, though. The Squeak people can compile their entire VM from
> Smalltalk to C, and their translator doesn't even do classes or objects! And
> they get a big, big win from it in terms of portability and ease of
> experimentation.

One of the big advantages they had of this system was that they could
run the VM inside Squeak and use Squeak's excellent debugging tools
(well, I think they were actually running it in another version of
Smalltalk when they were bootstrapping).  The language was a subset of
Smalltalk, which made this possible.  Pyrex isn't a subset, since it
introduces new keywords and static types -- the static types could
potentially be ignored, with the expectation that you might encounter
new type-related bugs when you moved to C.  This would have been an
issue for Squeak as well -- it's certainly no closer to C than Python --
but I don't know how they solved this problem.

Of course, Python has never had the insular ambitions that
Squeak/Smalltalk has, so a Pure Python World isn't as compelling.  But
that doesn't mean prototyping Pyrex code in interpreted Python wouldn't
be very nice -- especially since Pyrex runtime errors are unlikely to
have nice tracebacks (or even very readable translated C code).  Maybe
all that's required is a custom loader that filters out cdef and type
declarations.  Better yet if some of the type declarations could be
turned into asserts (though I don't know how you'd ensure that the
variable remained the correct type for the entire time -- just when a
function was first called).

  Ian






From philh at comuno.freeserve.co.uk  Sat Apr  6 12:42:35 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sat, 6 Apr 2002 18:42:35 +0100
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <slrnaaucsb.3j0.philh@comuno.freeserve.co.uk>

On 05 Apr 2002 19:49:54 -0800, Paul Rubin <phr-n2002a at nightsong.com> wrote:
>
>Most recent changes in Python are good (list comprehensions, x in dict,
>etc) but I'd rather that they appear only in major releases (2.0, 3.0, etc)
>which happen every 2 years or so, rather than in the minor releases
>that come out every few months.

Sounds good to me.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From harisri at bigpond.com  Tue Apr 23 22:16:41 2002
From: harisri at bigpond.com (Srihari Vijayaraghavan)
Date: 23 Apr 2002 19:16:41 -0700
Subject: stdin - seek() issue on Linux.
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com> <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com> <aa1bqb$lek$1@nntp6.u.washington.edu> <284b44c2.0204221908.1280cfc3@posting.google.com>
Message-ID: <284b44c2.0204231816.67c01ca8@posting.google.com>

Hello,
> > Quoth "Terry Reedy" <tejarex at yahoo.com>:
> > | "Srihari Vijayaraghavan" <no at spam.please> wrote in message
> > | news:WMTw8.46875$uR5.106037 at newsfeeds.bigpond.com...
>  
> > |> I am a new-bie python user. I have Python 2.2.1 installed on a Linux
> > |> computer. I get the following error message when I try to seek() in the
> > |> stdin, like:
> > |> #cat /var/log/messages | python test.py
> > |> (where /var/log/messages is over 1MB in size)
> > |>
> > |> Traceback (most recent call last):
> > |>   File "test.py", line 3, in ?
> > |>     sys.stdin.seek(-512, 2)
> > |> IOError: [Errno 29] Illegal seek
> > |
> > | By 'cat'ting the random access disk file, you turned it into a
> > | forward-access-only stream of bytes.  You definitely cannot seek
> > | backwards (and maybe not forward - don't know).  Instead, open() the
> > | file and use the resulting file object:
> > |
> > | msgs = open("/var/log/messages", 'r')
> > | msgs.seek(-512, 2)
> > | ...
> > 
> > Right, though it can be fixed without changing test.py at all -
> > "python test.py < /var/log/messages" will work fine.  This idiom
> > ("cat file | ...") used to regularly receive a UUOC award, "Useless
> > Use Of Cat", from notorious Perl hacker Randall Schwartz when I
> > followed comp.unix.shell a few years back.
> > 
> > Of course if the real program uses some filter, instead of "cat",
> > then there's no solution - you can't seek in a pipe, so the answer
> > has to be "don't do that".
> > 
Once again thanks for your responses. I really appreciate that.

OK. IMHO sys.stdin doesn't work as stated in PLR (Python Library
Reference), i.e., it doesn't work like a _real_ file object,
particularly not supporting seek() related operations. While your
technical explanation sounds OK to me, but I am still not convinced
why PLR advertises that sys.stdin is treated as a file object. Please
correct me if I am wrong here (as I am still a newbie user :-).

I have fixed the problem by using cStringIO.StringIO(sys.stdin.read())
method for now, which I think may not be the most optimised method.
(as I guess it might read the entire file into the memory, OTOH using
seek() method I could read only specific part of the file (sys.stdin)
that I want)

Thanks for your time.

Hari,
harisri at bigpond.com


From bryanjugglercryptographer at yahoo.com  Wed Apr 24 16:17:34 2002
From: bryanjugglercryptographer at yahoo.com (Bryan Olson)
Date: 24 Apr 2002 13:17:34 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <slrnacdape.60f.akuchlin@ute.mems-exchange.org> <7xu1q1mdwd.fsf@ruckus.brouhaha.com>
Message-ID: <1a517b5.0204241217.7400376b@posting.google.com>

Paul Rubin wrote:

> Yes, but this encryption module is supposed to work in 1.5.2.
> 
> Also, the 2.2 hmac module is written in python and looks pretty slow.
[...]
> And it actually might be faster to use two separate hash-based auth
> keys than deal with the fancy HMAC padding, if a simple double hash
> has some vulnerabilities.


HMAC's is only a small additive constant slower than the hash function.  
For messages more than a few blocks long, it's not significant.  The
extra padding and additional hash function calls only deal with two blocks.

Below is single-call HMAC in Python.


| import sha
| import md5
| import string
|
|
| def xor_strings(s1, s2):
|     """Pass two strings of the same length; returns their
|     bit-wise exclusive or.
|     """
|     char_list = map(lambda x, y: chr(ord(x) ^ ord(y)), s1, s2)
|     return string.join(char_list, "")
|
| def hmac(message, key, hash_function, block_size):
|     """Generic HMAC, as per RFC 2104.  The hash function must
|     follow the usual Python new..update..digest interface.
|     """
|     if len(key) > block_size:
|         key = hash_function.new(key).digest()
|     ipad = chr(0x36) * block_size
|     opad = chr(0x5C) * block_size
|     key = key + chr(0) * (block_size - len(key))
|     hash1 = hash_function.new(xor_strings(key, ipad))
|     hash1.update(message)
|     hash2 = hash_function.new(xor_strings(key, opad))
|     hash2.update(hash1.digest())
|     return hash2.digest()
|
| def hmac_sha1(message, key):
|     return hmac(message, key, sha, 64)
|
| def hmac_md5(message, key):
|     return hmac(message, key, md5, 64)
|


From ianb at colorstudy.com  Tue Apr  9 02:44:20 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 09 Apr 2002 01:44:20 -0500
Subject: multiple ranges to for statement?
In-Reply-To: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu>
References: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu>
Message-ID: <1018334661.472.964.camel@localhost>

On Tue, 2002-04-09 at 01:33, Damian Menscher wrote:
> I'm hoping to do something along the lines of:
> 
> for i in range(a,b),range(c,d):
>     stuff goes here

Most simply:

for i in range(a, b) + range(c, d):
    ...


--Ian






From Oschler at earthlink.net  Fri Apr  5 14:40:41 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Fri, 5 Apr 2002 14:40:41 -0500
Subject: Rule Engine in Python
References: <eb612768.0203101940.5565b282@posting.google.com> <m3it82d32l.fsf@wilma.localdomain> <I45r8.34886$0T6.10550048@e3500-atl1.usenetserver.com> <m3u1qq3ma8.fsf@wilma.localdomain>
Message-ID: <B4nr8.36171$ku5.3121156@e3500-atl2.usenetserver.com>

Harry George <hgg9140 at seanet.com> wrote in message
news:m3u1qq3ma8.fsf at wilma.localdomain...
> A few lines of python doing
> regualr expressions and some text manipulation takes the palce of a
> page or so of prolog.  That was all with ALS -- I've used SWI for a
> few simple examples, but never compared them for heavy processing.
>

I don't want to misinterpret anything, but are you indicating that if you
had your druthers you'd replace the prolog code wth python code?

thx





From nospam-readonly at getsoft.com  Sun Apr 21 21:35:17 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 11:35:17 +1000
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net>
Message-ID: <3cc368bf$0$15477$afc38c87@news.optusnet.com.au>

Hi David,
Ok this all starts to make some sense now. I did see some docs refer to
ndiff being in difflib but then other difflib doc's which didn't mention it.
I did try executing '/tools/scripts/ndiff' but have to admit I don't really
know what I'm doing just yet and didn't get anywhere. Is there some one to
run this in 2.1.2 or should I just install 2.2?

Thanks for your help.


"David Goodger" <goodger at users.sourceforge.net> wrote in message
news:B8E8DEB0.21E3D%goodger at users.sourceforge.net...
> Neville Franks wrote:
> > I still can't get ndiff to work. If I enter: 'difflib.ndiff' I get
> > 'AttributeError: 'difflib' module has no attribute 'ndiff'
>
> Tim Peters' ndiff.py is a script in the tools/scripts directory.
> Originally, it was self-contained, but at some point Tim migrated the
> SequenceMatcher class into the difflib module.  I extracted the rest of
the
> generic code and turned it into the difflib.Differ class and
difflib.ndiff()
> convenience function, Tim converted it to use generators, and it was
> released with Python 2.2.  You're running Python 2.1.2, so difflib.ndiff
> doesn't exist yet.  You'll have to upgrade to get the full functionality.
> You can't use the Python 2.2 difflib module in Python 2.1, because it uses
> the generator feature, new to Python 2.2.  However, you can get a Python
> 2.0/2.1-compatible version from
> http://docutils.sourceforge.net/test/difflib.py; this is the module after
I
> added Differ and ndiff(), but before Tim converted it to use generators.
>
> --
> David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
>   - Python Docutils: http://docutils.sourceforge.net/
>   - The Go Tools Project: http://gotools.sourceforge.net/
>




From stojek at part-gmbh.de  Tue Apr  2 17:16:06 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Tue, 02 Apr 2002 22:16:06 GMT
Subject: wxpython: problem with events and checklistbox
Message-ID: <3caa2be5.48482250@news.easynews.net>

HI,

in my GUI I have a wxCheckListBox (named CheckList) and an fuction
IfChecked(self,event) that process the checkevents (EVT_CHECLISTBOX).

Now, when unpickling all data at the start of my program,
I want to check some items with CheckList.check(index,true)
My problem is, that after each checking the function IfChecked is
called. In the docs I found something about PopEventHandler, but this
doesn't seem to solve my problem. Again, this should be an all day
problem, but I have no idea where to look.

Any hints?

Thanks,
Marcus


From mdk at mdk.com  Wed Apr 10 08:40:33 2002
From: mdk at mdk.com (mdk)
Date: 10 Apr 2002 12:40:33 GMT
Subject: Display current line number of running script
Message-ID: <Xns91EC58418DA12mdkmdkmdkmdk@130.133.1.4>

Hello,

Is there a function that would return the current line number that the 
running script is on?

Thanks!


From fredrik at pythonware.com  Wed Apr  3 12:30:41 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 17:30:41 GMT
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <5%Gq8.36874$l93.7328192@newsb.telia.net>

Roy Smith wrote:
> I've got a dictionary d, which may or may not have a value I'm looking
> for.  If it does, I want to do a few things.  If it doesn't, I want to
> do a few other things.  I could write:
>
> try:
>     value = d[key]
>     foo = stuff (value)
> except KeyError:
>     foo = otherStuff (key)

    try:
        value = d[key]
    except KeyError:
        foo = otherStuff(key)
    else:
        # found value, do stuff
        foo = stuff(value)

> The alternative would be something like:
>
> foundIt = 0
> try:
>     value = d[key]
>     foundIt = 1
> except KeyError:
>        foo = otherStuff (key)
>
> if (foundIt):
>    foo = stuff (value)
>
> but that's exceptionally ugly.  Any suggestions for how better to
> write this?

by following the Python style guide? ;-)

</F>




From geoff at gerrietts.net  Fri Apr  5 14:48:06 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 5 Apr 2002 11:48:06 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net>
Message-ID: <20020405194806.GA4774@isis.gerrietts.net>

Quoting Tim Peters (tim.one at comcast.net):
> [phil hunt, on language change]
> > ...
> > But is this changing going to be continual? Are the language
> > developers constantly going to tinker with the language, out of a
> > sense of fun? (Not that having fun is bad, it's just that I don't
> > want to code against a continually moving target).
> >
> > Is there a final destination in sight, which these changes are
> > leading to?
> 
> It's possible it's the same final destination that C, Fortran, Perl and Java
> (for examples) are approaching with their own never-ending changes, but, if
> so, I doubt we'll live long enough to celebrate arrival.  If you want a
> change-free language, you're best off looking for a dead or dying language.
> For example, Algol 60 hasn't changed in decades.  For that matter, Python
> 1.5.2 hasn't changed in years.  WRT any technology, you can jump off the
> train any time you like.  Unless you pick your exit point very carefully,
> though, you may end up alone in the middle of nowhere.

It's a little frightening to me how much this sounds like a threat.

I admire the changes made to the language over the last couple years,
and think of them as marked and pronounced improvements. I look
forward to 2.3 and 3.0.

I'm a little leery about the rate of change, though. It isn't too hard
to develop across versions these days, though it's harder than I
initially expected. I don't think I'm complaining yet, but I am
unsettled.

I'll just drop two thoughts into the discussion.

The first comes from some popular management-theory from a few years
ago -- might have been _Managing Change_ but it might have been
something else, I forget exactly. This book pointed out that human
nature responds better to one big, massive change than a dozen smaller
ones. That goes against what seems logically true, that the smaller
changes are easier to take. I guess the psychology, though, is that
the ambiguity and uncertainty caused by change has some constant
factor that is more disruptive than the actual change itself. One big
change, and people grapple with it and get back to "normal". With a
bunch of little changes, people are in a continual state of ambiguity.

I think the recommendation that comes from this principal is that any
change which breaks backward compatibility is worth a major version
number. You can disagree; I'm not arguing or complaining, just
suggesting a management approach.

The second is that I personally find little to admire in C++, but one
of the things I do admire is that Bjarne decided to take a couple
years off to let the compiler writers and application programmers
catch up. The result was a language that everyone more or less thought
they knew, a solid foundation for people to build on and move forward.

It may not have worked quite as well as it should have, because the
compilers available today vary pretty widely in implementation. But
there's some value to taking a day off.

I remember in the early days of Java, when 1.0 went to 1.1 and
everything broke, all the developers I was learning from expressed
disgust. That was the first big indicator that the Java's promise had
crossed fingers to some. To others it was a sign that Java was an
immature language.

I like the features. I like the ideas that drive the features. I have
nothing but respect for the python-dev crowd and the magnificent
language they've produced. I prefer programming in Python over all
other languages I know. I'm not trying to be critical. But I do think
the change-management strategy is flawed.

Thanks,
--G.

-- 
Geoff Gerrietts        <geoff at gerrietts dot net>
"Ordinarily he was insane, but he had lucid moments 
when he was merely stupid."        --Heinrich Heine




From ajs at ix.netcom.com  Sun Apr 14 10:49:59 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 14 Apr 2002 10:49:59 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000601c1e3c3$ca50b7e0$0334fea9@carol>

I happen to follow python-list through google and see
a post sent out by myself April 11 and lost in cyber 
showing up as a current post.

There being a time and place for everything - and
if there ever was a time and place for that post it
is certainly not now. 

Art





From billdozier at hotmail.com  Tue Apr 30 09:34:13 2002
From: billdozier at hotmail.com (Bill Dozier)
Date: 30 Apr 2002 06:34:13 -0700
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com>
Message-ID: <199334c5.0204300534.4dc4b4de@posting.google.com>

Erik Max Francis <max at alcyone.com> wrote in message news:<3CCE5F2A.1103E744 at alcyone.com>...

> Ralf Juengling wrote:

> 

> > when I want to make sure within a function that it can deal with

> > the arguments passed, the only choice is to check it's type via

> > 'isinstance()'.
> > So, what is the proper, 'pythonic' way of argument checking?

> 

> The "Pythonic" way of argument checking is to simply use it and catch

> the exceptions that will be thrown if it doesn't behave as planned.

> 


IMHO, "isinstance()" is usually not the right thing. Whether it's
"pythonic" or not, I can't say.

When you have code that checks the type of an object and then treats
it accordingly, it's usually a sign that you need to use some
polymorphism. Of course, you don't always have the luxury of fixing
the design.

Bill


From jeff at ccvcorp.com  Mon Apr 22 15:20:59 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 12:20:59 -0700
Subject: Mass Mailing help needed (Paul Rubin)
References: <mailman.1019325819.29016.python-list@python.org>
Message-ID: <MPG.172e025d803fc2009896bf@news.nwlink.com>

In article <mailman.1019325819.29016.python-list at python.org>, 
jvanasco at hotmail.com says...
> 
> I work mostly for large entertainment companies -- if you would like so, i 
> can send you to my portfolio and corporate site -- this particular list 
> problem is centered around a record label's website relaunch.  Instead of 
> inundating fans with 1 mailing from 20+  artist lists that they choose to 
> subscribe to, I've decided to merge multiple lists into 1.  Subscribers will 
> get 1 email with general label info, and specific artist news will be 
> provided based on the user's requests.  (Posts will only be about monthly, 
> aside from an 'album release' email that could be offschedule -- but that 
> specific list data is always available)

Well, I don't know Perl, so I can't say anything about how 
Mail::Bulkmail works, nor do I have experience with quite the 
scale that you're intending here, but...  One of my first Python 
projects was an automailer that emails various reports to 
clients, and it has worked fairly well for us.  I keep meaning to 
refactor it to use the new email package, but since it isn't 
broken, the priority on that is low...

I would avoid directly invoking sendmail at all.  My mailer uses 
smtplib to contact the mail server, and this *seems* fairly 
efficient.  Of course, our volume is less than 100 messages per 
day, so my experience may be different -- most of our overhead 
comes in converting raw text data to excel spreadsheets.

Since smtplib functions at a level of individual messages, you'd 
need to write your own bulk-handling utility that uses smtplib 
internally.  I'd think that it should be fairly simple to pass a 
list of email message instances, extract the destination address, 
and send them individually to smtplib.  This might not be 
extremely fast, but if you're sending out such mailings only once 
or twice a month, I suspect that an extremely short running time 
is not a hard constraint.  <wink>

-- 

Jeff Shannon
Technician/Programmer
Credit International


From wilk at flibuste.net  Mon Apr  8 06:32:51 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Mon, 8 Apr 2002 12:32:51 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
	<20020408120242.41456ace.wilk@flibuste.net>
	<a8rqh2$9h5$03$1@news.t-online.com>
Message-ID: <20020408123251.48eab8d6.wilk@flibuste.net>

Le Mon, 8 Apr 2002 12:20:00 +0200
"Harald Schneider" <h_schneider at marketmix.com> ?crivait:

> The stability of Python on Win32 is excellent. Even threadening is
> absolutely robust! Long running scripts are
> also stable ...
thanks

> 
> But what about the wxPython bindings? The only wxPython projects I know
> are some smaller applications, nothing
> really complex ....
When i see boa constructor i'm a litle bit afraid... but it doesn't mean
that it's a problem of wxPython. Also when i make some tests of wxPython
with idle or pythonwin i have often big scratch (whitout python log). But
i have never theses scratch if i run my script from the command line.

The opinion of an experienced user of wxPython would be welcome !

bye


(sorry for my bad english writing) 
> 
> Bye,
> Harald
> 
> "William Dod?" <wilk at flibuste.net> schrieb im Newsbeitrag
> news:20020408120242.41456ace.wilk at flibuste.net...
> > Le Mon, 8 Apr 2002 11:55:24 +0200
> > "Harald Schneider" <h_schneider at marketmix.com> ?crivait:
> >
> > > Hi,
> > >
> > > what are your experiences, using wxPython/wxWindows in large
> > > projects?
> > >
> > > Large means in my case a sales system, based on MySQL.
> > > I'm thinking about implementing it in wxPython, but I'm not sure
> > > about the stability of
> > > wxWindows and the wxPython wrapper ...
> > >
> > > The alternative way would be a project in Borland Delphi, which has
> > > proofed its
> > > stability over years ...
> > >
> > > So what do you think about it!
> > I think i have the same question in my head !
> > More, i would like to know how is the stability of python on windows,
> > compare to java or delphi for example.
> >
> > bye
> >
> > --
> > William Dod? - Informaticien Ind?pendant
> > http://www.flibuste.net
> 
> 


-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From aleax at aleax.it  Mon Apr  8 17:48:20 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 21:48:20 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8sf0k$2a5$0@216.39.172.122> <ahjs8.73608$S52.2597704@news2.tin.it> <a8sn8u$sj7$0@216.39.172.122>
Message-ID: <Eeos8.74753$S52.2651788@news2.tin.it>

Bengt Richter wrote:
        ...
> Well, I guess that depends on how you define "assignment" in Python.

I don't have to, as the Language Reference sees to that:
http://www.python.org/dev/doc/devel/ref/assignment.html

> ISTM there really is no such thing, in the sense of updating a named
> storage place with new content, 

"Assignment statements?are used to (re)bind names to values and to modify 
attributes or items of mutable objects" -- I guess one COULD at the
limit call an attribute "a named storage place", but it would not be a very
fruitful thing in Python.

> expression ;-) I'm not sure what that side effect should be called
> if not assignment, since it caused the identical binding that
> a='bee' would in that context.

It's a serious logical fallacy to reason "assignment statements may
(among other things) rebind names, therefore what rebinds a name
is an assignment".  Many bindings occur other than by assignment.
Would you call a for statement, a def statement, a class statement,
or an import statement "assignments", because they also rebind
names?!

> If it binds like a duck... ;-)

Should I try to display my imitatory abilities by quacking like a duck,
that does not justify your shooting me even if you have a hunting
license and ducks are in season.

> My criterion for judging whether it was an assignment was whether
> it accomplished the same name re/binding as a normal assignment,
> and I think it did, in the demonstrated context. So IMO Aahz's
> quote doesn't quite remain perfectly right ;-)

In many restricted contexts it's long been perfectly feasible to
perform a binding (and nothing but that) within an expression.
Your example was horrid -- it works "by accident" in a given
implementation, in a certain restricted context, because a
undefined behavior happens to be that way in said specific
implementation.  One can of course do much better than that:

        [x for x in (y,)][0]

is an expression perfectly valid in all contexts which binds or
rebinds name x to the same value as name y is bound to and
also has that value as its value-as-expression.  It doesn't
take all that much Python knowledge to build it, either.

So, whatever you could do in a language that allowed assignments
in expressions you could do in Python via that hard-to-read way, e.g.:

while [x for x in (getnext(23),)][0]:
    frobnicate(x)

is the Python (but NOT Pythonic) equivalent of C's

while(x=getnext(23))
    frobnicate(x);

That doesn't affect the "perfectly right" status of Aahz's quote, either.


Alex



From cabrera at wrc.xerox.com  Thu Apr 11 13:47:42 2002
From: cabrera at wrc.xerox.com (jic)
Date: 11 Apr 2002 17:47:42 GMT
Subject: ftplib and FTP Objects...
Message-ID: <a94i7u$6id$1@news.wrc.xerox.com>

Greetings!

I have tried all day to get the storbinary object to work and nothing.  

Here is the code:
  def FTPaFileToFP(self,mypath,FP_IP,user,pw):
    ftp = FTP(FP_IP)
    ftp.login(user,pw)   # connect to host, default port
    ftp.mkd('DOC.XSM') 
    ftp.pwd() 
    ftp.cwd('DOC.XSM') 
    WorkFiles = glob.glob (mypath + '\*.*')
    WorkFiles.sort()
    for file in WorkFiles:
      (junk,fname) = os.path.split(file)
      print fname
      f = open(file,'rb')
      ftp.storbinary('STOR ' + fname,f.read )
      f.close()
    ftp.quit()


Here is what I get:

Traceback (most recent call last):
  File "C:\myprograms\python\test\TIFs2DS.py", line 235, in ?
    c.FTPaFileToFP(mypath,'fpsdk','USER','pw')
  File "C:\myprograms\python\test\TIFs2DS.py", line 225, in FTPaFileToFP
    ftp.storbinary('STOR ' + fname,f.read )
  File "C:\Utils\Python21\lib\ftplib.py", line 371, in storbinary
    conn = self.transfercmd(cmd)
  File "C:\Utils\Python21\lib\ftplib.py", line 296, in transfercmd
    return self.ntransfercmd(cmd, rest)[0]
  File "C:\Utils\Python21\lib\ftplib.py", line 273, in ntransfercmd
    host, port = parse227(self.sendcmd('PASV'))
  File "C:\Utils\Python21\lib\ftplib.py", line 229, in sendcmd
    return self.getresp()
  File "C:\Utils\Python21\lib\ftplib.py", line 202, in getresp
    raise error_perm, resp
ftplib.error_perm: 500 Unrecognized command

I've tried setting up the buffer for read() and nothing.  Has anyone gotten the 
storbinary to work?

thanks much!

jos?



From mkayser at seas.upenn.edu  Mon Apr 29 12:39:57 2002
From: mkayser at seas.upenn.edu (Mike Kayser)
Date: Mon, 29 Apr 2002 12:39:57 -0400
Subject: caching of modules?
Message-ID: <aajsp7$b84$1@netnews.upenn.edu>

Hi,

I am writing a simple program (using the IDLE GUI and 2.2) that uses two
modules that I wrote. I'm in the debug phase right now, and I keep running
into the same problem: my module will have a runtime error, I go to the
module and correct the error, and then save and press ctrl-F5 (I don't even
know if the latter is necessary)... and then go back to my original program
and run again... and I get the SAME run-time error for the same line... but
I just changed the line so what it's now saying doesn't make sense.

for example:

(1) I get the error:
File "C:\Python22\project\optimize.py", line 435, in fminNCG
    print "FPRIME: " + fprime
TypeError: cannot concatenate 'str' and 'instance method' objects

(2) fprime is a function, so I change the line to
    print "FPRIME: " + str(fprime)

(3) I run again, and get the same runtime error, but with the new line. so
it will say:
  File "C:\Python22\project\optimize.py", line 435, in fminNCG
    print "FPRIME: " + str(fprime)
TypeError: cannot concatenate 'str' and 'instance method' objects


It looks to me like it's caching the module somewhere, but then when it
spits out error information it goes back to the file itself, even if it's
been modified.

Does anyone know if my hypothesis is right? If so, how can I force the
"module cache" to reload the updated file?

Thanks a million
Mike





From aahz at pythoncraft.com  Fri Apr 12 20:03:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Apr 2002 20:03:09 -0400
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <mailman.1018592665.14227.python-list@python.org> <3CB72CE6.DFAD5C10@python.org> <a97l29$hnv$0@216.39.172.122>
Message-ID: <a97sjt$dk2$1@panix1.panix.com>

In article <a97l29$hnv$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>
>ISTM what matters most of the time is how you can use it, not
>what its exact parentage is. E.g., people (ab)use type(x) to
>try to check if an argument to a function should be used one
>way or another way. I'm wondering if we're heading towards
>a hascapability(thing, cap) kind of general query, as a useful
>standard set of capabilities objects can be endowed with evolves?

http://www.python.org/peps/pep-0246.html

Needs much work before BDFL pronouncement.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From pinard at iro.umontreal.ca  Fri Apr  5 15:52:31 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 05 Apr 2002 15:52:31 -0500
Subject: The Topy project
Message-ID: <oqelhtq35c.fsf@carouge.sram.qc.ca>

Hi, people.  I merely feel like chatting a bit. :-)

Maybe nine months ago, I sent a message to this list about a `perl2py'
program I was working on.  After an intense burst, I put this `perl2py'
project on ice, partly because I was hitting some difficult problems, partly
because the program was significantly slower than I expected.  But I had
the vague intention of revisiting it one of these days...  Fun enough, a
Perl to Python compiler was announced earlier this week, yet the announce
was made on April 1st, and the description was giving simple things as
difficult, and difficult things as easy: I did not take it seriously! :-).

Two weeks ago, feeling overwhelmed by the tedious aspects of translating
a non-small Scheme project to Python, I decided to try automating the work
somewhat.  So, I took my unfinished `perl2py', ripped off the Perl processing
parts (while carefully putting them aside for later, who knows!), and added
code to read and process Scheme input.  I even tried to draw a clean line
between the front-end part for Scheme and the back-end part for Python.

I hit a few of the same walls than for `perl2py', but this time, I have been
able to traverse those I decided to, so far.  The fact that Scheme is, in
a sense, a less complex language than Perl, helps me at sorting out ideas.
Well, I would not say that Scheme is simple to process (at least for the
poor little me), but Perl is, euh, how should I put it, hum, "exhausting"...
A few walls remain for the Scheme to Python converter, and while none is
easy, I think I perceive angles that could successfully tackle each of them.

One sure thing is that Topy's speed is acceptable, even if not blazing. :-).
I wish myself good courage, as I would really like Topy by my side before
resuming the work on the Scheme biggie.  Moreover, if this tool ever works
satisfactorily for Scheme, this might invite us to work at other front-ends.

I might not be ready to publish even an alpha before a few weeks, but I'll
let you know when there is something (even dimly usable) to share.  My goal
is to produce legible and maintainable Python, _not_ at all to produce code
that might be blindly executed: serious post-edition and debugging are meant
to be unavoidable.  I roughly aim R4RS (without delays nor continuations),
Python approximations for various types and standard library functions,
and surely various limitations -- I'll make an inventory of them later :-).
Python 2.0 is needed to run the converter, the result is 1.5.2 compatible.

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





From andreas at mtg.co.at  Fri Apr 12 06:54:34 2002
From: andreas at mtg.co.at (Andreas Kostyrka)
Date: Fri, 12 Apr 2002 12:54:34 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <02041023313403.17153@arthur>
References: <mailman.1017466974.6751.python-list@python.org>
	<Ohxq8.45137$pT1.1283197@news1.tin.it>
	<20020410230535.674eef0a.andreas@mtg.co.at>
	<02041023313403.17153@arthur>
Message-ID: <20020412125434.5040465c.andreas@mtg.co.at>

On Wed, 10 Apr 2002 23:31:34 +0200
Alex Martelli <aleax at aleax.it> wrote:

> Super, but, I am not sure why is it a class method rather than a normal 
> instance method, accepting self and taking the table and column names
> from self (which of course could in turn implicitly delegate them to its
> class).  Wouldn't I normally be adding a row to a specific table, an
> instance of one of these classes, after all?  Maybe you can show us some
> more details because, so far, the superiority of a class method wrt a 
> normal instance method is still murky to me.  Thanks!
Well, basically it allows for alternative "constructors" by name instead of trying to overload on
set of arguments. Sometimes this is not needed. Sometimes this greatly clearifies the code.

For a table row there are 3 different ways to get it:
fetchByKeys        -> gets one instance by specifying the exact primary key.
fetchBySpec        -> gets a list of instances that satisfy the specification.
newRow             -> get a new row instance.

And yes, there are some warts, because there is no natural notation to get an "unbound" class method.
But Python relies upon exactly these to allow calling superclass methods.

Ideas would be something like:
superclass.classmethod.unbound(myclass,*args)
superclass.classmethod(myclass,*args)

Andreas




From netzapper at magicstar.net  Fri Apr  5 17:43:41 2002
From: netzapper at magicstar.net (A. Jones)
Date: Fri, 05 Apr 2002 22:43:41 GMT
Subject: file.getline()/cout problem
References: <3cae0463.46737479@mammoth.usenet-access.com>
Message-ID: <3cae287f.55982686@mammoth.usenet-access.com>

I apologize..

I seem to have posted to the wrong newsgroup... I'm truly sorry.


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From petera at weblink.com.au  Sun Apr  7 23:49:37 2002
From: petera at weblink.com.au (Peter Aylett)
Date: Mon, 8 Apr 2002 13:49:37 +1000
Subject: Win32 .pyc icon suggestion
Message-ID: <439969A9369AE4439637588ADB9B434767861B@weblink73.weblink.com.au>

Hi all,

Just a small idea:
I find that there is very little distinction between the icon for the
.py files and the .pyc files, which slows down the process of opening
and managing my program files.

Here's a replacement icon I suggest for the .pyc files that makes life
a bit easier, if anyone's interested.  (basically coloured yellow
instead of green).

rgds,
Peter Aylett




-------------- next part --------------
A non-text attachment was scrubbed...
Name: py-yellow.ico
Type: image/x-icon
Size: 766 bytes
Desc: py-yellow.ico
URL: <http://mail.python.org/pipermail/python-list/attachments/20020408/609adedb/attachment.bin>

From marklists at mceahern.com  Mon Apr 22 09:23:45 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 22 Apr 2002 08:23:45 -0500
Subject: Programming metaphors (Was: Programming as literature.)
In-Reply-To: <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEAHKMAA.marklists@mceahern.com>

[Jarno J Virtanen]
[...]
> When I think of the final products of software development and building
> construction I find it hard to compare them.  Yeah, sure there is some
> sort of an "architecture" that they're based on and they're both
> non-trivial combinations of some sort of building blocks.  But it's just
> that the software product seems so fragile.  [...]

Fragile is a good word.  I'd also say flexible.  In other words, soft.

// mark




From tim.one at comcast.net  Fri Apr 26 18:02:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 26 Apr 2002 18:02:52 -0400
Subject: [OT] What is Open Source?
In-Reply-To: <qPcy8kKkXMmP092yn@gnosis.cx>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEFIPBAA.tim.one@comcast.net>

[Tim]
> Me too, but at least one lawyer has told me it is a fantasy
> world, i.e. that an individual in the United States cannot put their
> work in the public domain:  you can't disclaim your copyright even if
> you want to, unless it's work done for the government and "they" put
> it in the public domain.

[Lulu of the Lotus-Eaters]
> All the Python PEP's seem to contain a disclaimer about being placed
> into the public domain.

Yup, and PEP 1 tells PEP authors to do this.

> Do you think Python has a problem because of this?

Not at all.  Perhaps it carries no legal force, perhaps it does:  IANAL, but
Python couldn't care less either way.  The only people it might bother are
the overly paranoid, fearing that if they reuse a PEP as if it were truly
public domain, one of the authors may come after them brandishing a lawyer's
opinion that the "public domain" declaration was without force.  Other
countries would probably laugh that out of court, but you never know what a
US court might decide.

> (I think some other Python-related documentation, but forget
> exactly which).

Code too.  For example, you can find lines like this in several modules in
the std distribution:

# Released to the public domain, by Tim Peters, 03 October 2000.

When I do that, I don't care whether it has legal force.  So long as people
*think* they know what "public domain" means, they're getting the message I
intended to convey.  That I can lull them into a false sense of security,
and sue them someday for believing me, is just gravy <wink>.

i'm-gonna-get-soooooooooooooooooo-rich-ly y'rs  - tim





From brian at sweetapp.com  Wed Apr 17 12:40:15 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 17 Apr 2002 09:40:15 -0700
Subject: Counterintuitive Python behavior
In-Reply-To: <15549.41721.616995.497834@12-248-41-177.client.attbi.com>
Message-ID: <00e401c1e62e$8d8c49b0$445d4540@Dell2>

Skip wrote:
>     Greg>     3 eq 3
>     Greg>     not (3 eq 4)
>     Greg>     a eq a            # PROVIDED a HAS A VALUE
>     Greg>     "foo" eq "foo"
>     Greg>     not ([] eq [])
>     Greg>     not ({} eq {})
>     Greg>     () eq ()
> 
> Guido's time machine strikes again:
> 
>     >>> 3 is 3
>     True

This is an implementation accident.

>>> a = 100000
>>> b = 100000
>>> a is b
0

I think that Greg's proposed operator would return true in this case.

Cheers,
Brian





From tismer at tismer.com  Mon Apr 29 06:42:08 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 29 Apr 2002 12:42:08 +0200
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org>
Message-ID: <3CCD2380.2040801@tismer.com>

Guido van Rossum wrote:
> Thomas Heller wrote:
> 
>>Do anything you want with my money (within resonable
>>bounds), but don't make me a member of another
>>organization. Please ;-)
> 
> 
> You can send your donation to the PSF.  Soon this will be tax-deductible
> (for US citizens).  http://www.python.org/psf/  Membership is free
> but only on invitation; but anyone can donate, much in the PSA spirit.
> We can talk about adding a starship account in the deal;

Oh, really? This is new to me. :-)

> Py subscriptions are not in the cards though.

I think you are slightly off-topic.
What we are talking about is a way to support PyZine,
while at the same time providing a missing structure
for Starship.

The overall positive reactions show that people are
missing something. Something simple.
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From anne_mc_murray at hotmail.com  Tue Apr 23 13:28:27 2002
From: anne_mc_murray at hotmail.com (Anne Mc Murray)
Date: 23 Apr 2002 10:28:27 -0700
Subject: FROM DIFFICULTIES WITH LEARNING AT SCHOOL, TO COMMUNICATION PROBLEMS IN THE FAMILY.,
Message-ID: <17c6c5e5.0204230928.1d051796@posting.google.com>

One of my sons is reluctant to produce any good school work, and has
turned into a frightful headache for the family's, despite being very
intelligent and creative, and a good reader too.
I've tried other alternatives, with psychologists, and
psychopedagoges, even private teachers, but nothing seems to work.
Please, has anyone a suggestion to make?

Anne


From robin at jessikat.fsnet.co.uk  Wed Apr 24 07:14:30 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 24 Apr 2002 12:14:30 +0100
Subject: Python call tree tool?
In-Reply-To: <31575A892FF6D1118F5800600846864DCBD042@intrepid>
References: <31575A892FF6D1118F5800600846864DCBD042@intrepid>
Message-ID: <c952IKAWOpx8EwR2@jessikat.demon.co.uk>

In message <31575A892FF6D1118F5800600846864DCBD042 at intrepid>, Simon
Brunning <SBrunning at trisystems.co.uk> writes
>> From:        Robin Becker [SMTP:robin at jessikat.fsnet.co.uk]
>> I searched in vain for a tool that can create a call tree or  inverted
>> call tree for python modules/packages. It seems inconceivable that the
>> Python community hasn't created such a tool so I must be looking in the
>> wrong places. 
>> 
>> Any one know where such a beast is?
> 
>Might it be worth looking at py2exe and/or Installer? I would have thought
>that these tools must create a call tree for internal use.
>
>Cheers,
>Simon Brunning
>TriSystems Ltd.
>sbrunning at trisystems.co.uk
....
it seems that both of these concentrate on finding imports ie are too
coarse to locate individual functions/methods which are dead or live. 
-- 
Robin Becker




From tom at studentuniverse.com  Mon Apr 29 18:37:19 2002
From: tom at studentuniverse.com (Russo, Tom)
Date: Mon, 29 Apr 2002 18:37:19 -0400
Subject: (no subject)
Message-ID: <AC310CEA8E38D611BE2F00D0B74A0DD6073046@sumailb01.studentuniverse.com>

> This PEP smacks of political requirements not practical ones. The
> builtin name space is only crowded if we want to use the names for
> something else.

I think that this hits the nail right on the head.  One of the reasons I
like python is that it is friendly to functional programmers like myself.
The goal of python _should not_ be to weed out "non-standard" features,
which seems to be the goal of this PEP.  What python _should_ try to do (in
my opinion at least) is take the best of all worlds, to the extent possible.
The functions at issue here (with the possible exception of input) represent
the best of functional programming integrated into python.  The fact that
certain people never learned to program in the function idiom is no basis
for removing them.

_t




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Wed Apr 17 15:41:34 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 17 Apr 2002 19:41:34 GMT
Subject: Unicode name questions
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com> <mailman.1018994314.14044.python-list@python.org> <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de> <Xns91F325AA52316RASXnewsDFE1@130.133.1.4>
Message-ID: <a9kj5e$ka3$1@nyheter.chalmers.se>

In article <Xns91F325AA52316RASXnewsDFE1 at 130.133.1.4>,
Philip Swartzleonard  <starx at pacbell.net> wrote:
>Martin v. Loewis || Wed 17 Apr 2002 01:25:23a:
>
>> Skip Montanaro <skip at pobox.com> writes:
>> 
>>> "Lambda" has been spelled with a "b" as long as I can remember.  
>> 
>> I think the Unicode charts use an ASCII transcription of the native
>> pronouncation of the letters, if possible. In any case, the Unicode
>> names for the characters are "official" in the context of Unicode.
>> The Unicode consortium refrains from renaming them. Between Unicode
>> 2.0 and Unicode 3.2, no character was renamed.
>> 
>> [Etc Etc Greekness]
>
>Hm. I was browing through some of their code charts for the heck of it, 
>and I noticed that they listed certain kata/hiragana characters's primary 
>names as SI, TI, TU, etc. I've listened to a considerable amount of 
>japanese media, and AFAICT the correct romanization of those sounds would 
>be SHI, CHI, TSU, etc.

Correct from who's point of view? Every language has its own transcriptions
of foreign scripts, and quite often several. In Sweden we have 2 major ways
of transcribing Cyrillic, and several minor ones.

You can be sure that a lot of effort has gone into making politically
acceptable compromises concerning the naming of every code in the Unicode
chart. Some are probably still the subject of contention. This is the
focus of very strong feelings among a lot of people in fields outside
the Computer Science one. Unless a person wants to create an artifical
load for burning time, that person should stay well clear of these issues
and accept what the standardisation people have arrived at as immovable
truths.

Jacob Hall?n

-- 


From lumber at jack.com  Tue Apr  9 19:03:52 2002
From: lumber at jack.com (Lumberjack)
Date: 9 Apr 2002 23:03:52 GMT
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org> <a8vnmr$r78$1@panix1.panix.com>
Message-ID: <Xns91EBA33E75BDlumberjack@216.148.53.96>

aahz at pythoncraft.com (Aahz) wrote:
> Unfortunately, I have to say that the stridency you exhibited in this
> thread (to my startlement) made me feel a bit that you *were* trying to
> score a point.  I'm not talking about your initial long post, but
> subsequent ones.

This from a person who is constantly trying to score points on Usenet. A 
check of google shows you have an inflated notion of yourself. Maybe you 
should stop measuring others by your own limited character.


From mlh at vier.idi.ntnu.no  Wed Apr  3 06:42:09 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Wed, 3 Apr 2002 11:42:09 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com>  <gIpq8.44728$pT1.1266378@news1.tin.it>
Message-ID: <slrnaalqkh.fa8.mlh@vier.idi.ntnu.no>

In article <gIpq8.44728$pT1.1266378 at news1.tin.it>, Alex Martelli wrote:
>Jeff Shannon wrote:
>        ...
>> (FWIW, this was the first time I've seen the Something vs.
>> Nothing distinction pointed out explicitly.  I think that this
>> explanation of how Python determines truth is exceptionally
>> clear, 
>
>Hear, hear.
>
>> and I think that the tutorial should be modified to
>> include this terminology.)
>
>Seconded.  _I_ am going to steal it (if I get Laura's permission,
>and if the PEP is rejected) for the Nutshell, for sure...

Yup. An excellent distinction. I'm tempted to re-edit the chapter on
conditionals in (the upcoming) Instant Python...

Laura: I think your post was brilliant.

>
>Alex
>

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From mwh at python.net  Tue Apr  9 08:13:35 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 9 Apr 2002 12:13:35 GMT
Subject: Newbie Problem Building 2.2 On SGI
References: <mailman.1018283656.17495.python-list@python.org>
Message-ID: <uwuvh9ibu.fsf@python.net>

Jeff Smith <smithj at kryos.colorado.edu> writes:

> I have installed both 2.2 and 2.2.1.c2 on an SGI O2 with IRIX6.5. 

I had reports of 2.2.1c1 working here (and 2.2.1c2 has no relavent
changes).

> I cannot import the modules array, math, and struct (out of the
> limited list of modules that I've tried to import). I have tried
> both the SGI C compiler and the GNU C compiler.

What goes wrong?  I probably can't help, but others can, and error
messages are easier to deal with than telepathy.

Cheers,
M.

-- 
  In that case I suggest that to get the correct image you look at
  the screen from inside the monitor whilst standing on your head.  
               -- James Bonfield, http://www.ioccc.org/2000/rince.hint


From BgPorter at NOartlogicSPAM.com  Thu Apr 11 10:08:39 2002
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Thu, 11 Apr 2002 14:08:39 GMT
Subject: Indenting lines 400..457 help
References: <a929tk$ik2$1@pcls4.std.com> <Fc2t8.90644$w7.7532194@bin6.nnrp.aus1.giganews.com> <a92f90$i57$1@pcls4.std.com>
Message-ID: <HNgt8.335984$Gf.29623810@bin2.nnrp.aus1.giganews.com>

"Ned Harvey" <rahvee at yaahhoo.com> wrote in message
news:a92f90$i57$1 at pcls4.std.com...
> By using the >> or << method below, vi uses the tab-characters, at least
in
> my account.  Is there any way you're aware of, that I can specify the
> characters to use?
>
See Geoff Gerrietts' response elsewhere.
The vim help sez:
     *'expandtab'* *'et'* *'noexpandtab'* *'noet'*
'expandtab' 'et' boolean (default off)
   local to buffer
   {not in Vi}
 In Insert mode: Use the appropriate number of spaces to insert a
 <Tab>.  Spaces are used in indents with the '>' and '<' commands and
 when 'autoindent' is on.  To insert a real tab when 'expandtab' is
 on, use CTRL-V<Tab>.  See also |:retab| and |ins-expandtab|.
 NOTE: This option is reset when 'compatible' is set.






From jb at cascade-sys.com  Mon Apr 29 18:08:54 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 15:08:54 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch> <MNiz8.32691$8D3.925320@news1.tin.it>
Message-ID: <3CCDC476.A5AC4144@cascade-sys.com>

Alex Martelli wrote:

> I'm surprised you miss the main advantage -- 0666, the Number
> of the Beast in octal, translates to decimal 438, the A.D. in which
> Emperor Theodosius II promulgated the Theodosian Code, the
> wildly antisemitic laws which, among other things, forbid Christians
> to even give alms to Jews, forbid Jews to hold State office or build
> synagogues, condemned to death any Jew converting a Christian.
>
> Surely this is therefore a number which is often handy in the everyday
> practice of programming, and even thinking of eventually making a
> non-builtin of the precious Python function that helps us work with a
> numbering system which has such wondrous properties is unholy.

It also opens your Unix files up to abuse by others.

;o)

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From geoff at gerrietts.net  Mon Apr  1 16:09:25 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 1 Apr 2002 13:09:25 -0800
Subject: GUI programming
In-Reply-To: <XP2q8.8178$vm6.1392106@ruti.visi.com>
References: <mailman.1017687607.4322.python-list@python.org> <XP2q8.8178$vm6.1392106@ruti.visi.com>
Message-ID: <20020401210925.GC6604@isis.gerrietts.net>

Quoting Grant Edwards (grante at visi.com):
> In article <mailman.1017687607.4322.python-list at python.org>, Gustavo Cordova wrote:
> 
> >> [..references to building wxGTK problems..]
> >> All in all, this can be very frustrating to new users.
> > 
> > Extremely.
> > 
> > I gave up after the n-th iteration.
> 
> It took me 5 or 6 tries, but I did get it to build and install.
> One of the things I found odd was that by default it would try
> to build stuff (GL, etc.) for which my system had no support.

I tried to build from source the first time, and had no luck at all. I
had lots of disagreements between versions that didn't make much sense
to me. I ended up installing a version of Mesa which really screwed up
all sorts of other dependencies on my system and still, I got no love.
I would venture that I dropped three to four hours trying to get all
the dependencies and install everything -- much of that effort appears
to have been unnecessary.

After this glaring failure, I installed from RPM, which was pretty
straightforward, and was able to develop in wxWindows with no problem.

Later, I built from CVS. I was able to do it on my stock RH7.2 system
after taking care to configure everything properly, though it took me
a few passes to get the make to go through to the end. Everything
compiled fine, including the GL support for wxGTK. It went smoothly
enough that I was surprised at my earlier failures.

I don't think that installing wxWindows on an RPM-based Linux system
is really terribly difficult, but at the same time, it's a dependency
that I would prefer not to create if I can avoid it.

These days, I develop for pygtk using a 1.5.2 interpreter. It's not
perfect, and the windows portability is a long way from perfect, but
it also doesn't have a dependency-management issue for most of my
audience -- RedHat 6.2+ users almost universally have python 1.5.2 and
pygtk pre-installed, and for most others, it's a very simple addition.

That said, I'm not sure my solution is the best. It just happens to
work for me.

Thanks,
--G.

-- 
Geoff Gerrietts             "People talk fundamentals and superlatives 
<geoff at gerrietts net>     and then make some changes of detail." 
http://www.gerrietts.net                  --Oliver Wendell Holmes Jr



From tejarex at yahoo.com  Tue Apr  9 12:24:11 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 09 Apr 2002 16:24:11 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no>
Message-ID: <LAEs8.299757$Gf.27435152@bin2.nnrp.aus1.giganews.com>

"Syver Enstad" <syver-en+usenet at online.no> wrote in message
news:upu18x4h3.fsf at online.no...
> Am I the only one in the entire world that thinks '\t'.join(mylist)
is
> elegant and natural?

no





From eric.brunel at pragmadev.com  Mon Apr 15 05:23:34 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 15 Apr 2002 11:23:34 +0200
Subject: Python installation on a Windows shared disk
Message-ID: <a9e5u2$1j0u$1@norfair.nerim.net>

Hi all,

I'm trying to do an installation of Python 2.1 on a shared Windows 2K disk.
People should then be able to use this installation from all PCs seeing this
disk. Since I didn't find any hint on how to do that in the Python distro, I
searched for it on the Python site and found:
http://starship.python.net/crew/jjkunce/netinstall/index.html
Unfortunately, it was written a long time ago and does not seem to work
anymore (some scripts have moved, so some imports fail, and even when
correcting this, the checking phase at the end fails, complaining about
invalid registered directories)

Is there a more "standard" way to do a shared installation for Win2K? Or
does anybody have done that quite recently and can give me some hints on
what to do, considering I still haven't understood what on earth a
"registry" is (and not really willing to learn anyway... ;-)? The objective
would be to have a single script that users could run from their PCs and
that would do all the installing job.

TIA
--
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com




From paul at fxtech.com  Sat Apr 27 15:44:17 2002
From: paul at fxtech.com (Paul Miller)
Date: Sat, 27 Apr 2002 14:44:17 -0500
Subject: handling different arguments depending on first argument
In-Reply-To: <20020427154240.F1588@eecs.tufts.edu>
References: <3CCAF77B.3BEBF259@fxtech.com>
 <3CCAF77B.3BEBF259@fxtech.com>
Message-ID: <5.1.0.14.2.20020427144305.044a1a80@cedar.he.net>

>   Sure. Try something like this:
>
>      def my_func (*args):

Thanks! I had actually JUST figured this out. It's a little weird when 
combined with keyword args, but putting it last does the trick.

Cheers!


-- 
Paul T. Miller | paul at fxtech.com | http://www.fxtech.com





From johnroth at ameritech.net  Mon Apr 22 08:59:51 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 22 Apr 2002 08:59:51 -0400
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <uc829fhdpe3581@news.supernews.com>

"Gon?alo Rodrigues" <op73418 at mail.telepac.pt> wrote in message
news:dit7cuo1nguiha34nq80m9q39igpjlr70d at 4ax.com...
> Just some clarifications.
>
> On Mon, 22 Apr 2002 06:10:02 GMT, Tim Daneliuk <tundra at tundraware.com>
> wrote:
>
>
> You should understand that this is a philosophical viewpoint. The
> mathematical field is roughly divided in the Platonist field, those
who
> believe that mathematical objects exist somewhere out there in space
(to
> quote Sun Ra) in a Wonderful World of Platonic Ideas, and the
formalist
> field of those who believe mathematics is essentialy a game we play on
> paper with symbols with no "real" meaning attached to it besides the
> convenience in describing natural phenomena.

That's my understanding of the philosophy as well. I normally
take the formalist side.

John Roth




From emile at fenx.com  Wed Apr  3 08:07:59 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 03 Apr 2002 13:07:59 GMT
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net> <mailman.1017810728.14199.python-list@python.org> <a8e7gv$fut$1@nntp9.atl.mindspring.net>
Message-ID: <P8Dq8.157972$Yv2.49617@rwcrnsc54>

Andrew Dalke
> What I didn't think about was Gerhard's "counter.cnt" mechanism.
> (My mind's still stuck in the 1.4 days :)  I really like that, and
> think I'll use it more often in the future.
>

Was this a post?  If so, I couldn't locate it.  Does it do something
like:

>>> def counter():
...     try: counter.cnt
...     except: counter.cnt= 0
...     counter.cnt += 1
...     return counter.cnt
...
>>> counter()
1
>>> counter()
2

--

Emile van Sebille
emile at fenx.com

---------



From b_mcerlean at yahoo.com  Thu Apr 25 12:20:20 2002
From: b_mcerlean at yahoo.com (Brian McErlean)
Date: 25 Apr 2002 09:20:20 -0700
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org> <aa1fe9$mn7$1@peabody.colorado.edu>
Message-ID: <f040c4f9.0204250820.2c6581d5@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aa1fe9$mn7$1 at peabody.colorado.edu>...
> > Second, math *is* an attempt to describe the physical world - that's what
> > makes it useful. Case in point: calculus, which was invented/discovered
> > specifically to deal with and describe the motion of physical bodies. Had
> > it not been useful in describing such motion, it would have been tossed
> > out, or at least not so widely accepted. Math is useful because of its
> > relevance to the world around us.
> 
> Sorry but no. The fact that certain physical problems may have served as 
> inspiration for some mathematical developments can be considered to be purely 
> accidental. The foundational structure of mathematics is entirely one of 
> logical self consistency, and massive amounts of work was done particularly 
> in the 20th century to build that framework. While it is true that up to the 
> 19th century most (if not all) significant advances in math were prompted by 
> challenges from the physical sciences, the work of the formalists in the 20th 
> century showed that mathematics could be built entirely on an abstract, 
> logical foundation.

But that abstract, formal foundation is built upon axioms that are
direct analogues of real-world properties.  The mathematical concept
of "addition" is directly mappable to the combining groups of discrete
objects.  All of maths is based on a framework which "just happens" to
be isomorphic to counting, measuring and comparing real-world objects.

We build more complex theorems on top of these, and way of combining
rules, (identical to logical rules that seem to hold in the real world
too.)  But all maths is based around a system that makes it applicable
in the real world, and this is no accident.

It would be possible to design a system with rules completely
different from anything "real", but no-one does this because its a
fundamentally pointless task.  Instead, mathematicians use the
existing framework, and while they may work in areas that seem obscure
and impractical, their work is still on some level relevant to a
similar "real" situation.  Since all our concepts are to some degree
shaped by the real world, actually creating such a system may be
harder than it sounds.

> That is not to say that it's not _useful_ to have a reference to the physical 
> world both for inspiration and for applications, but there's a crucial 
> difference between that and saying that 'math *is* an attempt to describe the 
> physical world'.

Real-world relevance may not be the _reason_ for math, but the
relevance is nonetheless there.  At its lower levels, math is an
attempt to descrive the world, and all else is build on top of this.

> What I do find most fascinating is how many seemingly super-abstract results 
> of mathematics turn up (sometimes decades later after they are found) having 
> physical implications which noone ever suspected. That 'mysterious 
> mathematical structure of the physical world' is one of the most beautiful, 
> in my opinion unanswerable questions of the foundation of science.

Beautiful yes, but I don't think unanswerable.  Its not really so
surprising that "real-world" relevance pops up in odd corners of math.
 I'd argue that the relevance was always there - our body of
mathematical knowledge is designed that way.  Perhaps its more
appropriate to speak of the 'mysterious physical structure of the
mathematical world'

> Cheers,
> 
> f.

Brian.


From chooper at umich.edu  Tue Apr  2 22:29:09 2002
From: chooper at umich.edu (CGH)
Date: Tue, 02 Apr 2002 22:29:09 -0500
Subject: Newbie: Classes and Lists
Message-ID: <8Guq8.1209$zT5.41108@news.itd.umich.edu>

Hi,

I am very new to Python and have a very simple question. I want to create
a 'chromosome' class for a genetic algorithm program. Essentially I just 
want a class that contains a list of n bits. That part seemed easy: here is 
what I tried:

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

import random

class Chromosome:

    data = []
    
    def __init__(self, length):
        for bit in range(length):
            self.data.append(random.randrange(0,2,1))

    def display(self):
        print self.data


population = []

for i in range(100):
    x = Chromosome(20)
    population.append(x)

population[50].display()

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

I want to define a population of 100 chromosomes. But instead of a list of 
Chromosome objects I seem to get one big (100 x 20) list of 0s and 1s. 
Clearly I've misunderstood something fundamental here. Any help would be 
appreciated.

Cameron


From barronsnyder at yahoo.com  Thu Apr 18 15:34:09 2002
From: barronsnyder at yahoo.com (Barron)
Date: 18 Apr 2002 12:34:09 -0700
Subject: PyQT, how to compile SIP on MAC? (long)
References: <3C9A62C2.ABA9AE5D@gmx.de> <mailman.1016790154.12480.python-list@python.org>
Message-ID: <de404184.0204181134.6bd4bc2c@posting.google.com>

Phil Thompson <phil at river-bank.demon.co.uk> wrote in message news:<mailman.1016790154.12480.python-list at python.org>...
> Ingo Linkweiler wrote:
> > 
> > Hi,
> > 
> > I am willing to use PyQt on Mac OS X
> > 
> > But I was not able to compile SIP --
> > http://www.river-bank.demon.co.uk/software/sip-3.0.tar.gz
> > 
> > Are there any hints you can give to compile SIP with success?
> > 
> > With windows and linux it works fine.
> 
> I've had reports of SIP building fine, but PyQt having problems with
> pyuic.
> 
> I can't test any of this myself because I don't have the tools (cue
> another begging letter for somebody to donate me a Mac and a development
> environment).
> 
> Phil


I, too, would love to get PyQt running on OS X! Any help is
appreciated.

Phil, I wish I could ship you a machine to do the work, but until I
win the lottery, my resources are quite limited <g>. However, I would
be willing to contribute a few dollars toward the effort....

It seems like the first step is compiling SIP.
So, I ran the .configure script. I received a few "no's" in the
output. Is this a problem? (full output at the end of this message) It
didn't seem to be, so I ran the "make check" to run the checks. (full
output at the end of this message) The only errors seemed to come at
the end:

...
/usr/bin/libtool: internal link edit command failed
make[1]: *** [libsip.la] Error 1
make: *** [check-recursive] Error 1

Do these errors prevent me from doing a "make install"?

Thanks,
Barron

Output from running ./configure and make check:

[localhost:/Applications/sip-3.1] barron% sh ./configure
checking for a BSD-compatible install... /usr/bin/install -c
checking whether build environment is sane... yes
checking for gawk... no
checking for mawk... no
checking for nawk... no
checking for awk... awk
checking whether make sets ${MAKE}... yes
checking for a BSD-compatible install... /usr/bin/install -c
checking for gcc... no
checking for cc... cc
checking for C compiler default output... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether cc accepts -g... yes
checking for style of include used by make... GNU
checking dependency style of cc... gcc
checking for g++... no
checking for c++... c++
checking whether we are using the GNU C++ compiler... yes
checking whether c++ accepts -g... yes
checking dependency style of c++... gcc
checking for bison... bison -y
checking for flex... flex
checking for yywrap in -lfl... no
checking for yywrap in -ll... yes
checking lex output file root... lex.yy
checking whether yytext is a pointer... yes
checking build system type... powerpc-apple-darwin5.3
checking host system type... powerpc-apple-darwin5.3
checking for ld used by GCC... /usr/bin/ld
checking if the linker (/usr/bin/ld) is GNU ld... no
checking for /usr/bin/ld option to reload object files... -r
checking for BSD-compatible nm... /usr/bin/nm -p
checking whether ln -s works... yes
checking how to recognise dependant libraries... file_magic Mach-O
dynamically linked shared library
checking command to parse /usr/bin/nm -p output... ok
checking how to run the C preprocessor... cc -E
checking for ANSI C header files... yes
checking for sys/types.h... yes
checking for sys/stat.h... yes
checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking dlfcn.h usability... no
checking dlfcn.h presence... no
checking for dlfcn.h... no
checking for ranlib... ranlib
checking for strip... strip
checking for objdir... .libs
checking for cc option to produce PIC... -fno-common
checking if cc PIC flag -fno-common works... yes
checking if cc static flag -static works... no
checking if cc supports -c -o file.o... yes
checking if cc supports -c -o file.lo... yes
checking if cc supports -fno-rtti -fno-exceptions... yes
checking whether the linker (/usr/bin/ld) supports shared libraries...
yes
checking how to hardcode library paths into programs... unsupported
checking whether stripping libraries is possible... no
checking dynamic linker characteristics... darwin5.3 dyld
checking if libtool supports shared libraries... yes
checking whether to build shared libraries... yes
checking whether to build static libraries... no
creating libtool
checking for ANSI C header files... (cached) yes
checking errno.h usability... yes
checking errno.h presence... yes
checking for errno.h... yes
checking stddef.h usability... yes
checking stddef.h presence... yes
checking for stddef.h... yes
checking for stdlib.h... (cached) yes
checking for string.h... (cached) yes
checking for cc option to accept ANSI C... none needed
checking for an ANSI C-conforming const... yes
checking for size_t... yes
checking for stdlib.h... (cached) yes
checking for working malloc... yes
checking for working strtod... yes
checking for vprintf... yes
checking for _doprnt... no
checking for strchr... yes
checking for strdup... yes
checking for strrchr... yes
checking for strtol... yes
checking for the Qt meta object compiler...
/Applications/QtMacEval/bin/moc
checking for the Qt include files... /Applications/QtMacEval/include
checking for the Qt library files... -L/Applications/QtMacEval/lib
-lqt-mt
checking for X... libraries /usr/X11R6/lib, headers /usr/X11R6/include
checking for gethostbyname... yes
checking for connect... yes
checking for remove... yes
checking for shmat... yes
checking for IceConnectionNumber in -lICE... yes
checking for -fno-exceptions... yes
checking for python... /sw/bin/python
checking Python version... 2.2
checking where the Python sip module will be installed...
/sw/lib/python2.2/site-packages
checking for Python include files... /sw/include/python2.2
configure: creating ./config.status
config.status: creating Makefile
config.status: creating sip/Makefile
config.status: creating siplib/Makefile
config.status: creating sip.spec
config.status: creating config.h
config.status: executing depfiles commands


[localhost:/Applications/sip-3.1] barron% make check
Making check in sip
source='main.c' object='main.o' libtool=no \
depfile='.deps/main.Po' tmpdepfile='.deps/main.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f main.c || echo
'./'`main.c
source='parser.c' object='parser.o' libtool=no \
depfile='.deps/parser.Po' tmpdepfile='.deps/parser.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f parser.c || echo
'./'`parser.c
source='lexer.c' object='lexer.o' libtool=no \
depfile='.deps/lexer.Po' tmpdepfile='.deps/lexer.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f lexer.c || echo
'./'`lexer.c
source='transform.c' object='transform.o' libtool=no \
depfile='.deps/transform.Po' tmpdepfile='.deps/transform.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f transform.c ||
echo './'`transform.c
source='gencode.c' object='gencode.o' libtool=no \
depfile='.deps/gencode.Po' tmpdepfile='.deps/gencode.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f gencode.c || echo
'./'`gencode.c
source='heap.c' object='heap.o' libtool=no \
depfile='.deps/heap.Po' tmpdepfile='.deps/heap.TPo' \
depmode=gcc /bin/sh ../depcomp \
cc -DHAVE_CONFIG_H -I. -I. -I..     -O2 -c `test -f heap.c || echo
'./'`heap.c
/bin/sh ../libtool --mode=link cc  -O2   -o sip  main.o parser.o
lexer.o transform.o gencode.o heap.o -ll
mkdir .libs
cc -O2 -o sip main.o parser.o lexer.o transform.o gencode.o heap.o 
-ll
Making check in siplib
source='siplib.c' object='siplib.lo' libtool=yes \
depfile='.deps/siplib.Plo' tmpdepfile='.deps/siplib.TPlo' \
depmode=gcc /bin/sh ../depcomp \
/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I.. 
-DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT -I/Applications/QtMacEval/include
 -I/usr/X11R6/include -I/sw/include/python2.2   -O2 -c -o siplib.lo
`test -f siplib.c || echo './'`siplib.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT
-I/Applications/QtMacEval/include -I/usr/X11R6/include
-I/sw/include/python2.2 -O2 -c siplib.c -Wp,-MD,.deps/siplib.TPlo 
-fno-common -DPIC -o siplib.lo
source='qtlib.cpp' object='qtlib.lo' libtool=yes \
depfile='.deps/qtlib.Plo' tmpdepfile='.deps/qtlib.TPlo' \
depmode=gcc /bin/sh ../depcomp \
/bin/sh ../libtool --mode=compile c++ -DHAVE_CONFIG_H -I. -I. -I.. 
-DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT -I/Applications/QtMacEval/include
 -I/usr/X11R6/include -I/sw/include/python2.2  -fno-exceptions  -O2 -c
-o qtlib.lo `test -f qtlib.cpp || echo './'`qtlib.cpp
c++ -DHAVE_CONFIG_H -I. -I. -I.. -DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT
-I/Applications/QtMacEval/include -I/usr/X11R6/include
-I/sw/include/python2.2 -fno-exceptions -O2 -c qtlib.cpp
-Wp,-MD,.deps/qtlib.TPlo  -fno-common -DPIC -o qtlib.lo
source='objmap.c' object='objmap.lo' libtool=yes \
depfile='.deps/objmap.Plo' tmpdepfile='.deps/objmap.TPlo' \
depmode=gcc /bin/sh ../depcomp \
/bin/sh ../libtool --mode=compile cc -DHAVE_CONFIG_H -I. -I. -I.. 
-DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT -I/Applications/QtMacEval/include
 -I/usr/X11R6/include -I/sw/include/python2.2   -O2 -c -o objmap.lo
`test -f objmap.c || echo './'`objmap.c
cc -DHAVE_CONFIG_H -I. -I. -I.. -DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT
-I/Applications/QtMacEval/include -I/usr/X11R6/include
-I/sw/include/python2.2 -O2 -c objmap.c -Wp,-MD,.deps/objmap.TPlo 
-fno-common -DPIC -o objmap.lo
if test -n "/Applications/QtMacEval/bin/moc"; then \
        /Applications/QtMacEval/bin/moc -o moc_sipQt.cpp ./sipQt.h; \
else \
        touch moc_sipQt.cpp; \
fi
source='moc_sipQt.cpp' object='moc_sipQt.lo' libtool=yes \
depfile='.deps/moc_sipQt.Plo' tmpdepfile='.deps/moc_sipQt.TPlo' \
depmode=gcc /bin/sh ../depcomp \
/bin/sh ../libtool --mode=compile c++ -DHAVE_CONFIG_H -I. -I. -I.. 
-DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT -I/Applications/QtMacEval/include
 -I/usr/X11R6/include -I/sw/include/python2.2  -fno-exceptions  -O2 -c
-o moc_sipQt.lo `test -f moc_sipQt.cpp || echo './'`moc_sipQt.cpp
c++ -DHAVE_CONFIG_H -I. -I. -I.. -DQT_THREAD_SUPPORT -DSIP_QT_SUPPORT
-I/Applications/QtMacEval/include -I/usr/X11R6/include
-I/sw/include/python2.2 -fno-exceptions -O2 -c moc_sipQt.cpp
-Wp,-MD,.deps/moc_sipQt.TPlo  -fno-common -DPIC -o moc_sipQt.lo
/bin/sh ../libtool --mode=link c++ -fno-exceptions  -O2   -o libsip.la
-rpath /sw/lib/python2.2/site-packages -version-info 8:0:0
-L/Applications/QtMacEval/lib -lqt-mt  -L/usr/X11R6/lib  -lSM -lICE
-lX11  siplib.lo qtlib.lo objmap.lo  moc_sipQt.lo
mkdir .libs
rm -fr .libs/libsip.la .libs/libsip.* .libs/libsip.*
(cd . && ln -s siplib.lo siplib.o)
(cd . && ln -s qtlib.lo qtlib.o)
(cd . && ln -s objmap.lo objmap.o)
(cd . && ln -s moc_sipQt.lo moc_sipQt.o)
../libtool: parse error: condition expected: xno = [3183]
c++ -dynamiclib -flat_namespace -undefined suppress -o
.libs/libsip.8.0.0.dylib  siplib.lo qtlib.lo objmap.lo moc_sipQt.lo 
-L/Applications/QtMacEval/lib -lqt-mt -L/usr/X11R6/lib -lSM -lICE
-lX11 -lc -install_name 
/sw/lib/python2.2/site-packages/libsip.8.dylib -compatibility_version
9 -current_version 9.0
ld: multiple definitions of symbol _sipQObjectClass
siplib.lo definition of _sipQObjectClass in section (__DATA,__common)
qtlib.lo definition of _sipQObjectClass in section (__DATA,__common)
objmap.lo definition of _sipQObjectClass in section (__DATA,__common)
/usr/bin/libtool: internal link edit command failed
make[1]: *** [libsip.la] Error 1
make: *** [check-recursive] Error 1
[localhost:/Applications/sip-3.1] barron%


From phr-n2002a at nightsong.com  Tue Apr  2 18:52:38 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 15:52:38 -0800
Subject: A faster Python?
References: <a8dcqr$p29$1@laurel.tc.umn.edu>
Message-ID: <7xn0wl3bg9.fsf@ruckus.brouhaha.com>

Scott Hennesy <henne021 at tc.umn.edu> writes:
> For what I do, 90% of the time Python is fast enough. But for that
> 10% of the time I find myself wishing for a faster Python.
> 
> What's the consensus out there? Do Psyco, NumPy, and SciPy fulfill
> the need for speed or is there a desire for a faster Python?

NumPy and SciPy are just for numerical computations, but they're
indispensible in those areas.

Psyco is a research project.  It would be great if it were developed
further, but in its current form it's not really useful.

> If so, what are the requirements for a faster Python? How much
> faster does it have to be? How much compatibility does it need with
> Python? How much resources can be devoted to optimization? Easy
> example: 20% faster, no compatibility, infinite resources. Hard
> example: Faster than optimized C, total compatibility, no resources
> ;-)

If you can get a 20% improvement without a lot of effort, go for it.
I think really big speedups will require a new compiler/interpreter
implementation (lots of work) and some changes to the language.

I don't think Python can be made to run faster than optimized C, but I
think it should evolve towards being able to run as fast as compiled
Lisp or Scheme.  

> I've considered doing something about maybe a half dozen
> times. Maybe with Python 2.2 the time is ripe?

2.2 isn't the time for sweeping changes but the issues should be kept
in mind constantly.  For now, I'd say it's a problem that the language
definition is so closely bound up with details of the CPython
implementation.  All the hacks like representing instance variables in
a __dict__ dictionary, or the particular layout of stack frame
objects, should be documented as specifics of a particular
implementation rather than as how the language works (this goes for
most everything surrounded by double underscores except for class
methods like __init__).  This is mostly a matter of how the
documentation is written, not the code.

If a new implementation gives much better performance (like a 5-10x
speedup, which I think is realizable) and stays reasonably compatible
for normal applications, but requires big changes in some low level
tools like debuggers, I think that's ok.

For now, there may not be the resources to implement such a thing, but
it's appropriate to begin paving the way for it to happen someday, by
trying not to add too many implementation-specific requirements to
the language definition.


From mscheuner at hfx.eastlink.ca  Wed Apr  3 16:00:18 2002
From: mscheuner at hfx.eastlink.ca (Marc Scheuner)
Date: 3 Apr 2002 13:00:18 -0800
Subject: NTLM authorization - compile into EXE
Message-ID: <9b309481.0204031300.70ce3c63@posting.google.com>

Folks,

I'm faced with the need to deal with an MS ISA proxy server. This
proxy users NTLM Authorization, which basically no internet tool
supports - except MS IE......

I downloaded the Python 1.5.2 source from here:
http://www.geocities.com/rozmanov/ntlm

and I also downloaded Python 1.5.2 itself - everything works okay.

But under Windows, having an extra command line window open just for
this NTLM proxy thingie is a bit "annoying" - so I thought I'd compile
it into an EXE, using Py2EXE. Downloaded all that stuff, too, and went
through all the steps recommended (distutils stuff and the like) - but
I still can't seem to compile that Python code into a Windows
EXEcutable.

The error message I get is very cryptic - at least to me as a Phyton
newbie. Don't really know what it tries to tell me....

---------------
File "C:\Program Files\Python\py2exe\resources\VersionInfo.py", line
56, in ?
  raise ImportError, "Could not import StringTables, no unicode
available"
ImportError: Could not import StringTables, no unicode available
---------------

Can anyone help? Or could someone do the Py2EXE step for me, and send
me the resulting EXE (in a ZIP - a .EXE attachment won't get through
the firewall and stuff).

Thanks!
Marc (mailto:mscheuner at hfx.eastlink.ca)


From ruediger.maehl_nospam at web.de  Wed Apr 17 02:28:36 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Wed, 17 Apr 2002 08:28:36 +0200
Subject: need help to choose an appropriate database
References: <fac97694.0204160518.6f6d79b7@posting.google.com>
Message-ID: <newscache$xi8pug$g0i$1@www-neu.dzsh.de>

"alexandre" wrote:
> I work in a company who wants to register informations concerning
> customers in a datebase. they suggest me to use CGI and Python to
> capt data. I must send these data in a database and i would like to
> know which ODBC is the most appropriate to that situation ?
>
> thank you

If you only need a simple solution you might check out FlatDB at
http://www.swoogan.com/flatdb.html

I am using SAPDB, but attention: that is a chapter of its own.
You find it at www.sapdb.org

R?diger




From mwh at python.net  Sat Apr 13 07:46:57 2002
From: mwh at python.net (Michael Hudson)
Date: Sat, 13 Apr 2002 11:46:57 GMT
Subject: inspect.getargspec() for built-in functions
References: <mailman.1018685007.21078.python-list@python.org>
Message-ID: <lkit6vizp1.fsf@pc150.maths.bris.ac.uk>

Tripp Scott <tripps81 at yahoo.com> writes:

> is there any?

Can't realistically be done[0].  A sometimes adhered to convention is
that the first line of the docstring contains a "prototype" (in the C
sense).

Cheers,
M.
[0] Is whether the function is METH_O exposed to Python code?  I don't
    think it is, but I'm not sure.[1]
[1] I think I can guess who will answer this question, if anyone...

-- 
  This is an off-the-top-of-the-head-and-not-quite-sober suggestion,
  so is probably technically laughable.  I'll see how embarassed I
  feel tomorrow morning.            -- Patrick Gosling, ucam.comp.misc


From mertz at gnosis.cx  Fri Apr 12 15:46:01 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Fri, 12 Apr 2002 15:46:01 -0400
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
 <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com>
 <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no>
 <mailman.1018592665.14227.python-list@python.org>
 <3CB72CE6.DFAD5C10@python.org>
Message-ID: <5lzt8kKkXANR092yn@gnosis.cx>                

"David Mertz, Ph.D." wrote:
|A list is a mutable sequence that has some items in it.  In contrast, an
|instance object--whether new style or old style (and w/ or w/o
|__slots__)--is a thing that has some attributes.  The basic shape of the
|data will always be different.

Guido van Rossum <guido at python.org> wrote previously:
|    class D(list):
|        __slots__ = ["foo"]
|List or instance?  The shape of a D instance is the same
|as the shape of a list, but with an extra slot added.

It seems like the shape of this data is different from either an
old-style list or an old-style instance.  There are several standard
modules (or even builtins) that have questionable--and probably
wrong--behavior around this:

    >>> d = D()
    >>> d.append('this')
    >>> d.foo = 'foo'
    >>> import pprint
    >>> pprint.pprint(d)
    ['this']
    >>> from pickle import dumps, loads
    >>> d2 = loads(dumps(d))
    >>> d2.foo
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    AttributeError: foo
    >>> repr(d)
    "['this']"

Moreover, xml_pickle in particular has an extra burden here, above and
beyond what pickle does.  The XML representation wants to be fairly
human-readable (which I think it is so far).  This requirement is not
really there for pickle itself.

But xml_pickle also wants to be cross-language; that's why there is a
DTD associated with the dialect, for example.  Other programming
languages should be able to read xml_pickle files, and create native
objects that provide at least a pretty good approximation of the data
structures contained in the Python objects.  And likewise going the
other direction.  As such, I take "sequence" or "map" to be general
families of data "shapes" that many languages use.  These basic shapes
should therefore be indicated in the XML (while also providing clues for
the more specific data containers used).

Yours, David...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From dave at rotwang.freeserve.co.uk  Sat Apr  6 04:21:38 2002
From: dave at rotwang.freeserve.co.uk (Dave Berkeley)
Date: Sat, 6 Apr 2002 10:21:38 +0100
Subject: location of a module's directory
References: <a8k1cb$s9v$1@news6.svr.pol.co.uk> <3CADBDB4.7090007@olen.to>
Message-ID: <a8memi$p0$1@news8.svr.pol.co.uk>

Yes, thanks. My mistake was only looking at the module sys, which is of
course built in, so it does not have a __file__ attribute. Thanks.

Dave

"Joonas Paalasmaa" <joonas at olen.to> wrote in message
news:3CADBDB4.7090007 at olen.to...
> Dave Berkeley wrote:
> > Is there a way to discover where a module was loaded from?
> >
> > I want to include a text configuration file with a .dll, and place them
in
> > the same directory. If the module knows where it was loaded from it will
be
> > able to find the config file.
> >
> > If you do (Python 2.1.1)
> >
> >
> >>import sys
> >>print sys.modules
> >
> >
> > you can get a dictionary of loaded modules, and a description, which
> > includes the path the module was loaded from. However, I can find no way
of
> > accessing this information programmatically.
>
> Are you looking for this?
>
> Python 2.2a3 (#23, Sep  7 2001, 01:43:22) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE Fork 0.8 -- press F1 for help
>  >>> import cgi
>  >>> cgi.__file__
> 'C:\\Python22\\lib\\cgi.py'
>




From vvainio at tp.spt.fi  Fri Apr 19 02:32:51 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 18 Apr 2002 23:32:51 -0700
Subject: ActivePython vs. Python
References: <WkGv8.38$Ud2.86702@news2.news.adelphia.net>
Message-ID: <ad496f8.0204182232.288e35e3@posting.google.com>

"Bob H" <bobx at linuxmail.org> wrote in message news:<WkGv8.38$Ud2.86702 at news2.news.adelphia.net>...
> What is the benefit of getting/using ActivePython over Python?

One drawback appears to be it's reliance on ActiveState (which has to
consider such things as profit - and it also provides ActivePerl ;-).
There was a long period when nothing happened with ActivePython, which
no doubt alarmed many people.

However, I do love the windows help format of the Python docs,
ActivePython might be worth it just because of them!


From pyth at devel.trillke.net  Mon Apr 22 14:24:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 22 Apr 2002 20:24:15 +0200
Subject: rlcompleter not calling __getattr__ on [ ]
In-Reply-To: <aa1hvn$o9i$1@peabody.colorado.edu>
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu>
Message-ID: <20020422182415.GI27395@devel.trillke>

On Mon, Apr 22, 2002 at 11:39:24AM -0600, Fernando P?rez wrote:
> Jaco Schieke wrote:
> 
> > Looks like the regexp in rlcompleter that searches the string does not
> > work correctly.  After changing the line
> > 
> >> m = re.match(r"(\w+(\.\w+)*)\.(\w*)", text)
> > 
> > to
> > 
> >> m = re.match(r"([\w\[\]]+(\.[\w\[\]]+)*)\.(\w*)", text)

or maybe even r"(\S+(\.\w+)*)\.(\w*)" ? This would principally
allow to add 

"".<tab>

to work as expected. Of course you need to remove " and ' 
from the delimiter list. 

But then again, i wonder if the rlcompleter should not simply
see the *whole* string and do the string-split according to delimiters
itself? This might be more flexible for 'general purpose' completion. 
I consider the delimiters to be a rlcompleter property and
not a readline property.

	holger




From grante at visi.com  Mon Apr 22 12:30:13 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 22 Apr 2002 16:30:13 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019489191.17192.python-list@python.org>
Message-ID: <pUWw8.48099$vm6.7526151@ruti.visi.com>

In article <mailman.1019489191.17192.python-list at python.org>, brueckd at tbye.com wrote:

>> > Is there a difference?
>>
>> Yes.  Science verifies it's theories by comparing them to the
>> physical world which they are attempting to describe.  It
>> doesn't matter how elegant, or internally consistent Science
>> is.  If it doesn't successfully describe the physical world,
>> it's wrong.
>>
>> Science has an external reference point. Mathematics does not.
>> Internal consistency is the only thing which mathematics can
>> attempt to verify.  Mathematics is not an attempt to describe
>> the physical world.
> 
> First, please back up, I was asking if there's a difference between
> discovery and invention, which you don't mention anywhere in your
> response.

Not literally, no.  I thought it obvious that since math
doesn't have a physical reference, it is pure invention, while
science due to it's ties to the physical world is a process of
discovery.

> Second, math *is* an attempt to describe the physical world

In my experience, that's a rather unique opinion, and one not
held by any of the math type people I've met.

> - that's what makes it useful.

The fact that it _can_ be used to quantify descriptions of the
physical world makes it useful.  That usefulness is not a
requirement of mathematics.

> Case in point: calculus, which was invented/discovered
> specifically to deal with and describe the motion of physical
> bodies. Had it not been useful in describing such motion, it
> would have been tossed out, or at least not so widely accepted.
> Math is useful because of its relevance to the world around us.

I agree that many fields of math are useful.  That is not,
however a fundamental requirement of math.

> Finally, in the more general sense, formal mathematical proofs
> and whatnot might not require an external reference point, but
> the foundation upon which they are built certainly does.

That's not the way I learned it.  You start with a set of
fundamental assumptions which you do not attempt to "prove".
You then build on that.

> Indeed, the fact that we've gotten as far as something like
> abstract algebra is largely due to the fact that the underlying
> building blocks *are* verifiable and applicable to our
> experiences outside of math and many of the "advances" in
> mathematics have been due to intuition or hunches provoked by
> real world phenomena.

Such hunches would imply that Euclidean geometry is an accurate
description of space...

-- 
Grant Edwards                   grante             Yow!  Did you find a
                                  at               DIGITAL WATCH in YOUR box
                               visi.com            of VELVEETA??


From logiplexsoftware at earthlink.net  Wed Apr  3 15:04:41 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 12:04:41 -0800
Subject: PEP 285: Adding a bool type
In-Reply-To: <uakpepcn1iir15@news.supernews.com>
References: <mailman.1017466974.6751.python-list@python.org>
	<2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk>
	<mailman.1017755045.4308.python-list@python.org>
	<uakpepcn1iir15@news.supernews.com>
Message-ID: <20020403120441.44846ff3.logiplexsoftware@earthlink.net>

On Tue, 2 Apr 2002 21:21:27 -0500
John Roth wrote:

> 
> "Laura Creighton" <lac at strakt.com> wrote in message
> news:mailman.1017755045.4308.python-list at python.org...
> > Dear Guido:
> >
> > I would first like to thank you for giving us an opportunity to
> respond.
> > I have spent most of the weekend thinking and writing a reply to this,
> > and I think that this has made me a better teacher.  For this I am
> > grateful.  I realise that this is rather long, but I have condensed
> > it substantially from my efforts earlier this weekend.   Thank you
> > for taking the time to read this.
> >
> > Laura Creighton
> >
> > -----------------------------
> >
> I'm not going to quote this, as it's already too long. I read up to
> the diatribe on the Swedish leap year, and I went AAAARRRRGGGH!
> 
> I really don't care if you want the silly function to return
> a Boolean, the number of leap days in the year, or the name
> of the idiot parliamentarian who decided to fix the calendar
> so that 1712 was different from the rest of Europe.
> 
> HOWEVER, if you want true or false as the return
> value, the name should be something like 'isLeapYear()'.
> If you want the number of leap days in the year, then
> it should be named something like 'numberOfLeapDays().'

> Sorry to make this sound like a flame, but I try to
> practice XP. The first time I ever come across a
> function named "isSomething" that returns a conceptual
> integer, it's going to get refactored, right then and
> there to a name that actually expresses what it does.

Laura was using the example given by Guido, since she was responding to
Guido's statements. 
I totally agree that a function name that implies a boolean should return a
boolean.  Guido was stating otherwise, that a boolean could also be used as
an int, which Laura felt (and apparently you and I agree) is wrong.  If you
had read the entire post you would have seen that that was (partially) her
point:

[GvR]
>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

[Laura]
> This is not an argument for allowing non-Boolean operations on bools();
> this is an argument for not writing functions that return Booleans.  Make
> them return numbers instead, so that you can use them as you did.  Last
> month we discussed why in 1712 February had 30 days in Sweden.  (See:

So what you are saying fits perfectly with what Laura asserted.  If you
have a function numberOfLeapDays() why on earth would you need
isLeapYear()?  The solution is simple: don't return a boolean when another
type can not only express true/false but also convey other information as
well.

I didn't necessarily agree with all of Laura's assertions regarding the
difficulty of understanding how booleans would fit into Python, but she
raised several compelling and thoughtful arguments.  I totally agree that
bools as return values of functions are less than useful.  Booleans have
the lowest information density of any type and thus would tend to increase
code verbosity if their use were to be encouraged.

Frankly Guido's dismissal of Laura's arguments as FUD was not only a bit
rude, but IMHO a bit short-sighted as well (given the fact that someday,
there may be a desire for a real boolean type).  Maybe Tim Peters has
borrowed his time-machine and not returned it ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From emile at fenx.com  Sun Apr 21 10:42:00 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 21 Apr 2002 14:42:00 GMT
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com> <a9ufpv$6hb$1@panix1.panix.com> <just-DF77C7.15571021042002@news1.xs4all.nl>
Message-ID: <YcAw8.44330$V44.923@rwcrnsc53>

Just van Rossum
> I'm already happy if they don't do
> 
>   if x in d.keys(): ...
> 
> :-)
>

and we shouldn't have to worry about seeing

if x in zip(*d.items())[0][:]:...

;-))

-- 

Emile van Sebille
emile at fenx.com

---------



From gbreed at cix.compulink.co.uk  Tue Apr 30 06:47:47 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Tue, 30 Apr 2002 10:47:47 +0000 (UTC)
Subject: Third Draft:  Pep for Deprecating Builtins
References: <mailman.1020109494.7386.python-list@python.org>
Message-ID: <aalsoj$dhj$1@thorium.cix.co.uk>

Sean 'Shaleh' Perry wrote:

> > 5.  oct() is another candidate based on the rarity of usage
> > 
> 
> so should we remove the left and right bit shift ops too then?  oct() 
> may be
> rarely used but having it keeps me from writing C or perl code.

If oct() were to be (re)moved, the few cases where it's needed could be 
replaced with %o formatting.  So if we are looking to sacrifice builtin 
functions that's one to consider.


                      Graham


From phd at phd.pp.ru  Thu Apr 11 05:42:46 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 11 Apr 2002 13:42:46 +0400
Subject: SMTP with attachments
In-Reply-To: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>; from j.ezequiel@spitech.com on Thu, Apr 11, 2002 at 05:34:11PM +0800
References: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>
Message-ID: <20020411134246.H30351@phd.pp.ru>

On Thu, Apr 11, 2002 at 05:34:11PM +0800, Ezequiel, Justin wrote:
> How do I send an e-mail with attachments?

   Download email 2.0.1 from http://sourceforge.net/projects/mimelib (it
will be incorporated into Python 2.3) and learn to use it. There are enough
docs and examples there.

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




From andersjm at dancontrol.dk  Wed Apr 10 08:38:19 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Wed, 10 Apr 2002 14:38:19 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <3CB427D3.7020404@geneva-link.ch>
Message-ID: <3cb432ca$0$231$edfadb0f@dspool01.news.tele.dk>

"Boris Borcic" <borcis at geneva-link.ch> wrote in message
news:3CB427D3.7020404 at geneva-link.ch...
> Anders J. Munch wrote:
>
> > Pythons truth values are not "Something or Nothing".  The phrase seems
> > to have caught on following Laura's post, but I believe it is
> > misleading.
> >
> > [] and {} are objects, they are objects with identity, and they are
> > mutable objects.  They are empty, but they are by no means "nothing".

> I could discuss the "Something" side of the matter, but I feel the point
> you raise here very much boils down to "the type of zero is a manifest
> information that disallows viewing the typed zero as a nothing".

No, I was specifically discussing how objects created by the expressions
[]and {} are poorly described by "nothingness", and how for this reason
equating pythonic falsity with nothingness can lead you astray, as shown in
the example.

and-I-was-specifically-NOT-entering-into-a-PEP285-flamewar-ly-y'rs,
Anders




From comdog at panix.com  Fri Apr  5 19:58:54 2002
From: comdog at panix.com (brian d foy)
Date: Fri, 05 Apr 2002 18:58:54 -0600
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050112.11d4ef89@posting.google.com> <slrnaaseso.99d.huaiyu@gauss.almadan.ibm.com>
Message-ID: <050420021858547595%comdog@panix.com>

In article <slrnaaseso.99d.huaiyu at gauss.almadan.ibm.com>, Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> wrote:

> On 5 Apr 2002 01:12:46 -0800, Alan Kennedy <alanmk at hotmail.com> wrote:
> >brian d foy <comdog at panix.com> wrote:

> >> for our April 2002 issue (April Fool!), The Perl Review replaced its
> >> Perl content with Python articles, so Python folks might want to read
> >> this issue:

> >>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf

> This is an interesting read.  Unfortunately, the first paragraph of Paul
> Prescod's article contains a glaring trivial error of Fibonacci numbers.  It
> must have been calculated by hand, as his program is correct.  

it's probably my fault.  i've fixed the problem (by running the python program
that Paul supplied).  a new version of the PDF file is now available, and i'll note 
my error in the next issue. sorry for the confusion and inconvenience.

> Wonder what
> impression that will make on the Perl audience.  

they'll probably think I need to do a better job converting what the 
author writes into LaTeX. :)

-- 
brian d foy <comdog at panix.com> - Perl services for hire
The Perl Review - a new magazine devoted to Perl 
<http://www.theperlreview.com>


From ak at silmarill.org  Tue Apr 30 14:07:00 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 30 Apr 2002 18:07:00 GMT
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com> <f0fd5987.0204300105.51529004@posting.google.com> <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk>
Message-ID: <slrnactndm.b2u.ak@ak.silmarill.org>

In article <extz8.43542$dh5.3425591 at news-binary.blueyonder.co.uk>, JohnO wrote:
> Garry Taylor wrote:
> 
>> Chris <chris at cmb-enterprises.com> wrote in message
>> news:<chris-4E0B1A.01281230042002 at corp.supernews.com>...
>>> In article <df30afd4.0204292111.31e5914a at posting.google.com>,
>>>  xiaotsing_79 at hotmail.com (David) wrote:
>>> 
>>> > I have a lot of experience in ANSI C/C++, I want to learn a new
>>> > language and which it can help me in the work.
>>> > I just want to clear which(python or perl) is more comfort for me?
>>> 
>>> In terms of how it looks, Perl might look more comfortable, but Python
>>> is likely to function more similarly to C++, while C will have more in
>>> common with Perl.
>>> 
>>> Learn both.
>> 
>> I use Perl at work day in day out, and use Python for my own personal
>> stuff, I would say that Python is better for pretty much everything,
>> particularly larger programs, where it's more structured syntax makes
>> code easier to read. I quite like Perl for CGI scripting, and little
>> 1-page scripts, but would never contemplate using it for a big
>> project.
>> 
>> Learning both is an ideal solution, but personally I wish I started on
>> Python far earlier than I did.
>> 
>> Garry
> 
> 
> I don't know if anyone else has noticed this but I've found that operators 
> and sys admins use perl and that developers tend towards python. Also sys 
> admins rarely utilise indirection but developers do.
>
What do you mean by indirection?


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From aleax at aleax.it  Sat Apr  6 09:29:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 06 Apr 2002 14:29:46 GMT
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org>
Message-ID: <uDDr8.66496$S52.2342405@news2.tin.it>

<posted & mailed>

Guido van Rossum wrote:
        ...
> About the change rate in general: it's hard to find the proper pace.
        ...
> for faster change.  I have to pick a middle ground (and it's easier to

Linus Thorvalds seems to have done pretty well with picking TWO middle 
grounds -- two parallel tracks for Linux, "stable" and "experimental".

Without the former he'd have lost the commercial interests, and "normal
people" who DO need SOME level of stability; without the latter, a crucial 
hard-core of neophile hackers might have gone away in search of other, more 
dynamic projects.  With both tracks, things seems to have been smoother.

Why don't we start thinking of a similarly dual-tracked Python -- one
track aiming at stability (in a "middle ground" sensible sense: strong
commitment to not breaking old code -- introducing small new features
and new library modules _is_ OK, else one can use a bugfix-only-"track"
for some previous release), one aiming at frequent releases, innovation,
etc.  You'd have to BDFL both tracks, I think (or else the "stable" one
would be perceived as "ugly stepsister", or, less likely, the experimental
one as an irrelevant fork should your attention be all on the stable)
and no doubt it would take experimentation and attention to decide on
such issues as average release rates, etc -- we can't just blindly
copy Linux's model, I just suggest stealing the dual-track idea.

The experimental track would give you lots of freedom (no guarantee
that what runs in e.g. 3.1.6 will keep running in 3.1.7), while the
stable one would prove reassuring to firms who are thinking of large
Python investments but quite reasonably wary of investing in a
technology they can't perceive as even _aiming_ to offer a stable
yet actively developed option.  Maybe when you release 3.0 (to pick
a nice round number) you could release 3.1 at the same time (just
about the same, perhaps) and decree 3.0.* the stable track until
further notice, 3.1.* the experimental one -- or, pick whatever
numbering scheme you prefer.


Alex



From mlh at vier.idi.ntnu.no  Wed Apr 10 22:21:00 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 02:21:00 +0000 (UTC)
Subject: [OT] Re: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>  <3CB2DB99.4080400@geneva-link.ch>
Message-ID: <slrnab9soc.6g8.mlh@vier.idi.ntnu.no>

In article <3CB2DB99.4080400 at geneva-link.ch>, Boris Borcic wrote:
>    PythonWin 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] [...]
>
>    >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
>    'BUSH'
>
>
>What's your problem ?

Nothing. Sorry -- thouhgt there was a syntax problem there. A bit
quick on the trigger ;)

(So, what you're basically saying is that Bush is the literate part of
bullshit? <wink>)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From jadestar at idiom.com  Mon Apr 29 23:00:09 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 30 Apr 2002 03:00:09 GMT
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com>
Message-ID: <aal1bp$1qf0$1@news.idiom.com>

Fortepianissimo <fortepianissimo at yahoo.com.tw> wrote:

> This question is so fundamental that I strongly suspect it's already
> answered - ig that's the case, please forgive this poor newbie, for he
> has done some work searching over Python FAQ and Google.

> Basically I want to have a subclass of list, like

> class MyList (list):


> then I want to

> l=MyList()
> ... do stuff to fill l
> s=" ".join(l)

> with the hope that s will be a nicely joined string with " " as the
> delimiter. Given my C++ background, I tried to define __str__/__repr__
> for both MyList and the user-defined types of the list members,
> assuming the type conversion would silently be invoked. But I was
> wrong!

> Even trying this failed:

> l=[1,2,3]
> s=" ".join(l)

> and I thought built-in types all have default __str__ and __repr__?

> What am I missing here? Many thanks to someone who would take time to
> rescue a poor soul.

> Ben

	foo.join() takes a list of strings; it doesn't seem to 
 	coerce other types into strings.

	Try this:

		l = [1,2,3]
		s = " ".join([repr(x) for x in l])
	
 	... or (some what more readable?):

		s = " ".join(map(repr,l))

	... or the (slightly strange but possibly more efficient):

		s = ( "%s " * len(l) ) % tuple(l)

	... if you don't mind the trailing space
	... or:

		s = ("%s " * (len(l) - 1) + "%s" ) % tuple(l)

	... those last two weirdos are based on a suggestion that
	% string interpolation under Python might be more efficient
	for long strings then looping over a list (successively building
	longer strings).  (I'm not sure that the performance hint 
	applies to join with map or list comprehensions; and there's always
	Knuth's opinion of premature optimization!)

	I'm not sure that any of these is really significantly better,
	more elegant or more readable than any of the others.  I don't
	like the list comprehension used in my first example (it's just
	ugly there) though I usually like list comprehensions in general.
	.join(map(...)) is the shortest and probably the most readable to
	any one with a clue what map does.  

	The last one is a bit much build a string of %s things that's 
	almost long enough, then add in another one for good measure 
	and then % that with a tuple!  

	On the other hand, if I can think of it off the cuff then it 
	can't be too hard for most Pythonistas to read. (I was able 
	to do all of these with only one gotchya and almost no thought 
	--- I initially thought I could get by without the explicit 
	call to "tuple()" since lists and tuples are sometimes sort of 
	interchangeable if they're in an immutable context --- well, 
	maybe that's a bad way to think of them. (Forget I said that!).

	On the bright side the last technique works in 1.5.x even without 
	any imports!  So it is quite portable, unlike any of the others.
	However, we can do:

		import string
		s = string.join(map(repr, l), " ")

	... which is quite similar to " ".join(map(...))

	Hope that helps.



From aleax at aleax.it  Thu Apr  4 06:56:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 11:56:24 GMT
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com>
Message-ID: <IbXq8.48892$pT1.1388420@news1.tin.it>

A. Jones wrote:
        ...
> I'll probably get blasted for this by Python gurus... but, I've used

*BLAST*!  No, seriously.  Don't do that.  You'll get hurt if you do.

> it with some success.  You can do
> 
> exec(myObjc + " = " + value)

You can do it without the parentheses, too (and that makes it
clearer what you're doing, because exec is a statement, while
you make it look as if it was a function).


> or, if it's a class instance you want to create
> 
> exec(myObjc + " = " + bar + "()").
> 
> Toss whatever variables you want sent on to the instance inside "("
> and ")" and you're homefree.
> 
> Unless there's some problem with that that an old-timer spots that I
> haven't yet.

The problem is basically exactly the specs as expressed: "please
help me trample all over my namespace by dirtying it with names
that I cannot control".  These specs are wrong and you are
guaranteed no end of grief if you do match them.

Suppose you do:

def wrongHeaded(name, value):
    exec name+'='+repr(value)
    print vars()

Seems fine, no?  Can hardly be simpler, right?

wrongHeaded('a', 23)

does indeed emit:
{'a': 23, 'name': 'a', 'value': 23}

proving it's done just what we expected.  BUT -- what if we
call it like...:

wrongHeaded('vars', 24)

OOPS!  Now it dies:

Traceback (most recent call last):
  File "d.py", line 6, in ?
    wrongHeaded('vars', 24)
  File "d.py", line 3, in wrongHeaded
    print vars()
TypeError: 'int' object is not callable

What's it taling about?  Answer: it's talking about 'vars'.
We just set it to 24, so we can't call it any more.

This comes to apply to ALL identifiers WITHOUT EXCEPTION.

You cannot use ANY identifier safely after that exec
statement, as you have no idea any more what the
identifier can refer to.  As a bonus, your code crawls,
since the Python compiler knows it does not know, and
therefore executes a full identifier look-up at runtime
rather than recognizing local variables at compile time.

What kind of useful code can you write without being
able to use ANY identifier safely?  WHY would you ever
WANT to trample all over your namespace to ensure every
identifier becomes an utter and total mystery?  Beats me.


WHATEVER the problem may be that you THINK you need to
solve by wilfully destroying your namespace, I ensure
you, it *ISN'T*.  There is ALWAYS a better approach,
whatever your application's exact needs might be.


Alex



From shalehperry at attbi.com  Tue Apr 16 02:43:31 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 15 Apr 2002 23:43:31 -0700 (PDT)
Subject: Did anyone have problems installing Numeric module on Debian
In-Reply-To: <3cbba193.2712952@news.cis.dfn.de>
Message-ID: <XFMail.20020415234331.shalehperry@attbi.com>

On 16-Apr-2002 Henry Lebowzki wrote:
> Hi-
> 
>       I've been having some problems installing the Numeric module
> on Debian Woody GNU/Linux.
> 
>       Whenever I use IDLE, it just can't find the correct path. I
> can see the Numeric module with MC (Midnight Commander) or another
> file viewer. However, IDLE's Path Browser doesn't "register" it.
> 
>       Should I be doing some "make" here? 
>       
>       The funny thing is that dpkg installed python-numeric without
> any problem.
> 

I have not encountered any strangeness here, however I do not use idle.  If you
just do:

$ python
>>> import Numeric

does it work?




From goodger at users.sourceforge.net  Thu Apr  4 23:32:54 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Fri, 05 Apr 2002 04:32:54 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <mailman.1017966793.11587.python-list@python.org>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204042338.XAA12847@crown.off.ekorp.com> <01c701c1dc35$5a2312f0$02010a0a@suxlap> <mailman.1017966793.11587.python-list@python.org>
Message-ID: <B8D2934A.20F7E%goodger@users.sourceforge.net>

Thank you to everyone who has commented on PEP 287, be they in favor,
well-meaning but opposed, or ill-wishing troll ;-) [1]_.  I have been
otherwise busy and haven't been able to keep up with replies, but I have
read and kept a record of all the posts so far (and will, for posts yet to
come) for consideration and reference when revising the PEP.  I may not be
able to offer timely replies to all the posts; ultimately the revised PEP
will serve as an umbrella reply.

.. [1] Actually, the posts opposed and even the trolls are often the most
   useful, since they point out weaknesses and help strengthen the
   arguments.  The number of people posting *in favor* of the PEP was
   roughly the same as the number opposed, but supporters tend to be
   silent while protesters tend to be quite vociferous and relentless.

Over the weeks to come, I will be revising PEP 287 and related PEPs 256-258,
merging the reStructuredText_ and DPS_ projects into the renamed Docutils_
project, and producing a "road map" describing the big picture.  Further
suggestions for improvement are welcome.

.. _reStructuredText: http://structuredtext.sourceforge.net/
.. _DPS: http://docstring.sourceforge.net/
.. _Docutils: http://docutils.sourceforge.net/

A special thanks must go to Richard Jones for taking up arms (or keyboard)
in my stead, and for taking the initiative to start work on `A
ReStructuredText Primer`_.  In hindsight it was much needed, and it's coming
along nicely.  Thanks Richard!

.. _A ReStructuredText Primer: http://mechanicalcat.net/tech/quickstart.html

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From gimbo at ftech.net  Thu Apr  4 14:46:12 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 4 Apr 2002 20:46:12 +0100
Subject: how to detect if a file is complete
In-Reply-To: <32f17666.0204040701.64dd0bfe@posting.google.com>
References: <32f17666.0204040701.64dd0bfe@posting.google.com>
Message-ID: <20020404194611.GA11526@andy.tynant.ftech.net>

On Thu, Apr 04, 2002 at 07:01:43AM -0800, harm wrote:

> I want to write a python script which checks a dir and copies all
> files to another dir. But there is the possibility that the file is
> not yet complete. (This are rather large Postscript files >100mb).
> Therefore I need to detect if the file is complete.
> 
> The files are copied on the samba-share by Adobe Acrobat Distiller,
> therefore i cannot save some check file like $FILENAME.complete...
> 
> Is there some other way (using python) to detect if the file is
> complete?

Define complete.

If the file has a well-defined structure (eg XML) or even just some
indication of its length, you might stand a chance.  If it's just
random plain text, how would you tell complete from incomplete?  I
don't know anything about Postscript internals but I imagine it uses
some concept of correct/broken, maybe this would help?

Alternative approach: use fuser to see what processes have the file
open.  I don't know how this would work with samba though...
Obviously the unix system will have no concept of the Distiller
process, so presumably it would be a process id associated with
samba...  Mumble mumble, try it and see I suppose.  :-)

Second alternative approach: don't copy any files created in the last
n minutes.  If you know Distiller never takes longer than 2 minutes to
do its stuff, set n to 3, or 5, or whatever, and you'll be sure you're
only copying complete files.  This may or may not be an acceptable
solution, depending on the time-criticality of the application.  If it
doesn't matter that it takes 10 minutes for the file to appear at its
destination, this is a very simple and effective solution to your
problem.

Good luck,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From op73418 at mail.telepac.pt  Wed Apr  3 11:52:22 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 03 Apr 2002 17:52:22 +0100
Subject: __name__ on new-style classes
References: <stcfaugkbh0u65d1euuofhs2b69ej7i34h@4ax.com> <3CA7C873.D1AC908A@python.org> <jvogausg16i1vgp93qkbb8f9sb3nicgb7u@4ax.com> <3CAA619C.A38F2DC9@python.org>
Message-ID: <t7cmau88q9knlq96ucseofcmtdbsn7viev@4ax.com>

On Wed, 03 Apr 2002 01:53:56 GMT, Guido van Rossum <guido at python.org>
wrote:

>"Gon?alo Rodrigues" wrote:
>> 
>> On Mon, 01 Apr 2002 02:35:47 GMT, Guido van Rossum <guido at python.org>
>> wrote:
>> 
>> >It's a feature.  The reason for the change is that I've never seen an
>> >example of a use for it (as opposed to changing an instance's __class__).
>> >If you have one, let us know.
>> 
>> I bumped into this "feature" a few days ago trying to understand
>> metaclasses in Python. And to understand them there is nothing better
>> than actually coding one, so I'll explain my problem/solution and let
>> you judge if this is a good use case for having a writable __name__.
>> 
>> My problem is modelling finite arithmetic in Python. Here by finite
>> arithmetic I mean all the quotients of the integer ring Z. There is a
>> countable number of them, one for each n > 1 (n=0 is Z and n=1 is the
>> trivial ring {0}. If n is negative then it is the same as -n). If n is
>> prime then we actually have a field.
>> 
>> The obvious way (to me, that is) to do this is to let each finite ring
>> Z_n be a class, inheriting from int. Since there are an infinite number
>> of these we have to have a metaclass, call it Field, that has as
>> intances the Z_n rings, e.g. Z_n would be Field(n). The skeleton of this
>> metaclass would be something like:
>> 
>> class Field(type):
>>     """The Finite Field metaclass."""
>> 
>>     def __new__(cls, n):
>>         if isinstance(n, int) and n > 1:
>>             #The template class.
>>             class Z(int):
>>                 pass
>> 
>>             #Generate proper name for the instance class, e.g. "Z_n"
>>             Z.__name__ = "Z_%d" % n
>> 
>>             #Return class.
>>             return Z
>>         else:
>>             raise TypeError, "%s is not an integer > 1." % str(n)
>> 
>> If I cant write to __name__ then all the instances will have the same
>> name... Not A Good Thing, IMHO.
>
>But you're doing the meta-instantiation all wrong.  Rather than using a
>class
>statement, you should be using cls.__new__(cls, "Z_%d"%n, (int,), {...}).
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)

Thanks for the correction. I should have figured that one out. Still,
shouldn't it be

type.__new__(cls, "Z_%d" % n, (int,), {...})

(instantiating the type class not the cls class) Btw, could you tell me
what is inherently wrong in generating classes with a class statement?
Could the generated class somehow behave improperly?

All the best,
Gon?alo Rodrigues


From marklists at mceahern.com  Tue Apr 30 23:07:49 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 22:07:49 -0500
Subject: crypto
In-Reply-To: <7xu1psshjy.fsf@ruckus.brouhaha.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEJCKNAA.marklists@mceahern.com>

[Paul Rubin]
> It will be GNU GPL, unless you'd like to fund development ;-).

Don't I wish!

How can I get my grubby little hands on it?

Cheers,

// mark




From gpropf at cfl.rr.com  Mon Apr  1 20:39:45 2002
From: gpropf at cfl.rr.com (Gregory Propf)
Date: Mon, 01 Apr 2002 20:39:45 -0500
Subject: Extending Python on Windows
Message-ID: <3CA90BE1.9010703@cfl.rr.com>

I am trying to learn how to make Python extensions for a cross platform 
project I'm thinking about.  Basically the CPU intensive stuff will be 
in C or C++ and the GUI will be done with wxWindows in python.  I will 
need to build python extensions for this.

Attached to this email is my version of the "spam" sample code that you 
find on the python website.  The instructions for compiling on *Linux* 
are in a comment at the end.  This procedure works fine and I am able to 
use the extension.  On Windows (NT 4.0 sp6) I have tried to build this 
extension using the version of gcc and ld that comes with cygwin.  The 
only change is that --shared is replaced with -c (--shared does not seem 
to work on Windows) and the .so extension is replaced with .dll.

Even on Windows the extension builds and links just fine.  I invoke 
python just the way the comments show (the LD_LIBRARY_PATH seems to be 
necessary on both Linux and Windows).  When I try to do 'import spam' in 
python I get an error that mentions dlopen and has a message like 
"Windows error 32??????????" (i.e. very big number).  I wish I had the 
full error text but I didn't save it and can't get to it from home. 
Just thought someone might know what this is and have a simple fix.  I 
will post again tomorrow with the full error text (but the "error 
number" is so huge I suspect it's bogus).


-- 
You see, "state of the art" Microsoft Office applications sport a
nifty feature called "smart quotes." (Rule of thumb--every time
Microsoft uses the word "smart," be on the lookout for something
dumb).  - John Walker, founder of Autodesk.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: spammodule.c
URL: <http://mail.python.org/pipermail/python-list/attachments/20020401/9812446c/attachment.c>

From jvanasco at hotmail.com  Sat Apr 20 14:02:06 2002
From: jvanasco at hotmail.com (Jonathan Vanasco)
Date: Sat, 20 Apr 2002 18:02:06 +0000
Subject: Mass Mailing help needed (Paul Rubin)
Message-ID: <F123CvvMSaQz8xckA1Q000146ce@hotmail.com>

>    3. Re: Mass Mailing help needed (Paul Rubin)
>Why should we think those 50,000 people have really opted in?
>  Are you sure you're not spamming?? What site is this for?

No.  Its not spamming.

I work mostly for large entertainment companies -- if you would like so, i 
can send you to my portfolio and corporate site -- this particular list 
problem is centered around a record label's website relaunch.  Instead of 
inundating fans with 1 mailing from 20+  artist lists that they choose to 
subscribe to, I've decided to merge multiple lists into 1.  Subscribers will 
get 1 email with general label info, and specific artist news will be 
provided based on the user's requests.  (Posts will only be about monthly, 
aside from an 'album release' email that could be offschedule -- but that 
specific list data is always available)

This solution is a little more processing time on the backend -- but only by 
20 or so seconds -- but it should cut down on mail sending time (larger 
messages, fewer envelopes).   Mail::Bulkmail provides for this -- was 
actually built for stuff like this -- but perl is ungodly slow at merging 
the list names / options compared to python.

_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com





From kevin at cazabon.com  Tue Apr  2 10:07:43 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 2 Apr 2002 07:07:43 -0800
Subject: PyArg_BuildValue, how to return a PyString object?
References: <5a4226f0.0204012004.24bb7b00@posting.google.com> <lkelhya3rg.fsf@elios.maths.bris.ac.uk>
Message-ID: <5a4226f0.0204020707.43486467@posting.google.com>

Thanks for the help Michael, I'll give it a try.  I did do some
mucking about last night, and the null bytes in the strings I'm using
are causing all sorts of problems.

I think I might try passing in an array instead, and using the buffer
interface to the data, I'm just not sure yet how to pass the
manipulated data back to python as an array (or convert it to an array
in Python)... more reading to do I guess.

Kevin.

Michael Hudson <mwh at python.net> wrote in message news:<lkelhya3rg.fsf at elios.maths.bris.ac.uk>...
> kevin at cazabon.com (Kevin Cazabon) writes:
> 
> > I'm writing a C extension dll for Python, and am having a bit of
> > trouble figuring out how to return a PyString_FromStringAndSize object
> > (along with some normal stuff).
> > 
> > (I need to use PyString_FromStringAndSize because the string may
> > contain null bytes, am I correct in this, or is there an easier way?)
> 
> Doubt it.
> 
> > Can I use something like this (it doesn't seem to work like this
> > though):
> > 
> > PyObject *myPyStringObject = PyString_FromStringAndSize(cstringarray,
> > size);
> > PyObject *myReturnObject =  Py_BuildValue("llo", long1, long2,
> > myPyStringObject);
> 
> I think you have to mae that "(llO)", don't you?
> 
> There might be a s# format code, I can't rememeber -- Py_BuildValue
> *is* documented, though so you can check for yourself:
> 
> > Or do I have to build a tuple manually?
> 
> No.  Rememebr to DECREF "O" args after calling Py_BuildValue --
> something I'm always forgetting.
> 
> Cheers,
> M.


From robin at jessikat.fsnet.co.uk  Fri Apr 12 13:16:08 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 12 Apr 2002 18:16:08 +0100
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com>
 <7E283A6EA4F0DBD6.B68D98BE8C574F5D.CE8B287F1A68C959@lp.airnews.net>
Message-ID: <LFGz2JAYZxt8EwR6@jessikat.demon.co.uk>

...
well there's certainly been some efforts with Tcl eg

s. "Tcl and Concurrent Object-Oriented Flight Software: Tcl on Mars",
by David E. Smyth, Mars Pathfinder Flight Software Team, Jet
Propulsion Laboratory.

        Describes the use of Tcl on the "Mars Pathfinder" project, in
        which Tcl (with several extensions) is used in several ways:
        as the language for inter-object messages within the
        object-oriented flight software; as the implementation
        language for the monitor and control environment; and as the
        initial implementation language for some of the flight
        software itself.

but I'm not sure that Tcl actually went to Mars.
-- 
Robin Becker


From anthony at interlink.com.au  Tue Apr 16 11:11:18 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 17 Apr 2002 01:11:18 +1000
Subject: Do you QA your Python? Was: 2.1 vs. 2.2 
In-Reply-To: Message from Tim Peters <tim.one@comcast.net> 
   of "Sun, 14 Apr 2002 18:59:19 -0400." <LNBBLJKPBEHFEDALKOLCOECFOOAA.tim.one@comcast.net> 
Message-ID: <200204161511.g3GFBIl31849@mbuna.arbhome.com.au>

>>> Tim Peters wrote
> > IMO until Python developers adopt a policy of being very reluctant
> > to do this,
> 
> Guido is very reluctant to break code, but it's not an absolute for him:


Note that this is a "in new major releases" statement. Breaking code in 
a point release isn't something that happens. I'm not aware of any code
changes in 2.1.2 or 2.1.3 that broke someone's 2.1.1 code. If someone 
can point me to these changes that caused problems, I'd love to see them.

If instead this whole discussion is just usenet noise, with no real point
or purpose, then ignore the previous paragraph. Needless to say, I wait
with baited breath for the collection of shocking incompatibilities to 
roll in. 

Anthony
(is it a fact of usenet that over time, all newsgroups become like 
comp.lang.perl.misc?)
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From fredrik at pythonware.com  Wed Apr  3 05:10:22 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 10:10:22 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <mailman.1017817926.31983.python-list@python.org>
Message-ID: <iyAq8.29179$n4.5840426@newsc.telia.net>

Delaney, Timothy wrote:

> The whole point is that this is the *input*, optionally to produce another
> format, and is also human-readable in the source code (or any other text
> file).

who cares, when it's not human-writable.

</F>




From deltapigz at telocity.com  Tue Apr 16 00:27:46 2002
From: deltapigz at telocity.com (Adonis Vargas)
Date: Tue, 16 Apr 2002 00:27:46 -0400
Subject: classes + modules
Message-ID: <3cbbac29$1_2@nopics.sjc>

is there any way to initialize a class then get the initialized class name
and put it into a list then later on calling it from the list as if it were
the class?

i.e.

class_list = []
from somemodule import *
x = someclass()
class_list.append(x)

then do

class_list[0].somefunction() # assuming x is position at 0


any help would greatly be appreciated.

Adoni




From jblazi at hotmail.com  Tue Apr  9 08:07:25 2002
From: jblazi at hotmail.com (Janos Blazi)
Date: Tue, 9 Apr 2002 14:07:25 +0200
Subject: Simple question! urgent
References: <62l5buo5do5eh0h93b4r9ua8skp61b7302@4ax.com>
Message-ID: <3cb2d8fc_9@news.newsgroups.com>

> How to get NUMBER OF ITEMS in array? I need it for checkin nr of items
> in argv[]

simply len(x)
J.B.




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From dalke at dalkescientific.com  Sun Apr 28 15:13:31 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 28 Apr 2002 13:13:31 -0600
Subject: Newbie: Problem with QTable
References: <3CCC4793.C3B2EE17@bellsouth.net>
Message-ID: <aahhjs$l85$1@slb4.atl.mindspring.net>

Jeff Layton:
>   I hate to bother everyone, but I'm trying to figure out how to
>use QTable. I cooked up a small code that doesn't work (natch).
  ...
>Traceback (most recent call last):
>  File "table1.py", line 17, in ?
>    table=MyTable(numRows, numCols)
>  File "table1.py", line 12, in __init__
>    self.setCaption("Table1")
>AttributeError: sipThis

Looks like you forgot to call the QTable constructor.  Looking at
your code:
    class MyTable( qttable.QTable ):
        def __init__(self, rows, cols):
            """ help me """
            self.setCaption("Table1")

yep, that's your problem.  Try this instead

    class MyTable( qttable.QTable ):
        def __init__(self, rows, cols):
            """ help me """
            qtable.QTable.__init__(self, rows, cols)
            self.setCaption("Table1")

                    Andrew
                    dalke at dalkescientific.com





From christophe.delord at free.fr  Mon Apr 15 18:43:08 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Tue, 16 Apr 2002 00:43:08 +0200
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
	<MtIu8.18713$b62.536574@news1.tin.it>
Message-ID: <20020416004308.48201c4a.christophe.delord@free.fr>

On Mon, 15 Apr 2002 22:27:56 GMT
Alex Martelli <aleax at aleax.it> wrote:

> Duncan Smith wrote:
>         ...
> >     U = P = 0
> >     for value in dict.values():
> >         if value == 1:
> >             U += 1
> >         elif value == 2:
> >             P += 1
> 
> Again I think it might be faster to avoid the if/else with yet
> another dictionary:
> 
> counts = {}
> for count in adict.values():
>     counts[count] = 1 + counts.get(count, 0)
> 
> U = counts.get(1, 0)
> P = counts.get(2, 0)
> 
> >     return U*S/(U*S+P*(1-S))
> 

As indices are integers (1 and 2), a simple array may be faster :

counts = [None, 0, 0]		# counts[0] not used
for count in adict.values():
	counts[count] += 1

U = counts[1]
P = counts[2]

-- 
Christophe Delord
http://christophe.delord.free.fr/


From shalehperry at attbi.com  Thu Apr 11 12:02:20 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 11 Apr 2002 09:02:20 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <Pine.LNX.4.33.0204110933410.26004-100000@ns0.tbye.com>
Message-ID: <XFMail.20020411090220.shalehperry@attbi.com>

> 
> Ah...I think I understand your perspective a little better now. I never
> think of a tuple as a list that's immutable, so I don't encounter very
> often the same population problem. To me, lists become immutable when I
> give them names in all caps, just like my variables become constants when
> I name them that way.
> 

I do similar things, but using the language to help prevent mistakes is handy
when I give people modules to use in their own code.




From maxm at mxm.dk  Thu Apr  4 12:47:36 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 04 Apr 2002 19:47:36 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk> <slrnaaok7j.4ei.jon+usenet@snowy.squish.net> <3CAC567F.6090407@mxm.dk> <slrnaaouim.h07.jon+usenet@snowy.squish.net>
Message-ID: <3CAC91B8.3050400@mxm.dk>

Jon Ribbens wrote:
> In article <3CAC567F.6090407 at mxm.dk>, Max M wrote:
> 
>>>http://foo/a/b/c + d       == http://foo/a/b/d
>>>http://foo/a/b/c + ../d    == http://foo/a/d
>>>http://foo/a/b/c + ../../d == http://foo/d
>>
>>Then what is:
>>
>>http://foo/ + d       == ?
> 
> 
> http://foo/d of course. I don't see your point.                                 
> See RFC 2396 for the rules on relative URL resolution.                          

Oh bugger ... I assumed that all path elements in this example where 
directories. Hence the difference. It's the old windows habbit kicking in.

If it doesn't end in ".htm" it *must* be a directory ... So much for 
abstract thinking :-s

Then my code doesn't work as is.

regards Max M



From gcordova at hebmex.com  Tue Apr  2 09:51:28 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Tue, 2 Apr 2002 08:51:28 -0600 
Subject: __slots__ query
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB41@mail.hebmex.com>

> 
> I notice that in 2.2 it's possible to do strange things to slots via
> instances.
> 

Strange to you, but pretty obvious to me:

1. You define a class with a specification of allowing
   for two (only two) attribute slots, named "a" and "b":

> Thus 
> >>> class A(object):
> ...     __slots__ = ['a','b']
> ... 

2. You instantiate such a class.
   Note: The object you used to define which slots you
   wanted was mutable, but after defining the class,
   __slots__ may no longer be needed.

> >>> a=A()
> >>> a.c=2
> ===> error
> >>> a.__slots__.append('c')
> >>> a.c=2
> ===> error
> >>> a.__slots__
> ['a', 'b', 'c']
> >>> A.__slots__
> ['a', 'b', 'c']
> >>> b=A()
> >>> b.c=2
> ===> error

Seems pretty natural to me. I'd be alarmed if there
actually *was* a .c attribute.

The __slots__ object might be mutable, it doesn't
mean that you can add attributes anytime after
defining the class.


-gustavo



From boud at valdyas.org  Thu Apr 11 17:08:27 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 11 Apr 2002 20:08:27 -0100
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
References: <oOjt8.8815$vF6.254710@news2.tin.it>
Message-ID: <3cb5d163$0$85433$e4fe514c@dreader3.news.xs4all.nl>

<posted & mailed>

Alex Martelli wrote:

> I'm a bit puzzled -- there must be something strange in the order of
> finalization going on.  I can't believe I have to "import *" to make
> my applications terminate gracefully -- perhaps just one or two
> things from the qt module might suffice?
> 

I think you've found a bug -- have you mailed Phil?

> If anybody has more insight or can suggest more graceful workarounds,
> please let me know!
> 

No, it should just work whichever way you want.

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


From Philippe.Faes at rug.ac.be  Mon Apr  8 09:26:22 2002
From: Philippe.Faes at rug.ac.be (Philippe Faes)
Date: Mon, 08 Apr 2002 15:26:22 +0200
Subject: C and Python
Message-ID: <3CB19A7E.4090600@rug.ac.be>

I want to use python for quick prototyping and convert to C/C++ later on.
Suppose I write some modules (say 2) in python that call eachothers 
functions. Then I port one module to C/C++ and later the second module. 
In the meanwhile I will want my hybrid project (python-C) to work.

solution1: use python calls between C and python.
disadvantage: when my code is transformed to pure C, I still need python 
calls between the function (extra overhead) and I still need the python 
interpreter to make the calls. Obviously I don't want to reprogram the C 
code once it works.

solution2: same as solution1, but I could use macros for the function 
calls, so that I can just recompile a C module that calls python 
functions into a C module that call C functions. This would (in a first 
phase) be limited to primitive datatypes.

My question:
Does any of this already exist?
Any hints?

thanks.

--
Philippe Faes
Student of computer science engineering
Ghent Universitiy http://www.ftw.rug.ac.be



From niemeyer at conectiva.com  Wed Apr  3 12:27:54 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Wed, 3 Apr 2002 14:27:54 -0300
Subject: PEP 235: role of PyType_GenericNew()
In-Reply-To: <xduwuvoamgg.fsf@leto.informatik.uni-freiburg.de>
References: <xduwuvoamgg.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <20020403142754.A2252@ibook.distro.conectiva>

Hi Ralf!

> I'm wondering whether the current implementation of PyType_GenericNew
> does, what it's supposed to do.
[...]
> Thus, only the allocator of the passed type is called.
> Shouldn't the base classes allocators be called as well?

If you're inheriting from another object, and you don't want to
do anything special at object initialization time, why would you
use this function in the first place? Just leave the slot empty
and allow the super class to initialize itself. If you implement
your special function, it's your business to call the super class.

Best regards!

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From lumber at jack.com  Mon Apr 29 20:07:17 2002
From: lumber at jack.com (Lumberjack)
Date: 30 Apr 2002 00:07:17 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2> <3CCD9649.98AED86F@cascade-sys.com> <mailman.1020120416.1514.python-list@python.org>
Message-ID: <Xns91FFAE356DC8lumberjack@216.148.53.84>

pinard at iro.umontreal.ca (Fran?ois Pinard) wrote:
> For me, sarcasm and strong wording are not welcome.  There are ways to
> disagree, and even disagree much, while staying friendly.

No doubt you would stay friendly with someone who dislikes the color of the 
electrical wire strung through the walls of your house and seriously 
proposes (against your wishes) to level your house and rebuild it simply to 
put in different colored wiring. And you are paying for all this, not them. 
That is either being very gentlemanly or very stupid. However, I see no 
reason follow your misguided attempt at suffering fools gladly.

>  Even a big
> accumulation of ironic statements are not going to make the shadow of a
> good argument.  I prefer intelligent discussions between nice people.

Well then, what do you think of the use of the "experimental method" to 
establish proof? Perhaps you should go and change the language as proposed 
and see where it takes you.


From Chris.Barker at noaa.gov  Mon Apr  1 13:05:16 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 01 Apr 2002 10:05:16 -0800
Subject: string copying
References: <mailman.1017474166.1514.python-list@python.org>
Message-ID: <3CA8A15A.8B2F12E6@noaa.gov>

Tim Peters wrote:
> Does python run out of memory
> gracefully?
> 
> Very.  You should get a clean MemoryError exception if you try.

Perhpas so, but with VM, you can really slow your system to a crawl
before running out of memory. I'm running Linux, and when I do something
silly that caused Python to need a lot of memory, it's quite impressive
how unresponsive my system gets. 120MB on a sytem with 128MB physical
memory is pretty darn bad, over 128 MB, and things REALLY slow down! If
I didn't know better, I'd think I had crashed teh machine. I have always
been able to telnet in from somewhere else and shut down the process if
I was really patient. The console was totally unresponsive. Is there no
way to restrict the memory usage of a process on NT? it's a safer way to
test these things.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 17:17:27 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 23:17:27 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <7xhemb744m.fsf@ruckus.brouhaha.com> <a9i41d$dqo$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9i4ef$e3b$1@nets3.rz.RWTH-Aachen.DE>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9i41d$dqo$1 at nets3.rz.RWTH-Aachen.DE...
>..
> However in the PHP version, and how I'm currently laying it out in the
> Python port, the session data is put into hidden form variables.
>..
> I still like this concept
> currently because it allows me to not be dependent on HTTP header cookies
> when I use the language internal session methods.
>..

(And what should be just as important, the complete game state can be saved
by simply saving the webpage on the client-side, and it can be called any
time later on without any server-side session time-outs ruining the fun. I
got the idea from http://www.ifiction.org .)




From gcordova at hebmex.com  Thu Apr  4 09:37:38 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 4 Apr 2002 08:37:38 -0600 
Subject: Threading problem
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB5E@mail.hebmex.com>

> > 
> > This is evil :-(
> 
> It's trade-offs.  It may be less than desirable for multi-processor
> utilization, but it actually carries quite a few implementation
> benefits for the Python core, as well as minimizes the performance
> overhead for maintaining native threading support.  In all but SMP
> cases, it works quite well (barring non-cooperative extensions).
> 

So it's a necessary evil, but evil nontheless. :-(

>
> The only way to change this would be to push all the interlocking down
> to the individual data structure/object levels, which would be no
> small feat (at least for the current CPython implementation), but one
> which I believe gets discussed from time to time.  Wouldn't hold my
> breathe though :-)
> 

Yeah, sounds like what's been happening to the Linux kernel
in the last iterations; locking has become steadily finer-grained,
so that multiprocessor setups will scale better.

I also believe that Python, eventually, will have to follow
that same path.

*Eventually*

And yes, it'll be a lot of work.

And it will impact it's internal architecture heavily.

Salutations!

-gustavo




From jeff at ccvcorp.com  Fri Apr 19 18:10:24 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 15:10:24 -0700
Subject: Trinary operator?
References: <mailman.1019240861.31256.python-list@python.org> <a9pvp0$kqj$1@nets3.rz.RWTH-Aachen.DE> <mailman.1019250162.24332.python-list@python.org>
Message-ID: <MPG.172a359921ba61609896bb@news.nwlink.com>

In article <mailman.1019250162.24332.python-list at python.org>, 
pyth at devel.trillke.net says...
> On Fri, Apr 19, 2002 at 10:47:09PM +0200, Philipp Lenssen wrote:
> 
> > (huge code fragments deleted :-)
> 
> really? wouldn't these few simple lines do what you/one want/s?
> 
> genderdict = { 'english': { 'm':'male', 'f':'female'},
>                'german' : { 'm':'m?nnlich','f': 'weiblich'} }
> 
> return genderdict.get(language,{}).get(gender,'UNKNOWN')
> 

Or, alternately:

genderdict = { ('english', 'm') : 'male',
               ('english', 'f') : 'female',
               ('german', 'm')  : 'm?nnlich',
               ('german', 'f')  : 'weiblich' }

return genderdict.get( (language, gender), 'UNKNOWN' )

using a dictionary keyed on tuples, instead of nested 
dictionaries.  I'm not sure which would be preferable -- I expect 
that it would depend on the typical usages.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mario at kyb291.biologie.uni-bielefeld.de  Wed Apr  3 11:54:38 2002
From: mario at kyb291.biologie.uni-bielefeld.de (Mario Wehbrink)
Date: 3 Apr 2002 11:54:38 MET
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de>
Message-ID: <3caad15e$1_2@news.uni-bielefeld.de>

Am  2 Apr 2002 18:09:53 MET  schrieb/wrote/  
Mario Wehbrink  < mario at bio58.uni-bielefeld.de > in comp.lang.python : 
 
> i wonder wether someone is using pygtk

Thanks for all the hints. I'll take a deeper look at Glade, and the
tutorials.

Thanks

Mario


-- 
Mario Wehbrink
wehbrink at web.de


From tejarex at yahoo.com  Mon Apr 22 09:56:37 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 22 Apr 2002 13:56:37 GMT
Subject: stdin - seek() issue on Linux.
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com>
Message-ID: <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com>

"Srihari Vijayaraghavan" <no at spam.please> wrote in message
news:WMTw8.46875$uR5.106037 at newsfeeds.bigpond.com...
> Hello,
>
> I am a new-bie python user. I have Python 2.2.1 installed on a Linux
> computer. I get the following error message when I try to seek() in
the
> stdin, like:
> #cat /var/log/messages | python test.py
> (where /var/log/messages is over 1MB in size)
>
> Traceback (most recent call last):
>   File "test.py", line 3, in ?
>     sys.stdin.seek(-512, 2)
> IOError: [Errno 29] Illegal seek

By 'cat'ting the random access disk file, you turned it into a
forward-access-only stream of bytes.  You definitely cannot seek
backwards (and maybe not forward - don't know).  Instead, open() the
file and use the resulting file object:

msgs = open("/var/log/messages", 'r')
msgs.seek(-512, 2)
...

Terry J. Reedy





From joe90 at rchrd1.on.wave.home.com  Wed Apr 10 13:05:10 2002
From: joe90 at rchrd1.on.wave.home.com (Joe Ninety)
Date: Wed, 10 Apr 2002 17:05:10 GMT
Subject: readline.insert_text() problem
Message-ID: <ah_s8.35488$cN1.11185@news01.bloor.is.net.cable.rogers.com>

I'm writing a script that prompts the user for information that has a 
default value.  When I call raw_input(), I would like the default value to 
be displayed, which the user can then edit or accept as is.

My understanding is that the following should do that, but it does not work 
on my system (Redhat 7.2, Python2.1):


import readline

readline.insert_text("default")
x = raw_input("Enter value: ")
print x


So when the user runs the program, he sees:

Enter value: default

and can edit the word 'default' and hit enter.

Is there any (other) way to do this??


Thanks,
Don




From philh at comuno.freeserve.co.uk  Tue Apr 16 18:25:40 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:25:40 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com> <a9fu9g$mj3$1@watserv3.uwaterloo.ca> <3CBBAB60.2CC717D7@alcyone.com>
Message-ID: <slrnabp974.ipr.philh@comuno.freeserve.co.uk>

On Mon, 15 Apr 2002 21:41:04 -0700, Erik Max Francis <max at alcyone.com> wrote:
>
>Removing pass seems like a terribly unuseful idea.

I'm not suggesting removing it, merely making it optional. People 
can still use it if they want to.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From marduk at python.net  Sun Apr 28 16:54:17 2002
From: marduk at python.net (marduk)
Date: 28 Apr 2002 15:54:17 -0500
Subject: ANNOUNCE: QNews 0.9 qmail<-->usenet gateway
Message-ID: <1020027258.2322.15.camel@berkowitz>

I am pleased to announce QNews 0.9 is now available for download.  This
is the first release of Qnews.

http://starship.python.net/crew/marduk/software/qnews/

QNews is a gateway between the qmail mail system and Usenet.  It allows
an ordinary user with an account on a system running Qmail to send and
receive articles on Usenet via email.  For example, you can post to
alt.test by sending an email to joeuser-qnews-alt.test at domain.com.  If like
your email program, but wish it had usenet capability then QNews might be
for you.

For more information on qmail, see http://www.qmail.org/

This software requires a system running qmail and Python 2.  It includes
a setup script for easy installation.  It's a small python module and is
fairly easy to use.  Documentation now is pretty sparse, but I am
releasing mainly to get feedback from users and other python
programmers.  I will release version 1.0 with documentation and bug
fixes if it accumulates interest.

QNews is free software released under the GPL.




 Posted Via Usenet.com Premium Usenet Newsgroup Services
----------------------------------------------------------
    ** SPEED ** RETENTION ** COMPLETION ** ANONYMITY **
----------------------------------------------------------        
                http://www.usenet.com


From pj at engr.sgi.com  Wed Apr 24 15:40:42 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 24 Apr 2002 19:40:42 GMT
Subject: stdin - seek() issue on Linux.
References: <mailman.1019617785.428.python-list@python.org> <1019625593.731162@yasure>
Message-ID: <aa71nq$tukai$1@fido.engr.sgi.com>

The seek semantics are lower than the stdio package.  They are
found in the semantics of the kernel.  Look at the open(2)
and lseek(2) man pages on most any Unix or Linux or Posix
emulation thereof.

Seek on a device that can seek is ok - though perhaps of no
affect to subsequent writes if the open was in append mode
(note that such a seek does affect subsequent reads, and hence
at the time of the seek there is no cause for error).  Seek on
a device (such as a pipe) that can't seek is an error.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From jyrinxat at mindspringdot.com  Sat Apr 13 23:51:17 2002
From: jyrinxat at mindspringdot.com (Jyrinx)
Date: Sat, 13 Apr 2002 20:51:17 -0700
Subject: Failure when profiling XML processing
Message-ID: <pan.2002.04.13.20.51.16.512445.4898@mindspringdot.com>

I'm trying to profile some code that parses an XML file. (I've never used
a profiler or done optimizations (this is my first big project) before.)
Here's what happens:

--- snip ---

>>> profile.run("import testCardSet; testCardSet.printAll()")
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/profile.py", line 71, in run
    prof = prof.run(statement)
  File "/usr/lib/python2.2/profile.py", line 404, in run
    return self.runctx(cmd, dict, dict)
  File "/usr/lib/python2.2/profile.py", line 410, in runctx
    exec cmd in globals, locals
  File "<string>", line 1, in ?
  File "testCardSet.py", line 7, in ?
    set = cardSet.LocalCardSet(filename)
  File "cardSet.py", line 26, in __init__
    self.loadFromFile(filename)
  File "cardSet.py", line 39, in loadFromFile
    loadCardSet.loadCardSet(self, filename)
  File "loadCardSet.py", line 241, in loadCardSet
    doc = xml.dom.minidom.parse(nrcsFile)
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/dom/minidom.py",
  line 962, in parse
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/dom/minidom.py",
  line 954, in _doparse
  File "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/dom/pulldom.py",
  line 255, in getEvent
  File
  "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/sax/expatreader.py",
  line 143, in feed
  File
  "/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/sax/expatreader.py",
  line 222, in start_element_ns
  File "/usr/lib/python2.2/profile.py", line 214, in trace_dispatch_i
    if self.dispatch[event](self, frame,t):
  File "/usr/lib/python2.2/profile.py", line 260, in trace_dispatch_call
    assert rframe.f_back is frame.f_back, ("Bad call", rfn,
AssertionError: ('Bad call',
('/var/tmp/python2-2.2.1-root/usr/lib/python2.2/xml/sax/expatreader.py',
132, 'feed'), <frame object at 0x81d1914>, <frame object at 0x8129f34>,
<frame object at 0x81d0bc4>, <frame object at 0x81d139c>)

--- snip ---

(I know that wrapped funny; hope it's understandable)

I don't really know, but this looks like a bug to me. The code runs fine
outside of the profiler.

Any ideas? Thanks!

Jyrinx
jyrinxat at mindspringdot.com

(Is there a more specific place where I should be asking about libraries?)


From dalke at dalkescientific.com  Sat Apr 20 04:59:57 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 02:59:57 -0600
Subject: old Python argument
Message-ID: <a9raig$7of$1@slb4.atl.mindspring.net>

I was delving through the back history of Python and came across
a longish thread in late 1994 about docstrings.  This was just
after they were discussed at the first Python conference.  There were
people who wanted explicit docstring assignment, like

def spam():
  __doc__ =  "blah"

"Is it really that painful to have to write '__doc__ = ' before
your docstring?"

Others who wanted document objects, instead of strings.

Also, one of the reasons against having the change to Python was
because of, and I quote, "the ad hoc nature of the changes."

Argument style sound familiar?

In Guido we trust.  :)

                    Andrew
                    dalke at dalkescientific.com






From frr at easyjob.net  Wed Apr 24 03:18:28 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Wed, 24 Apr 2002 09:18:28 +0200
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net>
Message-ID: <3umccu09j6dkui67hnppulv8ngk71g55s4@4ax.com>

On Wed, 24 Apr 2002 02:17:01 -0000, Jon Ribbens <jon+usenet at unequivocal.co.uk>
wrote:

>In article <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v at 4ax.com>, Fernando Rodr?guez wrote:
>> No. Step (b) processes the order and does give feedback. Step (a) saves in a
>> db some info conatined in a cookie (date of first visit and first referrer). I
>> want step (a) to be seamless.
>
>Must admit I don't have a clue what you mean. Why aren't scripts (a)
>and (b) in the same script?

Because I can't. Step (b) is a url to a third party payment processing company
and it's no tunder my control.

> If they really must be separate scripts,
>why doesn't script (a) just output nothing at all to stdout?

I didn't try it, but I think the browser wouldn't like a cgi url that doesn't
return any html content, no? O:-)



-----------------------
Fernando Rodriguez


From Chris.Barker at noaa.gov  Tue Apr  2 11:18:05 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 08:18:05 -0800
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <a84561$p5g$1@wanadoo.fr> <slrnaabbct.npb.cpbotha@crabtree.cpbotha.net> <4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com> <20020401020912.GE333@lilith.hqd-internal> <mailman.1017681605.21669.python-list@python.org> <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net> <slrnaaivjc.565.zeitlin@seth.lpthe.jussieu.fr> <slrnaaj2o9.a62.cpbotha@crabtree.cpbotha.net>
Message-ID: <3CA9D9A4.4A8AC33D@noaa.gov>

"Charl P. Botha" wrote:
> Tkinter is still the fastest to get up and running on all platforms
> and that that was part of the reason that it was the default GUI of Python.

Actually, the oposite is true: Tkinter is the fastest to get up and
running on all platforms because it is the default GUI of Python. IMHO,
It is the default GUI for Python because of historical reasons. None of
the other options have been mature enough for consideration until the
last couple of years.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From just at xs4all.nl  Fri Apr 19 03:35:04 2002
From: just at xs4all.nl (Just van Rossum)
Date: Fri, 19 Apr 2002 09:35:04 +0200
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <E4Iv8.2972$8D3.78501@news1.tin.it>
Message-ID: <just-2DF829.09350419042002@news1.xs4all.nl>

In article <E4Iv8.2972$8D3.78501 at news1.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> Everything IS "consistently object-oriented" -- polymorphism is the
> principal magic of object-orientedness.  There would be substantial
> loss of power without measurable conceptual simplification if Python
> had no methods (how would I store a bound-method just as any
> other callable then?) or if it had no functions (serious polymorphism
> loss).

I appreciate and agree with what you write, and I like how Python does 
things, too, BUT I can totally see why people find it strange that to 
get the length of an object you write

    len(o)

and to get (eg.) the keys from a dictionary you do

    o.keys()

This seems quite arbitrary, even to me, after maybe 6 years of working 
with Python. If o.__len__() were spelled o.len(), we could easily do 
without the len() builtin. (Note that I don't argue we could do without 
functions: I wouldn't want to miss them for the world...)

Just


From nhodgson at bigpond.net.au  Sun Apr  7 21:31:06 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Mon, 08 Apr 2002 01:31:06 GMT
Subject: Any reason why PythonWin wouldn't start at all?
References: <3cb0d423$1@news.sihope.com>
Message-ID: <up6s8.6799$vc3.26443@news-server.bigpond.net.au>

Spitfire 1500:

> I'm running Windows 2K, and each time i try to start pythonwin, absolutley
> nothing happens. It shows it as running on the task manager, but I can't
get
> to it whatsoever.

   It may be a dead window stopping DDE from working. Try starting with
pythonwin /nodde

   Neil





From gcordova at hebmex.com  Thu Apr 25 16:14:08 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 25 Apr 2002 15:14:08 -0500
Subject: AS/400
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339B25FB@mail.hebmex.com>

> 
> On 25 Apr 2002 10:34:33 -0700
> Terje Johan Abrahamsen wrote:
> 
> > I am trying to write some Python code that moves info from 
> > Excel into AS/400 throuhgt a display program called Rumba.
> > (www.walldata.com) Basically what I want it to do (short version)
> > is that I go to a cell in Excel, copy that cell, changes to the
> > Rumba window, goes to coordinate x,x pastes the number there.
> > Then goes back to Excel, gets a new number, goes to Rumba,
> > changes screens by typing F12, enters info from Excel again
> > in coordinate x,x. It is more to it than this, but if I can
> > figure out this, I can figure out the rest.
> 
> Not familiar with AS/400, but in general, this sort of thing 
> is very fragile. Unexpected user-interaction can cause the
> focus to get set to the wrong window, coordinate, put one of
> the programs in an unexpected state, etc, with possibly
> unhappy results.  Is it not possible to extract the data from 
> Excel via win32com, write it to a portable file and then import
> it on the AS/400 (perhaps transfering the file via ftp)?
> 

Sounds about right what Cliff says.

An alternative to a file would be a database, using ODBC or
something like that, and just using Python to coordinate the
process of reading off the Excel sheet into the AS/400 table.

Sounds like the most fluid option to me, either a file or a
database. Screen-scraping seems fragile.

Good luck !

-gustavo




From sholden at holdenweb.com  Tue Apr 23 10:12:29 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 10:12:29 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com>
Message-ID: <B%dx8.94879$T%5.77750@atlpnn01.usenetserver.com>

>
> >It may or may not make a difference. What's the difference between
> >CompatMySQLdb and MySQLdb?
>
> I'm not sure. I haven't been able to find MySQLdb. I found this Compat
thing
> in MySQL-python-0.9.1.win32-py2.2.exe at
>
http://sourceforge.net/project/showfiles.php?group_id=22307&release_id=51760
>
> Where can I find MySQLdb for win32 _and_ Linux? O:-)
>
Quick follow up: de-installing, and re-installing by running the Windows
installer, gives me a correctly-installed MySQLdb in
<python>/Libs/site-packages. Looks like there must have been a problem wiht
your install, you certainly appear to have the correct installer file. Does
it locate your python2.2 installation when you run it?

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From ball at invalid.null  Wed Apr 10 00:01:17 2002
From: ball at invalid.null (Andy Ball)
Date: Tue, 09 Apr 2002 23:01:17 -0500
Subject: Is Python Sticky?
References: <3cb3a8dd$1@nntp01.splitrock.net> <slrnab7aap.3ir.gerhard@lilith.my-fqdn.de>
Message-ID: <3CB3B90D.70103@invalid.null>

Hello Gerhard!

   GH> Python is an excellent choice both for writing web applications
     > and driving web-based interfaces.

Thanks, that's good to hear!

   GH> I'd head to the "Newbie" page at www.python.org and work thru
     > one of the tutorials.

Okay, I'll give that a try now.

Regards,
   - Andy Ball



From jeremy at alum.mit.edu  Wed Apr 10 23:41:55 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 10 Apr 2002 20:41:55 -0700
Subject: Easter Eggs
References: <mailman.1018391602.4247.python-list@python.org>
Message-ID: <b0f083db.0204101941.1f37632b@posting.google.com>

Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote in message news:<mailman.1018391602.4247.python-list at python.org>...
> Also, is anything else hidden in the Python interpreter that issues
> similar wisdom?  (I know...  I could read the source; but this is
> probably quicker).

I can think of one other small feature of this sort, but I don't think
I should tell you.

Jeremy


From greg at cosc.canterbury.ac.nz  Wed Apr 10 20:36:59 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 11 Apr 2002 12:36:59 +1200
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <a8vpp5$67d$1@panix1.panix.com> <yu99pu18lai2.fsf@europa.research.att.com> <uQUs8.2468$vF6.87600@news2.tin.it>
Message-ID: <3CB4DAAB.A1F9F05F@cosc.canterbury.ac.nz>

Alex Martelli wrote:
> 
> Given all this, if you find out and implement some Ruby Goldberg scheme
> that works (sort of) this way, you'll have code that is totally baffling to
> Pythonista users thereof, because it runs so deeply counter to our
> expectations, our usual social contracts.

Also, said scheme will be highly dependent on implementation
details of the current version of Python, and may well
break in future versions, or alternative implementations
such as Jython. Another good reason to avoid such tricks
if possible.

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


From tdelaney at avaya.com  Tue Apr 30 01:43:48 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 30 Apr 2002 15:43:48 +1000
Subject: Python vs. Perl,  which is better to learn?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2B5@natasha.auslabs.avaya.com>

> From: xiaotsing_79 at hotmail.com [mailto:xiaotsing_79 at hotmail.com]
> 
> I have a lot of experience in ANSI C/C++, I want to learn a new
> language and which it can help me in the work.
> I just want to clear which(python or perl) is more comfort for me?

http://www.python.org/doc/Comparisons.html

Tim Delaney




From dsavitsk at e-coli.net  Wed Apr 17 10:47:22 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 17 Apr 2002 14:47:22 GMT
Subject: Anyone into win32com IE, fetching a picture.. 
References: <m38z7mv6rd.fsf@flognat.myip.org>
Message-ID: <_Vfv8.7559$d7.2253309@newssrv26.news.prodigy.com>

it's a lame question, so here's a lame answer ... ;-)

i saw a post addressing how to do this on comp.lang.visual.basic.misc at
some point.

-d


"Andrew Markebo" <flognat at flognat.myip.org> wrote in message
news:m38z7mv6rd.fsf at flognat.myip.org...
> Hello!
>
> Sorry for asking probably a more windows-like question here, than on
> windows group.. but this is my home.. my.. well you know..
>
> Basically I am using win32com and InternetExplorer to fetch pages, and
> I would also like to fetch the pr0n^H^H^H^H ehm.. snake pictures, yeah
> snake pictures from the pages and store them..
>
> ie = win32com.client.Dispatch('InternetExplorer.Application')
> ie.Navigate(url)
> doc = ie.Document
> while doc.readyState != 'complete': time.sleep(0.1)
>
> Well now doc contains a bunch of info about the stuff in the url, if
> the url was an image, I can get the size, name and so on, but I
> haven't managed to get my hands on the data-part of the image.. Any
> hints??
>
> Yeah I know.. go urllib2 or similar.. but but.. now (using win32com) I
> have cookies, https and so on working good out of the box..
>
> Anyone who recognizes??
>
> Anyone who have a good http/url library that can handle this?
>
>         /Andy
>
> --
>  The eye of the beholder rests on the beauty!




From mgilfix at eecs.tufts.edu  Tue Apr 23 11:43:17 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 23 Apr 2002 11:43:17 -0400
Subject: Python & autoftp
In-Reply-To: <53b7d218.0204230720.686e473f@posting.google.com>; from davies@jkevin.freeserve.co.uk on Tue, Apr 23, 2002 at 08:20:09AM -0700
References: <53b7d218.0204230720.686e473f@posting.google.com>
Message-ID: <20020423114317.C10411@eecs.tufts.edu>

  You'll want to use the ftplib python module. It
allows you to establish ftp connections and control
them programmatically.  Check out the library docs at:
http://www.python.org/doc/current/lib/module-ftplib.html.
This is by far more robust than using bash and expect
since it doesn't rely on an external program and its
possibly changing interface.

                 -- Mike

On Tue, Apr 23 @ 08:20, kevin wrote:
> Hi there
> 
> I'm pretty new to Python and I would like to develop a script in
> Python to automatically FTP. I have already used bash and expect and
> now would like to replace bash entirely with Python.
> 
> Is there any merit in replacing bash with Python??
> Can anyone point me in the right direction.
> 
> 
> Thanks
> 
> 
> Kevin
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (davies)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From aahz at pythoncraft.com  Sun Apr 21 13:33:07 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 13:33:07 -0400
Subject: nested exceptions?
References: <mailman.1019401778.10328.python-list@python.org>
Message-ID: <a9ut4j$dis$1@panix3.panix.com>

In article <mailman.1019401778.10328.python-list at python.org>,
Nick Arnett <narnett at mccmedia.com> wrote:
> [mailto:python-list-admin at python.org]On Behalf Of Aahz
>>
>> That is, if you're checking N items against a list with M items, you're
>> gonna have poor performance algorithmically.  Note that I said "avoid";
>> using 'in' against a list is sometimes the simplest and most
>> straightforward approach, and if it's not being done in a loop, it won't
>> cause too much of a performance hit.
>
>So... although the following is such a short list that I don't think it
>really matters (except that, come to think of it, it is in a loop), it's bad
>form?
>
>if self.item[0:3] in ('Mon','Tue','Wed','Thu','Fri','Sat','Sun'):
>	self.item = self.item[4:]

Yup.  Use a dict, even if constructing the dict looks uglier.  There's
actually a double whammy in your example, because the tuple must be
constructed each time through the if.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From BgPorter at NOartlogicSPAM.com  Thu Apr 11 12:54:38 2002
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Thu, 11 Apr 2002 16:54:38 GMT
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83> <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de> <a91mdd$ck$1@panix1.panix.com> <slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>
Message-ID: <idjt8.103346$w7.8555424@bin6.nnrp.aus1.giganews.com>

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnab9jff.1kn.gerhard at lilith.my-fqdn.de...
> IMO using your real name is polite and whatever you say is not worth
> being said if you can't put your name under it.

I don't know. It worked here in the US for Madison, Hamilton and Jay when
they wrote "The Federalist Papers" under the single pseudonym 'Publius', and
the various anonymous authors who wrote "The Anti-Federalist Papers".






From martin at v.loewis.de  Fri Apr 19 16:38:26 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Apr 2002 22:38:26 +0200
Subject: GNU GPL & Python stand-alone packages from py2exe
References: <i0kit6nn3l4.fsf@Thales.MI.Uni-Koeln.DE> <j4hem7u2ol.fsf@informatik.hu-berlin.de> <MPG.1729fff0170354d59896b8@news.nwlink.com>
Message-ID: <m3pu0v4e65.fsf@mira.informatik.hu-berlin.de>

Jeff Shannon <jeff at ccvcorp.com> writes:

> Correct me if I'm wrong, but... I don't think that the GPL 
> requires that the source be distributed on the same media, just 
> that it be "available".

I think you are (slightly) wrong: You have three options for making
the source available: a) the source must "accompany" the binary, or b)
you must make a written offer to provide the source on request, or c)
you must include the offer you received yourself. Most people chose
option a), which is normally understood as "available through the same
channel as the binary". So if the binary is on CD, the source must also.

Regards,
Martin



From sholden at holdenweb.com  Mon Apr  8 07:55:20 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 07:55:20 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org> <Npor8.8281$q63.818@atlpnn01.usenetserver.com> <3CAE52CA.60F56E3D@python.org>
Message-ID: <oAfs8.26107$Sw5.2561@atlpnn01.usenetserver.com>

[posted & mailed]

"Guido van Rossum" <guido at python.org> wrote in message
news:3CAE52CA.60F56E3D at python.org...
[ ... ]
> By then, so many words had gone back
> & forth that I didn't feel like going back and responding to her piece in
> detail -- others had done that and the discussion had blown up to such a
> volume that I didn't think I could convince any of the PEP's detractors
> anyway.  I apologize to Laura that my judgement came across as rude.
>
Dismissive rather than outright rude, but given the storm of opinion raised
by this PEP, some of it less than well-informed, I can certainly understand
how this happened.

[ ... ]
> > >
> > However it happened, it left me feeling a little embarrassed that I had
> > originally defended Guido's inegrity and suggested that despite his
> > authorship the PEP's acceptance was not a foregone conclusion. I hope I
> > wasn't wrong, but that's not how it looks right now. Oh well.
>
> I'm sorry I embarrassed you.  Despite how it looks, acceptance of the PEP
> was *not* a foregone conclusion.  I have looked carefully for a killer
> reason against its acceptance, and did not find it.  I found a lot of
> disagreement on what people consider a "real" Boolean type.  I did find
> wide agreement on something that I had myself taken for granted, but which
> some proponents of "real" Booleans were challenging: the usefulness of
> Python's ability to test almost *any* object for its truth value, what
> Laura calls something/nothing.
>
No need for any apologies to me. I wouldn't dream of trying to have *my*
opinions constrain *your* behavior, and wouldn't want the responsibility if
they did ;-).

Certainly appearances can be deceptive, and I think anyone who's tracked
Python's development over a number of years knows you *do* take input
seriously, even when it exasperates the ... heck out of you.

> Finally, some folks believe that the new bool should not behave like an
> int.  I've thought about this too.  *Maybe*, if I had added a bool back
> in 1990, I would have agreed and introduced a "pure" bool type.  Right
now,
> if I added such a type, it would not be usable as the return value of
> comparisons and other operations that return a binary result, because of
> backwards compatibility.  And backwards compatibility is a key requirement
> for all language changes (see the division discussion).  Backwards
> compatibility would forever prevent returning a pure bool from existing
> built-in operations.  The only possible transition path has to use an
> "impure" bool like the PEP proposes.  Then the only open item is whether
> this should be considered an impurity that should be deprecated, or a
> permanent wart in the language.  I believe that it's a useful enough wart
> that it can be allowed to stay, and I don't believe that the language
> would be easier to learn or use with pure bools, so I think there's no
> point in deprecating it.  But if years from now our experience points
> out that there is in fact a problem with impure bools, I'm willing to
> reconsider.  However, since I see no way to get there without going
> through the impure bool, I see no reason to wait in implementing the PEP
> *now*, and that's what I have done.
>
> Summary: the impure bool makes it *possible* to transition to a pure bool
> *if* we find that a pure bool is needed; introducing a pure bool from day
> one would be impossible without breaking reams of existing code.
>

Understood. Code breakage is definitely to be avoided. I don't think Python
has too many warts overall (and won't repeat my prejudices on that topic
here).

regards
 Steve








From s.schwarzer at ndh.net  Sun Apr  7 15:42:49 2002
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Sun, 07 Apr 2002 21:42:49 +0200
Subject: [ANN] ftputil 1.1
Message-ID: <3CB0A139.C30BE9D@ndh.net>

ftputil 1.1 is released. You can find it at
http://www.ndh.net/home/sschwarzer/python/python_software.html .

ftputil provides a higher-level interface for FTP sessions than the
ftplib module. FTP servers can be accessed via an interface similar to
os and os.path. Remote files are accessible as file-like objects.

New since version 1.0:

- ftputil now runs under Python 2.1+ (not only 2.2+).
- documentation
- conditional upload/download (depending on local and remote file
  timestamps)
- FTPHost.stat follows links
- a session factory other than the default, ftplib.FTP, can be given
  in the FTPHost constructor; this allows to use classes derived from
  ftplib.FTP (like ftpslib.FTP_TLS from the M2Crypto package)
- several bugfixes (mostly regarding byte count in text mode
  transfers)
- unit test

Stefan


From Chris.Barker at noaa.gov  Fri Apr  5 16:08:16 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 05 Apr 2002 13:08:16 -0800
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <3cab56b5.61383044@news.laplaza.org>
Message-ID: <3CAE1240.D650F4AD@noaa.gov>

Mats Wichmann wrote:
> The simple fact is that they're not going to make any changes in this
> area (even to just fix a batch of scripts) until the next major
> version, which unfortunately is still a number of moons away.  The
> next release was going to be 8.0, but will now be 7.3, if they stories
> I've heard are correct, so a reasonble guess might be "no earlier than
> the end of the year".  Ouch.

Will they do it for 8.0 anyway??? I'd love to hear them say yes, but I
have heard nothing of the sort.

by the way, in regard to dueling pythons, I have found that If I try to
build a Makefile to buld an extension using: make -f Makefile,pre.in
boot, I get a makefile that is using Python 1.5, when what I want is
Python 2.1. Both are installed, as I want the Redhat tools to work.

How can I get this to work??? It's not obvious to me what I would change
in Makefile,pre.in

-thanks,
-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From rmj at slac.stanford.edu  Tue Apr 16 22:39:56 2002
From: rmj at slac.stanford.edu (Roger Jones)
Date: Tue, 16 Apr 2002 19:39:56 -0700
Subject: Problem Writing Files
Message-ID: <3CBCE07C.9A0E0C4@slac.stanford.edu>

I am using ActiveState Python 2.1 under Windows NT  and after opening a
file with:

afile=('test.dat','w')
afile.write('some random test text')
afile.read()

It fills the file with many x's and 0's and also with various other
text.
How can this action be fixed so that it writes the text properly?

Thanks!

-Roger Jones



From jepler at unpythonic.net  Thu Apr 11 15:17:23 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 11 Apr 2002 14:17:23 -0500
Subject: Strange, but interesting recursion problem in PythonWin.
In-Reply-To: <OFEE7F9AE7.905ECF89-ONC1256B98.004FFD45@Z-INDEX.NL>
References: <OFEE7F9AE7.905ECF89-ONC1256B98.004FFD45@Z-INDEX.NL>
Message-ID: <20020411141722.B31425@unpythonic.net>

On Thu, Apr 11, 2002 at 04:51:43PM +0200, F.Baart at sfk.nl wrote:
> I noticed something strange in the PythonWin
> 
> I have the following code:
> def countdown(n):
>  if n==0: print 'ready'
>  else: countdown(n-1)
> 
> I have a maximumrecursionlimit of 1000.
> When I run the countdown function with n=985 it works.
> When I run the countdown function with n=987 it raises a maxrecursionlimit
> error (which you would only expect at n>=999)
> But when I run the countdown function with n=986 pythonwin crashes.
> 
> Remarkable is that when I run the countdown function in a dos shell it runs
> up to n=999 and never crashes.

In Pythonwin, there are probably some execution frames "above" the one
where the interactive interpreter executes.

I don't know why it crashes, though.

Jeff




From dradul at yahoo.com  Wed Apr 17 23:01:58 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Wed, 17 Apr 2002 22:01:58 -0500
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019036193.29349.python-list@python.org>
Message-ID: <a9m5rn$4b8ru$2@ID-99513.news.dfncis.de>

"Hugh Sasse Staff Elec Eng" <hgs at dmu.ac.uk> escribi? en el mensaje
news:mailman.1019036193.29349.python-list at python.org...
On Tue, 16 Apr 2002, P. Alejandro Lopez-Valencia wrote:


On 17 Apr 2002, Martin v. Loewis wrote:

> "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:
>
> > And while you are at it... You have not compiled python itself as
PIC
> > code. No, gcc -shared *doesn't* work in Solaris unless you are using
GNU
        [...]
> Can you elaborate? gcc -shared works fine on Solaris, and binutils
> also works without problems.

>AFAIK this is true.  I've not had problems like this with other
>packages.

As said in other message, this is not the recommended procedure; the
fact
that you can does not mean you should. GNU as does not generate 100%
compatible binary object code, and GNU ld tends to botch Sparc object
code linked against libraries created with Sun tools (that is system
libraries, and you link libc, libnsl and libsocket with almost
everything like it or not).

>
> > Solaris' linker can't build a dynamic loading library/module using
> > object code compiled in position dependent mode.
>
> And rightly so. That should never be a problem, though.
>
> Actually, you *can* talk the Solaris linker to incorporate non-PIC
> text into a shared library. For that, it will make the text segment
> writable. The -mimpure-text option of gcc triggers that feature.

OK, but shouldn't this sort of information be in the configure script?

I don't think so. This part of the lore that makes you a true "Solaris
expert" whatever that may be. I have used Sun boxes for 12 years and
Solaris since it came out ( was that in '94?) and yet, I am no expert at
all :)

But seriously, Solaris is quirky, but no one contributes the appropriate
patches to CVS... (I am not going to, I have other things to worry about
presently).

>
> > This means, that you need to have all external libraries compiled
with
> > the -fPIC flag, even if they are static.
>
> If you link static libraries into Python, I recommend to make the
> modules that use them also static. Then you don't need PIC code for
> those.

Wouldn't this make the memory image of all python programs much larger?
        [...]

Nah. This is misinformed exaggeration. The culprit of unnecessary fluff
is gcc itself, and it becomes worse with each new release. If you are
concerned with executable size, you should add an "-s" flag to the gcc
front-end. This strips all debugging symbol code from the binary object
code. In the case of gcc 3 it can mean a size savings of 60% and even
more.

PIC code is inherently slower, but you won't notice any difference in
speed unless you are still using one of those 12 year old Sparc 1's
(Hell, those where slow even then).

Alejo



From peter at engcorp.com  Wed Apr  3 01:27:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 03 Apr 2002 01:27:11 -0500
Subject: Python compilers
References: <3CAA9786.2070700@moofed.com>
Message-ID: <3CAAA0BF.E8E4659F@engcorp.com>

Rob Hazlewood wrote:
> 
> I'm thinking of learning python, and am a little confused as to the
> overall structure of python programming.
> 
> Would I be right in saying that Python is an interpreted language?

Yes.

> I know it can be run as an interpreted language, however I seem to
> remember people talking about python compilers.

Yes.

What do those two things mean to you?  Python, as with Java, is 
compiled to "bytecode", which is like machine language for a
non-existent processor, sometimes called a virtual machine.
Java calls it that, while Python calls it "the interpreter".
They're basically the same thing.

Unlike with Java, Python compilation is transparent to the
user in most cases, and the resulting .pyc files are cached
on the hard drive and the compilation step is skipped the 
next time the program is run, if the source files do not change.

> Do Python compilers exist? are they efficient?

Yes, and no.  Mostly no, not in the sense you probably mean.
Partially yes, in that sense. :)  Definitely yes if you just
mean "can you get a standalone executable file?", using
several means.  At heart, Python is very dynamic and apparently
somewhat difficult to handle with a compiler.  There have
been and are various efforts along those lines but few
consider it a really high priority issue.

> If there is a python compiler, surely it would be able to generate code
> that is as efficient as a C compiler, resulting in what would be the
> best available OO language? (having simplicity, speed)

No, but relatively few people are overly concerned about that.
Most programmers who are greatly concerned about speed actually
have no particular requirement in mind, they just want "more".
Far too often this is a waste of their time.  Some people look
at ways to optimize programs that take 60 seconds to run and
get run once a day.  In C the same program might take 2 seconds,
but so what?  It might also crash, and it took ten times longer
to develop...

> What is the performance of python like compared to other languages such
> as C and java?

Similar to Java, between 10 and 100 times slower than C, depending
on the application.  The key thing to remember is that few programs
really need maximum performance, and most programs take far more
time to develop than they ever spend running.  Python accelerates
the development well beyond C, or even Java, and is considered
a highly maintainable language as well.  If raw performance
is required, one uses or writes extension modules which are
coded in C and integrated cleanly with Python.

-Peter


From sjmachin at lexicon.net  Mon Apr 29 18:06:14 2002
From: sjmachin at lexicon.net (John Machin)
Date: 29 Apr 2002 15:06:14 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <aaevge$emr$1@peabody.colorado.edu> <93d52e82.0204290618.349a0e5f@posting.google.com>
Message-ID: <c76ff6fc.0204291406.4bd58af5@posting.google.com>

TerryByrne1963 at yahoo.com (Terry Byrne) wrote in message news:<93d52e82.0204290618.349a0e5f at posting.google.com>...

[big snip]
> 
> At first I was just running the re.search on each line. Time: ~ 5 hrs
> for 60K line log file. Started testing with the string.find method to
> make sure a certain error code appeared on the line before running the
> "expensive" re.search, and that cut execution time from ~5 hrs to ~
> 1.5 hrs, good but "no cigar," as they say.
> 
> Then I realized that each line of content could be treated the same as
> you'd treat a switch() structure in C, or a case structure in Pascal:
> if one option is a "hit", then you needn't even check for all the
> other options. So I started pass-ing to the next line of the content
> whenever I got a "hit."

Please re-read your source of information on the pass statement.
"pass" does nothing. Zip. Zilch. It is only a place holder for cases
where you want to do nix but Python syntax insists on a statement
being present. Here are a couple of use cases:

class FubarError(Exception):
   pass

if a and b or c and d:
   pass
else:
   # Late at night, don't want to try inverting
   # the above in my head, don't like not (...)
   do_something()

You can experiment with pass's 'functionality', by putting something
like the following in a file and running it from the command-line:

for i in range(5):
   print i
   pass
   print 'Still here!'
   pass
   print 'Persistent little varmint ...'
   pass
   print 'Now do you believe Uncle John?'

Then change all occurences of 'pass' to '# pass' and see what happens.
Then change all the occurrences of '# pass' to 'continue' and see what
happens.

> That cut my time down to under a minute for a
> 60K-line log file.

Let's have a look at the following snippet of your code:

>      idx = aLine.find('Error Validator:2420')
>      if idx > -1:
>        LINKS_VNOQUERYHITS += 1
>        aLine += '<br>'
>        rfHandle.write(aLine)
>        aLine = None 
>        idx = -1
>        pass
>      idx = aLine.find('Error Validator:2412')
>      if idx > -1:

This just cannot possibly be working [or maybe this code isn't even
executed at all!!!]. You bind aLine to None then attempt to execute
idx = aLine.find('...') which will cause an exception:

AttributeError: 'NoneType' object has no attribute 'find'

... unless of course previously you have done

None = ""

which needless(?) to say is something you should *never* do.

If you can't find out why this is happening yourself, it's time to
head for the  tutor list.

All those lines which bind idx to -1 are pointless. Try changing all
of them firstly to idx = -99999999 and secondly to idx = 99999999. If
either of those makes any difference at all (except a tiny increase in
the run-time) then something drastic has happened: Python is stuffed,
your copy of Python is stuffed, ...

HTH,
John


From gerhard at bigfoot.de  Sun Apr  7 15:19:38 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 21:19:38 +0200
Subject: Regex question about \d
In-Reply-To: <LNBBLJKPBEHFEDALKOLCAELMOLAA.tim.one@comcast.net>
References: <20020407181149.GA28549@lilith.my-fqdn.de> <LNBBLJKPBEHFEDALKOLCAELMOLAA.tim.one@comcast.net>
Message-ID: <20020407191937.GA29202@lilith.my-fqdn.de>

* Tim Peters <tim.one at comcast.net> [2002-04-07 14:59 -0400]:
> [Gerhard H?ring]
> > Can somebody please explain this to me:
> >
> > >>> print re.search(r"\d*", "asdfsfafsw 453 dfsa").group(0)
> >
> > >>> print re.search(r"\d+", "asdfsfafsw 453 dfsa").group(0)
> > 453
> >
> > Why doesn't the first regex match the digits?
> 
> A Python/Perl regexp search matches at the leftmost position it can match.
> Any regexp of the form (whatever)* can match an empty string, so searching
> for any such regexp matches at the start of the string.  If the only thing
> it can match at the start of the string is the empty string, good enough:
> that is a match.

Ah! Ok.

Thanks,

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From gerhard at bigfoot.de  Mon Apr 29 23:57:28 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 30 Apr 2002 03:57:28 GMT
Subject: python2.2
References: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net> <slrnacr3io.nil.gerhard@lilith.my-fqdn.de> <A7hz8.22$Uw4.4633@newsread1.prod.itd.earthlink.net> <aak7qv$bf46e$1@ID-139865.news.dfncis.de>
Message-ID: <slrnacs5hb.p4p.gerhard@lilith.my-fqdn.de>

Dave Kuhlman wrote in comp.lang.python:
> The testing version is woody.  So change "stable" to "woody" in 
> your sources.list.  (Perhaps you could instead change "stable" to 
> "testing", but I'm not sure about that one.)

This would update the entire distribution to woody the next time a
"apt-get update; apt-get upgrade" is issued. This may not be the
desired effect.

> Not to distract you, but ... If you have gcc on your system,
> downloading, building, and installing the source is very straight
> forward. [...]

Indeed, building Python yourself has great benefit. One argument for
using Debian packages, however, is that you can access lots of useful
packaged Python libraries that have already been debianized:

gerhard at lilith:~$ apt-cache pkgnames|grep python|wc -l
    255

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From jepler at unpythonic.net  Tue Apr 23 23:06:27 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Tue, 23 Apr 2002 22:06:27 -0500
Subject: String handling bug in Python
In-Reply-To: <3cc6036f$1@news.nz.asiaonline.net>
References: <3cc5d5c8$1@news.nz.asiaonline.net> <3cc6036f$1@news.nz.asiaonline.net>
Message-ID: <20020423220627.B2102@unpythonic.net>

On Wed, Apr 24, 2002 at 12:57:02PM +1200, Colin Brown wrote:
> Thanks guys
> 
> I did not think to look up this  "documented feature", sorry.

Well, how else would you suggest that "raw" strings should work?  For
instance, if you want to let
    r'\' == '\\'
how will you write the raw string that equals "\\'"?

You can't write r'\'' under your system because that's the raw string r'\'
(your system) followed by another (unmatched) '.

I suspect it turns out that no matter what scheme you decided on, there
are a few strings you can't express raw-ly.  I suspect that "no odd number
of trailing backslashes" is the best that can be done.  As far as I can
tell, your proposed scheme would forbid "apostrophe following an odd number
of backslashes" which excludes a greater fraction of strings from be
written raw-ly than the one Python actually employed.

The shortest way to write an almost-raw string with an odd number of
trailing backslashes is probably
    r'.......' "\\"
due to the builtin string concatenation that the parser performs.

> It still looks like a cop-out to me and can join my "ugly list" along with
> leading-zero octal numbers.

This is a feature inherited from C, as you probably know.  I noticed the
other day to my surprise that int("0377") != 0377, though (1.5.2 through
2.3a0).  I'll have to admit that I had non-computer-savvy colleagues
run into this one though, when I worked in the USGS back in '96 or so.
("How come this script doesn't work when I search for month >= 09 in
the file?")  I've also run into it when tcl is used as a scripting
language, especially when the "normal" (not base-10) number parsing
machinery was used on the string contents of entry widgets.  This may
well bite us again as we make our next-generation software scriptable
in Python... (Of course, I'll be excited when we can tell the engineers
that the expression '8+3/4' will actually give them 8 3/4 inches and
not 8 inches .. this despite my personal misgivings about future_division)

Jeff




From aleax at aleax.it  Mon Apr 29 04:28:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 08:28:42 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <mailman.1020065634.29503.python-list@python.org>
Message-ID: <_u7z8.75194$vF6.2248951@news2.tin.it>

Delaney, Timothy wrote:
        ...
> I also personally don't think lamdas should be removed. However, if a
> lambda exceeds a (BDFL-pronounced) number of opcodes it should be a syntax
> error <0.5 wink>. This should reduce most lambdas to one or two
> expressions.

All lambdas are alredy "reduced" to exactly one expression.  Somehow
constraining that expression to be "simple" is a laudable goal, but -- I
dunno: it seems to me that putting further constraints, and moreover
ones hard to verify precisely for a programmer who wants to be thinking
about Python source code, not about bytecodes that come out, is hardly
the way of solving lambda's woes.


Alex



From jbwellsiv at yahoo.com  Sat Apr 13 16:49:23 2002
From: jbwellsiv at yahoo.com (JB)
Date: 13 Apr 2002 13:49:23 -0700
Subject: Python-tkinter and canvas tags
Message-ID: <ca58272d.0204131249.3f7e8663@posting.google.com>

Alright. I finally decided to give learning python a go this week and am
working on a rather simple gui app with Tkinter.

I have an odd problem and my hope is that someone can clear it up for me.

I have a main class (cfMain.py) that will serve as the container for the gui
portion of the application. cfMain.py defines a canvas widget on which I
will draw a number of shapes, each shape representing a different element in
a T1/T3 circuit.

A network element in the circuit is represented by a square. I've created
another class called gSNC.py to contain the gui logic for these network
elements.

Basically, when a user presses a button on the main app, a square is added
to the canvas. The gSNC constructor is called by an button event handler in
cfMain.py, passing a reference to the canvas widget. gSNC then uses this
canvas widget to render the square with the create_polygon method.

The problem arises when I try to handle any mouse entered/exited events. I
want the square to change color when a mouse enters or exits. If I move
everything (create_polygon call included) into cfMain.py and bind the event
handlers there, I'm able to change the color by using something like:

self.theCanvas.itemconfig(CURRENT,
fill=cfMain.SNCHLColor)

However, I'd like the event-handling methods to be defined in the class
gSNC.py, so I bind the events in the following manner:

# create a polygon

tmp = self.canvasRef.create_polygon(...omitted...)

self.canvasRef.tag_bind(tmp,
"<Any-Enter>", self.handleEnter)

handleEnter is defined as:

def handleEnter(self, event):
self.canvasRef.itemConfig(CURRENT,
fill=self.HLColor)

When I run it this way, the events are definitely being handled, but I get
an error stating that "NameError: Global name 'CURRENT' is not defined".

I've been trying to jog my memory of Tk's handling of canvas tags and
looking through references, but nothing has explained why this might be
happening yet. It may be a misunderstanding on my part, but I assumed that
since canvasRef is a reference the to canvas widget in cfMain.py, the
CURRENT tag should automatically be assigned to whatever is under the mouse
(as it is when I contain all polygon creation under the cfMain class). It
seems intuitive that you'd want to have this functionality, rather than
grouping all code that draws anything on a canvas in one class.

Can anyone clear my head?

Also, I've been trying to find info on how Tkinter handles event propagation
between classes. If anyone could point me in the right direction, I'd
appreciate it.

Thanks!
John


From neal at metaslash.com  Thu Apr  4 19:36:00 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 04 Apr 2002 19:36:00 -0500
Subject: 1 .__nonzero__() => memory fault (2.2 win32)
References: <a8ipmo$qq0$1@216.39.172.122> <3CACED9C.DF372C5E@alcyone.com>
Message-ID: <3CACF170.8F9A10C2@metaslash.com>

Erik Max Francis wrote:
> 
> Bengt Richter wrote:
> 
> >  Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> >  Type "help", "copyright", "credits" or "license" for more
> > information.
> >  >>> hasattr(1, '__nonzero__')
> >  1
> >  >>> 1 .__nonzero__()
> >
> > (You get a memory fault).
> 
> Whoops, I get the same thing.

This is fixed in 2.2.1, which is due out in about a week:

[neal at epoch 2.2]$ ./python 
Python 2.2.1 (#3, Apr  4 2002, 19:34:25) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> 1 .__nonzero__()
1
>>> 

Neal


From larry.tami at gte.net  Sat Apr 20 09:26:42 2002
From: larry.tami at gte.net (Larry Wright)
Date: Sat, 20 Apr 2002 13:26:42 GMT
Subject: Question about IDLE
References: <0p5w8.17527$CH.4180@rwcrnsc52.ops.asp.att.net> <a9r5cn$e2e$1@nsnmpen2-gest.nuria.telefonica-data.net>
Message-ID: <m0ew8.1284$bu.925@nwrddc03.gnilink.net>

I would agree with this. Fancy IDE's can tend to confuse things when 
learning a new language.

H. Safarzadeh wrote:

> Matt wrote:
> 
>> I'm trying to learn Python, and I was wondering if the IDLE program was
>> sufficient enough for basic programming, or should I find some sort of
>> text editor?
>> I'm currently using a win32 box.
>> 
>> Also, any links to tutorials or general information about Python that is
>> not listed on the Python website would be much appriciated.
>> 
>> -Matt
> 
> I think it is sufficient even for complex programs. You will be
> comfortable with it.



From h_schneider at marketmix.com  Mon Apr  8 05:54:31 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Mon, 8 Apr 2002 11:54:31 +0200
Subject: Read SFF (Structured Fax Format) files ..
Message-ID: <a8rp19$sto$06$1@news.t-online.com>

Hi,

is there a module, which is capable of reading or converting SFF files to
JPG ?
PIL has no native support for that ...

All the best,
Harald





From philh at comuno.freeserve.co.uk  Tue Apr 16 18:28:44 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:28:44 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu>
Message-ID: <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>

On Tue, 16 Apr 2002 10:17:30 -0700, Russell E. Owen <owen at nospam.invalid> wrote:
>In article <slrnabmoci.dhk.philh at comuno.freeserve.co.uk>,
> philh at comuno.freeserve.co.uk (phil hunt) wrote:
>
>>OK, I've decided to write a PEP. Preliminary to this, I'd like 
>>people's comments.
>>
>>I wish to make the pas startement optional. That is, in any Python 
>>program where there is a pass it can be left out and the meaning of 
>>the program is unchanged...
>
>I'd go along with it and offer a third rationale:
>
>The need for pass is a pain when commenting out entire blocks of code 
>(e.g. while debugging). Instead of just commenting out the code, one 
>must also add a "pass" statement, and then try to remember to take it 
>out again when re-enabling the code.

That's true, and one that has happened to me.

>In my opinion an optional pass is a very good thing -- sometimes you 
>want to be able to assert "this block intentionally left blank", but 
>sometimes it makes more sense to just accept a truly empty block.
>
>Thanks for proposing this. I'm sure you'll get lots of horrified 
>reactions

I don't see why: people who like pass can still use it.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From alanmk at hotmail.com  Fri Apr  5 05:05:29 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 5 Apr 2002 02:05:29 -0800
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com>
Message-ID: <f72dac0d.0204050205.4cb3e8ac@posting.google.com>

brian d foy wrote:
> for our April 2002 issue (April Fool!), The Perl Review replaced
> its Perl content with Python articles, so Python folks might want
> to read this issue:
> 
>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf

Forgot to mention: Does this mean that the Golf competition is open to
Python entries :-)

The golf competition is a neat idea. Python hasn't completely
suppressed the hacker in me :-)

Cheers,

Alan Kennedy.


From tejarex at yahoo.com  Mon Apr  8 20:39:31 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 09 Apr 2002 00:39:31 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com> <477762c2.0204081003.75867687@posting.google.com>
Message-ID: <7Lqs8.58234$%i.5437909@bin5.nnrp.aus1.giganews.com>

"Dave Brueck" <dave at pythonapocrypha.com> wrote in message
news:477762c2.0204081003.75867687 at posting.google.com...
> Paul Rubin <phr-n2002a at nightsong.com> wrote in message
news:<7xadshtrj1.fsf at ruckus.brouhaha.com>...
> > Stable doesn't just mean old scripts run without change in new
> > interpreters.  It also means new scripts usually run without
change in
> > old interpreters.
>
> How could that possibly work?

You declare the old interpreter to be bug free -- ie, that its
behavior *is* the (unchanging) specification.  Then you limit yourself
to invisible changes -- ie, those that only change (hopefully reduce)
time and space requirements.  (This allowance requires, of course,
that one define dependencies on running time and size to be outside
the specification and, in a sense, illegitimate.)

Terry J. Reedy





From martin at v.loewis.de  Wed Apr 10 13:38:44 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 10 Apr 2002 19:38:44 +0200
Subject: Building an extension that calls the Python interpreter recursively.
References: <5%Zs8.7857$de1.379233@news3.calgary.shaw.ca>
Message-ID: <m3it6z313v.fsf@mira.informatik.hu-berlin.de>

"Parzival Herzog" <parz at shaw.SpamBucket.ca> writes:

> I would like to know if this scheme is feasible

Yes, it sounds good.

> 1) Can the Python interpreter be re-invoked from an
> extension module?

Yes; anything that you can write in Python can also be invoked as a
series of API calls. This particular API is PyRun_String.

> 2) Will it process its evaluation of S in the same
> interpreter context that originally invoked ML1.process?

Yes, unless you play with threads (and even then this is easily
achieved).

> I.e. will results left by PyRun_String be accessible
> after process() returns?

Depends on what "results" is. If you execute statements, side effects,
such as modifying globals or locals, or changing attributes of objects
reachable through globals or locals, will persist.

If the string you run is an expression, you need to worry about
returning the result.

> 3) Are there gotchas or constraints on the global
> interpreter lock that I should be aware of under these
> circumstances?

As long as you hold the lock, no other Python threads will run. So if
this is a long-running action, you should release the lock inside
.process(). Then, before calling into Python, you need to create a new
thread state, acquire the lock for this thread state, and call into
Python. As a result, tracebacks will "end" in the callback.

As an alternative, you might consider just to keep the interpreter
lock.

Regards,
Martin


From just at xs4all.nl  Fri Apr 26 06:28:10 2002
From: just at xs4all.nl (Just van Rossum)
Date: Fri, 26 Apr 2002 12:28:10 +0200
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>
Message-ID: <just-9E8624.12281026042002@news1.xs4all.nl>

In article <C2cBaXANXSy8Ewbb at jessikat.demon.co.uk>,
 Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

> With 2.2.1 I'm puzzled why the first returns the local x and the second
> the global I suppose x=x rather than z=x is the problem, but don't
> understand why. I opposed local scopes originally as I liked the two
> scope rule, but this seems more complex than I imagined.
> 
> >>> x=2
> >>> def bingo(x):
> ...     class A:
> ...             z=x
> ...             def get(self):
> ...                     return self.z
> ...     return A().get()
> ... 
> >>> bingo(4)
> 4
> >>> def bango(x):
> ...     class A:
> ...             x=x
> ...             def get(self):
> ...                     return self.x
> ...     return A().get()
> ... 
> >>> bango(4)
> 2

What a coincidence:
   http://www.python.org/sf/532860
   http://mail.python.org/pipermail/python-dev/2002-April/023427.html

Just


From fdrake at acm.org  Mon Apr  1 15:33:00 2002
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 1 Apr 2002 15:33:00 -0500
Subject: New home for development version of docs
Message-ID: <15528.50172.390695.897521@grendel.zope.com>

The "in-developement" version of the documentation, previously found
at python.sourceforge.net, is now available from the development area
on python.org:

	http://www.python.org/dev/doc/


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From brian at sweetapp.com  Fri Apr 26 15:00:15 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 26 Apr 2002 12:00:15 -0700
Subject: Python's Lisp heritage
In-Reply-To: <20020426145328.A31190@node0.opengeometry.ca>
Message-ID: <001201c1ed54$99d56760$445d4540@Dell2>

William Park wrote:
> > Lisp:   (+ 1 2 3 4 5 6 7 8 9 10)
> > Python: 1+2+3+4+5+6+7+8+9+10
> 
> Behold...
> 	  seq -s+ 1 10
> :-)

Behold....
	55

:-)






From rmelson at gilia.nmsu.edu  Tue Apr 30 14:01:24 2002
From: rmelson at gilia.nmsu.edu (Bob Melson)
Date: 30 Apr 2002 18:01:24 GMT
Subject: Python vs. Perl, which is better to learn?
References: <mailman.1020147892.13578.python-list@python.org>
    <aamjat$qku$1@panix1.panix.com>
Message-ID: <aamm5k$kv3$1@bubba.NMSU.Edu>

In article <aamjat$qku$1 at panix1.panix.com>,
	aahz at pythoncraft.com (Aahz) writes:
> In article <mailman.1020147892.13578.python-list at python.org>,
> Andrew McNamara  <andrewm at object-craft.com.au> wrote:
>>
>>I'd go as far as to say the
>>sendmail.cf language is simpler, easier to learn and maintain than perl.
> 
> .sigfile?

Look at it this way -- you pick the best language tool to do the job at
hand.  If python is the better language for that job, you use python; if
it's perl, then you use perl; if it's some other language, you use _that_
language.  If you're asked to build a house, you certainly don't confine
yourself to a single tool -- you use the tools appropriate to the job you
have to do; why should it be different with scripting languages?

Bob Melson



From skip at pobox.com  Mon Apr 22 17:05:10 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 22 Apr 2002 16:05:10 -0500
Subject: using automake/autoconf to build extension module?
Message-ID: <15556.31494.400963.544446@beluga.mojam.com>

Can someone explain what automake incantations are necessary to build a
Python extension module?  I'd prefer not to use distutils if possible, since
the rest of the project is using automake/autoconf/libtool for build
configuration.

I tried the following:

    lib_LTLIBRARIES = alfmodule.la
    alfmodule_la_LDFLAGS = -module -avoid-version -export-symbols-regex initalf
    alfmodule_la_SOURCES = alf.h alfmodule.c
    alfmodule_CFLAGS = -I$(prefix)/include/python$(PYTHON_VERSION)

but it blows up during the compile because it has no -I flag referencing the
include directory.  (Obviously the last line above is incorrect in some
respect.  It never gets into the cc command.)

Thx,

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From wideopen at shinbiro.com  Wed Apr 17 20:23:11 2002
From: wideopen at shinbiro.com (Haegi Kim)
Date: 17 Apr 2002 17:23:11 -0700
Subject: Type Conversion to COM Variant
Message-ID: <a4488154.0204171623.2e1f7cd5@posting.google.com>

Hi, experts,

I have a COM client which gets a reference to a COM server.
The COM client has a exec expression in order to execute a script in
which the reference to the COM server is used.
It works find with the COM server in InProcServer mode, but doesn't in
LocalServer and raises the following error.
-----------------------------------------------------------------------------
exceptions.TypeError: Objects of type 'instance' can not be converted
to a COM VARIANT"
-----------------------------------------------------------------------------
In what case the type conversion causes the error? And what should I
do?
Thanks in advance.


From jb at cascade-sys.com  Mon Apr 29 17:18:28 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 14:18:28 -0700
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <20020429174425.GN16217@devel.trillke>
Message-ID: <3CCDB8A4.C410CFE4@cascade-sys.com>

holger krekel wrote:

> there are too solutions. choose yourself :-)

THREE if you include LAMBDA.

newnums = filter(lambda x: x is not None, nums )

This is like the first solution but does not require creating a function or
giving it a name.

> nums = [ 1,2,3,None,2,3]
>
> #version1
> def f(x): return x is not None
> newnums = filter(f, nums)
>
> #version2
> newnums = [ x for x in nums if x is not None]

This got me thinking, so I wrote a little test program to compare the speed of
the various alternatives.

With a large number of elements in the list (500000), Lambda/filter was the
fastest, function/filter was a close second and Comprehensions was a DISTANT
THIRD.

Here's my data on a Windows NT, Dual-P600 CPU machine:

    [C:/usr/jb]% python test.py
    lambda1: 1.29121285628
    lambda2: 1.31910440572
    lambda3: 1.31411255221
    func1: 1.43611083136
    func2: 1.44018521068
    compre: 4.75089779296
    loop: 4.33176244194

Geeze Louise, over 3.5X slower!!  Sheesh, even the loop was faster.

Results were similar, if less dramatic, on my dual P180MMX Linux server:

    [cascade jb]$ python test.py
    lambda1: 5.81
    lambda2: 5.78
    lambda3: 5.77
    func1: 5.91
    func2: 5.88
    compre: 6.47
    loop: 6.25

Also, the difference is less for smaller lists, though list comprehensions always
came out last.

The three lambda and two function forms included slight variations.

     Lambda1 the lambda expression was assigned to a global variable outside
     the loop.

     Lambda2 includes the lambda expression inline as an arg to filter.

     Lambda3 assigns the lambda expression to a local variable inside the
     loop.

     Func1 defines a local function to pass to filter.

     Func2 defines a global function to pass to filter.

     The comprehension and for loop were coded the only way they could be.

Try it yourself!  Code may be downloaded from:

    http://cascade-sys.com/~jb/Pythonetics/test.py

I think I heard somebody say that Lambda was faster but I was AMAZED at how much
more efficient it can be.  List comprehensions are SLOW by comparison.  On this
datum alone I would be bitterly opposed to getting rid of LAMBDA().

Finally, in the original application, if numeric values of 0 are impossible then
the tests in all cases can simply test the item in question, instead of "not is
None", which I expect would speed things up slightly.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From geoff at gerrietts.net  Tue Apr 16 20:38:45 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 16 Apr 2002 17:38:45 -0700
Subject: list_a-list_b
In-Reply-To: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>
References: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>
Message-ID: <20020417003845.GJ20804@isis.gerrietts.net>

Quoting John (nobody at nobody.com):
> 
> list_a=[1,2,4], list_b=[2,4]
> list_minus(list_a,list_b)
> print list_a #will print out a=[1] after removing all elements ([2,4])
> appearing in list_b

Lots of solutions for this problem, many of which are more general
than others. The simplest way is probably:

def list_minus(a, b):
  return [ x for x in a if not in b ]

That's probably not the best, if your measure of best is "takes the
least amount of time", once the lists get relatively large. But it's
probably acceptable for most cases, especially due to its relative
compactness and simplicity.

Thanks,
--G.

-- 
Geoff Gerrietts                <geoff at gerrietts net>
"I have read your book and much like it." --Moses Hadas




From nowhere at blue.com  Tue Apr 23 18:08:32 2002
From: nowhere at blue.com (RJ)
Date: Tue, 23 Apr 2002 22:08:32 GMT
Subject: Help w/extension for a single file ANSI C source?
Message-ID: <AXkx8.77$Ug5.68418669@newssvr13.news.prodigy.com>

Can anyone get me started on a Pyhton extension?
I'd like to be able to access the internal functions of an ANSI C program
which consists of just prog.h and prog.c (as below).
It compiles and runs w/Borland on MS and gcc on Linux, but I'm going around
in circles trying to make my first Python extension.
Otherwise I might next try making a dll with VS6 and using calldll.py, which
I at least have experience with.
Is SWIG the 'best' method for extensions? This doesn't need libraries and
does not have an __init function, just straight C.


/* *************************************** */
/*  prog.h - function prototypes  */
/* *************************************** */
#define HYPERDIAG           (int)'A'
#define RECTANGULAR         (int)'R'
#define LINEAR          (int)'L'
#define SPLINE          (int)'S'
#define CUSTOM          (int)'C'
int allocate_prog(int num_state,int *qnt_state,
           int num_resp, int num_cell,
           int memory, int rfield_shape, int collision_flag);
int train_prog(int prog_id, int *state, int *respns, int beta1, int beta2);
int prog_response(int prog_id, int *state, int *respns);
int clear_prog_weights(int prog_id);
int deallocate_prog(int prog_id);
int prog_memory_usage(int prog_id);
int get_prog(int prog_id, int index, int *buffer, int count);
int put_prog(int prog_id, int index, int *buffer, int count);


/*     prog.c */
#include <stdlib.h>
#include <fcntl.h>
#include <sys/stat.h>
#include <malloc.h>
#include "prog.h"
#ifndef O_BINARY
#   define O_BINARY 0
#endif
#define TRUE  1
#define FALSE  0
/* ***********************************************************************
The following definitions affect static data storage size, but NOT execution
time.
**************************************************************************/#
define NUM_NETWORKS 8
#define MAX_STATE_SIZE 64
#define MAX_RESPONSE_SIZE 8
#define MAX_GEN_SIZE 256
#define RF_TABLE_SIZE 128
/* ************************************************************************
/**************************************************************************
 The following store the major specs and pointers for the current prog
**************************************************************************/
static int first_call = TRUE;   /* is set to FALSE after init      */
static int *ap_a[NUM_NETWORKS+1];   /* pointers to all prog memories   */
static int *ap;                 /* pointer to current prog memory  */
static int ap_size[NUM_NETWORKS+1]; /* sizes in bytes of prog memories */
static int collide;         /* TRUE/FALSE control of hash collisions */
static int *qnt;                /* pointer to input quantization   */
static int *disp;               /* pointer to the displacement vector */
static int *rfieldt;            /* pointer to the rfield function table */
static unsigned int rndseq[2048];   /* Random number table for hashing */
static int indexes[MAX_GEN_SIZE];   /* prog mapping indexes            */
static long rfield[MAX_GEN_SIZE];   /* RF function magnitudes */
static long sum2_rfield = 0;    /* summed squared RF function magnitudes */
/**************************************************************************
The following functions are used internally by the prog code and can not
be called externally!
 **************************************************************************/
static void stoap(int *state)
{
    int i, j, k;
    long index, min_d, test_d, rf_index;
    long sum;
    int hash_tag;
    static long base[MAX_STATE_SIZE], qstate[MAX_STATE_SIZE];
...
}

static void setup(int prog_id)
{
...
}

/**************************************************************************
The following prog functions CAN be called externally!
**********************/
int allocate_prog(int num_state,int *qnt_state,int num_resp,
                int num_cell,int memory,int field_shape, int collide_flag)
{
    int i,j;
    int *pntr;

    /* initialize data structures on first call to driver */

    if (first_call) {
        genmap();
        for (i = 0; i <= NUM_NETWORKS; i++) ap_a[i] = 0;
        first_call = FALSE;
    }
...
}

etc...


Ray Schumacher
http://RJS.org




From charleshixsn at earthlink.net  Tue Apr 23 10:57:52 2002
From: charleshixsn at earthlink.net (Charles Hixson)
Date: Tue, 23 Apr 2002 07:57:52 -0700
Subject: PEP 276 -- What else could iter(5) mean?
References: <mailman.1015455997.18295.python-list@python.org><6qpu2gzktu.fsf@abnoba.intevation.de> <mailman.1015517623.2914.python-list@python.org> <u8g1dbkklg9i18@news.supernews.com>
Message-ID: <3CC57670.9060308@earthlink.net>

John Roth wrote:
>...
> I'm against the proposal, simply on the basis that I don't find
> 5.iter to have a blindingly obvious meaning. It's ok if you explain
>...
> 
> 

And to me the meaning *is* blindingly obvious ... and different!

to me 5.iter should return an unlimitedly long string of 5's.



From jblazi at hotmail.com  Mon Apr  8 15:21:10 2002
From: jblazi at hotmail.com (Janos Blazi)
Date: Mon, 8 Apr 2002 21:21:10 +0200
Subject: PyQt question (connecting signals)
Message-ID: <3cb1ed1f$1_3@news.newsgroups.com>

I have a very simple application. Can I connect a signal from a QPushButton
to a normal python function without using classes at all?

J.B.




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From mhammond at skippinet.com.au  Fri Apr 19 21:56:25 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 20 Apr 2002 01:56:25 GMT
Subject: Keyboard interrupts in Windows Python
References: <2830c89c.0204190721.633cd445@posting.google.com>
Message-ID: <3CC0CACB.8060406@skippinet.com.au>

Andrew Gregory wrote:
> I am using Python to control experiments from the interactive Window
> of Pythonwin (via some simple library routines in C to write data to
> interface chips). Sometimes I wish to interrupt an experiment before
> it has finished, ideally by using Control C to raise an exception. But
> Control C has no effect.
> How can I overcome this problem? Is there another control key which I
> should be using?
> 
> It would be equally acceptable to raise an exception by clicking on a
> button - but I am hoping to avoid Windows programming and work
> interactively for the moment.
> 
> To give an example, if I typed time.sleep(1000) how would I terminate
> this command under Windows?
> 
> Andrew Gregory, National Physical Lab., UK.

There are 2 relevant bugs:

https://sourceforge.net/tracker/?func=detail&aid=439992&group_id=5470&atid=105470

and

https://sourceforge.net/tracker/?func=detail&aid=231273&group_id=5470&atid=105470

I hope to look at them soon, although I did play with 
SetConsoleCtrlHandler() etc a few years ago and struggled :(

Mark.



From shalehperry at attbi.com  Mon Apr 22 18:46:26 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 22 Apr 2002 15:46:26 -0700 (PDT)
Subject: calendar.py doesn't use lists???
In-Reply-To: <aa2290$431$1@peabody.colorado.edu>
Message-ID: <XFMail.20020422154626.shalehperry@attbi.com>

> 
> In [1]: from calendar import month_abbr, month_name
> 
> In [2]: month_abbr ?
> Type:           instance
> Base Class:     calendar._localized_name
> String Form:    <calendar._localized_name instance at 0x81a8cb4>
> Namespace:      Interactive
> 
> In [3]: for m in month_abbr: print m
>    ...:
> Segmentation fault
> 
> Python shouldn't segfault ever, period (that is, unless it's calling your own
> extension modules and those are causing the segfault).
> 

2.2.1 works just fine here.




From achim.domma at syynx.de  Sat Apr 27 05:13:33 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Sat, 27 Apr 2002 11:13:33 +0200
Subject: problems redirecting unittest output to sys.stdout
Message-ID: <aadplo$rs7$06$1@news.t-online.com>

Hi,

I want to redirect the output of a unittest to a textfile (on Windows). The
default stream is sys.stderr which can not be redirected on windows. Until
now I started my tests like this :

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

To redirect the output I tried the following :

if __name__ == '__main__':
    import sys
    unittest.main(testRunner=unittest.TextTestRunner(stream=sys.stdout))

This way the output is redirected to sys.stdout, but the command line
parameters are ignored. If I call my script with the '-v' parameter, I get
no verbose output. I have no idee where the parameters get lost !?

any idea ?

greetings
Achim




From aleax at aleax.it  Fri Apr 12 17:25:21 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 12 Apr 2002 21:25:21 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com>
Message-ID: <5hIt8.10742$b62.296832@news1.tin.it>

Huaiyu Zhu wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> 
>>Personally, I'd rather see curry built-in -- it IS easy to write curry
>>in Python 2.2, but it's such a wonderful idiom (and would replace so
>>many lambda's;-) that I'd love to see it officially blessed.  Not that
>>I expect it to be (and nor will my heart be broken because of this:-).
> 
> Do you have in mind a special syntax for curry?

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549

see my curry-by-closure comments.  I think that having positional
args given at call-time concatenated after those given at curry-time
("if any" in either case), and kw args given at call-time taking precedence 
over those given at curry-time (ditto), is most natural and useful.
I think the syntax you propose is equivalent (for positional args --
you don't mention kw args).


Alex



From juenglin at informatik.uni-freiburg.de  Tue Apr 30 12:47:31 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 30 Apr 2002 18:47:31 +0200
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <199334c5.0204260645.3af87dc5@posting.google.com> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de> <vIxz8.81546$vF6.2424489@news2.tin.it>
Message-ID: <xdulmb56sm4.fsf@praline.informatik.uni-freiburg.de>

Alex Martelli <aleax at aleax.it> writes:

> Ralf Juengling wrote:
>         ...
> > BTW: It is not easy to add __iter__ to list's meta-type, is it?
> 
> And giving list a different metaclass than all the other types
> would no doubt be weird.

agree.

> Further, it would be utterly weird if 'take' was introduced,
> operated correctly on iterables (including iterators), but
> returned a list rather than an iterator!  It would freak me
> out... we finally have a 'lazy' paradigm, and we'd just
> throw it away like this?!  take and drop should both accept

I see what you mean from what I left out (i.e. that your version
of take(n, it) yields an iterator that raises StopIteration at the 
latest after n calls to it.next()).

However, I think it's good idea to let take() yield a list of
elements and drop() an iterator. Why this difference? Iterators 
are potentially "lists whithout end". But they always have *one* 
tail, i.e. a beginning.
I think of take() as of a convenient way to 'take' some elements
from the beginning of the list at once (which I would do otherwise 
by repetitive calls the iterators next method) and of drop() as of
a convenient way to drop some of these elements I'm not interested 
in.

So we wouldn't through away the 'lazy paradigm' with a take()
that yields a list. We would just use take() in cases where we 
actually want some elements at once, e.g. here:

> >>> List2D = list(take(5, iter(list, '')))
> >>> List2D
> [[], [], [], [], []]
> >>>

Ralf



From anthony at interlink.com.au  Tue Apr  9 01:57:25 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Tue, 09 Apr 2002 15:57:25 +1000
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ? 
In-Reply-To: Message from Matthew Baulch <matt@greenroom.com.au> 
   of "Tue, 09 Apr 2002 15:10:58 +1000." <3CB277E2.8060505@greenroom.com.au> 
Message-ID: <200204090557.g395vPf02009@mbuna.arbhome.com.au>

For what it's worth, I've seen email from someone @ redhat about this -
they're not making this change for 7.3, because 7.3 is a maintenance
release. Hopefully it will be in 8.x, whenever that is.

(As to which version they should run, well, there's 2.1.3 and 2.2.1 
to choose from now - if they want to run zope, then the former is 
needed, otherwise the latter is probably better. Dunno what they plan,
tho)

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From mgilfix at eecs.tufts.edu  Sat Apr 27 13:09:42 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 27 Apr 2002 13:09:42 -0400
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
In-Reply-To: <20020427171024.GA82912@roman.mobil.cz>; from neuhauser@mail.cz on Sat, Apr 27, 2002 at 07:10:24PM +0200
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org> <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com> <20020427171024.GA82912@roman.mobil.cz>
Message-ID: <20020427130942.B1588@eecs.tufts.edu>

  I disagree entirely. When writing production code, you want to model
the thought process that goes into creating the code as clearly if
possible.  If in your head you're thinking "if 'x' is equal to 0",
then you probably want to write it that way. While writing 0==x is
a better move in C, it's entirely against intuition and the natural
thought process.

  Python is smart to avoid such assignments...

               -- Mike

On Sat, Apr 27 @ 19:10, Roman Neuhauser wrote:
>     Anyone who writes "if (x == 0)" instead of "if (0 == x)" deserves
>     what he begs for.

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From gerhard at bigfoot.de  Sat Apr  6 17:44:59 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 6 Apr 2002 22:44:59 GMT
Subject: 2.2 schizo win32 linkage
References: <naBIBIAc1ur8EwVD@jessikat.fsnet.co.uk>
Message-ID: <slrnaauuj6.tl.gerhard@lilith.my-fqdn.de>

Robin Becker wrote in comp.lang.python:
> Once again I am being bitten by the choice of location for DLLs/pyds.
> 
> The standard location for win32 extensions is now lib/site-packages, but
> win32 has problems when DLLs live apart from things they link to.
> 
> As example
> 
> I have tcl83.dll & tk83.dll on the path at c:\tcl\bin and c:\Python\DLLs
> _tkinter.pyd is in the latter and presumably is built to work with
> those.
> 
> _imagingtk.pyd lives with PIL under c:\python\lib\site-packages.
> 
> If I import _tkinter first I get the Python Dlls, if I import _imagingtk
> first I get (by accident of environment) the ones from c:\tcl\bin.
> 
> This is a recipe for disaster and is certainly insecure.

Your path settings perhaps are, but that has nothing to do with
Python, only with Windows's search algorithm for DLLs.

> Surely it's not beyond the wit of the developers to come up with a
> scheme that allows us to put all binary extensions and dlls in one
> place.

That's certainly possible. Only I think it's a very bad idea.

The best solution I can think of is for dependencies of an extension
module to place them in the same directory as the extension module:
this is where Windows searches first when trying to satisfy the DLL
dependency.

Gerhard


From batman at daimi.au.dk  Tue Apr  9 11:32:04 2002
From: batman at daimi.au.dk (Torben Lauritzen)
Date: 09 Apr 2002 17:32:04 +0200
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <3CB2FDF3.BBBD99C0@dynamic.de>
Message-ID: <xm5vgb06g7f.fsf@decius.daimi.au.dk>

Matthias Hanitzsch <Matthias.Hanitzsch at dynamic.de> writes:

> Dennis wrote:
> 
> > x = 2
> >
> > def wrong(num):
> >     return num + 1
> >
> > print wrong(x)
> > print wrong(x)
> >
> > I cannot figure out why this (simplified) program won't increment x
> > and produce 3 and 4.
> 
> Hi,
> this is "call by value", that means that the parameter num of your
> function will get the VALUE of the variable x.
> Then num is incremented and returned.
> The variable x is not changed, as you never write to it. You just
> take its value.
> 
> HTH,
> Matthias
> 

As I see it, num is not incremented, before it is returned. So, even
if it had been a 'call by reference', x would not increase.

-- 
- Torben


From duncan at NOSPAMrcp.co.uk  Wed Apr 17 07:17:41 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 17 Apr 2002 11:17:41 +0000 (UTC)
Subject: pre-PEP for optional 'pass'
References: <mailman.1018972676.30534.python-list@python.org> <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1> <mailman.1019035355.21475.python-list@python.org>
Message-ID: <Xns91F37CE5BFE3Cduncanrcpcouk@127.0.0.1>

Andy Gimblett <gimbo at ftech.net> wrote in
news:mailman.1019035355.21475.python-list at python.org: 

> We're veering slightly from the original reason for this point, which 
> was how to write stub functions/methods which will be implemented 
> later.  The above certainly doesn't indicate any intent to actually 
> implement the methods (in fact, on the contrary, it suggests they 
> could be removed). 

I think you trimmed a bit too much when you followed up. I would have
described what the functions were intended to do if I had known that,
but as I said I have no idea what these particular functions do, so all
I could document was the lack of implementation. Since I was being
chastised for failing to produce meaningful docstrings for someone
else's imaginary functions I made them as meaningful as I could in the
circumstances. 

I suspect the actual methods in the example never will be implemented,
therefore the docstrings are probably accurate.

-- Duncan Booth                                            
duncan at rcp.co.uk int month(char
*p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3" 
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure? 


From ajs at ix.netcom.com  Mon Apr 29 20:54:06 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Mon, 29 Apr 2002 20:54:06 -0400
Subject: Third Draft: Pep for Deprecating Builtins
Message-ID: <000c01c1efe6$dcba54c0$0334fea9@carol>

>I'm not convinced Python 3 is a "new enough" language for this PEP's 
>ideas, although that would seem the obvious point for this sort of 
>tinkering

Perhaps I don't understand the transition path to new style classes.

Is the plan to have the 2 class structures co-exist indefinitely?
That would seem strange and more than a little ugly.

But isn't the concept of deprecating old style classes near 
impossible, as a real world matter?

Doesn't the Python of pure new style classes necessarily
need to be a new generation of the language?

If the issue here is my own confusion, I believe it
reasonable to assume it is confusion shared by enough 
other folks, so that some clarification of the intentions
would make sense.

Art









From terjej at mailandnews.com  Thu Apr 25 13:34:33 2002
From: terjej at mailandnews.com (Terje Johan Abrahamsen)
Date: 25 Apr 2002 10:34:33 -0700
Subject: AS/400
Message-ID: <1e140f34.0204250934.2fee1f26@posting.google.com>

I am trying to write some Python code that moves info from Excel into
AS/400 throuhgt a display program called Rumba. (www.walldata.com)
Basically what I want it to do (short version) is that I go to a cell
in Excel, copy that cell, changes to the Rumba window, goes to
coordinate x,x pastes the number there. Then goes back to Excel, gets
a new number, goes to Rumba, changes screens by typing F12, enters
info from Excel again in coordinate x,x. It is more to it than this,
but if I can figure out this, I can figure out the rest.

The Excel part I have no trouble with. The AS/400 part is a little
worse. Do you know some additions I have to install with Python? Where
can I find some info about the functions that moves around on the
AS/400 display and so forth. I have searched in google and on
python.org but haven't found much. I can't be the only one that needs
to do something like this? Thanks in advance for all answers and all
pointing in the right direction so I can find out more info...


From t.kloppenburg at billiton.de  Tue Apr  2 02:52:30 2002
From: t.kloppenburg at billiton.de (Tjabo Kloppenburg)
Date: Tue, 2 Apr 2002 09:52:30 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A240@natasha.auslabs.avaya.com>
References: <B43D149A9AB2D411971300B0D03D7E8BF0A240@natasha.auslabs.avaya.com>
Message-ID: <E16sJ5q-0003YK-00@mail.python.org>

U program da bad style :-),

> def f (names=None):
> def f( names=None ):

SCNR,
 tk.



From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sun Apr 14 23:25:20 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 20:25:20 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <eppstein-7156A5.13521614042002@news.service.uci.edu> <3cba0068@news.mhogaming.com> <vTou8.20361$8W5.17808@nwrddc02.gnilink.net>
Message-ID: <3cba4868@news.mhogaming.com>

"logistix" <logstx at bellatlantic.net> wrote in message
news:vTou8.20361$8W5.17808 at nwrddc02.gnilink.net...

| < snip! >
|
| So then the next question is, Are non-existant
| values considered False or should they throw an
| error?
|
| < snip! >

That is an EASY one (for even ME) to answer:

An undefined variable is a programming disaster waiting to
happen, and so should ALWAYS throw an exception if you try
to access it before it is defined.  If the variable has been
defined, but does not yet have a value, it should ALWAYS
return 'None' as a value, which has the meaning of 'False'
when used in a conditional statement.





From calves at coelce.com.br  Thu Apr 11 08:33:15 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Thu, 11 Apr 2002 09:33:15 -0300
Subject: RES: Create Excel Files without COM
Message-ID: <29A97D00F387D411AC7900902770E14804C1C16A@lcoeexc01.coelce.net>

Also it works at Excel 97.

> -----Mensagem original-----
> De: Thomas Guettler [mailto:st-newsgroups at thomas-guettler.de]
> Enviada em: quinta-feira, 11 de abril de 2002 04:55
> Para: python-list at python.org
> Assunto: Re: Create Excel Files without COM
> 
> 
> James Strater wrote:
> 
> > Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::
> > 
> > 
> >>Hi!
> >>
> >>Is there something like WriteExcel from perl in python?
> >>I need to generate an MS-Excel file. I don't want to use
> >>COM although it will run on windows.
> >>
> >>The data is simple. Just text in columns and rows.
> >>
> > 
> > 
> > You can format your output as HTML, and then put a .xls 
> extension on it.
> > When it is opened in excel, it will be fine.  To see what 
> your input needs
> > to look like, take your spreadsheet and 'save as' HTML.
> 
> 
> Thank you!
> 
> At least Excel 2002 and 2000 can open the created HTML.
> I pass it to the webbrowser with a content-type like this:
> REQUEST.RESPONSE.setHeader('Content-Type',
>     'application/vnd.ms-excel')
> and Excel opens it.
> 
>   Fine!
> 
> thomas
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020411/d2e9298f/attachment.html>

From mal at lemburg.com  Thu Apr 18 04:03:09 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 18 Apr 2002 10:03:09 +0200
Subject: UTC equivalent of time.mktime() ?
References: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com> <1019062680.653.398.camel@tp6>
Message-ID: <3CBE7DBD.12CCA9F1@lemburg.com>

Graham Ashton wrote:
> 
> On Wed, 2002-04-17 at 17:11, Mark McEahern wrote:
> > Why not just use the mxDateTime library and just do something like this:
> >
> >       mx.DateTime.utc().ticks()
> 
> I appreciate that I could (and thanks for the suggestion), but does this
> really mean that via the standard library there is no easy way to
> convert "1 Jan 1970" to 0 (i.e. string to seconds since epoch)?
> 
> Perhaps I should inspect the mxDateTime code...

Depends on whether your C lib supports leap seconds or not.

See the mxDateTime code for details: 
mxDateTime_AsGMTTicksWithOffset()

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From tl_news at nexgo.de  Mon Apr 15 06:09:22 2002
From: tl_news at nexgo.de (Tino Lange)
Date: Mon, 15 Apr 2002 12:09:22 +0200
Subject: problem running win32com.client application from commandline
References: <3hrjbu4qv9b9afjhr2nhnef025350k4lpd@4ax.com> <3CBA8463.30703@skippinet.com.au>
Message-ID: <a9e8si$47q$1@crusher.de.colt.net>

Mark Hammond wrote:

>> I just discovered a problem with a win32com.client based application.
>> [...]
>> Any idea? What's different, when I run the script via PythonWin?

> Pythonwin contains the code necessary to host true OCX controls.  You
> can't do this in python.exe or wpython.exe.  You may still be able to
> get what you need by using "pythonwin.exe /nodde /run foo.py"

Hi Marc,

thanks a lot for your reply!
Great solution - so I just run my script with a (hidden) PythonWin instead 
of python.exe and everything is fine.
Is there a list of those useful commandline-switches for PythonWin 
available? A "PythonWin.exe /?" or "PythonWin.exe --help" does not help.

Just one more question - it seems that this OCX I want to script via python 
has a form attached, that is filled under some circumstances - the return 
value of the called function via COM is then not enough for me. 

>From looking at the Word-Macro that uses the OCX originally I can see that 
there are some global variables shared between Word VBA and the OCX. Also 
there's this form that s the same in Word and in the OCX. Those global 
variables and the form are not accessible via the COM-Interface, but the 
Word Macro seems to have access. (at least the makepy generated file does 
not list them).

Is there a way to get the form-handle and have access to the form and for 
example some textual content?

Thanks a lot.

Tino




From peter at engcorp.com  Wed Apr 10 21:52:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 21:52:26 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer 
 division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org>
Message-ID: <3CB4EC5A.691F1F7B@engcorp.com>

Tim Peters wrote:
[in a dull moment]
> You users slay me <wink>.  For me, "my own entertainment" means I can use
> such "developer features" to get my work implementing features for you to
> use done quicker, have it work better, suffer fewer bugs, and give you more
> capabilities in your code than you had before.  If you're ever tempted to
> think that work devoted to improving language and library implementors'
> productivity is wasted on users, you should try not watering your garden
> during a drought too <wink>.

I'd like to take this opportunity to at least thank the developers,
bots included, and other contributors (in the most heartfelt way I
can without sending large cash contributions) for all the blood,
sweat and keypresses they've put into this phenomenal project.

I'm classified in the pathetic subset of users known as "commercial
developers building products using Python and to date not contributing
a damn thing back to the community".  My accomplishments in the
past two years using Python have been possible largely because of
the unpaid, thankless (well, until now by me anyway) efforts of 
a large number of mostly volunteer developers.  

We're leeches (so far) but at least we're very grateful leeches.

I'm dismayed by comments attacking the core developers when made
by people who have not contributed more than I or my team have.
Beggars can't be choosers, and nobody likes a whiner...

-Peter


From aahz at pythoncraft.com  Sat Apr 13 13:34:39 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Apr 2002 13:34:39 -0400
Subject: binding a reference to a variable
References: <mailman.1018682547.1577.python-list@python.org>
Message-ID: <a99q7f$ss6$1@panix1.panix.com>

In article <mailman.1018682547.1577.python-list at python.org>,
Tim Peters  <tim.one at comcast.net> wrote:
>
>This is a dodge the Pythonic mind may be too quick to take, though, as about
>once a year it suddenly (re)strikes me how silly the Python idiom
>
>    while 1:
>        line = f.readline()
>        if not line:
>            break
>
>really is <0.9 wink>.  The other 364 days I don't notice the hair.

But it *is* silly.  That's why we have iterators.  One thing I noticed
recently about how fragile the while 1 idiom is: it depends on
readline() returning linebreaks.  Writing a file-like class for which
this *isn't* true breaks the idiom.

I used to say instead, "Yeah, that's a crufty idiom, but it's worth the
pain for all the other features Python gives you, and I *don't* want to
allow 'while line = f.readline()'."
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From noah at noah.org  Tue Apr 16 12:47:22 2002
From: noah at noah.org (Noah)
Date: Tue, 16 Apr 2002 09:47:22 -0700
Subject: Backing up mail in Outlook Express using Python
In-Reply-To: <cGSu8.101$1kc.179320832@news.telia.no>
Message-ID: <PAENINBOOODFDEPOKPKCKEHJCLAA.noah@noah.org>

Hi,

How do you want to back up your Outlook email? 
If you want to keep it in Outlook format 
then you can just copy the BDX files.

Do you mean to convert your Outlook Express mail folder
into a plain text file? Say, into a standard UNIX mbox format?
The mbox format has the advantage that all email is stored
in one big, human readable text file and most (nay, all) UNIX
mail clients can read directly from this format.

The big trick is getting the email messages out of OE.
The OE DBX files are binary. I was unable to figure 
out how to do this from Python. What I did do was to 
find a Windows utility that would get me half-way there.

First, you need the DBXtract utility. You can find it here:
	http://chattanooga.net/~scochran/DBXtract.htm
This explodes your Outlook DBX file into a bunch of little files.
That may be all you need, but if you want to archive them
into a format that will let you read them on UNIX then you
need to convert the files into an mbox.

Second, you need my oesort script. This cleans up the
work done by DBExtract. I have attached a copy of the
script to this email. Hopefully the comments make it
somewhat clear as to how to use the script. This was
a quick ad hoc script, so it has not been carefully tested.
Use it at your own risk! Every once in a while I find that
an email gets left out of the final mbox file. I'm not sure
if that is the fault of my script or of DBXtract. It seems
to be 1 out of 100.

This person also has documentation on the OE format:
	http://oe5dbx.aroh.de/index.html

In the future do not store your mail in Outlook or any
other closed system. Make sure that your POP3 mail client
is configured to LEAVE MAIL ON SERVER (unfortunately, most
email clients default the wrong way). Or better yet, use IMAP.
I learned my lesson the hard way, which is why I wrote this script :-)

Yours,
Noah
-------------- next part --------------
A non-text attachment was scrubbed...
Name: oesort.py
Type: application/octet-stream
Size: 11255 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020416/cc98e793/attachment.obj>

From sholden at holdenweb.com  Wed Apr 17 09:23:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 09:23:31 -0400
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org> <a9icab$phq$1@nntp9.atl.mindspring.net> <Xns91F32355A8FA9RASXnewsDFE1@130.133.1.4>
Message-ID: <lJev8.12963$hl2.6959@atlpnn01.usenetserver.com>

"Philip Swartzleonard" <starx at pacbell.net> wrote in message
news:Xns91F32355A8FA9RASXnewsDFE1 at 130.133.1.4...
> Andrew Dalke || Tue 16 Apr 2002 04:35:54p:
>
[ Andrew Dalke defines a Strange() object ]
>
> In this case, wouldn't 3+7.0+x be a Typeerror or something? Int doesn't
> know how to add with a Strange, and Strange dosen't have an upgrading
> function... or do we assume addition is always commutitive and try '3+x+
> 7.0' (effectively), therefore breaking the optimization mentioned by the
> other... =)
>
In Python it takes very little time to check whether your assertion is true.
Readers are encouraged to perform the following experiment ;-) ...

>>> class Strange:
...  def __add__(self, other):
...   return self
...
>>> x = Strange()
>>> x+3+7.0
<__main__.Strange instance at 0x014CE610>

The point of Andrew's example is precisely that Strange() instances can have
values of *any* type added to them, since their __add__() method does not
ever refer to the second operand. So there's absolutely no reason why any
TypeError [or something] exception should be raised.

When you refer to an "upgrading function" I suspect you mean a coercion?
There's no need for a coercion, since for any Strange instance s, s+x == s
for any x (and, indeed, s+x is x as well)\

regards
 Steve






From jochen at jochen-kuepper.de  Sun Apr 28 20:46:25 2002
From: jochen at jochen-kuepper.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 28 Apr 2002 20:46:25 -0400
Subject: wordlist for spell-checker (was: <way too long>)
References: <aai16e$m3$1@newsg1.svr.pol.co.uk>
Message-ID: <m33cxfgwm6.fsf@box.home.de>

On Mon, 29 Apr 2002 00:41:33 +0100 G Willoughby wrote:

G> Is there a text file somewhere i could download that contains all the words
G> in the english language that i could read into python to create a spell
G> checker?

What's wrong with a google search? Then you could find, for example,
Kevin's word-list page:
,----
| http://wordlist.sourceforge.net/
`----

Greetings,
Jochen

PS Personally I think your subject line is a little /too/ long.
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From aleax at aleax.it  Sun Apr 28 03:29:48 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 07:29:48 GMT
Subject: local variable not found
References: <pan.2002.04.27.23.01.32.51309.30401@onesite.org> <HzEy8.28287$8D3.754996@news1.tin.it> <pan.2002.04.28.01.57.54.335756.30401@onesite.org>
Message-ID: <MxNy8.71694$vF6.2141942@news2.tin.it>

Alain Tesio wrote:
        ...
>> (in my experience) only when used with explicit
>> dictionaries, NOT by letting them default to locals()
>> and globals().
> 
> My problem is to make html templates where the user
> can insert python code.

I think there must be a zillion such beasts, and none
needs to use implicit dictionaries for exec / eval.  There
are a couple on the Cookbook, for example; see e.g.
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52305
which is of course my favourite since "my own baby".


Alex



From zeitlin at seth.lpthe.jussieu.fr  Mon Apr  1 10:02:11 2002
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Mon, 1 Apr 2002 15:02:11 +0000 (UTC)
Subject: wxPython  / wxSingleChoiceDialog ...
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de>
Message-ID: <slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr>

On 1 Apr 2002 13:54:42 GMT, Uwe Schmitt <uwe at rocksport.de> wrote:
> I tried to manipulate wxSinglChoiceDialog so that the presented 
> choices are printed with a fixed width font. I tried SetFont()
> but nothing changed...

 You should change the font of the listbox used by this dialog, not the dialog
itself. Although there is no way to access this listbox directly (and unlikely
to ever be), you can iterate over the dialog children until you find it.

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From mattdm at mattdm.org  Mon Apr  8 22:59:13 2002
From: mattdm at mattdm.org (Matthew Miller)
Date: 9 Apr 2002 02:59:13 GMT
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com>
Message-ID: <slrnab4m82.glr.mattdm@jadzia.bu.edu>

Anthony_Barker <anthony_barker at hotmail.com> wrote:
>Python 2.2x is stable why does redhat still insist on releasing 1.5x?
>All other distributions are now on 2.x.
>It seems from the Redhat's skipjack beta 2 that it will only ship the
>older version of python(1.5.2).

This is a good indication that the next release will indeed be 7.3, not 8.0.
(Note that the beta version is 7.2.9x, so the next version isn't
*necessarily* 7.3). Red Hat uses major version number changes to indicate
that they've broken compatiblity in some way (it's not just a marketing
thing like it is for some other complanies). Going to python 2 would break
stuff, so it has to wait. (On the other hand, there *is* the "python2"
package.)

-- 
Matthew Miller           mattdm at mattdm.org        <http://www.mattdm.org/>
Boston University Linux      ------>                <http://linux.bu.edu/>


From marklists at mceahern.com  Wed Apr  3 08:58:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 07:58:22 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <E16selR-0005Cn-00@tswings.swing.cluster>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>

[Christian Tanzer]
>   For me, Python's (old) way of handling boolean conditions is one of
>   its great strengths. This wasn't always the case. When I started
>   using Python, I was slightly put off by the lack of a boolean data
>   type. But quite quickly I learned to appreciate Python's way of
>   dealing with this issue.

How does the PEP change "Python's (old) way of handling boolean conditions"?

Correct me if I'm wrong, but under the PEP, you can still do this:

	l = range(10)
	if l:
		print "My, the list is not empty!"

So what's different?

// mark




From guido at python.org  Tue Apr  2 20:10:39 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 01:10:39 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it>
Message-ID: <3CAA5777.66E7C2D5@python.org>

Alex Martelli wrote:
> I _have_ seen old-time Fortranites rant quite convincingly against that
> useless, newfangled LOGICAL stuff.

And that is exactly what many of the responses in this thread
(including Laura's over-long FUD piece) sound like.

I'm reading (at least skimming) everything, but the more I read the
reactions, the more I believe that this is just the response I have to
expect for *every* change to the language I propose, and the more I'm
tempted to follow my gut instincts.  Long ago (right after lambda :-)
I learned to say "no" to proposals that didn't feel right to me, even
if I couldn't express very clearly why.  I think it's time I learn to
say "yes" even if the community doesn't see the light yet.

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


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 18:44:35 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 15:44:35 -0700
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <3cb8b52a@news.mhogaming.com>

"Jens Baader" <nospam at nospam.com> wrote in message
news:3cb8a755$0$12714$9b622d9e at news.freenet.de...

| Are there any plans to get Python "stable" or will
| the language continue to change with each major release
| of the interpreter?

A language that does not evolve is like "Latin" is now:  A
_DEAD_ language that is only used by academic type people.

There is nothing stopping you from getting (and staying
with) an older version of the language-- you do not always
have to be "on the bleeding edge" of Python evolution.  Many
people have chosen to stay with version 1.5.2-- it is
stable, there is a lot of documentation for it, and it has
been ported to the most platforms.  In fact, there is _NO_
reason to change to a newer version unless you need one or
more of the new features in the newer version to get work
done.

| I'm downloading Python 2.2.1 right now and I wonder
| why you still produce bugfix releases for the old 2.1
development line.

I think 2.1.3 was an "emergency" release of the 2.1 branch
to fix some bugs that caused Zope (a VERY important Python
application) to crash sometimes...

| It seems that Python 2.2 is somewhat broken. If not what's
the
| reason that keeps the people from upgrading to 2.2?

There were a LOT of changes in 2.2...  I think everyone was
waiting for the bug-fix release (2.2.1-- which is out now.)

| Another question: When will we see an official ISO/ANSI
standard
| for Python? I dislike working with nonstandardized
languages.

I think this would absolutely _RUIN_ Python (for me at
least.)  See my comment above about _DEAD_ languages...

| The fact that radical language changes are still made (the
| type/class unification) seem to suggest that Python is
still not
| mature enough for a real standard. Or maybe Python and
| the other "newschool" languages (like Ruby, Java) are not
| meant to by standardized? Don't know but I really would
| like to.

You don't _HAVE_TO_ use Python-- you can use any one of the
over 4000 "free" languages "out there"-- or you can write
your _OWN_ language.  If you write a language as good as
Python, and as stable as Python, without Python's warts, I
will be one of the first to download it and start using
it...  Until then, I will continue to use Python until
something better comes along--  (Thank you Guido for
creating a great language!)

Any more questions?

Very_Happy_With_Python_'ly yours,
--Ken Peek
http://www.SpiritSongDesigns.com

(Welcome to the "New World Order"!  --  "I just CAN'T wait
to get a "brain chip" in my head!"... "Oh, PLEASE!... At
least let me have one of those "Model 666" identification
microchips in my hand!!!" see: http://www.adsx.com/ )





From peter at engcorp.com  Mon Apr 15 21:37:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 15 Apr 2002 21:37:14 -0400
Subject: used ports - socket bind
References: <cf0ad9fb.0204150040.2f70bc4@posting.google.com>
Message-ID: <3CBB804A.6F2A20D0@engcorp.com>

Jesper Olsen wrote:
> 
> When binding a socket to a port, is there a way to test if the port
> is already used?
> 
> (Apart from catching the resulting exception if it is...)

Even if there was a way, it wouldn't be correct to use it.
This is no different from testing whether a file exists
before trying to open it, or many other examples.

Between the time you discover that the port is not in use
and the time you attempt to bind to it, something else might
already have bound to that port, so you'd still get the 
exception and still have to catch and handle it.

So save yourself the trouble and just do the try/except
approach.  (And I'm pretty sure the answer is "no".)

-Peter


From pyth at devel.trillke.net  Sat Apr 20 11:05:08 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 20 Apr 2002 17:05:08 +0200
Subject: Substitution with Hash Values
In-Reply-To: <slrnac2v9d.kv6.jon+usenet@snowy.squish.net>
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org> <slrnac2v9d.kv6.jon+usenet@snowy.squish.net>
Message-ID: <20020420150508.GB10574@devel.trillke>

On Sat, Apr 20, 2002 at 02:37:34PM -0000, Jon Ribbens wrote:
> In article <mailman.1019303438.1339.python-list at python.org>, holger krekel wrote:
> >> Need a raw string there, i.e.
> >> 
> >>   r"\{\{(.*?)\}\}"
> > 
> > yes, i am sometimes confused by 'raw strings' and 'strings'. 
> > Could you help by saying why exactly it is needed here?
> 
> Hmm, well, actually you don't, since \{ is not a string literal escape
> sequence and Python is weird and leaves in unknown escape sequences
> unchanged.

Actually my intention was to prevent the re module from thinking
i might mean 'repetition' with {...}. I thought that every special
regexp-character like '.' or '*' needs to be escaped if you want
it literally. I was surprised to see that you can omit the '\'
in the above case
 
> Basically, you always use raw strings when using regexps, and never
> use them anywhere else. You need them because both string literals and
> regexps treat \ as an escape character, so to get the \ through to the
> regexp unchanged you use a raw string.

thanks for the explanation! 

But is it true that you 'never' use them somewhere else? 
I would think that one would like to use 'r' whenever you don't 
want the escape-sequence-string-interpretation which is not really
restricted to regexp-strings. 

I have noticed that e.g. the docstring of the module os.py uses
r'...'

regards,

	holger




From gerhard at bigfoot.de  Wed Apr 17 18:36:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 17 Apr 2002 22:36:17 GMT
Subject: mod_python
References: <e1e40a36.0204171358.7af66fdf@posting.google.com>
Message-ID: <slrnabru6t.i6.gerhard@lilith.my-fqdn.de>

daniela eilers wrote in comp.lang.python:
> hi, experts
> currently, i'm a python beginner.
> normal python script run with apache, no probs.
> trying to run mod_python under apache i get the following error message:
> 
> premature end of script headers. 
> 
> there is the code of my cgi script:
> 
> from mod_python import apache 
>          def handler(req): 
>               req.content_type = "text/html" 
>               req.send_http_header() 
>               req.write("Hello World") 
>               return apache.OKimport apache 
                       ^^^^^^^^^^^^^^^^^^^^^^

Looks like here's the error. This should be simply apache.OK. Apart
from that, your example looks exactly like the one on  from

http://www.modpython.org/live/mod_python-2.7/doc/tut-what-it-do.html

and should be ok.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From cstoy at nc.rr.com  Tue Apr  9 00:18:46 2002
From: cstoy at nc.rr.com (Chris Stoy)
Date: Tue, 09 Apr 2002 04:18:46 GMT
Subject: Embedding Python with C++
Message-ID: <GYts8.26982$2J2.6830478@typhoon.southeast.rr.com>

Hey all,

I'm looking to embed Python as a scripting language in a game I am working
on.  I want to be able to call C++ methods from Python, and call Python from
C++ (for the usual code quick in Python, convert to C++ for speed later.)

I've looked into using the Boost.Python code and I'm fine with that.

My problem is, I have no clue how to compile Python into a static linked
library for Windows.  Actually, I need it to be cross-platform, which is
another issue.

I've never embedded Python before, and I'm fairly new to Python, so can
someone please give me a pointer to how to do this?  What are the files I
need to compile to build the static lib (using MS Visual Studio 6.0?) Also,
I don't need all of the standard Python components (like networking.)  Can
those be removed from the lib?

Thanks.

Chris.





From hove at phys.ntnu.no  Fri Apr 26 06:16:52 2002
From: hove at phys.ntnu.no (Joakim Hove)
Date: 26 Apr 2002 12:16:52 +0200
Subject: Converting a Python List to a C double *
Message-ID: <k0n4rhy3gtn.fsf@metropolis.phys.ntnu.no>

Hello,

I am writing a C-extension, and now I need to parse a list from python
into a c double *:

   static PyObject * power(PyObject *self, PyObject *args) {
     double *xdata;
   
     /*
       This routine shall be called from Python. It will always be
       called with a variable length list of floats: e.g. 
       module.power([0.1, 0.2, 0.45, 0.60, 0.70]).
     
       Within the c-function "power" I want to access the list in the
       pointer xdata. How to parse?
     */
     
     


     for (i=0; i<5; i++) 
       printf("x[%d] = %12.7f \n",i,xdata[i]);
     return PyNone();
   }  

Grateful for any suggestions - 

Joakim Hove


-- 
==== Joakim Hove      www.phys.ntnu.no/~hove/   =======================
|| Institutt for fysikk  (735) 93637 / E3-166  |  Sk?yensgate 10D    ||
|| N - 7491 Trondheim    hove at phys.ntnu.no     |  N - 7030 Trondheim ||
================================================= 73 93 31 68 =========


From max at alcyone.com  Fri Apr 26 00:24:30 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 25 Apr 2002 21:24:30 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com> <mailman.1019792457.31426.python-list@python.org>
Message-ID: <3CC8D67E.30DF7A32@alcyone.com>

"James J. Besemer" wrote:

> 8. I really gotta say -- some of you guys really need to lighten up
> and/or get a
> life.  There is, was and always will be a world aside from Python.

Yeah, you're absolutely right.  Could we have twenty or more points from
you outlining exactly how much they need to lighten up and get a life? 
Hint:  IRONY.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From gh_pythonlist at gmx.de  Mon Apr  1 23:54:15 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 2 Apr 2002 06:54:15 +0200
Subject: (Help) Embedding in Python 1.5
In-Reply-To: <slrnaahu3v.21k.id@c1299689-a.saltlk1.ut.home.com>
References: <slrnaahu3v.21k.id@c1299689-a.saltlk1.ut.home.com>
Message-ID: <20020402045415.GB1247@lilith.hqd-internal>

* HH <henrykhecht at my-deja.com> [2002-04-02 00:16 +0000]:
> I'm trying to embed some Python in some C. [...] Could someone supply
> me with a modern (but not too modern, i.e. 2.x) example makefile, or
> at least some advice?

Does my previous answer to a similar question help? URL is (long):

http://groups.google.com/groups?q=group:comp.lang.python+gerhard+distutils&hl=en&selm=mailman.1016129667.28593.python-list%40python.org&rnum=9

It requires distutils, which you'll need to download seperately for
Python 1.5.2: http://www.python.org/sigs/distutils-sig/download.html

I'd be very interested if this actually works for other people.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.3 ?C      Wind: 0.5 m/s



From geoff at gerrietts.net  Tue Apr 16 16:38:48 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 16 Apr 2002 13:38:48 -0700
Subject: problem negative lookahead assertion
In-Reply-To: <4b926dbb.0204160420.4e9e858d@posting.google.com>
References: <4b926dbb.0204151202.6d709633@posting.google.com> <mailman.1018902689.4821.python-list@python.org> <4b926dbb.0204160420.4e9e858d@posting.google.com>
Message-ID: <20020416203848.GB20804@isis.gerrietts.net>

Quoting sjoerd siebinga (ssiebinga at fa.knaw.nl):
> Thanks Geoff and Amk for your advise.
> 
> Sadly enough my input data requires nested {} due to specialist
> diacritics use.So the ([^}]*?) option gives more errors when
> processing the data through the latex interpreter. I used the
> ([\s,\W]) regex to grab all the curly braces in  the emph.
> 
> Now it is time to ask a silly question. I am a trained linguist and
> not a programmer.
> 
> > Note that this will break in a different situation, if you have {} inside  
> > the contents of \emph, as for example in \emph{ab^{cd}}, because this 
> > problem really needs a full parser to handle the general case.
> 
> What do you mean by a 'full parser'?

I think the way you translate that is "regular expressions aren't
going to be sufficient for what you're doing". You'd need a parser
that would look at each distinct token and/or character, and decide
what its disposition should be.

The problem is the nested {}'s. If you can't count on the first } you
encounter ending the \emph section, then you can't use [^}] to delimit
your \1.

I'm going to take some time and analyze your input document, but I
think the question you need to be asking right now is: how to
reasonably bound your \1 expression?

Let me give you an example, from your examples:


> `to wade through, ford', \emph{vadum} \index{unl~vadum}`ford'. It cannot

I believe this should be generated from a source line:

  `to wade through, ford', \emph{vadum}`ford'. It cannot

Now, what we want to retrieve is the "vadum" part.

For this particular case, it's easy:

  r"\\emph\{(\w+)\}"

This matches a string beginning with \emph{ followed by a group of
alphanumeric characters (and only alphanumeric characters), followed
by a closing brace }. The contents between the brace are extracted as
group 1.

However, this approach doesn't work if the "vadum" part -- the
emphasized term -- needs to be able to contain characters other than
alphanumerics. For example

> \emph{(ge)wolt}, \emph{(ge)walt}, \emph{ge]welt}

None of those would match, because (, ), and ] are not alphanumeric
characters.

So the next step is to try to figure out which characters are
permitted between the two enclosing { } brackets.

This line:

> OS \emph{gi]weldig}, OHG \emph{gi]welt\={\i}g}, \emph{gi]walt\={\i}g},

Is pretty revealing. From this, it looks like we can expect to see
just about anything. :) Further, it doesn't look like there's really
any way to say "a } followed by X" for any value of X where it will
(1) always match at the end of an \emph expression and (2) never
accidentally match in the middle of an \emph expression.

What you really need is a way to do brace-matching within your
expression, so that if \emph is followed by two {'s, you'll wait for
two }'s before you decide the \emph is over.

That is, you need to be able to look at:

> OHG \emph{gi]welt\={\i}g},

and you need to be able to extract gi]welt\={\i}g as the content of
the \emph expression, and to do that, you hafta know that the } after
the \i matches the { right before the \i, not the { after \emph.

I don't think you can't get this with a single regular expression, and
I don't believe that regular expressions in general are not
well-suited to this task. My certainty here is about 90%, and I only
express doubt because I may be ignorant of some techniques.


Below is a sort of pseudocode outline for how you would go about
solving this problem. It might not be the fastest or most efficient
way, but it's pretty readable and straightforward.

The complexity here is due to the fact that you can't trust your
pattern to identify only the conditions we are interested in. Instead,
we're thinking it's likely that the pattern will return lots of false
positives.

This is, by the way, some variation on the theme "a full parser".

- copy the input to a temporary string
- create an empty results buffer.

- while you still have characters in your temporary string:
  - if you match your current emph expression:
    - copy the characters up to the beginning of the match from the
      temp string to the results buffer
    - remove the characters up to the beginning of the match

    - extract group 0 (the whole match)
    - analyze it character-by-character to extract the contents of the
      \emph. this can be done by keeping a simple count of {'s and a
      count of }'s: when the two numbers are equal, you're at the end
      of the expression
    - if the next expression is an \index expression (check this
      here, this is necessary because your lookahead is not reliable):
      - copy all characters up to the start of \index into your "results" 
        buffer, with no transform. this was a false positive.
      - remove those characters from your temp string, and redo the loop.
    - if the next expression is not an \index expression:
      - put the \emph and the created \index into the results buffer.
        this is a true hit.
      - remove the characters up to the end of the \emph from the temp
        buffer.
  - if you don't match your expression, copy the remainder of your
    temporary string into the results buffer, and empty out your
    temporary string.

> As Geoff asked I will send  a couple of samples. My Python script is
> rather too large to be send to the newsgroup. If either of you want to
> see it, I can send it to you with a couple of sample lemmata. Another
> point is that the substitutions are preformed in turn much like an sed
> script. The emph regex is processed last.

I don't think we need more examples at this point, but one point that
would be helpful when picking your examples: pick one piece of input
text, preferably the shortest piece you can isolate that demonstrates
all the hard problems. Then include what happens when you process it,
and what you wish would happen when you processed it.

The examples you gave were pretty big (not a huge problem), and all
different, making it very difficult to compare what you have with what
you want.


-- 
Geoff Gerrietts             "There is no fate that cannot be 
<geoff at gerrietts net>     surmounted by scorn." --Albert Camus




From tegbert at abl.med.utah.edu  Tue Apr  2 12:50:24 2002
From: tegbert at abl.med.utah.edu (Tim Egbert)
Date: Tue, 2 Apr 2002 10:50:24 -0700
Subject: Problem Calling Python callbacks from C
In-Reply-To: <m3d6xigg6a.fsf@mira.informatik.hu-berlin.de>
References: <8351bb33.0204020854.272eaa7b@posting.google.com> <m3d6xigg6a.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Pine.LNX.4.10.10204021028590.32630-100000@abl.med.utah.edu>

I'm able to call a Python callback _function_ from C code by following the      
information in "Extending and Embedding the Python Interpreter," section
1.6, but I can't get the callback to work if I'm calling a Python object
_method_.  Does anyone know how to do that?

In other words, it works if I'm using a Python function such as:
   
    def callback_func(i):
        print i

and I set "callback_func" as the callback but not if the callback is an
object method such as:

    class callbackClass:
        def callback_method(self,i):
            print i
    cbc=callbackClass()

and I set the callback as "cbc.callback_method".

I'm using the "PyEval_CallObject(callback,arglist)" function in my C code
to make the callback.  I assume the problem has to do with not having the
"self" argument available in the arglist for the callback method, but I  
haven't been able to find any documentation on how to provide "self" or
what Python function to use for calling an object method.

Tim 



From vinfoley at iquebec.com  Mon Apr 29 17:42:35 2002
From: vinfoley at iquebec.com (Vincent Foley)
Date: Mon, 29 Apr 2002 21:42:35 GMT
Subject: Equivalent in Python?
Message-ID: <f7jz8.18268$Wd2.881746@wagner.videotron.net>

What's the Python equivalient of Perl's "a"++; ?


-- 

Vincent Foley-Bourgon
Email: vinfoley at iquebec.com
Homepage: http://darkhost.mine.nu:81


From quinn at bolivar.ugcs.caltech.edu  Tue Apr 23 23:04:28 2002
From: quinn at bolivar.ugcs.caltech.edu (Quinn Dunkan)
Date: 24 Apr 2002 03:04:28 GMT
Subject: calendar.py doesn't use lists???
References: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <slrnacc85r.id2.quinn@bolivar.ugcs.caltech.edu>

On 22 Apr 2002 15:05:16 -0700, Aaron Watters <aaron at reportlab.com> wrote:
>[CRASH AND BURN... WINDOWS XP SUGGESTS I SEND A COMPLAINT TO MICROSOFT...]

Wow, they suggest that when you get a segfault now?  Much different from
their usual finger-wagging "you should learn how to turn your computer off
properly" stuff.

>Seriously, why did you need to change these sequences from
>lists where they worked with my code, to somethingsimilartolists
>that broke my code?

>From a quick glance it looks like _localized_name is just a too-clever
but harmless way to make lists.  The real problem is that time.strftime
explodes when it gets handed certain values, and the regression tests missed
it.

>If there is a god, he or she sure liked to design
>beetles.
>  -- attribution forgotten

It was (supposedly) Haldane, a naturalist from way back when.  The context
was that he was asked what his view was, as a naturalist, of God's "character",
and the response was "an inordinate fondness for beetles".


From me at mikerobin.com  Wed Apr  3 12:34:55 2002
From: me at mikerobin.com (Michael Robin)
Date: Wed, 03 Apr 2002 17:34:55 GMT
Subject: readline() blocks after select() says there's data??
References: <2b57f654.0203141934.391d1bcc@posting.google.com> <mailman.1016168907.4720.python-list@python.org> <USnk8.25666$N7.5384674@ruti.visi.com> <uk7sdzqem.fsf@ctwd0143.fitlinxx.com> <_kqk8.25881$N7.5410299@ruti.visi.com> <ubsdpzd97.fsf@ctwd0143.fitlinxx.com> <83elihknoz.fsf@panacea.canonical.org>
Message-ID: <33Hq8.186075$uA5.150775@rwcrnsc51.ops.asp.att.net>

"Kragen Sitaker" <kragen at pobox.com> wrote in message
news:83elihknoz.fsf at panacea.canonical.org...
> But all of those approaches are really ways to try to work around the
> fact that WaitForMultipleObjects (sp?) only works on signaling of
> event objects...

Not sure if I'm reading you right, but (Msg)WaitForMultipleObjects() will
work on a plethora of
system objects.
>From the docs at
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dllproc/syn
chro_5gqb.asp and elsewhere:
------------------------------------
The MsgWaitForMultipleObjects function can specify handles of any of the
following object types in the pHandles array:
  a.. Change notification
  b.. Console input
  c.. Event
  d.. Job
  e.. Mutex
  f.. Process
  g.. Semaphore
  h.. Thread
  i.. Waitable timer
  j.. [mike: (1) sockets and (2) files, named pipes, serial comms - see
below]
(1) ...the native Windows function, WaitForMultipleObjects can also be used
for this purpose. This is especially useful if the thread needs to wait on
both socket and nonsocket events.
...A socket handle can optionally be a file handle in Windows Sockets 2. It
is possible to use socket handles with ReadFile, WriteFile, ReadFileEx,
WriteFileEx, DuplicateHandle...

(2) ...you can also use a file, named pipe, or communications device as a
synchronization object;
[...it is safer to use the event object because of the confusion that can
occur when multiple simultaneous overlapped operations are performed on the
same file, named pipe, or communications device. In this situation, there is
no way to know which operation caused the object's state to be signaled...]
------------

mike







From a.clarke11 at pop.ntlworld.com  Mon Apr  1 13:49:40 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Mon, 01 Apr 2002 19:49:40 +0100
Subject: Tkinter + import question
References: <3ca754d4$1_1@nopics.sjc>
Message-ID: <3CA8ABC3.3612B8B4@pop.ntlworld.com>

Hi

Try importing it first into your main program, then imported modules
should be able to activate it.
Adonis a dit:

i am writting a program utilizing Tk as my gui engine, and i have broken

down sections of the form into modules and loaded them through a main
script. my only question is in every module i am forced to import the
Tkinter module, and does this pose consume too much memory? or does
python
delegates itself? or is there a way to just load Tkinter into one module
and
have it received by all? i assumed this was done when i declared Tkinter
in
one module then all other modules would 'inherit' the Tkinter module,
but
that wasnt the case.

any sugguestions is greatly appreciated.

Adonis


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020401/f8487683/attachment.html>

From jeff at ccvcorp.com  Tue Apr  9 15:44:30 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 9 Apr 2002 12:44:30 -0700
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com>
Message-ID: <MPG.171be50c35e59a459896a1@news.nwlink.com>

In article <899f842.0204081714.25eab798 at posting.google.com>, 
anthony_barker at hotmail.com says...
> Python 2.2x is stable why does redhat still insist on releasing 1.5x?
> All other distributions are now on 2.x.
> 
> It seems from the Redhat's skipjack beta 2 that it will only ship the
> older version of python(1.5.2).

Because RedHat has a stated policy of shipping the same version 
of all major utilities with *all* minor releases within a 
specific major release of RedHat.  In other words, all of their 
major utilities will use the same version through each of RH 7.0, 
7.1, 7.2, 7.x ...

Python qualifies as a major utility in this sense because RH uses 
it in a significant number of its system configuration and 
administration tools.
 
> I find this annoying because it means I have to upgrade each redhat
> machine individually as it breaks a couple of my scripts.

Be aware that, since RH uses Python 1.5.2 for its tools, and in 
several cases uses deprecated modules that no longer function in 
Python 2.x, you'll need to be careful about upgrading -- you need 
to leave /usr/bin/python pointing to 1.5.2, and install 2.x as 
/usr/bin/python2 (RH does provide rpms that are configured to do 
this; those using RH should definitely get RH's rpms.)  It leaves 
you in an awkward position of having to watch that scripts get 
executed with the correct version of Python (since you can't 
presume that 'python myscript.py' will execute the latest 
version), but the only alternative is to change all of RH's 
scripts to explicitly point to /usr/bin/python1.5 (which could be 
quite a task).

[x-posts trimmed]

-- 

Jeff Shannon
Technician/Programmer
Credit International


From aahz at pythoncraft.com  Wed Apr 24 12:58:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Apr 2002 12:58:29 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <aa6o7l$nbi$1@panix3.panix.com>

In article <ad381f5b.0204240841.52333a3e at posting.google.com>,
Rad <zaka07 at hotmail.com> wrote:
>
>I still haven't received above mentioned files so I can't test the
>time needed to (for example) read a 15GB "file1", filter by few
>variables, and write a resulting subset as a "sub_file1".  Things
>would afterwards get more complicated cause I will have to pullout
>ID's from "sub_file1", remove duplicate ID's create
>"no_dup_sub_file1", match those to ID's in remaining 3 main files and
>pullout data linked with those ID's.
>
>I have a few weeks to prepare myself before data arrives and my
>question is: am I going the right way about the project, is Python
>(with humanly written code) capable of doing this kind of staff
>relatively quickly?

Python *can* handle this kind of task, but you'll be much better off if
you interface Python to a database.  The problem is that even 15GB (not
even talking about 80GB) is simply too big to fit in RAM, so you'll need
a way to process partial sets that do fit in RAM.  That's precisely what
a database is designed to do.  Python can help you transform the data
into formats that will fit better into the database, and Python can
drive the operation of the database, but you're going to need to learn
SQL to manage the actual database operation.

Python is certainly better than C for this; depending on the complexity
of the data and operations, you may be able to do this task entirely
within the database, skipping Python completely.

I strongly suggest that you push *VERY* *HARD* to get some small sample
files (100MB to 1GB range).  Get those samples ASAP.

Finally, make sure that you have at least four times the disk space as
the total size of all the files (that's probably a conservative guess,
but you'll definitely need at least 2.5 times).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From baas at ira.uka.de  Sun Apr 21 11:58:08 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Sun, 21 Apr 2002 17:58:08 +0200
Subject: ANN: Python Computer Graphics Kit v1.0beta2
Message-ID: <6ao5cu0cmv6pn4to0gpjfpmjjjlop4ar07@4ax.com>

The second beta release of the Python Computer Graphics Kit is
available at http://cgkit.sourceforge.net

What is it?
-----------

The kit is a collection of Python modules that contain the basic types
and functions required for creating 3D computer graphics images. The
kit includes several new types such as vectors, matrices and
quaternions. It contains a binding for Pixar's RenderMan interface
which is a renderer independent API to communicate 3D data to
renderers which will finally produce a 2D image. There are already
several RenderMan compliant renderers freely available (they are not
part of the kit). The kit also includes some of the functionality from
the RenderMan Shading Language which enables you to create procedural
models or textures from within Python.

Even though the kit focuses on RenderMan, the new types or the
Shading Language functionality can also be used for other rendering
mechanism such as OpenGL or other renderers such as POV-Ray.

The kit should run on any platform where Python (and a C/C++ compiler)
is available. Windows users can download a binary version for Python
2.1 or Python2.2.

What's new?
-----------

The noise module is far more complete than before. In addition to the
normal noise function there's a periodic noise function and a
cellnoise function which is constant between integer points. All
functions also have a vector version that return a vector instead of a
single float.

There has also been a small bugfix in the RenderMan binding 
concerning CSG, and a few other minor modifications (see the
changelog).

The kit should also compile under Linux now (the noise module couldn't
be imported in the previous version).

For more information, visit:

http://cgkit.sourceforge.net


- Matthias -



From srh232 at nyu.edu  Fri Apr 19 14:16:22 2002
From: srh232 at nyu.edu (Shane Hoversten)
Date: Fri, 19 Apr 2002 14:16:22 -0400
Subject: Class method ptrs and dictionaries
Message-ID: <3CC05EF6.2020101@nyu.edu>

Hi -

	I was screwing around with python and found, much to my surprise, that 
this does the "right" thing (or at least, the thing I wanted it to do) 
under these circumstances:

---

def nothing():
     print "I'm in nothing."

class foo:
     def bar(self):
         funcDict = {}

         funcDict["funcPtr"] = self.printOne
         func = funcDict["funcPtr"]
         func("Hello, what!")

         funcDict["funcPtr"] = self.printTwo
         func = funcDict["funcPtr"]
         func("Hello, what!", "I say!")

         funcDict["funcPtr"] = nothing
         func = funcDict["funcPtr"]
         func()

     # Class methods to get "pointers" to.

     def printOne(self, word):
         print "printOne: " + word

     def printTwo(self, word1, word2):
         print "printTwo: " + word1 + " " + word2


---

 >>> a = foo.foo()
 >>> a.bar()
printOne: Hello, what!
printTwo: Hello, what! I say!
I'm in nothing.

	My question: how does python "know" that the functions I'm putting in the 
dictionary are class methods that need to be handed a self pointer in 
the first two cases, and not handed one in the third case?  Are these 
issues described somewhere that somebody could point me to so I could 
read about how python handles this behind the scenes?  I did a couple 
quick searches but didn't turn anything up...

Thanks,

Shane



From tim.one at comcast.net  Tue Apr 16 02:10:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 02:10:01 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <a9g8ok$9a3$1@panix1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEJMOOAA.tim.one@comcast.net>

[Aahz]
> What's a "co-finite set"?  (Yeah, I could Google for it, but I usually
> understand *your* explanations -- eventually. ;-)

Well, David and Andrew gave you their highfalutin explanations, which were
technically accurate so far as they went.  Now for the Uncle Timmy
explanation:  a co-finite set is the color red!  That's all there is to it.
All the rest is a bunch of academic hot air.

there's-a-reason-people-understand-me<wink>-ly y'rs  - tim

PS:  Zorn's Lemmon is the conjecture that lemmons are yellow.  This
     has almost been proved, where by "proved" we mean talked about
     energetically on a newsgroup somewhere, and "conjecture" means
     somebody said so twice.





From sdm7g at Virginia.EDU  Mon Apr  1 16:08:47 2002
From: sdm7g at Virginia.EDU (Steven Majewski)
Date: Mon, 1 Apr 2002 16:08:47 -0500 (EST)
Subject: Files, Seek, and Readlines. A newbie question
In-Reply-To: <Ts%p8.149254$ZR2.64157@rwcrnsc52.ops.asp.att.net>
Message-ID: <Pine.OSX.4.43.0204011553210.1668-100000@d-128-61-180.bootp.virginia.edu>


* seek() always seeks to a byte location in a file.
  There is no seek to a line.

* readlines() reads ALL of the lines in a file. If you do:

 >>> lines = file.readlines()

  then the first line is  'lines[0]' and the 49th line is 'lines[48]'.
  (The optional argument to readlines() is the blocking size -- how
   much to read at a chunk, and is there for optimizing the reads.)
  readline() always reads the next line in the sequence.

* You should try reading the docs at <http://www.python.org/doc/>.
  The docs for file object methods are at:
  <http://www.python.org/doc/current/lib/bltin-file-objects.html>

* If you want to seek to a particular place in a file, you should
  write fixed length records rather than variable length lines of
  text. The reason you can't seek to a line in a file is that there's
  no way of knowing where the next line starts until you have read
  the line before it, so to read the 50th line, you have to read
  lines 1 to 49. If you use fixed length records, the to seek to
  the 50th record, you seek to the  50*size_of_record_in_bytes byte
  in the file.


Is this a homework assignment ? We seem to be getting variations on
this question a lot all of a sudden.


-- Steve Majewski






From akhar at videotron.ca  Tue Apr  9 04:51:38 2002
From: akhar at videotron.ca (akhar)
Date: Tue, 9 Apr 2002 01:51:38 -0700
Subject: porting python to embeded devices
Message-ID: <_lvs8.33252$lU2.1209375@wagner.videotron.net>

Has anyone ported Python to embedded devices? I heard of the deeply embedded
python porject but the link is dead. What would it take to port python to an
embedded platform (whether be it the mc68hc11 or PICs)?

Regards

Stephane




From marklists at mceahern.com  Thu Apr 25 18:38:20 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 25 Apr 2002 17:38:20 -0500
Subject: configuring mxCGIPython
In-Reply-To: <3CC87CAE.6B3F2FD5@lemburg.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEKHKMAA.marklists@mceahern.com>

[Marc-Andre Lemburg]
> Because Python always does on startup if not run with the
> -S switch (this is not available in cgipython though).
> Could you send the output of running cgipython
> with the environment variable PYTHONVERBOSE defined to 1 ?

I've appended the output below.  It looks like it's stuck here:

# site.py
95: if os.name == "posix" and os.path.basename(sys.path[-1]) == "Modules":

evaluating sys.path.  Hmm?  It'll take me a while to figure out what's going
on here.

> A simple work-around would be adding the site.py from the
> standard Python lib -- this should really be included in
> cgipython, though, and I think it is, so there must
> be something in site.py which fails.

If I put a site.py in the same path as cgipython, will that prevent it from
reading its built-in one?  I need to experiment with that.

Thanks for your help!

// mark

<pythonverbose=1 output>

cgipython exists
import site # frozen
import os # frozen
import posix # builtin
import posixpath # frozen
import stat # frozen
import UserDict # frozen
'import site' failed; traceback:
Traceback (most recent call last):
  File "/usr/local/src/Python/Python-2.2/Lib/site.py", line 95, in ?
IndexError: list index out of range
Python 2.2 (#1, Feb 19 2002, 17:44:05)
[GCC 2.95.2 20000220 (Debian GNU/Linux)]
Copyright (c) 2001 Python Software Foundation.
All Rights Reserved.

Copyright (c) 2000 BeOpen.com.
All Rights Reserved.

Copyright (c) 1995-2001 Corporation for National Research Initiatives.
All Rights Reserved.

Copyright (c) 1991-1995 Stichting Mathematisch Centrum, Amsterdam.
All Rights Reserved.
import __main__ # frozen
Traceback (most recent call last):
  File "../../CGIPython/cgipython.py", line 27, in ?
IOError: [Errno 2] No such file or directory
# clear __builtin__._
# clear sys.path
# clear sys.argv
# clear sys.ps1
# clear sys.ps2
# clear sys.exitfunc
# clear sys.exc_type
# clear sys.exc_value
# clear sys.exc_traceback
# clear sys.last_type
# clear sys.last_value
# clear sys.last_traceback
# restore sys.stdin
# restore sys.stdout
# restore sys.stderr
# cleanup __main__
# cleanup[1] signal
# cleanup[1] site
# cleanup[1] posix
# cleanup[1] exceptions
# cleanup[2] stat
# cleanup[2] posixpath
# cleanup[2] UserDict
# cleanup[2] os
# cleanup[2] os.path
# cleanup sys
# cleanup __builtin__
# cleanup ints: 3 unfreed ints in 1 out of 3 blocks
# cleanup floats





From brueckd at tbye.com  Mon Apr 22 15:12:27 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 22 Apr 2002 12:12:27 -0700 (PDT)
Subject: Science And Math  Was: Python's Lisp heritage
In-Reply-To: <pUWw8.48099$vm6.7526151@ruti.visi.com>
Message-ID: <Pine.LNX.4.33.0204221136540.25157-100000@ns0.tbye.com>

On Mon, 22 Apr 2002, Grant Edwards wrote:

> >> Science has an external reference point. Mathematics does not.
> >> Internal consistency is the only thing which mathematics can
> >> attempt to verify.  Mathematics is not an attempt to describe
> >> the physical world.
> >
> > First, please back up, I was asking if there's a difference between
> > discovery and invention, which you don't mention anywhere in your
> > response.
>
> Not literally, no.  I thought it obvious that since math
> doesn't have a physical reference, it is pure invention, while
> science due to it's ties to the physical world is a process of
> discovery.

Hmm... this sounds a lot like "invention and discovery are different
because I have defined them that way". Where does the requirement for a
physical reference come from? That seems orthogonal to invention versus
discovery.

Side note: if that were the case (that math is pure invention), then the
likelihood of two people inventing multiplication at roughly the same time
is quite small. The likelihood of them *discovering* it at the same time
is relatively much higher.

> > Second, math *is* an attempt to describe the physical world
>
> In my experience, that's a rather unique opinion, and one not
> held by any of the math type people I've met.

Perhaps, but as long as we're having a philisophical discussion, that is
completely irrelevant. :)

> > Case in point: calculus, which was invented/discovered
> > specifically to deal with and describe the motion of physical
> > bodies. Had it not been useful in describing such motion, it
> > would have been tossed out, or at least not so widely accepted.
> > Math is useful because of its relevance to the world around us.
>
> I agree that many fields of math are useful.  That is not,
> however a fundamental requirement of math.

Isn't it? Without some reference point to the real world you can't even
learn enough math to theorize that it is self-contained. Yes, it's nifty
that others have shown that math *can* be completely abstract, but they
can't even get to the point of discussing such a concept without first
crossing the bridge that connects it to our existence.

> > Finally, in the more general sense, formal mathematical proofs
> > and whatnot might not require an external reference point, but
> > the foundation upon which they are built certainly does.
>
> That's not the way I learned it.  You start with a set of
> fundamental assumptions which you do not attempt to "prove".
> You then build on that.

Unless you went to a very bizarre school, you learned it by relating it to
things you could actually experience. It's not that you didn't need to
prove that one apple plus one apple makes two apples, through your own
experience you could verify it with sufficient certainty to make it a
useful and safe building block (and then later you could learn enough math
to go back and prove it from a purely mathematical perspective).

Once you're up and running, so to speak, you can move away from this
bootstrap dependency, but there's nothing to suggest that a person can
forego this process.

> > Indeed, the fact that we've gotten as far as something like
> > abstract algebra is largely due to the fact that the underlying
> > building blocks *are* verifiable and applicable to our
> > experiences outside of math and many of the "advances" in
> > mathematics have been due to intuition or hunches provoked by
> > real world phenomena.
>
> Such hunches would imply that Euclidean geometry is an accurate
> description of space...

Well, an example of what might be considered a bad hunch doesn't dimish my
point in the least as there are plenty of cases where good hunches existed
too. :) And, for many purposes, Euclidean geometry *is* an accurate
description of space, and it may be nothing more than someone's hunch that
prompts us to refine our current understanding of space even further
beyond what we think of it today.  Hunch. (sorry, I desperately wanted to
reach my goal of using 'hunch' five times in single paragraph).

-Dave





From python at rcn.com  Wed Apr 24 14:24:59 2002
From: python at rcn.com (Raymond Hettinger)
Date: Wed, 24 Apr 2002 14:24:59 -0400
Subject: PEP 276 -- What else could iter(5) mean?
References: <mailman.1019576624.29992.python-list@python.org>
Message-ID: <aa6tbq$hg1$1@bob.news.rcn.net>

> > to me 5.iter should return an unlimitedly long string of 5's.
>
> Not 0..5 then?

or perhaps [0..5)


Raymond Hettinger




From phr-n2002a at nightsong.com  Fri Apr  5 07:37:05 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 04:37:05 -0800
Subject: PEP 285 and pickle compatibility
References: <7xsn6auy3i.fsf_-_@ruckus.brouhaha.com>
Message-ID: <7xg02ae2z2.fsf@ruckus.brouhaha.com>

Paul Rubin <phr-n2002a at nightsong.com> writes:
> Proposed fix: in new pickles, dump boolean True as "I01" and boolean
> False as "I00".  The current unpickler should load these as the integer
> values 1 and 0 and everything should work fine.  But these strings
> will never be generated by the current pickler, which wouldn't emit
> the leading 0's.  

I should have mentioned:

Note that if the old and new python versions are sending pickles to
each other, and the old version unpickles a bool under this scheme
and then repickles it and sends it back, the object's booleanness
will be lost.  The new version will send True and get back integer 1.

That's probably preferable default behavior to making a pickle that
old versions can't read; however, sometimes, it may not be the right
thing.  I guess a flag (i.e. the format code described earlier) passed
to the dump function should control what happens.

I can see some other possible ways to handle all this too.

It really should have been discussed in more detail in the PEP.


From Answer.via.news.please.prikrylp at nospam_skil.cz  Wed Apr 24 01:17:45 2002
From: Answer.via.news.please.prikrylp at nospam_skil.cz (Petr Prikryl)
Date: Wed, 24 Apr 2002 07:17:45 +0200
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com>
Message-ID: <aa5f7l$uck$1@news.nextra.cz>

Well, I have meant what quotes should I use in cases
when only single or only double are to be used (i.e.
when I do not want to include the opposite inside
the string.  Read it what *style* is prefered.

"Philipp Lenssen" wrote...
> "Petr Prikryl" wrote...
> > In other words, should I prefer writing:
> >     s1 = 'first string'
> > or
> >     s2 = "second string"
>
> For XHTML string output, I prefer single quotes
> because I don't have to escape double-quotes
> within the string, e.g.:
> s += '<img src="..." alt="" />'

Here you can also write:
s += "<img src='...' alt='' />"

... so this does not make it clearer ;-)

Petr




From clive at emultek.co.il  Thu Apr 18 11:27:43 2002
From: clive at emultek.co.il (Clive Bluston)
Date: Thu, 18 Apr 2002 17:27:43 +0200
Subject: ODBC and Excel
Message-ID: <3cbed86d@news.barak.net.il>

I am new to Python. I am trying to read from an Excel file and cannot even
get connected. The following code gives me:
dbi.internal-error: [Microsoft][ODBC Driver Manager] Invalid string or
buffer length in LOGIN
Has anyone any idea what I may be doing wrong?

import dbi, odbc;

        s = odbc.odbc('ODBC;Driver={Microsoft Excel Driver
(*.xls)};DBQ=d:\\clive\\python\\DEFAULT.XLS;DriverID=790')
 Timeout=5;')





From lac at strakt.com  Tue Apr  9 02:42:46 2002
From: lac at strakt.com (Laura Creighton)
Date: Tue, 09 Apr 2002 08:42:46 +0200
Subject: Off topic C question 
In-Reply-To: Message from ryanmorillo@hotmail.com (Ryan) 
   of "08 Apr 2002 19:58:05 PDT." <4a7fde99.0204081858.58baabdd@posting.google.com> 
References: <4a7fde99.0204081858.58baabdd@posting.google.com> 
Message-ID: <200204090642.g396gkdj021548@ratthing-b246.strakt.com>

> I know this is way off topic, but I figure the heavy hackers here will
> understand my mindset better than those at comp.lang.C
>     I have been doodling with python (nasty habit I know but I have to
> do somthing with my snake) and think I may be ready to learn a second
> language.  I am leaning towards C over C++ and Java even though I
> loose objects it will ground me firmly in functional programing.

If you want to learn 'functional programming' then C is not the place.
I recommend Haskell.  There are other options, and 
http://www.cs.nott.ac.uk/~gmh//faq.html is not a bad place to start
checking them all out.  C is a place to learn procedural programming, 
despite the fact that there are no procedures in C, only functions
that do not return a value.  If you only want to learn 'how to
program in a language that does not have classes' then there are
easier places to do that than C.  However, if your real goal is
to hack C, and all the above is just one misused word -- you want
K & R.  Everybody wants K & R.  If you find it tough going, you 
may want something else in addition to K & R, just to get you
started, but you also really want K & R.

Happy Hacking,
Laura Creighton




From henne021 at tc.umn.edu  Tue Apr  2 17:54:20 2002
From: henne021 at tc.umn.edu (Scott Hennesy)
Date: Tue, 2 Apr 2002 22:54:20 +0000 (UTC)
Subject: A faster Python?
Message-ID: <a8dcqr$p29$1@laurel.tc.umn.edu>

For what I do, 90% of the time Python is fast enough. But for that 10% of the time I find myself wishing for a faster Python. 

What's the consensus out there? Do Psyco, NumPy, and SciPy fulfill the need for speed or is there a desire for a faster Python? 

If so, what are the requirements for a faster Python? How much faster does it have to be? How much compatibility does it need with Python? How much resources can be devoted to optimization? Easy example: 20% faster, no compatibility, infinite resources. Hard example: Faster than optimized C, total compatibility, no resources ;-)

I've considered doing something about maybe a half dozen times. Maybe with Python 2.2 the time is ripe?

Scott



From mpoulin at verinet.com  Sat Apr  6 10:00:12 2002
From: mpoulin at verinet.com (Marc Poulin)
Date: Sat, 06 Apr 2002 10:00:12 -0500
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org>
Message-ID: <3caf28ce$0$69917$7586b60c@news.frii.com>

In article <mailman.1018081037.4211.python-list at python.org>, "David
McInnis" <david at dataovation.com> wrote:

> I need to alter the document properties of Word DOCS that our users

More info, please. What properties are you specifically trying to 
alter?


From bokr at oz.net  Wed Apr 24 18:38:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Apr 2002 22:38:26 GMT
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com> <slrnacc59a.tg8.jon+usenet@snowy.squish.net> <MCxx8.45058$0x3.38185@atlpnn01.usenetserver.com>
Message-ID: <aa7c52$ehu$0@216.39.172.122>

On Wed, 24 Apr 2002 08:31:09 -0400, "Steve Holden" <sholden at holdenweb.com> wrote:

>"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
>news:slrnacc59a.tg8.jon+usenet at snowy.squish.net...
>> In article <u3ex8.94967$T%5.10614 at atlpnn01.usenetserver.com>, Steve Holden
>wrote:
>> > This error was caused by Fernando's following a bogus suggestion of
>mine.
>> > His original SQL syntax was correct
>>
>> Uh, no it wasn't, he was using '%s' instead of %s. I'm confused now.
>
><sigh> Sorry, I should have had the courage of my own convictions. When
>Fernando told me he was still getting errors on his SQL I assumed that the
>MySQLdb module required delimiters around string parameters. You're right,
>this doesn't make any sense. You should have been confused. It eventually
>transpired that Fernando had an install problem, which a clean reinstall
>fixed.
>
>>>> import MySQLdb
>>>> c = MySQLdb.connect(db="pwpsite")
>>>> cc = c.cursor()
>>>> cc.execute("SHOW TABLES")
>5L
>>>> cc.fetchall()
>(('chapter',), ('comment',), ('part',), ('remark',), ('reviews',))
>>>> cc.execute("CREATE TABLE tmp (k integer primary key, s char(12))")
>0L
>>>> cc.execute("INSERT INTO tmp (k, s) VALUES (%s, %s)",
>...     [(1, "one"), (2, 'two'), (3, 'three')] )
>3L
>
>So, for the record, you should *not* put SQL string quotes around parameter
>markers in your parameterized SQL statements.
>
So is cc.execute(...) acting like
 >>> def dummyccx(fmt, params):
 ...     for tup in params:
 ...         print fmt % tup
 ...
 >>> dummyccx(
 ... "INSERT INTO tmp (k, s) VALUES (%s, %s)",
 ... [(1, "one"), (2, 'two'), (3, 'three')]
 ... )
 INSERT INTO tmp (k, s) VALUES (1, one)
 INSERT INTO tmp (k, s) VALUES (2, two)
 INSERT INTO tmp (k, s) VALUES (3, three)

or
 >>> def dummyccx(fmt, params):
 ...     for tup in params:
 ...         print fmt % tuple(map(repr,tup))
 ...
 >>> dummyccx(
 ... "INSERT INTO tmp (k, s) VALUES (%s, %s)",
 ... [(1, "one"), (2, 'two'), (3, 'three')]
 ... )
 INSERT INTO tmp (k, s) VALUES (1, 'one')
 INSERT INTO tmp (k, s) VALUES (2, 'two')
 INSERT INTO tmp (k, s) VALUES (3, 'three')

or
 >>> def dummyccx(fmt, params):
 ...     ins = fmt.split('VALUES')
 ...     out = []
 ...     for tup in params:
 ...         out.append(ins[1] % tuple(map(repr,tup)))
 ...     print ins[0]+'VALUES'+', '.join(out)
 ...
 >>> dummyccx(
 ... "INSERT INTO tmp (k, s) VALUES (%s, %s)",
 ... [(1, "one"), (2, 'two'), (3, 'three')]
 ... )
 INSERT INTO tmp (k, s) VALUES (1, 'one'),  (2, 'two'),  (3, 'three')

where print is taking the place of passing actual SQL to mySQL.

IOW, I think mySQL wants quotes around 'one' 'two' and 'three', so it's
probably being arranged somewhere?

Regards,
Bengt Richter


From thomas at gatsoft.no  Wed Apr 24 06:36:14 2002
From: thomas at gatsoft.no (Thomas Weholt)
Date: Wed, 24 Apr 2002 10:36:14 GMT
Subject: Quick question about Webware
Message-ID: <yUvx8.30$RNd.170832384@news.telia.no>

When you create a servlet using WebWare is it loaded on each request or is a
instance kept in memory?
I mean, what if the servlet uses some module or piece of code that takes a
while to load, how is this solved in WebWare?

Thomas




From joost_jacob at hotmail.com  Sat Apr 20 13:22:53 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 20 Apr 2002 10:22:53 -0700
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com>
Message-ID: <13285ea2.0204200922.6fb906f9@posting.google.com>

Here is your code modified so it works.
As you can see 'lambda' is used.  Read more about it in the manuals
and books, have fun learning things about Tkinter, i did like it.

#!/usr/bin/python

from Tkinter import *

def printNumber(number):
	print number
	
win = Tk()
Button(win, text='print 3', command=lambda: printNumber(3)).pack()
Button(win, text='print 5', command=lambda: printNumber(5)).pack()
win.mainloop()


From shalehperry at attbi.com  Tue Apr 16 12:08:56 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 16 Apr 2002 09:08:56 -0700 (PDT)
Subject: Where should 'import' statement be put?
In-Reply-To: <4620daca.0204160749.1f68a62a@posting.google.com>
Message-ID: <XFMail.20020416090856.shalehperry@attbi.com>

> 
> I believe imports are done only once regardless of the number of times
> the statement appears, but are there any other considerations here? I
> realize there may be scoping issues as well.
> 

def foo(input):
    import a_module

    a_module.do_things


if foo is never called a_module is never imported.  So if you could place
potentially unused modules inside functions this helps.  But this is also a
matter of style.




From max at alcyone.com  Thu Apr  4 14:14:37 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 04 Apr 2002 11:14:37 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com> <slrnaannkc.6ls.kzaragoza@localhost.localdomain> <3CABEA67.FD3F75C0@alcyone.com> <a8hq6o$o9a$1@panix1.panix.com>
Message-ID: <3CACA61D.C59BC837@alcyone.com>

Aahz wrote:

> Did you read the final version?  C'mon, don't start arguments where
> you
> have no clue what you're talking about.

Hence why I said "alternate choice."  Sheesh.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From pearu at cens.ioc.ee  Tue Apr 30 16:00:09 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Tue, 30 Apr 2002 23:00:09 +0300 (EEST)
Subject: import from a file in a subdirectory
In-Reply-To: <20020430193802.GF16217@devel.trillke>
Message-ID: <Pine.LNX.4.21.0204302249130.25862-100000@cens.ioc.ee>


On 30 Apr 2002, holger krekel wrote:

> On Tue, Apr 30, 2002 at 12:27:03PM -0700, rdack wrote:
> > i have a script running in a directory.
> > i want to call a subroutine in a file in a subdirectory.
> > 
> > in the first file i said  from subdir/ff.py import  subrtn
> > python doesn't like the slash.
> > 
> > how do i do it?
> 
> import sys
> sys.path.append('subdir')
> 
> import ff

I would suggest 

  sys.path.insert(0,'subdir')
  import ff
  del sys.path[0]

in case there also exists ff module/package in Python tree.

And note that if there is a ff Python module/package and it is imported
before, say, by other modules, then the above hooks are useless as

  import ff

uses already imported module, not the one in 'subdir'. I am not sure but
may be reload(ff) will fix this:

  sys.path.insert(0,'subdir')
  import ff
  reload(ff)
  del sys.path[0]


Regards,
	Pearu





From pyth at devel.trillke.net  Tue Apr 30 20:14:07 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 1 May 2002 02:14:07 +0200
Subject: Newbie: finding the key/index of the min/max element
In-Reply-To: <200204302300.g3UN0kkG002462@ratthing-b246.strakt.com>
References: <200204302300.g3UN0kkG002462@ratthing-b246.strakt.com>
Message-ID: <20020501001407.GJ16217@devel.trillke>

On Wed, May 01, 2002 at 01:00:46AM +0200, Laura Creighton wrote:
> > Benjamin Han wrote:
> > 
> > > In my quest of becoming a "Pythonistas" (where does this word come from BTW
> > ?),
> > 
> > The term derrives from the Sandinistas in Nicaragua, the Communist rebels who
> > fought against the US-installed conservative Somoza dictatorship (and others)
> > .
> > The suffix is sometimes used to refere to any contrarian group, and in cases
> > simply any group.
> 
> That is simple ignorant American narrow-minded foolishness talking.
> Normally, I would take extreme great care to not say anything so 
> offensive, but I have a deep suspcion that James J Besemer is one
> of the 'can't take the heat' folks, and will learn to like a kinder
> and gentler place than the one he proposes to make for Raymond Hettinger
> if he finds out what it is like to live in the place he proposed.

one moment, please. You are making some valid points but i don't
see any sense in this 'eye for an eye' attitude. 
After all beeing ignorant about something and the general
wish to collaboratively advance is the reason we are subscribed 
on this list, isn't it? Why not explain your 'ista' origins to the 
*poster of the original question*? 

If many pythonistas would react to peoples (dumb) questions
and (ignorant) suggestions (including mine) within the scheme
of your posting, many of us might have quit programming alltogether 
instead of enjoying the wonders of python and its community.

Sure enough i share your general criticism on US-politics and
especially their foreign politics.  But i don't think that
*anything* is won if you try to hammer it into somebodies head.

regards,

    holger




From mwh at python.net  Thu Apr  4 04:13:25 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 4 Apr 2002 09:13:25 GMT
Subject: run_pyc_file:  nested_scopes: 0
References: <Z6Mq8.229$c6.61189@typhoon.mn.ipsvc.net>
Message-ID: <lkzo0j3jyj.fsf@elios.maths.bris.ac.uk>

"joel8bit" <jwickard**nospam**@litriusgroup.com> writes:

> I juste exported a python program I was writing from my win2k box to a win98
> box.  I only exported the pyc files for the program.  When I try to run the
> program on the win98 machine I get this output:
> 
> run_pyc_file: nested_scopes: 0
> 
> I don't know exactly what's going on.

I *think* that's just a random debugging printf left in by mistake.
What version are you running?  If 2.1, you should probably upgrade to
2.1.2 (or wait a week and upgrade to 2.1.3).

Cheers,
M.


From theller at python.net  Fri Apr  5 07:40:59 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 5 Apr 2002 14:40:59 +0200
Subject: CGI GET vs. POST
Message-ID: <a8k613$suseq$1@ID-59885.news.dfncis.de>

I'm trying to use the cgi module in a CGI script,
but encountered a problem.

My form looks like this:

<form method=...>
<input name=keyword>
<input type=submit>
</form>

and this is part of the Python code in the script:

form = cgi.FieldStorage()
if form.has_key("keyword"):
    keyword = form["keyword"].value
    ... and so on.

If I use method=POST in the CGI-script, the
"keyword" is found as a field in FieldStorage,
but NOT if I use method=GET.
Is FieldStorage only usable with POST?

Thanks,

Thomas




From laz at strakt.com  Tue Apr 16 07:08:18 2002
From: laz at strakt.com (Fredrik Juhlin)
Date: Tue, 16 Apr 2002 13:08:18 +0200
Subject: Namespaces
In-Reply-To: <3cbc0275@news.comindico.com.au>
References: <3cbc0275@news.comindico.com.au>
Message-ID: <20020416110818.GI26982@strakt.com>

On Tue, Apr 16, 2002 at 08:52:29PM +1000, Graeme Matthew wrote:
<snip>
> #! C:\\python22\\python.exe
> 
> import sys
> 
> sys.path.append('C:\\www\\classes')
> import Template
> 
> template = Template.Template()
> 
> print "done"
> 
> I cannot get it to instantiate like
> 
> template = Template()

Try to replace your import row with:
from Template import Template

//FJ




From garrett at bgb-consulting.com  Mon Apr  1 13:42:03 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Mon, 01 Apr 2002 18:42:03 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <3CA8A9F3.62164864@bgb-consulting.com>

  One alternative approach does come to mind..... If there were really good
general support for enumerations, then boolean could be implmented as an
enumeration, not as it's own construct.

Guido van Rossum wrote:

>     1) Should this PEP be accepted at all.

  When I first started learning Python, I considered the lack of a Bool type a
horrible wart, and promptly added 'TRUE' 'FALSE' constants to everything I
write. Since then I've gotten over it, dropped the constants, and been
perfectly happy.

  The language doesn't NEED bool, and languages should never be amy bigger
than needed (look at Ada). However, this is a very common construct, and it's
what instructors and programmers new to the language expect. A language should
work the way you expect.

  My vote: +0
 
>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".)

  Return "True" and "False" in all cases. If you're going to do it, do it
right!

>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

  No, I don't think they cause problems, and often make code simpler.
 
>     5) Should operator.truth(x) return an int or a bool.  Tim Peters
>        believes it should return an int because it's been documented
>        as such.  I think it should return a bool; most other standard
>        predicates (e.g. issubtype()) have also been documented as
>        returning 0 or 1, and it's obvious that we want to change those
>        to return a bool.

  Again, if you're going to do it, do it right! Everything in the libraries
(and in new code) that is a boolean value should use a boolean type.

  That documentation was correct for the old versions of Python. The
documentation should be updated for the new versions!

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From db3l at fitlinxx.com  Thu Apr  4 16:16:01 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 04 Apr 2002 16:16:01 -0500
Subject: Downloading pages from a Windows Intranet
References: <mailman.1017936436.8304.python-list@python.org>
Message-ID: <uit77cgha.fsf@ctwd0143.fitlinxx.com>

Gustavo Cordova <gcordova at hebmex.com> writes:

> So, unless you're using some kind of authentication which is not
> part of the http protocol, we can't really help you much unless
> you give us a bit more information.

Microsoft's IIS has an "NTLM" authentication which is proprietary to
them that can be enabled to permit secure access to a site (typically
used on an Intranet) without requiring additional authentication
boxes.  In an all-MS shop, it works really well to let people use
their already authenticated domain information for Intranet security
as well, so it's probably reasonably common to turn on.

Note that while the actual authentication mechanism is proprietary (an
earlier post her did reference a Python module to help support this),
the HTTP protocol has generic support for authentication mechanisms in
general, so this doesn't violate HTTP itself.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aahz at pythoncraft.com  Sun Apr  7 16:02:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 16:02:01 -0400
Subject: python newbie
References: <d9ea2696.0204071156.4284b1f1@posting.google.com>
Message-ID: <a8q8jp$fob$1@panix1.panix.com>

In article <d9ea2696.0204071156.4284b1f1 at posting.google.com>,
distrex <distrex at hotmail.com> wrote:
>
>I'm trying my first script with a module and I can't get it to work
>right.  I'm using the time module to make part of the game random and
>i just want the number to be at a whole number (no decimals) and I
>can't figure out how to make it do that. If someone needs me to post
>the script I'd be happy to but if you can help me just from memory
>that'd be great too. Thanx in advance.

Use the random module.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From mwh at python.net  Fri Apr 12 06:27:37 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 10:27:37 GMT
Subject: Communication between a Python process and a Python cgi script
References: <mailman.1018590774.9993.python-list@python.org>
Message-ID: <lk8z7tjgoo.fsf@pc150.maths.bris.ac.uk>

Todd Warner <taw at redhat.com> writes:

> Another option:
> cgi script sends a kill and SIGALRM. *shrug* Depends on what the risk is.

But you can't send a signal to a process you don't own can you?
(setuid CGI scripts?  Run away!)

Cheers,
M.

-- 
  SPIDER:  'Scuse me. [scuttles off]
  ZAPHOD:  One huge spider.
    FORD:  Polite though.
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11


From gh_pythonlist at gmx.de  Tue Apr  2 12:42:06 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Tue, 2 Apr 2002 19:42:06 +0200
Subject: calling dll's functions with python
In-Reply-To: <8wmq8.52601$S52.1805120@news2.tin.it>
References: <8wmq8.52601$S52.1805120@news2.tin.it>
Message-ID: <20020402174206.GB7571@lilith.hqd-internal>

* Giovanni Laudicina <giovanni.laudicina at libero.it> [2002-04-02 18:12 +0000]:
> I'm a beginner with python and i did'n find this info on the FAQ and
> tutorials:
> How can I call a function included in a dll (I have only the .dll file, not
> the source code)?
 
There's an extension module that does exactly this. Google should find
it - it's name is "calldll".

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 13.3 ?C      Wind: 4.8 m/s



From fperez528 at yahoo.com  Thu Apr 11 20:30:12 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 11 Apr 2002 18:30:12 -0600
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu> <Xns91EDE505F5BD7cliechtigmxnet@62.2.16.82>
Message-ID: <a959s9$s08$1@peabody.colorado.edu>

Chris Liechti wrote:

> Rajarshi Guha <rxg218 at psu.edu> wrote in
> news:a94qk5$22mg at r02n01.cac.psu.edu:
> 
>> On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de Maria
>> wrote:
>> 
>>> Is there a way to save in a file the status, objects, functions of
>>> interpreter
>>> in order to restore it after the interpreter has been closed?
>>> 
>>> If it is possible, one can work on a project stop and then continue
>>> later adding functions, objects and so on.
> 
> thats a question for an IDE with such functionality. i don't know one but i
> think it should be easy to store all objects in the global namespace of the
> interpreter and load it again on next startup.
> the textual history of the interpreter with all the outputs must be saved
> separately.
> 

well, I've tried (not terribly hard) and failed. My ipython project is a fancy 
interpreter (http://www-hep.colorado.edu/~fperez/ipython/) and I wanted 
exactly that kind of functionality. It turns out pickle cant quite pickle 
_everything_ you throw at it, so there's bound to be things in your 
namespaces that you can't pickle.

I'd like to know if there's a way to make this work, because as I said my 
first attempts didn't get me anywhere. My kludgy semi-solution was to 
implement session logging/restore by replaying all input, but that's slow, 
error prone and has many other limitations.

Cheers,

f.


From phr-n2002a at nightsong.com  Sat Apr 13 13:54:08 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 13 Apr 2002 10:54:08 -0700
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com> <slrnabf5r2.s6b.grey@teleute.dmiyu.org> <7xr8lk46nl.fsf@ruckus.brouhaha.com> <3CB83354.90CB406@alum.mit.edu>
Message-ID: <7xn0w7a3i7.fsf@ruckus.brouhaha.com>

Will Ware <wware at alum.mit.edu> writes:
> This thread got me curious about it. There are SL-5000D's (the developer
> version) on eBay trading in the ballpark of $300. Is there any significant
> difference between the 5000D and the 5500?

The 5500 has 64M of memory and the 5000 has only 32M.


From reply at in_newsgroup.tks  Mon Apr  8 17:45:30 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 8 Apr 2002 21:45:30 GMT
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk> <p4is8.795$eJ6.210@atlpnn01.usenetserver.com>
Message-ID: <a8t31q$se$1@hfc.pacific.net.hk>

"Steve Holden" <sholden at holdenweb.com> wrote in
news:p4is8.795$eJ6.210 at atlpnn01.usenetserver.com: 

>> As far as books are concerned (a must) I'm thinking of buying
>> Web Programming in Python (this is written partly(?) by the folks
>> from the Slither project).
>> http://www.amazon.com/exec/obidos/ASIN/0130410659/qid=1017861581/sr=1-
>> 3/ref=sr_1_3/104-5555059-2897528
>>
> I can recommend this one, as the author is a close personal friend ;-)
> But, to set the record straight, apart from small acknowledged
> contributions, the writing (and therefore the mistakes) are all mine
> apart from David Beazley's contribution of the three XML chapters.

Nice try Steve but this isn't your book! The name is similar though :)

[snip]

> If you already have a server, just start writing CGI scripts for it.
> There is a LOT of material on the web explaining how to do this.
> 
> Once you are familiar with web technologies, then pick up Webware or
> similar and start to build applications using your framework of
> choice. Come to c.l.py with questions!

I think you're right. I was originally considering something along the 
lines of jumping into the deepend before learning to swim!

[snip]

> "Python Web Programming" is intended to explain enough of the
> technology, in enough detail, that you can build web systems in Python
> without having to use an external server. It doesn't try to argue that
> this is always the best way to go, though.

...and the (book) final is between "Python Web Programming" and the Slither 
folks' "Web Programming in Python"!! =)

She loves me, she loves me not, she... Geez, if I can't make up my mind 
about which book to buy, there's no hope for me, hehe ;-)

Cheers,
Scrumpy :)


From fredp at mygale.org.nospam  Mon Apr  8 03:38:20 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 8 Apr 2002 07:38:20 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it> <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20> <MMfr8.62526$S52.2173482@news2.tin.it>
Message-ID: <Xns91EA627D3A6D5PaCmAnRDLM@161.48.128.20>

Alex Martelli <aleax at aleax.it> said :

> Funny -- the developer edition WAS already for sale in UK & Ireland...
> which are part of Europe, of course.

Probably very different marketing dept. :-)
 
> I was hoping to order one in UK or Ireland (no customs within the
> EU...). 

True, although the price markup in europe is usually so bad that you end up 
paying the same, or more. The cheapest solution is when you know someone 
who is taking a trip to the US, and willing to buy/bring back one for you 
(especially for small stuff like that). The problem then will be with 
guarantee & repairs -- this is what I meant by "taking chances"...
 
> I go out of my way to get US-standard keyboards wherever I possibly
> can (and use 'US-international' to enter national characters when the
> OS supports that -- Windows and Linux do).  I never grew used to
> Italian layouts (nor French ones, Swedish ones, etc)...

For coding (as such a neat machine inevitably is for :-) that would not be 
a problem. But for day-to-day note-taking and data entry, I still prefer a 
frenchy keyboard with easily accessible accents & stuff. I'm not sure they 
can manage such a layout with the tiny keyboard on the Zaurus, however, so 
the point may be moot :)

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


From gerhard at bigfoot.de  Tue Apr 23 07:07:00 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 23 Apr 2002 11:07:00 GMT
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com>
Message-ID: <slrnacag2g.1u3.gerhard@lilith.my-fqdn.de>

Fernando Rodr?guez wrote in comp.lang.python:
> 
> I'm using CompatMysqldb from http://sourceforge.net/projects/mysql-python with
> Python 2.2
> 
> Whenever I try to insert some data into a mysql db, I get this error:
> 
>>>> curs.execute("insert into customer_info (email, url, referrer, first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
> Traceback (most recent call last):
>   File "<pyshell#26>", line 1, in ?
>     curs.execute("insert into customer_info (email, url, referrer,
> first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com',
> 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
>   File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line 227, in
> execute
>     af =af+self.__res.affectedrows()
> AttributeError: 'NoneType' object has no attribute 'affectedrows'
>>>> 
> 
> The data does get inserted... What am I doing wrong? O:-)

AFAIC, you should be using a tuple as second parameter to execute, not
a list containing a tuple.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From gerhard at bigfoot.de  Tue Apr  2 23:17:57 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 3 Apr 2002 04:17:57 GMT
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net> <slrnaaksu7.368.gerhard@lilith.my-fqdn.de>
Message-ID: <slrnaakt4i.368.gerhard@lilith.my-fqdn.de>

Gerhard H?ring wrote in comp.lang.python:
> Jim Jacobs wrote in comp.lang.python:
>> Is it possible to simulate "C" style static variables in functions?
> 
> No.
> 
>> In methods?
>  
> No.

Stupid me, sure you can: just assign attributes to the
functions/methods, like in:

def foo():
  foo.this_is_a_static_var = 5

You can access the function/method attributes from anywhere.

But I really can't see a good reason to actually use this feature.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From peter at engcorp.com  Sat Apr 13 12:15:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 13 Apr 2002 12:15:49 -0400
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <7E283A6EA4F0DBD6.B68D98BE8C574F5D.CE8B287F1A68C959@lp.airnews.net> <LFGz2JAYZxt8EwR6@jessikat.demon.co.uk> <459A680B92F06092.DD5B3EAA28701812.3A8CCB250DB66074@lp.airnews.net> <3CB853CA.8050507@cutter.rexx.com>
Message-ID: <3CB859B5.7618F570@engcorp.com>

Dave Kuhlman wrote:
> 
> Cameron Laird wrote:
> > While I understand the observations that Python, Tcl,
> > and so on surely aren't safe enough for medical devices,
> > nuclear plants, air traffic control, military devices,
> > and so on, I have reasonably certain knowledge of instances
> > of each of these.
> 
> The idea that Python, Tcl, and so on aren't safe enough for medical
> devices sounds like conventional wisdom to me.  
[...]
> Are there examples and evidence that Python fails when applied to
> long-running, mission critical tasks?

No, and in fact there is significant evidence that Python is quite
suitable for long-running, mission critical tasks, as a number of
us have indicated here.  My group is writing applications which
are expected to run for years without stopping and we're by no
means the extreme.

> By the way, I write C++ code.  And, the idea that C++ code is safer than
> Python code seems screwy to me.  Can you say "seg fault"?  I thought so.

I would definitely write safety-critical code in Python before I
would write it in C++.

I would also spend *significant* (read very expensive) effort
qualifying the interpreter as a component in such a system before
I released my product.

There is nothing wrong with Python the language in a safety 
(even life) critical application.  There is, however, something 
wrong with ignoring the fact that the CPython interpreter has 
not been designed with these types of applications in mind and, 
more to the point, has not been carefully analyzed with such 
a use in mind.  There's nothing except the cost stopping anyone 
who wants to do this from doing so.  It's quite possible,
as suggested by Cameron's comments above, that some groups
have already done so (but for perhaps obvious reasons will
not be releasing any of their changes back to the community).

-Peter


From rjgruet at yahoo.com  Thu Apr 18 12:05:35 2002
From: rjgruet at yahoo.com (Richard Gruet)
Date: Thu, 18 Apr 2002 18:05:35 +0200
Subject: why no 'length' method in sequences ?
Message-ID: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net>

Hi all

I wonder why there is no 'length' (or 'len') method - or attribute - in the
sequences types. Having to use the built-in function len() doesn't sound
very Object-Oriented to me. I understand that the built-in function len()
must be kept for compatibility reasons, but why not to add also the method
len() to the sequence types ? Is there a reason ?

(My apologies if this issue has already been raised. I searched the archives
but found nothing)

Richard Gruet

Advertisement ! Do you know the Python Quick Reference (see
http://rgruet.free.fr) ? It's currently for 2.0 and 2.1, but a version 2.2
should be available in the near future.




From claird at starbase.neosoft.com  Mon Apr  8 12:23:28 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 8 Apr 2002 11:23:28 -0500
Subject: Tkiner send command
References: <3cb0ae7d$1@news.cc.umr.edu>
Message-ID: <C9B1D945FE30F374.B5B450B89599A68A.8783B42B36F655C5@lp.airnews.net>

In article <3cb0ae7d$1 at news.cc.umr.edu>, Carl <kingprad at mail.com> wrote:
>I'm having trouble going from the formal specs in the Tcl/Tk Reference to
>writing an actual Tkinter send command. I just want to start a text editor,
>either the Windows default or just notepad.exe. Can someone please give me
>an example line of how to use it? Thanks very much!
			.
			.
			.
Ask the question again.  You're using Tkinter, and you
want to start notepad, is that right?  In that situa-
tion, send probably doesn't benefit you; instead, you
want something like
  os.system("notepad %s" % filename)
-- 

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


From merman at freenet.de  Tue Apr  9 04:35:45 2002
From: merman at freenet.de (merman)
Date: Tue, 9 Apr 2002 10:35:45 +0200
Subject: Embedding Python with C++
References: <GYts8.26982$2J2.6830478@typhoon.southeast.rr.com>
Message-ID: <3cb2a5dc$0$25705$9b622d9e@news.freenet.de>

"Chris Stoy" <cstoy at nc.rr.com> schrieb im Newsbeitrag
news:GYts8.26982$2J2.6830478 at typhoon.southeast.rr.com...
> Hey all,
>
> I'm looking to embed Python as a scripting language in a game I am working
> on.  I want to be able to call C++ methods from Python, and call Python
from
> C++ (for the usual code quick in Python, convert to C++ for speed later.)
>
> I've looked into using the Boost.Python code and I'm fine with that.
>
> My problem is, I have no clue how to compile Python into a static linked
> library for Windows.  Actually, I need it to be cross-platform, which is
> another issue.
>
> I've never embedded Python before, and I'm fairly new to Python, so can
> someone please give me a pointer to how to do this?  What are the files I
> need to compile to build the static lib (using MS Visual Studio 6.0?)
Also,
> I don't need all of the standard Python components (like networking.)  Can
> those be removed from the lib?
>
> Thanks.
>
> Chris.
>

Hi Chris,

I'm also a python-newbie. My problem is the same:

> I want to be able to call C++ methods from Python.

I have tried a few days with 'swig' (www.swig.org) - but there was no
success (under win2k & bcc55-compiler, commandline).
Maybe you can work with VS6 (recommand in the SWIG-doc) better with SWIG ...
.

Hope this helps

o-o

Thomas















From geoff at gerrietts.net  Tue Apr 30 22:24:57 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 30 Apr 2002 19:24:57 -0700
Subject: Python vs. Perl,  which is better to learn?
In-Reply-To: <9UHz8.23$s55.5@newsread2.prod.itd.earthlink.net>
References: <df30afd4.0204292111.31e5914a@posting.google.com> <chris-4E0B1A.01281230042002@corp.supernews.com> <f0fd5987.0204300105.51529004@posting.google.com> <extz8.43542$dh5.3425591@news-binary.blueyonder.co.uk> <g7Bz8.35106$8D3.1017931@news1.tin.it> <9UHz8.23$s55.5@newsread2.prod.itd.earthlink.net>
Message-ID: <20020501022457.GA28190@isis.gerrietts.net>

Quoting Billy Ng (kwokng at earthlink.net):
> I am also new to python.  I think python will be my replacement of
> perl.
> However, if I need to do very heavy regular expression, I still
> prefer perl.
> I just don't feel comfortable with the python's re syntax.

I understand your discomfort; I had it at first, too. I still
sometimes use Perl for a one-liner (eg perl -pi -e "s/^\s+//g;"), but
I've found that Python's regular expressions are just fine; it's a
matter of getting accustomed to it.

Of particular interest to me (and maybe Perl does this now, but it
didn't when I was deep into Perl Zen) is the way you can name
subpatterns in Python -- that allows you to use your parens for
grouping, and not hafta pick through the countless groups for the
actual pieces you want to extract.

I also appreciate the way Python actively encourages a person to
compile their regular expressions -- Perl doesn't, which leads to
continually re-constructing the state machine (unless they've
optimized that since I was deep into Perl Zen).

Luck,
--G.

-- 
Geoff Gerrietts             "There is no fate that cannot be 
<geoff at gerrietts net>     surmounted by scorn." --Albert Camus




From matthewm at zebrasoft.co.nz  Mon Apr 29 04:26:07 2002
From: matthewm at zebrasoft.co.nz (Matthew)
Date: Mon, 29 Apr 2002 20:26:07 +1200
Subject: Returning variable results from a function
References: <mailman.1020067191.11039.python-list@python.org>
Message-ID: <3CCD039F.7C5A01A1@zebrasoft.co.nz>

EXCELLENT! Thank-you. matthew.

"Delaney, Timothy" wrote:

> > From: Matthew [mailto:matthewm at zebrasoft.co.nz]
> >
> > Given something like:-
> >
> > def fred(size):
> >     return size, 1,2,3
> >
> > size,a,b,c=fred(3)
> >
> > how does one create a tuple to hold the results like:-
> >
> > size, (<a tuple holding the number of results?>) = fred(<no of
> > results?>)
>
> What you want to do is return the 1, 2, 3 as a tuple ...
>
> def spam (size):
>     return size, (1, 2, 3,)
>
> size, eggs = spam(3)
>
> Here you are returning a tuple, which has size and another tuple as its
> elements.
>
> However, if you mean "how do I take 'size' and produce a tuple of that
> length, already populated, and return it"
>
> def spam (size):
>     return size, tuple(range(1, size + 1))
>
> size, eggs = spam(3)
>
> Tim Delaney



From jeff at ccvcorp.com  Tue Apr  9 15:56:12 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 9 Apr 2002 12:56:12 -0700
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com> <slrnab4m82.glr.mattdm@jadzia.bu.edu> <3CB277E2.8060505@greenroom.com.au>
Message-ID: <MPG.171ce737fe06848a9896a3@news.nwlink.com>

In article <3CB277E2.8060505 at greenroom.com.au>, 
matt at greenroom.com.au says...

> In addition to that (i promise i'll stop soon), it is a bit annoying 
> that whenever i install red hat i have to
> 
> rm -f /usr/bin/python
> ln -s /usr/bin/python2 /usr/bin/python
> 
> because they believe that more of us will use 1.5 than 2.1.

No, they have no such belief.  It is simply that their 
administration tools use 1.5, and are broken under 2.x, and said 
tools all look for /usr/bin/python.  In other words, what you've 
just done (relinking /usr/bin/python to python2) has broken many 
of their admin tools.

[trimmed X-posts]

-- 

Jeff Shannon
Technician/Programmer
Credit International


From tim.one at comcast.net  Wed Apr 10 19:48:22 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 19:48:22 -0400
Subject: Easter Eggs
In-Reply-To: <a910e5$vnkga$1@ID-135695.news.dfncis.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEALONAA.tim.one@comcast.net>

[Bertrand Geston]
> That reminds me that I always wonder about the meaning of this one:
>
>    Flat is better than nested.
>
> Is it because of my OO background or do I miss something ?
> Any explanation ?

One guess as to how Guido feels about XML for his own use.  If you guessed
right, you'll also know why "lexical scoping" came so late in Python's life,
and 7th-Level Enlightenment is just a hefty donation (or two) beyond that.

enlightened-is-better-than-hungry-ly y'rs  - tim





From goodger at users.sourceforge.net  Tue Apr  2 23:11:46 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Wed, 03 Apr 2002 04:11:46 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <uaknnsdgu514a6@news.supernews.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <uaknnsdgu514a6@news.supernews.com>
Message-ID: <B8CFEB59.20D3E%goodger@users.sourceforge.net>

John Roth wrote:
> I don't particularly like the table syntax, (it's too much typing
> for my taste) but I don't have a reasonable alternative.

The readability of tables is far more important than the ease of writing.
Emacs table-mode is very useful for producing ASCII-art tables.

> The one question I've got is very, very basic:
> 
> why is it in a docstring?

The first sentence of the Abstract answers this:

    When plaintext hasn't been expressive enough for inline
    documentation, Python programmers have sought out a format for
    docstrings.

> Lots of people use docstrings for relatively short, online prompts.

I'm not sure what you mean.  I see three possibilities:

1. Are you using docstrings for something other than documentation [#]_?
   Then this PEP may not apply to your application.

   .. [#] After all, "docstring" does derive from "documentation string".

2. Are you referring to the "tooltips" that pop up in IDEs when you type
   a function/method name?  The ones I'm familiar with are generated
   from the parsed function/method signatures.

3. Are you objecting to extended, comprehensive docstrings, that may be
   unsuitable for ``print my_object.__doc__``, and may take up space in
   the running program?  Then the "Additional Docstrings" proposal from
   PEP 258 may reassure you.

If those don't address your concern, please elaborate.

> Some other mechanism should exist for extensive documentation that is
> intended for translation to some other format for eventual printing
> or display independent of the program.

That's the whole point of docstring extraction tools.  This PEP is merely
proposing a standard for the format of those docstrings.

> I much prefer not to have this stuff in the compiled module.

Then don't put it in.  Some don't want inline documentation.  Many do, and
this PEP is for them.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From pyth at devel.trillke.net  Sun Apr 21 14:31:24 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 20:31:24 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <0evu9a.621.ln@127.0.0.1>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <0evu9a.621.ln@127.0.0.1>
Message-ID: <20020421183124.GH10574@devel.trillke>

On Sun, Apr 21, 2002 at 02:12:16PM -0400, Carl Banks wrote:
> Allan Crooks wrote:
> > Hi,
> > 
> > I was wondering what the best approach is on providing libraries for
> > varying versions of Python?
> > (...)

> # Import stuff for version 2.1
> if sys.version[:2] >= (2,1):
>     from mymodule2_1 import *
> 
> # Import for version 2.2
> if sys.version[:2] >= (2,2):
>     from mymodule2_2 import *

this solves part of the 'code duplication' problem. Might get a bit tricky
if you just want to define a few additional methods for some classes
(for example 'generator versions' of methods otherwise returning lists).

Ah, and by the way how is

	sys.version[:2] >= (x,y) 

supposed to work? On my python 2.2 

	sys.version[:2] == '2.'   is True

am i missing something here?

	holger




From tdelaney at avaya.com  Tue Apr 30 20:19:44 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 1 May 2002 10:19:44 +1000 
Subject: Comment on draft PEP for deprecating six builtins
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2B8@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
> 
> Roman Neuhauser wrote:
>         ...
> >     Given the high tongue-in-cheek ration of this list, I'm 
> now really
> >     confused as to whether Fredrik Lundh is a bot, or if 
> you're "rocking
> >     me", as the Czech saying has it.
> 
> Of course /F is a bot.  So am I, and so is Tim.  So far, 
> we're the only
> contributors to c.l.py to have been _found out_ to be bots.

Note that the martellibot is referring to the timbot, not to me, who is a
real, live person as far as any of you know.

Tim Delaney




From jon+usenet at unequivocal.co.uk  Sat Apr  6 22:08:44 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sun, 07 Apr 2002 03:08:44 -0000
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>
Message-ID: <slrnaave1r.13h.jon+usenet@snowy.squish.net>

In article <40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk>,
  Alastair Nicol wrote:
> 1) Can only hosts which can see the interface your program is bound to 
> communicate with your program. In the "normal" case if you bind to 
> 127.0.0.1 for example ONLY your machine can see that interface. Other 
> machines cannot connect to it. Is this a safe / secure assumtion? (not just 
> for localhost)

No. Other machines on your local net can talk to your localhost by
sending ethernet packets with your machine's MAC address and 127.0.0.1
as the IP address.


From aahz at pythoncraft.com  Sat Apr  6 18:16:00 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 18:16:00 -0500
Subject: Pattern for permutations?
References: <mailman.1018131433.4336.python-list@python.org>
Message-ID: <a8nvjg$9jb$1@panix1.panix.com>

In article <mailman.1018131433.4336.python-list at python.org>,
Tim Churches  <tchur at optushome.com.au> wrote:
>
>I want to define a function which takes as its argument a
>list which contains one or more sequences, and then does stuff
>with every permutation of the elements of those sequences. For
>example, a function which works for a list containing three
>sequences:
>  [...]
>Is there a better way?

Recursion.  In Python 2.2, implemented with generators.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From lac at strakt.com  Fri Apr  5 01:40:24 2002
From: lac at strakt.com (Laura Creighton)
Date: Fri, 05 Apr 2002 08:40:24 +0200
Subject: Selecting cipher suites with socket.ssl 
In-Reply-To: Your message of "04 Apr 2002 18:30:10 -0800."
             <84cff98e.0204041830.7c0d0171@posting.google.com> 
References: <84cff98e.0204041830.7c0d0171@posting.google.com> 
Message-ID: <200204050640.g356eOrq032654@ratthing-b246.strakt.com>

>However, I'd like to know how I can specify which SSL cipher suites I
>allow i.e. the cipher suites that will be sent to the server in the
>cipher_suites field of the SSL ClientHello message.

You can't.  It's broken.  You need a different SSL implementation.
I recommend PyOpenSSL - http://pyopenssl.sourceforge.net

Laura Creighton




From max at alcyone.com  Sat Apr  6 14:53:51 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 06 Apr 2002 11:53:51 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se>
Message-ID: <3CAF524F.63E46C81@alcyone.com>

Jacob Hallen wrote:

> Python is a language that is weakly typed, ...

You mean dynamically typed, not weakly typed.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From djc at object-craft.com.au  Thu Apr 18 04:46:24 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 18 Apr 2002 18:46:24 +1000
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de> <m3g01um27o.fsf@ferret.object-craft.com.au> <m3ads2dkpy.fsf@mira.informatik.hu-berlin.de> <m38z7lrfh7.fsf@ferret.object-craft.com.au> <m3u1q9y0d6.fsf@mira.informatik.hu-berlin.de>
Message-ID: <m3n0w15r8f.fsf@ferret.object-craft.com.au>

>>>>> "Martin" == Martin v Loewis <martin at v.loewis.de> writes:

Martin> Dave Cole <djc at object-craft.com.au> writes:
>> I think that it is preferable to keep the exception and propagate
>> it back from the callback through the API to the thread which
>> called the API.  I am not sure why discarding the exception would
>> be a good idea.

Martin> That assumes that the call originated from Python. If some
Martin> library has created the thread (e.g. in response to an
Martin> incoming network connection which was accepted by the
Martin> library), there won't be anybody to propagate the exception
Martin> to.

The assumption is not an assumption in this case because the original
question was about this very situation.

Marcus> The thread is started like this:
Marcus> 
Marcus> T=threading.Thread(group=None,target=ExtModule.CFunction,name=None,
Marcus>                            args=(a,b,c,self.callback))
Marcus>         T.start()


>> IMO, from the point of view of the extension module user it should
>> not be visible that there was some non-Python code between the
>> Python code where the exception was raised and the Python code
>> which called the module API.

Martin> Again, assuming there is Python code on both sides, and also
Martin> assuming it is possible to pass something through from one end
Martin> to the other. *If* that is possible, you can just as well pass
Martin> the thread state through. If you cannot pass the thread-state
Martin> through, due to the design of the library you are wrapping,
Martin> tough luck: chances are low that you can pass the exception
Martin> back.

Agreed, but (I thought) we were talking about the situation where
Python created the threads.

- Dave

-- 
http://www.object-craft.com.au


From shalehperry at attbi.com  Thu Apr 25 12:28:56 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 25 Apr 2002 09:28:56 -0700 (PDT)
Subject: Dictionary that adds items to itself
In-Reply-To: <4620daca.0204250809.563a8c0b@posting.google.com>
Message-ID: <XFMail.20020425092856.shalehperry@attbi.com>

On 25-Apr-2002 gyromagnetic wrote:
> Hi,
> I would appreciate some advice on the following. I would like to
> create a dictionary-like object with the same properties as a normal
> dictionary except with regard to adding items. When items are added, I
> would like them to be processed by a function that may generate
> additional items. I would like these additional items to be added to
> the dictionary, have the function applied to them, have the resulting
> items added to the dictionary, etc.
> 
> Can this be done by subclassing UserDict? Is there another way?
> 

UserDict and a couple of __foo__ methods should get it going pretty easily. 
One thing to watch out for is your function adding items forever.




From starx at pacbell.net  Wed Apr 24 06:16:06 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 10:16:06 GMT
Subject: Accessing class-level variables
References: <mailman.1019621385.21532.python-list@python.org> <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4> <mailman.1019624445.12502.python-list@python.org> <Xns91FA105A11462RASXnewsDFE1@130.133.1.4> <slrnaccvoj.eo.gerhard@developer2.opus-gmbh.net>
Message-ID: <Xns91FA217759CC6RASXnewsDFE1@130.133.1.4>

Gerhard =?iso-8859-15?Q?H=E4ring?= || Wed 24 Apr 2002 02:48:12a:

> In article <Xns91FA105A11462RASXnewsDFE1 at 130.133.1.4>, Philip
> Swartzleonard wrote: 
>> Er, i think you're trying too hard. +) Now that i realized what you 
>> really want, I just tried this...:
>> 
>>>>> class A:
>>      d=5
>> 
>>>>> class B(A):
>>      pass
>> 
>>>>> B.d
>> 5
>>>>> class B(A):
>>      def p(self):
>>           B.d += 1
>>           print B.d
>>           print A.d
>> 
>>>>> B().p()
>> 6
>> 5
>>>>> B().p()
>> 7
>> 5
> 
> Heh. Thanks for the example code. But my needs are weirder. I really
> do want to access the class variables of a *derived* class from the
> *superclass* 8-) These are available for an instance method (it
> doesn't matter where the instance method is defined, be it in the
> superclass or in the derived class), by using self.__class__. One more
> reason why Python's explicit 'self' is useful. 

Ah, i guess I missed something. Anyway, for what you want to do, I don't 
think there is another way. I'm just riding off of the global variables 
declared in my module, the class names. You have no idea what the 
relevant class name is or where it will be, so some form of 
introspection is the only way to get whay you need.

> I also can't explain the behaviour your code shows using the Language
> Specification. http://www.python.org/doc/current/ref/class.html
> doesn't say how inheritance affects class variables. Anybody can
> explain me how this works? 

Hm. Well, it looks like the derived class just recives an extra binding 
of the inherited class. Here's a few tests on the subject:

>>> class C:
	a = [[1]]
>>> class D(C):
	pass
>>> C.a
[[1]]
>>> C.a[0] is D.a[0]
1
>>> C.a[0].append(2)
>>> D.a
[[1, 2]]
>>> C.a.append(3)
>>> C.a
[[1, 2], 3]
>>> D.a
[[1, 2], 3]

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From umar at cag.lcs.mit.edu  Fri Apr  5 17:06:33 2002
From: umar at cag.lcs.mit.edu (Umar Saif)
Date: Fri, 05 Apr 2002 17:06:33 -0500
Subject: icotl and Pointers
Message-ID: <200204052206.g35M6X304288@catfish.lcs.mit.edu>

Hi,
	I am trying to write a piece of python code to control a sound device (with 
an oss driver).
 The ioctls for the sound driver take in a pointer to an integer for the 
argument -- I was wondering whether it is possible to do this in python ? Or 
writing a native library is the only way around this ?

Thanks,
-U





From rxg218 at psu.edu  Wed Apr 10 19:24:56 2002
From: rxg218 at psu.edu (Rajarshi Guha)
Date: Wed, 10 Apr 2002 19:24:56 -0400
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org>
Message-ID: <a92hl1$1v4c@r02n01.cac.psu.edu>

On Wednesday 10 April 2002 18:55 in comp.lang.python Cliff Wells wrote:

> On Wed, 10 Apr 2002 18:42:34 -0400
> Rajarshi Guha wrote:
> 
>> Hi,
>>   I have a list like this:  l = [1,'s', 2,'t', 3,'d']
>> 
>> I'd like to loop through the list but within the loop I want
>> the current element and the element just after it:
>> I tried this

> 
> It would be better to organize your list into tuples before you start:
> 
> l = [(1,'s'), (2,'t'), (3,'d')]
> for a, b in l:
>     print a, b
> 
> However, if that isn't feasible, you could do:
> 
> l = [1,'s', 2,'t', 3,'d']
> for i in range(0, len(l), 2):
>     print l[i], l[i+1]
> 
> This is sort of fragile: if there aren't an even number of elements the
> loop will fail.


I got this code to work:

liter = iter(l)
while(1):
  try:
     print liter.next(), liter.next()
  except StopIteration:
     break

Works fine - is it an efficient way to handle this situation?
(I could arrange the data into tuples if necessary)

-- 
Rajarshi Guha
rajarshi at presidency.com


From jeff at ccvcorp.com  Wed Apr  3 14:10:56 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 3 Apr 2002 11:10:56 -0800
Subject: How to wait only so long for a server to respond?
References: <nwtq8.11707$ml2.924462@newsread1.prod.itd.earthlink.net>
Message-ID: <MPG.1714f39bc39de818989692@news.nwlink.com>

In article 
<nwtq8.11707$ml2.924462 at newsread1.prod.itd.earthlink.net>, 
ralph at penguin.com says...
> Pythoners:
> 
> I have some code that requests a response from a server on the Internet, 
> but I only want to wait so long for the server to respond.  How can I tell 
> python to wait x seconds/minutes, then move on to the next section of code 
> if the server has not responded?

Look for timeoutsocket.py -- I don't know the proper address to 
look at, but Google should find it.

(As I understand it, it's a drop-in replacement for the socket 
module that allows an optional timeout value to be specified.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From nestle at thepit.shacknet.nu  Tue Apr  9 17:12:01 2002
From: nestle at thepit.shacknet.nu (Matt Monacelli)
Date: Tue, 9 Apr 2002 17:12:01 -0400
Subject: using NumPy for histogram plot
Message-ID: <3cb358b4@news.isc.rit.edu>

I am trying to create a histogram plot of some data and am having a tough
time with the Numeric documentation on how to do this.  I am running on a
Windows machine (not Unix).  I am also trying to display this plot within a
wxPython GUI app.  THanks for any help...

Matt




From bergeston at yahoo.fr  Fri Apr 12 06:38:49 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Fri, 12 Apr 2002 12:38:49 +0200
Subject: try/except/else flexibility...
References: <slrnabcon3.g5r.mlh@vier.idi.ntnu.no>
Message-ID: <a96df6$ml74$1@ID-135695.news.dfncis.de>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnabcon3.g5r.mlh at vier.idi.ntnu.no...
> Just a thought: Given the prevalence of try/except/else in the "don't
> look before you leap" way of doing things, could it be useful to make
> the statement more flexible? For instance, instead of writing
>
>   try: foo + ''
>   except: pass
>   else: bar()
>
> (assuming I'm sloppy enough not to specify TypeError on the second
> line) one could simply write:
>
>   try: foo + ''
>   else: bar()
>
> and if I want something done only if it causes no exceptions, for
> instance pop off an element of a list unless it is empty, I could just
> do:
>
>   try: somelist.pop()
>
Just to remember you: ;-)))
>>> Explicit is better than implicit.
>>> Special cases aren't special enough to break the rules.
and if you want something more explicit:
>>> Errors should never pass silently.
>>> Unless explicitly silenced.
And then you could always use something like that (and let the console
remind you how you like to take risks)
>>> def tryAndForget(f, *args, **kw):
...  try:
...   return apply(f, args, kw)
...  except :
...   print('I like to live dangerously')


> Is this heresy? I see the harm in not specifying exactly which
> exceptions you're expecting, of course -- are there other arguments
> against this? (I'm sure that might be argument enough... Oh, well...)

More seriously, I think that:
try:
    ...
except:
    pass
means something like 'if there is something wrong that I didn't foresee,
please, please don't tell me' which remind me the joke of the guy falling
from a building and saying during the drop 'until now, it's all right, ...
until now, it's all right, ... until now, it's all right ...' (sorry for my
english, it's implemented as you like: every exception is silently absorbed
;-)

The perspective is completely different if you log evething that you didn't
foresee. A kinda tryAndLog function makes sense in some development phases
IMHO.

Have a nice WE

B.

>
> --
> Magnus Lie Hetland                                  The Anygui Project
> http://hetland.org                                  http://anygui.org





From tim.one at comcast.net  Sun Apr 14 16:51:23 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 16:51:23 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <Xns91F0276A4EF3JamesLLugojcom@216.148.53.85>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEBNOOAA.tim.one@comcast.net>

[James Logajan]
> ...
> Many firms I've worked with get support contracts to only get bug fixes
> _and_ priority response when problems happen.

As before, you can buy that kind of support for Python now.  What you can't
do now is get it for free.  I'm not sure there's a real paying market for
it, either, as people who uncover a critical bug now *generally* get a fix
for it within a day or two for free.  If I were running a business, the
amount I'd be willing to pay to put an upper bound on the outlier cases is
approximately nothing -- provided I understood how open source works.

> (E.g. Database systems like Oracle.  Applications are designed assuming
> the existing feature set and behavior and it is nontrivial to rewrite a
> mission critical app to take advantage of some new feature.)

Sure, it can be.

> Whether the support contract includes delivery of new releases was
> rarely a consideration.

Python 2.1.1 and 2.1.2 and 2.1.3 and 2.2.1 were all "pure bugfix" releases:
they added no features over their base versions, and strived like hell not
to change any visible behavior (although that can't be achieved 100% --
bugfixes can break things too, and it's almost always the case that
*someone* mistook a bug for a feature).  It's economically impossible for us
to offer custom bugfixes that aren't released to everyone.  So bugfix
releases are a subset of "new releases" for Python.  If the developers
weren't predominantly Unixoids, I suppose we could make the commercial world
more at ease by calling them "service packs" instead <0.5 wink>.

> We never bought tools because of their future potential, we bought them
> because we thought they were relatively stable and would solve our
> problems then and there.

I'm skeptical of that for a programming language.  You're really going to
base your *infrastructure* on tools that won't grow to encompass technical
change?  Fine, stick with 1.5.2 and try to build internationalized apps
without Unicode; ignore SSL; say "who needs it?" to XML; "64 bit servers?
no thank you", "a 2GB file should be plenty for anyone"; etc etc.  It sounds
like an efficient strategy, but for a company that intends to go out of
business in protest when the market dares to change.

> ...
> Well, you know business requirements: the product should fit in a shirt
> pocket, have a 23 inch screen, run on a 9V battery for 3 years, and be
> capable of towing your car. ;-) This is why I now always demand a
> _prioritized_ list of requirements from marketing (and they can't
> all have the same priority!)

I bet the priorities shift radically with every new sales prospect anyway --
the commercial world isn't unknown to me <wink>.





From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 16:57:44 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 22:57:44 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9i38s$d3i$1@nets3.rz.RWTH-Aachen.DE>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9i327$cvi$1 at nets3.rz.RWTH-Aachen.DE...
>..
> 1 == 2 and false or (10 + 3 * 2) == 4 or internal_harmlessFunction() ==
true
>..

(Please replace false with 0 and true with 1... I guess that's how to do it
in Python? Just reading through a help text here.)




From robin at jessikat.fsnet.co.uk  Fri Apr 19 08:25:57 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 19 Apr 2002 13:25:57 +0100
Subject: use of calldll
References: <84323230.0204190336.4dc51098@posting.google.com>
Message-ID: <hkopeKAVzAw8EwvB@jessikat.demon.co.uk>

In article <84323230.0204190336.4dc51098 at posting.google.com>, Piero <picodello at yahoo.it> writes

Variant is defined thusly according to my docs which are probably slightly out of date.
Just ensure that the union bit is long enough for the longest of the things in it.
I think that's 8 bytes, but others may know more. When assigning ensure the vt is
set to the corresponding value ie 0 for lVal, 1 for bVal etc etc.

When reading you need to check vt to see how to interpret the memory
at the address of the union.

struct tagVARIANT {
    VARTYPE vt;
    WORD wReserved1;
    WORD wReserved2;
    WORD wReserved3;
    union {
    //  C++ Type      Union Name   Type Tag                Basic Type
    //  --------      ----------   --------                ----------
        long          lVal;        // VT_I4                ByVal Long
        unsigned char bVal;        // VT_UI1               ByVal Byte
        short         iVal;        // VT_I2                ByVal Integer
        float         fltVal;      // VT_R4                ByVal Single
        double        dblVal;      // VT_R8                ByVal Double
        VARIANT_BOOL  boolVal;     // VT_BOOL              ByVal Boolean
        SCODE         scode;       // VT_ERROR
        CY            cyVal;       // VT_CY                ByVal Currency 
        DATE          date;        // VT_DATE              ByVal Date
        BSTR          bstrVal;     // VT_BSTR              ByVal String
        IUnknown      *punkVal;    // VT_UNKNOWN 
        IDispatch     *pdispVal;   // VT_DISPATCH          ByVal Object
        SAFEARRAY     *parray;     // VT_ARRAY|*           ByVal array
        // A bunch of other types that don't matter here...
        VARIANT       *pvarVal;    // VT_BYREF|VT_VARIANT  ByRef Variant
        void          * byref;     // Generic ByRef        
    };
};


......
>Something can help on using VARIANT type in Python and CALLDLL
>
>Thanks
>
>Piero Dell'OSte

-- 
Robin Becker


From graeme.matthew at unite.com.au  Tue Apr 16 07:10:33 2002
From: graeme.matthew at unite.com.au (Graeme Matthew)
Date: Tue, 16 Apr 2002 21:10:33 +1000
Subject: Python Classes and IDLE
Message-ID: <3cbc06b0@news.comindico.com.au>

Sorry to bug u all, but im at the point of giving up and going back to Perl
(much to my disgust as python is so clean !). I have been trying to get
classes going, even with two bloody books, one by Wesley Chun and its just a
nighmare, sometimes I have to close down IDLE and now I find out that it
must be reading the .pyc file thats why when I make changes they are not
taken into account. I eitehr have to delete the .pyc file or close down
IDLE, if this is the truth then this will take ages to do a project, anyone
got some advice ? thanks a mil




From goodger at users.sourceforge.net  Mon Apr 15 21:26:17 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Tue, 16 Apr 2002 01:26:17 GMT
Subject: Regular Expression Pattern Matching "State" Object
In-Reply-To: <mailman.1018895913.16705.python-list@python.org>
References: <mailman.1018895913.16705.python-list@python.org>
Message-ID: <B8E0F639.218FE%goodger@users.sourceforge.net>

Perhaps not exactly what you're looking for, but it may be a good start:

    http://docstring.sourceforge.net/dps/statemachine.py

"A finite state machine specialized for regular-expression-based text
filters"

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From opengeometry at yahoo.ca  Fri Apr 26 14:53:28 2002
From: opengeometry at yahoo.ca (William Park)
Date: Fri, 26 Apr 2002 14:53:28 -0400
Subject: Python's Lisp heritage
In-Reply-To: <001101c1ed52$47156450$445d4540@Dell2>; from brian@sweetapp.com on Fri, Apr 26, 2002 at 11:43:37AM -0700
References: <tyf1yd25zgs.fsf@pcitapi22.cern.ch> <001101c1ed52$47156450$445d4540@Dell2>
Message-ID: <20020426145328.A31190@node0.opengeometry.ca>

On Fri, Apr 26, 2002 at 11:43:37AM -0700, Brian Quinlan wrote:
> Actually, in your own example, the lisp example requires more typing.
> You disingenuously inserted unnecessary spaces in the Python example.
> Allow me to fix that:
> 
> Lisp:   (+ 1 2 3 4 5 6 7 8 9 10)
> Python: 1+2+3+4+5+6+7+8+9+10

Behold...
	  seq -s+ 1 10
:-)

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-cpu Cluster, Hosting, NAS, Linux, LaTeX, Python, vim, mutt, tin




From geoff at gerrietts.net  Tue Apr  2 14:25:02 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 2 Apr 2002 11:25:02 -0800
Subject: compiling Python with readline
In-Reply-To: <777AB66E-466E-11D6-A42D-0050E4857868@hhbrown.com>
References: <777AB66E-466E-11D6-A42D-0050E4857868@hhbrown.com>
Message-ID: <20020402192502.GA13827@isis.gerrietts.net>

Quoting Erik Price (pricee at hhbrown.com):
> Hello,
> 
> Sorry to trouble if this is not the right list -- I would like to 
> re-compile my Python 2.2 source (on Mac OS X 10.1.3) to support readline 
> (or whatever it takes to get emacs-style key bindings to work in the 
> interpreter, like I get in bash).
> 
> I was reading the README file in the source tree, and discovered that 
> there is no longer a --with-readline ./configure parameter.  The 
> instructions say "This option is no longer supported.  To use GNU 
> readline, enable module 'readline' in the Modules/Setup file."  I don't 
> seem to have the Modules/Setup file anywhere in my source tree.

The Modules/Setup file is automatically created from (I believe)
Modules/Setup.dist, but will not be overwritten if it exists.

In other words, if you need to use a nondefault Setup file, cd into
Modules, cp Setup.dist to Setup, and edit Setup.

When I built 2.2, the configure process automatically discovered
the existence of readline and built the interpreter properly. If you
haven't tried that, you might consider trying it. If it doesn't work,
you may have some configuration problems with the readline library (or
you may not, but it's possible).

Best of luck (if you need more help, feel free to ask).

Thanks,
--G.

-- 
Geoff Gerrietts             "That's it! I've had it with your sassy mouth!
<geoff at gerrietts net>     I didn't want to do this! (Well, actually, 
http://www.gerrietts.net/    I did....)"  -- Mojo Jojo, "Bubblevicious"



From dmmcf at uiuc.edu  Mon Apr 15 15:33:45 2002
From: dmmcf at uiuc.edu (D. Michael McFarland)
Date: Mon, 15 Apr 2002 14:33:45 -0500
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <3CBB29B9.95DED100@bgb-consulting.com>
Message-ID: <86ads43gfa.fsf@haggis.aae.uiuc.edu>

Don Garrett <garrett at bgb-consulting.com> writes:

>   I have a couple of Pentium 450s that can help. RedHat 7.x with Python 2.2
> but no additional modules yet. I can install them, but probably take me a day
> or two to get to it. So... 4-6 problems, I guess?

You, sir, have make my day.  Thank you.  I'll be in touch by private
mail.

Michael



From phr-n2002a at nightsong.com  Sun Apr  7 00:33:50 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 06 Apr 2002 21:33:50 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018152976.30380.python-list@python.org>
Message-ID: <7xit74vzr5.fsf@ruckus.brouhaha.com>

Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:
> > Stable, mature software should be able to run for that long without
> > needing to be upgraded.  I find just can't expect that from Python.
> 
> You appear to be saying that Python incorporates expiry logic, which
> causes a particular version to cease functioning correctly after some
> specified time.

Non sequtur.

> This of course is complete BS.  Likewise your whole argument is complete
> BS in the face of moves by a certain company towards subscription software
> with mandated upgrade schedules.

Another non sequitur.

> There are OS vendors who choose to imposing grossly out-of-date releases
> of 3rd party software on customers for their own selfish reasons.

Are you talking about Red Hat?  If a Python release as recent as 1.5.2
is already "grossly out-of-date", that confirms what I'm saying about
releases happening too fast.  The idea of stability and maturity is that
releases don't become grossly out of date anywhere near that quickly.

> You've said your piece to the point of being a broken record, and been
> informed that your POV is not accepted.  

That's YOUR opinion.  

> Please swallow future such utterances so that the positive
> contributions you've made to this community don't get lumped into
> irrelevence along with your whinging.

Stuff it.


From bgrotanLOVELYSPAM at itea.ntnu.no  Thu Apr 18 11:09:04 2002
From: bgrotanLOVELYSPAM at itea.ntnu.no (=?iso-8859-1?Q?Bj=F8rn?= Ove =?iso-8859-1?Q?Gr=F8tan?=)
Date: Thu, 18 Apr 2002 17:09:04 +0200
Subject: chopping csv-files (newbie)
References: <3CBE768C.844C90FF@itea.ntnu.no> <m3it6p5r17.fsf@ferret.object-craft.com.au>
Message-ID: <3CBEE190.52483E45@itea.ntnu.no>

Dave Cole wrote:
> 
> > I have a csv-file.. delimited by ;
> > example-line from the csv-file
> > "90383";"ooooxWOW   90383;";"Some discription";
> > "99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";
> >
> > In this file there is 3 fields extracted from a database.
> > The second field, can vary som 1 to 4 fields separated by ; but not with
> > "
> >
> > What I need to do is to build an array after opening a given file
> > with the information from the second and third fields.
> >
> > E.g:
> >
> > 90383 Some discription
> > 99875 New line-descriptor, Some discription
> >
> > You see where I'm going at?
> >
> > the second record in this file has a reference to the first record in
> > its second field.
> 
> You could use my CSV parser:
I'm already using it.. not quite confident on how to use it though
 
> >>> import csv
	     ^^^

> >>> p = csv.parser()
> >>> p.field_sep = ';'
> >>> p.parse('"90383";"ooooxWOW   90383;";"Some discription";')
> ['90383', 'ooooxWOW   90383;', 'Some discription', '']
> >>> p.parse('"99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";')
> ['99875', 'ooooxSomede90383;0000xNewli99875;', 'New line-descriptor', '']

in python.. how do I build a hash-table to get a list like this:

99875	New line-descriptor, Some discription

and so on and so forth...
I need some guidelines on how to get those numbers from the second
field.
The csv-file is sorted so I know that records with 4 id-records comes
last and 1 id-record
comes first.

I thought of making a hash-table with first the 1st-level idnr +
common-name is listed
and when the 2nd level takes the common-name from the idnr already
listet/stored in my hash-table.
then the 3rd and 4th level. How should I build such a hash-table? 

Best regards

Bj?rn Ove


From theoryboy at my-deja.com  Sat Apr 20 10:34:17 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 20 Apr 2002 07:34:17 -0700
Subject: TKinter Button widget
Message-ID: <ced73313.0204200634.251529e2@posting.google.com>

When I place some buttons on a frame I don't seem to be allowed to use
arguments in the function call attached to the button. If I do use
arguments, or even just empty brackets, the function is called when
the button is placed, not when the button is clicked on.

Below is some simple code which demonstrates this. Is there a way
around this problem?

Peter


#!/usr/bin/python

from Tkinter import *

def printNumber(number):
	print number
	
win = Frame()
win.pack()
Button(win, text='print 3', command=printNumber(3)).pack()
Button(win, text='print 5', command=printNumber(5)).pack()
 
win.mainloop()


From ffjhenon at club-internet.fr  Sat Apr 13 05:09:10 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Sat, 13 Apr 2002 11:09:10 +0200
Subject: libtk.so
Message-ID: <3CB7F5B6.12C8C5AA@club-internet.fr>

I have downloaded the latest version of Pyhon (2.2) for Linux ( Python
2.2 + tools + Tkinter 2.2 +devel)

But once I start an application that uses Tkinter, it asks me for
libtk.so.0 and libtcl.so.0.

I searched on my hard drive and could not find them. Where or how can I
get them.

Thanks




From robin at jessikat.fsnet.co.uk  Sat Apr 13 16:58:35 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 21:58:35 +0100
Subject: Python 2.2 unwise.exe
References: <ZkjnuIABLAu8EwYC@jessikat.fsnet.co.uk>
 <mailman.1018727067.4980.python-list@python.org>
Message-ID: <6Nd$EXA7vJu8EwPE@jessikat.fsnet.co.uk>

In article <mailman.1018727067.4980.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
>[Robin Becker]
>> Anyone else had a problem un-installing 2.2 for win32? When I tried to
>> un-install 2.2 prior to installing 2.2.1 I got an error saying it had
>> already been un-installed.
>>
>> Tried the link in start menu and it said no unwise.exe was there.
>
>Yes, I've seen this happen several times, but only under Win9X, not Win2K.
>Which flavor of Windows are you using?   So far you're the only other one to
>report it.  I can't account for it:  the installer definitely creates
>UNWISE.EXE in the right place (the root of your Python tree), but sometimes
>"it goes missing" later.  Exactly when and why remains a mystery to me.  It
>may be related to unplanned <ahem> system crashes.
....
Installed under C:\Python; system is Win2k. I installed win32all-146
after Python-2.2, but didn't save backups. 

I have a couple of other things installed eg Numeric-20.3 and PIL, but
nothing  that came with an installer. I don't remember anything untoward
about the Python-2.2 install.

The unwise.exe was just not there for some reason.
-- 
Robin Becker


From phr-n2002a at nightsong.com  Sat Apr  6 18:13:54 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 06 Apr 2002 15:13:54 -0800
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net> <MPG.17197f04a3f463e8989680@news.planet.nl>
Message-ID: <7xvgb4to7h.fsf@ruckus.brouhaha.com>

Johannes Gijsbers <taradino at softhome.net> writes:
> That's not really a good idea, as math.sqrt() returns a float. I'd say:
> 
> >>> import math
> >>> def isInt(number):
> ...     return int(number) == number
> ...
> >>> isInt(math.sqrt(289))
> 1
> >>> isInt(math.sqrt(315))
> 0

isInt(math.sqrt(1e100)) doesn't work so well.


From ianb at colorstudy.com  Wed Apr 17 16:00:16 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 17 Apr 2002 15:00:16 -0500
Subject: preferred syntax for list extend?
In-Reply-To: <a9kjh1$qc5$1@slb1.atl.mindspring.net>
References: <3CBCC196.466476CE@uniserve.com> 
	<a9jaah$493b$1@as201.hinet.hr>
	<mailman.1019068235.12319.python-list@python.org> 
	<a9kjh1$qc5$1@slb1.atl.mindspring.net>
Message-ID: <1019073616.1260.162.camel@localhost>

On Wed, 2002-04-17 at 14:51, Andrew Dalke wrote:
> 
> Ian Bicking:
> >+= is different from .extend()... for instance:
> >
> >q1 = [1, 2, 3]
> >q2 = q1
> >q3 = [4, 5, 6]
> >
> >q1 += q3   # then...
> >q2 == [1, 2, 3]
> 
> Really?

No, not really, you're right.

But that's weird...

q1 += q3
is not equivalent to:
q1 = q1 + q3

  Ian






From huaiyu at gauss.almadan.ibm.com  Wed Apr 24 18:08:17 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 24 Apr 2002 22:08:17 +0000 (UTC)
Subject: profile formats
Message-ID: <slrnaceb6g.e59.huaiyu@gauss.almadan.ibm.com>

Are there facilities in the profile module that identify the contribution to
the cumtime of a function f by each function called from f?

For example, if I have functions fa, fb, fc, with cumtime ta, tb, tc, I want
a table like
        
fa: ta      fa: taa  fb: tab  fc: tac  ...
fb: tb      fc: tbc  fa: tba  ...
...

where tij is the time fi spent in calling fj.  

Preferably each row and column would be sorted and formated, but this is not
crucial - nested dictionaries are easy to process anyway.

Such information cannot be deduced from ti or tj alone.  I need this to
identify not only which functions used the most time, but where they are
most heavily used.


Huaiyu


From theller at python.net  Tue Apr 23 07:13:11 2002
From: theller at python.net (Thomas Heller)
Date: Tue, 23 Apr 2002 13:13:11 +0200
Subject: Linker error when doing MSVC debug builds
References: <3CC535A7.2070607@gmx.net>
Message-ID: <aa3fkj$7h6s3$1@ID-59885.news.dfncis.de>

"Markus Ewald" <Markus_Ewald at gmx.net> wrote in message news:3CC535A7.2070607 at gmx.net...
> No this is not a cry for help :)
> I've got a little comment about the MSVC build setup:
>
> First, note that it is absolutely legal for a win32 project to compile
> in debug mode and link release libraries (if the intentioned is not to
> debug the specific library).
> As an example, I think noone outside Microsoft has ever possessed debug
> builds of user32.lib or kernel32.lib, yet everyone is compiling his
> debug projects with those :)
>
>
> The windows setup program only contains python22.lib
> However, if you do a debug build of your project (and _DEBUG is defined)
> pyconfig.h tries to link python22_d.lib, resulting in a linker error.
>
So why don't you install the source distribution and compile it yourself?
This way you have all the debug libraries. Occasionally stepping through
Python's source may be very instructive or even help you better locate
bugs in your extension.

Thomas




From gerhard at bigfoot.de  Fri Apr  5 13:43:44 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 5 Apr 2002 18:43:44 GMT
Subject: CGI GET vs. POST
References: <a8k613$suseq$1@ID-59885.news.dfncis.de> <a8kfse$teqtj$1@ID-59885.news.dfncis.de>
Message-ID: <slrnaars0t.og.gerhard@developer2.opus-gmbh.net>

In article <a8kfse$teqtj$1 at ID-59885.news.dfncis.de>, Thomas Heller wrote:
> [replying to my own post]
> "Thomas Heller" <theller at python.net> wrote in message news:a8k613$suseq$1 at ID-59885.news.dfncis.de...
>> I'm trying to use the cgi module in a CGI script,
>> but encountered a problem.
>>
> [snip}
>>
>> If I use method=POST in the CGI-script, the
>> "keyword" is found as a field in FieldStorage,
>> but NOT if I use method=GET.
>> Is FieldStorage only usable with POST?
>>
> 
> Problem solved, although I don't understand why.
> Maybe someone can enlighten me?
> 
> The script was in a file pyhelp.cgi on the server,
> and I called it with 'http://...../pyhelp' when it did
> not work.
> Calling it with 'http://..../pyhelp.cgi' made it work.
 
Ask yourself: why should this have worked? Just because not providing the
extension for executables does work on Windows? ;-)

Gerhard


From Patrick.Bussi at space.alcatel.fr  Fri Apr 19 04:31:06 2002
From: Patrick.Bussi at space.alcatel.fr (Patrick.Bussi at space.alcatel.fr)
Date: Fri, 19 Apr 2002 10:31:06 +0200
Subject: Newbie question: strange Re behaviour -- Many thanks
Message-ID: <C1256BA0.002F2B62.00@vzmta01.netfr.alcatel.fr>



Terrific. It works fine.

I understand now that I had missed something essential about regex.

Thank you Eric for this tutorial. Good bye.

---
Patrick Bussi
patrick.bussi at space.alcatel.fr


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

---------------------- Envoy? par Patrick Bussi/ALCATEL-SPACE le 19/04/2002
10:26 ---------------------------


Eric Brunel <eric.brunel at pragmadev.com> on 15/04/2002 19:26:42

Pour :    python-list at python.org
cc :   (ccc : Patrick Bussi/ALCATEL-SPACE)
Objet :   Re: Newbie question: strange Re behaviour


-------------- next part --------------

Patrick.Bussi at space.alcatel.fr wrote:
>
> Could someone help me understand what's wrong with this very simple regex,
> whose initial purpose was to extract the value of an HTML field. For
> demonstration, I have oversimplified the code below:
>
> -----start-----
> [pat]$ python
> Python 2.0 (#1, Apr 11 2001, 19:18:08)
> [GCC 2.96 20000731 (Linux-Mandrake 8.0 2.96-0.48mdk)] on linux-i386
> Type "copyright", "credits" or "license" for more information.
>>>>
>>>> import re
>>>> s='[^value=].*[^>]'

That means "any character not in the set ('v', 'a', 'l', 'u', 'e', '='),
followed by any string, followed by a character that is not a '>'". I guess
that's not what you want... but it explains why it ignores the leading 'a'
(which *is* in the set ('v', 'a', 'l', 'u', 'e', '=')).

Try:
s = "value=(.*)>"
and see what group(1) returns after compiling the re...

HTH
--
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com

--
http://mail.python.org/mailman/listinfo/python-list

From alf at logilab.fr  Mon Apr 29 04:57:40 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Mon, 29 Apr 2002 08:57:40 +0000 (UTC)
Subject: What is NotImplemented ?
References: <slrnacinls.j5.alf@virgo.logilab.fr> <lkn0vq34fi.fsf@pc150.maths.bris.ac.uk>
Message-ID: <slrnacq2o4.em.alf@virgo.logilab.fr>

Dans l'article <lkn0vq34fi.fsf at pc150.maths.bris.ac.uk>, Michael Hudson a ?crit :
> Alexandre Fayolle <alf at logilab.fr> writes:
> 
>> A quick question, here. What is the built in NotImplemented object? 
>> And what can you do with it? I thought I could raise it as an 
>> exception, but this is not the case (I use NotImplementedError instead).
> 
> I think you can return it from implementations of, say, __add__ to
> mean roughly "I don't know what to do here".
<snip>
> Returning NotImplemented from the __add__ makes Python go off and call
> the __radd__ method.

Ahah! Thanks a lot. 

The relevant documentation is in the Python Reference Manual, ?3.2,
while I was looking for it in the library reference. 

Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From gerson.kurz at t-online.de  Wed Apr 24 10:45:34 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Wed, 24 Apr 2002 14:45:34 GMT
Subject: Using 2.2 stdlib modules in 2.1
Message-ID: <3cc6c34f.3125812@news.t-online.de>

If I'm correct, the hmac module was added in 2.2, but is not present
in 2.1 or below. Now, for the next version of shicks! I want to
include SMTP AUTH, and for this I need the hmac module. For 2.2 users
everything is fine, but what if the user has an "older" python
version? (and many, including my webhoster, still do). 

Can I just distribute hmac.py with my BSD-licensed open source app, or
is there a website to point to where you can download lib sources
(hmac.py works without changes in 2.1 and presumably in 2.0 also) so
that I can ask users to manually download it if they need it (which I
think would be a quite bad solution, but still...). 



From martin at v.loewis.de  Thu Apr 18 02:46:11 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Apr 2002 08:46:11 +0200
Subject: Reading from a binary file in a portable way
References: <3cbe6892_11@news.newsgroups.com>
Message-ID: <m3pu0xy05o.fsf@mira.informatik.hu-berlin.de>

jb <egrossde at yahoo.de> writes:

> I should like to read a (binary) file x.y in a buffer and then do things 
> like getting the 16 bit integer that is at the offset n, n+1 in the file 
> (maybe lowbyte, then higyhbyte.
> 
> Previously I did it only under MS-Windows where for example I could open the 
> file with the flag O_BINARY but this is no longer possible under Linux.

If you can accept using the open() builtin, I recommend to pass it the
"b" flag; that works on all platforms.

Regards,
Martin



From aahz at pythoncraft.com  Sat Apr 27 10:51:21 2002
From: aahz at pythoncraft.com (Aahz)
Date: 27 Apr 2002 10:51:21 -0400
Subject: Functional Python (was Re: Steve Holden wrote:)
References: <20020419205101.17293.2839.Mailman@mail.python.org> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org> <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com>
Message-ID: <aaedt9$ovl$1@panix1.panix.com>

In article <Wnvy8.51932$z55.27618 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>
>So, you tend towards the functional. Nothing wrong with that, but you
>should be aware that Guido's on record as regretting allowing the
>modest functional constructs the language already has, and presumably
>leaves them in to avoid the anathema of code breakage.

Well, no, that's not quite true.  Guido regrets lambda and the style of
functional programming that relies on lambda, but he certainly doesn't
tar all functional programming with that brush or he wouldn't have
allowed list comprehensions and zip().
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From aleax at aleax.it  Fri Apr  5 06:21:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 11:21:16 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it> <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20>
Message-ID: <MMfr8.62526$S52.2173482@news2.tin.it>

Fred Pacquier wrote:
        ...
> the water. But just yesterday I read a review of the just-released (in the
> US) Zaurus, and the last line mentions that Sharp is "thinking about
> introducing the new Zaurus line in Europe"... next year. So if you want

Funny -- the developer edition WAS already for sale in UK & Ireland... which
are part of Europe, of course.

> one soon you'll probably have to take your chances and order one overseas,

I was hoping to order one in UK or Ireland (no customs within the EU...).

> and make do with the Qwerty keyboard (probably less of a problem for you
> than for us French though :-)

I go out of my way to get US-standard keyboards wherever I possibly
can (and use 'US-international' to enter national characters when the OS
supports that -- Windows and Linux do).  I never grew used to Italian
layouts (nor French ones, Swedish ones, etc)...


Alex



From phr-n2002a at nightsong.com  Sun Apr  7 20:51:10 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 07 Apr 2002 17:51:10 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <a8pqal$ftm$1@panix1.panix.com> <7xpu1bh481.fsf@ruckus.brouhaha.com> <a8q0g1$qu5$1@panix1.panix.com>
Message-ID: <7xzo0fui69.fsf@ruckus.brouhaha.com>

aahz at pythoncraft.com (Aahz) writes:
> That said, seems to me that booleans are the stuff of a minor release,
> unless you claim that minor releases should never contain language
> changes.

IMO, minor releases shouldn't contain language changes unless there's
a real sense that substantial numbers of users are yearning for the
change and don't want to wait for the next major release to get it.
The best reason for that is if the change makes some programming task
substantially easier than it was before.

That might be true of generators but it's certainly not true of bools.
Even users who favor adding bools aren't likely to say there's urgency
to adding them.  Bools are just a cosmetic change and people can
survive continuing to use 1 and 0 for a few more minor releases.

Think of a co-worker who needs to ask you a lot of questions in the
course of his work.  Would you rather have him interrupt you by
calling you every 5 minutes whenever he thinks of a new question, or
have him save the non-urgent questions for a daily meeting where you
can cover them all in one shot?  I'd rather minimize the interruptions.


From gerhard at bigfoot.de  Sun Apr 28 15:56:08 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 28 Apr 2002 21:56:08 +0200
Subject: Python Webpages
In-Reply-To: <lEXy8.34518$n4.7236390@newsc.telia.net>
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net>
Message-ID: <20020428195607.GA19697@lilith.my-fqdn.de>

* Magnus <nixx at telia.com> [2002-04-28 18:59 +0000]:
> [WebWare]
> Another powerful (not yet final though) module is Middlekit which e.g. makes 
> it easy to deploy a servlet that acts with some glueware to e.g. MySQL.
 
The more I think about it, the less I get the point of object-relational
mappers like MiddleKit. Especially the ones that require you to rebuild
the database schema every time you change the object schema. AFAIC
MiddleKit falls into this category.

I think I'd prefer going either straight SQL with a dictionary mapper
(pyPgSQL's PgResultSet has such a thing built in [1]) or using an object
database like ZODB.

[1] in pyPgSQL, you can say, for example:
cursor.execute("select id, name from test")
print cursor.id, cursor.name
print cursor['id'], cursor['name']
print cursor[0], cursor[1]

All forms are equivalent, but only the last one is DB-API2 compliant.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From jwbaxter at spamcop.net  Mon Apr 29 19:55:12 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Mon, 29 Apr 2002 16:55:12 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <mailman.1020109254.2687.python-list@python.org>
Message-ID: <jwbaxter-66FB6D.16551229042002@corp.supernews.com>

In article <mailman.1020109254.2687.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> Note that Guido's comment on Python-Dev was:
> 
>     I strongly suggest sticking to the status quo.  There are more
>     important fish to fry.
> 
> Relax.
> 

This PEP seems to be in the category of "lots of good ideas for a new 
language...but this isn't a new language".  

I'm not convinced Python 3 is a "new enough" language for this PEP's 
ideas, although that would seem the obvious point for this sort of 
tinkering.

  --John


From faizal at umland.com.my  Mon Apr 15 21:19:59 2002
From: faizal at umland.com.my (Faizal)
Date: Tue, 16 Apr 2002 09:19:59 +0800
Subject: Need help
Message-ID: <01C1E527.E2A601A0@faizal>

Hi there...

I am still new(no exprience in programming)...&  need help from you guys...
Can u teach me how to learn programming (with example....)


Thank you 

Faizal 

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.325 / Virus Database: 182 - Release Date: 19/02/02
 






From F.Baart at sfk.nl  Thu Apr 11 10:51:43 2002
From: F.Baart at sfk.nl (F.Baart at sfk.nl)
Date: Thu, 11 Apr 2002 16:51:43 +0200
Subject: Strange, but interesting recursion problem in PythonWin.
Message-ID: <OFEE7F9AE7.905ECF89-ONC1256B98.004FFD45@Z-INDEX.NL>

I noticed something strange in the PythonWin

I have the following code:
def countdown(n):
 if n==0: print 'ready'
 else: countdown(n-1)

I have a maximumrecursionlimit of 1000.
When I run the countdown function with n=985 it works.
When I run the countdown function with n=987 it raises a maxrecursionlimit
error (which you would only expect at n>=999)
But when I run the countdown function with n=986 pythonwin crashes.

Remarkable is that when I run the countdown function in a dos shell it runs
up to n=999 and never crashes.

I was wondering why pythonwin crashes with n=986, does anybody have any
ideas?

Many thanks,

Fedor


PythonWin 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32.





From brian at sweetapp.com  Tue Apr 16 17:10:38 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 16 Apr 2002 14:10:38 -0700
Subject: Unicode name questions
In-Reply-To: <15548.35951.239060.214140@12-248-41-177.client.attbi.com>
Message-ID: <009901c1e58b$28a5bd90$445d4540@Dell2>

Hi Skip,

>     _entities['&Lambda;'] = u"\N{GREEK CAPITAL LETTER LAMBDA}"
>     _entities['&lambda;'] = u"\N{GREEK SMALL LETTER LAMBDA}"
>     _entities['&thetasym;'] = u"\N{GREEK SMALL LETTER THETA SYMBOL}"
> 
> Have I got the names wrong or are there some missing character names
in
> Python's Unicode support?

This page is worth book marking:
http://www.unicode.org/charts

Your spelling of the first two isn't quite right (at least according to
the Unicode guys). Remove the "B" from "LAMBDA".

And change "GREEK SMALL LETTER THETA SYMBOL" to "GREEK SMALL LETTER
THETA"

Cheers,
Brian






From martin_byrne2 at yahoo.co.uk  Tue Apr 23 05:08:08 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 23 Apr 2002 02:08:08 -0700
Subject: writing a body using email module
References: <508cb8ae.0204220123.493c35be@posting.google.com> <wWSw8.72199$T%5.62931@atlpnn01.usenetserver.com>
Message-ID: <508cb8ae.0204230108.577fbe35@posting.google.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message news:<wWSw8.72199$T%5.62931 at atlpnn01.usenetserver.com>...
> "B" <martin_byrne2 at yahoo.co.uk> wrote in message
> news:508cb8ae.0204220123.493c35be at posting.google.com...
> > Hi,
> >
> > I've recently been looking at the relativly new email module. I've
> > been able to create the email with to,from,subject and attachments but
> > I cant for the life of me figure out how to write the body of the
> > mail.
> >
> > Can anyone help me out with this?
> >
> 
> Looks (from the docs, I've never used this library) like you need to call
> the add_payload() method with your message string as the argument.
> 
> regards
>  Steve

Yeah tried that, well I used set_payload()(add_payload() is
deprecated), still got nothing in the body of the email. Everything
else works fine.


From dalke at dalkescientific.com  Sat Apr 20 04:57:49 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 02:57:49 -0600
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org>
Message-ID: <a9raeg$co7$1@slb1.atl.mindspring.net>

James J. Besemer:
> It seems that Python started out (as a Lisp derivitave) with much less
> of an OO emphasis.  I expect len(o) has precedent dating back to the
> original language.

and in another post
> Python claims a Lisp heritage, which I imagine is why and/or work
> the way they do.   But for decision making, original Lisp most often

As I understand it, Python's Lisp heritage is pretty indirect.
The most direct one I know of is docstrings, based on a similar
feature in elisp.

If there was lisp heritage, why isn't there anything like car/cdr?

To back up my statement, here's some quotes from Guido I found in
the Google usenet archives (no URLs since the Google links are too
long)
] Also, the Python docs (to the extent that I wrote them)
] *never* attribute *anything* to Scheme or Lisp.  Python comes from the
] Algol family of languages.

] (I'm a rabid Lisp-hater -- syntax-wise, not semantics-wise! -- and
] excessive parentheses in syntax annoy me.  Don't ever write return(i)
] or if(x==y): in your Python code! :-)

] Common Lisp and Scheme.
]
] These languages are close to Python in their dynamic semantics, but so
] different in their approach to syntax that a comparison becomes almost
] a religious argument: is Lisp's lack of syntax an advantage or a
] disadvantage?

] My conjecture: writing lambdas in Python is a bad habit, brought on
] from exposure to Lisp or Scheme.  Most Python code looks better after
] rewriting without lambdas.


There's also a relevant quote from Tim Peters:
] The problem Python set out to solve is that ABC was useless for
] real-life programming.  If Guido knows anything substantial about
] Lisp or Scheme, he's done an excellent job of hiding it for more
] than a decade.  I've seen him struggle with simple elisp, for God's
] sake <wink>.


                    Andrew
                    dalke at dalkescientific.com





From kkto at csis.hku.hk  Mon Apr 29 02:52:32 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 29 Apr 2002 14:52:32 +0800
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <7ielgzro7j.fsf@enark.csis.hku.hk>

>>>>> "Stephen" == Stephen J Turnbull <stephen at xemacs.org> writes:

    Stephen> The consumer protection argument is the OSI line (a special
    Stephen> case thereof), and rms don't have much truck with anything so
    Stephen> economical and relativist.

If you think it that way, look at the GNU manifesto, which can be found on
the web at http://www.gnu.org/gnu/manifesto.html.  It clearly tells you
otherwise.  Here are some quotes...

   All intellectual property rights are just licenses granted by society
   because it was thought, rightly or wrongly, that society as a whole would
   benefit by granting them.  But in any particular situation, we have to
   ask: are we really better off granting such license?  What kind of act
   are we licensing a person to do?

   The case of programs today is very different from that of books a hundred
   years ago.  The fact that the easiest way to copy a program is from one
   neighbor to another, the fact that a program has both source code and
   object code which are distinct, and the fact that a program is used
   rather than read and enjoyed, combine to create a situation in which a
   person who enforces a copyright is harming society as a whole both
   materially and spiritually; in which a person should not do so regardless
   of whether the law enables him to.

In particular, RMS acknowledge that at some time the government should do
something like copyright, giving a limited monopoly in order to give
economic incentive to work.  He just argued that software development is not
an area in which this is a correct approach---because, in order to do that,
the general public have to give up too much.

So I would characterize OSI and free software by their point of attack.  The
OSS group mainly attacks from the side of software producers, trying to
convince them that open-source produces better software for smaller cost.
Free software community attack it from the users' side, telling consumers to
understand that they are supposed to have the right to see whatever inside
the program, and that it is a very important right that they should not sign
anything to give it up.  The only catch here is that in the eyes of free
software advocates, developers are themselves users of their software and
software produced by others.  The battle line of free software is simply
longer than that of OSS.

Currently, the OSS point of attack seems more effective.  But I like the
free software line of attack much better.  Users must not be at the mercy of
producers.

Regards,
Isaac.



From ngps at vista.netmemetic.com  Wed Apr 17 10:38:15 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 17 Apr 2002 14:38:15 GMT
Subject: amkCrypto installation on win32
References: <1103_1019028251@news.free.fr>
Message-ID: <a9k1cn$pmk$1@violet.singnet.com.sg>

According to Fr?d?ric Laurent  <liste_nospam_ at opikanoba.org>:
> I want to use RSA encoding/decoding, so I downloaded amkCrypto 0.1.3.
> I use python 2.2 on windows XP.
> 
> So, I just want to generate a pblic/private key, crypt and decrypt a text
> buffer... but it sounds to be not so simple to do this !

Try M2Crypto:

ngps at vista:~$ python
Python 2.2 (#7, Mar 31 2002, 23:17:41) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> from M2Crypto import RSA
>>> r = RSA.gen_key(1024, 3) 
...........++++++
...................++++++
>>> r
<M2Crypto.RSA.RSA instance at 0x8197d2c>
>>> ptxt = 'plain text'
>>> import sha              
>>> x = sha.sha(ptxt).digest()
>>> ctxt = r.public_encrypt(x, RSA.pkcs1_oaep_padding)
>>> ptxt2 = r.private_decrypt(ctxt, RSA.pkcs1_oaep_padding)
>>> ptxt2 == x
1
>>>  


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



From kirschh at lionbioscience.com  Wed Apr 24 02:42:50 2002
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 24 Apr 2002 08:42:50 +0200
Subject: Exceptions in Extensions: weird problem
Message-ID: <yv21yd5fvh1.fsf@lionsp093.lion-ag.de>

I have a C library which runs callbacks into user-supplied code. I
made a python extension to interface this library in a way that such a
callback ends up again in the extension which then executes python
scripts.

The following can happen:

1) the script code called via the callback throws an exception,

2) the extension catches the exception and translates it into an error
code for the calling C function,

3) the C function recognizes the error code and returns with its own
error code to the calling interfacing python extension,

4) the interfacing python extension recognizes the error code. It can
even identify the fact that the cause of the error was a
python-exception.

QUESTION:

How can step (4) initialize/augment the error traceback with the
traceback information generated in step (1)?

  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh bei lionbioscience punkt com
       *** Please do not send me copies of your posts. ***


From eppstein at ics.uci.edu  Fri Apr  5 16:35:26 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 05 Apr 2002 13:35:26 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net> <3CAC149F.3F1BA77B@lemburg.com> <mailman.1017930740.20674.python-list@python.org> <slrnaapbh8.4s7.huaiyu@gauss.almadan.ibm.com> <Iwor8.8425$q63.5655@atlpnn01.usenetserver.com>
Message-ID: <eppstein-83EBA8.13352605042002@news.service.uci.edu>

In article <Iwor8.8425$q63.5655 at atlpnn01.usenetserver.com>,
 "Steve Holden" <sholden at holdenweb.com> wrote:

> > BTW, many people mistake boolean algebra with two-element boolean algebra,
> > which is only its most trivial example.  The set operations (union, joint,
> > complement) on all subsets of a given set is a more generic example of
> > boolean algebra.
> >
> Don't I remember from my degree studies that a Boolean algebra can be
> defined on any lattice?

Lattices have union and intersection but not necessarily complement.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From nobody at nobody.com  Sun Apr  7 01:44:25 2002
From: nobody at nobody.com (John)
Date: Sun, 7 Apr 2002 00:44:25 -0600
Subject: _Pg module and CGI
References: <vKRr8.36988$tg4.439776@vixen.cso.uiuc.edu>
Message-ID: <sRRr8.36991$tg4.440046@vixen.cso.uiuc.edu>

Thanks for email answer. In cygwin, it was "createuser"

">
> Hi, assuming to build an application getting SQL statement from html form
> and processing the query in postgresql, how can I add this "web user" to
> access the database?
>
> Error I get:
> Traceback (most recent call last): File "c:\inetpub\scripts\sd.cgi", line
> 15, in ? db=_pg.connect('template1','localhost') _pg.error: FATAL 1: user
> "IUSR_VALUED-VYL4TJPA" does not exist
>
> Remedy I tried:
> adduser IUSR_VALUED-VYL4TJPA in UNIX prompt
> (but there wasn't such command)
>
> I'd appreciate your help very much!
>
>
>




From akuchlin at crystal.mems-exchange.org  Wed Apr  3 10:09:04 2002
From: akuchlin at crystal.mems-exchange.org (A.M. Kuchling)
Date: 03 Apr 2002 15:09:04 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <ualr7hmlhkemab@news.supernews.com>
Message-ID: <slrnaam6of.ldi.akuchlin@crystal.mems-exchange.org>

In article <ualr7hmlhkemab at news.supernews.com>, John Roth wrote:
> In fact, at this point I'd like everything substantive to be defered
> to Python 3000, and get started on that right away. I've got a
> few suggestions...

It seems really unlikely that anyone will ever start with a completely
blank slate and re-implement Python from the ground up in C/C++.  Such
efforts have to rebuild absolutely everything, from the parser to the
basic data structures to all the modules, and there would be weeks and
months of work before there was anything usable again.  

Personally I'm -0 on the PEP.  I don't think it clutters things up
much, because you could get 90% of this PEP by doing
__builtin__.True=1 and __builtin__.False=0.  The major thing the PEP
adds is a new type, mostly so the repr() can return True/False instead
of 1/0, and that's nothing shocking.

On the other hand, I never write 'TRUE = 1' at the top of my modules
and am happy with just using 0/1, so this proposal wouldn't buy me
anything and wouldn't save me from any of my bugs.  <shrug>

--amk                                                             (www.amk.ca)
  "He says he doesn't know who he is or why he has come."
  "Well, I admire a man with an open mind."
    -- Ruther and Shardovan, in "Castrovalva"




From tim.one at comcast.net  Fri Apr 19 15:49:02 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 19 Apr 2002 15:49:02 -0400
Subject: How to make regexes faster? (Python v. OmniMark)
In-Reply-To: <8EA7006D40B56719.DC7AB9A2B4317096.3110ECEC5F22E1FD@lp.airnews.net>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEGPDAAA.tim.one@comcast.net>

[Donn Cave]
> Part of the problem is that when you write something like a "grep"
> in Python and in Perl, the Perl program will naturally be written
> like while ($line = <STDIN>) {...}, and the Python program will
> naturally be written like while 1: line = sys.stdin.readline() ...
> That pits a lot of function calls against what must be an inline
> operation.  I think I decided that "I/O", in this practical sense

[Cameron Laird]
> Worse (or at least "more"):  Perl goes out of its way to optimize
>   while ($line = <>) {...}
> 'Least, it has in the past; I haven't looked lately.

Sorry, you both lose <wink>.  In older versions of Python, and on most
platforms, the single biggest speed difference in line-at-a-time input was
due to that Python's I/O is thread-safe but Perl's is not.  Call the
platform getch(), and it doesn't just grab the next char from the buffer, it
runs all over creation slapping locks on internal platform I/O structures
for the duration.  This locking overhead utterly swamps the time needed just
to get the next char.  Perl peeks and pokes the platform stdio _iobufs
directly, without locking, and in the presence of threads then "whatever
happens, happens".

Recent versions of Python endure some remarkable platform-specific pain to
reduce the platform locking overhead while remaining threadsafe.  The
new-in-2.2 idiom

    for line in fileobject:

also invokes the xreadlines "chunking" mechanism under the covers.






From ajs at ix.netcom.com  Tue Apr 30 21:46:18 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 30 Apr 2002 21:46:18 -0400
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
Message-ID: <000601c1f0b2$19eb9900$0334fea9@carol>

>What is the circle-ellipse problem

In my PyGeo implementation, circle is a sub-class
of plane.

And a circle represents in effect all geometric
objects projectively equivalent to a circle - the 
ellipse, hyperbola, and parabola - which are
drawn by projecting a circle from a point 
in space to a plane in space.

All of which follows concepts of a geometry that
can be considered a *root* geometry - generalized
projective geometry.

The programming point may be that only when
working from authentically root concepts 
can sub-classing be sound.

Art  




From aleax at aleax.it  Tue Apr 30 13:11:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 17:11:52 GMT
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <aamebj$c2olc$1@ID-59885.news.dfncis.de> <1bzz8.82059$vF6.2440100@news2.tin.it> <jCzz8.17994$xy4.10082@atlpnn01.usenetserver.com>
Message-ID: <sfAz8.82271$vF6.2450688@news2.tin.it>

Steve Holden wrote:
        ...
> What you say appears to mean is that iter(type, None), for example, should
> return an unbounded iterator of type. I'd be interested in what you have

type is not callable without arguments.  iter(XX, None) gives an unbounded 
iterator for any XX which is a type callable without arguments: each call
to the iterator's .next() yields [aka, each item in the iterator is] a
"default-initialized" instance of type XX (same as XX() returns).

> if-you-can't-call-a-callable-iterator-why-is-it-
>         -called-a-callable-iterator-ly y'rs  - steve

"Iterator that wraps a callable" might no doubt be a better name.  However:

Much like a burger flipper is not (necessarily) a burger, so is a callable 
iterator not (necessarily) a callable.  "Callable" here is used as a noun
(if any noun can be verbed, why can't adjectives be nouned...?).


Alex



From starx at pacbell.net  Thu Apr 11 01:33:28 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 11 Apr 2002 05:33:28 GMT
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <mailman.1018495402.12499.python-list@python.org>
Message-ID: <Xns91ECE5A88C0F8RASXnewsDFE1@130.133.1.4>

Geoff Gerrietts || Wed 10 Apr 2002 08:12:41p:

> Quoting Peter Hansen (peter at engcorp.com):
>> I'd like to take this opportunity to at least thank the developers,
>> bots included, and other contributors (in the most heartfelt way I
>> can without sending large cash contributions) for all the blood,
>> sweat and keypresses they've put into this phenomenal project.
> 
> I'll echo that.
> 
> I owe a lot more thanks than I've delivered, and probably a bunch more
> in the way of apologies than I've offered. Tim and Guido have both
> personally helped me out of tight spots when I've run into them,
> despite my semi-annual conservative whining.
> 
> I think that the community has a very high quality, despite the
> occassional disgruntled voice or fearful citizen, and that reflects on
> how easy the development team makes it to trust their judgement.
> 
> It's easy to get caught up in arguing the technical merits of this
> widget or that feature and lose sight of why we're all here: python is
> an excellent language with an amazingly clear syntax and so much
> expressive power you almost can't use it all.
> 
> It's a project that's worth contributing to, for sure.
> 

I must say I am impressed by a project with such a interactive crew out 
here in the public. Nothing fosters good feelings about a project's life 
as much as a good presense, i think.

Geoff again:
> We're leeches (so far) but at least we're very grateful leeches.

If programming lanugages were like cities and had sports teams...

In-rec-games-roguelike-devel-us-python-users-outnumber-ruby-users-2-to-
1-ly-yrz (two of us, one of them, afaict =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From bjc32 at hermes.cam.ac.uk  Tue Apr 16 09:16:41 2002
From: bjc32 at hermes.cam.ac.uk (Ben Carrington)
Date: 16 Apr 2002 06:16:41 -0700
Subject: ideas for the best way to do some math, please??
Message-ID: <693d5681.0204160516.541c0bfe@posting.google.com>

i am a complete beginner with python, please be gentle :I

i am trying to evaluate a reatively simple equation but i keep getting
this error:

overflowError: math range error

...pointing to math.log(p)
        ...
        #part3
	part3 = 1 - math.e**part2
	p = abs(part3)

	#part4
	part4i = math.log(p)
	part4ii = k*T*part4i
        ...

...how can i pass the value of "p" into math.log() and avoid this
problem? what is the best way to structure an equation, divide into
functions, use more brackets?? I have use L for long integers at the
top of the code and this corrected problems with large and very small
powers of numbers...but i don't know how to work the L into
math.log....hope this makes sense

thanks in advance

Ben


From ods at fep.ru  Mon Apr  8 11:24:33 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Mon, 8 Apr 2002 19:24:33 +0400 (MSD)
Subject: Runtime mix-ins with new-style classes
In-Reply-To: <20020408100657387-0400@braeburn.themorgue.org>
Message-ID: <Pine.LNX.4.44L.0204081917170.13708-100000@ods.pp.ru>

On 8 Apr 2002, Kevin Smith wrote:

KS> Before I started using the new-style classes (i.e. 'object'
KS> based), I was
KS> able to add mix-in base classes at runtime by modifying the
KS> __bases__
KS> attribute.  However, I recently tried to do the same thing
KS> with a new
KS> -style class and got an error telling me that __bases__ was
KS> a readonly
KS> attribute.  Is there a was to do runtime mix-ins with
KS> new-style classes?

1) You can automaticaly add mixins to all classes in module(s)
via new metaclasses approach (define global __metaclass__
variable before class definition is executed).

2) You can redefine variable that refer to class with new class
with mixin added:

module.MyClass = type(module.MyClass.__name__,
                      module.MyClass.__bases__+(MixIn,),
                      module.MyClass.__dict__)





From gtalvola at nameconnector.com  Wed Apr 24 10:47:29 2002
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Wed, 24 Apr 2002 10:47:29 -0400
Subject: Quick question about Webware
Message-ID: <BFFEB31BC23CCD44B8924E139E1D87E448EF57@MAILBOX.nameconnector.com>

Thomas Weholt wrote:
> When you create a servlet using WebWare is it loaded on each 
> request or is a
> instance kept in memory?
> I mean, what if the servlet uses some module or piece of code 
> that takes a
> while to load, how is this solved in WebWare?

Webware-specific questions should be directed to the webware-discuss mailing
list.  You can subscribe at
http://lists.sourceforge.net/lists/listinfo/webware-discuss .

But the answer to your question is that the module containing the servlet
code is imported only once, and a pool of servlet instances is maintained in
memory.  Instances are created and added to the pool as needed based on the
number of simultaneous requests for that servlet.  For example, if there are
currently 2 instances of a servlet in the pool, and 3 simultaneous requests
for that servlet come in, then one additional instance will be created and
added to the pool.

What does this all mean?  Basically, after the first access to a given
servlet, it runs fast :-)  Servlets are never unloaded from memory, but are
always reused.

If you have some time-consuming initialization steps, you can force them to
happen when the appserver first starts up by placing them into an
__init__.py in the same directory as your servlets.  This gets imported when
the appserver starts up.

- Geoff




From mal at lemburg.com  Thu Apr 25 07:01:00 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 25 Apr 2002 13:01:00 +0200
Subject: configuring mxCGIPython
References: <mailman.1019680127.28712.python-list@python.org> <23891c90.0204250140.7248c6ed@posting.google.com>
Message-ID: <3CC7E1EC.DC2E13FC@lemburg.com>

Paul Boddie wrote:
> 
> "Mark McEahern" <mark at mceahern.com> wrote in message news:<mailman.1019680127.28712.python-list at python.org>...
> > Hi, I'm trying to configure mxCGIPython:
> >
> >   http://www.egenix.com/files/python/mxCGIPython.html#Install
> >
> > I'm stuck at step 4, chmod 755 the path to the binary file cgipython.
> 
> [...]
> 
> > If I ftp to the server, I can chmod 755 the specific file.  If I change the
> > script just to reference the filename (not the full path), it works.
> 
> With the ability to perform a chmod using FTP, I don't see the reason
> for using the "install script" provided for mxCGIPython.

Note that in order for the web server to execute the cgipython
binary, the web server user must be able to cd to the directory
where the binary lives.

Depending on where you install the binary, this may not be the
possible. The only way around this is either finding a directory
which the web server user account can read or to install cgipython
in a sub-dir of cgi-bin/ which is protected against access through the
web server using e.g. an .htaccess file (depends on the web server
being used).

> > The problem is that the shebang line for the Python cgi files needs to have
> > the full path, I think.  If I don't chmod the full path, it doesn't work.
> > The full path would be:
> >
> >   #!/home/chain/pythons/cgipython
> 
> I found that it isn't always possible to deploy Python programs in a
> cgi-bin directory and to use the above notation to persuade the
> operating system to locate the Python executable. In the end, I used a
> wrapper shell script for each of my Python programs.

This should not be necessary if the directory permission
are right... but I may be wrong.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From cbrown at met.co.nz  Tue Apr  2 21:23:02 2002
From: cbrown at met.co.nz (Colin Brown)
Date: Wed, 3 Apr 2002 14:23:02 +1200
Subject: Win-2k [IOError 13] Permission denied
Message-ID: <TKtq8.80$qg.6538@ozemail.com.au>

Hi

I have an application that is moving a lot of files around on a local disk
and periodically I am getting an IOError 13 - Permission denied. Although
the program is multi-threaded, the errors seem to be occurring within the
same thread after two successive IO operations. Two code sequences that are
causing me problems at the moment are
(where x and y are files in non-shared directories):


        os.remove(y)        #  successful
        os.rename(x,y)      #  fails = IOError 13 - Permission denied


        shutil.copyfile(x,y)    #  successful
        os.remove(x)             #  fails = IOError 13 - Permission denied


I surmise that the operating system has not "completed" the first operation
before the second is requested.

The environment is Python 2.1 running under Win 2K Professional on a 1 GHz
Pentium. I would be grateful to anyone who can explain why these errors are
occurring and how to guarantee their elimination.

Thanks

Colin Brown
PyNZ




From stephen at xemacs.org  Mon Apr 15 09:59:27 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 15 Apr 2002 22:59:27 +0900
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp> <7xelhhmq5e.fsf@ruckus.brouhaha.com> <slrnablf4d.5k9.mlh@vier.idi.ntnu.no> <mailman.1018872513.23059.python-list@python.org>
Message-ID: <87zo05xdts.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Chris" == Chris Gonnerman <chris.gonnerman at newcenturycomputers.net> writes:

    Chris> "GPL Compatible" with respect to Python's license does NOT
    Chris> mean that anyone can change the license at will.  If

If by "change the license," you mean "sublicense under arbitrary
terms", I'm afraid that that is precisely what the FSF thinks it
means.  Otherwise they would not sanction it as GPL-compatible.  The
GPL says that the Work (not just explicitly GPLed parts) must be
redistributed "under the terms of this License" (section 2b).  And
this will apply recursively to those who receive copies from people
you distributed to under the GPL.

IANAL, but I don't see any other way to interpret it.

    Chris> someone incorporates GPL'd code into the Python core, the
    Chris> *aggregate* must be distributed under the GPL but the
    Chris> *core* is still under the normal Python license.

Yes, but you have to get a copy from someone who is allowed to
distribute it under the normal Python license first.  The existence of
that license doesn't exempt you from the terms of the GPL vis a vis
the copy you received under the GPL, strange as that may sound.

The GPL is based on the idea that a Work is a whole, and cannot be
subdivided.  Without that, the whole viral idea falls down.

    Chris> If you aren't the author / copyright owner, you have no
    Chris> right to change the license.

But this is what copyleft is about---preventing sublicensors from
changing the license terms.  See section 4.  No?  Clearly the authors
of the GPL believe it is possible to change license terms, and they go
to great pains to prevent it.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From lists.ASkwar at DigitalProjects.com  Fri Apr 26 14:05:19 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 20:05:19 +0200
Subject: Sending mail to an SMTP server that requires POP authentication ?
In-Reply-To: <20020426104354.6ed98204.logiplexsoftware@earthlink.net>
References: <aabrbv$2vg$1@norfair.nerim.net> <20020426104354.6ed98204.logiplexsoftware@earthlink.net>
Message-ID: <20020426180519.GU21205@teich.Garten.DigitalProjects.com>

?Cliff Wells? sagte am 2002-04-26 um 10:43:54 -0700 :
> Hm.  I think you'll have to log in with POP first, and then for some predefined
> period of time you'll be granted SMTP access.

It depends.  What you describe is often called SMTP-after-POP or
POP-before-SMTP.  However, he asked for SMTP authentication (smtp auth)
as described in RFC 2554.  Those two don't have much in common.

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 2 days 13 hours 3 minutes




From mal at lemburg.com  Wed Apr 10 16:58:35 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Wed, 10 Apr 2002 22:58:35 +0200
Subject: Working with untrusted coders
References: <NCBBLFCOHHDIKCAFGCFBEEEIKKAA.marklists@mceahern.com>
Message-ID: <3CB4A77B.D7EAD472@lemburg.com>

Mark McEahern wrote:
> 
> [Arnau Rossells]
> > [...]Python security
> > about private elements and others is merely a "don't step on the grass"
> > sign, everybody can change private functions and data in classes and even
> > modules. I'd like to know if there is some way of really securing python,
> > so parts of the code can't be overriden by the coders. Has anybody tried
> > it?
> 
> I haven't used this but it may be what you're looking for:
> 
>   http://www.lemburg.com/files/python/mxProxy.html

Right. mxProxy allows wrapping all kinds of Python objects with 
proxies and was designed for just this purpose. The underlying
objects can control which APIs are to be accessible and mxProxy
filters out any reference to the real object.

Of course, you have to make sure that the objects isn't 
referenced anywhere else to feel secure. Also note that you
can work around the mxProxy protection by writing a C
extension for this purpose.
 
> By the way, here's the Google search that resulted in the above as the first
> hit:
> 
>   http://www.google.com/search?hl=en&q=python+proxy+access
> 
> It's always a good idea to search for things first and Google is your
> friend.

Indeed :-)

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From mongiaamit at yahoo.com  Mon Apr  1 02:36:51 2002
From: mongiaamit at yahoo.com (Amit Mongia)
Date: Sun, 31 Mar 2002 23:36:51 -0800 (PST)
Subject: PyQt installation problems on windows
In-Reply-To: <3c9e25fa@nntp.server.uni-frankfurt.de>
Message-ID: <20020401073651.42513.qmail@web11305.mail.yahoo.com>

Hi Michael,
Thanks a ton for the prompt reply. Phil also replied
and it solved the problem. It was a version
incompatibility problem.
Regards,
Amit Mongia.

--- Michael 'Mickey' Lauer
<mickey at tm.informatik.uni-frankfurt.de> wrote:
> Amit Mongia <mongiaamit at yahoo.com> wrote:
> > Hi,
> > I have installed Python 2.2 , on my computer. I
> > created a gui using Qt Designer and used pyuic to
> make
> > a python file out of it. 
> > Now i want to run that file using the python
> > interpreter. Whenever i try doing that it gives an
> > error:
> > 
> > Traceback (most recent call last):
> >  File "WrapperInterface.py", line 10, in ?
> >    from qt import *
> >  File "F:\PYTHON22\lib\qt.py", line 44, in ?
> >    import libqtc
> > ImportError: DLL load failed: One of the library
> files
> > needed to run this applic
> > ation cannot be found.
> 
> Place the qt.dll from the non-commercial windows
> version
> into <windows>/system32.
> 
> :M:
> -- 
> http://mail.python.org/mailman/listinfo/python-list


__________________________________________________
Do You Yahoo!?
Yahoo! Greetings - send holiday greetings for Easter, Passover
http://greetings.yahoo.com/



From h.e.w.frater at cs.cf.ac.uk  Fri Apr 26 20:11:10 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 26 Apr 2002 17:11:10 -0700
Subject: looping through a list
Message-ID: <e2e9e807.0204261611.7514b387@posting.google.com>

I need help with a really simple algorithm that I can't get my head
around. I am looping through the list returned by a dcoracle2 execute
statement and need to add each item of data surrounded by some HTML
table tags to a string called text which I then pass to a page
generation method. At the moment I've got this:

result = c.fetchone()
  if (result != ""):
    i=0
    text = '<tr width = "80%" align = "center">'
    while (i != len(result)+1):
      i = i + 1
      text = text + '<td align = "center" valign =
"top">'+str(result[int(i)])  +'</td>'
    text = text + "</tr>"
    page.generate("","",sessionID,text)
  else:
    text = '<h4>No items found</h4>'
    page.generate("","",sessionID,text)


At the moment, it keeps saying list index out of bounds and I don't
know where I'm going wrong. I've tried all sorts of things in the
While statement and still can't get it going.

Hugh


From jeff at ccvcorp.com  Tue Apr 23 19:33:21 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 23 Apr 2002 16:33:21 -0700
Subject: Passing arguments to threading objects
References: <pan.2002.04.23.22.39.14.487844.11225@teich.garten.DigitalProjects.com>
Message-ID: <MPG.172f8f0792b30c7c9896c7@news.nwlink.com>

In article 
<pan.2002.04.23.22.39.14.487844.11225 at teich.garten.DigitalProject
s.com>, news.ASkwar at DigitalProjects.com says...

> def mache(name, max):
> 	for i in range(max):
> 		print `name` + ': ' + `i`
> 		time.sleep(random.random())
> 
> class myThread(threading.Thread):
> 	def run(self, name, max):
> 		for i in range(max):
> 			print `name` + ': ' + `i`

Aahz already answered your original question, but I wanted to add 
that you should really learn to use string formatting.  Your 
print statements above are much cleaner and more efficient if you 
instead use this:

    print "%s: %d" % (name, I)

The format specifiers embedded in the string (%s, %d) mark the 
location and the type of data to substitute.  Python will try to 
replace the %s with a string, and the %d with a (decimal) number.
Data is drawn from the tuple at the end of the expression, with 
the first specifier being replaced by the first item in the 
tuple, etc.  (The number of specifiers must match the length of 
the tuple.)

This is typically easier to follow than lots of string additions.  
It's also more efficient -- when you do multiple string 
additions, Python creates a new temporary string for *each* 
addition, whereas the string formatting requires only one new 
string no matter how many format specifiers you use.  In your 
example performance isn't a big issue, but especially if this is 
done in a tight loop, the resulting difference in efficiency can 
be huge.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From andersjm at dancontrol.dk  Thu Apr 11 08:00:26 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Thu, 11 Apr 2002 14:00:26 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <slrnab9s50.6g8.mlh@vier.idi.ntnu.no>
Message-ID: <3cb57b6b$0$78750$edfadb0f@dspool01.news.tele.dk>

"Magnus Lie Hetland" <mlh at vier.idi.ntnu.no> wrote in message
news:slrnab9s50.6g8.mlh at vier.idi.ntnu.no...
> In article <3cb3fbc0$0$78808$edfadb0f at dspool01.news.tele.dk>, Anders
> J. Munch wrote:
> >Pythons truth values are not "Something or Nothing".  The phrase seems
> >to have caught on following Laura's post, but I believe it is
> >misleading.
>
> I don't.
>
> >[] and {} are objects,
>
> Yes, as are None and 0.

The important part is that they are _mutable_ objects.  The identity of each
instance is by itself important information.

>
> > they are objects with identity, and they are
> >mutable objects.  They are empty, but they are by no means "nothing".
>
> Would you call them false? Would you call 0 empty? It's an intuitive
> way of explaining the concept, not a precise one.

With true~something and false~nothing, "if history:" in my example asks if
something was passed for the optional argument.  I'd call that intuitive.
Intuitively my function is correct as written.  Hence true~something and
false~nothing guides my intuition poorly.

And no, I wouldn't call 0 empty, although there is a reasonable association
in that 0 is the cardinality of an empty container.  I would call this stuff
"considered false" or "treated as false in Python".

As for intuitive ways of explaining concepts: Examples, examples, examples.
Accept no substitute.

and-that-goes-for-newsgroup-postings-too-ly-y'rs, Anders




From tdelaney at avaya.com  Wed Apr  3 23:41:31 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 4 Apr 2002 14:41:31 +1000 
Subject: [Python-Dev] ACCEPTED: PEP 285
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A250@natasha.auslabs.avaya.com>

> From: Andrew Koenig [mailto:ark at research.att.com]
>
> I'd like to suggest that comparisons with True are not only
> unnecessary but dangerous.  Comparisons with False are much
> less dangerous but still unnecessary.

Dangerous was the word I thought of first, but thought "may give incorrect
results" was better.

Comparing with False is just as dangerous as comparing with True ... after
all,

if [] == False:
    do_something

will not do anything ...

Tim Delaney




From mgilfix at eecs.tufts.edu  Sat Apr 20 14:34:18 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 20 Apr 2002 14:34:18 -0400
Subject: Pure python crypto
In-Reply-To: <a9s9bt$jj2$0@216.39.172.122>; from bokr@oz.net on Sat, Apr 20, 2002 at 05:43:25PM +0000
References: <20020417021703.H16954@eecs.tufts.edu> <mailman.1019257240.29930.python-list@python.org> <a9s9bt$jj2$0@216.39.172.122>
Message-ID: <20020420143418.A21270@eecs.tufts.edu>

  C extension modules already exist. There's another thread about
mkcrypto or others that have C extensions. So you should be all set
there. Good point 'bout the spelling mistake :)

                 -- Mike

On Sat, Apr 20 @ 17:43, Bengt Richter wrote:
> On Fri, 19 Apr 2002 18:59:12 -0400, Michael Gilfix <mgilfix at eecs.tufts.edu> wrote:
> 
> >  For those of you who were interested, the pure python blowfish
> >module can be found here: http://bofh.concordia.ca/blowfish.py.
> >
> >                    -- Mike
> >
> I missed the above thread, but thanks. Is there a c extension module version?
> BTW, a nit: I'm pretty sure Bruce Schneier's name doesn't have a 'd' in it ;-)
> 
> Regards,
> Bengt Richter
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (bokr)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From cliechti at gmx.net  Wed Apr  3 14:14:19 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 20:14:19 +0100
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <Xns91E5CE0E5919Dcliechtigmxnet@62.2.16.82>

roy at panix.com (Roy Smith) wrote in news:a8fba8$ln8$1 at panix2.panix.com:
> I've got a dictionary d, which may or may not have a value I'm looking
> for.  If it does, I want to do a few things.  If it doesn't, I want to
> do a few other things.  I could write:
> 
> try:
>     value = d[key]
>     foo = stuff (value)
> except KeyError:
>     foo = otherStuff (key)
> 
> which is fine as long as I can guarantee (hint: I can't) that stuff()
> doesn't raise a KeyError.  The alternative would be something like:

if d.has_key(key):
     foo = stuff(d[key])
else:
     foo = otherStuff(key)

chris

-- 
Chris <cliechti at gmx.net>



From joost_jacob at hotmail.com  Sun Apr 21 18:24:24 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 21 Apr 2002 15:24:24 -0700
Subject: Tkinter.Text.index() puzzlement
References: <3CC0B271.C76D3E95@nonesuch.com> <qB2w8.46153$vm6.6102555@ruti.visi.com> <3CC20EA2.1385A0E5@earthlink.net>
Message-ID: <13285ea2.0204211424.19cda7b1@posting.google.com>

Maybe you can try using CURRENT or INSERT instead of END ?


From huaiyu at gauss.almadan.ibm.com  Thu Apr 11 13:50:54 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 11 Apr 2002 17:50:54 +0000 (UTC)
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <slrnab9o7v.i8b.huaiyu@gauss.almadan.ibm.com> <ku3cy3ynn1.fsf@lasipalatsi.fi>
Message-ID: <slrnabbj7n.j2e.huaiyu@gauss.almadan.ibm.com>

On 11 Apr 2002 05:28:34 +0300, Erno Kuusela <erno-news at erno.iki.fi> wrote:
>In article <slrnab9o7v.i8b.huaiyu at gauss.almadan.ibm.com>,
>huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) writes:
>
>| On Wed, 10 Apr 2002 18:38:47 -0400, Brad Clements
>| <bkc at Murkworks.com> wrote:
>| >>>> class x(object):
>| >...  pass
>| >...
>| >>>> y = x()
>| >>>> type(y)
>| ><class '__main__.x'>
>| 
>| Is this a bug?  Is x a meta class?
>
>no, it is a regular (new-style) class.
>
>| >>> y.__class__ 
>| <class '__main__.x'>
>| >>> z = y()
>| Traceback (most recent call last):
>|   File "<stdin>", line 1, in ?
>| TypeError: 'x' object is not callable
>
>| Is y a class?  I'm puzzled ...
>
>no, it is an is an instance of x (and by implication, its type
>is x).

Ah, of course, type(y) is x.  I was excited about this just a few days ago,
but I guess playing with the old style class at the same time really
confuses one's mind.

I'm still finding my way around the new style classes, but I'd say that it
alone is worth the upgrade to 2.2.  Similarly nested scope is worth 2.1 and
list comprehension and augmented assignment is worth 2.0 for me.   

Huaiyu


From patmiller at llnl.gov  Thu Apr 11 11:56:32 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Thu, 11 Apr 2002 08:56:32 -0700
Subject: binding a reference to a variable
Message-ID: <3CB5B230.CDECEBFB@llnl.gov>

Bengt writes>
> >>> def set(name, val, ns=locals()):
>  ...     ns[name]=val
>  ...     return val

Bengt noted that 'ns' is local to the declaration module of the
'def set.'  

The workaround for that is

def set(name,val):
   import sys
   ns = sys._getframe().f_back.f_locals # local dictionary of caller
   ...

BUT WARNING!!!!! There is a bigger problem --  Changes to the local
dictionary are NOT reflected in the state of the execution.

The following does not work as intended....

def f():
   a = 7
   L = locals()
   print L.keys() # Prints ['a','L'] or some such
   L['a'] = 99  # Update the value in L
   print a      # But wait!  This prints out '7'

The issue is the implementation of the code object supporting f.
The value of the locals dictionary is created on the fly by
the call to locals().  The actual values of the local variables
are on a tiny, predefined stack frame that is used by the bytecode
interpreter wherein local variables are known only by an
integer offset.

Creating true "local" variables is astonishly tricky.

You can actually figure out where names map into the frame instance
(weave does (or will do) it), but it is messy and very implementation
dependent.

Pat

-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

All the world's a stage, / And the men and women merely players: / They
have their exits and their entrances; / And one man in his time plays
many parts.  -- William Shakespeare




From pyth at devel.trillke.net  Sun Apr 21 11:43:08 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 17:43:08 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <304d20df.0204210629.3f2b4d58@posting.google.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com>
Message-ID: <20020421154308.GF10574@devel.trillke>

On Sun, Apr 21, 2002 at 07:29:08AM -0700, Allan Crooks wrote:
> Hi,
> 
> I was wondering what the best approach is on providing libraries for
> varying versions of Python?
>  (...)
> 
> ---------------
> 
> <2.0 code>
> if sys.version[:2] < (2, 1): return
> 
> <2.1 code>
> if sys.version[:2] < (2, 2): return
> 
> <2.2 code>
> 
> ---------------
> 
> But that wouldn't work, because earlier versions of Python would still
> report SyntaxErrors, as it would have a problem with iterator
> statements.

there might a very interesting way to do it:

#include the next function into your module
def forversion(minversionstr, execstring):
	""" simple check if version is ok ... """
        from sys import version
        from string import split,find
        sysversion=split(version[:find(version,' ')],'.')
        minversion=split(minversionstr,'.')

        return sysversion < minversion and '' or execstring


# and this is what you can do inside your module 

exec forversion("2.2",
"""
#special hottest code for 2.2
from __future__ import generators

def fantastic_generator(str,step=2):
    for i in xrange(0,len(str),step):
        yield int(str[i:i+step])

class a:
    def forward(self,str):
        for obj in fantastic_generator(str):
            yield obj

""")

exec forversion("2.2",
"""
for obj in fantastic_generator('052342'):
        print "fantastic generator provides:",obj

for obj in a().forward('663311'):
    print 'forwarder yields', obj
""")

welcome to python's powerful execution facilities :-)
If you call this with python 2.1, 2.0 it does nothing 
(same for 1.5, i hope). 

Defining 'version-dependent' methods of objects with this
idiom basically works, too. Onyl __future__ makes
some problems. The problem here is that

  exec "from __future__ import generators"
  def f():
    yield 1

does not seem to work completly as expected. so you can't easily 
use generators in several 'forversion' places. Just use them 
at the beginning.

Everything else than __future__ should be ok, though!

regards,

	holger




From prova at qui.com  Sun Apr 28 05:46:19 2002
From: prova at qui.com (Utente)
Date: Sun, 28 Apr 2002 09:46:19 GMT
Subject: Help again: py2exe and win98
Message-ID: <LxPy8.72046$vF6.2149827@news2.tin.it>

I'm playing with python 2.2 and py2exe.
It works on a Windows XP system, generating the executable i need (except a
problem i posted in another message) but when i try to do the same thing on
a win 98 system (same files, same installations, same PYTHONPATH) it fails,
saying:

>changing back to 'C:\dati\Python\PrimaNota'
>creating dist\SMPNotaMain\SMPNotaMain.exe
>Traceback (most recent call last):
>  File "setup-exe.py", line 10, in ?
>    scripts = ['SMPNotaMain.pyw'])
>  File "C:\PYTHON22\lib\distutils\core.py", line 138, in setup
>    dist.run_commands()
>  File "C:\PYTHON22\lib\distutils\dist.py", line 893, in run_commands
>    self.run_command(cmd)
>  File "C:\PYTHON22\lib\distutils\dist.py", line 913, in run_command
>    cmd_obj.run()
>  File "C:\PYTHON22\Lib\site-packages\py2exe\build_exe.py", line 589, in
run
>    self.create_exe(exe_name, arcname, use_runw)
>  File "C:\PYTHON22\Lib\site-packages\py2exe\build_exe.py", line 799, in
create_
>exe
>    delete = self.add_versioninfo(exe_name, delete)
>  File "C:\PYTHON22\Lib\site-packages\py2exe\build_exe.py", line 844, in
add_ver
>sioninfo
>    add_resource(exe_name, str(vr), RT_VERSION, 1, delete)
>RuntimeError: BeginUpdateResource: La funzione ? valida soltanto in
modalit? Win
>32.
>
>C:\dati\Python\PrimaNota

Where "La funzione ? valida soltanto in modalit? Win32" is the italian for
something like "The function is valid only in Win32 mode".
Windows 98 is not a win 32 system???

Thanks in advance again, Max




From ktrinh at allmerica.com  Wed Apr  3 15:01:39 2002
From: ktrinh at allmerica.com (Khanh Trinh)
Date: 3 Apr 2002 12:01:39 -0800
Subject: PyException
Message-ID: <1a27883b.0204031201.79639b2f@posting.google.com>

Hi Everyone,

I've been having this problem for a week now but going nowhere with
this.
Any suggestion or answer is greatly appreciated.

The package was purchased from a vendor and they couldn't find the
answer for this at all.


Thank you very much.


Khanh Trinh




Message: 

Internal Error
Wed Apr 03 14:21:09 EST 2002  (PG3)

Traceback (innermost last):
  File "<string>", line 18, in ?
ImportError: No module named iclick

	at org.python.core.PyException.<init>(PyException.java)
	at org.python.core.PyException.<init>(PyException.java)
	at org.python.core.Py.ImportError(Py.java)
	at org.python.core.imp.import_logic(imp.java)
	at org.python.core.imp.import_name(imp.java)
	at org.python.core.imp.importName(imp.java)
	at org.python.core.ImportFunction.load(__builtin__.java)
	at org.python.core.ImportFunction.__call__(__builtin__.java)
	at org.python.core.PyObject.__call__(PyObject.java)
	at org.python.core.__builtin__.__import__(__builtin__.java(Compiled
Code))
	at org.python.core.imp.importFromAs(imp.java(Compiled Code))
	at org.python.core.imp.importFrom(imp.java)
	at org.python.pycode._pyx1.f$0(<string>:18)
	at org.python.pycode._pyx1.call_function(<string>)
	at org.python.core.PyTableCode.call(PyTableCode.java)
	at org.python.core.PyCode.call(PyCode.java)
	at org.python.core.Py.runCode(Py.java(Compiled Code))
	at org.python.core.Py.exec(Py.java(Compiled Code))
	at org.python.util.PythonInterpreter.exec(PythonInterpreter.java(Compiled
Code))
	at com.iclick.tools.script.plugins.PythonPlugin.evaluate(PythonPlugin.java(Compiled
Code))
	at com.iclick.tools.script.ScriptFacade.evaluate(ScriptFacade.java:266)
	at com.iclick.clickhr.handbook.W4Calculator.runScript(W4Calculator.java:331)
	at com.iclick.clickhr.handbook.W4Calculator.<init>(W4Calculator.java:196)
	at com.iclick.clickhr.handbook.MyPaycheckController.calculateW4(MyPaycheckController.java:271)
	at java.lang.reflect.Method.invoke(Native Method)
	at com.iclick.tools.expr.MethodCall.eval(MethodCall.java(Compiled
Code))
	at com.iclick.tools.expr.FormattedReflexp.eval(FormattedReflexp.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)
ScriptException
  wrapping PyException
	at com.iclick.tools.script.plugins.PythonPlugin.evaluate(PythonPlugin.java(Compiled
Code))
	at com.iclick.tools.script.ScriptFacade.evaluate(ScriptFacade.java:266)
	at com.iclick.clickhr.handbook.W4Calculator.runScript(W4Calculator.java:331)
	at com.iclick.clickhr.handbook.W4Calculator.<init>(W4Calculator.java:196)
	at com.iclick.clickhr.handbook.MyPaycheckController.calculateW4(MyPaycheckController.java:271)
	at java.lang.reflect.Method.invoke(Native Method)
	at com.iclick.tools.expr.MethodCall.eval(MethodCall.java(Compiled
Code))
	at com.iclick.tools.expr.FormattedReflexp.eval(FormattedReflexp.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)
java.lang.reflect.InvocationTargetException: Traceback (innermost
last):
  File "<string>", line 18, in ?
ImportError: No module named iclick

	at org.python.core.PyException.<init>(PyException.java)
	at org.python.core.PyException.<init>(PyException.java)
	at org.python.core.Py.ImportError(Py.java)
	at org.python.core.imp.import_logic(imp.java)
	at org.python.core.imp.import_name(imp.java)
	at org.python.core.imp.importName(imp.java)
	at org.python.core.ImportFunction.load(__builtin__.java)
	at org.python.core.ImportFunction.__call__(__builtin__.java)
	at org.python.core.PyObject.__call__(PyObject.java)
	at org.python.core.__builtin__.__import__(__builtin__.java(Compiled
Code))
	at org.python.core.imp.importFromAs(imp.java(Compiled Code))
	at org.python.core.imp.importFrom(imp.java)
	at org.python.pycode._pyx1.f$0(<string>:18)
	at org.python.pycode._pyx1.call_function(<string>)
	at org.python.core.PyTableCode.call(PyTableCode.java)
	at org.python.core.PyCode.call(PyCode.java)
	at org.python.core.Py.runCode(Py.java(Compiled Code))
	at org.python.core.Py.exec(Py.java(Compiled Code))
	at org.python.util.PythonInterpreter.exec(PythonInterpreter.java(Compiled
Code))
	at com.iclick.tools.script.plugins.PythonPlugin.evaluate(PythonPlugin.java(Compiled
Code))
	at com.iclick.tools.script.ScriptFacade.evaluate(ScriptFacade.java:266)
	at com.iclick.clickhr.handbook.W4Calculator.runScript(W4Calculator.java:331)
	at com.iclick.clickhr.handbook.W4Calculator.<init>(W4Calculator.java:196)
	at com.iclick.clickhr.handbook.MyPaycheckController.calculateW4(MyPaycheckController.java:271)
	at java.lang.reflect.Method.invoke(Native Method)
	at com.iclick.tools.expr.MethodCall.eval(MethodCall.java(Compiled
Code))
	at com.iclick.tools.expr.FormattedReflexp.eval(FormattedReflexp.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)
ScriptException
  wrapping PyException
	at com.iclick.tools.script.plugins.PythonPlugin.evaluate(PythonPlugin.java(Compiled
Code))
	at com.iclick.tools.script.ScriptFacade.evaluate(ScriptFacade.java:266)
	at com.iclick.clickhr.handbook.W4Calculator.runScript(W4Calculator.java:331)
	at com.iclick.clickhr.handbook.W4Calculator.<init>(W4Calculator.java:196)
	at com.iclick.clickhr.handbook.MyPaycheckController.calculateW4(MyPaycheckController.java:271)
	at java.lang.reflect.Method.invoke(Native Method)
	at com.iclick.tools.expr.MethodCall.eval(MethodCall.java(Compiled
Code))
	at com.iclick.tools.expr.FormattedReflexp.eval(FormattedReflexp.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)
ExprException:
    calculateW4 on com.iclick.custom.allmerica.clickhr.handbook.AMyPaycheckController at 446a4382
  wrapping InvocationTargetException
	at com.iclick.tools.expr.MethodCall.eval(MethodCall.java(Compiled
Code))
	at com.iclick.tools.expr.FormattedReflexp.eval(FormattedReflexp.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)
HtmlFormProcessor$SubmissionException
  wrapping ExprException:
    calculateW4 on com.iclick.custom.allmerica.clickhr.handbook.AMyPaycheckController at 446a4382
  wrapping InvocationTargetException
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processInvocation(HtmlFormProcessor.java(Compiled
Code))
	at com.iclick.servlet.ihtml.HtmlFormProcessor.processFormSubmission(HtmlFormProcessor.java:321)
	at com.iclick.servlet.ihtml.HtmlController.processFormSubmission(HtmlController.java:157)
	at com.iclick.servlet.ihtml.HtmlController.getCompiledPage(HtmlController.java:128)
	at com.iclick.servlet.ihtml.HtmlController.service(HtmlController.java:113)
	at com.iclick.servlet.Controller.serviceRequest(Controller.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.myservice(IClickServlet.java(Compiled
Code))
	at com.iclick.servlet.IClickServlet.doPost(IClickServlet.java:145)
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at javax.servlet.http.HttpServlet.service(HttpServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictServletInstance.doService(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet._service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.IdleServletState.service(StrictLifecycleServlet.java:287)
	at com.ibm.servlet.engine.webapp.StrictLifecycleServlet.service(StrictLifecycleServlet.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstance.service(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ValidServletReferenceState.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.ServletInstanceReference.dispatch(ServletManager.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.handleWebAppDispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.dispatch(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.webapp.WebAppRequestDispatcher.forward(WebAppRequestDispatcher.java(Compiled
Code))
	at com.ibm.servlet.engine.srt.WebAppInvoker.handleInvocationHook(WebAppInvoker.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CachedInvocation.handleInvocation(CachedInvocation.java(Compiled
Code))
	at com.ibm.servlet.engine.invocation.CacheableInvocationContext.invoke(CacheableInvocationContext.java:106)
	at com.ibm.servlet.engine.srp.ServletRequestProcessor.dispatchByURI(ServletRequestProcessor.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.OSEListenerDispatcher.service(OSEListener.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp$ServiceRunnable.run(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.SQEventListenerImp.notifySQEvent(SQEventListenerImp.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQEventSource.notifyEvent(SQEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.notifyService(SQWrapperEventSource.java(Compiled
Code))
	at com.ibm.servlet.engine.oselistener.serverqueue.SQWrapperEventSource$SelectRunnable.run(SQWrapperEventSource.java:220)
	at com.ibm.servlet.engine.oselistener.outofproc.OutOfProcThread$CtlRunnable.run(OutOfProcThread.java:248)
	at java.lang.Thread.run(Thread.java:481)


From cliechti at gmx.net  Fri Apr  5 14:38:02 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 5 Apr 2002 20:38:02 +0100
Subject: Can someone comment on my code?
References: <3CAD0719.6050707@NO-SPAM.bigfoot.com>
Message-ID: <Xns91E7D215AD096cliechtigmxnet@62.2.16.82>

Chris Burn <chris_burn at NO-SPAM.bigfoot.com> wrote in
news:3CAD0719.6050707 at NO-SPAM.bigfoot.com: 

> Hi, I've been writing perl, shell, c programs for a while but I
> thought I should move on to python.

welcome

> So this is my first real venture in python. Basically I work on an app
> that auto saves to your home dir every 10 mins, but I keep my working 
> files in a dir elsewhere. So this script is intended to be run after a
> crash, it searches for the last autosave and the last normal save in
> the current dir. It then copies the autosave to the current dir with
> _rec stuck on the end.
> 
> Anyway, it works, but I'm new to error handling in python and it would
> be good to know if I'm handling errors correctly and also if there are
> easier ways to do things or conventions I should be keeping to? It
> seems to be a bit long to me.
> 
> Oh it needs to run on python 1.5.2
> 
> Thanks
> Chris
>
>
> #!/usr/bin/python
> 
> import os, sys, re, shutil
> 
> class FileListError(Exception):
>      def __init__(self, message):
>              self.message = message
>      def __str__(self):
>           return self.message
...

no need to redefine __init__ and __str__ here. this will work as expected:

class FileListError(Exception): pass
class NoAutoError(Exception): pass
class NoSavesError(Exception): pass
class DontWriteError(Exception): pass


> def fileList(dir):
>      flist = []
>      try:
>           for file in os.listdir(dir):
>                flist.append(os.path.join(dir, file))
>           return flist
>      except:
>           raise FileListError(dir)

catch specific exceptions. just catchig everything makes debuging and code 
maintenace a _lot_ harder.

> def isSave(fname):
>      return re.search('autoSave\d\.shk', fname)

note that '\' in re are escaped twice. once by the python parser and once 
py the re module. this doesn't hurt as '\d' and '\.' are no valid python 
escapes. you can avoid problems by using raw strings:
 
r'autoSave\d\.shk'

> def isScript(fname):
>      return re.search('\.shk', fname)
> 
> def getLastAuto():
>      files = fileList("/home/users/cburn/nreal/autosave")
>      files = filter(isSave, files)
> 
>      if files == []:
>           raise NoAutoError

no need to test for emptyness explicit. "if not files: ..." will do.

>      filetime = []
>      for fname in files:
>           filetime.append([os.path.getmtime(fname), fname])
> 
>      filetime.sort()
>      filetime.reverse()
>      return filetime[0][1]
> 
> def getLastSave():
>      files = fileList(os.getcwd())
>      files = filter(isScript, files)
> 
>      if files == []:
>           raise NoSavesError
> 
>      filetime = []
>      for fname in files:
>           filetime.append([os.path.getmtime(fname), fname])
> 
>      filetime.sort()
>      filetime.reverse()
>      return filetime[0][1]
> 
> def okWrite(file):
>      if not os.path.exists(file):
>           raise DontWriteError

when you add

if __name__ == '__main__':
    	...

here and indent to following codeblocks, you can use the same file as a 
script and as a module for an otherone. makes code reuse very easy.

> try:
>      lastAuto = getLastAuto()
>      lastSave = getLastSave()
> 
>      newSave = os.path.split(lastSave)
>      newSave = os.path.join(newSave[0], re.match('([^\s]+)\.shk',
>      newSave[1]).group(1) + '_rec.shk') 
> 
>      try:
>           okWrite(newSave)
>           print 'Copying %s to %s' % (lastAuto, newSave)
>           ok = raw_input("OK? ")
>           if (ok == "y"):
>                shutil.copyfile(lastAuto, newSave)
>      except:
>           raise
if you just re-raise the exception you don't need the try: ... except: 
block at all...

> except FileListError, err:
>      print 'Error: Cannot read %s directory' % err
> except NoAutoError:
>      print 'Error: Cannot find any autosaves!'
> except NoSavesError:
>      print 'Error: Cannot find any shake scripts!'
> except DontWriteError:
>      print "Can't copy, file exists: %s" % newSave
> except:
>      print "Unexpected error:", sys.exc_info()[0]

you don't need to handle all errors with execptions...

if okWrite(newSave):
    	...

is very readable and okWrite is shorter too:

def okWrite(file):
    return os.path.exists(file)


HTH, chris


-- 
Chris <cliechti at gmx.net>



From wavy_gravy at worldnet.att.net  Thu Apr 18 10:13:27 2002
From: wavy_gravy at worldnet.att.net (Wavy Gravy)
Date: 18 Apr 2002 07:13:27 -0700
Subject: PyType_GenericAlloc
Message-ID: <5e603766.0204180613.523d8bb2@posting.google.com>

Greetings,
 I am trying to get familiar with the new Type/Class API. Browsing
through various sources (only few exist), I am trying to figure out
what exactly this function returns? I notice that this is the most
common fuction in the "tp_alloc" slot for your own builtin type.
Probably an example shoud clarify, ex. say you have your object as:

typedef struct {
  char *name;
} PyCatObject;

staticforward PyTypeObject PyCat_Type;

OLD_WAY Creation
-----------------
 PyObject_New(<PyCatObject pointer),(PyCat_Type pointer). And this
make sense
 because I can see the return type being PyCatObject *.

NEW_WAY (2.2)
------------
 fill the tp_new slot with your function and then in general fill the
tp_alloc slot with PyType_GenericAlloc. And then in your tp_new
function in general make a call to type->tp_alloc(type,0).
 So in the example above, what I am not able to figure out is how does
the PyType_GenericAlloc know to return PyCatObject * ? I can see that
it knows the size because we fill it in the appropriate slot.
 
Maybe i don't understand the old way to clearly? Is the "actual"
object the first thing in the returned malloced bytes? Is there a any
doc on the actual layout of the PyObject* that is returned by this
calls?

Thanks


From Chris.Barker at noaa.gov  Wed Apr 10 13:05:03 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 10 Apr 2002 10:05:03 -0700
Subject: Help: Python for a big commercial project?
References: <mailman.1018437980.14078.python-list@python.org>
Message-ID: <3CB470BE.8D1CB58C@noaa.gov>


Cereal Muesli wrote:
> I want to propose Python as
>   the language of choice for internal maintenance tasks, high level control
> tasks for the loading procedures, etc...
>   the people here is not likely to use any free stuff for "serious"
> development (you know...no money = no quality, sigh!),

> I need to know
>   some things:

These are all important questions, but I'm curious what the other
options being considered are, and how they stack up in these ways, and
of course, others.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From gimbo at ftech.net  Thu Apr 11 04:20:18 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 11 Apr 2002 09:20:18 +0100
Subject: PEP 282: A Logging System
In-Reply-To: <000501c1e0b8$d1136cc0$652b6992@alpha>
References: <Pine.LNX.4.44L.0204101618300.13708-100000@ods.pp.ru> <000501c1e0b8$d1136cc0$652b6992@alpha>
Message-ID: <20020411082017.GB1261@andy.tynant.ftech.net>

On Wed, Apr 10, 2002 at 06:54:48PM +0100, Vinay Sajip wrote:

> The logging module is not designed to use threading, or for thread safety.

Really?  It's _not_ thread safe?  Oh.  Will it be, ever?

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From aleax at aleax.it  Tue Apr  9 06:04:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 10:04:44 GMT
Subject: changing local dictionnary for functions
References: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr> <vIos8.3287$vT6.670@atlpnn01.usenetserver.com>
Message-ID: <01zs8.62383$pT1.1793993@news1.tin.it>

Steve Holden wrote:

> "Boris Boutillier" <boris at cantal.lip6.fr> wrote in message
> news:pan.2002.04.08.18.15.31.115122.19669 at cantal.lip6.fr...
>> Is there a way to pass a local dictionnary to a function as it is
>> possible to do with a file with the execfile statement ?
>>
>> In the python reference manual, in the section 4.1 : Code
>> Blocks,execution frames and namespace, there is a table that say that
>> function body uses globals namespace of the containing block and a new
>> locals namespace. And there is a note that say global and local can be
>> overrident, how is it possible ? I think it would be the answer to my
>> first question.
> 
> I suspect this is an arror in the documentation, since the namespaces
> provided to a function call are (unless I am mistaken) determined solely
> by the context of the call. Sorry.
> 
> If  I'm wrong, you can guarantee that three different people will point
> this out within three minutes of my hitting the "send" key :-)

The only way that occurs to _me_ for "changing a function's
dictionaries" would be to extract the code object from the function
and build a new function (with module new) using that code object
and whatever dictionary you want for globals.  Even so, the locals
(not really a dictionary... they only halfway-masquerade as one
through the built-in function of that name) would be created anew
at each call to the functions, so it still wouldn't meet the
request by Boris as I understand it.


Alex



From martin at v.loewis.de  Tue Apr 16 02:36:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:36:54 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de> <3cbb7e33@news.mhogaming.com>
Message-ID: <m31ydgm9o9.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> | That is only meaningful for positive numbers - it just doesn't work
> | for twos complement.
[...]
> Ahhhh!!!  Got'cha!!!!
> 
> Python DOESN'T HAVE UNSIGNED NUMBERS!!!

Why are you shouting at me? I didn't say "unsigned numbers", I said
"positive numbers". Python has infinitely many of those - assuming you
have enough memory to represent them :-)

> Since there is no way to know this happened to or object, we either
> have to make an overly complex "Rube Goldberg" type exploration, if
> we want to be able to do a logical right shift of an int-long
> object.

Just restrict yourself to positive numbers, and the logical shift is
the same as the arithmetic shift.

> THAT is the reason for the '>>>' operator-- the designers of Java
> got rid of unsigned integers, because many programming mistakes
> happened in 'C' with them in existence-- this (in turn) necessitated
> the creation of a different operator for 'arithmetic right shift',
> and 'logical right shift'.

The other reason, of course, is that they still wanted to represent 32
logical bits in a single int value. In Python, you can represent
arbitrary many bits in a positive number, and the right-shift will be
"logical".

> Quite frankly, I don't give a damn HOW the long (or int) is represented
> internally to the machine

Good, neither do I. I wonder why you bring up internal representation
then all the time.

> Is what I want so EVIL that Python can't do things that way?

No, it is impossible to do what you want - you have not presented a
single consistent specification, yet, which is also implementable.

Regards,
Martin


From narnett at mccmedia.com  Wed Apr 24 22:30:15 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 24 Apr 2002 19:30:15 -0700
Subject: Constructing an object from another subclass of the same parent?
Message-ID: <NDBBLAEMOMEHKALAMKKJEEPLKDAA.narnett@mccmedia.com>

I think that the answer to this is somewhere in a part of my brain that's
off-line at the moment (long day)... but I'm having trouble even composing a
reasonable search for the answer.

I can't remember how to construct an object that is a subclass of the parent
of the subclass I'm in.

E.g.,

class knight:
	blah, blah

class french(knight):
	blah, blah in your general direction

class english(knight):
	joker=french('Renee')

Neither the stuff above nor self.french('Renee') work, which really didn't
surprise me, since they shouldn't, but what should?

Something really awful along the lines of importing the module I'm in came
to mind.

And in case anyone cares, the parent's main purpose is to open a database,
create a handle and give me some methods for shortcuts to run certain types
of queries.  It has various subclasses that perform operations on the
database.  And now one of them needs one of the others.  Usually by the time
I've asked a question this clearly, the answer is equally clear and I throw
away the message... but apparently this one will actually survive.

Thanks... and surely it's time to quit for the day.

Nick

--
narnett at mccmedia.com
(408) 904-7198





From martin at v.loewis.de  Mon Apr  1 17:53:38 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 00:53:38 +0200
Subject: Initialization problem of shared library in HP-UX11
References: <7aafbf19.0204010929.62d17d6a@posting.google.com> <m37knr71pl.fsf@mira.informatik.hu-berlin.de> <7aafbf19.0204011414.25acdd0@posting.google.com>
Message-ID: <m3d6xjxc7h.fsf@mira.informatik.hu-berlin.de>

flz at cfdrc.com (gother) writes:

> I know windows do the initialization when it load the shared lib. In
> HP, when the shared lib links to an main function, it do the same
> thing. But somehow HP-UX11 failed to do it inside the python which
> called shl_load function. So I guess inside the executable the shared
> lib is loaded with different system function call instead of shl_load
> or different arguments of shl_load. If anybody be familar with the
> internal loading process of HP, he must have the answer.

Perhaps the answer is: It just can't work; it is a known deficiency of
HP-UX. Then you are back to my suggestion. I don't know whether this
would be a correct answer.

Now that you seem to want understanding this, we should debug it it
step by step. 

First question: What is the command line that you use to link the
extension module?

If it starts with 'ld', that could explain it. Many systems require
special options to the linker (in particular addition of object
files) to support initialization of shared libraries.

If it starts with 'gcc', that still could explain it. Try the system
compiler; it might be a bug in gcc.

If it is the bundled C compiler, this definitely explains it.

If it is the unbundled compiler, next question:

What is the documentation for all the command line option you are
giving? What other options related to shared libraries are mentioned
in the documentation of ld(1) and c89(1)?

And so on. It is unlikely that you find a lot of HP-UX expertise in
this group - I recommend to contact HP.

Regards,
Martin


From bokr at oz.net  Mon Apr  8 12:55:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 16:55:12 GMT
Subject: Bitwise/Binary operations in Python < 2.1
References: <slrnab2sb5.j99.petro@lists.bounty.org> <Kgfs8.59392$pT1.1709038@news1.tin.it>
Message-ID: <a8si1g$bn7$0@216.39.172.122>

On Mon, 08 Apr 2002 11:36:10 GMT, Alex Martelli <aleax at aleax.it> wrote:

>petro wrote:
>        ...
>> * priorities/facilities are encoded into a single 32-bit quantity, where
>> the * bottom 3 bits are the priority (0-7) and the top 28 bits are the
>> facility
That leaves one bit out, right? If you really mean _top_ 28 bits starting
with bit 31 (numbering lsb..msb as 0..31), then you will need >> 4 in
place of >> 3 in Alex's code below.

>> * (0-big number).  Both the priorities and the facilities map roughly
>> * one to strings in the syslogd(8) source code.  This mapping is
>> * included in this file.
>> 
>> The problem is, well, basically I'm pretty stupid, and can for the life
>> of me figure out how to do this in such a way that it's portable from
>> Python 1.5.2 to 2.x
>
>what about:
>
>def extractPriorityAndFacility(single32bitQuantity):
>    priority = single32bitQuantity & 7
>    facility = (single32bitQuantity >> 3) & 0xFFFFFFF
Depending on the meaning of "top" ...   ^-- might need 4 here??
>    return priority, facility
>
Just a nit, but I suppose it could make a difference.

Regards,
Bengt Richter


From hopfgartner at rolmail.net  Mon Apr 15 04:11:58 2002
From: hopfgartner at rolmail.net (Peter Hopfgartner)
Date: Mon, 15 Apr 2002 08:11:58 GMT
Subject: SAP DCOM Connector, return type of DimAs
References: <pan.2002.04.15.09.02.11.851376.559@rolmail.net>
Message-ID: <pan.2002.04.15.09.09.30.365898.559@rolmail.net>

On Mon, 15 Apr 2002 10:04:27 +0200, Peter Hopfgartner wrote:

> Dear members of the Python community,
> 
> is there anybody, who used Python successfully to connect to SAP through
> a DCOM Connector?
> 
> My current problem is:
> 
> I get the COM project through:
> 
> 
>         # Rufe die COM-Komponente auf
>         self.sap_com = win32com.client.Dispatch('SAP.BAPI.1')
> (... connect successfully to SAP ... )
> 
>         self.porders = self.sap_com.GetSAPObject('PlannedOrder')
>         self.header =  self.sap_com.DimAs(self.porders, 'Create',
>                                                                    'HeaderData')
> 
>         self.header =  self.porders.DimAs('Create','HeaderData')
> 
> Then, I would like to insert some data through self.header.Value, but
> this is seen only as an integer from Python.
> 
> In Visual Basic the data is inserted as
> 
> header.Value("Material") = 123456
> 
> In Visual Basic everything works fine.
> 
> All COM objects are early bounded, as described in Hammonds (excellent)
> Python on Win32 book.
> 
> I'm using ActiveState Python 2.2 on Windows 2000 Server.
> 
> If anybody could help me he will earn my eternal gratitude.
> 
> Peter


Sorry, the second DimAs invocation was only some dispaered test.

Peter, still looking for help.


From mlh at vier.idi.ntnu.no  Tue Apr  9 06:28:27 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 9 Apr 2002 10:28:27 +0000 (UTC)
Subject: Private attributes
Message-ID: <slrnab5gib.ho6.mlh@vier.idi.ntnu.no>

Just playing with the property descriptor API (and being absolutely
horrid with the inspect module <wink>)... In case anyone else might
find it intersting (or amusing):

-----------------------------------------------------
__metaclass__ = type
import inspect

def called_from(obj):
    try:
        frame = inspect.stack()[2][0]
        caller = frame.f_locals['self']
    except:
        return 0
    return caller is obj

class private:
    def __init__(self, name):
        self.name = name
    def __get__(self, obj, type=None):
        if called_from(obj):
            return obj.__dict__[self.name]
        else:
            raise AttributeError
    def __set__(self, obj, value):
        if called_from(obj):
            obj.__dict__[self.name] = value
        else:
            raise AttributeError
    def __delete__(self, obj):
        if called_from(obj):
            del obj.__dict__[self.name]
        else:
            raise AttributeError
-----------------------------------------------------

Example use:


class Test:
    x = private('x')
    def __init__(self):
        self.x = 42
    def test(self):
        print self.x

>>> t = Test()
>>> t.test()
42
>>> t.x = 43
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "private.py", line 27, in __set__
    raise AttributeError
AttributeError
>>> t.x

Having to supply the attribute name is a bit clunky, but could easily
be fixed by using a custom metaclass. Then one could just have a
marker constant or something for the metaclass to detect when creating
the class, e.g:

class Test2:
    x = private
    ...

- Magnus (who knows how naughty he has been, and who will never
  actually use this stuff ;)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From narnett at mccmedia.com  Wed Apr 24 23:11:51 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 24 Apr 2002 20:11:51 -0700
Subject: Constructing an object from another subclass of the same parent?
In-Reply-To: <NDBBLAEMOMEHKALAMKKJEEPLKDAA.narnett@mccmedia.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJCEPNKDAA.narnett@mccmedia.com>

Never mind.

Seems that (for the first time, at least) I made the error of re-using the
class name as a variable.  I was thinking that it should have simply worked
and now it does.

Suddenly, the convention of distinguishing names with capitalization looms
important.

Nick





From lumber at jack.com  Tue Apr 23 17:52:38 2002
From: lumber at jack.com (Lumberjack)
Date: 23 Apr 2002 21:52:38 GMT
Subject: Kindly Get Back To Me Please  24/04/02 21:32:21
References: <mailman.1019590182.14138.python-list@python.org>
Message-ID: <Xns91F997455191Dlumberjack@216.148.53.84>

"brendowean" <bramk at arabia.com> wrote:
> This letter may come to you as a surprise due to the fact that we have
> not yet met.

Shucks, no surprise at all. I meet your kind all the time. In fact, you 
look familiar. Aren't you the infamous Bram Stoker in disguise? Why last we 
met, I thought you had died in the clutches of that Python!

> The message could be strange but reel if you pay some 
> attention to it.

Not nearly as strange as how you managed to escape the clutches of that 
creature!

> I could have notified you about it at least for the sake of your
> integrity. Please accept my sincere apologies. In bringing this message
> of goodwill to you, I have to say that I have no intentions of causing
> you any pains.

I spit on your good will! And damn your eyes!

> I am  Mr Bram Khumalo,son of the late rebel leader  
> Maubane Khumalo
> of Angola who was killed on the 22nd of febuary 2002 .

Died: 2002-02-22. 2 bad. Python: the bane or Maubane.

> I managed to get
> your contact details through "The World Business Journal", a journal
> of the Johannesburg Chamber of Commerce in South Africa in the time I
> was desperately looking for a trustworthy person to assist me in this
> confidential business.

Yes yes, everyone knows I am trustworthy! But you, what of you!?

> my late father, Maubane Khumalo was able to deposit a large sum of
> money in differnt banks in europe My father is presently death and the 
> movement
> of his family members (including me) is restricted.

I warned you about losing your account number! Did you try: "import * from 
Netherlands"?

> We are forbidden
> to either travel abroad or out of our localities. Presently, the 
> US$8,500,000.00
> EIGHT, MILLION, FIVE HUNDRED DOLLARS my father transfered to 
> Netherlands
> is safe and is in a security firm. Before you can get access to it i
> have to give you the password I am therefore soliciting your help to
> have this money transfered into your account.

Does this firm use the Python crypt module? Have you submitted a bug 
report?

> before my government get
> wind of this fund .You know my father was a rebel leader in Angola 
> before
> his death My reason for doing this is because it will be difficult for
> the Angolan government to trace my father's money to an individual's
> account, especially when such an individual has no relationship ,I 
> decided
> to keep that money for my family use. At present the money is
> kept in a Security Company in nertherland.
> I am currently and temporarily living in Angola with my husband my 
> brother
> has a refugee status, Moreover the political climate
> in Angola at the moment being so sensitive and unstable.

I tried to warn Guido about instability, but does he listen? Noooo! But 
another release of Python should fix any Angola module exceptions.

> With this 
> password
> and information I will send to you, and power of attorney to the 
> security
> firm, When you are ready i will give you the information needed before
> you can get access to the fund you will then proceed to Netherlands 
> where
> the US$8,500,000.00 EIGHT, MILLION, FIVE HUNDRED DOLLARS will be given
> to you as payment. Alternatively, you can have the fund transferred 
> into
> any account that suits you.

I will be transferring the funds via a standard socket connection to the 
PSA account if it is all the same to you. And I just checked, the amount in 
the account is US$8,499,998.50. It looks like you've been hit with a 
US$1.50 Angolan transfer fee.

> Kindly get back to me through my email address,with all your coatacts
> addresses and send me your name as appear in your passport,and your
> Telephone and fax number.waiting to hear from you, God bless you.

Okay, here is the information you need:
1600 Pensylvania Avenue
Washington DC

Lumber Jack
Voice and fax: country code 1, 1-800-555-1212

> Yours sincerely,
> Bram Khumalo. 

I'm okay, I sleep all night and I work all day!


From alan.salmoni at ntlworld.com  Fri Apr 12 09:46:52 2002
From: alan.salmoni at ntlworld.com (Alan)
Date: Fri, 12 Apr 2002 14:46:52 +0100
Subject: saving contents of wxHtmlWindow
Message-ID: <vxBt8.12401$C21.2528592@news6-win.server.ntlworld.com>

apologies if this is the wrong place, but when using a wxHtmlWindow, how 
can I asve the contents? I have checked the wxWindows docs, but found 
nothing, and I am relatively new at this.

Al.



From lac at strakt.com  Mon Apr 29 12:03:28 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 29 Apr 2002 18:03:28 +0200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...) 
In-Reply-To: Message from "Stephen J. Turnbull" <stephen@xemacs.org> 
   of "29 Apr 2002 23:07:01 +0900." <87adrm38fu.fsf@tleepslib.sk.tsukuba.ac.jp> 
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> <mailman.1020071811.14821.python-list@python.org>  <87adrm38fu.fsf@tleepslib.sk.tsukuba.ac.jp> 
Message-ID: <200204291603.g3TG3SkG027830@ratthing-b246.strakt.com>

> >>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
> 
> > >>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
> > 
> >     Laura> People here seem to be thinking that the GPL is there to
> >     Laura> protect creators rights to their software.
> > 
> > I'm pretty sure that that is what rms has in mind, as long as both
> > "creators" and "software" are inclusive collective nouns.
> > 
> > The consumer protection argument is the OSI line (a special case
> > thereof), and rms don't have much truck with anything so economical
> > and relativist.
> 
>     Laura> 1. Why did you remove my prior line, 'as a thought
>     Laura> experiment?'
> 
> I apologize for the rudeness, none was intended.  I did not, and still
> don't, see that its presence or absence has any effect on the
> semantics of the discussion.  So I took it out.

Apology accepted.  The reason I put this in was because I don't want to
get involved with 'what was RMS thinking' and 'What does RMS believe'.
If we want to find that out, the only way is to ask RMS.  Going to his
writings and attempting to work things out, unless he explicitly says
he believes something is frought with great peril.  Being wrong about
what RMS means when he says something is as easy as falling off a
log.  And I find all discussions of 'what was RMS thinking' to be
more or less pointless.

When I frame this as a thought experiment, then the people who think
that RMS intended a Consumer Protection Proclamation, and the people
who think that if the GPL is a Consumer Protection Proclamation then 
he got one by accident can have discussions on the merit of crafting
a Consumer Protection policy without bringing RMS and his beliefs into
it.

> 
>     Laura> I put it in in order to avoid having this conversation
>     Laura> right now?
> 
> But it didn't.  I don't see how the GPL is about consumer protection.
> 
> Not as "original intent" (you _did_ use the word "intended"), and not
> as a "thought experiment" in social engineering.  I planned to address
> both, but I see I completely missed the mark on the "thought
> experiment" (and further confused the issue by mentioning rms again).
> Again, I'm sorry for creating misunderstanding.
> 
> There are plenty of ways to address the goal of consumer protection,
> many of which might very well be more profitable (eg, the US DOD's
> second source policy) for the developer of the software under license
> than the GPL.  None of these are satisfactory substitutes for the GPL
> (or free software in general) because they violate freedom.  Another
> example: the Aladdin Free Public License has all the benefits of the
> GPL in protecting customers from broken software: you can always hire
> someone to fix your broken Ghostscript.  You just can't redistribute
> the fixed version for money, and thus amortize the cost.
> 
> But you don't want to go there, do you?  rms definitely doesn't.

Me, personally -- of course I want to go there.  But I am leaving the
'get things done with the US legal system' aspects to Americans who 
live there.

> If one is _only_ interested in consumer protection, there are plenty
> of very attractive alternatives to the GPL that could be imagined.  As
> you point out, some might have to be enforced by changes in IP law in
> practice, but most could (in theory) be achieved simply by customers
> insisting on a reasonable contract.  (The OSI line again.  "You don't
> need to abolish IP or to demand the GPL; there are other ways to
> protect yourself.  But you should protect yourself by insisting on
> access to source.  Preferably by using free software, which benefits
> all software users, usually without hurting you.")
> 
>     Laura> 2. I learned this from RMS himself, when he was living in
>     Laura> my house about 15 years ago.  Look up 'hoarder' in his
>     Laura> writings.
> 
> I've read them.  I understand the concept.
> 
> My mother doesn't.  She will never notice how much software has been
> "hoarded" by Microsoft, only how much has been delivered at remarkably
> low prices.  My mother has no need or desire for a copy of the source
> herself.  She recognizes the usefulness of _someone_ having access to
> source, but "second-sourcing" or "source escrow" would be plenty to
> protect that wily consumer in her own opinion.
> 
> My point is that rms developed the GPL to protect/enforce sharing of
> software ideas, not to protect consumers' economic interests in
> working software.  The protection of consumers is a very desirable
> side effect, but according to rms _economic benefits in general must
> never be considered on a par with software freedom_.  The GPL wasn't
> intended to protect consumers, and it is not tuned to doing that
> effectively, since it surely results in less software being developed.
> There's a very good reason for rms's denial of consumer sovereignty:
> it is easy to show under the standard microeconomic models that "a
> little bit of IP in the right places" will surely benefit consumers
> relative to "no IP," and the GPL imposes _more_ frictions than "no IP"
> does.

You are presenting consumer protection as a purely economic issue.  
While I believe that every issue is to some extent an economic issue,
I don't see that consumer protection is primarily about economic benefits.
Is it seen that way in Japan?  My observations of the USA lead me to
believe that it is seen that way there, but clearly not totally or
Bill Gates wouldn't have to argue that consumers were protected,
anti-trust was not an issue, because prices were low.  It wouldn't
be arguable in that case.

> 
> So I don't think an argument starting from consumer protection will
> lead to the GPL, or anything that looks like it, in many important
> cases.  rms certainly doesn't.  He says so every time he's asked to
> comment on "open source," and he's personally asked me not to
> formalize, not to measure, not to mention, and most especially not to
> publish what "is easy to show...".[1]
> 
> Thus my point that the GPL (and software freedom in general) is
> primarily about protecting the freedom of developers to develop.
> 
>     Laura> I thought compelling people was a worse sin than software
>     Laura> hoarding.  [...]  RMS does not see things this way at all.
>     Laura> He thinks that software hoarding is a worse sin than
>     Laura> compelling,
> 
> I've seen him write similar things, also the moral equation of
> intellectual property to human slavery.  But I don't understand what
> he means by this, as I would deduce a need to close the "ASP loophole"
> in the GPL.  Eg with a "must publish `deployed' modifications" clause
> a la the Apple Public Source License.  But he firmly rejects that.[2]

Ok.  I don't speak for RMS but tomorrow, on my holiday, I will begin
to write up why I want a world where the right to close source your
program is greeted with the same horror as the right to sell your
children into slavery.  But I want a new thread for this, and I need
some time.  It is hard to do this one properly.  Since tomorrow and
the day after are holidays here in Sweden, I should probably get
enough time in those 2 days, but no promises.

> 
> By the way, there's a small technical error in your consumer
> protection argument.  You wrote
> 
>     Laura> The thing is you will be able to tell (or hire somebody to
>     Laura> tell you) before you buy
> 
> There's nothing in the GPL that requires that you get source _before_
> you buy the software.  Cygnus and The Kompany both have made money on
> GPLed software for which the source is not public.  It just happens
> that the customers had no interest in redistributing source
> themselves.  In fact, it seems plausible that the GPL was helpful in
> convincing some customers not to redistribute the source, as it very
> likely would have quickly ended up in the hands of a competitor.

This is very true.  I just over-simplified, in the interests of brevity.
But as long as I can return software whose source I have evaluated under
some 'unfit for its stated purpose' clause, or just as part of an
automatic 'right to return after so many days clause', then getting to
see the source before or after is less important.  (And remember, everybody,
I am in Sweden -- a lesson in how the American Consumer is protected 
under American law is of extremely limited use for me.  Professionally,
when I am selling things to Americans, I have a real lawyer who makes
sure that these details are handled correctly.)

> 
> I doubt this is all that important in practice yet---most free
> software seems to be public source as well as open source---but I
> suspect it will grow over time, especially in market segments
> developing for ASP-like applications.[3]

(hah! did you see the kicking that the Kompany received for that!
 <reads footnote> <grins> -- yes, I see you did.)

> Footnotes: 
> [1]  The "standard microeconomic models" are far more applicable than
> most free software advocates would like to believe.  Showing what the
> limits and appropriate modifications are is a project I guess I should
> dust off and get published.

This sort of analysis is part of the talk that Jacob Hall?n intends to
present at EuroPython, tentatively called _Open Source Business Models_.
But with only 35 minutes, I doubt we can show much detailed modelling.
Papers like yours need to get published.  Where do you intend to 
publish this?

> 
> [2]  For reasons I understand, I think, and agree with.  So evidently
> I don't understand the hoarding v. compelling and IP v. slavery ideas.

Ok, as long as you understand that what you are getting is a paper about
why _I_ think hoarding is worse than compelling and that Closed Source
is exploitative.  I am not trying to pass my ideas off as RMS's ideas.
You can show them to him, of course, if you like, but I make no
predictions about whether he will end up saying that he agrees with me
or not.

> [3]  Though it may not.  The Kompany now avoids the GPL where
> possible, as they found that bad PR from not publishing source to
> GPLed software is actually more unpleasant than bad PR from
> distributing software under proprietary licenses.

This may not be the only reason, of course.  Some businesses hear GPL
and run away screaming, in any case.

Laura Creighton



From ykingma at accessforall.nl  Sat Apr 13 18:07:33 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sat, 13 Apr 2002 23:07:33 +0100
Subject: try/except/else flexibility...
References: <slrnabcon3.g5r.mlh@vier.idi.ntnu.no>
Message-ID: <3CB8AC11.758685D1@accessforall.nl>


Magnus Lie Hetland wrote:
> 
> Just a thought: Given the prevalence of try/except/else in the "don't
> look before you leap" way of doing things, could it be useful to make
> the statement more flexible? For instance, instead of writing
> 
>   try: foo + ''
>   except: pass
>   else: bar()
> 
> (assuming I'm sloppy enough not to specify TypeError on the second
> line) one could simply write:
> 
>   try: foo + ''
>   else: bar()
> 
> and if I want something done only if it causes no exceptions, for
> instance pop off an element of a list unless it is empty, I could just
> do:
> 
>   try: somelist.pop()
> 
In the pop() case there is a check that can be done before anything
needs to be done that might not be easily undone, ie. I suppose that

  try: somelist.pop()

is equivalent to:

  if somelist: somelist.pop()

In the general case:

  try: whatever()

a lot may have been done before an exception is evt. thrown
and I don't think there is a way to undo that in python without
a lot of programmer support in try/finally statements.
Databases have commit/rollback, but this requires expensive logging.


Regards,
Ype


P.S.
> Is this heresy?

If you look for a language that supports undoing assignments
try Sartre. Sartre supports unassignment, which might
come in handy. However "Unlike traditional programming languages
(or maybe very much like them), nothing in Sartre is guaranteed
except maybe for the fact that nothing is guaranteed",
so YMWV, ie. your milage will vary:
http://www.catseye.mb.ca/esoteric/sartre/index.html


-- 
email at xs4all.nl


From elian at inbrief.net  Mon Apr 22 15:17:50 2002
From: elian at inbrief.net (Edmund Lian)
Date: Mon, 22 Apr 2002 15:17:50 -0400
Subject: Overriding methods in classes that use property()
Message-ID: <OFA4A5E3D1.8ADFD95C-ON85256BA3.0069E6A4@juscom.com>

Thomas wrote:

>You can achive what you want with a metaclass, see code below.<

I see why people say that metaclasses will make your brain explode!!!

...Edmund.





From kkaustub at hotmail.com  Mon Apr 22 00:10:39 2002
From: kkaustub at hotmail.com (k k)
Date: Mon, 22 Apr 2002 09:40:39 +0530
Subject: help needed :about Blt configuration 
Message-ID: <F157rbK8cp3Yh0lekJo00002291@hotmail.com>

Hi,
I have python 2.2 installed on my machine.I have also tried it on
python 2.1.1. I downloaded Pmw ran the demo the demo ran fine
except of BLTgraph for which it said BLT pakage not installed so i
downloaded one "blt24y-for-tcl83.exe" installed it in
c:\Python22\tcl the installation went through but it is still
giving same error . i am working with windows 2000 please tell me
if any further configuration is required

Thanks
Kaustubh

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





From phd at phd.pp.ru  Fri Apr 12 07:57:40 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 12 Apr 2002 15:57:40 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <lkvgaxhyuu.fsf@pc150.maths.bris.ac.uk>; from mwh@python.net on Fri, Apr 12, 2002 at 11:38:03AM +0000
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk> <mailman.1018608627.2395.python-list@python.org> <lkvgaxhyuu.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020412155740.L12997@phd.pp.ru>

On Fri, Apr 12, 2002 at 11:38:03AM +0000, Michael Hudson wrote:
> >    I'll try. What is worse, Python 2.2 failed the same way :( It seems
> > upgrading gcc from 2.8.0 to 2.95.3 did more harm than good :(((
> 
> Oh dear.  I apologise for encouraging you to break you system...

   Not a big problem, in any case the system was in a fragile state for a
couple of years. I cannot recompile Postgres after 7.0 - 7.1 was dumping
core, too.
   Right now I am compiling Pg 7.2.1. If it will fail in any way - this
will be the last straw. I'll find resources to upgrade that damn system
starting with OS. So, after all, the upgarde did some good by destroying
the too fragile installation :)

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




From phr-n2002a at nightsong.com  Wed Apr  3 19:22:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 16:22:24 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <7xbsd0qpmn.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
>     True and False will properly round-trip through pickling and
>     marshalling; for example pickle.loads(pickle.dumps(True)) will
>     return True, and so will marshal.loads(marshal.dumps(True)).

It occurs to me that the array module should also support the new bool
type (though the letters 'b' and 'B' are already used, so a different
onewould have to be chosen).  Then at least we get an efficient way of
handling bit vectors that we didn't have before.


From fperez528 at yahoo.com  Thu Apr 18 22:58:43 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 18 Apr 2002 20:58:43 -0600
Subject: traceback output to a network connection
References: <a9npmv$en6$05$1@news.t-online.com>
Message-ID: <a9o18d$1va$1@peabody.colorado.edu>

Henning Peters wrote:

> hi NG,
> 
> for debugging a network server i would like to show the parser's traceback
> output on my network client. I figured out that the built-in traceback
> module would be a good choice, but it seems, that the traceback-methods are
> limited to write only to open file handlers.
> 
> is there a known way to relay this stuff to an open network socket
> connection?
> thanks...
> 
> bye,
> henning

define your own traceback handler and assign it to sys.excepthook. Then you 
can do whatever you want in it: print, log, whatever. Look at the code for 
the standard cgitb module for an example which generates html tracebacks 
instead of plain text.

f.


From philh at comuno.freeserve.co.uk  Wed Apr  3 18:18:20 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Thu, 4 Apr 2002 00:18:20 +0100
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017851533.5911.python-list@python.org>
Message-ID: <slrnaan3ds.n0n.philh@comuno.freeserve.co.uk>

On Wed, 3 Apr 2002 11:31:03 -0500 (EST), Kevin Jacobs <jacobs at penguin.theopalgroup.com> wrote:
>On Wed, 3 Apr 2002, Fredrik Lundh wrote:
>> Andrew Koenig wrote:
>> > Specifically: If I want to write a function that answers a yes/no
>> > question, I have lots of possible ways of spelling yes (1, 2, "yes",
>> > and so on) and lots of possible ways of spelling no (0, {}, None,
>> > and so on).  There isn't a single preferred way.
>> 
>> if you study real python code, you'll find that there is.
>
>Having managed 200k+ LOC Python projects with at times 20+ developers, I
>find that there isn't.  At least not without getting out a baseball bat.
>
>The first thing many of our programmers want to do at the top of their
>modules is to define:
>
>False = 0
>True = not False

If the proposal was for the equivalent of these declarations to be 
built into python, i would have no objections to it.

But the proposal has the possibilities of messing up serialisation 
between different versions, thus creating incompatibilities.

IMO Python should look towards stability between versions, and not 
introduce changes that cause backward incompatibilities unless there 
is a *very* good reason. And IMO "It'll make the code look prettier" 
*isn't* a good reason.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From ophirb at hotmail.com  Tue Apr 16 05:50:13 2002
From: ophirb at hotmail.com (ophir)
Date: 16 Apr 2002 02:50:13 -0700
Subject: installation probs on SGI: _socket package removed.
References: <6a3d3529.0204151211.7f38f354@posting.google.com> <a9fghs$rfu3t$1@fido.engr.sgi.com>
Message-ID: <6a3d3529.0204160150.50924ca5@posting.google.com>

offer at sgi.com (Nhi Vanye) wrote in message news:<a9fghs$rfu3t$1 at fido.engr.sgi.com>...
> $ from ophirb at hotmail.com -#194894 | sed "1,$s/^/> /"
> >
> >
> >When building python on SGI IRIX646 I've notice that although the
> >_socket library compiled and linked ok, the 'make' process failed to
> >import _socket and removed it.
> >Following the advice I found here at the newsgroup I attempted to
> >manually compile and link _socket.so, then tried importing it from
> >python.
> >The results: An exception was thrown saying the symbol __eprintf was
> >not defined in the shared object.
> >Has anyone encountered this problem and have any help on this?
> >What library is the symbol __eprintf defined in?
> 
> Is this the SGI Freeware distribution or are you building from scratch ?
> 
> richard.

I'm building from scratch, I couldn't find a pre-built version, could
you point me to one?.
Anyway, I managed to compile and link manually by adding -lgcc and
adding the relevant -L... (apparently the up-to-date libgcc in our
system was installed outside /usr/local/lib...).  Any better way to do
this (have the 'make' work somehow)?  I've noticed that many people
reported problems related to the _socket being removed at build time
(often without the installer noticing, so the problem was found only
when trying to use a package the imports _socket).  It seems to me the
procedure I had to go through to discover the source of the problem
and fix it is a bit too complex...
Thanks.


From cliechti at gmx.net  Sun Apr 14 14:46:15 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 14 Apr 2002 20:46:15 +0200
Subject: Jython help pls
References: <3cb9c6c3.115241899@news.mweb.co.za>
Message-ID: <Xns91F0D381895A5cliechtigmxnet@62.2.16.82>

sqrt_2 wrote in news:3cb9c6c3.115241899 at news.mweb.co.za:
> Can anyone help with Jython compiler error messages. I find it
> difficult to understand what the errors are as I have only just
> started with this product.

maybe you want to work through a python tutorial? the one one python.org
is for CPython but 90% or more will apply to Jython too.

> Messages  like: 
> 'NameError: Java'
note that python and java are case sensitive. you probly meant "from java 
import ..." (small 'J')

> What does that mean?? Is there a reasonable site where I can find
> explanations for common error messages? The Jython site FAQ is no help
> at all.

The python docs should help 



-- 
Chris <cliechti at gmx.net>



From phr-n2002a at nightsong.com  Wed Apr 24 11:36:17 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 08:36:17 -0700
Subject: Slight irritation with try/finally indentation
References: <mailman.1019657853.14965.python-list@python.org>
Message-ID: <7x662hm7m6.fsf@ruckus.brouhaha.com>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:
> is the only solution I see, but this is not much appealing either, maybe
> the cure is not worse than the illness.  I am not sure.  Are others on
> this list bothered with this issue?  And if yes, what do you choose to do?

Just live with the unaligned indentation.  The same thing happens
in lisp:

(setup)
(unwind-protect
    (process)
  (cleanup))


From andrewm at object-craft.com.au  Mon Apr 22 21:52:57 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Tue, 23 Apr 2002 11:52:57 +1000
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...) 
In-Reply-To: Your message of "Mon, 15 Apr 2002 17:35:56 GMT."
             <0cEu8.47362$iw.4776263@bin4.nnrp.aus1.giganews.com> 
Message-ID: <20020423015257.10FC438F50@coffee.object-craft.com.au>

>Interersting.  If so, then RMS is 'guilty' of an almost humorous
>contradiction.  The GPL *is* an assertion of property rights in code,
>backed up by FSF lawyers.  It explicity extracts a price from those
>who would use GPL software.  It is not directly monetary, but it is a
>price nonetheless, as evidenced by the fact that many do not choose to
>pay, even though they might be willing to pay a reasonable amount of
>money to get the same software without the non-monetary price
>attached.

Speculation - maybe he's aware of this and simply sees it as playing
what he sees as a fatally flawed system against itself? 8-)

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/




From aahz at pythoncraft.com  Wed Apr  3 17:16:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 17:16:59 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017851533.5911.python-list@python.org>
Message-ID: <a8fv0r$587$1@panix1.panix.com>

In article <mailman.1017851533.5911.python-list at python.org>,
Kevin Jacobs  <jacobs at penguin.theopalgroup.com> wrote:
>On Wed, 3 Apr 2002, Fredrik Lundh wrote:
>> Andrew Koenig wrote:
>>>
>>> Specifically: If I want to write a function that answers a yes/no
>>> question, I have lots of possible ways of spelling yes (1, 2, "yes",
>>> and so on) and lots of possible ways of spelling no (0, {}, None,
>>> and so on).  There isn't a single preferred way.
>> 
>> if you study real python code, you'll find that there is.
>
>Having managed 200k+ LOC Python projects with at times 20+ developers, I
>find that there isn't.  At least not without getting out a baseball bat.
>
>The first thing many of our programmers want to do at the top of their
>modules is to define:
>
>False = 0
>True = not False
>
>  [...]

Here's the tricky question: will you have to wield the bat any less to
get them to stop writing

if x == True:
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From max at alcyone.com  Wed Apr  3 14:31:34 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 11:31:34 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <mailman.1017843494.13847.python-list@python.org>
Message-ID: <3CAB5896.6ACD6AA6@alcyone.com>

Fredrik Lundh wrote:

> Mark McEahern wrote:
> 
> > So what's different?
> 
> more than one way to do it.  more than one concept to
> explain to newcomers.  more than one "obviously correct"
> way to do things.  less is more.  etc.

But there already _is_ "more than one way to do it."  Python has a wide
variety of kinds of objects which are treated as false in truth tests,
this just adds one more whose purpose is indicate an exclusive attempt
to distinguish between true and false (except for all the others which
can take on a wide variety of "true" values).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From ldw at us.ibm.com  Wed Apr  3 12:10:19 2002
From: ldw at us.ibm.com (Larry Whitley)
Date: Wed, 3 Apr 2002 11:10:19 -0600
Subject: Threading problem
References: <a8f8f6$vdu$1@news.rchland.ibm.com>
Message-ID: <a8fd1r$vbu$1@news.rchland.ibm.com>

Nevermind, I see it now.  Funny how you can fail to see something so simple,
come back to it in an hour and there it is.  I should have said:

  theading.Thead.__init__(self, name=thName)

L

"Larry Whitley" <ldw at us.ibm.com> wrote in message
news:a8f8f6$vdu$1 at news.rchland.ibm.com...
> I have a PC with two processors at work and have decided to try to use
> threading to run some long running console applications.  My basic idea is
> to start two threads, one for each processor, and give them a list of
things
> to do.  As each thread finishes it one job will go to the list and pick
off
> the next thing to do.  When the list is empty, both threads are done.
>
> Here's a snippet of the code that inherits from threading.Thread:
>
> class MyThread(threading.Thread):
>     def __init__(self, thName, jobList, jobLock):
>         threading.Thread().__init__(self, name=thName)
>         self.jobList = jobList
>         self.jobLock = jobLock
>
>
> My code is failing in first statement of the constructor, you can see the
> error below.
>
>
> X:\Projects\Pecos\ttc6\thsp>run.py
> Traceback (most recent call last):
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 59,
>     main()
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 46,
>     th1 = MyThread( "Thread1", jobList, jobLock )
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 9, i
>     threading.Thread().__init__(self, name=thName)
>   File "C:\Python22\lib\threading.py", line 355, in __
>     assert group is None, "group argument must be None
> AssertionError: group argument must be None for now
>
> The thing that puzzles me is that threading.Thread() is insisting that
> group=None.  That's ok with me, and I haven't changed it, yet it's
> complaining as if I did.
>
> What am I doing wrong?
>
> Larry
> Running Python 2.2
>
>




From dsavitsk at e-coli.net  Mon Apr  8 00:16:39 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 08 Apr 2002 04:16:39 GMT
Subject: Any reason why PythonWin wouldn't start at all?
References: <3cb0d423$1@news.sihope.com>
Message-ID: <HQ8s8.5400$d7.1424773@newssrv26.news.prodigy.com>

what version of python/win32all are you using?

"Spitfire 1500" <s1500 at sihope.nospam.com> wrote in message
news:3cb0d423$1 at news.sihope.com...
> I'm running Windows 2K, and each time i try to start pythonwin, absolutley
> nothing happens. It shows it as running on the task manager, but I can't
get
> to it whatsoever.
>
> I've reinstalled it several times, killed the reg keys, but nothing seems
to
> do it. It runs just fine on my laptop, which runs WIndows 98.
>
> With the numerous problems of the freezing IDLE, I'd really like to use
> Pythonwin instead. Anyone?
>
> --
> Kids love the rich taste of web content!
> http://british.nerp.net
>
>




From jbublitzNO at SPAMnwinternet.com  Mon Apr 29 10:00:38 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Mon, 29 Apr 2002 14:00:38 GMT
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net> <3CCCD547.6090405@SPAMnwinternet.com> <6x6z8.30735$8D3.846221@news1.tin.it>
Message-ID: <3CCD5216.4080402@SPAMnwinternet.com>

Alex Martelli wrote:
> <posted & mailed>

> Jim wrote:

>>to have sip and PyQt installed, but I assume you do, since PyKDE2 won't
>>build and install without them. Make sure, however, that they all match
>>in version (latest is sip3.1, PyQt3.1, PyKDE2-3.1 - the 3.1 is the *sip*
>>version *not* the KDE version).

> Having KDE 2.2.2, Qt 3.0.3, PyQt 3.1, and Python 2.2.1, I'd like to take
> PyKDE for a spin, but can't figure out what to download from where, and/or
> how to configure it properly.  All my attempts so far (admittedly I haven't
> given it much focuses attention) end up with something like:

> checking for Qt version >= 230 and < 233... configure: error: "incompatible 
> Qt version"

> and indeed I don't have such an old Qt (and can't/won't downgrade).  Is
> there a way to do what I'm looking for -- try some PyKDE on an installation
> with all the "latest stable releases" of the various other tools, as
> listed -- or do I just have to wait for a stable KDE 3 or whatever...?

Ack! I screwed up! It should be sip 3.0, PyQt3.0 and PyKDE2-3.0.
Either way, I don't think KDE2.x.x will work with Qt3.0 (at
least not wrt to PyKDE2 - there are some class changes in the
QTable area), so you'll need to wait for sip 3.2, PyQt 3.2
(available shortly) and PyKDE2-3.2 (available shortly after the
previous two).

PyKDE2 won't work with sip 3.1 - the distributed tarballs are
generated with sip 3.0 and are not compatible with 3.1, and
some other changes (most notably threading) make it more
reasonable just to jump over 3.1 to 3.2 . The versioning is
getting awfully confusing so here's a recap:

   PyKDE2-3.0-x (where x is a KDE version from 2.1.1 to 3.0beta2
   inclusive) needs sip 3.0 and PyQt3.0. All are available on the
   PyKDE download page at http://www.riverbankcomputing.co.uk.
   The 3.0beta2 version will *not* work with KDE3.0 final.

   PyKDE2-3.2 (single tarball supports all KDE versions from 2.1.1 to
   3.0 final release) needs sip 3.2 and PyQt3.2 and all should be
   available in the next week or two.

Jim



From loewis at informatik.hu-berlin.de  Thu Apr 18 07:36:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 13:36:42 +0200
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de> <m3g01um27o.fsf@ferret.object-craft.com.au> <m3ads2dkpy.fsf@mira.informatik.hu-berlin.de> <m38z7lrfh7.fsf@ferret.object-craft.com.au> <m3u1q9y0d6.fsf@mira.informatik.hu-berlin.de> <m3n0w15r8f.fsf@ferret.object-craft.com.au>
Message-ID: <j4wuv56xx1.fsf@informatik.hu-berlin.de>

Dave Cole <djc at object-craft.com.au> writes:

> The assumption is not an assumption in this case because the original
> question was about this very situation.

I see. I did not understand that the OP was already concerned with a
callback from his C function to Python; I thought that he only wanted
to call into C code and wondered why the other threads are blocked.

Regards,
Martin



From duncan at NOSPAMrcp.co.uk  Tue Apr 23 05:55:14 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 23 Apr 2002 09:55:14 +0000 (UTC)
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org>
Message-ID: <Xns91F96D5FC9D06duncanrcpcouk@127.0.0.1>

Graham Ashton <gashton at cmedltd.com> wrote in 
news:mailman.1019552082.8541.python-list at python.org:

> This means that I'd like to be able to detect whether my class is being
> subclassed by a new or classic class. To do that I'm testing self to see
> if it has a __new__ attribute. Is there a better way to do it?

You could test that type(self)==types.InstanceType

A better way to do it would be to write the code so that it works for both 
or only for new style classes (since if you make your base class newstyle 
then all subclasses are automatically newstyle).

> 
> I find that the behaviour varies depending on the order in which a
> class's super classes are defined. In the example below, I'd like the
> instantiation of class B to throw the same RuntimeError as A:
> 
> Python 2.2 (#2, Mar 11 2002, 13:24:00) 
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> from AbstractSingleton import AbstractSingleton
>>>> class A(AbstractSingleton, object):
> ...   pass
> ... 
>>>> a = A()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "AbstractSingleton.py", line 60, in __init__
>     self.__check_for_new_style_class()
>   File "AbstractSingleton.py", line 66, in __check_for_new_style_class
>     raise RuntimeError, "New style classes are not supported"
> RuntimeError: New style classes are not supported
>>>> class B(object, AbstractSingleton):
> ...   pass
> ... 
>>>> b = B()
>>>> 
> 
Perhaps if B was to call AbstractSingleton.__init__ this would do what you 
want, but since only object.__init__ is called here, your checks are never 
done.

You need to make B explicitly call AbstractSingleton here or make 
AbstractSingleton a new style class in which case its __init__ will be 
called.


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From aleax at aleax.it  Tue Apr  2 16:46:57 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 21:46:57 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl>
Message-ID: <lFpq8.44725$pT1.1265790@news1.tin.it>

Ype Kingma wrote:
        ...
> The problem is that a lot of code has been written assuming 1/0 behaviour
> of boolean values.

The "problem" is that behavior is VERY useful.  So, taking it out would
reduce productivity (by a little) as well as making the language harder to
teach (by a little or by a lot -- I used to think "by a little", Laura's 
post is making me thing it may well be "by a lot").


> This Something/Nothing feature is indeed very useful.
> However, it is not part of the python language, it is part of the way in
> which the standard types are implemented.

The Language Reference manual of the Python language disagrees
with you.  E.g. at http://www.python.org/dev/doc/devel/ref/Booleans.html :

"""
In the context of Boolean operations, and also when expressions are
used by control flow statements, the following values are interpreted
as false: None, numeric zero of all types, empty sequences
(strings, tuples and lists), and empty mappings (dictionaries).  All
other values are interpreted as true.
"""

This is the Python LANGUAGE we know and love -- the LANGUAGE
Reference says so, and us Language Lawyers know that.  You want to
change Python, that's one thing.  But don't tell us that this part of the
Language Reference "is not part of the python language".

> if something_there:
>     ...
> else: # nothing_there
>     ...
> 
> A real boolean is required to control the conditional branch.

If you say so.  In this case, since we DO "control the conditional branch"
in the current version of Python, then we HAVE "A real boolean" whatever
that may be (as defined by the language reference, see above).  Every
value in Python is "A real boolean", since any can "control the conditional
branch".


> As I said, it's not in the language, it's in the standard types.
> The millisecond is spent on the implicit derivation of a boolean
> in the readers' head. The  interpreter does the same thing, although
> faster.

There's no delay at all in my mind as I parse "if zebras:" -- there
WOULD be if that had to carry excess verbiage (to get a "real",
"pristine" boolean type -- which fortunately the PEP does not
propose, thanks be).  There's no delay in the minds of students
for which the Python something/nothing concept has 'clicked'.  I
read Laura's point as it being crucial to make that concept click,
and the existence of True/False being a substantial obstacle to
making it click.


> After having my head set straight on this by Martin v. Loewis
> (see the thread 'but not as int subtype') I concluded in favour
> of PEP 285, with the remark that what can in the future
> not be implemented using the special methods of bool and current
> standard types should be deprecated at the same time the PEP
> gets into force. The only example of such code I know of
> is user implementations of __getitem__, ie. objects
> indexable with square brackets, that are actually indexed with bool
> arguments. When all such user code has been fixed, there is nothing in the
> way of a real boolean type that is not an int-in-new-hat.

Do you propose to make your "real boolean type" non-hashable,
so that it can't be used to index into a dictionary?  Can you please
clarify the reason for *THAT*?

If you do NOT want to forbid usage of your newfangled booleans
as keys into dictionaries, then you're at least using imprecise language
in talking about "objects indexable with square brackets", since
dicts are indeed thus indexable (and more generally so are all
mappings).  Besides, would adict[True] be the very same dict entry
as adict[1], or not?  If they aren't to be the same entry, then how do
you propose to transition smoothly from a present in which they are
to a future in which they are not?  If they are, you need hash(True)
to be 1 and True==1 to hold (or else dicts would be seriously slowed
down -- a pythonic nono! -- and user-coded mappings would have
serious problems too), so SOME level of confusability must be kept.

I think the best levels of confusability (interchangeability, polymorphism, 
name it as you will) between concepts are "all" and "none", in general.  If
"all", the concepts are practically coincident even though there may be
some _conceptual_ separation that has no reflection in practice; if "none",
the concepts are utterly separate in practice even though there may be
some _conceptual_ common ground.  Either of these is easy to teach.
Middle-ground is harder to teach, though there may be very good (real,
practical -- or even philosophical) reasons to pay this price in some cases.

I don't think "none" is feasible here, and I much prefer the current status 
quo of "all" anyway.  The PEP has a lot of confusability (thanks to 
inheritance and True==1, perhaps even to str(True) being '1' depending on 
how that's decided); your proposal has very little (but, I think, not 
"none" -- see above).  So there's a conceptual price to pay, as well
as a pragmatic one (low for the PEP, high indeed for your proposal).  And
I _still_ don't see that the claimed benefits would actually amount to a
substatial hill of beans.  I have nothing against beans, particularly in my
soup, but they seem to come too dear here -- for a price I might perhaps
be willing to pay for good steak, if that.  High price, tiny benefit, my 
personal mental scales slant more and more towards a 'nay'.


> In fact this could be seen as a good example of the usefulness of
> object orientation: it allows separation of concerns to the extent
> that the standard types are documented separately from the language.

But that's NOT what the Python Language Reference chooses to do
for "Booleans" -- it documents them right smack in the middle of
the Language Reference itself.  So?


Alex



From shalehperry at attbi.com  Fri Apr  5 17:35:58 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 05 Apr 2002 14:35:58 -0800 (PST)
Subject: compare function in sort function of lists
In-Reply-To: <0C7CA8D8DF75494EB09AB6016990107F011B1982@NOXMAIL.noetixad.com>
Message-ID: <XFMail.20020405143558.shalehperry@attbi.com>

On 05-Apr-2002 Gabe Newcomb wrote:
> Hi all:
>       I'm trying to sort a list numerically, e.g.,  so that 11 comes
> AFTER 2. The documentation says I need to pass sort() a compare
> function...but I can't find an example of this.
> 
> This should be an easy one for you folks :)
> 

>>> l = [1,5,11,2,67,23]
>>> l.sort()
>>> l
[1, 2, 5, 11, 23, 67]
>>> def comp(a,b): # sort in reverse
...   if b < a: return -1
...   elif b == a: return 0
...   else: return 1
... 
>>> l.sort(comp)
>>> l
[67, 23, 11, 5, 2, 1]

p.s. try the tutor at python.org list for simple things like this




From bvdpoel at uniserve.com  Sat Apr 13 17:05:16 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Sat, 13 Apr 2002 14:05:16 -0700
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com> <WW0u8.27978$iw.2540856@bin4.nnrp.aus1.giganews.com>
Message-ID: <3CB89D8C.CC7A0FBA@uniserve.com>

Terry Reedy wrote:
> 
> <bvdpoel at uniserve.com> wrote in message
> news:3CB884C7.EE7E04D at uniserve.com...
> > I have a list like this:
> >
> > [ [0, x, [0, y, [4, n] ...]
> 
> I am assuming that you mean like this:
> 
> [ [0,x], [0,y], [4,n], ...]
> 
> If so, it is *possible* that the built in [].sort(cmpfunc) will work
> for you *if* you supply the optional custom compare function cmpfunc()
> such as:
> 
> def mycompare(a,b): return cmp(a[0],b[0])
> 
> However, read the reference manual for the Python version you are
> using and test.  I believe (but may be wrong) that .sort() is now
> order preserving but was not always.  If not, you will have to find
> one or devise a hack to make the sort keys unique.  I believe that one
> of a[0]<b[0] or a[0]<=b[0] may work, but again check specific
> reference.
> 
> Terry J. Reedy

Thanks all. After posting a question it's always appropriate to think
about it some more:) Of course, my problem was that old thing DATA
STRUCTURES. I changed to a dict of lists using the event offset as the
dict key. That way multiple events at one offset are all stored togther
and I can parse them later in the proper order.

Sorry for wasting your time with my dumb idea! NOTE TO SELF: Think more
before posting.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel


From jeff at ccvcorp.com  Fri Apr 19 14:00:52 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 11:00:52 -0700
Subject: ActivePython vs. Python
References: <mailman.1019201079.9672.python-list@python.org>
Message-ID: <MPG.1729fb33f8af93529896b7@news.nwlink.com>

In article <mailman.1019201079.9672.python-list at python.org>, 
brian at sweetapp.com says...
>
> For Windows, I would certainly choose ActivePython over PythonLabs
> Python, if ActivePython is available for the Python version that I want.

Very much agreed.  I'm quite happy that ActiveState has finally 
released ActivePython 2.2.1 -- now I can finally upgrade all the 
machines here...  :)

And I must say that the Windows Help docs are an incredible boon!

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jeff at ccvcorp.com  Tue Apr  2 14:54:13 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 11:54:13 -0800
Subject: [Newbie] embed c++ in Python
References: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <MPG.1713ac4070a1ea0a98968a@news.nwlink.com>

In article 
<3ca8901c$0$27282$6e365a64 at newsreader02.highway.telekom.at>, 
mstern at aon.at says...
> Hi!
> 
> I'm new to Python and I'm looking for a good tutorial how to embed a C/C++
> function into Python. I found some tutorials but no one describes what I
> need:
> 
> I want to know how to write a function which receives some variables as
> argument changes them and sends the results back to Python. 
[....] 
> Is there a tutorial? Has someone got a simple example for this?

Look in the standard documentation for the section "Extending and 
Embedding" ...

(As a minor point, what you want to do is typically called 
"extending" -- the term "embedding" is used to refer to running a 
Python interpreter as part of a larger C/C++ application.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jkraska at san.rr.com  Thu Apr  4 20:25:47 2002
From: jkraska at san.rr.com (Courageous)
Date: Fri, 05 Apr 2002 01:25:47 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <mailman.1017869530.3706.python-list@python.org> <b5cnaucrm55rpm8kgt0nibnd4lr43neljs@4ax.com> <a8h83p$rqigt$1@ID-135695.news.dfncis.de>
Message-ID: <lmvpauoae9pgdr359v7h1g4hmrcl9rv0g3@4ax.com>

>Did you compare it with JavaCC or SableCC ? I really would like to read your
>comments about that. Any link to an article comparing compiler-compilers is
>welcome as well.

IIRC, both JavaCC and SableCC are LALR(k) compiler technologies.
ANTLR is LALL(k) with predication; likewise, it is an _integrated
suite_ of tools, for lexing, parsing, treewalking, and AST generation.
It also contains special facilities for translation.

As far as I can tell, there simply isn't anything out there like
it. 

C//



From bokr at oz.net  Mon Apr 15 13:19:04 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 17:19:04 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <3CBA64AD.AE4C563D@cosc.canterbury.ac.nz>
Message-ID: <a9f228$kq0$0@216.39.172.122>

On Mon, 15 Apr 2002 17:27:09 +1200, Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

>Bengt Richter wrote:
>> 
>> On thinking about it, I think this is ugly, though probably
>> convenient in the implementation. I.e., looking at the bits
>> of the absolute value is not usually what you really want when
>> you convert to hex, IMO.
>
>The way I see it, if you're doing bit twiddling
>using unified ints/longs, you should always be
>working with *positive* integers, in which case
>the hex representation will always be what
>you want.
>
>The only reason negative integers come into
>the picture now is that setting bit 31 of a
>non-long int has the side effect of making it
>negative. Once ints and longs are unified,
>this presumably will no longer happen.
>
But surely there will still be negative integers,
whether generated by a wart mechanism or simply by
unary minus etc. So I take it you would like to use the
current sign-magnitude long hex representation, with
the L dropped for unification?

Bit-twiddling operations on integers don't make sense in terms
terms of the actual numerical values (except perhaps 0 and 1
cast as bools), so there's really an implicit mapping from the
integers to something else where it can make sense.

One 'something else' is an array of booleans whose logical values
are represented by the 0/1 bits of the integers. You can do
this with negative numbers too. You just have to have some
convention for normalizing the boolean arrays to the same size,
given two integers. One consistent convention would be to
say the bit array extends indefinitely by sign extension, with the
extended sign represented by as many bits as it takes to normalize
the two arrays to the same length. Operations then proceed between
corresponding array elements, and then the result maps back to integer,
with the topmost bit still being interpreted as representing indefinitely
extendeded sign bits.

For this kind of interpretation, I would prefer to have the hex based
on the bit arrays, not sign/magnitude. In particular, the ~ operation
also changes the sign bit, and to me,

 >>> hex(~4)
 '0xfffffffb'

is more informative and useful than
 >>> hex(~4L)
 '-0x5L'

The former just needs a convention for eliminating redundant high
end bits in the representation (which I think have to be assumed
indefinitely extended unless you subclass to a fixed width).
(Perhaps int could become a platform dependent 32 or 64-bit wide
subclass of unified integer? (just an instant idea, not thought through ;-))

Regards,
Bengt Richter


From guido at python.org  Mon Apr  8 13:05:55 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 08 Apr 2002 17:05:55 GMT
Subject: Booleans, integer division, backwards compatibility; where is 
 Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com>
Message-ID: <3CB1CEC0.F324F480@python.org>

> Guido van Rossum wrote:
> > You clearly fall in the most conservative camp.  I just had a long
> > exchange with someone who pleaded strongly for moving forward at a
> > higher pace, breaking more existing code so that the perfect language
> > can be obtained sooner.
> 
Paul Boddie replied:
> Do such people actually do any real work with the language, though?

Yes, my correspondent is a heavy Python user.

> Wouldn't they be better served by a "big bang, super upgrade" some
> time when all aspects of perfection are ready?

No, he explicitly pleaded against that (pointing to MacOS X as a
counterexample).

> For your amusement, perhaps you should insist that anyone voting (*)
> for particular changes must demonstrate competence with the features
> introduced by such changes (and prove that they're a real user of
> them) before being allowed to ask for any more.

I wish we could insist on some kind of competency test before accepting
someone's contributions, but this is Usenet, even bona fide mental
patients have a chance to be taken seriously by *someone* in the audience.
:-)

> Anyway, I'm still using 2.0 for most things because Unicode was an
> important feature that 2.0 provided. As for the other features
> introduced in 2.0 and later, I can't say that I use any of them apart
> from the += operator - I hardly ever even use the -= operator! I would
> also guess that a reasonable number of people who use the language
> still regard these newer features as advanced or magical, especially
> those people who learned the language in the pre-1.6 days.

And there's nothing wrong with that.  At least, += and list comprehensions
don't break any existing code.

> As far as the 2.2 features are concerned, they seem to have a
> reputation in certain circles as being badly-defined, immature and
> "best steered clear of until later".

Huh?  I'm not sure where you heard that.  Alex Martelli himself just
told me that he is using it and considers it stable.  Maybe the people
who say this themselves deserve a reputation as immature and "best steered
clear of"?

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


From skip at pobox.com  Mon Apr 22 18:34:37 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 22 Apr 2002 17:34:37 -0500
Subject: calendar.py doesn't use lists???
In-Reply-To: <9a6d7d9d.0204221405.7d394207@posting.google.com>
References: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <15556.36861.487833.812444@12-248-41-177.client.attbi.com>

    Aaron> Hi guys.  I got an irksome problem.  It's possible that it's
    Aaron> fixed in python 2.2.1, but even if it is, it irks me.

Aaron,

In 2.2 month_abbr, month_name, day_abbr and day_name were changed to become
instances that computed their values on-the-fly instead of static lists.
This allowed them to be locale-sensitive.  It's possible the problem you
tripped over has been fixed.  The initial version that moved away from
static instances was pretty brutal on the strftime() C library function.
It's quite possible that's what your crash is from.  In 2.2.1 Tim Peters
implemented several changes that were meant to address the problem.  Please
download the latest version from

    http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/*checkout*/python/python/dist/src/Lib/calendar.py?rev=HEAD&content-type=text/x-python

and give it a try.  If it still fails, please file a bug report with SF.

    Aaron> Seriously, why did you need to change these sequences from lists
    Aaron> where they worked with my code, to somethingsimilartolists that
    Aaron> broke my code?

It was strictly an internationalization issue.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From Oschler at earthlink.net  Wed Apr  3 00:39:21 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 3 Apr 2002 00:39:21 -0500
Subject: Zope internal db or MySQL?
Message-ID: <Gswq8.20256$0T6.6662133@e3500-atl1.usenetserver.com>

I'm a Zope newbie.  Would it be better for me to initially set up my site
using Zope's internal database, and switch to MySQL later?  Or should I just
bite the bullet and setup MySQL initially.  My database needs won't be very
heavy, maybe a hundred transactions a day.

thx






From hweaver at pinetel.com  Fri Apr 26 20:26:54 2002
From: hweaver at pinetel.com (Hal Weaver)
Date: Sat, 27 Apr 2002 00:26:54 -0000
Subject: Python on Oberon?
References: <3cc9c8d8$1_5@goliath.newsgroups.com>
Message-ID: <slrnacjs5t.16p.hweaver@tp.new>

On Fri, 26 Apr 2002 18:05:04 -0400, Brad Clements <bkc at Murkworks.com> wrote:
> Just wondering if anyone has thought about this possibility..
 
I've thought about it, partly in the context of using the OS called
Native Oberon, and also development of OPython (Python source coded in
Oberon).

Unfortunately, this would be a daunting project, the Oberon user base
has evaporated, and the convenient functionality of Unix and CPython has
won me away from Oberon.  

Hal


From sholden at holdenweb.com  Mon Apr 15 09:01:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 09:01:56 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <edAu8.10842$Ji7.6909@atlpnn01.usenetserver.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
> On Sun, 14 Apr 2002 01:02:17 -0400, Tim Peters <tim.one at comcast.net>
wrote:
> >> ...
[ talk on leaving things out left out :-)]
>
> It is not new features that are the problem, it is new features that
> break existing code.
>
And these would be ... ?

> IMO until Python developers adopt a policy of being very reluctant
> to do this, many IT managers will be wary of using it in anything
> other thsan toy projects, and Python's user base will not grow as
> quickly as (IMO) it should.
>
And your evidence that they have not yet adopted such a policy is ... ?

> >  Quite the
> >contrary, business wants bug fixes for older releases, and they *also*
want
> >new features in old releases, especially in the libraries (like, e.g.,
SSL
> >support for Windows, and support for hot new protocols
>
> Indeed. And the best way to make sure that Python gets support for
> new protocols is to make sure the user base is big enough that
> someone is bound to write an open-source library to support any new
> protocol. And the best way to achieve that is to ensure backwards
> compatibility.
>
Indeed. And the best way to ensure growth in the user base is to stop
broadcasting mis-perceptions of the True (sorry, 1) State of Affairs.

> In summary:
>
>    New features => good
>
>    New features that break existing code => bad
>
While this last assertion is uncontentious (though hardly surprising or
original), you have totally failed to provide any real evidence of code
breakage due to the recent changes in the language. On the other hand,
Andrew Kuchling recently reported on this list (in message
<slrnabba3h.h8h.akuchlin at crystal.mems-exchange.org>) that it took *minimal
effort* to port over 170,000 lines of Python code to 2.2.

So, I don't disagree with your final premise, but I would categorise your
posting and similar messages as the worst kind of FUD: stuff that's actually
put out by Python *supporters*! Andrew's remarks, however, are useful if for
no other reason than they are based on experience rather than supposition.
To quote him directly: "This means that from my point of view Python *has*
been managed for stability, and the job has been done reasonably well."

It's no use going around screaming "the sky is falling, the sky is falling".
What the developers need to know is, just exactly who is it falling on, and
where do they need to prop it up. If their investigations prove that the sky
isn't really falling at all, just drooping a little in the far left corner,
then a lot of time and energy has been wasted by people who could otherwise
have been adding really super new backward-compatible features to your
favorite language. You might be surprised about how much time the developers
have recently spent trying to ensure a better perception of Python's
stability.

In the absence of any indication that things are likely to change, perhaps a
more positive attitude would encourage others to take up Python rather than
shy away from it for completely spurious reasons. The people developing
Python are not idiots, and are just as familiar as we are with the damaging
impact of code breakage due to language design change.

the-sky-bluddywell-isn't-falling-ly y'rs  - steve







From peter at engcorp.com  Sun Apr  7 18:52:41 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 18:52:41 -0400
Subject: purging file buffers - to make new data be written to file
References: <a8qhjp$7go$1@news6.svr.pol.co.uk>
Message-ID: <3CB0CDB9.218301C2@engcorp.com>

ddoc wrote:
> 
> Can I do that from within a python script, or do I have to give a command
> line argument to make Python start without buffers?
> 
> On Linux.  The program is to save a memo in a slightly organised format,
> but I find the file is not written until Python exits.

Can you post a code snippet that reproduces the problem?  Maybe
you just need to close() the file...


From Steven.dot.Suchting. at computer.dot.org  Mon Apr  1 16:32:53 2002
From: Steven.dot.Suchting. at computer.dot.org (StevenViCy)
Date: Mon, 1 Apr 2002 21:32:53 GMT
Subject: How to deal with relational data?
Message-ID: <3CA8D205.3C9E742B@computer.dot.org>

How do we Python programmers typically deal with creating and using
objects in which the source data is stored in a relational data base?

I know and use the standard DB-API to access the data in relational DB,
and I can create the classes that I want to store the record data as
Python objects.  What's got me concerned is that I seem to be going down
the road of having to have a cursor in most of my routines that access
the data in the Python objects, and I'm wondering if there is a way to
avoid that.

In my application, I have objects (or tuples in a table), that reference
1-to-many other objects (tuples in another table), each of which
references 1-to-many other objects (or tuples in either the same or
another table), and so on in a sort of hierarchy of references to
objects of various classes.

One approach (the first that I tried), created very large and
complicated objects that were hard to comprehend, but worse, were
structured so that it was not very easy to use the various classes in
the heirarchy as building blocks when the classes needed to be used or
referenced by other objects.  To create these large objects, it was very
straightforward to just get the cursor and use it to pull the data
together, and then for subsequent operations using the class, I never
had to use the cursor again.  But, as I said, the objects became
unweildly, and some of the lower leaves of class ended up getting copied
(instead of just referenced) by multiple instances of the top level
object.  I couldn't really reuse the references to the lower tier
objects.

The next approach was to define the relationship between the top-level
object and the lower level objects using what were essentially key
identifiers.  I kept the lower level objects more logically separate
from the parent object, intending to eventually create a cache of some
kind to store the lower level objects so that I would not have
duplicates of the lower level objects in memory.  But in my attempts to
create methods for the top-level object that would need to access data
from the lower tier objects, I couldn't figure out a way to do the
access without passing a cursor into the top-level objects methods - and
this is what seems like a waste. It doesn't seem to make for a very
cohesive, standalone object.

Any recommendations?

Thanks in advance,
Steve S.



From phd at phd.pp.ru  Sat Apr 13 07:45:59 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Sat, 13 Apr 2002 15:45:59 +0400
Subject: python raw socket / ping
In-Reply-To: <c9b1b5f8.0204121049.3bc6e949@posting.google.com>; from hak@iki.fi on Fri, Apr 12, 2002 at 11:49:12AM -0700
References: <c9b1b5f8.0204121049.3bc6e949@posting.google.com>
Message-ID: <20020413154558.A25622@phd.pp.ru>

On Fri, Apr 12, 2002 at 11:49:12AM -0700, hankoifi wrote:
> I have been searching for raw socket examples for python.

  http://py.vaults.ca/apyllo.py?find=socket

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




From jeff at ccvcorp.com  Thu Apr  4 13:52:38 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 4 Apr 2002 10:52:38 -0800
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <mailman.1017877754.545.python-list@python.org>
Message-ID: <MPG.171640d03943a93a989696@news.nwlink.com>

In article <mailman.1017877754.545.python-list at python.org>, 
tdelaney at avaya.com says...

> # only generated in Python 2.3 or higher

I think that warnings should be generated regardless of the 
version of Python being used.  Odds are good that code being 
written for 2.2 or 2.1 will be run under 2.3 someday.  Might as 
well code for that possibility now. 

-- 

Jeff Shannon
Technician/Programmer
Credit International


From frithiof.jensen at removethis.ted.ericsson.dk  Thu Apr  4 03:50:16 2002
From: frithiof.jensen at removethis.ted.ericsson.dk (Frithiof Andreas Jensen)
Date: Thu, 4 Apr 2002 10:50:16 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <a8eve0$kuo$1@perki.connect.com.au> <3CAB16BA.EEED07DE@sympatico.ca> <j48z84zp5z.fsf@informatik.hu-berlin.de> <3CAB60E7.8BEF40DC@sympatico.ca>
Message-ID: <1017910237.53325@robert>

"vio" <vmilitaru at sympatico.ca> wrote in message
news:3CAB60E7.8BEF40DC at sympatico.ca...
> I basically want to re-use all the o-o logic currently buried into the
java code
> (which is not really Java-specific, rather program-specific). So
recuperating
> "most" of that already coded logic would be the main goal.

Then use a UML tool to capture the java design and re-implement it in
Python - http://www.objectdomain.com/index.html does a UML tool that support
Java and Python, although I have not used it.  (In my opinion, that is the
only sensible way - apart from starting completely from scratch!)

> In my view, the bigger the java code,
> the bigger the appeal for such an automated approach. Because however
imperfect,
> it's still less expensive than a complete rewrite.

I will bet against that - a complete rewrite is often the most effective and
the least expensive way to port a program to another programming language.
Often the requirements have changed too and the design wants refinement etc.
so one is already rewriting to some extent.

> And who wants to reinvent the
> wheel "in another language"?

You  - you want to program a translator when The Job is porting to another
language....

> It all comes down to cost-benefit ratio. The
> automated translation could also be viewed as a good jump-start, upon
which
> breathing programmers could maybe easier improve upon, rather than
starting from
> scratch. In this sense, we're now talking philosophy.

No, thats (PHB) management speaking: Automatically translated code is shite
to look at and even worse to understand, to debug and to extend - and when
you add the requirements that you want to preserve the *design* as well, you
just up the requirements on the translator to that of a human - since the
design/architecture is really only there to make humans able to understand
and work with the code.

The compiler/interpreter do not care.

If you really are not convinced, then run a Fortran program through f2c on a
Linux box and gaze upon the resulting C code. Then try and modify that. If
you want to learn Flex/Bison then get the tools & docs off the net and build
yourself a calculator. You will then appreciate what a huge task a decent
translator is and why the code generated by one is so hackish.




From johnroth at ameritech.net  Mon Apr  8 15:59:59 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 8 Apr 2002 15:59:59 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com> <eppstein-CB6351.08534907042002@news.service.uci.edu> <ub315bi2fkdube@news.supernews.com> <3CB1CA65.ECE0437C@python.org>
Message-ID: <ub3tb6o43pkgdc@news.supernews.com>

"Guido van Rossum" <guido at python.org> wrote in message
news:3CB1CA65.ECE0437C at python.org...
> John Roth wrote:
> > Unfortunately, I don't see a PEP on enums. Maybe I should
> > dust off the rewrite of the set PEP I made that included them.
>
> You could reuse some of the implementation techniques I used
> for bool.  Then bool could later be retrofitted as an enum.
> Like bool, I expect that the key benefit of enums is that
> when you print them you'll get their name instead of their
> value.  Currently I use string literals for that, but it's
> not always optimal, and doesn't work for C APIs.
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)

I'm not certain that's a real good idea (the make bools
an enum part.) The reason is that, conceptually, bools
are an answer to the question: is it or isn't it? Enums
are an enumeration of cases. An instance of an enum
is a specific case from the list; a set based on an enum
is the union of all cases that are present.

We have (hopefully) finished bashing that question into
a pulp; I simply don't see modeling either concept on the
integers as promoting anything but confusion. I'm reasonably
happy with the compromises that had to be made with the
bool implementation - after all, I come from an IBM
mainframe background, and compatibility uber alles!
However, since I don't see the need to retrofit enums
into anything existing, I don't see why we would need
the same level of compatibility hack. I certainly don't
see the need of another upheaval by changing bool a
second time.

This isn't to say that mapping enums and integers can't
be useful, and I intend to carry on with Pascal's ideas
in that direction. I see performing arithmetic on enums
to be a useful subset of enums in general: when you need
it, you need it, but otherwise it's simply useless. For
example, an enum of days of the week can usefully
be set up for arithmetic, but an enum of error codes
probably would have no need for that. However,
a mapping would be useful in both cases for sequence
indexes, and also for the standard representation of
sets as bit strings.

In fact, I intend to extend them a bit
by allowing the enum definition to specify whether the
integer sequence is open or closed (that is, whether
addition off the end causes an exception, or whether
it wraps.)

Also, I'm not certain there are a large number of places
in Python where an enum is appropriate as a return
value. In most of those cases, the appropriate return
technique would be to return an instance of a class,
not an enum. For example, the change to the exception
mechanism that replaced string exceptions with instances
of exception objects. Although string exceptions are
still valid, I don't see anything to be gained by replacing
them with enums.

As a final note, how do I submit an acceptable PEP?
That is, what are the restrictions on formatting it? If I
have to format it myself like the rest of the PEPs and
documentation, it's going to be a while.

John Roth




From tim.one at comcast.net  Thu Apr 11 00:08:02 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 11 Apr 2002 00:08:02 -0400
Subject: Easter Eggs
In-Reply-To: <roy-DBB765.21261310042002@news1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEBMONAA.tim.one@comcast.net>

[Roy Smith]
> Actually, backward compatibility can be harder to achieve than
> you seem to think.  By adding the "this" module, you broke existing
> code which relied on:
>
> try:
>    import this
> except ImportError:
>    do stuff
>
> doing stuff.  Stick that in your PEP and smoke it :-)

It's actually worse than that:  if some poor suck^H^H^H^Huser had a this.py
in one of their own directories, and we suddenly put one under Lib/ before
their directory on the Python search path, they'd pick up a wrong module by
mistake.  I tried to keep this.py out of the maintenance releases for that
reason; indeed, at first this.py was scrupulously kept out of CVS
altogether, and inserted into the pre-conference pre-release tarball and
installers by hand.  Nobody even noticed, let alone complained, so tolerance
for me bitching about this.py vanished.  In fairness, I became an
enthusiastic fan after I realized exactly who it would and would not piss
off <wink>.





From claird at starbase.neosoft.com  Mon Apr 22 13:44:18 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 22 Apr 2002 12:44:18 -0500
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at>
Message-ID: <EC7694DBF1581DFC.5E45D82C67327D60.BDDD4EB8581664F1@lp.airnews.net>

In article <3CC43FEC.EE94FFEB at kfunigraz.ac.at>,
Siegfried Gonzi  <siegfried.gonzi at kfunigraz.ac.at> wrote:
>Lulu of the Lotus-Eaters wrote:
>> 
>> |> mathematics, including the study of algorithms, is a natural science.
>> 
>> Tim Daneliuk <tundra at tundraware.com> wrote previously:
>> |I disagree strongly...  All of mathematics is a formal construct of the
>> |human mind created with the intent of removing ambiguity and enhancing
>> |our ability to describe what we *think* and what we *observe*.
>> 
>> To my mind, Daneliuk present here the strongest argument that
>> mathematics (and therefore CS) *IS* indeed a science.
>
>How do you observe infinity in mathematics?: n+1,n+2,...
>
>> Alex may disagree with me here, but I think my opinion here is quite
>> close to the middle Wittgenstein, or the _Philosohy of Mathematics_ and
>> the _Blue and Brown Books_.  It is also very close to some of Godel's
>> remarks, in his philosophical moods.
>
>Wittgenstein dissmised the pure behavior of mathematicians: proof n, do
>not rest and proof n+1, do not rest and proof n+2,...
>Wittgenstein and Goedel met in the so called "Wiener Kreis" (an exquiste
>circle --on a regular basis-- for intellectuals in Vienna/Austria).
>
>Goedel had direct influence (via v. Neumann) into: I dare to say
>computer science, instead I write "computer engineering".
>
>
>S. Gonzi

I'll say what I think is the same, in somewhat different
language:  do not draw conclusions from mathematicians'
philosophy.  There's reasonably strong evidence that
mathematicians are, in the aggregate, unreliable witnesses
as to the epistemology of their expertise.  They do good
mathematics; they recognize good mathematics; they hold
personal beliefs about what good mathematics is; but they
are FAR more trustworthy about the former (their capacity
to generate quality output) than the latter (philosophically
valuable explanations).
-- 

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


From neuhauser at mail.cz  Fri Apr 26 17:36:54 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Fri, 26 Apr 2002 23:36:54 +0200
Subject: slices: foo[i:j:k] => TypeError
Message-ID: <20020426213654.GY82912@roman.mobil.cz>

Hi there,

I'm playing with Python, and found something that seems to contradict
the description in Language Reference:

roman at roman ~ > python
Python 2.1.1 (#1, Oct 12 2001, 20:13:31) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "copyright", "credits" or "license" for more information.
>>> foo = list ('abcdefghijkl')
>>> foo
['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l']
>>> foo[0]
'a'
>>> foo[0:6]
['a', 'b', 'c', 'd', 'e', 'f']
>>> foo[0:6:2]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: sequence index must be integer
>>> 

What's up? Is that a bug, or am I dumb?

-- 
FreeBSD 4.4-STABLE
11:31PM up 11 days, 12:43, 14 users, load averages: 0.06, 0.12, 0.10




From logiplexsoftware at earthlink.net  Mon Apr  1 13:47:14 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 10:47:14 -0800
Subject: How to close wxGrid located inside wxFrame
In-Reply-To: <3CA89E10.AF2C93B3@vip.fi>
References: <3CA89E10.AF2C93B3@vip.fi>
Message-ID: <20020401104714.04aabd54.logiplexsoftware@earthlink.net>

On Mon, 01 Apr 2002 20:51:12 +0300
Pekka Niiranen wrote:

> Hi,
> 
>     how should I define the OnExit -method in MainFrame -class below
>     so that wxWindows would not leak memory, when "Exit" is selected
>     from the menu. I should somehow close the grid before the frame, but
> how ?

The grid should be automatically destroyed when its parent is.  If for some
reason you do need to manually destroy it, this should work:

    def OnExit(self, event):
        # Close program
	self.pn_grid.Destroy()
        self.Close(true)

You'll of course have to change the line where you instantiate pn_grid
from:
        # Create grid
        pn_grid = PnGrid(self)

To:
        # Create grid
        self.pn_grid = PnGrid(self)



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From noah at noah.org  Wed Apr 10 11:02:17 2002
From: noah at noah.org (Noah Spurrier)
Date: Wed, 10 Apr 2002 08:02:17 -0700
Subject: Create Excel Files without COM
References: <3CB44516.60005@thomas-guettler.de> <slrnab8hp2.3r2.stefan.heimann@kunz.ratzer>
Message-ID: <008301c1e0a0$b6b3f240$0200a8c0@saysyou>

CSV is good because Excel does not make you through an import process.
The .CSV extension is registered with Excel when you install office. It
opens up almost
as if it were a regular spreadsheet (although it does give you warnings when
you try to save).
Tab and pipe | delimited files are not quote as good. Excel can read them
easily, but you have
to go through the import process every time... None of this is good if you
want to store
formulas, graphs, or formatted reports. It's good for basic data
interchange. Also I have
found that most naive users can be trained to export Excel into CSV format.

There are some rules for being Excel compatible.
    1. Any fields that contain commas must be quoted with double quotes.
    2. If a field contains a double quote then first you must double quote
        the entire FIELD and then the double quotes
        must be DOUBLE DOUBLE quoted.
            "ABC"  -->  """ABC"""
            AB"C"D  -->  "AB""C""D"
    3. If you EXPORT CSV format note that Excel has a weird property or bug
        where it will truncate trailing enpty fields (but not always! I
don't know if there is a rule).
        So a table like this:
            A, B, C, D
            A, B, ,
            A, B, C, D
        May get exported like this:
            A, B, C, D
            A, B
            A, B, C, D

Example.csv:
    A,"B,C","""D""",E,"F,G,""H""",I,"""",J,"""""",K
Will be read into these fields (converted into rows):
    A
    B,C
    "D"
    E
    F,G,"H"
    I
    "
    J
    ""
    K

Yours,
Noah

----- Original Message -----
From: "Stefan Heimann" <stefan.heimann at web.de>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, April 10, 2002 7:07 AM
Subject: Re: Create Excel Files without COM


> Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::
>
> > Hi!
> >
> > Is there something like WriteExcel from perl in python?
> > I need to generate an MS-Excel file. I don't want to use
> > COM although it will run on windows.
> >
> > The data is simple. Just text in columns and rows.
>
> What about CSV?
>
> Bye
>   Stefan
>





From mario at bio58.uni-bielefeld.de  Tue Apr  2 18:09:53 2002
From: mario at bio58.uni-bielefeld.de (Mario Wehbrink)
Date: 2 Apr 2002 18:09:53 MET
Subject: Someone using PyGtk
Message-ID: <3ca9d7d1$1_2@news.uni-bielefeld.de>

Hallo,

i wonder wether someone is using pygtk and if it is nice toolkit to
learn. I already took a look at tkInter, but i did not like it. wxPython
looked good, but it is so damn slow on my machine (PPro 200) even on
very small programs (hello - world).

OK, what i'm looking for is an easy to learn (i'm not very experienced
in gui-programming), quite fast(small) toolkit.

Any suggestions?


Mario

-- 
Mario Wehbrink
wehbrink at web.de


From bgrotanLOVELYSPAM at itea.ntnu.no  Mon Apr 22 07:22:11 2002
From: bgrotanLOVELYSPAM at itea.ntnu.no (=?iso-8859-1?Q?Bj=F8rn?= Ove =?iso-8859-1?Q?Gr=F8tan?=)
Date: Mon, 22 Apr 2002 13:22:11 +0200
Subject: strip/trim a string
References: <3CC3D89E.9115E1A5@itea.ntnu.no> <mailman.1019470179.5340.python-list@python.org>
Message-ID: <3CC3F263.CFC33956@itea.ntnu.no>

Martin Franklin wrote:
> 
> On Monday 22 Apr 2002 9:32 am, Bj?rn OveGr?tan wrote:
> > yes, I'm a newbie at python.
> >
> > and I need to strip/trim a string... e.g. I have a string of unknown
> > length, and only care
> > for the last five(5) characters.
> >
> > An example string:
> > xxxNOGBJ  94873
> >
> > where 94873 is the only part of the string I want to keep
> 
> If it is allways the last five then :
> 
> s="xxxNOGBJ  94873"
> 
> s[-5:]
> 
> at the python int.:
> $ python
> Python 2.0 (#11, Nov  6 2000, 09:01:06)
> [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> s="xxxNOGBJ  94873"
> >>> s[-5]
> '9'
> >>> s[:-5]
> 'xxxNOGBJ  '
> >>> s[-5:]
> '94873'
> >>>

Now this get a bit more complicated...
eksample-data I'm processing:
"76543";"4";"000X???KOST998765;XXXX??ORG
98765;XXXX?NONSET876545;XXXX?SIGNAL76543;";"Some information";"Even more
information";

This is excerpt from python-script:
---------------------------------------------------------
        line = oufil.readline()
        if not line:
                break
        fields = p.parse(line)		# this parses the 5 main fields
        hierarki = p.parse(fields[2])   # this parses the N fields in of
fields[2] which can be everything from 1 to 4 fields
					
        if not fields:
        	continue
----------------------------------------------------------

What I'm struggeling for is this hash-table.. 
12345	Parent1-name
23456	Parent2-name, Parent1-name
34567	Parent3-name, Parent2-name, Parent3-name
45678	Parent4-name, Parent3-name, Parent2-name, Parent1-name

the csv-data is sorted before starting the script...

>From the hash-table I'm going to build ldap-files (ldif) for importing
into an ldap-server
typically:
dn= ou=Parent4-name, ou=Parent3-name, ou=Parent2-name,
ou=Parent1-name,dc=no
    objectClass: top
    objectClass: organizationalunit
ou: Parent4-name

(though I have to build ou's for those with no parent(parent1) first,
then parent2 and so on)

How do you suggest the design of such a system (even in pseudo-code)


----

BOG


From martin at v.loewis.de  Tue Apr 16 02:46:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:46:52 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com> <a9fcut$e9l$0@216.39.172.122> <3cbb8f83@news.mhogaming.com>
Message-ID: <m3sn5wkun7.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> When I print it out in hexadeciaml, I want the bit representation of
> a 'binary twos-complement' number, of a size (in bytes) that will
> represent the number without truncation (regardless of HOW the
> number is stored inside the machine.)

That is impossible to implement. A negative number has an infinite
amount of leading ones, so you *have* to truncate if you want a finite
hex representation.

> Let me clarify the above-- the '>>' should behave in a manner such
> that the sign bit is COPIED to the high bit during a right shift.

An integer does not have a high bit.

> So-- if the number starts out as a positive number, it STAYS a
> positive number after the shift.  The proposed '>>>' operator forces
> a zero into the high bit

Again, an integer does not have a high bit - it is infinitely many
bits. Of course, the internal representation is finite - but we agreed
to ignore the internal representation.

> # b is now a long, equal to: 0x008642130c, or the
> # equivalent of 5 bytes (no matter HOW it is
> # represented internally.)  This is because we need
> # the extra byte to convey the sign.  

If you think that numbers are conceptually represented in two's
complement, we need infinitely many bytes to represent this number,
not just five. Which number would 0x08642130c represent? 0x8642130c?

> # THIS is what _I_ expect--  the number of "bytes"
> # (or at least the equivalent) is still the same,
> # and we can see that the number is now negative,

Why do you think that abstraction should use units of 8 bits, if we
agree that the internal representation should not matter? Why not
units of, say, 11 bits? 16?

> This is how _I_ believe things should work.

It can't work meaningfully that way.

Regards,
Martin


From mgilfix at eecs.tufts.edu  Mon Apr 29 16:36:53 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 29 Apr 2002 16:36:53 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <XFMail.20020429124231.shalehperry@attbi.com>; from shalehperry@attbi.com on Mon, Apr 29, 2002 at 12:42:31PM -0700
References: <aak1uq$bgt$1@bob.news.rcn.net> <XFMail.20020429124231.shalehperry@attbi.com>
Message-ID: <20020429163653.F9076@eecs.tufts.edu>

  For an interesting use of apply, check out the way the pygtk
bindings are generated. While it can easily be argued that apply could
be avoided, here's an example where a large code-base will quickly be
broken.

             -- Mike

On Mon, Apr 29 @ 12:42, Sean 'Shaleh' Perry wrote:
> > 4. apply() was repeatedly suggested as another candidate for
> >     deprecation since the * and ** forms make it obsolete

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From claird at starbase.neosoft.com  Sun Apr 21 23:34:40 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Apr 2002 22:34:40 -0500
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019381679.10995.python-list@python.org> <uc5aajf0m2dmc5@news.supernews.com> <roy-F25CD7.09321421042002@oldnews.panix.com>
Message-ID: <F793FAB307B21C67.1A11EEF4EC6BC775.AD956936B336761D@lp.airnews.net>

In article <roy-F25CD7.09321421042002 at oldnews.panix.com>,
Roy Smith  <roy at panix.com> wrote:
			.
			.
			.
>What's truly mind boggling is that those same machines parsed fortran.

Painfully.  We were also extracting teeth without anaesthesia
about that same century.
-- 

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


From rahvee at yaahhoo.com  Wed Apr 10 18:50:13 2002
From: rahvee at yaahhoo.com (Ned Harvey)
Date: Wed, 10 Apr 2002 18:50:13 -0400
Subject: Indenting lines 400..457 help
References: <a929tk$ik2$1@pcls4.std.com> <mailman.1018473981.17738.python-list@python.org>
Message-ID: <a92fji$tro$1@pcls4.std.com>

Damn, you guys are great!

"Geoff Gerrietts" <geoff at gerrietts.net> wrote in message
news:mailman.1018473981.17738.python-list at python.org...
> Quoting Ned Harvey (rahvee at yaahhoo.com):
> > For the life of me, I can't find any easy way to indent (or unindent)
lines
> > numbered 400..457 or whatever.  Does anyone have any good suggestions?
> >
> > I'm usually a vi user, but I'll happily learn a command of some other
> > format.  I've tried
> >     :400,457s/(.*)/   \1/g
> > and a ton of variations of the same idea, but it seems that the
> > search/replace doesn't do what I expect it to.
>
> In vim, the easiest way is to hit V, scroll down a few lines, then hit
> the > key to shift everything right one indent level. If the indent
> level is too large, :set shiftwidth=4. I also recommend :set
> tabstop=8, :set softtabstop=4, :set expandtab.
>
> In oldschool vi, try:
> :400
> 57>>
>
> Another common trick I use is:
>
> :400,457s/^/    /
>
> Best of luck,
> --G.
>
> --
> Geoff Gerrietts             "Overspecialize and you breed in weakness.
> <geoff at gerrietts net>     It's slow death." -- Maj. Motoko Kusanagi
> http://www.gerrietts.net                          (Ghost in the Shell)
>
>




From jon+usenet at unequivocal.co.uk  Mon Apr 29 18:45:36 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 22:45:36 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <aakelj$do6$1@maud.ifi.uio.no>
Message-ID: <slrnacrj8f.n56.jon+usenet@snowy.squish.net>

In article <aakelj$do6$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> I just thought of great idea:  How about using one file and not mix html and
> python code at the same time ?
>
><html>
></html>
> 
><!--PYTHON Code
> -->

You could do that, but why on earth would you want to?


From mlh at vier.idi.ntnu.no  Tue Apr 16 09:55:25 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 16 Apr 2002 13:55:25 +0000 (UTC)
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no>  <s4zu8.20814$vF6.690664@news2.tin.it>
Message-ID: <slrnabobad.8ho.mlh@vier.idi.ntnu.no>

In article <s4zu8.20814$vF6.690664 at news2.tin.it>, Alex Martelli wrote:
>Magnus Lie Hetland wrote:
[snip]
>I think the best name for what's now raw_input might be something
>like getline.

Or perhaps readline (? la file.readline)?

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From phd at phd.pp.ru  Fri Apr 19 10:44:32 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 19 Apr 2002 18:44:32 +0400
Subject: Should exceptions have an __init__ method?
In-Reply-To: <1019227035.1583.5.camel@tp6>; from gashton@cmedltd.com on Fri, Apr 19, 2002 at 03:37:14PM +0100
References: <1019227035.1583.5.camel@tp6>
Message-ID: <20020419184432.G29496@phd.pp.ru>

On Fri, Apr 19, 2002 at 03:37:14PM +0100, Graham Ashton wrote:
> I'm unable to determine the difference between these two exception
> classes though (where class E is essentially empty):
> 
> Python 2.2 (#2, Mar 11 2002, 13:24:00) 
> [GCC 2.95.3 20010315 (release)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> class E(Exception):

   Class Exception already has __init__, so you are used it. You don't have
to write your own __init__.

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




From jeff at ccvcorp.com  Thu Apr 18 13:59:09 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 18 Apr 2002 10:59:09 -0700
Subject: [OT] What is Open Source?
References: <20020418002823.C4068@phd.pp.ru> <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov> <mailman.1019085816.3890.python-list@python.org> <a9lp2o$nfv$1@watserv3.uwaterloo.ca>
Message-ID: <MPG.1728a93cfcd0fc8e9896b1@news.nwlink.com>

In article <a9lp2o$nfv$1 at watserv3.uwaterloo.ca>, 
jcmason at student.math.uwaterloo.ca says...
> In article <mailman.1019085816.3890.python-list at python.org>,
> holger krekel  <pyth at devel.trillke.net> wrote:
> >> As I understand it, the GPL does not give you right to use the software
> >> as you see fit.  You can only use it by linking with other GPL'd
> >> software.
> >
> >cant be true: gpled software under windows wouldnt be possible...
> 
> And yet it happened.

Indeed, it happened because the claimed restriction is *not* 
true.  There are no limits to the usage of GPL software.  (How 
many businesses are using Linux?)  What the GPL limits, is 
derivative works.  You can *use* the software, but when you 
*modify* it, then you come under the restrictions that the GPL 
lays out.  (Particularly, when you modify it and then distribute 
your modified version.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From tim.one at comcast.net  Fri Apr 19 19:29:38 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 19 Apr 2002 19:29:38 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <LNBBLJKPBEHFEDALKOLCMENOOOAA.tim.one@comcast.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKBOPAA.tim.one@comcast.net>

[Tim]
> ...
> (Wise uses a different mechanism to get rid of UNWISE.EXE -- they
> shouldn't be trying to do a redundant at-reboot delete of this
> specific file)

Wise agreed, and quickly built a new version of UNWISE.EXE that doesn't have
this problem.  This leaves one other small problem:  how do we manage to
magically replace a hundred thousand or so vulnerable UNWISE.EXE's already
installed in Python directories on Windows boxes?  It's times like this I'm
glad Python installs secret backdoors on Windows boxes so Guido can control
them from the comfort of home.

it's-typing-in-all-your-names-that's-the-bitch-ly y'rs  - tim





From mkelly2002 at adelphia.net  Tue Apr 16 15:50:51 2002
From: mkelly2002 at adelphia.net (Michael Kelly)
Date: Tue, 16 Apr 2002 19:50:51 GMT
Subject: Win32 Open Source or free .NET integration?
References: <RSJu8.3093$MM2.1226076@news1.news.adelphia.net> <uhemc4ano.fsf@ctwd0143.fitlinxx.com>
Message-ID: <e00pbugs69lk497ga1u2j5oh8bllbcf7a8@4ax.com>

On 15 Apr 2002 22:52:59 -0400, David Bolen <db3l at fitlinxx.com> wrote:

>That actually seems reasonably unlikely at this point.  The "Visual
>Python" project at ActiveState is to tie into Visual Studio for .NET
>(e.g., the IDE) and not tying Python itself into the CLR (Common
>Language Runtime) of .NET.

That's a bummer.  Reading the stuff on their site
"full integration" leads one to believe if you plug
Visual Python in it's plugged in just as good as
C# and that's obviously not the case.  The CLR
part is what I'm interested in.  Right now I play
around with COM stuff on Windows to mix and
match components from various languages
and if this CLR in a sandbox became cross
platform and esp. if it was free it could be
really cool.

>From what I've seen you're freed up from
all those COM low-level headaches for
the most part.


Mike

--

"Only choice is an oxymoron."

--


From jafo at tummy.com  Sun Apr 14 00:57:09 2002
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 13 Apr 2002 22:57:09 -0600
Subject: RELEASED: Python 2.1.3
In-Reply-To: <200204090255.g392tex17317@pcp742651pcs.reston01.va.comcast.net>; from guido@python.org on Mon, Apr 08, 2002 at 10:55:40PM -0400
References: <200204090255.g392tex17317@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <20020413225709.B18327@tummy.com>

On Mon, Apr 08, 2002 at 10:55:40PM -0400, Guido van Rossum wrote:
>Today, April 8th 2002, we've released Python 2.1.3, the latest version

I have recently put up RPMs for 2.1.3, sources and bnaries compiled
against Red Hat 7.2.

See:

   http://www.python.org/2.1.3/rpms.html
or
   ftp://ftp.python.org/pub/python/2.1.3/rpms/

for more information.

Sean
-- 
 Do bad programmers wake up on Christmas morning to find coal in
 their sockets?  -- Sean Reifschneider
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python




From robin at jessikat.fsnet.co.uk  Sat Apr  6 07:43:08 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 6 Apr 2002 13:43:08 +0100
Subject: 2.2 schizo win32 linkage
Message-ID: <naBIBIAc1ur8EwVD@jessikat.fsnet.co.uk>

Once again I am being bitten by the choice of location for DLLs/pyds.

The standard location for win32 extensions is now lib/site-packages, but
win32 has problems when DLLs live apart from things they link to.

As example

I have tcl83.dll & tk83.dll on the path at c:\tcl\bin and c:\Python\DLLs
_tkinter.pyd is in the latter and presumably is built to work with
those.

_imagingtk.pyd lives with PIL under c:\python\lib\site-packages.

If I import _tkinter first I get the Python Dlls, if I import _imagingtk
first I get (by accident of environment) the ones from c:\tcl\bin.

This is a recipe for disaster and is certainly insecure.

Surely it's not beyond the wit of the developers to come up with a
scheme that allows us to put all binary extensions and dlls in one
place.

Previously when I asked that DLLs be the one and only place we were told
that DLLs is reserved for Python. Please decide otherwise, that attitude
seems unhelpful.
-- 
Robin Becker


From johnroth at ameritech.net  Sat Apr  6 10:14:28 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 10:14:28 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>
Message-ID: <uau3rsge16bp1e@news.supernews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:gJDq8.55255$S52.1904212 at news2.tin.it...
> John Roth wrote:
>         ...
> > Maybe my post in response didn't propagate? As I said in that
> > post, the root of her problem is a poorly chosen name for the
> > function. If she paid attention to constructing the name properly,
>
> She adopted the name *Guido* had given to the function right in
> his pep (actually clarifying it a bit from 'isleap' in his PEP
> to 'isLeapYear', more readable, in her post).

Sorry I'm late in responding to both posts - I had a problem
with my newsgroup vendor and I'm just back on.

I thought the original example was very bad, but I didn't
say so at the time because I hadn't figured out what I didn't
like about it.

The original example was something like

number of days in year = 365 + isLeap(year).

To put it bluntly, that's a very confusing hack that depends
on conceptual True being 1 and conceptual False being
zero.

Again, to put it very bluntly, Laura swallowed a camel
while straining at a gnat. She had a great deal to say about
the difficulty of teaching that example, and I agree completely.
It would be very difficult to teach, but the reason is that
the name is simply wrong. The fact that Guido himself
used that as an example doesn't make it right, either.

If she had corrected the name, then her entire basis for
the objection would have evaporated.

John Roth
>
>
> Alex
>




From news at datatailors.xs4all.nl  Thu Apr 11 07:10:05 2002
From: news at datatailors.xs4all.nl (Peter van Kampen)
Date: 11 Apr 2002 11:10:05 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr>
Message-ID: <slrnabarp0.sn7.news@datatailors.com>

Ante,

>>>print "a="+str(a)
a=1

Hth,

PterK

In article <a93oqk$jf6$1 at bagan.srce.hr>, Ante Bagaric wrote: 

>Is there a way to prevent the print statement to automatically add a space
>between two objects?
>
>>>>print "a=",a
>a= 1
>
>how to make that cursed blank character dissappear?
>Other than print "a=%d" % a because it wouldnt work for iteration of prints..
>
>for i in range(5):
>    print i,
>	
>gives 0, 1, 2, 3, 4 and I want 0,1,2,3,4
>
>Yes, I know it's possible to circumvent this behaviour, but I'm curiuous
>whether it can be done within the print statement alone.
>
>
>


From stefanfaltz at web.de  Tue Apr  2 09:22:28 2002
From: stefanfaltz at web.de (Stefan Faltz)
Date: 2 Apr 2002 06:22:28 -0800
Subject: User-Agent, version and other mysteries
Message-ID: <fce95053.0204020622.3c1f2f27@posting.google.com>

I try to change the "User-Agent" header, i tried it first by something
like

urllib.FancyURLopener.version = "Whatever Agent you like"

that didn't work, what a surprise

then i ripped some code, and i think it is reasonable

class AppURLopener(urllib.FancyURLopener):
    def __init__(self, *args):
        apply(urllib.FancyURLopener.__init__, (self,) + args)
        self.version = "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT
5.0; T312461; Q312461)"

urllib._urlopener = AppURLopener

agent=urllib.FancyURLopener()
or
agent=AppURLopener()
or
agent=urllib._urlopener
even
agent=urllib.AppURLopener()

and now i am lost, changing the code upside/down
i was thinking of doing a
urllib.addheader('User-Agent','whateveragent')
but that can't be the right solution, that is just tooo, i don't like
that idea
or changing the Version string right in the urllib, but that is even
worse

maybe anyone has solved the problem allready, i was looking for any
readme or tutorial, but wasn't able to find one


From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 16:45:22 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 20:45:22 +0000 (UTC)
Subject: How to avoid mutual imports
References: <mailman.1018638085.8824.python-list@python.org>
Message-ID: <slrnabehr2.mfc.huaiyu@gauss.almadan.ibm.com>

Brian Quinlan <brian at sweetapp.com> wrote:
>Huaiyu wrote:
>> I need advice on how to handle the following situation:
>
>[long description snipped]
>
>If interface.py only defines data_initialization, then you could move
>your imports into it i.e.
>
>def data_initialization(...):
> 	from utils1 import a, b, c
> 	from utils2 import x, y, z
>
>	# do real work

This, plus a little rearrangement elsewhere, does the trick.  Thank you.

It is a little bit tedious, though, since the data_initialization is in fact
many separate functions, and the import has to be repeated.  The good thing
is that the imports are not all the same, so the new arrangement makes clear
where each one is needed.

>You could also move data_initialization into its own module (might be a
>good idea anyway).

Well, that has the same problem if I want to use the data in the testing,
unless testing is also in a separate module.

Huaiyu


From fredp at mygale.org.nospam  Thu Apr 11 03:24:26 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 11 Apr 2002 07:24:26 GMT
Subject: Baffled by CGI cross-browser issues... help sought :)
References: <Xns91ECB3C2132FBPaCmAnRDLM@161.48.128.20> <a92b7c$lmb$0@216.39.172.122>
Message-ID: <Xns91ED60234E2A0PaCmAnRDLM@161.48.128.20>

bokr at oz.net (Bengt Richter) said :

> On 10 Apr 2002 15:37:38 GMT, Fred Pacquier <fredp at mygale.org.nospam>
> wrote: [... cgi problems ...]
> 
>>* on Netscape 4.x things sort of work, except that each time I submit
>>a form I get an error message like "A network error occurred while
>>receiving data (connection reset by peer)". After dismissing the
>>dialog the resulting page usually displays correctly (although
>>sometimes truncated). 
>>
> Possibly your CGI output is set up as buffered and not getting
> flushed? 

Could be. As Chris Liechti suggested, I have verified that the problem only 
happens with the POST method. If I use GET with the exact same forms, 
everything works great with all browsers. He also confirms that the POST 
versions run normally under his version of Opera, so the issue seems to be 
only with POST _and_ Netscape/MSIE. I'll need to look into the POST handler 
of CGIHTTPServer .py and see if I can understand what's going on there :)

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


From logiplexsoftware at earthlink.net  Tue Apr 16 17:20:26 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 16 Apr 2002 14:20:26 -0700
Subject: O_O <- Re: Magic words
In-Reply-To: <3CBC94B0.58505F48@prescod.net>
References: <45e6545c.0204120147.3707d410@posting.google.com>
	<Xns91EE1FA914746RASXnewsDFE1@130.133.1.4>
	<45e6545c.0204130449.d6cfd69@posting.google.com>
	<Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4>
	<1104_1018820620@news.dsl-only.net>
	<Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1>
	<Xns91F2905ED8DAARASXnewsDFE1@130.133.1.4>
	<3CBC94B0.58505F48@prescod.net>
Message-ID: <20020416142026.6ba4106b.logiplexsoftware@earthlink.net>

On Tue, 16 Apr 2002 14:16:32 -0700
Paul Prescod wrote:

> Yes, I will probably change the name of that. I would have put that in
> long before minidom was in the Python standard library!

Like nearly 5 minutes before? <wink>

> Philip Swartzleonard wrote:
> > 
> > Duncan Booth || Mon 15 Apr 2002 04:35:13a:
> > > [...]
> > > __paul_prescod_hack_constant__
> > > [...]
> > 
> > Now there's an interesting variable.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From Janssen at rz.uni-frankfurt.de  Mon Apr  1 16:26:51 2002
From: Janssen at rz.uni-frankfurt.de (Michael Janssen)
Date: Mon, 01 Apr 2002 23:26:51 +0200
Subject: long integers
References: <mailman.1017540760.21438.python-list@python.org>
Message-ID: <3ca8d11f@nntp.server.uni-frankfurt.de>


converting long integer to integer strips off the "L":
>>> int(3L)
>>> 3


Michael Hall wrote:

> 
> G'day:
> 
> I'm using Python 1.5.2 and am having trouble with long integers when I
> pull data out of a MySQL database using MySQLdb. When I retrieve a
> numeric value from a field with an `int' data type, the value has an `L'
> stuck on the end.
> 
> In one script, I worked around this by stripping the L off with:
> 
> var = numeric_id # ("2L")
> var = var[:-1]
> print var
>>> 2
> 
> In this case, numeric_id was passed through a hidden field in an HTML
> form. But this doesn't work in another script for some reason, where
> the value is passed straight from a database query. I tried
> converting the data type from number to string, but that doesn't work
> either:
> 
> var = result[0] # ("2L")
> var = 'var[:-1]'
> print var
>>> var[:-1]
[snip]



From robin at jessikat.fsnet.co.uk  Tue Apr 16 15:21:29 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 16 Apr 2002 20:21:29 +0100
Subject: Python 2.2 unwise.exe
References: <GSYATKAqUGv8EwRX@jessikat.demon.co.uk>
 <mailman.1018983273.20776.python-list@python.org>
Message-ID: <0j8cfLA5mHv8EwF5@jessikat.fsnet.co.uk>

In article <mailman.1018983273.20776.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
>[Robin Becker]
>> On the office machine well I installed Python 2.2.1 last Friday evening.
>> My unwise.exe is now gone as of Tuesday 10:00 am.
>>
>> I had a hard crash Monday morning.
>
>Did you have a previous version of 2.2 on your machine, and uninstall it
>before installing 2.2.1?
>

sadly no. On the office Win2k machine I had 2.1.2 previously. On the
home Win2k machine I had 2.2 and discovered that unwise.exe was missing
on Saturday when I tried to uninstall in order to reinstall with 2.2.1. 

>On a clean Win2K Pro box, I'm finally able to provoke this (or at least one
>of its causes) reproducibly, via:
>
>1. Boot the box.
>
>2. Install 2.2.1.
>
>3. Uninstall 2.2.1.
>
>4. Install 2.2.1 again.  Note that \Python22\UNWISE.EXE is there.
>
>5. Reboot the box (just logging off and on again isn't enough).
>
>6. Now \Python22\UNWISE.EXE no longer exists.
>
>
>

-- 
Robin Becker


From sholden at holdenweb.com  Fri Apr 19 19:56:33 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 19 Apr 2002 19:56:33 -0400
Subject: The Python Manual Index
References: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de> <uc0d5d68i0vhf1@corp.supernews.com> <Xns91F5997F18E9ARASXnewsDFE1@130.133.1.4>
Message-ID: <nb2w8.4665$T%5.1737@atlpnn01.usenetserver.com>

"Philip Swartzleonard" <starx at pacbell.net> wrote in message
news:Xns91F5997F18E9ARASXnewsDFE1 at 130.133.1.4...
>
[Thomas Heller's new search interface]
> >>
> DeepBleu || Fri 19 Apr 2002 08:16:23a:
>
> > WOW!
> > Great Job!
>
> Indeed. Something like this needs to be added to the python.org
> interface... i've been wondering for a long time why they don't have a
> document search function =).
>
Well, all they really needed to do was link to

    http://web.pydoc.org/

though there's no denying that Thomas' code is effective and useful.


regards
 Steve






From marklists at mceahern.com  Tue Apr  9 10:33:16 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 9 Apr 2002 09:33:16 -0500
Subject: Beginnger Question
In-Reply-To: <c4a65c5f.0204090624.710228ff@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEAEKKAA.marklists@mceahern.com>

[Dennis]
> My current problem is this:
>
> x = 2
>
> def wrong(num):
>     return num + 1
>
> print wrong(x)
> print wrong(x)
>
> I cannot figure out why this (simplified) program won't increment x
> and produce 3 and 4.  This is probably indicative of a bigger problem
> with understanding functions (and their limitations).  Are there any
> sources out there that explain functions more in-depth and clearly?!
> Thanks for any and all help.

You are not assigning the return value to x, that's why x is unchanged.

Consider:

x = wrong(x)
x = wrong(x)
print x

By the way, even for exercises like this where you're exploring the
language, it's helpful to use names that indicate the intention of the code:

def addOne(num):
	return num + 1

// mark





From ballabio at mac.com  Thu Apr 11 04:42:18 2002
From: ballabio at mac.com (Luigi Ballabio)
Date: Thu, 11 Apr 2002 09:42:18 +0100
Subject: Easter Eggs
In-Reply-To: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
References: <fm2s8kKkXAiR092yn@gnosis.cx>
 <mailman.1018391839.6535.python-list@python.org>
Message-ID: <5.1.0.14.0.20020411093924.00a5eec0@mail.mac.com>

At 12:17 AM 4/11/02 +0100, Robin Becker wrote:
>Am I the only person that finds the whole concept of Easter eggs a slap
>in the face from the developers.
>
>When I find this rubbish in M$oft's expensive software I feel outraged.
>Is it better because it's open source? I think not. On the other hand
>perhaps the Python development process is itself a joke. When one
>questions the idiots in charge one ends in the kill file. No one expects
>the Spanish inquisition, but perhaps if rather trite philosophy is
>hidden in python and we have geniuses for developers, what else may be
>embedded therein?

Robin,
         I'd like to try your coffee brand (ok, so I like it strong. Sue 
me.) Which one is it?

Bye,
         Luigi





From phr-n2002a at nightsong.com  Tue Apr  2 18:37:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 15:37:24 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
Message-ID: <7xr8lx3c5n.fsf@ruckus.brouhaha.com>

David Goodger <goodger at users.sourceforge.net> writes:
> Here's a serious proposal, safe to post now that April Fool's is over.
> Please read & comment.

Man, I thought it was an April fools joke when I saw it last night.
All those `different` :kinds: 'of' "quoting" are much more confusing
than explicit markup.  So I think it's better to use explicit markup.
The ultimate explicit system is tangle/weave but that's a bit extreme.
Javadoc has been very successful in practice and you could do a lot
worse than just copying it lock stock and barrel.

The reStructuredText stuff looks like POD all over again.  If you want
POD, I'd say just use POD, rather than coming up with yet another
incompatible format.

The thing I like least about the PEP is that it's written like
marketing material--it sets out to push a particular solution and
looks for things to bash about the alternatives rather than treating
them evenhandedly and being willing to accept the idea that
reStructuredText may not be the answer.  Javadoc and POD are
widely-deployed, practical systems; a lot of people use them and like
them and there are REASONS for that.  The PEP should acknowledge their
success and see if there's something to be learned from it, rather
than dismissing them so airily.

Finally, among the explicit markup languages examined in the PEP,
Texinfo should be included.  It's much less ugly than html/sgml, and
its markup commands are directly designed for documentation rather
than typesetting.  I think it answers the criticisms made of the other
explicit markup systems mentioned, maybe not totally, but at least
partially.


From dietmar at schwertberger.de  Thu Apr 25 16:23:57 2002
From: dietmar at schwertberger.de (Dietmar Schwertberger)
Date: Thu, 25 Apr 2002 22:23:57 +0200
Subject: Problem with Oracle interfaces and non-ASCII characters
Message-ID: <ant252057bc82%2U@schwertberger.freenet.de>

Hi!

I'm accessing an Oracle database via cx_Oracle, DCOracle2 and odbc from the
win32 extensions.
Unfortunately both native Oracle interfaces don't return non-ASCII
characters.
E.g. I'm getting a '?' instead of the degree symbol '\xb0' or 'A' instead of
the A umlaut '\xc4' (in ISO latin 1 encoding).
The odbc module returns these characters as intended, but I'd like to avoid
my users having to set up ODBC first.

The network traffic with the native interfaces contains the non-ASCII
characters, so the conversion to ASCII happens in either the oracle oci.dll
or the Python interfaces.

The database is version 8i, but I'm not sure about that; client is 8.1.7
and an older version (7.x) for ODBC.
I can't test the native oracle interfaces with the 7.x client as I don't
have binaries for that (Python 2.2 or 2.1/Zope).

Any suggestions?


Regards,

Dietmar



From max at alcyone.com  Tue Apr 16 14:37:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 11:37:28 -0700
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca> <a9hjvc$cdq$1@as201.hinet.hr>
Message-ID: <3CBC6F68.57A8D59F@alcyone.com>

Ante wrote:

> Um.. and shouldn't one read the article before replying?
> 
> If this is solely a scope problem, then why does it work if you
> replace
> 'num += 1' with 'print num'?

Because Python looks for variables differently if a name is being bound
(or rebound) instead of simply referenced.

If you simply reference the name (e.g., print num), the interpreter has
no problem figuring out you mean the global num.  If you bind it (num =
...) however, the interpreter can't know whether you mean the global num
or a new, local num.  So you have to tell the interpeter you mean the
global one with the use of the global keyword (global num).

My advice to people new to these scopes issues is to always use the
global keyword when you're referencing global variables, even when it's
not strictly required.  This helps make code more self-documenting,
makes it clear to others (including yourself at a later time) what you
had intended to do, and means you don't have to modify your code later
should you decide to change a simple reference into a binding later on.

> Shouldn't it say 'referencing local before assignment yada yada yada'
> anyway?

No, because num += 1 is determines to refer to a local variable, and (in
this case) expands to num = num + 1; it's the second num -- the
reference -- that it's complaining about, since it's already determined
that you mean the local binding, not the global one.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From phr-n2002a at nightsong.com  Sun Apr  7 20:39:31 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 07 Apr 2002 17:39:31 -0700
Subject: ACCEPTED: PEP 285
References: <mailman.1018222875.17263.python-list@python.org>
Message-ID: <7x4rinvxa4.fsf@ruckus.brouhaha.com>

"Arthur Siegel" <ajs at ix.netcom.com> writes:
> Has Paul (with whom, to repeat, I only had one exchange with - nasty
> as hell) somehow exhibited less devotion to the "cause"?

I don't remember any exchanges at all with you (I don't mean they
didn't happen, I just mean I forgot them), much less "nasty" ones.
I don't think I've had any really nasty exchanges with anyone on 
this ng.  Maybe I have overly high standards of nastiness :).

> But the fact that Aahz has Paul on the hotseat seems at the moment
> as more of the same kind of stuff that has be off the bus.

I'm on the hotseat?  Somehow I didn't notice.  Shrug.


From siegfried.gonzi at kfunigraz.ac.at  Thu Apr 25 10:12:14 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 25 Apr 2002 16:12:14 +0200
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at> <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com>
Message-ID: <3CC80EBE.E9735284@kfunigraz.ac.at>

Steve Holden wrote:

> Python is usually a reliable consumer of memory, and a reliable collector of
> garbage (although a few pathological cases have been observed). You are
> seeing the behavior you describe because the interpreter is somehow
> retaining pointers to the data structures created during your simulations.
> Without a bit more information about the program structure and your usage
> it's difficult to say what might really be causing your problem.

a) Honestly speaking: I have programmed without any clear structure (the
simulation is more an ad hoc solution, it even includes an external mie
calculation program (written in C) and it uses also some satellite
binary datas): I make heavy use of returning lists with multiple
members: [[122.23,344.9,...],[2.23,4455.34,455.56,...]]. I also use
Numeric but it is not uncommon that not every problem can be solved with
slices alone.

But remark a) is not the problem:

I just start the script from the idle editor: 'Run script'.

I even tried to use del(list1), del(list2),... but it won't help. I also
write the results to a file (and not to the output screen) in order not
to interference the calculation.

After 4 simulations it comes to a crash.


S. Gonzi


From fredp at mygale.org.nospam  Fri Apr  5 05:12:03 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 5 Apr 2002 10:12:03 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it>
Message-ID: <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20>

Alex Martelli <aleax at aleax.it> said :

> Reports of pystones and other Zaurus/SL5000 Python/PyQT experiences
> are VERY welcome -- it's a really sexy little thingy and I'm SORELY
> tempted to have a fling, but getting it in Italy isn't easy as yet...

So am I, since my preferred platform of past years (Psion) is now dead in 
the water. But just yesterday I read a review of the just-released (in the 
US) Zaurus, and the last line mentions that Sharp is "thinking about 
introducing the new Zaurus line in Europe"... next year. So if you want one 
soon you'll probably have to take your chances and order one overseas, and 
make do with the Qwerty keyboard (probably less of a problem for you than 
for us French though :-)

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


From root at [127.0.0.1]  Wed Apr 24 08:38:25 2002
From: root at [127.0.0.1] (nobody)
Date: Wed, 24 Apr 2002 22:38:25 +1000
Subject: Php vs Python
References: <mailman.1019615325.15799.python-list@python.org>
Message-ID: <3cc6a741$0$15474$afc38c87@news.optusnet.com.au>

Personally, I think the PHP language itself is fugly.  Somehow it manages to
inherit the worst of C and Perl, namely the fetish with
$(excessive->punctuation{characters});

However, PHP was made for HTML, is well supported and is easily inlined in
HTML whereas, AFAIK, Python pages have to be entirely Python.

Unless you're writing a major web application, use PHP.





From nospam-readonly at getsoft.com  Mon Apr 22 02:00:02 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 16:00:02 +1000
Subject: A simply newbie question about ndiff
References: <mailman.1019450919.19653.python-list@python.org>
Message-ID: <3cc3a6cd$0$15475$afc38c87@news.optusnet.com.au>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1019450919.19653.python-list at python.org...
> [Neville Franks]
> > One thing I've noticed now is for larger files (73,000 lines)
> > ndiff can take several minutes on my 1G PC, but that is another issue
> > again. I assume the fact that it is being interpreted has quite a bit
> > to do with this.
>
> Not so.  I've been reimplementing ndiff since the early 80's in a variety
of
> languages, and the Python version is by far the fastest, due to using
fancy
> algorithms that were just too painful to code in lower-level languages.
> ndiff is expensive because it tries very hard to produce an "intuitive"
> diff.  If you don't need that kind of diff, there are any number of diff
> variants that run much faster.

Actually my entire reason for getting into Python right now was only to look
at your ndiff code to see the sort of results it delivered and also get a
feel for its space/time usage. I only discovered your ndiff over the
weekend, which is a little surprising as I've also been working on Diff's
for quite a few years now as part of my programmer's editor, ED for Windows.

I've been working on a major rewrite of ED4W over the past 3+ years and I
did look at replacing my diff code with something else which would hopefully
be a bit better (results/time/memory), but didn't find anything that fitted
the bill. I did have a look at fcomp which produced good results quickly for
small files, but managed to use all available memory and time for larger
files. Hence the interest in ndiff.

It will be very interesting to compare the results of ED's Diff and ndiff
with some of the more taxing files I get from time to time.

> ndiff can be incredibly expensive when it's faced with large blocks of
> mismatching lines (meaning the blocks have no lines at all in common).
Then
> it does a full cross-product search, for each line in one block computing
a
> similarity score between it and every line in the other block.  The
> best-matching pair is used as a synch point, and then the whole business
is
> done twice again for the mismatcking sub-blocks on each side of the synch
> pair.  To keep memory use finite, it doesn't even try to remember the
> similarity computations across iterations of this, and does a ton of
> redundant work instead.  In the end, that can take time cubic in the
number
> of mismatching lines.
>
> So, like most tools, it works best when it isn't really needed <wink>.

Yes indeed.

I wonder how it would perform in C++. From what you've said it doesn't sound
like their would be much of performance gain.

FYI on the 73,000 line file I diffed there was only one line different in
the two files. ED's diff took approx. 16 seconds and ndiff I think around
2.5 minutes.

--
Neville Franks, Author of ED for Windows - http://www.getsoft.com




From aahz at pythoncraft.com  Tue Apr  9 18:16:12 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 18:16:12 -0400
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net>
Message-ID: <a8vp7c$3v1$1@panix1.panix.com>

In article <tnys8.11$p56.2333 at newsb.telia.net>,
Fredrik Lundh <fredrik at pythonware.com> wrote:
>Sean 'Shaleh' Perry wrote:
>>
>> " ".join(mylist) # is this perl or python?
>
>*why* string objects have join method has been explained over and
>over and over again, ...

True.

> ...and should also be obvious for anyone with some
>knowledge how strings and sequences work in Python.

Maybe it "should" be, but I like to think I rank reasonbly high on the
ability to reason Pythonically, and I still think that "".join() is a
wart.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From sholden at holdenweb.com  Mon Apr  1 18:38:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 18:38:39 -0500
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <1d6q8.118888$Ou.77410@atlpnn01.usenetserver.com>

"Boudewijn Rempt" <boud at valdyas.org> wrote in message
news:3ca8a8fe$0$45639$e4fe514c at dreader3.news.xs4all.nl...
> Well, I'd like to finally officially announce that my book, _GUI
> Programming with Python using the Qt Toolkit_ is now officially available
> on paper. Whew!
>
> You buy it directly from the publisher's website: http://www.opendocs.org,
>
> or, preferably, at your favourite local or on-line bookshop. For instance,
> Amazon lists it as:
>
>
http://www.amazon.com/exec/obidos/ASIN/0970033044/qid=1017685892/sr=8-1/ref=
sr_8_1_1/103-3219259-3855046
>
> (And I wonder how that's going to come across.) The ISBN is
0-97003300-4-4.
>
> Anyway, I've written about 160.000 words of English prose and code, all
> about Qt, PyQt, sip, Qt Designer and BlackAdder. I cover both Qt 2 and Qt
3
> in a lot of detail.
>
> All source code to the examples is included on a cd-rom, and Opendocs will
> shortly open a webforum where I'll post updated code and where people can
> discuss problems.
>
> You can read the book online at: http://www.opendocspublishing.com/pyqt.
>
Congratulations! I remember your original posts asking for good examples and
choices for the graphics subnsystem. You must feel a weight has been lifted
...

regards
 Steve






From p.magwene at snet.net  Thu Apr 11 00:02:49 2002
From: p.magwene at snet.net (Paul Magwene)
Date: Thu, 11 Apr 2002 04:02:49 GMT
Subject: ANN: disipyl - a 2D and 3D Python and Jython plotting library
Message-ID: <mailman.1018559646.27720.clpa-moderators@python.org>

DESCRIPTION:
-----------
    disipyl is an object-oriented wrapper around the DISLIN plotting
library.  DISLIN is a powerful and flexible multiplatform (Win32, Unix,
Linux, etc.) library designed for displaying scientific data. DISLIN's
author, Helmut Michels, has made available a DISLIN plotting extension for
the Python programming language (see http://www.linmpi.mpg.de/dislin/ for
more details).

    disipyl provides a set of classes which represent various aspects
of DISLIN plots, as well as providing some easy to use classes for
creating commonly used plot formats (e.g. scatter plots, histograms, 3-D
surface plots).  A major goal in designing the library was to facilitate
interactive data exploration and plot creation.

    Documentation, a tutorial, and a demo program are included.  The
    library
has been tested on Win32, Linux, and FreeBSD, but I anticipate that it
should work on any platform which can make use of Python, NumPy, and the
DISLIN python extensions.

    Feedback, comments, and critique are gladly accepted (email:
paul.magwene at yale.edu).



VERSION:
-------
    This is release 0.6 of disipyl.

CHANGES:
-------
New in this release:

* support for plot interaction (record plot coordinates via mouse
clicks)

* new mechanism for quering disipyl objects about the options they support

* cleanup of data/option setting mechanisms for disiypl objects

* Jython support (when DISLIN java library available - most but not all classes)

* updated documentation and tutorial


URL:
----
    You can find disipyl at:
http://pantheon.yale.edu/~pmm34/disipyl.html




From googlegroups at sixtyten.org  Mon Apr 22 10:33:15 2002
From: googlegroups at sixtyten.org (Allan Crooks)
Date: 22 Apr 2002 07:33:15 -0700
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org>
Message-ID: <304d20df.0204220633.50bf23bf@posting.google.com>

> But if you are truly set on using incompatible constructs  
> *everywhere* for the fun of it, you are going to have a problem
> with code duplication :-)

I wouldn't actually mind having code which would do the following:

--------

try:
    import warning
except ImportError:
    warning = None

....

if warning is not None:
    <do something with warning module>

--------

The main problem is when I want to use 'yield', there's no real way of
compensating for a language change... I suppose it is my fault for
trying too much. :) The more I think about it, the more maintaining
concurrent versions seems appealing... at least there is no 'hacking'
involved...

Allan.


From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Apr 30 21:30:16 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Wed, 1 May 2002 03:30:16 +0200
Subject: Importing dynamically loaded modules (in packages)?
References: <aanbio$pk3$1@news1.xs4all.nl> <mailman.1020213066.536.python-list@python.org>
Message-ID: <aangf8$6ev$1@news1.xs4all.nl>

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1020213066.536.python-list at python.org...

> > "import Package.TestModule" still doesn't work: ""no module named
TestModule""
> > How can I correct this?
>
>     sys.modules['Package.TestModule'] = TestModule
>
> after that you should be able to do
>
>     import Package.TestModule

Ouch... isn't that cheating? I mean, the Module name is "TestModule",
not "Package.TestModule".
Anyway I've tried
    import test.pystone
and sys.modules.keys() indeed contains not only "test" but also "test.pystone".
So my guess is, for module "a.b.c.m" we need to add
"a", "a.b", "a.b.c", "a.b.c.m" (etc) to sys.modules.
I still find this a bit weird, but if it works, who cares.

I'll have to try this. Thanks.

Irmen



From jochen at unc.edu  Sat Apr  6 12:26:52 2002
From: jochen at unc.edu (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 06 Apr 2002 12:26:52 -0500
Subject: 2.2.1c2 test_fcntl hangs
Message-ID: <lysn69qrp7.fsf@bock.chem.unc.edu>

With latest cvs release22-maint (April 6, 1am EST) test_fcntl hangs.
(I couldn't find any postings about this.  I think I have seen this
before myself, though...)

,----
| python Lib/test/test_fcntl.py
| struct.pack: '\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00'
| Status from fnctl with O_NONBLOCK:  0

Ok, next morning I press C-c...

| Traceback (most recent call last):
|   File "Lib/test/test_fcntl.py", line 38, in ?
|     rv = fcntl.fcntl(f.fileno(), fcntl.F_SETLKW, lockdata)
| KeyboardInterrupt
`----

This is RedHat-7.0 with all updates + gcc-3.0.4. Python is
,----
| Python 2.2.1 (#2, Apr  6 2002, 00:53:20) 
| [GCC 3.0.4] on linux2
| Type "help", "copyright", "credits" or "license" for more information.
`----

Any ideas?  Please cc me.

Greetings,
Jochen
-- 
University of North Carolina                       phone: +1-919-962-4403
Department of Chemistry                            phone: +1-919-962-1579
Venable Hall CB#3290 (Kenan C148)                    fax: +1-919-843-6041
Chapel Hill, NC 27599, USA                            GnuPG key: 44BCCD8E


From aleax at aleax.it  Tue Apr  2 16:50:04 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 21:50:04 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com>
Message-ID: <gIpq8.44728$pT1.1266378@news1.tin.it>

Jeff Shannon wrote:
        ...
> (FWIW, this was the first time I've seen the Something vs.
> Nothing distinction pointed out explicitly.  I think that this
> explanation of how Python determines truth is exceptionally
> clear, 

Hear, hear.

> and I think that the tutorial should be modified to
> include this terminology.)

Seconded.  _I_ am going to steal it (if I get Laura's permission,
and if the PEP is rejected) for the Nutshell, for sure...


Alex



From rjones at ekit-inc.com  Tue Apr 16 19:27:34 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 17 Apr 2002 09:27:34 +1000
Subject: File locking using fcntl.flock()
Message-ID: <200204170927.34133.rjones@ekit-inc.com>

I can't get any locking to work on Solaris 5.7 using python 2.1.3. On Linux 
2.4.9, I get an exception when I try to re-lock something that's locked. On 
Solaris, I don't get the exception (and without LOCK_NB, I don't block 
either). I've also tried the lockf and posixfile.lock methods to no avail.

>>> import fcntl
>>> f=open('foo', 'w')
>>> fcntl.flock(f.fileno(), fcntl.LOCK_EX)
>>> f2=open('foo', 'w')
>>> fcntl.flock(f2.fileno(), fcntl.LOCK_EX|fcntl.LOCK_NB)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
IOError: [Errno 11] Resource temporarily unavailable
>>>

I just don't get that exception on Solaris... anyone?



    Richard





From mwh at python.net  Wed Apr  3 08:15:21 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 3 Apr 2002 13:15:21 GMT
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
Message-ID: <lkhems53fa.fsf@elios.maths.bris.ac.uk>

Jan Kybic <kybic at ieee.org> writes:

> Hello,
>         this is a post in favour of a faster Python:
> 
> 
> > Or are there any hints that Numric is really way slower than Matlab?
> 
> In my field (signal and image processing) and on high-level tasks,
> Python+Numeric runs at the same speed as the equivalent Matlab code.
> I have one big application that I basically wrote twice for both
> platforms and that's the conclusion I made.
> 
> On the other hand, all the calculation that you cannot express using
> high-level primitives (using matrices, vectors,...) and has to be
> performed with loops element by element, is painfully slow. Equally
> slow in Python, as in IDL, Matlab,... It is so slow that it is
> unusable, it is very difficult to develop an algorithm if it takes
> days to process any reasonably sized input. My solution is to rewrite
> the critical part of the code in C/C++, I did this for Matlab,
> Python+Numeric, as well as IDL, but this is awkward - it forces you to 
> work simultaneously with two languages and to deal besides the algorithm
> itself with a lot of implementation and interface details.
> 
> In fact, it has become such a burden, that I am seriously considering
> switching to another language. The only problem is that I still have
> not found anything high-level, elegant, fast, and free at the same time.

If you're at INRIA, why aren't you using ocaml already? <wink>

> It would be so much easier if I could just compile parts of the
                                        ^^^^
You make it sound so simple...

> Python code into the machine code. This would enable to use Python
> both for the prototyping and for the production code.

Have you looked at psyco?

> One poster said that a true Python compiler is not feasible due to
> Python's dynamic nature. My reponse to this is that I am quite willing 
> to add (optional) typing information on variables to assist the
> compiler to generate an efficient code. I have recently came across
> the language Dylan which permits just that. I think this is a feasible
> way of improving the Python performance.
> 
> I would love to hear your thoughts on the subject. 

Well, there was boundless (it made me exceed my quota on the mail
server I was using ot the time) discussion on the types-sig on
more-or-less this sort of topic a couple of years back, although
perhaps more of the focus was on guaranteeing correctness than
optimization.  You could read that, but you don't want to.

It's possible, perhaps even desirable, that psyco could be extended in
this fashion.  Got a spare man-month or so?  I haven't.

Cheers,
M.


From tanzer at swing.co.at  Thu Apr  4 01:12:54 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 04 Apr 2002 08:12:54 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 09:37:02 CDT."
             <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <E16t0UQ-0007a4-00@tswings.swing.cluster>

> Python is not trying to follow the text books. (In fact, maybe I
> should say "Python is trying not to follow the text books.")

Maybe you could add that gem to the PEP? <wink>

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





From Oschler at earthlink.net  Mon Apr  8 10:42:30 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 8 Apr 2002 10:42:30 -0400
Subject: For SuSE Linux users who want Zope
References: <Dfas8.19447$gA5.2718940@e3500-atl2.usenetserver.com>
Message-ID: <Y_hs8.23162$gA5.2920554@e3500-atl2.usenetserver.com>

Robert Oschler <Oschler at earthlink.net> wrote in message
news:Dfas8.19447$gA5.2718940 at e3500-atl2.usenetserver.com...

<snip>

Please ignore the "Yast2" uninstall problem portion of the message, appended
to the end of the message.  I made a wrong assumption regarding the
"Select/Deselect" button in Yast2.  You have to double-click to uninstall a
package (then you will get the coveted 'd' next to the package name), not
single-click like I was.  Leftover Windoze uninstall module habit plus "late
night fuzzy-eyed blues".

thx






From mattmsykes at yahoo.co.uk  Mon Apr  8 21:18:48 2002
From: mattmsykes at yahoo.co.uk (Matt M Sykes)
Date: 8 Apr 2002 18:18:48 -0700
Subject: Carbonized embedded MacPython
References: <f69ccee4.0204080600.21d8ac87@posting.google.com> <just-984F55.17183508042002@news1.xs4all.nl>
Message-ID: <f69ccee4.0204081718.6869a20@posting.google.com>

Just van Rossum <just at xs4all.nl> wrote
> You're much better off asking such specialized questions on the 
> MacPython mailing list:
>    http://www.python.org/sigs/pythonmac-sig/
> 
> Just
> 
> 
> In article <f69ccee4.0204080600.21d8ac87 at posting.google.com>,
>  mattmsykes at yahoo.co.uk (Matt M Sykes) wrote:
> 
> > Hi,
> > 
> > I would like to embed Python into an already-existing Carbon
> > application.
> > 

Thanks.  Somehow I didn't notice the link to the SIG on the
MacPython page.

--Matt


From ajs at ix.netcom.com  Wed Apr 10 21:50:45 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Wed, 10 Apr 2002 21:50:45 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000101c1e0fb$d5b7f320$0334fea9@carol>

Tim writes -

>Focusing on personalities just doesn't work in this medium.

I'm aware I'm on the "group dynamics" thread, not the
who's an asshole thread.

For you, I'll give Aahz a big group hug if I ever meet him.

And when we're back on the who's an asshole thread I'll volunteer for the
job again.

Guido, putting things to bed, worded well enough to protect the dignity of
those who opposed the PEP:

"...no killer arguments, therefore..."

Aahz thereafter, re-opening wounds, gratuitously insulting
folk who took the time to give their opinions at Guido's request -  and all
with some unmistakable "us guys" undertone.

"Guido, having heard no good arguments against the PEP."

I'm simply suggesting maybe looking at a few more resumes before signing him
for the PR gig, is all.

Just might not be his line of work.

But its your guys call.

Art









From sholden at holdenweb.com  Fri Apr 12 10:42:48 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 12 Apr 2002 10:42:48 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer   division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <wTit8.1381$F43.347@atlpnn01.usenetserver.com> <a95cqa$adr$1@panix1.panix.com> <3CB6494E.E051CBA5@engcorp.com>
Message-ID: <wKEt8.107$rs4.65@atlpnn01.usenetserver.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CB6494E.E051CBA5 at engcorp.com...
> Aahz wrote:
> >
[ ... ]
> > Besides, you never know when a bunch of Usenet posts might turn into a
> > book contract.
>
> Hah!  If I had time to write books, I'd definitely have time to
> contribute to the Python core in some way.
>
> Or, I could stop spending so bloody much time posting to Usenet... :-)
>
> Thanks for the good thoughts Steve.  I guess we contribute some.
>
> On the other hand, I think I could post ten messages a night and
> still not contribute as much value as someone who manages to submit
> a few patches a year to the core.  I really can't emphasize strongly
> enough how much I think we should value those who actually write
> or fix core code.
>
Nah, they're just a bunch of slefish bastards who spend all their time
writing code because that's how they get their jollies. What did the Python
core team ever do for us?

Oh, sorry, wrong thread ;-)

life-of-who-ly y'rs  - steve





From ods at fep.ru  Tue Apr  9 06:18:11 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Tue, 9 Apr 2002 14:18:11 +0400 (MSD)
Subject: CGI multiple select?
In-Reply-To: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>
Message-ID: <Pine.LNX.4.44L.0204091407470.13708-100000@ods.pp.ru>

On Mon, 8 Apr 2002, John wrote:

J> In my html form, I have 'multiple select'
J>
J> <select name=spred multiple>
J> <option..>
J> <option..>
J>
J>
J> This form invokes CGI routine, where I try to read selected
J> entries from
J> form["spred"]
J>
J> My dilemma is here:
J>
J> When only one entry is selected:
J> form["spred"].value (will be the object selected)
J>
J> When multiple entries were selected:
J> for o in form["spred"]:
J>   o.value (--> each of these will be the object selected)
J>
J> Thus, I need to handle according to situation.. I though my
J> previous
J> question on 'isList' can solve this matter, but somehow, all
J> solution
J> suggested (isSequenceType, type, ..) cannot distinguish these
J> two cases.

If you are using apache and set LimitRequestBody to reasonable
small value (for security reasons), the most safe way is to use
getlist() of FieldStorage _always_.  Then you keep yourself form
doing many "if ... else" or "try ... except".

If you need the list of options -- just call it and you always
recieve a list (may be empty).  If there is the only field:

field = form.getlist('field')
if field:
    # use field[0]
else:
    # field is empty

Or just this way (put any default value you want in list below):

field = (form.getlist('field')+[''])[0]





From jdake at worldonline.nl  Mon Apr  1 15:13:47 2002
From: jdake at worldonline.nl (Jacq)
Date: Mon, 01 Apr 2002 22:13:47 +0200
Subject: benchmarking question (from newbie)
Message-ID: <3CA8BF7B.BE99768B@worldonline.nl>

Hello,

I am a student of "language and artificial intelligence".
I would be very grateful if anyone could advise me on a
benchmarking question I have:

What I have is a script that repeatedly calls a machine-learning
program each time with different parameters/algorithms.
What I need is to measure the time the machine-learning-programme
takes running, *independent* of other processes running on the
same machine.
(I have no control over what processes are started up
by other users, and I want to compare the speed of the different
algorithms/parameter-combinations.)

I thought I could use something like this:
    start = time.clock()
    os.system("call to machine-learning program")
    stop = time.clock()
    cpuTime = stop - start
But testing this on my own machine, I get different values for the same
program-call depending on what other processes I have started on my
machine.

I use Python 1.5 on linux (RedHat 6.1)

Thanks,

Jacqueline Dake




From sholden at holdenweb.com  Mon Apr  8 09:30:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 09:30:02 -0400
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <aZgs8.28133$Sw5.18510@atlpnn01.usenetserver.com>

"vio" <vmilitaru at sympatico.ca> wrote in message
news:3CAAC372.792A0F9A at sympatico.ca...
> Greetings,
>
> I need to translate a (rather voluminous 100MB) java program to python.
After a
> few searches on google, apparently the best tools for this task are
Lex-Yacc
> (Flex-Bison on linux). While I am working on the Lex-YACC HOWTO, I would
be very
> interested in hearing of others' successful experiences/feedback at this.
Any
> pointers of interest would similarily be appreciated.
>
A 100MB program? Perhaps you should have posted this two days earlier?

regards
 Steve






From jeff at ccvcorp.com  Thu Apr  4 14:15:01 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 4 Apr 2002 11:15:01 -0800
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com>
Message-ID: <MPG.171645fed640672d989697@news.nwlink.com>

In article <3cabf705.91920985 at mammoth.usenet-access.com>, 
netzapper at magicstar.net says...
> On 4 Apr 2002 00:58:10 GMT, Marco Herrn <herrn at gmx.net> wrote:
> 
> >Hello. I know that it is possible to turn a string into a funtion call by
> >using getattr(). But what I want to achieve is turning a string into an
> >object name. 
> >I would have a method like
> >
> >  foo(self, objname, objtype)
> >
> >Then I would call this method like
> >  
> >  foo(myObj, bar)
> >
> >Then the method foo should create an object with the name "myObj" from the
> >type bar.
> >
> >Is such an thing possible? And if, how?
> 
> I'll probably get blasted for this by Python gurus... but, I've used
> it with some success.  You can do
> 
> exec(myObjc + " = " + value)

You *can* do this, but using exec is asking for trouble someday.  
It's like pointers in C -- can be very useful, but can also cause 
lots of problems when somehow, something that you don't expect 
ends up getting in there (and if you use it much, that *will* 
happen).  There's a reason that C++ introduces (numerous flavors 
of) safe pointers, and new/delete to hide malloc()... and why 
languages like Python hide all forms of memory management.  
Typically, in Python, there are few things that can be done with 
exec/eval(), that can't be done better some other way.  (Not to 
mention that exec/eval() typically is ugly and hard to read, and 
therefore difficult to maintain...)

In the O.P.'s case, I'd ask, why do you want to create a top-
level variable?  I don't know your specific application, so this 
may not be the best solution, but I'd most likely try saving any 
objects created like this into a dictionary, like so:

MyObj = 'MyDesiredName'
class bar:
    pass

objects = {}
objects[MyObj] = bar()

You could also accomplish much the same thing by making objects 
into an empty class, and using setattr():

class ObjectHolder:
    pass

obj = ObjectHolder()
setattr(obj, MyObj, bar())

This requires a little bit more indirection than directly 
creating a variable from the string in MyObj, but it's safer, and 
often more convenient -- if I'm creating variables from strings, 
then I'm probably creating a group of them, and it makes sense to 
keep track of them as a group.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From mlh at vier.idi.ntnu.no  Sun Apr 14 09:26:22 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 14 Apr 2002 13:26:22 +0000 (UTC)
Subject: Checking whether an async* dispatcher is "alive"
References: <slrnabggo4.o80.mlh@vier.idi.ntnu.no>  <Oj5u8.34280$%A.2510727@news1.west.cox.net>
Message-ID: <slrnabj0ru.rf.mlh@vier.idi.ntnu.no>

In article <Oj5u8.34280$%A.2510727 at news1.west.cox.net>, Van Gale wrote:
[snip]
>Hmm, this is a bit tricky and depends some on how your protocol works.

Oh... I was mainly after a way of checking whether I could write to
the socket or not. It seems that exceptions are generated if is has
been closed (fine by me) but that the asyncore dispatcher intercepts
these and just prints a log message. I could override the error
handling methods to pass through the exception (so I could handle it
myself) -- I just though there might be another way.

[snip]
>I presume however, that your protocol is asynchronous over a persistent
>connection since you are implying you want to send data to a client without
>having just recieved a request.

Yes. I have a list of connected clients and would like to remove the
dead ones.

> In this case you really only have 2
>choices: either just trust TCP's built-in reliability, tr-try, and
>keep-alive features or you can implement some kind of heartbeat message.

I don't mind trusting TCP, if that means that I'll get an exception
when trying to send something to a dead (as determined by TCP) socket.
It just seems that I then have to override the exception handling of
asynchat.async_chat. I had just hoped that there would be some
ready-made way of checking this.

>See pg 581 of [1] for a discussion and sample C code for using TCP's ugent
>mode to implement heartbeats.  His implementation has the client send the
>heartbeats and the server raises a signal if it hasn't recieved any
>heartbeats after a specified interval.  I don't know how portable this is,
>or even how easy it would be to implement in Python if you're using async*
>classes.

Yes, this does sound like a very safe way of doing things, but kind of
overkill in my case, I think. (Possible extension, sure...)

>My personal preference (atm) is just trusting TCP and handling errors when
>the connection is broken because you're going to get a send error on the
>heartbeat anyway :)

Yes, that sounds just fine -- but how do I do that?

I've tried something like:

  alive = []
  for session in self.sessions:
      try:
          session.push(message)
          alive.append(session)
      except socket.error: pass
  self.sessions = alive

but since the push() simply logs the error, this doesn't work. If I
override handle_error, I could probably make this work (or?). But are
there other ways of doing it? Something in the API that I've missed
(like Thread.isAlive)?

>If you're implementing something like a server for an
>IM protocol (e.g. Jabber) which needs up to date "presence" information,
>then it should have a heartbeat built-in to the protocol (as does Jabber).

I don't really need that. Actually, I could live with having a bunch
of dead sessions too, I guess. But somehow, it seems that the default
exception handling wreaks havoc somehow (making it difficult to
re-connect or something -- not exactly sure what happens).

>Btw, the Medusa http_server, status_handler, and the monitor_server, are
>probably worth checking out even if your protocol is fully asynchronous.  I
>really like having a port I can connect to with a browser that gives me a
>nice picture of the internal status of the running server, and
>monitor_server takes the ultimate step: it gives you an interactive Python
>session inside your running server.

Yeah, that is cool :)

>Van
>[1] W. Richard Stevens "Unix Network Programming: Networking APIs: Sockets
>and XTI Volume 1"
>

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From maug at poczta.onet.pl  Fri Apr  5 12:31:47 2002
From: maug at poczta.onet.pl (Marek Augustyn)
Date: Fri, 5 Apr 2002 19:31:47 +0200
Subject: Nested scopes hitch
Message-ID: <a8kn23$ikq$1@zeus.polsl.gliwice.pl>

Hi,

Would somebody explain the following behavior (Python 2.2)?

Function nested in class:
(why nested scoping doesn't work?)

class A:
    a = 1
    def f():
        print a
    f()

result:

Traceback (most recent call last):
  File "c.py", line 1, in ?
    class A:
  File "c.py", line 5, in A
    f()
  File "c.py", line 4, in f
    print a
NameError: global name 'a' is not defined

and version with function nested in function:
(nested scoping works perfectly)

def f1():
    a = 1
    def f2():
        print a
    f2()
f1()

result:

1
---------------

A bit of explanation:
I want a singleton class object (no object instances), and a function to do
some initialization when class object is created. Is it (first example) the
proper way to achieve it?

Regards,
August




From mertz at gnosis.cx  Thu Apr 18 16:40:56 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 18 Apr 2002 16:40:56 -0400
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com>
 <I6Fv8.36158$vF6.1093396@news2.tin.it>
Message-ID: <Y9yv8kKkXsic092yn@gnosis.cx>         

Alex Martelli <aleax at aleax.it> wrote previously:
|class FunkyReprMeta(type):
|    def __repr__(cls):
|        return "My wonderful class %s" % cls.__name__
|class FunkyReprObject:
|    __metaclass__ = FunkyReprMeta
|class C(FunkyReprObject):
|    "whatever you want here"

I don't really understand what a metaclass buys you here.  I think I
might be missing something important, since Alex proposes it.

However, it appears that one could simply define the __repr__ method in
the regular class FunkyReprObject, ignore metaclasses altogether, and
wind up with the same results.  What is better about Alex' approach?

Yours, Lulu...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From mcherm at destiny.com  Mon Apr  8 14:33:04 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 08 Apr 2002 14:33:04 -0400
Subject: Is current integer for-loop syntax a wart?
Message-ID: <3CB1E260.7070000@destiny.com>

> PEP 276 (Simple Iterator for ints) and PEP 284 (Integer for-loops), among
> others, are written from the perspective that the current integer for-loop
> syntax
> 
>     for i in range(10):
> 
> is a wart.  Is this true?  That is, where does the Python community stand on
> this issue?

I think it's a wart.

Actually, I have only two complaints. The first is a minor, almost 
conceptual point... just because I want to loop from 0 to 9 doesn't mean 
I should have to allocate memory for the list [0,1,2,3,4,5,6,7,8,9]. The 
objection is more meaningful for longer loops. Of course, it's not a 
very strong objection, because I have options like:
               for i in xrange(10):
... but if I'm a beginner I might not know about that. And sometimes 
using range() is faster than using xrange() (I'm not clear on why). But 
this is really NOT an important objection and the more I consider it the 
less I care.


The second complain is significant (I think). That is the following:
        for i in range(len(mySequence)):
This is something that I use all the time, and more relevently it is 
something that BEGINNERS use all the time. I want Python to be REALLY 
easy for beginners. I argue all the time that Python is better than Java 
even on pure syntax. In java you have to do this all the time:
     Iterator i = myCollection.iterator();
     while( i.hasNext() ) {
         MyClass x = (MyClass) i.next();
         <code using x>
     }
In python, that's as easy as this:
     for x in myCollection:
         <code using x>
or sometimes even this:
     [<code using x> for x in myCollection]
This is a real win for Python... having to type all that boilerplate 
every time I want to go through a list is painful... having syntax which 
makes COMMON things EASY, and UNUSUAL things EXPLICIT is a very 
important feature of Python.

Well, in my opinion, having to write
     for i in range(len(myList)):
begins to smack of too much boilerplate. Frankly, I'd be satisfied with 
option 2 from PEP 212 http://python.sourceforge.net/peps/pep-0212.html), 
making it as simple as:
     for i in indices(myList):
Or I'd be satisfied with PEP 276 
(http://python.sourceforge.net/peps/pep-0276.html), although the fact 
that it would affect contexs other than loops makes me somewhat queasy.

----

It has been suggested that the awkwardness in using range() or xrange() 
for loops which are not half-open, or which have a step size other than 
1 is a problem in need of a solution. I say "Wart?" -- ie, I've never 
found it to be a problem myself. If I get confused while writing it, I 
just find a way to be more explicit... often resorting to a "while" loop 
instead of "for". Never thought it was any big deal.

---

Finally, the BDFL has proposed that release 2.3 avoid introducing new 
syntax. GREAT IDEA. While I do claim to see some warts, I wouldn't say 
that these are in any way a priority! Let's go for solidity first, and 
language design second... because the language design is already excellent!

-- Michael Chermside



Final commen







From aahz at pythoncraft.com  Mon Apr  1 09:29:40 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 09:29:40 -0500
Subject: Most important extensions?
References: <a8614l$dfa$1@panix1.panix.com> <a87gnj$e75$1@panix1.panix.com> <slrnaafsqi.8mf.gerhard@lilith.my-fqdn.de>
Message-ID: <a89qsk$b65$1@panix1.panix.com>

In article <slrnaafsqi.8mf.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>
>This post is the first one coming thru my own nntp-proxy :-)

Cool!  Written in Python?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From mwh at python.net  Mon Apr  1 12:19:11 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 1 Apr 2002 17:19:11 GMT
Subject: __name__ on new-style classes
References: <stcfaugkbh0u65d1euuofhs2b69ej7i34h@4ax.com> <3CA7C873.D1AC908A@python.org>
Message-ID: <lkd6xjtjzl.fsf@elios.maths.bris.ac.uk>

Guido van Rossum <guido at python.org> writes:

> "Gon?alo Rodrigues" wrote:
> > 
> > One can change the name of a classic class but not of a new-style one
> > (see below for example). Is this a bug or a feature, and if the latter
> > is there any reason for the change?
> 
> It's a feature.  The reason for the change is that I've never seen an
> example of a use for it (as opposed to changing an instance's __class__).
> If you have one, let us know.

I've done it, in code that went like:

def wrap(klass):
    class Wrapped(klass):
        ... stuff ...
    Wrapped.__name__ = 'Wrapped(%s.%s)'%(klass.__module__.__name__,
                                         klass.__name)
    return Wrapped

It makes for slightly better error messages.

Cheers,
M.


From joonas at olen.to  Thu Apr 11 15:21:08 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Thu, 11 Apr 2002 22:21:08 +0300
Subject: Communication between a Python process and a Python cgi script
Message-ID: <3CB5E224.4E3DC5BA@olen.to>

I have a Python program started by root that does some magic stuff
every 10 minutes. So the program in an infinite loop of sleep, do stuff,
sleep, do stuff... How can I make a cgi-script to end the process's
sleep phase and make it to "do stuff" immediately. I tried signals, but
it is propably too risky because the process is run as root. Any
suggestions?


From Chris.Barker at noaa.gov  Tue Apr  2 12:18:32 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 09:18:32 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA8E37A.E89CA002@noaa.gov> <7x4rivj5ya.fsf@ruckus.brouhaha.com>
Message-ID: <3CA9E7CF.52D6CD37@noaa.gov>


Paul Rubin wrote:
> > if string.find(...):
> >       do something.
> 
> Are you proposing to start making 0 resolve as true?  If not, returning
> False has the same problem as returning None.

Yes, I am. I realise that there is far too much history in both Computer
Science in general and Python in particular for it to ever happen, but I
do think that that is how it should be in the ideal language. The
concept of True or False is distinct from the concept of a number, and
I'd like to see it stay that way in the language.

Yes, you would be writing a lot of 

if bool(expression):

rather than

if expression:

But I think it would be much more clear and less awkward.

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From tismer at tismer.com  Mon Apr 29 12:40:07 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 29 Apr 2002 18:40:07 +0200
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org> <mailman.1020076925.25849.python-list@python.org> <aajncv$gfj$1@panix2.panix.com>
Message-ID: <3CCD7767.3000008@tismer.com>

Aahz wrote:
> In article <mailman.1020076925.25849.python-list at python.org>,
> Christian Tismer  <tismer at tismer.com> wrote:
> 
>>What we are talking about is a way to support PyZine, while at the same
>>time providing a missing structure for Starship.
> 
> *Is* that what we're talking about?  Maybe you missed it on python-dev,

Could not find it on python-dev, but there was a larger
thread on the crew list that I in fact missed (shame).

> but there's been some recent discussion about resurrecting an active
> Starship membership policy, so I'd bet that people other than myself
> assumed that was your primary thrust.
> 
> If the goal is to push PyZine, I'd prefer to make that a separate issue
> from Starship.

The first idea was to use PyZine subscription as membership
policy. Then I got the idea to do better by creating
a group which serves as a replacement of the PSA role
for Starship. I want to use the created money to
do sponsoring, where PyZine was one of them. But
we can of course do donations to PSF.
While I don't want to make Starship a part of PSF;
instead I think it should be something on its own.

ciao - chris
-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From ping at lfw.org  Wed Apr  3 17:53:55 2002
From: ping at lfw.org (Ka-Ping Yee)
Date: Wed, 3 Apr 2002 16:53:55 -0600 (CST)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <200204031856.g33Iuqr22751@odiug.zope.com>
Message-ID: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org>

Guido van Rossum wrote:
> False and True are numbers [...]

That's exactly where this falls down.

Patrick O'Brien wrote:
> [...] True and False could be dictionary keys separate from 0 and 1.

Guido van Rossum wrote:
> No.  That would break backwards compatibility.  False==0, and True==1;
> everything else follows from that.  (But False is not 0, and True is
> not 1!)

This is a strange distinction to make -- one that is not made in any
other programming language i have ever encountered.  Of course *i* get
it and *you* get it -- but it will be so hard to teach this particular
weirdness that i consider it fatal to the proposal.


-- ?!ng





From nhodgson at bigpond.net.au  Sun Apr  7 01:24:37 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sun, 07 Apr 2002 06:24:37 GMT
Subject: Compatibilty table between BOA, wxPython, Python?
References: <11e94203.0204062147.5d2281b9@posting.google.com>
Message-ID: <FCRr8.4389$vc3.16639@news-server.bigpond.net.au>

Jose Rivera:

> Does somebody knows where is a table to know wich version of wxPython
> is compatible with wich version of Python, and BOA, I  have tried to
> find this with no success.
>
> When I try start BOA using wxPython 2.3.2.1 and python 2.2, I get an
> error.

   What does the error message say? Which version of Boa are you using?

   I am able to run Boa v0.1.0-alpha on wxPython 2.3.2.1 on Python 2.2 on
Windows 2000.

   Neil






From jeff.pitman at brooks.com  Wed Apr 24 12:35:54 2002
From: jeff.pitman at brooks.com (Jeff Pitman)
Date: Thu, 25 Apr 2002 00:35:54 +0800
Subject: urllib2, cookies and redirect.
References: <mailman.1019643164.21771.python-list@python.org>
Message-ID: <aa6mtb$q2t@netnews.hinet.net>

Jean-Marc Hoeltzli wrote:
> any idea ??

I have battled with this problem for a little bit.  I think I have a 
preliminary solution and will help you get what you need to do now instead 
of waiting for the next python version.  Check out this patch submitted to 
sf.net:

http://sourceforge.net/tracker/index.php?func=detail&aid=548197&group_id=5470&atid=305470

The explanation is on the site there.  Give it a try and let me know if it 
works for you.  If it doesn't work, see if you can hack it to make it work 
and send me a patch! :)

Good Luck!

-jeff


From rasnick1 at att.net  Thu Apr 18 12:58:03 2002
From: rasnick1 at att.net (Jack Rasnick)
Date: Thu, 18 Apr 2002 16:58:03 GMT
Subject: Backup Script
References: <i3%u8.29433$Rw2.2216172@bgtnsc05-news.ops.worldnet.att.net>
Message-ID: <vWCv8.32982$Rw2.2487191@bgtnsc05-news.ops.worldnet.att.net>

This message is for Paul Casteels,
Paul, I tried to email you this return but my ISP (att.net) says that it
doesn't deliver to ISPs that send spam mail.  Your ISP is University of
Antwerp?  Anyhow let me know you got this email.  Had you had this problem
before or is it unique to att.net (which is very good at spam filtering).
Sorry I couldn't email you as what you said about NTFS is very interesting.

Jack

Paul,
I'm working on some bash scripts to do this for windows 98 right now.  I
hope to have it working by the end of this weekend.  I'm new to linux and
bash seem too finicky as a language.  So I'm looking for a good general
purpose, but higher-level language.  Guess it might be pretty hard to create
a linux boot floppy with python on it<grin>.  Is perl too large for floppy?
Guess I better figure out how to create a linux boot CD with python.

Anyhow, sorry to hear about linux NTFS problems.  Maybe I should consider
FAT32 instead.  I plan on getting a new pc with WinXP very soon and want a
backup of the clean install ASAP.  On Win98 I use a second minimum install
of Win98 to winzip up the original as a backup.  Works great, but its labor
intensive.  Its saved my butt many a time!

I'll send you a copy for you to look at if you want.  Wouldn't mind seeing
yours also.  Be interesting to see how you found those empty sectors to zero
out (defrag?)

Jack


----- Original Message -----
From: Paul Casteels <Paul.Casteels at ua.ac.be>
To: <rasnick at att.net>
Sent: Thursday, April 18, 2002 2:00 AM
Subject: Re: Backup Script

> Hello,
>
> I am also interested in this.
> The main problem I found is that Linux cannot reliably handle reading
> and writing of NTFS partitions.
> What I do now is, under Windows2000, fill the empty space on my disk
> with zeros, boot Linux, treat the Windows disk as a device : gzip/dd can
> work efficiently now. Simple shell scripts can do, I don't need Python
> for this : I boot Linux from a floppy.
> Can you let me know, if you can come up with something better ?
>
> Best regards,
>
>
> --
> Paul Casteels Paul.Casteels at ua.ac.be Tel: +32.3.8202455
> Fax: +32.3.8202470
> University of Antwerp Dpt.Physics
> Universiteitsplein 1
> B-2610 Wilrijk
> Belgium
>








From tripps81 at yahoo.com  Fri Apr 19 05:36:51 2002
From: tripps81 at yahoo.com (Tripp Scott)
Date: Fri, 19 Apr 2002 16:36:51 +0700
Subject: source code cross reference tool
Message-ID: <5.1.0.14.2.20020419163449.25556a78@202.95.131.154>

does there exist any? or, has anyone used lxr to cross reference 
a python source code tree? it would be cool to have the source 
code in a set of html pages that allows a user to click around 
and jump to function/class definitions, etc.

t





From cstyphoon at yahoo.com.cn  Fri Apr  5 00:32:56 2002
From: cstyphoon at yahoo.com.cn (=?gb2312?q?=B3=DB=20=C0=EE?=)
Date: Fri, 5 Apr 2002 13:32:56 +0800 (CST)
Subject: I like python
Message-ID: <20020405053256.34386.qmail@web15004.mail.bjs.yahoo.com>



























































































































































































_________________________________________________________
Do You Yahoo!? 
???????????????????????
http://cn.worldcup.yahoo.com/




From jblazi at hotmail.com  Tue Apr  9 07:49:59 2002
From: jblazi at hotmail.com (Janos Blazi)
Date: Tue, 9 Apr 2002 13:49:59 +0200
Subject: PyQt question (colors)
Message-ID: <3cb2d4e7_11@news.newsgroups.com>

If I say

    stat_lab1.setBackgroundColor(QColor("DarkGreen"))

and stat_lab is a QLabel, this has the required effect. But if I say

    pb.setBackgroundColor(QColor("DarkGreen"))

and pb is a QProgressBar object, so this has no effect at all. Can anybody
tell me the reason for this? I use PyQt with Qt 2.3 on a MS-Win2000 system.

J.B.





-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From max at alcyone.com  Thu Apr  4 00:53:43 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 21:53:43 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com> <slrnaannkc.6ls.kzaragoza@localhost.localdomain>
Message-ID: <3CABEA67.FD3F75C0@alcyone.com>

"Kris J. Zaragoza" wrote:

> Maybe, but I'm not sure this is strictly necessary, since according to
> the PEP:
> 
>     ("%s" % str(True)) => "True"

Unless the alternate choice is accepted where str(True) == '1'.

> It's not as pretty as having a format specifier for it, but I'd
> imagine it will do for the short term.  Then, of course, you have to
> wonder what exactly
> 
>     ("%d" % True) => ?
> 
> should print.  After all, bools are ints too.  Should this be "1" or
> an error of some sort?

Since, as you say, bools are ints, it only makes sense that "%d" % True
== '1', since int(True) == 1.

> And what if you've done some arithmetic on a
> bool and end up with another value?  What would your proposed %b
> format specifier print for
> 
>    val = int.__new__(bool,42)
>    ("%b" % val) => ?
> 
> I really should stop thinking about these things and just go to
> bed... :-)

Just like %s does an implicit call to str, %b should do an implicit call
to bool.  So this should be just the same as "%b" % bool(val).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From brueckd at tbye.com  Wed Apr  3 12:43:35 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 3 Apr 2002 09:43:35 -0800 (PST)
Subject: How best to write this try/except block?
In-Reply-To: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <Pine.LNX.4.33.0204030936320.5663-100000@ns0.tbye.com>

On 3 Apr 2002, Roy Smith wrote:

> I've got a dictionary d, which may or may not have a value I'm looking
> for.  If it does, I want to do a few things.  If it doesn't, I want to
> do a few other things.  I could write:
>
> try:
>     value = d[key]
>     foo = stuff (value)
> except KeyError:
>     foo = otherStuff (key)
>
> which is fine as long as I can guarantee (hint: I can't) that stuff()
> doesn't raise a KeyError.

Sounds like stuff() is poorly-implemented or you're not using it the
right way. Why would stuff() be letting a KeyError get by uncaught? (seems
odd to me) You have to decide one of three ways:

1 - decide that it doesn't make sense for stuff() to not catch the
KeyError, so fix stuff().

2 - decide that it's ok for stuff to raise a KeyError because you have
some other action to take, so your resulting code would look like:
  try:
    value = d[key]
    try:
      foo = stuff(value)
    except KeyError:
      foo = somethingelse(value)
  except KeyError:
    foo = otherstuff(key)

3 - decide that if stuff raises a KeyError, it means this whole operation
you're doing failed, so leave the code as it is (and let either KeyError
be caught by your one except clause).

-Dave





From pclinch at internet-glue.co.uk  Wed Apr  3 16:42:51 2002
From: pclinch at internet-glue.co.uk (Paul Clinch)
Date: 3 Apr 2002 13:42:51 -0800
Subject: How to wait only so long for a server to respond?
References: <mailman.1017800895.19820.python-list@python.org> <n1Dq8.12893$ml2.1014180@newsread1.prod.itd.earthlink.net>
Message-ID: <8cf2994e.0204031342.7c7432c9@posting.google.com>

Ralph <ralph at penguin.com> wrote in message news:<n1Dq8.12893$ml2.1014180 at newsread1.prod.itd.earthlink.net>...
> Sean 'Shaleh' Perry wrote:
> 
> > 
> > On 03-Apr-2002 Ralph wrote:
> >> Pythoners:
> >> 
> >> I have some code that requests a response from a server on the Internet,
> >> but I only want to wait so long for the server to respond.  How can I
> >> tell python to wait x seconds/minutes, then move on to the next section
> >> of code if the server has not responded?
> >> 
> >> Thanks as always.
> >> 
> > 
> > a select loop is a common way to wait for data on a socket and specify a
> > timeout.  Look for documentation on select().
> Sean:
> 
> Thanks for the hint.  I'll take a look.
> 
> Ralph

You can also try timeoutsocket.py from the vaults of Parnassus.
I have used it with 1.52 successfuly.

Paul


From phr-n2002a at nightsong.com  Mon Apr 15 19:38:19 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 16:38:19 -0700
Subject: Looking for GUI Form Builder for Python
References: <mailman.1018882348.11577.python-list@python.org>
Message-ID: <7xlmbomt1w.fsf@ruckus.brouhaha.com>

"Paul Sage" <psage at ncaustin.com> writes:
> Have you investigated QT?  I believe it might be able to do what you are
> looking for.  However, maybe someone on this list who has worked with it
> will be able to confirm that for you.  I have not worked with it, but I
> believe it has a Python wrapper available.

There's also Blade.


From siegfried.gonzi at kfunigraz.ac.at  Wed Apr  3 07:01:27 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 03 Apr 2002 14:01:27 +0200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
Message-ID: <3CAAEF17.32333986@kfunigraz.ac.at>

Jan Kybic wrote:

> In my field (signal and image processing) and on high-level tasks,
> Python+Numeric runs at the same speed as the equivalent Matlab code.
> I have one big application that I basically wrote twice for both
> platforms and that's the conclusion I made.
> 
> On the other hand, all the calculation that you cannot express using
> high-level primitives (using matrices, vectors,...) and has to be
> performed with loops element by element, is painfully slow. Equally
> slow in Python, as in IDL, Matlab,... 

Not splitting hairs here: but IDL and Yorick are faster than Matlab when
it comes to loops!

>It is so slow that it is
> unusable, it is very difficult to develop an algorithm if it takes
> days to process any reasonably sized input. My solution is to rewrite
> the critical part of the code in C/C++, I did this for Matlab,
> Python+Numeric, as well as IDL, but this is awkward - 

Good to see that my assumption that Numeric is not slower than Matlab or
IDL can be confirmed.

> it forces you to
> work simultaneously with two languages and to deal besides the algorithm
> itself with a lot of implementation and interface details.
 
Dubois dubbed this "steering your code". I do not have got much
experience at this field (foreign functions). But why do you think that
steering code is in the long run not the panacea? 

> I would love to hear your thoughts on the subject.

I am with (who not would acknowledge a faster Python?) you on the
subject.  I am not a computer scientist and will not go to propose how
to make Python faster. There were threads which discussed the topic.
Often Common Lisp is strained as a model for a faster Python. Common
Lisp (compiled) normaly is about 10 times slower than an equivalent C
code (you can declare Lisp variables --and these would boost you in the
2 times slower than realm of C or Fortran--but declaring variables in
Lisp is a different story and science and for normal users quite hard to
manage).

I think the question actually is more or less: would you be contented
with a Python which produces compiled code which in turn executes 10
times slower than an equivalent C code? Normally, Python executes code
about 100 times slower than C.

And the next question: It is feasible to improve Python to that level?
You may not forget that people -- even when Python improves from a 100
times to 10 times performance penalty -- then will also compare it to C
or Fortran and the 10 times performance hit remains!

It would be a pity if one is investing much time in improving Python but
has to recognise that people always will point to the 10 times
performance hit compared to lets say C (Common Lisp coders are faced
with that problem every day).


> Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France

What about your experience with Scilab? Or even due to INRIA: OCaml?


S. Gonzi


From hancock at anansispaceworks.com  Fri Apr 12 13:01:23 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Fri, 12 Apr 2002 10:01:23 -0700
Subject: Inspecting multiple-inheritance
Message-ID: <3CB712E3.B26D6F27@anansispaceworks.com>

Hi All,

I'm trying to understand a portion of Zope, which involves
highly multiply-inherited objects.  There are several
instances of methods being overloaded, and so I'm trying
to figure out "who wins".  In some cases there appear to
be as many as six classes inherited from, and each of
those may have three-to-six as well, so that in the end
there may be 20 or more superclasses to search.

I know I can find out what methods/attributes an object
has by using dir() -- but is it possible to figure out
what arguments they expect, or what superclass they
came from?  In this particular case, I'm trying to
determine where __init__ comes from, and what arguments
it expects, as I expect that I need to call the appropriate
__init__ from superclasses.

But in general, it seems like this would be an important
thing to be able to do, as this would seem to be a
hazard of object-oriented programming in general. I tried
to search the onsite documentation and Learning Python,
but I don't really know what I'm looking for.

Thanks for any suggestions or comments!

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------




From ngreco at softhome.net  Mon Apr  8 11:17:53 2002
From: ngreco at softhome.net (Nahuel Greco)
Date: Mon, 8 Apr 2002 12:17:53 -0300
Subject: Python programmer available in Argentina
Message-ID: <20020408121753.2289fdd3.ngreco@softhome.net>

Hi, im looking for a python/linux related job in Argentina, more info
about me in:

http://nahuel.codelarvs.com.ar/Curriculum.html



------------
Nahuel Greco     




From emile at fenx.com  Wed Apr 10 07:25:31 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 10 Apr 2002 11:25:31 GMT
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <u4rijals0.fsf@python.net>
Message-ID: <LiVs8.199903$ZR2.102340@rwcrnsc52.ops.asp.att.net>

"Michael Hudson" <mwh at python.net> wrote in message
news:u4rijals0.fsf at python.net...
> "Emile van Sebille" <emile at fenx.com> writes:
> > Oooh... that's almost as much fun as saying 'xyzzy' the first
time...
> > ;-)
>
> That's not as much fun as "syzygy" (which is a real word,
> astonishingly enough).
>

That may still be fun for some, but the advent of speaking dictionaries
has taken mankind from the caves.  ;-)

http://www.m-w.com/cgi-bin/audio.pl?syzygy01.wav



--

Emile van Sebille
emile at fenx.com

---------



From sholden at holdenweb.com  Wed Apr 24 15:29:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 15:29:24 -0400
Subject: mxCGIPython
References: <mailman.1019673524.11680.python-list@python.org>
Message-ID: <HKDx8.6600$z55.6042@atlpnn01.usenetserver.com>

"Mark McEahern" <mark at mceahern.com> wrote ...
> Hi, I found this wonderful site:
>
>   http://www.egenix.com/files/python/mxCGIPython.html
>
> However, when I try to gunzip this file, I get this message:
>
>   $ gunzip -f cgipython-2.2-Linux-2.2.19-i686-with-debian-2.2.gz
>   gunzip: cgipython-2.2-Linux-2.2.19-i686-with-debian-2.2.gz: not in gzip
> format
>
> This happens both on Cygwin and on Linux.  Any ideas?
>

$ file cgi*
cgipython-2.2-Linux-2.2.19-i686-with-debian-2.2.gz: ELF 32-bit LSB
executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared
libs), stripped


tells me it's an executable. Your browser has probably expanded it (without
renaming) on the fly. Try removing the .gz extension and running it (I
can't - I did the "file" under cygwin, which won't execute it, and quite
right too).

but-don't-blame-me-if-it's-a-bomb-ly y'rs  - steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From loewis at informatik.hu-berlin.de  Thu Apr 11 08:02:49 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 11 Apr 2002 14:02:49 +0200
Subject: distutils
References: <GTet8.4573$Dw4.8174@news1.bredband.com>
Message-ID: <j4sn62xx1y.fsf@informatik.hu-berlin.de>

Magnus Heino <magnus.heino at pleon.sigma.se> writes:

> Im using distutils to create an rpm for me, 
> 
> python2.2 setup.py bdist_rpm
> 
> Now, this installs the module in /usr/lib/python2.2/site-packages/
> 
> How can I install it somewhere else instead?

See

http://www.python.org/doc/current/inst/alt-install-windows.html

Regards,
Martin


From s.schwarzer at ndh.net  Mon Apr  1 17:45:38 2002
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Mon, 01 Apr 2002 23:45:38 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com>
Message-ID: <3CA8E312.78175D08@ndh.net>

Erik Max Francis wrote:
> Stefan Schwarzer wrote:
> > But, if I understand your argument correctly, "True == 1" and
> > "False == 0" shouldn't be true either because True and False are
> > bools, and 1 and 0 are ints. (The problem with this is, that, as a
> > consequence to the reasoning in the PEP, bools are ints in disguise.)
> 
> Yes, bools are really ints underneath the hood, and for backward
> compatibility it makes sense that there should be implicit conversions
> in numeric contexts from bools to ints, so True == 1 and False == 1 are
> both true.  However, it won't go the other way around -- there won't be
> (and shouldn't be) implicit conversion from _ints_ (or anything else) to
> bools, so True == 2 or False == [] shouldn't be true.  (Because of
> Python's typing system, they won't be errors, either, just untrue.)  If
> you want conversion, you have to make it implicit with a bool
> conversion:  True == bool(2) and False == bool([]) will both be true.

Yes, that is (and was) all well understood. :-)

> > Example: "(val1, val2)[0]" is pretty clear, but "(val1, val2)[False]"
> > looks quite strange. If that's the price of the change I would still
> > prefer that "2 > 1" gives 1, not True.
> 
> But that's simply a case of bad style, not a demonstration of the
> inappropriateness of Booleans.  After all, even now you could have
> written
> 
>         (x, y)[operator.truth([])]
> 
> This is ugly to be certain, but doesn't indicate a flaw in the language.

Sorry, I don't understand what you want to say. Perhaps I can make explain
what I meant ...

> > 1 Have a bool type without these strange interactions with ints (i. e.
> >   a bool should never equal an int, as an int should never equal a
> >   string, as you explain above), which breaks a lot of code.
> >
> > 2 Or, as the PEP suggests, introduce a bool type which behaves quite
> >   unintuitive, but keeps existing programs running.
> 
> I don't see why you think the behavior (I presume you're referring to
> syntaxes like True == 2) is counterintuitive.  Why _would_ you expect
> this to do what you want? It doesn't for any other Python types that
> I'm aware of.  == tests for equality, it doesn't do automatic type
> coercion.

Yes, exactly that's my point. I meant, "True == 2" doesn't (IMHO) make
sense. But "True == 1" (comparing a bool with an int) doesn't make sense
either. Neverless this expression would be true, according to the PEP.

I would like to have bools in Python, rather than not. But only if
bools and ints are different types, so that "True == 1" would be false,
similar to that "'1' == 1" is false. And _my_ conclusion was, that I
would rather have no bool type than a mixed-up concept in which bools
are really ints.

What's the point in saying "There should be a bool type to distinguish
between truth values and integers" when this distinction is so shallow?
This seems kind of "illogical" to me. Why bother that "2 > 1" should
give "True" or "1" when I would have to explain "Of cource an int is
different from a bool, but, erm, True == 1 nonetheless". I can't see
how this should make the language easier to understand, both for novices
and for experienced programmers.

Stefan


From tool_man at spamcop.net  Thu Apr 11 11:37:09 2002
From: tool_man at spamcop.net (Tim Lavoie)
Date: Thu, 11 Apr 2002 15:37:09 GMT
Subject: Sound programming?
References: <3CB50DBC.1B0E533F@alum.mit.edu>
Message-ID: <slrnabbbd4.8fn.tool_man@user22.beyondtv.net>

In article <3CB50DBC.1B0E533F at alum.mit.edu>, Will Ware wrote:
> Somebody was asking me this evening if there's an easy way to
> play a tone of a specified frequency and duration in Python.
> I didn't know off the top of my head. I've since poked around
> the web a bit and it appears like the Snack Sound Toolkit
> (http://www.speech.kth.se/snack/) might be a good way to go.
> One of its virtues is being cross-platform, which I suspect is
> among his unstated wishes. Are there any other options?

Another is Csound, which is a wickedly powerful software synth. It's
written in C, runs on multiple platforms, and has lots of extras like
front-ends written, including some Python. It might be waaay too much
for a simple task, but the extra power is rather intoxicating.

  Csound: http://www.csounds.com

  Python bits:
    http://pythonsound.sourceforge.net/
    http://web.tiscali.it/mupuxeddu/software/CSFE/index.html

-- 
I finally went to the eye doctor.  I got contacts.  I only need them to
read, so I got flip-ups.
                -- Steven Wright


From jlamanna at its.caltech.edu  Sun Apr 28 19:33:37 2002
From: jlamanna at its.caltech.edu (James Lamanna)
Date: Sun, 28 Apr 2002 16:33:37 -0700
Subject: PyQt 3.1 Compilation Problems
References: <3CCB01F5.1090607@its.caltech.edu> <3CCBD041.F2A703CF@river-bank.demon.co.uk>
Message-ID: <3CCC86D1.2020503@its.caltech.edu>

Phil Thompson wrote:

>
>This is a similar bug in the Mandrake 8.1 Qt package. They removed the
>the QSGIStyle and QCDEStyle classes for some reason, but didn't make the
>corresponding change to qconfig.h.
>
>Add the following to $QTDIR/include/qconfig.h and try again...
>
>#define	QT_NO_STYLE_SGI
>#define QT_NO_STYLE_CDE
>
>Phil
>
>.
>
Just thought I'd let you know that apparently the Debian Qt package also 
leaves out the Windows and Motif styles, so you have to add
#define QT_NO_STYLE_WINDOWS
#define QT_NO_STYLE_MOTIF
also.

--James Lamanna






From tchendricks at lucent.com  Wed Apr 24 11:11:16 2002
From: tchendricks at lucent.com (Tom Hendricks)
Date: 24 Apr 2002 08:11:16 -0700
Subject: Finding scripts
Message-ID: <af916b72.0204240711.41f125f8@posting.google.com>

I am a newbie to the python world.  I am in the process of configuring
about 300 switches with basically the same information except for IP
address and VLAN information.

It was suggested to me that I download python and use expect to create
a script to perform this work more efficiently and faster.  This is
great in theory if you have programmed before but I think it would be
faster to manually configure the switches than to learn python at this
time, short-term thinking, of course.

Can someone direct me to where I might find scripts for configuring
Enterasys Smartswitch 6000, Smartswitch 2200, and Matrix E7's?

In the meantime, back to my Learn to Program using Python: Lesson 1,
Getting Started by Richard Baldwin.


Thanks,

Tom


From loewis at informatik.hu-berlin.de  Thu Apr 18 07:22:05 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 13:22:05 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019036193.29349.python-list@python.org> <a9m5rn$4b8ru$2@ID-99513.news.dfncis.de>
Message-ID: <j4662p8d5u.fsf@informatik.hu-berlin.de>

"P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:

> As said in other message, this is not the recommended procedure; the
> fact that you can does not mean you should. GNU as does not generate
> 100% compatible binary object code, and GNU ld tends to botch Sparc
> object code linked against libraries created with Sun tools (that is
> system libraries, and you link libc, libnsl and libsocket with
> almost everything like it or not).

This is FUD. Software does not "tend to" do things. It either does
them, or it doesn't. Use the version of binutils that is recommended
in the gcc release notes, and all will work just fine.

> > Actually, you *can* talk the Solaris linker to incorporate non-PIC
> > text into a shared library. For that, it will make the text segment
> > writable. The -mimpure-text option of gcc triggers that feature.
> 
> OK, but shouldn't this sort of information be in the configure script?

No. Using -mimpure-text has undesirable side effects, such as making
the text segment writable (which in turn means that the shared library
won't be shared across address spaces).

There is only so much that configure can find out automatically. If
you have a slightly broken system, or one that lacks certain features,
you need to give explicit instructions to your tools.

> But seriously, Solaris is quirky, but no one contributes the appropriate
> patches to CVS... (I am not going to, I have other things to worry about
> presently).

Patches that propose to replace -shared with -G when using gcc will be
rejected, as they are plain wrong. Patches that always add
-mimpure-text will be rejected as the cause performance degradation
for extension modules that don't need relocations in their text
segment.

> > If you link static libraries into Python, I recommend to make the
> > modules that use them also static. Then you don't need PIC code for
> > those.
> 
> Wouldn't this make the memory image of all python programs much larger?

What are "all python programs"? There is only one Python
executable. 

If you are concerned about the memory consumption of python processes:
Since Solaris uses demand-paged memory, code of extensions won't be
loaded into memory until somebody touches the page that has the code,
which usually won't happen until the module is imported.

> Nah. This is misinformed exaggeration. The culprit of unnecessary fluff
> is gcc itself, and it becomes worse with each new release. If you are
> concerned with executable size, you should add an "-s" flag to the gcc
> front-end. This strips all debugging symbol code from the binary object
> code. In the case of gcc 3 it can mean a size savings of 60% and even
> more.

Of course, debugging information does not contribute a bit to the
in-memory consumption, since the dynamic loader skips the debugging
sections when starting a program.

Regards,
Martin


From goldinf_obases3728_1y at yahoo.com  Mon Apr  1 00:58:21 2002
From: goldinf_obases3728_1y at yahoo.com (GOLD INFO)
Date: Mon, 1 Apr 2002 09:58:21 +0400
Subject: =?windows-1251?Q?=CB=E5=E3=EA=EE=F1=F2=FC_=EF=F0=E8=E2=EB=E5=F7=E5=ED=E8=FF_=ED=EE=E2=FB=F5_=EA=EB=E8=E5=ED=F2=EE=E2?=
Message-ID: <17367200241155821496@yahoo.com>




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020401/c949f681/attachment.html>

From mtaylor at lowebroadway.com  Tue Apr 30 10:52:46 2002
From: mtaylor at lowebroadway.com (Moray Taylor)
Date: 30 Apr 2002 07:52:46 -0700
Subject: Backdrop in QListView using PyQt, possible?
Message-ID: <d906b2c1.0204300652.5d7fac93@posting.google.com>

Hi,

Is it possible to apply a backdrop to a QListView, i.e. a PNG or
whatever, loaded into a QImage and used as a tiled (or otherwise)
backdrop, like you can on a QCanvas?

Please send replies to mtaylor at lowebroadway.com aswell please.

Thanks in advance

Moray


From magnus.heino at pleon.sigma.se  Thu Apr 11 07:58:20 2002
From: magnus.heino at pleon.sigma.se (Magnus Heino)
Date: Thu, 11 Apr 2002 13:58:20 +0200
Subject: distutils
Message-ID: <GTet8.4573$Dw4.8174@news1.bredband.com>

Hi.

Im using distutils to create an rpm for me, 

python2.2 setup.py bdist_rpm

Now, this installs the module in /usr/lib/python2.2/site-packages/

How can I install it somewhere else instead?

/Magnus Heino


From theller at python.net  Mon Apr 15 13:56:37 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 15 Apr 2002 19:56:37 +0200
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net> <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1> <a9f30t$mmo$0@216.39.172.122>
Message-ID: <a9f494$2lltn$1@ID-59885.news.dfncis.de>

"Bengt Richter" <bokr at oz.net> wrote in message news:a9f30t$mmo$0 at 216.39.172.122...
> On Mon, 15 Apr 2002 11:35:13 +0000 (UTC), Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:
>
> >Scott David Daniels <Scott.Daniels at Acm.Org> wrote in
> >news:1104_1018820620 at news.dsl-only.net:
> >
> >>> Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:
> >>> > Thank you very much for pointers, but I can think of something not
> >>> > listed there. For example, __all__ for "from module import *",
> >>> > pickle protocols, __new__ for subclassing from a type and so on. Is
> >>> > there a complete (or near-complete) list?
> >>> ...
> >>
> >> You could also take this lack of a list as a charter for your
> >> collecting _all_ (not __all__) such names, and creating a document for
> >> others who, like you want to know the list of such names.  This could
> >> include package references if you so desire (others will complain if
> >> you leave something out, providing you with more fodder for your
> >> document).
> >>
> >
> >To get you started, here is a list built by grepping over the Python
> >sources. Some of the names are probably red herrings, and there are
> >probably some used by libraries not in this list, but it should give you a
> >start:
> >
> [... grepped sorted list ...]
>
> Cool. Do you feel motivated to do the same on the .../doc/* tree and
> then do a diff? ;-)
>
There's a tool which does this: Doc/tools/undoc_symbols.py.
(It has to be tweaked slightly).

Thomas




From bernie at 3captus.com  Fri Apr  5 21:55:05 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Sat, 06 Apr 2002 02:55:05 GMT
Subject: I like python
References: <mailman.1018014552.1132.python-list@python.org>
Message-ID: <3CAE63C8.9B5FE1B8@3captus.com>

Hello cstyphoon,

There are people in this group who can read Chinese (I am one of those)
so you can post your question in Chinese.  In doing so, however, will
limited the number of people who can help you (Chinese are minority in
the group).  Also, be careful of the encoding you use.  I suggest you
use Big5.  Right now I can't read the Chinese in your message.

Welcome to the world of python!

Bernie


?? ?? wrote:
> 
>                               styphoon at yahoo.com.cn
>                               Shichuan Province China
> 
> Hello,comrades!
> I am a highschool-student in Chinese, and I am fond of
> computer, especial programming.But now I have some
> difficulties.Are you ready to help me? If it is
> possible,then I will tell you what difficulties I have
> met.I know only a little English, and, do you know
> Chinese? The next paragraph is written in Chinese, can
> you understand? I wonder whether you can write to me
> in Chinese? In  that way, I will be more grateful to
> you.
> 
> 
> 
> ????????????????????,????????????????,??????????,????????????????????,??????????????????????.????????????????,?????????????????????????????????????????????,????????????.??????????????????,??????????????????????????.
> ??????,??????????????,????,????!
>                           ??????  ????
>                          2002.4.5.
>                          Best wishes!
> 
> 
> _________________________________________________________
> Do You Yahoo!?
> ??????????????????????????????????????????????
> http://cn.worldcup.yahoo.com/

-- 
There are three schools of magic.  One:  State a tautology, then ring
the changes on its corollaries; that's philosophy.  Two:  Record many
facts.  Try to find a pattern.  Then make a wrong guess at the next
fact; that's science.  Three:  Be aware that you live in a malevolent
Universe controlled by Murphy's Law, sometimes offset by Brewster's
Factor; that's engineering.
                -- Robert A. Heinlein


From henne021 at tc.umn.edu  Tue Apr  2 16:43:17 2002
From: henne021 at tc.umn.edu (Scott Hennesy)
Date: Tue, 2 Apr 2002 21:43:17 +0000 (UTC)
Subject: ANN: Pytalk version 0.20 released
Message-ID: <a8d8l5$ndv$1@laurel.tc.umn.edu>

Available at http://sourceforge.net/projects/pytalk

Pytalk is a natural language understanding program written in Python. It includes an English language parser, an English dictionary tagged with parts of speech and including phrasal verbs and common first names, a spelling corrector, and an indexed file module.

Changes for version 0.20:

Much more robust parsing (and more maintainable newparser.py code)

Many improvements to generation

Improvements to inference.py for better semantics

Indexed.py refactored somewhat and experimental use of indexed files instead of shelves (performance suffers however)




From pyth at devel.trillke.net  Sat Apr 27 17:11:00 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 27 Apr 2002 23:11:00 +0200
Subject: Update locals() / resolved
In-Reply-To: <20020427202326.GB16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke>
Message-ID: <20020427211100.GC16217@devel.trillke>

On Sat, Apr 27, 2002 at 10:23:26PM +0200, holger krekel wrote:
> On Sat, Apr 27, 2002 at 03:06:08PM -0500, jepler at unpythonic.net wrote:
> > On Sat, Apr 27, 2002 at 09:57:25PM +0200, holger krekel wrote:
> > > Hello,
> > > 
> > > how can you dynamically update the locals() namebindings?
> > 
> > You don't.  See the library documentation:
> >     locals()
> > 	Return a dictionary representing the current local symbol table. Warning:
> > 	The contents of this dictionary should not be modified; changes may not
> > 	affect the values of local variables used by the interpreter.

ok. i found a nice workaround (if one dare call it that:-)

class dictlookup(type({})):
    def __getattr__(self, name):
        return self[name]

then i can do:

def func(...,somedict,...):
    m = dictlookup(somedict)
    ...
       m.key1 
    ...

which is reasonably short and additionally seems more robust.
I think it's generally quite nice for dictionaries containing 
'strings' as keys which you know in advance and want to access
without noise. 

    holger




From nospam-readonly at getsoft.com  Mon Apr 22 17:28:07 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Tue, 23 Apr 2002 07:28:07 +1000
Subject: A simply newbie question about ndiff
References: <mailman.1019493956.28198.python-list@python.org>
Message-ID: <3cc4804f$0$15473$afc38c87@news.optusnet.com.au>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1019493956.28198.python-list at python.org...
> [Neville Franks]

> snip...

> > I only discovered your ndiff over the weekend, which is a little
> > surprising as I've also been working on Diff's for quite a few
> > years now as part of my programmer's editor, ED for Windows.
>
> Cool.  ndiff will do a fine job for you, but if you're slinging giant
files
> you'll need to compromise.
>
>..
> Now SequenceMatcher uses a much cleverer implementation than that
> brute-force triply-nested loop (which, as written, has worst-case cubic
> time), but it can still take substantial time, and 70K lines is
ridiculously
> out of the range of sizes ndiff was intended to handle.  70K is in the
range
> of a very long novel; the ndiff algorithm was first used to create
> human-readable patches for source code files, and later for plain-ASCII
> design documents.  Those typically take a few seconds.


Well believe it or not but there are quite a few people using ED that have
source files in excess of 70K. Not something I would advocate but ...

On a more serious note I do have folks who compare large'ish data (ascii
text) files. These are typically dumps of some sort. Apart from being large
they can also get less than optimal results as they can have large numbers
of repeated lines which throws my diff out.

I've organized a copy of ED for you to play with and would be interested to
hear what you think of it's Diff capabilities. ED doesn't have Python
support just yet, but it is planned. It does support some 34+ languages
though. A Free Trial is available at www.getsoft.com for anyone else who may
be interested.

--
Neville Franks, Author of ED for Windows - http://www.getsoft.com




From ak at silmarill.org  Fri Apr 19 15:37:53 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Fri, 19 Apr 2002 19:37:53 GMT
Subject: urllib2: pwd protected urls
References: <slrnac0r2p.ol1.ak@ak.silmarill.org>
Message-ID: <slrnac0sk8.ovb.ak@ak.silmarill.org>

In article <slrnac0r2p.ol1.ak at ak.silmarill.org>, Andrei Kulakov wrote:
> Hello,
> 
> I can't figure out how to get pwd-protected urls using urllib2. Here's
> what I'm doing:
> 
>>>> from urllib2 import *
>>>> a = HTTPBasicAuthHandler()
>>>> a.add_password("default", "host.net", "login", "password")
>>>> o = build_opener(a)
>>>> install_opener(o)
>>>> f = urlopen("http://host.net/file.zip")
> 
> I get Error 401: Auth required
> 
> What's 'realm'? I set it to 'default' and I also tried None.
> 

Geoff Gerrietts told me how to do this:

urlobj = urllib.urlopen("http://login:password at host.net/file.zip")


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From philh at comuno.freeserve.co.uk  Mon Apr 22 18:57:15 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 22 Apr 2002 23:57:15 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <3CBB83B1.8EC2E2F6@python.org> <slrnabp8ub.ipr.philh@comuno.freeserve.co.uk> <cpofgbwnfa.fsf@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <slrnac95ab.sc.philh@comuno.freeserve.co.uk>

On Mon, 22 Apr 2002 19:17:53 GMT, Guido van Rossum <guido at python.org> wrote:
>> >phil hunt wrote:
>> >> IMO until Python developers adopt a policy of being very reluctant
>> >> to do this, many IT managers will be wary of using it in anything
>> >> other thsan toy projects, and Python's user base will not grow as
>> >> quickly as (IMO) it should.
>
>[Guido]
>> >If in your opinion Python should grow faster, perhaps you could
>> >help out by contributing rather than by telling us what to do...
>
>philh at comuno.freeserve.co.uk (phil hunt) writes:
>
>> I would contribute, but you have just said that you won't accept my 
>> proposed changes.
>
>There are other ways to contribute besides by proposing language
>features.  We need help doing code cleanup, backporting patches to
>bugfix releases, documentation, and so on.

There have certainly been times when I have read the documentation,
and thought "that point could have been better explained". Having
said that, I'm still using 2.0, and I dare say the latest docs are
better. Buit the next time I get this feeling, I will have a look at
the latest docs and submit a patch (assuming that's the correct way
to submit changes to documentation).

>Also, I only told you about one very specific (and silly) 

IMO it wasn't silly. To some extent, you must agree with me, because 
otherwise you would have required stuff like:

   import nothing from system
   define no classes in this program
   def myFunction(void):
      a = [ empty ]    
      b = "\empty"

instead of:

   def myFunction():
      a = [ ]
      b = ""

Indeed there are many places in Python (and probably every other 
programming language as well) were one can specify nothing simply by 
not saying anything, rather than by explicitly having to say "I 
don't want anything here".

>change
>you've proposed that I wouldn't accept it.  But I'd rather see you
>help out in other ways than proposing language changes, since the
>latter (even if they don't break anything) cause complaints about the
>pace of change.

These could be mitigated by making non-bugfix releases less often, 
perhaps using a system of even/odd version numbers like Linux uses.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From philh at comuno.freeserve.co.uk  Tue Apr 16 19:15:41 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 17 Apr 2002 00:15:41 +0100
Subject: pre-PEP for optional 'pass'
References: <mailman.1018914818.28508.python-list@python.org>
Message-ID: <slrnabpc4s.ipr.philh@comuno.freeserve.co.uk>

On Mon, 15 Apr 2002 18:52:14 -0500, Mark McEahern <marklists at mceahern.com> wrote:
>[phil hunt]
>> I wish to make the pass statement optional. That is, in any Python
>> program where there is a pass it can be left out and the meaning of
>> the program is unchanged.
>[...]
>> Thoughts?
>
>I imagine you'll need to provide working code that demonstrates what's
>required to accomplish this proposed PEP--or at least some analysis of the
>scope of the work and the potential for breakage.  It would be really
>interesting to see you post something about that.
>
>Having never really delved into the C source code for Python itself, I have
>no real sense of the scope of the change you're proposing.

I've looked at it briefly.

> Would existing
>code break? 

Not unless my mod contained a bug.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From jacek.generowicz at cern.ch  Sat Apr 27 05:13:48 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 27 Apr 2002 11:13:48 +0200
Subject: Python's Lisp heritage
References: <mailman.1019846689.31490.python-list@python.org>
Message-ID: <tyfd6wl33n7.fsf@pcitapi22.cern.ch>

"Brian Quinlan" <brian at sweetapp.com> writes:

> Jacek wrote:
> > No, as Paul said, they are used for "function calls". Admittedly, in
> > mathematical expressions function call often coincide with
> > sub-expressions, but it does depend on the sub-expression in
> > question. Compare
> > 
> >    (+ 1 2 3 4 5 6 7 8 9 10)
> > 
> > with
> > 
> >    1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10
> > 
> > Gosh, what a lot of typing I have to do in Python, compared to lisp.
> 
> Actually, in your own example, the lisp example requires more typing.
> You disingenuously inserted unnecessary spaces in the Python example.
> Allow me to fix that:
> 
> Lisp:   (+ 1 2 3 4 5 6 7 8 9 10)
> Python: 1+2+3+4+5+6+7+8+9+10
> 
> You could also write it like:
> 
> >>> reduce(operator.add, (1,2,3,4,5,6,7,8,9,10))
> 55

(reduce #'+ '(1 2 3 4 5 6 7 8 9 10))

What, you mean that in python, if I want to refer to the funcion +, I
first have to import it from a separate module, and then call it
`add'? What a lot of typing ! :-) :-) :-)

(i.e. Let's not go off on completely pointless tangents.)

> Your example is not really interesting, except to show that lisp has a
> multi-argument add function and Python doesn't.

My example was indended to disabuse James of the notion that
parentheses are used in lisp "around sub expressions".

As I said in my reply to him, the example was partly in jest.

> Implementing one, as I demonstrated above, is a one liner e.g.
> 
> def add(*args): return reduce(operator.add, args)
> add(1,2,3,4,5,6,7,8,9,10)

Now go and implement it for all the other operators. Now go and
implement a CL environment in Python, and you can claim that every
feature of Common Lisp is available in Python (or whatever languague
you choose). :-)

But if you do that, then allow me to implement a Python in CL (I'm not
going to participatie in any ensuing discussions about which of the
two tasks is easier), thus proving that all computer languages are
absolutely identical, and I'm not talking about Turing equivalence :-)

Or maybe we should just stick to the point in question.


From NO_SPAM_jmpurser2 at attbi.com  Mon Apr  1 14:22:38 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Mon, 01 Apr 2002 19:22:38 GMT
Subject: readlines problems - Newbie - Settled
References: <k_1q8.146744$Yv2.44641@rwcrnsc54> <Xns91E3D6A2BD74Ccliechtigmxnet@62.2.16.82>
Message-ID: <2s2q8.150053$ZR2.64667@rwcrnsc52.ops.asp.att.net>

Thanks to Chris's link to the documentation I found that I can't do what I
was trying to do the way I was trying to do it.  Bottom line I've got to
pick a different path.  Fair enough.

Thanks for the help.

John Purser

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns91E3D6A2BD74Ccliechtigmxnet at 62.2.16.82...
> "John Purser" <NO_SPAM_jmpurser2 at attbi.com> wrote in
> news:k_1q8.146744$Yv2.44641 at rwcrnsc54:
> > I've got a little function:
> > def FindLine(Record):
> >      im1_Dif.seek(0)
> >      im1_Dif.readlines(Record - 1)
> >      test = im1_Dif.readline()
> >      print test
> >
> > When I call it I wind up with the same record in "test" whether I pass
> > it 1 or 10.  The readlines line reads a huge chunk of text with no
> > regard for lines at all.  I have no idea what is causing this.  Where
> > can I find the documentation on the file methods?  I can't find them
> > in the documentation. Can anyone tell me what's going on here?
>
> in the docs:
> http://www.python.org/doc/current/lib/bltin-file-objects.html
>
> you can see that this is a limit in bytes if even implemented.
>
> you can read all the lines in a buffer:
> >>> all = im1_Dif.readlines()
> and just work with the last one
> >>> test = all[-1]
>
> chris
> --
> Chris <cliechti at gmx.net>
>




From aleax at aleax.it  Mon Apr 29 11:59:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 15:59:16 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org>
Message-ID: <o5ez8.77579$vF6.2303394@news2.tin.it>

Roman Neuhauser wrote:

>> The problem with int.chr() is that a lot of people are going to get
>> confused when they find that "42.chr()" doesn't work (although "42
>> .chr()" does).
> 
>     why? letters aren't valid decimal digits, are they?

Decimal digits have nothing to do with it.  The lexer does maximal-munch,
so it snags off "42." as one first valid token (a float), "chr" as a
second one, and so on -- unless you separate the digits from the dot,
e.g. with whitespace, this is more or less inevitable.  And making the
lexer complicated/slower/less predictable to allow "calling methods on
integer literals" without space doesn't appear an attractive trade-off.


Alex



From pyth at devel.trillke.net  Tue Apr 30 09:17:20 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 15:17:20 +0200
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <kNvz8.12234$xy4.11902@atlpnn01.usenetserver.com>
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <kNvz8.12234$xy4.11902@atlpnn01.usenetserver.com>
Message-ID: <20020430131720.GD16217@devel.trillke>

On Tue, Apr 30, 2002 at 08:04:01AM -0400, Steve Holden wrote:
> > As it is, we can no longer look at a fragment of code containing a
> > division operator and know what it means without checking version and
> > import declarations (which may appear ANYWHERE in the source).  I
> > frankly find this to be BIZARRE.
> >
> If you're talking about "from __future__ ...", these can only appear before
> any other source lines, and so (bugs notwithstanding) they aren't as
> difficult to detect as you believe. Remember that some imports from
> __future__ modify how the compiler acts, so you can't just switch them on
> and off at random.

from __future__ even modifies what the *parser* accepts. 
It might help to educate us and others to mention - as needed -
the basic way of how python-sourcecode gets processed:

- parsing into an abstract syntax tree (AST)

- compiling the AST into bytecode 

- running/executing the bytecode

Feel free to correct me :-)

Otherwise i second the statements in your posting.

    holger




From starx at pacbell.net  Mon Apr  1 16:17:53 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 1 Apr 2002 21:17:53 GMT
Subject: Most important extensions?
References: <a8614l$dfa$1@panix1.panix.com>
Message-ID: <Xns91E3877851015RASXnewsDFE1@130.133.1.4>

Aahz || Sat 30 Mar 2002 07:51:49p:

> I'm teaching Python for [Perl] Programmers at OSCON this July.  I've
> already decided to mention mxODBC, NumPy, and PIL, plus PythonPoint
> (aka ReportLab) gets a mention on my title page.  I'd like this group
> to vote on the most important extensions for me to add to this list. 
> Please vote for no more than three.

I've seen just about everything already mentioned that I could think of, 
except these two:

PyOpenGL - Interface to OpenGL, the only (?) widely known, popular, and 
extensivly used (by corporations) graphics toolkit not owned by MS. And 
the interface is such that you can use it with nothing aside from the GL 
Redbook 95% of the time, and the differences are mostly ovbious 
(comprimises for functions that want pointers, mostly)

PyGame - Interface to SDL, which is a high-quality 2d grahpics kit that 
seems to be rising these days...

Oh, and since there's three slots, PyOpenGLContext, which nicely ties the 
first one into the second, and into WX and TK.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From sholden at holdenweb.com  Sat Apr 27 08:16:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 27 Apr 2002 08:16:34 -0400
Subject: looping through a list
References: <e2e9e807.0204261611.7514b387@posting.google.com>
Message-ID: <4Hwy8.52567$z55.33321@atlpnn01.usenetserver.com>

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in message
news:e2e9e807.0204261611.7514b387 at posting.google.com...
> I need help with a really simple algorithm that I can't get my head
> around. I am looping through the list returned by a dcoracle2 execute
> statement and need to add each item of data surrounded by some HTML
> table tags to a string called text which I then pass to a page
> generation method. At the moment I've got this:
>
> result = c.fetchone()

Remember that the result will be a tuple. It will, therefore, never be equal
to "". When empty it will be equal to ().

>   if (result != ""):
>     i=0
>     text = '<tr width = "80%" align = "center">'
>     while (i != len(result)+1):
>       i = i + 1
>       text = text + '<td align = "center" valign =
> "top">'+str(result[int(i)])  +'</td>'
>     text = text + "</tr>"
>     page.generate("","",sessionID,text)
>   else:
>     text = '<h4>No items found</h4>'
>     page.generate("","",sessionID,text)
>
If would seem simpler to do the following:

if result:
    text = '''
    <tr width="80%%" align="center">
    %s
    </tr>''' % "\n    ".join(['<td align="center" valig="top">%s</td>' % c
for c in result])

>
> At the moment, it keeps saying list index out of bounds and I don't
> know where I'm going wrong. I've tried all sorts of things in the
> While statement and still can't get it going.
>
Not using a while statement will cure all those blues :-)

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From jb at cascade-sys.com  Fri Apr 26 00:38:20 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 21:38:20 -0700
Subject: Dynamic Binding
References: <LNBBLJKPBEHFEDALKOLCCEAPPBAA.tim.one@comcast.net>
Message-ID: <3CC8D9BC.48F0D9E9@cascade-sys.com>

Tim Peters wrote:

> [Tim]
> > It seems this would have been a surprise to McCarthy <wink>:
> >
> >     http://www-formal.stanford.edu/jmc/history/lisp/node4.html

> [James J. Besemer]
> > I disagree with your interpretation.
>
> I take it you didn't follow the link and read the referenced paragraph.

No, I must confess I did not.  I started to and then got drawn into doing some
real work for a couple days.  Also, it looked familiar and I was thinking that
maybe I had read it long ago.

When finally I sat back down to reply, my objective was merely to establish that
dynamic scoping originally was SOP in Lisp.  From that perspective, the quote
does speak for itself.  I actually thought you cited the quote in contradiction
of my premise and I said I disagreed because it clearly does not.

But actually, in a way, you were agreeing with me, right?  ;o)  And the quote
does provide additional fodder.

So I was wrong to disagree with you because you were agreeing with me, right?
:o)

Mistakes like this can happen in "heated agreement."

Please excuse my misunderstanding.

> > Rather he is saying that the implementation DID use dynamic "scoping,"
> > that it worked just fine in most circumstances and that when people
> > encountered a situation where "lexical scoping" would help McCarthy
> > regarded their "difficulty" simply to be a bug.
>
> I'll just finish the sentence I partially quoted:
>
>      I must confess that I regarded this difficulty as just a bug and
>      expressed confidence that Steve Russell would soon fix it.

Clarifies the ambiguity about who's bug it was.

> The differences between
> lexical and dynamic scoping weren't clearly understood at the time.  Yes, it
> used dynamic scoping, but the full implications of that implementation
> choice weren't understood in advance, and some came as unwelcome surprises.

I think we're in complete agreement here.

As I said elsewhere in my previous post, I think dynamic scoping a natural
progression that many language designers go through.  The associated problems
are fairly subtle and complex.

> Reading the paper remains a recommended alternative to presumption <wink>.

It does look worthwhile.  I'll definitely make a point of reading it.

Thanks for the history pointer, the clarifications, an interesting discussion
generally and your patience.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From mark.charsley at REMOVE_THIS.radioscape.com  Mon Apr 15 06:30:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Mon, 15 Apr 2002 11:30 +0100 (BST)
Subject: Beginner's performance problem
References: <mailman.1018596868.16860.python-list@python.org>
Message-ID: <memo.20020415113053.1532A@a.a>

In article <mailman.1018596868.16860.python-list at python.org>, 
shalehperry at attbi.com (Sean 'Shaleh' Perry) wrote:

> test and triple test your \bfoo\b regex's.  They sometimes mean things 
> you did
> not realize.  For instance re.compile(r'\bengine') would match 
> 'regex-engine'.

Ooh. Good point. Fortunately, as the stuff being changed in SQL, the names 
only contain alphanumerics and underscores. Even so, the changes are 
tested by
- checking that the files are the same size after the changes
- being recognized as valid SQL when rebuilding the database
- running all the automated unit tests against the new database

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From t_woerner at el-equip.com  Wed Apr  3 16:15:11 2002
From: t_woerner at el-equip.com (trevor)
Date: Wed, 03 Apr 2002 16:15:11 -0500
Subject: Deeply embedded Python still alive?
References: <SX1o8.573$vm6.188350@ruti.visi.com> <a7qtqg$u3f$0@216.39.172.122> <U27o8.610$vm6.225166@ruti.visi.com> <3CA143B7.DED16DF5@engcorp.com>
Message-ID: <XfKq8.552$cC5.347644@newsfeed.slurp.net>

thank you Peter for that update.

i would like to point out that i've been successful at getting a very 
minimal python-2.2 compiled, installed, and working on a 4MB zflinux board 
using uClibc-0.9.10, busybox-0.60.2, and gcc-3.0.4. it isn't the fastest 
thing around but it sure is neat! i'm still trying to decide if it will be 
adequate (performance-wise) for our purposes.

the fileset clocks in at 3036KB which, when compressed to create an initrd, 
becomes 1036KB. the python part of which takes up 2038KB (uncompressed). a 
basic test script would be something like:

---
for num in range(10):
        print num,
print
---

when the system boots up running "python test.py" gives:
0 1 2 3 4 5 6 7 8 9

unfortunately python doesn't make it very easy to build a minimal system, i 
had to do a lot of shuffling around to get it to work. additionally i had 
to play with the Makefile and edit some code too.

i would certainly appreciate if someone could point me to any docs that 
would make building a minimal system easier. for instance, how can i 
compile just a very basic python system without it going and trying to 
compile the _tk module on me?

anyway if you would like instructions i'd be only too happy to post them, 
but if nobody's interested then i'll save myself the effort.

best regards,
        trevor
-- 
remove underscore if you want to email me


From h.e.w.frater at cs.cf.ac.uk  Wed Apr 17 20:19:12 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 17 Apr 2002 17:19:12 -0700
Subject: DCOracle2 importing problems
Message-ID: <e2e9e807.0204171619.46de6db1@posting.google.com>

Hi All, I can import the DCOracle2 module fine from the interactive
interpreter but my CGI script causes an internal server error. The
interactive interpreter is running on the same machine as my scripts.
A tail of the apache error log reveals that my script can't import the
DCOracle2 module.
Any ideas anyone?

Hugh


From greg at cosc.canterbury.ac.nz  Mon Apr  8 21:00:40 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 09 Apr 2002 13:00:40 +1200 (NZST)
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <lbhy9g3gyse.fsf@xihu.inria.fr>
Message-ID: <200204090100.NAA00593@s454.cosc.canterbury.ac.nz>

Jan Kybic <kybic at ieee.org>:

> Once [Pyrex] compiles a little more of Python (notably classes), ideally
> any valid Python code, it should definitely become a tool of choice
> for people wanting to speed up their Python code.

I'm not sure that Pyrex will ever compile *all* valid Python
code. It probably wouldn't be very effective at speeding up
pure Python code anyway. The reason it deals with Python code
at all is so that you can write code to convert between
Python and C data structures -- the fact that you can
compile pure Python with it is sort of a side effect.

Given that, I'm not particularly motivated to put in large
amounts of effort to support features which would
be difficult to translate into C and which don't have
clear utility for writing conversion code. Generators
and nested scopes come to mind.

I might tackle these one day if there's enough demand,
but they're a long way off.

Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+





From harm at laatje.dhs.org  Thu Apr  4 10:01:43 2002
From: harm at laatje.dhs.org (harm)
Date: 4 Apr 2002 07:01:43 -0800
Subject: how to detect if a file is complete
Message-ID: <32f17666.0204040701.64dd0bfe@posting.google.com>

Hi all,

I want to copy some files from one dir to another using python. 
The files are copied onto my linux box via a samba-share.

I want to write a python script which checks a dir and copies all
files to another dir. But there is the possibility that the file is
not yet complete. (This are rather large Postscript files >100mb).
Therefore I need to detect if the file is complete.

The files are copied on the samba-share by Adobe Acrobat Distiller,
therefore i cannot save some check file like $FILENAME.complete...

Is there some other way (using python) to detect if the file is
complete?

Thanks,

Harm de Laat
Informatiefabriek
The Netherlands


From brian_l at yahoo.com  Sat Apr 20 03:55:14 2002
From: brian_l at yahoo.com (Brian Lenihan)
Date: 20 Apr 2002 00:55:14 -0700
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <afe3b146.0204192355.2a4b42b0@posting.google.com>

Shane Hoversten <srh232 at nyu.edu> wrote in message news:<3CC0D892.2060006 at nyu.edu>...
> Hi -
> 
> 	This is really dumb, but I just want to print out a character without a 
> newline and without a space behind it.  You can't do:

import sys
sys.write(c)


From martin at v.loewis.de  Mon Apr  1 03:20:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 10:20:32 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <3CA5F90F.BF37B528@engcorp.com> <m3ofh593zv.fsf@mira.informatik.hu-berlin.de> <mailman.1017618769.426.python-list@python.org>
Message-ID: <m3y9g7rfsf.fsf@mira.informatik.hu-berlin.de>

Kirill Simonov <kirill_simonov at mail.ru> writes:

> * Martin v. Loewis <martin at v.loewis.de>:
> > Can you show examples of (real!) code that will break? Preferably code
> > you've written yourself.
> 
> The changes will break the code like this:
> 
> if type(arg) == type(''):
>     ...
> elif type(arg) == type(0):
>     ...
> 
> You can find such code even in the standard library.

As a general example, I agree. I even know a place in PyXML that I
will need to change: For the better, it turns out, since it dispatches
by the type, but needed a special TruthValue class to support the
notion of boolean values in the underlying XML DTD.

I'm asking about *specific* code that will need to change. There are 8
occurrences of "type(0)" in the standard library. Of those *none* will
break. 

The occurrence in RExec could stand some improvement: either support
for the old signature could be entirely removed, or the type check
should be changed to an isinstance(args[-1],int) check. In the current
code, breakage would only occur if somebody passed a bool as the
verbose argument. Without seeing an application, this is hard to
establish; all usages within Python itself either let the paramater
default, or pass an integer (rexec.test does that).

So it is not clear that anything in the standard library will break.

Regards,
Martin


From bokr at oz.net  Mon Apr 22 16:01:54 2002
From: bokr at oz.net (Bengt Richter)
Date: 22 Apr 2002 20:01:54 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <mailman.1019483442.32389.python-list@python.org> <MqVw8.47908$vm6.7506303@ruti.visi.com>
Message-ID: <aa1q7i$nu7$0@216.39.172.122>

On Mon, 22 Apr 2002 14:50:20 GMT, grante at visi.com (Grant Edwards) wrote:

>In article <mailman.1019483442.32389.python-list at python.org>, brueckd at tbye.com wrote:
>>
>>> > You should understand that this is a philosophical viewpoint. The
>> [snip]
>>> To put this another way, ask yourself where you stand on this
>>> question: is mathematics discovered or invented?
>
Too big a question ;-) Was pi discovered or invented? I'd say discovered ;-)
I.e., I think there are fundamental things, and there are compositions. I would
tend to say that we discover fundamental things, but that we can make/invent endless
compositions. However, when we notice that one of our inventions has an interesting
property, that is a discovery ;-) Or perhaps a re-cognition.

>Very well put.
>
>> Is there a difference?
>
>Yes.  Science verifies it's theories by comparing them to the
Who is that anthropomorphized "Science" ?
>physical world which they are attempting to describe.  It
IMO, when you get to implementation details, comparing theories
to the physical world involves representing both in some medium
where they can be "inspected" for correspondence. This inspection
involves either a programmed process operating on representations
and yielding a true/false indication, or a careful mental version
of the same with associated consciousness of the process.

>doesn't matter how elegant, or internally consistent Science
>is.  If it doesn't successfully describe the physical world,
>it's wrong.
What "physical world" though? And what is a successful description?

IMHO describing the physical world is putting your description into
correspondence with some representation of observations of the physical
world. Those observation representations are not the physical world itself.
They are a product of a special relationship between observer, observation process
and world.

My point is that convincing yourself that something is true "about the world"
is really an operation with representations, and so is convincing yourself
that something is true in mathematics.

It is interesting to speculate on how our ability operate with mental
representations is implemented, and how consciousness arises in conjunction
with some of it, but that would make for a long thread ;-)

>
>Science has an external reference point. Mathematics does not.
How is a cup more like a donut than a bowl? IOW, ISTM many mathematical
concepts owe a lot to relationships we abstract from our perceptions of
the world. Indeed, it would not surprise me if we are hardwired for some
abstract operations, and that the hard-wiring depends on a mapping between
laws of the physical world and laws of thinking. E.g., I suspect that the continuity
of fields under the influence of discrete particles/effects is fundamental
to our unified conscious experience of discretely sensed world and body elements.

>Internal consistency is the only thing which mathematics can
>attempt to verify.  Mathematics is not an attempt to describe
>the physical world.
>
>> One could argue that the only difference between discovery and
>> invention is the arrogance of the person reporting it.
>
>One could.  But one would be missing the point.
>
One could also be missing _a_ point by dismissing a thought ;-)

IOW, when I invent something, I don't know how I do it, any more than
I know how "I" select and activate the muscles it takes for me to walk
or circulate my blood[1]. So, to say that "I" invent something is a kind of
foolish arrogance. It's more like a happy privilege to discover that my
mind/brain has produced a (to me) new combination of thought elements.
Of course, if someone else is playing with similar mental Lego pieces,
it's not unlikely that they will have a similar experience. (USPTO take note ;-)

[1] That's not to say they can't be identified objectively and scientifically
investigated. It's saying that I don't consciously know how "I" control my
body (if and when I do ;-)

Regards,
Bengt Richter


From aahz at pythoncraft.com  Wed Apr 17 19:26:48 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Apr 2002 19:26:48 -0400
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net>
Message-ID: <a9l0bo$h1j$1@panix1.panix.com>

In article <a9kjh1$qc5$1 at slb1.atl.mindspring.net>,
Andrew Dalke <dalke at dalkescientific.com> wrote:
>
>while += is defined as
>
>list_inplace_concat(PyListObject *self, PyObject *other)
>{
> other = PySequence_Fast(other, "argument to += must be iterable");
> if (!other)
>  return NULL;
>
> if (listextend_internal(self, other) < 0)
>  return NULL;
>
> Py_INCREF(self);
> return (PyObject *)self;
>}
>
>The code is almost identical, except for that return self part.
>(I don't know why that's needed.)

Heh.  That's because of the way __iadd__ works.  Immutable objects must
return a new object with __iadd__; mutable objects *MAY* return self
(and usually do).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From tim.one at comcast.net  Tue Apr 23 22:03:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 23 Apr 2002 22:03:11 -0400
Subject: Python's Lisp heritage
In-Reply-To: <3CC500AF.8BAE6F7E@cascade-sys.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEJAPAAA.tim.one@comcast.net>

[James J. Besemer]
> ...
> The original semantics (as formally defined in Lisp itself
> [McCarthy]) called for dynamic binding.

It seems this would have been a surprise to McCarthy <wink>:

    http://www-formal.stanford.edu/jmc/history/lisp/node4.html

See paragraph "d. Free variables".  When a user bumped into a function where
the difference between dynamic and lexical scoping mattered, and lexical
scoping was what they needed, "I must confess that I regarded this
difficulty as just a bug".  The understanding of these issues was much
muddier at the time.

> ...
> I find it highly ironic that Perl originally did it the wrong way
> ("local"), although later they introduced a now-recommended static
> binding ("my").

More recently, REBOL went down the same path.  Dynamic scoping is attractive
for implementation reasons; everyone who succumbs eventually regrets it
(except for RMS in elisp ...).





From cjw at sympatico.ca  Sat Apr  6 16:55:50 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Sat, 06 Apr 2002 16:55:50 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it>
Message-ID: <3CAF6EE6.7564CDF7@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020406/2928b1e0/attachment.html>

From bokr at oz.net  Thu Apr  4 19:01:59 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Apr 2002 00:01:59 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org> <a8gq78$hg$0@216.39.172.122> <3CAC658E.51B3A912@python.org>
Message-ID: <a8iphn$qq0$0@216.39.172.122>

On Thu, 04 Apr 2002 14:35:26 GMT, Guido van Rossum <guido at python.org> wrote:

>Bengt Richter wrote:
>> What if I wanted to subtype str and have a bool signifying URL-escaped or not,
>> yet still have normally functioning __len__() etc.?
>
>That's an absurd idea, right there with redefining + to mean *.
>
I thought absurd examples were a good thing ;-)

Regards,
Bengt Richter


From jon+usenet at unequivocal.co.uk  Wed Apr 24 19:11:45 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Wed, 24 Apr 2002 23:11:45 -0000
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com> <slrnacc59a.tg8.jon+usenet@snowy.squish.net> <MCxx8.45058$0x3.38185@atlpnn01.usenetserver.com> <aa7c52$ehu$0@216.39.172.122>
Message-ID: <slrnaceetc.3k8.jon+usenet@snowy.squish.net>

In article <aa7c52$ehu$0 at 216.39.172.122>, Bengt Richter wrote:
> So is cc.execute(...) acting like
> >>> def dummyccx(fmt, params):
>  ...     for tup in params:
>  ...         print fmt % tuple(map(repr,tup))
>  ...
> >>> dummyccx(
>  ... "INSERT INTO tmp (k, s) VALUES (%s, %s)",
>  ... [(1, "one"), (2, 'two'), (3, 'three')]
>  ... )
>  INSERT INTO tmp (k, s) VALUES (1, 'one')
>  INSERT INTO tmp (k, s) VALUES (2, 'two')
>  INSERT INTO tmp (k, s) VALUES (3, 'three')

Sort of that one. Except it isn't 'repr', it's a function which
examines the type of the parameter and converts it into appropriate
MySQL representation (i.e. for strings put "'" at the beginning and
end and quote various characters with a "\", for numbers convert them
to strings, etc etc).

Oh, except you shouldn't be relying on the 'for tup in params', you
should imagine it is doing:

  def dummyccx(fmt, params):
    print fmt % tuple(map(sql_quote, params))

Except it's more complicated than this because you might be using a
mapping not a sequence for the parameters. See
http://www.python.org/topics/database/DatabaseAPI-2.0.html
for details.


From ajs at ix.netcom.com  Tue Apr  2 00:06:08 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 2 Apr 2002 00:06:08 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <000701c1da04$1a4f1800$0334fea9@carol>

>>Inventing new, non-standard terminology in order to be >>"simpler"  is
rarely helpful; it doesn't substantially help  >>newbies, and it only serves
to confuse people actually >>involved.

>This is born out by experimental psychologists, e.g.

Also interesting is group psychology.

I was pillared on Edu-Sig for making this point.

Of course I was referencing Alice and Guido's Linux
journal article - not chiming in support of his insight.

Art




From philh at comuno.freeserve.co.uk  Wed Apr 17 08:38:50 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 17 Apr 2002 13:38:50 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu> <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk> <kq5v8.59911$3L2.5192980@bin7.nnrp.aus1.giganews.com>
Message-ID: <slrnabqr6q.ll2.philh@comuno.freeserve.co.uk>

On Wed, 17 Apr 2002 02:50:56 GMT, Terry Reedy <tejarex at yahoo.com> wrote:
>
>"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
>news:slrnabp9cs.ipr.philh at comuno.freeserve.co.uk...
>> >Thanks for proposing this. I'm sure you'll get lots of horrified
>> >reactions
>>
>> I don't see why: people who like pass can still use it.
>
>I am not horrified, however, I do have two objections:
>
>1. We will have two styles of Python to read: with and without pass.
>I expect this would be confusing, especially since one would not know
>for sure until one sees a pass.

I don't see that as a serious problem. We already have as many 
styles of python as there are python programmers. For example, I 
have never used lambda. As another example, some people prefer short
variable names nad others long ones, t versus totalCount. Etc.

>2. Newcomers will wonder why we have it if it is not required.  Then
>they will wonder "should I use it or not" and post requests for
>advice.

Do we get newbies posting asking which is better of:

   if some_condition:
      do_something()

or:

   if some_condition: do_something()

If not, why would they ask for pass?

>It would have been ok with me, I think, if the language had allowed
>null suites from the beginning, but it is now to late, practically
>speaking, to remove pass.

That's true, but I AM NOT PROPOSING TO REMOVE IT.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From ldw at us.ibm.com  Wed Apr  3 10:52:07 2002
From: ldw at us.ibm.com (Larry Whitley)
Date: Wed, 3 Apr 2002 09:52:07 -0600
Subject: Threading problem
Message-ID: <a8f8f6$vdu$1@news.rchland.ibm.com>

I have a PC with two processors at work and have decided to try to use
threading to run some long running console applications.  My basic idea is
to start two threads, one for each processor, and give them a list of things
to do.  As each thread finishes it one job will go to the list and pick off
the next thing to do.  When the list is empty, both threads are done.

Here's a snippet of the code that inherits from threading.Thread:

class MyThread(threading.Thread):
    def __init__(self, thName, jobList, jobLock):
        threading.Thread().__init__(self, name=thName)
        self.jobList = jobList
        self.jobLock = jobLock


My code is failing in first statement of the constructor, you can see the
error below.


X:\Projects\Pecos\ttc6\thsp>run.py
Traceback (most recent call last):
  File "X:\Projects\Pecos\ttc6\thsp\run.py", line 59,
    main()
  File "X:\Projects\Pecos\ttc6\thsp\run.py", line 46,
    th1 = MyThread( "Thread1", jobList, jobLock )
  File "X:\Projects\Pecos\ttc6\thsp\run.py", line 9, i
    threading.Thread().__init__(self, name=thName)
  File "C:\Python22\lib\threading.py", line 355, in __
    assert group is None, "group argument must be None
AssertionError: group argument must be None for now

The thing that puzzles me is that threading.Thread() is insisting that
group=None.  That's ok with me, and I haven't changed it, yet it's
complaining as if I did.

What am I doing wrong?

Larry
Running Python 2.2




From opengeometry at NOSPAM.yahoo.ca  Tue Apr 23 13:28:36 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 23 Apr 2002 17:28:36 GMT
Subject: Mass Mailing help needed (Paul Rubin)
References: <mailman.1019325819.29016.python-list@python.org> <MPG.172e025d803fc2009896bf@news.nwlink.com> <mailman.1019553582.18769.python-list@python.org> <slrnacarct.3s6.akuchlin@ute.mems-exchange.org>
Message-ID: <aa45k4$7kjio$2@ID-99293.news.dfncis.de>

A.M. Kuchling <akuchlin at ute.mems-exchange.org> wrote:
> In article <mailman.1019553582.18769.python-list at python.org>, 
>        Graham Ashton wrote:
>> With that approach you have to open one TCP connection per message. This
>> is expensive, especially as your SMTP daemon should be doing all sorts
>> of checks on your IP to see if it'll happily accept mail from you. Not
> 
> It's certainly possible to send multiple messages over a single SMTP
> connection, and IP checks can be loosened by tweaking the SMTP
> server's configuration.  For example, it would be easiest and most
> secure to disable address checking on connections from localhost, and
> then run your bulk mailing software on the same machine as the SMTP
> server.  We've had to do this in order to make a 1700-member mailing
> list run acceptably; without disabling the address checks, each
> message to the list resulted in 1700 DNS lookups before the message
> was even accepted.

If 1700 recipients are all different domain, then Sendmail will check for
MX or A record.  But, you can run your own DNS server (either cache-only or
forward-only) to speed up lookup for previously encountered domains.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8 CPU cluster, NAS, (Slackware) Linux, Python, LaTeX, Vim, Mutt, Tin


From newsfroups at jerf.org  Thu Apr 25 12:20:31 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Thu, 25 Apr 2002 16:20:31 GMT
Subject: Freeing memory
References: <3CC7F1F7.8B987146@kfunigraz.ac.at> <y2Ux8.13578$z55.7087@atlpnn01.usenetserver.com> <3CC80EBE.E9735284@kfunigraz.ac.at>
Message-ID: <3CC82D49.9070208@jerf.org>

Siegfried Gonzi wrote:
> I just start the script from the idle editor: 'Run script'.
> 
> I even tried to use del(list1), del(list2),... but it won't help. I also
> write the results to a file (and not to the output screen) in order not
> to interference the calculation.

You may have some cyclic references in your dataset. What version of 
Python are you using?

IDLE always has the same Python interpreter instance. If you can run the 
script with no parameters like that, I'd re-recommend running directly 
from a command line or from a double-click on some script file (as Steve 
Holden did). Then there's no chance that Python will hold the memory 
after execution (or the OS wouldn't be doing it's job).

The alternative is to figure out what's going on and make careful use of 
  weak references and more careful memory handling; just making sure to 
run in a seperate instance should be enough.



From akuchlin at ute.mems-exchange.org  Tue Apr 23 13:15:06 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 23 Apr 2002 17:15:06 GMT
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org> <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com>
Message-ID: <slrnacb5ko.4i6.akuchlin@ute.mems-exchange.org>

In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>, 
	Steve Holden wrote:
> Since urllib knows nothing of cookies, you will need to integrate some sort
> of a cookie jar into the library, with a new API  for the clients to
> retrieve and store the cookies.

This is worthwhile, but I don't think it belongs in urllib.  It
belongs in a module or package of its own that provides general
Web-browser features such as cookies, remembering authentication
usernames and passwords, and a cache.  This package could then be used
for implementing HTML-scraping scripts, spiders, or a Web browser.

--amk                                                             (www.amk.ca)
Ph-nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn. 
    -- H.P. Lovecraft, "The Call of Cthulhu"



From siegfried.gonzi at kfunigraz.ac.at  Mon Apr 22 12:53:00 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 22 Apr 2002 18:53:00 +0200
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
	 <mailman.1019299357.11513.python-list@python.org>
	 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
	 <Xns91F67C40EB0E0lumberjack@216.148.53.83>
	 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
	 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
	 <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org>
Message-ID: <3CC43FEC.EE94FFEB@kfunigraz.ac.at>

Lulu of the Lotus-Eaters wrote:
> 
> |> mathematics, including the study of algorithms, is a natural science.
> 
> Tim Daneliuk <tundra at tundraware.com> wrote previously:
> |I disagree strongly...  All of mathematics is a formal construct of the
> |human mind created with the intent of removing ambiguity and enhancing
> |our ability to describe what we *think* and what we *observe*.
> 
> To my mind, Daneliuk present here the strongest argument that
> mathematics (and therefore CS) *IS* indeed a science.

How do you observe infinity in mathematics?: n+1,n+2,...

> Alex may disagree with me here, but I think my opinion here is quite
> close to the middle Wittgenstein, or the _Philosohy of Mathematics_ and
> the _Blue and Brown Books_.  It is also very close to some of Godel's
> remarks, in his philosophical moods.

Wittgenstein dissmised the pure behavior of mathematicians: proof n, do
not rest and proof n+1, do not rest and proof n+2,...
Wittgenstein and Goedel met in the so called "Wiener Kreis" (an exquiste
circle --on a regular basis-- for intellectuals in Vienna/Austria).

Goedel had direct influence (via v. Neumann) into: I dare to say
computer science, instead I write "computer engineering".


S. Gonzi


From gerhard at bigfoot.de  Tue Apr  9 22:54:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 10 Apr 2002 02:54:17 GMT
Subject: Is Python Sticky?
References: <3cb3a8dd$1@nntp01.splitrock.net>
Message-ID: <slrnab7aap.3ir.gerhard@lilith.my-fqdn.de>

ball at righty.eiu-newman.org wrote in comp.lang.python:
> Hello!
> 
> For a while "Learn Python" has been on my list of things to do.  Right
> now I have an application that could use some glue between a web based
> interface (in its simplest form a single page with a couple of buttons
> on it), and a shell script (I'm running NetBSD).  Would Python be
> suitable for this kind of thing?

Python is an excellent choice both for writing web applications and
driving web-based interfaces.

> Are there other pieces of the puzzle that I'll need?

All you'll need should be in the standard Python library.

> Any pointers would be most welcome.

I'd head to the "Newbie" page at www.python.org and work thru one of
the tutorials. From the standard Python library, you'll need the cgi
moduel for writing CGI scripts, and urllib or urllib2 for controlling
a CGI script from your Python script.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From abagaric at rest-art.hr  Tue Apr 16 12:33:49 2002
From: abagaric at rest-art.hr (Ante)
Date: Tue, 16 Apr 2002 18:33:49 +0200
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca>
Message-ID: <a9hjvc$cdq$1@as201.hinet.hr>

>
> do some reading about scope... add the following line to dummy.py:
>

Um.. and shouldn't one read the article before replying?

If this is solely a scope problem, then why does it work if you replace
'num += 1' with 'print num'?

Shouldn't it say 'referencing local before assignment yada yada yada'
anyway?

My question was: why does it work if you only use the variable, and doesn't
work if you assign to it? I guess it would actually make sense for plain
assignment, but for += operator, it doesn't seem so.





From jeff at ccvcorp.com  Tue Apr  2 13:59:17 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 10:59:17 -0800
Subject: memory usage
References: <mailman.1017604009.31777.python-list@python.org>
Message-ID: <MPG.17139f61d4f2a565989689@news.nwlink.com>

In article <mailman.1017604009.31777.python-list at python.org>, 
abou at cam.org says...
> Is there a specific way to clean the memory used by python scripts?
> While debugging my project, I always get a 'not enough memory to run
> this program' after running it a few times (Win98).

Unfortunately, this is probably a problem with the Windows memory 
manager, having nothing to do with Python.  Python does its own 
garbage collection, and once you've released all references to 
your objects (which should happen automatically when Python shuts 
down), all that memory is freed.  However, Python can't guarantee 
that the underlying memory manager will actually *do* anything 
with that freed memory.  

So, your options are:  1) Make sure that you're actually 
terminating each debugging session before trying to start a new 
one.  2) When this error happens, reboot Windows and then try 
again.  3)  Buy more memory -- it's cheap.  (I've never run out 
of memory, even when I only had 64MB, but then, I haven't been 
running many memory-intensive scripts.)  4) Use a different OS.  
The Win9x memory manager is particularly poor.  If you can 
upgrade to an NT-based Windows, you'll be much better off.  

-- 

Jeff Shannon
Technician/Programmer
Credit International


From paul at boddie.net  Wed Apr 24 11:42:08 2002
From: paul at boddie.net (Paul Boddie)
Date: 24 Apr 2002 08:42:08 -0700
Subject: Quick question about Webware
References: <yUvx8.30$RNd.170832384@news.telia.no>
Message-ID: <23891c90.0204240742.4caa32d4@posting.google.com>

"Thomas Weholt" <thomas at gatsoft.no> wrote in message news:<yUvx8.30$RNd.170832384 at news.telia.no>...
> When you create a servlet using WebWare is it loaded on each request or is a
> instance kept in memory?

That's "Webware for Python", by the way. ;-)

As far as I know, several threads are instantiated on startup and when
a "request path" is resolved inside one of those threads, the servlet
factory creates a servlet instance which I would hope would stick
around after the request has been processed. Having said that, I
wasn't sure whether this was implemented within the framework itself,
and when I wrote my own servlet factory once, I decided to implement
"instance caching" within that factory.

I don't know whether other servlet factories implement caching
themselves, although a quick glance at the PythonServletFactory class
indicates that some kind of cache is in use.

> I mean, what if the servlet uses some module or piece of code that takes a
> while to load, how is this solved in WebWare?

Per-servlet-factory-instance (ie. per-thread) caches seem to be used.
One of the Webware mailing lists (webware-discuss, perhaps) might be a
better place to ask these kinds of questions, though.

Paul


From josegomez at gmx.net  Mon Apr  8 07:26:00 2002
From: josegomez at gmx.net (josegomez at gmx.net)
Date: Mon, 8 Apr 2002 11:26:00 +0000 (UTC)
Subject: How do I generate 1/f (pink) noise in Python
References: <3CB17354.60609@mxm.dk>
Message-ID: <a8ruo8$bd0$1@hermes.shef.ac.uk>

Max M <maxm at mxm.dk> wrote:
> I am dabling in algorithmical composition and at times I need to 
> generate random values. Often the values should have the characteristics 
> of pink noise (1/f).

    I am not too sure of this. Certainly, it looks more like an    
application of Numeric, but you can get a relatively easy to
impelement algorithm  from
<http://www.firstpr.com.au/dsp/pink-noise/>. Basically, you generate
white noise (uniform distribution), and then you filter it to produce
pink noise.

    Hope that is of some help,
    Jose

-- 
Jos? L G?mez Dans			PhD student
Tel: +44 114 222 5582			Radar & Communications Group
FAX; +44 870 132 2990			Department of Electronic Engineering
					University of Sheffield UK


From greg at cosc.canterbury.ac.nz  Tue Apr 23 01:47:26 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 23 Apr 2002 17:47:26 +1200
Subject: Proper syntax for adding methods to a class instance dynamically?
References: <aa2ahm$8pr$1@peabody.colorado.edu>
Message-ID: <3CC4F56E.DD7B4BC6@cosc.canterbury.ac.nz>

Fernando P?rez wrote:
> 
> Docstring:
>     Functions to create new objects used by the interpreter.
> 
>     You need to know a great deal about the interpreter to use this!
> 
> Is there a simpler, perhaps 'safer' way to achieve the same result?

Nothing simpler, and there isn't really anything "dangerous"
about new.instancemethod. If you're worried that the details
might change in a future version of Python (which is possible,
since the functions in the new module depend somewhat on the
internals) you could do this:

   def create_instance_method(inst, func):
     c = inst.__class__
     c.temp_method = func
     meth = inst.temp_method
     del c.temp_method
     return meth

   def f(self):
     print "Hi, I'm", self

   class C:
     pass

   i = C()
   i.new_method = create_instance_method(i, f)
   i.new_method()

> I'd also like to know how to add them to the whole class.

That's easy -- just assign a plain function to a class
attribute, e.g.

  C.some_method = f

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


From phr-n2002a at nightsong.com  Tue Apr 30 22:55:29 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 30 Apr 2002 19:55:29 -0700
Subject: crypto
References: <mailman.1020218462.23651.python-list@python.org>
Message-ID: <7xu1psshjy.fsf@ruckus.brouhaha.com>

"Mark McEahern" <marklists at mceahern.com> writes:
> >   http://www.nightsong.com/phr/crypto/crypto.txt
> 
> Wow, I really like the philosophy behind this approach.  What kind of
> license are you envisioning for this?  I'd be interested to try it out.

It will be GNU GPL, unless you'd like to fund development ;-).

Regards

Paul


From hgs at dmu.ac.uk  Thu Apr 18 10:55:47 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 15:55:47 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <23891c90.0204180534.5b3c8542@posting.google.com>
Message-ID: <Pine.GSO.4.40.0204181549540.24683-100000@neelix>

On 18 Apr 2002, Paul Boddie wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> wrote in message news:<mailman.1019123256.19883.python-list at python.org>...
> >
> > What else does it look for?  If I symling /usr/local/bin/ld to gld will
> > configure find it?
>
> You might want to look at config.log to see what is going on "behind
> the scenes". It might help even more if you were to pass the -v flag
> to configure. In the log, I think it should indicate at which line the

Seems to be using gnu ld.

configure:2083: checking for opendir in -ldir
configure:2102: gcc -o conftest -g -O2   conftest.c -ldir   1>&5
/usr/local/sparc-sun-solaris2.7/bin/ld: cannot find -ldir
collect2: ld returned 1 exit status
configure: failed program was:
--
configure:3355: checking for shl_load in -ldld
configure:3374: gcc -o conftest -g -O2   conftest.c -ldld  -ldl  1>&5
/usr/local/sparc-sun-solaris2.7/bin/ld: cannot find -ldld
collect2: ld returned 1 exit status
configure: failed program was:
--

[further examples trimmed]

GNU ld version 2.12
Copyright 2002 Free Software Foundation, Inc.

>
> I just hope that the whole linking thing isn't some
> dark-side-of-the-force libtool-determined configuration, otherwise you

I don't know.
>
> Paul
>
        Hugh





From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Apr 30 19:51:09 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Wed, 1 May 2002 01:51:09 +0200
Subject: Pyro mailing list
Message-ID: <aanald$nph$1@news1.xs4all.nl>

Hi

If you're interested in Pyro (see http://pyro.sourceforge.net - it is
a distributed objects system for Python), you might also want to
subscribe to the Pyro mailing list, to be the first to
get all the news, discussions, suggestions and tips!

Please visit   http://lists.sourceforge.net/lists/listinfo/pyro-core

Thank you
Irmen de Jong

PS   only a few people have subscribed until now, that's
   the reason for this little promotion ;-)



From c.f.a.johnson at rogers.com  Wed Apr 10 14:47:27 2002
From: c.f.a.johnson at rogers.com (Chris F.A. Johnson)
Date: 10 Apr 2002 18:47:27 GMT
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <u4rijals0.fsf@python.net>
Message-ID: <slrnab925v.ej4.c.f.a.johnson@xword.rogers.com>

In article <u4rijals0.fsf at python.net>, Michael Hudson wrote:
> "Emile van Sebille" <emile at fenx.com> writes:
> 
>> Kalle Svensson
>> >
>> > "import this", IIRC.
>> >
>> 
>> 
>> Oooh... that's almost as much fun as saying 'xyzzy' the first time...
>> ;-)
> 
> That's not as much fun as "syzygy" (which is a real word,
> astonishingly enough).

    But still not as much fun as [the results of] zymurgy!

-- 
	Chris F.A. Johnson                              bq933 at torfree.net
	=================================================================
	c.f.a.johnson at rogers.com                http://cfaj.freeshell.org
	cfaj at freeshell.org        http://members.rogers.com/c.f.a.johnson


From sag at hydrosphere.com  Mon Apr 15 14:48:12 2002
From: sag at hydrosphere.com (Sue Giller)
Date: Mon, 15 Apr 2002 12:48:12 -0600
Subject: COM-Python thread callbacks
Message-ID: <20020415184227497.AAA287@mail.climatedata.com@SUEW2000>

Well, I have another question about using python COM objects with 
VB.

My application uses a VB front end that creates a Python com 
object.  The VB app passes to the Python object a VB callback 
object (class object with a single function that displays a message).

The python COM object is then called to do something. The Python 
COM object then creates a thread (threading.Thread) object that 
creates a file and writes some stuff to it, passing to the thread a 
python callback function.  When the thread ends, it uses the python 
callback fn in the python COM object to pass up a message about 
what it did.  The python callback function then passes that message 
back to the VB front end via the VB callback object, using the VB 
object's single function.  What is being passed around in a simple 
text string.

When the VB front end gets the message, it displays it in a status 
box.

Now for the problem.  If I run this application from within the VB 
IDE, everything works just fine.  But, when I compile the VB 
application into an exe and run it from the exe, the application 
abruptly closes without putting up the returned message, sometimes 
leaving the file in an open state.  The same abrupt ending will occur 
in the VB IDE if I put a break at the status box assignment line and 
then step thru the line.  It will display the msg and then close 
abruptly.

I am not an expert on COM or threads, so I am wondering if my 
design for the callback functions is the correct one, and if anyone 
has any ideas why this works in the IDE and not as an EXE.  I use 
the two levels of callbacks because I got GPFs from VB when I tried 
to pass the VB callback directly on to the thread.

Pseudo code (real code is long):

' in VB app
set VBCallBackObj = new clsCallBack
set PYWorkObj = CreateObject("Python.worker")
' pass the callback once to the python object
PYWorkObj.SetCallBack(VBCallBackObj) 
' now have COM object do something
PYWorkObj.DoSomething(inputArgs)
# python object creates the thread and does its thing, passing in its 
callback function
athread = theThread(inputArgs, self.CallBack)
athread.start()
# the thread gets done and calls back up to python object
PYWorkObj.CallBack(msg)
# the python call back function calls up to the VB object using known 
callback function name in the object, CallMe()
VBCallBackObj.CallMe(msg)


# python code snippets
# in the python com side, in the PYWorkObj
def SetCallBack(self, CallBackObj):
    """ called from outside with VB object """
    # callback obj is a vb object with a callback function named CallMe
    import win32com.client
    self._callBack = win32com.client.Dispatch(CallBackObj)

def CallBack(self, msg):
    """ a callback for the thread to signal when its done """
    self._callBack.CallMe(msg)

def DoSomething(inputArgs):
    inputargs = ProcessSomehow(inputArgs)
    currentThread = theThread(inputargs, self.CallBack)
    currentThread.start()

class theThread(threading.Thread):
    """ one thread for each job """        
    def __init__(self, args, CallBackFn):
        threading.Thread.__init__(self)
        self.args = args
        self._callBack = CallBackFn

    def run(self):
        import pythoncom
        pythoncom.CoInitialize()
        file, value = doJob(self.args)
        pythoncom.CoUninitialize()
        self._callBack("Job Done")




From ljz at asfast.com  Tue Apr 16 14:24:21 2002
From: ljz at asfast.com (Lloyd Zusman)
Date: Tue, 16 Apr 2002 14:24:21 -0400
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca>
 <a9hjvc$cdq$1@as201.hinet.hr>
 <mailman.1018979135.19771.python-list@python.org>
Message-ID: <m2wuv7ijsa.fsf@asfast.com>

Slight correction ...

Paul Osman <paul.osman at sympatico.ca> writes:

> [ ... ]
>
> <code>
> num = 5
> def dummy():
>     newnum = num + 5
>     return num
> </code>
> 
> [ ... ]

I think that you meant to type this:

    return newnum

-- 
 Lloyd Zusman
 ljz at asfast.com


From aahz at pythoncraft.com  Wed Apr 10 15:56:46 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 15:56:46 -0400
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <yu99662zmufg.fsf@europa.research.att.com> <rxZs8.254$p56.69633@newsb.telia.net> <yu99vgazifhp.fsf@europa.research.att.com>
Message-ID: <a925du$lj6$1@panix1.panix.com>

In article <yu99vgazifhp.fsf at europa.research.att.com>,
Andrew Koenig  <ark at research.att.com> wrote:
>
>Fredrik> you could of course use a RE-like approach:
>
>Yes, that's one possibility.

You say that as though you don't really like the idea.  But that *is*
the Pythonic approach: return an object that you can probe to get the
appropriate information.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From loewis at informatik.hu-berlin.de  Tue Apr 16 05:32:03 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Apr 2002 11:32:03 +0200
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE> <m3wuv8iras.fsf@mira.informatik.hu-berlin.de> <a9fj34$3mi$1@nets3.rz.RWTH-Aachen.DE> <m3ofgkkugm.fsf@mira.informatik.hu-berlin.de> <a9gkn6$nft$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <j4y9fokmzw.fsf@informatik.hu-berlin.de>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:

> > > What do I need to do? I already installed PyXML.
> >
> > What do you want to do today?
> >..
> 
> I want to port the QML Interpreter ( http://questml.com ) to Python for a
> start -- if I find a working XPath handler.

I see. In that case, you don't need to install anything else:
xml.xpath is at your service.

Regards,
Martin



From tl_news at nexgo.de  Wed Apr 10 01:40:38 2002
From: tl_news at nexgo.de (Tino Lange)
Date: Wed, 10 Apr 2002 07:40:38 +0200
Subject: RELEASED email 2.0 (standalone)
References: <mailman.1018328523.25609.clpa-moderators@python.org>
Message-ID: <eqj7bu0uc91hp5obvs0e7geuqo33usdfo7@4ax.com>

On Mon, 8 Apr 2002 17:43:36 -0400, barry at zope.com (Barry A. Warsaw)
wrote:

>
>I've just released version 2.0 of the standalone email package.  This
>will likely be the last official standalone release, as I intend to
>complete the merge of the package into the Python 2.3 cvs tree.

>The email package is compatible with both Python 2.1 and Python 2.2.
>Online documentation is available at

Hi Barry!

I've installed it on my Python 2.1.x - boxes (Win and Linux).
But the test fails and there are deprecation warnings?!

Here's the test_email.py - output:

tinux:/usr/local/src/email-2.0/tests # python test_email.py
........................./usr/local/lib/python2.1/site-packages/email/MIMEText.py:43:
DeprecationWarning: _encoder argument is obsolete.
  DeprecationWarning, 1)
..........test_email.py:483: DeprecationWarning: add_payload() is
deprecated, use attach() instead.
  self.msg.add_payload("""\
.............test_email.py:664: DeprecationWarning: add_payload() is
deprecated, use attach() instead.
  container.add_payload(intro)
test_email.py:665: DeprecationWarning: add_payload() is deprecated,
use attach() instead.
  container.add_payload(image)
......./usr/local/lib/python2.1/site-packages/email/Utils.py:145:
DeprecationWarning: Use Header.Header.encode() instead.
  warnings.warn('Use Header.Header.encode() instead.',
DeprecationWarning, 1)
.test_email.py:816: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s), 'this is some text')
test_email.py:818: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s), u'Keld J\xf8rn Simonsen')
test_email.py:821: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s), 'If you can read this you understand the
example.')
test_email.py:823: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s),
test_email.py:826: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s), u'this issome text')
test_email.py:828: DeprecationWarning: Use Header.decode_header()
instead.
  eq(Utils.decode(s), u'this is some text')
..test_email.py:965: DeprecationWarning: add_payload() is deprecated,
use attach() instead.
  msg.add_payload(msg1)
test_email.py:966: DeprecationWarning: add_payload() is deprecated,
use attach() instead.
  msg.add_payload(msg2)
.test_email.py:869: DeprecationWarning: add_payload() is deprecated,
use attach() instead.
  m.add_payload('Here is the body of the message.\n')
....................FF..........................test_email.py:1634:
RuntimeWarning: test_japanese_codecs() skipped
  warnings.warn('test_japanese_codecs() skipped', RuntimeWarning, 1)
.....
======================================================================
FAIL: test_formatdate (__main__.TestMiscellaneous)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test_email.py", line 1180, in test_formatdate
    time.gmtime(now)[:7])
  File "/usr/local/lib/python2.1/site-packages/unittest.py", line 273,
in failUnlessEqual
    raise self.failureException, (msg or '%s != %s' % (first, second))
AssertionError: (2002, 4, 10, 5, 35, 32, 0) != (2002, 4, 10, 5, 35,
32, 2)
======================================================================
FAIL: test_formatdate_localtime (__main__.TestMiscellaneous)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test_email.py", line 1186, in test_formatdate_localtime
    time.localtime(now)[:7])
  File "/usr/local/lib/python2.1/site-packages/unittest.py", line 273,
in failUnlessEqual
    raise self.failureException, (msg or '%s != %s' % (first, second))
AssertionError: (2002, 4, 10, 7, 35, 32, 0) != (2002, 4, 10, 7, 35,
32, 2)
----------------------------------------------------------------------
Ran 112 tests in 1.968s

FAILED (failures=2)
tinux:/usr/local/src/email-2.0/tests #

What's wrong? Why those failures and warnings?
Best regards

Tino



From calves at coelce.com.br  Wed Apr 10 10:26:27 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Wed, 10 Apr 2002 11:26:27 -0300
Subject: RES: Create Excel Files without COM
Message-ID: <29A97D00F387D411AC7900902770E14804BE52DE@lcoeexc01.coelce.net>

If so, why not write a text file, wich could read by MS-Excel?

> -----Mensagem original-----
> De: Thomas Guettler [mailto:st-newsgroups at thomas-guettler.de]
> Enviada em: quarta-feira, 10 de abril de 2002 10:59
> Para: python-list at python.org
> Assunto: Create Excel Files without COM
> 
> 
> Hi!
> 
> Is there something like WriteExcel from perl in python?
> I need to generate an MS-Excel file. I don't want to use
> COM although it will run on windows.
> 
> The data is simple. Just text in columns and rows.
> 
>   thomas
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020410/b23f642e/attachment.html>

From duncan at NOSPAMrcp.co.uk  Tue Apr 30 03:59:27 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 07:59:27 +0000 (UTC)
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1> <mailman.1020100431.16386.python-list@python.org>
Message-ID: <Xns92005AC7EB928duncanrcpcouk@127.0.0.1>

[posted and mailed]

holger krekel <pyth at devel.trillke.net> wrote in
news:mailman.1020100431.16386.python-list at python.org: 

>> > Btw, what is the best/shortest way to express 'filter' and 'reduce'
>> > with list comprehensions?
<snip>

> You really feel that these free floating strings with
> no clear separations and lots of micro-management where
> you dont't even know what does what and where the parts
> start to float around and you ask yourself 
> when the sentence will ever end 
> 
I didn't say that. The question was what was the best/shortest way to 
replace the functions with list comprehensions. I answered the question. I 
did not say anywhere that they were always better than the functional 
alternative.

There are times when map/filter/reduce are the obvious ways to do things. 
What I did say, but you trimmed, was that I believe that if writing an 
explicit for loop removes the use of a lambda, then that is an improvement.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From phr-n2002a at nightsong.com  Fri Apr  5 20:32:00 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 17:32:00 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org>
Message-ID: <7x1ydt8ve7.fsf@ruckus.brouhaha.com>

Cliff Wells <logiplexsoftware at earthlink.net> writes:
> Ah, you must have chose APL or Forth, since they're pretty much in the
> feature-lock stage.  It can't be C since they've also added the damn
> boolean. 
> 
> Seriously though, if this is a big concern for you (and it might well be),
> target the 1.5.2 language spec and the chances of your code not working
> even in 2.2 are slim, so I'm not sure what would be compelling enough to
> make you abandon Python.  I myself don't like every change in Python (or
> even some things that have been in place since 1.5.2) but it's still an
> excellent language.  I suspect you are masking some other complaint with
> this one.

The issue isn't that change happens, but that it happens too damn
frequently.  Python has added new language features in 2.0, 2.2, and
now 2.3, and I think they're all within a year or so of each other.
C goes for a decade at a time without changes like that.

I like Python a lot but I think it falls a little bit short of truth
in advertising, when it calls itself a stable, mature language.
Mature to me means I can download someone's interesting new Python
script and not have to install a new version of Python in order to run
it.  That means the language shouldn't sprout new features so quickly
(new library modules is fine) and should try hard to not make
backward-incompatible changes.

Most users NEVER upgrade their system software and I try to never
upgrade mine.  The only time I normally get new system software is
when I buy a new computer (every 2-3 years, say) and install a current
OS on it from scratch.  Stable, mature software should be able to run
for that long without needing to be upgraded.  I find just can't
expect that from Python.


From jamescalthoff at yahoo.com  Sat Apr  6 14:12:18 2002
From: jamescalthoff at yahoo.com (James Althoff)
Date: Sat, 6 Apr 2002 11:12:18 -0800
Subject: PEP 285: Adding a bool type
Message-ID: <001b01c1dd9e$fab63160$54dac13f@Pacbell.net>

[Laura Creighton]
> Your boss comes in and says, oh, Mark, we need a change to that
> isValid function.  We need to return 3 states, Valid, Invalid, and
> Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
>
> What do you do now?

Returning a third value from isValid, "-1 indicates that an account has been
suspended", for example, seems like a bad idea because all code that tests
"if isValid" (as you recommend) would still allow suspended accounts to work
since -1 passes the "is something" test.  Since the calling code has to be
changed anyway, why not design it better?

Personally, I'm not a big fan of special codes: 0 means this, 1 means
that, -1 means something else, 999 means yet another thing, etc.

Jim







From david at thekramers.net  Fri Apr  5 09:57:24 2002
From: david at thekramers.net (David Kramer)
Date: Fri, 5 Apr 2002 09:57:24 -0500
Subject: newbie: declare function before definition?
Message-ID: <Pine.LNX.4.44.0204050953040.17812-100000@kramer.thekramers.net>

Disclaimer: 16 years programming experience, 1 week experience with 
Python.

It appears to me that in Python, functions need to be defined before they 
are called (physically in the file, I mean).  This means any piddly little 
utility functions have to be before my main routine, and the reader has to 
wade through them before getting to the main routine.

In some other languages, one can declare a function, and define it later, 
and the compiler is happy.  Is there a way to do this in Python?

Here is an example of what I'm talking about:

[dkramer at fbsddev02 devel]$ cat deftest1
#!/usr/remote/bin/python

def MyFuncBefore ():
    print "In Before"

print "Before"
MyFuncBefore()
print "After"
MyFuncAfter()

def MyFuncAfter ():
    print "In After"



[dkramer at fbsddev02 devel]$ deftest1
Before
In Before
After
Traceback (most recent call last):
  File "./deftest1", line 9, in ?
    MyFuncAfter()
NameError: name 'MyFuncAfter' is not defined


-------------------------------------------------------------------
DDDD   David Kramer                           http://thekramers.net
DK KD
DKK D  As a SysAdmin, yes, I CAN read your e-mail, 
DK KD  but I DON'T get that bored!
DDDD   



From tejarex at yahoo.com  Sat Apr 13 16:55:18 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 13 Apr 2002 20:55:18 GMT
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com>
Message-ID: <WW0u8.27978$iw.2540856@bin4.nnrp.aus1.giganews.com>

<bvdpoel at uniserve.com> wrote in message
news:3CB884C7.EE7E04D at uniserve.com...
> I have a list like this:
>
> [ [0, x, [0, y, [4, n] ...]

I am assuming that you mean like this:

[ [0,x], [0,y], [4,n], ...]

If so, it is *possible* that the built in [].sort(cmpfunc) will work
for you *if* you supply the optional custom compare function cmpfunc()
such as:

def mycompare(a,b): return cmp(a[0],b[0])

However, read the reference manual for the Python version you are
using and test.  I believe (but may be wrong) that .sort() is now
order preserving but was not always.  If not, you will have to find
one or devise a hack to make the sort keys unique.  I believe that one
of a[0]<b[0] or a[0]<=b[0] may work, but again check specific
reference.

Terry J. Reedy





From skip at pobox.com  Wed Apr  3 07:14:14 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 3 Apr 2002 06:14:14 -0600
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
References: <mailman.1017795845.3464.python-list@python.org>
        <r1tq8.161131$af7.78459@rwcrnsc53>
        <3CAAB238.290440C5@kfunigraz.ac.at>
        <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
Message-ID: <15530.61974.34351.747550@12-248-41-177.client.attbi.com>

    Jan> On the other hand, all the calculation that you cannot express
    Jan> using high-level primitives (using matrices, vectors,...) and has
    Jan> to be performed with loops element by element, is painfully
    Jan> slow. Equally slow in Python, as in IDL, Matlab,... It is so slow
    Jan> that it is unusable, it is very difficult to develop an algorithm
    Jan> if it takes days to process any reasonably sized input. My solution
    Jan> is to rewrite the critical part of the code in C/C++, I did this
    Jan> for Matlab, Python+Numeric, as well as IDL, but this is awkward -
    Jan> it forces you to work simultaneously with two languages and to deal
    Jan> besides the algorithm itself with a lot of implementation and
    Jan> interface details.

Have you tried weave?  It's part of SciPy.  From the SciPy front page:

    Weave 0.2.1 Now Available: weave is a new package that provides tools
    for including C/C++ code within Python code.  The weave.inline()
    function executes C code directly within Python, weave.blitz()
    translates Python Numeric expressions to C++ for fast execution...

More info is here:

    http://www.scipy.org/

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)



From bbollenbach at shaw.ca  Fri Apr 12 18:05:43 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 22:05:43 GMT
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org>
Message-ID: <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca>

In article <mailman.1018626392.16328.python-list at python.org>, Graham Ashton 
wrote:
> On Fri, 2002-04-12 at 13:28, Steve Holden wrote:
>> "Brad Bollenbach" <bbollenbach at shaw.ca> wrote ...
>> >
>> > This doesn't actually make sense. To call software "Open Source" is to
>> > acknowledge it as being distributed under a license that is defined as
>> > compatible with what Richard Stallman calls "Free Software". 
> 
> This isn't really the right place for this discussion, but the above is
> incorrect. Stallman didn't coin the term "Open Source", the OSI
> (www.opensource.org) did. IIRC, Stallman doesn't agree with the term
> Open Source, as it can dilute the "free as in speech" aspect of GPL
> compatible licenses, which the FSF stands for. It's all explained
> here....

This is a semantic debate. I know that Stallman chose the term "Free
Software" because he felt it was the best (and most concise) way to
express what he was thinking of.

As it turns out, *lots* of hackers talk about "Open Source" when they
often mean exactly the same thing as what Stallman calls "Free
Software". From the context of this discussion, I gathered the same
interchange is going on here.

>> > So you don't have the "choice" of licensing Open Source software. It got 
>> > to be called "Open Source" *because* of the license you've already
>> > chosen for it.
> 
> So you're saying that you can't license *your own* software to different
> people under different agreements, if you so choose?

Hmmm? Who said that?

You misread, or misunderstood (or both) what I said. It's not "optional"
to license Open Source/Free Software. It got to be *called* Open
Source/Free _because_ of the license you chose for it.

If you have a piece of software with no license whatsoever attached to
it, it's not proprietary, open source/free, shareware, "freeware", or
whatever. It's nothing more than a piece of software with your copyright
on it. You've not made any allowances, nor any limitations on the usage
of it. (However, the implicit copyright you get from creating an
original work does put some limitations on what people can legally do
without your permission).

If you attach the GPL, it's now "Free Software".

If you attach the MS-Style EULA, it's now "Proprietary Software".

If you attached an MS-Style license, that permits some kind of
evaluation period at no charge, it's now "shareware".

And so on.

In other words to think: 

"Hmmmm...should I license my Free Software (or Open Source, if that's 
the terminology you're using to refer to what Stallman calls "Free 
Software")?"

is already redundant because you had to have already attached a Free
Software compatible license to *call it free software* to begin with.


--
Brad


From vze3myxh at verizon.net  Sun Apr 21 08:29:59 2002
From: vze3myxh at verizon.net (d506)
Date: Sun, 21 Apr 2002 12:29:59 GMT
Subject: more oracle problems!
References: <e2e9e807.0204201959.7232ae91@posting.google.com>
Message-ID: <3CC2B081.20900@verizon.net>

Hugh wrote:
>     db = dco2.connect(u, p, dbname)
> dco2.DatabaseError: (12154, 'Error while trying to retrieve text for
> error ORA-1 2154')

This might happen if your ORACLE_HOME is not set properly, or if your 
oracle client software is not installed fully. In your python script, 
try to print the environment variable ORACLE_HOME and see if it has a value.

-vsb







From jrup at mail.com  Sun Apr 14 13:32:44 2002
From: jrup at mail.com (jrup)
Date: Sun, 14 Apr 2002 17:32:44 GMT
Subject: Specifying a time from previous day
Message-ID: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net>

Hi,

This is my first post to this group and I am new to python. I have some
experience with perl and vbscript for sysadmin-type scripting, but am not a
programmer.

I need to search some log files, but only if they were created after 4:00pm
the previous day.

I've looked at the python time module and its documentation, and have
skimmed mxDateTime documentation. I'm using activestate python 2.1.1 and
would like to stay with it's native library since it's the one sanctioned by
our senior nt person (also not sure how it would interact with the
activestate install, or I'd try it).

Can someone please point me in the right direction? Is this doable using the
time module? I haven't had much luck figuring it out, yet.

Thanks very much,
Jim R




From just at xs4all.nl  Mon Apr 15 11:41:52 2002
From: just at xs4all.nl (Just van Rossum)
Date: Mon, 15 Apr 2002 17:41:52 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com>
Message-ID: <just-59A848.17415215042002@news1.xs4all.nl>

In article <3cbaefb5 at news.mhogaming.com>,
 "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:

> When printing hexadecimal numbers, the bit pattern is what is needed-- not the
> sign and magnitude.

Here's a snippet from the CVS log of intobject.c:
----------------------------
revision 2.28
date: 1997/01/12 19:48:03;  author: guido;  state: Exp;  lines: +2 -7
Changed hex() and oct() again, to never emit a '-' sign.
----------------------------

I vaguely remember that hex(-1) indeed once printed -0x1, and that this 
was later fixed. Ah, the diff from 2.27 to 2.28 shows that this is 
indeed the case:


Index: intobject.c
===================================================================
RCS file: /cvsroot/python/python/dist/src/Objects/intobject.c,v
retrieving revision 2.27
retrieving revision 2.28
diff -c -r2.27 -r2.28
*** intobject.c 10 Jan 1997 17:39:30 -0000      2.27
--- intobject.c 12 Jan 1997 19:48:03 -0000      2.28
***************
*** 725,734 ****
        long x = v -> ob_ival;
        if (x == 0)
                strcpy(buf, "0");
-       else if (-x < 0)
-               sprintf(buf, "0%lo", x);
        else
!               sprintf(buf, "-0%lo", -x);
        return newstringobject(buf);
  }
  
--- 725,732 ----
        long x = v -> ob_ival;
        if (x == 0)
                strcpy(buf, "0");
        else
!               sprintf(buf, "0%lo", x);
        return newstringobject(buf);
  }
  
***************
*** 738,747 ****
  {
        char buf[20];
        long x = v -> ob_ival;
!       if (-x <= 0)
!               sprintf(buf, "0x%lx", x);
!       else
!               sprintf(buf, "-0x%lx", -x);
        return newstringobject(buf);
  }
  
--- 736,742 ----
  {
        char buf[20];
        long x = v -> ob_ival;
!       sprintf(buf, "0x%lx", x);
        return newstringobject(buf);
  }
  



Just


From marklists at mceahern.com  Fri Apr 12 09:32:01 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 08:32:01 -0500
Subject: Beginner's performance problem
In-Reply-To: <a95d1v$avb$1@panix1.panix.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCELHKKAA.marklists@mceahern.com>

[Aahz]
> I'm familiar with TVP, but what's TVM?

texturized vegetable manicotti, of course.  ;-)

// m




From cbrown at metservice.com  Thu Apr 25 19:09:22 2002
From: cbrown at metservice.com (Colin Brown)
Date: Fri, 26 Apr 2002 11:09:22 +1200
Subject: Why you sometimes get Windows Permission Denied errors
Message-ID: <3cc88d2f$1@news.nz.asiaonline.net>

Hi

I have been struggling to solve why an occasional "Permission Denied" error
popped up from the following code fragment in one of my Win2K program
threads:

...
input_file = preprocess(raw_file)
os.system('third_party input_file output_file > error_file')
if os.path.exists(saved_file):
    os.remove(saved_file)
os.rename(input_file,saved_file)
...

The error occurs on the os.rename(). The third_party executable was closing
input_file before terminating and anyway the subshell process has finished
before the os.rename is called! Most baffling.

With the aid of handle.exe from www.sysinternals.com I finally resolved the
problem. To see the problem at first hand try the following script:

import time, thread, os

def rm(file):
    print 'delete x.x'
    os.remove(file)

def wait():
    if os.name == 'nt':
        os.system('pause')
    elif os.name == 'posix':
        os.system('sleep 3')
    print 'end wait'

print 'create x.x'
f = open('x.x','w')
thread.start_new_thread(wait,())
time.sleep(1)
print '\nclose x.x'
f.close()
rm('x.x')

Although this works fine on Linux, I get a Permission Denied error on
Windows. I surmise that an os.system call uses a C fork to create the
subshell  - an exact duplicate of the current process environment including
OPEN FILE HANDLES. Because the os.system call has not returned before the
os.remove is called a "Permission Denied" error occurs. Quite simple really.

Now going back to my original problem, I had other threads doing os.system
calls. When one of these calls happens during the preprocess file write of
my above thread AND takes longer to complete than the os.system call in the
above thread then the file will still be open and the os.rename will return
the Permission Denied error.

What can be done to prevent the problem? I am open to suggestions here. My
solution was to use a threading.Semaphore() lock around all my file
operations to ensure they could only happen sequentially and also around my
os.system calls as below:

mainline>
lock=threading.Semaphore()

threads>
lock.acquire()
try:
    file_operation()
finally:
    lock.release()

subprocess_call>

def child(cmd):
    os.system(cmd)

...
lock.acquire()
try:
    thread.start_new_thread(child,(cmd,))
    time.sleep(1)
finally:
    lock.release()
...

This is definitely not elegant. If you know that your subprocess call will
always finish quickly then maybe you could just trap the error, delay and
retry until the operation suceeds.

I hope this has been helpful.

Colin Brown
PyNZ





From jb at cascade-sys.com  Mon Apr 29 21:00:55 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 18:00:55 -0700
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net> <mailman.1020118433.15459.python-list@python.org> <Xns91FFB1BDEE1BCRASXnewsDFE1@130.133.1.4>
Message-ID: <3CCDECC7.A55D77AF@cascade-sys.com>

Philip Swartzleonard wrote:

> James J. Besemer || Mon 29 Apr 2002 03:11:25p:
>
> But if you just want to increment something, use 'foo+=1'.

Good point.  It may fill the immediate need.

> Exactly the
> same, except one more character and only works on a line of it's own

Well, it's not at all exactly the same.

It's more akin to Perl's "+=" operator, but even then still quite a bit
different, being a statement instead of an operator.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From starx at pacbell.net  Mon Apr 29 20:27:17 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 30 Apr 2002 00:27:17 GMT
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net> <mailman.1020118433.15459.python-list@python.org>
Message-ID: <Xns91FFB1BDEE1BCRASXnewsDFE1@130.133.1.4>

James J. Besemer || Mon 29 Apr 2002 03:11:25p:

> 
> Vincent Foley wrote:
> 
>> What's the Python equivalient of Perl's "a"++; ?
> 
> You can't get there from here.
> 
> I.e. there is no practical equivalant.
> 
> Including it would allow deadly side-effects within an expression and
> thus would percipitate the end of the world, or at least mankind.
> 
> Regards
> 
> --jb

But if you just want to increment something, use 'foo+=1'. Exactly the 
same, except one more character and only works on a line of it's own 
(avoiding above said doom... man pointer-walking funcitons in C are nuts 
with that... =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From phr-n2002a at nightsong.com  Sun Apr 28 20:41:11 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 28 Apr 2002 17:41:11 -0700
Subject: Is there a text file somewhere i could download that contains all the words in the english language that i could read into python to create a spell checker?
References: <aai16e$m3$1@newsg1.svr.pol.co.uk>
Message-ID: <7x662bnxp4.fsf@ruckus.brouhaha.com>

"G. Willoughby" <never at mind.info> writes:
> Is there a text file somewhere i could download that contains all the words
> in the english language that i could read into python to create a spell
> checker?

There's a bunch of these linked from <http://wordlist.sourceforge.net>.


From brueckd at tbye.com  Wed Apr  3 12:50:54 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 3 Apr 2002 09:50:54 -0800 (PST)
Subject: Java to Python translation
In-Reply-To: <Pine.LNX.4.33.0204031154280.28311-100000@bagend>
Message-ID: <Pine.LNX.4.33.0204030944450.5663-100000@ns0.tbye.com>

On Wed, 3 Apr 2002, Scherer, Bill wrote:

> On Wed, 3 Apr 2002 brueckd at tbye.com wrote:
>
>
> > (setting up Oracle on
> > Linux so I can use the Python DB API is a mess).
>
> How so? We've been using Python with DCOracle on Linux boxes to
> talk to Oracle on Sun and Linux boxes for years now, and I've
> got no complaints...

Reread my post: setting up Oracle is the problem, not DCOracle, not
Python. The nice thing about the pure Java Oracle JDBC driver is that it
requires no Oracle installation/setup/configuration whatsoever outside of
your program - I can drop our app onto any Windows or Linux machine and
run it right out of the box; there's no installing Oracle client libs, no
setting up data sources, etc, etc.

-Dave





From alwyn at alwyn.demon.co.uk  Mon Apr 15 09:41:53 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Mon, 15 Apr 2002 13:41:53 GMT
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu>
Message-ID: <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk>

In <eppstein-A27785.23594014042002 at news.service.uci.edu> David Eppstein
<eppstein at ics.uci.edu> wrote:
>In article <3CBA5EDB.C60BF821 at cosc.canterbury.ac.nz>,
> Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
>
>> MacPython has a facility for creating an "applet" from
>> a Python script. This is a small Macintosh application
>> that can be launched from the Finder. You can drop
>> files on the applet, and their pathnames will appear in
>> sys.argv.
>
>Which is great for your own use -- it's what I do with Python programs 
>that I'm going to run with much frequency.  But it's not so good for 
>distributing programs to others, unless you feel comfortable telling 
>them to install the appropriate version of Python before using your 
>applet.
>
>MacPython does also have a facility for creating a standalone 
>application, but these are much bigger (several megabytes?) and I 
>haven't been able to get this to work.

But if the target platform is Mac OS X, you can run Python scripts from
the command line, as you would on any other Unix-type system. 


Alwyn


From jochen at unc.edu  Sat Apr  6 17:22:42 2002
From: jochen at unc.edu (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 06 Apr 2002 17:22:42 -0500
Subject: 2.2.1c2 test_fcntl hangs
In-Reply-To: <m3r8lsfq1x.fsf@mira.informatik.hu-berlin.de>
References: <lysn69qrp7.fsf@bock.chem.unc.edu>
	<m3r8lsfq1x.fsf@mira.informatik.hu-berlin.de>
Message-ID: <ly1ydspj0u.fsf@bock.chem.unc.edu>

On 06 Apr 2002 23:57:46 +0200 Martin v Loewis wrote:

Martin> "Jochen K?pper" <jochen at unc.edu> writes:

>> | Traceback (most recent call last):
>> |   File "Lib/test/test_fcntl.py", line 38, in ?
>> |     rv = fcntl.fcntl(f.fileno(), fcntl.F_SETLKW, lockdata)
>> | KeyboardInterrupt
>> `----
>> 
>> This is RedHat-7.0 with all updates + gcc-3.0.4. Python is

Martin> On a local file system, or through NFS? 

Through NFS, mounted with /default/ from a RedHat-7.1 machine...

Martin> Is there any chance that anybody else was holding locks to the
Martin> test file at the same time?

No, I don't think anybody else was holding a lock on that file, as
this is inside my $HOME hierarchy.

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From db3l at fitlinxx.com  Thu Apr  4 16:13:04 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 04 Apr 2002 16:13:04 -0500
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com> <3ouoaukmu32afgatil65cmcb7pvb8e5pcb@4ax.com> <a8i0pv$bu6$1@news.rchland.ibm.com> <MPG.171653199552480c989699@news.nwlink.com>
Message-ID: <ulmc3cgm7.fsf@ctwd0143.fitlinxx.com>

Jeff Shannon <jeff at ccvcorp.com> writes:

> A word of warning, though -- you may not be getting much usage of 
> your second processor.  Python itself has a Global Interpreter 
> Lock, forcing only one Python thread to be running at any given 
> time.  Secondary threads are not run on multiple processors.  You 
> may not be having this problem, since you're executing external 
> programs -- I'd imagine that those could be scheduled on the 
> second processor, but you should verify this.

Actually, threads can certainly run on any of multiple processors,
since they should be scheduled by the OS, but they can't be
interpreting Python code simultaneously.  But that doesn't stop them
from executing on different processors when they do get control of the
GIL.  Whether or not it will actually happen will depend on the
underlying OS SMP implementation, but shouldn't be limited (aside from
total CPU that can be spent interpreting code) by Python's GIL.

In this case, since external processes are being executed (it wasn't
mentioned but I assume through os.system(), os.popen() or some such
call), the GIL is going to be released before the OS-level execution
API call is made and execution by other threads on other processors
can continue.

So you'll still end up with at most one CPU's worth of time spent
interpreting Python code, but the external processes can use any
remaining CPU time.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aahz at pythoncraft.com  Sat Apr  6 17:07:08 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 17:07:08 -0500
Subject: Incomplete draft: Namespaces
Message-ID: <a8nric$sbb$1@panix1.panix.com>

I've been wrestling with the way namespaces really work, and I've been
discussing it a bit on python-dev.  I'm moving the discussion here to
get some broader viewpoints.  Here's an incomplete draft of what I've
got so far:

In Python, everything is an object: basic data types, classes, class
instances, functions, modules, and so on.  Objects are accessed through
bindings; a binding is a reference to an object, but the reference is
always implicit (unlike pointers).  Targets contain bindings.  Each time
an object gets bound to a target, the object's refcount gets increased.

Targets can be accessed through names or the __getitem__ / __setitem__
protocol.  A name is primary if it exists in execution scope.  A name is
an attribute if it is accessed via dot notation (or the __getattr__ /
__setattr__ protocol -- this gets a bit messy with new-style classes, so
we'll ignore that complication for now).  Computed bindings are accessed
through the __getitem__ / __setitem__ protocol.  There are also implicit
targets, such as the return statement.

A binding can refer to any Python object.  Multiple bindings can refer
to the same object.  Rebinding a target does not affect the original
object unless the original object's refcount goes to zero (and the
object gets automatically deleted).

A namespace is a collection of names that all exist within the same
execution scope or object.

The three execution scopes are function local (hereafter referred to as
local), module global (hereafter referred to as global), and builtin
scope.  Local scope exists any time Python's instruction pointer is
inside a function/method.  Global scope refers to the currently
executing module; explicitly accessing a function in another module
through attributes changes the current module:

    from M import f
    f()                 # f()'s global scope is the current module
    import M
    M.f()               # f()'s global scope is now the module M

Builtins are the same across all modules.  (It's possible to break this
explicitly through namespace games, but it's always true absent any
tricks.)

XXX Execution vs. object namespace

XXX Namespace lookups

Python only does indirect object lookups when retrieving an object
through a binding, not when creating bindings.  In other words

    a = 1
    def f():
        print a
    f()

will print 1, but

    def f():
        a = 1
    f()
    print a

generates a NameError.  In similar fashion

    class A:
        x = 1

    class B(A):
        pass

    B.y = 2
    print B.x

sets up a situation where B.x will look up in A's namespace to retrieve
x, but y goes into B's namespace.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From marklists at mceahern.com  Tue Apr  9 16:50:10 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 9 Apr 2002 15:50:10 -0500
Subject: binding a reference to a variable
In-Reply-To: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com>

[Andrew Koenig]
> What I would really like is to be able to define a function called,
> say, `set' such that set(x, 42) has the same effect as `x = 42'.

What about this?

  def set(name, value):
    globals()[name] = value

  set("x", 5)

  print x

The only difference is that you pass set a string representation of the
variable you want to set.  Close, but no cigar?

There's also this recipe:

  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/80373

I'm curious why you need this?

Cheers,

// mark






From sholden at holdenweb.com  Wed Apr 17 10:10:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 10:10:37 -0400
Subject: shipping a python app
References: <a9js5c$2m9$1@panix2.panix.com>
Message-ID: <upfv8.14124$hl2.9868@atlpnn01.usenetserver.com>

"Roy Smith" <roy at panix.com> wrote in message
news:a9js5c$2m9$1 at panix2.panix.com...
> I've been given a task to write a small gui-based app which needs to
> run on both NT and Unix platforms.  About all I know about NT is how
> to spell it, and I've heard it's got some cool flight simulator games.
>
[how Tcl is currently used]
>
> So, that brings us around to what, exactly, would be involved in
> shipping a stand-alone python app?  All my python development so far
> has assumed that the run-time environment already exists, or can be
> installed by the user.  That will not be the case here.  We'll need to
> assume that nothing is there and package up whatever we'll need.  I'm
> sure people have done this before.  Anybody willing to share your
> experiences?
>
> Fortunately, for this particular app, there's no intellectual property
> being protected so we don't care if the source code is visible.

Gordon McMillan's installer package is almost certainly what you want.
Gordon's web page on

    http://www.mcmillan-inc.com/install1.html

also has an intelligent discussion of the alternatives.

regards
 Steve






From phr-n2002a at nightsong.com  Wed Apr 24 09:20:34 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 06:20:34 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <slrnacdape.60f.akuchlin@ute.mems-exchange.org>
Message-ID: <7xu1q1mdwd.fsf@ruckus.brouhaha.com>

"A.M. Kuchling" <akuchlin at ute.mems-exchange.org> writes:
> > Although, HMAC in Python using string multiplication instead of loops
> > might not be so slow.  I'll see if I can code it that way.
> 
> Note that Python 2.2's standard library includes an hmac module.

Yes, but this encryption module is supposed to work in 1.5.2.

Also, the 2.2 hmac module is written in python and looks pretty slow.

I should re-read Krawczyk's HMAC paper to see what that fancy padding
is really supposed to do.  Remember that the underlying encryption
here (SHA-OFB) probably isn't the greatest in the world either.

And it actually might be faster to use two separate hash-based auth
keys than deal with the fancy HMAC padding, if a simple double hash
has some vulnerabilities.


From l0819m0v0smfm001 at sneakemail.com  Sat Apr  6 13:29:16 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Sat, 06 Apr 2002 18:29:16 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org>
Message-ID: <3CAF04F2.8020300@sneakemail.com>

Laura Creighton wrote:

> Note:  don't tell me that because python bools are just ints, it
>        doesn't matter -- people who were using bools for
>        DEMOCRAT and REPUBLICAN can now add 
>        THE_DAMN_VOTE_WASNT_PUNCHED_THROUGH - (the integer 3)
>        when to their horror they discover 2 values was not enough.
>        I need to teach students to not use bools in languages which
>        are not so forgiving.
> 


   Python's a terrible language to try to teach students how to use less 
  forgiving languages.  So what?  Should Python add type declarations 
and compile-time type checking so that it's easier to use Python to 
teach students not to do things that would get them in trouble in Java?

   Personally, I think PEP 285 does the right thing *for Python*, given 
its history, and absolutely do not care whether it might give someone 
the wrong idea about how some other hypothetical language works.  (Hey, 
at least your students will be ready for C99, which has the same notion 
of bool, so it's not a dead loss. <wink>)


   Joshua





From jdadson at ix.netcom.com  Sat Apr 27 21:49:15 2002
From: jdadson at ix.netcom.com (Jive Dadson)
Date: 28 Apr 2002 01:49:15 GMT
Subject: input/output through socket or COM port
References: <3CCB1CE9.3E08F027@ix.netcom.com> <Xns91FE207CF4D44cliechtigmxnet@62.2.16.82>
Message-ID: <3CCB557C.5BDB495E@ix.netcom.com>

Thanks a lot.  I snarfed the pserial zip file.  Do you know a similar
thing for
a TCP/IP host?

Jive


From huaiyu at gauss.almadan.ibm.com  Fri Apr 26 16:15:06 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 26 Apr 2002 20:15:06 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz>
Message-ID: <slrnacjda9.2uv.huaiyu@gauss.almadan.ibm.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
>Julian Tibble wrote:
>> 
>> erm... If there is no copyright then there is no *need* for copyleft.
>
>But if RMS would be happy with a world in which there
>was simply no copyright, he wouldn't bother with the
>GPL -- he could get the same effect just by putting
>all the stuff he writes in the public domain.

In a noncopyright world, everything and their derivatives are in public
domain, forever.  In a copyright world, public domain is only for a specific
version.

GPL essentially says: You can treat this stuff as public domain as long as
you guarantee, as far as in your capacity, that all derived versions are
also in public domain.

GPL is an attempt to simulate a noncopyright world within the legal
confinement of a copyright world.  It is a better simulation than releasing
into public domain.

>The fact that he hasn't done that means he
>*does* want *some* kind of intellectual property
>laws.

You cannot logically conclude that RMS wants copyright.  

However, your conclusion is correct by a silly accident - RMS supports
trademarks, which could be classified as *some* IP.  This just shows the
fallacy of lumping all kinds of IP together.

Huaiyu


From aleax at aleax.it  Thu Apr  4 01:28:09 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 06:28:09 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com> <3CAB0015.6563C468@python.org> <Xns91E5DC39566ADcliechtigmxnet@62.2.16.82>
Message-ID: <ZnSq8.57346$S52.1997588@news2.tin.it>

Chris Liechti wrote:
        ...
> 6 == True and 7 == True does _not_ infer 6 == 7. when using an 'if' nobody

Sure does, IF == has its fundamental mathematical properties: reflexive,
symmetric, transitive.  If it doesn't, then it does not even define an 
equivalence relationship among objects -- an absurd issue for the
concept of "equality".

Pragmatically, among other absurdities, such an == would imply that
booleans are not hashable, as otherwise x==y -> hash(x)==hash(y)
and hash(True) cannot equal both hash(6) AND hash(7).  Or else
you'd have to revolutionize the definition of hash vs eq and the whole
set of concepts underlying dictionaries.  Thinking of what this would
imply for < and > is also fun (presumably you'd have to make bools
non-comparable, like complex numbers -0- what fun!).

I think this detail of your proposal is really balmy.


Alex



From fredrik at pythonware.com  Wed Apr  3 10:08:03 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 3 Apr 2002 17:08:03 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBKEPHKIAA.marklists@mceahern.com>
Message-ID: <02a201c1db21$5c8ddfd0$0900a8c0@spiff>

mark wrote:

> Hmm, I'm not sure I follow.
> 
> This is Python without the PEP:
> 
> falseValues = [{}, [], 0, None]
> for x in falseValues:
>     if not x:
>         print "%s is 'false'." % x
> 
> This is Python with the PEP:
> 
> falseValues = [{}, [], 0, None, False]
> for x in falseValues:
>     if not x:
>         print "%s is 'false'." % x
> 
> I don't see this vast conceptual chasm between pre and post PEP that you
> appear to see.  Please help me understand.

can you explain how your second example covers
everything that's in the PEP?

can you explain why the second version is "better",
in any sense of that word?

</F>




From martin at v.loewis.de  Tue Apr 23 02:46:33 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 23 Apr 2002 08:46:33 +0200
Subject: using automake/autoconf to build extension module?
References: <mailman.1019509602.30761.python-list@python.org> <m34ri34d5w.fsf@mira.informatik.hu-berlin.de> <mailman.1019515722.23735.python-list@python.org>
Message-ID: <m3adrugbee.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

>     Martin> That does not necessary contradict each other. You can easily
>     Martin> embed Python fragments into a configure.in, to determine the
>     Martin> appropriate command line options.
> 
> I'll take your word for it, but it irks me to no end that distutils can't do
> dependencies.  Having that means I have to hack things to force files to be
> deleted I shouldn't have to.  If I can use make as the build tool I'd much
> prefer that.

If you want to use traditional make for dependencies and to invoke
commands, that's fine. All I'm saying is that you should use the same
command line options that distutils would use, instead of trusting,
say, libtool, to get it right.

Regards,
Martin


From aahz at pythoncraft.com  Sun Apr 21 10:22:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 10:22:45 -0400
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com> <a9ufpv$6hb$1@panix1.panix.com> <roy-D4669D.09550721042002@oldnews.panix.com>
Message-ID: <a9uhvl$8v7$1@panix2.panix.com>

In article <roy-D4669D.09550721042002 at oldnews.panix.com>,
Roy Smith  <roy at panix.com> wrote:
>aahz at pythoncraft.com (Aahz) wrote:
>>
>> Hmmmm....  it occurs to me that those of us teaching Python will need to
>> be *very* careful in making sure that people avoid 'in' over lists.
>
>Why?

O(N*M)

That is, if you're checking N items against a list with M items, you're
gonna have poor performance algorithmically.  Note that I said "avoid";
using 'in' against a list is sometimes the simplest and most
straightforward approach, and if it's not being done in a loop, it won't
cause too much of a performance hit.

It wasn't such a big deal earlier, because it used to be we could just
say, "Don't use 'in' with loops," but now you *do* want to use for dicts
just about everywhere.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From roy at panix.com  Thu Apr  4 08:13:26 2002
From: roy at panix.com (Roy Smith)
Date: Thu, 04 Apr 2002 08:13:26 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285 (truth pickles)
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com> <mailman.1017895453.22508.python-list@python.org>
Message-ID: <roy-857582.08132604042002@news1.panix.com>

Gustavo Niemeyer <niemeyer at conectiva.com> wrote:
> Testing if "a is True" is ok. True and False are singletons

If I pickle a bool, and then restore it, do I get the same singleton back?


From fperez528 at yahoo.com  Sat Apr 13 19:05:00 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 13 Apr 2002 17:05:00 -0600
Subject: Tools/Scripts/texi2html.py
References: <g4t3bugaplcn104b10ip18nghfer4uia89@4ax.com> <rccfbu0gbdfrq41866shbmc6jt9pj2a364@4ax.com>
Message-ID: <a9adki$s9j$1@peabody.colorado.edu>

Robert Pyron wrote:

> By the way, I would be pleased and honored if my modifications were
> to be included in the standard distribution. How do I submit them
> for consideration?
> 

Go to the python sourceforge site and submit them as a patch. It sounds like 
you did a good job, and the html help generation stuff I'm sure would be much 
appreciated by the Windows crowd. After sumbission your code will get 
reviewed, perhaps changes will be required and discussed, and if ultimately 
it really is good enough, it will go into a future release.

Cheers,

f.


From gerhard at bigfoot.de  Mon Apr 29 15:57:08 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 29 Apr 2002 21:57:08 +0200
Subject: python2.2
In-Reply-To: <A7hz8.22$Uw4.4633@newsread1.prod.itd.earthlink.net>
References: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net> <slrnacr3io.nil.gerhard@lilith.my-fqdn.de> <A7hz8.22$Uw4.4633@newsread1.prod.itd.earthlink.net>
Message-ID: <20020429195708.GA25045@lilith.my-fqdn.de>

* Billy Ng <kwokng at earthlink.net> [2002-04-29 19:26 +0000]:
> This is my source.list, would you please tell me how to change this file to
> allow me to install the testing packages?
> 
> deb ftp://ftp.debian.org/debian/ stable main
> deb-src ftp://ftp.debian.org/debian/ stable main
> deb http://non-us.debian.org/debian-non-US stable/non-US main
> deb-src http://non-us.debian.org/debian-non-US stable/non-US main
> 
> deb ftp://ftp.debian.org/debian/ stable main
> deb-src ftp://ftp.debian.org/debian/ stable main
> deb http://non-us.debian.org/debian-non-US stable/non-US main
> deb-src http://non-us.debian.org/debian-non-US stable/non-US main

I'm not sure installing testing packages on stable will work at all, but
you could try adding this to your sources.list (you'll probably have
mirrors that work better for you):


deb ftp://ftp.uni-erlangen.de/pub/Linux/debian/ testing main non-free contrib
deb-src ftp://ftp.uni-erlangen.de/pub/Linux/debian/ testing main non-free contrib
deb http://non-us.debian.org/debian-non-US testing/non-US main contrib non-free
deb-src http://non-us.debian.org/debian-non-US testing/non-US main contrib non-free

Then, you should also modifiy your /etc/apt/preferences to:

Package: *
Pin: release a=stable
Pin-Priority: 900

Package: *
Pin: release a=testing
Pin-Priority: -10


To install a package with its dependencies from the testing
distribution, you can then use:

apt-get -t testing install python2.2

I don't really understand all about the "pinning", etc. I've my mixed
testing/sid setup for apt from a HOWTO somewhere. But unfortunately, I'm
not able to find it at the moment :-(

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From mwh at python.net  Fri Apr 12 11:54:28 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 15:54:28 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk>

Ralf Juengling <juenglin at informatik.uni-freiburg.de> writes:

> Shouldn't I find the special methods '__iadd__' et al, when
> asking for the attributes of 'int'?

ints are immutable, so they don't implement __iadd__.  lists do:

>>> [].__iadd__
<method-wrapper object at 0x8145f24>

> I couldn't find them either in ther operator-module:

Hmm, I think the issue here is that 

x += y

translates (in the presence of an __iadd__ method) roughly to 

x = x.__iadd__(x, y)

and you can't do that in the form of 

operator.blah(x,y)

.  Or it might just be oversight.

Cheers,
M.

-- 
  Remember - if all you have is an axe, every problem looks 
  like hours of fun.                                        -- Frossie
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From dsavitsk at e-coli.net  Sat Apr  6 13:32:26 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 06 Apr 2002 18:32:26 GMT
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org> <pan.2002.04.06.11.23.11.491490.3709@thomas-guettler.de>
Message-ID: <_aHr8.4202$d7.1292137@newssrv26.news.prodigy.com>

if you do use a win32 COM solution, note that you need to find a way only
open one document at a time.  If one is open and a second one is opened,
they will start to fight, that is, the original will be forgotten and Word
will do all modification to the most recently open doc.

i have written (am still working on) a spooler for word doc generation on a
server.  it will appear at ecpspooler.sourceforge.net one of these days, but
if you are interested let me know.  it is win32 only, however.

-d


"Thomas Guettler" <pan-newsreader at thomas-guettler.de> wrote in message
news:pan.2002.04.06.11.23.11.491490.3709 at thomas-guettler.de...
> On Sat, 06 Apr 2002 10:15:33 +0200, David McInnis wrote:
>
> > I need to alter the document properties of Word DOCS that our users
> > upload to our Web site.  Does anyone know how to do this?  (I am python
> > installed on a Linux box).
>
> You could do this with python and COM if you would use windows with
> MS-Office installed.
>
> If you have a Windows server around you could use some RPC method
> to do this form Linux.
>
> Wvware can change your MSWord files to HTML or an other format on unix.
> But it can't change the MSWord file as far as I know.
>
> BTW: I think it is called "I have python installed". But I am not a
> native speaker
>
>  thomas
>
> --
> Thomas Guettler <guettli at thomas-guettler.de>
> http://www.thomas-guettler.de




From srh232 at nyu.edu  Fri Apr 19 22:55:14 2002
From: srh232 at nyu.edu (Shane Hoversten)
Date: Fri, 19 Apr 2002 22:55:14 -0400
Subject: An extraordinarily stupid question
Message-ID: <3CC0D892.2060006@nyu.edu>

Hi -

	This is really dumb, but I just want to print out a character without a 
newline and without a space behind it.  You can't do:

print c,

	Because Python puts a space behind it.  I can't find anywhere in the dox 
that actually talk about "print" in any detail at all.  Surely there 
must be some way to do this?

Thanks,

Shane



From nospam at nospam.com  Sat Apr 13 20:10:14 2002
From: nospam at nospam.com (Jens Baader)
Date: Sun, 14 Apr 2002 02:10:14 +0200
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com> <3cb8c690$0$19942$9b622d9e@news.freenet.de>
Message-ID: <3cb8c8d5$0$12708$9b622d9e@news.freenet.de>

"Jens Baader" <nospam at nospam.com> schrieb im Newsbeitrag
news:3cb8c690$0$19942$9b622d9e at news.freenet.de...
>
> "Terry Reedy" <tejarex at yahoo.com> schrieb im Newsbeitrag
> news:WL2u8.26958$GS6.2775752 at bin3.nnrp.aus1.giganews.com...
>
> I see. Thank you for your answers.
>
> BTW, try this in interactive mode:
>
> "The Python interpreter is very stable"*30000
>
> At least at my system the interpreter promptly
> hangs itself. I guess I just found my first bug..
> (not bad I just read the first few pages of the
> tutorial)

It works well if you use the command
line interface so the bug seem to be related to
IDLE (I'm using Python 2.2.1 with WinXP)

HTH,
Jens






From matt at mondoinfo.com  Mon Apr 29 14:39:57 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Mon, 29 Apr 2002 18:39:57 GMT
Subject: Python sous UNIX
References: <aak1q3$fgl$1@wanadoo.fr>
Message-ID: <1sgz8.58771$vm6.9978812@ruti.visi.com>

On Mon, 29 Apr 2002 20:01:11 +0200, infocap <infocap at wanadoo.fr>
wrote:

>Bonjour ? tous,

Bonjour!

Please forgive me for replying in English. My French is sufficient for
reading but not for writing. In addition to The Vaults of Parnassus,
you may find some small and interesting Python programs at Useless
Python (it's not really useless):

http://www.lowerstandard.com/python/pythonsource.html

and at ActiveState's Python cookbook:

http://aspn.activestate.com/ASPN/Cookbook/Python

Regards,
Matt



From Chris.Barker at noaa.gov  Fri Apr 19 12:10:55 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 19 Apr 2002 09:10:55 -0700
Subject: Selecting Polygons in an image
References: <a9mn30$kbq$1@hermes.shef.ac.uk>
Message-ID: <3CC04F4D.1A0721AA@noaa.gov>

josegomez at gmx.net wrote:
> The process is fairly simple: you load your image,
> and then you select with them mouse a number of polygons of interest
> (PoI). These can then be saved on to a file as a linked list of
> coordinates. You then use these coordinates to extract information
> from each frame (they are co-registered).

>I was
> thinking about Python, as Numeric would be a terrific tool to use. I
> had a look at scipy, which uses (among others!) some WxWindows
> code to do plotting. However, I don't really know how flexible this is
> to basically treat a canvas region and select points in it and all
> that.

I don't think the SciPy stuff is nearly complete enough to do that yet,
and I don't know if it's headed that way anyway. I'd look further into that.

It would be greatly useful to know of similar efforts before I
> actually get down to do some work on thi!

I am working on a Canvas (with wxPtyhon) that sound very much like what
you need. It displays assorted objects defined in floating point
coordinates, and lets you zoomi in an out and move around the resulting
image. IO'm workong on implimenting the ability to click on objects to
interact with them, but havn't actuallu put that in yet. drop me an
email if you are interested in workig with me on it. My goal is to
create a generally useful tool, so I'd love another user.

Also, there has been recent discussion on the wxPython mailing list
about a general purpose canvas. Chekc the archives, asn get on the list
to join the discussion.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From just at xs4all.nl  Sun Apr 28 02:39:47 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sun, 28 Apr 2002 08:39:47 +0200
Subject: Update locals()
References: <mailman.1019937529.7684.python-list@python.org>
Message-ID: <just-BA1361.08394728042002@news1.xs4all.nl>

In article <mailman.1019937529.7684.python-list at python.org>,
 holger krekel <pyth at devel.trillke.net> wrote:

> how can you dynamically update the locals() namebindings?

It's not supported. It's not _supposed_ to work.

Just


From dale at riverhall.NOTHANKS.co.uk  Tue Apr 23 05:26:48 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 23 Apr 2002 10:26:48 +0100
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com> <mailman.1019532041.13113.python-list@python.org>
Message-ID: <i5aacuou9hq6rolki5fan3ceaj65loo5k7@4ax.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> wrote:

>  If you're using a password to store information, then you should
>select a symmetric key encryption algorithm. I suggest searching
>google for the Mcrypto module or here's an implementation of mine of
>blowfish in python: http://bofh.concordia.ca/blowfish.py.
>

That looks very compact. Well documented too. Nicely done.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From fperez528 at yahoo.com  Sat Apr 13 21:26:31 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 13 Apr 2002 19:26:31 -0600
Subject: Weak references and tracebacks - NEVER MIND
References: <a9albr$3ha$1@peabody.colorado.edu>
Message-ID: <a9altu$3rp$1@peabody.colorado.edu>

Fernando P?rez wrote:

> Is it ok to use weak references to hold tracebacks without generating
> circular references? From the documentation:

Never mind, you simply can't make those weak refs. Sorry (tested with other 
types. dumb me).

f.


From hancock at anansispaceworks.com  Sun Apr 21 10:48:00 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 21 Apr 2002 07:48:00 -0700
Subject: word stemming / conflation
References: <3CC2C23B.DAF9057D@anansispaceworks.com> <003d01c1e9a1$7f7512f0$02010a0a@suxlap> <004701c1e9a1$e87ff170$02010a0a@suxlap>
Message-ID: <3CC2D120.EEF3A438@anansispaceworks.com>

Andreas Jung wrote:
> http://sourceforge.net/projects/pystemmer/
> From: "Andreas Jung" <andreas at andreas-jung.com>
> To: "Terry Hancock" <hancock at anansispaceworks.com>; <python-list at python.org>
> Sent: Sunday, April 21, 2002 22:00
> Subject: Re: word stemming / conflation
> 
> > I have written  wrappers for the snowball stemmers of

Actually I just found that after I posted -- I was about
to post an "oh nevermind I found it post" :-D  Turns out
I searched sourceforge for "stemming" not "stemmer". Looks
like just what I was looking for, anyway.  Thank you for
what looks like a lot of work!

Ironically, I guess they need a better stemmer
too, huh? ;-D

Is "Snowball" any relation to "SNOBOL"?  I'm not really
familiar with either, being mostly a scientific programmer
(until recently, I guess -- I find myself doing a lot
of information storage tasks now and less number
crunching).  

Thanks!
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
------------------------------------------------------




From borcis at geneva-link.ch  Sun Apr  7 14:21:34 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Sun, 07 Apr 2002 20:21:34 +0200
Subject: [OT] Re: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org> <a8gq78$hg$0@216.39.172.122> <3CAC658E.51B3A912@python.org>
Message-ID: <3CB08E2E.3C9DFD51@geneva-link.ch>

Guido van Rossum wrote:
> 
> > What if I wanted to subtype str and have a bool signifying URL-escaped or not,
> > yet still have normally functioning __len__() etc.?
> 
> That's an absurd idea, right there with redefining + to mean *.

Well, wouldn't doing just that while replacing prices by their exponentials
allow the economy to continue without a hitch (in principle, that is) ?

The legal aspects are interesting : could laws written with money in
mind be deemed to apply to accounting done by *outwardly* multiplying
prices and change instead of adding them. Anybody knows multiplication
isn't addition...

Regards,

Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")


From mwh at python.net  Thu Apr 11 10:00:42 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 11 Apr 2002 14:00:42 GMT
Subject: Python 2.2.1: no readline support on Linux
References: <m3g022d7q0.fsf@yak.ethz.ch>
Message-ID: <lkofgqwa0z.fsf@pc150.maths.bris.ac.uk>

tinu <tinu at email.ch> writes:

> Dear Pythoneers
> 
> I just installed Python 2.2.1 on SuSE Linux 7.3, (standard
> build from source) and lost all readlines support. The same
> feature worked perfect with 2.1. Has anybody advice on this?

Do you need to install some kind of "readline-devel" package?

Cheers,
M.

-- 
  Reading Slashdot can [...] often be worse than useless, especially
  to young and budding programmers: it can give you exactly the wrong
  idea about the technical issues it raises.
 -- http://www.cs.washington.edu/homes/klee/misc/slashdot.html#reasons


From gerhard at bigfoot.de  Sun Apr 21 11:47:44 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 21 Apr 2002 15:47:44 GMT
Subject: c++ vs python!
References: <3CC2D8D7.20601@student.kun.nl>
Message-ID: <slrnac5noq.9en.gerhard@lilith.my-fqdn.de>

husam wrote in comp.lang.python:
> If we compare c++ with python, which one is faster when it comes to 2D, 
> 3D graphics and math. And which one is faster when it comes to design 
> and implementation of "heavy" applications?

Python is allows for far faster development speed, while using C++
results in faster code for the same algorithm. OTOH, using a language
like C++ slows down the development process, in some cases
considerably.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From tejarex at yahoo.com  Tue Apr  2 13:36:19 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 02 Apr 2002 18:36:19 GMT
Subject: ANN: Twisted 0.16.0
References: <mailman.1017763034.29511.python-list@python.org>
Message-ID: <DSmq8.162733$7b.16317690@bin7.nnrp.aus1.giganews.com>

"Michal Wallace" <sabren at manifestation.com> wrote in message
news:mailman.1017763034.29511.python-list at python.org...
> On Mon, 1 Apr 2002, Itamar Shtull-Trauring wrote:
>
> > For more information on how Twisted helps you run licensed,
unpirated
> > Open-Source software, visit:
> > http://twistedmatrix.com/documents/howto/register
>
> Okay, I see this was posted yesterday (April Fools day), but
> the site's still up today.
>
> Is this for real?

The Twisted Web Application Framework is very much for real.  I
suspect the license page is intended to be somewhat satiric.

Terry

> If so, why? How can piracy be a problem for an open source
> project?  How does this licensing scheme prevent it?





From hancock at anansispaceworks.com  Mon Apr 15 13:58:33 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Mon, 15 Apr 2002 10:58:33 -0700
Subject: Extension class problem
Message-ID: <3CBB14C9.A934DA18@anansispaceworks.com>

Hi,
I'm running into something a little confusing. This must
be a fairly basic question, but I haven't been able to
find an answer to it in the documentation I've got (including
the online stuff, of course).

Anyway, I'm attempting a multiple-inheritance subclass
with some Zope objects (DTMLDocument  and Image if that
matters, which I doubt -- except that both are "Extension
Classes" rather than normal Python classes):

I need to bind some extra attributes and methods into
my object, so I made a helper class to hold those:

class NaryaContrib:
	def __init__(self, a, b, c):
		self.a = a
		self.b = b
		# ... simplified for brevity of course

Then I have the classes I actually want to use:

class GalleryArt(NaryaContrib, OFS.Image.Image):
	def __init__(self, a, b, c, id, title, file):
		NaryaContrib.__init__(self, a, b, c)
		OFS.Image.Image.__init__(self, id, title, file)
		#...

Except that that doesn't work -- this causes a runtime error
reminding me that an unbound method requires the first
argument to be the class instance. But it seems to me that
I've done that (i.e. the "self" should be a valid instance
since I've inherited from these classes (?) ).  At least
this seems to work for normal Python classes, when I test
it.

A little research reveals that OFS.Image.Image is an "extension
class" by virtue of inheriting from a basic extension class
in Zope, "Persistent".  My impression is that extension classes
impose some extra limitations, and I must be running into one
of them, but I can't figure out what to do about it.  Is there
a standard solution for this?  I probably just need a little
help finding the right place to look for help.

Thanks for any help,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------




From chris.gonnerman at newcenturycomputers.net  Wed Apr 17 00:07:09 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Tue, 16 Apr 2002 23:07:09 -0500
Subject: Problem Writing Files
References: <3CBCE07C.9A0E0C4@slac.stanford.edu>
Message-ID: <002801c1e5c5$5a4b8020$0101010a@local>

----- Original Message ----- 
From: "Roger Jones" <rmj at slac.stanford.edu>


> I am using ActiveState Python 2.1 under Windows NT  and after opening a
> file with:
> 
> afile=('test.dat','w')
> afile.write('some random test text')
> afile.read()

This surely is not a screen capture from an actual session or a sample
of an actual script... the first line is obviously wrong, as it assigns
a tuple of ('test.dat', 'w') to afile rather than opening a file.  It 
would correctly be:

    afile = open('test.dat', 'w')

Next, you write some data without a line ending (did you intend that?)
and then switched to using .read() on the file handle without further 
action.  Since you just did a write, the file pointer is at EOF, and 
assuming it's OK to follow a .write() with a .read() you should get a
null (empty) string.

I don't know about NT but under many Unixoid OS's you must have an 
intervening .seek() call between the .write() and .read() (including
of course a .rewind() call as it is basically an alias to .seek()).

Further the file mode 'w' should be 'w+' if you intend to both write
and read on the same stream.

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





From mark.charsley at REMOVE_THIS.radioscape.com  Thu Apr 11 09:32:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Thu, 11 Apr 2002 14:32 +0100 (BST)
Subject: Beginner's performance problem
References: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de>
Message-ID: <memo.20020411143211.1400B@a.a>

In article <slrnab68mh.15p.gerhard at lilith.my-fqdn.de>, gerhard at bigfoot.de 
(Gerhard =?iso-8859-15?Q?H=E4ring?=) wrote:

> I haven't tried but an approach using the re module should be really
> fast, and shorter. Untested code follows:
> 
> import re
> names = ["ALGORITHM_CLASS", ...]
> 
> text = open("ThirteenThousandLines.sql").read()
> for name in names:
>     regex = re.compile(name, re.I) # case insensitive search for name
>     text = regex.sub(name, text)   # replace all occurences
> print text

Aha! I'd put off reading up on the regular expression stuff until later as 
I thought it just did searches, as opposed to also doing in-place 
replaces. Half an hour spent on the HOW-TO and I'm a lot better informed. 
Add a couple of r'\b' to the above, and I ought to be laughing. 

TVM to everyone who helped.

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From nospam-readonly at getsoft.com  Sun Apr 21 20:32:21 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 10:32:21 +1000
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net>
Message-ID: <3cc359ff$0$15478$afc38c87@news.optusnet.com.au>

"Emile van Sebille" <emile at fenx.com> wrote in message
news:DtIw8.49773$HH5.4310 at rwcrnsc51.ops.asp.att.net...
> Neville Franks
> > I enter 'import difflib' which I assume is necessary and get no
> complaints.
> > Then if I type 'ndiff' I get: NameError: name 'ndiff' is not defined.
> If I
> > type: 'ndiff file1 file2' I get: SyntaxError: invalid systax.
> >
> > I've also tried:
> > >>> s = SequenceMatcher( None, "abcd", "bcde" )
> > and I get:
> > NameError: name 'SequenceMatcher' is not defined.
>
> Those names live in the difflib namespace.  When using them like this,
> you'll need to prefix them.  Try difflib.SequenceMatcher

Hi Emile, Thanks that resolved my problem with SequenceMatcher but I still
can't get ndiff to work. If I enter: 'difflib.ndiff' I get 'AttributeError:
'difflib' module has no attribute 'ndiff''

Is there a way to bring the difflib namespace into the normal? namespace.

--
Neville Franks, Author of ED for Windows - http://www.getsoft.com




From peter at engcorp.com  Sun Apr 14 23:04:59 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 14 Apr 2002 23:04:59 -0400
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za>
Message-ID: <3CBA435B.EEA59C7C@engcorp.com>

sqrt_2 wrote:
> 
> For those of you who develop apps using wxPython there is this great
> product I know of.
> It is called Boa-Constructor and also incorporates Zope.
> Have a look at:
> http://boa-constructor.sourceforge.net

Of course, those products are well known around here.

What's less well known is the relationship between Boa-Constructor and Zope.


From mscheuner at hfx.eastlink.ca  Thu Apr  4 11:41:32 2002
From: mscheuner at hfx.eastlink.ca (Marc Scheuner)
Date: 4 Apr 2002 08:41:32 -0800
Subject: NTLM authorization - compile into EXE
References: <9b309481.0204031300.70ce3c63@posting.google.com> <MPG.171563d334a3f2cf989695@news.nwlink.com>
Message-ID: <9b309481.0204040841.61960b51@posting.google.com>

Jeff Shannon <jeff at ccvcorp.com> wrote in message 
>> But under Windows, having an extra command line window open just
for
>> this NTLM proxy thingie is a bit "annoying" 
> 
> Actually, while creating an EXE might have other advantages, it's 
> not the right way to go about solving your problem.

Well, it was hoping I would be able to put the EXE into my startup
group, and "massage" it so that it would run without showing up in the
taskbar - either by moving it to the Windows tray area, or by hiding
it alltogether.

So running it under Pythonw.exe doesn't really cut it either, I think.
I'll give it a try, though.

Still - I was hoping someone could help me convert this bugger into an
.EXE on Windows.

Thanks!
Marc


From claxtonr at beer.com  Mon Apr  8 00:37:30 2002
From: claxtonr at beer.com (hejduk)
Date: 7 Apr 2002 21:37:30 -0700
Subject: Problem with os.spawnv, system, execv etc.
Message-ID: <da8db604.0204072037.644facd6@posting.google.com>

Hi, I'm trying to make a small menu widget to go on my Linux 
Enlightenment desktop, using pyGTK. Bascally it's a window that is
populated with a few buttons that our supposed to be shortcuts to some
of my favourite apps. I've got the buttons to work except for one
problem: when I press a button to start an app the menu widget becomes
unusable until that app is shut down, at which time control is
returned to it. Needless to say this doesn't make for a very useful
menu program. I'm using callbacks to call the following function
whenever a button is pressed:

def progExec(self, file):
       import os
       os.spawnv("P_NOWAIT", file, ('' '', '' ''))

I've tried using "P_WAIT, P_DETACH," etc. I've also tried using the
os.system and os.execv, but the result is always the same. Same story
when I try starting the menu widget with Enlightenment so as to free
up the command console. Is there any way to set up the menu program to
work independent of any of the programs that it starts, ie: so I can
start several programs at once using it and perhaps add some other
features.


Also, why do the os.spawnv and os.execv functions require that extra
arguement to supply arguements to the starting program. What if I
don't want to supply any arguements to the programs I'm starting? My
example function above causes a bunch of errors to be listed when the
program starts up. Is there any way to get around this?

Thanks in advance,


From ralph at penguin.com  Wed Apr  3 07:59:20 2002
From: ralph at penguin.com (Ralph)
Date: Wed, 03 Apr 2002 12:59:20 GMT
Subject: How to wait only so long for a server to respond?
References: <nwtq8.11707$ml2.924462@newsread1.prod.itd.earthlink.net> <3CAAAB3A.8020504@volkswagen.de>
Message-ID: <I0Dq8.12892$ml2.1014180@newsread1.prod.itd.earthlink.net>

Maik Hertha wrote:

> Ralph wrote:
> 
>>Pythoners:
>>
>>I have some code that requests a response from a server on the Internet,
>>but I only want to wait so long for the server to respond.  How can I tell
>>python to wait x seconds/minutes, then move on to the next section of code
>>if the server has not responded?
>>
>>Thanks as always.
>>
>>Raplh
>>
> you should have a look on the signal module.
> example is here: http://www.python.org/doc/current/lib/node214.html
> 
> 
Many thanks


From lenssen at hitnet.rwth-aachen.de  Mon Apr 15 17:19:25 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Mon, 15 Apr 2002 23:19:25 +0200
Subject: XPath?
Message-ID: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE>

Is it possible to use XPath in Python, if so are there any small samples? (I
already saw some XML DOM and SAX mentioned in the context of Python, but
little of XPath.) I don't mean using XSLT files, but opening an XML file in
Python, and executing something like evaluate/ match/
selectNodes(xPathString), returning a set of nodes.




From idbaxter at semdesigns.com  Sat Apr 13 19:43:37 2002
From: idbaxter at semdesigns.com (Ira Baxter)
Date: 13 Apr 2002 16:43:37 -0700
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <mailman.1017869530.3706.python-list@python.org> <b5cnaucrm55rpm8kgt0nibnd4lr43neljs@4ax.com> <a8h83p$rqigt$1@ID-135695.news.dfncis.de> <lmvpauoae9pgdr359v7h1g4hmrcl9rv0g3@4ax.com>
Message-ID: <6a625eea.0204131543.714e0cb7@posting.google.com>

Courageous <jkraska at san.rr.com> wrote in message news:<lmvpauoae9pgdr359v7h1g4hmrcl9rv0g3 at 4ax.com>...
> >Did you compare it with JavaCC or SableCC ? I really would like to read your
> >comments about that. Any link to an article comparing compiler-compilers is
> >welcome as well.
> 
> IIRC, both JavaCC and SableCC are LALR(k) compiler technologies.
> ANTLR is LALL(k) with predication; likewise, it is an _integrated
> suite_ of tools, for lexing, parsing, treewalking, and AST generation.
> It also contains special facilities for translation.
> 
> As far as I can tell, there simply isn't anything out there like
> it. 
> 
> C//

You might be very interested in the DMS Software Reengineering Toolkit,
then.  It provides a UNICODE-capable lexer,
context-free parsing engine, automatic tree building,
conditional source-to-source rewrites, source-fragment tree generation,
attribute evalution, symbol-table support, and, of course,
a procedural API for walking/modifying trees if you absolutely
must.    Yes, all these pieces are nicely integrated.
See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.

However, this is a commercial tool.   For free, I'll agree,
ANTLR is hard to beat.


From python at rcn.com  Thu Apr 25 15:56:00 2002
From: python at rcn.com (Raymond Hettinger)
Date: Thu, 25 Apr 2002 15:56:00 -0400
Subject: ftplib and attributes
References: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>
Message-ID: <aa9n2h$2rh$1@bob.news.rcn.net>

"LoGan" <giannivolontosky at TOGLIyahoo.it> wrote
> hi, im new bie and im trying to use ftplib, but my problem is to get the
> attributes of the files, how can i do ?
> LoGan

>>> from ftplib import *
>>> f = FTP('ftp.myisp.com','python','mypassword')
>>> print f.dir()
drwxrwsr-x   5 python   www          8192 Mar  1 16:54 .
drwxrwsr-x   5 python   www          8192 Mar  1 16:54 ..
drwxrwsr-x   2 python   www          8192 Aug  7  2001 art
-rw-r--r--   1 python   www          2100 Jan  4  1998 atom.gif
-rw-r--r--   1 python   www         97293 Jan  4  1998 boat1.jpg
-rw-r--r--   1 python   www        113540 Jan  4  1998 boat2.jpg

Arrivederci,


Raymond Hettinger





From ashimali at btinternet.com  Mon Apr  8 05:09:45 2002
From: ashimali at btinternet.com (ashimali at btinternet.com)
Date: Mon, 8 Apr 2002 10:09:45 +0100 (BST)
Subject: newbie question about event based programming
Message-ID: <6203240.1018256985203.JavaMail.root@127.0.0.1>

Hello All, 
I'm wondering if there is a way to put an "event listener" (sorry for the Java terminology), on a directory. What I'm trying to do is have a python script be invoked on files whenever new file/s are  copied/moved to a particular directory.
Any tips gratefully recieved!
Cheers
Adam




From max at alcyone.com  Mon Apr  1 14:28:56 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 01 Apr 2002 11:28:56 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <roy-502972.07485701042002@news1.panix.com>
Message-ID: <3CA8B4F8.8F58DC0C@alcyone.com>

Roy Smith wrote:

> Does it look any stranger than "False : val1 ? val2" ?

I presume you mean False ? val1 : val2, which is better written val2
:-).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From mwh at python.net  Fri Apr  5 08:42:56 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 5 Apr 2002 13:42:56 GMT
Subject: Theoretical threshold of improving Python's performance
References: <3CADA276.D91B7544@kfunigraz.ac.at>
Message-ID: <ug02as1e2.fsf@python.net>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:

> I got a private email where I can read that Python can be never made
> as fast as Common Lisp due to Python's dynamic nature. May I ask
> without any ulterior-motive: "Has the beformentioned proposition any
> value?".

Never is a long time.  Python's dynamism would seem to make it harder
to get to CL-type speeds, but impossible?  I don't think that's a
question that can be meaningfully answered.

> I know there were discussion concerning the topic; but I would be really
> interested in, from a theoretical point of few, whether it is a *fact*
> that Python can *never* become as fast as Common Lisp* or not.

See above.  Note the woolly vagueness of my comments :)

> I think it is documented that I am not really (maybe until yet)
> striving for a faster Python (clearly, would not say no) but I am
> really interested in and a little bit clueless whether Python is
> really much more dynamic than Common Lisp and this is an upper
> threshold.

Python probably is somewhat more dynamic that Common Lisp.  For
example, in Python, dicts use __eq__ methods; in Common Lisp, hash
tables are restricted to eql, equal and equalp (I think -- a bunch of
equality predicates that you can't override on your own types, in any
case).

Also, cl:+ similarly isn't extensible (and doesn't concatenate
strings), etc.

I don't know how much practical difference this makes -- some, but not
a lot, would be my guess.  Avoiding runtime method resolution is
something CL compilers work hard at, I think.

> *[I got some complaints that I sometimes stress the 10 times Common Lisp
> performance hit; if some experienced profi programmer can tell me that
> he is using Common Lisp in combination with declaring variables on a
> daily basis I am going to quit my rants immediately]

Not sure what you mean by "profi", but when I wrote a uni assignment
in CL, I declared most of my variables (possibly too many, in fact).
It's the only way to get CMUCL's compiler to shut up when you (declare
(optimize (speed 3) (safety 0)), for one thing...

Cheers,
M.

-- 
  Not only does the English Language borrow words from other
  languages, it sometimes chases them down dark alleys, hits
  them over the head, and goes through their pockets.   -- Eddy Peters


From skip at pobox.com  Wed Apr  3 10:55:25 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 3 Apr 2002 09:55:25 -0600
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <lbh7knolreu.fsf@xihu.inria.fr>
References: <mailman.1017795845.3464.python-list@python.org>
        <r1tq8.161131$af7.78459@rwcrnsc53>
        <3CAAB238.290440C5@kfunigraz.ac.at>
        <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
        <3CAAEF17.32333986@kfunigraz.ac.at>
        <lbh8z84nake.fsf@xihu.inria.fr>
        <15531.7864.698585.485436@12-248-41-177.client.attbi.com>
        <lbh7knolreu.fsf@xihu.inria.fr>
Message-ID: <15531.9709.924953.953487@12-248-41-177.client.attbi.com>

    >> You're thinking of weave.inline().  I was talking about
    >> weave.blitz().  Reread the last statement in the text I quoted, then
    >> take a look here:
    >> 
    >> http://www.scipy.org/site_content/weave/tutorial.html#Blitz

    Jan> OK, I agree this is a nice piece of work. However, this is still
    Jan> limited to relatively simple, one line expressions. No way of
    Jan> adding more complex calculations (with ifs, loops,...).

Not to belabor the point, but you can simply use Python's control
structures:

    if today == "Monday":
        expr = "a[1:-1,1:-1] =  (b[1:-1,1:-1] + b[2:,1:-1] + b[:-2,1:-1]" \
                              "+ b[1:-1,2:] + b[1:-1,:-2]) / 5."
        weave.blitz(expr)

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From stefan.heimann at web.de  Wed Apr 10 10:07:38 2002
From: stefan.heimann at web.de (Stefan Heimann)
Date: 10 Apr 2002 14:07:38 GMT
Subject: Create Excel Files without COM
References: <3CB44516.60005@thomas-guettler.de>
Message-ID: <slrnab8hp2.3r2.stefan.heimann@kunz.ratzer>

Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::

> Hi!
> 
> Is there something like WriteExcel from perl in python?
> I need to generate an MS-Excel file. I don't want to use
> COM although it will run on windows.
> 
> The data is simple. Just text in columns and rows.

What about CSV? 

Bye
  Stefan


From emile at fenx.com  Tue Apr  2 20:37:27 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 03 Apr 2002 01:37:27 GMT
Subject: A faster Python?
References: <mailman.1017795845.3464.python-list@python.org>
Message-ID: <r1tq8.161131$af7.78459@rwcrnsc53>

Nick Arnett
> To be specific, I do a lot of time series, correlation, regression and
> clustering analysis.  Those can really bog down.  While I'm fine with
> exporting to Excel, SPSS, etc., for heavy-duty analysis,

This surprised me.  I like python with Numeric exactly because it's
faster and processes more data than Excel.  I'm now regularly
calculating and delivering pivot-style results from a million row source
with 240 columns.

--

Emile van Sebille
emile at fenx.com

---------



From syver-en+usenet at online.no  Sat Apr  6 15:23:32 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sat, 06 Apr 2002 20:23:32 GMT
Subject: New classes in 2.2
Message-ID: <un0wglgqf.fsf@online.no>

I have to admit that I am excited about the revisions to pythons
object model. I hope that the current state of affairs where there are
two kinds of classes will eventually fade away, as it seems quite ugly
at the time being (wouldn't want to be a new to Python just now). It
would be sad if Python we're to develop along the lines as C++, where
there seems to be no obvious way to do even the simplest of things (in
my view, even though I think it's a fascinating language, my main
language in fact until I discovered Python).

I've started looking more closely at the new style classes (object
derived). And it struck me as strange that in Guido's Unifying types
and classes in Python 2.2, the new super function and proper
classmethods are discussed. I soon found out that super would (in
addition to ordinary methods) only work for class data members and
static functions, and not for class methods were they in my view would
be most elegant.

I am sure someone has got a better version of super for class methods
than I can present here, but I insert it below anyway as a motivation
to improve or post a link to a better implementation. 

(Shamelessly stolen from guido's Super class in his article, and
blindly hacked till it did what I wanted it to do then and there :-):

class ClassMethodSuper(cls):
    def __init__(self, cls):
        self.__cls__ = cls
    def __getattr__(self, attr):
        starttype = self.__cls__
        mro = iter(starttype.__mro__)
        for cls in mro:
            if cls is self.__cls__:
                break
        for cls in mro:
            if attr in cls.__dict__:
                x = getattr(cls, attr)
                x = Command(x.im_func, self.__cls__)
                return x
        raise AttributeError, attr


class Command(object):
    def __init__(self, func, *args, **kwargs):
        self.func = func
        self.args = args
        self.kw = kwargs

    def __call__(self, *args, **kwargs):
        args = self.args + args
        kwargs.update(self.kw)
        self.func(*args, **kwargs)


-- 

Vennlig hilsen 

Syver Enstad


From huaiyu at gauss.almadan.ibm.com  Tue Apr  9 13:22:41 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 9 Apr 2002 17:22:41 +0000 (UTC)
Subject: Function question
References: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu>
Message-ID: <slrnab68r0.dvl.huaiyu@gauss.almadan.ibm.com>

On Sun, 7 Apr 2002 23:55:47 -0500, John <nobody at nobody.com> wrote:
>
>Hi I have a question:
>
>Let's say there is a function with two arguments.
>def add(a,b): return a+b
>
>What I want to do is:
>I want to get 'a' dynamically..
>But once a value is decided, I want a separate function (say, new_add) that
>does
>new_add(b) (equals to) add(a,b)
>
>That is, I want the function definition to be dynamic.
>
>Would that be possible? Any comment would be greatly appreciated.

The easiest way With Python 2.2:

>>> def f1(a):
...    def f2(b):
...       return a+b
...    return f2
... 
>>> f = f1(3)
>>> f(3)
6
>>> f(2)
5
>>> f(-1)
2

With Python 2.1 put this at the beginning of module:

from __future__ import nested_scopes

With earlier Python, use default arguments, like def f2(b, a=a), but this
has risks.

Or, if there are more than a few arguments, you might want to use a callable
object (an instance of a class with a __call__ method).

Huaiyu


From mertz at gnosis.cx  Wed Apr  3 22:55:48 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Wed, 03 Apr 2002 22:55:48 -0500
Subject: Static Variables (fwd)
Message-ID: <E78q8kKkXYcD092yn@gnosis.cx>                

"Jim Jacobs" <jim.jacobs at jacobshome.com> wrote previously:
|Is it possible to simulate "C" style static variables in functions?
|In methods?

>From the other responses, it seems like a lot of users don't know the
"Python static trick."  I use this moderately often... it is a slightly
strange spelling, but not hard once you learn it:

    >>> def counter(msg, cnt=[0]):
    ...   print cnt[0], msg
    ...   cnt[0] += 1
    ...
    >>> counter('first call')
    0 first call
    >>> counter('second call')
    1 second call
    >>> counter('third call')
    2 third call
    >>> counter('counter override', [37])
    37 counter override

Anything passed with a mutable type default argument gets defined along
with the function itself.  The value doesn't get redefined on calls
unless the caller explicitly forces it.

We-don't-need-no-stinkin-closures-ly yours, David...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From eric.brunel at pragmadev.com  Mon Apr 15 13:26:42 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 15 Apr 2002 17:26:42 +0000
Subject: Newbie question: strange Re behaviour
References: <mailman.1018883258.24889.python-list@python.org>
Message-ID: <a9er6q$2gdh$1@norfair.nerim.net>

Patrick.Bussi at space.alcatel.fr wrote:
> 
> Could someone help me understand what's wrong with this very simple regex,
> whose initial purpose was to extract the value of an HTML field. For
> demonstration, I have oversimplified the code below:
> 
> -----start-----
> [pat]$ python
> Python 2.0 (#1, Apr 11 2001, 19:18:08)
> [GCC 2.96 20000731 (Linux-Mandrake 8.0 2.96-0.48mdk)] on linux-i386
> Type "copyright", "credits" or "license" for more information.
>>>>
>>>> import re
>>>> s='[^value=].*[^>]'

That means "any character not in the set ('v', 'a', 'l', 'u', 'e', '='), 
followed by any string, followed by a character that is not a '>'". I guess 
that's not what you want... but it explains why it ignores the leading 'a' 
(which *is* in the set ('v', 'a', 'l', 'u', 'e', '=')).

Try:
s = "value=(.*)>"
and see what group(1) returns after compiling the re...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From allexpro at myrealbox.com  Thu Apr 11 18:13:48 2002
From: allexpro at myrealbox.com (allexpro)
Date: Thu, 11 Apr 2002 18:13:48 -0400
Subject: sockets
Message-ID: <uVnt8.8550$jD4.264772@weber.videotron.net>

 I'm trying to make a python MUD for the moment, but i don't know what to
use
for sockets. i've heard alot of things about socket, serverSoket, medusa,
simpleHTTPServer,
asyncore, etc. but i just don't know which to use. Could anyone help me
please?
(I'm using python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on
win32)




From lac at strakt.com  Sat Apr  6 14:13:38 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 21:13:38 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from Guido van Rossum <guido@python.org> 
   of "Sat, 06 Apr 2002 13:54:34 CDT." <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> 
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com>  <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> 
Message-ID: <200204061913.g36JDcdj017764@ratthing-b246.strakt.com>

Thank you very much for the reply.
I just want to correct something you ascribed to me.

> I guess we'll have to agree to disagree here: I don't think that bools
> are bad, either Python's variety or any other language's kind.  I've
> seen your arguments, and they just don't convince me.  So be it.

I think there is a time and a place for bools --  Describing things
that have two states that can never, ever, have 3 or more states.  
These things are vanishingly close to non-existent, unless you are
a mathematician in which case you may be playing with these sorts
of things all the time.  What I am trying to do is save people
like Mark McEahern from his isValid function, which returns whether
a username is valid or not.  On the day his boss demands a third
value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
he will thank me.

That is what I mean by teaching people not to use bools -- teaching
when not to use them. This is very hard to teach.  I know it because
I have been doing it for years and years.   I even know _why_ it is
hard to teach, but that doesn't make it any easier.

Laura




From lac at strakt.com  Mon Apr  8 14:43:15 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 08 Apr 2002 20:43:15 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from Ken Seehof <kseehof@neuralintegrator.com> 
   of "Mon, 08 Apr 2002 07:34:28 PDT." <EFEFIHAINIINNELLHFAIMECBCAAA.kseehof@neuralintegrator.com> 
References: <EFEFIHAINIINNELLHFAIMECBCAAA.kseehof@neuralintegrator.com> 
Message-ID: <200204081843.g38IhFdj020386@ratthing-b246.strakt.com>

Lots of people have come to this discussion late.  So you don't
get the wrong idea ... in the style of the 'have you come for an
argument' sketch ....

This is the 'using booleans for something that is not intrinsically
a 2 state thing is bad' class.  The 'using objects is often better
than integers class' is 2 doors down, on the right hand side.  We
had a 'mangling your boolean returing functions to return ints instead'
class in the building next door, but it self-destructed last week,
so we don't offer that any more...

Laura Creighton




From clspence at one.net  Thu Apr 25 19:35:04 2002
From: clspence at one.net (Chris Spencer)
Date: 25 Apr 2002 18:35:04 -0500
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com>
Message-ID: <gb4hcu4b55gq99blg08eudh14pm8ivju0k@4ax.com>

	I want to thank everyone who responded to this.  I read the floating
point tutorial and now understand the issues.  I wasn't really looking for that
much precision anyways (2 decimal places was all I really need).  So now that I
know, I shant ask again.

BTW, this is one thing I love about the Python community.  Ask a stupid question
and you get good and helpful answers.

Chris.

On 25 Apr 2002 18:40:45 -0400, David Bolen <db3l at fitlinxx.com> wrote:

<snip>

>The results you are seeing aren't precisely "wrong" (certainly not in
>terms of binary floating point precision), it's just more detail than
>you probably expect by default.  If you thought that computers using
>binary floating point (the most typical nowadays) in lieu of BCD or
>other decimal based formats could precisely represent base 10 numbers
>the same way you might with pencil and paper, that was an erroneous,
>but forgiveable, assumption :-)

<snip>

>-- David



From loewis at informatik.hu-berlin.de  Fri Apr 19 11:29:46 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Apr 2002 17:29:46 +0200
Subject: GNU GPL & Python stand-alone packages from py2exe
References: <i0kit6nn3l4.fsf@Thales.MI.Uni-Koeln.DE>
Message-ID: <j4hem7u2ol.fsf@informatik.hu-berlin.de>

Ulrich Goertz <u at g0ertz.de> writes:

> 1) Since the Python license is "GPL compatible" this should be 
> no problem, 

Correct.

> but what does that mean in practice? Can/should I put
> the whole package under the GPL and omit the Python license?

No, you cannot put the entire package under the GPL; you need to
include the PSF license agreement and notice of copyright (see clause
2 of the PSF license).

> Or can I just put my own code under the GPL, and use the 
> Python license for the Python part, although I bundle both
> into one .exe-file and a few dll's?

Yes, that is what you gain by a "GPL-compatible" license: the GPL
would not allow to distribute your software in this way if the Python
license was not GPL compatible.

> 2) Would I have to put the source code of the Python interpreter
> on the CD, too? 

Upon reading the GPL, that indeed appears to be the case:

  If identifiable sections of that work are not derived from the
  Program, and can be reasonably considered independent and separate
  works in themselves, then this License, and its terms, do not apply
  to those sections when you distribute them as separate works.

This section would apply to Python: Python is an identifiable section
of your work that can be considered independent, so the Python
licensing terms apply when distributing Python separately. However:

  But when you distribute the same sections as part of a whole which
  is a work based on the Program, the distribution of the whole must
  be on the terms of this License, whose permissions for other
  licensees extend to the entire whole, and thus to each and every
  part regardless of who wrote it.

So the GPL applies to the work 'as a whole', and therefore, the same
source distribution clause applies.

> 3) Would it be more suitable to release the source code of my
> application under the GPL, and release the stand-alone package
> under another ("closed") license which would not request the 
> source code to be included

If you own all of the code, or have permission of all other authors to
distribute their code without providing source, you can certain
operate that way.

> As a side note, I do understand that the GPL is not as free
> as other licenses (the Python license, say), but I do not 
> want to give others the possibility to do what they want with
> my code without any restrictions. So I think the GPL is
> suitable for me. My questions are more about how to proceed
> in practice in this case.

I think 'freeness' is difficult to measure, so it is hard to say what
is 'more free' and 'less free'. I think RMS argues that the GPL is
actually 'more free', since it not only gives freedom to the author of
the software, but also freedom to the community of users. Some people
agree, and some disagree.

Regards,
Martin



From fperez528 at yahoo.com  Mon Apr  8 22:13:11 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 08 Apr 2002 20:13:11 -0600
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com> <1018304015.46066@cswreg.cos.agilent.com> <MPG.171bbba6ee73df8c98969f@news.nwlink.com> <mailman.1018310800.31349.python-list@python.org> <MPG.171bd911b760bfa19896a0@news.nwlink.com>
Message-ID: <a8tf8h$i5g$1@peabody.colorado.edu>

Jeff Shannon wrote:

> I obviously can't speak to your specific designs, not having seen
> them, but I can't imagine how globals()/locals() would be
> necessary unless you have a large amount of interdependencies in
> your code, and it is generally accepted that interdependent code
> is not a good thing.  If you can explain to me how you can
> require access to large portions of another scope's namespace
> without having large interdependencies, then perhaps I'll change
> my mind about this method.

Here's an example:

    def __call__(self,header='',local_ns=None,global_ns=None,dummy=None):
        """Activate the interactive interpreter.

        __call__(self,header='',local_ns=None,global_ns,dummy=None) -> Start
        the interpreter shell with the given local and global namespaces, and
        optionally print a header string at startup.

        The shell can be globally activated/deactivated using the
        set/get_dummy_mode methods. This allows you to turn off a shell used
        for debugging globally.

        However, *each* time you call the shell you can override the current
        state of dummy_mode with the optional keyword parameter 'dummy'. For
        example, if you set dummy mode on with IPShell.set_dummy_mode(1), you
        can still have a specific call work by making it as IPShell(dummy=0).

        The optional keyword parameter dummy controls whether the call
        actually does anything.  """


        # Allow the dummy parameter to override the global __dummy_mode
        if dummy:
            return
        else:
            if dummy != 0 and self.__dummy_mode:
                return

        # Set global subsystems (display,completions) to our values
        sys.displayhook = self.sys_displayhook_embed
        try:
            import readline
            readline.set_completer(self.IP.Completer.complete)
        except:
            pass

        # Get locals and globals from caller
        call_frame = sys._getframe().f_back

        if local_ns is None:
            local_ns = call_frame.f_locals
        if global_ns is None:
            global_ns = call_frame.f_globals

        try:
            if self.banner:
                if header:
                    self.IP.embed_mainloop('%s\n%s' % (self.banner,header),
                                           local_ns,global_ns)
                else:
                    self.IP.embed_mainloop(self.banner,local_ns,global_ns)
            else:
                self.IP.embed_mainloop(self.banner,local_ns,global_ns)
                pass
        except SystemExit:
            pass

        if self.exit_msg:
            print self.exit_msg
            
        # Restore global systems (display, completion)
        sys.displayhook = self.sys_displayhook_ori
        try:
            readline.set_completer(self.sys_ipcompleter_ori)
        except:
            pass # not embedded, or not using readline.


This is an embedded interpreter, which _needs_ to know the namespace structure 
of where it's being called, but can't expect to be passed parameters (since 
it simply needs to know _all_ of the caller's namespace). This discussion has 
made me realize that in fact this system fails to properly recongize nested 
scopes. 

So tell me how such a system could be built otherwise and I'll be happy to 
change my design.

Cheers,

f.


From synthespian at uol.com.br  Sat Apr 27 23:21:14 2002
From: synthespian at uol.com.br (synthespian)
Date: Sun, 28 Apr 2002 00:21:14 -0300
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s7el$sr7$1@slb2.atl.mindspring.net>
Message-ID: <pan.2002.04.28.00.20.56.555436.3624@uol.com.br>

On Sat, 20 Apr 2002 14:12:17 -0300, Andrew Dalke wrote:

> Siegfried Gonzi:
>>What Python --from my pragmatic point of view-- has in common with Lisp
>>is the nice feature that you can put in your return  list different
>>objects,
>>
>>e.g. [[1,2,3],["334.233,"Ein bloeder Oesterreicher",777],array]
> 
> This is also present in Perl and Tcl, and also I believe in Smalltalk,
> Mathematica, and even APL.
> 
>>And it is possible in Python (sorry I know only the idle) to build your
>>functions in the command-line (every Common Lisp environment provides
>>you this).
> 
> I don't know what it's called in Common Lisp, but in Python I think you
> are refering to the interactive mode.  The command-line is what's used
> to start Python from, say, the unix shell.
> 
>>I often have heard that CLOS (Common Lisp Object System) lets you change
>>your class objects on the fly, I do not have any experience with
> that
>>and I am not sure how this relates to Python (my Python object oriented
>>programming experience is zero).
> 
> You can modify Python classes and instances as well, though I don't know
> how similar it is to CLOS.  I do know there are differences between how
> Python does it and how Smalltalk does it.  (Eg, reload() doesn't
updatetheir
> old versions of newly redefined classes.)
> 
>                     Andrew
>                     dalke at dalkescientific.com

<sarcasm> No kidding! You can modify your classes?! Gee, I wonder if you can do *
that * in C++ </sarcasm>
But can you modify them * in runtime * ?! Without halting your
application and, say, recompiling? <smirk>
Answer: NO
Only in LISP can you do that.
That's why NASA is so fond of using LISP on their probes.
And what's this discussion Python X LISP? I thought Peter Norvig's
article already had cleared that up (that Python is a LISP from people
who never understood CONS).

Cheers

H
synthespian at uol.com.br


From mwh at python.net  Thu Apr 18 08:18:45 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 12:18:45 GMT
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019130760.16514.python-list@python.org>
Message-ID: <lkofghqjq4.fsf@pc150.maths.bris.ac.uk>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> I don't know enough about the internals to assert anything about this
> myself, so I tried a build with CC=cc and CXX=CC because we fortunately
> have the Forte6 compiler set available.
> 
> Script started on Thu Apr 18 11:53:08 2002
>         [...]
> 160 tests OK.
> 4 tests failed:
>     test_largefile test_socket test_sundry test_urllib2
> 23 tests skipped:
>     test___all__ test_al test_asynchat test_bsddb test_cd test_cl
>     test_curses test_dl test_gdbm test_gl test_imgfile
>     test_linuxaudiodev test_minidom test_nis test_ntpath test_openpty
>     test_pyexpat test_sax test_socket_ssl test_socketserver
>     test_sunaudiodev test_winreg test_winsound
> Ask someone to teach regrtest.py about which tests are
> expected to get skipped on sunos5.
> *** Error code 1
> make: Fatal error: Command failed for target `test'
> neelix hgs 13 %> exit
> neelix hgs 14 %>
> script done on Thu Apr 18 12:33:51 2002
> 
> How concerned should I be about these?

test_socket failing is bad.  The test_sundry & test_urllib2 failures
would be connected to this, I imagine.

More is hard to say (I must admit my head is starting to spin like a
top with all these reports).

Cheers,
M.

-- 
  Just point your web browser at http://www.python.org/search/ and
  look for "program", "doesn't", "work", or "my". Whenever you find
  someone else whose program didn't work, don't do what they
  did. Repeat as needed.    -- Tim Peters, on python-help, 16 Jun 1998


From phr-n2002a at nightsong.com  Thu Apr  4 09:34:33 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 06:34:33 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com> <3CAC6283.424163FB@python.org>
Message-ID: <7xu1qr4jnq.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
> > I think there will need to be a format code for truth values:
> > 
> >   ("%b" % True) => "True"
> 
> There's no need.  "%s" % True already yields 'True'.  if you want
> coercion to bool, use "%s" % bool(b).

In that case there's also no need for %d, since "%s" % 17 already
yields '17'.  If ints have a format code and bools don't, one is
left with the impression that bools aren't as respectable a type
as ints.  That doesn't seem in the spirit of the PEP.


From phr-n2002a at nightsong.com  Wed Apr 24 01:04:28 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 23 Apr 2002 22:04:28 -0700
Subject: SHA-based encryption function in Python
Message-ID: <7x3cxlk7qb.fsf@ruckus.brouhaha.com>

I've put together an encryption function written in Python using the
SHA module to provide a keystream in output feedback mode.  It's
nowhere near as good as AES, but should be a big improvement over the
rotor module.  It's at

  http://www.nightsong.com/phr/crypto/p2.py

if anyone wants to look at it and make comments.  Just say

   import p2

to load it, and then to encrypt, say

  ciphertext = p2.p2_encrypt(plaintext, passphrase)

and to decrypt, say

  plaintext = p2.p2_decrypt(ciphertext, passphrase)

If you decrypt an altered ciphertext or use an incorrect key, the
decryption function raises a CryptError exception.

There is 24 bytes of plaintext expansion and it's slow for short
strings, but the module provides an integrity check, and is reasonably
fast on long strings (by pure-Python standards).  It encrypts about
500k bytes/sec on my P3-750 on long strings, which is about 5x as fast
as RC4 coded in pure Python.  I've tested it in Python 1.5.2 and 2.2.

Note, this is NOT the AES module discussed in the "Message encryption
function" sci.crypt thread recently (that module will be written in C
and use AES).  However, following that discussion, this function makes
a nonce based on the key and plaintext as well as the current time and
an internal state based on the time of initialization and the values
of previously issued nonces.  It currently doesn't try to use any
other system-dependent entropy.  I might see if I can portably add
some to the next version.

It's possible that a nonce could get re-used if two separate Python
instances simultaneously encrypt the same plaintext with the same key.
That could happen, for example, on a cgi web server where two requests
come in very close together and each starts its own Python instance.
That could be a security failure in some applications, but hopefully
not many.

This version is intended for testing and review ONLY, and contains a
date check that stops it from running after June 1 (it tells you to
get a new version instead, in case there's security bugs in this test
version).  Please don't remove the date check.

Please let me know of any problems/errors that you notice.  If nobody
spots anything within maybe a week or two, then I'll release another
version without the date check.  Future versions are obviously NOT
guaranteed to be compatible--if this version has a design error, the
error shouldn't be propagated.

I believe this code is exportable from the US under BXA exemption TSU
and that I've properly notified BXA of its publication on the web.


From loewis at informatik.hu-berlin.de  Mon Apr  8 06:51:18 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 08 Apr 2002 12:51:18 +0200
Subject: executing an external program
References: <3CB14C40.402@beth.uniforum.org.za>
Message-ID: <j4r8lqxy3d.fsf@informatik.hu-berlin.de>

mixo <mixo at beth.uniforum.org.za> writes:

> How can I continue execution of my script, and run only one instance
> of the external program and keep track of what is doing so as to
> avoid confusion?

On Unix, you can use os.fork/os.exec to create the subprocess
asynchronously. You can then use os.wait to wait for the completion of
the child process when you want to start the next process. For that,
you have three modi operandi:
1. you invoke wait blockingly.
2. you invoke wait non-blockingly, and operate in a polling fashiong.
3. you install a SIGCLD handler, and start the next process when the
   previous one terminates.

HTH,
Martin



From dradul at yahoo.com  Fri Apr 26 20:54:00 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Fri, 26 Apr 2002 19:54:00 -0500
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com> <3CC7889D.2080002@nowhere.org> <7x8z7cvusy.fsf@ruckus.brouhaha.com> <3CC8BCD1.7090804@nowhere.org> <7xwuuvatiz.fsf@ruckus.brouhaha.com>
Message-ID: <aact9l$9odgq$3@ID-99513.news.dfncis.de>

"Paul Rubin" <phr-n2002a at nightsong.com> escribi? en el mensaje
news:7xwuuvatiz.fsf at ruckus.brouhaha.com...

> However, I'm planning to write an AES module in C that would be less
> kludgy (plus a lot faster) than this SHA thing.  So the fancy module
> should be able to use AES in most situations.
>

See the project at http://pyaes.sourceforge.net/[1] for a working
implementation of AES and SHA as C extensions that are unfortunately
abandoned. Don't need to reinvent the wheel, make it better.

[1] I may be wrong about the project name, but the code is there
somewhere.



From Chris.Barker at noaa.gov  Tue Apr  2 12:01:33 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 09:01:33 -0800
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com> <slrnaaftai.8mf.gerhard@lilith.my-fqdn.de> <w20q8.1572$0T6.38985@e3500-atl1.usenetserver.com> <3CA8EB72.F6E7973C@noaa.gov> <Xns91E3B2015AB65RASXnewsDFE1@130.133.1.4>
Message-ID: <3CA9E3D4.EA29A7DC@noaa.gov>

Philip Swartzleonard wrote:
> Chris Barker || Mon 01 Apr 2002 03:21:29p:
> > Robert Oschler wrote:
> >> Also, will it coexists with Konqueror (which shells Netscape I think)
> >> and Netscape?
> >
> > NO, it doesn't. Konqueror is it's own beast, written from scratch, I
> > think. It'a a pretty nice web brower, I use it most of the time.
> >
> > You can have any number oif browsers on your Linux box, this isn't
> > Monopoly style software...
> 
> It's a little confusing to answer that way. Yes they do coexist, as in you
> can install both of them at the same time and nothing bad will happen. No,
> they don't ... communicate? Neither does anything special because the
> other is there, or uses it in any way? (I think that's what you meant by
> your first no).

You're right, that was confusing.

No, Konqueror does not shell Netscape.

>Neither does anything special because the
> other is there, or uses it in any way? 

Mostly true, although Konqueror can access netscape bookmarks, and they
can both use some of the same browser plug ins.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From guido at python.org  Wed Apr  3 17:32:52 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 17:32:52 -0500
Subject: ACCEPTED: PEP 285
Message-ID: <200204032232.g33MWqJ28933@odiug.zope.com>

Here's one final posting of PEP 285.  I've read (or at least skimmed)
all feedback that was on our news server by 9:30am EST on April 2nd,
plus anything mailed directly to me or to python-dev until now.

There was a lot of negative feedback, but there was also a lot of
support for the PEP.  The feedback made it easy to decide on most of
the smaller issues in the review section of the PEP as previously
posted: even reviewers who were against the PEP mostly supported my
preferences for the various details.

Despite the negative feedback, I've decided to accept the PEP.  The
most important perceived problem is that newbies tend to write

    if x == True: ...

where they should write

    if x: ...

I believe this problem should be solved by education (the Zen master
hits the student on the head with a stick, and the student gets
enlightenment) rather than by holding back what I believe will be a
useful feature.  You can expect the bool type in Python 2.3; it will
be in CVS tonight.

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

PEP: 285
Title: Adding a bool type
Version: $Revision: 1.20 $
Last-Modified: $Date: 2002/04/03 22:11:05 $
Author: guido at python.org (Guido van Rossum)
Status: Accepted
Type: Standards Track
Created: 8-Mar-2002
Python-Version: 2.3
Post-History: 8-Mar-2002, 30-Mar-2002, 3-Apr-2002


Abstract

    This PEP proposes the introduction of a new built-in type, bool,
    with two constants, False and True.  The bool type would be a
    straightforward subtype (in C) of the int type, and the values
    False and True would behave like 0 and 1 in most respects (for
    example, False==0 and True==1 would be true) except repr() and
    str().  All built-in operations that conceptually return a Boolean
    result will be changed to return False or True instead of 0 or 1;
    for example, comparisons, the "not" operator, and predicates like
    isinstance().


Review

    I've collected enough feedback to last me a lifetime, so I declare
    the review period officially OVER.  I had Chinese food today; my
    fortune cookie said "Strong and bitter words indicate a weak
    cause."  It reminded me of some of the posts against this
    PEP... :-)

    Anyway, here are my BDFL pronouncements.  (Executive summary: I'm
    not changing a thing; all variants are rejected.)

    1) Should this PEP be accepted?

    => Yes.

       There have been many arguments against the PEP.  Many of them
       were based on misunderstandings.  I've tried to clarify some of
       the most common misunderstandings below in the main text of the
       PEP.  The only issue that weighs at all for me is the tendency
       of newbies to write "if x == True" where "if x" would suffice.
       More about that below too.  I think this is not a sufficient
       reason to reject the PEP.

    2) Should str(True) return "True" or "1"?  "1" might reduce
       backwards compatibility problems, but looks strange.
       (repr(True) would always return "True".)

    => "True".

       Almost all reviewers agree with this.

    3) Should the constants be called 'True' and 'False' (similar to
       None) or 'true' and 'false' (as in C++, Java and C99)?

    => True and False.

       Most reviewers agree that consistency within Python is more
       important than consistency with other languages.

    4) Should we strive to eliminate non-Boolean operations on bools
       in the future, through suitable warnings, so that for example
       True+1 would eventually (in Python 3000) be illegal?

    => No.

       There's a small but vocal minority that would prefer to see
       "textbook" bools that don't support arithmetic operations at
       all, but most reviewers agree with me that bools should always
       allow arithmetic operations.

    5) Should operator.truth(x) return an int or a bool?

    => bool.

       Tim Peters believes it should return an int, but almost all
       other reviewers agree that it should return a bool.  My
       rationale: operator.truth() exists to force a Boolean context
       on its argument (it calls the C API PyObject_IsTrue()).
       Whether the outcome is reported as int or bool is secondary; if
       bool exists there's no reason not to use it.  (Under the PEP,
       operator.truth() now becomes an alias for bool(); that's fine.)

    6) Should bool inherit from int?

    => Yes.

       In an ideal world, bool might be better implemented as a
       separate integer type that knows how to perform mixed-mode
       arithmetic.  However, inheriting bool from int eases the
       implementation enormously (in part since all C code that calls
       PyInt_Check() will continue to work -- this returns true for
       subclasses of int).  Also, I believe this is right in terms of
       substitutability: code that requires an int can be fed a bool
       and it will behave the same as 0 or 1.  Code that requires a
       bool may not work when it is given an int; for example, 3 & 4
       is 0, but both 3 and 4 are true when considered as truth
       values.

    7) Should the name 'bool' be changed?

    => No.

       Some reviewers have argued for boolean instead of bool, because
       this would be easier to understand (novices may have heard of
       Boolean algebra but may not make the connection with bool) or
       because they hate abbreviations.  My take: Python uses
       abbreviations judiciously (like 'def', 'int', 'dict') and I
       don't think these are a burden to understanding.  To a newbie,
       it doesn't matter whether it's called a waffle or a bool; it's
       a new word, and they learn quickly what it means.

       One reviewer has argued to make the name 'truth'.  I find this
       an unattractive name, and would actually prefer to reserve this
       term (in documentation) for the more abstract concept of truth
       values that already exists in Python.  For example: "when a
       container is interpreted as a truth value, an empty container
       is considered false and a non-empty one is considered true."

    8) Should we strive to require that Boolean operations (like "if",
       "and", "not") have a bool as an argument in the future, so that
       for example "if []:" would become illegal and would have to be
       writen as "if bool([]):" ???

    => No!!!

       Some people believe that this is how a language with a textbook
       Boolean type should behave.  Because it was brought up, others
       have worried that I might agree with this position.  Let me
       make my position on this quite clear.  This is not part of the
       PEP's motivation and I don't intend to make this change.  (See
       also the section "Clarification" below.)


Rationale

    Most languages eventually grow a Boolean type; even C99 (the new
    and improved C standard, not yet widely adopted) has one.

    Many programmers apparently feel the need for a Boolean type; most
    Python documentation contains a bit of an apology for the absence
    of a Boolean type.  I've seen lots of modules that defined
    constants "False=0" and "True=1" (or similar) at the top and used
    those.  The problem with this is that everybody does it
    differently.  For example, should you use "FALSE", "false",
    "False", "F" or even "f"?  And should false be the value zero or
    None, or perhaps a truth value of a different type that will print
    as "true" or "false"?  Adding a standard bool type to the language
    resolves those issues.

    Some external libraries (like databases and RPC packages) need to
    be able to distinguish between Boolean and integral values, and
    while it's usually possible to craft a solution, it would be
    easier if the language offered a standard Boolean type.  This also
    applies to Jython: some Java classes have separately overloaded
    methods or constructors for int and boolean arguments.  The bool
    type can be used to select the boolean variant.  (The same is
    apparently the case for some COM interfaces.)

    The standard bool type can also serve as a way to force a value to
    be interpreted as a Boolean, which can be used to normalize
    Boolean values.  When a Boolean value needs to be normalized to
    one of two values, bool(x) is much clearer than "not not x" and
    much more concise than

        if x:
            return 1
        else:
            return 0

    Here are some arguments derived from teaching Python.  When
    showing people comparison operators etc. in the interactive shell,
    I think this is a bit ugly:

        >>> a = 13
        >>> b = 12
        >>> a > b
        1
        >>>

    If this was:

        >>> a > b
        True
        >>>

    it would require a millisecond less thinking each time a 0 or 1
    was printed.

    There's also the issue (which I've seen baffling even experienced
    Pythonistas who had been away from the language for a while) that
    if you see:

        >>> cmp(a, b)
        1
        >>> cmp(a, a)
        0
        >>> 

    you might be tempted to believe that cmp() also returned a truth
    value, whereas in reality it can return three different values
    (-1, 0, 1).  If ints were not (normally) used to represent
    Booleans results, this would stand out much more clearly as
    something completely different.


Specification

    The following Python code specifies most of the properties of the
    new type:

        class bool(int):

            def __new__(cls, val=0):
                # This constructor always returns an existing instance
                if val:
                    return True
                else:
                    return False

            def __repr__(self):
                if self:
                    return "True"
                else:
                    return "False"

            __str__ = __repr__

            def __and__(self, other):
                if isinstance(other, bool):
                    return bool(int(self) & int(other))
                else:
                    return int.__and__(self, other)

            __rand__ = __and__

            def __or__(self, other):
                if isinstance(other, bool):
                    return bool(int(self) | int(other))
                else:
                    return int.__or__(self, other)

            __ror__ = __or__

            def __xor__(self, other):
                if isinstance(other, bool):
                    return bool(int(self) ^ int(other))
                else:
                    return int.__xor__(self, other)

            __rxor__ = __xor__

        # Bootstrap truth values through sheer willpower
        False = int.__new__(bool, 0)
        True = int.__new__(bool, 1)

    The values False and True will be singletons, like None.  Because
    the type has two values, perhaps these should be called
    "doubletons"?  The real implementation will not allow other
    instances of bool to be created.

    True and False will properly round-trip through pickling and
    marshalling; for example pickle.loads(pickle.dumps(True)) will
    return True, and so will marshal.loads(marshal.dumps(True)).

    All built-in operations that are defined to return a Boolean
    result will be changed to return False or True instead of 0 or 1.
    In particular, this affects comparisons (<, <=, ==, !=, >, >=, is,
    is not, in, not in), the unary operator 'not', the built-in
    functions callable(), hasattr(), isinstance() and issubclass(),
    the dict method has_key(), the string and unicode methods
    endswith(), isalnum(), isalpha(), isdigit(), islower(), isspace(),
    istitle(), isupper(), and startswith(), the unicode methods
    isdecimal() and isnumeric(), and the 'closed' attribute of file
    objects.  The predicates in the operator module are also changed
    to return a bool, including operator.truth().

    Because bool inherits from int, True+1 is valid and equals 2, and
    so on.  This is important for backwards compatibility: because
    comparisons and so on currently return integer values, there's no
    way of telling what uses existing applications make of these
    values.

    It is expected that over time, the standard library will be
    updated to use False and True when appropriate (but not to require
    a bool argument type where previous an int was allowed).  This
    change should not pose additional problems and is not specified in
    detail by this PEP.


C API

    The header file "boolobject.h" defines the C API for the bool
    type.  It is included by "Python.h" so there is no need to include
    it directly.

    The existing names Py_False and Py_True reference the unique bool
    objects False and True (previously these referenced static int
    objects with values 0 and 1, which were not unique amongst int
    values).

    A new API, PyObject *PyBool_FromLong(long), takes a C long int
    argument and returns a new reference to either Py_False (when the
    argument is zero) or Py_True (when it is nonzero).

    To check whether an object is a bool, the macro PyBool_Check() can
    be used.

    The type of bool instances is PyBoolObject *.

    The bool type object is available as PyBool_Type.


Clarification

    This PEP does *not* change the fact that almost all object types
    can be used as truth values.  For example, when used in an if
    statement, an empty list is false and a non-empty one is true;
    this does not change and there is no plan to ever change this.

    The only thing that changes is the preferred values to represent
    truth values when returned or assigned explicitly.  Previously,
    these preferred truth values were 0 and 1; the PEP changes the
    preferred values to False and True, and changes built-in
    operations to return these preferred values.


Compatibility

    Because of backwards compatibility, the bool type lacks many
    properties that some would like to see.  For example, arithmetic
    operations with one or two bool arguments is allowed, treating
    False as 0 and True as 1.  Also, a bool may be used as a sequence
    index.

    I don't see this as a problem, and I don't want evolve the
    language in this direction either.  I don't believe that a
    stricter interpretation of "Booleanness" makes the language any
    clearer.

    Another consequence of the compatibility requirement is that the
    expression "True and 6" has the value 6, and similarly the
    expression "False or None" has the value None.  The "and" and "or"
    operators are usefully defined to return the first argument that
    determines the outcome, and this won't change; in particular, they
    don't force the outcome to be a bool.  Of course, if both
    arguments are bools, the outcome is always a bool.  It can also
    easily be coerced into being a bool by writing for example "bool(x
    and y)".


Resolved Issues

    (See also the Review section above.)

    - Because the repr() or str() of a bool value is different from an
      int value, some code (for example doctest-based unit tests, and
      possibly database code that relies on things like "%s" % truth)
      may fail.  It is easy to work around this (without explicitly
      referencing the bool type), and it is expected that this only
      affects a very small amount of code that can easily be fixed.

    - Other languages (C99, C++, Java) name the constants "false" and
      "true", in all lowercase.  For Python, I prefer to stick with
      the example set by the existing built-in constants, which all
      use CapitalizedWords: None, Ellipsis, NotImplemented (as well as
      all built-in exceptions).  Python's built-in namespace uses all
      lowercase for functions and types only.

    - It has been suggested that, in order to satisfy user
      expectations, for every x that is considered true in a Boolean
      context, the expression x == True should be true, and likewise
      if x is considered false, x == False should be true.  In
      particular newbies who have only just learned about Boolean
      variables are likely to write

          if x == True: ...

      instead of the correct form,

          if x: ...

      There seem to be strong psychological and linguistic reasons why
      many people are at first uncomfortable with the latter form, but
      I believe that the solution should be in education rather than
      in crippling the language.  After all, == is general seen as a
      transitive operator, meaning that from a==b and b==c we can
      deduce a==c.  But if any comparison to True were to report
      equality when the other operand was a true value of any type,
      atrocities like 6==True==7 would hold true, from which one could
      infer the falsehood 6==7.  That's unacceptable.  (In addition,
      it would break backwards compatibility.  But even if it didn't,
      I'd still be against this, for the stated reasons.)

      Newbies should also be reminded that there's never a reason to
      write

          if bool(x): ...

      since the bool is implicit in the "if".  Explicit is *not*
      better than implicit here, since the added verbiage impairs
      redability and there's no other interpretation possible.  There
      is, however, sometimes a reason to write

          b = bool(x)

      This is useful when it is unattractive to keep a reference to an
      arbitrary object x, or when normalization is required for some
      other reason.  It is also sometimes appropriate to write

          i = int(bool(x))

      which converts the bool to an int with the value 0 or 1.  This
      conveys the intention to henceforth use the value as an int.


Implementation

    A complete implementation in C has been uploaded to the
    SourceForge patch manager:

        http://python.org/sf/528022

    This will soon be checked into CVS for python 2.3a0.


Copyright

    This document has been placed in the public domain.



Local Variables:
mode: indented-text
indent-tabs-mode: nil
fill-column: 70
End:




From johnroth at ameritech.net  Tue Apr  2 20:52:10 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 2 Apr 2002 20:52:10 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net>
Message-ID: <uaknnsdgu514a6@news.supernews.com>

"David Goodger" <goodger at users.sourceforge.net> wrote in message
news:B8CEABB2.20C5A%goodger at users.sourceforge.net...
> Here's a serious proposal, safe to post now that April Fool's is over.
> Please read & comment.
>

Given the complexity of what has to be done, it seems
reasonably minimal, and reasonably readable. I don't
particularly like the table syntax, (it's too much typing
for my taste) but I don't have a reasonable alternative.

The one question I've got is very, very basic:

why is it in a docstring? Lots of people use docstrings for
relatively short, online prompts. Some other mechanism should
exist for extensive documentation that is intended for translation
to some other format for eventual printing or display independent
of the program.

I much prefer not to have this stuff in the compiled
module.

John Roth




From jeff at ccvcorp.com  Mon Apr 22 18:23:42 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 15:23:42 -0700
Subject: Interacting with other Windoze apps
References: <mailman.1019506545.30012.python-list@python.org>
Message-ID: <MPG.172e2d3bdb060f4d9896c2@news.nwlink.com>

In article <mailman.1019506545.30012.python-list at python.org>, 
brad-g-johnson at uiowa.edu says...
> Hello,
> 
> I'm a new Python programmer and have a question about "interacting" with
> other Windoze applications.  I'd like to send key strokes/characters to an
> application window, but have not been able to find any documentation on
> this.  Is it possible?  For example, suppose a user has Microsoft Word open.
> Is it possible to send an "Alt + F" then an "O" to the Word window to
> activate the *F*ile menu and select *O*pen?

This is not possible directly through Python.  However, using the 
Win32 extensions, it is possible to get access to the Windows 
Scripting Host (WSH), which includes (among many other things) 
the ability to activate another application, and to send 
keystrokes to the currently active application.  In order to do 
this, you'll need to find references for the WSH, most of which 
will probably be aimed at Visual Basic users -- WSH can be driven 
by any language with an ActiveScripting interface, but MS 
(unsurprisingly) focuses mostly on its own languages, VB and (to 
a much lesser extent) JScript/ECMAScript.  You'll also need to 
read up a bit on Python's COM framework (part of the Win32 
extensions), through which you can access the WSH.

Another alternative, which may or may not be easier, is to 
directly control applications that expose a COM automation 
interface (which many programs, including the major MS Office 
apps, do).  Thus, instead of asking Windows to send keystrokes to 
Word that result in  _F_ile/Open being selected, you can directly 
send Word a request to load a particular file.  You'll need to 
determine the object model for each application individually -- 
MS typically includes fairly good documentation for this within 
each application, but again, it focuses primarily on Visual 
Basic, and you'll have to translate to Python yourself.  If 
you're only interested in controlling a small number of different 
programs, this is probably the most effective way to go, but if 
you need to control arbitrary applications (or an app which 
specifically does not expose a COM interface) then you'll have to 
use the WSH as an intermediary.
 
-- 

Jeff Shannon
Technician/Programmer
Credit International


From lenssen at hitnet.rwth-aachen.de  Fri Apr 19 16:47:09 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Fri, 19 Apr 2002 22:47:09 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE><NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com><mailman.1019068175.11920.python-list@python.org><99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com><mailman.1019233076.19513.python-list@python.org><3CC0578E.822EB961@metaslash.com> <mailman.1019240861.31256.python-list@python.org>
Message-ID: <a9pvp0$kqj$1@nets3.rz.RWTH-Aachen.DE>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1019240861.31256.python-list at python.org...
>..
> It hadn't occurred to me that exceptions would be any faster than
> .get().  I didn't use a predefined dictionary since the OP was looking for
an
> equivalent to C's ternary operator (which I took to mean "one line")
>..

The OP has to add that his ternary operators actually can span several lines
for readability. (I'm not a big fan of saving vertical space when
readability suffers.)

language = "german";
gender = "m";

verboseGenderMale = (language == "english") ?
        "male" : "m?nnlich";
verboseGenderFemale = (language == "english") ?
        "female" : "weiblich";
verboseGender = (gender == "m") ?
        verboseGenderMale : verboseGenderFemale;


And no, this is not very realistic code and I would likely do it all
different.

function getVerboseGender(gender)
{
    verboseGender = "";

    isMale = (gender == "m");
    isFemale = (gender == "f");

    if (isMale || isFemale)
    {
        switch (self->language)
        {
            case "german":
                verboseGender = (isMale) ?
                        "m?nnlich" : "weiblich";
                break;
            default: // english
                verboseGender = (isMale) ?
                        "male" : "female";
        }
    }
    else
    {
        message = "Gender exception: asexual";
        error->raise(message);
    }

    return verboseGender;
}


(Nope, I guess in the end I would put that data where it belongs, outside
the source.)






From sholden at holdenweb.com  Tue Apr 30 09:37:09 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 09:37:09 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>	<3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca> <mailman.1020126468.7915.python-list@python.org>
Message-ID: <G8xz8.14199$xy4.5322@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote
> > [James J. Besemer]
> >
> > > I don't interpret this strongly worded and slightly sarcastic critique
to
> > > be hostile.
>
Fran?ois Pinard wrote:
>
> > For me, sarcasm and strong wording are not welcome.  There are ways to
> > disagree, and even disagree much, while staying friendly.
>
> > Americans tend to value straight talk.  I put it all under the
> >  category of effective communication.  Being French
> > Canadian, this may be alien to you.

Well, speaking as a US-based Englishman (from Yorkshire, where sarcasm comes
as naturally as calling your best friend a stupid bastard) I have to say my
experience is that *some* Americans will *say* they value straight talk
until they actually receive some, Then one is being "aggressive".

Seriously, little mild-mannered me had to undergo self-administered behavior
therapy to stop me from saying what I really thought, just in the interests
of "getting along with my colleagues". The fact that they were building crap
web sites in a piss-poor language with no architecture and not even the
slightest acknowledgement that security might be important cut little ice
with the team, but we all got on a lot better after my lobotomy.

We got on even better after I left, and now I haven't spoken to any of them
for three years we are the best of pals.

> Although I agree with you that sarcasm and rhetorical questions are no
> substitute for a logical argument, they can put an effective emphasis on
an
> otherwise factual and logical post such as Logjam's (or whatever his name
is).
>
They can, but they can also alienate someone who is genuinely trying to
reach a meeting of the minds over some topic. Particularly if there's any
ambiguity about the intent of the sarcasm, which is after all a double-edged
sword. What's a rhetorical question?

> Nor is Sarcasm or frank talk necessarily unfriendly.  My close friends
tease
> each other mercilessly.  After an exchange that grew a bit tense at
points,
> Steve Holden and I emerged better friends and better acquainted with each
other
> than I think we would have otherwise.  Many people rely on close friends
to
> utter the bare truths that mere acquaintences would avoid.
>
So would this be a good time to say "Shut up, you stupid bastard"? In fun,
of course. Seriously, I'm joking. [See what I mean about the ambiguity?]

> I think you need to... [all together now]... Lighten Up.
>
> > I prefer intelligent discussions between nice people.
>
> Yeah, "Mean People Suck."
>
Indeed they do. And no matter where you go, there you are.

> Incidentally, by my count, *I* am the only victim of an unprovoked
personal
> attack on this list today.
>
stop-counting-it-will-only-depress-you-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/





From whitney at pk3i.com  Tue Apr  2 02:46:17 2002
From: whitney at pk3i.com (Whitney Williams)
Date: 1 Apr 2002 23:46:17 -0800
Subject: Zope query: importing packages from External Methods
Message-ID: <934d5a7e.0204012346.262d975b@posting.google.com>

I have created a package and placed it into a directory which I
believe should be visible to Zope and/or it's associated python
installation.

However, the normal package import semantics don't seem to be
available -- i.e., the sub-modules are not added as attributes to the
root module, although I have placed __init__ files in the relevant
directories (and precompiled the modules with mixed results).

I have confirmed this package does work as expected if invoked via an
external python interpreter.

I have tried ZOPE\lib\python\Shared, PYTHON\site-packages, and even
ZOPE\Extensions.

=> I don't want to litter my Extensions directory with 50 separate
module files
=> I want to enforce a hierarchy on the module usage
=> Since I want to reuse the package(s) from outside code, I would
prefer not to use a Zope-specific mechanism (i.e., a Product).

I am certain I am overlooking the obvious, but I can't find anything
in the documentation.

PS: In the course of experimentation, I have noticed that
ZOPE\lib\python\Shared\DC behaves as desired -- but I cannot replicate
this behavior in my own packages placed here (I have even tried 'cp
-Rap DC/ DCK/', preserving attributes, but the clone directory does
not behave as expected?!).

What am I missing? 

Any ideas would be appreciated.


From aahz at pythoncraft.com  Mon Apr 15 10:57:51 2002
From: aahz at pythoncraft.com (Aahz)
Date: 15 Apr 2002 10:57:51 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <a9eppf$ghv$1@panix1.panix.com>

In article <a9cv63$qhu$1 at bob.news.rcn.net>,
Raymond Hettinger <python at rcn.com> wrote:
>
>The help and tutor maillists indicate that the use of input() rather than
>raw_input() is widespread.  I think this function is so harmful and so
>likely to be used that it should be deprecated hasta pronto (right away).
>When really needed, eval(raw_input()) is easy enough to code and explicit
>about what it does.

Write a PEP.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From chris.myers at ingenta.com  Fri Apr 26 08:58:42 2002
From: chris.myers at ingenta.com (Christopher Myers)
Date: Fri, 26 Apr 2002 08:58:42 -0400
Subject: Is there a __start__ method?
References: <3CC86462.72244834@ingenta.com> <20020425165056.E14065@unpythonic.net>
Message-ID: <3CC94F02.F6B68490@ingenta.com>

Jeff Epler wrote:
> 
> What good will it do you for your GUI to appear on the screen if it's
> frozen anyway until this process of "accessing 3 separate URLs"
> completes?
> 

Good point, but the answer is, the first time I tried to run the
"atLaunch" method I created, I didn't know what was going on while I
waited for the gui to post.  Maybe I could make a splash screen with a
status bar, hourglass, or watch face or something . . .

The other ideas below seem liek too much consdering the simplicity of my
app, but thanx for the quick response.

> You can use the .after_idle() method in tk to schedule something to
> happen when the gui becomes "idle".  From the tk documentation:
>     after idle script ?script script ...?
>         Concatenates the script arguments together with space separators
>         (just as in the concat command), and arranges for the resulting
>         script to be evaluated later as an idle callback. The script will
>         be run exactly once, the next time the event loop is entered
>         and there are no events to process.  The command returns an
>         identifier that can be used to cancel the delayed command using
>         after cancel. If an error occurs while executing the script then
>         the bgerror mechanism is used to report the error.
> however, while the "script" executes, the GUI will block.
> 
> You could arrange a system using Tcl file handlers, but last I knew
> _tkinter.createfilehandler did not exist on Windows. (The underlying Tcl
> API is marked "unix only" in the Tcl documentation)
> 
> Lastly, you could use a thread for this processing.  However, as another
> thread recently said, only the main thread can access Tk stuff....
> 
> Jeff

-- 
Christopher Myers, Graduate Software Developer 
Ingenta, Inc.
12 Bassett St.
Providence, RI  02903
ph:  401.331.2014 x 102
em:  chris.myers at ingenta.com
aim: chrismyers001




From tanzer at swing.co.at  Thu Apr 11 03:01:39 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 11 Apr 2002 09:01:39 +0200
Subject: binding a reference to a variable
In-Reply-To: Your message of "Wed, 10 Apr 2002 15:42:43 GMT."
             <yu99662zmufg.fsf@europa.research.att.com>
Message-ID: <E16vYaR-0005FL-00@tswings.swing.cluster>

Andrew Koenig <ark at research.att.com> wrote:

> The point of this example is the . operator, which takes a pattern
> (by value) as its left argument and a variable (by reference) as its
> right argument.  The result is a pattern that matches the same strings
> as the left argument, but if the match is successful, it assigns
> the substring matched to the variable that is the right argument.
>
> As it happens, I can implement much the same idea in C++, though
> the exact details of operator overloading make it look somewhat
> different:
>
>     string id, domain;
>     match(name, pos(0) + (break("@") >> id) + "@" + (rpos(0) >> domain));
>
> and I can also implement a similar idea in ML, though I don't
> think it's worth going into details here.
>
> So I was thinking about how I would do something similar in Python.
>
> Focusing just on the  (break("@") . id)  subexpression, how would
> one implement it?  That's what led me on this path.

How about using a class to implement the matcher and assigning the
matched substrings to a namespace provided by the matcher instance in
question?

Using `* args, ** kw` trickery, you might be able to trade readability
of the matcher code for ease of the matcher clients. For instance:

    m(name, m.pos(0) + m.break(id="@") + "@" + m.rpos(domain=0))

YMMV,

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





From arthur.siegel at rsmi.com  Thu Apr  4 12:49:15 2002
From: arthur.siegel at rsmi.com (arthur.siegel at rsmi.com)
Date: Thu, 4 Apr 2002 12:49:15 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <OFA343891A.563BA6ED-ON85256B91.00608529@RSMi.com>

>[Guido]
>> I don't think Boolean algebra is a big subject.

>[David Eppstein]
>> There are advanced graduate math books on Boolean algebra.
> >[...]
> >But these books' content is almost completely irrelevant for day-to-day
> >computer programming.

>[Mark]
>>Don't you figure that's what Guido meant by saying it's not a 'big' subject?
>>I mean, there are advanced philosophy books on Heidegger...

Damn willies I tell you.

Guido will always be right to a following that will work to find
the meaning that makes him right.

Difficult *not* to be able to do that.

Probably a known phenom.

A scary thing to watch -  as in any phenom watched
from the outside with the participants having
a lowered self-awareness.

And not a criticism of Guido himself.

Of which I *do* have a few.

Art






From cimarron+google at taylors.org  Sat Apr 20 17:35:33 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 20 Apr 2002 14:35:33 -0700
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <29e28c51.0204201335.6e0856da@posting.google.com>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message news:<a9kci2$dv0$1 at nets3.rz.RWTH-Aachen.DE>...
> Not that it's a big deal since there's many ways to do that, but can I use
> something like
>     verboseGender = (gender == 'm') ? 'male' : 'female'
> in Python?

Will this do what you want?

verboseGender = ('female','male')[gender=='m']

Cim


From emile at fenx.com  Thu Apr 18 17:40:12 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 18 Apr 2002 21:40:12 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <I6Fv8.36158$vF6.1093396@news2.tin.it> <mailman.1019162916.19664.python-list@python.org>
Message-ID: <03Hv8.35853$HH5.2764@rwcrnsc51.ops.asp.att.net>

"Lulu of the Lotus-Eaters" <mertz at gnosis.cx> wrote in message
news:mailman.1019162916.19664.python-list at python.org...
> Alex Martelli <aleax at aleax.it> wrote previously:
> |class FunkyReprMeta(type):
> |    def __repr__(cls):
> |        return "My wonderful class %s" % cls.__name__
> |class FunkyReprObject:
> |    __metaclass__ = FunkyReprMeta
> |class C(FunkyReprObject):
> |    "whatever you want here"
>
> I don't really understand what a metaclass buys you here.  I think I
> might be missing something important, since Alex proposes it.
>
> However, it appears that one could simply define the __repr__ method
in
> the regular class FunkyReprObject, ignore metaclasses altogether, and
> wind up with the same results.  What is better about Alex' approach?
>

I think you missed the point of OP's question...

>>> class C:
...     pass
...
>>> C
<class __main__.C at 0x007A9548>

>>> class FunkyReprMeta(type):
...     def __repr__(cls):
...             return "My wonderful class %s" % cls.__name__
...
>>> class FunkyReprObject:
...     __metaclass__ = FunkyReprMeta
...
>>> class C(FunkyReprObject):
...     "whatever you want here"
...
>>>
>>> C
My wonderful class C



--

Emile van Sebille
emile at fenx.com

---------



From fperez528 at yahoo.com  Sat Apr 27 09:49:38 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 27 Apr 2002 07:49:38 -0600
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com>
Message-ID: <aaevge$emr$1@peabody.colorado.edu>

Terry Byrne wrote:

> One of my log files was ~60,000 lines long. At first I was using
> re.search to find the relevant error codes. But it took ~5 hours to
> process this log file -- too slow. Knowing that the string module is
> much faster than re, I switched to string.find(), and only if that
> found a certain error code would I run the re operation. (re is needed
> to obtain more specific error information.) That cut my time down to
> ~1 hr and 45 minutes, a big improvement but still too slow.
> 
> Then I used the keyword "pass" whenever I ran an re since each line
> can contain only one error; after running an re it's not necessary to
> look for other error codes at all. Now that 60,000-line log file takes
> less than a minute for Python to handle. Good enough.
> 

why don't you post some code snippet illustrating the before/after? I'm sure 
quite a few would like to see the actual details.

cheers,

f.


From andy47 at halfcooked.com  Mon Apr  8 19:29:28 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Mon, 8 Apr 2002 23:29:28 +0000 (UTC)
Subject: Need help with distutils
References: <mailman.1018083739.10368.python-list@python.org> <Xns91EA988AA8260andy47halfcookedcom@203.109.252.31> <slrnab3gei.1i0.akuchlin@crystal.mems-exchange.org>
Message-ID: <Xns91EB605BFDDADandy47halfcookedcom@203.109.252.31>

"A.M. Kuchling" <akuchlin at crystal.mems-exchange.org> wrote in 
news:slrnab3gei.1i0.akuchlin at crystal.mems-exchange.org:

> In article <Xns91EA988AA8260andy47halfcookedcom at 203.109.252.31>, 
>    Andy Todd wrote:
>> notice that the pattern you pass to global-exclude has to be a Unix-
style 
>> "glob" pattern. In that notation '.' is a special character so using 
>> '.cvsignore' doesn't actually match your files ;-) 
> 
> No, '.' is not special in glob notation.

Which means that this is a bug? I can verify that ".cvsignore" doesn't work 
and "*cvsignore" does.
  
> 
>> Also note that this isn't properly implemented for distutils on Windows 
or 
>> Mac. If you try and package a source distribution on Windows, for 
instance, 
>> you will get all of your CVS directories in the .zip file as well. 
> 
> Can you please file a bug for this?
> 
> --amk
> 

Certainly can, we've found a couple of other issues with distutils when 
trying to package PythonCard. I'll switch this discussion to the distutils-
sig.

Regards,
Andy
-- 
Contents free posts a speciality


From fperez528 at yahoo.com  Wed Apr 17 01:20:00 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 16 Apr 2002 23:20:00 -0600
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <slrnabmtk0.a8q.ak@ak.silmarill.org> <slrnabp92n.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <a9j0nn$s7n$1@peabody.colorado.edu>

phil hunt wrote:

>>I think the idea behind pass is that usually not having anything in
>>a conditional is a mistake, so we want to be informed about it. Putting
>>pass is like saying "I'm sure that's what I want."
> 
> I'm sure what I want, without needing to explicitly say it.
> 

And that would put python in the land of the dreaded C construct:
if(this) 
        do_that();
        and_that();

Was the second function meant to be in the if and the developer forgot braces? 
Or is it meant to be truly outside and it's misindented? 

You may have known what you wanted without being explicit when you wrote the 
code, but what about the poor sob who has to deal with it two years later 
when you're not around? That's why explicitness is good, particularly when 
implictness opens the way to ambiguity.

Cheers,

f.


From gerson.kurz at t-online.de  Mon Apr  1 04:52:35 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Mon, 01 Apr 2002 09:52:35 GMT
Subject: [ANN] PYNOSPAM 0.1 (and some Embedded-Python-Questions)
References: <3ca4b9e8.10558859@news.t-online.de> <slrnaach4v.2gs.philh@comuno.freeserve.co.uk>
Message-ID: <3ca82861.13359171@news.t-online.de>

On Sat, 30 Mar 2002 23:05:03 +0000, philh at comuno.freeserve.co.uk (phil
hunt) wrote:

>The fact that you've written this demonstrates that Outlook has a 
>way of enabling user-written filters for incoming email. To your 
>knowledge, does Outlook allow filters for outgoing email too?

Yes, but its a really *really* terrible. Don't expect to come out with
a sane mind.

- First of all, the solution works for exactly two mail clients:
Microsoft Outlook, and Microsoft Exchange. Great interoperability, eh?
Whether or not it works for even Microsoft Outlook Express I don't
know. You can see from that that it is not in any way a solution you
can reuse for Eudora, or Netscape, or The Bat. And, worse yet, on the
MSDN site they say MAPI is no longer updated with the advent of .NET,
which is great because to this day there exists no Microsoft Mail
Client using .NET rather than MAPI. 

- Secondly, there are two types of extensions Outlook supports: OLE
Add-Ins, and Exchange Add-Ins. ("Exchange" is basically an old version
of "Outlook"). Neither of which you can write using Pythons COM tools!
Furthermore, OLE Add-Ins are "generic" (the concept of "COM addin" is
present in all Microsoft Office 2000 apps, but "hidden" - check MSDN
if you don't believe it). Plus, AFAIK they only work in Outlook
2000/XP, not in Outlook 98. (Its Microsoft, what did you expect). The
Exchange Extensions are easier to implement (in C++) and work both in
older and newer Outlook clients (until the happy day microsoft decides
no longer to support that API).

- As for your question, there is one interface your extension can
implement (IExchExtMessageEvents), that will give you a chance to
modify the mail before sending. This actually might be a good idea,
because if the end user send their mail using a Microsoft Exchange
Server, you can forget about the SMTP proxy approach. (<irony>And many
people do, because Exchange Server is a socalled "killer app" because
it has shared contacts</irony>). 

- modifying the mail using MAPI is hell. You have a "property storage"
and you can get/set one or more properties for a message from it. Now,
say you want to strip HTML mails (as I do). You find that MAPI defines
a PR_BODY_HTML, (but of course you won't find that in the includes
that come with devstudio6, you need *updated headers* for that:
Download the latest platform SDK). You test with small HTML mails and
it works fine - PR_BODY_HTML is the HTMLized body, and you can change
that. Then you test it and promptly receive some HTML spam that is 17K
in size ("The bitter network administrator"). You query the HTML body
and get MAPI_E_OUT_OF_MEMORY back (get that: on a machine that has 1
GB RAM, and nearly 700MB RAM free, you get out of memory for querying
a 17K message!!!). You look for help in microsoft newsgroups and find
to your utter disbelieve that, if the HTML mail is "too large", you
cannot use the "normal" get/set property method, but must implement a
stream object that reads data from a property stream, and from a
DIFFERENT property none the less called COMPRESSED_RTF. WTF!!!

- To use python from the addin, the C++ written DLL embedds the python
interpreter and runs the spam detection there. 

HTH, Gerson


From vinay_sajip at yahoo.co.uk  Thu Apr 11 11:22:56 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 11 Apr 2002 08:22:56 -0700
Subject: ANN: Logging Module v0.4.2 released
Message-ID: <2e37dc1.0204110722.311fe4b0@posting.google.com>

A new version of the proposed Python standard logging module (as per
PEP 282) has been released. You can get all the information from

http://www.red-dove.com/python_logging.html

There are "download" and "recent changes" links at the top of that
page. The new stuff includes some bug fixes for the Tkinter GUI
configurator, one significant performance optimization, a bug in the
logging receiver example application, an example showing writing
logging records to an RDBMS and changes to the base Filter class to
add some useful functionality.

As always, your feedback is most welcome (especially bug reports,
patches and suggestions for improvement).

Cheers

Vinay Sajip
Red Dove Consultants Ltd.

Changes since the last version:
=================================
Bug fix fileConfig() - setup of MemoryHandler target and errors when
loggers have no handlers set or handlers have no formatters set
logconf.py - seems to hang if window closed when combo dropdown is
showing - added code to close popup on exit
Some tweaks to _srcfile computation (normpath added)
findCaller() optimized, now a lot faster!
Logger.removeHandler now closes the handler before removing it
fileConfig() removes existing handlers before adding the new set, to
avoid memory leakage when repeated calls are made
Fixed logrecv.py bug which hogged CPU time when TCP connection was
closed from the client
Added log_test14.py to demonstrate/test a DBHandler which writes
logging records into an RDBMS using the Python Database API 2.0 (to
run, you need something which supports this already installed - I
tested with
mxODBC)
Made getLogger name argument optional - returns root logger if omitted
Altered Filter to take a string initializer, filtering a sub-hierarchy
rooted at a particular point (idea from Denis S. Otkidach).
Added log_test15.py to test Filter initializer
Minor docstring changes


From SBrunning at trisystems.co.uk  Wed Apr 24 11:59:37 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 24 Apr 2002 16:59:37 +0100
Subject: Probably I'm just too stupid, but...
Message-ID: <31575A892FF6D1118F5800600846864DCBD05E@intrepid>

> From:	redflag at directbox.com [SMTP:redflag at directbox.com]
> 
	(snip)

> why didn't "c" work as I imagined?
 
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> def c():
...     for i in range(10):
...         print i
...
>>> c()
0
1
2
3
4
5
6
7
8
9
>>>

Note the parentheses when calling the function.

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




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




From starx at pacbell.net  Tue Apr  9 02:49:36 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 9 Apr 2002 06:49:36 GMT
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com> <3CB274D2.A7554C79@engcorp.com>
Message-ID: <Xns91EAF2930CA4ERASXnewsDFE1@130.133.1.4>

Peter Hansen || Mon 08 Apr 2002 09:57:54p:

> Chris Spencer wrote:
>> 
>> 4.  Did I mention that the documentation is both incomplete and
>> odious?  I should not have to constantly do a "find in files" to find
>> out what parameters a damned method call takes because the way
>> wxPython implemented a method is different than the wxWindows way and
>> someone forgot to make a note of it in the rubbish passed off as a
>> "manual". 
> 
> Although I share your pain, I don't complain much about it since I'm
> well aware of the effort required to put together something like this
> and maintain it, and I don't believe the authors would appreciate
> my condemnation as much as they would my assistance.
> 
> Maybe you could contribute some of what you learned back to the
> project and offer to help improve the documentation?

Hm. We have the wxpython wiki thing*, a user-updated website. Perhaps we 
could use this. We could translate the wx documentation to something 
like this, and let people make comments about what they need and what 
they don't understand and what isn't explained properly and what's left 
out while they work with the documentation that we already have. 
Something to make contributing an ongoing and natural and imediate 
process.

The only exapmle I've seen like this so far is the PHP refrence... much 
useful information is contained in the comments on each page. 

*http://wiki.wxpython.org/

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From phd at phd.pp.ru  Tue Apr 30 05:02:05 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 30 Apr 2002 13:02:05 +0400
Subject: How to retrive attached file from newsgroup mail
In-Reply-To: <63170f57.0204300045.788bc634@posting.google.com>; from sanjay2kind@yahoo.com on Tue, Apr 30, 2002 at 01:45:42AM -0700
References: <63170f57.0204300045.788bc634@posting.google.com>
Message-ID: <20020430130205.G15421@phd.pp.ru>

On Tue, Apr 30, 2002 at 01:45:42AM -0700, sanjay wrote:
>   Im using NNTP to retrive the newsgroups mails to display on web
> pages.The email is multipart i,e text/html with doc,pdf,gif file has
> attached. Is there any python script to get all the plain text/html
> message with attached file from mail.

   http://phd.pp.ru/Software/Python/#extract_mime,
http://phd.pp.ru/Software/Python/#mimedecode and faster mirrors at
http://phd.by.ru/Software/Python/ and http://phd2.chat.ru/Software/Python/

   I am working on rewriting mimedecode.py to use email package. Actually
I've finished rewriting it, but the code requires Python 2.3 (i.e. it
requires my latest patches for the email package).

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




From gerhard at bigfoot.de  Sat Apr 13 18:41:37 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 14 Apr 2002 00:41:37 +0200
Subject: 2.1 vs. 2.2
In-Reply-To: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <20020413224137.GB6470@lilith.my-fqdn.de>

* Jens Baader <nospam at nospam.com> [2002-04-13 23:46 +0200]:
                      ^^^^^^^^^^

If you don't have permission of the nospam.com domain owner to use their
domain, you should stop abusing their services *now*. If you don't want
to get spam, I'd recommend you get a throwaway email address at a
freemailer.

> I'm downloading Python 2.2.1 right now and I wonder why you still
> produce bugfix releases for the old 2.1 development line.

Some projects are still using the Python 2.1 line, for example Zope.
Also, bugfix releases for older Python versions are done by volunteers,
not by the Python core team.

> It seems that Python 2.2 is somewhat broken.

It isn't. Bugfix releases are a service for those people that
standardized on a certain Python version for the current stable release
of their product. For example, AFAIK all Zope 2.4.x releases use Python
2.1.x. So a bugfix release for Python 2.1 is what Zope users need (I'll
ignore Zope 2.5 for the moment, dunno how stable it is).

In fact the main reason for releasing Python 2.1.3 was a fix for a bug
experienced in Zope.

> If not what's the reason that keeps the people from upgrading to 2.2?

Don't know. All the FUD that's currently spread about Python's
stability? ;-)

> Another question: When will we see an official ISO/ANSI standard for
> Python? [...]

Most probably: never.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From logiplexsoftware at earthlink.net  Fri Apr 19 12:15:52 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 19 Apr 2002 09:15:52 -0700
Subject: Trinary operator?
In-Reply-To: <99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com>
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
	<NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>
	<mailman.1019068175.11920.python-list@python.org>
	<99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com>
Message-ID: <20020419091552.671afc71.logiplexsoftware@earthlink.net>

On Fri, 19 Apr 2002 10:37:23 +0100
Dale Strickland-Clark wrote:

> >
> >Or it might be preferable to use a dictionary:
> 
> But several orders of magnitude more processing. Certainly not for
> tight, efficient loops.
> 

It does take considerably longer, but longer is still relative:

import time
gender = 'f'

t = time.time()
for i in range(100000):
    verboseGender = (gender == 'm') and 'male' or 'female'
print time.time() - t

t = time.time()
for i in range(100000):
    verboseGender = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
print time.time() - t

0.178238987923
0.567307949066

So for 100,000 iterations, and/or takes around .2s and a dict takes around .6s
(PIII 800MHz).  I'd still weigh in favor of the dict for most applications,
although there may be situations where this would make a difference.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From gbreed at cix.compulink.co.uk  Tue Apr  9 06:24:26 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Tue, 9 Apr 2002 10:24:26 +0000 (UTC)
Subject: Discussion of Python midi library
References: <mailman.1018286064.26816.python-list@python.org>
Message-ID: <a8ufgq$n1u$1@thorium.cix.co.uk>

Mike Brenner wrote:

> Here is a suggestion as to what to encapsulate (each one is harder than 
> the next.
> 
> 1. Midi Instruments (preferably the hundred general midi instruments 
> from grand-piano up through gunshot and the hundreds of xg midi 
> instruments available on the more advanced synthesizers and 
> sequencers). The attributes of the instruments would include:

There are 128 General MIDI instruments.  This suggestion sounds like 
overkill, but as you're being ambitious it might be a good idea.  Make 
sure you have the basic library in place first.

> 2. Midi Selection Sequences. Each midi hardware will have certain 
> selection sequences, bank selection commands, patch selection commands, 
> controllers, etc. There must be a way of doing the machine independent 
> stuff in a machine independent manner and yet there must be a way to 
> insert the device-specific stuff. The currently selected hardware must 
> be part of the model. There must be ways to insert, delete, modify, 
> move, and copy notes and other commands into channels, including 
> setting their velocity (loudness), frequency (pitch), duration, etc. A 
> single command will enter a chord, a triple, a comment such as key, a 
> meta event, etc.

This all sounds like it makes sense.  Treating chords and triples might 
run into problems as the aren't treated as such by MIDI.  If we're doing 
hardware abstraction, I suggest going beyond some of the MIDI limitations 
like with channels and number of notes.

> 3. Image. A string image function must represent every object, possibly 
> in a standard format like xml, musicml, etc.)

This goes beyond what I'd expect from a MIDI library.

> 4. Time. There must be a consistent time model used throughout, which 
> easily represents notes, rests, start, finish, etc.

Certainly!  This is the main encapsulation you need over a verbatim 
translation of the MIDI.

> 5. Graphical representations in staff format, say, in jpeg or png 
> format.
> 
> 6. Reading in a bit-mapped graphical representation (say a jpeg or png 
> file) of a staff into midi format.

More stuff that goes beyond what a MIDI library should concern itself 
with.  Adding hooks for a score editing program makes sense, but staff 
notation shouldn't be special-cased in this way.  It differs from MIDI in 
some fundamental ways, particularly timing, and the two should be clearly 
distinguished.

> 7. Pattern recognition of a wav file into a midi format, choosing the 
> right notes, durations, and instruments.

And this is an extremely ambitious project in itself.


                    Graham


From martin_byrne2 at yahoo.co.uk  Wed Apr 17 12:37:03 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 17 Apr 2002 09:37:03 -0700
Subject: Accessing mdb(access 2000) files on linux using python
Message-ID: <508cb8ae.0204170837.6f800a24@posting.google.com>

Hi,
Just wondering if anyone knows how to manipulate mdb files stored on a
linux webserver using python.Using win32com on windows is fine but on
linux can i manipulate the database using pythons odbc connectivity or
is there another way.

Any suggestions, links or sample code would be of great help.

Cheers
B


From tejarex at yahoo.com  Mon Apr 22 13:23:51 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 22 Apr 2002 17:23:51 GMT
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83> <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>
Message-ID: <HGXw8.48383$ro5.4407911@bin6.nnrp.aus1.giganews.com>

"Jarno J Virtanen" <jajvirta at cc.helsinki.fi> wrote in message

> When I think of the final products of software development and
building
> construction I find it hard to compare them.  Yeah, sure there is
some
> sort of an "architecture" that they're based on and they're both
> non-trivial combinations of some sort of building blocks.  But it's
just
> that the software product seems so fragile.  There are of course
methods
> and ways to build high quality software, but no matter how
protectively
> it is built, there's always the chance that some part of the
software
> fails which makes the service provided by the software unusable.

Fragility results from lack of redundancy.  We can understand speech
with noise and prose with typos (which were more common when
everything was written by hand) because normal language is about 50%
redundant.  Its the original error-correcting code.  Engineered
structures are commonly engineered to be at least 50% redundant (ie,
have a safety factor of 2 or more).

Terry J. Reedy




From boud at valdyas.org  Tue Apr 23 15:56:52 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 23 Apr 2002 18:56:52 -0100
Subject: Destroying instances with PyQT
References: <f0fd5987.0204230240.56804d3f@posting.google.com>
Message-ID: <3cc59504$0$41709$e4fe514c@dreader4.news.xs4all.nl>

Garry Taylor wrote:

> I have attached a QTimer to one of my objects, but I cannot seem to
> get rid of it when I'm finished, which cause the whole object not to
> be destroyed, normally my instances are erased simply by erasing the
> Dictionary in which they reside, which has always worked before the
> QTimer was added. In the QT manual it says you can destruct an
> instance like this:
> 
> QTimer::~QTimer ()
> 
> How can I translate this into Python?
> 

Simple remove the reference to the QObject that is the parent
of the QTimer. The following code removes the owning object
after ten ticks:


from qt import *
import sys

class Q(QObject):

        def __init__(self):
                QObject.__init__(self)
                self.o=O(self)
                
        def delete(self):
                self.o = None

class O(QObject):

        def __init__(self, parent):
                QObject.__init__(self)
                self.parent = parent
                self.i = 0
                self.t = QTimer(self)
                QObject.connect(self.t, SIGNAL("timeout()"), self.p)
                self.t.start(0, 0)

        def p(self):
                self.i += 1
                print "bla", self.i
                if self.i == 10:
                        self.parent.delete()

q = Q()
app=QApplication(sys.argv)
app.exec_loop()


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


From donn at u.washington.edu  Fri Apr 26 19:27:37 2002
From: donn at u.washington.edu (Donn Cave)
Date: 26 Apr 2002 23:27:37 GMT
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to do math?)
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com> <slrnacjga3.2uv.huaiyu@gauss.almadan.ibm.com>
Message-ID: <aacnp9$1b34$1@nntp6.u.washington.edu>

Quoth huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu):
| David Bolen <db3l at fitlinxx.com> wrote:
|> instead.  The default output at the interpreter is to use repr() to
|> print an object, which shows the highest precision possible to best
|> assure that its output could be used to re-create the object.  The
|> print statement will use str() which is designed for "prettier" output
|> and will do some rounding.
|
| The current arrangement between repr and str has some shortcomings:
|
| - nested structures use repr for elements
| - repr is constrained by the desire to be inverse of eval for simple types
| - default repr for instances are optimized for object id, not content
| - str cannot differentiate between 2 and "2"
|
| It appears that another display level is needed, somewhat between repr and
| str.  Let's call it 'disp' here.  It is intended to be used for display in a
| controlled way, esp for nested structures.

Minus the floating point problem, is this function not identical to repr()?

It seems to me the crux of the problem is the notion that
repr is the inverse of eval.  It's fine with me if people
want to use this feature, but it is not by any stretch of
the imagination a reliable characteristic of Python objects.
I mean, eval(repr(sys.stdin))?  Class instances? etc.  We're
not looking at just an occasional exception, they're all over.
Again, if it works for someone, then that's great, but it's
a lost cause as a reliable feature.

Meanwhile, I really think str is poorly served by the common
notion that it's just prettier.  In my opinion, str is most
precisely suited to string conversion:  example that comes
to mind is Message.__str__ in rfc822.py, which returns its
header data.  repr() returns the usual information that it's
a Message instance.  Which is "prettier"?  The one that gives
you the kind of information you wanted!

In a context where I expect strings, I will use str(), in case
a data type conversion is necessary.  I use repr() when I want
a string that represents the object as such.  Now I'm not _using_
the object as data that must be a string, I'm generating a
description of it as an object.

So what about those damned floats?  Well, apparently we're generating
a more detailed description than we used to, and obviously that's
good for the people who want to marshal that way, but it's bad for
everyone.  I don't think this marshalling feature is worth it.

	Donn Cave, donn at u.washington.edu

| Here's what I think it should do:
|
| - disp(0.05)    -> "0.05", just like str(0.05).
| - disp("abc")   -> "'abc'", just like repr('abc').
| - disp([a,b,c]) -> '['+', '.join(map(disp, [a,b,c]))+']'.
| - disp(obj)     ->  obj.__disp__().
|
| Having and additional display method also reduces the need to override
| __repr__, which is the source of very confusing diagnostics when not done
| right (ie, when diagnostics are needed).
|
| Huaiyu


From donn at drizzle.com  Tue Apr 30 02:35:12 2002
From: donn at drizzle.com (Donn Cave)
Date: Tue, 30 Apr 2002 06:35:12 -0000
Subject: Update locals()
References: <20020427195725.GA16217@devel.trillke> <20020428205757.GH16217@devel.trillke> <02042900255602.25256@arthur> <mailman.1020042230.9019.python-list@python.org> <q7bz8.76797$vF6.2277789@news2.tin.it>
Message-ID: <1020148511.232812@yasure>

Quoth Alex Martelli <aleax at aleax.it>:
...
| Right, and THAT is the issue: exec-without-explicit-dicts DOES muck
| with the local namespace in ways that cannot be controlled and thus
| lead to unmaintainable code, hard-to-find bugs, and so on.

I have to say, I don't think I've seen this much horror expressed
about "exec" before today.  I certainly don't reach for exec as the
tool for every problem, but I do use it in one application in
class statements where its intent is precisely to muck with the
local namespace.  It works fine, isn't hurting anything and isn't
going to randomly write zeros across my namespace.  I'm not saying
its uses are many, but ideas like "abomination" are for the religious.

I haven't read every word of this exchange, but I did take a look
at compile() and new, and if this is the approved way to do things
we need a little more documentation ("You need to know a great deal
about the interpreter to use this!", as new.__doc__ says, is a
starter, but perhaps could be more specific.  I was stumped, anyway.

	Donn Cave, donn at drizzle.com


From garry at sage.att.com  Tue Apr  2 16:56:15 2002
From: garry at sage.att.com (Garry Hodgson)
Date: Tue, 2 Apr 2002 21:56:15 GMT
Subject: Why do Windows sockets break after exactly 4 minutes?
References: <mailman.1016754947.17945.python-list@python.org>
Message-ID: <3CAA28FF.9E94F7F5@sage.att.com>

Geoffrey Talvola wrote:
> 
> I've noticed that Windows sockets seem to close all by themselves if there
> is no activity for 4 minutes.

your windows machine stays up that long?  wow!

-- 
Garry Hodgson                   Let my inspiration flow
Senior Hacker                      in token rhyme suggesting rhythm
Software Innovation Services    that will not forsake me
AT&T Labs                          'til my tale is told and done.
garry at sage.att.com


From cliechti at gmx.net  Mon Apr  1 16:38:26 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Apr 2002 22:38:26 +0100
Subject: benchmarking question (from newbie)
References: <3CA8BF7B.BE99768B@worldonline.nl>
Message-ID: <Xns91E3E67B8D0CAcliechtigmxnet@62.2.16.82>

Jacq <jdake at worldonline.nl> wrote in news:3CA8BF7B.BE99768B at worldonline.nl:
> I am a student of "language and artificial intelligence".
> I would be very grateful if anyone could advise me on a
> benchmarking question I have:
> 
> What I have is a script that repeatedly calls a machine-learning
> program each time with different parameters/algorithms.
> What I need is to measure the time the machine-learning-programme
> takes running, *independent* of other processes running on the
> same machine.
> (I have no control over what processes are started up
> by other users, and I want to compare the speed of the different
> algorithms/parameter-combinations.)
> 
> I thought I could use something like this:
>     start = time.clock()
>     os.system("call to machine-learning program")
>     stop = time.clock()
>     cpuTime = stop - start
> But testing this on my own machine, I get different values for the same
> program-call depending on what other processes I have started on my
> machine.
> 
> I use Python 1.5 on linux (RedHat 6.1)

appart from time.clock beeing the right one, you could use the
'time' command on *nix/Linux as you anyway start an external app.

you can just add 'time' infront of an other command (shell):
$ time ls -l
and then it prints a statistics at the end.

chris

-- 
Chris <cliechti at gmx.net>



From fredrik at pythonware.com  Wed Apr  3 05:10:21 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 10:10:21 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com> <mailman.1017795008.700.python-list@python.org>
Message-ID: <hyAq8.29178$n4.5840455@newsc.telia.net>

Richard Jones wrote

> Again, have you _read_ the ReST documentation, specifically the discussion
> of  alternatives? It's extensive, and very well written.

do you realize that the ReST documentation is much
longer than the complete Ruby language reference?

I'd rather learn a new programming language than yet
another markup language.  especially when the latter
sucks.

</F>




From stojek at part-gmbh.de  Tue Apr  9 09:05:00 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Tue, 09 Apr 2002 13:05:00 GMT
Subject: how to read hardware info on Win NT
Message-ID: <3cb2e686.21131515@news.easynews.net>

Hi,

could anyone please give me a hit how to get hardware information with
Python? I need the id of my ethernet card and my harddisk.

Thanks,
Marcus


From eppstein at ics.uci.edu  Sat Apr 27 15:16:16 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 27 Apr 2002 12:16:16 -0700
Subject: Removing dupes from a list
References: <mailman.1019853649.15936.python-list@python.org> <Jvjy8.66361$vF6.2030499@news2.tin.it> <a2972632.0204271058.72ac51cf@posting.google.com>
Message-ID: <eppstein-618AFD.12161627042002@news.service.uci.edu>

In article <a2972632.0204271058.72ac51cf at posting.google.com>,
 les_ander at yahoo.com (les ander) wrote:

> > http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560
> > 
> > 
> > Alex
> 
> 
> def unique(List):
>   return list(dict(zip(List,[None]*len(List))))

The ASPN recipe also works for mutable list members.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From Gabe.Newcomb at noetix.com  Fri Apr  5 17:21:14 2002
From: Gabe.Newcomb at noetix.com (Gabe Newcomb)
Date: Fri, 5 Apr 2002 14:21:14 -0800
Subject: compare function in sort function of lists
Message-ID: <0C7CA8D8DF75494EB09AB6016990107F011B1982@NOXMAIL.noetixad.com>

Hi all:
	I'm trying to sort a list numerically, e.g.,  so that 11 comes
AFTER 2. The documentation says I need to pass sort() a compare
function...but I can't find an example of this.

This should be an easy one for you folks :)

TIA,

Gabe Newcomb
Software Test Automation Engineer
425.372.2732
Noetix Corporation
www.noetix.com






From akuchlin at ute.mems-exchange.org  Thu Apr 25 12:08:38 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 25 Apr 2002 16:08:38 GMT
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <j4helzdfbl.fsf@informatik.hu-berlin.de> <aa98g0$9i4$1@panix3.panix.com>
Message-ID: <slrnacgag6.98m.akuchlin@ute.mems-exchange.org>

In article <aa98g0$9i4$1 at panix3.panix.com>, Aahz wrote:
> What HOWTO?  There isn't one at http://py-howto.sourceforge.net/

http://pyxml.sourceforge.net/topics/docs.html links to
http://py-howto.sourceforge.net/xml-howto/xml-howto.html, but the text
should really live on pyxml.sourceforge.net, because the master copy
of the text lives in the PyXML CVS tree.

--amk                                                             (www.amk.ca)
  "What is it?"
  "It's a perfectly ordinary electric dog."
    -- Stott and the Doctor meet K9, in "Nightmare of Eden"



From h.jehadalwan at student.kun.nl  Sun Apr 21 11:20:55 2002
From: h.jehadalwan at student.kun.nl (husam)
Date: Sun, 21 Apr 2002 17:20:55 +0200
Subject: c++ vs python!
Message-ID: <3CC2D8D7.20601@student.kun.nl>

If we compare c++ with python, which one is faster when it comes to 2D, 
3D graphics and math. And which one is faster when it comes to design 
and implementation of "heavy" applications?

thanks in advance.



From pyth at devel.trillke.net  Tue Apr 23 13:04:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 23 Apr 2002 19:04:28 +0200
Subject: rlcompleter not calling __getattr__ on [ ]
In-Reply-To: <aa3222$lna$1@peabody.colorado.edu>
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org> <aa2460$534$1@peabody.colorado.edu> <mailman.1019533244.22752.python-list@python.org> <aa2u72$k0a$1@peabody.colorado.edu> <mailman.1019545127.27780.python-list@python.org> <aa3222$lna$1@peabody.colorado.edu>
Message-ID: <20020423170428.GL27395@devel.trillke>

On Tue, Apr 23, 2002 at 01:19:38AM -0600, Fernando P?rez wrote:
> holger krekel wrote:
> >> Currently ipython seems to do these just fine, after removing the following
> >> from the delimiters:
> >> 
> >> readline_remove_delims -/'"[]{}()
> > 
> > does
> > 
> >   b=[]
> >   a=a-b.<tab>
> > 
> > work then? And
> > 
> >   [[].<tab>
> >   
> > 
> > really works, too?
> 
> Well, almost :)
> 
> In [2]: b=[]
> 
> In [3]: a=a- b.<TAB>
> b.__add__           b.__getslice__      b.__ne__            b.count
> b.__class__         b.__gt__            b.__new__           b.extend
> [snip]

:-)

i think that a proper solution should set delimiters to '' and let
rlcompleter do any interpreting magic. To be honest, what we have now
is not sufficient to handle even everyday cases.

i am starting to look into writing completer routines
which handle stuff like

	a=a-[].<tab>
        a=[].<tab>
        a=["".<tab>
        a,b,c.  <tab>
        a,b=c,"""()[]'""".<tab>

correctly. and maybe even 

	().<tab>
	([abc("[]()",()).<tab>

i check back if it starts to work :-)

	holger




From Chris.Barker at noaa.gov  Tue Apr  2 14:17:57 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 11:17:57 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org>
Message-ID: <3CAA03E4.A79BA593@noaa.gov>

WOW! fabulous post, Laura. As a result of mainly this post, I have
changed my thinking somewhat.

Laura Creighton wrote:

> Python does not distinguish between True and
> False -- Python makes the distinction between something and nothing.

> So I out-and-out tell people this.  {} is a dictionary-shaped nothing.
> [] is a list-shaped nothing. 0 is an integer-shaped nothing.
> 0.0 is a float shaped nothing.

Thinking in terms of something vs. nothing is an excellent way to
explain how Python treats these things, but I still have a problem with
it. Who is to say that an empty list is nothing? or 0 for that matter. 0
particularly: it is an integer, and a very useful one at that, I find it
very inconvenient that 0 is always interpreted as nothing (see the
string.find() example). In Python the only object I want consistently
interpreted as nothing is None.

That being said, the proposed "int-in-a-hat" (nice coinage, Laura), does
not solve my problem at all. Everything that is currently interpreted as
nothing will continue to be so interpreted, which is must be so as not
to break probably every module over ten or so lines of code long.

The Python language is very much set in its ways: it relies on the
concept of something vs. nothing, and it has pre-defined what nothing is
for all the standard types in a very useful way. I'd be happier if that
legacy didn't exist, but like many people posting, I feel this halfway
measure is worse than what we have now.

Perhaps for Py3k, if we accept code breakage galor, a real Bool can be
introduced, and we can get rid of de-facto definitions of particular
values of an object being nothing. For now, let's keep it the way it is.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From niemeyer at conectiva.com  Thu Apr  4 11:49:11 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Thu, 4 Apr 2002 13:49:11 -0300
Subject: PEP 235: role of PyType_GenericNew()
In-Reply-To: <xdusn6bamcn.fsf@leto.informatik.uni-freiburg.de>
References: <xduwuvoamgg.fsf@leto.informatik.uni-freiburg.de> <mailman.1017854949.20733.python-list@python.org> <xdusn6bamcn.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <20020404134911.A1046@ibook.distro.conectiva>

> Write a base-class _A in C, which fills the type's tp_alloc-slot with
> its special allocator, for the tp_new slot I simply use PyType_GenericNew.

Ok..

> At the Python level I write a class A which is a subclass of _A. Since

No problems..

> A does not introduce new data members, there is no need to overwrite
> __new__. This way, however, the superclass-allocator (_A.alloc) never
> gets called when I create instances of A...

And that's ok as well. You can't allocate memory area for the same
object instance two times. Indeed, I'm not sure why you're mentioning
that __new__ was not overwritten in python level. The allocator function
called will be the same, no matter if it's overwritten or not (you can't
allocate memory at the python level). Also, introducing data members is
not a reason for overwriting __new__. Ralf, you seem to be confused
about the uses of tp_init (__init__), tp_new (__new__), and tp_alloc
slots. Have you read PEP 253?

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From sholden at holdenweb.com  Mon Apr  8 18:26:57 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 18:26:57 -0400
Subject: CGI multiple select?
References: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>
Message-ID: <xQos8.3303$vT6.1896@atlpnn01.usenetserver.com>

"John" <nobody at nobody.com> wrote in message
news:56ns8.37728$tg4.448055 at vixen.cso.uiuc.edu...
>
> Hi,
>
> In my html form, I have 'multiple select'
>
> <select name=spred multiple>
> <option..>
> <option..>
>
>
> This form invokes CGI routine, where I try to read selected entries from
> form["spred"]
>
> My dilemma is here:
>
> When only one entry is selected:
> form["spred"].value (will be the object selected)
>
> When multiple entries were selected:
> for o in form["spred"]:
>   o.value (--> each of these will be the object selected)
>
> Thus, I need to handle according to situation.. I though my previous
> question on 'isList' can solve this matter, but somehow, all solution
> suggested (isSequenceType, type, ..) cannot distinguish these two cases.
>
I believe the following, from the Library Reference Manual, section 11.2.2,
will apply (but I haven't tested this today):

"""
If the submitted form data contains more than one field with the same name,
the object retrieved by "form[key]" is not a FieldStorage or
MiniFieldStorage instance but a list of such instances. Similarly, in this
situation, "form.getvalue(key)" would return a list of strings. If you
expect this possibility (when your HTML form contains multiple fields with
the same name), use the type() built-in function to determine whether you
have a single instance or a list of instances. For example, this code
concatenates any number of username fields, separated by commas:

from types import ListType

value = form.getvalue("username", "")
if isinstance(value, ListType):
    # Multiple username fields specified
    usernames = ",".join(value)
else:
    # Single or no username field specified
    usernames = value
"""

regards
 Steve





From maxm at mxm.dk  Mon Apr 15 02:14:51 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 15 Apr 2002 08:14:51 +0200
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za> <3CBA43B9.22850DF1@engcorp.com>
Message-ID: <3CBA6FDB.9030706@mxm.dk>

Peter Hansen wrote:

> Of course, those products are well known around here.
> 
> What's less well known is the relationship between Boa-Constructor 
> and Zope.  I strongly suspect there isn't one.
> 
> Can you elaborate on this supposed "incorporation of Zope
> into Boa-Constructor"?


It can, eventually, be used as an editor for Zope. I beleive either 
through ftp or webdav.

regards Max M



From claird at starbase.neosoft.com  Sun Apr 28 16:28:21 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 28 Apr 2002 15:28:21 -0500
Subject: Scripting lanuages for laboratory tests
References: <aagkfm$6kf$1@news5.svr.pol.co.uk> <aahh9q$a1k$1@nyheter.chalmers.se>
Message-ID: <6A0FEBB23B453624.BF22EF486940710C.18CC1559478D96F0@lp.airnews.net>

In article <aahh9q$a1k$1 at nyheter.chalmers.se>,
Jacob Hallen <jacob at boris.cd.chalmers.se.cd.chalmers.se> wrote:
>In article <aagkfm$6kf$1 at news5.svr.pol.co.uk>,
>Etta Makwetta <etienne at coetzeee.freeserve.co.uk> wrote:
			.
			.
			.
>"Everything is a string"-paradigm painful to use.
			.
			.
			.
>Tkinter is much easier to use than plain tk. I have built fairly substantial
>applications in both. You pay a small price in the form of having to have
>both Python and tcl installed on your machine, and there are a few very
>small spots where the tcl base of tkinter shows through.
>
>Jacob Hall?n
>
>-- 

There are wise, experienced people who share these
opinions.  There are similarly qualified people on
the other side.  I think it's most accurate to ad-
vise that "everything is a string" and Tkinter vs.
Tk are close enough issues that there's a large
subjective component to their evaluation.  For me,
"subjective" is not a euphemism for "bad"; it just
means, "likely to depend on personal factors."

The nice part of this whole story is that both
Tcl/Tk and Python/Tkinter are sufficiently light-
weight that a good engineer can get a reasonable
sense of their operation in a day.  *That*'s the
counsel we should give:  take these languages out
for a spin, giving special considerations to the
aspects already mentioned.

Incidentally, the "everything is a string" model
can be quite entertaining, in the hands of a master; 
see <URL: http://wiki.tcl.tk/everything >.
-- 

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


From Gareth.McCaughan at pobox.com  Wed Apr 17 20:33:50 2002
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Thu, 18 Apr 2002 01:33:50 +0100
Subject: Does Python need a '>>>' operator?
References: <mailman.1018937566.11432.python-list@python.org>
Message-ID: <slrnabs53e.1kfm.Gareth.McCaughan@g.local>

Tim Peters wrote:

> PS:  Zorn's Lemmon is the conjecture that lemmons are yellow.  This
>      has almost been proved, where by "proved" we mean talked about
>      energetically on a newsgroup somewhere, and "conjecture" means
>      somebody said so twice.

A friend of mine has a project of classifying all finite
chromatic mathematical fruit jokes. He knows about

    "What's purple and commutes?"
    "What's yellow and equivalent to the axiom of choice?"
    "What's green and uniquely determined by its first Chern class?"

and I think those are all he's found. He conjectures that
there are no more, but the conjecture appears difficult to
prove without tiresome case-splitting. Perhaps a computer
proof is in order. (It would be nicely ironic if it could
find one involving an Appel.)

If physics is allowed, we can have "What's green and described
by Bose-Einstein statistics?". And "What's brown and shrivelled
and carried by the Z particle?".


(Oh, all right. An abelian grape; Zorn's lemon; a lime bundle;
a Cooper pear; a neutral current.)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc


From lac at strakt.com  Wed Apr 10 08:42:29 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 10 Apr 2002 14:42:29 +0200
Subject: What has PEP 285 done to us? 
In-Reply-To: Message from Boris Borcic <borcis@geneva-link.ch> 
   of "Wed, 10 Apr 2002 12:05:35 +0200." <3CB40E6F.90509@geneva-link.ch> 
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org>  <3CB40E6F.90509@geneva-link.ch> 
Message-ID: <200204101242.g3ACgTdj023540@ratthing-b246.strakt.com>

I agree with you that we are seeing things in terms of aesthetics, and
it appears we share a sense of its value, even though we may differ
in the details. But in the last article I switched context without 
leaving a sign-post, a common failing of mine for which I apologise.  
I used your message to launch an appeal to everybody.  That unfairly
focused attention on you, which was an unforseen consequence.  So I
was negligent.  I will try to do better.

Once more with feeling ---
Everybody who casts what happened as a 'struggle between GvR and
myself' rather than a flawed bit of failed communication is either
misunderstanding, or coopting me for their own purposes.  I don't
take kindly to being used like that, people.

re: what is in nature vs what is in perception -- see David Pye
The Nature and Aesthetics of Design -- especially Chapter 11
(function and fiction) and  the latter part of Chapter 13 (the
aesthetics of design) and all of chapter 14 (perception and
looking).

I can't discuss aesthetics here now.  I am not sure that an newsgroup
is ever an appropriate place for such discussions, but this one
right now is not.   Maybe in 6 months or so.  No promises.

Laura Creighton




From lord_elendil at hotmail.com  Fri Apr 12 12:58:47 2002
From: lord_elendil at hotmail.com (Bryce C. Mason)
Date: Fri, 12 Apr 2002 16:58:47 GMT
Subject: Macintosh Development
Message-ID: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>

Hi folks.

I am doing my development on a PC (activeware's distribution) and the final
project will be running on a macintosh.  Umm...I am a little worried that
there is not a command line interpreter in the Mac, at least one I have not
seen.  Help.  How do I port my project to the mac?

Bryce Mason





From hfoffani at yahoo.com  Thu Apr  4 06:41:46 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Thu, 4 Apr 2002 13:41:46 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <a8h2n6$ff2$1@nntp9.atl.mindspring.net>
Message-ID: <20020404064150.752$Gx@news.newsreader.com>

"Andrew Dalke"
> ....
>         if self.connections.has_key(occasion) == 0:
>
> when it should have said
>         if not self.connection.has_key(occasion):
>
> I know this will still work correctly since False == 0,
> but I suspect this means there's many people who are doing
> explicit checks for 0/1 (/whatever?) for truth tests.
>

While at this moment it's not possible to realiable check
for those constructions, it's interesting to note that
pychecker (in Python 2.3) will issue a warning!

I'm not convinced that PEP285 will hassle newbies.
Actually, I even think that it can improve code style.

Regards,
-Hernan





From aahz at pythoncraft.com  Mon Apr  8 16:05:32 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Apr 2002 16:05:32 -0400
Subject: Division change in 3.0! (was Re: Booleans, integer division, backwards compatibility; where is Python  going?)
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com>
Message-ID: <a8st6c$jlp$1@panix1.panix.com>

In article <23891c90.0204080156.27f36d22 at posting.google.com>,
Paul Boddie <paul at boddie.net> wrote:
>Guido van Rossum <guido at python.org> wrote in message news:<3CAEFC4E.2778AC55 at python.org>...
>> 
>> About the change rate in general: it's hard to find the proper pace.
>> Most new features are requested by users (even the division change!)
>
>...which I eventually came to see as a good idea, although I
>personally believe that the way it was announced caused unnecessary
>panic. Postponing mandatory changes in semantics even to the release
>after next, when the frequency of releases appears to be doubling over
>time, might not have been enough to satisfy the most conservative
>users.

Once again, for the record: the mandatory change in the way division
works is going to be in 3.0.

This is not the way it was originally presented (and I agree with you
about the presentation), but it's a good example of Guido changing his
mind after getting input.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From marklists at mceahern.com  Wed Apr  3 11:59:49 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 10:59:49 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <15531.12122.601076.455659@anthem.wooz.org>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEPOKIAA.marklists@mceahern.com>

[Barry A. Warsaw]
> The strongest con for PEP 285 seems to be the natural inclination for
> people to want to write "if cond == True:".

I don't see why you want to blame that inclination on PEP 285.  Folks are
perfectly capable of monstrosities like this without it:

	x = isinstance(foo, bar)
	if x == 1:
		...

If, as Fredrik says, there's already one obvious way to return True (and
that's the int 1), then we already have the problem that people will be
inclined to compare something to this canonical truth in conditional
statements.

I suppose you're saying that PEP 285 makes it worse?  I have no idea whether
you're right.  It certainly doesn't make ME more likely to do it.  ;-)

Cheers,

// mark





From mgilfix at eecs.tufts.edu  Sat Apr 27 15:42:40 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 27 Apr 2002 15:42:40 -0400
Subject: handling different arguments depending on first argument
In-Reply-To: <3CCAF77B.3BEBF259@fxtech.com>; from paul@fxtech.com on Sat, Apr 27, 2002 at 07:09:45PM +0000
References: <3CCAF77B.3BEBF259@fxtech.com>
Message-ID: <20020427154240.F1588@eecs.tufts.edu>

  Sure. Try something like this:

     def my_func (*args):

       mode = args.pop (0)
       if mode == MODE1:
         # Do mode1 stuff with remainder of args
       elif mode == MODE2:
         # Do mode2 stuff
       else:
         raise Exception

              -- Mike

On Sat, Apr 27 @ 19:09, Paul Miller wrote:
> I've written a C function (callable from Python) that can take a
> variable number of arguments, and the arguments after the first are
> dependent on what the first was. For example, the signatures for 2
> "modes" might look like this:
> 
> 	my_func(MODE1, var1, var2, var3=None)
> 	my_func(MODE2, var1, var2=None)
> 
> 	etc.
> 
> Is it possible to handle this kind of different calling convention in
> pure Python? I'd like to write the equivalent of an "overloaded" Python
> object constructor, which takes various arguments depending on the
> first.
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (paul)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From boud at valdyas.org  Thu Apr 11 17:35:55 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Thu, 11 Apr 2002 23:35:55 +0200
Subject: ANN: Source code for GUI Programming with Python Using the Qt Toolkit
Message-ID: <3cb60427$0$71652$e4fe514c@dreader4.news.xs4all.nl>

I've cleaned up the source code for all the examples in the book, and
ported it to PyQt 3.1. While awaiting the Opendocs webforum for the
book, I've put up the packages on my own webserver.

http://www.xs4all.nl/~bsarempt/python/pyqt_qt2.tgz

     Source for Python 1.x, 2.x, Qt 2.x, 3.x and PyQt
     2.x. (Grab this if you use the PyQt that comes with
     your Linux distribution.)

http://www.xs4all.nl/~bsarempt/python/pyqt_qt2.tgz

     Source for Python 2.x, Qt 3.x and PyQt 3.x. (Grab this
     if you're current with the PyQt from 
     http://www.riverbankcomputing.co.uk

The book can be read online at:
    http://www.opendocspublishing.com/pyqt

Or bought, for instance from Amazon:  
 
http://www.amazon.com/exec/obidos/ASIN/0970033044/qid=1016980570/sr=8-1/ref=sr_8_1_1/102-2842252-0596931


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


From brian at sweetapp.com  Wed Apr  3 11:52:49 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 3 Apr 2002 08:52:49 -0800
Subject: How best to write this try/except block?
In-Reply-To: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <006101c1db2f$fcc77340$445d4540@Dell2>

> try:
>     value = d[key]
>     foo = stuff (value)
> except KeyError:
>     foo = otherStuff (key)
>

try:
    value = d[key]
except KeyError:
    foo = otherStuff (key)
else:
    foo = stuff (value)

OR:

if value.has_key(key):
    foo = stuff (value)
else:
    foo = otherStuff (key)





From zayats at blue.seas.upenn.edu  Mon Apr  1 17:10:03 2002
From: zayats at blue.seas.upenn.edu (Salim Zayat)
Date: 1 Apr 2002 22:10:03 GMT
Subject: Arabic bidirectionality in Tkinter
Message-ID: <a8alrr$sa$1@netnews.upenn.edu>

Hey all.  I have a question about Tkinter and internationalization.  I 
have a unicode string:

s = u'\u0631\u062f\u062f\u0627\u063a'

When I send this to a canvas to be drawn as text:

canvas.create_text(x, y, text=s)

I get all the letters treated as individual characters, instead of 
changing their presentation in comparison to the characters around it 
(glyphs I believe theyre called).

Qt deals with this bidirectionality automatically, but Tkinter doesnt 
appear to.  What can I do?

Thanks for any help!

Salim


From timr at probo.com  Thu Apr 25 23:39:21 2002
From: timr at probo.com (Tim Roberts)
Date: Thu, 25 Apr 2002 20:39:21 -0700
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net> <0kiccu46a2580bbohtpavbla241q95etj0@4ax.com> <1nmccuogt7t9ler1eqq4kec95bhr0mcchr@4ax.com>
Message-ID: <ptihcukfjtv1r7jrnhc9uil9f0aqiocdi0@4ax.com>

Fernando Rodr?guez <frr at easyjob.net> wrote:

>On Tue, 23 Apr 2002 23:05:59 -0700, Tim Roberts <timr at probo.com> wrote:
>
>>Theoretically, it would be possible to have the <FORM> fire a URL on the
>>home system with the order processing URL, and have that script generate
>>its own HTTPS connection to the order processing site, but that means the
>>web site user will not see that his connection is secure. 
>
>Why? If everything is done in a SSL connection...

You are correct.  As long as your <form> is fired under https, there would
be no such problem.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From jeremy at alum.mit.edu  Mon Apr 29 11:51:34 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 29 Apr 2002 08:51:34 -0700
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk> <just-9E8624.12281026042002@news1.xs4all.nl> <ucjicvlrf7djd3@news.supernews.com>
Message-ID: <b0f083db.0204290751.52c145f4@posting.google.com>

"John Roth" <johnroth at ameritech.net> wrote in message news:<ucjicvlrf7djd3 at news.supernews.com>...
> The discussion in the bug report indicates that the 'x'
> following the '=' should be compiled using the 'LOAD_NAME'
> op code, which implements the old rules. If it in fact did this,
> I would expect the first example to return 4, not 2,
> because 'LOAD_NAME' is supposed to ignore the
> parameter.

(I hope this answer is clear without quoting the original code
example.)

LOAD_NAME is an old-fashioned beast that looks in the current
namespace, the global namespace, and the builtin namespace.  It never
raises an exception unless the name can't be found in any of those
namespaces. Note in particular that it's always three namespaces; in
the case of a nested block, enclosing namespaces are not searched.

Since LOAD_NAME is used for class namespaces, you'll never see an
unbound local error.  If a variable 'x' is a local but has not been
bound yet, LOAD_NAME will next search the globals and builtins.  If
the local happens to shadow a global or builtin, you'll get its value.
 Otherwise, you'll get a NameError.
 
> Clearly, it isn't. At a guess, I'd think the problem is that
> the compiler is trying to make 'z' a local variable of the
> def, so it's using a different op code in the first case than
> in the second.

No.  The compiler never "tries to make" a local variable bound in one
block into a local variable in a different block.

This stuff isn't documented, so it's no surprise that it's confusing. 
But I think the documentation will mostly be of the form:  The results
of XXX are undefined.  (The CPython implements currently does this
...)

Jeremy


From merman at freenet.de  Mon Apr 15 08:44:29 2002
From: merman at freenet.de (merman)
Date: Mon, 15 Apr 2002 14:44:29 +0200
Subject: newbie-question: extending python with c++
Message-ID: <3cbac8f9$0$12315$9b622d9e@news.freenet.de>

Hi there,

I have a c++ -class; I can compile this class successful as standalone-app.
No problem.
Now I want to work with swig.

There are three files: Tree.cpp, Tree.h, Tree.i

first step:
swig -python -c++ -o Tree_wrap.cpp Tree.i (output is Tree_wrap.cpp - it's
ok)

second step:
g++ -c Tree.cpp Tree_wrap.cpp -IC:\python22\include (also ok - output are
Tree.o and Tree_wrap.o)

third step - *build*:
g++ Tree.o Tree_wrap.o -shared -o Tree.dll (this line throws lots of
failures - whats wrong?)

failure-examples:

Tree_wrap.o(.text+0x311):Tree_wrap.cpp: undefined reference to
`_imp__PyString_FromString'
Tree_wrap.o(.text+0x441):Tree_wrap.cpp: undefined reference to
`_imp__PyExc_NameError'
Tree_wrap.o(.text+0x44a):Tree_wrap.cpp: undefined reference to
`_imp__PyErr_SetString'
Tree_wrap.o(.text+0x4c1):Tree_wrap.cpp: undefined reference to
`_imp__PyExc_NameError'
Tree_wrap.o(.text+0x4ca):Tree_wrap.cpp: undefined reference to
`_imp__PyErr_SetString'

... and so on. Is there a llinker-problem?

I work on win2k with MINGW (gcc for win32). My python-ver. is python2.2.1c1.

Can anybody help?

regards thomas






From corbet-python at lwn.net  Fri Apr 26 11:52:37 2002
From: corbet-python at lwn.net (Jonathan Corbet)
Date: Fri, 26 Apr 2002 09:52:37 -0600
Subject: A SYSV IPC module?
Message-ID: <20020426155237.4009.qmail@eklektix.com>

I have a task which, it appears, requires hooking up to SYSV shared memory
segments and semaphores.  Surprisingly, there appears to be no Python
module which exports these system calls.  Surely somebody must have had to
solve this problem before...does anybody have a pointer to where I could
find such module?

Thanks,

jon

Jonathan Corbet
Executive editor, LWN.net
corbet at lwn.net




From ffjhenon at club-internet.fr  Sun Apr 21 18:50:31 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Mon, 22 Apr 2002 00:50:31 +0200
Subject: console redirection to Tkinter
Message-ID: <3CC34237.CE6F21F1@club-internet.fr>

I want to use Tkinter to launch an external raytracer(POVRAY).
Like Python, it reads a script,
it spits out an image in a box and displays the line number which
is calculated to the console.
When the raytrace is over(after 1 second or 1 day), it outputs the stats
to the console as well.
I would like both outputs (line number and stats) to be redirected
to my Tkinter application.(That's where I get stuck)
When I uncomment the outr=pout.rea.....my app gets stuck.
The t.insert('1.0',pout) and t.insert('end',perr) do not give me what I
expect.


What shall I do ?
Thanks for any advice.


------------------cut here--------------------------
#!/usr/bin/env python
from Tkinter import *
import popen2

def lance():
  cmd= "x-povray +i/mnt/win_c/Utile/pyvon/7.pov +w320 +h240 -f +dgt +p"
  run = popen2.Popen3(cmd)
  pout=sys.stdout
  perr=sys.stderr
#  outr=pout.readlines() # when this
  t.insert('1.0',"---out----")
  t.insert('1.0',pout)
  t.insert('end',"---err----")
  t.insert('end',perr)

mframe=Frame()
b=Button(mframe,text="Raytrace !",command=lance)
mframe.pack()
b.pack()
t=Text(mframe,width=20)
t.pack()
b.mainloop()
------------------cut here--------------------------


From ark at research.att.com  Wed Apr 10 17:30:08 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 21:30:08 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <1018450381.103820@cswreg.cos.agilent.com> <yu991ydnmtzc.fsf@europa.research.att.com> <a927vf$euh$0@216.39.172.122>
Message-ID: <yu99u1qjfdi7.fsf@europa.research.att.com>

Bengt> What about x() as the spelling of get-x's-value and x(v) as the
Bengt> spelling of set-x's-value-to-v-and-return-v?

That still assumes that the value of x was set previously.

If I make that assumption, there are lots of ways of solving the
problem -- the first one I suggested was simply to set x to [].

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From amapy at snafu.de  Sat Apr 13 12:39:46 2002
From: amapy at snafu.de (Andreas Maurer)
Date: Sat, 13 Apr 2002 18:39:46 +0200
Subject: Some troubles with COMObject of the ClearCaseAutomation Library 4.1 (ccauto.dll)
Message-ID: <3CB85F52.40000@snafu.de>

Hi all,

I've written a little script using the ClearCase Automation Library 4.1
with python 2.1.2

When I'm trying to get the collection of HistoryRecords of one
CC-element this error message appears:
------------------

 >>> hist = Elem.HistoryRecords(0)

Traceback (most recent call last):
   File "C:\win32app\Python21\lib\site-packages\win32com\client\dynamic.py", line 279, in _make_method_
codeObject = compile(methodCode, "<COMObject %s>" % self._username_,"exec")
   File "<string>", line 1
      def HistoryRecords(self, pICCBranchType=pythoncom.Missing, Since=<time object at 013BB240>, User='', Minor=0, ExcludeCheckOutEvents=0, Recurse=0, DirectoryOnly=0):
                                                                       ^
  SyntaxError: invalid syntax
Traceback (most recent call last):
   File "<interactive input>", line 1, in ?
TypeError: object of type 'None' is not callable
--------------
It seems, there is some trouble with the time object.

The IDL-Syntax for the HistoryRecords Method looks like this:
--------------------------------
         [id(0x00000005), propget, helpstring("Gets the collection of history records for this object"), helpcontext(0x00000c85)]
         HRESULT HistoryRecords(
                         [in] ICCBranchType* pICCBranchType,
                         [in, optional, defaultvalue(00:00:00)] DATE Since,
                         [in, optional, defaultvalue("")] BSTR User,
                         [in, optional, defaultvalue(0)] VARIANT_BOOL Minor,
                         [in, optional, defaultvalue(0)] VARIANT_BOOL ExcludeCheckOutEvents,
                         [in, optional, defaultvalue(0)] VARIANT_BOOL Recurse,
                         [in, optional, defaultvalue(0)] VARIANT_BOOL DirectoryOnly,
                         [out, retval] ICCHistoryRecords** pHistoryRecords);
         [id(0x00000006), propget, helpstring("Gets the collection of hyperlinks associated with this VOB object"), helpcontext(0x00000c86)]
--------------

I even tried a static dispatch, but the result was not that what I expected.

If you need more information, please let me know

Many Thanks

Andreas



From bokr at oz.net  Mon Apr  1 16:52:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 21:52:17 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <3CA62CD8.6ABD44D0@accessforall.nl> <m3adsp8zie.fsf@mira.informatik.hu-berlin.de> <3CA6F012.E2631003@accessforall.nl> <m3adspm1kl.fsf@mira.informatik.hu-berlin.de> <3CA79401.CFA41B65@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain>
Message-ID: <a8akqh$8bk$0@216.39.172.122>

On Mon, 01 Apr 2002 18:33:14 GMT, kzaragoza at attbi.com (Kris J. Zaragoza) wrote:

>In article <3CA79401.CFA41B65 at python.org>, Guido van Rossum wrote:
>> But see also my comments on these two in
>> http://www.python.org/doc/essays/pepparade.html; neither PEP is likely to
>> have any effect on Python's future.
>> 
>> --Guido van Rossum (home page: http://www.python.org/~guido/)
>
>Guido,
>
>I have to ask one question that I haven't seen brought up yet: What
>does PEP 285 imply for the future of Python?  Do you intend to
>eventually change the concept of what is true?
>
>Specifically, I refer to the Python Language Reference
>(http://www.python.org/doc/2.2/ref/lambda.html):
>
>"""In the context of Boolean operations, and also when expressions are
>used by control flow statements, the following values are interpreted
>as false: None, numeric zero of all types, empty sequences (strings,
>tuples and lists), and empty mappings (dictionaries). All other values
>are interpreted as true."""
>
>Do you intend to eventually (NOTE: I say eventually, I know it's not
>immediate) modify this to read that only the value True is true and
>the value False is false?  If this is the case, I believe this adds
>needless complexity to code whereas (in a simple case)
>
>if foo:
>    ...do whatever...
>
>would now have to read
>
>if bool(foo):
>    ...do whatever...
>
Perhaps this could be discussed in terms of hidden properties. I.e.,

    if foo:
    ... whatever

results is accessing foo's hidden boolean property because of the
'if'-context.

    bar = foo

binds bar to the foo object, not the value of the hidden property
this time, because of the '='-context. Some people get confused when
both types of access are used in the same expression, as in

    baz = foo or bar

Here, first foo's hidden boolean property is tested. It that is true,
then the value is foo as in '='-context. Similarly for bar.

Of course the 'hidden property' is only rhetoric here, but why not
make this a clear and controllable unified mechanism instead of a
collection of special effects? I would like to see an overridable
__bool__ method (not actually a programmatic property) for the base
object class, perhaps making deprecation of __nonzero__ possible.

Anyway, x.__bool__() would be the standard access mechanism for the
'hidden boolean property' of x, and plain x as usual in '='-contexts.

>in order to convert None, empty sequences/mappings, and non-integer
>zeros to boolean values.  It may seem like a small change, but it
>would add unnecessary verbiage to what is already fairly obvious in
>control constructs and other boolean expressions.
No verbiage would be necessary if the base classes had appropriate
__bool__ methods, and appropriate implicit conversions were done
for arithmetic contexts (e.g. indexing operators etc.)
>
>If you don't intend to modify this concept, then what's the real gain
>for the new boolean type?  The concept is already fairly clear, and
>would be needlessly muddied by this extra type that holds one of two
>values.  (Try explaining to a newbie that it's actually one of two
>pre-allocated instances!)
The benefit would be that you could derive from any object-based class
and customize your class's definition of instances' boolean values
in a standard way, as you saw fit. IOW, each class gets to modify the
concept for its own purposes.

bool(x) would call x.__bool__(), which would return a bool. If that were
done right, it should also be possible to return a subtype of bool that
had different str/repr so that print bool(x) might say, e.g.,
On/Off instead of True/False, and yet have subtype instance work
properly as a logic value.

>
>The only real advantage I can see in the latter case is to make some
>code a bit more self-documenting.  If you assign a value of False or
>True to a variable, it's pretty clear what that variable is intended
>to hold.  However, given the firestorm of discussion that has erupted
>around the details of this PEP, is this small gain really worth the
>effort?  Is there some other advantage I'm missing?
Well, what do you think of the above?

Regards,
Bengt Richter


From bokr at oz.net  Thu Apr 11 09:34:30 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Apr 2002 13:34:30 GMT
Subject: decision tables Re: Another stab at a "switch/case" construct (new  proposal):
References: <mailman.1017466547.3800.python-list@python.org>
Message-ID: <a943d6$3ad$0@216.39.172.122>

On Sat, 30 Mar 2002 08:33:13 +0300 (MSK), Roman Suzi <rnd at onego.ru> wrote:

>On Fri, 29 Mar 2002, Chris Gonnerman wrote:
>
>>----- Original Message -----
>>From: "Cliff Wells" <logiplexsoftware at earthlink.net>
>>>
>>> Now which is cleaner?   I think the perceived need for this type of
>>> construct is a holdover from thinking in other programming languages
>>versus
>>> thinking in Python.
>
>if-elif-else is a complete functional substitute for any kind
>of switch/case. I can't see why this topic is still discussed.
>Python has dictionaries, so many simple things could be
>made by them. More complex decision making pieces of code
>could be elegantly done with if's.
>
>*
>
>However, I'd liked to see a good paradigm for expressing
>decision tables like this one:
>
>F1 F2 F3 R
>0  0  *  do1
>0  1  0  do2
>0  1  1  do3
>1  *  *  do4
>
>in Python. (With lazy evaluation, of course).
>
>Any ideas?
>

See if this does what you want (what did you want to do with it?)

The DT class constructor will read your decision table just as is above, and make
an instance that has a decision tree in it. The test prints out the table and tree
and runs all the combinations of inputs, but you don't have to do all that of course.

The test code for the following is (snipped from below):
    # first test
    if testno==1:
        print '\n--< Test 1, with complete table and list of functions >--'
        print dtable
        dt = DT(dtable, dofuns(4), on_error=errf)
        dt.ptree()
        for i,j,k in [(i,j,k) for i in 0,1 for j in 0,1 for k in 0,1]:
            print dt(i,j,k)('args:',i,j,k)

dtable is your table with triple quotes around it, otherwise verbatim. It ignores blank
lines before and after, and white space left and right, so you can lay it out prettily ;-)
dofuns(4) generates a list of four functors to use ad do1-do4, but
DT accepts keyword specification of functions overriding or in place of the function list.
The keys must match the right column in the dtable if used (e.g., do2=foo2) and an error function
is optional via keyword on_error=whatever.
The print statement shows how the parameters are passed to the decoder like dt(i,j,k) and
the arguments passed can be anything, since they are captured as *args, **kwds and passed
through to the selected function. ('args:',i,j,k) was just to print something recognizable
using returns from the test functions.
I assumed '*' meant "don't care" in the table. I think it will generate a tree if you
use other than trailing '*'s, but the tree won't be as efficient space wise. It should
be the same speed though. The latter (heck, all) is largely untested, so no doubt there
are bugs somewhere.

A test (#1) with a complete table makes this test result (only tested on NT4 Python 2.2):

[ 6:28] C:\pywk\dtable>python dtable.py 1

--< Test 1, with complete table and list of functions >--

    F1 F2 F3 R
    0  0  *  do1
    0  1  0  do2
    0  1  1  do3
    1  *  *  do4

[
    [
        <dofun instance do_1>,
        [
            <dofun instance do_2>,
            <dofun instance do_3>,
        ],
    ],
    <dofun instance do_4>,
]

<dofun instance do_1> was called with (('args:', 0, 0, 0),{})
<dofun instance do_1> was called with (('args:', 0, 0, 1),{})
<dofun instance do_2> was called with (('args:', 0, 1, 0),{})
<dofun instance do_3> was called with (('args:', 0, 1, 1),{})
<dofun instance do_4> was called with (('args:', 1, 0, 0),{})
<dofun instance do_4> was called with (('args:', 1, 0, 1),{})
<dofun instance do_4> was called with (('args:', 1, 1, 0),{})
<dofun instance do_4> was called with (('args:', 1, 1, 1),{})

===< dtable.py >===================================================================
#!/usr/bin/python
# dtable.py (C) 2002 by Bengt Richter. All rights reserved and transferred to PSF
# This is alpha code. NO WARRANTY OR GUARANTEE OF ANY KIND
# The DT class constructor takes a decision table and a set of functions and defines
# an instance with a call method for decoding per the table and returning the
# appropriate function. The instance contains a decision tree generated from the
# decision table. This file can be run from the command line to execute one of
# four tests (1-4) thus: python dtable testno
# see test code at end for usage patterns. Sorry about doc shortage. More possible
# if any interest. You can reach me at bokr at oz.net.
#
class DT(object):
    def tree_upd(subtree, conds, fun):
        # subtree can be:
        # [] empty, which can receive a terminal [fun] or a choice [x,y] or
        # a don't-care-but-do-care-further-on where x and y are both []
        # 
        # [fun] -- already a terminal, anything more is multiple definition
        #
        # [x,y] -- a '*' should update both x and y subtrees
        condc = conds and conds[0] or '<error if used>'
        if conds.count('*') == len(conds):
            if subtree == []: subtree.append(fun)
            else: raise ValueError, '*-tail conflicts for %s: %s' % (fun.__name__,conds)
        elif condc == '*':
            # current is don't care, but do care later
            # split current node if not already split
            if subtree == []: subtree.extend([[],[]]) #operate destructively, don't rebind!
            if not isinstance(subtree[0], list) or not isinstance(subtree[1], list):
                raise ValueError,'Conflicting "*" in tail "%s" for %s' % (cond, fun.__name__)
            DT.tree_upd(subtree[0],conds[1:],fun)
            DT.tree_upd(subtree[1],conds[1:],fun)

        elif condc == '0':
            if subtree == []: subtree.extend([[],[]]) # make sure it's split
            if not isinstance(subtree[0], list):
                raise ValueError,'Conflicting condition tail "%s" for %s' % (cond, fun.__name__)
            DT.tree_upd(subtree[0],conds[1:],fun) # update subtree

        elif condc == '1':
            if subtree == []: subtree.extend([[],[]]) # make sure it's split
            if not isinstance(subtree[1], list):
                raise ValueError,'Conflicting condition tail "%s" for %s' % (cond, fun.__name__)
            DT.tree_upd(subtree[1],conds[1:],fun) # update subtree

        else:
            raise ValueError,'Bad condition code "%s" for %s' % (condc, fun.__name__)
    tree_upd = staticmethod(tree_upd)

    def tree_clean(t):
        if not isinstance(t, list): return t
        for i in xrange(len(t)):
            t[i] = DT.tree_clean(t[i])
        if len(t)==1 and not isinstance(t[0], list): return t[0]
        return t
    tree_clean = staticmethod(tree_clean)

    def _ptree(t,ind):
        sind = ' '*4*ind
        if isinstance(t, list):
            print '%s[' % sind
            for st in t:
                DT._ptree(st,ind+1)
            print '%s]%s' % (sind, ind and ',' or '')
        else:
            print '%s%s,' % (sind,t)
    _ptree = staticmethod(_ptree)

    def defaultErrf(*args, **kwds):
        return 'defaultErrf: no function specified for:', args, kwds
    defaultErrf = staticmethod(defaultErrf)

    def __init__(self, dt, *flist,**fdir):
        if flist and isinstance(flist[0], list):
            flist = flist[0] # flist tuple contains a single list of args instead of arg list

        # allow pretty table formatting by stripping leading/trailing blank lines and spaces
        dtlines = [ s.strip() for s in dt.split('\n') if s and not s.isspace()]

        self.kwk = dtlines[0].split()[:-1]    # for kw=>arg list translation overriding flist
        self.dtree = []                       # dtree[0] is the root
                                              # * * * foo => dtree == [foo]
                                              # 1 * * bar => dtree == [[],bar]
                                              # tree is cleaned up after building is completed
        nlines = len(dtlines)
        for fx in xrange(1,nlines):           # walk table by lines
            line = dtlines[fx]
            ls = line.split()
            ls, fname = ls[:-1], ls[-1]       # cond list, function name
            #print '-----------------\nfname:',fname,fdir,flist,locals(),globals()
            fun = fdir and fdir.get(fname) or flist and flist[fx-1] or DT.defaultErrf

            # add to tree as necessary for new choices
            # call recursively descending tree update
            DT.tree_upd(self.dtree, ls, fun)
        # set optional on_error function to call with (args,kwargs) like others
        self.on_error = (
            fdir and fdir.get('on_error') or
            flist and len(flist)>nlines and flist[nlines]
        ) # optional on_error function
        self.dtree = DT.tree_clean(self.dtree) # change [fun] to fun at leaves

    # define dispatch based on decision tree and input, returning selected function
    def __call__(self, *args, **kw):
        if kw:
            args = [kw.get(k,0) for k in self.kwk] # default 0
        else:
            args = list(args)
        t = self.dtree
        while len(args)<3: args.append(0)
        for fx in args:
            # choose between t[1] and t[2] per fx
#           print 't => ',t
            try:
                if not isinstance(t, list): return t
                t = t[fx]
                if t == []: raise IndexError
            except IndexError:
                if self.on_error: return self.on_error
                raise ValueError,'Decision table does not define %s' % args
        return t

    def ptree(self):
        DT._ptree(self.dtree, 0)
        print

##########################
##        tests         ##
##########################
def test(testno=0):
    # test example
    dtable="""
    F1 F2 F3 R
    0  0  *  do1
    0  1  0  do2
    0  1  1  do3
    1  *  *  do4
    """

    # the above dtable should lead to this decision tree
    """
    tree =
    [                                   # * * * 
        [                               # 0 * *
            <dofun instance do_1>,      # 0 0 *
            [                           # 0 1 *
                <dofun instance do_2>,  # 0 1 0
                <dofun instance do_3>,  # 0 1 1
            ],
        ],
        <dofun instance do_4>,          # 1 * *
    ]

    # it is traversed with args list [F1,F2,F2] of integer 0's & 1's
    t = tree
    for fx in args:
        ... etc., see __call__ method above
    """

    # test functions
    def foo1(*args): return 'foo1', args
    def foo2(*args): return 'foo2', args
    def foo3(*args): return 'foo3', args
    def foo4(*args): return 'foo4', args
    def errf(*args): return 'errf', args

    # test functions as functors in list
    def dofuns(n):
        class dox:
            def __init__(self,v): self.v = v
            def __repr__(self): return '<dofun instance %s>' % self.v
            def __call__(self,*args,**kw):
                return '<dofun instance %s> was called with (%s,%s)' % (self.v, args, kw)
        return [dox('do_%d' % (i+1,)) for i in xrange(n)]

    # first test
    if testno==1:
        print '\n--< Test 1, with complete table and list of functions >--'
        print dtable
        dt = DT(dtable, dofuns(4), on_error=errf)
        dt.ptree()
        for i,j,k in [(i,j,k) for i in 0,1 for j in 0,1 for k in 0,1]:
            print dt(i,j,k)('args:',i,j,k)
    else:
        # second etc tests with 0 1 0 undefined
        dtable2 = """
        F1 F2 F3 R
        0  0  *  do1
        0  1  1  do3
        1  *  *  do4
        """
        if testno==2:
            print '\n--< Test 2, table missing 0 1 0 do2, functions in list >--'
            print '\n',dtable2
            dt = DT(dtable2, dofuns(4), on_error=errf)
            dt.ptree()
            for i,j,k in [(i,j,k) for i in 0,1 for j in 0,1 for k in 0,1]:
                print dt(i,j,k)('args:',i,j,k)

        # third test, with 0 1 0 undefined
        elif testno==3:
            print '\n--< Test 3, table missing 0 1 0 do2, functions kw specified >--'
            print '\n',dtable2
            dt = DT(dtable2, do1=foo1,do3=foo3,do4=foo4,on_error=errf) # get functions named as in table (do1,do3,do4) per kwds
            dt.ptree()
            for i,j,k in [(i,j,k) for i in 0,1 for j in 0,1 for k in 0,1]:
                print dt(i,j,k)('args:',i,j,k)

        # fourth test, with 0 1 0 undefined and do1 function missing
        elif testno==4:
            print '\n--< Test 4, table missing 0 1 0 do2, functions kw specified except do1 >--'
            print '\n',dtable2
            dt = DT(dtable2, do3=foo3,do4=foo4,on_error=errf) # get functions named as in table (do1,do3,do4) per kwds
            dt.ptree()
            for i,j,k in [(i,j,k) for i in 0,1 for j in 0,1 for k in 0,1]:
                print dt(i,j,k)('args:',i,j,k)
        else:
            print 'Undefined test number: %d' % testno

if __name__ == '__main__':
    import sys
    try:
        n = int(sys.argv[1])
        test(n) # do it/them
    except:
        print "Usage: python dtable.py testno\n    where testno can be 1,2,3,4"

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

Anyway, this sat as a saved reply with nothing in it for a long time. I finally
thought to get it out, in case you or someone was interested. I just cut and pasted
to make a few different tests, as can be seen ;-)

Regards,
Bengt Richter


From christophe.delord at free.fr  Thu Apr 11 05:09:22 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 11 Apr 2002 11:09:22 +0200
Subject: Pattern for permutations?
References: <mailman.1018131433.4336.python-list@python.org>
Message-ID: <3CB552C2.3050603@free.fr>

Hello,

I have a simple solution using generators (python 2.2).
(think recursive!)


from __future__ import generators

def prod(*seqs):
	if seqs:
		for x in seqs[0]:
			for xs in prod(*seqs[1:]):
				yield (x,)+xs
	else:
		yield ()

for x,y,z in prod((1,2,3),('a','b','c'),(4,5,6)):
	print x,y,z



Hope it helps!

Christophe.


Tim Churches wrote:

> This may be an elementary question but the solution is
> not obvious to me.
> 
> I want to define a function which takes as its argument a
> list which contains one or more sequences, and then does stuff
> with every permutation of the elements of those sequences. For
> example, a function which works for a list containing three
> sequences:
> 
> def nestedloops(list_of_sequences):
> 	for element0 in list_of_sequences[0]:
> 		for element1 in list_of_sequences[1]:
> 			for element2 in list_of_sequences[2]:
> 				# do stuff with each permutation
> 				print element0, element1, element2
> 
> nestedloops([(1,2,3),('one','two','three'),('isa','dalawa','tatlo')])
> 
> How does one generalise such a function so that it handles any number 
> of sequences in the passed list of sequences i.e. a dynamic number of
> levels of nesting? 
> 
> It occurred to me that the function could emit dynamic code for the 
> "for" loops and then eval() that, but that seems a bit messy and
> possibly slow.
> 
> Is there a better way?
> 
> Tim C
> 
> 
> 


-- 
Christophe Delord
http://christophe.delord.free.fr/



From roy at panix.com  Sat Apr  6 15:27:34 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 06 Apr 2002 15:27:34 -0500
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <200204061913.g36JDcdj017764@ratthing-b246.strakt.com> <mailman.1018122494.10925.python-list@python.org>
Message-ID: <roy-9E63B6.15273306042002@news1.panix.com>

Paul Sidorsky <paulsid at shaw.ca> wrote:
> The proper course of action here would be to add a second check,
> isPaid(), to wherever the new information is needed.

Or, instead of isValid() and isPaid(), have a single method, getStatus(), 
which could return invalidAccount, paymentOverdue, or accountUpToDate.  Not 
all things fit into a boolean, but those that don't shouldn't be accessed 
by methods whose names begin with "is".  I guess the die-hard lispers would 
say the same about names ending in "p" :-)

The most fascinating (and often, frustrating) example I know of "almost 
boolean" is SQL's concept of NULL.  In what other language can "A == B" and 
"A != B" both be false?  The problem is that they tried to cram two 
distinct concepts ("are the operands valid?" and "are the operands equal?") 
into a single conceptual slot.


From max at alcyone.com  Sun Apr 28 18:07:49 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 28 Apr 2002 15:07:49 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com> <3CC449BF.8030807@alcyone.com> <pan.2002.04.28.00.32.39.955778.3624@uol.com.br>
Message-ID: <3CCC72B4.1DA12B8C@alcyone.com>

synthespian wrote:

>         You go now and post that opinion on comp.lang.lisp and you're
> going to
> get beaten up till you drop.
>         For one thing LISP is a huge language, with various
> programming
> paradigms, and Scheme insists on being small and functional - what the
> hell for, nobody at comp.lang.lisp seems to know.

That's Common Lisp.  "Lisp" is a word that can either mean Common Lisp,
or any other Lisp-like language, or the general family of similar
languages, which includes Scheme; Common Lisp is the name of the "huge
language with various programming paradigms"; "Lisp" is ambiguous.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From mats at laplaza.nospam.org  Fri Apr 26 09:34:23 2002
From: mats at laplaza.nospam.org (Mats Wichmann)
Date: Fri, 26 Apr 2002 13:34:23 GMT
Subject: configuring mxCGIPython
References: <NCBBLFCOHHDIKCAFGCFBEEJEKMAA.marklists@mceahern.com> <mailman.1019766529.8485.python-list@python.org>
Message-ID: <3cc956c3.90050515@news.laplaza.org>

On Thu, 25 Apr 2002 22:28:57 +0200, "M.-A. Lemburg" <mal at lemburg.com>
wrote:

:Mark McEahern wrote:

:>   ./cgipython: error in loading shared libraries: libxmlparse.so.1: cannot
:> open shared object file: No such file or directory
:> 
:> I'm not sure yet what that means yet, but I think it means I may be able to
:> set LD_LIBRARY_PATH, if: 1) that file resides on the server; and 2) I know
:> where it is.
:
:Right.
: 
:> Is there a way I can tell cgipython not to bother loading that particular
:> shared object?
:
:No. Could be that libxmlparse is referenced by pyexpat which is part
:of cgipython. On most systems the expat libs should be part
:of the system, so setting LD_LIBRARY_PATH to e.g. /usr/lib
:should help. If not, you can try getting that file from some
:other source and install it in the same dir as cgipython.

Not helping to solve the problem at hand, but FYI:

The dynamic linker should always be looking in its "trusted"
directories, /lib and /usr/lib - so setting LD_LIBRARY_PATH to contain
one of these is not useful.
Mats Wichmann



From rasmussn at lanl.gov  Tue Apr  2 19:24:51 2002
From: rasmussn at lanl.gov (rasmussn at lanl.gov)
Date: Tue, 2 Apr 2002 17:24:51 -0700
Subject: [Newbie] embed c++ in Python
In-Reply-To: <3ca95375$0$3698$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <36D98F4A-4699-11D6-AC67-003065D61546@lanl.gov>

On Monday, April 1, 2002, at 11:47 PM, Matthias Stern wrote:

>>> I want to know how to write a function which receives some variables as
>>> argument changes them and sends the results back to Python.
>
> Thank you for the links I got (http://www.scipy.org/site_content/weave
>  and http://www.acl.lanl.gov/siloon/ ). I've visited the sites and
> discovered a new problem: Both solutions require some additional
> stuff to be installed. Maybe it's much simpler to use them but it
> is quite hard to add some additional stuff like this to our project. It's
> a cross-platform project with many operating systems and we want as
> little additional programs as possible.

You really only have to install additional packages to create the Python
extensions (for Siloon at least and I assume for weave).  This is normally
a one time event (unless you plan on changing you API frequently).  Then
you can package up the Siloon generated code (along with maybe a few Siloon
header files) in your project and be done with it.

Unless you want everyone on the project to be able to generate the 
extensions.
Then you would need to distribute the additional packages.

Regards,
Craig




From boud at valdyas.org  Tue Apr 23 15:57:37 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 23 Apr 2002 18:57:37 -0100
Subject: Python, PyQt make install problems on IRIX
References: <f0fd5987.0204220106.7673e345@posting.google.com> <3cc45c8d$0$307$e4fe514c@dreader1.news.xs4all.nl> <f0fd5987.0204230057.25fdc747@posting.google.com> <aa3r2a$tiabi$1@fido.engr.sgi.com>
Message-ID: <3cc59530$0$41709$e4fe514c@dreader4.news.xs4all.nl>

offer at sgi.com wrote:

> $ from gtaylor at lowebroadway.com -#196241 | sed "1,$s/^/> /"
>>
>>
>>Irix certainly is'nt all that common, in a way I think that kind of
>>helps people make installers, because there are not many hardware
>>configurations to support or variations of the OS, like Linux, there
>>are loads of variations on the same thing.
> 
> Hey, Irix is pretty common here :-)

I would think so :-).

> 
> Anyway its my intention to ship PyQt in the next release of Irix Freeware
> (Aug), so I'll keep an eye out for this problem (I have no idea what's
> causing it).
> 

That's very cool!
-- 
Boudewijn Rempt | http://www.valdyas.org


From bokr at oz.net  Mon Apr  1 15:21:42 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 20:21:42 GMT
Subject: Debugging flags [was: Re: PEP 285: Adding a bool type]
References: <mailman.1017466974.6751.python-list@python.org> <7xhemy1jo4.fsf@ruckus.brouhaha.com> <slrnaac6kf.2gs.philh@comuno.freeserve.co.uk> <slrnaacj3a.2gs.philh@comuno.freeserve.co.uk> <lXYp8.108382$Ou.84158@atlpnn01.usenetserver.com>
Message-ID: <a8afgm$o81$0@216.39.172.122>

On Mon, 1 Apr 2002 08:05:24 -0500, "Steve Holden" <sholden at holdenweb.com> wrote:

>"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
>[...]
>> I've gone through some of my herbivore code, looking to see where
>> I've used 0 or 1 as a boolean. Example 1:
>>
>> debug = 0  # debugging this module?
>>
>> Frankly, if that is unclear to anyone, they are not cut out to be a
>> programmer. Bear in mind that most non-programmers are aware of the
>> 0/1 paradigm for boolean values: it appears on much electrical
>> equipment.
>>
>I find it much more helpful to use
>
>    debug = (1, 2, 7, 93) # debugging parsing, tree build, code generation
>and error messages
>
>with corresponding tests like
>
>    if 7 in debug:
>        ...
I wonder if that would be faster as
     if debug and 7 in debug

>
>Given that when I'm not debugging the list is empty, these tests are not
>unduly time-consuming. But then I've often gone over the top to allow
>selective testing in complex developments.
>
I haven't had occasion to do it in Python, but I think magic letters
are easier to remember than magic numbers ;-)

debug = raw_input('Enter a string of characters for tests to enable: ')
...
if debug and 'x' in debug:
   # do x-test
...
etc.

Hm, for full mnemonics, you could make debug a list of words:

debug = raw_input('Enter names of tests to enable (separated by spaces): ').split()
...
if debug and 'foo-test' in debug:
...
etc.

I've used letters to do it in C, a little differently.
It can be useful, especially to avoid recompiling. And you
can do it so you can recompile with all the tests excluded, by
changing just one line, and without using #if/#endif clutter.

(Untested reconstruction from memory, XXX for concise example ;-)

#define IF_DEBUG_ENABLED(x) if(debug & (1<<(ord(x)-ord('a')))
// to compile out debug code (make it dead code), use the next line instead
// #define IF_DEBUG_ENABLED(x) if(0)
long debug = 0;
for(char *p = argv[1]; *p; ++p){ // XXX assuming available on cmd line
   debug |= (1<<(ord(*p)-ord('a'))); // XXX no warning if unhandled chars
}
...
IF_DEBUG_ENABLED('x'){	// test x enabled?
...
}
etc.

Regards,
Bengt Richter


From aahz at pythoncraft.com  Tue Apr 23 14:21:46 2002
From: aahz at pythoncraft.com (Aahz)
Date: 23 Apr 2002 14:21:46 -0400
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org>
Message-ID: <aa48nq$a61$1@panix1.panix.com>

In article <mailman.1019552082.8541.python-list at python.org>,
Graham Ashton  <gashton at cmedltd.com> wrote:
>
>I find that the behaviour varies depending on the order in which a
>class's super classes are defined. In the example below, I'd like the
>instantiation of class B to throw the same RuntimeError as A:
>
>Python 2.2 (#2, Mar 11 2002, 13:24:00) 
>[GCC 2.95.3 20010315 (release)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> from AbstractSingleton import AbstractSingleton
>>>> class A(AbstractSingleton, object):
>...   pass

Maybe you should tell us a bit more about what you're trying to do so
that you needn't rely on crufty internal stuff to get the correct result.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From ods at fep.ru  Mon Apr 15 05:59:27 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Mon, 15 Apr 2002 13:59:27 +0400 (MSD)
Subject: Deprecate input()
In-Reply-To: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <Pine.LNX.4.44L.0204151352410.13708-100000@ods.pp.ru>

On Sun, 14 Apr 2002, Raymond Hettinger wrote:

RH> The help and tutor maillists indicate that the use of
RH> input() rather than
RH> raw_input() is widespread.  I think this function is so
RH> harmful and so
RH> likely to be used that it should be deprecated hasta pronto
RH> (right away).
RH> When really needed, eval(raw_input()) is easy enough to code
RH> and explicit
RH> about what it does.
RH>
RH> <no wink -- this function needs to go away>

Certanly.  I was wondered to see input() documented as function
that accept only integer input (but not evaluate it!) in Russian
translation of Andre S. Lessa's "Developer's Handbook" (I do not
know if this bug is in original book too).





From aleax at aleax.it  Tue Apr 30 10:18:03 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 14:18:03 GMT
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <199334c5.0204260645.3af87dc5@posting.google.com> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <vIxz8.81546$vF6.2424489@news2.tin.it>

Ralf Juengling wrote:
        ...
> BTW: It is not easy to add __iter__ to list's meta-type, is it?

list's "metatype" (metaclass) is the built-in type named 'type'.
like for all other built-in types and new-style classes.  Adding
__iter__ to type would thus make almost 'everything' iterable.

And giving list a different metaclass than all the other types
would no doubt be weird.

Further, it would be utterly weird if 'take' was introduced,
operated correctly on iterables (including iterators), but
returned a list rather than an iterator!  It would freak me
out... we finally have a 'lazy' paradigm, and we'd just
throw it away like this?!  take and drop should both accept
and return iterators.  You can make any (bounded) iterator
into a list by calling list on it, just like you can make
a bounded iterator from any sequence by calling iter.

If we had 'take', you could already use it for such purposes:

>>> from __future__ import generators
>>> def take(n, iterable):
...     for item in iterable:
...         if n<=0: raise StopIteration
...         n -= 1
...         yield item
...
>>> List2D = list(take(5, iter(list, '')))
>>> List2D
[[], [], [], [], []]
>>>

Remember the two-arguments form of iter lets you make ANY
callable-without-arguments into an iterator -- no need to
muck around with 'type' or anything like that.

I don't see what's more elegant in this way to build List2D
than the obvious list-comprehension:

List2D = [ [] for i in xrange(5) ]

but that's another issue.


Alex



From jepler at unpythonic.net  Tue Apr 23 22:50:48 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Tue, 23 Apr 2002 21:50:48 -0500
Subject: stdin - seek() issue on Linux.
In-Reply-To: <284b44c2.0204231816.67c01ca8@posting.google.com>
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com> <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com> <aa1bqb$lek$1@nntp6.u.washington.edu> <284b44c2.0204221908.1280cfc3@posting.google.com> <284b44c2.0204231816.67c01ca8@posting.google.com>
Message-ID: <20020423215048.A2102@unpythonic.net>

On Tue, Apr 23, 2002 at 07:16:41PM -0700, Srihari Vijayaraghavan wrote:
> Once again thanks for your responses. I really appreciate that.
> 
> OK. IMHO sys.stdin doesn't work as stated in PLR (Python Library
> Reference), i.e., it doesn't work like a _real_ file object,
> particularly not supporting seek() related operations. While your
> technical explanation sounds OK to me, but I am still not convinced
> why PLR advertises that sys.stdin is treated as a file object. Please
> correct me if I am wrong here (as I am still a newbie user :-).

Please try the following C program with STDIN both as a terminal and as a
file:
    cat > test_seek.c <<EOF
    #include <stdio.h>     /* perror() */
    #include <sys/types.h> /* lseek() */
    #include <unistd.h>    /* lseek() */
    #include <errno.h>     /* errno */
    int main(void) {
	errno = 0;
	lseek(0, 0, SEEK_SET);
	perror("lseek");
	return 0;
    }
    EOF
    cc test_seek.c -o test_seek
    ./test_seek  # fails 'lseek: Illegal seek'
    ./test_seek < /dev/null  # succeeds 'lseek: Success'
    
or this Perl program:
    perl -e 'seek STDIN, 0, 0 or die $!' # fails: Illegal seek at -e line 1.'
    perl -e 'seek STDIN, 0, 0 or die $!' < /dev/null # sicceeds

Not all files are seekable.  On Unix, notable exceptions are terminals,
pipes, and sockets.

If you don't know in advance whether a particular "file" is seekable, you
can test it without any side-effects by
    try:
	f.seek(0, 1)
    except IOError:
	f is not seekable (wrap with CStringIO or temporary file if
	    seeking is required)
    else:
	f is seekable
f.seek(0, 1) seeks 0 bytes from the current position.  This should succeed
without changing anything on a normal file, and fail without changing
anything on a non-seekable file.

A note to people coming from a DOS background: In DOS, a pipeline "a|b"
writes the output from a to a temporary file, then invokes b with its input
from that temporary file.  In Unix, a pipeline "a|b" executes a and b at
the same time with a "pipe" connecting the two processes.  This difference
can be noticed when trying to seek, and sometimes when there are buffering
effects in the pipe between A and B.

One last note -- the file descriptor returned by popen*() is also
non-seekable.

Jeff




From dalke at dalkescientific.com  Thu Apr 18 02:15:05 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Thu, 18 Apr 2002 00:15:05 -0600
Subject: Does Python need a '>>>' operator?
References: <mailman.1018937566.11432.python-list@python.org> <slrnabs53e.1kfm.Gareth.McCaughan@g.local>
Message-ID: <a9lo2f$9ji$1@nntp9.atl.mindspring.net>

Gareth McCaughan:
>A friend of mine has a project of classifying all finite
>chromatic mathematical fruit jokes. He knows about
  [three omitted for sake of humanity]

>and I think those are all he's found. He conjectures that
>there are no more, but the conjecture appears difficult to
>prove without tiresome case-splitting. Perhaps a computer
>proof is in order.

I did a computer proof.  Google helped me find a few that
were at least close

First off, there are plenty of grape variations, starting
with green instead of purple.

http://www.xs4all.nl/~jcdverha/scijokes/1_6.html
Q: What's purple, commutes, and is worshiped by a limited number
   of people?
A: A finitely venerated abelian grape.
Q: What's green, dangerous and commutative?
A: An abelian grape with a machine gun.
   (and that one's just plain silly).
Q: What's yellow, linear, normed and complete?
A: A Bananach space.


http://www.math.washington.edu/~plochin/Humor/mathjokes.txt
 What's green and homeomorphic to the open unit interval?
       The real lime.
 What's lavender and commutes?
       An abelian semigrape.


http://www.math.utah.edu/~cherk/mathjokes.html
The cherry theorem (a puzzle that reminds some of calculus theorems)
Q: What is a small, red, round thing that has a cherry pit inside?
A: A cherry.

http://ctl.unbc.ca/CMS/dailyjokes.html
Q:   What's green and very, very far away?
A:   The lime at infinity.

http://www.materialisations.com/humour/humour-math.html
Q: What is yellow and covered in blue paint?
A: The pumping lemon.

http://www.cs.stir.ac.uk/~scu/Humour/Science/maths.html
Q: What do you get when you cross an elephant with a banana?
A: Elephant banana sine theta in a direction mutually perpendicular
to the two as determined by the right hand rule.

(eh, but there's no color in that last one)


Excuse me while I now go crazy from reading too many of these pages.

                    Andrew
                    dalke at dalkescientific.com





From aleax at aleax.it  Tue Apr  2 10:54:58 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 15:54:58 GMT
Subject: using (was Re: The Python Way)
References: <5515F24370BCB753.DA7CF91464346EF6.E3A577CFA6882D82@lp.airnews.net> <a7qqhr$m5$1@panix1.panix.com> <jjb5-0A3399.14320327032002@newsstand.cit.cornell.edu> <a80nej$jlp$1@panix1.panix.com>
Message-ID: <mvkq8.43706$pT1.1223399@news1.tin.it>

Aahz wrote:
        ...
>     using MyLock():
>         <code>
> 
> The using construct would automatically call the start() and finish()
> methods of the MyLock() instance, no matter what exceptions got raised
> in <code>.

I definitely wouldn't mind a slightly more compact equivalent of

try:
    <code>
finally:
    <some guaranteed termination>

particularly with the ability of displaying the termination code
that's guaranteed to run BEFORE the code that's being tried (when
it's important to let the reader know at once what code he or she
can take as guaranteed-to-run-at-end).  But is the new keyword
'using' and the specific convention of .start() and .finish()
calls all warranted?  Wouldn't the ability to write the finally
clause before the try one suffice for the "display the key part
first" stylistic purpose, for example?  And you might then (just
maybe) omit the try: line, i.e. have a new 'finally:' statement
where:

finally:
    <guaranteed termination stuff>

<code>

has the same semantics as:

try:
    <code>
finally:
    <guaranteed termination stuff>


I dunno -- maybe this isn't clear enough, or error-prone wrt
indentation errors of some try clause and/or its body.  But I
suspect that if we need a new keyword for this, we ain't gonna
get it (just pessimistic, maybe...:-)


Alex



From akuchlin at ute.mems-exchange.org  Sun Apr 14 15:27:12 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 14 Apr 2002 19:27:12 GMT
Subject: 2.1 vs. 2.2 
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <As2u8.17016$vF6.537198@news2.tin.it>
Message-ID: <slrnabjm0g.c5v.akuchlin@ute.mems-exchange.org>

In article <As2u8.17016$vF6.537198 at news2.tin.it>, Alex Martelli wrote:
> I don't see anything "broken" about 2.2.  The changes to the object
> model invalidated extensions which relied on very deep characteristics
> of Python 2.1 and earlier, such as ExtensionClass.  The next major

Actually, current versions of ExtensionClass work fine with 2.2.  Zope
itself is only tested with Python 2.1, but we use ExtensionClass under
2.2 every day and have no problems.

--amk


From pan-newsreader at thomas-guettler.de  Thu Apr  4 16:54:54 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Thu, 04 Apr 2002 23:54:54 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk> <MPG.17164ee731d2609f989698@news.nwlink.com>
Message-ID: <pan.2002.04.04.23.54.54.239035.1346@thomas-guettler.de>

On Thu, 04 Apr 2002 21:52:57 +0200, Jeff Shannon wrote:

> In article <3CAC1665.8040209 at mxm.dk>, maxm at mxm.dk says...
>> Thomas Guettler wrote:
>> 
>> 
>> > Is there already a function in the standard python modules which
>> > returns the a relative url given two absolute urls?
>> > 
>> > Example:
>> > 
>> > relative_url("http://foo/a/b/c", "http://foo/a/d")
>> >  --> return: "../d"
>> 
>> I hope you mean that the result should be "../../d" ??
> 
> Actually, that's not quite correct (and neither are your code examples
> below, for the same reason).
> 
> In the URL 'http://foo/a/b/c', we could presume that 'foo' is a domain
> name, 'a' and 'b' are directories, and 'c' is a filename. Thus, the
> 'current directory' ('.') for this URL is //foo/a/b, and the parent
> directory ('..') is //foo/a.  Therefore, in this case, the O.P.'s
> initial relative URL ('../d') is correct.
> 
> The issue is complicated somewhat in that '//foo/a/b/c' can *also* be
> shorthand for '//foo/a/b/c/index.html' (or whatever the server's
> default-filename is).  If (and *only* if) the original URL is
> representing a default document, then your result would be the correct
> one.

If you try to access //foo/a and "a" is a directory the webserver
redirects you to /foo/a/:

--guettli at sonne:~/python$ telnet localhost 80
Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.
GET /~guettli/test
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<HTML><HEAD>
<TITLE>301 Moved Permanently</TITLE>
</HEAD><BODY>
<H1>Moved Permanently</H1>
The document has moved <A HREF="http://sonne.heaven/~guettli/test/">here</A>.<P>
</BODY></HTML>
Connection closed by foreign host.

Since there doesn't seem to be a function, I did it like this:


import urlparse
import re
import string

def relative_url(source, target):
	su=urlparse.urlparse(source)
	tu=urlparse.urlparse(target)
	junk=tu[3:]
	if su[0]!=tu[0] or su[1]!=tu[1]:
		#scheme (http) or netloc (www.heise.de) are different
		#return absolut path of target
		return target
	su=re.split("/", su[2])
	tu=re.split("/", tu[2])
	su.reverse()
	tu.reverse()

	#remove parts which are equal   (['a', 'b'] ['a', 'c'] --> ['c'])
	while len(su)>0 and len(tu)>0 and su[-1]==tu[-1]:
		su.pop()
		last_pop=tu.pop()
	if len(su)==0 and len(tu)==0:
		#Special case: link to itself (http://foo/a http://foo/a -> a)
		tu.append(last_pop)
	if len(su)==1 and su[0]=="" and len(tu)==0:
		#Special case: (http://foo/a/ http://foo/a -> ../a)
		su.append(last_pop)
		tu.append(last_pop)
	tu.reverse()
	relative_url=[]
	for i in range(len(su)-1): 
		relative_url.append("..")
	rel_url=string.join(relative_url + tu, "/")
	rel_url=urlparse.urlunparse(["", "", rel_url, junk[0], junk[1], junk[2]])
	return rel_url

def test_relative_url(source, target, result):
	res=relative_url(source, target)
	if res!=result:
		print "Test FAILED: result is:", res, "should be:", result, \
			  "source:", source, "target:", target
	else:
		print "Test ok: result is:", res, "source:", source, "target:", target
		
	
test_relative_url("http://foo/a/b", "http://foo/c", "../c")
test_relative_url("http://foo/a/b", "http://foo/c/d", "../c/d")
test_relative_url("http://foo/a/b", "ftp://foo/c", "ftp://foo/c")
test_relative_url("http://foo/a", "http://foo/b", "b")
test_relative_url("http://foo/a/", "http://foo/b", "../b")
test_relative_url("http://foo:80/a/", "http://foo/b", "http://foo/b")
test_relative_url("http://foo:8080/a/", "http://foo/b", "http://foo/b")
test_relative_url("http://foo/a", "http://foo/a", "a")
test_relative_url("http://foo/a/", "http://foo/a", "../a")
test_relative_url("http://foo/a", "http://foo/b/c", "b/c")
test_relative_url("http://foo/a/", "http://foo/b/c", "../b/c")
test_relative_url("http://foo/a/b", "http://foo/c/d", "../c/d")
test_relative_url("http://foo/a/b/", "http://foo/c/d", "../../c/d")
test_relative_url("http://foo/a", "http://foo/b", "b")
test_relative_url("http://foo/a;para?query#frag", "http://foo/a", "a")
test_relative_url("http://foo/a", "http://foo/a;para?query#frag",
				  "a;para?query#frag")


Feedback welcome
 thomas


-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From borcis at geneva-link.ch  Mon Apr  8 13:51:41 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 08 Apr 2002 19:51:41 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk> <3CAFAAA4.41EA7CF9@python.org> <+CcOX7A2gWs8EwPR@jessikat.fsnet.co.uk> <3CB1846B.5050109@geneva-link.ch> <roy-995FAF.10470208042002@news1.panix.com>
Message-ID: <3CB1D8AD.AE06866F@geneva-link.ch>

Roy Smith wrote:
> 
> If he wanted to be truly elegant and economical, he'd have simply written
> "<ploink>".

Smells like a rodeo.

Boris
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")


From emile at fenx.com  Wed Apr 10 23:31:07 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 11 Apr 2002 03:31:07 GMT
Subject: Tuple comprehension
References: <slrnaba052.efc.mlh@vier.idi.ntnu.no>
Message-ID: <%r7t8.37204$CU.41841@rwcrnsc51.ops.asp.att.net>

Magnus Lie Hetland
> As far as I can see, the PEP on dict comprehension is still open... I
> think both dictionary comprehension and tuple comprehension seem like
> natural parallels to list comprehension. But then I would, wouldn't I?
> <wink>
>

Why not just tuple([my list comp]) ?


--

Emile van Sebille
emile at fenx.com

---------



From gimbo at ftech.net  Fri Apr 12 04:13:27 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 12 Apr 2002 09:13:27 +0100
Subject: Help: Python for a big commercial project?
In-Reply-To: <MJdt8kKkXAtS092yn@gnosis.cx>
References: <F194tIpqqJgSt51NDa200002509@hotmail.com> <mailman.1018464082.15807.python-list@python.org> <roy-EB08BB.20470710042002@news1.panix.com> <60F7BB039BFDD68F.7FE6889DFED7EA98.93812C99068FD653@lp.airnews.net> <MJdt8kKkXAtS092yn@gnosis.cx>
Message-ID: <20020412081327.GA13919@andy.tynant.ftech.net>

On Thu, Apr 11, 2002 at 02:13:32PM -0400, Lulu of the Lotus-Eaters wrote:

> The other important aspect is the motto/mantra that all Pythonistas[*]
> should repeat to them upon waking in the morning:  "Tim Peters is
> smarter than I am."
> 
> Yours, Lulu...
> 
> [*] Nope! There is not an off-by-one error here.

Yeah...  Everybody loves my baby, but my baby loves nobody but me.

:-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From mgilfix at eecs.tufts.edu  Thu Apr 18 01:33:47 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 18 Apr 2002 01:33:47 -0400
Subject: Pure python crypto
In-Reply-To: <3CBE53E8.8080707@verio.net>; from vlindberg@verio.net on Wed, Apr 17, 2002 at 11:04:40PM -0600
References: <20020417021703.H16954@eecs.tufts.edu> <02041718383201.00171@toad> <mailman.1019063074.28268.python-list@python.org> <3CBE53E8.8080707@verio.net>
Message-ID: <20020418013346.O924@eecs.tufts.edu>

  Heh. So would I. I'm definitely interested in the project. Perhaps
we could launch one? It *is* a large undertaking though. What crypto
modules are already out there though (stuff that's been implemente
that we could use)...?

            -- Mike

On Wed, Apr 17 @ 23:04, VanL wrote:
> [Snip Various]
> 
> If you are implementing a sshlib module (for lack of a better term) I 
> would be *very* interested.
> 
> VanL
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (vlindberg)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From marklists at mceahern.com  Fri Apr 12 09:45:40 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 08:45:40 -0500
Subject: emacs python mode completion
In-Reply-To: <871ydl75j7.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMELHKKAA.marklists@mceahern.com>

[Stephen J. Turnbull]
> Although it's not exactly what you what, something that works in all
> modes is dabbrev (bound to M-/ in XEmacs by default).  This completes
> words according to context.  So if I type M c M-/ here I get "McEahern."

Thanks!

// mark




From w3miller at yahoo.com  Tue Apr  9 01:54:25 2002
From: w3miller at yahoo.com (Walter Miller)
Date: Mon, 8 Apr 2002 22:54:25 -0700 (PDT)
Subject: SSL error:  2.1.1 works but 2.2 doesn't
Message-ID: <20020409055425.53965.qmail@web21411.mail.yahoo.com>

Apologies if this has been discussed/solved already. 
I looked for prior postings but haven't found a
solution for fixing this error I'm getting with
_socket.pyd on 2.2.

Using Gerhard H?ring's "socket.pyd with SSL support
for Python 2.2 on Windows"
(http://www.cs.fhm.edu/~ifw00065/downloads/python22-win32-ssl.zip):

"""
Traceback (most recent call last):
  File "C:\test.py", line 3, in ?
    doc =
urllib2.urlopen("https://www.medi-cal.ca.gov").read()
  File "C:\Python22\lib\urllib2.py", line 138, in
urlopen
    return _opener.open(url, data)
  File "C:\Python22\lib\urllib2.py", line 322, in open
    '_open', req)
  File "C:\Python22\lib\urllib2.py", line 301, in
_call_chain
    result = func(*args)
  File "C:\Python22\lib\urllib2.py", line 792, in
https_open
    return self.do_open(httplib.HTTPS, req)
  File "C:\Python22\lib\urllib2.py", line 774, in
do_open
    code, msg, hdrs = h.getreply()
  File "C:\Python22\lib\httplib.py", line 728, in
getreply
    response = self._conn.getresponse()
  File "C:\Python22\lib\httplib.py", line 572, in
getresponse
    response = self.response_class(self.sock)
  File "C:\Python22\lib\httplib.py", line 98, in
__init__
    self.fp = sock.makefile('rb', 0)
  File "C:\Python22\lib\httplib.py", line 607, in
makefile
    buf = self.__ssl.read()
socket.sslerror: (5, 'EOF occurred in violation of
protocol')
""""


Here's the code:

"""
import urllib2

doc =
urllib2.urlopen("https://www.medi-cal.ca.gov").read()
print doc
"""


This works using Python 2.1.1 and Al Dass's
_socket.pyd: 
http://home.attbi.com/~al.dass/dev/python/2.1.1/win32_socket_ssl/python_2_1_1_ssl_socket.zip


any ideas?  BTW, using Win2k Pro.  I also verified
this on FreeBSD 4.5: Python 2.1.1 "works" (returns the
document) but all subsequent versions produce errors.

Walter

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/




From aleax at aleax.it  Sat Apr 27 18:34:20 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 22:34:20 GMT
Subject: Python's Lisp heritage
References: <mailman.1019922475.8361.python-list@python.org> <tyfsn5h13mh.fsf@pcitapi22.cern.ch> <ZADy8.70845$vF6.2115921@news2.tin.it> <pan.2002.04.27.17.15.43.188185.9182@uchicago.edu>
Message-ID: <MHFy8.28398$8D3.758370@news1.tin.it>

Ben Wolfson wrote:
        ...
>> Anyone for Haskell's f a b c ...?  Now THAT is clean and elegant... the
        ...
> No parentheses in the simple case, maybe, but at least in SML, which is
> similar, the parentheses can multiply pretty quickly if one of the
> parameters is the result of a function, or a function returns a function:
> (f a (g b)) (h c) isn't all that pretty, unless you like Lisp.  Haskell
> may be smarter about that kind of thing, though.

No, you do need SOME parentheses, if the normal priority is not the 
one you want -- I'd *hate* a language so "smart" as to parse
        f a g b h c
in many different ways depending on declared or inferred types for
the identifiers, or anything like that (usability == 0).

Still, I don't see much in it either way between the parenthesization
you deem "not all that pretty":
        (f a (g b))(h c)
and the Python equivalent:
        f(a, g(b))(h(c))

The former may be obviously reduced to:
        f a (g b) (h c)
"a function returning a function" is the USUAL case after all, for all
situations that in other languages you'd consider "a function with N>1
arguments".  So, you don't need the parentheses around
        f a (g b)
although you may of course use redundant parentheses if and when
you feel they make an expression prettier, as in any other language:-).

The parentheses around (g b) ARE needed because you want to
indicate that g applies to b and the RESULT of that is an argument
to the unnamed function returned by (f a) -- otherwise, g would be
the argument to (f a), and b the argument to the resulting unnamed
function f a g , aka ( (f a) g ) if you want redundant parentheses.


Alex



From aahz at pythoncraft.com  Tue Apr 30 17:34:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 30 Apr 2002 17:34:45 -0400
Subject: Python vs. Perl, which is better to learn?
References: <mailman.1020147892.13578.python-list@python.org> <aamjat$qku$1@panix1.panix.com> <aamm5k$kv3$1@bubba.NMSU.Edu>
Message-ID: <aan2ll$ib6$1@panix1.panix.com>

In article <aamm5k$kv3$1 at bubba.NMSU.Edu>,
Bob Melson <"melsonr at earthlink.net"> wrote:
>
>Look at it this way -- you pick the best language tool to do the job at
>hand.  If python is the better language for that job, you use python; if
>it's perl, then you use perl; if it's some other language, you use _that_
>language.  If you're asked to build a house, you certainly don't confine
>yourself to a single tool -- you use the tools appropriate to the job you
>have to do; why should it be different with scripting languages?

Fine sentiment.  Now, under what circumstances would you recommend Perl
over Python?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From zooko at zooko.com  Wed Apr  3 12:03:43 2002
From: zooko at zooko.com (Zooko)
Date: Wed, 03 Apr 2002 09:03:43 -0800
Subject: Monitor class (was: The Python Way)
Message-ID: <E16soAi-0000nK-00@imp>

There was discussion about a general lock-manager meta-class:

http://groups.google.com/groups?th=1e465bd53977862f

I think this is what is called a "Monitor".  It's an object which contains a 
lock, and whenever you invoke any method of that object the lock is acquired 
before the method is run and then released afterward.

I implemented something like this a couple of years ago and it has been in 
continuous use ever since: [1].  It comes with unit tests that show what it does
and give some limited assurance that it works [2].

There are several reason why you shouldn't trust this code too much:

1.  I don't use it much anymore as I've moved to an event-based paradigm with 
very little need for locking.  (See [3].)

2.  This Python meta-programming stuff is *funky*.  I tried at least three other 
designs for this purpose which never quite worked right before I finally hit 
upon this one.  Mistakes on my part or changes in new versions of Python might 
introduce subtle errors.

3.  The code has been moved around and massaged a bit and it might have broken 
in some subtle way.

On the other hand, it does pass three unit tests and if you try using it and 
have some problems (including the problem of "I don't understand how to use 
this!"), then I'll fix it for you in the interests of keeping my pyutil 
library [4] will be up to date and useful.

Regards,

Zooko

[1] http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pyutil/pyutil/Monitor.py?rev=HEAD&content-type=text/vnd.viewcvs-markup
[2] http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pyutil/pyutil/Monitor_unittest.py?rev=HEAD&content-type=text/vnd.viewcvs-markup
[3] http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/pyutil/pyutil/DoQ.py?rev=HEAD&content-type=text/vnd.viewcvs-markup
[4] http://sf.net/projects/pyutil

---
                 zooko.com
Security and Distributed Systems Engineering
---




From mlh at vier.idi.ntnu.no  Sat Apr 13 10:39:00 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 13 Apr 2002 14:39:00 +0000 (UTC)
Subject: Checking whether an async* dispatcher is "alive"
Message-ID: <slrnabggo4.o80.mlh@vier.idi.ntnu.no>

In a server I'm working on, I need to check whether an asynchat
dispatcher is still connected, i.e. whether I can send it data. I've
tried to use try/except (with socket.error) but since async* catches
that itself, It's no good.

Are there any methods/attributes that indicate whether I can send data
on a dispatcher? (Possibly some method/attribute on its socket
object?)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From woodsplitter at rocketmail.com  Thu Apr 25 18:54:32 2002
From: woodsplitter at rocketmail.com (David Rushby)
Date: Thu, 25 Apr 2002 18:54:32 -0400
Subject: sql but... without server
References: <7fvfcu0vkc610fma8545v0nhdm7c6hem7q@4ax.com>
Message-ID: <qM%x8.21681$z55.13474@atlpnn01.usenetserver.com>

> I  would like to use sql database with python but I do not want to
> install any sql server.
>
> I would like to use sql engine as a standalone file (e.g. dll).

SQLite is probably a good solution:
http://www.hwaci.com/sw/sqlite/

There are at least two *very* immature Python bindings floating around, so
your mileage may vary.





From aleax at aleax.it  Thu Apr 11 04:39:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 08:39:44 GMT
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
Message-ID: <kZbt8.5601$b62.137821@news1.tin.it>

Brad Clements wrote:

> Could someone explain this to me.
> 
> Lots of code does  type(x) in [InstanceType]

I'll explain to you why lots of code does typetesting if you explain to me 
lemmings' alleged propensity for hurling themselves off a cliff.  You go 
first -- I suspect I can then just copy and paste your explanation right 
back to you:-).


Alex



From loewis at informatik.hu-berlin.de  Wed Apr 17 11:33:47 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Apr 2002 17:33:47 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019055757.10130.python-list@python.org>
Message-ID: <j4pu0y9w6c.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> WARNING: removing "math" since importing it failed
>         [...]

> So even building statically doesn't work.

You did not build statically. Instead, all extension modules are build
by setup.py now.

It would be crucial to understand *why* the imports fail. For that, it
would be crucial to perform the following actions:

1. Re-invoke the commands that are used to build the math module
   manually (without invoking make; just invoking the compiler)

2. Invoke ./python

3. 'import math'

4. Report the errors you get when invoking math that way

5. Report the command line used to link python.

Regards,
Martin



From loewis at informatik.hu-berlin.de  Wed Apr 24 06:32:40 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 24 Apr 2002 12:32:40 +0200
Subject: xml question?
References: <aa6121$bgf$03$1@news.t-online.com>
Message-ID: <j4it6hjsjb.fsf@informatik.hu-berlin.de>

"Alex Adrian" <Alex.Adrian at gmx.de> writes:

> i would like to make this string xml conform? is there an easy
> filterfunction in python to convert characters like "<>??" .... to xml
> conform characters??

Sure. XML supports arbitrary characters - you just have to declare the
encoding.

Replace

> s = """?xml version="1.0"?>

with

s = """<?xml version="1.0" encoding="iso-8859-1"?>

Regards,
Martin


From mfranklin1 at gatwick.westerngeco.slb.com  Wed Apr  3 07:49:13 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 3 Apr 2002 13:49:13 +0100
Subject: Newbie: issuing shell commands with mod_python
In-Reply-To: <lFCq8.906$d45.95879@monolith.news.easynet.net>
References: <lFCq8.906$d45.95879@monolith.news.easynet.net>
Message-ID: <200204031249.g33CnUR26768@helios.gatwick.geco-prakla.slb.com>

On Wednesday 03 Apr 2002 1:32 pm, neil wrote:
> import os
> def main(req):
> ? ? child = os.popen("/usr/bin/htpasswd -cb /usr/local/apache/users/myfile
> name1 passwd1")
> ? ? data = child.read()
> ? ? err = child.close()
> ? ? return data


Try this from a python shell:-

[$ python
Python 2.0 (#11, Nov  6 2000, 09:01:06)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import os
>>> def main(req):
...     child = os.popen("/usr/bin/htpasswd -cb   BLAHBLAH martin martin1")
...     data = child.read()
...     err = child.close()
...     return data
...
>>> main('foo')
Adding password for user martin
''


Cheers
Martin




From clpy at snakefarm.org  Tue Apr 23 05:26:32 2002
From: clpy at snakefarm.org (Carsten Gaebler)
Date: Tue, 23 Apr 2002 11:26:32 +0200
Subject: Why won't this POST
References: <8f8ffe67.0204222047.47d661ec@posting.google.com>
Message-ID: <3CC528C8.FDE414CE@snakefarm.org>

Sunit Joshi wrote:

> I'm trying to post data to IIS5.0 but keep getting error message about
> "host not found".

> # begin HTTP request
> req = HTTP('localhost/eubix')
> #req.putrequest("GET" ,'/PyCgi/formTest.py?' + postData)
> req.putrequest('POST' ,'/cgi-bin/ProcessInput.py') #Here's where it

"localhost/eubix" is not a valid hostname. What's the "/eubix" part
supposed to mean?

cg.
-- 
Why settle for snake oil if you can have the whole snake?
http://python.org/


From neuhauser at mail.cz  Sat Apr 27 14:07:27 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Sat, 27 Apr 2002 20:07:27 +0200
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
In-Reply-To: <20020427130942.B1588@eecs.tufts.edu>
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com> <mailman.1019801261.30122.python-list@python.org> <Wnvy8.51932$z55.27618@atlpnn01.usenetserver.com> <20020427171024.GA82912@roman.mobil.cz> <20020427130942.B1588@eecs.tufts.edu>
Message-ID: <20020427180727.GB82912@roman.mobil.cz>

> From: Michael Gilfix <mgilfix at eecs.tufts.edu>
> To: python-list at python.org
> Subject: Re: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
> Date: Sat, 27 Apr 2002 13:09:42 -0400

> On Sat, Apr 27 @ 19:10, Roman Neuhauser wrote:
> >     Anyone who writes "if (x == 0)" instead of "if (0 == x)" deserves
> >     what he begs for.
> 
>   I disagree entirely. When writing production code, you want to model
> the thought process that goes into creating the code as clearly if
> possible.  If in your head you're thinking "if 'x' is equal to 0",
> then you probably want to write it that way. While writing 0==x is
> a better move in C, it's entirely against intuition and the natural
> thought process.

    I disagree entirely.

    When I write code @work, I think "if 'x' is equal to 0", because
    one of my colleagues couldn't wrap his head around the safe way.
    When I write my own code, I think "if 0 is equal to 'x'". No, wait,
    I don't know what I think: the thought is abstract, and can be
    written both ways. But I actually have to remember to put the
    variable on the left, and the constant on the right side.

>   Python is smart to avoid such assignments...

    Python thinks I'm too dumb to take care of myself.

-- 
FreeBSD 4.4-STABLE
8:02PM up 12 days, 9:14, 10 users, load averages: 0.06, 0.04, 0.01




From aleax at aleax.it  Wed Apr  3 02:39:23 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 07:39:23 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <Xns91E4EF89BD4EDcliechtigmxnet@62.2.16.82> <pTpq8.53356$S52.1844974@news2.tin.it> <eppstein-E6E1AD.15234102042002@news.service.uci.edu>
Message-ID: <Lkyq8.45231$pT1.1286029@news1.tin.it>

David Eppstein wrote:

> In article <pTpq8.53356$S52.1844974 at news2.tin.it>,
>  Alex Martelli <aleax at aleax.it> wrote:
> 
>> > yes. but a real bool could be introduced as module for now (for the RPC
>> > and marshalling people that need one) i think its not needed otherwise
>> 
>> A standard library module sounds good.  Or maybe it could go into
>> operator.
> 
> So you'd be happy with a Boolean class, as long as it isn't convertable
> to integers and none of the standard operators are changed to return it?

I don't think it would do any damage if int(x) accepted an x of such
a class, just as it accepts, e.g., an x of type str.


> That doesn't seem very nice either from the aesthetic purity-of-design
> point of view (non-convertable is more pure than convertable maybe but

There's no "loss of purity" in int('23') returning 23, so why should there
be any in, say, int(Boolean.True) returning 1?

> (Booleans would be too cumbersome to use except in unusual situations).

They'd come in handy for those peculiar situations where you HAVE
to distinguish between integers and booleans, basically only when you are
interfacing some external subsystem such as a DB or RPC system that
does require such distinctions (as opposed to ones that _don't_ support
it, such as, we're told, MySql).  And, by not affecting built-ins, 
operators, etc, they wouldn't be in the way in everyday life, nor, most
particularly, would they impose conceptual burdens on teaching Python.


> In comparison the actual PEP is similarly not-pure but gives better

Far more "not-pure", since the PEP requires _implicit_ conversion (or
else tons of code would break).

> expressivity and possibly adds minor functionality (ability to make code
> that behaves differently for ints and for bools, just like now you can

It would be quite possible to distinguish between int and Booleans if
you needed to, even if Booleans came from a specific module rather
than from builtins, of course.  Such a need is a very rare issue, thus
it should not be allowed to get in the way for everyday usage.

> behave differently for lists and for tuples even though they are
> interconvertable).

[1,2]+(3,4) raises an exception.  All sequences are convertable
_explicitly_ (tuple([1,2]) and so forth).  But very rare needs, such
as homogeneous arrays, are quite properly shunted off to their
own standard modules, they DON'T get in the way of our common
usage of the general builtin sequences.


Alex



From mwh at python.net  Wed Apr  3 16:40:45 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 3 Apr 2002 21:40:45 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org> <a8ed9i$rnodf$1@ID-41653.news.dfncis.de> <3cab55f4.61190667@news.laplaza.org> <a8fn61$s1nk7$1@ID-41653.news.dfncis.de>
Message-ID: <lky9g4o3z7.fsf@elios.maths.bris.ac.uk>

Tobias Klausmann <klausman-un030402 at tuts.net> writes:

> So far, I've only written libraries/modules (no shebang) or console
> apps that are at least semi-interactive. But if I'd write some such
> utility that might run cronned, I'd surely stumble across the system
> where cron doesn't know about /usr/local/bin/python15 =)

The distutils can help here.

Cheers,
M.


From daniel.dittmar at sap.com  Fri Apr 12 07:34:41 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 12 Apr 2002 13:34:41 +0200
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com>
Message-ID: <a96goi$oe9$1@news1.wdf.sap-ag.de>

>   File "C:\Utils\Python21\lib\ftplib.py", line 273, in ntransfercmd
>     host, port = parse227(self.sendcmd('PASV'))
>   File "C:\Utils\Python21\lib\ftplib.py", line 229, in sendcmd
>     return self.getresp()
>   File "C:\Utils\Python21\lib\ftplib.py", line 202, in getresp
>     raise error_perm, resp
> ftplib.error_perm: 500 Unrecognized command

It seems as if the FTP server doesn't know the PASV command. You could try
def set_pasv(self, val):
        '''Use passive or active mode for data transfers.
        With a false argument, use the normal PORT mode,
        With a true argument, use the PASV command.'''
to use normal instead of passive mode.

You should also pass the open file to storbinary instead of the read method.

Daniel





From timr at probo.com  Tue Apr 30 01:26:18 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 29 Apr 2002 22:26:18 -0700
Subject: fileinput.input() fails on Windows 2K
References: <aaha32$heu$05$1@news.t-online.com>
Message-ID: <ujascusubjtf5a48hcbbaqnnf6uh9eafhf@4ax.com>

"Ingo Blank" <spam at yourself.com> wrote:
>
>this tiny script gives a traceback under Windows 2K if/when input is piped
>in to stdin, while it runs as expected under Linux.
>
>## MakeDef.py
>## Read ouput of DUMPBIN /EXPORTS <file.DLL> and create a .DEF file
>
>from string import *
>import fileinput
>
>print "EXPORTS"
>hit = 0
>for line in fileinput.input():
> words = split(line)
> if hit and len(words) == 4:
>  print "\t%s" % words[3]
> hit |= find(line,"RVA") >= 0
>
>----------------
>
>C:>DUMPBIN /EXPORTS libjpeg.dll | MakeDEF.py
>
>Traceback (most recent call last):
>  File "C:\MakeDEF.py", line 6, in ?
>    for line in fileinput.input():
>  File "E:\Python21\lib\fileinput.py", line 181, in __getitem__
>    line = self.readline()
>  File "E:\Python21\lib\fileinput.py", line 262, in readline
>    self._buffer = self._file.readlines(self._bufsize)
>IOError: [Errno 9] Bad file descriptor
>
>-----------------
>
>Am I missing something, or is this a bug in fileinput.py ?
>I'm using ActiveState Python 2.1.

It is a known bug, apparently in the Windows 2000 shell itself.  You can
work around the problem by invoking Python instead of relying on the .py
extension:

C:>DUMPBIN /EXPORTS libjpeg.dll | python MakeDEF.py

You can use popen from inside MakeDEF.py to invoke the DUMPBIN command and
parse the output, if you want.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From hancock at anansispaceworks.com  Sun Apr 21 09:44:27 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sun, 21 Apr 2002 06:44:27 -0700
Subject: word stemming / conflation
Message-ID: <3CC2C23B.DAF9057D@anansispaceworks.com>

I was working on an algorithm to condense strings of
text into mnemonic labels (hopefully this will produce
more meaningful labels than simply assigning a number!).
I got the idea from the way later versions of
latex2html generate filenames.

Anyway, after tinkering with this for awhile I've
discovered that it can be a bottomless pit! A little
web research reveals various MS / PhD papers for doing
this task (which apparently is called "conflation" as
well as "word stemming") in English and lots of other
languages. Looks like it's a lot harder than you
might at first think. I didn't find much in Python,
though.

So, I'm wondering if there are some Python resources
(or better-yet already written modules) for doing
this sort of thing.  I don't really need a completely
robust system -- the occasional error will be quite
tolerable. So I'm more interested in quick and simple
solutions that work most of the time, rather than
the ultra-robust library-science type.

Thanks for any information or leads you might be
able to suggest,

Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
------------------------------------------------------




From sholden at holdenweb.com  Tue Apr 16 10:26:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 10:26:10 -0400
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no> <s4zu8.20814$vF6.690664@news2.tin.it> <slrnabobad.8ho.mlh@vier.idi.ntnu.no> <SmWu8.26250$vF6.834282@news2.tin.it>
Message-ID: <2yWu8.26048$Ji7.10720@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in ...
> Magnus Lie Hetland wrote:
>         ...
> >>I think the best name for what's now raw_input might be something
> >>like getline.
> >
> > Or perhaps readline (? la file.readline)?
>
> I slightly prefer getline, because raw_input differs from a
> somefile.readline -- the latter does not strip off the '\n',
> while raw_input does -- each of the two slightly different
> behaviors being optimal in its context.  I think it would be
> harder to teach if the names coincided.
>
There's also the minor difference that raw_input() takes an argument to be
used as an interactive prompt!

regards
 Steve






From cliechti at gmx.net  Sat Apr 27 21:10:07 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Apr 2002 03:10:07 +0200
Subject: input/output through socket or COM port
References: <3CCB1CE9.3E08F027@ix.netcom.com>
Message-ID: <Xns91FE207CF4D44cliechtigmxnet@62.2.16.82>

Jive Dadson <jdadson at ix.netcom.com> wrote in 
news:3CCB1CE9.3E08F027 at ix.netcom.com:
> I want to configure Python so it will read and write standard input and
> output to and from either a TCP/IP socket or a serial COM port data
> link. There will be no terminal or keyboard.

my serial lib migh be of interest for you http://pyserial.sf.net
 
> I compiled Python under VC++ and used the Visual Studio debugger to step
> through the code, but as yet I have not even been able to find where in
> the code Python does input/output!

just assign your own file like object to sys.stdin,stderr and stdout. you 
don't even need a compiler for that...

for out and err you only need the "write" method. 
don't know whats all needed for the stdin object. at least "readline". i'd 
provide "read" and "readlines" too.
you might need to overload "input" and "raw_input" in the builtins too.

when you receive code from an other machine you might be interested in the 
rexec module, if security is an issue.

chris

-- 
Chris <cliechti at gmx.net>



From tejarex at yahoo.com  Tue Apr 16 22:50:56 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 17 Apr 2002 02:50:56 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu> <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <kq5v8.59911$3L2.5192980@bin7.nnrp.aus1.giganews.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
news:slrnabp9cs.ipr.philh at comuno.freeserve.co.uk...
> >Thanks for proposing this. I'm sure you'll get lots of horrified
> >reactions
>
> I don't see why: people who like pass can still use it.

I am not horrified, however, I do have two objections:

1. We will have two styles of Python to read: with and without pass.
I expect this would be confusing, especially since one would not know
for sure until one sees a pass.

2. Newcomers will wonder why we have it if it is not required.  Then
they will wonder "should I use it or not" and post requests for
advice.

It would have been ok with me, I think, if the language had allowed
null suites from the beginning, but it is now to late, practically
speaking, to remove pass.

Terry J. Reedy





From grante at visi.com  Mon Apr  8 13:55:38 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 08 Apr 2002 17:55:38 GMT
Subject: Reading/writing binary data.
References: <mailman.1018285938.26500.python-list@python.org>
Message-ID: <uQks8.20420$vm6.3149576@ruti.visi.com>

In article <mailman.1018285938.26500.python-list at python.org>, Jeff Smith wrote:

>> I've been searching the Python (1.5) docs, but am unable to find any
>> functions to read/write binary data from/to a file.  Under the "open"
>> function I see that you can append a "b" to the mode to specify binary
>> mode, but that doesn't help me any.
>> 
>> For example, if I want to take a list of numbers like [64,66,67] and
>> write them to a file as binary, how would I do that?  I also need to
>> be able to control whether the data is written in big endian or little
>> endian format.

> The array module has everything you need. You can use the tofile and
> fromfile methods to read and write binary data, the tolist and fromlist
> methods to switch data to/from lists, and the byteswap method for swapping
> between big- and little-endian.

The struct module also allows you to convert between binary and
"internal" formats.

-- 
Grant Edwards                   grante             Yow!  The Korean War must
                                  at               have been fun.
                               visi.com            


From cjensen at bioeng.ucsd.edu  Tue Apr  9 13:30:34 2002
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Tue, 09 Apr 2002 10:30:34 -0700
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com>
Message-ID: <3CB3253A.6060304@bioeng.ucsd.edu>

Dennis wrote:
> If there's a better forum for asking newb questions, please point me
> there!
> 
> I'm new to programming.  My only experience up to this point is some
> aborted attempts to learn C and Java.  I'm finding Python much easier
> to learn than those.  Currently, I'm pushing my way through
> "Non-Programmers Tutorial for Python" by Josh Cogliati.
> 
> My current problem is this:
> 
> x = 2
> 
> def wrong(num):
>     return num + 1
> 
> print wrong(x)
> print wrong(x)
> 
> I cannot figure out why this (simplified) program won't increment x
> and produce 3 and 4.  This is probably indicative of a bigger problem
> with understanding functions (and their limitations).  Are there any
> sources out there that explain functions more in-depth and clearly?! 
> Thanks for any and all help.
> 
> Dennis
> dennis at horsethieves.com


In this example, the problem is that you are not actualy modifying x. 
You probably want to do something like:

x = 2
x = wrong(x)
print x  #should print 3
x = wrong(x)
print x  # should print 4


In general you might want to look into "pass by reference" and "pass by 
value" function arguments.  This will help in understanding how function 
arguments are handled in general.  Though, Python has it's own argument 
passing mechanism generaly called "pass by object"  I suggest doing a 
groups.google search for these terms in the Python news group.


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



From offer at sgi.com  Mon Apr 15 17:26:20 2002
From: offer at sgi.com (Nhi Vanye)
Date: 15 Apr 2002 21:26:20 GMT
Subject: installation probs on SGI: _socket package removed.
References: <6a3d3529.0204151211.7f38f354@posting.google.com>
Message-ID: <a9fghs$rfu3t$1@fido.engr.sgi.com>

$ from ophirb at hotmail.com -#194894 | sed "1,$s/^/> /"
>
>
>When building python on SGI IRIX646 I've notice that although the
>_socket library compiled and linked ok, the 'make' process failed to
>import _socket and removed it.
>Following the advice I found here at the newsgroup I attempted to
>manually compile and link _socket.so, then tried importing it from
>python.
>The results: An exception was thrown saying the symbol __eprintf was
>not defined in the shared object.
>Has anyone encountered this problem and have any help on this?
>What library is the symbol __eprintf defined in?

Is this the SGI Freeware distribution or are you building from scratch ?

richard.

-- 

-----------------------------------------------------------------------
Richard Offer                         Technical Lead, Trust Technology.
"Specialization is for insects"
__________________________________________http://reality.sgi.com/offer/



From dmmcf at uiuc.edu  Tue Apr 16 20:18:44 2002
From: dmmcf at uiuc.edu (D. Michael McFarland)
Date: Tue, 16 Apr 2002 19:18:44 -0500
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu> <a9f9mq$lss$1@peabody.colorado.edu>
 <861ydg3ff3.fsf@haggis.aae.uiuc.edu> <3CBC5686.BCBCF663@noaa.gov>
 <a9holn$5t1$2@peabody.colorado.edu>
Message-ID: <86g01vrxcr.fsf@haggis.aae.uiuc.edu>

Fernando P?rez <fperez528 at yahoo.com> writes:

> This is another one for which weave can be a godsend. The blitz
> module in weave can for many numeric expressions generate equivalent
> C++ code which avoids all the intermediate copies. I don't want to
> sound like a cheerleader (I'm not a developer of weave, only a
> user), but weave is really amazing in how easily it allows you to
> integrate Numeric pyhton code with C snippets.

I played with weave 0.2.3 shortly after that version was released, but
never got weave.test() to run cleanly.  I said to myself, "Something
here has great promise," and went back to my old ways of doing
things.  Should I try the CVS version?

Re cheerleading, well, if the student in question hadn't listened to
my proselytizing about Python, he wouldn't have quite his current
problem.  (He'd have different and worse problems, but that's another
post.)

I've snipped away your mention of example files, but by all means,
please send me a few.

Cheers,
Michael


From guido at python.org  Sat Apr  6 22:54:05 2002
From: guido at python.org (Guido van Rossum)
Date: Sun, 07 Apr 2002 03:54:05 GMT
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com> <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <200204061913.g36JDcdj017764@ratthing-b246.strakt.com>  <200204062313.g36NDV428946@pcp742651pcs.reston01.va.comcast.net> <mailman.1018144574.9452.python-list@python.org>
Message-ID: <3CAFC3B1.70797E11@python.org>

[Guido]
> > Laura, you've said this over and over, and I don't buy it.
> >
> > Two arguments from XP apply here:
> >
> > - YAGNI (You Ain't Gonna Need It), or also "do the simplest thing that
> >   can possibly work".  If all you need today is a bool, use a bool.

[Laura]
> Aha, now I understand a large part of why we disagree.  A bool is not
> simpler than an int.  'Something that can change state' is simpler than
> 'something that can change state and must only have one of two values'.
> Do you go around making functions that return bool all the time, only
> then to change them to return int every time you have the idea that
> there actually are 3 states here you want to model?  Then you will
> be littered with one obsolete 2-valued function every time this happens.
> They pile up, and since your wretched customers have all started using
> the 2-value functions, you are stuck having to maintain these things,
> even though you now hate them, and want everybody to use the 3-value
> versions.

Sigh.  So much rhetoric, worthy a better cause.  Laura, you
can believe whatever you want, but I'll continue to define
functions that return bool whenever I think a bool makes sense.
I'm not responding to your diatribe about XP, which isn't
relevant (or if it is, you forgot to post the punch line).

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


From kamikaze at kuoi.asui.uidaho.edu  Wed Apr  3 21:48:37 2002
From: kamikaze at kuoi.asui.uidaho.edu (Mark 'Kamikaze' Hughes)
Date: 4 Apr 2002 02:48:37 GMT
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org>
Message-ID: <slrnaanfo4.1jcm.kamikaze@kuoi.asui.uidaho.edu>

Fri, 29 Mar 2002 17:39:14 -0800, Umesh Persad <upersad at tstt.net.tt> spake:
> felds and buttons. I see there are many GUI toolkits
> available. I am wondering which one should I use
> (tkinter, wxpython?) that will be simple enough
> to get me up and running qiuckly. Also are there
> any examples that I can look at? 

  If you want A) ease of installation on all platforms, and B) a really
good book on the subject, there is only one serious choice: Tkinter.
It's the standard, it's easy to work with, and John Grayson's book
_Python and Tkinter Programming_ is superb.  I've run my Umbra game
<http://kuoi.asui.uidaho.edu/~kamikaze/Umbra/> on every platform I could
get my hands on (Windows, Linux, Mac, FreeBSD, HP-UX, and Solaris), and
it just works (though I had to tweak the font sizes on Windows but not
on other platforms, I dunno why).

  If you're willing to accept some installation hassles on some
machines, PyQt is a superior GUI toolkit (Qt uses events and listeners,
which Qt call "signals" and "slots" just to be different, rather than
callbacks, and has better components), and the book _GUI Programming
with Python: QT Edition_ <http://www.opendocspublishing.com/pyqt/> seems
to be quite good so far (I'm still in the middle of reading it).

  wxPython has a handful of fanatical believers, but I was never able to
get it working on my Linux box, and the Mac and non-Linux Unix support
is basically nil.  I can't recommend it for serious use.

  The fourth option is PyGame <http://www.pygame.org/>, which is based
on SDL, is fairly portable and easy to install, and is nigh-perfect for
developing games in.  However, the documentation is not as good as it
should be, it has a fairly non-Pythonic interface, and it's not a good
general-purpose GUI solution.

  For your project in particular, you could get by just fine with
Tkinter, but you may want to take a look at PyGame.

> Also, if any one is familiar with Tkinter,
> how do I put up a window with a size
> that I want? Also how do I place a canvas
> on the window at a predefined size?

  Windows can be positioned with code like:
win.geometry("+%d+%d" % (x, y) )
  I'm sure there are other ways, but that's how I do it.  If you want to
resize it, use:
win.geometry("%dx%d+%d+%d" % (w, h, x, y) )

  Positioning a canvas, OTOH, I don't recall.  I know there are ways to
do absolute positioning, but I almost always use the packer or grid
layout managers.  I think this is covered in Grayson's book.

-- 
 <a href="http://kuoi.asui.uidaho.edu/~kamikaze/"> Mark Hughes </a>
"No one is safe.  We will print no letters to the editor.  We will give no
space to opposing points of view.  They are wrong.  The Underground Grammarian
is at war and will give the enemy nothing but battle." -TUG, v1n1


From pinard at iro.umontreal.ca  Mon Apr 29 10:19:40 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 10:19:40 -0400
Subject: Python Webpages
In-Reply-To: <slrnacq7jk.gav.jon+usenet@snowy.squish.net>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net>
	<aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no>
	<slrnacpc38.obc.jon+usenet@snowy.squish.net>
	<23891c90.0204290051.3edcc62d@posting.google.com>
	<slrnacq7jk.gav.jon+usenet@snowy.squish.net>
Message-ID: <oqadrmr3ib.fsf@titan.progiciels-bpi.ca>

[Jon Ribbens]

> Can you think of an example where the "extra power" provided by a more
> complicated templating system actually makes something easier to do than
> with jonpy? ;-)

To really reply, one would have to study `jonpy'.  Yet, I got from an earlier
message that `jonpy' does not allow arbitrary Python in the HTML pages.

Like many others, I made my own templating system, but did not overly
publish it as there are so many already.  Surely in my case, I would loose
big if I was not allowed to link Web pages with installed Python modules,
these take care of database access, specialised HTML formatting (that
would only encumber the templating language if it was done there), etc.

In fact, before writing my own, I glanced around to see what was written,
and my overall feeling was that systems were either weakish or over-bloated.
If you nicely marry Python with your templating system, you get an extremely
powerful templating system that is also very simple.  The power _is_ needed.

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




From seb_haase at yahoo.com  Thu Apr  4 20:08:38 2002
From: seb_haase at yahoo.com (Sebastian Haase)
Date: 4 Apr 2002 17:08:38 -0800
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <f445a2fe.0204041708.6fd6fd0d@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3n0wjdyzr.fsf at mira.informatik.hu-berlin.de>...
> seb_haase at yahoo.com (Sebastian Haase) writes:
> 
> > Solution 1)  -->> BAD!! <<-
>  [Solution 2)]
> > =====
> > The REAL 
> > Solution !  --> "we" are not really interested in debugging the python
> > lib !
> > ========    ==> we only want to debug _our_ own C/C++ code!!!
> 
> Solution 3) Don't define _DEBUG in your debug builds.
> 
> Regards,
> Martin

I think that wouldn't work - since I'm still interested in debugging
_my_ stuff - so that needs me to #define _DEBUG

BTW - I forgot to list as 1. c) that  starting from source leaves you
with an "uninitialized" registry - meaning other programms/modules
might think you don't have any python on your system... (not to
mention the missing start-menu entries ;-)

Regards,
Sebastian


From sholden at holdenweb.com  Mon Apr  1 07:56:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 07:56:06 -0500
Subject: PEP 285: Adding a bool type
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <mailman.1017469127.15681.python-list@python.org> <Pine.LNX.4.21.0203301456390.16587-100000@cens.ioc.ee>
Message-ID: <EOYp8.108214$Ou.56262@atlpnn01.usenetserver.com>

"Pearu Peterson" <pearu at cens.ioc.ee> wrote in message
news:Pine.LNX.4.21.0203301456390.16587-100000 at cens.ioc.ee...
>
> > ----- Original Message -----
> > From: "Guido van Rossum" <guido at python.org>
> >
> > >     Dear reviewers:
> > >
> > >     I'm particularly interested in hearing your opinion about the
> > >     following three issues:
> > >
> > >     1) Should this PEP be accepted at all.
>
> It seems that this is already decided as Guido is the author and also the
> only one who accepts PEPs.
>
That's not fair. Guido's record on accepting language change may not be
perfect, but the fact that he's prepared to submit a PEP rather than
institute a unilateral change is to be praised. Remember the division
brouhaha? As a result of user comments there were significant changes to the
proposed policy.

> > >     The standard bool type can also serve as a way to force a value to
> > >     be interpreted as a Boolean, which can be used to normalize
> > >     Boolean values.  Writing bool(x) is much clearer than "not not x"
> > >     and much more concise than
> > >         if x:
> > >             return 1
> > >         else:
> > >             return 0
>
> This one does not convince me at all as one could always define
>
>   bool = lambda x: not not x
>
> and I would never write such a verbose code fragment as above just to
> return whether x is false or true. However, using constructs like
>
>   if x:
>      ...
>
>   while x:
>      ...
>
> for x being a sequence or dictionary can be very useful and efficient.
>
> In future (Python 2.>=3, >3), are these constructs valid or must they
> be replaced with
>
>   while bool(x):
>     ...
>
> ? :(
>
If you seriously think that a proposal from the language's principal
designer would cause such massive code breakage, not to mention
contradicting almost every published Python book, then you aren't reading
the same newsgroup, or using the same language, as I am!

>
> Ka-Ping Yee has raised related conserns in
>
>   http://mail.python.org/pipermail/python-dev/2002-March/020893.html
>
> that never got a proper answer from Guido nor others.
>
Possibly so, but is seems absolutely clear from the posted proposal that the
answers to Ka-Ping's questions are "No", "No" and "No" respectively.

> So, I am (-1)*0 on this PEP.
>
> -1 --- I have succesfully programmed in Python without needing the bool
> function and I share Ka-Ping's concerns about Python future. To me the
> only reasonable point in rationale seems to be that "other languages have
> or will have boolean". That also sounds quite weak to me --- other
> languages have many useful concepts that would be more worth for
> introducing to Python than this boolean concept.
>
The most reasonable point in the rationale to *me* was the "experience from
teaching Python...".

> 0  --- this PEP will be hardly rejected, it seems to me.
>
Again, I think you're being unfair to Guido. If the PEP is accepted ist will
be because you are in a minority who think it will not result in improvement
to the language (or at least will not degrade it at all).

Did it catch you on a bad day?

regards
 Steve






From tim.one at comcast.net  Tue Apr 16 02:10:02 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 02:10:02 -0400
Subject: Why don't strings share data in Python?
In-Reply-To: <87d6x09trz.fsf@mathdogs.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEJMOOAA.tim.one@comcast.net>

[Mike Coleman]
> Does anyone know why strings (i.e., those of length >1) don't
> share their data in Python?

There are many reasons, but one of them is invisible to Python users:
because Python is intended to play nice with C libraries, all Python strings
end with a \0 byte.  There's nothing you can do in pure Python code to
confirm or deny that, but it's true anyway under the covers.  Tons of (C)
code relies on that for simplicity and speed.

Somtimes it peeks thru in a bug; e.g., this very old bug just got fixed a
day or two ago:

>>> complex("1+1j\0buncha junk")
(1+1j)
>>>

It reveals that the internal code that parses a string passed to complex()
mistakenly stopped at the first \0 byte instead of paying attention to the
*length* of the string (which Python stores separately from the string
guts).





From offer at sgi.com  Wed Apr  3 14:02:26 2002
From: offer at sgi.com (Nhi Vanye)
Date: 3 Apr 2002 19:02:26 GMT
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <slrnaaivjc.565.zeitlin@seth.lpthe.jussieu.fr> <slrnaaj2o9.a62.cpbotha@crabtree.cpbotha.net> <slrnaam8ss.4mg.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <a8fjk2$oavq0$1@fido.engr.sgi.com>

$ from zeitlin at dptmaths.ens-cachan.fr -#192302 | sed "1,$s/^/> /"
>
>
>> I did NOT say that wxWindows doesn't build on these architectures.
>
> Your statement implied that there were problems with building it under SGI
>and/or Solaris without saying what those problems were. I wanted to know more
>details about this.

I haven't looked at wxWindows for years, but here's an out-of-date opinion :-)

Back in '98-99 I went through an tidied the wxWindows code up to accept
the stricter requirements that the MIPS Pro compilers have on SGI (ie
identified the gcc specific behaviour and modified it/wrapped it with
ifdefs). And fed all my changes back.

Three months later it was broken again---in different places. At that
point I got distracted with real work and let it go.

One of the things that tk has (and Qt for that matter) is that it works
out of the box without using gcc as its platform compatability
layer. That's not an unreasonable approach---I've known commercial
products that took the same approach, but it can raise issues for those
who don't use gcc).


> Regards,
>VZ


richard.

-- 
Richard Offer                         Technical Lead, Trust Technology.
"Specialization is for insects"
__________________________________________http://reality.sgi.com/offer/



From stephen at xemacs.org  Sat Apr 13 03:32:15 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 13 Apr 2002 16:32:15 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Huaiyu" == Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> writes:

    Huaiyu> Whether they have changed the definition [of free
    Huaiyu> software] in recent years, or whether I have misunderstood
    Huaiyu> this issue from the beginning, I don't know.

You have misunderstood from the beginning.

rms's fundamental position is radical: property rights in software are
morally wrong.  Unfortunately, even rms himself seems incapable of
behaving in a way consonant with this belief.  (Ie, he clearly
believes that he has a special right to manage development of projects
he has initiated, such as Emacs, gcc, or glibc.  I imagine he would
clarify as the EU does, by distinguishing an author or maintainer's
_moral_ rights in a work from _economic_ rights, but I've never seen
him comment on precisely that issue.)  It's clearly a political
non-starter, too: more than most people, programmers tend to resist
alienating their rights in their work.  This is not surprising, given
that they _can_ have it and give it away, too.

For these reasons, rms's explanations of free software itself tend to
bounce off the minds of most people, and they get warped into
something else more familiar.

Developers tend to get quite confused.  They recognize that in the
real world, they are grateful for the use of lots of wonderful free
software, but on the other hand reaching high levels of quality in
their own projects requires resources that seem (and often are)
impossible to assemble without proprietary licensing.  rms is not
sympathetic; he feels developers are morally required to develop
alternative sources of revenue (cf. The GNU Manifesto for some
models).

    Huaiyu> So now my mental picture is 

    Huaiyu>    (copyleft) is subset of (free) is subset of (open source)

The second is not a proper inclusion.

The first is a proper inclusion.

[Aside: One of the litmus tests for membership in the "free software
movement" as opposed to the "open software movement" is preference for
copyleft over other forms of free software.  Both that and Bill
Gates's fear of the GPL are somewhat ironic, given that GPL is the
most anti-competitive of the common free software licenses.]


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From vinay_sajip at yahoo.co.uk  Fri Apr 12 13:11:14 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 12 Apr 2002 10:11:14 -0700
Subject: Logging Module 0.4.2 - Minor update to an HTML file
References: <2e37dc1.0204110722.311fe4b0@posting.google.com>
Message-ID: <2e37dc1.0204120911.24b8c005@posting.google.com>

Unaccountably, the copy of logging_pydoc.html released with 0.4.2 of
the logging system was corrupted so that it could not be viewed with
Netscape 4.7. I have posted an update, see

http://www.red-dove.com/logging_pydoc.html

People who install the module should not need it, as they can use
pydoc on logging.py itself. But it contains some minor mods which
facilitate links from the main page on logging, and in doing these
mods I must have inadvertently introduced the problem :-(. The updated
page looks OK (at least on Netscape 4.73/Win98).

Thanks to Joonas Paalasmaa for letting me know.

Regards

Vinay Sajip


From me at mikerobin.com  Tue Apr 16 00:31:02 2002
From: me at mikerobin.com (Michael Robin)
Date: Tue, 16 Apr 2002 04:31:02 GMT
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com>
Message-ID: <aONu8.50259$CH1.45659@sccrnsc02>

If think you could ask the same thing for tuples (also r/o), or even for
lists, etc., if
you had copy-on-write semantics. (Like a "hashed lisp".) I assume the answer
is that it would take
extra time to see if the string already existed, and space to hold pointers
to strings you had no way of knowing would ever need to be referenced again.
You can always intern() strings yourself if you know your particular program
will take advantage of multiple references.

m

"Mike Coleman" <mkc+dated+1021521407.f909ec at mathdogs.com> wrote in message
news:87d6x09trz.fsf at mathdogs.com...
> Does anyone know why strings (i.e., those of length >1) don't share their
data
> in Python?  Since their immutable, it seems like this would be the obvious
> thing to do.  So, for example, the space behavior of this code could be
linear
> rather than quadratic/horrific:
>
> d = {}
> for i in xrange(100000000):
>     d[mybigstring[i:]] = mybigstring[i:]
>
>
> Mike




From aleax at aleax.it  Thu Apr  4 05:05:05 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 10:05:05 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CAB16BA.EEED07DE@sympatico.ca> <j48z84zp5z.fsf@informatik.hu-berlin.de> <3CAB60E7.8BEF40DC@sympatico.ca> <1017910237.53325@robert>
Message-ID: <lzVq8.48605$pT1.1381394@news1.tin.it>

Frithiof Andreas Jensen wrote:
        ...
> If you really are not convinced, then run a Fortran program through f2c on
> a Linux box and gaze upon the resulting C code. Then try and modify that.

That's basically what I did many years ago to move cad.lab's main code
base from Fortran to C -- except that it was on Ultrix (one of Digital's
attempts at Unix:-), not Linux, and that of course I heavily hacked f2c
to suit our specific needs, the specific idioms we used, and so on.  The
result was quite good; substantial amounts of that code, both parts
that were later heavily modified and ones that didn't need to be, were
still in production when I left the firm (renamed in the meantime to
think3 -- www.think3.com) two months ago.  I still shudder to think how
much of the then-new-and-exciting language Perl I used to write various
auxiliary scripts for pre-processing and post-processing tasks...:-).

But the roughly 6 months it took me (with about 3 months' worth of help
from each of 3 other colleagues) still vastly beat the alternative of
recoding _every_thing (part of the push was porting to a platform that
lacked any real Fortran compiler).  At about 1 million lines of code,
roughly 10,000 function points, an optimistic estimate for recoding it
all would have been 5 man-years (at 1000 loc/day and 200 days/year),
while this way it took us little more than 1 man-year.

If I hadn't had f2c or something like it to start with, of course,
then the project would not even have begun (we also considered some
commercial alternatives, but f2c's open-source nature was a must to
enable me to perform the many necessary hacks).  I don't know of
anything equivalent for java->python migration.  But I couldn't let
this aspersion on f2c stand unchallenged...!-)



Alex



From chris.myers at ingenta.com  Thu Apr 25 16:17:38 2002
From: chris.myers at ingenta.com (Christopher Myers)
Date: Thu, 25 Apr 2002 16:17:38 -0400
Subject: Is there a __start__ method?
Message-ID: <3CC86462.72244834@ingenta.com>

I figured I'd try here first.

I have a gui app that I wrote mainly using Pmw with a few plain tkinter
widgets thrown in.  In my __init__method, I have another method that I
want to execute at startup that I call at the very end.  This method
sometimes takes quite a while to execute (it's accessing 3 separate
URL's), so the main window doesn't post until this is complete.  

What I want is something java-like, where there is an __init__ method
that builds the app, then a __start__ method (is that what it is in
Java?) is called which does other stuff on launch after the GUI is
built.

Anything like this available?

Thanks,
-- 
Christopher Myers, Graduate Software Developer 
Ingenta, Inc.
12 Bassett St.
Providence, RI  02903
ph:  401.331.2014 x 102
em:  chris.myers at ingenta.com
aim: chrismyers001


From DavidA at ActiveState.com  Thu Apr  4 19:09:04 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Thu, 04 Apr 2002 16:09:04 -0800
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de> <d7fcf8b4.0204041558.1ecf5f27@posting.google.com>
Message-ID: <3CACEB20.EE9E4918@activestate.com>

Just as an FYI, if you're debugging your extension modules and want the
debug libraries, you can get them as a downloadable add-on to
ActivePython using pyppm:

   pyppm install ActivePython-Debug

However, this will not solve the problem that if you have third-party
extension modules installed (Numeric, PIL, or whatnot), that you need
the debug versions of those as well.

Sam's/Tim's code should work if you don't want anyone's debug code but
your own.

--david




From aleax at aleax.it  Tue Apr 30 17:30:08 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 21:30:08 GMT
Subject: cPickle problem
References: <Ssiz8.63$bK3.1993791@newnews.cc.stevens-tech.edu> <FAiz8.78905$vF6.2347612@news2.tin.it> <bQDz8.64$y34.1997856@newnews.cc.stevens-tech.edu>
Message-ID: <A1Ez8.35824$8D3.1036311@news1.tin.it>

Vojin Jovanovic wrote:
        ...
> can I replace addequation with __setattr__ so that I can type
> 
>>>>k.c='self.a+self.b'
> 
> to achieve the same effect.

Yes.

> My attempts with __setattr__ were unsuccessful.

Probably because you kept using instructions such as self.y=4 in
other methods.  When you add __setattr__ you must also
translate such instructions into self.__dict__['y']=4 and the like.


Alex



From claird at starbase.neosoft.com  Fri Apr 19 14:03:10 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Apr 2002 13:03:10 -0500
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <a9pj29$pdm$1@nntp6.u.washington.edu>
Message-ID: <8EA7006D40B56719.DC7AB9A2B4317096.3110ECEC5F22E1FD@lp.airnews.net>

In article <a9pj29$pdm$1 at nntp6.u.washington.edu>,
Donn Cave <donn at u.washington.edu> wrote:
			.
			.
			.
>Part of the problem is that when you write something like a "grep"
>in Python and in Perl, the Perl program will naturally be written
>like while ($line = <STDIN>) {...}, and the Python program will
>naturally be written like while 1: line = sys.stdin.readline() ...
>That pits a lot of function calls against what must be an inline
>operation.  I think I decided that "I/O", in this practical sense
			.
			.
			.
Worse (or at least "more"):  Perl goes out of its way to optimize
  while ($line = <>) {...}
'Least, it has in the past; I haven't looked lately.
-- 

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


From jb at cascade-sys.com  Tue Apr 30 14:04:42 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 11:04:42 -0700
Subject: Third Draft:  Pep for Deprecating Builtins
References: <mailman.1020108957.30689.python-list@python.org> <YUvz8.12366$xy4.11240@atlpnn01.usenetserver.com>
Message-ID: <3CCEDCBA.5498152B@cascade-sys.com>

Steve Holden wrote:

> "Sean 'Shaleh' Perry" ?shalehperry at attbi.com? wrote in message
> news:mailman.1020108957.30689.python-list at python.org...
> [ ... ]
> ?
> ? The problem is compounded when you are not in control of the python on the
> ? system.  Imaging CGIs broken when the person's ISP decides to upgrade
> their
> ? machines to a new OS rev which has a new python.
> ?
>
> Thankfully, for the most part that's precisely what you have to do ...
> imagine them.

I think Sean makes a valid point and you're wrong to suggest the problem is
imaginary.

I happen to be developing an e-commerce CGI in Python on a machine where I don't
control the Python installation.  Who knows when some other customer will ask
the (somewhat clueless) IP to upgrade in a way that breaks my CGI?  I guess I
will have to go the extra lengths to protect myself but what a pain in the ass.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From tejarex at yahoo.com  Wed Apr 17 17:49:34 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 17 Apr 2002 21:49:34 GMT
Subject: preg_replace?
References: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE> <odlv8.24535$b62.683481@news1.tin.it> <a9kp6q$qch$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <O5mv8.21222$PV.2437888@bin6.nnrp.aus1.giganews.com>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9kp6q$qch$1 at nets3.rz.RWTH-Aachen.DE...
>. Now that I'm told the regex is correct another look makes
> it seem obvious.

The main possible glitch is nested brackets in the input.  You may or
may not like the result if there are.





From logiplexsoftware at earthlink.net  Mon Apr  1 14:29:07 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 11:29:07 -0800
Subject: GUI programming
In-Reply-To: <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net>
References: <mailman.1017437568.25402.python-list@python.org>
	<a84561$p5g$1@wanadoo.fr>
	<slrnaabbct.npb.cpbotha@crabtree.cpbotha.net>
	<4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com>
	<20020401020912.GE333@lilith.hqd-internal>
	<mailman.1017681605.21669.python-list@python.org>
	<slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net>
Message-ID: <20020401112907.3b0be834.logiplexsoftware@earthlink.net>

On Mon, 01 Apr 2002 19:04:16 GMT
Charl P. Botha wrote:

> In article <mailman.1017681605.21669.python-list at python.org>, Cliff Wells
wrote:

> > wxPython is remarkably easy to install on both Windows and RPM-based
Linux
> > systems.  I haven't tried it on anything else, but I doubt even a
> > tarball-based install would be very difficult.
> > 
> > Building from CVS would be more difficult, but that's rarely necessary.
> 
> Try getting a running wxPython installation on SGI Irix or SUN Solaris
for
> instance.
> 
> Secondly, why would building from CVS be more difficult than building
from
> tarballs?  The tarballs also contain oodles of source that have to be
> configured, built and installed.  If you haven't actually done this, you
> should give it a try before actually commenting. :)

I qualified my statements in what seemed a fairly clear manner, so I'm not
sure what your point is, except maybe to vent about your bad experience on
legacy platforms ;) 

Building from CVS would most likely (although not certainly) be more
difficult since the tarballs represent a _release_ and have at least been
tested on something besides Robin Dunn's PC, whereas the CVS represents
what Robin was doing last night and so may introduce bugs and/or
incompatibilities on a particular platform.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From wortelslaai at yahoo.com  Thu Apr 18 13:02:54 2002
From: wortelslaai at yahoo.com (Jaco Schieke)
Date: 18 Apr 2002 10:02:54 -0700
Subject: rlcompleter not calling __getattr__ on [ ]
Message-ID: <b5ddd682.0204180902.2866a9c0@posting.google.com>

Being new to interactive python, I have the problem that rlcompleter
does not seem to be able to resolve the items in a list:

This works as expected:
======================================
>>> import readline, atexit, rlcompleter
>>> readline.parse_and_bind('tab: complete')
>>> a=[[],"",1]
>>> a.<tab>
a.append   a.count    a.extend   a.index    a.insert   a.pop     
a.remove   a.reverse  a.sort


Now is this also supposed to work?
==================================
>>> a[0].<tab>
.   ..         <---- this is the output I receive
                     instead of the properties of an []


'cause this works:
==================================
>>> b=a[0]
>>> b.<tab>
b.append   b.count    b.extend   b.index    b.insert   b.pop     
b.remove   b.reverse  b.sort


From starx at pacbell.net  Wed Apr 17 06:21:46 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 17 Apr 2002 10:21:46 GMT
Subject: list_a-list_b
References: <mailman.1019022993.8455.python-list@python.org>
Message-ID: <Xns91F32263AA919RASXnewsDFE1@130.133.1.4>

Tim Peters || Tue 16 Apr 2002 10:54:56p:

> [David Eppstein]
>> ...
>> Better is to convert b to a dictionary
>> (where are dict comprehensions when you need them):
>>
>>     dict_b = {}
>>     for x in list_b: dict_b[x] = 1
> 
> Looks clear enough to me, although
> 
>     for x in list_b:
>         dict_b[x] = 1
> 
> is clearer.  If you like, you can do this:
> 
>     dict([(x, 1) for x in list_b])
> 
> or this:
> 
>     dict(zip(list_b, [1]*len(list_b)))
> 
> or even:
> 
>     dict(zip(list_b, list_b))
> 
> since you don't really care what the dict's values are, just the keys.
>  I draw the line before:
> 
>     dict(zip(*[list_b]*2))
> 
> though.  Contrary to popular belief, there's no Fabulous Prize
> awaiting those who minimize either vertical or horizontal space
> <wink>. 

Minimizing relative vertical space is great; that's why list 
comprehensions are better than nested for loops for expressing simple 
ideas ('i want var to be an x by y matrix of foozles' =), and why moving 
stuff into subfunctions is nice. Horizontal space though... feh. Unless 
you're using emacs default python mode, in a two or three level deep 
looping structre in a function in a class, and you need to write 
'self.data[x][y].variable = self.long_function_name( lots, of, arguments 
)' ... =)




-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From m_mariappanX at trillium.com  Wed Apr 24 09:11:23 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Wed, 24 Apr 2002 06:11:23 -0700
Subject: atoi
Message-ID: <53A7943A5BD8D411B6930002A5073155022D748F@bgsmsx90.iind.intel.com>

hi everybody.... thanks i was looking for int(string) or sting.atoi(strung)
routines only....

thx again, maharajan


>-----Original Message-----
>From: Artur Lipowski [mailto:lal at pro.onet.pl]
>Sent: Wednesday, April 24, 2002 6:25 PM
>To: python-list at python.org
>Subject: Re: atoi
>
>
>In article <mailman.1019649779.5458.python-list at python.org>, 
>Mariappan, MaharajanX wrote:
>...
>> is there any python buildin method whihc will convert 
>strings into integer,
>> like atoi in c lib.?
>> "234" should be converted into interger value 234
>...
>int() and company.
>Look at "Python Library Reference" chapter 2.1 "Built-in 
>Functions". Also 
>using searchable Python manual at 
>http://starship.python.net/crew/theller/pyhelp.cgi 
>with "conversion" keyword will find out what you need.
>In other words RTFM 8-)
>
>Regards,
>--
>Artur Lipowski
>-- 
>http://mail.python.org/mailman/listinfo/python-list
>




From duncan at NOSPAMrcp.co.uk  Wed Apr 10 05:12:15 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 10 Apr 2002 09:12:15 +0000 (UTC)
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org>
Message-ID: <Xns91EC67743463Bduncanrcpcouk@127.0.0.1>

Kalle Svensson <kalle at gnupung.net> wrote in 
news:mailman.1018391839.6535.python-list at python.org:

> [Lulu of the Lotus-Eaters]
>> Who can tell me what's in there.  Someone on c.l.py recently let drop
>> the hint about finding the Zen of Python at the interactive prompt in
>> 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
>> 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
>> magic command was.  Can someone remind me, please?
> 
> "import this", IIRC.
> 
The code that in this.py is a terrible advert for Python though. 
It does:
   d = {}
   for c in (65, 97):
       for i in range(26):
           d[chr(i+c)] = chr((i+13) % 26 + c)

   print "".join([d.get(c, c) for c in s])

when all it actually needs is:
   print s.decode('rot13')

Patch submitted to sourceforge.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From hoel at germanlloyd.org  Thu Apr 11 09:56:50 2002
From: hoel at germanlloyd.org (Berthold =?iso-8859-15?q?H=F6llmann?=)
Date: 11 Apr 2002 15:56:50 +0200
Subject: Strange segfault in extension
References: <sewuved1h7.fsf@pc970145.GermanLloyd.de>
Message-ID: <seu1qi5of1.fsf@pc970145.GermanLloyd.de>

hoel at germanlloyd.org (Berthold H?llmann) writes:

> I have an extension that, where calling a function, generates an
> segmentatin fault. This is nothing new for me, but I stumbled over the
> backtrace in gdb. Why is "args" getting the address 0x1? Is this
> possible?
> 
> Here is the top of my Backtrace
> 
> #0  0xfe4069f0 in LexSrt_method (self=0x0, args=0x1) at module/LexSrt.c:73
> #1  0xaea38 in PyCFunction_Call (func=0x1c43f0, arg=0x274c30, kw=0x0)
>     at Objects/methodobject.c:80
> #2  0x4abb4 in eval_frame (f=0x30d3f8) at Python/ceval.c:2004
> 
> Thanks for your help
> 
> Berthold

It was'nt that strange really. My module uses NumPy arrays and the
actual segfault was in a line

#0  0xfe726aa8 in LexSrt_method (self=0x0, args=0x2) at module/LexSrt.c:75
75              coor = (PyArrayObject*)PyArray_ContiguousFromObject

and my module was spread among some C source files. We are in the
process of switching from Python 1.5.2 to 2.2.1 and so we need to
define PY_ARRAY_UNIQUE_SYMBOL in our C files. Not that we didn't came
accross this problem earlier, but the symptom was different.

So this problem seems to be solved for now.

Greetings
-------------- next part --------------

Berthold
-- 
Dipl.-Ing. Berthold H?llmann   __   Address:       
hoel at germanlloyd.org        G /  \ L Germanischer Lloyd
phone: +49-40-36149-7374    -+----+- Vorsetzen 32/35    P.O.Box 111606
fax  : +49-40-36149-7320      \__/   D-20459 Hamburg    D-20416 Hamburg

From eddie at holyrood.ed.ac.uk  Fri Apr  5 06:14:31 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 5 Apr 2002 11:14:31 +0000 (UTC)
Subject: telnetlib question
References: <1017951810.138264@seven.kulnet.kuleuven.ac.be>
Message-ID: <a8k0un$r6p$1@scotsman.ed.ac.uk>

"NightOwl" <nightowl at uk2.net> writes:

>Hi,

>I would like to be able to read the output of the last command in a telnet
>session, without shutting it down.

>What I would like to do is

>- log in (can do that)
>- cat the contents of a couple of files (while the content <> searchstring)
>- if the search string is found, echo something.

>I managed to log in and cat a file, I managed to do read_all(), however,
>that outputs everything AND it seems like I needed to close the connection
>first. The doc's are not clear at all at this point. What function do I
>need?

>Can somebody give me some example code, or the correct function to read the
>output of the last command (the last cat in my case)?

>I found read_eager(), read_lazy etc, however, the explanation is very
>unclear to me!

Yup, those docs are somewhat on the terse side.

As a suggestion, what I normally use is read_until (prompt) to package up the
output from each command.  It's not so easy if the prompt keeps changing but
as long as there is something you know for sure you can match on (eg stick an
'echo blah' command after the cat).

I can e-mail you some examples if you want.

Eddie


From bokr at oz.net  Wed Apr 10 23:25:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Apr 2002 03:25:44 GMT
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu> <a92kfi$dhn$1@peabody.colorado.edu>
Message-ID: <a92vno$1o3$0@216.39.172.122>

On Wed, 10 Apr 2002 18:13:02 -0600, Fernando =?ISO-8859-1?Q?P=E9rez?= <fperez528 at yahoo.com> wrote:

>Joel Bender wrote:
>
>> I would like to process the elements of a list that match a condition,
>> then remove them.  I'll use integers in my example, but my list items
>> are actually objects.
>> 
>> What's the "right" way to do this?  For example:
>> 
>>     >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>>     >>> dothese = [i for i in lst if i < 5]
>>     >>> lst = [i for i in lst if i >= 5]
>>     >>> for item in dothese:
>>     ...     print item
>> 
>
>I don't know if it's 'right', but the following seems natural to me:
>
>In [36]: def flt(x):
>   ....:   if x<5:
>   ....:     print x
>   ....:     return 0
>   ....:   else:
>   ....:     return 1
>   ....:
>
>In [37]: lst
>Out[37]:
>[3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>
>In [38]: filter(flt,lst)
>3
>2
>1
>0
>4
>Out[38]:
>[5, 7, 8, 6, 9]
>
>You can define the flt() function to suit your taste, add parameters to it, 
>etc.
>
>It's single pass, goes in the right order and returns the cleaned list. The 
>only problem it has is that it generates a secondary list instead of 
>modifying the existing one in-place. You're the only one who knows if that's 
>a problem or not (it basically depends on the size of your lists and if you 
>can afford the extra copy, even if just temporarily).
>
You might repack this way (it feels C-ish though ;-):

 >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
 >>> i=0
 >>> for j in xrange(len(lst)):
 ...     if lst[j]<5:
 ...         print lst[j]
 ...     else:
 ...         lst[i] = lst[j]
 ...         i += 1
 ...
 3
 2
 1
 0
 4
 >>> lst[i:]=[]
 >>> lst
 [5, 7, 8, 6, 9]

I'm not sure if the slice assignment is done in place,
but I think it could be.


Regards,
Bengt Richter


From phd at phd.pp.ru  Tue Apr  9 04:00:27 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 12:00:27 +0400
Subject: What has PEP 285 done to us?
In-Reply-To: <20020408134207.1af35256.logiplexsoftware@earthlink.net>; from logiplexsoftware@earthlink.net on Mon, Apr 08, 2002 at 01:42:07PM -0700
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net>
Message-ID: <20020409120027.B8297@phd.pp.ru>

On Mon, Apr 08, 2002 at 01:42:07PM -0700, Cliff Wells wrote:
> I have to say that the most damaging thing introduced by PEP 285 has little to
> do with the technical merits of the discussion.  The level of respect accorded
> others in this community has taken the most damage.  It seems that almost every
> discussion on this topic has quickly gone down in flames.

   I do not think the PEP *did* damage - it only revealed that our
community has problems. I don't understand what problems.

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




From sholden at holdenweb.com  Tue Apr 16 16:19:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 16:19:16 -0400
Subject: Help me learn Python !
References: <4t7pbu4u83fe7et66lqdvhobq0hv975rn0@4ax.com>
Message-ID: <3J%u8.868$hl2.215@atlpnn01.usenetserver.com>

"peter" <REMOVEpeterp100 at hotmail.com> wrote ...
> I'm working out of the Python Web Programming book.  I'm trying to
> figure out how to D/L some of the modules that I need to get my code
> to work (like: lrwp, lrwplib, etc) ???
>
> I'm been to http://www.freecode.com/projects/xitamiwebserver/ and have
> copied some of the module code, but the indentations are all wrong, so
> it doesn't work.  Is there anyway to download all the modules (in
> their correct format).
>
>
> I'm using Windows 2000, and the Xitami webserver.
>
>
I don't ever remember having these problems on my Win98 system. Although
I've now got Win2k I have not yet tried any LRWP stuff. Hold on ...

Nope, don't see any indentation errors -- are you sure you haven't mangled
the code? Note, by the way, you need to change the call

            self.sock.connect(self.host, self.port)

to

            self.sock.connect((self.host, self.port))

before it will run - a change in the library brought it into line with the
documentation, breaking code that took advantage of the opriginal bug. For
the record, I just ran the code and everything works fine: I get five
responses from testlrwp (which responds at "/testapp1"), and then "The URL
you asked for does not exist on this website". All as expected.

Hope this helps.

regards
 Steve






From gimbo at ftech.net  Wed Apr 17 12:38:18 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Wed, 17 Apr 2002 17:38:18 +0100
Subject: Readline/command history w/ interactive prompt
Message-ID: <20020417163818.GA19289@andy.tynant.ftech.net>

Just a thought...

One of the things I find "warty" about the interactive prompt is that
the command-line history retains multiple copies of identical
successive lines, rather than trimming them down to a single instance.

Slightly fictional example:

>>> import my_mysql_wrapper
>>> d = my_mysql_wrapper.connection(parms)
>>> d.select('select count(*) from some_table')
>>> reload(my_mysql_wrapper)
<module 'my_mysql_wrapper' from 'my_mysql_wrapper.py'>
>>> d.select('select count(*) from some_table')
>>> d.select('select count(*) from some_table')
>>> d.select('select count(*) from some_table')
>>> d.select('select count(*) from some_table')
>>> d.select('select count(*) from some_table')
... another 15 instances of that same .select() call

At this point, I go and fix a bug in my_mysql_wrapper, so I want to
reload it.  Alas, I have to hit the up arrow 20 times to make that
happen, and I have to watch carefully to see when I get the line I
want.  If only it could be done with a mere 2 hits instead...

Now, I'm thinking this is _probably_ a limitation of GNU readline, not
python per se, because I notice this behaviour in bash too.  Can
anyone tell me if this is the case or not?  And is there any good
reason for having this behaviour?  It seems well crufty to me - the
Amiga shell did this right, why can't GNU/we?  ;-)

Hmmm... /me wonders if python on the Amiga gets this right...  :-)

As I said, just a thought...

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From niemeyer at conectiva.com  Mon Apr  1 10:12:04 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Mon, 1 Apr 2002 12:12:04 -0300
Subject: PEP 285: Adding a bool type
In-Reply-To: <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020401121204.A9299@ibook.distro.conectiva>

Hi Martin!

> The standard library has 110 occurrences of "return 0", 70 occurrences
> of "return 1", and 136 occurrences of "return None". Optimistically,
> this makes 300 functions. It would help users of these 300 functions
> if they would know whether those functions return a boolean or an
> integral value (if integral, what range?); for the "return None"
> cases, the boolean type would help to clarify whether this is "return
> no value", "return False", or "return nil object";

I hope nobody is planning to change "return None" to "return False".
This would break a lot of code doing "ret == None".

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]



From brian_l at yahoo.com  Thu Apr 18 03:19:37 2002
From: brian_l at yahoo.com (Brian Lenihan)
Date: 18 Apr 2002 00:19:37 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz> <3CBC5C04.8F2C3D5A@noaa.gov> <afe3b146.0204171407.6e5e4b0e@posting.google.com> <ad053a43.0204171823.54f54b99@posting.google.com>
Message-ID: <afe3b146.0204172319.20022281@posting.google.com>

tony-clpy at lownds.com (Tony Lownds) wrote in message news:<ad053a43.0204171823.54f54b99 at posting.google.com>...
> > Tony Lownds has instructions here:
> > http://mail.python.org/pipermail/pythonmac-sig/2002-April/005420.html
> > 
> 
> Here too: http://tony.lownds.com/macosx/tkinter.html
> 
> Someday Tkinter and IDLE will work nicely OOTB on OSX!
> 
> > AquaTk seems to have some problems with path separators (try file open
> > in IDLE for an example).
> 
> 
> AquaTk's File open/save doesn't handle unix paths well. Since python
> on Darwin (a.k.a. MachoPython) uses unix paths natively, IDLE's Open
> and Save methods don't work. I believe this is posted as a bug in Tk's
> Tracker, but it could be fixed in Python code too.

Yes, we ran into the same problem while porting our build system to OS
X.  To make the transition easier, the build system needs to be able
to run on both Mach-o and Carbon Python depending on the context.

We wanted to avoid changing the Applescripts which create and run
Codewarrior projects because we think we will be able to avoid
Applescript completely in the near future, assuming there is a version
of CW which can do Carbon builds from the command line.

The easiest solution was write some functions which always create a
correct path for the given environment, even when the input contains
both Unix and Mac path seps.

We generally try to avoid overriding things like os.path.join to avoid
any chance things breakng if we upgrade Python.  os.path.join is not a
very good example as I doubt it is going to move to a different module
or disappear anytime soon.  A better example is some of the constants
in the fcntl module which went away in 2.2 because they are already
defined in the os module.


From geoff at gerrietts.net  Tue Apr 16 16:42:52 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 16 Apr 2002 13:42:52 -0700
Subject: Newbie problem: global variables
In-Reply-To: <a9i1uk$1is$1@narses.hrz.tu-chemnitz.de>
References: <a9i1uk$1is$1@narses.hrz.tu-chemnitz.de>
Message-ID: <20020416204252.GC20804@isis.gerrietts.net>

Quoting Mirko Opitz (opm at hrz.tu-chemnitz.de):
> Hi,
> 
> I've a small problem I need a variable which is present in all subfunctions
> an also in the main prog but I don't want do give it in the header of all
> the functions. Is there anyway to declaire it as Public or global like in
> other languages?

def funcy(arg):
  global my_global
  my_global = arg

def not_so_funcy():
  global my_global
  print my_global

funcy("spam")
no_so_funcy()


Do what you want?

--G.
 

-- 
Geoff Gerrietts <geoff at gerrietts dot net>      http://www.gerrietts.net/
   "If I were two-faced, would I be wearing this one?" --Abraham Lincoln




From aahz at pythoncraft.com  Thu Apr 11 21:15:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 21:15:04 -0400
Subject: old vs new iteration protocols (was Re: Booleans, &c)
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <C%Xs8.3139$b62.77265@news1.tin.it> <a925qq$ne4$1@panix1.panix.com> <b9ct8.6406$vF6.199796@news2.tin.it>
Message-ID: <a95ceo$9oo$1@panix1.panix.com>

In article <b9ct8.6406$vF6.199796 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>Aahz wrote:
>> In article <C%Xs8.3139$b62.77265 at news1.tin.it>,
>> Alex Martelli  <aleax at aleax.it> wrote:
>>>
>>>With the previous iterator protocol, "for item in somedict:" COULD
>>>NOT "work smoothly and seamlessly".  Nor, at all.  Think about it.
>> 
>> Oh, come on, Alex; with your twisted imagination, I'm sure you can think
>> of ways in which a dictionary could be forced to work with the old
>> iterator protocol -- it would just be very, very messy.
>
>Twisted imagination, me?  I'm a plodding infantry type and I take this
>as a challenge.  How does e.g, somedict[0] return some "first thingy
>in somedict" AND at the same time raise KeyError if 0 ain't a key in
>somedict, just to start with?  The latter is part of the fundamental
>social contract of dictionaries -- the former is part of the old
>iteration protocol.

<blink>  You're right, never mind.  I was getting things confused in my
head.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From phr-n2002a at nightsong.com  Sun Apr  7 01:15:56 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 06 Apr 2002 22:15:56 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1018153694.532.python-list@python.org>
Message-ID: <7xelhsvxsz.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> > Doesn't matter what you call it, if a 2.1 script can't run at all or as
> > expected on 2.0, that is a major change in the _colloquial_ sense.
> 
> I expect you phrased that backwards, yes?  If you use a new-in-2.1
> feature, then of course it's not going to run under 2.0.  The other
> direction is almost always tractable, although in some cases it does
> take foresight, or the active application of hindsight, to pull it off.

I don't think he phrased it backwards.  "If you use a new-in-2.1
feature" presupposes that 2.1 has features that aren't in 2.0, and I
think James is arguing that it should not.  The issue is just how
often you're willing to tell people they have to upgrade.

If I'm running 2.0, and 2.1 comes out, it's pretty similar, so I don't
upgrade.  Then someone sends me a script that using some new-in-2.1
feature and I have to upgrade.  Then 2.2 comes out with new features
of its own, someone sends me a 2.2 script and I have to upgrade again.
Next, 2.3 comes out and I want to run a script that uses bools.
Whoops, another upgrade.  It gets tedious!  I can deal that every
18-24 months (i.e. when I buy a new computer) but I don't want to do
it every 3-6 months.  Of course if I was really serious about security
I'd have to do a line by line code review of every one of those damn
releases before installing it, making it even worse.  As it is, I
mostly just shut my eyes and install, but I feel like I'm stepping off
into a vast and dangerous unknown every time I do it.

The stability-valuing approach would be DON'T PUT NEW FEATURES into
releases like 2.1 without good reason to get the particular features
out fast.  New library modules, or enhancements of existing ones, are
generally ok (I mean stuff like network protocols, not string
methods).  Bug fixes are ok.  Stuff like bools (even if you think
they're a good idea) should be added only in major releases, not minor
ones like 2.3.  Stuff like dict comprehensions should be added at the
same time as list comprehensions, or not at all; again, not in a minor
release.  If 3.0 comes out and it has a lot of cool new stuff that's
not in 2.0, that's great, I'll upgrade, but I want the new features
and upgrade pain in a single release, not drop-drop-drop like the
Chinese water torture.

The proposal of having separate "development" and "stable" release
tracks like Linux has sounds like a promising direction for keeping
both camps (language development enthusiasts and would-be production
users) happier than they are in the current situation.


From aahz at pythoncraft.com  Mon Apr  1 09:24:44 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 09:24:44 -0500
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <slrnaacavs.2gs.philh@comuno.freeserve.co.uk> <a85od6$9i$1@panix1.panix.com> <DlTp8.48032$S52.1614804@news2.tin.it>
Message-ID: <a89qjc$ah1$1@panix1.panix.com>

In article <DlTp8.48032$S52.1614804 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>Actually, this specific problem is there, like several others, only
>*for MUTABLE objects*.  Take an "immutable" rectangle:
>
>class ImmutableRectangle(object):
>    def __init__(self, width, heigth):
>        object.__setattr__(self, 'width', width)
>        object.__setattr__(self, 'heighth', heigth)
>    def __setattr__(self, name, value):
>        raise TypeError, "cannot mutate ImmutableRectangle"
>    def __hash__(self):
>        return hash(self.width) ^ hash(self.heigth)
>    def area(self):
>        return self.width*self.heigth
>    def perimeter(self):
>        return 2*(self.width+self.heigth)
>    # etc, etc
>
>Now, subclassing this is perfectly fine:
>
>class ImmutableSquare(ImmutableRectangle):
>    def __init__(self, side):
>        ImmutableRectangle.__init__(self, side, side)

Where is the side attribute?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From aahz at pythoncraft.com  Mon Apr  8 15:51:33 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Apr 2002 15:51:33 -0400
Subject: Never say never again (was Re: PEP 285: Adding a bool type)
References: <mailman.1017466974.6751.python-list@python.org> <a8sf0k$2a5$0@216.39.172.122> <ahjs8.73608$S52.2597704@news2.tin.it> <a8sn8u$sj7$0@216.39.172.122>
Message-ID: <a8ssc5$gqo$1@panix1.panix.com>

In article <a8sn8u$sj7$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>
>Well, I guess that depends on how you define "assignment" in Python.
>ISTM there really is no such thing, in the sense of updating a named
>storage place with new content, so what goes by the name "assignment"
>needs some explanation. If the statement a=b is an assignment,
>then binding the name 'a' to object b is the meaning of Python
>assignment as I see it. 

I define "assignment" in Python as being precisely equal to "=".
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From h.e.w.frater at cs.cf.ac.uk  Wed Apr 24 20:43:46 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: Thu, 25 Apr 2002 00:43:46 +0000
Subject: sql statement generation problem
Message-ID: <aa7g4o$djf$1@mannews.swan.ac.uk>

when using the following line to create an sql statement:  query2 =    
"select * from tblitem where itemtypeid = %(itemTypeID)s and %(searchIn)s
like '%(query)s'" % sqlValues    

it works, but it need to have this  like '%%(query)s%' so that oracle will
 perform the search properly. However, the extra %'s seem to mess things  
up. Is there any way to get round this. Also, DCOracle2 is returning 1.0 
for a value of 1 in the database. This is totally messing things up, and I
need to find a way round this.

Any ideas,
Hugh



From aahz at pythoncraft.com  Sat Apr  6 09:42:07 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Apr 2002 09:42:07 -0500
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <a8n1fv$1p4$1@panix1.panix.com>

In article <uDDr8.66496$S52.2342405 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>Linus Thorvalds seems to have done pretty well with picking TWO middle 
>grounds -- two parallel tracks for Linux, "stable" and "experimental".
>
>Without the former he'd have lost the commercial interests, and "normal
>people" who DO need SOME level of stability; without the latter, a crucial 
>hard-core of neophile hackers might have gone away in search of other, more 
>dynamic projects.  With both tracks, things seems to have been smoother.
>
>Why don't we start thinking of a similarly dual-tracked Python -- one
>track aiming at stability (in a "middle ground" sensible sense: strong
>commitment to not breaking old code -- introducing small new features
>and new library modules _is_ OK, else one can use a bugfix-only-"track"
>for some previous release), one aiming at frequent releases, innovation,
>etc.  

PEP 6 does allow for the possibility of a non-bugfix patch release, but
I think the Python developer community is still too small to permit a
full two-track process.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From phr-n2002a at nightsong.com  Wed Apr 24 19:42:17 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 16:42:17 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com>
Message-ID: <7xadrsslye.fsf@ruckus.brouhaha.com>

Richard Parker <richard at electrophobia.com> writes:
> > How about if I just hash a second time, i.e. H(H(K || x)) or
> > H(K || H(K || X)), rather than coding up the full HMAC machinery?
> 
> H(K1 || H(K2 || X)) is secure.  The key-length equivalent security of this
> construction is given by the length of one key, not both.  This construction
> is essentially HMAC, except that the HMAC construction simplifies the keying
> and makes the key-length equivalent security more obvious by adding a simple
> key-separation technique to derive both keys from a single key.

Do the keys really need to be independent?
How is H(K || (H || 'a' || K || X))?  

I read the HMAC paper years ago and don't remember much, except that
the reasoning given for every detail of HMAC including the magic
constants was fairly elaborate.  Some the attacks these details were
trying to stop didn't seem practical though.

Given the not-so-hot SHA-OFB encryption that I'm using, I'm mostly
only worried about practical attacks, though I did switch the MAC to
use encrypt-then-MAC since that didn't slow the function down noticably.

Thanks

Paul


From skip at pobox.com  Tue Apr 16 16:41:19 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 16 Apr 2002 15:41:19 -0500
Subject: Unicode name questions
Message-ID: <15548.35951.239060.214140@12-248-41-177.client.attbi.com>

I'm trying to build a dictionary that maps HTML entities to the appropriate
Unicode string literals.  I found three entities whose purported Unicode
names, when used, generate 

    UnicodeError: Unicode-Escape decoding error: unknown Unicode character name

They are:

    _entities['&Lambda;'] = u"\N{GREEK CAPITAL LETTER LAMBDA}"
    _entities['&lambda;'] = u"\N{GREEK SMALL LETTER LAMBDA}"
    _entities['&thetasym;'] = u"\N{GREEK SMALL LETTER THETA SYMBOL}"

Have I got the names wrong or are there some missing character names in
Python's Unicode support?

Thanks,

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Sun Apr 28 15:17:36 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 28 Apr 2002 19:17:36 GMT
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com> <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com>
Message-ID: <aahhsg$a1k$2@nyheter.chalmers.se>

In article <z4Xy8.72324$z55.71416 at atlpnn01.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>"J.Jacob" <joost_jacob at hotmail.com> wrote in message
>news:13285ea2.0204270857.2f34f8a4 at posting.google.com...
... Complicated suggestion by Steve deleted.
>>
>> # file fortune.py
>> import random, os
>>
>> def randlistindex(yourlist):
>>     "Return a random index number into yourlist"
>>     return random.randint(0, len(yourlist)-1)
>>
>> def quote():
>>     """
>>     Return a quote <string> from file allfortunes,
>>     assume the file allfortunes is located in the same
>>     directory as this module."""
>>     datafile = os.path.join(
>>       os.path.dirname(__import__(__name__).__file__),
>>       'allfortunes' )
>>     quotesfile = open(datafile)
>>     quoteslist, quote = [], ''
>>     for s in quotesfile.readlines():
>>         if s[0] == '%':
>>             if quote:
>>                 quoteslist.append(quote)
>>                 quote = ''
>>         else: quote = quote + s
>>     if quote: quoteslist.append(quote)
>>     print quoteslist[randlistindex(quoteslist)],

Simple suggestion by me added:

1. Get the size of the file.
2. Get a random number 0 - size-1.
3. Index to this point in the file.
4. Iterate back to the nearest previous '%'.
5. Print fortune from there.

This will of course skew the probability of getting printed towards longer
cookies, but nobody will notice that.

Jacob Hall?n


-- 


From flz at cfdrc.com  Mon Apr  1 17:14:57 2002
From: flz at cfdrc.com (gother)
Date: 1 Apr 2002 14:14:57 -0800
Subject: Initialization problem of shared library in HP-UX11
References: <7aafbf19.0204010929.62d17d6a@posting.google.com> <m37knr71pl.fsf@mira.informatik.hu-berlin.de>
Message-ID: <7aafbf19.0204011414.25acdd0@posting.google.com>

This works in other platform. This is other thing. Another thing is
the global variables with used-defined class type belong to the big
software package. We only export a few functions to the python and
cannot modify the original package. Sometime, we even don't know how
many this kind of variables inside the packages. The global variables
are used inside the package, not explcitly used by the function we
exported to python. So no way we can put those variables inside the
module's init function.

I know windows do the initialization when it load the shared lib. In
HP, when the shared lib links to an main function, it do the same
thing. But somehow HP-UX11 failed to do it inside the python which
called shl_load function. So I guess inside the executable the shared
lib is loaded with different system function call instead of shl_load
or different arguments of shl_load. If anybody be familar with the
internal loading process of HP, he must have the answer.

Thanks of Martin's reply  


martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m37knr71pl.fsf at mira.informatik.hu-berlin.de>...
> flz at cfdrc.com (gother) writes:
> 
> > Anybody know how to resolve this problem?
> 
> I recommend to initialize the variable inside the module's init
> function.
> 
> Regards,
> Martin


From loewis at informatik.hu-berlin.de  Fri Apr 19 12:58:01 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Apr 2002 18:58:01 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019230778.30753.python-list@python.org>
Message-ID: <j4y9fjsk12.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> 20 tests were skipped but that would be less if I could enable that
> regrtest.py thing you told me about.  Whereabouts is the correct place
> to add the options to enable those tests?  

Neither, nor: invoke regrtest with appropriate options.

Regards,
Martin


From yaser at lambenttek.com  Wed Apr 24 03:06:15 2002
From: yaser at lambenttek.com (Yaser Ansari)
Date: Wed, 24 Apr 2002 12:36:15 +0530
Subject: (no subject)
Message-ID: <DD0D2E55A4BC0A4DBE4D125B2867A0DB141C0C@lambentserver.LAMBENT.COM>

 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020424/8bc8aed4/attachment.html>

From mpoulin at verinet.com  Thu Apr  4 09:06:32 2002
From: mpoulin at verinet.com (Marc Poulin)
Date: Thu, 04 Apr 2002 09:06:32 -0500
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org> <3CAC19D8.142A918@sympatico.ca> <7xhemru1pg.fsf@ruckus.brouhaha.com>
Message-ID: <3cac7941$0$69916$7586b60c@news.frii.com>

In article <7xhemru1pg.fsf at ruckus.brouhaha.com>, "Paul Rubin"
<phr-n2002a at nightsong.com> wrote:

> This makes no sense.  WHY do you want to migrate all that Java to
> Python? Why not leave it in Java and keep using it?  That's what none of
> us are understanding.  What do you GAIN from making an unmaintainable
> machine translation?

And how much will it cost?
And who's paying for it?


From segreteria at studiomasson.it  Tue Apr 30 10:30:52 2002
From: segreteria at studiomasson.it (Studio Masson)
Date: Tue, 30 Apr 2002 16:30:52 +0200
Subject: Problem with module
Message-ID: <aam9sa$bvfkp$1@ID-131182.news.dfncis.de>

Hello everybody.
I have a strange "problem" (well, it's not a real problem but...). I'm
developing a program in Python (2.2) and wxPython (well, it's my first
python program...). It works as i expect on two pc, one with Windows XP
professional, one with Windows 98. I simply "copy" the directory where the
sources are, and then i work, using IDLE. The program is always into the
same directory ("c:\dati\python") in all the computers. If i do the same
(copy the program's directory) on a windows ME system, and try to run the
program (CTRL-F5 with IDLE) it stops with the following message:

>>>
Traceback (most recent call last):
  File "C:\Dati\Python\PrimaNota\SMPNotaMain.py", line 1200, in ?
    PrimaNota = PNotaApp(0)
  File "C:\PYTHON22\lib\site-packages\wxPython\wx.py", line 1587, in
__init__
    _wxStart(self.OnInit)
  File "C:\Dati\Python\PrimaNota\SMPNotaMain.py", line 1193, in OnInit
    self.frame = MainFrame()
  File "C:\Dati\Python\PrimaNota\SMPNotaMain.py", line 90, in __init__
    self.PNDBCorrente = MemoryDB.MemoryDB()
AttributeError: 'module' object has no attribute 'MemoryDB'

>>>

I repeat, the exact same source works in both other computers!!! The module
"MemoryDB" exists, and is in the same subdir of the main script. It's
imported simply with "import MemoryDB".
The most strange thing is that if i "double-click" on the program's icon, it
still works correctly!
Anyone has some suggestion on how to run the program whitout leavin' the
IDLE on a Win ME system???

Thanks.

p.s. i have set a PYTHONPATH environment variable to c:\dati\python, and
added c:\python22 to the path in all the systems. However, this shouldn't be
necessary, right... ?




From loewis at informatik.hu-berlin.de  Fri Apr 19 11:13:22 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 19 Apr 2002 17:13:22 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019220338.385.python-list@python.org>
Message-ID: <j4lmbju3fx.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> I remembered: I forgot to uncomment all the places where the Tk stuff
> lives.
[...]
> # *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
> 	-L/usr/local/lib \
[...]
> ld.so.1: ./python: fatal: libtk8.3.so: open failed: No such file or directory

That is an old Solaris problem: the dynamic loader searches on /lib
and /usr/lib for shared libraries, no matter what directories where
given with -L to the compiler. You have the following options (again
all alternative):

1. Set LD_LIBRARY_PATH when running Python. This is most common, but
   also ugliest, since it requires every Python user to set
   LD_LIBRARY_PATH.
2. Pass -R/usr/local/lib when linking python; you can do that by adding
   this into Modules/Setup. Alternative spellings of this option are
   -Wl,-R/usr/local/lib, 
   -Xlinker -R/usr/local/lib
   -Wl,--rpath /usr/local/lib
   -Xlinker --rpath /usr/local/lib
3. Set LD_RUN_PATH when linking python; ld(1) will treat LD_RUN_PATH
   as implicitly specifying -R options.
4. Use crle(1) to extend the system-wide list of directories searched
   for shared libraries; you will need to do that on all systems where
   you run the python executable.
5. Recompile Tcl and Tk as static libraries (or explicitly request
   that the static versions - which you already have - are used in
   linking; you do that by giving the full .a file path, instead of
   using -l options)

Contrary to my earlier claims: dependence on a shared library *is* a
good reason to build an extension as a shared library also. Currently,
each start of python needs to locate libtk8.3.so. If _tkinter.so was a
shared module, then libtk8.3.so would be located only if _tkinter is
imported.

I'd personally go with option 2, and build _tkinter.so, or go with
option 5.

HTH,
Martin


From andrew at acooke.org  Fri Apr 26 22:49:54 2002
From: andrew at acooke.org (andrew cooke)
Date: 26 Apr 2002 19:49:54 -0700
Subject: XML, Python, XML-SIG, documentation, PyXml, 4
References: <aef33705.0204231902.40694af4@posting.google.com> <m3znzth8fy.fsf@mira.informatik.hu-berlin.de>
Message-ID: <aef33705.0204261849.83541b3@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m3znzth8fy.fsf at mira.informatik.hu-berlin.de>...
> andrew at acooke.org (andrew cooke) writes:
> > The web pages at http://pyxml.sourceforge.net/topics/ seems to be out
> > of date.  In particular, the download talks about 0.6.6 when at
[...]
> No, that version is 0.7.0. Using 0.6.6 is still reasonable. I don't
> think the Web pages are out of date.
[...]

Rather than argue about whether or not someone else's work is any
good, I've written my own page that might be useful to others:
http://www.acooke.org/andrew/writing/python-xml.html

Cheers,
Andrew


From lumber at jack.com  Thu Apr 11 12:19:29 2002
From: lumber at jack.com (Lumberjack)
Date: 11 Apr 2002 16:19:29 GMT
Subject: Notes on 2.2's changes
References: <slrnabba3h.h8h.akuchlin@crystal.mems-exchange.org>
Message-ID: <Xns91ED5ED8EF7A0lumberjack@216.148.53.96>

"A.M. Kuchling" <akuchlin at crystal.mems-exchange.org> wrote:
> This means that from my point of view Python *has* been managed for
> stability, and the job has been done reasonably well.  (I'd be
> interested in seeing a similar list from people who find there's too
> much change.  It would be useful for figuring out what sorts of
> changes cause the most problems.)

Sometimes a single bug can take several man days to fix or cause financial 
hardship. It is impossible to tell from your list how much _time_ (and 
money) was consumed in locating and fixing those problems. Basically, you 
haven't provided data on the more important metric. And it isn't clear why 
you felt it was necessary to be compatible with 2.2. Wouldn't it have been 
simpler to just say "run these on version 2.0.x of Python"? The alleged 
underlying philosophy seems to be that you can pick whatever version 
that suits you and stick with it; you are under no compunction to track 
changes to Python. So why track it if there is no compunction? I'd like to 
see you address that.

P.S. Frankly I'm surprised that the list was as long as it was.


From robin at jessikat.fsnet.co.uk  Tue Apr 30 14:36:13 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 30 Apr 2002 19:36:13 +0100
Subject: Is it me or a python wart
References: <C2cBaXANXSy8Ewbb@jessikat.demon.co.uk>
 <just-9E8624.12281026042002@news1.xs4all.nl>
 <ucjicvlrf7djd3@news.supernews.com>
 <b0f083db.0204290751.52c145f4@posting.google.com>
 <Io2j9JAgacz8EwST@jessikat.fsnet.co.uk> <uct01lso59lu7d@news.supernews.com>
Message-ID: <HKJAFeAdQuz8Ewwr@jessikat.demon.co.uk>

In article <uct01lso59lu7d at news.supernews.com>, John Roth
<johnroth at ameritech.net> writes
>
......
>> Robin Becker
>
>If you do that, you can get some very non-intuitive behavior.
>
>For example:
>
>c = 2
>def foobar(a):
>    b = c
>    c = a
>
>In Python, the first assignment fails because 'c' is an
>undefined local variable. I've worked in languages
>where 'c' would be resolved as 2, and then suddenly
>change in the middle of the function to be resolved
>as the parameter 'a'. It's a very fertile source of bugs.
>
>Consider this loop:
>
>a = 2
>for x in foobar:
>    c = a
>    a = x
>
>The first time through the loop, 'c' is set to a
>value global to the loop, the rest of the time,
>it's set to the loop variable (on the preceding pass!)
>This is also very non-intuitive behavior.
>

in a lexical language the first ref(a) is the global, later ones are the
local, but I suppose any such rule is breakable/variable. I suppose
the static rule is easier to check statically, but here we seem
to have a mix of static and dynamic. 


>However, I suspect the primary motivation for this
>is optimization. The variable 'dictionary' in functions
>is not a dictionary, but a structure that allows faster
>access. Given that, the 'if it's assigned to, its a local
>variable' rule is quite reasonable.
>

That is the root of the problem

>In any case, any function or method that's big
>enough to cause a problem is probably too big.
>The 'one page rule' was in existence thirty years
>ago when we had to look at programs on paper!
>
>John Roth
>
>
No matter how small the fact that we have to look down & right to check
is a problem. People make assumptions about the way things work. Of
course we shouldn't and get bitten time and time again. If the rules
didn't have so many teeth we'd be better off.
-- 
Robin Becker


From sholden at holdenweb.com  Sun Apr  7 11:39:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 7 Apr 2002 11:39:59 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca>
Message-ID: <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com>

Sorry abt the format of this posting: OE foundit very difficult to remove
Colin's HTML...

"Colin J. Williams" <cjw at sympatico.ca> wrote in message
news:3CAF6EE6.7564CDF7 at sympatico.ca...

Guido asks seven questions:
[Colin provides seven answers]

> Perhaps we should also be rethinking the meaning of None.

Or the meaning of Life ... :-)

> Maybe it would be better to think of it as a NULL in relational
> algebra.
> Suppose we have a person record (p) with attributes age and sex.
> p.age= 22 and p.sex= F       (F asumed to be part of an enumeration)
> then we know something about this person.
> On the other hand, suppose we have p.age= None and p.sex= M
> We know that we have a male of unknown age.
> Suppose now that we ask whether this person's age is greater
> than 25:
>         if p.age > 25: ...
> Currently, the response is False, but really we don't know.
> Thus, it might be better to think in terms of a three value
> system False, Unknown, True.

If you really wanted to go this way (and there are worse options), the
"logical" (no pun intended) value to use for Unknow id clearly None itself.
Relational systems mostly seem to return NULL as the result of operations on
NULL, whether they be comparisons or arithmetic operations.

Note this would introduce three-valued logic into Python, which would be
helluva confusing to beginners, who tend to understand the law of the
excluded middle implicitly.

> The PEP 285 does not appear to envisage sub-classes, but
> they could be useful for people who would prefer to obtain
> No, Yes or N, Y from repr().
> To sum up, having True and False as defined constant values
> would help a little, but in view of the vociferous response
> and the degree of interest, a pause for further review might
> be best.

It's becoming clear to me that Guido's biggest mistake was proposing this as
a PEP. Since none of the arguments expressed as a result have swayed his
opinion that the change is a good thing, he might as well have simply
introduced the change by fiat.

That way he could have ignored the resulting howls of protest rather than
being forced to respond to them. Not much else would be different...

regards
 Steve







From mertz at gnosis.cx  Mon Apr  1 14:31:13 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Mon, 01 Apr 2002 14:31:13 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <idVp8.48143$S52.1619383@news2.tin.it> <7xk7rr3a38.fsf@ruckus.brouhaha.com>
Message-ID: <BWLq8kKkXkTX092yn@gnosis.cx>                

Paul Rubin <phr-n2002a at nightsong.com> wrote previously:
|The point about pickle/marshal is especially good.  I can probably
|think of a other dozen other serialization and remote object Python
|modules that would also have to be updated.

Actually, as auther of one of those dozen serializiation modules
(xml_pickle), I would vote +1/2 for this very reason.

One of the goals (not really realized yet) of xml_pickle, and especially
of XML-RPC or SOAP, is to communicate data bundles/objects across
different programming langugages.  If you can serialize objects, it is
sort of obvious to want to restore them later (maybe even in
heterogeneous environments).  A number of the potential recipients of
communication with xml_pickle are languages with a Boolean/Truth data
type.  While the impedence mismatch is not all that difficult to handle
as-is, the PEP makes it a tiny bit easier.

Overall, however, my feeling on the PEP is very close to neutral.  I can
see occassional problems and confusions coming up; and occassional other
confusions being clarified.  Nothing huge in either direction.

Yours, David...


--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/




From robin at jessikat.fsnet.co.uk  Sat Apr 13 06:04:49 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 13 Apr 2002 11:04:49 +0100
Subject: Python 2.2 unwise.exe
Message-ID: <ZkjnuIABLAu8EwYC@jessikat.fsnet.co.uk>

Anyone else had a problem un-installing 2.2 for win32? When I tried to
un-install 2.2 prior to installing 2.2.1 I got an error saying it had
already been un-installed.

Tried the link in start menu and it said no unwise.exe was there.

Luckily I had a copy of unwise.exe on another machine. Copied into the
right place and it seems to have done the trick.
-Flat cats are not better than nested birds-ly yrs.- 
Robin Becker


From bokr at oz.net  Tue Apr 16 08:06:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 12:06:12 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com> <a9fcut$e9l$0@216.39.172.122> <3cbb8f83@news.mhogaming.com>
Message-ID: <a9h43k$3nc$0@216.39.172.122>

On Mon, 15 Apr 2002 19:40:49 -0700, "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
[...]
>
>What I want, is an object that "looks and feels" like a 'binary twos-complement
>integer', that can be extended a byte at a time.  I don't _CARE_ how the machine
>represents it internaly.  It might even be in BCD for some weird CPU (for all we
>know)-- but the interface that _I_ see, is a 'binary twos-complement integer'
>made up of a finite number of bytes.  I don't want to know HOW the machine
>STORES it-- I just want a consistent interface to it.
>
No prob.

>When I print it out (in decimal)-- I want it to appear the way it does now--
>sign and magnitude.
>
>When I print it out in hexadeciaml, I want the bit representation of a 'binary
>twos-complement' number, of a size (in bytes) that will represent the number
>without truncation (regardless of HOW the number is stored inside the machine.)
No prob either.

>
>| >An int type simply shifts a zero into bit 31, so that
>| >this works the same as it will on a long.
>
>Let me clarify the above-- the '>>' should behave in a manner such that the sign
>bit is COPIED to the high bit during a right shift.  So-- if the number starts
>out as a positive number, it STAYS a positive number after the shift.  The
No prob.
>proposed '>>>' operator forces a zero into the high bit, having no effect on the
>sign bit of a positive number, but forcing a negative (twos-complement) number
>to now be positive after the shift.  Again-- I don't care HOW the number is
>represented internally-- this is the way it should always appear to the
>programmer...
Ok, but which sign bits should I keep when I shift in zeroes? I have to normalize
to something. A single sign bit is one way, not keeping to the left. Whatever sign
bits are required to fill out k*4 or k*8 bits is another. Or you can specify a
fixed width w and have it work like the 32 of in works now. But one way or another,
you have to provide a rule that results in a width to use for a particular negative
number, when you want to shift in zeroes. What's it going to be?

[...]
>|
>| In any case, you can define a function to whatever you actually want.
>| (I suspect you'll wind up specifying width explicitly). But if
>| you want a >>> operation, what width should it use? Remember, different
>| platforms may use different internal long representations.
>
>You CANNOT do this in a portable way from OUTSIDE of the object.  Only the
Sure you can. See below.

>OBJECT "knows" how many bytes in an equivalent twos-complement binary number are
>required to represent the number contained in the object.  Even if the internal
No, the required number of bits can be determined exactly from the numeric value.

>representation is in BCD, or a funny "17-bit per word" [an old IBM machine did
>this] format, the programmer should always "see" an array of BYTES that
>represent a twos-complement binary integer.  If the long class overloads the
>'>>>' operator, and it "knows" how many "pseudo-bytes" are needed to represent
>the number, then it can damn well "do the good thing" when this method is called
>(as an operator on the long.)  YOU however, are not supposed to be peeking
>inside the object (OOP rule #1-- encapsulation)-- let the object figure out how
>to do the '>>>' (and the '>>' and '<<' for that matter.)
Fine, but you have to define the meaning of >>> without "peeking inside".
Below I padded to full bytes before >>> shifting. Is that what you wanted?
(I'm not proposing this, I'm just trying to interpret your >>> so you can
see whether it makes more sense to you than it does to me ;-)

See if this does n >>> rsh the way you want:

 >>> def kshr(n, rsh):
 ...     nbytes = (len('%x' % ((n<<1)^n,))+1)/2
 ...     return (n&((1L<<(nbytes*8))-1)) >> rsh
 ...
 >>> hex(kshr( 0x80000000,1))
 '0x40000000L'
 >>> hex(kshr(-0x80000000,1))
 '0x40000000L'
 >>> hex(kshr( 0xf0000000,1))
 '0x78000000L'
 >>> hex(kshr(-0xf0000000,1))
 '0x7F88000000L'

note the extra byte needed before the shift:
 >>> hex(kshr(-0xf0000000,0))
 '0xFF10000000L'

If I did it as intended, it figures how many bytes are necessary to hold
the sign and significant bits of n, masks n to that size, and shifts right rsh.
(Note that I didn't rewrite hex to prepend zero at any time).

See if this does it:

 >>> for i in xrange(36): print '%16s%16s' % (hex(kshr(-1L<<i,0)),hex(kshr(-1L<<i,1)))
 ...
            0xFFL           0x7FL
            0xFEL           0x7FL
            0xFCL           0x7EL
            0xF8L           0x7CL
            0xF0L           0x78L
            0xE0L           0x70L
            0xC0L           0x60L
            0x80L           0x40L
          0xFF00L         0x7F80L
          0xFE00L         0x7F00L
          0xFC00L         0x7E00L
          0xF800L         0x7C00L
          0xF000L         0x7800L
          0xE000L         0x7000L
          0xC000L         0x6000L
          0x8000L         0x4000L
        0xFF0000L       0x7F8000L
        0xFE0000L       0x7F0000L
        0xFC0000L       0x7E0000L
        0xF80000L       0x7C0000L
        0xF00000L       0x780000L
        0xE00000L       0x700000L
        0xC00000L       0x600000L
        0x800000L       0x400000L
      0xFF000000L     0x7F800000L
      0xFE000000L     0x7F000000L
      0xFC000000L     0x7E000000L
      0xF8000000L     0x7C000000L
      0xF0000000L     0x78000000L
      0xE0000000L     0x70000000L
      0xC0000000L     0x60000000L
      0x80000000L     0x40000000L
    0xFF00000000L   0x7F80000000L
    0xFE00000000L   0x7F00000000L
    0xFC00000000L   0x7E00000000L
    0xF800000000L   0x7C00000000L
 >>> for i in xrange(36): print '%16s%16s' % (hex(kshr(+1L<<i,0)),hex(kshr(+1L<<i,1)))
 ...
             0x1L            0x0L
             0x2L            0x1L
             0x4L            0x2L
             0x8L            0x4L
            0x10L            0x8L
            0x20L           0x10L
            0x40L           0x20L
            0x80L           0x40L
           0x100L           0x80L
           0x200L          0x100L
           0x400L          0x200L
           0x800L          0x400L
          0x1000L          0x800L
          0x2000L         0x1000L
          0x4000L         0x2000L
          0x8000L         0x4000L
         0x10000L         0x8000L
         0x20000L        0x10000L
         0x40000L        0x20000L
         0x80000L        0x40000L
        0x100000L        0x80000L
        0x200000L       0x100000L
        0x400000L       0x200000L
        0x800000L       0x400000L
       0x1000000L       0x800000L
       0x2000000L      0x1000000L
       0x4000000L      0x2000000L
       0x8000000L      0x4000000L
      0x10000000L      0x8000000L
      0x20000000L     0x10000000L
      0x40000000L     0x20000000L
      0x80000000L     0x40000000L
     0x100000000L     0x80000000L
     0x200000000L    0x100000000L
     0x400000000L    0x200000000L
     0x800000000L    0x400000000L

Regards,
Bengt Richter


From cliechti at gmx.net  Tue Apr  2 17:44:19 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Apr 2002 23:44:19 +0100
Subject: portable subprocess with return value and path search
References: <5fa44575.0204021305.600a987@posting.google.com>
Message-ID: <Xns91E4F1A829CD2cliechtigmxnet@62.2.16.82>

joshualevy at yahoo.com (Joshua Levy) wrote in 
news:5fa44575.0204021305.600a987 at posting.google.com:

> I'm trying desperately to get python to do the following
> four things at the same time:
>     1. Run a subprocess (like "java com.foo.bar arg")
>     2. Get the process return value from that subprocess
>     3. Work on both Windows and UNIX  (I'll settle for 
>           post-Win2000 and Solaris/Linux.)
>     4. Resolve the executable, based on the current path.
>        (So I only need "java", and not "c:\jdk1.2.2\bin\java",
>        in the example above.)
> Os.system does not return values from Windows.  The versions of
> Os.spawn on Widnows do not resolve executables.
> Thanks in advance for any help.

popen seems to fulfill the above points:
>>> os.popen("exit 77").close()
77

(win2k, py2.2)

you get a file like object from popen. maybe you need to read() away the 
output so that the running app does not block because of a full stdout 
pipeline.

-- 
Chris <cliechti at gmx.net>



From gustav at morpheus.demon.co.uk  Sun Apr 28 08:40:29 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Sun, 28 Apr 2002 13:40:29 +0100
Subject: slices: foo[i:j:k] => TypeError
References: <mailman.1019856470.541.python-list@python.org> <aCky8.66523$vF6.2038703@news2.tin.it> <mailman.1019924208.19241.python-list@python.org> <fFDy8.70854$vF6.2116208@news2.tin.it>
Message-ID: <cqqncugq00loh16lp8hgl6stslqt36ifg1@4ax.com>

On Sat, 27 Apr 2002 20:15:07 GMT, Alex Martelli <aleax at aleax.it> wrote:

>>     What's the expected date for the builtin types to accept slice
>>     objects?
>
>No idea.  The original intro was before PEPs, so there's not even a
>PEP to refer to, I believe.  You may want to ask on Python-Dev --
>if the BDFL OK's it for 2.3, I can write the code, if you or some other
>volunteer can handle the rest of the patch submissal process (sorry,
>can't volunteer more time than what it takes to write & debug the code).

Hmm, I've always been annoyed by this wart. I've never had the time to
do any coding, though...

Let me put out some feelers on python-dev, and see what response I get.
If it looks promising, I'll do the bureaucracy for you.

Paul.


From trentm at ActiveState.com  Wed Apr  3 15:48:34 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 3 Apr 2002 12:48:34 -0800
Subject: IDE Wars, the prequel!
In-Reply-To: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com>; from Oschler@earthlink.net on Wed, Apr 03, 2002 at 03:37:24PM -0500
References: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com>
Message-ID: <20020403124834.B1387@ActiveState.com>

[Robert Oschler wrote]
> In a galaxy far, far away..
> 
> I've been getting some IDE suggestions on the COLA forum.  The current batch
> seems to be:
> 
> Gvim
> Emacsen
> ActiveState's ActivePython

A clarification here. Mark Hammond's set of Win32 extensions for Python
includes an IDE called PythonWin. The Win32 extensions are installed by
default with ActivePython, hence you have PythonWin is you installed
ActivePython. (You can still get PythonWin with other Python distributions --
you just have to install win32all after installing Python. See
http://starship.python.net/crew/mhammond/)

ActiveState also has an IDE called Komodo (for Python, Perl, PHP, XSLT, etc)
that runs on Windows and Linux.
    http://www.ActiveState.com/Products/Komodo/

You should look at: http://www.python.org/editors/ 

Aside: Anybody know who maintains "http://www.python.org/editors/". Its
comments on PythonWin are misleading.

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From gimbo at ftech.net  Mon Apr 22 10:25:35 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Mon, 22 Apr 2002 15:25:35 +0100
Subject: Science And Math  Was: Python's Lisp heritage
In-Reply-To: <Pine.LNX.4.33.0204220729500.25157-100000@ns0.tbye.com>
References: <20020422124126.GD23706@andy.tynant.ftech.net> <Pine.LNX.4.33.0204220729500.25157-100000@ns0.tbye.com>
Message-ID: <20020422142535.GA25461@andy.tynant.ftech.net>

On Mon, Apr 22, 2002 at 07:35:26AM -0700, brueckd at tbye.com wrote:
> On Mon, 22 Apr 2002, Andy Gimblett wrote:
> > To put this another way, ask yourself where you stand on this
> > question: is mathematics discovered or invented?
> 
> Is there a difference?

Well, in the context of the debate, there is.  Perhaps you personally
don't believe there's any difference between inventing or discovering
maths, and hence that the debate is invalid, but there are others in
the world who disagree with you, believe the debate is a valid and
possibly important one, and may even go so far as to hold one of the
following to be true, and the other to be false:

Discovery: Mathematics is intrinsically "there" and "true" regardless
           of our intervention.  It's basically an empirical science,
           and is useful precisely because it represents reality
           directly and accurately.

Invention: We're making mathematics up as we go along and it bears no
           direct relation to actual reality.  It's useful because we
           construct it in such a way as to be useful, but to say our
           maths and only our maths is a proper reflection of reality
           is a monumental act of hubris upon our part.

I can't decide which I believe, or even which I want to believe.  :-)

> One could argue that the only difference between discovery and
> invention is the arrogance of the person reporting it.

Interesting point of view...  Unfortunately I can't decide whether you
mean claiming discovery is more arrogant, or claiming invention.  I'd
guess invented, though.  Which would make you a Platonist.  :-)

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From jbperez808 at yahoo.com  Sun Apr 21 22:07:54 2002
From: jbperez808 at yahoo.com (Jonathan P.)
Date: 21 Apr 2002 19:07:54 -0700
Subject: compiled Jython vs. CPython
Message-ID: <f57664b9.0204211807.2235f3bf@posting.google.com>

Has anyone out there tried to compare the
speed of a jythonc-compiled app to that of a
CPython one?

It is possible that if such an app does not
depend on a runtime interpreter and is run
by a JIT-JVM that it might give better performance
than one which runs interpreted (i.e. CPython).

JIT-compilation techniques have come a long way 
and in fact most benchmarks I've seen show that 
JIT-compiled Java code run faster than 
statically-compiled ones by a substantial 
margin!

Could jythonc be *the* Python compiler we wish
we had?


From jeff at ccvcorp.com  Tue Apr  2 13:49:44 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 10:49:44 -0800
Subject: Tkinter + import question
References: <3ca754d4$1_1@nopics.sjc>
Message-ID: <MPG.17139d115e6884f4989687@news.nwlink.com>

In article <3ca754d4$1_1 at nopics.sjc>, deltapigz at telocity.com 
says...
> i am writting a program utilizing Tk as my gui engine, and i have broken
> down sections of the form into modules and loaded them through a main
> script. my only question is in every module i am forced to import the
> Tkinter module, and does this pose consume too much memory?

You do have to import Tkinter (or any other module) into every 
module of yours that (directly) uses it.  Note that, if you 
create a Tkinter object X in Module A, then pass that object to 
Module B, Module B doesn't necessarily need to import Tkinter in 
order to use object X -- only if you need to use 
functions/constants/etc from the Tkinter namespace do you need to 
import Tkinter.  Object X's methods are in object x's own 
namespace, and can be used on their own.  (But if you're passing 
constants into those methods, you may need to import Tkinter to 
get the constants.)

Don't worry about memory consumption, though.  Once a module is 
imported once in your program, any further imports are *very* 
quick and simple.  The initial import runs the module and creates 
an entry in the sys.modules dictionary.  Subsequent imports 
simply create a new reference to that sys.modules entry -- they 
are *very* lightweight.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From klausman-un070202 at tuts.net  Wed Apr  3 05:34:31 2002
From: klausman-un070202 at tuts.net (Tobias Klausmann)
Date: 3 Apr 2002 10:34:31 GMT
Subject: date / time
References: <3CAAD0D1.2040309@beth.uniforum.org.za>
Message-ID: <a8elrm$r7du9$3@ID-41653.news.dfncis.de>

mixo <mixo at beth.uniforum.org.za> wrote:
> 
> Suppose there is a given date, 2002-03-01 (YYYY-MM-DD).
> How can you get the day (2002-02-28 ) before the given date and
> check that it is valid?

(I replace the interpreters >>> prompt with ### for better
 legibility in USENET posts)

### import time
### date="2002-03-01"
### time.strptime(date,"%Y-%m-%d")
(2002, 3, 1, 0, 0, 0, 4, 60, 0)
### time.mktime(time.strptime(date,"%Y-%m-%d"))
1014937200.0
### time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60)
1014850800.0
### time.ctime(time.mktime(time.strptime(date,"%Y-%m-%d"))-(24*60*60))
'Thu Feb 28 00:00:00 2002'
###

Just remember that time.ctime() will give you local time
while time.gmtime() will give you UTC. The input date is
assumed to be the local time. If you drift between TZs, make
use of the time.tzname variable.

By the way: what do you mean by "valid"?

Greets/HTH,
Tobias



From fperez528 at yahoo.com  Mon Apr 15 15:14:15 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 13:14:15 -0600
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org> <3CBB1AE8.E2843017@kfunigraz.ac.at>
Message-ID: <a9f8ru$lef$2@peabody.colorado.edu>

Siegfried Gonzi wrote:

> I ran into a similar problem: including an external C function in
> Python. The SWIG manual is somewhat behind time and my first attempts
> failed. I just then wrote to the developers (or mailinglist?) and they
> explained me a "how to" with the assistance of distutils.
> 

Could you share that mini-howto here with the rest of us? I'm sure that more 
than one would find it useful.

cheers,

f


From pyth at devel.trillke.net  Wed Apr 24 11:54:12 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 24 Apr 2002 17:54:12 +0200
Subject: Probably I'm just too stupid, but...
In-Reply-To: <34e7f752.0204240746.7d2c41b0@posting.google.com>
References: <34e7f752.0204240746.7d2c41b0@posting.google.com>
Message-ID: <20020424155412.GB20209@devel.trillke>

On Wed, Apr 24, 2002 at 08:46:06AM -0700, redflag wrote:
> >>> def c():
> 	for i in range(10):
> 		print i
> 
> 		
> >>> c
> <function c at 0x00AA93B0>

this tells you that you have a function at hand
(which is an callable object in python semantics).

just do

>>> c()

and it will work.

	holger




From clpy at snakefarm.org  Tue Apr 23 11:45:33 2002
From: clpy at snakefarm.org (Carsten Gaebler)
Date: Tue, 23 Apr 2002 17:45:33 +0200
Subject: Why won't this POST
References: <8f8ffe67.0204222047.47d661ec@posting.google.com> <3CC528C8.FDE414CE@snakefarm.org> <8f8ffe67.0204230502.261eeca9@posting.google.com>
Message-ID: <3CC5819D.FB70344D@snakefarm.org>

Sunit Joshi wrote:
> 
> Carsten Gaebler <clpy at snakefarm.org> wrote in message news:<3CC528C8.FDE414CE at snakefarm.org>...

> > "localhost/eubix" is not a valid hostname. What's the "/eubix" part
> > supposed to mean?
> >
> > cg.
> 
> Actually localhost/eubix is the virtual directory in IIS5.0. I can
> connect to localhost/eubix, thru. req = HTTP('localhost/eubix')
> But after that everything fails.

Nope, HTTP() does not establish a connection. It just initializes some
internal variables (hostname and port). The connection is made within
putrequest(), which is what you see in your traceback.

cg.
-- 
Python is taken from python site. What to do? What is that?
                            -- scenes from comp.lang.python


From zopestoller at thomas-guettler.de  Tue Apr 16 11:08:09 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 16 Apr 2002 17:08:09 +0200
Subject: xml2dict - dict2xml
References: <a9gudi$964$04$1@news.t-online.com>
Message-ID: <3CBC3E59.5090908@thomas-guettler.de>

Henning Peters wrote:

> hi NG,
> 
> i am searching for the best solution to convert a small xml fragment in a
> python dictionary and vice versa.
> speed is a very important factor, so i have tried the xml.parsers.expat
> interface, but i did not found out, how to map a tree with it (i.e. to parse
> only one branch).
> 


Try this link:

http://www-106.ibm.com/developerworks/xml/library/x-matters11.html


I asked a similar question some time ago. I haven't tried the
code from this article.

  thomas



From danb_83 at yahoo.com  Sat Apr 20 02:31:15 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 19 Apr 2002 23:31:15 -0700
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <ad052e5c.0204192231.693c095b@posting.google.com>

Shane Hoversten <srh232 at nyu.edu> wrote in message news:<3CC0D892.2060006 at nyu.edu>...
> Hi -
> 
> 	This is really dumb, but I just want to print out a character without a 
> newline and without a space behind it.  You can't do:
> 
> print c,
> 
> 	Because Python puts a space behind it.  I can't find anywhere in the dox 
> that actually talk about "print" in any detail at all.  Surely there 
> must be some way to do this?

sys.stdout.write(c)


From marklists at mceahern.com  Tue Apr 30 18:40:40 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 17:40:40 -0500
Subject: crypto
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEIGKNAA.marklists@mceahern.com>

Hi, I need to:

1.  Generate public/private key pairs.
2.  Encrypt data with the public key.
3.  Decrypt it with the private key.

Step 1 is likely to be manual.  (You're probably thinking, "No shit,
sherlock.")

I'm using Linux and I've seen how I can create a key with gpg.  I'm just not
sure how to use that properly from Python.  I suppose I could write things
out to files and use os.popen() or something, but that seems like it would
increase my risk, not decrease it.  ;-)

I'm a relative novice with cryptography, so I'd appreciate any and all
pointers.

I've downloaded amk's pycrypt package:

  http://www.amk.ca/python/writing/pycrypt/

Sadly, the Crypt.PublicKey module has been removed.

Thanks for any feedback.

Cheers,

// mark





From johanfo at essay.org  Sun Apr 28 22:48:37 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Mon, 29 Apr 2002 04:48:37 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net>
Message-ID: <aaicbc$9jt$1@maud.ifi.uio.no>

Yes, jonpy does look promesing.  Looks very professional also :)  I have teste
various solutions now including Webware, I'll test jonpy tomorrow.

--
JF?

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnacpc38.obc.jon+usenet at snowy.squish.net...
> In article <aai5oo$99l$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> > Just realized that these Servlet based development models really limits the
> > usage of software like Dreamweaver.  Since all code is in the program these
> > utilities are almost worthless.  Is there a good way to combine Webware with
> > templates.  I think of templates as pages where certain %words% are
> > substituted etc...
>
> This is exactly why I created jonpy (http://jonpy.sourceforge.net/).
> It puts *no* Python code inside your HTML templates, you simply mark
> sections using comments (e.g. <!--wt:newsitem-->) and replacements
> using dollar signs (e.g. $$title$$). The idea is that even if you are
> suddenly given an entirely new design with new HTML files with none of
> the modifications you made to make the pages work, it will only take a
> few minutes to fix them up again.




From pinard at iro.umontreal.ca  Fri Apr 12 22:48:51 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 12 Apr 2002 22:48:51 -0400
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
In-Reply-To: <MPG.17211dcc4bd939939896af@news.nwlink.com>
References: <mailman.1017656355.5887.python-list@python.org>
	<12bb92e6.0204091223.738e0798@posting.google.com>
	<i_nt8.15754$ir6.679622@news1.calgary.shaw.ca>
	<GrAt8.11681$F43.8566@atlpnn01.usenetserver.com>
	<eGIt8.21408$de1.1000064@news3.calgary.shaw.ca>
	<MPG.172101b4b7f13f639896ae@news.nwlink.com>
	<1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca>
	<MPG.17211dcc4bd939939896af@news.nwlink.com>
Message-ID: <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca>

> > Public domain software is [...] a
> > special case of non-copylefted free software, [...]

[Jeff Shannon]

> Which *does*, in fact, mean that it is free software.  Unless perhaps
> free software means something other than free software?

There are many definitions of free software, and some people fight over
which definition is the real one.  Some even want monopoly of freedom! :-)

How I understand the GPL is that a software is not free, if it can be
used to create a bound between parties, in such a way that one of the
parties has lost part of its freedom on that software through that bound,
in particular the freedom to modify the software and/or share it around.

A public domain software can be modified by anyone, and the result can be
made available to someone else under a binding license.  So, that software
was indeed free, but that quality of being "free" was totally fragile...

The GPL wants to help programs at being free in more permanent ways.

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




From mlh at vier.idi.ntnu.no  Sat Apr 27 18:08:08 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 27 Apr 2002 22:08:08 +0000 (UTC)
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org>
Message-ID: <slrnacm8a7.mru.mlh@vier.idi.ntnu.no>

In article <mailman.1019792457.31426.python-list at python.org>, James J.
Besemer wrote:
[snip]
>4. Actually, this all makes me wonder if your web source is accurate in the first
>place?  I've come to place a lot of trust in my well worn OAD.  Shit.  Now I'll be
>pouring over different dictionaries all evening.

While you're at it you might want to look the difference between
"pore" and "pour", unless you want to ruin your dictionaries <wink>.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From aleax at aleax.it  Wed Apr 10 06:53:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 10:53:14 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <a8vpp5$67d$1@panix1.panix.com> <yu99pu18lai2.fsf@europa.research.att.com>
Message-ID: <uQUs8.2468$vF6.87600@news2.tin.it>

<posted & mailed>

Andrew Koenig wrote:

>>> Now, f is an object that I can use to rebind the name x.  For
>>> example, if I execute f(42), that sets x to 42.
> 
> Aahz> You really don't want to do this in Python.
> 
> One thing that fascinates me is how the nature of a language affects
> the preferred idioms therein.  To help me further my understanding,
> would you mind explaining to me why this particular idea is undesirable

Guess you could describe it as a social contract about "who/what binds
a function's local variables".  The Python answer is, only the code of
the function itself binds the function's local variables -- the very
fact of binding a name ("unqualified" name, if you will -- dots apart!) in 
a function is what defines that name as being local (unless a global
statement explicitly says otherwise).  Local variables are bound by
assignment statements (plain ones, with just '='; augmented ones may
sometimes RE-bind, but never give a name its first binding... pity
that using an augmented assignment ONLY, *still* tags the variable as
local, but I guess either choice would have problems) or one of a few
other statements that bind names (def, class, import, from).  Binding
with an exec statement is possible but "considered harmful" -- exec
is too powerful and hard to control, particularly if you don't use
it with explicit dictionaries, and thus tends to produce subtle, hard
to find bugs too often.  There are almost always better alternatives.

Part of the social contract comes from higher level issues, but part
of it stems from a long-standing optimization: locals() is NOT a
real rebindable dictionary, but just masquerades as one in some
ways; KNOWING that local variables are only bound by the code of
the function itself lets the compiler perform a crucial optimization.

What you could see as a further social contract is: any call such
as f(x) NEVER has any effect on the BINDING of name x -- possibly
on the OBJECT to which name x is bound, if that object is mutable,
but not on the binding itself.  Knowing that constraint allows a
lot of reasoning about what's going on, that would be harder or
impossible if that constraint wasn't there.  Here, there are really
no implementation issues in play -- this holds for global names
(which _are_ generally rebindable from just about everywhere) and
"composite" names too (such as x.y) -- it IS a social contract that
puts such "rebindings as a side effect of function calls on
names used as arguments" out of bounds.  Function f receives no
information at all about what names or other syntax forms were
used to obtain the actual-arguments it receives -- it just gets
the values of the actual-arguments.  So for example after:
    x = y = whatever_expression_you_want
    f(x)
there is NO way for f to ascertain whether it was called as
f(x) or rather as f(y) -- short, I guess, of rooting through the
call stack and looking for source files or bytecode of its
caller to find out.  Thus, it's socially "unthinkable" that
there would be any specific effect on the exact name used by
f's caller... given that f cannot even find out about that
name (not portably, not with a bounded amount of effort).


Given all this, if you find out and implement some Ruby Goldberg scheme 
that works (sort of) this way, you'll have code that is totally baffling to 
Pythonista users thereof, because it runs so deeply counter to our
expectations, our usual social contracts.


Alex



From phr-n2002a at nightsong.com  Fri Apr 26 23:57:19 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 26 Apr 2002 20:57:19 -0700
Subject: [OT] What is Open Source?
References: <mailman.1019853592.15502.python-list@python.org>
Message-ID: <7xwuutssio.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> Me too, but at least one lawyer has told me it is a fantasy world,
> i.e. that an individual in the United States cannot put their work
> in the public domain: you can't disclaim your copyright even if you
> want to, unless it's work done for the government and "they" put it
> in the public domain.

This actually doesn't surprise me, if you think about situations where
someone might want to stop you from disclaiming a copyright :-(.


From mrenxo at mail.com  Fri Apr 26 12:52:48 2002
From: mrenxo at mail.com (Enxo)
Date: 26 Apr 2002 09:52:48 -0700
Subject: Killing a process under Windows NT/2K
References: <aab8hn$o4s$05$1@news.t-online.com>
Message-ID: <6affa25.0204260852.3e1a4f4a@posting.google.com>

> I need to kill a process under Windows (NT/2K), whose PID is known.
> Unfortunately os.kill() doesn't work under Windows.
> How can I achieve this anyway?

Try this:

      import win32process
      import win32api
      import win32con
               
      pid = win32api.GetCurrentProcessId()  
      handle = win32api.OpenProcess(win32con.PROCESS_TERMINATE, 0, pid)
      win32process.TerminateProcess(handle, 1)

Good Luck!


From jarvin24 at tutor.cc.tut.fi  Sat Apr 13 08:50:01 2002
From: jarvin24 at tutor.cc.tut.fi (=?iso-8859-1?Q?J=E4rvinen?= Petri)
Date: Sat, 13 Apr 2002 12:50:01 +0000 (UTC)
Subject: FileUpload via CGI by Python
Message-ID: <slrnabgabq.gjo.jarvin24@tutor.cc.tut.fi>

Hey, i'm trying to do form that uploads file to server.

Is there any ready solutions?

- Petri


From pinard at iro.umontreal.ca  Mon Apr 29 22:14:02 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 22:14:02 -0400
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCDE443.CA2525FF@cascade-sys.com>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
	<oqpu0inmxa.fsf@titan.progiciels-bpi.ca>
	<3CCDE443.CA2525FF@cascade-sys.com>
Message-ID: <oq7kmqndat.fsf@titan.progiciels-bpi.ca>

[James J. Besemer]

> Nor is Sarcasm or frank talk necessarily unfriendly.

"Frank talk" might be an English idiom I do not know.  Reading the words
for what they genuinely mean, I feel like sharing that one may express
his ideas very frankly, without any need to be sarcastic or strong-worded.

The Python community is more appealing _without_ sarcasm, and _with_ frank
and articulated speech.  Let's all collaborate at keeping it a nice place
to be.

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




From bokr at oz.net  Sat Apr  6 16:41:43 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Apr 2002 21:41:43 GMT
Subject: Python-list digest, Vol 1 #10324 - 13 msgs
References: <E16tGLp-0007ou-00@mail.python.org> <mailman.1018014793.1604.python-list@python.org> <axir8.52642$pT1.1485864@news1.tin.it>
Message-ID: <a8nq2n$8s3$0@216.39.172.122>

On Fri, 05 Apr 2002 14:29:26 GMT, Alex Martelli <aleax at aleax.it> wrote:

>Jonathan Claggett wrote:
>        ...
>> Actually, I really wasn't thinking in terms of functional programming (I
>> can't Lisp my way out of a paper bag :-). My immediate motive was only
>> skin deep in that I'd like to replace the (unique) lambda syntax in my
>> code with the (standard) def syntax.
>
>You can do that now!  def is a statement and you can put it wherever you
>need.  Just choose a good name, as you would for some complicated
>sub-expression in order to clarify things.  For example, instead of
>
>Button(buttonsFrame, text='Squirk', command=lambda: self.squirkIt(23, 45))
>
>you can code:
>
>def specificSquirker(): self.squirkIt(23, 45)
                        ^-- return 
if you really want equivalent code, though maybe only side effects matter
in the the example context?

>Button(buttonsFrame, text='Squirk', command=specificSquirker)
>
>
>It's generally not hard to come up with a name which helps readability.
>I think un-lambdaing your code in this way is a good thing...
>

Regards,
Bengt Richter


From db3l at fitlinxx.com  Sat Apr 20 00:27:38 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 20 Apr 2002 00:27:38 -0400
Subject: Python 2.2 unwise.exe
References: <mailman.1019259099.12047.python-list@python.org>
Message-ID: <uofgfkn9h.fsf@ctwd0143.fitlinxx.com>

Tim Peters <tim.one at comcast.net> writes:

> Wise agreed, and quickly built a new version of UNWISE.EXE that doesn't have
> this problem.  This leaves one other small problem:  how do we manage to
> magically replace a hundred thousand or so vulnerable UNWISE.EXE's already
> installed in Python directories on Windows boxes?  It's times like this I'm
> glad Python installs secret backdoors on Windows boxes so Guido can control
> them from the comfort of home.

I don't see how to fix the current installed base, but since they'll
only get hit - thus it really only matters - when they go to install a
new version (and thus will be downloading it), perhaps publishing a
copy of UNWISE.EXE along with the Python downloads would at least
provide a workaround if they don't have another copy somewhere on
their system to use.

Either that or try to see if the installation script could be made to
clean out any pending delete operation as part of the installation?

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aahz at pythoncraft.com  Thu Apr  4 16:27:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Apr 2002 16:27:52 -0500
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com>
Message-ID: <a8iggo$28k$1@panix1.panix.com>

In article <a8hq64$114s$1 at news.rchland.ibm.com>,
Larry Whitley <ldw at us.ibm.com> wrote:
>
>On a windows 2k system I have a single master thread, spawning two worker
>threads and ending. The master thread is started from the command prompt
>(run.py) and the worker threads use "start /belownormal /wait ..." to run a
>sequence of C++ programs that I have writen that take a long time to execute
>(they're simulation performance models). 

If you're using start, you're not really using threads.  You'll need to
create a much more robust architecture, whether it's by embedding the
C++ programs in Python extensions (and learning how to release the GIL)
or by turning the C++ programs into COM servers and connecting them to
Python through win32all, or some other mechanism I haven't thought of.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From jeffbi at jetsoft.com  Tue Apr  2 18:07:28 2002
From: jeffbi at jetsoft.com (Jeff Bienstadt)
Date: Tue, 2 Apr 2002 15:07:28 -0800
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net> <3ca8f000$1@news.nwlink.com> <JLcfwNF0zPq8EwJm@nym.alias.net> <3ca9e2bd@news.nwlink.com> <MPG.1713c1ae2a7ecc1698968d@news.nwlink.com>
Message-ID: <3caa39b1$1@news.nwlink.com>

"Jeff Shannon" <jeff at ccvcorp.com> wrote in message
news:MPG.1713c1ae2a7ecc1698968d at news.nwlink.com...
> In article <3ca9e2bd at news.nwlink.com>, jeffbi at jetsoft.com says...
> >
> > One of my other goals is to have a newsreader that works (as much as
> > possible) the same on both Linux and on Windows, so I was thinking
> > that using Tkinter for the GUI would probably be the best way to
accomplish
> > that.
>
> My personal recommendation would be to look at wxPython.  Tkinter
> is a bit easier to pick up, especially if you have no previous
> GUI experience, but it becomes difficult to do more complex
> things with it.  Also, Tkinter apps always *look* like Tkinter
> apps.  By contrast, wxPython (and wxWindows, which it wraps) uses
> the platform-native widgets wherever possible, so a wxPython app
> looks like a Windows app on Windows and a GTK app on Linux.
> wxPython is also more flexible than Tkinter, and it doesn't rely
> on another scripting language like Tkinter does (Tkinter relies
> on TCL).
>
> --

Thanks, Jeff.  I'll look into this.

I had heard of wxPython (seen it mentioned in various posts here)
but I hadn't made the connection (doh!) --- I didn't know it was a
wrapper around wxWindows, I just assumed it was another Win32
implementation of Python.

Thanks again,

---jkb





From jepler at unpythonic.net  Sat Apr 27 16:06:08 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 27 Apr 2002 15:06:08 -0500
Subject: Update locals()
In-Reply-To: <20020427195725.GA16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke>
Message-ID: <20020427150607.A2052@unpythonic.net>

On Sat, Apr 27, 2002 at 09:57:25PM +0200, holger krekel wrote:
> Hello,
> 
> how can you dynamically update the locals() namebindings?

You don't.  See the library documentation:
    locals()
	Return a dictionary representing the current local symbol table. Warning:
	The contents of this dictionary should not be modified; changes may not
	affect the values of local variables used by the interpreter.

Whenever a name is not bound inside a function, it is a global (it could
never have a definition inside the function, so if the name comes from
anywhere, it's global).  If there's a "global" declaration, then it's
global.  If there's an assignment, then it's a local -- but changing a key
in locals still doesn't work (or is at least permitted not to work).

You probably need to rethink the way you're approaching the problem.

Jeff




From juenglin at informatik.uni-freiburg.de  Fri Apr 26 04:57:10 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 26 Apr 2002 10:57:10 +0200
Subject: Initializing a list with copies
Message-ID: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de>

Hi,

does anybody see an easy way to do the following:
I need a list of initially empty lists that are to be
filled, l say. The parameter n says how many empty
lists l should contain. Someone tried 

>>> l = [[]]*n

to create l. However, this is not what he wanted:
>>> n
2
>>> l
[[], []]
>>> l[0].append(1)
>>> l
[[1], [1]]


To create an l with inidividual empty lists, he now
does
>>> l = [[] for i in range(n)]


Can this be done more elegant? By some expression that 
was similar to '[[]]*n' ?

Thanks,
Ralf


From fakeaddress at nowhere.org  Thu Apr 25 22:34:35 2002
From: fakeaddress at nowhere.org (Bryan Olson)
Date: Fri, 26 Apr 2002 02:34:35 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com> <3CC7889D.2080002@nowhere.org> <7x8z7cvusy.fsf@ruckus.brouhaha.com>
Message-ID: <3CC8BCD1.7090804@nowhere.org>

Paul Rubin wrote:

 > I thought of adding a function that would allow encrypting in segments
 > but don't currently have applications that need that.  Remember that
 > the main goal of this code (and I didn't expect it to become such a
 > complicated project) was to just have something in portable Python
 > that could be used instead of the silly rotor module.

I'm surprised such a trivial goal interests you.  If this isn't worth
doing well, then I think it's just another case of "hey look - I
invented a cipher", and personally I respond to those with "O.K. throw
it on the 'unused' pile."

The reason I'm enthusiastic about this project is that it actually meets
a need.  Due to silly politics, a variety of programming environments
ship with SHA-1 and without a decent cipher.  The value here is in
creating a strong and efficient cipher without stepping outside the
native programming languages.

The definition of the cryptosystem is much more valuable than any
particular implementation.  If you don't have a need to encrypt in
segments while preserving state in-between, then fine don't implement
it.  The definition of the system should still have that potential.

The reference implementation I offered doesn't offer a piece-by-piece
facility either, but by separating the IV and MAC from the ciphertext,
it defines what such a system should produce.

 > It's slow
 > enough that it would be painful to encrypt large files with it.

Sure, but stand-alone (en/de)cryption a large file at one time is
usually a mistake.  Even my reference implementation, on my 'budget'
Celeron 1200, is fast enough to saturate my DSL line many times over.

A Python-only cipher is of little, maybe negative, value.  A cipher that
is very fast in C (or ASM) and moderately fast in a variety of scripting
languages could be a serious player.  We should aim for the big-time.


 > And
 > in an SSL-like protocol it still seems ok to encrypt-and-sign each
 > message.  (You couldn't implement real SSL with this module because of
 > the nonstandard cipher).

Well, I kind of disagree with the principle.  If this cipher is to be of
real value, we should definitely not assume it will always be
"nonstandard".  It's unlikely to be a FIPS, but it could be at least as
standard as RC4.

SSL/TLS is designed to support multiple ciphers.  Certainly support for
AES is coming.  There's no reason this cipher can't be included.  I see
little use for an "SSL-like" protocol.  Protocols are valuable if and
only if they are widely implemented.


--Bryan



From yee at uclink.berkeley.edu  Fri Apr 19 19:49:17 2002
From: yee at uclink.berkeley.edu (Raymond Yee)
Date: 19 Apr 2002 16:49:17 -0700
Subject: Unable to use win32com on Ecco COM server despite success with VB and Perl COM
Message-ID: <e5eab55e.0204191549.41985d9a@posting.google.com>

I've been trying unsuccesfully to use Python to access the
functionality of the "Visual Basic Toolkit"
(http://www.ivitar.com/vbecco/). The toolkit is a COM InProc
Server (.exe).

I'm puzzled as to how to proceed because of two things:  1) My
Python program is able to use some functions of the toolkit but
not others and 2) I've had success with both Visual Basic and Perl
in accessing the toolkit via COM.

Below are the three versions of the program.  Ecco is a personal
information manager.  This program is supposed to print out the
text of any selected outline and the corresponding URL in a folder
called "Net Location"  (The developer key has beeen x'ed out)

In the Perl and VB versions, I am able to get both the "text
value" of the selected item ($item->Text in Perl) and the
corresponding URL ($item->FolderValue("Net Location")).  However,
in the Python version, while EccoItem.Text is what I expect,
EccoItem.FolderValue("Net Location") returns

(u'Net Location', None)

-- a tuple, whose first item is type unicode.  However, I don't
see anything in the return value remotely related to the URL I'm 
trying to access.

Any suggestions on how to get Python to work with the toolkit? (Am
I using the wrong syntax?  Are there options I'm setting right?
How should I debug this?)

Thanks in advance,

-Raymond Yee

---------------------------------------------------------------
Python version:

# build OPML from the selected outline (if more than one is selected,
use the first item

import win32com.client
Ecco = win32com.client.Dispatch("VBEcco.EccoApp")
Ecco.Register("xxxxxxxx", -1, -1)
EccoSelection = Ecco.Selection

#print value of item in Net Location folder

for EccoItem in EccoSelection:
    print "Text: " + EccoItem.Text
    print EccoItem.FolderValue("Net Location")

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

Perl version:

# establish COM connection to Ecco

use Win32::OLE qw(in);
#use Win32::OLE::in;

my($ecco, $sel, $item);

$ecco = Win32::OLE->new("VBEcco.EccoApp", 'Quit');
$ecco->Register("xxxxxxxx", -1, -1);

$sel = $ecco->Selection;

foreach $item (in($sel)) {

    print $item->Text, "\n";
    print $item->FolderValue("Net Location") , "\n";

}

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

Visual Basic version:

Option Explicit

'Place in Declarations
Dim Ecco As EccoApp
Dim EccoSelection As EccoItems
Dim EccoItem As EccoItem

Private Sub Form_Load()

Set Ecco = New EccoApp
Ecco.Register "xxxxxxxx", -1, -1

' Get Selections
Set EccoSelection = Ecco.Selection

'if there is a first selection, print out the text, and figure out
whether there is an associated URL in the Net Location folder

For Each EccoItem In EccoSelection
    MsgBox ("Text: " + EccoItem.Text)
    MsgBox ("NL: " + EccoItem.FolderValue("Net Location"))
Next


End Sub


From mjais at web.de  Fri Apr 12 07:03:25 2002
From: mjais at web.de (Markus Jais)
Date: Fri, 12 Apr 2002 13:03:25 +0200
Subject: Python and databases
References: <UByt8.1138$4D6.194@rwcrnsc52.ops.asp.att.net>
Message-ID: <k8zt8.1$vr4.340@news.ecrc.de>

John Purser wrote:

> I'm working on my first real project with python and it's a client
> server/database program.  I'm looking  for a good tutorial on python and
> databases.  Anybody got a good link they'd care to share?
> 
> Thanks,
> 
> John Purser

check out the homepage of the db sig

http://www.python.org/sigs/db-sig/

there you will find some useful stuff

markus


From jacob at boris.cd.chalmers.se.cd.chalmers.se  Thu Apr 11 11:23:06 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 11 Apr 2002 15:23:06 GMT
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com> <p34s8.5367$d7.1394275@newssrv26.news.prodigy.com>
Message-ID: <a949oq$it2$1@nyheter.chalmers.se>

The really useful thing would be to have a platform independent way of
invoking a print job. Here are some ideas for how to do it:

1. Assume Acrobat reader available.

2. Use the Reportlab tools or some other way of generating pdf.

3. Have a pythonlib function that handles the platform specific invocation,
   causing Acrobat Reader to print the document.

Under Unix, you could have a number of fallback options, like xpdf and
pdf2ps.
How to handle this on a PC or a Mac is way beyond my level of knowledge.

I'd be willing to work on the Unix side of such a library. Any takers for
Mac and PC?

There are of course alternatives that do not involve generating PDF, but
from what I understand, they will be way more complex.

Jacob Hall?n
-- 


From shippy at nmt.edu  Fri Apr 26 18:49:13 2002
From: shippy at nmt.edu (Jeff Shipman)
Date: Fri, 26 Apr 2002 16:49:13 -0600
Subject: cgi module question
References: <3CC9A3E2.9060803@nmt.edu> <kud6wmb6wt.fsf@lasipalatsi.fi> 	<3CC9BC7F.5020305@nmt.edu> <1019860498.2058.1458.camel@localhost>
Message-ID: <3CC9D969.2040508@nmt.edu>

Whew. Thanks for telling me about StringIO! I
was getting ready to open pipes and stuff...

Just to make sure, this is the code I'm using.
Please tell me if I'm making any incorrect
assumptions:

# Grab query string and process form
if os.environ['REQUEST_METHOD'] == 'POST':
    qs = sys.stdin.readline()
    sio = StringIO.StringIO(qs)
    form = cgi.FieldStorage(sio)
    sio.close()
else:
    qs = os.environ['QUERY_STRING']
    form = cgi.FieldStorage()


Thanks again.

Ian Bicking wrote:
> On Fri, 2002-04-26 at 15:45, Jeff Shipman wrote:
> 
>>Uh, no, not really because it's consumed by
>>FieldStorage(). If I consume it before that, then
>>there's FieldStorage() can't grab it. I still
>>need to use the cgi module, but I also need
>>to get the POST query string.
> 
> 
> You can read in the POST string, and then pass it to the cgi module (as
> a file-like object with StringIO).  There's an optional argument to
> FieldStorage -- look at the source to find it.
> 
>   Ian
> 
> 
> 


-- 
Jeff "Shippy" Shipman     E-Mail: shippy at nmt.edu
Computer Science Major    ICQ: 1786493
New Mexico Institute of Mining and Technology
Homepage: http://www.nmt.edu/~shippy



From aahz at pythoncraft.com  Fri Apr 26 00:53:03 2002
From: aahz at pythoncraft.com (Aahz)
Date: 26 Apr 2002 00:53:03 -0400
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz>
Message-ID: <aaamff$am5$1@panix2.panix.com>

In article <3CC8D767.599720 at cosc.canterbury.ac.nz>,
Greg Ewing  <greg at cosc.canterbury.ac.nz> wrote:
>Julian Tibble wrote:
>> 
>> erm... If there is no copyright then there is no *need* for copyleft.
>
>But if RMS would be happy with a world in which there was simply no
>copyright, he wouldn't bother with the GPL -- he could get the same
>effect just by putting all the stuff he writes in the public domain.
>
>The fact that he hasn't done that means he *does* want *some* kind of
>intellectual property laws.

Not quite.  The problem with public domain is that someone could make a
small modification and then assert IP ownership over the corpus.  The
Timbot rightly says people over-worry about this, but it's definitely a
problem in some cases.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From neuhauser at mail.cz  Tue Apr 30 09:35:05 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 15:35:05 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1>
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1>
Message-ID: <20020430133505.GN325@roman.mobil.cz>

> From: Duncan Booth <duncan at NOSPAMrcp.co.uk>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Tue, 30 Apr 2002 12:39:30 +0000 (UTC)
> 
> Bernhard Herzog <bh at intevation.de> wrote in 
> news:6qn0vlbcpk.fsf at abnoba.intevation.de:
> 
> >>     Ugh, that looks to me as appropriate as join() being a string
> >>     method / module function.
> > 
> > Exactly. I still think that join as a function makes more sense than as
> > a string method and for the same reasons as stated for chr above. Even
> > if join is implemented as a string method.
> 
> I prefer writing
>    str.join(' ', aList)
> rather than
>    ' '.join(aList)
> 
> When the string is a literal I find the former reads much better. YMMV

    IMO that's because join being a string method is just awkward. How
    does this read:
    aList.join(' ')
    After all, you tell the list to give you it's concatenation, using
    the argument as the glue.
 
> Sometimes though I take advantage of it being a method:
>    joinlines = '\n'.join
>    ...
>    s = joinlines(aList)

    This is not a method. You masked it as a global function.
 
> Note that there was a suggestion on the python-dev list to move the 
> remaining objects in the string module into the str type and then deprecate 
> the string module. So str.hexdigits instead of string.hexdigits etc.

    That should be done, of course.

-- 
FreeBSD 4.4-STABLE
3:32PM up 1 day, 5:57, 11 users, load averages: 0.01, 0.03, 0.00




From nospam at nonesuch.com  Fri Apr 19 20:12:33 2002
From: nospam at nonesuch.com (David Smith)
Date: Fri, 19 Apr 2002 17:12:33 -0700
Subject: Tkinter.Text.index() puzzlement
Message-ID: <3CC0B271.C76D3E95@nonesuch.com>

I am trying to use the "index" method to remember places in a Text window, but
it doesn't seem to work properly.  Here's a sample.


from Tkinter import *
root = Tk()
text = Text(root)
text.pack()
print "Beginning: end at %s" % text.index(END)
text.insert(END, "Line 1")
print "After line 1: end at %s" % text.index(END)
text.insert(END, "\nLine 2")
print "After line 2: end at %s" % text.index(END)
text.insert(END, "\nLine 3")
print "After line 3: end at %s" % text.index(END)
text.insert(END, "\n")
print "Beginning line 4: end at %s" % text.index(END)
root.mainloop()


This fragment puts up the text window, which displays:
Line 1
Line 2
Line 3

The printed output says:
Beginning: end at 2.0
After line 1: end at 2.0
After line 2: end at 3.0
After line 3: end at 4.0
Beginning line 4: end at 5.0

I don't understand why the first line is claimed to be number 2,
nor why the character positions are all zero.  Is this normal?

This is with Python 2.1 under Debian.

-- 
David Smith
drs at labs dot agilent dot com  <- hopefully doesn't trigger spambots


From lists.ASkwar at DigitalProjects.com  Thu Apr 25 19:17:32 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 01:17:32 +0200
Subject: shipping a python app
In-Reply-To: <upfv8.14124$hl2.9868@atlpnn01.usenetserver.com>
References: <a9js5c$2m9$1@panix2.panix.com> <upfv8.14124$hl2.9868@atlpnn01.usenetserver.com>
Message-ID: <20020425231732.GJ21205@teich.Garten.DigitalProjects.com>

?Steve Holden? sagte am 2002-04-17 um 10:10:37 -0400 :
> Gordon McMillan's installer package is almost certainly what you want.
> Gordon's web page on
> 
>     http://www.mcmillan-inc.com/install1.html

Hm, I'm also looking for a way to ship a Python app.  I installed the
5b3.1 package on Linux running Python 2.2.  When I start the Build
phase, I just get this error message:

[askwar at teich Installer]$ python Build.py SPEC/DeDuper.spec 
Traceback (most recent call last):
  File "Build.py", line 688, in ?
    build(sys.argv[1])
  File "Build.py", line 23, in build
    rthooks = eval(open(os.path.join(HOMEPATH, 'rthooks.dat'), 'r').read())
  File "<string>", line 1
    {
     ^
SyntaxError: invalid syntax

Well, why is that?

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 18 hours 16 minutes




From bokr at oz.net  Sat Apr 13 22:51:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 02:51:16 GMT
Subject: dos promt simulation
References: <6ac1ecad.0204130435.93d0ad3@posting.google.com> <Xns91EFBDC8F2FCcliechtigmxnet@62.2.16.82> <6ac1ecad.0204131339.56928640@posting.google.com>
Message-ID: <a9aqr4$moh$0@216.39.172.122>

On 13 Apr 2002 14:39:58 -0700, theivo at abv.bg (I.C.) wrote:

>> there is also popen2, popen3 etc. they give you a file for stdin where you 
>> can write to.
>
>Maybe popen2\3 could help, but I think, I don't use them correctly :
>
>import os
>dos_input, dos_output = os.popen2('time')
>
>for line in dos_output.readlines(): # Current time is 22:14:43,24
>        print line                  # Enter new time:
>
>dos_input.write(raw_input())        # User enters the new time
     ^^^^^--??
>
>Where is my mistake?
>
I would guess that's probably one.

Regards,
Bengt Richter


From jyrinxat at mindspringdot.com  Sun Apr 14 22:35:22 2002
From: jyrinxat at mindspringdot.com (Jyrinx)
Date: Sun, 14 Apr 2002 19:35:22 -0700
Subject: Failure when profiling XML processing
References: <pan.2002.04.13.20.51.16.512445.4898@mindspringdot.com> <m3lmbqwrtl.fsf@mira.informatik.hu-berlin.de>
Message-ID: <pan.2002.04.14.19.35.21.795236.13012@mindspringdot.com>

>> (Is there a more specific place where I should be asking about
>> libraries?)
> 
> The XML-SIG would be the appropriate place.

Gotcha; I'm forwarding my post there. Thanks!

Jyrinx
jyrinxat at mindspringdot.com


From arthur.siegel at rsmi.com  Wed Apr 10 14:38:27 2002
From: arthur.siegel at rsmi.com (arthur.siegel at rsmi.com)
Date: Wed, 10 Apr 2002 14:38:27 -0400
Subject: What has PEP 285 done to us?
Message-ID: <OFB45E86D1.7C0EEDEB-ON85256B97.0065F037@RSMi.com>

</F> asks

  >>Arthur Siegel wrote:

   >> never has and never will speak for the core development
    >> team.

>and who exactly are you speaking for?

Do you mean who signs my paycheck?

Let me look....

P...A..U..L  R..U..B..I..N

But its made out for $0.00

Art








From mathieu543 at hotmail.com  Tue Apr  9 16:56:09 2002
From: mathieu543 at hotmail.com (Imer)
Date: 9 Apr 2002 13:56:09 -0700
Subject: class callable with any method name
Message-ID: <9d095c5a.0204091256.4d88a246@posting.google.com>

Hi, I'm trying to find a way to implement a class that would allow any
method name to be called.
When that happens, a special method would be called and the name of
the method that was originally called would be passed as an argument.

For instance:
########
class A:
  def __method_called__ (self, method_name, method_argument):
    print "The method", method_name, "was called with the argument",
method_argument
########

Then the following code:
########
a=A()
a.this_is_a_random_method(arg=1)
a.yet_another_random_method(param=3)
########

Would print out:
########
The method this_is_a_random_method was called with the argument
{'arg':1}
The method yet_another_random_method was called with the argument
{'param':3}
########

Is there a way to do that ?

Thanks.

Mathieu.


From jdhunter at nitace.bsd.uchicago.edu  Wed Apr 24 10:03:09 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Wed, 24 Apr 2002 09:03:09 -0500
Subject: flatten tuple
References: <m2adrvnxiv.fsf@mother.paradise.lost> <mailman.1019594863.1719.python-list@python.org>
Message-ID: <m23cxl8a8y.fsf@video.paradise.lost>

>>>>> "holger" == holger krekel <pyth at devel.trillke.net> writes:

    holger> Hello John, because i have related interests i have taken
    holger> the time to write a very small (one screen) and robust
    holger> more general 'SubstDict'. you can use it like this:

Thanks to all for really helpful suggestions.  Duncan's code worked
great -- Bengt's even does the recursive check.  I have upgraded to
2.2, and will study your examples closely to gleam some knowledge from
them.

Thanks again,
John Hunter


From anthony at interlink.com.au  Tue Apr  2 22:26:10 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 03 Apr 2002 13:26:10 +1000
Subject: 2.1.3 release plan for next week...
Message-ID: <200204030326.g333QAW16029@mbuna.arbhome.com.au>

With the recent evil evil bugfix that can cause Zope to crash
with 2.1.2 squished, I plan to cut a version 2.1.3 early next
week. So, if you know of any critical bugs or major major 
annoyances* in 2.1.2, now is the chance to speak up!

Anthony

* the lack of type/class unification in 2.1.2 is an annoyance,
yes, but that's not going into 2.1.3.



From nospam at bigfoot.com  Fri Apr  5 13:25:18 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 5 Apr 2002 20:25:18 +0200
Subject: rowcount in dbi w/ odbc
References: <a8iv0i$56r$1@news.ycc.yale.edu>
Message-ID: <a8kpne$1b5b$1@norfair.nerim.net>

Hi,

Google for "mxOdbc" and use it in place of the ActiveState/win32all odbc
(unsupported) subpackage.
This package supports DBI 2.0 API spec (richer) and is still supported by
its author.
mxOdbc is free for non commercial use.

HTH

--Gilles

"don" <donald.braman at yale.edu> a ?crit dans le message news:
a8iv0i$56r$1 at news.ycc.yale.edu...
> I'm new to Python programming. I am trying to gain access to specific





From chris.gonnerman at newcenturycomputers.net  Wed Apr 24 08:24:39 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 24 Apr 2002 07:24:39 -0500
Subject: atoi
References: <53A7943A5BD8D411B6930002A5073155022D748D@bgsmsx90.iind.intel.com>
Message-ID: <005d01c1eb8b$0594bda0$0101010a@local>

----- Original Message -----
From: "Mariappan, MaharajanX" <m_mariappanX at trillium.com>


> Hi all,
>
> simple......but need to know
>
> is there any python buildin method whihc will convert strings into
integer,
> like atoi in c lib.?
>
> "234" should be converted into interger value 234
>
> but the string function ord(character) convert it into equvalent ASCII
code

Python 2.1.2 (#31, Jan 15 2002, 17:28:11) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> int("234")
234

Is that what you're looking for?

The float() function works similarly but returns a floating-point value.

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






From vvainio at tp.spt.fi  Wed Apr 17 04:06:41 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 17 Apr 2002 01:06:41 -0700
Subject: bug in build? configuration --with-cxx ends up using PY_CFLAGS instead of CPPFLAGS
Message-ID: <ad496f8.0204170006.596b6818@posting.google.com>

I need to create a c++ - friendly version of python 2.2.1 (long story
short: hp-ux11, omniORB[py]). When I do ./configure --with-cxx=aCC,
the Makefile that gets generated has the following rule:

Modules/ccpython.o: $(srcdir)/Modules/ccpython.cc
	$(CXX) -c $(PY_CFLAGS) -o $@ $(srcdir)/Modules/ccpython.cc

So it uses PY_CFLAGS, while it should IMHO use CPPFLAGS (it appears to
work properly with CPPFLAGS).

So, the Makefile.pre.in should apparently be altered? Why has this not
been done? Would it break something on other platforms?


From mgilfix at eecs.tufts.edu  Thu Apr 18 14:00:21 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 18 Apr 2002 14:00:21 -0400
Subject: problems with Perl; changing to Python
In-Reply-To: <MPG.17291f2bfcb85301989684@news.rz.uni-karlsruhe.de>; from merkosh@hadiko.de on Thu, Apr 18, 2002 at 07:22:04PM +0200
References: <MPG.17291f2bfcb85301989684@news.rz.uni-karlsruhe.de>
Message-ID: <20020418140021.I3177@eecs.tufts.edu>

  Hiya. I did the same switch myself not too long ago. I tend to
use Python now for my general purpose programming and am much
happier with it than perl. I've written large programs in perl and
manage an OS projects that has over 10000 lines of perl (check out
http://peep.sourceforge.net) or Net::Peep on CPAN. I'll interperse
comments through out...

On Thu, Apr 18 @ 19:22, Uwe Mayer wrote:
> Perl is great for scripts. Perl is great for programs with less than 
> perhaps 200 lines. A program should be aware of its memory usage; for a 
> script I don't care.

  I'm not sure I agree with this statement so you should be weary of
that assertion. Perl can be fine for large programs but you have to
be disciplined (i.e, always 'use strict' and avoid using ridiculous
constructs). However, perl cannot currently compile to bytecode and
thus has a long start up time for large applications.

  Perl has also done a lot of stuff right that Python is still solving
in its infancy. Take the perldoc system for example or the maturity
of something like CPAN. The beauty of perl is that there is a CPAN
module for almost anything... not so for python.  However, the python
standard library is considerably richer than the perl standard library
IMO.

  The nice thing about python is that the syntax is always clear and
causes people to use a similar style when writing scripts, making
things easier to understand. Your programs tend to be much more
maintainable as a result. Python also compiles to bytecode (can even
do some optimizations) and executes that byte code in the interpreter,
much the way java processes byte code as well. And there's more but
I'll address your specific issues.

> I need a language that can be used to write programs with. GUI, and 
> commandline tools. When the program run in background they shouldn't 
> take up more memory than necessary.
> F.e. I wrote a Perl script checking for changes on homepages, expiry 
> date, etc in regular intervals. This "program" is 600 lines long and 
> uses 7mb of ram... too much in my oppinion.
> Can I do that in Python?

  I have written several gui tools in python. Never in Perl
though.  There are several gui packages. Some come with python as
standard, such as Tkinter (which is very portable but not very
featured). Alternatively, you can use wrappers around packages such as
pygtk (around GTK) or wxPython (around wxWindows).

  I don't find that python programs hog memory. It really depends on
how you construct your app. Like all interpreted languages, you will
have the overhead of loading the intepreter into memory but I don't
think that's really of any import.

  Using lines of code to gauge memory usage isn't very
useful. Depending on what you're doing, a simple 5 line program can
use up gobs of ram. I think python is more efficient than perl though
as everything in python uses references, and not so in perl.

> I'll need to write programs which have to be extensible. The program 
> may, during calculation require up to 500 mb ram. Fine with me. On the 
> other hand I'll neet to write something working in background and then 
> (depending on its task and size) 1-2 MB should be enough. 
> What about Python here?

  You'll love python here. Python classes are easily extensible and very
maintainable. Easy to add new features to classes. I'm not sure what you
mean about ram vs. extensibility though. If you mean memory management,
then that all depends on what sort of task you're doing and is very
much program dependent.

> Installing Perl modules was a torture. I am sitting behind a firewall 
> and in order to install the modules (compiling them did never work on my 
> Windows 2k mashine) I had to run my own DNS server and tunnel the 
> connection. Will that work better with Python? (f.e. Proxy support would 
> be sufficient!)

  Well, python supports modules but doesn't have a CPAN. Packaging
of python modules was introduced recently (roughly as of 2.0 or so)
and we're still getting that stuff straight. But you can very easily
download any package and install it. Many programs work this way
(install their libraries into the site-package directory). I've also
never had any problems using python on win32. Never tried perl though.

> And for the GUI programming: I used Perl/Tk previously, so Tkinter would 
> be most fammiliar, I guess. What other GUI packages can Python offer me 
> that will work (perhaps even in compiled .exe format) on windows 
> mashines? What about features like "minimize to tray" or "always on 
> top" that I wasn't able to accomplish in Perl/Tk?

  Think I answered this but again: pygtk and wxPython are obvious choices.
If you're programming in windows (and you sem to want to), then wxPython
is the way to go. It has all sorts of features that Tk doesn't include.
I highly recommend it.

> I need a language where I can rely on OOP if I need to. A language where 
> there's a difference between class- and instance- varriables. 
> How about multiple inheritance (perl, c++) or interfaces (jave) in 
> python?

  Python uses OOP extensively and using clean syntax (not like the
hack that is OOP in perl). Inheritance is very straight-forward and
understable.  You can also do multiple inheritance and although there
is no "interface" concept, it's easy to write an abstract class that
accomplishes the same thing.

> Are there any applications written in Python?
> Am I right here with Python or will I have to stick to a hybid C++/Java 
> solution?

  Sure. Go to sourceforge.net and search for python apps. I have
a 5000 GUI ftp program myself (ftpcube.sourceforge.net) that uses
pygtk. Has some outdated screenshots if you want to check it out.
Doesn't run on windows yet since the windows port of GTK doesn't
support threading but runs nicely on all *nixes. I have some others
that I've written for research - particularly a graphing program using
wxPython.

  Obviously with my propaganda, I'm supporting python over
C++/JAva. I'm not a big fan of Java since I feel that the language is
clunky (from a programming perspective). You might also want to
check out the Jython project as it might interest you.

  Hope that helps.

              -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From lac at strakt.com  Mon Apr 22 10:23:50 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 22 Apr 2002 16:23:50 +0200
Subject: Science And Math Was: Python's Lisp heritage 
In-Reply-To: Message from Andy Gimblett <gimbo@ftech.net> 
   of "Mon, 22 Apr 2002 13:41:26 BST." <20020422124126.GD23706@andy.tynant.ftech.net> 
References: <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>  <20020422124126.GD23706@andy.tynant.ftech.net> 
Message-ID: <200204221423.g3MENpdj026909@ratthing-b246.strakt.com>

> On Mon, Apr 22, 2002 at 12:57:38PM +0100, Gon?alo Rodrigues wrote:
> 
> > You should understand that this is a philosophical viewpoint. The
> > mathematical field is roughly divided in the Platonist field, those who
> > believe that mathematical objects exist somewhere out there in space (to
> > quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
> > field of those who believe mathematics is essentialy a game we play on
> > paper with symbols with no "real" meaning attached to it besides the
> > convenience in describing natural phenomena.
> 
> To put this another way, ask yourself where you stand on this
> question: is mathematics discovered or invented?
> 
> -Andy

Those aren't the only two choices out there.  From the 'saving your
cake and eating it too department' -- All mathematics is
invented, but the human mind is only capable of inventing things that
in some way correspond to 'things out there' (whatever that means,
given that this perfectly corresponds to some religious thought on 
the nature of the soul, which is generally though of as 'in here'
rather than 'out there').  Here we start approaching Steven Pinker's
Theory of Language, as expounded in 'the Language Instinct' or Chomsky's -
see 'Language and Problems of Knowledge' or 'Rules and Representations'.

Given that people have been attacking Chomsky as practicing something
which is fundamentally non-scientific (by which they mean that it
is impossible to construct experiments to falsify his claims) for
half a century now, it is possible that the 'saving your cake and
eating it too' assertion is both false and not provably false.

Laura Creighton




From akuchlin at crystal.mems-exchange.org  Mon Apr  8 12:14:11 2002
From: akuchlin at crystal.mems-exchange.org (A.M. Kuchling)
Date: 08 Apr 2002 16:14:11 GMT
Subject: Need help with distutils
References: <mailman.1018083739.10368.python-list@python.org> <Xns91EA988AA8260andy47halfcookedcom@203.109.252.31>
Message-ID: <slrnab3gei.1i0.akuchlin@crystal.mems-exchange.org>

In article <Xns91EA988AA8260andy47halfcookedcom at 203.109.252.31>, 
   Andy Todd wrote:
> notice that the pattern you pass to global-exclude has to be a Unix-style 
> "glob" pattern. In that notation '.' is a special character so using 
> '.cvsignore' doesn't actually match your files ;-)

No, '.' is not special in glob notation.  

> Also note that this isn't properly implemented for distutils on Windows or 
> Mac. If you try and package a source distribution on Windows, for instance, 
> you will get all of your CVS directories in the .zip file as well.

Can you please file a bug for this?

--amk


From eric at enthought.com  Wed Apr  3 17:15:11 2002
From: eric at enthought.com (eric)
Date: Wed, 3 Apr 2002 17:15:11 -0500
Subject: CLAPACK Build on Mac w/ NumPy?  Help.
References: <030420021516186189%pecora@anvil.nrl.navy.mil>
Message-ID: <0f0001c1db5d$06173430$6b01a8c0@ericlaptop>

Hey Lou,

You might look at SciPy.  The linalg2 module (in CVS) uses f2py to wrap both
Fortran Lapack and C lapack.  If linalg2 (its looking pretty good, but still
under development) isn't suitable for your purposes, then at least they might
provide some hints.

eric

----- Original Message -----
From: "Louis M. Pecora" <pecora at anvil.nrl.navy.mil>
Newsgroups: comp.lang.python
Sent: Wednesday, April 03, 2002 3:16 PM
Subject: CLAPACK Build on Mac w/ NumPy? Help.


> Has anyone been successful getting CLAPACK working with Numeric Python?
> I have the source and a codewarrior compiled library for CLAPACK, but I
> have not clue how to integrate the two and have the real CLAPACK
> replace the lite version that comes with Numeric.Py.  I checked the
> docs, but found nothing.  Maybe I'm overlooking something.
>
> Any clues greatly appreciated.
>
> --
> -- Lou Pecora
>
>   - My views are my own.






From huaiyu at gauss.almadan.ibm.com  Tue Apr  9 14:34:10 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 9 Apr 2002 18:34:10 +0000 (UTC)
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk>
Message-ID: <slrnab6d12.dvl.huaiyu@gauss.almadan.ibm.com>

On Tue, 09 Apr 2002 10:48:31 +0200, Max M <maxm at mxm.dk> wrote:
>So when Guido wrote that the post was an overlong piece of FUD I more or 
>less agreed. But I was rather surprised by his choice of words, and 
>found it to be a rather rude brush-of. And I do believe that if I had 
>made a lot of work making an argument like Lauras, that I would feel a 
>bit bad about it being turned down that way.

I was also surprised by Guido's unfortunate choice of word and feel it rude.
But, just to give him the benefit of doubt, it is likely that he meant
something like

"""This is fear, uncertainty and doubt that is not important since I have
already thought about all these issues thoroughly"""

instead of something like

"""This is fear, uncertainty and doubt that is of the same variety that comes
regularly out of a company in Redmond""".

Whatever his meaning is, this is unfortunate, since most people read the
word FUD with a hint of bad intention.

Let's get over this.

Huaiyu


From sjmachin at lexicon.net  Thu Apr 18 20:17:58 2002
From: sjmachin at lexicon.net (John Machin)
Date: 18 Apr 2002 17:17:58 -0700
Subject: help splitting strings
References: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com> <mailman.1019147737.28803.python-list@python.org>
Message-ID: <c76ff6fc.0204181617.3a79df1e@posting.google.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> wrote in message news:<mailman.1019147737.28803.python-list at python.org>...
> You definitely want to use the re module and finall
> here.  That way you can have tighter control over what
> constitutes a field.  The difficulty here is crafting your
> regular expression so it matches what you want. 
> 

Sorry to be flatly contradictory, but you definitely *don't* want
to use regular expressions to try to unravel CSV format, whether (de
facto) standard or not. Therein lie both madness and slow processing.
Take Emile's advice and use Dave Cole's CSV module. If you have data
that it doesn't handle (i.e. not "standard") then post again or send
me e-mail -- I am lashing together a module that handles non-standard
variations of CSV format.


From sholden at holdenweb.com  Mon Apr  8 09:08:40 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 09:08:40 -0400
Subject: ANN: Twisted 0.16.0
References: <mailman.1017787747.12194.python-list@python.org> <3CAA8CFF.B6EDDCDE@alcyone.com>
Message-ID: <8Fgs8.27735$Sw5.8757@atlpnn01.usenetserver.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CAA8CFF.B6EDDCDE at alcyone.com...
> Michal Wallace wrote:
>
> > Thanks for trying to answer my question. I'm afraid I still
> > don't follow the logic of the registrations though.
>
> Are you aware the announcement was posted on April Fool's Day, and all
> the responses you've been getting are playful and unclear?

Which, IMHO, simply mirrors the documentation ;-) Much is implicit.

>  You've been had.
>
Well, a good April Fool's joke should obviously *be* an April Fool's joke,
so ultimately the joke's on the Twisted team.

regards
 Steve






From logiplexsoftware at earthlink.net  Mon Apr  1 12:08:00 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 09:08:00 -0800
Subject: GUI programming
In-Reply-To: <20020401020912.GE333@lilith.hqd-internal>
References: <mailman.1017437568.25402.python-list@python.org>
	<a84561$p5g$1@wanadoo.fr>
	<slrnaabbct.npb.cpbotha@crabtree.cpbotha.net>
	<4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com>
	<20020401020912.GE333@lilith.hqd-internal>
Message-ID: <20020401090800.3353ad9b.logiplexsoftware@earthlink.net>

On Mon, 1 Apr 2002 04:09:12 +0200
Gerhard H?ring wrote:

> * Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> [2002-03-31 12:31
+0100]:
> > "Charl P. Botha" <cpbotha at i_triple_e.org> wrote:
> > 
> > >With wxPython, one has to get, configure and build wxWindows first
(and this
> > >does take quite a while).  
> > 
> > Nonsense.
> > 
> > 1. Download
> > 2. Run Setup
> > 3. Use it.
> > 
> > The whole operation takes less than 5 minutes.
> > 
> > Well - it does on Windows, anyway.
>  
> Obviously you've never installed glib, gtk, wxGTK then wxPython all in
> /usr/local on a Unix system. Not speaking all the dependencies if
wxPython if
> you really want to run all of the demos. This can easily take hours.
> 
> If you're lucky, however, your distribution already provides all the
required
> packages in an easy-to-use package. But from the ones I know, only
FreeBSD has
> the most current wxPython and wxGTK versions in the ports tree. Even
Debian
> sid, which is quite bleeding edge, still had wxPython 2.2.7 last time I
checked
> (a few days ago).

wxPython is remarkably easy to install on both Windows and RPM-based Linux
systems.  I haven't tried it on anything else, but I doubt even a
tarball-based install would be very difficult.

Building from CVS would be more difficult, but that's rarely necessary.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From nhodgson at bigpond.net.au  Mon Apr  8 08:31:48 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Mon, 08 Apr 2002 12:31:48 GMT
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com><20020408120242.41456ace.wilk@flibuste.net><a8rqh2$9h5$03$1@news.t-online.com> <20020408123251.48eab8d6.wilk@flibuste.net>
Message-ID: <U4gs8.10597$vc3.34524@news-server.bigpond.net.au>

William Dod?:

> When i see boa constructor i'm a litle bit afraid... but it doesn't mean
> that it's a problem of wxPython.

   Boa Constructor is still in a pre-release state - version 0.1.0 Alpha
downloadable and 0.1.1 Alpha in CVS. Riaan is quite conservative in his
version numberings - I'd call the current code version 0.9. At this stage of
development, large features are still appearing such as an out-of-process
debugger so you shouldn't expect great stability.

> Also when i make some tests of wxPython
> with idle or pythonwin i have often big scratch (whitout python
> log). But have never theses scratch if i run my script from the
> command line.

   There are often problems caused by running a program that uses a GUI
toolkit within the process of an IDE using the same or another toolkit. The
solution here is to use an IDE that features out-of-process debugging such
as HAP, Wing IDE, or Komodo. PythonWin and IDLE currently only support
running programs inside the IDE process.

   Neil




From jessica at ikotsi.ac.at  Thu Apr 25 16:29:17 2002
From: jessica at ikotsi.ac.at (Jessica Felix)
Date: Thu, 25 Apr 2002 20:29:17 GMT
Subject: Art and all that Jazz:  until Mustapha pours the cats inadvertently, Afif won't look any easy shores
Message-ID: <comp.lang.python.c64ed659@ikotsi.ac.at>

It's very blank today, I'll reject daily or Rachel will seek the jugs.

"ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
that incorporates the two passions of its owner, Jessica Felix -
art and jazz.

Jessica is an artist whose medium is jewelry. She creates what she calls
"spirited art," specializing in images from indigenous cultures around the world.

http://www.artandallthatjazz.com/index-2.shtml

119A Plaza Street
Healdsburg, CA, USA 95448
1-707-433-7900

They generally order inside Moammar when the ugly carrots measure through the rude arena.



















When will you pull the bizarre younger tickets before Osama does?






From spam at yourself.com  Fri Apr 26 06:01:27 2002
From: spam at yourself.com (Ingo Blank)
Date: Fri, 26 Apr 2002 12:01:27 +0200
Subject: Killing a process under Windows NT/2K
Message-ID: <aab8hn$o4s$05$1@news.t-online.com>

Hi,

I need to kill a process under Windows (NT/2K), whose PID is known.
Unfortunately os.kill() doesn't work under Windows.
How can I achieve this anyway?

Thanks in advance

--ingo
voy NG arktb QBG qr [ rot13 ]

PGP PublicKey
http://keys.pgp.com:11371/pks/lookup?op=get&search=0x18B44974





From xx758 at cam.ac.uk  Tue Apr 30 06:00:23 2002
From: xx758 at cam.ac.uk (Xiao-Qin Xia)
Date: Tue, 30 Apr 2002 11:00:23 +0100
Subject: Global Variable
Message-ID: <aalpqc$h9s$1@pegasus.csx.cam.ac.uk>

Hello,

If a module (for example, gmpy) is imported in the main.py, how can it be 
used by other modules (for example, mymodule) without being imported by 
each of the other modules? 

#---------main.py---------
import gmpy
import mymodule
...


#--------mymodule.py---------
#I want use gmpy here, do I have to import it again? can main.py add gmpy 
into a global dict?


Thanks,
Xiao-Qin


From jason at tishler.net  Fri Apr 12 08:11:41 2002
From: jason at tishler.net (Jason Tishler)
Date: Fri, 12 Apr 2002 08:11:41 -0400
Subject: python on cygwin
In-Reply-To: <01d801c1e1b8$ac962e40$6b01a8c0@ericlaptop>
References: <01d801c1e1b8$ac962e40$6b01a8c0@ericlaptop>
Message-ID: <20020412121141.GA948@tishler.net>

Eric,

On Thu, Apr 11, 2002 at 08:26:22PM -0400, eric wrote:
> I read a few mails about a rebase solution at:
> 
>     http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
> 
> However, after building and running this on the percribed files, I get:
> 
> $ rebase -d -b 0x68000000 -o 0x10000 cygXpm-X4.dll ...
> [snip]
> cygXpm-X4.dll: new base = 67fe0000, new size = 20000
> cygXpm-noX4.dll: new base = 67fc0000, new size = 20000
> cygbz21.0.dll: new base = 67f90000, new size = 30000
> cygcrypto.dll: new base = 67ee0000, new size = b0000
> ReBaseImage() failed with last error = 2
> 
> Can anyone point out the error in my ways, or point me to a fix?

Replace the argument list above with the cyg*.dll files in *your*
/usr/bin.  The "last error = 2" error message is caused by a file not
found condition.  A later version of rebase displays better error messages
that make this more obvious.

See the following for more (possibly useful) information:

    http://www.zope.org/Members/dgeorgieff/howto_zope_cvs_on_cygwin/

Otherwise, just follow Michael's suggestion and build _socket statically.

I will try to release a pre-built 2.2.1 ASAP.

Jason




From dsavitsk at e-coli.net  Mon Apr  8 00:21:00 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 08 Apr 2002 04:21:00 GMT
Subject: Newbie:  How to sound an alarm
References: <ucFMmso3BHA.1332@cpimsnntpa03>
Message-ID: <LU8s8.5403$d7.1425212@newssrv26.news.prodigy.com>

below are 2 solutions that showed up.  try google groups for getting old
postings.

-d


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:5aRq8.3571$d7.1106160 at newssrv26.news.prodigy.com...
> on win32 also try
>
> >>> import winsound
> >>> winsound.Beep(440,250)
>
> the first number is the frequency, the second is the duration.
>
> -d
>
>
> "Trond Gaarder" <trondga at -NOSPAM-ifi.uio.no> wrote in message
> news:a8gfr8$h4u$1 at maud.ifi.uio.no...
> > Don't know if this is the best/correct way of doing it in Python, but it
> > works for me (on a PC):
> >
> > print "\a"
> >
> > This should emit a beep from the pc. What it sounds like is highly
> > machine-dependent, however...
> >
> >
> > -Trond

"billludden" <billludden at email.msn.com> wrote in message
news:ucFMmso3BHA.1332 at cpimsnntpa03...
> Hi,
>     I am attempting to learn to program using Python (first language).
>      I would like to know how to make my box sound a tone or beep when my
> program finishes.
> I posted this query last week but I cannot access messages posted during
> that time (must be Windows).
> Thank You,
> Bill
>
>
>




From ajs at ix.netcom.com  Thu Apr  4 02:29:21 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Thu, 4 Apr 2002 02:29:21 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <000f01c1dbaa$70fc7100$0334fea9@carol>

Guido flame bait ending with -

>I think it's time I learn to say "yes" even 
>if the community doesn't see the light yet.

Do what you got to do.

Gee do I like to play with Python.

Gee does this place give me the willies.

Art
 









From fperez528 at yahoo.com  Mon Apr 15 13:57:34 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 11:57:34 -0600
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it> <a9a7sa$qd3$1@reader05.wxs.nl> <Xns91F179739F715duncanrcpcouk@127.0.0.1>
Message-ID: <a9f4c5$ihg$1@peabody.colorado.edu>

Duncan Booth wrote:

>>> def relatively_prime(a, b): return gcd(a, b) == 1
>> 
>> Thats funny, I wanted to do that but I don't know how to use filter
>> with a function that takes 2 variables. E.g., to get a list if all
>> numbers less than 10 relatively prime to 3, I would write
>> 
>>      filter(relatively_prime, range(1, 10)) # aaah, where does the 3
>>      go???
>> 
> If you want this readable then use a list comprehension:
>    [ n for n in range(1,10) if relatively_prime(n, 3) ]
> 

And if you insist on a lambda solution, it's not that bad (but at least to me, 
less readable than the list comprehension):

filter(lambda n:relatively_prime(n,3),range(1,10))

cheers,

f


From phd at phd.pp.ru  Thu Apr 25 10:30:33 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 25 Apr 2002 18:30:33 +0400
Subject: sql but... without server
In-Reply-To: <14Ux8.13604$z55.3648@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Thu, Apr 25, 2002 at 10:04:11AM -0400
References: <7fvfcu0vkc610fma8545v0nhdm7c6hem7q@4ax.com> <14Ux8.13604$z55.3648@atlpnn01.usenetserver.com>
Message-ID: <20020425183033.E14696@phd.pp.ru>

On Thu, Apr 25, 2002 at 10:04:11AM -0400, Steve Holden wrote:
> www.chordate.com

   The site died long ago. Gadfly sources are now in the Zope (www.zope.org)
sources.

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




From pyth at devel.trillke.net  Tue Apr 30 20:30:46 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 1 May 2002 02:30:46 +0200
Subject: Importing dynamically loaded modules (in packages)?
In-Reply-To: <aanbio$pk3$1@news1.xs4all.nl>
References: <aanbio$pk3$1@news1.xs4all.nl>
Message-ID: <20020501003046.GK16217@devel.trillke>

On Wed, May 01, 2002 at 02:06:47AM +0200, Irmen de Jong wrote:
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82234
> more or less confirms the code that I've created myself (in Pyro).
> But this code only works if the module does not come from a package.
> i.e. loading the .py file of "TestModule" and then "import TestModule"
> works fine, but loading "Package.TestModule.py" and then
> "import Package.TestModule" doesn't work.

> I've been toying around with creating package/modules structures
> with new.module, mimicing the package structure on disk, but
> even if I have this:
> 
> sys.modules['Package']= module 'Package'
>     module 'Package' . TestModule = the loaded module
> 
> "import Package.TestModule" still doesn't work: ""no module named TestModule""
> How can I correct this?

    sys.modules['Package.TestModule'] = TestModule 

after that you should be able to do

    import Package.TestModule

regards,

    holger




From h_schneider at marketmix.com  Wed Apr 10 09:41:19 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Wed, 10 Apr 2002 15:41:19 +0200
Subject: PIL + McMillan Installer
Message-ID: <a91f3r$4o7$06$1@news.t-online.com>

Hi,

I use Activestate Python 2.1 and McMillan Installer 5.
A script that uses PIL runs fine. But the one-file-distribution seems to
have a problem
with Image.open()

----
img = Image.open(File)
  File "C:\usr\python\PIL\Image.py", line 902, in open
    raise IOError, "cannot identify image file"
IOError: cannot identify image file
----

The .spec file has following settings:

---
a = Analysis(['C:\\usr\\python\\Installer\\support\\useUnicode.py',
'C:\\WORK\\Software\\Python\\test\\test.py'],
             pathex=['..\\lib','c:\\usr\python\pil'])
pyz = PYZ(a.pure)
exe = EXE( pyz,
          a.scripts,
          a.binaries,
          name=test.exe',
          debug=0,
          console=1,
          icon='myicon.ico')
---

Any ideas?

Thanks a lot,
Harald Schneider





From sjmachin at lexicon.net  Fri Apr 19 22:06:10 2002
From: sjmachin at lexicon.net (John Machin)
Date: 19 Apr 2002 19:06:10 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net>
Message-ID: <c76ff6fc.0204191806.406ec53b@posting.google.com>

claird at starbase.neosoft.com (Cameron Laird) wrote in message news:<48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4 at lp.airnews.net>...
> In article <3CBF9168.AAE0FC46 at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
> >"Frederick H. Bartlett" wrote:
> >> 
> >> I was recently introduced to OmniMark. One of our exercises was to take
> >> a plain text file of Hamlet and convert it to SGML.
> >> 
> >> So I did it in Python, too. But the best time I could get from Python
> >> was .57 sec, while OmniMark came in at .20 sec. What's the most
> >> efficient technique for Pythonesque regex-based text processing?
> >
> >Hmmm... how fast do you need it to be?  Sounds to me like 0.57 seconds
> >is pretty darned fast.
> >
> >Do you have specific goals, or are you just on a search for 
> >something faster?  Remember, "better is the enemy of good"
> >and the grass is always greener.
> >
> >(See http://www.seds.org/~chrisl/akin.html )
> >
> >-Peter
> 
> I was counting on Peter to write this.
> 
> Because it's correct, of course.  Moreover, you should
> know that Peter has payroll responsibilities.  He writes
> from a more practical perspective than other acquaintances
> might afford you.
> 
>

Another practical perspective:

When people mention excessive run-time to me, I ask them to rank the
severity of their problem on this scale:
1. Job doesn't finish before they get back to their desk with a fresh
mug of coffee.
2. Job has (exclusive access to database or excessive impact on
on-line response time) and doesn't finish by sun-up.
3. Job doesn't finish before the weekly cold back-up is due.


From DeepBleu at DeepBleu.org  Thu Apr 11 13:18:43 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Thu, 11 Apr 2002 12:18:43 -0500
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com> <3CB2FE25.86A37EA1@imk.fraunhofer.de> <ub6dji5plutpc5@corp.supernews.com> <SKYs8.19881$GM.6702@atlpnn01.usenetserver.com> <ubb56d6fvmpjf0@corp.supernews.com> <93jt8.1445$F43.775@atlpnn01.usenetserver.com>
Message-ID: <ubbhakehkh1l29@corp.supernews.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:93jt8.1445$F43.775 at atlpnn01.usenetserver.com...
> "DeepBleu" <DeepBleu at DeepBleu.org> wrote in message
> news:ubb56d6fvmpjf0 at corp.supernews.com...
> >
> [ ... ]
> > >
> > > Do you close the zipfile after you've finished writing it?
> > >
> > Do you mean text1.zip (from the code above) as the zipfile to be closed?
> > No, I am not!  I thought it was enough to close the ZipFile instance
> > (zip_archive).  Since text1.zip makes its appearance (first used) in:
> > >>>zip_archive = ZipFile("text1.zip", "w", 8)
> > How do I close the text1.zip (if that is the one you are refering to)?
> > DeepBleu
> >
> Ignore me. It was obvious in the code you gave that you are indeed
correctly
> closing the file after writing it.
>
> Since the zipfile format is binary, you should *not* be running Unix2DOS
to
> convert it -

Steve
I am not running Unix2DOS on the zipped text file (after zipping it).  I
would never really attempt to run Unix2DOS on a binary.  I am runnig
Unix2DOS on the text file itself.  Then, after that, I am zipping that text
file.
What I am doing is that I am taking a list of text files.  Run Unix2DOS on
each one of them, zip then, then FTP them to another Sun machine where
people can download them through a web broswer into their Windows PC's where
they can open them using WinZip usually.
Every text file is unzipping fine by the end users *except one*  Nothing  is
special about that one.  I looked at it.  There is just text.  This thing is
just freaky :-)
I applied the Python script on that text file only (unix2dos, zip, ftp) and
*it works*
I applied the Python script on it included with the other text files:  it
does not work and users report the 2-byte missing error!  The others:  They
work without a problem!
So for some reason, this text file in particular does not like running in a
batch process.  It is a text file with an attitude :)  What bugs me is that
I do not like such random things to arise.
Anyway, I know the above is not of much help in order to diagnose the
problem.  I'll just have to play with the script and the order by which the
files are being zipped to see if it makes any difference.
Thanks,
DeepBleu




From narnett at mccmedia.com  Sun Apr 21 11:08:08 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Sun, 21 Apr 2002 08:08:08 -0700
Subject: nested exceptions?
In-Reply-To: <a9uhvl$8v7$1@panix2.panix.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJGECJKDAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Aahz

...

> That is, if you're checking N items against a list with M items, you're
> gonna have poor performance algorithmically.  Note that I said "avoid";
> using 'in' against a list is sometimes the simplest and most
> straightforward approach, and if it's not being done in a loop, it won't
> cause too much of a performance hit.

So... although the following is such a short list that I don't think it
really matters (except that, come to think of it, it is in a loop), it's bad
form?

if self.item[0:3] in ('Mon','Tue','Wed','Thu','Fri','Sat','Sun'):
	self.item = self.item[4:]

If it's not obvious, this is stripping the day abbreviation from the
beginning of a timestamp that I'm converting to another format.

I can't quite see how to use "while... not" to do what I want, however.

Nick





From syver-en+usenet at online.no  Fri Apr 12 13:08:10 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 12 Apr 2002 17:08:10 GMT
Subject: A simpler super.
References: <u662x3u1j.fsf@online.no> <mailman.1018622124.15759.python-list@python.org>
Message-ID: <uzo083kxs.fsf@online.no>

Gustavo Niemeyer <niemeyer at conectiva.com> writes:

> > def Super(yourself):
> >     if isinstance(yourself, type): # for classes
> >         return super(yourself, yourself)
> >     return super(yourself.__class__, yourself)
> 
> I had the same idea a while ago. This can't be done. Think about what
> happens here:
> 
> class A:
> 	def __init__(self):
> 		pass
> class B(A):
> 	def __init__(self):
> 		Super(self).__init__(self) # <-- problem
> class C(B):
> 	def __init__(self):
> 		Super(self).__init__(self)
> 
> In this case, you'll get a loop, because self.__class__ is C, no
> matter where it is called.

Ahhh, I see. Shame, I guess it was too good to be true then :-)

-- 

Vennlig hilsen 

Syver Enstad


From grey at despair.dmiyu.org  Wed Apr 10 19:38:28 2002
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Wed, 10 Apr 2002 23:38:28 -0000
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <slrnab9j7m.hf2.grey@teleute.dmiyu.org>

On Thu, 11 Apr 2002 00:17:40 +0100, Robin Becker <robin at jessikat.fsnet.co.uk>
wrote:
> When I find this rubbish in M$oft's expensive software I feel outraged.
> Is it better because it's open source? I think not. 

    Why not?  From what I've read, in this case, you have to specify a compile
time option to have it be compiled in.  That is the exact opposite of
Microsoft which does the compiling and there is NO way for you to get it out.
IE, opt-in with a default of out vs. in with no way to opt-out.

    World of difference.  If you can't see that I suggest you reexamine your
priorities.

> hidden in python and we have geniuses for developers, what else may be
> embedded therein? 

    That's the joy, you can go look.  Why don't you just do that and let us
know, ok?

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------


From arturs at iidea.pl  Sat Apr  6 04:44:58 2002
From: arturs at iidea.pl (Artur Skura)
Date: Sat, 6 Apr 2002 09:44:58 +0000 (UTC)
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org> <pan.2002.04.06.11.23.11.491490.3709@thomas-guettler.de>
Message-ID: <slrnaath72.r4v.arturs@aph.waw.pdi.net>

Thomas Guettler wrote:
> BTW: I think it is called "I have python installed". But I am not a
> native speaker

You never know with these English-speaking people. When they've done
something, they say they're done. Maybe it's a kind of elision:
I'm (a ,,) python installed('' person).

Regards,
Artur
-- 
http://www.wolneprogramy.org
http://mf.gnome.pl


From ufechner at csi.com  Mon Apr 15 17:11:01 2002
From: ufechner at csi.com (Uwe Fechner)
Date: Mon, 15 Apr 2002 23:11:01 +0200
Subject: Looking for GUI Form Builder for Python
References: <a9en77$4vf8@imsp212.netvigator.com>
Message-ID: <3CBB41E5.8080901@csi.com>

Calvin wrote:
> Dear all,
>  Can I build Python TK GUI interface using a form builder. Anyone know such
> builder. I just want to build my interface in a visual way.
> -Calvin
> 
> 

Hello,

I think, that there is only one good solution:

Boa Constructor:

http://sourceforge.net/projects/boa-constructor/

It is a little bit hard to install, (you should use the newest version 
from cvs), but it's a really good IDE and open source.

So give it a try.

Regards:

Uwe



From starx at pacbell.net  Wed Apr 24 05:19:21 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 09:19:21 GMT
Subject: weak references and threads
References: <3CC653B1.9060701@vip.fi>
Message-ID: <Xns91FA17D7E4F8ARASXnewsDFE1@130.133.1.4>

pekka niiranen || Tue 23 Apr 2002 11:41:55p:

> Could somebody explain me in SIMPLE english what are:
> 
> 1)  weak references

A name that does not force the object it points to to exist.

Normally, our objects stay around for as long as we have a name for them. 
If we write:

a = [1]

That particular list (i used a list because defining a list guarentees a 
unique ojbect) will stay around for as long as the name 'a' exists. If 
this is in a function, the name will be removed at the end of the 
function, if it is in the first level of the file, it will stay around 
untill told to do otherwise. That is, we say:

del a

And the name 'a' is removed. If this variable has no other names at this 
point, it will be removed as well. If on the other hand, we do this:

a = [2]
b = a

We now have two names for the same object (python never copies unless 
asked. Any apparent copying done by an assignemnt is due to the observable 
properties of immutable types (strings, numbers, tuples), and expectations 
of behavior similar to other programming languages. This is the pointer-
ish stumbling block for python, it seems). If we then:

del a

The name 'a' is removed, but because the name 'b' still points at that 
object, the object is unaffected.

Now, let's take a hypothetical function weak(var), which returns a weak 
reference to whatever you pas to it as var. If we write:

a = [3]
w = weak(a)

We again have two names for one variable. But now if we:

del a

The system wil reclaim the [3] list object as if w did not exist. This is 
because weak references complety ignore and bypass the magic-behind-the-
scenes system called 'reference counting' that lets Python keep track of 
when objects acutally need to be removed. This means you genereally have 
to check and see if the object still exists each time you use it, or set 
up some object to be notified when it gets destroied, and do your own 
bookkeeping.

There is only one reason I know that you would want to use weak 
references, and that is in the creation of advanced, customized data 
structures. The point is to avoid something called 'circular references', 
which is probably best explained by an example. Let's take a realy basic 
class (this is right from my 2.2 IDLE prompt):

>>> class pointer:
	def __del__(self):
		print 'Object being removed now!'

All this does is print a message when the object is deleted. A simple 
control case shows how this works:

>>> a = pointer()
>>> del a
Object being removed now!

But if we create two of these objects:

>>> a = pointer()
>>> b = pointer()

And add a regular name to each that points to the other:

>>> a.target = b
>>> b.target = a

We can then remove both objects, and:

>>> del a
>>> del b
>>> 

Nothing. The system does not delete either one. It's like they've fallen 
into another dimension: they are still there, they still take up space in 
memory, but because we have no name for them we cannot ever reach them 
from the outside world. (It would be interesting to see if we could create 
a thread running on a set of objects and then remove all of those objects 
from the initating part of the program, but that's beyond what i'm trying 
to explain here). In reality, cyclic problems would be much more subtle, 
and tend to slowly leak memory until (if the program runs long enough), 
the whole thing fails. (At which point it would probably all be freed as 
the interperter exits: it cannot delete it earlier because it is not 
gaurenteed to be safe for the program, but it still watches over things 
enough to prevent permanent system memory pollution).

The best analogy for the problem that I can think of is this. In every 
released version of the old Infocom text adventure Zork (one), there was a 
problem like this, but this is what comes to mind. It goes something like 
this:

>look
You are in some room. There is an inflated life raft and a heavy golden 
coffin here.
>put coffin in raft
[Done].
>look
You are in some room. There is an inflated life raft here. The raft seems 
to contain a heavy golden coffin.
>put raft in coffin
[Done].
>look
You are in some room.

Each object can only exist in one other object's contents list, so each 
put operation is essentially a move instead of a copy and delete (so much 
the difference =). The principle is the same, however.

> 2)  threads

Er... it's late. I better get going =).

(Shortest answer: they are a way of making the computer run your program 
in two places at once, at the same time. They are often used to do long 
(longer than 1/2 a second) processes in GUI based apps, so the core of the 
program can continue to respond to user interaction while the work gets 
done.)

> and why or when are they needed.
> 
> Threads I have not used but the use of weak references
> I have seen recommended with wxpython's grid component.

If you have used pointers in C or such, weak refs share many of the 
hassles (minus worrying about segfaults). But both of these things are at 
least moderatly advanced, so it's ok not to have used them yet. Threading 
especially can be tricky if you don't have everything else down pat-- it 
has enough trickyness to it when you do.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From max at alcyone.com  Wed Apr  3 00:02:55 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 02 Apr 2002 21:02:55 -0800
Subject: ANN: Twisted 0.16.0
References: <mailman.1017787747.12194.python-list@python.org>
Message-ID: <3CAA8CFF.B6EDDCDE@alcyone.com>

Michal Wallace wrote:

> Thanks for trying to answer my question. I'm afraid I still
> don't follow the logic of the registrations though.

Are you aware the announcement was posted on April Fool's Day, and all
the responses you've been getting are playful and unclear?  You've been
had.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From pj at engr.sgi.com  Mon Apr  8 20:21:56 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 9 Apr 2002 00:21:56 GMT
Subject: newbie question about event based programming
References: <mailman.1018257016.24202.python-list@python.org> <%des8.59197$pT1.1704440@news1.tin.it>
Message-ID: <a8tc74$pm65o$3@fido.engr.sgi.com>

|> Event-driven programming is a great thing, but it's not supported
|> as well as it could and should be across every platform and toolkit
|> in the world:-(.

Seems like an opportunity for an enterprising coder ...

    Develop a Python module that provides file/directory
    change notification, in a reasonably operating system
    independent way, that is coded (in O.S. specific ways)
    to work on most any system, including the fall back to
    polling, if need be.  Should work with at least Windows
    and Linux fam.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From j.ezequiel at spitech.com  Thu Apr 11 05:34:11 2002
From: j.ezequiel at spitech.com (Ezequiel, Justin)
Date: Thu, 11 Apr 2002 17:34:11 +0800
Subject: SMTP with attachments
Message-ID: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>

How do I send an e-mail with attachments?
I am a newbie so be gentle.
I found a couple of examples (mfetch.py and MailTest.py) but could not get them to work.
MailTest.py raises an exception re "instructions.txt" not found.
mfetch.py imports "win32com.client" which I don't have and may not want to use as my program most likely will be run on a Linux box.
Please help.
 <<MailTest.py>> 
 <<mfetch.py>> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MailTest.py
Type: application/octet-stream
Size: 1641 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020411/93f140ef/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mfetch.py
Type: application/octet-stream
Size: 11849 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020411/93f140ef/attachment-0001.obj>

From cliechti at gmx.net  Thu Apr 11 16:29:38 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Apr 2002 22:29:38 +0200
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu>
Message-ID: <Xns91EDE505F5BD7cliechtigmxnet@62.2.16.82>

Rajarshi Guha <rxg218 at psu.edu> wrote in
news:a94qk5$22mg at r02n01.cac.psu.edu: 

> On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de Maria
> wrote: 
> 
>> Is there a way to save in a file the status, objects, functions of
>> interpreter
>> in order to restore it after the interpreter has been closed?
>> 
>> If it is possible, one can work on a project stop and then continue
>> later adding functions, objects and so on.

thats a question for an IDE with such functionality. i don't know one but i 
think it should be easy to store all objects in the global namespace of the 
interpreter and load it again on next startup.
the textual history of the interpreter with all the outputs must be saved 
separately.

> I have a related question - does the OP refer to object persistence or
> object serialization? Is there a difference between the two?

just serializing an object does not mean that its persistent. you can send 
a serialized object through the net or you can use to store it on disk.
 
> I looked at the shelve and pickle modules - the former uses a dbm
> database whereas the latter just dumps an ASCII representation of the
> object. The fact that shelve uses a dbm indicates that it should be
> more efficient. Is this true? When would I choose one over the other?

i think pickle is more versatile. the serialized object can be stored in a 
file, sent over the net (see e.g. PYRO.sf.net), put in a database etc.

chris

-- 
Chris <cliechti at gmx.net>



From Oschler at earthlink.net  Wed Apr  3 15:37:24 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 3 Apr 2002 15:37:24 -0500
Subject: IDE Wars, the prequel!
Message-ID: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com>

In a galaxy far, far away..

I've been getting some IDE suggestions on the COLA forum.  The current batch
seems to be:

Gvim
Emacsen
ActiveState's ActivePython

I'm going to look at them all this weekend, but I'd like to hear from those
of you who have experiences they would like to share.

thx






From peter at engcorp.com  Sun Apr  7 23:47:48 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 23:47:48 -0400
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com> <a8q9bi$2vka$1@news.idiom.com> <d9ea2696.0204071920.1b9d75c4@posting.google.com>
Message-ID: <3CB112E4.B5884860@engcorp.com>

distrex wrote:
> Jim Dennis wrote:
> > distrex wrote:
> > >I understand most of python. But I still don't get modules. I know
> > >they are just basicaly a function but i don't know how to use them
> > >effectively and what each one does.  [...]
> >
> >  Modules are things that you import into your namespace.  
[...snip rest of Jim's reply]
> 
> Don't take this the wrong way but I am a newbie and didn't understand
> 86% of what you just said. All I got was modules are classes functions
> and variables. and a few other things but that was to complicated for
> me (the newbie)

I can see how it might have been above the level you are at just yet,
so don't take this the wrong way: what are you going to do about it?

Have you read the Python tutorial yet?  It's on the web site.  Have
you read the documentation on modules?  Or is that where you got the
(wrong) impression they were "basically functions"?  If so, there may be
some improvements to be made since I don't believe the documentation
authors intended for the material to be incomprehensible to newbies.

If you've tried to learn what they are, but are still having a hard
time, maybe you could post some specific questions to test your 
understanding.  You ask "how to use them effectively", which is
a good one, but without more of an idea what you are trying to 
accomplish, it might be hard to provide a good answer.

Your question about "what each one does" is also strange to us.  
It almost sounds like you know there are dozens of standard modules
and were asking what each one does.  There is full documentation
on each module already, so that must not have been what you meant.
But in that case the question is bizarre. "Each one"?  What "each"
are you talking about?

And in case it helps any, here's a shorter answer than Jim's:

Modules are files which contain collections of functions, variables,
and classes.  They help organize your code so it doesn't all 
have to be in one large file, and they let you put code in 
libraries to make it easier to re-use, so you don't have to
copy-and-paste all the time to re-use other people's code.
(This is a somewhat simplified description but may be more 
useful to a newbie.)

Help us a little more, and we can help you a little more. :-)

-Peter


From pyth at devel.trillke.net  Wed Apr 24 17:04:54 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 24 Apr 2002 23:04:54 +0200
Subject: python and very large data sets???
In-Reply-To: <3CC6ED95.569DA16A@metaslash.com>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <3CC6ED95.569DA16A@metaslash.com>
Message-ID: <20020424210454.GA22555@devel.trillke>

On Wed, Apr 24, 2002 at 01:38:29PM -0400, Neal Norwitz wrote:

> To give you an idea of Python's speed, I was able to write out 600MB
> of data in 450 seconds.  I could read the data, modify it, and write
> it back out in 700 seconds.

i am not sure if you code would run much faster if coded the same
way in c...

> 
> By box is 650 Mhz Athlon, 256 MB RAM
> 
> time to write  input file: 450.9 seconds
> time to modify input file: 702.2 seconds
> first line of output file: A B C
> file sizes (input/output): 600000000/600000000

well, this really depends on how you do it (even in c). 
my code example below is about 6 times faster on
a 750MHZ Athlon, 256 MB RAM with an old harddisk (!):

time to write input file[600000 kb]:                  69.1 seconds
time to read input file, upper(), write to new file: 129.8 seconds
file sizes (input/output): 600000kb /600000kb

This seems to be much more IO-bound than python-bound.
I am pretty sure you can improve the modification time
time by using two separate modern disks, MMAP, Threads and what not.

Anyway, Bengt Richter correctly pointed out that one needs
a better specification to come up with a reasonable
estimation.

I just don't happen to see the advantages of bringing
a database into the picture. It seems like a classical
batch job and it 'random access many times' is not needed,
so why?

	holger


-- snip 

#!/usr/bin/env python2.2

import os,sys,time

kb=1024

fn1='/net/projects/bigtestfile'
fn2='/net/projects/bigtestfile.out'

def filesize(fn):
    return os.stat(fn)[6] 

def filekbsize(fn):
    return filesize(fn)/kb
    
def create(kbnum):
    f=open(fn1, 'wb+')
    print "writing file with size %d kb" % kbnum
    for i in xrange(kbnum):
        f.write('a'*kb)
    f.close()

def modify():
    f1=open(fn1,'rb+')
    f2=open(fn2,'wb+')
    f1size=filesize(fn1)

    while 1:
        kbbuf = f1.read(kb)
        if len(kbbuf)==0:
            break
            
        kbbuf = kbbuf.upper()
        f2.write(kbbuf)

    f2.close()
    f1.close()

if __name__=='__main__':
    size = len(sys.argv)>1 and int(sys.argv[1]) or 1000

    start = time.time()
    create(size)
    print 'time to write input file[%d kb]: %.1f seconds' % (filekbsize(fn1),time.time() - start)
    
    start = time.time()
    modify()
    print 'time to read input file, upper(), write to new file: %.1f seconds' % (time.time() - start)

    print 'file sizes (input/output): %dkb /%dkb' % \
                    (filekbsize(fn1),filekbsize(fn2))


-- snip











From gleki at gol.ge  Tue Apr  9 04:03:04 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Tue, 09 Apr 2002 10:03:04 +0200
Subject: calling class methods as functions
Message-ID: <3CB2A038.2468F842@gol.ge>

Hello!

I wonder if someone explaines tom how can I overcome the following
problem in Python2.0 (I have to use this interpreter):
suppose, we have 2 classes, B, and C.

in class B we have a method "write"

in class C we have a method "write"

def write (self, arg):
    import B
    res=B.B.write(self, arg)    #res is a string being either "ok" or
"failure"
    del B
    return res


The interpreter gives an error, that class isntance must be given as 1st
argument.

What's wrong?

Thanx,
Giorgi

PS. I know that the best solution is to make a baseclass and put the
method "write" there. Of course. But I would like to know why this
solution fails.



From narnett at mccmedia.com  Tue Apr  9 11:12:10 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Tue, 9 Apr 2002 08:12:10 -0700
Subject: Creating objects in thread created with subclass
In-Reply-To: <a8st6c$jlp$1@panix1.panix.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJMEIKKAAA.narnett@mccmedia.com>

Aahz & all,

I'm working from Aahz's example of a thread pool spider
(http://starship.python.net/crew/aahz/IPC9/ThreadPoolSpider.py) and I'm
uncertain about whether I should continue with the approach that I used in a
single-threaded version.  In my original, I create an object that knows how
to do the retrieval of pages and some processing (this spider is not
recursive in the usual sense of following links, but it is somewhat
similar).  After it retrieves a series of pages and extracts the data I
want, it returns a list of tuples that I pass to another object that either
directly inserts it into a MySQL database or writes a bulk insert file.  So
I already have a producer-consumer pattern of sorts.

In my original, I'd create the retriever object, set attributes, then call
the function to go do its work, which returns the data to pass to the
database insertion object.  Here's my question -- in the "run" method of
Aahz's example, if I create my retriever object in an ordinary way, then all
of the threads will be sharing the same object, right?  And that most
definitely will not work.  So, I could create objects with 'eval' or 'exec'
to generate unique names for each thread -- but is that the right way to do
it?  Or should I stop using objects there and directly use the functions
from my classes.  I haven't looked at a whole lot of examples, but so far,
the latter seems to be the way people do it.

The retrieval object is a subclass of a subclass; the top-level class knows
a lot about generating URLs, analyzing data, etc.  Thus, I think it'll be a
bit of work to rearrange it into functions.  And I'm going to lose some of
the elegancy of the subclassing, which I don't particularly like.

If it helps, the reason for subclassing is that this spider knows how to
retrieve some specific types of Web pages.  The rules for recursing and
analyzing them vary by Web site, but they have a lot in common.

Thanks!

Nick





From bh at intevation.de  Tue Apr 30 08:19:51 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 30 Apr 2002 14:19:51 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org>
Message-ID: <6qn0vlbcpk.fsf@abnoba.intevation.de>

Roman Neuhauser <neuhauser at mail.cz> writes:

> > From: Bernhard Herzog <bh at intevation.de>
> > Why should chr be an int method in the first place? It's not tied to
> > ints all that much[1]. IMO it sits somewhere between ints (including
> > int-like objects !) and strings so it shouldn't be a method of either
> > class. Leaving it a function is perfectly fine, therefore, although
> > putting it into the string module might be a good idea. The same would
> > apply to ord, IMO.
> 
>     Ugh, that looks to me as appropriate as join() being a string
>     method / module function.

Exactly. I still think that join as a function makes more sense than as
a string method and for the same reasons as stated for chr above. Even
if join is implemented as a string method.

> > [1] In fact, chr today accepts any object that can be converted
> > successfully to an int, so that chr(42.5) does work, but that's not
> > always desirable.
> 
>     Any object that can be converted to a character should have a chr()
>     method.

Would the same reasoning apply to objects that can be converted to ints,
for example. I.e. should strings have a method .int() instead of int
taking a string as argument (among other types)? (forgetting for the
moment, that it's actually implemented through the __int__ method :) )

I see numbers and characters as quite different datatypes and a
conversion between the two is right in the middle and neither more on
the int side than on the str side. 

The correspondence between characters depends on an encoding, something
which chr and ord do not take into account at all, for example, and one
couldn't expect that e.g. int knows about encodings, so that a chr
method could take the encoding as argument (perhaps defaulting to
ascii).

Of course, whether a conversion operation should be a method, a function
or a constructor is largely a matter of taste and is influenced by
whatever other languages one is coming from. For someone with a
C-background, characters are ints, so making chr an int method may seem
appropriate from that perspective.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From aleax at aleax.it  Tue Apr 16 14:56:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 20:56:42 +0200
Subject: python threading
In-Reply-To: <BFFEB31BC23CCD44B8924E139E1D87E448EF1F@MAILBOX.nameconnector.com>
References: <BFFEB31BC23CCD44B8924E139E1D87E448EF1F@MAILBOX.nameconnector.com>
Message-ID: <02041620564200.30447@arthur>

On Tuesday 16 April 2002 20:39, Geoffrey Talvola wrote:
	...
> Don't most GUI toolkits have some way to post an event to the main GUI
> thread's event loop from another thread?  wxPython has wxPostEvent()
> which does this.

I don't know of any Tkinter or PyQt equivalent.  I hope somebody can
enlighten me about those, as they're the toolkits I need to use.


> So you could write a wrapper class around Queue that posts a "wakeup"
> event to the main GUI thread after an item is added to the queue.  The
> event handler for the "wakeup" event would process whatever items had
> been added to the Queue and then return control to the event loop.

<nod> yep.  And figure out something similar for asyncore/asynchat or
other packages that use a select -- that would be just the ticket.  All
the modification needed to Queue would be an optional callable to call
when the queue transitions from empty to non-empty -- different ones
could be supplied for various GUI kits, async*, etc.


Alex




From neal at metaslash.com  Tue Apr  9 17:17:03 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 09 Apr 2002 17:17:03 -0400
Subject: class callable with any method name
References: <9d095c5a.0204091256.4d88a246@posting.google.com>
Message-ID: <3CB35A4F.5E57FEC3@metaslash.com>

Imer wrote:
> 
> Hi, I'm trying to find a way to implement a class that would allow any
> method name to be called.
> When that happens, a special method would be called and the name of
> the method that was originally called would be passed as an argument.
> 
> For instance:
> ########
> class A:
>   def __method_called__ (self, method_name, method_argument):
>     print "The method", method_name, "was called with the argument",
> method_argument
> ########
> 
> Then the following code:
> ########
> a=A()
> a.this_is_a_random_method(arg=1)
> a.yet_another_random_method(param=3)
> ########
> 
> Would print out:
> ########
> The method this_is_a_random_method was called with the argument
> {'arg':1}
> The method yet_another_random_method was called with the argument
> {'param':3}
> ########

class _Ref:
    def __init__(self, method, name):
        self.method = method
        self.name = name
    def __call__(self, *args, **kw):
        apply(self.method, (self.name,) + args, kw)
        
class A:
    def __getattr__(self, attr):
        return _Ref(self.__method_called__, attr)
    def __method_called__(self, name, *args, **kw):
        print '%s() was called with %s, %s' % (name, args, kw)

Neal


From syver-en+usenet at online.no  Tue Apr  9 20:40:13 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Wed, 10 Apr 2002 00:40:13 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <a8vp7c$3v1$1@panix1.panix.com>
Message-ID: <ud6x8wfnx.fsf@online.no>

aahz at pythoncraft.com (Aahz) writes:
> 
> Maybe it "should" be, but I like to think I rank reasonbly high on
> the ability to reason Pythonically, and I still think that "".join()
> is a wart.

Maybe what is Pythonic should change then? This would only be
natural (that what is pythonically should change), because the language
itself seems to be changing to a more pervasive object
orientedness (Which I applaud).

-- 

Vennlig hilsen 

Syver Enstad


From pete at dexterslab.de  Thu Apr 18 20:54:30 2002
From: pete at dexterslab.de (Henning Peters)
Date: Fri, 19 Apr 2002 02:54:30 +0200
Subject: traceback output to a network connection
Message-ID: <a9npmv$en6$05$1@news.t-online.com>

hi NG,

for debugging a network server i would like to show the parser's traceback
output on my network client. I figured out that the built-in traceback
module would be a good choice, but it seems, that the traceback-methods are
limited to write only to open file handlers.

is there a known way to relay this stuff to an open network socket
connection?
thanks...

bye,
henning





From trentm at ActiveState.com  Thu Apr 18 12:22:51 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Thu, 18 Apr 2002 09:22:51 -0700
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
In-Reply-To: <Xns91F4680F3221cliechtigmxnet@62.2.16.82>; from cliechti@gmx.net on Thu, Apr 18, 2002 at 12:37:02AM +0200
References: <mailman.1019071474.3193.python-list@python.org> <Xns91F4680F3221cliechtigmxnet@62.2.16.82>
Message-ID: <20020418092251.K28973@ActiveState.com>

(Reply privately be mistake in first response.)

[Chris Liechti wrote]
> Trent Mick <TrentM at ActiveState.com> wrote in 
> news:mailman.1019071474.3193.python-list at python.org:
> > In addition, PyPPM (the Python Package Manager) enables you to easily
> > install a number of other popular Python packages. For example, try
> > this:
> >     > pyppm install Numeric
> > to install Numeric Python. Use this command to list all available
> > packages:
> >     > pyppm search
> 
> is that also available separately and can i set up my own packages and 
> server?

No, PyPPM is not available separately. There is new PPM tech in the
works that may allow some of what you want but no guarantees on when
that might be available.

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From bbollenbach at shaw.ca  Wed Apr 10 21:02:22 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Thu, 11 Apr 2002 01:02:22 GMT
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu>
Message-ID: <yg5t8.10013$Kq4.394311@news2.calgary.shaw.ca>

In article <a92hl1$1v4c at r02n01.cac.psu.edu>, Rajarshi Guha wrote:
[snip]
> (I could arrange the data into tuples if necessary)

There's your answer then. :)

When the choice lies between smart data structures, or poor data
structures that force you to think of ways to code around them, go with
the smart data structures.

e.g.

l = [(1,'s'), (2,'t'), (3,'d')]

for x, y in l:
    print x, y

Can't get any simpler than that.

HTH,

Brad


From mgash at trisystems.co.uk  Tue Apr  9 10:43:24 2002
From: mgash at trisystems.co.uk (Mark Gash)
Date: Tue, 9 Apr 2002 15:43:24 +0100 
Subject: how to read hardware info on Win NT
Message-ID: <31575A892FF6D1118F5800600846864DCC77AA@intrepid>

>>On Tue, Apr 09, 2002 at 01:05:00PM +0000, Marcus Stojek wrote:
>> could anyone please give me a hit how to get hardware information with
>> Python? I need the id of my ethernet card and my harddisk.

>On a similar topic, are there any modules to give you Win NT/2k/etc
performance
>statistics, process lists, network stats, disk usage, etc ?

>Any work being done on such a module ?

There are two modules that will assist you in this query; 
For Marcus, you could investigate the _winreg module - you will be able to
script a query of the registry keys and data values. The only down side is
that you will have to know where and what you are looking at.
For Chris, investigate the win32pdh and win32pdhutil modules - this will
enable you to extract data from the performance data collectors.

A search on google for all of these modules will result in some useful
tutorials in these subjects.

If you want any further help, I'll try ;-)

Hope this helps

-- 
http://mail.python.org/mailman/listinfo/python-list




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




From martin at v.loewis.de  Tue Apr  9 16:09:01 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Apr 2002 22:09:01 +0200
Subject: xml encoding in minidom
References: <db8a9aa8.0204090941.3b518901@posting.google.com>
Message-ID: <m38z7w7hya.fsf@mira.informatik.hu-berlin.de>

joakim.storck at home.se (Joakim Storck) writes:

> I've been playing around with the xml.dom.minidom, but I have some
> problems with the document encoding. What I'm trying to do is
> basically to take care of data from a web form using the cgi-module,
> then insert data into a xml-document using the minidom. However, the
> data does contain swedish characters (?, ?, ?) which messes things up
> more than I could have imagined.

The easiest thing should be to use the XML default encoding, UTF-8.

> Since the minidom has no way of setting document encoding directly, i
> tried the following:
> 
> doc = xml.dom.minidom.Document()
> pi = doc.createProcessingInstruction('xml','version="1.0"
> encoding="ISO-8859-1" ')
> doc.appendChild(pi)

Notice that the <?xml header is *not* a processing instruction,
although it looks like one.

> There is also some kind of problem with the minidom.toxml()-method.
> From what I've read in other postings it has to do with file encoding,
> which seems to be set to UTF-8 by default.

That is not the default file encoding, it is the default encoding
mandated by the XML recommendation. minidom does not support emission
of XML in any other encoding.

> I'm no expert, neither at xml nor Python, but it seems I'm not the
> only one who experienced these problems.

It might be easiest if you stop fighting the defaults. If you insist
on generating XML documents in Latin-1, you need to write the XML
header explicitly, then use writexml of the root node.

Regards,
Martin


From stephen at xemacs.org  Mon Apr 29 10:07:01 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 29 Apr 2002 23:07:01 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> <mailman.1020071811.14821.python-list@python.org>
Message-ID: <87adrm38fu.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:

> >>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
> 
>     Laura> People here seem to be thinking that the GPL is there to
>     Laura> protect creators rights to their software.
> 
> I'm pretty sure that that is what rms has in mind, as long as both
> "creators" and "software" are inclusive collective nouns.
> 
> The consumer protection argument is the OSI line (a special case
> thereof), and rms don't have much truck with anything so economical
> and relativist.

    Laura> 1. Why did you remove my prior line, 'as a thought
    Laura> experiment?'

I apologize for the rudeness, none was intended.  I did not, and still
don't, see that its presence or absence has any effect on the
semantics of the discussion.  So I took it out.

    Laura> I put it in in order to avoid having this conversation
    Laura> right now?

But it didn't.  I don't see how the GPL is about consumer protection.

Not as "original intent" (you _did_ use the word "intended"), and not
as a "thought experiment" in social engineering.  I planned to address
both, but I see I completely missed the mark on the "thought
experiment" (and further confused the issue by mentioning rms again).
Again, I'm sorry for creating misunderstanding.

There are plenty of ways to address the goal of consumer protection,
many of which might very well be more profitable (eg, the US DOD's
second source policy) for the developer of the software under license
than the GPL.  None of these are satisfactory substitutes for the GPL
(or free software in general) because they violate freedom.  Another
example: the Aladdin Free Public License has all the benefits of the
GPL in protecting customers from broken software: you can always hire
someone to fix your broken Ghostscript.  You just can't redistribute
the fixed version for money, and thus amortize the cost.

But you don't want to go there, do you?  rms definitely doesn't.

If one is _only_ interested in consumer protection, there are plenty
of very attractive alternatives to the GPL that could be imagined.  As
you point out, some might have to be enforced by changes in IP law in
practice, but most could (in theory) be achieved simply by customers
insisting on a reasonable contract.  (The OSI line again.  "You don't
need to abolish IP or to demand the GPL; there are other ways to
protect yourself.  But you should protect yourself by insisting on
access to source.  Preferably by using free software, which benefits
all software users, usually without hurting you.")

    Laura> 2. I learned this from RMS himself, when he was living in
    Laura> my house about 15 years ago.  Look up 'hoarder' in his
    Laura> writings.

I've read them.  I understand the concept.

My mother doesn't.  She will never notice how much software has been
"hoarded" by Microsoft, only how much has been delivered at remarkably
low prices.  My mother has no need or desire for a copy of the source
herself.  She recognizes the usefulness of _someone_ having access to
source, but "second-sourcing" or "source escrow" would be plenty to
protect that wily consumer in her own opinion.

My point is that rms developed the GPL to protect/enforce sharing of
software ideas, not to protect consumers' economic interests in
working software.  The protection of consumers is a very desirable
side effect, but according to rms _economic benefits in general must
never be considered on a par with software freedom_.  The GPL wasn't
intended to protect consumers, and it is not tuned to doing that
effectively, since it surely results in less software being developed.
There's a very good reason for rms's denial of consumer sovereignty:
it is easy to show under the standard microeconomic models that "a
little bit of IP in the right places" will surely benefit consumers
relative to "no IP," and the GPL imposes _more_ frictions than "no IP"
does.

So I don't think an argument starting from consumer protection will
lead to the GPL, or anything that looks like it, in many important
cases.  rms certainly doesn't.  He says so every time he's asked to
comment on "open source," and he's personally asked me not to
formalize, not to measure, not to mention, and most especially not to
publish what "is easy to show...".[1]

Thus my point that the GPL (and software freedom in general) is
primarily about protecting the freedom of developers to develop.

    Laura> I thought compelling people was a worse sin than software
    Laura> hoarding.  [...]  RMS does not see things this way at all.
    Laura> He thinks that software hoarding is a worse sin than
    Laura> compelling,

I've seen him write similar things, also the moral equation of
intellectual property to human slavery.  But I don't understand what
he means by this, as I would deduce a need to close the "ASP loophole"
in the GPL.  Eg with a "must publish `deployed' modifications" clause
a la the Apple Public Source License.  But he firmly rejects that.[2]

By the way, there's a small technical error in your consumer
protection argument.  You wrote

    Laura> The thing is you will be able to tell (or hire somebody to
    Laura> tell you) before you buy

There's nothing in the GPL that requires that you get source _before_
you buy the software.  Cygnus and The Kompany both have made money on
GPLed software for which the source is not public.  It just happens
that the customers had no interest in redistributing source
themselves.  In fact, it seems plausible that the GPL was helpful in
convincing some customers not to redistribute the source, as it very
likely would have quickly ended up in the hands of a competitor.

I doubt this is all that important in practice yet---most free
software seems to be public source as well as open source---but I
suspect it will grow over time, especially in market segments
developing for ASP-like applications.[3]


Footnotes: 
[1]  The "standard microeconomic models" are far more applicable than
most free software advocates would like to believe.  Showing what the
limits and appropriate modifications are is a project I guess I should
dust off and get published.

[2]  For reasons I understand, I think, and agree with.  So evidently
I don't understand the hoarding v. compelling and IP v. slavery ideas.

[3]  Though it may not.  The Kompany now avoids the GPL where
possible, as they found that bad PR from not publishing source to
GPLed software is actually more unpleasant than bad PR from
distributing software under proprietary licenses.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN



From phr-n2002a at nightsong.com  Fri Apr 12 19:04:50 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 16:04:50 -0700
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net>
Message-ID: <7x1ydkikml.fsf@ruckus.brouhaha.com>

"akhar" <akhar at videotron.ca> writes:
> Has anyone ported Python to embedded devices? I heard of the deeply embedded
> python porject but the link is dead. What would it take to port python to an
> embedded platform (whether be it the mc68hc11 or PICs)?

I'd say it's hopeless or at least impractical.  Those processors are
much too constrained to run Python.  It would be like trying to run an
SQL database on them.  Python has many virtues but being able to run
on very tiny processors isn't one of them.



From dustym at riseup.net  Thu Apr 18 16:09:38 2002
From: dustym at riseup.net (Dusty)
Date: 18 Apr 2002 13:09:38 -0700
Subject: Curses for windows.
References: <f8e5025.0204170953.570cfb43@posting.google.com> <m3it6qyxr5.fsf@mira.informatik.hu-berlin.de> <slrnabtbg7.8h5.look@pikkukorppi.cs.tut.fi>
Message-ID: <7b7daf0d.0204181209.112e3740@posting.google.com>

Or you can write the program in UNIX and ssh in using putty.

Dusty

Simo Salminen <look at sig.invalid> wrote in message news:<slrnabtbg7.8h5.look at pikkukorppi.cs.tut.fi>...
> * Martin v. Loewis [17 Apr 2002 20:40:30 +0200]
> > alx78 at hotmail.com (alex) writes:
> > 
> >> Where is "curses" module for python 2.2 in windows NT.
> > 
> > Curses is not supported on Windows NT.
> 
> but you can use curses on NT if you use cygwin-python.


From neal at metaslash.com  Wed Apr  3 18:13:29 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Wed, 03 Apr 2002 18:13:29 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com>
Message-ID: <3CAB8C99.232A8571@metaslash.com>

Guido van Rossum wrote:
> 
> Despite the negative feedback, I've decided to accept the PEP.  The
> most important perceived problem is that newbies tend to write
> 
>     if x == True: ...
> 
> where they should write
> 
>     if x: ...
> 
> I believe this problem should be solved by education (the Zen master
> hits the student on the head with a stick, and the student gets
> enlightenment) rather than by holding back what I believe will be a
> useful feature.  You can expect the bool type in Python 2.3; it will
> be in CVS tonight.

And the warnings have already been added to pychecker.
I know I'm going to wish I never said this, but ...
If you find REAL problems with code that breaks, let me know.
I will try to address problems that may arise.

This code:
	True = 1
	False = 0

	def test(a):
	    if a is True: print 'True'
	    if a == False: print 'False'

generates these warnings:
	test.py:2: Should not assign to True, it is (or will be) a builtin
	test.py:3: Should not assign to False, it is (or will be) a builtin
	test.py:6: Comparisions with True are not necessary
	test.py:7: Comparisions with False are not necessary

Neal




From sholden at holdenweb.com  Wed Apr 17 10:07:40 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 10:07:40 -0400
Subject: Classes Again !
References: <3cbd7527$1@news.comindico.com.au>
Message-ID: <Jmfv8.13955$hl2.1855@atlpnn01.usenetserver.com>

"Graeme Matthew" <graeme.matthew at unite.com.au> wrote ...
> Hi all  I am going mad again, but making progress, been on a problem for 1
> hour already and cannot see the cause ! please could someone help
>
As long as you don't post again in another hour saying "It's been an hour
since I posted and nobody's helped me yet!" ;o)

> The getSubDocument method in my template class is failing suddenly , i
dont
> have a clue why ! thanks a mil
>
> ERROR:
>
> Traceback (most recent call last):
>   File "C:\www\cgi-bin\main.cgi", line 10, in ?
>     data = t.getSubDocument('C:/www/templates/template_org_query.html')
> AttributeError: Template instance has no attribute 'getSubDocument'
>
Technically your method isn't failing, it simply isn't being found. This
might lead me to suspect that the web server was caching classes, and using
an old definition. Was there ever a time when you had Template.Template
definitions without that method? Are you sure there aren't two copies of the
module, and you are perhaps importing an older copy?

You don't say what web server you are using, nor how Python is integrated
with it. There might be one reason for Apache+mod_snake to fail, another for
IIS+ActiveScripting.

I can't see anything obviously wrong. Maybe a little more information would
help someone spot the problem?

regards
 Steve
[CGI code omitted]






From loewis at informatik.hu-berlin.de  Thu Apr 18 11:02:15 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 17:02:15 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019123256.19883.python-list@python.org> <23891c90.0204180534.5b3c8542@posting.google.com>
Message-ID: <j4bsch6oeg.fsf@informatik.hu-berlin.de>

paul at boddie.net (Paul Boddie) writes:

> I just hope that the whole linking thing isn't some
> dark-side-of-the-force libtool-determined configuration, otherwise
> you may need to go in even deeper than I've dared to in the past.

It turns out that it was simpler than that: the autoconf test for GNU
ld fails for binutils 2.12; I've corrected this for configure of
Python 2.1, 2.2, and 2.3.

Regards,
Martin


From shalehperry at attbi.com  Mon Apr 29 15:42:31 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 29 Apr 2002 12:42:31 -0700 (PDT)
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <aak1uq$bgt$1@bob.news.rcn.net>
Message-ID: <XFMail.20020429124231.shalehperry@attbi.com>

On 29-Apr-2002 Raymond Hettinger wrote:
> From the comments, I've learned:
> 
> 1. map, filter, divmod, and pow may be immovable and should
>     be left alone.
> 
> 2. reduce has fewer friends, but should probably also be
>     left alone
> 

thank you, thank you.  Gah, I could not imagine being forced to use list
comprehensions everywhere.

> 4. apply() was repeatedly suggested as another candidate for
>     deprecation since the * and ** forms make it obsolete
> 
> 5.  oct() is another candidate based on the rarity of usage
> 

so should we remove the left and right bit shift ops too then?  oct() may be
rarely used but having it keeps me from writing C or perl code.




From fperez528 at yahoo.com  Tue Apr  9 20:53:17 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 09 Apr 2002 18:53:17 -0600
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de> <uvgb0eefo.fsf@turangalila.harmonixmusic.com>
Message-ID: <a902f3$a7m$1@peabody.colorado.edu>

Dan Schmidt wrote:
> Pymacs allows you to script Emacs with python.  I haven't used it, though.
> 
> http://www.iro.umontreal.ca/~pinard/pymacs/

Since the included examples are very sparse, here's some sample code to get 
you going (some of it is useful, other will at least show you how to do basic 
stuff):
"""Functions accessible from inside Emacs.

Filename: pym.py

Usage in Emacs: import them with
M-x pymacs-load pym

And then all functions here become available in Emacs as
pym-function-name with every _ replaced by - in the names.  """

__author__ = "Fernando Perez. <fperez at pizero.colorado.edu>"
__license__= "GPL"

import pymacs,re

# emacs is the global which handles lisp-based interaction with the active
# Emacs buffer from which any function is called

emacs = pymacs.lisp

# interactions is a global dict which MUST be updated for each new function
# defined which we want to be visible to Emacs. Each function must have an
# entry in it with the function as key (the function *object*, NOT its name as
# a string) and a string as value. At a minimum, this string will be empty,
# but it can contain the names of variables to be read interactively by the
# function, lisp-style.

# Functions meant to be used internally only (not exposed to Emacs) don't need
# an entry in interactions.

interactions = {}

#***************************************************************************
# WARNING: things from genutils copied verbatim here. For some reason pymacs
# does not import other modules correctly (my own, it seems ok with system
# stuff).

def indent(str,nspaces=4,ntabs=0):
    """Indent a string a given number of spaces or tabstops.
    
    indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces.
    """
    
    ind = '\t'*ntabs+' '*nspaces
    outstr = '%s%s' % (ind,str.replace('\n','\n'+ind))
    if outstr.endswith('\n'+ind):
        return outstr[:-len(ind)]
    else:
        return outstr

# End of genutils copy/paste job.

#***************************************************************************
# Lisp utility functions, snatched from elsewhere.

def clean_undo_after(checkpoint):
        """\
Remove all intermediate boundaries from the Undo list since CHECKPOINT.
"""
        emacs("""
(let ((undo-list %s))
  (if (not (eq buffer-undo-list undo-list))
      (let ((cursor buffer-undo-list))
        (while (not (eq (cdr cursor) undo-list))
          (if (car (cdr cursor))
              (setq cursor (cdr cursor))
            (setcdr cursor (cdr (cdr cursor)))))))
  nil)
"""
             % (checkpoint or 'nil'))

#***************************************************************************
# Utility functions, none of which need an interactions[] entry.

def lisp_obj_info(obj):
    """Return various details about a lisp object as a string.

    Useful mainly for debugging purposes."""
    
    info = [obj,obj.__class__,obj.index,type(obj.index),repr(obj)]
    info = map(str,info)
    info = '\n'.join(info)
    return info

#---------------------------------------------------------------------------
def lisp_char(lisp_obj):
    """Return a single character string from a lisp char object.

    Used to extract characters from their lisp form as obtained in interactive
    functions with the c code. """
    text_form = repr(lisp_obj)
    try:
        return re.search(r"'\?(.)'",text_form).group(1)
    except:
        return None

#---------------------------------------------------------------------------
def is_yes(lisp_obj):
    """Check whether an interactive lisp character reply is a yes (y/Y)"""

    try:
        return lisp_char(lisp_obj).lower() == 'y'
    except:
        return 0

#---------------------------------------------------------------------------
def cut_region(mode='string'):
    """Return the active region and remove it from Emacs.

    The mode parameter (default 'string') defines whether to return the region
    as a string or as a list of lines (mode='list').

    It is the caller's responsibility to insert the updated text at the
    end back in the Emacs buffer with a call to emacs.insert(...)."""

    start, end = emacs.point(), emacs.mark(emacs.t)
    # BUG: buffer_substring() can't extract regions with dos line endings 
(\r\n)
    # It dumps a traceback.
    region = emacs.buffer_substring(start, end)
    if mode == 'list':
        region = region.splitlines()
    emacs.delete_region(start, end)
    return region
    # cut_region() doesn't need an entry in interactions[] b/c it's meant to
    # be # used internally by other functions in this module, not directly
    # from Emacs

#---------------------------------------------------------------------------
def insert_text(text,offset=0):
    """Insert text in buffer and move cursor to a certain offset.

    If called with no offset, leaves the cursor at the current position."""

    # save undo state so we can roll everything into a single operation for 
undo
    checkpoint = emacs.buffer_undo_list.value()
    user_pos = emacs.point()
    emacs.insert(text)
    emacs.goto_char(user_pos+offset)
    # Collapse all operations into a single one, for Undo.
    clean_undo_after(checkpoint)

#---------------------------------------------------------------------------
def insert_indented_text(text,offset):
    """Insert indented text in buffer and move cursor to a certain offset."""
    
    # save undo state so we can roll everything into a single operation for 
undo
    checkpoint = emacs.buffer_undo_list.value()
    # figure out if we are indented or not, and adapt text accordingly
    indent_level = get_line_offset()
    if indent_level > 0:
        text = indent(text,indent_level)
    # perform actual insertion with proper cursor positioning
    offset += indent_level
    emacs.beginning_of_line()
    user_pos = emacs.point()
    emacs.insert(text)
    emacs.goto_char(user_pos+offset)
    # Collapse all operations into a single one, for Undo.
    clean_undo_after(checkpoint)

#---------------------------------------------------------------------------
def get_line_offset():
    """Return number of characters cursor is offset from margin.    """
    user_pos = emacs.point()
    emacs.beginning_of_line()
    line_start = emacs.point()
    emacs.goto_char(user_pos)
    return user_pos - line_start
# end get_line_offset()

#---------------------------------------------------------------------------
def newfn_string(name,sep,end,args=''):
    """Template for a new function definition.

    Returns the string containing the definition and the integer offset for
    cursor positioning."""
    
    # prepare text
    out = ''
    sep = lisp_char(sep)
    if sep is not None:
        out += '#'+sep*77+'\n'
    out += 'def '+name+'('+args
    offset = len(out)
    out += '):\n'
    out += '    """\n'*2
    if is_yes(end):
        out += '# end '+name+'()\n'
    return out,offset

#***************************************************************************
# 'Public' functions (exposed to Emacs). All these MUST have an interactions[]
# entry

def bow():
    """Break a region replacing all whitespace with newlines.

    Originally an example in Pymacs' README."""

    region = cut_region()
    emacs.insert('\n'.join(region.split()))

# Update interactions[] for functions meant to be visible in Emacs.

# Interaction strings follow some funny emacs-lisp conventions, with the first
# letter being a code and the rest a prompt. Use `C-h f interactive' in Emacs
# to get a description.  The simplest one is a prompt for a string, which is
# given as a string of the form 's<prompt>'.

# Will print 'name ' in the minibuffer and get a string:
#interactions[deft] = 'sNew function name? '

# The c code is for characters, and the Pymacs readme says they are returned
# to python as ints, but that doesn't seem to be the case. Instead I'm getting
# Pymacs.Lisp objects, which have a repr() of the form "lisp('?<char>')" where
# <char> is the returned character.

interactions[bow] = ''

# Note that trying to set interactions as a function attribute:
# bow.interactions = ''
# is NOT WORKING. The module loads in Emacs, but no functions are actually
# recognized. Tested with Python 2.1, it might work with Python 2.2

#-----------------------------------------------------------------------------
def dos2unix():
    """Remove DOS line endings from a region.
    """
    # Save undo state so we can roll everything into a single operation for 
undo
    checkpoint = emacs.buffer_undo_list.value()
    region = cut_region('list')
    emacs.insert('\n'.join(region)+'\n')
    # Collapse all operations into a single one, for Undo.
    clean_undo_after(checkpoint)

# BUG: it's not working b/c of a bug in emacs.buffer_substring(), so let's not
# activate it for now.
#interactions[dos2unix] = ''

#---------------------------------------------------------------------------
def newfn(name,sep,end,args=''):
    """Insert a template for a new function definition."""

    insert_indented_text(*newfn_string(name,sep,end))

new_template = 'sNew %s name? \n'\
               'cEnter separator (RET for none): \n'\
               'cPrint end marker (y/[N])? '

interactions[newfn] = new_template % 'function'

#-----------------------------------------------------------------------------
def newweave(name,sep,end,use_blitz):
    """Insert a template for a new weave function definition.
    """

    blitz,ending = '',''
    if is_yes(use_blitz):
        blitz = ",type_factories = blitz_type_factories"
    if is_yes(end):
        ending = "\n# end %s()" % (name,)

    head,offset = newfn_string(name,sep,0)
    head += \
'''
    code = \\
"""

"""
    return weave.inline(code,[]%(blitz)s)%(ending)s
''' % locals()
    insert_indented_text(head,offset)

interactions[newweave] = new_template % 'weave function'
interactions[newweave] += '\ncUse blitz type factories (y/[N])? '

#---------------------------------------------------------------------------
def newmeth(name,sep,end):
    """Insert a template for a new method definition.    """
    
    insert_indented_text(*newfn_string(name,sep,end,'self'))

interactions[newmeth] = new_template % 'method'

#---------------------------------------------------------------------------
def newclass(name,sep,end):
    """Template for new class definition.    """
    out =  ('class %s:\n' % (name,)) + ('    """\n'*2) + '\n'
    offset = get_line_offset()+len(out) + len ("    def __init__(self")
    new_str = newfn_string('__init__',None,None,'self')[0]
    out += indent(new_str)
    if is_yes(end):
        out += '# end class '+name+'\n'
    insert_indented_text(out,offset)

interactions[newclass] = new_template % 'class'




From erno-news at erno.iki.fi  Wed Apr 10 22:28:34 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Apr 2002 05:28:34 +0300
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <slrnab9o7v.i8b.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ku3cy3ynn1.fsf@lasipalatsi.fi>

In article <slrnab9o7v.i8b.huaiyu at gauss.almadan.ibm.com>,
huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) writes:

| On Wed, 10 Apr 2002 18:38:47 -0400, Brad Clements
| <bkc at Murkworks.com> wrote:
| >>>> class x(object):
| >...  pass
| >...
| >>>> y = x()
| >>>> type(y)
| ><class '__main__.x'>
| 
| Is this a bug?  Is x a meta class?

no, it is a regular (new-style) class.

| >>> y.__class__ 
| <class '__main__.x'>
| >>> z = y()
| Traceback (most recent call last):
|   File "<stdin>", line 1, in ?
| TypeError: 'x' object is not callable

| Is y a class?  I'm puzzled ...

no, it is an is an instance of x (and by implication, its type
is x).

  -- erno


From jon+usenet at unequivocal.co.uk  Mon Apr 29 16:10:43 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 29 Apr 2002 20:10:43 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org>
Message-ID: <slrnacra60.v3g.jon+usenet@snowy.squish.net>

In article <mailman.1020102500.7281.python-list at python.org>, Fran?ois Pinard wrote:
>> Code and HTML are separate things that belong in separate files
>> rather than muddled together.
> 
> I presume this is debatable.  Up to now, I found easier having single files,
> for the case you are adjusting both aspects in a single editing session.

Hmm. In my experience they are different languages for different
purposes created and edited at different times by different people
with different tools.

> Yet, seeing the whole thing, both the designer and the programmer can leave
> the page in a consistent state, even if not fluent in the _other_ field.

Woah, *big* disagreement here. Not sure what kind of designers you are
used to, but in my experience they, and/or the tools they use are
incapabale of editing the HTML without destroying the code.

> My experience taught me that it is attractively simple to use a
> single file instead of two,

Heh. I would say deceptively attractive ;-)

> For better style, and a bit for better speed, the amount of Python code
> in HTML pages should stay small.

If smaller is better, why is smallest not best?

If you take a look at the following URL you can see examples of jonpy
template use:

  http://jonpy.sourceforge.net/wt-examples.html#templatecode

I must admit I have never yet seen a system which integrates HTML and
code in the same file without the file ending up an unreadable mess.

Cheers


Jon


From pyth at devel.trillke.net  Sun Apr 21 14:05:36 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 20:05:36 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <a9ut8d$dql$1@panix3.panix.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com>
Message-ID: <20020421180536.GG10574@devel.trillke>

On Sun, Apr 21, 2002 at 01:35:09PM -0400, Aahz wrote:
> In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
> Allan Crooks <googlegroups at sixtyten.org> wrote:
> >
> >Another thought I've had is doing it by specifying, say for module
> >'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
> >what versions of Python it is running on, and then compiling all
> >modules it can deal with, and then importing all objects into it's
> >namespace.
> 
> Yup.  But do it with a package.

But doesn't this lead to lots of duplicated -hard to maintain-  code?

	Holger




From aleax at aleax.it  Wed Apr  3 02:20:49 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 07:20:49 GMT
Subject: Subclass sadness (was Re: [Python-Dev] PEP 285: Adding a bool type)
References: <200203300539.g2U5dAR17743@pcp742651pcs.reston01.va.comcast.net> <a8ajpi$i36$1@lust.ihug.co.nz> <k2dq8.50144$S52.1717644@news2.tin.it> <a8dmgf$3ms$1@lust.ihug.co.nz>
Message-ID: <l3yq8.53787$S52.1863266@news2.tin.it>

Gary Stephenson wrote:
        ...
>> > Given that our objects
>> > _are_ allowed to mutate their state, wouldn't it be preferable for them
> to
>> > do so in more intelligent ways?
>>
>> No.  "Change" is such a troublesome notion already, that to have it
>> train behind itself a "keeping track of uncountably infinite sets" is
>> quite a conceptual mountain.
> 
> I'm not sure we are actually talking about uncountably infinite sets here

Not in programming (the number of different programs is _countably_
infinite, since we write them with a finite alphabet), only in maths.

> are we?.  Isn't it rather the finite and (probably) small set of heir
> classes that need to be kept track of?  (I do agree that even this is

Not conceptually, in a language where the definition of classes also
changes continually.  And if you do want change, why shouldn't it?

> probably way too much book-keeping overhead to make it practicable in the
> real world.)  And modelling "Change" well is surely something that we all
> need to continually improve upon.

There's no practical usefulness at all in having all objects recompute
their set memberships at each change in the objects and/or the set
definitions, of course -- and there's no theoretical interest in it at all
for me, either, when checking membership on the fly "just in time" is
so obviously workable and conceptually clean.  The "tracking" of set
memberships is useless conceptual overhead, just as it would be
absurd pragmatical overhead.

>> Objects that do not change are perfectly well implementable and quite
>> usable.  Anybody looking for a closer match to maths should IMHO turn
>> right to lazy FP (Haskell...) rather than tilt agains the windmill of
>> 'proper' modeling of change.
> 
> I have yet to really come to grips with Haskell - but look forward to
> doing
> so.  However, I think what we are looking for here is not so much a closer
> match to maths, but a closer match to reality - which has always been one

"typing" has no real match to reality.  Set membership does, but
the "tracking" you seem to desire doesn't.

> of the O-O's great marketing points - that it (supposedly) models reality
> more
> closely.  Besides, just because nobody has yet come up with a really good
> imperative system for subtyping doesn't necessarily mean that it's all
> "tilting at windmills" (though it may very well prove to be so!).

But what would that "good system" accomplish that is of any interest
whatsoever?  The very idea of letting object changes but NOT allowing
the same dynamicity for types, on which so many current languages are
based, is half-cocked anyway.

>> Yes.  If you realize that 'behavior' is never intrinsic to a set of
> values,
>> and systematically _superimpose_ behavior on value sets (Haskell's
>> typeclasses being of course the right way to do that:-), then you
>> complete the conceptual trip to systematic purity.
> 
> Err, as I said, I'm not terribly Haskell enlightened (yet).  Does this
> perhaps relate in some way to Dylan's approach of eschewing methods in
> favour of generic functions (multi-methods)?

In a way, yes, although Dylan is smoother in keeping parallelism
between the various objects involved, while Haskell has different
pluses.


>> The set/bag example under immutability depends on functions such as
>> put and remove returning new sets/bags -- a conceptual approach that
>> simplifies things a lot, even though implementation-wise it may be too
>> heavy.  But the insights you gain from there are still valuable and to
> some
>> extent applicable to impure worlds with mutating objects.  Among bag's
>> axioms (w has/put/remove) should be something like the material
> implication:
>>     (has item bag) => (has item (remove item (put item bag)))
>> and among set's axioms w has/put/remove should be:
>>     () => (not (has item (remove item set)))
>> so the impossibility of any one entity X satisfying both sets of axioms
>> is trivial to prove, no?
> 
> Yes. Extremely lucid and succinct!  I wonder how often such axiomatic
> analysis
> would identify pathology in our inheritance relationships.  Or does

Quite often, if we wrote complete axioms.  We don't, because even
our "assertion languages" (such as Eiffel's require/ensure clauses) are
oriented to execution, so the universal quantifiers that are such a must
for good axiomatization (in the above notation there are several implied
"for any":-) can't even be conceived within the confines of those
languages, AND because axiomatization under change is so much more
of a bother (you keep having to introduce alien ideas about "before"
and "after", seemingly intuitive to humans but extra baggage for
formal manipulation).

> "pragmatism beats purity" mean we shouldn't worry overly much about such
> things?  "If it gets the job done .." and all that. <g>

The very substantial investment in axiomatization of imperative
languages over the last 40-ish years has certainly yielded small
pragmatic benefits so far (particularly when compared to the
higher cost/benefit ratio for the effort to make FP languages
practical).  "Small" is not 0, of course.


Alex



From jason at tishler.net  Fri Apr 12 08:26:52 2002
From: jason at tishler.net (Jason Tishler)
Date: Fri, 12 Apr 2002 08:26:52 -0400
Subject: python on cygwin
In-Reply-To: <lkhemhjiyr.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1018591376.23499.python-list@python.org>
 <lkhemhjiyr.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020412122652.GB948@tishler.net>

Michael,

On Fri, Apr 12, 2002 at 09:38:22AM +0000, Michael Hudson wrote:
> (recent NT escapee -- so no more cygwin help, I'm afraid, Jason)

Congratulations on your new found freedom. :,)  I'm sorry to see you go,
but thanks for your help.

Jason




From andreas at andreas-jung.com  Sun Apr 21 22:00:36 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Sun, 21 Apr 2002 22:00:36 -0400
Subject: word stemming / conflation
References: <3CC2C23B.DAF9057D@anansispaceworks.com>
Message-ID: <003d01c1e9a1$7f7512f0$02010a0a@suxlap>

I have written  wrappers for the snowball stemmers of
Dr. Porter: http://pystemmer.sourceforge.net.

Andreas


----- Original Message ----- 
From: "Terry Hancock" <hancock at anansispaceworks.com>
To: <python-list at python.org>
Sent: Sunday, April 21, 2002 09:44
Subject: word stemming / conflation


> I was working on an algorithm to condense strings of
> text into mnemonic labels (hopefully this will produce
> more meaningful labels than simply assigning a number!).
> I got the idea from the way later versions of
> latex2html generate filenames.
> 
> Anyway, after tinkering with this for awhile I've
> discovered that it can be a bottomless pit! A little
> web research reveals various MS / PhD papers for doing
> this task (which apparently is called "conflation" as
> well as "word stemming") in English and lots of other
> languages. Looks like it's a lot harder than you
> might at first think. I didn't find much in Python,
> though.
> 
> So, I'm wondering if there are some Python resources
> (or better-yet already written modules) for doing
> this sort of thing.  I don't really need a completely
> robust system -- the occasional error will be quite
> tolerable. So I'm more interested in quick and simple
> solutions that work most of the time, rather than
> the ultra-robust library-science type.
> 
> Thanks for any information or leads you might be
> able to suggest,
> 
> Terry
> 
> -- 
> ------------------------------------------------------
> Terry Hancock
> hancock at anansispaceworks.com       
> ------------------------------------------------------
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 





From db3l at fitlinxx.com  Thu Apr 25 18:40:45 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 25 Apr 2002 18:40:45 -0400
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>
Message-ID: <uit6f1jwy.fsf@ctwd0143.fitlinxx.com>

Chris Spencer <clspence at one.net> writes:

> 	I tested this using the Python command interpreter on a Windows XP box
> running:
> Python 2.1 (#15, Apr 16 2001, 18:25:49) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> >>> float(5)/float(100)
> 0.050000000000000003
> >>> 5.0/100.0
> 0.050000000000000003
> 
> 	Okay, I might not be a math genius, but this is most definately WRONG.
> What gives?  Anyone have insight into this little "feature" of Python floating
> point operations?

See http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.098.htp

With Python 2.2, a new appendix (B) was added to the tutorial covering
floating point issues.  You can also get to it online at
http://www.python.org/doc/current/tut/node14.html

The results you are seeing aren't precisely "wrong" (certainly not in
terms of binary floating point precision), it's just more detail than
you probably expect by default.  If you thought that computers using
binary floating point (the most typical nowadays) in lieu of BCD or
other decimal based formats could precisely represent base 10 numbers
the same way you might with pencil and paper, that was an erroneous,
but forgiveable, assumption :-)

If you want Python to "fib" to you about the result (as earlier
versions did by default), you can do:

   >>> print float(5)/float(100)
   0.05

instead.  The default output at the interpreter is to use repr() to
print an object, which shows the highest precision possible to best
assure that its output could be used to re-create the object.  The
print statement will use str() which is designed for "prettier" output
and will do some rounding.

But don't fool yourself into thinking that regardless of what is being
printed, the computer is thinking of the precise value 0.05 - it
isn't.  That's the nature of floating point computations in general.
And if you're going to be writing code dependent on floating point
computations, it's important that you recognize these limitations,
since it can cause you problems (e.g., a classic example is testing
explicitly against 0.0 following a floating point computation).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From wurmy at earthlink.net  Sat Apr  6 13:41:22 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sat, 06 Apr 2002 18:41:22 GMT
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com> <3cadd841.225619243@news.laplaza.org> <Xns91E81370BB5FERASXnewsDFE1@130.133.1.4>
Message-ID: <3CAF41F5.83060882@earthlink.net>

Philip Swartzleonard wrote:

> Well, i'm not someone to ask about the detailed meanings of japanese
> words, but i'm bored, so i did some dictionary diving on these terms.
> Here's a couple hits (kanji meanings in parens).
> 
> keishikibi (form/style - system/expression - beauty) 'beauty of form'
> kotan (wither/die - faint/pale O_O) 'elegant simplicity'
> tanryoubi (simple - clear - beautiful) 'i just made this word up'

That last one looks suspiciously like "Ruby". You're not trying
to feed us subliminal messages are you? :-)

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA==')) 
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/


From peter at engcorp.com  Tue Apr 16 02:25:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 16 Apr 2002 02:25:50 -0400
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za> <3CBA43B9.22850DF1@engcorp.com> <3CBBB8A9.32CE5B9@earthlink.net>
Message-ID: <3CBBC3EE.B44E1EF6@engcorp.com>

Hans Nowak wrote:
> 
> Peter Hansen wrote:
> 
> > > http://boa-constructor.sourceforge.net
> >
> > Of course, those products are well known around here.
> >
> > What's less well known is the relationship between Boa-Constructor
> > and Zope.  I strongly suspect there isn't one.
> >
> > Can you elaborate on this supposed "incorporation of Zope
> > into Boa-Constructor"?
> 
> I downloaded the CVS version of Boa the other day, and it
> has a new tab on its components toolbar called "Zope", with
> components like DTMLDocument, DTMLMethod, ZCatalog, GadflyDA,
> etc. Not that I use it (yet)...

Well, I stand corrected! :-)  Not sure quite what the appeal
is yet, but I'll have to check it out sometime.  I suppose
it might replace WebFolders under Windows?

-Peter


From gerhard at bigfoot.de  Mon Apr  8 01:36:44 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 8 Apr 2002 05:36:44 GMT
Subject: IsList?
References: <9F9s8.37463$tg4.444710@vixen.cso.uiuc.edu>
Message-ID: <slrnab2b3b.1gd.gerhard@lilith.my-fqdn.de>

John wrote in comp.lang.python:
> Hi, How can one know wheter var 'a' is list or not? Is there any
> IsList-like function?

Python 2.2:
    type(a) is list

    but I think it would be better to use

    isinstance(a, list)

    to also test for subclasses of list

Python < 2.2:
    type(a) is type([])

        or

    from types import ListType
    type(a) is ListType

That being said, you normally shouldn't test for the type directly.
It's much better to think which "interface" you want to match, and
then test for this interface. Interfaces are implicit in Python, they
are fulfilled if the methods that make up the interface are present in
a class. So the first step is to decide what your lists have in
common, i. e. which features of them you use. Then test for the list
features you use in your application.

For example, a list is a sequence, just like a tuple is, only that
this sequence is mutable, so you could test for this with a check
like:

#!/usr/bin/env python2.2
# Requires Python >= 2.2
def is_sort_of_a_list(l):
    for method in ['__getitem__', '__setitem__']:
        if method not in dir(l):
            return 0
    return 1

print is_sort_of_a_list([5])    # list
print is_sort_of_a_list((5,))   # tuple
print is_sort_of_a_list("asdf") # string

Of course you need to change the attributes to be tested for the
application you have in mind.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From neuhauser at mail.cz  Sat Apr 27 12:31:09 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Sat, 27 Apr 2002 18:31:09 +0200
Subject: slices: foo[i:j:k] => TypeError
In-Reply-To: <aCky8.66523$vF6.2038703@news2.tin.it>
References: <mailman.1019856470.541.python-list@python.org> <aCky8.66523$vF6.2038703@news2.tin.it>
Message-ID: <20020427163109.GZ82912@roman.mobil.cz>

> From: Alex Martelli <aleax at aleax.it>
> Subject: Re: slices: foo[i:j:k] => TypeError
> To: python-list at python.org
> Date: Fri, 26 Apr 2002 22:34:46 GMT
> 
> Roman Neuhauser wrote:
>         ...
> >>>> foo[0:6:2]
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > TypeError: sequence index must be integer
> >>>> 
> > 
> > What's up? Is that a bug, or am I dumb?
> 
> http://www.python.org/dev/doc/devel/lib/typesseq.html only
> mentions x:y slices; the x:y:z form is supported only by
> Numeric, I believe.  The underlying problem is that list &c
> still use the deprecated __getslice__ (or equivalent thereof)
> rather than accepting slice objects in __getitem__ (&c for
> set and del).

    Ok. What I don't understand then is why is extended slicing
    described among primaries in the language reference. And I'm not
    talking about a development version of the manual, I'm talking about
    the 2.1 version.
    What's the expected date for the builtin types to accept slice
    objects?

-- 
FreeBSD 4.4-STABLE
6:26PM up 12 days, 7:38, 10 users, load averages: 0.04, 0.04, 0.01




From mlh at vier.idi.ntnu.no  Tue Apr  9 07:02:33 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 9 Apr 2002 11:02:33 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1018339899.2981.python-list@python.org>
Message-ID: <slrnab5ii9.ho6.mlh@vier.idi.ntnu.no>

In article <mailman.1018339899.2981.python-list at python.org>, Christian
Tanzer wrote:
[snip]
>There are several dimensions to typing, including:
>
>- static vs. dynamic
>
>- strong vs. weak
>
[snip]
>
>Another important point with typing is whether typing applies to
>variables or to objects. In Python, objects are (strongly) typed, not
>variables. Thinking `variable` when considering typing (or objects in
>general) will not work very well in Python.

Hm. I thought this was what dynamic typing was all about, i.e. having
typed values but untyped variables/parameters (which means that
variable types need not be determined statically, at compile time)...
What am I missing here?

>Names-don't-matter-except-when-they-do-ly yr's

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From jajvirta at cc.helsinki.fi  Mon Apr 22 06:50:37 2002
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 22 Apr 2002 10:50:37 GMT
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83>
Message-ID: <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>

22 Apr 2002 06:29:05 GMT Lumberjack wrote:
> Tim Daneliuk <tundra at tundraware.com> wrote:
>> Carl Banks wrote:
>>> Computer programming is not science.
>> 
>> Agreed - it is craft.
> 
> Computer programming has all the properties of literature, including 
> subjective analysis, the existence of "schools of thought," and genres, 
> among other things.

This has been a significant metaphor in the past. It's just that it 
doesn't cover software development fully or adequately. For instance,
literature is often a one (wo)man activity whereas software development
involves several contributors (sometimes even several hundred or more).
Furthermore, writing a piece is an activity that has a clearly 
identified end point. It doesn't even matter whether there's few 
typos in the final product. Software development is more of a 
maintenance activity where bugs and shortcomings are fixed throughout
the lifetime of a product. (Taken to the extreme, literally, it is
actually only about fixing bugs and shortcomings (ie. adding features)
which starts as soon as the initial release is made which is made
very early in case of extreme programming.)

I find every other metaphor lacking something essential too.  One
popular metaphor is building construction or engineering activity in
general.  Personally I just don't get the feeling that there's any of
those so called components that the final product is made of in computer
programming.  These components are supposed to be realiable (and in a
way simple) yet provide something significant to the product.  I can't
figure any non-trivial and substantial programming component which can
be viewed having the same reliability (or reliability properties, such
as "this bolt holds the pressure of N") as the compontents that
buildings are made of. 

When I think of the final products of software development and building
construction I find it hard to compare them.  Yeah, sure there is some
sort of an "architecture" that they're based on and they're both
non-trivial combinations of some sort of building blocks.  But it's just
that the software product seems so fragile.  There are of course methods
and ways to build high quality software, but no matter how protectively
it is built, there's always the chance that some part of the software
fails which makes the service provided by the software unusable. Even
in the case of formal methods there is a possibility of failure. Now,
whether it's just that software development is such a young discipline
or that there's some inherent factor that makes it uncertain and 
undeterministic is beoynd my comprehension. 

Having said all that I'll just have to remind that metaphors are of
course useful if they shed light on some aspect on the subject.  They
just shouldn't be extended too far.  (Hmh, It seems that I've been
stating the obvious all the time, well, that's just me ;-)



From h.e.w.frater at cs.cf.ac.uk  Fri Apr 19 21:05:44 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 19 Apr 2002 18:05:44 -0700
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com>
Message-ID: <e2e9e807.0204191705.6569c572@posting.google.com>

I'm importing it in the interactive interpreter using:
>>> import cgi #runs fine

here's the cgi script:

import DCOracle2
import md5 
import time 
import base64
import cgi
import rotor

connectString = "'scm1hewf/carthage at cs2000'"

def makeSessionId(st):
  string = ""
  m = md5.new()
  m.update('this is a test of the emergency broadcasting system')
  m.update(str(time.time()))
  m.update(str(st))
  return string.replace(base64.encodestring(m.digest())[:-3], '/', '$')

def enc(password):
  rt = rotor.newrotor('this is a test of the emergency broadcast system', 20)
  return rt.encrypt(password)

def checkid(id, password):
  db = DCOracle2.connect(connectString)
  c = db.cursor()
  query = "'select password from tblborrower where id = " + formid + "'"
  c.execute(query)
  return c.fetchone()

form = cgi.FieldStorage()
id = form.getvalue('id')
password = form.getvalue('password')

if (form.getvalue('id')):
  encrytedpassword = enc(password)
  sessionId = makeSessionId(id)
  result = checkid(id, encryptedpassword)


From max at alcyone.com  Mon Apr  1 14:28:12 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 01 Apr 2002 11:28:12 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net>
Message-ID: <3CA8B4CC.88C35BB0@alcyone.com>

Stefan Schwarzer wrote:

> But, if I understand your argument correctly, "True == 1" and
> "False == 0" shouldn't be true either because True and False are
> bools, and 1 and 0 are ints. (The problem with this is, that, as a
> consequence to the reasoning in the PEP, bools are ints in disguise.)

Yes, bools are really ints underneath the hood, and for backward
compatibility it makes sense that there should be implicit conversions
in numeric contexts from bools to ints, so True == 1 and False == 1 are
both true.  However, it won't go the other way around -- there won't be
(and shouldn't be) implicit conversion from _ints_ (or anything else) to
bools, so True == 2 or False == [] shouldn't be true.  (Because of
Python's typing system, they won't be errors, either, just untrue.)  If
you want conversion, you have to make it implicit with a bool
conversion:  True == bool(2) and False == bool([]) will both be true.

> Example: "(val1, val2)[0]" is pretty clear, but "(val1, val2)[False]"
> looks quite strange. If that's the price of the change I would still
> prefer that "2 > 1" gives 1, not True.

But that's simply a case of bad style, not a demonstration of the
inappropriateness of Booleans.  After all, even now you could have
written

	(x, y)[operator.truth([])]

This is ugly to be certain, but doesn't indicate a flaw in the language.

> So I see these choices:
> 
> 1 Have a bool type without these strange interactions with ints (i. e.
>   a bool should never equal an int, as an int should never equal a
>   string, as you explain above), which breaks a lot of code.
> 
> 2 Or, as the PEP suggests, introduce a bool type which behaves quite
>   unintuitive, but keeps existing programs running.

I don't see why you think the behavior (I presume you're referring to
syntaxes like True == 2) is counterintuitive.  Why _would_ you expect
this to do what you want?  It doesn't for any other Python types that
I'm aware of.  == tests for equality, it doesn't do automatic type
coercion.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From robin at jessikat.fsnet.co.uk  Thu Apr 18 05:31:56 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 18 Apr 2002 10:31:56 +0100
Subject: pyRXP Memory Likage???
References: <c5f48de3.0204161113.14919d25@posting.google.com>
Message-ID: <fdB2AUAMKpv8EwiV@jessikat.demon.co.uk>

In article <c5f48de3.0204161113.14919d25 at posting.google.com>, Jose
Manuel Esteve <chema at imdea.com> writes
>I have been playing with pyRXP and is incredibly fast as compared with
>minidom, also the memory footprint is smaller for one file, however if
>you loop for parsing a lot of xml strings the memory usage gets higher
>and higher as compared to minidom
>
>import pyRXP
>xml='<start><tag1>text here</tag1></start>'
>P=pyRXP.Parser()
>for i in range(0,100000):
>    tup=P.parse(xml)
>
>will show the behaviour
yikes, better get my debugging hat on :(
-simple failure is better than complex failure somtimes-ly yrs- 
Robin Becker


From guido at python.org  Sat Apr  6 10:45:37 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 06 Apr 2002 15:45:37 GMT
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it> <just-89BA7A.17235906042002@news1.xs4all.nl>
Message-ID: <3CAF18F8.56821C0B@python.org>

Just van Rossum wrote:
> 
> In article <uDDr8.66496$S52.2342405 at news2.tin.it>,
>  Alex Martelli <aleax at aleax.it> wrote:
> 
> > Linus Thorvalds seems to have done pretty well with picking TWO middle
> > grounds -- two parallel tracks for Linux, "stable" and "experimental".
> 
> It seems 2.1.x and 2.2.x currently serve as these two tracks,
> respectively. I think it's a very good idea to think about how to
> formalize this for future releases.

Agreed.  I've moved this to python-dev; see you all there.

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


From max at alcyone.com  Tue Apr  2 17:37:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 02 Apr 2002 14:37:42 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it>
Message-ID: <3CAA32B6.92D26E8C@alcyone.com>

Alex Martelli wrote:

> The Language Reference manual of the Python language disagrees
> with you.  E.g. at
> http://www.python.org/dev/doc/devel/ref/Booleans.html :
> 
> """
> In the context of Boolean operations, and also when expressions are
> used by control flow statements, the following values are interpreted
> as false: None, numeric zero of all types, empty sequences
> (strings, tuples and lists), and empty mappings (dictionaries).  All
> other values are interpreted as true.
> """

Note that's not complete; it doesn't mention instances of custom classes
which override the __nonzero__ method to return one of these types.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From lac at strakt.com  Sat Apr 27 16:37:00 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 27 Apr 2002 22:37:00 +0200
Subject: inline assignments in conditionals 
In-Reply-To: Message from jaf <jaf@lcsaudio.com> 
   of "Sat, 27 Apr 2002 12:33:47 PDT." <2143388891-BeMail@sugoi> 
References: <2143388891-BeMail@sugoi> 
Message-ID: <200204272037.g3RKb0kG019255@ratthing-b246.strakt.com>

Jeremy:
[snip]
> Writing things that way is really unnatural, at least for English
> speakers.  It makes the intent of the code harder to understand (at
> least for me).  For example, if you were talking to a friend, you
> wouldn't say:
> 
>    "If today is Sunday, I'm going to go shopping"
> 
> but rather:
> 
>    "If Sunday is today, I'm going to go shopping"

Interesting.  I would never use the second construct, and I use the
first one all the time.  I wonder if this is a difference between
people whose first language is English, and some for whom it was not.
I don't think that anybody has had any trouble understanding me when I
say 'If today is Saturday, then we are going to the concert house,
otherwise we are going to hack.'  But then I have been writing
if 0 == x for a long time as well.  (It was great -- Borodin Quartet
playing Sjostakovitj and Beethoven, by the way.)

Laura Creighton




From aleax at aleax.it  Tue Apr  2 11:00:17 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 16:00:17 GMT
Subject: simulated annealing
References: <3CAA1FDC.7FEF351D@gol.ge>
Message-ID: <lAkq8.43728$pT1.1224293@news1.tin.it>

Giorgi Lekishvili wrote:

> Hi!
> Has somebody written a free module for simulated annealing?

http://margaretmorgan.com/wesley/python/


Alex



From cbbrowne at acm.org  Wed Apr 24 00:14:50 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: Wed, 24 Apr 2002 00:14:50 -0400
Subject: Python's Lisp heritage
References: <mailman.1019613886.1762.python-list@python.org>
Message-ID: <m3662henr9.fsf@chvatal.cbbrowne.com>

Centuries ago, Nostradamus foresaw when Tim Peters <tim.one at comcast.net> would write:
> [James J. Besemer]
>> ...
>> The original semantics (as formally defined in Lisp itself
>> [McCarthy]) called for dynamic binding.
>
> It seems this would have been a surprise to McCarthy <wink>:
>
>     http://www-formal.stanford.edu/jmc/history/lisp/node4.html
>
> See paragraph "d. Free variables".  When a user bumped into a function where
> the difference between dynamic and lexical scoping mattered, and lexical
> scoping was what they needed, "I must confess that I regarded this
> difficulty as just a bug".  The understanding of these issues was much
> muddier at the time.
>
>> ...
>> I find it highly ironic that Perl originally did it the wrong way
>> ("local"), although later they introduced a now-recommended static
>> binding ("my").
>
> More recently, REBOL went down the same path.  Dynamic scoping is attractive
> for implementation reasons; everyone who succumbs eventually regrets it
> (except for RMS in elisp ...).

Dynamic scope buys you the ability to not need to specify _all_ the
values that you might be customizing.

If a new Emacs mode requires adding in a bunch of additional
parameters, dynamic scope lets them be visible throughout the scope
during which they are "live" as opposed to just within the environment
in which they were defined.

>From a paper on Emacs:

  "Some language designers believe that dynamic binding should be
  avoided, and explicit argument passing should be used
  instead. Imagine that function A binds the variable FOO, and calls
  the function B, which calls the function C, and C uses the value of
  FOO. Supposedly A should pass the value as an argument to B, which
  should pass it as an argument to C.

  This cannot be done in an extensible system, however, because the
  author of the system cannot know what all the parameters will
  be. Imagine that the functions A and C are part of a user extension,
  while B is part of the standard system. The variable FOO does not
  exist in the standard system; it is part of the extension. To use
  explicit argument passing would require adding a new argument to B,
  which means rewriting B and everything that calls B. In the most
  common case, B is the editor command dispatcher loop, which is
  called from an awful number of places.

  What's worse, C must also be passed an additional argument. B
  doesn't refer to C by name (C did not exist when B was written). It
  probably finds a pointer to C in the command dispatch table. This
  means that the same call which sometimes calls C might equally well
  call any editor command definition. So all the editing commands must
  be rewritten to accept and ignore the additional argument. By now,
  none of the original system is left!"

If all you have is lexical scoping, parameters have to get explicitly
passed down the chain in order to get from function A to function C.
That means introducing additional parameters to function B, which
didn't actually care about those extra values.

Most of the time, lexical scope is likely to be more useful.  But
there are places where dynamic scope is to be preferred.
-- 
(concatenate 'string "cbbrowne" "@cbbrowne.com")
http://www.cbbrowne.com/info/spreadsheets.html
How does the guy who drives the snowplow get to work in the mornings?


From bokr at oz.net  Wed Apr 17 01:12:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 05:12:57 GMT
Subject: Why don't strings share data in Python?
References: <mailman.1018937610.11880.python-list@python.org> <87n0w37z0n.fsf@mathdogs.com>
Message-ID: <a9j08p$nj6$0@216.39.172.122>

On 16 Apr 2002 23:03:04 -0500, Mike Coleman <mkc+dated+1021607974.93924d at mathdogs.com> wrote:

>Tim Peters <tim.one at comcast.net> writes:
>> There are many reasons, but one of them is invisible to Python users:
>> because Python is intended to play nice with C libraries, all Python strings
>> end with a \0 byte.
>
>Well, that's certainly a bug in the ointment.
>
>I almost thought I might be able to use NumPy arrays, as it looks like they do
>share data.  They're not hashable, though, unfortunately, which kills the
>usage I had in mind.
>
So what did you have in mind?

Regards,
Bengt Richter


From magnus.heino at sigma.pleon.se  Fri Apr 26 16:37:07 2002
From: magnus.heino at sigma.pleon.se (Magnus Heino)
Date: Fri, 26 Apr 2002 22:37:07 +0200
Subject: Simple bar graphs
References: <Xns91FCD43535458Latzikatz@139.174.2.56>
Message-ID: <hUiy8.8930$WJ4.11848@news1.bredband.com>

> I'm looking for a _pure_ Python module which generates some very simple
> bar graphs and outputs them as gif or jpg. I looked through Parnassus and
> other sources but could not get enough information to make a decision.

Well, this isnt python, but if the bar graphs are simple, I guess you can 
use this?

http://developer.netscape.com/docs/technote/javascript/graph/

-- 

  /Magnus


From mal at lemburg.com  Tue Apr  9 16:32:47 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 09 Apr 2002 22:32:47 +0200
Subject: Thread/Process-safety of functions in the Standard Library, eg the 
 dbm classes
References: <3cb32a63.17540718@news.t-online.de> <a8vamd$vp443$1@ID-59885.news.dfncis.de>
Message-ID: <3CB34FEF.3AF483D9@lemburg.com>

Thomas Heller wrote:
> 
> > Furthermore, do you have any suggestions for these requirements: I
> > want to be able to store some data (<100K), and access it from
> > multiple processes in a safe way. Currently I'm using mySQL with
> > mysqldb, which seems like heavy overkill for that amount of data.
> > Plus, I want users to use my program without having to first install
> > MySQL (or any other dbm for that matter).
> >
> > [A hardcore idea would be to make a "anydbm-server" that you can
> > communicate with using sockets ;]
> 
> Sound like a job for Gadfly ;-)
> Gadfly includes a server which communicates via sockets with the
> backend, IIRC.

Or have a look at mxBeeBase (the beta which I announced on the
egenix-users mailing list, not the current release one (0.6.0)).
It uses locks for accessing the data files.

-- 
Marc-Andre Lemburg
________________________________________________________________________
Business:                                        http://www.lemburg.com/
Python Pages:                             http://www.lemburg.com/python/




From eppstein at ics.uci.edu  Wed Apr 24 20:45:29 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 24 Apr 2002 17:45:29 -0700
Subject: An IMHO interesting optimization question...
References: <3cc73c81@nntp.server.uni-frankfurt.de>
Message-ID: <eppstein-F7AFD1.17452824042002@news.service.uci.edu>

In article <3cc73c81 at nntp.server.uni-frankfurt.de>,
 Michael 'Mickey' Lauer <mickey at tm.informatik.uni-frankfurt.de> wrote:

> stats = {}
...
> 
> # initialize dictionary, so we don't have to check for key-existance later
> for i in xrange( 1, longWord+1 ):
>     stats[i] = 0

Why are you using a dictionary when you know the indices are a 
predetermined range of integers?  Wouldn't it be better to do

    stats = [0]*(longWord+1)

?

Also, if speed is really a problem, you might think about replacing the 
count() function with inline code.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From Chris.Barker at noaa.gov  Mon Apr  1 18:21:29 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 01 Apr 2002 15:21:29 -0800
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com> <slrnaaftai.8mf.gerhard@lilith.my-fqdn.de> <w20q8.1572$0T6.38985@e3500-atl1.usenetserver.com>
Message-ID: <3CA8EB72.F6E7973C@noaa.gov>

Robert Oschler wrote:
> Also, will it coexists with Konqueror (which shells Netscape I think) and
> Netscape?

NO, it doesn't. Konqueror is it's own beast, written from scratch, I
think. It'a a pretty nice web brower, I use it most of the time.

You can have any number oif browsers on your Linux box, this isn't
Monopoly style software...


-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From richard at electrophobia.com  Wed Apr 24 02:33:53 2002
From: richard at electrophobia.com (Richard Parker)
Date: Wed, 24 Apr 2002 06:33:53 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com>
Message-ID: <B8EB9FE2.EF8C%richard@electrophobia.com>

in article 7x3cxlk7qb.fsf at ruckus.brouhaha.com, Paul Rubin at
phr-n2002a at nightsong.com wrote on 4/23/02 10:04 PM:

> I've put together an encryption function written in Python using the
> SHA module to provide a keystream in output feedback mode.

I just took a quick look at your Python code.  I'd encourage you to not use
the secret prefix method to construct a MAC from a hash function, i.e.
MAC(x) = H(K || x).  This method is generally considered to be insecure.
Use the HMAC construction instead.

> ...this function makes a nonce based on the key and plaintext as well
> as the current time and an internal state based on the time of
> initialization and the values of previously issued nonces.  It
> currently doesn't try to use any other system-dependent entropy.

The ciphertext and key passed to the decryption function could be used as
additional sources of entropy for your "_state" variable.

> It's possible that a nonce could get re-used if two separate Python
> instances simultaneously encrypt the same plaintext with the same key.
> That could happen, for example, on a cgi web server where two requests
> come in very close together and each starts its own Python instance.
> That could be a security failure in some applications, but hopefully
> not many.

Perhaps you could add a function to your API that would allow the programmer
using your API to provide additional an source of entropy.  This would
enable the hypothetical programmer of the CGI using your library to fix the
problem by using information from the HTTP requests as a source of entropy.

-Richard



From phr-n2002a at nightsong.com  Thu Apr  4 09:27:12 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 06:27:12 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <3CABDF6B.5020808@jerf.org> <3CAC19D8.142A918@sympatico.ca> <7xhemru1pg.fsf@ruckus.brouhaha.com> <3CAC5182.F6CB3F07@sympatico.ca>
Message-ID: <7xzo0j4jzz.fsf@ruckus.brouhaha.com>

vio <vmilitaru at sympatico.ca> writes:
> > This makes no sense.  WHY do you want to migrate all that Java to Python?
> > Why not leave it in Java and keep using it?  That's what none of us
> > are understanding.  What do you GAIN from making an unmaintainable
> > machine translation?
> 
> Focus.

> Focus on one language and one environment. Which happens to be
> python. Migrating legacy code is a strategic decision. Obviously,
> the goal is to NOT end up with an unmaintaibable machine
> translation. In that case the "solution" becomes worse than the
> "problem" (one of python's big appeal is its ease of
> maintenance).

You won't get that from an automatic translation.  Part of the ease
of maintenance comes from the code for an app just being smaller in
Python than it would be in Java.  But a translation would be bigger
and not smaller.  And the code output will be awful, because the
translation system won't have any way to figure out what the programmer
was actually thinking.  Look at a Babelfish translation of French
into English to see what you'll be dealing with code-wise.  This
is what I mean when I say that the translation project might fail
disastrously, but if it succeeds, you'll have an even bigger disaster.

What the heck kind of application has 100 MB of Java code anyway?
It sounds like it may already be the output of some automatic process.


From herrn at gmx.net  Wed Apr  3 19:58:10 2002
From: herrn at gmx.net (Marco Herrn)
Date: 4 Apr 2002 00:58:10 GMT
Subject: turning a string into an object name
Message-ID: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de>

Hello. I know that it is possible to turn a string into a funtion call by
using getattr(). But what I want to achieve is turning a string into an
object name. 
I would have a method like

  foo(self, objname, objtype)

Then I would call this method like
  
  foo(myObj, bar)

Then the method foo should create an object with the name "myObj" from the
type bar.

Is such an thing possible? And if, how?

Thanks in advance
Marco

-- 
Marco Herrn             herrn at gmx.net
(GnuPG/PGP-signed and crypted mail preferred)
Key ID: 0x94620736



From mats at laplaza.org  Wed Apr  3 14:43:32 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Wed, 03 Apr 2002 19:43:32 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net>
Message-ID: <3cab56b5.61383044@news.laplaza.org>

On 01 Apr 2002 21:56:44 -0500, nbecker at fred.net wrote:

:Keep lobbying Redhat.  The correct solution is for them to put
:
:#! /usr/bin/python1.5
:
:If they need a specific version.  I argued this with them already.
:Keep up the pressure.  Maybe they will get it.


The simple fact is that they're not going to make any changes in this
area (even to just fix a batch of scripts) until the next major
version, which unfortunately is still a number of moons away.  The
next release was going to be 8.0, but will now be 7.3, if they stories
I've heard are correct, so a reasonble guess might be "no earlier than
the end of the year".  Ouch.


Mats Wichmann



From tanzer at swing.co.at  Fri Apr  5 01:38:35 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 05 Apr 2002 08:38:35 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "Thu, 04 Apr 2002 23:49:25 BST."
             <zpL9tKA1hNr8EwkQ@jessikat.fsnet.co.uk>
Message-ID: <E16tNMp-0001Ob-00@tswings.swing.cluster>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

> In article <mailman.1017946214.8160.python-list at python.org>,
> arthur.siegel at rsmi.com writes
(snip)
> >Guido will always be right to a following that will work to find
> >the meaning that makes him right.
>
> In any large subject a 'Guido' will always claim to be bigger than 'God'
> thus leading to Sergeant Pepper phenomena.

Robin & Arthur,

Why don't the two of you go and design and use a language of your own?

Your repeated ad-hominem attacks on Guido make me sick.

Or-go-get-a-life,

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





From dale at riverhall.NOTHANKS.co.uk  Mon Apr  1 14:13:32 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Mon, 01 Apr 2002 20:13:32 +0100
Subject: Access to main namespace from imported module?
References: <re8haugrbcchbephckl9t1p1s27tvg18hr@4ax.com> <Xns91E3D4A144441cliechtigmxnet@62.2.16.82>
Message-ID: <3achauk064f59svh1qbem3q2iarlcurumr@4ax.com>

Chris Liechti <cliechti at gmx.net> wrote:

>Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote in 
>news:re8haugrbcchbephckl9t1p1s27tvg18hr at 4ax.com:
>
>> Just curious, like, but is it possible for an included module to
>> reference the namespace of the outer-most module, "__main__".
>
>just do an import
>>>> import __main__
>>>> __main__.var = 7
>etc.
>
>chris


Aha!

Cheers.
--
Dale Strickland-Clark
Riverhall Systems Ltd


From sandskyfly at hotmail.com  Wed Apr 10 19:30:20 2002
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 10 Apr 2002 16:30:20 -0700
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83>
Message-ID: <b03e80d.0204101530.2483536d@posting.google.com>

Lumberjack wrote 

> Amazing to see so much slow learning taking place. So why don't you all do 
> it again some time. Like the next PEP and the next one and the next one, 
> and so on ad infinitum. Just like previous PEPs. The old Chinese water 
> torture would be more humane than continuing these unending so-called 
> "enhancements" to the language. Drip drip drip... Neither you nor Guido 
> have a clue; unless this is his secret plan to destroy the appeal of Python 
> by setting Python programmers at each other's throats, or they begin to 
> associate emotional discord with the language and decide to keep their 
> distance. If so, the plan is well executed.
> 
> People aren't going to change, so Guido needs to adjust to the reality of 
> the political consequences of his current strategy.

Trollshit.

If you don't have faith in Guido's and Tim's judgement calls why don't
you buzz off and create your own language.

Sandy


From ffjhenon at club-internet.fr  Sun Apr 14 12:17:47 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Sun, 14 Apr 2002 18:17:47 +0200
Subject: Lambda  added to release 2.2 ?
Message-ID: <3CB9ABAB.E37E685B@club-internet.fr>

When I start my python script, here is what I get on the console from
Linux (Python 2.1) :
a26.py:51: SyntaxWarning: local name 'self' in '__init__' shadows use of
'self' as global in nested scope 'lambda'
  def __init__(self, master):


When I use the lines below :
>    insertmenumaterial.add_command(label="interior", command = lambda :
self.insert_example("interior.txt"))

>     def insert_example(self,nom):
>        chemin="insert"
>       contents=fd.read()
>         self.text.insert(INSERT,contents)
 >        fd.close()


I get the following message :
>>>     Exception in Tkinter callback
>>> Traceback (most recent call last):
>>>   File "/usr/lib/python2.1/lib-tk/Tkinter.py", line 1285, in
__call__
>>>     return apply(self.func, args)
>>>   File "a26.py", line 219, in <lambda>
>>>     insertmenumaterial.add_command(label="interior", command =
lambda : self.insert_example("interior.txt"))
>>> NameError: global name 'self' is not defined

I don't have any problem under W2000 ( Python 2.2)
Is lambda function a new add-on to 2.2 ? I thought it was added ages
ago.


Thanks





From marklists at mceahern.com  Tue Apr 30 13:28:25 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 12:28:25 -0500
Subject: remove read only flag
In-Reply-To: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEHGKNAA.marklists@mceahern.com>

> How can one remove the read-only flag from a file
> (e.g. c:\temp\file.txt) using python in win32.

http://aspn.activestate.com/ASPN/Python/Reference/Products/ActivePython/Pyth
onWin32Extensions/win32api__SetFileAttributes_meth.html

// mark





From m_mariappanX at trillium.com  Mon Apr 22 02:33:44 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Sun, 21 Apr 2002 23:33:44 -0700
Subject: building python in win2k with BLT libraries?
Message-ID: <53A7943A5BD8D411B6930002A5073155022D745D@bgsmsx90.iind.intel.com>

Hi all,

I've successfully build python-2.2 with tkinter + BLT libraries in linux and
solaris and is working fine.
how can i build pyhton-2.2 with tkinter+BLT libraries in win2k? Anybody have
tried this combination of build before.

I've never tried before to build python in win2k.  
Any pointer to docs is enough for me.

Thanks in advance.
Maharajan





From safarzadeh at yahoo.com  Sat Apr 20 12:00:35 2002
From: safarzadeh at yahoo.com (H. Safarzadeh)
Date: Sat, 20 Apr 2002 12:00:35 -0400
Subject: Question about IDLE
References: <0p5w8.17527$CH.4180@rwcrnsc52.ops.asp.att.net>
Message-ID: <a9r5cn$e2e$1@nsnmpen2-gest.nuria.telefonica-data.net>

Matt wrote:

> I'm trying to learn Python, and I was wondering if the IDLE program was
> sufficient enough for basic programming, or should I find some sort of
> text editor?
> I'm currently using a win32 box.
> 
> Also, any links to tutorials or general information about Python that is
> not listed on the Python website would be much appriciated.
> 
> -Matt

I think it is sufficient even for complex programs. You will be comfortable 
with it.



From logstx at bellatlantic.net  Thu Apr  4 00:41:36 2002
From: logstx at bellatlantic.net (logistix)
Date: Thu, 04 Apr 2002 05:41:36 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com>
Message-ID: <kIRq8.203$8r5.191@nwrddc01.gnilink.net>

Both repr() and str() will return "True" or "False", so:

>>> "1 < 2 is %s" % (1 < 2)
1 < 2 is False

should be all you need.
Unless you want a non-bool to throw an error, but I don't see why you would.

--
-

"Paul Rubin" <phr-n2002a at nightsong.com> wrote in message
news:7xy9g4nk9w.fsf at ruckus.brouhaha.com...
> I think there will need to be a format code for truth values:
>
>   ("%b" % True) => "True"
>
> etc.




From gtalvola at nameconnector.com  Tue Apr 16 14:39:52 2002
From: gtalvola at nameconnector.com (Geoffrey Talvola)
Date: Tue, 16 Apr 2002 14:39:52 -0400
Subject: python threading
Message-ID: <BFFEB31BC23CCD44B8924E139E1D87E448EF1F@MAILBOX.nameconnector.com>

Alex Martelli wrote:
> On Tuesday 16 April 2002 07:35 pm, Justin Sheehy wrote:
> 	...
> > I've wrapped Queues with Events before, and it makes things a little
> > nicer.  It is quite straightforward to make a Queue that exposes a
> > get method that will simply block until an Event is set by another
> > thread or an optional timeout has elapsed.
> >
> > It doesn't solve the whole problem, but waiting on an Event is much
> > nicer than a get/sleep loop.
> 
> Yes, but a GUI thread can't do that either -- nor can a network one;
> they must be ready for user-input events and netword-data-arriving
> events too.  The problem is how to seamlessly merge waiting for
> THOSE kinds of things and for Python-thread-stuff kinds...
> 

Don't most GUI toolkits have some way to post an event to the main GUI
thread's event loop from another thread?  wxPython has wxPostEvent() which
does this.

So you could write a wrapper class around Queue that posts a "wakeup" event
to the main GUI thread after an item is added to the queue.  The event
handler for the "wakeup" event would process whatever items had been added
to the Queue and then return control to the event loop.

- Geoff




From aleax at aleax.it  Tue Apr  9 05:00:55 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 09:00:55 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <slrnab2inr.8qu.mlh@vier.idi.ntnu.no> <Lpgs8kKkXgpE092yn@gnosis.cx> <mailman.1018301543.25726.python-list@python.org>
Message-ID: <b5ys8.62201$pT1.1789842@news1.tin.it>

Lulu of the Lotus-Eaters wrote:

> |the Lotus-Eaters wrote:
> |>    fs = (can_be_moved, can_be_sized, can_be_stacked,
> |>          can_be_coloured, can_be_hidden)
> |>    major = len([1 for f in fs if f(o)]))
> 
> |from operators impor add
> |fs = ...
> |major = reduce(add, [f(o) for f in fs])
> 
> Well... my version is two bytes shorter :-) (more if you count the extra
> import).

I personally prefer:

major = len(filter(None, [f(o) for f in fs]))

but I understand your reluctance to mix older features, such as filter,
with newer ones, such as list comprehension, when the latter can do
the job by themselves.  With old features only:

major = len(filter(None, map(apply, fs, len(fs)*[o])))

is arguably a bit clunky.  However, both of these filter-based
alternatives share the advantages of your solution vs Magnus'.


Alex



From stojek at part-gmbh.de  Wed Apr 10 11:05:20 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 10 Apr 2002 15:05:20 GMT
Subject: how to read hardware info on Win NT
References: <3cb2e686.21131515@news.easynews.net>
Message-ID: <3cb45107.23060437@news.easynews.net>

Thanks for the __winreg hint. Looks like I have to get some infos
about the registry because I have no idea how to find the right key.

If anyone knows, please tell me.

I need the id of my ethernet card for a license documentation tool we
want to write in Python. As many commercial softwares are using this
id for generating the license keys etc. I thought there should be
module in Python.  Maybe there is a modul for commercial distribution
of Python software?

bye,
marcus



From brueckd at tbye.com  Mon Apr 22 10:35:26 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 22 Apr 2002 07:35:26 -0700 (PDT)
Subject: Science And Math  Was: Python's Lisp heritage
In-Reply-To: <20020422124126.GD23706@andy.tynant.ftech.net>
Message-ID: <Pine.LNX.4.33.0204220729500.25157-100000@ns0.tbye.com>

On Mon, 22 Apr 2002, Andy Gimblett wrote:

> On Mon, Apr 22, 2002 at 12:57:38PM +0100, Gon?alo Rodrigues wrote:
>
> > You should understand that this is a philosophical viewpoint. The
[snip]

> To put this another way, ask yourself where you stand on this
> question: is mathematics discovered or invented?

Is there a difference? One could argue that the only difference between
discovery and invention is the arrogance of the person reporting it.

-Dave





From jimd at vega.starshine.org  Sun Apr  7 16:32:58 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 7 Apr 2002 20:32:58 GMT
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4> <AkIr8.5651$Sa2.1837@nwrddc01.gnilink.net> <MPG.17197f04a3f463e8989680@news.planet.nl>
Message-ID: <a8qadq$3di$1@news.idiom.com>

In article <MPG.17197f04a3f463e8989680 at news.planet.nl>, Johannes Gijsbers wrote:

>logistix wrote:
>> Use python's type system:

>> >>> import types
>> >>> a, b = 1, 1.0
>> >>> if type(a) == types.IntType: print "int"
>> ...
>> int
>> >>> if type(b) == types.IntType: print "int"
>> ...
>> >>>

>That's not really a good idea, as math.sqrt() returns a float. I'd say:
 ...
>Johannes Gijsbers

 Worse, it doesn't account for subclasses of integer.

 In Python it is considered poor form to check for types;  it's 
 better to test for attributes or properties.  So one might
 take an anonymous object and do something like:

	def deal_with_unknown(a)
		if hasattr(a,'__mod__') and not a % 1:
			....

 ... since we can reasonably assert that all subclasses of numbers
 and integers should have (or inherit) an __mod__ (modulo) attribute.
 This example should short circuit so that any non-number that's
 passed to it will not raise an exception because a % 1 will be
 skipped if hasattr() is false.



From martin at v.loewis.de  Thu Apr  4 14:48:51 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 04 Apr 2002 21:48:51 +0200
Subject: minidom.parseString and =?iso-8859-1?q?=E5=E4=F6?=
References: <Bm1r8.1138$bB4.399@news2.bredband.com>
Message-ID: <m3r8lvdz30.fsf@mira.informatik.hu-berlin.de>

Magnus Heino <magnus.heino at sigma.pleon.se> writes:

> What is the best way to solve this?
> 
> [magnus at daysleeper magnus]$ python2.2
> Python 2.2 (#1, Mar  1 2002, 14:43:56)
> [GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-104)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> mystring='<foo>???</foo>'

The best solution is to use well-formed xml, i.e. add a 

<?xml version="1.0" encoding="iso-8859-1"?>

header. Alternatively, encode your document in the XML default
encoding, i.e. UTF-8.

Regards,
Martin



From dalke at dalkescientific.com  Mon Apr  8 01:01:54 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 7 Apr 2002 23:01:54 -0600
Subject: Function question
References: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu>
Message-ID: <a8r8b5$3ra$1@slb2.atl.mindspring.net>

John wrote:
>Let's say there is a function with two arguments.
>def add(a,b): return a+b
>
>What I want to do is:
>I want to get 'a' dynamically..
>But once a value is decided, I want a separate function (say, new_add) that
>does
>new_add(b) (equals to) add(a,b)

What you want is often known as "currying".  From the ActiveState Cookbook
at http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549
are several implementations, the most recent being Alex Martelli quote
of Nick Perkins for a 2.2-specific version

def curry(*args, **create_time_kwds):
    func = args[0]
    create_time_args = args[1:]
    def curried_function(*call_time_args, **call_time_kwds):
        args = create_time_args + call_time_args
        kwds = create_time_kwds.copy()
        kwds.update(call_time_kwds)
        return func(*args, **kwds)
    return curried_function



Used, for example, as (not tested):
>>> import operator
>>> add_two = curry(operator.add, 2)
>>> add_two(4)
6
>>>

See also google groups for commentary on this topic.

                    Andrew
                    dalke at dalkescientific.com






From lists.ASkwar at DigitalProjects.com  Wed Apr 24 17:57:52 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Wed, 24 Apr 2002 23:57:52 +0200
Subject: Faster md5/1-way encryption?
Message-ID: <20020424215752.GC15371@teich.Garten.DigitalProjects.com>

Hi!

I'm developing an application in Python 2.2 which will read a file and
then calculate the md5 for each line.  This is app is goint to use
wxPython as the GUI frontend.  The app should finally run on Linux and
Windows.

The files which should be processed are rather large (100,000+ lines)
and on my machine this takes ages to run.  I'm currently doing it like
this:


import os
import xreadlines
import threading
import md5

class md5Thread(threading.Thread):
    def __init__(self, Dateiname):
        threading.Thread.__init__(self)
        self.done = 0
        self.Zeilennummer = None
        self.Zeilenanzahl = None
        self.Dateiname = Dateiname
        self.Inhalt = None
        self.max = 1000
        
    def run(self):
        dateiname = self.Dateiname
        self.Zeilenanzahl = os.path.getsize(dateiname)
        linesep = os.linesep
        datei = file(dateiname, 'r', 1)
        
        self.Zeilennummer = 0
        self_Inhalt = ''
        i = 0
        for zeile in xreadlines.xreadlines(datei):
            md5_zeile = md5.new(zeile).hexdigest()
            self_Inhalt += md5_zeile + linesep
            self.Zeilennummer = datei.tell()
            self.ZeilennummerZeile = i
            i += 1
            
        self.Inhalt = self_Inhalt
        self.done = 1

As you can see, I've put the md5 generation stuff into a seperate
thread.  This thread is started when a button is pressed, like this:

        mt = md5Thread(filename)
        mt.start()

Further, I use a wxTimer to poll the current byteposition of the thread
and display it in a wxGauge and wxTextctrl.

Since this is my first Python app, there are for sure some things that
can be optimized.

I suppose what's taking so very long is 
a) The md5 generation itself
b) That I instantiate a new md5 object for every single line and thus
will have 100,000+ objects.

Could somebody please tell me, how I could speed the whole thing up?  I
don't have to calculate the md5 sum of every line, but I need to use
some sort of 1-way encryption.  So, if there are faster alternatives
than md5, I'd be happy as well.

Thanks,

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 16 hours 47 minutes




From tdelaney at avaya.com  Tue Apr  2 21:50:18 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 3 Apr 2002 12:50:18 +1000 
Subject: PEP 285: Adding a bool type
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A246@natasha.auslabs.avaya.com>

> From: Erno Kuusela [mailto:erno-news at erno.iki.fi]
> 
> In article <3CAA5777.66E7C2D5 at python.org>, Guido van Rossum
> <guido at python.org> writes:
> 
> | I think it's time I learn to say "yes" even if the community doesn't
> | see the light yet.

FWIW, I haven't posted a vote yet, so ...

I am very much in favour of a boolean type. I've read every single message
on the topic, and still believe that a boolean type, with True and False
would be a Good Thing(TM).

I do have a couple of issues with it though.

1. I don't know that bool should inherit from int. bool should definitely
interact with int transparently, but I feel isinstance(obj, int) and
isinstance(obj, bool) should never return true for the same object. I know
that inheriting from int is the easiest way to get the behaviour we want,
but I would prefer coercion to inheritance.

2. I think the documentation of the boolean type, and True and False
objects, should *very* strongly state that comparing to True and False is a
Bad Thing(TM) - that the values should only be used in assignment and other
appropriate places. Likewise, the tutorial will need to stress this.

3. I think it will need to be introduced with a __future__ statement as
there will be code breakage (very little perhaps, but some). This will also
give people time to get used to the idea.

> the following language changes come immediately
> to mind as feeling superfluous:
> - list comprehensions (map/filter)

I find I always use map/filter. I just don't seem to think in
comprehensions. However, *many* people use and like list comprehensions ...
so it hardly seems superfluous.

> - f(*args, **kw) (apply)

I really like this syntax, but don't get to use it much (many of my scripts
need to be 1.5.2-compatible).

The only thing added to Python since 1.5.2 that I strongly disagree with is
print >> which I consider a real wart and will never use.

And of course I'd really like built-in support for rationals :)

Tim Delaney



From stom at hotpop.com  Thu Apr 25 01:23:31 2002
From: stom at hotpop.com (Stom Fairecroire)
Date: Thu, 25 Apr 2002 13:23:31 +0800
Subject: Python v. Euphoria
Message-ID: <394fcus92g8rt0e2ektsd36s9qu7cq2cij@4ax.com>

Hi, All
Many, many years ago I did a lot of programming in HPL (Hewlet Packard
Language) then did lots using PICK basic. Now I want to start some
very low key programming again - just for my own amusement. Anyone
have any views on the pros and cons  of Python v. Euphoria ? Which
would be the better option?
Regards and thanks,
Stom


From dalke at dalkescientific.com  Tue Apr  2 04:59:28 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 2 Apr 2002 02:59:28 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <3CA95E0A.3070507@mxm.dk>
Message-ID: <a8bvd0$rg3$1@slb3.atl.mindspring.net>

Max M wrote:
>This is a common idiom:
>
>def f(names=None):
>     if names == None:
>         names = []
>     ...

Hmmm.  I would have done "if names is None".  No chance of confusing
the object with an object that overrides __cmp__.

I rarely, if ever, need tests of the form "ret == None".

                    Andrew
                    dalke at dalkescientific.com





From aleax at aleax.it  Sun Apr 28 02:30:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 08:30:29 +0200
Subject: Update locals()
In-Reply-To: <20020427233027.GE16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke> <ySFy8.28416$8D3.759092@news1.tin.it> <20020427233027.GE16217@devel.trillke>
Message-ID: <0204280830290C.16995@arthur>

On Sunday 28 April 2002 01:30, holger krekel wrote:
	...
> speaking of it, why is it possible (read: works) to say
>
> class dictlookup(dict):
>     __getattr__ = dict.get
>
> but not
>
> class dictlookup(dict):
>     __getattr__ = dict.__getitem__
>
> am i missing something?

Types:

>>> dict.__getitem__
<slot wrapper '__getitem__' of 'dict' objects>
>>> dict.get
<method 'get' of 'dict' objects>

Not all callable types function identically in all positions (at
least not yet, probably not ever) and in particular a 'slot wrapper'
instance need not be usable in lieu of a 'method' instance.


> > why ever should the dirty, terrible hacks be allowed
>
> i would answer "for use with exec" but you already said that
> this works. You are probably going to say that "exec" is the
> next candidate for evil hacks. 

At least 'exec' (without explicit dictionaries) is an immediate signal to 
the compiler and to the human reader of the program -- it means "i, the 
author of this program, have not had time to think the problem through
carefully and come up with a decent solution, so I kludged up this fragile
and unmaintainable hack".  So the compiler knows it can turn off any
attempt at optimization, as this cannot possibly be a production program,
and the human reader knows where to look for any mysterious
misbehavior.

> That may be true but it certainly
> has its good uses and i appreciate that assigning to locals from
> within "exec" is allowed. 

Please give ONE example which, in your opinion, is best handled
with 'exec' without explicit dictionaries rather than in any other way.

For extra credit, please make sure that the advantage is worth a 
3200% slowdown:

>>> import time
>>> def woe(x):
...   return x+1
...
>>> def wie(x):
...   exec ''
...   return x+1
...
>>> def timeit(f):
...   a=time.clock()
...   for i in range(10000): f(i)
...   b=time.clock()
...   return b-a
...
>>> timeit(woe)
0.009999999999999995
>>> timeit(wie)
0.33000000000000002
>>>

The exec statement in wie tells the compiler that it has to turn
all optimization off.

I'm reasonably confident that, whatever use example you can come
up with, I can suggest better alternatives that do not use exec without
explicit dictionaries.

> When you want to have e.g. templated
> parts of your program (see c++) this really helps. Or would
> you go as far as to say that templating is not really needed
> with python?

Speaking as a C++ expert (Brainbench MVP for C++), ever since I started 
with Python I found all coding in Python interestingly close to coding C++ 
templates.  The key similarity is signature-based polymorphism, that's 
always on in Python (until and unless one type-tests, a practice that is 
best avoided over 9 times in 10) but needs templates in C++.

On the other hand, Python's polymorphism works at runtime, so in terms of 
_performance_ C++ can run rings around Python in the most favourable (to
C++) cases.  Indeed, one uses C++ or C to code highly optimized extensions
to Python exactly to exploit this performance gain.

Neither issue is affected in the least by exec without explicit 
dictionaries -- except that performance further plummets.

Note that C++'s templates are never usable to define local variables with
arbitrary names unknown at compile-time: template parameters are
typenames or constants.  In Python you can pass type objects as arguments
(and similarly for the values, of course) and, apart from performance, get
exactly the same effect (and then some) WITHOUT using exec.


Alex




From tejarex at yahoo.com  Sat Apr  6 10:38:31 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 06 Apr 2002 15:38:31 GMT
Subject: Destructors
References: <Y7Ar8.5367$Tf4.2431098@news2.news.adelphia.net>
Message-ID: <XDEr8.143333$VJ1.11980660@bin3.nnrp.aus1.giganews.com>

"Robert Oschler" <Oschler at earthlink.net> wrote in message
news:Y7Ar8.5367$Tf4.2431098 at news2.news.adelphia.net...
> My favorite object construct/destruct paradigm is the Boost
Library's
> shared_ptr template class for C++ objects.  When the reference count
of the
> shared pointer reaches zero, the destructor of the object it wraps
is
> called.  Delphi has "interfaces" which do the same, although they
behave
> differently than classes do when it comes to inheritance and other
subtle
> programming aspects so it can be a little awkward at times.
>
> Python seems to follow the Java finalize/gc() model where
destructors (or
> finalize() in Java) aren't called until the garbage collector frees
the
> object.

The language standard does not specify when destructors/finalizers are
called.  The standard CPython implementation uses reference counting
(with optional gc to clean up circular references).  The Jython
(Python in Java instead of C) implementation does use the Java model.

 >As in Java, the Python book I'm reading stresses that you have no
> idea when the destructor will be called, or if it will happen at
all.

If your code will or might be run on Jython, then this is true.

Terry J. Reedy





From lac at strakt.com  Sat Apr 27 17:27:51 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 27 Apr 2002 23:27:51 +0200
Subject: Stackless goes Limbo 
In-Reply-To: Message from "Donn Cave" <donn@drizzle.com> 
   of "Sat, 27 Apr 2002 17:34:40 -0000." <1019928879.371845@yasure> 
References: <mailman.1019911009.16159.python-list@python.org> <aaee44$p73$1@panix1.panix.com>  <1019928879.371845@yasure> 
Message-ID: <200204272127.g3RLRpkG019604@ratthing-b246.strakt.com>

> Quoth aahz at pythoncraft.com (Aahz):
> | In article <mailman.1019911009.16159.python-list at python.org>,
> | Laura Creighton  <lac at strakt.com> wrote:
> |>
> |> For those of you who can't live without a Limbo introduction written
> |> by a real C inventor:
> |>
> |> http://www.vitanuova.com/inferno/papers/limbo.html
> |
> | Ugh.  I've gotten so corrupted by Python I can't even read well-written
> | C code.
> 
> Tsk, that ain't C.  As I read on, I was surprised to find out that the
> sys->print notation doesn't imply sys is a pointer, rather that it's
> a module.  (Don't know why not sys.print, then.)  In some places it
> could remind you of Python - uses "self" more or less the same way,
> has garbage collection that apparently supports timely finalization,
> tuple unpacking assignment like x, y, z = t.  Also vaguely reminiscent
> of ML or something.
> 
> The important thing to me is that the system looks like it would work
> well for modular function libraries, thanks mostly to the garbage
> collection storage model, and who knows, maybe with less function call
> overhead.  The thing I miss here is exceptions - nothing like that?
> Maybe I skimmed it too fast.
> 
> I think I like the channels.  Not sure I like the Unicode strings.
> I wouldn't mind trying it out, if it ever escapes from Inferno.
> 
> 	Donn Cave, donn at drizzle.com

Limbo has exceptions.  I am not sure what you mean by 'escapes from
Inferno' -- but check out the free download page at vitanuovo
http://cgi.www.vitanuova.com/cgi-bin/www.vitanuova.com/idown.pl

Quoting from it: 
A Binary & Limited Source release of Inferno is now available for downloading 
from Vita Nuova which contains: 

        The Inferno operating system running in hosted mode for 
        Windows 95/98, Windows NT, Windows 2000, Solaris (sparc), 
        Linux (x86), FreeBSD and Plan 9 which includes 

<snip>

	Native and Inferno versions of the Limbo compiler for use inside 
	and outside of Inferno and with which all of the applications and 
	library modules can be compiled.
 
<snip>
-------------------
So what you want may already have happened.

Laura Creighton




From nightowl at uk2.net  Thu Apr  4 15:22:28 2002
From: nightowl at uk2.net (NightOwl)
Date: Thu, 4 Apr 2002 22:22:28 +0200
Subject: telnetlib question
Message-ID: <1017951810.138264@seven.kulnet.kuleuven.ac.be>

Hi,

I would like to be able to read the output of the last command in a telnet
session, without shutting it down.

What I would like to do is

- log in (can do that)
- cat the contents of a couple of files (while the content <> searchstring)
- if the search string is found, echo something.

I managed to log in and cat a file, I managed to do read_all(), however,
that outputs everything AND it seems like I needed to close the connection
first. The doc's are not clear at all at this point. What function do I
need?

Can somebody give me some example code, or the correct function to read the
output of the last command (the last cat in my case)?

I found read_eager(), read_lazy etc, however, the explanation is very
unclear to me!

Thx




From peter at engcorp.com  Sun Apr  7 23:06:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 23:06:07 -0400
Subject: Java and Python
References: <3C96DBB4.6040109@pobox.com> <3C97FB8A.5B9E9247@engcorp.com> <3C98108D.5010002@pobox.com> <3C981961.9F5E5383@engcorp.com> <eedf05f0.0204071808.451e23da@posting.google.com>
Message-ID: <3CB1091F.9A04FA08@engcorp.com>

Geometric Engineer wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> >
[...comments about hiring practices, and ignoring certifications...]
> > Not all companies hire in this way, of course, but all
> > the best ones do. :-)

>   How has your experience been anyway with tests of any kind to bring
> out actual programming ability? Do you rely mainly on "look and feel"
> from the interviews? 

Not sure what you mean by "look and feel", but if my inference 
is correct, then I believe that is closer to what I look at than
tests...  As I said, communication skills are by far the most 
important.  Look, as in appearance, has little or nothing to 
do with it, but I think you just mean do I rely on the "GUI"
rather than attempting to do a black box test to infer what
is inside.  I think that's what I'm saying.  Hard to tell with
all the metaphors in the way. :-)

> I've found generally the people who know about a
> subject will be able to explain simply and clearly how to work at it,
> and that the people who don't know about a subject will talk lots but
> give little real information, or worse, be actively confusing.

Exactly so.  In fact, every poor hiring decision I've ever made
was something I could have avoided had I paid more attention 
to my "spider sense" during the interview.  In most of those 
cases, a programming test would have done nothing to
dissuade me (certainly all these candidates would have passed
a basic C certification test, for example).  I plan to continue 
my process, but spend even more care thinking about the responses 
and putting faith in my gut feeling, rather than going against 
my intuition as I have once too often in the past, just because 
the candidate looked good on paper or came with glowing 
technical references.

-Peter


From borcis at geneva-link.ch  Mon Apr 15 08:41:40 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 15 Apr 2002 14:41:40 +0200
Subject: Does Python need an 'xor' operator?
References: <mailman.1018819588.582.python-list@python.org> <3CBABDBA.5010807@geneva-link.ch>
Message-ID: <3CBACA84.10104@geneva-link.ch>

> assert 304 in 340343, ParadiseMismatchError


oops, this should perhaps be :

     assert 304 in 340343, "P424DI5E M15M47CH E2202"


Boris

--
"Art is a trigram"



From tmckernan at dyrectmedia.com  Thu Apr  4 16:48:28 2002
From: tmckernan at dyrectmedia.com (Tim McKernan)
Date: 4 Apr 2002 13:48:28 -0800
Subject: parameter limit in zope xml-rpc
References: <43077ab.0204040801.1f96a36a@posting.google.com>
Message-ID: <43077ab.0204041348.5920a198@posting.google.com>

tmckernan at dyrectmedia.com (Tim McKernan) wrote in message news:<43077ab.0204040801.1f96a36a at posting.google.com>...
> I'm running Zope 2.4.1 and python 2.1, on an NT box.  I'm using Zope
> and python as an xml-rpc server to pass functions on to a group of C
> functions.  Based on the error message below it seems that there is
> either 1) a limit to how many parameters can be sent over the wire
> using xml-rpc, or 2) a limit to how many parameters zope will accept. 
> Can anyone explain why I am seeing this?
> 
> 
> 1) I create a method called OfferAppts with 9 parameters.  I test it
> with a Java xml-rpc client that sends in 9 parameters and it returns
> an appropriate response from the server.
> 
> 2) I create a method called OfferAppts with 10 parameters.  I test it
> and with a Java xml-rpc client that sends in 10 parameters and this is
> what the server returns:
> 
>   <h2>Zope Error</h2>
>   <p>Zope has encountered an error while publishing this resource.</p>
> 
>   <p>
>   <strong>Error Type: TypeError</strong><br>
>   <strong>Error Value: too many arguments</strong><br>
>   </p>
> 
> 
> Thanks,
> Tim


Please disregard this as I've found the error.  I kept forgetting to
delete and re-add the external method to Zope, causing the error
above.

-T


From dwelch91 at nospam.attbi.com  Sat Apr  6 12:21:02 2002
From: dwelch91 at nospam.attbi.com (djw)
Date: Sat, 06 Apr 2002 17:21:02 GMT
Subject: Modify Word Doc with Python
References: <mailman.1018081037.4211.python-list@python.org>
Message-ID: <3CAF2E80.4030403@nospam.attbi.com>

David McInnis wrote:
> I need to alter the document properties of Word DOCS that our users
> upload to our Web site.  Does anyone know how to do this?  (I am python
> installed on a Linux box).
> 
> David
> 
> 
> 

Might be overkill, but I've heard of this package being used to do 
conversions between Word and other formats...

http://www.outsideinserver.com/

D



From tim.one at comcast.net  Sat Apr 13 03:20:48 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 03:20:48 -0400
Subject: binding a reference to a variable
In-Reply-To: <yu99hemj7xa0.fsf@europa.research.att.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEMPONAA.tim.one@comcast.net>

[Andrew Koenig, on spelling SNOBOL4-ish unevaluated expressions]
> Yup.  But why not
>
>         arb = null | (len(1) + indirect(lambda: arg))
>
> ?

[as opposed to creating a class to model "SNOBOL4 variable"]

It depends on all you're trying to do.  At the time, I'm not sure Python
even had a lambda expression; if it did, it certainly didn't have lexical
scoping, so it would have been more like

         arb = null | (len(1) + indirect(lambda arb=arb: arb))

to break dependence on global names -- but that wouldn't have worked without
more tricks, as it requires a binding for arb first in order to establish
the default-argument value.

Now that it's possible to rely on lexical closures, I gotta say I look at

    indirect(lambda: arb)

and shudder a little, especially when plain "arb" *can* serve by playing
(previously sketched) tricks with the spelling of "=" instead.  I'd like the
pattern part to be as easy to read as possible; whatever gimmick attaches a
name to a pattern is dull boilerplate, and I can easily learn not to see it.
This is a dodge the Pythonic mind may be too quick to take, though, as about
once a year it suddenly (re)strikes me how silly the Python idiom

    while 1:
        line = f.readline()
        if not line:
            break

really is <0.9 wink>.  The other 364 days I don't notice the hair.

> ...
> Indeed.  But I want more Snobolish than Iconish patterns.

Me too, most of the time.  The paper introducing Icon argued strongly for
the benefits of integrating SNOBOL4's pattern-matching sublanguage with the
rest of the language, but in practice I stare at an old Icon program and
have a hard time finding "the pattern" part!  Matching and computation are
so seamlessly interleaved it can be hard to change what either part is doing
without screwing the other, or just to see what either part is doing on its
own.

BTW, from the start there was one beyond-SNOBOL4 gimmick I wanted that drove
me to think modeling variables as class instances would be best:  I
mentioned I had a var.ass(pattern, option) method to model conditional and
immediate assignment.  I also had another option:  accumulated immediate
assigment, meaning that all assignments to var during a match attempt were
appended to a list of matches.  So, e.g., given a pattern to match a run of
digits, then

    var.ass(that_pattern, ACCUM) + fail

is a pattern to find all runs of digits (in unanchored mode), and afterwards

    var.matches()

is a list of the matches.  There were some other gimmicks attached to var
instances too, like returning slice indices instead of actual strings (the
engine worked with slice indices under the covers anyway, and it was
sometimes more important in *uses* to know that a match spanned positions
subject[14:17] than that 'foo' was the substring that matched -- and giving
the user a way to know the indices is a lesson we rammed into Python's
regexp facility).





From logiplexsoftware at earthlink.net  Fri Apr 26 15:12:25 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 12:12:25 -0700
Subject: Sending mail to an SMTP server that requires POP authentication ?
In-Reply-To: <20020426180519.GU21205@teich.Garten.DigitalProjects.com>
References: <aabrbv$2vg$1@norfair.nerim.net>
	<20020426104354.6ed98204.logiplexsoftware@earthlink.net>
	<20020426180519.GU21205@teich.Garten.DigitalProjects.com>
Message-ID: <20020426121225.2e809961.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 20:05:19 +0200
Alexander Skwar wrote:

> ?Cliff Wells? sagte am 2002-04-26 um 10:43:54 -0700 :
> > Hm.  I think you'll have to log in with POP first, and then for some
predefined
> > period of time you'll be granted SMTP access.
> 
> It depends.  What you describe is often called SMTP-after-POP or
> POP-before-SMTP.  However, he asked for SMTP authentication (smtp auth)
> as described in RFC 2554.  Those two don't have much in common.

The subject said "SMTP server that requires POP authentication".  Is that part
of smtp-auth?  I looked at the RFC, but see no mention of such (and am not
familiar enough with mail protocols to know).  I was guessing that since he
mentioned POP authentication that he was referring to POP-before-SMTP (and
mistakenly using smtp-auth to try to authenticate).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 21:12:01 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 13 Apr 2002 01:12:01 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org>
Message-ID: <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com>

Gerhard H?ring <gh_pythonlist at gmx.de> wrote:
>Quoting http://www.fsf.org/licenses/license-list.html
>
>    """
>    The License of Python 2.0.1, 2.1.1, and newer versions.  This is a
>    free software license and is compatible with the GNU GPL.  Please
>    note, however, that intermediate versions of Python (1.6b1, through
>    2.0 and 2.1) are under a different license (see below).
>    [...]
>    """

Thank you for pointing this out.  I thought it was classified by FSF as
compatible with GPL.  I did not know that it was actually classified as free
software.

I had the impression that the FSF only admits a software as free if its
licence guarantees it to be free forever in all derived versions (copyleft).

Reading that page again, the criteria appears to be just that the current
version is free.  

Whether they have changed the definition in recent years, or whether I have
misunderstood this issue from the beginning, I don't know.

So now my mental picture is 

   (copyleft) is subset of (free) is subset of (open source)

It is also refreshing to see that

   (public domain) is subset of (free)


Huaiyu


From aahz at pythoncraft.com  Fri Apr 26 12:13:38 2002
From: aahz at pythoncraft.com (Aahz)
Date: 26 Apr 2002 12:13:38 -0400
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <slrnacgag6.98m.akuchlin@ute.mems-exchange.org> <aaaknh$3uk$1@panix2.panix.com> <m3r8l3roiv.fsf@mira.informatik.hu-berlin.de>
Message-ID: <aabubi$1le$1@panix3.panix.com>

In article <m3r8l3roiv.fsf at mira.informatik.hu-berlin.de>,
Martin v. Loewis <martin at v.loewis.de> wrote:
>aahz at pythoncraft.com (Aahz) writes:
>> Martin? :
>>>
>>>http://pyxml.sourceforge.net/topics/docs.html links to
>>>http://py-howto.sourceforge.net/xml-howto/xml-howto.html, but the text
>>>should really live on pyxml.sourceforge.net, because the master copy
>>>of the text lives in the PyXML CVS tree.
>> 
>> Ah.  And conversely, py-howto should link to pyxml.  Well, I don't have
>> access to either of those, so someone else needs to fix it.  Martin?
>
>I've now removed all traces of the XML howto from the py-howto
>project. It is not clear to my why the py-howto pages should have a
>link, so I have not added one (in any case, I could not locate the
>corresponding index.ht, anyway).

http://py-howto.sourceforge.net/ says

    Python HOWTOs are documents that cover a single, specific topic, and
    attempt to cover it fairly completely. Modelled on the Linux
    Documentation Project's HOWTO collection, this collection is an
    effort to foster documentation that's more detailed than the Python
    Library Reference.

Included on that page are documents about advocacy, editor
configuration, and Curses programming.  I think that a link to PyXML from
there is entirely appropriate; you're free to disagree, of course.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From pyth at devel.trillke.net  Tue Apr 16 23:11:51 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 17 Apr 2002 05:11:51 +0200
Subject: pre-PEP for optional 'pass'
In-Reply-To: <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu> <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <20020417031151.GA12470@devel.trillke>

> 
> >In my opinion an optional pass is a very good thing -- sometimes you 
> >want to be able to assert "this block intentionally left blank", but 
> >sometimes it makes more sense to just accept a truly empty block.
> >
> >Thanks for proposing this. I'm sure you'll get lots of horrified 
> >reactions
> 
> I don't see why: people who like pass can still use it.

i come in late to this thread, but could you list (or point
me to a list of) uses assuming an implicit "pass" except
that it saves 5 characters? Or is it a matter of taste (which
i think is perfectly valid).

	holger




From uwe at rocksport.de  Tue Apr  2 05:15:24 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 2 Apr 2002 10:15:24 GMT
Subject: wxPython  / wxSingleChoiceDialog ...
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de> <slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr> <a89vu0$8bm2c$1@hades.rz.uni-sb.de> <mailman.1017686284.32405.python-list@python.org> <a8ahju$8dmek$2@hades.rz.uni-sb.de> <mailman.1017696188.28944.python-list@python.org>
Message-ID: <a8c0bs$8eofb$3@hades.rz.uni-sb.de>

Cliff Wells <logiplexsoftware at earthlink.net> wrote:
| On 1 Apr 2002 20:57:34 GMT
| Uwe Schmitt wrote:
|> o.k  i used that and found a wxListBox as children. How do I now
|> manipulate the font of the entries of the ListBox ???

| Assuming you have a reference to the wxListBox at this point, you should be
| able to simply use listbox.SetFont().  Does that not work?

no that does not work, I tried it....

greetings, uwe



-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From ianb at colorstudy.com  Sun Apr  7 15:05:32 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 07 Apr 2002 14:05:32 -0500
Subject: Modify Word Doc with Python
In-Reply-To: <009601c1dd43$398e9f00$9865fea9@bigdaddy>
References: <009601c1dd43$398e9f00$9865fea9@bigdaddy>
Message-ID: <1018206332.2378.20.camel@localhost>

On Sat, 2002-04-06 at 02:15, David McInnis wrote:
> I need to alter the document properties of Word DOCS that our users
> upload to our Web site.  Does anyone know how to do this?  (I am python
> installed on a Linux box).

I've done some exporting from Word using wvWare (wvware.com).  It works
reasonably well if you don't mind loosing some layout.  However, it
doesn't seem appropriate for what you want.  wvSummary will read the
properties, but not write it.

The document properties you are talking about (NOT custom document
properties, just the normal ones) are part of the OLE file format --
i.e., they aren't specific to Word, and code that manipulates Excel
documents would work identically.  It's also well-documented compared to
the Word format that's contained in the OLE format.

Unfortunately, I've looked around a bit and haven't found such a thing
separated out as a single command-line tool or library -- I assume
support exists in some programs (Gnumeric, perhaps), but I haven't had
that much motivation to dig through such large amounts of source code. 
Such a program might yet exist, however, as "console" contains "ole",
and so you have to sort through a million console applications to fine
an OLE application :P

  Ian






From deltapigz at telocity.com  Thu Apr 18 22:59:27 2002
From: deltapigz at telocity.com (Adonis Vargas)
Date: Thu, 18 Apr 2002 22:59:27 -0400
Subject: importing modules question
Message-ID: <3cbf8c02$1_2@nopics.sjc>

how can i load modules from a different path ?

i.e.

a_python_script.py is in /tmp
but X user created Y.py script in a folder called /tmp/scripts
then i want load the script in /tmp/scripts

how do go about this?

any help would greatly be appreciated.

Adonis





From egrossde at yahoo.de  Thu Apr 18 05:17:20 2002
From: egrossde at yahoo.de (jb)
Date: Thu, 18 Apr 2002 11:17:20 +0200
Subject: Slightly OT Linux newbie question
Message-ID: <3cbe8b02_1@news2.newsgroups.com>

I installed Python 2.2.1 under Linux (Suse 8.0) and use KDE 3.0 (i.e. when 
it is not hanging; somtimes it is).

Whe I worked under MS-Win everything was very simple and comfortable: There 
was a menu where I could select the IDLE or the documentation or the Python 
prompt, whatever I wanted.

Can anybody give me a hint as to how to achieve the same thing under Linux?

-- 
Janos Blazi


-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From bmason at rand.org  Mon Apr 15 15:42:33 2002
From: bmason at rand.org (Bryce C. Mason)
Date: Mon, 15 Apr 2002 12:42:33 -0700
Subject: Thanks
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net>
Message-ID: <a9fadq$f7r$1@lumberjack.rand.org>

Thanks for all the suggestions.  It is not Mac OS X, so I will look into the
"drag and drop" application manager.

Bryce

"Bryce C. Mason" <lord_elendil at hotmail.com> wrote in message
news:bnEt8.130789$QB1.31528888 at typhoon3.we.ipsvc.net...
> Hi folks.
>
> I am doing my development on a PC (activeware's distribution) and the
final
> project will be running on a macintosh.  Umm...I am a little worried that
> there is not a command line interpreter in the Mac, at least one I have
not
> seen.  Help.  How do I port my project to the mac?
>
> Bryce Mason
>
>
>




From johnroth at ameritech.net  Sun Apr 14 07:56:06 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 14 Apr 2002 07:56:06 -0400
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122>
Message-ID: <ubirifj8o6fu48@news.supernews.com>

"Bengt Richter" <bokr at oz.net> wrote in message
news:a9aoqc$j94$0 at 216.39.172.122...
> On Sat, 13 Apr 2002 15:58:19 -0700, "Ken Peek"
<Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
>
> >"Martin v. Loewis" <martin at v.loewis.de> wrote in message
> >news:m3ofgn1avt.fsf at mira.informatik.hu-berlin.de...
> >| "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:
> >|
> >| > I think an 'xor' operator is needed.  The 'xor'
> >| > operator means 'logical-exclusive-or', and
> >| > returns a 1 or 0.
> >|
> >| In Python 2.3, you will be able to write
> >|
> >|    bool(a) ^ bool(b)
> >|
> >| Seems pretty clear to me.
> >|
> >| Regards,
> >| Martin
> >
> >You didn't read my post well enough-- we can already _DO_
> >that with Python v2.2.x:
> >
> >(not a) ^ (not b)
> >
> >IMHO, neither the above, nor the future v2.3.x idiom:
> >
> >bool(a) ^ bool(b)
> >
> Hm, ^ should probably be != to return bools instead of integers?
>
> >is as clear as:
> >
> >a xor b
> >
> >...
> >
> >Everyone_is_entitled_to_have_my_opinion'ly yours,
> >
> Well, to tie in with existing spelled-out 'and' and 'or' operations,
> it should probably do something along those lines (i.e., evaluate to
> one or other of the objects, not one of the boolean values used in the
logic),
> but I can't think of something sensible.

If I remember my logic and the description of what the
short circuit 'and' and 'or' operators actually do, it would
have to return the second operand in all cases. That's
logically incorrect - the second operand doesn't
necessarally have the correct truth value. Nor does
the first operand. In fact, there is one case (both operands
false) where neither has the correct truth value!

So it's not possible to return one of the operands, we have
to return True or False. However, now the operation
is not consistent with 'and' and 'or,' which imposes an
issue with teaching the language.

John Roth
>
> Regards,
> Bengt Richter




From mjais at web.de  Fri Apr 12 02:05:55 2002
From: mjais at web.de (Markus Jais)
Date: Fri, 12 Apr 2002 08:05:55 +0200
Subject: 4Suite for Python 2.2
References: <zxft8.37$Mh5.1049@nreader1.kpnqwest.net> <3CB64FFD.6D0D6AD7@engcorp.com>
Message-ID: <rNut8.1$xV3.411@news.ecrc.de>

Peter Hansen wrote:

> Paulo Pinto wrote:
>> 
>> Hello,
>> 
>> does anyone know when 4Suite for Python 2.2
>> would be available ?
>> 
>> I can only find support up to version 2.1 :(
> 
> Doesn't it run under 2.2?  (I haven't tried and don't know --
> this is a serious question.)
> 
> I thought 4Suite was simply a very large collection of XML
> processing tools in native Python (which I would have
> expected would still run under 2.2).  Or are there binary
> libraries which are keyed to python21.dll or something?
> 
> -Peter

hello
I have it running under python 2.2 and it seems to work fine.
have just used the 4xslt command line tool until now
but this works good.

markus


From phr-n2002a at nightsong.com  Fri Apr  5 08:56:12 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 05:56:12 -0800
Subject: Theoretical threshold of improving Python's performance
References: <3CADA276.D91B7544@kfunigraz.ac.at>
Message-ID: <7xn0wi5jwj.fsf@ruckus.brouhaha.com>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:
> I got a private email where I can read that Python can be never made as
> fast as Common Lisp due to Python's dynamic nature. May I ask  without
> any ulterior-motive: "Has the beformentioned proposition any value?".

I think there may be some truth to it even without declarations in CL.
For example, in CL you'd often use a defstruct to do something that
you'd use a class instance for in Python.  Accessing a member variable
is then like an array reference in CL, but takes a dictionary lookup
in Python.  

I think Python can evolve to be about where CL now is in terms of
compiled code speed, though.  And as it is, compiled CL would probably
beat Python (compiled by some hypothetical compiler), but not by 10x
or anything like that.


From pyth at devel.trillke.net  Tue Apr 30 15:05:53 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 21:05:53 +0200
Subject: Update locals() / resolved
In-Reply-To: <slrnactp07.pv.mlh@vier.idi.ntnu.no>
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org> <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no>
Message-ID: <20020430190553.GE16217@devel.trillke>

On Tue, Apr 30, 2002 at 06:35:51PM +0000, Magnus Lie Hetland wrote:
> In article <fjbz8.7570$M63.5673 at atlpnn01.usenetserver.com>, Steve Holden wrote:
> [snip]
> >> > Why use type({}) when you can use dict?
> >>
> >[ ... good reasons ...]
> >>
> >> 3. And 'dict' is something i often use in local definitions
> >>    when working with dictionaries.
> >>
> >Yeah, well we all have to cut that out and start using "adict" or "thisdict"
> >pretty sharpish. This one bites now.
> 
> How so? It only "bites" if you actually need access to the dict type
> in the same scope where you bind your variable... I've got plenty of
> code with stuff like
> 
>   file = open('foo.bar')
> 
> and it doesn't bother me one bit (except in the cases where 'file'
> should have been something more descriptive, I guess ;)
> 
> Thes names aren't reserved in any way...

right. i do it all the time for file,str,dict,string,list...
they are simply to short and descriptive to be left bound
to something i don't access directly. and using type('') or
friends is usually a very precise way of expressing that
you want the type object for that particular thing.

regards,

    holger




From jwilhelm at outsourcefinancial.com  Tue Apr 23 11:51:52 2002
From: jwilhelm at outsourcefinancial.com (Joseph Wilhelm)
Date: 23 Apr 2002 08:51:52 -0700
Subject: Developing and ircd with python
In-Reply-To: <aa3v5n$8jg$1@nsnmrro2-gest.nuria.telefonica-data.net>
References: <aa278j$8c8$1@nsnmrro2-gest.nuria.telefonica-data.net>
	<3CC4A648.3070100@jerf.org> <Xns91F8E339D9A04RASXnewsDFE1@130.133.1.4> 
	<aa3v5n$8jg$1@nsnmrro2-gest.nuria.telefonica-data.net>
Message-ID: <1019577112.16116.2.camel@jwilhelm.ofsloans.com>

I would suggest looking at www.irchelp.org
You can get links to RFC's, other IRCD's, and various other bits of
information.

On Tue, 2002-04-23 at 08:38, Raphael Ribeiro wrote:
> Do you know any place where i can find info about that?
> 
> "Philip Swartzleonard" <starx at pacbell.net> wrote in message
> news:Xns91F8E339D9A04RASXnewsDFE1 at 130.133.1.4...
> > Jeremy Bowers || Mon 22 Apr 2002 05:07:52p:
> >
> > > Raphael Ribeiro wrote:
> > >> Is it possible?
> > >
> > > Yes.
> > >
> >
> > There's something entirely perfect about this conversation...
> >
> > --
> > Philip Sw "Starweaver" [rasx] :: www.rubydragon.com
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list






From fredrik at pythonware.com  Tue Apr 30 10:01:42 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 14:01:42 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org>
Message-ID: <atxz8.34774$n4.7370773@newsc.telia.net>

Roman Neuhauser wrote:

> having spotty knowledge of python

that's pretty obvious from your posts.

maybe you should spend a little more time learning, and
a little less time telling us about your "opinions" on issues
you don't really understand?

further reading:

http://www.google.com/search?q=cache:opG_4HjTnbIC:www.theonion.com/onion3811/child-rearing_advice.html

</F>




From logiplexsoftware at earthlink.net  Wed Apr 10 12:27:17 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 09:27:17 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <000001c1e067$76889380$0334fea9@carol>
References: <000001c1e067$76889380$0334fea9@carol>
Message-ID: <20020410092717.0044363e.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 03:58:26 -0400
Arthur Siegel wrote:

> Cliff writes -
> 
> >This was a direct attack on Guido's integrity, which was >uncalled for. You
> *assumed* much about what he meant
> >(perhaps he didn't make himself clear, but
> >that's hardly grounds for labeling him a dictator) and when he >clarified
> his position he became a liar.
> 
> I didn't label him a dictator. As far I understand he has labeled himself
> such.  And the fact that he has is wonderful and honest - because he is and
> should say so, and there is general agreement that software of this nature
> cannot be designed by committee.

Well, I suppose this is true to some extent, much as Linus is arguably the
dicatator of Linux (note the recent VM change).  However, I read your post as
being a bit overly aggressive/resentful.

> And being dictator he has the unique ability to issue the LION PEP. The LIKE
> IT OR NOT PEP.
> 
> If Guido is concerned  - lets say -  about any code breakage issues he might
> have missed lets limit discussion to that.

That sounds good.  Some people were undoubtedly misled by his PEP proposal into
thinking that he was neutral on the issue when clearly he wasn't (although you
could take the fact that he submitted the PEP as an indication that he was for
it and was merely looking for overwhelming reasons not to implement it).

> And thanks, Cliff, for your thread.

Thank you for being gracious.

Regards,


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From anthony_barker at hotmail.com  Wed Apr 17 12:40:52 2002
From: anthony_barker at hotmail.com (Anthony_Barker)
Date: 17 Apr 2002 09:40:52 -0700
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org>
Message-ID: <899f842.0204170840.29e579ae@posting.google.com>

For me open source is simply the ability to "look under the hood" of
any program.
Almost every other type of human endevor allows this. 

David: What do you think of a open source license similar to what
books or other  types of media have?

1) the source is open
2) any derivative works must be open
3) any derivative works that are for educational or non-profit are
allowed
4) derivative works that are used for a profit or that substatially
harm the ability of the original work to produce a profit are not
allowed

This is from the Fair Use provision is in the Copyright Act,

more info:
http://www.benedict.com

Anthony

 -o)
 /\\
_\_V   Anthony Barker
Linux & Python addict
http://xminc.com/anthony/


mertz at gnosis.cx (David Mertz, Ph.D.) wrote in message news:<mailman.1018897470.30703.python-list at python.org>...
> |Do you think the set of "open source software" as defined by OSI and the
> |set of "free software" as defined by FSF are
> |1) the same set,
> |2) two different sets,
> |3) none of the above;
> 
> Well, both the OSI and the FSF are quite unambiguous in stating that
> these are two different sets.
> 
> And both groups are equally unambiguous in stating that there is a lot
> of overlap between the sets.  And moreover, the FSF is awfully clear in
> stating exactly what they don't like about the set of things that are
> Open Source but not Free Software (i.e. exactly the same the same thing
> they don't like about all non-Free software).
> 
> It is somewhat up-in-the-air whether the set of Open Source Software is
> a strict superset of Free Software.  At the least, Open Source includes
> generally more software, and -very nearly- all software that is Free
> Software is also Open Source Software.
> 
> There is nothing very difficult (and nothing -at all- ambiguous) in
> here.  However, it *IS* true--as with any area of human cognition--that
> there are people who fail to understand the categories.  Big deal.  Some
> people are wrong, and some things are so.
> 
> Yours, David...


From sanner at scripps.edu  Sat Apr 20 21:30:48 2002
From: sanner at scripps.edu (Michel F. SANNER)
Date: Sun, 21 Apr 2002 01:30:48 GMT
Subject: [ANN] PMV, ADT and ViPEr release
Message-ID: <3CC21676.F304B614@scripps.edu>

We are happy to announce the release 1.1alpha of PMV, ADT and ViPER.

PMV is a Python-based molecular visualization and manipulation program
ADT is a graphical front end to the automated docking program AutoDock
ViPEr is a visual programming environment for Python.

These 3 programs are all built on top a several python packages such as
DejaVu, a general purpose 3D geometry visualization package, MolKit a
component to deal with molecules etc ...

We apologize for the long delay in puting ViPEr online after its
presentation at the last Python conference. Since it shares many
components with previously released tools (PMV and ADT) we had to update

all the tools before releasing ViPEr. I hope the fun you'll have with
ViPEr will make up for the waiting :)

Download at:

http://www.scripps.edu/~sanner/python/packager.html

All of our programs and components are free for non commercial use. Only

a small
subset of the components require a license for commercial use.
You can either download the complete release which contains:
  - a python interpreter version 1.5.2
 - python packages developped in the MGL lab (platform dependent and
independent)
 - Third party python packages (platform dependent and independent)
 - Applications : PMV, ADT and ViPEr.

Or only the packages which contains:
 - python packages developped in the MGL lab (platform dependent and
independent)
 - Third party python packages (platform dependent and independent)
 - Applications : PMV, ADT and ViPEr.

The download "A la carte" is of course still available for the most
advanced users.

You can find more information on our web site :
http://www.scripps.edu/~sanner/python/index.html

Two mailing lists have been created pmv and pmvdev (see README.html file

once you downloaded and installed our packages).

For any question concerning :
        - the installation or PMV please contact Sophie COON
(sophiec at scripps.edu)
        - ADT please contact Ruth Huey (rhuey at scripps.edu)
        - ViPEr please contact Michel Sanner (sanner at scripps.edu)

Have Fun !

Ruth Huey, Sophie Coon, Daniel Stoffler and Michel Sanner






From spamtrash at softhome.net  Mon Apr 22 08:21:57 2002
From: spamtrash at softhome.net (Johannes Gijsbers)
Date: 22 Apr 2002 05:21:57 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org> <kulmbimohe.fsf@lasipalatsi.fi>
Message-ID: <e87ce949.0204220421.3579bd47@posting.google.com>

Erno Kuusela <erno-news at erno.iki.fi> wrote in message news:<kulmbimohe.fsf at lasipalatsi.fi>...
 
> it doesn't show up in many newsreaders, because your message lacks a
> "references" header. that's the usual mechanism that newsreaders use
> to do threading. your message does have a "in-reply-to" header, which
> your newreader could be using. it's for mail and not news, so not
> standard.

Maybe the "in-reply-to" header should be used by the news/mail client
if there's no "references" header. The "in-reply-to" header is already
used to form the "references" header when replying(if no "references"
header is present), so why shouldn't it be used to thread? I know
there could be a problem with "in-reply-to" headers containing
multiple parents, but I'm sure the implementation could choose a
solution for that(using only one parent, or not threading at all).

Johannes


From jwbaxter at spamcop.net  Mon Apr  8 16:21:30 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Mon, 08 Apr 2002 13:21:30 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1018205719.28293.python-list@python.org>
Message-ID: <jwbaxter-22DAC9.13212908042002@corp.supernews.com>

In article <mailman.1018205719.28293.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> In the case of PEP 285, Guido asked for community advice on 8 specific
> issues, one of which was "Should this PEP be accepted?".  He didn't find the
> contra arguments to that one persuasive.

I see two fallacies repeated regularly, in computer contexts and 
othewise; in Usenet and in other venues (eg Newspaper letters columns).

1.  "They" didn't do it the way I would have done it; therefore it was 
done incorrectly.
  I fell into that one myself in the integer division 
"discussion"...what I finally took away from that--after making far too 
much noise--was learning that indeed dynamic typing means some ideas 
which are perfectly reasonable in a static typed environment (eg int / 
int --> int) aren't reasonable in the presence of dynamic typing 
(continuing the example...because the output varies "too much" depending 
on the input).

2.  I made a wonderously-beautiful suggestion, and they didn't 
gratefully pick it up and use it, therefore I was ignored.
  Well, *possibly* I was ignored.  Possibly--more likely--my suggestion 
didn't stand up to close inspection (eg, in the political arena, 
specialized to the US, there is a Federal regulation which clearly 
prevents doing what I suggested).  And of course, possibly my suggestion 
was the best idea and "they" liked it but it didn't fit some larger 
agenda (in the political arena, again, perhaps it conflicted with 
rewarding friends and being re-elected).

With regard to PEP 285...I thought "oh well, a Boolean type will be 
added...Guido will do it such that it doesn't hurt me enough to worry 
about...it might even help sometime."  Given that, I read some of 
Guido's postings and samples of others, and concluded that indeed I 
didn't need to worry about it.


From boris.boutillier at lip6.fr  Wed Apr 24 11:09:53 2002
From: boris.boutillier at lip6.fr (Boris Boutillier)
Date: Wed, 24 Apr 2002 17:09:53 +0200
Subject: Subclass of int and initialisation with not integer values
Message-ID: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr>

Hi all,

I've got a little problem with subclass of int.

Here is a sample of what I want to do :

class toto(int):
	true  = 'O'
	false = 'N'
	def __init__(self,str):
		self.str = str
		if str=='O':
			int.__init__(self,1)
		elif str == 'N':
			int.__init__(self,0)
		else:
			raise ValueError, 'Not valid value %s'%(str)
	def __str__(self):
		return self.str
>>> x = toto('N')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ValueError: invalid literal for int(): N
>>> x = toto(1)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 11, in __init__
ValueError: Not valid value 1

In some way it seems that the int initialisation is called before __init__
I want toto to be a subclass of int because I need toto objects to be
used as index for lists.

Do you know some way to solve the problem ? Here true and false are 'O'
and 'N' for the example but in my example I use a metaclass to create
numerous classes of this kind, with different true letter and false
letter.

Boris
.


From patmiller at llnl.gov  Mon Apr  8 16:56:28 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Mon, 08 Apr 2002 13:56:28 -0700
Subject: Pyrex and weave.accelerate
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1
	 <00e101c1dd14$5620a930$6b01a8c0@ericlaptop> <3CB1FA63.6ADACEDE@llnl.gov> <034701c1df34$36d24d50$777ba8c0@ericlaptop>
Message-ID: <3CB203FC.2AA8F4D8@llnl.gov>

eric wrote:
> I'm also alluding to creating C/C++ versions of Python functions
> that have a standard C calling convention (instead of the self, args)
> stuff that has to be parsed on function calls.  If a pure C version

The other issue is the "callback" issue.  A common pattern in
scipy is to pass in a function as an argument

e.g.
>>> integrate(0.0,1.0, sin)

Now, there is a HUGE speed gain to be had if

def foo(x):
   return x*x - 2*x + 1

could be called without any of the Python overhead.  Right now,
the internal bit of the integration function has to look like

   PyObject* result = PyObject_CallFunction(f,"d",x);
   double fx = PyFloat_AsDouble(result);

This in a really tight loop....  We're looking at
a way that one could query an object for its
underlying C function and call that

   PyObject* functionAddress = PyObject_CallMethod(f,"asCFunction","");
   long functionAddressAsLong = PyInt_AsLong(functionAddress);
   double (*function)(double) = reinterpret_cast<....>(functionAddressAsLong);
   for(...) {
       double fx = function(x);
   }

Which is noticably faster.

-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

Knowing is not enough; we must apply. Willing is not enough; we must do.
-- Johann Wolfgang von Goethe (1749-1832)




From johnroth at ameritech.net  Sat Apr  6 20:36:10 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 20:36:10 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018040171.16545.python-list@python.org> <3CAE1D58.4050500@mxm.dk>
Message-ID: <uav89gbesd3d89@news.supernews.com>

"Max M" <maxm at mxm.dk> wrote in message news:3CAE1D58.4050500 at mxm.dk...
> brueckd at tbye.com wrote:
>
> >>because of the problems I have
> >>actually encountered with making code work with the many extant
versions of
> >>Python, I will no longer be doing any new projects in the language
and am
> >>not recommending Python to anyone.
>
> > So, what language did you move on to?
>
> 386 assembler probably. It's a little less portable, but the specs
> haven't changed in over a decade.

Oh? A whole slew of 64 bit instructions, including 64 bit
addressing? And it "hasn't changed?"

Of course, old programs will still run, and continue to
run indefinitely - by design.

John Roth
>
> regards Max M
>




From donn at u.washington.edu  Thu Apr 11 18:16:21 2002
From: donn at u.washington.edu (Donn Cave)
Date: 11 Apr 2002 22:16:21 GMT
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
References: <mailman.1018558717.13308.python-list@python.org>
Message-ID: <a951vl$2a6s$1@nntp6.u.washington.edu>

Quoth "Noah" <noah at noah.org>:
| Hmm... This sounds suspicious... If FTP won't write its buffered
| output then no application could see it -- that includes the shell!

When you use ftp in a normal session, the shell doesn't see its
output.  You're interacting with the ftp process through the
tty driver.  This phenomenon occurs when output is a pipe, not
when it's a tty.

| Python shouldn't be any different in what it sees.
| If I start FTP from the shell, then how does FTP know 
| when to flush the buffer so that you can see it in a shell?

C stdio I/O, which ftp presumably uses, buffers its output differently
depending on the device.  If it sees that it's writing to a tty, then
it "line buffers" output.  Otherwise, it "block buffers".  That's why
the solution is a device that acts like a pipe but looks like a tty.
(Of course all this is assuming that ftp actually does use stdio.  It
doesn't have to, and I haven't checked.)

	Donn Cave, donn at u.washington.edu


From SBrunning at trisystems.co.uk  Wed Apr 24 10:52:37 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 24 Apr 2002 15:52:37 +0100
Subject: variable substitution
Message-ID: <31575A892FF6D1118F5800600846864DCBD05A@intrepid>

> From:	claird at starbase.neosoft.com [SMTP:claird at starbase.neosoft.com]
> There ought to be a more canonical-but-readable Web reference
> for what David Beazley calls the string "modulo operator"; I'm
> open to suggestions.
 
Try <http://www.brunningonline.net/simon/python/PQR2.1.html#stringMethods>.
(Page down a couple of times.)

I'm (*very* slowly) working on a 2.2 update to this document. I'm planning
to add a fuller TOC at that time, to make navigation easier.

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




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




From db3l at fitlinxx.com  Mon Apr 15 22:52:59 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 15 Apr 2002 22:52:59 -0400
Subject: Win32 Open Source or free .NET integration?
References: <RSJu8.3093$MM2.1226076@news1.news.adelphia.net>
Message-ID: <uhemc4ano.fsf@ctwd0143.fitlinxx.com>

Michael Kelly <mkelly2002 at adelphia.net> writes:

> I haven't used .NET myself yet but another programmer
> using it raves about the back end integration among
> the languages.  Also I noticed ActiveState is doing
> a Visual Python that's supposed to integrate fully
> into .NET.

That actually seems reasonably unlikely at this point.  The "Visual
Python" project at ActiveState is to tie into Visual Studio for .NET
(e.g., the IDE) and not tying Python itself into the CLR (Common
Language Runtime) of .NET.

There was a Python.NET project (more of an investigation really) also
back a while at ActiveState - they may still have Mark Hammond's white
paper available on their site.  It achieved some interoperability, but
was a decent amoutn away from a full integration into .NET and some of
the results of the experimentation were not that promising in terms of
performance or good fit of the CLR versus some of Python's dynamic
model.

>             I know there are some open source project
> for .NET for Linux/unix but I was wondering if anyone
> has heard of a free Win32 implementation in the works?

Depends on what you mean by "Win32 implementation", or just what an
"open source project for .NET" is meant to mean?  The term .NET is
wroth with over-use and ambiguity (probably the way Microsoft likes
it).

If you mean the fundamentals - the .NET framework (the CLR along with
the command line C# compiler and associated documentation) is freely
available from Microsoft to download for Windows.  That's roughly
equivalent to what they are providing as the ECMA:

    http://msdn.microsoft.com/netframework/prodinfo/getdotnet.asp

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From gimbo at ftech.net  Thu Apr  4 04:48:17 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 4 Apr 2002 10:48:17 +0100
Subject: How best to write this try/except block?
In-Reply-To: <xIGq8.25783$VQ2.8306517@twister.socal.rr.com>
References: <a8fba8$ln8$1@panix2.panix.com> <xIGq8.25783$VQ2.8306517@twister.socal.rr.com>
Message-ID: <20020404094817.GA1779@andy.tynant.ftech.net>

On Wed, Apr 03, 2002 at 05:10:53PM +0000, Jeff Youel wrote:

> You can avoid the exception handler altogether with:
> 
>     value = d.get(key, None)
>     if value:
>         foo = stuff(value)
>     else:
>         foo = otherStuff(key)

Nasty.  What if foo[key] is 0?  Or (evil cackle) False...  ;-)

Use has_key() or try...except...else instead.

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From op73418 at mail.telepac.pt  Wed Apr  3 18:20:38 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Thu, 04 Apr 2002 00:20:38 +0100
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <Xns91E5D13ADC09cliechtigmxnet@62.2.16.82> <mailman.1017860168.5415.python-list@python.org>
Message-ID: <tf3nau4nocqut2sol33ja4iqpsr9rv90qg@4ax.com>

On Wed, 3 Apr 2002 12:55:26 -0600, "Mark McEahern"
<marklists at mceahern.com> wrote:

>[Chris Liechti]
>> i disagree. comparing True/False with other values should give the same 
>> result as when it were used with an 'if':
>> >>> True == 1, True == 6
>> True, True
>
>So 6 == True == 1, but 6 != 1?
>
>// mark
>

Hmmm I think Chris just busted up Transitivity of Equality. No longer an
equivalence relation...

Gon?alo Rodrigues


From logiplexsoftware at earthlink.net  Wed Apr 10 18:55:45 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 15:55:45 -0700
Subject: looping throuhg a list 2 items at a time
In-Reply-To: <a92f5j$250g@r02n01.cac.psu.edu>
References: <a92f5j$250g@r02n01.cac.psu.edu>
Message-ID: <20020410155545.1dc5f074.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 18:42:34 -0400
Rajarshi Guha wrote:

> Hi,
>   I have a list like this:  l = [1,'s', 2,'t', 3,'d']
> 
> I'd like to loop through the list but within the loop I want 
> the current element and the element just after it:
> I tried this
> 
> liter = iter(l)
> for i in l:
>   print i, liter.next()
> 
> But I get:
> 
> 1 1
> s s
> 2 2
> t t
> 3 3
> d d
> 
> whereas I want
> 
> 1 s
> 2 t
> 3 d
> 
> I can see I'm wrong - any ideas on how to fix it?

It would be better to organize your list into tuples before you start:

l = [(1,'s'), (2,'t'), (3,'d')]
for a, b in l:
    print a, b

However, if that isn't feasible, you could do:

l = [1,'s', 2,'t', 3,'d']
for i in range(0, len(l), 2):
    print l[i], l[i+1]

This is sort of fragile: if there aren't an even number of elements the loop
will fail. 

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From nodzie at net2000.com.au  Wed Apr 17 05:26:18 2002
From: nodzie at net2000.com.au (nodz)
Date: Wed, 17 Apr 2002 00:26:18 -0900
Subject: IDLE Errors
Message-ID: <3CBD3FBA.D8F616F5@net2000.com.au>

I am currently using SuSE distribution of Linux (v6.2).  It comes
bundled with python 1.5.2 which I believe should come with IDLE v0.1.
Howerver, no matter what I seem to do I cannot configure idle to run.

I thought that I would therefore download the latest version of Python
(v2.2.1).  I downloaded this to my home directory user at localhost.  I
then tar -zxvf Python-2.2.1.tgz.  This creates the directory
Python-2.2.1.  I then changed to su root and did the following
./configure, make clean, make install as instructed in the README file.
Once completed, I changed back from su to user and changed to the
Python-2.2.1/Tools/idle subdirectory.  I then type idle to run the
executable file.  I get the following error

Traceback (most recent call last):
  File "./idle", line 5, in ?
    from idlellib import IdleConf
ImportError: No module named idlelib

Can anyone advise me of what I'm doing wrong and how to correct the
problem?
If I have installed the program incorrectly, can anyone tell me how to
remove the package?

Chris

PS  I also have another problem.  If I run python from the root
directory the version that runs is the 1.5.2, whereas if I run it from
the user at localhost home directory the version of python that runs is
2.2.1.  How do I correct this problem, can I remove 1.5.2 and then set
up the computer so that only the 2.2.1 version is envoked?





From tim.one at comcast.net  Mon Apr 15 21:03:45 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 15 Apr 2002 21:03:45 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <a9frf6$d6u$0@216.39.172.122>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEIHOOAA.tim.one@comcast.net>

[Bengt Richter]
> ...
> I just think the signed hex is not terribly useful.

In practice, if I'm working with 300-bit bitsets represented as longs, I
never do hex(x), but hex(x & mask) instead, where mask = ~(-1L << 300) (a
string of 300 low-order bits).  Regardless of x's sign, that "shows the
bits" in a 1-means-on 0-means-off sense.  The key is that I *know* the width
I have in mind, much as hex(int) knows the width of a platform C long today,
so can pull the same trick for regular ints.





From goodger at users.sourceforge.net  Sun Apr 21 23:01:12 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Mon, 22 Apr 2002 03:01:12 GMT
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net> <3cc3765f$0$15478$afc38c87@news.optusnet.com.au>
Message-ID: <B8E8F587.21E4A%goodger@users.sourceforge.net>

Neville Franks wrote:
> I thought that Tim's original ndiff compared files, but the one in difflib
> compares lists of strings. Does this mean I somehow need to read the files
> into lists of strings and then call ndiff.

Yes.  It's easy to turn a file into a list of strings.  Read with
afile.readlines(), or split a multi-line string with astring.splitlines(1)
(the 1 keeps the line endings).

> A short example of how to use
> ndiff to compare two files would be greatly appreciated.

Use the source, Luke!  There are painfully detailed examples in there.

Here's a quickie (untested, but oughta work)::

    import difflib
    list1 = open('file1').readlines()
    list2 = open('file2').readlines()
    diff = difflib.ndiff(list1, list2)
    print ''.join(diff)

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
  - The Go Tools Project: http://gotools.sourceforge.net/



From ChrisE at lantech.com  Fri Apr 26 13:50:19 2002
From: ChrisE at lantech.com (Christopher Encapera)
Date: Fri, 26 Apr 2002 13:50:19 -0400
Subject: File sizes from MSWin...
Message-ID: <BC74F8F7E87DD511990C000103E2FEFA3000@XCHG>

When I try:

>>> os.path.getsize('x:\\exchsrvr\\mdbdata\\PRIV.EDB')
1475354624
>>> 

(1,475,354,624 bytes)
The actual size is 14,360,256,512 bytes.  When I try:

>>> import win32file
>>> a = win32file.GetFileSize('X:\\exchsrvr\\mdbdata\\PRIV.EDB')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: The object is not a PyHANDLE object

This is the MS Exchange private information store, and ours is large (13.3
G).  Any suggestions on how to get an accurate size?  I have ActiveState
ActivePython 2.1 installed, and X: is a mapped drive to a WinNT machine from
a Win2k workstation...
Thanks in advance!

Christopher Encapera
Lantech.Com
chrise at lantech.com
502-267-4200





From pj at engr.sgi.com  Mon Apr  8 20:28:32 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 9 Apr 2002 00:28:32 GMT
Subject: Newbie Problem Building 2.2 On SGI
References: <mailman.1018283656.17495.python-list@python.org>
Message-ID: <a8tcjg$pm65o$4@fido.engr.sgi.com>

Jeff wrote:
|> I have installed both 2.2 and 2.2.1.c2 on an SGI O2 with
|> IRIX6.5. I cannot import the modules array, math, and struct

>From my Irix 6.5.11m system "sam":

    (sam) $ python
    Python 2.2b2 (#2, Dec  7 2001, 13:35:03) [C] on irix6
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import array
    >>> import math
    >>> import struct
    >>> ^D

I see no errors or problems importing array, math and struct,
so I can't immediately reproduce your problem.

Your best bet would be to explain why you can't import them,
or to cut+paste what happens if you try.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From bokr at oz.net  Wed Apr 24 13:24:14 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Apr 2002 17:24:14 GMT
Subject: Accessing class-level variables
References: <mailman.1019621385.21532.python-list@python.org>
Message-ID: <aa6pnu$6vt$0@216.39.172.122>

On Wed, 24 Apr 2002 06:04:04 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?= <gh_pythonlist at gmx.de> wrote:

>I'm trying to access a class variable in a derived class. I sincerly
>hope there's a less kludgy way than what I'm currently using:
>
>class X:
>    a = 5
>
>class Y(X):
>    def p(self):
>        print self.__class__.a
>
>Y().p()
>
Unless a is shadowed, seems like this would work,
or am I missing something?

 >>> class X:
 ...     a = 5
 ...
 >>> class Y(X):
 ...     def p(self):
 ...         print self.a
 ...
 >>> Y().p()
 5
 >>> import sys; sys.version
 '2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)]'

Regards,
Bengt Richter


From bokr at oz.net  Mon Apr  8 12:03:32 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 16:03:32 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8pqal$ftm$1@panix1.panix.com> <7xpu1bh481.fsf@ruckus.brouhaha.com> <a8q0g1$qu5$1@panix1.panix.com> <a8qqf7$p4h$0@216.39.172.122> <Epes8.59265$pT1.1705378@news1.tin.it>
Message-ID: <a8sf0k$2a5$0@216.39.172.122>

On Mon, 08 Apr 2002 10:37:24 GMT, Alex Martelli <aleax at aleax.it> wrote:

>Bengt Richter wrote:
>
>> On 7 Apr 2002 13:43:29 -0400, aahz at pythoncraft.com (Aahz) wrote:
>> [...]
>>>
>>>Assignments in expressions will never happen.
>>>
>> I guess you mean they won't be spelled concisely ;-)
>> 
>>  >>> a='eigh'
>>  >>> a
>>  'eigh'
>>  >>> locals().__setitem__('a','bee') and 0 or a
>>  'bee'
>
>Now try that same, identical expression inside a 
>function, will you...?
>
Once is not never. Aahz just pushed my never-say-never button,
and I found one expression that worked ;-)

But why "identical" ;-) If I can find another spelling that works
everywhere (and it wouldn't surprise me if you had one in mind ;-)
wouldn't that be interesting?

How about globals() instead of locals()? (Not that that is exactly
the same. The implied goal was an expression that would have the
effect of the statement a=b' as a side effect when evaluated
in the same context, while having an expression value of b,
analogous to the way it works in C).

>You can't assume locals() is modifiable at all:
>attempts to modify it may fail silently and/or
>they may raise exceptions ("undefined behavior").
>
It did _seem_ to work (and did, at least well enough, if you're
in a hurry to refute "never" ;-), and you do have to look for
the documentation of restrictions:

 >>> help(locals)
 Help on built-in function locals:

 locals(...)
     locals() -> dictionary

     Return the dictionary containing the current scope's local variables.

Do you think that should be modified to say it doesn't necessarily work
like an ordinary dictionary, as is mentioned in a footnote of
Python22/Doc/ref/execframes.html? :

"""
The current implementations return the dictionary actually used to implement
the namespace, except for functions, where the optimizer may cause the local
namespace to be implemented differently, and locals() returns a read-only dictionary.
"""

Regards,
Bengt Richter


From peter at engcorp.com  Thu Apr 18 23:43:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 18 Apr 2002 23:43:34 -0400
Subject: importing modules question
References: <3cbf8c02$1_2@nopics.sjc>
Message-ID: <3CBF9266.131A6C74@engcorp.com>

Adonis Vargas wrote:
> 
> how can i load modules from a different path ?
> 
> i.e.
> 
> a_python_script.py is in /tmp
> but X user created Y.py script in a folder called /tmp/scripts
> then i want load the script in /tmp/scripts

Use the PYTHONPATH environment variable, or if you are
okay with hardcoding something into the script itself,
you can do

  import sys
  sys.path.append('/tmp/scripts')

before the other imports.

-Peter


From ajacques at morphics.com  Tue Apr 16 14:42:30 2002
From: ajacques at morphics.com (Alex Jacques)
Date: Tue, 16 Apr 2002 14:42:30 -0400
Subject: Problems building Python tkinter module on Solaris
Message-ID: <BHEEJDEEADLNNIEEIBMAAEILCDAA.ajacques@morphics.com>

I'm having a problem building the tkinter module on Solaris from source (all
other parts of Python build/run ok). It doesn't successfully link against
the tcl/tk libs, so from the "make" output I get:

gcc -shared build/temp.solaris-2.7-sun4u-2.2/_tkinter.o
build/temp.solaris-2.7-sun4u-2.2/tkappinit.o -L/usr/openwin/lib -L/usr/local
/lib -ltk8.0 -ltcl8.0 -lX11 -o build/lib.solaris-2.7-sun4u-2.2/_tkinter.so
Text relocation remains                         referenced
    against symbol                  offset      in file
nameTable                           0x254
/usr/local/lib/libtk8.0.a(tkColor.o)
nameTable                           0x34
/usr/local/lib/libtk8.0.a(tkColor.o)
...
string.6                            0x274
/usr/local/lib/libtk8.0.a(tkColor.o)
...
valueTable                          0x184
/usr/local/lib/libtk8.0.a(tkColor.o)
...
[and 30+k more lines like this]

This was against the tcl/tk 8.0 in my /usr/local.
I read the Python FAQ 3.18 and other mentions of trying another version of
tcl/tk. Since I cannot remove the tcl/tk in my /usr/local (insufficent
privilege, and other folks use it), I built tcl/tk 8.3.4 in my /usr/freeware
tree (same place where I build/install Python).

My Python build "procedure" is

./configure --prefix=/usr/freeware
make

No luck: the make file still grabs the tcl/tk stuff from /usr/local.

How do I tell Make to use the headers and libs from /usr/freeware instead of
(or prior to) the stuff from /usr/local?

My environment:
Python 2.2.1 (never tried any others on Solaris)
Solaris 2.7
gcc 2.95.3
ld Solaris/ELF (3.0)





From fperez528 at yahoo.com  Fri Apr 19 16:03:30 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 19 Apr 2002 14:03:30 -0600
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <a9pj29$pdm$1@nntp6.u.washington.edu>
Message-ID: <a9pt8a$3nh$1@peabody.colorado.edu>

Donn Cave wrote:

> Part of the problem is that when you write something like a "grep"
> in Python and in Perl, the Perl program will naturally be written
> like while ($line = <STDIN>) {...}, and the Python program will
> naturally be written like while 1: line = sys.stdin.readline() ...
> 

I don't know how far back (version wise) this works, but I'd rather use:

for line in sys.stdin.xreadlines():

which should be pretty good. And I guess if the name lookups aren't optimized 
away (which I don't know) you can always do

xread = sys.stdin.xreadlines
for line in xread():

I'm not saying that perl can't be faster than python at some things. Simply 
that I wouldn't be surprised if the factor of 10 loss was more due to someone 
writing python code using perlisms than to python's limitations. It's 
probably similar to the kind of problems you'd get in perl if you tried to 
coerce it into doing things the python ways. 

Each tool has its own optimal way of being used, and such way must be learned 
before claims such as the aboved carry any weight at all.

Cheers,

f.



From altis at semi-retired.com  Sat Apr 27 20:29:08 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 27 Apr 2002 17:29:08 -0700
Subject: wxPython Mac on OS X preview
References: <SBgy8.46$EE5.53683@news.uswest.net>
Message-ID: <6mHy8.1442$6U3.103528@news.uswest.net>

I've added wiki pages for wxPython Mac and MachoPython.

http://wiki.wxpython.org/index.cgi/wxPythonMac

http://wiki.wxpython.org/index.cgi/MachoPython

These pages are "in the rough" so comments, corrections, additions, etc. are
welcome.

ka

"Kevin Altis" <altis at semi-retired.com> wrote in message
news:SBgy8.46$EE5.53683 at news.uswest.net...
> There are disk images (binary installers) of MachoPython 2.2.1 and the
> wxMacPython 2.3.3 preview build available for OS X at:
>
http://sourceforge.net/project/showfiles.php?group_id=10718&release_id=56880
>
> This software should be considered beta quality. It is not the final
> wxPython Mac 2.3.3 release. MachoPython 2.2.1 is required to use wxPython
> Mac.
>
> I made an announcement on my blog, but forgot to post to c.l.py.
> http://radio.weblogs.com/0102677/2002/04/24.html#a49
>
> The most recent information is available on the wxPython-mac mailing list
> http://lists.wxwindows.org/mailman/listinfo/wxpython-mac
>
> If you want to help with the project or just join in the fun please
> subscribe.
>
> Bug fixes and new binaries should be appearing frequently as a lot of
> testing and work is being done on wxMac (the C++ wxWindows sources) and
> wxPython Mac right now. Frank Vercruesse has provided some instructions
for
> building from cvs.
>
> http://lists.wxwindows.org/pipermail/wxpython-mac/2002-April/000227.html
>
> Finally, as might be expected, I'm pretty excited about this because it
> means PythonCard works on OS X. More info on that topic is available on
the
> pythoncard-users mailing list
> http://lists.sourceforge.net/lists/listinfo/pythoncard-users
>
> and the wiki page
> http://wiki.wxpython.org/index.cgi/PythonCardOnMac
>
> ka
> ---
> Kevin Altis
> altis at semi-retired.com
>
>




From stephen at xemacs.org  Mon Apr 15 23:37:43 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 16 Apr 2002 12:37:43 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp> <slrnabktvn.20b.hzhu@mars.localdomain> <87wuv9xdr8.fsf@tleepslib.sk.tsukuba.ac.jp> <0cEu8.47362$iw.4776263@bin4.nnrp.aus1.giganews.com>
Message-ID: <87sn5wwby0.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Terry" == Terry Reedy <tejarex at yahoo.com> writes:

    Terry> "Stephen J. Turnbull" <stephen at xemacs.org> wrote in message
    Terry> news:87wuv9xdr8.fsf at tleepslib.sk.tsukuba.ac.jp...

    >> I do have issues with some GPL advocates, because their
    >> arguments amount to claiming that nobody should have property
    >> rights in his own code.  (rms says this explicitly, but very
    >> few are so courageous as he.)

    Terry> Interersting.  If so, then RMS is 'guilty' of an almost
    Terry> humorous contradiction.

If you like.  But know that rms is well aware of the paradox, and
considers that the joke is on the society which attempts to implement
an immoral legal system.

rms manages to create an arena within society which implements a close
approximation to his moral scheme despite the fact that most people do
not agree with him, and within the framework of society's explicit
conventions.  The use of property rights to allow a body of software
to be simultaneously publically available to all, yet unappropriatable
by those who would profit by selectively withholding it, is a glorious
legal hack.

I disagree with his moral principle, but applaud his (and his
lawyers') creativity.  And the GPL is a useful tool for free software
and open source software advocates alike.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From kevin at cazabon.com  Tue Apr 23 17:04:21 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 23 Apr 2002 14:04:21 -0700
Subject: non-blocking sockets
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>
Message-ID: <5a4226f0.0204231304.32e9676b@posting.google.com>

Why not use a separate thread to monitor the socket, and post an event
when data is read from the socket and available to the rest of the
program?  This way, the socket thread can block all it wants, as its
only job is to monitor for incoming data, and the rest of your program
can happily do other stuff in the meantime.

Kevin Cazabon.



so very tired <g0riaman at cdf.toronto.edu> wrote in message news:<Pine.GSO.4.30.0204222138510.19452-100000 at eddie.cdf>...
> When trying to read from a socket, I don't want it hanging, so I set it to
> non-blocking by calling
> 
> setblocking(0)
> 
> but then when I try to read from it and there is no data in the socket, I
> get an error. I tried this in both windows and linux and in both I get an
> error and the program stops. It says there's an exception but it doesn't
> say the name of the exception so I can't even try to catch.
> Does anyone know how I can check if there's data coming from a socket
> connection without hanging or having the program barf on me?
> Thanks.
> mRiaz


From mgilfix at eecs.tufts.edu  Sat Apr 27 16:43:35 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 27 Apr 2002 16:43:35 -0400
Subject: inline assignments in conditionals
In-Reply-To: <200204272037.g3RKb0kG019255@ratthing-b246.strakt.com>; from lac@strakt.com on Sat, Apr 27, 2002 at 10:37:00PM +0200
References: <2143388891-BeMail@sugoi> <jaf@lcsaudio.com> <200204272037.g3RKb0kG019255@ratthing-b246.strakt.com>
Message-ID: <20020427164335.O1588@eecs.tufts.edu>

  I think this example is backwards. Today is the variable.  I'd
usually say: "If today is Sunday, I'm going to go shopping". So much
for that :)

            -- Mike

On Sat, Apr 27 @ 22:37, Laura Creighton wrote:
> Jeremy:
> [snip]
> > Writing things that way is really unnatural, at least for English
> > speakers.  It makes the intent of the code harder to understand (at
> > least for me).  For example, if you were talking to a friend, you
> > wouldn't say:
> > 
> >    "If today is Sunday, I'm going to go shopping"
> > 
> > but rather:
> > 
> >    "If Sunday is today, I'm going to go shopping"
> 
> Interesting.  I would never use the second construct, and I use the
> first one all the time.  I wonder if this is a difference between
> people whose first language is English, and some for whom it was not.
> I don't think that anybody has had any trouble understanding me when I
> say 'If today is Saturday, then we are going to the concert house,
> otherwise we are going to hack.'  But then I have been writing
> if 0 == x for a long time as well.  (It was great -- Borodin Quartet
> playing Sjostakovitj and Beethoven, by the way.)
> 
> Laura Creighton
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (lac)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From logiplexsoftware at earthlink.net  Mon Apr  8 16:42:07 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 8 Apr 2002 13:42:07 -0700
Subject: What has PEP 285 done to us?
Message-ID: <20020408134207.1af35256.logiplexsoftware@earthlink.net>

I have to say that the most damaging thing introduced by PEP 285 has little to
do with the technical merits of the discussion.  The level of respect accorded
others in this community has taken the most damage.  It seems that almost every
discussion on this topic has quickly gone down in flames.

If you are trying to win an argument, and not simply argue for argument's sake,
it usually goes against your best interest to start that argument by insulting
the other party.  I know when I read a response to any of my posts that start
out with "That's stupid!" or "You call that code?? Why I oughta..." or some
such nonsense, I have a difficult time properly assessing the respondent's
arguments.  In fact, because I'm human (as are most people on this list,
Timbots notwithstanding), it would be very difficult for me to reconsider my
position at this point (no matter how wrong I might be) as it's become a
personal issue rather than a technical one.

We are all familiar with the tendency of Usenet to perpetuate
misunderstandings.  Many of the flames I have seen have been the result of the
respondent not fully understanding the poster's meaning and getting
self-righteous about a non-issue (often re-stating the original poster's
argument in different terms yet somehow making it seem like a disagreement).

It is possible to argue heatedly without resorting to personal insults and
insinuations.  Being unable to do so is a good indication that your argument
can't stand on its own.  Such techniques are grounds for a default in a formal
debate.

I'd also like to point out that insulting someone and apologizing for it in the
same post is worthless. If you sincerely meant that apology you could as easily
have edited the post as add some empty words at the end (the "sorry if this
sounds like a flame" variety).

This list is known for being particularly helpful and friendly.  Let's keep it
that way.  If you want to argue technical issues and programming techniques,
this is a good place.  If you want to fight and bicker, take it to the
playground (or maybe Yahoo!'s Fight Room - the people there won't know what the
hell a boolean is, but they'll be more than willing to exchange insults with
you over it, because you don't know either, you moron ;)

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From phd at phd.pp.ru  Tue Apr 16 03:54:32 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 16 Apr 2002 11:54:32 +0400
Subject: py script web page problem ??
In-Reply-To: <61fe942.0204152214.3cbe6c0a@posting.google.com>; from aries@bitstream.net on Mon, Apr 15, 2002 at 11:14:08PM -0700
References: <61fe942.0204152214.3cbe6c0a@posting.google.com>
Message-ID: <20020416115432.D11374@phd.pp.ru>

On Mon, Apr 15, 2002 at 11:14:08PM -0700, linux_on_wood wrote:
> <FORM action="friends1.py" method=get>
> 
> my python script code is:    friends1.py  as chmod 755 on linux box
> 
> #! /usr/bin/env python
> 
> This script code looks very correct to me but it
> gets called --friends1.py DOES NOT render the
> html, it just shows the source code as text.

   A script (any script) should not "rebder" HTML - it will just pass the
HTML back to a browser. Do I understand your question right that instead of
HTML you get the source code of the script?

> ANY IDEAS, HELP , WHY  has a problem ???????????

   It seems your web-server was not configured to recognize .py extension
as a CGI extension. Reconfigure your server.

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




From brian at sweetapp.com  Mon Apr 15 13:09:56 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 15 Apr 2002 10:09:56 -0700
Subject: newbie-question: extending python with c++
In-Reply-To: <3cbac8f9$0$12315$9b622d9e@news.freenet.de>
Message-ID: <006e01c1e4a0$5e3079a0$445d4540@Dell2>

Merman wrote: 
> I have a c++ -class; I can compile this class successful as
standalone-app.
> No problem.
> Now I want to work with swig.
> 
> There are three files: Tree.cpp, Tree.h, Tree.i
> 
> first step:
> swig -python -c++ -o Tree_wrap.cpp Tree.i (output is Tree_wrap.cpp -
it's
> ok)
> 
> second step:
> g++ -c Tree.cpp Tree_wrap.cpp -IC:\python22\include (also ok - output
are
> Tree.o and Tree_wrap.o)
> 
> third step - *build*:
> g++ Tree.o Tree_wrap.o -shared -o Tree.dll (this line throws lots of
> failures - whats wrong?)

Where need to link with Pythonxx.lib. But instead of this manual crap,
you should just write a simple distutils setup script.

Cheers,
Brian





From aleax at aleax.it  Mon Apr 15 17:20:26 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 21:20:26 GMT
Subject: String interpolation question
References: <a9fcmo$npi$1@peabody.colorado.edu>
Message-ID: <uuHu8.23410$vF6.768198@news2.tin.it>

Fernando P?rez wrote:

> y=[5,6]
> i=1
> 
> I want to print y[i]. The following fails:

If you're a string-interpolation junky:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66018

and particularly Skip Montanaro's variant (acknowledgedly inspired
by a post by Steve Majewski) may be to your taste.  With that:

        print "%$(y[i])s" % Eval()

works basically as you wish (I could quibble with a few details, but
it _would_ be quibbling).


Personally, I don't use string interpolation so widely.  If and when I
don't want by-position correspondence, I use explicit tags:

        print "%(nice_item)s" % {'nice_item': y[i]}

or actually (a question of tastes)

        print "%(nice_item)s" % dodict(nice_item=y[i])

with one of my favourite Python oneliners somewhere above:

def dodict(**kwargs): return kwargs


But again I'm quibbling.  The point is that I want to be able to i18n's
my format-strings, e.g via gettext.  That's what the explicit tags are
most useful for, IMHO -- allow natural language dependent reordering.

But I most definitely do NOT want general Python expressions in there,
which a translator might well accidentally botch.  I do want tags that
are easily recognizable as such, it's easy to explain that anything of
the form %(whateverishere)s stands for the runtime-determined
string "named" 'whateverishere' and should be used as part of the
whole format-string's translation in whatever place is appropriate in
the target natural-language.  By using a purpose-built dictionary on
the RHS of %, I control what values can be formatted at that point --
no accidents (if an unavailable tag is asked for I get an exception that
I can catch or propagate -- exactly as it should be).

For a particularly involved, perhaps multi-string message, I could also
build the dictionary once and use it on several % ops, I guess (never
happened yet, but maybe some day), letting each translated format
string get whatever subset of the available tagged values it needs.


For me, this does all I need.  In simple throw-away code I don't need
anything more than positional-correspondence of values.  In code of
some worth and weight, I want to keep the possibility of doing i18n when
needed without too much effort, and tagged values and the right
dict on the RHS of % provides very well for that. The idea of passing
locals() on the RHS is, well, something I think is nice to show a newbie
to wow him or her, but not something I'd ever use in my code...:-).


Alex



From borcis at geneva-link.ch  Tue Apr  9 05:38:27 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 09 Apr 2002 11:38:27 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1018124969.17602.python-list@python.org>
Message-ID: <3CB2B693.80903@geneva-link.ch>

Ken Seehof wrote:

replying to Jacob Hall?n :

>>...The hard thing to
>>teach is how to interpret a "Type mismatch error" and why unexpected
>>automatic conversions happen. Automatic conversions are a pitfall even for
>>very seasoned programmers.


Who can deny this in good faith ? Reminds me that I reported
SNOBOL reporting "String Arithmetic Not Implemented" instead of
"Integer Overflow". Very confusing, but almost OO :)


> 
> Properly used bools will tend not to interact with other numeric types in
> practice.  Anyway, a TypeError won't occur.


As I get it, the PEP's bools are not intended to change anything to 
homeland python, but simply to endow it with proper diplomatic 
credentials when dispatching secretaries of State to the middle east.

TypeErrors might still occur if the resident programmer then insists on 
dispatching the wrong types for the mission, I guess.


>>the complexity is going to drive Pythons core supporters - the biologists,
>>physicists, astronomers, sociologists and other scientists - to look for
>>other alternatives, that require less work to understand.


I guess industrial age scientists will not find it so hard to
understand a feature destined to better translate the quirks
of "remotely controlled devices" to a transparent exception.

Boris Borcic
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From jeff at ccvcorp.com  Fri Apr  5 17:55:37 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 5 Apr 2002 14:55:37 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <mailman.1017757086.9920.python-list@python.org> <KBor8.8537$q63.3756@atlpnn01.usenetserver.com> <3CAE1CC8.7D18F4A9@alcyone.com>
Message-ID: <MPG.1717cb31e97aaec798969c@news.nwlink.com>

In article <3CAE1CC8.7D18F4A9 at alcyone.com>, max at alcyone.com 
says...
> Steve Holden wrote:
> 
> > *My* point, however badly I originally made it, was that writing
> > 
> >     if ret == None:
> > 
> > is *sometimes* like writing
> > 
> >     if (a == 3) == True:
> > 
> > assuming this new wart appears in Python.
> 
> Comparing to None (preferably via is rather than ==, since == can be
> overridden to do strange things) is perfectly legitimate when an API
> declares that a function or method returns "an object or None."  After
> all, the __nonzero__ method can be overridden to make perfectly valid
> objects evaluate to false in a truth test.

I would suspect that anyone defining such an object would be 
rather surprised if you refused to test it with 'if not object:' 
-- after all, that's the whole *point* of defining __nonzero__().  
There's no reason for them to define it, if not to explicitly 
allow such a test to succeed/fail in the appropriate 
circumstance, because that's what the method *does*.  And in any 
case in which a function is defined as returning "an object or 
None", if the function returns an object that claims to be 
false/zero, then nobody should be surprised if the caller of that 
function treats the object as if it were None.

It *could* be desirable to compare to None in some cases, where a 
function is specifically declared to return something that might 
be false, or None.  For instance, string.find() could have 
reasonably been defined to return None if no match was found 
(where currently it returns -1).  In that case, you would need to 
compare to None in order to distinguish from the legitimate 
return of 0.  However, trying to protect yourself from objects 
that define __nonzero__() is counterproductive -- if they define 
that, they're intending to be used as true/false values.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From tdelaney at avaya.com  Mon Apr 29 03:32:50 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 17:32:50 +1000
Subject: Comment on draft PEP for deprecating six builtins
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A5@natasha.auslabs.avaya.com>

> From: Raymond Hettinger [mailto:python at rcn.com]
> 
> "Martin v. L?wis" <loewis at informatik.hu-berlin.de> wrote in message
> news:j4elgz9e9a.fsf at informatik.hu-berlin.de...
> > "Raymond Hettinger" <python at rcn.com> writes:
> >
> > >     The module additions would go into Python 2.3.  All six
> > > functions would be left intact and raise deprecation warnings in
> > > Python 2.3 and 2.4.  The six functions would be removed in Python
> > > 2.5 (having allowed two versions and a full year for code 
> updates).
> >
> > I believe the phasing out of the old names should be even slower. In
> > 2.3, introduce the new names; leave the old names intact 
> unchanged. In
> > 2.4 and 2.5, raise DeprecationWarnings; in 2.6, remove them.

Regardless of anything else, Guido has guaranteed no (intentional) code
breakage by default until Python 3.0. So this would mean that, if this PEP
were accepted, the names could not be removed until Python 3.0.

BTW, I'm someone who uses map() and filter() extensively - I find it very
natural to think in tems of those functions. I don't tend to use reduce()
much though. I don't agree that these are natural first candidates for
removal.

I also personally don't think lamdas should be removed. However, if a lambda
exceeds a (BDFL-pronounced) number of opcodes it should be a syntax error
<0.5 wink>. This should reduce most lambdas to one or two expressions.

Tim Delaney




From tdelaney at avaya.com  Mon Apr 29 03:58:19 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 17:58:19 +1000
Subject: Returning variable results from a function
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A7@natasha.auslabs.avaya.com>

> From: Matthew [mailto:matthewm at zebrasoft.co.nz]
> 
> Given something like:-
> 
> def fred(size):
>     return size, 1,2,3
> 
> size,a,b,c=fred(3)
> 
> how does one create a tuple to hold the results like:-
> 
> size, (<a tuple holding the number of results?>) = fred(<no of
> results?>)

What you want to do is return the 1, 2, 3 as a tuple ...

def spam (size):
    return size, (1, 2, 3,)

size, eggs = spam(3)

Here you are returning a tuple, which has size and another tuple as its
elements.

However, if you mean "how do I take 'size' and produce a tuple of that
length, already populated, and return it"

def spam (size):
    return size, tuple(range(1, size + 1))

size, eggs = spam(3)

Tim Delaney




From mikael.lexen at stop.spam.volvo.com  Thu Apr 18 07:37:45 2002
From: mikael.lexen at stop.spam.volvo.com (Mikael =?iso-8859-1?Q?Lex=E9n?=)
Date: Thu, 18 Apr 2002 13:37:45 +0200
Subject: chopping csv-files (newbie)
References: <3CBE768C.844C90FF@itea.ntnu.no> <m3it6p5r17.fsf@ferret.object-craft.com.au>
Message-ID: <3CBEB009.8B4B7A56@stop.spam.volvo.com>

Hi

I get the following when I try to import your module

PythonWin 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) - see
'Help/About PythonWin' for further copyright information.
>>> import csv
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ImportError: DLL load failed with error code 193

What's wrong? I'am on NT4.

/Mikael

Dave Cole wrote:

> > I have a csv-file.. delimited by ;
> > example-line from the csv-file
> > "90383";"ooooxWOW   90383;";"Some discription";
> > "99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";
> >
> > In this file there is 3 fields extracted from a database.
> > The second field, can vary som 1 to 4 fields separated by ; but not with
> > "
> >
> > What I need to do is to build an array after opening a given file
> > with the information from the second and third fields.
> >
> > E.g:
> >
> > 90383 Some discription
> > 99875 New line-descriptor, Some discription
> >
> > You see where I'm going at?
> >
> > the second record in this file has a reference to the first record in
> > its second field.
>
> You could use my CSV parser:
>
> >>> import csv
> >>> p = csv.parser()
> >>> p.field_sep = ';'
> >>> p.parse('"90383";"ooooxWOW   90383;";"Some discription";')
> ['90383', 'ooooxWOW   90383;', 'Some discription', '']
> >>> p.parse('"99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";')
> ['99875', 'ooooxSomede90383;0000xNewli99875;', 'New line-descriptor', '']
>
> The module is available at:
>
>         http://www.object-craft.com.au/projects/csv/
>
> - Dave
>
> --
> http://www.object-craft.com.au

--
/Mikael

_________________________________________________
Mikael Lex?n
Volvo Information Technology AB
Dept 9514, HD2N
SE-405 08 G?teborg, Sweden

Telephone: +46 31 3215757
E-mail: mikael.lexen at volvo.com




From sholden at holdenweb.com  Fri Apr  5 16:50:17 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 5 Apr 2002 16:50:17 -0500
Subject: The Icon language [was: Re: PEP 285: Adding a bool type]
References: <mailman.1017466974.6751.python-list@python.org> <3CA8E37A.E89CA002@noaa.gov> <7x4rivj5ya.fsf@ruckus.brouhaha.com> <3CA9E7CF.52D6CD37@noaa.gov> <QPoq8.44596$pT1.1256392@news1.tin.it>
Message-ID: <P%or8.9190$q63.7986@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:QPoq8.44596$pT1.1256392 at news1.tin.it...
> Chris Barker wrote:
>         ...
> > do think that that is how it should be in the ideal language. The
> > concept of True or False is distinct from the concept of a number, and
> > I'd like to see it stay that way in the language.
>
> Have a look at Icon (or Snobol, but that was goto-based... Icon's
> much nicer).  It CENTERS on this concept.
>
[...]
>
> If you want "booleans" given the respect you seem to think they
> deserve, you may actually be happier with Icon than with Python.
> Me, I think that Python's decision regarding "concrete nothings",
> although less pure, was more pragmatically useful than Icon's
> purity in distinguishing "something or nothing" from any actual
> kind of "somethings", and contributed to Python's greater success.
> But - study Icon, try it out on some concrete programming task,
> and make up your own mind.
>
Or take a look at this old program. Note especially the "/owid := 72" idiom,
used to default a variable, and the "every" construct, which forces a
context in which all values oif an expression are used. Icon was very
advanced for its time, but I don't think it really took the object-oriented
paradigm on board. Certainly the standard OO system was klugey, and never
became part of the language.

regards
 Steve

#
#    classify.icn:    read a mail file and produce a subject-sorted
#            list of mail items and replies, or an index to
#            the mail file.
#
#    (c) Copyright Steve Holden 1991
#
#    Identification:    @(#)classify.icn    1.2
#    SCCS source:    /export/home/steve/dev/tools/src/SCCS/s.classify.icn
#    Last change:    10:37:32    92/09/14
#    Retrieved on:    20:46:13    94/02/15
#
#    Bugs:        Sort is case-sensitive.
#            Craps out on overlength index lines.
#            (Possibly elsewhere) subjects are not always
#                well-trimmed.
#
#
link mlib                # mail i/o library

procedure main(arg)

while arg[1][1] == "-" do        # handle arguments
    case (o :=pop(arg))[2] of {
    "i":        indexrun := ""
    default:    stop("Unrecognised classify option: ",o)
    }

if *arg = 0 then arg := [ "-" ]
if *arg > 1 then stop("Bug: can only handle one file at a time. Sorry.")

ttls := table([])
refs := table([])

msg := []
while filename := pop(arg) do {
    ifn := OpenMail(filename) | stop("Cannot open file ",filename)
    while put(msg,ReadMail(ifn))
}

every m := 1 to *msg do {
    trim( \(msg[m].headers["Subject"][1]) | "** Untitled Mail" ) ?
        if (="Re:" & tab(upto(~':\t ')))
        then refs[ms := tab(0)] |||:= [m]
        else ttls[ms := tab(0)] |||:= [m]
}

tlst := sort(ttls)
rlst := sort(refs)


while (*tlst > 0) | (*rlst > 0) do {        # for each subject...
#    write("## ",*tlst,"+",*rlst)
    if (rlst[1][1] << tlst[1][1]) | (*tlst = 0)
    then {                    # orphaned reference
        subject := rlst[1][1]
        refs := pop(rlst)[2]
        ttls := []
    } else {                # message
        subject := tlst[1][1]
        ttls := pop(tlst)[2]
        if subject == rlst[1][1]    # handle references, if any
        then refs := pop(rlst)[2]
        else refs := []
    }
    if \indexrun then {
        present(subject,ttls,refs)
    } else
    every WriteMail(msg[!ttls | !refs])
}

end

procedure present(s, t, r, owid)

local    ts        # temp string for message numbers

/owid := 72
ts := ""
if *t > 0 then {
    every i := 1 to *t-1 do ts ||:= t[i] || ", "
    ts ||:= t[-1]
} else ts := "-"
if *r > 0 then {
    ts ||:= "; "
    every i := 1 to *r-1 do ts ||:= r[i] || ", "
    ts ||:= r[-1]
}
s ||:= repl(" ",3-(*s % 3))
write(s,right(ts,owid-*s,".. "))
return
end







From kamikaze at kuoi.asui.uidaho.edu  Tue Apr  2 22:26:12 2002
From: kamikaze at kuoi.asui.uidaho.edu (Mark 'Kamikaze' Hughes)
Date: 3 Apr 2002 03:26:12 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <slrnaaktij.kr1.kamikaze@kuoi.asui.uidaho.edu>

Sat, 30 Mar 2002 00:39:10 -0500, Guido van Rossum <guido at python.org> spake:
> PEP: 285
> Title: Adding a bool type
>     1) Should this PEP be accepted at all.

  Excellent - I have TRUE and FALSE defined in most of my Python code
for exactly this reason, and I'd much rather have a system-defined pair
of constants with some language support instead.

>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".)

  str(True) should be "True", str(False) should be "False" - the main
annoyance I had with enumerated types in C was that they printed as
ints, not as their names.  What bloody good was that?

  Backwards compatibility is a problem, though.  You may want to have an
optional warning in 2.3 - str(bool) will be difficult to find with just
a grep, you need to examine every line of the code, or test it at
runtime.

  If everyone had comprehensive unit tests, they wouldn't have this
problem, but that's not often the case.

>     3) Should the constants be called 'True' and 'False'
>        (corresponding to None) or 'true' and 'false' (as in C++, Java
>        and C99).

  While I'm more comfortable typing true and false from Java programming
habits or TRUE and FALSE from old C habits, True and False are more
consistent with Python's style, and you should use those.

>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

  The central problem with that is that Python doesn't have a ternary
operator, which means that Python programmers use math on booleans more
than usual.  I don't see a problem with this, anyway, as long as people
know what the int values of True and False are.

>     5) Should operator.truth(x) return an int or a bool.  Tim Peters
>        believes it should return an int because it's been documented
>        as such.  I think it should return a bool; most other standard
>        predicates (e.g. issubtype()) have also been documented as
>        returning 0 or 1, and it's obvious that we want to change those
>        to return a bool.

  truth() should return a bool, and the documentation should be changed,
to make the current use of 0 and 1 implicitly meaning false and true
explicit.

>     It has been suggested that, in order to satisfy user expectations,
>     for every x that is considered true in a Boolean context, the
>     expression x == True should be true, and likewise if x is
>     considered false, x == False should be true.  This is of course
>     impossible; it would mean that e.g. 6 == True and 7 == True, from
>     which one could infer 6 == 7.  Similarly, [] == False == None
>     would be true, and one could infer [] == None, which is not the
>     case.

  It's also not necessary - bool(6) == bool(7) is the expression being
reached for there, and it works just fine.

-- 
 <a href="http://kuoi.asui.uidaho.edu/~kamikaze/"> Mark Hughes </a>
"No one is safe.  We will print no letters to the editor.  We will give no
space to opposing points of view.  They are wrong.  The Underground Grammarian
is at war and will give the enemy nothing but battle." -TUG, v1n1


From h.e.w.frater at cs.cf.ac.uk  Sat Apr 20 00:28:47 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 19 Apr 2002 21:28:47 -0700
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com> <ubu2a0dbso0u66@corp.supernews.com> <mailman.1019153737.24037.python-list@python.org> <slrnabu7rm.e84.gerhard@lilith.my-fqdn.de>
Message-ID: <e2e9e807.0204192028.76e43129@posting.google.com>

OK, so the way I am reading this is as follows:

Beacause the script is being run as a 'nobody' user, who doesn't have
a .cshrc.mine file which I've got to set all my paths so all my
modules work when run in my home directory - the user can't find the
paths to the modules.

So I need to set the environment variable so that DCOracle2 can find
the oracle compiled whatever language modules. These reside in
/u04/app/oracle/product/OraHome1/lib on our systems here at UNI.

So I need to add this to my LD_LIBRARY_PATH. But I need to do this in
the start of my CGI script. But what should I use to set this:

import sys
sys.path.append(PATH TO LIBS)
or:

import os
os.environ[LD_LIBRARY_PATH] = 'PATH TO LIBS'

I need an answer quite soon so please help.
Thanks in advance,
Hugh


From tim.one at comcast.net  Sat Apr 13 22:27:29 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 22:27:29 -0400
Subject: 2.1 vs. 2.2
In-Reply-To: <roy-56DE95.20201113042002@news1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEPLONAA.tim.one@comcast.net>

[Roy Smith]
> ISO certification is a wonderful thing.  It lets people who worry about
> that sort of stuff check off the "ISO certification" checkbox and
> be happy.

Which is why it's rare to see a thing go thru the expense of ISO
standardization unless vendors stand to make more money back by selling to
the "ISO certification checkbox" crowd; or unless that crowd has been so
burned by gratuitous variations in a thing that it's worth the expense to
them to get an ISO checkbox to beat their vendors over the head with
(although, in the case of languages, they're likely to find that the
eventual standard is carefully worded to allow all variations that already
exist).

> Other than that, it's not clear it has any value.

Oh, there's real value in having a precise standard -- even one so
weasel-worded as to allow getting away with anything.  Last I looked, the
Fortran 77 standard didn't guarantee anything about

     I = 3

because nothing in the standard required a conforming processor to supply
integers as large as 3 <wink>.





From hzhu at mars.localdomain  Mon Apr 15 03:47:24 2002
From: hzhu at mars.localdomain (Huaiyu Zhu)
Date: Mon, 15 Apr 2002 07:47:24 GMT
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com> <mailman.1018655308.18781.python-list@python.org> <slrnabf1f0.mfc.huaiyu@gauss.almadan.ibm.com> <87zo083vgg.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <slrnabktvn.20b.hzhu@mars.localdomain>

Stephen J. Turnbull <stephen at xemacs.org> wrote:

>You have misunderstood from the beginning.

Well, what you wrote does not help in any way to clarify the situation,
except to make it clear that you have a big issue with GPL.

>    Huaiyu> So now my mental picture is 
>
>    Huaiyu>    (copyleft) is subset of (free) is subset of (open source)
>
>The second is not a proper inclusion.

That's interesting.  In your other follow up to my post, you said

>    Huaiyu> Using two terms to mean the same concept when we do have
>    Huaiyu> two distinct concepts to name leaves us one term short.
>
>But we only have one concept.  What we have two of is _movements_.

How come they are not equal if they are also the same?  

<rant>
I'm interested in what people mean when they say "free software" or "open
source software".  I might also be interested in hearing the justification
given for different preferences, but not before the speaker makes clear what
he means by these terms.

If you know that there are at least two movements that do not even agree
on the usage of terms, it is somewhat pointless in starting a sales pitch
without identifying which movement you are pitching for and how its terms
differ from the others.
</rant>

Huaiyu


From aahz at pythoncraft.com  Thu Apr 25 11:51:18 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Apr 2002 11:51:18 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com>
Message-ID: <aa98lm$afj$1@panix3.panix.com>

In article <ad381f5b.0204250629.5144d196 at posting.google.com>,
Rad <zaka07 at hotmail.com> wrote:
>
>I agree that the actual work is not that complex even though there are
>quite a few other bits involved in the project that I thought were not
>relevant to mention in my earlier post, I also agree that it could
>be done in a relational database but if the files were smaller, for
>example Access (which I know quite well), I also know some SQL, but I
>never used MySQL, which is perhaps a possible route to take.  However,
>as I just mentioned I don't know anything about MySQL (except that one
>has to pay for it) + having been learning Python in the last few weeks
>I kind of like it + I was encouraged by your posts and I'm now pretty
>convinced that Python2.2 is the way to go. I won't have to access data
>randomly, just pull out the data fitting certain criteria from original
>4 files and create resulting files.  I might decide (depending on the
>size) to put those into a database though.  I don't understand how I
>can interface Python to a database?

MySQL is free; it's support that costs.  Python has libraries that allow
you to access a database through Python functions and classes.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From jepler at unpythonic.net  Thu Apr 25 17:50:56 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 25 Apr 2002 16:50:56 -0500
Subject: Is there a __start__ method?
In-Reply-To: <3CC86462.72244834@ingenta.com>
References: <3CC86462.72244834@ingenta.com>
Message-ID: <20020425165056.E14065@unpythonic.net>

What good will it do you for your GUI to appear on the screen if it's
frozen anyway until this process of "accessing 3 separate URLs"
completes?

You can use the .after_idle() method in tk to schedule something to
happen when the gui becomes "idle".  From the tk documentation:
    after idle script ?script script ...?
	Concatenates the script arguments together with space separators
	(just as in the concat command), and arranges for the resulting
	script to be evaluated later as an idle callback. The script will
	be run exactly once, the next time the event loop is entered
	and there are no events to process.  The command returns an
	identifier that can be used to cancel the delayed command using
	after cancel. If an error occurs while executing the script then
	the bgerror mechanism is used to report the error.
however, while the "script" executes, the GUI will block.

You could arrange a system using Tcl file handlers, but last I knew
_tkinter.createfilehandler did not exist on Windows. (The underlying Tcl
API is marked "unix only" in the Tcl documentation)

Lastly, you could use a thread for this processing.  However, as another
thread recently said, only the main thread can access Tk stuff....

Jeff




From sholden at holdenweb.com  Tue Apr 30 11:17:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 11:17:24 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>	<3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca> <mailman.1020126468.7915.python-list@python.org>
Message-ID: <ECyz8.16652$xy4.10715@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1020126468.7915.python-list at python.org...


> [James J. Besemer]
[ ... ]
> Nor is Sarcasm or frank talk necessarily unfriendly.  My close friends
tease
> each other mercilessly.  After an exchange that grew a bit tense at
points,
> Steve Holden and I emerged better friends and better acquainted with each
other
> than I think we would have otherwise.  Many people rely on close friends
to
> utter the bare truths that mere acquaintences would avoid.


While I don't wish to disagree with you here, one point I forgot to mention
in my last posting: very few people have the thick skin, intelligence, wit,
humor and modesty :-) that I do.

You might therefore wish to ... dare I say ... lighten up yourself when
dealing with these wimpy foreign types.

leave-xenophobia-to-the-timbot-he-does-it-so-well-ly y'rs  - steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From sqrt_2  Sun Apr 14 14:15:54 2002
From: sqrt_2 (sqrt_2)
Date: Sun, 14 Apr 2002 18:15:54 GMT
Subject: Jython help pls
Message-ID: <3cb9c6c3.115241899@news.mweb.co.za>

Can anyone help with Jython compiler error messages. I find it
difficult to understand what the errors are as I have only just
started with this product.
Messages  like: 
'NameError: Java'
What does that mean?? Is there a reasonable site where I can find
explanations for common error messages? The Jython site FAQ is no help
at all.


From philh at comuno.freeserve.co.uk  Sat Apr  6 12:48:02 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sat, 6 Apr 2002 18:48:02 +0100
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net> <mailman.1018036693.6922.python-list@python.org>
Message-ID: <slrnaaud6i.3j0.philh@comuno.freeserve.co.uk>

On Fri, 5 Apr 2002 11:48:06 -0800, Geoff Gerrietts <geoff at gerrietts.net> wrote:
>
>The second is that I personally find little to admire in C++, but one
>of the things I do admire is that Bjarne decided to take a couple
>years off to let the compiler writers and application programmers
>catch up. The result was a language that everyone more or less thought
>they knew, a solid foundation for people to build on and move forward.

Indeed so. Stroustrup has been concerned not to break backwards 
compatibility with changes. This is probably one of the reasons C++ 
is as popular as it is.

>I remember in the early days of Java, when 1.0 went to 1.1 and
>everything broke, all the developers I was learning from expressed
>disgust. That was the first big indicator that the Java's promise had
>crossed fingers to some. To others it was a sign that Java was an
>immature language.

It was immature.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From mark.charsley at REMOVE_THIS.radioscape.com  Mon Apr 15 06:30:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Mon, 15 Apr 2002 11:30 +0100 (BST)
Subject: Beginner's performance problem
References: <a95d1v$avb$1@panix1.panix.com>
Message-ID: <memo.20020415113054.1532C@a.a>

In article <a95d1v$avb$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> I'm familiar with TVP, but what's TVM?

"Thanks very much". What's TVP? "That's very python-esque"?

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From philh at comuno.freeserve.co.uk  Mon Apr  8 06:49:21 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 8 Apr 2002 11:49:21 +0100
Subject: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> <mailman.1018118665.30693.python-list@python.org>
Message-ID: <slrnab2tdh.2ot.philh@comuno.freeserve.co.uk>

On Sat, 06 Apr 2002 20:39:41 +0200, Laura Creighton <lac at strakt.com> wrote:
>Mark McEahern writes:
>> [Laura Creighton]
>> > [...]  I am rejecting the existence of booleans in the language
>> > because it is too hard to teach how to use them properly.
>> 
>> How do you use them properly?
>> 
>> Suppose I want to validate users.  I might write something like this:
>> 
>> 	def isValid(user, password):
>> 		if ...:
>> 			return True
>> 		else:
>> 			return False
>
>My pleasure, especially since this is such a cool example.  You have
>just shot yourself in the foot, and do not know it.  When you have
>shipped your isValid function, across the world to your 10,000 clients,
>you will discover this.
>
>Your boss comes in and says, oh, Mark, we need a change to that
>isValid function.  We need to return 3 states, Valid, Invalid, and
>Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
>
>What do you do now?  

You need 2 functions: isUserValid(user, password) and
hasUserPaidUpToDate(user).



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From akuchlin at ute.mems-exchange.org  Tue Apr 23 10:20:15 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 23 Apr 2002 14:20:15 GMT
Subject: Mass Mailing help needed (Paul Rubin)
References: <mailman.1019325819.29016.python-list@python.org> <MPG.172e025d803fc2009896bf@news.nwlink.com> <mailman.1019553582.18769.python-list@python.org>
Message-ID: <slrnacarct.3s6.akuchlin@ute.mems-exchange.org>

In article <mailman.1019553582.18769.python-list at python.org>, 
	Graham Ashton wrote:
> With that approach you have to open one TCP connection per message. This
> is expensive, especially as your SMTP daemon should be doing all sorts
> of checks on your IP to see if it'll happily accept mail from you. Not

It's certainly possible to send multiple messages over a single SMTP
connection, and IP checks can be loosened by tweaking the SMTP
server's configuration.  For example, it would be easiest and most
secure to disable address checking on connections from localhost, and
then run your bulk mailing software on the same machine as the SMTP
server.  We've had to do this in order to make a 1700-member mailing
list run acceptably; without disabling the address checks, each
message to the list resulted in 1700 DNS lookups before the message
was even accepted.

--amk                                                             (www.amk.ca)
  "What do we say if the door is opened?"
  "'Hello'?"
    -- Richard Mace and Adric, in "The Visitation"


From pyth at devel.trillke.net  Tue Apr 30 08:06:44 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 14:06:44 +0200
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <200204292311.57363.bhan@andrew.cmu.edu>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org> <200204292311.57363.bhan@andrew.cmu.edu>
Message-ID: <20020430120644.GZ16217@devel.trillke>

On Mon, Apr 29, 2002 at 11:11:57PM -0400, Benjamin Han wrote:
> >
> > s = " ".join([str(i) for i in l])
> 
> This surely looks nicer. I'm just wondering how far Python has fared in terms 
> of this concept of OO? This example seems to show a very different character 
> of OO in Python, in that __str__ is not automatically called, which seems 
> rather counter-intuitive.

It doesn't seem counterintuitive to me (any more). You might be interested in 
reading about the design-principles of python here:

http://www.python.org/dev/culture.html

You will find that 'expliciteness' is valued very high. That's
a design decision which works out greatly, in my opionion.
There are always ways to increase convenience explicitely
without interfering with *the whole system*. Take your
example. You have a list that inherits from the list-type
in python. You want to join the elements using
a specified character. The *generic* string.join method
does not autoconvert for you (like c++ might do).  
But it is very easy to provide convenience:

class Mylist(type([])):
    def strjoin(self, char):
        return char.join(map(str, self))

and then calling

print Myline.strjoin(',')

Compare this to the pain of inheriting from
the C++-STL-sequence classes and defining appropriate
auto-conversion methods. Btw, i prefer to
inherit from 'type([])' to make it visible
that i am extending a basic type. 'list' could
be anything IMO. The Unification of Types and 
Objects is a recent development. See here
for reference:

http://www.python.org/peps/pep-0253.html

> Without private members, and with seemingly complex ways of dealing with class 
> members and methods - so what's the state of OO in Python? Is it still 
> rapidly evolving? How much is an effort in converting Python into C++? Anyone 
> would like to comment on some of these questions, or point to some article, 
> FAQ (esp. helpful from a C++ programmer's point of view)?

Coming from c++ i find it very useful to work my way through the various
PEPs (Python Enhancement Proposals). From time to time, not all at once :-)

http://www.python.org/peps/

There are probably other resources which others can point you too.

have fun,

    holger




From jacob at boris.cd.chalmers.se.cd.chalmers.se  Sat Apr  6 14:14:58 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 6 Apr 2002 19:14:58 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org>
Message-ID: <a8nhfi$gbt$1@nyheter.chalmers.se>

1) Should this PEP be accepted at all.

No.

I come from a background where I have taught students hands-on programming
in the saltmines of my local university, taught engineers in the industry
how to program and where I have implemented actual programs for various uses,
from the humble MCU that controls a valve, to systems that manage the
collected literature of all academia in Sweden. I have taught using about
a dozen different programming languages and last I counted, I had written
programs in about 30 languages, including some assembler dialects.

My humble opinion is that new features in a language should make it easier
to accomplish what you set out to do - create a working program.

Python is a language that is weakly typed, essentially allowing a statement
like

x = f(y)

or

x = a + b

to assign a value of any type to x. This value may be an int, a float,
a string, a list or a dictionary (and an imaginary).

If we ignore the imaginary, which has a very specialized scope, we have 5
base types. These can interact in approximately 5! (120) different ways in a
statement. If we add one more base type we get 6! (720) different ways
that our types can interact.

For a student as well as an active implementor, it is necessary to learn
what the results of all (or at least most of) these interactions are.
Add one more base type and you have increased the complexity of the type
system by a large factor.

As a contrast, teaching how the current type system works is a matter of
a 15 minute lecture, that no student will misunderstand. The hard thing to
teach is how to interpret a "Type mismatch error" and why unexpected
automatic conversions happen. Automatic conversions are a pitfall even for
very seasoned programmers.

Closely related types also make things more difficult than highly ortogonal
ones. Booleans, like ints and floats are scalar, so relations between
these 3 will be more important than say, the relationship between an int and
a dictionary.

If we take the example of C++, one of the reasons it is really hard to teach
and to use is the very close relationship between references and pointers.
Of course it is very nice to be able to take a reference, cast it into a
pointer, subtract it from another pointer and get an int, or a long, or an
unsigned long, or whatever. (Yes, multiple length, closely related types
add a lot of complexity. (And I didn't even mention unsigned ;-))) 

You can learn to master complexity as you become proficient in a language and
you will select a subset of its features that you know and understand.
People who write C code and have never used a function pointer or an enum or
a const declaration are in vast majority, even though these are all great
tools in the right context. C is only moderately more complex than Python.


So, is a boolean always a waste of time?

No, I don't think so. In a statically typed language, they actually catch
some errors at compile time. Python is dynamically typed.

In a language for embedded applications, you can save execution time and
memory space by using a single bit for your boolean variable. Python
is good for many things, but it is not optimised for embedded applications.

It does add readability to some code to use the literals 'true' and 'false'.
Python contains ample constructs for the user to do this. You can even add
a module in the standard library called "bool", exporting "true" and "false".


Guido thinks this is a small change to the language. I strongly disagree.
My conclusion is that the addition of a boolean type to Python would
hit it where it hurts most. Python is a simple language, because it has
a very small and simple type system. Making it harder to learn by increasing
the complexity is going to drive Pythons core supporters - the biologists,
physicists, astronomers, sociologists and other scientists - to look for
other alternatives, that require less work to understand.

Jacob Hall?n
AB Strakt

(Yes, I know that booleans are implemented in a cheat way according to the PEP,
 making them almost aliases for int 0 and int 1. I does not ease the burden
 of knowing the type interaction information.)
-- 


From gcordova at hebmex.com  Thu Apr  4 16:00:32 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 4 Apr 2002 15:00:32 -0600 
Subject: cgi.py performance
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB68@mail.hebmex.com>

> 
> Hey, that looks pretty cool.  I do have one question before I 
> try it out: in
> mime.py, Entity.__init()__, which splits up the data on the 
> boundaries,
> loads the entire file at once, if I am not mistaken ("msg = 
> fp.read()").
> What will happen if I am transferring a very large file, 50-100MB?
> 
> Thanks,
> Jim
> 

Well, for a moment there, you're gonna have instances
of huge strings, really.

A good multipart/form-data parser should never load
the whole data into memory at once, but load it up
in small chunks while splitting and parsing it.

I once had a C module that did precisely that,
both with url-encoded and with multipart/form-data,
except I can't remember what I did with that.

Anyhow, it's perfectly doable from python, also.

-gustavo




From mgilfix at eecs.tufts.edu  Tue Apr 23 00:53:26 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 23 Apr 2002 00:53:26 -0400
Subject: need encryption algorithm
In-Reply-To: <e2e9e807.0204222041.5fa92f05@posting.google.com>; from h.e.w.frater@cs.cf.ac.uk on Mon, Apr 22, 2002 at 09:41:30PM -0700
References: <e2e9e807.0204221741.597acd60@posting.google.com> <e2e9e807.0204222041.5fa92f05@posting.google.com>
Message-ID: <20020423005326.D21282@eecs.tufts.edu>

  I hope you're not a CS major or planning on being a CS professional.
Otherwise, that's gonna suck.

On Mon, Apr 22 @ 21:41, Hugh wrote:
> I fully understand what everyone is going on about with regard to me
> saying that the quality of the encryption is not that important. The
> fact of the matter is, I'm only writing the software for my 1st year
> university project and in reality it's only a theoretical system, no
> one is actually going to use it. I'm using the exercise to learn
> python and basic database design although to be quite honest I loathe
> internet programming. Don't know why, just don't like it.
> 
> HUgh
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (h.e.w.frater)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From dkuhlman at rexx.com  Sun Apr 14 13:54:53 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: 14 Apr 2002 17:54:53 GMT
Subject: Reading/writing binary data.
References: <mailman.1018285938.26500.python-list@python.org> <uQks8.20420$vm6.3149576@ruti.visi.com> <mailman.1018302086.27758.python-list@python.org>
Message-ID: <a9cfpc$237in$1@ID-139865.news.dfncis.de>

Ben Logan <ben at wblogan.net> wrote:
> 
> Thanks, Grant and Jeff, for your help.  Both of those modules look
> very promising and I never would have found them. :)
> 
> Regards,
> Ben
> 

Not sure what your requirements are, but if you are not stuck on
finding a *binary* representation, you might also want to look at
xdrlib and pickle in the Python standard library.  They both
support serializing and writing Python data structures to files.

And, if you want a persistent, dictionary-like object, see the
shelve module in the Python standard library.

Also, and especially if you are an XML person. you might want to be
aware that there are ways to write (and read) (some? there are
restrictions here, I'm pretty sure.) Python data structures to XML. 
If this is what interests you, go to
http://www.python.org/sigs/xml-sig/, get the PyXML distribution,
then look at:

    /w2/Python/PyXML-0.7/xml/marshal/wddx.py
    /w2/Python/PyXML-0.7/xml/marshal/generic.py

A google search for "wddx faq" will lead you to plenty of
information about WDDX.

  - Dave


-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman


From jacobs at penguin.theopalgroup.com  Wed Apr  3 11:31:03 2002
From: jacobs at penguin.theopalgroup.com (Kevin Jacobs)
Date: Wed, 3 Apr 2002 11:31:03 -0500 (EST)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <00af01c1db2a$e7316d10$ced241d5@hagrid>
Message-ID: <Pine.LNX.4.44.0204031119240.29628-100000@penguin.theopalgroup.com>

On Wed, 3 Apr 2002, Fredrik Lundh wrote:
> Andrew Koenig wrote:
> > Specifically: If I want to write a function that answers a yes/no
> > question, I have lots of possible ways of spelling yes (1, 2, "yes",
> > and so on) and lots of possible ways of spelling no (0, {}, None,
> > and so on).  There isn't a single preferred way.
> 
> if you study real python code, you'll find that there is.

Having managed 200k+ LOC Python projects with at times 20+ developers, I
find that there isn't.  At least not without getting out a baseball bat.

The first thing many of our programmers want to do at the top of their
modules is to define:

False = 0
True = not False

    or

false = 0
true = 1

   or 

TRUE = 1
FALSE = 0

  or 

from my_local_trashcan_module import true, false

  or some fairly extreme object implementations in Python or C.

Of course, I have to break their fingers and force them to stop re-inventing
their own true and false literals, and to use 1's and 0's like the rest of
the disadvantaged Python world.  Ah, the joys of code review.... Believe it
or not, this, naming conventions, and consistent exception handling styles
are among the biggest bones of contention we have during our code review
sessions.

The moral of the story is that I will be extremely happy once there is a
globally blessed way of spelling true and false other than 1 and 0.  Of
course, I and my team aren't exactly typical Python developers.  Though
wouldn't it be nice if we were?

-Kevin

--
Kevin Jacobs
The OPAL Group - Enterprise Systems Architect
Voice: (216) 986-0710 x 19         E-mail: jacobs at theopalgroup.com
Fax:   (216) 986-0714              WWW:    http://www.theopalgroup.com





From duncan at NOSPAMrcp.co.uk  Mon Apr 15 07:35:13 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 15 Apr 2002 11:35:13 +0000 (UTC)
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net>
Message-ID: <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1>

Scott David Daniels <Scott.Daniels at Acm.Org> wrote in
news:1104_1018820620 at news.dsl-only.net: 

>> Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:
>> > Thank you very much for pointers, but I can think of something not
>> > listed there. For example, __all__ for "from module import *",
>> > pickle protocols, __new__ for subclassing from a type and so on. Is
>> > there a complete (or near-complete) list?
>> ...
> 
> You could also take this lack of a list as a charter for your
> collecting _all_ (not __all__) such names, and creating a document for
> others who, like you want to know the list of such names.  This could
> include package references if you so desire (others will complain if
> you leave something out, providing you with more fodder for your
> document). 
> 

To get you started, here is a list built by grepping over the Python 
sources. Some of the names are probably red herrings, and there are 
probably some used by libraries not in this list, but it should give you a 
start:

__GNUC__
_____________undefined______________
__abs__
__across__
__add__
__all__
__and__
__assert__
__base__
__bases__
__basicsize__
__builtin__
__builtins__
__call__
__check__
__class__
__cmp__
__coerce__
__complex__
__console__
__contains__
__copy__
__copyright__
__debug__
__declare__
__deepcopy__
__del__
__delattr__
__delete__
__delitem__
__delslice__
__dict__
__dictoffset__
__displayhook__
__div__
__divmod__
__doc__
__domain__
__eq__
__excepthook__
__file__
__flags__
__float__
__floordiv__
__future__
__ge__
__get__
__getattr__
__getattribute__
__getinitargs__
__getitem__
__getslice__
__getstate__
__gt__
__hash__
__hello__
__hex__
__iacross__
__iadd__
__iand__
__idiv__
__ifloordiv__
__ilshift__
__imod__
__import__
__imul__
__init__
__int__
__interfaces__
__invert__
__ior__
__ipow__
__irshift__
__isub__
__itemsize__
__iter__
__itruediv__
__ixor__
__le__
__len__
__lltrace__
__long__
__lshift__
__lt__
__main__
__members__
__metaclass__
__methods__
__mod__
__module__
__mro__
__mul__
__name__
__ne__
__neg__
__new__
__nonzero__
__objclass__
__oct__
__or__
__path__
__paul_prescod_hack_constant__
__phello__
__pos__
__pow__
__racross__
__radd__
__rand__
__rccheck__
__rdiv__
__rdivmod__
__reduce__
__repr__
__return__
__rfloordiv__
__rlshift__
__rmod__
__rmul__
__ror__
__rpow__
__rrshift__
__rshift__
__rsub__
__rtruediv__
__rxor__
__safe_for_unpickling__
__self__
__set__
__setattr__
__setitem__
__setslice__
__slots__
__stderr__
__stdin__
__stdout__
__str__
__sub__
__subclasses__
__test__
__thisclass__
__truediv__
__types__
__unicode__
__version__
__warningregistry__
__weakref__
__weakrefoffset__
__xor__


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bokr at oz.net  Thu Apr 11 17:33:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Apr 2002 21:33:58 GMT
Subject: binding a reference to a variable
References: <mailman.1018540703.7281.python-list@python.org>
Message-ID: <a94vg6$7f9$0@216.39.172.122>

On Thu, 11 Apr 2002 08:56:32 -0700, Patrick Miller <patmiller at llnl.gov> wrote:

>Bengt writes>
>> >>> def set(name, val, ns=locals()):
>>  ...     ns[name]=val
>>  ...     return val
>
>Bengt noted that 'ns' is local to the declaration module of the
>'def set.'  
>
I also noted
"Of course, def set ... must be executed where it can capture ns=locals()
for use inside the function, where locals() is another space (and read-only)."
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

... i.e., that locals() inside a function is different and read-only,
which would predict the following.

>The workaround for that is
>
>def set(name,val):
>   import sys
>   ns = sys._getframe().f_back.f_locals # local dictionary of caller
>   ...
>
>BUT WARNING!!!!! There is a bigger problem --  Changes to the local
>dictionary are NOT reflected in the state of the execution.
>
unless locals() is globals() => 1
as is apparently normally the case at module scope, e.g.,

 +---< locagloba.py >-------------
 |print 'locagloba id(locals())=%s, id(globals())=%s' % (id(locals()), id(globals()))
 |def set(name, val, ns=locals()):
 |    ns[name]=val
 |    return val
 |set('z',456)
 |print z, locals()['z']
 +--------------------------------
 Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
 Type "help", "copyright", "credits" or "license" for more information.
 >>> import locagloba
 locagloba id(locals())=8029232, id(globals())=8029232
 456 456
 >>> locagloba.z
 456

 >>> print 'interactive id(locals())=%s, id(globals())=%s' % (id(locals()), id(globals()))
 interactive id(locals())=7989472, id(globals())=7989472
 >>> locals() is globals()
 1
 >>> locagloba.z
 456

OTOH, might as well use globals() in the first place ;-)

>The following does not work as intended....
>
>def f():
>   a = 7
>   L = locals()
>   print L.keys() # Prints ['a','L'] or some such
>   L['a'] = 99  # Update the value in L
>   print a      # But wait!  This prints out '7'
>
>The issue is the implementation of the code object supporting f.
>The value of the locals dictionary is created on the fly by
>the call to locals().  The actual values of the local variables
>are on a tiny, predefined stack frame that is used by the bytecode
>interpreter wherein local variables are known only by an
>integer offset.
>
>Creating true "local" variables is astonishly tricky.
>
>You can actually figure out where names map into the frame instance
>(weave does (or will do) it), but it is messy and very implementation
>dependent.
>
But hiding that mess is the job of an implementation ;-)

>From what's been posted, ISTM that making locals() read/write
inside a function is feasible, with a performance hit. Probably
no one has wanted it badly enough to do it at all, never mind finding
a way to penalize only the locals()-modifying users.

Regards,
Bengt Richter


From googlegroups at sixtyten.org  Mon Apr 22 10:55:01 2002
From: googlegroups at sixtyten.org (Allan Crooks)
Date: 22 Apr 2002 07:55:01 -0700
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <0evu9a.621.ln@127.0.0.1> <mailman.1019413960.21990.python-list@python.org> <gs6v9a.vb1.ln@127.0.0.1>
Message-ID: <304d20df.0204220654.3924d79c@posting.google.com>

Carl Banks <imbosol at vt.edu> wrote in message news:<gs6v9a.vb1.ln at 127.0.0.1>...
> holger krekel wrote:
> > On Sun, Apr 21, 2002 at 02:12:16PM -0400, Carl Banks wrote:
> >> Allan Crooks wrote:
> >> > Hi,
> >> > 
> >> > I was wondering what the best approach is on providing libraries for
> >> > varying versions of Python?
> >> > (...)
>  
> >> # Import stuff for version 2.1
> >> if sys.version[:2] >= (2,1):
> >>     from mymodule2_1 import *
> >> 
> >> # Import for version 2.2
> >> if sys.version[:2] >= (2,2):
> >>     from mymodule2_2 import *
> > 
> > this solves part of the 'code duplication' problem. Might get a bit tricky
> > if you just want to define a few additional methods for some classes
> > (for example 'generator versions' of methods otherwise returning lists).
> 
> I wouldn't say so.  It is straightforward enough to add a method to a
> class any time:
> 
> >>> class a:
> ...    pass
> ...
> >>> def p(self,x):
> ...    print "hello, %s" % x
> ...
> >>> a.p = p
> >>> b = a()
> >>> b.p("world")
> hello, world

That would still be problematic, how would you refer to class "a" that
is defined in another module? Say you had the following code in
module_20:

class a:
    def hello(self): return 'hello there!'

And you had this code in module_22:

from module_20 import a
a.hello = lambda self: 'hello there again!'

If module_21 then defines it's own 'hack' of module_20's 'a', then
module_22 will lose the change, unless you are aware of the import in
module_21.

I know it's a reasonably simple change, but if you forget about it, it
could be a bit confusing. I think the best thing to do would be to
have the 'newer' module inheriting the entire namespace from the older
module.

So module_21 will have a "from module_20 import *" line, module_22
will have "from module_21 import *" and so on, with the frontend
module doing the correct import of whichever module corresponds to the
version of Python being used.

I'm also a bit reluctant of namespace hacking of classes, since it
just seems too... 'hacky'. It would also make inheritance a problem.

Maintaining concurrent versions doesn't seem as bad as I first
thought...

> I just did it how the OP did it, without bothering to check.  It
> should be sys.version_info[:2] >= (x,y)

Ooops, my mistake.

I like the ideas people come up with though...

Allan.


From bokr at oz.net  Sat Apr 13 21:34:51 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 01:34:51 GMT
Subject: Python 2.2 unwise.exe
References: <mailman.1018734147.26303.python-list@python.org>
Message-ID: <a9ambr$edo$0@216.39.172.122>

On Sat, 13 Apr 2002 17:40:32 -0400, Tim Peters <tim.one at comcast.net> wrote:
[...]
>[Robin Becker, on unwise.exe vanishing]
>> Installed under C:\Python; system is Win2k. I installed win32all-146
>> after Python-2.2, but didn't save backups.
>>
>> I have a couple of other things installed eg Numeric-20.3 and PIL, but
>> nothing  that came with an installer. I don't remember anything untoward
>> about the Python-2.2 install.
>>
>> The unwise.exe was just not there for some reason.
>
>Crud, I was hoping it was unique to Win9x.  "Just not there for some reason"
>is the same thing I've seen.  I build the PLabs Windows installer, so I get
>more practice with it than anyone else.  "Just not there" has definitely
>happened more than once after I've run an install and verified that
>UNWISE.EXE is indeed exactly what and where it should be after the install
>completes.  Yet two hours or two weeks or two months later, it's plain gone.
>Nothing else is missing when this happens, just UNWISE.EXE.  That's all I
>know for sure about it ... OK, I'm going to change UNWISE.EXE in my install
>trees to read-only.  Maybe that will catch whatever is deleting it in the
>act.  Heh:  Just did that.  UNWISE.EXE still existed in my 2.1 and 2.2
>trees, but had already gone missing in my experimental 2.3 tree.
>
>windows-is-an-adventure-ly y'rs  - tim
>
>PS: The PLabs Windows installer is a zip file of a kind WinZip knows how to
>read.  That is, e.g., you can open Python-2.2.1.exe directly from WinZip.
>If UNWISE.EXE goes missing for you, you can extract it from the installer
>.exe, using WinZip.  Extract the file UNWISE32.EXE into the root of your
>Python installation, and rename it to UNWISE.EXE.  Then the "Uninstall
>Python" shortcut should work fine again.  Using an older version of
>UNWISE.EXE from somewhere else may or may not work completely; the newer
>version we ship knows how to perform rollbacks, and does a much more
>complete job of undoing *all* Windows install fiddling than older versions
>are capable of doing.
>
If it happens in a Windows (eg NT) where you can set up access audit on a file so
_anything_ done to/with unwise.exe gets logged in the security log, that might
be even more effective than just read-only (which afaik won't prevent a rename or move
(or attrib ;-) BTW, could it have gone invisible rather than missing?). You could scan
the whole disk (including normally hidden files) for a file with identical data, I suppose.

You should also be able to control access better than simple read-only, e.g., by giving
ownership to a special user that reserves all the permissions except read. The most privileged
admin can always get around it, but a ISTM a program would have to have been written to do
the special steps, and I don't know that it could avoid getting caught in the log anyway.
In any case, such a prog doesn't seem like the kind of cause being hunted here.

Regards,
Bengt Richter


From martin at v.loewis.de  Tue Apr 23 15:24:36 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 23 Apr 2002 21:24:36 +0200
Subject: reading unix files from NT using SSH2
References: <mailman.1019572193.18219.python-list@python.org>
Message-ID: <m3bscafcaz.fsf@mira.informatik.hu-berlin.de>

James Strater <James.Strater at exolink.com> writes:

> I have a python script running on an NT box.  I need to connect to a  unix
> box using SSH2 and read some files sitting on that box.  Can anyone point me
> in the right direction on how to go about doing this?   

I recommend to use scp. When that completes, the file will be locally
available.

Regards,
Martin



From phd at phd.pp.ru  Tue Apr  2 03:27:38 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 2 Apr 2002 12:27:38 +0400
Subject: CD Device Name
In-Reply-To: <1017720021.720264@cheroke.optimation.co.nz>; from Russell.Every@MentumGroup.com on Tue, Apr 02, 2002 at 02:59:36PM +1000
References: <1017720021.720264@cheroke.optimation.co.nz>
Message-ID: <20020402122738.C25705@phd.pp.ru>

On Tue, Apr 02, 2002 at 02:59:36PM +1000, Russell wrote:
> Anyone ideas (prefereably Python code) on how to get the label of a CD under
> Windows 2000?

   http://csl.cse.ucsc.edu/~ben/python/

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



From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sun Apr 14 23:44:47 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 20:44:47 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu>
Message-ID: <3cba4cf7@news.mhogaming.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-BBF667.20005414042002 at news.service.uci.edu:

| < snip! >
|
| >>> 0x40000000<<1
| -2147483648
|
| This is inconsistent with int-long unification.  In
| fact, I think it is a bug.  The answer should, of
| course, be 2147483648L, just as you would get by
| multiplying by 2 instead of shifting.
|
| < snip! >

Thanks Dave-- that IS what I was trying to say, but I wasn't
sure enough to use the "B.." word-- you do have a way with
words...

The other thing I was trying to say, is that if this _BUG_
(gee- that wasn't so hard!) in Python is fixed, then we need
a way to consistently do all of this bit shifting.  Now that
ints and longs are 'unified', you are supposed to be able to
write code that doesn't care what the internal
representation is-- that should also apply to the '<<' and
'>>' operators.  BUT, since we NOW don't know where the high
bit is (but the class that is managing the long object
DOES), we need a '>>>' operator to consistently handle a
logical right shift-- (i.e.: there is no way to code this
behavior in Python.)

So-- who wants to report this as a bug?  (I don't even know
how to do that...)





From phr-n2002a at nightsong.com  Wed Apr  3 23:48:56 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 20:48:56 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org> <mailman.1017893479.16809.python-list@python.org>
Message-ID: <7xu1qsnk5j.fsf@ruckus.brouhaha.com>

Andrew Koenig <ark at research.att.com> writes:
> >> No.  That would break backwards compatibility.  False==0, and True==1;
> >> everything else follows from that.  (But False is not 0, and True is
> >> not 1!)
> 
> Ping> This is a strange distinction to make -- one that is not made in
> Ping> any other programming language i have ever encountered.
> 
> I would be very surprised if none of the Lisp family of languages
> has similar anomalies in the relationship between eq and equal.

Python already has a bunch of eq but not equal numeric types:

  1 == 1L      1 is not 1L
  1 == 1.0     1 is not 1.0
  1 == 1+0j    1 is not 1+0j

I felt better about the PEP after Guido pointed this out.


From dkuhlman at rexx.com  Sun Apr 14 23:37:05 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: 15 Apr 2002 03:37:05 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <a9dhsv$28c2c$1@ID-139865.news.dfncis.de>

phil hunt <philh at comuno.freeserve.co.uk> wrote:

[snip]

> 
> In summary:
> 
>    New features => good
> 
>    New features that break existing code => bad
> 

Two points:

1. A language that does not progress and change will wither ... and
die ... slowly ...

2. The progression of Python versions has been, for me, very smooth
and orderly.  I have had plenty of time to adapt to and learn to
take advantage of the changes from one version to the next.  I have
been given plenty of information that enabled me to adapt.  Most
changes have been backward compatible.  Very little old code has
broken.

In summary, stop whining.  OK.  I apologize.  And, actually I'm
grateful that there are people like Phil, here, who keep reminding
the Python dev team to think changes and their consequences through
carefully.

But, if you *really* want something to complain about, look at what
happened in the Java community when they had to switch collection
libraries and had to adapt to Swing.  And, I've read that Visual
Basic programmers are going through some very wrenching changes
because of Visual Basic .Net.  Try walking a mile in their shoes. 
Actually, I apologize again.  I wouldn't wish Java or Visual Basic
on anyone. Forcing a Pythonista to program in Java or VB
would be cruel.

  - Dave

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman


From jaf at lcsaudio.com  Sat Apr 27 17:04:30 2002
From: jaf at lcsaudio.com (jaf)
Date: Sat, 27 Apr 2002 14:04:30 -0700 (-0700)
Subject: inline assignments in conditionals
Message-ID: <7586506669-BeMail@sugoi>

Aggh!  Of course I meant it the other way around.  :^P
It would have been a nice example too, if I hadn't completely
stuffed it up.  So for the record, "If today is Sunday" is my
preferred form (i.e. if day == 0), and I should never again
be allowed to post without proofreading first.

Jeremy

>   I think this example is backwards. Today is the variable.  I'd
> usually say: "If today is Sunday, I'm going to go shopping". So much
> for that :)
>
>             -- Mike
>
> On Sat, Apr 27 @ 22:37, Laura Creighton wrote:
> > Jeremy:
> > [snip]
> > > Writing things that way is really unnatural, at least for English
> > > speakers.  It makes the intent of the code harder to understand
> > > (at
> > > least for me).  For example, if you were talking to a friend, you
> > > wouldn't say:
> > >
> > >    "If today is Sunday, I'm going to go shopping"
> > >
> > > but rather:
> > >
> > >    "If Sunday is today, I'm going to go shopping"
> >
> > Interesting.  I would never use the second construct, and I use the
> > first one all the time.  I wonder if this is a difference between
> > people whose first language is English, and some for whom it was
> > not.
> > I don't think that anybody has had any trouble understanding me
> > when I
> > say 'If today is Saturday, then we are going to the concert house,
> > otherwise we are going to hack.'  But then I have been writing
> > if 0 == x for a long time as well.  (It was great -- Borodin
> > Quartet
> > playing Sjostakovitj and Beethoven, by the way.)
> >
> > Laura Creighton






From lac at strakt.com  Tue Apr  9 11:36:42 2002
From: lac at strakt.com (Laura Creighton)
Date: Tue, 09 Apr 2002 17:36:42 +0200
Subject: What has PEP 285 done to us? 
In-Reply-To: Message from mlh@vier.idi.ntnu.no (Magnus Lie Hetland) 
   of "Tue, 09 Apr 2002 11:23:56 -0000." <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> 
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2CB16.8080103@geneva-link.ch>  <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> 
Message-ID: <200204091536.g39Fagdk022126@ratthing-b246.strakt.com>

> In article <3CB2CB16.8080103 at geneva-link.ch>, Boris Borcic wrote:
> [snip]
> >(b) Laura made a lengthy point, overlooking the improvable while clear 
> >to the language designer's POV, that it very much boiled down to : "I've 
> >stacked my didactics of truth values in Python on the name of None, and 
> >the PEP makes me feel I've been telling lies".
> 
> That's not how I saw it. The main point I extracted was: Until now,
> the Pythonic "truth values" have been "something" and "nothing" (i.e.
> the empty list, zero, None, etc.) With the bool type this is still
> true, but we get a type which will give newbies the impression that it
> isn't.
> 
> >python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
> 
> Hm. What programming language is this supposed to be?
> 
> --
> Magnus Lie Hetland                                  The Anygui Project
> http://hetland.org                                  http://anygui.org
> 

Magnus understood me very well.  I only want to add is that the 
something-nothing is the real empirical truth of Nature, and all rest
is formalism.  Thus, if your teaching job is to train for realism
and to reject empty formalism (learning when formalism is empty and when 
it is worthwhile is exactly what I am teaching and this is hard, hard,
hard, hard, hard) then your job just got harder.  

In contrast --

Boris' phrasing indicates that I was 'trying to make a point' as if
this was some sort of ego-driven competition between myself and Guido
with some sort of pay-off in terms of whether the PEP passes, or
whether more big names on comp.lang.python come out in favour of me
than Guido or some other piece of nonsense.  I did not want this.
Boris has revealed a flaw in our process, because that he took it that
way means that the PEP process is being perceived as an ego-game.

This _really_ stinks, people.  I trusted Guido to be able to take off his
'I am an implementor and I just implemented a cool new feature' hat
and put on his 'I am a langauge designer, and I am now looking at the
language implications' hat.  This trust we must have in Guido or else
he will not be able to launch his own new features into Python.  This
is the faith we must have or else we think that we ONLY GOT LUCKY
in that Python is great.  

Stop worshipping or demonising the man.  Only in bad American movies
do the leaders of the military want blind discipline and worship.
Leadership is hard, and way more fascinating and creative than that.
But this list is not the place to discuss this, either.

Laura Creighton




From pyth at devel.trillke.net  Mon Apr 22 12:33:49 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 22 Apr 2002 18:33:49 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <aa1bh8$frm$1@news1.wdf.sap-ag.de>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org> <304d20df.0204220633.50bf23bf@posting.google.com> <aa1bh8$frm$1@news1.wdf.sap-ag.de>
Message-ID: <20020422163349.GG27395@devel.trillke>

On Mon, Apr 22, 2002 at 05:51:04PM +0200, Daniel Dittmar wrote:
> Allan Crooks wrote:
> > The main problem is when I want to use 'yield', there's no real way of
> > compensating for a language change... I suppose it is my fault for
> > trying too much. :) The more I think about it, the more maintaining
> > concurrent versions seems appealing... at least there is no 'hacking'
> > involved...
> 
> You could do 'yield (expr, ...)', which looks like a yield to 2.2 and like a
> routine call to anyone else.
> 
> There is still the problem how to declare
> from __future__ import generators
> It has to be the very first statement, so it can't be inside a try-except.

yes. actually i looked up the exact condition up in the C-Source. 
It doesn't need to be the first statement in your *module*. It needs
to be the first statement in any *parsing tree*. So

  import os
  from __future__ import generators

does not work. but 

  import os
  exec "from __future__ import generators"

works because the exec-string part leads to a new parsing tree.
the __future__ stuff modifies what the parsers sees as 'valid' python.
that's why it has to be at the beginning of building the parsing tree.

It follows that while you can call any generator defined inside the exec-string
you cannot use "yield"-sytanx outside the exec.

	holger




From alf at logilab.fr  Mon Apr  8 08:16:46 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Mon, 8 Apr 2002 12:16:46 +0000 (UTC)
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns?  [LONG!])
References: <mailman.1018025712.3333.python-list@python.org>
Message-ID: <slrnab32he.au.alf@sagittarius.logilab.fr>

Gustavo Cordova wrote:

> 
> Andy hadn't caught on that a module is loaded only once,
> the first time it's imported. Any subsequent imports are
> only accesses to the original imported version, so that
> makes them effectively singletons.

Stop me if I'm wrong, but this is actually not quite true. 
Suppose you have a package called 'p' containing modules
m1 and m2. You also have another module called 'n' somewhere 
else. m2 is your wanabe singleton, and has a function 
called 'foo'. 

In m1, you can access m2 with a simple 'from m2 import foo', 
but in n, you have to 'from p.m2 import foo'. In this case, 
m2 is imported twice, and changes to global variables are not 
made. 

The solution is to use 'from p1.m2 import foo' in module m1. 

Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From rpyron at alum.mit.edu  Mon Apr  8 16:40:09 2002
From: rpyron at alum.mit.edu (Robert Pyron)
Date: Mon, 08 Apr 2002 20:40:09 GMT
Subject: Tools/Scripts/texi2html.py
Message-ID: <g4t3bugaplcn104b10ip18nghfer4uia89@4ax.com>

I have made some changes to Tools/Scripts/texi2html.py that I think
might be useful for others...

I wanted to produce a Windows HTMLHelp (.CHM) file for MySQL.
To that end, I made the following modifications:

1. BUGFIX: In function makefile(), strip blanks from the nodename. 
   This is necesary to match the behavior of parser.makeref() and 
   parser.do_node().

2. I modified the original texi2html code to add support for
the following items:
     @uref       URL reference
     @image      image file reference
     @multitable output an HTML table
     accents     partial support for accents, to match MAKEINFO output

3. I added a new command-line option, '-hh basename', to specify
HTML Help output. This will cause three files to be created
in the current directory:
      `basename`.hhp  HTML Help Workshop project file
      `basename`.hhc  Contents file for the project
      `basename`.hhk  Index file for the project
When fed into HTML Help Workshop, the resulting file will be
named `basename`.chm.

4. A new class, HTMLHelp, to accomplish item 3.

My version of texi2html.py can be found at:
http://www.dullandboring.com/pystuff/index.htm

Bob Pyron





From tim.one at comcast.net  Sun Apr  7 14:54:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 07 Apr 2002 14:54:11 -0400
Subject: PEP 285: Adding a bool type
In-Reply-To: <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELLOLAA.tim.one@comcast.net>

[Steve Holden]
> ...
> It's becoming clear to me that Guido's biggest mistake was
> proposing this as a PEP. Since none of the arguments expressed as a
> result have swayed his opinion that the change is a good thing, he
> might as well have simply introduced the change by fiat.

Asking for arguments doesn't obligate one to agree with the arguments
presented.  It doesn't even obligate one to respond in detail to every
argument presented (let alone spend 10 hours a day for a month doing so).
It does obligate one to consider all arguments presented, or at least it
does in my opinion -- and in Guido's.

In the case of adding a ternary conditional operator (C ?: workalike), Guido
rejected his own implementation patch, which didn't get to PEP stage because
the arguments against it from people trying the patch were sufficient.

If something gets to the PEP stage, then of course the author (Guido or not)
believes it's a good idea, unless like Laura's PEP 666 it was introduced for
the purpose of rejection.  It's still not a foregone conclusion, as
evidenced by Guido's PEP 259 (in which case the arguments presented
convinced him to reject his PEP).

In the case of PEP 285, Guido asked for community advice on 8 specific
issues, one of which was "Should this PEP be accepted?".  He didn't find the
contra arguments to that one persuasive.  So what?  He found arguments on
some of the other issues persuasive.  I believe the ultimate pronouncements
all matched his initial views, but this didn't surprise me, as backward
compatibility forced most smaller decisions.  At the time he wrote the PEP,
the disposition of str() and repr() were contentious in pre-PEP Python-Dev
debate, and it was quite helpful to see that virtually nobody on c.l.py
favored str(True) returning '1', while several presented arguments against
it doing so.

> That way he could have ignored the resulting howls of protest rather
> than being forced to respond to them.

If Guido's chief motivation were avoiding howls of protest, he'd never say
anything.  He also got howls of agreement, although it's always the case
that partisans on either side have a much better memory for the howls they
agree with.  If Guido would like to avoid anything, I expect it's that
there's howling at all, whether or pro or con -- Usenet debate routinely
gets childish and abusive (not mentioning Robin Becker by name, but if I had
to, I probably would).

 Not much else would be different...

As it turned out in this case, yes, but that this case turned out as it did
was not a necessary outcome, just a likely one (and PEP 259 looked darned
likely the day it was introduced too).





From bokr at oz.net  Fri Apr 12 15:43:57 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Apr 2002 19:43:57 GMT
Subject: ASCII to HEX conversion
References: <34e98ec2.0204111415.504d246f@posting.google.com> <Xns91EE3AF2DC0cliechtigmxnet@62.2.16.82> <292c8da4.0204120356.3709c4f1@posting.google.com>
Message-ID: <a97ddt$8k$0@216.39.172.122>

On 12 Apr 2002 04:56:27 -0700, westernsam at hotmail.com (sam) wrote:

>The following code will give binary from decimal. (N.B. I cannot claim
>credit for this as I didn't write it and just found it somewhere)
>
>import math, operator
>
>def binary_repr(self, number, max_length = 32):
>        assert number < 2L << max_length
>        shifts = map (operator.rshift, max_length * [number], range
>(max_length - 1, -1, -1))
>        digits = map (operator.mod, shifts, max_length * [2])
>        if not digits.count (1): return "00000000"
>        digits = digits [digits.index (1):]
>        returnString = string.join (map (repr, digits), '')
>        while (len(returnString) < 8):
>            returnString = "0%s" % returnString
>        return returnString

A little less importing etc:

 >>> def binstr(n, maxbits=None):
 ...     absn = abs(n)
 ...     s=[chr(((n>>b)&1)+48) for b in xrange(maxbits or len(hex(n))*4)
 ...         if maxbits or absn>>b or not b or (n<0 and absn>>(b-1))]
 ...     s.reverse()
 ...     return ''.join(s)
 ...
 >>> binstr(1)
 '1'
 >>> binstr(0)
 '0'
 >>> binstr(1,8)
 '00000001'
 >>> binstr(0xaaaaaaaa)
 '10101010101010101010101010101010'
 >>> binstr(0xcccc, 32)
 '00000000000000001100110011001100'
 >>> binstr(-1L<<55)
 '110000000000000000000000000000000000000000000000000000000'
 >>> binstr( 1L<<55,57)
 '010000000000000000000000000000000000000000000000000000000'
 >>> binstr(-5)
 '1011'
 >>> binstr(-5,8)
 '11111011'
 >>> binstr(-5,2)
 '11'

Note that I preserved a sign bit as such instead of generating the bits
of the absolute value with a minus sign. This makes an ambiguity, but it
shows the bits of the negative value. You can easily change it as desired.

Regards,
Bengt Richter


From fperez528 at yahoo.com  Sat Apr 13 21:16:53 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sat, 13 Apr 2002 19:16:53 -0600
Subject: Weak references and tracebacks
Message-ID: <a9albr$3ha$1@peabody.colorado.edu>

Is it ok to use weak references to hold tracebacks without generating circular 
references? From the documentation:

Warning:
Assigning the traceback return value to a
  local variable in a function that is handling an exception will
  cause a circular reference.  This will prevent anything referenced
  by a local variable in the same function or by the traceback from
  being garbage collected.  Since most functions don't need access to
  the traceback, the best solution is to use something like
  type, value = sys.exc_info()[:2] to extract only the
  exception type and value. If you do need the traceback, make sure
  to delete it after use (best done with a try ... finally statement) or to 
call exc_info() in
  a function that does not itself handle an exception. Note:
Beginning
  with Python 2.2, such cycles are automatically reclaimed when garbage
  collection is enabled and they become unreachable, but it remains more
  efficient to avoid creating cycles.

So I'm wondering if the following is ok:

type,value,tb=sys.exc_info()
tb_ref = weakref.ref(tb)
del tb

... and then I can keep tb_ref around without this being a problem.

Thanks for any help,

f.


From JamesL at Lugoj.com  Sat Apr 13 23:04:49 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 14 Apr 2002 03:04:49 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <3cb8a755$0$12714$9b622d9e@news.freenet.de> <WL2u8.26958$GS6.2775752@bin3.nnrp.aus1.giganews.com>
Message-ID: <Xns91EFCC39E5237JamesLLugojcom@216.148.53.84>

"Terry Reedy" <tejarex at yahoo.com> wrote:
> 
> "Jens Baader" <nospam at nospam.com> wrote in message
>> If not what's the
>> reason that keeps the people from upgrading to 2.2?
> 
> Perhaps you should ask before jumping to foolish conclusions.  Some
> (overlapping) answers based on newsgroup posts as I remember them:
[ List of alleged reasons... ]
> * ???

Just FYI, I believe you missed a couple of the largest factors why frequent 
upgrades of compilers and interpreters are frowned upon in some places: the 
cost of Quality Assurance (QA) testing and lack of commercial support. QA 
takes time and money. It may take someone only a few minutes to download 
and install the latest version of Python, but a properly run software 
development shop will run QA tests on their applications when changing 
something as important as a compiler or interpreter. Besides, the adage "if 
it ain't broke, don't fix it" is also followed a lot, with justifiable 
reasons. "Hey boss, let's move to version X+1 of this compiler/interpreter 
cause it has these new features," is a tough sell anyplace that has a 
budget and other higher priorities.

Another important criteria in commercial environments is what versions are 
supported commercially. The "core competencies" of many businesses is in 
the specific applications they write, not compiler or interepter 
development. And open source can't change that limitation. Business people 
believe in contractual obligations and monetary incentives as their 
insurance that support will be there when they need it.


From arthur.siegel at rsmi.com  Mon Apr 22 17:28:06 2002
From: arthur.siegel at rsmi.com (arthur.siegel at rsmi.com)
Date: Mon, 22 Apr 2002 17:28:06 -0400
Subject: Science And Math Was: Python's Lisp heritage
Message-ID: <OF51572BC3.924AFD3F-ON85256BA3.0072F219@RSMi.com>

>>>Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
>>>would-be applications to the natural sciences. I think it was P. Erdos
>>>who remarked once that all natural science is conceivably finite - e.g.
>>>there would be a time when there would be nothing more to learn about
>>>the universe - but mathematics is literally infinite.
>
>>I'm not so sure about the latter assertion, and I'm *definitely* not
>>certain about the assertion that natural science is finite.

>The first assertion by Erdos is indeed problematic. The second is not.

My sense, intution is of mathemetics as infinite only in the sense of allowing
infinite homologous descriptions of/perspectives on  the One Thing.

The perspectives are invented.

The Thing is not.

The modern assumption that increased abstractness brings better perspective
perhaps reaches a point not only of diminishing, but negative returns.

But as the opinion of someone who finds the more abstract mathematical
realms beyond his reach, maybe this represents a Philosophy of Sour Grapes.

Art






From jwickard**nospam** at litriusgroup.com  Tue Apr  2 10:16:08 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Tue, 02 Apr 2002 15:16:08 GMT
Subject: Quick return question
References: <0Ujq8.18$c6.7717@typhoon.mn.ipsvc.net>
Message-ID: <YWjq8.19$c6.7993@typhoon.mn.ipsvc.net>

"joel8bit" <jwickard**nospam**@litriusgroup.com> wrote in message
news:0Ujq8.18$c6.7717 at typhoon.mn.ipsvc.net...
> Hello all:
> given the function:
>
> def adder(**args):
>     for key in args.keys():
>         print args[key]
>
> print adder(fName="sam", lName="scott", mName="wayne")
>
> I get:
>
> sam
> scott
> wayne
> None
>
> Why am I getting None as well, because there's no return value for the
> function?
>
>

Disregard I understand now,
I told it to print the function call and there was no return value.




From tdelaney at avaya.com  Mon Apr 22 22:29:47 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 23 Apr 2002 12:29:47 +1000
Subject: flatten tuple
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A28A@natasha.auslabs.avaya.com>

> From: John Hunter [mailto:jdhunter at nitace.bsd.uchicago.edu]
> 
> >>>>> "Delaney," == Delaney, Timothy <tdelaney at avaya.com> writes:
> 
>     >> I want to be able to create a new tuple, which is the expanded
>     >> and flattened version x, so that if an element of x is a key of
>     >> m, then the new list has that value instead of the key value .
>     >> In the example
> 
>     Delaney> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/121294
>     Delaney> 
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/118845
> 
> Thanks for the tips -- these look helpful.  
> 
> It appears that I can not use the yield features of generators with
> 2.1.3.  I tried 'from __future__ import generators' but still got a
> Syntax Error on each of the yield lines.  Is this feature only
> supported in the 2.2 line, even with the __future__ ?

That's correct - the other recipe though should help you.

Tim Delaney




From logiplexsoftware at earthlink.net  Wed Apr 24 16:59:41 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 24 Apr 2002 13:59:41 -0700
Subject: Quick: Format Specs in %
In-Reply-To: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
References: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
Message-ID: <20020424135941.7ede5497.logiplexsoftware@earthlink.net>

On 24 Apr 2002 20:41:14 GMT
Philip Swartzleonard wrote:

> 
> '%2f' % some_number pads the number with 2 blank spaces
> '%.2f' % some_number fixes the number to 2 decimal places
> 
> the problem:
> '%2.2f' % some_number only does the second of those two things. I need 
> both. I tried looking up the % operator in the index of the language 
> reference, but couldn't find it.

The first digit in %2.2f specifies how wide the field is, not how many spaces
to pad with.

>>> f
3.1415926535897931
>>> print "%2f" % f
3.141593
>>> print "%.2f" % f
3.14
>>> print "%10.2f" % f
      3.14
^^^^^^  <- 6 spaces + 4 characters in the number = 10 

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From op73418 at mail.telepac.pt  Mon Apr 22 07:57:38 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 12:57:38 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com>
Message-ID: <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>

Just some clarifications.

On Mon, 22 Apr 2002 06:10:02 GMT, Tim Daneliuk <tundra at tundraware.com>
wrote:

>Carl Banks wrote:
>> 
><SNIP>
>
>> 
>> Sorry, but I don't agree that algorithms, or any mathematical concept,
>> is an invention of the human mind.  The algorithms have existed since
>> the beginning of time.  The quicksort had the property of requiring ~
>> n^2 comparisons in the worst case before it had ever been used or
>> applied.  We merely discovered this algorithm; we didn't create it.
>> 
>> So I say mathematics, including the study of algorithms, is a natural
>> science.  And, for now, that's all I have to say about it.
>
>I disagree strongly (but am willing to be convinced otherwise).  All of
>mathematics is a formal construct of the human mind created with the
>intent of removing ambiguity and enhancing our ability to describe what
>we *think* and what we *observe*.  We can (and do) create all kinds of new
>calculi that are problem- or domain-specific.  That is, we *invent*
>mathematics to suit our needs - mathematics is not one-for-one
>correspondent to the natural universe.

You should understand that this is a philosophical viewpoint. The
mathematical field is roughly divided in the Platonist field, those who
believe that mathematical objects exist somewhere out there in space (to
quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
field of those who believe mathematics is essentialy a game we play on
paper with symbols with no "real" meaning attached to it besides the
convenience in describing natural phenomena.

>
>e.g., Calling the upper-bound of Quicksort N^2 is merely one way of looking at 
>the time complexity of this "idea".  There is not some hardwired constant
>in the universe that declares heapsort to be O*(N^2) - this is merely an
>artifact of the calculus used for Big-O analysis.
>
>The reason I take this position is pretty simple.  We are able to construct
>mathematical systems which describe intellectual abstractions for which
>there is no natural analog.   That is, mathematics can embrace way more than
>just the natural world and its workings.  (Theoretical mathematicians are famous
>for their contempt of "mere" Applied mathematicians whom they view as operating in a
>very limited, mechanical intellectual domain.)  For instance, even the theoretical

You should be careful when you say this. There is *virtually no*
mathematical field that has not been applied.

>physicists only need about 5 dimensions to describe Life, The Universe, and
>Everything, but mathematicians routinely work in arbitrary "n-space" for which
>there is not physical/natural correspondent.  Similarly, a mathematician
>can introduce ideas like an "Incompleteness Theorem" which is a meta-mathematical
>commentary on mathematics itself - something far beyond the domain and range of
>mere natural science.

This is plainly not true. String theory models live in 10 and 24
dimensions. You would not want to disparage the biggest fad in current
theoretical physics, would you?

Even so let me reiterate, that almost everything that mathematicians
have concocted is currently used in theoretical physics. To give you
just an example, my own, I use homotopy and higher order category
theory.

Even more when mathematics advances by theoretical physics promptings. I
could give you numerous examples, but just remind you that quantum
mechanics contributed a LOT (the celebrated Von Neumann, for example) to
the growth of abstract mathematics (functional analysis, Hilbert spaces,
operator algebras, etc.). Some of the more active work in mathematics
today is also prompted by theoretical physics.

Somehow (I do not intend to make this more precise now) new mathematical
formalisms and the discover of new laws in the Universe go hand in hand.

>
>I'm not choosing sides here on the math vs. science debate - both are important
>artifacts of the human intellect.  I merely take umbrage with the notion that
>mathematics is innately wired into the universe somehow.  

See above to understand that what you are saying is really not that
obvious.

>
>
>------------------------------------------------------------------------------
>Tim Daneliuk
>tundra at tundraware.com

best regards,
Gon?alo Rodrigues



From sholden at holdenweb.com  Mon Apr  1 18:05:14 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 18:05:14 -0500
Subject: long integers
References: <mailman.1017540760.21438.python-list@python.org>
Message-ID: <HJ5q8.118538$Ou.59702@atlpnn01.usenetserver.com>

"Michael Hall" <olc at ninti.com> wrote in message
news:mailman.1017540760.21438.python-list at python.org...
>
> G'day:
>
> I'm using Python 1.5.2 and am having trouble with long integers when I
> pull data out of a MySQL database using MySQLdb. When I retrieve a
> numeric value from a field with an `int' data type, the value has an `L'
> stuck on the end.
>
> In one script, I worked around this by stripping the L off with:
>
> var = numeric_id # ("2L")
> var = var[:-1]
> print var
> >> 2
>
Clearly the value in var isn't coming from a numeric database field, because
if you tried this with a long integer you would be told it is an
unsubscriptable object. What you have is, I suspect, a string.

> In this case, numeric_id was passed through a hidden field in an HTML

HTML form values are *always* passed as strings. The MySQLdb module will
obligingly store such a value in an integer database column, since it's
smart enough to do the conversion (and to complain if the string isn't a
valid integer).

> form. But this doesn't work in another script for some reason, where
> the value is passed straight from a database query. I tried
> converting the data type from number to string, but that doesn't work
> either:
>
> var = result[0] # ("2L")
> var = 'var[:-1]'
> print var
> >> var[:-1]
>
If you had removed the single quotes around the right-hand side of the
expression you would probably have seen the "unsubscriptable value"
exception I mentioned above. As it is you are just setting var to the
eight-character string "var[:-1]", which won't help you.

> So now I'm confused. I don't yet know enough Python to know what to try
> now. Is there some easy way around this problem? Can this be fixed in
> MySQL by changing or modifying the data type, or should I upgrade to a
> later version of Python?

The bottom line is that you should only see this behavior when printing
these values out, because the print statement invokes the __repr__() method
of the objects it is called on to print.  Ordinary string representations of
long integers don't (nowadays) include the trailing "L". Explicit string
conversion is all you need, since repr(str(x)) == str(x):

>>> str(1L)
'1'
>>> repr(1L)
'1L'

Hope this helps. Python database usage is pretty easy, and MySQL is
extremely good value for money (particulary the freeware versions).

regard=s
 Steve







From max at alcyone.com  Tue Apr 30 13:54:00 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 30 Apr 2002 10:54:00 -0700
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com> <xdur8kx1i1i.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <3CCEDA38.2559BE98@alcyone.com>

Ralf Juengling wrote:

> Thanks, I wasn't aware of these (looked for them in the type module).

I agree, operator seems a strange place for them.

> *BUT*, these (eg. isSequenceType) work only for C extension types,
> right?

No, they work with any object:

>>> import operator
>>> operator.isSequenceType([1, 2, 3])
1
>>> class MyList(list): pass
... 
>>> operator.isSequenceType(MyList())
1

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From ods at fep.ru  Thu Apr 11 04:38:54 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 11 Apr 2002 12:38:54 +0400 (MSD)
Subject: PEP 282: A Logging System
In-Reply-To: <20020411082017.GB1261@andy.tynant.ftech.net>
Message-ID: <Pine.LNX.4.44L.0204111235590.13708-100000@ods.pp.ru>

On Thu, 11 Apr 2002, Andy Gimblett wrote:

AG> On Wed, Apr 10, 2002 at 06:54:48PM +0100, Vinay Sajip wrote:
AG>
AG> > The logging module is not designed to use threading, or
AG> for thread safety.
AG>
AG> Really?  It's _not_ thread safe?  Oh.  Will it be, ever?

Thread safety of logging module is very important. If it won't,
every module thread safe itself will become unsafe when using
logging.





From nospam-readonly at getsoft.com  Sun Apr 21 22:33:25 2002
From: nospam-readonly at getsoft.com (Neville Franks)
Date: Mon, 22 Apr 2002 12:33:25 +1000
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net>
Message-ID: <3cc3765f$0$15478$afc38c87@news.optusnet.com.au>

Ok thanks. I've installed 2.2 and ndiff is available but I'm still confused.
I thought that Tim's original ndiff compared files, but the one in difflib
compares lists of strings. Does this mean I somehow need to read the files
into lists of strings and then call ndiff. A short example of how to use
ndiff to compare two files would be greatly appreciated.


"David Goodger" <goodger at users.sourceforge.net> wrote in message
news:B8E8E4D6.21E42%goodger at users.sourceforge.net...
> Neville Franks wrote:
> > Is there some one to run this in 2.1.2 or should I just install 2.2?
>
> Unless you have a reason to stay with 2.1, you might as well install 2.2.1
> (the latest).  If you don't want to, then as I said:
>
> >> you can get a Python 2.0/2.1-compatible version from
> >> http://docutils.sourceforge.net/test/difflib.py
>
> --
> David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
>   - Python Docutils: http://docutils.sourceforge.net/
>   - The Go Tools Project: http://gotools.sourceforge.net/
>




From gimbo at ftech.net  Mon Apr 22 15:04:57 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Mon, 22 Apr 2002 20:04:57 +0100
Subject: Interesting topics
In-Reply-To: <20020422165134.GD25778@andy.tynant.ftech.net>
References: <mailman.1019468028.28617.python-list@python.org> <Xns91F85897F9A6DRASXnewsDFE1@130.133.1.4> <20020422165134.GD25778@andy.tynant.ftech.net>
Message-ID: <20020422190456.GG25778@andy.tynant.ftech.net>

On Mon, Apr 22, 2002 at 05:51:35PM +0100, Andy Gimblett wrote:

> ... in mutt I find Ctrl-D (delete entire thread) and Meta-D (delete
> subthread) _indispensable_ around here.  :-)

And they're getting lots of usage on the maths thread right now...

;-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From tejarex at yahoo.com  Tue Apr  2 19:31:18 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 03 Apr 2002 00:31:18 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>              <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com>
Message-ID: <q3sq8.535469$pN4.38175074@bin8.nnrp.aus1.giganews.com>

 Guido van Rossum wrote:

> I don't see the introduction of the bool type as painful at all.
> I find getting it accepted painful though. :-(

Part of the concern is 'why bother? or 'would't adding two builtin
names do as much good?'.

Another large part is the fear that you might someday be persuaded to
cripple the new type.  You made this fear seem reasonable by tossing
it out as a real possibility and by half-conceding the argument behind
such a move:

>         4) Should we strive to eliminate non-Boolean operations on
bools
>            in the future, through suitable warnings, so that e.g.
True+1
>            would eventually (e.g. in Python 3000 be illegal).

(Your follow-on sentence seems like a pretty weak endorsement of the
status quo:
>
Personally,
>            I think we shouldn't; 28+isleap(y) seems totally
reasonable to me.
)

My point about the name 'bool' is that this question and its
justification are at least partly an artifact of this proposed name.
If the sometimes non-boolean-acting values were called something else,
such as forb*, complaints that "Forbs do things 'non-Boolean!' " could
be met with "So what?" and "Yes, we know, that is why we did not call
them bools."

* FORced Binary choice.

Terry J. Reedy





From alwyn at alwyn.demon.co.uk  Thu Apr 18 18:57:56 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Thu, 18 Apr 2002 22:57:56 GMT
Subject: PyQT, how to compile SIP on MAC? (long)
References: <3C9A62C2.ABA9AE5D@gmx.de> <mailman.1016790154.12480.python-list@python.org> <de404184.0204181134.6bd4bc2c@posting.google.com>
Message-ID: <alwyn-20BEEF.23575218042002@news-text.blueyonder.co.uk>

In article <de404184.0204181134.6bd4bc2c at posting.google.com>,
 barronsnyder at yahoo.com (Barron) wrote:
> ...
> /usr/bin/libtool: internal link edit command failed
> make[1]: *** [libsip.la] Error 1
> make: *** [check-recursive] Error 1
> 
> Do these errors prevent me from doing a "make install"?

Um, have you tried the compiler's '-flat_namespace' option? That might 
eliminate the duplicate symbols for you. 


Alwyn


From tim.one at comcast.net  Sat Apr 13 15:42:41 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 15:42:41 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <ZkjnuIABLAu8EwYC@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOIONAA.tim.one@comcast.net>

[Robin Becker]
> Anyone else had a problem un-installing 2.2 for win32? When I tried to
> un-install 2.2 prior to installing 2.2.1 I got an error saying it had
> already been un-installed.
>
> Tried the link in start menu and it said no unwise.exe was there.

Yes, I've seen this happen several times, but only under Win9X, not Win2K.
Which flavor of Windows are you using?   So far you're the only other one to
report it.  I can't account for it:  the installer definitely creates
UNWISE.EXE in the right place (the root of your Python tree), but sometimes
"it goes missing" later.  Exactly when and why remains a mystery to me.  It
may be related to unplanned <ahem> system crashes.





From gerson.kurz at t-online.de  Mon Apr 29 11:02:42 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Mon, 29 Apr 2002 15:02:42 GMT
Subject: Lambda Scoping
Message-ID: <3ccd5f99.5559046@news.t-online.de>

The following code
...
funcs = []
for instance in range(10):
    funcs.append( lambda: instance )
print map(apply,funcs)

funcs = []
for instance in range(10):
    funcs.append( lambda instance=instance: instance )
print map(apply,funcs)
...
prints

[9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

which I sort-of understand (I figure, in "lambda: instance" instance
is a lazy expression; "lambda instance=instance: instance" the
expression itself - "instance" - is a lazy expression, too, but the
argument list is not.) However, can anybody please explain this more
detailed? 


From westernsam at hotmail.com  Fri Apr 12 07:56:27 2002
From: westernsam at hotmail.com (sam)
Date: 12 Apr 2002 04:56:27 -0700
Subject: ASCII to HEX conversion
References: <34e98ec2.0204111415.504d246f@posting.google.com> <Xns91EE3AF2DC0cliechtigmxnet@62.2.16.82>
Message-ID: <292c8da4.0204120356.3709c4f1@posting.google.com>

The following code will give binary from decimal. (N.B. I cannot claim
credit for this as I didn't write it and just found it somewhere)

import math, operator

def binary_repr(self, number, max_length = 32):
        assert number < 2L << max_length
        shifts = map (operator.rshift, max_length * [number], range
(max_length - 1, -1, -1))
        digits = map (operator.mod, shifts, max_length * [2])
        if not digits.count (1): return "00000000"
        digits = digits [digits.index (1):]
        returnString = string.join (map (repr, digits), '')
        while (len(returnString) < 8):
            returnString = "0%s" % returnString
        return returnString


From tim.one at comcast.net  Mon Apr 15 12:31:19 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 15 Apr 2002 12:31:19 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <3CBA64AD.AE4C563D@cosc.canterbury.ac.nz>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEFMOOAA.tim.one@comcast.net>

[Greg Ewing]
> The way I see it, if you're doing bit twiddling
> using unified ints/longs, you should always be
> working with *positive* integers, in which case
> the hex representation will always be what
> you want.

Viewing longs as characteristic bit vectors, they model finite and co-finite
sets.  Only finite sets "look positive"; co-finite sets "look negative".  It
doesn't take more than the unary ~ operator to go from one kind to the
other, i.e. if x is a bit set represented as a long, ~x is its complement.





From peter at engcorp.com  Tue Apr  9 00:57:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 00:57:54 -0400
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>
Message-ID: <3CB274D2.A7554C79@engcorp.com>

Chris Spencer wrote:
> 
> 4.  Did I mention that the documentation is both incomplete and odious?  I
> should not have to constantly do a "find in files" to find out what parameters a
> damned method call takes because the way wxPython implemented a method is
> different than the wxWindows way and someone forgot to make a note of it in the
> rubbish passed off as a "manual".

Although I share your pain, I don't complain much about it since I'm
well aware of the effort required to put together something like this
and maintain it, and I don't believe the authors would appreciate
my condemnation as much as they would my assistance.

Maybe you could contribute some of what you learned back to the project
and offer to help improve the documentation?

-Peter


From cjw at sympatico.ca  Wed Apr 10 16:04:45 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 10 Apr 2002 16:04:45 -0400
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com>
Message-ID: <3CB49ADD.89B242A4@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020410/5c401798/attachment.html>

From peter at engcorp.com  Tue Apr  2 20:15:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 20:15:40 -0500
Subject: style Q: global vars in CAPS?
References: <slrnaak4m7.u14.ak@ak.silmarill.org>
Message-ID: <3CAA57BC.F5144DF2@engcorp.com>

Andrei Kulakov wrote:
> 
> Hello,
> 
> What do you think about global variables spelled in caps? Is that what
> most people do?

More commonly ALL CAPS is a convention that signifies "constant",
not "global".  If the globals in question are constants and not
variables, I'd say "yes".

If they're really variables, a different convention I've seen
more often is to prefix "g_" or just "g", as in gSomeGlobal.

And of course you should minimize the use of globals as much
as possible, since they can lead to all kinds of difficulties
if you are not careful.  (Although they are sometimes necessary,
as discussion in the XP mailing list has recently covered.)

-Peter


From jrup at mail.com  Sun Apr 14 17:06:54 2002
From: jrup at mail.com (jrup)
Date: Sun, 14 Apr 2002 21:06:54 GMT
Subject: Specifying a time from previous day
References: <03ju8.28136$%s3.9832152@typhoon.ne.ipsvc.net> <pRku8.15554$b62.440441@news1.tin.it>
Message-ID: <Obmu8.28168$%s3.9942967@typhoon.ne.ipsvc.net>

Thank you both for your replies.

I worked on this some more after I'd posted and it turns out I'd ended up
doing basically what Alex suggested. I'd wondered about DST, so I'll work on
redoing it using UCT. I'd pretty much had the file stats part worked out.

Here's what I did, it's not as elegant, but I think it's the same idea:

# current local time in seconds
time.time()

# time 24 hours prior to now in seconds
yesterday = time.time() - 24 * 60 * 60

# put time 24 hours ago into a tuple
YesterdayLocal = time.localtime(yesterday)

# reassign the tuple elements to another tuple so they can be changed
YesterdayTuple = YesterdayLocal[0:]

# then change the hour and everything after it using the info from the
previous day
YesterdayFourPm = (YesterdayTuple[0], YesterdayTuple[1], YesterdayTuple[2],
16, 0, 0, 0, 0, 0)

# change the new yesterday 4:00pm time into seconds to be the constant
# against which the times of files in the log directory will be compared
YesterdayFourPmInSecs = time.mktime(YesterdayFourPm)

# then change the file creation times into seconds and compare with the
# new constant

Aside from the wordiness, and lack of UTC, is this the right idea?
Thanks again,
Jim


"Alex Martelli" <aleax at aleax.it> wrote in message
news:pRku8.15554$b62.440441 at news1.tin.it...
> jrup wrote:
>
> > Hi,
> >
> > This is my first post to this group and I am new to python. I have some
> > experience with perl and vbscript for sysadmin-type scripting, but am
not
> > a programmer.
> >
> > I need to search some log files, but only if they were created after
> > 4:00pm the previous day.
>
> >>> import time
> >>> time.time()
> 1018811834.620291
>
> this gives us the time in seconds since some arbitrary starting point
> ("the epoch").  24 hours ago was clearly 24*60*60 seconds ago, i.e.:
>
> >>> time.time()-24*60*60
> 1018725558.410575
>
> Given this representation for time, we can obtain one in terms of
> year-month-day-hour-minute-second-etc (9-items tuple) -- say
> in local time (GMT, aka UTC, is the alternative):
>
> >>> yesterday = time.localtime(time.time()-24*60*60)
> >>> yesterday
> (2002, 4, 13, 21, 20, 29, 5, 103, 1)
> >>>
>
> As you want a given time, you'll only get the first three items of
> this (year, month, day), then your specified time (16, 0, 0), then
> 2 zeros, and finally the same last item as yesterday (DST flag: this will
> break by 1 hour when DST has just started or finished -- if you could have
> your specs in UTC...); time.mktime can rebuild the "time in seconds
> since the epoch" from this:
>
> >>> threshold = time.mktime(yesterday[:3]+(16,0,0)+(0,0,yesterday[-1]))
> >>> threshold
> 1018706400.0
>
> When you examine a file, os.stat gives you a lot of data about it:
>
> >>> os.stat('/home/alex/.bashrc')
> (33188, 32453L, 776L, 1, 501, 501, 403L, 1018648057, 1018267187,
1018267187)
>
> The last three numbers are times in the seconds-since-the-epoch
> form -- last access, modification, inode change.  It appears that what
> you want is to look at files such that os.stat(thefilename)[-2] >
threshold,
> where threshold is computed as above.
>
>
> I hope this information is enough for you to complete your program.
>
>
> Alex
>




From tdelaney at avaya.com  Wed Apr  3 18:08:47 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 4 Apr 2002 09:08:47 +1000 
Subject: [Python-Dev] PEP 285: Adding a bool type
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A24B@natasha.auslabs.avaya.com>

> From: Guido van Rossum [mailto:guido at python.org]
>
> The first round of discussion (on python-dev) brought up some
> contentious issues where people disagreed strongly with particular
> details proposed in the PEP, and I put those contentious issues at the
> top of the review list.  If you go to the latest version of the PEP on
> the web (http://www.python.org/peps/pep-0285.html) you'll see that
> I've added some other issues that were brought up in the more recent
> discussion.

I have only one other issue - the PEP does not deal with the case of sending
data to an older version of Python (which I believe Alex Martelli brought
up, though I could be wrong).

Perhaps the section that states int(bool(x)) (resolved issues) should
include that the structure should be used for backwards-compatibility with
previous versions of Python.

Tim Delaney




From bgrotanLOVELYSPAM at itea.ntnu.no  Thu Apr 18 03:32:28 2002
From: bgrotanLOVELYSPAM at itea.ntnu.no (=?iso-8859-1?Q?Bj=F8rn?= Ove =?iso-8859-1?Q?Gr=F8tan?=)
Date: Thu, 18 Apr 2002 09:32:28 +0200
Subject: chopping csv-files (newbie)
Message-ID: <3CBE768C.844C90FF@itea.ntnu.no>

I have a csv-file.. delimited by ;

example-line from the csv-file
"90383";"ooooxWOW   90383;";"Some discription";
"99875";"ooooxSomede90383;0000xNewli99875;";"New line-descriptor";

In this file there is 3 fields extracted from a database.
The second field, can vary som 1 to 4 fields separated by ; but not with
"

What I need to do is to build an array after opening a given file
with the information from the second and third fields.

E.g:

90383	Some discription
99875	New line-descriptor, Some discription

You see where I'm going at?

the second record in this file has a reference to the first record in
its second field.


I have no choice to rebuild the database for my needs since I only have
read-access to it and have nothing to do with the database itself.

I also want to save the records in the array (hash-table?) to a file 

Could someone please help me with an example on how to do this?
Saving the hash-table to a file I can manage.. but the above part is a
bit tricky for me since I don't know enough python..

Best regards

Bj?rn Ove Gr?tan


From martin at v.loewis.de  Mon Apr 15 02:20:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:20:52 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com>
Message-ID: <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> When "hex(x)" creates a string from a long, it should not have a '-'
> sign if the long is negative, but rather should print out the
> hexadecimal representation of the number in using the number of
> bytes that the object currently occupies.

That can't possibly work. That way, hex(-1L) would be "1", or perhaps
"0001" (since it will use 2 bytes to represent the value). This is
indistinguishable from hex(1L) - something that is clearly
undesirable.

> If we are printing a long as a decimal number, then we DO add the
> '-' sign.

So why should hexadecimal be different? It is just the base-16
representation of the number, nothing more.

Regards,
Martin


From les_ander at yahoo.com  Sun Apr 28 15:42:23 2002
From: les_ander at yahoo.com (les ander)
Date: 28 Apr 2002 12:42:23 -0700
Subject: email images
Message-ID: <a2972632.0204281142.30eb1e8@posting.google.com>

Hi,
to avoid being a victim of spam bots, i would like to have a image 
representation of my email to post on my web site.

is there a way to convert string to images using python?


From neal at metaslash.com  Thu Apr 11 23:09:08 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 11 Apr 2002 23:09:08 -0400
Subject: Built in file copy utility
References: <Ozrt8.4709$c6.399223@typhoon.mn.ipsvc.net>
Message-ID: <3CB64FD4.62FA1AAA@metaslash.com>

joel8bit wrote:
> 
> I just wanted to check and see if there was a built in module to copy files
> from one directory to another.

>>> import shutil
>>> print dir(shutil)  
['__all__', '__builtins__', '__doc__', '__file__', '__name__', '_build_cmdtuple', 'copy', 'copy2', 'copyfile',
'copyfileobj', 'copymode', 'copystat', 'copytree', 'os', 'rmtree', 'stat', 'sys']
>>> print shutil.copyfile.__doc__
Copy data from src to dst

Neal


From phr-n2002a at nightsong.com  Tue Apr 30 04:07:12 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 30 Apr 2002 01:07:12 -0700
Subject: Cookie encoding
References: <mailman.1020145372.5420.python-list@python.org>
Message-ID: <7xd6whehjj.fsf@ruckus.brouhaha.com>

Terry Hancock <hancock at anansispaceworks.com> writes:
> So I need to encode the result into something safer. I've
> tried the uu.encode, base64.encode, etc. but I'm just
> taking shots in the dark (and these didn't work) -- I
> figure there has to be a standard answer to this
> problem.  

Try binascii.hexlify.  The result should be safe for just
about anything.

> There is a "Cookie" module, which looks
> interesting, but it seems to want to make the whole
> HTTP set-cookie line, not just the legal value.

WARNING!  DANGER WILL ROBINSON!  The default ("smart") cookie class
defined by that module uses pickle to encode arbitrary Python objects
into cookies.  This creates a security hole at the server side since
attackers can create cookies that instantiate class instances with
malicious data passed to the class initializers.  Don't use smart cookies.


From h_schneider at marketmix.com  Tue Apr  9 04:10:56 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Tue, 9 Apr 2002 10:10:56 +0200
Subject: Creating a DB app with mySql and Windows GUI
References: <a8rhml$1id$03$1@news.t-online.com> <wrps8.20370$Yt.763733@read1.cgocable.net>
Message-ID: <a8u7cb$590$06$1@news.t-online.com>

Hi,

I can't recommend BOA at the moment, since it is in a too early state. But
it will become a great
piece of software ... Your first choice for an IDE should be WingIDE. It's
very robust and it's a real pleasure to work with.
For GUI design/wxWindows you should consider wxDesigner. The handling might
look strange, if you are used to Visual Basice etc., but it does its job
very well ...

Sources:
http://archaeopteryx.com/
http://www.roebling.de/

All the best,
Harald


> As for the GUI builder and IDE, I don't really know if I can help you. I
> used a wide variety of editors such as EditPlus, Vim, Ultraedit, Glimmer,
> etc. I rarely use IDE's that have debugging built in. As for building
> GUI's, I use wxPython. I hear that Boa Constructor is great for easily
> building wxPython GUI's, but I have never used it.
>
> --
> - Mark Nenadov (http://www.freelance-developer.com)




From reply at in_newsgroup.tks  Thu Apr  4 14:11:19 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 4 Apr 2002 19:11:19 GMT
Subject: Newbie's 1st Web App decisions!
Message-ID: <a8i8gn$rbe$1@hfc.pacific.net.hk>

I want to learn Python for use in creating web apps and possibly to create 
stand-alone apps with wxPython.

To help you Python scholars understand my limited programming/scripting 
background I've;
1) Created apps within the following environments - dBASE IV, Lotus 
Approach 2 & MS Access (an old version) with limited use of their related 
languages and no standalone exes and;
2) Created/coded HTML4.01+CSS2 static freebie websites which included pages 
of formatted data created by merging HTML templates with CSV databases.

Mainly my background is in IT/Network mgmt.

I'm a quick learner (self taught from books and web-based info. and a Ye 
Olde college course ions ago!) and would like to get into Python web app 
development as quickly as possible, even though I have not yet written a 
significant line of Python code :)

There seem to be various projects which provide web app 
servers/components/development frameworks. Quite a few seem to overlap ;-) 
Here are one's I've looked at:
Slither http://slither.toolsofcomputing.com/
Webware for Python http://webware.sourceforge.net/
Skunkweb http://skunkweb.sourceforge.net/about.html
OSE http://ose.sourceforge.net/

There are also various projects which 'integrate' Python with 
Apache/AOLserver:
Mod_Python
Mod_Snake
PyApache
Cheetah
PyWX

It look like I'm spoilt for choice and it's hard to decide where I should 
direct my efforts.

As a side note, some of the comments made by Paul Boddie ( 
http://mail.python.org/pipermail/python-list/2001-September/064133.html ) 
seem justified.

At the moment I'm considering using either Webware or Slither (not an easy 
choice) and/or Mod_Snake (instead of Mod_Python - but hey, I'm a newbie 
so...).

As far as books are concerned (a must) I'm thinking of buying
Web Programming in Python (this is written partly(?) by the folks from the 
Slither project).
http://www.amazon.com/exec/obidos/ASIN/0130410659/qid=1017861581/sr=1-
3/ref=sr_1_3/104-5555059-2897528

and possibly
Learn to Program Using Python (in case I'm not as quick a learner as I 
thought, hehe)
http://www.amazon.com/exec/obidos/ASIN/0201709384/103-9771277-6190237

Webware seems to be one of the more respected projects and I get the 
impression that it improves the productivity/time to completion for people 
using it.

On the surface, Slither seems to cover more web app server bases than 
Webware. I am interested in persistance with backend db's but am more drawn 
to using Slither with mod_snake than mod_python. So far though, I have not 
found any independant info on Slither (and nothing at all on Slither with 
mod_snake) so input from users with experience of using Slither and Webare 
would be most welcomed.

What I would be most appreciative of, is your helpful suggestions on how to 
start in Python Web app development and on relevant books.

Finally (at last you cry!), although Zope is one way to go, as this is for 
me primarily a learning experience, I would prefer to stay closer to Python 
and 'perhaps' enjoy better web app performance. Zope does seem to be better 
suited for creating CMS apps but I am not creating a CMS app :)

Bows and leaves the stage,
Scrumpy :)


From max at alcyone.com  Mon Apr 22 14:33:46 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 22 Apr 2002 11:33:46 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com> <3CC449BF.8030807@alcyone.com> <aa1jt0$hgs$1@panix1.panix.com>
Message-ID: <3CC4578A.80000@alcyone.com>

Aahz wrote:

> In article <3CC449BF.8030807 at alcyone.com>,
> Erik Max Francis  <max at alcyone.com> wrote:
> 
>>Here's a Web page outlining some of the differences between Scheme and 
>>Common Lisp:
>>
>>	http://dept-info.labri.u-bordeaux.fr/~strandh/Teaching/MT
> 
> 
> 404

Weird, that URL got truncated for some reason.

http://dept-info.labri.u-bordeaux.fr/~strandh/Teaching/MTP/Common/Strandh-Tutorial/diff-scheme.html

-- 
  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
  A new, virtual planet, every day.



From merman at freenet.de  Mon Apr 15 16:26:56 2002
From: merman at freenet.de (merman)
Date: Mon, 15 Apr 2002 22:26:56 +0200
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org>
Message-ID: <3cbb355a$0$166$9b622d9e@news.freenet.de>

"Brian Quinlan" <brian at sweetapp.com> schrieb im Newsbeitrag
news:mailman.1018890724.31468.python-list at python.org...
> Merman wrote:
> > I have a c++ -class; I can compile this class successful as
> standalone-app.
> > No problem.
> > Now I want to work with swig.
> >
> > There are three files: Tree.cpp, Tree.h, Tree.i
> >
> > first step:
> > swig -python -c++ -o Tree_wrap.cpp Tree.i (output is Tree_wrap.cpp -
> it's
> > ok)
> >
> > second step:
> > g++ -c Tree.cpp Tree_wrap.cpp -IC:\python22\include (also ok - output
> are
> > Tree.o and Tree_wrap.o)
> >
> > third step - *build*:
> > g++ Tree.o Tree_wrap.o -shared -o Tree.dll (this line throws lots of
> > failures - whats wrong?)
>
> Where need to link with Pythonxx.lib. But instead of this manual crap,
> you should just write a simple distutils setup script.
>
> Cheers,
> Brian

Ok Brian, and how can I write a disutils setup script? I don't know that!
Read the subject Brian: newbie-question!!!

o-o

thomas





From aahz at pythoncraft.com  Tue Apr 16 00:29:06 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 00:29:06 -0400
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <3CBB8ABB.57BAA189@engcorp.com>
Message-ID: <a9g9ai$a23$1@panix1.panix.com>

In article <3CBB8ABB.57BAA189 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>
>(For what it's worth, we've more or less outlawed any internal tools
>not written in Python as being a waste of the developer's time, and
>NONE of our utilities are "too slow".  (No, we're not quite as anal
>about it as that sounds, but you get the point.))

Where's your company based, again?                                          ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From phr-n2002a at nightsong.com  Tue Apr  2 22:06:29 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 19:06:29 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com> <uakogq1r0i0524@news.supernews.com>
Message-ID: <7xpu1h4h1m.fsf@ruckus.brouhaha.com>

"John Roth" <johnroth at ameritech.net> writes:
> I absolutely agree. For me, the entire point is providing
> a **standard** way of expressing **in the code** the
> notion that a particular variable only takes two values:
> True and False. This is such a common case that it deserves
> to be promoted to be a part of the language.

Why not ask for enumeration types then?


From nospam at bigfoot.com  Fri Apr 26 11:30:40 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 26 Apr 2002 17:30:40 +0200
Subject: Sending mail to an SMTP server that requires POP authentication ?
Message-ID: <aabrbv$2vg$1@norfair.nerim.net>

With this kind of server, only holders of a pop account can send mail
through SMTP using their POP credentials.

I did not find how to do this with smtplib.

Did someone succeed ? How ?

Thanks in advance.

--Gilles
glenfant AT bigfoot DOT com




From imbosol at vt.edu  Tue Apr 30 01:39:03 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 30 Apr 2002 01:39:03 -0400
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org> <200204292311.57363.bhan@andrew.cmu.edu>
Message-ID: <nlalaa.ko2.ln@127.0.0.1>

Benjamin Han wrote:
> On Monday 29 April 2002 10:02 pm, Cliff Wells wrote:
>> On 29 Apr 2002 18:54:12 -0700
>>
>> Fortepianissimo wrote:
>> > This question is so fundamental that I strongly suspect it's already
>> > answered - ig that's the case, please forgive this poor newbie, for he
>> > has done some work searching over Python FAQ and Google.
>> >
>> > Basically I want to have a subclass of list, like
>> >
>> > class MyList (list):
>> >     ...
>> >
>> > then I want to
>> >
>> > l=MyList()
>> > ... do stuff to fill l
>> > s=" ".join(l)
>>
>> s = " ".join([str(i) for i in l])
> 

> This surely looks nicer. I'm just wondering how far Python has fared
> in terms of this concept of OO? This example seems to show a very
> different character of OO in Python, in that __str__ is not
> automatically called, which seems rather counter-intuitive.

It may seem counterintuitive for someone accustomed to a language
using typecast operators.  But, despite appearances, __str__ isn't
typecast operator.  In fact, Python doesn't have typecast operators.
If you want a type changed, you have to do it by hand.

But even if __str__ was a typecast, it still wouldn't be called, since
Python doesn't have overloaded functions.  The interpreter has no way
of knowing that " ".join is expecting a list of strings.  The join
method could have, as a convenience, converted non-string items to
strings itself, but there is never any implicit conversion.

(Some of the built-in numberic types might have implicit conversion;
not sure of the details.)


But I don't see what all that has to do with OO, though.


> Without private members, and with seemingly complex ways of dealing
> with class members and methods - so what's the state of OO in
> Python? 

There are certainly some oddities in the object/class scheme.  On the
whole, I don't see much practical difference between C++ and Python
classes.  I use them pretty much the same way.

Concerning private members, I would say they're not the hallmark of OO
you might think they are.  C++ was my first OO language, and, perhaps
like you, I about freaked out when I saw a language that claimed to be
"object oriented" but didn't have private data members.




> Is it still rapidly evolving?

Well, they recently added this new-style class thingie.  I imagine
it'll be tuned up along the way, but I think it's the end of the
revolutionary changes for awhile.  Someone else who knows better than
I can answer.


> How much is an effort in
> converting Python into C++? Anyone would like to comment on some of
> these questions, or point to some article, FAQ (esp. helpful from a
> C++ programmer's point of view)?

Well, let me give you a little heads up on the differences you might
see.  Many C++ programmers are brought up to believe that C++ is
synonymous with OOP.  You seem to tend that way, with your idea that
implicit typecasting is an OO feature.  (If not, I apologize I seem
facetious.)

There's a whole world of OO stuff out there that doesn't do it the way
C++ does.  And that's not necessarily (and not probably) a bad thing.

Personally, I find C++'s OO to be the most constraining of all
languages I've tried, excepting Ada 95 (no interfaces, no MI, ugh).
Why?  Because you are forced to construct elaborate class hierarchies
to get the desired polymorphic behavior.

Imagine implementing a hierarchy of file-like classes.  File-like
objects have similarities, but there are a lot of options: input,
output, or both?  seekable or nonseekable?  buffered or nonbuffered?
lockable or nonlockable?  text or binary?  is it a file, a string, a
pipe, a socket, or something else?  are there any other exotic
options?

How do you implement all this in a class hierarchy?  Well, for
example, you could define a class such as
BufferedSeekableLockableInputBinaryFileWithIOCTL, which inherits from
classes such as InputFile, BufferedSeekableObject, ObjectWithIOCTL,
LockableObject, etc., which in turn inherit from deeper classes.
Then, your classes will implement only the methodd it actually needs,
and the compiler will make sure nothing lacking the appropriate method
can be called.

No one's going to do that, though.  Instead, they'll pick one or two
option (usually input/output and file/string) and build a heirarchy on
that.  Then, they'll have a member function such as isseekable to
determine whether you can call the seek method.

But isn't the point of OOP so that you can avoid this kind of thing?
Don't you want the compiler to keep track of whether a file-like
object is seekable or not?  Yes, you do.  You might not think you
don't, but that's probably only because C++ makes it really difficult
to do it right.


Python makes it easy to do it right.  You don't need to define all
those elaborate classes in Python, because, in Python, you can get
polymorphic behavior without inheritance.

This is, IMHO, THE BEST THING about Python.  You want a file-like
object?  No problem: just write a class whose objects look like a
file, and act like a file, and you have a file-like object.  Because
your class it doesn't inherit from anything, it doesn't have to
implement methods it doesn't need.  If seek makes sense for a
particular type, add it; if not, don't.  And if someone tries to call
seek on an unseekable file-like object, you set an exception.

Python frees you from the contraints of the class hierarchies.  You
can reserve inheritance for when you actually need to share code
between several types, and otherwise leave the class interfaces up to
the individual class.


What I'm trying to say here is dynamic typing is so much better and
FUN than static typing it isn't even funny.


I defintely suggest you try Python out and keep and open mind, and try
not to use inheritance unless your classes really need to share code.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From tim.one at comcast.net  Mon Apr  1 20:41:20 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 01 Apr 2002 20:41:20 -0500
Subject: PEP 285: Adding a bool type
In-Reply-To: <001601c1daa9$16d572a0$0334fea9@carol>
Message-ID: <LNBBLJKPBEHFEDALKOLCAENHOJAA.tim.one@comcast.net>

[Arthur Siegel]
> ...
> Efforts to further gloss Python's air of simplicity are going
> too, too far.

The PEP doesn't mention simplicity, except perhaps indirectly and partially
via its claim that the existence of a standard, distinguishable true/false
type could make life easier for a particular class of external library
(which has been confirmed by people maintaining such libraries).

> We've accomplished that Python can be taught while
> hiding the existence of numeric types.
>
> If the "teaching community" demands more of Python - I do suspect Guido
> faces pressures - let them find something better for their purposes.

AFAIK, this PEP hasn't even been mentioned on the Edu-SIG.  It certainly
didn't come from the teaching community; historically, Barry Warsaw has been
the biggest advocate of adding a true/false type, and most educators would
almost certainly object to the "impurity" of this proposal.  The little
Guido said about teaching in the PEP comes from his own experience
introducing to Python to others; that should be clear from the phrasing
alone:

    When showing people comparison operators etc. in the interactive
    shell, I think this is a bit ugly:

All of that part of the PEP was written in the first person.

> Of course they won't.
>
> Its called calling a bluff.

AFAICT, "they" aren't even aware of this PEP.  If they were, there would be
a lot more complaints about mixing bools with ints.




From aleax at aleax.it  Wed Apr 17 12:20:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 16:20:29 GMT
Subject: pre-PEP for optional 'pass'
References: <mailman.1019050054.11833.python-list@python.org>
Message-ID: <hhhv8.23713$b62.658633@news1.tin.it>

brueckd at tbye.com wrote:
        ...
> No, you were being chastised for promoting a horrible standard - that
> *all* classes and *all* methods should have a docstring. It's just plain
> wrong.

Even if a method is "so obvious as not to need a docstring" it's hardly
the end of the world to put an empty docstring there.  Few methods would
thus qualify IMHO -- just a few special methods whose semantics is entirely
Python-dictated, when there's absolutely nothing to say that's of interest.

This stance is partly based on my belief that boilerplate code should not
exist, of course -- if your style is to have classes full of methods named
getBlah whose body is "return blah", that would change things -- but I think
you'd be better advised to use properties for such things, instead:-).


Alex




From gerhard at bigfoot.de  Mon Apr 29 15:07:04 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 29 Apr 2002 19:07:04 GMT
Subject: Python sous UNIX
References: <aak1q3$fgl$1@wanadoo.fr>
Message-ID: <slrnacr6em.nil.gerhard@lilith.my-fqdn.de>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

infocap wrote in comp.lang.python:
> Bonjour ? tous,
> 
> Je travaille pour une entreprise d'?dition mais voil? que depuis quelques
> temps je m'int?resse beaucoup ? l'informatique et particuli?rement au
> langage de programmation PYTHON  sous UNIX et j'aimerai approfondir mes
> connaissances au travers de petits programmes int?ressants, alors, si
> quelqu'un peut satisfaire ma demande, je le remercie d'avance.

Pi?ces de Python:
http://sourceforge.net/snippet/browse.php?by=lang&lang=6

Ceux l? sont plus avanc?:
http://aspn.activestate.com/ASPN/Cookbook/Python

Et, naturellement, la librairie standard de Python qu'on peut trouver
? /usr/(local)/lib/python2.2.

Il y a aussi un group francais pour ce sujet: fr.comp.lang.python

Gerhard

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE8zZndeyJldYarQ8ARApngAJ9EfHt8iu9TRY/5UaM5B68kXv4GywCgj2Xa
hl8TsaQKuyVs8UiFlOiamR4=
=w/ke
-----END PGP SIGNATURE-----


From aleax at aleax.it  Tue Apr  9 12:53:51 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 16:53:51 GMT
Subject: Beginner's performance problem
References: <memo.20020409174302.1584B@a.a>
Message-ID: <z0Fs8.446$vF6.11027@news2.tin.it>

Mark Charsley wrote:

> For reasons best not gone into, I needed to correct the case of a whole
> bunch of SQL code. As such I created the little script below...
> 
> It reads in a source file, then for each line it does a case-insensitive
> search for each TableName and ColumnName (as contained in the "names"
> collection), checks that it's not just matching a substring, and then
> corrects the case of the word if necessary.


Something like (warning, untested):


substitutionDictionary = {}
for correctName in names:
    substitutionDictionary[correctName.upper()] = correctName

def substitutor(matchObject):
    return substitutionDictionary[matchObject.group().upper()]

import re
correctorRE = re.compile(r'\b' + '|'.join(names) + r'\b', re.I)

newLines = []
for originalLine in theFile.readlines():
    correctedLine = correctorRE(originalLine, substitutor)
    newLines.append(correctedLine)

someFile.writelines(newLines)


looks likely to go MUCH faster.  In general, in scripting languages,
reasoning and programming more abstractly can lead to faster code,
as well as higher productivity, than getting into minute details.
Regular expressions may often help with that (although they're easy
to fall in love with and overuse), here by offering case insensitive
matching as well as "bulk matching" with the | operator and "bulk
substitution" with the sub method.


Alex



From alwyn at alwyn.demon.co.uk  Mon Apr 22 12:46:56 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Mon, 22 Apr 2002 16:46:56 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org>
Message-ID: <alwyn-0BB274.17465022042002@news-text.blueyonder.co.uk>

In article <mailman.1019491732.12046.python-list at python.org>,
 Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote:
> 
> Alex may disagree with me here, but I think my opinion here is quite
> close to the middle Wittgenstein, or the _Philosohy of Mathematics_ and
> the _Blue and Brown Books_. 

Is it? I think different people understand radically different things 
when they read Wittgenstein. 

> It is also very close to some of Godel's
> remarks, in his philosophical moods.

I recall that Russell reports having had discussions with Goedel, 
Einstein and other distinguished emigres in America. He found their 
ideas impossibly 'Platonist', and he could not abide them. 


Alwyn


From jimw1956 at cs.comnojunk  Sun Apr  7 11:05:24 2002
From: jimw1956 at cs.comnojunk (Jim Williams)
Date: 07 Apr 2002 15:05:24 GMT
Subject: Can I print from Python?
References: <slrnab0n45.rrn.ak@ak.silmarill.org>
Message-ID: <20020407110524.09532.00002024@mb-cs.news.cs.com>

<< t's very platform-dependant; on unix, if your lp command is set up
correctly, you can save to a text file and then do os.system("lp
file.txt"). I don't know about windows. >>
Ahhh..that sheds a little light then.  probably could do something like 
os.system("copy file.txt lpt1") maybe.  Thanks will experiment with it.





From bkc at Murkworks.com  Thu Apr 11 12:50:02 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 11 Apr 2002 12:50:02 -0400
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it>
Message-ID: <3cb5bd03$1_6@goliath.newsgroups.com>

Heh.  3 posts lambasting me for using type(), and none of them actually
answered my question. ;-)

The point of my post is: why does type(newstyleinstance) return <class
__main__.x> rather than something like <instance __main__.x>?

---

aaaanyway.. "type(thing) in [InstanceType]" is what xml_pickle uses to
figure out how to pickle stuff.

Since xml_pickle dies trying to pickle instances of new style classes, I
went through and added isinstance(thing,ObjectType) which seems to work.

However you also say its wrong, which I agree is the wrong solution.

I just want to be able to xml_pickle any Python object. I don't write this
code, I just use it!

So what's needed is a reliable mechanism for determining that "something" is
an instance of a class. But in 2.2, looks like everything is an instance.
Fortunately I can just wait for someone smarter than I am to solve this
problem.

--
Novell DeveloperNet Sysop #5

_
"Alex Martelli" <aleax at aleax.it> wrote in message
news:kZbt8.5601$b62.137821 at news1.tin.it...
> Brad Clements wrote:
>
> > Could someone explain this to me.
> >
> > Lots of code does  type(x) in [InstanceType]
>
> I'll explain to you why lots of code does typetesting if you explain to me
> lemmings' alleged propensity for hurling themselves off a cliff.  You go
> first -- I suspect I can then just copy and paste your explanation right
> back to you:-).
>
>
> Alex
>




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From wware at alum.mit.edu  Thu Apr 25 00:21:33 2002
From: wware at alum.mit.edu (Will Ware)
Date: Thu, 25 Apr 2002 00:21:33 -0400
Subject: cgi script to study tarballs
Message-ID: <3CC7844D.ED73E83C@alum.mit.edu>

The following is a CGI script that peeks into gzipped tarballs
and lets you study the files in them. I've set this up on my
server and you can waste lotsa CPU on my server by sniffing
around this URL:
http://willware.net:8080/cgi-bin/studycode.py?tarball=http://www.python.org/ftp/python/src/py152.tgz
This script is incredibly slow with large tarballs, but it
does work.

###########################################################
#!/usr/bin/python
import string, sys, os

def ouch(arg):
    print "Content-type: text/html\n"
    print "<HTML><TITLE>Ouch</TITLE>"
    print "<BODY BGCOLOR=\"WHITE\">"
    print "<H1>Ouch: %s</H1>" % repr(arg)
    print "Don't try to call this CGI script directly."
    print "</BODY></HTML>"
    sys.exit(0)

Env = os.environ

try:
    for thing in string.split(Env['QUERY_STRING'], '&'):
        try:
            a, b = string.split(thing, '=')
            Env[a] = b
        except:
            pass

except KeyError:
    ouch(1)

if Env.has_key('tarball'):
    tarball = Env['tarball']
else:
    ouch(2)

if Env.has_key('file'):
    file = Env['file']
else:
    file = None

url = "http://willware.net:8080" + Env["REQUEST_URI"]

if file == None:
    # overview of the whole tarball
    print "Content-type: text/html\n"
    print "<HTML><TITLE>File list for %s</TITLE>" % tarball
    print "<BODY BGCOLOR=\"WHITE\">"
    print "<H1>File list for %s</H1>" % tarball

    print "<ul>"
    inf = os.popen("lynx -source %s | gunzip | tar tf -" % tarball)
    for L in inf.readlines():
        L = L[:-1]
        if L[-1] == "/":
            print '<li>' + L
        else:
            print ('<li><a href="%s&file=%s">%s</a>' %
                   (url, L, L))
    inf.close()
    print "</ul>"

    print "</BODY></HTML>"
    sys.exit(0)

else:
    # overview of the whole tarball
    print "Content-type: text/html\n"
    print "<HTML><TITLE>%s from %s</TITLE>" % (file, tarball)
    print "<BODY BGCOLOR=\"WHITE\">"
    print "<H1>%s <i>from</i> %s</H1>" % (file, tarball)

    cmd = ("lynx -source %s | gunzip | tar -O -x -f - %s" %
           (tarball, file))
    print "<pre>"
    inf = os.popen(cmd)
    i = 1
    print "<pre>"
    for L in inf.readlines():
        print ("%5d " % i) + L[:-1]
        i = i + 1
    print "</pre>"

    print "</BODY></HTML>"
    sys.exit(0)
###########################################################


From ralph at inputplus.demon.co.uk  Sun Apr  7 05:11:07 2002
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 7 Apr 2002 09:11:07 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com>
Message-ID: <a8p2fb$3da$1@inputplus.demon.co.uk>

Hi John,

> I understood your point the first time. I disagree with it. If you
> want to change the definition of a function to do something else,
> then you may well need to change the name. This is called
> refactoring.

I think we all perfectly understand your opinion too and no matter how
many times you repeat it there will still be some of us that like to do

    a = b + is_movable(c)

where is_movable will only ever return 0 or 1 meaning false or true.
That's because we have experience where this is possible, works well,
and is expressive.  We also have experience of languages where we loose
that simplicity so know we don't want to loose it again.

Just so you know your message is loud and clear  :-)

Cheers,


Ralph.



From ajs at ix.netcom.com  Tue Apr  9 23:43:24 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Tue, 9 Apr 2002 23:43:24 -0400
Subject: What has PEP 285 done to us?
Message-ID: <000f01c1e041$e0a832a0$0334fea9@carol>

Cliff writes -

>This was a direct attack on Guido's integrity, which was >uncalled for.
You *assumed* much about what he meant
>(perhaps he didn't make himself clear, but
>that's hardly grounds for labeling him a dictator) and when he >clarified
his position he became a liar.

I didn't label him a dictator. As far I understand he has labeled himself
such.  And the fact that he has is wonderful and honest - because he is and
should say so, and there is general agreement that software of this nature
cannot be designed by committee.

And being dictator he has the unique ability to issue the LION PEP. The LIKE
IT OR NOT PEP.

If Guido is concerned  - lets say -  about any code breakage issues he might
have missed lets limit discussion to that.

People don't waste time making arguments that will have no impact, Aahz
(oops, forgot we were suppose to be being nice) doesn't get to play friend
of the court, and at the end of the day folks decide to use or not use
Python, LION PEP included.

Clean and simple - and easier on everyone's treads.

And thanks, Cliff, for your thread.

Art





From loewis at informatik.hu-berlin.de  Tue Apr  9 05:09:35 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Apr 2002 11:09:35 +0200
Subject: exception question
References: <a8ua60$7nd$06$1@news.t-online.com>
Message-ID: <j47knhw84w.fsf@informatik.hu-berlin.de>

"Henning Peters" <pete at dexterslab.de> writes:

> normally, the second argument of except takes the variable name. the problem
> is, how i should get the variable, if i want to catch ALL exceptions. 

You should use sys.exc_info().

HTH,
Martin


From emile at fenx.com  Thu Apr 25 20:01:24 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 26 Apr 2002 00:01:24 GMT
Subject: AS/400
References: <1e140f34.0204250934.2fee1f26@posting.google.com>
Message-ID: <nN0y8.6207$ao1.8009@rwcrnsc54>

"Terje Johan Abrahamsen" <terjej at mailandnews.com> wrote in message
news:1e140f34.0204250934.2fee1f26 at posting.google.com...
> I am trying to write some Python code that moves info from Excel into
> AS/400 throuhgt a display program called Rumba. (www.walldata.com)
> Basically what I want it to do (short version) is that I go to a cell
> in Excel, copy that cell, changes to the Rumba window, goes to
> coordinate x,x pastes the number there. Then goes back to Excel, gets
> a new number, goes to Rumba, changes screens by typing F12, enters
> info from Excel again in coordinate x,x. It is more to it than this,
> but if I can figure out this, I can figure out the rest.
>
> The Excel part I have no trouble with. The AS/400 part is a little
> worse. Do you know some additions I have to install with Python? Where
> can I find some info about the functions that moves around on the
> AS/400 display and so forth. I have searched in google and on
> python.org but haven't found much. I can't be the only one that needs
> to do something like this? Thanks in advance for all answers and all
> pointing in the right direction so I can find out more info...

It likely will be fragile, but if you must, make sure the users know the
rules...

Take a look at mSched at http://www.mjtnet.com/  I've used it now a
couple of times now to automate processes with systems that didn't want
to play nice.  You can create macros on the fly with python and execute
them on demand, again using python to control the process.

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Wed Apr 17 22:31:31 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 18 Apr 2002 04:31:31 +0200
Subject: pre-PEP for optional 'pass'
In-Reply-To: <3CBE215D.F146ED46@alcyone.com>
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com> <a9gbhj$dhm$1@panix1.panix.com> <slrnabs3mm.1kfm.Gareth.McCaughan@g.local> <3CBE215D.F146ED46@alcyone.com>
Message-ID: <20020418023131.GD17361@devel.trillke>

> > (These days there ought to be a computer scientist in the
> > mix too. Suggestions?)
> 
> "I'll just write a program to exhaustively find them all.  What's a
> prime number again?"

or

  "That should be easy to do. Sorry, i am to busy with other stuff
   to do it myself."

	holger





From gimbo at ftech.net  Wed Apr 17 04:13:22 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Wed, 17 Apr 2002 09:13:22 +0100
Subject: pre-PEP for optional 'pass'
In-Reply-To: <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <a9hmba$2p26$1@nntp6.u.washington.edu> <slrnabp9cs.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <20020417081322.GA2063@andy.tynant.ftech.net>

On Tue, Apr 16, 2002 at 11:28:44PM +0100, phil hunt wrote:
> On Tue, 16 Apr 2002 10:17:30 -0700, Russell E. Owen <owen at nospam.invalid> wrote:
> >The need for pass is a pain when commenting out entire blocks of code 
> >(e.g. while debugging). Instead of just commenting out the code, one 
> >must also add a "pass" statement, and then try to remember to take it 
> >out again when re-enabling the code.
> 
> That's true, and one that has happened to me.

But as has already been pointed out, you can simply wrap the entire
block up in """.  Thus, this isn't a reason for getting rid of pass.

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From steve at ferg.org  Mon Apr 29 21:26:26 2002
From: steve at ferg.org (Stephen Ferg)
Date: 29 Apr 2002 18:26:26 -0700
Subject: String handling bug in Python
References: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com> <mailman.1019857431.12024.python-list@python.org>
Message-ID: <b16e4ef7.0204291726.79c9b596@posting.google.com>

> It seems if you imposed the limitation that the z-string not contain the quote
> character used to delimit the string then you could simply use concatenation 
> to get around that limitation:
> 
> z"'" z'"' z'\'

Yes. Exactly.


From sjoshi at ingr.com  Tue Apr 23 09:02:56 2002
From: sjoshi at ingr.com (Sunit Joshi)
Date: 23 Apr 2002 06:02:56 -0700
Subject: Why won't this POST
References: <8f8ffe67.0204222047.47d661ec@posting.google.com> <3CC528C8.FDE414CE@snakefarm.org>
Message-ID: <8f8ffe67.0204230502.261eeca9@posting.google.com>

Carsten Gaebler <clpy at snakefarm.org> wrote in message news:<3CC528C8.FDE414CE at snakefarm.org>...
> Sunit Joshi wrote:
> 
> > I'm trying to post data to IIS5.0 but keep getting error message about
> > "host not found".
>  
> > # begin HTTP request
> > req = HTTP('localhost/eubix')
> > #req.putrequest("GET" ,'/PyCgi/formTest.py?' + postData)
> > req.putrequest('POST' ,'/cgi-bin/ProcessInput.py') #Here's where it
> 
> "localhost/eubix" is not a valid hostname. What's the "/eubix" part
> supposed to mean?
> 
> cg.

Actually localhost/eubix is the virtual directory in IIS5.0. I can
connect to localhost/eubix, thru. req = HTTP('localhost/eubix')
But after that everything fails.

Sunit
sjoshi at ingr.com


From christophertavares at earthlink.net  Fri Apr 26 01:37:46 2002
From: christophertavares at earthlink.net (Chris Tavares)
Date: Fri, 26 Apr 2002 05:37:46 GMT
Subject: Why you sometimes get Windows Permission Denied errors
References: <3cc88d2f$1@news.nz.asiaonline.net>
Message-ID: <KI5y8.5446$iU4.458895@newsread2.prod.itd.earthlink.net>

"Colin Brown" <cbrown at metservice.com> wrote in message
news:3cc88d2f$1 at news.nz.asiaonline.net...
> Hi
>
> I have been struggling to solve why an occasional "Permission Denied"
error
> popped up from the following code fragment in one of my Win2K program
> threads:
>
> ...
> input_file = preprocess(raw_file)
> os.system('third_party input_file output_file > error_file')
> if os.path.exists(saved_file):
>     os.remove(saved_file)
> os.rename(input_file,saved_file)
> ...
>
> The error occurs on the os.rename(). The third_party executable was
closing
> input_file before terminating and anyway the subshell process has finished
> before the os.rename is called! Most baffling.
>
> With the aid of handle.exe from www.sysinternals.com I finally resolved
the
> problem. To see the problem at first hand try the following script:
>
> import time, thread, os
>
> def rm(file):
>     print 'delete x.x'
>     os.remove(file)
>
> def wait():
>     if os.name == 'nt':
>         os.system('pause')
>     elif os.name == 'posix':
>         os.system('sleep 3')
>     print 'end wait'
>
> print 'create x.x'
> f = open('x.x','w')
> thread.start_new_thread(wait,())
> time.sleep(1)
> print '\nclose x.x'
> f.close()
> rm('x.x')
>
> Although this works fine on Linux, I get a Permission Denied error on
> Windows. I surmise that an os.system call uses a C fork to create the
> subshell  - an exact duplicate of the current process environment
including
> OPEN FILE HANDLES. Because the os.system call has not returned before the
> os.remove is called a "Permission Denied" error occurs. Quite simple
really.
>

Not quite what's happening - Windows doesn't do fork. CreateProcess (the
underlying system call) will start a new process, but it doesn't do the
virtual memory copy thing that Unix does on a fork. However, there is
something called file handle inheritance, so that could be what's happening
here.

> Now going back to my original problem, I had other threads doing os.system
> calls. When one of these calls happens during the preprocess file write of
> my above thread AND takes longer to complete than the os.system call in
the
> above thread then the file will still be open and the os.rename will
return
> the Permission Denied error.
>
> What can be done to prevent the problem? I am open to suggestions here. My
> solution was to use a threading.Semaphore() lock around all my file
> operations to ensure they could only happen sequentially and also around
my
> os.system calls as below:
>

[... snip ... ]

Hmm... one possible solution would be to use win32process.CreateProcess
(part of the win32 extensions) instead of os.system. Depending on how
complex your command lines are (do you do a lot of io redirection?) that may
be either trivial or very difficult. That way you can make sure handles
don't get inherited, and you won't have this problem.

I suspect that the underlying C library system call just isn't set up to be
called in this manner - os.system isn't the Windows Way.

-Chris





From jimd at vega.starshine.org  Thu Apr  4 04:02:38 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 4 Apr 2002 09:02:38 GMT
Subject: Catalog of Python Patterns? [LONG!]
References: <mailman.1017886389.26908.python-list@python.org>
Message-ID: <a8h4re$274k$1@news.idiom.com>

In article <mailman.1017886389.26908.python-list at python.org>, Delaney, 
Timothy wrote:

>> From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
>> jimd at vega.starshine.org (Jim Dennis) writes:

>>>  At a minimum it should list each of the 23 GoF patterns, and
>>>  at least one minimal Python implementation of that pattern.

>> Yucch, I've never been a fan of that book, it seems too cultish to me.

> More importantly, many of the patterns are not applicable to Python. _Design
> Patterns_ comes from a C/C++ point of view ... there is also _Design
> Patterns for Smalltalk_ (or something along those lines) which is apparently
> much more applicable.

 Actually the GoF book has some C++ and some Smalltalk examples.  They
 aren't separate.

> It is useful to *know* the GoF patterns, but it is important not to be a
> slave to them.

>Tim Delaney

 I agree that the book has somewhat of a cult following, and I'm
 naturally wary of cults.  I also agree that one should not be a 
 slave to fashion, and that some of the hype of "patterns" (and
 XP, and ...) is "fashion" more than practical.

 However, I'm trying to understand them well enough to distinguish
 which parts of useful versus cosmetic.

 I disagree (with one followup that's not quoted herein) that
 patterns are not applicable to Python.  The author of that comment
 claimed that GoF was a C++ specific book (which is untrue, as I 
 point out above).  Python is much more like Smalltalk than C++,
 and the GoF specific chose it because it is a dynamic language
 --- for its contrast to C++ and its ilk.

 As I pointed out in my post, Python employs several patterns in
 its core and standard libraries.  Every class is a Factory, the
 Bastion module returns Proxy objects, we see Iterators all over
 the place, etc.  So I still want to see more.




From phd at phd.pp.ru  Fri Apr 26 11:58:52 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 26 Apr 2002 19:58:52 +0400
Subject: A SYSV IPC module?
In-Reply-To: <20020426155237.4009.qmail@eklektix.com>; from corbet-python@lwn.net on Fri, Apr 26, 2002 at 09:52:37AM -0600
References: <20020426155237.4009.qmail@eklektix.com>
Message-ID: <20020426195852.D28624@phd.pp.ru>

On Fri, Apr 26, 2002 at 09:52:37AM -0600, Jonathan Corbet wrote:
> I have a task which, it appears, requires hooking up to SYSV shared memory
> segments and semaphores.  Surprisingly, there appears to be no Python
> module which exports these system calls.  Surely somebody must have had to
> solve this problem before...does anybody have a pointer to where I could
> find such module?

   http://py.vaults.ca/parnassus/apyllo.py?find=shared+memory

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




From brian at sweetapp.com  Sat Apr  6 13:57:19 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 6 Apr 2002 10:57:19 -0800
Subject: "Memory could not be Read/Written" crashes with C extension DLL... HELP PLEASE!
In-Reply-To: <5a4226f0.0204061049.c2b1026@posting.google.com>
Message-ID: <003901c1dd9c$e2096df0$445d4540@Dell2>

> What is the difference though, shouldn't the code below work too?  Is
> there maybe something wrong in the PyTuple code?  (As mentioned
> before, it only crashed when called multiple times (40+ usually), but
> worked fine otherwise.

PyTuple_SetItem does not increment the reference count of the object.
That's exactly what you want for pLong* but probably not for inBuffer
(you didn't show enough code for me to be sure). 

You should develop with a debug version of Python. That way you will see
the reference count after each command you type in the interactive
shell. 

Cheers,
Brian

> 
> >   // build pp1, inBufferSize, pp3 into PyObjects for returning
> >   pLong1 = PyLong_FromLong(pp1);
> >   pLong2 = PyLong_FromLong(inBufferSize);
> >   pLong3 = PyLong_FromLong(pp3);
> >
> >   // build pLong1, pLong2, pLong3, pString into a tuple to return
> >
> >   pReturn = PyTuple_New(4);
> >   PyTuple_SetItem(pReturn, 0, pLong1);
> >   PyTuple_SetItem(pReturn, 1, pLong2);
> >   PyTuple_SetItem(pReturn, 2, pLong3);
> >   PyTuple_SetItem(pReturn, 3, inBuffer);
> >
> >   return pReturn;
> > }
> --
> http://mail.python.org/mailman/listinfo/python-list





From novaluz at sp.mailbr.com.br  Mon Apr 15 11:07:17 2002
From: novaluz at sp.mailbr.com.br (NovaLuz)
Date: Mon, 15 Apr 2002 12:07:17 -0300
Subject: Concorra a uma Luminaria - Novaluz - GRATÏS!!
Message-ID: <101455-22002411515717420@sp.mailbr.com.br>

                                                                  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020415/78b13c89/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: application/octet-stream
Size: 5298 bytes
Desc: image001.gif
URL: <http://mail.python.org/pipermail/python-list/attachments/20020415/78b13c89/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.gif
Type: application/octet-stream
Size: 6686 bytes
Desc: image002.gif
URL: <http://mail.python.org/pipermail/python-list/attachments/20020415/78b13c89/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image003.gif
Type: application/octet-stream
Size: 3254 bytes
Desc: image003.gif
URL: <http://mail.python.org/pipermail/python-list/attachments/20020415/78b13c89/attachment-0002.obj>

From erict at millfilm.co.uk  Thu Apr 25 13:39:37 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 25 Apr 2002 18:39:37 +0100
Subject: Tix install question
Message-ID: <3CC83F59.C6BF1027@millfilm.co.uk>

I am trying to install Tix on a linux machine. I don't have access
privilege to write above my user directory.
There is already a python1.5 on the machine.
I installed  tcl8.3, tk8.3, tix8.1 and python2.2 with  --prefix =
/usr/people/erict/bin
It was 'succefull'
I included the following line in my .cshrc:
setenv PYTHONHOME /usr/people/erict/bin
setenv PYTHONPATH /usr/people/erict/bin/lib
setenv TIX_LIBRARY /usr/people/erict/bin/lib
setenv TCL_LIBRARY /usr/people/erict/bin/lib
setenv TK_LIBRARY /usr/people/erict/bin/lib

now here what I am getting:
erict_fairlop > python
Python 2.2.1 (#3, Apr 25 2002, 15:25:49)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import Tix
>>> root = Tix.Tk()
/usr/people/erict/bin/lib/libtix8.1.8.3.so     ## print out that I had
to see what it was trying to load
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/people/erict/bin/lib/python2.2/lib-tk/Tix.py", line 213, in
__init__
    self.tk.eval('package require Tix')
TclError: couldn't load file "/usr/lib/libtix": /usr/lib/libtix.so:
undefined symbol: XLowerWindow
>>>


Any Idea
Thanks
Eric



From s.schwarzer at ndh.net  Mon Apr  1 06:12:37 2002
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Mon, 01 Apr 2002 12:12:37 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com>
Message-ID: <3CA840A5.A522501D@ndh.net>

Erik Max Francis wrote:
> Stefan Schwarzer wrote:
> > but on the other hand have to keep in mind subtle interactions between
> > bools and ints. For example, I consider it very unintutive that
> > "True == 1" should by true but "True == 2" should be false.
> 
> Why?  Implicit conversions seem un-Pythonic, so why would this be
> unfortunate?  When you write (True == 2) you're asking, "Is the value
> true equal to the integer value 2?" the answer to which is clearly no.
> Same thing with something like (2 == "2") -- no conversions are going on
> here so the test is false.

But, if I understand your argument correctly, "True == 1" and
"False == 0" shouldn't be true either because True and False are
bools, and 1 and 0 are ints. (The problem with this is, that, as a
consequence to the reasoning in the PEP, bools are ints in disguise.)

Example: "(val1, val2)[0]" is pretty clear, but "(val1, val2)[False]"
looks quite strange. If that's the price of the change I would still
prefer that "2 > 1" gives 1, not True.

So I see these choices:

1 Have a bool type without these strange interactions with ints (i. e.
  a bool should never equal an int, as an int should never equal a
  string, as you explain above), which breaks a lot of code.

2 Or, as the PEP suggests, introduce a bool type which behaves quite
  unintuitive, but keeps existing programs running.

Number 1 isn't practical, number 2 is confusing. I think it does more
harm than good.

A bool type in itself would be nice, but in fact I have never missed it
in Python. It's not worth the price which emerged during the discussion.

Stefan


From aleax at aleax.it  Thu Apr 11 17:45:02 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 21:45:02 GMT
Subject: Tuple comprehension
References: <mailman.1018540222.3383.python-list@python.org> <oGit8.8539$vF6.244299@news2.tin.it> <slrnabbk8q.j2e.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ytnt8.9887$vF6.289331@news2.tin.it>

Huaiyu Zhu wrote:
        ...
> One ancillary question: why xrange(n) is not iterator?

The Time Machine must have malfunctioned, I guess, so that, strangely 
enough, they were unable go back and design xrange as an iterator years 
before iterators were in the language...?


Alex



From sholden at holdenweb.com  Tue Apr 16 08:14:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 08:14:24 -0400
Subject: Backing up mail in Outlook Express using Python
References: <cGSu8.101$1kc.179320832@news.telia.no>
Message-ID: <JCUu8.23257$Ji7.8640@atlpnn01.usenetserver.com>

"Thomas Weholt" <thomas at gatsoft.no> wrote in message
news:cGSu8.101$1kc.179320832 at news.telia.no...
> I need to backup all my mail stored in Outlook Express. Can I use Python
to
> do this?
>
No reason why not, though I'm not sure why you'd want to, since the message
are just stored in files.

You can find out where they are from the Maintenance tab on Tools | Options
(click the "Store Folder" button).

I did adapt some code from Delphi to read the mailbox files, and can let you
have that if it would help, but sounds to me like a file copy is all you're
looking for.

regards
 Steve






From max at alcyone.com  Tue Apr 16 06:44:08 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 16 Apr 2002 03:44:08 -0700
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
		<3CBBD31D.AA2E653C@alcyone.com> <20020416115225.1d9fa2d2.wilk@flibuste.net>
Message-ID: <3CBC0078.C10C5F4@alcyone.com>

William Dode wrote:

> thanks, i write now to google to ask why they didnt spoke about.

Huh?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From maxm at mxm.dk  Tue Apr  9 07:06:12 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 09 Apr 2002 13:06:12 +0200
Subject: Discussion of Python midi library
References: <mailman.1018286064.26816.python-list@python.org> <a8ufgq$n1u$1@thorium.cix.co.uk>
Message-ID: <3CB2CB24.7000607@mxm.dk>

gbreed at cix.compulink.co.uk wrote:

> Mike Brenner wrote:

>>Here is a suggestion as to what to encapsulate (each one is harder than 
>>the next.


Oh boy ... I guess that you will get to implement a few of those things 
yourself if you want to use them ;-)

With my knowledge of midi and music production I expect the project to 
consist of several layers in the end:

     1. PrimitiveMidi
     Reads in the file, converts from the binary format to some sensible 
Python objects like BigEndian, VarLength, Byte, Code. All in all just a 
list of objects. I want to keep it as simple and low level as possible. 
After a read there will be a single list of simple objects that mainly 
represents string and ints and can converto those to/from the fileformat.

     2. Midi
     The library that programmers actually want to use. It should 
encapsulate the primitive data or convert it to some more relevant and 
userfriendly objects. Like: midi_event, time_signature, tempo_event and 
so on. But pretty much I expect it to be an implementation of the midi spec.

     3. MidiPlayer
     Plays music in realtime through the midiport. Hereby also the 
soundcard.

     4. MidiSequencer
     This should be a library that is independant of the Midi library, 
but can work together with that and the MidiPlayer. Here objects for 
setting things like tempo, time signatures, musical events etc. should 
be. My ideal here is Cubase, but without any kind of graphics.

     5. AlgoComposer
     Well this is actually what I am trying to get at. The rest is just 
filling :-) It will be completely independant of the other libraries but 
will use them. I expect it to be general enough to also generate output 
to software composition systems like c-sound. Perhaps I will skip 3 and 
4 completely.

Of all this I have made a first version of "PrimitiveMidi". But it was a 
bad design so I am rewriting it now.

I have a very strong idea as to how I want the "Midi" layer to work.

I also have some prototypes of the "AlgoComposer", so I know where I 
want to end.

generally I want both ends to meet nicely. The least important part for 
me is probably the "MidiPlayer" as I can allways save my music as a midi 
file and then play that. Basically I am afraid that it will take to long 
to write, but I know that interactivity is great in making music so I 
also fin



>>1. Midi Instruments (preferably the hundred general midi instruments 


Belongs in "MidiSequencer"



>>2. Midi Selection Sequences. Each midi hardware will have certain 


Belongs in "MidiSequencer"



>>3. Image. A string image function must represent every object, possibly 
>>in a standard format like xml, musicml, etc.)


Belongs in none of my library modules. But feel free to write your own :-)

>>4. Time. There must be a consistent time model used throughout, which 
>>easily represents notes, rests, start, finish, etc.


Belongs in "MidiSequencer"



>>5. Graphical representations in staff format, say, in jpeg or png 
>>6. Reading in a bit-mapped graphical representation (say a jpeg or png 


Not in any of my modules.


>>7. Pattern recognition of a wav file into a midi format, choosing the 
>>right notes, durations, and instruments.

> And this is an extremely ambitious project in itself.


Yes and I won't do it in my lifetime  ;-)

regards Max M



From jacek.generowicz at cern.ch  Fri Apr 26 10:03:31 2002
From: jacek.generowicz at cern.ch (Jacek Generowicz)
Date: 26 Apr 2002 16:03:31 +0200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org>
Message-ID: <tyf1yd25zgs.fsf@pcitapi22.cern.ch>

"James J. Besemer" <jb at cascade-sys.com> writes:

> Paul Foley wrote:

[snip]

> > In any case, parentheses are not used around sub-expressions, or to
> > separate or group statements.  They're only used (in evaluated
> > contexts) for "function" calls, and most other languages, including
> > Python, use parentheses for that purpose, too:
> >
> >  Python: foo(x)
> >  Lisp: (foo x)
> >
> > Same number of parentheses, right?
> 
> Besides, this is an entirely contrived, intellectually dishonest example.

I disagree. It accurately illustrates the point being made. It is
important that the point be made, because all too many people believe
the lies they are told about parentheses in lisp.

> First off, parens ARE used around sub expressions last I checked,

No, as Paul said, they are used for "function calls". Admittedly, in
mathematical expressions function call often coincide with
sub-expressions, but it does depend on the sub-expression in
question. Compare

   (+ 1 2 3 4 5 6 7 8 9 10)

with

   1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10

Gosh, what a lot of typing I have to do in Python, compared to lisp.

> And, oh yeah, I forgot, aside from the parentheses, the pervasive
> prefix notation can be a drag.  And I'm the type that still finds
> HP's stack notation less confusing in a calculator than infix.  But
> having to mentally convert everything back and forth between infix
> and prefix notation also can make you nuts.

> That being said, I acknowledge that arithmetic is a small part of
> some large systems and once you rise above the lower levels of
> abstraction, "everything" is function (or member) calls, and your
> original example begins to be more valid.  Still, arithmetic crops
> up and awful lot even then and can confound even the most
> enthusiastic programmer.

I don't think you will find many people who will argue that prefix
notation is easier to type or read than infix, when dealing with
mathematical formlae, particularly as those mathematical formulae are
likely to be copied from or compared to ones written on paper, with
all the usual operator precedences taken for granted. But, if you are
dealing with mathematical expressions, then, obviously, you will use
some sort of infix notation macro:

  [2]> (load "utilities/infix.lisp")
  ;; Loading file utilities/infix.lisp ...
  ;;; ******************************************************************
  ;;;   Infix notation for Common Lisp.
  ;;;   Version 1.3  28-JUN-96.
  ;;;   Written by Mark Kantrowitz, CMU School of Computer Science.
  ;;;   Copyright (c) 1993-95. All rights reserved.
  ;;;   May be freely redistributed, provided this notice is left intact.
  ;;;   This software is made available AS IS, without any warranty.
  ;;; *******************************************************************
  ;; Loading of file utilities/infix.lisp is finished.
  T
  [3]> #i(1+2*3)
  7
  [4]> #i(2*asin(1))
  3.1415927
  [5]> #i(1+2*3/4-5)
  -5/2

Incidentally, it would be interesting to see how many people would
correctly work out in their heads (on first attempt) what 1+2*3/4-5
evaluates to (leaving aside the fact that in Python it evaluates to
-3[*]) . . . and then compare it to the success rate when considering
the equivalent (+ 1 (/ (* 2 3) 4) -5). Or maybe how long it takes them
to get the right answer in each case.

> > I'd expect Lisp syntax to be
> > rather less error-prone, if anything, since it's much simpler
> 
> I think that's debatable.  Certainly simplicity makes it easier to
> learn, so that's a plus.  But it's still very tedious to get all the
> parentheses matched up just right.

It seems that many (most?) people who actually try to write lisp
without making a big issue of the parentheses, very rapidly come to a
conclusion that disagrees with yours.




[*]
I'm sorry, I can't remember what was concluded about new behaviour of
Python in this respect (and I only have access to 1.5 at present :-(


From mark at mceahern.com  Wed Apr 24 16:25:41 2002
From: mark at mceahern.com (Mark McEahern)
Date: Wed, 24 Apr 2002 15:25:41 -0500
Subject: configuring mxCGIPython
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEHPKMAA.mark@mceahern.com>

Hi, I'm trying to configure mxCGIPython:

  http://www.egenix.com/files/python/mxCGIPython.html#Install

I'm stuck at step 4, chmod 755 the path to the binary file cgipython.  I'm
using this cgi script:

  #!/bin/sh
  echo content-type: text/html
  echo
  path=`pwd`
  cgifile=$path/cgipython

  if [ -f $cgifile ]
  then
    echo Making $cgifile executable...
    if chmod -v 755 $cgifile
    then
      echo Successful.
    else
      echo Failed.
    fi
  else
    echo $cgifile does not appear to exist.
  fi

I get these results:

  Making /home/chain/pythons/cgipython executable...
  failed to change mode of /home/chain/pythons/cgipython to 0755 (rwxr-xr-x)
  Failed.

If I ftp to the server, I can chmod 755 the specific file.  If I change the
script just to reference the filename (not the full path), it works.

The problem is that the shebang line for the Python cgi files needs to have
the full path, I think.  If I don't chmod the full path, it doesn't work.
The full path would be:

  #!/home/chain/pythons/cgipython

I've also tried using a relative path for the shebang line.  The cgi-bin
folder and the folder where cgipython is stored are siblings, so relative to
the Python cgi files in cgi-bin, the path to the cgipython file should be:

  ../pythons/cgipython

But I've tried:

  #!../python/cgipython
  #!/python/cgipython

None of these work.  I simply get the internal server error.

Any ideas?

Thanks,

// mark





From dalke at dalkescientific.com  Wed Apr 17 14:20:15 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 17 Apr 2002 12:20:15 -0600
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9kebg$ub9$1@slb5.atl.mindspring.net>

Philipp Lenssen:
>can I use something like
>    verboseGender = (gender == 'm') ? 'male' : 'female'
>in Python?

No.  Search the archives for "ternary operator" for plenty of
past discussion.  Google lists 65 posts for comp.lang.python.

Besides, for this specific case you should have some error checking,
because there are genders other than 'male' and 'female' (eg,
asexual bacteria and hermaphradite earthworms).

  verboseGenderTable = {
     "m": "male",
     "f": "female",
  }
  verboseGender = verboseGenderTable[gender]

Then if something other than 'm' or 'f' is used, this will raise
a KeyError.

                Andrew
                dalke at dalkescientific.com





From eric at scipy.org  Thu Apr 11 20:26:22 2002
From: eric at scipy.org (eric)
Date: Thu, 11 Apr 2002 20:26:22 -0400
Subject: python on cygwin
Message-ID: <01d801c1e1b8$ac962e40$6b01a8c0@ericlaptop>

Hello,

I'm trying to build python from scratch on cygwin and am running into "remap"
errors such as:

    C:\cygwin\tmp\scipy_test\cygwin\Python-2.2.1\python.exe: *** unable
    to remap C:\cyggdbm.dll to same address as parent -- 0x420000

I'm using the latests cygwin (1.3.10) on W2K.  When trying to build python,
THings chug along just fine, and then I get the following:

gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -DUSE_DL_IMPORT -I. -I/tmp/scipy_t
/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/scipy_test/cygwi
.2.1/Modules/gdbmmodule.c -o build/temp.cygwin-1.3.10-i686-2.2/gdbmmodule.o
gcc -shared -Wl,--enable-auto-image-base build/temp.cygwin-1.3.10-i686-2.2/gdbmm
/usr/local/lib -L. -lgdbm -lpython2.2 -o build/lib.cygwin-1.3.10-i686-2.2/gdbm.d
Warning: resolving _gdbm_errno by linking to __imp__gdbm_errno (auto-import)
building 'termios' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -DUSE_DL_IMPORT -I. -I/tmp/scipy_t
/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/scipy_test/cygwi
.2.1/Modules/termios.c -o build/temp.cygwin-1.3.10-i686-2.2/termios.o
C:\cygwin\tmp\scipy_test\cygwin\Python-2.2.1\python.exe: *** unable to remap C:\
\cyggdbm.dll to same address as parent -- 0x420000
C:\cygwin\tmp\scipy_test\cygwin\Python-2.2.1\python.exe: *** unable to remap C:\
\cyggdbm.dll to same address as parent -- 0x420000
     25 [main] python 1684 sync_with_child: child 1908(0x144) died before initia
ith status code 0x1
   6684 [main] python 1684 sync_with_child: *** child state child loading dlls
error: Resource temporarily unavailable
make: *** [sharedmods] Error 1

I read a few mails about a rebase solution at:

    http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html

However, after building and running this on the percribed files, I get:

$ rebase -d -b 0x68000000 -o 0x10000 cygXpm-X4.dll cygXpm-noX4.dll     cygbz21.
0.dll cygcrypto.dll cygform5.dll cygform6.dll     cyggdbm.dll cyghistory4.dll c
yghistory5.dll cygintl.dll     cygitcl30.dll cygitk30.dll cygjbig1.dll cygjpeg6
b.dll cygmenu5.dll     cygmenu6.dll cygncurses++5.dll cygncurses++6.dll cygncur
ses5.dll     cygncurses6.dll cygpanel5.dll cygpanel6.dll cygpcre.dll     cygpcr
eposix.dll cygpng2.dll cygreadline4.dll cygreadline5.dll     cygregex.dll cygss
l.dll cygtcl80.dll cygtclreg80.dll cygtiff3.dll     cygtk80.dll cygz.dll
cygXpm-X4.dll: new base = 67fe0000, new size = 20000
cygXpm-noX4.dll: new base = 67fc0000, new size = 20000
cygbz21.0.dll: new base = 67f90000, new size = 30000
cygcrypto.dll: new base = 67ee0000, new size = b0000
ReBaseImage() failed with last error = 2

Can anyone point out the error in my ways, or point me to a fix?

thanks,
eric


--
Eric Jones <eric at enthought.com>
Enthought, Inc. [www.enthought.com and www.scipy.org]
(512) 536-1057







From michael at stroeder.com  Wed Apr  3 05:18:55 2002
From: michael at stroeder.com (=?ISO-8859-1?Q?Michael_Str=F6der?=)
Date: Wed, 03 Apr 2002 12:18:55 +0200
Subject: How to check whether email /domain exists
References: <mailman.1017823807.18628.python-list@python.org>
Message-ID: <3CAAD70F.8050708@stroeder.com>

Denis S. Otkidach wrote:
> On Wed, 3 Apr 2002, A wrote:
> 
> A> What is the simplest way how to check
> A>  whether a given email address( or domain for this email)
> A> exists?
> A> Thanks for help
> 
> Check for MX record, e.g. with
> http://www.interlink.com.au/anthony/python/dns.tar.gz

This module moved to: http://pydns.sf.net

Ciao, Michael.



From bbollenbach at shaw.ca  Thu Apr 11 21:37:59 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 01:37:59 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <3CB58302.5090200@mxm.dk>
Message-ID: <XTqt8.15956$Kq4.675799@news2.calgary.shaw.ca>

In article <3CB58302.5090200 at mxm.dk>, Max M wrote:
> btw: it would be nice if join() had a switch so that it would 
> automatically do a str() on each object it joins.
> 
> like::
> 
> >>> print ','.join(range(5), toStr=1)
> 0,1,2,3,4
> 
> I keep having to do this repeatedly in my code when formatting output.
> 
> Perhaps join() should even do this as the default, as it makes the most 
> sense in the join() context.
> 
> like::
> 
> def newJoin(self, items):
>      return self.join([str(i) for i in items])
> 
> Is there any case where it isn't applicable at all?
 
There's lots of places where it's not applicable:

>>> amounts = [1.234, 5.281, 45.23421]
>>> ','.join(["%.2f" % x for x in amounts])
'1.23,5.28,45.23'

or whatever. The point is, the possibilities are limited only by your
imagination.


HTH,

Brad Bollenbach


From phr-n2002a at nightsong.com  Wed Apr 24 19:36:09 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 16:36:09 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <slrnacdape.60f.akuchlin@ute.mems-exchange.org> <7xu1q1mdwd.fsf@ruckus.brouhaha.com> <1a517b5.0204241217.7400376b@posting.google.com>
Message-ID: <7xelh4sm8m.fsf@ruckus.brouhaha.com>

bryanjugglercryptographer at yahoo.com (Bryan Olson) writes:
> > And it actually might be faster to use two separate hash-based auth
> > keys than deal with the fancy HMAC padding, if a simple double hash
> > has some vulnerabilities.
> 
> HMAC's is only a small additive constant slower than the hash function.  
> For messages more than a few blocks long, it's not significant.  The
> extra padding and additional hash function calls only deal with two blocks.
> 
> Below is single-call HMAC in Python.

The additive constant is pretty painful for short strings.  In
applications I've worked on, short strings (like account numbers in a
database, or session ID's written into browser cookies) get encrypted
a lot and it's worthwhile to be able to handle them fast.

The SHA encryption function is already pretty slow for short strings
(implementing string-xor by 32-bit operations with the array module
helps for long strings).  HMAC pads the key out to the block length of
the underlying hash function, which also means the compression
function is called an extra time on each of the two hashes.  That may
not matter much when called from Python though.  And I think I can
speed up the HMAC string-xor's either with the array module or with
the string.translate function.

I'm looking for other ways to speed up the encryption function for
short strings but need to make more measurements.

Meanwhile, the whole function has mysteriously slowed down by about
15% even for long strings, even when I back out to old versions from
my RCS depot.  It's like my computer has suddenly gotten slightly
slower.  I'll reboot later and see if that changes anything, but it
makes me wonder if my older benchmarks are any good either.


From frr at easyjob.net  Fri Apr 26 11:37:17 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Fri, 26 Apr 2002 17:37:17 +0200
Subject: Thesaurus with Python
References: <3CC96DDF.6010405@thomas-guettler.de>
Message-ID: <70ticuo90v74ko0tq64hcs0mg5ogo92ima@4ax.com>

On Fri, 26 Apr 2002 17:10:23 +0200, Thomas Guettler
<zopestoller at thomas-guettler.de> wrote:

>I want to design a thesaurus in python.
>
>If I search for thesaurus and pyhton with google I get
>no reasonable result.

You should search for 'python wordnet' instead. ;-)

Try this: http://pywordnet.sourceforge.net/



-----------------------
Fernando Rodriguez


From wilk at flibuste.net  Tue Apr  9 06:56:32 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Tue, 9 Apr 2002 12:56:32 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
	<vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>
	<3CB274D2.A7554C79@engcorp.com>
	<a8u6tj$quj$02$1@news.t-online.com>
Message-ID: <20020409125632.40965c79.wilk@flibuste.net>

Le Tue, 9 Apr 2002 10:03:04 +0200
"Harald Schneider" <h_schneider at marketmix.com> ?crivait:

> Thats the fact! If the community gives back the experience, people like
> me would stop thinking about using
> Delphi or VB ... we soon would have an excellent and up-to-date
> documentation.
Me to !

> 
> What about posting them to
> http://wiki.wxpython.org/index.cgi/wxPython_20Cookbook :)
Is-there an equivalent for python (langage and library) ?

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From tim.one at comcast.net  Tue Apr 16 16:38:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 16:38:52 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <BIEJKCLHCIOIHAGOKOLHMENPCPAA.tim.one@comcast.net>
Message-ID: <BIEJKCLHCIOIHAGOKOLHOEOGCPAA.tim.one@comcast.net>

[Tim]
> On a clean Win2K Pro box, I'm finally able to provoke this (or at
> least one of its causes) reproducibly, via:
>
> 1. Boot the box.
>
> 2. Install 2.2.1.
>
> 3. Uninstall 2.2.1.
>
> 4. Install 2.2.1 again.  Note that \Python22\UNWISE.EXE is there.
>
> 5. Reboot the box (just logging off and on again isn't enough).
>
> 6. Now \Python22\UNWISE.EXE no longer exists.

For anyone playing along, this particular way of provoking the problem
appears stoppable via inserting step 2a:

2a: Delete this line from near the end of \Python22\INSTALL.LOG:

    Delete in-use files: On





From martin at v.loewis.de  Mon Apr 15 02:42:18 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:42:18 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <m34rif19hj.fsf@mira.informatik.hu-berlin.de> <slrnabjgst.741.philh@comuno.freeserve.co.uk>
Message-ID: <m3n0w5layd.fsf@mira.informatik.hu-berlin.de>

philh at comuno.freeserve.co.uk (phil hunt) writes:

> To me, the idea that ^ means xor is totally unclear.
> 
> What's wrong with xor meaning xor?

As a function, I have no problem with it. As an operator, the problem
is that it introduces a new keyword.

Regards,
Martin



From martin at v.loewis.de  Fri Apr 19 03:21:26 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Apr 2002 09:21:26 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019154174.26611.python-list@python.org>
Message-ID: <m31ydc2lxl.fsf@mira.informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> test test_largefile crashed -- exceptions.IOError: [Errno 27] File too large
> 
>     How much space does it want?  I don't know this one anyway.

It creates a file with 2500000000L. Depending on the file system you
are using, that may or may not be possible (it tries to create it with
a hole in it, so it normally takes only little disk space).

> test test_nis skipped -- No NIS master found for any map
> 
>     I didn't expect that.  We use nisplus....

You may want to run the test verbosely: 

./python Lib/test/regrtest.py -v test_nis

> test test_socket_ssl skipped -- Use of the `network' resource not enabled
> 
>     What network resouce?

You need to explicitly give regrtest.py permission to use certain
resources; see 'regrtest.py -h' for details. In the case, the default
'make test' doesn't run the test to avoid triggering a dial-up
connection.

> neelix hgs 11 %> ./python
> Python 2.2.1 (#2, Apr 18 2002, 18:23:04) [C] on sunos5
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import tkine ter
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: No module named tkinter

That should have been either Tkinter, or _tkinter - not tkinter.

> gcc -v -shared build/temp.solaris-2.7-sun4u-2.2/_tkinter.o build/temp.solaris-2.7-sun4u-2.2/tkappinit.o -L/usr/openwin/lib -L/usr/local/lib -ltk8.3 -ltcl8.3 -lX11 -o build/lib.solaris-2.7-sun4u-2.2/_tkinter.so
> gcc: build/temp.solaris-2.7-sun4u-2.2/_tkinter.o: No such file or directory
> gcc: build/temp.solaris-2.7-sun4u-2.2/tkappinit.o: No such file or directory

It seems that you have not put grabbed enough commands: there
definitely *was* a command to build both _tkinter.o and tkappinit.o.
When setup.py decides that building the module failed, it removes not
only the resulting .so, but also all involved object files.

Regards,
Martin



From gashton at cmedltd.com  Tue Apr 23 04:53:14 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 23 Apr 2002 09:53:14 +0100
Subject: Detecting a new style class
Message-ID: <1019551994.576.3.camel@tp6>

Hi. I've got some code tha only works properly for classic classes.
Before I get round to fixing it, I'd like to put some protection in
place to prevent it from being used with a new style class.

This means that I'd like to be able to detect whether my class is being
subclassed by a new or classic class. To do that I'm testing self to see
if it has a __new__ attribute. Is there a better way to do it?

I find that the behaviour varies depending on the order in which a
class's super classes are defined. In the example below, I'd like the
instantiation of class B to throw the same RuntimeError as A:

Python 2.2 (#2, Mar 11 2002, 13:24:00) 
[GCC 2.95.3 20010315 (release)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from AbstractSingleton import AbstractSingleton
>>> class A(AbstractSingleton, object):
...   pass
... 
>>> a = A()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "AbstractSingleton.py", line 60, in __init__
    self.__check_for_new_style_class()
  File "AbstractSingleton.py", line 66, in __check_for_new_style_class
    raise RuntimeError, "New style classes are not supported"
RuntimeError: New style classes are not supported
>>> class B(object, AbstractSingleton):
...   pass
... 
>>> b = B()
>>> 

Thanks.

-- 
Graham Ashton





From aahz at pythoncraft.com  Wed Apr 10 11:40:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 11:40:29 -0400
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <Xns91EBD90A7449Dlumberjack@216.148.53.83> <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de>
Message-ID: <a91mdd$ck$1@panix1.panix.com>

In article <slrnab7hn9.4tj.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>
>Why would anybody care if you aren't man/woman enough to post with
>your real name?

Because many people, respectable people, *do* post to Usenet without
using their "real name".  What the heck is a real name, anyway?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From tanzer at swing.co.at  Sat Apr  6 02:23:54 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Sat, 06 Apr 2002 09:23:54 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "Sat, 06 Apr 2002 01:40:01 GMT."
             <3CAE52CA.60F56E3D@python.org>
Message-ID: <E16tkYF-0003Zj-00@tswings.swing.cluster>

Guido van Rossum <guido at python.org> wrote:

> I agree.  I can assure you that I *did* read Laura's objections carefully,
> but ended up disagreeing with her conclusion that accepting the PEP was a
> bad idea.  I was baffled that she and Alex used so many words to explain
> their position ("do I have to read and respond to all *that*?!" were my
> exact
> thoughts) and this made me speechless in response.  Eventually I recovered
> from the blow and decided that Laura's fear that adding bool would make her
> teaching job harder was ungrounded.  By then, so many words had gone back
> & forth that I didn't feel like going back and responding to her piece in
> detail -- others had done that and the discussion had blown up to such a
> volume that I didn't think I could convince any of the PEP's detractors
> anyway.  I apologize to Laura that my judgement came across as rude.

Thank you.

Did-your-PEP-secretly-try-to-change-Alex's-verbosity-level-too-?-ly yrs
Christian

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





From mlh at vier.idi.ntnu.no  Fri Apr 12 00:30:27 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 04:30:27 +0000 (UTC)
Subject: try/except/else flexibility...
Message-ID: <slrnabcon3.g5r.mlh@vier.idi.ntnu.no>

Just a thought: Given the prevalence of try/except/else in the "don't
look before you leap" way of doing things, could it be useful to make
the statement more flexible? For instance, instead of writing

  try: foo + ''
  except: pass
  else: bar()

(assuming I'm sloppy enough not to specify TypeError on the second
line) one could simply write:

  try: foo + ''
  else: bar()

and if I want something done only if it causes no exceptions, for
instance pop off an element of a list unless it is empty, I could just
do:

  try: somelist.pop()

Is this heresy? I see the harm in not specifying exactly which
exceptions you're expecting, of course -- are there other arguments
against this? (I'm sure that might be argument enough... Oh, well...)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From phd at phd.pp.ru  Fri Apr 12 08:36:42 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 12 Apr 2002 16:36:42 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <200204121227.g3CCRSd04250@localhost.localdomain>; from dreed@capital.edu on Fri, Apr 12, 2002 at 08:27:28AM -0400
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk> <20020412144929.B12997@phd.pp.ru> <200204121227.g3CCRSd04250@localhost.localdomain>
Message-ID: <20020412163642.N12997@phd.pp.ru>

On Fri, Apr 12, 2002 at 08:27:28AM -0400, Dave Reed wrote:
> Have you compiled earlier versions of Python on this system?

   Many times.

links -dump http://www.lemburg.com/files/python/mxCGIPython.html | \
   grep Solaris | wc -l
:)

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




From erno-news at erno.iki.fi  Tue Apr  9 23:51:52 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Apr 2002 06:51:52 +0300
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <a8vpp5$67d$1@panix1.panix.com> <yu99pu18lai2.fsf@europa.research.att.com>
Message-ID: <kun0wcmcrr.fsf@lasipalatsi.fi>

In article <yu99pu18lai2.fsf at europa.research.att.com>, Andrew Koenig
<ark at research.att.com> writes:

| One thing that fascinates me is how the nature of a language affects
| the preferred idioms therein.  To help me further my understanding,
| would you mind explaining to me why this particular idea is undesirable
| in Python?  I can think of several possible reasons, but I don't know
| which reason is most plausible to the Python community.

the way i see it, knowing that your callee can't swap another object
into a variable makes life easier as you don't have to worry about
that happening.

this is a bigger help in python than it might be in some other
languages, because python doesn't have the option of passing
function parameters by value.

  -- erno


From phr-n2002a at nightsong.com  Wed Apr 24 23:05:14 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 20:05:14 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com>
Message-ID: <7x7kmwjx5h.fsf@ruckus.brouhaha.com>

Richard Parker <richard at electrophobia.com> writes:
> > Do the keys really need to be independent?
> 
> As I recall, it was essential to the HMAC proof of security that the keys be
> independent.  I don't know if is possible to prove HMAC secure with
> dependent keys, but it would almost certainly require stronger assumptions
> about the underlying hash function.

But HMAC's keys are already dependent (they're the same key xor'd with
two different constants), so I'm not sure how concatenating something is
really worse.  I'm back to wanting to re-read the HMAC paper, which I'll
try to do sometime.

> > How is H(K || (H || 'a' || K || X))?
> 
> I am concerned about it.  You're basically using HMAC with a weaker
> pseudorandom key derivation.  If you absolutely can not use the xor of ipad
> and opad, make sure you use a key derivation for both K1 and K2.  Perhaps
> the following:
> 
>   H(c1 || K || H(c2 || K || x))

OK, I can try that.  I'm resisting HMAC because in the perverse economics
of interpreted languages, calling the library SHA function can be faster
than xor'ing two 64-character strings in the interpreter (I'll have to
run actual timings though).  

> I strongly recommend you just use HMAC, rather than inventing your own MAC
> construction.

I'll experiment with some ways of coding HMAC to run fast in Python.
The implementation that comes with the 2.2 library is pretty slow.

Currently I'm using

  K1 = H('auth1' + K)
  K2 = H('auth2' + K)
  MAC = H(K1 + H(K2 + ciphertext))

which I hope is at least as good as HMAC (the keys are more
independent) and it uses fewer interpreter operations (though more SHA
calls).  It's speed isn't too bad, but I always like to look for
improvements.

Note that there's a trivial O(2**64) attack on my authentication since
I'm truncating the MAC to 8 bytes.  So if a security difference
between HMAC and what I'm doing needs more than O(2**64) work to
exploit, it's not really useful to an attacker.  Is that a reasonable
way to think of this?

Thanks as usual.  I've pushed new code up to my web site:

  http://www.nightsong.com/phr/crypto/p2.py


From doru-cat at ifi.uio.no  Thu Apr 18 03:52:12 2002
From: doru-cat at ifi.uio.no (Doru-Catalin Togea)
Date: Thu, 18 Apr 2002 09:52:12 +0200 (MET DST)
Subject: What webhosting
In-Reply-To: <3CBDE7B4.3348.4EB76@localhost>
Message-ID: <Pine.GSO.4.21.0204180908001.12591-100000@fjorir.ifi.uio.no>

Hi!

This is unrelated to Python programming.

I get som unsolicited emails sometimes and I often wonder what's in a
spammers mind?

Like this guy who used to send me everyday for two weeks an add for using
some Credit Card service of his? Lately he cooled down to once a
week. Does he really think I'm ever going to use his service when he
promotes it like that?

Then they hide behind all these "random generated - jump on the
keyboard - bunch of letters" usernames, they fake the header of the
e-mails, they provide no way for one to remove oneself from the list or to
contact them. Now that is indeed manly and demands respect! But then
again, they do not seem to care for either privacy or respect.

And maybe the most useless of them all are the ones sending
porn-spam. Really, porn is everywhere on the internett, who needs their
adds? Besides, sex is such an outworn issue. Event though this is not
relevant to this discussion, who has anything to offer which has not
been out there for ages? "Let me show you how someone 
put his dick in some hole." Preferably for a few bucks of course. Neither
the ones puting their dicks nor the ones providing the holes care in the
least for your or my sex-life, all they want is a valid credit card
number.

Who knows, maybe with the evolution of the internett protocols it might be
possible in the future for users to guard themselves against unsolicited
e-mail.

Some of these comments may not apply to the intentions of Ladislav, the
originator of this thread. I have no idea what kind of unsolicited e-mail
he wants to send.

Best regards,
Catalin


	<<<< ================================== >>>>
	<<     We are what we repeatedly do.      >>
	<<  Excellence, therefore, is not an act  >>
	<<             but a habit.               >>
	<<<< ================================== >>>>





From robinjim at earthlink.net  Fri Apr  5 19:05:58 2002
From: robinjim at earthlink.net (robin and jim)
Date: Sat, 06 Apr 2002 00:05:58 GMT
Subject: single-quoted string conversion to triple-quoted string
Message-ID: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>

Is there a way to convert a single quoted string to a triple quoted string?

For example, given:

''' abc %d xyz '''

the result of ''' abc %d xyz''' % 10 is:

' abc 10 xyz '

which is a single quoted string.

I would like to convert the result to a triple-quoted string.




From aahz at pythoncraft.com  Tue Apr 16 13:30:35 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 13:30:35 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018969952.6438.python-list@python.org>
Message-ID: <a9hn3r$ri9$1@panix1.panix.com>

In article <mailman.1018969952.6438.python-list at python.org>,
Anthony Baxter  <anthony at interlink.com.au> wrote:
>
>If instead this whole discussion is just usenet noise, with no real point
>or purpose, then ignore the previous paragraph. Needless to say, I wait
>with baited breath for the collection of shocking incompatibilities to 
>roll in. 

What are you using for bait?

>(is it a fact of usenet that over time, all newsgroups become like 
>comp.lang.perl.misc?)

Mu.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From clspence at one.net  Fri Apr  5 16:40:12 2002
From: clspence at one.net (Chris Spencer)
Date: 5 Apr 2002 15:40:12 -0600
Subject: Memory Leaks with wxwindow and grid
References: <3cab5418@news.vip.fi>
Message-ID: <616sau4beeq51j1u271qcaist53pafs83p@4ax.com>

	This was discussed on the wxPython mailing list a while back.  Evidently
this is "expected behaviour".
	Good luck on using the grid control.  I went through so many GPFs using
it that I'll *never* touch the ugly thing again.

Chris.

On Wed, 3 Apr 2002 22:10:04 +0300, "Pekka Niiranen" <krissepu at vip.fi> wrote:

>I get the following lines in dos prompt in win2000,
>when exiting the main frame:
>
>E:\koodit\python\grid>python pngrid.py
>21:39:42: There were memory leaks.
>21:39:42: ----- Memory dump -----
>21:39:42: wxEvtHandler at $1067DB8, size 48
>21:39:42:
>21:39:42:
>21:39:42: ----- Memory statistics -----
>21:39:42: 1 objects of class wxEvtHandler, total size 48
>21:39:42:
>21:39:42: Number of object items: 1
>21:39:42: Number of non-object items: 0
>21:39:42: Total allocated size: 48
>
>Note that if I comment line:
>            self.pn_grid = PnGrid(self)
>
>
>no memory leaks will occur then "exit" is selected from menu !!!
>
>What am I missing here ?
>
>----------code starts here-------------------
>
>from wxPython.wx import *
>from wxPython.grid import *
>from wxPython.lib.mixins.grid import wxGridAutoEditMixin
>import sys, pprint
>ID_EXIT  = 102
>
>class PnGrid(wxGrid):
>    def __init__(self, parent):
>        wxGrid.__init__(self, parent, -1)
>
>        self.moveTo = None
>
>        EVT_IDLE(self, self.OnIdle)
>
>        self.CreateGrid(25, 25)
>        #self.EnableEditing(false)
>
>        # simple cell formatting
>        self.SetColSize(3, 200)
>        self.SetRowSize(4, 45)
>        self.SetCellValue(0, 0, "First cell")
>        self.SetCellValue(1, 1, "Another cell")
>        self.SetCellValue(2, 2, "Yet another cell")
>        self.SetCellValue(3, 3, "This cell is read-only")
>        self.SetCellFont(0, 0, wxFont(12, wxROMAN, wxITALIC, wxNORMAL))
>        self.SetCellTextColour(1, 1, wxRED)
>        self.SetCellBackgroundColour(2, 2, wxCYAN)
>        self.SetReadOnly(3, 3, true)
>
>        self.SetCellEditor(5, 0, wxGridCellNumberEditor())
>        self.SetCellValue(5, 0, "123")
>        self.SetCellEditor(6, 0, wxGridCellFloatEditor())
>        self.SetCellValue(6, 0, "123.34")
>
>        self.SetCellValue(6, 3, "You can veto editing this cell")
>
>
>        # attribute objects let you keep a set of formatting values
>        # in one spot, and reuse them if needed
>        attr = wxGridCellAttr()
>        attr.SetTextColour(wxBLACK)
>        attr.SetBackgroundColour(wxRED)
>        attr.SetFont(wxFont(10, wxSWISS, wxNORMAL, wxBOLD))
>
>        # you can set cell attributes for the whole row (or column)
>        self.SetRowAttr(5, attr)
>
>        self.SetColLabelValue(0, "Custom")
>        self.SetColLabelValue(1, "column")
>        self.SetColLabelValue(2, "labels")
>
>        self.SetColLabelAlignment(wxALIGN_LEFT, wxALIGN_BOTTOM)
>
>        # test all the events
>        EVT_GRID_CELL_LEFT_CLICK(self, self.OnCellLeftClick)
>        EVT_GRID_CELL_RIGHT_CLICK(self, self.OnCellRightClick)
>        EVT_GRID_CELL_LEFT_DCLICK(self, self.OnCellLeftDClick)
>        EVT_GRID_CELL_RIGHT_DCLICK(self, self.OnCellRightDClick)
>
>        EVT_GRID_LABEL_LEFT_CLICK(self, self.OnLabelLeftClick)
>        EVT_GRID_LABEL_RIGHT_CLICK(self, self.OnLabelRightClick)
>        EVT_GRID_LABEL_LEFT_DCLICK(self, self.OnLabelLeftDClick)
>        EVT_GRID_LABEL_RIGHT_DCLICK(self, self.OnLabelRightDClick)
>
>        EVT_GRID_ROW_SIZE(self, self.OnRowSize)
>        EVT_GRID_COL_SIZE(self, self.OnColSize)
>
>        EVT_GRID_RANGE_SELECT(self, self.OnRangeSelect)
>        EVT_GRID_CELL_CHANGE(self, self.OnCellChange)
>        EVT_GRID_SELECT_CELL(self, self.OnSelectCell)
>
>        EVT_GRID_EDITOR_SHOWN(self, self.OnEditorShown)
>        EVT_GRID_EDITOR_HIDDEN(self, self.OnEditorHidden)
>        EVT_GRID_EDITOR_CREATED(self, self.OnEditorCreated)
>
>
>
>    def OnCellLeftClick(self, evt):
>        pprint.pprint("OnCellLeftClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnCellRightClick(self, evt):
>        pprint.pprint("OnCellRightClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnCellLeftDClick(self, evt):
>        pprint.pprint("OnCellLeftDClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnCellRightDClick(self, evt):
>        pprint.pprint("OnCellRightDClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnLabelLeftClick(self, evt):
>        pprint.pprint("OnLabelLeftClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnLabelRightClick(self, evt):
>        pprint.pprint("OnLabelRightClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnLabelLeftDClick(self, evt):
>        pprint.pprint("OnLabelLeftDClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnLabelRightDClick(self, evt):
>        pprint.pprint("OnLabelRightDClick: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>
>    def OnRowSize(self, evt):
>        pprint.pprint("OnRowSize: row %d, %s\n" %
>                       (evt.GetRowOrCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnColSize(self, evt):
>        pprint.pprint("OnColSize: col %d, %s\n" %
>                       (evt.GetRowOrCol(), evt.GetPosition()))
>        evt.Skip()
>
>    def OnRangeSelect(self, evt):
>        if evt.Selecting():
>            pprint.pprint("OnRangeSelect: top-left %s, bottom-right %s\n" %
>                           (evt.GetTopLeftCoords(),
>evt.GetBottomRightCoords()))
>        evt.Skip()
>
>
>    def OnCellChange(self, evt):
>        pprint.pprint("OnCellChange: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>
>        # Show how to stay in a cell that has bad data.  We can't just
>        # call SetGridCursor here since we are nested inside one so it
>        # won't have any effect.  Instead, set coordinants to move to in
>        # idle time.
>        value = self.GetCellValue(evt.GetRow(), evt.GetCol())
>        if value == 'no good':
>            self.moveTo = evt.GetRow(), evt.GetCol()
>
>
>    def OnIdle(self, evt):
>        if self.moveTo != None:
>            self.SetGridCursor(self.moveTo[0], self.moveTo[1])
>            self.moveTo = None
>        evt.Skip()
>
>
>    def OnSelectCell(self, evt):
>        pprint.pprint("OnSelectCell: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>
>        # Another way to stay in a cell that has a bad value...
>        row = self.GetGridCursorRow()
>        col = self.GetGridCursorCol()
>        if self.IsCellEditControlEnabled():
>            self.HideCellEditControl()
>            self.DisableCellEditControl()
>        value = self.GetCellValue(row, col)
>        if value == 'no good 2':
>            return  # cancels the cell selection
>        evt.Skip()
>
>
>    def OnEditorShown(self, evt):
>        if evt.GetRow() == 6 and evt.GetCol() == 3 and \
>           wxMessageBox("Are you sure you wish to edit this cell?",
>                        "Checking", wxYES_NO) == wxNO:
>            evt.Veto()
>            return
>
>        pprint.pprint("OnEditorShown: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>
>    def OnEditorHidden(self, evt):
>        if evt.GetRow() == 6 and evt.GetCol() == 3 and \
>           wxMessageBox("Are you sure you wish to  finish editing this
>cell?",
>                        "Checking", wxYES_NO) == wxNO:
>            evt.Veto()
>            return
>
>        pprint.pprint("OnEditorHidden: (%d,%d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetPosition()))
>        evt.Skip()
>
>
>    def OnEditorCreated(self, evt):
>        pprint.pprint("OnEditorCreated: (%d, %d) %s\n" %
>                       (evt.GetRow(), evt.GetCol(), evt.GetControl()))
>
>
>
>class MainFrame(wxFrame):
>    """ Creates and initializes the main window"""
>    def __init__(self, parent, id, title, style):
>        # First, call the base class' __init__ method to create the frame
>        wxFrame.__init__(self, parent, id, title, wxPoint(100, 100),
>wxSize(640, 480), style)
>
>        self.CreateStatusBar()
>        self.SetStatusText("This is the statusbar")
>
>        # Make menu with separate MakeMenus -method
>        menu = self.MakeMenus()
>        self.SetMenuBar(menu)
>
>        # Associate some events with methods of this class
>        EVT_MENU(self, ID_EXIT,  self.OnExit)
>
>        # Create grid
>        self.pn_grid = PnGrid(self)
>
>    def MakeMenus(self):
>        """Creates the menus of the main window"""
>
>        # Filemenu contents
>        filemenu = wxMenu()
>        filemenu.Append(ID_EXIT, "Exit", "Terminate the program")
>
>        # Collect all menus
>        main = wxMenuBar()
>        main.Append(filemenu, "File")
>        return main
>
>    def OnExit(self, event):
>        # Close program
>        self.pn_grid.Destroy()
>        self.Close(true)
>
>
>class MainApp(wxApp):
>    # wxWindows calls this method to initialize the application
>
>    def OnInit(self):
>    # Create an instance of our customized Frame class
>        # Create an instance of our customized Frame class
>        frame = MainFrame(NULL, -1, " PnGrid", wxSTAY_ON_TOP | wxCAPTION)
>        frame.Show(true)
>
>    # Tell wxWindows that this is our main window
>        self.SetTopWindow(frame)
>        return true
>
>if __name__ == "__main__":
>    app = MainApp(0)
>    app.MainLoop()
>
>



From oyving+news at pvv.org  Wed Apr 24 20:02:54 2002
From: oyving+news at pvv.org (Oyvind Gronnesby)
Date: 25 Apr 2002 02:02:54 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org> <mailman.1019682973.29139.python-list@python.org>
Message-ID: <871yd4odap.fsf@pvv.org>

* Cliff Wells
|
| On 24 Apr 2002 22:57:20 +0200
| Oyvind Gronnesby wrote:
| 
| > * Cimarron Taylor
| > |
| > | Will this do what you want?
| > | 
| > | verboseGender = ('female','male')[gender=='m']
| > 
| > Wouldn't this break in a PEP 285 Python world?
| 
| No.  That was the whole point of subclassing bool from int.

But wouldn't that just imply that True/False would turn into 1/0 if coerced
into it by int()?

-- 
?yvind Gr?nnesby
have castle, need dragon


From chbu at hrz.tu-chemnitz.de  Thu Apr 11 02:12:42 2002
From: chbu at hrz.tu-chemnitz.de (Chris Burkert)
Date: Thu, 11 Apr 2002 08:12:42 +0200
Subject: Phyton and SQL
Message-ID: <3CB5295A.4070803@hrz.tu-chemnitz.de>

Hi

I just started yesterday and took a look at the Phyton-Manual the first 
time. It's a very interesting and powerful language. But what I didn't 
find was a comfortable way to communicate with Databases like Oracle or 
MySQL through the SQL language.

Could you send me some Informations, Links, and more about my problem ?

thanks chris burkert

=> Sorry if I don't use the english language correctly.



From max at alcyone.com  Fri Apr 26 17:37:14 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 26 Apr 2002 14:37:14 -0700
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to do 
 math?)
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com> <slrnacjga3.2uv.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3CC9C88A.9B447B18@alcyone.com>

Huaiyu Zhu wrote:

> It appears that another display level is needed, somewhat between repr
> and
> str.  Let's call it 'disp' here.  It is intended to be used for
> display in a
> controlled way, esp for nested structures.

Just derive a new class and override __str__ and/or __repr__ to do what
you want.

Adding _another_ stringifying builtin function doesn't seem anywhere
near worth it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From dalke at dalkescientific.com  Wed Apr 17 21:00:26 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 17 Apr 2002 19:00:26 -0600
Subject: [OT] What is Open Source?
References: <20020418002823.C4068@phd.pp.ru> <BA2C21F8-5254-11D6-BEC7-0030657EBC3E@lanl.gov> <mailman.1019085816.3890.python-list@python.org>
Message-ID: <a9l5kf$k0e$1@slb3.atl.mindspring.net>

holger krekel:
>On Wed, Apr 17, 2002 at 04:44:50PM -0600, rasmussn at lanl.gov wrote:
>> As I understand it, the GPL does not give you right to use the software
>> as you see fit.  You can only use it by linking with other GPL'd
>> software.
>
>cant be true: gpled software under windows wouldnt be possible...

There's a clause in the GPL to allow GPL'ed software to link with
system libraries.  After all, GPL software started as code for non-GPL'ed
OSes, else no one would have been able to use the code until there was
a working kernel.

                    Andrew
                    dalke at dalkescientific.com





From aleax at aleax.it  Tue Apr 16 03:00:02 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 07:00:02 GMT
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <MtIu8.18713$b62.536574@news1.tin.it> <20020416004308.48201c4a.christophe.delord@free.fr>
Message-ID: <SZPu8.19037$b62.543870@news1.tin.it>

Christophe Delord wrote:
        ...
>> >     U = P = 0
>> >     for value in dict.values():
>> >         if value == 1:
>> >             U += 1
>> >         elif value == 2:
>> >             P += 1
>> 
>> Again I think it might be faster to avoid the if/else with yet
>> another dictionary:
>> 
>> counts = {}
>> for count in adict.values():
>>     counts[count] = 1 + counts.get(count, 0)
>> 
>> U = counts.get(1, 0)
>> P = counts.get(2, 0)
>> 
>> >     return U*S/(U*S+P*(1-S))
>> 
> 
> As indices are integers (1 and 2), a simple array may be faster :

I don't see any indication that a count may not be whatever
positive integer it pleases.  The OP's code singles out 1
and 2, but why shouldn't a certain tuple turn up 3497 times?
So, I think in your code below:

> counts = [None, 0, 0]         # counts[0] not used
> for count in adict.values():
> counts[count] += 1
> 
> U = counts[1]
> P = counts[2]

you need to guard the increment with either an if count<3 or
(probably better if large counts are unusual) a try/except
IndexError.  Once you've done that you might see some
performance increase, yes.  I tend to think of dicts for
any typical "sparse array" need in Python, but that's because
in most cases where I'm histogramming I can't ignore counts
greater than some threshold -- in this special case where
ignoring is what I want, a list and some care may indeed
be faster (you gotta benchmark and profile it of course).


Alex



From logiplexsoftware at earthlink.net  Mon Apr 29 21:48:26 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 18:48:26 -0700
Subject: Access violation trying to run Python over Win NT network.
In-Reply-To: <ucrs88tkj2ttc7@corp.supernews.com>
References: <ucrs88tkj2ttc7@corp.supernews.com>
Message-ID: <20020429184826.0eefdecb.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 21:17:55 -0400
Patrick wrote:

> I'm trying to run Python from a network drive.  On most machines it works,
> but on some machines it generates an access violation about 60% of the time,
> (just by typing 'python' with no arguments).  If I install Python locally it
> runs fine on all the machines.

Even on the machines where you get the access violations it works about 40% of
the time?

> I would really appreciate any help.  I've been trying to get the people at
> work to use Python and they have finally agreed to give it a try, but they
> don't want to install it on everyones machines, so we have to run from a
> network drive.

I've done this before without too many problems.  The usual problem was the
generation of .pyc files.  If the user doesn't have write permission on the
file/directory then it fails.  It seems odd that you would get access
violations simply by typing "python", but perhaps when python runs it tries to
byte-compile some modules and fails (very unsure of this, although I'm guessing
there are platform modules that might need to be compiled at startup).  You
might try compiling all of the modules yourself and see if that fixes it.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From aleax at aleax.it  Sat Apr 13 03:58:34 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 07:58:34 GMT
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl>
Message-ID: <KyRt8.11643$b62.313960@news1.tin.it>

Jozef wrote:

> I was reading on functional programming in Python and wanted to test
> myself by writing a function:
> 
> functional_euler = lambda n: \
> len(filter(lambda i: gcd(i, n) == 1, range(1, n + 1)))
> 
> (gcd is the greatest common divisor function). It gives the following
> error: <stdin>:1: SyntaxWarning: local name 'n' in 'lambda' shadows use of
> 'n' as global in nested scope 'lambda'
> 
> I am not really sure even what this means... does lambda create a new
> scope or something? Is there another way to write this?

You're clearly using Python 2.1.something and don't have a:

    from __future__ import nested_scopes

at the very top of your module.  Therefore, the Python compiler does
not give you lexically nested scopes but does warn you about "shadowing"
of locals vs global nested-scope identifiers -- i.e., spots where
lexically nested scopes will change your program's semantics when
you enable them.

Either insert the 'from __future__', or move to Python 2.2.1 -- or both.

Yes, lambda (like all other function definitions) does create a new
scope.  List comprehensions don't, and they're just as 'fp' -- indeed,
Python copied them right from one of the "functionalest" programming
languages to be reasonably widespread, Haskell, just adjusting the
syntax a little bit (more keywords, less punctuation).

I like FP a lot, and I practise it -- but I eschew lambda in Python
in most cases: it's limited and goes much against the grain of the
language.  All it boils down to is, give your functions a name -- in
most cases a well-chosen name will clarify your code.  E.g.,

def relatively_prime(a, b): return gcd(a, b) == 1

and so on...


Alex



From lac at strakt.com  Sat Apr  6 13:39:41 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 06 Apr 2002 20:39:41 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from "Mark McEahern" <marklists@mceahern.com> 
   of "Sat, 06 Apr 2002 11:57:07 MDT." <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> 
References: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> 
Message-ID: <200204061839.g36Idfdj017724@ratthing-b246.strakt.com>

Mark McEahern writes:
> [Laura Creighton]
> > [...]  I am rejecting the existence of booleans in the language
> > because it is too hard to teach how to use them properly.
> 
> How do you use them properly?
> 
> Suppose I want to validate users.  I might write something like this:
> 
> 	def isValid(user, password):
> 		if ...:
> 			return True
> 		else:
> 			return False

My pleasure, especially since this is such a cool example.  You have
just shot yourself in the foot, and do not know it.  When you have
shipped your isValid function, across the world to your 10,000 clients,
you will discover this.

Your boss comes in and says, oh, Mark, we need a change to that
isValid function.  We need to return 3 states, Valid, Invalid, and
Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.

What do you do now?  

> Using isValid, that returns one of two possible answers, is as clear and
> bright as day to me.  If later on I want a function that does something
> different, it would most definitely be named something different.  Why on
> earth would I want a function named isFoo to return one of more than two
> possible values?  (Your answer is probably that you never want a function
> named isFoo.  Hmm, I still don't see why.)

Do you now? Use Booleans when you want to return ON and OFF -- the
state of that bit in memory, of something that is constrained by
physical reality, or mathematical reality to only have 2 values, ever,
cross my heart and the world will end if this ever ceases to be the
case.  Don't artificially limit yourself to 2 values because your
vision at the time only extended to 2 values.  The one thing constant
in this world is change.   If you don't need a third value, somebody 
else will, later.  Integers are your friends.

Laura Creighton




From aleax at aleax.it  Wed Apr 10 06:07:18 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 10:07:18 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1018301543.25726.python-list@python.org> <b5ys8.62201$pT1.1789842@news1.tin.it> <mailman.1018388430.20332.python-list@python.org>
Message-ID: <q9Us8.2287$vF6.84177@news2.tin.it>

Lulu of the Lotus-Eaters wrote:

> |Lulu of the Lotus-Eaters wrote:
> |major = len([1 for f in fs if f(o)]))
> 
> Alex Martelli <aleax at aleax.it> wrote:
> |I personally prefer:
> |major = len(filter(None, [f(o) for f in fs]))
> 
> Why?  I had thought of basically that one.  But the short "if f(o)"
> clause seemed to read better than the longer "filter(None..)" construct.

Musing along in introspection...:

If it wasn't for the need of having meaningless contents in the
list built by the comprehension, maybe.  It takes an epsilon of
reflection to realize "the contents are irrelevant", which for
some reason the idiom starting with "len(filter(None, " does not
seem to require of me (maybe because I've seen that idiom so many
other times).

Funny enough, [f for f in fs if f(o)] reads better to me -- maybe
because it DOES a more similar job to filter, rather than choosing 
"arbitrary" contents.  A subtle issue indeed, given the len( ...
clarifies at once that the contents don't matter, and yet that's
how it feels to me.

> I don't really care one way or another about mixing filter() with
> listcomps, but the few extra keystrokes in Alex' don't seem to add to
> clarity.

Certainly not to a hypothetical reader not familiar with filter(None, ...
(who'd have to look it up).  Just as apply is close to obsolete due
to the * and ** forms for actual arguments, so maybe I should try to
think of filter and map similarly -- and yet while I love list
comprehensions, filter and map appear so natural to me when they don't
need a (yecch) lambda.


> Another option is to squeeze a reduce(operator.add...) into there.  But
> practically, len() seems to do everything that needs doing.

Basically MLH's suggestion, right?  No, it seems to me that len is
quite superior to +/ (as we spelled reduce(add... back in APL) here --
simpler, more obvious, and not needing ones and zeroes specifically
(faster, too -- minor consideration of course, but...).


Alex



From pieroul at attglobal.net  Tue Apr 30 22:59:03 2002
From: pieroul at attglobal.net (Pierre Rouleau)
Date: Tue, 30 Apr 2002 22:59:03 -0400
Subject: Any CD-ROM writer support in python?
Message-ID: <3CCF59F7.D5AA5EE4@attglobal.net>

I am looking for python based CD-ROM writer support.
Does anyone know where I could find some python code or libraries to do
that?

I want to write a script to control a CD-ROM writer to store data
(really to create a software installation CD-ROM).
I have a complete software build system controlled by Python script and
the last (and missing) step is to integrate
the CD-ROM burning.  Our software build runs at night and I'd like to
complete the process by automating the
creation of a re-writable CD-ROM in the script.

I could use a command-line based CD-ROM writer application and control
it from the Python script, or have the whole thing done in Python (with
possibly some library support).  I have found one command-line
application so far, but it's not open source.  I don't have any pure
python-based solution.

I am open to any suggestion!

Thanks!


--
        Pierre Rouleau




From roy at panix.com  Tue Apr 30 07:45:34 2002
From: roy at panix.com (Roy Smith)
Date: Tue, 30 Apr 2002 07:45:34 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1>
Message-ID: <roy-E96293.07453430042002@news1.panix.com>

Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:
> The problem with int.chr() is that a lot of people are going to get 
> confused when they find that "42.chr()" doesn't work (although "42 .chr()" 
> does).

What about "(42).chr"?  Wouldn't that work?  Seems a lot more natural than 
putting a space in there (which I didn't even realize was legal, and looks 
horrible to me).


From jepler at unpythonic.net  Wed Apr 24 12:14:54 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 24 Apr 2002 11:14:54 -0500
Subject: Subclass of int and initialisation with not integer values
In-Reply-To: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr>
References: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr>
Message-ID: <20020424111450.B9269@unpythonic.net>

On Wed, Apr 24, 2002 at 05:09:53PM +0200, Boris Boutillier wrote:
> Hi all,
> 
> I've got a little problem with subclass of int.

You must create a __new__ method.  See the descrinfo.html talk about
how it works.


class toto(int):
    true, false = 'O', 'N'
    def __init__(self, str):
	self.str = str

    def __new__(cls, val):
	if val == cls.true: 
	    return int.__new__(cls, 1)
	elif val == cls.false:
	    return int.__new__(cls, 0)
	else:
	    raise ValueError

    def __str__(self):
	return self.str

print toto('N')
print toto('O')
print toto(0)

Jeff




From distrex at hotmail.com  Mon Apr  8 21:24:12 2002
From: distrex at hotmail.com (distrex)
Date: 8 Apr 2002 18:24:12 -0700
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com> <a8q9bi$2vka$1@news.idiom.com> <d9ea2696.0204071920.1b9d75c4@posting.google.com> <a8r62k$20b$1@peabody.colorado.edu>
Message-ID: <d9ea2696.0204081724.62a1038f@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<a8r62k$20b$1 at peabody.colorado.edu>...
> distrex wrote:
> 
> > Don't take this the wrong way but I am a newbie and didn't understand
> > 86% of what you just said. All I got was modules are classes functions
> > and variables. and a few other things but that was to complicated for
> > me (the newbie)
> > 
> 
> Don't despair. The above was one of the best descriptions of modules I've ever 
> read (but I'm not a newbie), to the point where I'd strongly encourage its 
> author to contact the doc team and push for its inclusion in the official 
> docs.
> 
> Just have a bit of patience, read it over again, slowly, go through the whole 
> tutorial doing the examples and it will make sense. You may be a bit green 
> still but it will come quickly, I'm sure. Python is one of the easiest 
> languages on earth to learn: you'll be up and running in no time at all, and 
> having fun programming!
> 
> Cheers,
> 
> f.


I've had fun with python I just want to see if Ican make it more fun
with the modules so I'll just keep looking for a good explaination of
ALL the modules.
thanx for all your help


From ole.fostad at netcom.no  Sun Apr 14 06:17:32 2002
From: ole.fostad at netcom.no (Ole Fostad)
Date: Sun, 14 Apr 2002 12:17:32 +0200
Subject: can't see list text of a wxPython ListCtrl widget
References: <6bd9f01b.0204131604.6031da9e@posting.google.com>
Message-ID: <a9bkvu$bht$1@lise.netcom.no>

Hi!

Replace
"self.list.SetStringItem(0, 0, 'ho')" with "self.list.InsertStringItem(0,
'ho')" and
"self.list.SetStringItem(1, 0, 'ho')" with "self.list.InsertStringItem(1,
'ho')"

Ole

"sameer" <sameer_ at email.com> wrote in message
news:6bd9f01b.0204131604.6031da9e at posting.google.com...
> hi,
>   I am playing around with the wxPython ListCtrl widget, and for some
> reason I can't figure out, I am not seeing the text I add to the list.
>  I can see the Column Names just fine, just not any item in the list.
> I am pasting the important code below which is part of an init method
> of a wxPanel.
>
>         tID = wxNewId()
>         self.list = wxListCtrl(self, tID, pos = (5, 5),
> style=wxLC_REPORT)
>         self.list.SetDimensions(5, 5, 400, 200)
>         self.list.InsertColumn(0, "Subject")
>         self.list.InsertColumn(1, "Entry")
>         self.list.InsertColumn(2, "Date")
>         self.list.SetColumnWidth(0, 100)
>         self.list.SetColumnWidth(1, 100)
>         self.list.SetColumnWidth(2, 100)
>         self.list.SetStringItem(0, 0, 'ho')
>         self.list.SetStringItem(0, 1, 'b...')
>         self.list.SetStringItem(0, 2, '0-2002')
>         self.list.SetStringItem(1, 0, 'ho')
>         self.list.SetStringItem(1, 1, 'b...')
>         self.list.SetStringItem(1, 2, '0002')
>
>
>
> Thanks in advance for the help.




From donn at drizzle.com  Thu Apr 11 03:06:10 2002
From: donn at drizzle.com (Donn Cave)
Date: Thu, 11 Apr 2002 07:06:10 -0000
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer   division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <Xns91ECDF3AC49F7lumberjack@216.148.53.85> <3CB5221D.EE464B76@engcorp.com>
Message-ID: <1018508767.552240@yasure>

Quoth Peter Hansen <peter at engcorp.com>:
...
| In any case what's your point?  You have every right, and 
| apparently the strong motivation, to fork off your own branch
| of the language and continue development in your own fashion.
| Surely that will let you demonstrate how to bring legitimacy
| in ways the current project leads have not been able...

His point might have been more clearly articulated, but there
certainly was a point.  Look for "will undermine the legitimacy
of Python", in the part you omitted.

"Love it or leave it" bumper stickers used to be more common in the US,
a few years back, in response to vociferous criticism of US foreign
policy at the time.  The criticism was often unfair, but then US foreign
policy was too easy a target.  I hope I never see another one of those
bumper stickers.

The hostility directed towards Mr. Lumberjack suggests that maybe he
was smart to use a nom de plume for this message.  Tomorrow he can
post under his real name and interact on amicable terms with the
people who want to chase him off today for criticizing our leader.
That's fine, I guess, as long as you all are having fun, but don't
kid yourselves that this is any kind of a service to the community.
It's actually kind of a sorry picture.

	Donn Cave, donn at drizzle.com


From chris at psychofx.com  Tue Apr 30 16:41:17 2002
From: chris at psychofx.com (Chris Miles)
Date: Wed, 1 May 2002 06:41:17 +1000
Subject: PSA/PSF, pyZine international payments
In-Reply-To: <aamrko$a8v$1@reader1.tiscali.nl>; from gmuller@worldonline.nl on Tue, Apr 30, 2002 at 09:34:18PM +0200
References: <aamrko$a8v$1@reader1.tiscali.nl>
Message-ID: <20020501064117.B2904@psychofx.com>

I am an Australian, currently living in London, and bought my issue of
Py very easily using Paypal.  Paypal supports international accounts now,
and it is very handy for paying for things in US currency.

CM


On Tue, Apr 30, 2002 at 09:34:18PM +0200, GerritM wrote:
> Living in Europe it is still amazingly difficult to transfer money from a
> local bank account to an USA account (Note that despite the introduction of
> the Euro even transfers crossing borders is still difficult). It would be
> nice to have some easy way to pay PSF (PSA?) and order pyZine from Europe.
> 
> For Dutch readers would the PSF be seen as charity and therefor as tax
> deducable?

-- 
Chris Miles
chris at psychofx.com
http://www.psychofx.com/chris/




From python at dejazzd.com  Wed Apr 17 20:43:41 2002
From: python at dejazzd.com (Glenn Stauffer)
Date: Wed, 17 Apr 2002 20:43:41 -0400
Subject: DCOracle2 importing problems
In-Reply-To: <e2e9e807.0204171619.46de6db1@posting.google.com>
References: <e2e9e807.0204171619.46de6db1@posting.google.com>
Message-ID: <200204172043.41851.python@dejazzd.com>

If you are having the same problem I've had, it has something to do with the 
linking of the DCOracle2 library, I think.  I didn't figure out what I needed 
to do during compile to statically link in the Oracle libraries and instead 
use os.environ to set the LD_LIBRARY_PATH and ORACLE_HOME:

os.environ['ORACLE_HOME'] = '/app/oracle/product/8.1.6'
os.environ['LD_LIBRARY_PATH'] = '/app/oracle/product/8.1.6/lib'

Not sure if I tested this, but you may only need to set the LD_LIBRARY_PATH.

I only had this problem on a Compaq Tru64 server (not on Linux or Sun).  
Eventually I decided to use cx_oracle instead of DCOracle2 for most of my 
programs.

Glenn Stauffer

On Wednesday 17 April 2002 08:19 pm, Hugh wrote:
> Hi All, I can import the DCOracle2 module fine from the interactive
> interpreter but my CGI script causes an internal server error. The
> interactive interpreter is running on the same machine as my scripts.
> A tail of the apache error log reveals that my script can't import the
> DCOracle2 module.
> Any ideas anyone?
>
> Hugh





From frr at easyjob.net  Wed Apr 24 03:16:08 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Wed, 24 Apr 2002 09:16:08 +0200
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net> <0kiccu46a2580bbohtpavbla241q95etj0@4ax.com>
Message-ID: <1nmccuogt7t9ler1eqq4kec95bhr0mcchr@4ax.com>

On Tue, 23 Apr 2002 23:05:59 -0700, Tim Roberts <timr at probo.com> wrote:

>Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote:
>
>>In article <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v at 4ax.com>, Fernando Rodr?guez wrote:
>>> No. Step (b) processes the order and does give feedback. Step (a) saves in a
>>> db some info conatined in a cookie (date of first visit and first referrer). I
>>> want step (a) to be seamless.
>>
>>Must admit I don't have a clue what you mean. Why aren't scripts (a)
>>and (b) in the same script? If they really must be separate scripts,
>>why doesn't script (a) just output nothing at all to stdout?
>
>My guess is that the (b) option is a URL at a third-party order processing
>site that is not under his control. 

Yep. :-P

> If my guess is correct, that means the
>original poster is actually hoping to have a single <FORM> fire to two
>different URLs.  

Or anything similar. O:-)

>Theoretically, it would be possible to have the <FORM> fire a URL on the
>home system with the order processing URL, and have that script generate
>its own HTTPS connection to the order processing site, but that means the
>web site user will not see that his connection is secure. 

Why? If everything is done in a SSL connection...



-----------------------
Fernando Rodriguez


From zeitlin at seth.lpthe.jussieu.fr  Tue Apr  2 04:48:28 2002
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Tue, 2 Apr 2002 09:48:28 +0000 (UTC)
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <a84561$p5g$1@wanadoo.fr> <slrnaabbct.npb.cpbotha@crabtree.cpbotha.net> <4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com> <20020401020912.GE333@lilith.hqd-internal> <mailman.1017681605.21669.python-list@python.org> <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net>
Message-ID: <slrnaaivjc.565.zeitlin@seth.lpthe.jussieu.fr>

On Mon, 01 Apr 2002 19:04:16 GMT, Charl P. Botha <cpbotha at i_triple_e.org> wrote:
> Try getting a running wxPython installation on SGI Irix or SUN Solaris for
> instance.

 And just what exactly is the problem with it? I don't have access to SGI but
it did build without any problems on a Solaris box here. If you have
encountered any problems, have you reported them to us (wxWindows developers)
as bugs? I'm sorry to say this, but you have to consider spending some of your
time on writing (useful) bug reports - without the users help there is
absolutely no chance that wxPython or wxGTK builds out of the box on the
platforms we don't have access to. It's amazing by how much the number of
complains about the difficulties of building wxGTK exceeds the number of the
bug reports (of which there are maybe 1 or 2 right now).

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From a.clarke11 at ntlworld.com  Sat Apr 20 14:38:34 2002
From: a.clarke11 at ntlworld.com (tony clarke)
Date: Sat, 20 Apr 2002 19:38:34 +0100
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com>
Message-ID: <3CC1B5A9.E5E1FDBF@ntlworld.com>

Hi

Yes, there is a way out: defining the function outside the button
definition. This unfortunately can require careful workarounds when the
function is detailed, eg with specific arguments, but unfortunately that
is what Tkinter asks us to do, this version at least.
tonyclarke(nospam)@www.com

Peter Saffrey wrote:

> When I place some buttons on a frame I don't seem to be allowed to use
> arguments in the function call attached to the button. If I do use
> arguments, or even just empty brackets, the function is called when
> the button is placed, not when the button is clicked on.
>
> Below is some simple code which demonstrates this. Is there a way
> around this problem?
>
> Peter
>
> #!/usr/bin/python
>
> from Tkinter import *
>
> def printNumber(number):
>         print number
>
> win = Frame()
> win.pack()
> Button(win, text='print 3', command=printNumber(3)).pack()
> Button(win, text='print 5', command=printNumber(5)).pack()
>
> win.mainloop()



From steingrd at ifi.uio.no  Tue Apr 16 15:54:22 2002
From: steingrd at ifi.uio.no (Steingrim Dovland)
Date: Tue, 16 Apr 2002 19:54:22 GMT
Subject: Counting the number of elements in a list - help.
References: <a9hu8e$2fg$1@lumberjack.rand.org>
Message-ID: <m3n0w3fmhf.fsf@localhost.localdomain>

"Bryce C. Mason" <bmason at rand.org> writes:

> 1) a function which returns the number of elements in a list
> 2) a function which takes a list and an index and returns a zero when that
> list index is out of range.

I think you're looking for len(), a built in function.

>>> x = [1,2,3]
>>> len(x)
3

-- 
Here the sea spray give
I was with her
We're under the ship so get me over


From logiplexsoftware at earthlink.net  Tue Apr  9 16:51:18 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 9 Apr 2002 13:51:18 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <000001c1dfce$4a2ffb00$0334fea9@carol>
References: <000001c1dfce$4a2ffb00$0334fea9@carol>
Message-ID: <20020409135118.6fe5438e.logiplexsoftware@earthlink.net>

On Tue, 9 Apr 2002 09:51:01 -0400
Arthur Siegel wrote:

> Where my own sensibilities happen to explode is when I perceive a good
> portion of the community defending Guido's
> position when it is based on thin evidence, is non-technical, and represents
> no more than one guy's take based on the amount of attention he happens to
> have been able to give the matter, as they do when it is based on an
> articulated position on technical matters - where I would agree the burden
> of proof most definitely shifts.

And this is a valid position (one that I tend to take).  However some of your
earlier remarks were not of this nature:

[from your earlier post]
> The fact as to who's football we happening to be playing
> with now being a little shoved in our face.  Guido seemed
> willing to accept the sense of the community, along as the
> community told him what he wanted to hear. When
> it didn't, he simply retroactively rephrased his question.

This was a direct attack on Guido's integrity, which was uncalled for.  You
*assumed* much about what he meant (perhaps he didn't make himself clear, but
that's hardly grounds for labeling him a dictator) and when he clarified his
position he became a liar.

While it is clear that much misunderstanding went on, it seems you went too far
in your interpretation of what Guido meant (as did I, but I prefer to confront
people personally and give them a chance to explain rather than hang them in
public over a misunderstanding).  Apparently there will be no winning for Guido
on this one:  you acknowledge (as does he) that he doesn't always make himself
clear, but when he does try to clarify he gets accused of "retroactively
rephrasing" which I take as a euphemism for "lying" in this context (but I'll
be willing to reconsider in case you want to "retroactively rephrase" that
statement ;)

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mcfletch at rogers.com  Tue Apr 23 11:04:35 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 23 Apr 2002 11:04:35 -0400
Subject: Detecting a new style class
References: <1019551994.576.3.camel@tp6>
Message-ID: <3CC57803.70302@rogers.com>

For determining whether an instance' class is a new-style class, check 
for explicit sub-classing of object.

	issubclass( self.__class__, object )

 >>> class R(object):
... 	pass
...
 >>> issubclass( R, object )
1
 >>> class R(x, object):
... 	pass
...
 >>> issubclass( R, object )
1
 >>> class R(object,x):
... 	pass
...
 >>> issubclass( R, object )
1
 >>> class R:
... 	pass
...
 >>> issubclass( R, object )
0
 >>> issubclass( property, object )
1

(Note that isinstance won't work, as it always seems to return 1 even 
for instances of classes which return 0 for issubclass).

HTH,
Mike

Graham Ashton wrote:
> Hi. I've got some code tha only works properly for classic classes.
> Before I get round to fixing it, I'd like to put some protection in
> place to prevent it from being used with a new style class.
> 
> This means that I'd like to be able to detect whether my class is being
> subclassed by a new or classic class. To do that I'm testing self to see
> if it has a __new__ attribute. Is there a better way to do it?
> 
> I find that the behaviour varies depending on the order in which a
> class's super classes are defined. In the example below, I'd like the
> instantiation of class B to throw the same RuntimeError as A:
> 
...
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From nobody at nobody.com  Mon Apr  8 16:35:07 2002
From: nobody at nobody.com (John)
Date: Mon, 8 Apr 2002 15:35:07 -0500
Subject: CGI multiple select?
Message-ID: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>

Hi,

In my html form, I have 'multiple select'

<select name=spred multiple>
<option..>
<option..>


This form invokes CGI routine, where I try to read selected entries from
form["spred"]

My dilemma is here:

When only one entry is selected:
form["spred"].value (will be the object selected)

When multiple entries were selected:
for o in form["spred"]:
  o.value (--> each of these will be the object selected)

Thus, I need to handle according to situation.. I though my previous
question on 'isList' can solve this matter, but somehow, all solution
suggested (isSequenceType, type, ..) cannot distinguish these two cases.






From mwh at python.net  Tue Apr  9 12:00:33 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 9 Apr 2002 16:00:33 GMT
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no>
Message-ID: <usn64ame1.fsf@python.net>

Syver Enstad <syver-en+usenet at online.no> writes:

> "Fredrik Lundh" <fredrik at pythonware.com> writes:
> 
> > Sean 'Shaleh' Perry wrote:
> > > " ".join(mylist) # is this perl or python?
> > 
> > *why* string objects have join method has been explained
> > over and over and over again, and should also be obvious for
> > anyone with some knowledge how strings and sequences work
> > in Python.
> > 
> > > whoever decided this is the new style should be shot, a lot
> 
> Am I the only one in the entire world that thinks '\t'.join(mylist) is
> elegant and natural?

No.

Cheers,
M.

-- 
 Very clever implementation techniques are required to implement this
 insanity correctly and usefully, not to mention that code written
 with this feature used and abused east and west is exceptionally
 exciting to debug.       -- Erik Naggum on Algol-style "call-by-name"


From jeff at ccvcorp.com  Tue Apr  9 15:45:03 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 9 Apr 2002 12:45:03 -0700
Subject: Redhat 7.3 skipjack Beta2- why still python 1.5x ?
References: <899f842.0204081714.25eab798@posting.google.com>
Message-ID: <MPG.171ce49d11c6086a9896a2@news.nwlink.com>

In article <899f842.0204081714.25eab798 at posting.google.com>, 
anthony_barker at hotmail.com says...
> Python 2.2x is stable why does redhat still insist on releasing 1.5x?
> All other distributions are now on 2.x.
> 
> It seems from the Redhat's skipjack beta 2 that it will only ship the
> older version of python(1.5.2).

Because RedHat has a stated policy of shipping the same version 
of all major utilities with *all* minor releases within a 
specific major release of RedHat.  In other words, all of their 
major utilities will use the same version through each of RH 7.0, 
7.1, 7.2, 7.x ...

Python qualifies as a major utility in this sense because RH uses 
it in a significant number of its system configuration and 
administration tools.
 
> I find this annoying because it means I have to upgrade each redhat
> machine individually as it breaks a couple of my scripts.

Be aware that, since RH uses Python 1.5.2 for its tools, and in 
several cases uses deprecated modules that no longer function in 
Python 2.x, you'll need to be careful about upgrading -- you need 
to leave /usr/bin/python pointing to 1.5.2, and install 2.x as 
/usr/bin/python2 (RH does provide rpms that are configured to do 
this; those using RH should definitely get RH's rpms.)  It leaves 
you in an awkward position of having to watch that scripts get 
executed with the correct version of Python (since you can't 
presume that 'python myscript.py' will execute the latest 
version), but the only alternative is to change all of RH's 
scripts to explicitly point to /usr/bin/python1.5 (which could be 
quite a task).

[x-posts trimmed]

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jochen at jochen-kuepper.de  Sun Apr 28 14:45:12 2002
From: jochen at jochen-kuepper.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 28 Apr 2002 14:45:12 -0400
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com>
Message-ID: <m3y9f7hdc7.fsf@box.home.de>

On 27 Apr 2002 10:01:21 -0700 Terry Byrne wrote:

Thanks for your posting, but I don't understand the following:

Terry> Then I used the keyword "pass" whenever I ran an re since each
Terry> line can contain only one error; after running an re it's not
Terry> necessary to look for other error codes at all.

You are passing 'pass' to compile or search, or what are you doing
here?  Is this documented anywhere?  I could neither find it in the
manual nor the re HOWTO; maybe it's just me, though:O

Btw: An index would serve the HOWTO well:)) It is pretty hard to find
information in there in a non-linear way (that is, without reading all
of it over and over).  It's a nice document, but it is a little
unorganized.

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From rgwright at ux11.cso.uiuc.edu  Wed Apr  3 12:46:54 2002
From: rgwright at ux11.cso.uiuc.edu (robert garretson wright)
Date: Wed, 03 Apr 2002 17:46:54 GMT
Subject: Breaking String into Values
References: <HD5p8.33836$tg4.398811@vixen.cso.uiuc.edu> <MPG.170ea2709d1d2965989685@news.nwlink.com>
Message-ID: <ieHq8.35805$tg4.422078@vixen.cso.uiuc.edu>

Jeff Shannon <jeff at ccvcorp.com> wrote:
> rgwright at ux.cso.uiuc.edu says...
>> I am working on reading in a data file format which is set up as a series
>> of lines that look like this:
>> 
>> 3500035000010104A Foo 45
>> 
>> I want to break up into a variables as follows:
>> a = 35000, b = 35000, c = 10104, d = 'A', e = 'Foo', f = 45
>
> So, I would have something like this:
>
> format = { a: (0,5),
>            b: (5,10),
>            c: (10,15),
>            ....        }
>
> class Record:
>     def __init__(self, data)
>         for field,zone in format.keys():
>             setattr(self, field, data[zone[0]:zone[1]])
>
> If you want to do error checking of most of the fields, I'd 
> define a series of functions that do whatever checking and/or 
> converting you need, and include those in the dictionary:
>
> def ConfirmValidId():
>     ....
>
> format = { a: (ConfirmValidId, 0, 5),
>            b: (int, 5, 10),   ... }
>
> Then your setattr line becomes:
>
>         setattr(self, field, zone[0]( data[zone[1]:zone[2]] ) )

Thanks for the help everyone (this is sort of a group reply). What I am
working on is the Census Tiger/LINE database which is quite large. Each
is composed of a number of region (Virginia has 135). Each region has 17 
database tables. 

The first table (RT1) is:
10301 103802945 K  Loch                          Cir   A41        299          1
        298          200002366923669              51516506509382793827          
350003500001010401010410051000 -76314472+37052074 -76312226+37052312
followed by 8000+ similar lines.

The region I am testing on has 5 Megs of database tables that have to be read
in. The entire Tiger/LINE database is about 1 gig compressed.

I cleaned up my original code (removing all but a few of the temp variables)
into something like this. ToInt changes a string to an integer with error
checking and returning None when the string is blank.

class RT1Line:
    def __init__(self,line):
        TI = ToInt
        # Record Type
        temp = line[0:1].rstrip()
        if not temp.isalnum:
            raise ParseError('Blank value not allowed')
        self.rt = temp
        # Version Number
        temp = line[1:5]
        if not temp.isalnum:
            raise ParseError('Blank value not allowed')
        self.version = TI(temp,'version')
        # TIGER/LINE ID, Permanent Record Number
        temp = line[5:15]
        if not temp.isalnum:
            raise ParseError('Blank value not allowed')
        self.tlid = TI(temp,'tlid')
        # Single Side Source Code
        self.side1 = TI(line[15:16],'side1')
        # Linear Segment Source Code
        self.source = line[16:17].rstrip()
        ...

This read in the RT1 file in 14.0 seconds.

Next I abused the struct module to split up the string. This also got rid
of all of the temp variables. The struct unpack line is horrible but the file
is generated by another program. The error checking and cleanup code looks
better though.

class RT1Line:
    def __init__(self,line):
        TI = ToInt
        v = struct.unpack('1s4s10s1s1s2s30s4s2s3s11s11s11s11s1s1s1s1s5s5s5s5s1s1
s1s1s2s2s3s3s5s5s5s5s5s5s6s6s4s4s10s9s10s9s',line[:-2])
        # Record Type
        if not v[0].isalnum:
            raise ParseError('Blank value not allowed')
        self.rt = v[0]
        ...

This read in the RT1 file in 13.1 seconds.

Finally I tried the your method of keeping the format information in a 
dict, and using setattr. This required that I make a cleanup / error checking
function for every type of value.

format = { 'rt': (NoBlankLeftAlpha,0,1),
           'version': (NoBlankInt,1,5),
           'tlid': (NoBlankInt,5,15),
           'side1': (Int,15,16),
           ...

class RT1Line:
    def __init__(self,line):
        for field,zone in format.items():
            setattr(self, field, zone[0]( line[zone[1]:zone[2]] ) )

This read in the RT1 file in 23 seconds.

I think that I will go with the struct.unpack way. It is relatively clean
and since I generate these files with other code the nasty unpack line
is not so bad.

Thanks to everyone who responded.

-- 
Robert Wright
rgwright at uiuc dot edu


From ponderor at lycos.com  Tue Apr  9 11:56:50 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 9 Apr 2002 10:56:50 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr  9)
Message-ID: <0E2F4026AED42F15.EC9A6128A5A36F18.229A5A824D55F289@lp.airnews.net>

QOTW:  "If you have numerous objects that are conceptually similar, so
that it makes sense to name them sequentially, then you *really* want
to put them in a list or dictionary instead.

The same thing is true of almost *every* use of exec or eval() -- it
seems handy, but if you look for it, there's a better solution to your
real problem."   Jeff Shannon

    
    PEP 285: during a week in which sqrt(False) was computed in all
    seriousness, we can safely conclude bool is not a black and white issue.
        http://groups.google.com/groups?q=%22PEP%20285&hl=en&sa=N&tab=wg
        http://mail.python.org/pipermail/python-list/2002-April/097312.html
    
    A quick round of options for printer output
        http://mail.python.org/pipermail/python-list/2002-April/097043.html
    
    Once again, for the record: the mandatory change in the way division
    works is going to be in 3.0.
        http://mail.python.org/pipermail/python-list/2002-April/097302.html
    
    A survey contrast of Python and C++
        http://mail.python.org/pipermail/python-list/2002-April/096465.html
    
    Large project wxPython testimonials
        http://mail.python.org/pipermail/python-list/2002-April/097175.html
    
    Alex Martelli: "Why don't we start thinking of a similarly dual-tracked Python -- one track aiming at stability ...  one aiming at frequent releases, innovation,
    etc."
        http://mail.python.org/pipermail/python-list/2002-April/096866.html
    **Discussion moved to Python-Dev:**
        http://mail.python.org/mailman/listinfo/python-dev
    
    _The Perl Review_ does Python
        http://mail.python.org/pipermail/python-list/2002-April/096602.html
    Python interest group for the Los Angeles area
        http://mail.python.org/pipermail/python-list/2002-April/096896.html
    
    
========================================================================

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

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

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

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

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

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

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


-- The Python-URL! Team--

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

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


From kalle at gnupung.net  Tue Apr  9 18:35:50 2002
From: kalle at gnupung.net (Kalle Svensson)
Date: Wed, 10 Apr 2002 00:35:50 +0200
Subject: Easter Eggs
In-Reply-To: <fm2s8kKkXAiR092yn@gnosis.cx>
References: <fm2s8kKkXAiR092yn@gnosis.cx>
Message-ID: <20020409223550.GC16263@proton.lysator.liu.se>

[Lulu of the Lotus-Eaters]
> Who can tell me what's in there.  Someone on c.l.py recently let drop
> the hint about finding the Zen of Python at the interactive prompt in
> 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
> 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
> magic command was.  Can someone remind me, please?

"import this", IIRC.

Med st?rsta h?gaktning,
  Kalle
-- 
Kalle Svensson (kalle at gnupung.net) - Laziness, impatience, hubris: Pick two!
English: http://www.gnupung.net/  Svenska: http://www.lysator.liu.se/~kalle/
Stuff: ["http://www.%s.org/" % x for x in "gnu debian python emacs".split()]




From ak at silmarill.org  Mon Apr 15 20:53:33 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 16 Apr 2002 00:53:33 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <slrnabmtk0.a8q.ak@ak.silmarill.org>

In article <slrnabmoci.dhk.philh at comuno.freeserve.co.uk>, phil hunt wrote:
> 
> OK, I've decided to write a PEP. Preliminary to this, I'd like 
> people's comments.
> 
> I wish to make the pas startement optional. That is, in any Python 
> program where there is a pass it can be left out and the meaning of 
> the program is unchanged.
> 

[snip]

I think the idea behind pass is that usually not having anything in
a conditional is a mistake, so we want to be informed about it. Putting
pass is like saying "I'm sure that's what I want."

I don't think pass is ugly. In fact, empty space after : feels wrong.


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From robin at jessikat.fsnet.co.uk  Tue Apr  9 08:15:15 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 9 Apr 2002 13:15:15 +0100
Subject: PEP 285: Adding a bool type
References: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk>
 <a8uh5p$nu7$1@thorium.cix.co.uk>
Message-ID: <zpF9N5ATtts8Ewc0@jessikat.demon.co.uk>

In article <a8uh5p$nu7$1 at thorium.cix.co.uk>, gbreed at cix.compulink.co.uk
writes
>Robin Becker wrote:
>
>> This article in salon 
>>         http://www.salon.com/tech/feature/2002/04/08/lehman/print.html
>> 
>> about Manny Lehman, my old department head, gives an insight into why
>> people like myself are often afraid of software change.
>
>And why Python is an exception
>
>"""
>Michael Godfrey, a University of Waterloo scientist, is equally hesitant 
>but still finds the Lehman approach useful. In 2000, Godfrey and a fellow 
>Waterloo researcher, Qiang Tu, released a study showing that several 
>open-source software programs, including the Linux kernel and fetchmail, 
>were growing at geometric rates, breaking the inverse squared barrier 
>constraining most traditionally built programs. 
>"""
>
>
>                                 Graham

Lines of code is a poor measure if that's what's growing exponentially. 

Anyhow growth isn't the problem, if the bugs are growing exponentially
as well then nothing is achieved. M$ Windows has certainly grown
exponentially in usage if not in code size at least of the last decade.

Python isn't growing exponentially, but the existence of multiple
versions of python with small incompatibilities in odd corners of the
language is certainly an example of what leads to uncertain behaviour.

The Python developers seem unhappy that many want to stick with 1.5.2
(eg Redhat), but users of any software are not required to follow the
research interests of others. 

Scripts in sed or awk that are 20 years old still run because their
languages are stable. Python scripts that are 5 years old don't run. The
rate of change is increasing precisely because product developers think
continually changing them is advantageous. New versions always create
update fever further down the food chain which is why intermediary users
get so fed up with continual minor adjustments that cause software
uncertainty.
-- 
Robin Becker


From jeff at ccvcorp.com  Wed Apr  3 13:30:30 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 3 Apr 2002 10:30:30 -0800
Subject: Someone using PyGtk
References: <3ca9d7d1$1_2@news.uni-bielefeld.de> <MPG.1713c8b646c14e7098968e@news.nwlink.com> <Xns91E4ADCD75B59RASXnewsDFE1@130.133.1.4>
Message-ID: <MPG.17140471d0caf8bc989691@news.nwlink.com>

In article <Xns91E4ADCD75B59RASXnewsDFE1 at 130.133.1.4>, 
starx at pacbell.net says...
> Jeff Shannon || Tue 02 Apr 2002 01:55:39p:
> 
> > Hm, wxPython seemed to perform adequately for me, on a (Win95) 
> > P200 with only 64MB RAM.  
> 
> I use WX too, and it seems that you pretty much pay a fixed price for 
> importing the modlue, no matter how much you need or how big your program 
> is. I'd think most other frameworks would have the same problem, an easy 
> way to tell would be something like:
> 
> print "Loading wxPython...",
> from wxPython.wx import *
> print "done"
> 
> ... I bet most of your execuation time for a hello world program would be 
> between those prints.

Indeed, testing with this simple script: 

from time import time
start = time()
from wxPython.wx import *
print "import took %f seconds" % (time() - start)

seems to indicate a 2 1/2 - 4 1/2 second import time on my (as 
noted, extremely slow) machine.  (Faster times happen when 
repeated in quick succession, slower times after some time has 
passed; I suspect caching and virtual memory paging to be the 
significant variables.) This, added to the time to start the 
Python interpreter, means that there is a significant startup 
time for even a very simple wxPython program.  But that startup 
time increases fairly slowly with the complexity of the program.  
And keep in mind that starting Notepad.exe, an optimized C 
program that consists almost entirely of an OS-builtin control, 
takes almost a second on this same machine, and any complex 
application (such as Excel or Photoshop) can take 45 seconds to 
a minute to load.  Also keep in mind that slow startup time is 
not necessarily correlated with poor responsiveness once the 
application is running.  Therefore, I would say that, unless 
there are *serious* timing constraints, complaints about wxPython 
being slow are on questionable ground.  I'd consider algorithm 
optimization to be a more fruitful area to work on, than worrying 
about the toolkit.

(I'm sure that Philip is in agreement on this; I'm clarifying my 
point for the O.P.'s benefit here.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From amuys at shortech.com.au  Tue Apr 16 21:57:59 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 16 Apr 2002 18:57:59 -0700
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <87vgaswc1s.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <7934d084.0204161757.72ff525f@posting.google.com>

"Stephen J. Turnbull" <stephen at xemacs.org> wrote in message news:<87vgaswc1s.fsf at tleepslib.sk.tsukuba.ac.jp>...
> >>>>> "David" == David Mertz <mertz at gnosis.cx> writes:
> 
>     David> There is nothing very difficult (and nothing -at all-
>     David> ambiguous) in here.
> 
> Nothing is difficult or ambiguous for any True Believer.  It is true
> that if you define the issues narrowly as you have, there is nothing
> difficult or ambiguous.
> 
> Those of us seeking truth, on the other hand, often must accept that
> multiple definitions and opinions are current, and try to adapt to
> others' points of view to further communication.

On the other hand, in this particular instance the parties concerned
have spent substantial effort defining the terms.  There is very
little ambiguity in the terms themselves.  Opinions on the relative
worth, and sustainability of Open Source and Free Software are
subjective and continue to be discussed extensively.  However to
attempt to suggest that the terms themselves have not been well
defined is dishonest.  To attempt to cast someone who objects to this
as some raving idealogue, and thereby dismiss them with casual wave
handling as a self declared 'seeker of truth' (tm) is the most
appalling piece of intellectual hypocracy I have seen for some time.

Andrae Muys


From neuhauser at mail.cz  Tue Apr 30 12:15:27 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 18:15:27 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com>
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net> <mailman.1020176703.26327.python-list@python.org> <ELyz8.16820$xy4.289@atlpnn01.usenetserver.com>
Message-ID: <20020430161526.GT325@roman.mobil.cz>

> From: "Steve Holden" <sholden at holdenweb.com>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Tue, 30 Apr 2002 11:27:00 -0400
> 
> "Roman Neuhauser" <neuhauser at mail.cz> wrote in message
> news:mailman.1020176703.26327.python-list at python.org...
> > > From: "Fredrik Lundh" <fredrik at pythonware.com>
> > > Subject: Re: Comment on draft PEP for deprecating six builtins
> > > To: python-list at python.org
> > > Date: Tue, 30 Apr 2002 14:01:42 GMT
> > >
> > > Roman Neuhauser wrote:
> > >
> > > > having spotty knowledge of python
> > >
> > > that's pretty obvious from your posts.
> > >
> > > maybe you should spend a little more time learning, and
> > > a little less time telling us about your "opinions" on issues
> > > you don't really understand?
> > >
> > > further reading:
> > >
> > > http://www.google.com/search?q=cache:opG_4HjTnbIC:www.theonion.com/onion3811/child-rearing_advice.html
> >
> >     how does that relate to python or OO?
>
> I  think it actually relates a little more to giving advice on subjects you
> confess not to know much about.

    Not advice. I was expressing my opinion as a python user.
 
> If you haven't come across the effbot before, you should know that he tends
> to be rude in this way. I think it happened when the parity unit was fitted
> to his main memory. Fortunately it does not stop him from being a valuable
> source of information on topics Pythonic. Every now and then he jumps out
> and bites a poster on the ankles due to simple crotchetiness, and then he
> will subside into inactivity once more, contemplating the innards of his
> software in silent satisfaction.

    Given the high tongue-in-cheek ration of this list, I'm now really
    confused as to whether Fredrik Lundh is a bot, or if you're "rocking
    me", as the Czech saying has it.

-- 
FreeBSD 4.4-STABLE
6:11PM up 1 day, 8:36, 12 users, load averages: 0.10, 0.06, 0.01




From gerhard at bigfoot.de  Mon Apr  1 02:30:11 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Mon, 1 Apr 2002 09:30:11 +0200
Subject: PyOS -- was: Re: ANN: cage 1.0 -- Cellular automata engine in Python
In-Reply-To: <20020401100837.GA1536@blackscarab>
References: <3CA5194E.D5AB00E1@alcyone.com> <20020401021435.GF333@lilith.hqd-internal> <20020401100837.GA1536@blackscarab>
Message-ID: <20020401073011.GA9427@lilith.hqd-internal>

* cruciatuz <sasoft at gmx.de> [2002-04-01 10:08 +0000]:
> On Mon, Apr 01, 2002 at 04:14:35AM +0200, Gerhard H?ring wrote:
> > * Erik Max Francis <max at alcyone.com> [2002-03-29 17:47 -0800]:
> > > Here's the latest PyOS installment:
> > > [...]
> > 
> > For one second I thought: "Hey cool - somebody wrote an operating system in
> > Python." :-)
> >  
> > This would be a nice exercise, IMO: use an existing kernel and C library, like
> > Linux and glibc, then implement everything starting, including /sbin/init and
> > the standard network servers (echo, finger, ...) in Python. Add a little Python
> > shell and voila - the PyOS :-)
> there was (the link is dead): "spyLinux" at
> http://dilbert.physast.uga.edu/~andy/minilinux.html
> 
> it was a boot-disk which booted linux and came with a small
> python-interpreter. it was quite nice.

Yes, indeed. I used that once. It was basically kernel 2.0.something with
busybox and a stripped-down Python 1.5.2.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))



From starx at pacbell.net  Fri Apr 12 05:36:48 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 12 Apr 2002 09:36:48 GMT
Subject: emacs python mode completion
References: <mailman.1018595366.4741.python-list@python.org>
Message-ID: <Xns91EE1AC7689E8RASXnewsDFE1@130.133.1.4>

Sean 'Shaleh' Perry || Fri 12 Apr 2002 12:08:08a:

> 
> On 11-Apr-2002 Mark McEahern wrote:
>> Python mode in emacs has the nice syntax coloring and all, but what
>> would be really great is if I could press tab or something and have
>> it complete statements.  I found this thread where Guido replied and
>> pointed out rlcompleter and the like for the interactive interpreter:
>> 
>> 
>> http://groups.google.com/groups?hl=en&selm=199801111855.NAA21082%40eri
>> c.CNRI .Reston.Va.US
>> 
>> Is there a tool for emacs that does Python command completion?
>> 
> 
> Doing that would require a python interpreter unless you limit it to
> language keywords and perhaps __builtins__.

We don't need a C interperter to use ETAGS. A modified version of this 
would do most of the work I think. (It would help to parse and keep track 
of the global namespace of the current module, too, so 'interesting' 
imports work properly. Use this to figure out what module/name we are 
talking about, and ETAGS to find that. Or something. I couldn't do this. 
Not now, not yet. =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From fgeiger at datec.at  Sun Apr 14 09:11:50 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Sun, 14 Apr 2002 15:11:50 +0200
Subject: Want to read images from a WebCAM
Message-ID: <a9bv72$772$1@rex.ip-plus.net>

Hi all,

does anybody know of an alternative to VideoCapture reading out a TerraTec
Pro USB WebCam on W2k?

My question probably should be "Is there an alternative to DirectX reading
out a WebCam", because VideoCapture likely is not to blame here.

I ask this, because the whole thing too often crashes with a Blue Screen Of
Death. I tried to do some investigations on this and installed the DirectX
8.1 SDK/Debug in a hope to see some diagnostic messages before the whole
machine crashes.

Indeed, there are some in case the machine does not crash:

"ASSERT Failed! Executable: python.exe Pid 614 Tid 4d4. Module qedit.dll, 3
objects left active! At line (...)\dllentry.cpp. Continue?".

There are quite a few of those messages.

Then I decided to try to debug vidcap.pyd. No success - I always end-up
somewhere in the OS kernel.

What's the next step? Installing the symbols of W2k?

I thought, before I go into *that*, I rather ask, whether I'm the only one
in such an unpleasant situation and if there are alternatives on W2k.

Any hint is appreciated.

Best reagrds
Franz GEIGER






From adeptus at earthlink.net  Mon Apr 22 17:44:16 2002
From: adeptus at earthlink.net (adeptus)
Date: Mon, 22 Apr 2002 21:44:16 GMT
Subject: Big Bucket O [Python,Zope,Stackless,OpenBSD] questions.
Message-ID: <Qu%w8.1925$iQ5.165333@newsread2.prod.itd.earthlink.net>

Just got openBSD up and running, and was very dissapointed when I tried to
get the standalone ZODB up and running... Then I found out about the
threading issues...

There's not much info out there on these questions, so anything anybody has
to say on the subject would be appreciated.

Python 2.2, Zope 2.5, StandaloneZODB all working on a BSD3.0 installation?

Zodb under stackless python? Is it possible?


I need the new features in 2.2 (generators, metaclasses) for a very
ambitious Zope product...


Thanks...




From jmarshal at mathworks.com  Tue Apr  9 16:50:57 2002
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 9 Apr 2002 20:50:57 GMT
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org>
Message-ID: <a8vk7h$apk$1@news.mathworks.com>

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

>    In Python *everything* is a pointer. Really.

That's an implementation detail--it would be possible to use non-pointer
bit-patterns for some objects.  For example, some implementation of Python might
store ints that fit into 29-bits as a single word consisting of the integer
followed by three tag bits.  The other seven tag-bit patterns could be used to
designate pointers (now forced to lie on 8-byte boundaries), small strings, etc.

To emulate the semantics of pointers, I like using single-element lists, as some
others have suggested.


From kzaragoza at attbi.com  Mon Apr  1 17:27:54 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: Mon, 01 Apr 2002 22:27:54 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017691324.15119.python-list@python.org>
Message-ID: <slrnaahnu8.nsu.kzaragoza@localhost.localdomain>

In article <mailman.1017691324.15119.python-list at python.org>, Tim Peters wrote:
> I'll channel him:  he intends to add False to that list.  That's all.

Gotta love those digital seances. :) Thanks, Tim.  This does make me
feel a bit better about what direction this is taking.

>> If you don't intend to modify this concept, then what's the real gain
>> for the new boolean type?
> 
> Have you read the PEP?  I can assure you he's not hiding a killer good
> reason he hasn't already mentioned <wink>.

I have read the PEP, but based on that alone, I don't see any reasons
that are terribly compelling.  Perhaps it's just that I'm an
experienced developer or that I have a CS degree.  I just don't find
the basic concepts as they are that hard to grasp.  That's why I keep
wondering how much aid this will really be to a new or inexperienced
programmer.

Kris

-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From dalke at dalkescientific.com  Sun Apr 28 15:18:49 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 28 Apr 2002 13:18:49 -0600
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s7el$sr7$1@slb2.atl.mindspring.net> <pan.2002.04.28.00.20.56.555436.3624@uol.com.br>
Message-ID: <aahhtq$7lq$1@slb4.atl.mindspring.net>

synthespian:
><sarcasm> No kidding! You can modify your classes?! Gee, I wonder if you
can do *
>that * in C++ </sarcasm>
>But can you modify them * in runtime * ?! Without halting your
>application and, say, recompiling? <smirk>
>Answer: NO
>Only in LISP can you do that.
>That's why NASA is so fond of using LISP on their probes.
>And what's this discussion Python X LISP? I thought Peter Norvig's
>article already had cleared that up (that Python is a LISP from people
>who never understood CONS).

I'm not sure what your point is here.  In more recent Pythons (1.5 and
newer, I think) you can do

>>> class Spam:
...  def what(self): print "I'm spam"
...
>>> class Eggs:
...  def what(self): print "I'm eggs"
...
>>> x = Spam()
>>> x.what()
I'm spam
>>> x.__class__ = Eggs
>>> x.what()
I'm eggs
>>>

                    Andrew
                    dalke at dalkescientific.com





From paul at prescod.net  Tue Apr 16 12:57:02 2002
From: paul at prescod.net (Paul Prescod)
Date: Tue, 16 Apr 2002 09:57:02 -0700
Subject: Python for .NET vs. Jython
References: <f57664b9.0204152215.57312efa@posting.google.com>
Message-ID: <3CBC57DE.3B027E70@prescod.net>

"Jonathan P." wrote:
> 
> Jython is undeniably a very successful implementation of a language on
> the JVM.
> 
> Considering Python for .NET is still very far from the level of
> maturity that Jython enjoys and the fact that the CLR is supposed to
> be designed from the ground up to support multiple languages (MS even
> commissioned and asked feedback from Mark Hammond and the others
> working with him on Python for .NET regarding the CLR), could it be
> the CLR is (ironically) less of a fit for Python than the JVM?

I think it is safer to say that magnitudes more effort have been spent
on Jython than on Python for .NET.

> If you read Jim Hugunin's original paper on JPython, it seems that
> from the beginning, Python on the JVM was a good idea and that,
> despite Sun's not designing the JVM for different languages, it
> actually turned out to be a really good fit for our favorite language
> after all.

Personally, I would say that if it were a "really good fit" then Jython
would be almost as fast as CPython. But I guess I'll agree that the fit
is surprisingly good for something that came about accidentally.

 Paul Prescod




From tomchance at gmx.net  Mon Apr 29 08:11:49 2002
From: tomchance at gmx.net (Tom Chance)
Date: Mon, 29 Apr 2002 12:11:49 +0000
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net> <3CCCD547.6090405@SPAMnwinternet.com>
Message-ID: <3CCD3885.7020301@gmx.net>

> I do PyKDE2 (the bindings for KDE), and I do them on SuSE7.1/7.3 
> (depending on the version). Find your python directory (/usr/lib/python 
> is normal for SuSE, although that should link to something like 
> /usr/lib/python2.1). Next look for the site-packages subdirectory - if 
> you have PyKDE2 installed, there should be a number of libk*cmodule.so 
> entries there (somewhere between 9 and 11 different names), as well as 
> some corresponding .py files.
> 
> Next, at the command line, open the Python interpreter and do:
> 
>  >>> import sys
>  >>> print sys.path
> 
> You should see 'site-packages' included in the list that prints out.
> (Ctrl-D to exit)

OK I did that and found bizarrely that the "print sys.path" statement 
printed out directories in the /usr/local tree, whilst I'd been working 
with python in the /usr tree, so I had somewhere along the line got two 
installs?! No problemo though, I symlinked 
/usr/local/lib/python/site-packages to /usr/lib/python/site-packages and 
it imported the gnome modules without a hitch :)



> etc. The module names correspond to the KDE library names.You also need 
> to have sip and PyQt installed, but I assume you do, since PyKDE2 won't 
> build and install without them. Make sure, however, that they all match 
> in version (latest is sip3.1, PyQt3.1, PyKDE2-3.1 - the 3.1 is the *sip* 
> version *not* the KDE version).

As for KDE, I didn't realise you needed to download those extra 
packages! From a tutorial I'd read from the python site about oding 
Python with KDE it made it seem like the KDEBindings package 
automatically set-up all the python bindings, including modules, for 
you. Oh well. I'm going to wait until either PyKDE2-3.1 comes out (it's 
on 3.0 atm), or better still until the packages come out for KDE3.0 final.

Thanks for the help! :)

Tom



From wolfgang.grafen at gmx.de  Sun Apr 21 19:02:49 2002
From: wolfgang.grafen at gmx.de (Wolfgang Grafen)
Date: Mon, 22 Apr 2002 01:02:49 +0200
Subject: newbie dictionary question
References: <TiHw8.99936$CH1.86776@sccrnsc02>
Message-ID: <3CC34519.E6C1063A@gmx.de>

"James Chapman <" schrieb:

> Hi,
>
> I have a question about dictionaries. I've been coding up a system of
> objects that represents the reuslts of different auctions. One thing I've
> been doing constantly though is iterating over dictionaries where the keys
> are the prices of the bids and the values are the bids at that amount.
>
> I was wondering if there is an easy way to consolidate to dictionaries into
> one, where if they both have the same key the resulting value can be
> concatenated.
>
> Also is there any similar syntax for dictionaries like the list syntax
> "[some_function(x) for x in other_list]"?
>
> Thanks,
>
> James Chapman

>>> d1={'a':1,'b':2}
>>> d2={'c':3,'b':4}
>>> d1.update(d2)
>>> d1
{'b': 4, 'c': 3, 'a': 1}

Hope this is what you mean but consider that the value of 'b' in d1 will be
overridden by the value of 'b' in d2.

wolfgang



From tim.one at comcast.net  Sun Apr 14 19:48:54 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 19:48:54 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <3cba0944@news.mhogaming.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOECIOOAA.tim.one@comcast.net>

[Ken Peek]
> As part of the "Grand Unification of the 'int' and the
> 'long int' types", the "new Python way" (as provided to
> us by Guido,) of viewing 'long ints' is that if they are
> a positive number, then the 'long int' should behave in
> a manner as if there is an infinite number of zeros
> above the highest bit in the number, and if the 'long
> int' is negative, then the 'long int' should behave as
> if there is an infinite number of ones above the highest
> bit (i.e.: in so called "two's complement" fashion-- the
> same way a signed 'regular int' behaves.)

Yes, but that's not new:  that's *always* been how long ints have been
viewed in Python, at least since 1.0 (long before that, it went thru a brief
confusing stage of viewing long ints as being in an unbounded 1's-complement
notation; that was a mess).

> Thus-- my "expectation" is that 0x8000000000000000 >> 1
> is: 0x4000000000000000 (and that is the way Python now
> works for positive 'long ints'.)

Right.

> Given the above description of the way NEGATIVE 'long
> ints' should behave, my 'expectation' is that:
> -0x8000000000000000 >> 1 _SHOULD_ return:
> -0xC000000000000000,

Oh no -- that would be plain wrong.  You're not accounting for the "-" sign
at the start of the result.   Recalling that -x == (~x) + 1, -0x80 is an
infinite string of 1 bits followed by 7 zero bits (hex notation doesn't
really help with the visualzation).  If you shift that right by 1 bit, then
you should (clearly) get an infinite string of one bits followed by 6 zero
bits.  In binary:

...111000000

Take the complement:

...000111111

Add 1:

...001000000

Back to hex:

-0x40

In exactly the same way, for short ints today:

    (-0x80) >> 1 == -0x40

and ditto for longs:

    (-0x80L) >> 1 == -0x40L

Those are both true in every version of Python since 1.0, and will continue
to be true after int/long unification is complete.  Right shifts of a Python
integral type have never needed to care whether the input was a short or
long.  It can only make a difference for left shifts now.

> (_NOT_ -0x4000000000000000 as Python _NOW_ behaves.)

Python may not be doing what you expect, but it is doing what you're arguing
for it to do <wink>.

> The way Python _NOW_ works, is to give a different
> behavior for the '>>' operator depending on whether you
> are working on an 'int', or a 'long int'.

No -- you're getting confused by the "-" notation, and that's all.  Every
other confusion follows from that.





From jwinkelman at movaz.com  Tue Apr  2 09:38:54 2002
From: jwinkelman at movaz.com (Jeff Winkelman)
Date: 2 Apr 2002 06:38:54 -0800
Subject: Help with COM and Type Library
Message-ID: <90164d1a.0204020638.7de0af02@posting.google.com>

I need some help with creating a type library for a Python COM Server
I am trying to implement.

I've read through previous postings on comp.lang.python regarding
creating type libraries and Python COM Servers, and have made some
progress.

I followed Alex Martelli's suggestion (Jan. 2001) to perform the
following steps:

1.  Create and IDL file
2.  Generate a type library using MIDL.
3.  Run makepy.py on the type library which generates a module with
some useful code.

My problem now is that I don't know what to do with this generated
file.

Here is my generated file:

# Created by makepy.py version 0.4.0
# By python version 2.1.1 (#20, Jul 26 2001, 11:38:51) [MSC 32 bit
(Intel)]
# From type library 'HelloWorld.tlb'
# On Mon Apr 01 16:02:09 2002
"""HelloWorld 1.0 Type Library"""
makepy_version = '0.4.0'
python_version = 0x20101f0

import win32com.client.CLSIDToClass, pythoncom

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

CLSID = pythoncom.MakeIID('{D71160DA-3B79-49A5-9E6E-FA23F458FBFB}')
MajorVersion = 1
MinorVersion = 0
LibraryFlags = 8
LCID = 0x0

from win32com.client import DispatchBaseClass
class IHelloWorld(DispatchBaseClass):
	"""IHelloWorld Interface"""
	CLSID = pythoncom.MakeIID('{DA3F71EA-ADFA-44F0-A890-1F88120C8309}')

	def Hello(self, who=defaultNamedNotOptArg):
		"""method Hello"""
		return self._oleobj_.InvokeTypes(0x1, LCID, 1, (24, 0), ((12,
0),),who)

	_prop_map_get_ = {
	}
	_prop_map_put_ = {
	}

class CoClassBaseClass:
	def __init__(self, oobj=None):
		if oobj is None: oobj = pythoncom.new(self.CLSID)
		self.__dict__["_dispobj_"] = self.default_interface(oobj)
	def __repr__(self):
		return "<win32com.gen_py.%s.%s>" % (__doc__,
self.__class__.__name__)

	def __getattr__(self, attr):
		d=self.__dict__["_dispobj_"]
		if d is not None: return getattr(d, attr)
		raise AttributeError, attr
	def __setattr__(self, attr, value):
		if self.__dict__.has_key(attr): self.__dict__[attr] = value; return
		try:
			d=self.__dict__["_dispobj_"]
			if d is not None:
				d.__setattr__(attr, value)
				return
		except AttributeError:
			pass
		self.__dict__[attr] = value

class HelloWorld(CoClassBaseClass): # A CoClass
	# HelloWorld Class
	CLSID = pythoncom.MakeIID("{35B9B525-B4E8-4953-A0E9-680B99FAF11C}")
	coclass_sources = [
	]
	coclass_interfaces = [
		IHelloWorld,
	]
	default_interface = IHelloWorld

IHelloWorld_vtables_dispatch_ = 1
IHelloWorld_vtables_ =  [('Hello', 1, ((12, 0, None),), (3, 0, None),
('who',))]

RecordMap = {
}

CLSIDToClassMap = {
	'{DA3F71EA-ADFA-44F0-A890-1F88120C8309}' : IHelloWorld,
	'{35B9B525-B4E8-4953-A0E9-680B99FAF11C}' : HelloWorld,
}
CLSIDToPackageMap = {}
win32com.client.CLSIDToClass.RegisterCLSIDsFromDict( CLSIDToClassMap )
VTablesToPackageMap = {}
VTablesToClassMap = {
}


VTablesNamesToCLSIDMap = {
	'IHelloWorld' : '{DA3F71EA-ADFA-44F0-A890-1F88120C8309}',
}


At this point, I'm stuck.  I really don't even know where to start
modifying this generated file and pulling out the stuff I don't need. 
If I look at example COM Servers implemented in Python, several of the
things I need are missing from this file.

Here is my COM Server all by itself.

class HelloWorld:
	# HelloWorld Class
	_public_methods_ = ['Hello']
	_public_attrs_ = ['softspace', 'noCalls']
	_readonly_attrs_ = ['noCalls']
	# All registration stuff to support fully automatic
register/unregister
	_reg_verprogid_ = "PYTHONLib.HelloWorld.1"
	_reg_progid_ = "PYTHONLib.HelloWorld"
	_reg_desc_ = "Hello World Example"
	_reg_clsid_ = "{35B9B525-B4E8-4953-A0E9-680B99FAF11C}"
	_reg_class_spec_ = "HelloWorld"

	def __init__(self):
		self.softspace = 1
		self.noCalls = 0
	def Hello(self, who):
		self.noCalls += 1
		return "Hello" + " "*self.softspace + who

Do I even need the CoClassBaseClass and IHelloWorld classes?

What significance does the vtable stuff have?  It seems to me that I
would not need this either since I am not wrapping up an existing
server, but rather implementing my own.

Can anyone give me some pointers on where to go from here?

Thanks in advance for any help.

Jeff


From imbosol at vt.edu  Tue Apr  9 19:33:33 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 9 Apr 2002 19:33:33 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <a8tov6$484$1@panix1.panix.com> <231u8a.lc.ln@127.0.0.1> <a8vnt9$roa$1@panix1.panix.com>
Message-ID: <dotv8a.un.ln@127.0.0.1>

Aahz wrote:
> In article <231u8a.lc.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
>>Aahz wrote:
>>>
>>> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
>>> free to buy another book if K&R turns out to be too heavy for you right
>>> now, but recognize that you'll need it eventually.
>>
>>Really?  I've been programming in C for ten years, and I've never even
>>seen the K&R book.  I think I learned it just fine.  Not that I have a
>>better suggestion.
> 
> I'm of the opinion that if you used K&R, you'd agree with me.

Well, since I have been programming in C for ten years, and have never
needed K&R, I have already proven your original statments, that I
"*MUST* get K&R. Period," and that I'll "need it eventually," wrong.

So if I agreed with you, I'd be wrong.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From picodello at yahoo.it  Fri Apr 19 07:36:08 2002
From: picodello at yahoo.it (Piero)
Date: 19 Apr 2002 04:36:08 -0700
Subject: use of calldll
Message-ID: <84323230.0204190336.4dc51098@posting.google.com>

I use the very good calldll module to interface to Windows DLL.

Now I have a problem with this type of function 'C 'prototype :

BOOL ReadOPCItem(HANDLE hConnect,HANDLE hGroup,HANDLE hItem,VARIANT
*pVar,FILETIME * pTimeStamp, DWORD quality)

I make this Python routine  :

def ReadOPCItem(self,hGroup,hItem):
	#~ ' May be used by an application to read the current value of an
item.
	#~ ' Is only valid if EnableDLLBuffering has been set to TRUE when
the
	#~ ' connection was established.
	#~ Public Type FileTime
	#~ dwLowDateTime As Long
	#~ dwHighDateTime As Long
	#~ End Type
	#~ Declare Function ReadOPCItem Lib "WTClient" Alias
"_ReadOPCItem at 24" (ByVal hConnect As Long, ByVal hGroup As Long, ByVal
hItem As Long, Var As Variant, pTimeStamp As FileTime, Quality As
Long) As Long
	ReadOPCItem =calldll.get_proc_address(self.opcclient,"_ReadOPCItem at 24")
		var_buf = calldll.membuf(8)
		filetime_buf = calldll.membuf(8)
		quality_buf = calldll.membuf(4)
		stato = calldll.call_foreign_function   (ReadOPCItem,'llllll', 'i',
			(self.handle,hGroup,hItem,var_buf.address(),filetime_buf.address(),quality_buf.address()))
		quality = struct.unpack ('l', quality_buf.read())
		value = struct.unpack ('bbbbbbb', var_buf.read())
		filetime = struct.unpack ('ll', filetime_buf.read())
		
		return stato,value,quality,filetime


Something can help on using VARIANT type in Python and CALLDLL

Thanks

Piero Dell'OSte


From sixtysix at inwind.it  Thu Apr  4 11:24:39 2002
From: sixtysix at inwind.it (Massimo Valentini)
Date: Thu, 4 Apr 2002 18:24:39 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk> <eR_q8.117876$%g7.3768326@twister2.libero.it>
Message-ID: <i4%q8.117950$%g7.3769873@twister2.libero.it>

"Massimo Valentini" wrote
:
: I would suggest to consider even a case like this:
:
: url1 = "http://foo/a/b/images"
: url2 = "http://foo/a/b/c/d/images"
: print "Expect: '../../../images'"
: print repr(relativePath(url1,url2))
: print ""
:

Sorry, the case to consider is this one:

url1 = "http://foo/a/b/images"
url2 = "http://foo/a/c/images/b"
print "Expect: '../../c/images/b'"
print repr(relativePath(url1,url2))
print ""

Massimo




From spam at yourself.com  Fri Apr 26 09:29:59 2002
From: spam at yourself.com (Ingo Blank)
Date: Fri, 26 Apr 2002 15:29:59 +0200
Subject: Killing a process under Windows NT/2K
References: <aab8hn$o4s$05$1@news.t-online.com> <ur8l2wsch.fsf@online.no>
Message-ID: <aabkon$tte$03$1@news.t-online.com>

"Syver Enstad" <syver-en+usenet at online.no> schrieb im Newsbeitrag
news:ur8l2wsch.fsf at online.no...
> "Ingo Blank" <spam at yourself.com> writes:
>
> > Hi,
> >
> > I need to kill a process under Windows (NT/2K), whose PID is known.
> > Unfortunately os.kill() doesn't work under Windows.
> > How can I achieve this anyway?
> >
> > Thanks in advance
> >
>
> This might help. You have to have the win32all extensions installed,
> but so should *everybody* that uses python on win32 anyway.
>
> # Utility functions that should have been implemented in the os module
> import signal
> signal.SIGKILL = 120 # cludge: try to mimic the standard library kill
>
> def kill(pid, sig):
>     assert sig == signal.SIGKILL # cludge: try to mimic the standard
library kill
>     handle = win32api.OpenProcess(win32con.PROCESS_TERMINATE, 0, pid)
>     win32api.TerminateProcess(handle, 1) # exit status
>
>
> def waitpid(pid):
>     handle =
win32api.OpenProcess(win32con.SYNCHRONIZE|win32con.PROCESS_QUERY_INFORMATION
, 0, pid)
>     win32event.WaitForSingleObject(handle, win32event.INFINITE)
>     exitCode = win32process.GetExitCodeProcess(handle)
>     return pid, exitCode
>

Thanks for your help, Syver!
That works perfectly...

--ingo

>
> Vennlig hilsen
>
> Syver Enstad




From tismer at tismer.com  Wed Apr 24 11:34:22 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 24 Apr 2002 17:34:22 +0200
Subject: Stackless goes Limbo
Message-ID: <3CC6D07E.409@tismer.com>

Hi all,

after some time of consideration,
I have decided to move the interface to Stackless
Python's multitasking towards the ideas behind
the Limbo Language.

Besides the tasklets, there are "channel" objects,
which allow for communication between tasklets.
This approach is simple and powerful. Blocking of
tasklets takes place on writes when no tasklet
is listening, or on reads, when no data is available.

There is some nice introduction to Limbo at
http://www.vitanuova.com/inferno/papers/descent.html
written by Brian W. Kernighan, one of the inventors
of the C language.

I hope to have a prototype ready by end of April.

See the stackless homepage for more news,
including the new sponsorship honor page at:

<http://www.stackless.com/>

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From neal at metaslash.com  Fri Apr 19 12:27:55 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 19 Apr 2002 12:27:55 -0400
Subject: os.system() with redirection
References: <mailman.1019229521.16923.python-list@python.org>
Message-ID: <3CC0458B.A27B476D@metaslash.com>

Alice NgarKit Ko wrote:
> 
> Hi,
> I want to call a program with os.system().  The program redirects output
> to file only through >& outputfile.   Here is my script (temp.py).
> 
> line = 'protomol temp.conf.PME.Ewald >& run.temp.PME.Ewald'
> os.system(line)
> 
> >python temp.py
> >sh:  run.temp.PME.Ewald:  bad number

 >& is not in the standard Bourne shell.

Do:  app > file 2>&1
or specifically for your example:

	protomol temp.conf.PME.Ewald > run.temp.PME.Ewald 2>&1

Neal


From max at alcyone.com  Fri Apr 12 18:14:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 12 Apr 2002 15:14:16 -0700
Subject: ANN: Munch 0.1 -- Embed Python in text as markup
Message-ID: <3CB75C38.757A849D@alcyone.com>

Summary

    Munch is a system for embedding Python expressions and statements
    in text (or any other unexecuted content); it takes source files,
    processes them, and produces output.  These are done via
    expansions, which are snippets of Python code delimited by special
    characters involving the at sign (@), and are substituted in the
    output with the return value (for expressions) or with any
    in-Python printing to stdout that takes place (for both
    expressions and statements).  Textual data not explicitly
    delimited in this way is sent unmolested to the output file,
    allowing Python to be used effectively as a markup language.  Explicit
    expansion within Python code is possible.  Also supported are
    m4-style diversions, where output can be systematically deferred
    and recalled.

    Expressions are embedded in text with the '@(...)' notation; as a
    shortcut, simple variables and expressions can be abbreviated as
    '@variable', '@object.attribute', '@array[index]', or
    '@function(arguments)'.  Full-fledged statements are embedded with
    '@{...}'.


Getting the software

    The software is available in a tarball here:
    http://www.alcyone.com/pyos/munch/munch-0.1.tar.gz.

    The official URL for this Web site is
    http://www.alcyone.com/pyos/munch/.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From jwickard**nospam** at litriusgroup.com  Tue Apr  2 10:13:00 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Tue, 02 Apr 2002 15:13:00 GMT
Subject: Quick return question
Message-ID: <0Ujq8.18$c6.7717@typhoon.mn.ipsvc.net>

Hello all:
given the function:

def adder(**args):
    for key in args.keys():
        print args[key]

print adder(fName="sam", lName="scott", mName="wayne")

I get:

sam
scott
wayne
None

Why am I getting None as well, because there's no return value for the
function?




From boris at cantal.lip6.fr  Mon Apr  8 12:15:31 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Mon, 08 Apr 2002 18:15:31 +0200
Subject: changing local dictionnary for functions
Message-ID: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr>

Is there a way to pass a local dictionnary to a function as it is
possible to do with a file with the execfile statement ?

In the python reference manual, in the section 4.1 : Code
Blocks,execution frames and namespace, there is a table that say that
function body uses globals namespace of the containing block and a new
locals namespace. And there is a note that say global and local can be
overrident, how is it possible ? I think it would be the answer to my
first question.

Boris


From Oschler at earthlink.net  Wed Apr  3 18:50:23 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Wed, 3 Apr 2002 18:50:23 -0500
Subject: IDE Wars, the prequel!
References: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com> <mailman.1017867011.28827.python-list@python.org>
Message-ID: <zrMq8.25970$0T6.7657756@e3500-atl1.usenetserver.com>

Trent Mick <trentm at ActiveState.com> wrote in message
news:mailman.1017867011.28827.python-list at python.org...
> A clarification here. Mark Hammond's set of Win32 extensions for Python
> includes an IDE called PythonWin. The Win32 extensions are installed by
> default with ActivePython, hence you have PythonWin is you installed
> ActivePython. (You can still get PythonWin with other Python
distributions --
> you just have to install win32all after installing Python. See
> http://starship.python.net/crew/mhammond/)
>
> ActiveState also has an IDE called Komodo (for Python, Perl, PHP, XSLT,
etc)
> that runs on Windows and Linux.
>     http://www.ActiveState.com/Products/Komodo/
>

A little confused here.  So should I look at Komodo, or are you saying that
I can install win32all on Linux and then run ActivePython?

thx






From philh at comuno.freeserve.co.uk  Sat Apr  6 12:39:38 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sat, 6 Apr 2002 18:39:38 +0100
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <a8km9i$7t6$1@panix1.panix.com>
Message-ID: <slrnaaucmq.3j0.philh@comuno.freeserve.co.uk>

On 5 Apr 2002 12:18:42 -0500, Aahz <aahz at pythoncraft.com> wrote:
>In article <slrnaapc23.r7n.philh at comuno.freeserve.co.uk>,
>phil hunt <philh at comuno.freeserve.co.uk> wrote:
>>
>>Recently on this newsgroup there's been a discussion about adding a 
>>bool datatype to Python; and Guido has decided to do this.
>>
>>Some time ago, there was a rather contentious discussion about 
>>changing the semantics of the / operator. To be precise, what should 
>>it do when the arguments are integers: should 1/2 be 0 or 0.5?
>>
>>In both cases, there are good arguments in favour of the change. 
>>(Though having said that, in both cases Python has done perfectly 
>>well for years without it). The problem with making these changes is 
>>that they can cause currently-working Python code to fail on new 
>>versions of the language.
>
>Note that in the case of division, the incompatible change has been
>officially postponed to Python 3.0.  In both cases, the argument is that
>Python has *not* worked "perfectly well"; both cases in current Python
>generate ambiguities of the sort that Guido loathes.

What ambiguities? There is nothing ambiguous about current behavour 
of truth values. Furthermore, there is nothing that, in practise, 
hampers programmers: a programmer can easily learn to use 1 and 0 
for true and false.

The only advantage for the new scheme, as I see it, is it looks 
prettier and is a bit easier for newbies. Ditto for integer 
division.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From sholden at holdenweb.com  Wed Apr 24 08:14:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 08:14:56 -0400
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org> <aa48nq$a61$1@panix1.panix.com> <mailman.1019640045.27779.python-list@python.org>
Message-ID: <znxx8.44829$0x3.4841@atlpnn01.usenetserver.com>

"Graham Ashton" <gashton at cmedltd.com> wrote ...
> On Tue, 2002-04-23 at 19:21, Aahz wrote:
> > In article <mailman.1019552082.8541.python-list at python.org>,
> > Graham Ashton  <gashton at cmedltd.com> wrote:
> > >
> > >>>> class A(AbstractSingleton, object):
> > >...   pass
> >
> > Maybe you should tell us a bit more about what you're trying to do so
> > that you needn't rely on crufty internal stuff to get the correct
result.
>
> Okay, good point. I'm using the Alex's Borg pattern in place of a
> singleton, as shared state will do quite nicely. I'm also trying to
> protect myself from getting caught out by new style classes (in other
> words it doesn't work -- in general -- for classes derived from, for
> example, dict).
>
> So, until I need singletons that support subclasses of dict in all their
> glory, I decided to do the "simplest thing that will possibly work" and
> raise an exception if I forgot about the limitation (hence the need to
> detect whether it's been inherited by a new style class).
>
> An alternative question would be -- does anybody know how to do the Borg
> so that it works for new style classes?
>
> P.S. Thanks to those who replied to my earlier posts -- both solutions
> work nicely.
>

You could do worse than read Alex's conference paper from Python10:

    http://www.python10.org/p10-papers/13/index.htm

in which he shows how Borg came about, and how it can be carried forward
into the brave new world of type/class unification.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From python at rcn.com  Tue Apr 30 18:22:49 2002
From: python at rcn.com (Raymond Hettinger)
Date: Tue, 30 Apr 2002 18:22:49 -0400
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
Message-ID: <aan5hv$aor$1@bob.news.rcn.net>

Reading all of the posts on the deprecation pep shows that there is a
tension between ability to remove cruft versus breakage of old programs.
My idea for getting the best of both is psuedo-deprecation.

Let's take anything that shouldn't be there anymore (read as anything that
wouldn't be put in again if we were doing this from scratch) in a separate
section of the library reference called Obsolete Constructs.  We'll pull it
out of the main body of the docs and the tutorial; remove uses of it from
the Python library; and ask authors not to even mention the
psuedo-deprecates
(except in a separate section devoted to being able to read old code).

That way, we can publish a short, clean list of built-ins; create new
modules for functionals and iterttools; on go on merrily as if the language
had been truly cleaned.  Everyone can start using the language, not as Guido
had intended, but as he would intend if he had it to do over again.

I no longer think map(), filter(), and reduce() should even be
psuedo-deprecated,
but apply(), oct() and hex() are prime candidates.  The input() function
should
still be considered for genuine deprecation and deletion, but that reflects
my
view of input() as a bug.

We can add iterzip() and iterrange() to the builtins; psuedo-deprecate zip()
and range(); and live in a world with more uniform interfaces and less
unnecessary memory consumption.

Alex Martelli pointed out that this was the original concept of deprecation,
"it's there but we're sorry it is".

Neal Norwitz has already started building out PyChecker to issue warnings
for deprecated constructs.

Alex also pointed out that this form of "moral 'suasion" was tried with <>
vs !=.  Still, we have not tried isolating redundant, obsolete, or rarely
used functions in a separate, back-of-the-book section.

I think there is a clear advantage for creating a strong, central core of
what you need to know and de-emphasizing what was part of the past.

Though all the functions would still be in a dir() listing of __builtins__,
there is a powerful intellectual advantage to mentally whittling down
the list of tools to a well integrated, orthogonal, best-of-class set.

The quality of programs will improve, newbies will learn faster, the
language can advance, AND old programs will run fine.

'nuff said,


Raymond Hettinger




From rouse at sbcglobal.net  Tue Apr 23 13:33:47 2002
From: rouse at sbcglobal.net (Chris Rouse)
Date: Tue, 23 Apr 2002 10:33:47 -0700
Subject: Newbie: Running a script on WindowsXP
In-Reply-To: <#dmZtdu6BHA.1188@cpimsnntpa03>
Message-ID: <000001c1eaed$07019ee0$5cfea8c0@truebrujah>

Your issue with this is that you don't have an Autoexec any more in
Windows XP. What you have to do is right click on My Computer then click
on properties. Then go to the advanced tab and hit the Environment
button. In the upper list box you will find the path variable. Click on
the path variable and click end. You should find the end of the line and
add it like you would Autoexec. Now click ok and ok.. and you should be
up and running.


Chris Rouse

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of billludden
Sent: Tuesday, April 23, 2002 9:54 AM
To: python-list at python.org
Subject: Newbie: Running a script on WindowsXP

Hi,

I am trying to figure out how to run a Python21 script
from the "Run" box on WindowsXP Home edition.
When I run it, the Command Prompt box just briefly
flashes on the screen and closes itself.  I have tried
using the -i option like you would use in Windows 98
but it doesn't seem to work.  I also tried appending
Autoexec with:
    SET PATH=.........;C:\PROGRA~1\PYTHON21

Does anyone understand what I'm doing wrong?
(incidentally, I had the same problem with "hello world"
 when I tried it in C)

Thank you!
Bill




-- 
http://mail.python.org/mailman/listinfo/python-list





From aahz at pythoncraft.com  Fri Apr  5 18:29:25 2002
From: aahz at pythoncraft.com (Aahz)
Date: 5 Apr 2002 18:29:25 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net> <mailman.1018036693.6922.python-list@python.org>
Message-ID: <a8lc0l$ikk$1@panix1.panix.com>

In article <mailman.1018036693.6922.python-list at python.org>,
Geoff Gerrietts  <geoff at gerrietts.net> wrote:
>
>The second is that I personally find little to admire in C++, but one
>of the things I do admire is that Bjarne decided to take a couple
>years off to let the compiler writers and application programmers
>catch up. The result was a language that everyone more or less thought
>they knew, a solid foundation for people to build on and move forward.

That was Python 1.5.2.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From m.faassen at vet.uu.nl  Sat Apr 27 06:54:36 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Apr 2002 10:54:36 GMT
Subject: zope and jboss or weblogic
References: <e786d63d.0204262051.14ebe470@posting.google.com> <mailman.1019890132.10241.python-list@python.org>
Message-ID: <aae01c$lau$1@newshost.accu.uu.nl>

Gerhard H?ring <gerhard at bigfoot.de> wrote:
> * Wenshan Du <python at dohao.org> [2002-04-26 21:51 -0700]: > hi,
>> I like zope, I want to know what's the difference between zope and
>> jboss or weblogic? which is better? Can python & zope replace J2EE?
>> 
> Not only do Zope and J2EE play in a different league, but it's also
> different kinds of sports.

> J2EE is really only a bunch of so-called enterprise-APIs thrown together
> and made sure the versions of the selected APIs work well together.
> These APIs include data persistence, distributed computing, others I
> don't remember, and last but not least, a web application framework
> (servlets) with a everybody-has-it-so-we-too code-in-HTML solution
> (JSP). OTOH, I don't know (yet), what ZOPE really is about.

Well, Zope is a bunch of different things:

  * data persistence (the ZODB)

  * distributed computing (ZEO)
 
  * web application framework (ZPublisher, whole bunch of surrounding
    stuff such as session tracking)

  * web-based management/development UI

  * HTML templating (DTML, ZPT)

  * relational database integration (ZSQL)

  * indexing and searching (ZCatalog)

  * security infrastructure (who can do what)

  * additions on top of Zope, such as the content management framework (CMF)

And so on. It all looks surprisingly innocuous through the web UI, but
there's a lot there. It's cool.

One problem of the current generation of Zope (Zope 2) is that the APIs are 
often grown, not designed, and documentation, while these days rather
abundant, isn't that easy to work your way through (though with the
advent of several Zope books this has improved a lot).

Zope 3, currently under development in open source fashion and in
vaporware stage is promising to clean up things a lot, and establishes
a nice component architecture. In general things are becoming way more
Pythonic than Zope 2 is. When it's ready, Zope 3 looks to become very
very very cool.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From greg at cosc.canterbury.ac.nz  Mon Apr 15 22:34:38 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 16 Apr 2002 14:34:38 +1200
Subject: split() off last substring
References: <3CBA41CB.27DCA084@uniserve.com>
Message-ID: <3CBB8DBE.91C10232@cosc.canterbury.ac.nz>

bvdpoel at uniserve.com wrote:
> 
> I guess I want something like:
> 
>         b,c=a.rsplit(1)

It would be nice if you could do this with

  b, c = a.split(-1)

PEP, anyone?

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


From jeff at jeffyouel.com  Fri Apr  5 17:36:44 2002
From: jeff at jeffyouel.com (Jeff Youel)
Date: Fri, 05 Apr 2002 22:36:44 GMT
Subject: crc32 function
References: <Xns91E821BC599Fvarialgmxde@130.133.1.4>
Message-ID: <0Gpr8.27781$zN.10554406@twister.socal.rr.com>

"Robert Spielmann" <varial at gmx.de> wrote in message
news:Xns91E821BC599Fvarialgmxde at 130.133.1.4...
> I have a program where I need to calculate the crc32 values of fibonacci
> numbers, is there a way to get the crc32 of int variables or do I need to
> first write the number to a file, then read it back as a string to do the
> crc32 stuff?

import zlib
n = 314
crc32 = zlib.crc32(str(n))

HTH,
Jeff




From huaiyu at gauss.almadan.ibm.com  Tue Apr 30 20:34:46 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 1 May 2002 00:34:46 +0000 (UTC)
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu> <3kjz8.32779$8D3.928578@news1.tin.it> <aam88o$oec$1@netnews.upenn.edu> <g8yz8.81675$vF6.2428847@news2.tin.it> <aamt6p$dfa$1@netnews.upenn.edu>
Message-ID: <slrnacue16.9l1.huaiyu@gauss.almadan.ibm.com>

Paul Hughett <hughett at mercur.uphs.upenn.edu> wrote:
>Alex Martelli <aleax at aleax.it> wrote:
>
>:> Alex Martelli <aleax at aleax.it> wrote:
>
>:>>> r=Numeric.array(range(13))
>
>:>>> r[-3::]
>: array([10, 11, 12])
>:>>> r[-3::-1]
>: array([10,  9,  8,  7,  6,  5,  4,  3,  2,  1,  0])
>:>>>
>
>Hmm.  Is the value of r[-1:-1:-1] equal to array([12, 11, 10, 9, 8, 7,
>6, 5, 4, 3, 2, 1, 0]) or to an empty array?  It could plausibly be
>either, depending on whether the middle -1 is interpreted as a lower
>limit or as 1 from the end.

Some quick experiments:

>>> r = arange (7)
>>> r[::-1]
array([6, 5, 4, 3, 2, 1, 0])
>>> r[5::-1]
array([5, 4, 3, 2, 1, 0])
>>> r[5:2:-1]
array([5, 4, 3])
>>> r[:2:-1]
array([6, 5, 4, 3])
>>> r[7:2:-1]
array([0, 6, 5, 4, 3])
>>> r[17:2:-1]
array([0, 6, 5, 4, 3])
>>> r[17:-1:-1]
array([0])
>>> r[:-1:-1]
zeros((0,), 'l')
>>> r[:-11:-1]
array([6, 5, 4, 3, 2, 1, 0])
>>> r[:-3:-1]
array([6, 5])
>>> r[:-5:-1]
array([6, 5, 4, 3])
>>> r[-2:-5:-1]
array([5, 4, 3])
>>> 

Summarization of the rules left as an exercise to the reader.  :-)
When the bounds are out of range, the rules are not very obvious, in fact.

Huaiyu


From loewis at informatik.hu-berlin.de  Thu Apr 25 10:26:54 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 25 Apr 2002 16:26:54 +0200
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc>
Message-ID: <j4helzdfbl.fsf@informatik.hu-berlin.de>

"Adonis Vargas" <deltapigz at telocity.com> writes:

> i am just starting my venture into XML as a means of document output
> and soon as a form of internet protocols. and i would like to know
> how i can parse XML in Python? now i know i have access to SAX and
> DOM, but how do i use them? i have looked into the documentation,
> and it seems to be written in a martian language, and i am looking
> in doing basic stuff just regular nodes, in particular. any
> link/sample usage just anything to get my feet wet in this subject,
> im a decent Python programmer, so there shouldnt be too much
> problems, just need something i can work with.

Have you looked at the HOWTO? Which part did you not understand (just
indicate the first unclear sentence)?

Regards,
Martin



From dalke at dalkescientific.com  Sat Apr 13 21:32:00 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 13 Apr 2002 19:32:00 -0600
Subject: Need customised sort
References: <3CB884C7.EE7E04D@uniserve.com>
Message-ID: <a9alvs$13t$1@slb3.atl.mindspring.net>

bvdpoel at uniserve.com:
>I have a list like this:
>

(from elsewhere)
> [ [0,x], [0,y], [4,n]...]. ...]

>and I want to sort it a bit. The sort should only work on the first
>element AND, retain the order of list elements if 2 adjacent elements
>have identical first elements.

suppose you have a list 'x' containing a set of terms

  x = [ [0, 5], [0, 4], [4, 1]]

then you could get the sort you wanted by

>>> y = [(term[0], i, term) for i, term in zip(range(len(x)), x)]
>>> y
[(0, 0, [0, 5]), (0, 1, [0, 4]), (4, 2, [4, 1])]
>>> y.sort()
>>> [term[2] for term in y]
[[0, 5], [0, 4], [4, 1]]
>>>

What this does is make a new list 'y' where each term in y is:
  - the first field from each term in x, which is the primary sort value

  - the index of the term, as in 0, 1, 2, 3, ..., which is the secondary
      sort value, used to reserve the original order if there are ties in
      the first sort value

  - the original term in x ; note, because the secondary sort keys are
      unique this value will never affect the sorting

This new list, 'y', is easily sorted by Python, and the result is in
the sort order you want.  You just need to get the original data out
of the list, by getting the 3rd term (the [2] term) from y.

                    Andrew
                    dalke at dalkescientific.com





From emile at fenx.com  Fri Apr 12 11:07:27 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 12 Apr 2002 15:07:27 GMT
Subject: Variables common to all objects of a class
References: <3CB6CD99.70409@thomas-guettler.de>
Message-ID: <PKCt8.13308$CH1.7391@sccrnsc02>

Thomas Guettler
> Hi!
> 
> How do you store variables which should be common to all
> objects of a class?
> 

class Foo:
    class_var_count_foos = 0 # shared by all instances
    def __init__(self, var):
        self.instance_var = var
        Foo.class_var_count_foos += 1

foo1 = Foo(1)
foo2 = Foo(2)
foo3 = Foo(3)

print foo1.class_var_count_foos
print foo1.instance_var

print foo2.instance_var


--

Emile van Sebille
emile at fenx.com

---------



From dale at riverhall.NOTHANKS.co.uk  Tue Apr 16 07:26:20 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 16 Apr 2002 12:26:20 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <0f2obug5v5dhuvlur9hanj2ecb5mbds2ou@4ax.com>

This is a non-issue, as far as I'm concernced.

There are far more pressing adjustments to the language that I'd like
to see, like 'unless' and 'until', but don't get me started.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From emile at fenx.com  Sun Apr 21 23:13:13 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 22 Apr 2002 03:13:13 GMT
Subject: A simply newbie question about ndiff
References: <3cc33cc2$0$15474$afc38c87@news.optusnet.com.au> <DtIw8.49773$HH5.4310@rwcrnsc51.ops.asp.att.net> <3cc359ff$0$15478$afc38c87@news.optusnet.com.au> <B8E8DEB0.21E3D%goodger@users.sourceforge.net> <3cc368bf$0$15477$afc38c87@news.optusnet.com.au> <B8E8E4D6.21E42%goodger@users.sourceforge.net> <3cc3765f$0$15478$afc38c87@news.optusnet.com.au>
Message-ID: <ddLw8.50678$HH5.4019@rwcrnsc51.ops.asp.att.net>

Neville Franks
> Ok thanks. I've installed 2.2 and ndiff is available but I'm still
confused.
> I thought that Tim's original ndiff compared files, but the one in
difflib
> compares lists of strings. Does this mean I somehow need to read the
files
> into lists of strings and then call ndiff. A short example of how to
use
> ndiff to compare two files would be greatly appreciated.
>

David pointed you to python22/tools/scripts/difflib.py.  In it you'll
find this example of how to use it:

def fopen(fname):
    try:
        return open(fname, 'r')
    except IOError, detail:
        return fail("couldn't open " + fname + ": " + str(detail))

# open two files & spray the diff to stdout; return false iff a problem
def fcompare(f1name, f2name):
    f1 = fopen(f1name)
    f2 = fopen(f2name)
    if not f1 or not f2:
        return 0

    a = f1.readlines(); f1.close()
    b = f2.readlines(); f2.close()
    for line in difflib.ndiff(a, b):
        print line,

    return 1


You could then say something like:

fcompare(r'c:\CrunchyFrog.txt', r'c:\CrunchyFfruug.txt')

Assuming-you-were-casting-Peter-Sellers-in-a-skit-ly y'rs,


--

Emile van Sebille
emile at fenx.com

---------



From grante at visi.com  Sat Apr 20 17:35:25 2002
From: grante at visi.com (Grant Edwards)
Date: Sat, 20 Apr 2002 21:35:25 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83>
Message-ID: <slrnac3nqj.1ec.grante@tuxtop.visi.com>

In article <Xns91F67C40EB0E0lumberjack at 216.148.53.83>, Lumberjack wrote:
> aahz at pythoncraft.com (Aahz) wrote:
>> What's amazing about Lisp is that it's still in many ways on the cutting
>> edge of computer science.
> 
> Perhaps you should name these "many ways". Taken at face value, your 
> statement pretty much claims that computer science hasn't progressed in 40 
> some years. I can at least agree with that statement. Of course the 
> "science" in "computer science" is pure bovine excrement. If you can't or 
> wont measure it, it isn't science. "Computer religion" or "computer cults" 
> would better capture the reality.

It's part math, part crafstmanship.  I agree it isn't science.
As a general rule, no field with "science" in its name is.

-- 
Grant Edwards                   grante             Yow!  .. My pants just went
                                  at               on a wild rampage through a
                               visi.com            Long Island Bowling Alley!!


From jwickard**nospam** at litriusgroup.com  Wed Apr  3 18:20:25 2002
From: jwickard**nospam** at litriusgroup.com (joel8bit)
Date: Wed, 03 Apr 2002 23:20:25 GMT
Subject: run_pyc_file:  nested_scopes: 0 
Message-ID: <Z6Mq8.229$c6.61189@typhoon.mn.ipsvc.net>

I juste exported a python program I was writing from my win2k box to a win98
box.  I only exported the pyc files for the program.  When I try to run the
program on the win98 machine I get this output:

run_pyc_file: nested_scopes: 0

I don't know exactly what's going on.




From mkellner at rpvreport.com  Tue Apr 30 17:39:17 2002
From: mkellner at rpvreport.com (Martin Kellner)
Date: 30 Apr 2002 14:39:17 -0700
Subject: Free tool for reports development
References: <6a5f9873.0204291439.41156c92@posting.google.com> <8ubscu8qagjjk8dkc2t06eb5n7s2n26f82@4ax.com>
Message-ID: <6a5f9873.0204301339.6c1aa697@posting.google.com>

Actually we're not planning to release the source code, but we're
improving it to be able to do lots of things that actually doesn't do.
For example graphics. What I know is that the tool will remain free.

Tim Roberts <timr at probo.com> wrote in message news:<8ubscu8qagjjk8dkc2t06eb5n7s2n26f82 at 4ax.com>...
> mkellner at rpvreport.com (Martin Kellner) wrote:
> > 
> >Sorry for disturbing, but I would like to notify you that we're
> >offering a FREE TOOL for report development: Rpv Printing System.
> > 
> >This tool lets you develop Windows reports from any programming
> >language and we've been noticed that Python community could be glad to
> >hear about it.
> 
> This is a very interesting concept.
> 
> It would be fun to play with the source code to make the input format more
> like HTML/XMLish.  A tool that could take an HTML <TABLE>, format it, and
> paginate it would be valuable.
> 
> Do you plan on releasing the Delphi source code for this tool?


From jeff at ccvcorp.com  Tue Apr  9 19:17:51 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 9 Apr 2002 16:17:51 -0700
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8vk7h$apk$1@news.mathworks.com>
Message-ID: <MPG.171d167b861e1d669896a4@news.nwlink.com>

In article <a8vk7h$apk$1 at news.mathworks.com>, 
jmarshal at mathworks.com says...
> 
> To emulate the semantics of pointers, I like using single-element
> lists, as some others have suggested.
> 

I prefer to *not* emulate the semantics of pointers.  If I have 
data that needs to be changed in such a way, I'll put it in a 
class and change it through a method.  If I wanted to fiddle with 
pointers, I'd program in C.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From fredrik at pythonware.com  Wed Apr  3 06:07:16 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 11:07:16 GMT
Subject: PyInt_AsLong()
References: <xdu7knpaz87.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <EnBq8.29186$n4.5840811@newsc.telia.net>

Ralf Juengling wrote:

> PyInt_AsLong() returns '-1? to indicate an error, this meets the convetion
> for PyXXX()-functions which return integers. But how do you distinguish an
> '-1' indicating an error from a the legal long-value '-1'?

hint: when in doubt, grep the Modules directory in the
Python source distribution:

    v = PyInt_AsLong(obj);
    if (v == -1 && PyErr_Occurred())
        return NULL; /* error */
    /* v is valid */

hope this helps!

</F>




From guido at python.org  Wed Apr 17 16:39:02 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 17 Apr 2002 16:39:02 -0400
Subject: pre-PEP for optional 'pass'
In-Reply-To: Your message of "Wed, 17 Apr 2002 20:48:29 +1100."
             <Pine.OS2.4.32.0204172037220.16906-100000@tenring.andymac.org> 
References: <Pine.OS2.4.32.0204172037220.16906-100000@tenring.andymac.org> 
Message-ID: <200204172039.g3HKd2C30985@odiug.zope.com>

> On Tue, 16 Apr 2002, Guido van Rossum wrote:
> 
> > phil hunt wrote:
> >
> > > I wish to make the pas startement optional. That is, in any Python
> > > program where there is a pass it can be left out and the meaning of
> > > the program is unchanged.
> >
> > I expect I may be too late here, but I'd like to point out that I
> > will never accept such a proposal, so you can stop wasting your time.
> > I've seen your arguments, and they don't convince me.  Sorry. :-)
> 
> That is your prerogative.  However I understood one of the roles of
> PEPs was that when rejected, they recorded for posterity your
> reasons for rejecting them.  Such records would help quench the
> recurring bouts of foaming-at-the-mouth featuritis that clogs
> c.l.py/python-list from time to time, as the afflicted could be
> directed to the PEP - it may not cure them, but it would tell them
> why they were beyond help...

In this particular case, it seems to me that all the reasons to reject
this idea have already been rehashed ad infinitum in the newsgroup.
If Phil still wants to write the PEP, that's fine, but he should then
be prepared to record the rejection reasons, even if he disagrees with
them.

> Alas several rejected PEPs (don't recall which off the top of my head)
> have just been recorded as "rejected" with no rationale for rejection.

That's a bug in the PEP process.  Rejected PEPs *should* have a reason
for rejection, and the authors of those PEPs should be reminded of
their responsibility.

> Which _is_ a waste of effort, IMO.

Half true.  It still tells people not to try proposing the same idea
again.

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




From sholden at holdenweb.com  Sat Apr 20 12:05:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 20 Apr 2002 12:05:50 -0400
Subject: The Python Manual Index
References: <a9pc0t$54o6t$1@ID-59885.news.dfncis.de> <uc0d5d68i0vhf1@corp.supernews.com> <Xns91F5997F18E9ARASXnewsDFE1@130.133.1.4> <nb2w8.4665$T%5.1737@atlpnn01.usenetserver.com> <a9rj1l$5gq1n$1@ID-59885.news.dfncis.de>
Message-ID: <Tngw8.26161$T%5.11592@atlpnn01.usenetserver.com>

"Thomas Heller" <theller at python.net> wrote ...
>
> "Steve Holden" <sholden at holdenweb.com> wrote ...
[ ... ]
> > Well, all they really needed to do was link to
> >
> >     http://web.pydoc.org/
> >
> > though there's no denying that Thomas' code is effective and useful.
> >
> One difference at least is that my index search not only searches the
module
> index, but also the C-api index.
> Try http://starship.python.net/crew/theller/pyhelp.cgi?keyword=Py_INCREF
> for example.
>

That's really neat. Though actually I'll have to set my version *not* to
search the C-api index, since I try to avoid that rather complex part of
Python known as "the implementation".

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From martin at v.loewis.de  Thu Apr 18 08:59:45 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Apr 2002 14:59:45 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019131683.24506.python-list@python.org>
Message-ID: <m3ofghp3ge.fsf@mira.informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> Yes.  Does this mean that my Solaris2.7 system is broken, or is this a
> problem with Solaris2.7 generally?

I'm still not sure what happens on your system - this is the first
report in a long time where linking fails to export symbols from the
Python binary - the last report was a few years ago, and lead to the
add of -Wl,-export-dynamic.

In general, Python builds on Solaris out-of-the-box, both with GCC and
with the SunPRO compiler. The problems with non-PIC code only occurs
in the following case:
- users have installed additional non-Sun packages, such as ncurses,
  or GNU readline;
- they use GCC
- they do not use GNU binutils

For that case, the default installation is a best-effort scenario: all
extensions that can be built successfully will be; for the other
(usually few) extensions, manual intervention is necessary and
possible.

There is a recent new kind of failure on Solaris, where people use gcc
3. This installs libgcc_s.so.1, and links binaries with it, but the
Solaris ld.so.1 won't find libgcc_s.so at run time. Again, there are
different ways to solve this - but I'd consider this a problem of the
GCC 3 installation.

Regards,
Martin


From phr-n2002a at nightsong.com  Tue Apr  2 22:04:01 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 19:04:01 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org>
Message-ID: <7xy9g54h5q.fsf@ruckus.brouhaha.com>

David Ascher <DavidA at ActiveState.com> writes:
> IIRC, there wasn't nearly this much heat in the discussion over the
> long/int unification or the type/class unification proposals, even
> though the latter especially was much more complicated, ambitious, and
> compatibility-challenging.


It's a cost/benefit thing.  Type/class unification may have cost more
in complexity than bools cost, but it has high benefits.  The bool
thing would a complicated change with limited benefit, so it's provoked
some sharp reactions.

> PS: I wish that this much heat got generated discussing changes and
> additions to the standard library...

It's usually easier to justify ADDING stuff than CHANGING stuff.
Additions to the standard library almost never mess up the language
guts, so if they have benefits they should generally be added without
needing a lot of discussion.  

Changing old libraries in ways that break exisiting programs is
usually bad, unless the change is expected to remove more bugs than it
causes.


From maxx at easynerws.com  Tue Apr 16 12:27:43 2002
From: maxx at easynerws.com (maxx at easynerws.com)
Date: Tue, 16 Apr 2002 16:27:43 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com>

On Sun, 14 Apr 2002 19:36:27 +0100, philh at comuno.freeserve.co.uk (phil hunt)
wrote:

>IMO until Python developers adopt a policy of being very reluctant 
>to do this, many IT managers will be wary of using it in anything 
>other thsan toy projects, and Python's user base will not grow as 
>quickly as (IMO) it should.

Why? 

At my company we are stuck in the realm of Windows, and the IT staff must
constantly patch the webservers to keep up with security and bugfixes. A quick
check of the IT media will show management's displeasure over this continual
process, but it does not appear to have hampered the deployment of Windows
servers.

As a matter of fact, what seems to be more reliable is that IT managers will
deploy more of what they already have, what their developers are most
productive, or what their required component requires. If change rates were a
critical factor, no one would be deploying Microsoft's IIS.

As mentioned in elsewhere this thread, Java and VB, have and are undergoing
complete architecture restructuring, and they still are expanding their
territories.

Keep in mind, IT managers often do not think and act like developers.


From andy47 at halfcooked.com  Mon Apr  8 21:01:46 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Tue, 9 Apr 2002 01:01:46 +0000 (UTC)
Subject: Creating a DB app with mySql and Windows GUI
References: <a8rhml$1id$03$1@news.t-online.com>
Message-ID: <Xns91EB700172FE0andy47halfcookedcom@203.109.252.31>

"Alexander v. Zitzewitz" <a.zitzewitz at generic-home.de> wrote in news:a8rhml
$1id$03$1 at news.t-online.com:

> Hello Python Freaks !
> 
> What components should I use for building a GUI Database App with a MySQL
> Server and a GUI on Windows platforms with Python. The following 
components
> are needed:
> 
> - a feasable IDE with Debugger interface
> - a GUI Builder
> - a MySQL inding
> 
> Please let me know, if anyone already did a similar project and what
> experiences he/she made with using Python
> 
> Best regards
> 
> Alexander v. Zitzewitz
> 
> 

<obligatory plug>
Take a look at PythonCard (http://pythoncard.sourceforge.net). It doesn't 
have absolutely everything you require, but its pretty good for alpha 
software. In particular, have a look at the dbBrowser sample application 
which shows how to interact with MySQL.
</obligatory plug>

Regards,
Andy
-- 
Contents free posts a speciality


From huaiyu at gauss.almadan.ibm.com  Wed Apr 17 19:49:31 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 17 Apr 2002 23:49:31 +0000 (UTC)
Subject: [OT] What is Open Source?
References: <mailman.1019083777.25904.python-list@python.org>
Message-ID: <slrnabs2gb.smi.huaiyu@gauss.almadan.ibm.com>

rasmussn at lanl.gov <rasmussn at lanl.gov> wrote:
>
>On Wednesday, April 17, 2002, at 02:28 PM, Oleg Broytmann wrote:
>
>>    It is not enough to have the code open. One needs the rights to read, 
>> to
>> use, to copy, to modify, to distribute modified and unmodified versions 
>> (so
>> everyone benefits from the open code). Only Free Software gives you such
>> freedom and rights.
>
>As I understand it, the GPL does not give you right to use the software
>as you see fit.  You can only use it by linking with other GPL'd
>software.

That is quite opposite of the truth.

You can use it _any way_ you want for your own purpose.  No problem at all.
Only when you distribute it, you are bound by GPL to licence it and other
linked software exactly the same way as you got it.

So if I want to use any GPL software together with any MS software using any
kind of linking, there is nothing in GPL to prevent me to do so, although
there are some new MS licenses that do explicitly prevent me to do so.

Huaiyu


From guido at python.org  Wed Apr  3 14:07:21 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 14:07:21 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 20:18:50 +0200."
             <031001c1db3c$0460ad90$ced241d5@hagrid> 
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff><yu99wuvovkmm.fsf@europa.research.att.com><00af01c1db2a$e7316d10$ced241d5@hagrid><200204031619.g33GJXm02577@europa.research.att.com><016b01c1db31$8da51010$ced241d5@hagrid> <yu99ofh0vg4x.fsf@europa.research.att.com>  
            <031001c1db3c$0460ad90$ced241d5@hagrid> 
Message-ID: <200204031907.g33J7Lo22817@odiug.zope.com>

> (Python docs tend to use "true" to mean anything which isn't
> "false", and where operator.truth returns the right thing.  this
> won't change post-PEP -- if you start interpreting "return true" as
> "return True", you'll be asking for trouble)

Unclear.  I intend to change as many functions as I reasonably can to
return True/False, and I intend to change their documentation to
match.

While some of the documentation indeed talks of "true" and "false",
those are not well-defined values.  They are properties of values, and
I find it bad style to say "f() returns true if X and false
otherwise".  I would have written that either as "f() returns a true
value if X and a false value otherwise" (in case I don't want to
commit to returning 0 or 1), or "f() returns 1 if X, 0 otherwise" (in
case I do).

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




From andrew at acooke.org  Tue Apr 23 23:02:54 2002
From: andrew at acooke.org (andrew cooke)
Date: 23 Apr 2002 20:02:54 -0700
Subject: XML, Python, XML-SIG, documentation, PyXml, 4
Message-ID: <aef33705.0204231902.40694af4@posting.google.com>

I was hoping to do some quick + dirty XML manipulation using Python,
but I'm having a hard time working out exactly what is available, what
is maintained, and where the documentation is.

The web pages at http://pyxml.sourceforge.net/topics/ seems to be out
of date.  In particular, the download talks about 0.6.6 when at
http://sourceforge.net/projects/pyxml there's a version of PyXml at
version 7.0 from December last year.  Also, the documentation at
http://pyxml.sourceforge.net/topics/docs.html is terribly incomplete.

Looking on the XML-SIG mailing list, there seems to still be activity,
with discussion of PyXml and 4Suite (http://4Suite.org and
http://sourceforge.net/projects/foursuite/).  There seems to have
documentation in the sourceforge download area.

Which of these packages should I use, and what documentation will help
me, if I want to:

- automatically generate a tree from the XML file (I believe this is
related to the DOM)
- supply a class to some (already existing) tree walking code that
does local rearrangements of the XML data
- automatically generate an XML file from the modified tree

(I've already looked at XSLT and I don't believe that is the right
tool for this kind of job).

Also, I know that it's boring updating web sites, and that this is all
a volunteer effort, but the current XML pages are not very useful.  It
would be less confusing to simply point people to the mailing list
archives, rather than to incomplete HOWTO documents that were last
changed years ago.  Perhaps the links to the sourceforge pages would
be useful too.

Thanks,
Andrew


From sholden at holdenweb.com  Mon Apr  8 15:19:29 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 15:19:29 -0400
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>
Message-ID: <S4ms8.1264$vT6.701@atlpnn01.usenetserver.com>

"Alastair Nicol" <calaban.madness at blueyonder.co.uk> wrote in message
news:40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk...
> Hi,
>   I've been enjoying the wonders of socket programming with Python. I have
> read a lot of information on the subject, but still have a few unanswered
> questions which is why, out of desperation, i'm asking here.
>
> 1) Can only hosts which can see the interface your program is bound to
> communicate with your program. In the "normal" case if you bind to
> 127.0.0.1 for example ONLY your machine can see that interface. Other
> machines cannot connect to it. Is this a safe / secure assumtion? (not
just
> for localhost)
>
Any address in the 127.0.0.0 to 127.255.255.255 space is by convention
reserved for the so-called "loopback" network. The transmit buffer filled by
the sending network layer is copied into a receive buffer and processed as
an incoming datagram by the receive portion of the stack. This means that
you can (for example) perform client/server tests on a single machine in the
absence of any network hardware at all.

By definition, therefore, any other machine trying to communicate with an
address on network 127 will see its own interfaces, not those of some remote
machine.

> 2) Also when a client connects to a socket the server gets the IP address
> of the client. Is this the "real" ip number? Ive heard you can spoof IP
> numbers so what hoops are needed to validate that the client is really
that
> client.
>
It's the IP address that your IP network layer found in the source address
field of the incoming datagram. This is normally the sender's IP address,
but of course on a DOS machine there's nothing to stop any user from putting
bigus packets out on the network, since there's no hardware protection. If
the sender has maliciously generated a datagram with some other host's IP
address as the source address there's nothing you can do to detect that.

If you need authentication then IPSec (or IPv6) contains features which
allow you to be fairly certain that IP address spoofing isn't taking place.

> openSSL is an option i would like to use, but for reasons to numerious to
> list here, i cant.
>
Never mind. Ideally, every network's external boundary router would filter
out IP datagrams that came from bogus addresses (i.e. those not inside
thenetwork), but alas we are a long way away from that desirable situation
as yet.

regards
 Steve






From duncan at NOSPAMrcp.co.uk  Tue Apr 30 08:39:30 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 30 Apr 2002 12:39:30 +0000 (UTC)
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de>
Message-ID: <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1>

Bernhard Herzog <bh at intevation.de> wrote in 
news:6qn0vlbcpk.fsf at abnoba.intevation.de:

>>     Ugh, that looks to me as appropriate as join() being a string
>>     method / module function.
> 
> Exactly. I still think that join as a function makes more sense than as
> a string method and for the same reasons as stated for chr above. Even
> if join is implemented as a string method.

I prefer writing
   str.join(' ', aList)
rather than
   ' '.join(aList)

When the string is a literal I find the former reads much better. YMMV

Sometimes though I take advantage of it being a method:
   joinlines = '\n'.join
   ...
   s = joinlines(aList)

Note that there was a suggestion on the python-dev list to move the 
remaining objects in the string module into the str type and then deprecate 
the string module. So str.hexdigits instead of string.hexdigits etc.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From jmarshal at mathworks.com  Wed Apr 10 14:48:02 2002
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 10 Apr 2002 18:48:02 GMT
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8vk7h$apk$1@news.mathworks.com> <MPG.171d167b861e1d669896a4@news.nwlink.com>
Message-ID: <a921d2$2og$1@news.mathworks.com>

Jeff Shannon <jeff at ccvcorp.com> wrote:
> In article <a8vk7h$apk$1 at news.mathworks.com>, 
> jmarshal at mathworks.com says...
>> 
>> To emulate the semantics of pointers, I like using single-element
>> lists, as some others have suggested.
>> 

> I prefer to *not* emulate the semantics of pointers.  If I have 
> data that needs to be changed in such a way, I'll put it in a 
> class and change it through a method.  If I wanted to fiddle with 
> pointers, I'd program in C.

But then you weren't the one asking about pointers in Python :)

I do agree though.  Other non-pointery approaches are cleaner.


From sholden at holdenweb.com  Sun Apr  7 13:30:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 7 Apr 2002 13:30:42 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com><mailman.1017865089.20676.python-list@python.org><Npor8.8281$q63.818@atlpnn01.usenetserver.com> <mailman.1018049233.13141.python-list@python.org>
Message-ID: <Lo%r8.5289$Sw5.2558@atlpnn01.usenetserver.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1018049233.13141.python-list at python.org...
> On Fri, 5 Apr 2002 16:09:41 -0500
> Steve Holden wrote:
>
> Steve, please be a bit more careful.  Your post sort of made it look like
> you were responding to things *I* had said (which I've snipped here) which
> were in fact said by John Roth.  I was even confused for a second =)
>
Mea culpa. I'll be more careful.

> [snip]
>
[snip snip]
>
> > > Frankly Guido's dismissal of Laura's arguments as FUD was not only a
> bit
> > > rude, but IMHO a bit short-sighted as well (given the fact that
> someday,
> > > there may be a desire for a real boolean type).  Maybe Tim Peters has
> > > borrowed his time-machine and not returned it ;)
> > >
> > However it happened, it left me feeling a little embarrassed that I had
> > originally defended Guido's inegrity and suggested that despite his
> > authorship the PEP's acceptance was not a foregone conclusion. I hope I
> > wasn't wrong, but that's not how it looks right now. Oh well.
>
> While I disagree with Guido's decision, I don't really think it has
> anything to do with his integrity (although it certainly could give that
> appearance).  I was fairly irritated at how this change went about (enough
> to contact Guido personally about it) and despite not reaching agreement
on
> the decision, I do feel certain that the issue was properly considered on
> his part and that he simply came to a different conclusion than many
others
> (myself included).
>
I don't think it has anything to do with his integrity either, but somebody
did suggest that the PEP was a foregone conclusion, and I tried to counter
that argument.

regards
 Steve






From bokr at oz.net  Tue Apr 16 16:31:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 20:31:16 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no> <s4zu8.20814$vF6.690664@news2.tin.it> <slrnabobad.8ho.mlh@vier.idi.ntnu.no> <SmWu8.26250$vF6.834282@news2.tin.it> <2yWu8.26048$Ji7.10720@atlpnn01.usenetserver.com>
Message-ID: <a9i1mk$hq6$1@216.39.172.122>

On Tue, 16 Apr 2002 10:26:10 -0400, "Steve Holden" <sholden at holdenweb.com> wrote:

>"Alex Martelli" <aleax at aleax.it> wrote in ...
>> Magnus Lie Hetland wrote:
>>         ...
>> >>I think the best name for what's now raw_input might be something
>> >>like getline.
>> >
>> > Or perhaps readline (? la file.readline)?
>>
>> I slightly prefer getline, because raw_input differs from a
>> somefile.readline -- the latter does not strip off the '\n',
>> while raw_input does -- each of the two slightly different
>> behaviors being optimal in its context.  I think it would be
>> harder to teach if the names coincided.
>>
>There's also the minor difference that raw_input() takes an argument to be
>used as an interactive prompt!
>
get_line_promptly ? ;-)

Regards,
Bengt Richter


From egrossde at yahoo.de  Thu Apr 18 18:43:19 2002
From: egrossde at yahoo.de (jb)
Date: Fri, 19 Apr 2002 00:43:19 +0200
Subject: OT Qt problem (but happens with PyQt as well)
Message-ID: <3cbf47ee_1@news2.newsgroups.com>

Sorry for this off topic question.

I use Linux (Suse 8.0). I compiled the example program "statistics". The
text content of the cells is  hardly legible and some of the cells, which
should have the background "red" have the background "black" instead (and so
their content is not legible at all).

I use Qt 3.03 which I downloaded and installed today.

Can anybody give mie a hint as to how can I have a nice spreadsheet?

Janos Blazi





-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From paul at boddie.net  Thu Apr 25 05:40:49 2002
From: paul at boddie.net (Paul Boddie)
Date: 25 Apr 2002 02:40:49 -0700
Subject: configuring mxCGIPython
References: <mailman.1019680127.28712.python-list@python.org>
Message-ID: <23891c90.0204250140.7248c6ed@posting.google.com>

"Mark McEahern" <mark at mceahern.com> wrote in message news:<mailman.1019680127.28712.python-list at python.org>...
> Hi, I'm trying to configure mxCGIPython:
> 
>   http://www.egenix.com/files/python/mxCGIPython.html#Install
> 
> I'm stuck at step 4, chmod 755 the path to the binary file cgipython.

[...]

> If I ftp to the server, I can chmod 755 the specific file.  If I change the
> script just to reference the filename (not the full path), it works.

With the ability to perform a chmod using FTP, I don't see the reason
for using the "install script" provided for mxCGIPython.

> The problem is that the shebang line for the Python cgi files needs to have
> the full path, I think.  If I don't chmod the full path, it doesn't work.
> The full path would be:
> 
>   #!/home/chain/pythons/cgipython

I found that it isn't always possible to deploy Python programs in a
cgi-bin directory and to use the above notation to persuade the
operating system to locate the Python executable. In the end, I used a
wrapper shell script for each of my Python programs.

Paul


From sholden at holdenweb.com  Tue Apr 30 09:18:21 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 09:18:21 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Xns91FF62618790Flumberjack@216.148.53.85> <b03e80d.0204291719.1f36c42b@posting.google.com> <Xns91FFDFB10888lumberjack@216.148.53.85>
Message-ID: <2Twz8.13741$xy4.9902@atlpnn01.usenetserver.com>

"Lumberjack" <lumber at jack.com> wrote in message
news:Xns91FFDFB10888lumberjack at 216.148.53.85...
> sandskyfly at hotmail.com (Sandy Norton) wrote:
> > I'm sure you know you can communicate 100% of your point with less
> > than 1% of your sarcasm. I'd hate to see people like you dissuade
> > future pep writers from articulating their ideas to the community.
>
> Nothing will dissuade fools; certainly not sarcasm. But for the non-fools
> who inadvertently do foolish things, a sharp prick, metaphorically
> speaking, will train them to ponder more carefully any future ideas before
> presenting them publicly and wasting people's time. (I'd change my nom de
> guerre to Ivan Petrovich Pavlov, but I don't recall him being portrayed in
> any Monty Python skit.)
>
Nom-de-guerre? Says something about your approach to dialog.

> Besides, I'm a lumberjack and I'm okay.

Clearly a matter of opinion.

not-in-need-of-training-just-now-thankyou-ly y'rs  - steve
--
Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From steve at ferg.org  Fri Apr 26 13:23:30 2002
From: steve at ferg.org (Stephen Ferg)
Date: 26 Apr 2002 10:23:30 -0700
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc>
Message-ID: <b16e4ef7.0204260923.b9ae868@posting.google.com>

This looks promising.  Amazon.com says that it will be published on May 21, 2002. 

Definitive XML Application Development (With CD-ROM)
by Lars Marius Garshol


From just at xs4all.nl  Sat Apr 13 04:38:17 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sat, 13 Apr 2002 10:38:17 +0200
Subject: New classes, __slots__, __dict__
References: <mailman.1018639826.18289.python-list@python.org>
Message-ID: <just-2F729A.10381713042002@news1.xs4all.nl>

In article <mailman.1018639826.18289.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> Another link:
> 
>     http://www.python.org/sf/520644
> 
> That's bug report "[520644] __slots__ are not pickled".  The short course is
> that someclass.__slots__ is unreliable because mutable, and so how to find
> all the actual slots in all the classes (direct and base) by magic is
> something of a puzzle and/or PITA.  Unless/until that does get done by
> magic, you need to define a __getstate__ to spell out which slots do and
> don't want to be pickled.

Is making __slots__ immutable no solution? Seems almost too easy...

Just


From bokr at oz.net  Wed Apr 17 17:52:47 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 21:52:47 GMT
Subject: pre-PEP for optional 'pass'
References: <Pine.OS2.4.32.0204172037220.16906-100000@tenring.andymac.org> <mailman.1019076217.19792.python-list@python.org>
Message-ID: <a9kqrf$1g3$0@216.39.172.122>

On Wed, 17 Apr 2002 16:39:02 -0400, Guido van Rossum <guido at python.org> wrote:
[...]
>That's a bug in the PEP process.  Rejected PEPs *should* have a reason
>for rejection, and the authors of those PEPs should be reminded of
>their responsibility.
>
>> Which _is_ a waste of effort, IMO.
>
>Half true.  It still tells people not to try proposing the same idea
>again.
The danger is that if they haven't seen _their_ rationale recorded and rejected,
they will delude themselves into thinking they have an irresistible rationale
that should be presented to sway you for the good of the universe ;-)

Regards,
Bengt Richter


From sanjay2kind at yahoo.com  Tue Apr 30 10:55:17 2002
From: sanjay2kind at yahoo.com (sanjay)
Date: 30 Apr 2002 07:55:17 -0700
Subject: How to retrive attached file from newsgroup mail
References: <63170f57.0204300045.788bc634@posting.google.com> <mailman.1020157433.29195.python-list@python.org>
Message-ID: <63170f57.0204300655.3ae78971@posting.google.com>

Oleg Broytmann <phd at phd.pp.ru> wrote in message news:<mailman.1020157433.29195.python-list at python.org>...
> On Tue, Apr 30, 2002 at 01:45:42AM -0700, sanjay wrote:
> >   Im using NNTP to retrive the newsgroups mails to display on web
> > pages.The email is multipart i,e text/html with doc,pdf,gif file has
> > attached. Is there any python script to get all the plain text/html
> > message with attached file from mail.
> 
>    http://phd.pp.ru/Software/Python/#extract_mime,
> http://phd.pp.ru/Software/Python/#mimedecode and faster mirrors at
> http://phd.by.ru/Software/Python/ and http://phd2.chat.ru/Software/Python/
> 
>    I am working on rewriting mimedecode.py to use email package. Actually
> I've finished rewriting it, but the code requires Python 2.3 (i.e. it
> requires my latest patches for the email package).
> 
> Oleg.


 extract_mime is helpful for 7/8 bit, not for base64(attached file).
any python way to  solve it.


From alex at NoSpamThankYouMam.com  Fri Apr 26 23:42:09 2002
From: alex at NoSpamThankYouMam.com (Alexander Whillas)
Date: Sat, 27 Apr 2002 13:42:09 +1000
Subject: email 2.0.1
Message-ID: <41py8.49965$uR5.117139@newsfeeds.bigpond.com>

hey y'all,

I just tried to install the email 2.0.1 package with the usual:

python setup.py install

on a windows2000 box running Python 2.2.1 (#34, Apr  9 2002, 19:34:33) and
I still get version 1.0 when I check the version. I've added:

"C:\Program Files\Apache Group\Python22\Lib\site-packages\email" and
"C:\Program Files\Apache Group\Python22\Lib\site-packages"

to my Pythonpath but still no joy. Anybody have the trick to installing on
winNT machines?

xelA




From DeepBleu at DeepBleu.org  Mon Apr 22 10:56:58 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Mon, 22 Apr 2002 09:56:58 -0500
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com> <ubu2a0dbso0u66@corp.supernews.com> <mailman.1019153737.24037.python-list@python.org> <slrnabu7rm.e84.gerhard@lilith.my-fqdn.de> <e2e9e807.0204192028.76e43129@posting.google.com>
Message-ID: <uc8954chc95taa@corp.supernews.com>

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in message
news:e2e9e807.0204192028.76e43129 at posting.google.com...
> OK, so the way I am reading this is as follows:
>
> Beacause the script is being run as a 'nobody' user, who doesn't have
> a .cshrc.mine file which I've got to set all my paths so all my
> modules work when run in my home directory - the user can't find the
> paths to the modules.
>
> So I need to set the environment variable so that DCOracle2 can find
> the oracle compiled whatever language modules.

dco2.so will look for libclntsh.so.  It needs to find this file.  This is in
your oracle directory.  In turn the oracle directory should be configured
such that you have ORACLE_HOME set up correctly.

>These reside in
> /u04/app/oracle/product/OraHome1/lib on our systems here at UNI.
>
> So I need to add this to my LD_LIBRARY_PATH. But I need to do this in
> the start of my CGI script. But what should I use to set this:
>
This part is better set up in the profile used.

> import sys
> sys.path.append(PATH TO LIBS)
> or:
>
Check the post I sent on the above and how to use it with append or insert.
 This should have the path to the DCO2 directory.

> import os
> os.environ[LD_LIBRARY_PATH] = 'PATH TO LIBS'
>
NO!  Do not do that!  Set this up in the profile.
And PLEASE send the ORGINAL ERROR message you encountered before moving
anything into CGI directory.
If you expect help, please help us help you.
DeepBleu





From Markus_Ewald at gmx.net  Tue Apr 23 11:39:04 2002
From: Markus_Ewald at gmx.net (Markus Ewald)
Date: Tue, 23 Apr 2002 17:39:04 +0200
Subject: Linker error when doing MSVC debug builds
References: <3CC535A7.2070607@gmx.net> <aa3fkj$7h6s3$1@ID-59885.news.dfncis.de>
Message-ID: <3CC58018.5010205@gmx.net>

Thomas Heller wrote:

> "Markus Ewald" <Markus_Ewald at gmx.net> wrote in message news:3CC535A7.2070607 at gmx.net...
>>The windows setup program only contains python22.lib
>>However, if you do a debug build of your project (and _DEBUG is defined)
>>pyconfig.h tries to link python22_d.lib, resulting in a linker error.
>>
>>
> So why don't you install the source distribution and compile it yourself?
> This way you have all the debug libraries. Occasionally stepping through
> Python's source may be very instructive or even help you better locate
> bugs in your extension.
> 


Because that's what the SDK is for, imho.

The 'extension' in this case is a script server module which /embeds/ 
python as scripting language into a 3D engine, "The Nebula Device" 
(www.radonlabs.de). Anyone interested in working with the CVS distro of 
the engine is currently required to download and compile Python himself 
in order to be able to debug the engine's code.

I don't mind compiling python myself, but neither do I see any reason to 
prevent windows users from doing debug builds with the SDK.

Plus, making _Py_DeAlloc() public in release builds would not influence 
any existing project at all. Debug/Debug, Release/Release would stay 
as is, only would Debug/Release builds become possible.


> Thomas
> 

-Markus-



From pecora at anvil.nrl.navy.mil  Wed Apr  3 15:16:18 2002
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Wed, 03 Apr 2002 15:16:18 -0500
Subject: CLAPACK Build on Mac w/ NumPy?  Help.
Message-ID: <030420021516186189%pecora@anvil.nrl.navy.mil>

Has anyone been successful getting CLAPACK working with Numeric Python? 
I have the source and a codewarrior compiled library for CLAPACK, but I
have not clue how to integrate the two and have the real CLAPACK
replace the lite version that comes with Numeric.Py.  I checked the
docs, but found nothing.  Maybe I'm overlooking something.  

Any clues greatly appreciated.

-- 
-- Lou Pecora

  - My views are my own.


From bokr at oz.net  Sat Apr 13 20:30:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 00:30:52 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCOEOBONAA.tim.one@comcast.net> <mailman.1018723227.15055.python-list@python.org> <V01u8.27065$043.2467943@bin6.nnrp.aus1.giganews.com>
Message-ID: <a9aijs$6ha$0@216.39.172.122>

On Sat, 13 Apr 2002 21:01:41 GMT, "Terry Reedy" <tejarex at yahoo.com> wrote:

>
>"Mike C. Fletcher" <mcfletch at rogers.com> wrote in message
>news:mailman.1018723227.15055.python-list at python.org...
>> I was unaware that there was no guarantee for 0-length tuples always
>> being unique objects.  There are lots of places in my code where I
>use
>> 0-length tuples in this pattern:
>>
>> NULL = ()
>>
>> class A:
>> def b( self, object=NULL ):
>> if object is NULL:
>> #was not specified
>> elif object is None:
>> #specified to be object None
>>
>> Of course, I can use:
>>
>> NULL = []
>>
>> But I'd always avoided that because of bad associations with having
>> mutable objects as default parameters (yes, it doesn't actually
>matter
>> in this pattern, but it's a thirty-times-burned-thirty-first-shy
>thing).
>>   Suppose I'll wind up using singleton instances instead.
>
>Given that None is a possible input, so it cannot be used as the
>no-input flag, you could also try something like NULL="OpTioNaL ArG
>NoT GiVeN"
>
You could also do:

 >>> class A:
 ...     def b(self, *obj):
 ...         if not obj: return 'obj not specified'
 ...         return 'did something with "%s"' % (obj[0],)
 ...
 >>> a=A()
 >>> a.b()
 'obj not specified'
 >>> for x in [None,(),[],{},'',0]: print a.b(x)
 ...
 did something with "None"
 did something with "()"
 did something with "[]"
 did something with "{}"
 did something with ""
 did something with "0"

Regards,
Bengt Richter


From python at rcn.com  Mon Apr 29 03:18:58 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 03:18:58 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <d06z8.12562$iU4.1091750@newsread2.prod.itd.earthlink.net>
Message-ID: <aais78$lfu$1@bob.news.rcn.net>

"Chris Tavares" <christophertavares at earthlink.net> wrote in message
news:d06z8.12562
> >         from functionals import map, filter, reduce
> >         from math import pow, divmod
> >         from deprecated import input
<snip>
>
> The "directive" needs to be in the code. And you've already provided the
> form of this directive: the import statements above.

Agreed.

>
> I would think the reasonable migration would be to add the modules listed
> above (or add the functions as mentioned), but leave the builtins for a
> version or two. Then remove the builtins.

Agreed.


Raymond Hettinger




From ffjhenon at club-internet.fr  Tue Apr  9 16:02:44 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Tue, 09 Apr 2002 22:02:44 +0200
Subject: stdout redirect : Howto with popen2.popen3
References: <3CAF8AC1.90704077@club-internet.fr> <a8rl4s$15sk$1@norfair.nerim.net>
Message-ID: <3CB348E4.487C606F@club-internet.fr>

I'll give it a try.

I may still do dumb mistakes . I am still a beginner.

Merci for the advice

Fabien

Eric Brunel a ?crit :

> Fabien Henon wrote:
> [snip]
> >   2. How can I tell when the raytrace is over ? The poll() gets out of
> > the loop.
> > What is command to listen to a program to tell if it's running or not?
> [snip]
> > while run.poll() <>-1:
> >      print "attente"   # // I never got to here and goes straight to
> > fd=open.....
>
> Errr, maybe you just inverted the condition? The library reference says for
> the poll method on the Popen3 class: "poll() Returns -1 if child process
> hasn't completed yet, or its return code otherwise.". So, if I understand
> correctly, that means that while the child is running, poll() returns -1,
> and when the child has exited, poll() returns its exit code. So your loop
> should be:
>
> while run.poll() == -1:
>   # etc...
>
> Didn't try it though. HTH however...
>  - eric -



From peter at engcorp.com  Thu Apr 11 22:41:18 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 11 Apr 2002 22:41:18 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  
 division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <wTit8.1381$F43.347@atlpnn01.usenetserver.com> <a95cqa$adr$1@panix1.panix.com>
Message-ID: <3CB6494E.E051CBA5@engcorp.com>

Aahz wrote:
> 
> In article <wTit8.1381$F43.347 at atlpnn01.usenetserver.com>,
> Steve Holden <sholden at holdenweb.com> wrote:
> >
> >Personally I regard your frequent posts on this group as a significant
> >contribution to "the community" (maybe because I'd like my own to be
> >regarded in the same way), so I think you're being entirely too
> >self-denigrating.
> 
> Besides, you never know when a bunch of Usenet posts might turn into a
> book contract.

Hah!  If I had time to write books, I'd definitely have time to 
contribute to the Python core in some way.

Or, I could stop spending so bloody much time posting to Usenet... :-)

Thanks for the good thoughts Steve.  I guess we contribute some.

On the other hand, I think I could post ten messages a night and
still not contribute as much value as someone who manages to submit
a few patches a year to the core.  I really can't emphasize strongly
enough how much I think we should value those who actually write
or fix core code.

-Peter


From borcis at geneva-link.ch  Thu Apr 25 17:31:27 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 25 Apr 2002 23:31:27 +0200
Subject: [OT] Re: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com>
Message-ID: <3CC875AF.98D5557E@geneva-link.ch>

Aahz wrote:

> What if there were no rhetorical questions?

If they were no rhetorical questions...

..questions would be believed either more or less
related to answers, than they currently are on
the average.

Boris Borcic
---
assert BOA in BAOBAB, "Paradise Mismatch Error"


From fdrake at acm.org  Mon Apr  1 11:37:06 2002
From: fdrake at acm.org (Fred L. Drake)
Date: Mon,  1 Apr 2002 11:37:06 -0500 (EST)
Subject: [development doc updates]
Message-ID: <20020401163706.D36FF18ED64@grendel.zope.com>

The development version of the documentation has been updated:

    http://python.sourceforge.net/maint-docs/

Documentation being prepared for Python 2.2.1.  This version includes
Andrew Kuchling's "What's New in Python 2.2" alongside the original
documents.




From lac at strakt.com  Wed Apr  3 12:18:15 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 03 Apr 2002 19:18:15 +0200
Subject: Threading problem 
In-Reply-To: Message from "Larry Whitley" <ldw@us.ibm.com> 
   of "Wed, 03 Apr 2002 09:52:07 MDT." <a8f8f6$vdu$1@news.rchland.ibm.com> 
References: <a8f8f6$vdu$1@news.rchland.ibm.com> 
Message-ID: <200204031718.g33HIGrq029056@ratthing-b246.strakt.com>

> I have a PC with two processors at work and have decided to try to use
> threading to run some long running console applications.  My basic idea is
> to start two threads, one for each processor, and give them a list of things
> to do.  As each thread finishes it one job will go to the list and pick off
> the next thing to do.  When the list is empty, both threads are done.
> 
> Here's a snippet of the code that inherits from threading.Thread:
> 
> class MyThread(threading.Thread):
>     def __init__(self, thName, jobList, jobLock):
>         threading.Thread().__init__(self, name=thName)
>         self.jobList = jobList
>         self.jobLock = jobLock
> 
> 
> My code is failing in first statement of the constructor, you can see the
> error below.
> 
> 
> X:\Projects\Pecos\ttc6\thsp>run.py
> Traceback (most recent call last):
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 59,
>     main()
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 46,
>     th1 = MyThread( "Thread1", jobList, jobLock )
>   File "X:\Projects\Pecos\ttc6\thsp\run.py", line 9, i
>     threading.Thread().__init__(self, name=thName)
>   File "C:\Python22\lib\threading.py", line 355, in __
>     assert group is None, "group argument must be None
> AssertionError: group argument must be None for now
> 
> The thing that puzzles me is that threading.Thread() is insisting that
> group=None.  That's ok with me, and I haven't changed it, yet it's
> complaining as if I did.
> 
> What am I doing wrong?
> 
> Larry
> Running Python 2.2

Take out the self in threading.Thread().__init__(self, name=thName)

Queue.Queue() was designed for exactly what you want to do.  Its dead
easy to use, as opposed to managind your own List and locks.
See http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
for an example of how to use it.  Or go read Aahz's tutorial.

Laura Creighton




From philh at comuno.freeserve.co.uk  Sun Apr 14 14:36:27 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 19:36:27 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org>
Message-ID: <slrnabjj1b.741.philh@comuno.freeserve.co.uk>

On Sun, 14 Apr 2002 01:02:17 -0400, Tim Peters <tim.one at comcast.net> wrote:
>> ...
>> In that particular debate, a lot of people were asking for certain new
>> features to be LEFT OUT of Python. I don't understand how someone can
>> volunteer to contribute work, time, or money toward leaving something
>> out.
>
>That was barely mentioned in the "crushing amount of debate over 'stability'
>[that] has gone by on Python-Dev the last week", so I don't know what "that
>particular debate" means to you, but it surely doesn't refer to anything in
>the paragraph you were responding to.  The debate on Python-Dev has about
>backward compatibility, and especially about producing bugfix releases for
>older releases for a much longer time.  Those things take work, and lots of
>it.  Note that James (to whom I was replying) was making the case for why
>commercial support is important to businesses -- nobody is going to buy a
>support contract that only guarantees not to add new features.

It is not new features that are the problem, it is new features that 
break existing code.

IMO until Python developers adopt a policy of being very reluctant 
to do this, many IT managers will be wary of using it in anything 
other thsan toy projects, and Python's user base will not grow as 
quickly as (IMO) it should.

>  Quite the
>contrary, business wants bug fixes for older releases, and they *also* want
>new features in old releases, especially in the libraries (like, e.g., SSL
>support for Windows, and support for hot new protocols

Indeed. And the best way to make sure that Python gets support for 
new protocols is to make sure the user base is big enough that 
someone is bound to write an open-source library to support any new 
protocol. And the best way to achieve that is to ensure backwards 
compatibility.

In summary:

   New features => good

   New features that break existing code => bad

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From slackness at caramail.com  Sat Apr 13 18:44:44 2002
From: slackness at caramail.com (slack)
Date: 13 Apr 2002 15:44:44 -0700
Subject: Developping for Windows CE??
Message-ID: <ab96a890.0204131444.19b35a3f@posting.google.com>

Hello... I've recently posted a message here but it's about something
else now... I program applications for Windows CE based handheld
computers like Ipaq or Jornada. I principally use eVB and sometimes
eVC++ for Ras connections for example. As I want to know more about
Python, I would like to know what benefits I could have in using
Python in Windows CE development? Could I develop applications for
using IrDA port? Could I program Ras connections? If anyone has more
informations to tell me about Python and Windows CE don't hesitate to
contact me...

d_vally at hotmail.com


From johnroth at ameritech.net  Tue Apr  2 21:05:28 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 2 Apr 2002 21:05:28 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>              <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com>
Message-ID: <uakogq1r0i0524@news.supernews.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CAA335F.4A17F6FF at alcyone.com...
> Guido van Rossum wrote:
>
> > I don't see the introduction of the bool type as painful at all.
> >
> > I find getting it accepted painful though. :-(
>
> I think of it in very simple terms:  True and False are just
> self-documenting values that indicate that you're talking about a
> Boolean value rather than something else -- and enabling the ability
to
> write self-documenting code is always good.  For backward
compatibility
> concerns, bool must be implemented as (effectively) a subclass of int.
> Very little code will break with the introduction of the PEP.  That is
> all.

I absolutely agree. For me, the entire point is providing
a **standard** way of expressing **in the code** the
notion that a particular variable only takes two values:
True and False. This is such a common case that it deserves
to be promoted to be a part of the language.

That is, after all, what programming languages are for:
to make my job as a programmer easier. When they
include standard ways of doing the things that need doing
in standard ways, they accomplish this.

Linguistic purity is a great topic for discussion over
pizza and beer, but if it gets in the way of adopting
something as simple as this, then there is something
radically wrong with our priorities.

John Roth

>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.
> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.




From ak at silmarill.org  Mon Apr 15 20:48:51 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Tue, 16 Apr 2002 00:48:51 GMT
Subject: Going through folders...
References: <mailman.1018913559.19895.python-list@python.org>
Message-ID: <slrnabmtb6.a8q.ak@ak.silmarill.org>

In article <mailman.1018913559.19895.python-list at python.org>, FREDRIK
HULDTGREN wrote:
> I want to go through a catalogue structure and make a textfile that
> lists the files in each catalogue/undercatalogue. I know that there is
> a function for doing this somewhere, I just can't remember what it is
> right now...  all help is appreciated...
> 
> /fredrik

import os

f = open("filelist.txt", 'w')

def visit(arg, dirname, names):
    f.write(dirname + '\n')
    for name in names:
        f.write("\t%s\n" % name)

os.path.walk(path, visit, None)

where path is path to walk through..

 - Andrei

> 
> 


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From paul at boddie.net  Tue Apr  2 09:44:29 2002
From: paul at boddie.net (Paul Boddie)
Date: 2 Apr 2002 06:44:29 -0800
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de>
Message-ID: <23891c90.0204020644.3ea4bc40@posting.google.com>

"Ruediger Maehl" <ruediger.maehl_nospam at web.de> wrote in message news:<newscache$d8oxtg$iv4$1 at www-neu.dzsh.de>...
>
> "Trung Hoang" <thoa0025 at mail.usyd.edu.au> schrieb im Newsbeitrag
> news:pRzp8.4373$hi7.15765 at news-server.bigpond.net.au...
> > What relational database is easy to setup and use with python? besides
> > msql (im having some problems)
> 
> SAPDB, a full featured database for free at www.sapdb.org
> with Python interface.

I'm not sure I would second that recommendation unless SAP DB has
changed dramatically over the past few weeks. I found that the
installation was quite a bit more complicated than, for example, MySQL
(at least on Windows) and upon issuing the arcane commands for
restarting database instances, the software would complain that the
instances needed restarting, which was hardly very reassuring with
respect to either the stability or the usability of the software.

As for the Python DB-API module, I couldn't get any version of it I
found to work with the database system. Admittedly, my objective was
to get a JDBC connection up and running, which was eventually
successful, and I therefore cut that particular side-investigation
short.

Despite it's often-cited shortcomings, MySQL and the MySQLdb module
are quite easy to set up. On UNIX systems, PostgreSQL and pgPySQL are
possibly as easy, although one doesn't need to suffer a manual with
PostgreSQL which consists of a single >1.5MB HTML file. I've also had
success with Sybase Adaptive Server Anywhere and mxODBC (without a
driver manager), but that software is commercially licensed and may
not even be available any more.

Another interesting issue that one might want to consider later on is
Unicode support with database systems and Python DB-API modules. Some
modules don't seem to support Unicode (pyPgSQL, for example), whereas
others apparently rely on the database system for such support to
actually function (eg. mxODBC). Some people may find this an
irritation earlier than others, however.

Paul


From ykingma at accessforall.nl  Fri Apr  5 17:26:56 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 05 Apr 2002 23:26:56 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it>
Message-ID: <3CAE24A2.7ECBB78@accessforall.nl>

Alex,


> >> This is the Python LANGUAGE we know and love -- the LANGUAGE
> >> Reference says so, and us Language Lawyers know that.  You want to
> >> change Python, that's one thing.  But don't tell us that this part of the
> >> Language Reference "is not part of the python language".
> >
> > I do not want to change that part of the language. I tried to
> > say that the backward compatible 1/0 behaviour and
> > the Something/Nothing behaviour can be implemented in the special methods
> > of the standard types, even when a non int-a-new-hat bool would be added
> > to the language.
> 
> Even if having to modify the __getitem__, __setitem__, __delitem__ (at
> the least) of a zillion user-defined types out there was feasible (and it's
> not), I think this would not fully exhaust, as I indicated.

The users writing such code were knowingly taking a risk by using
a name that starts with a double underscore. For these cases I think
deprecation is quite acceptable.

(Off course one could simply iirc :1,$s/\<__//g in all concerned code,
but that would cause too many other problems.)

> I use slices in every sequence-like type I code -- a slice object
> is what gets passed to my __getitem__ (etc) methods when
> client code uses myobject[3:] or whatever (special methods
> __getslice__ etc are deprecated for new objects, although for
> backwards compatibility they're still used if present, and quite
> unfortunately list &c still use them).  My __getitem__ &c need
> to know whether they're called with an index that has .start
> and .stop attributes (I feature-test for those, obviously, NOT
> typetest for slice!) or with an index that is a number.  I feel
> absolutely no need to know whether an argument passed to
> my object's method comes from -- and on the other side of
> things, to know whether my object is being tested for truth,
> why, I already have the (unfortunately named) __nonzero__.

When the object passed to your __getitem__ method does not
have .start and .stop attributes you could simply call int()
on it from within your __getitem__ implementation.
But you would only need to do this in certain cases, see below.

> > Indexing a dict with both int and bool keys might involve some confusion
> > when True is not 1 anymore. This case should be rare, however.
> 
> Why should it be rare?  All hashables are used uniformly as keys
> into dictionaries.

It should be rare because the confusion would only occur when both
True and 1 are used as keys and they are intended to be the same key.
That should only happen when True and 1 are used interchangeably,
which might happen by programmer mistake or by using 
old (1/0) and new (True/False) style code together.

> > For standard types, ie. normal dictionaries there is no problem since
> > one can simply add the int() method call in the implementation of
> > __getitem__ of the standard dictionary for backwards compatibility.
> 
> You propose a substantial slow-down of standard dictionaries "for
> backwards compatibility" against an incompatibility that need never
> be introduced in the first place -- you can't "simply add the int()
> method call", you first have to typetest against boolean on EVERY
> dictionary access to know if the "int()" kludge is needed at all (you
> can't just call int() on EVERY key without breaking a megazillion
> lines of code -- i hope you realize that? -- and if you could that
> would be a slow-down too).  I don't think you realize at all how
> CRUCIAL dictionary look-up performance is to Python.  The number
> of dictionary accesses is huge in every real Python program.  If there
> is one thing that MUST be always optimized, that's the one.
> 
> The incompatibility is needless: apart from un-engineeringly yearnings
> for "purity" (which, in Python, is trumped by pragmatics) there's no
> reason for hash(True) to differ from 1, and True==1 is similarly not
> a problem at all.  Easiest would be not to distinguish booleans from
> ints at all, second easiest, as the PEP proposes, to make booleans
> a subclas of ints, but even without the latter pragmatic fix a sensible
> strategy would still be feasible.  Slowing down all dictionary access
> is NOT a sensible strategy, nor acceptable at all.
> 
> Slowing down all list accesses as you want to do is not quite as
> horrible, but still a pretty bad move.  What Python most needs right
> now is a bit more attention to performance issues, NOT slow-downs
> in its core to satisfy some people's desires for "purity".

The slowdown would be there for backwards compatibility only.
In the pure case an int() call would not be needed at all: 
no one would index a list directly with a bool and a dict might repr as:

{True:'Used to be like one', 1:'The only one'}

ie. True and 1 would be different keys.
When adding the int() call to the __setitem__ method of the standard
dicts for backward compatibility the dict shown above would only a have
a single key 1 and the value would be the last value put there.

> > User code trying to implement a dictionary will need a similar
> > fix as I pointed out above. I can only hope there is no more breakage in
> > user code.
> 
> "only hope" is not acceptable in this context.  Fortunately, we can
> do better than that (and I'm confident that the unacceptability of
> slowing down dictionaries and lists is quite obvious to decision
> makers in this case, so I don't fear it will happen).

I'm still putting my hopes on deprecation (as above) after PEP 285 enters
into force. During the deprecation phase with non int bools such cases
could be warned for eg. by a special verion of int() that would also cause
a slowdown, and after the deprecation a pure version with normal speed
would remain.

Regards,
Ype


P.S. I havn't followed the change of behaviour of the / operator in detail,
but wasn't/isn't this done using a deprecation period with warnings, too?

-- 
email at xs4all.nl


From dswegen at software.plasmon.com  Thu Apr 18 06:40:52 2002
From: dswegen at software.plasmon.com (Dave Swegen)
Date: Thu, 18 Apr 2002 11:40:52 +0100
Subject: Python signal handling
In-Reply-To: <lkwuv6qtxa.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1019049215.30218.python-list@python.org> <lkwuv6qtxa.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020418104052.GI30292@software.plasmon>

On Wed, Apr 17, 2002 at 02:26:05PM +0000, Michael Hudson wrote:
> Dave Swegen <dswegen at software.plasmon.com> writes:
> 
> > I've recently been bitten by a nasty (undocumented) feature in Perl,
> > where installing a signal handler to catch SIGCHLD will eventually cause
> > a core dump. So before I venture into any daemon handling in python I
> > was wondering if Python too suffers from problems with signal handling?
> 
> Python suffers from problems with signal handling, but it sounds like
> they're not the same problems as those suffered by perl.
> 
> > I believe the reason for perl's problem is that it isn't re-entrant, and
> > thus two rapidly incoming signals will cause perl to fo *Pooof*.
> 
> In Python, signal handlers (usually) get run in between opcodes, so if
> an opcode takes a particularly long time to run, your signal handler
> may not run until some time after the OS delivers the signal to your
> process.

Which is how I believe perl is going to do it. Having an implementation
which is slow (and behaves somewhat unexpectededly) is far better than
one which is fast, but fatally broken (and BLEEP! undocumented!).

Thanks for the info.

Cheers
    Dave




From theller at python.net  Fri Apr  5 14:11:38 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 5 Apr 2002 21:11:38 +0200
Subject: CGI GET vs. POST
References: <a8k613$suseq$1@ID-59885.news.dfncis.de> <a8kfse$teqtj$1@ID-59885.news.dfncis.de> <slrnaars0t.og.gerhard@developer2.opus-gmbh.net>
Message-ID: <a8ksti$t2jhl$1@ID-59885.news.dfncis.de>

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message news:slrnaars0t.og.gerhard at developer2.opus-gmbh.net...
> In article <a8kfse$teqtj$1 at ID-59885.news.dfncis.de>, Thomas Heller wrote:
> > [replying to my own post]
> > "Thomas Heller" <theller at python.net> wrote in message news:a8k613$suseq$1 at ID-59885.news.dfncis.de...
> >> I'm trying to use the cgi module in a CGI script,
> >> but encountered a problem.
> >>
> > [snip}
> >>
> >> If I use method=POST in the CGI-script, the
> >> "keyword" is found as a field in FieldStorage,
> >> but NOT if I use method=GET.
> >> Is FieldStorage only usable with POST?
> >>
> >
> > Problem solved, although I don't understand why.
> > Maybe someone can enlighten me?
> >
> > The script was in a file pyhelp.cgi on the server,
> > and I called it with 'http://...../pyhelp' when it did
> > not work.
> > Calling it with 'http://..../pyhelp.cgi' made it work.
>
> Ask yourself: why should this have worked?
>  Just because not providing the
> extension for executables does work on Windows? ;-)


It was a typical beginners mistake in this case (not with
Python for me, but with CGI on a *linux* machine ;-):
In addition to the pyhelp.cgi script there was a pyhelp script
around in this directory.

Thanks anyway,

Thomas




From aleax at aleax.it  Thu Apr 11 03:57:03 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 07:57:03 GMT
Subject: a list from lists
References: <3CB53D35.5030308@beth.uniforum.org.za>
Message-ID: <jlbt8.6127$vF6.195339@news2.tin.it>

mixo wrote:

> I have a list of lists (which are not of the same size but can be).
> How can I join lists to for a new list with elements from
> the list?
> 
> i.e. from,
>    listOfList = [ [1,2,3],[4,5],[6,7,8,9]]
> 
> I want to get,
>    NewList= [1,2,3,4,5,6,7,8,9]
> 
> The new list does not need to be in any particular order.

If you really have only one level of nesting,

newList = [ x for sublist in listOfLists for x in sublist ]

meets the bill.  On the "do the simplest thing that can
possibly work" principle, this is what you should do IF
only one level of nesting does indeed suffice for you.

More interesting is the ability to handle different and
arbitrary level of nesting.

First you have to decide what is an "atom", i.e., something
that's acceptable as an item in your final list.  "Anything
that's not a list" is most likely wrong (UserList should be
treated like a list, and why not tuples, ...?).  I suggest,
in Python 2.2:

def isAtom(item):
    try: iter(item)
    except: return 1   # not iterable -> surely an atom
    else:              # iterable -> atom if string-like
        try: 1 / (item in item)
        except: return 0         # normal iterable
        else: return 1           # string-like

but it makes no difference to the main issue if you want
to use some other definition such as the unpleasant:

def isAtom(item): return not isinstance(item, list)


Once you do have this, Python 2.2 makes it simple with
generators:

from __future__ import generators

def flatten(sequence, isAtom):
    if isAtom(sequence): yield sequence
    else:
        for item in sequence:
            for subitem in flatten(item, isAtom):
                yield subitem

Should you really need a list rather than (as more
often happens) an iterator,

newList = list(flatter(listOfLists))

will work,


If you need to run on Python 2.1, you can, by
producing a list directly, e.g.:


def flatten(sequence, isAtom):
    if isAtom(sequence): 
        return [ sequence ]
    else:
        return [ x for item in sequence
            for x in flatten(item, isAtom) ]


You can even do without list comprehensions
if you need to use Python 1.5.2 -- just change
any return [ x for a in b for x in a ] into the
equivalent nested loop construct:

    result = []
    for a in b:
        for x in a:
            result.append(x)
    return result


Alex



From grante at visi.com  Wed Apr 17 14:45:50 2002
From: grante at visi.com (Grant Edwards)
Date: Wed, 17 Apr 2002 18:45:50 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <a9kebg$ub9$1@slb5.atl.mindspring.net>
Message-ID: <ypjv8.42546$vm6.5594931@ruti.visi.com>

In article <a9kebg$ub9$1 at slb5.atl.mindspring.net>, Andrew Dalke wrote:
> Philipp Lenssen:
>>can I use something like
>>    verboseGender = (gender == 'm') ? 'male' : 'female'
>>in Python?
> 
> No.  Search the archives for "ternary operator" for plenty of
> past discussion.  Google lists 65 posts for comp.lang.python.
> 
> Besides, for this specific case you should have some error checking,
> because there are genders other than 'male' and 'female' (eg,
> asexual bacteria and hermaphradite earthworms).

And in German, there's "neutral".

Unless we _are_ talking about bioligical stuff, then the term
you want is probably "sex" rather than "gender" (which I read
somewhere has always been a grammatical term not a biological
one until recently?), and you've may have to differentiate
between phenotype and genotype depending on the eventual use of
the data....

[somebody else has already posted a useful answer]

-- 
Grant Edwards                   grante             Yow!  Being a BALD HERO
                                  at               is almost as FESTIVE as a
                               visi.com            TATTOOED KNOCKWURST.


From zopestoller at thomas-guettler.de  Tue Apr 23 11:23:11 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 23 Apr 2002 17:23:11 +0200
Subject: Class Methods with Python 2.1
Message-ID: <3CC57C5F.7040804@thomas-guettler.de>

Is it possible to have methods in classes
which don't take a self argument, because
they only change attributes common to all
instances of a class?

class car:
     available=["porsche", "volvo", "trabbi"]
     def __init__(self, name):
        if name not in available:
            raise "Name not in available cars"
        ....
     def add_available_car(name):
         availalble.append(name)

One solutionn is to define this method
at module level. But that's not nice.



Unfortunately I have to stick to Python2.1

   thomas



From skip at pobox.com  Mon Apr  8 10:19:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 8 Apr 2002 09:19:18 -0500
Subject: Carbonized embedded MacPython
In-Reply-To: <f69ccee4.0204080600.21d8ac87@posting.google.com>
References: <f69ccee4.0204080600.21d8ac87@posting.google.com>
Message-ID: <15537.42726.212806.915165@12-248-41-177.client.attbi.com>

    Matt> I would like to embed Python into an already-existing Carbon
    Matt> application.

Without commenting on anything else, I direct you to the Python Mac SIG:

    http://www.python.org/sigs/pythonmac-sig/

The email address is

    pythonmac-sig at python.org

You'll get much more useful help with Mac issues there than on the general
list, especially since what you want to do seems a bit out of the ordinary.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From aleax at aleax.it  Fri Apr 26 13:07:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 17:07:27 GMT
Subject: Converting a Python List to a C double *
References: <k0n4rhy3gtn.fsf@metropolis.phys.ntnu.no>
Message-ID: <jPfy8.65668$vF6.2006478@news2.tin.it>

<posted & mailed>

Joakim Hove wrote:

> 
> Hello,
> 
> I am writing a C-extension, and now I need to parse a list from python
> into a c double *:
> 
>    static PyObject * power(PyObject *self, PyObject *args) {
>      double *xdata;
>    
>      /*
>        This routine shall be called from Python. It will always be
>        called with a variable length list of floats: e.g.
>        module.power([0.1, 0.2, 0.45, 0.60, 0.70]).
>      
>        Within the c-function "power" I want to access the list in the
>        pointer xdata. How to parse?
>      */

/* warning, untested code, so there might be bugs, but the general
   idea would be something like: */

    PyObject* thelistofdoubles;
    int listlength;
    int i;
    PyObject* temp;

    if(!PyArg_ParseTuple(args, "O", &thelistofdoubles))
        return 0;
    if(!PySequence_Check(thelistofdoubles)) {
        PyErr_SetString(PyExc_TypeError, "argument must be a sequence");
        return 0;
    }
    listlength = PySequence_Size(thelistofdoubles);
    if(listlength<0)
        return 0;
    xdata = malloc(listlength * sizeof(double));
    for(i=0; i<listlength; i++) {
        temp = PySequence_GetItem(thelistofdoubles, i);
        if(!temp) {
            free(xdata);
            return 0;
        }
        if(!PyFloat_Check(temp)) {
            Py_DECREF(temp);
            free(xdata);
            PyErr_SetString(PyExc_TypeError, "all items must be float");
            return 0;
        }
        xdata[i] = PyFloat_AS_DOUBLE(temp);
        Py_DECREF(temp);
    }    


>      for (i=0; i<5; i++)
>        printf("x[%d] = %12.7f \n",i,xdata[i]);

Of course, you'll need to let i go no higher than listlength, which may
well be < 5, ot else risk a crash.

>      return PyNone();
>    }
> 
> Grateful for any suggestions -

Hope this helps...


Alex



From lac at strakt.com  Mon Apr  8 11:52:07 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 08 Apr 2002 17:52:07 +0200
Subject: PEP 285: Adding a bool type 
In-Reply-To: Message from Ken Seehof <kseehof@neuralintegrator.com> 
   of "Mon, 08 Apr 2002 07:34:28 PDT." <EFEFIHAINIINNELLHFAIMECBCAAA.kseehof@neuralintegrator.com> 
References: <EFEFIHAINIINNELLHFAIMECBCAAA.kseehof@neuralintegrator.com> 
Message-ID: <200204081552.g38Fq7dj019911@ratthing-b246.strakt.com>

Ken Seehof:

(I agree with everything in his article, and I am snipping it to this point
 because of this.)

> You have repeatedly argued that adding booleans to python is a bad thing
> specifically because it will make it more difficult to change functions
> such as IsRed to return more than two different values.  Well, changing
> IsRed to return more than two different values (without refactoring) is
> always a bad thing.  I don't see what's wrong with making bad things more
> difficult.

No.  This is  where we have a misunderstanding.  this is because of
the isleap function.  You have conceptualized that as 'a function 
that was designed to return a boolean' and therefore think that I want
to encourage people to write functions called issomethings to return
ints so you can extend them to return 3 values.

My understanding of the isleap function came with the baggage of
and Guido wants to let people be able to do arithmetic.  We never
had booleans before.  Therefore my conceptualization of the function
was that it was a function designed to return the integer number
of leap days.  Clearly its designer thinks that this can only be
0 or 1.  Thus its designer gave it the rotten name isleap. I think
both the name is poor and that doing arithmetic on bools is poor.

I think that the function that returns the number of leap days should
not be called isleap, but I fear that anybody who thinks, hah, leap
days are 0 or 1, therefore I should get out a bool, will make this
error.  This error is using a bool for something that can have more
than 2 states to return, confusing 'that this year 1, is a leap
year' with 'we get 1 more day'.  And I have oceans of sympathy for
whatever poor soul did that.  It is only when the function comes
to Sweden that the poor soul will go 'oops'.

> You have also argued that writing yes/no functions without foresight is a
> bad thing.  I agree with you.  And I am in favor of decreasing the total
> number of yes/no functions and variables.  However, I don't see any
> reason to think that adding booleans to python will encourage writing
> more yes/no functions.  This is like the proposition that making condoms
> available causes teenagers to have more sex.

(er, I think that you are claiming that making them available causes
teenagers to think about sex more often, or want sex more, or something
like that, which I think, like you, is foolish.  I sure hope that making 
condoms available causes teenagers to have more sex -- Every teenager who 
wants to have sex, has a willing partner, but doesn't because he or she 
doesn't want a pregnancy and there is no birth control, can now have sex.  
To my mind this is a _big_ _win_.

> In short, here's my advice:
> 
> 1. Do not implement functions or variables with yes/no semantics unless
> you are quite certain that the yes/no semantics are appropriate.  If
> you are a teenager, don't have sex.
> 
> 2. Once you have implemented yes/no semantics, you are committed, until
> you change your model and refactor accordingly.  Always use a condom.
> 
> This advice applies equally whether booleans are part of the language
> or not.  Adding booleans to python has no real effect other than to
> make any commitment to yes/no semantics more clear.  This helps people
> follow both pieces of advice.
> 
> I am glad booleans were added, and I recommend that they should usually
> be avoided.  :-)
> 
> - Ken Seehof


I agree 100% with your recommendation, and actually what I was saying
is that its hard to get students to abide by that recommendation.  It's
a hard recommendation to teach.

ps -- to others -- no, curse it all, I am not leaving my students with
   the idea that they should never use booleans or that they should
   write functions that return 500 integer ret vals.  Give me credit
   for some sense, please.  This is a training exercise to get you
   over the bad problem of writing bools when you shouldn't, nothing
   more.  My freind Kenji Okamoto, who has been teaching Python in
   Japan for 8+ years now says that this is a function of the English
   language, and Indo-European languages in general which have an 'is'
   verb.  People whose langages do not have them do not make too many
   booleans -- the trick is to get them to make any at all some times.
   But naive programmers in the Western language, he says, think always
   in terms of two valued logic 'is' or 'is not' properties.  This
   makes for bad programming practice.  It would be nice if he would
   discuss this more here, but his English is not great and he doesn't
   like to do so.  We must repect his wishes.

Laura Creighton




From BPettersen at NAREX.com  Sat Apr 27 14:10:55 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 27 Apr 2002 12:10:55 -0600
Subject: does readlines stop at 255?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192B9836E@admin56.narex.com>

> From: v.wehren [mailto:v.wehren at home.nl] 
> 
> It appears that readlines() stops at 255...when reading a line..
> 
> How can I work around this if my file has lines longer than 
> 255 characters (and making them shorter is not an option...)??

readline() can handle any length lines (up to available memory). There
must be something else wrong. Post code.

-- bjorn




From hgs at dmu.ac.uk  Thu Apr 18 05:43:56 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 10:43:56 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <m3elheyxhf.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181033450.8851-100000@neelix>

On 17 Apr 2002, Martin v. Loewis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > > You did not build statically. Instead, all extension modules are build
> > > by setup.py now.
> >
> > I left *shared* commented out.  So that was not enough, and neither was
> > *static* apparently.
>
> To build all modules statically, you need to to add *static* (or just
> not add *shared*), and remove the comments for all modules that you
> want to build statically.

So:

#
# In addition, *static* reverses this effect (negating a previous
# *shared* line).

is ambiguous, because it does not reverse all the effects of *shared*?
>
> > gcc   -o python \
> > 		Modules/python.o \
> > 		libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm
>
> Now we are getting somewhere. This line is wrong; it ought to have the
> option -Wl,-export-dynamic *if* you are using GNU ld. Can you tell
> whether you are using GNU ld?

I have tried to configure this in different ways:
configure
configure -- with-gnu-ld
changning the path to have /usr/local/bin at the front so gnu ld will be
found first.
In all these cases I can find no trace of /usr/local/bin/ld in the
config or make files resulting from running configure.

What else does it look for?  If I symling /usr/local/bin/ld to gld will
configure find it?
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Hugh





From nico at aries.logilab.fr  Thu Apr 25 08:36:33 2002
From: nico at aries.logilab.fr (Nicolas Chauvat)
Date: Thu, 25 Apr 2002 12:36:33 +0000 (UTC)
Subject: Creation of the Python-Logic Project/SIG
Message-ID: <slrnacfu2h.f56.nico@aries.logilab.fr>

Hello Pythonistas,

Since too few people voted +1 on the meta-sig mailing to agree with the
creation of the logic-sig, I decided to create it and host it at logilab.org.

Web site:

    http://www.logilab.org/python-logic/

Mailing-list:

    http://lists.logilab.org/mailman/listinfo/python-logic/

If you are interested in logic and constraint programming, please join the 
mailing list and help us move this forward. Logilab started to implement
some things internally and we will (slowly) publish them along the way, but
as usual, the more hands the more work gets done the faster :-)

See you there !

-- 
Nicolas Chauvat

http://www.logilab.com - "Mais o? est donc Ornicar ?" - LOGILAB, Paris (France)


From distrex at hotmail.com  Tue Apr  9 18:10:55 2002
From: distrex at hotmail.com (distrex)
Date: 9 Apr 2002 15:10:55 -0700
Subject: port scan
Message-ID: <d9ea2696.0204091410.7755db60@posting.google.com>

I don't know if this is possible but is there a way to create a port
scanner in python? If there is is there anyone who has successfuly
done it that can point me in the right direction. I know there is a
way to do it with VB and C++ but I'd like to stick with making
programs with python for now untill I understand it.


From oliver.erlewein at sqs.de  Mon Apr 15 05:17:27 2002
From: oliver.erlewein at sqs.de (Oliver Erlewein)
Date: 15 Apr 2002 02:17:27 -0700
Subject: Building Python 2.1.x on Solaris
Message-ID: <32dd230a.0204150117.3f4a5bcc@posting.google.com>

Hi all

I've got a large Zope-site. To run Zope I need to build python 2.1.x
on my Solaris (sparc & intel) boxes. But if I do a plain vanilla build
I have the following problem on the Zope side:

Zope is multithreading but if I run a big job then all other executed
jobs wait for the first big job before executig one after the other.
In short Zope runs single threaded.

I've also experienced similar behaviour on Linux. The thing is, that
as soon as I use precompiled stuff (RPM's) all works fine and
threading is no problem. Therefore I suppose that my standard build of
Python is missing something. So I'd like some comments on my build.

> make clean
> rm config.cache
> LD_LIBRARY_PATH=/usr/lib/lwp
> export LD_LIBRARY_PATH
> ./configure &#8211;-with-threads 
> make
> make install

I'm no pro what installing is concerned so please kee answers
understandable (sort of ....for dummies). Thanx a lot in advance!

Oh ... and I've searched the web & groups on the subject and the only
things that seemed interesting to me were OPT="-O2 -fPIC" and
CC="gcc". Do I have to use them and what do they do?

Please also respond to oliver.erlewein at sqs.de because I don't get this
group. Thanx.

Oliver Erlewein


From Chris.Barker at noaa.gov  Mon Apr  1 17:47:29 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 01 Apr 2002 14:47:29 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <3CA8E37A.E89CA002@noaa.gov>

Well, I've spent a great deal of time reading this thread, so I might as
well spend a little time putting in my $.02.

It seems to me that we have a fundamental problem here: a boolean type
is a great idea, and would be an excelent part of the Python language,
but it has not, as of yet, existed. In order to introduce one that
doesn't break a lot of existing code, it needs to be highly compatable
with the old system, which I think is a fundementally bad one. Frankly,
using 1 and 0 to mean true and false is bad enough, but every type (most
types?) also have values that can represent false ([], "", etc.). Having
the new boolean type mimic integers is probably the most pragmatic
solution, but fundamentally., using 0 and 1, rather than "" and "true"
is just a matter of convention. It is the most popular convention, by
far, but not defined in the current language. This makes it even harder
to come up with a backwards compatable solution.

I believe it was Alex Martelli who wrote the most convicing argument
that a boolean type that behaves almost exactly like the integers 0 and
1 is a bad idea. However, I am convinced that having a real boolean type
is such a good idea that it is worth a little pain in the transition.

By the way, here is an example of why the current sytem is bad:

string.find() returns the index of the found string, and 0 being a
perfectly valid index, it can NOT be used to return an indication that
the substring was not found, so -1 is returned instead. None could not
be returned to mean not found, because both 0 and None would resolve as false.
I'd love to see False be returned, and then you could do:

if string.find(...):
	do something.

There have been numerous threads from newbies (and not so newbies) who
tried to use this construct, and had it fail on them. Making the concept
of True and False separate from the concept of a value is a VERY good
idea. The number zero is a perfectly valid integer, having it stand in
for the concept of False is a bad idea. 

Of course, it is a major incompatabilty to have 0 not evaluate as false,
but I'd love to see the language head that way...Py3k maybe?

Here are my answers and comments to the original questions:

>     1) Should this PEP be accepted at all.
YES, YES, YES, ... but only if the intent is to make a purer boolean
type in the future.
 
>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".)

True.  str() is supposed to return an nice, human readable form of the
object. True is the only option for this. Anyone parsing the result of
str() should have been using repr(), and that will break anyway because
of the goal for eval(repr(x)) == x

>     3) Should the constants be called 'True' and 'False'
>        (corresponding to None) or 'true' and 'false' (as in C++, Java
>        and C99).

True and False, it fits better with Python and the rest of the languages
be damded
 
>     Most other details of the proposal are pretty much forced by the
>     backwards compatibility requirement; e.g. True == 1 and
>     True+1 == 2 must hold, else reams of existing code would break.

which is too bad...

>     Minor additional issues:
> 
>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

I absolutely think we should!!! In fact, if we're not going to head
toward a more pure Boolean, then we shouldn't put one in at all. Others
have called this a slippery slope, and I agree, but it is one we want to
be at the bottom of, so we should get there slow, but not try not to get
there. 28 + True is just rediculous, it's only about saving a little
typing, a looks like a hack to me. What the heck is wrong with:

if isleap(y) :
	days == 29
else:
	days == 28

or, if you want to be more compact:

days = 28
if isleap(y): days+=1

for a language that doesn't have a "case" statement, this kind of
trickery just doesn't fit. By the way, using a bool as an index is even
worse: 

x = list[a == b]

is that clear??? I can't even use a float as an index:

x = list[math.floor(y)]

Why should I be able to use a bool.

>     5) Should operator.truth(x) return an int or a bool.

A bool. It could even be defined as

def truth(x): return bool(x)

yes, it's completely redundant with bool(), but it could be depricated
as well.

int(bool(x)) 
							
makes it quite clear what you want.

> Rationale
> 
>     Most languages eventually grow a Boolean type; even C99 (the new
>     and improved C standard, not yet widely adopted) has one.

People have dismissed this as irreleant, but I only somewhat agree. If
most langages do something, it doesn't mean Python should, but it does
mean that it probably has some good reasons to be considered.

>     Many programmers apparently feel the need for a Boolean type; most
>     Python documentation contains a bit of an apology for the absence
>     of a Boolean type.  I've seen lots of modules that defined
>     constants "False=0" and "True=1" (or similar) at the top and used
>     those.

I've never done that, but I have always wanted a boolean type, I'd only
use it if it was a standard part of the language. As I made clear above,
I want it to be a lot different that just an integer that happens to be
valued 0 or 1.

>     I don't see this as a problem, and I don't want evolve the
>     language in this direction either; I don't believe that a stricter
>     interpretation of "Booleanness" makes the language any clearer.

Here's where I disagree. A stricter interpretation of "Booleanness"
would make the language clearer, and a type that sometimes act like an
integer, but isn't really, makes the language less clear. I think it's a
fine compromise during a transistion, but as it stands it just make
things more confusing, so if there is no intention of cleaning it up in
the future, don't make the mess now.


-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From evan at 4-am.com  Mon Apr  8 21:01:31 2002
From: evan at 4-am.com (Evan Simpson)
Date: Tue, 09 Apr 2002 01:01:31 GMT
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org>
Message-ID: <3CB23D6F.2030006@4-am.com>

Greg Ewing wrote:
> Pyrex is more or less Python with C data types
> added. You can write functions which freely
> intermix operations on Python and C data, with
> all Python reference counting and error checking
> completely automated.

It's SPython! Yaay! (If you don't know what I'm on about, see 
http://groups.google.com/groups?selm=HeIo2.78%24u5.3610%40news-dal.corridex.com
and it's thread).

Thanks,

Evan @ 4-am



From daniel at dittmar.net  Wed Apr 10 17:42:46 2002
From: daniel at dittmar.net (Daniel Dittmar)
Date: Wed, 10 Apr 2002 23:42:46 +0200
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <3CB4B1D6.1030504@dittmar.net>

Joel Bender wrote:

> I would like to process the elements of a list that match a condition, 
> then remove them.  I'll use integers in my example, but my list items 
> are actually objects.

[...]

> I'd rather not scan the list twice, particularly if there's nothing to 
> do.  I'd rather not build another double-linked list.
> 
> Is there something like this?
> 
>     >>> for item in lst:
>     ...     if (item < 5):
>     ...         print item
>     ...         del lst[ xxxx ]
> 
> But the "trick" is figuring out what xxxx should be.  Using lst.remove() 
> would work if the internals used "is" and not "is equal to" and the 
> removal didn't modify the list where it would break the loop somehow.


Using .remove () is similar to scanning the list, so it doesn't buy you anything in terms of performance.


How about:

for i in xrange (len (lst) - 1, -1, -1):
     item = lst [i]
     if item < 5:
         print item
         del lst [i]

This makes it possible to delete items from the list as the access is 
back to front.

Daniel



From daniel at dittmar.net  Thu Apr  4 14:56:54 2002
From: daniel at dittmar.net (Daniel Dittmar)
Date: Thu, 04 Apr 2002 21:56:54 +0200
Subject: Downloading pages from a Windows Intranet
References: <8a27e309.0204040754.5073e0b2@posting.google.com>
Message-ID: <3CACB006.6000602@dittmar.net>

 > Trying to use urllib or urllib2 to get a page from our Intranet
 > (Windows) gives me an Access Denied error.  The funny thing is that on
 > some pages it works (?).  Any way, someone found a Perl module

MS has implemented an extension to HTTP where the client passes your 
Windows user to the Web server and accesses pages using this user. This 
works only when using the functions in WININET.DLL (I suppose).

 > from our intranet, even asp pages.  Is there an equivalent module or
 > functions that I could use from Python?  Or can anyone give me an
 > example of how to work around this?  Or should I just settle for a
 > Perl/Python solution?

1. You could pas your Windows user + password as an additional HTTP 
header, although this (basic authentication) might be disabled on the server

2. you could look into the Win32 extensions if the functions in 
WININET.DLL are made accessible from Python

3. You could make the web pages readable by 'Everyone'

4. You could indeed use the stuff ftom Perl

Daniel Dittmar



From ftobin at neverending.org  Mon Apr 29 19:40:02 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 29 Apr 2002 23:40:02 GMT
Subject: comparing str's with ints
In-Reply-To: <ucr7gdps6vuh0a@news.supernews.com>
References: <20020429115856.R1840-100000@palanthas.neverending.org>
 <aak5n9$efr$1@slb6.atl.mindspring.net> <ucr7gdps6vuh0a@news.supernews.com>
Message-ID: <20020429154445.V3184-100000@palanthas.neverending.org>

John Roth, on 2002-04-29, wrote:

> If you really, truely have to do Perl type comparisons between integers
> and strings, you might consider deriving your own integer and string
> types, and providing a custom __cmp__() method.

Oh, I certainly didn't want to do weakly-typed Perl-style comparisions.

It was a type error on my part to be comparing a str to an int (I had
expecting to be comparing an int to an int).  Since no exception was
thrown, I expected that 'foo > 0' (foo was accidentally a string instead
of an int) was at least doing something sane, but obviously it wasn't (it
was doing the 'arbitrary but consistent' comparison).

-- 
Frank Tobin		http://www.neverending.org/~ftobin/





From phil.edwards at mistral.co.uk  Mon Apr 22 12:04:01 2002
From: phil.edwards at mistral.co.uk (Phil Edwards)
Date: Mon, 22 Apr 2002 17:04:01 +0100
Subject: Python DNS Utilities
Message-ID: <3cc43b20@peer1.news.newnet.co.uk>

Hi Everybody:

I'm looking at putting together some kind of generic abstraction layer for 
handling DNS configuration files (adding entries to /etc/named.conf, 
adding/removing RR's to/from zone files, etc, etc) using Python.

Has anybody done any similar work in this area who might be willing to 
share ideas/code? I have some basic stuff working already (thanks to the 
guy who was doing my job before me) but this needs some work to improve the 
performance and extend the functionality in certain areas.

I've already looked (v. briefly!) at the pyDNS module, but this is 
concerned mainly with implementing a DNS client using Python, and I'm 
interested in the admin tasks involved in running a DNS server.

My particular issue at the moment is finding an efficient way to parse the 
torturous syntax which is permitted in the 'named.conf' file. I'm currently 
using the 'shlex' module to do this, but the code is a bit of a nightmare.

I'm even beginning to wonder if it might not be quicker to just put the 
whole DNS system into a Postgres database and maintain it from 
there...seems a shame to waste all the hard work my predecessor did though.

-- 

Regards

Phil Edwards
UNIX Systems Administrator
==========================================================================
Mistral Internet                                  http://www.mistral.co.uk
Tel: +44 (0)1273 747432                           Fax   +44 (0)1273 747452

Running Mandrake Linux release 8.1 (Vitamin) for i586 Kernel 2.4.8-34.1mdk
System uptime : 34 days, 22:58,  3 users,  load average: 1.09, 1.07, 1.08 
==========================================================================



From ruthenium at altavista.net  Fri Apr 12 06:09:48 2002
From: ruthenium at altavista.net (Ben)
Date: 12 Apr 2002 03:09:48 -0700
Subject: Efficient wxOGL redraws
Message-ID: <fe1e9dc5.0204120209.b20ad58@posting.google.com>

Hi all,

I hope one or other of the experts here can give me some advice...

In my wxPython (oh how I love it) app under Win32, I am wanting to use
little wxOGL wxShapes as slider bars, for defining start and stop
positions.  The idea is very simple - the user clicks on the bar,
drags it to the required position, and releases it.  The trick is that
the bar must have a fixed Y position.  That is, when the user drags
the bar, it only moves horizontally.


Now I've managed to get it working after a fashion, but it's VERY
inefficient.  My event handler includes the following:

def OnMovePre(self,dc,x,y,oldX,oldY,display):
		if y == oldY:
			return true			
		return false

def OnDragLeft(self,draw,x,y,keys, attachment):
		shape = self.GetShape()
		canvas = shape.GetCanvas() 
		dc = wxClientDC(canvas)
        	canvas.PrepareDC(dc) 
		self.GetShape().Erase(dc)
		self.GetShape().SetX(x)	
		self.GetShape().SetY(18)
		self.canvas.Redraw(dc)

Doing this causes horrid flicker, and steals lots of CPU cycles.

Can anyone think of a more elegant way of doing this using wxOGL?  I
could probably use GDI primitives, but I don't fancy all the schlep of
doing hit testing etc.

Thanks & regards,

Ben


From spinwing at inwind.it  Fri Apr  5 01:38:07 2002
From: spinwing at inwind.it (Stefano Vedovelli)
Date: Fri,  5 Apr 2002 08:38:07 +0200
Subject: how to detect if a file is complete
Message-ID: <GU313J$434708AE96A987644BA715C106B24D4F@inwind.it>

Hello

in my opinion is more simple. Just try to obtain a lock to the file. Under Unix you can use the fcntl module, under Windows the win32file extension. If the file is in use by another process, the lock will just fail. You catch the exception, wait for a while, and give it another try, maybe with a thread.

Hope this helps
Stefano





From aleax at aleax.it  Wed Apr  3 08:46:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 13:46:52 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com>
Message-ID: <gJDq8.55255$S52.1904212@news2.tin.it>

John Roth wrote:
        ...
> Maybe my post in response didn't propagate? As I said in that
> post, the root of her problem is a poorly chosen name for the
> function. If she paid attention to constructing the name properly,

She adopted the name *Guido* had given to the function right in
his pep (actually clarifying it a bit from 'isleap' in his PEP
to 'isLeapYear', more readable, in her post).


Alex



From niemeyer at conectiva.com  Tue Apr 30 17:42:31 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Tue, 30 Apr 2002 18:42:31 -0300
Subject: os.environ.copy()['COPY_HUH'] = "not really"
In-Reply-To: <20020430152022.I17488@unpythonic.net>
References: <20020430152022.I17488@unpythonic.net>
Message-ID: <20020430184231.C30307@ibook.distro.conectiva>

> Under Python2.1 and earlier (back to 1.5.2):
> >>> import os
> >>> os.environ.copy()['COPY_HUH'] = "not really"
> >>> print os.environ['COPY_HUH']
> not really
> 
> Under even 2.3a0 (CVS):
> >>> import os
> >>> os.environ.copy()['COPY_HUH'] = "not really"
> >>> os.system("echo $COPY_HUH")
> not really
> 0

Here's the problem:

>>> print os.environ.__class__
os._Environ
>>> copy = os.environ.copy()
>>> print copy.__class__
os._Environ

It's not a real dictionary, but an os._Environ class. Copying it
creates instances of the same class, and that class' behavior is
to insert the item in the environment. I belive this should be
avoided in this case, returning a true dictionary from copy(). I'll
send a patch to SourceForge, if you haven't done so yet.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From fperez528 at yahoo.com  Mon Apr 29 17:47:55 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 29 Apr 2002 15:47:55 -0600
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu>
Message-ID: <aakf70$pm9$1@peabody.colorado.edu>

Paul Hughett wrote:

> I am writing a Python extension in which it has become useful to
> define slice objects that run backwards (i.e. that have a negative
> increment) and want to use the existing convention, if any, for
> such objects.  But I don't recall ever seeing such a convention.
> 
> Hence my question: Is there an established meaning within Python for
> slice objects with a negative increment?  E.g. slice(1, 10, -2) ?

Yes, the standard Numeric already has well established conventions:

In [6]: z=arange(10)

In [7]: z
Out[7]: array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

In [8]: z[::-3]
Out[8]: array([9, 6, 3, 0])

Numeric is standard enough that I wouldn't even consider breaking its 
conventions (esp. since it's the only widely known Python module which fully 
implements slice objects).

hth,

f


From gimbo at ftech.net  Thu Apr 11 05:33:29 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 11 Apr 2002 10:33:29 +0100
Subject: initializing list with number of elements given by a variable
In-Reply-To: <3CB555E6.3080207@web.de>
References: <3CB555E6.3080207@web.de>
Message-ID: <20020411093329.GC1261@andy.tynant.ftech.net>

On Thu, Apr 11, 2002 at 11:22:46AM +0200, Oliver Pust wrote:

> So something like this:
> 
> >>> length_of_list = 6
> >>> list = ????????? # what to place here, is the question
> >>> print list
> >>> [0,0,0,0,0,0]    # the resulting list should look like this

Here ya go:

>>> l = [0] * 6   
>>> l
[0, 0, 0, 0, 0, 0]

Cheers,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From xiaotsing_79 at hotmail.com  Tue Apr 30 01:11:31 2002
From: xiaotsing_79 at hotmail.com (David)
Date: 29 Apr 2002 22:11:31 -0700
Subject: Python vs. Perl,  which is better to learn?
Message-ID: <df30afd4.0204292111.31e5914a@posting.google.com>

I have a lot of experience in ANSI C/C++, I want to learn a new
language and which it can help me in the work.
I just want to clear which(python or perl) is more comfort for me?


From gh_pythonlist at gmx.de  Sun Apr  7 16:54:43 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 22:54:43 +0200
Subject: I'm going crazy using cgi module. HELP PLEASE!!!!
In-Reply-To: <3CB0A940.63F504E1@ndh.net>
References: <9c5a942b.0204050836.58595d84@posting.google.com> <3CB0A940.63F504E1@ndh.net>
Message-ID: <20020407205443.GA29692@lilith.my-fqdn.de>

* Stefan Schwarzer <s.schwarzer at ndh.net> [2002-04-07 22:17 +0200]:
> I didn't investigate your code in detail but it may be that there's an
> defect in it which causes an exception. You may use
> 
> import sys
> sys.stderr = sys.stdout
> 
> at the top of your module to receive a traceback in the browser window
> or catch the exception with
> 
> try:
>     ...
> except:
>     # do something with sys.exc_info()

If you're using Python >= 2.2 (*), you may want to use the cgitb module
for debugging by putting this on top of your script:

import cgitb
cgitb.enable()

Be sure to have the Content-type set to text/html to display the nicely
formatted tracebacks.

Gerhard

(*) Else you'll have to get the cgitb module yourself.
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 4.2 ?C      Wind: 3.0 m/s




From iwk_removethis at xs4all_removethisalso.nl  Sat Apr 27 01:55:59 2002
From: iwk_removethis at xs4all_removethisalso.nl (iwk)
Date: Sat, 27 Apr 2002 07:55:59 +0200
Subject: Compiled socket.pyd available for Win32 Python 2.2.1?
Message-ID: <3CCA3D6F.7020506@xs4all_removethisalso.nl>

Hi there,

Does any benevolent soul have socket.pyd compiled for Win32 Python 2.2.1 
available for download? I need SSL support and the default Win install 
doesn't have ssl support compiled (why? US export regulations?) in. I 
don't have access to Visual C and lower versions of Python or Cygwin are 
unfortunaltely not an option (stalinist system management :-)

Regards,

Iwan



From martin at v.loewis.de  Mon Apr  1 18:31:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 01:31:49 +0200
Subject: Cannot find modules on SuSE Linux 7.3
References: <3CA8EE17.1050006@orange.removethis.net>
Message-ID: <m3lmc7vvve.fsf@mira.informatik.hu-berlin.de>

Colin Pinkney <colin.pinkney at orange.removethis.net> writes:

> I have $PYTHONHOME set to /usr/lib/python which is a link to
> /usr/lib/python2.1 and I've tried setting it to /usr/lib/python2.1 too.

Don't set PYTHONHOME. If you set it, it should be set to /usr.

Regards,
Martin


From googlegroups at sixtyten.org  Mon Apr 22 10:25:47 2002
From: googlegroups at sixtyten.org (Allan Crooks)
Date: 22 Apr 2002 07:25:47 -0700
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <mailman.1019403879.24573.python-list@python.org>
Message-ID: <304d20df.0204220625.6d1430ba@posting.google.com>

holger krekel <pyth at devel.trillke.net> wrote in message news:<mailman.1019403879.24573.python-list at python.org>...

> exec forversion("2.2",

<snip>

I had considered using exec as a way of getting round it, but it seems
like quite a hack to get what I want (wrapping code in a string to
prevent Python accessing it properly).

Mind you, considering what I want to do, I suppose hacking is really
the only feasible solution. :)

Doing things via exec would ultimately mean that I couldn't take
advantage of PyChecker...

Thanks for the suggestion though...
Allan.


From grante at visi.com  Tue Apr 30 12:17:55 2002
From: grante at visi.com (Grant Edwards)
Date: Tue, 30 Apr 2002 16:17:55 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2> <3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca> <mailman.1020126468.7915.python-list@python.org> <G8xz8.14199$xy4.5322@atlpnn01.usenetserver.com>
Message-ID: <Tszz8.59520$vm6.10213286@ruti.visi.com>

In article <G8xz8.14199$xy4.5322 at atlpnn01.usenetserver.com>, Steve Holden wrote:

> We got on even better after I left, and now I haven't spoken to
> any of them for three years we are the best of pals.

Now _that's_ one for the quote file...

-- 
Grant Edwards                   grante             Yow!  VICARIOUSLY
                                  at               experience some reason
                               visi.com            to LIVE!!


From eddieborder2001 at yahoo.co.uk  Fri Apr 12 14:48:50 2002
From: eddieborder2001 at yahoo.co.uk (Ed)
Date: 12 Apr 2002 11:48:50 -0700
Subject: wxImage problem
Message-ID: <d5534716.0204121048.3a666b3c@posting.google.com>

how can i put an image into a wxPanel?
i read on the net that the code is ...

wxImage_AddHandler(wxGIFHandler())
gif = wxImage('bitmaps/image.gif', wxBITMAP_TYPE_GIF).ConvertToBitmap()
wxStaticBitmap(panel, -1, bmp, wxPoint(10, pos),wxSize(bmp.GetWidth
(),bmp.GetHeight()))


From johahn at delta.telenordia.se  Sun Apr 14 05:10:57 2002
From: johahn at delta.telenordia.se (Johan Hahn)
Date: Sun, 14 Apr 2002 11:10:57 +0200
Subject: C API Q II : __main__.__dict__.keys() not same as dir()
Message-ID: <a9bl5l$2aq$2@green.tninet.se>

Hi

I'm writing a console for a game using the c api.

I figured that since the InteractiveConsole hangs after calling interact()
you have to have atleast 2 threads to use that, right? I thought the
documentation was a little vague on that.

Anyway, for now I use the InteractiveInterpreter and do the input buffering
in C and it works great. So great I decided to implement completion of
Python commands, just like when you hit tab in Unix.

For example if you type...
>>> import code
>>> code.I
...and press <tab> you should be presented the list:
['InteractiveInterpreter', 'InteractiveConsole']

That went fairly easy too (thanks to Python), but one problem remains.

When you import a module using python shell, it appears in
__main__.__dict__. In my program however, that is not the case. The module
is added to sys.modules and can be used without any problem but I need to
see it for my completion code to work.
Here is an example of how it looks in the console:

>>> import __main__, sys, os
>>> __main__.__dict__.keys()
['__builtins__', '__name__', '__doc__']
>>> dir()
['__builtins__', '__doc__', '__main__', '__name__', 'os', 'sys']


Finally, here is my question.
Where does the user imported modules (and all other objects too for that
matter) hide??




Here is part of my code (error checking and reference counting left out)

/* Python code, module b4 */
import code
class jhCon(code.InteractiveInterpreter):
    def complete(self, codeline):
        # First we have to parse the input around the dots
        import string
        parsedCode = string.split(codeline, '.')

        # BUG: dir(__main__) misses user-added objects
        import __main__
        availableCommands = dir(__main__)
        dict = __main__.__dict__

        # traverse the hierarchy
        word = parsedCode[0]
        count = 0
        while word in availableCommands:
            count += 1
            dict = dict[word]
            word = parsedCode[count]
            availableCommands = dir(dict)

        # Check if the incomplete word matches the beginning of
        # any of the availableCommands. If no match do nothing.
        possibleCommands = []
        for command in availableCommands:
            if command[:len(word)] == word:
                possibleCommands.append(command)

        # If we have only one possible command we 'complete' the
        # command by returning only the part of it that remains
        # to be typed in.
        # If there are more than one possible command we print them
        if len(possibleCommands) == 1:
            return possibleCommands[0][len(word):]
        print '\n',possibleCommands

iii = jhCon()
def complete(codeline):
    iii.complete(codeline)




/* C++ code */

Note: I let the constructor and destructor take care of initializing python
and the m_complete variable.

void init()
{
PyObject* moduleName = PyString_FromString("b4");
PyObject* module = PyImport_Import(moduleName);
PyObject* dict = PyModule_GetDict(module);
m_complete = PyDict_GetItemString(dict, "complete");
}


std::string complete(std::string s)
{
PyObject* rslt = PyObject_CallFunction(m_complete, "(s)", s.c_str());
if (!rslt) {           /* Python exception raised */
    PyErr_Print();
    return "";
}
else {
    if (rslt == Py_None)
        return "";
    std::string retVal(PyString_AsString(rslt));
    Py_XDECREF(rslt);
    return retVal;
}
}



mvh,
Johan






From bkc at Murkworks.com  Fri Apr 26 18:05:04 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Fri, 26 Apr 2002 18:05:04 -0400
Subject: Python on Oberon?
Message-ID: <3cc9c8d8$1_5@goliath.newsgroups.com>

Just wondering if anyone has thought about this possibility..

--
Novell DeveloperNet Sysop #5

_




-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From smithj at kryos.colorado.edu  Mon Apr  8 12:33:05 2002
From: smithj at kryos.colorado.edu (Jeff Smith)
Date: Mon, 08 Apr 2002 10:33:05 -0600
Subject: Newbie Problem Building 2.2 On SGI
Message-ID: <3CB1C641.6CF81ECE@kryos.colorado.edu>

I have installed both 2.2 and 2.2.1.c2 on an SGI O2 with IRIX6.5. I
cannot import the modules array, math, and struct (out of the limited
list of modules that I've tried to import). I have tried both the SGI C
compiler and the GNU C compiler.

Jeff







From glchapman at earthlink.net  Fri Apr 12 23:24:52 2002
From: glchapman at earthlink.net (Greg Chapman)
Date: Sat, 13 Apr 2002 03:24:52 GMT
Subject: metaclass
References: <mailman.1018633601.14441.python-list@python.org>
Message-ID: <r88fbu8f86i6hmave4g2v9dnedmm84g14k@4ax.com>

On Fri, 12 Apr 2002 12:45:21 -0500, "Mark McEahern" <mark at mceahern.com> wrote:

>I'm trying to translate the tracing metaclass provided here:
>
>  http://www.python.org/doc/essays/metaclasses/Trace.py
>
>to use the new __metaclass__ construct in Python 2.2, but I can't seem to
>get off the ground.
>
>For starters, suppose I simply want the metaclass to switch out one class
>with another.
>
>  # This should return an instance of Proxy instead of Switched.
>  # Eventually, Proxy will be able to trace calls to Switched
>  # before passing them on to Switched.
>  s = Switched()

I got a bit boggled looking at the Trace metaclass.  However, if I understand it
correctly, it may not be necessary in 2.2 to use a metaclass to trace all method
calls.   I believe the following class should work (though it will not log
syntactic operations which result in method calls -- e.g., getting an attribute
will not log calls to __getattribute__.  It also doesn't log calls to __init__).
Not tested (beyond the test routine in Trace.py):

class Traced(object):
    def __trace_call__(self, fp, fmt, *args):
        fp.write((fmt+'\n') % args)

    def __getattribute__(self, name):
        res = super(Traced, self).__getattribute__(name)
        if isinstance(res, types.MethodType):
            if self.__trace_output__ and (name != '__trace_call__'):
                fullname = self.__class__.__name__ + "." + name
                res = TracingWrapper(fullname, res, self)
        return res

class TracingWrapper(object):
    def __init__(self, name, func, inst):
        self.__name__ = name
        self.func = func
        self.inst = inst

    def __call__(self, *args, **kw):
        self.inst.__trace_call__(self.inst.__trace_output__,
                                 "calling %s, inst=%s, args=%s, kw=%s",
                                 self.__name__, self.inst, args, kw)
        try:
            rv = self.func(*args, **kw)
        except:
            t, v, tb = sys.exc_info()
            self.inst.__trace_call__(self.inst.__trace_output__,
                                     "returning from %s with exception %s: %s",
                                     self.__name__, t, v)
            raise t, v, tb
        else:
            self.inst.__trace_call__(self.inst.__trace_output__,
                                     "returning from %s with value %s",
                                     self.__name__, rv)
            return rv


class C(Traced):
    def __init__(self, x=0): self.x = x
    def m1(self, x): self.x = x
    def m2(self, y): return self.x + y
    __trace_output__ = sys.stdout

(etc.)
---
Greg Chapman


From fredp at mygale.org.nospam  Mon Apr  8 03:39:06 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 8 Apr 2002 07:39:06 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it> <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20> <MMfr8.62526$S52.2173482@news2.tin.it> <mailman.1018026973.8164.python-list@python.org>
Message-ID: <Xns91EA629E151BDPaCmAnRDLM@161.48.128.20>

Gregor Hoffleit <gregor at hoffleit.de> said :

> It definitely sells in Germany: 699,- ? for the SL-5500G. There was an
> advertisement yesterday from our local discounter in the newspaper.
> Have a look at www.promarkt.de to confirm this; you should be able to
> navigate to it using the pictures ;-)

Hmm. So maybe the article I read was either misinformed or very unclear. 
Always glad to be wrong in such cases :-)

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


From geoff at gerrietts.net  Fri Apr 19 15:28:47 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 19 Apr 2002 12:28:47 -0700
Subject: why no 'length' method in sequences ?
In-Reply-To: <just-2DF829.09350419042002@news1.xs4all.nl>
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <E4Iv8.2972$8D3.78501@news1.tin.it> <just-2DF829.09350419042002@news1.xs4all.nl>
Message-ID: <20020419192847.GD13249@isis.gerrietts.net>

Quoting Just van Rossum (just at xs4all.nl):
> I appreciate and agree with what you write, and I like how Python does 
> things, too, BUT I can totally see why people find it strange that to 
> get the length of an object you write
> 
>     len(o)
> 
> and to get (eg.) the keys from a dictionary you do
> 
>     o.keys()

I'm interested in this thread only because of the special case of the
len() function -- I still find myself, even after a couple years of
using it, occassionally writing object.len() instead of len(object).

I think it's because in my head, most of the functions in Python
operate /on/ objects, while I think of a length as more of an
attribute /of/ an object.

Maybe that distinction is artificial, or maybe it's less about Python
and more about how I think about functions vs. methods. It's certainly
a more OO-centric way of thinking about the two.

I don't know that this situation calls for correction, or whether it
calls for education, or whether it's just a strange corner-case that
might bear consideration when thinking about the broader language
design issues.

It's just interesting to me, that others are expressing a similar
sentiment.

Thanks,
--G.

-- 
Geoff Gerrietts             "There is no fate that cannot be 
<geoff at gerrietts net>     surmounted by scorn." --Albert Camus




From eppstein at ics.uci.edu  Mon Apr 15 11:01:21 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 08:01:21 -0700
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk>
Message-ID: <eppstein-7165E1.08012115042002@news.service.uci.edu>

In article 
<20020415144153325+0100 at pc-62-30-160-65-hw.blueyonder.co.uk>,
 Alwyn <alwyn at alwyn.demon.co.uk> wrote:

> >MacPython does also have a facility for creating a standalone 
> >application, but these are much bigger (several megabytes?) and I 
> >haven't been able to get this to work.
> 
> But if the target platform is Mac OS X, you can run Python scripts from
> the command line, as you would on any other Unix-type system. 

If you've installed a *different* version of Python (Unix Python, not 
MacPython), you can do this.  It doesn't come installed by default, and 
I have no idea whether it can do GUI stuff (unless you also install 
X-windows, or something).

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From ssiebinga at fa.knaw.nl  Mon Apr 15 16:02:19 2002
From: ssiebinga at fa.knaw.nl (sjoerd siebinga)
Date: 15 Apr 2002 13:02:19 -0700
Subject: problem negative lookahead assertion
Message-ID: <4b926dbb.0204151202.6d709633@posting.google.com>

I am currently working on indexing an Old Frisian etymological
dictionary typeset in LaTeX2e. I use a Python script to insert the
index (\index{}) marker for 45 languages. My problem is the following.
I have marked up all the words which have a fixed structure - language
abbreviation (MDu.) \emph{middle dutch  word}. 


mdu = re.compile(r'MDu\.\s\\emph\{(.*?)\}([\s,\W])')
data = mdu.sub('MDu. \emph{\\1}\\2 \index{mdu~\\1}', data)


Now sometimes these words have more that one variant such as shown
below in the sample text excerpt or are mentioned in running text. I
tried to round up these forms with the following regular expression.


emph = re.compile(r'\s.\\emph\{(.*?)\}([\s,\W])(?!\\index)')
data = emph.sub('\emph{\\1}\\2 \index{unl~\\1}', data)


When this was applied to the text below all the \emph{} phrases are
replaced instead of only those not followed by the \index{} commando.

<sample text>

\begin{germdata} ON \emph{va\th a} \index{on~va\th a}`wade, rush, walk
through', OE \emph{wadan}, \index{oe~wadan} OHG, MHG \emph{watan}
\index{mhg~watan}`wade, stride', MLG \emph{w\=aden},
\index{mlg~w\=aden} MDu. \emph{waden}, \index{mdu~waden} \emph{waeyen}
`wade, go' \end{germdata}

</sample text>


I have tried all the variants I could think of on the emph regular
expression but they all had the same outcome. I also used the
redemo.py script that comes with the python distribution to identify
the problem.

Could somebody help me out.

regards sjoerd 

PS I am using python 2.2.1 on a suse linux 7.3 box.


From jgoerzen at complete.org  Sat Apr 20 16:46:26 2002
From: jgoerzen at complete.org (John Goerzen)
Date: 20 Apr 2002 15:46:26 -0500
Subject: [ANN] dictclient
Message-ID: <87adrycd3x.fsf@complete.org>

This is an announcement for dictclient.  dictclient is a Python
library for communicating with DICT (RFC2229) servers.  It provides
access to the full command suite in a nice OO interface.

Dictclient's home is:

gopher://quux.org/1/devel/dictclient

Or via HTTP:

http://quux.org/devel/dictclient

Docs available at
http://quux.org/devel/gopher/Downloads/LiveSource/dictclient/dictclient.html

My homepage:
http://www.complete.org/

Thanks!

-- John


From huaiyu at gauss.almadan.ibm.com  Mon Apr 15 14:14:06 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Mon, 15 Apr 2002 18:14:06 +0000 (UTC)
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <roy-0BE2A8.19181913042002@news1.panix.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de> <Prau8.17347$vF6.558738@news2.tin.it>
Message-ID: <slrnabm63e.nkd.huaiyu@gauss.almadan.ibm.com>

Alex Martelli <aleax at aleax.it> wrote:
>Martin v. Loewis wrote:
>
>> Indeed, and so it is desirable to have infix shorthand notations for
>> the frequent operations. However, as the number of infix operators
>> grows, the convenience loses: you now have to start remembering line
>> noise.
>
>Very well put.  Indeed, given the small amount of bit-twiddling in
>typical Python usage, I think & ~ | ^ are already overboard -- they
>make a lot of sense in C, a very low-level language that is used a
>LOT for bit-twiddling, but, in Python, named built-in functions might
>have been more appropriate in hindsight (fewer precedence levels to
>remember) -- if not, even, functions in a separate module.
>
>A simple histogram of non-alphameric tokens in Lib/*.py only partially
>conforts this thesis:
[snip statistics show small usage of & ~ | ^]

Do you plan to revive the bitwise module you wrote sometime ago?

Huaiyu


From jvanasco at hotmail.com  Tue Apr 23 02:38:39 2002
From: jvanasco at hotmail.com (Jonathan Vanasco)
Date: Tue, 23 Apr 2002 06:38:39 +0000
Subject: mass mailing
Message-ID: <F226goNrRJBJ1qCVxqY00006c76@hotmail.com>

>Is MailMan an option?
yes and no.. i've been pilfering code from it to get things working -- but 
having a mailman interface wont work.. everything needs to be custom and 
modding mailmain is too much a chore.  there's also a lot of mailman 
features that wont be supported, and i'd have to remove.

>Since smtplib functions at a level of individual messages, you'd need to 
>write your own bulk-handling utility that uses smtplib internally.?

yeah.. thats what i've figured.. i've been playing around with it lately, 
trying to move some of perl's Mail::Bulkmail features into python, based on 
the mailman code...  i think what is going to end up happening, is sorting a 
list of emails by domain name, then processing the domains into chunks of 
100 or less, and feeding those, via smtplib to postfix as a , delimited list 
(or so the rfc specks say will work)

ideally, that will mean 1 envelope/message per chunk -- so my bandwidth and 
qfiles could go down from anywhere between 100 and 0 %

the extra coding is worth it though -- checking through the mailman 
archives, someone from gnu.org had major problems handling 50,000 emails.  
granted -- this project may never receive more than 100 emails.  but should 
it go higher, i want to be prepared.  and not recoding something under a 
shitty support rate vs an inflated development one :)

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





From bbollenbach at shaw.ca  Fri Apr 12 20:02:14 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Sat, 13 Apr 2002 00:02:14 GMT
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <aAKt8.22045$ir6.979928@news1.calgary.shaw.ca>

In article <slrnabeqkg.mfc.huaiyu at gauss.almadan.ibm.com>, Huaiyu Zhu wrote:
> Brad Bollenbach <bbollenbach at shaw.ca> wrote:
>>
>>This is a semantic debate. I know that Stallman chose the term "Free
>>Software" because he felt it was the best (and most concise) way to
>>express what he was thinking of.
> 
> That's right.  A large portion of OSS is FS.  But another large 
> portion is  not.  FS is a proper subset of OSS.

Fine, I guess I'll have to cite Stallman directly to back up what I
said. :) From http://www.fsf.org/philosophy/categories.html#OpenSource:

<begin gnu>
Open Source software
====================

 The term ``open source'' software is used by some people to mean more
 or less the same thing as free software. We prefer the term ``free
 software''; follow that link to see the reasons.
<end gnu>


--
Brad


From siegfried.gonzi at kfunigraz.ac.at  Sat Apr 20 09:32:02 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 20 Apr 2002 14:32:02 +0100
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org>
Message-ID: <a9rn4i$mm4$1@newsreader1.netway.at>

"James J. Besemer" <jb at cascade-sys.com> schrieb im Newsbeitrag
news:mailman.1019299357.11513.python-list at python.org...

> Seems to me the more prominent genetic similarities are the overall
semantics: late binding, garbage collection, lists and >sequences as a way
of life.  Also, the fact that functions are values that may be free rebound
to other names.  The >semantics of and/or are rarely
> encountered outside of lisp and derivatives.  MAP, APPLY, and especially
LAMBDA also are pretty unique to Lisp.  >This is VERY like Lisp and very
UNLIKE "Algol like" languages.

It is correct that they are not Algol-like; but map, apply,lambda are
typical for functional languages like Clean, Haskell,...

What Python --from my pragmatic point of view-- has in common with Lisp is
the nice feature that you can put in your return  list different objects,

e.g. [[1,2,3],["334.233,"Ein bloeder Oesterreicher",777],array]

In Clean for example you can not easily construct that, except you write
some form of a struct (or special tuple constructs). Clean is a strict
(strongly typed) language and this behavior is visible in every situation.

And it is possible in Python (sorry I know only the idle) to build your
functions in the command-line (every Common Lisp environment provides you
this). I often have heard that CLOS (Common Lisp Object System) lets you
change your class objects on the fly, I do not have any experience with that
and I am not sure how this relates to Python (my Python object oriented
programming experience is zero).

Python shows very often its C heritage, whereas Common Lisp is devoid of
that impression. From this point of view Lisp is much different from Python;
Lisp really went its own way.

S. Gonzi







From pyth at devel.trillke.net  Tue Apr 30 16:28:22 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 22:28:22 +0200
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <200204301613.44775.bhan@andrew.cmu.edu>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <200204292311.57363.bhan@andrew.cmu.edu> <nlalaa.ko2.ln@127.0.0.1> <200204301613.44775.bhan@andrew.cmu.edu>
Message-ID: <20020430202822.GI16217@devel.trillke>

On Tue, Apr 30, 2002 at 04:13:44PM -0400, Benjamin Han wrote:
> On Tuesday 30 April 2002 01:39 am, you wrote:
> > Benjamin Han wrote:
> >
> > > How much is an effort in
> > > converting Python into C++? Anyone would like to comment on some of
> > > these questions, or point to some article, FAQ (esp. helpful from a
> > > C++ programmer's point of view)?
> >
> > Well, let me give you a little heads up on the differences you might
> > see.  Many C++ programmers are brought up to believe that C++ is
> > synonymous with OOP.  You seem to tend that way, with your idea that
> > implicit typecasting is an OO feature.  (If not, I apologize I seem
> > facetious.)
> >
> 
> I guess you're right - I did learn a lot from so many immensely useful 
> replies. Time to shed my C++ skin and learn something new. Thank you guys!

Funny to notice, that at least ANSI-C++ is *newer* than python as far
as i know.

> (wouldn't it be nice if someone writes up a comparison list between Python and 
> the other major OO languages, e.g., SmallTalk, C++, Java, etc.?)

you might want to look at:

http://www.python.org/doc/Comparisons.html

have fun,

    holger




From mertz at gnosis.cx  Tue Apr  9 17:29:36 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Tue, 09 Apr 2002 17:29:36 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <slrnab2inr.8qu.mlh@vier.idi.ntnu.no> <Lpgs8kKkXgpE092yn@gnosis.cx>
 <mailman.1018301543.25726.python-list@python.org>
 <b5ys8.62201$pT1.1789842@news1.tin.it>
Message-ID: <A11s8kKkXUPH092yn@gnosis.cx>         

|Lulu of the Lotus-Eaters wrote:
|major = len([1 for f in fs if f(o)]))

Alex Martelli <aleax at aleax.it> wrote:
|I personally prefer:
|major = len(filter(None, [f(o) for f in fs]))

Why?  I had thought of basically that one.  But the short "if f(o)"
clause seemed to read better than the longer "filter(None..)" construct.
I don't really care one way or another about mixing filter() with
listcomps, but the few extra keystrokes in Alex' don't seem to add to
clarity.

Another option is to squeeze a reduce(operator.add...) into there.  But
practically, len() seems to do everything that needs doing.

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------





From mhammond at skippinet.com.au  Mon Apr 15 03:41:16 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Apr 2002 07:41:16 GMT
Subject: problem running win32com.client application from commandline
References: <3hrjbu4qv9b9afjhr2nhnef025350k4lpd@4ax.com>
Message-ID: <3CBA8463.30703@skippinet.com.au>

Tino Lange wrote:
> Hi All!
> 
> I just discovered a problem with a win32com.client based application.
> 
> If I initialize the object 
> acheck = win32com.client.Dispatch(com_name)
> from pythonwin it's no problem at all. The application runs and does
> everything I want it to do for me :-)
> 
> But if I use it from the commandline (either python.exe or
> pythonw.exe) then the above initialization of the COM-Object fails. A
> MessageBox appears (from the COM-Object, Method
> "UserControl_Initialize") telling that there's not enough memory to
> initialize.
> 
> Unfortunately I just have the OCX - I cannot look in the COM-Object
> :-(
> 
> Any idea? What's different, when I run the script via PythonWin?
> Is there a possibility to run commandline-scripts?
> 

Pythonwin contains the code necessary to host true OCX controls.  You 
can't do this in python.exe or wpython.exe.  You may still be able to 
get what you need by using "pythonwin.exe /nodde /run foo.py"

Mark.



From kalle at gnupung.net  Wed Apr 10 19:51:37 2002
From: kalle at gnupung.net (Kalle Svensson)
Date: Thu, 11 Apr 2002 01:51:37 +0200
Subject: Easter Eggs
In-Reply-To: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <20020410235137.GI1577@proton.lysator.liu.se>

[Robin Becker]
> In article <mailman.1018391839.6535.python-list at python.org>, Kalle
> Svensson <kalle at gnupung.net> writes
> >[Lulu of the Lotus-Eaters]
> >> Who can tell me what's in there.  Someone on c.l.py recently let drop
> >> the hint about finding the Zen of Python at the interactive prompt in
> >> 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
> >> 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
> >> magic command was.  Can someone remind me, please?
> >
> >"import this", IIRC.
> ...
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers. 
[snip rant]

Would you mind explaining why you feel this way, instead of calling
names?

Med utm?rkt h?gaktning,
  Kalle
-- 
Kalle Svensson, http://www.juckapan.org/~kalle/
Student, root and saint in the Church of Emacs.




From grante at visi.com  Mon Apr 29 14:28:49 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 29 Apr 2002 18:28:49 GMT
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1>
Message-ID: <Bhgz8.58769$vm6.9975333@ruti.visi.com>

In article <Xns91FF9515A57Eduncanrcpcouk at 127.0.0.1>, Duncan Booth wrote:
> holger krekel <pyth at devel.trillke.net> wrote in 
> news:mailman.1020082610.7467.python-list at python.org:
> 
>> Btw, what is the best/shortest way to express 'filter' and 'reduce' with 
>> list comprehensions?
> 
>  filter(fn, lst)
> becomes:
>  [ v for v in lst if fn(v) ]

I find the former far more readable.

>  filter(None, lst)
> becomes:
>  [ v for v in lst if v ]

Same here.

> If the second argument to filter is a string or tuple then you have to 
> convert the resulting list back to the appropriate type:
> 
> e.g.
>  filter(fn, aString)
>  str.join('', [ v for v in aString if fn(v) ])

IMO, claiming that the latter is not less clear is pretty bogus.

> reduce doesn't return a list, so a list comprehension is not an
> appropriate replacement.
> 
>  reduce(fn, seq, initial)
> becomes:
>  res = initial
>  for v in seq: res = fn(res, v)

I find the "reduce" spelling simpler and easier to both read
and write.

> All of these replacements are big winners if they remove a use of the 'l' 
> word.

If getting rid of "lambda" is the goal, then why not state that
openly and propose elimination of "lambda".

> I don't think any of them lose in clarity even if there are no 'l's 
> around.

You've got to be kidding -- especially looking at the "reduce"
and "filter a string" examples.

If you want to move map/filter/reduce into a module, fine.  I
think it's too much breakage for too little gain, but whatever...

However, claiming that they aren't useful because the above
"replacements" are just as clear is just silly.

-- 
Grant Edwards                   grante             Yow!  NOT fucking!! Also
                                  at               not a PACKAGE of LOOSE-LEAF
                               visi.com            PAPER!!


From ark at research.att.com  Wed Apr  3 23:09:59 2002
From: ark at research.att.com (Andrew Koenig)
Date: 03 Apr 2002 23:09:59 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org>
References: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org>
Message-ID: <yu99pu1gdrzc.fsf@europa.research.att.com>

>> No.  That would break backwards compatibility.  False==0, and True==1;
>> everything else follows from that.  (But False is not 0, and True is
>> not 1!)

Ping> This is a strange distinction to make -- one that is not made in
Ping> any other programming language i have ever encountered.

I would be very surprised if none of the Lisp family of languages
has similar anomalies in the relationship between eq and equal.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark




From jochen at jochen-kuepper.de  Mon Apr 29 22:26:55 2002
From: jochen at jochen-kuepper.de (Jochen =?iso-8859-1?q?K=FCpper?=)
Date: 29 Apr 2002 22:26:55 -0400
Subject: re HOWTO
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <m3y9f7hdc7.fsf@box.home.de> <93d52e82.0204290637.60a65a9e@posting.google.com>
Message-ID: <m3g01dgbv4.fsf_-_@box.home.de>

On 29 Apr 2002 07:37:24 -0700 Terry Byrne wrote:

>> Btw: An index would serve the HOWTO well:)) It is pretty hard to find
>> information in there in a non-linear way (that is, without reading all
>> of it over and over).  It's a nice document, but it is a little
>> unorganized.

Terry> I'm nobody in the Python world, wish I could help you with the
Terry> documentation! I agree with you that it's a good reference, but I too
Terry> didn't find it helpful as a tutorial.

Well, I thought the re HOWTO is a nice tutorial, but needs an index to
be a truely useful reference, as the content is soemwhat scarmbled and
the toc doesn't get one (at least not me) to the information.

[...Python books...]

Well, I actually liked Lutz: /Programming Python/.  But then, I start
missing the stuff I am looking for too often now:)) It is a great
introduction, though!


I think I misunderstood your original posting:( I thought you were
using a "pass" option in the regexp compile or match/search calls, and
of course I never heard of it, as it apparently doesn't exit:O

Terry> pass lets me treat the long list of lines like a C switch or a
Terry> Pascal case:

Well, a switch-case construct is modeled by if...elif...else...endif
in python, AFAICT.

Terry> once a certain error code is found, pass skips to the next line
Terry> in the list without even testing for the other codes.

And here others have pointed out already that you wanna use continue.
See sections 4.4 and 4.5 of the Python turorial.

Greetings,
Jochen
-- 
Einigkeit und Recht und Freiheit                http://www.Jochen-Kuepper.de
    Libert?, ?galit?, Fraternit?                GnuPG key: 44BCCD8E
        Sex, drugs and rock-n-roll


From db3l at fitlinxx.com  Mon Apr 15 14:15:35 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 15 Apr 2002 14:15:35 -0400
Subject: Python 2.2 unwise.exe
References: <mailman.1018727067.4980.python-list@python.org>
Message-ID: <upu104ym0.fsf@ctwd0143.fitlinxx.com>

Tim Peters <tim.one at comcast.net> writes:

> [Robin Becker]
> > Anyone else had a problem un-installing 2.2 for win32? When I tried to
> > un-install 2.2 prior to installing 2.2.1 I got an error saying it had
> > already been un-installed.
> >
> > Tried the link in start menu and it said no unwise.exe was there.
> 
> Yes, I've seen this happen several times, but only under Win9X, not Win2K.
> Which flavor of Windows are you using?   So far you're the only other one to
> report it.  I can't account for it:  the installer definitely creates
> UNWISE.EXE in the right place (the root of your Python tree), but sometimes
> "it goes missing" later.  Exactly when and why remains a mystery to me.  It
> may be related to unplanned <ahem> system crashes.

It happened to me under WinNT 4.0 SP4.  As far as I can recall I
haven't had any abnormal shutdown of this system since installing 2.2.

I ended up just copying the unwise.exe from a Python 2.1 directory and
it un-installed fine.  UnWisePW32.exe was present - just unwise.exe
was missing.

It seems to be in place ok at the moment from the 2.2.1 install.  But
I can't say that I ever checked or noticed if it was properly present
after first installing 2.2.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From gimbo at ftech.net  Tue Apr 16 07:12:33 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Tue, 16 Apr 2002 12:12:33 +0100
Subject: Namespaces
In-Reply-To: <3cbc0275@news.comindico.com.au>
References: <3cbc0275@news.comindico.com.au>
Message-ID: <20020416111233.GF15240@andy.tynant.ftech.net>

On Tue, Apr 16, 2002 at 08:52:29PM +1000, Graeme Matthew wrote:

> import Template
> template = Template.Template()
> 
> I cannot get it to instantiate like
> 
> template = Template()

template = Template.Template() is the correct way to do it.

You could do what you want with:

from Template import *
template = Template()

but such "namespace pollution" is generally seen as a bad thing (see
http://py-howto.sourceforge.net/doanddont/doanddont.html for more on
this).  Slightly less harmful would be:

from Template import Template
template = Template()

but really you're doing the right thing anyway - don't knock it.  :-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From DavidA at ActiveState.com  Wed Apr  3 17:47:08 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Wed, 03 Apr 2002 14:47:08 -0800
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com>
Message-ID: <3CAB866C.4DF51F72@activestate.com>

Guido van Rossum wrote:

> Despite the negative feedback, I've decided to accept the PEP.  The
> most important perceived problem is that newbies tend to write
> 
>     if x == True: ...
> 
> where they should write
> 
>     if x: ...
> 
> I believe this problem should be solved by education (the Zen master
> hits the student on the head with a stick, and the student gets
> enlightenment) rather than by holding back what I believe will be a
> useful feature.  

Would it not be "relatively" easy to add a compiler-time warning for
most uses of this dangerous idiom?

At the very least, the boolean type could do the equivalent of:

  def __eq__(self, other):
     if __debug__ and other not in (0,1,True,False):
         warnings.warn("Testing of equality is best done by asking
objects whether they're true, not comparing with a boolean",
RunTimeWarning)
     return int.__eq__(self, other)

or something similar?

--david




From imbosol at vt.edu  Tue Apr 30 19:02:53 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 30 Apr 2002 19:02:53 -0400
Subject: Python vs. Perl, which is better to learn?
References: <mailman.1020147892.13578.python-list@python.org>    <aamjat$qku$1@panix1.panix.com> <aamm5k$kv3$1@bubba.NMSU.Edu>
Message-ID: <tq7naa.ik1.ln@127.0.0.1>

"Bob Melson" wrote:
> In article <aamjat$qku$1 at panix1.panix.com>,
>        aahz at pythoncraft.com (Aahz) writes:
>> In article <mailman.1020147892.13578.python-list at python.org>,
>> Andrew McNamara  <andrewm at object-craft.com.au> wrote:
>>>
>>>I'd go as far as to say the
>>>sendmail.cf language is simpler, easier to learn and maintain than perl.
>> 
>> .sigfile?
> 
> Look at it this way -- you pick the best language tool to do the job at
> hand.  If python is the better language for that job, you use python; if
> it's perl, then you use perl; if it's some other language, you use _that_
> language.  If you're asked to build a house, you certainly don't confine
> yourself to a single tool -- you use the tools appropriate to the job you
> have to do; why should it be different with scripting languages?


So you agree that he shouldn't bother with Perl.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From andrewm at object-craft.com.au  Tue Apr 30 22:34:10 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 01 May 2002 12:34:10 +1000
Subject: Python vs. Perl, which is better to learn? 
In-Reply-To: Your message of "Wed, 01 May 2002 01:53:09 GMT."
             <9UHz8.23$s55.5@newsread2.prod.itd.earthlink.net> 
Message-ID: <20020501023410.43D0638F50@coffee.object-craft.com.au>

>I am also new to python.  I think python will be my replacement of perl.
>However, if I need to do very heavy regular expression, I still prefer perl.
>I just don't feel comfortable with the python's re syntax.

Well, the RE syntax itself should be (almost?) identical - I guess
you're refering to way the re module is used (that it's not part of
Python's syntax)?

I'd suggest you not let this discourage you - while I initially avoided
using RE's in Python, I quickly came to love the rest of the language,
and then found that not having RE's in the language's syntax was not the
problem I thought it was (you rapidly discover better ways of working -
superior, I would argue, to working in perl).

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/




From logiplexsoftware at earthlink.net  Wed Apr 10 15:13:38 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 12:13:38 -0700
Subject: Python FREE IDE for windows
In-Reply-To: <005b01c1e0c3$707dc620$01c8a8c0@raphaelofizs8j>
References: <005b01c1e0c3$707dc620$01c8a8c0@raphaelofizs8j>
Message-ID: <20020410121338.7ae5cc64.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 16:10:53 -0300
Raphael Ribeiro wrote:

> does anyone know?

Only Google does, but apparently the secret is safe there.

http://www-106.ibm.com/developerworks/linux/library/l-pide/

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From aleax at aleax.it  Mon Apr 29 13:31:39 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 17:31:39 GMT
Subject: Operator overloading is much slower than a method call?
References: <c2f51c49.0204290851.2061078a@posting.google.com>
Message-ID: <%rfz8.32177$8D3.903364@news1.tin.it>

Scott Harris wrote:

> I did some simple timing experiments last night that seem to indicate that
> operator overloading is about 35% slower than a method call on an object.

Yes, in the current implementation, it IS a bit slower

> where the overloaded + operator simply calls the add method.

So you're paying for TWO method calls rather than ONE, that's one factor.
But there's more.  Let's measure.  Script ol.py:

class Overloader:
    def add(self, other):
        return self.__class__()
    def ind(self, other):
        return self.add(other)
    def __add__(self, other):
        return self.add(other)

class NSOverloader(object, Overloader):
    pass

a = b = Overloader()
nsa = nsb = NSOverloader()

import time
r = range(100000)

start=time.clock()
for i in r: pass
stend=time.clock()
over = stend-start

start=time.clock()
for i in r: c = a + b
stend=time.clock()
oper = stend-start

start=time.clock()
for i in r: c = a.add(b)
stend=time.clock()
meth = stend-start

start=time.clock()
for i in r: c = a.ind(b)
stend=time.clock()
indi = stend-start

start=time.clock()
for i in r: c = nsa + b
stend=time.clock()
nsoper = stend-start

start=time.clock()
for i in r: c = nsa.add(b)
stend=time.clock()
nsmeth = stend-start

start=time.clock()
for i in r: c = nsa.ind(b)
stend=time.clock()
nsindi = stend-start


print over, oper, meth, indi, nsoper, nsmeth, nsindi



Results on my box:

[alex at lancelot alex]$ python -O ol.py
0.02 1.05 0.45 0.54 0.81 0.61 0.7
[alex at lancelot alex]$ python -O ol.py
0.03 1.04 0.45 0.54 0.81 0.61 0.7

So: loop overhead negligible for 100,000 operations.

Old-style classes: 4.5 microseconds per operation
for the direct call to .add; 5.4 for the indirect
one; 10.4 for the overloaded-indirection -- a net
cost of 5 microseconds/operation for the overload
itself, plus of course the normal costs of the
indirect-call.

New-style classes: 6.1 for direct call, 7.0 for
indirection, 8.1 for overloaded-indirection -- a
net cost of 1.1 microseconds/operation for the
overload itself, plus etc.

One to five extra microseconds per operation ain't
hay, if you are indeed performing millions of
operations each of which does nothing or very little
(as you talk about vector operations in your program,
though, I wonder if each operation couldn't perform
enough work to make this overhead negligible).


Alex



From mgilfix at eecs.tufts.edu  Wed Apr 17 23:09:12 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 23:09:12 -0400
Subject: list_a-list_b
In-Reply-To: <a9klfs$n5t$0@216.39.172.122>; from bokr@oz.net on Wed, Apr 17, 2002 at 08:21:16PM +0000
References: <mailman.1019022993.8455.python-list@python.org> <Xns91F32263AA919RASXnewsDFE1@130.133.1.4> <a9klfs$n5t$0@216.39.172.122>
Message-ID: <20020417230912.B924@eecs.tufts.edu>

  I think I'd be in favor of allowing some other grammer to extend
beyong a single line (not just () and [ ]). Something like = and ','
make sense to extend beyond one line and the extra indent is pretty
clear. I really like the idea, especially when assigning to object
instances with long names like:

	self.long_object.something[ index + 1 % something ] =
		blah

  It sure beats '\'.

              -- Mike

On Wed, Apr 17 @ 20:21, Bengt Richter wrote:
> If you want the form but not the looong one-liner:
> 
>     self.data[x][y].variable = (
>         self.long_function_name(
>             lots,
>             of,
>             arguments 
>         )
>     )
> 
> Hm, this makes me wonder if '=' could introduce an optionally indented expression block,
> using ident/dedent where the above is using = ( .. )

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From skip at pobox.com  Wed Apr 24 15:10:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 24 Apr 2002 14:10:18 -0500
Subject: Why don't strings share data in Python?
In-Reply-To: <wzd6wp480e.fsf@cs.uu.nl>
References: <87d6x09trz.fsf@mathdogs.com>
        <mailman.1018931022.12502.python-list@python.org>
        <wzd6wp480e.fsf@cs.uu.nl>
Message-ID: <15559.794.483924.745059@12-248-41-177.client.attbi.com>

    CG> If you mean, why don't substrings share space with parent strings,
    CG> hmm, that's an interesting question.  I can see that a string object
    CG> implemented something like this struct:

    CG>   struct ExampleString {
    CG>     int len;
    CG>     char *buf;
    CG>   }

    CG> would allow that, and it sounds like an interesting optimization.

It probably wouldn't turn out to be an optimization at all.  It would slow
down string creation in many situations and make passing strings between
Python's C implementation and external C functions much slower.  Suppose I
had

    s = "abcdefg"
    t = s[2:5]

In theory, with a data structure like above, t->buf could just refer to
&(s->buf[2]).  Unfortunately, t->buf would not be NULL-terminated.  To pass
it to most C lib functions you'd first have to strncpy() it.  The current
implementation, while it wastes some storage, exchanges string data with
external functions without copying.

String object creation uses a variable length object header.  This works
because strings are immutable.  The bytes for the raw string are allocated
along with the object header itself.  This saves a malloc() call for each
string allocation, which is quite noticeable because Python allocates so
many string objects.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From mlh at vier.idi.ntnu.no  Sat Apr 13 06:31:56 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 13 Apr 2002 10:31:56 +0000 (UTC)
Subject: Thread-safe sockets?
Message-ID: <slrnabg28s.g9m.mlh@vier.idi.ntnu.no>

I'm working with a ThreadedTCPServer and the following question came
up: Is it safe for two different threads to use the
StreamRequestHandler's rfile and wfile simultaneously? (I.e. one
thread reads and another writes -- no simultaneous reading or
writing.) These file-like objects are created with the socket.makefile
method.

Since the two operations seem so unrelated, it would also seem that
this is safe -- but I'm worried that there may be (perhaps
paltform-dependent) thread-unsafe interference between the two at some
lower level. Does anyone know?

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From ponderor at lycos.com  Mon Apr 22 20:26:32 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 22 Apr 2002 19:26:32 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr 22)
Message-ID: <E6CCBFF070E5A288.3FEE8A1DFA481C1A.F6F8164C20E5B097@lp.airnews.net>

QOTW:  Paul Rubin:  "Comparing Python with Perl, generally I find
Python better designed but its implementation more likely to take
short cuts."
       Tim Peters:  "Heh -- I take it you haven't looked much at
either implementation."
		  

    Calling all Python SU members...
        http://mail.python.org/pipermail/python-list/2002-April/098790.html
    
    Python 2.1 vs. 2.2 Thread: "Why fix bugs to older versions"*30000  
        http://mail.python.org/pipermail/python-list/2002-April/098435.html
    *30000: Bonus silly string metaphor!
    
    Options on developing a module to support functionality tailored to
    multiple Python versions
        http://mail.python.org/pipermail/python-list/2002-April/099836.html
    
    Bit twiddling discussion shifts to finite chromatic mathematical fruit
    jokes
        http://mail.python.org/pipermail/python-list/2002-April/098420.html
    
    ---
    Announcements:
    Python Sidebar for Mozilla
        http://mail.python.org/pipermail/python-list/2002-April/099389.html
        http://starship.python.net/crew/mhammond/mozilla
    ...with a little help from the on-line Python Manual Search:
        http://starship.python.net/crew/theller/pyhelp.cgi
    
    PythonCard 0.6.5 Released
        http://mail.python.org/pipermail/python-announce-list/2002-April/001395.html
    
    
========================================================================

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

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

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

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

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

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

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


-- The Python-URL! Team--

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

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


From aahz at pythoncraft.com  Sun Apr  7 12:00:12 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 12:00:12 -0400
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <just-89BA7A.17235906042002@news1.xs4all.nl> <3CAF18F8.56821C0B@python.org> <mailman.1018172897.27393.python-list@python.org>
Message-ID: <a8pqec$g3r$1@panix1.panix.com>

In article <mailman.1018172897.27393.python-list at python.org>,
Gustavo Niemeyer  <niemeyer at conectiva.com> wrote:
>Guido:
>>
>> Agreed.  I've moved this to python-dev; see you all there.
>
>Bye.

Just in case it's not clear, you're welcome to join the discussion on
python-dev; it's not a restricted mailing list the way it used to be.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From tejarex at yahoo.com  Thu Apr 11 11:08:52 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 11 Apr 2002 15:08:52 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it> <a941v1$1dr$1@bagan.srce.hr>
Message-ID: <8Ght8.23772$_A.2251838@bin5.nnrp.aus1.giganews.com>

"Ante Bagaric" <abf at fly.srk.fer.hr> wrote in message
news:a941v1$1dr$1 at bagan.srce.hr...
> um, yep... I don't know where I got those commas myself :)
> I wasn't copying it from the actual output, I was producing it right
out
> of my memory and I guess I was a little too quick. No commas there.

Copy and paste is our best friend for not looking foolish.

I recent got complacent too, in spite of knowing this, and got bit
similarly.  Join the club.

Terry





From ikjos at email.uophx.edu  Sun Apr 28 19:57:14 2002
From: ikjos at email.uophx.edu (Ian Kjos)
Date: Sun, 28 Apr 2002 18:57:14 -0500
Subject: GC and resource finalization
Message-ID: <019901c1ef10$6c4d3940$0dce4542@austin.rr.com>

>>> class x(file):
 pass

>>> y=x('morethreading.py')
>>> y.foo=y
>>> y=None
>>>

In the slightly more general case, when (if ever) can the interpretter prove
that it is safe to close that file? Will the file be closed?











From buzzard at urubu.freeserve.co.uk  Mon Apr 15 17:48:05 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Mon, 15 Apr 2002 22:48:05 +0100
Subject: A little advice please? (Convert my boss to Python)
Message-ID: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>

My boss is considering moving to Python (from Poplog), so I've coded up
something to read in records from a text file (as below) and return a
statistic based on the numbers of unique and paired records (as he
requested).  Easy enough (code below), but now he wants to compare it for
speed against some existing code (Poplog?).  He also wants to be able to
calculate the statistic for only a subset of variables.

So what I'm looking for is speed, and some advice so that I don't end up
trying too many alternatives.  Maybe I should be using Numeric arrays (slice
out the superfluous columns)?  Maybe an n-dimensional array (for n
variables) and just count the cells with 1 and 2?  (Then slice / marginalise
and recount for queries on subsets of variables.  I like the sound of this,
but maybe there are limitations on array size / number of dimensions?)
Maybe I could avoid reading the data for superfluous variables and compare
records without the need to 'line.split()'?  Maybe my approach of using the
record as a dictionary key and incrementing the values is not the best way
of counting uniques and pairs?

Anyone done anything similar?  Any advice?  TIA.

Duncan


Identification        Var1    Var2    Var3    Var4    Var5 ...

0000000000001    N        0           2           3        0
0000000000002    N        0           2           2        0
0000000000003    N        1           3           3        1
0000000000004    Y        0           2           2        2
0000000000005    N        0           2           1        0

def do_stuff(filename, S):  #(S is floating point)
    f = open(filename, 'r')
    lines = f.readlines()
    f.close()
    lines = [line.split()[1:] for line in lines]
    lines = [line for line in lines if line != []]
    lines = lines[1:]
    dict = {}
    for line in lines:
        my_key = tuple(line)
        if dict.has_key(my_key):
            dict[my_key] += 1
        else: dict[my_key] = 1
    U = P = 0
    for value in dict.values():
        if value == 1:
            U += 1
        elif value == 2:
            P += 1

    return U*S/(U*S+P*(1-S))




From bokr at oz.net  Wed Apr 10 19:04:41 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 23:04:41 GMT
Subject: how to read hardware info on Win NT
References: <3cb2e686.21131515@news.easynews.net> <3cb45107.23060437@news.easynews.net> <MPG.171e2b58f71bb5039896a6@news.nwlink.com>
Message-ID: <a92ge9$2jn$0@216.39.172.122>

On Wed, 10 Apr 2002 11:59:07 -0700, Jeff Shannon <jeff at ccvcorp.com> wrote:

>In article <3cb45107.23060437 at news.easynews.net>, stojek at part-
>gmbh.de says...
> 
>> I need the id of my ethernet card for a license documentation tool we
>> want to write in Python. As many commercial softwares are using this
>> id for generating the license keys etc. 
>
>ISTM that this method has potential problems.  What happens if 
>the software is installed on a machine with no ethernet card?  
>What happens if the software is installed, and then the ethernet 
>card is upgraded?  What happens if the entire machine is 
>replaced, and the licensee wishes to install the software on 
>their new machine?
>
More people will migrate to free/open software ;-)

>I have always been leery about the idea of tying software 
>licenses to hardware IDs, because in an upgradable system, there 
>is no guarantee that those IDs will remain constant.
>
Yes, it's a damn nuisance. And end users don't like nuisances, so
hopefully the market will sort it out ;-)

Regards,
Bengt Richter


From dalke at dalkescientific.com  Tue Apr 16 17:30:07 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 16 Apr 2002 15:30:07 -0600
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9i4uv$k3h$1@slb4.atl.mindspring.net>

Philipp Lenssen:
>Is it possible to have a safe eval() on the server-side even when the
string
>consists of submitted user data?

What do you call 'safe'?  Ahh, your next post

> all that should be allowed is arithmetics (boolean as
> well) plus a limited set of functions that I write & define.

Will you allow "blahblahblahblah" * 2 ** 31 ? (which is a possible denial
of service attack if you don't have a few GB of RAM handy, since it makes
an 8GB string.)

If so, take a look at 'eval' in the documentation, under
  http://www.python.org/doc/current/lib/built-in-funcs.html
Call it with an empty locals and an almost empty globals

  eval("2+3", {"__builtins__": {}}, {})

The "__builtins__" is needed because if an empty globals dictionary
is used then eval will set it to the current __builtins__, which
allows all sorts of hacks, like

  eval("__import__('os').system('ls')")

                    Andrew
                    dalke at dalkescientific.com





From cowo at braveheart.cowo.net  Wed Apr  3 04:01:49 2002
From: cowo at braveheart.cowo.net (Giuseppe Corbelli)
Date: Wed, 3 Apr 2002 11:01:49 +0200
Subject: IOError 4 with pipes
Message-ID: <slrnaalh7t.g8a.cowo@braveheart.cowo.net>

Hi
I've done a pygtk GUI to cmdline program (global source indexing) as a
plugin for glimmer text editor.
It uses os.popen to exec the prog and read the output.
I developed it with Python 2.1 without problems.
Updated Linux distro do Mdk 8.2 and began getting
Traceback (most recent call last):
  File "/usr/lib/python2.2/site-packages/libglade.py", line 28, in __call__
    ret = apply(self.func, a)
  File "/usr/local/share/glimmer/scripts/Global.py", line 510, in
__on_searchbutton_cb
    completionlist = self.search (tag, self.__searchtype)
  File "/usr/local/share/glimmer/scripts/Global.py", line 162, in search
    for line in xreadlines.xreadlines (pipe):
IOError: [Errno 4] Interrupted system call

Offending line is

pipe = os.popen (self.__global_path+" -ax '"+initial_string+"'", 'r')

Of course exe path is correct.

Tried other pipe methods (popen2,3,popen3 class) without luck, keep getting
same error.
I am unable to understand why.

-- 
  Giuseppe "Cowo" Corbelli \/\ My software: http://cowo.yoda2000.net
        **ANTI-SPAM** -- **Substitute IHATESPAM with cowo to reply by mail**
 [cowo at braveheart cowo]$ make sex
    make: *** No rule to make target 'sex'.  Stop. 


From borcis at geneva-link.ch  Tue Apr 16 02:42:30 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 16 Apr 2002 08:42:30 +0200
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <3CBABA71.3040909@geneva-link.ch> <YwCu8.32367$vm6.5103753@ruti.visi.com>
Message-ID: <3CBBC7D6.8050203@geneva-link.ch>


Grant Edwards wrote:

> In article <3CBABA71.3040909 at geneva-link.ch>, Boris Borcic wrote:
> 
> 
>>For the shortcut semantics, we'd like to have :
>>     a xor b
>>
>>behave exactly like :
>>     (not b and a) or (not a and b)
>>
> 
> I don't understand how that's a shortcut (assuming that means
> the same thing as short-circuit).  You have to evaluate both
> operands of an xor operator.


That's right, the shortcut is in confusing short-circuit semantics
with returning operands when possible. In any case, my question
was elsewhere as the renaming of the thread suggested.

Boris Borcic
--
python >>> assert 304 in 340343, "P424D15E M15M47CH E2202"



From jeff at ccvcorp.com  Tue Apr  2 16:25:37 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 13:25:37 -0800
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net> <3ca8f000$1@news.nwlink.com> <JLcfwNF0zPq8EwJm@nym.alias.net> <3ca9e2bd@news.nwlink.com>
Message-ID: <MPG.1713c1ae2a7ecc1698968d@news.nwlink.com>

In article <3ca9e2bd at news.nwlink.com>, jeffbi at jetsoft.com says...
> 
> One of my other goals is to have a newsreader that works (as much as
> possible) the same on both Linux and on Windows, so I was thinking
> that using Tkinter for the GUI would probably be the best way to accomplish
> that.

My personal recommendation would be to look at wxPython.  Tkinter 
is a bit easier to pick up, especially if you have no previous 
GUI experience, but it becomes difficult to do more complex 
things with it.  Also, Tkinter apps always *look* like Tkinter 
apps.  By contrast, wxPython (and wxWindows, which it wraps) uses 
the platform-native widgets wherever possible, so a wxPython app 
looks like a Windows app on Windows and a GTK app on Linux.  
wxPython is also more flexible than Tkinter, and it doesn't rely 
on another scripting language like Tkinter does (Tkinter relies 
on TCL).

-- 

Jeff Shannon
Technician/Programmer
Credit International


From aleax at aleax.it  Mon Apr 15 03:05:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 15 Apr 2002 07:05:59 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <rZuu8.16169$b62.460568@news1.tin.it>

Raymond Hettinger wrote:

> The help and tutor maillists indicate that the use of input() rather than
> raw_input() is widespread.  I think this function is so harmful and so
> likely to be used that it should be deprecated hasta pronto (right away).
> When really needed, eval(raw_input()) is easy enough to code and explicit
> about what it does.
> 
> <no wink -- this function needs to go away>

Seconded.  'input' is an attractive nuisance (technical term of law, I 
believe) for novices.  The earliest phases of teaching Python would
be easier if using input started giving a warning, looking forwards to
eventual removal.


Alex



From bokr at oz.net  Wed Apr 17 00:28:07 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 04:28:07 GMT
Subject: Problem Writing Files
References: <3CBCE07C.9A0E0C4@slac.stanford.edu>
Message-ID: <a9itkn$g17$0@216.39.172.122>

On Tue, 16 Apr 2002 19:39:56 -0700, Roger Jones <rmj at slac.stanford.edu> wrote:

>I am using ActiveState Python 2.1 under Windows NT  and after opening a
>file with:
>
>afile=('test.dat','w')
>afile.write('some random test text')
>afile.read()
>
>It fills the file with many x's and 0's and also with various other
>text.
>How can this action be fixed so that it writes the text properly?
>
>Thanks!
>
Do you know how to copy from an interactive session and paste to your newreader
for posting? In the "DOS" window running Python, you can use Alt-Space>e>k
and then use the mouse to left-button-down on the top left character of a box area
that you create by dragging towards bottom right. When sized as desired, release
the mouse button and press Enter to copy to clipboard.

>From IDLE it's more normal windows text selection, so you should know.
For best response here, please try one of the above.

 >>> afile=('test.dat','w')
 >>> afile.write('some random test text')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 AttributeError: 'tuple' object has no attribute 'write'
 >>> afile.read()
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 AttributeError: 'tuple' object has no attribute 'read'

Suspicion is you were just theorizing, or didn't transcribe carefully ;-)

RTFD also applies:

 >>> print file.__doc__
 file(name[, mode[, buffering]]) -> file object
Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
writing or appending.  The file will be created if it doesn't exist
when opened for writing or appending; it will be truncated when
opened for writing.  Add a 'b' to the mode for binary files.
Add a '+' to the mode to allow simultaneous reading and writing.
If the buffering argument is given, 0 means unbuffered, 1 means line
buffered, and larger numbers specify the buffer size.
Note:  open() is an alias for file().

I'd do it for you, but thought you'd get more out of this ;-)
Maybe the next go won't work either, but things'll go faster if you
post the real thing.

Regards,
Bengt Richter


From pearu at cens.ioc.ee  Tue Apr  2 15:55:46 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Tue, 2 Apr 2002 23:55:46 +0300 (EEST)
Subject: Something and nothing (Re: PEP 285: Adding a bool type)
In-Reply-To: <3CAA03E4.A79BA593@noaa.gov>
Message-ID: <Pine.LNX.4.21.0204022326250.8932-100000@cens.ioc.ee>

On 2 Apr 2002, Chris Barker wrote:

> WOW! fabulous post, Laura.

I agree.

> Laura Creighton wrote:
> 
> > Python does not distinguish between True and
> > False -- Python makes the distinction between something and nothing.
> 
> > So I out-and-out tell people this.  {} is a dictionary-shaped nothing.
> > [] is a list-shaped nothing. 0 is an integer-shaped nothing.
> > 0.0 is a float shaped nothing.
> 
> Thinking in terms of something vs. nothing is an excellent way to
> explain how Python treats these things, but I still have a problem with
> it. Who is to say that an empty list is nothing? or 0 for that matter. 0

You seem missed essential words in the definitions. Laura defined that
an empty list is a "list-shaped nothing", not just a "nothing".
You can think of a None as a true "nothing". 

Note that one can add certain "nothing" to an object (with the same type
as this "nothing") with an effect that there is no effect. 

This holds for strings, lists, tuples, numbers, in some sense also for
dictionaries, and trivially for None: there is no object with the same
type as None (except None itself, addition of None to None is yet
unimplemented feature in Python;-) that would be subject to the addition
of nothing.

Pearu




From cbrown at metservice.com  Tue Apr 23 17:42:14 2002
From: cbrown at metservice.com (Colin Brown)
Date: Wed, 24 Apr 2002 09:42:14 +1200
Subject: String handling bug in Python
Message-ID: <3cc5d5c8$1@news.nz.asiaonline.net>

I have entered the following at the command line in Python 2.1.1 on Linux
and in Python 2.2 on NT and get the same results:

>>> r'\'
  File "<string>", line 1
     r'\'
        ^
SyntaxError: invalid token
>>>

It appears as if the r option for strings does not handle a backslash as the
last string character properly.

Colin Brown
PyNZ





From steffen at molli15.org  Tue Apr  9 09:41:56 2002
From: steffen at molli15.org (Steffen Kirschke)
Date: Tue, 09 Apr 2002 15:41:56 +0200
Subject: Simple question! urgent
References: <62l5buo5do5eh0h93b4r9ua8skp61b7302@4ax.com>
Message-ID: <a8ur48$506$1@news.ilk.de>

RobW wrote:

> x = [1,2,3,4]
> 
> How to get NUMBER OF ITEMS in array? I need it for checkin nr of items
> in argv[]

len(x)
> 
> --
> RW



From billludden at email.msn.com  Sun Apr  7 19:23:53 2002
From: billludden at email.msn.com (billludden)
Date: Sun, 7 Apr 2002 17:23:53 -0600
Subject: Newbie:  How to sound an alarm
Message-ID: <ucFMmso3BHA.1332@cpimsnntpa03>

Hi,
    I am attempting to learn to program using Python (first language).
     I would like to know how to make my box sound a tone or beep when my
program finishes.
I posted this query last week but I cannot access messages posted during
that time (must be Windows).
Thank You,
Bill





From peter at engcorp.com  Tue Apr  9 09:26:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 09:26:27 -0400
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net>
Message-ID: <3CB2EC03.B78593B1@engcorp.com>

akhar wrote:
> 
> Has anyone ported Python to embedded devices? I heard of the deeply embedded
> python porject but the link is dead. What would it take to port python to an
> embedded platform (whether be it the mc68hc11 or PICs)?

I believe a better choice would be Lua, which as far as I understand
is intended for such things (although not sure about HC11 and PICS).

Python pretty much requires more horsepower, memory, and other
resources than either of those two platforms support.  It does
not have as a goal the ability to run on 16-bit platforms, and
generally needs on the order of 200K (that's a pretty rough guess)
for a real system.  I believe there have been attempts, with some
success, to cram it into smaller packages, but those efforts have
not come to fruition yet.  (It would be a subset of Python, for sure.)

-Peter


From ChrisE at lantech.com  Sun Apr 28 16:35:48 2002
From: ChrisE at lantech.com (Christopher Encapera)
Date: Sun, 28 Apr 2002 16:35:48 -0400
Subject: File sizes from MSWin...
Message-ID: <BC74F8F7E87DD511990C000103E2FEFA3003@XCHG>


>If you upgrade to 2.2 you should be fine.  That includes large file
>(64-bit) support under Windows by default, and the os.path.getsize()
>call will return a long integer rather than an integer.  The bonus is
>that it's then also portable across systems rather than limited to
>Windows.


Upgrading to 2.2 did the trick.  Thanks again for the info!


Christopher Encapera
Lantech.Com
chrise at lantech.com
502-267-4200





From Matthias.Hanitzsch at dynamic.de  Fri Apr 19 08:57:28 2002
From: Matthias.Hanitzsch at dynamic.de (Matthias Hanitzsch)
Date: Fri, 19 Apr 2002 14:57:28 +0200
Subject: [Pymacs] Regular expressions
Message-ID: <3CC01438.96A7FB88@dynamic.de>

Hallo,

I use Pymacs to extend Emacs.
What is the syntax of the regular expression <REGEX> if I use
the following ?:

from Pymacs import lisp
...
lisp.looking_at("<REGEX>")
...

Is it the EmacsLisp syntax or Python syntax ?


Thank you for your help,
Matthias



From a.clarke11 at ntlworld.com  Sun Apr 14 16:17:29 2002
From: a.clarke11 at ntlworld.com (tony clarke)
Date: Sun, 14 Apr 2002 21:17:29 +0100
Subject: Neural Networks
Message-ID: <3CB9E3D9.121EA05C@ntlworld.com>

Hi

Does anyone know any sites, or have any code or other info about
artificial intelligence. I am specifically interested in simple neural
network programming, being a simple soul myself!

Thanks in advance

Tony



From martin at v.loewis.de  Wed Apr  3 02:24:08 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Apr 2002 09:24:08 +0200
Subject: Python compilers
References: <3CAA9786.2070700@moofed.com>
Message-ID: <m3ofh1nt2f.fsf@mira.informatik.hu-berlin.de>

Rob Hazlewood <rob at moofed.com> writes:

> I'm thinking of learning python, and am a little confused as to the
> overall structure of python programming.
> 
> Would I be right in saying that Python is an interpreted language?

No: there are no "interpreted languages" and "compiled languages";
there are only interpters for languages, and compilers for languages.
The same language may be executed under each model, or in a mixed
approach (like Java).

> I know it can be run as an interpreted language, however I seem to
> remember people talking about python compilers.
> 
> Do Python compilers exist? are they efficient?

Python interpreters (both CPython and Jython) generate byte code
before executing anything - this is a compilation step.

There have been attempts to write compilers to compile Python to C
code; those gave only minimal speedup.

There is also an ongoing attempt to perform just-in-time compilation
(Psyco); this gives measurable speed-up, for selected applications.

> If there is a python compiler, surely it would be able to generate
> code that is as efficient as a C compiler, resulting in what would be
> the best available OO language? (having simplicity, speed)

The Python interpreter is written in C, so it already runs at C speed.

A Python compiler most likely would *not* be able to produce code that
is as fast as code that a C compiler could produce for an "equivalent"
C program - because both programs can't possibly be "equivalent".

Just consider the program

int add(int x, int y){return x+y;}

In C, the compiler can compile this to a machine program that uses the
processor's integer addition, and expects arguments on the stack. The
"equivalent" Python program is

def add(x,y):return x+y

Right? Wrong. The Python function can also be used to do

print add("Hello,","World")

So a compiler compiling the Python version needs to take the
possibility into account that you are not passing integers, but
strings - you can't know, at compile time, what "+" is.

HTH,
Martin


From cbbrowne at acm.org  Fri Apr 12 11:20:46 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: Fri, 12 Apr 2002 11:20:46 -0400
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de>
Message-ID: <m31ydlndtd.fsf@chvatal.cbbrowne.com>

Uwe Schmitt <uwe at rocksport.de> wrote:
> I'd like to send "raw" data to a printer, how can I do this,
> using Python ???

It's terribly platform dependent.

On Unix-like platforms, it would involve opening output to some file
like /dev/lp, /dev/lp0, or such, which might require special
permission configuration.

On Win32, it likely involves throwing data at some Win32 function;
heaven only knows what.

If you haven't even figured out what platform you're on, and the
implications of that, I'd suggest the thought that you've really not
thought out what you want...
-- 
(concatenate 'string "cbbrowne" "@ntlug.org")
http://www.ntlug.org/~cbbrowne/sap.html
Visualize whirled peas.


From mlh at vier.idi.ntnu.no  Fri Apr  5 22:53:19 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 03:53:19 +0000 (UTC)
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com>  <mailman.1017895453.22508.python-list@python.org>
Message-ID: <slrnaass9f.178.mlh@vier.idi.ntnu.no>

In article <mailman.1017895453.22508.python-list at python.org>, Gustavo
Niemeyer wrote:
>
>Hi Neal!
>
>> 	    if a is True: print 'True'
>[...]
>> 	test.py:6: Comparisions with True are not necessary
>
>Testing if "a is True" is ok. True and False are singletons this is
>the best way to know if 'a' is really 'True' (and not 1, [0], "foo",
>etc). This test will be necessary sometimes.

It is neither worse nor better than using '==', IMO. Yes, it is a
singleton, but the danger is using 'a is True' instead of simply 'a'.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From eppstein at ics.uci.edu  Sun Apr  7 11:53:49 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 07 Apr 2002 08:53:49 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com>
Message-ID: <eppstein-CB6351.08534907042002@news.service.uci.edu>

In article <ub0k49taktp243 at news.supernews.com>,
 "John Roth" <johnroth at ameritech.net> wrote:

> > By the way, this is conceptually very similar to ints (at least,
> > nonnegative ints): ints answer a specific question: how many things?
> > They do not answer the question: which things are they.
> 
> Not exactly. Consider the common case of 'take a number'
> played at various burocracies. The manager can tell how many
> there are by inspecting the next number on the roll, but each person
> waiting has his own number, and they can be called out of order.

I was talking about cardinal numbers, you're talking about ordinal 
numbers.

Ordinal numbers in programming are often used for converting objects 
into positions to look up data in an array.  They're less often needed 
in Python because we have dictionaries instead.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From huaiyu at gauss.almadan.ibm.com  Mon Apr 29 22:15:26 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 30 Apr 2002 02:15:26 +0000 (UTC)
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> <mailman.1020071811.14821.python-list@python.org> <87adrm38fu.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <slrnacrvhu.6u4.huaiyu@gauss.almadan.ibm.com>

Stephen J. Turnbull <stephen at xemacs.org> wrote:
>>>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
>>     Laura> People here seem to be thinking that the GPL is there to
>>     Laura> protect creators rights to their software.
>> 
>> I'm pretty sure that that is what rms has in mind, as long as both
>> "creators" and "software" are inclusive collective nouns.
[snip]
>
>But it didn't.  I don't see how the GPL is about consumer protection.
[snip]
>I've read them.  I understand the concept.
>
>My mother doesn't.  She will never notice how much software has been

One problem with your postings in this thread is that you are constantly
arguing in three lines without clear distinction: what is the intent of GPL,
what you think it would achieve, and what is the impression left on the
unwashed mass.

If you stick to RMS's intent, you'll see that it is indeed about consumer
protection from the very beginning.  There are plenty of evidence all over
the net, from GNU, FSF and RMS writings, and also summarized by several
posters here.

Whether we agree this is its actual effect is a different issue.  

Whether it is something known to the mass media is yet another issue.

Not distinguishing these issues makes your posts sound rhetorical.  Not that
rhetorics are not useful.  They may be the only valid tool in certain
situations, esp for advocacy - see Alex's thought-provoking analysis.  

However, injecting rhetorics into discussions that are intended to be
technical by most discussants does often produce irritating effects.  

BTW, My impression on the distinction between OSI/FSF is just about the
opposite of yours: I think OSI is arguing from the producer line, while FSF
is from the consumer line.  But I do not intend to argue this point here,
but rather to use it as a data point to show that you cannot just assume
that your view is *obviously* correct.

Huaiyu


From ggr at qualcomm.com  Wed Apr 24 12:50:59 2002
From: ggr at qualcomm.com (Gregory G Rose)
Date: 24 Apr 2002 09:50:59 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <slrnacdape.60f.akuchlin@ute.mems-exchange.org> <7xu1q1mdwd.fsf@ruckus.brouhaha.com>
Message-ID: <aa6npj$3r@qualcomm.com>

In article <7xu1q1mdwd.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2002a at nightsong.com> wrote:
>I should re-read Krawczyk's HMAC paper to see what that fancy padding
>is really supposed to do.  Remember that the underlying encryption
>here (SHA-OFB) probably isn't the greatest in the world either.

The proof of security of NMAC (that HMAC is based
on) requires that the inner and outer hash
functions be independent. For HMAC, this is
approximated by using the padding so that the keys
are effectively different, making the hash
operations effectively independent (wave hands
here).

>And it actually might be faster to use two separate hash-based auth
>keys than deal with the fancy HMAC padding, if a simple double hash
>has some vulnerabilities.

Two different keys would certainly satisfy the
conditions.

Greg.
-- 
Greg Rose                                       INTERNET: ggr at qualcomm.com
Qualcomm Australia          VOICE:  +61-2-9817 4188   FAX: +61-2-9817 5199
Level 3, 230 Victoria Road,                http://people.qualcomm.com/ggr/ 
Gladesville NSW 2111    232B EC8F 44C6 C853 D68F  E107 E6BF CD2F 1081 A37C


From abagaric_NOSPAM at rest-art.hr  Thu Apr 18 11:44:16 2002
From: abagaric_NOSPAM at rest-art.hr (Ante Bagaric)
Date: Thu, 18 Apr 2002 17:44:16 +0200
Subject: Assignment operator?
References: <a9mksa$8dqs$1@as201.hinet.hr> <WiBv8.15852$eS3.5828@atlpnn01.usenetserver.com>
Message-ID: <a9mpq4$940c$1@as201.hinet.hr>

> Obviously, immutables must implement augmented assignment by re-binding
the
> name to a new object, otherwise
>
> >>> a = 3
> >>> a += 2
> >>> a
> 5

Umm, indeed. Somehow I overlooked this problem...

> would not be possible.
> I think what you haven't understood is the update in-place semantics that
> mutable object implementations *may* choose to adopt. Otherwise you seem
to
> have it down pat.

Hmm, and the compiler doesn't know in advance what object a name references,
so it has no choice but to flag '+=' as rebinding operation even if it is
not
the case. It doesn't sound nice, but I admit there's no obvious alternative
that
comes to mind. This whole namespace and local/global issue has an aura of
'unfinishness' around it :)







From claird at starbase.neosoft.com  Mon Apr 29 09:28:17 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 29 Apr 2002 08:28:17 -0500
Subject: Scripting lanuages for laboratory tests
References: <aagkfm$6kf$1@news5.svr.pol.co.uk>
Message-ID: <4BC38B9E40BBBF12.D92E4F7F41DFEC2F.A65AFEBC15784948@lp.airnews.net>

In article <aagkfm$6kf$1 at news5.svr.pol.co.uk>,
Etta Makwetta <etienne at coetzeee.freeserve.co.uk> wrote:
			.
			.
			.
>Neither Tcl nor Python supports calling an arbitrary dll.  Tcl supports
			.
			.
			.
Alex and I posted references to extensions which allow both
these languages to support use of arbitrary DLLs.

If this requirement is particularly vital for you, you might
be most comfortable with a scripting language based on C
<URL: http://www.itworld.com/AppDev/710/swol-0818-regex/ >.
-- 

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


From donn at u.washington.edu  Tue Apr  2 17:39:21 2002
From: donn at u.washington.edu (Donn Cave)
Date: 2 Apr 2002 22:39:21 GMT
Subject: ctrl-c and os.system problem
References: <mailman.1017777544.11640.python-list@python.org>
Message-ID: <a8dbup$hec$1@nntp6.u.washington.edu>

Quoth Noom Knews <noomnews at yahoo.com>:
| I have a simple program like this
|
| os.system("ls -lR / > /tmp/ls.out")
| print "end"
|
| When I press ctrl-c, the "end" message is still print.
|  I'd like to terminate the whole process when pressing
| ctrl-c instead of just "end".
|
| I tried to catch the signal already but can't catch
| the SIGINT or anything else at all during the
| os.system call.   This does not work either
...
| This was tried on Redhat Linux with python 1.5.2

NetBSD's system(3) man page explains that system ignores SIGINT
and SIGQUIT.  I assume the same must be true on Red Hat Linux,
and Python calls C system().

Maybe you could use spawnv()?

	Donn Cave, donn at u.washington.edu


From patmiller at llnl.gov  Mon Apr  8 16:15:31 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Mon, 08 Apr 2002 13:15:31 -0700
Subject: Pyrex and weave.accelerate
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1 <00e101c1dd14$5620a930$6b01a8c0@ericlaptop>
Message-ID: <3CB1FA63.6ADACEDE@llnl.gov>

The Pyrex stuff is really cool.. The focus on my work
on weave.accelerate is a little different (though
same end goal, easier bridge from C/C++).  The nice
bit about weave.accelerate is that it allows a fallback
position to just run the native Python in cases
where the compiler can't run (system resource limits,
Jython, Pippy, etc...)

It appears that the premise of Pyrex (assumption warning!
I'm not reading Greg's mind!) is that C types have
Python equivalents expressible through the API
(e.g. short/int/long are expressible through
the PyInt_AsLong and PyInt_FromLong methods).

The weave accelerate works somewhat the other way,
it presumes the existence of a C/C++ API to Python object types.
Weave.accelerate allows programmers to teach weave
the API for a type and then have it build the
extension module from that knowledge.


Where Eric alludes to scientific needs, he is
referring to array/field acceleration.  Now
the initial target of weave is Numpy arrays...


def div(u):
  d = zeros(len(u),'d')
  for i in range(len(u)-1):
     d[i] = u[i+1]-u[i]
  return d

So, weave.accelerate tries to replace function calls with
equivalent C/C++ code.  Now the nice thing is for the
simulation code I work with that has its own wrapping
scheme for a field class derived from C++ STL vectors<double>.
I can describe the API for this special field type and
then stub in the correct code for Numpy arrays, vector<double>,
specialized field classes, or even normal Python lists.

Weave accelerate uses type inference based on types in real
use, so

def foo(n):
  v = n + 1
  return v

If you call foo(3), weave.accelerate examines the value 3
and determines that it matches the C++ type, long.  It
derives the type of "v" by examining the return type
of a long + long (hence it is long).

If you later call foo(3.3), weave will compiler another version
of the function (this time, v will have C++ type double).

Pyrex looks to extend the language, I was looking at
doing something similar, but with type assertions for
cases where the types were well known.

So as opposed to Pyrex language extensions

def foo(int n):
   cdef int v
   v = n + 1
   return vf


I can encode the types as

def foo(n):
   assert type(n) == IntType, weave.TypeError
   v = n + 1
   return v

Now, I get full fallback to Python (in case weave fails to compile)
with proper typechecking.....

Eric and I will have more details for the O'Reilly conference
paper and talk.... Stay tuned.

Pat

-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

Change is not made without inconvenience, even from worse to
better -- 16th Century British theologian Richard Hooker

eric wrote:
> 
> 017935351.5088.python-list at python.org> <slrnaasqc7.178.mlh at vier.idi.ntnu.no>
> Subject: Re: A faster Python?, Python compiler, Dylan,...
> Date: Fri, 5 Apr 2002 21:39:55 -0500
> Organization: enthought
> MIME-Version: 1.0
> Content-Type: text/plain;
>         charset="Windows-1252"
> Content-Transfer-Encoding: 7bit
> X-Priority: 3
> X-MSMail-Priority: Normal
> X-Mailer: Microsoft Outlook Express 5.50.4522.1200
> X-Mimeole: Produced By Microsoft MimeOLE V5.50.4522.1200
> 
> Hey Magnus,
> 
> That is exactly what Pat Miller of LLNL is working on.  The function
> weave.accelerate() (in CVS and quite limited right now) is aimed at this
> purpose.  It is a little different than Python2C to satisfy some needs in the
> scientific community.  Also, Greg Ewing's new Pyrex project is a good candidate
> for this sort of thing -- hopefully we can learn from each other because Pyrex
> looks very promising.
> 
> So, the short answer is "good idea -- we're working on it." :-)




From Chris.Barker at noaa.gov  Tue Apr  9 16:37:22 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 09 Apr 2002 13:37:22 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org> <Npor8.8281$q63.818@atlpnn01.usenetserver.com> <3CAE52CA.60F56E3D@python.org> <oAfs8.26107$Sw5.2561@atlpnn01.usenetserver.com>
Message-ID: <3CB35101.AF4FA8F7@noaa.gov>

"Guido van Rossum" <guido at python.org> wrote in message
> I found a lot of
> disagreement on what people consider a "real" Boolean type.  I did find
> wide agreement on something that I had myself taken for granted, but which
> some proponents of "real" Booleans were challenging: the usefulness of
> Python's ability to test almost *any* object for its truth value, what
> Laura calls something/nothing.

I am one of those people, and I do see the "usefulness of Python's
ability to test almost *any* object for its truth value". It is
fantasically useful. What I don't like is that it is implicit, rather
than explicit.

I would love being able to do:

if bool(anyobject):  #(or maybe istrue(), or something like that)
	whatever

What I don't like so much is:

if anyobject:
	whatever

because unless anyobject is of a class I have defined myself, Python has
decided ahead of time what an untrue (or nothing) value is for that
object. That means I can't have a function return an integer, for
example, and have it evaluate as true, when a zero is not "nothing" in
this context.

and yes, I know that it is WAY too late in Python develpment to
impliment this, and I do love the laguage anyway.

By the way, can I now subclass from int, and use objects of the
resulting class as an index, and also redefine __nonzero__ or whatever,
so that I can get a 0 to be both an index and evaluate as true? I'm not
running 2.2, so I can't test it myslef. If I can, then I guess I can
achieve most of what I want, but I'm sure it would confuse people if
anyone else were to use my class!

-Chris





-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From m.faassen at vet.uu.nl  Mon Apr 22 07:46:05 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 22 Apr 2002 11:46:05 GMT
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org> <eppstein-2C56E0.23265219042002@news.service.uci.edu>
Message-ID: <aa0t5t$bb6$1@newshost.accu.uu.nl>

David Eppstein <eppstein at ics.uci.edu> wrote:
> In article <mailman.1019279977.28194.python-list at python.org>,
> Tim Peters <tim.one at comcast.net> wrote:

>> > I'm really delighted to hear about performance improvements in file
>> > object input in recent Python versions (in a message that would
>> > have been in this thread if Tim Peters used a newsreader instead
>> > posting to the list.)
>> 
>> Hmm.  Shows up fine in the newsreader I use -- this, like hearing about the
>> speedups for the first time from that msg, may be another symptom of General
>> Upgrade Fear <0.9 wink>.

> The point is that your responses don't have the appropriate References: 
> header, so they don't get threaded with other messages on the same topic.

Right, this is a longstanding timbot bug; the timbot consistently breaks
threads. It must be due to some extremely funky space time continuumisms
otherwise you'd think the timbot would've noticed it and fixed it by now.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From loewis at informatik.hu-berlin.de  Fri Apr 12 11:31:12 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 12 Apr 2002 17:31:12 +0200
Subject: How to test tp_print?
References: <20020412112023.201$FO@news.newsreader.com>
Message-ID: <j4662xszlr.fsf@informatik.hu-berlin.de>

"Hernan M. Foffani" <hfoffani at yahoo.com> writes:

> My question is:
> Is this code portable across different C stdio
> libraries? I'm thinking in line delimiters here.
> I suspect it is not :-(

You are right, this is not portable.

> Any other approach to this?

Open the file in binary mode.

Regards,
Martin


From aleax at aleax.it  Mon Apr  8 06:20:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 10:20:44 GMT
Subject: import and function question
References: <d906b2c1.0204080155.77b87631@posting.google.com>
Message-ID: <0aes8.59173$pT1.1704009@news1.tin.it>

Moray Taylor wrote:

> Hi,
> 
> I have two questions..
> 
> a) Can I 'import' a module with the full path, i.e. I don't want to
> add it to sys.path, just something like 'import /home/me/mod.py', is
> this possible.
> I have tried adding it to the path, importing, and immediatly removing
> it from the path, but the next time I do an 'import' with a different
> module of the same name, it picks up the old one, even though it isn't
> in the sys.path.

The import statement will always pick up sys.modules[name] when you
import name.  If that's not what you need (and for plugins you may
not want that, indeed) see the imp module in the standard library.
imp.find_module lets you find a module on a directory-path of your
choice, then imp.load_module lets you load the module you've found.

> b) I need to overwrite a function with a new one of the same name, but
> 'import'ed from a module. Is this possible, I really just need to
> destroy this function before I import the new one, can this be done?

del thefunctionname

will unbind the name from the function object ("destroy" only if there
are no other references -- you cannot destroy an object to which there
are outstanding references, Python saves you from dangling reference
problems; if you have subtle needs e.g. for caching, see the weak
references module).  Most often, though, just:

thefunctionname = themodulename.thefunctionname

(after importing the module and binding the module object to name
themodulename -- whether by an import statement or by more devious
means) will suffice.


Alex



From phr-n2002a at nightsong.com  Sun Apr  7 12:18:38 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 07 Apr 2002 09:18:38 -0700
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <a8pqal$ftm$1@panix1.panix.com>
Message-ID: <7xpu1bh481.fsf@ruckus.brouhaha.com>

aahz at pythoncraft.com (Aahz) writes:
> I think it might be possible to make changes in the way Guido-introduced
> PEPs get presented to reduce the amount of shouting from people like
> Laura.  I hold no hopes that anything would get better results from
> people like Arthur and Paul Rubin.

I'm mildly opposed to the bool PEP but it's not that big a deal for
me.  I'm more unhappy with the process that results in adding
something like that to a minor version (2.3).  If it was saved for
3.0, I think that would make more sense, for reasons I've explained.
There are other features that I would LIKE to see added, like
assignments in expressions; but those, too, shouldn't first appear
in a minor release.



From pc_olsen at yahoo.com  Tue Apr  9 15:38:33 2002
From: pc_olsen at yahoo.com (Peter Olsen)
Date: 9 Apr 2002 12:38:33 -0700
Subject: Python-Aware Modeler/Engineer Seeks Position/Contacts
Message-ID: <12bb92e6.0204091138.23741368@posting.google.com>

Hello,

I am an experienced engineer/mathematician/modeler who has 
built models and done computational mathematics in Python.  

I offer:

* Advanced education in Engineering, Operations Research, 
  Computer Science, and Applied Mathematics

  ** BS, United States Coast Guard Academy
  ** MS Aeronautical Systems, University of West Florida
  ** MS Operations Research, MS Aeronautical Engineering, both
     from the United States Naval Postgraduate School
  ** Professional Degree of Aeronautical Engineer (the 
     terminal degree for practicing engineers not
     engaged in research), US Naval Postgraduate School
  ** Diploma (with highest distinction) United States Naval
     War College, College of Command and Staff.
  ** Additional graduate work in mathematics, statistics, and
     computer science at Johns Hopkins University,
     and the University of Maryland Baltimore County. 

* Extensive experience in modeling, simulation, and analysis.
  (I built the quantitative work-progress model used to allocate
  resources for the EXXON VALDEZ oil-spill cleanup.)

* Wide-ranging technical achievement
  
  ** National Security Agency "Master Engineer"
  ** Principal Member of Technical Staff, Draper Laboratory
  ** Principal Scientist, Information Extraction and Transport

* Experienced leadership, particularly in technical teams

  ** Selected as National Security Agency "distinguished-leader-
     in-fact" for 1996
  ** Experience as a Coast Guard Commissioned Officer and Search-
     and Rescue Aircraft Commander (leading crews 250 miles offshore
     in bad weather).

* Licensure as a Professional Engineer (Maryland Licence 12283)

* Fluency in Java, C, FORTRAN, and Lisp. Acquaintance with Smalltalk,
  S-Plus, APL, and R --- and a STRONG preference for Python

*  Current TOP SECRET clearance

I'm looking for a position in which I can use my skills, experience, 
and creativity to a bring substantial value to an organization that 
encourages teamwork, values leadership, and rewards achievement.

If your organization doesn't need my skills, perhaps you know 
of another that might.  I will appreciate any additions to my
professional network.

Complete resume emailed promptly upon request.

Peter Olsen, P.E., Ae.E.
pc_olsen at yahoo.com


From johnroth at ameritech.net  Sun Apr 21 07:58:34 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 21 Apr 2002 07:58:34 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org>
Message-ID: <uc5aajf0m2dmc5@news.supernews.com>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1019381679.10995.python-list at python.org...
>
>
> That being said, I guess I just don't get the strong reaction to this
secondary
> point.  I'm new here but I get the impression that "lisp" is a naughty
word,
> for some unfathomable reason (unless it's simply that it's syntax
sucks).  For
> me to say Python shares many concepts from Lisp is intended as praise.
You
> react as if it's some kind of an insult.

I think what people are trying to tell you is that many of the
features that Lisp pioneered are now features of so many
languages that it simply isn't necessary to quote Lisp as a
direct ancestor of anything, unless you're doing historical
research.

Python has a well understood ancestry, starting with the
experimental ABC language, and continuing on with
suggestions from a wide variety of sources.

One of the things that is most unusual about Python is
it's rejection of bracketing where it isn't needed:
using indentation for block structure is one example.
This is diametrically opposed to Lisp.

John Roth




From mwh at python.net  Thu Apr 25 07:12:29 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 25 Apr 2002 11:12:29 GMT
Subject: is there any dedicated mailing list fo python PMW project?
References: <mailman.1019720565.10727.python-list@python.org>
Message-ID: <lk4ri0kozp.fsf@pc150.maths.bris.ac.uk>

"Mariappan, MaharajanX" <m_mariappanX at trillium.com> writes:

> Python user  comunity,
> 
> is there any dedicated mailing list fo python Pmw project?

There seems to be a list here:

http://sourceforge.net/mail/?group_id=10743

It looks pretty quiet, though.

Cheers,
M.

-- 
  I have a feeling that any simple problem can be made arbitrarily
  difficult by imposing a suitably heavy administrative process
  around the development.       -- Joe Armstrong, comp.lang.functional


From guido at python.org  Thu Apr  4 09:22:26 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:22:26 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xy9g4nk9w.fsf@ruckus.brouhaha.com>
Message-ID: <3CAC6283.424163FB@python.org>

Paul Rubin wrote:
> 
> I think there will need to be a format code for truth values:
> 
>   ("%b" % True) => "True"

There's no need.  "%s" % True already yields 'True'.  if you want
coercion to bool, use "%s" % bool(b).

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


From db3l at fitlinxx.com  Mon Apr 22 15:39:36 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Apr 2002 15:39:36 -0400
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org> <slrnac2v9d.kv6.jon+usenet@snowy.squish.net> <mailman.1019315200.29890.python-list@python.org>
Message-ID: <uvgaj7cav.fsf@ctwd0143.fitlinxx.com>

holger krekel <pyth at devel.trillke.net> writes:

> But is it true that you 'never' use them somewhere else? 
> I would think that one would like to use 'r' whenever you don't 
> want the escape-sequence-string-interpretation which is not really
> restricted to regexp-strings. 

Exactly.

I use them all the time when I need a \ in my strings for some reason
or another.  Regex's are probably the most obvious cross-platform
case, but I think the next most likely case in our code is when
encoding Windows file paths (so this is less common for Unix
environments).  Sometimes you want a native looking path for display
purposes (otherwise in many cases you can just use forward slashes
even under Windows), or sometimes (as with some UNC paths) you really
need the backslashes.

> I have noticed that e.g. the docstring of the module os.py uses
> r'...'

I've done it in docstrings too when I used examples or other text that
included characters that regular strings would interpret.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From claird at starbase.neosoft.com  Wed Apr 24 10:52:57 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2002 09:52:57 -0500
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at> <EC7694DBF1581DFC.5E45D82C67327D60.BDDD4EB8581664F1@lp.airnews.net> <3CC4549E.B63D3E8A@tundraware.com>
Message-ID: <A0880FA4391D6351.8D6B4E946BCB6481.B58F11AD2969D518@lp.airnews.net>

In article <3CC4549E.B63D3E8A at tundraware.com>,
Tim Daneliuk  <tundra at tundraware.com> wrote:
>Cameron Laird wrote:
>
>> I'll say what I think is the same, in somewhat different
>> language:  do not draw conclusions from mathematicians'
>> philosophy.  There's reasonably strong evidence that
			.
			.
			.
>You're right, of course, but I would expand this to include anyone
>of great ability.  Precious few geniuses can ever explain the
>'epistemology of their expertise' be it in mathematics, boat building,
>brick laying, or square dancing.  This is, to me, another reason to
>dismiss the mechanist/meathematical explanations for the human mind and
>spirit.  There is something transcendental about genius that I doubt can
>ever be described in purely mechanical terms...
			.
			.
			.
Oh, absolutely.  Yes:  geniuses and even more
mundane experts in all sorts of fields have
proven to be unreliable guides to their own
success.  Given this general proposition, one
might hope for better from mathematicians for
several reasons--that philosophical precision
*looks* like mathematics, for example--but it
turns out not to be so; mathematicians have no
systematic ability to do ontology that's more
consequential than their boat-building peers.

I'll make it explicit:  I'm not denying the
value of the foundational work of Poincar?,
Brouwer, and so on.  I am only celebrating
them as happy exceptions.
-- 

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


From loewis at informatik.hu-berlin.de  Wed Apr 10 10:20:51 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 10 Apr 2002 16:20:51 +0200
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org>
Message-ID: <j4wuvfzlbw.fsf@informatik.hu-berlin.de>

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

> gcc -c -DNDEBUG -O2 -I. -I./Include -DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
> In file included from Python/thread_pthread.h:9,
>                  from Python/thread.c:113:
> /usr/include/pthread.h:238: parse error before `destructor'
> /usr/include/pthread.h:238: parse error before `)'
> make: *** [Python/thread.o] Error 1
> 
>    The line in question /usr/include/pthread.h:238 is
> int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
>    Nothing suspicious.
> 
>    Any idea what can be wrong? 

I suggest to produce preprocessor output, via -E (not sure whether
2.8.1 already had --save-temps).

Regards,
Martin



From gregory.p.green at boeing.com  Tue Apr  2 15:12:21 2002
From: gregory.p.green at boeing.com (Greg Green)
Date: Tue, 2 Apr 2002 20:12:21 GMT
Subject: runtime location discovery
References: <20020330014832.GF7602@isis.gerrietts.net> <mailman.1017700143.8870.python-list@python.org>
Message-ID: <x6mg02dx3kq.fsf@skorch.rt.cs.boeing.com>

I know this doesn't answer the question, but finding a program's
path __is__ covered in the Unix FAQ:

http://www.erlenstar.demon.co.uk/unix/faq_2.html#SEC23

-- 

Greg Green


From mwh at python.net  Wed Apr 17 10:26:05 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 17 Apr 2002 14:26:05 GMT
Subject: Python signal handling
References: <mailman.1019049215.30218.python-list@python.org>
Message-ID: <lkwuv6qtxa.fsf@pc150.maths.bris.ac.uk>

Dave Swegen <dswegen at software.plasmon.com> writes:

> I've recently been bitten by a nasty (undocumented) feature in Perl,
> where installing a signal handler to catch SIGCHLD will eventually cause
> a core dump. So before I venture into any daemon handling in python I
> was wondering if Python too suffers from problems with signal handling?

Python suffers from problems with signal handling, but it sounds like
they're not the same problems as those suffered by perl.

> I believe the reason for perl's problem is that it isn't re-entrant, and
> thus two rapidly incoming signals will cause perl to fo *Pooof*.

In Python, signal handlers (usually) get run in between opcodes, so if
an opcode takes a particularly long time to run, your signal handler
may not run until some time after the OS delivers the signal to your
process.

Cheers,
M.

-- 
  ARTHUR:  Why should a rock hum?
    FORD:  Maybe it feels good about being a rock.
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 8


From gmcm at hypernet.com  Fri Apr 12 08:54:34 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 12 Apr 2002 12:54:34 GMT
Subject: Include zlib.dll in McMillan Installer
References: <a93ld3$jb4$03$1@news.t-online.com> <Xns91ED608AADBDEgmcmhypernetcom@199.171.54.213> <a968re$qp9$07$1@news.t-online.com>
Message-ID: <Xns91EE5ABFB9F91gmcmhypernetcom@199.171.54.214>

[posted and mailed]

Harald Schneider wrote:

> Hello Gordon,
> 
> removing zlib.dll from excludes[] in bindepend.py adds zlib.dll to the
> archive.
> Can I do this by a command in the .spec file?

No, you can't override that from the .spec file.

However, you should be fine removing it. The exclude dates from
the Python 1.5 timeframe - in certain situations you could end
up with 2 copies of zlib.dll (run*.exe also used it, so it was
always included anyway).

-- Gordon
http://www.mcmillan-inc.com/



From peter at engcorp.com  Thu Apr 18 07:50:59 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 18 Apr 2002 07:50:59 -0400
Subject: Hostname Resolution not working in Explorer
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net> <3CBE4DAA.E710D618@engcorp.com> <200204181437.26855.rjones@ekit-inc.com> <mailman.1019105135.13010.python-list@python.org>
Message-ID: <3CBEB323.BFC82570@engcorp.com>

Richard Jones wrote:
> 
> On Thu, 18 Apr 2002 14:37, Richard Jones wrote:
> > This is a common problem with Windows - it doesn't know how to resolve
> > "localhost" in some situations. Usually it coincides with you changing the
> > network configuration (unplugging a network cable or something).
> 
> Oh, I forgot to add - I believe in _some_ versions of Windows you can
> hard-code an entry in a hosts file you place in the Windows System directory
> (IIRC - it's been a while - there'll be a sample file in there I think). The
> entry would read
> 
> 127.0.0.1 localhost
> 
> just the one line in a file called "hosts" in the Windows System dir, and it
> should fix things. Well, it did once for me :)

Well, he _did_ say that he'd already added that line to the HOSTS file....

-Peter


From tdelaney at avaya.com  Sun Apr 28 21:11:02 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 11:11:02 +1000
Subject: inline assignments in conditionals 
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A1@natasha.auslabs.avaya.com>

> From: Laura Creighton [mailto:lac at strakt.com]
> > 
> >    "If today is Sunday, I'm going to go shopping"
> > 
> > but rather:
> > 
> >    "If Sunday is today, I'm going to go shopping"
> 
> I don't think that anybody has had any trouble understanding me when I
> say 'If today is Saturday, then we are going to the concert house,

Not at all - this is exactly how I say things.

> otherwise we are going to hack.'  But then I have been writing
> if 0 == x for a long time as well.

The one doesn't follow the other. "If today is Sunday" corresponds to "if x
is 0" - "today" is the variable, "Sunday" is a constant.

Tim Delaney




From lists.ASkwar at DigitalProjects.com  Thu Apr 25 18:45:03 2002
From: lists.ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 00:45:03 +0200
Subject: Faster md5/1-way encryption?
In-Reply-To: <7xhem0vw83.fsf@ruckus.brouhaha.com>
References: <20020424215752.GC15371@teich.Garten.DigitalProjects.com> <3CC7317B.3040208@rogers.com> <mailman.1019718106.22365.python-list@python.org> <7xhem0vw83.fsf@ruckus.brouhaha.com>
Message-ID: <20020425224503.GH21205@teich.Garten.DigitalProjects.com>

?Paul Rubin? sagte am 2002-04-25 um 04:44:28 -0700 :
> I don't understand this.  Do you really want to accumulate a string
> containing 100,000 hash results (1.6 megabytes of hashes)?  Or do you
> want to hash all the lines together and get one result?

I want to get 100,000 hash results.  

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 17 hours 43 minutes




From tchur at optushome.com.au  Sat Apr  6 02:25:12 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sat, 06 Apr 2002 17:25:12 +1000
Subject: Booleans, integer division, backwards compatibility; where is 
 Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <3CAEA2D8.46D2F3CF@optushome.com.au>

Paul Rubin wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > No, that's "stagnant"...
> >
> > Stable is also a good adjective for it, since most of the older
> > code still runs unchanged, and Python from ten years ago is
> > still very recognizable to new coders today.
> 
> Stable doesn't just mean old scripts run without change in new
> interpreters.  It also means new scripts usually run without change in
> old interpreters.  That's what keeps breaking in Python.

I sometimes use a statistics package called Stata. It has been around
for 
well over a decade, with major upgrades to the core language released
about every two years or so - hence there are three or four versions in
common
use. In order to allow users who are still running Stata V4
to write routines which will behave correctly in Stata V7 (and vice
versa),
Stata has a "version" statement. If I am running Version 7, I can put
something like
"USE VERSION 4" (I don't recall the exact syntax) at the top of my
programs 
and the program will be executed as if it were running in Stata V4.
Hence, a user 
who still runs Stata V4 can put "USE VERSION 4" in every program and be
sure these 
will work correctly when run in all subsequent versions of Stata. 

This is the equivalent, I think, of having a "from __past__ import *"
statement 
in Python, or more precisely, a "from __v1.5.2__ import *" statement. 

Obviously it is impossible to retrospectively provide support for such a
statement,
but if there were a "from __v2.2.2__ import *" statement, one could put
that at the 
beginning of every Python module one writes and be certain that the Bool
type or any other
newfangled feature would not trip up one's code at any time in the
future, no matter how
recent the version of Python used to execute it. Similarly a user of a
recent version could 
verify that their code runs OK with older versions by using such a
statement, which would
of course disable all new language features and library modules, as well
as making any new 
language behaviour revert to the old.

However, it may not be possible to avoid bloat because successive
versions of Python would need
to support more and more old behaviour, but if this backwards support
could be relegated to
the libraries, it shouldn't be too bad. Bloat in the Python core in
order to support
historical behaviour would be a lot less desirable, though.  

And of course, such support for historical features and behaviour may
result in ever 
increasing intervals between releases of new versions of Python, as the
developers 
struggle to accomodate ever more historical baggage.

Hence I don't think that any of the above is a good idea.

Tim C




From bernie at 3captus.com  Mon Apr  1 16:53:33 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Mon, 01 Apr 2002 21:53:33 GMT
Subject: IDLE for SUSE Linux?
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com>
Message-ID: <3CA8D724.336A22BA@3captus.com>


Robert Oschler wrote:
> 
> I'm looking for a good of binaries, or a good URL concerning setting up IDLE
> under SUSE Linux 7.3 Pro.  Got URL?
> 

SuSE Linux 7.3 Pro comes with Python 2.1.1, so IDLE comes with it.  IDLE
is located at /usr/share/doc/packages/python/Tools/idle/idle.py.

> Also, if anyone has any experience working with Python under KDeveloper
> (KDE) I'd love to hear about it.  As soon as I get Glimpse/GlimseIndex
> installed I'm going to give it a whirl.
> 
> thx

As for GUI development I am using Tkinter but am in the process of
switching to Qt (Qt is the best GUI toolkit I've seen so far).  I am
impress with Qt Designer that comes with SuSE.  Since I needed only an
easy to use GUI designer.  I can then generate C++ implementation class
from it and use PyQt (comes with SuSE as well) to hook up with Python.

If you wanted a heayweight IDE, take a look at Black Adder
(http://www.thekompany.com/products/blackadder/).

The only issue I have with Qt is the distibution of run-time libraries
to Windows.  In order to do so it cost you US$ 400.


Bernie

-- 
There are three schools of magic.  One:  State a tautology, then ring
the changes on its corollaries; that's philosophy.  Two:  Record many
facts.  Try to find a pattern.  Then make a wrong guess at the next
fact; that's science.  Three:  Be aware that you live in a malevolent
Universe controlled by Murphy's Law, sometimes offset by Brewster's
Factor; that's engineering.
                -- Robert A. Heinlein


From greg at cosc.canterbury.ac.nz  Tue Apr  9 22:29:02 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 14:29:02 +1200 (NZST)
Subject: ANN: Pyrex - a language for writing Python extension modules
In-Reply-To: <15532.51903.323557.147303@12-248-41-177.client.attbi.com>
Message-ID: <200204100229.OAA00739@s454.cosc.canterbury.ac.nz>

Skip Montanaro <skip at pobox.com>:

> Interesting, but why not simply attach the type information to function
> attributes so the Pyrex code really is just Python?

I imagine that would look extremely ugly. Also, functions aren't the
only places that type information, and other kinds of information, is
needed in Pyrex. Have you read the Language Overview? There's much,
much more to Pyrex than just annotating things with types.

>    You can defer Pyrex compilation until you realize you need it.  Naively
>    compiling all your code to C makes for a lot of (object) code bloat and
>    doesn't necessarily speed anything up (if, for example, the code is I/O
>    bound or not executed frequently enough).  This was one of the problems
>    with Python2C.

You're not meant to compile all your Python code with Pyrex.
You only put Python code in a Pyrex module if it's needed to
interface with non-Python code. Pyrex isn't designed to speed
up Python code, but to bridge the gap between Python and
non-Python. Speed, if any, comes from the non-Python code
(Pyrex code operating on C data, or external C code).

Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+




From aleax at aleax.it  Wed Apr 10 10:37:04 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 14:37:04 GMT
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool type)
References: <mailman.1018446401.19134.python-list@python.org>
Message-ID: <k6Ys8.3581$vF6.110965@news2.tin.it>

Ken Seehof wrote:
        ...
>> I disagree.  For example, in C I can use a float to index into an
>> array and get it converted to an int implicitly and silently; in
>> Python, using a float to index into a list is a TypeError.  What
>> kind of compensating examples do you have in mind, where C would
>> be stricter than Python "for using the wrong type of number"?
> 
> Not with my compiler...
> 
>   float x = 1.0;
>   int a[4];
>   a[x] = 4;
> 
>   error C2108: subscript is not of integral type
> 
> Hmm.  I thought you could do that in python, but I was wrong too...

So was I -- I misremembered the way C is defined (I _thought_ the
inability to use float as indices was a C89 / C++ innovation, but
it appears it's in good old K&R too).

> You win.

Let's call it a draw, since we were both wrong.


> Well, after the grand numeric unification, (python 3.0 right?) you
> will be able to index by float if the value is exactly equal to
> an integer.  But C/C++ still will not permit it.

After said unification (if it ever happens -- from the PEP Parade
I'm not sure it will), there will basically be ONE "number type"
in Python.  How "weakly typed" will then be, if that single type
can be used wherever that single type can be used?-)


Alex



From guido at python.org  Mon Apr  8 12:47:20 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 08 Apr 2002 16:47:20 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com> <eppstein-CB6351.08534907042002@news.service.uci.edu> <ub315bi2fkdube@news.supernews.com>
Message-ID: <3CB1CA65.ECE0437C@python.org>

John Roth wrote:
> Unfortunately, I don't see a PEP on enums. Maybe I should
> dust off the rewrite of the set PEP I made that included them.

You could reuse some of the implementation techniques I used
for bool.  Then bool could later be retrofitted as an enum.
Like bool, I expect that the key benefit of enums is that
when you print them you'll get their name instead of their
value.  Currently I use string literals for that, but it's
not always optimal, and doesn't work for C APIs.

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


From jeff at ccvcorp.com  Fri Apr 12 15:15:35 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 12 Apr 2002 12:15:35 -0700
Subject: wxImage problem
References: <d5534716.0204121048.3a666b3c@posting.google.com>
Message-ID: <MPG.1720d21b7d478cb39896ad@news.nwlink.com>

In article <d5534716.0204121048.3a666b3c at posting.google.com>, 
eddieborder2001 at yahoo.co.uk says...
> how can i put an image into a wxPanel?
> i read on the net that the code is ...
> 
> wxImage_AddHandler(wxGIFHandler())
> gif = wxImage('bitmaps/image.gif', wxBITMAP_TYPE_GIF).ConvertToBitmap()
> wxStaticBitmap(panel, -1, bmp, wxPoint(10, pos),wxSize(bmp.GetWidth
> (),bmp.GetHeight()))

Here, you're loading the image and converting it to a wxBitmap, 
and assigning that bitmap to 'gif'.  Then you're creating a  
wxStaticBitmap, and passing it an identifier 'bmp' -- perhaps 
this is your problem?  

Other than that, and that I believe it's considered somewhat 
preferable to use 'wxBitmapFromImage( wxImage(...) )' rather than 
'wxImage(...).ConvertToBitmap()' as you do here, this looks right 
at first glance, but I haven't tested or compared with the 
docs...

If you're having any other problems, it's very helpful if you 
post the complete error messages that you're getting, or a 
specific description of the problematic behavior, so that we're 
not left trying to second-guess the issue...  

-- 

Jeff Shannon
Technician/Programmer
Credit International


From gashton at cmedltd.com  Tue Apr 23 05:18:41 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 23 Apr 2002 10:18:41 +0100
Subject: Mass Mailing help needed (Paul Rubin)
In-Reply-To: <MPG.172e025d803fc2009896bf@news.nwlink.com>
References: <mailman.1019325819.29016.python-list@python.org> 
	<MPG.172e025d803fc2009896bf@news.nwlink.com>
Message-ID: <1019553521.574.12.camel@tp6>

On Mon, 2002-04-22 at 20:20, Jeff Shannon wrote:

> I would avoid directly invoking sendmail at all.  My mailer uses 
> smtplib to contact the mail server, and this *seems* fairly 
> efficient.

With that approach you have to open one TCP connection per message. This
is expensive, especially as your SMTP daemon should be doing all sorts
of checks on your IP to see if it'll happily accept mail from you. Not
only that, if you want to send a lot of messages you'll soon run out of
TCP ports on most unix boxes, which can have nasty side effects.

Calling sendmail directly is like giving it an intravenous injection.
Sending messages in via TCP port 25 is like giving it a couple of pills
and a sip of water.

> Since smtplib functions at a level of individual messages ....

You'll run out of connections in no time. I've never had to send more
than 10,000 messages in a customer mailing, but I always spread them out
over quite a long period of time (i.e. four hours or so) in order to
reduce the negative effect on our servers. Optimising it too much was
never really a priority, though I'm sure it would have been possible.

-- 
Graham Ashton





From geoff at gerrietts.net  Tue Apr 16 18:05:04 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Tue, 16 Apr 2002 15:05:04 -0700
Subject: Safe eval?
In-Reply-To: <a9i4ms$ece$1@nets3.rz.RWTH-Aachen.DE>
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <a9i327$cvi$1@nets3.rz.RWTH-Aachen.DE> <mailman.1018991313.16792.python-list@python.org> <a9i4ms$ece$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <20020416220504.GE20804@isis.gerrietts.net>

Quoting Philipp Lenssen (lenssen at hitnet.rwth-aachen.de):
> "Geoff Gerrietts" <geoff at gerrietts.net> wrote in message
> news:mailman.1018991313.16792.python-list at python.org...
> > Quoting Philipp Lenssen (lenssen at hitnet.rwth-aachen.de):
> > > I just realized, as opposed to PHP, there's much stricter namespaces in
> > > Python and maybe not using "import" on anything harmful could be the
> > > solution?
> >
> > You can hack on the rexec module to make it do what you want to do;
> > that's what I do. It may or may not provide ample security for your
> > purposes; it's certainly better than a raw eval().
> >
> > The code I use is attached; under 2.something,
> 
> Thanks! I will try out your class and simply use it if it does what I want.
> Please email me if you a) don't want me to use it xor b) don't want to get
> credited in the source file or the http://questml.com webpage.
> 
> > this breaks unless you
> > override make_re to pass.
> >..
> 
> Could you explain that? I didn't understand...

Sorry, I went and did it in the code I sent, then forgot to remove the
straggler lines from the message. Under 2.1 or 2.2 (I forget which, I
think 2.2) the RExec class calls make_re() in its __init__, and
make_re() breaks if sre isn't in the list of allowed imports.

I'd give some serious consideration to Andrew Dalke's solution, too.
That may be a quicker route to exactly where my complicated route gets
you -- I think that when I wrote this, I'd tried eval(string, {}, {})
and never even considered replacing the __builtins__ namespace with an
empty dictionary.

I'm considering refactoring, myself....

--G.

-- 
Geoff Gerrietts            "Democracy is a form of government that substitutes 
geoff at gerrietts dot net  election by the incompetant many for appointment 
http://www.gerrietts.net/   by the corrupt few." --George Bernard Shaw




From maxx at easynews.com  Wed Apr 17 14:53:44 2002
From: maxx at easynews.com (maxx at easynews.com)
Date: Wed, 17 Apr 2002 18:53:44 GMT
Subject: need help to choose an appropriate database
References: <fac97694.0204170759.35b66092@posting.google.com>
Message-ID: <u8hrbuklva06jnfpa3hdarbe7al0s6lql1@4ax.com>

On 17 Apr 2002 08:59:20 -0700, alexandre.herbiniere at caramail.com (alexandre)
wrote:

>I use zope to access to a database, but I have to choose it.
>I mustn t store a lot of data so should i choose MySQL, PostGreSQL or
>anther one ?
>Which one is the most efficient ( speed...)

Look into SAPDB (http://www.sapdb.org)


From aleax at aleax.it  Wed Apr 10 06:34:22 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 10:34:22 GMT
Subject: static, dynamic, strong and weak (was RE: PEP 285: Adding a bool type)
References: <mailman.1018392859.8707.python-list@python.org>
Message-ID: <OyUs8.2398$b62.59604@news1.tin.it>

Ken Seehof wrote:
        ...
> as a float, int, longint, or bool, depending on context.  All numbers
> of equal value are pretty much interchangable (this becomes more true
> as we add more numeric unification features).  In c/c++ numbers are
> more strongly typed, in that there are many ways to get error messages
> for using the wrong type of number without explicit conversion.

I disagree.  For example, in C I can use a float to index into an
array and get it converted to an int implicitly and silently; in
Python, using a float to index into a list is a TypeError.  What
kind of compensating examples do you have in mind, where C would
be stricter than Python "for using the wrong type of number"?


> The fact that an object's type doesn't actually change seems like an
> implementation detail to me.  Numbers in python -feel- as if they are
> weakly typed.

Not to me.  Even in Haskell I can write 1+2.0 - and Haskell is the
strongest-typed and feeling-strongest-typed GOOD language I know
(let's PLEASE forget that unnamed language where you must use + only
to add integers and +. to add floats... have they changed it yet?-).

The first time I (accidentally) used a float as a list index and
got a TypeError has impressed Python on me as strongly (albeit dynamically)
typed, better than any other study or experience so far.


Alex



From loewis at informatik.hu-berlin.de  Sun Apr 14 10:42:44 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 14 Apr 2002 16:42:44 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <slrnabj2g8.1gj.grante@tuxtop.visi.com>
Message-ID: <j48z7qpcij.fsf@informatik.hu-berlin.de>

grante at visi.com (Grant Edwards) writes:

> Point is that "^" operates bitwise, and there is no
> exclusive-or operation for "truth" values.

That depends on the argument types. If both arguments are of type
bool, ^ performs exclusive-or.

Regards,
Martin



From newsfroups at jerf.org  Mon Apr 22 20:07:52 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Tue, 23 Apr 2002 00:07:52 GMT
Subject: Developing and ircd with python
References: <aa278j$8c8$1@nsnmrro2-gest.nuria.telefonica-data.net>
Message-ID: <3CC4A648.3070100@jerf.org>

Raphael Ribeiro wrote:
> Is it possible?

Yes.



From mvcc_cr at yahoo.com  Sat Apr 13 19:37:12 2002
From: mvcc_cr at yahoo.com (Marco Caamano)
Date: 13 Apr 2002 17:37:12 -0600
Subject: httplib drops ssl connection to MS IIS Server
Message-ID: <1018741036.919.2.camel@prowler>

Hi, I am new to Python and new to the list.

I've been experiencing some problems with ssl connections using httplib.
I have the following function:
----------------
def testurl2():
        print '\n\nUsing HTTPS request'
        body='\202test'
        host=("209.154.200.218:443")
        hs=httplib.HTTPS()
        hs.set_debuglevel(4)
        hs.connect(host)
        hs.putrequest('POST','/scripts/gateway.dll?Transact')
        hs.putheader('Content-Type','x-Visa-II/x-auth')
        hs.putheader('Content-Length', str(len(body)))
        hs.endheaders()
        hs.send(body)
        status, reason, header= hs.getreply()
        data=hs.getfile().read()
        print data
        print status
        return
----------------

When executed it connects to the server and then drops on the following 
error:

----------------
Using HTTPS request
send: 'POST /scripts/gateway.dll?Transact HTTP/1.0\r\n'
send: 'Content-Type: x-Visa-II/x-auth\r\n'
send: 'Content-Length: 5\r\n'
send: '\r\n'
send: '\x82test'
Traceback (most recent call last):
  File "test.py", line 4, in ?
    misc_func.testurl2()
  File "misc_func.py", line 180, in testurl2
    status, reason, header= hs.getreply()
  File "/usr/lib/python2.2/httplib.py", line 728, in getreply
    response = self._conn.getresponse()
  File "/usr/lib/python2.2/httplib.py", line 570, in getresponse
    response = self.response_class(self.sock, self.debuglevel)
  File "/usr/lib/python2.2/httplib.py", line 98, in __init__
    self.fp = sock.makefile('rb', 0)
  File "/usr/lib/python2.2/httplib.py", line 607, in makefile
    buf = self.__ssl.read()
socket.sslerror: (5, 'EOF occurred in violation of protocol')
----------------

I like to note that this happens when I try to connect to this specific
server, it is a MS IIS Server. When I use the same funtion on my local
Apache server it works fine. Since I don't have access to the IIS Server
changing it is not an option.

I've searched the web for similar situations but have found no
references so far. 

I'll really apreciate a pointer here, what I am doing wrong? what should
I try next?

thanks!


Marco Caama?o C.





_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com





From loewis at informatik.hu-berlin.de  Wed Apr  3 12:07:36 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Apr 2002 19:07:36 +0200
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <a8eve0$kuo$1@perki.connect.com.au> <3CAB16BA.EEED07DE@sympatico.ca>
Message-ID: <j48z84zp5z.fsf@informatik.hu-berlin.de>

vio <vmilitaru at sympatico.ca> writes:

> From a technical point of view, I don't see anything magical or
> impossible about a java2python tool. I am aware of at least a
> software product that does it, but it's a commercial offering. This
> comercial tool's existance is proof of technical feasibility. I also
> have coded myself an early java2py translator, in python. But it was
> cheezy, and implements only a fraction of the java grammar.  I'm not
> a java programmer, and I do not want to spend much time to learn it
> either (again!) to create rules for the translation.

Translating the source code on a statement-by-statement basis might
indeed be possible. However, the resulting Python program will hardly
work, as you need all the Java libraries that it uses. Furthermore,
naive translation may lose semantic information: e.g. you have
overloading in Java but not in Python. So information which operation
to call can easily be lost during translation.

Regards,
Martin



From sholden at holdenweb.com  Tue Apr 30 08:56:49 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 08:56:49 -0400
Subject: Global Variable
References: <aalpqc$h9s$1@pegasus.csx.cam.ac.uk>
Message-ID: <Sywz8.13212$xy4.221@atlpnn01.usenetserver.com>

"Xiao-Qin Xia" <xx758 at cam.ac.uk> wrote in message
news:aalpqc$h9s$1 at pegasus.csx.cam.ac.uk...
> Hello,
>
> If a module (for example, gmpy) is imported in the main.py, how can it be
> used by other modules (for example, mymodule) without being imported by
> each of the other modules?
>
Let me count the ways ... but the *real* answer is, just import it into the
other modules as well! This will hardly take any time, since the code of the
module is only executed when it is first imported (or on subsequent use of
the relaod() function).

> #---------main.py---------
> import gmpy
> import mymodule
> ...
>
So, for example, you can get at gmpy in your other modules as __main__.gmpy.
Butt-ugly, isn't it?
>
> #--------mymodule.py---------
> #I want use gmpy here, do I have to import it again? can main.py add gmpy
> into a global dict?

It actually *does* add gmpy to the sys.modules dict:

>>> import sys
>>> for m in sys.modules:
...     print m
...
stat
posixpath
UserDict
signal
site
__builtin__
sys
posix
__main__
exceptions
readline
os
os.path
>>>

In practice it's much easier to say

    import mymodule

than it is to say

    import sys
    mymodule = sys.modules["mymodule"]

and there isn't even any excessive cost, so just go ahead and import the
modules you need!

A common technique for functions which aren't always going to be called is
to perform the imports inside the function. That way the import is only
performed if the function gets called, and the additional per-call imports
really don't add much overhead.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From sholden at holdenweb.com  Mon Apr 15 13:15:14 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 13:15:14 -0400
Subject: wich odbc is stable ?
References: <20020413141754.6d14e50d.wilk@flibuste.net>
Message-ID: <IWDu8.13736$Ji7.3891@atlpnn01.usenetserver.com>

"William Dod?" <wilk at flibuste.net> wrote ...
> hi,
>
> I need to connect to an MSAccess database for simple query (no lock, no
> transaction), so i would like to know if the standard odbc provided by
> win32extension is enough stable even if it's written to be no supported...
> Or must i use callDll or ado ?
>
> thanks to read my so bad english writing
>
As Dale Strickland-Clark has already suggested, mxODBC might be a more
comprehensive alternative.

The main mistake made with the odbc module is omitting the import of the
associated dbi module.
If you always use

    import dbi, odbc

the chance of errors will be reduced.

Please don't bother about your English - at least I can read that, which I
can't do when someone posts in (say) German. Those questions I must leave to
the more competent linguists :-)

regards
 Steve






From aahz at pythoncraft.com  Wed Apr 10 11:43:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 11:43:59 -0400
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <a90rv2$fvh$1@slb4.atl.mindspring.net>
Message-ID: <a91mjv$112$1@panix1.panix.com>

In article <a90rv2$fvh$1 at slb4.atl.mindspring.net>,
Andrew Dalke <dalke at dalkescientific.com> wrote:
>Emile van Sebille:
>>
>>Oooh... that's almost as much fun as saying 'xyzzy' the first time...
>
>I just said it, but nothing happened here.

Well, duh!  You need to be in the cave.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From paul at palydon.demon.co.uk  Tue Apr  9 04:01:38 2002
From: paul at palydon.demon.co.uk (paul at palydon.demon.co.uk)
Date: Tue, 09 Apr 2002 09:01:38 +0100
Subject: Off topic C question
Message-ID: <E16uqZO-000Ec4-0C@finch-post-12.mail.demon.net>

> I know this is way off topic, but I figure the heavy hackers here will
> understand my mindset better than those at comp.lang.C
>     I have been doodling with python (nasty habit I know but I have to
> do somthing with my snake) and think I may be ready to learn a second
> language.  I am leaning towards C over C++ and Java even though I
> loose objects it will ground me firmly in functional programing.  I am
> wondering what book would be the best for a begginer?  The two I have
> seen that look interesting are
> C for Dummies (parts one and two) or 
> The C Programming Language
> by Brian W. Kernighan, Dennis M. Ritchie.

Have a look at http://www.accu.org/bookreviews/public/index.htm , Beginners C






From bkc at Murkworks.com  Tue Apr  9 17:17:42 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 9 Apr 2002 17:17:42 -0400
Subject: property() type that only protects set?
References: <3cb34fb8$1_4@goliath.newsgroups.com> <3CB357E3.95EFC8E1@metaslash.com>
Message-ID: <3cb358cb$1_2@goliath.newsgroups.com>

If you leave out fget (that is, let it default to None) then you have an
attribute that cannot be read.

This isn't what I want.. I want an attr that can be read without a special
getter function

(something like

myprop = property(self.__getattr,my_set)

But of course this syntax doesn't work.



--
Novell DeveloperNet Sysop #5

_
"Neal Norwitz" <neal at metaslash.com> wrote in message
news:3CB357E3.95EFC8E1 at metaslash.com...
> Brad Clements wrote:
> >
> > Anyone have an implementation of property() that doesn't require a
getter,
> > only a setter (and optional del)
> >
> > I only need to protect set operations, and want to allow get to go
straight
> > to __dict__
> >
> > (or object when I'm using slots)
>
> From the docstring:
>
> >>> print property.__doc__
> property(fget=None, fset=None, fdel=None, doc=None) -> property attribute
>
> fget is a function to be used for getting an attribute value, and likewise
> fset is a function for setting, and fdel a function for del'ing, an
> attribute.  Typical use is to define a managed attribute x:
> class C(object):
>     def getx(self): return self.__x
>     def setx(self, value): self.__x = value
>     def delx(self): del self.__x
>     x = property(getx, setx, delx, "I'm the 'x' property.")
> --------------------------------------------------------------------------
--
>
> so use:
>
>     x = property(fset=setx)
>
> Neal




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From emile at fenx.com  Tue Apr  9 23:54:07 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 10 Apr 2002 03:54:07 GMT
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org>
Message-ID: <zHOs8.201693$Yv2.65602@rwcrnsc54>

Kalle Svensson
>
> "import this", IIRC.
>


Oooh... that's almost as much fun as saying 'xyzzy' the first time...
;-)

Plugh!-ly y'rs

--

Emile van Sebille
emile at fenx.com

---------



From peter at engcorp.com  Tue Apr  9 09:38:46 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 09:38:46 -0400
Subject: how to read hardware info on Win NT
References: <3cb2e686.21131515@news.easynews.net>
Message-ID: <3CB2EEE6.69A86E06@engcorp.com>

Marcus Stojek wrote:
> 
> Hi,
> 
> could anyone please give me a hit how to get hardware information with
> Python? I need the id of my ethernet card and my harddisk.

Can you tell us how you would do it without Python?

Then we can show you the way to accomplish the same with Python.


From aleax at aleax.it  Thu Apr 11 05:03:06 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 09:03:06 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204100552.6b822307@posting.google.com> <C%Xs8.3139$b62.77265@news1.tin.it> <a923g0$ks2$1@news.island.liu.se>
Message-ID: <ejct8.5680$b62.139324@news1.tin.it>

Thomas Bellman wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> 
>> Once you're using (e.g.) iterators in your package, or subclassing
>> a standard class, there's no extra cost in having the modules of
>> your package use (e.g.) generators or true-division as well, so
>> why shouldn't a package author freely do so?
> 
> For the specific case of iterators, it is fairly easy to add a
> backwards compatible interface to your classes, so they work even
> in an older interpreter:
        ...
> Then just let your classes inherit from IteratorEmulator.  This,
> of course, assumes that you don't need __getitem__() for other

Too big an assumption, IMHO - it rules out all those of your
classes which are mappings (maybe half of your containers).

And that's not all -- you may be USING iterators quite apart
from EXPOSING iterators for your classes.  "for x in adict:"
is faster as well as smoother than "for x in adict.keys():".

That's one more backwards-compatibility kludge you'd need to
build -- e.g. your own iter function when there's no builtin
one.  Oh, and, your own dict class when there's no builtin 
dict type -- can't just use UserDict if anywhere you are
calling dict with a sequence of pairs (not strictly speaking
an iterator issue, but close).

> purposes, or that you are returning infinite sequences.
> 
> Generators, inheriting from builtin types, and true division, are
> not as easy to get running in an older interpreter, so if you are
> using one of *them*, you might as well use iterators too.

Wrapping all of your meant-to-be-true divisions into your own
function truediv(a,b) is no klunkier than wrapping all of
your iterations-over-dictionaries into iter(somedict) (with
your own implementation of iter if it ain't builtin).   Both
are clunky aplenty.  Depending on what builtin type you are
inheriting from, that may be easier (UserDict in lieu of
dict, for example -- that's pretty easy to arrange in some
backwards-compatibility preable) or hardish (defining file as
an inheritable class that autodelegates to a file object) --
in either case arranging to pass your objects to the Python
framework can be a real problem (some pieces want true file
object, true dictionaries, and the like -- admittedly even
in 2.2 some internal uses bypass your method overrides and
do the equivalent of dict.__getitem__(theobj, thekey) and
so on:-).


Alex



From cmueller at onebox.com  Wed Apr 10 15:15:39 2002
From: cmueller at onebox.com (rockmuelle)
Date: 10 Apr 2002 12:15:39 -0700
Subject: IClassFactory2, Python and run-time licensing
Message-ID: <aaa059f1.0204101115.76a26f4e@posting.google.com>

I've just finished developing an application in Python that uses a 3rd
party ActiveX control.  The control requires a design-time license
which I have purchased and a run-time license, created through
IClassFactory2.RequestLicKey and embedded in the distributed code.  It
uses IClassFactory2.CreateInstanceLic to provide the run-time instance
of the control.

pythoncom does not support this interface and doesn't appear to expose
any functions for creating licensed ActiveX controls.

Has anyone encoutered this situation before and come up with a
solution?

Thanks.

-Chris


From peter at engcorp.com  Sun Apr 14 23:38:51 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 14 Apr 2002 23:38:51 -0400
Subject: split() off last substring
References: <3CBA41CB.27DCA084@uniserve.com>
Message-ID: <3CBA4B4B.797E5275@engcorp.com>

bvdpoel at uniserve.com wrote:
> 
> Can I tell split() to split off the LAST seqment? I guess I want
> something like:
> 
>         a="aaa.bbb.ccc"
>         b,c=a.rsplit(1)
>         print b,c
>         # "aaa.bbb", "ccc"

How about this:

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
>>> a = 'aaa.bbb.ccc'
>>> a.split('.')
['aaa', 'bbb', 'ccc']
>>> b = a.split('.')[:-1]
>>> b = '.'.join(b)
>>> c = a.split('.')[-1]
>>> b,c
('aaa.bbb', 'ccc')

-Peter


From jaf at lcsaudio.com  Sun Apr 28 12:31:17 2002
From: jaf at lcsaudio.com (jaf)
Date: Sun, 28 Apr 2002 09:31:17 -0700 (-0700)
Subject: struct module - '<q' equivalent for old python?
Message-ID: <42079966076-BeMail@sugoi>

Hi all,

In my application, I need to be able to read and write a 64-bit integer
(in little-endian format) from/to a file.  For Python 2.2, the struct
module has a nice '<q' datatype argument that will do just that, but it
isn't implemented in older versions and throws an exception.

Does anyone know of a good alternative method I could use as a fallback
in this case?  Some of the platforms I need to support don't have
Python2.2 readily available, so I'd like to be able support older
versions of Python, if possible.

Thanks,
Jeremy





From boud at valdyas.org  Fri Apr 19 09:06:59 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Fri, 19 Apr 2002 15:06:59 +0200
Subject: Is there a PyQT 'send'?
References: <d906b2c1.0204190350.2d5ae684@posting.google.com>
Message-ID: <3cc01920$0$41693$e4fe514c@dreader4.news.xs4all.nl>

Moray Taylor wrote:

> Hi,
> 
> Is there an equivalent command to Tk's 'send' in Python/QT?
> 

What does Tk's send do again... Page, page, page, 438 in Grayson, ah! Yes. 
No -- I don't think so. Every Qt app is on its own, and there's nothing 
that I know of that can send a command from one app to another. 

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


From paulsid at shaw.ca  Tue Apr 23 13:46:14 2002
From: paulsid at shaw.ca (Paul Sidorsky)
Date: Tue, 23 Apr 2002 11:46:14 -0600
Subject: non-blocking sockets
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>
Message-ID: <3CC59DE6.115129A4@shaw.ca>

so very tired wrote:

> When trying to read from a socket, I don't want it hanging, so I set it to
> non-blocking by calling
> 
> setblocking(0)
> 
> but then when I try to read from it and there is no data in the socket, I
> get an error. I tried this in both windows and linux and in both I get an
> error and the program stops. It says there's an exception but it doesn't
> say the name of the exception so I can't even try to catch.
> Does anyone know how I can check if there's data coming from a socket
> connection without hanging or having the program barf on me?

The error you need to catch is called "error" in the socket module, i.e.
socket.error.  It's described at the top of section 7.2 in the library
reference.  

Depending on what you're doing you might want to check the errno to see
that the problem is only that there's no data available and not
something more serious.  A blanket "except socket.error:  pass" will
work well enough to get you started, though.

BTW for anything heavy-duty, using select as others have recommended is
probably a better solution.

-- 
======================================================================
Paul Sidorsky                                          Calgary, Canada
paulsid at shaw.ca                        http://members.shaw.ca/paulsid/




From starx at pacbell.net  Tue Apr  9 06:33:14 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 9 Apr 2002 10:33:14 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <23891c90.0204080156.27f36d22@posting.google.com> <3CB1CEC0.F324F480@python.org> <23891c90.0204090012.7fc80c1f@posting.google.com>
Message-ID: <Xns91EB245A0BCD9RASXnewsDFE1@130.133.1.4>

Paul Boddie || Tue 09 Apr 2002 01:12:17a:

> Guido van Rossum <guido at python.org> wrote in message
> news:<3CB1CEC0.F324F480 at python.org>... 
>>
>> Paul wrote:
>> >
>> > As far as the 2.2 features are concerned, they seem to have a
>> > reputation in certain circles as being badly-defined, immature and
>> > "best steered clear of until later".
>> 
>> Huh?  I'm not sure where you heard that.  Alex Martelli himself just
>> told me that he is using it and considers it stable.  Maybe the
>> people who say this themselves deserve a reputation as immature and
>> "best steered clear of"?
> 
> Perhaps their views were tainted by the need to support releases of
> Python lower than 2.2. I'm sure that there are many nice features
> which reduce the amount of code (along with the complexity of the
> code) substantially in Python 2.2, but it isn't always fair to expect
> everyone to upgrade Python every time someone releases a new piece of
> software.
> 
> Even though I use 2.0 and only have 2.1 to hand because an increasing
> number of packages required it over time (most of which I only
> evaluated anyway), I couldn't see why I also had to install 2.2 when
> it was released because the odd package required it, arguably on an
> unnecessarily frivolous basis.
> 
> I suppose one argument against a "super upgrade" is that backwards
> compatibility is readily abandoned (by package authors), but then,
> regardless of the scale of the differences between releases, I would
> estimate that there is still a fair amount of coding for compatibility
> on a number of the larger Python projects. (Although PyXML's support
> for Python 1.5.2 is possibly too heroic to be practical any more.)

Huh? Your argument is all well and good, but what does supporing 
releases earlier than 2.2 have to do with it's new features being badly 
defined and immature themselves?

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From kybic at ieee.org  Thu Apr  4 12:26:57 2002
From: kybic at ieee.org (Jan Kybic)
Date: 04 Apr 2002 19:26:57 +0200
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <101f01c1dbe6$41a555c0$6b01a8c0@ericlaptop>
References: <mailman.1017795845.3464.python-list@python.org>
	<r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
	<lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
	<lbh8z84nake.fsf@xihu.inria.fr>
	<15531.7864.698585.485436@12-248-41-177.client.attbi.com>
	<lbh7knolreu.fsf@xihu.inria.fr>
	<15531.9709.924953.953487@12-248-41-177.client.attbi.com>
	<mailman.1017850275.2134.python-list@python.org>
	<0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop>
	<lbh8z83kiez.fsf@xihu.inria.fr>
	<101f01c1dbe6$41a555c0$6b01a8c0@ericlaptop>
Message-ID: <lbhy9g3gyse.fsf@xihu.inria.fr>

> >criticism remains: You have to code your function twice, first in
> >Python for prototyping, then in C. But I agree that weave saves you a
> >lot of work.
> 
> If this is the only criticism, than we're doing quite well. :-) The
> alternative to Python right now if you need speed is to write your
> entire program (or extension) in C/C++.  Of course, then the complaint

Well, not the entire program. If possible, I am gradually
reimplementing Python functions in C one by one, creating interface
with Swig to them. This way, I can always just plug in the old one and
the new one, to verify that the behaviour rests the same.


> becomes "Man, it sure does run fast, but I sure wish C had dynamic
> typing and didn't have all this memory management stuff." ;-) When I

Of course. :-)

> That said, I just saw another project similar to weave come across the
> line on python-dev called Pyrex.  very cool.  It allows C typing in

I have just had a look. Very cool indeed! Actually that's about what I had
in mind when I called for a Python compiler. Once it compiles a little
more of Python (notably classes),  ideally any valid Python code, it should
definitely become a tool of choice for people wanting to speed up
their Python code. An interface to Numeric should not be difficult.
Greg, please keep me posted about the new versions.

> a look at Pysco, Pyrex, and weave, choose the one you like, and pitch
> in!  We all know that a faster Python is desired.  It even looks very

Indeed. Another well worth project on the list.

J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/




From mhammond at skippinet.com.au  Tue Apr 16 00:19:02 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 16 Apr 2002 04:19:02 GMT
Subject: problem running win32com.client application from commandline
References: <3hrjbu4qv9b9afjhr2nhnef025350k4lpd@4ax.com> <3CBA8463.30703@skippinet.com.au> <a9e8si$47q$1@crusher.de.colt.net>
Message-ID: <3CBBA67F.6050209@skippinet.com.au>

Tino Lange wrote:


> Is there a list of those useful commandline-switches for PythonWin 
> available? A "PythonWin.exe /?" or "PythonWin.exe --help" does not help.

Nope - look in pythonwin\pywin\framework\intpyapp.py

> Just one more question - it seems that this OCX I want to script via python 
> has a form attached, that is filled under some circumstances - the return 
> value of the called function via COM is then not enough for me. 
> 
> From looking at the Word-Macro that uses the OCX originally I can see that 
> there are some global variables shared between Word VBA and the OCX. Also 
> there's this form that s the same in Word and in the OCX. Those global 
> variables and the form are not accessible via the COM-Interface, but the 
> Word Macro seems to have access. (at least the makepy generated file does 
> not list them).
> 
> Is there a way to get the form-handle and have access to the form and for 
> example some textual content?

Sorry, I have no idea - that is completely up to the OCX.

Mark.



From theivo at abv.bg  Sun Apr 14 04:19:23 2002
From: theivo at abv.bg (I.C.)
Date: 14 Apr 2002 01:19:23 -0700
Subject: dos promt simulation
References: <6ac1ecad.0204130435.93d0ad3@posting.google.com> <Xns91EFBDC8F2FCcliechtigmxnet@62.2.16.82> <6ac1ecad.0204131339.56928640@posting.google.com> <a9aqr4$moh$0@216.39.172.122>
Message-ID: <6ac1ecad.0204140019.54fff51@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<a9aqr4$moh$0 at 216.39.172.122>...
> On 13 Apr 2002 14:39:58 -0700, theivo at abv.bg (I.C.) wrote:
> 
> >> there is also popen2, popen3 etc. they give you a file for stdin where you 
> >> can write to.
> >
> >Maybe popen2\3 could help, but I think, I don't use them correctly :
> >
> >import os
> >dos_input, dos_output = os.popen2('time')
> >
> >for line in dos_output.readlines(): # Current time is 22:14:43,24
> >        print line                  # Enter new time:
> >
> >dos_input.write(raw_input())        # User enters the new time
>   ^^^^^--??
> >
> >Where is my mistake?
> >
> I would guess that's probably one.
> 
> Regards,
> Bengt Richter

Can somebody give me an example, which uses popen2?
Thank you!

Ivo


From dale at riverhall.NOTHANKS.co.uk  Tue Apr  2 04:21:21 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 02 Apr 2002 10:21:21 +0100
Subject: The fastest XML parser around
References: <mailman.1017654423.12312.clpa-moderators@python.org>
Message-ID: <vatiauojb7geb1l9jbij95agp9rgjn78vp@4ax.com>

"Andy Robinson" <andy at reportlab.com> wrote:

<snip>

>XML document into memory, and validated, as quickly and efficiently
>as possible.   You can parse and validate Hamlet in a tenth of a second on a
>standard PC.
>

This looks really promising, Andy. I'll take a look at it shortly.

However, your announcement begs the question: 
What's a standard PC? :-) 

Also, your discovery of Hamlet in XML is going to cause some real
problems for the historians.

--
Dale Strickland-Clark
Riverhall Systems Ltd


From scottrharris at earthlink.net  Mon Apr 29 12:51:33 2002
From: scottrharris at earthlink.net (Scott Harris)
Date: 29 Apr 2002 09:51:33 -0700
Subject: Operator overloading is much slower than a method call?
Message-ID: <c2f51c49.0204290851.2061078a@posting.google.com>

I did some simple timing experiments last night that seem to indicate that
operator overloading is about 35% slower than a method call on an object.

Here's what I was doing:

I have a class Vector that implements standard 3-vector operations like
dot products and additions, etc...

If a and b are vectors, I can write

c=a.add(b) #approach 1
  
or 

c=a+b #approach 2

where the overloaded + operator simply calls the add method.

Anyhow, I find that second approach is about 35% slower on my machine than the
direct method call.


I timed this by doing about 60K vector additions with the only change being
the direct call to the method versus the operator approach.

I'm pretty suprised by these results. I had only expected a slight performance
hit. Since my code makes pretty extensive use of overloading for readability
I guess I'm taking a big hit.

Is this a well known result? Is there anything that can be done about it?
Is operator overloading really that much more work for the interpreter to
figure out?

Comments would be appreciated.

I'm using python2.2 on Linux. 

Thanks,
-Scott Harris
scottrharris.earthlink.net


From aleax at aleax.it  Thu Apr  4 00:46:03 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 04 Apr 2002 05:46:03 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain> <a8akqh$8bk$0@216.39.172.122> <a8gk26$b16$0@216.39.172.122>
Message-ID: <vMRq8.57284$S52.1995953@news2.tin.it>

Bengt Richter wrote:
        ...
>>in the C code (and I haven't checked), then __nonzero__() (or
>>__len__()) is called if defined.  This gives us the control we're
>>looking for.
> 
> Well, I wouldn't want to mess with __len__ if I e.g., were subtyping str

Of course not.  But __nonzero__ exists for the SOLE purpose of
defining the object's truth value.

> for something where I wanted to be able to write
> 
>     if s: print '"%s" (length %d) is special.' % (s, len(s))
> 
> I.e., in a logical context( if x..., x and ..., etc.), I think
> x.__bool__() should have priority over x.__nonzero__() or x.__len__().

__nonzero__ is used for nothing else except obtaining the object's
truth value.  You don't appear to understand this...?

>>I certainly don't object to implementing the bool type.  I guess I
>>just don't see it as a significant advantage.
> 
> I'd like the semantic clarity. And I'd like to define a class or subtype
> whose boolean logical value doesn't depend on anyones concept of Nothing
> or Empty if those don't fit. 

You CAN do that, TODAY.  Therefore, your major motivation for
desiring booleans seems to be an insufficient knowledge of Python/

> use when I poll. Or take a whimsical example: What if I want to subclass
> int and be able to say if n: print n,'is a prime', expecting that
> n.__bool__() will be called so I can calculate primeness and cache the
> result in a boolean attribute for faster return next time, etc.

You CAN do that, TODAY.  Whether it's wise or not is another issue,
but, for example:

class intWithPredicate(int):

    def __init__(self, value=0, predicate=None):
        int.__init__(self, value)
        self.setPredicate(predicate)

    def setPredicate(self, predicate=None):
        self.predicate = predicate
        self.isResultKnown = 0

    def __nonzero__(self):
        if self.isResultKnown: 
            return self.result
        if self.predicate is None:
            self.result = self != 0
        else:
            self.result = self.predicate(self)
        self.isResultKnown = 1
        return self.result

You don't need any change to Python to implement this functionality,
whether it makes sense or not to have the functionality in question.


> The point is to have freedom to use a useful general mechanism without

The point is you HAVE the freedom, NOW -- the mechanism is
general and there are no semantic contortions.

> semantic contortions, or giving up the ability if __nonzero__ and/or
> __len__ already have uses for the class.

What "use for the class" do you expect __nonzero__ to have, as it
is ONLY used to define the object's truth value?!


Alex



From donn at u.washington.edu  Wed Apr 17 13:51:58 2002
From: donn at u.washington.edu (Donn Cave)
Date: 17 Apr 2002 17:51:58 GMT
Subject: UTC equivalent of time.mktime() ?
References: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com> <mailman.1019062773.24881.python-list@python.org>
Message-ID: <a9kcnu$11v6$1@nntp6.u.washington.edu>

Quoth Graham Ashton <gashton at cmedltd.com>:
| On Wed, 2002-04-17 at 17:11, Mark McEahern wrote:
|> Why not just use the mxDateTime library and just do something like this:
|> 
|> 	mx.DateTime.utc().ticks()
|
| I appreciate that I could (and thanks for the suggestion), but does this
| really mean that via the standard library there is no easy way to
| convert "1 Jan 1970" to 0 (i.e. string to seconds since epoch)?

I have not been following, so pardon me if this has already been
suggested, but on a platform where you have 2.0 or later and strptime,
this might do that:

  lt = time.mktime(time.strptime(sys.argv[1], '%d %b %Y'))
  utoffset = time.mktime(time.gmtime(lt)) - lt
  print lt - utoffset

	Donn Cave, donn at u.washington.edu


From phr-n2002a at nightsong.com  Wed Apr  3 01:46:00 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 22:46:00 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com> <mailman.1017795008.700.python-list@python.org>
Message-ID: <7xzo0lth3r.fsf@ruckus.brouhaha.com>

Richard Jones <rjones at ekit-inc.com> writes:

> For a _practical_ example of the "markup", please see the Roundup
> documentation mentioned in my response. I think you'll find that the
> original text is perfectly readable and significantly easier to edit
> and maintain that any "explicit" markup language. Apologies for the
> long
>   
> http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/roundup/roundup/doc/installation.txt?rev=HEAD

That looks pretty similar to an Info file (texinfo output).  I think
I'd rather type "@section{Overview}" than have to supply those special
underscores directly.

> Note that my manual Table of Contents at the top is now possible using a 
> directive::
> 
>    .. contents:: **Table of Contents**
> 
> The output is at http://roundup.sf.net/doc/installation.html

So the ink isn't even dry on this PEP and already it's sliding down
the slippery slope into explicit markup <wink>.

> >  So I think it's better to use explicit markup.
> > The ultimate explicit system is tangle/weave but that's a bit extreme.
> > Javadoc has been very successful in practice and you could do a lot
> > worse than just copying it lock stock and barrel.
> 
> Javadoc. Bleah. Completely unreadable in its source form - which is in the 
> code. You need to be able to read the documentation when viewing the code!

I don't find Javadoc unreadable (it's just HTML) but it's certainly
ugly and I'd rather read a formatted version.

> Please, read the justification text, and the anlysis of existing formats. 
> POD, while not mentioned, shares a lot of the shortcomings of other STX 
> formats in that it was grown over time.

I did read the justification text but found it scanty and
unpersuasive.  It read too much like a sales pitch and not enough like
an objective analysis.

> Again, have you _read_ the ReST documentation, specifically the
> discussion of alternatives? It's extensive, and very well
> written. It discusses many existing STX systems and their strengths
> and weaknesses.

I looked briefly at the ReST documentation linked from the PEP, but
if I'm supposed to read it in detail to be able to use it, I might
as well use a more traditional markup language.

> >  Javadoc and POD are
> > widely-deployed, practical systems; a lot of people use them and like
> > them and there are REASONS for that.
> 
> They are defacto standards. That's not always the sign of a well-designed 
> system.

They have good and bad points, and their good points were attractive
enough that a lot of people were willing to write docs in them instead
of in plain text or as separate docs rather than inline as is
traditional for C programs.  That says something in their favor.


From pyth at devel.trillke.net  Tue Apr 30 16:06:02 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 22:06:02 +0200
Subject: Update locals() / resolved
In-Reply-To: <eppstein-346BE0.12401530042002@news.service.uci.edu>
References: <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org> <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no> <mailman.1020193732.29753.python-list@python.org> <eppstein-346BE0.12401530042002@news.service.uci.edu>
Message-ID: <20020430200602.GH16217@devel.trillke>

On Tue, Apr 30, 2002 at 12:40:15PM -0700, David Eppstein wrote:
> In article <mailman.1020193732.29753.python-list at python.org>,
>  holger krekel <pyth at devel.trillke.net> wrote:
> 
> > > Thes names aren't reserved in any way...
> > 
> > right. i do it all the time for file,str,dict,string,list...
> > they are simply to short and descriptive to be left bound
> > to something i don't access directly. and using type('') or
> > friends is usually a very precise way of expressing that
> > you want the type object for that particular thing.
> 
> So what do you do when you have overridden type?

I don't do it. Or i do it only in very limited
local scopes. type and the keywords are holy :-)

regards,
    holger




From max at malva.com.uaREMOVE.IT  Mon Apr 15 09:22:07 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Mon, 15 Apr 2002 16:22:07 +0300
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs> <a9e2g8$nhh$00$1@news.t-online.com>
Message-ID: <v5ke9a.ufm.ln@cvs>

 Achim Domma wrote:

>> I'd like to have the algorithm moved out from Python source.
>> I've heard about XSLT language that is used to describe transformation of
>> XML data into HTML. I'm seeking for something similar. The difference is
>> that target language is C++ and not HTML.
>>
>> Any suggestions?

> XSLT can transform XML into any kind of text, so you are not restricted to
> HTML. In fact, XSLT seems to be the solution you are looking for.

I'd elaborate it, thanks.



-- 
Shit happens.


From gusraan at terra.com.br  Wed Apr 17 18:56:59 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: Wed, 17 Apr 2002 19:56:59 -0300
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
References: <mailman.1019071474.3193.python-list@python.org> <Xns91F4680F3221cliechtigmxnet@62.2.16.82>
Message-ID: <a9kuj8$bsa$1@nsnmrro2-gest.nuria.telefonica-data.net>

Do i have to remove the last version of active python i have?

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns91F4680F3221cliechtigmxnet at 62.2.16.82...
> Trent Mick <TrentM at ActiveState.com> wrote in
> news:mailman.1019071474.3193.python-list at python.org:
> > In addition, PyPPM (the Python Package Manager) enables you to easily
> > install a number of other popular Python packages. For example, try
> > this:
> >     > pyppm install Numeric
> > to install Numeric Python. Use this command to list all available
> > packages:
> >     > pyppm search
>
> is that also available separately and can i set up my own packages and
> server?
>
> chris
>
>
> --
> Chris <cliechti at gmx.net>
>




From sholden at holdenweb.com  Mon Apr 22 10:03:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 10:03:37 -0400
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org>
Message-ID: <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com>

"Jonathan Hseu" <vomjom at vomjom.org> wrote ...
> Urllib will always have problems with some sites.  Notably, sites with
> cookies that also redirect.
>
> Here's what happens:
> Say, perhaps after POSTing, you want to grab a cookie from the headers.
> The website puts the cookie in a 302 message (redirection for those of
> you who don't know HTTP responses), and then redirects you elsewhere.
> Now, to grab the cookie, you need to be able to get the headers of that
> message before getting redirected.
>
It's true that urllib will not allow the client code to capture cookie
values in the circumstances you describe.

You do not mention it, but I would imagine (without having performed any
testing) that in the event it's redirected to another domain/path
combination a browser would send back a new set of cookie headers,
appropriate to the new URI.

> This is certainly not possible without overriding some functions in
> FancyURLopener or declaring some in URLopener (which creates exceptions
> for all HTTP responses != 200).  This isn't a big deal (I did it myself
> for my program), but this can be simply fixed by adding support for
> cookies within urllib.
>
> Shall I take it upon myself to make such a patch?  Comments/Flames?
>
Since urllib knows nothing of cookies, you will need to integrate some sort
of a cookie jar into the library, with a new API  for the clients to
retrieve and store the cookies.

> I personally think easy cookie handling within urllib is a good thing.

I agree, but you'll probably have to spend some time finding out what the
existing browser community does, and emulating that.

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From hgs at dmu.ac.uk  Wed Apr 17 13:54:36 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Wed, 17 Apr 2002 18:54:36 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4pu0y9w6c.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204171643370.18149-100000@neelix>

On 17 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > WARNING: removing "math" since importing it failed
> >         [...]
>
> > So even building statically doesn't work.
>
> You did not build statically. Instead, all extension modules are build
> by setup.py now.

I left *shared* commented out.  So that was not enough, and neither was
*static* apparently.
>
> It would be crucial to understand *why* the imports fail. For that, it
> would be crucial to perform the following actions:

OK, back to the complete beginning, (make distclean; configure)
not touching Modules/Setup at all.
I collect the commnands and re-invoke them by hand ...
>
> 1. Re-invoke the commands that are used to build the math module
>    manually (without invoking make; just invoking the compiler)
>
> 2. Invoke ./python
>
> 3. 'import math'
>
> 4. Report the errors you get when invoking math that way

Script started on Wed 17 Apr 2002 18:32:26 BST
neelix hgs 11 %> gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude -c /home/hgs/Python-2.2.1/Modules/mathmodule.c -o build/temp.solaris-2.7-sun4u-2.2/mathmodule.o
neelix hgs 12 %> gcc -shared build/temp.solaris-2.7-sun4u-2.2/mathmodule.o -L/usr/local/lib -lm -o build/lib.solaris-2.7-sun4u-2.2/math.so
neelix hgs 13 %> ./python
Python 2.2.1 (#1, Apr 17 2002, 17:39:20)
[GCC 2.95.3 20010315 (release)] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import math
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/build/lib.solaris-2.7-sun4u-2.2/math.so: symbol PyExc_OverflowError: referenced symbol not found
>>>
neelix hgs 14 %> exit
neelix hgs 15 %>
script done on Wed 17 Apr 2002 18:43:06 BST
>
> 5. Report the command line used to link python.
Script started on Wed Apr 17 17:31:21 2002
neelix hgs 11 %> gmake
        [...]
ranlib libpython2.2.a
gcc   -o python \
		Modules/python.o \
		libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm
case [....]
>
> Regards,
> Martin
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you,
        Hugh





From alwyn at alwyn.demon.co.uk  Sat Apr 20 11:19:59 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Sat, 20 Apr 2002 15:19:59 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org>
Message-ID: <20020420161950604+0100@pc-62-30-160-65-hw.blueyonder.co.uk>

In <mailman.1019299357.11513.python-list at python.org> "James J. Besemer"
<jb at cascade-sys.com> wrote:
>
>Note that CAR and CDR stand for Contents Address Register and Contents
>Data Register.  The original Lisp implementation (before some of you
>were born) was very specific to the IBM 709X architecture, in which it
>was easy to pack two addresses into a single word (with a
>few bits to spare). 

It was an IBM 704 to start with, actually. CAR stood for 'Contents of the
Address part of Register number' and CDR for 'Contents of the Decrement
part of Register number'. See:
http://www
-formal.stanford.edu/jmc/history/lisp/node2.html#SECTION00020000000000000000


Alwyn


From zopestoller at thomas-guettler.de  Fri Apr 19 08:36:04 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 19 Apr 2002 14:36:04 +0200
Subject: (Win32) Mapping WebDAV to Drive Letter
Message-ID: <3CC00F34.8000807@thomas-guettler.de>

Hi!

Does someone know how you could use the win32 api
with python to map a webdav server to a drive letter?

There is a C library for the webdav part. But I have
no clue how you could map it to a drive letter.

thomas



From jacob at boris.cd.chalmers.se.cd.chalmers.se  Tue Apr 30 16:48:42 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 30 Apr 2002 20:48:42 GMT
Subject: Tix Vs Pmw which is better to use?
References: <3CCE6C0D.FA96F79A@millfilm.co.uk>
Message-ID: <aamvva$l6a$1@nyheter.chalmers.se>

In article <3CCE6C0D.FA96F79A at millfilm.co.uk>,
Eric Texier  <erict at millfilm.co.uk> wrote:
>My application needs a very dynamic widget package, It needs to add and
>remove  widget constently.
>I am looking also for something that will let me create more organic
>widget like tree but on a free canvas.
>
>- Do I have to be aware of some major difference before I should choose
>one or the other?
>- Is there an other package to be considered ?

I have used Pmw for quite a while, and while it is a very nice and very well
coded package, it was not written with inheritance in mind. For this reason,
if what you want to do (as it turned out it was what I wanted to do) is to
do something slightly different, it is easier to write your own widget from 
scratch.

I had a brief look at Tix but was not all that impressed.

For a nice tree widget to work with and expand, use the one that is a part of 
Idle. It is simple and fairly cleanly implemented.

Jacob Hall?n


-- 


From jon+usenet at unequivocal.co.uk  Tue Apr 30 11:46:54 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 30 Apr 2002 15:46:54 -0000
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net> <23891c90.0204300033.16549c52@posting.google.com> <3CCEAE2B.8000300@gmx.net>
Message-ID: <slrnactf3d.lf5.jon+usenet@snowy.squish.net>

In article <3CCEAE2B.8000300 at gmx.net>, Tom Chance wrote:
><TMPL_VAR NAME=TITLE>
> 
> $template->param(TITLE => 'Hello there!');
> 
> Perhaps Python merely needs an equivalent of this?

Python already has several equivalents of this. I think you can do
better though, which is what jonpy does. You instead do simply:

  $$title$$

and then in your Python code:

  class main(wt.TemplateCode):
    def title(self):
      return "Hello there!"

This more object-oriented approach is both simpler and more powerful,
since you can use class inheritance to very easily make re-usable
configurable objects. Have repeated sections such as table rows is
also very easy using 'sections':

  <!--wt:newsitem-->
    <tr><td><a href="newsitem.html?id=$$%ID$$">$$title$$</a></td></tr>
  <!--wt:/newsitem-->

  class main(wt.TemplateCode):
    class newsitem(rowaccess.RowAccess):
      def main(self, template):
        dbc = website.dbh.cursor()
        dbc.execute("SELECT ID,title FROM newsitems")
        for self.row in dbc.fetchall()
          self.process(template)

Cheers


Jon


From cliechti at gmx.net  Wed Apr 10 15:25:08 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 10 Apr 2002 21:25:08 +0200
Subject: Baffled by CGI cross-browser issues... help sought :)
References: <Xns91ECB3C2132FBPaCmAnRDLM@161.48.128.20>
Message-ID: <Xns91ECDA15DD485cliechtigmxnet@62.2.16.82>

Fred Pacquier <fredp at mygale.org.nospam> wrote in
news:Xns91ECB3C2132FBPaCmAnRDLM at 161.48.128.20: 
> a form I get an error message like "A network error occurred while
> receiving data (connection reset by peer)". After dismissing the
> dialog the resulting page usually displays correctly (although
> sometimes truncated). 

maybe you're not receiving not all the data of the request? i guess that 
happens with POST?
 
> If anyone has an idea of what and where to investigate I'd be grateful
> for that, as I'm nonplussed. The code is tiny and non-secret (although
> probably ugly as hell) so if anyone wants to take a look, that's easy
> too. 

if you had put in on your page and posted a link here, i might already have 
looked at it ;-)

chris

-- 
Chris <cliechti at gmx.net>



From mattmsykes at yahoo.co.uk  Mon Apr  8 10:00:30 2002
From: mattmsykes at yahoo.co.uk (Matt M Sykes)
Date: 8 Apr 2002 07:00:30 -0700
Subject: Carbonized embedded MacPython
Message-ID: <f69ccee4.0204080600.21d8ac87@posting.google.com>

Hi,

I would like to embed Python into an already-existing Carbon
application.

The application is quite large, so removing/replacing all the Carbon
API calls would be infeasible at this point due to time constraints.
That is, I must use Carbonized Python libs.

I downloaded and installed MacPython22full.bin which contains some
Carbon libs.  I soon found that I didn't have any header files.

On to the source.  There are two places to obtain sources: python.org
and the MacPython homepage, http://www.cwi.nl/~jack/macpython.html.
Note ftp://ftp.cwi.nl/pub/jack/python/mac/GUSI2-for-MacPython22.sit is
a broken link at the MacPython homepage.

The source at the MacPython homepage does not have a configure script
nor a project file; how do I build it?  It does have *some* prj files
but none for building python.  Hoping for the best, I tried using
those header files to compile an embedded-python test program with my
Carbon libs.  My machine promptly crashed when I ran it.  To me this
means the header files don't match the libs I have.  The exact same
code runs perfectly on Windows and FreeBSD.

Next: sources from the ftp site mentioned at python.org:
ftp://ftp.python.org/pub/python/2.2.

This source appears to require the unix-type build environment on OS
X; the instructions are to run the configure script and compile with
the unix-type (GNU?) compiler using make.

>From what I understand, I must use the same compiler (same run-time
environment) to build both python and my target application if I wish
to embed python in the target application.

For political and practical reasons beyond my control, the target
application must use the Metrowerks compiler (I am not in charge of
the Mac builds, I'm just helping with the embedded python side of it).

There is a mwerks directory in the sources, but the README there is
clearly out of date as none of the files it mentions exist.  There is
one project file called 'frozen.prj' which appears to be out of date
also.  I am not convinced this was the source used to build the Carbon
MacPython at the MacPython website.  I believe Jack did some magic
which is not included here.

All I really need (I think) are the exact header files which were used
to build the Carbon libs which are available at the MacPython website,
as well as info on the precise build environment and settings.  The
good news is that it says the Carbon libs were built with the
Metrowerks compiler.

Is there anyone who has embedded Python into a Carbon app?  Is there
anyone who is able to build Carbon-MacPython from sources (using the
Metrowerks compiler)?

Any help would be appreciated.

--Matt


From matt at mondoinfo.com  Sat Apr 20 22:50:22 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sun, 21 Apr 2002 02:50:22 GMT
Subject: Tkinter.Text.index() puzzlement
References: <3CC0B271.C76D3E95@nonesuch.com> <qB2w8.46153$vm6.6102555@ruti.visi.com> <3CC20EA2.1385A0E5@earthlink.net>
Message-ID: <ONpw8.46823$vm6.6727402@ruti.visi.com>

On Sun, 21 Apr 2002 00:59:04 GMT, David R. Smith
<DSmithSprint05 at earthlink.net> wrote:

Dear David,

>Matthew, thanks for your response.

You're very welcome.

>I'm trying to add a textual read-eval-print cycle onto a program
>which is mainly GUI-driven.  I thought I'd set a marker at the end of
>the computer-generated output, e.g.,

>    marker = text.index(END)

>and then when the <Return> event strikes, pick up the text between
>marker and text.index(END) as of that time, and send it off to eval.
>This keeps resulting in null strings, even though I entered
>"123+456<Return>".  Then the string might belatedly come through.  Is
>it hopeless to use text.index() to do useful work?

I suspect that that depends on your definition of hopeless <wink>. As
part of a larger project, I've been working on an undo function for
the Text widget for some time and I still haven't gotten it to figure
out just how the thing behaves in all situations.

On the other hand, it seems that what you're describing ought to be
possible:

>>> from Tkinter import *
>>> r=Tk()
>>> t=Text(r)
>>> t.pack()
>>> t.insert(END,"Wibble")
>>> i=t.index(END) # Entered some text after this line
>>> t.get(i,END)
'Text Text Text\n\n'

You might also want to have a look at the code for IDLE. It seems to
do something like what you're describing.

Regards,
Matt


From tavis at calrudd.com  Fri Apr 26 20:46:59 2002
From: tavis at calrudd.com (Tavis Rudd)
Date: 26 Apr 2002 17:46:59 -0700
Subject: Killing a process under Windows NT/2K
References: <aab8hn$o4s$05$1@news.t-online.com> <6affa25.0204260852.3e1a4f4a@posting.google.com>
Message-ID: <f67cb2b7.0204261646.20f9595@posting.google.com>

Syver Enstad (syver-en+usenet at online.no) wrote:
> mrenxo at mail.com (Enxo) wrote: 
> > I need to kill a process under Windows (NT/2K), whose PID is known.
> > Unfortunately os.kill() doesn't work under Windows.
> > How can I achieve this anyway?
> 
> Try this:
> 
>       import win32process
>       import win32api
>       import win32con
>                
>       pid = win32api.GetCurrentProcessId()  
>       handle = win32api.OpenProcess(win32con.PROCESS_TERMINATE, 0, pid)
>       win32process.TerminateProcess(handle, 1)

Is this a hard kill (like kill -9) or a soft kill that triggers the
appropriate signal handler inside the process?

Cheers,
Tavis


From clubdelpelosuperfluo at yahoo.it  Mon Apr 29 05:09:42 2002
From: clubdelpelosuperfluo at yahoo.it (Stefano)
Date: 29 Apr 2002 02:09:42 -0700
Subject: Numeric Python, COM and Eigenvalues
References: <9e1f166b.0204220437.57a087de@posting.google.com> <aa1ft2$mn7$2@peabody.colorado.edu> <9e1f166b.0204222329.6f5685a6@posting.google.com> <3CC66B94.3000101@skippinet.com.au>
Message-ID: <9e1f166b.0204290109.c92476d@posting.google.com>

Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<3CC66B94.3000101 at skippinet.com.au>...
> Stefano wrote:
> > Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aa1ft2$mn7$2 at peabody.colorado.edu>...
> > 
> >>Stefano wrote:
> >>
> >>
> >>>I am working with Numpy and trying to create a COM server.
> >>>I managed to call this server by Excel and get results like Inverse
> >>>matrix or Cholesky decomposition but the server doesn't work as I ask
> >>>for Eigenvalues/vectors (Excel does not respond anymore).
> >>>Has anyone heard of a problem of this kind?
> >>
> >>Well, with zero example code and details provided that's going to be a hard 
> >>one to get help on ;)
> >>
> >>Do you know for a fact you're feeding it a diagonalizable matrix? Does numpy 
> >>compute the eigenvals/vects nicely if you test it with the same matrix on its 
> >>own (not via excel)? Start testing with a symmetric 2x2 real matrix for which 
> >>you can hand-compute the eigenvalues trivially, that might tell you where the 
> >>problem is.
> >>
> >>Good luck,
> >>
> >>f.
> > 
> > 
> > I have already tried with a trivial matrix via Shell.
> > The same matrix makes Excel freeze.
> 
> Can you send the code for a trivial demo of the problem.  If you can get 
> a 10 line script that fills a 2x2 matrix and hangs Excel it would help.
> 
> Mark.

Mark,
I made not only Excel but also Doubletalk browser hang.
To make Doubletalk hang I wrote:

from Numeric import *
from Linearalgebra import *
a = array(([1,0],[0,1]))
b = eigenvalus(a)

and Doubletalk browser does not respond
Can this help you?
Otherwise, from VBA:
I have posted in the previous message the python COM.
You can make it freeze calling this sub from VBA:

Sub ItHangs()
    Dim numpy As Object
    Dim matrice(1 To 2, 1 To 2) As Variant
    Dim a as Variant
    matrice(2, 2) = 1
    matrice(1, 1) = 1
    Set numpy = CreateObject("PythonNumeric.Utilities")
    a = numpy.Eigenvalues(matrice)
    Set numpy = Nothing
End Sub

Thanks

Stefano


From noah at noah.org  Wed Apr 24 21:03:34 2002
From: noah at noah.org (Noah)
Date: Wed, 24 Apr 2002 18:03:34 -0700
Subject: Is  fcntl.ioctl broken?
Message-ID: <PAENINBOOODFDEPOKPKCIEIACMAA.noah@noah.org>

The documentation for ioctl in the fcntl module says:
	"ioctl(fd, op, arg) 
	This function is identical to the fcntl() function, 
	except that the operations are typically defined in 
	the library module IOCTL."

The only problem I can see is that there is no IOCTL module.
I'm looking to get and set the winsize structure for the
controlling terminal.

Yours,
Noah







From brian at sweetapp.com  Mon Apr  1 17:57:19 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 1 Apr 2002 14:57:19 -0800
Subject: unicode string problems
In-Reply-To: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com>
Message-ID: <003401c1d9d0$9408bc70$445d4540@Dell2>

Gon?alo Rodrigues wrote:

> f.write("Mar?o 2002" + march.Name())
> 
> where march.Name() returns a unicode string I get a unicode error. I
> tried converting both unicodes to strings via str but obviously I got
an
> error (in the first string the culprit is the "?" character).
> 
> Can someone help me out here and show me the way to write these
strings
> to the file?

Maybe there should be a FAQ for this...

The problem that you are having is due to the fact that there are many
possible string encodings for the same Unicode string (and vise-versa).
For example:

>>> u'?'.encode('iso-8859-1')
'\x87'
>>> u'?'.encode('utf-8')
'\xc2\x87'
>>> u'?'.encode('utf-16-le')
'\x87\x00'

So, when you add a string and Unicode object together, Python attempts
to convert the string into a Unicode object. But it refuses to guess
what encoding you mean and rejects all non-ASCII characters.

Here is a simple solution:

f.write("Mar?o 2002" + march.Name().encode('latin-1'))

This will convert the Unicode name into a string object using the
Latin-1 encoding.

Cheers,
Brian




From josegomez at gmx.net  Thu Apr 18 11:00:48 2002
From: josegomez at gmx.net (josegomez at gmx.net)
Date: Thu, 18 Apr 2002 15:00:48 +0000 (UTC)
Subject: Selecting Polygons in an image
Message-ID: <a9mn30$kbq$1@hermes.shef.ac.uk>

Hi all,
    I am using an old in-house software package to load up HIPS files
(it's some sort of graphic formst that stores a header, plus a number
of different images or "frames". It is used in Remote Sensing and
medical imaging). The process is fairly simple: you load your image,
and then you select with them mouse a number of polygons of interest
(PoI). These can then be saved on to a file as a linked list of
coordinates. You then use these coordinates to extract information
from each frame (they are co-registered).

    The problem is that the program I am using is very old. I can't
find the source for it to compile it for Linux, and well, it's just
so annoying to use that a full re-write could be useful. I was
thinking about Python, as Numeric would be a terrific tool to use. I
had a look at scipy, which uses (among others!) some WxWindows
code to do plotting. However, I don't really know how flexible this is
to basically treat a canvas region and select points in it and all
that. It would be greatly useful to know of similar efforts before I
actually get down to do some work on thi!

    Cheers,
    Jose

-- 
Jos? L G?mez Dans			PhD student
Tel: +44 114 222 5582			Radar & Communications Group
FAX; +44 870 132 2990			Department of Electronic Engineering
					University of Sheffield UK


From hgs at dmu.ac.uk  Thu Apr 18 08:17:50 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 13:17:50 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j41ydd8cp1.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181304460.24683-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > So:
> >
> > #
> > # In addition, *static* reverses this effect (negating a previous
> > # *shared* line).
> >
> > is ambiguous, because it does not reverse all the effects of *shared*?
>
> Maybe. This is meant in the context of the Setup file: all modules
> listed after shared are compiled statically. If no modules are listed,
> this has no effect.
>
> If you can propose a rewording, please submit a patch to
> sf.net/projects/python.

I can't yet: I don't know which properties are reversed:
 * Modules included in the config.c file
 * List of objects to be added to library archive
 * linker options added to linker options
 * Rules added to Makefile
 * Names collected in SHAREDMODS variable.
>
> > I have tried to configure this in different ways:
> > configure
> > configure -- with-gnu-ld
>
> Python's configure does not support --with-gnu-ld, so this option has
> no effect.
>
> > What else does it look for?
>
> Use 'gcc --print-prog-name=ld' to find out which ld binary gcc uses.
/usr/local/sparc-sun-solaris2.7/bin/ld
/usr/local/sparc-sun-solaris2.7/bin/ld --version gives

GNU ld version 2.12
Copyright 2002 Free Software Foundation, Inc.
>
> > If I symling /usr/local/bin/ld to gld will configure find it?
>
> No. If you create an ld link in

Oh, ok, I thought configure might look for gld first.

> /usr/local/lib/gcc-lib/<system>/<version>; that will make gcc use a
> different ld binary.
>
> *However*, there is no need to do that; the system linker should work
>  as well. Please invoke the python linker line line manually, adding
>  -v:

I think I will delay doing that for now, because most of the tests
worked without using GCC.  Another post suggests that this is not common
to Solaris7, (referring to pre-releases of Python-2.2.1).  Therefore I
may not really be helping anyone else by exploring that.
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you again,
        Hugh





From eppstein at ics.uci.edu  Mon Apr 29 12:10:29 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 29 Apr 2002 09:10:29 -0700
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1>
Message-ID: <eppstein-AA3306.09102929042002@news.service.uci.edu>

In article <Xns91FF9515A57Eduncanrcpcouk at 127.0.0.1>,
 Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:

> All of these replacements are big winners if they remove a use of the 'l' 
> word. I don't think any of them lose in clarity even if there are no 'l's 
> around.

If you want to get rid of lambda, say you're getting rid of lambda.
But all of these functional idioms are useful with nested def:s, and 
probably more clear that way than with lambdas.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From mwh at python.net  Wed Apr 10 06:22:11 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 10 Apr 2002 10:22:11 GMT
Subject: binding a reference to a variable
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com> <mailman.1018385900.7252.python-list@python.org>
Message-ID: <u8z7valyh.fsf@python.net>

Andrew Koenig <ark at research.att.com> writes:

> Mark> I'm curious why you need this?
> 
> I want to be able to stuff an object into a larger data
> structure and have that object reach out into a variable and leave a
> trail of breadcrumbs.

Can you come up with a more fleshed out example?  There probably are
some ways of doing what you want but they will be hackish and fragile
(grovelling around in frame objects, etc).  OTOH, if you explain what
you are trying to acheive, perhaps we can come up with a more elegant
way (it strikes me that 2.2's properties may help here, for instance).

Having random variables change their values at random times strikes me
as an idea that might come from - ooh - C++? <wink>

Cheers,
M.

-- 
  I don't have any special knowledge of all this. In fact, I made all
  the above up, in the hope that it corresponds to reality.
                                            -- Mark Carroll, ucam.chat


From maxm at mxm.dk  Mon Apr  8 06:39:16 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 08 Apr 2002 12:39:16 +0200
Subject: How do I generate 1/f (pink) noise in Python
Message-ID: <3CB17354.60609@mxm.dk>

I am dabling in algorithmical composition and at times I need to 
generate random values. Often the values should have the characteristics 
of pink noise (1/f).

My guess is that the random module can allready do this with one of the 
built in functions and the right parameters, but I must admit that my 
mathematical foundation isn't strong enough to deduct how to do it.

Speed is not of the essence here.

Google turns up nothing so does anybody know?

regards Max M



From guido at python.org  Fri Apr 26 22:04:12 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 27 Apr 2002 02:04:12 GMT
Subject: [Crew] PSA revival - kind of?
References: <mailman.1019860411.18377.python-list@python.org>
Message-ID: <3CCA078E.970DD66D@python.org>

> [Guido]
> > You can send your donation to the PSF.

[Tim]
> No, you cannot.  Not yet.  The PSF needs a favorable advance ruling from the
> Internal Revenue Service before it can accept any donations, and we haven't
> yet completed the mountains of paperwork needed for that.

Actually, according to our accountant, we can claim tax deductibility
as soon as we've *filed* the 501(c)(3) application form.  That's at most
another month.

> >  http://www.python.org/psf/  Membership is free but only on invitation;
> 
> Note that invitations aren't offered by Guido, so bribing him won't help.

That's what Tim thinks. :-)

> Dates with Uncle Timmy are, although somehow that didn't make it into the
> mission statement.

Not for minors though...

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


From sheershion at mailexpire.com  Tue Apr 30 19:22:44 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Tue, 30 Apr 2002 23:22:44 -0000
Subject: Problem with module
References: <aam9sa$bvfkp$1@ID-131182.news.dfncis.de> <D5yz8.15957$xy4.10514@atlpnn01.usenetserver.com>
Message-ID: <Xns9201DF771440rcamesz@amesz.demon.nl>

Steve Holden wrote:

> I'm sorry I can't be any direct help. However, you might wish to
> know that because IDLE uses the Tkinter graphical troolkit, you
> might expect trouble when mixing it with wxPython programs - you
> can get two graphical main loops struggling for control over your
> windows. 

IIRC, the main problem with Idle is that it doesn't spawn a separate 
Python interpreter for executing scripts.

My advice: don't use Idle, use SciTE.


Robert Amesz


From aleax at aleax.it  Tue Apr  2 15:49:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 02 Apr 2002 20:49:52 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA8E37A.E89CA002@noaa.gov> <7x4rivj5ya.fsf@ruckus.brouhaha.com> <3CA9E7CF.52D6CD37@noaa.gov>
Message-ID: <QPoq8.44596$pT1.1256392@news1.tin.it>

Chris Barker wrote:
        ...
> do think that that is how it should be in the ideal language. The
> concept of True or False is distinct from the concept of a number, and
> I'd like to see it stay that way in the language.

Have a look at Icon (or Snobol, but that was goto-based... Icon's
much nicer).  It CENTERS on this concept.


> Yes, you would be writing a lot of
> 
> if bool(expression):
> 
> rather than
> 
> if expression:

Tut, tut -- in Icon you do write "if expression" (well, syntax's a bit
different and my memory rusty, but, basically).  There's no concept
of true or false or a boolean type: rather, any expression succeds
or fails. Only if it succeeds does it return anything (and there's no
such thing as 'returning a boolean', of course).  'if' tests whether
the expression succeeds (and doesn't care about what it returns,
as long as it does -- but you can save it somewhere with a :=,
since x:=expression succeeds iff expression does).

There's more, of course, such as the ability to backtrack (trying to
make the expression succeed by pumping sub-expressions for
multiple results), the ability to return multiple results one after the
other (and control structures to pump an expression for all of its
results), and so on.  Often you'll find these things reminiscent of
Python concepts (exceptions, iterators), but Icon was there first
(there may have been some indirect inspiration from Icon in some
of Python's aspects).

See: http://www.cs.arizona.edu/icon/ .

Icon's a stimulating language, and far from dead (version 9.4.1 for
Unix was released on March 27, although the Windows version is
still current at 9.3.2).  However, it has had nothing like Python's
success.  I think it's worth studying if only to understand what DID
Python get right, or luckier, than Icon.  (And anyway, Icon's
paradigms are different enough from most other language's
to make its study worthwhile as a mind-widening experience).

If you want "booleans" given the respect you seem to think they
deserve, you may actually be happier with Icon than with Python.
Me, I think that Python's decision regarding "concrete nothings",
although less pure, was more pragmatically useful than Icon's
purity in distinguishing "something or nothing" from any actual
kind of "somethings", and contributed to Python's greater success.
But - study Icon, try it out on some concrete programming task,
and make up your own mind.


Alex



From daw at mozart.cs.berkeley.edu  Thu Apr 25 14:57:30 2002
From: daw at mozart.cs.berkeley.edu (David Wagner)
Date: Thu, 25 Apr 2002 18:57:30 +0000 (UTC)
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com> <B8ECCE7D.F941%richard@electrophobia.com> <7xu1q0vwqi.fsf@ruckus.brouhaha.com>
Message-ID: <aa9jiq$9f6$1@agate.berkeley.edu>

Paul Rubin  wrote:
>Anyway, I begin to miss my original scheme of hashing under the
>encryption (MAC-then-encrypt):
>  ciphertext = encrypt(plaintext + H(plaintext))
>The Bellare/Namprempre article list this scheme as insecure but I
>don't understand the reasons yet.

Yeah.  A lot of people find this counter-intuitive.
(The question seems to come up about once every 6 months here
on sci.crypt.)  Fortunately, I saved my answer from the last time. :-)
The following attack assumes encrypt() uses CBC mode encryption,
though the comments apply to most of the common modes of
operation.

Suppose you want to make the receiver think message M was sent,
even thought the sender would never authorize this.  Then you
should construct M' = M || H(M) || X, where X is arbitrary and
where || denotes concatenation (it better be at block boundaries).
Ask the sender to transmit M'; he will form M' || H(M'), encrypt
it with CBC mode, and transmit the result.  Now you can truncate
the ciphertext, snipping it just before the "X" part, and the
receiver will think M was sent, which is an integrity failure.


From marklists at mceahern.com  Thu Apr 25 09:23:16 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 25 Apr 2002 08:23:16 -0500
Subject: configuring mxCGIPython
In-Reply-To: <23891c90.0204250140.7248c6ed@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEJEKMAA.marklists@mceahern.com>

[Paul Boddie]
> I found that it isn't always possible to deploy Python programs in a
> cgi-bin directory and to use the above notation to persuade the
> operating system to locate the Python executable. In the end, I used a
> wrapper shell script for each of my Python programs.

Here's the wrapper shell script I'm using.  I'd be very interested in any
feedback, suggestions you might have:

  #!/bin/bash

  echo content-type: text/plain
  echo

  echo Trying to capture the stderr of cgipython

  if [ -f cgipython ]
  then
    echo cgipython exists
    # Trying to send stderr to stdout.
    ./cgipython test.py 2>&1
  else
    echo cgipython does not exist
  fi

test.py, fwiw, is simply:

  import sys
  print sys.version

When I run this, I get:

  Trying to capture the stderr of cgipython
  cgipython exists
  ./cgipython: error in loading shared libraries: libxmlparse.so.1: cannot
open shared object file: No such file or directory

I'm not sure yet what that means yet, but I think it means I may be able to
set LD_LIBRARY_PATH, if: 1) that file resides on the server; and 2) I know
where it is.

Is there a way I can tell cgipython not to bother loading that particular
shared object?

Thanks,

// mark





From tundra at tundraware.com  Mon Apr 22 14:10:05 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 18:10:05 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>  <20020422124126.GD23706@andy.tynant.ftech.net> <mailman.1019485540.16255.python-list@python.org>
Message-ID: <3CC451B1.C274099C@tundraware.com>

Laura Creighton wrote:
> 
> > On Mon, Apr 22, 2002 at 12:57:38PM +0100, Gon?alo Rodrigues wrote:
> >
> > > You should understand that this is a philosophical viewpoint. The
> > > mathematical field is roughly divided in the Platonist field, those who
> > > believe that mathematical objects exist somewhere out there in space (to
> > > quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
> > > field of those who believe mathematics is essentialy a game we play on
> > > paper with symbols with no "real" meaning attached to it besides the
> > > convenience in describing natural phenomena.
> >
> > To put this another way, ask yourself where you stand on this
> > question: is mathematics discovered or invented?
> >
> > -Andy
> 
> Those aren't the only two choices out there.  From the 'saving your
> cake and eating it too department' -- All mathematics is
> invented, but the human mind is only capable of inventing things that
> in some way correspond to 'things out there' (whatever that means,

Yeah, there is an analagous argument for the existence of God called
the 'ontological argument' that more-or-less  says 'since you can
conceive the notion of God, he must exist in some form'.  Now I'm
a theist, but I think this argument is baloney both in mathematics
and theology.  I can conceive any number of things for which there
is no correspondent in the physical universe: flying donkeys, unearned
wealth, honest politicians, and so on.  In other words, I claim you
can always conceive more than that which is real which is why mathematics
is primarily a contruct not a discovery.



-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From brueckd at tbye.com  Thu Apr 11 12:38:26 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Thu, 11 Apr 2002 09:38:26 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <XFMail.20020411084028.shalehperry@attbi.com>
Message-ID: <Pine.LNX.4.33.0204110933410.26004-100000@ns0.tbye.com>

On Thu, 11 Apr 2002, Sean 'Shaleh' Perry wrote:

> > d = empty list or dict
> > for item in someInputSet:
> >     do something to item and add it to d
> >
>
> since tuples are immutable there is no
> initialize-to-empty-then-loop-to-populate idiom, you are correct.
> Because of this when a tuple would work we have to use lists.
> Numerous times while coding I have started using a tuple and had to
> fall back to lists because I eventually needed a way to give all of
> the values in my list a value based on a function.
>
> I personally like tuples because they say "this list is immutable, it
> won't grow or shrink and try not to touch its contents".  Just like
> using 'const' and 'private' in other languages.

Ah...I think I understand your perspective a little better now. I never
think of a tuple as a list that's immutable, so I don't encounter very
often the same population problem. To me, lists become immutable when I
give them names in all caps, just like my variables become constants when
I name them that way.

-Dave





From brian at sweetapp.com  Fri Apr 19 03:22:46 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Fri, 19 Apr 2002 00:22:46 -0700
Subject: ActivePython vs. Python
In-Reply-To: <ad496f8.0204182232.288e35e3@posting.google.com>
Message-ID: <010701c1e773$03726b10$445d4540@Dell2>

Ville wrote:
> One drawback appears to be it's reliance on ActiveState (which has to
> consider such things as profit - and it also provides ActivePerl ;-).
> There was a long period when nothing happened with ActivePython, which
> no doubt alarmed many people.

It's not like ActivePython locks you in; it's just a nicer packaging of
PythonLabs Python. 

For Windows, I would certainly choose ActivePython over PythonLabs
Python, if ActivePython is available for the Python version that I want.

Cheers,
Brian





From mwh at python.net  Thu Apr 18 06:00:27 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 18 Apr 2002 10:00:27 GMT
Subject: Installation problem
References: <5.1.0.14.2.20020417144022.00b55a48@postoffice4.mail.cornell.edu> <mailman.1019076282.20400.python-list@python.org>
Message-ID: <lkads1mif8.fsf@pc150.maths.bris.ac.uk>

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

> On Wed, Apr 17, 2002 at 02:47:24PM -0400, Eric Evans wrote:
> > I'm having a problem trying to install Python 2.2.1 and I'm hoping someone 
> > out there may be able to offer some helpful advice.  I'm on Solaris 8 with 
> 
>    It seems Python 2.2.1 has got a lot of compilation problems on different
> versions of Solaris :(

Also that I go *no* such reports during the c1, c2 process :(

Quite a few people got it to build and pass the test suite...

Cheers,
M.

-- 
  In the 1950s and 60s there was a regular brain drain of young
  Australians from the cities to London, but it was because of 
  money, culture and opportunity, not spiders.
           -- Al Grant, ucam.chat, from Owen Dunn's review of the year


From missive at frontiernet.net  Sun Apr  7 17:23:45 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 07 Apr 2002 21:23:45 -0000
Subject: help:problem using the RawPen class/turtle module.
References: <8374ee10.0204060303.3333260a@posting.google.com> <slrnaauvbv.3v.missive@whave.frontiernet.net> <8374ee10.0204070114.56260848@posting.google.com> <8374ee10.0204070523.3bbf801b@posting.google.com>
Message-ID: <slrnab1e81.4b.missive@whave.frontiernet.net>

> I think I've found the problem,
> it seems that the call to 
> 
> ...
> self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
>                                   width=self._width,
>                                   arrow="last",
>                                   capstyle="round",
>                                   fill=self._color)
> 
> ...
> 
> in the turtle.py ( _draw_turtle() method)
> doesn't work because the _canvas to refer to is
> the one in the 'self' instance,  
> so with the following change:
> 
> self._arrow = self._canvas.create_line(x-dx,y+dy,x,y,....
>               ^^^^^
> 
> everything is ok.
> 
> Without it the _canvas refers to a global that it
> isn't been initialized to a Canvas when there is a
> call to the RawPen class.
> 


Nice work.

In the version I have, that line does not exist, but all of
the calls to things like _canvas_create_*
are called as self._canvas_create_*
so you may have uncovered a bug.

I think the turtle module is not really used very much,
though I think it is pretty cool.

Please file a bug report!



From jimd at vega.starshine.org  Mon Apr  1 04:21:42 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 1 Apr 2002 09:21:42 GMT
Subject: Most important extensions?
References: <mailman.1017557628.22081.python-list@python.org> <a877p1$sns$1@panix1.panix.com> <20020331112707.353$Hl@news.newsreader.com> <a87gso$eh9$1@panix1.panix.com>
Message-ID: <a898r6$2e6r$1@news.idiom.com>

In article <a87gso$eh9$1 at panix1.panix.com>, Aahz wrote:
>In article <20020331112707.353$Hl at news.newsreader.com>,
>Hernan M. Foffani <hfoffani at yahoo.com> wrote:
>>"Aahz" <aahz at pythoncraft.com> escribi? en el mensaje

>>> Pmw does, but that's the only one.  Sorry for not making clear that I
>>> was specifically referring to 3rd-party extensions.

>> I'd think win32all may apply then.  But you said OSCON, so.... ;-)

> I'm adding it to my list of proposed extensions (and it'll likely be one
> of them, in the end).  OSCON is pro-Open Source, not anti-Windows.
> (There are plenty of presentations on using Open Source effectively under
> Windows.)

 Rpy, Medusa, PythonCard. 

 (Diversity, power, and potential)


From starx at pacbell.net  Wed Apr 24 00:27:51 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 04:27:51 GMT
Subject: Accessing class-level variables
References: <mailman.1019621385.21532.python-list@python.org>
Message-ID: <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4>

Gerhard =?iso-8859-15?Q?H=E4ring?= || Tue 23 Apr 2002 09:04:04p:

> I'm trying to access a class variable in a derived class. I sincerly
> hope there's a less kludgy way than what I'm currently using:
> 
> class X:
>     a = 5
> 
> class Y(X):
>     def p(self):
>         print self.__class__.a
> 
> Y().p()
> 
> Gerhard

Well, i knew this worked for dispatching __init__'s, but... well here 
=):

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> class A:
	d = 5

>>> class B:
	def p(self):
		print A.d

>>> z = B()
>>> z.p()
5
-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From m_mariappanX at trillium.com  Tue Apr 23 03:13:33 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Tue, 23 Apr 2002 00:13:33 -0700
Subject: FW: A Q on using one of the Pmw widget Dialog
Message-ID: <53A7943A5BD8D411B6930002A5073155022D746F@bgsmsx90.iind.intel.com>

> Hi all,
> 
> A Q on using one of the Pmw widget Dialog   I wud like to display a dialog
> box[global modal mode]. The Dialog box  have only a label saying "Wait a
> moment..". There is no btns in the buttonbox. Till this step working....
> And  some time after, a method on the 2nd thread will try to
> deactivate[withdraw] this dialog box. and it cudn't do like that Anyidea
> on using Dialog widget in this manner.
> 
> Maharajan




From bokr at oz.net  Wed Apr 10 18:52:25 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 22:52:25 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <3CB3F0FE.B5382A7C@kfunigraz.ac.at>
Message-ID: <a92fn9$kr$0@216.39.172.122>

On Wed, 10 Apr 2002 09:59:58 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>phil hunt wrote:
>
>> I disagree. K&R has a mixture of clarity and concision that makes
>> it a delight to read.
>> 
That's true IMO also.
>> IMO, anyone for whom K&R isn't to their taste, probably won't make a
>> very good programmer anyway.
>
Why make that bet, though? There are always people that will surprise you
in their combinations of attributes (which makes the world more interesting ;-)

>The last thing which I am  bothered with: whether to become a good
>programmer or not. For example, I did never use a struct in C; that
>means my C code is more or less a big spaghetti one. I do not care as
>long as I get the correct result.
I take it you never write anything that isn't faster to reinvent
than to get the benefit of reusing in any way ;-)

>
>And I have really, really not the intention to waste my life as a
>professional programmer (there I think one should adapt to a better
>programming style). Though programming is sometimes interesting and
>indispensible.
>
If you are going to do much programming, I suspect you will ultimately
waste more of your life if you habitually write "big spaghetti" ;-)

Regards,
Bengt Richter


From sandysj at juno.com  Wed Apr  3 12:01:06 2002
From: sandysj at juno.com (Jeff Sandys)
Date: Wed, 3 Apr 2002 17:01:06 GMT
Subject: Python message board application
References: <mailman.1017767705.13411.python-list@python.org>
Message-ID: <3CAB3552.D5F12ACB@juno.com>

Richard Walkington wrote:
> 
> Does anyone know of a cgi based message board app written in Python.
> I've had a look on Freshmeat, Google etc but haven't come up with much.

The Seattle Python Interest Group uses a Python based wiki called
MoinMoin:
	http://www.seapig.org
	http://twistedmatrix.com/users/jh.twistd/moin/moin.cgi/MoinMoin
Thanks,
Jeff Sandys


From jon+usenet at unequivocal.co.uk  Tue Apr 16 13:23:35 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 16 Apr 2002 17:23:35 -0000
Subject: httplib drops ssl connection to MS IIS Server
References: <mailman.1018741166.19994.python-list@python.org> <slrnabltto.f08.jon+usenet@snowy.squish.net> <a9hgc7$r0d$1@coco.singnet.com.sg>
Message-ID: <slrnabongh.5mb.jon+usenet@snowy.squish.net>

In article <a9hgc7$r0d$1 at coco.singnet.com.sg>, Ng Pheng Siong wrote:
>> It is a bug in MS IIS, it just closes the SSL socket rather than going
>> through the SSL shutdown negotiation. I don't think you can work
>> around this using httplib.
> 
> Nobody is accusing MS IIS of being bug-free, but I think in this instance
> the bug is with Python's SSL.

If it is, I can't find it. I also remember finding information about
this bug on the Internet back when it was originally causing me
problems, but I can't find anything now :(

Patch #500311 on SourceForge fixes the problem for me.


From martin at v.loewis.de  Wed Apr  3 02:14:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 03 Apr 2002 09:14:13 +0200
Subject: Buffer Interface, can't get PyObject_AsReadBuffer to work?
References: <5a4226f0.0204022145.49010b2e@posting.google.com>
Message-ID: <m3sn6dntiy.fsf@mira.informatik.hu-berlin.de>

kevin at cazabon.com (Kevin Cazabon) writes:

>   // Convert the pInBuffer from PyObject to a PyBuffer (readable),
> THIS DOESN'T WORK (the Python object type that is passed in as
> inBuffer is a 'B' array)
>   if (PyObject_CheckReadBuffer(inBuffer)) {
>     //this ALWAYS returns 0, meaning unsuccessful... what's wrong???
>     ok = PyObject_AsReadBuffer(inBuffer, &pInBuffer, &p4size);

0 means success, -1 means error. Use the source, Luke.

HTH,
Martin


From stephen at xemacs.org  Mon Apr 15 11:02:00 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 16 Apr 2002 00:02:00 +0900
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com> <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca> <MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca> <mailman.1018674988.22523.python-list@python.org> <87g01x1r6o.fsf@tleepslib.sk.tsukuba.ac.jp> <7xelhhmq5e.fsf@ruckus.brouhaha.com>
Message-ID: <87u1qdxaxj.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Paul" == Paul Rubin <phr-n2002a at nightsong.com> writes:

    Paul> "Stephen J. Turnbull" <stephen at xemacs.org> writes:
    >> And the (rms-deprecated!) LGPL is surely good enough for
    >> anybody who only cares about freedom of their own code.  So the
    >> GPL is an advocacy tool, pure and simple.  It's all about
    >> affecting licensing decisions for code that others write, it
    >> has nothing to do with protecting freedom of yours.

My bad.  s/yours/your code/.  Since it seems likely that you'd have
written the same response had I written that, I'll continue.

    Paul> Um, I value the freedom to use code that I wrote, including
    Paul> derivative versions.

I don't see a problem.  Since you wrote the derivative versions, you
are free to use them.<wink>  If you mean derivative versions written
by others, that incorporate your code, well, the code that you wrote
is still available in the original, and you're free to use it.  I
still see no problem.  If you want to use code that you wrote, as
others have modified it, then you need the LGPL, of course.  But still
no problem that requires the GPL.

But if you want to use all of the code that _somebody else writes_,
and just happens to be _linked_ with code based on a few lines of
yours, well, this is where the GPL shines![1]  The GPL is a legal tool
that often increases your freedom to use their code by constraining
their freedom to use yours.  But it's your freedom to use _their_ code
that requires GPL "protection" (ie, the viral provision), not your
(or anyone's) freedom to use _your_ code in any sense, not modified
and definitely not verbatim.

This is generally a good thing for society, of course, since it's not
just the GPL author's freedom that is protected, but everybody's.  But
it's not about protecting your own creations.

    Paul> It promotes cooperation by providing an incentive to release
    Paul> free software.

"I'll make you an offer you will find rather expensive to refuse."
I'm not really sure I consider that cooperation.  Sounds a bit like a
commercial transaction to me.

There is no question that both the GNU- and BSD/MIT/Perl/Apache-style
communities are founded on voluntary inter-developer cooperation.  The
core groups seem to work together because, shucks, "we like working on
this software with each other" (except when we hate each other<wink>).
Nobody at GNU needs the GPL to get them to share!  Nor have the non-
copyleft communities fallen apart because of occasional proprietary
defections.

So it looks to me like the GPL is aimed at people who insist on quid
pro quo, not at people who are interested in, and value, cooperation
per se.


Footnotes: 
[1]  This is not at all implausible.  Google for "Aladdin Ghostscript
GNU readline" (case seems to be significant).

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From rjs at blue-cove.com  Tue Apr 16 16:24:41 2002
From: rjs at blue-cove.com (rjs)
Date: 16 Apr 2002 13:24:41 -0700
Subject: Python extension or dll and calldll?
Message-ID: <1c988afc.0204161224.21ee76bd@posting.google.com>

After reading through http://www.python.org/doc/ext/intro.html, and
not being a good C programmer (although I have access to MS VS 6), and
since I have not written either a Python extension or ddl before, and
anticipating having to spend some time learning, which is easier: turn
the ANSI C code into a dll and call it (which I do for another project
using ioport) or write a Pyhton  extension? The code is a fairly
straight-forward neural network, the makefile and header are below.
I'm writing wxPython and would just like to do direct calls to the
functions. I have a server/socket based version of the NN I run on
Linux, but that would require socket(), and I would like a desktop app
(for MS for now...)

Suggestions/experiences appreciated...

Ray

# Makefile
CC  = cc
SRCS    = iRACE.c serverb.c
OBJS    = iRACE.o serverb.o
INCLUDES= iRACE.h
LIBS    = -lm
TARGET  = server
all: $(TARGET)
$(TARGET): $(OBJS)
    $(CC) -o $(TARGET) $(OBJS) $(LIBS)
clean:
    rm -f $(OBJS) $(TARGET)



/*  rbfnet.h - function prototypes  */
/* Define some commonly used NETWORK receptive fields geometries */
#define HYPERDIAG           (int)'A'
#define RECTANGULAR         (int)'R'
#define LINEAR          (int)'L'
#define SPLINE          (int)'S'
#define CUSTOM          (int)'C'

int allocate_rbfnet(int num_state,int *qnt_state,
           int num_resp, int num_cell,
           int memory, int rfield_shape, int collision_flag);
int train_rbfnet(int rbfnet_id, int *state, int *respns, int beta1,
int beta2);
int rbfnet_response(int rbfnet_id, int *state, int *respns);
int clear_rbfnet_weights(int rbfnet_id);
int deallocate_rbfnet(int rbfnet_id);
int rbfnet_memory_usage(int rbfnet_id);
int get_rbfnet(int rbfnet_id, int index, int *buffer, int count);
int put_rbfnet(int rbfnet_id, int index, int *buffer, int count);
int adjust_rbfnet(int rbfnet_id,int *state,int *drespns,int beta1);
int map_rbfnet_input(int rbfnet_id,int *state,int *weights[], int
*rfmags);
int set_rbfnet_rf_displacement(int rbfnet_id,int *buffer);
int set_rbfnet_rf_magnitude(int rbfnet_id,int *buffer);
int save_rbfnet(int rbfnet_id,char *filename);


From gleki at gol.ge  Tue Apr  2 16:17:16 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Tue, 02 Apr 2002 13:17:16 -0800
Subject: simulated annealing
Message-ID: <3CAA1FDC.7FEF351D@gol.ge>

Hi!
Has somebody written a free module for simulated annealing?
I have an experiense of making this, just I'd like to avoid inventing a
weel:(

Thanx,
Greetings,
Giorgi



From brueckd at tbye.com  Fri Apr  5 16:38:51 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Fri, 5 Apr 2002 13:38:51 -0800 (PST)
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
In-Reply-To: <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96>
Message-ID: <Pine.LNX.4.33.0204051337230.9744-100000@ns0.tbye.com>

On 5 Apr 2002, James Logajan wrote:

> > Is there a final destination in sight, which these changes are
> > leading to?
>
> The answer to this isn't as interesting to me as _when_ will this tinkering
> end? Something that isn't up for debate: because of the problems I have
> actually encountered with making code work with the many extant versions of
> Python, I will no longer be doing any new projects in the language and am
> not recommending Python to anyone. The language tinkering is one of several
> reasons I left Python and this newsgroup many months ago. (I have
> resubscribed recently to find out if there was a way to rework an old
> Tkinter app of mine that people have asked for some changes to, and will be
> exiting here in a few days.)

So, what language did you move on to?

-Dave





From theller at python.net  Thu Apr 25 02:26:00 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 25 Apr 2002 08:26:00 +0200
Subject: Quick: Format Specs in %
References: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
Message-ID: <aa87i8$8j83h$1@ID-59885.news.dfncis.de>

"Philip Swartzleonard" <starx at pacbell.net> wrote in message news:Xns91FA8B766EF98RASXnewsDFE1 at 130.133.1.4...
>
> '%2f' % some_number pads the number with 2 blank spaces
> '%.2f' % some_number fixes the number to 2 decimal places
>
> the problem:
> '%2.2f' % some_number only does the second of those two things. I need
> both. I tried looking up the % operator in the index of the language
> reference, but couldn't find it.
>
See the 'Python Manual Search':

http://starship.python.net/crew/theller/pyhelp.cgi?keyword=%25&version=current

Thomas




From bokr at oz.net  Mon Apr 15 20:32:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 00:32:38 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9frf6$d6u$0@216.39.172.122>

On 15 Apr 2002 23:06:43 +0200, martin at v.loewis.de (Martin v. Loewis) wrote:

>bokr at oz.net (Bengt Richter) writes:
>
>> IMO 'internal representation' is a red herring. We have an abstract
>> concept of number representation (twos complement) that corresponds
>> to many hardware devices' usage, but it is the abstraction that is
>> important IMO, not the hardware.
>
>Yes, but this notion is inherently bound with the notion of "size" of
>an integer, atleast in the signed variant (for unsigned integers,
>there is no need for twos complement). If you create the complement,
>you have to know how many bits to invert.
All of them ;-) I don't need an infinite number of representation bits
if they are all the same value. Thus sign can be considered to be
indefinitely extended in the abstraction, and you don't need to specify
width.
>
>I absolutely agree with your (earlier) conclusion that bit arithmetic
>is best done with unsigned numbers.
I don't remember using those words ;-) In any case, in the abstract, you
don't do bit arithmetic on the numbers per se (i.e., taking a number
as a point on the real axis). We first have to get to bits. And I think
you can get to bits just as well from negative numbers as from positive.
What's the difference between an infinite number of extended 0 sign bits
vs the same number of 1 sign bits? Either way, they're all the same and
can be represented by a single bit.

>
>> When looking at integers cast as arrays of booleans, hex provides a
>> representation, but w.r.t. boolean values it is not very useful to
>> print it as base-16 integer with sign+absolute value.
>
>I also agree.
>
>> If I infer correctly, Ken is arguing that hex is more often used to
>> view the boolean aspect of integers than their numeric
>> aspect. That's true for me too.
>
>And so it is for me. Still, I disagree that a display of a hex number
>with a sign is illogical or incorrect. In fact, when talking about
>"unbounded" numbers, it is illogical to assume twos complement.
>
I don't think I said it was illogical. I didn't mean to. As you pointed out,
base 16 is just another base that can be used like base 10, and analogous
signed representations of numbers are logical. I just think the signed
hex is not terribly useful.

>> >Yes-- I read the PEP too-- when this happens, I will consider it a
>> >Python "_WART_", and will set about inventing ways around it for my
>> >own use.  I have forgotten more computer languages than most people
>> >will ever learn-- an I have NEVER found printing a hexadecimal
>> >number in sign and magnitude format to be useful...
>>
>> I'll second that.
>
>At a minimum, you need to propose an alternative behaviour then. I
>must say that I don't find the proposed behaviour of mandatory leading
>zeroes (the 0h proposal) very intuitive, or following tradition.
Actually, my intent was to follow tradition, but tradition in hex
representation needed an adaptation to work for non-fixed width integers.
It occurred to me that if you included an actual sign bit the way the
traditional integer hex representation does, then all you'd have to
do was to make sure positive numbers had a zero sign bit represented.
That would imply a leading zero hex digit if the next digit was binary 1xxx.
Pretty simple.

>
>Instead, there is a simple solution: use unsigned numbers if you want
>to represent bit arrays.
>
Well, that assumes I want to use bit arrays without transformations
to/from general signed integers. That's a bit restrictive.

>> I would guess so. I do. But for language purposes my preference is
>> to refer to an abstract bit pattern, not hardware (even though
>> hardware mappings will exist and be useful).
>
>That is only meaningful for positive numbers - it just doesn't work
>for twos complement.
>
I works fine the way I'm thinking of it ;-) You must be looking at
it differently, and I'm not sure what view it doesn't work for.
My view is an abstraction where the sign bit is indefinitely duplicated
to the left, but since they're all the same, you can think of them
as represented by a single sign bit. If there is more than one bit
in the whole number, the top two bits will be different. In any case,
the top bit's value will be -(0|1)*2**N and the rest (0|1)*2**n, where
n is >=0 and <N. Ie.e, if the bits are numbered 0..N from the lsb,
and have values b[n], the numerical value would be
-b[N]*2**N plus b[n]*2**n for n=0..N-1.

Duplication of the topmost bit (changing N) doesn't change the numeric value,
whether + or -, so you can consider it to extend indefinitely, if convenient.
To represent such a number in hex, you would duplicate the
sign bit to fill the top hex digit. Thus +3and -3would be

+3   (0)011  0h7 -0*2**2 + 1*2**1 + 1*2**0    0+2+1 =  3  or  0+0+2+1
-3   (1)101  0hd -1*2**2 + 0*2**1 + 1*2**0   -4+0+1 = -3  or -8+4+0+1

The (x) are duplications to fill out the hex digit
but +8 and -8 would be

+8   01000 => 0001000 => 0h08  -0*2**4 + 1*2**3 + 0*2**2 + 0*2**1 + 0*2**0   0+8+0+0+0 =  0
-8    1000 =>    1000 => 0h8            -1*2**3 + 0*2**2 + 0*2**1 + 0*2**0    -8+0+0+0 = -8
or with a redundant digit
-8   11000 => 1111000 => 0hf8
 => -1*2**7 +  1*2**6 + 1*2**5 + 1*2**4 +1*2**3 + 0*2**2 + 0*2**1 + 0*2**0
       -128       +64      +32      +16      +8       +0       +0       +0  = -8

Of course the "two" in twos complement is scaled to match whatever point you want to stop at,
even infinity, by induction ;-)

Regards,
Bengt Richter


From boud at valdyas.org  Wed Apr 24 01:36:06 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Wed, 24 Apr 2002 07:36:06 +0200
Subject: PyGtk TestRunner for PyUnit?
References: <aa54p6$g2h@dispatch.concentric.net>
Message-ID: <3cc64474$0$338$e4fe514c@dreader1.news.xs4all.nl>

Phlip wrote:


> Has anyone written a PyGtk TestRunner for PyUnit?

No, but translating the existing Tkinter one to PyQt
proved to be really easy, not more than an evening's
work.

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


From shalehperry at attbi.com  Mon Apr  8 16:00:50 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 08 Apr 2002 12:00:50 -0800 (PDT)
Subject: html parsing
In-Reply-To: <3CB1E20A.B3300AA4@ncsa.uiuc.edu>
Message-ID: <XFMail.20020408120050.shalehperry@attbi.com>

On 08-Apr-2002 Randy Heiland wrote:
> I'm a newbie at parsing html in Python... could someone offer a snippet
> of code to:
> 1) specify a URL
> 2) parse the html and return the hyperlink associated with a unique
> anchor string
> 

import htmllib # and read the docs in python-lib about it, tells you all you
               # need in a clear manor




From guido at python.org  Fri Apr  5 13:38:55 2002
From: guido at python.org (Guido van Rossum)
Date: Fri, 05 Apr 2002 18:38:55 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96> <u7knmdimg.fsf@ctwd0143.fitlinxx.com> <Xns91E6BF93F68E0JamesLLugojcom@216.148.53.83>
Message-ID: <3CADF01A.7F0974B2@python.org>

James Logajan wrote:
> When I looked at it I saw a "class bool(int)" definition, so I
> wasn't sure how instances of such a (pseudo?) class would be handled in a
> dictionary.

It wasn't a pseudo class.  It was working code, and designed to answer
such questions precisely, without having to use ambiguous English words.

> The statement "The values False and True will be singletons,
> like None," further added to my confusion, since None is a valid dictionary
> key, distinct from all other values, which when combined with the class
> definition pretty much sealed the confusion. I assumed the "ACCEPTED: PEP
> 285" posting would have subsumed previous discussions, including any
> concerning the class instances as dictionary keys. I couldn't find a clear
> answer there, hence my question.

The PEP specifies the answer, by stating that True==1 and False==0.  The
answer really *does* follow from that (and even this statement is redundant,
as it follows from the way True and False are created in the code block).
You could have typed the code in to Python 2.2 and gotten your answer.

> My apologies to those who have sent me angry e-mails for not realizing that
> PEPs do not stand alone from the lengthy discussion threads on Usenet. The
> fault is entirely mine.

A PEP's specification *does* stand alone (besides dependence on the Python
spec) but it doesn't have to be a tutorial.

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


From psage at ncaustin.com  Mon Apr 15 10:51:01 2002
From: psage at ncaustin.com (Paul Sage)
Date: Mon, 15 Apr 2002 09:51:01 -0500
Subject: Looking for GUI Form Builder for Python
Message-ID: <03F53405C6ED434A986221919785A6A7106C10@nca-postal.ncaustin.com>

Have you investigated QT?  I believe it might be able to do what you are
looking for.  However, maybe someone on this list who has worked with it
will be able to confirm that for you.  I have not worked with it, but I
believe it has a Python wrapper available.

-----Original Message-----
From: Calvin [mailto:hoiwai930 at hotmail.com] 
Sent: Sunday, April 14, 2002 7:55 AM
To: python-list at python.org
Subject: Looking for GUI Form Builder for Python

Dear all,
 Can I build Python TK GUI interface using a form builder. Anyone know
such
builder. I just want to build my interface in a visual way.
-Calvin


-- 
http://mail.python.org/mailman/listinfo/python-list




From flyguysyl at hotmail.com  Sat Apr 20 07:57:29 2002
From: flyguysyl at hotmail.com (flyguy)
Date: 20 Apr 2002 04:57:29 -0700
Subject: Newbie: creating an <li> list
Message-ID: <c73096a5.0204200357.f7b9169@posting.google.com>

Hy. I'd like to know how you can send a new html page with a group of
radio button. Here's my attempt:

  r1 is the value of a radio button that a user checks so that he can
be offered a format of response (in this case radio buttons) to a
question he composed while constructing a survey:

if not (form.has_key("question") and form.has_key("r1") and
form.has_key("r2")):
    MaPage=MaPage +  "<H1>Error</H1>"
    MaPage=MaPage + "<h3>Veuillez construire votre question et choisir
un format de reponse</h3>"
    

else:
 if (form.has_key("question") and form.has_key("r1") and
form.has_key("r2")):
        Q=form["question"].value
        MaPage=MaPage + "<center><h2>Voici votre question ainsi que
votre choix de reponse</h2></center>"
        MaPage=MaPage + "<br><br><table width=80%><tr><td><h2>" + Q +
"</h2></td></tr></table>"
        
        if(form["r1"].value==1):
             MaPage=MaPage + "<ol><li><input type=radio name=rep1
value=1>"
             MaPage=MaPage + "<li><input type=radio name=rep1
value=2>"
             MaPage=MaPage + "<li><input type=radio name=rep1
value=3>"
             MaPage=MaPage + "<li>input type=radio name=rep1
value=4></ol>"
             
MaPage=MaPage + "</body></HTML>"
print MaPage


From johnroth at ameritech.net  Sat Apr  6 21:30:53 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 21:30:53 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org>
Message-ID: <uavbgcabvl642b@news.supernews.com>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1018111094.10533.python-list at python.org...
> +>
> > "Alex Martelli" <aleax at aleax.it> wrote in message
> > news:gJDq8.55255$S52.1904212 at news2.tin.it...
> > > John Roth wrote:
> > >         ...
> > > > Maybe my post in response didn't propagate? As I said in that
> > > > post, the root of her problem is a poorly chosen name for the
> > > > function. If she paid attention to constructing the name
properly,
> > >
> > > She adopted the name *Guido* had given to the function right in
> > > his pep (actually clarifying it a bit from 'isleap' in his PEP
> > > to 'isLeapYear', more readable, in her post).
> >
> > Sorry I'm late in responding to both posts - I had a problem
> > with my newsgroup vendor and I'm just back on.
> >
> > I thought the original example was very bad, but I didn't
> > say so at the time because I hadn't figured out what I didn't
> > like about it.
> >
> > The original example was something like
> >
> > number of days in year = 365 + isLeap(year).
> >
> > To put it bluntly, that's a very confusing hack that depends
> > on conceptual True being 1 and conceptual False being
> > zero.
> >
> > Again, to put it very bluntly, Laura swallowed a camel
> > while straining at a gnat. She had a great deal to say about
> > the difficulty of teaching that example, and I agree completely.
> > It would be very difficult to teach, but the reason is that
> > the name is simply wrong. The fact that Guido himself
> > used that as an example doesn't make it right, either.
> >
> > If she had corrected the name, then her entire basis for
> > the objection would have evaporated.
> >
> > John Roth
> > >
> > >
> > > Alex
> > >
>
> Still haven't found the time to read my original post, notice that
> I _did_ correct the name in the original post, and understand what
> I was trying to say, I see.  I will correct your misconception
> because you might confuse somebody else.
>
> I am objecting to the fact that isleap, whatever the hell it is
> called, returns a boolean.  I am rejecting the existence of
> booleans in the language because it is too hard to teach how
> to use them properly.  Most people never learn.  Most code that
> uses booleans should not.
>
> Proper use of booleans requires not only that you have a
> function that returns one of two values, but THAT YOU NEVER
> EVER WANT THAT FUNCTION TO EVER BE MODIFIED TO RETURN ONE OF THREE,
> (or more).  What I tell my students is that you really want
> a hole to open in the ceiling and a Warner Brothers safe to fall
> on the head of somebody who attempts to extend, subclass, or
> in any other way change the function to return one of three values.

I understood your point the first time. I disagree with it. If you
want to change the definition of a function to do something else,
then you may well need to change the name. This is called
refactoring.

> And, by God, if the function that returns _the number of leap days
> in the year_ isn't safe, at 0 or 1, then damn near close to nothing
> that my students are going to be tempted to use 'True' and 'False'
> for is.  This example makes my course notes, under Why using booleans
> is bad, and you should never do it.  (unless you are a mathematician
> who wants to do real 2-valued algebra, or somebody interfacing with
> some other piece of software that needs those sort of bools, or
> some other, EXTREMELY RARE,  person.  Don't you guys interrupt here.
> This is hard enough to teach as it is.)
>
> It does not help that it appears to the outside that Guido just went
> through all the python functions that returned 1 or 0 and changed
> them to return True or False.  I know that Guido would not be
> so stupid, and that there is a reason why he wants, for instance,
> isinstance to return only 0 or 1 for evermore, but I can't see it.
> I would like an isinstance that returned 0 for not an instance of
> this class, 1 for a direct instance, and 2 for a subclass.  There
> must be a lot of code out there I have never seen that relies on
> this never happening, and now, alas, I have to find out why.
>
> Because my students will ask, of course.  If True and False are
> good enough for Guido, then it's good enough for them.  I would
> love it if a huge comment was added to python files that now
> return booleans saying why it was necessary to restrict the
> return values to one of 2 forever more, so people are not confused
> into thinking that returning booleans is a good idea in general,
> but that may be too much to ask for.  I would add the comments
> myself, but I am having problems seeing what the reasons are.
> Maybe after I get some more sleep, it will come clearer to me.
>
> I truly liked telling students, especailly ones that have never used
> Python while we are using some other language,   'You don't need
> True and False. Look at Python.'
>
> Now I have to say, to the students who are using
> Python: 'Python has True and False.  It is not for you little
>          worms, and if I catch you using it I will make you do
>          the assignment over.'
>
> I know in all the Python code I have ever written I have only 5 times
> wanted a boolean True or False -- all times when I was talking to a
> database that required it.  The rest of the time I wanted an integer,
> 0 or 1.  Even if I couldn't see a way that somebody was likely to
> extend it to return 3, there most certainly was never ever going to be
> any code that required that it only have one of 2 values, so now I
> will have to change all my True and False there to be 'Yes' or 'No' or
> something.  (grumble, now I wish I had lost the argument to the person
> who wanted the constants we picked to be TRUE and FALSE).
>
> Teaching people to not use booleans is hard.  They grab for them
> whenever they have a function that returns one of two values.  It
> was a lot easier to teach this concept when the fool things were not
in
> the language.
>
> Note:  don't tell me that because python bools are just ints, it
>        doesn't matter -- people who were using bools for
>        DEMOCRAT and REPUBLICAN can now add
>        THE_DAMN_VOTE_WASNT_PUNCHED_THROUGH - (the integer 3)
>        when to their horror they discover 2 values was not enough.
>        I need to teach students to not use bools in languages which
>        are not so forgiving.

Here I thought that teaching about bools was quite easy.
Bools answer a specific question: is it x? True or False.
They do not answer the question: which X is it?

In other words, bools select one possible case out of
however many there are; they are not a substitute for a
universe which has only two cases.

People who were using bools for your Democrat or Republican
example were simply using them incorrectly. In most states, there
were way more than two choices: the Green party got enough votes
to throw the election to Bush. If he hadn't run, Gore would have
won!

It may very well be that bools are overused. That doesn't mean
that they should never be used.

John Roth

>
> Laura Creighton
>
>




From emile at fenx.com  Wed Apr 10 22:17:56 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 11 Apr 2002 02:17:56 GMT
Subject: Easter Eggs
References: <mailman.1018485864.13992.python-list@python.org>
Message-ID: <on6t8.206591$Yv2.67543@rwcrnsc54>

Tim Peters
> We take t-shirts very seriously at
> PythonLabs.
>
> they're-all-backward-compatible-too-ly y'rs  - tim
>

I don't know about that... mine seems too big, backward compatibly
speaking...

--

Emile van Sebille
emile at fenx.com

---------



From ftobin at neverending.org  Mon Apr 29 15:40:00 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 29 Apr 2002 19:40:00 GMT
Subject: comparing str's with ints
In-Reply-To: <20020429115856.R1840-100000@palanthas.neverending.org>
References: <20020429115856.R1840-100000@palanthas.neverending.org>
Message-ID: <20020429145902.M2252-100000@palanthas.neverending.org>

Frank Tobin, on 2002-04-29, wrote:

> BUT, *this* does not make sense:
> >>> '5' > 3
> 1
>
> What in the world is supposed to be the semantics of comparing a string
> with an integer?  Seems like a type trap...

I just read the Python language spec:

  The operators <, >, ==, >=, <=, and != compare the values of two
  objects. The objects need not have the same type. If both are numbers,
  they are converted to a common type. Otherwise, objects of different
  types always compare unequal, and are ordered consistently but
  arbitrarily.

'arbitrarily'.  Lovely, eh?

-- 
Frank Tobin		http://www.neverending.org/~ftobin/



From bokr at oz.net  Sat Apr 20 13:43:25 2002
From: bokr at oz.net (Bengt Richter)
Date: 20 Apr 2002 17:43:25 GMT
Subject: Pure python crypto
References: <20020417021703.H16954@eecs.tufts.edu> <mailman.1019257240.29930.python-list@python.org>
Message-ID: <a9s9bt$jj2$0@216.39.172.122>

On Fri, 19 Apr 2002 18:59:12 -0400, Michael Gilfix <mgilfix at eecs.tufts.edu> wrote:

>  For those of you who were interested, the pure python blowfish
>module can be found here: http://bofh.concordia.ca/blowfish.py.
>
>                    -- Mike
>
I missed the above thread, but thanks. Is there a c extension module version?
BTW, a nit: I'm pretty sure Bruce Schneier's name doesn't have a 'd' in it ;-)

Regards,
Bengt Richter


From shalehperry at attbi.com  Wed Apr 10 18:15:45 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 10 Apr 2002 15:15:45 -0700 (PDT)
Subject: Extracting from a list
In-Reply-To: <20020410145430.4432d2be.logiplexsoftware@earthlink.net>
Message-ID: <XFMail.20020410151545.shalehperry@attbi.com>

On 10-Apr-2002 Cliff Wells wrote:
> On Wed, 10 Apr 2002 23:42:46 +0200
> Daniel Dittmar wrote:
> 
>> How about:
>> 
>> for i in xrange (len (lst) - 1, -1, -1):
>>      item = lst [i]
>>      if item < 5:
>>          print item
>>          del lst [i]
>> 
>> This makes it possible to delete items from the list as the access is 
>> back to front.
> 
> Good answer.  Why didn't I think of that...
> 

because python coding does not make you think about counting loops ...




From pricee at hhbrown.com  Tue Apr  2 14:18:51 2002
From: pricee at hhbrown.com (Erik Price)
Date: Tue, 2 Apr 2002 14:18:51 -0500
Subject: compiling Python with readline
Message-ID: <777AB66E-466E-11D6-A42D-0050E4857868@hhbrown.com>

Hello,

Sorry to trouble if this is not the right list -- I would like to 
re-compile my Python 2.2 source (on Mac OS X 10.1.3) to support readline 
(or whatever it takes to get emacs-style key bindings to work in the 
interpreter, like I get in bash).

I was reading the README file in the source tree, and discovered that 
there is no longer a --with-readline ./configure parameter.  The 
instructions say "This option is no longer supported.  To use GNU 
readline, enable module 'readline' in the Modules/Setup file."  I don't 
seem to have the Modules/Setup file anywhere in my source tree.

If anyone knows how to enable the emacs key bindings, would you mind 
sending me a note about it?  I'm not subscribed to the list yet 
(normally I subscribe to tutor at python from another account, but didn't 
find my answer there), so a CC would be gratefully appreciated.

Thanks in advance,

Erik





----

Erik Price
Web Developer Temp
Media Lab, H.H. Brown
pricee at hhbrown.com




From eje4 at cornell.edu  Wed Apr 17 14:47:24 2002
From: eje4 at cornell.edu (Eric Evans)
Date: Wed, 17 Apr 2002 14:47:24 -0400
Subject: Installation problem
Message-ID: <5.1.0.14.2.20020417144022.00b55a48@postoffice4.mail.cornell.edu>

Hi,

I'm having a problem trying to install Python 2.2.1 and I'm hoping someone 
out there may be able to offer some helpful advice.  I'm on Solaris 8 with 
gcc 3.0, and during the installation of Python it compiles along just fine 
for a while until it gets to the place where it is running build_ext.  Then 
I start getting error messages such as

building 'struct' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. 
-I/usr/local/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c 
/usr/local/Python-2.2.1/Modules/structmodule.c -o 
build/temp.solaris-2.8-sun4u-2.2/structmodule.o
gcc -shared build/temp.solaris-2.8-sun4u-2.2/structmodule.o 
-L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/struct.so
WARNING: removing "struct" since importing it failed

and so on for all the rest of the extensions.  All of them fail.  What is 
going on here that is causing these compiles to fail?  I'd appreciate any 
guidance on this....

Thanks a lot,
Eric Evans
Cornell Phonetics Lab





From jb at cascade-sys.com  Sun Apr 28 18:25:20 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 28 Apr 2002 15:25:20 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <m2znzpwc7a.fsf@mycroft.actrix.gen.nz>
Message-ID: <3CCC76D0.BCF906A9@cascade-sys.com>

Paul Foley wrote:

> I think you're missing the point: Lisp is not defined in terms of
> text; instead, it's the objects that result from calling READ on that
> text.  There are no parentheses in that result; there are only lists,
> symbols, strings, integers, etc., etc.  That's why you can define
> things like the infix syntax reader Jacek Generowicz mentioned.
>
> READ can call arbitrary user-defined functions to parse the textual
> representation into internal objects, so it can read any surface
> syntax you want.

I appreciate that Lisp can be customized to process any alternative input that you like.  And
that the default S-expression notation may be abstracted away from the underlying language
semantics.

However, seems to me that out of the box there is this default S-expression notation that is
rather pervasive to programming in Lisp.  If you pick up a book on a particular dialect of
Lisp or a manual on a particular implementation they all begin with this same default syntax.
This default notation, it seems to me and many others, requires more parentheses than you
normally have occasion to use in Algol-like languages such as Pascal, C or Python.

Although I personally did not find these differences to be a problem when I used to program in
Lisp I sympathize with those who find them strange and unfamiliar.  I feel the default
S-expression syntax represent a legitimate barrier to entry, however slight, to new people
approaching and learning the language for the first time.

You and others obviously do not agree.

So be it.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From ronald at forte-intl.com  Sat Apr 20 18:36:04 2002
From: ronald at forte-intl.com (Ronald Cole)
Date: 20 Apr 2002 15:36:04 -0700
Subject: Problem embedding python on Red Hat Linux 7.2
References: <a9q49n$tq7$1@news.datasync.com>
Message-ID: <m3662mngkr.fsf@yakisoba.forte-intl.com>

*newbie alert*

I have a program that calls Py_Initialize() and I cannot compile it
because I cannot seem to find the python library (-lpython?) in
/usr/lib to link it with.  Here's what Red Hat figured I needed:

$ rpm -qa | grep python
python-xmlrpc-1.5.1-7.x.3
python-1.5.2-35
python-devel-1.5.2-35
rpm-python-4.0.4-7x
python-popt-0.8.8-7.x.2

and I've scanned distribution cd's 1 and 2 looking for it but can't
seem to find any sign of a lib*.{a,so} in the rpms.

Please, someone take pity and clue me!

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald at forte-intl.com>      Phone: (760) 499-9142
President, CEO                             Fax: (760) 499-9152
My GPG fingerprint: C3AF 4BE9 BEA6 F1C2 B084  4A88 8851 E6C8 69E3 B00B


From mwh at python.net  Fri Apr 12 06:33:40 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 10:33:40 GMT
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org>
Message-ID: <lk4rihjgek.fsf@pc150.maths.bris.ac.uk>

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

> On Wed, Apr 10, 2002 at 07:22:32PM +0400, Oleg Broytmann wrote:
> > On Wed, Apr 10, 2002 at 02:54:20PM +0000, Michael Hudson wrote:
> > > >    Environment: SUN Sparc Ultra-1, Solaris 2.5.1, gcc 2.8.1.
> > >                                                         ^^^^^
> > > Is this negotiable?  I.e. could you try upgrading?  This is pretty old now.
> 
>    I upgarded my software, including gcc (to 2.95.3). Not much help:
> 
> gcc -s  -Xlinker --export-dynamic -o python \
>                 Modules/python.o \
>                 libpython2.2.a -lsocket -lnsl -ldl  -lpthread -lthread   -lm
> case $MAKEFLAGS in \
> *-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py -q build;; \
> *) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py build;; \
> esac
> running build
> running build_ext
> building 'struct' extension
> Segmentation Fault - core dumped

Yikes!  Can you dig into this?  I can't, not having access to a
relavent machine... (you did run make clean, didn't you?)

Cheers,
M.

-- 
  I also fondly recall Paris because that's where I learned to
  debug Zetalisp while drunk.                          -- Olin Shivers


From philh at comuno.freeserve.co.uk  Sat Apr  6 18:33:20 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 7 Apr 2002 00:33:20 +0100
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <slrnaav1e0.4vh.philh@comuno.freeserve.co.uk>

On Sat, 06 Apr 2002 14:29:46 GMT, Alex Martelli <aleax at aleax.it> wrote:
><posted & mailed>
>
>Guido van Rossum wrote:
>        ...
>> About the change rate in general: it's hard to find the proper pace.
>        ...
>> for faster change.  I have to pick a middle ground (and it's easier to
>
>Linus Thorvalds seems to have done pretty well with picking TWO middle 
>grounds -- two parallel tracks for Linux, "stable" and "experimental".
>
>Without the former he'd have lost the commercial interests, and "normal
>people" who DO need SOME level of stability; without the latter, a crucial 
>hard-core of neophile hackers might have gone away in search of other, more 
>dynamic projects.  With both tracks, things seems to have been smoother.
>
>Why don't we start thinking of a similarly dual-tracked Python -- one
>track aiming at stability (in a "middle ground" sensible sense: strong
>commitment to not breaking old code -- introducing small new features
>and new library modules _is_ OK, else one can use a bugfix-only-"track"
>for some previous release), one aiming at frequent releases, innovation,
>etc.

An excellent idea, IMO !!!

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From tim.one at comcast.net  Sun Apr 14 01:02:17 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 01:02:17 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <7xr8lic2yb.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEADOOAA.tim.one@comcast.net>

[Paul Rubin]
> A lot of stuff gets done because it needs to get done, rather than
> scratch an itch.

Do you have a concrete example?  I can't think of one.  In a world of
volunteers, "need" is in the judgment of each volunteer, and counts as
scratching their own itches to me.  In contrast, Guido said 1.5.2 "needed" a
bugfix release about 6 months after it was released, and nobody volunteered,
and 1.5.2 never got a bugfix release.  Volunteers can't be forced to do
anything; they do what they want to do, and for people playing the game
that's great for everyone.

> ...
> In that particular debate, a lot of people were asking for certain new
> features to be LEFT OUT of Python. I don't understand how someone can
> volunteer to contribute work, time, or money toward leaving something
> out.

That was barely mentioned in the "crushing amount of debate over 'stability'
[that] has gone by on Python-Dev the last week", so I don't know what "that
particular debate" means to you, but it surely doesn't refer to anything in
the paragraph you were responding to.  The debate on Python-Dev has about
backward compatibility, and especially about producing bugfix releases for
older releases for a much longer time.  Those things take work, and lots of
it.  Note that James (to whom I was replying) was making the case for why
commercial support is important to businesses -- nobody is going to buy a
support contract that only guarantees not to add new features.  Quite the
contrary, business wants bug fixes for older releases, and they *also* want
new features in old releases, especially in the libraries (like, e.g., SSL
support for Windows, and support for hot new protocols -- this is a
fast-moving field, if you haven't noticed <wink>).





From aleax at aleax.it  Sun Apr 28 03:22:06 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 28 Apr 2002 07:22:06 GMT
Subject: IMAP2IMAP Synchronization -- iis [Was: Re: iis: comparsion of two lists]
References: <20020426224924.GA1318@rehor.surfbest.net> <20020426232802.GA1927@rehor.surfbest.net> <JTry8.26517$8D3.706486@news1.tin.it> <20020428034636.GB1279@rehor.surfbest.net>
Message-ID: <yqNy8.71678$vF6.2141429@news2.tin.it>

Matej Cepl wrote:

> On Sat, Apr 27, 2002 at 06:51:21AM +0000, Alex Martelli wrote:
>> It would, but the first assignment is useless and the approach
>> needlessly complicated IMHO.
>> 
>> erase = [ m for m in proc_old if m not in loc_old ]
> 
> Concerning the first assignement, rather be (slightly) slower
> than sorry and your solution seems to work only with Python 2.*.
> I am still on 1.5.2. Do I have to _really_ download multimegabyte
> beast just to get above command?

Of course not, you can keep using a 4-years-old language version for
another 4 years (or 44).

Even in 1.5.2, in your use of:

    erase = ()
    erase = something else

the first assignment is useless.  Not an issue of performance, but of
Occam's razor: entities must not be multiplied needlessly, and a
statement that is totally useless is is a good candidate for pruning.

In 1.5.2, you didn't have list comprehensions, so the idiomatic approach
would use filter and (yech) a lambda with an injected name:

erase = filter(lambda m, loc_old=loc_old: m not in loc_old, proc_old)

My favorite would be to forego the hard-to-read lambda for a named
function (the name injection is inevitable in 1.5.2 or 2.0):

def not_in_loc_old(m, loc_old=loc_old): return m not in loc_old
erase = filter(not_in_loc_old, proc_old)

> Therefore, could I ask somebody for review of the below script,
> especially concerning the safety of messages and speed, please?

Code that claims it must be as fast as possible should NOT use
completely and obviously unnecessary assignments, as yours does.
Take out either the unneeded assignments or the speed claims.

It's silly to use two operators in expressions such as
        not (x in y)
when a single operator called not in suffices:
        x not in y

The in and not in operators with a list RHS are O(N) [take time
roughly proportional to the number of items in the RHS].  This is
likely to produce the speed bottleneck unless the lists involved
are very, VERY small.  For membership checking, prepare, not
lists, but dictionaries, and (in 1.5.2 and up to 2.1) use has_key.

Access to global names is quite a bit slower than access to
local names.

So the code you consider "ugly but as fast as possible" should become:

    def makeSet(msgDict):
        result = {}
        for m in msgDict.values():
            result[m[0]] = 1
        return result

    proc_new = makeSet(processed.new_msg)
    proc_old =  makeSet(processed.old_msg)
    loc_new = makeSet(local.new_msg)
    loc_old =  makeSet(local.old_msg)

    def filterSet(aSet, avoidSet):
        result = {}
        for m in aSet.keys():
            if not avoidSet.has_key(m):
                result[m]=1
        return result

    e = filterSet(proc_old, loc_old)
    d = filterSet(proc_new, loc_new)

    def valueInE(x, e=e, p=processed.old_msg):
        return e.has_key(p[x][0])
    erase = filter(valueInE, processed.old_msg.keys())

    def valueInD(x, d=d, p=processed.new_msg):
        return d.has_key(p[x][0])
    download = filter(valueInE, processed.new_msg.keys())

    return erase, download

By modestly increasing elegance, we should also get substantial
increases in performance unless your sets are all tiny (in which
case performance doesn't matter all that much).  Try -- performance
issues are always to be checked by measurement.


Alex



From cnd at ecn.purdue.edu  Mon Apr  1 10:22:29 2002
From: cnd at ecn.purdue.edu (Christopher N. Deckard)
Date: Mon, 01 Apr 2002 10:22:29 -0500
Subject: creating a module
Message-ID: <3CA87B34.CF1128F7@ecn.purdue.edu>

I'm having some difficulties making my python module available to
other programs besides itself.

Say I have a module called Thinger, and a basic file layout like:

Thinger (a directory)
       |- Thinger.py
          from Thinger_definitions import *
          from Commands.Command import Command
          from Commands.Command1 improt Command1
          class Thinger:
            methods...

       |-  Thinger_definitions.py (global definitions for Thinger 
                                   services.  Every other file has
                                   import * from
Thinger_definitions)

       |-  ThingerConnection.py (handles connections to a server)
           class ThingerConnection:
              methods...

       |-  Commands (a directory)
                   |- Command.py (base command class)
                      from Thinger_definitions import *
                      class Command:
                        methods...
                   |- Command1.py
                      from Thinger_definitions import *
                      class Command1(Command):
                        methods...

If I create a script in the Thinger directory, which uses the
Thinger class, then everything works fine.  If I make a script
outside of Thinger (parent directory) and do something like

test_script:
  from Thinger.Thinger import Thinger
  my_thing = Thinger()

Then this fails when the files in the Command directory try to
import Thinger_definitions.

Really, what I'd like to do is to be able to say in Thinger.py

from Commands import *

and that would import all of the Commands that were in there and
make them directly useable as Command(), Command1(), etc.

Another question along these lines is what shouls go in the
__init__.py script in each directory?  Right now I just have the
keyword 'pass' and that seems to do fine for most things.

Thanks,
-Chris

P.S.  If you could, please send mail to me as well as the group.


From tanzer at swing.co.at  Thu Apr  4 02:19:06 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 04 Apr 2002 09:19:06 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 07:58:22 MDT."
             <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
Message-ID: <E16t1WU-0007iC-00@tswings.swing.cluster>

"Mark McEahern" <marklists at mceahern.com> wrote:

> [Christian Tanzer]
> >   For me, Python's (old) way of handling boolean conditions is one of
> >   its great strengths. This wasn't always the case. When I started
> >   using Python, I was slightly put off by the lack of a boolean data
> >   type. But quite quickly I learned to appreciate Python's way of
> >   dealing with this issue.
>
> How does the PEP change "Python's (old) way of handling boolean conditions"?
>
> Correct me if I'm wrong, but under the PEP, you can still do this:
>
> 	l = range(10)
> 	if l:
> 		print "My, the list is not empty!"
>
> So what's different?

Fortunately Guido closed the case before I got around to reading your
question <wink>.

Let's move on.

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





From kosh at aesaeion.com  Wed Apr  3 16:01:26 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Wed, 3 Apr 2002 14:01:26 -0700 (MST)
Subject: IDE Wars, the prequel!
In-Reply-To: <wCJq8.24678$0T6.7328999@e3500-atl1.usenetserver.com>
Message-ID: <Pine.LNX.4.32.0204031359450.705-100000@aesaeion.com>

On Wed, 3 Apr 2002, Robert Oschler wrote:

> In a galaxy far, far away..
>
> I've been getting some IDE suggestions on the COLA forum.  The current batch
> seems to be:
>
> Gvim
> Emacsen
> ActiveState's ActivePython
>
> I'm going to look at them all this weekend, but I'd like to hear from those
> of you who have experiences they would like to share.
>
> thx

I like using kdevelop under KDE. I find it does a nice job of managing
projects and it allows simple committing of changes to CVS. Overall I have
not had problems with kdevelop and since it uses the kde io slaves you
can edit items over ftp/http/webdav/etc.

Designing and building web applications http://webme-eng.com





From grante at visi.com  Fri Apr  5 17:42:42 2002
From: grante at visi.com (Grant Edwards)
Date: Fri, 05 Apr 2002 22:42:42 GMT
Subject: file.getline()/cout problem
References: <3cae0463.46737479@mammoth.usenet-access.com>
Message-ID: <slrnaasa43.145.grante@tuxtop.visi.com>

In article <3cae0463.46737479 at mammoth.usenet-access.com>, A. Jones wrote:

> In trying to understand how file access might work in my instance, I
> wrote a small test program with the expectation that it would print
> off 35 lines of arbitrarily length of a file.
[...]
> #include <iostream.h>
> #include <string.h>
> #include <fstream.h>
> 
> ifstream testfile("QOD.qdb");
[...]

Perhaps my Python installation is broken, but this program
won't run at all!

$ python
Python 2.1.1 (#1, Aug 13 2001, 19:37:40) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> #include <iostream.h>
... #include <string.h>
... #include <fstream.h>
... 
>>> ifstream testfile("QOD.qdb");
  File "<stdin>", line 1
    ifstream testfile("QOD.qdb");
                    ^
SyntaxError: invalid syntax
>>> 



-- 
Grant Edwards                   grante             Yow!  An INK-LING? Sure --
                                  at               TAKE one!! Did you BUY any
                               visi.com            COMMUNIST UNIFORMS??


From paul at boddie.net  Wed Apr 10 09:52:44 2002
From: paul at boddie.net (Paul Boddie)
Date: 10 Apr 2002 06:52:44 -0700
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204090526.18675d02@posting.google.com> <0NCs8.63131$pT1.1810689@news1.tin.it> <23891c90.0204100109.4a44932b@posting.google.com> <RYTs8.2219$b62.57325@news1.tin.it>
Message-ID: <23891c90.0204100552.6b822307@posting.google.com>

Alex Martelli <aleax at aleax.it> wrote in message news:<RYTs8.2219$b62.57325 at news1.tin.it>...
> 

[Iterators]

> I've noticed they immensely help Python beginners that come to them
> without preconcepts.  Such beginners are surprised to learn that
> these are extremely recent additions to Python, so well they fit in.

Well, I was excited a few years ago when iterators appeared in C++, so
I suppose I should be even more excited now. :-) However, it's
arguably a bigger deal for the implementer of the iterator (or
generator, even) than the "users" or "clients" of them. Many beginners
just expect things to work in the way that iterators themselves
promote, and it was previously up to the implementer of a framework to
provide nice interfaces to "users"/"clients" - now, it's just easier
for the implementer.

[...]

> They're great progress over the previously murky "iteration protocol".

What protocol was that again? ;-)

[Applications which subclass built-in types]

> Or maybe not being used to them you just don't think of cases where
> they'd be handy (if you need to keep compatibility with 2.1, then
> of course you'd better NOT start thinking of such cases:-).  I.e.,
> Sapir-Whorf hypothesis somewhere between its weakest and strongest
> forms (though I still prefer the way Wittgenstein independently
> put it earlier -- "The limits of my language are the limits of
> my world").

It's probably best not to bring semi-controversial linguistic themes
into the discussion - Whorf allegedly had less of a grasp of the
languages he used to illustrate his hypothesis than I do of the most
recent Python features. ;-)

Seriously, I can't think of cases off the top of my head where
inheriting from an integer, floating point, complex number, string,
tuple, list or dictionary are so compelling. An "ordered dictionary"
can be useful, but we already had UserDict to define the interface for
that kind of thing.

[New features in the tutorial]

> Yes, just not in the tutorial (at least not until they're in the
> "current" language rather than needing import from the future).

That's part of my original argument, though. I get the impression that
some packages not only require Python 2.2 because they use new,
standard features, but also so that the author can drag users further
into the future by making use of not-yet-standard features, if only
for their own entertainment.

[...]

> AMK's mission is not to tell you WHY you can be interested in
> major stuff such as subclassing builtins, iterators, generators,
> though -- just to ensure you DO know easily what IS new.

It wouldn't be a bad thing to have such a source of brief rationales
for new features, however.

Paul


From garrett at bgb-consulting.com  Mon Apr  1 19:59:43 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Tue, 02 Apr 2002 00:59:43 GMT
Subject: PEP 262 (database of installed Python packages)
References: <3wgg8.685$Qo2.70120@news02.tsnz.net>
Message-ID: <3CA902AF.EF9C710E@bgb-consulting.com>


Andy Elvey wrote:
> 
> Hi all.
>   I've spent a _frustrating_ amount of time with a _Ruby_ package that can't
> find various files.  Now, I knew about PEP 262, so I was thinking in general
> terms,  it would be *truly* wonderful if there were a (standard?) database
> of  *all*  installed packages (Python, Ruby, and whatever else) - maybe in a
> directory called "search".   In Linux, this could be "/search" ,  in Win32
> "C:\search".

  If this is an approach worth taking, then broaden it further. Take advantage
of the native install DBs that already exist. RPMs on Linux, the MS Installer
on Windows, and some cheap hack where there is nothing to build on.

  You are correct that this is a general problem, but I really think that a
solution should either be fully generic, or fully specific. IE: Use the system
wide facilities, or use a language specific tool.

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From dman at dman.ddts.net  Thu Apr  4 09:29:23 2002
From: dman at dman.ddts.net (dman)
Date: Thu, 4 Apr 2002 08:29:23 -0600
Subject: [Tutor] How to check whether email /domain exists
In-Reply-To: <3CAADA7B.25443.509AE1@localhost>
References: <3CAADA7B.25443.509AE1@localhost>
Message-ID: <20020404142923.GA18719@dman.ddts.net>

On Wed, Apr 03, 2002 at 10:33:31AM +0200, A wrote:
| Hello,
| What is the simplest way how to check 
|  whether a given email address( or domain for this email) exists?
| Thanks for help

$ host -t mx <domain>
$ host -t a <domain>

You need to do a DNS lookup.  Sites SHOULD have a MX records that look
like :
    $ host -t mx dman.ddts.net
    dman.ddts.net       	MX	1 dman.ddts.net
    dman.ddts.net       	MX	10 dman2.ddts.net
The name on the left is the domain, the name on the right MUST point
to an A (or AAAA or A6 for IPv6) record, not a CNAME.  For example :
    $ host -t a dman.ddts.net
    dman.ddts.net       	A	65.107.69.216

I don't know what python modules/functions there are for performing
DNS queries.

You can _try_ to test for the existence of an email address by opening
an SMTP session with an MX for the domain and trying the "EXPN" or
"VRFY" commands, but many sites disable those because spammers can use
that to harvest valid addresses.

HTH,
-D

-- 

A)bort, R)etry, D)o it right this time





From fdrake at acm.org  Mon Apr  1 17:06:24 2002
From: fdrake at acm.org (Fred L. Drake)
Date: Mon,  1 Apr 2002 17:06:24 -0500 (EST)
Subject: [development doc updates]
Message-ID: <20020401220624.423C818EAD1@grendel.zope.com>

The development version of the documentation has been updated:

    http://www.python.org/dev/doc/devel/

Updated documentation on lexical scoping in the language reference.



From logiplexsoftware at earthlink.net  Tue Apr  9 15:40:28 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 9 Apr 2002 12:40:28 -0700
Subject: What has PEP 285 done to us?
In-Reply-To: <3CB2AADF.9050908@mxm.dk>
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net>
	<mailman.1018339297.32067.python-list@python.org>
	<3CB2AADF.9050908@mxm.dk>
Message-ID: <20020409124028.43ec4361.logiplexsoftware@earthlink.net>

On Tue, 09 Apr 2002 10:48:31 +0200
Max M wrote:

> The whole bool thing as seen from my point of view happened something 
> like this:

[snip]
> So when Guido wrote that the post was an overlong piece of FUD I more or 
> less agreed. But I was rather surprised by his choice of words, and 
> found it to be a rather rude brush-of. And I do believe that if I had 
> made a lot of work making an argument like Lauras, that I would feel a 
> bit bad about it being turned down that way.
> 
> It was hinted that discussions had taken place on python-dev, but I 
> don't subscribe to it so all I, and I guess most others, got was the 
> short reply from Guido.
> 
> So even though I agree that Guido didn't have to reply seperately to 
> each of Lauras points, I believe that a longer softer and more humane 
> reply would not have hurt in this case.
> 
> Online you often need to state what you find obvious so others won't 
> miss the crucial.
>
> So perhaps Guido need a little work with the people skills ;-) and the 
> rest of us need to act properly, which includes not getting upset when 
> others don't.

I agree and discussed this with Guido offlist and he admitted that perhaps he
was a bit too terse (mostly due to being busy and not wanting to respond to
every single post, especially when he felt that most posts were adequately
discussed by others).

Anyway, I'd just like to clarify that when I started this particular thread I
wasn't aiming my remarks at Guido, but the community in general. There were
several other people who were posting fairly inflammatory remarks on both
sides, some in particular being purely derogatory and not appropriate to the
discussion.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From jeff at ccvcorp.com  Mon Apr 22 16:14:04 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Mon, 22 Apr 2002 13:14:04 -0700
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com> <efb63dac.0204221151.6d866042@posting.google.com>
Message-ID: <MPG.172e0ed9611cc5119896c1@news.nwlink.com>

In article <efb63dac.0204221151.6d866042 at posting.google.com>, 
juenglin at informatik.uni-freiburg.de says...

> As I found out today, these operations are advertised as in-place
> operations in PEP 203 (see below). 

> [Cut out from PEP 203]
> 
>       They implement the same operator as their normal binary form,
>       except that the operation is done `in-place' when the left-hand
>       side object supports it...

The key point here being "when the left-hand side object supports 
it" -- which clearly implies that immutable objects will not be 
modified in place, since they do not support modification.

While there might be some theoretical argument for considering 
augmented assignment on immutable types to be an error, keep in 
mind that practicality beats purity.  Making augmented assignment 
on integers into an error would, in practice, be quite silly.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From max at alcyone.com  Sat Apr  6 15:52:58 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 06 Apr 2002 12:52:58 -0800
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com>  <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net> <mailman.1018120753.6552.python-list@python.org>
Message-ID: <3CAF602A.9FC57B55@alcyone.com>

Laura Creighton wrote:

> What I am trying to do is save people
> like Mark McEahern from his isValid function, which returns whether
> a username is valid or not.  On the day his boss demands a third
> value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
> he will thank me.

I really don't see what you think the big benefit is.  A Boolean is a
logical return value for a test function; I use them all the time, and
have seen them used very often on large projects.  In my experience,
it's pretty rare that a simple yes-or-no question needs to be expanded
to a multiple choice question; in those questions, at any rate, the
introduction of another function which returns an enum (or equivalent)
is probably more appropriate.

I also don't know of many cases where bosses get down to the level of
what individual functions return; I can't imagine the hell of a
micromanaging environment where a typical supervisor is really concerned
at that level of detail.

Further, let's take the case at face value:  You have a function that
returns a bool, and you realize later in the design that you need a
three-state function -- say, "yes," "no," and "unknown."  You're saying
that the person who uses a Boolean is punished for the implementation
change, but why?  Not only does the function in question need to be
changed to return the three-state value now, but now other code which
calls it _also_ needs to be changed, totally independent of whether or
not the original formulation was a Boolean, just an integer, or
something else.

An API change means that code which uses it needs to be touched; whether
you started by returning a Boolean or not doesn't change that fact.  If
you had started by using integers (say, 0, and 1) and need to add a
third value (maybe -1 or 2), if you started by using a Python bool then
the impact will be just as minimal as if you had used 0 and 1, since
True == 1 and False == 0.  I really don't see what benefit you're
getting at.

> That is what I mean by teaching people not to use bools -- teaching
> when not to use them. This is very hard to teach.  I know it because
> I have been doing it for years and years.   I even know _why_ it is
> hard to teach, but that doesn't make it any easier.

I really don't see where teaching bools is hard, particularly
considering it's a lot more straightforward than teaching them to return
arbitrary values for different conditions.  "If a function is a
yes-or-no question, return a bool."  Is that _really_ that hard?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From newsfroups at jerf.org  Fri Apr 12 17:52:07 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Fri, 12 Apr 2002 21:52:07 GMT
Subject: currying (was: Re: A trivial question about print)
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3CB756E4.5030404@jerf.org>

Huaiyu Zhu wrote:
> def f(a, b, c): return a+b+c
> f1 = curry(f, a, b)
> f1(c)
> map(curry(f, x, y), a)  # map(lambda z:f(x, y, z), a)
> 
> Huaiyu

Has anyone done any work on creating a more general curry function? One 
that might for instance curry the second of three positional arguments? 
Using the above definitions:

f2 = curry2(f, a, c)
f2(b)

I've used one-off curry functions for this, but I'm surprised that 
nobody seems to have considered this. (Maybe everybody just writes 
wrappers around functions?

def f2(a, c, b): return f(a,b,c)
f3 = curry(f, a, c)

)

Not a biggie, just curious. But I have had instances where I needed this 
type of currying.



From wilk at flibuste.net  Tue Apr 16 05:52:25 2002
From: wilk at flibuste.net (William Dode)
Date: Tue, 16 Apr 2002 11:52:25 +0200
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
	<3CBBD31D.AA2E653C@alcyone.com>
Message-ID: <20020416115225.1d9fa2d2.wilk@flibuste.net>

Le Tue, 16 Apr 2002 00:30:37 -0700
Erik Max Francis <max at alcyone.com> ?crivait:

> William Dode wrote:
> 
> > Is there a python api somewhere ?
> > Google is known to use python, maybe we must write to them to ask a
> > python
> > wrapper for their api ?
> 
> 	http://diveintomark.org/projects/#pygoogle
thanks, i write now to google to ask why they didnt spoke about.

bye

> 
> -- 
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.
> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.


-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From cliechti at gmx.net  Tue Apr 16 14:42:20 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 16 Apr 2002 20:42:20 +0200
Subject: input() vs raw_input()
References: <slrnabm6im.a8q.ak@ak.silmarill.org> <slrnabobfb.8ho.mlh@vier.idi.ntnu.no>
Message-ID: <Xns91F2D2D8A7229cliechtigmxnet@62.2.16.82>

mlh at vier.idi.ntnu.no (Magnus Lie Hetland) wrote in 
news:slrnabobfb.8ho.mlh at vier.idi.ntnu.no:
> In article <slrnabm6im.a8q.ak at ak.silmarill.org>, Andrei Kulakov wrote:
>>input() must not be mentioned in any tutorials for new users. It belongs
>>somewhere in the depths of the thickest reference manual. Where do new
>>users *learn* about it? I don't remember seeing it in any tutorial and
>>yet it pops up all the time.
> 
> I use it in my tutorials. I don't see what the fuss is about. You have
> to explain how it works (at least in a limited way) but so what?

do an
>>> input()

and then enter: '__import__("os").system("dir")'
without quotes. or replace "dir" with "echo y|format c:" or "rm -r *" when 
you have a couple of hours to spend ;-)

"input" is as evil as "eval" and "exec". (i don't want them to go away, but 
they should only be used by those who know what they're doing. maybe the 
docs should make that clearer.)

chris

-- 
Chris <cliechti at gmx.net>



From rjones at ekit-inc.com  Mon Apr  1 00:24:47 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 1 Apr 2002 16:24:47 +1100
Subject: Python/Linux
In-Reply-To: <xORp8.63088$Q7.49964592@e3500-atl1.usenetserver.com>
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <mailman.1017628802.8651.python-list@python.org> <xORp8.63088$Q7.49964592@e3500-atl1.usenetserver.com>
Message-ID: <200204010524.FAA03500@crown.off.ekorp.com>

On Mon, 1 Apr 2002 15:07, Robert Oschler wrote:
> So what then do you think?  BlueFish, Quanta, WebSphere Homepage builder?

I'm using Quanta Gold at the moment, and it rocks. Really good Zope 
functionality. Quite a nice editor all-round. Very responsive developers. 
Very cheap software :)

     http://www.thekompany.com/products/quanta/


   Richard



From peter at engcorp.com  Thu Apr 18 23:39:20 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 18 Apr 2002 23:39:20 -0400
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net>
Message-ID: <3CBF9168.AAE0FC46@engcorp.com>

"Frederick H. Bartlett" wrote:
> 
> I was recently introduced to OmniMark. One of our exercises was to take
> a plain text file of Hamlet and convert it to SGML.
> 
> So I did it in Python, too. But the best time I could get from Python
> was .57 sec, while OmniMark came in at .20 sec. What's the most
> efficient technique for Pythonesque regex-based text processing?

Hmmm... how fast do you need it to be?  Sounds to me like 0.57 seconds
is pretty darned fast.

Do you have specific goals, or are you just on a search for 
something faster?  Remember, "better is the enemy of good"
and the grass is always greener.

(See http://www.seds.org/~chrisl/akin.html )

-Peter


From guido at python.org  Wed Apr  3 14:54:37 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 14:54:37 -0500
Subject: [Python-Dev] Cross-posting
In-Reply-To: Your message of "Wed, 03 Apr 2002 14:36:42 EST."
             <20020403193642.GA2076@panix.com> 
References: <20020403193642.GA2076@panix.com> 
Message-ID: <200204031954.g33Jsbo26224@odiug.zope.com>

> Please be careful to avoid cross-posting between python-list and
> python-dev.  I really hate reading messages twice.

Too bad.

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




From phil at mrinfo.de  Tue Apr 16 12:11:13 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 18:11:13 +0200
Subject: How does xml.dom.minidom compare?
Message-ID: <a9hidh$j94$1@swifty.westend.com>

I saw some differences listed regarding xml.dom and xml.dom.minidom at
http://www.python.org/dev/doc/maint22/lib/dom-objects.html . I will probably
start to collect experiences soon but does anybody already have some
performance experiences? And are they practically interchangeable, so that
one could replace the one with the other at a later point without much
hassle?




From rasnick1 at att.net  Tue Apr 16 15:36:46 2002
From: rasnick1 at att.net (Jack Rasnick)
Date: Tue, 16 Apr 2002 19:36:46 GMT
Subject: Backup Script
Message-ID: <i3%u8.29433$Rw2.2216172@bgtnsc05-news.ops.worldnet.att.net>

Hi,
I'm looking for some example backup scripts to backup my Windows O/S.  What
I want to do is boot up Linux on a dual boot PC and run a python script that
backs up the Windows directories to a compressed file (zip, bzip, etc).

Does anyone know of any good scripts on the web or have a good script
example that I could use as a starting point.  I'm new to python, so I'm not
sure where to look.  I've tried google searches, but have had no luck.

Thanks - Jack

--
p. s. - To reply to me via email, remove the '1' from rasnick1.att.net  -
spam prevention at work here..




From stojek at part-gmbh.de  Wed Apr  3 03:48:35 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 03 Apr 2002 08:48:35 GMT
Subject: wxpython: problem with events and checklistbox
References: <3caa2be5.48482250@news.easynews.net> <Xns91E534FF3A655rcamesz@amesz.demon.nl>
Message-ID: <3caabb5e.1387250@news.easynews.net>

>Well the first hint would be to ask yourself if you really need that 
>event handler in the first place. If it can't cope with the program 
>setting or removing the checkmarks in the control, chances are it won't 
>handle user input any better or more gracefully. You might want to 
>reconsider your design.

Believe it of not, i changed the design ;-)
Anyway I think there should be a way to distiguish between events
triggered by the user (check with mouseclick) and triggered by a
function (List.Check(i,true)). At least there should be a way to kill
an event right after it had been raised by a function.

>
>But apart from that, there are two very simple solutions:
>
>1. Don't set the event handler in the __init__ of your class, but do it 
>after you've set all the checks you want.
Doesn't work for this problem, but is a very good hint for another
one.
>
>2. Use a flag somewhere in your class, and check (in another sense of 
>the word, no pun intended) that flag at the start of your event 
>handler. Simply do a return if the flag indicates the event handler 
>should be temporarily disabled. (You could even do event.Skip() prior 
>to that return if you want that event to propagate to the parent 
>window, but somehow I don't think that's needed here.)
That's what I did prior to my posting. But as I don't know very much
about eventhandling I thought there should be a better way.

Thanks to all of you.
Marcus



From greg at cosc.canterbury.ac.nz  Thu Apr 11 01:17:04 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 11 Apr 2002 17:17:04 +1200
Subject: ANN: Plex 1.1.2
Message-ID: <3CB51C50.B447D05A@cosc.canterbury.ac.nz>

I have released a minor update to Plex, a lexical
analysis module for Python:

http://www.cosc.canterbury.ac.nz/~greg/python/Plex/

The yield() method of the Scanner class has been
renamed to produce() for compatibility with Python
2.2 and later. (The old name is still available
when running under 2.1 or earlier, or under 2.2
when generators are not being used.)

I have also changed one of the tests slightly so
that it doesn't get confused by recent changes
in the way repr() displays strings containing
unprintable characters.

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


From lac at strakt.com  Fri Apr  5 08:59:12 2002
From: lac at strakt.com (Laura Creighton)
Date: Fri, 05 Apr 2002 15:59:12 +0200
Subject: Selecting cipher suites with socket.ssl 
In-Reply-To: Message from martin@v.loewis.de (Martin v. Loewis) 
    of "05 Apr 2002 09:23:09 +0200." <m3y9g21uea.fsf@mira.informatik.hu-berlin.de 
References: <84cff98e.0204041830.7c0d0171@posting.google.com <mailman.1017978613.9042.python-list@python.org  <m3y9g21uea.fsf@mira.informatik.hu-berlin.de 
Message-ID: <200204051359.g35DxCrq000593@ratthing-b246.strakt.com>

> Gerhard H<E4>ring <gh_pythonlist at gmx.de> writes:
>
> > It's not possible. For everything but the most simple client-side SSL, I
> > recommend you use a different SSL implementation for Python. m2crypto
> > seems to be quite mature, but there are others, too:
>
> My recommendation would be slightly different: Please modify the
> existing SSL support to better suite your needs, and contribute your
> changes. I believe this will help Python users better in the long run.
>
> Regards,
> Martin

Why should the OP do this rather than just get our PyOpenSSL, or m2crypto
(which also didn't work a year ago when we tried it, either, but I believe
does now)?  Why not put our PyOpenSSL (or m2crypto) into Python?  This is
a serious question which nobody has answered for nearly a year now.

Laura Creighton
   




From doyle at acrossthebigpond.com  Tue Apr  9 21:36:16 2002
From: doyle at acrossthebigpond.com (Kevin Doyle)
Date: Tue, 09 Apr 2002 20:36:16 -0500
Subject: Get Button Text
References: <3CB0D763.3090901@acrossthebigpond.com> <Xns91E9B00BFBC38JamesLLugojcom@216.148.53.96>
Message-ID: <3CB39710.7010208@acrossthebigpond.com>

Thanks James that worked great, exactly what I was looking for and a 
really nice straight forward example.

James Logajan wrote:
> Kevin Doyle <doyle at acrossthebigpond.com> wrote:
> 
>>I have code that creates x amount of Buttons on the fly.  Each buttons 
>>will have a unique Text displayed on it.
>>
>>I want to be able to get the text vale of the button when it is pressed.
>>
>>I also want to know is it possible to create just one function and bind 
>>this one function to all the buttons, and call the function with a 
>>reference to the button clicked as an argument.
> 
> 
> The secret is the use of lambda to define the command (there is another 
> way, but this has worked for me). Try this; hopefully it will become clear.
> 
> #-----------------------
> from Tkinter import *
> 
> buttonWidgets = []
> buttonLabels = [ "OK", "Cancel", "Whatever" ]
> 
> def YourFunctionCall(button_text):
>     print "You pressed ", button_text
>     # Just for the heck of it, delete the button pressed.
>     buttonLabels.remove(button_text)
>     UpdateButtons()
> 
> def UpdateButtons():
>     global buttonWidgets
>     global root
>     for button in buttonWidgets:
>         button.destroy()
>     buttonWidgets = []
>     for buttonLabel in buttonLabels:
>         button = Button(root, text = buttonLabel, command =
>                     lambda txt = buttonLabel: YourFunctionCall(txt))
>         button.pack(side = LEFT)
>         buttonWidgets.append(button)
> 
> root = Frame()
> root.pack()
> UpdateButtons()
> root.mainloop()



From sholden at holdenweb.com  Mon Apr 15 09:08:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 09:08:02 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <Xns91F0276A4EF3JamesLLugojcom@216.148.53.85>
Message-ID: <YiAu8.10898$Ji7.9263@atlpnn01.usenetserver.com>

"James Logajan" <JamesL at Lugoj.com> wrote ...
>
[ ... ]
> Well, you know business requirements: the product should fit in a shirt
> pocket, have a 23 inch screen, run on a 9V battery for 3 years, and be
> capable of towing your car. ;-) This is why I now always demand a
> _prioritized_ list of requirements from marketing (and they can't all have
> the same priority!)

Well, you forgot it should cost next to nothing and be hugely profitable.

inventing-the-23-inch-shirt-pocket-won't-be-easy-ly yrs  - steve






From goodger at users.sourceforge.net  Tue Apr  2 00:28:01 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Tue, 02 Apr 2002 05:28:01 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
Message-ID: <B8CEABB2.20C5A%goodger@users.sourceforge.net>

Here's a serious proposal, safe to post now that April Fool's is over.
Please read & comment.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net


PEP: 287
Title: reStructuredText Standard Docstring Format
Version: $Revision: 1.3 $
Last-Modified: $Date: 2002/04/02 03:50:38 $
Author: goodger at users.sourceforge.net (David Goodger)
Discussions-To: doc-sig at python.org
Status: Draft
Type: Informational
Created: 25-Mar-2002
Post-History: 02-Apr-2002
Replaces: 216


Abstract

    When plaintext hasn't been expressive enough for inline
    documentation, Python programmers have sought out a format for
    docstrings.  This PEP proposes that the reStructuredText markup
    [1]_ be adopted as the standard markup format for structured
    plaintext documentation in Python docstrings, and for PEPs and
    ancillary documents as well.  reStructuredText is a rich and
    extensible yet easy-to-read, what-you-see-is-what-you-get
    plaintext markup syntax.

    Only the low-level syntax of docstrings is addressed here.  This
    PEP is not concerned with docstring semantics or processing at
    all.  Nor is it an attempt to deprecate pure plaintext docstrings,
    which are always going to be legitimate.  The reStructuredText
    markup is an alternative for those who want more expressive
    docstrings.


Benefits

    Programmers are by nature a lazy breed.  We reuse code with
    functions, classes, modules, and subsystems.  Through its
    docstring syntax, Python allows us to document our code from
    within.  The "holy grail" of the Python Documentation Special
    Interest Group (Doc-SIG) [2]_ has been a markup syntax and toolset
    to allow auto-documentation, where the docstrings of Python
    systems can be extracted in context and processed into useful,
    high-quality documentation for multiple purposes.

    The proposed format (reStructuredText) is entirely readable in
    plaintext format, and many of the markup forms match common usage
    (e.g., ``*emphasis*``), so it reads quite naturally.  Yet it is
    rich enough to produce complex documents, and extensible so that
    there are few limits.

    The reStructuredText parser is available now.  The Docutils
    project is at the point where standalone reStructuredText
    documents can be converted to HTML; other output format writers
    will become available over time.  Work is progressing on a Python
    source "Reader" which will implement auto-documentation.  Authors
    of existing auto-documentation tools are encouraged to integrate
    the reStructuredText parser into their projects, or better yet, to
    join forces to produce a world-class toolset for the Python
    standard library.

    Tools will become available in the near future, which will allow
    programmers to generate HTML for online help, XML for multiple
    purposes, and perhaps eventually PDF/DocBook/LaTeX for printed
    documentation, essentially "for free" from the existing
    docstrings.  The adoption of a standard will, at the very least,
    benefit docstring processing tools by preventing further
    "reinventing the wheel".

    Eventually PyDoc, the one existing standard auto-documentation
    tool, could have reStructuredText support added.  In the interim
    it will have no problem with reStructuredText markup, since it
    treats all docstrings as plaintext.


Goals

    These are the generally accepted goals for a docstring format, as
    discussed in the Doc-SIG:

    1. It must be readable in source form by the casual observer.

    2. It must be easy to type with any standard text editor.

    3. It must not need to contain information which can be deduced
       from parsing the module.

    4. It must contain sufficient information (structure) so it can be
       converted to any reasonable markup format.

    5. It must be possible to write a module's entire documentation in
       docstrings, without feeling hampered by the markup language.

    reStructuredText meets and exceeds all of these goals, and sets
    its own goals as well, even more stringent.  See "Features" below.

    The goals of this PEP are as follows:

    1. To establish reStructuredText as a standard docstring format by
       attaining "accepted" status (Python community consensus; BDFL
       pronouncement).  Once reStructuredText is a Python standard,
       effort can be focused on tools instead of arguing for a
       standard.  Python needs a standard set of documentation tools.

    2. To address any related concerns raised by the Python community.

    3. To encourage community support.  As long as multiple competing
       markups are out there, the development community remains
       fractured.  Once a standard exists, people will start to use
       it, and momentum will inevitably gather.

    4. To consolidate efforts from related auto-documentation
       projects.  It is hoped that interested developers will join
       forces and work on a joint/merged/common implementation.

    5. To adopt reStructuredText as the standard markup for PEPs.  One
       or both of the following strategies may be applied:

       a) Keep the existing PEP section structure constructs (one-line
          section headers, indented body text).  Subsections can
          either be forbidden or supported with underlined headers in
          the indented body text.

       b) Replace the PEP section structure constructs with the
          reStructuredText syntax.  Section headers will require
          underlines, subsections will be supported out of the box,
          and body text need not be indented (except for block
          quotes).

       Support for RFC 2822 headers will be added to the
       reStructuredText parser (unambiguous given a specific context:
       the first contiguous block of a PEP document).  It may be
       desired to concretely specify what over/underline styles are
       allowed for PEP section headers, for uniformity.

    6. To adopt reStructuredText as the standard markup for
       README-type files and other standalone documents in the Python
       distribution.


Rationale

    The lack of a standard syntax for docstrings has hampered the
    development of standard tools for extracting and converting
    docstrings into documentation in standard formats (e.g., HTML,
    DocBook, TeX).  There have been a number of proposed markup
    formats and variations, and many tools tied to these proposals,
    but without a standard docstring format they have failed to gain a
    strong following and/or floundered half-finished.

    Throughout the existence of the Doc-SIG, consensus on a single
    standard docstring format has never been reached.  A lightweight,
    implicit markup has been sought, for the following reasons (among
    others):

    1. Docstrings written within Python code are available from within
       the interactive interpreter, and can be 'print'ed.  Thus the
       use of plaintext for easy readability.

    2. Programmers want to add structure to their docstrings, without
       sacrificing raw docstring readability.  Unadorned plaintext
       cannot be transformed ('up-translated') into useful structured
       formats.

    3. Explicit markup (like XML or TeX) is widely considered
       unreadable by the uninitiated.

    4. Implicit markup is aesthetically compatible with the clean and
       minimalist Python syntax.

    Proposed alternatives have included:

    - XML [3]_, SGML [4]_, DocBook [5]_, HTML [6]_, XHTML [7]_

      XML and SGML are explicit, well-formed meta-languages suitable
      for all kinds of documentation.  XML is a variant of SGML.  They
      are best used behind the scenes, because they are verbose,
      difficult to type, and too cluttered to read comfortably as
      source.  DocBook, HTML, and XHTML are all applications of SGML
      and/or XML, and all share the same basic syntax and the same
      shortcomings.

    - TeX [8]_

      TeX is similar to XML/SGML in that it's explicit, not very easy
      to write, and not easy for the uninitiated to read.

    - Perl POD [9]_

      Most Perl modules are documented in a format called POD -- Plain
      Old Documentation.  This is an easy-to-type, very low level
      format with strong integration with the Perl parser.  Many tools
      exist to turn POD documentation into other formats: info, HTML
      and man pages, among others.  However, the POD syntax takes
      after Perl itself in terms of readability.

    - JavaDoc [10]_

      Special comments before Java classes and functions serve to
      document the code.  A program to extract these, and turn them
      into HTML documentation is called javadoc, and is part of the
      standard Java distribution.  However, the only output format
      that is supported is HTML, and JavaDoc has a very intimate
      relationship with HTML, using HTML tags for most markup.  Thus
      it shares the readability problems of HTML.

    - Setext [11]_, StructuredText [12]_

      Early on, variants of Setext (Structure Enhanced Text),
      including Zope Corp's StructuredText, were proposed for Python
      docstring formatting.  Hereafter these variants will
      collectively be call 'STexts'.  STexts have the advantage of
      being easy to read without special knowledge, and relatively
      easy to write.

      Although used by some (including in most existing Python
      auto-documentation tools), until now STexts have failed to
      become standard because:

      - STexts have been incomplete.  Lacking "essential" constructs
        that people want to use in their docstrings, STexts are
        rendered less than ideal.  Note that these "essential"
        constructs are not universal; everyone has their own
        requirements.

      - STexts have been sometimes surprising.  Bits of text are
        unexpectedly interpreted as being marked up, leading to user
        frustration.

      - SText implementations have been buggy.

      - Most STexts have have had no formal specification except for
        the implementation itself.  A buggy implementation meant a
        buggy spec, and vice-versa.

      - There has been no mechanism to get around the SText markup
        rules when a markup character is used in a non-markup context.

    Proponents of implicit STexts have vigorously opposed proposals
    for explicit markup (XML, HTML, TeX, POD, etc.), and the debates
    have continued off and on since 1996 or earlier.

    reStructuredText is a complete revision and reinterpretation of
    the SText idea, addressing all of the problems listed above.


Features

    Rather than repeating or summarizing the extensive
    reStructuredText spec, please read the originals available from
    http://structuredtext.sourceforge.net/spec/ (.txt & .html files).
    Reading the documents in following order is recommended:

    - An Introduction to reStructuredText [13]_

    - Problems With StructuredText [14]_ (optional for those who have
      used StructuredText; it explains many markup decisions made)

    - reStructuredText Markup Specification [15]_

    - A Record of reStructuredText Syntax Alternatives [16]_ (explains
      markup decisions made independently of StructuredText)

    - reStructuredText Directives [17]_

    There is also a "Quick reStructuredText" user reference [18]_.

    A summary of features addressing often-raised docstring markup
    concerns follows:

    - A markup escaping mechanism.

      Backslashes (``\``) are used to escape markup characters when
      needed for non-markup purposes.  However, the inline markup
      recognition rules have been constructed in order to minimize the
      need for backslash-escapes.  For example, although asterisks are
      used for *emphasis*, in non-markup contexts such as "*" or "(*)"
      or "x * y", the asterisks are not interpreted as markup and are
      left unchanged.  For many non-markup uses of backslashes (e.g.,
      describing regular expressions), inline literals or literal
      blocks are applicable; see the next item.

    - Markup to include Python source code and Python interactive
      sessions: inline literals, literal blocks, and doctest blocks.

      Inline literals use ``double-backquotes`` to indicate program
      I/O or code snippets.  No markup interpretation (including
      backslash-escape [``\``] interpretation) is done within inline
      literals.

      Literal blocks (block-level literal text, such as code excerpts
      or ASCII graphics) are indented, and indicated with a
      double-colon ("::") at the end of the preceding paragraph (right
      here -->)::

          if literal_block:
              text = 'is left as-is'
              spaces_and_linebreaks = 'are preserved'
              markup_processing = None

      Doctest blocks begin with ">>> " and end with a blank line.
      Neither indentation nor literal block double-colons are
      required.  For example::

          Here's a doctest block:

          >>> print 'Python-specific usage examples; begun with ">>>"'
          Python-specific usage examples; begun with ">>>"
          >>> print '(cut and pasted from interactive sessions)'
          (cut and pasted from interactive sessions)

    - Markup that isolates a Python identifier: interpreted text.

      Text enclosed in single backquotes is recognized as "interpreted
      text", whose interpretation is application-dependent.  In the
      context of a Python docstring, the default interpretation of
      interpreted text is as Python identifiers.  The text will be
      marked up with a hyperlink connected to the documentation for
      the identifier given.  Lookup rules are the same as in Python
      itself: LGB namespace lookups (local, global, builtin).  The
      "role" of the interpreted text (identifying a class, module,
      function, etc.) is determined implicitly from the namespace
      lookup.  For example::

          class Keeper(Storer):

              """
              Keep data fresher longer.

              Extend `Storer`.  Class attribute `instances` keeps track
              of the number of `Keeper` objects instantiated.
              """

              instances = 0
              """How many `Keeper` objects are there?"""

              def __init__(self):
                  """
                  Extend `Storer.__init__()` to keep track of
                  instances.  Keep count in `self.instances` and data
                  in `self.data`.
                  """
                  Storer.__init__(self)
                  self.instances += 1

                  self.data = []
                  """Store data in a list, most recent last."""

              def storedata(self, data):
                  """
                  Extend `Storer.storedata()`; append new `data` to a
                  list (in `self.data`).
                  """
                  self.data = data

      Each piece of interpreted text is looked up according to the
      local namespace of the block containing its docstring.

    - Markup that isolates a Python identifier and specifies its type:
      interpreted text with roles.

      Although the Python source context reader is designed not to
      require explicit roles, they may be used.  To classify
      identifiers explicitly, the role is given along with the
      identifier in either prefix or suffix form::

          Use :method:`Keeper.storedata` to store the object's data in
          `Keeper.data`:instance_attribute:.

      The syntax chosen for roles is verbose, but necessarily so (if
      anyone has a better alternative, please post it to the Doc-SIG).
      The intention of the markup is that there should be little need
      to use explicit roles; their use is to be kept to an absolute
      minimum.

    - Markup for "tagged lists" or "label lists": field lists.

      Field lists represent a mapping from field name to field body.
      These are mostly used for extension syntax, such as
      "bibliographic field lists" (representing document metadata such
      as author, date, and version) and extension attributes for
      directives (see below).  They may be used to implement docstring
      semantics, such as identifying parameters, exceptions raised,
      etc.; such usage is beyond the scope of this PEP.

      A modified RFC 2822 syntax is used, with a colon *before* as
      well as *after* the field name.  Field bodies are more versatile
      as well; they may contain multiple field bodies (even nested
      field lists).  For example::

          :Date: 2002-03-22
          :Version: 1
          :Authors:
              - Me
              - Myself
              - I

      Standard RFC 2822 header syntax cannot be used for this
      construct because it is ambiguous.  A word followed by a colon
      at the beginning of a line is common in written text.

    - Markup extensibility: directives and substitutions.

      Directives are used as an extension mechanism for
      reStructuredText, a way of adding support for new block-level
      constructs without adding new syntax.  Directives for images,
      admonitions (note, caution, etc.), and tables of contents
      generation (among others) have been implemented.  For example,
      here's how to place an image::

          .. image:: mylogo.png

      Substitution definitions allow the power and flexibility of
      block-level directives to be shared by inline text.  For
      example::

          The |biohazard| symbol must be used on containers used to
          dispose of medical waste.

          .. |biohazard| image:: biohazard.png

    - Section structure markup.

      Section headers in reStructuredText use adornment via underlines
      (and possibly overlines) rather than indentation.  For example::

          This is a Section Title
          =======================

          This is a Subsection Title
          --------------------------

          This paragraph is in the subsection.

          This is Another Section Title
          =============================

          This paragraph is in the second section.


Questions & Answers

    Q1: Is reStructuredText rich enough?

    A1: Yes, it is for most people.  If it lacks some construct that
        is required for a specific application, it can be added via
        the directive mechanism.  If a common construct has been
        overlooked and a suitably readable syntax can be found, it can
        be added to the specification and parser.

    Q2: Is reStructuredText *too* rich?

    A2: For specific applications or individuals, perhaps.  In
        general, no.

        Since the very beginning, whenever a markup syntax has been
        proposed on the Doc-SIG, someone has complained about the lack
        of support for some construct or other.  The reply was often
        something like, "These are docstrings we're talking about, and
        docstrings shouldn't have complex markup."  The problem is
        that a construct that seems superfluous to one person may be
        absolutely essential to another.

        reStructuredText takes the opposite approach: it provides a
        rich set of implicit markup constructs (plus a generic
        extension mechanism for explicit markup), allowing for all
        kinds of documents.  If the set of constructs is too rich for
        a particular application, the unused constructs can either be
        removed from the parser (via application-specific overrides)
        or simply omitted by convention.

    Q3: Why not use indentation for section structure, like
        StructuredText does?  Isn't it more "Pythonic"?

    A3: Guido van Rossum wrote the following in a 2001-06-13 Doc-SIG
        post:

            I still think that using indentation to indicate
            sectioning is wrong.  If you look at how real books and
            other print publications are laid out, you'll notice that
            indentation is used frequently, but mostly at the
            intra-section level.  Indentation can be used to offset
            lists, tables, quotations, examples, and the like.  (The
            argument that docstrings are different because they are
            input for a text formatter is wrong: the whole point is
            that they are also readable without processing.)

            I reject the argument that using indentation is Pythonic:
            text is not code, and different traditions and conventions
            hold.  People have been presenting text for readability
            for over 30 centuries.  Let's not innovate needlessly.

        See "Section Structure via Indentation" in "Problems With
        StructuredText" [14]_ for further elaboration.

    Q4: Why use reStructuredText for PEPs?  What's wrong with the
        existing standard?

    A4: The existing standard for PEPs is very limited in terms of
        general expressibility, and referencing is especially lacking
        for such a reference-rich document type.  PEPs are currently
        converted into HTML, but the results (mostly monospaced text)
        are less than attractive, and most of the value-added
        potential of HTML is untapped.

        Making reStructuredText the standard markup for PEPs will
        enable much richer expression, including support for section
        structure, inline markup, graphics, and tables.  In several
        PEPs there are ASCII graphics diagrams, which are all that
        plaintext documents can support.  Since PEPs are made
        available in HTML form, the ability to include proper diagrams
        would be immediately useful.

        Current PEP practices allow for reference markers in the form
        "[1]" in the text, and the footnotes/references themselves are
        listed in a section toward the end of the document.  There is
        currently no hyperlinking between the reference marker and the
        footnote/reference itself (it would be possible to add this to
        pep2html.py, but the "markup" as it stands is ambiguous and
        mistakes would be inevitable).  A PEP with many references
        (such as this one ;-) requires a lot of flipping back and
        forth.  When revising a PEP, often new references are added or
        unused references deleted.  It is painful to renumber the
        references, since it has to be done in two places and can have
        a cascading effect (insert a single new reference 1, and every
        other reference has to be renumbered; always adding new
        references to the end is suboptimal).  It is easy for
        references to go out of sync.

        PEPs use references for two purposes: simple URL references
        and footnotes.  reStructuredText differentiates between the
        two.  A PEP might contain references like this::

            Abstract

                This PEP proposes adding frungible doodads [1] to the
                core.  It extends PEP 9876 [2] via the BCA [3]
                mechanism.

            References and Footnotes

                [1] http://www.example.org/

                [2] PEP 9876, Let's Hope We Never Get Here
                    http://www.python.org/peps/pep-9876.html

                [3] "Bogus Complexity Addition"

        Reference 1 is a simple URL reference.  Reference 2 is a
        footnote containing text and a URL.  Reference 3 is a footnote
        containing text only.  Rewritten using reStructuredText, this
        PEP could look like this::

            Abstract
            ========

            This PEP proposes adding `frungible doodads`_ to the
            core.  It extends PEP 9876 [#pep9876]_ via the BCA [#]_
            mechanism.

            .. _frungible doodads: http://www.example.org/

            .. [#pep9876] `PEP 9876`__, Let's Hope We Never Get Here

            __ http://www.python.org/peps/pep-9876.html

            .. [#] "Bogus Complexity Addition"

        URLs and footnotes can be defined close to their references if
        desired, making them easier to read in the source text, and
        making the PEPs easier to revise.  The "References and
        Footnotes" section can be auto-generated with a document tree
        transform.  Footnotes from throughout the PEP would be
        gathered and displayed under a standard header.  If URL
        references should likewise be written out explicitly (in
        citation form), another tree transform could be used.

        URL references can be named ("frungible doodads"), and can be
        referenced from multiple places in the document without
        additional definitions.  When converted to HTML, references
        will be replaced with inline hyperlinks (HTML <A> tags).  The
        two footnotes are automatically numbered, so they will always
        stay in sync.  The first footnote also contains an internal
        reference name, "pep9876", so it's easier to see the
        connection between reference and footnote in the source text.
        Named footnotes can be referenced multiple times, maintaining
        consistent numbering.

        The "#pep9876" footnote could also be written in the form of a
        citation::

            It extends PEP 9876 [PEP9876]_ ...

            .. [PEP9876] `PEP 9876`_, Let's Hope We Never Get Here

        Footnotes are numbered, whereas citations use text for their
        references.

    Q5: Wouldn't it be better to keep the docstring and PEP proposals
        separate?

    A5: The PEP markup proposal may be removed if it is deemed that
        there is no need for PEP markup, or it could be made into a
        separate PEP.  If accepted, PEP 1, PEP Purpose and Guidelines
        [19]_, and PEP 9, Sample PEP Template [20]_ will be updated.

        It seems natural to adopt a single consistent markup standard
        for all uses of structured plaintext in Python, and to propose
        it all in one place.

    Q6: The existing pep2html.py script converts the existing PEP
        format to HTML.  How will the new-format PEPs be converted to
        HTML?

    A6: One of the deliverables of the Docutils project [21]_ will be
        a new version of pep2html.py with integrated reStructuredText
        parsing.  The Docutils project will support PEPs with a "PEP
        Reader" component, including all functionality currently in
        pep2html.py (auto-recognition of PEP & RFC references).

    Q7: Who's going to convert the existing PEPs to reStructuredText?

    A7: A call for volunteers will be put out to the Doc-SIG and
        greater Python communities.  If insufficient volunteers are
        forthcoming, I (David Goodger) will convert the documents
        myself, perhaps with some level of automation.  A transitional
        system whereby both old and new standards can coexist will be
        easy to implement (and I pledge to implement it if necessary).

    Q8: Why use reStructuredText for README and other ancillary files?

    A8: The reasoning given for PEPs in A4 above also applies to
        README and other ancillary files.  By adopting a standard
        markup, these files can be converted to attractive
        cross-referenced HTML and put up on python.org.  Developers of
        Python projects can also take advantage of this facility for
        their own documentation.

    Q9: Won't the superficial similarity to existing markup
        conventions cause problems, and result in people writing
        invalid markup (and not noticing, because the plaintext looks
        natural)?  How forgiving is reStructuredText of "not quite
        right" markup?

    A9: There will be some mis-steps, as there would be when moving
        from one programming language to another.  As with any
        language, proficiency grows with experience.  Luckily,
        reStructuredText is a very little language indeed.

        As with any syntax, there is the possibility of syntax errors.
        It is expected that a user will run the processing system over
        their input and check the output for correctness.

        In a strict sense, the reStructuredText parser is very
        unforgiving (as it should be; "In the face of ambiguity,
        refuse the temptation to guess" [22]_ applies to parsing
        markup as well as computer languages).  Here's a design goal
        from "An Introduction to reStructuredText" [13]_:

            3. Unambiguous.  The rules for markup must not be open for
               interpretation.  For any given input, there should be
               one and only one possible output (including error
               output).

        While unforgiving, at the same time the parser does try to be
        helpful by producing useful diagnostic output ("system
        messages").  The parser reports problems, indicating their
        level of severity (from least to most: debug, info, warning,
        error, severe).  The user or the client software can decide on
        reporting thresholds; they can ignore low-level problems or
        cause high-level problems to bring processing to an immediate
        halt.  Problems are reported during the parse as well as
        included in the output, often with two-way links between the
        source of the problem and the system message explaining it.

    Q10: Will the docstrings in the Python standard library modules be
         converted to reStructuredText?

    A10: Over time, with the help of the developer community, many
         modules will be converted.  Some modules may never be
         converted.  A future toolset will have to allow for
         incompleteness.


References & Footnotes

    [1] http://structuredtext.sourceforge.net/

    [2] http://www.python.org/sigs/doc-sig/

    [3] http://www.w3.org/XML/

    [4] http://www.oasis-open.org/cover/general.html

    [5] http://docbook.org/tdg/en/html/docbook.html

    [6] http://www.w3.org/MarkUp/

    [7] http://www.w3.org/MarkUp/#xhtml1

    [8] http://www.tug.org/interest.html

    [9] http://www.perldoc.com/perl5.6/pod/perlpod.html

    [10] http://java.sun.com/j2se/javadoc/

    [11] http://docutils.sourceforge.net/mirror/setext.html

    [12] http://dev.zope.org/Members/jim/StructuredTextWiki/FrontPage

    [13] An Introduction to reStructuredText
         http://structuredtext.sourceforge.net/spec/introduction.txt

    [14] Problems with StructuredText
         http://structuredtext.sourceforge.net/spec/problems.txt

    [15] reStructuredText Markup Specification
         http://structuredtext.sourceforge.net/spec/reStructuredText.txt

    [16] A Record of reStructuredText Syntax Alternatives
         http://structuredtext.sourceforge.net/spec/alternatives.txt

    [17] reStructuredText Directives
         http://structuredtext.sourceforge.net/spec/directives.txt

    [18] Quick reStructuredText
         http://structuredtext.sourceforge.net/docs/quickref.html

    [19] PEP 1, PEP Guidelines, Warsaw, Hylton
         http://www.python.org/peps/pep-0001.html

    [20] PEP 9, Sample PEP Template, Warsaw
         http://www.python.org/peps/pep-0009.html

    [21] http://docutils.sourceforge.net/

    [22] From "The Zen of Python (by Tim Peters)",
         http://www.python.org/doc/Humor.html#zen

    [23] PEP 216, Docstring Format, Zadka
         http://www.python.org/peps/pep-0216.html


Copyright

    This document has been placed in the public domain.


Acknowledgements

    Some text is borrowed from PEP 216, Docstring Format [23]_, by
    Moshe Zadka.

    Special thanks to all members past & present of the Python Doc-SIG.



Local Variables:
mode: indented-text
indent-tabs-mode: nil
sentence-end-double-space: t
fill-column: 70
End:



From abagaric_NOSPAM at rest-art.hr  Mon Apr 29 10:40:10 2002
From: abagaric_NOSPAM at rest-art.hr (Ante Bagaric)
Date: Mon, 29 Apr 2002 16:40:10 +0200
Subject: Tkinter question
Message-ID: <aajm5s$4vn$1@as201.hinet.hr>

Is there anything in Tkinter (or tcl/tk in general) similar to
java.awt.Robot class?







From aleax at aleax.it  Fri Apr 26 18:34:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 22:34:46 GMT
Subject: slices: foo[i:j:k] => TypeError
References: <mailman.1019856470.541.python-list@python.org>
Message-ID: <aCky8.66523$vF6.2038703@news2.tin.it>

Roman Neuhauser wrote:
        ...
>>>> foo[0:6:2]
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: sequence index must be integer
>>>> 
> 
> What's up? Is that a bug, or am I dumb?

http://www.python.org/dev/doc/devel/lib/typesseq.html only
mentions x:y slices; the x:y:z form is supported only by
Numeric, I believe.  The underlying problem is that list &c
still use the deprecated __getslice__ (or equivalent thereof)
rather than accepting slice objects in __getitem__ (&c for
set and del).


Alex



From huangcg at 163.net  Wed Apr  3 21:16:14 2002
From: huangcg at 163.net (huangcg)
Date: 3 Apr 2002 18:16:14 -0800
Subject: OK
Message-ID: <e8d5abee.0204031816.2bfec035@posting.google.com>

Is python out of fashion?


From claird at starbase.neosoft.com  Thu Apr 18 14:27:05 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Apr 2002 13:27:05 -0500
Subject: problems with Perl; changing to Python
References: <MPG.17291f2bfcb85301989684@news.rz.uni-karlsruhe.de>
Message-ID: <ACDB0B346DCD7075.DAF98D35E08A2E98.866BF0CF32FAFE8C@lp.airnews.net>

In article <MPG.17291f2bfcb85301989684 at news.rz.uni-karlsruhe.de>,
Uwe Mayer  <merkosh at hadiko.de> wrote:
			.
			.
			.
>Perl is great for scripts. Perl is great for programs with less than 
>perhaps 200 lines. A program should be aware of its memory usage; for a 
>script I don't care.
>
>I need a language that can be used to write programs with. GUI, and 
>commandline tools. When the program run in background they shouldn't 
>take up more memory than necessary.
>F.e. I wrote a Perl script checking for changes on homepages, expiry 
>date, etc in regular intervals. This "program" is 600 lines long and 
>uses 7mb of ram... too much in my oppinion.
>Can I do that in Python?
>
>I'll need to write programs which have to be extensible. The program 
>may, during calculation require up to 500 mb ram. Fine with me. On the 
>other hand I'll neet to write something working in background and then 
>(depending on its task and size) 1-2 MB should be enough. 
>What about Python here?
>
>Installing Perl modules was a torture. I am sitting behind a firewall 
>and in order to install the modules (compiling them did never work on my 
>Windows 2k mashine) I had to run my own DNS server and tunnel the 
>connection. Will that work better with Python? (f.e. Proxy support would 
>be sufficient!)
>
>And for the GUI programming: I used Perl/Tk previously, so Tkinter would 
>be most fammiliar, I guess. What other GUI packages can Python offer me 
>that will work (perhaps even in compiled .exe format) on windows 
>mashines? What about features like "minimize to tray" or "always on 
>top" that I wasn't able to accomplish in Perl/Tk?
>
>I need a language where I can rely on OOP if I need to. A language where 
>there's a difference between class- and instance- varriables. 
>How about multiple inheritance (perl, c++) or interfaces (jave) in 
>python?
>
>Are there any applications written in Python?
>Am I right here with Python or will I have to stick to a hybid C++/Java 
>solution?
			.
			.
			.
Quite a few people are successful writing large
applications in Perl.  Python's better in this
role, though, I believe any objective observer
will agree.  In fact, I claim Python is better
than C++ or Java (!) for large-scale programming.
Please, whatever you do, don't regard C++ and
Java as paragons of effectiveness.

The run-time memory footprint of Python is
roughly comparable to Perl's.  It's good that
you're sensitive to this; however, most people
find modern equipment runs moderate applications
coded in either Perl or Python just fine, with-
out hitting memory limits, and we have
relatively few measurements from real applica-
tions.  My summary:  I speculate that memory
limits are less of a practical problem than you
appear to believe they are.

Python's about the most flexible language in the
world in its bindings with GUI toolkits.  C
barely has as many.  <URL: http://
starbase.neosoft.com/~claird/comp.lang.python/python_GUI.html >
has more on the subject.

Installation of non-standard modules is a juve-
nile art.  You might have much better, or worse,
experience with Python, as compared to your Perl
history.  It's hard to predict.  I generally
have been happier with Python, in this regard,
but there's quite a range of results.

It *is* possible to "minimize to tray" and "keep
on top" programmatically in Perl/Tk.  I'm working
on these for the official FAQ, by the way.

OOP in Python is quite pleasant.
-- 

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


From phr-n2002a at nightsong.com  Mon Apr 29 01:33:34 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 28 Apr 2002 22:33:34 -0700
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <7xlmb7gjbl.fsf@ruckus.brouhaha.com>

"Stephen J. Turnbull" <stephen at xemacs.org> writes:
> The consumer protection argument is the OSI line (a special case
> thereof), and rms don't have much truck with anything so economical
> and relativist.

RMS has many times described proprietary software as being like a
house with a locked basement, which only the original vendor could
access to fix the plumbing or furnace.  The freedom to have your
programs customized and maintained by anyone you want is an important
aspect of free software.  It's not just an "open source" thing; it's
an issue of liberty.


From mstern at aon.at  Tue Apr  9 09:36:41 2002
From: mstern at aon.at (Matthias Stern)
Date: Tue, 9 Apr 2002 15:36:41 +0200
Subject: Pass Python-List to C++ array
Message-ID: <3cb2f622$0$3698$6e365a64@newsreader02.highway.telekom.at>

I want to pass a a Python-list to a simple C++ array. Can I use the Python
function "PyArg_ParseTuple" somehow? Or is there another way?

thx, Mac




From kragen at pobox.com  Mon Apr 29 18:06:41 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 29 Apr 2002 18:06:41 -0400
Subject: comparing str's with ints
References: <20020429115856.R1840-100000@palanthas.neverending.org>
Message-ID: <83adrmrwge.fsf@panacea.canonical.org>

Frank Tobin <ftobin at neverending.org> writes:
> BUT, *this* does not make sense:
> >>> '5' > 3
> 1
> 
> What in the world is supposed to be the semantics of comparing a string
> with an integer?  Seems like a type trap...

Thus, heterogeneous collections are sortable.  This might be a wart;
once, > couldn't raise exceptions, and now we're stuck with its old
behavior in this case for compatibility.  It's kind of nice that
heterogeneous collections are sortable, though.



From mhammond at skippinet.com.au  Sat Apr 13 10:09:29 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 13 Apr 2002 14:09:29 GMT
Subject: KeyboardInterrupt errors in ASP pages
References: <508cb8ae.0204110045.431dadee@posting.google.com> <3CB550C9.1000902@mxm.dk>
Message-ID: <3CB83C5C.9040704@skippinet.com.au>

Max M wrote:
> B wrote:
> 
>> I am using Python as the language for creating server side asp pages.
>> The problem is that when using Response.Redirect(page), I seem to get
>> the message below randomly but frequently enough to be a problem.
>>
>> Has anyone else found this problem and is there a way around it. Is it
>> a bug in the Response.Redirect code?
>>
>> Python ActiveX Scripting Engine error '80020009'
>> Traceback (innermost last): KeyboardInterrupt
>> /Default.asp, line 0
> 
> 
> As far as I understand it's a known bug, but without any known solution. 
> I get it too at times.
> 
> I have just signed over a project to a customer on .asp/iis so I am a 
> bit worried about it. But I don't really know what to do about it.
> 
> Maybe I should write a test harness to pound the IIS to see if I can 
> provoke the error.

Yes please - that is exactly what I need.  I need some way to provoke 
the error often enough that I can afford to "wait" for it in a debugger. 
  Anything you can do to get a reliable repro would be a big help!

Mark.



From jimw1956 at cs.comnojunk  Tue Apr 16 02:21:15 2002
From: jimw1956 at cs.comnojunk (Jim Williams)
Date: 16 Apr 2002 06:21:15 GMT
Subject: input() vs raw_input()
References: <slrnabm6im.a8q.ak@ak.silmarill.org>
Message-ID: <20020416022115.24190.00003679@mb-mi.news.cs.com>

>input() must not be mentioned in any tutorials for new users. It belongs
>somewhere in the depths of the thickest reference manual. Where do new
>users *learn* about it?

I have a couple of books that do explain it...Python 2.1 Bible, and the Python
QuickStart guide to be exact.

Basically the input statement will accept ONLY a valid python expression as
input whereas the raw_input will accept whatever the user types in ( or at
least that's what I get from what I've read).  If you are wanting a statement
to get input from the user, then raw_input would be what you want to use.



From joachim.kaeber at imk.fraunhofer.de  Tue Apr  9 10:43:49 2002
From: joachim.kaeber at imk.fraunhofer.de (Joachim Kaeber)
Date: Tue, 09 Apr 2002 16:43:49 +0200
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com>
Message-ID: <3CB2FE25.86A37EA1@imk.fraunhofer.de>

DeepBleu wrote:

> Yes, I am using zipfile (from zipfile import ZipFile).  Sorry I stated zlib.
> Still with one text file in particular, WinZip reports "missing 2 bytes"
> DeepBleu

is the script run on Windows? If yes, is the zipfile opened
in binary mode?

-- 
Joachim Kaeber                Fraunhofer Institut Medienkommunikation
Schloss Birlinghoven                   phone: +49 (0) 2241 - 14 25 46
D-53754 Sankt Augustin                   fax: +49 (0) 2241 - 14 24 49
Germany                                  http://www.imk.fraunhofer.de


From jason at tishler.net  Tue Apr 16 10:36:54 2002
From: jason at tishler.net (Jason Tishler)
Date: Tue, 16 Apr 2002 10:36:54 -0400
Subject: python on cygwin
In-Reply-To: <056901c1e485$53125c20$6b01a8c0@ericlaptop>
References: <01d801c1e1b8$ac962e40$6b01a8c0@ericlaptop>
 <20020412121141.GA948@tishler.net> <056901c1e485$53125c20$6b01a8c0@ericlaptop>
Message-ID: <20020416143654.GC1368@tishler.net>

Eric,

On Mon, Apr 15, 2002 at 09:56:21AM -0400, eric wrote:
> Thanks Jason,

You are welcome.

> Is this something that needs to be fixed in the python disribution,

No.

> or something that needs to be fixed in cygwin?

Yes.

> Any prospect of this eventually working out of the box,

Yes, but see below...

> or is this something that will always need tweaking?

Hopefully not.

I have submitted a (large) patch to Cygwin's setup.exe to perform rebasing
automatically.  Although it was not accepted in its original form, it
should be (eventually) after some (major) rework.

Rebasing a set of files is easy.  Unfortunately, maintaining and
persisting free and used lists of DLL base addresses to enable multiple
rebases over time is not.  This is only exacerbated by the unavailability
of STL in Cygwin's setup.exe since it is built with gcc -mno-cygwin.

Since the above requires an effort on the order of tens of hours
(instead of just hours), I am having trouble finding the spare time to
implement this functionality.  However, I will eventually complete this
work since it is important to me, generally useful, and very needed by
the Cygwin community.

Offers to help will be greatly appreciated.  If interested, please review
the past discussions on the cygwin-apps list:

    http://cygwin.com/cgi-bin/htsearch?method=and&format=builtin-long&sort=time&config=htdig&restrict=%2Fml%2Fcygwin-apps%2F&exclude=&words=rebase

and specifically the first version of my Cygwin setup.exe rebase patch:

    http://sources.redhat.com/ml/cygwin-apps/2002-02/msg00010.html

Thanks,
Jason




From SBrunning at trisystems.co.uk  Thu Apr 18 04:18:02 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 18 Apr 2002 09:18:02 +0100
Subject: chopping csv-files (newbie)
Message-ID: <31575A892FF6D1118F5800600846864DCBCFE2@intrepid>

> From:	Bj?rn Ove Gr?tan [SMTP:bgrotanLOVELYSPAM at itea.ntnu.no]
> I have a csv-file..
> 
(snip details of requirement)
 
First rule is - don't re-invent the wheel.

You might suppose that you aren't the first person to need to parse CSV
files. A quick search at Parnassus (<http://www.vex.net/parnassus/>) would
prove you correct.

I have used a couple of the modules that you find here. Dave Cole's CSV is
faster, being in C. Laurence Tratt's ASV module is in pure Python. Both work
just fine.

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




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




From aahz at pythoncraft.com  Wed Apr  3 16:56:06 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 16:56:06 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <slrnaaktij.kr1.kamikaze@kuoi.asui.uidaho.edu>
Message-ID: <a8ftpm$pi$1@panix1.panix.com>

In article <slrnaaktij.kr1.kamikaze at kuoi.asui.uidaho.edu>,
Mark 'Kamikaze' Hughes <kamikaze at kuoi.asui.uidaho.edu> wrote:
>
>  The central problem with that is that Python doesn't have a ternary
>operator, which means that Python programmers use math on booleans more
>than usual.  I don't see a problem with this, anyway, as long as people
>know what the int values of True and False are.

That's a good point.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From mertz at gnosis.cx  Fri Apr 26 17:17:30 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Fri, 26 Apr 2002 17:17:30 -0400
Subject: [OT] What is Open Source?
References: <mailman.1019853592.15502.python-list@python.org>
 <iNcy8kKkX8fK092yn@gnosis.cx>
Message-ID: <qPcy8kKkXMmP092yn@gnosis.cx>         

Tim Peters <tim.one at comcast.net> wrote previously:
|Me too, but at least one lawyer has told me it is a fantasy world, i.e. that
|an individual in the United States cannot put their work in the public
|domain:  you can't disclaim your copyright even if you want to, unless it's
|work done for the government and "they" put it in the public domain.

All the Python PEP's seem to contain a disclaimer about being placed
into the public domain.  Do you think Python has a problem because of
this?  (I think some other Python-related documentation, but forget
exactly which).

Yours, Lulu...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s




From mlh at vier.idi.ntnu.no  Sun Apr  7 14:09:49 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 7 Apr 2002 18:09:49 +0000 (UTC)
Subject: Argument against iterable ints...
References: <mailman.1018117586.26522.python-list@python.org>
Message-ID: <slrnab12rd.cde.mlh@vier.idi.ntnu.no>

In article <mailman.1018117586.26522.python-list at python.org>, James
Althoff wrote:
[snip]
>Magnus,
>
>I don't see how the above is even recursive, much less infinitely recursive.
>Could you elaborate?

def flatten(seq):
    try: iter(seq)
    except: yield seq
    else:
        for sub in seq:
            for item in flatten(sub):
                yield item

Anyway, I guess I shouldn't frame this as an argument against the
rejected PEP -- but rather _for_ a char type (to avoid having the
string "s" consist of the string "s" which again consists of the
string "s" etc., instead of consisting of the char 'c' which is not a
sequence...) But with the uproar against bool, I dare not do so ;)

>Jim

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From shalehperry at attbi.com  Sat Apr 13 10:46:22 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 13 Apr 2002 07:46:22 -0700 (PDT)
Subject: Checking whether an async* dispatcher is "alive"
In-Reply-To: <slrnabggo4.o80.mlh@vier.idi.ntnu.no>
Message-ID: <XFMail.20020413074622.shalehperry@attbi.com>

On 13-Apr-2002 Magnus Lie Hetland wrote:
> In a server I'm working on, I need to check whether an asynchat
> dispatcher is still connected, i.e. whether I can send it data. I've
> tried to use try/except (with socket.error) but since async* catches
> that itself, It's no good.
> 
> Are there any methods/attributes that indicate whether I can send data
> on a dispatcher? (Possibly some method/attribute on its socket
> object?)
> 

The usual tactic is to send the smallest possible operation and see if you get
a response aka "heartbeating".  Several protocols have a noop command you can
send that the server acknowledges.




From christophertavares at earthlink.net  Tue Apr  9 01:32:42 2002
From: christophertavares at earthlink.net (Chris Tavares)
Date: Tue, 09 Apr 2002 05:32:42 GMT
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1018314090.9984.python-list@python.org>
Message-ID: <_1vs8.1787$CA6.114998@newsread1.prod.itd.earthlink.net>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:mailman.1018314090.9984.python-list at python.org...
[ ... snip ... ]
>
> I'm not sure that Pyrex will ever compile *all* valid Python
> code. It probably wouldn't be very effective at speeding up
> pure Python code anyway. The reason it deals with Python code
> at all is so that you can write code to convert between
> Python and C data structures -- the fact that you can
> compile pure Python with it is sort of a side effect.
>
[ ... snip ... ]

It could be interesting to see how far you could get with just the subset
you do compile, though. The Squeak people can compile their entire VM from
Smalltalk to C, and their translator doesn't even do classes or objects! And
they get a big, big win from it in terms of portability and ease of
experimentation.

So, you don't need to compile all of Python syntax to have a very
interesting project on your hands.

-Chris





From mgilfix at eecs.tufts.edu  Wed Apr 17 10:29:41 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 10:29:41 -0400
Subject: tcp/ip server
In-Reply-To: <3CBD2C81.B8F23E03@mathworks.fr>; from boris.huart@mathworks.fr on Wed, Apr 17, 2002 at 10:04:17AM +0200
References: <3CBD2C81.B8F23E03@mathworks.fr>
Message-ID: <20020417102941.L16954@eecs.tufts.edu>

  Check out the struct module (import struct). It'll probably help you
with what you want. Remember in general to convert any ints and stuff
to network order.

                 -- Mike

On Wed, Apr 17 @ 10:04, Boris Huart wrote:
> Hi all,
> 
> I am developping a TCP/IP server in python 1.5.2 and I would like it to
> send data (mainly floats) to a Java applet. Does anyone know how to make
> it work out ? For the moment the server understands the requests (in
> strings) from the applet but the
> applet doesn't understand anything from the server...
> 
> TIA,
> 
> Boris
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (boris.huart)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/dee4ced1/attachment.sig>

From stephen at xemacs.org  Thu Apr 18 03:24:12 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 18 Apr 2002 16:24:12 +0900
Subject: Importance of trademarks (was RE: [OT] What is Open Source?....)
References: <mailman.1019068721.15404.python-list@python.org>
Message-ID: <87n0w1qxk3.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Nick" == Nick Arnett <narnett at mccmedia.com> writes:

    Nick> The main reason patents don't work for software is that the
    Nick> length of protection is too long in most cases, so that by
    Nick> the time software patents expire, the technology is often
    Nick> worthless.

I don't think reducing term would help as much as you might hope.
There are two problems.  First is that in most cases of patented
software, clean-room reimplementation is trivial simply from seeing
the results.  In practice, they fail the non-obvious test (the PTO's
definition of "non-obvious" is itself non-obvious to sane people,
beware!), and distressingly often, the no-prior-art test.  This
reduces the set of genuinely patentable software enough that one
wonders if it's worth bothering at all.

Second, the transactions costs (for example, finding out there is a
patent, determining that you need to license it, and negotiating the
license) are far too high compared to the cost of development.  So
this means that in most cases it is _impossible_ to strike a sensible
balance between providing incentive for original development of
genuinely patentable software and reducing monopoly losses simply by
adjusting the term of the patent.

And the lack of a "clean-room" defense creates a chilling effect on
the whole industry because of the enormous cost of defending yourself
against accusations of infringement.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From jjb5 at cornell.edu  Wed Apr 10 16:35:10 2002
From: jjb5 at cornell.edu (Joel Bender)
Date: Wed, 10 Apr 2002 16:35:10 -0400
Subject: Extracting from a list
Message-ID: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>

Hi,


I would like to process the elements of a list that match a condition, 
then remove them.  I'll use integers in my example, but my list items 
are actually objects.

What's the "right" way to do this?  For example:

    >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
    >>> dothese = [i for i in lst if i < 5]
    >>> lst = [i for i in lst if i >= 5]
    >>> for item in dothese:
    ...     print item

I'd rather not scan the list twice, particularly if there's nothing to 
do.  I'd rather not build another double-linked list.

Is there something like this?

    >>> for item in lst:
    ...     if (item < 5):
    ...         print item
    ...         del lst[ xxxx ]

But the "trick" is figuring out what xxxx should be.  Using lst.remove() 
would work if the internals used "is" and not "is equal to" and the 
removal didn't modify the list where it would break the loop somehow.

I could rebuild the list for things that should not be processed:

    >>> newlst = []
    >>> for item in lst:
    ...     if (item < 5):
    ...         print item
    ...     else
    ...         newlst.append(item)
    >>> lst = newlst

But these seems like a lot of work when there's nothing in the list to 
process.


Joel


From rich at worldsinfinite.com  Tue Apr  2 07:43:46 2002
From: rich at worldsinfinite.com (Rich Harkins)
Date: Tue, 2 Apr 2002 07:43:46 -0500
Subject: PEP 285: Adding a bool type
In-Reply-To: <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>
References: <mailman.1017486169.15848.python-list@python.org> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>
Message-ID: <200204021243.g32ChkQ02118@localhost.localdomain>

On Monday 01 April 2002 04:49 pm, Steve Holden wrote:
> "Gustavo Niemeyer" <niemeyer at conectiva.com> wrote in message
> news:mailman.1017673984.1773.python-list at python.org...
>
> > Hi Martin!
> >
[snip]
>
> Arguably, most of that code should have read
>
>     if ret:
>         ...
>
> rather than
>
>     if ret == None:
>         ...
>
> since the latter has the flavor of "if expr==True".
>
> So, is breaking bad code more excusable than breaking good code ;-)?
>
> regards
>  Steve

Hmmm... this is slightly OT for the thread but I have made a bit of a habit 
of doing the following:

if result is None:		# As opposed to result == None
	stuff

This is somewhat out of paranoia, and slightly out of interest in speed (my 
recollection is that previous Python versions could end up calling __coerce__ 
on comparisons with None and in many of these cases I didn't really want 
that).  Is there any compelling reason not to do this?  Isn't None to be a 
singleton in perpetuity?

Rich



From member at dbforums.com  Thu Apr 11 16:12:16 2002
From: member at dbforums.com (indira)
Date: 11 Apr 2002 20:12:16 GMT
Subject: Monitoring running Windows Applications
References: <3c13b5b2.56161165@127.0.0.1>
Message-ID: <3cb5ee20$1@usenetgateway.com>

I have a similar problem. How can I read the contenets of an editbox in
a webpage using python? The edit boxes are contained in a frame. I tried
using Enumchildwindows, but it didn't work.



--
Posted via dBforums
http://dbforums.com


From phd at phd.pp.ru  Wed Apr 10 09:59:24 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 10 Apr 2002 17:59:24 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
Message-ID: <20020410175924.D19899@phd.pp.ru>

Hello!

   Environment: SUN Sparc Ultra-1, Solaris 2.5.1, gcc 2.8.1.

   Python 1.5.2, 2.0, 2.1{.1,.2,.3} and Python 2.2 all had been compiled
successfully.

   Python 2.2.1 would not compile. The problem:

gcc -c -DNDEBUG -O2 -I. -I./Include -DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
In file included from Python/thread_pthread.h:9,
                 from Python/thread.c:113:
/usr/include/pthread.h:238: parse error before `destructor'
/usr/include/pthread.h:238: parse error before `)'
make: *** [Python/thread.o] Error 1

   The line in question /usr/include/pthread.h:238 is
int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
   Nothing suspicious.

   Any idea what can be wrong? Python/thread_pthread.h has

#ifdef __APPLE__
#define destructor xxdestructor
#endif

   before

#include <pthread.h>

   but it looks ok.

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




From brian at sweetapp.com  Wed Apr  3 19:37:09 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Wed, 3 Apr 2002 16:37:09 -0800
Subject: IDE Wars, the prequel!
In-Reply-To: <zrMq8.25970$0T6.7657756@e3500-atl1.usenetserver.com>
Message-ID: <000401c1db70$daa34dc0$445d4540@Dell2>

Robert wrote: 
> A little confused here.  So should I look at Komodo, or are you saying
> that I can install win32all on Linux and then run ActivePython?

>From the user's perspective, ActivePython and Komodo have nothing to do
with each other.

Komodo is an IDE for developing scripting languages, including Python
(and Perl, PHP, Tcl and XSLT). It works on Linux and Windows.

ActivePython is ActiveState's Python distribution, which is available
for several platforms (at least Windows and Linux). ActivePython is 99+%
compatible with PythonLabs Python but it has a nicer installer and some
extra modules. One of those extra modules is win32all, which is a Python
extension written by Mark Hammond. This extension, which includes the
PythonWin IDE, only works on Windows.

So you can install ActivePython on Linux but it will not include the
PythonWin IDE (though it does include IDLE).

Cheers,
Brian





From netzapper at magicstar.net  Tue Apr  2 03:50:56 2002
From: netzapper at magicstar.net (A. Jones)
Date: Tue, 02 Apr 2002 08:50:56 GMT
Subject: file.write[lines]() weirdness
Message-ID: <3ca96e30.52835380@mammoth.usenet-access.com>

I'm sure I'm a fool... but, I've spent about three hours now playing
with this code, trying to get it all worked out.

The code will execute correctly the first time it's run (with the
program creating all the files fresh).  However, each time after that
(whether in the same session or if I close the module and reload it),
it garbles what it's written before, and only the newly created
entries into qodDB are readable.  

It all works excellently, except for this one little problem (it's
even fast enough for me).  The file lengths come up with "correct"
numbers, so it's seeking to (what I think is) the correct place before
it starts writing.

Any insight would be much appreciated.

Aubrey Jones

(code below:
default values for global variables:
qodIndex : {}
localCount : 0)

def chunkSynch(start, stop):
    global qodIndex
    global localCount
    global qodDB
    del qodDB #closes local database from __main__ call open(,'r')
    bmeCount =
int(urllib.urlopen("http://cgi.bmezine.com/qod3.exe?search=/totalqa").readline()[1:-3])
#gets QA count
    if bmeCount > 0: #up to date?
            try: #retrieves QODv3.0 online database to filelike object
                qodNT =
urllib.urlopen("http://cgi.bmezine.com/qod3.exe?search=/retrieveqa&start="
+ `start` + "&finish=" + `stop`)
            except IOError: #error handler for connection problems
                print "Problem connecting to database."
                return None
    else: #already up to date.
        print "Local database up to date."
        return None  
###TROUBLE STARTS HERE
    curLength = os.stat("QOD.qdb")[6]
    print curLength
    cumLength = 0
    qodDB = file("QOD.qdb", 'w') #opens QOD local database file in
write mode
    qodDB.seek(curLength)
    while 1: #writes local entries
        temp = qodNT.readline() #holds the readline value temporarily
        if temp == "": #checks for end of file
            break
        qodDB.writelines([temp])
        temp1 = eval("(" + temp[0:-2] + ")") #I'm proud of this
workaround
        qodIndex[int(temp1[0])] = curLength + cumLength
        cumLength = cumLength + len(temp) + 1
        print cumLength
### I THINK THE TROUBLE ENDS HERE
    localCount = stop
    del qodDB
    del qodNT
    qodDB = file("QOD.qdb", 'r')


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From nospam at nospam.com  Mon Apr 29 09:06:17 2002
From: nospam at nospam.com (iserver7.com)
Date: Mon, 29 Apr 2002 14:06:17 +0100
Subject: Register domains at only $14.50
Message-ID: <1020085436.9521@newsfront1>

Register domains at only $14.50
Visit: http://www.iserver7.com


From sholden at holdenweb.com  Sat Apr 27 06:56:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 27 Apr 2002 06:56:11 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <tyf1yd25zgs.fsf@pcitapi22.cern.ch> <mailman.1019840809.5035.python-list@python.org>
Message-ID: <Kvvy8.51992$z55.49903@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote ...
>
[ ... ]
> 1. I think that's a fair statement.  Earlier in the week, someone
suggested I'd
> be more happy with Python after I was "reprogrammed" to think
"Pythonically"
> (which sounds like something L. Ron Hubbard would say.)  I understand that
> after a period of adjustment, people can function normally while wearing
> glasses that flip their entire field of vision upside down.  It takes a
few
> days to adjust after putting them on and another couple days to adjust
after
> taking them off.  So, although I can't disagree with your statement
(people
> adapt), I don't think it's a very strong point.  Humans are adept at
overcoming
> much more severe handicaps.
>
Don't quote me out of context, dammit. Just because you didn't detect the
humour didn't mean it wasn't there. And it waas "reeducated", not
"reprogrammed to think Pythonically".

People can also learn to ride bicycles where turning the handlebars to the
left turns the wheel to the right.

People can even learn to program in C, dammit.

Your problem is you just won't take those damned glasses off again now
you've got used to them.

[ ... ]
> 5. Finally, I think the problem here is that we're all making passionate,
> definitive statements about items that ultimately really boil down to
matters
> of personal preference.  And there's no accounting for taste.
>
True. But passion is a relative judgement. Programming languages and
operating systems stopped being a religious issue for me many years ago, and
so I tend to reserve passion for more appropriate activities.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From logiplexsoftware at earthlink.net  Thu Apr 25 14:56:31 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 25 Apr 2002 11:56:31 -0700
Subject: AS/400
In-Reply-To: <1e140f34.0204250934.2fee1f26@posting.google.com>
References: <1e140f34.0204250934.2fee1f26@posting.google.com>
Message-ID: <20020425115631.67c3ba8c.logiplexsoftware@earthlink.net>

On 25 Apr 2002 10:34:33 -0700
Terje Johan Abrahamsen wrote:

> I am trying to write some Python code that moves info from Excel into
> AS/400 throuhgt a display program called Rumba. (www.walldata.com)
> Basically what I want it to do (short version) is that I go to a cell
> in Excel, copy that cell, changes to the Rumba window, goes to
> coordinate x,x pastes the number there. Then goes back to Excel, gets
> a new number, goes to Rumba, changes screens by typing F12, enters
> info from Excel again in coordinate x,x. It is more to it than this,
> but if I can figure out this, I can figure out the rest.

Not familiar with AS/400, but in general, this sort of thing is very fragile. 
Unexpected user-interaction can cause the focus to get set to the wrong window,
coordinate, put one of the programs in an unexpected state, etc, with possibly
unhappy results.  Is it not possible to extract the data from Excel via
win32com, write it to a portable file and then import it on the AS/400 (perhaps
transfering the file via ftp)?


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From phr-n2002a at nightsong.com  Wed Apr 24 00:37:31 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 23 Apr 2002 21:37:31 -0700
Subject: Python's Lisp heritage
References: <3CC500AF.8BAE6F7E@cascade-sys.com> <LNBBLJKPBEHFEDALKOLCEEJAPAAA.tim.one@comcast.net> <mailman.1019619764.15284.python-list@python.org>
Message-ID: <7xsn5lg19w.fsf@ruckus.brouhaha.com>

Neil Schemenauer <nas at python.ca> writes:
> There's still time. :-)  Erik Naggum has suggested that Emacs should be
> re-implemented in Common Lisp.  That would be cool but I probably still
> wouldn't use it.

Emacs has already been re-implemented in Common Lisp and the CL
version is called Hemlock.  You can probably find it through Google.

GNU Emacs one of these years will supposedly change its extension
language from Emacs Lisp to Scheme (Guile).  Scheme of course supports
dynamic scope with the fluid-let function.


From martin at v.loewis.de  Mon Apr  1 12:44:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 19:44:54 +0200
Subject: Initialization problem of shared library in HP-UX11
References: <7aafbf19.0204010929.62d17d6a@posting.google.com>
Message-ID: <m37knr71pl.fsf@mira.informatik.hu-berlin.de>

flz at cfdrc.com (gother) writes:

> Anybody know how to resolve this problem?

I recommend to initialize the variable inside the module's init
function.

Regards,
Martin


From emile at fenx.com  Sun Apr 21 23:13:13 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 22 Apr 2002 03:13:13 GMT
Subject: correct Numeric behavior?
References: <m3hem4si40.fsf@loco.ucdavis.edu>
Message-ID: <ddLw8.50679$HH5.4150@rwcrnsc51.ops.asp.att.net>

"Rick Ree" <rick at bioinformatics.org> wrote in message
news:m3hem4si40.fsf at loco.ucdavis.edu...
> The following seems counterintuitive:
>
> >>> import Numeric
> >>> a = Numeric.array([0])
> >>> a
> array([0])
> >>> not a
> 1
> >>> b = Numeric.array([1])
> >>> not b
> 0
>


Oh GREAT!  Another way to test for False!

;-)


--

Emile van Sebille
emile at fenx.com

---------



From borcis at geneva-link.ch  Sun Apr 28 04:16:09 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Sun, 28 Apr 2002 10:16:09 +0200
Subject: Python's Lisp heritage
References: <mailman.1019846689.31490.python-list@python.org>
Message-ID: <3CCBAFC9.C95E39A1@geneva-link.ch>

Brian Quinlan wrote:
> 
> def add(*args): return reduce(operator.add, args)

or 0.02$ :

def add(sum=0,*args) : return sum + (args and add(*args) or 0)


> add(1,2,3,4,5,6,7,8,9,10)
> 
> Cheers,
> Brian


From theller at python.net  Tue Apr  9 14:07:53 2002
From: theller at python.net (Thomas Heller)
Date: Tue, 9 Apr 2002 20:07:53 +0200
Subject: Thread/Process-safety of functions in the Standard Library, eg the dbm classes
References: <3cb32a63.17540718@news.t-online.de>
Message-ID: <a8vamd$vp443$1@ID-59885.news.dfncis.de>

"Gerson Kurz" <gerson.kurz at t-online.de> wrote in message news:3cb32a63.17540718 at news.t-online.de...
> I wonder if it is documented which functions of the Python Standard
> Library are threadsafe, and which not. More explicitly, I was having a
> look at the anydbm interface and am pretty sure that it is not safe to
> access from multiple-processes at the same time, and I don't even know
> whether or not it is safe for multiple threads.
>
> So, is there such a documentation?
>
> Furthermore, do you have any suggestions for these requirements: I
> want to be able to store some data (<100K), and access it from
> multiple processes in a safe way. Currently I'm using mySQL with
> mysqldb, which seems like heavy overkill for that amount of data.
> Plus, I want users to use my program without having to first install
> MySQL (or any other dbm for that matter).
>
> [A hardcore idea would be to make a "anydbm-server" that you can
> communicate with using sockets ;]

Sound like a job for Gadfly ;-)
Gadfly includes a server which communicates via sockets with the
backend, IIRC.

Thomas




From tdelaney at avaya.com  Tue Apr 30 23:00:02 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 1 May 2002 13:00:02 +1000 
Subject: Palm
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2BD@natasha.auslabs.avaya.com>

> From: Terje Johan Abrahamsen [mailto:terjej at mailandnews.com]
> 
> Is it possible to make Python programs that run on the Palm OS?
> Preferably stand alone programs, without a xxx kb interpreter?

http://www.python.org/download/download_other.html

or

http://www.google.com/search?hl=en&q=python+palm

In particular, look at the second entry.

Tim Delaney




From paulsid at shaw.ca  Sat Apr  6 14:46:45 2002
From: paulsid at shaw.ca (Paul Sidorsky)
Date: Sat, 06 Apr 2002 12:46:45 -0700
Subject: PEP 285: Adding a bool type
References: <200204061706.g36H6Udj017546@ratthing-b246.strakt.com>
 <200204061854.g36IsYH28765@pcp742651pcs.reston01.va.comcast.net>
 <200204061913.g36JDcdj017764@ratthing-b246.strakt.com>
Message-ID: <3CAF50A5.F7841D3C@shaw.ca>

Laura Creighton wrote:

> I think there is a time and a place for bools --  Describing things
> that have two states that can never, ever, have 3 or more states.
> These things are vanishingly close to non-existent, unless you are
> a mathematician in which case you may be playing with these sorts
> of things all the time.  What I am trying to do is save people
> like Mark McEahern from his isValid function, which returns whether
> a username is valid or not.  On the day his boss demands a third
> value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
> he will thank me.

I haven't been following this thread closely, but this caught my eye.  

I sincerely hope this isn't the approach you are using to teach
programming.  If it is, it's no wonder you have a problem teaching
bools.

An isValid() function answers the question "is this valid"?.  valid-but
is not an answer to that question.  There are only two answers:  yes and
no.  Or, if you like, True and False.  Returning valid-but in an
isValid() function would be like returning 6-if-you-add-1 as the answer
to 3+2.

The proper course of action here would be to add a second check,
isPaid(), to wherever the new information is needed.  This has the
following advantages:

1) Adding a new function call is generally far easier than modifying a
bunch of existing calls.
  
2) If any checks do not care about the Paid status, they need never be
modified.

3) The clarity of the old function is preserved.

4) The functionality of the old function is preseved, i.e. there is no
chance of introducing bugs in a working function caused by ripping its
guts apart.

You don't demolish a house when you just need to add a bedroom, so why
rebuild a perfectly good isValid() function?

> That is what I mean by teaching people not to use bools -- teaching
> when not to use them. This is very hard to teach.  I know it because
> I have been doing it for years and years.   I even know _why_ it is
> hard to teach, but that doesn't make it any easier.

Teach people instead that half-answers like valid-but do not have a
place in programming (and why they don't), and then maybe you won't even
have to teach them when not to use bools.

-- 
======================================================================
Paul Sidorsky                                          Calgary, Canada
paulsid at shaw.ca                        http://members.shaw.ca/paulsid/




From neuhauser at mail.cz  Tue Apr 30 09:31:40 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 15:31:40 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <6qn0vlbcpk.fsf@abnoba.intevation.de>
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de>
Message-ID: <20020430133140.GM325@roman.mobil.cz>

> From: Bernhard Herzog <bh at intevation.de>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: 30 Apr 2002 14:19:51 +0200
> 
> Roman Neuhauser <neuhauser at mail.cz> writes:
> 
> > > From: Bernhard Herzog <bh at intevation.de>
> > > Why should chr be an int method in the first place? It's not tied to
> > > ints all that much[1]. IMO it sits somewhere between ints (including
> > > int-like objects !) and strings so it shouldn't be a method of either
> > > class. Leaving it a function is perfectly fine, therefore, although
> > > putting it into the string module might be a good idea. The same would
> > > apply to ord, IMO.
> > 
> >     Ugh, that looks to me as appropriate as join() being a string
> >     method / module function.
> 
> Exactly. I still think that join as a function makes more sense than as
> a string method and for the same reasons as stated for chr above. Even
> if join is implemented as a string method.

    What I wanted to say was that join() should be a list method, not
    that it should be in __builtin__.
 
> > > [1] In fact, chr today accepts any object that can be converted
> > > successfully to an int, so that chr(42.5) does work, but that's not
> > > always desirable.
> > 
> >     Any object that can be converted to a character should have a chr()
> >     method.
> 
> Would the same reasoning apply to objects that can be converted to ints,
> for example. I.e. should strings have a method .int() instead of int
> taking a string as argument (among other types)? (forgetting for the
> moment, that it's actually implemented through the __int__ method :) )

    IMO yes.
 
> I see numbers and characters as quite different datatypes and a
> conversion between the two is right in the middle and neither more on
> the int side than on the str side. 
 
    chr() pukes on anything different than int or float AFAICT. int()
    insists on getting a single character. quite simple, IMO.

> The correspondence between characters depends on an encoding, something
> which chr and ord do not take into account at all, for example, and one
> couldn't expect that e.g. int knows about encodings, so that a chr
> method could take the encoding as argument (perhaps defaulting to
> ascii).
 
    I didn't say that chr(), whether it's in __builtin__ or int/float
    method should be charset aware. That's out of the scope of this
    discussion.

> Of course, whether a conversion operation should be a method, a function
> or a constructor is largely a matter of taste and is influenced by
> whatever other languages one is coming from. For someone with a
> C-background, characters are ints, so making chr an int method may seem
> appropriate from that perspective.

    I don't know whether my preference for chr() being a method of
    int/float objects is influenced by the fact that I like C(-like
    languages). I know that I feel that the ability to express itself as
    a string does belong to int/float objects. Vice versa for strings
    and int() etc.

-- 
FreeBSD 4.4-STABLE
2:49PM up 1 day, 5:15, 11 users, load averages: 0.07, 0.08, 0.05




From greg at turtleprod.com  Mon Apr 29 15:42:31 2002
From: greg at turtleprod.com (Greg Pierce)
Date: Mon, 29 Apr 2002 14:42:31 -0500
Subject: PythonCOM syntax question
Message-ID: <B0000071394@buford.turtleprod.com>

I'm trying to interact with the Navision Axapta ERP package via it's COM connector...and I'm having trouble with particular types of calls.  I'm hoping someone can help me if there's some syntax problem I'm having.

I'm able to create the Axapta COM objects and can call certain methods on them, but not others.  I'm no COM expert, but there is a difference in the method specs that I can see.  Those that I _cannot_ call all have a "get" in the API documentation for Axapta -- like this:

HRESULT get_Field( [in] VARIANT field, [out, retval] VARIANT *pVal );

Those that I _can_ call don't -- like this: 

HRESULT Call(  [in] BSTR bstrMethod, [in, optional] VARIANT v1,  [out, retval] VARIANT *pResult);

Is there some different syntax I should need to use in Python to access those "get" methods?

This code in VB works ( where "buffer" is an already created COM object) :

MyValue = buffer.Call( "openBalanceMST" )
AccountNumber	= buffer.Field("AccountNum")

But in Python, the "call" line works, and then the "field" line returns a "com_error: (-2147352573, 'Member not found.', None, None)"

Any ideas?  Did that explanation make sense?

Thanks,

g.






From sholden at holdenweb.com  Wed Apr 24 08:41:03 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 08:41:03 -0400
Subject: Asynchronous cgi that returns nothing. Possible? O:-)
References: <olvacuoe4oo1aa5gjj6sa4vsimdom3pabv@4ax.com> <MPG.172f49fb3aa52b3c9896c6@news.nwlink.com> <rphbcu8qlh4qgc1eonmrgts5ug7nv0u41v@4ax.com> <slrnacc5cs.tg8.jon+usenet@snowy.squish.net> <3umccu09j6dkui67hnppulv8ngk71g55s4@4ax.com> <slrnacctn8.s1r.jon+usenet@snowy.squish.net>
Message-ID: <z5yx8.275$z55.113@atlpnn01.usenetserver.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnacctn8.s1r.jon+usenet at snowy.squish.net...
> In article <3umccu09j6dkui67hnppulv8ngk71g55s4 at 4ax.com>, Fernando
Rodr?guez wrote:
> >> If they really must be separate scripts, why doesn't script (a)
> >> just output nothing at all to stdout?
> >
> > I didn't try it, but I think the browser wouldn't like a cgi url that
doesn't
> > return any html content, no? O:-)
>
> I thought they were running on the same machine, in which case the
> entire sequence of [run script a + run script b] would produce the
> correct CGI output.
>
> You are going to have to do as someone else suggested, use the Python
> standard library routines to issue another https request yourself from
> the server to run script (b).

You *could* use some clever client-side Javascript so that an onClick()
event on the form's submit button (or some other suitable GUI element) calls
a procedure that targets script (a) into an invisible frame and then submits
the form that triggers script (b) targeted into the main window.

Generally, browsers don't like null outputs, and some will complain visibly.
Using this scheme, however, script (a) can send a well-formed HTML response
that the user will never see. You could even include some more clever
Javascript in *that*, if you wanted, to check the output from script (b) as
it came in. Yawn.

tired-of-working-round-web-inadequacies-ly y'rs  - steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From cbrown at metservice.com  Thu Apr 25 16:33:16 2002
From: cbrown at metservice.com (Colin Brown)
Date: Fri, 26 Apr 2002 08:33:16 +1200
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net> <3cc6036f$1@news.nz.asiaonline.net> <mailman.1019617724.32183.python-list@python.org> <3cc67564$0$11927$edfadb0f@dspool01.news.tele.dk>
Message-ID: <3cc86899$1@news.nz.asiaonline.net>

On Wed, Apr24, 2002 at 21:02, Anders J. Munch wrote

> Python raw strings are delineated like this:
>
>   A raw string ends at the first matching quote character which is not
>   preceded by an odd number of consecutive backslashes.
>
Thanks Anders, a definition that I can understand.

> It would be great if I could have written this instead:
>   A raw string ends at the first matching quote character.
>
Indeed, intuitively that is how I expected raw string handling to work.

> Much simpler, every bit as useful and - dare I say it - more pythonic
> than the current semantics.
>
My sentiments exactly.





From magnus.heino at pleon.sigma.se  Wed Apr 17 08:32:54 2002
From: magnus.heino at pleon.sigma.se (Magnus Heino)
Date: Wed, 17 Apr 2002 14:32:54 +0200
Subject: minidom and =?ISO-8859-15?Q?=E5=E4=F6?= once again :-P
Message-ID: <WXdv8.2962$ZN4.5053@news2.bredband.com>


Hi.

I am (still) trying to get minidom and ??? to play together. And I guess 
the main problem really is me ;-)

But why doesnt this work? Shouldnt parseString eat what toxml produce?

/Magnus

[magnus at bombardier magnus]$ python2.2
Python 2.2 (#1, Jan 23 2002, 14:50:45)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import xml.dom.minidom
>>> d=xml.dom.minidom.parseString('<?xml version="1.0" 
encoding="iso-8859-15" ?><test>???</test>')
>>> d.toxml()
'<?xml version="1.0" ?>\n<test>\xe5\xe4\xf6</test>'
>>> d=xml.dom.minidom.parseString(d.toxml())
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 
965, in parseString
    return _doparse(pulldom.parseString, args, kwargs)
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/minidom.py", line 
952, in _doparse
    toktype, rootNode = events.getEvent()
  File "/usr/lib/python2.2/site-packages/_xmlplus/dom/pulldom.py", line 
256, in getEvent
    self.parser.feed(buf)
  File "/usr/lib/python2.2/site-packages/_xmlplus/sax/expatreader.py", line 
143, in feed
    self._parser.Parse(data, isFinal)
UnicodeError: UTF-8 decoding error: invalid data
>>>


From matt at mondoinfo.com  Fri Apr 19 20:27:02 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sat, 20 Apr 2002 00:27:02 GMT
Subject: Tkinter.Text.index() puzzlement
References: <3CC0B271.C76D3E95@nonesuch.com>
Message-ID: <qB2w8.46153$vm6.6102555@ruti.visi.com>

On Fri, 19 Apr 2002 17:12:33 -0700, David Smith <nospam at nonesuch.com>
wrote:

Dear David,

>I am trying to use the "index" method to remember places in a Text
>window, but it doesn't seem to work properly.
[. . .]

>from Tkinter import *
>root = Tk()
>text = Text(root)
>text.pack()
>print "Beginning: end at %s" % text.index(END)
[. . .]

>The printed output says:
>Beginning: end at 2.0
[. . .]

>I don't understand why the first line is claimed to be number 2, nor
>why the character positions are all zero.  Is this normal?

That depends on your definition of normal <wink>. It's the way that
the widget is intended to work so it's normal in that sense. But the
person who wrote the Tk text widget may have a definition of the
normal behavior of a text widget that's different from yours and mine.

What's happening is that the widget is "helpfully" adding a newline
to the widget even if it's empty:

>>> from Tkinter import *
>>> r=Tk()
>>> t=Text(r)
>>> t.pack()
>>> t.get("1.0",END)
'\n'

Regards,
Matt


From phd at phd.pp.ru  Mon Apr  8 08:03:07 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 8 Apr 2002 16:03:07 +0400
Subject: ACCEPTED: PEP 285
In-Reply-To: <3CB1846B.5050109@geneva-link.ch>; from borcis@geneva-link.ch on Mon, Apr 08, 2002 at 01:52:11PM +0200
References: <mailman.1017873312.18135.python-list@python.org> <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk> <3CAFAAA4.41EA7CF9@python.org> <+CcOX7A2gWs8EwPR@jessikat.fsnet.co.uk> <3CB1846B.5050109@geneva-link.ch>
Message-ID: <20020408160307.M4949@phd.pp.ru>

On Mon, Apr 08, 2002 at 01:52:11PM +0200, Boris Borcic wrote:
> Robin Becker wrote:
> >>>
> >>You just made my kill file.
> >>
> >>--Guido van Rossum (home page: http://www.python.org/~guido/)
> >>
> > wow, I thought I was already there since the division rumble.
> > Interesting how the irreverent spirit of Monty Python seems singularly
> > lacking these days.
> 
> I'd rather note the typical elegance and economy of the BDFL's designs;
> did you notice how he wrote "made my kill file" and not "made it to my
> kill file" ?

   That's why Python is Python as we know now :)

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




From cliechti at gmx.net  Sat Apr  6 15:44:15 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 21:44:15 +0100
Subject: beginners syntax question
References: <3caf47a5@news1.homechoice.co.uk> <3CAF4E4B.ECE28871@ntlworld.com>
Message-ID: <Xns91E8DD4F8F366cliechtigmxnet@62.2.16.82>

tony clarke <a.clarke11 at ntlworld.com> wrote in 
news:3CAF4E4B.ECE28871 at ntlworld.com:

> Hi
> 
> Try listing the objects  l=["object1','object2'..]
> then
> for i in l:
>  i.update()

almost, leave away the quotes:
l=[object1, object2]
 
>> I'm trying to call the update() method of several objects ie:
>>
>> object1.update()
>> object2.update()
>> etc
>>
>> I'd really like to do
>>
>> for x in range(7):
>>   "object"+str(x).update
>>
>> but can't figure outthe syntax, is this possible?

yes it is, but there are better ways.
why are you not storing the objects in a list
instead of naming them with numbers?

calling methods on objects in a list is shown above.

chris

-- 
Chris <cliechti at gmx.net>



From tanzer at swing.co.at  Wed Apr 10 03:35:45 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Wed, 10 Apr 2002 09:35:45 +0200
Subject: Off topic C question
In-Reply-To: Your message of "Tue, 09 Apr 2002 08:42:46 +0200."
             <200204090642.g396gkdj021548@ratthing-b246.strakt.com>
Message-ID: <E16vCdt-0003BS-00@tswings.swing.cluster>

Laura Creighton <lac at strakt.com> wrote:

> However, if your real goal is to hack C, and all the above is just one
> misused word -- you want K & R. Everybody wants K & R. If you find it
> tough going, you may want something else in addition to K & R, just to
> get you started, but you also really want K & R.

In addition, if you want to use C, I'd really recommend that you get
Andrew Koenig's `C Traps and Pitfalls'.

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





From bokr at oz.net  Thu Apr 11 18:02:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Apr 2002 22:02:12 GMT
Subject: Beginner's performance problem
References: <slrnab68mh.15p.gerhard@lilith.my-fqdn.de> <memo.20020411143211.1400B@a.a>
Message-ID: <a95154$ans$0@216.39.172.122>

On Thu, 11 Apr 2002 14:32 +0100 (BST), mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley) wrote:

>In article <slrnab68mh.15p.gerhard at lilith.my-fqdn.de>, gerhard at bigfoot.de 
>(Gerhard =?iso-8859-15?Q?H=E4ring?=) wrote:
>
>> I haven't tried but an approach using the re module should be really
>> fast, and shorter. Untested code follows:
>> 
>> import re
>> names = ["ALGORITHM_CLASS", ...]
>> 
>> text = open("ThirteenThousandLines.sql").read()
>> for name in names:
>>     regex = re.compile(name, re.I) # case insensitive search for name
>>     text = regex.sub(name, text)   # replace all occurences
>> print text
>
>Aha! I'd put off reading up on the regular expression stuff until later as 
>I thought it just did searches, as opposed to also doing in-place 
>replaces. Half an hour spent on the HOW-TO and I'm a lot better informed. 
>Add a couple of r'\b' to the above, and I ought to be laughing. 
>
>TVM to everyone who helped.
>
Best reread Alex's post carefully, until you understand the difference from the above ;-)

Regards,
Bengt Richter


From aahz at pythoncraft.com  Wed Apr  3 17:31:42 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Apr 2002 17:31:42 -0500
Subject: Yes and no (was Re: PEP 285: Adding a bool type)
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
Message-ID: <a8fvse$7rp$1@panix1.panix.com>

In article <3CAA5777.66E7C2D5 at python.org>,
Guido van Rossum  <guido at python.org> wrote:
>
>I'm reading (at least skimming) everything, but the more I read the
>reactions, the more I believe that this is just the response I have to
>expect for *every* change to the language I propose, and the more I'm
>tempted to follow my gut instincts.  Long ago (right after lambda :-)
>I learned to say "no" to proposals that didn't feel right to me, even
>if I couldn't express very clearly why.  I think it's time I learn to
>say "yes" even if the community doesn't see the light yet.

I note that you've certainly done that in the past.  ;-)  One critical
difference between "yes" and "no" in this context is that a "no" can be
retracted, but "yes" rarely can.  So I think that a bias toward "no" is
overall good.

But yes, we have to trust your instincts in the end; that's why we're
all using Python in the first place.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From maxm at mxm.dk  Mon Apr  8 10:38:41 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 08 Apr 2002 16:38:41 +0200
Subject: Discussion of Python midi library
References: <3CB19732.8060304@mxm.dk> <a8s8ru$qej$1@thorium.cix.co.uk>
Message-ID: <3CB1AB71.3040904@mxm.dk>

gbreed at cix.compulink.co.uk wrote:
> Max M wrote:
> 
>>The biggest challenge here is getting the internal representation of 
>>the midi data right.
> 
> You mean getting it to match the standard, or encapsulating it in the most 
> convenient way?

Well the standard is a pretty dumb iff format, so encapsulating it the 
most convenient way will be the key.

>>So generally I want to be able to load in the data into a Python 
>>specific objects representation that can be easily modified and then 
>>saved back out as a new midi file.
> 
> I'd like it to be generic enough that the code could be re-used to 
> transform live MIDI streams as well.

I believe that the parts of the project could be used for something like 
that.

> I think a few of us are interested, but we may not have the time to 
> contribute much.  I did get half way through converting my old C++ code, 
> but I don't think it's really good enough for anybody else to work with.  
> Unless you want an example of something that can read and write the files 
> successfully.

I expect to do all the work myself. At least until it is usefull enough 
that other might pick it up for their own use.

>>I allready have some code and some ideas.

> Where

I will do a writeup of my ideas, put the code online and get back to you.

regards Max M



From lellinghaus at yahoo.com  Thu Apr 18 14:17:32 2002
From: lellinghaus at yahoo.com (Lance Ellinghaus)
Date: Thu, 18 Apr 2002 13:17:32 -0500
Subject: DCOracle2 importing problems
References: <e2e9e807.0204171619.46de6db1@posting.google.com> <ubu2a0dbso0u66@corp.supernews.com>
Message-ID: <002b01c1e705$63606900$3201040a@angel1>

This could be caused by the CGI environment not being able to locate the
shared libraries for Oracle. Under Solaris, make sure the CGI environment is
exporting LD_LIBRARY_PATH that contains the path to the Oracle shared
libraries.

Lance Ellinghaus

----- Original Message -----
From: "DeepBleu" <DeepBleu at DeepBleu.org>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, April 18, 2002 12:58 PM
Subject: Re: DCOracle2 importing problems


> How are you importing the DCOracle2 module, and what is *the exact* error
> that you are getting?
> DeepBleu
>
> "Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote in message
> news:e2e9e807.0204171619.46de6db1 at posting.google.com...
> > Hi All, I can import the DCOracle2 module fine from the interactive
> > interpreter but my CGI script causes an internal server error. The
> > interactive interpreter is running on the same machine as my scripts.
> > A tail of the apache error log reveals that my script can't import the
> > DCOracle2 module.
> > Any ideas anyone?
> >
> > Hugh
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From aleax at aleax.it  Mon Apr  8 06:51:20 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 10:51:20 GMT
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it> <Xns91E98AAF1B9ABJamesLLugojcom@216.148.53.85>
Message-ID: <ICes8.72426$S52.2558330@news2.tin.it>

<posted & mailed>

James Logajan wrote:

> Alex Martelli <aleax at aleax.it> wrote:
>> Why don't we start thinking of a similarly dual-tracked Python -- one
>> track aiming at stability (in a "middle ground" sensible sense: strong
>> commitment to not breaking old code -- introducing small new features
>> and new library modules _is_ OK, else one can use a bugfix-only-"track"
>> for some previous release), one aiming at frequent releases,
>> innovation, etc.
> 
> This proposal seems to have good promise and if well executed would
> probably mollify a curmudgeon such as myself.

I do not think so, considering all of your recent remarks.  I'm trying
to point this out because [a] I may be misunderstanding the remarks and
I'd like to be put right in that case and [b] there is a risk of such a
scheme being considered as a magic bullet and causing disappointment
when it turns out it's not.

You had a script written on a given release, 1.5.2, which also needed
to work with completely different major releases, such as 2.0.  The
scheme I'd like to see would not ensure that a change of major release
number breaks no existing code: that would be too constraining, "lock
in" any existing implementation bug -- not the Python I dream of.

Further: I've repeatedly seen you complaining, quite apart from stuff
written for 1.5.2 not working identically in 2.0, about a very
different issue: that (e.g) 2.1 would let one use NEW things that
2.0 didn't have.  Again, as long as the new things don't break
previously correct code, I don't envisage the dual-tracking idea
as removing such occurrences.  As I see things, along the stable
track a minor release would try to ensure *backwards* compatibility,
but NOT freeze features, much less bug-fixes.

Linux's scheme surely has execution problems too (it IS a HUGE base
of code and a LOT of contributors after all), but even excluding
that, and judging by intention, you should see it does not meet your
objections (I don't know what WOULD meet them).  The idea is that a
program which runs correctly on (e.g.) 2.4.20 will keep running on
2.4.21 (no similar guarantee going from 2.5.20 to 2.5.21, i.e., on
the experimental track), but it's *FINE* if a program is written
that neets "2.4.21 or more" and DOESN'T run on 2.4.20 because of
some bug that was only fixed, or some small feature that was only
added, in 2.4.21.  As I read your remarks, this would not be
satisfactory to you.


Alex



From shippy at nmt.edu  Tue Apr 16 17:23:39 2002
From: shippy at nmt.edu (Jeff Shipman)
Date: Tue, 16 Apr 2002 15:23:39 -0600
Subject: using httplib
References: <3CBC9027.3010603@nmt.edu>
Message-ID: <3CBC965B.7030704@nmt.edu>

Nevermind. I was forgetting to calculate and
send the 'Content-length' header for my send()
string. Once I did that, I got back the INVALID
response.

-- 
Jeff "Shippy" Shipman     E-Mail: shippy at nmt.edu
Computer Science Major    ICQ: 1786493
New Mexico Institute of Mining and Technology
Homepage: http://www.nmt.edu/~shippy



From nospam at nonesuch.com  Mon Apr 22 14:27:11 2002
From: nospam at nonesuch.com (David Smith)
Date: Mon, 22 Apr 2002 11:27:11 -0700
Subject: Tkinter.Text.index() puzzlement
References: <3CC0B271.C76D3E95@nonesuch.com> <qB2w8.46153$vm6.6102555@ruti.visi.com> <3CC20EA2.1385A0E5@earthlink.net> <13285ea2.0204211424.19cda7b1@posting.google.com>
Message-ID: <3CC455FF.8E287462@nonesuch.com>

That seems to be the ticket.  Thank you.


"J.Jacob" wrote:
> 
> Maybe you can try using CURRENT or INSERT instead of END ?

-- 
David Smith
drs at labs dot agilent dot com  <- hopefully doesn't trigger spambots


From aleax at aleax.it  Tue Apr  9 05:47:16 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 09:47:16 GMT
Subject: calling class methods as functions
References: <3CB2A038.2468F842@gol.ge>
Message-ID: <EMys8.75730$S52.2688885@news2.tin.it>

Giorgi Lekishvili wrote:

> Hello!
> 
> I wonder if someone explaines tom how can I overcome the following
> problem in Python2.0 (I have to use this interpreter):

2.0 didn't have class methods : class methods were introduced in 2.2.

What you are calling a class method is an ordinary method.


> suppose, we have 2 classes, B, and C.

Unrelated.  OK.


> in class B we have a method "write"
> 
> in class C we have a method "write"
> 
> def write (self, arg):
>     import B
>     res=B.B.write(self, arg)    #res is a string being either "ok" or
> "failure"
>     del B
>     return res
> 
> 
> The interpreter gives an error, that class isntance must be given as 1st
> argument.
> 
> What's wrong?

B.B.write is an *unbound method* and it carries along the information
that it was extracted from class B, exactly in order to avoid you
accidentally calling it on objects it wasn't meant to be called on.

If you want to extract the method's underlying function, that's
B.B.write.im_func.  It's a very peculiar thing to do -- if write
COULD indeed be called on anything, you should not code it as a
method, within class B, but as a function within module B.  Still,
for whatever peculiar needs you may have, the language lets you
call B.B.write.im_func(self, arg) -- "on your head be it":-).


Alex



From pyth at devel.trillke.net  Sun Apr 28 21:02:16 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 03:02:16 +0200
Subject: Update locals()
In-Reply-To: <02042900255602.25256@arthur>
References: <20020427195725.GA16217@devel.trillke> <200204281716.TAA10331@prim.merlinux.de> <20020428205757.GH16217@devel.trillke> <02042900255602.25256@arthur>
Message-ID: <20020429010216.GJ16217@devel.trillke>

On Mon, Apr 29, 2002 at 12:25:56AM +0200, Alex Martelli wrote:
> > Thanks for this transformation. But of course a 'but' follows :-)
> > As soon as
> >
> > - you want to adapt the definition to a filter's method signature
> >
> > - have templated or parametrized code in the function body
> >
> > - want to see what actually gets defined
> >   (with my version i can easily write or print the
> >    actual function's definition and look at a real definition)
> >
> > it might get difficult or impossible to do it your way.
> 
> It might, but I asked you to provide an example, you did, and I recoded
> it in a way that I consider preferable (without any exec or eval).  If you
> want to continue this exercise, despite its proving rather sterile, please
> keep providing examples for me to recode, rather than handwaving
> generalizations as in those first two points.

Why is "adapting to the definition of a methods signature" handwaving?
Previously we were both using '*args,**kwargs' while
the adapted function definition ('access') does not have any elipsis.

ok. so the code would *basically* look like this:

   from inspect import formatargspec, getargspec
   for name in statefulList:
        sig = apply(formatargspec, getargspec(getattr(filters,name))) 
        sig = '(self,'+sig[1:]
        exec """
            def %(name)s %(sig)s:
                    ...
        """

Please note that i said *basically*. This does not work if 
filters.name is a callable class in which case the __init__ signature
is the reference.

Point two (templated/parametrized code) is clarified below.

> Point three is murky to me.  "The actual function definition" you can't
> either see or print with your method -- 
> ...
> Source code can be seen or printed at runtime without needing it to
> come from a string exec'd with default dictionaries, as is pretty
> obvious.  So, if this point has any importance, you should restate it.

Oh well. (Of course) i mean the templated function definition 
which i want to see expanded just as the parser&compiler sees it. 
Not the function which it delegates to (which i already showed
in source code...). Strange misunderstanding.

> > Writing a reusable helper function i could say something like
> >
> >     exec make_adapters(filters, statefulList, '''
> >     def %name (self, %args):
> >         """%__doc__"""
> >         return %sourcescope.%name (%args) (self)
> >     ''')
> >
> > or even better make the templated adapter definition a defaulted
> > argument. This provides more flexibility for the *caller* of the helper
> > function without ever having to modify the helper function.
> 
> I _thought_ I knew what the term "unmaintainable" meant.  This, however, 
> takes this important term to levels I had never dreamed of.  Any bug at
> all in this complicated, tightly-coupled nest of function and data, any 
> little discrepancy, will be hell to find out and fix.

1. It's seems at least as maintainable as - say - the various html-templating
   engines and embedded code stuff. Or DTML for that matter :-)
   (Agreed, it's not pretty)

2. I can write out (in make_adapters) the actual (expanded) definitions 
   to stdout or a file. There i can look at it as if i had written an 
   ordinary function (read: no templating/weird constructions). Easy to debug.

3. it's not that my whole program consists of these things. It's kept
   only to very few places. 

> ...
> strings of sourcecode are not the best ways to
> sling code around in a language where functions and classes are first-class 
> objects).  

What hinders make_adapters to return code objects or in other words:
since when is exec restricted to strings? Sorry, i don't get it.
exec can execute code objects as well as strings. 

> > Also i use the template 'exec in current scope' method to allow
> >
> > resultfilters = "isfile and nolink and fnmatch('data*')"
> >
> > to be compiled at run time into a function and then be
> > applied to - say - 15000 files. This technique is similar
> > to what the regular expressions module does with re.compile.
> 
> *NO*!  re.compile NEVER mucks with your local namespace.
> Any "similarity" may be with functions such as the built-in compile
> (see below), never with exec-in-local-scope, which *does*
> have uncontrollable alterations of namespace as its only reason to exist.

[ i must, i must, i must be much more careful with respect to 
  possible (evil) misinterpretations ]

OF COURSE i didn't mean that re.compile mucks with the local namespace.
This thread started (and still has this as the subject line !!!)
about "updating locals" with a re-pattern's match.groupdict(). 
Remember that we both came up with a 3-liner to achive the goal
AND that i AGREED that this is better than updating locals.

re.compile compiles a pattern into a pattern object which
can be used for searches. In the above case the helper
uses exec to dynamically construct a function object from
a pattern. That's why i said 'similar' technique. Why are
you often assuming the worst of all interpretations? 

The code that gets "isfile and nolink and fnmatch('data')" 
makes a function object using 'exec' in local scope.
It could use a dedicated dictionary, agreed. But why should it?

> > This may be a better example than the original one?!
> > At least i don't see how you can apply your techniques
> > to it without using slow 'evals' or a slow interpretation
> 
> You seem to be unaware that built-in function 'compile' is able to make 
> code objects, that eval is not slow at all when passed a code object, 
> and/or that the functions of module new also let you build functions.
> These are not "my" technique, btw, but Python's.

i am not unaware of this fact. Not at all. I just thought that providing
strings is more readable and didn't bother to send it through a 
'compile' statement.  

> > None.strip() in
> >
> > ...__dict__.get(name).__doc__.strip()
> >
> > would fail at run time if a filters's doc has not been defined.
> > Same with your code.
> 
> Yes, but later you access filter.name unconditionally, and filter
> is what is at the point of the ellipsis.  The 'get', with or without a
> default 2nd argument, has nothing to do with whether a specific
> named filter has or lacks a docstring, but only with whether a
> filter of that name exists in the filter module.  And if no filter of
> that name exists, the method is going to fail at runtime anyway.
> So I still don't see why you're using this peculiar construct.

sorry. my mistake. Wrong translation from thought to code.
i meant to say 

getattr(filters.__dict__.get(name), '__doc__', undoc).strip()

the name will be there and a runtime-error is ok otherwise. 
It's just the undefined __doc__ which i didn't want to
cause a runtime error. Let's drop this issue.

> > The only better thing i can imagine is if you tried a tiny bit
> > harder to see some *valid* points rather than only pin down invalid
> > points. Otherwise you give me the feeling that i am completly on the
> > wrong path. (or is this intended?!)
> 
> I do believe you are "completely on the wrong path" in wanting to
> use exec in local scope.  I am certainly open to being convinced,
> should some example be brought that I don't find a way to recode
> which seems preferable to me; it just appears to me to be highly
> unlikely for such an example to emerge.

You may be right. 

> > I just try to communicate a use for "exec" where it is preferrable
> > to not use explicit dictionaries.
> 
> I think "exec in local scope" is invariably overkill, and the more highly
> disciplined tools that Python provides in abundance afford better,
> faster and more maintainable ways to perform even tasks so dynamic
> that their very wisdom is dubious.

yes, let's hope so. 

> I think one good language, quite close to Python in most respects, that may
> support better your specific quest of runtime compilation of source in
> order to alter namespaces, is Ruby.  In Python, you're fighting against
> the grain of the language when you do this.  In Ruby, you are closer
> to the mainstream of the language -- you may also alter in such ways
> built-in types akin to Python's object, string, dict, etc, while Python does
> not let you do that.  I think Python's approach is wiser, and Ruby's leads
> to a programming style damaging the ability to understand and maintain
> large programs (but I haven't tackled 'big' projects in Ruby, so this is
> only a working hypothesis for me so far).  But I suspect you might disagree,
> and perhaps find yourself more at home in Ruby.  Do give it a look.

sorry. i am not going away from python :-/ Like i said earlier
it's not that altering namespaces is my prefered technique everywhere. 
I just wanted (and am obviously failing) to discuss a possible use of 
execing in current scope. 

If you are coming to EuroPython maybe we find half an hour where we 
can discuss it on the screen? Doing this in mail makes it hard to not 
misunderstand each other. 

It's especially frustrating if you continue to assume that i would think 
're.compile modifies the callers namespace' or 'exec only takes strings' 
or 'i always want to alter namespaces from everywhere' or
'source code cannot be seen at runtime' or ...
and finally try to get me to Ruby. I am not a speaking japanese :-)

regards,

    holger




From sholden at holdenweb.com  Mon Apr  8 18:21:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 18:21:35 -0400
Subject: html parsing
References: <3CB1E20A.B3300AA4@ncsa.uiuc.edu>
Message-ID: <wLos8.3293$vT6.2991@atlpnn01.usenetserver.com>

[posted & mailed]

"Randy Heiland" <heiland at ncsa.uiuc.edu> wrote in message
news:3CB1E20A.B3300AA4 at ncsa.uiuc.edu...
> I'm a newbie at parsing html in Python... could someone offer a snippet
> of code to:
> 1) specify a URL
> 2) parse the html and return the hyperlink associated with a unique
> anchor string
>
Take a look at the "webchecker" web crawler application included in the
Tools directory of your Python distribution -- tells you everything you need
to know.

regards
 Steve






From gh_pythonlist at gmx.de  Mon Apr 29 06:25:13 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 29 Apr 2002 12:25:13 +0200
Subject: CGI and client's IP?
In-Reply-To: <1a6qcuc92l8ac0tcegqj95fqemdctda21v@4ax.com>
References: <1a6qcuc92l8ac0tcegqj95fqemdctda21v@4ax.com>
Message-ID: <20020429102512.GB23223@lilith.my-fqdn.de>

* Fernando Rodr?guez <frr at easyjob.net> [2002-04-29 11:58 +0200]:
> Can I get the client's IP froma a cgi script? O:-)
 
This might do it: os.environ["REMOTE_ADDR"]

It's instructive to look at the output of this little CGI script:

import cgi
cgi.test()

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 12.7 ?C      Wind: 3.9 m/s




From peter at engcorp.com  Wed Apr 10 20:04:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:04:47 -0400
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <slrnab8v1o.emj.philh@comuno.freeserve.co.uk> <MPG.171e33893fed6df09896a7@news.nwlink.com>
Message-ID: <3CB4D31F.70E055AB@engcorp.com>

Jeff Shannon wrote:
> 
> In article <slrnab8v1o.emj.philh at comuno.freeserve.co.uk>,
> philh at comuno.freeserve.co.uk says...
> > What have the PEPs ever done for us?
> >
> > (Apart from the roads. The roads go without saying.)
> 
> ... and the Aqueduct...

And sanitation.  The roads, the aquaduct, and sanitation are three 
things the PEPs have done for us.


From bkc at Murkworks.com  Tue Apr 16 12:24:10 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 16 Apr 2002 12:24:10 -0400
Subject: I got my Py!
Message-ID: <3cbc4e4e$1_5@goliath.newsgroups.com>

Not bad for $3


--
Novell DeveloperNet Sysop #5

_




-----=  Posted via Newsfeeds.Com, Uncensored Usenet News  =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
 Check out our new Unlimited Server. No Download or Time Limits!
-----==  Over 80,000 Newsgroups - 19 Different Servers!  ==-----


From eppstein at ics.uci.edu  Wed Apr  3 18:45:21 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 03 Apr 2002 15:45:21 -0800
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com> <mailman.1017874091.20985.python-list@python.org>
Message-ID: <eppstein-541795.15452103042002@news.service.uci.edu>

In article <mailman.1017874091.20985.python-list at python.org>,
 David Ascher <DavidA at ActiveState.com> wrote:

> Would it not be "relatively" easy to add a compiler-time warning for
> most uses of this dangerous idiom?
> 
> At the very least, the boolean type could do the equivalent of:
> 
>   def __eq__(self, other):
>      if __debug__ and other not in (0,1,True,False):
>          warnings.warn("Testing of equality is best done by asking
> objects whether they're true, not comparing with a boolean",
> RunTimeWarning)
>      return int.__eq__(self, other)
> 
> or something similar?

I'd rather see a compiler time warning than the runtime warning you 
suggest, because it's likely that some completely innocuous "if x==y" 
code will trigger the runtime warning when one but not both of the 
variables ends up being a bool.  You only want to warn about comparisons 
to the constants True and False, not about comparisons to expressions 
that happen to take those values.

If you're going to consider compiler time warnings, another one should 
be for the syntax "if bool(x):" deprecated in the latest PEP 285 draft.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From aleax at aleax.it  Sat Apr 27 17:17:27 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 21:17:27 GMT
Subject: local variable not found
References: <pan.2002.04.27.23.01.32.51309.30401@onesite.org>
Message-ID: <HzEy8.28287$8D3.754996@news1.tin.it>

Alain Tesio wrote:
        ...
> NameError: global name 'y' is not defined
> 
> But it works if I use print locals()["y"]
> 
> Is this an expected behaviour ? Why ??

It's fully expected, because the real locals are realized
at compile-time unless the compiler sees that your
function contains an exec statement -- and yours
does not, since you use an eval function instead (which
SHOULD fail when given a statement rather than an
expression, but in reality currently succeeds by mistake).

Change your eval(...) to an exec ... and it should work.

It's close to impossible that such an approach is
anywhere like halfway-decent, by the way.  eval and
exec make sense a tiny percentage of the time, but
(in my experience) only when used with explicit
dictionaries, NOT by letting them default to locals()
and globals().


Alex



From u at g0ertz.de  Fri Apr 19 10:52:07 2002
From: u at g0ertz.de (Ulrich Goertz)
Date: 19 Apr 2002 16:52:07 +0200
Subject: GNU GPL & Python stand-alone packages from py2exe
Message-ID: <i0kit6nn3l4.fsf@Thales.MI.Uni-Koeln.DE>

Hi,

sorry to bring up licensing issues once again, but I need some
clarification on this. Say I wrote a Python application which
I want to publish under the GNU General Public License. Assume
I want to create a stand-alone distribution for Windows with
py2exe, put it on a CD ROM and distribute it. Now my questions:

1) Since the Python license is "GPL compatible" this should be 
no problem, but what does that mean in practice? Can/should I put
the whole package under the GPL and omit the Python license?
Or can I just put my own code under the GPL, and use the 
Python license for the Python part, although I bundle both
into one .exe-file and a few dll's?

2) Would I have to put the source code of the Python interpreter
on the CD, too? As far as I understand the wording of the GPL,
since I include the Python interpreter, that would indeed be
necessary. On the other hand, my understanding of the spirit
of the GPL (possibly wrong...) tells me that it might be
superfluous. In addition, it would be pretty cumbersome.

3) Would it be more suitable to release the source code of my
application under the GPL, and release the stand-alone package
under another ("closed") license which would not request the 
source code to be included, and would not allow to modify the 
program. That would enable me to distribute the stand-alone 
without the source code of the Python interpreter, and ensure 
that people that want to modify my program would have to take 
the source code distribution and thus are bound by the GPL.

As a side note, I do understand that the GPL is not as free
as other licenses (the Python license, say), but I do not 
want to give others the possibility to do what they want with
my code without any restrictions. So I think the GPL is
suitable for me. My questions are more about how to proceed
in practice in this case.

Regards, Ulrich


From roy at panix.com  Wed Apr 10 20:26:08 2002
From: roy at panix.com (Roy Smith)
Date: Wed, 10 Apr 2002 20:26:08 -0400
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <roy-E50C96.20260810042002@news1.panix.com>

Robin Becker <robin at jessikat.fsnet.co.uk> wrote:
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers. 

Software is my job (both writing it and useing it).  If I can't have a 
little fun doing my job, it's time for a new career.


From tim.one at comcast.net  Wed Apr 10 20:36:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 20:36:32 -0400
Subject: Booleans, integer division,
 backwards compatibility; where is  Python  going?
In-Reply-To: <23891c90.0204100552.6b822307@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEBAONAA.tim.one@comcast.net>

[Paul Boddie]
> ...
> Well, I was excited a few years ago when iterators appeared in C++, so
> I suppose I should be even more excited now. :-) However, it's
> arguably a bigger deal for the implementer of the iterator (or
> generator, even) than the "users" or "clients" of them.

> ...

> That's part of my original argument, though. I get the impression that
> some packages not only require Python 2.2 because they use new,
> standard features, but also so that the author can drag users further
> into the future by making use of not-yet-standard features, if only
> for their own entertainment.

You users slay me <wink>.  For me, "my own entertainment" means I can use
such "developer features" to get my work implementing features for you to
use done quicker, have it work better, suffer fewer bugs, and give you more
capabilities in your code than you had before.  If you're ever tempted to
think that work devoted to improving language and library implementors'
productivity is wasted on users, you should try not watering your garden
during a drought too <wink>.

even-elves-need-to-eat-ly y'rs  - tim





From martin at v.loewis.de  Sun Apr 21 06:35:54 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 21 Apr 2002 12:35:54 +0200
Subject: IDLE Help problem
References: <a9t4ui0cnm@enews4.newsguy.com>
Message-ID: <m3k7r1tk39.fsf@mira.informatik.hu-berlin.de>

Jim <jimb417 at totalNOSPAMspeed.net> writes:

> I have Opera, Konquerer, and Netscape available. What do I need to do to 
> get rid of this error. I'm using Mandrake 8.0. Thanks in advance.

Do you have an environment variable BROWSER set? (you don't need to,
but having it set incorrectly could cause the problems you see)

If not, do you have "netscape" or "kfm" *in your path*?

Regards,
Martin


From mwh at python.net  Thu Apr 11 09:55:28 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 11 Apr 2002 13:55:28 GMT
Subject: Binding a reference to a variable (solved?!?)
References: <E16v1a2-0005a6-00@mail.python.org> <mailman.1018463002.9069.python-list@python.org>
Message-ID: <lksn62wa9o.fsf@pc150.maths.bris.ac.uk>

Patrick Miller <patmiller at llnl.gov> writes:

> Here's a scary way to do what Andrew was looking for.  It has
> the proper form and fewer (but still significant) restrictions.
> 
> >>> x = 7
> >>> set(x,99)
> >>> print x
> 
> I'll likely submit the following as a ASPN recipe, though
> it clearly needs a better description of motivation.  Just
> because you CAN do a thing doesn't mean you SHOULD
> do a thing.

Do you know about my bytecodehacks?  I think you might like
them... (though they only really work with 1.5.2).

[...]
> x = 3
> y = [1,2,3,4,]
> z = {"something": 0}
> 
> 
> def set(lhs,rhs,G=globals()):
>     address = id(lhs)

This is a pretty bad way of identifying the passed in value.  What you
"want" to do is look at
sys._getframe().f_back.{f_lasti,f_code.co_code}, poke around until you
find the

    LOAD_FAST 'x'

instruction or whatever, and go from there.

Of course, this is all absurdly hacky, but kind of fun.  It should
come with a kind of "explosive licence" -- if you use it in production
code, your shoes explode.

Cheers,
M.

-- 
  Programming languages should be designed not by piling feature on
  top of feature, but by removing the weaknesses and restrictions
  that make the additional features appear necessary.
               -- Revised(5) Report on the Algorithmic Language Scheme


From ball at righty.eiu-newman.org  Tue Apr  9 22:52:13 2002
From: ball at righty.eiu-newman.org (ball at righty.eiu-newman.org)
Date: 10 Apr 2002 02:52:13 GMT
Subject: Is Python Sticky?
Message-ID: <3cb3a8dd$1@nntp01.splitrock.net>

Hello!

For a while "Learn Python" has been on my list of things to do.  Right
now I have an application that could use some glue between a web based
interface (in its simplest form a single page with a couple of buttons
on it), and a shell script (I'm running NetBSD).  Would Python be
suitable for this kind of thing?  Are there other pieces of the puzzle
that I'll need?  Any pointers would be most welcome.

Regards,
  - Andy Ball.


From SBrunning at trisystems.co.uk  Thu Apr 11 04:31:28 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 11 Apr 2002 09:31:28 +0100
Subject: What has PEP 285 done to us?
Message-ID: <31575A892FF6D1118F5800600846864DCBCF7B@intrepid>

> From:	Peter Hansen [SMTP:peter at engcorp.com]
> Jeff Shannon wrote:
> > 
> > In article <slrnab8v1o.emj.philh at comuno.freeserve.co.uk>,
> > philh at comuno.freeserve.co.uk says...
> > > What have the PEPs ever done for us?
> > >
> > > (Apart from the roads. The roads go without saying.)
> > 
> > ... and the Aqueduct...
> 
> And sanitation.  The roads, the aquaduct, and sanitation are three 
> things the PEPs have done for us.
 
And it's safe to walk in the streets at night now.

Yeah, they certainly know how to keep order. Let's face it; they're the only
ones who could in a place like this.

The-irony-is-thick-enough-to-cut-ly y'rs,
Simon B.




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




From magnus.heino at pleon.sigma.se  Fri Apr 26 07:35:43 2002
From: magnus.heino at pleon.sigma.se (Magnus Heino)
Date: Fri, 26 Apr 2002 13:35:43 +0200
Subject: mx.BeeBase.BeeDict
Message-ID: <tYay8.8691$WJ4.11752@news1.bredband.com>

Could anyone please show me an example usage of mx.BeeBase?

Thanks!


[magnus at daysleeper magnus]$ python2.2
Python 2.2.1 (#2, Apr 13 2002, 14:35:50)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from mx.BeeBase.BeeDict import BeeStringDict
>>> s=BeeStringDict('foo')
>>> s['foo']='bar'
>>> s['foo']
'bar'
>>> s.commit()
2000 2002-04-26 13:33:54.74 [p4474] Committing all changes for "foo"...
2000 2002-04-26 13:33:54.74 [p4474]  writing key "foo":
-------------------------------------------------------------------------------
'bar'
-------------------------------------------------------------------------------
not found; cc=-1
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/opt/lib/python2.2/site-packages/mx/BeeBase/BeeDict.py", line 1067, 
in commit
    index[key] = address
TypeError: record address must be an integer
>>>



From andrewm at object-craft.com.au  Wed Apr  3 06:05:18 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 03 Apr 2002 21:05:18 +1000
Subject: Java to Python translation 
In-Reply-To: Your message of "03 Apr 2002 01:46:25 PST."
             <7xbsd15d3i.fsf@ruckus.brouhaha.com> 
Message-ID: <20020403110518.C6DB238F50@coffee.object-craft.com.au>

>> > I need to translate a (rather voluminous 100MB) java program to
>> > python. After a few searches on google, apparently the best tools
>> > for this task are Lex-Yacc (Flex-Bison on linux). While I am working
>> > on the Lex-YACC HOWTO, I would be very interested in hearing of
>> > others' successful experiences/feedback at this. Any pointers of
>> > interest would similarily be appreciated.
[...]
>I think for a project like that, even a success would be disastrous.
[...]

Indeed.

But I bet a complete re-write in python would less than 10MB, and take
less than a tenth of the time of the original project... 8-)

The suggestion of using Jython would be good one - subsystems can be
rewritten and tested as time and resources allow.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/



From claird at starbase.neosoft.com  Thu Apr  4 11:31:49 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2002 10:31:49 -0600
Subject: Python compilers
References: <3CAA9786.2070700@moofed.com> <3CAAA0BF.E8E4659F@engcorp.com>
Message-ID: <F3A8543E0B42CE8C.87ADBC36DD108823.18818E360FF95C6A@lp.airnews.net>

In article <3CAAA0BF.E8E4659F at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
			.
		[articulate lucidity]
			.
			.
>> If there is a python compiler, surely it would be able to generate code
>> that is as efficient as a C compiler, resulting in what would be the
>> best available OO language? (having simplicity, speed)
>
>No, but relatively few people are overly concerned about that.
>Most programmers who are greatly concerned about speed actually
>have no particular requirement in mind, they just want "more".
>Far too often this is a waste of their time.  Some people look
>at ways to optimize programs that take 60 seconds to run and
>get run once a day.  In C the same program might take 2 seconds,
>but so what?  It might also crash, and it took ten times longer
>to develop...
			.
			.
			.
Peter's answer is consummate.  On the off-chance that
Mr. Hazlewood truly wants to pursue his search for "the
best available OO language" as one which emphasizes
performance (in a naive one of its several senses) and
simplicity, I recommend that he consider Smalltalk (and
Squeak in particular).

Depending on how you conceive "simplicity", such languages
as Eiffel, Dylan, Self, Perl (!), and Sather also empha-
size performance in interesting ways.

OK, so does Ada, but not even *I* have the energy to 
argue for its simplicity.
-- 

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


From mal at lemburg.com  Thu Apr 25 06:53:40 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 25 Apr 2002 12:53:40 +0200
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019279977.28194.python-list@python.org> <a9rbi7$fr2$1@slb4.atl.mindspring.net>
Message-ID: <3CC7E034.EF1A564@lemburg.com>

Andrew Dalke wrote:
> 
> Now, mxTextTools.  Three attempts.  First two failures took a week,
> thereabouts.  The final success took another couple weeks to get a
> working regexp parser (yes, on top of mxTextTools) going.  Let's
> call it three weeks.  And this was after knowing a lot about how
> those As work.
> 
> But I gotta say, mxTextTools is amazingly fast and flexible.  On
> one test it was only 50% slower than fgrep for a simple text search.
> (Not quite a direct comparison because the Python code was doing
> other things as well.)

Side-note: fgrep uses assembler on some platforms whereas mxTextTools
relies on the quality of the C optimizer to reach top speeds.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From simon at uggs.demon.co.uk  Thu Apr  4 18:29:21 2002
From: simon at uggs.demon.co.uk (Simon Foster)
Date: Thu, 04 Apr 2002 23:29:21 GMT
Subject: how to detect if a file is complete
References: <32f17666.0204040701.64dd0bfe@posting.google.com>
Message-ID: <3cade180.760813@news.dsl.pipex.com>

On 4 Apr 2002 07:01:43 -0800, harm at laatje.dhs.org (harm) wrote:

>Hi all,
>
>I want to copy some files from one dir to another using python. 
>The files are copied onto my linux box via a samba-share.
>
>I want to write a python script which checks a dir and copies all
>files to another dir. But there is the possibility that the file is
>not yet complete. (This are rather large Postscript files >100mb).
>Therefore I need to detect if the file is complete.
>
>The files are copied on the samba-share by Adobe Acrobat Distiller,
>therefore i cannot save some check file like $FILENAME.complete...
>
>Is there some other way (using python) to detect if the file is
>complete?
>

Why not check the size of the file and consider it complete when
it stops growing?  
--
Simon Foster
Somewhere in the West of England


From mcfletch at rogers.com  Wed Apr 24 18:28:11 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 24 Apr 2002 18:28:11 -0400
Subject: Faster md5/1-way encryption?
References: <20020424215752.GC15371@teich.Garten.DigitalProjects.com>
Message-ID: <3CC7317B.3040208@rogers.com>

Something that hits me right off the bat:

 >             md5_zeile = md5.new(zeile).hexdigest()
 >             self_Inhalt += md5_zeile + linesep

Would likely be much better as:

	myResults = []
	...
	myResults.append( md5.new(zeile).hexdigest() )
	...
	myResults = string.join( myResults, os.linesep )

Which would save you creating and destroying huge numbers of (at the end 
of processing) very large strings.  If you can avoid that last step 
(join) and just use writelines to save to file, even better.

Oh, and:
	lines = len( myResults)

can replace your 'i' counter in that approach (won't be a noticable 
savings, I just hate counters :) ).

Finally, file.tell() may be fairly slow (depends on implementation), as 
it may require flushing all buffers so that the file can give an 
accurate result as to current position.  Unless you really need that 
accuracy, consider just accumulating the length of the lines you're 
processing and adding 1 or 2 each time for the length of os.linesep.

MD5 can process gigabytes of information pretty quickly, so I'd be 
surprised if a 100,000 line file is a huge problem.

HTH,
Mike

Alexander Skwar wrote:
> Hi!
...
> The files which should be processed are rather large (100,000+ lines)
> and on my machine this takes ages to run.  I'm currently doing it like
> this:
> 
> 
> import os
> import xreadlines
> import threading
> import md5
> 
> class md5Thread(threading.Thread):
...
>     def run(self):
>         dateiname = self.Dateiname
>         self.Zeilenanzahl = os.path.getsize(dateiname)
>         linesep = os.linesep
>         datei = file(dateiname, 'r', 1)
>         
>         self.Zeilennummer = 0
>         self_Inhalt = ''
>         i = 0
>         for zeile in xreadlines.xreadlines(datei):
>             md5_zeile = md5.new(zeile).hexdigest()
>             self_Inhalt += md5_zeile + linesep
>             self.Zeilennummer = datei.tell()
>             self.ZeilennummerZeile = i
>             i += 1
>             
>         self.Inhalt = self_Inhalt
>         self.done = 1
...
> Further, I use a wxTimer to poll the current byteposition of the thread
> and display it in a wxGauge and wxTextctrl.
> 
> Since this is my first Python app, there are for sure some things that
> can be optimized.
> 
> I suppose what's taking so very long is 
> a) The md5 generation itself
> b) That I instantiate a new md5 object for every single line and thus
> will have 100,000+ objects.
> 
> Could somebody please tell me, how I could speed the whole thing up?  I
> don't have to calculate the md5 sum of every line, but I need to use
> some sort of 1-way encryption.  So, if there are faster alternatives
> than md5, I'd be happy as well.
> 
> Thanks,
> 
> Alexander Skwar


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From m_mariappanX at trillium.com  Thu Apr 25 06:20:49 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Thu, 25 Apr 2002 03:20:49 -0700
Subject: a ? on using Pmw dialog widget?
Message-ID: <53A7943A5BD8D411B6930002A5073155022D749F@bgsmsx90.iind.intel.com>

Hi python list,

I've already psotign this question. This time I've attache the test
secnario.

I try to deactivate the global modal dialog box from another thread. But
cudn't

Any idea how can do that?

Thx
Maharajan
 <<test.ZIP>> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.ZIP
Type: application/octet-stream
Size: 1064 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020425/01284385/attachment.obj>

From johnroth at ameritech.net  Wed Apr  3 06:40:00 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:40:00 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <slrnaakn16.17k.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ualq61a0br5839@news.supernews.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnaakn16.17k.huaiyu at gauss.almadan.ibm.com...
> I don't think Guido follows this thread. I hope someone at least refer
to
> him Laura's and Alex' excellent articles.
>
He saw it. His response was that Laura's article was FUD.
My take on it is that the root of her problem is a poorly
chosen name - there would be exactly **no** pedagogical
issue if the function was named "numberOfLeapDays()"
instead of "isLeapYear()"

John Roth






From greg at cosc.canterbury.ac.nz  Tue Apr 23 22:51:54 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 24 Apr 2002 14:51:54 +1200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org>
Message-ID: <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz>

Andrew McNamara wrote:
> 
> >Interersting.  If so, then RMS is 'guilty' of an almost humorous
> >contradiction.
> 
> Speculation - maybe he's aware of this and simply sees it as playing
> what he sees as a fatally flawed system against itself? 8-)

I wonder what he would do if there were no intellectual
property laws. Would he be campaigning for the introduction
of a law that enabled him to enforce the price that the
GNU copyleft exacts?

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


From pyth at devel.trillke.net  Sun Apr 21 16:38:07 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 22:38:07 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <gs6v9a.vb1.ln@127.0.0.1>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <0evu9a.621.ln@127.0.0.1> <mailman.1019413960.21990.python-list@python.org> <gs6v9a.vb1.ln@127.0.0.1>
Message-ID: <20020421203807.GJ10574@devel.trillke>

On Sun, Apr 21, 2002 Carl Banks wrote:
> >> # Import stuff for version 2.1
> >> if sys.version[:2] >= (2,1):
> >>     from mymodule2_1 import *
> >> 
> >> # Import for version 2.2
> >> if sys.version[:2] >= (2,2):
> >>     from mymodule2_2 import *
> > 
> > this solves part of the 'code duplication' problem. Might get a bit tricky
> > if you just want to define a few additional methods for some classes
> > (for example 'generator versions' of methods otherwise returning lists).
> 
> I wouldn't say so.  It is straightforward enough to add a method to a
> class any time:
> 
> >>> class a:
> ...    pass
> ...
> >>> def p(self,x):
> ...    print "hello, %s" % x
> ...
> >>> a.p = p
> >>> b = a()
> >>> b.p("world")
> hello, world

True. But it splits source-code where you don't really want it to split.
Usually i prefer to keep all object-functions at one place.

> > supposed to work? On my python 2.2 
> > 
> >        sys.version[:2] == '2.'   is True
> >
> > am i missing something here?
> 
> I just did it how the OP did it, without bothering to check.  It
> should be sys.version_info[:2] >= (x,y)

thanks for clarification.

	holger




From eppstein at ics.uci.edu  Sun Apr 14 23:00:54 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 20:00:54 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122>
Message-ID: <eppstein-BBF667.20005414042002@news.service.uci.edu>

In article <a9dbc9$bge$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
wrote:

> All 32-bit ints are represented in hex without a sign, so

This signed-unsigned confusion bit me when I needed a routine to count 
the nonzero bits of some input.  Normally one would write

    def popCount(x):
        count = 0
        while x:
            count += 1
            x &= x - 1

but of course this gets into an infinite loop when the high bit is set.  
Arguably, this is the correct behavior: a negative number in 2's 
complement should be thought of as having an infinite sequence of 1-bits.

But in my case, my input didn't ever have the high bit set, and a good 
thing too, or I would have gotten ValueError when I tried 
int(hexstring,16).
The reason I was getting into trouble anyway was because of the way 
right-shifting can change sign:

>>> 0x40000000<<1
-2147483648

This is inconsistent with int-long unification.  In fact, I think it is 
a bug.  The answer should, of course, be 2147483648L, just as you would 
get by multiplying by 2 instead of shifting.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From marklists at mceahern.com  Wed Apr  3 08:49:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 07:49:17 -0600
Subject: PEP 285: Adding a bool type
In-Reply-To: <ualq61a0br5839@news.supernews.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEPDKIAA.marklists@mceahern.com>

[John Roth]
> He saw it. His response was that Laura's article was FUD.
> My take on it is that the root of her problem is a poorly
> chosen name - there would be exactly **no** pedagogical
> issue if the function was named "numberOfLeapDays()"
> instead of "isLeapYear()"

I thought she was trying to make the stronger point that even without PEP
285, any function that is designed to return true/false (1/0) is badly
designed.  Moreover, PEP 285 would only encourage people to write such
functions.

Does that apply to, say, isinstance()?

// mark




From ahorvath at szgti.bmf.hu  Tue Apr 16 11:30:05 2002
From: ahorvath at szgti.bmf.hu (Horvath, Arpad)
Date: 16 Apr 2002 08:30:05 -0700
Subject: Brackets in brackets, braces in braces
Message-ID: <2e1a250.0204160730.637139f1@posting.google.com>

How can I find strings like this with regular expressions:\something
[something1] {something2} {something3}Where in the braces can be also
another braces. For example:\frac{1\times10^{-16}}{4}   can be at the
place, where something2 is.(braces in pairs)\interval [J]
{\frac{1\times10^{-16}}{4}} {1..5}My half solution is at
www.szgti.bmf.hu/~ahorvath/regexp.txt.


From gcordova at hebmex.com  Fri Apr  5 11:44:00 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 5 Apr 2002 10:44:00 -0600 
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? 
	[LONG!])
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB74@mail.hebmex.com>

> 
> Andy Gimblett <gimbo at ftech.net> writes:
> > > Use a module.  Don't use a class unless you need a class.
> > 
> > /me closes his eyes tight, opens his mouth, thinks about it for a
> > while, shaking his head around slightly, feels something click, says
> > "wow", opens his eyes, says "wow" again, and hits reply.
> > 
> > Wow.  Man, I love python.
> > 
> > Cheers for the epiphany
> 
> Care to explain it?  I missed something.
> 

Andy hadn't caught on that a module is loaded only once,
the first time it's imported. Any subsequent imports are
only accesses to the original imported version, so that
makes them effectively singletons.

So if you keep lotsa utility functions, variables and
constants in a certain module, and import it from a lot
of diferent places in a single application, then you're
really referencing the originally imported version, all
others are only references.

Neat, eh?

:-)

-gustavo




From martin at v.loewis.de  Fri Apr  5 13:51:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Apr 2002 20:51:13 +0200
Subject: Filename Too Long
References: <be230158.0204050948.7fa6a62f@posting.google.com>
Message-ID: <m3d6xe9dy6.fsf@mira.informatik.hu-berlin.de>

rudi_sass at yahoo.com (Peter D) writes:

> I'm using windows and trying to call os.path.getmtime() after using
> os.path.walk... However, I'm choking with "[Errno 38] Filename too
> long" on paths with len > ~260
> 
> Is there any way around this error?

Since you are asking for a work-around: cd into one of the more nested
directories when the path gets longer (os.chdir), and use relative
paths from then on.

If you want to study how to solve the problem: the relevant code
snippet is in Modules/posixmodule.c

	/* the library call can blow up if the file name is too long! */
	if (pathlen > MAX_PATH) {
		PyMem_Free(pathfree);
		errno = ENAMETOOLONG;
		return posix_error();
	}

There might be different ways to approach this:

- challenge the correctness of the comment:
   - try to establish why the author of that code thought that the C
     library could blow up
   - analyse whether these arguments are still true with the current
     compiler version
  -or-
   - prove the argument wrong by analysing the source code of the C
     library
- then remove the constraint
-or-
- use different API to achieve the same effect without suffering from
  the constraint.

I'm serious about these suggestions: users would appreciate if this
gets fixed somehow - apparently, the system allows longer file names,
and apparently, the system itself can deal with that quite well. This
can be only true if the system either doesn't use its C library, or if
the C library does not have this restriction.

HTH,
Martin


From jim.jacobs at jacobshome.com  Tue Apr  2 22:30:59 2002
From: jim.jacobs at jacobshome.com (Jim Jacobs)
Date: Tue, 2 Apr 2002 22:30:59 -0500
Subject: Static Variables
Message-ID: <a8dt35$a2i$1@slb2.atl.mindspring.net>

Is it possible to simulate "C" style static variables in functions?  In
methods?

Thanks,

JDJ




From mlh at vier.idi.ntnu.no  Fri Apr  5 22:51:33 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 03:51:33 +0000 (UTC)
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com> <mailman.1017875711.27201.python-list@python.org>  <yu99lmc4drov.fsf@europa.research.att.com>
Message-ID: <slrnaass65.178.mlh@vier.idi.ntnu.no>

In article <yu99lmc4drov.fsf at europa.research.att.com>, Andrew Koenig wrote:
[snip]
> I'd like to suggest that comparisons with True are not only
> unnecessary but dangerous.

Indeed.

> Comparisons with False are much less dangerous but still
> unnecessary.

Why do you consider them less dangerous? Using "foo==False" instead of
"not foo" will be just as wrong as using "foo==True" for "foo" as a
Boolean value...

>-- Andrew Koenig, ark at research.att.com,
>http://www.research.att.com/info/ark

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From egrossde at yahoo.de  Sun Apr 28 04:04:31 2002
From: egrossde at yahoo.de (jb)
Date: Sun, 28 Apr 2002 10:04:31 +0200
Subject: Indexing cells in a table
Message-ID: <3ccbac7c_3@news.newsgroups.com>

I have a base class x with attributes a,b,c,..., that is x.a=0 is o.k.

Now T is a rectangular table of elements of the type x, that is I can say
T[4][3].c=1.

Is there a generic way to address an area of the table and assign values to 
one of the attributes in the area?
That is, can I write a generic function setValue that works like 

T.setValue( (0,0) ,(m,n) , c , 1 ) 

and assigns the value 1 to the x.c-s in the specified area.

Example:

  setValue( (1,1), (2,3), c, 1)

is the same as

  T[1][1].c=1
  T[1][2].c=1
  ...
  T[2][3].c=1

TIA,

-- 
Janos Blazi


-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From phd at phd.pp.ru  Wed Apr 10 05:57:48 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 10 Apr 2002 13:57:48 +0400
Subject: Sending a signal to a process started by another user
In-Reply-To: <11aed8f9.0204100138.20c54afe@posting.google.com>; from joonas@olen.to on Wed, Apr 10, 2002 at 02:38:01AM -0700
References: <3CB31C2C.9010503@olen.to> <mailman.1018377258.31626.python-list@python.org> <11aed8f9.0204100138.20c54afe@posting.google.com>
Message-ID: <20020410135748.O2883@phd.pp.ru>

On Wed, Apr 10, 2002 at 02:38:01AM -0700, Joonas Paalasmaa wrote:
> Oleg Broytmann <phd at phd.pp.ru> wrote in message news:<mailman.1018377258.31626.python-list at python.org>...
> > On Tue, Apr 09, 2002 at 04:53:58PM +0000, Joonas Paalasmaa wrote:
> > > How can I send a signal to a process started by another user. In this 
> > > case my intention is to send a signal 14 to a process started by user jp 
> > > from a cgi script. The script raises an "Operation not permitted"- 
> > > OSError perhaps because the script is run by user "apache" and the 
> > > process by user "jp". How can In make this work properly?
> > 
> >    Using some form of setuid programs.
> 
> Can you give me an example how to use setuid scripts with CGI and Python.

   Just compile a program, that does what you need, and ask your sysadmin
to install it setuid.
   Please note, that setuid programs are security risks, and setuid scripts
are disabled on all modern OSes. If you want setuid script - you need a
setuid wrapper for it.

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




From alain at onesite.org  Sat Apr 27 17:01:32 2002
From: alain at onesite.org (Alain Tesio)
Date: Sat, 27 Apr 2002 23:01:32 +0200
Subject: local variable not found
Message-ID: <pan.2002.04.27.23.01.32.51309.30401@onesite.org>

Hi,

I have a problem trying to use code compiled on the fly,
python looks in the globals variables only.

Look at this example :

===
import codeop

def f():
	x=[1,2,3]
	eval(codeop.compile_command("y=x"))
	print y

f()
==

It cannot find the variable y (I don't know its name at runtime):

Traceback (most recent call last):
  File "/tmp/testlocal.py", line 8, in ?
    f()
  File "/tmp/testlocal.py", line 6, in f
    print y
NameError: global name 'y' is not defined


But it works if I use print locals()["y"]

Is this an expected behaviour ? Why ??

Thanks
Alain


From aleax at aleax.it  Wed Apr 10 05:53:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 10 Apr 2002 09:53:53 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204090526.18675d02@posting.google.com> <0NCs8.63131$pT1.1810689@news1.tin.it> <23891c90.0204100109.4a44932b@posting.google.com>
Message-ID: <RYTs8.2219$b62.57325@news1.tin.it>

Paul Boddie wrote:
        ...
>> Nothing is more natural to a beginner than "if x in somedict:" and
>> "for x in thefile:" -- having had the opportunity to teach some
>> beginners with 2.2, I've seen them slurp these up instantly (just
>> try to contrast with "if somedict.has_key(x):" and
>> "for x in thefile.xreadlines():" ...!).
> 
> Well, both examples you gave seemed to be controversial when up for
> discussion in the past, but if you say they help people, then I'll
> take your word for it.

I've noticed they immensely help Python beginners that come to them
without preconcepts.  Such beginners are surprised to learn that
these are extremely recent additions to Python, so well they fit in.

Human beings can grow used to anything, and sometimes get set in their
ways, so I can't judge how well people already expert of Python are
taking them.  Still, "for x in thedict:" is faster than "for x in 
thedict.keys():", so it's not only syntax sugar.


>> And the new features are the right way to implement such obvious progress
>> -- iterators, for example, let you loop on (e.g.) dictionaries or files,
>> without requiring such types to define unnatural (or unfeasible)
>> "indexing" then go out of their way to stop misuse.
> 
> Iterators are indeed nice to use and a very welcome feature.

They're great progress over the previously murky "iteration protocol".


>> > advanced/experienced Python developer.
>> 
>> You MUST be joking -- inheriting from builtin types is SUCH a win,
>> and so obviously so, for advanced and experienced developers!
> 
> Perhaps I'm not writing the kinds of applications where this is the
> case, then.

Or maybe not being used to them you just don't think of cases where
they'd be handy (if you need to keep compatibility with 2.1, then
of course you'd better NOT start thinking of such cases:-).  I.e.,
Sapir-Whorf hypothesis somewhere between its weakest and strongest
forms (though I still prefer the way Wittgenstein independently
put it earlier -- "The limits of my language are the limits of
my world").  Mappings and sequences are such useful concepts that
I doubt you don't have such objects in your applications -- and
it's such a frequent occurrence nowadays for the best implementation
strategy for mappings or sequences to involve subclassing builtins,
that, while possible, it would be peculiar if none of your cases
fell into this category (always IF you don't need to care about
Python 2.1 and earlier, of course!).


>> Generators need a "from __future__ import", so I applaud the tutorial's
>> choice not to burden beginners with them YET.
> 
> It would be nice for some coverage of that particular feature, though.

Yes, just not in the tutorial (at least not until they're in the
"current" language rather than needing import from the future).

> I can't say that I've ever written one myself, although they do seem
> useful for certain kinds of applications.

Just basically all of those where you have loops; if your applications
have no loops (including ones obtained by recursion), then generators
are not of much interest:-).

Seriously: generators are a way to factor most looping control
structures' logic into specific reusable objects.  You can always
do that with iterators built in other ways (just classes with
a 'next' method), but generators are simpler and more concise
in a reasonably wide number of cases, basically because they
directly map the reasonably frequent case where it's natural to
keep part of a control structure's state in terms of "_where_
in this source code am I executing now".

Iterators (and generators in particular) are also a nice canonic
way to model many kinds of "lazy evaluation" -- I find that
part especially exciting -- but that's a rather different way
of looking at things.


>> > other things to do during working hours) it becomes increasingly hard
>> > to keep track of what's great in the more recent Python releases.
>> 
>> I think AMK's "what's new" articles are great for that purpose.
> 
> I must find the time to review them at my leisure, then.

I think you'll be glad you do -- they're concise and to the point.

AMK's mission is not to tell you WHY you can be interested in
major stuff such as subclassing builtins, iterators, generators,
though -- just to ensure you DO know easily what IS new.


Alex



From max at alcyone.com  Sun Apr 28 18:09:14 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 28 Apr 2002 15:09:14 -0700
Subject: how to pickle a lambda?
References: <3ccc58db$0$14003$afc38c87@news.optusnet.com.au>
Message-ID: <3CCC730A.FC60D13F@alcyone.com>

Alejandro Dubrovsky wrote:

> i've got a class that defines something like the following:
> 
> def t(self,a,b):
>         ...
> 
> def h(self):
>         self.thelambdafunctiontobepickled = lambda x: self.t(x,4)
> 
> is there any simple way of pickling that lambda function above so that
> when i unpickle it will be set to the right function?

You can't pickle a lambda function.  However, you could for instance
pickle a string representation of it that when eval'd gives the desired
lambda.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From mcherm at destiny.com  Thu Apr  4 12:42:36 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Thu, 04 Apr 2002 12:42:36 -0500
Subject: The Funniest thing about PEP 285
Message-ID: <3CAC908C.8@destiny.com>

The funniest thing about PEP 285 is this:

    I thought it was an elaborate April Fools joke.

After all, I reasoned, having a bool type might be a good idea, but it's 
the sort of thing that is raised over and over again, and has (up until 
now) been rejected just as often.

There were even a few "give-aways" -- things which were just a little 
bit too incredible to believe. For instance, Guido had made it quite 
clear that 2.3 was to be a "no new syntax" release, so announcing it 
would appear in 2.3 couldn't be true. And Tim's claim that 
operator.truth should return an int instead of a bool (when it's NAMED 
"truth"!) seemd a bit beyond the credible.

Considering the date, I decided it was an elaborate (and REALLY WELL 
DONE) April Fool's Day joke.

Of course, if such a joke appeared in Slashdot (anyone remember 
"Parrot"?) or most other forums, there would be a horde of people 
posting to say "This is just an April Fool's joke.", but I figured that 
the Python community had (appropriately) a better developed sense of 
humor, and knew better than to spoil the joke by revealing it.

In the end, I think the joke was on me.

<grin>

-- Michael Chermside


PS: The most AMAZING thing about PEP 285 is this:

Despite my thinking that it was a joke, I completely support the 
proposal. In fact, I think that the version Guido declared as accepted 
is the best possible solution with regard to every issue raised during 
the discussion. And, ammusingly, I think of that more as a validation of 
MY language design skills than of Guido's.





From andy47 at halfcooked.com  Tue Apr  9 00:19:25 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Tue, 9 Apr 2002 04:19:25 +0000 (UTC)
Subject: Need help with distutils
References: <mailman.1018083739.10368.python-list@python.org> <Xns91EA988AA8260andy47halfcookedcom@203.109.252.31> <slrnab3gei.1i0.akuchlin@crystal.mems-exchange.org> <Xns91EB605BFDDADandy47halfcookedcom@203.109.252.31>
Message-ID: <Xns91EB91841173Eandy47halfcookedcom@203.109.252.31>

Andy Todd <andy47 at halfcooked.com> wrote in 
news:Xns91EB605BFDDADandy47halfcookedcom at 203.109.252.31:

> "A.M. Kuchling" <akuchlin at crystal.mems-exchange.org> wrote in 
> news:slrnab3gei.1i0.akuchlin at crystal.mems-exchange.org:
> 
>> In article <Xns91EA988AA8260andy47halfcookedcom at 203.109.252.31>, 
>>    Andy Todd wrote:
>>> notice that the pattern you pass to global-exclude has to be a Unix-
>>> style "glob" pattern. In that notation '.' is a special character so
>>> using  '.cvsignore' doesn't actually match your files ;-) 
>> 
>> No, '.' is not special in glob notation.
> 
> Which means that this is a bug? I can verify that ".cvsignore" doesn't 
work 
> and "*cvsignore" does.
>   

Apologies, I spoke too soon. I had "*cvsignore" excluded for another reason 
and thought it was because of glob. So I ended up combining two 
misunderstandings and confused them for a bug. I have changed my 
MANIFEST.in to read

global-exclude .cvsignore

I can verify that this does work. The problem the op had (I believe) was 
that he tried

recursive-exclude . .cvsignore

Which by my reading of the documentation *should* do the same thing, but I 
can confirm doesn't (at least on Windows2000 with distutils running under 
Cygwin). I observed the same problem - namely a message stating;

warning: no previously-included files matching '.cvsignore' found under 
directory '.'

Ah, the joys of distutils. I don't know if this is a feature or a bug, but 
at least we can workaround it.

>> 
[snip]


Regards,
Andy
-- 
Contents free posts a speciality


From aleax at aleax.it  Fri Apr  5 04:46:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 09:46:37 GMT
Subject: Singleton?  No -- Borg! (was Re: Catalog of Python Patterns? [LONG!])
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com> <20020405014732.GB17563@isis.gerrietts.net> <mailman.1017996733.25069.python-list@python.org>
Message-ID: <1oer8.51722$pT1.1468384@news1.tin.it>

Andy Gimblett wrote:
        ...
> Really?  I've used Singleton a fair bit (probably too much in fact)
> for stuff like configuration information.  I'm curious as to how else
> I could/should do this.

Have a look at the paper I presented at IPC10, e.g. at:
http://www.aleax.it/Python/5ep.html


Alex



From nikolai.kirsebom at siemens.no  Tue Apr  2 06:10:25 2002
From: nikolai.kirsebom at siemens.no (Nikolai Kirsebom)
Date: Tue, 02 Apr 2002 11:10:25 GMT
Subject: ASP / IIS problems
Message-ID: <3ca99102.602353265@news.mch.sni.de>

I have 2 NT 4.0 (SP6) servers running IIS 4.0 where I make use of
Python in ASP pages.  One of the machines works ok, "Python asp pages"
are displayed correctly (in web-browser).  However on the other
machine I have problems in that I have to "change" the content of the
asp-file in order for the content to be presented correctly.  This is
maybe an IIS issue, however the system failed to work after I
installed version 2.1 (of Python).  Now it is running Python 2.2 with
win32all-142.  If I have and asp-file containing VBScript code, there
is no problem - a refresh of the page in the web-browser
displays the correct result (as an example a hit counter).

As an example, the start of my asp-file (Python) looks as follows:

<%@ LANGUAGE = Python %>
<%
import Test
reload(Test)
o = Test.Test()
o.ProcessRequest(Request)
%>
<html> ....

    <%= o.Attr('Name') %>
    <%= o.Attr('Email') %>

</html>


So, the first time after having changed the file, the content shown in
the web-browser is correct (values from 'Name' and 'Email' are
correct). If I then press ALT+Refresh (in the web-browser), the page
displayed is all empty (white) - no error messages.  If I save the
file again (update date in file system), and press ALT+Refresh,
content is again ok.

Does anyone have any idea where to look.
I've compared the settings for the IIS server (running Python 2.0)
which works with the one which does not work, and all are identical.

Thanks for any help.

Nikolai

PS: My son has the same problem on an XP professional - all
newly installed.






From h_schneider at marketmix.com  Tue Apr 16 07:36:07 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Tue, 16 Apr 2002 13:36:07 +0200
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
Message-ID: <a9h20k$nb4$00$1@news.t-online.com>

If you have Active State Python, just launch pyppm and type in

search google

Now you can download the api classes from their repository.

Bye,
Harald


"William Dode" <wilk at flibuste.net> schrieb im Newsbeitrag
news:20020416092517.5a9e5845.wilk at flibuste.net...
> hi,
>
> I just read this :
>
> With the Google Web APIs service, software developers can query more than
> 2 billion web documents directly from their own computer programs. Google
> uses the SOAP and WSDL standards so a developer can program in his or her
> favorite environment - such as Java, Perl, or Visual Studio .NET.
>
> Is there a python api somewhere ?
> Google is known to use python, maybe we must write to them to ask a python
> wrapper for their api ?
>
> bye
>
> --
> William Dod? - Informaticien Ind?pendant
> http://www.flibuste.net
>




From noah at noah.org  Tue Apr 23 02:30:33 2002
From: noah at noah.org (Noah Spurrier)
Date: Mon, 22 Apr 2002 23:30:33 -0700
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com> <e2e9e807.0204222041.5fa92f05@posting.google.com> <20020423005326.D21282@eecs.tufts.edu>
Message-ID: <0f0e01c1ea90$5f3b8150$0200a8c0@saysyou>

Way to go, Mentor! ;-)

Noah

----- Original Message ----- 
From: "Michael Gilfix" <mgilfix at eecs.tufts.edu>
To: "Hugh" <h.e.w.frater at cs.cf.ac.uk>
Cc: <python-list at python.org>
Sent: Monday, April 22, 2002 9:53 PM
Subject: Re: need encryption algorithm


>   I hope you're not a CS major or planning on being a CS professional.
> Otherwise, that's gonna suck.
> 
> On Mon, Apr 22 @ 21:41, Hugh wrote:
> > I fully understand what everyone is going on about with regard to me
> > saying that the quality of the encryption is not that important. The
> > fact of the matter is, I'm only writing the software for my 1st year
> > university project and in reality it's only a theoretical system, no
> > one is actually going to use it. I'm using the exercise to learn
> > python and basic database design although to be quite honest I loathe
> > internet programming. Don't know why, just don't like it.
> > 
> > HUgh
> > -- 
> > http://mail.python.org/mailman/listinfo/python-list
> `-> (h.e.w.frater)
> 
> -- 
> Michael Gilfix
> mgilfix at eecs.tufts.edu
> 
> For my gpg public key:
> http://www.eecs.tufts.edu/~mgilfix/contact.html
> 
> 
> 




From l0819m0v0smfm001 at sneakemail.com  Tue Apr  9 18:40:14 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Tue, 09 Apr 2002 22:40:14 GMT
Subject: Yeah whatever... (was RE: PEP 285: Adding a bool type)
References: <mailman.1018296200.32231.python-list@python.org> <a8ufgq$n1v$1@thorium.cix.co.uk>
Message-ID: <3CB3346D.9030801@sneakemail.com>

gbreed at cix.compulink.co.uk wrote:

> The real world's no fun.  A peek at the types module shows that Python 
> already has 33 fundamental types, which means there are already 
> 33!=8683317618811886495518194401280000000 ways of combining them by this 
> strange factorial rule.  It's trivial to show that Python is already 
> impossible to use as a programming language.  Unless perhaps we can all 
> give ourselves super intelligence as well.


    Fortunately, the very act of using Python grants you super 
intelligence.  Unfortunately it tends to erode unless you update 
frequently.<wink>

   Joshua





From cliechti at gmx.net  Mon Apr  1 14:53:09 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Apr 2002 20:53:09 +0100
Subject: Access to main namespace from imported module?
References: <re8haugrbcchbephckl9t1p1s27tvg18hr@4ax.com>
Message-ID: <Xns91E3D4A144441cliechtigmxnet@62.2.16.82>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> wrote in 
news:re8haugrbcchbephckl9t1p1s27tvg18hr at 4ax.com:

> Just curious, like, but is it possible for an included module to
> reference the namespace of the outer-most module, "__main__".

just do an import
>>> import __main__
>>> __main__.var = 7
etc.

chris

-- 
Chris <cliechti at gmx.net>



From mrchameleon at hotmail.com  Tue Apr  9 01:06:16 2002
From: mrchameleon at hotmail.com (Chris Reay)
Date: 8 Apr 2002 22:06:16 -0700
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org> <3CB23D6F.2030006@4-am.com>
Message-ID: <7652139e.0204082106.633363f2@posting.google.com>

>>
>> Not to criticise your project, but this is not a good example to use -- 
>> you've started with a very unnecessarily slow algorithm.  For finding 
>> prime numbers, the Sieve of Eratosthenes is much better.  See e.g.
>> <http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/117119> which I 
>> implemented using simple generators so you don't need the kmax parameter.
>> 
>> -- 
>> David Eppstein       UC Irvine Dept. of Information & Computer Science

Are you kidding, David? That is "very unnecessarily" (sic) banal.


From johannh at uclink.berkeley.edu  Sun Apr 21 14:30:59 2002
From: johannh at uclink.berkeley.edu (Johann Hibschman)
Date: 21 Apr 2002 11:30:59 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com>
Message-ID: <m2adrwzyxo.fsf@uclink.berkeley.edu>

"John Roth" <johnroth at ameritech.net> writes:

> "Carl Banks" <imbosol at vt.edu> wrote in message
> news:tnos9a.de.ln at 127.0.0.1...
> >
> > The study of algorithms, their properties, etc., is science.
> > Implementing a quicksort algorithm is not science, but studying its
> > properties is.
> 
> I would dispute that. The study of algorithms may very well
> be a branch of mathematics, but mathematics is not science.

Exactly.  Computer science got to be a science as some unholy average
between mathematics and engineering.  :-)

Math certainly isn't a science.

--Johann



From marklists at mceahern.com  Thu Apr 25 17:43:36 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 25 Apr 2002 16:43:36 -0500
Subject: configuring mxCGIPython
In-Reply-To: <3CC86709.30433BD3@lemburg.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEKEKMAA.marklists@mceahern.com>

[M.-A. Lemburg]
> No. Could be that libxmlparse is referenced by pyexpat which is part
> of cgipython. On most systems the expat libs should be part
> of the system, so setting LD_LIBRARY_PATH to e.g. /usr/lib
> should help. If not, you can try getting that file from some
> other source and install it in the same dir as cgipython.

That seems to be the case.  I found and uploaded:

  libxmlparse.so.1
  libxmltok.so.1

Now, when I run the shell script that wraps cgipython, I get:

  'import site' failed; use -v for traceback
  Traceback (most recent call last):
    File "../../CGIPython/cgipython.py", line 27, in ?
  IOError: [Errno 2] No such file or directory

Why would it be trying to do an import site?  Hmm.

Thanks,

// mark




From jidanni at hotmail.com  Tue Apr 30 03:30:28 2002
From: jidanni at hotmail.com (Dan Jacobson)
Date: 30 Apr 2002 00:30:28 -0700
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <aact9h$9odgq$2@ID-99513.news.dfncis.de>
Message-ID: <c3d20384.0204292330.ea14313@posting.google.com>

> This will help you get started:
> 
>         http://pythonzh.sourceforge.net/

I see nothing there.

[mail: see jidanni.org]


From tim.one at comcast.net  Sat Apr 13 17:40:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 17:40:32 -0400
Subject: Python 2.2 unwise.exe
In-Reply-To: <6Nd$EXA7vJu8EwPE@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOOONAA.tim.one@comcast.net>

[Robin Becker, on unwise.exe vanishing]
> Installed under C:\Python; system is Win2k. I installed win32all-146
> after Python-2.2, but didn't save backups.
>
> I have a couple of other things installed eg Numeric-20.3 and PIL, but
> nothing  that came with an installer. I don't remember anything untoward
> about the Python-2.2 install.
>
> The unwise.exe was just not there for some reason.

Crud, I was hoping it was unique to Win9x.  "Just not there for some reason"
is the same thing I've seen.  I build the PLabs Windows installer, so I get
more practice with it than anyone else.  "Just not there" has definitely
happened more than once after I've run an install and verified that
UNWISE.EXE is indeed exactly what and where it should be after the install
completes.  Yet two hours or two weeks or two months later, it's plain gone.
Nothing else is missing when this happens, just UNWISE.EXE.  That's all I
know for sure about it ... OK, I'm going to change UNWISE.EXE in my install
trees to read-only.  Maybe that will catch whatever is deleting it in the
act.  Heh:  Just did that.  UNWISE.EXE still existed in my 2.1 and 2.2
trees, but had already gone missing in my experimental 2.3 tree.

windows-is-an-adventure-ly y'rs  - tim

PS: The PLabs Windows installer is a zip file of a kind WinZip knows how to
read.  That is, e.g., you can open Python-2.2.1.exe directly from WinZip.
If UNWISE.EXE goes missing for you, you can extract it from the installer
.exe, using WinZip.  Extract the file UNWISE32.EXE into the root of your
Python installation, and rename it to UNWISE.EXE.  Then the "Uninstall
Python" shortcut should work fine again.  Using an older version of
UNWISE.EXE from somewhere else may or may not work completely; the newer
version we ship knows how to perform rollbacks, and does a much more
complete job of undoing *all* Windows install fiddling than older versions
are capable of doing.





From Answer.via.news.please.prikrylp at nospam_skil.cz  Tue Apr 23 06:43:00 2002
From: Answer.via.news.please.prikrylp at nospam_skil.cz (Petr Prikryl)
Date: Tue, 23 Apr 2002 12:43:00 +0200
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
Message-ID: <aa3dti$n0$1@news.nextra.cz>

Hi,

Being quite new to Python -- is there any preference of
single or double quotes when writing strings?  I am thinking
about cases when both may be used, not when I need to
include the alternative inside the string.

In other words, should I prefer writing:
    s1 = 'first string'
or
    s2 = "second string"

In yet other words, should I prefer the Pascal look of
the strings or the C/C++ look of the strings?

For me, single quotes look less noisy in the source
text (yes I work with C++, but there you have no choice).
Also, Python returns single quotes as the result of repr():

>>> print repr("double")
'double'
>>> print repr('single')
'single'

On the other hand, when the Python code is occasionally
to be converted to C/C++ code, then double quotes
may be appreciated.

While I do not consider this a problem for me (I can sleep
quite well), I think that proposing some prefered way
could make the look of Python sources more similar
each other, i.e. slightly more understandable.  I guess
that it may be important for beginners.  The experienced
user often have reasons to use both styles.

Should this be the part of the Python Style Guide (PEP8)?
I think that it would be nice to add some explanation on this
theme even when there is no prefered way.

Thanks for your opinions,
  Petr

--
Petr Prikryl (prikrylp at skil dot cz)







From borcis at geneva-link.ch  Tue Apr 30 10:40:49 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 30 Apr 2002 16:40:49 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1>
Message-ID: <3CCEACF1.95BE9A2B@geneva-link.ch>

Duncan Booth wrote:
> 
> I prefer writing
>    str.join(' ', aList)
> rather than
>    ' '.join(aList)
> 
> When the string is a literal I find the former reads much better. YMMV

I think the method syntax would be much clearer if the method was
named something like "glue".

  eg   '; '.glue(aList)

Boris Borcic
--
assert BOA in BAOBAB, "Paradise Mismatch Error"


From mlerner at NOSPAM.PLEASE.umich.edu  Mon Apr 29 22:35:16 2002
From: mlerner at NOSPAM.PLEASE.umich.edu (Michael George Lerner)
Date: Tue, 30 Apr 2002 02:35:16 GMT
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com>
Message-ID: <Epnz8.831$SY6.24713@news.itd.umich.edu>

Are you sure you're putting appropriate things into the list?

PythonWin 2.2 (#28, Mar 28 2002, 12:10:20) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2001 Mark Hammond (mhammond at skippinet.com.au) - see 'Help/About PythonWin' for further copyright information.
>>> class MyList(list):
... 	pass
... 
>>> m = MyList()
>>> m.append(1)
>>> m.append(2)
>>> m
[1, 2]
>>> 'apple'.join(m)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: sequence item 0: expected string, int found
>>> m2 = MyList()
>>> m2.append('one')
>>> m2.append('two')
>>> m2
['one', 'two']
>>> 'apple'.join(m2)
'oneappletwo'
>>> 'flonk'.join([1,2])
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: sequence item 0: expected string, int found
>>> 



Fortepianissimo <fortepianissimo at yahoo.com.tw> wrote:
> This question is so fundamental that I strongly suspect it's already
> answered - ig that's the case, please forgive this poor newbie, for he
> has done some work searching over Python FAQ and Google.

> Basically I want to have a subclass of list, like

> class MyList (list):
>     ...

> then I want to

> l=MyList()
> ... do stuff to fill l
> s=" ".join(l)

> with the hope that s will be a nicely joined string with " " as the
> delimiter. Given my C++ background, I tried to define __str__/__repr__
> for both MyList and the user-defined types of the list members,
> assuming the type conversion would silently be invoked. But I was
> wrong!

> Even trying this failed:

> l=[1,2,3]
> s=" ".join(l)

> and I thought built-in types all have default __str__ and __repr__?

> What am I missing here? Many thanks to someone who would take time to
> rescue a poor soul.


> Ben


From ks at da.ru  Wed Apr 24 07:23:08 2002
From: ks at da.ru (Sergey Konopatov)
Date: 24 Apr 2002 04:23:08 -0700
Subject: SMPP(Simple Message Peer to Peer) or UCP module for Python.
Message-ID: <a16e887c.0204240323.4899ae73@posting.google.com>

Hi All!

Does any one heard about subject ?

I don`t wanna use Jython with java module from www.logica.com.
Best Regards, Sergey


From johanfo at essay.org  Mon Apr 29 17:41:06 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Mon, 29 Apr 2002 23:41:06 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net>
Message-ID: <aakelj$do6$1@maud.ifi.uio.no>

> I must admit I have never yet seen a system which integrates HTML and
> code in the same file without the file ending up an unreadable mess.

I just thought of great idea:  How about using one file and not mix html and
python code at the same time ?
Example:
==================================================
<html>
    <body>
    %content%
    </body>
</html>

<!--PYTHON Code
from smartTemplates import *

page.content =  "Isn't this cool ;)"

-->
==================================================

In this example you dont mix the content and code, but keep it in the same
file.
By embedding the code in html comment tags it becomes trivial to edit the
file in most
WYSIWYG editors like Dreamweaver.

--
JF?




From phr-n2002a at nightsong.com  Mon Apr 22 17:44:25 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 22 Apr 2002 14:44:25 -0700
Subject: Encryption
References: <70h8cugniao40148h76fv4o7bgci2j3ddd@4ax.com>
Message-ID: <7x662j76iu.fsf@ruckus.brouhaha.com>

Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> writes:
> Mcrypt can use a number of encryption techniques and I'm not too
> bothered which we use. Just as long as I can decypher it at our end.
> 
> Any suggestions?

Why do you especially want to use mcrypt, in that case?


From hydef at hydef.com  Thu Apr  4 15:02:47 2002
From: hydef at hydef.com (hydef at hydef.com)
Date: Thu, 04 Apr 2002 20:02:47 GMT
Subject: HDTV - DTV - PROJECTION - PLASMA TV'S
Message-ID: <Hj2r8.2360$cC5.641920@newsfeed.slurp.net>

HYDEFINITION
800-233-2430
9AM-5PM M-F

HTTP://WWW.HYDEF.COM

GE 32GT720 $477.36 HDTV-DTV-TV	
GE 35GT720 $826.36 HDTV-DTV-TV	
GE 35GT740 $635.42 HDTV-DTV-TV	
GE 46GW948 $1,006.70 PROJECTION TV	
JVC AV27D302 $294.84 HDTV-DTV-TV	
JVC AV27D502 $351.00 HDTV-DTV-TV	
JVC AV27F702 $486.72 HDTV-DTV-TV	
JVC AV27F802 $589.68 HDTV-DTV-TV	
JVC AV32D302 $477.36 HDTV-DTV-TV	
JVC AV32D502 $542.88 HDTV-DTV-TV	
JVC AV32F702 $776.88 HDTV-DTV-TV	
JVC AV32F802 $851.76 HDTV-DTV-TV	
JVC AV36D302 $720.72 HDTV-DTV-TV	
JVC AV36D502 $786.24 HDTV-DTV-TV	
JVC AV36F702 $1,160.64 HDTV-DTV-TV	
JVC AV36F802 $1,235.52 HDTV-DTV-TV	
JVC AV36PF902 $1,872.00 HDTV-DTV-TV	
JVC AV50D502 $1,272.96 PROJECTION TV	
JVC AV60D502 $1,450.80 PROJECTION TV	
JVC AV61S902 $6,817.50 PROJECTION TV	
JVC GDV4210PZW $5,706.50 PLASMA	
JVC GDV500PZU $11,319.00 PLASMA	
PANASONIC AG527D $509.60 COMBO UNIT	
PANASONIC ATH3015W $28,280.00 HDTV-DTV-TV	
PANASONIC CT27DC50 $480.48 COMBO UNIT	
PANASONIC CT27SX11 $429.00 HDTV-DTV-TV	
PANASONIC CT27SX31 $507.00 HDTV-DTV-TV	
PANASONIC CT32D11 $480.48 HDTV-DTV-TV	
PANASONIC CT32D31 $527.28 HDTV-DTV-TV	
PANASONIC CT32G6 $468.00 HDTV-DTV-TV	
PANASONIC CT32HX41 $1,164.80 HDTV-DTV-TV	
PANASONIC CT32SX31 $790.40 HDTV-DTV-TV	
PANASONIC CT34WX50 $2,060.40 HDTV-DTV-TV	
PANASONIC CT36D11 $673.92 HDTV-DTV-TV	
PANASONIC CT36D31 $720.72 HDTV-DTV-TV	
PANASONIC CT36HX41 $1,456.00 HDTV-DTV-TV	
PANASONIC CT36SX31 $1,164.80 HDTV-DTV-TV	
PANASONIC DTM3050W $3,508.80 HDTV-DTV-TV	
PANASONIC PT42PD3P $5,009.60 PLASMA	
PANASONIC PT47WX49 $1,497.60 PROJECTION TV	
PANASONIC PT50PD3P $9,130.40 PLASMA	
PANASONIC PT51HX41 $1,518.40 PROJECTION TV	
PANASONIC PT52DL10 $7,070.00 PROJECTION TV	
PANASONIC PT56HX41 $1,664.64 PROJECTION TV	
PANASONIC PT56WX51 $2,121.00 PROJECTION TV	
PANASONIC PT56WXF95 $2,121.00 PROJECTION TV	
PANASONIC PT61HX41 $1,927.80 PROJECTION TV	
PANASONIC PT65WX51 $2,403.80 PROJECTION TV	
PANASONIC PVC2541 $305.76 COMBO UNIT	
PANASONIC PVC2580 $399.36 COMBO UNIT	
PANASONIC PVC2780 $499.20 COMBO UNIT	
PANASONIC PVDF2700 $790.40 COMBO UNIT	
PANASONIC PVDM2791 $648.96 COMBO UNIT	
PANASONIC PVQ2511 $282.88 COMBO UNIT	
PANASONIC PVSS2710 $648.96 COMBO UNIT	
PANASONIC TH37PWD4UZ $4,393.50 PLASMA	
PANASONIC TH42PWD4UY $4,999.50 PLASMA	
PANASONIC TH50PHD3U $8,837.50 PLASMA	
PHILIPS 27DV60S $941.35 COMBO UNIT	
PHILIPS 27PT81S $543.39 HDTV-DTV-TV	
PHILIPS 30PW9817 $1,890.08 HDTV-DTV-TV	
PHILIPS 32PT71B $545.30 HDTV-DTV-TV	
PHILIPS 32PT81S $862.36 HDTV-DTV-TV	
PHILIPS 32RF50S $755.50 HDTV-DTV-TV	
PHILIPS 34PW9817 $2,283.69 HDTV-DTV-TV	
PHILIPS 34PW9847 $3,000.33 HDTV-DTV-TV	
PHILIPS 36PT41B $750.89 HDTV-DTV-TV	
PHILIPS 36PT71B $781.92 HDTV-DTV-TV	
PHILIPS 42FD9932 $6,099.98 PLASMA	
PHILIPS 50P8341 $1,348.86 PROJECTION TV	
PHILIPS 55P8341 $1,478.19 PROJECTION TV	
PHILIPS 55P9271 $1,664.13 PROJECTION TV	
PHILIPS 55PP9401 $2,091.91 PROJECTION TV	
PHILIPS 55PP9701 $3,148.93 PROJECTION TV	
PHILIPS 60P8341 $1,626.32 PROJECTION TV	
PHILIPS 60PP9401 $2,214.96 PROJECTION TV	
PHILIPS 60PP9601 $2,751.77 PROJECTION TV	
PHILIPS 60PP9701 $3,474.57 PROJECTION TV	
PHILIPS 64P8341 $1,830.94 PROJECTION TV	
PHILIPS PA9032C $854.69 HDTV-DTV-TV	
PHILIPS PPC932C $807.91 HDTV-DTV-TV	
PHILIPS SC3932C $877.44 HDTV-DTV-TV	
QUASAR SP3234 $463.84 HDTV-DTV-TV	
RCA D36120 $1,286.75 HDTV-DTV-TV	
RCA D36TF30 $1,513.76 HDTV-DTV-TV	
RCA D52120 $1,513.76 HDTV-DTV-TV	
RCA D52130 $1,513.76 PROJECTION TV	
RCA D61130 $2,044.16 PROJECTION TV	
RCA F32648 $461.45 HDTV-DTV-TV	
RCA F32649 $432.81 HDTV-DTV-TV	
RCA F32669 $455.08 HDTV-DTV-TV	
RCA F32689 $509.18 HDTV-DTV-TV	
RCA F32715 $509.18 HDTV-DTV-TV	
RCA F36649 $648.15 HDTV-DTV-TV	
RCA F36669 $702.25 HDTV-DTV-TV	
RCA F36689 $700.13 HDTV-DTV-TV	
RCA F36715 $700.13 HDTV-DTV-TV	
RCA F38310 $1,962.48 HDTV-DTV-TV	
RCA G27648 $508.12 HDTV-DTV-TV	
RCA G27649 $508.12 HDTV-DTV-TV	
RCA G32705 $865.61 HDTV-DTV-TV	
RCA G36705 $1,135.06 HDTV-DTV-TV	
RCA J32430 $583.44 HDTV-DTV-TV	
RCA J32530 $689.52 HDTV-DTV-TV	
RCA J36530 $970.63 HDTV-DTV-TV	
RCA MM32110 $1,211.43 HDTV-DTV-TV	
RCA MM36110 $1,484.08 HDTV-DTV-TV	
RCA MM52110 $1,855.36 PROJECTION TV	
RCA MM61110 $1,899.89 PROJECTION TV	
RCA P46921 $1,080.96 PROJECTION TV	
RCA P52936 $1,135.06 PROJECTION TV	
RCA P52955 $1,157.33 PROJECTION TV	
RCA P52960 $1,157.33 PROJECTION TV	
RCA P61310 $2,574.56 PROJECTION TV	
RCA P61939 $1,351.46 PROJECTION TV	
RCA P61960 $1,362.07 PROJECTION TV	
RCA PR42300 $5,104.72 PLASMA	
RCA SCENIUM HD65W20 $2,999.00 PROJECTION TV	
RCA SCENIUM L50000 $5,437.40 PROJECTION TV	
RCA SCENIUM PHD50300 $9,720.02 PLASMA



From reply at in_newsgroup.tks  Mon Apr  8 13:35:09 2002
From: reply at in_newsgroup.tks (Scrumpy)
Date: 8 Apr 2002 17:35:09 GMT
Subject: Newbie - web app deployment issues
Message-ID: <a8skcd$mr5$1@hfc.pacific.net.hk>

Hi all,

Thanks for your feedback to my earlier post "Newbie's 1st Web App 
decisions!"

As a Python newbie who wants to try developing web apps, I want to avoid as 
much as possible, the need to deploy to specialized hosting environments. 
For example, out of the many hosting companys in the world, only a very 
small percentage provide Zope hosting. I haven't confirmed this but there's 
probably a premium to pay for Zope hosting (which may be warranted but it's 
not something I want to pay for at this point in time).

Some web hosting companys are more accomodating than others i.e. after 
evaluation, some may add customer requested/coded python modules, others 
will refuse and others will not install Python at all.

At least in the near future, hosting my own internet site or having a whole 
server rented/co-located at an ISP/hosting company is not an option so I 
have to rely on what is provided by, or will be accepted by hosting 
companys on their existing web servers.

***
My question for those who have faced a similar situation (hopefully that 
means many of you ;-) ) is:

How did you get your Python based web app hosted i.e. what 
modules/components/frameworks did you use to run your app, which were 
allowed by a 'non-specialized' web host i.e. a host that probably runs 
Apache and has Python and a few modules preinstalled?
***

I would be 'most' interested to hear from people using the project(s) 
listed below but I would also appreciate feedback if you have a good 
solution which doesn't include them.

The web app servers/frameworks that I am looking into in more detail are:

Slither http://slither.toolsofcomputing.com/
Webware for Python http://webware.sourceforge.net/
Skunkweb http://skunkweb.sourceforge.net/about.html

Thanks a bunch!

Cheers,
Scrumpy :)

This has been a Python newbie production!


From peter at engcorp.com  Tue Apr 16 01:18:33 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 16 Apr 2002 01:18:33 -0400
Subject: classes + modules
References: <3cbbac29$1_2@nopics.sjc>
Message-ID: <3CBBB429.913C2294@engcorp.com>

Adonis Vargas wrote:
> 
> is there any way to initialize a class then get the initialized class name
> and put it into a list then later on calling it from the list as if it were
> the class?
> 
> class_list = []
> from somemodule import *
> x = someclass()
> class_list.append(x)
> class_list[0].somefunction() # assuming x is position at 0

Hey, come on, this is _Python_, man!  Try it!  :-)

C:\>python
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> class SomeClass:
...   def someMethod(self):
...     print "It works!"
...
>>> classList = []
>>> x = SomeClass()
>>> classList.append(x)
>>> classList[0].someMethod()
It works!


-Peter


From aahz at pythoncraft.com  Fri Apr 26 00:23:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 26 Apr 2002 00:23:13 -0400
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc> <j4helzdfbl.fsf@informatik.hu-berlin.de> <aa98g0$9i4$1@panix3.panix.com> <slrnacgag6.98m.akuchlin@ute.mems-exchange.org>
Message-ID: <aaaknh$3uk$1@panix2.panix.com>

In article <slrnacgag6.98m.akuchlin at ute.mems-exchange.org>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>In article <aa98g0$9i4$1 at panix3.panix.com>, Aahz wrote:
>>
>> What HOWTO?  There isn't one at http://py-howto.sourceforge.net/
>
>http://pyxml.sourceforge.net/topics/docs.html links to
>http://py-howto.sourceforge.net/xml-howto/xml-howto.html, but the text
>should really live on pyxml.sourceforge.net, because the master copy
>of the text lives in the PyXML CVS tree.

Ah.  And conversely, py-howto should link to pyxml.  Well, I don't have
access to either of those, so someone else needs to fix it.  Martin?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From pinard at iro.umontreal.ca  Mon Apr 29 13:45:30 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 13:45:30 -0400
Subject: Python Webpages
In-Reply-To: <slrnacqrps.sdd.jon+usenet@snowy.squish.net>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net>
	<aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no>
	<slrnacpc38.obc.jon+usenet@snowy.squish.net>
	<23891c90.0204290051.3edcc62d@posting.google.com>
	<slrnacq7jk.gav.jon+usenet@snowy.squish.net>
	<mailman.1020090053.19284.python-list@python.org>
	<slrnacqrps.sdd.jon+usenet@snowy.squish.net>
Message-ID: <oqhelupfet.fsf@titan.progiciels-bpi.ca>

[Jon Ribbens]

> [...] I do not think that code of any sort should be in the HTML
> pages. Code and HTML are separate things that belong in separate files
> rather than muddled together.

I presume this is debatable.  Up to now, I found easier having single files,
for the case you are adjusting both aspects in a single editing session.

This is a bit like if a programming language was forcing comments in a
file and code in another.  It is more easy editing comments and the code
together.  Or course, we have separate manuals for bigger things -- but I
like to think that active HTML pages are like very small projects, edited as
needed by both by the HTML designer (who is not necessarily a programmer)
and the application programmer (who is not necessarily an HTML designer).
Yet, seeing the whole thing, both the designer and the programmer can leave
the page in a consistent state, even if not fluent in the _other_ field.

The same applies for `awk' or `yacc' which mixes patterns or grammars with
actions related to them.  And for all kind of literate programming.

> Linking web pages with installed Python modules is exactly what jonpy does.
> [...] Yes, it is my belief that jonpy, while being very simple, is also
> very powerful.

Good! :-) I felt a need for simplicity, leaving the bulk of the power to
Python itself.  My experience taught me that it is attractively simple to
use a single file instead of two, for linking HTML and Python into active
pages.  Of course, if we have lengths of Python code, these are best kept
in separate (and compiled) modules, merely imported from the HTML page.
For better style, and a bit for better speed, the amount of Python code
in HTML pages should stay small.

> Hence my question to Paul about what features of other more complicated
> templating systems make him feel that they have "extra power" in
> comparison, because personally I can't think of any.

PHP is a clear example of bloat.  There are others!  PHP has a lot of things,
and is a little monstrous to learn fully.  I agree with you that the extra
power is not worth all the time it takes for one to study the tool, and
easily defeated with only a few lines of Python, here and there, as needed.

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




From aleax at aleax.it  Wed Apr 17 10:22:19 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 14:22:19 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <3CBD7406.3050909@mxm.dk> <3CBD80B2.91A7D596@yahoo.com>
Message-ID: <vyfv8.30552$vF6.951821@news2.tin.it>

dominikush at yahoo.com wrote:
        ...
> I understand that! But my main objection remains: Why should
> I care about the type of object being handled by a dictionary?

You shouldn't: whatever types of objects are the dictionary's values,
method .values() will ALWAYS return a new list object whose items
refer to EXACTLY the same objects as the dictionary's values.


> This could not only break unit tests but also make maintenance
> of ``old'' code somewhat painful.

I don't understand what you're talking about.  My sentence above
was true at least since Python 1.5.2 (probably earlier, but that's
when I hopped aboard).


Alex



From Chris.Barker at noaa.gov  Fri Apr 26 15:19:34 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 26 Apr 2002 12:19:34 -0700
Subject: Numeric question
References: <pan.2002.04.25.21.32.03.310021.2578@hooper.net> <YM2y8.71958$HH5.5067@rwcrnsc51.ops.asp.att.net>
Message-ID: <3CC9A844.F1BE7E7B@noaa.gov>


Emile van Sebille wrote:

> OK.  It's x.  So let's build a contiguous x and try it again:

> >>> x = array(inputs[0][:])

or

x = inputs[0][:].copy()

The copy method isn't in the version of the docs I have, but it's very
handy and quite clear as to what it is doing.

or

x = inputs[0,:].copy()

Numpy arrays are N-dimensional, you might as well index them that way.

come to think of it, why not:

inputs[0].copy() ?


-Chris







-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From roy at panix.com  Sat Apr 13 20:20:11 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 13 Apr 2002 20:20:11 -0400
Subject: 2.1 vs. 2.2
References: <mailman.1018739908.9275.python-list@python.org>
Message-ID: <roy-56DE95.20201113042002@news1.panix.com>

Tim Peters <tim.one at comcast.net> wrote:
> When enough people who feel as you do volunteer all the time and money it
> takes to produce an ISO/ANSI standard.  If you haven't been involved in that
> process, I can tell you that the effort it requires is staggering.  I know
> of no plans to do anything of the sort for Python.

ISO certification is a wonderful thing.  It lets people who worry about 
that sort of stuff check off the "ISO certification" checkbox and be happy.  
Other than that, it's not clear it has any value.


From gimbo at ftech.net  Fri Apr  5 09:43:26 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Fri, 5 Apr 2002 15:43:26 +0100
Subject: file rewind?
In-Reply-To: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>
References: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>
Message-ID: <20020405144326.GF796@andy.tynant.ftech.net>

> when I run the module from the comman line by itself, it counts the lines
> alright, but does not count the characters.  do I need to "rewind" the file
> reference after the first function call to countLines()?

As has been suggested, you could use seek to rewind to the start of
the file.  But why read the file twice when you could read it once?
You could rewrite countLines so they accept a list of strings instead
of a file object, then do this:

def test(name):
    file = open(name, "r")
    filelines = file.readlines()
    countLines(filelines)
    countChars(filelines)

However, you can do away with countLines and countChars entirely:

def wc(name):
    file = open(name, "r")
    data = file.read()
    chars = len(data)
    lines = string.count(data, "\n")
    return (chars, lines)

The standard library is a wonderful thing - explore it and use it!
I'm pretty sure string.count is actually implemented in C in which
case I promise you it'll be faster than countLines().  End result:
faster, shorter code without reinventing the wheel.

One thing though: if the file could be very large, you probably don't
want to read it all into memory at once, in which case you should read
bite-sized chunks instead.  The trade off is code complexity vs memory
usage, but it's still readily comprehensible:

import string
import sys

def wc(filename, chunksize=1024):
    (chars, lines) = (0, 0)
    file = open(filename, "r")
    while 1:
        data = file.read(chunksize)
        if not data:
            break
        chars = chars + len(data)
        lines = lines + string.count(data, "\n")
    return (chars, lines)

if __name__ == '__main__':
    filename = sys.argv[1]
    (chars, lines) = wc(filename)
    print "%s has %s characters and %s lines" % (filename, chars, lines)

HTH,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From mlh at vier.idi.ntnu.no  Thu Apr 11 22:19:40 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 02:19:40 +0000 (UTC)
Subject: Tuple comprehension
References: <slrnaba052.efc.mlh@vier.idi.ntnu.no>  <%r7t8.37204$CU.41841@rwcrnsc51.ops.asp.att.net>
Message-ID: <slrnabch1s.c8l.mlh@vier.idi.ntnu.no>

In article <%r7t8.37204$CU.41841 at rwcrnsc51.ops.asp.att.net>, Emile van
Sebille wrote:
>Magnus Lie Hetland
>> As far as I can see, the PEP on dict comprehension is still open... I
>> think both dictionary comprehension and tuple comprehension seem like
>> natural parallels to list comprehension. But then I would, wouldn't I?
>> <wink>
>>
>
>Why not just tuple([my list comp]) ?

I thought I mentioned that in my post. Oh, well.

Of course you can do that. Just like you can use a for-loop instead of
list comprehensions -- they're not essential at all either. Just
compact and readable.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From mhaag at pacific.net  Sun Apr 21 17:17:07 2002
From: mhaag at pacific.net (MH)
Date: Sun, 21 Apr 2002 14:17:07 -0700
Subject: Newbie programmer here...
Message-ID: <uc73k44mopm050@corp.supernews.com>

I'm a novice programmer. I wrote my first program, a simple database, in 
BASIC a long time ago, and have used Visual Basic mostly since then. I have 
some experience with SQL, also.

I have recently been wrestling with PERL as a means to create interactive 
websites, but I find PERL incredibly inconsistent and confusing, and the 
debugging features I think are sorely lacking. I spend HUGE amounts of time 
solving very basic problems.

I am looking for ANYTHING but PERL to accomplish basic tasks such as 
collecting, processing, and displaying data via CGI/HTML. Is PYTHON a good 
candidate for this? What about PHP? Suggestions?




From aahz at pythoncraft.com  Sat Apr 27 20:02:57 2002
From: aahz at pythoncraft.com (Aahz)
Date: 27 Apr 2002 20:02:57 -0400
Subject: Stackless goes Limbo
References: <mailman.1019911009.16159.python-list@python.org> <aaee44$p73$1@panix1.panix.com> <1019928879.371845@yasure>
Message-ID: <aafe7h$kfb$1@panix1.panix.com>

In article <1019928879.371845 at yasure>, Donn Cave <donn at drizzle.com> wrote:
>Quoth aahz at pythoncraft.com (Aahz):
>| In article <mailman.1019911009.16159.python-list at python.org>,
>| Laura Creighton  <lac at strakt.com> wrote:
>|>
>|> For those of you who can't live without a Limbo introduction written
>|> by a real C inventor:
>|>
>|> http://www.vitanuova.com/inferno/papers/limbo.html
>|
>| Ugh.  I've gotten so corrupted by Python I can't even read well-written
>| C code.
>
>Tsk, that ain't C.  

Looks like C, smells like C.  ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From ianb at colorstudy.com  Mon Apr 29 01:42:28 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 29 Apr 2002 00:42:28 -0500
Subject: Python Webpages
In-Reply-To: <20020428195607.GA19697@lilith.my-fqdn.de>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net> 
	<20020428195607.GA19697@lilith.my-fqdn.de>
Message-ID: <1020058948.2284.275.camel@localhost>

On Sun, 2002-04-28 at 14:56, Gerhard H?ring wrote:
> * Magnus <nixx at telia.com> [2002-04-28 18:59 +0000]:
> > [WebWare]
> > Another powerful (not yet final though) module is Middlekit which e.g. makes 
> > it easy to deploy a servlet that acts with some glueware to e.g. MySQL.
>  
> The more I think about it, the less I get the point of object-relational
> mappers like MiddleKit. Especially the ones that require you to rebuild
> the database schema every time you change the object schema. AFAIC
> MiddleKit falls into this category.
> 
> I think I'd prefer going either straight SQL with a dictionary mapper
> (pyPgSQL's PgResultSet has such a thing built in [1]) or using an object
> database like ZODB.
> 
> [1] in pyPgSQL, you can say, for example:
> cursor.execute("select id, name from test")
> print cursor.id, cursor.name
> print cursor['id'], cursor['name']
> print cursor[0], cursor[1]

The advantage of a more complete middleware is that you can add
functionality to the objects -- if you are creating MiddleKit objects
from the database schema, and you never add anything to those objects,
then it isn't nearly as interesting.

For instance, in some of my applications I associate images with rows,
but I don't store the images in the database, they are just on disk
somewhere.  My middle object has methods to manipulate that image, give
the URL for it, etc. -- and it has the same basic interface as database
access.  Or you might have a derived value -- say a case where you
require a user give an email, and they may also give their real name. 
So you change the name method to return their given name, if they
provided it, or their email if they didn't.

That sort of thing is the "middle" in middleware.  You don't need
something like MiddleKit to make your own middleware, of course -- you
can write your own objects that fetch rows from the database and provide
an interface to that information.  That's perfectly viable, but it tends
to feel a tad tedious.  You'll probably find yourself recreating
something like MiddleKit.

  Ian






From mal at lemburg.com  Thu Apr  4 03:53:51 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 04 Apr 2002 10:53:51 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com>  
	            <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>
Message-ID: <3CAC149F.3F1BA77B@lemburg.com>

Guido van Rossum wrote:
> 
> > Right, but my point is that if you want to make booleans first class
> > objects in Python, you should consider making them behave as defined
> > in the text books rather than trying to make them feel like integers
> > but look like truth values.
> 
> I reject this reasoning.  Python is not trying to follow the text
> books.  (In fact, maybe I should say "Python is trying not to follow
> the text books.")

Uhm, I can't follow you here -- which text books you are talking about 
here (Python has always tried to conform to standards in most areas). 
FWIW, I was talking about the text books students use to learn boolean 
algebra.

BTW, I wonder why you didn't follow up on my other comments...

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From jb at cascade-sys.com  Tue Apr 23 02:35:27 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 22 Apr 2002 23:35:27 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz>
Message-ID: <3CC500AF.8BAE6F7E@cascade-sys.com>

Paul Foley wrote:

> On Sun, 21 Apr 2002 02:32:29 -0700, James J Besemer wrote:
>
> A lot of people think "Lisp" means "Scheme".
> Scheme sucks, so they think "Lisp" sucks.

That complaint seems specific to the Lisp community.  I don't sense that a lot of
people here know either language well enough to be making that particular complaint.

> Well, that's not entirely true.  Lisp has always used lexical scoping
> in compiled code.  There used to be a difference between compiled and
> interpreted code, though, where the interpreter used (only) dynamic
> scope.

My statement was 100% accurate, as "original Lisp" had no notion of compiled code.
The original semantics (as formally defined in Lisp itself [McCarthy]) called for
dynamic binding.  Most original implementations did it that way, many later
experimented with alternative approaches, and ultimately static binding emerged as
the preferred way to do things.

I find it highly ironic that Perl originally did it the wrong way ("local"), although
later they introduced a now-recommended static binding ("my").

> Hardly.  Syntax is perhaps /the/ single /best/ characteristic of Lisp.

Well, I have to back down and admit that good or bad is a value judgement.

I actually was conceeding to most peoples' negative view of the matter.  Personally I
am neutral on Lisp syntax and appreciate its several benefits.  With the right pretty
printer (regular old Vi's lisp mode works just fine) the parentheses pretty much
cease to be an issue.

However, you have to admit that having to surround virtually each and every
statement, expression, sub-expression, etc. in a pair of parentheses is at the very
least more tedious and error prone for the programmer than, say, using operator
precedence to organize arithmetic, semicolons or newlines to separate statements and
braces or indentation to group statements.  So although I don't fully agree, I have
to sympathise with people who say the syntax is it's worst feature.  And I very much
prefer controlling the semantics with an "Algol-like" syntax.

> [And it's worth noting that, as with other aspects of Lisp, people
> just keep reinventing it, badly --

Hmmm.... I was recently accused of making this claim, which is entirely untrue.
Evidently you've argued this point previously in this forum?

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From phr-n2002a at nightsong.com  Wed Apr  3 04:15:09 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 01:15:09 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <7xelhx3zz6.fsf@ruckus.brouhaha.com>

vio <vmilitaru at sympatico.ca> writes:
> I need to translate a (rather voluminous 100MB) java program to
> python. After a few searches on google, apparently the best tools
> for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> on the Lex-YACC HOWTO, I would be very interested in hearing of
> others' successful experiences/feedback at this. Any pointers of
> interest would similarily be appreciated.

That's crazy.  Don't even think about it.  If you want Python-Java
integration, use Jython.


From starx at pacbell.net  Sat Apr  6 16:34:42 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 6 Apr 2002 21:34:42 GMT
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com> <3cadd841.225619243@news.laplaza.org> <Xns91E81370BB5FERASXnewsDFE1@130.133.1.4> <3CAF41F5.83060882@earthlink.net>
Message-ID: <Xns91E88A4AB1DBRASXnewsDFE1@130.133.1.4>

Hans Nowak || Sat 06 Apr 2002 10:41:22a:

> Philip Swartzleonard wrote:
> 
>> Well, i'm not someone to ask about the detailed meanings of japanese
>> words, but i'm bored, so i did some dictionary diving on these terms.
>> Here's a couple hits (kanji meanings in parens).
>> 
>> keishikibi (form/style - system/expression - beauty) 'beauty of form'
>> kotan (wither/die - faint/pale O_O) 'elegant simplicity'
>> tanryoubi (simple - clear - beautiful) 'i just made this word up'
> 
> That last one looks suspiciously like "Ruby". You're not trying
> to feed us subliminal messages are you? :-)
> 

No, no. Just stuck three kanji together, ryou and bi being seperate ones. 
I honsetly don't know anything at all about ruby except the occsainal 
mumbel on here or rec.games.roguelike.devel =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From ykingma at accessforall.nl  Wed Apr 24 15:12:08 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Wed, 24 Apr 2002 20:12:08 +0100
Subject: Slight irritation with try/finally indentation
References: <mailman.1019657853.14965.python-list@python.org>
Message-ID: <3CC70381.50E25C28@accessforall.nl>


Fran?ois Pinard wrote:
> 
> Hello, people.
> 
> There is something which I find slightly irritating with the try/finally
> construct, and wonder if someone would not have a good advice to offer.
> I guess it is very usual that one has, within a suite, something like:
> 
>     [...]
>     setup()
>     process()
>     cleanup()
>     [...]
> 
> with `cleanup' undoing the effect of `setup', more or less, really seeing
> `setup()' and `cleanup()' as a kind of big parentheses around `process()'.
> In a context where `process()' may raise exceptions, `cleanup()' might be
> mandatory once `setup()' has completed, so one rewrites the above as:
> 
>     [...]
>     setup()
>     try:
>         process()
>     finally:
>         cleanup()
>     [...]
> 
> What I find irritating is that `cleanup()' is not aligned anymore with
> `setup()', as it was originally, so we loose on the legibility of the
> parenthetical idiom we wanted to stress.  Surely, I may _not_ write:
 
It's a bit irritating indeed. I occasionaly use the non indented form:

    setup()
    try: process()
    finally: cleanup()

but it only works for single lines and gives a bit more work when
lines have to be added, so this form tends to disappear over time...

Have fun,
Ype


-- 
email at xs4all.nl


From noah at noah.org  Tue Apr 23 13:57:00 2002
From: noah at noah.org (Noah)
Date: Tue, 23 Apr 2002 10:57:00 -0700
Subject: non-blocking sockets
In-Reply-To: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf>
Message-ID: <PAENINBOOODFDEPOKPKCOEELCMAA.noah@noah.org>

You probably get an IOError or an OSError.
You should be able to catch any generic exception
(see the code at the end).

Getting an exception on a non-blocking read is
normal behavior. This is how Python tells you that
the there was no data to read. Generally you should
use 'select' before you try to call a non-blocking read,
but there are times when you may want to do this.

If we assume that this is the case then what you are getting 
is an IOError or OSError because an behind-the-scenes system call
set errno to be not zero. I am not sure how setblocking(0) works, 
but it probably sets the O_NONBLOCK flag on the file descriptor
(which you can also do using the fcntl module).
At the system level when you have O_NONBLOCK set on a socket and 
you try to read from an empty socket then read() returns -1 and 
sets errno to EAGAIN.
The problem is that Python interprets errno as an exception, 
(well, not really a problem, just a better way of looking at it...)
When a Python function calls the underlying system function
if that system function sets errno to be not zero 
then Python will throw an exception. The trick is that you can 
get different exceptions depending on how you call read -- 
depsite the fact that the same low-level system function is being called. 
If you call os.read() then you may get an OSError.
If you call Python's regular read then you may get an IOError.
I am not sure what you may get if you call the Socket Object methods.

Fortunately we have the errno module, so when you get an
exception you can check the errno value. The exception object
that you catch will have an errno attribute. Just compare that value
to the contstants in the errno module. For example:
	except IOError, ex:
		if ex.errno == errno.EAGAIN

Try this code. Let me know if this makes things more clear:

	import errno

        try:
                read_foo (my_socket)   ### YOUR STUFF HERE
        except OSError, e:
                print e.str()
                print 'errno:', e.errno
                if e.errno == errno.EAGAIN:
                        print 'OSError: EAGAIN'
        except IOError, e:
                print e.str()
                print 'errno:', e.errno
                if e.errno == errno.EAGAIN:
                        print 'IOError: EAGAIN'
        except Exception, e:
                print e.str()
                print 'errno:', e.errno
                print 'Something else happened!'

Yours,
Noah



-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of so very tired
Sent: Tuesday, April 23, 2002 8:01 AM
To: python-list at python.org
Subject: non-blocking sockets


When trying to read from a socket, I don't want it hanging, so I set it to
non-blocking by calling

setblocking(0)

but then when I try to read from it and there is no data in the socket, I
get an error. I tried this in both windows and linux and in both I get an
error and the program stops. It says there's an exception but it doesn't
say the name of the exception so I can't even try to catch.
Does anyone know how I can check if there's data coming from a socket
connection without hanging or having the program barf on me?
Thanks.
mRiaz







From Answer.via.news.please.prikryl at skil.nospam.cz  Fri Apr 19 08:08:40 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Fri, 19 Apr 2002 14:08:40 +0200
Subject: How to save bitmap from MS Windows clipboard to xxx.png?
Message-ID: <a9p1e3$efb$1@news.nextra.cz>

Hi,

I am quite new to Python to now what was done already
and what was not.

What I need is to save the window snapshot (in MS
Windows) into PNG file.  It just came to my mind that
some Python utility may already exist for that task
or that some PNG support is implemented.

How I am doing it now
=====================

Say, some error message box is displayed by a MS
Windows application.  I do press `Alt-Print Screen`
to copy the bitmap of the window into the clipboard.

Then I run the graphic editor (CorelXara in my case),
paste the content of the clipboard into the new
document, and export the result into PNG file.

I am converting the bitmap to the form with 4 bits
per pixel, optimized palette, no dithering, no
interlace.  This way the snapshots files are
reasonably small and very well displayed on HTML
pages (yet another bug tracing + doc system ;-).


How would I like to do it
=========================

It would be nice if I could simply capture the bitmap
into clipboard and then just run some Python script
(command line) with filename as the only argument.


If there is nothing like this ready to run, could you
point me to some resources that would help me to
write such utility?  Any other comments?


Thanks for your time and the experience,
  Petr

--
Petr Prikryl (prikrylp at skil dot cz)





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 15:44:26 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 12:44:26 -0700
Subject: Does Python need an 'xor' operator?
Message-ID: <3cb88af1@news.mhogaming.com>

I think an 'xor' operator is needed.  The 'xor' operator means
'logical-exclusive-or', and returns a 1 or 0.

Here is some code that does what I want (in all the cases I can think of) for
'xor':

# logical-exclusive-or operator:
def xor(a,b):
    return (not a) ^ (not b)

This will work, but I think it is less clear (in an expression) than: 'a xor b'.





From steffen at molli15.org  Sat Apr 13 12:50:58 2002
From: steffen at molli15.org (Steffen Kirschke)
Date: Sat, 13 Apr 2002 18:50:58 +0200
Subject: Looking for doh!.. simple sequence type
References: <3CB6BED8.2000502@mxm.dk>
Message-ID: <a99k6c$g9c$1@news.ilk.de>

Max M wrote:

> Hmm ... I am trying to create a simple sequence type that I cant iterate
> over like (Python 2.1)::
> 
> lst = Lst()
> for i in lst:
>      print i
> 
> And I thought I could do it like:
> 
> class Lst:
> 
>      def __len__(self):
>          return 4
> 
>      def __getitem__(self, idx):
>          return 42
> 
> But I have to write it like below to avoid an endless loop..
> 
> lst = Lst()
> for i in range(len(lst)):
>      print i,raise StopIteration
> 
>  >>>42 42 42 42

--<schnipp>--

class Lst:
        def __init__(self):
                self.iter = 0
                self.len = 4

        def __getitem__(self,key):
                if self.iter == self.len:
                        self.iter = 0
                        raise StopIteration
                else:
                        self.iter += 1
                        return 'The answer'


or, instead of __getitem__

        def __iter__(self):
                self.iter = 0
                return self

        def next(self):
                if self.iter == self.len:
                        raise StopIteration
                else:
                        self.iter += 1
                        return 42
--<schnapp>--

The second one works afaik from 2.2 up only..

hth

Steffen



From max at alcyone.com  Thu Apr  4 19:19:40 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 04 Apr 2002 16:19:40 -0800
Subject: 1 .__nonzero__() => memory fault (2.2 win32)
References: <a8ipmo$qq0$1@216.39.172.122>
Message-ID: <3CACED9C.DF372C5E@alcyone.com>

Bengt Richter wrote:

>  Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>  Type "help", "copyright", "credits" or "license" for more
> information.
>  >>> hasattr(1, '__nonzero__')
>  1
>  >>> 1 .__nonzero__()
> 
> (You get a memory fault).

Whoops, I get the same thing.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From tejarex at yahoo.com  Fri Apr 12 09:49:28 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Fri, 12 Apr 2002 13:49:28 GMT
Subject: Variables common to all objects of a class
References: <3CB6CD99.70409@thomas-guettler.de>
Message-ID: <IBBt8.9234$K5.826062@bin5.nnrp.aus1.giganews.com>

"Thomas Guettler" <zopestoller at thomas-guettler.de> wrote in message
news:3CB6CD99.70409 at thomas-guettler.de...
> Hi!
>
> How do you store variables which should be common to all
> objects of a class?

You should call instances of a class instances since 'object' is a
generic term that in Python includes everything..  Class variables are
usually defined within the class statement.  They can also be assigned
after creation of the class, as in

class.attribute = value

> I would do it like this. Is this the right pyhton-way?

To my knowledge, no.

> #foo.py
> class_variable=0

This is a module (aka global) variable.

> def class_method():
>      print "I work only on class variables"

Putting class methods outside the class is a kludge was previously
necessary since methods within the class are taken to be instance
methods.  The 2.2 release seeks to remedy this with the new mechanism
for static and class methods.

> class foo:
>      object_variable=1

This is a class variable (altough some people define instance
variables as class variables to document or give default values).  It
can be accessed directly as foo.object_variable or indirectly as
instance.object_variable if there is no instance variable of the same
name.

>      def my_class_variable(self):
>          print "cv", class_variable, "ov", self.object_variable
>
> if __name__=="__main__":
>      f=foo()
>      f.my_class_variable()
>      import foo
>      foo.class_method()

Importing a module into itself strikes me as problematic.

Terry J. Reedy





From aahz at pythoncraft.com  Sat Apr 20 20:33:06 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Apr 2002 20:33:06 -0400
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <mailman.1017817926.31983.python-list@python.org> <iyAq8.29179$n4.5840426@newsc.telia.net>
Message-ID: <a9t1c2$pmt$1@panix1.panix.com>

In article <iyAq8.29179$n4.5840426 at newsc.telia.net>,
Fredrik Lundh <fredrik at pythonware.com> wrote:
>Delaney, Timothy wrote:
>>
>> The whole point is that this is the *input*, optionally to produce another
>> format, and is also human-readable in the source code (or any other text
>> file).
>
>who cares, when it's not human-writable.

Mind providing at least some anecdotal evidence for this assertion?  I'm
speaking as someone who uses vi to write XML and HTML with some
proficiency, but who does find XML/HTML slightly awkward.  I have no
experience yet with reST.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From mwh at python.net  Wed Apr 10 10:54:20 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 10 Apr 2002 14:54:20 GMT
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
References: <mailman.1018447222.24371.python-list@python.org>
Message-ID: <uit6zvbur.fsf@python.net>

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

> Hello!
> 
>    Environment: SUN Sparc Ultra-1, Solaris 2.5.1, gcc 2.8.1.
                                                        ^^^^^
Is this negotiable?  I.e. could you try upgrading?  This is pretty old now.

>    Python 1.5.2, 2.0, 2.1{.1,.2,.3} and Python 2.2 all had been compiled
> successfully.
> 
>    Python 2.2.1 would not compile. The problem:

I have a hard time believing 2.2 compiled but 2.2.1 does not.  The
only change the "#ifdef __APPLE__" block you mention below.

> gcc -c -DNDEBUG -O2 -I. -I./Include -DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
> In file included from Python/thread_pthread.h:9,
>                  from Python/thread.c:113:
> /usr/include/pthread.h:238: parse error before `destructor'
> /usr/include/pthread.h:238: parse error before `)'
> make: *** [Python/thread.o] Error 1
> 
>    The line in question /usr/include/pthread.h:238 is
> int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
>    Nothing suspicious.
> 
>    Any idea what can be wrong? Python/thread_pthread.h has
> 
> #ifdef __APPLE__

Try making that 

#if defined(__APPLE__) || defined(__SUN__)

(or whatever).

> #define destructor xxdestructor
> #endif

Cheers,
M.

-- 
  Lisp does badly because we refuse to lie.  When people ask us if 
  we can solve insoluble problems we say that we can't, and because 
  they expect us to lie to them, they find some other language
  where the truth is less respected.   -- Tim Bradshaw, comp.lang.lisp


From jicman at cinops.xerox.com  Fri Apr 12 14:33:44 2002
From: jicman at cinops.xerox.com (Jose Isaias Cabrera)
Date: Fri, 12 Apr 2002 14:33:44 -0400
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com> <20020412101611.GB15466@andy.tynant.ftech.net> <20020412131633.GA17830@andy.tynant.ftech.net>
Message-ID: <00d701c1e250$93d5d4d0$2ec4e70d@mountain.dssc.mc.xerox.com>

thanks Andy! This, with setting the mode from PASV to normal, worked!
Thanks for all the help.

back to python...:-)

> Sorry, I just realised that should be:
>
> ftp.storbinary('STOR ' + fname, f)
>
> (ie ftplib takes care of the read() for you).






From thoa0025 at mail.usyd.edu.au  Tue Apr  2 23:12:14 2002
From: thoa0025 at mail.usyd.edu.au (Trung Hoang)
Date: Wed, 03 Apr 2002 04:12:14 GMT
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com> <mailman.1017769807.20362.python-list@python.org>
Message-ID: <yivq8.14048$hi7.49995@news-server.bigpond.net.au>

So would GadFly be the easiest to setup out of all the databases?
cheers


"M.-A. Lemburg" <mal at lemburg.com> wrote in message
news:mailman.1017769807.20362.python-list at python.org...
> Paul Boddie wrote:
> >
> > "Ruediger Maehl" <ruediger.maehl_nospam at web.de> wrote in message
news:<newscache$d8oxtg$iv4$1 at www-neu.dzsh.de>...
> > >
> > > "Trung Hoang" <thoa0025 at mail.usyd.edu.au> schrieb im Newsbeitrag
> > > news:pRzp8.4373$hi7.15765 at news-server.bigpond.net.au...
> > > > What relational database is easy to setup and use with python?
besides
> > > > msql (im having some problems)
> > >
> > > SAPDB, a full featured database for free at www.sapdb.org
> > > with Python interface.
> >
> > I'm not sure I would second that recommendation unless SAP DB has
> > changed dramatically over the past few weeks. I found that the
> > installation was quite a bit more complicated than, for example, MySQL
> > (at least on Windows) and upon issuing the arcane commands for
> > restarting database instances, the software would complain that the
> > instances needed restarting, which was hardly very reassuring with
> > respect to either the stability or the usability of the software.
>
> Oh, SAP DB is quite usable. In fact, we switched from MySQL
> to SAP DB for most of our work. SAP DB is fast, full-featured,
> open-source and very stable. It does have it's share of bugs, but
> then so do MySQL and all the others.
>
> > As for the Python DB-API module, I couldn't get any version of it I
> > found to work with the database system. Admittedly, my objective was
> > to get a JDBC connection up and running, which was eventually
> > successful, and I therefore cut that particular side-investigation
> > short.
>
> You should probably try zxODBC which is a JDBC-based version
> of mxODBC written by Brian Zimmer. It's part of Jython since
> version 2.1.
>
> > Despite it's often-cited shortcomings, MySQL and the MySQLdb module
> > are quite easy to set up. On UNIX systems, PostgreSQL and pgPySQL are
> > possibly as easy, although one doesn't need to suffer a manual with
> > PostgreSQL which consists of a single >1.5MB HTML file. I've also had
> > success with Sybase Adaptive Server Anywhere and mxODBC (without a
> > driver manager), but that software is commercially licensed and may
> > not even be available any more.
> >
> > Another interesting issue that one might want to consider later on is
> > Unicode support with database systems and Python DB-API modules. Some
> > modules don't seem to support Unicode (pyPgSQL, for example), whereas
> > others apparently rely on the database system for such support to
> > actually function (eg. mxODBC). Some people may find this an
> > irritation earlier than others, however.
>
> mxODBC is certainly well alive. Note that Unicode support in mxODBC
> is available both as emulation (mxODBC does the conversion to e.g.
> UTF-8) and using native support (the ODBC driver has to support
> Unicode).
>
> MS SQL Server and MS Access both support native Unicode mode
> via mxODBC.
>
> --
> Marc-Andre Lemburg
> CEO eGenix.com Software GmbH
> ______________________________________________________________________
> Company & Consulting:                           http://www.egenix.com/
> Python Software:                   http://www.egenix.com/files/python/
>




From sholden at holdenweb.com  Fri Apr 19 22:04:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 19 Apr 2002 22:04:42 -0400
Subject: using get() with Entry()
References: <12f756b9.0204191747.7acb2b74@posting.google.com>
Message-ID: <334w8.7131$T%5.6943@atlpnn01.usenetserver.com>

"Eric Brake" <ebrake2002 at yahoo.com> wrote in message
news:12f756b9.0204191747.7acb2b74 at posting.google.com...
> I've looked through all of the documentation I can find. I can't make
> get() work with Entry().
>
> Here's my code:
> ###
> guess = StringVar()
> guessbox = 0
> guessbox= Entry(root, width=20, textvariable=guess).pack()
> guess.set(guessbox.get(guessbox))
> ###
>
> It keeps giving me a error message saying:
> Traceback (most recent call last):
>   File "D:\My Documents\mymodules\gui\guessthenumber.py", line 43, in
> ?
>     guess.set(guessbox.get(guessbox))
> AttributeError: 'NoneType' object has no attribute 'get'
>
> What am I not doing? I just now learning this so break it down,
> Please.
> Thank you for your help

It's a while since I used Tkinter, but isn't the point of associating a
textvariable with the Entry widget to make it unnecessary to synchronize the
contents yourself?

Therefore, I should have thought your last line should have been something
like

    mystring = guess.get()

Tkinter itself handles the updaying of the StringVar's value for you, you
just pull the value out. You can also use its set() method to put a value in
to the TextEntry. A Google search for "python tkinter stringvar" yields
David Mertz' useful page

    http://www-106.ibm.com/developerworks/library/l-tkprg/

Find "stringvar" in that and you're almost home.

I agree it's annoying that Fredrik Lundh didn't complete the public version
of his Tkinter reference manual. I'd complain, except that he's done far
more for the Python system than I'm ever likely to...

regards
 Steve
--
home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From cliechti at gmx.net  Sat Apr  6 13:35:24 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 19:35:24 +0100
Subject: BaseHTTPServer
References: <a8ljkn$6ht$1@lacerta.tiscalinet.it>
Message-ID: <Xns91E8C7772E577cliechtigmxnet@62.2.16.82>

"FrancescO" <francesco.contini at tiscalinet.it> wrote in
news:a8ljkn$6ht$1 at lacerta.tiscalinet.it: 
> but... if i try to stop and restart the server i have an exception:
> port already in use
...
> after a minutes, or more,  the port is free... and i can restart the
> program, but, obviously,
...
> I don't run a particular code to stop.. (kill -9 ;P ) and i like to

so don't use SIGKILL! when you kill a proces it is terminated by the OS. it 
keeps sockets closed for some time so that hanging connections time out and 
don't give problems with a newly started server on the same port.

if you stop the program with an other signal like SIGTERM python has the 
chance to clean up and the socket is freed, so that you can immediately use 
the port again.

> use the same way used
> by the /etc/init.d/... scripts to start and stop.

i have a daemon wrapper script and init.d script for a Debian/Linux system 
if you're interested.

chris

-- 
Chris <cliechti at gmx.net>



From gcordova at hebmex.com  Mon Apr 22 12:47:31 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 22 Apr 2002 11:47:31 -0500
Subject: constructors in python
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062338E656E@mail.hebmex.com>

> > 
> > >
> > >C++ doesn't permit not to call the base c-or.
> > > 
> > 
> > And thanks to all the Gods and Goddesses for that.
> 
> But it does allow you to call any of the Base's ctors, which is an
> excellent way to introduce bugs <wink>
> 
> -- bjorn
> 

Actually, I was thanking the Gods and Goddesses for Python
not being more like C++.  :-)

polytheisticly-yours,

-gustavo




From aleax at aleax.it  Fri Apr 26 18:26:39 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 22:26:39 GMT
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org> <aa48nq$a61$1@panix1.panix.com> <mailman.1019640045.27779.python-list@python.org> <aachl4$ltf$1@panix2.panix.com>
Message-ID: <zuky8.66510$vF6.2037973@news2.tin.it>

Aahz wrote:
        ...
> What do you need Borg for?  Alex himself has agreed that using a module
> as a singleton is the correct answer 80-90% of the time.

More than that: it's 80%-90% of the time *in which people DO think about
Singleton*, plus a few times in which that doesn't even come to mind.  Add
in a class instance masquerading as a module, as for example in
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65207 , and
it must be over 90%.

But if you DO need to subclass the One And Only -- most of the remaining
8-9% of the time is covered by that -- then Borg is good.


Alex



From phd at phd.pp.ru  Thu Apr 11 04:30:05 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 11 Apr 2002 12:30:05 +0400
Subject: PDCurses
In-Reply-To: <a92e9v$bun$1@pegasus.tiscalinet.it>; from f.onorati@tiscalinet.it on Thu, Apr 11, 2002 at 12:29:42AM +0200
References: <a92e9v$bun$1@pegasus.tiscalinet.it>
Message-ID: <20020411123005.C30351@phd.pp.ru>

On Thu, Apr 11, 2002 at 12:29:42AM +0200, Abacuc wrote:
> How can i use PDCurses onWin32 with Python?

   Once there was a wrapper at
http://www.mikroplan.com.pl/~mak/python/#Win32, but Grzegorz Makarewicz
retired and removed the site. If he still reads the newsgroup/mailist he
may republish the thing somewhere.

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




From phr-n2002a at nightsong.com  Tue Apr 16 16:58:01 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 16 Apr 2002 13:58:01 -0700
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <7xhemb744m.fsf@ruckus.brouhaha.com>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
> Is it possible to have a safe eval() on the server-side even when the string
> consists of submitted user data?

Well, look at the docs for the rexec (restricted execution) and Bastion
classes, but basically no, there's no safe way.  Rexec doesn't stop an
infinite loop from hosing your cpu.

Why do you want to eval strings from untrusted users?



From twisted at itamarst.org  Tue Apr  2 14:45:51 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Tue, 02 Apr 2002 21:45:51 +0200
Subject: ANN: Twisted 0.16.0
References: <Pine.LNX.4.44.0204021043340.31094-100000@mercury.sabren.com>
Message-ID: <3CAA0A6F.1040600@itamarst.org>

Michal Wallace wrote:

 > Okay, I see this was posted yesterday (April Fools day), but
 > the site's still up today.
 >
 > Is this for real?

bin/twistd can load license files, yes. If you don't have one it prints
a nice long error message (which I removed in CVS, so it won't be in
0.16.1).

Glyph has some insane reason he wants it to be in (rumor has it a rogue
Pavlovian has linked the registration form to a Certain Person's cage,
giving him a food pellet every time someone registers. Try not to
overfeed him.)

 > If so, why? How can piracy be a problem for an open source
 > project?  How does this licensing scheme prevent it?

Well, let's quote one the BSA's latest press release:

"""
Washington, DC (March 21) ?In its ongoing effort to protect consumers
from Internet software pirates and support a safe, secure and legal
online community, the Business Software Alliance (BSA) announced that
last month it executed three unannounced court-ordered seizures on
individuals whom BSA investigated for selling unlicensed software out of
their homes in Kentucky, Missouri and Pennsylvania. Under the
supervision of the U.S. Marshals, BSA carried away computers, equipment
and CDs used in the mail-order operations.

...

The following are helpful tips on safe shopping for software online:

   * Trust Your Instincts -- Whether it is being sold as new or used,
     if a price for software seems too good to be true, it probably is.

...
"""

Now, you wouldn't want any Sikrit PSU Agents showing up at your house,
asking to see *your* Twisted license, and not being able to show one,
now would you? And do you really trust a bunch of people giving away
software for FREE?





From martin at v.loewis.de  Wed Apr 17 15:27:22 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 21:27:22 +0200
Subject: Making a C extension compatible with 2.2
References: <mailman.1019067094.1410.python-list@python.org>
Message-ID: <m3ads2yvl1.fsf@mira.informatik.hu-berlin.de>

Hamish Lawson <hbl at st-andrews.ac.uk> writes:

> Aha! It so happens it's also ingmod that my problem relates to! :-)  I
> recently notified Holger Meyer about it and he said he would try to
> address it soon, but I thought I would see what I could do myself in the
> meantime (though I'm far from expert on writing C extensions for
> Python).

That module appears to lack checks for exceptions in a number of
places. For example:

#define	STR(obj)			(PyString_AsString(PyObject_Str(obj)))

This can easily fail: If PyObject_Str raises an exception, it will
return NULL. That would crash the interpreter, though.

Also, when allocating things, it often does not check for
errors. E.g. in sqlda_description:

	list = PyList_New(0);
	for (i = 0, v = sqlda->sqlvar; i < sqlda->sqld; ++i, ++v) {
...
		tuple = PyTuple_New(7);

The code should check both list and tuple for NULL.

			prec = PyInt_FromLong((long)IISQL_PRCSN(v->sqllen));
			scale = PyInt_FromLong((long)IISQL_SCALE(v->sqllen));

Here, prec and scale may be NULL.

In turn, callers of sqlda_description don't check for exceptions:

	self->descr = sqlda_description(self->out);

or

		if (self->out) {
			if (!self->descr) {
				self->descr = sqlda_description(self->out);
			}
			if (self->descr) {
				Py_INCREF(self->descr);
				return(self->descr);
			}
		}
		Py_INCREF(Py_None);
		return(Py_None);

If sqlda_description returns NULL to report an exception, this
fragment will return None, instead of reporting the exception further
out.

It will be hard to find all these problems, but I think Just was right
that the likely cause is an unreported exception, which will show up
the next time somebody calls PyErr_Occurred.

Regards,
Martin



From mcfletch at rogers.com  Sat Apr 13 14:39:14 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sat, 13 Apr 2002 14:39:14 -0400
Subject: __iadd__ and fellows missing (Python 2.2)
References: <LNBBLJKPBEHFEDALKOLCOEOBONAA.tim.one@comcast.net>
Message-ID: <3CB87B52.9010702@rogers.com>

I was unaware that there was no guarantee for 0-length tuples always 
being unique objects.  There are lots of places in my code where I use 
0-length tuples in this pattern:

NULL = ()

class A:
	def b( self, object=NULL ):
		if object is NULL:
			#was not specified
		elif object is None:
			#specified to be object None

Of course, I can use:

NULL = []

But I'd always avoided that because of bad associations with having 
mutable objects as default parameters (yes, it doesn't actually matter 
in this pattern, but it's a thirty-times-burned-thirty-first-shy thing). 
  Suppose I'll wind up using singleton instances instead.

Sigh.  Will need to poke through every package I own and update them 
someday.

Have fun,
Mike

Tim Peters wrote:
...
> The implementation can choose, or choose not, to share memory blobs for any
> value of an immutable type, and exactly when & which objects get shared can
> (and does) vary across releases.  The only objects guaranteed to be
> singletons (meaning you'll always get exactly the same object no matter how
> its value is computed) are None, type objects (things the type() function
> returns), and in 2.3 True and False.  So, in particular, "is" should never
> be used as a test for integer equality, or equality with an empty tuple or
> string.
...





From apeak at softhome.net  Wed Apr 17 08:34:15 2002
From: apeak at softhome.net (apeak at softhome.net)
Date: Wed, 17 Apr 2002 12:34:15 GMT
Subject: financial software
Message-ID: <20020417123415.10128.qmail@softhome.net>

I'm considering writing a financial management program, but before I
reinvent another wheel, I wonder if there's a good one out there, something
easy for a newbie to understand.  If there are no such projects, then is
there other real-world Python code I could study that would help me roll my
own?  Currently, I'm a Python beginner.




From jepler at unpythonic.net  Sun Apr 28 09:50:13 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 28 Apr 2002 08:50:13 -0500
Subject: Update locals()
In-Reply-To: <20020428115221.GF16217@devel.trillke>
References: <20020427195725.GA16217@devel.trillke> <ySFy8.28416$8D3.759092@news1.tin.it> <20020427233027.GE16217@devel.trillke> <0204280830290C.16995@arthur> <20020428115221.GF16217@devel.trillke>
Message-ID: <20020428085009.A983@unpythonic.net>

On Sun, Apr 28, 2002 at 01:52:21PM +0200, holger krekel wrote:
> Thanks for the explanations. 
> 
> There is no question that C++-like templates often are
> not needed in python. I regret mentioning C++ because
> it is only a very distant analogy as you pointed
> out precisely. 
> 
> > Please give ONE example which, in your opinion, is best handled
> > with 'exec' without explicit dictionaries rather than in any other way.
> 
> i try to make it short and answer questions (if any of you gurus
> ever have any :-) as needed. Obviously you can always pass explicit 
> dictionaries to exec (globals, locals). That can't be the point. 
> 
> I cite some of my code.
> 
> class filename:
>     """ Abstraction for filename ... """
>     # ...
> 
>     for name in filters._stateful_all:
>         doc=filters.__dict__.get(name,undoc).__doc__.strip()+'\n'
>         exec indentedcode("""
> 
>     def %(name)s (self,*args,**kargs):
>         '''%(doc)s'''
>         return filters.%(name)s (*args,**kargs) (self)
>     """ % locals()) 
> 
>     # ...

I don't quite see what you're doing here, but you might try something like
(given nested scopes) (untested):

    class filename:
	def __getattr__(self, attr):
	    if not hasattr(filters, attr): raise AttributeError, attr
	    filter_func = getattr(filters, attr)
	    ret = lambda *args, **kargs: filter_func(*args, **kargs)(self)
	    ret.__doc__ = getattr(filter_func, __doc__, undoc)
	    #setattr(self, attr, ret)
	    return ret

My version will also work properly if "filters" is extended at runtime with
new attributes... It's reasonbly efficient to construct a function at
runtime with 'lambda', and you can trade time for space by caching the
attribute on the instance once it's been computed once

Of course, given that you always immediately call the instance of the
filters. function, I don't see why you can't simply write

    class filename:
	def access(self, mode='r'):
	    flags=0
	    if 'r' in mode: flags|=os.R_OK
	    if 'w' in mode: flags|=os.W_OK
	    if 'x' in mode: flags|=os.X_OK
	    return access(str(self), self.flags) and self

There's probably some reason which isn't as apparent in this example as
some of the more complex ones you've written, though...

Jeff




From JamesL at Lugoj.com  Sun Apr  7 16:38:07 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 7 Apr 2002 20:38:07 GMT
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <Xns91E98AAF1B9ABJamesLLugojcom@216.148.53.85>

Alex Martelli <aleax at aleax.it> wrote:
> Why don't we start thinking of a similarly dual-tracked Python -- one
> track aiming at stability (in a "middle ground" sensible sense: strong
> commitment to not breaking old code -- introducing small new features
> and new library modules _is_ OK, else one can use a bugfix-only-"track"
> for some previous release), one aiming at frequent releases,
> innovation, etc.

This proposal seems to have good promise and if well executed would 
probably mollify a curmudgeon such as myself.


From neal at metaslash.com  Wed Apr  3 21:05:21 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Wed, 03 Apr 2002 21:05:21 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <mailman.1017877754.545.python-list@python.org>
Message-ID: <3CABB4E1.E5E0A79E@metaslash.com>

"Delaney, Timothy" wrote:
> 
> test.py:6: Comparisons with [TRUE|True|true] may give incorrect results
>                    ^ extra i ;)
> test.py:7: Comparisons with [FALSE|False|false] may give incorrect results
>                    ^ extra i ;)

Already caught and fixed. :-)

> and would also add a warning for code in 2.3+ for code like
> 
> TRUE = <something>
> FALSE = <something>
> true = <something>
> false = <something>
> 
> test.py:2: TRUE appears to be a boolean constant - builtin True should be
> used.
> test.py:3: FALSE appears to be a boolean constant - builtin False should be
> used.
> test.py:4: true appears to be a boolean constant - builtin True should be
> used.
> test.py:5: false appears to be a boolean constant - builtin False should be
> used.

This is already done, but not in the example.  pychecker will even
bitch about tRuE, since the code always normalizes, then checks.
It was easier than writing all the cases, and I'm lazy.

Should not assign to true, it is similar to builtin True
Same for comparisons (or comparisions :-).

The error messages probably should be improved.
The check should be available in any version, since part 
of pychecker's purpose is to help with upgrades,
finding problems before upgrading.

After everything settles out with booleans, it would be good
to revisit this and other issues as they arise.
(So remind me later or submit a bug on SF.)

Neal


From phil at mrinfo.de  Tue Apr 23 09:42:14 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Tue, 23 Apr 2002 15:42:14 +0200
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com> <Uscx8.93527$T%5.72611@atlpnn01.usenetserver.com>
Message-ID: <aa3oa1$1no$1@swifty.westend.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:Uscx8.93527$T%5.72611 at atlpnn01.usenetserver.com...
> "Philipp Lenssen" <phil at mrinfo.de> wrote in message
>..
> > For XHTML string output, I prefer single quotes because I don't have to
> > escape double-quotes within the string, e.g.:
> > s += '<img src="..." alt="" />'
> >
>
> For XHTML and HTML output you would find the triple-quoted forms much
> easier. Then you can use EITHER quotes in your output, and include
newlines
> as well, as in:
>
> s = """%s
> <img src="..." alt="A comment">""" % s
>

Yes, for long lines without any variables I do that.
What if you have to do an if-check every second line and dynamic attribute
values?
As you can see it depends on the needs, and the s += '<img .../>' is often
easier.
The more database/ XML driven the XHTML output is in the first place, the
less you will find yourself outputting static portions of XHTML in the first
place.




From rpm1deletethis at frontiernet.net  Mon Apr 29 23:23:03 2002
From: rpm1deletethis at frontiernet.net (Patrick)
Date: Mon, 29 Apr 2002 23:23:03 -0400
Subject: Access violation trying to run Python over Win NT network.
References: <ucrs88tkj2ttc7@corp.supernews.com> <mailman.1020131453.4624.python-list@python.org>
Message-ID: <ucs3ir64u584e@corp.supernews.com>

Cliff Wells <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1020131453.4624.python-list at python.org...
> On Mon, 29 Apr 2002 21:17:55 -0400
> Patrick wrote:
>
> > I'm trying to run Python from a network drive.  On most machines it
works,
> > but on some machines it generates an access violation about 60% of the
time,
> > (just by typing 'python' with no arguments).  If I install Python
locally it
> > runs fine on all the machines.
>
> Even on the machines where you get the access violations it works about
40% of
> the time?

Yes.

In fact, I can run python over the network fine on my machine, then take my
harddrive,
 (it's removable), to a different machine, and it won't run over the
network, (most of
the time).

>
> > I would really appreciate any help.  I've been trying to get the people
at
> > work to use Python and they have finally agreed to give it a try, but
they
> > don't want to install it on everyones machines, so we have to run from a
> > network drive.
>
> I've done this before without too many problems.  The usual problem was
the
> generation of .pyc files.  If the user doesn't have write permission on
the
> file/directory then it fails.  It seems odd that you would get access
> violations simply by typing "python", but perhaps when python runs it
tries to
> byte-compile some modules and fails (very unsure of this, although I'm
guessing
> there are platform modules that might need to be compiled at startup).
You
> might try compiling all of the modules yourself and see if that fixes it.
>

I'll check this, but the fact that I can log on some machines as myself and
it always
runs, and on other machines it rarely runs.

One other thing I forgot to mention, it seems to always work on Windows 2000
drives.

Thanks,
Patrick







From emile at fenx.com  Thu Apr 18 12:43:36 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 18 Apr 2002 16:43:36 GMT
Subject: help splitting strings
References: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com>
Message-ID: <YICv8.34916$HH5.3504@rwcrnsc51.ops.asp.att.net>

"Tony Case" 
> I'm pretty new to Python which probably explains my problems,
> I'm trying to read data from a number of comma separated variable
> files.  These are fairly large, 5000 lines plus.
> I'm reading the files and using split to separate the
> individual values, some numeric some string.  The problem is that a
> number of the lines have commas and double quotes embeded in the

Sounds like you need the CSV module.  Take a look at

http://www.object-craft.com.au/projects/csv/

--

Emile van Sebille
emile at fenx.com

---------



From guido at python.org  Mon Apr 15 21:36:35 2002
From: guido at python.org (Guido van Rossum)
Date: Tue, 16 Apr 2002 01:36:35 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <3CBB80CC.9C57B055@python.org>

phil hunt wrote:

> I wish to make the pas startement optional. That is, in any Python
> program where there is a pass it can be left out and the meaning of
> the program is unchanged.

I expect I may be too late here, but I'd like to point out that I
will never accept such a proposal, so you can stop wasting your time.
I've seen your arguments, and they don't convince me.  Sorry. :-)

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


From aahz at pythoncraft.com  Mon Apr 29 15:33:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Apr 2002 15:33:09 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net>
Message-ID: <aak75l$1ff$1@panix1.panix.com>

In article <aak1uq$bgt$1 at bob.news.rcn.net>,
Raymond Hettinger <python at rcn.com> wrote:
>
>From the comments, I've learned:
>
> [...]

9.  You need to do better formatting.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From bokr at oz.net  Sun Apr  7 03:14:39 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Apr 2002 07:14:39 GMT
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com>
Message-ID: <a8orkv$2cj$0@216.39.172.122>

On 6 Apr 2002 17:07:08 -0500, aahz at pythoncraft.com (Aahz) wrote:

>I've been wrestling with the way namespaces really work, and I've been
>discussing it a bit on python-dev.  I'm moving the discussion here to
>get some broader viewpoints.  Here's an incomplete draft of what I've
>got so far:
>
[...]

Thanks for a summary of notes on names, binding, etc., but so that
I/we can respond more appropriately, what is your goal?
I.e., to document things as they are, to explore the fringes
in terms of what might be, to compare to other languages, or what?

Regards,
Bengt Richter


From tim.one at comcast.net  Tue Apr 30 23:24:45 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 30 Apr 2002 23:24:45 -0400
Subject: Fatal Python error: GC object already in linked list
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F5659@MAYOR.retzlaff.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEFAPCAA.tim.one@comcast.net>

[Jimmy Retzlaff]
> I'm working on an application in Python and my application occasionally
> dies with the error:
>
> Fatal Python error: GC object already in linked list
>
> I'm quite unfamiliar with Python's implementation, but I tried studying
> a little in this case. That error message only appears in a macro in
> objimpl.h (_PyObject_GC_TRACK), and that macro only seems to be used in
> a function that creates traceback objects in traceback.c
> (newtracebackobject).

You need to get a better string-searching program <wink>.  For example, you
missed this key use in gcmodule.c:

void
PyObject_GC_Track(void *op)
{
	_PyObject_GC_TRACK(op);
}

If you possibly can, build a debug-mode Python from source, and run under a
debugger.  The message is produced by a call to Py_FatalError(), and on
Windows that in turn calls the Win32 DebugBreak() function.

> ...
> Looking at CVS it appears the use of that macro was added to prevent
> the old memory leak that could happen when calling sys.exc_info().

All objects of all container types that can appear in reference cycles
eventually invoke _PyObject_GC_TRACK.  This is what lets cycle garbage
collection know that such objects need to be watched.  You get the error
message if you try to tell gc to pay attention to an object it's already
paying attention to.  IOW, somebody is calling _PyObject_GC_TRACK on a given
object twice.  In a debugger you'll see exactly who's doing it, and on which
object.

> My program is a non-trivial (~15K lines of Python) GUI app on Windows
> based on wxPython. I've been developing this app for 6-8 months and the
> error just started showing up about 3 weeks ago. I haven't updated any
> major components (other than my own code) for a few months, so I'm
> guessing my own code changes have triggered this.

If you don't have any C code of your own in this mix, you're probably
tickling a memory-management bug in wxPython (or some other C extension).
Such problems are unlikely to be in the core Python code, but that's not
impossible either.

> Occurrences seem quite random, which can indicate timing issues of
> course.

Try

import gc
gc.disable()

at the start.  If the problem goes away then, it doesn't prove anything, but
at least it would be amusing <wink>.

> Windows XP Pro (I've also seen the problem on Windows 2000)
> Python 2.2 (the standard Windows binary distribution)

Move to Python 2.2.1 ASAP, please.  That did fix one excruciatingly rare and
subtle bug in the interaction between cyclic gc and a delicate gimmick that
lets you deallocate multi-hundred deep nested containers without blowing
your stack space.

> Win32all Build 146
> wxPython 2.3.2.1 Hybrid (also seen it with a Normal build)
> Numeric 20.3

Ouch.





From jeff at ccvcorp.com  Thu Apr  4 14:52:57 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 4 Apr 2002 11:52:57 -0800
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de> <3CAC1665.8040209@mxm.dk>
Message-ID: <MPG.17164ee731d2609f989698@news.nwlink.com>

In article <3CAC1665.8040209 at mxm.dk>, maxm at mxm.dk says...
> Thomas Guettler wrote:
> 
> 
> > Is there already a function in the standard python modules
> > which returns the a relative url given two absolute urls?
> > 
> > Example:
> > 
> > relative_url("http://foo/a/b/c", "http://foo/a/d")
> >  --> return: "../d"
> 
> I hope you mean that the result should be "../../d" ??

Actually, that's not quite correct (and neither are your code 
examples below, for the same reason).

In the URL 'http://foo/a/b/c', we could presume that 'foo' is a 
domain name, 'a' and 'b' are directories, and 'c' is a filename.  
Thus, the 'current directory' ('.') for this URL is //foo/a/b, 
and the parent directory ('..') is //foo/a.  Therefore, in this 
case, the O.P.'s initial relative URL ('../d') is correct.

The issue is complicated somewhat in that '//foo/a/b/c' can 
*also* be shorthand for '//foo/a/b/c/index.html' (or whatever the 
server's default-filename is).  If (and *only* if) the original 
URL is representing a default document, then your result would be 
the correct one.
 
There doesn't seem to be a way to determine whether a given URL 
represents a default document or not, however, without hitting 
the server in question to find out.  (I certainly wouldn't want 
to rely on the type of, or even existence of, a file extension 
for this purpose.  No rule says that HTML documents *must* use a 
.htm/.html extension...)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From kseehof at neuralintegrator.com  Fri Apr 12 10:25:23 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Fri, 12 Apr 2002 07:25:23 -0700
Subject: post event from C-extension
In-Reply-To: <3cb6f29f.23943234@news.easynews.net>
Message-ID: <EFEFIHAINIINNELLHFAIKEOICAAA.kseehof@neuralintegrator.com>

I'm curious to see if you get a solution.  Try the wxPython list too.

How about if you just use PyObject_Call to call a python function that
then calls wxPostEvent on the python side.

Another approach would be to pass a wxPython window object in an entry
point in your extension.  Perhaps if you grab the right header file
from the wxPython sources, you could access the C++ pointer to the
actual C++ window.  From there you would call the wxWindows C++ version
of the wxPostEvent method.  I haven't tried this myself, but I'm
curious if it would work.

> Hi,
> 
> my program has a couple of time consuming functions. With the help
> of the wxPython Cookbook I use Threads to keep my GUI alive. When
> finished, each Thread uses wxPostEvent to notify the application.
> 
> One function is a C-extension and I have no idea how to make this
> thing sending an event. Any hint's, examples, links ?
> 
> Thanks
> Marcus
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From erno-news at erno.iki.fi  Mon Apr  8 01:50:10 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 08 Apr 2002 08:50:10 +0300
Subject: purging file buffers - to make new data be written to file
References: <a8qhjp$7go$1@news6.svr.pol.co.uk>
Message-ID: <ku6632oi25.fsf@lasipalatsi.fi>

In article <a8qhjp$7go$1 at news6.svr.pol.co.uk>, ddoc
<news at 92tr.freeserve.co.uk> writes:

| Can I do that from within a python script, or do I have to give a command 
| line argument to make Python start without buffers?

you can open files with buffering disabled by passing 0 as the
3rd argument to open().

f.flush() will commit any data that python and the c library might
have not yet written to the operating system. this will let
other processes in the system see the new data.

the above two things just make sure the os knows about the data.  if
you want to make sure the data ends up on disk physically, use
os.fsync(f.fileno()).

  -- erno


From gerhard at bigfoot.de  Mon Apr 29 14:17:56 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 29 Apr 2002 18:17:56 GMT
Subject: python2.2
References: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net>
Message-ID: <slrnacr3io.nil.gerhard@lilith.my-fqdn.de>

Billy Ng wrote in comp.lang.python:
> I am using debian linux.

Which distribution? stable (potato), testing (woody) or sid?

> I am trying to install the python2.2 that is an unstable package
> onto the system, but it says "Couldn't find package python2.2"

If you're using testing or sid,

    apt-get update
    apt-get install python2.2

should install it.

> Wouls anyone know how to install the unstable package via apt-get

You don't need the unstable package. python2.2 is already in testing
(woody).

If this doesn't work for you, most probably you haven't set up your
/etc/apt/sources for packages from testing or sid, yet.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From idbaxter at semdesigns.com  Wed Apr 24 10:38:12 2002
From: idbaxter at semdesigns.com (Ira D. Baxter)
Date: Wed, 24 Apr 2002 09:38:12 -0500
Subject: Python call tree tool?
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk><lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk><7dD$gUA$uZx8EwWr@jessikat.demon.co.uk> <mailman.1019588025.27016.python-list@python.org>
Message-ID: <3cc6c10d$1@giga.realtime.net>

"Fran?ois Pinard" <pinard at iro.umontreal.ca> wrote in message
news:mailman.1019588025.27016.python-list at python.org...
> [Robin Becker]
>
> > I guess I can use inspect to figure out the functions/classes that are
> > declared in an arbitrary module, but it seems harder to determine if
> > these things are actually referenced in other modules. For that I assume
> > I will have to do some code analysis.
>
> It may be difficult to do in general.
> Such usages do not appear uncommon at all to me, and it surely requires an
> extremely clever and sophisticated code analyser to find out what really
> happens at run-time.

Most analyses of real programs are hard to do exactly,
even if you take into account inter-procedurel (inter object)
interactions.   In any case, that doesn't stop people from
implementing pretty reasonable analyzers.   But you are
right, you need strong foundations to do it.

I can't offer an off-the-shelf call tree analyzer.
But our DMS Software Reengineering Toolkit would
be good for implementing one.  In a similar vein,
we have done "Deactive" code analysis ("defined but not
referenced" code as defined by FAA DO187B
flight software requirements) for large Java systems.
See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.

--
Ira D. Baxter, Ph.D. CTO Semantic Designs, Inc.
http://www.semdesigns.com




From slackness at caramail.com  Sat Apr 13 18:38:50 2002
From: slackness at caramail.com (slack)
Date: 13 Apr 2002 15:38:50 -0700
Subject: Python newbie
Message-ID: <ab96a890.0204131438.5cce8f66@posting.google.com>

Hello,
I'm interested in Python programming language since one around one
year. I think it is an interesting language because it's very simple
to use this language to program windows... We don't have to think
about event handling instead of the purpose of the program itself...
well... I don't want to explain why I find it interesting... I've
installed a version of Python on Windows computers but I find it very
slow compared to Linux version of Python. I've noticed that the first
time I used Python on Windows. Well. This is not really the problem. I
want to use Python on Windows but I don't know if I can convert .py
files into .exe files? Is it possible thanks to some utility?

thank you.

slack


From martin at v.loewis.de  Sat Apr 13 18:34:50 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 00:34:50 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl> <m3662wypqt.fsf@mira.informatik.hu-berlin.de> <3cb7f8a7$0$85437$e4fe514c@dreader3.news.xs4all.nl> <m37knb9axo.fsf@mira.informatik.hu-berlin.de> <3cb80a78$0$43852$e4fe514c@dreader1.news.xs4all.nl>
Message-ID: <m3wuvb1b3p.fsf@mira.informatik.hu-berlin.de>

Boudewijn Rempt <boud at valdyas.org> writes:

> Martin v. Loewis wrote:
> > No, not in any released version. Meanwhile, I recommend to stay away
> > from Unicode literals if you need "funny" characters, and write
> > 
> >   unicode("<string containing funny characters>", "utf-8")
> > 
> > instead.
> 
> Yes. Although I might be going to try the relevant patch.

Please do, by any means. If you have any comments, please put them
into the SF issue. Notice that you can chose between a phase 1 and a
phase 2 implementation already.

Regards,
Martin



From bokr at oz.net  Wed Apr 24 22:00:09 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Apr 2002 02:00:09 GMT
Subject: An IMHO interesting optimization question...
References: <3cc73c81@nntp.server.uni-frankfurt.de>
Message-ID: <aa7nv9$aju$0@216.39.172.122>

On 25 Apr 2002 01:15:14 +0200, Michael 'Mickey' Lauer <mickey at tm.informatik.uni-frankfurt.de> wrote:

>Hello,
>
>I have an optimiziation question. Given the following code
>which generates a word length statistic up to words of the length
><longWord>.
>
>=-------------------------------------------------=
>#!/usr/bin/python
>
>def count( word ):
>   "Increments the wordcounter"
>   length = len( word )
>   if length > longWord:
>       length = longWord
>   stats[length] += 1
>
>def printout():
>    "Formatted output of dictionary"
>    for i in xrange( 1, longWord ):
>        print "Length %d = #%d" % ( i, stats[i] )
>    print "Length >%d = #%d" % ( longWord, stats[longWord] )
>
># import libraries and check for help request
>
>import sys, re
>if len( sys.argv ) != 1:
>    print """
>wordStatistic.py 1.0
>--------------------
>Reads from stdin and prints a word statistic.
>Usage: cat <file> | wordStatistic.py
>"""
>    sys.exit( 0 )
>
>longWord = 15
>defWord = "[A-Za-z]+"
>
>stats = {}
>wordExp = re.compile( defWord )
>
># initialize dictionary, so we don't have to check for key-existance later
>for i in xrange( 1, longWord+1 ):
>    stats[i] = 0
>
># main loop. iterates over standard input as long as there's
># something coming in...
>while 1:
>    line = sys.stdin.readline()
>    if line:
>        for word in wordExp.findall( line ):
>            count( word )
>    else:
>        break
>=-------------------------------------------------=
>
>For a given document, this program takes 20 seconds on my machine.
>I now try to optimize it by eliminating the while loop. The 2nd version looks like:
>
>=-------------------------------------------------=
># main loop. iterates over standard input as long as there's
># something coming in...
>lines = sys.stdin.readlines()
>for line in lines:
>    for word in wordExp.findall( line ):
>        count( word )
>=-------------------------------------------------=
>
>This takes 14 seconds on my machine which seems to be quite an achievement.
>
>Now on to the next. I try to eliminate the for loop by substituting with a map statement,
>which is supposed to be faster (because it's looping in C). The 3rd version looks like:
>
>=-------------------------------------------------=
># main loop. iterates over standard input as long as there's
># something coming in...
>lines = sys.stdin.readlines()
>for line in lines:
>    map( count, wordExp.findall( line ) )
>=-------------------------------------------------=
>
>Now this takes 17 seconds on my machine. Wow? Why that? Why does it take longer?
>
>On to the final version, which transforms the outer loop into a map statement:
>
>=-------------------------------------------------=
># main loop. iterates over standard input as long as there's
># something coming in...
>map( lambda line: map( count, wordExp.findall( line ) ), sys.stdin.readlines() )
>=-------------------------------------------------=
>
>This is even worse - it takes 18 seconds on my computer. What am I missing? And do you
>know a version which would be even faster than the 2nd version?
>
For starters, I would put the loop inside a function, and use a local list and index it with
the integer lengths instead of using a globally referenced dictionary. Why hash values that
can only be 1..15? And I would do the count logic right in the loop, to avoid the function
call overhead. When the loop is done, return the list of counts.

I have to leave, but I'm sure there's more speedup possible. How big is your file?

Regards,
Bengt Richter


From pyth at devel.trillke.net  Thu Apr 18 17:46:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 18 Apr 2002 23:46:15 +0200
Subject: PEP for an unrestricted __init__ ?!
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com>
Message-ID: <20020418214615.GB28515@devel.trillke>

On Thu, Apr 18, 2002 at 01:59:19PM -0700, Jimmy Retzlaff wrote:
> > holger krekel [mailto:pyth at devel.trillke.net] writes:
> > I'd would like to know if anybody thinks it's
> > worthwile to start a formal PEP
> ...
> > to bring full factory-pattern power to the __init__
> > constructor of a class.
> 
> I believe that's exactly what __new__ is for in new-style classes. See
> PEP 285 for an example of __new__ usage for exactly this purpose (in the
> Specification section).

good pointer. It seems like PEP 253 is the original PEP 
for the use described in PEP 285. Guido's time machine :-)

But __new__ or __init__ still restrict you to return instances
of their respective class? This does sound rather sane,
but i imagined some other uses for an unrestricted __init__ (or __new__)
such as making functions-objects a special case of classes:

   def func(args):
      "compute some stuff
      return result
and
   class func:
       def __init__(self, args):
          "compute some stuff"
          return result
	
start to look very similar. This opens up some interesting uses,
like letting the functions "constructor" decide if it has enough arguments and
returning a function instance (with some prebound arguments)
otherwise. Currently functions are always singletons...

but i might be terribly on the wrong path ...

	holger




From shalehperry at attbi.com  Thu Apr 11 11:40:28 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 11 Apr 2002 08:40:28 -0700 (PDT)
Subject: Tuple comprehension
In-Reply-To: <Pine.LNX.4.33.0204110859270.25928-100000@ns0.tbye.com>
Message-ID: <XFMail.20020411084028.shalehperry@attbi.com>

> 
> I guess the reason list and dictionary comprehensions make sense to me and
> tuple comprehensions do not is because for lists and dictionaries they are
> direct replacements for the initialize-to-empty-then-loop-to-populate
> task, for which there is no tuple equivalent:
> 
> d = empty list or dict
> for item in someInputSet:
>     do something to item and add it to d
> 

since tuples are immutable there is no
initialize-to-empty-then-loop-to-populate idiom, you are correct.  Because of
this when a tuple would work we have to use lists.  Numerous times while coding
I have started using a tuple and had to fall back to lists because I eventually
needed a way to give all of the values in my list a value based on a function.

I personally like tuples because they say "this list is immutable, it won't
grow or shrink and try not to touch its contents".  Just like using 'const' and
'private' in other languages.

If there was a way to create a tuple whose contents were programmatically
created and avoid the list -> tuple conversion I suspect tuples would get more
use.




From siegfried.gonzi at kfunigraz.ac.at  Thu Apr 25 08:09:27 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Thu, 25 Apr 2002 14:09:27 +0200
Subject: Freeing memory
Message-ID: <3CC7F1F7.8B987146@kfunigraz.ac.at>

The topic has emerged quite often in the past but never completely
answered.

I have an application which consumes at every run 60MB RAM. Every new
start of the script ('Run scipt') augments the memory up to old_memory +
60MB. After a few runs I am at 400MB.

I always just thought Python has a garbage collector and will dispose
the memory after the last reference (for me this means: end of my
script).

In Clean or Common Lisp or R-language or Yorick I never had any dire
need to
consult the Windows Task-Manager. The explanation that Python really
gets rid of the memory but the Windows OS does not show it sounds a
little bit strange to me. 

I am now a little bit more accustomed to Python, but I cannot imagine to
use it further for larger projects. My machine has got 256MB RAM and I
just use Windows XP (80MB RAM) and Python2.2; there is absolutely no
excuse
for Python not to find any free successive memory block.

There is the chance that  memory becomes free; but I am not willing to
wait about 10 minutes (after a 10 minutes calculation) in order to see a
possible disposing of memory. I want to start the second simulation
right after the first, but after a few simulations I end up at 400MB and
Python chrashes.


Any suggestions? 

S. Gonzi


From jeff at ccvcorp.com  Thu Apr  4 15:10:34 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 4 Apr 2002 12:10:34 -0800
Subject: No output from threads
References: <a8hq64$114s$1@news.rchland.ibm.com> <3ouoaukmu32afgatil65cmcb7pvb8e5pcb@4ax.com> <a8i0pv$bu6$1@news.rchland.ibm.com>
Message-ID: <MPG.171653199552480c989699@news.nwlink.com>

In article <a8i0pv$bu6$1 at news.rchland.ibm.com>, ldw at us.ibm.com 
says...
> I made the command look like: "start /belownormal /wait cmd /c
> myprog.exe -opt 3 >out3.txt".  No joy.
> Then I tried: "cmd /c start /belownormal /wait myprog.exe -opt 3 >out3.txt".
> Also no joy.
> 
> Finally I went for simplicity and tried just: "myprog.exe -opt 3 >out3.txt".
> And it works.  The only problem with this is that messages printed out while
> the threads are running are a little confusing.  I have only one window for
> both threads and can't tell one message with another.  I guess I'll put up
> with that for now.

A word of warning, though -- you may not be getting much usage of 
your second processor.  Python itself has a Global Interpreter 
Lock, forcing only one Python thread to be running at any given 
time.  Secondary threads are not run on multiple processors.  You 
may not be having this problem, since you're executing external 
programs -- I'd imagine that those could be scheduled on the 
second processor, but you should verify this.

In general, Python doesn't make effective use of SMP.  Hopefully 
this will be fixed eventually, but it's a huge undertaking.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jeff at ccvcorp.com  Wed Apr  3 15:32:09 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Wed, 3 Apr 2002 12:32:09 -0800
Subject: date / time
References: <3CAAD0D1.2040309@beth.uniforum.org.za> <a8elrm$r7du9$3@ID-41653.news.dfncis.de>
Message-ID: <MPG.171506a638d9ca2c989694@news.nwlink.com>

In article <a8elrm$r7du9$3 at ID-41653.news.dfncis.de>, klausman-
un070202 at tuts.net says...
> mixo <mixo at beth.uniforum.org.za> wrote:
> > 
> > Suppose there is a given date, 2002-03-01 (YYYY-MM-DD).
> > How can you get the day (2002-02-28 ) before the given date and
> > check that it is valid?
> 
> (I replace the interpreters >>> prompt with ### for better
>  legibility in USENET posts)
> 
> ### import time
> ### date="2002-03-01"
> ### time.strptime(date,"%Y-%m-%d")
> (2002, 3, 1, 0, 0, 0, 4, 60, 0)

A word of warning about time.strptime(), from the library 
reference:

: Note: This function relies entirely on the underlying
: platform's C library for the date parsing, and some of these 
: libraries are buggy. There's nothing to be done about this 
: short of a new, portable implementation of strptime(). 
: 
: Availability: Most modern Unix systems. 

So, this will not work on Windows (I believe there is a 
strptime() implementation for Windows, but you'd have to hunt it 
down), and depending on your system libraries, even if it *does* 
exist it may be buggy.  (Perhaps this explains your 1-sec-off 
problem?)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From eppstein at ics.uci.edu  Sat Apr 20 12:15:41 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 20 Apr 2002 09:15:41 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <mailman.1019286100.3207.python-list@python.org>
Message-ID: <eppstein-D28D3E.09154120042002@news.service.uci.edu>

In article <mailman.1019286100.3207.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> > It affects other things than threading, too -- e.g. I use References:
> > to pick out responses to my own messages, and yours don't get caught.
> 
> Then you'll be spared this <wink>.

No such luck.  I also highlight and read your posts...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From phd at phd.pp.ru  Thu Apr 11 04:06:44 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 11 Apr 2002 12:06:44 +0400
Subject: Easter Eggs
In-Reply-To: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>; from robin@jessikat.fsnet.co.uk on Thu, Apr 11, 2002 at 12:17:40AM +0100
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <20020411120644.B30351@phd.pp.ru>

On Thu, Apr 11, 2002 at 12:17:40AM +0100, Robin Becker wrote:
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers.

   You are not, but most peeple (including me) find them funny.

   "I think it's extraordinary important that we in computer science keep
fun in computing. When it started out, it was an awful lot of fun. Of
course, the paying customers got shafted every now and then, and after a
while we began to take their complaints seriously. We began to feel as if
we really were responsible for the successful, error-free perfect use of
these machines. I don't think we are. I think we're responsible for
stretching them, setting them off in new directions, and keeping fun in the
house. I hope the field of computer science never loses its sense of fun.
Above all, I hope we don't become missionaries. Don't feel as if you're
Bible salesman. The word has too many of those already. What you know about
computing other people will learn. Don't feel as if the key to successful
computing is only in your hands. What's in your hands, I think and hope, is
intelligence: the ability to see the machine as more than when you were
first led up to it, that you can make it more."

     Alan J. Perlis (April 1, 1922 - February 7, 1990)

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




From just at xs4all.nl  Sat Apr 20 03:18:32 2002
From: just at xs4all.nl (Just van Rossum)
Date: Sat, 20 Apr 2002 09:18:32 +0200
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net> <zwFv8.36236$vF6.1097456@news2.tin.it> <a9nbuq$916$1@aphrodite.grec.isp.9tel.net> <uc14cjalpru3b5@news.supernews.com> <a9q8f4$svn$1@aphrodite.grec.isp.9tel.net> <mailman.1019269858.27293.python-list@python.org>
Message-ID: <just-47A290.09183220042002@news1.xs4all.nl>

In article <mailman.1019269858.27293.python-list at python.org>,
 com-nospam at ccraig.org (Christopher A. Craig) wrote:

> I have the utmost respect for Alex, so I don't understand why I can't
> figure out what he is saying, but why couldn't len(a) be equivalent to
> a.len()?

In a reply to my post, that either was private or never reached the 
group, he has one good argument *for* len as a function: you can't do 
the following easily without it:
   map(len, seq)
The equivalent with a method would be something like
   [i.len() for i in seq]
Indeed not as nice.

Still, I remain unconvinced also.

[ ... ]

> The main problem I see with this is that it wouldn't have worked
> before you got the ability to subtype in Python2.2, so historically it
> had to be a method before.  Plus adding it now means you either have a
              ^^^^^^
              "function", right?
> stub "len" method which may not work, or you have a major backward
> incompatibility.  

I don't think it's neccesarily the missing ability to subtype, but 
rather the huge difference between type and classes that simply would've 
made it much harder to give builtin types a [__]len[__] method 
accessible from Python. Not impossible, but harder. But I agree with you 
that len-as-a-function is probably a historic artefact.

> I personally disagree with adding it now, but not because it can't be
> done.

Well, it can't be done for b/w compatibility... But I know what you 
mean, and I agree on both counts.

Just


From eric.brunel at pragmadev.com  Tue Apr  9 13:12:13 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 9 Apr 2002 17:12:13 +0000
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com>
Message-ID: <a8v03v$b8o$1@norfair.nerim.net>

Henning Peters wrote:
> is there anything similar to the pointer concept from c/c++ in python?
>
> i would like to store values in two different locations. both values
> should always be equal. the one could be a pointer on the other. is this
> possible to do this in python?

I'd use an attribute in an object. Try this:

class C:
  pass
o = C()
o.a = 'whatever'
oo = o
print oo.a
oo.a = 'yeah'
print oo.a
print o.a

You have two names ("o" and "oo") pointing on the same object, so changing 
the value of the attribute "a" on one changes the value on the other one.

HTH
 - eric -



From donn at u.washington.edu  Wed Apr  3 13:35:06 2002
From: donn at u.washington.edu (Donn Cave)
Date: 3 Apr 2002 18:35:06 GMT
Subject: ctrl-c and os.system problem
References: <mailman.1017777544.11640.python-list@python.org> <a8dbup$hec$1@nntp6.u.washington.edu> <e27b96c5.0204022055.74c3d570@posting.google.com>
Message-ID: <a8fi0q$oqm$2@nntp6.u.washington.edu>

Quoth noomnews at yahoo.com (Noom News):
| "Donn Cave" <donn at u.washington.edu> wrote in message news:<a8dbup$hec$1 at nntp6.u.washington.edu>...
...
|> NetBSD's system(3) man page explains that system ignores SIGINT
|> and SIGQUIT.  I assume the same must be true on Red Hat Linux,
|> and Python calls C system().
|> 
|> Maybe you could use spawnv()?

| Thanks!  The strange thing is that SIGINT kill the os.system but was
| not propated to kill the whole python process.  This behavior made
| python different from perl where the whole process would be killed.

I'm not really familiar with perl, but I believe its system function
may not call the C system(3) function as Python's does.  In fact,
it's supposed to have some of the benefits that spawnv() provides.

Maybe you could use spawnv()?

	Donn Cave, donn at u.washington.edu


From peter at engcorp.com  Tue Apr  2 20:09:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 20:09:02 -0500
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com>
Message-ID: <3CAA562E.C5D592A2@engcorp.com>

Paul Rubin wrote:
> 
> David Goodger <goodger at users.sourceforge.net> writes:
> > Here's a serious proposal, safe to post now that April Fool's is over.
> > Please read & comment.
> 
> Man, I thought it was an April fools joke when I saw it last night.
> All those `different` :kinds: 'of' "quoting" are much more confusing
> than explicit markup.  So I think it's better to use explicit markup.

I had the same impression.  I was reading it for the first time,
as I showed it to a coworker who was interested.  By the time we'd
scrolled through the ReStructured Text documentation I had to say
to him, "Hmm... that's so bloody complicated it might be easier
writing XML tags directly.  Sort of defeats the purpose of structured
text, doesn't it?"

When structured text becomes so complicated to type that one has to
actually work at it, and it needs a large manual to use, it has
become too unwieldy to be useful.

My suggestion would be go with a simpler Structured Text for the 80%
of documentation needs it, and allow an escape to DocBook or 
something for more complicated docs...

-Peter


From peter at engcorp.com  Wed Apr 10 20:41:19 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:41:19 -0400
Subject: Help: Python for a big commercial project?
References: <F194tIpqqJgSt51NDa200002509@hotmail.com> <mailman.1018464082.15807.python-list@python.org>
Message-ID: <3CB4DBAF.CC7AE350@engcorp.com>

Cliff Wells wrote:
> 
> On Wed, 10 Apr 2002 13:24:59 +0200 Cereal Muesli wrote:
> 
> > - First of all; do you think that Python is suitable for this
> >   kind of "serious, big" project?
> 
> Absolutely.  I would not hesitate to use Python for *any* project.  At a prior
> job, I used Python develop the primary application for a contract worth over $3
> million a year.  In my current position, I am moving our applications from C to
> Python and these applications will be responsible for monitoring and
> controlling security systems in some high-risk/top-secret facilities (overseas
> military installations, government research facilities, etc - if I told you
> more, well, you know the line ;)

Complete, total, and 100% redundant agreement with Cliff on all the issues
he mentions here.  My company is writing applications from scratch in Python
to be used to monitor and control communication systems in similar facilities 
(would be fun someday to find that our stuff went into some of the same places) 
and for industrial use and if I told you more I'd have to ask Cliff to get one 
of the people he knows in his line of business to kill you...

Hey Cliff, did Bruno call you back yet?  I'm still kicking myself for that
accidental disclosure on the XP mailing list.  It's going to take so long for
him to track them all down through their email addresses ...  :-(

-Peter


From anotherdeadbard at yahoo.com  Mon Apr 29 00:14:18 2002
From: anotherdeadbard at yahoo.com (Jerry Beasley)
Date: Mon, 29 Apr 2002 04:14:18 GMT
Subject: snippet source code
References: <20020427232434.13edc1ca.wilk@flibuste.net>
Message-ID: <MPG.173693a7e183fb7b989680@netnews.insightbb.com>

Try : http://www.vex.net/parnassus/ - "The Vaults of Parnassus".
Good Luck - 
 J Beasley


In article <20020427232434.13edc1ca.wilk at flibuste.net>, wilk at flibuste.net 
says...
> hi,
> 
> when i learn i make a lot of little examples for me. Now i want to share
> my work, where can i put them ? I know activestate cookbook, sourceforge,
> but is there something else ?
> 
> bye
> 
> 


From djc at object-craft.com.au  Sun Apr 21 10:50:30 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 22 Apr 2002 00:50:30 +1000
Subject: Sybase module 0.35pre1 released
Message-ID: <m34ri5f6mh.fsf@ferret.object-craft.com.au>

What is it:
The Sybase module provides a Python interface to the Sybase relational
database system.  It supports all of the Python Database API, version
2.0 with extensions.

Notes:

In this release the module uses callback instead of inline error
handling from the Sybase CT library.  This has caused quite extensive
changes to the threading support inside the low level extension
module.

One of the nice side effects of using callback error handling is that
server errors while executing stored procedures will now be reported
correctly.

FreeTDS support is much improved in this version.  You will still
experience segfaults while using FreeTDS though.  These issues are
still being worked on.

For the adventurous, you can build for FreeTDS like this:

     python setup.py build_ext -D HAVE_FREETDS -U WANT_BULKCOPY
     python setup.py install

The module is available here:

    http://www.object-craft.com.au/projects/sybase/sybase-0.35pre1.tar.gz

The module home page is here:

    http://www.object-craft.com.au/projects/sybase/

- Dave

-- 
http://www.object-craft.com.au


From maik.hertha at volkswagen.de  Wed Apr  3 01:08:59 2002
From: maik.hertha at volkswagen.de (Maik Hertha)
Date: Wed, 03 Apr 2002 08:08:59 +0200
Subject: compiling Python with readline
References: <mailman.1017782947.29025.python-list@python.org>
Message-ID: <3CAA9C7B.20005@volkswagen.de>

Erik Price wrote:

>
> <quote source="Modules/Setup">
> readline -L /sw/lib/ readline.c -lreadline -ltermcap
> </quote>
>
> But this time my ./configure failed.
>
> My libreadline.* files are located in /sw/lib, they are 
> /sw/lib/libreadline.4.dylib, /sw/lib/libreadline.a, and 
> /sw/lib/libreadline.dylib.  I am guessing that I am giving the wrong 
> syntax for the -L option.  Can anyone see what I'm doing wrong here?
>
you should have a look on line 19 of Module/Setup:
------------>8---------

# Lines have the following structure:
#
# <module> ... [<sourcefile> ...] [<cpparg> ...] [<library> ...]
#
# <sourcefile> is anything ending in .c (.C, .cc, .c++ are C++ files)
# <cpparg> is anything starting with -I, -D, -U or -C
# <library> is anything ending in .a or beginning with -l or -L
# <module> is anything else but should be a valid Python
# identifier (letters, digits, underscores, beginning with non-digit)
#
-----------8<-----------
you should change your line to

readline readline.c -L/sw/lib -lreadline -ltermcap

then you should add your library path to LD_LIBRARY_PATH before 
execution (or ldconfig on linux).

m.

-- 

mit freundlichem Gru? /
best regards

Maik Hertha

--------------------------------------------------- h+h
EBSP Anwenderbetreuung, +49 5361 9-74950
Volkswagen AG / Brieffach 1721 / D-38436 Wolfsburg
http://ebsp.wob.vw.de              
maik.hertha at volkswagen.de
---------------------------------------------------
hartmann+hertha
it (beratung / entwicklung / support)
http://www.hartmann-hertha.de
mhertha at hartmann-hertha.de
--------------------------------------------------- h+h




From NO_SPAM_jmpurser2 at attbi.com  Thu Apr 25 11:28:03 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Thu, 25 Apr 2002 15:28:03 GMT
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com>
Message-ID: <7gVx8.69911$HH5.4733@rwcrnsc51.ops.asp.att.net>

I did a not dissimilar project recently and like you I didn't feel I had
time to stop and learn how to interface with MySQL before I started.  In my
case I had a 22 meg binary data file and an ASCII file with 156,000 records
of edit information for the data in the binary file.  My first "brute force"
method looked like it was going to take 80 hours to finish!  I tinkered
around with some indexing schemes and then looked into the anydbm module.
What I wound up doing was writing the ASCII file into an anydbm file by key
value on the first pass, then running through the binary file using the key
value to search for edit material in the anydbm file.  Much faster than
brute force and more reliable than my index routines.

Now that customer is happy with the solution and the time pressure is off
I'm looking into using a relational database for projects like these but in
a pinch the anydbm module can get you up and running quickly and for free.

Have fun.

John Purser

"Rad" <zaka07 at hotmail.com> wrote in message
news:ad381f5b.0204250629.5144d196 at posting.google.com...
> Thanks for all the suggestions and ideas, I'll try to answer to all
> your questions in this one post.
>
> It is a longish term project but after a few initial data extractions
> I'll have some time to re-implement the procedures if there is a need
> to do so.
> At this time I don't know for sure how many unique ID's are going to
> be, not less than 15 million I guess.
> Don't have any options on the way data is delivered, it's going to
> come on tape/s and all relevant files will be in fixed width format.
> I am OK with that, shouldn't have any problems reading it in and
> storing it on the HD unless DDS starts clicking :(.  I have four 73 GB
> SCSI disks on the computer that will run win2k, 2 GB of RAM, dual xeon
> processor and three power units.
>
> I agree that the actual work is not that complex even though there are
> quite a few other bits involved in the project that I thought were not
> relevant to mention in my earlier post, I also agree that it could be
> done in a relational database but if the files were smaller, for
> example Access (which I know quite well), I also know some SQL, but I
> never used MySQL, which is perhaps a possible route to take.  However,
> as I just mentioned I don't know anything about MySQL (except that one
> has to pay for it) + having been learning Python in the last few weeks
> I kind of like it + I was encouraged by your posts and I'm now pretty
> convinced that Python2.2 is the way to go. I won't have to access data
> randomly, just pull out the data fitting certain criteria from
> original 4 files and create resulting files.  I might decide
> (depending on the size) to put those into a database though.
> I don't understand how I can interface Python to a database?
>
> All four files have dates in them but they do come in YYYYMMDD format
> and I was planning to use string comparisons, same for the rest of
> data,I was thinking to treat it all as strings.
>
> My initial plan is to either split files into smaller files that could
> fit into the memory or read chunks of data from original big file
> using file_object.redlines(some_size_that_ can_ fit_ in_ the_ memory).
>
> I'm stuck with it (the project) and I'm looking forward to the
> challenge.
> Hope I don't regret saying this publicly!?
>
> I believe that initial filtering will reduce file size for at least
> 50% so I'll do some research on awk (for windows).  I guess there is
> info about it on the Internet.
>
> Thanks again for all your suggestions and ideas.
> Probably "talk" to you soon.




From sholden at holdenweb.com  Tue Apr 30 10:42:07 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 10:42:07 -0400
Subject: Problem with module
References: <aam9sa$bvfkp$1@ID-131182.news.dfncis.de>
Message-ID: <D5yz8.15957$xy4.10514@atlpnn01.usenetserver.com>

"Studio Masson" <segreteria at studiomasson.it> wrote ...
> Hello everybody.
> I have a strange "problem" (well, it's not a real problem but...). I'm
> developing a program in Python (2.2) and wxPython (well, it's my first
> python program...). It works as i expect on two pc, one with Windows XP
> professional, one with Windows 98. I simply "copy" the directory where the
> sources are, and then i work, using IDLE. The program is always into the
> same directory ("c:\dati\python") in all the computers. If i do the same
> (copy the program's directory) on a windows ME system, and try to run the
> program (CTRL-F5 with IDLE) it stops with the following message:
>

I'm sorry I can't be any direct help. However, you might wish to know that
because IDLE uses the Tkinter graphical troolkit, you might expect trouble
when mixing it with wxPython programs - you can get two graphical main loops
struggling for control over your windows.

regards
 Steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From mhammond at skippinet.com.au  Wed Apr 17 08:14:07 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 17 Apr 2002 12:14:07 GMT
Subject: Type Conversion to COM Variant
References: <a4488154.0204162329.1f7ec7f@posting.google.com>
Message-ID: <3CBD675C.9040904@skippinet.com.au>

Haegi Kim wrote:
> Hi, experts,
>  
> I have a COM client which gets a reference to a COM server.
> The COM client has a exec expression in order to execute a script in
> which the reference to the COM server is used.
> It works find with the COM server in InProcServer mode, but doesn't in
> LocalServer and raises the following error.
> -----------------------------------------------------------------------------
> exceptions.TypeError: Objects of type 'instance' can not be converted
> to a COM VARIANT"
> -----------------------------------------------------------------------------
> In what case the type conversion causes the error? And what should I
> do?
> Thanks in advance.

It looks like you are trying to pass a class instance as a param to a 
COM call.  Only COM objects can be passed in this way - arbitrary Python 
objects can not.  I can not give any more details without extra 
information, such as the traceback.

Mark.



From jriveramerla at yahoo.com  Sun Apr  7 01:03:56 2002
From: jriveramerla at yahoo.com (Jose Rivera)
Date: 6 Apr 2002 22:03:56 -0800
Subject: wxPython Book ?
References: <11e94203.0204031526.f11bfe6@posting.google.com> <3caef28e@nntp.server.uni-frankfurt.de>
Message-ID: <11e94203.0204062203.57889ae8@posting.google.com>

Thanks for your answers... I hope soon will be a book... I don't know
why QT and Tkinter has books and wxPython ... I hope that gap will
close in short time.


From brian_l at yahoo.com  Wed Apr 17 18:07:08 2002
From: brian_l at yahoo.com (Brian Lenihan)
Date: 17 Apr 2002 15:07:08 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz> <3CBC5C04.8F2C3D5A@noaa.gov>
Message-ID: <afe3b146.0204171407.6e5e4b0e@posting.google.com>

Chris Barker <Chris.Barker at noaa.gov> wrote in message news:<3CBC5C04.8F2C3D5A at noaa.gov>...
> Greg Ewing wrote:
> 
> > In theory, it would seem that you could install Unix python
> > on Mac OS X together with the Aqua version of Tcl/Tk, and
> > run Tkinter stuff.
> > 
> > When I tried this recently, however, it didn't work. I
> > got a Tkinter window, but it was "dead" -- it didn't
> > respond to mouse clicks. 
> 
> Check out the MacPython SIG mailing list, I think this is now working.
> Also, wxPthon on MacOS-X is getting very close.

Tony Lownds has instructions here:
http://mail.python.org/pipermail/pythonmac-sig/2002-April/005420.html

AquaTk seems to have some problems with path separators (try file open
in IDLE for an example).


From clubdelpelosuperfluo at yahoo.it  Mon Apr 22 08:37:03 2002
From: clubdelpelosuperfluo at yahoo.it (Stefano)
Date: 22 Apr 2002 05:37:03 -0700
Subject: Numeric Python, COM and Eigenvalues
Message-ID: <9e1f166b.0204220437.57a087de@posting.google.com>

I am working with Numpy and trying to create a COM server.
I managed to call this server by Excel and get results like Inverse
matrix or Cholesky decomposition but the server doesn't work as I ask
for Eigenvalues/vectors (Excel does not respond anymore).
Has anyone heard of a problem of this kind?

Thanks
 
Stefano


From geoff at gerrietts.net  Wed Apr 10 23:12:41 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 10 Apr 2002 20:12:41 -0700
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer division,backwards compatibility; where is  Python  going?)
In-Reply-To: <3CB4EC5A.691F1F7B@engcorp.com>
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com>
Message-ID: <20020411031241.GC30655@isis.gerrietts.net>

Quoting Peter Hansen (peter at engcorp.com):
> I'd like to take this opportunity to at least thank the developers,
> bots included, and other contributors (in the most heartfelt way I
> can without sending large cash contributions) for all the blood,
> sweat and keypresses they've put into this phenomenal project.

I'll echo that.

I owe a lot more thanks than I've delivered, and probably a bunch more
in the way of apologies than I've offered. Tim and Guido have both
personally helped me out of tight spots when I've run into them,
despite my semi-annual conservative whining.

I think that the community has a very high quality, despite the
occassional disgruntled voice or fearful citizen, and that reflects on
how easy the development team makes it to trust their judgement.

It's easy to get caught up in arguing the technical merits of this
widget or that feature and lose sight of why we're all here: python is
an excellent language with an amazingly clear syntax and so much
expressive power you almost can't use it all.

It's a project that's worth contributing to, for sure.

-- 
Geoff Gerrietts             "Overspecialize and you breed in weakness.
<geoff at gerrietts net>     It's slow death." -- Maj. Motoko Kusanagi
http://www.gerrietts.net                          (Ghost in the Shell)




From juenglin at informatik.uni-freiburg.de  Tue Apr 30 08:17:28 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 30 Apr 2002 14:17:28 +0200
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <199334c5.0204260645.3af87dc5@posting.google.com>
Message-ID: <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de>

billdozier at hotmail.com (Bill Dozier) writes:

> Ralf Juengling <juenglin at informatik.uni-freiburg.de> wrote in message news:<xduy9fa4z2x.fsf at leto.informatik.uni-freiburg.de>...
> 
> 
> 
> This looks like the old "mutable initializer" thingie.
> 

yeah probably old traps, newbies are falling into in any season.


> > >>> l = [[] for i in range(n)]
> 
> > Can this be done more elegant? 
> "[[] for i in range(n)]" looks fine to me.
> 
> > By some expression that was similar to '[[]]*n' ? 
> Do you mean, "with few characters to type?" That's not the same as
> "with more elegance."

That's admittedly a bit finical, but I don't like to have to create a
temporary list to create the list l. That's why it looks unelegant to
me. (Okay, I could use 'xrange' instead of 'range'...)

Here's a proposal, that is 'more elegant' in my eyes. If the builtin 
'list' type was an *iterable* factory function and we'd borrow the 
function 'take' from a more functional language, say

def take(n, i):
    if not n:
        return []
    else:
        i = iter(i)
        return [i.next()] + take(n-1, i)

then 

l = take(n, list)

would do it.

BTW: It is not easy to add __iter__ to list's meta-type, is it?

Ralf



From peter at engcorp.com  Tue Apr 23 23:59:58 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 23 Apr 2002 23:59:58 -0400
Subject: Php vs Python
References: <mailman.1019615325.15799.python-list@python.org>
Message-ID: <3CC62DBE.EC54FFE6@engcorp.com>

"Ramil D. Tolentino" wrote:
> 
> Could someone give me a run down between python and php. 
> Which is easier to use? which is better or more powerful? Thanks

Uh, PHP is better but Python is more powerful?

Or was it the other way around?

(You really need to explain something about your intended use
before the answers will mean anything to you.  What do you
want to accomplish?)

-Peter


From mlh at vier.idi.ntnu.no  Thu Apr 11 23:28:52 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 03:28:52 +0000 (UTC)
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it>  <3cb5bd03$1_6@goliath.newsgroups.com>
Message-ID: <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no>

In article <3cb5bd03$1_6 at goliath.newsgroups.com>, Brad Clements wrote:
>Heh.  3 posts lambasting me for using type(), and none of them actually
>answered my question. ;-)

Sorry about that. It is a very important point, though (not to use
type checking). If you give the problem where you are using it, I'm
sure we can suggest a more Pythonic alternative (if you're interested,
of course).

>The point of my post is: why does type(newstyleinstance) return <class
>__main__.x> rather than something like <instance __main__.x>?

Because <class __main__.x> is the type of newstyleinstance, also
called its class (the class named "__main__.x" or simply "x").

See? In Python 2.2 there is no sharp divide between types and classes
anymore.

>aaaanyway.. "type(thing) in [InstanceType]" is what xml_pickle uses to
>figure out how to pickle stuff.

Well, that doesn't make it any better.

>Since xml_pickle dies trying to pickle instances of new style classes, I
>went through and added isinstance(thing,ObjectType) which seems to work.

Hm. How about simply removing the type check and see what happens?
Such a check might have been necessary with old-style classes
(although I think I'd rather check for the presence of a __class__
attribute or something) it might well be that you can pickle
non-instance objects the same way as instances with newstyle classes.
After all, there is no big difference anymore.

>However you also say its wrong, which I agree is the wrong solution.
>
>I just want to be able to xml_pickle any Python object. I don't write this
>code, I just use it!

I see. It seems to rely on the division between instances and
non-instances, which is becoming more and more meaningless. I suggest
submitting a bug report to the author (suggesting a major rewrite...)

>So what's needed is a reliable mechanism for determining that "something" is
>an instance of a class. But in 2.2, looks like everything is an instance.

More or less.

>Fortunately I can just wait for someone smarter than I am to solve this
>problem.
>
>--
>Novell DeveloperNet Sysop #5

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From Markus_Ewald at gmx.net  Tue Apr 23 06:21:27 2002
From: Markus_Ewald at gmx.net (Markus Ewald)
Date: Tue, 23 Apr 2002 12:21:27 +0200
Subject: Linker error when doing MSVC debug builds
Message-ID: <3CC535A7.2070607@gmx.net>

No this is not a cry for help :)
I've got a little comment about the MSVC build setup:

First, note that it is absolutely legal for a win32 project to compile 
in debug mode and link release libraries (if the intentioned is not to 
debug the specific library).
As an example, I think noone outside Microsoft has ever possessed debug 
builds of user32.lib or kernel32.lib, yet everyone is compiling his 
debug projects with those :)


The windows setup program only contains python22.lib
However, if you do a debug build of your project (and _DEBUG is defined) 
pyconfig.h tries to link python22_d.lib, resulting in a linker error.

This leads to another problem:
To avoid modifying pyconfig.h, most people would simply rename/copy 
python22.lib to python22_d.lib.

Result: _DEBUG enables Py_DEBUG enables Py_TRACE_REFS which then
           uses the _Py_DeAlloc() function instead of the same-named
           macro. _Py_DeAlloc() is not exported in the release build
           of Python22.lib.
           -> Unresolved external symbol __imp___Py_DeAlloc()

           I guess there were already one or two newbies asking
           why they got unresolved externals while they
           obviously linked python22(_d).lib in their projects ;)


Solutions: -Do not define Py_DEBUG when _DEBUG is defined
               Requires anyone who wants to debug python code to
               manually #define Py_DEBUG in his projects.

              -Export _Py_DeAlloc() in Release mode as well
               Also done in STLport, where stldebug support can
               be used in release mode as well.

              -Use an ugly hack when including Python.h:
               #ifdef _DEBUG
                 #undef _DEBUG
                 #include <Python.h>
                 #define _DEBUG
               #else
                 #include <Python.h>
               #endif
               ...no words needed.


Thx,
-Markus-



From robin at jessikat.fsnet.co.uk  Thu Apr 11 18:36:56 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 11 Apr 2002 23:36:56 +0100
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx>
 <mailman.1018391839.6535.python-list@python.org>
 <mailman.1018509927.741.python-list@python.org>
Message-ID: <SVS1yIAIAht8Ewzm@jessikat.fsnet.co.uk>

In article <mailman.1018509927.741.python-list at python.org>, Luigi
Ballabio <ballabio at mac.com> writes
>Robin,
>        I'd like to try your coffee brand (ok, so I like it strong. Sue me.) 
>Which one is it?
>
>Bye,
>        Luigi
Capstan Full Strength
-Beautiful lies are not better than ugly truths-ly yrs- 
Robin Becker


From dale at riverhall.NOTHANKS.co.uk  Mon Apr  1 13:12:47 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Mon, 01 Apr 2002 19:12:47 +0100
Subject: Access to main namespace from imported module?
Message-ID: <re8haugrbcchbephckl9t1p1s27tvg18hr@4ax.com>

Just curious, like, but is it possible for an included module to
reference the namespace of the outer-most module, "__main__".

I can't see a way.

For example:

Module A contains:

-----
import moduleb

def wibble():
    """Handy docstring."""

-----
Module B contains:
-----
def wobble():
    print __main__.wibble.__doc__
-----

Obviously, this doesn't work but the way that __main__ is described in
the docs suggests that it might.

Thanks
--
Dale Strickland-Clark
Riverhall Systems Ltd


From ShantelS at ActiveState.com  Tue Apr  2 14:12:58 2002
From: ShantelS at ActiveState.com (Shantel Shave)
Date: Tue, 2 Apr 2002 11:12:58 -0800
Subject: ActiveState Programming Survey - Win $1000 O'Reilly Bookshelf
Message-ID: <01bc01c1da7a$66f0d8d0$7503a8c0@activestate.ca>

ActiveState is excited to offer you the chance to share your opinion, and win a
$1000 O'Reilly Bookshelf!

As open source programmers and users ourselves, we really believe in the power
of collaboration. Your feedback is invaluable in helping us to better serve the
Python community through our open source projects and product development.

Please share your thoughts by completing our online Programming Survey.

All respondents have the option of entering a draw to win a $1000 O'Reilly
Bookshelf. If you win, you'll be able to customize your prize from all available
titles in the O'Reilly catalog, to a total value of $1000!

Survey must be completed by April 8, 2002. Draw to be held April 10, 2002.

To take the survey, please click here:
http://www.ActiveState.com/AS_Programming_Survey

Thank you!
ActiveState




From kwokng at earthlink.net  Mon Apr 29 15:26:24 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Mon, 29 Apr 2002 19:26:24 GMT
Subject: python2.2
References: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net> <slrnacr3io.nil.gerhard@lilith.my-fqdn.de>
Message-ID: <A7hz8.22$Uw4.4633@newsread1.prod.itd.earthlink.net>

This is my source.list, would you please tell me how to change this file to
allow me to install the testing packages?

deb ftp://ftp.debian.org/debian/ stable main
deb-src ftp://ftp.debian.org/debian/ stable main
deb http://non-us.debian.org/debian-non-US stable/non-US main
deb-src http://non-us.debian.org/debian-non-US stable/non-US main

deb ftp://ftp.debian.org/debian/ stable main
deb-src ftp://ftp.debian.org/debian/ stable main
deb http://non-us.debian.org/debian-non-US stable/non-US main
deb-src http://non-us.debian.org/debian-non-US stable/non-US main

Thanks!

Billy Ng

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnacr3io.nil.gerhard at lilith.my-fqdn.de...
> Billy Ng wrote in comp.lang.python:
> > I am using debian linux.
>
> Which distribution? stable (potato), testing (woody) or sid?
>
> > I am trying to install the python2.2 that is an unstable package
> > onto the system, but it says "Couldn't find package python2.2"
>
> If you're using testing or sid,
>
>     apt-get update
>     apt-get install python2.2
>
> should install it.
>
> > Wouls anyone know how to install the unstable package via apt-get
>
> You don't need the unstable package. python2.2 is already in testing
> (woody).
>
> If this doesn't work for you, most probably you haven't set up your
> /etc/apt/sources for packages from testing or sid, yet.
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>




From James.Strater at exolink.com  Wed Apr 10 10:32:06 2002
From: James.Strater at exolink.com (James Strater)
Date: Wed, 10 Apr 2002 09:32:06 -0500
Subject: Create Excel Files without COM
Message-ID: <F666B26E8EF5B84C88E3DF34E5A1E49CF0818D@venus.exolink.com>

Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::

> Hi!
> 
> Is there something like WriteExcel from perl in python?
> I need to generate an MS-Excel file. I don't want to use
> COM although it will run on windows.
> 
> The data is simple. Just text in columns and rows.


You can format your output as HTML, and then put a .xls extension on it.
When it is opened in excel, it will be fine.  To see what your input needs
to look like, take your spreadsheet and 'save as' HTML.


James






From aleax at aleax.it  Fri Apr 26 17:21:07 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 21:21:07 GMT
Subject: How do I format this
References: <8f8ffe67.0204261312.77787e94@posting.google.com>
Message-ID: <7xjy8.66363$vF6.2030761@news2.tin.it>

Sunit Joshi wrote:
        ...
> However it fails at <td width="1\%"> eventhough I'm escaping the '%'
> with a message ValueError: unsupported format character '"' (0x22) at

You don't "escape" % with a \ in a format string: rather, you double it
up.  Change that \ into another % and be happy.


Alex



From cepl.m at neu.edu  Sun Apr 28 17:14:50 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Sun, 28 Apr 2002 21:14:50 +0000 (UTC)
Subject: IMAP2IMAP Synchronization -- iis [Was: Re: iis: comparsion of two lists]
In-Reply-To: <20020428115225.GA999@rehor.surfbest.net>
References: <20020426224924.GA1318@rehor.surfbest.net> <20020426232802.GA1927@rehor.surfbest.net> <JTry8.26517$8D3.706486@news1.tin.it> <20020428034636.GB1279@rehor.surfbest.net> <yqNy8.71678$vF6.2141429@news2.tin.it> <20020428115225.GA999@rehor.surfbest.net>
Message-ID: <20020428211449.GB1467@rehor.surfbest.net>

On Sun, Apr 28, 2002 at 11:52:27AM +0000, Matej Cepl wrote:
> Thanks a lot, that's what I was exactly interested in.

Hi,

just let everybody know, that I am giving up. The last version is 
available on http://www.volny.cz/cepls/ps-pdf/iis.py and if 
anybody wants to finish it and debug it, is very welcome (please 
let me know though).

	Have a nice day
	
		Matej

-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
A nation or civilization that continues to produce soft-minded
men purchases its own spiritual death on an installment plan.
	--Martin Luther King, Jr. 



From greg at cosc.canterbury.ac.nz  Mon Apr 15 22:29:29 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 16 Apr 2002 14:29:29 +1200
Subject: Does Python need a '>>>' operator?
References: <mailman.1018888350.13078.python-list@python.org>
Message-ID: <3CBB8C89.150B3E01@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> [Greg Ewing]
> > The way I see it, if you're doing bit twiddling
> > using unified ints/longs, you should always be
> > working with *positive* integers
> 
> Viewing longs as characteristic bit vectors, they model finite and co-finite
> sets.

Well, I meant finite bit strings,
which is what people twiddle most of the time.

> Only finite sets "look positive"; co-finite sets "look negative".  It
> doesn't take more than the unary ~ operator to go from one kind to the
> other

Yes, that's a bit of a nuisance. Whenever you
do a complement, you really need to specify how
many bits you're complementing somehow, e.g.

  y = ~x & 0x0ffffffff

The neatest solution would be to have a special
fixed-length-bit-string type. Maybe if array.array
grows a boolean type option, we could give it
& | ~ << >> operators too.

Come to think of it, bit operators would apply
quite nicely to array.arrays of char/short/long
too.

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


From jamescalthoff at yahoo.com  Sun Apr  7 16:03:42 2002
From: jamescalthoff at yahoo.com (James Althoff)
Date: Sun, 7 Apr 2002 13:03:42 -0700
Subject: Argument against iterable ints...
Message-ID: <000701c1de6f$51d16ca0$54dac13f@Pacbell.net>

[Magnus Lie Hetland]
> def flatten(seq):
>     try: iter(seq)
>     except: yield seq
>     else:
>         for sub in seq:
>             for item in flatten(sub):
>                 yield item
>
> Anyway, I guess I shouldn't frame this as an argument against the
> rejected PEP <snip>

Indeed.  Especially since the referred to infinite recursion is *not* a
problem with iterable ints!

I believe you are referring to the following issue:

>>> iter('a').next() == 'a'
1
>>> iter('a').next() == iter(iter('a').next()).next()
1
>>>

This situation does not arise for iterable ints -- which we can simulate
with xrange as in:

>>> iter(xrange(0)).next() == 0
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
StopIteration
>>>

This is because iter(n) returns (hypothetically speaking, of course ;-) 0,
1, . . ., n - 1, "raise StopIteration" -- meaning *only* "raise
StopIteration"  happens for iter(0).next().  (And other positive ints
eventually "count down" to 0).

Note the following simulation:

>>> from __future__ import generators
>>>
>>> def flatten(seq):
...     try: isinstance(seq,int) or iter(seq)
...     except: yield seq
...     else:
...         if isinstance(seq,int):
...             iterator = iter(xrange(seq))
...         else:
...             iterator = iter(seq)
...         for sub in iterator:
...             for item in flatten(sub):
...                 yield item
...
>>> list(flatten(4))
[]
>>>

Notice that there is no infinite recursion for integers.

But you did uncover an interesting issue with *strings* and iterators!  :-)

Jim





From starx at pacbell.net  Sat Apr 27 05:19:27 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 27 Apr 2002 09:19:27 GMT
Subject: problems redirecting unittest output to sys.stdout
References: <aadplo$rs7$06$1@news.t-online.com>
Message-ID: <Xns91FD17D69F41ARASXnewsDFE1@130.133.1.4>

Achim Domma || Sat 27 Apr 2002 02:13:33a:

> Hi,
> 
> I want to redirect the output of a unittest to a textfile (on
> Windows). The default stream is sys.stderr which can not be redirected
> on windows. Until now I started my tests like this :
> 
> [..code..]
> 
> This way the output is redirected to sys.stdout, but the command line
> parameters are ignored. If I call my script with the '-v' parameter, I
> get no verbose output. I have no idee where the parameters get lost !?
> 
> any idea ?

This is one of the reasons i keep cygwin around. Link your windows 
python binary something like /bin/wpy, and run 'wpy foo.py 2>&1 > 
out.txt' or however you redirect streams under unix (haven't done it in 
a while =).

There are also some free little 'bootstrap' kinda programs for windows, 
which do something along the lines of '1. copy stderr to stdout, 2. run 
whatever you pass to it as an argument' (like some slowdown programs). 
This would be a significantly smaller way of getting what you want =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From gpropf at cfl.rr.com  Tue Apr  2 01:56:29 2002
From: gpropf at cfl.rr.com (Gregory Propf)
Date: Tue, 02 Apr 2002 01:56:29 -0500
Subject: [Fwd: Re: Extending Python on Windows]
Message-ID: <3CA9561D.8040207@cfl.rr.com>

Accidentally sent this to Gerhard instead of the list itself.

-- 
You see, "state of the art" Microsoft Office applications sport a
nifty feature called "smart quotes." (Rule of thumb--every time
Microsoft uses the word "smart," be on the lookout for something
dumb).  - John Walker, founder of Autodesk.
-------------- next part --------------
An embedded message was scrubbed...
From: Gregory Propf <gpropf at cfl.rr.com>
Subject: Re: Extending Python on Windows
Date: Tue, 02 Apr 2002 01:07:37 -0500
Size: 2615
URL: <http://mail.python.org/pipermail/python-list/attachments/20020402/d1ff2c97/attachment.mht>

From gimbo at ftech.net  Wed Apr 24 05:26:38 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Wed, 24 Apr 2002 10:26:38 +0100
Subject: urllib + headers
In-Reply-To: <aa5rnk$68k$1@news.iucc.ac.il>
References: <aa5rnk$68k$1@news.iucc.ac.il>
Message-ID: <20020424092638.GA13701@andy.tynant.ftech.net>

On Wed, Apr 24, 2002 at 11:52:26AM +0200, eli li wrote:
> I open url with urllib.urlopen, open method is POST. How is it
> possible in request to specify headers like 'Referer',
> 'Accept-Language', etc?

Don't use urllib.urlopen()  :-)

Using urllib2, create a urllib2.Request object, call its add_header()
method to add the headers, then pass the object to urllib2.urlopen()

Hope this helps,

Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From jeff at ccvcorp.com  Tue Apr  2 17:17:12 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 2 Apr 2002 14:17:12 -0800
Subject: wxpython: problem with events and checklistbox
References: <3caa2be5.48482250@news.easynews.net>
Message-ID: <MPG.1713cdc7548837a698968f@news.nwlink.com>

In article <3caa2be5.48482250 at news.easynews.net>, stojek at part-
gmbh.de says...
> HI,
> 
> in my GUI I have a wxCheckListBox (named CheckList) and an fuction
> IfChecked(self,event) that process the checkevents (EVT_CHECLISTBOX).
> 
> Now, when unpickling all data at the start of my program,
> I want to check some items with CheckList.check(index,true)
> My problem is, that after each checking the function IfChecked is
> called. 

If I'm understanding your problem correctly (you don't want to 
call the usual handler while you're unpickling), then the quick 
solution would be to delay calling EVT_CHECKLISTBOX() until after 
you've unpickled things.  At that point, no handler will be 
registered; once the item is unpickled, register the handler in 
the usual way (by calling the EVT_* function).

Of course, determining when it's safe for you to register the 
handler, and making it work both for new instances and unpickled 
instances, is up to you...  ;)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From maxm at mxm.dk  Tue Apr  9 12:13:41 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 09 Apr 2002 18:13:41 +0200
Subject: yet another "Here's to 1.5.2" post
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no> <usn64ame1.fsf@python.net>
Message-ID: <3CB31335.2040501@mxm.dk>

Michael Hudson wrote:
> Syver Enstad <syver-en+usenet at online.no> writes:

>>Am I the only one in the entire world that thinks '\t'.join(mylist) is
>>elegant and natural?
> 
> No.

Well I would have found mylist.join('\t') more natural, but I can live 
with the other.

It's probably just old JavaScript habbit that creeps up.

regards Max M



From boud at valdyas.org  Fri Apr 12 16:41:27 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Fri, 12 Apr 2002 22:41:27 +0200
Subject: problems writing utf8
Message-ID: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl>

It started when I wanted to write utf-8 data to standard output.
I've got an utf-8 enabled konsole, so I didn't want any of the
default escapes print uses.

However, that didn't work -- giving errors because some of my data
went way beyond ascii ordinal 128.

Then I tried to write the utf-8 data to a file. I have tried to
construct that file with two methods:

    f = open("syllables", "w+")
    d2 = codecs.EncodedFile(f, "unicode_internal", "utf-8")
    f2.write(u"a")
    f2.close()

This segfaults when Python exits:

Python 2.2.1 (#1, Apr 12 2002, 22:26:33)
[GCC 2.95.3 20010315 (SuSE)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import codecs
>>> f = open("syllables", "w+")
>>> f2 = codecs.EncodedFile(f, "unicode_internal", "utf-8"
... )
>>> f2.write(u"a")
>>> f2.close()
>>>
Segmentation fault

The second method writes garbage-encoded data to the file:

    f3 = codecs.open("syllables2", "w+", "utf-8")
    f3.write(u"?")
    f3.close()

(Where u"?" contains any Unicode character you like -- in this
case a glottal stop.)

I'm sure I must be doing something very silly -- in the second case.
As for the first, I rather thought Python didn't segfault... 

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


From db3l at fitlinxx.com  Wed Apr 24 13:37:44 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 24 Apr 2002 13:37:44 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <uofg9huaf.fsf@ctwd0143.fitlinxx.com>

zaka07 at hotmail.com (Rad) writes:

(working with 80GB worth of text file data)

> I have a few weeks to prepare myself before data arrives and my
> question is: am I going the right way about the project, is Python
> (with humanly written code) capable of doing this kind of staff
> relatively quickly?

If you're comparing manipulating the files with Python versus
"learning" C (even if you're also learning Python), I think that even
with any reduction in performance in Python you'll likely have a net
gain in time.  My bet is you'd easily eat up any time savings by the
final C application in terms of building and debugging the code.  And
in this sort of application, you may find your overall performance
swamped by I/O handling rather than computation, so Python may not
be at that much of a disadvantage.  And some of your correlation
requirements would require much more coding in C than in Python due
to the native structures such as dictionaries in Python.

This does depend somewhat on the frequency of these manipulations
(e.g., are you building something that will be used daily on data sets
of this size or just this once), but even if this is a long term
project, you sound like you have some immediate deadlines for this
specific data set - and you can always choose to re-implement for
repeated runs later if you find you need to.
  
A few suggestions:

* While you don't have the files yet it sounds like you have the specs
  for the files - write up some small test files and use them to work
  on the code.  And even when you initially get the files, extract a
  small portion of them to develop against to streamline testing.
  Assure yourself as much as possible that you're comfortable with
  your code before wasting time by turning it over on the real 15GB
  files.

* Some of the cross-referencing you seem to need to do will most
  likely be easy matches to Python dictionaries (perhaps keeping an
  index of key fields and where they exist in the source file) so
  practice up on them.

* Use the latest Python (2.2) - many improvements have been made on
  the basic I/O performance in the common file processing loop (e.g.,
  "for line in file:") that you'll get the advantage of.

* Don't skimp on your hardware.  Find the beefiest machine you've got
  handy, with the fastest drive, and ensure you have plenty of memory.
  Spend a few bucks on memory (it's cheap) if you need to.  It's quite
  likely that in final processing, you'll be heavily dependent on I/O
  performance - just copying a 15GB file is a time consuming process.
  
* I don't know your platform, but since these are text files, make
  sure you gain some familiarity with any text processing utilities
  you may have.  If you're under Unix, you want to be very comfortable
  with stuff like head, tail, split, grep, and diff.  Tools like these
  can make it much easier to manipulate the files during development
  testing or even as part of an overall process of handling them.  If
  you can find the time to check out awk you may even find that if
  this is a one shot task, a few quick awk scripts are all you need.

  If you aren't under Unix, I'd suggest getting ahold of equivalent
  tools (personally, I'd just install Cygwin and get all the same Unix
  ones).

* Be sure you have a good text editor or browser (e.g., less) that
  you're comfortable with for examining the raw files if you run into
  questions.  You're unlikely to really want to load a 15GB file into
  any editor or OS, regardless of virtual memory handling, but some of
  the previous text tools can chop up the file into more manageable
  pieces if you need to check something out.  And a simple browser
  like less can page through a file without keeping much of it memory.
  
--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From borcis at geneva-link.ch  Thu Apr 18 04:59:29 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 18 Apr 2002 10:59:29 +0200
Subject: Importance of trademarks (was RE: [OT] What is Open Source?....)
References: <mailman.1019068721.15404.python-list@python.org> <87n0w1qxk3.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <3CBE8AF1.1020901@geneva-link.ch>

Stephen J. Turnbull wrote:


> And the lack of a "clean-room" defense creates a chilling effect on
> the whole industry because of the enormous cost of defending yourself
> against accusations of infringement.


After medicine, software development appears doomed to finance 
arbitrarily large populations of lawyers, at least under US law. Maybe 
this means physicians and software engineers should ally against that 
obfuscation industry. After all, due process should be a pathology 
correcting activity, and physicians and software engineers are both 
chiefly diagnostic-and-cure professionals.

Regards, Boris Borcic
--
assert BOA in BAOBAB, "Paradise Mismatch Error"



From erno-news at erno.iki.fi  Sat Apr  6 18:12:01 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 07 Apr 2002 02:12:01 +0300
Subject: socket security
References: <40sr8.3765$rg3.345703@news-binary.blueyonder.co.uk>
Message-ID: <kuhemol8vy.fsf@lasipalatsi.fi>

In article <40sr8.3765$rg3.345703 at news-binary.blueyonder.co.uk>,
Alastair Nicol <calaban.madness at blueyonder.co.uk> writes:

| 2) Also when a client connects to a socket the server gets the IP address 
| of the client. Is this the "real" ip number? Ive heard you can spoof IP 
| numbers so what hoops are needed to validate that the client is really that 
| client.

yes, it is trivial to forge the source address of ip packets.

if the attacker wants to do 2 way communication with the forged
address, he will also need to intercept the packets that are sent to
the phony address. (this is easy if the attacker is on the same lan,
by using eg arp spoofing.)

many tcp implementations have fairly unpredictable initial sequence
numbers, so 2 way communication is required to set up a tcp
connection.

| openSSL is an option i would like to use, but for reasons to numerious to 
| list here, i cant.

if you only need authentication, it's easy to roll your own
shared-secret system by using sha or md5 from the standard
library. this requires you to handle distributing the secrets to
clients somehow.  see eg "digest authentication" in
<URL: http://www.faqs.org/rfcs/rfc2617.html>.

  -- erno



From sholden at holdenweb.com  Tue Apr 23 10:08:04 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 10:08:04 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com>
Message-ID: <tXdx8.94618$T%5.47937@atlpnn01.usenetserver.com>

"Fernando Rodr?guez" <frr at easyjob.net> wrote in message
news:uloacus4pml1r7f7eac9ts3u20oroj34qb at 4ax.com...
> On Tue, 23 Apr 2002 08:21:30 -0400, "Steve Holden" <sholden at holdenweb.com>
> wrote:
[incorrect suggestion about string quotes]
>
> If I add the string quotes again, I get the same AttributeError exception
I
> was getting before:
>
> >>> curs.execute("insert into customer_info (email, url, referrer,
first_visit) values ( '%s', '%s', '%s', '%s')", ('billg at microsoft.com',
'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
> Traceback (most recent call last):
>   File "<pyshell#56>", line 1, in ?
>     curs.execute("insert into customer_info (email, url, referrer,
> first_visit) values ( '%s', '%s', '%s', '%s')", ('billg at microsoft.com',
> 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00'))
>   File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line
227, in
> execute
>     af =af+self.__res.affectedrows()
> AttributeError: 'NoneType' object has no attribute 'affectedrows'
> >>>
>
> >It may or may not make a difference. What's the difference between
> >CompatMySQLdb and MySQLdb?
>
> I'm not sure. I haven't been able to find MySQLdb. I found this Compat
thing
> in MySQL-python-0.9.1.win32-py2.2.exe at
>
http://sourceforge.net/project/showfiles.php?group_id=22307&release_id=51760
>
> Where can I find MySQLdb for win32 _and_ Linux? O:-)
>

Sorry I was wrong about the quotes: this apparently only works with the "?"
paramstyle (or possibly it's module-dependent: mxODBC is fine with just
question marks).

I'm confused about why you don't get a proper MySQLdb installation from the
same file I used, though. When I ran it, it installed a MySQLdb directory
ujnder <python>/Lib/site-packages, and I have no problems importing MySQLdb.

In your "Add/Remove programs" control panel item do you see "Python 2.2
MySQL-python-0.9.1"? Maybe a de-install/reinstall would help. My own system
has no reference at all to CompatMySQLdb that I can see, but I notice that
when I open the installer .exe file in WinZip there it is. Did you perchance
treat the installer as an archive rather than running it?

Just in case I'm wrong on this one I'm going to try de-installing and
re-installing the Windows version of the package. It's pretty robust,
though: with Andy Dustman's help I have got it to install from the source
under 2.2 on cygwin as well as under my Windoze installation. It's nice
having MySQLdb and mxODBC under both Windows *and* cygwin.

I'll report back if I discover anything more.

regards
 Steve
--
home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/





From tejarex at yahoo.com  Fri Apr 19 21:34:21 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sat, 20 Apr 2002 01:34:21 GMT
Subject: GNU GPL & Python stand-alone packages from py2exe
References: <i0kit6nn3l4.fsf@Thales.MI.Uni-Koeln.DE> <j4hem7u2ol.fsf@informatik.hu-berlin.de> <MPG.1729fff0170354d59896b8@news.nwlink.com> <m3pu0v4e65.fsf@mira.informatik.hu-berlin.de>
Message-ID: <xA3w8.52172$XV5.4661697@bin4.nnrp.aus1.giganews.com>

> channel as the binary". So if the binary is on CD, the source must
also.

I am a bit puzzled why the OP is making such a big fuss about this:
CDrom = 640M, Python source about 1M.  Tuck it away in obscure
subdirectory.  Mention existence in one sentence where required.  and
forget about it.

Terry J. Reedy





From parz at shaw.SpamBucket.ca  Wed Apr 10 12:45:53 2002
From: parz at shaw.SpamBucket.ca (Parzival Herzog)
Date: Wed, 10 Apr 2002 16:45:53 GMT
Subject: Building an extension that calls the Python interpreter recursively.
Message-ID: <5%Zs8.7857$de1.379233@news3.calgary.shaw.ca>

I am planning to build a Python extension module for the
ML/I macro processor. I am not too experienced at
building Python extensions, and I have some questions
about the design of this extension. I would appreciate
any comments.

ML/I is a text processor, reading a stream of input text
and producing (zero or more) streams of output text.

During ML/I's processing, I want ML/I's macro language
to be able to evaluate Python statements and expressions,
the purpose being 1) to compute strings not easily
computable with ML/I's macro language; 2) to access
external information via Python's facilities; and 3) to
record information parsed by ML/I into Python data
structures.

What I plan to do is to implement an extension type
whose Python prototype would look something like this:

class ML1:
    def __init(self,...)
        ...
    def SetInputStream(self, stream):
    ''' record file-like object representing the input stream'''

    def SetErrorStream(self, stream):
    ''' record file-like object representing the error output stream'''

    def AddOutputStream(self, stream):
    ''' record an additional file-like object representing an output
stream'''

    def process(self, globals,locals)
    '''ML/I will consume input, and produce output,
      evaluating Python expressions in the context of
      globals, locals.
    '''

Now during process(), ML/I will compute strings S representing
Python expressions or statements, and evaluate them with

        PyRun_String(S,type,globals,locals)

I would like to know if this scheme is feasible, i.e.

1) Can the Python interpreter be re-invoked from an
extension module?
2) Will it process its evaluation of S in the same
interpreter context that originally invoked ML1.process?
I.e. will results left by PyRun_String be accessible
after process() returns?
3) Are there gotchas or constraints on the global
interpreter lock that I should be aware of under these
circumstances?

Thanks,
   Parzival




From bokr at oz.net  Sun Apr  7 23:48:09 2002
From: bokr at oz.net (Bengt Richter)
Date: 8 Apr 2002 03:48:09 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017466974.6751.python-list@python.org> <slrnaaha69.nn5.kzaragoza@localhost.localdomain> <a8akqh$8bk$0@216.39.172.122> <a8gk26$b16$0@216.39.172.122> <vMRq8.57284$S52.1995953@news2.tin.it>
Message-ID: <a8r3tp$dno$0@216.39.172.122>

On Thu, 04 Apr 2002 05:46:03 GMT, Alex Martelli <aleax at aleax.it> wrote:

>Bengt Richter wrote:
>        ...
>>>in the C code (and I haven't checked), then __nonzero__() (or
>>>__len__()) is called if defined.  This gives us the control we're
>>>looking for.
>> 
>> Well, I wouldn't want to mess with __len__ if I e.g., were subtyping str
>
>Of course not.  But __nonzero__ exists for the SOLE purpose of
>defining the object's truth value.
>
I didn't know that was its SOLE purpose. Its name has nothing to do with
"truth value," so when I first saw it, I thought to myself that it was
something I'd have to look into if I defined a class that had some numeric
aspect to expose. I assumed it would be about numbers and/or perhaps indexing,
or sequence access control, etc., but certainly not that its ONLY purpose
was representing a custom boolean value.

IOW, I should have RTFM better, but I think TFM could have used less misleading
names, and I do think __nonzero__ would be better named __bool__. At least that
would have snagged my attention in a different way.

>> for something where I wanted to be able to write
>> 
>>     if s: print '"%s" (length %d) is special.' % (s, len(s))
>> 
>> I.e., in a logical context( if x..., x and ..., etc.), I think
>> x.__bool__() should have priority over x.__nonzero__() or x.__len__().
>
>__nonzero__ is used for nothing else except obtaining the object's
>truth value.  You don't appear to understand this...?
>
You're right. I didn't. The name totally fooled me when I first saw it.
I assumed that it had some unknown numeric use that might be broken if
I wrote a kludge to misuse it as a bool. At the time, my normal curiosity
was busy with something else, so I didn't look at the implementation
(or even the official doc ;-/) to check my assumption. I really should
have done better homework before talking about it, I grant you.

Now __nonzero__ is specified to return 0 or 1 as the "truth value."
I assume that will change with PEP 285? But whether True/False or 1/0, surely
__nonzero__ suggests that what is true or not is whether the thing is zero or not.

I think 'paper jammed' or 'connected' or 'tainted' etc. don't fit well with "nonzero".
That may seem a minor quibble, but ISTM, not unrelated to why Guido wants a bool
type in the first place.

>>>I certainly don't object to implementing the bool type.  I guess I
>>>just don't see it as a significant advantage.
>> 
>> I'd like the semantic clarity. And I'd like to define a class or subtype
>> whose boolean logical value doesn't depend on anyones concept of Nothing
>> or Empty if those don't fit. 
>
>You CAN do that, TODAY.  Therefore, your major motivation for
>desiring booleans seems to be an insufficient knowledge of Python/
>
I can do part, that's true. But I can't have the semantic clarity of returning
an official bool whose str value will be True or False, or something that I can
control through subtyping bool. And I can't do it through an interface whose name
reflects the general bool that it's delivering. I can build my own world that does
this, but that's not the same as using official infrastructure. So saying I CAN do
that TODAY is focusing on a narrow aspect.

I did mistake __nonzero__ for something I shouldn't use arbitrarily for my own purposes.
I thought from the name that __nonzero__ depended on someone's concept of zero and
nonzero, and I felt that returning a boolean through __nonzero__ would not be appropriate
unless my class had a state that could meaningfully be described as zero or non-zero.

That's why I was trying to propose something that ties in better with bool and bool(x).
What better than __bool__ ?

But thank you for pushing me to RTFM:
"""
__nonzero__(self) 
Called to implement truth value testing; should return 0 or 1.
When this method is not defined, __len__() is called, if it is
defined (see below). If a class defines neither __len__() nor __nonzero__(),
all its instances are considered true.  
"""

I am assuming, but haven't seen explicitly in the PEP, that bool(x) will result
in bool(x.__nonzero__()) if the __nonzero__ method is defined. Unfortunately
(ISTM), the code spec seems to preclude use of bool subtypes via x.__nonzero__().
IOW, str(bool(x)) will be limited to False/True, because bool(x) always returns
a singleton of the base class, never passing through possible compatible subtypes
(however unusual a constructor that might be ;-)

            def __new__(cls, val=0):
                      # This constructor always returns an existing instance
                      if val:
                          return True
                      else:
                          return False

I'd like to do something like the following (though this does not work, apparently
for several reasons, so please take it as pseudo-code to convey the idea):

            def __new__(cls, val=0):
                      # This constructor always returns an existing instance
                      try:
                          vbool = val.__nonzero__() # will crash if val is an integer[1]
                      except:
                          vbool = val
                      if isinstance(vbool, bool):
                          return vbool
                      elif vbool:
                          return True
                      else:
                          return False
 
[1] More directly illustrated by the following, which seems[2] like a bug:

 Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
 Type "help", "copyright", "credits" or "license" for more information.
 >>> hasattr(1, '__nonzero__')
 1
 >>> 1 .__nonzero__()

(You get a memory fault).

[2] I now know it's a known bug, and fixed in cvs.

>> use when I poll. Or take a whimsical example: What if I want to subclass
>> int and be able to say if n: print n,'is a prime', expecting that
>> n.__bool__() will be called so I can calculate primeness and cache the
>> result in a boolean attribute for faster return next time, etc.
>
>You CAN do that, TODAY.  Whether it's wise or not is another issue,
>but, for example:
>
>class intWithPredicate(int):
>
>    def __init__(self, value=0, predicate=None):
>        int.__init__(self, value)
>        self.setPredicate(predicate)
>
>    def setPredicate(self, predicate=None):
>        self.predicate = predicate
>        self.isResultKnown = 0
>
>    def __nonzero__(self):
>        if self.isResultKnown: 
>            return self.result
>        if self.predicate is None:
>            self.result = self != 0
>        else:
>            self.result = self.predicate(self)
>        self.isResultKnown = 1
>        return self.result
>
The predicate generalization is a nice touch. And your example demonstrates
that I can do part of what I'd like now. Now can you also get the result to
show up when you print bool(x) and result is a subtype of bool
(which I don't think is possible with the current bool definiton, but that's
what I'd like). (It would be a strange constructor, but how else to get
a bool subtype from an object via its __nonzero__ method? It would be ugly to
write print bool(x) for ordinary True/False, and print y.__nonzero__() for
bool subtypes, and not be able to write [str(bool(x)) for x in objlist]
without losing custom bool strs.

>You don't need any change to Python to implement this functionality,
>whether it makes sense or not to have the functionality in question.
>
You captured a good part of what I'd like to do, but not all of it.
>
>> The point is to have freedom to use a useful general mechanism without
>
>The point is you HAVE the freedom, NOW -- the mechanism is
>general and there are no semantic contortions.
Well, stretching the meaning of nonzero to mean true is a semantic contortion, IMO ;-)

BTW, I think it might be helpful to mention in the PEP how bool(x) relates to __nonzero__
and __len__.  I expect that it will be ok for __nonzero__ to return 0/1 or False/True and
that all will be normalized to False/True in any case.

>
>> semantic contortions, or giving up the ability if __nonzero__ and/or
>> __len__ already have uses for the class.
>
>What "use for the class" do you expect __nonzero__ to have, as it
>is ONLY used to define the object's truth value?!
>
Well, I thought it might be used to indicate that there was something
non-zero about an object. I wasn't willing to assume that there was no
such independent number-specific use, given the suggestiveness (to me)
of the name. But now I know better ;-)

Regards,
Bengt Richter


From pinard at iro.umontreal.ca  Fri Apr 12 21:48:25 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 12 Apr 2002 21:48:25 -0400
Subject: Forcing quote characters in repr(STRING), how?
In-Reply-To: <Xns90968DA40B8A3duncanrcpcouk@127.0.0.1>
References: <mailman.988841104.7613.python-list@python.org>
	<Xns90968DA40B8A3duncanrcpcouk@127.0.0.1>
Message-ID: <oqit6w2wt2.fsf@titan.progiciels-bpi.ca>

Hi!  Cleaning old mail, I stumbled on that old conversation:

---------------------------------------------------------------------->
[Fran?ois Pinard]

> When using `repr()' on a string, Python automatically selects if it
> will use single or double quotes to enclose the produced representation
> of the string.  Is there a clean way to force that choice to be, under
> user control, a particular quote character (either simple or double)?

[Tim Peters]

> Not a clean way, nor even a dirty way:  the logic is hardcoded.

[Duncan Booth]

> Never say there isn't a dirty way.  The following works, although I
> really wouldn't recommend it: [...]

> def quote(s, dquote=0):
>     """Return repr of the string argument forcing single or double 
> quotes."""
>     if dquote:
>         nasty="'\0"
>     else:
>         nasty='"\0'
>     r = repr(nasty+s)
>     bad = len(`nasty`)-1
>     r = r[0]+r[bad:]
>     return r

> if __name__=='__main__':
>     def test(s):
>         print "single quote(%s) -> %s" % (`s`, quote(s, 0))
>         print "double quote(%s) -> %s" % (`s`, quote(s, 1))
>     test("ab'cd")
>     test("ab'c\"d")
>     test('ab"cd')
----------------------------------------------------------------------<

Despite not recommended, I nevertheless followed the trick above in Pymacs.
Preceded by a long comment, the little piece of code looks almost clean :-).

    [...]
    elif type(value) == types.StringType:
        # Python delimits a string it by single quotes preferably, unless
        # single quotes appear within the string while double quotes do
        # not, in which case it uses double quotes for string delimiters.
        # Checking the string contents, the C code stops at the first NUL.
        # We prefix the string with a single quote and a NUL, this forces
        # double quotes as delimiters for the whole prefixed string.  Then,
        # we get rid of the representation of the single quote and the NUL.
        write('"' + repr("'\0" + value)[6:])
    [...]

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




From andrewm at object-craft.com.au  Wed Apr  3 21:44:10 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Thu, 04 Apr 2002 12:44:10 +1000
Subject: OK 
In-Reply-To: Your message of "03 Apr 2002 18:16:14 PST."
             <e8d5abee.0204031816.2bfec035@posting.google.com> 
Message-ID: <20020404024410.91D5138F50@coffee.object-craft.com.au>

>Is python out of fashion?

No, but posting trolls is.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/




From news1 at ns2.freenix.org  Mon Apr 15 11:31:56 2002
From: news1 at ns2.freenix.org (Olivier Ricou)
Date: 15 Apr 2002 15:31:56 GMT
Subject: tcl/tk panel in tkinter
Message-ID: <a9erpc$1nk$1@wanadoo.fr>

Hi,

 I have a tcl/tk program which I would like to include in a panel of
my python/tkinter application. Of course I don't want to redo the
all tcl/tk program and I want that the 2 programs communicate. 

 To fix the idea, imagine that the tcl/tk program do computation when 
you give it data thru its tk interface and the python/tkinter use the
results of the tcl/tk program to do draws. So I want to have a
tcl/tk panel from which I give the data and another panel on which
I have the draws in the same python/tkinter application.

Is there is an easy way to do that ? Does someone have an example ?

					Regards,

					   Olivier.


From aahz at pythoncraft.com  Sat Apr 27 10:55:00 2002
From: aahz at pythoncraft.com (Aahz)
Date: 27 Apr 2002 10:55:00 -0400
Subject: Stackless goes Limbo
References: <mailman.1019911009.16159.python-list@python.org>
Message-ID: <aaee44$p73$1@panix1.panix.com>

In article <mailman.1019911009.16159.python-list at python.org>,
Laura Creighton  <lac at strakt.com> wrote:
>
>For those of you who can't live without a Limbo introduction written
>by a real C inventor:
>
>http://www.vitanuova.com/inferno/papers/limbo.html

Ugh.  I've gotten so corrupted by Python I can't even read well-written
C code.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From jeff at jonesj.com  Fri Apr 19 17:48:12 2002
From: jeff at jonesj.com (Jeff Jones)
Date: Fri, 19 Apr 2002 16:48:12 -0500
Subject: Substitution with Hash Values
Message-ID: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com>

I am a long time perl user who is trying to get a grip on python.  I
really like what I have seen so far, but I have hit a little snag
regarding substitution (like perl).

I use html templates for all of my CGI programming, and everything that needs
to be replaced is contained within double curly braces ( {{value}} ).  In
perl I use a hash to keep all of the variables that are going to replace
the placeholders in the template, and replace them the following way:

	....
	s/{{(.*?)}}/$hash{$1}/g;
	....

That substiution is accomplished as I read in the template file.

Can anyone tell me how I might do this in python?  I have tried many
different ways, but none seem to work.

Thanks.

Jeff Jones


From mwh at python.net  Fri Apr 12 06:37:00 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 10:37:00 GMT
Subject: Problem in 'Doc' building in 2.2.1 (OK in 2.2.1c2)
References: <Pine.GSO.4.44.0204120620180.13652-100000@step.polymtl.ca>
Message-ID: <lkzo09i1om.fsf@pc150.maths.bris.ac.uk>

Vincent Cojot <coyote at step.polymtl.ca> writes:

> Hello everyone,
> 
> 	Python 2.2.1 (release) appears to be missing something in its
> documentation, thereby preventing a sucessfull build of the Documentation:
> 
> [root at toffee Doc]# gmake -n
> gmake: *** No rule to make target `whatsnew/whatsnew22.tex', needed by
> `html/whatsnew/whatsnew22.html'.  Stop.
> 
> This problem isn't present in 2.2.1c2 because the Make dependency for
> the whatsnew stuff isn't there..

Yes, this was me finding out that you have to write "update" in your
.cvsrc file -- I had "up -dP" in there, but that's not enough (it
turns out).

Sorry.  It's possible that we'll release a new tarball to amend this,
but not certain.

> Additionnally, is there -anything- that helps latex2html -not- to complain
> about "dots" in the command name:

latex2html?  Brrr.  I don't know.

Cheers,
M.

-- 
  Hey, if I thought I was wrong, I'd change my mind.  :)
                                    -- Grant Edwards, comp.lang.python


From unendliche at hanmail.net  Sat Apr 13 08:49:17 2002
From: unendliche at hanmail.net (Seo Sanghyeon)
Date: 13 Apr 2002 05:49:17 -0700
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4>
Message-ID: <45e6545c.0204130449.d6cfd69@posting.google.com>

Philip Swartzleonard <starx at pacbell.net> wrote

> The first section of the index[1] for the language reference manual lists
> all of these things (the underscore section), with links.

Thank you very much for pointers, but I can think of something not listed
there. For example, __all__ for "from module import *", pickle protocols,
__new__ for subclassing from a type and so on. Is there a complete (or
near-complete) list?


From ark at research.att.com  Wed Apr 10 11:52:23 2002
From: ark at research.att.com (Andrew Koenig)
Date: Wed, 10 Apr 2002 15:52:23 GMT
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com> <1018450381.103820@cswreg.cos.agilent.com>
Message-ID: <yu991ydnmtzc.fsf@europa.research.att.com>

Greg> Andrew Koenig (ark at research.att.com) wrote:

Greg> : My motivation is to be able to write something that is analogous
Greg> : to call-by-reference in C++.

Greg> *Not* having call-by-reference means that you can read code without
Greg> wondering whether a particular function call is or is not by reference.  I
Greg> consider this a blessing, and I'd like to think that the developers of
Greg> Lisp, C, Java, and Python feel the same way.

Of course, C arrays behave as if they are pass-by-reference
(only!), even though we both know that's not strictly true
under the skin.

Anyway, I would be quite content even if the references were
explicit at the point of call, which would meet your objection.
And there is even a straightforward way to achieve that state
now:  Pass a pair of get/set functions instead of the variable
itself.

So in fact there is a way to do what I want; it's just that it's
more verbose than I'd like.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From philh at comuno.freeserve.co.uk  Wed Apr 17 08:31:57 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 17 Apr 2002 13:31:57 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com> <3CBC5B6D.70427378@noaa.gov>
Message-ID: <slrnabqqpt.ll2.philh@comuno.freeserve.co.uk>

On Tue, 16 Apr 2002 10:12:14 -0700, Chris Barker <Chris.Barker at noaa.gov> wrote:
>maxx at easynerws.com wrote:
>> Why?
>> 
>> At my company we are stuck in the realm of Windows, and the IT staff must
>> constantly patch the webservers to keep up with security and bugfixes. A quick
>> check of the IT media will show management's displeasure over this continual
>> process, but it does not appear to have hampered the deployment of Windows
>> servers.
>
>If Python had Microsoft's other advantages in the Marketplace, it would
>be a Whole lot more popular. Imagine if Microsoft decided to drop VBasic
>and start selling MSPython!

It'd bprobably be incompatible, with lots of features designed to 
lock you in to the MS world.

>> As mentioned in elsewhere this thread, Java and VB, have and are undergoing
>> complete architecture restructuring, and they still are expanding their
>> territories.
>
>Again, with the marketing put into JAVA, Python would be very popular indeed.

Quite likely. 


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From loewis at informatik.hu-berlin.de  Tue Apr 16 05:29:47 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 16 Apr 2002 11:29:47 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018875990.21271.python-list@python.org>
Message-ID: <j43cxwm1o4.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

> > But this is really strange; mathmodule.so must have imported at some
> > point, or setup.py would have deleted it...
> 
> Hmm.  I don't know.  Anyway, here is the result of re-making the
> Python, but not the make test, after removing that file...
> 
> Script started on Mon Apr 15 13:57:01 2002
> neelix hgs 1 %> gmake
> gcc -shared  Modules/mathmodule.o   -o Modules/mathmodule.so

That explains it: he has enabled mathmodule.c in the *shared* section
of Setup.

I recommend to build Python from original sources, in a fresh
directory.

Regards,
Martin


From matthewm at zebrasoft.co.nz  Mon Apr 29 03:53:29 2002
From: matthewm at zebrasoft.co.nz (Matthew)
Date: Mon, 29 Apr 2002 19:53:29 +1200
Subject: Returning variable results from a function
Message-ID: <3CCCFBF9.2908C052@zebrasoft.co.nz>

Hi

Given something like:-

def fred(size):
    return size, 1,2,3

size,a,b,c=fred(3)

how does one create a tuple to hold the results like:-

size, (<a tuple holding the number of results?>) = fred(<no of
results?>)

Thanks so much, matthew.




From senux at senux.com  Wed Apr  3 08:16:17 2002
From: senux at senux.com (Brian Lee)
Date: Wed, 3 Apr 2002 22:16:17 +0900
Subject: audio stream server by python?
Message-ID: <20020403131617.GA6755@jupiter.senux.com>

Hi, is there any simple code to transfer audio stream data 
to remote host by Python?

-- 
 ____       
 |o | i   / Brian Lee    <senux at senux.com>
 #######|<  Web - http://www.senux.com/en/
(x-x-x-x) \ GnuPG public key - 0x46C763A3



From neal at metaslash.com  Wed Apr 24 13:38:29 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Wed, 24 Apr 2002 13:38:29 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <3CC6ED95.569DA16A@metaslash.com>

Learning C is probably not a good way to go given your situation.
Python is questionable.  Depends a lot on what you really need to do.

To give you an idea of Python's speed, I was able to write out 600MB
of data in 450 seconds.  I could read the data, modify it, and write
it back out in 700 seconds.

By box is 650 Mhz Athlon, 256 MB RAM:

time to write  input file: 450.9 seconds
time to modify input file: 702.2 seconds
first line of output file: A B C
file sizes (input/output): 600000000/600000000

The program I used is below.

Neal
--
import os
import time

BIG_MOMMA = 100*1000*1000
IFILENAME = '/home/neal/build/bigfile'
OFILENAME = IFILENAME + '.out'

def create_file(filename, count):
    f = open(filename, 'a')
    for i in xrange(count):
        f.write('a b c\n')
    f.close()

def modify_file(ifilename, ofilename):
    ifile = open(ifilename)
    ofile = open(ofilename, 'w+')
    for x in ifile:
        ofile.write(x.upper())
    ifile.close()
    ofile.close()

def main():
    start = time.time()
    create_file(IFILENAME, BIG_MOMMA)
    print 'time to write  input file: %.1f seconds' % (time.time() - start)

    start = time.time()
    modify_file(IFILENAME, OFILENAME)
    print 'time to modify input file: %.1f seconds' % (time.time() - start)

    f = open(OFILENAME)
    print 'first line of output file:', f.readline(),
    f.close()

    print 'file sizes (input/output): %d/%d' % \
                    (os.stat(IFILENAME)[6], os.stat(OFILENAME)[6])

    os.unlink(IFILENAME)
    os.unlink(OFILENAME)

main()


From robin at jessikat.fsnet.co.uk  Mon Apr 15 04:07:58 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 15 Apr 2002 09:07:58 +0100
Subject: Python 2.2 unwise.exe
References: <A7B5C22F20EE3C4E95744E2F03F1887F5640@MAYOR.retzlaff.com>
 <mailman.1018827089.1679.python-list@python.org>
Message-ID: <880uYVAepou8Ewdy@jessikat.fsnet.co.uk>

In article <mailman.1018827089.1679.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
.....
>I have noticed that unwise.exe is at least much more likely to vanish if
>Win98 crashes between the time I do an install and get around to doing a
>normal reboot.  That suggests Wise is doing something special with it --
>although I've no idea what (or, really, even "if").
.....
that would be consistent with my case as my home machine suffered a TV
video card induced hard crash (with file-system check) during the
relatively short period between installing 2.2 and attempting the
deinstall.


Googling on missing unwise32 I notice one report from 1995 about
unwise32 being virus infected. I doubt that's the case, but I run an
automatic virus checker and if unwise32 looks like a virus perhaps it's
being quarantined or something.
-Explicit nonsense is not better than implicit sense-ly yrs- 
Robin Becker


From l0819m0v0smfm001 at sneakemail.com  Sat Apr  6 18:41:53 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Sat, 06 Apr 2002 23:41:53 GMT
Subject: yet another "Here's to 1.5.2" post
References: <1018131203.756898@cswreg.cos.agilent.com>
Message-ID: <3CAF4E35.3050806@sneakemail.com>

Greg Weeks wrote:

> Now, if only the energy that was put into adding the above blemishes had
> been placed into making Python more robustly installable.  On my system
> (HP-UX 10.2) with my reasonable efforts at building a system:
> 
>     in 1.5.2, there is no FCNTL module
> 
>     in 2.2, "configure" fails due to erroneous C pre-processor code
> 
>     in 2.2, "make test" fails for reasons I haven't had the time to pin down
> 

   Instead of volunteering other people's time and energy, maybe you 
should contribute some of your own.  If fixing "make test" and 
"configure" on your HP-UX system isn't high enough priority for *you* 
who actually uses it, why should anyone else leap to do your bidding? 
Particularly when you don't even seem to be thankful for the software 
that's already been donated to you by their efforts....

   Joshua



From max at alcyone.com  Mon Apr  1 17:11:02 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 01 Apr 2002 14:11:02 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com> <3CA8E312.78175D08@ndh.net>
Message-ID: <3CA8DAF6.4C6B2D51@alcyone.com>

Stefan Schwarzer wrote:

> I would like to have bools in Python, rather than not. But only if
> bools and ints are different types, so that "True == 1" would be
> false,
> similar to that "'1' == 1" is false. And _my_ conclusion was, that I
> would rather have no bool type than a mixed-up concept in which bools
> are really ints.

Does it bother you that 1.0 == 1, or that 1L == 1?  I certainly see the
argument that a Boolean should not be an integer, but with the current
PEP -- and to have any hope of maintaining some backward compatibility
-- a bool must be a degenerate integer, and so True == 1 makes sense. 
If you consider this evil (which I don't), then it's a necessary evil. 
Some compromises must be made for the sake of backward compatibility.

I'm not sure I see the big problem, since explicit comparisons to true
are rarely useful anyway.  i.e., this is another example of code you
_can_ write, but not necessarily code you _should_.  No matter what the
language, there is always accounting for poor programmers and bad style;
Python is no exception.

Python is not even alone in this category, other languages have (what
amounts to) implicit conversion from Booleans to integers.  

> What's the point in saying "There should be a bool type to distinguish
> between truth values and integers" when this distinction is so
> shallow?

Because the purpose is for distinguishing between _intent_ and not
value.  If I have a function that returns 0, I don't know whether that's
a count, or a Boolean, or a trinary value (say, 1, 0, and -1, like cmp),
or otherwise.  If I see one that returns False, I damn sure know what's
going on.  The inclusion of Booleans allows one to write better
self-documented code, because it indicates what the programmers is
intending to do with values, not just the values.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From gregLANDRUM at mindspring.com  Tue Apr  9 12:15:47 2002
From: gregLANDRUM at mindspring.com (Greg Landrum)
Date: Tue, 9 Apr 2002 09:15:47 -0700
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <a8v43b$nes$1@slb0.atl.mindspring.net>

I used to do stuff in Tkinter, then I switched to wxPython because it looked
better, was faster, and had the Grid (spreadsheet-like) widget.

wxPython has a lot of functionality and, once you know how to use it, is
mostly pretty easy to use.  As others have said, the documentation is in a
sorry state.  The wxPython docs are slight edits of the C++ docs, which are
incomplete (at best).  Last time I checked the new Grid widget (the reason I
switched) still didn't have up-to-date documentation (more than a year after
its release).  This is somewhat mitigated by the wonderfully supportive and
open mailing list, but only somewhat.

Add to this the fact that the API has changed (in code-breaking ways) from
version to version and you get an unhappy user (me).  Despite the advanced
version numbers (2.2.x and 2.3.x), I never escaped from the feeling that I
was using beta-quality code.

Please bear in mind that many of my criticisms (particularly the ones about
documentation) are due to wx itself, not wxPython.

I am currently looking at qt as a replacement.  The underlying C++ is well
documented and supported.  The Python bindings are definitely still
beta-level (maybe even alpha on windows), but they are at least building on
a stable and capable core.

-greg





From tim.one at comcast.net  Sun Apr 21 23:50:55 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 21 Apr 2002 23:50:55 -0400
Subject: A simply newbie question about ndiff
In-Reply-To: <3cc3765f$0$15478$afc38c87@news.optusnet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPAOPAA.tim.one@comcast.net>

[Neville Franks]
> Ok thanks. I've installed 2.2 and ndiff is available but I'm
> still confused.  I thought that Tim's original ndiff compared files,
> but the one in difflib compares lists of strings.

Tools/scripts/ndiff.py still exists, and still compares files.  If that's
all you want to do, just run it -- no need to program anything.  Or if you
waht to see how to use difflib to do it yourself, again
Tools/scripts/ndiff.py is the answer:  in Python 2.2, ndiff.py is a tiny
program that invokes difflib facilities to do all the hard work.

> Does this mean I somehow need to read the files into lists of strings
> and then call ndiff. A short example of how to use ndiff to compare two
> files would be greatly appreciated.

See function fcompare() in 2.2's ndiff.py.  The part that compares and
prints is 4 lines, including reading the files <wink>:

    a = f1.readlines(); f1.close()
    b = f2.readlines(); f2.close()
    for line in difflib.ndiff(a, b):
        print line,





From akuchlin at crystal.mems-exchange.org  Thu Apr 11 11:14:58 2002
From: akuchlin at crystal.mems-exchange.org (A.M. Kuchling)
Date: 11 Apr 2002 15:14:58 GMT
Subject: Notes on 2.2's changes
Message-ID: <slrnabba3h.h8h.akuchlin@crystal.mems-exchange.org>

I'd like to provide a point of view on how many changes 2.2 required
in a sizable amount of code.

The MEMS Exchange codebase I work with is, as of today, 171,437 lines
of Python code.  Looking through the CVS logs, I find the following
messages relating to Python 2.2:

In Grouch:
===============
Tweak is_instance_object to (hopefully) work with Python 2.2.

Fixed InstanceType._check_attrs() so it doesn't blow up on encountering
an instance with no __dict__.  (vars() in Python 2.2 raises an exception
in this situation.)

Modified AtomicType so the 'type' attribute can be None until it's
actually needed, ie. in get_type() and check_value().  This lets me
completely jettison most of the complicated pickle-support code -- no
more _fixup_atomic_types() methods, and (more importantly) no more tests
failing under Python 2.2.  (The code was equally broken under 2.1, it
was just hidden because of blind luck; a change in dictionary order
revealed it under 2.2.)
 
Fix InstanceType._check_attrs() to use vars() instead of dir() to fetch
the list of actual instance attributes.  This fixes one of the
incompatibilities between Grouch and Python 2.2.

Split a couple of tests up so that failures (caused by type name changes
in Python 2.2) are easier to read.
===============

In Quixote's:
===============
(ptl_compile.py)
Support Python 2.2.  (Changes the .ptlc magic number, and adds a 
                      version-specific check for a change to the
                      compiler package.) 

doc/INSTALL:
Complicate instructions for installing the compiler package, since
doing so is only necessary for 2.0 <= Python < 2.2.
Simplify installation instructions for Quixote itself.
===============

In the rest of our application code:
===============
Rename __debug__ to DEBUG for 2.2 compat. 

Removed private copy of asyncore.py; 2.2 contains the needed fixes
===============

And that's it.  Only one application code change, the __debug__ one,
would have actually prevented code from running with a SyntaxError,
and that's only because the file was only previously run under 1.5.2
(so it was going straight from 1.5.2 to 2.2, missing the intermediate
releases where assigning to __debug__ caused a warning). The Quixote
change was to the module that parses PTL and outputs Python bytecode
from the parse tree.  Grouch had the most changes, but it does
introspection on Python objects and types, so it's not surprising that
2.2 caused some upheaval.

Looking at the changes required for 2.0 and 2.1, Quixote again took
the most work.  The actual application code had one or two
multi-argument .append() calls, and that was the extent of the fixes.

This means that from my point of view Python *has* been managed for
stability, and the job has been done reasonably well.  (I'd be
interested in seeing a similar list from people who find there's too
much change.  It would be useful for figuring out what sorts of
changes cause the most problems.)

--amk                                                             (www.amk.ca)
I tolerate this century but I don't enjoy it.
    -- The Doctor, in "An Unearthly Child"


From nobody at nobody.com  Mon Apr  8 00:55:47 2002
From: nobody at nobody.com (John)
Date: Sun, 7 Apr 2002 23:55:47 -0500
Subject: Function question
Message-ID: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu>

Hi I have a question:

Let's say there is a function with two arguments.
def add(a,b): return a+b

What I want to do is:
I want to get 'a' dynamically..
But once a value is decided, I want a separate function (say, new_add) that
does
new_add(b) (equals to) add(a,b)

That is, I want the function definition to be dynamic.

Would that be possible? Any comment would be greatly appreciated.




From see at below  Sun Apr 28 23:08:38 2002
From: see at below (Paul Foley)
Date: 29 Apr 2002 15:08:38 +1200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s7el$sr7$1@slb2.atl.mindspring.net> <pan.2002.04.28.00.20.56.555436.3624@uol.com.br> <aahhtq$7lq$1@slb4.atl.mindspring.net>
Message-ID: <m2d6wjw6a1.fsf@mycroft.actrix.gen.nz>

On Sun, 28 Apr 2002 13:18:49 -0600, Andrew Dalke wrote:

> I'm not sure what your point is here.  In more recent Pythons (1.5 and
> newer, I think) you can do

> >>> class Spam:
> ...  def what(self): print "I'm spam"
> ...
> >>> class Eggs:
> ...  def what(self): print "I'm eggs"
> ...
> >>> x = Spam()
> >>> x.what()
> I'm spam
> >>> x.__class__ = Eggs
> >>> x.what()
> I'm eggs

OK, but now try this:

  _spams = WeakList()
  _eggs = WeakList()
  
  class Spam:
     def __init__(self):
         self.y = 42
         _spams.append(self)

  class Eggs:
     def __init__(self):
         self.z = 17
         _eggs.append(self)

  x = Spam()
  x.__class__ = Eggs


Problems: x.z isn't present, x.y is present but unwanted, and the
instance is a member of _spams, not of _eggs.

You changed the class, but you didn't do anything to ensure that the
object is actually a _valid_ instance of the new class.

In CL, (change-class x 'eggs) can do whatever needs to be done to make
x into a valid egg (there are various generic functions invoked along
the way that you can write methods on to make it do the right thing;
the main one being update-instance-for-different-class, which chooses
the appropriate method to run based on both the old and new classes)

-- 
In modern physics, the more logical you are, the more wrong you are.
                                                           -- Lama Govinda
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From BgPorter at NOartlogicSPAM.com  Mon Apr 22 09:37:39 2002
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Mon, 22 Apr 2002 13:37:39 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz>
Message-ID: <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com>

"Paul Foley" <see at below> wrote in message
news:m2wuv0xr1u.fsf at mycroft.actrix.gen.nz...
> On Sun, 21 Apr 2002 02:32:29 -0700, James J Besemer wrote:
>
> > That being said, I guess I just don't get the strong reaction to this
secondary
> > point.  I'm new here but I get the impression that "lisp" is a naughty
word,
> > for some unfathomable reason
>
> A lot of people think "Lisp" means "Scheme".
> Scheme sucks, so they think "Lisp" sucks.
>
> [Scheme isn't Lisp.  Scheme are Lisp are only distantly related, and
> Scheme is about as closely related to Algol as to Lisp...]

I'd be interested to learn more about this -- the only ref I have in hand is
Abelson & Sussman * 2's "Structure and Interpretation of Computer Programs",
which sez (on p 3 of my hard copy, but the link here is to the online
version):

http://mitpress.mit.edu/sicp/full-text/book/book-Z-H-9.html#%_chap_1
"Lisp was not the product of a concerted design effort. Instead, it evolved
informally in an experimental manner in response to users' needs and to
pragmatic implementation considerations. Lisp's informal evolution has
continued through the years, and the community of Lisp users has
traditionally resisted attempts to promulgate any ``official'' definition of
the language. This evolution, together with the flexibility and elegance of
the initial conception, has enabled Lisp, which is the second oldest
language in widespread use today (only Fortran is older), to continually
adapt to encompass the most modern ideas about program design. Thus, Lisp is
by now a family of dialects, which, while sharing most of the original
features, may differ from one another in significant ways. The dialect of
Lisp used in this book is called Scheme"

In what ways are Scheme and Lisp only distantly related?

--
// Brett g Porter * Lead Engineer, Development Practices
// BgPorter at artlogic.com * www.artlogic.com
// Art & Logic, Inc. * software engineering and design
// Desktop * Embedded * Web




From cepl.m at neu.edu  Fri Apr 26 21:41:29 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Sat, 27 Apr 2002 01:41:29 +0000 (UTC)
Subject: iis -- IMAP to IMAP Synchronization [Was: Re: comparsion of two lists]
In-Reply-To: <16my8.10553$Oq2.1509222@news1.west.cox.net>
References: <20020426224924.GA1318@rehor.surfbest.net> <16my8.10553$Oq2.1509222@news1.west.cox.net>
Message-ID: <20020427014126.GA2517@rehor.surfbest.net>

On Sat, Apr 27, 2002 at 12:17:01AM +0000, Van Gale wrote:
> "Matej Cepl" wrote:
> > I would like to compare two
> > list of messages (in the local and remote folders) and create
> > list of messages, to be deleted/downloaded. In plain Python, the
> > part of deletion would go like this:

OK, I have created ALPHA version of my script!! Yeah (my first
big script in Python :-)!!!

However, because the script should work with my real emails, I am 
rather concerned, whether I have not made some stupid mistake. 
Would you be so kind and review the attached script and comment 
on it (especially safety of my emails and speed), please?

Also, do you know about any website, where I should post the 
script, so that it would available to everybody?

Hopefully, it might be usefull for other people as well.

	Thanks a lot
	
		Matej Cepl

-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
They that can give up essential liberty to obtain a little
temporary safety deserve neither liberty nor safety.
	-- Benjamin Franklin, Historical Review
	   of Pennsylvania, 1759.

-------------- next part --------------
#!/usr/bin/env python
# $Id: iis.py,v 0.7 2002/04/27 01:32:52 matej Exp $
"""
IMAP-IMAP Synchronization - iis

(always use lower case to avoid terrible misunderstanding :-)

Copyright (c) 2002 Matej Cepl <matej at ceplovi.cz>
   
Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following
conditions:

The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT.  IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS IN THE SOFTWARE.

"""
import ConfigParser
import imaplib
from sys import stderr, exit
from string import split, join
from os.path import expanduser
from time import time, gmtime

# constants
# return errorlevels
DOWNERR = 1
UPERR = 2
SELECTERR = 3
LISTERR = 4
LOGINERR = 5
CONFIGERR = 10

# other constants
IMAPPORT = "143"
LOCALSTR = 'local'

def chatch_error(errmsg,stat_num):
	stderr.write("IMAP Error: %s !\n" % errmsg)
	exit(stat_num)

class Conifg(ConfigParser.ConfigParser):
	"""
	Object containing configuration.
	
	It reads configuration file (in WIN .ini syntax) like this
	
	[localuser1]
	local = password,last_sync
	server1.com = username,password,last_sync (in IMAP shape)
	
	[localuser2]
	local = password,last_sync
	server2.com = username,password,last_sync (in IMAP shape)
	"""
	def __init__(self,fname=".iisrc"):
		ConfigParser.ConfigParser.__ini__(self)
		self.filename = expanduser("~/"+fname)
		self.options = {}
		try:
			self.read(self.filename)
		except ParsingError,e:
			chatch_error(e,CONFIGERR)
		for sect in self.sections():
			tmp_dict = {}
			for opt in self.options(sect)[:-1]:
				tmp_dict[opt] = split(self.get(sect,opt),',')
			self.options[sect] = tmp_dict

	def close(self)
		try:
			f = open(self.filename,'w')
			for sect in self.options.keys():
				f.write('[' + sect + ']\n')
				for opt in self.options[sect].keys():
					f.write(opt + '=')
					f.write(self.optionsp[sect][opt] + '\n')
			f.close()

def comp_lists(processed,local):
	"""
	Compare two lists of messages and select for download/delete.

	INPUT: two objects Server -- one to be processed and other
	OUTPUT: erase, download -- list of messages to delete from
			  the processed server, and list of messages to be
			  downloaded on other server
	What is newer than self.last and it isn't in other, should be
	downloaded from processed, and what is older and isn't on
	other, should be deleted. The same other way around.

	OK, I admitt, that code below is ugly, but it should be as
	fast as possible -- which is what we need.
	"""
	proc_old = map(lambda x: processed.old_msg[x][0],processed.old_msg.keys())
	proc_new = map(lambda x: processed.new_msg[x][0],processed.new_msg.keys())
	loc_old = map(lambda x: local.old_msg[x][0],local.old_msg.keys())
	loc_new = map(lambda x: local.new_msg[x][0],local.new_msg.keys())
	
	erase = e = download = d = ()
	e = filter(lambda m: return not(m in loc_old), proc_old)
	d = filter(lambda m: return not(m in loc_new), proc_new)

	erase = filter(lambda x: processed.old_msg[x][0] in e, \
	processed.old_msg.keys())
	download = filter(lambda x: processed.new_msg[x][0] in d, \
	processed.new_msg.keys())

	return erase,download

def sync(loc,rem):
	"""
	Synchronize two folders on servers.

	## INPUT:
	## OUTPUT:
	## description
	"""
	# delete old on the remote and download new messages
	delete,down = comp_lists(rem,loc)
	try:
		ok, answer = rem.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Remote messages were not deleted!\n")
		# rather let it be -- it is better not to have
		# something deleted, than to have deleted more than
		# we want
	for msg in down:
		try:
			ok, stuff = rem.fetch(msg,'(RFC822)')
			tmp_msg = rem.new_msg[msg]
			ok, answer = loc.append(loc.curr_folder,tmp_msg[2],tmp_msg[1],stuff)
		except Exception,e:
			catch_error(e,DOWNERR)
	
	# delete old on local and upload new ones
	delete,up = comp_lists(loc,rem)
	try:
		ok, answer = loc.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Local messages were not deleted!\n")
	for msg in up:
		try:
			stuff = loc.fetch(msg,'(RFC822)')
			tmp_msg = loc.old_msg[msg]
			ok, answer = rem.append(rem.curr_folder,tmp_msg[2],tmp_msg[1],stuff)
		except Exception,e:
			catch_error(e,UPERR)

class Server(imaplib.IMAP4):
	"""
	Main class for managing IMAP servers and folders.
	
	## some description
	"""
	def __init__(self,server='',port=IMAPPORT,user,password,last_sync):
		"""
		Connect to the server and login.
		"""
		self.host=server
		self.port=port
		self.user=user
		self.passwd=password
		self.last=last_sync
		self.new_msg={}
		self.old_msg={}
		self.curr_folder=''
		try:
			imaplib.IMAP4.__init__(self,self.host,self.port)
			self.login(self.user,self.passwd)
		except Exception,e:
			catch_error(e,UPERR)

	def message_lists(self,folder="")
		"""
		Fetch list of messages, that are older than self.last and
		newer than that.
		INPUT: correctly opened IMAP server, name of the folder,
		correctly set self.last (in IMAP representation).
		OUTPUT: sets self.new_msg and self.old_msg.
		"""
		if folder=="":
			folder="INBOX"
		try:
			self.select(folder)
		except Exception,e:
			catch_error(e,SELECTERR)
		self.curr_folder=folder
		try:
			ok, stuff = self.search(None,'( UNDELETED SINCE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					else
						raise ValueError, 'The field UID is not fetched!',
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					else
						raise ValueError, 'The field INTERNALDATE is not fetched!',
					ok, header = self.fetch(msg,'(FLAGS)')
					if ok == 'OK':
						flags = split(header[0])[2]
					else
						raise ValueError, 'The FLAGS are not fetched!',
					self.new_msg[msg]=uid,date,flags
			ok, stuff = self.search(None,'( UNDELETED BEFORE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					self.old_msg[msg]=uid,date
		except Exception,e:
			catch_error(e,LISTERR)
		self.close()
	def complete(self):
		"""
		leave server
		"""
		self.logout()

if __name__ == '__main__':
	## General outline
	conf = Config()
	for usr in conf.options.keys():
		opts = conf.options[usr][LOCALSTR]
		local = Server(user=usr,password=opts[0],last_sync=opts[1])
		local.message_lists()
		for svr in conf.options[usr].keys():
			if svr != LOCALSTR:
				ropts = conf.options[usr][svr]
				remote = Server(user=ropts[0],password=ropts[1],last_sync=ropts[2])
				remote.message_lists()
				sync(local,remote)
				sync_time = remote.Time2Internaldate(gmtime(time())
				conf.options[usr][svr][2] = sync_time
		conf.options[usr][LOCALSTR][1] = sync_time
	conf.close()

# vim: set ts=3:

From claird at starbase.neosoft.com  Fri Apr 19 12:12:16 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Apr 2002 11:12:16 -0500
Subject: Python scripting to dynamically generate 3D human models
References: <mailman.1019218239.16981.python-list@python.org> <gWWv8.109596$K5.9156921@bin5.nnrp.aus1.giganews.com>
Message-ID: <2456318AC06B003F.07004C8158D8635D.43C7EAA88E326885@lp.airnews.net>

In article <gWWv8.109596$K5.9156921 at bin5.nnrp.aus1.giganews.com>,
Terry Reedy <tejarex at yahoo.com> wrote:
>
>"Woon Shan Chin" <woonshan at time.net.my> wrote in message
>news:mailman.1019218239.16981.python-list at python.org...
>> Hi,
>>
>> I heard about 'Python scripting' from a member experts-exchange.com;
>an
>> IT knowledge sharing community. I'm working on a virtual fitting
>room
>> program, that needs to dynamically generate 3D human models during
>run-
>> time (based on users' input of their body measurements for eg.). So,
			.
			.
			.
>There are two ways to interface the standard version of Python, which
>is written in portable C with other C or C++ functions.  One is to
			.
		[good explanation]
			.
			.
One thing we lack, though, is an introduction to the
*idea* of scripting, especially for those destined to
embed Python.  How do we make clear to them the bene-
fits of a scripting engine?  Presumably in the case
at hand, Woon Shan Chin might be in a position to ex-
pose some of his calculations so that users can exploit
them in novel combinations.  *That*'s what a scripting
engine facilitates.
-- 

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


From alanmk at hotmail.com  Thu Apr 11 08:38:47 2002
From: alanmk at hotmail.com (Alan Kennedy)
Date: 11 Apr 2002 05:38:47 -0700
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com>
Message-ID: <f72dac0d.0204110438.5c8d3e3@posting.google.com>

Peter Hansen wrote:

> Thanks and kudos to Python core team 
             ^^^^^

I propose a new word for python lexicon: "kuidos": The rightfully
deserved thanks and praise of the Python community at large for a
group of very nice, very clever and very humble people who do a lot
work on other people's behalf.

I've been involved in some non-profit projects for the last five years
(www.crohns.org, www.paratuberculosis.org) and the ability of some
people (a very few, Thank Guido) to whine and complain about the hard
work that other people do on their behalf never ceases to amaze me.

Thanks guys, I *really* appreciate it.

Cheers,

Alan Kennedy.


From grante at visi.com  Fri Apr 12 13:29:37 2002
From: grante at visi.com (Grant Edwards)
Date: Fri, 12 Apr 2002 17:29:37 GMT
Subject: how to print to printer ?
References: <a96q8o$9cd5d$1@hades.rz.uni-sb.de>
Message-ID: <5QEt8.26820$vm6.4051596@ruti.visi.com>

In article <a96q8o$9cd5d$1 at hades.rz.uni-sb.de>, Uwe Schmitt wrote:

> I'd like to send "raw" data to a printer, how can I do this,
> using Python ???

Under Unix:

  f = open("/dev/lp0")  # or whatever device is your printer 
  f.write(data)
  f.close()

I've no idea how to do it under Windows.

-- 
Grant Edwards                   grante             Yow!  I feel partially
                                  at               hydrogenated!
                               visi.com            


From greg at cosc.canterbury.ac.nz  Wed Apr 10 02:37:45 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 18:37:45 +1200
Subject: multiple ranges to for statement?
References: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu> <mailman.1018334719.16002.python-list@python.org>
Message-ID: <3CB3DDB9.E0712D39@cosc.canterbury.ac.nz>

Ian Bicking wrote:
> 
> for i in range(a, b) + range(c, d):
>     ...

Alternatively:

  for seq in range(a, b), range(c, d):
    for i in seq:
      ...

which avoids building two copies of the
sequences.

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


From cliechti at gmx.net  Sun Apr 28 11:56:40 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Apr 2002 17:56:40 +0200
Subject: input/output through socket or COM port
References: <3CCB1CE9.3E08F027@ix.netcom.com> <Xns91FE207CF4D44cliechtigmxnet@62.2.16.82> <3CCB557C.5BDB495E@ix.netcom.com>
Message-ID: <Xns91FEB6CCA14DBcliechtigmxnet@62.2.16.82>

Jive Dadson <jdadson at ix.netcom.com> wrote in 
news:3CCB557C.5BDB495E at ix.netcom.com:

> Thanks a lot.  I snarfed the pserial zip file.  Do you know a similar
> thing for
> a TCP/IP host?

s = socket.socket(...)
s.connect(...)
fi = s.makefile()

-- 
Chris <cliechti at gmx.net>



From shalehperry at attbi.com  Mon Apr 22 18:44:00 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 22 Apr 2002 15:44:00 -0700 (PDT)
Subject: calendar.py doesn't use lists???
In-Reply-To: <15556.36861.487833.812444@12-248-41-177.client.attbi.com>
Message-ID: <XFMail.20020422154400.shalehperry@attbi.com>

# This module used to have hard-coded lists of day and month names, as
# English strings.  The classes following emulate a read-only version of
# that, but supply localized names.  Note that the values are computed
# fresh on each call, in case the user changes locale between calls.

the comment in calendar.py.




From op73418 at mail.telepac.pt  Mon Apr  1 09:03:41 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 01 Apr 2002 15:03:41 +0100
Subject: __name__ on new-style classes
References: <stcfaugkbh0u65d1euuofhs2b69ej7i34h@4ax.com> <3CA7C873.D1AC908A@python.org>
Message-ID: <jvogausg16i1vgp93qkbb8f9sb3nicgb7u@4ax.com>

On Mon, 01 Apr 2002 02:35:47 GMT, Guido van Rossum <guido at python.org>
wrote:

>"Gon?alo Rodrigues" wrote:
>> 
>> One can change the name of a classic class but not of a new-style one
>> (see below for example). Is this a bug or a feature, and if the latter
>> is there any reason for the change?
>
>It's a feature.  The reason for the change is that I've never seen an
>example of a use for it (as opposed to changing an instance's __class__).
>If you have one, let us know.
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)

I bumped into this "feature" a few days ago trying to understand
metaclasses in Python. And to understand them there is nothing better
than actually coding one, so I'll explain my problem/solution and let
you judge if this is a good use case for having a writable __name__.

My problem is modelling finite arithmetic in Python. Here by finite
arithmetic I mean all the quotients of the integer ring Z. There is a
countable number of them, one for each n > 1 (n=0 is Z and n=1 is the
trivial ring {0}. If n is negative then it is the same as -n). If n is
prime then we actually have a field.

The obvious way (to me, that is) to do this is to let each finite ring
Z_n be a class, inheriting from int. Since there are an infinite number
of these we have to have a metaclass, call it Field, that has as
intances the Z_n rings, e.g. Z_n would be Field(n). The skeleton of this
metaclass would be something like:

class Field(type):
    """The Finite Field metaclass."""

    def __new__(cls, n):
        if isinstance(n, int) and n > 1:
            #The template class.
            class Z(int):
                pass

            #Generate proper name for the instance class, e.g. "Z_n"
            Z.__name__ = "Z_%d" % n

            #Return class.
            return Z
        else:
            raise TypeError, "%s is not an integer > 1." % str(n)


If I cant write to __name__ then all the instances will have the same
name... Not A Good Thing, IMHO.

All the best,
Gon?alo Rodrigues



From phd at phd.pp.ru  Tue Apr 16 09:01:55 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 16 Apr 2002 17:01:55 +0400
Subject: Mp3
In-Reply-To: <scbc338b.035@Riksvasa.syd.kth.se>; from tdi01fhu@syd.kth.se on Tue, Apr 16, 2002 at 02:21:28PM +0200
References: <scbc338b.035@Riksvasa.syd.kth.se>
Message-ID: <20020416170155.A15453@phd.pp.ru>

On Tue, Apr 16, 2002 at 02:21:28PM +0200, FREDRIK HULDTGREN wrote:
> I would like to extract the bitrate from an mp3 file using python. Anyone know how?

   http://csl.cse.ucsc.edu/~ben/python/
   http://pyid3lib.sourceforge.net/
   https://sourceforge.net/projects/pyid3/

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




From phr-n2002a at nightsong.com  Fri Apr  5 11:37:55 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 05 Apr 2002 08:37:55 -0800
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? [LONG!])
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com> <20020405014732.GB17563@isis.gerrietts.net> <mailman.1017996733.25069.python-list@python.org> <a8kdde$12t$1@panix1.panix.com> <mailman.1018020913.20887.python-list@python.org>
Message-ID: <7xr8lunlss.fsf@ruckus.brouhaha.com>

Andy Gimblett <gimbo at ftech.net> writes:
> > Use a module.  Don't use a class unless you need a class.
> 
> /me closes his eyes tight, opens his mouth, thinks about it for a
> while, shaking his head around slightly, feels something click, says
> "wow", opens his eyes, says "wow" again, and hits reply.
> 
> Wow.  Man, I love python.
> 
> Cheers for the epiphany

Care to explain it?  I missed something.


From carole at autosim.no  Tue Apr 23 11:21:29 2002
From: carole at autosim.no (Carole Valentin)
Date: 23 Apr 2002 08:21:29 -0700
Subject: wxPython: wxDC problem
Message-ID: <9d8d2e93.0204230721.7ae6d8a0@posting.google.com>

Hello,

I have a road map drawn in a wxDC and some cars in the same wxDC.
When the cars are moving, I would like to redraw only them and not the
entire map.
So I don't want to change a part of the map, but I want to have an
independant wxDC
for my cars which are moving on the map.
Do you think that what I am asking for is possible?
How can I do that?
Does the wxDC::SetClippingRegion method create an independant wxDC?

Thank you very much,
Carole.


Here is a part of my code:
*******************

class MapWindow(wxWindow):


#-----------------------------------------------------------------------

    def __init__(self, parent, ID_MAP, pos=wxPoint(10,10),\
                 size=wxDefaultSize):

        EVT_PAINT(self, self.onPaint)


#-----------------------------------------------------------------------

    def onPaint(self, event):

        dc = wxPaintDC(self)
        self.drawMap(dc)


#-----------------------------------------------------------------------

    def drawMap(self, dc = None):

        dc.Clear()

        # I draw each lane of each road
        for i in range(len(self.myTerrain.roads)):
             for j in range(len(self.myTerrain.roads[i].lanes)):
                 for k in
range(len(self.myTerrain.road[i].lanes[j].lanePointsCentral) - 1):
                      line =
self.coefZoom*self.myTerrain.roads[i].lanes[j].width
                      dc.SetPen(wxPen(wxNamedColour('grey'), line ))
                      dc.DrawLine(
self.coefZoom*self.myTerrain.roads[i].lanes[j].lanePointsCentral[k].x
+
self.xGap ,\

self.coefZoom*self.myTerrain.roads[i].lanes[j].lanePointsCentral[k].y
+
self.yGap ,\

self.coefZoom*self.myTerrain.roads[i].lanes[j].lanePointsCentral[k+1].x
+ self.xGap ,\

self.coefZoom*self.myTerrain.roads[i].lanes[j].lanePointsCentral[k+1].y
+ self.yGap )


        # I draw only one car for the moment
        # carCorners = [ (x1,y1), (x2,y2), (x3,y3), (x4,y4) ]
        # carCorners = each corner of the rectangle which 
        # represents the car

        dc.SetPen(wxPen(wxNamedColour('BLACK'), 3))
        dc.DrawPolygon([
[self.coefZoom*self.myVehicle.carCorners[0][0]
+ self.xGap,\
                         
self.coefZoom*self.myVehicle.carCorners[0][1]
+ self.yGap],\
                        
[self.coefZoom*self.myVehicle.carCorners[1][0]
+ self.xGap,\
                         
self.coefZoom*self.myVehicle.carCorners[1][1]
+ self.yGap],\
                        
[self.coefZoom*self.myVehicle.carCorners[2][0]
+ self.xGap,\
                         
self.coefZoom*self.myVehicle.carCorners[2][1]
+ self.yGap],\
                        
[self.coefZoom*self.myVehicle.carCorners[3][0]
+ self.xGap,\
                         
self.coefZoom*self.myVehicle.carCorners[3][1]
+ self.yGap] ])


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 03:06:34 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 00:06:34 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cba7c41@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m34ridmq00.fsf at mira.informatik.hu-berlin.de...

| That's what I'm trying to tell you all the time: the
| '>>>' operator is meaningless if it is defined as
| "fill in zeroes". How does it know where to start
| inserting zeroes?

No, it isn't meaningless:

The 'long' class has an internal representation for the
long number.  The number of bytes that are currently
being used to contain the number are known (internally)
to the object.  The zeroes get shifted into the high bit
of the number, no matter how many bytes are used to
contain the number.

An int type simply shifts a zero into bit 31, so that
this works the same as it will on a long.

The '>>' operator should also work the same for a long
as it does for an int.

The '<<' operator should also work the same for a long
as it does for an int.

The 'hex()' method should work the same for a long as
it does for and int.

Why is this so confusing to you?  Java programmers have
been putting the difference between the '>>' and the
'>>>' operators to good use for years!





From sjoshi at ingr.com  Tue Apr 23 00:47:44 2002
From: sjoshi at ingr.com (Sunit Joshi)
Date: 22 Apr 2002 21:47:44 -0700
Subject: Why won't this POST
Message-ID: <8f8ffe67.0204222047.47d661ec@posting.google.com>

Hello
I'm trying to post data to IIS5.0 but keep getting error message about
"host not found". The python script is in /cgi-bin/ProcessInput.py. I
can connect to the host but cannot execute the script.
Here's the error message:

Will POST  43 bytes: txtName=guest&txtPwd=guest&txtPyInput=login
Traceback (most recent call last):
  File "C:\Sunit\Web\eubix\eubix_Local\cgi-bin\postTest.py", line 17,
in ?
    req.putrequest('POST' ,'/cgi-bin/ProcessInput.py')
  File "C:\Python21\lib\httplib.py", line 437, in putrequest
    self.send(str)
  File "C:\Python21\lib\httplib.py", line 379, in send
    self.connect()
  File "C:\Python21\lib\httplib.py", line 363, in connect
    self.sock.connect((self.host, self.port))
  File "<string>", line 1, in connect
socket.error: host not found

Here's the postTest.py script

from httplib import HTTP
from urllib import quote

quoteUsername = 'guest'
quotePwd = 'guest'
quoteInput = 'login'
postData = 'txtName=' + quote(quoteUsername) + '&txtPwd=' \
				+ quote(quotePwd) + '&txtPyInput=' + quote(quoteInput)

print 'Will POST ', str(len(postData)) + ' bytes:', postData

# begin HTTP request
req = HTTP('localhost/eubix')
#req.putrequest("GET" ,'/PyCgi/formTest.py?' + postData)
req.putrequest('POST' ,'/cgi-bin/ProcessInput.py') #Here's where it
stalls
req.putheader("Accept", "text/html")
req.putheader("User-Agent", "postTest.py")

# Need this otherwise Only GET works
req.putheader("Content-type", "application/x-www-form-urlencoded")

# Set Content-length to length of PostData
req.putheader("Content-length", str(len(postData)))

req.endheaders()

# Send POST data after ending headers
req.send(postData)

replycode, message, headers = req.getreply()
print "HTTP RESPONSE: ", replycode, message
print 'HEADERS: ', str(headers)

# Get file-like object from HTTP response
# and print received HTML to screen.
fd = req.getfile()
txtlines = fd.read()
fd.close()
print '\nReceived following HTML:\n'
print txtlines


thanks
Sunit
sjoshi at ingr.com


From eric.brunel at pragmadev.com  Tue Apr 23 08:27:41 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 23 Apr 2002 12:27:41 +0000
Subject: console redirection to Tkinter
References: <3CC34237.CE6F21F1@club-internet.fr>
Message-ID: <aa3cls$ip$1@wanadoo.fr>

Fabien Henon wrote:
> When I uncomment the outr=pout.rea.....my app gets stuck.
> The t.insert('1.0',pout) and t.insert('end',perr) do not give me what I
> expect.
> 
> 
> What shall I do ?
> Thanks for any advice.
> 
> 
> ------------------cut here--------------------------
> #!/usr/bin/env python
> from Tkinter import *
> import popen2
> 
> def lance():
>   cmd= "x-povray +i/mnt/win_c/Utile/pyvon/7.pov +w320 +h240 -f +dgt +p"
>   run = popen2.Popen3(cmd)
>   pout=sys.stdout
>   perr=sys.stderr
> #  outr=pout.readlines() # when this

If you do that, sys.stdout and sys.stderr get the standard output and the 
standard error on the *current* process, i.e. your Python script. This is 
indeed not what you want. To get the standard and error output of the child 
process, you should do:

pout = run.fromchild
perr = run.childerr

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From djc at object-craft.com.au  Wed Apr 17 09:11:36 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 17 Apr 2002 23:11:36 +1000
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de>
Message-ID: <m3g01ulbav.fsf@ferret.object-craft.com.au>

>>>>> "Martin" == Martin v Loewis <martin at v.loewis.de> writes:

Martin> stojek at part-gmbh.de (Marcus Stojek) writes:
>> The C-extension does what it is supposed to do, the callback
>> function works, but the C-function does not start as a parallel
>> thread but is freezing my UI. Anything basic I miss here ?

Martin> Yes, the Global Interpreter Lock (GIL). When C code runs that
Martin> was invoked from Python code, it holds the GIL, and only one
Martin> such C code can run at any time. So you need to release the
Martin> lock, using Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS.

If the API which is being wrapped by the extension module utilises
callbacks, and you want to make those callbacks available to Python
code, and you want to do this in a threading environment where the GIL
is released when calling the API, then the above will not work.

I tried this for my Sybase module and received a segfault for my
troubles.

In cases like this you have to do something a teeny bit more
complicated.  In my case I was lucky in that the callback functions
from the Sybase library pass a pointer to a connection structure.  In
the callback I then use the pointer to locate my Python object which
wraps the Sybase connection structure where I store all sorts of
Python related info including:

1) A PyThread_type_lock lock which ensures that all thread
   access to a Sybase API's on a connection is serialised.

2) A PyThreadState* which stores the thread state of the thread
   currently operating on the connection.

Then each call to a connection related Sybase API (ct_cancel() for
example) where I want to release the GIL looks like this:

    SY_CONN_BEGIN_THREADS(self);
    status = ct_cancel(self->conn, NULL, type);
    SY_CONN_END_THREADS(self);

which expands (more or less) to this:

    PyThread_acquire_lock(self->lock, WAIT_LOCK);
    self->thread_state = PyEval_SaveThread();

    status = ct_cancel(self->conn, NULL, type);

    PyEval_RestoreThread(self->thread_state);
    self->thread_state = NULL;
    PyThread_release_lock(self->lock);

So when a get a callback on a connection all I have to do locate the
associated Python connection object (conn) and do this:

    PyEval_RestoreThread(conn->thread_state);
    conn->thread_state = NULL;
    :
    result = PyEval_CallObject(servermsg_cb, args);
    :
    conn->thread_state = PyEval_SaveThread();

When the Sybase API invokes the callback internal to my extension
module I reacquire the GIL with the correct thread state and then call
up to the Python callback function registered by the module user.
Once the Python callback function returns I release the GIL again
because it will be reacquired when the Sybase API returns (in this
case ct_cancel()).

This seems to be working fine so far (fingers crossed).

- Dave

-- 
http://www.object-craft.com.au


From phr-n2002a at nightsong.com  Wed Apr  3 20:55:38 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 17:55:38 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xbsd0qpmn.fsf@ruckus.brouhaha.com> <3CABABCA.EB446FEE@python.org>
Message-ID: <7x7kno5isl.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
> I wouldn't say "should", but it sure would be nice.  Unfortunately the
> internal architecture of the array module doesn't make this easy; it
> would probably require a major refactoring, or perhaps a less pervasive
> switch to new-style classes so that subclassing can be used.

I'm not sure I understand the issue about new-style classes, but if
old-style classes are going away then new-style has to be supported,
right?

> Maybe you and a few others can get together and start working on the
> code?  Or do I have to do everything myself?

I've been wanting for a while to add some features to the array
module, mainly to support efficient slice operations for text editing.
That means updating the code to use the multi-segment buffer
interface.  I guess bit vectors complicate that too.

Anyway, I'll put it on my list, but don't hold your breath.


From kevin at cazabon.com  Wed Apr  3 00:45:58 2002
From: kevin at cazabon.com (Kevin Cazabon)
Date: 2 Apr 2002 21:45:58 -0800
Subject: Buffer Interface, can't get PyObject_AsReadBuffer to work?
Message-ID: <5a4226f0.0204022145.49010b2e@posting.google.com>

More growing pains... I can't seem to get PyObject_AsReadBuffer to
work for me, any suggestions based on the following code?  "ok" always
ends up as 0, meaning that the operation failed...

Thanks!  Kevin.

----------------------------------
static PyObject *
pyHasp(PyObject *self, PyObject *args)
{
  // declarations
  PyObject *inBuffer = NULL;
  void *pInBuffer = NULL;
  ULONG Service = 0;
  ULONG SeedCode = 0;
  ULONG LptNum = 0;
  ULONG Pass1 = 0;
  ULONG Pass2 = 0;
  ULONG p1 = 0;
  ULONG p2 = 0;
  ULONG p3 = 0;
  int p4size = 0;
  int ok = 0;
  void *EncodeDecodeBuffer;

  EncodeDecodeBuffer = malloc(1001);
  if (EncodeDecodeBuffer == NULL) {
    return Py_BuildValue("llls", -9994, -9994, -9994, "Could not
allocate memory for the EncodeDecodeBuffer!");
  }
  memset (EncodeDecodeBuffer, 0, 1001);

  // parse the PyObject arguments, assign to variables accordingly,
this works fine, inBuffer is a Python array ('B'), and ParseTuple
succeeds (returns 1)
  ok = PyArg_ParseTuple(args, "llllllllOl", &Service, &SeedCode,
&LptNum, &Pass1, &Pass2, &p1, &p2, &p3, &inBuffer, &p4size);

  if (!ok) { 
    printf("Could not parse the Python Tuple in pyHasp.dll!\n");
    return Py_BuildValue("llls", -9999, -9999, -9999, "Could not
access the passed string as a Python buffer!");
  }

  // Convert the pInBuffer from PyObject to a PyBuffer (readable),
THIS DOESN'T WORK (the Python object type that is passed in as
inBuffer is a 'B' array)
  if (PyObject_CheckReadBuffer(inBuffer)) {
    //this ALWAYS returns 0, meaning unsuccessful... what's wrong???
    ok = PyObject_AsReadBuffer(inBuffer, &pInBuffer, &p4size);
    if (!ok) {
      printf("PyObject_AsReadBuffer failed\n");
      return Py_BuildValue("llls", -9995, -9995, -9995, "Could not
access the passed array as a Python buffer!");
    }
  }

/* The rest of the code is here, never gets there though. */


From robin at jessikat.fsnet.co.uk  Fri Apr  5 05:36:13 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 5 Apr 2002 11:36:13 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017946214.8160.python-list@python.org>
 <zpL9tKA1hNr8EwkQ@jessikat.fsnet.co.uk> <6qhemqa2cf.fsf@abnoba.intevation.de>
Message-ID: <gzxZLgAd4Xr8Ewmq@jessikat.demon.co.uk>

In article <6qhemqa2cf.fsf at abnoba.intevation.de>, Bernhard Herzog
<bh at intevation.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> In any large subject a 'Guido' will always claim to be bigger than 'God'
>
>Well, 'Guido' *is* bigger than 'God':
>

heh heh, at last someone with a sense of humour. Of course now we can
argue whether it's 1 or True that 'Guido' > 'God'

and at least no one mentioned Hitler. duh!
 
>Python 2.3a0 (#2, Apr  4 2002, 18:58:40) 
>[GCC 2.95.4 20011006 (Debian prerelease)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> 'Guido' > 'God'
>True
>>>> 
>
>However, he's still not bigger than 'Larry':
>
>>>> 'Guido' > 'Larry'
>False
>>>> 
>
>But 'Tim' is:
>
>>>> 'Tim' > 'Larry'
>True
>>>> 
>
>
>   Bernhard
>

-- 
Robin Becker


From maxm at mxm.dk  Tue Apr  9 10:26:28 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 09 Apr 2002 16:26:28 +0200
Subject: Problem with packages and modules
References: <117880a1.0204090542.258f0f51@posting.google.com>
Message-ID: <3CB2FA14.1050708@mxm.dk>

J?rgen Hansen wrote:

> This is an example of directory hierachy:
> 
> company/                   Top-level package
>     __init__.py            Initialize the company package
>     plot/                  A plot package
>         __init__.py
>         axes.py            Modules in the plot package
>         axis.py
>         bar.py
>         figure.py
>         line.py
>         plotobject.py
>         points.py
>         tick.py
>         utils.py    
>     log.py                 Modules in the company package
>     utils.py
> 
> 
> If I were to use the plot package I would type in
>    >>> from company import plot
> 
> and assume that I could access modules as:
>    >>> plot.figure.Figure()    # where Figure is a class name.


You will need to add the full path to a module.

from company.plot.figure import Figure
fig = Figure()

or

import company.plot.figure
fig = company.plot.figure.Figure()

or

from company.plot import figure
fig = figure.Figure()


> Further if I typed in:
>    >>> import company
> 
> I would assume that I could access all subpackages and their modules
> with a dotted name (company.plot.figure.Figure() for instance), but
> this is obviously not the case.
>

You would need to import the modul in company/__init__.py:
import plot

to do:

import company
fig = company.plot.figure.Figure()


Hmm... I hope this helps a bit.

regards Max M





From boud at valdyas.org  Sun Apr 14 08:39:18 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sun, 14 Apr 2002 14:39:18 +0200
Subject: problems writing utf8
References: <3cb746ca$0$43877$e4fe514c@dreader1.news.xs4all.nl> <m3662wypqt.fsf@mira.informatik.hu-berlin.de> <3cb7f8a7$0$85437$e4fe514c@dreader3.news.xs4all.nl> <m37knb9axo.fsf@mira.informatik.hu-berlin.de> <3cb80a78$0$43852$e4fe514c@dreader1.news.xs4all.nl> <m3wuvb1b3p.fsf@mira.informatik.hu-berlin.de> <3cb977b6$0$71669$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <3cb97afa$0$71650$e4fe514c@dreader4.news.xs4all.nl>

Boudewijn Rempt wrote:

> 
> I found only one diff file -- and it fails to patch a current cvs
> checkout :-(, in the file unicodeobject.c. Should I have tried
> applying it against a the source of 2.2.1?
> 

It's not my lucky day today -- I can't patch current cvs, and while
patching against 2.2.1 works, compiling gives me an infinite recursion
error upon compiling:

...

  File "/data/src/Python-2.2.1/Lib/warnings.py", line 95, in warn_explicit
    showwarning(message, category, filename, lineno)
RuntimeError: maximum recursion depth exceeded
make: *** [sharedmods] Error 1

And now I'm non-plussed again :-).
-- 
Boudewijn Rempt | http://www.valdyas.org


From neal at metaslash.com  Thu Apr  4 20:32:11 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 04 Apr 2002 20:32:11 -0500
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96>
Message-ID: <3CACFE9B.78AE24BC@metaslash.com>

James Logajan wrote:
> 
> Eri Ma Franci <max at alcyone.com> wrote:
> > Could you find something better to complain about?
> 
> Then I'd like to complain that neither you nor Guid va Rossu addressed the
> question I asked in my previous message concerning bools as keys. If a bool
> is to be treated as an int sometimes and sometimes not, I don't recall
> seeing anything about whether it will be treated as an int as a dictionary
> key. If I do:
> 
> x = ["is", "are"]
> y = {0: "", 1:"s"}
> 
> for n in range(3):
>    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
> 
> Will the behavior be backword compatible and equivalent or not? Simple
> question I assume; waiting for an answer.

You could try it yourself, but it looks the same to me.

Python 2.1.1 (#1, Aug 13 2001, 19:37:40) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> x = ["is", "are"]
>>> y = {0: "", 1:"s"}
>>> 
>>> for n in range(3):
...    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
... 
There are 0 items.
There is 1 item.
There are 2 items.

******************

Python 2.3a0 (#238, Apr  4 2002, 20:30:33) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> x = ["is", "are"]
>>> y = {0: "", 1:"s"}
>>> 
>>> for n in range(3):
...    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
... 
There are 0 items.
There is 1 item.
There are 2 items.


From pecora at anvil.nrl.navy.mil  Thu Apr  4 11:16:50 2002
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Thu, 04 Apr 2002 11:16:50 -0500
Subject: Help running Piddle in Interpreter on a Mac.
Message-ID: <040420021116502698%pecora@anvil.nrl.navy.mil>

Whenever I try to use Piddle plotting in the Python Interpreter I get
the following message:

Traceback (most recent call last):
  File "Drive:Applications:Python 2.2:Lib:site-packages:piddletest.py",
line 398, in ?
    mainLoop()
  File "Drive:Applications:Python 2.2:Lib:site-packages:piddletest.py",
line 392, in mainLoop()
    runtest(backends[backend], tests[test])
  File "Drive:Applications:Python 2.2:Lib:site-packages:piddletest.py",
line 329, in runtest
    canvas = testfunc(canvasClass)
  File "Drive:Applications:Python 2.2:Lib:site-packages:piddletest.py",
line 20, in minimal
    canvas = canvasClass(pagesizes.A6, "testA")  # A6 is a quarter page
  File "Drive:Applications:Python
2.2:Lib:site-packages:piddle:piddleQD.py", line 166, in __init__
    self._window = _QDCanvasWindow(self, size, name)
  File "Drive:Applications:Python
2.2:Lib:site-packages:piddle:piddleQD.py", line 91, in __init__
    W.Window.__init__(self, size, title)
  File "Drive:Applications:Python 2.2:Mac:Tools:IDE:Wwindows.py", line
19, in __init__
    fontsettings = W.getdefaultfont()
  File "Drive:Applications:Python 2.2:Mac:Tools:IDE:W.py", line 28, in
getdefaultfont
    prefs = getapplication().getprefs()
  File "Drive:Applications:Python 2.2:Mac:Tools:IDE:W.py", line 24, in
getapplication
    raise WidgetsError, 'W not properly initialized: unknown
Application'
Wbase.WidgetsError: W not properly initialized: unknown Application
Exception exceptions.AttributeError: "QDCanvas instance has no
attribute 'picture'" in <bound method QDCanvas.__del__ of
<piddleQD.QDCanvas instance at 0x12483a30>> ignored



"Drive" is my HD name and Python is kept in the "Applications" folder. 
This problem happens even with the original piddletest.py module which
works ok until attempting to open a plot window -- meaning I don't get
far.  Then I get the above msg.  

I am running Python 2.2 on a Macintosh G4, system 9.1.  Same problem on
my G3 and PB.

Anyone have success in getting Piddle to work with the Interpreter on a
Mac?  Any help appreciated.

-- 
-- Lou Pecora

  - My views are my own.


From sholden at holdenweb.com  Mon Apr  1 18:10:01 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 18:10:01 -0500
Subject: selecting files from MySQL
References: <a87o30$qabnh$1@fu-berlin.de>
Message-ID: <aO5q8.118589$Ou.53288@atlpnn01.usenetserver.com>

"amy" <amylang at zedat.fu-berlin.de> wrote in message
news:a87o30$qabnh$1 at fu-berlin.de...
> hello python-knowing,
> can somebody help me to solve a little problem in a script, sc selecting
> files from a mysql database?
> so far, it doesn't want to work:
>
Wow, you can tell what a computer wants to do and doesn't!!! Do you have a
D-type connector in your neck ;-) ?
>
>     if fileid['translation'] == 'none':
>        actie = suchen fileid == 'none'
>     elif fileid['translation'] == 'de-orig':
>        actie = suchen fileid == 'de-orig'
>     elif fileid['translation'] == 'nl-orig':
>        actie = suchen fileid == 'nl-orig'
>     elif fileid['translation'] == 'no-orig':
>        actie = suchen fileid == 'no-orig'
>     else:
>        search = opt['such']
>
Seriously, although there are some people on this group who are almost
psychic in their debug skills, a little hard information can help a lot. The
error message you saw was ... ?

regards
 Steve






From jdadson at ix.netcom.com  Sat Apr 27 17:47:52 2002
From: jdadson at ix.netcom.com (Jive Dadson)
Date: 27 Apr 2002 21:47:52 GMT
Subject: input/output through socket or COM port
Message-ID: <3CCB1CE9.3E08F027@ix.netcom.com>

I hope this is an easy question, for people who know about Python
internals, or Posix stuff.  This is for an industrial control
application.  Assume that a suitable subset of MS Windows API calls and
Posix API calls are available.

I want to configure Python so it will read and write standard input and
output to and from either a TCP/IP socket or a serial COM port data
link. There will be no terminal or keyboard.

Scenario 1: The Python system will run in standard interactive mode,
accepting input from a client computer on the data link, and compiling
and executing it, just as if a human typed had it into a Python
console.  Prompts and such, plus text generated by "print" commands will
go back to the client computer.

That prompts that it writes in that mode might prove too verbose, so...

Scenario 2: The Python engine reads from the data link and compiles and
executes each line it receives, but does not echo anything back other
than what the user's programs generate with "print" commands.  The
client computer should be able to send new function definitions, just as
in interactive mode. 

I compiled Python under VC++ and used the Visual Studio debugger to step
through the code, but as yet I have not even been able to find where in
the code Python does input/output!

Anyone know how to do this?  Does python use stdio (printf, scanf, etc)
for interactive mode?

Best regards,
Jive


From mark at mceahern.com  Fri Apr 12 13:45:21 2002
From: mark at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 12:45:21 -0500
Subject: metaclass
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEMCKKAA.mark@mceahern.com>

I'm trying to translate the tracing metaclass provided here:

  http://www.python.org/doc/essays/metaclasses/Trace.py

to use the new __metaclass__ construct in Python 2.2, but I can't seem to
get off the ground.

For starters, suppose I simply want the metaclass to switch out one class
with another.

  # This should return an instance of Proxy instead of Switched.
  # Eventually, Proxy will be able to trace calls to Switched
  # before passing them on to Switched.
  s = Switched()

I try this:

#! /usr/bin/env python
# junk.py

class Proxy(object):

  pass

class SwitcherooMeta(type):

  def __new__(cls, name, bases, dict):
    # Normally, this should just do this:
    # return type.__new__(cls, name, bases, dict)
    # Instead, try to return the proxy type.
    # This results in 'Proxy' object is not callable:
    return Proxy()
    # This doesn't work either:
    object.__new__(Proxy, name, bases, dict)

class Switched(object):

  __metaclass__ = SwitcherooMeta

s = Switched()

When I run that, I get:

$ junk.py
Traceback (most recent call last):
  File "./junk.py", line 19, in ?
    s = Switched()
TypeError: 'Proxy' object is not callable

I've tried to follow the documentation provided here:

  http://www.python.org/2.2/descrintro.html#metaclasses

Any pointers on meta classes?

Thanks,

// mark





From tomchance at gmx.net  Tue Apr 30 13:18:03 2002
From: tomchance at gmx.net (Tom Chance)
Date: Tue, 30 Apr 2002 17:18:03 +0000
Subject: Python Webpages
References: <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net> <23891c90.0204300033.16549c52@posting.google.com> <3CCEAE2B.8000300@gmx.net> <mailman.1020176632.25665.python-list@python.org>
Message-ID: <3CCED1CB.3050505@gmx.net>

> 
>     I always thought Perl's system was
> 
>     print "Content-Type: text/html\n\n";
>     print "<html><body>Hello world!</body></body>\n";

That is Perl's built in system, simply printing HTML to STDOUT. But 
every Perl web developer with half a brain uses the CGI module to vastly 
extend Perl's capabilities, and perhaps also CGI::Application and 
HTML::Template as well.

use CGI;
my $cgi = new CGI;
my $name = $cgi->param("name");
print $cgi->header(-type => "text/html");

and so on...

Tom




From emile at fenx.com  Wed Apr 24 15:32:35 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 24 Apr 2002 19:32:35 GMT
Subject: Window problem in windows!
References: <mailman.1019653396.7431.python-list@python.org> <Ozzx8.64548$HH5.4797@rwcrnsc51.ops.asp.att.net> <Xns91FA740A8E146RASXnewsDFE1@130.133.1.4>
Message-ID: <nLDx8.65836$HH5.4550@rwcrnsc51.ops.asp.att.net>

Philip Swartzleonard
> I'm fairly sure he means that double-clicking the script in windows
causes
> two command-promptish output windows to come up. Sounds wierd.
>

OK.  This test.py only displays one window here when double-clicked...

import os
command = "pause \r\n"
command += "pause"
os.system(command)

So we still need an example the recreates the problem...

--

Emile van Sebille
emile at fenx.com

---------



From quinn at vomit.ugcs.caltech.edu  Mon Apr  1 03:07:33 2002
From: quinn at vomit.ugcs.caltech.edu (Quinn Dunkan)
Date: 1 Apr 2002 08:07:33 GMT
Subject: interactive python shell
References: <mailman.1017254623.8656.python-list@python.org>
Message-ID: <slrnaag5a4.qe.quinn@vomit.ugcs.caltech.edu>

On Wed, 27 Mar 2002 13:33:37 -0500, ian reinhart geiser <geiseri at yahoo.com>
wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>Greetings
>	I have an python application and I would like to provide a "console" in the 
>application where users can have an interactive session with python that has 
>all of the applications enviroment.  This is for debugging the system in real 
>time.
>
>	Is there an easy way to do this?  All it needs to do is execute commands and 
>return data.  I am using PyQt for my GUI, so it would be cool to fit into 
>there, but I can get away with a input line and a display view if that is 
>easier.  This is not for production, only to make a more useful debug tool.

Other people have pointed out some modules, but when I want to talk to python
for debugging, I do the following.  It's simple and basically works.
You can pass eval and exec namespaces if you want.

def repl():
    try:
        while 1:
            c = raw_input(sys.ps1)
            if not c:
                continue
            if c[-1] == ':':
                while 1:
                    s = raw_input(sys.ps2)
                    if not s:
                        c = c + '\n'
                        break
                    c = c + '\n' + s
            try:
                print repr(eval(c))
            except SyntaxError:
                try:
                    exec c
                except:
                    traceback.print_exc()
            except:
                traceback.print_exc()
    except:
        print

I don't know about PyQt, but it should be easy to adapt to whatever it likes...
replace raw_input with reading from a stream in a seperate thread if it
does that or take out the line reading loop and have it process incomplete
blocks if it likes callbacks.


From claird at starbase.neosoft.com  Fri Apr 19 09:40:57 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Apr 2002 08:40:57 -0500
Subject: ActivePython vs. Python
References: <WkGv8.38$Ud2.86702@news2.news.adelphia.net> <ad496f8.0204182232.288e35e3@posting.google.com>
Message-ID: <28F039E1FDED5C39.5D7D9F37745B34F9.13A8F643C308FEFF@lp.airnews.net>

In article <ad496f8.0204182232.288e35e3 at posting.google.com>,
Ville Vainio <vvainio at tp.spt.fi> wrote:
>"Bob H" <bobx at linuxmail.org> wrote in message news:<WkGv8.38$Ud2.86702 at news2.news.adelphia.net>...
>> What is the benefit of getting/using ActivePython over Python?
>
>One drawback appears to be it's reliance on ActiveState (which has to
>consider such things as profit - and it also provides ActivePerl ;-).
>There was a long period when nothing happened with ActivePython, which
>no doubt alarmed many people.
>
>However, I do love the windows help format of the Python docs,
>ActivePython might be worth it just because of them!

I'll gratuitously frame this remark in a context I prefer:
yes, it's important to be aware that ActivePython is a pro-
prietary product delivered by a for-profit corporation.  In
some situations and for some people, that is a drawback;
for others, it is a positive asset.  One of the delights of
open source is that we can package up ideas in different
wrappers to meet the needs of different end-uses.
-- 

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


From paul at boddie.net  Thu Apr 25 07:55:30 2002
From: paul at boddie.net (paul at boddie.net)
Date: 25 Apr 2002 11:55:30 -0000
Subject: configuring mxCGIPython
Message-ID: <20020425115530.11456.qmail@www3.nameplanet.com>

On Thu, 25 Apr 2002 13:01:00 +0200 "M.-A. Lemburg" <mal at lemburg.com> wrote:
>
>Note that in order for the web server to execute the cgipython
>binary, the web server user must be able to cd to the directory
>where the binary lives.

This may explain a number of things.

>Depending on where you install the binary, this may not be the
>possible. The only way around this is either finding a directory
>which the web server user account can read or to install cgipython
>in a sub-dir of cgi-bin/ which is protected against access through the
>web server using e.g. an .htaccess file (depends on the web server
>being used).

Yes, this would appear to be a solution, and I'm tempted to use that instead of 
my existing wrapper approach which is also forced to use an .htpasswd protected 
directory, albeit one in the public HTML documents space. That's certainly a 
problem with budget hosting - it's probably rare to be given non-public 
directories where programs can be stored.

>> I found that it isn't always possible to deploy Python programs in a
>> cgi-bin directory and to use the above notation to persuade the
>> operating system to locate the Python executable. In the end, I used a
>> wrapper shell script for each of my Python programs.
>
>This should not be necessary if the directory permission
>are right... but I may be wrong.

Could there also be some mechanism which prevents arbitrary paths from being 
used after #! - a list of trusted "shells", for example?

Paul




From tim.one at comcast.net  Tue Apr 16 20:52:02 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 16 Apr 2002 20:52:02 -0400
Subject: pre-PEP for optional 'pass'
In-Reply-To: <a9hnt6$5t1$1@peabody.colorado.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCCENIOOAA.tim.one@comcast.net>

[Fernando P?rez]
> ...
> Besides, you already saw Guido's unambiguous response, so this
> discussion is strictly of academic interest.

Actually not:  academics aren't attracted to discussions like this <wink>.






From sjmachin at lexicon.net  Wed Apr 10 10:29:31 2002
From: sjmachin at lexicon.net (John Machin)
Date: 10 Apr 2002 07:29:31 -0700
Subject: Extension Question
References: <cab7ec6a.0204100140.6b141d16@posting.google.com>
Message-ID: <c76ff6fc.0204100629.55b2b2@posting.google.com>

ics1010 at yahoo.com (sbp) wrote in message news:<cab7ec6a.0204100140.6b141d16 at posting.google.com>...
> In an extension I'm writing, I have a main import module which works
> fine. I'm a little confused about trying to include an object type in
> the same dll. From the import module's init I call an init function
> for the type which calls Py_InitModule with the name and method array
> for the type,

Whoops ... Py_InitModule should be called once, to initialise the
*module*, specifying the name and the table of module-level
functions/methods.
For old-style types, one or more of those functions will be
constructors for instances of the type(s) you have defined in your
module. For the moment, we'll draw a veil over the new-style types
introduced in version 2.2.

For the standard methods, you fill in functions in the slots of the
type's descriptor table. For the methods that you invent, you need a
tp_gettattr slot pointing to a function like the following, where spam
is your type.

static struct PyMethodDef spam_methods[] = {
   {"add_eggs", (PyCFunction)spam_add_eggs, METH_VARARGS,
    spam_add_eggs_doc},
   /* etc */
   {NULL, NULL}         /* sentinel */
};

static PyObject *
spam_getattr(self, name)
   spamobject *self;
   char *name;
{
   return Py_FindMethod(spam_methods, (PyObject *)self, name);
}
/* There are more modern ways of doing this but if you can find the
docs
   tell the world :-)
*/

So what happens when you do this:
 >>> import mymodule
 >>> spobj = mymodule.spam_factory(some_args) # your constructor
 >>> spobj.add_eggs(2)

is that Python finds the type of spobj, looks in the type's tp_getattr
slot, calls your spam_gettatr function looking for "add_eggs"; this
calls Py_FindMethod which rummages in the spam_methods table, finds
that "add_eggs" is implemented by the spam_add_aggs C function, does a
little more magic ("binding the method to the instance"), and then
calls spam_add_eggs ...

> but when I try to call a method from this array, I get
> an exception stating there is no attribute for the method name I'm
> calling. I have functions for the various type protocols (i.e.
> attributes) which work fine.

It is not apparent to me exactly what you mean by "functions for the
various type protocols (i.e. attributes)" nor how this could "work
fine" when you don't seem to be able to invoke a constructor to create
an object.

What source(s) of information have you been using to get so far??? I
don't have any book other than the venerable 1st edition (1996) of
"Programming Python" by Mark Lutz but this technique is in that book
(see example 14-2 stacktyp.c) and I can't see how/why it wouldn't be
in later books that have a serious chapter on extending.

I suggest that you find some practical working examples of extension
modules and follow their lead -- try (1) the Vaults of Parnassus (2)
files in the modules and objects directory of a pre-2.2 Python source
distribution e.g. listobject.c and arraymodule.c (3)
http://object-craft.com.au/projects/csv/ which is a well-written
not-too-large module.

HTH,
John


From loewis at informatik.hu-berlin.de  Wed Apr 17 11:24:46 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 17 Apr 2002 17:24:46 +0200
Subject: minidom and =?iso-8859-15?q?=E5=E4=F6?= once again :-P
References: <WXdv8.2962$ZN4.5053@news2.bredband.com>
Message-ID: <j47kn6bb5t.fsf@informatik.hu-berlin.de>

Magnus Heino <magnus.heino at pleon.sigma.se> writes:

> >>> d.toxml()
> '<?xml version="1.0" ?>\n<test>\xe5\xe4\xf6</test>'

This is not what I'm getting. I get

...
    writer.write(data)
  File "/usr/local/lib/python2.2/StringIO.py", line 139, in write
    s = str(s)
UnicodeError: ASCII encoding error: ordinal not in range(128)

Apparently, you've changed the system default encoding, so that the
conversion from Unicode strings to byte strings silently converted the
Unicode string.

The .toxml() method currently does not expect this to happen; instead,
it means to return a Unicode object as output. The problem is that
StringIO attempts a str() conversion of the Unicode object. That is a
bug in Python 2.2 which has been corrected in Python 2.2.1 (where
StringIO supports Unicode again).

In any case, your best bet is to use an explicitly-encoding stream:

>>> s=StringIO.StringIO()
>>> d.writexml(codecs.getwriter('utf-8')(s))  
>>> s.getvalue()
'<?xml version="1.0" ?>\n<test>e\xc3\xa4\xc3\xb6</test>'

HTH,
Martin


From ods at fep.ru  Sat Apr 27 09:33:46 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Sat, 27 Apr 2002 17:33:46 +0400 (MSD)
Subject: name of the current function ?
In-Reply-To: <aae7kf$apv$06$1@news.t-online.com>
Message-ID: <Pine.LNX.4.44L.0204271732590.20682-100000@ods.pp.ru>

On Sat, 27 Apr 2002, Solveig Wiesmayr wrote:

SW> I would like to get the name 'my_name' of the current
SW> function to pass it
SW> to some other (debugging-) function.

>>> import sys
>>> def my_name():
...     return sys._getframe(1).f_code.co_name
...
>>> my_name()
'?'
>>> def f():
...     return my_name()
...
>>> f()
'f'






From grante at visi.com  Sun Apr 14 09:53:33 2002
From: grante at visi.com (Grant Edwards)
Date: Sun, 14 Apr 2002 13:53:33 GMT
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnabj2g8.1gj.grante@tuxtop.visi.com>

In article <m31ydiy7b9.fsf at mira.informatik.hu-berlin.de>, Martin v. Loewis wrote:
> grante at visi.com (Grant Edwards) writes:
> 
>> I agree.  Since the other boolean operators are "not" "and"
>> "or", it would be a bit of a surprise that one doesn't use
>> "xor". And having to convert a boolean to a "chunk of bits"
>> first and do a bitwise xor on it is just plain ugly.
> 
> As Bengt points out, it is tricky to define this analogous to 'and'
> and 'or'. Both have a shortcut meaning, evaluating the second one only
> if the first does not determine the outcome. Furthermore, they both
> return the last expression evaluated. This is not possible for 'xor'

Obviously.  That's implied by the definition of the operation.

> you have to evaluate both, and there is no clear choice as to which
> one is better as a result - in fact, the OP wants it to return bool.
> 
> So xor would be very much *unlike* 'and' and 'or'.

Point is that "^" operates bitwise, and there is no
exclusive-or operation for "truth" values.  Returning a bool
seems like the expected behavior to me.

-- 
Grant Edwards                   grante             Yow!  .. someone in DAYTON,
                                  at               Ohio is selling USED
                               visi.com            CARPETS to a SERBO-CROATIAN


From billdozier at hotmail.com  Tue Apr  9 14:08:12 2002
From: billdozier at hotmail.com (Bill Dozier)
Date: 9 Apr 2002 11:08:12 -0700
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com> <slrnab5pur.ngm.zeitlin@seth.lpthe.jussieu.fr>
Message-ID: <199334c5.0204091008.2efbc49b@posting.google.com>

Vadim Zeitlin <zeitlin at seth.lpthe.jussieu.fr> wrote in message news:<slrnab5pur.ngm.zeitlin at seth.lpthe.jussieu.fr>...

>  The problem with having wxPython-only documentation would be an even

> increased risk of inconsistencies between wxWindows and wxPython docs. 

Since Robin uses SWIG to generate his wrapper code, could this
situation improve once generated documentation is re-implemented in
SWIG?

I've had my eyes on wxPython for a long time, waiting for a decent
MacOS (and now Mac OS X) port. I've got high hopes now....

Bill


From JamesL at Lugoj.com  Fri Apr  5 11:48:57 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 5 Apr 2002 16:48:57 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
Message-ID: <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96>

philh at comuno.freeserve.co.uk (phil hunt) wrote:
> Another problem is that when a language is constantly changing, it 
> becomes necessary to constantly relearn it. Python has changed a lot 
> in the last few years, for example new operators such as +=, and 
> allowing users to subclass from builtin types, are big (and useful) 
> changes. Python 1.5.1 was an immature language that could benefit 
> from changing.

Version 1.5.2 was sufficiently complete and long-lived enough that I 
established it as the version I would develop against.

> Is there a final destination in sight, which these changes are 
> leading to?

The answer to this isn't as interesting to me as _when_ will this tinkering 
end? Something that isn't up for debate: because of the problems I have 
actually encountered with making code work with the many extant versions of 
Python, I will no longer be doing any new projects in the language and am 
not recommending Python to anyone. The language tinkering is one of several 
reasons I left Python and this newsgroup many months ago. (I have 
resubscribed recently to find out if there was a way to rework an old 
Tkinter app of mine that people have asked for some changes to, and will be 
exiting here in a few days.)


From tdelaney at avaya.com  Mon Apr 29 03:36:55 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 17:36:55 +1000
Subject: Python newbie
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A6@natasha.auslabs.avaya.com>

> From: slackness at caramail.com [mailto:slackness at caramail.com]
> 
> Thank you but I know that. I would like my programs to be launched by
> people who don't have Python interpreter on their machine. I think
> that Py2exe (mentionned before by.. I don't remember, sorry, but
> thanks) could resolve this problem. I'll try it.

Well, technically no - your users will still have a Python VM on their
machine. Your users just won't know it.

> I would like to know one more thing... Is it possible to use Python to
> program for Windows CE devices??

http://www.python.org/download/download_windows.html

Tim Delaney




From jwinkelman at movaz.com  Thu Apr  4 09:44:00 2002
From: jwinkelman at movaz.com (Jeff Winkelman)
Date: 4 Apr 2002 06:44:00 -0800
Subject: Passing arguments to COM Server
Message-ID: <90164d1a.0204040644.5a076db1@posting.google.com>

I am having some difficulty with a COM object I'm trying to use from
Python.

Background:
I have an application that I am trying to call a Python COM Server
from.  One limitation of this application is that it cannot access any
COM Server that doesn't have a type library associated with it.

So, for example, it does not support the equivalent VB code:

set obj = CreateObject("Library.Class")

I have struggled with getting a type library built that represents my
Python COM object, so I have resorted to building a COM DLL in Visual
Basic that is a wrapper around my Python COM Server.  The Visual Basic
DLL is visiable to my front end application, and at a basic level,
this seems to work.

So in summary, the front end app. calls a visual basic DLL that in
turn calls a Python COM Server.

The next step requires that I pass a COM Object from the front end app
all the way down to python.

This, on the surface, seems to work as well.  When I get to my Python
COM Server, I turn the PyIDispatch object into a COM object that's
usable by Python with the following line of code:

obj = win32com.client.Dispatch(obj)

The problem I have is that when I try to call methods of this passed
in COM object, the front end object generates an error stating "Wrong
Number of Positional Parameters".

For example,

obj.SetValNumber("xxxx", 0, 10.0) generates a COM error from the front
end application with the above description.

I realize that this is an error being generated by the front end app,
however, I am not clear why the correct number of parameters is not
making it to the object.

One concern that I have is if I print out the object in python, I get
the following:

The object is <COMObject <unknown>>

It seems that Python knows this is a COM object, but doesn't know what
kind.

I have run makepy on the Type Library for the front end application,
but that didn't seem to make any differnce.

I am positive that I am attempting to pass the right arguments to the
COM object, however, something is happening along to create this
problem.

Does anyone have any suggestions on things I might try?

Here's my Python COM Server Code.

import win32com
import win32com.client

class IPython:
    _public_methods_ = ['runScript']
    _reg_progid_ = "Movaz.IPython"
    _reg_desc_ = "Python Inteface"
    _reg_clsid_ = "{FF2AA877-5F2B-4F28-A283-DB0C772442E9}"

    def __init__(self):
        print 'Creating IPython Object'
    
    def runScript(self, script, context):
        # Use these commands in Python code to auto generate .py
support
        from win32com.client import gencache
        gencache.EnsureModule('{B2794EF3-C0B6-11D0-939C-0020AF68E893}',
0, 1, 0)
        o = win32com.client.Dispatch(context)
        
        print 'The context object is', `o`
        
        o.SetValNumber("Locals.Gregg", 0, 100)
        return "Running Script " + str(script) + " " #+ str(name)

def Register():
    import win32com.server.register
    return win32com.server.register.UseCommandLine(IPython)
    
if __name__=='__main__':
    print "Registering COM server..."
    Register()

And a dump from Python when the error occurs.

Object with win32trace dispatcher created (object=None)
Creating IPython Object
in <IPython.IPython instance at 02E5B024>._QueryInterface_ with
unsupported IID IPersistStreamInit
({7FD52380-4E07-101B-AE2D-08002B2EC713})

in <IPython.IPython instance at 02E5B024>._QueryInterface_ with
unsupported IID IPersistPropertyBag
({37D84F60-42CB-11CE-8135-00AA004BB851})

in _GetIDsOfNames_ with '('runScript',)' and '1033'

in _Invoke_ with 1000 1033 3 (u'Test', <PyIDispatch at 0x2e5cf9c with
obj at 0x2b458e8>)
The object is <COMObject <unknown>>
Traceback (most recent call last):
  File "C:\Python21\win32com\server\dispatcher.py", line 40, in
_Invoke_
    return self.policy._Invoke_(dispid, lcid, wFlags, args)
  File "C:\Python21\win32com\server\policy.py", line 274, in _Invoke_
    return self._invoke_(dispid, lcid, wFlags, args)
  File "C:\Python21\win32com\server\policy.py", line 279, in _invoke_
    return S_OK, -1, self._invokeex_(dispid, lcid, wFlags, args, None,
None)
  File "C:\Python21\win32com\server\policy.py", line 509, in
_invokeex_
    return apply(func, args)
  File "C:\Python21\IPython.py", line 25, in runScript
    o.SetValNumber("Locals.Gregg", 0, 100)
  File "C:\Python21\win32com\client\dynamic.py", line 432, in
__getattr__
    raise pythoncom.com_error, details
com_error: (-2147352567, 'Exception occurred.', (0, 'TSAPI', 'Expected
1 positional (non-named) parameters; found 0.', None, 0, -17808),
None)
    

Thanks in advance for any help.

Jeff


From jb at cascade-sys.com  Tue Apr 30 15:10:47 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 12:10:47 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2>	<3CCD9649.98AED86F@cascade-sys.com> <oqpu0inmxa.fsf@titan.progiciels-bpi.ca> <mailman.1020126468.7915.python-list@python.org> <ECyz8.16652$xy4.10715@atlpnn01.usenetserver.com>
Message-ID: <3CCEEC37.73118261@cascade-sys.com>

Steve Holden wrote:

> While I don't wish to disagree with you here, one point I forgot to mention
> in my last posting: very few people have the thick skin, intelligence, wit,
> humor and modesty :-) that I do.

You also forgot grace, tact and your remarkably large penis.

> You might therefore wish to ... dare I say ... lighten up yourself when
> dealing with these wimpy foreign types.

Rather, seems I should need to TIGHTEN up when dealing with those weenies.  ;o)

Nevertheless, point taken.  Though I have to suppress a gag reflex to say it, I
shall strive to be more "Politically Correct" in my future posts to this forum.

That being said, however, I'll beg everyone's indulgence in advance in the
event I occasionally lapse back into old habits.  I frankly can't recall the
last time I intentionally insulted anyone.  If I ever have occasion to do so it
will be shockingly unambiguous.  The ground will shake and birds may fall
stunned from the sky.  When in doubt you'll virtually always be correct in
assuming my comment was a failed attempt at humor or a clumsy American's gaff
rather than an intentional insult.  Nevertheless, my professional background
includes working with a number of people who were perfect assholes but also
were supremely competent and talented, technically.  Before starting my
company, my last boss was an ex Navy officer.  Frank talk was not only the norm
you lost points for being overly deferential.  I personally continue to think
speaking frankly and directly is an effective means to communicate, so I may
lapse from time to time.  Also, my language probably suffers from watching too
much premium cable TV, Dennis Miller Live being a particular favorite.  I
humbly beg all y'all's indulgence.

Oops, gotta go.  The voices are saying it's time again to clean all the guns.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From fredrik at pythonware.com  Tue Apr 30 17:38:06 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 21:38:06 GMT
Subject: inner classes
References: <mailman.1020200515.3993.python-list@python.org>
Message-ID: <29Ez8.34862$n4.7403406@newsc.telia.net>

"Hermann Hesse" wrote:

> Can someone give an example of how you'd use bound methods to replace inner
> classes in python?

quoting from some random Java tutorial:

public class SomeGUI extends JFrame
{
   // inner class definitions
   class Button1Handler implements ActionListener
   {
      public void actionPerformed(ActionEvent e)
   {
         // do something
      }
   }

   protected void buildGUI()
   {
      button1 = new JButton();
      button1.addActionListener(new Button1Handler());

here's the python version:

class SomeGUI(PyFrame):

    def Button1Handler(self, e):
        # do something

    def buildGUI(self):
        button1 = PyButton()
        button1.addActionListener(self.Button1Handler)

where "self.Button1Handler" evaluates to a callable object that
represents the Button1Handler method for a specific SomeGUI
instance.

</F>




From aleax at aleax.it  Wed Apr  3 05:10:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 10:10:36 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <eppstein-E6E1AD.15234102042002@news.service.uci.edu> <Lkyq8.45231$pT1.1286029@news1.tin.it> <eppstein-F70646.00210603042002@news.service.uci.edu>
Message-ID: <wyAq8.54383$S52.1879050@news2.tin.it>

David Eppstein wrote:

> In article <Lkyq8.45231$pT1.1286029 at news1.tin.it>,
>  Alex Martelli <aleax at aleax.it> wrote:
> 
>> > expressivity and possibly adds minor functionality (ability to make
>> > code that behaves differently for ints and for bools, just like now you
>> > can
>> 
>> It would be quite possible to distinguish between int and Booleans if
>> you needed to, even if Booleans came from a specific module rather
>> than from builtins, of course.  Such a need is a very rare issue, thus
>> it should not be allowed to get in the way for everyday usage.
> 
> Sure, but it's only *useful* to distinguish between them if you're
> likely to actually see Booleans in your input, which will only happen if
> they aren't hidden in some back corner where only the
> external-language-interfacers can find them.

You'd see them when client-code using your external-interfacing module
need to show them to you, specifically.  And, basically, nowhere else.
Ah, bliss.  I doubt that's the solution we'll see, but that's quite
another issue -- it would be by far the least of evils.


Alex



From guido at python.org  Thu Apr  4 09:35:15 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 09:35:15 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Thu, 04 Apr 2002 10:53:51 +0200."
             <3CAC149F.3F1BA77B@lemburg.com> 
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>  
            <3CAC149F.3F1BA77B@lemburg.com> 
Message-ID: <200204041435.g34EZFa10532@pcp742651pcs.reston01.va.comcast.net>

[MAL]
> > > Right, but my point is that if you want to make booleans first class
> > > objects in Python, you should consider making them behave as defined
> > > in the text books rather than trying to make them feel like integers
> > > but look like truth values.

[me]
> > I reject this reasoning.  Python is not trying to follow the text
> > books.  (In fact, maybe I should say "Python is trying not to follow
> > the text books.")

[MAL]
> Uhm, I can't follow you here -- which text books you are talking about 
> here (Python has always tried to conform to standards in most areas). 

The textbooks that say that a Boolean value does not support
arithmetic operations.  In tis discussion several people have tried to
argue for such bools and claimed that that is the textbook definition.

> FWIW, I was talking about the text books students use to learn boolean 
> algebra.

I don't think Boolean algebra is a big subject.  I bet that most
people only learn about it explicitly when they learn computer
programming.

> BTW, I wonder why you didn't follow up on my other comments...

This thread is way too long to comment on everything.  I probably felt
that I'd already responded to the same issue in a different message.

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




From eeide at cs.utah.edu  Tue Apr  9 11:22:42 2002
From: eeide at cs.utah.edu (Eric Eide)
Date: Tue, 09 Apr 2002 09:22:42 -0600 (MDT)
Subject: Response to Guido's essay on graph representation
References: <eppstein-87C133.12464904042002@news.service.uci.edu> <pM9r8.11156$FU6.2998356006@newssvr10.news.prodigy.com>
Message-ID: <ywrpu1828xp.fsf@ioka.cs.utah.edu>

"Paul" == Paul M <p.magwene at snet.net> writes:

	Paul> I've recently been fooling around with various graph algorithms,
	Paul> [...]

Is your code available?  This wasn't clear to me from your previous message.

A well-written Python library of graph data structures and algorithms sounds
like something I would want to add to my arsenal!

Eric.

-- 
-------------------------------------------------------------------------------
Eric Eide <eeide at cs.utah.edu>  .         University of Utah School of Computing
http://www.cs.utah.edu/~eeide/ . +1 (801) 585-5512 voice, +1 (801) 581-5843 FAX


From maxm at mxm.dk  Wed Apr 17 09:09:26 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 17 Apr 2002 15:09:26 +0200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <3CBD7406.3050909@mxm.dk>

dominikush at yahoo.com wrote:

> Who is wrong here: my intuition or Python (2.2)? If it's
> my intuition, how can I train my thinking about Python's
> execution model, so that my intuition get's better ;-)

Your intuition!

Objects like lists gets copied by reference. So Python behaves exactly 
like it should.

Maybe it is clearer with another notation:

 >>> dict = {'a':<list ref1>,'b':<list ref2>}
 >>> list = dict.values()
 >>> list

[<list ref1>, <list ref2>]

 >>> dict['a'].append(3)
 >>> dict

{'a': <list ref1>, 'b': <list ref2>}

 >>> list

[<list ref1>, <list ref2>]


regards Max M



From dsavitsk at e-coli.net  Tue Apr  9 14:30:44 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 09 Apr 2002 18:30:44 GMT
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk>
Message-ID: <orGs8.5669$d7.1571810@newssrv26.news.prodigy.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
news:slrnab617e.97a.philh at comuno.freeserve.co.uk...
* * *
> IMO, anyone for whom K&R isn't to their taste, probably won't make a
> very good programmer anyway.
>

Yow! IMO, anyone who thinks that only one sort of person (a sort who likes a
particular book no less) makes a good programmer needs to relax a bit in
their judgmentalness.

-d

> --
> <"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
> "I would guess that he really believes whatever is politically
> advantageous for him to believe."
>                         -- Alison Brooks, referring to Michael
>                               Portillo, on soc.history.what-if




From aahz at pythoncraft.com  Sat Apr 20 13:49:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Apr 2002 13:49:02 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at>
Message-ID: <a9s9me$890$1@panix1.panix.com>

In article <a9rn4i$mm4$1 at newsreader1.netway.at>,
Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
>
>Python shows very often its C heritage, whereas Common Lisp is devoid
>of that impression. From this point of view Lisp is much different from
>Python; Lisp really went its own way.

Um.  Lisp is older than C.  In fact, the only older language still in
common use is Fortran.  (COBOL came after Lisp.)

What's amazing about Lisp is that it's still in many ways on the cutting
edge of computer science.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From donald.braman at yale.edu  Thu Apr  4 13:06:14 2002
From: donald.braman at yale.edu (don)
Date: Thu, 4 Apr 2002 10:06:14 -0800
Subject: dbi odbc error
Message-ID: <a8i4hh$htk$1@news.ycc.yale.edu>

I'm new to python. I'm trying to update tables in an MS Access database. I
can connect and retrieve data just fine. The following *does* work well:

                import dbi, odbc
                conn = odbc.odbc(mybase)
                cur = conn.cursor()
                cur.execute('SELECT * FROM mytable WHERE myfield < 1830')
                print cur.description
                for tup in cur.description:
                        print tup[0],
                print
                while 1:
                        rec = cur.fetchmany(10)
                        if not rec: break
                        print rec

But when I try to update a field the in table, I get an error. So, the
following does *not* work:

                import dbi, odbc
                conn = odbc.odbc(mybase)
                cur = conn.cursor()
                cur.execute('UPDATE mytable SET mytable.myfield= "test";')
                conn.commit()
                cur.close()
                conn.close()

The error I get is:

  File "F:\simple update.py", line 4, in ?
    cur.execute('UPDATE mytable SET mytable.myfield = "test";')
dbi.program-error: [Microsoft][ODBC Microsoft Access Driver] Too few
parameters. Expected 1. in EXEC







From sholden at holdenweb.com  Mon Apr 22 19:10:52 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 19:10:52 -0400
Subject: calendar.py doesn't use lists???
References: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <fO0x8.82707$T%5.833@atlpnn01.usenetserver.com>

"Aaron Watters" <aaron at reportlab.com> wrote in message
news:9a6d7d9d.0204221405.7d394207 at posting.google.com...
> Hi guys.  I got an irksome problem.  It's possible that
> it's fixed in python 2.2.1, but even if it is, it irks me.
>
In order to be fixed it has to be broken. Please read on ...

> I tried the following script:
>
> """
> import sys
> print sys.version
> if 1:
>     from calendar import month_abbr, month_name
>     # hey python guys, i want lists!!
>     for x in (month_abbr, month_name):
>         print type(x)
>     month_abbr = list(month_abbr)
>     month_name = list(month_name)
>     print month_abbr
>     print month_name
> """
>
> Using python 2.1 I get this output
>
> """
> 2.1.1 (#20, Jul 26 2001, 11:38:51) [MSC 32 bit (Intel)]
> <type 'list'>
> <type 'list'>
> ['   ', 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep',
'Oct', 'N
> ov', 'Dec']
> ['', 'January', 'February', 'March', 'April', 'May', 'June', 'July',
'August', '
> September', 'October', 'November', 'December']
> """
>
> (completing normally)
>
> Using Python 2.2 I get this
>
> """
> 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)]
> <type 'instance'>
> <type 'instance'>
> [CRASH AND BURN... WINDOWS XP SUGGESTS I SEND A COMPLAINT TO MICROSOFT...]
> """
>
Interestingly, both cygwin2.2 and Windows2K 2.2 also fail. There's clearly
some drastic problem here...

> Of course it may be that this is fixed in python 2.2.1 but...
>
> WHO'S GOOFY IDEA WAS THIS?
>
Well, reading the source gives co clue to the guilty party. We could find
out in SourceForge if we really wanted. It wasn't me, honest.

> Seriously, why did you need to change these sequences from
> lists where they worked with my code, to somethingsimilartolists
> that broke my code?
>
Well, let's take a look at the documentation, shall we. In "5.13 calendar --
General calendar-related functions" for Python 2.2 there is no mention of
these attributes. Nothing in 2.1 or 1.5.2 either. So offhand I'd say the
author was free to replace those sequences with
somethingbearingnodamnedresemblanceatalltolists without anyone having any
right to complain.

> Irked.  In general I've noticed a tendancy with recent python
> releases towards changing stuff for the fun of changing stuff,
> I'm sorry to say.  Please remember that when you fix something
> that isn't broken, you might just be breaking somebody else's code.
>
> Sorry to whine.
>
It's easy to understand why you are pissed, but having broken encapsulation
the way you did it's a little unfair to whine at the developers. There's
been a lot of complaining about language change recently, but precious
little hard evidence of real code breakage. As a well-known programmer in
the Python community your opinion counts, so please be careful not to spread
FUD. I have to say in this case that the code breakage you report is not
Python's fault (though the crash and burn behavior should be fixed:
segfaults are a bit OTT).

Suppose I were to mail you and complain that gadfly.privatevar no longer
contains a floogle instance since you updated it. Would I have a leg to
stand on? Answer: only if your documentation included a description of that
attribute.

Thanks for Gadfly, by the way. A brilliant piece of software that helps
Python-only users get into relational databases.

regards
 Steve
--
home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/





From paul.osman at sympatico.ca  Tue Apr 16 13:40:48 2002
From: paul.osman at sympatico.ca (Paul Osman)
Date: Tue, 16 Apr 2002 13:40:48 -0400
Subject: Newbie question
References: <a9hhbs$1ksn$1@as201.hinet.hr> <3CBC4EC4.4080401@sympatico.ca> <a9hjvc$cdq$1@as201.hinet.hr>
Message-ID: <3CBC6220.7010004@sympatico.ca>

Ante wrote:

>Um.. and shouldn't one read the article before replying?
>
>If this is solely a scope problem, then why does it work if you replace
>'num += 1' with 'print num'?
>
>Shouldn't it say 'referencing local before assignment yada yada yada'
>anyway?
>
>My question was: why does it work if you only use the variable, and doesn't
>work if you assign to it? I guess it would actually make sense for plain
>assignment, but for += operator, it doesn't seem so.
>
Okay, let me try this again... Python detects local names statically. 
Variables assigned in a function are classified as local by default. 
therefore in the following script:

<code>
num = 1

def dummy():
    num = 5
    return num

print dummy()
print num
</code>

The output will be:
5
1
This is because there are actually two variables named "num", one 
global, and one local to the function dummy(). Now, if we were to do 
take out the assignment from the dummy() function, the output would be:
1
1
Hence in this case, 'num' represents only one variable (with global 
scope). This is why your print statement worked inside the function. The 
print was referring to the 'num' with global scope.

Now, as for why an assignment won't work, well, if you did this:
<code>
num = 5
def dummy():  
    num = 1
    return num
</code>
You would be creating one variable with global scope called num and one 
with function level scope called num. Note that the function level 
variable is destroyed once the function is finished. Therefore the += 
operation you were attempting before wouldn't make much sense... if the 
variable only exists within the function? global variables are immutable 
from within functions. Consider this:
<code>
num = 5
def dummy():
    newnum = num + 5
    return num
</code>
the function dummy() will now return 6. Because you were not changing 
the value of 'num'.

Any of this clear? Forgive me, I'm not the best at explaining things, 
but hopefully this should shed some light..

-Paul






From jimmy at retzlaff.com  Tue Apr 30 14:15:38 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 30 Apr 2002 11:15:38 -0700
Subject: Fatal Python error: GC object already in linked list
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F5659@MAYOR.retzlaff.com>

I'm working on an application in Python and my application occasionally
dies with the error:

Fatal Python error: GC object already in linked list

I'm quite unfamiliar with Python's implementation, but I tried studying
a little in this case. That error message only appears in a macro in
objimpl.h (_PyObject_GC_TRACK), and that macro only seems to be used in
a function that creates traceback objects in traceback.c
(newtracebackobject). Looking at CVS it appears the use of that macro
was added to prevent the old memory leak that could happen when calling
sys.exc_info().

My program is a non-trivial (~15K lines of Python) GUI app on Windows
based on wxPython. I've been developing this app for 6-8 months and the
error just started showing up about 3 weeks ago. I haven't updated any
major components (other than my own code) for a few months, so I'm
guessing my own code changes have triggered this.

Occurrences seem quite random, which can indicate timing issues of
course. If my app is started and left alone, the error will occur
anywhere from 10 seconds to several hours later. I do some idle and
timer processing. That processing does regularly raise and handle
exceptions (which I assume tickles newtracebackobject), but it all seems
to be pretty vanilla stuff. The timer processing was added only a few
days before I discovered the problem.

Anyone have any ideas about what kind of thing might cause such a
problem?

Thanks in advance,
Jimmy


p.s. -- some version information of heavily used components:

Windows XP Pro (I've also seen the problem on Windows 2000)
Python 2.2 (the standard Windows binary distribution)
Win32all Build 146
wxPython 2.3.2.1 Hybrid (also seen it with a Normal build)
Numeric 20.3




From jeff at ccvcorp.com  Fri Apr 19 13:33:59 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 19 Apr 2002 10:33:59 -0700
Subject: Trinary operator?
References: <mailman.1019066614.29792.python-list@python.org> <just-4227D7.21004717042002@news1.xs4all.nl> <ubvub71mto0d04@news.supernews.com>
Message-ID: <MPG.1729f4d41b01ab809896b4@news.nwlink.com>

In article <ubvub71mto0d04 at news.supernews.com>, 
johnroth at ameritech.net says...
> 
> "Just van Rossum" <just at xs4all.nl> wrote in message
> news:just-4227D7.21004717042002 at news1.xs4all.nl...
> > In article <mailman.1019066614.29792.python-list at python.org>,
> >  "Mark McEahern" <marklists at mceahern.com> wrote:
> >
> > > > Not that it's a big deal since there's many ways to do that, but
> can I use
> > > > something like
> > > >     verboseGender = (gender == 'm') ? 'male' : 'female'
> > > > in Python?
> > >
> > > People seem to use and/or for that:
> >
> > But they shouldn't.
> 
> Why not? There's no other reason for the behavior
> of 'and' and 'or' returning the object that caused the
> final result.

I would say that they shouldn't do that because it's unclear, or 
at least unintuitive.  I know this trick exists, but any time I 
see 'a and b or c' I have a heck of a time remembering which gets 
returned in what case.  Of course, I never was thrilled about C's 
ternary operator, either.

I definitely think that using a small dictionary, even one 
created in-place (I.e., "{0:'female', 1:'male'}[gender=='m']"), 
is a more maintainable solution.  It still looks a bit ugly, but 
I can at least follow what's happening.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From aleax at aleax.it  Wed Apr  3 08:48:51 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 13:48:51 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <gIpq8.44728$pT1.1266378@news1.tin.it> <slrnaalqkh.fa8.mlh@vier.idi.ntnu.no>
Message-ID: <7LDq8.55269$S52.1904449@news2.tin.it>

Magnus Lie Hetland wrote:
        ...
> Yup. An excellent distinction. I'm tempted to re-edit the chapter on
> conditionals in (the upcoming) Instant Python...

Speaking as an unofficial tech reviewer for that book -- please do.
I think Laura will give you permission just as she gave it to me,
and you're luckier than me -- you don't have to cater for Python 2.3...


Alex



From logiplexsoftware at earthlink.net  Mon Apr  1 13:25:31 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 10:25:31 -0800
Subject: wxPython  / wxSingleChoiceDialog ...
In-Reply-To: <a89vu0$8bm2c$1@hades.rz.uni-sb.de>
References: <a89or2$8dn6m$1@hades.rz.uni-sb.de>
	<slrnaagtjj.d7o.zeitlin@seth.lpthe.jussieu.fr>
	<a89vu0$8bm2c$1@hades.rz.uni-sb.de>
Message-ID: <20020401102531.23cdfc19.logiplexsoftware@earthlink.net>

On 1 Apr 2002 15:55:44 GMT
Uwe Schmitt wrote:

> Vadim Zeitlin <zeitlin at seth.lpthe.jussieu.fr> wrote:
> | On 1 Apr 2002 13:54:42 GMT, Uwe Schmitt <uwe at rocksport.de> wrote:
> |> I tried to manipulate wxSinglChoiceDialog so that the presented 
> |> choices are printed with a fixed width font. I tried SetFont()
> |> but nothing changed...
> 
> |  You should change the font of the listbox used by this dialog, not the
dialog
> | itself. Although there is no way to access this listbox directly (and
unlikely
> | to ever be), you can iterate over the dialog children until you find
it.
> 
> I inspectetd the __dict__ attribute of my singlechoicedialog object and
> tried dir() but found no hints how to iterate over the dialog children...

dialog.GetChildren().  Look at the docs for wxWindow.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From gcordova at hebmex.com  Wed Apr 24 12:57:29 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Wed, 24 Apr 2002 11:57:29 -0500
Subject: FROM DIFFICULTIES WITH LEARNING AT SCHOOL, TO COMMUNICATION P
	ROBLEMS IN THE FAMILY.,
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339535EE@mail.hebmex.com>

> On 23 Apr 2002 10:28:27 -0700, Anne Mc Murray 
> <anne_mc_murray at hotmail.com> wrote:
> >One of my sons is reluctant to produce any good school work, and has
> >turned into a frightful headache for the family's, despite being very
> >intelligent and creative, and a good reader too.
> >I've tried other alternatives, with psychologists, and
> >psychopedagoges, even private teachers, but nothing seems to work.
> 
> Does it *matter* that his school work isn't good? IMO, if his work 
> isn't marked as part of an exam assessment that will affect his 
> career or further education choices, it isn't such a big deal.
> 
> >Please, has anyone a suggestion to make?
> 
> Try posting to an appropriate newsgroup?
> 

Taking out that black leather belt and putting it to good
use on the boy's ass would work wonders too. But first
talk to him and tell him *why*. Then let him have some.

Worked wonders in the olden times.

hmmm... and not so olden too.

This-is-gonna-hurt-you-more-than-him'ly yours,

-gustavo

pd: IT'S THOSE DAMNED VIDEO GAMES I TELL YOU!!!
    AND THAT ROCK AND ROLL TRASH!!!
    AND THE INTERNET!!! IT'S THE WORK OF THE DEVIL IT IS!!!
    (of the devil and B1FF!!!)
:-)




From tundra at tundraware.com  Mon Apr 22 18:30:02 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 22:30:02 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <mailman.1019510980.7097.python-list@python.org>
Message-ID: <3CC48EDD.2BEE5215@tundraware.com>

arthur.siegel at rsmi.com wrote:
> 
> >>>Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
> >>>would-be applications to the natural sciences. I think it was P. Erdos
> >>>who remarked once that all natural science is conceivably finite - e.g.
> >>>there would be a time when there would be nothing more to learn about
> >>>the universe - but mathematics is literally infinite.
> >
> >>I'm not so sure about the latter assertion, and I'm *definitely* not
> >>certain about the assertion that natural science is finite.
> 
> >The first assertion by Erdos is indeed problematic. The second is not.
> 
> My sense, intution is of mathemetics as infinite only in the sense of allowing
> infinite homologous descriptions of/perspectives on  the One Thing.
> 
> The perspectives are invented.
> 
> The Thing is not.

This is entirely cogent and on-the-mark, IMO, but it still sidesteps the
question of whether or not mathematics can operate on things which do
not exist in the natural realm.  My (unproven/unprovable) claim is that
mathematics is capable of operating on a larger problem set than
that posed by natural science no matter how large/finite/infinite natural
science, in fact, is.

> 
> The modern assumption that increased abstractness brings better perspective
> perhaps reaches a point not only of diminishing, but negative returns.
> 
> But as the opinion of someone who finds the more abstract mathematical
> realms beyond his reach, maybe this represents a Philosophy of Sour Grapes.
> 

I feel your pain ;)  I'm not sure that the assumption is really that
increased abstraction gives better perspective.  I suspect increased
abstraction is pursued by mathematicians because it is intellectually
interesting and new territory to conquer...



-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From mertz at gnosis.cx  Thu Apr 18 21:27:42 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 18 Apr 2002 21:27:42 -0400
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com>
 <I6Fv8.36158$vF6.1093396@news2.tin.it>
 <mailman.1019162916.19664.python-list@python.org>
 <7cIv8.36928$vF6.1122050@news2.tin.it>
Message-ID: <OK3v8kKkXUJW092yn@gnosis.cx>         

Alex Martelli <aleax at aleax.it> wrote previously:
|Thanks for the compliment, but you're only missing *the whole point*,
|it seems to me.

At least I was right in knowing I was missing something :-(.

I think I was reading the subject as something more like "changing the
repr() *IN* a class".  And I didn't bother actually running Alex' code
to make my misreading stand out.

Sorry for the think-o... back to the regularly scheduled dicussion.

Yours, Lulu...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From starx at pacbell.net  Sat Apr 13 07:09:47 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 13 Apr 2002 11:09:47 GMT
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <Xns91EF2A8A1CF47RASXnewsDFE1@130.133.1.4>

Ralf Juengling || Sat 13 Apr 2002 03:27:20a:

> Michael Hudson <mwh at python.net> writes:
> 
>> Ralf Juengling <juenglin at informatik.uni-freiburg.de> writes:
>> 
>> > Shouldn't I find the special methods '__iadd__' et al, when
>> > asking for the attributes of 'int'?
>> 
>> ints are immutable, so they don't implement __iadd__.  lists do:
>> 
> 
> Hm. But then an expression 'i+=1' should cause an error if 'i'
> is an int...
> 
>>>> i = 1
>>>> type(i) 
> <type 'int'>
>>>> id(i)
> 134526912
>>>> i+=1
>>>> id(i)
> 134526852
>>>> 
> 
> I understand now, that 'i+=1' is no real in-place operation but
> just a shortcut for 'i=i+1' since int is immutable. The language
> reference states that this is okay, anyhow, I find it confusing.
> 
> For the sake of clarity, an 'in-place' operation should really 
> work in-place, don't you think?

Even though it can't? I'm guessing that you mean that id(i) should be 
the same before and after i+=1? My first reaction would be that you'd 
need to create another variable anyway, before doing the add, and moving 
it back to where the original was would be too expensive for such a 
simple operaiton.

However it appears the real reason is far wierder; observe:

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> a = 1
>>> b = 2
>>> c = 2
>>> id(a)
7692256
>>> id(b)
7692232
>>> id(c)
7692232
>>> b is c
1
>>> a += 1
>>> id(a)
7692232
>>> a is b
1

It appears that python only kees one copy of any given int around, and 
any name that happens to gain the status of being that integer really 
just gets another copy of it's pointer. It seems like it'd be less 
effort, but...

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From andreas at andreas-jung.com  Tue Apr 16 08:26:58 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Tue, 16 Apr 2002 08:26:58 -0400
Subject: Mp3
References: <scbc338b.035@Riksvasa.syd.kth.se>
Message-ID: <014301c1e542$00f0fc40$02010a0a@suxlap>

the bitrate is usally encoded as ID3 tag. try one of the ID3 modules
for Python (-> search google).

Andreas

----- Original Message -----
From: "FREDRIK HULDTGREN" <tdi01fhu at syd.kth.se>
To: <python-list at python.org>
Sent: Tuesday, April 16, 2002 08:21
Subject: Mp3


I would like to extract the bitrate from an mp3 file using python. Anyone
know how?
:)

/Fredrik


--
http://mail.python.org/mailman/listinfo/python-list






From johnroth at ameritech.net  Tue Apr 30 07:37:23 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 30 Apr 2002 07:37:23 -0400
Subject: Third Draft: Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch> <MNiz8.32691$8D3.925320@news1.tin.it> <mailman.1020117911.9648.python-list@python.org> <slrnacsuc9.vg.gerhard@developer2.opus-gmbh.net>
Message-ID: <uct0e7jiv6pt3a@news.supernews.com>

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnacsuc9.vg.gerhard at developer2.opus-gmbh.net...
> In article <mailman.1020117911.9648.python-list at python.org>, Laura
Creighton wrote:
> > Alex, Alex, surely you know that the main advantage of oct is that
> > dec 25 == oct 31  which you can use to prove that all Christians are
> > Satanists.
>
> I don't remember the proof for this obvious fact. Could you tell me
how it
> works again?
>
> Gerhard

3 * 8 = 24 + 1 = 25.




From sholden at holdenweb.com  Mon Apr 15 13:05:28 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 15 Apr 2002 13:05:28 -0400
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <3CBABA71.3040909@geneva-link.ch> <YwCu8.32367$vm6.5103753@ruti.visi.com>
Message-ID: <yNDu8.13611$Ji7.6515@atlpnn01.usenetserver.com>

"Grant Edwards" <grante at visi.com> wrote in message
news:YwCu8.32367$vm6.5103753 at ruti.visi.com...
> In article <3CBABA71.3040909 at geneva-link.ch>, Boris Borcic wrote:
>
> > For the shortcut semantics, we'd like to have :
> >      a xor b
> >
> > behave exactly like :
> >      (not b and a) or (not a and b)
>
> I don't understand how that's a shortcut (assuming that means
> the same thing as short-circuit).  You have to evaluate both
> operands of an xor operator.  There is no possibly way to
> short-circuit the expression so that only one is evaluated.
>
But they each might be evaluated *twice* in the given equivalent expression
if the existing Boolean operators didn't short-circuit.

regards
 Steve






From aleax at aleax.it  Thu Apr 11 06:19:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 12:19:29 +0200
Subject: [Python-Help] SMTP with attachments
In-Reply-To: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>
References: <0F757892D113D611BD2E0002559C1FF4637FE0@EMAIL>
Message-ID: <E16vbg1-00027Q-00@mail.python.org>

On Thursday 11 April 2002 11:34 am, Ezequiel, Justin wrote:
> How do I send an e-mail with attachments?
> I am a newbie so be gentle.

OK, but please don't cross-post so widely -- addressing only one of the
mailing lists you chose would be best.

If you use Python 2.2 (or have installed the separately available email
package), the email package is best.

At:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/86674

you'll find an example of attaching all the files in the current directory.

I hope it's reasonably easy for you to edit this recipe if you want to
be more selecting or whatever.  Note that this recipe doesn't SEND
the email, it just BUILDS it (and writes it into a file called 
dirContentsMail); there are other recipes on the Cookbook for
SENDING, e.g.:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52243

(which you can also take as an example of using the older MimeWriter
way of preparing multipart mail, such as mail with attachments, should
you be unable to use the newer email module for whatever reason).


Alex




From pyth at devel.trillke.net  Mon Apr 29 20:36:47 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 02:36:47 +0200
Subject: Remove masked elements from an array
In-Reply-To: <aaknh8$1c4$1@peabody.colorado.edu>
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <20020429174425.GN16217@devel.trillke> <mailman.1020115270.9666.python-list@python.org> <aaknh8$1c4$1@peabody.colorado.edu>
Message-ID: <20020430003647.GU16217@devel.trillke>

On Mon, Apr 29, 2002 at 06:09:55PM -0600, Fernando P?rez wrote:
> James J. Besemer wrote:
> 
> > holger krekel wrote:
> > 
> >> there are too solutions. choose yourself :-)
> > 
> > THREE if you include LAMBDA.
> 
> No, please heed other advice and use the MA approach. Please note that I'm not 
> dissing this post on its own merits, it's a nice discussion, it just is 
> ill-adapted to numeric processing. Numeric provides its own _fast_ way to 
> manipulate arrays with invalid data, by all means use it. After all, speed is 
> one of the big reasons why you use Numeric to begin with (you could do most 
> things in it with regular lists, if you had the age of the universe to wait).

Yes, this is the probably the best solution. I have never used Numeric
but it seems to be a very useful module.

Though it cant hurt to get to know basic techniques :-)

    holger




From hfoffani at yahoo.com  Thu Apr  4 08:00:49 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Thu, 4 Apr 2002 15:00:49 +0200
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <_vUq8.552234$pN4.39754796@bin8.nnrp.aus1.giganews.com> <xdupu1fae7h.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <20020404080053.340$Yd@news.newsreader.com>

"Ralf Juengling"
> this might have been brought up already (I didn't follow the whole
> discussion): Currently, when an object x is tested as in
> 'if x:' or 'if not x:'
>
> the interpretation of x as a boolean is done by PyObject_IsTrue()
> at the very end. PyObject_IsTrue() itself tries to interpret the
> given object as a number, then (if this fails) as a mapping,
> finally as a sequence. Here is where the something/nothing semantic
> roots in.
>
> But what about user-defined types?
>

__nonzero__() ?


class YepNop:
    def __init__(self, arg):
        self.val = arg

    def __nonzero__(self):
        if self.val == "yep":
            return 1
        else:
            return 0

yes = YepNop("yep")
no  = YepNop("nop")
if yes:
    print "yes"
if not no:
    print "no"


Regards,
-Hernan





From chris at psychofx.com  Tue Apr  9 08:44:08 2002
From: chris at psychofx.com (Chris Miles)
Date: Tue, 9 Apr 2002 22:44:08 +1000
Subject: how to read hardware info on Win NT
In-Reply-To: <3cb2e686.21131515@news.easynews.net>; from stojek@part-gmbh.de on Tue, Apr 09, 2002 at 01:05:00PM +0000
References: <3cb2e686.21131515@news.easynews.net>
Message-ID: <20020409224408.A15950@psychofx.com>

On Tue, Apr 09, 2002 at 01:05:00PM +0000, Marcus Stojek wrote:
> could anyone please give me a hit how to get hardware information with
> Python? I need the id of my ethernet card and my harddisk.

On a similar topic, are there any modules to give you Win NT/2k/etc performance
statistics, process lists, network stats, disk usage, etc ?

Any work being done on such a module ?

Regards,
Chris.

-- 
Chris Miles




From logiplexsoftware at earthlink.net  Mon Apr 29 17:45:51 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 14:45:51 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCDB559.A1BF0E47@cascade-sys.com>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
	<20020429130136.5b0c34f7.logiplexsoftware@earthlink.net>
	<3CCDB559.A1BF0E47@cascade-sys.com>
Message-ID: <20020429144551.55f5d67a.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 14:04:25 -0700
James J. Besemer wrote:

> 
> Cliff Wells wrote:
> 
> > Well that's a stupid statement - *your* head is in the clouds.
> 
> Oh yeah?  Fuck you.

Heh.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From timr at probo.com  Tue Apr 30 01:49:08 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 29 Apr 2002 22:49:08 -0700
Subject: Free tool for reports development
References: <6a5f9873.0204291439.41156c92@posting.google.com>
Message-ID: <8ubscu8qagjjk8dkc2t06eb5n7s2n26f82@4ax.com>

mkellner at rpvreport.com (Martin Kellner) wrote:
> 
>Sorry for disturbing, but I would like to notify you that we're
>offering a FREE TOOL for report development: Rpv Printing System.
> 
>This tool lets you develop Windows reports from any programming
>language and we've been noticed that Python community could be glad to
>hear about it.

This is a very interesting concept.

It would be fun to play with the source code to make the input format more
like HTML/XMLish.  A tool that could take an HTML <TABLE>, format it, and
paginate it would be valuable.

Do you plan on releasing the Delphi source code for this tool?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From tim at vegeta.ath.cx  Tue Apr 30 20:16:36 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Wed, 01 May 2002 00:16:36 GMT
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net> <mailman.1020118433.15459.python-list@python.org> <Xns91FFB1BDEE1BCRASXnewsDFE1@130.133.1.4> <slrnacu417.pma.tim@vegeta.ath.cx> <Xns9200AD794AD56RASXnewsDFE1@130.133.1.4>
Message-ID: <slrnacudbr.qst.tim@vegeta.ath.cx>

Philip Swartzleonard graced us by uttering:
> Tim Hammerquist || Tue 30 Apr 2002 02:38:50p:
> 
>>> But if you just want to increment something, use 'foo+=1'. Exactly
>>> the same, except one more character and only works on a line of it's
>>> own (avoiding above said doom... man pointer-walking funcitons in C
>>> are nuts with that... =)
>> 
>> Well, yes and no. The type being incremented is a string, not an
>> int. For example: [..]
> 
> Oh... =). I figured it was just a wierd way of stating a variable
> name.  But then, i have <2 minutes of total perl experience (if
> anything, changing shebangs and 'global settings' at the top of a few.
> maybe.)

No harm. It is one of the (much) less self-explanatory examples of
syntactic sugar.

Ironic that with all the talk of Perl's extraneous non-alphanumeric
symbols throughout its syntax, in this case the quotes were actually
meant to quote! =)

As a recovering Perler, I just appreciate you're not blindly
Perl-bashing. Thx. =)

Tim Hammerquist
-- 
If there's anything more important than my
ego around, I want it caught and shot now.
    -- Zaphod Beeblebrox, "The Hitchhiker's Guide to the Galaxy"


From cliechti at gmx.net  Thu Apr  4 14:51:30 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 4 Apr 2002 20:51:30 +0100
Subject: turning a string into an object name
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <3cabf705.91920985@mammoth.usenet-access.com> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com>
Message-ID: <Xns91E6D45CAAFEcliechtigmxnet@62.2.16.82>

netzapper at magicstar.net (A. Jones) wrote in
news:3cac943f.132176618 at mammoth.usenet-access.com: 

> On Thu, 04 Apr 2002 11:56:24 GMT, Alex Martelli <aleax at aleax.it>
> wrote:
> 
>>You cannot use ANY identifier safely after that exec
>>statement, as you have no idea any more what the
>>identifier can refer to.  As a bonus, your code crawls,
>>since the Python compiler knows it does not know, and
>>therefore executes a full identifier look-up at runtime
>>rather than recognizing local variables at compile time.
>>
>>What kind of useful code can you write without being
>>able to use ANY identifier safely?  WHY would you ever
>>WANT to trample all over your namespace to ensure every
>>identifier becomes an utter and total mystery?  Beats me.
> 
> I agree with you if it's an arbitrary string being passed to your exec
> call .  However, if you have some idea of what the form of the string
> is going to be, it doesn't seem so dangerous.  For instance:
> 
> for x in range(0, 100):
>      exec foo + `x` + " = " + "myClass()"
> 
> allows you to easily make a large list of sequentially named
> variables, with no chance of stepping on a system call.
> 
> Furthermore, you can wrap it inside a class.
> 
> class test:
>      def wrongHeaded(self, name, value):
>          exec "self."+name+'='+repr(value)
>          print vars()
> 
> this produces no exceptions, and doesn't cause any problems, since
> vars is never defined... only self.vars is.

until someone has the idea to pass some fancier names like
t.wrongHeaded('a; del None', 57)

and then have fun watching your entire program fail...

when you wrap it in a class there is no reason to not use getattr/setattr.

exec/eval is always critical like os.system() and the like. security holes 
due to unmotivated eval/os.system etc. use come up every now and then and 
you wound't want to be the author of that software.

chris

-- 
Chris <cliechti at gmx.net>



From m_mariappanX at trillium.com  Wed Apr 24 08:06:21 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Wed, 24 Apr 2002 05:06:21 -0700
Subject: atoi
Message-ID: <53A7943A5BD8D411B6930002A5073155022D748D@bgsmsx90.iind.intel.com>

Hi all,

simple......but need to know

is there any python buildin method whihc will convert strings into integer,
like atoi in c lib.?

"234" should be converted into interger value 234

but the string function ord(character) convert it into equvalent ASCII code

thanks....
maharajan






From boud at valdyas.org  Mon Apr  8 16:18:18 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 08 Apr 2002 19:18:18 -0100
Subject: Python + MySQL + Qt/Gtk(or somethink like that)
References: <d9641a00.0204080848.4d80d2b0@posting.google.com>
Message-ID: <3cb1d117$0$85432$e4fe514c@dreader3.news.xs4all.nl>

Wander Stabenow Weberling wrote:

> Hi,
> 
> I'm brazilian programmer and I realy want to know how to use python
> with MySQL and a GUI toolkit like Qt or Gtk. All tutorials that I
> found in the web (maybe I dont know how to search) are very basical.
> 

For a non-basic introduction to Python and Qt, check my book 
(www.opendocs.org). It doesn't however cover database applications. 
I didn't have time or space to cover that.

However, if you want a complicated example, you might want to download
Kura, which is a PyQt application that accesses a MySQL database.
You can find it at: http://www.valdyas.org/linguistics/index.html.


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


From bokr at oz.net  Sun Apr 14 21:45:45 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 01:45:45 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com>
Message-ID: <a9dbc9$bge$0@216.39.172.122>

On Sun, 14 Apr 2002 15:55:55 -0700, "Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
[...]
>
>Thus-- my "expectation" is that 0x8000000000000000 >> 1
>is: 0x4000000000000000 (and that is the way Python now
>works for positive 'long ints'.)
>
>Given the above description of the way NEGATIVE 'long
>ints' should behave, my 'expectation' is that:
>-0x8000000000000000 >> 1 _SHOULD_ return:
>-0xC000000000000000,
>
>(_NOT_ -0x4000000000000000 as Python _NOW_ behaves.)
>
I think you are touching on an inconsistency in representation,
but I think you may be misinterpreting the hex representation as
it now is. The minus sign means that the hex is representing the
bits of the absolute value, not the bits of the negative
number in its twos complement form.

On thinking about it, I think this is ugly, though probably
convenient in the implementation. I.e., looking at the bits
of the absolute value is not usually what you really want when
you convert to hex, IMO.

All 32-bit ints are represented in hex without a sign, so
you get this ugliness:
 >>> 1<<31
 -2147483648
 >>> hex((1<<31))
 '0x80000000'
 >>> hex((1<<31)*1L)
 '-0x80000000L'

but
 >>> hex(1L<<31)
 '0x80000000L'

or perhaps more instructive, since 8 appears both ways above:
 >>> hex(15<<28)
 '0xf0000000'
 >>> hex((15<<28)*1L)
 '-0x10000000L'
 >>> hex(15L<<28)
 '0xF0000000L'

or simply
 >>> hex(-1)
 '0xffffffff'
 >>> hex(-1L)
 '-0x1L'

If hex long representation were interpreted to represent actual twos
complement bits, with the top hex digit 8-bit being extended indefinitely
to the left, you could have hex(-1L) be represented as 0xFL (which would
require that hex(15L) be represented as 0x0FL).

This might make for too much breakage, so an alternative hex prefix could
be usedm e.g., 0s instead of 0x, s meaning the the msb is interpreted as
sign, the way 32-bit ints are done now, but independent of bit width.

Using this format, 0xC would be equal to 12, but 0sC would be equal to -4
which is equal to -0x4 (which you recognize from your example). To use
0s prefix for a positive number with a msb in the 8-bit of the ms hex digit,
you'd have to prefix a hex 0 digit, e.g., 0xc == 12 == 0s0c.

Thus your -0xC000000000000000 is currently wrong as sign+hex(abs(value)),
but it could be right if we dropped the sign and used the new representation, e.g.,
0sC000000000000000 meaning that the msb was a sign bit.

>The way Python _NOW_ works, is to give a different
>behavior for the '>>' operator depending on whether you
>are working on an 'int', or a 'long int'.  This was OK
>before, because the 2 object types were treated
>differently by the language.  I thought we were trying
>_NOW_ (versions >= 2.2.x) to "unify" the behavior of
>'ints' and 'long ints'.
>
>Am I wrong?
>
Yes, I think it's not the >> operator that's at fault. It's that
0x80000000 does not numerically equal 0x80000000L, and the rest follows.
 >>> 0x80000000 == 0x80000000L
 0
 >>> 0x80000000
 -2147483648
 >>> 0x80000000L
 2147483648L

>Now, if I am right about this, then we don't really
>know if we are right shifting an 'int', or a 'long int'
We don't need to know, we just need to understand what numeric
values are really being represented internally and as hex.
The numeric values are being shifted right consistently.
It's just that negative hex longs are represented as
signed absolute values, whereas negative ints are represented
with actual bits in 8 digits. This makes ugly confusion, but
not invalid arithmetic.

>and so it would be 'useful' to have a '>>>' operator in
>this case, to guarantee that zeroes are shifted into
>the highest bit.  JAVA has this facility, why must
>Python _NOT_ have it?
Python effectively does shift in 0's for positive numbers and 1's
for negative, but if you want 0x80000000 with zeroes above, you
have to spell it 0x80000000L, 'cause 0x80000000 is a negative int
(at least on platforms with 32-bit ints. 64-bit probably works
analogously). If you want to shift in zeroes to a negative int,
you're implicitly asking for a reinterpret cast on the int number
representation from signed to unsigned (which you can accomplish
various ways).

Regards,
Bengt Richter


From gcordova at hebmex.com  Thu Apr 25 09:57:02 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 25 Apr 2002 08:57:02 -0500
Subject: a ? on using Pmw dialog widget?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233989693@mail.hebmex.com>

Tkinter is only usable from the main thread; it's a design and
implementation restriction.

Good luck.

-gustavo




From jon+usenet at unequivocal.co.uk  Thu Apr  4 08:07:50 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Thu, 04 Apr 2002 13:07:50 -0000
Subject: cgi.py performance
References: <a8ght4$o8o$1@web1.cup.hp.com>
Message-ID: <slrnaaok15.4ei.jon+usenet@snowy.squish.net>

In article <a8ght4$o8o$1 at web1.cup.hp.com>, jimh wrote:
> I am using cgi.py to handle the transfer of large incoming files.  Because
> of the way that cgi.py handles multipart/form-data, the performance is VERY
> poor.

You could try jon.cgi from jonpy, it might be more efficient.
http://jonpy.sourceforge.net/


From boud at valdyas.org  Tue Apr 30 12:03:05 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 30 Apr 2002 18:03:05 +0200
Subject: Backdrop in QListView using PyQt, possible?
References: <d906b2c1.0204300652.5d7fac93@posting.google.com>
Message-ID: <3ccec07b$0$322$e4fe514c@dreader1.news.xs4all.nl>

<posted & mailed>

Moray Taylor wrote:

> Is it possible to apply a backdrop to a QListView, i.e. a PNG or
> whatever, loaded into a QImage and used as a tiled (or otherwise)
> backdrop, like you can on a QCanvas?
> 

It should be possible, but it won't be as easy as with those myriad Visual 
Basic grid controls, that can put a complete avi flick in the backdrop 
(poor secretary who finds that in his productivity app!)

Perhaps you can fake it by creating your own widget style. Otherwise you're 
up against some serious re-implementation of the paint method of both 
QListView and QListViewItem.

Or, but I don't know whether it'll work, you might want to investigate the 
setBackgroundPixmap (or the modern equivalent) of QWidget: a QListView is a 
QWidget, even if a QListViewItem isn't.

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


From noomnews at yahoo.com  Tue Apr  2 14:58:40 2002
From: noomnews at yahoo.com (Noom Knews)
Date: Tue, 2 Apr 2002 11:58:40 -0800 (PST)
Subject: ctrl-c and os.system problem
Message-ID: <20020402195840.36791.qmail@web20208.mail.yahoo.com>

I have a simple program like this

os.system("ls -lR / > /tmp/ls.out")
print "end"

When I press ctrl-c, the "end" message is still print.
 I'd like to terminate the whole process when pressing
ctrl-c instead of just "end".

I tried to catch the signal already but can't catch
the SIGINT or anything else at all during the
os.system call.   This does not work either

try:
   os.system("ls -lR / > /tmp/ls.out")
   print "end"
except:
   print "get all exception"

This will keep printing "end" when ctrl-c is pressed
too. I think this could be os.system problem. If I use
other statement instead of os.system, the program
works as I expected.

This was tried on Redhat Linux with python 1.5.2

Thanks,
Nick

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://http://taxes.yahoo.com/



From colin.pinkney at orange.removethis.net  Mon Apr  1 18:32:39 2002
From: colin.pinkney at orange.removethis.net (Colin Pinkney)
Date: Mon, 01 Apr 2002 23:32:39 +0000
Subject: Cannot find modules on SuSE Linux 7.3
Message-ID: <3CA8EE17.1050006@orange.removethis.net>

Hi,

Just installed SuSE 7.3 but there is something wrong with the python 
installation. Anything that uses python falls over with errors like this 
(this came from xrpm):

<<
'import site' failed; use -v for traceback
Traceback (most recent call last):
   File "/usr/lib/xrpm/xrpm.py", line 33, in ?
     import os
ImportError: No module named os
 >>

I have $PYTHONHOME set to /usr/lib/python which is a link to 
/usr/lib/python2.1 and I've tried setting it to /usr/lib/python2.1 too.

If I run python outside this directory and type something like 'import 
os' it can't find any module. If I run python in /usr/lib/python2.1 then 
it still starts up with:

<<
'import site' failed; use -v for traceback
Python 2.1.1 (#1, Sep 24 2001, 05:28:47)
[GCC 2.95.3 20010315 (SuSE)] on linux2
 >>

But then typing 'import site' or importing any other basic module seems 
to succeed.

Can anyone see what is wrong?

TIA

-- 
Colin Pinkney



From jb at cascade-sys.com  Mon Apr 29 20:51:38 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 17:51:38 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <B43D149A9AB2D411971300B0D03D7E8BF0A2AB@natasha.auslabs.avaya.com>
Message-ID: <3CCDEA9A.DBF6E351@cascade-sys.com>

LOL!!  ;o)

--jb

"Delaney, Timothy" wrote:

> From: James J. Besemer [mailto:jb at cascade-sys.com]
>
> > Incidentally, by my count, *I* am the only victim of an
> > unprovoked personal attack on this list today.
>
> Damn!
>
> Alex - Your mother was a hamster, and your father smelled of elderberries.
>
> Steve - I fart in your general direction.
>
> Laura - Go away or I shall taunt you another time!
>
> Tim Delaney
>
> --
> http://mail.python.org/mailman/listinfo/python-list

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From tim.one at comcast.net  Wed Apr 17 01:30:50 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 17 Apr 2002 01:30:50 -0400
Subject: Tuple/List identity guarantees Was(Re: __iadd__ and fellows
 missing (Python 2.2))
In-Reply-To: <3CBB3117.7090309@rogers.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOFOOAA.tim.one@comcast.net>

[Mike C. Fletcher]
> This little mental exercise all started when Tim pointed out the tuples
> _won't_ work for the pattern (well, he pointed out that there's no
> guarantee that imutable values (particularly empty tuples) will always
> be seperate objects).  So using an empty tuple for the pattern is no
> longer suitable.

It was never suitable in theory, and since it so happens that the
implementations to date usually do return the same object for "an empty
tuple", in practice it hasn't even worked by accident.

> To the best of my knowledge, there _is_ an implicit guarantee (at least,
> such a guarantee is assumed in a lot of code) that two different
> _mutable_ objects (such as lists or instances) will _always_ return
> different id values, and thus always compare unequal with "is".

This is explicitly guaranteed by the language reference manual, and, indeed,
by the same (long <wink>) sentence that also explicitly warns it's not
guaranteed for operations returning objects of immutable types.  See the
"Objects, value and types" section.

> ...
> Oh, I care deeply for the Timbot's feelings.  Six years of fighting a
> bloody guerilla war against him ;) has made me come to love the little
> floating ball of light :) (though the strange looking guy that always
> hangs out around him kinda freaks me out ;) ).

This is a burden for the three of us, Mike.

> ...
> Reminds me, where have all the optimisation challenges gone these days?
> Tim, Christian and I used to spend whole minutes on those babies, and
> now all we can do is bat around recycled debates on object identity :).

Well, the fun when out of it when people starting taking it seriously.   The
fun of the old speed battles was based on pouring unbounded effort into
shaving microseconds nobody could sanely care about.  A magnificent effort
in pursuit of an inherently worthless goal.  This was my generation's
substitute for the giddy senselessness of war <wink>.

all's-well-so-long-as-you-wake-up-drunk-ly y'rs  - tim





From fperez528 at yahoo.com  Mon Apr  8 21:55:21 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 08 Apr 2002 19:55:21 -0600
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com> <b0f083db.0204080837.625d270f@posting.google.com> <1018304015.46066@cswreg.cos.agilent.com> <MPG.171bbba6ee73df8c98969f@news.nwlink.com>
Message-ID: <a8te6h$hae$1@peabody.colorado.edu>

Jeff Shannon wrote:

> If you want access to variables inside of a function, you should
> pass those specific variables as arguments.  If you've got lots
> of arguments in common between lots of functions, assemble the
> lot into a class.  If neither of these is appropriate, then you
> should redesign -- you're using a very ugly approach.
> 

Not necessarily. I can show you several examples of functions which need to 
know the context of the caller in a generic enough fashion that argument 
passing isn't a viable option. Fortunately so far I've been able to get what 
I've needed by manipulating the stack, but the point remains. Yes, in _most_ 
cases a clean, explicit passing of arguments is the right way to do things. 
But there are perfectly valid cases where more complex solutions are needed, 
don't dismiss them so quickly.

Cheers,

f.


From geo_engineer at yahoo.com  Sun Apr  7 22:08:04 2002
From: geo_engineer at yahoo.com (Geometric Engineer)
Date: 7 Apr 2002 19:08:04 -0700
Subject: Java and Python
References: <3C96DBB4.6040109@pobox.com> <3C97FB8A.5B9E9247@engcorp.com> <3C98108D.5010002@pobox.com> <3C981961.9F5E5383@engcorp.com>
Message-ID: <eedf05f0.0204071808.451e23da@posting.google.com>

BUMBLEBEE BUZZ:

  I found this long message by you (snipped in quoted material) to be
very interesting and to confirm suspicions from past research. I plan
to carefully ignore pieces of paper and to look for those factors you
mention, among which especially is a "lack of snotty programming ego"
(as I'd put it undiplomatically), aside from my sad prejudice against
Perl (the most damnably unmaintainable language I've seen, other than
GW-Basic laced with uncontrolled out-of-scope GOTO's).

  How has your experience been anyway with tests of any kind to bring
out actual programming ability? Do you rely mainly on "look and feel"
from the interviews? I've found generally the people who know about a
subject will be able to explain simply and clearly how to work at it,
and that the people who don't know about a subject will talk lots but
give little real information, or worse, be actively confusing.

  Muttering Engineer

Peter Hansen <peter at engcorp.com> wrote in message news:<3C981961.9F5E5383 at engcorp.com>...
> 
> The point is, I would never hire you or anyone _because_
> you knew Python, or Java.  I would hire you because
> I thought you were capable of learning it and using it
> effectively.  I actually tend to shy away from people 
> with great lists of "certifications" in languages because
> it suggests to me they have difficulty learning and 
> think it's a very impressive thing that they managed
> to learn enough to write an exam which tests knowledge
> of syntax but little about programming skill.
> 
> My suggestion then is to "get over" the comparison thing,
> and make sure you know both, and throw in others as you
> come across them and have opportunity to learn. This will
> make you a better programmer, because of the learning and
> generalization your brain will do, not because of the 
> specific skills you have.
> 
> Not all companies hire in this way, of course, but all
> the best ones do. :-)


From sandskyfly at hotmail.com  Mon Apr 29 21:19:11 2002
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 29 Apr 2002 18:19:11 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Xns91FF62618790Flumberjack@216.148.53.85>
Message-ID: <b03e80d.0204291719.1f36c42b@posting.google.com>

<snip unecessarily hostile comments from lumberjack>

I'm sure you know you can communicate 100% of your point with less
than 1% of your sarcasm. I'd hate to see people like you dissuade
future pep writers from articulating their ideas to the community.

regards,

Sandy


From h_schneider at marketmix.com  Mon Apr  8 06:20:00 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Mon, 8 Apr 2002 12:20:00 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <20020408120242.41456ace.wilk@flibuste.net>
Message-ID: <a8rqh2$9h5$03$1@news.t-online.com>

The stability of Python on Win32 is excellent. Even threadening is
absolutely robust! Long running scripts are
also stable ...

But what about the wxPython bindings? The only wxPython projects I know are
some smaller applications, nothing
really complex ....

Bye,
Harald

"William Dod?" <wilk at flibuste.net> schrieb im Newsbeitrag
news:20020408120242.41456ace.wilk at flibuste.net...
> Le Mon, 8 Apr 2002 11:55:24 +0200
> "Harald Schneider" <h_schneider at marketmix.com> ?crivait:
>
> > Hi,
> >
> > what are your experiences, using wxPython/wxWindows in large projects?
> >
> > Large means in my case a sales system, based on MySQL.
> > I'm thinking about implementing it in wxPython, but I'm not sure about
> > the stability of
> > wxWindows and the wxPython wrapper ...
> >
> > The alternative way would be a project in Borland Delphi, which has
> > proofed its
> > stability over years ...
> >
> > So what do you think about it!
> I think i have the same question in my head !
> More, i would like to know how is the stability of python on windows,
> compare to java or delphi for example.
>
> bye
>
> --
> William Dod? - Informaticien Ind?pendant
> http://www.flibuste.net




From peter at engcorp.com  Tue Apr 23 23:58:46 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 23 Apr 2002 23:58:46 -0400
Subject: XML, Python, XML-SIG, documentation, PyXml, 4
References: <aef33705.0204231902.40694af4@posting.google.com>
Message-ID: <3CC62D76.2E0D3A54@engcorp.com>

andrew cooke wrote:
> 
> Which of these packages should I use, and what documentation will help
> me, if I want to:
> 
> - automatically generate a tree from the XML file (I believe this is
> related to the DOM)
> - supply a class to some (already existing) tree walking code that
> does local rearrangements of the XML data
> - automatically generate an XML file from the modified tree
> 
> (I've already looked at XSLT and I don't believe that is the right
> tool for this kind of job).

Why do you say that?  What you describe above is pretty much
what XSLT is for...

-Peter


From bulk_trap at yahoo.com  Fri Apr 19 15:30:13 2002
From: bulk_trap at yahoo.com (bulk_trap)
Date: 19 Apr 2002 12:30:13 -0700
Subject: What webhosting
References: <mailman.1019071656.4705.python-list@python.org>
Message-ID: <9b55104d.0204191130.387c3f39@posting.google.com>

"A" <printers at sendme.cz> wrote in message news:<mailman.1019071656.4705.python-list at python.org>...
> Hi,
> Does anybody know about a webhosting provider ( with Python and 
> Telnet access) that does not remove a user's account if the user 
> sends occasionally some unsolicited emails?

  I've heard Rackspace is pretty spam-friendly.


From phr-n2002a at nightsong.com  Tue Apr  2 21:03:27 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 18:03:27 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
Message-ID: <7xg02dh72o.fsf@ruckus.brouhaha.com>

Guido van Rossum <guido at python.org> writes:
> I'm reading (at least skimming) everything, but the more I read the
> reactions, the more I believe that this is just the response I have to
> expect for *every* change to the language I propose, and the more I'm
> tempted to follow my gut instincts.  Long ago (right after lambda :-)
> I learned to say "no" to proposals that didn't feel right to me, even
> if I couldn't express very clearly why.

I think if you want to take the experience from Python and use it to
redesign the language, that's fine, it's always possible to think of
improvements to put into a successor language no matter how good
the original is.

Introducing all these gradual changes in Python is pretty painful
though.  Bool is one of those features that might have been ok if it
was present from the beginning.  Adding it now just creates another
speed bump for programmers to get used to, and program maintainers to
deal with, and another big hassle for users who have to once again
download a new Python release as programs start using the new
features.

> I think it's time I learn to say "yes" even if the community doesn't
> see the light yet.

I don't think it's appropriate to introduce a change just because
it's a good idea ("good" means > 0, like +0 or +1).  For something
as pervasive as this, it should be +3 or better.  It's ok to say
no to good ideas as well as to bad ones.

I haven't followed this thread too much since Alex's initial long post
on it, which is the best post on the thread that I've seen and made me
feel like not a lot more needed to be said about the subject.  So I
hope you'll take a look at that post if you haven't yet.


From bokr at oz.net  Mon Apr 15 02:06:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 06:06:12 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu>
Message-ID: <a9dqkk$5kf$0@216.39.172.122>

On Sun, 14 Apr 2002 20:00:54 -0700, David Eppstein <eppstein at ics.uci.edu> wrote:

>In article <a9dbc9$bge$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
>wrote:
>
>> All 32-bit ints are represented in hex without a sign, so
>
>This signed-unsigned confusion bit me when I needed a routine to count 
>the nonzero bits of some input.  Normally one would write
>
>    def popCount(x):
>        count = 0
>        while x:
>            count += 1
>            x &= x - 1
>
>but of course this gets into an infinite loop when the high bit is set.  
>Arguably, this is the correct behavior: a negative number in 2's 
>complement should be thought of as having an infinite sequence of 1-bits.
>
>But in my case, my input didn't ever have the high bit set, and a good 
>thing too, or I would have gotten ValueError when I tried 
>int(hexstring,16).
>The reason I was getting into trouble anyway was because of the way 
>right-shifting can change sign:
 ^^^^^ ITYM left?
>
>>>> 0x40000000<<1
>-2147483648
>
>This is inconsistent with int-long unification.  In fact, I think it is 
>a bug.  The answer should, of course, be 2147483648L, just as you would 
>get by multiplying by 2 instead of shifting.
>
I think you're right. I wonder how much code will actually break.

 >>> 0x040000000<<1
 -2147483648

but
 >>> 0x040000000<<1L
 2147483648L

but**2
 >>> 0x80000000>>1L
 -1073741824L

It makes perfect sense when you know what's happening ;-/

With int-long unification, arguably 0x80000000 written plain should be promoted
to 0x8000000L, and you should have to write -0x80000000 if that's what you mean,
unless we distinguish by msb as 0s80000000 negative vs 0s080000000 positive. Maybe 0h is
a better prefix for hex: 0h80000000 + 0h080000000 => 0, and 0hf + 0h1 => 0

I like this representation better than sign+hex(magnitude), because you can read
the actual bits (assuming twos complement representation). The analogous binary
representation would have a mandatory 0 msb for positive numbers too:
 >>> binstr(0x40000000<<1)
 '110000000000000000000000000000000'
 >>> binstr(0x40000000<<1L)
 '010000000000000000000000000000000'
 >>> binstr(0x80000000>>1L)
 '11000000000000000000000000000000'

note that that last line didn't need to be
 '111000000000000000000000000000000'

in fact, it could have been
 '1000000000000000000000000000000'

guess maybe I could fix that, but I think I like the symmetry in
 >>> for n in xrange(8): print '%-6s%-6s' % (binstr(n), binstr(-n))
 ...
 0     0
 01    11
 010   110
 011   101
 0100  1100
 0101  1011
 0110  1010
 0111  1001

Regards,
Bengt Richter


From ykingma at accessforall.nl  Mon Apr 22 16:57:28 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Mon, 22 Apr 2002 21:57:28 +0100
Subject: compiled Jython vs. CPython
References: <f57664b9.0204211807.2235f3bf@posting.google.com>
Message-ID: <3CC4790C.B7495A5E@accessforall.nl>

Jonathan,

"Jonathan P." wrote:
> 
> Has anyone out there tried to compare the
> speed of a jythonc-compiled app to that of a
> CPython one?

This is the best I could find (from June 2001,
I hope someone has more recent info):

http://www.geocrawler.com/archives/3/7018/2001/6/0/6016132/

and the benchmarks comparing Jython+JIT to CPython
are mostly in favour of CPython.

>
> It is possible that if such an app does not
> depend on a runtime interpreter and is run
> by a JIT-JVM that it might give better performance
> than one which runs interpreted (i.e. CPython).

It's possible to compile using CPython to an executable,
and it's possible to compile using jythonc to a jar
including everything needed for execution in a JVM.
However, due to the dynamic nature of python
the compilation results need to include large
parts of the interpreter itself, eg. consider
how one would compile:

exec someStringWithPythonCode

There is no choice for this: here an interpreter has
to be called even from compiled code.

> JIT-compilation techniques have come a long way
> and in fact most benchmarks I've seen show that
> JIT-compiled Java code run faster than
> statically-compiled ones by a substantial
> margin!

An advantage of jython is that it profits from
speedups of the JVM's. Since Java is more widespread
than CPython, there is in general more interest in
speeding up JVM's than in speeding up the CPython
byte code interpreter.

> Could jythonc be *the* Python compiler we wish
> we had?

In both CPython and Jython all python code is compiled
on the fly to byte code before being executed. For imported
modules this byte code is also saved.

Jythonc is used to compile python code to java classes
that can be used by other java code. It allows eg.
to define java method signatures for python methods
and to define a java package for the generated java class.
This is needed for calling jython from java code.
AFAIK the method/function body code (ie. not dependent
on the method signature) generated by jythonc is
the same as the method body code generated on the fly.

However compilation to byte code is only one factor
that determines execution speed.
The speed of byte code execution is another factor,
and this is where Java's JIT helps a lot, but AFAIK
currently not enough to wish we all had java bytecodes
instead of CPython bytecodes.


Regards,
Ype

-- 
email at xs4all.nl


From tanzer at swing.co.at  Wed Apr 10 04:08:34 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Wed, 10 Apr 2002 10:08:34 +0200
Subject: pointer
In-Reply-To: Your message of "Tue, 09 Apr 2002 16:45:08 +0200."
             <a8uukg$fuu$04$1@news.t-online.com>
Message-ID: <E16vD9e-0003IF-00@tswings.swing.cluster>

"Henning Peters" <pete at dexterslab.de> wrote:

> is there anything similar to the pointer concept from c/c++ in python?
>
> i would like to store values in two different locations. both values should
> always be equal. the one could be a pointer on the other. is this possible
> to do this in python?

http://effbot.org/guides/python-objects.htm

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





From nospam at bigfoot.com  Tue Apr 16 09:47:05 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 16 Apr 2002 15:47:05 +0200
Subject: need help to choose an appropriate database
References: <fac97694.0204160518.6f6d79b7@posting.google.com>
Message-ID: <a9h9he$2i1i$1@norfair.nerim.net>

Alexandre,

You forgot to tell on what OS you are currently working.
Start learning how to play with databases in python here...

http://www.python.org/sigs/db-sig/

Personnally, I suggest you to play with MySQL :

* free and open source
* scalable (support for huge databases)
* many 3rd party tools available
* available for Unix/Windows
* python adaptor API identical for Unix and Windows
* free ODBC driver available

Database server and client (for all languages) http://www.mysql.org
Python adapter http://sourceforge.net/project/showfiles.php?group_id=22307

But perhaps you'll make yourself another opinion and choose another solution
better suited to your config and problems.

HTH

--Gilles

"alexandre" <alexandre.herbiniere at caramail.com> a ?crit dans le message
news: fac97694.0204160518.6f6d79b7 at posting.google.com...
> I work in a company who wants to register informations concerning
> customers in a datebase. they suggest me to use CGI and Python to
> capt data. I must send these data in a database and i would like to
> know which ODBC is the most appropriate to that situation ?
>
> thank you



From wware at alum.mit.edu  Thu Apr 11 00:14:52 2002
From: wware at alum.mit.edu (Will Ware)
Date: Thu, 11 Apr 2002 00:14:52 -0400
Subject: Sound programming?
Message-ID: <3CB50DBC.1B0E533F@alum.mit.edu>

Somebody was asking me this evening if there's an easy way to
play a tone of a specified frequency and duration in Python.
I didn't know off the top of my head. I've since poked around
the web a bit and it appears like the Snack Sound Toolkit
(http://www.speech.kth.se/snack/) might be a good way to go.
One of its virtues is being cross-platform, which I suspect is
among his unstated wishes. Are there any other options?


From peter at engcorp.com  Fri Apr  5 22:23:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 05 Apr 2002 22:23:25 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python 
 going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com>
Message-ID: <3CAE6A2D.291C35F2@engcorp.com>

Paul Rubin wrote:
> 
> I like Python a lot but I think it falls a little bit short of truth
> in advertising, when it calls itself a stable, mature language.
> Mature to me means I can download someone's interesting new Python
> script and not have to install a new version of Python in order to run
> it.  

No, that's "stagnant".  

Mature is certainly a word that applies to Python.  The language 
is over a decade old, very rich, well proven in very diverse 
application areas, and on and on.

Stable is also a good adjective for it, since most of the older
code still runs unchanged, and Python from ten years ago is
still very recognizable to new coders today.

Evolving is another word we need to use, though, and that's what 
leads some to feel uncomfortable or downright scared of the situation.

Embrace change!  The changes make Python that much more effective
and strong.  

On the other hand, beware change!  At least for the sake of change.  
I think those driving the changes need to be ever more cautious in 
allowing changes to make it through, and have an ever-rising threshold 
of added value before something is accepted.

-Peter


From roy at panix.com  Sun Apr 21 09:55:07 2002
From: roy at panix.com (Roy Smith)
Date: Sun, 21 Apr 2002 09:55:07 -0400
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com> <a9t177$pi5$1@panix1.panix.com> <a9ufpv$6hb$1@panix1.panix.com>
Message-ID: <roy-D4669D.09550721042002@oldnews.panix.com>

aahz at pythoncraft.com (Aahz) wrote:
> Hmmmm....  it occurs to me that those of us teaching Python will need to
> be *very* careful in making sure that people avoid 'in' over lists.

Why?


From philh at comuno.freeserve.co.uk  Mon Apr  1 19:04:57 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 2 Apr 2002 01:04:57 +0100
Subject: Anyone need a program written?
References: <3ca43ef4.3666136@news.tudelft.nl> <7o_o8.17314$Ou.7311@atlpnn01.usenetserver.com> <slrnaa93j2.kqo.ak@ak.silmarill.org> <3ca4a322@news.nwlink.com> <RHSZTyiDU6p8EwiF@nym.alias.net>
Message-ID: <slrnaahtd9.ddm.philh@comuno.freeserve.co.uk>

On Mon, 1 Apr 2002 01:08:35 +0100, Mindbomb <Mindbomb at localhost.net> wrote:
>>
>>I'm doing this primarily as a way to learn Python (why start small :->),
>>but also so that I can have a news reader that
>>   a) works
>>   b) runs on both my Linux and Windows boxen
>>   c) isn't Outlook Express
>
>If you carry on with this, I'd for one feel there is a gaping hole of 
>newsreaders which actually make it east to post anonymously. If it could 
>use anonymous re-mailers and interface with PGP/GPG it'd be even better!

That's a feature that I'm seriously considering putting 
in Herbivore. Or rather 2 features:

(1) chainer anonymous remailers

(2) anonymous posting to usenet

The problem with the latter is that it might encourage spammers.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From Ajay.Kapoor at cosinecom.com  Wed Apr  3 17:57:58 2002
From: Ajay.Kapoor at cosinecom.com (Ajay Kapoor)
Date: Wed, 3 Apr 2002 14:57:58 -0800 
Subject: Probelm in marshaling Any with typecode/value of struct
Message-ID: <69BCCDDC980B4641BFC908D7BF95F18401FC4266@exchsrv-eng>


Hi ,

      I am trying to pass a structure as CORBA.Any. I am constructing the
TypeCode using the stringified Fnorb id.
      But at the runtime while sending the corba request I get the
marshaling error for CORBA.Any
     Here is the code snippet for what I am doing

    idl definition
-----------------------------------
     module agent {

             interface GenericAgent
                { 

                      struct NameValuePair
                        {
                         string name;
                          any value;
		    };
 		struct Subnet
		{
		    string address;  /* 32 bit mask  */
		    short mask;      /* 0..32        */
		};
           } };

Now I am trying this

  # got the corba object reference some how
     agent =  reference
     sn_tc =CORBA.TypeCode("IDL:cosinecom/agent/common/Subnet:1.0")  #
string is the Fnorb generated ID
     srcSubnet =
agent.NameValuePair("sourceSubnet",CORBA.Any(sn_tc,Subnet("3.3.3.3",24)))
......
     # making the request
      agent.createObject(srcSubnet)



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

I have also tried using
tc=  CORBA.TypeCode(CORBA.id(Subnet) )  to construct the typecode and use it
with Any . But that also failed.

I am getting the following error in all the cases.

 File
"/vobs/orion/cosine/agent_cli/generated/fnidl/cosinecom/agent/__init__.py",
line 85, in createObject
    apply(request.invoke, args, kw)
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/DII.py", line
160, in invoke
    (forwarded, self.__results) = client.synchronous(self, args)
  File
"/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/GIOPClient.py", line
117, in synchronous
    self.__marshal_parameters(cursor, request.inputs(), parameters)
  File
"/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/GIOPClient.py", line
461, in __marshal_parameters
    typecode._fnorb_marshal_value(cursor, parameters[i])
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 2121, in _fnorb_marshal_value
    self._element_type._fnorb_marshal_value(cursor, element)
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 1476, in _fnorb_marshal_value
    member.type._fnorb_marshal_value(cursor, dict[py_member_name])
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 794, in _fnorb_marshal_value
    value._fnorb_marshal(cursor)
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/Any.py", line
77, in _fnorb_marshal
    self.__typecode._fnorb_marshal(cursor)
  File "/dvlp/akapoor/Python-1.5.2/Lib/site-packages/Fnorb/orb/TypeCode.py",
line 216, in _fnorb_marshal
    cursor.marshal('L', self._kind.value()) 
AttributeError: 'string' object has no attribute 'value'


Can somebody help me.

Thanx
Ajay Kapoor

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020403/72754ed0/attachment.html>

From geoff at gerrietts.net  Thu Apr  4 20:47:32 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Thu, 4 Apr 2002 17:47:32 -0800
Subject: Catalog of Python Patterns? [LONG!]
In-Reply-To: <a8h4re$274k$1@news.idiom.com>
References: <mailman.1017886389.26908.python-list@python.org> <a8h4re$274k$1@news.idiom.com>
Message-ID: <20020405014732.GB17563@isis.gerrietts.net>

Quoting Jim Dennis (jimd at vega.starshine.org):
>  I disagree (with one followup that's not quoted herein) that
>  patterns are not applicable to Python.  The author of that comment
>  claimed that GoF was a C++ specific book (which is untrue, as I 
>  point out above).  Python is much more like Smalltalk than C++,
>  and the GoF specific chose it because it is a dynamic language
>  --- for its contrast to C++ and its ilk.

I think -- unless you are referring to a post that isn't in this
thread -- that you're misconstruing the comment. The idea is that some
of the patterns in the GoF don't really apply or aren't particularly
useful when you're talking about Python and not C++ or some other
strongly-typed language.

I think that you could make a case for the limited applicability of
Command pattern, the Singleton pattern, the Prototype pattern, and I'm
sure several others. You could implement them, but there's not a solid
reason to do so, because Python doesn't impose the restrictions that
the patterns are designed to circumvent.

Others have recommended Thinking In Python. I'm also enjoying that
book; you probably will as well.

Thanks,
--G.

-- 
Geoff Gerrietts                <geoff at gerrietts net>
"I have read your book and much like it." --Moses Hadas




From guido at python.org  Thu Apr  4 09:16:15 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:16:15 GMT
Subject: [Python-Dev] ACCEPTED: PEP 285 (truth pickles)
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com> <mailman.1017895453.22508.python-list@python.org> <roy-857582.08132604042002@news1.panix.com>
Message-ID: <3CAC610F.D20A25BB@python.org>

Roy Smith wrote:
> If I pickle a bool, and then restore it, do I get the same singleton back?

Yes.  Ditto for marshal.

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


From hst at empolis.co.uk  Wed Apr  3 07:49:18 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Wed, 3 Apr 2002 13:49:18 +0100
Subject: PEP 285: Adding a bool type
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220B58@hendrix.empolisuk.com>

> In article <ualr7hmlhkemab at news.supernews.com>, John Roth
> <johnroth at ameritech.net> writes
> .....
> >Consequently, I think of it (0 and  1 as false and true) as a
> >hack carried over from C and similar languages.
> >
> >John Roth
> C is a very successful language. I don't think of its conditional
> testing as a hack, but then I'm an engineer. What works is good and
> Python's current conditional testing seems to work very well and
> moreover efficiently in many cases of interest.
> 
> Forcing all these into the (true, false) set using a Boole (if we're
> going to honour the man let's spell him right) variable seems 
> to put too
> much emphasis on formalism. It also ignores the eminent sense of not
> fixing unbroken things and not adding yet another current to 
> the Python
> type cake. 
> -I wish I had Kalmanian filter types-ly yrs- 
> Robin Becker

Couldn't agree more with you Robin. The PEP is definitely a -1 for me as well

 http://mail.python.org/mailman/listinfo/python-list

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.



From shalehperry at attbi.com  Tue Apr  2 21:26:38 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 02 Apr 2002 18:26:38 -0800 (PST)
Subject: How to wait only so long for a server to respond?
In-Reply-To: <nwtq8.11707$ml2.924462@newsread1.prod.itd.earthlink.net>
Message-ID: <XFMail.20020402182638.shalehperry@attbi.com>

On 03-Apr-2002 Ralph wrote:
> Pythoners:
> 
> I have some code that requests a response from a server on the Internet, 
> but I only want to wait so long for the server to respond.  How can I tell 
> python to wait x seconds/minutes, then move on to the next section of code 
> if the server has not responded?
> 
> Thanks as always.
> 

a select loop is a common way to wait for data on a socket and specify a
timeout.  Look for documentation on select().



From aahz at pythoncraft.com  Wed Apr 24 00:41:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Apr 2002 00:41:45 -0400
Subject: Python's Lisp heritage
References: <mailman.1019613886.1762.python-list@python.org> <m3662henr9.fsf@chvatal.cbbrowne.com>
Message-ID: <aa5d29$nhi$1@panix1.panix.com>

In article <m3662henr9.fsf at chvatal.cbbrowne.com>,
Christopher Browne  <cbbrowne at acm.org> wrote:
>
>If all you have is lexical scoping, parameters have to get explicitly
>passed down the chain in order to get from function A to function C.
>That means introducing additional parameters to function B, which
>didn't actually care about those extra values.

The Pythonic response to this need is to pass a single mutable object
that itself is extensible: list, dict, class/instance, module (this
would actually be a global, mostly).

Oh, I just realized something.  You can use a *class* as a Singleton.
<smirk>  That's probably a better idea than imp.new_module().
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From missive at frontiernet.net  Sat Apr  6 17:57:35 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sat, 06 Apr 2002 22:57:35 -0000
Subject: help:problem using the RawPen class/turtle module.
References: <8374ee10.0204060303.3333260a@posting.google.com>
Message-ID: <slrnaauvbv.3v.missive@whave.frontiernet.net>

On 6 Apr 2002 03:03:14 -0800, alex <ale_xp at yahoo.com> wrote:
> Hi all, I'm a Python beginner, and I'm trying to use
> the module "turtle" to do some turtle graphics.
> But I have a problem when I try to use the turtle
> class RawPen in order to use it with a Canvas. 
> In particular when I try to run the following code: ...
> 
> ...
> root = Tkinter.Tk()
> c       =   Tkinter.Canvas(root,heigh=200,width=300)
                                  ^^^^^
            just a transcription typo, right?

> c.pack()
> mypen   =   turtle.RawPen(c)
> ... 
> 

I tried this at the interactive python prompt:

import turtle
import Tk
root=Tkinter.Tk()
c=Tkinter.Canvas(root)
c.pack(expand=1)
mypen=turtle.RawPen(c)


and got no errors.

This was with python2.1 configured with Tkinter

So it may be a configuration (rather than a
usage or syntax) problem.




From mark.charsley at REMOVE_THIS.radioscape.com  Tue Apr  9 12:43:00 2002
From: mark.charsley at REMOVE_THIS.radioscape.com (Mark Charsley)
Date: Tue, 9 Apr 2002 17:43 +0100 (BST)
Subject: Beginner's performance problem
Message-ID: <memo.20020409174302.1584B@a.a>

For reasons best not gone into, I needed to correct the case of a whole 
bunch of SQL code. As such I created the little script below...

It reads in a source file, then for each line it does a case-insensitive 
search for each TableName and ColumnName (as contained in the "names" 
collection), checks that it's not just matching a substring, and then 
corrects the case of the word if necessary.

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

import os
import string

names = [
    "ALGORITHM_CLASS",
    "ALGORITHM_CONSTRAINT",
# 200-odd other names snipped...
    "Width",
    "WorkUnitHandle"
    ]

def ProcessFile(filename) :
    sourceFile = open(filename)
    sourceLines = sourceFile.readlines()
    sourceFile.close()


    numChanges = 0

    for lineNo in range(len(sourceLines)) :

        changed = 0
        upperSource = string.upper(sourceLines[lineNo])

        for name in names :

            upperName = string.upper(name)
            startPos = 0
            while (startPos != -1) :
                startPos = string.find(upperSource,upperName,startPos +1)
                if (startPos >0) :
                    previousChar = upperSource[startPos-1:startPos]
                    nextChar = upperSource[startPos + len(upperName) :\
			startPos + len(upperName)+1]
                    
                    if  not previousChar.isalnum() and \
                        not previousChar == "_" and \
                        not nextChar.isalnum() and \
                        not nextChar == "_" :
                        # OK upperName isn't a substring of the string 
			# we've found
                        if not (sourceLines[lineNo][startPos: startPos + \
                          len(upperName)] == name) :
                            myNewLine = sourceLines[lineNo][:startPos] + \
                              name + sourceLines[lineNo][startPos + \
                              len(upperName):]
                            
                            sourceLines[lineNo] = myNewLine
                            changed = 1
                            print "#",

        if changed : numChanges += 1

    print
    print filename + " : " + ("%d" % numChanges) + " changed lines"

    sourceFile = open(filename,"w+")
    sourceFile.writelines(sourceLines)
    sourceFile.close()


ProcessFile("ThirteenThousandLines.sql")

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

It's not a complicated (or particularly elegant) script but it runs rather 
slowly - taking about a minute to run on a development spec box.

I (an experienced C++ developer, beginning python programmer) have two 
main questions:

1) why is it so slow? I wouldn't have thought that processing 13,000 lines 
* 200 names shouldn't take more than a few seconds on a modern PC. PLaying 
around with simplified versions seems to indicate that calling 
string.find() repeatedly on a list of 13000 strings is a lot slower than 
calling string.find() the same number of times on a single large string 
containing all the source file. I avoided that approach however because 
I'd expect calling "bigString = bigString[:pos] + replacement + 
bigString[pos + len(replacement):]" repeatedly _would_ slow things down.

2) what's the best way of finding inefficiencies in python code (other 
than asking here?). Are there any profilers out there for python?


Many TIA

-- 

Mark - personal opinion only, could well be wrong, not representing 
company, don't sue us etc. etc.


From billludden at email.msn.com  Tue Apr 23 22:26:19 2002
From: billludden at email.msn.com (billludden)
Date: Tue, 23 Apr 2002 20:26:19 -0600
Subject: Newbie: Running a script on WindowsXP
References: <#dmZtdu6BHA.1188@cpimsnntpa03>
Message-ID: <OrwUhez6BHA.1572@cpimsnntpa03>

Thank you all for your help.

Bill




From Oschler at earthlink.net  Tue Apr  2 17:22:42 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 2 Apr 2002 17:22:42 -0500
Subject: Associate PY files with Python (KDE/Linux)
Message-ID: <uaqq8.12467$ku5.1078856@e3500-atl2.usenetserver.com>

I'm running KDE 2.x under SUSE 7.3 Pro.  I'd like to set up a file
association so that when I click on a .py file in Konqueror it loads Python
with the py file ready to go.  Anyone know how?

thx






From ssiebinga at fa.knaw.nl  Tue Apr 16 08:20:01 2002
From: ssiebinga at fa.knaw.nl (sjoerd siebinga)
Date: 16 Apr 2002 05:20:01 -0700
Subject: problem negative lookahead assertion
References: <4b926dbb.0204151202.6d709633@posting.google.com> <mailman.1018902689.4821.python-list@python.org>
Message-ID: <4b926dbb.0204160420.4e9e858d@posting.google.com>

Thanks Geoff and Amk for your advise.

Sadly enough my input data requires nested {} due to specialist
diacritics use.So the ([^}]*?) option gives more errors when
processing the data through the latex interpreter. I used the
([\s,\W]) regex to grab all the curly braces in  the emph.

Now it is time to ask a silly question. I am a trained linguist and
not a programmer.

> Note that this will break in a different situation, if you have {} inside  
> the contents of \emph, as for example in \emph{ab^{cd}}, because this 
> problem really needs a full parser to handle the general case.

What do you mean by a 'full parser'?

As Geoff asked I will send  a couple of samples. My Python script is
rather too large to be send to the newsgroup. If either of you want to
see it, I can send it to you with a couple of sample lemmata. Another
point is that the substitutions are preformed in turn much like an sed
script. The emph regex is processed last.


Input before processing.

\begin{lemma}\begin{entry}wald\end{entry}\begin{seman}approval\end{seman}\begin{grammar}subst.f.\end{grammar}
\begin{level}PIE\end{level} \begin{misc} On the Ds B \emph{weld(e)},
cf. Aofr. 143-6.\end{misc} \begin{pfris}*wald\end{pfris}
\begin{germdata} ON \emph{vald}, OE \emph{weald}, OS \emph{gi]wald},
OHG \emph{gi]walt}, MHG \emph{(ge)walt}, MLG \emph{(ge)walt}, 
\emph{(ge)w\=o lt}, ODu. \emph{ge]walt}, MDu. \emph{(ge)wout},
\emph{(ge)wolt}, \emph{(ge)walt}, \emph{ge]welt} `power', cf.  also \#
str. VII: Goth. \emph{waldan}, ON \emph{valda}, OE \emph{wealdan}, OS
\emph{waldan}, OHG \emph{waltan},  MHG \emph{walten}, MLG
\emph{w\textbrevemacron{a}lden}, \emph{w\=olden}, OFris. \emph{walda}
(q.v.), ODu.  \emph{waldan}, MDu. \emph{wouden} `rule'; \# *waldiga-:
OS \emph{gi]weldig}, OHG \emph{gi]welt\={\i}g}, \emph{gi]walt\={\i}g},
MHG \emph{(ge)waltec}, \emph{-waltic}, \emph{-weltic}, MLG
\emph{(ge)w\textbrevemacron{a}ldich}, \emph{-weldich},
\emph{-w\=oldich}, \emph{-w\"oldich}, OFris.  \emph{weldech} (q.v.),
MDu. \emph{(ge)weldich} `powerful', with denom. OFris. 
\emph{weldegia} (q.v.), MLG \emph{weldigen}, \emph{waldigen}
`adjudge', etc.\end{germdata} \begin{pgerm}*waldi-\end{pgerm}  
\begin{pie}*ulh$_2$-d\textsuperscript{h}-\end{pie} \begin{pokpage}
1111-2\end{pokpage} \begin{piemisc} Toch. A \emph{w\"al}, B
\emph{walo} `king', Lat. \emph{val\=e\=o} `be strong'; with a dental
extension as  in Gmc. Lith. \emph{v\'{e}ldu}, \emph{veld\d{e}?ti}
`rule'\end{piemisc} \begin{biblio}Seebold 536-7\end{biblio}
\begin{seealso} walda ,  weldich ,  weldegia \end{seealso}  
\end{lemma}

Input after processing without the emph substitution

\begin{lemma}
\begin{entry}wada \index{r1~wada} \end{entry}
\begin{seman}wade\end{seman}
\begin{grammar}str.vb.VI\end{grammar}
 \begin{level}PIE [?]\end{level}  
\begin{pfris}*wada \index{pfris~*wada} \end{pfris} 
\begin{germdata} ON \emph{va\symbol{'360}a}
\index{on~va\symbol{'360}a}`wade, rush, walk through', OE
\emph{wadan}, \index{oe~wadan} OHG, MHG \emph{watan} 
\index{mhg~watan}`wade, stride', MLG \emph{w\=aden},
\index{mlg~w\=aden} MDu. \emph{waden}, \index{mdu~waden} \emph{waeyen}
`wade, go'\end{germdata}
\begin{pgerm}*wadanaN \index{pgerm~*wadanaN} \end{pgerm}  
\begin{pie}*uh$_2$d\textsuperscript{h}- [?]
\index{pie~*uh$_2$d\textsuperscript{h}- [?]} \end{pie}
\begin{piemisc} Semantically close is Lat. \emph{vad\=are} 
\index{lat~vad\=are}`to wade through, ford', \emph{vadum} `ford'. It
cannot be established with certainty whether Alb. \emph{va} 
\index{alb~va}`ford' is historically cognate with \emph{vadum} or is a
borrowing of this Lat. form (Demiraj 1997: 405). The Latin forms seem
to be cognate with \emph{v\=adere} `go, walk, rush'. If we allow a
connection with words denoting a less specific way of moving than
wading in particular, like in this Lat. instance, Arm. \emph{gam} 
\index{arm~gam}`come' could furthermore be compared, pointing to a PIE
heritage [POK 1109]. Note also that less specific meanings are
attested within Gmc., where the ON verb can also mean `rush, walk
through' and the OE, OHG, MHG, MLG and MDu. verbs can also denote the
notion `go, stride'.\end{piemisc}
\begin{biblio}Schrijver 170; Seebold 530-1\end{biblio}   
\end{lemma}

After processing with the emph regex

\begin{lemma}
\begin{entry}wald \index{r1~wald} \end{entry}
\begin{seman}approval\end{seman}
\begin{grammar}subst.f.\end{grammar}
 \begin{level}PIE\end{level}
 \begin{misc} On the Ds Toch. B \emph{weld(e)}, \index{unl~weld(e)}
\index{tochb~weld(e)} \index{tochb~weld(e)} cf. Aofr. 143-6.\end{misc}
 \begin{pfris}*wald \index{pfris~*wald} \end{pfris}
 \begin{germdata} ON \emph{vald} \index{on~vald} OE \emph{weald},
\index{unl~weald} \index{oe~weald} OS \emph{gi]wald},
\index{unl~gi]wald} \index{os~gi]wald} OHG \emph{gi]walt},
\index{unl~gi]walt} \index{ohg~gi]walt} MHG \emph{(ge)walt},
\index{unl~(ge)walt} \index{mhg~(ge)walt} MLG \emph{(ge)walt},
\index{unl~(ge)walt} \index{mlg~(ge)walt} \emph{(ge)w\=o lt},
\index{unl~(ge)w\=o lt} ODu. \emph{ge]walt}, \index{unl~ge]walt}
\index{odu~ge]walt} MDu. \emph{(ge)wout}, \index{unl~(ge)wout}
\index{mdu~(ge)wout} \emph{(ge)wolt}, \index{unl~(ge)wolt}
\emph{(ge)walt}, \index{unl~(ge)walt} \emph{ge]welt} 
\index{unl~ge]welt}`power', cf. also \# str. VII: Goth. \emph{waldan},
\index{unl~waldan} \index{goth~waldan} ON \emph{valda}
\index{on~valda} OE \emph{wealdan}, \index{unl~wealdan}
\index{oe~wealdan} OS \emph{waldan}, \index{unl~waldan}
\index{os~waldan} OHG \emph{waltan}, \index{unl~waltan}
\index{ohg~waltan} MHG \emph{walten}, \index{unl~walten}
\index{mhg~walten} MLG \emph{w\textbrevemacron{a}lden},
\index{mlg~w\textbrevemacron{a}lden} \emph{w\=olden},
\index{unl~w\=olden} OFris. \emph{walda}  \index{unl~walda}
\index{ofris~walda} \index{ofris~walda}(q.v.), ODu. \emph{waldan},
\index{unl~waldan} \index{odu~waldan} MDu. \emph{wouden} 
\index{unl~wouden} \index{mdu~wouden}`rule'; \# *waldiga-: OS
\emph{gi]weldig}, \index{unl~gi]weldig} \index{os~gi]weldig} OHG
\emph{gi]welt\={\i}g}, \index{ohg~gi]welt\={\i}g}
\emph{gi]walt\={\i}g}, MHG \emph{(ge)waltec}, \index{unl~(ge)waltec}
\index{mhg~(ge)waltec} \emph{-waltic}, \index{unl~-waltic}
\emph{-weltic}, \index{unl~-weltic} MLG
\emph{(ge)w\textbrevemacron{a}ldich},
\index{mlg~(ge)w\textbrevemacron{a}ldich} \emph{-weldich},
\index{unl~-weldich} \emph{-w\=oldich}, \index{unl~-w\=oldich}
\emph{-w\"oldich}, \index{unl~-w\"oldich} OFris. \emph{weldech} 
\index{unl~weldech} \index{ofris~weldech} \index{ofris~weldech}(q.v.),
MDu. \emph{(ge)weldich}  \index{unl~(ge)weldich}
\index{mdu~(ge)weldich}`powerful', with denom. OFris. \emph{weldegia} 
\index{unl~weldegia} \index{ofris~weldegia}
\index{ofris~weldegia}(q.v.), MLG \emph{weldigen},
\index{unl~weldigen} \index{mlg~weldigen} \emph{waldigen} 
\index{unl~waldigen}`adjudge', etc.\end{germdata}
 \begin{pgerm}*waldi- \index{pgerm~*waldi-} \end{pgerm}  
\begin{pie}*ulh$_2$-d\textsuperscript{h}-
\index{pie~*ulh$_2$-d\textsuperscript{h}-} \end{pie}
\begin{pokpage} 1111-2\end{pokpage} 
\begin{piemisc} Toch. A \emph{w\"al}, \index{unl~w\"al}
\index{tocha~w\"al} Toch. B \emph{walo}  \index{unl~walo}
\index{tochb~walo} \index{tochb~walo}`king', Lat. \emph{val\=e\=o} 
\index{unl~val\=e\=o} \index{lat~val\=e\=o}`be strong'; with a dental
extension as in Gmc. Lith. \emph{v\'{e}ldu}, \index{lith~v\'{e}ldu}
\index{lith~v\'{e}ldu} \emph{veld\d{e}? \index{unl~veld\d{e}ti}
`rule'\end{piemisc}
\begin{biblio}Seebold 536-7\end{biblio} 
\begin{seealso} walda , weldich , weldegia \end{seealso}  
\end{lemma}

and what I wanted to see was

\begin{lemma}
\begin{entry}wada \index{r1~wada} \end{entry}
\begin{seman}wade\end{seman}
\begin{grammar}str.vb.VI\end{grammar}
 \begin{level}PIE [?]\end{level}  
\begin{pfris}*wada \index{pfris~*wada} \end{pfris} 
\begin{germdata} ON \emph{va\symbol{'360}a}
\index{on~va\symbol{'360}a}`wade, rush, walk through', OE
\emph{wadan}, \index{oe~wadan} OHG, MHG \emph{watan} 
\index{mhg~watan}`wade, stride', MLG \emph{w\=aden},
\index{mlg~w\=aden} MDu. \emph{waden}, \index{mdu~waden} \emph{waeyen}
<<<\index{unl~waeyen}>>> `wade, go'\end{germdata}
\begin{pgerm}*wadanaN \index{pgerm~*wadanaN} \end{pgerm}  
\begin{pie}*uh$_2$d\textsuperscript{h}- [?]
\index{pie~*uh$_2$d\textsuperscript{h}- [?]} \end{pie}
\begin{piemisc} Semantically close is Lat. \emph{vad\=are} 
\index{lat~vad\=are}`to wade through, ford', \emph{vadum}
<<<\index{unl~vadum}>>>`ford'. It cannot be established with certainty
whether Alb. \emph{va}  \index{alb~va}`ford' is historically cognate
with \emph{vadum} <<<\index{unl~vadum}>>> or is a borrowing of this
Lat. form (Demiraj 1997: 405). The Latin forms seem to be cognate with
\emph{v\=adere} <<<\index{unl~v\=adere}>>> `go, walk, rush'. If we
allow a connection with words denoting a less specific way of moving
than wading in particular, like in this Lat. instance, Arm. \emph{gam}
 \index{arm~gam}`come' could furthermore be compared, pointing to a
PIE heritage [POK 1109]. Note also that less specific meanings are
attested within Gmc., where the ON verb can also mean `rush, walk
through' and the OE, OHG, MHG, MLG and MDu. verbs can also denote the
notion `go, stride'.\end{piemisc}
\begin{biblio}Schrijver 170; Seebold 530-1\end{biblio}   
\end{lemma}

The forms that should have been substituted are inclosed between <<<
and >>>.

regards sjoerd


From pyth at devel.trillke.net  Mon Apr 29 13:12:04 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 19:12:04 +0200
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
In-Reply-To: <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1>
References: <aaiu6e$pfr$1@bob.news.rcn.net> <mailman.1020082610.7467.python-list@python.org> <Xns91FF9515A57Eduncanrcpcouk@127.0.0.1>
Message-ID: <20020429171204.GM16217@devel.trillke>

On Mon, Apr 29, 2002 at 01:40:18PM +0000, Duncan Booth wrote:
> > Btw, what is the best/shortest way to express 'filter' and 'reduce' with 
> > list comprehensions?
> 
>  filter(fn, lst)
> becomes:
>  [ v for v in lst if fn(v) ]
> 
>  filter(None, lst)
> becomes:
>  [ v for v in lst if v ]
> 
> If the second argument to filter is a string or tuple then you have to 
> convert the resulting list back to the appropriate type:
> 
> e.g.
>  filter(fn, aString)
>  str.join('', [ v for v in aString if fn(v) ])

You really feel that these free floating strings with
no clear separations and lots of micro-management where
you dont't even know what does what and where the parts
start to float around and you ask yourself 
when the sentence will ever end 

come even close to 

    filter(fn,list)  
    
???

We must be living in different worlds, then.

Just because list comprehension is a useful tool
doesn't make it 'more readable' than the often
used filter/map/reduce functions. 

This has *nothing* to do with lambda and its
reasons for existence and i would rather like
to keep the concerns separate.

    holger




From logstx at bellatlantic.net  Tue Apr  2 00:44:54 2002
From: logstx at bellatlantic.net (logistix)
Date: Tue, 02 Apr 2002 05:44:54 GMT
Subject: PEP 285: Adding a bool type: yes, but not as int subtype
References: <mailman.1017704644.22152.python-list@python.org>
Message-ID: <qzbq8.3238$tW2.2164@nwrddc04.gnilink.net>

>
> >>> 2-1  # doesn't change
> 1
> >>> 2<1  # changes to False
> 0
> >>>
>
> ?  Probably, but I don't really care -- nobody tried to make computers
easy
> for me, and suffering builds character <wink>.  I've been at this for
> decades now, and I'll need a millisecond less thought.
>

So this PEP will break the ternary hack (a and b or c) and other crazy
exploitation of short-circuiting tricks in lambdas? Not that I mind. ;)





From starx at pacbell.net  Wed Apr 24 16:41:14 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 20:41:14 GMT
Subject: Quick: Format Specs in %
Message-ID: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>

'%2f' % some_number pads the number with 2 blank spaces
'%.2f' % some_number fixes the number to 2 decimal places

the problem:
'%2.2f' % some_number only does the second of those two things. I need 
both. I tried looking up the % operator in the index of the language 
reference, but couldn't find it.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From jepler at unpythonic.net  Wed Apr 10 16:42:32 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 10 Apr 2002 15:42:32 -0500
Subject: Creating objects in thread created with subclass
In-Reply-To: <NDBBLAEMOMEHKALAMKKJIEKOKAAA.narnett@mccmedia.com>
References: <MPG.171d20a7416584ee9896a5@news.nwlink.com> <NDBBLAEMOMEHKALAMKKJIEKOKAAA.narnett@mccmedia.com>
Message-ID: <20020410154232.C27429@unpythonic.net>

On Tue, Apr 09, 2002 at 06:18:36PM -0700, Nick Arnett wrote:
> self.dbh.execute("INSERT INTO Foo (fieldnames) VALUES (%s,[%s])",(data[1:]))

Construct a string of as many "%s,..." as you need, substitute it into
the command you pass to .execute():

    format = ",".join(["%s"] * len(data[1:]))
    command = "INSERT INTO Foo (fieldnames) VALUES (%s)" % format
    self.dbh.execute(command, data[1:])

Jeff




From guido at python.org  Wed Apr  3 09:37:02 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 03 Apr 2002 09:37:02 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Wed, 03 Apr 2002 16:03:45 +0200."
             <3CAB0BC1.20B90FED@lemburg.com> 
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com>  
            <3CAB0BC1.20B90FED@lemburg.com> 
Message-ID: <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>

> Right, but my point is that if you want to make booleans first class
> objects in Python, you should consider making them behave as defined
> in the text books rather than trying to make them feel like integers
> but look like truth values.

I reject this reasoning.  Python is not trying to follow the text
books.  (In fact, maybe I should say "Python is trying not to follow
the text books.")

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



From philh at comuno.freeserve.co.uk  Mon Apr 22 09:39:01 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 22 Apr 2002 14:39:01 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <slrnac84jl.1c6.philh@comuno.freeserve.co.uk>

On Mon, 22 Apr 2002 12:57:38 +0100, Gon?alo Rodrigues <op73418 at mail.telepac.pt> wrote:
>
>You should understand that this is a philosophical viewpoint. The
>mathematical field is roughly divided in the Platonist field, those who
>believe that mathematical objects exist somewhere out there in space (to
>quote Sun Ra) in a Wonderful World of Platonic Ideas,

This is true to the extent that different mathematical cultures 
re-invent the same ideas. E.g. both middle eastern and central 
American societies independently invented multiplication.

> and the formalist
>field of those who believe mathematics is essentialy a game we play on
>paper with symbols with no "real" meaning attached to it besides the
>convenience in describing natural phenomena.

This is true, too.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From tundra at tundraware.com  Mon Apr 22 01:50:03 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 05:50:03 GMT
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
Message-ID: <3CC3A2E0.B1AA0F70@tundraware.com>

Carl Banks wrote:

<SNIP>

> Computer programming is not science.

Agreed - it is craft.

> The study of algorithms, their properties, etc., is science.

Disagree - it is Mathematics.  Science, almost by definition, involves
hypothesis, experiment, and analysis of a sort almost never done in
algorithmics, complexity theory, or combinatorics.

Even in the more abstruse corners of modern science, like Physics, while
the majority of the work is almost entirely theoretical, there is still
the underlying belief in the importance of empirical verification.

I realize this all is splitting hairs to some degree because it is essentially
impossible to decouple mathematics from science at this point in human
history, but this still does not mean they are one and the same thing.

So, in my view, programming is *craft* but might someday become an
*engineering discipline* (very doubtful IMHO), thinking about how programs
work is *mathematics*, and the only when computers are applied for some
empirical purpose can we really call what they do *science*.  

Then again, its much easier to mooch grant money if you call something 
"science", I guess...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From logiplexsoftware at earthlink.net  Mon Apr 29 21:32:33 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 18:32:33 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <b03e80d.0204291719.1f36c42b@posting.google.com>
References: <aaim48$7fd$1@bob.news.rcn.net>
	<Xns91FF62618790Flumberjack@216.148.53.85>
	<b03e80d.0204291719.1f36c42b@posting.google.com>
Message-ID: <20020429183233.753e3455.logiplexsoftware@earthlink.net>

On 29 Apr 2002 18:19:11 -0700
Sandy Norton wrote:

> <snip unecessarily hostile comments from lumberjack>
> 
> I'm sure you know you can communicate 100% of your point with less
> than 1% of your sarcasm. I'd hate to see people like you dissuade
> future pep writers from articulating their ideas to the community.

Especially since lumberjacks are supposed to be "okay".

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From PoulsenL at capecon.com  Mon Apr 29 13:26:25 2002
From: PoulsenL at capecon.com (PoulsenL at capecon.com)
Date: Mon, 29 Apr 2002 13:26:25 -0400
Subject: Remove masked elements from an array
Message-ID: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>

I have a Numeric array with missing elements from which I wish to calculate
a histogram.  However, in order to do so I must first remove the 'None'
values from the array.  I have a kludgy solution, but I was wondering if
someone could point me in the right direction on what is the proper method
of collapsing a masked array to just the non-masked values.

Thanks 

Loren




From hgg9140 at seanet.com  Thu Apr  4 21:35:27 2002
From: hgg9140 at seanet.com (Harry George)
Date: 04 Apr 2002 18:35:27 -0800
Subject: Rule Engine in Python
References: <eb612768.0203101940.5565b282@posting.google.com> <m3it82d32l.fsf@wilma.localdomain> <I45r8.34886$0T6.10550048@e3500-atl1.usenetserver.com>
Message-ID: <m3u1qq3ma8.fsf@wilma.localdomain>

"Robert Oschler" <Oschler at earthlink.net> writes:

> Harry George <hgg9140 at seanet.com> wrote in message
> news:m3it82d32l.fsf at wilma.localdomain...
> >
> > If you are willing to use prolog as the rules language, there are
> > python/prolog bindings (e.g., SWI prolog).  We use python and prolog
> > jointly in a C++ CADenvironment.
> >
> 
> Harry,
> 
> I'm a long term user of LPA Prolog, how does SWI Prolog compare and do you
> know what the cost would be if I used SWI Prolog on a single Linux box
> acting as a web server (that was consulting the Prolog Rule engine, accessed
> via Python scripts running under Zope).
> 
> thx
> 
>

For cost=money: In our mixed C++/prolog/python code, we are using ALS
prolog (commercial).  I don't know the current SWI license (I think it
went open source, but check).

For cost=cpu resources: We've never done profiling -- it was always
fast enough once we got it to work at all.  That is on SGI
workstations.  Getting it to work at all of course is the payoff for
python over prolog for many tasks.  A few lines of python doing
regualr expressions and some text manipulation takes the palce of a
page or so of prolog.  That was all with ALS -- I've used SWI for a
few simple examples, but never compared them for heavy processing.
 
> 
> 

-- 
Harry George
hgg9140 at seanet.com


From l0819m0v0smfm001 at sneakemail.com  Wed Apr  3 23:39:48 2002
From: l0819m0v0smfm001 at sneakemail.com (Joshua Macy)
Date: Thu, 04 Apr 2002 04:39:48 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017874455.22890.python-list@python.org>
Message-ID: <3CAB9F6D.9050600@sneakemail.com>

Ka-Ping Yee wrote:

>
> Guido van Rossum wrote:
> 
>>No.  That would break backwards compatibility.  False==0, and True==1;
>>everything else follows from that.  (But False is not 0, and True is
>>not 1!)
>>
> 
> This is a strange distinction to make -- one that is not made in any
> other programming language i have ever encountered.  Of course *i* get
> it and *you* get it -- but it will be so hard to teach this particular
> weirdness that i consider it fatal to the proposal.
> 
>


   Nobody has any business testing using "is" without a clear 
understanding of object identity (that's its whole purpose)--and with 
that understanding it's blindingly obvious why False is not 0 and True 
is not 1.

If you want weird, try:

 >>> 1 is 1
1
 >>> a = 1
 >>> a is 1
1
 >>> 1001 is 1001
1
 >>> b = 1001
 >>> b is 1001
0


Is it obvious without explanation what's going on here?  Of course not, 
but that doesn't make it a fatal flaw in Python. I don't even believe 
it's hard to teach, but I reiterate that until it's taught new Python 
programmers should avoid testing with "is"--it will bite them almost 
immediately. Stick to == and there aren't any problems.


Joshua



From roy at panix.com  Sat Apr  6 09:39:17 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 06 Apr 2002 09:39:17 -0500
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <roy-B0CC5C.09391706042002@news1.panix.com>

Alex Martelli <aleax at aleax.it> wrote:
>  Maybe when you release 3.0 (to pick
> a nice round number) you could release 3.1 at the same time (just
> about the same, perhaps) and decree 3.0.* the stable track until
> further notice, 3.1.* the experimental one -- or, pick whatever
> numbering scheme you prefer.

Isn't this what happened with 1.6 and the 2.0 alpha?  I seem to remember 
they were announced on the very same day.  As I recall, however, the 1.6 
release was advertised as "Well, we're contractually obligated to do it, so 
here it is, but we don't actually expect anybody will use it".


From mark at freelance-developer.com  Wed Apr 17 08:37:33 2002
From: mark at freelance-developer.com (Mark Nenadov)
Date: Wed, 17 Apr 2002 08:37:33 -0400
Subject: Problem Writing Files
References: <3CBCE07C.9A0E0C4@slac.stanford.edu>
Message-ID: <KXdv8.23501$Yt.1175935@read1.cgocable.net>

Roger Jones wrote:

> I am using ActiveState Python 2.1 under Windows NT  and after opening a
> file with:
> 
> afile=('test.dat','w')
> afile.write('some random test text')
> afile.read()
> 
> It fills the file with many x's and 0's and also with various other
> text.
> How can this action be fixed so that it writes the text properly?
> 
> Thanks!
> 
> -Roger Jones
> 
> 

You are missing something in your first line. Instead of doing 
"afile=('test.dat', 'w')" you should do "afile=open('test.dat', 'w')".

The problem is that you are creating a tuple (with the elements 'test.dat' 
and 'w') instead of an actual file object.

-- 
-- Mark Nenadov (http://www.freelance-developer.com)


From phd at phd.pp.ru  Tue Apr  9 05:10:50 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 13:10:50 +0400
Subject: exception question
In-Reply-To: <a8ua60$7nd$06$1@news.t-online.com>; from pete@dexterslab.de on Tue, Apr 09, 2002 at 10:56:03AM +0200
References: <a8ua60$7nd$06$1@news.t-online.com>
Message-ID: <20020409131050.G8297@phd.pp.ru>

On Tue, Apr 09, 2002 at 10:56:03AM +0200, Henning Peters wrote:
> i am asking me, if it is possible to catch an exception and to save the
> exception value in a variable?!
> 
> try:
> ...
> except: __all_exceptions__, error:
> ...

   Two ways to do it:

1:
try:
   ...
except:
   e1, e2, e3 = sys.exc_info()

2:
try:
   ...
except Exception, e:
   ...

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




From lac at strakt.com  Sat Apr 27 08:32:55 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 27 Apr 2002 14:32:55 +0200
Subject: Stackless goes Limbo 
In-Reply-To: Message from Christian Tismer <ct@gras.de> 
   of "Fri, 26 Apr 2002 09:11:01 +0200."
Message-ID: <200204271232.g3RCWtkG017950@ratthing-b246.strakt.com>

For those of you who can't live without a Limbo introduction written
by a real C inventor:

http://www.vitanuova.com/inferno/papers/limbo.html

(by Ritchie). 

Laura Creighton




From andersjm at dancontrol.dk  Wed Apr 10 05:36:51 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Wed, 10 Apr 2002 11:36:51 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <Xns91EC16845A7E9RASXnewsDFE1@130.133.1.4>
Message-ID: <3cb40842$0$78789$edfadb0f@dspool01.news.tele.dk>

"Philip Swartzleonard" <starx at pacbell.net> wrote:
> Hm, interesting.
>
> Right at this molment (and i'm a little tired), i'm thinking that None
> is only nothing by itself. Think of this, if we say:
>
> x = None
>
> x still exists. It isn't nothing.

Of course; strictly speaking None is something, it's an object in its own
right.  It is only by convention that we use it to indicate the absense of a
more meaningful value.

- Anders




From nas at python.ca  Tue Apr 23 23:44:55 2002
From: nas at python.ca (Neil Schemenauer)
Date: Tue, 23 Apr 2002 20:44:55 -0700
Subject: Python's Lisp heritage
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEJAPAAA.tim.one@comcast.net>; from tim.one@comcast.net on Tue, Apr 23, 2002 at 10:03:11PM -0400
References: <3CC500AF.8BAE6F7E@cascade-sys.com> <LNBBLJKPBEHFEDALKOLCEEJAPAAA.tim.one@comcast.net>
Message-ID: <20020423204455.B28240@glacier.arctrix.com>

Tim Peters wrote:
> Dynamic scoping is attractive for implementation reasons; everyone who
> succumbs eventually regrets it (except for RMS in elisp ...).

There's still time. :-)  Erik Naggum has suggested that Emacs should be
re-implemented in Common Lisp.  That would be cool but I probably still
wouldn't use it.

  vi-forever-ly-y'rs Neil




From barry at zope.com  Thu Apr  4 14:08:29 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Thu, 4 Apr 2002 14:08:29 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285
References: <200204032232.g33MWqJ28933@odiug.zope.com>
	<3CAB8C99.232A8571@metaslash.com>
	<m3bsczfgic.fsf@mira.informatik.hu-berlin.de>
Message-ID: <15532.42157.607160.633720@anthem.wooz.org>

>>>>> "MvL" == Martin v Loewis <martin at v.loewis.de> writes:

    MvL> While Gustavo suggests to let pass 'a is True', I recommend
    MvL> to give a warning for this - people who want to see whether
    MvL> something is True should write 'True is a'.

"True is a"... what?  hippopotamus? <wink>.

-Barry




From gcordova at hebmex.com  Tue Apr 30 10:50:32 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Tue, 30 Apr 2002 09:50:32 -0500
Subject: Tix Vs Pmw which is better to use?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233AA50C5@mail.hebmex.com>

Hmmm...

Tix and Pmw attack different domains: Tix is native Tcl/Tk,
and Pmw is a Python attempt to create something akin to Tix
but in Python. The problem is that Python doesn't have access
to Tk internals, like Tix would.

So, if all the widgets you need are available from Tix, use
Tix. You could also write Pmw wrappers arround the widgets
you use (be them Tix or whatever); and then you could use
Pmw "exclusively".

I write "Exclusively" quoted, because although your app would
only be based on Pmw, internally it's also using Tix, and whatever
other widgets you need.

Personally, I really don't like Pmw all that much, because
Tk already has it's own megawidget protocol and packages,
so you're using a double layer over your application's GUI.

I simply write wrapper classes around the extension widgets
that I use, which work just like normal, ordinarly Tkinter
widgets (Tktable, Tkhtml, and looking for more :-)

Good luck!

-gustavo




From tdi01fhu at syd.kth.se  Tue Apr 16 08:21:28 2002
From: tdi01fhu at syd.kth.se (FREDRIK HULDTGREN)
Date: Tue, 16 Apr 2002 14:21:28 +0200
Subject: Mp3
Message-ID: <scbc338b.035@Riksvasa.syd.kth.se>

I would like to extract the bitrate from an mp3 file using python. Anyone know how?
:)

/Fredrik




From lac at strakt.com  Tue Apr 30 19:00:46 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 01 May 2002 01:00:46 +0200
Subject: Newbie: finding the key/index of the min/max element 
In-Reply-To: Message from "James J. Besemer" <jb@cascade-sys.com> 
   of "Tue, 30 Apr 2002 14:36:58 PDT."
Message-ID: <200204302300.g3UN0kkG002462@ratthing-b246.strakt.com>

> 
> Benjamin Han wrote:
> 
> > In my quest of becoming a "Pythonistas" (where does this word come from BTW
> ?),
> 
> The term derrives from the Sandinistas in Nicaragua, the Communist rebels who
> fought against the US-installed conservative Somoza dictatorship (and others)
> .
> The suffix is sometimes used to refere to any contrarian group, and in cases
> simply any group.

That is simple ignorant American narrow-minded foolishness talking.
Normally, I would take extreme great care to not say anything so 
offensive, but I have a deep suspcion that James J Besemer is one
of the 'can't take the heat' folks, and will learn to like a kinder
and gentler place than the one he proposes to make for Raymond Hettinger
if he finds out what it is like to live in the place he proposed.

At any rate, idiot fool deluded with the idea that the American perspective
is somehow comprehensive,

'ista' by those of us who speak Spanish means one hell of a lot
broader than that.  And not the same broadness in every Spanish
speaking nation.  My Spanish comes from where I lived in childhood,
San Pedro Sula, Honduras, and thus is fairly close to that of
Nicauragua (the country next door).  <except I lived in Cuenca Equador
also, complicated stuff my childhood>.  What 'ista' means varies from
sub-region to sub-region, and I get into confusion using it as I am
used to when I speak with Spaniards, where the rules are different --
but -- at any rate, in Honduras, it is used for everything from
'following the teaching of' to 'liking'.  You get 'istas' for every
damn fad on the planet, every popular song, let alone songwriter, and
just preferring the colour blue.

All of this has been going on for vastly longer than Super-Power
sponsored hell in small countries.

The Super-Power sponsored hell was such a vileness, that nobody would
be particularily interested in invoking that image.  Pythonista may
resonate to ignorant Americans with only that sound, but the rest of
us have a richer cultural experience in which to ground such an
expression.  At least by God I hope so -- else we have finally
reached the Nadir of Disneyism, where violent repression followed with
anarchy, and the dying hopes and dreams of a generation are _glorified_
because they are BRANDED.

Yours most disgustedly,
Laura Creighton




From loewis at informatik.hu-berlin.de  Wed Apr 24 04:23:28 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 24 Apr 2002 10:23:28 +0200
Subject: StringIO vs array.tofile() ?
References: <mailman.1019633945.8958.python-list@python.org>
Message-ID: <j4y9fdld33.fsf@informatik.hu-berlin.de>

jaf <jaf at lcsaudio.com> writes:

> Hi all,
> 
> Why is that?  Isn't a StringIO a valid "open file"?

The array's tofile method requires a file object (i.e. an object whose
type is types.FileType); a StringIO object is only "file-like".

Since this limitation is documented (it says file, not file-like),
this is a bug in your code. I recommend to use the .tostring method
instead.

Regards,
Martin



From martin at v.loewis.de  Mon Apr 22 17:49:15 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 22 Apr 2002 23:49:15 +0200
Subject: using automake/autoconf to build extension module?
References: <mailman.1019509602.30761.python-list@python.org>
Message-ID: <m34ri34d5w.fsf@mira.informatik.hu-berlin.de>

Skip Montanaro <skip at pobox.com> writes:

> Can someone explain what automake incantations are necessary to build a
> Python extension module?  

I personally recommend to stay away from libtool, if you can. It is
terribly broken.

> I'd prefer not to use distutils if possible, since
> the rest of the project is using automake/autoconf/libtool for build
> configuration.

That does not necessary contradict each other. You can easily embed
Python fragments into a configure.in, to determine the appropriate
command line options.

You need to extract PY_CFLAGS from Python's Makefile; one way is, e.g.

eval `python - <<END
import distutils.sysconfig
flags=distutils.sysconfig.get_config_var("PY_CFLAGS")
print 'PY_CFLAGS="%"' % flags
END`

Then, you can integrate PY_CFLAGS into your own build process.

Notice that this still won't get the right -I option; your computation
of the include directory will be still needed.

Regards,
Martin


From sholden at holdenweb.com  Mon Apr 22 12:28:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 12:28:39 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org> <304d20df.0204220633.50bf23bf@posting.google.com> <mailman.1019491242.2145.python-list@python.org>
Message-ID: <xVWw8.76795$T%5.8026@atlpnn01.usenetserver.com>

"holger krekel" <pyth at devel.trillke.net> wrote ...
[ ... ]
> If so, let me note that "execing a string" is what python does itself.
> e.g. when you import a module the interpreter is execing a module-string
or
> code-object. It's some weird hack but the regular python way.
>
Well, there's the minor point that the efficiency of the interpreter's
regular operation is hugely affected by the pre-compiled ".pyc" file that
will frequently be created. When you import a module, therefore, usually you
will be executing a code object.

Writing a module to "exec" one or other of multiple strings gets you pretty
much the worst of all worlds -- when the (compiled) module is loaded, it
will be executed as a code object, but then one selected version-specific
string must still be compiled by an "exec" before being executed each time
the module is run.

If some misguided user tries to take advantage of this "portability"
technique by using the same directory for multiple versions of Python the
situation gets a whole lot worse, as the ".pyc" files will be re-created
each time the Python interpreter is different from the one that imported the
module the last time!

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From aleax at aleax.it  Fri Apr  5 04:52:50 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 09:52:50 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAD4CD8.2090700@mxm.dk> <7Adr8.1447$%A.217236@news1.west.cox.net>
Message-ID: <Ster8.62071$S52.2165674@news2.tin.it>

Van Gale wrote:

> 
> "Max M" wrote:
> 
>> Oh stop whining!
> 
> Was that really necessary?

Indispensable!  Have you never taken Flamewars 101?  Starting with an
offensive tone is one of the best-known techniques in the business.
Deadpan is a more advanced subject, of course, although worthwhile.


>> Just do a search and replace in your codebase::
>>
>>      s/1/True/gi
>>
>> That should solve all those problems in a few seconds.
> 
> Yeah... it'll solve **those** problems, but think about it for a sec.
> Maybe, just maybe, this will create a slew of entirely different problems?

Actually, the VIM-only case-insensitivity flag is crucial here,
since it lets you substitute both uppercase 1 and lowercase 1
in one sweep (pity that, as written, it works for one line only,
but, if one's Python code IS just one line in all, that's OK).


Alex



From boris at cantal.lip6.fr  Mon Apr  8 12:45:23 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Mon, 08 Apr 2002 18:45:23 +0200
Subject: changing local dictionnary for functions
References: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr>
Message-ID: <pan.2002.04.08.18.45.23.620630.19669@cantal.lip6.fr>

To be more precise, I'm looking for this kind of function :
execfunction(func,*args,globals,locals)

func is a function
*args : is a list of arguments to pass to the function
globals is a dictionnary (or subclass of it) being used as global dict
locals is a dictionnary (or subclass of it) being used as local dict

I think this kind of function can be written by playing with the
func.func_code bytecode, but before trying to do it this way I'd like to
be sure there is no simple way to do it.

Boris


On Mon, 08 Apr 2002 18:15:31 +0200, Boris Boutillier wrote:

> Is there a way to pass a local dictionnary to a function as it is
> possible to do with a file with the execfile statement ?
> 
> In the python reference manual, in the section 4.1 : Code
> Blocks,execution frames and namespace, there is a table that say that
> function body uses globals namespace of the containing block and a new
> locals namespace. And there is a note that say global and local can be
> overrident, how is it possible ? I think it would be the answer to my
> first question.
> 
> Boris


From ASkwar at DigitalProjects.com  Fri Apr 26 08:34:50 2002
From: ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Fri, 26 Apr 2002 14:34:50 +0200
Subject: python and very large data sets???
In-Reply-To: <20020426003428.GD22555@devel.trillke>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6o7l$nbi$1@panix3.panix.com> <20020424172403.GE20209@devel.trillke> <20020425223536.GG21205@teich.Garten.DigitalProjects.com> <20020426003428.GD22555@devel.trillke>
Message-ID: <20020426123450.GM21205@teich.Garten.DigitalProjects.com>

?holger krekel? sagte am 2002-04-26 um 02:34:28 +0200 :
> it is quite easy to set up a database e.g. for a website but 
> is the same true if you have 100 GB of data being processed?
> Easy as pie? Maybe i am underestimating mysql...

Yes, I'd think so.  He will have 100 GB, that's true.  And on some
systems he will not be able to store more than (2 GB - 1 byte) in one
table.  But even that's only a problem when doing the inserts; thanks to
merge tables, it's easy to have one table which can select data from
multiple tables without joins.

But other than that, I don't think there'll be a problem.  He
doesn't need ultra high performance, as he'll most likely only have 1
thread accessing the database at any given time.

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 2 days 7 hours 31 minutes




From narnett at mccmedia.com  Wed Apr 17 14:37:03 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Wed, 17 Apr 2002 11:37:03 -0700
Subject: Importance of trademarks (was RE: [OT] What is Open Source?....)
In-Reply-To: <20020417135054.A88552@doublegemini.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJCEHPKCAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Clark C . Evans

...

> There is a difference between "trademark" and "copyright".  RMS firmly
> believes in trademarks; and he believes that he has a special right to
> manage development of projects (named by a trademark) he has initiated.

Trademark law was what allowed Sun to stop Microsoft from applying its
"embrace and extend (and destroy)" approach to Java.  I think that's an
excellent example of how trademarks can be used to prevent what I like to
call "unfair cooperation."

Anti-trust law never really contemplated the importance of open standards as
we know them today, so it protects against actions that tend to diminish
fair competition.  So, anyone who wants to mess with a standard could defend
their actions by arguing that they are enhancing competition.  But in this
day and age, behavior that diminishes fair cooperation -- as Microsoft was
trying to do to Java -- is at least as important to stifle as the
traditional notions of unfair competition, in my opinion.

I'd love to see some public discussion of what kind of laws might strengthen
protection and opening up of standards as they emerge.  Seems to me that
when something like AIM becomes widely used, its protocols should become
open and standardized.  That's what patent law does -- assures that
inventions become open (that's what 'patent' means, after all) eventually.
The main reason patents don't work for software is that the length of
protection is too long in most cases, so that by the time software patents
expire, the technology is often worthless.

Meanwhile, we can all hope (and act so) that the PSF and affiliates will use
the trademark to prevent anyone from trying to create a proprietary version
with the same or similar name.

Nick





From achim.domma at syynx.de  Mon Apr 29 10:28:18 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Mon, 29 Apr 2002 16:28:18 +0200
Subject: os.path like library for urls ?
Message-ID: <aajkrn$o0a$07$1@news.t-online.com>

Hi,

I read a html page via urllib2 and want to make the links in the page
absolute. Is there a os.path like library for manipulation urls ?

greetings
Achim




From ffjhenon at club-internet.fr  Sat Apr  6 18:54:41 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Sun, 07 Apr 2002 01:54:41 +0200
Subject: stdout redirect : Howto with popen2.popen3
Message-ID: <3CAF8AC1.90704077@club-internet.fr>

I want to use Tkinter to launch a raytracer that can last from 1 second
to several days.
That raytracer displays an image in an external box and the line which
is being raytraced to the console.
Once the raytracing process is over, it outputs stats or an error
message if something goes
is wrong in the script.

For the time being, all I managed to get is the image box and a file
with the stats.

  1? I would like the stats to be returned straight to a Tkinter window.

I use popen as shown below.

  2? How can I tell when the raytrace is over ? The poll() gets out of
the loop.
What is command to listen to a program to tell if it's running or not?



###     pin,pout = os.popen2(cm )       ## these two lines
###     pin,pout,perr = os.popen3(cm )  ## do not work



cm="command 2> sortie.txt"
run=popen2.Popen3(cm)
pi=run.pid
pid=pi+1
while run.poll() <>-1:
     print "attente"   # // I never got to here and goes straight to
fd=open.....


fd=open("sortie.txt")  # // reads the file created by the pipe
for line in fd.readlines():
     if line.endswith("\r"):
         line=line[:-2]+"\n"
         self.text_out.insert(END,line)  # // inserts the file in the
Tkinter box
                           // from the previous raytrace .......




Thanks for any help




From loewis at informatik.hu-berlin.de  Mon Apr 29 03:04:49 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 29 Apr 2002 09:04:49 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net>
Message-ID: <j4elgz9e9a.fsf@informatik.hu-berlin.de>

"Raymond Hettinger" <python at rcn.com> writes:

>     The module additions would go into Python 2.3.  All six
> functions would be left intact and raise deprecation warnings in
> Python 2.3 and 2.4.  The six functions would be removed in Python
> 2.5 (having allowed two versions and a full year for code updates).

I believe the phasing out of the old names should be even slower. In
2.3, introduce the new names; leave the old names intact unchanged. In
2.4 and 2.5, raise DeprecationWarnings; in 2.6, remove them.

Regards,
Martin



From matt at greenroom.com.au  Tue Apr  9 03:39:19 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Tue, 09 Apr 2002 17:39:19 +1000
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net>
Message-ID: <3CB29AA7.5020409@greenroom.com.au>

akhar wrote:

> Has anyone ported Python to embedded devices? I heard of the deeply embedded
> python porject but the link is dead. What would it take to port python to an
> embedded platform (whether be it the mc68hc11 or PICs)?
> 
> Regards
> 
> Stephane
> 


What platforms though?

A version of python for the pocket pc eg. win ce can be found
here: http://www.murkworks.com/Research/Python/PocketPCPython/Overview

Python for palmos:
   http://www.isr.uci.edu/projects/sensos/python/

PS: I haven't found either of them to be very reliable though
as they are pre-release. If you want a real embedded solution, look at java



From geoff at gerrietts.net  Mon Apr 15 19:23:28 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 15 Apr 2002 16:23:28 -0700
Subject: String interpolation question
In-Reply-To: <a9fhad$qbt$1@peabody.colorado.edu>
References: <a9fcmo$npi$1@peabody.colorado.edu> <mailman.1018904544.23487.python-list@python.org> <a9fhad$qbt$1@peabody.colorado.edu>
Message-ID: <20020415232328.GC12476@isis.gerrietts.net>

Quoting Fernando P?rez (fperez528 at yahoo.com):
> Geoff Gerrietts wrote:
> 
> print "$i arr1[$i]=$arr1[$i] arr2[$i]=$arr2[$i] arr3[$i]=$arr3[$i] ....\n";
> 
> To me, the above line is clear, unambiguous and requires zero contortions to 
> either write or read (by the coder). I have been so far unable to write in 
> python anything that comes anywhere within even shooting distance (with an 
> ICBM) of this (but I'd love to be proven wrong ;)

I see where you're coming from here, and I don't argue too strenuously
with it -- in this case the perl syntax doesn't bite back, which is a
pleasant place to find yourself. ;)

I would solve this in one of several ways, depending on which way
struck me at the time of writing. I think the most likely solution
would be something like:

>>> pieces = ["%d" %(i,)]
>>> for x in range(1,maxarray+1):
>>>   arr = "arr%d" % x
>>>   val = locals()[arr][i]
>>>   pieces.append("%(arr)s[%(i)d]=%(val)s}"%locals())
>>> print " ".join(pieces)

Because of who I am, and how I code, I think it's more likely that it
would turn up like:

>>> pieces = ["%d" %(i,)]
>>> for x in range(1,maxarray+1):
>>>   arr = "arr%d" % x
>>>   val = locals()[arr][i]
>>>   bit = "%s[%d]=%s" % (arr,i,val)
>>>   pieces.append(bit)
>>> print " ".join(pieces)

And if I stepped even farther back, it's likely that I'd have
descriptive names for my lists, and consequently it'd look more like:

>>> pieces = ["%d" % (i,)]
>>> arrays = ["names", "dates", "faces"]
>>> for arr in arrays:
>>>   arr = "arr%d" % x
>>>   val = locals()[arr][i]
>>>   bit = "%s[%d]=%s" % (arr,i,val)
>>>   pieces.append(bit)
>>> print " ".join(pieces)

I usually prefer to have a collection of objects instead of a bunch of
synchronized lists, though, so it's even more likely that I'd be
writing a __str__ method on a class.

Again, best of luck -- sounds like the ASPN recipe has made you happy,
and I concede that it's cool -- Eval'ing anything always makes me
nervous, though :)

--G.

-- 
Geoff Gerrietts             "People talk fundamentals and superlatives 
<geoff at gerrietts net>     and then make some changes of detail." 
http://www.gerrietts.net                  --Oliver Wendell Holmes Jr




From skip at pobox.com  Tue Apr 23 12:07:33 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 23 Apr 2002 11:07:33 -0500
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
In-Reply-To: <aa3dti$n0$1@news.nextra.cz>
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <15557.34501.747759.818289@12-248-41-177.client.attbi.com>

    Petr> Being quite new to Python -- is there any preference of single or
    Petr> double quotes when writing strings?

Not that I'm aware of.  Coming from a C background, I personally tend to use
"..." for multi-character strings and '.' for single-character strings (in
the absence of any other constraints), but I don't adhere to it closely.  In
particular, if I'm modifying someone else's code I try to mimic their usage.
When the strings contain apostrophes I try to use quotation marks and vice
versa.  If strings contain both, I switch to triple-quoted strings.  The
most hard-and-fast rule I adhere to is to try and avoid escaping embedded
apostrophes or quotation marks.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From donn at u.washington.edu  Mon Apr 15 16:05:09 2002
From: donn at u.washington.edu (Donn Cave)
Date: 15 Apr 2002 20:05:09 GMT
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3cb8a237@news.mhogaming.com> <MPG.1724b5c6afb62a419896b0@news.nwlink.com>
Message-ID: <a9fbpl$56u$1@nntp6.u.washington.edu>

Quoth Jeff Shannon <jeff at ccvcorp.com>:
| In article <3cb8a237 at news.mhogaming.com>, 
| Ken.Peek at SpiritSongDesigns.comNOSPAM says...
|
| > ...  If you do not intend to operate the
| > spacecraft in 'interactive mode', then why use a scripting
| > language at all?
|
| For the same reasons that one uses a "scripting" language in any 
| other project -- speed of development, plus clearer, easier-to-
| read code (thanks to less boilerplate).  Which leads directly to 
| easily maintainable code and simpler software upgrades.
|
| Note that the term "scripting language" is a very imprecise term, 
| and it's arguable that it doesn't strictly apply to Python.  
| Python is an interpreted general-purpose language, which is 
| useful for scripting (among other things).  Then again, there are 
| some people who use C as a scripting language...

Sure - especially if you take it in the most useful sense,
application scripting, in which case the nature of the language
is more or less irrelevant once you have access to the host
platform's scripting API.

If I were promoted to spacecraft programmer, Python's attraction
for me would be mainly its storage (reference counted "garbage
collection") and exceptions.  The storage policy (which by the
way would favor C Python over Java Python, since the latter's
implementation is as I understand it less predictable) solves a
lot of problems that make library interfaces gnarly (and subject
to bugs like buffer over-runs.)  The exceptions also simplify
library interfaces, though here I would be more worried about
surprises - it's an attraction, but it does carry its own problems.

None of these requires an interpreter, which I think might have
been the point of the question.  I can't imagine anyone would
have a reason to specifically want an interpreter for an application
like this, given the choice, but we usually aren't given the choice.

	Donn Cave, donn at u.washington.edu


From lac at strakt.com  Mon Apr 29 17:58:56 2002
From: lac at strakt.com (Laura Creighton)
Date: Mon, 29 Apr 2002 23:58:56 +0200
Subject: Third Draft: Pep for Deprecating Builtins 
In-Reply-To: Message from Alex Martelli <aleax@aleax.it> 
   of "Mon, 29 Apr 2002 21:19:40 GMT." <MNiz8.32691$8D3.925320@news1.tin.it> 
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch>  <MNiz8.32691$8D3.925320@news1.tin.it> 
Message-ID: <200204292158.g3TLwukG029504@ratthing-b246.strakt.com>

> Boris Borcic wrote:
>         ...
> > Octal has some virtues, e.g. 365 == 0555 making it easy
> > to remember the number of days in a usual (non-leap) year :)
> 
> I'm surprised you miss the main advantage -- 0666, the Number
> of the Beast in octal, translates to decimal 438, the A.D. in which
> Emperor Theodosius II promulgated the Theodosian Code, the
> wildly antisemitic laws which, among other things, forbid Christians
> to even give alms to Jews, forbid Jews to hold State office or build
> synagogues, condemned to death any Jew converting a Christian.
> 
> Surely this is therefore a number which is often handy in the everyday
> practice of programming, and even thinking of eventually making a
> non-builtin of the precious Python function that helps us work with a
> numbering system which has such wondrous properties is unholy.
> 
> 
> Alex

Alex, Alex, surely you know that the main advantage of oct is that 
dec 25 == oct 31  which you can use to prove that all Christians are
Satanists.

Laura 




From aahz at pythoncraft.com  Wed Apr 24 18:26:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Apr 2002 18:26:09 -0400
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <3CC6ED95.569DA16A@metaslash.com> <mailman.1019682346.19715.python-list@python.org>
Message-ID: <aa7be1$i3v$1@panix1.panix.com>

In article <mailman.1019682346.19715.python-list at python.org>,
holger krekel  <pyth at devel.trillke.net> wrote:
>
>I just don't happen to see the advantages of bringing a database into
>the picture. It seems like a classical batch job and it 'random access
>many times' is not needed, so why?

>From the original post:

    Things would afterwards get more complicated cause I will have to
    pullout ID's from "sub_file1", remove duplicate ID's create
    "no_dup_sub_file1", match those to ID's in remaining 3 main files and
    pullout data linked with those ID's.

This screams "*JOIN*" to me.  Now, if sub_file1 is less than 100MB,
*maybe* Python can handle it.  IMO, that is true IIF the records are
strictly fixed-length.  But IME requirements will change such that joins
over larger and larger datasets will be needed -- and why re-invent a
database that's designed precisely for this purpose?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From giannivolontosky at TOGLIyahoo.it  Thu Apr 25 15:39:32 2002
From: giannivolontosky at TOGLIyahoo.it (LoGan)
Date: Thu, 25 Apr 2002 19:39:32 GMT
Subject: ftplib and attributes
Message-ID: <pan.2002.04.25.21.35.03.529909.14114@TOGLIyahoo.it>

hi, im new bie and im trying to use ftplib, but my problem is to get the
attributes of the files, how can i do ?
LoGan


From sholden at holdenweb.com  Fri Apr  5 16:09:41 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 5 Apr 2002 16:09:41 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org>
Message-ID: <Npor8.8281$q63.818@atlpnn01.usenetserver.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1017865089.20676.python-list at python.org...
> On Tue, 2 Apr 2002 21:21:27 -0500
> John Roth wrote:
>
> >
> > "Laura Creighton" <lac at strakt.com> wrote in message
> > news:mailman.1017755045.4308.python-list at python.org...
> > > Dear Guido:
> > >
> > > I would first like to thank you for giving us an opportunity to
> > respond.
> > > I have spent most of the weekend thinking and writing a reply to this,
> > > and I think that this has made me a better teacher.  For this I am
> > > grateful.  I realise that this is rather long, but I have condensed
> > > it substantially from my efforts earlier this weekend.   Thank you
> > > for taking the time to read this.
> > >
> > > Laura Creighton
> > >
> > > -----------------------------
> > >
> > I'm not going to quote this, as it's already too long. I read up to
> > the diatribe on the Swedish leap year, and I went AAAARRRRGGGH!
> >
> > I really don't care if you want the silly function to return
> > a Boolean, the number of leap days in the year, or the name
> > of the idiot parliamentarian who decided to fix the calendar
> > so that 1712 was different from the rest of Europe.
> >
> > HOWEVER, if you want true or false as the return
> > value, the name should be something like 'isLeapYear()'.
> > If you want the number of leap days in the year, then
> > it should be named something like 'numberOfLeapDays().'
>
Just so John know he isn't the only person going AARRGGHH, I read this went
AARRGGHH myself. It should not be named "something like
'numberOfLeapDays()'". It should be named LeapDays(), or possibly
LeapDayCount() if you also have a function that returns the actual dates
which ARE the leap days. Or, I suppose, you could call it
"theNumberOfLeapDaysInTheSwedishYear()" ;-)

> > Sorry to make this sound like a flame, but I try to
> > practice XP. The first time I ever come across a
> > function named "isSomething" that returns a conceptual
> > integer, it's going to get refactored, right then and
> > there to a name that actually expresses what it does.
>
People who live in glass houses ...

> Laura was using the example given by Guido, since she was responding to
> Guido's statements.
> I totally agree that a function name that implies a boolean should return
a
> boolean.  Guido was stating otherwise, that a boolean could also be used
as
> an int, which Laura felt (and apparently you and I agree) is wrong.  If
you
> had read the entire post you would have seen that that was (partially) her
> point:
>
> [GvR]
> >     4) Should we strive to eliminate non-Boolean operations on bools
> >        in the future, through suitable warnings, so that e.g. True+1
> >        would eventually (e.g. in Python 3000 be illegal).  Personally,
> >        I think we shouldn't; 28+isleap(y) seems totally reasonable to
> >        me.
>
> [Laura]
> > This is not an argument for allowing non-Boolean operations on bools();
> > this is an argument for not writing functions that return Booleans.
Make
> > them return numbers instead, so that you can use them as you did.  Last
> > month we discussed why in 1712 February had 30 days in Sweden.  (See:
>
> So what you are saying fits perfectly with what Laura asserted.  If you
> have a function numberOfLeapDays() why on earth would you need
> isLeapYear()?  The solution is simple: don't return a boolean when another
> type can not only express true/false but also convey other information as
> well.
>
It seems plain (now I've read so much of the discussion) that if book isn't
going to behave as a full-blown Boolean object it would be better to leave
well enough alone.

> I didn't necessarily agree with all of Laura's assertions regarding the
> difficulty of understanding how booleans would fit into Python, but she
> raised several compelling and thoughtful arguments.  I totally agree that
> bools as return values of functions are less than useful.  Booleans have
> the lowest information density of any type and thus would tend to increase
> code verbosity if their use were to be encouraged.
>
I too thought Laura made a thoughtful contribution to the discussion, that
deserved to be considered more respectfully.

> Frankly Guido's dismissal of Laura's arguments as FUD was not only a bit
> rude, but IMHO a bit short-sighted as well (given the fact that someday,
> there may be a desire for a real boolean type).  Maybe Tim Peters has
> borrowed his time-machine and not returned it ;)
>
However it happened, it left me feeling a little embarrassed that I had
originally defended Guido's inegrity and suggested that despite his
authorship the PEP's acceptance was not a foregone conclusion. I hope I
wasn't wrong, but that's not how it looks right now. Oh well.

regards
 Steve






From wware at alum.mit.edu  Wed Apr 24 23:06:24 2002
From: wware at alum.mit.edu (Will Ware)
Date: Wed, 24 Apr 2002 23:06:24 -0400
Subject: wiki helper cgi script
Message-ID: <3CC772B0.684CAFD4@alum.mit.edu>

I have a few wikis on my site, all using UseMod. The following Python script
can be shared by all my wikis, and it lets me put a link on the wiki that
generates an alphabetized list of all the wiki's pages with links to them.
It doesn't care what directory or subdirectory the wiki in is. I just put the
following in the wiki:

   [http://willware.net:8080/cgi-bin/pagelist.py List] of all pages

I'm running this on a Linux box with Apache. YMMV otherwise.

Bon appetit, Will Ware

#######################################
#!/usr/bin/python
import os, sys, string
 
def ouch(arg):
    print "Content-type: text/html\n"
    print "<HTML><TITLE>Ouch</TITLE>"
    print "<BODY BGCOLOR=\"WHITE\">"
    print "<H1>Ouch: %s</H1>" % repr(arg)
    print "Don't try to call this CGI script directly."
    print "</BODY></HTML>"
    sys.exit(0)
 
if os.environ.has_key('HTTP_REFERER'):
    referer = os.environ['HTTP_REFERER']
else:
    ouch(1)
 
referer = referer[7:]    # strip off "http://"
referer = string.split(referer, "/")
 
if referer[0] != 'willware.net:8080':
    ouch(2)
 
if referer[1] != 'cgi-bin':
    ouch(3)
 
referer = string.join(referer[2:-1], "/")
 
inf = os.popen("find %s/data/page -type f" % referer)
L = inf.readlines()
inf.close()
L.sort()
 
print """Content-type: text/html
 
<html>
<title>List of all pages for WifiWiki</title>
<body bgcolor=white>
<h1>List of all pages for WifiWiki</h1>
<ul>
"""
 
for x in L:
    x = x[13+len(referer):]
    x = x[:-4]
    url = "http://willware.net:8080/cgi-bin/%s/wiki.cgi?%s" % (referer, x)
    item = "<li><a href=\"%s\">%s</a>" % (url, x)
    print item
 
print """</ul>
</body>
</html>
"""
#######################################


From sholden at holdenweb.com  Fri Apr  5 16:17:05 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 5 Apr 2002 16:17:05 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net>              <3CAC149F.3F1BA77B@lemburg.com> <mailman.1017930740.20674.python-list@python.org> <slrnaapbh8.4s7.huaiyu@gauss.almadan.ibm.com>
Message-ID: <Iwor8.8425$q63.5655@atlpnn01.usenetserver.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnaapbh8.4s7.huaiyu at gauss.almadan.ibm.com...
[...]
>
> BTW, many people mistake boolean algebra with two-element boolean algebra,
> which is only its most trivial example.  The set operations (union, joint,
> complement) on all subsets of a given set is a more generic example of
> boolean algebra.
>
Don't I remember from my degree studies that a Boolean algebra can be
defined on any lattice?

regards
  Steve






From borcis at geneva-link.ch  Thu Apr 11 05:16:29 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 11 Apr 2002 11:16:29 +0200
Subject: [OT] Re: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>  <3CB2DB99.4080400@geneva-link.ch> <slrnab9soc.6g8.mlh@vier.idi.ntnu.no>
Message-ID: <3CB5546D.8010802@geneva-link.ch>

Magnus Lie Hetland wrote:

>>
>>   >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
>>   'BUSH'
>>
>>What's your problem ?
> 
> Nothing. Sorry -- thouhgt there was a syntax problem there. A bit
> quick on the trigger ;)


Revealing :)


> 
> (So, what you're basically saying is that Bush is the literate part of
> bullshit? <wink>)


More concretely, providing support to the idea that the "USA" in 
"CRUSADE" explains the general pattern of some, er, crude presidential 
choice of words.

Blind-spots-are-consubstantial-to-visions-ly yours,

Boris Borcic
--
"Hope achieves the square root of the impossible"




From dthierbach at gmx.de  Sat Apr 20 08:08:31 2002
From: dthierbach at gmx.de (Dirk Thierbach)
Date: 20 Apr 2002 12:08:31 GMT
Subject: Host a Gtk widget inside a Tk frame
References: <63604d2.0204171324.3b8c4d6f@posting.google.com>
Message-ID: <a9rlnv$i8p$1@sun27.hrz.tu-darmstadt.de>

In comp.os.linux.x Phlip <phlip_cpp at yahoo.com> wrote:
> Xists:

> I have this friend who has written a large GUI using only Tkinter.

> Then he discovered a must-have widget. But unfortunately this widget
> binds to Gtk. (The Widget in question is VPython, but that's not
> important right now.)

> My friend thinks the best way would be to extract the X windows ID
> from a Tk Frame, wrap this in some way that Gtk understands, and
> present it to the Gtk control as its parent handle. 

It is possible to do that, either when creating the widget in Gtk,
or even after creation using XReparentWindow. Howover, more natural
and probably also

> But my friend's customers expect a seamless interface.

more to your friend's customers liking, would be to extend Tcl/Tk with
some glue code that links to the widget he wants. Tcl/Tk was *made*
for incorporating C extensions (a fact most people forget).

There is a lot of documentation how to do that; a quick search turns
up for example the "Engineering Guide for the Tcl core implementation
and C-level extensions." by John Ousterhout himself, on 

http://tcl.activestate.com/doc/

- Dirk

[F'up to comp.os.linux.x]


From hfoffani at yahoo.com  Fri Apr 12 12:35:16 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Fri, 12 Apr 2002 18:35:16 +0200
Subject: How to test tp_print?
References: <20020412112023.201$FO@news.newsreader.com> <j4662xszlr.fsf@informatik.hu-berlin.de>
Message-ID: <20020412123523.718$pw@news.newsreader.com>

> > Any other approach to this?
>
> Open the file in binary mode.

Doh!
What I was thinking of?
Thanks and sorry.
-Hernan




From erno-news at erno.iki.fi  Tue Apr  9 16:40:54 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 09 Apr 2002 23:40:54 +0300
Subject: binding a reference to a variable
References: <yu994rikfzh5.fsf@europa.research.att.com>
Message-ID: <kur8lomwq1.fsf@lasipalatsi.fi>

In article <yu994rikfzh5.fsf at europa.research.att.com>, Andrew Koenig
<ark at research.att.com> writes:

| I would like to be able to write an expression that yields an
| object that I can subsequently use to rebind a name that I mention
| only in that expression.  How do I do it?

[...]

| So far, the closest I've been able to come is this:

|         def set(var, val):
|                 var[:] = [val]

[...]
                
| Is it possible to do better (i.e., to solve this problem more succinctly)?

no - python likes to keep variables private to a scope.

  -- erno


From robin at jessikat.fsnet.co.uk  Tue Apr 16 13:53:46 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 16 Apr 2002 18:53:46 +0100
Subject: Python 2.2 unwise.exe
References: <mailman.1018727067.4980.python-list@python.org>
 <upu104ym0.fsf@ctwd0143.fitlinxx.com>
Message-ID: <GSYATKAqUGv8EwRX@jessikat.demon.co.uk>

On the office machine well I installed Python 2.2.1 last Friday evening.
My unwise.exe is now gone as of Tuesday 10:00 am.

I had a hard crash Monday morning.
-- 
Robin Becker


From bokr at oz.net  Mon Apr  1 17:03:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 22:03:58 GMT
Subject: Files, Seek, and Readlines. A newbie question
References: <Ts%p8.149254$ZR2.64157@rwcrnsc52.ops.asp.att.net>
Message-ID: <a8alge$8bk$1@216.39.172.122>

On Mon, 01 Apr 2002 15:58:43 GMT, "John Purser" <NO_SPAM_jmpurser2 at attbi.com> wrote:

>Hello,
>
>I'm using a large (>120,000 lines) sorted ASCII text file as an index.  I
>started running into problems as soon as I began to use the .seek() to
>position before a read.  Even when not opened as a binary file seek moves by
>bytes not lines.  Is there a line oriented seek version?  Is there a better
>way to get to line 50 than:
>myFile.seek(0)
>myFile.readlines(49)
>Record = myFile.readline()
>?
>
I recently posted a class definition that would create an index
of line positions so that later you could access the lines randomly.

See post titled "Re: seek in a file"
Email me if you don't find it. But I'll be gone a couple of days.

Regards,
Bengt Richter


From sholden at holdenweb.com  Tue Apr 23 10:16:41 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 10:16:41 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net>
Message-ID: <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnacar0d.kd0.jon+usenet at snowy.squish.net...
> In article <uloacus4pml1r7f7eac9ts3u20oroj34qb at 4ax.com>, Fernando
Rodr?guez wrote:
> >>curs.execute(
> >>    """insert into customer_info (email, url, referrer, first_visit)
> >>    values ( %s, %s, %s, %s)""",
> >>    ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com',
'2001-03-07
> >>12:32:00')
> >>)
> >
> > mysqldb.error: (1064, "You have an error in your SQL syntax near
> > '@microsoft.com, www.easyjob.net, www.google.com, 2001-03-07 12:32:00)'
at
> > line 1")
>
> Woah, you have *seriously* screwed something up there. The code shown
> is correct. I suggest you post the earlier part of your code where
> your are connecting to the database and creating 'curs'.
>
> Are you sure you really are doing dbc.execute(foo, bar) and not
> dbc.execute(foo % bar)?

This error was caused by Fernando's following a bogus suggestion of mine.
His original SQL syntax was correct, and we are currently tracking down what
looks like an install problem.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From noah at noah.org  Wed Apr 24 20:02:26 2002
From: noah at noah.org (Noah)
Date: Wed, 24 Apr 2002 17:02:26 -0700
Subject: os.forkpty vs. pty module
Message-ID: <PAENINBOOODFDEPOKPKCCEHPCMAA.noah@noah.org>

The documentation for os.forkpty() says
to use pty module instead because it's more portable;
yet the pty module appears to be fairly immature
and in the words of the documentation:
	"Because pseudo-terminal handling is 
	highly platform dependant, there is code 
	to do it only for SGI and Linux."

os.forkpty() seems to work on OpenBSD.
Is there a way to evaluate which to use?
Should I actually avoid os.forkpty?
Could this be deprecated in the future?

Yours,
Noah




From jadestar at idiom.com  Mon Apr 29 23:12:57 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 30 Apr 2002 03:12:57 GMT
Subject: Asynchronous Socket with the ServerSocket module (??)
References: <8Xgz8.730$5y4.17760@castor.casema.net>
Message-ID: <aal23p$1u2g$1@news.idiom.com>

guyon moree <cheezus at wanadoo.nl> wrote:

> hi all,

> i'm trying to write some sort of server program on linux. i stumbled into
> the ServerSocket module and it does pretty much what i want. i create a
> requesthandler class and pass it to the module and call server_forever() By
> using ForkingTCPServer i am able to make it handle multiple connecties.

> all fine...

> but...

> how can i keep a list of connections and pass for example messages from one
> client to the other through the server?

> thanx in advance, i am a newbie though.

> regards,
> Guyon

	Sorry I don't have time to dig up the modules and write a
	working sample, but I suspect that you'd do something like:

	* create your own class (to hold the list of active connections)
	and a couple of methods to add and remove to/from it (accessors)
	(alternatively just make a global list or dictionary --- possibly
	of tuples --- or of socket descriptors, at least)

	* over-ride the accept/handle_connection methods in your socket
	server to add new connections to your list/dictionary/class instance

	* over-ride the close/teardown method in your socket server to
	remove closed connections.

	* finally modify the method (which you surely most have already 
	over-ridden) to implement your connection lookup and message
	passing in the receiver handler of your connects.  (In other words
	when you receive data from one of the connections you look up
	where it's going in your dictionary/list/instance and then 
	try to perform the relay.  

	Naturally there's lots of fussy bits that you have to handle
	in there.  First, you have to find the actual method (names and
	prototypes) that have to be over-ridden.  Also you have to figure
	out how you want to handle exceptions and what you blocking 
	semantics should be (you'd like to avoid deadlock so you probably
	need to use non-blocking reads or non-blocking writes, pick one
	and be consistent).  Most importantly you have to figure out what
	sorts of exception handling are appropriate to your needs.

	I'd consider whipping up something (it would be a good learning
	exercise for me) but my wife will be very unhappy if I don't get
	her out of the house and off to dinner soon!

domestic-bliss-over-rules-hacking-ly yrs?




From aleax at aleax.it  Thu Apr 18 18:58:11 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 22:58:11 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <I6Fv8.36158$vF6.1093396@news2.tin.it> <mailman.1019162916.19664.python-list@python.org>
Message-ID: <7cIv8.36928$vF6.1122050@news2.tin.it>

Lulu of the Lotus-Eaters wrote:

> Alex Martelli <aleax at aleax.it> wrote previously:
> |class FunkyReprMeta(type):
> |    def __repr__(cls):
> |        return "My wonderful class %s" % cls.__name__
> |class FunkyReprObject:
> |    __metaclass__ = FunkyReprMeta
> |class C(FunkyReprObject):
> |    "whatever you want here"
> 
> I don't really understand what a metaclass buys you here.  I think I
> might be missing something important, since Alex proposes it.

Thanks for the compliment, but you're only missing *the whole point*,
it seems to me.

> However, it appears that one could simply define the __repr__ method in
> the regular class FunkyReprObject, ignore metaclasses altogether, and
> wind up with the same results.

Obviously not: when you define a special method in a class, you control
operation of the *instances* of the class, NOT of the class object itself.

When you want to control operations on the class object itself, you must
then obviously define special methods in the class's class -- i.e., the
metaclass.  That's the main reason for custom metaclasses to exist at all:
the special methods that they define control operations on the classes that
instantiate them (just as a key reason for a class to exist is to define
special methods to control operation on the class's instances).


Alex



From robin at jessikat.fsnet.co.uk  Wed Apr 10 19:17:40 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 11 Apr 2002 00:17:40 +0100
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx>
 <mailman.1018391839.6535.python-list@python.org>
Message-ID: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>

In article <mailman.1018391839.6535.python-list at python.org>, Kalle
Svensson <kalle at gnupung.net> writes
>[Lulu of the Lotus-Eaters]
>> Who can tell me what's in there.  Someone on c.l.py recently let drop
>> the hint about finding the Zen of Python at the interactive prompt in
>> 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
>> 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
>> magic command was.  Can someone remind me, please?
>
>"import this", IIRC.
...
Am I the only person that finds the whole concept of Easter eggs a slap
in the face from the developers. 

When I find this rubbish in M$oft's expensive software I feel outraged.
Is it better because it's open source? I think not. On the other hand
perhaps the Python development process is itself a joke. When one
questions the idiots in charge one ends in the kill file. No one expects
the Spanish inquisition, but perhaps if rather trite philosophy is
hidden in python and we have geniuses for developers, what else may be
embedded therein? 
-- 
Robin Becker


From mwh at python.net  Thu Apr 25 07:08:59 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 25 Apr 2002 11:08:59 GMT
Subject: Is  fcntl.ioctl broken?
References: <mailman.1019696694.20193.python-list@python.org>
Message-ID: <lk8z7ckp5k.fsf@pc150.maths.bris.ac.uk>

"Noah" <noah at noah.org> writes:

> The documentation for ioctl in the fcntl module says:
> 	"ioctl(fd, op, arg) 
> 	This function is identical to the fcntl() function, 
> 	except that the operations are typically defined in 
> 	the library module IOCTL."
> 
> The only problem I can see is that there is no IOCTL module.
> I'm looking to get and set the winsize structure for the
> controlling terminal.

Oops, somethings out of date here (like the docs you're reading).

The constants you're looking for were never in the IOCTL module,
though.  They *were* TERMIOS.TIOCGWINSZ and TERMIOS.TIOCSWINSZ but now
they're in the termios module.

>>> struct.unpack('hhhh', fcntl.ioctl(0, termios.TIOCGWINSZ, "\000"*8))
(55, 80, 499, 719)

Cheers,
M.

-- 
  The meaning of "brunch" is as yet undefined.
                                             -- Simon Booth, ucam.chat


From kingprad at mail.com  Mon Apr  8 16:12:21 2002
From: kingprad at mail.com (Carl)
Date: Mon, 8 Apr 2002 15:12:21 -0500
Subject: Tkiner send command
References: <3cb0ae7d$1@news.cc.umr.edu> <C9B1D945FE30F374.B5B450B89599A68A.8783B42B36F655C5@lp.airnews.net>
Message-ID: <3cb1f89f@news.cc.umr.edu>

Oh yeah! <smacks forehead> That's obviously the best solution - so simple I
completely overlooked it. Thanks a lot!

Carl


"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
news:C9B1D945FE30F374.B5B450B89599A68A.8783B42B36F655C5 at lp.airnews.net...
> In article <3cb0ae7d$1 at news.cc.umr.edu>, Carl <kingprad at mail.com> wrote:
> >I'm having trouble going from the formal specs in the Tcl/Tk Reference to
> >writing an actual Tkinter send command. I just want to start a text
editor,
> >either the Windows default or just notepad.exe. Can someone please give
me
> >an example line of how to use it? Thanks very much!
> .
> .
> .
> Ask the question again.  You're using Tkinter, and you
> want to start notepad, is that right?  In that situa-
> tion, send probably doesn't benefit you; instead, you
> want something like
>   os.system("notepad %s" % filename)
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html




From Matthias.Hanitzsch at dynamic.de  Thu Apr 11 08:52:39 2002
From: Matthias.Hanitzsch at dynamic.de (Matthias Hanitzsch)
Date: Thu, 11 Apr 2002 14:52:39 +0200
Subject: Python extensible Emacs clone ?
Message-ID: <3CB58717.4BF3035B@dynamic.de>

Hallo,

is there a free Emacs clone that uses Python as an extension language ?

Matthias



From h.e.w.frater at cs.cf.ac.uk  Tue Apr 23 00:41:30 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 22 Apr 2002 21:41:30 -0700
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com>
Message-ID: <e2e9e807.0204222041.5fa92f05@posting.google.com>

I fully understand what everyone is going on about with regard to me
saying that the quality of the encryption is not that important. The
fact of the matter is, I'm only writing the software for my 1st year
university project and in reality it's only a theoretical system, no
one is actually going to use it. I'm using the exercise to learn
python and basic database design although to be quite honest I loathe
internet programming. Don't know why, just don't like it.

HUgh


From gerhard at bigfoot.de  Sat Apr 20 02:43:02 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 20 Apr 2002 08:43:02 +0200
Subject: How to make regexes faster? (Python v. OmniMark)
In-Reply-To: <eppstein-2C56E0.23265219042002@news.service.uci.edu>
References: <mailman.1019279977.28194.python-list@python.org> <eppstein-2C56E0.23265219042002@news.service.uci.edu>
Message-ID: <20020420064302.GA720@lilith.my-fqdn.de>

* David Eppstein <eppstein at ics.uci.edu> [2002-04-19 23:26 -0700]:
> [to Tim Peters]
> The point is that your responses don't have the appropriate References: 
> header, so they don't get threaded with other messages on the same topic.
> It appears as if you are using a mail agent rather than a news agent, 
> since References: is a news header (and your other headers say you're 
> using MS Outlook).

This might also be a bug/missing feature in Mailman's mail<->news
gateway, but if this, my response shows up correctly for you, then it's
another proof that Mutt rocks and Outlook sucks :-P

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From greg at cosc.canterbury.ac.nz  Wed Apr 10 02:41:23 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 18:41:23 +1200
Subject: [Python-Dev] Re: ANN: Pyrex - a language for writing Python 
 extension modules
References: <200204100229.OAA00739@s454.cosc.canterbury.ac.nz> <mailman.1018411525.28033.python-list@python.org>
Message-ID: <3CB3DE93.6EED687F@cosc.canterbury.ac.nz>

Paul Prescod wrote:
> 
> It would be clearer if the canonical[1] PyRex example was a bridge
> example instead of a "pure pyrex" example.

It would be good to have examples of both kinds.
Any suggestions on what should be the "hello world"
of Pyrex bridging examples? Preferably something
simple, instantly understandable to a wide range
of people, and showing off Pyrex's strengths
compared with other things like SWIG.

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


From max at malva.com.uaREMOVE.IT  Mon Apr 15 04:03:04 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Mon, 15 Apr 2002 11:03:04 +0300
Subject: C++ code generation from xml
Message-ID: <of1e9a.4vg.ln@cvs>

Hi there.

I have a Python program that uses data stored in XML file to generate some
C++ code. The problem (or, more precisely, code smell) I have is that the
actual generation algorithm is hardwired into python and therefore is
cryptic, difficult to understand and maintain.

Small example:
	def genBody(self, c):
		name = c.cxxClassName
		new = 'new %s(%s);' % (name, c.factoryCtorArg or '')
		if c.factoryNargs == 0:
			return '\treturn %s' % new
		body = '\t%s * c = %s\n' % (name, new)
		body = body + '\tc->setCmdArgs(args);\n'
		return '%s\treturn c;' % body

I'd like to have the algorithm moved out from Python source.
I've heard about XSLT language that is used to describe transformation of
XML data into HTML. I'm seeking for something similar. The difference is
that target language is C++ and not HTML.

Any suggestions?

Thanks in advance.


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


From jonathan at claggetts.net  Fri Apr  5 08:58:15 2002
From: jonathan at claggetts.net (Jonathan Claggett)
Date: Fri, 05 Apr 2002 08:58:15 -0500
Subject: Python-list digest, Vol 1 #10324 - 13 msgs
In-Reply-To: <E16tGLp-0007ou-00@mail.python.org>
References: <E16tGLp-0007ou-00@mail.python.org>
Message-ID: <1018015095.1486.141.camel@localhost.localdomain>

On Thu, 2002-04-04 at 18:08, python-list-request at python.org digested:

> >I would like to be able to define anonymous functions and classes by
> >omitting the function or class name normally found after the def or
> >class keyword. 
> 
> Feel free to write a PEP so that this abomination can be rejected once
> and for all.  ;-)  (I don't think there have been any large threads on
> this in 2002, but there certainly were at least two or three in 2001.)

I might create the PEP for posterity's sake. Should I dub it the
'sacrificial PEP'?

> (Yes, I can be quite certain it will be rejected, given Guido's regret
> over including lambda in the first place.  Python has rightly included
> many useful bits from functional programming languages, but Python is
> also emphatically *NOT* a functional language itself.)

This is interesting news to me since I've been picking Python habits up
from the latest edition of Programming Python and which uses lambda a
fair amount. Specifically, I've been using them in my Tkinter programs
to pass extra parameters to my event handlers. Is it more proper
(Pythonic) to def the helper function first and pass the helper function
name second?

Actually, I really wasn't thinking in terms of functional programming (I
can't Lisp my way out of a paper bag :-). My immediate motive was only
skin deep in that I'd like to replace the (unique) lambda syntax in my
code with the (standard) def syntax.







From jeff at ccvcorp.com  Fri Apr 12 18:38:43 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 12 Apr 2002 15:38:43 -0700
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca>
Message-ID: <MPG.172101b4b7f13f639896ae@news.nwlink.com>

In article <eGIt8.21408$de1.1000064 at news3.calgary.shaw.ca>, 
bbollenbach at shaw.ca says...
> 
> To "put it into the public domain" the only thing that
> is implicit is the fact that you hold the copyright. When you start
> adding clauses like "feel free to modify and distribute it as you like",
> etc. you've begun to establish it as Open Source (or "Free" depending on
> whose terminology you're more interested in using).

Hm, I always thought that the definition of "Public Domain" was 
that anyone is free to make use of it as they see fit.  For 
instance, books that are in the public domain can be edited and 
republished by whomever so wishes.

Note that the GPL places additional restrictions that Public 
Domain doesn't -- the intent of the GPL is to prevent material 
from being later used in a proprietary product, whereas Public 
Domain has no such restriction.  This doesn't change the fact, 
though, that Public Domain software is Free Software in the FSF 
sense of the word.  

(See http://www.fsf.org/philosophy/categories.html)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From lac at strakt.com  Wed Apr 10 13:00:02 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 10 Apr 2002 19:00:02 +0200
Subject: What has PEP 285 done to us? 
In-Reply-To: Message from Max M <maxm@mxm.dk> 
   of "Wed, 10 Apr 2002 16:09:52 +0200." <3CB447B0.1030800@mxm.dk> 
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no> <mailman.1018366940.13473.python-list@python.org> <3CB40E6F.90509@geneva-link.ch> <mailman.1018442912.3164.python-list@python.org>  <3CB447B0.1030800@mxm.dk> 
Message-ID: <200204101700.g3AH02dj023795@ratthing-b246.strakt.com>

> Laura Creighton wrote:
> 
> > re: what is in nature vs what is in perception -- see David Pye
> > The Nature and Aesthetics of Design -- especially Chapter 11
> > (function and fiction) and  the latter part of Chapter 13 (the
> > aesthetics of design) and all of chapter 14 (perception and
> > looking).
> 
> You aren't by any chance a philosophy major ;-)
> 
> regards Max M

No. Moreover, while the Philosophy section wouldn't be a poor place to
look for this book, Architecture and Civil Engineering would be my
first stops.

Laura




From gerhard at bigfoot.de  Tue Apr  2 23:14:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 3 Apr 2002 04:14:34 GMT
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net>
Message-ID: <slrnaaksu7.368.gerhard@lilith.my-fqdn.de>

Jim Jacobs wrote in comp.lang.python:
> Is it possible to simulate "C" style static variables in functions?

No.

> In methods?
 
No.

This concept sucks anyway, IMNSHO.

But you can have class-level variables.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From aleax at aleax.it  Sun Apr  7 11:43:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 07 Apr 2002 15:43:37 GMT
Subject: PEP 285: Adding a bool type (partially OT)
References: <mailman.1017466974.6751.python-list@python.org> <3CAE24A2.7ECBB78@accessforall.nl> <onpr8.64923$S52.2260694@news2.tin.it> <3CAF91A9.EE3607CC@accessforall.nl>
Message-ID: <JOZr8.69884$S52.2468332@news2.tin.it>

Ype Kingma wrote:
        ...
> The language spec does not specify the type of the index or key
> argument to __getitem__.

It does for sequences (int and slice only) and for dictionaries (any
hashable) but not for other mappings (each mapping define its
own rules: something like "only a string is accepted" is quite
common in various mappings implemented in the standard
library, for example).


>>     - testing membership with aset.has_key(item),
>> or:
>>     - indicating membership with aset[item] = 1, then
>>     - testing membership with aset.get(item, 0)
>> 
>> Why would you think this is "rare"?  It's common currency.  How many tens
>> of thousands of lines of python have you written, reviewed, or
>> maintained, that you presume to teach me what is common and what is rare
>> in python "as she is spoken", pray?
> 
> In how many of these cases would the 'item' key actually be confused
> when it could be either a bool or an int? So far you failed to convince me
> that this is not a rare case.

What does it matter what type the item is?  has_key must now return
a Boolean, while the get idiom that always was equivalent to it must
keep returning int.  Any existing code that has tests coded both ways,
for ANY type of item, will return "mixed" Booleans and int's.  Modern
equivalent idiom "item in aset" also returns Boolean, but THAT is quite
likely rare since it only became possible in Python 2.2.  Still, where 2.2
was accepted as the used language version, whenever the get idiom
was used I refactored it into the 'in' operator on code I was revising
anyway (but not on parts of code I wasn't revising) -- clearer, more
concise, AND faster... irresistible (the "one obvious way to do it", once
it became acceptable).

> Even though I don't see the relevance of the second question
> the answer is: a few, and I'm not teaching here, I'm explaining
> my opinion.

Based on WHAT experience of Python, that's what I'm asking --
maintenance responsibility for how many tens of thousands of lines,
etc.  How can you have a sound opinion on what is rare or frequent if
not on the basis of some reasonably representative sample?


>> > That should only happen when True and 1 are used interchangeably,
>> > which might happen by programmer mistake or by using
>> > old (1/0) and new (True/False) style code together.
>> 
>> Old-style only will do perfectly well, since lots of extremely common
>> idioms therein (a<b, d.has_key(k), x in y, and so on) change return
>> values silently.  Neither any mistake nor any 'together' is needed,
>> unless you're somehow trying to imply that anybody coding some kind
>> of "advanced" (?!) construct such as "a<b" was using "new style code"
>> unwittingly years ago...
> 
> By new style I meant explicit use of True/False. 

So do I: and there is absolutely no "explicit use of True/False" in any
of the several expressions I enumerted, nor the many more whose
return type change silently from int to Boolean.  "True and 1 are used
interchangeably" in any existing function that can return any of these
expressions OR an explicit 0 or 1, as in:

def isValueAtIndex(alist, anindex, avalue):
    if -len(alist)<=anindex<len(alist):
        return alist[ainindex] == avalue
    else:
        return 0

or equivalently and more Pythonically:

def isValueAtIndex(alist, anindex, avalue):
    try: return alist[ainindex] == avalue
    except IndexError: return 0

There is absolutely no programmer mistake here, nor any use of old
style and new style code together.  Yet such existing code may very
well return int or Boolean interchangeably _because the return type
of very common operations has changed silently in 2.3_ ... !!!

> As for "advanced (?!): I don't have a time machine.

So you'll have to admit that claiming "interchangeable return of int
or Boolean" can only be caused by mistakes or "mixing old and
new style" was a wildly unfounded assertion, yes?


>> > In the pure case an int() call would not be needed at all:
>> 
>> But to check whether it's needed or not, a type-check would be
>> needed, and that IS a substantial slow-down.  You don't seem to
> 
> If it's not needed at all, there is also no need for the check.

How do you *KNOW* whether to check or not, without a check?


> I am not talking about an actual implementation.
> I'm talking about a specification, while trying to keep an eye on
> implementABILITY.
> Since you didn't say it cannot be done, I'm actually encouraged.

It IS quite possible to implement a language where the semantics
of, say, + and - operators, are switched in any source line which has a
prime number of characters.  It IS totally absurd, but clearly feasible.

Why such technical feasibility would encourage anybody to consider
such absurdities is beyond me, but please feel free to go ahead and
implement such a beauty.  It will still be less absurd and abstruse
than your cherished "implementable" specification, anyway.


Alex



From narnett at mccmedia.com  Tue Apr  2 20:03:17 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Tue, 2 Apr 2002 17:03:17 -0800
Subject: A faster Python?
In-Reply-To: <a8dcqr$p29$1@laurel.tc.umn.edu>
Message-ID: <NDBBLAEMOMEHKALAMKKJKEOKJOAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Scott Hennesy
> Sent: Tuesday, April 02, 2002 2:54 PM
> To: python-list at python.org
> Subject: A faster Python?
>
>
> For what I do, 90% of the time Python is fast enough. But for
> that 10% of the time I find myself wishing for a faster Python.

Generally speaking, the times I most often feel constrained by processing
speed have to do with analyzing large amounts of data.  I imagine that as
bandwidth increases and there are more and more automated methods of
generating data that can be analyzed, more and more people will feel the
same pain.  People do want to analyze more and more data -- log files,
e-mail and far more.  I'd very much welcome any kind of speed improvements
that would help in that area.

To be specific, I do a lot of time series, correlation, regression and
clustering analysis.  Those can really bog down.  While I'm fine with
exporting to Excel, SPSS, etc., for heavy-duty analysis, the more I can do
without leaving Python, the happier I am.  And I'm considering a distributed
version of some of the things I'm working on, which I'll only do with open
tools that won't cost users anything.

Nick




From anthony_barker at hotmail.com  Tue Apr 16 00:19:59 2002
From: anthony_barker at hotmail.com (Anthony_Barker)
Date: 15 Apr 2002 21:19:59 -0700
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <29e28c51.0204090215.38f92899@posting.google.com>
Message-ID: <899f842.0204152019.3c13a1e5@posting.google.com>

my 2 cents...

Look at the K&R book and see if you like it and then buy it or 
the Harbison and Steele as your main reference book

Then get another dumbies style book from the Library. I recommend
"Learn c programming in Linux in 21 days" from SAMS..

What this book does is give you the discipline to plow through the
topic in 3 weeks (especially usefull if you are working). Also their
descriptions and diagrams of pointers are more intuitive."C
Programming: A Modern Approach" - a textbook also has good
explanations of pointers...

If you have some programming background, pointers (pointers to
pointers to structs etc), defensive programming (due to pointers) and
perhaps data structures are the initial challenging things in c.

Once you are done with those and if you still want more I recommend 
"Expert C Programming" by Peter van der Linden. It is one of the most
entertaining programming books I have read - his love of c programming
shines through every word.

I also just picked up a used copy of "Advanced Linux Programming"
which is available as a pdf at codesourcery. There are several good
references to the standard c library available on the web.

Anthony

 -o)
 /\\
_\_V   Anthony Barker
Linux & Python addict
http://xminc.com/anthony/

Pointers
http://www.howstuffworks.com/c.htm
http://home.netcom.com/~tjensen/ptr/pointers.htm

Library
http://www.neilstuff.com/index.php?p=80
search google for more..........


cimarron+google at taylors.org (Cimarron Taylor) wrote in message news:<29e28c51.0204090215.38f92899 at posting.google.com>...
> > I am wondering what book would be the best for a begginer? 
> 
> If you want the best overall value, I recommend three books: 
> 1. K&R 
> 2. "The Standard C Library" by P.J. Plauger, and 
> 3. "The Dictionary of Standard C" by Rex Jaeschke.  
> 
> Cim


From peter at engcorp.com  Wed Apr 10 21:01:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 21:01:55 -0400
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3CB4BB04.1B8DD2A6@bgb-consulting.com>
Message-ID: <3CB4E083.7EC2EDC@engcorp.com>

Don Garrett wrote:
> 
> "M. J. Mix" wrote:
> >
> > Is anyone using Python as a spacecraft ground system scripting
> > language, or onboard (embedded) language?
> >
> > Pros and cons when compared with TCL or JAVAScript.  Just
> > some real life, honest opinions.
> 
>   I'm not aware of ANY scripting languages that have been developed in a manor
> suitable for life critical fail-safe situations. I like Python a lot, and find
> it very reliable for day to day stuff, but I'd hate to have it embedded in my
> pace maker. 

Agree.  Unsuitable for pacemakers, or any other life-critical
application, at this time.  Not inherently unsuitable mind you.

> I also don't think it's suitable for serious real-time processing.

Don't agree at all.  Python is completely suited for real-time processing
if you can handle two constraints: it is not as fast as some other 
languages, so you have to accept latency on the order of milliseconds to
seconds as opposed to microseconds, and no one (to my knowledge) has
done a thorough analysis of the performance characteristics of the
interpreter so you have no guarantees unless you do that yourself.
(Translation of that last: can't be used for truly "hard" realtime
applications of any time scale without someone doing the proofs.)

>   Fail-safe and Real-time processing are both rather specialized and arcane
> arts.  

I tend to think of them as more hard-core engineering than art, but
maybe that's a matter of opinion.  Or background?

> Embedded systems in spacecraft often fall into these categories
> (just like pacemakers). Standard Python is just not suitable for these tasks,
> though specialized versions could be developed (given the types of
> cost/expense usually associated with such systems).

Pacemakers no, embedded systems even such as in spacecraft, why not?
Other than the need to do a more thorough risk analysis because of 
the incredibly high cost of spacecraft failure, what would need to
change in Python for it to be suitable?

(The reason I say all this is that I'm using Python for what
are considered embedded, mission critical applications with some 
near-hard realtime response requirements on the order of seconds 
and I don't agree with the "just not suitable" claim at all.  
It depends entirely on the degree of risk deemed acceptable.
Because of the expected lower defect rate, higher productivity,
and lower cost of maintenance of Python code we feel its use is 
not only acceptable, but almost mandated.)

-Peter


From starx at pacbell.net  Tue Apr 16 17:10:32 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 16 Apr 2002 21:10:32 GMT
Subject: O_O <- Re: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net> <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1>
Message-ID: <Xns91F2905ED8DAARASXnewsDFE1@130.133.1.4>

Duncan Booth || Mon 15 Apr 2002 04:35:13a:
> [...]
> __paul_prescod_hack_constant__
> [...]

Now there's an interesting variable.

Also interesting that _..._undefined_..._ has so many underscores... 


-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From prova at qui.com  Sun Apr 28 05:33:29 2002
From: prova at qui.com (Utente)
Date: Sun, 28 Apr 2002 09:33:29 GMT
Subject: Help: py2exe and os.path
Message-ID: <JlPy8.72016$vF6.2149385@news2.tin.it>

Hello everybody. I need some help using py2exe. I'm building a program using
Python 2.2, wxPython and a module i'm developing.
I would like to distribute the program as a windows exe, and i'm playing
with py2exe 0.3.1 (very nice!). The program is getting bigger every day.
The problem is that i can't understand two things:
py2exe works fine, but i get the following error (this is the only error in
a long source!!!):

warning: py2exe:
***************************************************************
**********
warning: py2exe: * The following modules were not found:
warning: py2exe: *   os.path
warning: py2exe:
***************************************************************
**********

I'm using many modules, and with IDLE it works fine! Why doesn't it find the
os.path module? Actually i only need a function from os.path to see if a
file exists (if so, i load data from it), but the "compiled" version never
tests this and always creates it from scratch. :(

Thanks in advance for any help or suggestion...
Max




From tim.one at comcast.net  Thu Apr  4 21:13:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 04 Apr 2002 21:13:25 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEACOLAA.tim.one@comcast.net>

[James Logajan]
> ...
> If a bool is to be treated as an int sometimes and sometimes not,

bool is a subclass of int that overrides only the constructor, __str__ and
__repr__ methods.

> I don't recall seeing anything about whether it will be treated as
> an int as a dictionary key.

Provided you don't convert it to string via str() or repr() first before
using it as a dict key, yes.

> If I do:
>
> x = ["is", "are"]
> y = {0: "", 1:"s"}
>
> for n in range(3):
>    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
>
> Will the behavior be backword compatible and equivalent or not? Simple
> question I assume; waiting for an answer.

C:\Code\python\PCbuild>python
Python 2.3a0 (#29, Apr  3 2002, 20:49:03) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> x = ["is", "are"]
>>> y = {0: "", 1:"s"}
>>> for n in range(3):
...     print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
...
There are 0 items.
There is 1 item.
There are 2 items.
>>>





From h_schneider at marketmix.com  Tue Apr  9 03:42:47 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Tue, 9 Apr 2002 09:42:47 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <mailman.1018291584.15367.python-list@python.org>
Message-ID: <a8u5nj$alq$01$1@news.t-online.com>

Hi Cliff,

thanks for your answer! On what kind of clients do you run this software?
My customer has an mixed Win2K, Win98-environment. Especially the Win98
machines might cause stability probs when running wxPython. Do you have any
experiences in this direction ?

All the best,
Harald

"Cliff Wells" <logiplexsoftware at earthlink.net> schrieb im Newsbeitrag
news:mailman.1018291584.15367.python-list at python.org...
> On Mon, 8 Apr 2002 11:55:24 +0200
> Harald Schneider wrote:
>
> > Hi,
> >
> > what are your experiences, using wxPython/wxWindows in large projects?
> >
> > Large means in my case a sales system, based on MySQL.
> > I'm thinking about implementing it in wxPython, but I'm not sure about
the
> > stability of
> > wxWindows and the wxPython wrapper ...
>
> While working as a developer at a call center (telephone sales and
surveys) I
> developed a system using Python, wxPython and MySQL.  I did encounter the
> occasional problem when upgrading wxPython versions, but nothing that
wasn't
> readily fixed.  It was a moderately complex system, having a custom markup
> language for prompting the telemarketer (so different sales regions could
use a
> different "script" without code changes) that was parsed and passed
through as
> HTML to the wxPython HTML widget.  Besides the database access, the
software
> also had to control an autodialer via telnet and have a fallback dialing
system
> in case the autodialer failed.  The system ran on 90 seats, 16 hours a
day, 7
> days a week for over two years. It was fast, stable, and any of the
problems we
> had could be usually be directly attributed to me =)
>
> Also, if you have doubts about whether the lack of a standard/stable/* IDE
for
> wxPython is a problem, consider that this system went live less than 30
days
> after the initial spec. There was only one developer (me) using only
XEmacs
> (although the first couple of days in production were a special sort of
hell ;)
>
> Regards,
>
> --
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
>
>




From logiplexsoftware at earthlink.net  Mon Apr 29 13:06:30 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 10:06:30 -0700
Subject: Super-newbie syntax error question
In-Reply-To: <Cg%y8.5752$9F5.405701@typhoon.austin.rr.com>
References: <Cg%y8.5752$9F5.405701@typhoon.austin.rr.com>
Message-ID: <20020429100630.3c64c6ec.logiplexsoftware@earthlink.net>

On Sun, 28 Apr 2002 23:07:14 GMT
Charlie Wood wrote:

> I have some code that I've cobbled together, and have just introduced a
> syntax error. I've read through the docs, done lots of trial and error, and
> I'm left scratching my head. Any ideas?
> 
> Here's the code:
> 
>     def endElement(self, name):
>         if self._parent == "item":
>             if name == "title":
>                 self._title = self._text
>             elif name == "link":
>                 self._link = self._text
>             elif name == "description":
>                 self._descr = self._text
>             elif name == "category":
>                 self._category = self._text
>             elif name == "item":
>                 if not self._list_started:
>                     self._out.write("\n")
>                     self._list_started = 1
> 
>                 self._out.write('%s: %s\n%s\n%s\n\n' %
>                                 (self._category, self._title, self._descr,
> self._link))
> 
>     f = file("/article.txt", "w")
>     f.write('Category: %s\nTitle: %s\nDescription: %s\nLink: %s\n\n' %
>       (self._category, self._title, self._descr, self._link))
>     f.close()
> 
>                 self._title = None
>                 self._link = None
>                 self._descr = ""
>                 self._category = ""

Other people have addressed your syntax error, but I'd like to point out that
you may want to use a dictionary rather than assigning to attributes as this
could shorten your code quite a bit:

class SomeClass(...):
    def __init__(self, ...):
        self._initAttribs()
        ...

    def _initAttribs(self):
        """explicitly initialize dictionary since it isn't clear 
           whether all values will be set later"""
        self.attribs = {
            'category': '',
            'description': '',
            'title': None,
            'link': None
        }

    ...

    def endElement(self, name):
        if self._parent = "item":
            if name != "item":
                self.attribs[name] = self._text
            else:
                if not self._list_started:
                    self._out.write("\n")
                    self._list_started = 1
                self._out.write('%(category)s:'
                                '%(title)s\n'
                                '%(description)s\n'
                                '%(link)s\n\n' 
                                 % self.attribs)

                # unsure of the indentation level of this bit since it appears
                # to be incorrect in your original code (also note using "open"
                # rather than "file")
                f = open("/article.txt", "w")
                f.write('Category: %(category)s\n'
                        'Title: %(title)s\n'
                        'Description: %(description)s\n'
                        'Link: %(link)s\n\n'
                        % self.attribs)
                f.close()
                self._initAttribs()

Whether you can take this approach of course depends upon your particular
application, but whenever I see if/elif/else start to stretch out, I start
thinking "lookup table".  Not only does it eliminate long if/elif/else
constructs, but you can use it in format strings as well, which I find to be a
bit clearer.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From martin at v.loewis.de  Wed Apr 17 03:26:11 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 17 Apr 2002 09:26:11 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018875990.21271.python-list@python.org> <j43cxwm1o4.fsf@informatik.hu-berlin.de> <a9id9a$33oqt$1@ID-99513.news.dfncis.de>
Message-ID: <m37kn6wzu4.fsf@mira.informatik.hu-berlin.de>

"P. Alejandro Lopez-Valencia" <dradul at yahoo.com> writes:

> And while you are at it... You have not compiled python itself as PIC
> code. No, gcc -shared *doesn't* work in Solaris unless you are using GNU
> Binutils instead of the native linker and assembler, something *not
> recommended at all*. 

Can you elaborate? gcc -shared works fine on Solaris, and binutils
also works without problems.

> Solaris' linker can't bild a dynamic loading library/module using
> object code compiled in position dependent mode.

And rightly so. That should never be a problem, though.

Actually, you *can* talk the Solaris linker to incorporate non-PIC
text into a shared library. For that, it will make the text segment
writable. The -mimpure-text option of gcc triggers that feature.

> This means, that you need to have all external libraries compiled with
> the -fPIC flag, even if they are static. 

If you link static libraries into Python, I recommend to make the
modules that use them also static. Then you don't need PIC code for
those.

> Ahh! And make sure that you give the linker the actual library paths
> you use with an -R flag, else the dynamic loader won't find them and
> you'll have to use an explicit LD_LIBRARY_PATH environment variable
> for each invocation invocation of the internpreter.

That is the prime reason why I dislike shared libraries in the first
place - statically linked binaries are much easier to manage; use of
shared libraries should be restricted to system libraries, IMO.

Regards,
Martin


From aahz at pythoncraft.com  Tue Apr  2 09:56:07 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Apr 2002 09:56:07 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <3CA95E0A.3070507@mxm.dk> <a8bvd0$rg3$1@slb3.atl.mindspring.net>
Message-ID: <a8cgq7$pta$1@panix1.panix.com>

In article <a8bvd0$rg3$1 at slb3.atl.mindspring.net>,
Andrew Dalke <dalke at dalkescientific.com> wrote:
>Max M wrote:
>>
>>This is a common idiom:
>>
>>def f(names=None):
>>     if names == None:
>>         names = []
>>     ...
>
>Hmmm.  I would have done "if names is None".  No chance of confusing
>the object with an object that overrides __cmp__.

Yup.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From bokr at oz.net  Tue Apr 16 17:46:37 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Apr 2002 21:46:37 GMT
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>  <slrnablens.5k9.mlh@vier.idi.ntnu.no> <mailman.1018897170.28523.python-list@python.org>
Message-ID: <a9i63t$vje$0@216.39.172.122>

On 15 Apr 2002 13:58:11 -0500, Ian Bicking <ianb at colorstudy.com> wrote:

>On Mon, 2002-04-15 at 06:35, Magnus Lie Hetland wrote:
>> But (in the (very?) long run) I think that the name "input" should
>> then be used for what is now raw_input. I know -- breakage, and all
>> that... I just think it fairly odd to have *only* raw_input, since the
>> "raw" qualification then becomes fairly meaningless.
>
>Why not just make input into what it should be, a moderately-evaluating
>raw_input... like:
>
>def input(prompt=''):
>    val = raw_input(prompt)
>    for converter in [int, float, complex]:
>        try:
>            return converter(val)
>        except ValueError:
>            pass
>    return val
>
>
>Of course, it gets more complex if you allow lists, dictionaries, and
>tuples... which you probably should.  But that's certainly doable while
>still being safe.
>

You could also just add a sequence of validation conditions, e.g., (pseudocode, not tested):

    def prompted_input(prompt='', *validation_seq):
        val = [raw_input(prompt)]
        for valcond in validation_seq:
            valcond[0](val, valcond[1:]) #raises exception if cond violated
        return val[0]

to be called something like (example shows more than usual checks ;-):

        prompted_input(
            'Enter a number: ',
            (vfun_conv, [int], 'Error: %(theValue) is not an integer.'),
            (vfun_min, 3, 'Need at least %(theMin)'),
            (vfun_max, 8, 'Must be no more than %(theMax)'),
            (vfun_sel, (3,7), 'Must select from %(theSet)'),
            (mySpecialCheck, myArgs)
            )

with vfun_xxx being builtin standard validation functions, but the list not
limited to those. Custom error text would be passed through via ValidationCheck exception.
Using [val] vs val allows possibility of repairing/modifying something in mySpecialCheck.

Regards,
Bengt Richter


From bokr at oz.net  Sat Apr  6 23:58:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Apr 2002 04:58:38 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <0rsr8.1752$CU.3728@rwcrnsc51.ops.asp.att.net> <a8odr2$rdh$0@216.39.172.122>
Message-ID: <a8ojlu$bg9$0@216.39.172.122>

On 7 Apr 2002 03:18:58 GMT, bokr at oz.net (Bengt Richter) wrote:

>On Sat, 06 Apr 2002 01:45:32 GMT, "Emile van Sebille" <emile at fenx.com> wrote:
>
>>"robin and jim" <robinjim at earthlink.net>
>>> Is there a way to convert a single quoted string to a triple quoted
>>string?
>>>
>>> For example, given:
>>>
>>> ''' abc %d xyz '''
>>>
>>> the result of ''' abc %d xyz''' % 10 is:
>>>
>>> ' abc 10 xyz '
>>>
>>> which is a single quoted string.
>>>
>>> I would like to convert the result to a triple-quoted string.
>>>
>>>
>>
>>What do you want to do with it?  Once a string is created, it doesn't
>>matter which style quoting was used -- it's simply a string.
>>
>True, but notice:
>
> >>> """ ' """
> " ' "
> >>> """ " """
> ' " '
> >>> """ ' " """
> ' \' " '
> >>>
>
>perhaps he would have liked """ ' " """ back instead of ' \' " '
>Seems like just as valid as switching the other quotes, e.g., a quick hack
>to simulate extending the quote-choosing of repr:
>
> >>> def repr3(x):
> ...     rep = repr(x)
> ...     if rep.find('\\n')<0 and rep[1:-1].find("\\'")<0: return rep
> ...     rep=rep[1:-1].replace('\\n','\n')
> ...     rep=rep.replace("\\'","'")
> ...     return '"""' + rep + '"""'
> ...
> >>> tests = [""" ' """, """ " """, """ ' " """, """Line 1
> ... Line 2"""]
> >>> for t in tests: print 'plain: %s\n repr: %s\nrepr3: %s\n'% (t, repr(t), repr3(t))
> ...
> plain:  '
>  repr: " ' "
> repr3: " ' "
>
> plain:  "
>  repr: ' " '
> repr3: ' " '
>
> plain:  ' "
>  repr: ' \' " '
> repr3: """ ' " """
>
> plain: Line 1
> Line 2
>  repr: 'Line 1\nLine 2'
> repr3: """Line 1
> Line 2"""
>
Well, it needs some logic for handling embedded triple quotes, and preventing
unescaping from creating unintended triples quotes in the representation, but
I don't have time right now ... it was a "quick hack" ;-)

Regards,
Bengt Richter


From aahz at pythoncraft.com  Tue Apr  2 10:07:31 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Apr 2002 10:07:31 -0500
Subject: Python style (was Re: PEP 285: Adding a bool type)
References: <B43D149A9AB2D411971300B0D03D7E8BF0A240@natasha.auslabs.avaya.com> <mailman.1017733985.13744.python-list@python.org>
Message-ID: <a8chfj$rij$1@panix1.panix.com>

In article <mailman.1017733985.13744.python-list at python.org>,
Tjabo Kloppenburg  <tk at taponet.de> wrote:
>>
>> def f (names=None):
>> def f( names=None ):
>
>U program da bad style :-),

You should read http://python.sourceforge.net/peps/pep-0008.html
to learn about Python style.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From aleax at aleax.it  Tue Apr  9 10:20:44 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 14:20:44 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <23891c90.0204090012.7fc80c1f@posting.google.com> <Xns91EB245A0BCD9RASXnewsDFE1@130.133.1.4> <23891c90.0204090526.18675d02@posting.google.com>
Message-ID: <0NCs8.63131$pT1.1810689@news1.tin.it>

Paul Boddie wrote:
        ...
> sentiments.) However, I do feel that the more recent features seem
> increasingly remote to the beginner and even the more

This can only come from a very selective view of "recent features".

Nothing is more natural to a beginner than "if x in somedict:" and
"for x in thefile:" -- having had the opportunity to teach some
beginners with 2.2, I've seen them slurp these up instantly (just
try to contrast with "if somedict.has_key(x):" and
"for x in thefile.xreadlines():" ...!).  And the new features are
the right way to implement such obvious progress -- iterators, for
example, let you loop on (e.g.) dictionaries or files, without
requiring such types to define unnatural (or unfeasible) "indexing"
then go out of their way to stop misuse.

> advanced/experienced Python developer.

You MUST be joking -- inheriting from builtin types is SUCH a win,
and so obviously so, for advanced and experienced developers! 

> List comprehensions (introduced
> in 2.0) do appear in the standard tutorial, but I can't easily find
> generators, even though they are documented in the language reference.

Generators need a "from __future__ import", so I applaud the tutorial's
choice not to burden beginners with them YET.

> other things to do during working hours) it becomes increasingly hard
> to keep track of what's great in the more recent Python releases.

I think AMK's "what's new" articles are great for that purpose.



Alex



From starx at pacbell.net  Thu Apr 11 03:31:23 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 11 Apr 2002 07:31:23 GMT
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <mailman.1018495402.12499.python-list@python.org> <Xns91ECE5A88C0F8RASXnewsDFE1@130.133.1.4> <3CB523FD.7D83C0C@engcorp.com>
Message-ID: <Xns91ED580658BARASXnewsDFE1@130.133.1.4>

Peter Hansen || Wed 10 Apr 2002 10:49:49p:

> Philip Swartzleonard wrote:
>> 
>> Geoff again:
>> > We're leeches (so far) but at least we're very grateful leeches.
> 
> Just for the record, 'twas I not Geoff who said that.  Geoff may
> be a leech, he may even be a veritable wart on the bottom side of
> society, but so far he hasn't admitted it here.  <wink>
> 
> 'course-he-may-also-be-a-really-smashingly-nice-guy-ly yr's,
> -Peter

Oops, I looked at both articles, figured out who wrote what, stuck that on 
the end, and wrote down the other name. Doh %D.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From paul.osman at sympatico.ca  Thu Apr 18 00:31:21 2002
From: paul.osman at sympatico.ca (Paul Osman)
Date: Thu, 18 Apr 2002 00:31:21 -0400
Subject: Hostname Resolution not working in Explorer
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net>
Message-ID: <3CBE4C19.9050106@sympatico.ca>

Gary,

Zope by default runs on port 8080. This is why you type :8080 after the 
name of the server/domain. If you leave the port blank in a browser 
location bar, the browser defaults the request to port 80 (which is the 
default port for most web servers -- i.e. IIS, Apache). If you want zope 
to run on port 80 by default, you will have to play around with the 
configuration (I'd point you in the right direction but I don't actually 
know much about Zope).
And btw:

http://127.0.0.1:8080/mange
http://localhost:8080/mange
http://hostname:8080/mange

will all mean the same thing when you're on the same machine as the server.

Hope this helps.

-Paul Osman

Gary Spivey wrote:

>I have windows XP Home, and IE6. I have a zope server running on my box.
>The server is running fine, and I can access if it I access
>
>http://127.0.0.1:8080/mange
>
>However, if I use
>
>http://localhost/mange
>
>the page isn't found. The same holds true if I use my hostname rather than
>localhost - also, the actual IP address works.
>
>I have made sure that the name entries exist in the
>window/system32/drivers/etc/HOSTS file, and I can ping both localhost and my
>hostname from a command window (with the correct address resolution).
>
>So - why can I not get the resolution to work from within Internet Explorer?
>
>Cheers
>Gary
>
>






From loewis at informatik.hu-berlin.de  Tue Apr  2 11:51:39 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Apr 2002 18:51:39 +0200
Subject: shelve files
References: <89a9b6ce.0204020830.7934adf0@posting.google.com>
Message-ID: <j4pu1i11t0.fsf@informatik.hu-berlin.de>

jkring at calbay.com (Jim) writes:

> I moved a shelve file from one machine to another and I cannot open it
> on the new machine.  Is there something that I should be aware of?  

You need the same version of bsddb on both machines (of the database
package, not the Python module). If you cannot arrange for using the
same BSDDB version, you need to use a different underlying dbm
package, such as gdbm or dumpdbm (although the latter is not
recommended).

It may be that you can transfer the data successfully with
db_dump/db_load; I haven't tried that.

Regards,
Martin


From phr-n2002a at nightsong.com  Wed Apr  3 19:49:58 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 16:49:58 -0800
Subject: [Python-Dev] PEP 285: Adding a bool type
References: <mailman.1017875889.27696.python-list@python.org>
Message-ID: <7xzo0kfft5.fsf@ruckus.brouhaha.com>

"Delaney, Timothy" <tdelaney at avaya.com> writes:
> I have only one other issue - the PEP does not deal with the case of sending
> data to an older version of Python (which I believe Alex Martelli brought
> up, though I could be wrong).
> 
> Perhaps the section that states int(bool(x)) (resolved issues) should
> include that the structure should be used for backwards-compatibility with
> previous versions of Python.

I think the pickle and cPickle constructors should take a flag to do
this conversion automatically.  For marshal, the flag isn't needed
since marshal doesn't claim to be compatible across versions.


From claird at starbase.neosoft.com  Tue Apr 16 09:11:40 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 16 Apr 2002 08:11:40 -0500
Subject: Python and spacecraft onboard or ground control scripting
References: <mailman.1018941435.32338.python-list@python.org>
Message-ID: <01DC83A0B79E6132.0A4063AD62DFC5DE.BB706949341DB607@lp.airnews.net>

In article <mailman.1018941435.32338.python-list at python.org>,
Christian Tanzer <tanzer at swing.co.at> wrote:
>
>claird at starbase.neosoft.com (Cameron Laird) wrote:
>
>> Yes, for example, Python objects are dynamically created, and there is
>> the potential for "out of memory". I have a personal belief, however,
>> that the memory analysis of general C programs is no easier to compute
>> than that of Python programs.
>
>IIRC, the OP talked about mission critical systems. In many such
>systems, you are not allowed to allocate memory dynamically. While I
>agree that memory errors are much more likely in a C program than in a
>Python program, this doesn't hold when there isn't any runtime memory
>allocation going on in the corresponding C program.
			.
			.
			.
Entirely true.  I welcome all help making the context
of these propositions sufficiently precise so that it's
meaningful to consider whether they're true or not.  I
recognize that I'm a bit touchy about how the interpre-
tation of "mission-critical" can wander about in the
course of a conversation.
-- 

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


From phr-n2002a at nightsong.com  Fri Apr 12 19:11:11 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 16:11:11 -0700
Subject: big ints & elliptic curves
References: <a8v338$4sl$1@watserv3.uwaterloo.ca>
Message-ID: <7xwuvch5rk.fsf@ruckus.brouhaha.com>

"John" <none at none.com> writes:
> Is there any good library for multi-precision arithmetics in Python? Also
> looking for elliptic curve implementations.

Python already includes a long int type (implemented in C) whose performance
isn't so bad.  For more speed you can use gmpy.  

Using either the built-in arithmetic or gmpy to implement elliptic curve
arithmetic in GF(p) should give acceptable results.  For GF(2**N),

    http://sourceforge.net/projects/cryptkit/  

includes a Python port of Mike Rosing's package.


From stephen at xemacs.org  Wed Apr  3 22:20:07 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 04 Apr 2002 12:20:07 +0900
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: <3CAB8C99.232A8571@metaslash.com>
References: <200204032232.g33MWqJ28933@odiug.zope.com>
	<3CAB8C99.232A8571@metaslash.com>
Message-ID: <87k7roxi8o.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Neal" == Neal Norwitz <neal at metaslash.com> writes:

    Neal> This code [...]  generates these warnings:

    Neal> 	test.py:6: Comparisions with True are not necessary

That doesn't look like "shinai"[1] to me.  Is it strong enough?

Footnotes: 
[1]  The split-bamboo sword preferred by Zen masters because it hurts
more, with less permanent damage---to the bamboo, anyway.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.




From gyromagnetic at excite.com  Thu Apr 25 12:09:27 2002
From: gyromagnetic at excite.com (gyromagnetic)
Date: 25 Apr 2002 09:09:27 -0700
Subject: Dictionary that adds items to itself
Message-ID: <4620daca.0204250809.563a8c0b@posting.google.com>

Hi,
I would appreciate some advice on the following. I would like to
create a dictionary-like object with the same properties as a normal
dictionary except with regard to adding items. When items are added, I
would like them to be processed by a function that may generate
additional items. I would like these additional items to be added to
the dictionary, have the function applied to them, have the resulting
items added to the dictionary, etc.

Can this be done by subclassing UserDict? Is there another way?

I am using Python2.1.3.

Thanks.

-gyro


From Oschler at earthlink.net  Mon Apr  8 01:54:08 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 8 Apr 2002 01:54:08 -0400
Subject: For SuSE Linux users who want Zope
Message-ID: <Dfas8.19447$gA5.2718940@e3500-atl2.usenetserver.com>

This message is for any SuSE Linux users (or perphaps other distros that
bundle Zope) that are having trouble getting Zope running.  I am running
SuSE Linux 7.3 Pro.  I installed Zope and Python using the Yast2
installation program.  Although the Linux install went beautifully, when I
tried to get Zope running I got absolutely nowhere (permission problems,
directory path problems, configuration problems, etc).  If you are not
having trouble getting Zope running disregard this message, it doesn't apply
to you.

If you installed Zope, remove it.  If you have trouble uninstalling Zope
under SuSE as I did, see the end of this message for help on that:

Here's what works well:

Get a copy of "Zope: Web Application Construction Kit". (It's a great book,
Amazon has it).  Follow the instructions on pp. 20 and 21 with the following
important fix:

Do NOT do:

./install -u nobody -g nobody

as the book tells you to on p. 21.  Under SuSE 7.3 there is no group called
'nobody', instead use:

./install -u nobody -g nogroup

If you accidentally use the book's version, you'll see a ton of change
group/owner error messages since 'nobody' is not a valid group name.  You'll
have to kill the entire install directory and re-unpack the tarball.  (If
you just rerun "./install" there will be lots of files left with the bad
group setting, that's why you have to kill the tainted directory first).

That's it!  Everything goes very easily from then on.  Of course, YMMV.
Hope this helps someone.

thx

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

(These messages are copied from another thread I started on
alt.os.linux.suse regarding trouble uninstalling Zope.

I'm trying to uninstall Zope from my SuSE 7.3 Pro Linux box.  I use Yast2's
Install/Remove Software module.  I search for Zope and it shows it as
installed (with 'i' next to it).  I then select it for uninstall (now has an
'X' next to it).  It asks for CD4, says its uninstalling, and then rebuilds
the system configuration.  I exit Yast2, reenter the Install/Remove Software
module and search for Zope.  It still shows Zope as being installed.

What am I doing wrong?

Nevermind, turns out that; zope-mysql*, zope-doc*, and zope-doc-pdf* were
dependent on it.  Apparently Yast2 in 7.3 doesn't warn you and just ignores
the request.  I had to go to an xterm window, and then I typed "rpm -e
zope".  It told me the names of the dependent packages, so I removed each
one of those too using the "rpm -e" command, and then reissued the "rpm -e
zope" command which executed fine.

Note, I discovered that you do not use the version info following the
package name with the "rpm" command, just the package name.  For example, to
remove "zope-doc-2.4.1-18" I typed "rpm -e zope" _not_ "rpm -e
zope-doc.2.4.1-18".  Be very careful not to ruin your system by uninstalling
anything important, rpm does _not_ issue a confirmation question before
eliminating a package.  (At least on SuSE 7.3 Pro)

<end>







From aleax at aleax.it  Mon Apr  1 03:52:30 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 01 Apr 2002 08:52:30 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <idVp8.48143$S52.1619383@news2.tin.it>

Guido van Rossum wrote:
        ...
>     I'm particularly interested in hearing your opinion about the
>     following three issues:
> 
>     1) Should this PEP be accepted at all.

No, -1, please don't.  It seems to me that the advantages are small:

a. We don't have to model _every_ important concept with a type.  A
_protocol_, yes, but we do already have a Truth 'protocol' just as we
have (say) an Iterator one without needing an Iterator _type_.

b. The hoped-for uniformity of notation won't happen -- we've already 
seen on these threads people who want every constant to be uppercased
(to make it 'stand out') and so will use TRUE=True etc anyway.

c. The ability to model Booleans separately from integers when interfacing
external stuff (databases, rpc systems, etc) cuts both ways -- we may be a 
little better placed regarding external systems that do draw this 
"distinction without a difference" (although it seems to me that we're
doing fine already), we'll be a little worse placed regarding external 
systems that don't (such as, we're told, MySql) -- and the transition
costs will, I think, balance out and nullify any hoped-for gain.

d. Other languages have (in some cases) their own reasons to introduce
boolean types -- e.g., in C++, the ability to define an 'operator bool' that
is distinguished from 'operator int' (although this hoped-for gain did NOT
in practice materialize due to the transparent bool->int conversion...) --
but in Python we already have __int__ and __nonzero__ specials (and we
hear the latter is not due for renaming anyway) so, no advantage here; in
Java, the need to diagnose errors for such constructs as 'if(x=0)' -- but
in Python, we already forbid that, so, no advantage here; and so on.

e. normalization to 0 or 1 might better be accomplished by making
operator.truth better known (we've heard about its introducer, Jim Fulton,
having forgotten about its existence and still pining for it) and more
available, e.g. by hoisting it up to builtins (name it bool or name it truth
or whatever, but have it always available) -- a minimal change that has
some small but real benefits at very small cost (for compatibility I guess
operator.truth would stay available, just like module exceptions stays
around even though exception classes are built-ins).

f. the claimed conceptual benefits in easing the lives of programmers and
newbies in particular I consider as illusions.  Programmers want booleans
iff they're used to them from previous languages -- I've never seen people
used to (e.g.) APL, good old (uncommon?-) Lisp, or (dare I say it) Perl,
feel sad and deprived by the lack of booleans in their respective languages.
I _have_ seen old-time Fortranites rant quite convincingly against that
useless, newfangled LOGICAL stuff.  And since we can't spell the type in
every way at once, we're not going to soothe all unjustified desires due to
previous habits anyway -- if it's bool the Javaites will pine for boolean, 
and viceversa, and those Fortranites (if any) who've come to love their
LOGICAL's won't find them anyway.  Just as the constants won't be
spelled 'True' AND 'true' AND 'TRUE' and (particularly:-) '.TRUE.' too...

Newbies in my teaching experience grok 0/1 at once.  Look at the on/off
switch of many modern appliances and what do you see there?  0 for off,
1 for on.  I think it's even some kind of international standard, but in any
case there IS a lot of precedent for it and I've gotten perfect mileage by
using it in teaching.  Many fewer people have heard of George Boole than
have used appliances with 0 and 1 on their switches.  _Particularly_ since,
by this PEP, we're going to keep using False and True for arithmetic, with
them being equal to 0 and 1 respectively for all matters arithmetic, giving
them this dual identity and dual names won't help beginners at all.

Which leads me right onto disadvantages of the PEP -- because that's
the main one I see: extra conceptual load in teaching the language.  "0 can
be spelled either 0 or False" with subtly (but generally irrelevantly:-) 
different consequences... _please_ don't force me to have to teach that.
I'm not a staunch paladin of Python's simplicity when that simplicity has
substantial costs -- I fully agree that practicality beats purity and all 
that -- but simplicity IS still an important Pythonic value.  Diminishing 
that simplicity by a little bit for no substantial benefit would seem a 
wrong decision to me -- and I've already argued, I hope, that the benefits
we expect from this change _are_ insubstantial... "pretty small" at best.
When in doubt, I'd pump for choosing simplicity and stability.  Not that I'm
really in doubt here (I dislike this bool introduction:-) but even if I 
were I'd use the simplicity criterion to decide:-).

I've wasted too much of my life trying to explain to people why they must
NOT code "if some_condition==True:" or the equivalent thereof in various
languages, but just "if some_condition:".  The "== True" appendix, useless
at best, often damaging, seems one of the most natural errors by newbies.
In Python I was blessedly spared that waste of teaching time.  If the PEP
is approved, I'll be back to wasting that time and effort (perhaps more,
given how seductive "is True" will appear to many newbies...!).

Breakage will be modest but will be there, in doctest usage and other
kinds of tests -- wherever some output is compared with some expected
output, I'll get false errors over and over again depending on what choice
is made for str and repr.  Unless str and repr are both '0' and '1'.  If 
they are, the point of the PEP becomes murkier.  And what happens to
other forms of output, such as marshaling and pickling -- does the bool
or int nature of a 0 or 1 get preserved or lost in a round trip?  I can see
disadvantages and breakage possibilities in either case.

There are other minor issues too, such as the breakage of (already
rather broken) legacy code that typetests with type(a)==type(0), or
type(a)==types.IntType; and the likely bickering on c.l.p about being
able to do True*6 &c, which I fear will be more substantial than the
very occasional bleating about "not having booleans"; but these _are_
admittedly minor.  Still, the overall balance seems negative to me.

It won't break Python if you pass this PEP.  It will, however, make it
a little bit worse, IMHO -- and just when I was hoping we'd see one
release without substantial changes to the core language, which I
think would have been a nice public-relations plus with those circles
that are worried about language stability.  Oh well.


If this change does go ahead, here's my opinion on the sub-issues:


>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".)

Breakage would be a little bit less with '1'.  Yes, it does look
strange -- but so does the whole PEP, to me:-).  If you hadn't
posted it weeks ago to python-dev one might be inclined to suspect 
an April Fool's joke...:-).

>     3) Should the constants be called 'True' and 'False'
>        (corresponding to None) or 'true' and 'false' (as in C++, Java
>        and C99).

Uppercasing seems preferable here.

>     Most other details of the proposal are pretty much forced by the
>     backwards compatibility requirement; e.g. True == 1 and
>     True+1 == 2 must hold, else reams of existing code would break.

Yes.  We can't have "real" booleans (a la Pascal/Java) -- thanks be.

>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.

I agree with you.  "real" booleans (on which one cannot do arithmetic)
basically just force insertion of a lot of extra int() calls and do nothing
but reduce readability thereby.

>     5) Should operator.truth(x) return an int or a bool.  Tim Peters

An int.  There are other ways to get a bool: why make operator.truth
identical with the proposed bool builtin?  Better to have a direct way
to get an int 0 or 1 and one to get a bool, than two ways to get a
bool and nothing but multiple ops (int(bool(blah))) to get an int.

>        believes it should return an int because it's been documented
>        as such.  I think it should return a bool; most other standard
>        predicates (e.g. issubtype()) have also been documented as
>        returning 0 or 1, and it's obvious that we want to change those
>        to return a bool.

If this PEP passes, docs and functions throughout the library will
indeed have to be changed (not a point in its favor:-).  But there
are substantial reasons to keep operator.truth unchanged (give a
direct way to get 0/1) and none that I can see to have it become a
useless duplicate of the new bool builtin.


Alex



From goodger at users.sourceforge.net  Tue Apr  2 22:11:22 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Wed, 03 Apr 2002 03:11:22 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <200204030123.BAA15136@crown.off.ekorp.com> <3CAA60A5.F88697D5@engcorp.com> <mailman.1017799677.14060.python-list@python.org>
Message-ID: <B8CFDD11.20D30%goodger@users.sourceforge.net>

Richard Jones wrote:
> On Wed, 3 Apr 2002 11:53, Peter Hansen wrote:
>> I would suggest an even shorter version of the quickref, focused on
>> only those items needed in 80% of embedded documentation.  It does
>> _look_ complicated at first.
> 
> I agree - I think the quickref definitely needs:
> 
> 1. reorganising to filter the most common, simple markup to the top, and
> 2. much higher status/metioning on the ReST website.

You're both talking about a "Gentle Introduction" document, I think.
There's certainly room for one -- any volunteers?

I like the quickref as it is though, a true "Quick Reference".  It should be
reasonably complete, for quick lookups (that's why I added the table of
contents).  It's not really suitable as a primer though, and that document
doesn't exist yet.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From bokr at oz.net  Wed Apr 24 21:34:08 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Apr 2002 01:34:08 GMT
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com> <slrnacc59a.tg8.jon+usenet@snowy.squish.net> <MCxx8.45058$0x3.38185@atlpnn01.usenetserver.com> <aa7c52$ehu$0@216.39.172.122> <slrnaceetc.3k8.jon+usenet@snowy.squish.net>
Message-ID: <aa7meg$84u$0@216.39.172.122>

On Wed, 24 Apr 2002 23:11:45 -0000, Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote:

>In article <aa7c52$ehu$0 at 216.39.172.122>, Bengt Richter wrote:
>> So is cc.execute(...) acting like
>> >>> def dummyccx(fmt, params):
>>  ...     for tup in params:
>>  ...         print fmt % tuple(map(repr,tup))
>>  ...
>> >>> dummyccx(
>>  ... "INSERT INTO tmp (k, s) VALUES (%s, %s)",
>>  ... [(1, "one"), (2, 'two'), (3, 'three')]
>>  ... )
>>  INSERT INTO tmp (k, s) VALUES (1, 'one')
>>  INSERT INTO tmp (k, s) VALUES (2, 'two')
>>  INSERT INTO tmp (k, s) VALUES (3, 'three')
>
>Sort of that one. Except it isn't 'repr', it's a function which
>examines the type of the parameter and converts it into appropriate
>MySQL representation (i.e. for strings put "'" at the beginning and
>end and quote various characters with a "\", for numbers convert them
>to strings, etc etc).
>
No surprise that it wasn't repr, though with subclassed types instantiated
just for passing as SQL parameters, I guess you could make it come close now.
I just wanted to make it clear that leaving "'" out of the format string
was not directly for the sake of SQL itself, but more for some intermediating
software's purposes.

>Oh, except you shouldn't be relying on the 'for tup in params', you
>should imagine it is doing:
>
>  def dummyccx(fmt, params):
>    print fmt % tuple(map(sql_quote, params))
>
I guess I could RTFM, but how do you do a single INSERT with many value tuples?
Does internal optimizing take care of that when it uses the list of tuples or
other sequence?

>Except it's more complicated than this because you might be using a
>mapping not a sequence for the parameters. See
>http://www.python.org/topics/database/DatabaseAPI-2.0.html
>for details.
Next time I want to do something with a DB, I'll have to try it with Python.
(I used the mySQL C API before).

Regards,
Bengt Richter


From Bill.Scherer at VerizonWireless.com  Wed Apr  3 11:56:41 2002
From: Bill.Scherer at VerizonWireless.com (Scherer, Bill)
Date: Wed, 3 Apr 2002 11:56:41 -0500 (EST)
Subject: Java to Python translation
In-Reply-To: <Pine.LNX.4.33.0204030923150.5663-100000@ns0.tbye.com>
Message-ID: <Pine.LNX.4.33.0204031154280.28311-100000@bagend>

On Wed, 3 Apr 2002 brueckd at tbye.com wrote:


> (setting up Oracle on
> Linux so I can use the Python DB API is a mess). 

How so? We've been using Python with DCOracle on Linux boxes to 
talk to Oracle on Sun and Linux boxes for years now, and I've 
got no complaints...


-- 
Bill.Scherer at Verizon Wireless
RHCE 807101044903581





From Chris.Barker at noaa.gov  Tue Apr  2 11:29:24 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 02 Apr 2002 08:29:24 -0800
Subject: [Newbie] embed c++ in Python
References: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at> <a8a62r$57u$1@peabody.colorado.edu> <3ca95375$0$3698$6e365a64@newsreader02.highway.telekom.at>
Message-ID: <3CA9DC4B.11D7031A@noaa.gov>

> > > I want to know how to write a function which receives some variables as
> > > argument changes them and sends the results back to Python.

This is a very straightforward oridinary extension problem. Look on the
web for "Extending Python C". The Python API is C so even if you use
C++, you are pretty much just writing C. The exception to this is the
Boost library, which looks very nice. It wraps the C Python objects in
C++. If you have one or two simple extensions to make, I'm not sure it's
worht learning Boost, but ifyou are doing alot it probably is. For
simple C extensions, start with the standard docs: the extending and
embedding tutorial, and then the C API reference.

Note: Python Integers are imutable, so it may not be possible to do
exactly what you specified. It is very easy to pass a mutable python
object (List, dictionary) to a C extension and change it there. You
could also easily write one that you would call this way:

(a,b) = Cfunction(a,b)

Look through teh docs, you'll figure it out.

Another source I used for learning to write extensions is the NumPy
docs, there are some examples there that are more involved than the
examples in the standard docs. Even if you have no use for NumPy arrays,
the examples might be usefull.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From wortelslaai at yahoo.com  Mon Apr 22 13:19:53 2002
From: wortelslaai at yahoo.com (Jaco Schieke)
Date: 22 Apr 2002 10:19:53 -0700
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu>
Message-ID: <b5ddd682.0204220919.32ce27d@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<a9mv1o$7jh$1 at peabody.colorado.edu>...
> Jaco Schieke wrote:
> 
> > Being new to interactive python, I have the problem that rlcompleter
> > does not seem to be able to resolve the items in a list:
> > 
> > This works as expected:
> > ======================================
> >>>> import readline, atexit, rlcompleter
> >>>> readline.parse_and_bind('tab: complete')
> >>>> a=[[],"",1]
> >>>> a.<tab>
> > a.append   a.count    a.extend   a.index    a.insert   a.pop
> > a.remove   a.reverse  a.sort
> > 
> > 
> > Now is this also supposed to work?
> > ==================================
> >>>> a[0].<tab>
> > .   ..         <---- this is the output I receive
> >                      instead of the properties of an []
> > 
> > 
> > 'cause this works:
> > ==================================
> >>>> b=a[0]
> >>>> b.<tab>
> > b.append   b.count    b.extend   b.index    b.insert   b.pop
> > b.remove   b.reverse  b.sort
> 
> I was going to ask the same thing as I'm inclined to think this is a bug: 
> readline has '[' and ']' as part of its internal list of word delimiters, so 
> when you do 
> a[0].<tab>
> it breaks the word at ']' and sees a bare '.', hence the completions you get. 
> So far so good.
> 
> BUT: if you change the list of delimiters (with readline.set_delimiters()) to 
> remove '[' and ']', the problem persists. That's the part I think to be a bug 
> (I could be wrong, but I'd like to understand why if that's the case). At 
> this point it shouldn't break at ']' anymore, and the completion should be 
> done on the full 'a[0]'. I put print statements in the rlcompleter code to 
> see this and indeed, readline _is_ breaking on ']' even after removing the 
> character from the list of delimiters.
> 
> Unfortunately the completion code is in C and I don't have the sources around, 
> so I haven't had the chance to take a look at this.
> 
> I'd love to hear from someone who knows the code as to this behavior really is 
> a bug or not before shooting off to SF and looking like an idiot :)
> 
> Cheers,
> 
> f.


Looks like the regexp in rlcompleter that searches the string does not
work correctly.  After changing the line

> m = re.match(r"(\w+(\.\w+)*)\.(\w*)", text)

to 

> m = re.match(r"([\w\[\]]+(\.[\w\[\]]+)*)\.(\w*)", text)

it matched on the brackets too and could get to the objects.  I also
had to change the delimiters to exclude [].


From max at alcyone.com  Fri Apr 12 18:43:02 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 12 Apr 2002 15:43:02 -0700
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca>
Message-ID: <3CB762F6.18578A15@alcyone.com>

Brad Bollenbach wrote:

> To "put it into the public domain" the only thing that
> is implicit is the fact that you hold the copyright.

No, when you release something into the public domain you explicitly
waive your copyright.  "Public domain" has a very specific legal meaning
in licenses; it doesn't just mean a public release.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From juenglin at informatik.uni-freiburg.de  Wed Apr 24 12:17:34 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 24 Apr 2002 18:17:34 +0200
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com> <efb63dac.0204221151.6d866042@posting.google.com> <MPG.172e0ed9611cc5119896c1@news.nwlink.com>
Message-ID: <xduznzt3wbl.fsf@popeline.informatik.uni-freiburg.de>

Jeff Shannon <jeff at ccvcorp.com> writes:

> 
> > [Cut out from PEP 203]
> > 
> >       They implement the same operator as their normal binary form,
> >       except that the operation is done `in-place' when the left-hand
> >       side object supports it...
> 
> The key point here being "when the left-hand side object supports 
> it" -- which clearly implies that immutable objects will not be 
> modified in place, since they do not support modification.
> 
> While there might be some theoretical argument for considering 
> augmented assignment on immutable types to be an error, keep in 
> mind that practicality beats purity.  Making augmented assignment 
> on integers into an error would, in practice, be quite silly.

The practical argument for purity is, that consistent semantics of 
operations is easier getting used to. Even if the type supporting
 __iadd__ is mutable, an in-place operation is not demanded by the
language spec. It reads there

"... These methods should attempt to do the operation in-place
(modifying self) and return the result (which could be, but 
does not have to be, self)..."

I've realized that Python is much less strict than other 
languages and builds more often on convention rather than 
specification. It's probably just a matter of taste that I 
don't appreciate Python's freedom in any case (here I don't).

Thanks for your comments,
Ralf

PS: I also would miss the idiom "i+=1" for integer increments...


From hoel at germanlloyd.org  Thu Apr 25 04:08:39 2002
From: hoel at germanlloyd.org (Berthold =?iso-8859-15?q?H=F6llmann?=)
Date: 25 Apr 2002 10:08:39 +0200
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org> <mailman.1019682973.29139.python-list@python.org> <871yd4odap.fsf@pvv.org> <MPG.1730fb06f5bab3e99896c8@news.nwlink.com>
Message-ID: <sepu0okxo8.fsf@pc970145.GermanLloyd.de>

Jeff Shannon <jeff at ccvcorp.com> writes:

> In article <871yd4odap.fsf at pvv.org>, oyving+news at pvv.org says...
> > * Cliff Wells
> > |
> > | On 24 Apr 2002 22:57:20 +0200
> > | Oyvind Gronnesby wrote:
> > | 
> > | > * Cimarron Taylor
> > | > |
> > | > | Will this do what you want?
> > | > | 
> > | > | verboseGender = ('female','male')[gender=='m']
> > | > 
> > | > Wouldn't this break in a PEP 285 Python world?
> > | 
> > | No.  That was the whole point of subclassing bool from int.
> > 
> > But wouldn't that just imply that True/False would turn into 1/0 if coerced
> > into it by int()?
> 
> Yes, they will.  This is by design, to maintain backwards 
> compatibility.  The only effective difference between 1 and True, 
> and 0 and False, is how it's written, but they add, subtract, 
> multiply, and hash the same.

But there is a difference:

>python
Python 2.2.1 (#1, Apr 10 2002, 17:11:02) 
[GCC 3.0.4] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> a = 1
>>> b = 1
>>> a is b
1
>>> a = True
>>> a is b
0
>>> 

Greetings

Berthold
-- 
Dipl.-Ing. Berthold H?llmann   __   Address:
hoel at germanlloyd.org        G /  \ L Germanischer Lloyd
phone: +49-40-36149-7374    -+----+- Vorsetzen 32/35    P.O.Box 111606
fax  : +49-40-36149-7320      \__/   D-20459 Hamburg    D-20416 Hamburg


From noah at noah.org  Thu Apr 11 14:15:57 2002
From: noah at noah.org (Noah)
Date: Thu, 11 Apr 2002 11:15:57 -0700
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
In-Reply-To: <a94d75$hde$1@nntp6.u.washington.edu>
Message-ID: <PAENINBOOODFDEPOKPKCKEMPCKAA.noah@noah.org>

Damn... The only problem I see with this explaination is
that this would imply that popen is almost totally useless.
Why bother opening a pipe with both in and out streams
if can't actually read and write from both streams?
I could have sworn that I have seen interactive programs
such as telnet and ftp scripted from Python.

I looked at the pty stuff, but I was going to implement that
AFTER I got pipes working. Pipes seem to be the most portable
and intuitive way to solve this problem.

Damnit! If I can do it in Perl I'm sure I can make it work in Python!

Noah


-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Donn Cave
Sent: Thursday, April 11, 2002 9:22 AM
To: python-list at python.org
Subject: Re: Nonblocking IO on popen pipe (isn't nonblocking)


Quoth "Noah" <noah at noah.org>:
| I'm having headaches with nonblocking IO on a pipe.
| I opened a UNIX command through os.popen4.
| Then I set O_NONBLOCK on the output stream of the pipe.
| Finally I did a select() on the output pipe to wait for data.
| If select returns I would try to read the data.
| The problem is that select() is blocking. Select should block 
| until some data is available on the pipe, so I assume that
| select is never even seeing any data on the pipe.
| I added a timeout to select() and then tried to read the
| pipe, but it was empty (read() returned '').
|
| One program that I am attempting to talk to via a pipe is ftp.
| Any interactive program I open will block. 
| If I try open a program that terminates like "ls" or "uname -a" 
| then my script works as expected.
|
| What am I doing wrong that is causing interactive programs to block?

It isn't you.  I'd argue that your O_NONBLOCK is unnecessary, as long
as you're using select anyway, but it won't make any difference if
the command on the other end doesn't send the data.  It doesn't send
the data because its output goes through C library stdio buffers, which
treat a pipe like a disk file and buffer up whole blocks.  There could
be other problems, like some of the data is actually going to stderr,
but when you say the only programs that work are the ones that terminate,
that's a hallmark of this problem (buffers are flushed on exit.)

The only solution is to get away from popen2 and pipes, and use a
pseudotty instead.  2.1 and up have a usable posix.openpty() on some
UNIX platforms, and there's an older pty.py module that may or may
not be useful.  Ptys are sort of tricky and cross-platform support is
poor, but they'll work better in a couple of ways for applications
that are intended for interactive use.

	Donn Cave, donn at u.washington.edu





From sholden at holdenweb.com  Fri Apr  5 16:22:23 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 5 Apr 2002 16:22:23 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <mailman.1017757086.9920.python-list@python.org>
Message-ID: <KBor8.8537$q63.3756@atlpnn01.usenetserver.com>

"Gustavo Niemeyer" <niemeyer at conectiva.com> wrote ...
> > Arguably, most of that code should have read
> >
> >     if ret:
> >         ...
> >
> > rather than
> >
> >     if ret == None:
> >         ...
> >
> > since the latter has the flavor of "if expr==True".
> >
> > So, is breaking bad code more excusable than breaking good code ;-)?
>
> It should be "if not ret", and most of the time this is a matter of
> taste (some times you just have to do this, since you want to check for
> "None", and not for 0). Nevertheless, this is a valid construct, and
> not uncommon (indeed, I have no problems with this). If you don't
> like this, don't use for yourself, but language design is much more than
> "I don't like this so everyone using it should burn".
>
I certainly don't remember making any suggestion that required programmer
combustibility, and I take your point (and agree with it) entirely.

*My* point, however badly I originally made it, was that writing

    if ret == None:

is *sometimes* like writing

    if (a == 3) == True:

assuming this new wart appears in Python. Clearly there are cases where an
explicit test for None (or [], or some other value treated as false) is
required. Equally clearly there are places where the explicit test can be
omitted, and where it can be omitted it usually should be.

I believe I already pointed out myself that it should be "if not ret". Makes
no essential difference to my argument (apart from being correct instead of
incorrect :-).

regards
 Steve






From thomas.heller at ion-tof.com  Fri Apr 26 15:47:13 2002
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Fri, 26 Apr 2002 21:47:13 +0200
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com>
Message-ID: <0d9001c1ed5b$29939830$e000a8c0@thomasnotebook>

> --------------------------------------
> I would like to start the SPA:
> 
>     Starship Python Activity
> 
> as a non-profit organisation.
> --------------------------------------
> 
> Here the proposed layout:
> SPA membership costs USD 25 by the year.
> This price includes Starship membership
> and a continous subscription to PyZine.
> 

Do anything you want with my money (within resonable
bounds), but don't make me a member of another
organization. Please ;-)

Thomas





From eric.brunel at pragmadev.com  Mon Apr  8 06:46:39 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 8 Apr 2002 10:46:39 +0000
Subject: stdout redirect : Howto with popen2.popen3
References: <3CAF8AC1.90704077@club-internet.fr>
Message-ID: <a8rl4s$15sk$1@norfair.nerim.net>

Fabien Henon wrote:
[snip]
>   2. How can I tell when the raytrace is over ? The poll() gets out of
> the loop.
> What is command to listen to a program to tell if it's running or not?
[snip]
> while run.poll() <>-1:
>      print "attente"   # // I never got to here and goes straight to
> fd=open.....

Errr, maybe you just inverted the condition? The library reference says for 
the poll method on the Popen3 class: "poll() Returns -1 if child process 
hasn't completed yet, or its return code otherwise.". So, if I understand 
correctly, that means that while the child is running, poll() returns -1, 
and when the child has exited, poll() returns its exit code. So your loop 
should be:

while run.poll() == -1:
  # etc...

Didn't try it though. HTH however...
 - eric -



From tim.one at comcast.net  Thu Apr 25 18:50:40 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 25 Apr 2002 18:50:40 -0400
Subject: Dynamically changing an object's class - still to be allowed?
In-Reply-To: <3CC7ED36.9B1E57A0@st-andrews.ac.uk>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAENODAAA.tim.one@comcast.net>

[Hamish Lawson]
> I came across a debate that ran on the list while Python 2.2 was being
> developed that suggested (if I understood correctly) that Guido was
> planning to limit the ability to dynamically change an object's class by
> assigning to its __class__ attribute. It was indicated that these
> limitations might be implemented for Python 2.2, but they don't appear
> to have been; might they yet?

Well, not in 2.2 <wink>.  Some 2.2 prereleases disallowed changing __class__
at all, but Guido backed down from this after enough people showed arguably
valid use cases, and 2.2 tries to allow it *when it can*.  That's the plan
for 2.3 & onward too.

Use of some features of new-style classes can make it impossible, though.
All instances of old-style classes had exactly the same memory layout, but
instances of new-style class *may* have radically different memory layouts.
As an obvious example, "int" is a new-style class, and you're never going to
be able to change the __class__ of your own new-style class's instances to
"int":

>>> class C(object):
...    pass
...
>>> c = C()
>>> c.__class__ = int
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __class__ assignment: 'int' object layout differs from 'C'
>>>

Not that it's not complaining because 'int' is a builtin class, it's
complaining because the instance memory layouts don't jibe.

I wouldn't blame you if you thought this was a bug in 2.3 CVS <wink>:

>>> (3).__class__ = bool
>>> 3
True
>>>





From gusraan at terra.com.br  Wed Apr 24 19:05:28 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: Wed, 24 Apr 2002 20:05:28 -0300
Subject: Twisted ircd is written in python
References: <slrnac9ak3.fho.ak@ak.silmarill.org>
Message-ID: <aa7dn4$402$1@nsnmrro2-gest.nuria.telefonica-data.net>

didn't find it :(

"Andrei Kulakov" <ak at silmarill.org> wrote in message
news:slrnac9ak3.fho.ak at ak.silmarill.org...
> google for twisted matrix
>
>
> --
> Cymbaline: intelligent learning mp3 player - python, linux, console.
> get it at: cy.silmarill.org




From jeff at ccvcorp.com  Tue Apr  9 20:01:44 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Tue, 9 Apr 2002 17:01:44 -0700
Subject: Creating objects in thread created with subclass
References: <mailman.1018365202.7991.python-list@python.org>
Message-ID: <MPG.171d20a7416584ee9896a5@news.nwlink.com>

In article <mailman.1018365202.7991.python-list at python.org>, 
narnett at mccmedia.com says...

> [...] Here's my question -- in the "run" method of
> Aahz's example, if I create my retriever object in an ordinary way, then all
> of the threads will be sharing the same object, right?  And that most
> definitely will not work.  

No, that's not correct.  If you create an object in the Run 
method of your threading.Thread subclass, then you will create a 
new instance of that object for *each* instance of your subclass, 
so each thread will have a unique object created for it.  Since 
you're using a thread pool, then you'll be re-using that object 
each time that you re-use that thread, but ThreadA will have a 
different instance of it than ThreadB.

> So, I could create objects with 'eval' or 'exec'
> to generate unique names for each thread -- but is that the right way to do
> it?  

No, that is definitely *not* the right way to do it -- as I keep 
saying, exec/eval() are very rarely the right way to do it, 
whatever "it" may be.  ;)

> Or should I stop using objects there and directly use the functions
> from my classes.  I haven't looked at a whole lot of examples, but so far,
> the latter seems to be the way people do it.

You should be fine using your objects.  Just remember that, 
unless you explicitly create a new object when the thread is 
pulled from the thread pool, you'll have a one-to-one 
correspondence between your threads and your object.  (Presuming 
that you're creating only one such object in the Run() method...)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From hgs at dmu.ac.uk  Thu Apr 18 09:22:53 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 14:22:53 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <lkofghqjq4.fsf@pc150.maths.bris.ac.uk>
Message-ID: <Pine.GSO.4.40.0204181420250.24683-100000@neelix>

On Thu, 18 Apr 2002, Michael Hudson wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > Script started on Thu Apr 18 11:53:08 2002
> >         [...]
> > 160 tests OK.
> > 4 tests failed:
> >     test_largefile test_socket test_sundry test_urllib2
> > 23 tests skipped:
        [...]
> > *** Error code 1
> > make: Fatal error: Command failed for target `test'
        [...]
> > How concerned should I be about these?
>
> test_socket failing is bad.  The test_sundry & test_urllib2 failures
> would be connected to this, I imagine.

It is worse

Traceback (most recent call last):
  File "./Lib/test/test_socket.py", line 12, in ?
    import socket
  File "/home/hgs/Python-2.2.1/Lib/socket.py", line 41, in ?
    from _socket import *
ImportError: No module named _socket
>
> More is hard to say (I must admit my head is starting to spin like a
> top with all these reports).
>
> Cheers,
> M.
>





From ssweeney at idiosync.net  Wed Apr  3 19:24:43 2002
From: ssweeney at idiosync.net (Scott Sweeney)
Date: Wed, 3 Apr 2002 16:24:43 -0800
Subject: ANN: SLiP and SLIDE - a quick XML shorthand syntax and tool for editing
Message-ID: <PLEFIJEECKLKCPNJFOFAOEOECBAB.ssweeney@idiosync.net>

Announcing...

SLiP 0.1: a "Sorta Like Python" shorthand syntax for XML
SLIDE 0.1: a "Sorta Like an IDE" lightweight windows exe for SLiP editing
SLiPpy 0.1: a SLiP package for Python, including a SLiP->XML library
(sliplib.py) and source for SLIDE (slide.py)

	http://www.scottsweeney.com/projects/slip


[Description]

	SLiP is a quick, alternative shorthand syntax for creating and editing XML
data by hand and if you know Python, it should also be familiar.  Great for
taking notes.


[Benefits]

	* Less typing - provides a quicker and easier means for creating or editing
XML

	* Easier to read - creates a pleasant layout for the data with indentation
and spacing

	* Benefits of XML - allow for hierarchical data, unlike tabbed or other
character delimited data files

	* Cheap - provides a free XML editing alternative

	* Editor-agnostic and Keyboard-oriented - provide a quicker, more
hacker-friendly way to create XML over point-and-click alternatives

	* Works in existing editors - most Python editors (IDLE, Scintilla, etc)
should accept and make working with this syntax easy




[Example]

This portion of SLiP:

	my:addresses(xmlns:my="http://example.org/my"):

		#some random comment

		address(type="home"):

			street(): "123 Sesame Street"
			city(): "Wonderland"
			state(): "CA"
			zipCode(): "90012"
			comment():

				"""
				Please leave packages with Grouch
		 		in garbage can next door.

				"""


.. would convert to this XML:

	<my:addresses xmlns:my="http://example.org/my">
		<!-- random comment -->
		<address type="home">
			<street>123 Sesame Street</street>
			<city>Wonderland</city>
			<state>CA</state>
			<zipCode>90012</zipCode>
			<comment>

			Please leave packages with Grouch
			in garbage can next door.

			</comment>
		</address>
	</my:addresses>




I've created some free (MIT license) tools including a Python library for
translating between XML and SLiP (SLiPpy), a lightweight editor wrapper
(SLIDE), and some brief documentation - all available at the website,
http://www.scottsweeney.com/projects/slip

This is my first release of these tools so they are a little rough and they
need more input, so let me know what you think.

Thanks,
Scott Sweeney


[Contact]
  ssweeney at idiosync.net


<P><A HREF="http://www.scottsweeney.com/projects/slip">SLiP</a> - a
"Sorta Like Python" shorthand for XML (3-Apr-2002).






From mlh at vier.idi.ntnu.no  Tue Apr  9 07:23:56 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Tue, 9 Apr 2002 11:23:56 +0000 (UTC)
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch>
Message-ID: <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>

In article <3CB2CB16.8080103 at geneva-link.ch>, Boris Borcic wrote:
[snip]
>(b) Laura made a lengthy point, overlooking the improvable while clear 
>to the language designer's POV, that it very much boiled down to : "I've 
>stacked my didactics of truth values in Python on the name of None, and 
>the PEP makes me feel I've been telling lies".

That's not how I saw it. The main point I extracted was: Until now,
the Pythonic "truth values" have been "something" and "nothing" (i.e.
the empty list, zero, None, etc.) With the bool type this is still
true, but we get a type which will give newbies the impression that it
isn't.

>python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")

Hm. What programming language is this supposed to be?

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From gh_pythonlist at gmx.de  Wed Apr  3 05:34:24 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 3 Apr 2002 12:34:24 +0200
Subject: verifying ssl certificates
In-Reply-To: <7xu1qt4h3h.fsf@ruckus.brouhaha.com>
References: <3CA88D15.140053DC@ecn.purdue.edu> <7x7knrdv9k.fsf@ruckus.brouhaha.com> <3CA9C2E0.EA8A2B15@ecn.purdue.edu> <7xu1qt4h3h.fsf@ruckus.brouhaha.com>
Message-ID: <20020403103424.GC463@lilith.hqd-internal>

* Paul Rubin <phr-n2002a at nightsong.com> [2002-04-02 19:05 -0800]:
> "Christopher N. Deckard" <cnd at ecn.purdue.edu> writes:
> > In other words:
> > 
> > import socket
> > 
> > cert = open('my_cert')
> > s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> > s.connect(('host', 'port'))
> > ss = ssl(s, cert, None)
> > 
> > That last line being the key.  Does that do the certificate
> > validation?

I'd recommend to not use the standard Python SSL for security-critical
applications at this point. Lots of bugs have been fixed, but it's not very
good by design.

> I'm not familiar with how the socket module does ssl, but if ssl() is
> supposed to do an SSL session negotiation, then yes, it should check
> the cert.

It should, however it doesn't check the certificates at all (Modules/_ssl.c):

        SSL_CTX_set_verify(self->ctx,
                                   SSL_VERIFY_NONE, NULL); /* set verify lvl */

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 15.4 ?C      Wind: 5.9 m/s



From sholden at holdenweb.com  Tue Apr 23 08:21:30 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 08:21:30 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com>
Message-ID: <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com>

"Fernando Rodr?guez" <frr at easyjob.net> wrote ...
>
> I'm using CompatMysqldb from http://sourceforge.net/projects/mysql-python
with
> Python 2.2
>
> Whenever I try to insert some data into a mysql db, I get this error:
>
> >>> curs.execute("insert into customer_info (email, url, referrer,
first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com',
'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
> Traceback (most recent call last):
>   File "<pyshell#26>", line 1, in ?
>     curs.execute("insert into customer_info (email, url, referrer,
> first_visit) values ( '%s', '%s', '%s', '%s')", [('billg at microsoft.com',
> 'www.easyjob.net', 'www.google.com', '2001-03-07 12:32:00')])
>   File "D:\ARCHIV~1\Python22\Lib\site-packages\CompatMysqldb.py", line
227, in
> execute
>     af =af+self.__res.affectedrows()
> AttributeError: 'NoneType' object has no attribute 'affectedrows'
> >>>
>
> The data does get inserted... What am I doing wrong? O:-)
>

I don't think you need the string quotes around the parameter markers - the
module determines the required type, and in many cases will perform
conversions for you as necessary. Since you are using a list of tuples, you
should really be calling executemany() rather than execute, though most DB
API modules will accept lists of tuples. Try this:

curs.execute(
    """insert into customer_info (email, url, referrer, first_visit)
    values ( %s, %s, %s, %s)""",
    ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07
12:32:00')
)

or, equivalently (but redundantly, unless MySQdb requires executemany()):

curs.execute(
    """insert into customer_info (email, url, referrer, first_visit)
    values ( %s, %s, %s, %s)""",
    [('billg at microsoft.com', 'www.easyjob.net', 'www.google.com',
'2001-03-07 12:32:00')]
)


It may or may not make a difference. What's the difference between
CompatMySQLdb and MySQLdb?

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From donn at u.washington.edu  Tue Apr 23 12:24:22 2002
From: donn at u.washington.edu (Donn Cave)
Date: 23 Apr 2002 16:24:22 GMT
Subject: stdin - seek() issue on Linux.
References: <WMTw8.46875$uR5.106037@newsfeeds.bigpond.com> <pEUw8.85206$XV5.7705262@bin4.nnrp.aus1.giganews.com> <aa1bqb$lek$1@nntp6.u.washington.edu> <284b44c2.0204221908.1280cfc3@posting.google.com>
Message-ID: <aa41rm$34ri$1@nntp6.u.washington.edu>

Quoth harisri at bigpond.com (Srihari Vijayaraghavan):
(quoting me)
|> Of course if the real program uses some filter, instead of "cat",
|> then there's no solution - you can't seek in a pipe, so the answer
|> has to be "don't do that".
|> 

| My mistake, I haven't given the complete context. The above mentioned
| method doesn't throw exceptions in Windows using the same Python
| version (neither does it print anything, which is the ultimate problem
| I would like to be solved).

That's too bad.  Something that doesn't/can't work should raise an
exception, in my opinion.  I don't know how Microsoft's stdin works,
but it must either fail all the time, or part of the time.  If they
allow it to pass without error because it occasionally may work,
that's appalling.

| Are there any better method of redirecting the stdout of a program to
| stdin of my program in both windows and linux? (specifically I should
| be able to seek() on it)

Write to a disk file, then run the second program.  Or if you can
store enough data in your program to satisfy your requirement,
then you can use a pipe.

	Donn Cave, donn at u.washington.edu


From jmarshal at mathworks.com  Wed Apr 17 17:10:47 2002
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 17 Apr 2002 21:10:47 GMT
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com> <mailman.1019068235.12319.python-list@python.org> <a9kjh1$qc5$1@slb1.atl.mindspring.net> <mailman.1019073733.27071.python-list@python.org> <Filv8.32338$vF6.998647@news2.tin.it>
Message-ID: <a9kocn$6mu$1@news.mathworks.com>

Alex Martelli <aleax at aleax.it> wrote:
> Ian Bicking wrote:
>         ...
>> But that's weird...
>> 
>> q1 += q3
>> is not equivalent to:
>> q1 = q1 + q3

> If it were, there would be little reason for it to exist -- saving
> a couple of keystrokes would be a silly reason to introduce a
> bunch of new operators.

To me there's a more compelling reason for the existence of "+=": you don't need
to use a temporary variable to prevent the lvalue from being computed twice.
Eg:

  x[f()] += 1

vs

  x[f()] = x[f()] + 1

or

  idx = f()
  x[idx] = x[idx] + 1

Granted, since in Python assignment is a statement and not an expression, this
is less of a win than the existence of += in C, where it might be inconvenient
to introduce a temporary.


From grey at despair.dmiyu.org  Tue Apr 16 19:18:48 2002
From: grey at despair.dmiyu.org (Steve Lamb)
Date: Tue, 16 Apr 2002 23:18:48 -0000
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <slrnabmtk0.a8q.ak@ak.silmarill.org> <slrnabp92n.ipr.philh@comuno.freeserve.co.uk>
Message-ID: <slrnabpcc0.88d.grey@teleute.dmiyu.org>

On Tue, 16 Apr 2002 23:23:19 +0100, phil hunt <philh at comuno.freeserve.co.uk>
wrote:
> I'm sure what I want, without needing to explicitly say it.

    Fortunately we are not borg.  We are not in your head.  Unfortunately
we're in your code.  We're not sure what you meant.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
    To email: Don't despair!   |  -- Lenny Nero, Strange Days
-------------------------------+---------------------------------------------


From Oschler at earthlink.net  Tue Apr  9 12:15:14 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 9 Apr 2002 12:15:14 -0400
Subject: zope: Zahlen-Index
References: <3cb28f90$1@netnews.web.de>
Message-ID: <zkEs8.33957$VM5.15749060@e3500-atl1.usenetserver.com>

<abele>; STephan <abele at wlb-stuttgart.de> wrote in message
news:3cb28f90$1 at netnews.web.de...
> Liebe Experten,
>
>
> ich weis nun nicht, ob ich hier in der Liste richtig bin.
> Trotzdem m?chte ich folgendes Zope-Problem posten:
>

Stephan,

You'll probably have much better luck with Zope questions on one of the Zope
mailing lists.

http://www.zope.org/Resources/MailingLists

German Zope mailing lists:

: zope-de at onelist.com

To subscribe:

 mailto:zope-de-subscribe at onelist.com

> Ich habe einen ZCatalog erstellt und habe die relevanten Objekte
> katalogisiert

Regarding the problems you are having with indexes and finding objects in
the database, an excellent book is "The Zope Construction kit".  Check out
the ZCatalog example.  Unfortunately it is only in English:

http://www.amazon.de/exec/obidos/ASIN/0672321335/qid=1018367360/sr=2-4/ref=s
r_sp_bow/302-1434818-7754414

German Zope book:

http://www.beehive.de/news/dpunktannounce_D.html

There were a couple of books in German too though, don't know how good they
are.  Good luck.  Sorry I couldn't be of more help.  I am a Zope newbie, and
I don't understand German.

thx





From jcmason at student.math.uwaterloo.ca  Tue Apr 16 16:43:05 2002
From: jcmason at student.math.uwaterloo.ca (Joe Mason)
Date: 16 Apr 2002 20:43:05 GMT
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com> <ku4ricgeve.fsf@lasipalatsi.fi>
Message-ID: <a9i2cp$52m$1@watserv3.uwaterloo.ca>

In article <ku4ricgeve.fsf at lasipalatsi.fi>,
Erno Kuusela  <erno-news at erno.iki.fi> wrote:
>In article <87d6x09trz.fsf at mathdogs.com>, Mike Coleman
><mkc+dated+1021521407.f909ec at mathdogs.com> writes:
>
>| Does anyone know why strings (i.e., those of length >1) don't share
>| their data in Python?  Since their immutable, it seems like this
>| would be the obvious thing to do.
>
>in addition to reasons presented i other replies, this would
>also result in much less obvious memory usage. saving a short
>slice of a huge string would hang on to the huge string. java
>does this, and i've heard people make less than flattering
>comments about this feature.

I suppose it would be possible to implement copy-on-collect as well as
copy-on-write (wow, how's that for a self-contradiction?)  When the GC
notices that the only reference to the string is actually a substring, it
makes a copy and throws out the original.

But, yeah, it'd be a little complicated.

Joe 


From paul at boddie.net  Tue Apr  2 03:47:14 2002
From: paul at boddie.net (Paul Boddie)
Date: 2 Apr 2002 00:47:14 -0800
Subject: Python embedded like PHP
References: <mailman.1015988122.18668.python-list@python.org> <mailman.1016619871.27287.python-list@python.org> <7xk7s7mpiy.fsf@ruckus.brouhaha.com> <mailman.1016626151.9950.python-list@python.org> <slrna9haud.e6b.akuchlin@ute.mems-exchange.org> <slrnaa3bi8.lam.jon+usenet@snowy.squish.net> <23891c90.0203290655.25bbcab9@posting.google.com> <slrnaacob1.bu2.jon+usenet@snowy.squish.net> <23891c90.0203310820.60696fb1@posting.google.com> <slrnaahnai.i1t.jon+usenet@snowy.squish.net>
Message-ID: <23891c90.0204020047.4d4ae046@posting.google.com>

Jon Ribbens <jon+usenet at unequivocal.co.uk> wrote in message news:<slrnaahnai.i1t.jon+usenet at snowy.squish.net>...
> 
> Ah, I see. I must admit I have never seen a web application where
> the benefits this would provide would outweigh the extra complexities
> it would entail.

I think it depends a fair amount on the application. I've worked with
"form-heavy" applications myself, and it's not exactly a secret that
database-driven applications often have their presentation determined
by the underlying database schema.

> From an HTML point of view, the only two input attributes which could be
> magically generated are 'size' and 'maxlength'. 'size' is more dependent on
> the form design than the input requirements, and hence it is not appropriate
> to base this value on the input type, and the benefits of automagically
> setting maxlength would seem to be extremely minor.

Yes, I probably got confused between "length" and "size" - I always
need to refer to the HTML specification in that case. Still, I
wouldn't underestimate the benefit of making sure that even such
simple properties are synchronised at all levels of an application.

> From a coding point of view, I would think the benefits are fairly
> small too. In most applications I can think of, input data is usually
> being looked up from rather than entered into a database. For example,
> if the user enters a username the question is not normally "is this a
> valid username?", but "is this an existing username?", so no input
> validation is required. I guess the common exception is email
> addresses.

Some common validation cases are numbers, dates and monetary amounts,
and I've also seen systems which validated measurements. To an extent,
one can break down complicated fields into collections of simpler
ones, although some people dislike such interfaces. Nevertheless,
being able to declaratively associate form fields with the means of
validation, without the explicit need to code it, could well be
beneficial, in my opinion.

> The nearest jonpy gets to this sort of thing is MultiForm, a class
> for creating multi-page input forms. It provides a 'check' method
> which you override to validate the input data and produce appropriate
> error messages.

That's another area where automation gives some notable advantages,
too.

Paul


From tanzer at swing.co.at  Wed Apr  3 01:00:54 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Wed, 03 Apr 2002 08:00:54 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "02 Apr 2002 02:22:12 GMT."
             <a8b4kk$je2$0@216.39.172.122>
Message-ID: <E16sdpG-00052o-00@tswings.swing.cluster>

bokr at oz.net (Bengt Richter) wrote:

> >If you called it "binary" instead of "bool" would it help you think of
> >it as being more like a number?
> >
> No. Because numbers are ordered, and bools are not. bools are selected
> elements from an unordered set of two possibilities.

FWIW, Ada and Pascal define booleans as ordered with False < True. In
both languages, you have to use `ord(x)` to get the integer equivalent
of a boolean value.

Both languages' booleans suck in comparison to Python's concept of
truth.

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




From phr-n2002a at nightsong.com  Tue Apr 16 15:20:24 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 16 Apr 2002 12:20:24 -0700
Subject: Python webmail packages?
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com> <7876a8ea.0204151614.138853be@posting.google.com> <mailman.1018918474.3622.python-list@python.org> <7xsn5wjk85.fsf@ruckus.brouhaha.com> <mailman.1018941693.2296.python-list@python.org>
Message-ID: <7xsn5ve9hj.fsf@ruckus.brouhaha.com>

Ian Bicking <ianb at colorstudy.com> writes:
> > Hmm, how much code is involved?  I didn't want to deal with installing
> > a huge package like Zope.
> 
> Installation isn't hard, like I said it's CGI.  It happens to reuse some
> code from Zope, but all that code is contained in the package itself.

Thanks.  Any impressions of how it compares to squirrelmail?  It
sounds like it isn't being actively maintained, which is not so
attractive.  Squirrelmail on the other hand has bugs that I can't seem
to convince the developers are really bugs.  I can patch my local copy
with fixes, but I'd rather do that with something written in Python.


From woonshan at time.net.my  Fri Apr 19 08:04:05 2002
From: woonshan at time.net.my (Woon Shan Chin)
Date: Fri, 19 Apr 2002 20:04:05 +0800
Subject: Python scripting to dynamically generate 3D human models
Message-ID: <2a82e2fbdf.2fbdf2a82e@time.net.my>

Hi,

I heard about 'Python scripting' from a member experts-exchange.com; an 
IT knowledge sharing community. I'm working on a virtual fitting room 
program, that needs to dynamically generate 3D human models during run-
time (based on users' input of their body measurements for eg.). So, I 
heard from a member in that community that I could import Poser's 3D 
human models into my program and use Python script to read users' input 
of their measurements at runtime and make changes in the parameters of 
the basic model such as wasit, height, etc., so as to regenerate the 3D 
human model that matches the users' input.

Eventually, Python scripting is something new to me and I would like to 
know how I can use Python script to perform the functionality mentioned 
above. For instance, the program I am developing is a MFC application 
written in Visual C++. How could I link the Python scripts to my 
program then?

Hope you could provide me some pointers as to how I can perform those 
functionality in my program.

Thanks a lot! I need your advice urgently! Your help is appreciated!

-sweetgirl-






From altis at semi-retired.com  Fri Apr 26 23:34:05 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 26 Apr 2002 20:34:05 -0700
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <zZoy8.110$EE5.139514@news.uswest.net>

PythonCard http://pythoncard.sourceforge.net/ contains a simple textEditor
sample, which is basically a Windows Notepad clone. There is a screenshot of
it at http://pythoncard.sourceforge.net/samples2.html

ka

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:aacdeg$7g$1 at nets3.rz.RWTH-Aachen.DE...
> Is there a simple, open source text-editor project available in Python?
>
>




From logiplexsoftware at earthlink.net  Wed Apr  3 15:20:45 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 3 Apr 2002 12:20:45 -0800
Subject: duelling pythons
In-Reply-To: <a8ed9i$rnodf$1@ID-41653.news.dfncis.de>
References: <1eed326b.0203312131.2ef13b62@posting.google.com>
	<mailman.1017695170.25885.python-list@python.org>
	<m3d6xibyfn.fsf@nbecker.fred.net>
	<mailman.1017770346.22212.python-list@python.org>
	<a8ed9i$rnodf$1@ID-41653.news.dfncis.de>
Message-ID: <20020403122045.48b5bd87.logiplexsoftware@earthlink.net>

On 3 Apr 2002 08:08:18 GMT
Tobias Klausmann wrote:
> In a related matter: how common is /usr/bin/env ?
> I really like that tool to run python wherever it is: 
> 
> #!/usr/bin/env python
> 
> As it is part of sh-utils, I guess it's pretty widely used.
> 
> Anybody know of specific environments where it is not commonly
> available?

Well, it exists on Linux and SCO Unix (which implies that it's fairly
common since SCO is next to useless [right below useless, actually]). 
However, SCO puts it in /bin, so it unless SCO is unusual in this respect
(wouldn't be a surprise - I'd actually expect it to be found in /etc or
something given their wonderful filesystem layout) it isn't much better
than specifying the path for Python directly.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From noah at noah.org  Tue Apr 16 12:47:24 2002
From: noah at noah.org (Noah)
Date: Tue, 16 Apr 2002 09:47:24 -0700
Subject: Backing up mail in Outlook Express using Python
In-Reply-To: <JCUu8.23257$Ji7.8640@atlpnn01.usenetserver.com>
Message-ID: <PAENINBOOODFDEPOKPKCMEHJCLAA.noah@noah.org>

I'd be interested to see that Delphi code.
As far as I could tell the messages in OE
are binary -- and not just simple binary
headers around the message texts.
Does your Delphi code use COM to get the
messages out of Outlook?

Yours,
Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Steve Holden
Sent: Tuesday, April 16, 2002 5:14 AM
To: python-list at python.org
Subject: Re: Backing up mail in Outlook Express using Python


"Thomas Weholt" <thomas at gatsoft.no> wrote in message
news:cGSu8.101$1kc.179320832 at news.telia.no...
> I need to backup all my mail stored in Outlook Express. Can I use Python
to
> do this?
>
No reason why not, though I'm not sure why you'd want to, since the message
are just stored in files.

You can find out where they are from the Maintenance tab on Tools | Options
(click the "Store Folder" button).

I did adapt some code from Delphi to read the mailbox files, and can let you
have that if it would help, but sounds to me like a file copy is all you're
looking for.

regards
 Steve









From tony-clpy at lownds.com  Wed Apr 17 22:23:47 2002
From: tony-clpy at lownds.com (Tony Lownds)
Date: 17 Apr 2002 19:23:47 -0700
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu> <3CBB9217.8A54E272@cosc.canterbury.ac.nz> <3CBC5C04.8F2C3D5A@noaa.gov> <afe3b146.0204171407.6e5e4b0e@posting.google.com>
Message-ID: <ad053a43.0204171823.54f54b99@posting.google.com>

> Tony Lownds has instructions here:
> http://mail.python.org/pipermail/pythonmac-sig/2002-April/005420.html
> 

Here too: http://tony.lownds.com/macosx/tkinter.html

Someday Tkinter and IDLE will work nicely OOTB on OSX!

> AquaTk seems to have some problems with path separators (try file open
> in IDLE for an example).


AquaTk's File open/save doesn't handle unix paths well. Since python
on Darwin (a.k.a. MachoPython) uses unix paths natively, IDLE's Open
and Save methods don't work. I believe this is posted as a bug in Tk's
Tracker, but it could be fixed in Python code too.

-Tony


From bokr at oz.net  Thu Apr  4 19:04:40 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Apr 2002 00:04:40 GMT
Subject: 1 .__nonzero__() => memory fault (2.2 win32)
Message-ID: <a8ipmo$qq0$1@216.39.172.122>

 Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
 Type "help", "copyright", "credits" or "license" for more information.
 >>> hasattr(1, '__nonzero__')
 1
 >>> 1 .__nonzero__()

(You get a memory fault).

Regards,
Bengt Richter


From bh at intevation.de  Fri Apr  5 05:04:16 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 05 Apr 2002 12:04:16 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017946214.8160.python-list@python.org> <zpL9tKA1hNr8EwkQ@jessikat.fsnet.co.uk>
Message-ID: <6qhemqa2cf.fsf@abnoba.intevation.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> In any large subject a 'Guido' will always claim to be bigger than 'God'

Well, 'Guido' *is* bigger than 'God':

Python 2.3a0 (#2, Apr  4 2002, 18:58:40) 
[GCC 2.95.4 20011006 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> 'Guido' > 'God'
True
>>> 

However, he's still not bigger than 'Larry':

>>> 'Guido' > 'Larry'
False
>>> 

But 'Tim' is:

>>> 'Tim' > 'Larry'
True
>>> 


   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From aleax at aleax.it  Sat Apr 27 16:10:33 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 20:10:33 GMT
Subject: Python's Lisp heritage
References: <mailman.1019922475.8361.python-list@python.org> <tyfsn5h13mh.fsf@pcitapi22.cern.ch>
Message-ID: <ZADy8.70845$vF6.2115921@news2.tin.it>

Jacek Generowicz wrote:
        ...
> The purpose of the parentheses (in this context) is to call functions
> and associate parameters with those function calls. Python has a not
> too dissimilar use of parentheses:

Anyone for Haskell's f a b c ...?  Now THAT is clean and elegant... the
concept of "a function with > 1 parameter" is unnecessary (as Haskell
Curry pointed out), so there's no need to associate a function with any
more than its first (and only) parameter: f a returns an unnamed
function which takes argument b, returning another unnamed function
which takes argument c.  Look, ma, no parentheses -- and no tricks
needed to achieve that, just some deep mathematical reality:-).

>   that you would really like to extend the set of infix operators,
>   with say @, to allow you to call splark without having to type those
>   infernal parentheses. Is it going to be easier in Python or in Lisp?

Haskell -- it has definition of new infix operators as a built-in.  But
THAT, IMHO, is a design error: languages with mutable syntax are
spending far too much complexity on syntax sugar -- and not to
good overall effect either.  The language's "new improved" syntax
may LOOK spiffier when you first see a new op introduced in some
article or tutorial about the language -- but when the rubber hits
the road, some poor devil of a programmer is going to have a very
hard time decrypting expressions using the new op -- where does
he or she go look up precedence and associativity, and how much
work does it take to parse them mentally...?

Prefix (and postfix) only need define arity (and Curry's result makes
that a don't-care for prefix:-), infix needs precedence and
associativity, and THAT is hard going.  I don't even _remember_
how a `f` b `g` c parses in Haskell (where stropping `f` is one simple
way to use a function f)... (f a) (b `g` c), or (f a b) `g` c ...?


Alex



From dave at pythonapocrypha.com  Mon Apr  8 14:03:23 2002
From: dave at pythonapocrypha.com (Dave Brueck)
Date: 8 Apr 2002 11:03:23 -0700
Subject: Booleans, integer division, backwards compatibility; where is Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <477762c2.0204081003.75867687@posting.google.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote in message news:<7xadshtrj1.fsf at ruckus.brouhaha.com>...
> Peter Hansen <peter at engcorp.com> writes:
> > No, that's "stagnant"...
> > 
> > Stable is also a good adjective for it, since most of the older
> > code still runs unchanged, and Python from ten years ago is
> > still very recognizable to new coders today.
> 
> Stable doesn't just mean old scripts run without change in new
> interpreters.  It also means new scripts usually run without change in
> old interpreters.

How could that possibly work? I must be misunderstanding what you mean
- could you please give an example from another language that
illustrates your point?

-Dave


From mwh at python.net  Fri Apr 12 05:41:14 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 12 Apr 2002 09:41:14 GMT
Subject: New classes, __slots__, __dict__
References: <mailman.1018592622.13221.python-list@python.org>
Message-ID: <lkd6x5jitz.fsf@pc150.maths.bris.ac.uk>

mertz at gnosis.cx (David Mertz, Ph.D.) writes:

> | Also, my admittedly simpleminded thought is that if handling slots
> | is a 2-liner, why doesn't pickle.py just do that? I have to believe that
> | the pickle authors are smart people and there is some deep technical
> | reason why getstate MUST exist if slots are used, and not just that the
> | pickle authors were too lazy and punted.
> 
> So what do you think... smart people who wrote pickle?  What's wrong
> with the (slightly-more-than-two) line solution?  What is the deep
> technical mojo that argues against this approach?

Time to read python-dev archives, I think.

Start here:

    http://mail.python.org/pipermail/python-dev/2002-February/020072.html

Then read this thread

    http://mail.python.org/pipermail/python-dev/2002-February/020129.html

HTH,
M.

-- 
  I would hereby duly point you at the website for the current pedal
  powered submarine world underwater speed record, except I've lost
  the URL.                                         -- Callas, cam.misc


From Tjabo.Kloppenburg at unix-ag.org  Mon Apr  1 20:03:56 2002
From: Tjabo.Kloppenburg at unix-ag.org (Tjabo Kloppenburg)
Date: Tue, 2 Apr 2002 00:03:56 -0100
Subject: IDLE for SUSE Linux?
In-Reply-To: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com>
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com>
Message-ID: <E16t1Dw-0003KV-00@mail.python.org>

hi,

> I'm looking for a good of binaries, or a good URL concerning setting up
> IDLE under SUSE Linux 7.3 Pro.  Got URL?

where's the problem?

In Suse 7.2 IDLE can be found on CDROM #4 in
  /suse/d3/python-demo.rpm

you can find the package by doing a 
  zgrep -i "idle.py" /cdrom/ARCHIVES.gz
with CDROM 1.

tk.
(better copy the .gz on your harddisk and write a small shell script to 
search in the copied file)





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sun Apr 14 18:18:06 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 15:18:06 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <eppstein-7156A5.13521614042002@news.service.uci.edu>
Message-ID: <3cba0068@news.mhogaming.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-7156A5.13521614042002 at news.service.uci.edu:

| < snip >
|
| If (a xor b) is true, it must be because exactly one of
| a and b is true. So, in that case, it seems reasonable
| to return the one that is true. I'm not sure what value
| would make most sense to return when (a xor b) is false
| though.
|
| < snip >

Ahhhh!!!  that's a GREAT idea!  That way, you could write
something like:

# give me either a or b, but only if one
# of them is true, else give me false:
c = a xor b

# the above would be equivalent to:
def xor(a,b):
    if   a and not b: return a
    elif b and not a: return b
    else:             return false

c = xor(a,b) # get either a or b, or false

This would indeed make 'xor' work very similar to the 'and'
(as well as the 'or') operator(s)...

Bengt Richter pointed out that if we want a boolean result
we can just use the '!=' operator (and he is right), but
the above behavior of 'xor' would be also useful...

Good thinking Dave!





From psage at ncaustin.com  Tue Apr 30 12:08:00 2002
From: psage at ncaustin.com (Paul Sage)
Date: Tue, 30 Apr 2002 11:08:00 -0500
Subject: Window question for Tkinter
Message-ID: <03F53405C6ED434A986221919785A6A7106C2F@nca-postal.ncaustin.com>

How do you get a window to pop back to the front?  For instance, I am
running PythonWin and I start my application.  The app runs and pops up
a new window.  The new window forces the main window behind PythonWin.
This is not the behavior I want.  Any ideas?




From bh at intevation.de  Sun Apr  7 16:36:55 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 07 Apr 2002 22:36:55 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com> <mailman.1017757086.9920.python-list@python.org> <KBor8.8537$q63.3756@atlpnn01.usenetserver.com> <3CAE1CC8.7D18F4A9@alcyone.com> <MPG.1717cb31e97aaec798969c@news.nwlink.com>
Message-ID: <6qelhri6u0.fsf@abnoba.intevation.de>

Jeff Shannon <jeff at ccvcorp.com> writes:

> In article <3CAE1CC8.7D18F4A9 at alcyone.com>, max at alcyone.com 
> says...
> > Comparing to None (preferably via is rather than ==, since == can be
> > overridden to do strange things) is perfectly legitimate when an API
> > declares that a function or method returns "an object or None."  After
> > all, the __nonzero__ method can be overridden to make perfectly valid
> > objects evaluate to false in a truth test.
> 
> I would suspect that anyone defining such an object would be 
> rather surprised if you refused to test it with 'if not object:' 

To illustrate the point Erik made, here's how I learned to compare to
None as explicitly as possible and not just testing whether a certain
object is true.

Consider a class like this:

class Xyzzy:

    def __init__(self, parent = None):
        """Instances of Xyzzy may have an optional parent"""
        if parent:
            # if we're instantiated with a parent add
            # ourselves to its list of children.
            parent.add_child(self)


The parents of the Xyzzy isntances defined some protocol for managing
children and the children always called a certain method of the parent
when you instantiated them with a parent.

At first, those parents were always true so that that if statement
always worked correctly, i.e. called the parent's add_child method if
and only if Xyzzy was instantiated with a parent.

Later, the parents were extended to support the sequence protocol, so
that you could use len(parent) to find out how many children there were,
for isntance. Unfortunately, that meant that suddenly all (potential)
parents that didn't have children yet were false! After all, an object
that doesn't implement __nonzero__ (which didn't yet exist in the python
version I was using back then IIRC) but implements __len__ is considered
false if __len__ returns 0.

The solution was to switch to an explicit test for None.

And testing for None with "is" is a good idea. Not only can any object
be equal to None if it implements the various __*__ methods related to
comparison accordingly, "is" is also faster. And when the objects you
compare to None implements __getattr__ but not __eq__ or __cmp__
comparing for equality can be quite slow because Python tries to get
hold of several "magic" methods before it reverts to testing for
identity anyway:

>>> class C:
...     def __getattr__(self, attr):
...             print "__getattr__", attr
...             raise AttributeError, attr
... 
>>> c = C()
>>> c == None
__getattr__ __eq__
__getattr__ __coerce__
__getattr__ __cmp__
0
>>> if c: pass
... 
__getattr__ __nonzero__
__getattr__ __len__
>>> 


Whether the speed matters depends on your particular application, of
course.


   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From martin at v.loewis.de  Tue Apr 16 14:58:59 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 20:58:59 +0200
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net>
Message-ID: <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de>

stojek at part-gmbh.de (Marcus Stojek) writes:

> The C-extension does what it is supposed to do, the callback function
> works, but the C-function does not start as a parallel thread but is
> freezing my UI. Anything basic I miss here ?

Yes, the Global Interpreter Lock (GIL). When C code runs that was
invoked from Python code, it holds the GIL, and only one such C code
can run at any time. So you need to release the lock, using
Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS.

Regards,
Martin



From ajs at ix.netcom.com  Fri Apr  5 21:27:16 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Fri, 5 Apr 2002 21:27:16 -0500
Subject: PEP 285: Adding a bool type
Message-ID: <000101c1dd13$a8c9d040$0334fea9@carol>

>In any large subject a 'Guido' will always claim to be bigger >than 'God'
>thus leading to Sergeant Pepper phenomena.

We all must trust our intuitions.

I had a moment of self-doubt.

Certainly there was nothing tremendously
unreasonable about Guido's remark that

"I don't think Boolean algebra is a big subject."

But David's - a PEP285 supporter -  remark
that indicating there might be more interest
in the subject than Guido thought - was purely
conversational and friendly.

That Mark felt compelled to intercede in  Guido's
defense I intuited as strange - a example of the
Phenom.

But maybe not.  Maybe it had nothing to do with
Guido being Guido. Maybe I had mis-intuited,

But I look in at Mark's weblog:

"my instant reaction to pep 285 (yes, i'm still thinking about this) was,
"rock on, guido!" it just seemed obviously right. then came the fud (labeled
such by guido-i wish i were that clever)."

For some, Guido's cleverness goes well beyond some proven
talent at computer language design.

But he ain't the only one with intuition.

Art







From aahz at pythoncraft.com  Mon Apr 22 01:22:34 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 01:22:34 -0400
Subject: Newbie programmer here...
References: <uc73k44mopm050@corp.supernews.com>
Message-ID: <aa06mq$cka$1@panix3.panix.com>

In article <uc73k44mopm050 at corp.supernews.com>, MH  <mhaag at pacific.net> wrote:
>
>I have recently been wrestling with PERL as a means to create interactive 
>websites, but I find PERL incredibly inconsistent and confusing, and the 
>debugging features I think are sorely lacking. I spend HUGE amounts of time 
>solving very basic problems.
>
>I am looking for ANYTHING but PERL to accomplish basic tasks such as 
>collecting, processing, and displaying data via CGI/HTML. Is PYTHON a good 
>candidate for this? What about PHP? Suggestions?

Well, at least if you don't like Python, you won't waste much of your
time.  Start with the tutorial at
http://www.python.org/doc/current/tut/tut.html
It should take you a day or three to work through it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From jorgenhansen at hotmail.com  Wed Apr 10 05:41:57 2002
From: jorgenhansen at hotmail.com (=?ISO-8859-1?Q?J=F8rgen_Hansen?=)
Date: 10 Apr 2002 02:41:57 -0700
Subject: Problem with packages and modules
References: <117880a1.0204090542.258f0f51@posting.google.com> <3CB2FA14.1050708@mxm.dk> <117880a1.0204092207.7e51ff80@posting.google.com> <Xns91EBEF972B8BFRASXnewsDFE1@130.133.1.4>
Message-ID: <117880a1.0204100141.63915d90@posting.google.com>

Philip Swartzleonard <starx at pacbell.net> wrote in message news:<Xns91EBEF972B8BFRASXnewsDFE1 at 130.133.1.4>...
> J?rgen Hansen || Tue 09 Apr 2002 11:07:37p:
> 
> > Max M <maxm at mxm.dk> wrote in message news:<3CB2FA14.1050708 at mxm.dk>...
> >> J?rgen Hansen wrote:
> >> 
> >> > This is an example of directory hierachy:
> > <snip example hierachy>
>  [ Snip some other stuff ]
> >> > Further if I typed in:
> >> >    >>> import company
> >> > 
> >> > I would assume that I could access all subpackages and their
> >> > modules with a dotted name (company.plot.figure.Figure() for
> >> > instance), but this is obviously not the case.
> >> >
> >> 
> >> You would need to import the modul in company/__init__.py:
> >> import plot
> >> 
> >> to do:
> >> 
> >> import company
> >> fig = company.plot.figure.Figure()
> >> 
> >> 
> >> Hmm... I hope this helps a bit.
> > 
> > It does. Thanks a lot. After I wrote, I got thinking, that my approach
> > would have meant that I would import the entire hierachy. Not very
> > smart if the hierachy, where larger.
> > 
> > Well thanks again, for the prompt answer.
> > 
> > regards
> > Jorgen
> > 
> 
> No one's mentioned this particular permutation; if you want to use 
> 'plot' as a logical module in your code, you can just do 'from company 
> import plot' =).

But that would still require that I import the required modules of
'plot' in the __init__.py file of the 'plot' package. At least that is
my experience, correct me if I'm wrong.

/Jorgen


From kybic at ieee.org  Wed Apr  3 10:53:01 2002
From: kybic at ieee.org (Jan Kybic)
Date: 03 Apr 2002 17:53:01 +0200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <lkhems53fa.fsf@elios.maths.bris.ac.uk>
Message-ID: <lbhy9g4kcde.fsf@xihu.inria.fr>

> > It would be so much easier if I could just compile parts of the
>                                         ^^^^
> You make it sound so simple...

Well, translating something like:


def fib(a):
        if a <= 2: return 1
        else: return fib(a-2) + fib(a-1)

into:

 int fib(int a) {                   
     if(a <= 2) return 1;
        else return fib(a-2) + fib(a-1);  }

does indeed seem to be simple, once you know that 'a' is an integer.
         
> Have you looked at psyco?

Yes, very impressive once it does what it promises. It is more
ambitious than what I had in mind in that it does not use any
assistance (type information) from the programmer.

> Well, there was boundless (it made me exceed my quota on the mail
> server I was using ot the time) discussion on the types-sig on
> more-or-less this sort of topic a couple of years back, although
> perhaps more of the focus was on guaranteeing correctness than
> optimization.  You could read that, but you don't want to.

Yes, I remember. I did not quite buy the correctness argument, runtime
checking is fine with me, but for a compiler it might be very useful, IMO.

> It's possible, perhaps even desirable, that psyco could be extended in
> this fashion.  Got a spare man-month or so?  I haven't.

No, neither have I, sorry. :-( Maybe later.

J.

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/


From dradul at yahoo.com  Tue Apr 16 11:02:50 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Tue, 16 Apr 2002 10:02:50 -0500
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1018875990.21271.python-list@python.org> <j43cxwm1o4.fsf@informatik.hu-berlin.de>
Message-ID: <a9id9a$33oqt$1@ID-99513.news.dfncis.de>

"Martin v. L?wis" <loewis at informatik.hu-berlin.de> escribi? en el
mensaje news:j43cxwm1o4.fsf at informatik.hu-berlin.de...
> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > > But this is really strange; mathmodule.so must have imported at
some
> > > point, or setup.py would have deleted it...
> >
> > Hmm.  I don't know.  Anyway, here is the result of re-making the
> > Python, but not the make test, after removing that file...
> >
> > Script started on Mon Apr 15 13:57:01 2002
> > neelix hgs 1 %> gmake
> > gcc -shared  Modules/mathmodule.o   -o Modules/mathmodule.so
>
> That explains it: he has enabled mathmodule.c in the *shared* section
> of Setup.
>
> I recommend to build Python from original sources, in a fresh
> directory.

And while you are at it... You have not compiled python itself as PIC
code. No, gcc -shared *doesn't* work in Solaris unless you are using GNU
Binutils instead of the native linker and assembler, something *not
recommended at all*. Solaris' linker can't bild a dynamic loading
library/module using object code compiled in position dependent mode.

This means, that you need to have all external libraries compiled with
the -fPIC flag, even if they are static. Ahh! And make sure that you
give the linker the actual library paths you use with an -R flag, else
the dynamic loader won't find them and you'll have to use an explicit
LD_LIBRARY_PATH environment variable for each invocation invocation of
the internpreter.



From kwokng at earthlink.net  Mon Apr 29 11:10:11 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Mon, 29 Apr 2002 15:10:11 GMT
Subject: python2.2
Message-ID: <nndz8.13032$iU4.1143455@newsread2.prod.itd.earthlink.net>

I am using debian linux.  I am trying to install the python2.2 that is an
unstable package onto the system, but it says "Couldn't find package
python2.2"  Wouls anyone know how to install the unstable package via
apt-get

Thanks!

Billy Ng




From martin at v.loewis.de  Tue Apr 16 02:50:49 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:50:49 +0200
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE> <m3wuv8iras.fsf@mira.informatik.hu-berlin.de> <a9fj34$3mi$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <m3ofgkkugm.fsf@mira.informatik.hu-berlin.de>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:

> Yes, I've been trying to reach their site, but various links didn't return
> anything.
> What do I need to do? I already installed PyXML.

What do you want to do today?

Regards,
Martin


From marklists at mceahern.com  Thu Apr  4 12:38:24 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 4 Apr 2002 11:38:24 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <eppstein-AA64FA.09075504042002@news.service.uci.edu>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEDHKJAA.marklists@mceahern.com>

[Guido]
> I don't think Boolean algebra is a big subject.

[David Eppstein]
> There are advanced graduate math books on Boolean algebra.
> [...]
> But these books' content is almost completely irrelevant for day-to-day
> computer programming.

Don't you figure that's what Guido meant by saying it's not a 'big' subject?
I mean, there are advanced philosophy books on Heidegger...

Cheers,

// mark





From senux at senux.com  Thu Apr 11 08:33:14 2002
From: senux at senux.com (Brian Lee)
Date: Thu, 11 Apr 2002 21:33:14 +0900
Subject: appeding string to file
Message-ID: <20020411123314.GA23460@jupiter.senux.com>

I tried follow test code to append string to a file. 
When list t is size 10000, second time is slower fourth
But list t is size 30000, second time is slower twenty three
times than first time.

Why does second test is more slower? Is the load to copy
string to a variable too high?

Any advice?

from Python newbie...

#!/usr/bin/env python

import time
import os

try:
    os.remove('test1.log')
except:
    pass    
try:
    os.remove('test2.log')
except:
    pass    

# prepare
t = []
for x in range(30000):
    t.append('Test string (' + str(x) + ')')
print 'Completed string t'

# first
start = time.time()
for x in t: 
    open('test1.log', 'a').write(x + '\n') 
stop = time.time()
print 'First test:', float(stop - start)

tt = ''
# second
start = time.time()
for x in t: 
    tt += x + '\n'
print 'start writing...'
open('test2.log', 'w').write(tt)
stop = time.time()
print 'Second test:', float(stop - start)

'''
RESULT:

[senux at porsche test]$ python2 write.py 
Completed string t
First test: 1.02917897701
start writing...
Second test: 4.56314098835
[senux at porsche test]$ python2 write.py 
Completed string t
First test: 3.14635396004
start writing...
Second test: 71.6274340153
'''

-- 
 ____       
 |o | i   / Brian Lee    <senux at senux.com>
 #######|<  Web - http://www.senux.com/en/
(x-x-x-x) \ GnuPG public key - 0x46C763A3




From gh_pythonlist at gmx.de  Wed Apr 24 00:04:04 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 24 Apr 2002 06:04:04 +0200
Subject: Accessing class-level variables
Message-ID: <20020424040404.GA1001@lilith.my-fqdn.de>

I'm trying to access a class variable in a derived class. I sincerly
hope there's a less kludgy way than what I'm currently using:

class X:
    a = 5

class Y(X):
    def p(self):
        print self.__class__.a

Y().p()

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 9.7 ?C      Wind: 2.2 m/s




From phil at mrinfo.de  Tue Apr 23 07:09:30 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Tue, 23 Apr 2002 13:09:30 +0200
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <aa3fbk$pfb$1@swifty.westend.com>

"Petr Prikryl" <Answer.via.news.please.prikrylp at nospam_skil.cz> wrote in
message news:aa3dti$n0$1 at news.nextra.cz...
>..
> In other words, should I prefer writing:
>     s1 = 'first string'
> or
>     s2 = "second string"
>..

For XHTML string output, I prefer single quotes because I don't have to
escape double-quotes within the string, e.g.:
s += '<img src="..." alt="" />'




From aahz at pythoncraft.com  Tue Apr 30 10:28:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 30 Apr 2002 10:28:52 -0400
Subject: OT: Happy Birthday Larry Niven
References: <3cce9f3a$1_6@goliath.newsgroups.com>
Message-ID: <aam9n4$kh2$1@panix2.panix.com>

In article <3cce9f3a$1_6 at goliath.newsgroups.com>,
Brad Clements <bkc at Murkworks.com> wrote:
>
>I think Python is a language that Motie Engineers would use...

<snicker>  I know a few Perl-lovers who'd disagree -- and I might agree
with them.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From phil at mrinfo.de  Mon Apr 22 07:35:38 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Mon, 22 Apr 2002 13:35:38 +0200
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83> <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>
Message-ID: <aa0sgl$9om$1@swifty.westend.com>

"Jarno J Virtanen" <jajvirta at cc.helsinki.fi> wrote in message
news:slrnac7qns.dh0.jajvirta at sirppi.helsinki.fi...
>..
> I find every other metaphor [for programming] lacking something
> essential too.  One popular metaphor is building construction or
> engineering activity in general.
>..

I once read about gardening as such metaphor. This was in the context of
refactoring software constantly, like making sure nothing grows too big.
Plant a seed here, and there, then go to another part, take a bit off the
earth and put it in somewhere else, and so on, see it grow over time and
manage to keep it living. I have to say I certainly think e.g. architecture
has much more planning before the actual "product" is build, and the process
of building that product isn't quite as coupled with the design phase, like
growing in organic ways. (Of course if you believe that you can have the
perfect product specification, the programmer's are simply builders in a
process of creating this fixed thing, but I haven't experienced that so far
in my own job, nor would I work in programming if that would be the case.)






From phd at phd.pp.ru  Mon Apr 29 09:33:12 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 29 Apr 2002 17:33:12 +0400
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Mon, Apr 29, 2002 at 09:15:57AM -0400
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <5yEp8.5051$hi7.18667@news-server.bigpond.net.au> <3CA7192C.E1C8974A@engcorp.com> <mailman.1017612966.7043.python-list@python.org> <uco40vl11ipj04@corp.supernews.com> <mailman.1020024596.32196.python-list@python.org> <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com>
Message-ID: <20020429173312.C4776@phd.pp.ru>

On Mon, Apr 29, 2002 at 09:15:57AM -0400, Steve Holden wrote:
> >    Gadfly now lives in the Zope sources at www.zope.org.
> 
> Oleg, hi! I also remember seeing somewhere patches that moved gadfly to re

   I think this is exactly that - in the Zope sources. Gadfly, after all,
is bundled with Zope.

> usage? gadfly.sf.net is a placeholder, it seems to me.

   Tried to look at this page? :)

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




From wware at alum.mit.edu  Sat Apr 13 09:32:04 2002
From: wware at alum.mit.edu (Will Ware)
Date: Sat, 13 Apr 2002 09:32:04 -0400
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com> <slrnabf5r2.s6b.grey@teleute.dmiyu.org> <7xr8lk46nl.fsf@ruckus.brouhaha.com>
Message-ID: <3CB83354.90CB406@alum.mit.edu>

Paul Rubin wrote:
> The SL-5500 hardware is similar to an Ipaq 36/37xx plus it has a cute
> little slide-out keyboard that's not all that useful.  I got the
> developer version for $199.00 which was a good deal at the price.  I
> wouldn't recommend it at $500 though.  I'll post a review to my website soon.

This thread got me curious about it. There are SL-5000D's (the developer
version) on eBay trading in the ballpark of $300. Is there any significant
difference between the 5000D and the 5500?


From tundra at tundraware.com  Mon Apr 22 14:30:02 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 18:30:02 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org> <3CC43FEC.EE94FFEB@kfunigraz.ac.at> <EC7694DBF1581DFC.5E45D82C67327D60.BDDD4EB8581664F1@lp.airnews.net>
Message-ID: <3CC4549E.B63D3E8A@tundraware.com>

Cameron Laird wrote:

> I'll say what I think is the same, in somewhat different
> language:  do not draw conclusions from mathematicians'
> philosophy.  There's reasonably strong evidence that
> mathematicians are, in the aggregate, unreliable witnesses
> as to the epistemology of their expertise.  They do good
> mathematics; they recognize good mathematics; they hold
> personal beliefs about what good mathematics is; but they
> are FAR more trustworthy about the former (their capacity
> to generate quality output) than the latter (philosophically
> valuable explanations).
> --

You're right, of course, but I would expand this to include anyone
of great ability.  Precious few geniuses can ever explain the
'epistemology of their expertise' be it in mathematics, boat building,
brick laying, or square dancing.  This is, to me, another reason to
dismiss the mechanist/meathematical explanations for the human mind and
spirit.  There is something transcendental about genius that I doubt can
ever be described in purely mechanical terms...

------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From paul at boddie.net  Tue Apr 23 06:19:27 2002
From: paul at boddie.net (Paul Boddie)
Date: 23 Apr 2002 03:19:27 -0700
Subject: DBAPI
References: <87adrvambl.fsf@complete.org> <Xns91F975562E137andy47halfcookedcom@203.109.252.31>
Message-ID: <23891c90.0204230219.76559977@posting.google.com>

John Goerzen <jgoerzen at complete.org> wrote in 
news:87adrvambl.fsf at complete.org:
> >
> > I have used things like JDBC for Java and DBI for Perl.  To me, DBAPI
> > for Python looks rather week.  For instance, if the code is to be
> > truly portable across databases, I'd have to check the paramstyle for
> > the driver and write each query 5 times!

I suppose you could always parameterize the parameter style. ;-) But I
think you have a valid point in this area.

> > Is there any "meta-DBAPI" module that I can use that would present me
> > with a standardized interface, and then (for instance) convert my
> > queries to whatever format the backend wants?

Someone did have some code to do more ambitious abstraction over
different database systems - search the DB-SIG mailing list archives
for more information. Bear in mind, however, that database-neutral
layers of all kinds (particularly those based on SQL) are non-trivial
to implement and it's arguably typical that many of those that get
started only get to cover about two or three different database
systems.

Andy Todd <andy47 at halfcooked.com> wrote in message news:<Xns91F975562E137andy47halfcookedcom at 203.109.252.31>...
>
> As far as I am aware there is no meta-DBAPI. In my experience when people  
> talk about standard database interfaces that enable them to write 
> portable code they haven't actually tried to write any. SQL isn't 
> standard so it is impossible to write database independent code for all 
> but the most trivial of applications.

Certain SQL standards do exist and it would be possible to write to
those and then have some kind of layer to translate the statements
into vendor-specific versions. If only database vendors adhered to the
standards...

[...]

> On the other hand it is very hard to write Python db modules that will 
> work in exactly the same way on several databases anyway. The example you 
> give of parameter style is a case in point, most Python db modules stick 
> close to the implementation in the supplied database libraries rather 
> than try and wrestle them into some abritrary standard.

Well, JDBC specifies an "arbitrary standard" and I don't think that
there's much need to complain about how it is now impossible to use
Proprietary Vendor Corporation's NonStandardParameterEncoding(TM) with
JDBC - most developers really don't want the "choice" that the vendors
give in this area. Moreover, the DB-API seems to confuse the issue
with its "Python-style" encodings, which I'm fairly sure is
responsible for part of the misunderstanding amongst certain
developers about what parameters are in SQL statements.

[...]

> For more detailed discussion may I suggest moving to the DB-SIG mailing 
> list, for details see - http://mail.python.org/mailman/listinfo/db-sig

See you there!

Paul


From greg at cosc.canterbury.ac.nz  Mon Apr  8 01:57:27 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 08 Apr 2002 17:57:27 +1200
Subject: ANN: Pyrex - a language for writing Python extension modules
References: <mailman.1017952983.647.clpa-moderators@python.org> <eppstein-4311CC.17190504042002@news.service.uci.edu>
Message-ID: <3CB13147.89F02706@cosc.canterbury.ac.nz>

David Eppstein wrote:
> 
> Not to criticise your project, but this is not a good example to use --
> you've started with a very unnecessarily slow algorithm.

The point of the example isn't to show how to find
prime numbers quickly. It's just a representative
CPU-intensive task that demonstrates what Pyrex
can achieve. Using a more efficient algorithm
wouldn't make that point any more effectively.

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


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 01:18:41 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 22:18:41 -0700
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org>
Message-ID: <3cba62f8@news.mhogaming.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1018828228.8551.python-list at python.org...

| < snip! >
|
| [Ken Peek]:
| > Given the above description of the way NEGATIVE
| > 'longs' should behave, my 'expectation' is that:
| > -0x8000000000000000 >> 1 _SHOULD_ return:
| > -0xC000000000000000,
|
| < snip! >
|
| No -- you're getting confused by the "-" notation, and
that's all.  Every
| other confusion follows from that.

You may be right that I am getting confused about the hex
display of longs-- That's because the display of a hex
number with a '-' sign prepended is such a ridiculous
notion, that it didn't even dawn on me that was what was
going on.  This is very confusing and I think this is a
_BUG_.  I don't care if "Python has always done it this
way"-- it is _WRONG_, and it is a _BUG_!  This should be
fixed so that decimal display of longs uses the '-' sign,
but hex display does not (you would just get the number of
hex digits in their raw "twos complement" form that the long
object's hex display method "knows" there is...)

There is another _BUG_ in Python:

>>> 0x40000000 << 1
-2147483648

This should have been automatically converted to a long, not
a negative int.

If we _ARE_ going to unify the long and int types, then
let's bite the bullet, and DO IT RIGHT.  Even if it breaks
code that depends on this bug being there.  I shouldn't have
to worry whether it's a long or an int when I am doing ANY
operation in Python-- they should both give the expected
result with _NO_ surprises...  (In fact, there is nothing
wrong with demoting a long to an int automatically, if the
value of a result can now be held in an int.)





From eppstein at ics.uci.edu  Mon Apr 15 17:38:28 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 14:38:28 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <eppstein-922653.14382815042002@news.service.uci.edu>

In article <m3662sk6xo.fsf at mira.informatik.hu-berlin.de>,
 martin at v.loewis.de (Martin v. Loewis) wrote:

> Yes, but this notion is inherently bound with the notion of "size" of
> an integer, atleast in the signed variant (for unsigned integers,
> there is no need for twos complement). If you create the complement,
> you have to know how many bits to invert.

Not if you have a representation which (when interpreted as a bitvector) 
allows bitvectors with infinitely many one-bits.  Which in fact is what 
we have with Python's long integer type.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From gerhard at bigfoot.de  Mon Apr  8 04:58:31 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 8 Apr 2002 08:58:31 GMT
Subject: Hiding the python source from the user !
References: <5e430dc.0204080028.2443f026@posting.google.com>
Message-ID: <slrnab2mtn.21k.gerhard@lilith.my-fqdn.de>

Arun KUmar wrote in comp.lang.python:
> hai,
> 
> i am creating an UI application in python, which i need to be given
> for demo purposes to my customers,
> so how can hide the source of my application?
> i don't to show the python code to the customers until he buys the
> product.
> 
> so how to do this?

Use a license.

If your product was good enough to be worth buying, you would already
know the alternatives or have found them on groups.google.com
<0.3 wink>.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From wurmy at earthlink.net  Tue Apr 16 01:34:48 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Tue, 16 Apr 2002 05:34:48 GMT
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za> <3CBA43B9.22850DF1@engcorp.com>
Message-ID: <3CBBB8A9.32CE5B9@earthlink.net>

Peter Hansen wrote:

> > http://boa-constructor.sourceforge.net
> 
> Of course, those products are well known around here.
> 
> What's less well known is the relationship between Boa-Constructor
> and Zope.  I strongly suspect there isn't one.
> 
> Can you elaborate on this supposed "incorporation of Zope
> into Boa-Constructor"?

I downloaded the CVS version of Boa the other day, and it
has a new tab on its components toolbar called "Zope", with
components like DTMLDocument, DTMLMethod, ZCatalog, GadflyDA,
etc. Not that I use it (yet)...

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA==')) 
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/


From boud at valdyas.org  Tue Apr 30 03:05:16 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 30 Apr 2002 09:05:16 +0200
Subject: Newbie: QTable Questions - Part II
References: <3CCDD95C.280F7DC6@bellsouth.net>
Message-ID: <3cce426d$0$37910$e4fe514c@dreader3.news.xs4all.nl>

Jeff Layton wrote:

> Hello,
> 
>    I'm in the midst of learning PyQT (and python more in-depth).
> I haven't found much if any documentation about QTable for
> PyQT. Andrew Dalke helped me get started (and pointing out
> that in my panic I forgot to sit down and think :). Now I want to
> modify the appearance of my table created with QTable. In
> particular, I want to change the :
> 
> (1) column headers (via QHeader I think)

QHeader.addLabel.

> (2) change the column width and row height

I think that you need QHeader.resizeSection or 

QTable.setColumnWidth, QTable.setRowHeight

> (3) Make the cells read-only

QTableItem.EditType or QTableItem.setEnabled

> (4) Change the color of text in the cells (and maybe the background
>     of the cells as well).

Subclass QTableItem.paint, and pass a different colour group to the
parent function.

> 
> Could a kind soul give me a kick in the pants to get me started?
> 

A kick, eh? Well, I must say I haven't tried all this; it's just
straight from the Qt documentation. You might run into some bugs
because this is relatively little testes. I still want to use QTable
in my Kura project, but I got side-tracked by a little language 
generator project and an editor :-).

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


From mwh at python.net  Wed Apr  3 12:07:48 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 3 Apr 2002 17:07:48 GMT
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <lkhems53fa.fsf@elios.maths.bris.ac.uk> <lbhy9g4kcde.fsf@xihu.inria.fr>
Message-ID: <lkhemsk8ws.fsf@elios.maths.bris.ac.uk>

Jan Kybic <kybic at ieee.org> writes:

> > > It would be so much easier if I could just compile parts of the
> >                                         ^^^^
> > You make it sound so simple...
> 
> Well, translating something like:
> 
> 
> def fib(a):
>         if a <= 2: return 1
>         else: return fib(a-2) + fib(a-1)
> 
> into:
> 
>  int fib(int a) {                   
>      if(a <= 2) return 1;
>         else return fib(a-2) + fib(a-1);  }
> 
> does indeed seem to be simple, once you know that 'a' is an integer.

Just picking, but the C has different semantics wrt. overflow...

> > Have you looked at psyco?
> 
> Yes, very impressive once it does what it promises. 

I'm not sure what promises you mean, but it gets a 4x speedup on
pystone today.

> It is more ambitious than what I had in mind in that it does not use
> any assistance (type information) from the programmer.

Yes.

> > Well, there was boundless (it made me exceed my quota on the mail
> > server I was using ot the time) discussion on the types-sig on
> > more-or-less this sort of topic a couple of years back, although
> > perhaps more of the focus was on guaranteeing correctness than
> > optimization.  You could read that, but you don't want to.
> 
> Yes, I remember. 

Ah.

> I did not quite buy the correctness argument, runtime checking is
> fine with me, but for a compiler it might be very useful, IMO.

I was more interested in the optimization too..

> > It's possible, perhaps even desirable, that psyco could be extended in
> > this fashion.  Got a spare man-month or so?  I haven't.
> 
> No, neither have I, sorry. :-( Maybe later.

Indeedy.  Maybe after 2.2.1 is done...

Cheers,
M.


From mertz at gnosis.cx  Thu Apr 11 23:49:45 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Thu, 11 Apr 2002 23:49:45 -0400
Subject: type() new style class instance says "class", not "ObjectType"
In-Reply-To: <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no>
References: <3cb4bd45$1_1@goliath.newsgroups.com>
 <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com>
 <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no>
Message-ID: <Zllt8kKkXAoD092yn@gnosis.cx>                

|>aaaanyway.. "type(thing) in [InstanceType]" is what xml_pickle uses to
|>figure out how to pickle stuff.

mlh at vier.idi.ntnu.no (Magnus Lie Hetland) wrote:
|Hm. How about simply removing the type check and see what happens?
|Such a check might have been necessary with old-style classes
|(although I think I'd rather check for the presence of a __class__
|attribute or something) it might well be that you can pickle
|non-instance objects the same way as instances with newstyle classes.
|After all, there is no big difference anymore.
|I see. It seems to rely on the division between instances and
|non-instances, which is becoming more and more meaningless.

Unfortunately, there's more to it than this.  There still is--and
presumably always will be--a difference between "instances" and other
types of things.  For example, a list is a mutable sequence that has
some items in it.  In contrast, an instance object--whether new style or
old style (and w/ or w/o __slots__)--is a thing that has some
attributes.  The basic shape of the data will always be different.  The
type/class unification doesn't change that.

Here's a simple example of xml_pickle at work:

    >>> from gnosis.xml.pickle import dumps
    >>> class C: pass
    ...
    >>> o, o2 = C(), C()
    >>> o.lst = [1,2,3]
    >>> o.obj = o2
    >>> print dumps(o)
    <?xml version="1.0"?>
    <!DOCTYPE PyObject SYSTEM "PyObjects.dtd">
    <PyObject module="__main__" class="C" id="1312268">
    <attr name="lst" type="list" id="1309612" >
      <item type="numeric" value="1" />
      <item type="numeric" value="2" />
      <item type="numeric" value="3" />
    </attr>
    <attr name="obj" type="PyObject" id="1312492"
          module="__main__" class="C">
    </attr>
    </PyObject>

It doesn't make sense to treat an instance-like thing the same way as a
list-like thing.  Or for that matter, a number-like thing needs to be
different also (it is a simple value, not a kind of container).

So somewhere in the logic of xml_pickle (however it might be
refactored), we need to ask the question "Is this attribute list-like or
instance-like (or dictionary-like, or number-like, etc)?"  That question
might not be spelled like:

    if type(item) in (ListType, TupleType):

But it needs to get asked one way or another.

Mind you, I know there are some borderline cases.  In fact, xml_pickle
has a lot more flexibility in dealing with borderline cases in
customizable ways than I have indicated.  I raise a question at this
borderline in another note I just sent to the list, in fact.  For
example, is a new style object that inherits from 'list' list-like or
instance-like?  Or an old style object that inherits from UserList?  Or
a NumPy array that is shaped a lot like a list?

These sorts of questions take some thought to answer, but the basic fact
that collections come in different shapes remains true.

Yours, David...

--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From nhodgson at bigpond.net.au  Sat Apr  6 17:20:45 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Sat, 06 Apr 2002 22:20:45 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net> <mailman.1018036693.6922.python-list@python.org> <slrnaaud6i.3j0.philh@comuno.freeserve.co.uk>
Message-ID: <1xKr8.2993$vc3.11896@news-server.bigpond.net.au>

phil hunt:.

> Indeed so. Stroustrup has been concerned not to break backwards
> compatibility with changes. This is probably one of the reasons C++
> is as popular as it is.

    The C++ version compatibility story is quite messy. If you just look at
it as an upgrade from standard C to standard C++ then it is simple, but C++
was a living language for 15 years before it was standardised. During this
period, the language was defined by diverse and often incompatible
implementations. During the standardisation process, implementations tracked
some changes to the language but not others.

   Every keyword added caused some code to break. I watched code break
because of the addition of "bool" and "typename". The change in for loop
variable scope "for (int i..." is *still* having repercussions. There is as
yet no complete implementation of the standard.

   Python is mostly defined by one implementation and so the version
compatibility issue is simpler and more visible. None of my own Python code
has broken because of Python changes between 1.5 to 2.2. This is without me
needing to worry about or code around the incompatibilities possible between
these versions. With my C++ code, which has been developed over a longer
period and is a larger body of work, I have had to code quite defensively
(for loop scope that works on both new and old, avoidance of the standard
library because of incompatibilities, avoidance of *exceptions* because of
Windows CE) and have still been bitten by the gradual progress of C++
implementations.

   C++ vendors are already talking about adding the features 'likely' in the
next standard in preference to full compliance with the current standard.
http://www.codeproject.com/interview/herbsutter3032002.asp
"But if I have my way we'll do that whole unified wish-list of C++98 and
pre-C++0x standard features in the order that best serves the community,
because that community priority is what I'm being paid to push inside the
company."

   Neil





From gbreed at cix.compulink.co.uk  Wed Apr  3 09:06:07 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Wed, 3 Apr 2002 14:06:07 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <Lkyq8.45231$pT1.1286029@news1.tin.it>
Message-ID: <a8f28f$7p8$1@thorium.cix.co.uk>

Alex Martelli wrote:

> They'd come in handy for those peculiar situations where you HAVE
> to distinguish between integers and booleans, basically only when you 
> are
> interfacing some external subsystem such as a DB or RPC system that
> does require such distinctions (as opposed to ones that _don't_ support
> it, such as, we're told, MySql).  And, by not affecting built-ins, 
> operators, etc, they wouldn't be in the way in everyday life, nor, most
> particularly, would they impose conceptual burdens on teaching Python.

If I were interfacing with an RPC system that happened to use dynamic 
typing between languages with different boolean implementations, it'd be 
pretty darned convenient if 1+1==2 happened to convert magically to the 
system's boolean type.  That'd also be the first time the proposal would 
make any difference to me.  My everyday life isn't really affected by the 
type of 1+1==2, so long as it evaluates true in situations where that's 
relevant.

> Far more "not-pure", since the PEP requires _implicit_ conversion (or
> else tons of code would break).

AIUI the PEP makes booleans a subclass of integers.  Which means they 
already are integers so no conversion needs to take place for them to 
behave as integers.

> It would be quite possible to distinguish between int and Booleans if
> you needed to, even if Booleans came from a specific module rather
> than from builtins, of course.  Such a need is a very rare issue, thus
> it should not be allowed to get in the way for everyday usage.

Again, is there a suggestion that the PEP *would* get in the way of 
everyday usage?  For that first time you do need to distinguish, it'd save 
a lot of trouble going through the documentation if a comparison like x<10 
already worked.


                      Graham

          <http://www.microtonal.co.uk>


From glchapman at earthlink.net  Thu Apr 25 09:57:25 2002
From: glchapman at earthlink.net (Greg Chapman)
Date: Thu, 25 Apr 2002 13:57:25 GMT
Subject: An IMHO interesting optimization question...
References: <3cc73c81@nntp.server.uni-frankfurt.de> <mailman.1019693386.30078.python-list@python.org>
Message-ID: <ad2gcukre8q2e5gcm86m2gd6krdgsii96a@4ax.com>

On Wed, 24 Apr 2002 19:08:53 -0500, Skip Montanaro <skip at pobox.com> wrote:

>The cost to run a loop is the startup/tear down cost plus the cost of
>running all the iterations.  I believe the startup/tear down cost of map()
>is fairly high, while it's per iteration cost is low (high and low being
>relative to the cost of the same work using a for loop).  You have to
>amortize that startup cost over a large number of iterations for it to pay
>off.  My guess is that your wordExp.findall() call only returns a list of at
>most a few elements.

My understanding is that the real problem with map is the overhead involved with
calling a Python function (i.e., a function implemented in Python).  Map called
with a function implemented in C (e.g., repr) will generally be faster than the
equivalent Python for loop, while map with a Python function will generally be
slower.  For example, on my system both of the for-loop variants below are
faster than the map with python function version (even though the second for
loop version does a global function lookup every time through the loop).  The
map with C function is the fastest:

import time

def pyrepr(obj):
    return `obj`

def map_cfunc(objs):
    return map(repr, objs)

def map_pyfunc(objs):
    return map(pyrepr, objs)

def for_operator(objs):
    res = [''] * len(objs)
    i = 0
    for obj in objs:
        res[i] = `obj`
        i += 1
    return res

def for_globalfunc(objs):
    res = [''] * len(objs)
    i = 0
    for obj in objs:
        res[i] = repr(obj)
        i += 1
    return res

def test():
    tests = [map_cfunc, map_pyfunc, for_operator, for_globalfunc]
    items = range(10000)
    for test in tests:
        start = time.clock()
        for i in range(10):
            test(items)
        stop = time.clock()
        print ('%14s:' % test.__name__), (stop-start)*1000


---
Greg Chapman



From sholden at holdenweb.com  Wed Apr 24 08:31:09 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 08:31:09 -0400
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <slrnacar0d.kd0.jon+usenet@snowy.squish.net> <u3ex8.94967$T%5.10614@atlpnn01.usenetserver.com> <slrnacc59a.tg8.jon+usenet@snowy.squish.net>
Message-ID: <MCxx8.45058$0x3.38185@atlpnn01.usenetserver.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnacc59a.tg8.jon+usenet at snowy.squish.net...
> In article <u3ex8.94967$T%5.10614 at atlpnn01.usenetserver.com>, Steve Holden
wrote:
> > This error was caused by Fernando's following a bogus suggestion of
mine.
> > His original SQL syntax was correct
>
> Uh, no it wasn't, he was using '%s' instead of %s. I'm confused now.

<sigh> Sorry, I should have had the courage of my own convictions. When
Fernando told me he was still getting errors on his SQL I assumed that the
MySQLdb module required delimiters around string parameters. You're right,
this doesn't make any sense. You should have been confused. It eventually
transpired that Fernando had an install problem, which a clean reinstall
fixed.

>>> import MySQLdb
>>> c = MySQLdb.connect(db="pwpsite")
>>> cc = c.cursor()
>>> cc.execute("SHOW TABLES")
5L
>>> cc.fetchall()
(('chapter',), ('comment',), ('part',), ('remark',), ('reviews',))
>>> cc.execute("CREATE TABLE tmp (k integer primary key, s char(12))")
0L
>>> cc.execute("INSERT INTO tmp (k, s) VALUES (%s, %s)",
...     [(1, "one"), (2, 'two'), (3, 'three')] )
3L

So, for the record, you should *not* put SQL string quotes around parameter
markers in your parameterized SQL statements.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/





From tdelaney at avaya.com  Tue Apr  2 23:24:46 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 3 Apr 2002 14:24:46 +1000 
Subject: Static Variables
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A248@natasha.auslabs.avaya.com>

> From: Gerhard H?ring [mailto:gerhard at bigfoot.de]
> 
> Jim Jacobs wrote in comp.lang.python:
> > Is it possible to simulate "C" style static variables in functions?
> 
> No.
> 
> > In methods?
>  
> No.

Of course it's possible.

def function (a, b, c=[]):
    pass

c is a static variable (however, could be overridden by the caller).

> This concept sucks anyway, IMNSHO.

Generally, yes, but it can be very useful. For example, for function-level
synchronisation (ignoring the use of Queue.Queue here ;)

def function (a, b, lock=threading.RLock()):
    
    lock.acquire()

    try:
        pass
    finally:
        lock.release()

Tim Delaney



From tejarex at yahoo.com  Tue Apr 23 18:14:05 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Tue, 23 Apr 2002 22:14:05 GMT
Subject: file_ref = open(string_ref)
References: <1019579845.565223@cswreg.cos.agilent.com>
Message-ID: <N0lx8.150362$GS6.14733110@bin3.nnrp.aus1.giganews.com>

"Greg Weeks" <weeks at vitus.scs.agilent.com> wrote in message
news:1019579845.565223 at cswreg.cos.agilent.com...
>.  So, in my
> opinion, our view of data is wrong.  To align our view of data to
our code,
> we need to replace the view
>
> V1: Objects are regions of memory.
>
> with:
>
> V2: Objects are addresses (of regions of memory).

I consider both views to be very wrong.  In Python, an object is an
information unit with an identity, a type, and a value.  When Python
code is executed by a computer, objects are stored in one or more
blocks of linear memory.  When executed by a human, 'blocks of linear
memory' almost certain does not apply and 'region' of memory' may be
problematical.  In either case, neither a container nor its location
is the same as its contents, and it is a mistake to confuse the three.

Terry J. Reedy





From jepler at unpythonic.net  Tue Apr 30 12:09:21 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 30 Apr 2002 11:09:21 -0500
Subject: Tkinter Text widget reading input (how to do overwrite mode?)
In-Reply-To: <13285ea2.0204300303.1343dc44@posting.google.com>
References: <13285ea2.0204290624.4fbe8cba@posting.google.com> <20020429100034.A12526@unpythonic.net> <mailman.1020104511.3714.python-list@python.org> <13285ea2.0204300303.1343dc44@posting.google.com>
Message-ID: <20020430110921.B17488@unpythonic.net>

Here's another version of the code, with a small enhancement plus 
a number of comments.


import Tkinter, sys

def comparable_index(index):
    """ Convert a Text index like '37.15' to a tuple that can be compared
    with Python's normal comparison operators """
    return map(int, index.split("."))

class OverwriteText(Tkinter.Text):
    def __init__(self, *args, **kw):
	Tkinter.Text.__init__(self, *args, **kw) # Create the widget

	# Insert the OverwriteText bindings in the widget tags.
	# When an event happens, Tk starts with the most specific tag
	# and unless processing is stopped, it moves to the next
	# less specific tag.  At each step, the best match for the
	# event is determined, and that binding is executed.

	# Usually a widget's tags look like
	#  .full.path WidgetClass .toplevel all
	# where .full.path is the Tk name for the widget, and .toplevel
	# is the name for the toplevel window containing it (often ".").
	# Because we want to modify the behavior of the Text widget,
	# we insert this new tag, OverwriteText just before Text.
	# It *should* always be between indices 0 and 1, but we go through
	# this activity anyway.
	tags = list(self.bindtags())
	idx = tags.index("Text")
	tags[idx:idx] = ["OverwriteText"]
	self.bindtags(tuple(tags))
	self.is_overwrite = 0

	# bind_class() with one argument returns all the bindings for that
	# class.  If none exist, then create them.  This notion of the "class"
	# is distinct from Python classes
	if not self.bind_class("OverwriteText"):
	    self._init_bindings()

    def _init_bindings(self):
	# Create two bindings, one that will change the way <KeyPress> works,
	# and one that will replace the old behavior of the Insert key with
	# a new one (toggle overwrite mode)
	
	# Just writing self.bind_class(..., OverwriteText.KeyPress) would not
	# work, because the first (only) arg would be "evt" not an instance
	# of OverwriteText

	# The following bindings are necessary due to the Tk event matching
	# machinery.  We don't want to treat Alt-X the same as X, and the way
	# Tk does this is by creating a more specific binding which performs
	# no action.  The whole thing is complicated, and explained in the
	# Tk manual page for "bind" under the "MULTIPLE MATCHES" section.
	#     http://tcl.activestate.com/man/tcl8.2.3/TkCmd/bind.htm#M41
	nop = "# nothing"
	self.bind_class("OverwriteText", "<Alt-KeyPress>", nop)
	self.bind_class("OverwriteText", "<Meta-KeyPress>", nop)
	self.bind_class("OverwriteText", "<Control-KeyPress>", nop)
	self.bind_class("OverwriteText", "<Escape>", nop)
	self.bind_class("OverwriteText", "<KP_Enter>", nop)
	if sys.platform == "mac":
	    self.bind_class("OverwriteText", "<Command-KeyPress>", nop)

	self.bind_class("OverwriteText", "<KeyPress>",
	    lambda evt: OverwriteText.KeyPress(evt.widget, evt))
	self.bind_class("OverwriteText", "<Insert>",
	    lambda evt: OverwriteText.Insert(evt.widget, evt))

    def toggle_overwrite(self):
	self.is_overwrite = not self.is_overwrite

    def KeyPress(self, evt):
	# This is called when a key has been pressed, whether or not
	# overwrite mode is enabled
	if self.is_overwrite:
	    # We find where the end-of-line is
	    eol = self.index("insert lineend")
	    # And we also find the index for the characters we would ovwrrite
	    nextchar = self.index("insert + %dc" % len(evt.char))
	    eol_num = comparable_index(eol)
	    nextchar_num = comparable_index(nextchar)
	    if eol_num < nextchar_num:
		# If the EOL is closer to the insert point, then delete only
		# that far
		index = eol
	    else:
		# Otherwise, delete as many chars as we'll insert
		index = nextchar
	    # Actually perform the deletion, from the insertion point to
	    # the index determined above
	    self.delete("insert", index)

    def Insert(self, evt):
	self.toggle_overwrite()
	# Returning "break" from the binding function means that
	# Tk will not pass this event on to the next tag.  Otherwise,
	# pressing Insert would toggle between insert and overwrite modes,
	# and then insert the text from the clipboard
	return "break"

def _test():
    import sys
    t = OverwriteText()
    t.pack()
    t.insert(0.0, open(sys.argv[0]).read())
    t.mainloop()

# If you run this as a script (as opposed to importing it), create a sample
if __name__ == '__main__': _test()




From hamish_lawson at yahoo.co.uk  Wed Apr  3 04:17:03 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 3 Apr 2002 01:17:03 -0800
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net> <3CAA7FD0.56C285E0@engcorp.com>
Message-ID: <915a998f.0204030117.3078a735@posting.google.com>

Peter Hansen wrote:

> Jim Jacobs wrote:
> > 
> > Is it possible to simulate "C" style static variables in functions?  In
> > methods?
> 
> Maybe ...  what do you mean by simulate, and what characteristics
> of the C statics would you like to have?
> [...]
> In Python, you would use either an instance variable, inside
> a method of an object, or the answer would be "no, but can
> you get by with using 'global'?".

If the requirement is to be able to get a different result for each
call of the function, then a generator might be the answer.

Hamish Lawson


From tundra at tundraware.com  Mon Apr 22 14:20:02 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 18:20:02 GMT
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
  <mailman.1019299357.11513.python-list@python.org>
  <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
  <Xns91F67C40EB0E0lumberjack@216.148.53.83>
  <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
  <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
  <3CC3A89C.BE2BF65@tundraware.com> <mailman.1019491732.12046.python-list@python.org>
Message-ID: <3CC4536A.81C9B57D@tundraware.com>

Lulu of the Lotus-Eaters wrote:
> 
> |> mathematics, including the study of algorithms, is a natural science.
> 
> Tim Daneliuk <tundra at tundraware.com> wrote previously:
> |I disagree strongly...  All of mathematics is a formal construct of the
> |human mind created with the intent of removing ambiguity and enhancing
> |our ability to describe what we *think* and what we *observe*.
> 
> To my mind, Daneliuk present here the strongest argument that
> mathematics (and therefore CS) *IS* indeed a science.
> 
> The wrinkle I would add to it is in describing exactly *which* science
> mathematics is.  Rather than being something close to Physics, and
> describing the behavior of the (large-scale) physical universe, I see
> Mathematics as a special branch of Psychology.  Mathematics describes
> the laws of thought, but at a level of abstraction that applies to any
> thinking being.  If we find martians, or teach dolphins, or invent
> computers sophisticated enough to "think", their thought patterns will
> be constrained by the rules of Mathematics--even though they quite
> likely will not be characterized by uniquely human Psychological
> features like, say Oedipal complexes, structured 3-scale memory,
> prototype-centered reasoning, and so on.

Hmm, this is a really interesting argument.  However appealing it is, I
find no compelling evidence to support the idea that "...thought patterns 
will be constrained by the rules of Mathematics."  I'm not saying it is
wrong (though my instinct is that it is, indeed, wrong).  I am saying it
is heretofore unproven.  If you take this position, you are arguing that
*all* human intellectual activitiy like creative action, analysis,
emotion, need, and so on can, at least in priciple, be described
mathematically.  This is a very mechanical view of the human intellect,
which I find doubtful...

>
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From gcordova at hebmex.com  Mon Apr  1 13:49:03 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 1 Apr 2002 12:49:03 -0600 
Subject: GUI programming
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB33@mail.hebmex.com>

> [..references to building wxGTK problems..]
> All in all, this can be very frustrating to new users.
>

Extremely.

I gave up after the n-th iteration.

And let's not talk about enabling GL. :-(

By far, Tkinter is the defacto standard precisely
because it's ubiquitous. Not because it's the
fasters, nicest, easiest (although it *is* quite
simple to learn), efficientest, skinnablest,
or whatever. But because it's the easiest to
obtain (on most platforms), and ubiquitous.

Salutations!

-gustavo



From wilk at flibuste.net  Fri Apr 12 10:26:08 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Fri, 12 Apr 2002 16:26:08 +0200
Subject: 2.1 stable and 2.2 dev ?
Message-ID: <20020412162608.22381c75.wilk@flibuste.net>

hi,

I would like to know if the version 2.1.x is the stable version and the
2.2.x the dev version ? If it's like that, it's oposite of linux version
where odd number are for dev...

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From starx at pacbell.net  Wed Apr 10 05:17:35 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 10 Apr 2002 09:17:35 GMT
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54> <a90rv2$fvh$1@slb4.atl.mindspring.net>
Message-ID: <Xns91EC17861D8C7RASXnewsDFE1@130.133.1.4>

Andrew Dalke || Wed 10 Apr 2002 01:00:02a:

> Emile van Sebille:
>>Oooh... that's almost as much fun as saying 'xyzzy' the first time...
> 
> I just said it, but nothing happened here.

Wouldn't a more likly WoP around here be something like 'NI!' =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From tim.one at comcast.net  Wed Apr 10 19:46:10 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 10 Apr 2002 19:46:10 -0400
Subject: Easter Eggs
In-Reply-To: <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAKONAA.tim.one@comcast.net>

[Robin Becker]
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers.

Oh yes, and that was the point.  The idiots in charge stayed up an extra
minute to dream up something that would only upset you.  I'm not sure what
it says about who that they succeeded so easily <wink>.





From sameer_ at email.com  Sat Apr 13 20:04:30 2002
From: sameer_ at email.com (sameer)
Date: 13 Apr 2002 17:04:30 -0700
Subject: can't see list text of a wxPython ListCtrl widget
Message-ID: <6bd9f01b.0204131604.6031da9e@posting.google.com>

hi, 
  I am playing around with the wxPython ListCtrl widget, and for some
reason I can't figure out, I am not seeing the text I add to the list.
 I can see the Column Names just fine, just not any item in the list. 
I am pasting the important code below which is part of an init method
of a wxPanel.

        tID = wxNewId()
        self.list = wxListCtrl(self, tID, pos = (5, 5),
style=wxLC_REPORT)
        self.list.SetDimensions(5, 5, 400, 200)
        self.list.InsertColumn(0, "Subject")
        self.list.InsertColumn(1, "Entry")
        self.list.InsertColumn(2, "Date")
        self.list.SetColumnWidth(0, 100)
        self.list.SetColumnWidth(1, 100)
        self.list.SetColumnWidth(2, 100)
        self.list.SetStringItem(0, 0, 'ho')
        self.list.SetStringItem(0, 1, 'b...')
        self.list.SetStringItem(0, 2, '0-2002')
        self.list.SetStringItem(1, 0, 'ho')
        self.list.SetStringItem(1, 1, 'b...')
        self.list.SetStringItem(1, 2, '0002')



Thanks in advance for the help.


From niemeyer at conectiva.com  Wed Apr  3 23:42:55 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Thu, 4 Apr 2002 01:42:55 -0300
Subject: [Python-Dev] ACCEPTED: PEP 285
In-Reply-To: <3CAB8C99.232A8571@metaslash.com>
References: <200204032232.g33MWqJ28933@odiug.zope.com> <3CAB8C99.232A8571@metaslash.com>
Message-ID: <20020404014255.A1388@ibook.distro.conectiva>

Hi Neal!

> 	    if a is True: print 'True'
[...]
> 	test.py:6: Comparisions with True are not necessary

Testing if "a is True" is ok. True and False are singletons
this is the best way to know if 'a' is really 'True' (and not 1,
[0], "foo", etc). This test will be necessary sometimes.

> 	    if a == False: print 'False'
[...]
> 	test.py:7: Comparisions with False are not necessary

This should probably say something more agressive, like
"Comparisions with boolean values may be dangerous", since this
may give undesirable results (if the user was trying to check
for False, and not for 0, [], "", etc).

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From newsfroups at jerf.org  Thu Apr  4 00:07:31 2002
From: newsfroups at jerf.org (Jeremy Bowers)
Date: Thu, 04 Apr 2002 05:07:31 GMT
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca>
Message-ID: <3CABDF6B.5020808@jerf.org>

vio wrote:
> Greetings,
> 
> I need to translate a (rather voluminous 100MB) java program to python. After a
> few searches on google, apparently the best tools for this task are Lex-Yacc
> (Flex-Bison on linux). While I am working on the Lex-YACC HOWTO, I would be very
> interested in hearing of others' successful experiences/feedback at this. Any
> pointers of interest would similarily be appreciated.

Translating amoung computer languages is not quite as hard as 
translating amoung human languages, but it's not exactly trivial either. 
It's the little things that will trip you up: The legality and meaning 
of negative indices, the subtle differences in rules for what can go in 
a dictionary, subtle differences in precision, order of operations, 
facilities that don't quite work in one language or the other that 
require thousands of lines of code to precisely emulate, but that a 
*human* could tell could be fixed with just two carefully chosen lines. 
Features in Java with no direct equivalents in Python (overloading's a 
big one, but it's not the only one). Differences in propogation of 
exceptions.

And you say you want this because you have a 100MB of code. By the same 
token, that 200+MB of code you'll still need to verify after this 
hypothetical operation, and 200+MB of code that will be throwing new and 
subtle errors... if you have unit tests. (If you don't have unit tests, 
quit now; you don't stand a chance in hell of testing 200MB of code. 
Even if it's 195MB comments!)

(Doubling in size is not unreasonable. Don't expect it to shrink in 
size, or improve in readibility. Entrophy requires your code to decrease 
in legibility; the more you automate, the more true this is. Ever looked 
at the assembly output of your C compiler? Or looked at C++ translated 
into C, back when compilers worked that way? I'd actually guess 3-5x, 
what with all the little error checking and behavior fixing code that 
will probably have to be added every time you use a data structure or 
library call that almost, but doesn't quite, exist in the other 
language. Add all this together, and the code you're verifying won't 
exactly be the paragon of readibility, either; this will be as close to 
obfuscated Python as you can expect to see in your actual life.)

And the fact that you will experience tens of thousands of instances of 
"the little things", each multiplying and playing off of each other.

Bear in mind that even if you get the grammer right, you essentially 
have to write a complete and total interface in Python to match the 
libraries you use in Java. Use Java streams? You will need to write an 
emulator for that stream in Python. (Even if you think you can convert 
the use of streams to some python equivalent, you need to write that 
too! Syntax analysis will only be the first step of a long journey here!)

God help you if this is a GUI project!

Use the network? More subtle differences, even after you code the 
translation layer. Data structures? You'll need to emulate those. How 
many 3rd party Java libraries do you use? (Are you sure?) You'll have to 
re-implement those too. Even if it's just writing a wrapper around a 
Python library with similar (but not identical!) behavior, that's still 
a non-trivial task, because you need to test it, and test it, and test 
it some more, because *100MB* of Java code may use it in every 
conceivable way! You can't even get away with capability-for-capability 
compatibility, you need bug-for-bug!

When all's said and done, you essentially need to write Java inside 
Python, at least all of it you use. Somebody wrote C++ in C. Somebody 
wrote C in assembler. Somebody wrote assembler in machine code. Somebody 
wrote Java in C++. (Sure, the projects have since migrated, but that 
doesn't really apply to your case.) Actually, lots of someones.

You are faced with a virtually impossible task; how many people will be 
helping you translate this 100MB of code, and how long do you have? 
Trying to replace it with an actually impossible task is not a step 
forward, even if they both look equally unlikely to be doable now.  If 
there's no other option, perhaps you should simply seek employment 
elsewhere; it's likely to be more rewarding in the long run.

Maybe, just *maybe*, if this is 100MB of **pure** computational code, 
with a small input module, a small output module, and 99MB of 
mathematical-type manipulations that never ever *ever* calls out to any 
sort of library in Java, you *might* be able to get away with this. 
Maybe. Maybe. I'd be several orders of magnitude more optimistic if 
those "M"'s were "K"'s, though.



From gusraan at terra.com.br  Mon Apr 22 19:44:23 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: Mon, 22 Apr 2002 20:44:23 -0300
Subject: Developing and ircd with python
Message-ID: <aa278j$8c8$1@nsnmrro2-gest.nuria.telefonica-data.net>

Is it possible?





From dfan at harmonixmusic.com  Mon Apr  1 10:09:12 2002
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 01 Apr 2002 10:09:12 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xhemy1jo4.fsf@ruckus.brouhaha.com>
Message-ID: <uadsncv6v.fsf@turangalila.harmonixmusic.com>

Paul Rubin <phr-n2002a at nightsong.com> writes:

| Guido van Rossum <guido at python.org> writes:
|
| >     There's also the issue (which I've seen puzzling even experienced
| >     Pythonistas who had been away from the language for a while) that if
| >     you see:
| > 
| >         >>> cmp(a, b)
| >         1
| >         >>> cmp(a, a)
| >         0
| >         >>> 
| > 
| >     you might be tempted to believe that cmp() also returned a truth
| >     value.  If ints are not (normally) used for Booleans results, this
| >     would stand out much more clearly as something completely
| >     different.
| 
| If cmp returns a boolean, it can no longer distinguish the 3 cases
| a < b, a == b, or a > b.  That is seriously broken.  

Guido is not proposing any change to cmp.  He's proposing that other
functions that actually do mean to return truth values return True or
False, rather than 0 or 1, so that people will not look at cmp output
and think "Oh, that looks just like boolean results."

Dan

-- 
http://www.dfan.org


From python at rcn.com  Mon Apr 29 14:03:02 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 14:03:02 -0400
Subject: Third Draft:  Pep for Deprecating Builtins
Message-ID: <aak1uq$bgt$1@bob.news.rcn.net>

>From the comments, I've learned:

1. map, filter, divmod, and pow may be immovable and should
    be left alone.

2. reduce has fewer friends, but should probably also be
    left alone

3. input() has almost no supporters and should probably be
    deprecated.  it is security risk, a trap for new users, easily
    replaced with eval(raw_input), and is named in a way that
    doesn't suggest that eval is occurring.

4. apply() was repeatedly suggested as another candidate for
    deprecation since the * and ** forms make it obsolete

5.  oct() is another candidate based on the rarity of usage

6. command line options and environment variables are
    universally hated, so this idea is dropped

7. the language needs some means of deprecating so that clutter
    does accumulate.  that means needs to be well documented,
    have a long, slow phase-in, and have a mechasism for
    restoring old behavior if it is ever needed.

8. there is some lumberjack person who is very angry

9. putting new functional tools in a separate module was
    well received.

Raymond Hettinger





From phil at mrinfo.de  Wed Apr 24 05:22:01 2002
From: phil at mrinfo.de (Philipp Lenssen)
Date: Wed, 24 Apr 2002 11:22:01 +0200
Subject: Php vs Python
References: <mailman.1019615325.15799.python-list@python.org>
Message-ID: <aa5te4$c4r$1@swifty.westend.com>

"Ramil D. Tolentino" <ai.matrix at eudoramail.com> wrote in message
news:mailman.1019615325.15799.python-list at python.org...
> Could someone give me a run down between python and php. Which is
> easier to use? which is better or more powerful?
>..

PHP will make you put a $ preceding every variable, which is annoying (and
can cause errors, especially if you switch between languages). Python has
more restricted namespaces and private functions in classes, in other words
it's more organized than PHP. Python lacks a switch and ternary operator
which PHP doesn't. The syntax of Python seems to a bit more eccentric
compared to several languages, but I find the approach elegant and easy to
pick up. As for features I can't give you good insight as there's so many
functions and libraries available for both languages. The main PHP website/
online documentation is better than the official Python one. For newsgroup
support Python wins, compared to the german PHP group. The support for
Python on Apache servers seems to be a bit lower. The security in Python
seems to be a bit higher (no global post-variables, even though they're
fixing that in new PHP releases). XML support seems to be about the same at
the moment. Python wins in error handling. Python can be used for developing
offline GUIs.

In the end, this is just all my opinion from web development perspective.
and a matter of taste, and if you like the one or the other better and it
gets your job done that it's the best tool for you.




From mcherm at destiny.com  Mon Apr  8 11:10:03 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 08 Apr 2002 11:10:03 -0400
Subject: PEP 284, Integer for-loops (code bash?)
Message-ID: <3CB1B2CB.4050601@destiny.com>

Laura Creighton writes:
> I think that it
> would have been nice if when Guido was designing the language he
> used Haskall's syntax for loops.  But he didn't.

I *still* haven't found time to learn Haskall, and a quick browse of the 
docs didn't get me anywhere. Can you summarize what Haskall's syntax
for loops IS? I realize this is a little off topic, but I'm curious.

-- Michael Chermside





From robin at jessikat.fsnet.co.uk  Sat Apr 13 20:14:57 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 14 Apr 2002 01:14:57 +0100
Subject: cStringIO write
References: <QHxVGIALCDu8EwLs@jessikat.fsnet.co.uk>
 <krYt8.13007$b62.354393@news1.tin.it> <fTIWTKAuNHu8Ewq$@jessikat.fsnet.co.uk>
 <Ne2u8.13850$b62.382109@news1.tin.it>
Message-ID: <6oTvtIABoMu8Ew9I@jessikat.fsnet.co.uk>

In article <Ne2u8.13850$b62.382109 at news1.tin.it>, Alex Martelli
<aleax at aleax.it> writes
>Robin Becker wrote:
>        ...
>> strange in my docs I see no mention of these special types. I see for
>
>You may not be looking at the latest version of the docs, at
>
>http://www.python.org/dev/doc/devel/lib/module-cStringIO.html
>
>
>Alex
>
thanks
-- 
Robin Becker


From vjovanov at stevens-tech.edu  Mon Apr 29 16:57:26 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Mon, 29 Apr 2002 16:57:26 -0400
Subject: cPickle problem
Message-ID: <Ssiz8.63$bK3.1993791@newnews.cc.stevens-tech.edu>

I thought that cPickle can dump almost any kind of object, but it
doesn't work on this one


>>> import cPickle
>>> class foo:
 def __init__(self):
  self.e={'a':'4+5'}
 def __getattr__(self,arg):
  try:
   exec 'result =' + self.e[arg]
  except:
   return '#U'
  return result


>>> A=foo()
>>> A.a
9
>>> file=open('test','w')
>>> cPickle.dump(A,file)
Traceback (most recent call last):
  File "<pyshell#62>", line 1, in ?
    cPickle.dump(A,file)
TypeError: object of type 'string' is not callable


Any idea what is wrong?

Vin




From max at alcyone.com  Fri Apr 26 05:36:20 2002
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 26 Apr 2002 02:36:20 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org>
Message-ID: <3CC91F94.4BAEEA62@alcyone.com>

"James J. Besemer" wrote:

> Seriously, I agree, the underlying abstraction -- the semantics --
> does not require
> parentheses.  But that's not really the language, per se.  I'm of the
> school that a
> particular language, curly-L, consists of a tuple: ( Syntax x
> Semantics ).

That's a Cartesian product, not a tuple.  :-)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is one long process of getting tired.
\__/ Samuel Butler
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From phd at phd.pp.ru  Wed Apr 10 11:22:32 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 10 Apr 2002 19:22:32 +0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <uit6zvbur.fsf@python.net>; from mwh@python.net on Wed, Apr 10, 2002 at 02:54:20PM +0000
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net>
Message-ID: <20020410192232.G19899@phd.pp.ru>

On Wed, Apr 10, 2002 at 02:54:20PM +0000, Michael Hudson wrote:
> >    Environment: SUN Sparc Ultra-1, Solaris 2.5.1, gcc 2.8.1.
>                                                         ^^^^^
> Is this negotiable?  I.e. could you try upgrading?  This is pretty old now.

   I don't care if it is old, I love old and proven software :) Really, I
don't understand the rush for all new versions.

> > gcc -c -DNDEBUG -O2 -I. -I./Include -DHAVE_CONFIG_H  -o Python/thread.o Python/thread.c
> > In file included from Python/thread_pthread.h:9,
> >                  from Python/thread.c:113:
> > /usr/include/pthread.h:238: parse error before `destructor'
> > /usr/include/pthread.h:238: parse error before `)'
> > make: *** [Python/thread.o] Error 1
> > 
> >    The line in question /usr/include/pthread.h:238 is
> > int pthread_key_create(pthread_key_t *key, void (*destructor)(void *));
> >    Nothing suspicious.
> > 
> >    Any idea what can be wrong? Python/thread_pthread.h has
> > 
> > #ifdef __APPLE__
> 
> Try making that 
> 
> #if defined(__APPLE__) || defined(__SUN__)

   The define is "sun", and it seems it helped. At least it is being
compiled now.
   Thank you for the help. Would you put it in the CVS or at least file a
bug report?

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




From peter at engcorp.com  Wed Apr 10 20:27:05 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:27:05 -0400
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx>
	 <mailman.1018391839.6535.python-list@python.org> <14lsJVAUgMt8Ewk9@jessikat.fsnet.co.uk>
Message-ID: <3CB4D859.4F795B1A@engcorp.com>

Robin Becker wrote:
> 
> >"import this", IIRC.
> ...
> Am I the only person that finds the whole concept of Easter eggs a slap
> in the face from the developers.

You're probably not the only one, but I do not share that view at all.

> When I find this rubbish in M$oft's expensive software I feel outraged.

I'm outraged by their unreliable software, expensive or with Easter
eggs or otherwise.  I have absolutely no difficulties with the idea
that the developers, in taking pride in their creation, implemented
an Easter egg.  (If it contributed to the unreliability that would
be a different story.)

> Is it better because it's open source? I think not. On the other hand
> perhaps the Python development process is itself a joke. When one
> questions the idiots in charge one ends in the kill file. No one expects
> the Spanish inquisition, but perhaps if rather trite philosophy is
> hidden in python and we have geniuses for developers, what else may be
> embedded therein?

In the case of Python's "import this", I was *very* pleased to find
it there, as it makes it very convenient to pull up that list 
without having to resort to a web browser (or even have one handy).

What, you mean you don't refer to that list frequently?  Why not?

;-)

-Peter


From shalehperry at attbi.com  Fri Apr 26 22:14:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 26 Apr 2002 19:14:12 -0700 (PDT)
Subject: looping through a list
In-Reply-To: <e2e9e807.0204261611.7514b387@posting.google.com>
Message-ID: <XFMail.20020426191412.shalehperry@attbi.com>

> 
> result = c.fetchone()
>   if (result != ""):
>     i=0
>     text = '<tr width = "80%" align = "center">'
>     while (i != len(result)+1):
>       i = i + 1
>       text = text + '<td align = "center" valign =
> "top">'+str(result[int(i)])  +'</td>'
>     text = text + "</tr>"
>     page.generate("","",sessionID,text)
>   else:
>     text = '<h4>No items found</h4>'
>     page.generate("","",sessionID,text)
> 
> 
> At the moment, it keeps saying list index out of bounds and I don't
> know where I'm going wrong. I've tried all sorts of things in the
> While statement and still can't get it going.
> 

In python you almost never want to use a counting loop.  The above is better
written as:

for piece in result:
    text = text + '<td align="center" valign="top"> + piece + '</td>'

the string append is slow, so this:

table_entries = []
for piece in result:
    tables_entries.append('<td align="center" valign="top">%s</td>' % piece)

text = string.join(tables_entries, '')

will work better still.  Although you may find the first version more readable.




From sjmachin at lexicon.net  Sat Apr 20 17:55:13 2002
From: sjmachin at lexicon.net (John Machin)
Date: 20 Apr 2002 14:55:13 -0700
Subject: How to make regexes faster? (Python v. OmniMark)
References: <3CBF71D4.5668CF1A@optonline.net> <3CBF9168.AAE0FC46@engcorp.com> <48F07A11B1D3ADCA.2691EAF64BEB7B96.D086A2379D2D43D4@lp.airnews.net> <c76ff6fc.0204191806.406ec53b@posting.google.com> <1019278636.737636@yasure>
Message-ID: <c76ff6fc.0204201355.21f4d4f6@posting.google.com>

"Donn Cave" <donn at drizzle.com> wrote in message news:<1019278636.737636 at yasure>...
> Quoth sjmachin at lexicon.net (John Machin):
> ...
> | Another practical perspective:
> |
> | When people mention excessive run-time to me, I ask them to rank the
> | severity of their problem on this scale:
> | 1. Job doesn't finish before they get back to their desk with a fresh
> | mug of coffee.
> | 2. Job has (exclusive access to database or excessive impact on
> | on-line response time) and doesn't finish by sun-up.
> | 3. Job doesn't finish before the weekly cold back-up is due.
> 
> How would you score a CGI program on a busy server in that scale?
> Or for that matter any program that runs on a busy server.  And
> I drink enough coffee, without refilling the cup every time I run
> some software that makes me wait.

The scale is for batch processes, and was tendered in context of the
OP's
comparison of two methods of performing a batch process, the longer of
which took 0.57 SECONDS.

> 
> I'm really delighted to hear about performance improvements in file
> object input in recent Python versions (in a message that would
> have been in this thread if Tim Peters used a newsreader instead
> posting to the list.)  Mostly because it's a good thing in itself,
> assuming the numbers are there, but also because it shows that
> someone must have felt that performance was important enough to
> work pretty hard on it - that's a healthy sign.

You can become delighted *earlier* and more comprehensively if,
instead of waiting for sporadic messages, you read the "What's New in
Python x.y" summary when a new version is released. In this particular
case:

http://www.amk.ca/python/2.1/index.html#SECTION0001500000000000000000


From quinn at hurl.ugcs.caltech.edu  Thu Apr  4 16:29:18 2002
From: quinn at hurl.ugcs.caltech.edu (Quinn Dunkan)
Date: 4 Apr 2002 21:29:18 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <3CAAC1FB.2050105@mxm.dk> <JGFq8.544000$pN4.38878349@bin8.nnrp.aus1.giganews.com>
Message-ID: <slrnaaphdc.39o.quinn@hurl.ugcs.caltech.edu>

On Wed, 03 Apr 2002 16:00:42 GMT, Terry Reedy <tejarex at yahoo.com> wrote:
>"Max M" <maxm at mxm.dk> wrote in message news:3CAAC1FB.2050105 at mxm.dk...
>> Through Zope I have used Structured text extensively. Also I have
>good
>> experience teaching stuctured text to my customers. After a very
>short
>> while they start to use it insted of html when generating content
>for
>> their sites due to ease of use.
>
>Just curious: are you willing to (consider, at least) switching to
>ReST, or are we going to end up with two similar standards used by
>different parts of the Python community?  IE, do Goodger's arguments
>about why the differences from ST are improvements cut any ice with
>you?

I'm not speaking for Max, but my answer is "no."

I agree with Goodger that a convention for docstrings would be nice, and that
structured text is the most reasonable candidate, but I don't agree that we
need a complicated extensible format for it.  I use paragraphs, lists, verbatim
code, and internal links (in the format I use, verbatim code goes in ''s and if
it happens to be the name of a documented symbol in the current module it gets
replaced with a link to the docs, so actually the format doesn't support links
at all, but the translator can insert them), and anything with many more
features, even if I personally don't use them, is likely to go over my "too
complicated" threshold.

A quick look at the python stdlib docs reveals that they could do just fine
with the above features, except the use of footnotes, which I'm happy without.

Why not take Bobo's StructuredText and simplify it to the essentials rather
than generalizing it to "the ultimate extensible structured text format"?


From pinard at iro.umontreal.ca  Fri Apr  5 20:08:15 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 05 Apr 2002 20:08:15 -0500
Subject: The Topy project
In-Reply-To: <7xit75wwcy.fsf@ruckus.brouhaha.com>
References: <mailman.1018039992.15896.python-list@python.org>
	<Xns91E7EF76BA86Acliechtigmxnet@62.2.16.82>
	<mailman.1018046786.6596.python-list@python.org>
	<7xit75wwcy.fsf@ruckus.brouhaha.com>
Message-ID: <oqwuvlocqo.fsf@carouge.sram.qc.ca>

[Paul Rubin]

> What do you do about mutable strings?

Nothing, besides raising a translation diagnostic over mutating procedures.
At run-time, an un-edited conversion aborts with a Python run-time error.

The goal of Topy is to help porting a Scheme program to Python, not to
produce an equivalent program.  My feeling is that Scheme strings are
best translated as Python strings while porting an application, this is
the most legible choice.  Some adjustment is required on the result by
the programmer, who ought to re-evaluate the algorithms on such points.

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





From philh at comuno.freeserve.co.uk  Wed Apr 10 18:28:13 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 10 Apr 2002 23:28:13 +0100
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <slrnab8v1o.emj.philh@comuno.freeserve.co.uk> <MPG.171e33893fed6df09896a7@news.nwlink.com>
Message-ID: <slrnab9f3t.q7.philh@comuno.freeserve.co.uk>

On Wed, 10 Apr 2002 12:34:01 -0700, Jeff Shannon <jeff at ccvcorp.com> wrote:
>In article <slrnab8v1o.emj.philh at comuno.freeserve.co.uk>, 
>philh at comuno.freeserve.co.uk says...
>> What have the PEPs ever done for us?
>> 
>> (Apart from the roads. The roads go without saying.)
>
>... and the Aqueduct...

... and reducing crime...

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 03:42:58 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 09:42:58 +0200
Subject: XPath?
References: <a9fg5u$19t$1@nets3.rz.RWTH-Aachen.DE> <m3wuv8iras.fsf@mira.informatik.hu-berlin.de> <a9fj34$3mi$1@nets3.rz.RWTH-Aachen.DE> <m3ofgkkugm.fsf@mira.informatik.hu-berlin.de>
Message-ID: <a9gkn6$nft$1@nets3.rz.RWTH-Aachen.DE>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3ofgkkugm.fsf at mira.informatik.hu-berlin.de...
> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
>
> > Yes, I've been trying to reach their site, but various links didn't
return
> > anything.
> > What do I need to do? I already installed PyXML.
>
> What do you want to do today?
>..

I want to port the QML Interpreter ( http://questml.com ) to Python for a
start -- if I find a working XPath handler.




From johnroth at ameritech.net  Mon Apr  1 19:33:25 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 1 Apr 2002 19:33:25 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com>
Message-ID: <uahuoa33sg6mfc@news.supernews.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CA8B4CC.88C35BB0 at alcyone.com...
> Stefan Schwarzer wrote:
>
>
> I don't see why you think the behavior (I presume you're referring to
> syntaxes like True == 2) is counterintuitive.  Why _would_ you expect
> this to do what you want?  It doesn't for any other Python types that
> I'm aware of.  == tests for equality, it doesn't do automatic type
> coercion.

Actually, "==" does whatever the class wants, if the class
has defined the appropriate magic method. Of course,
if you want other ordering functions to work that depend on
the "==" operator, then the class designer has to be reasonable
about what he has it do, but he doesn't have to.

And yes, he can include automatic type coercions if he wants to.

John Roth

>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.
> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.




From NO_SPAM_jmpurser2 at attbi.com  Mon Apr  1 10:58:43 2002
From: NO_SPAM_jmpurser2 at attbi.com (John Purser)
Date: Mon, 01 Apr 2002 15:58:43 GMT
Subject: Files, Seek, and Readlines. A newbie question
Message-ID: <Ts%p8.149254$ZR2.64157@rwcrnsc52.ops.asp.att.net>

Hello,

I'm using a large (>120,000 lines) sorted ASCII text file as an index.  I
started running into problems as soon as I began to use the .seek() to
position before a read.  Even when not opened as a binary file seek moves by
bytes not lines.  Is there a line oriented seek version?  Is there a better
way to get to line 50 than:
myFile.seek(0)
myFile.readlines(49)
Record = myFile.readline()
?

Thanks,

John Purser






From pete at dexterslab.de  Tue Apr  9 04:56:03 2002
From: pete at dexterslab.de (Henning Peters)
Date: Tue, 9 Apr 2002 10:56:03 +0200
Subject: exception question
Message-ID: <a8ua60$7nd$06$1@news.t-online.com>

hi NG,

i am asking me, if it is possible to catch an exception and to save the
exception value in a variable?!

try:
...
except: __all_exceptions__, error:
...

normally, the second argument of except takes the variable name. the problem
is, how i should get the variable, if i want to catch ALL exceptions. i
know, that this is not a fine way, but for my project (sending debug
messages over network) it is essential...

any idea?

bye,
henning





From skip at pobox.com  Tue Apr  2 12:11:34 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 2 Apr 2002 11:11:34 -0600
Subject: wxPython redefines string.letters?
In-Reply-To: <8351bb33.0204020854.272eaa7b@posting.google.com>
References: <8351bb33.0204020854.272eaa7b@posting.google.com>
Message-ID: <15529.58950.763933.194933@12-248-41-177.client.attbi.com>

    Jon> I realise this should go to the wxPython list, and after I've
    Jon> joined it will do ;-), but in the meantime:

    ...

Search google for wxpython and locale.  I believe the latest version fixes
this problem.  (It's getting funky with your locale settings.)

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)



From bokr at oz.net  Fri Apr 12 19:06:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Apr 2002 23:06:12 GMT
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com> <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no> <mailman.1018592665.14227.python-list@python.org> <3CB72CE6.DFAD5C10@python.org> <a97l29$hnv$0@216.39.172.122>
Message-ID: <a97p94$ros$0@216.39.172.122>

On 12 Apr 2002 21:54:17 GMT, bokr at oz.net (Bengt Richter) wrote:
[...]

>hasattr(c, '__getitem__') etc. can answer some of those questions,
>but if you had an official enumerated set of capabilities, you could
>have a __capabilities__ slot and check with a mask (e.g., against
>constants in __builtins__, e.g., INDEXING).
>
Sorry to follow myself up, but contrary to what I literally wrote, I don't
want to imply a literal integer mask for this, except maybe as part of hidden
implementation. That's too C-ish for a Python.

>For your class above people might write hascapability(c, INDEXING) or
>c.__capabilities__&INDEXING rather than type(c) is list (which fails)
Instead of __capabilities__ one could have a __hascapability__ method.

>or isinstance(c, list) which might be unnecessarily narrow for the
>generic capability (which a non-list-based custom object might support).
>
>BTW, A standard capabilities/protocol query method might be something to
>reserve a mask bit for, for custom extension of the concept.
A __hascapability__ method could serve that purpose.

Regards,
Bengt Richter


From duncan at NOSPAMrcp.co.uk  Mon Apr  8 07:20:41 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 8 Apr 2002 11:20:41 +0000 (UTC)
Subject: Nested scopes hitch
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl> <1018044224.815568@cswreg.cos.agilent.com>
Message-ID: <Xns91EA7D4BE5CABduncanrcpcouk@127.0.0.1>

weeks at vitus.scs.agilent.com (Greg Weeks) wrote in
news:1018044224.815568 at cswreg.cos.agilent.com: 

> Anyway, it increased visibility enough so that it is no longer
> possible to call a function and pass it the scope as two dictionaries,
> which I had found handy on occasion.

I don't think this is correct.
If your function accesses variables in a nested scope, then those variables 
are present in the dictionary locals().

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bokr at oz.net  Sun Apr 14 07:26:48 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Apr 2002 11:26:48 GMT
Subject: dos promt simulation
References: <6ac1ecad.0204130435.93d0ad3@posting.google.com> <Xns91EFBDC8F2FCcliechtigmxnet@62.2.16.82> <6ac1ecad.0204131339.56928640@posting.google.com> <a9aqr4$moh$0@216.39.172.122> <6ac1ecad.0204140019.54fff51@posting.google.com>
Message-ID: <a9bp1o$6tf$0@216.39.172.122>

On 14 Apr 2002 01:19:23 -0700, theivo at abv.bg (I.C.) wrote:

>bokr at oz.net (Bengt Richter) wrote in message news:<a9aqr4$moh$0 at 216.39.172.122>...
>> On 13 Apr 2002 14:39:58 -0700, theivo at abv.bg (I.C.) wrote:
>> 
>> >> there is also popen2, popen3 etc. they give you a file for stdin where you 
>> >> can write to.
>> >
>> >Maybe popen2\3 could help, but I think, I don't use them correctly :
>> >
>> >import os
>> >dos_input, dos_output = os.popen2('time')
>> >
>> >for line in dos_output.readlines(): # Current time is 22:14:43,24
>> >        print line                  # Enter new time:
>> >
>> >dos_input.write(raw_input())        # User enters the new time
>>   ^^^^^--??
>> >
>> >Where is my mistake?
>> >
>> I would guess that's probably one.
>> 
Sorry, that _is_ the child process stdin. I wasn't paying very good attention ;-/
After my comment, I felt obligated to try a few things, but could not get a volley going
between the interactive python and a python script executed via popen2('python rwrw.py').
I did find that writing all that the other script wanted as input into the pipe first made
it do its thing and produce everything back in one gob.

I tried writing to stdout and flushing in the other script, and still couldn't read anything
before supplying all its input, even though it suposedly wrote first. I tried to control the
buffer size to 1, but that didn't do anything either. It looks almost like you have to close
the pipe at the child end to have the output available to popen. Maybe sockets is a better idea
for communicating back and forth with another process. Mostly just have to decide on a message
delimiter, presumably.

I remember a large hassle getting pipes to work in C on windows, so it would not surprise me
to find limitations the way you can use them in windows popen2. It ought to be possible to do
something quite flexible under NT though. Just a lot of work to find the squirrely bits
of info you need. It wouldn't surprise me if some windows versions just fake pipes via hidden
temp files that have to be closed before they're opened by the recipient. At least it's acting
like that. But NT wouldn't have to, so I don't know what's happening. And I can't chase it now.

Regards,
Bengt Richter


From bryanjugglercryptographer at yahoo.com  Wed Apr 24 17:01:34 2002
From: bryanjugglercryptographer at yahoo.com (Bryan Olson)
Date: 24 Apr 2002 14:01:34 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com>
Message-ID: <1a517b5.0204241301.1e97c170@posting.google.com>

Paul Rubin wrote:
> I've put together an encryption function written in Python using the
> SHA module to provide a keystream in output feedback mode.  It's
> nowhere near as good as AES, but should be a big improvement over the
> rotor module.  It's at
> 
>   http://www.nightsong.com/phr/crypto/p2.py

Not so good.  The attacker only has to guess one block to break
all subsequent blocks.  For a demo, put the code below into 
the p2.py file

--Bryan


import string

def xor_strings(s1, s2):
    char_list = map(lambda x, y: chr(ord(x) ^ ord(y)), s1, s2)
    return string.join(char_list, "")


plain = ("This is Fixed Header"
        + "And this is the the real secret message.")


cipher = p2_encrypt(plain, "this is the key")
cipher = cipher[_ivlen:] # Trim off the IV

#  Figure out the first block based on knowing the
#  Fixed header.
seed = xor_strings(cipher[:20], "This is Fixed Header")


#   Now decrypt the message
for i in range(len(cipher) / 20):
    print xor_strings(seed, cipher[i * 20 : i * 20 + 20])
    seed = sha.new(seed).digest()


From pj at engr.sgi.com  Fri Apr 26 20:11:16 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 27 Apr 2002 00:11:16 GMT
Subject: Changing the repr() of a class?
References: <3cbf1a10_3@mk-nntp-1.news.uk.worldonline.com> <mailman.1019212237.13240.python-list@python.org> <NETv8.45693$eS3.10263@atlpnn01.usenetserver.com> <Dmjy8.66344$vF6.2029394@news2.tin.it>
Message-ID: <aacqb4$ujhso$3@fido.engr.sgi.com>

|> (as long as the rate of change does not
|> reduce the acceptance of the technology in the market)

Ah - so we should keep up, even accelerate the rate of
Python change.  Just don't do a very good job of documenting
it, so that the rate of change doesn't _appear_ excessive.

(just kidding, to what end isn't clear ...)

-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From cepl.m at neu.edu  Fri Apr 26 18:49:24 2002
From: cepl.m at neu.edu (Matej Cepl)
Date: Fri, 26 Apr 2002 18:49:24 -0400
Subject: iis: comparsion of two lists
Message-ID: <20020426224924.GA1318@rehor.surfbest.net>

Hi,

I am newbie in Python world and non-programmer (and totally
enjoying power of Python). I have begun to write a small script
for synchronization of two IMAP folders (it seems, there is
nothing like that all over the Net) and everything goes well (if
anybody wants to comments, he is very welcomed; unfinished draft
is attached), but for one operation. I would like to compare two
list of messages (in the local and remote folders) and create
list of messages, to be deleted/downloaded. In plain Python, the
part of deletion would go like this:

   erase = ()
   for msg in proc_old:
   		if not(msg in loc_old):
   			erase.append(msg)

Of course, it is crucial part of the script, so I would like to 
optimize it by using map() function. Could I do something like 
this?:

   map(lambda m: if not(msg in loc_old): erase.append(msg), proc_old)

Thanks for any comments

	Matej Cepl

-- 
Matej Cepl, cepl.m at neu.edu
138 Highland Ave. #10, Somerville, Ma 02143, (617) 623-1488
 
Opinions founded on prejudice are always sustained with the
greatest violence.
	-- Hebrew Proverb

-------------- next part --------------
#!/usr/bin/env python
# $Id: iis.py,v 0.6 2002/04/26 22:42:01 matej Exp $
"""
IMAP-IMAP Synchronization - iis

(always use lower case to avoid terrible misunderstanding :-)

Copyright (c) 2002 Matej Cepl <matej at ceplovi.cz>
   
Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following
conditions:

The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT.  IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
OTHER DEALINGS IN THE SOFTWARE.

"""
import ConfigParser
import imaplib
from sys import stderr, exit
from string import split, join
from os.path import expanduser

# constants
# return errorlevels
DOWNERR = 1
UPERR = 2
SELECTERR = 3
LISTERR = 4
LOGINERR = 5
CONFIGERR = 10

# other constants
IMAPPORT = "143"

def chatch_error(errmsg,stat_num):
	stderr.write("IMAP Error: %s !\n" % errmsg)
	exit(stat_num)

class Conifg(ConfigParser.ConfigParser):
	"""
	Object containing configuration.
	
	It reads configuration file (in WIN .ini syntax) like this
	
	[localuser1]
	local = password
	server1.com = username,password,last_sync (in IMAP shape)
	
	[localuser2]
	local = password
	server2.com = username,password,last_sync (in IMAP shape)
	"""
	def __init__(self,fname=".iisrc"):
		ConfigParser.ConfigParser.__ini__(self)
		self.filename = expanduser("~/"+fname)
		try:
			self.read(self.filename)
		except ParsingError,e:
			chatch_error(e,CONFIGERR)
		## read configuration file and fill appropriate data
		## structures
	def self.close()
		## write changed values to configuration file

def comp_lists(processed,local):
	"""
	Compare two lists of messages and select for download/delete.

	INPUT: two objects Server -- one to be processed and other
	OUTPUT: erase, download -- list of messages to delete from
			  the processed server, and list of messages to be
			  downloaded on other server
	What is newer than self.last and it isn't in other, should be
	downloaded from processed, and what is older and isn't on
	other, should be deleted. The same other way around.
	"""
	proc_old = processed.old_msg.keys()
	proc_new = processed.new_msg.keys()
	loc_old  = local.old_msg.keys()
	loc_new  = local.new_msg.keys()
	erase = ()
	download = ()
	map(lambda m: if not(msg in loc_old): erase.append(msg), proc_old)
	map(lambda m: if not(msg in loc_new): download.append(msg), proc_new)
	return erase,download

def sync(loc,rem):
	"""
	Synchronize two folders on servers.

	## INPUT:
	## OUTPUT:
	## description
	"""
	# delete old on the remote and download new messages
	delete,down = comp_lists(rem,loc)
	try:
		ok, answer = rem.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Remote messages were not deleted!\n")
		# rather let it be -- it is better not to have
		# something deleted, than to have deleted more than
		# we want
	for msg in down:
		try:
			stuff = rem.fetch(msg,'(RFC822)')
			# tohle musim nejak dodelat
			# loc.append(mailbox,flags,date,message)
		except Exception,e:
			catch_error(e,DOWNERR)
	
	# delete old on local and upload new ones
	delete,up = comp_lists(loc,rem)
	try:
		ok, answer = loc.store(join(delete), 'FLAGS', '(\Deleted)')
	except Exception,e:
		stderr.write("Local messages were not deleted!\n")
	for msg in up:
		try:
			stuff = loc.fetch(msg,'(RFC822)')
			# I have to add something here ... :-)
			# rem.append(mailbox,flags,date,message)
			# date - expected tuple according to library time
		except Exception,e:
			catch_error(e,UPERR)

class Server(imaplib.IMAP4):
	"""
	Main class for managing IMAP servers and folders.
	
	## some description
	"""
	def __init__(self,server,port=IMAPPORT,user,password,last_sync):
		"""
		Connect to the server and login.
		"""
		self.host=server
		self.port=port
		self.user=user
		self.passwd=password
		self.last=last_sync
		self.new_msg={}
		self.old_msg={}
		try:
			imaplib.IMAP4.__init__(self,self.host,self.port)
			self.login(self.user,self.passwd)
		except Exception,e:
			catch_error(e,UPERR)

	def message_lists(self,folder="")
		"""
		Fetch list of messages, that are older than self.last and
		newer than that.
		INPUT: correctly opened IMAP server, name of the folder,
		correctly set self.last (in IMAP representation).
		OUTPUT: sets self.new_msg and self.old_msg.
		"""
		if folder=="":
			folder="INBOX"
		try:
			self.select(folder)
		except Exception,e:
			catch_error(e,SELECTERR)
		try:
			ok, stuff = self.search(None,'( UNDELETED SINCE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					self.new_msg[msg]=uid,date
			ok, stuff = self.search(None,'( UNDELETED BEFORE ' + self.last + ' )')
			if ok == 'OK':
				for msg in split(stuff):
					ok, header = self.fetch(msg,'(UID)')
					if ok == 'OK':
						uid = split(header[0])[2][:-1]
					ok, header = self.fetch(msg,'(INTERNALDATE)')
					if ok == 'OK':
						temptup = split(header[0])[2:]
						tstr = temptup[0][1:] + temptup[1] + temptup[2][:-2]
						date = imaplib.Internaldate2tuple(tstr)
					self.old_msg[msg]=uid,date
		except Exception,e:
			catch_error(e,LISTERR)
		self.close()
	def complete(self):
		"""
		leave server
		"""
		self.logout()

if __name__ == '__main__':
	## General outline
	conf = Config()
	## for users:
	##   local = Server(...)
	##   local.message_lists()
	##   for servers:
	##      remote = Server(...)
	##      remote.message_lists()
	##      sync(local,remote)
	conf.close()

# vim: set ts=3:

From maxm at mxm.dk  Mon Apr 15 05:29:37 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 15 Apr 2002 11:29:37 +0200
Subject: split() off last substring
References: <mailman.1018860315.14961.python-list@python.org>
Message-ID: <3CBA9D81.50003@mxm.dk>

Noah wrote:

> Can I tell split() to split off the LAST seqment? I guess I want
> something like:
> 
> 	a="aaa.bbb.ccc"
> 	b,c=a.rsplit(1)
> 	print b,c
> 	# "aaa.bbb", "ccc"
> 

It would be esier to just::


 >>> hmm = "aaa.bbb.ccc"
 >>> hmm[hmm.rfind('.')+1:]
'ccc'


regards Max M



From pyth at devel.trillke.net  Mon Apr 29 15:58:29 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 29 Apr 2002 21:58:29 +0200
Subject: Third Draft:  Pep for Deprecating Builtins
In-Reply-To: <XFMail.20020429124231.shalehperry@attbi.com>
References: <aak1uq$bgt$1@bob.news.rcn.net> <XFMail.20020429124231.shalehperry@attbi.com>
Message-ID: <20020429195829.GP16217@devel.trillke>

On Mon, Apr 29, 2002 at 12:42:31PM -0700, Sean 'Shaleh' Perry wrote:
> 
> On 29-Apr-2002 Raymond Hettinger wrote:
> > From the comments, I've learned:
> > 
> > 1. map, filter, divmod, and pow may be immovable and should
> >     be left alone.
> > 
> > 2. reduce has fewer friends, but should probably also be
> >     left alone
> > 
> 
> thank you, thank you.  Gah, I could not imagine being forced to use list
> comprehensions everywhere.

This is unfair. The PEP didn't want to eliminate filter and friends (ff).
It just suggested to move it to a module which would be included
in the core python-distribution. You could then *always forever* do:

from funcs import filter, reduce, niftyotheroperator

and use the ff functions as before! And additionally the author
suggested that even *new* friends for filter should be introduced.
Neat!

I would  like to add that a user's perspective
is very important but should by no means the only  
measure. The developers do a great job IMO and of course
want to advance. They just happen to be very fast :-)

regards,

    holger




From peter at engcorp.com  Tue Apr 16 00:58:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 16 Apr 2002 00:58:27 -0400
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com>
Message-ID: <3CBBAF73.D1343917@engcorp.com>

Erik Max Francis wrote:
> 
> jepler at unpythonic.net wrote:
> 
> >     [2, 3, 5, 7, 9, ...]
> > would generate successive primes.
> 
> Pretty lousy code, since 9 is not prime.  :-)

No, it just generates some things that are not primes
as well.  You have to admit, though, it _would_ generate
all primes, given enough time... :-)


From aleax at aleax.it  Tue Apr 30 12:47:25 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 18:47:25 +0200
Subject: Third Draft: Pep for Deprecating Builtins
In-Reply-To: <20020430111828.C17488@unpythonic.net>
References: <mailman.1020129837.31112.python-list@python.org> <CEsz8.80366$vF6.2381303@news2.tin.it> <20020430111828.C17488@unpythonic.net>
Message-ID: <E172anG-0000RZ-00@mail.python.org>

On Tuesday 30 April 2002 06:18 pm, Jeff Epler wrote:
> On Tue, Apr 30, 2002 at 08:32:34AM +0000, Alex Martelli wrote:
> > So, types.ClassType COULD stay around "forever" from my POV, even long
> > after the default metaclass has become type -- at worst, the way to
> > enable the alternative metaclass as the default one would be 2 lines at
> > module scope:
> >
> > import types
> > __metaclass__ = types.ClassType
>
> Why not
>     from __past__ import old_style_classes

*THIS* one would be pretty easy to implement at almost no cost, if so
wished when and if the default __metaclass__ becomes the buit-in
'type'.  I don't particularly care for the style, but <shrug> -0.

> this would extend well to
>     from __past__ import integer_division
>     from __past__ import no_yield_token

These would be somewhat messier, requiring the interpreter to carry
around old code forever, I think (offhand I'm not sure what it would take
to implement the 'no_yield_token' one, actually).  No doubt the purpose
of "from __past__" is indeed to enable such further additions, which is
part of what makes me -0 on it.  But it IS still something of a <shrug>.


Alex




From donn at drizzle.com  Sat Apr 27 13:34:40 2002
From: donn at drizzle.com (Donn Cave)
Date: Sat, 27 Apr 2002 17:34:40 -0000
Subject: Stackless goes Limbo
References: <mailman.1019911009.16159.python-list@python.org> <aaee44$p73$1@panix1.panix.com>
Message-ID: <1019928879.371845@yasure>

Quoth aahz at pythoncraft.com (Aahz):
| In article <mailman.1019911009.16159.python-list at python.org>,
| Laura Creighton  <lac at strakt.com> wrote:
|>
|> For those of you who can't live without a Limbo introduction written
|> by a real C inventor:
|>
|> http://www.vitanuova.com/inferno/papers/limbo.html
|
| Ugh.  I've gotten so corrupted by Python I can't even read well-written
| C code.

Tsk, that ain't C.  As I read on, I was surprised to find out that the
sys->print notation doesn't imply sys is a pointer, rather that it's
a module.  (Don't know why not sys.print, then.)  In some places it
could remind you of Python - uses "self" more or less the same way,
has garbage collection that apparently supports timely finalization,
tuple unpacking assignment like x, y, z = t.  Also vaguely reminiscent
of ML or something.

The important thing to me is that the system looks like it would work
well for modular function libraries, thanks mostly to the garbage
collection storage model, and who knows, maybe with less function call
overhead.  The thing I miss here is exceptions - nothing like that?
Maybe I skimmed it too fast.

I think I like the channels.  Not sure I like the Unicode strings.
I wouldn't mind trying it out, if it ever escapes from Inferno.

	Donn Cave, donn at drizzle.com


From gerhard at bigfoot.de  Wed Apr 17 19:23:21 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 17 Apr 2002 23:23:21 GMT
Subject: mod_python
References: <e1e40a36.0204171358.7af66fdf@posting.google.com> <slrnabru6t.i6.gerhard@lilith.my-fqdn.de>
Message-ID: <slrnabs0v6.o3.gerhard@lilith.my-fqdn.de>

Gerhard H?ring wrote in comp.lang.python:
> daniela eilers wrote in comp.lang.python:
>> hi, experts
>> currently, i'm a python beginner.
>> normal python script run with apache, no probs.
>> trying to run mod_python under apache i get the following error message:
>> 
>> premature end of script headers. 
>> 
>> there is the code of my cgi script:
>> 
>> from mod_python import apache 
>>          def handler(req): 
>>               req.content_type = "text/html" 
>>               req.send_http_header() 
>>               req.write("Hello World") 
>>               return apache.OKimport apache 
>                        ^^^^^^^^^^^^^^^^^^^^^^
> 
> Looks like here's the error. This should be simply apache.OK. Apart
> from that, your example looks exactly like the one on  from
> 
> http://www.modpython.org/live/mod_python-2.7/doc/tut-what-it-do.html
> 
> and should be ok.

Except of course you should really fix the indentation of the handler
function, which should of course start at column 0. Otherwise, your
Python script will not even compile, which you could check by
importing it.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From phr-n2002a at nightsong.com  Wed Apr  3 03:01:22 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 00:01:22 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017795008.700.python-list@python.org> <7xzo0lth3r.fsf@ruckus.brouhaha.com> <mailman.1017817388.30724.python-list@python.org>
Message-ID: <7xd6xhyzvx.fsf@ruckus.brouhaha.com>

Richard Jones <rjones at ekit-inc.com> writes:
> > That looks pretty similar to an Info file (texinfo output).  I think
> > I'd rather type "@section{Overview}" than have to supply those special
> > underscores directly.
> 
> Yeah, I guess at this point it's down to personal preference. I think we can 
> summarise by saying that you're happy to have your documentation source in 
> the texinfo format [#]_, whereas the Structure Text camp would rather see it 
> more looking like actual documentation. Is that fair?

I'm not sure.  The PEP examines a bunch of different markup-based
alternatives rather than comparing markup vs. non-markup.  The problem
it finds with Javadoc source isn't that it's merely non-beautiful, but
that it's actually quite hard to read without practice and HTML
familiarity.  So I don't see it as a matter of a markup vs. a
non-markup camp.  If there were a markup system that was easier to
read directly, that might be a better answer than either Javadoc or
ReST.  I'm pretty comfortable with Texinfo, for example.

I wonder what other systems with inline docs, like Parc Smalltalk or
the MIT Lisp machine, use for syntax?  Emacs Lisp uses plain text with
some typographic conventions, but these don't go as far as ReST, and
it uses a separate system for structured manuals.

> > >    .. contents:: **Table of Contents**
> 
> I think it's justifiable in this case - it saves a lot of typing :)
> 
> Also, the other commonly used explicit markup is images::
> 
>   .. image:: graphic.png
> 
> ... that's pretty hard to do without an explicit markup (short of 
> ASCII-art'ing the PNG :)

You might also want to add a directive for enumerated lists, so that
if you insert a new item in the numbered list in your Roundup docs,
you don't have to manually renumber the rest of them.  And you might
want to add something for multi-column tables (useful for describing
functions and args, for example) so you don't have to draw them as
ascii art.  And so on.  You end up with a hybrid system, which isn't
necessarily so bad, but I don't think it's what you set out to do.

> > I looked briefly at the ReST documentation linked from the PEP, but
> > if I'm supposed to read it in detail to be able to use it, I might
> > as well use a more traditional markup language.
> 
> I've started writing a "primer" that should be the first port of
> call. It's a somewhat more verbose version of selected parts of the
> quick reference. Just the basics, explained in a little more
> detailed and human way than the formal documentation. I'd hope that
> that helps a little...

That will help, but I thought the idea of a PEP was to present a complete
specification of the proposed feature, which is why I didn't follow the
external references too carefully.

Btw, could those responding to me please just respond here in the
newsgroup-- I don't need the email carbon copies since I'm reading
here.  Thanks.


From guido at python.org  Tue Apr  2 16:45:10 2002
From: guido at python.org (Guido van Rossum)
Date: Tue, 02 Apr 2002 16:45:10 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Tue, 02 Apr 2002 20:49:41 +0200."
             <3CA9FD45.E6E33129@lemburg.com> 
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>  
            <3CA9FD45.E6E33129@lemburg.com> 
Message-ID: <200204022145.g32LjAK20525@odiug.zope.com>

> I haven't followed this thread for quite some time, but since the
> PEP still seems alive,

:-)

> let me add some experience I've had with
> using the already existing singletons (Py_True and Py_False)
> in Python for recognizing truth values.
> 
> Py_True and Py_False can be exposed in Python via
> 
> True = (1==1)
> False = (1!=1)

Yuck.  What a horrible implementation detail.

> and most comparison operations and quite a few other APIs
> returning truth values make use of these singletons.

This works purely by accident.  Given that at least Py_False is also
known as &_Py_ZeroStruct, I'm surprised that the code that optimizes
ints doesn't initialize small_ints[NSMALLNEGINTS] with Py_False.

> I thought it would be a good idea to use those two
> singletons for protocols which use booleans such as
> XML-RPC. My experiences with this approach are, well,
> not so good :-/
> 
> The reason is that True and False are really integers
> and not of a special new type. Now they are singletons,
> which is good, since they represent the only states
> a boolean can have and in many cases work reasonably
> well as boolean representative, but not always 
> (e.g. arithmetic operations such as True - True == False). 
> Also, when trying to recognize the two singletons you have 
> to use the "is" comparison -- "==" will fail to 
> differentiate between 1 and True...
> 
> def isboolean(x):
>     return x in (True, False)
> 
> ...doesn't work...
> 
> def isboolean(x):
>     return (x is True) or (x is False)
> 
> ..does.

The correct way if PEP 285 is accepted would of course be
isinstance(x, bool).

> As a conclusion, I think it would be better to make bool() a 
> new type which does *not* inherit from integers, but which 
> does know how deal with other types which are commonly
> used together with booleans such as integers. However, the
> type should implement boolean algebra and not try to
> mimic integer arithemtic, i.e. True - True raises an
> exception.

That would break more code.  E.g. (a!=0) + (b!=0) + (c!=0) counts how
many of (a, b, c) are nonzero; this would break.

> Py_True and Py_False should then be made singletons
> of this new type (possibly after a transition phase which
> replaces the singletons with a version that doesn't raise
> exceptions but instead issues warnings).

That's what the PEP does (but with a different type).

> This may sound painful at first, but in the long run,
> I believe, it'll result in the same benefits as other 
> painful changes have or will (e.g. the change from integer 
> division to floating point division).

I don't see the introduction of the bool type as painful at all.

I find getting it accepted painful though. :-(

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



From krissepu at vip.fi  Wed Apr 24 02:41:55 2002
From: krissepu at vip.fi (pekka niiranen)
Date: Wed, 24 Apr 2002 06:41:55 GMT
Subject: weak references and threads
Message-ID: <3CC653B1.9060701@vip.fi>

Could somebody explain me in SIMPLE english what are:

1)  weak references
2)  threads

and why or when are they needed.

Threads I have not used but the use of weak references
I have seen recommended with wxpython's grid component.

-pekka-



From python at rcn.com  Sun Apr 14 18:16:43 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sun, 14 Apr 2002 18:16:43 -0400
Subject: Deprecate input()
Message-ID: <a9cv63$qhu$1@bob.news.rcn.net>

The help and tutor maillists indicate that the use of input() rather than
raw_input() is widespread.  I think this function is so harmful and so
likely to be used that it should be deprecated hasta pronto (right away).
When really needed, eval(raw_input()) is easy enough to code and explicit
about what it does.

<no wink -- this function needs to go away>


Raymond Hettinger






From dalf at nospam.org  Fri Apr  5 12:05:51 2002
From: dalf at nospam.org (dalf)
Date: 5 Apr 2002 17:05:51 GMT
Subject: Theoretical threshold of improving Python's performance
References: <3CADA276.D91B7544@kfunigraz.ac.at>
Message-ID: <a8klhf$54a$1@ites.inria.fr>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
> I got a private email where I can read that Python can be never made as
> fast as Common Lisp due to Python's dynamic nature. May I ask  without
> any ulterior-motive: "Has the beformentioned proposition any value?".

> I know there were discussion concerning the topic; but I would be really
> interested in, from a theoretical point of few, whether it is a *fact*
> that Python can *never* become as fast as Common Lisp* or not.

  No.
  For static compilation (i.e. you completly compile the program before
running it), it is likely that Common Lisp is faster.

  But, thanks to Just-In-Time compilation methods, you are not bound
to static compilation ; then you could go arbitrary close to Lisp or
faster. It amounts to, among other things, what amount of type information
the JIT Python is able to guess. Example, if your code is doing 99% of its
processing on lists of size 4 with ints (and 1% with more arbitrary lists),
in the most ideal case, your JIT engine could detect this, generate code
using SIMD Pentium III+ assembly instructions, with vectors of 4 x 32 bits,
using a fast test and bailout for the 1% of the "arbitrary lists" cases.
  Not that it is easy, but it isn't impossible. Look how efficient are some
Java JIT.

.dalf


From tripps81 at yahoo.com  Wed Apr  3 19:20:09 2002
From: tripps81 at yahoo.com (Tripp Scott)
Date: Thu, 04 Apr 2002 07:20:09 +0700
Subject: Portable runtime
Message-ID: <5.1.0.14.2.20020404071247.04bb6210@202.95.131.154>

I was just wondereing whether there is a Python binding to APR, 
NSPR, or perhaps plans to sit CPython on top of one of these libraries?

With Python I can already do:
- portable path manipulation and file operation (os.path, open())
- portable threads (threading)
- portable sockets (socket)

But how do I deal with these?
- character set conversion
- natural order management of strings
- portable generation of UUID
- portable shared memory

Btw, does Python emulate threads on platforms where threads are not available?

t





From guido at python.org  Sun Apr  7 21:46:18 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 08 Apr 2002 01:46:18 GMT
Subject: help:problem using the RawPen class/turtle module.
References: <8374ee10.0204060303.3333260a@posting.google.com> <slrnaauvbv.3v.missive@whave.frontiernet.net> <8374ee10.0204070114.56260848@posting.google.com> <8374ee10.0204070523.3bbf801b@posting.google.com> <slrnab1e81.4b.missive@whave.frontiernet.net>
Message-ID: <3CB0F739.D69DFBD2@python.org>

Somebody wrote:
> 
> > I think I've found the problem,
> > it seems that the call to
> >
> > ...
> > self._arrow = _canvas.create_line(x-dx,y+dy,x,y,
> >                                   width=self._width,
> >                                   arrow="last",
> >                                   capstyle="round",
> >                                   fill=self._color)
> >
> > ...
> >
> > in the turtle.py ( _draw_turtle() method)
> > doesn't work because the _canvas to refer to is
> > the one in the 'self' instance,
> > so with the following change:
> >
> > self._arrow = self._canvas.create_line(x-dx,y+dy,x,y,....
> >               ^^^^^
> >
> > everything is ok.
> >
> > Without it the _canvas refers to a global that it
> > isn't been initialized to a Canvas when there is a
> > call to the RawPen class.
> 
Lee Harr wrote:
> 
> Nice work.
> 
> In the version I have, that line does not exist, but all of
> the calls to things like _canvas_create_*
> are called as self._canvas_create_*
> so you may have uncovered a bug.
> 
> I think the turtle module is not really used very much,
> though I think it is pretty cool.
> 
> Please file a bug report!

No need, this was already filed (python.org/sf/536117) and will be fixed
both in 2.2.1 and in 2.3 when it comes out.

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


From fperez528 at yahoo.com  Wed Apr 10 20:13:02 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 10 Apr 2002 18:13:02 -0600
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <a92kfi$dhn$1@peabody.colorado.edu>

Joel Bender wrote:

> I would like to process the elements of a list that match a condition,
> then remove them.  I'll use integers in my example, but my list items
> are actually objects.
> 
> What's the "right" way to do this?  For example:
> 
>     >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>     >>> dothese = [i for i in lst if i < 5]
>     >>> lst = [i for i in lst if i >= 5]
>     >>> for item in dothese:
>     ...     print item
> 

I don't know if it's 'right', but the following seems natural to me:

In [36]: def flt(x):
   ....:   if x<5:
   ....:     print x
   ....:     return 0
   ....:   else:
   ....:     return 1
   ....:

In [37]: lst
Out[37]:
[3, 2, 5, 1, 0, 7, 4, 8, 6, 9]

In [38]: filter(flt,lst)
3
2
1
0
4
Out[38]:
[5, 7, 8, 6, 9]

You can define the flt() function to suit your taste, add parameters to it, 
etc.

It's single pass, goes in the right order and returns the cleaned list. The 
only problem it has is that it generates a secondary list instead of 
modifying the existing one in-place. You're the only one who knows if that's 
a problem or not (it basically depends on the size of your lists and if you 
can afford the extra copy, even if just temporarily).

Cheers,

f.


From fredrik at pythonware.com  Wed Apr  3 09:17:47 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 3 Apr 2002 16:17:47 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
Message-ID: <016701c1db1a$569543e0$0900a8c0@spiff>

Mark McEahern wrote:

> So what's different?

more than one way to do it.  more than one concept to
explain to newcomers.  more than one "obviously correct"
way to do things.  less is more.  etc.

</F>




From dale at riverhall.NOTHANKS.co.uk  Tue Apr 23 12:25:03 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Tue, 23 Apr 2002 17:25:03 +0100
Subject: Encryption
References: <70h8cugniao40148h76fv4o7bgci2j3ddd@4ax.com> <7x662j76iu.fsf@ruckus.brouhaha.com> <359acu8n2qode4gnnmh38rkekp204dss3n@4ax.com> <7xn0vu4rma.fsf@ruckus.brouhaha.com>
Message-ID: <jm2bcu4jud8vrhhogffrfu94ch9l13gp23@4ax.com>

Paul Rubin <phr-n2002a at nightsong.com> wrote:

>If the data volume isn't too large and the security requirements
>aren't too intense, you can code RC4 in Python and get the other end
>to configure Arcfour encryption into Mcrypt, and do all the decryption
>in pure Python.  That's probably the easiest approach.  Otherwise you
>have to either get python-mcrypt working under Windows, or find some
>other library that works under Windows.
>
>I coded RC4 in Python a few months ago and it ran at about 100
>kbytes/sec on my P3-750, if that's of any help.  I don't know if
>that's fast enough for your application.  A C implementation would run
>at tens of megabytes/sec.

Thanks. Stuff to think about.
--
Dale Strickland-Clark
Riverhall Systems Ltd



From pyth at devel.trillke.net  Wed Apr 24 13:24:03 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 24 Apr 2002 19:24:03 +0200
Subject: python and very large data sets???
In-Reply-To: <aa6o7l$nbi$1@panix3.panix.com>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <aa6o7l$nbi$1@panix3.panix.com>
Message-ID: <20020424172403.GE20209@devel.trillke>

On Wed, Apr 24, 2002 at 12:58:29PM -0400, Aahz wrote:
> In article <ad381f5b.0204240841.52333a3e at posting.google.com>,
> Rad <zaka07 at hotmail.com> wrote:
> >
> >I still haven't received above mentioned files so I can't test the
> >time needed to (for example) read a 15GB "file1", filter by few
> >variables, and write a resulting subset as a "sub_file1".  Things
> >would afterwards get more complicated cause I will have to pullout
> >ID's from "sub_file1", remove duplicate ID's create
> >"no_dup_sub_file1", match those to ID's in remaining 3 main files and
> >pullout data linked with those ID's.
> >
> 
> Python *can* handle this kind of task, but you'll be much better off if
> you interface Python to a database. 

I disagree. Don't be so humble :-)
Using a database requires

- setting up/configuring an appropriate database
  for the task (may not be easy)

- getting the info from the files into a database
  (requires reading the file anyway!)

- reading from the database with partial result sets
  and convert again to a file.

i think this is errorprone and quite complex.
why not a more pythonic way like this:

- use module mmap to map source and destination files 
  into addressspace.

- write a method to parse a string that runs over your file
  does some substitutions and writes to a mmaped file.
  python is fast enough here i guess.

if mmap does not work good enough (on windows) you can
resort to read 100MB chunks of the file with file.seek 
or other methods.

it is important to know what your time constraints are.
10 minutes, an hour, a night?

	holger




From starx at pacbell.net  Mon Apr  1 20:28:44 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 2 Apr 2002 01:28:44 GMT
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com> <slrnaaftai.8mf.gerhard@lilith.my-fqdn.de> <w20q8.1572$0T6.38985@e3500-atl1.usenetserver.com> <3CA8EB72.F6E7973C@noaa.gov>
Message-ID: <Xns91E3B2015AB65RASXnewsDFE1@130.133.1.4>

Chris Barker || Mon 01 Apr 2002 03:21:29p:

> Robert Oschler wrote:
>> Also, will it coexists with Konqueror (which shells Netscape I think)
>> and Netscape?
> 
> NO, it doesn't. Konqueror is it's own beast, written from scratch, I
> think. It'a a pretty nice web brower, I use it most of the time.
> 
> You can have any number oif browsers on your Linux box, this isn't
> Monopoly style software...

It's a little confusing to answer that way. Yes they do coexist, as in you 
can install both of them at the same time and nothing bad will happen. No, 
they don't ... communicate? Neither does anything special because the 
other is there, or uses it in any way? (I think that's what you meant by 
your first no).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From ods at fep.ru  Mon Apr  8 08:47:56 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Mon, 8 Apr 2002 16:47:56 +0400 (MSD)
Subject: Modules are Singletons! (was Re: Catalog of Python Patterns? 
 [LONG!])
In-Reply-To: <slrnab32he.au.alf@sagittarius.logilab.fr>
Message-ID: <Pine.LNX.4.44L.0204081645230.13708-100000@ods.pp.ru>

On Mon, 8 Apr 2002, Alexandre Fayolle wrote:

AF> Stop me if I'm wrong, but this is actually not quite true.
AF> Suppose you have a package called 'p' containing modules
AF> m1 and m2. You also have another module called 'n' somewhere
AF> else. m2 is your wanabe singleton, and has a function
AF> called 'foo'.
AF>
AF> In m1, you can access m2 with a simple 'from m2 import foo',
AF> but in n, you have to 'from p.m2 import foo'. In this case,
AF> m2 is imported twice, and changes to global variables are
AF> not
AF> made.

You are not right:

>>> import sys
>>> filter(lambda m: m.find('m1')>=0 or m.find('m2')>=0, sys.modules.keys())
[]
>>> from p import m1
>>> filter(lambda m: m.find('m1')>=0 or m.find('m2')>=0, sys.modules.keys())
['p.m1', 'p.m2']
>>> from p import m2
>>> filter(lambda m: m.find('m1')>=0 or m.find('m2')>=0, sys.modules.keys())
['p.m1', 'p.m2']





From loewis at informatik.hu-berlin.de  Sat Apr  6 05:40:51 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Apr 2002 12:40:51 +0200
Subject: Selecting cipher suites with socket.ssl
References: <84cff98e.0204041830.7c0d0171@posting.google.com> <mailman.1017978613.9042.python-list@python.org> <m3y9g21uea.fsf@mira.informatik.hu-berlin.de> <200204051346.g35Dkerq000532@ratthing-b246.strakt.com> <m3it769ee9.fsf@mira.informatik.hu-berlin.de> <mailman.1018087333.20495.python-list@python.org>
Message-ID: <j4k7rl2jpo.fsf@informatik.hu-berlin.de>

Laura Creighton <lac at strakt.com> writes:

> But I don't like the assumption that the OP can do certificate generation
> better than Martin Sj?gren did.  

I didn't suggest that. Instead, I suggested that the OP changes the
Python SSL support - something that Martin Sj?gren didn't.

Also, the OP was only asking for selecting cipher suites.

I was mainly objecting the assertion "It's broken." I don't think that
is true - it is mainly incomplete.

> So naturally, there is no big push here to get pyOpenSSL in the standard
> library.  But I don't think that the OP should waste his time hacking
> on the standard SSL module; pick something that other people want to see.

I don't think this effort would be wasted. He only needs a relatively
small feature; if he implements that, all Python 2.3 users would have
it readily available.

> Besides, whatever he comes up with may be rejected on the grounds
> that the code isn't good enough, or it adds the sort of complexity
> you don't want in the standard library.  Modifying an existing
> library shouldn't be by definition easier to do than adding a new
> one -- foul code is foul code

I disagree that the current SSL support is foul code.

Regards,
Martin


From see at below  Sat Apr 27 08:36:09 2002
From: see at below (Paul Foley)
Date: 28 Apr 2002 00:36:09 +1200
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org>
Message-ID: <m2znzpwc7a.fsf@mycroft.actrix.gen.nz>

On Thu, 25 Apr 2002 22:26:30 -0700, James J Besemer wrote:

> Paul Foley wrote:

>> No I don't have to admit that...I don't even have to admit that there
>> are any parentheses in Lisp :-)

> What color is the sky in your world?  ;o)

> Seriously, I agree, the underlying abstraction -- the semantics -- does not require
> parentheses.  But that's not really the language, per se.

I think you're missing the point: Lisp is not defined in terms of
text; instead, it's the objects that result from calling READ on that
text.  There are no parentheses in that result; there are only lists,
symbols, strings, integers, etc., etc.  That's why you can define
things like the infix syntax reader Jacek Generowicz mentioned.

READ can call arbitrary user-defined functions to parse the textual
representation into internal objects, so it can read any surface
syntax you want.

>> In any case, parentheses are not used around sub-expressions, or to
>> separate or group statements.  They're only used (in evaluated
>> contexts) for "function" calls, and most other languages, including
>> Python, use parentheses for that purpose, too:
>> 
>> Python: foo(x)
>> Lisp: (foo x)
>> 
>> Same number of parentheses, right?

> Besides, this is an entirely contrived, intellectually dishonest example.

> First off, parens ARE used around sub expressions last I checked, unless I don't get how
> you define sub expression.  E.g.,

>     (foo (+ x (/ 1 n)))

> vs.

>     foo( x + 1 / n ).

+ and / are just ordinary functions, in Lisp.  If they were ordinary
functions in Python you'd have to write

      foo(+(x,/(1,n)))

which, again, has the same number of parentheses.  Functions are the
general case -- there are only a handful of infix operators in general
use, and you don't get to define new ones in Python anyway -- so I
don't think I'm being "intellectually dishonest".

When you talk about parenthesising subexpressions, I think of
something like

  x + (1 / n)

requiring the parentheses; i.e., having to write

  (+ x ((/ 1 n)))  or perhaps even  (+ (x) ((/ (1) (n))))

in Lisp (which aren't even legal in Lisp: you /can't/ parenthesise
subexpressions!)

-- 
In modern physics, the more logical you are, the more wrong you are.
                                                           -- Lama Govinda
(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From tejarex at yahoo.com  Wed Apr 10 11:03:26 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 10 Apr 2002 15:03:26 GMT
Subject: multiple ranges to for statement?
References: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu> <mailman.1018334719.16002.python-list@python.org> <3CB3DDB9.E0712D39@cosc.canterbury.ac.nz>
Message-ID: <2vYs8.299022$2q2.27543450@bin4.nnrp.aus1.giganews.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3CB3DDB9.E0712D39 at cosc.canterbury.ac.nz...
> Ian Bicking wrote:
> > for i in range(a, b) + range(c, d):

> Alternatively:
>
>   for seq in range(a, b), range(c, d):
>     for i in seq:
>       ...
> which avoids building two copies of the sequences.

Nice idiom ("Why didn't I thing of that!"): it scales to k subranges
much better.
To also avoid building all at once (and retyping 'range'), factor
further:

for lo,hi in (a,b),(c,d),(e,f):
  for i in range(lo,hi):
    ...

In 2.2+, one could also write a multirange() generator that would
interprete a string like '1-3,5,8-15,37,88,101-133'.  (In 2.1-, use
class instead.)

Terry J. Reedy





From martin at v.loewis.de  Sun Apr  7 01:32:27 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 07 Apr 2002 08:32:27 +0200
Subject: 2.2.1c2 test_fcntl hangs
References: <lysn69qrp7.fsf@bock.chem.unc.edu> <m3r8lsfq1x.fsf@mira.informatik.hu-berlin.de> <ly1ydspj0u.fsf@bock.chem.unc.edu> <slrnaauun4.tl.gerhard@lilith.my-fqdn.de>
Message-ID: <m3it74c938.fsf@mira.informatik.hu-berlin.de>

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

> AFAIK some filesystems have problems when being exported via NFS.
> (Older) reiserfs come to mind.

I share the same theory. I also found problems with locking over NFS
in test_fcntl - but it turned out that other programs trying to use
file locking on the same server blocked indefinitely as well.

If you merely want to run the test suite, you should disable
test_fcntl.

Regards,
Martin



From cliechti at gmx.net  Tue Apr 30 15:42:45 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Apr 2002 21:42:45 +0200
Subject: import from a file in a subdirectory
References: <644f6688.0204301127.5d3372d@posting.google.com>
Message-ID: <Xns9200DD22B5809cliechtigmxnet@62.2.16.82>

rdacker at pacbell.net (rdack) wrote in 
news:644f6688.0204301127.5d3372d at posting.google.com:
> i have a script running in a directory.
> i want to call a subroutine in a file in a subdirectory.
> 
> in the first file i said  from subdir/ff.py import  subrtn
> python doesn't like the slash.
> 
> how do i do it?
 
this is simple:
>>> sys.path.append(os.path.abspath("subdir"))
>>> from ff import subrtn
the abspath is only needed when your programm changes the current
directory.


".py" is never needed when writing "import".

a package can be a subdirectory and then you can use
>>> import package.submodule
but then the "package" directory needs an "__init__.py"

HTH
chris

-- 
Chris <cliechti at gmx.net>



From tdi01fhu at syd.kth.se  Wed Apr 17 10:08:50 2002
From: tdi01fhu at syd.kth.se (FREDRIK HULDTGREN)
Date: Wed, 17 Apr 2002 16:08:50 +0200
Subject: Div crypto stuff
Message-ID: <scbd9e21.061@Riksvasa.syd.kth.se>

>From:   Michael Gilfix <mgilfix at eecs.tufts.edu> 
>Date:   Wednesday - April 17, 2002 8:17 AM 
>Subject:   Pure python crypto 
>   
> Would anyone have any use for a pure python implementation of
>Blowfish? Or any other crypto algorithms for that matter.  As a matter
>of interest, I found myself implementing the blowfish module recently
>but was unsure if any audience would be interested. No C compiler
>is required, which is nice. Could use some testing though.
>             -- Mike 

>From:   Fr?d?ric Laurent <liste_nospam_ at opikanoba.org> 
>Subject:   amkCrypto installation on win32 
   
>Hello,
>I want to use RSA encoding/decoding, so I downloaded amkCrypto 0.1.3.
>I use python 2.2 on windows XP.

>Must I have to download and install swig in order to make amkCrypto work
>(for the mxCrypto module) ?
>I don't have any C++ compiler installed, and it seems to be quite complicated >:
>amkCrypto -> mxCrypto -> swig -> c++ and so on...
>moreover the link http://starship.skyport.net/crew/amk/python/crypto.html, >given
>in the documentation, on the pycrypt package is broken...>>

>So, I just want to generate a pblic/private key, crypt and decrypt a text
>buffer... but it sounds to be not so simple to do this !>

>Did I miss something ???

>thanks in advance

>--
>Fr?d?ric Laurent
>http://www.opikanoba.org  


These type of questions are more likely to be answered on the Python-Crypto mailing list.

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

/crast




From aleax at aleax.it  Wed Apr 17 10:01:17 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 17 Apr 2002 14:01:17 GMT
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <Nefv8.30383$vF6.948633@news2.tin.it>

dominikush at yahoo.com wrote:

> Hello,
> 
> one thing I like very much about Python is that statements
> work like you would expect them to work. Take for example
> the use of dict.values() for dictionaries: If you store the
> result of dict.values(), and change the dictionary after-
> wards, the previously stored result remains untouched.

The .values() method of a dictionary is defined to return
a new list of the values.  That's more or less inevitable, 
since a dictionary doesn't _have_ a list of its value 
normally, so it must build it on the fly when you ask for it.
It's not a copy -- it's a new list object.

However, Python does NOT copy except in situations where
a copy is specifically defined to happen.  The .values()
method being in a vague sense such a situation, as mentioned...
a new object, rather than a copy of any existing one.

In general, whenever possible, Python returns references 
to the same objects it already had around, rather than 
copying; if you DO want a copy you ask for it -- see module 
copy if you want to do so in a general way.  Of course,
building new objects is a different case.

If this is counteintuitive, so be it -- there is really
no alternative in the general case without imposing huge
overhead, making copies of everything "just in case".
MUCH better to get copies only on explicit request (and
new objects, when there's no existing object that could
either be copied or referred-to).

Of course there are in-between cases -- such as slices.

The standard sequences give you a new object when you 
ask for a slice; this only matters for lists (for immutable
objects you shouldn't care if you get copies or what).
A list is not able to "share a part of itself", so when
asked for a slice it gives out a copy, a new list (for
generality, of course, it then also does when asked for
a slice-of-everything, thelist[:] -- so in that limit case
the new object can be seen as a copy of the existing one).

The justly popular Numeric package, on the other hand,
defines an array type which IS able to share some or all 
data among several array objects -- so a slice of a Numeric 
array does share data with the array it's sliced from.  It's
a new object, mind you:

>>> import Numeric
>>> a=Numeric.array(range(6))
>>> b=a[:]
>>> id(a)
136052568
>>> id(b)
136052728
>>>

but the two distinct objects a and b do share data:

>>> a
array([0, 1, 2, 3, 4, 5])
>>> b
array([0, 1, 2, 3, 4, 5])
>>> a[3]=23
>>> b
array([ 0,  1,  2, 23,  4,  5])
>>>


Each behavior has excellent pragmatics behind it -- lists 
are _way_ simpler by not having to worry about data sharing, 
arrays have different use cases by far -- but it's hard to 
be unsurprising when two somewhat similar objects differ 
in such details.

But all of the copies which do "happen", e.g. by the
limit case of list slicing or whatever else (with ONE 
exception of which more later) are always SHALLOW copies.

NEVER does Python embark on the HUGE task of _deep_ copying 
unless you very specifically ask it to -- specifically with 
function deepcopy of module copy.  DEEP copying is a serious 
matter -- function deepcopy has to watch out for cycles, 
reproduce any identity of references, potentially follow 
references to any depth, recursively -- it has to reproduce
faithfully a graph of objects referencing each other
with unbounded complexity.  It works, but of course it
can never be as fast as the mundane business of shallow
copying (which in turn is never as fast as just handing
out one more reference to an existing object, whenever
the latter course of action is feasible).


So, that's what has apparently snagged you here:

> However, if a dictionary has lists as value entries, I get
> a counterintuitive behavior (which, just recently, broke my
> code): If you change the dict, the list you previously
> created via dict.values() gets automagically updated. A nice
> feature, but nothing I would have expected!

Not really -- if you change _objects to which the dict refers_
(rather than changing the dict in se), then other references
to just-the-same-objects remain references to just the same
objects -- if the objects mutate, you see the mutated objects
from whatever references to them you may be using.


>>>> dict = {'a':[1],'b':[2]}
>>>> list = dict.values()
>>>> list
> [[1], [2]]

Don't use the names of built-in types as variables: you WILL
be burned one day if you do this.  dict, list, str, tuple, file,
int, long, float, unicode... do NOT use these identifiers for 
your own purposes, tempting though they may be (an "attractive
nuisance", to be sure).  If you don't get into the habit of
avoiding them, one day you'll be trying to (e.g.) build a
list with x=list('ciao') and get puzzling errors... because
you have rebound identifier 'list' to refer to a certain list
object rather than to the list type itself.

Use alist, somedict, myfile, whatever... nothing to do with
your problem here, just some other simple advice!-)


>>>> dict['a'].append(3)

This does not "change the dictionary" -- the dictionary object
still contains exactly the same references, to objects with
the same id's (two string objects, the keys, and two list
objects, the values).  You're changing (mutating) one of those
objects, but that's quite another issue.  You could be
mutating said list object through any reference to it
whatsoever, e.g.:

>>> alist=list('ciao')
>>> adict={'a':alist}
>>> adict
{'a': ['c', 'i', 'a', 'o']}
>>> alist.pop()
'o'
>>> adict
{'a': ['c', 'i', 'a']}
>>>

If you wanted dictionary adict to refer to a COPY (a "snapshot",
if you will) of the contents of alist, you could have said so:

>>> import copy
>>> alist=list('ciao')
>>> adict={'a':copy.copy(alist)}
>>> adict
{'a': ['c', 'i', 'a', 'o']}
>>> alist.pop()
'o'
>>> adict
{'a': ['c', 'i', 'a', 'o']}
>>>

and then the dictionary object's string-representation would
be isolated from whatever changes to the list to which name
alist refers.  The string representation delegates part of its
job to the objects to which the dictionary object refers, so,
if you want to isolate it, you do need copies -- maybe deep
ones, in fact (<shudder>... well no, not really, but...:-).


>>>> dict
> {'a': [1, 3], 'b': [2]}
>>>> list
> [[1, 3], [2]]
> 
> Looks like that in the first case a copy is returned while
> in the latter case list references are returned. Ok, but

Nope.  ALWAYS references.  .values() doesn't return a reference
to an existing object NOR a copy of an existing object, because
there's no "existing object" in this case -- so it always
returns a NEW object, suitably built as per its specs.

> according to Python's philosophy I shouldn't mind if I work
> with lists in the dictionary or anything else. If the
> behavior depends on the knowledge of the type of values I
> put into a dictionary, I find that somehow counterintuitive.

There is no such dependence.  Just a huge difference
between changing an object, and changing (mutating) some
OTHER object to which the first refers.

In Bologna over 100 years ago we had a statue of a local hero
depicted pointing forwards with his finger -- presumably to
the future, but given where exactly it was placed, the locals
soon identified it as "the statue that points to Hotel
Belfiore".  The one day some enterprising developer bought
the hotel's building and restructured it -- in particular,
where the hotel used to be was now a restaurant, Da Carlo.

So, "the statue that points to Hotel Belfiore" had suddenly
become "the statue that points to Da Carlo"...!  Amazing
isn't it?  Considering that marble isn't very fluid and the
statue had not been moved or disturbed in any way...?

This is a real anecdote, by the way (except that I'm not
sure of the names of the hotel and restaurant involved --
I could be wrong on those), but I think it can still help
here.  The dictionary, or statue, has not changed at all,
even though the objects it refers/points to may have been
mutated beyond recognition, and the name people know it
by (the dictionary's string-representation) may therefore
change.  That name or representation was and is referring
to a non-intrinsic, non-persistent, "happenstance"
characteristic of the statue, or dictionary...


> Who is wrong here: my intuition or Python (2.2)? If it's
> my intuition, how can I train my thinking about Python's
> execution model, so that my intuition get's better ;-)

Your intuition, which led you astray here (Python does just
what it should do), can be trained in several ways.  The
works of J. L. Borges and I. Calvino, if you like fiction
that's reasonably sophisticated but still quite pleasant,
are good bets.  If you like non-fiction written by
engineers fighting hard to dispell some of the errors of
philosophers, Wittgenstein and Korzibsky are excellent.

I'm not kidding, but I realize that many Pythonistas don't
really care for either genre.  In which case, this group
and its archives, essays by GvR and /F, and the Python
sources, may also prove interesting reading.


Alex




From phd at phd.pp.ru  Wed Apr 17 16:30:45 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 18 Apr 2002 00:30:45 +0400
Subject: Installation problem
In-Reply-To: <5.1.0.14.2.20020417144022.00b55a48@postoffice4.mail.cornell.edu>; from eje4@cornell.edu on Wed, Apr 17, 2002 at 02:47:24PM -0400
References: <5.1.0.14.2.20020417144022.00b55a48@postoffice4.mail.cornell.edu>
Message-ID: <20020418003045.D4068@phd.pp.ru>

On Wed, Apr 17, 2002 at 02:47:24PM -0400, Eric Evans wrote:
> I'm having a problem trying to install Python 2.2.1 and I'm hoping someone 
> out there may be able to offer some helpful advice.  I'm on Solaris 8 with 

   It seems Python 2.2.1 has got a lot of compilation problems on different
versions of Solaris :(

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




From aahz at pythoncraft.com  Mon Apr 22 14:13:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Apr 2002 14:13:52 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com> <3CC449BF.8030807@alcyone.com>
Message-ID: <aa1jt0$hgs$1@panix1.panix.com>

In article <3CC449BF.8030807 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>
>Here's a Web page outlining some of the differences between Scheme and 
>Common Lisp:
>
>	http://dept-info.labri.u-bordeaux.fr/~strandh/Teaching/MT

404
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From starx at pacbell.net  Thu Apr 18 00:32:55 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 18 Apr 2002 04:32:55 GMT
Subject: list_a-list_b
References: <mailman.1019022993.8455.python-list@python.org> <Xns91F32263AA919RASXnewsDFE1@130.133.1.4> <a9klfs$n5t$0@216.39.172.122>
Message-ID: <Xns91F3DB614E822RASXnewsDFE1@130.133.1.4>

Bengt Richter || Wed 17 Apr 2002 01:21:16p:

> On 17 Apr 2002 10:21:46 GMT, Philip Swartzleonard <starx at pacbell.net>
> wrote: [...]
>>Tim Peters || Tue 16 Apr 2002 10:54:56p:
> [...]
>>> 
>>> though.  Contrary to popular belief, there's no Fabulous Prize
>>> awaiting those who minimize either vertical or horizontal space
>>> <wink>. 
>>
>>Minimizing relative vertical space is great; that's why list 
>>comprehensions are better than nested for loops for expressing simple 
>>ideas ('i want var to be an x by y matrix of foozles' =), and why
>>moving stuff into subfunctions is nice. Horizontal space though...
>>feh. Unless you're using emacs default python mode, in a two or three
>>level deep looping structre in a function in a class, and you need to
>>write 'self.data[x][y].variable = self.long_function_name( lots, of,
>>arguments )' ... =)
> 
> If you want the form but not the looong one-liner:
> 
>     self.data[x][y].variable = (
>         self.long_function_name(
>             lots,
>             of,
>             arguments 
>         )
>     )
> 
> Hm, this makes me wonder if '=' could introduce an optionally indented
> expression block, using ident/dedent where the above is using = ( .. )

Well, the version of python-mode i've go t'would still put that as 

    self.data[x][y].variable = (
        self.long_function_name(
        lots,
        of,
        arguments 
        )
        )

Which is better than i thought it would be:

    self.data[x][y].variable = (
                                self.long_function_name(
                                                        lots,
                                                        of,
                                                        arguments 
                                                        )
                               )

But I still need to change it sometime =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From tomchance at gmx.net  Tue Apr 30 10:37:35 2002
From: tomchance at gmx.net (Tom Chance)
Date: Tue, 30 Apr 2002 14:37:35 +0000
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com>
Message-ID: <3CCEAC2F.10802@gmx.net>

I started off by learning Perl a few years ago, and I'm now starting to 
learn python. I've found that learning both is extremely valuable, both 
because learning python has really strengthened and deepened my 
understanding of perl, and because they both perform very different 
functions. Perl excels at things like administration scripts and 
programs, web pages and anything requiring complex text processing. 
Python excels at developing full programs, and prototypes for C/C++ 
programs. If you can put both under your belt, you'd be a lot better off 
for it.

As for which first, well that depends slightly on your priorities, but 
if you have none, I'd learn Perl first, because the syntax will be more 
familiar to you, and then when you've got past the more advanced topics 
like references and writing perl modules, move onto python. O'Reilly's 
"Learning Python" would be an excellent introduction for you, as it 
constantly compares Python to C/C++.

Tom


David wrote:
> I have a lot of experience in ANSI C/C++, I want to learn a new
> language and which it can help me in the work.
> I just want to clear which(python or perl) is more comfort for me?



From joost_jacob at hotmail.com  Thu Apr 25 17:36:06 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 25 Apr 2002 14:36:06 -0700
Subject: console redirection to Tkinter
References: <3CC34237.CE6F21F1@club-internet.fr>
Message-ID: <13285ea2.0204251336.3346e853@posting.google.com>

Argh, pasted the wrong buffer, really !  The line with fout.readline()
should have fin instead of fout.
Here is the correct example:


from Tkinter import *
import popen2

cmd = "echo 'hello world'"
fin, fout, ferr = popen2.popen3(cmd)
root = Tk()
t = Text(root)
t.insert('1.0', fin.readline()) 
t.pack()
root.mainloop()


dutch-people-would-spot-the-error-immediately-yrs


From zen at shangri-la.dropbear.id.au  Sat Apr  6 22:53:55 2002
From: zen at shangri-la.dropbear.id.au (Stuart Bishop)
Date: Sun, 7 Apr 2002 13:53:55 +1000
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204061839.g36Idfdj017724@ratthing-b246.strakt.com>
Message-ID: <153DF589-49DB-11D6-9BF0-000393031882@shangri-la.dropbear.id.au>

On Sunday, April 7, 2002, at 04:39  AM, Laura Creighton wrote:

> Mark McEahern writes:
>> [Laura Creighton]
>>> [...]  I am rejecting the existence of booleans in the language
>>> because it is too hard to teach how to use them properly.
>>
>> How do you use them properly?
>>
>> Suppose I want to validate users.  I might write something like this:
>>
>> 	def isValid(user, password):
>> 		if ...:
>> 			return True
>> 		else:
>> 			return False
>
> My pleasure, especially since this is such a cool example.  You have
> just shot yourself in the foot, and do not know it.  When you have
> shipped your isValid function, across the world to your 10,000 clients,
> you will discover this.
>
> Your boss comes in and says, oh, Mark, we need a change to that
> isValid function.  We need to return 3 states, Valid, Invalid, and
> Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
>
> What do you do now?

You correct them. isValid is returning a perfectly valid
boolean state - either the username, password combination
is valid or it isn't. This makes this function useful.
At some point, my application needs to make a decision
'can I log in or not', and it has to get a boolean answer.

If I make my isValid function return a state rather than
a boolean, then every single piece of code that uses
isValid will then need to encode the business logic about
what to do if it gets these states. Everytime I add a new
state ('Valid-but-only-on-weekends','Valid-if-they-are-on-campus'),
every bit of code that calls isValid now needs to be updated with
the new business logic. By changing the possible return values,
the API has been changed

If my isValid function contains the business logic and can only
return True/False, then I can update it whenever it is needed
and not need to update every single call of this function to
handle the new state.

> case.  Don't artificially limit yourself to 2 values because your
> vision at the time only extended to 2 values.  The one thing constant
> in this world is change.   If you don't need a third value, somebody
> else will, later.  Integers are your friends.

I constantly see functions that should return boolean. Looking
at the builtin functions page of the Python library for example,
hasattr, callable, isinstance and issubclass should all return
a boolean. Saying they return 0,1 and maybe something additional
in the future makes my code unnecessarily complex:

if callable(foo) == 1:
	result = foo()
elif callable(foo) == 0:
	result = foo
else:
	raise RuntimeError('Oops! I havn't a clue what is happening!')

--
Stuart Bishop <zen at shangri-la.dropbear.id.au>
http://shangri-la.dropbear.id.au/





From stephen at xemacs.org  Sun Apr 28 22:40:05 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 29 Apr 2002 11:40:05 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org>
Message-ID: <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:

    Laura> People here seem to be thinking that the GPL is there to
    Laura> protect creators rights to their software.

I'm pretty sure that that is what rms has in mind, as long as both
"creators" and "software" are inclusive collective nouns.

The consumer protection argument is the OSI line (a special case
thereof), and rms don't have much truck with anything so economical
and relativist.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


From bvdpoel at uniserve.com  Sun Apr 14 22:58:19 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Sun, 14 Apr 2002 19:58:19 -0700
Subject: split() off last substring
Message-ID: <3CBA41CB.27DCA084@uniserve.com>

Can I tell split() to split off the LAST seqment? I guess I want
something like:

	a="aaa.bbb.ccc"
	b,c=a.rsplit(1)
	print b,c
	# "aaa.bbb", "ccc"

Thx.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From s.schwarzer at ndh.net  Mon Apr 22 06:14:57 2002
From: s.schwarzer at ndh.net (Stefan Schwarzer)
Date: Mon, 22 Apr 2002 12:14:57 +0200
Subject: DB-API 2.0
References: <aa0lig$asv$1@news.tpi.pl>
Message-ID: <3CC3E2A1.8691283A@ndh.net>

Hello Piotr

"Piotr Czaba?ski" wrote:
> 1.
>     Fetch all (remaining) rows of a query result, returning them as a
> sequence of sequences (e.g. a list of tuples).

If I understand correctly you would use

conn = ...
cursor = conn.cursor()
cursor.execute(...)
...
more_items = cursor.fetchall()

>     J think the proper result would be a list of lists - specilally for
> people working with wxPython.

I consider it sensible that the inner structure is a tuple because it
represents a "row" from a query that should probably not be modified
(modifying would seem unintuitive to me).

Stefan


From philh at comuno.freeserve.co.uk  Tue Apr 16 18:24:50 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:24:50 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <3CBB76CA.A423AAD8@alcyone.com>
Message-ID: <slrnabp95i.ipr.philh@comuno.freeserve.co.uk>

On Mon, 15 Apr 2002 17:56:42 -0700, Erik Max Francis <max at alcyone.com> wrote:
>phil hunt wrote:
>
>> OK, I've decided to write a PEP. Preliminary to this, I'd like
>> people's comments.
>> 
>> I wish to make the pas startement optional. That is, in any Python
>> program where there is a pass it can be left out and the meaning of
>> the program is unchanged.
>
>The purpose of pass to explicitly disclaim what you're intending to do;
>it reduces programmer error.  In cases in code where you see something
>like:
>
>	if condition:
>	doSomething
>
>It is far more likely that this is an indentation mistake rather than a
>deliberate null statement. 

I don't recall ever having made that sort of mistake. In any case, 
that sort of thing (i.e. legal but possibly wrong code) is what lint 
tools are for. The classic example being if (a = b) in C.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From starx at pacbell.net  Sat Apr 13 00:16:23 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 13 Apr 2002 04:16:23 GMT
Subject: Initilizing MD lists w/ objects.
References: <Xns91EEC9BB14699RASXnewsDFE1@130.133.1.4> <slrnabf9cb.nub.gerhard@lilith.my-fqdn.de>
Message-ID: <Xns91EED895931BDRASXnewsDFE1@130.133.1.4>

Gerhard =?iso-8859-15?Q?H=E4ring?= || Fri 12 Apr 2002 08:27:07p:

> Philip Swartzleonard wrote in comp.lang.python:
>> I'm looking for a reasonable way to initilize a multidimensional list
>> with a bunch of blank objects of some type (already been burned by
>> using [obj ()] * 400 or similar =). This is the best i've come up
>> with, but it seems icky:
>> 
>>         self.data = [None]*40
>> 
>>         for x in range(40):
>>             self.data[x] = []
>>             for y in range(40):
>>                 self.data[x].append(Vis_cell())
> 
> What about list comprehensions, like:
> 
> self.data = [[Vis_cell() for column in range(40)] for row in
> range(40)] 

Oh. That's perfect. Thanks =).
(Just having also just found out that
for x in data: for cell in x: cell.wahatever
works as well as a double for in range =)

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From narnett at mccmedia.com  Tue Apr  9 21:18:36 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Tue, 9 Apr 2002 18:18:36 -0700
Subject: Creating objects in thread created with subclass
In-Reply-To: <MPG.171d20a7416584ee9896a5@news.nwlink.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJIEKOKAAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Jeff Shannon

...

> No, that's not correct.  If you create an object in the Run
> method of your threading.Thread subclass, then you will create a
> new instance of that object for *each* instance of your subclass,
> so each thread will have a unique object created for it.  Since
> you're using a thread pool, then you'll be re-using that object
> each time that you re-use that thread, but ThreadA will have a
> different instance of it than ThreadB.

I have to say that I'm relieved to hear that... and equally relieved to find
that now that I fixed a couple of things elsewhere in the spider, it works
fine just as you say.  (And I thought it worked that way, but the other
errors were confusing me.)

Thanks!

> > So, I could create objects with 'eval' or 'exec'
> > to generate unique names for each thread -- but is that the
> right way to do
> > it?
>
> No, that is definitely *not* the right way to do it -- as I keep
> saying, exec/eval() are very rarely the right way to do it,
> whatever "it" may be.  ;)

Ah.  Well.  Then as long as we're on the subject... I couldn't figure out a
way to build an insert command for MySQLdb without it.  Given a list of
tuples to insert, I need to give the database handle a statement that looks
like this:

self.dbh.execute("INSERT INTO Foo (fieldnames) VALUES (%s,[%s])",(data[1:]))

The first tuple in the list of tuples is the fieldnames, which can vary.
Thus, the number of "%s" occurrences in the statement has to be calculated.
Is there a way to create this insertion statement without using exec?

Nick





From roy at panix.com  Sat Apr 20 19:19:38 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 20 Apr 2002 19:19:38 -0400
Subject: nested exceptions?
Message-ID: <roy-AAA3A1.19193820042002@news1.panix.com>

I've got a dictionary whose keys are oids.  An oid is a string which looks 
like '1.3.6.1.4.3.5.66.3.4', and represents a tree-structured hierarchy.

I want to look an oid up in my dictionary, and if it's not there, I want to 
keep working my way up the tree until I find it.  Thus, if 
'1.3.6.1.4.3.5.66.3.4' isn't a key, I want to try '1.3.6.1.4.3.5.66.3', and 
then '1.3.6.1.4.3.5.66', and then '1.3.6.1.4.3.5', and so on until I find 
it (if nothing else, '1' is guaranteed to be a key).

What I've come up with uses recursion:

    def getBaseObject (self, objects, oid):
        try:
           return objects[oid]
       except KeyError:
           return self.getBaseObject (objects, self.getParentOid (oid)) 

    def getParentOid (self, oid):
        parts = oid.split ('.')
        parentParts = parts[:-1]
        return '.'.join (parentParts)

The only thing I'm not sure about is that as you recurse your way up the 
tree, you'll have multiple KeyError exceptions in the process of being 
handled, stacked on top of each other.  Are there any problems with that?

I suppose the alternative would be to write this using 
objects.has_key(oid), but the way I've got it now seems neater.

As far as efficiency goes, I expect the vast majority of the oids to exist 
as keys.  For those that don't, I believe all of them will require only a 
single exception to resolve, but I want to make sure I handle the general 
case.


From gcordova at hebmex.com  Fri Apr 26 10:55:24 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 26 Apr 2002 09:55:24 -0500
Subject: python and very large data sets???
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062339B30E6@mail.hebmex.com>

> 
> it is quite easy to set up a database e.g. for a website but 
> is the same true if you have 100 GB of data being processed?
> Easy as pie? Maybe i am underestimating mysql...
> 
> 	holger
> 

Yes indeed, you are.

MySQL holds it's own managing *huge* datasets, and it's *VERY*
(yes, in capital letters and surrounded with asterisks) fast
about doing it's work.

Just don't expect transactions, nor references. Strictly flat
table stuff.

Bloody-evilly-fast, though.

:-)

-gustavo




From gimbo at ftech.net  Wed Apr 17 04:24:49 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Wed, 17 Apr 2002 09:24:49 +0100
Subject: preferred syntax for list extend?
In-Reply-To: <a9jaah$493b$1@as201.hinet.hr>
References: <3CBCC196.466476CE@uniserve.com> <a9jaah$493b$1@as201.hinet.hr>
Message-ID: <20020417082449.GB2063@andy.tynant.ftech.net>

On Wed, Apr 17, 2002 at 10:01:48AM +0200, Ante wrote:
> 
> Even if it was faster, I would still prefer +=.  foo.extend(yada)
> doesn't look beautiful to me at all.  After all, why do we have,
> e.g. == operator so we can say if a == b and not a.equals(b)?
> Because the former is simply better looking, and to me the same
> applies to += versus .extend() or even .append()

True, but note that += won't work with old python interpreters, which
might be one reason for using extend()

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From tim.one at comcast.net  Tue Apr 23 15:51:54 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 23 Apr 2002 15:51:54 -0400
Subject: dictionary method question
In-Reply-To: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEJPDAAA.tim.one@comcast.net>

[dsavitsk]
> for some dictionary d is there any relationship between the returns from
> d.keys() and d.values()?  that is, to get them in some arbitrary but
> matching order, do i need to loop over d.items() or can i count on a
> similarity of the above? in small tests this seems to be the case.

So long as you don't mutate the dict between steps, yes, it's guaranteed
that

    d.keys()   == [k for k, v in d.items()]
and
    d.values() == [v for k, v in d.items()]

In 2.2. it's also guaranteed that

    d.items()  == list(d.iteritems())
    d.keys()   == list(d.iterkeys())
    d.values() == list(d.itervalues())





From tim.one at comcast.net  Thu Apr  4 21:13:24 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 04 Apr 2002 21:13:24 -0500
Subject: ACCEPTED: PEP 285
In-Reply-To: <a8ig6m$mb$1@panix1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEABOLAA.tim.one@comcast.net>

[Aahz]
> There is no alternate choice following BDSL pronouncement.

I saw Guido today, and surprisingly enough he's not *quite* yet a Benevolent
Dictator Sick of Life.

a-few-more-complaints-from-pluto-about-breaking-"if-x:"-may-do-
    the-trick-though-ly y'rs  - tim





From pinard at iro.umontreal.ca  Mon Apr 29 19:02:30 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 29 Apr 2002 19:02:30 -0400
Subject: Python Webpages
In-Reply-To: <aakelj$do6$1@maud.ifi.uio.no>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net>
	<aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no>
	<slrnacpc38.obc.jon+usenet@snowy.squish.net>
	<23891c90.0204290051.3edcc62d@posting.google.com>
	<slrnacq7jk.gav.jon+usenet@snowy.squish.net>
	<mailman.1020090053.19284.python-list@python.org>
	<slrnacqrps.sdd.jon+usenet@snowy.squish.net>
	<mailman.1020102500.7281.python-list@python.org>
	<slrnacra60.v3g.jon+usenet@snowy.squish.net>
	<aakelj$do6$1@maud.ifi.uio.no>
Message-ID: <oqlmb6nm61.fsf@titan.progiciels-bpi.ca>

[Johan Fredrik ?hman]

> > I must admit I have never yet seen a system which integrates HTML and
> > code in the same file without the file ending up an unreadable mess.

> I just thought of great idea: How about using one file and not mix html
> and python code at the same time ?

This is almost exactly what we do.  Here is a short but real example
(phone number changed, that's all).  This page controls what the user
see if there is a database error, or such things.


<!--: Faire
  from Local import commun
  config = commun.config()
  production = config.getboolean('traiter', 'Production')
  adresse = config.get('traiter', 'MailTo')
  entretien = os.path.exists('/usr/local/etc/entretien')
:-->
<html>
 <head><title>erreur.html</title></head>
 <body>
<!--: Si production :-->
<!--: Si entretien :-->
  Le syst?me ?CHO est temporairement non disponible.  Veuillez r?-essayer
  un peu plus tard en journ?e.  Nous nous excusons de cet inconv?nient.

  Robert Lachance, SRAM.
<!--: Sinon :-->
  Une erreur est survenue lors du traitement de votre requ?te.  Veuillez
  contacter madame Louise Groulx ou monsieur Robert Lachance au num?ro :
  (123) 456-7890.

  Vous pouvez, si vous le pr?f?rez, envoyer un courriel ? l'adresse :
  %(adresse)s.  Vous devrez alors pr?ciser les circonstances
  sous lesquelles l'erreur s'est produite.
<!--: FinSi :-->
<!--: Sinon :-->
  Oups!  Probl?me inattendu durant le traitement...
<pre>
%(gros_diagnostic)s
</pre>
<!--: FinSi :-->
 </body>
</html>

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




From DeepBleu at DeepBleu.org  Tue Apr  9 14:44:02 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Tue, 9 Apr 2002 13:44:02 -0500
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com> <3CB2FE25.86A37EA1@imk.fraunhofer.de>
Message-ID: <ub6dji5plutpc5@corp.supernews.com>

"Joachim Kaeber" <joachim.kaeber at imk.fraunhofer.de> wrote in message
news:3CB2FE25.86A37EA1 at imk.fraunhofer.de...
> DeepBleu wrote:
>
> > Yes, I am using zipfile (from zipfile import ZipFile).  Sorry I stated
zlib.
> > Still with one text file in particular, WinZip reports "missing 2 bytes"
> > DeepBleu
>
> is the script run on Windows? If yes, is the zipfile opened
> in binary mode?
>
The script is running on Sun Solaris with Python 2.1

>>>from zipfile import ZipFile
>>>zip_archive = ZipFile("text1.zip", "w", 8)
>>>zip_archive.write(PathToTextFile, os.path.split(PathToTextFile)[1])
>>>zip_archive.close()

Before starting this process, I am running Unix2DOS on the text file:
Unix2Dos TextFile TextFile

Reason is that the zip file will be opened later in windows (using WinZip
for example).
Hope this answers your question,
DeepBleu




From dsavitsk at e-coli.net  Sat Apr 20 14:12:09 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 20 Apr 2002 18:12:09 GMT
Subject: Unable to use win32com on Ecco COM server despite success with VB and Perl COM
References: <e5eab55e.0204191549.41985d9a@posting.google.com>
Message-ID: <Zbiw8.8083$d7.2500042@newssrv26.news.prodigy.com>

I don't know the answer to this, but (a) noone else seems to be answering,
and (b) i can give a little bit of information that might be helpful.

If one creates a method in a VB class like

Public Function X (a as String, b as String)
    X = 1
End Function

Then calls it from Python like
>>> y = com_object.X('a', 'b')
>>> print y
(1, u'a', u'b')

whereas if a and b are specified ByVal

Public Function X (ByVal a as String, ByVal b as String)
    X = 1
End Function

then
>>> y = com_object.X('a', 'b')
>>> print y
1

So, you might be able to just use
>>> y[0] to get the return value.

This doesn't seem to address your problem exactly, but maybe it will shed
some light on what is going on.

-doug


"Raymond Yee" <yee at uclink.berkeley.edu> wrote in message
news:e5eab55e.0204191549.41985d9a at posting.google.com...
> I've been trying unsuccesfully to use Python to access the
> functionality of the "Visual Basic Toolkit"
> (http://www.ivitar.com/vbecco/). The toolkit is a COM InProc
> Server (.exe).
>
> I'm puzzled as to how to proceed because of two things:  1) My
> Python program is able to use some functions of the toolkit but
> not others and 2) I've had success with both Visual Basic and Perl
> in accessing the toolkit via COM.
>
> Below are the three versions of the program.  Ecco is a personal
> information manager.  This program is supposed to print out the
> text of any selected outline and the corresponding URL in a folder
> called "Net Location"  (The developer key has beeen x'ed out)
>
> In the Perl and VB versions, I am able to get both the "text
> value" of the selected item ($item->Text in Perl) and the
> corresponding URL ($item->FolderValue("Net Location")).  However,
> in the Python version, while EccoItem.Text is what I expect,
> EccoItem.FolderValue("Net Location") returns
>
> (u'Net Location', None)
>
> -- a tuple, whose first item is type unicode.  However, I don't
> see anything in the return value remotely related to the URL I'm
> trying to access.
>
> Any suggestions on how to get Python to work with the toolkit? (Am
> I using the wrong syntax?  Are there options I'm setting right?
> How should I debug this?)
>
> Thanks in advance,
>
> -Raymond Yee
>
> ---------------------------------------------------------------
> Python version:
>
> # build OPML from the selected outline (if more than one is selected,
> use the first item
>
> import win32com.client
> Ecco = win32com.client.Dispatch("VBEcco.EccoApp")
> Ecco.Register("xxxxxxxx", -1, -1)
> EccoSelection = Ecco.Selection
>
> #print value of item in Net Location folder
>
> for EccoItem in EccoSelection:
>     print "Text: " + EccoItem.Text
>     print EccoItem.FolderValue("Net Location")
>
> ---------------------------------------------------------------
>
> Perl version:
>
> # establish COM connection to Ecco
>
> use Win32::OLE qw(in);
> #use Win32::OLE::in;
>
> my($ecco, $sel, $item);
>
> $ecco = Win32::OLE->new("VBEcco.EccoApp", 'Quit');
> $ecco->Register("xxxxxxxx", -1, -1);
>
> $sel = $ecco->Selection;
>
> foreach $item (in($sel)) {
>
>     print $item->Text, "\n";
>     print $item->FolderValue("Net Location") , "\n";
>
> }
>
> ---------------------------------------------------------------
>
> Visual Basic version:
>
> Option Explicit
>
> 'Place in Declarations
> Dim Ecco As EccoApp
> Dim EccoSelection As EccoItems
> Dim EccoItem As EccoItem
>
> Private Sub Form_Load()
>
> Set Ecco = New EccoApp
> Ecco.Register "xxxxxxxx", -1, -1
>
> ' Get Selections
> Set EccoSelection = Ecco.Selection
>
> 'if there is a first selection, print out the text, and figure out
> whether there is an associated URL in the Net Location folder
>
> For Each EccoItem In EccoSelection
>     MsgBox ("Text: " + EccoItem.Text)
>     MsgBox ("NL: " + EccoItem.FolderValue("Net Location"))
> Next
>
>
> End Sub




From theivo at abv.bg  Sat Apr 13 17:39:58 2002
From: theivo at abv.bg (I.C.)
Date: 13 Apr 2002 14:39:58 -0700
Subject: dos promt simulation
References: <6ac1ecad.0204130435.93d0ad3@posting.google.com> <Xns91EFBDC8F2FCcliechtigmxnet@62.2.16.82>
Message-ID: <6ac1ecad.0204131339.56928640@posting.google.com>

> there is also popen2, popen3 etc. they give you a file for stdin where you 
> can write to.

Maybe popen2\3 could help, but I think, I don't use them correctly :

import os
dos_input, dos_output = os.popen2('time')

for line in dos_output.readlines(): # Current time is 22:14:43,24
        print line                  # Enter new time:

dos_input.write(raw_input())        # User enters the new time

Where is my mistake?

I.C.


From phr-n2002a at nightsong.com  Wed Apr  3 17:11:15 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 14:11:15 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017853609.14796.python-list@python.org>
Message-ID: <7xvgb8o2kc.fsf@ruckus.brouhaha.com>

"Mark McEahern" <marklists at mceahern.com> writes:
> [Kevin Jacobs]
> > The moral of the story is that I will be extremely happy once there is a
> > globally blessed way of spelling true and false other than 1 and 0.  Of
> > course, I and my team aren't exactly typical Python developers.  Though
> > wouldn't it be nice if we were?
> 
> I'd like to consciously go out of my way and violate all rules of netiquette
> and say simply this:
> 
> Me too.

There are already some predefined constants in Python (or at least some
modules), like math.pi=3.14159... and so forth.

If the constants True=1 and False=0 were added to Python (just plain
integer values, not a bool type with any special semantics), that
would take care of what you're asking for, no?


From logiplexsoftware at earthlink.net  Mon Apr  1 11:44:21 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 1 Apr 2002 08:44:21 -0800
Subject: runtime location discovery
In-Reply-To: <20020330014832.GF7602@isis.gerrietts.net>
References: <20020330014832.GF7602@isis.gerrietts.net>
Message-ID: <20020401084421.3aa73878.logiplexsoftware@earthlink.net>

On Fri, 29 Mar 2002 17:48:32 -0800
Geoff Gerrietts wrote:
>     Image.open(os.path.join(root_location, "image/naked_chicks.png"))

Maybe if you post this image we could come up with some ideas ;)


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From andy at reportlab.com  Mon Apr  1 18:00:04 2002
From: andy at reportlab.com (Andy Robinson)
Date: Mon, 01 Apr 2002 23:00:04 GMT
Subject: The fastest XML parser around
References: <mailman.1017654423.12312.clpa-moderators@python.org> <a89pnc$7j9$1@panix1.panix.com>
Message-ID: <UD5q8.874$aO4.4453379@news-text.cableinet.net>

> Do you have information available on using pyRXP with PythonPoint?

Not yet.  The simplest thing would be to add a 3-line hack to validate
the document if pyRXP is present :-)

I plan to rewrite pythonpoint using our XML toolkit - indeed, use it
as the exercise in the manual - but we need to finalize two more components
first:  a 'lazy accessor' giving friendly syntax for navigating the tree,
and
something to convert attributes to numbers/lengths/colors en masse.
After that the Pythonpoint code could get a lot shorter.  Some broad
plans are in the back of the pyRXP documentation, referenced from
the web page my announcement pointed at.

- Andy




From syver-en+usenet at online.no  Sun Apr  7 15:02:03 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sun, 07 Apr 2002 19:02:03 GMT
Subject: New classes in 2.2
References: <un0wglgqf.fsf@online.no> <a8no98$leu$1@panix1.panix.com>
Message-ID: <ud6xbfi4w.fsf@online.no>

aahz at pythoncraft.com (Aahz) writes:

> 
> Python 3.0 will get rid of "classic" classes entirely.  

That's a relief, I hope it will have a super that works on class
methods too.

-- 

Vennlig hilsen 

Syver Enstad


From zeitlin at seth.lpthe.jussieu.fr  Wed Apr  3 10:45:32 2002
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Wed, 3 Apr 2002 15:45:32 +0000 (UTC)
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <a84561$p5g$1@wanadoo.fr> <slrnaabbct.npb.cpbotha@crabtree.cpbotha.net> <4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com> <20020401020912.GE333@lilith.hqd-internal> <mailman.1017681605.21669.python-list@python.org> <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net> <slrnaaivjc.565.zeitlin@seth.lpthe.jussieu.fr> <slrnaaj2o9.a62.cpbotha@crabtree.cpbotha.net>
Message-ID: <slrnaam8ss.4mg.zeitlin@seth.lpthe.jussieu.fr>

On Tue, 02 Apr 2002 10:37:45 GMT, Charl P. Botha <cpbotha at crabtree.cpbotha.net> wrote:
> Please read the whole thread before replying.

 As a matter of fact I did. But I didn't find any specific mention of the
problems which anybody has encountered under either of those platforms, sorry
if I missed this somewhere.

> I did NOT say that wxWindows doesn't build on these architectures.

 Your statement implied that there were problems with building it under SGI
and/or Solaris without saying what those problems were. I wanted to know more
details about this.

> To clarify, and I've said this previously in this thread, I
> have the highest regard for wxWindows and wxPython.

 Although it surely is a pleasure to hear this, I'm more interested in making
wxGTK and wxPython build out of the box on as many platforms as possible. And
for this we need the users bug reports which was the only reason I replied to
this thread. So I'd simply like to reiterate my plea to all [potential]
wxPython users to please let us know about any problems with building and
installing it.

> What I DID say was that getting a working wxPython installation on one of
> those boxes can take quite a while.  One has to get GTK and all the packages
> it depends on and configure and build all of these.  Then one has to
> configure and build wxGTK and wxPython... all of this can take some time.

 Yes, but this is not a bug, it's a feature. Or at least there is nothing we
can do about it (except making a wxPython/X11 version).

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From johanfo at essay.org  Sun Apr 28 20:56:17 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Mon, 29 Apr 2002 02:56:17 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no>
Message-ID: <aai5oo$99l$1@maud.ifi.uio.no>

Just realized that these Servlet based development models really limits the
usage of software like Dreamweaver.  Since all code is in the program these
utilities are almost worthless.  Is there a good way to combine Webware with
templates.  I think of templates as pages where certain %words% are substituted
etc...

--
JF?

"Johan Fredrik ?hman" <johanfo at essay.org> wrote in message
news:aahiq2$6t3$1 at maud.ifi.uio.no...
> I have just had a look at Webware myself,  and it seems like one of the more
> promesing projects.  At least it is not abandoned.
> --
> JF?
>
> > I tested Webware the other week and it looks very promising. It is very
> easy
> > to e.g. develop a template as a servlet and let other servlets inherit the
> > layout and functionality which makes Webware extremly powerful and easy to
> > use.
> >
> > There is also support for PSP which I haven't tried (embedded Python in
> > HTML).
> >
> > Another powerful (not yet final though) module is Middlekit which e.g.
> makes
> > it easy to deploy a servlet that acts with some glueware to e.g. MySQL.
> >
> > I am going to follow the work of Webware since I like the freedom it
> offers.
> > It seems more convenient to use comapred to Zope.
>
>
>




From mlh at vier.idi.ntnu.no  Wed Apr  3 07:16:48 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Wed, 3 Apr 2002 12:16:48 +0000 (UTC)
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017795008.700.python-list@python.org> <7xzo0lth3r.fsf@ruckus.brouhaha.com> <mailman.1017817388.30724.python-list@python.org>  <7xd6xhyzvx.fsf@ruckus.brouhaha.com>
Message-ID: <slrnaalslg.fa8.mlh@vier.idi.ntnu.no>

In article <7xd6xhyzvx.fsf at ruckus.brouhaha.com>, Paul Rubin wrote:
[snip]
>I wonder what other systems with inline docs, like Parc Smalltalk or
>the MIT Lisp machine, use for syntax?  Emacs Lisp uses plain text with
>some typographic conventions, but these don't go as far as ReST, and
>it uses a separate system for structured manuals.

How about a syntax like Yodl? It's a markup language I used a bit a
*long* time ago. Heh -- funny -- it's on xs4all, just like Python:

  http://www.xs4all.nl/~jantien/yodl/

I'm not proposing to use the language (at least not necessarily...)
but perhaps the main markup syntax. Yeah, I know -- it's explicit
markup... But it ought to be quite familiar to Pythoneers...

It basically works like em(this). I.e. macros (== markup tags) are
used like function calls. (Hey, they could even be implemented
directly as functions/methods in Python :)

There are a couple of extra syntax additions. I think you can write
foo+(bar) to get the string foo(bar) in your text (without actually
calling a macro) for instance.

Oh, well. Just a thought.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From donn at u.washington.edu  Wed Apr  3 13:30:15 2002
From: donn at u.washington.edu (Donn Cave)
Date: 3 Apr 2002 18:30:15 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <ualr7hmlhkemab@news.supernews.com> <slrnaam6of.ldi.akuchlin@crystal.mems-exchange.org>
Message-ID: <a8fhnn$oqm$1@nntp6.u.washington.edu>

Quoth "A.M. Kuchling" <akuchlin at crystal.mems-exchange.org>:
| In article <ualr7hmlhkemab at news.supernews.com>, John Roth wrote:
|> In fact, at this point I'd like everything substantive to be defered
|> to Python 3000, and get started on that right away. I've got a
|> few suggestions...
|
| It seems really unlikely that anyone will ever start with a completely
| blank slate and re-implement Python from the ground up in C/C++.  Such
| efforts have to rebuild absolutely everything, from the parser to the
| basic data structures to all the modules, and there would be weeks and
| months of work before there was anything usable again.  

Are you saying there will never be a 3000, whatever it's called, a
substantially new version that sacrifices some compatibility?  Whether
it's accomplished from a blank slate or reuses a lot of the existing
core, I thought that was the plan.  All the modules that survive that
would obviously have to be at least looked at.

If so, I thought it was a good point - why not just get started on it?
Why wrestle and compromise with these compatibility problems, when
that big overhaul is on the horizon anyway?

	Donn Cave, donn at u.washington.edu


From mertz at gnosis.cx  Tue Apr  2 23:58:40 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Tue, 02 Apr 2002 23:58:40 -0500
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net>
Message-ID: <Awoq8kKkX44c092yn@gnosis.cx>                

"Jim Jacobs" <jim.jacobs at jacobshome.com> wrote previously:
|Is it possible to simulate "C" style static variables in functions?
|In methods?

>From the other responses, it seems like a lot of users don't know the
"Python static trick."  I use this moderately often... it is a slightly
strange spelling, but not hard once you learn it:

    >>> def counter(msg, cnt=[0]):
    ...   print cnt[0], msg
    ...   cnt[0] += 1
    ...
    >>> counter('first call')
    0 first call
    >>> counter('second call')
    1 second call
    >>> counter('third call')
    2 third call
    >>> counter('counter override', [37])
    37 counter override

Anything passed with a mutable type default argument gets defined along
with the function itself.  The value doesn't get redefined on calls
unless the caller explicitly forces it.

We-don't-need-no-stinkin-closures-ly yours, David...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s




From lumber at jack.com  Thu Apr 11 00:56:55 2002
From: lumber at jack.com (Lumberjack)
Date: 11 Apr 2002 04:56:55 GMT
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com>
Message-ID: <Xns91ECDF3AC49F7lumberjack@216.148.53.85>

Peter Hansen <peter at engcorp.com> wrote:
> I'm dismayed by comments attacking the core developers when made
> by people who have not contributed more than I or my team have.
> Beggars can't be choosers, and nobody likes a whiner...

This incorrectly labeled "troll" has made small monetary and other 
contributions to the advancement of Python. And a larger investment in 
using the language.

I'd hate to see that investment thrown away over a failure by the core 
developers to recognize the normally undeniable evidence that a never-
ending flow of frequent PEPs and changes to the language will undermine the 
legitimacy of Python in a host of obvious and not so obvious ways. Change 
isn't synonymous with progress; changing your underwear every day isn't 
progress (well, for anyone doing it on a weekly basis I guess it would be).
These maxims or precepts should be kept in mind:

1) One should have a well defined set of prioritized requirements and a 
recognizable goal. Saying you are designing the "perfect language" isn't a 
requirement or a goal; it is meaningless and not measurable. Nor is saying 
it should be easy for nonprogrammers. If you don't have a well defined idea 
of your goal, you are never going to know when you've arrived or how to get 
there.

2) One shouldn't ask "is there any compelling reason not to make this 
change" but rather "are there any compelling reasons with good supporting 
evidence that justify making this change?" Quite a different thing indeed. 
And the minor release strategy doesn't yield evidence since no one is doing 
any data collection to measure the success of the changes, and even if they 
were, the goal of backward compatibility would be in conflict with the need 
to remove failed changes from the language.

3) Changes should be much much less frequent, and previously minor release 
changes should be incorporated into the major releases. People had no 
choice in accepting Python at the start; it was a take-it or leave-it 
issue. The psychological impact was focussed, not dragged out for a long 
period of time. There are also commercial reasons for minimizing the 
frequency of changes. Frequent minor releases telegraph indecision and lack 
of discipline.


From robin at jessikat.fsnet.co.uk  Wed Apr  3 07:24:22 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 3 Apr 2002 13:24:22 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
 <mailman.1017799668.13959.python-list@python.org>
 <ualr7hmlhkemab@news.supernews.com>
Message-ID: <J7qT5iA2Rvq8EwLD@jessikat.fsnet.co.uk>

In article <ualr7hmlhkemab at news.supernews.com>, John Roth
<johnroth at ameritech.net> writes
.....
>Consequently, I think of it (0 and  1 as false and true) as a
>hack carried over from C and similar languages.
>
>John Roth
C is a very successful language. I don't think of its conditional
testing as a hack, but then I'm an engineer. What works is good and
Python's current conditional testing seems to work very well and
moreover efficiently in many cases of interest.

Forcing all these into the (true, false) set using a Boole (if we're
going to honour the man let's spell him right) variable seems to put too
much emphasis on formalism. It also ignores the eminent sense of not
fixing unbroken things and not adding yet another current to the Python
type cake. 
-I wish I had Kalmanian filter types-ly yrs- 
Robin Becker


From h_schneider at marketmix.com  Thu Apr  4 01:24:54 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Thu, 4 Apr 2002 08:24:54 +0200
Subject: wxPython Book ?
References: <11e94203.0204031526.f11bfe6@posting.google.com>
Message-ID: <a8gr9l$c29$04$1@news.t-online.com>

"Jose Rivera" <jriveramerla at yahoo.com> schrieb im Newsbeitrag
news:11e94203.0204031526.f11bfe6 at posting.google.com...
> Hi...
> Is there a wxPython book available? that gives a depth tutorial?

There is a book on the way:

> > Hi,
> >
> > We're a computer book publisher seeking reviewers for a book proposal
> > on wxPython we currently have in hand. The proposal is approximately
> > 2,913 words in length. We're seeking 3-6 people to review it and
> > make comments that will help us make a decision on whether or not to
> > publish.
> >
> > If you're interested or can recommend someone who might be qualified
> > and willing, please feel free to contact us.
> >
> > Thanks in advance!
> >
> > Sincerely,

> >

> >

> >

> > =======================================

> > Susan W. Capparelle

> > Assistant Publisher

> > Manning Publications Co.

> > 209 Bruce Park Avenue, Greenwich, CT 06830

> > suca at manning.com

> > tel. 203.629.2211 www.manning.com

> > fax. 203.629.2084

> > =======================================





From jeff at ccvcorp.com  Thu Apr 18 15:32:13 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Thu, 18 Apr 2002 12:32:13 -0700
Subject: Counting the number of elements in a list - help.
References: <a9hu8e$2fg$1@lumberjack.rand.org>
Message-ID: <MPG.1728bf07ecc579969896b3@news.nwlink.com>

In article <a9hu8e$2fg$1 at lumberjack.rand.org>, bmason at rand.org 
says...
> I need to know these things; may someone help me out?
> 
> 1) a function which returns the number of elements in a list
> 2) a function which takes a list and an index and returns a zero when that
> list index is out of range.

It seems that your problem has been solved, but... just 'cause...

Your second function could easily be written like this:

def InRange(sequence, index):
    try:
        sequence[index]
    except IndexError:
        return 0
    else:
        return 1

(Note that the 'return 1' isn't strictly required to be in an 
else: clause here, but I think that using else, in this case, is 
a clearer expression of intent.)

-- 

Jeff Shannon
Technician/Programmer
Credit International


From jeff at jeffyouel.com  Wed Apr  3 12:10:53 2002
From: jeff at jeffyouel.com (Jeff Youel)
Date: Wed, 03 Apr 2002 17:10:53 GMT
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <xIGq8.25783$VQ2.8306517@twister.socal.rr.com>

"Roy Smith" <roy at panix.com> wrote in message
news:a8fba8$ln8$1 at panix2.panix.com...
> I've got a dictionary d, which may or may not have a value I'm looking
> for.  If it does, I want to do a few things.  If it doesn't, I want to
> do a few other things.  I could write:
<snip>

You can avoid the exception handler altogether with:

    value = d.get(key, None)
    if value:
        foo = stuff(value)
    else:
        foo = otherStuff(key)


Jeff




From phil.harris at zope.co.uk  Mon Apr 22 17:54:13 2002
From: phil.harris at zope.co.uk (Phil Harris)
Date: Mon, 22 Apr 2002 22:54:13 +0100
Subject: Big Bucket O [Python,Zope,Stackless,OpenBSD] questions.
References: <Qu%w8.1925$iQ5.165333@newsread2.prod.itd.earthlink.net>
Message-ID: <009001c1ea48$40563160$e1e8403e@harrisfamily>

I think you're a little too ambitious as Zope 2.5 and above won't work  with
Python 2.2 (yet).

The recommended version of Python for Zope 2.5.x is Python 2.1.3.


----- Original Message -----
From: "adeptus" <adeptus at earthlink.net>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, April 22, 2002 10:44 PM
Subject: Big Bucket O [Python,Zope,Stackless,OpenBSD] questions.


> Just got openBSD up and running, and was very dissapointed when I tried to
> get the standalone ZODB up and running... Then I found out about the
> threading issues...
>
> There's not much info out there on these questions, so anything anybody
has
> to say on the subject would be appreciated.
>
> Python 2.2, Zope 2.5, StandaloneZODB all working on a BSD3.0 installation?
>
> Zodb under stackless python? Is it possible?
>
>
> I need the new features in 2.2 (generators, metaclasses) for a very
> ambitious Zope product...
>
>
> Thanks...
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>






From tim.one at comcast.net  Sat Apr 20 01:17:42 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 20 Apr 2002 01:17:42 -0400
Subject: How to make regexes faster? (Python v. OmniMark)
In-Reply-To: <1019278636.737636@yasure>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKOOPAA.tim.one@comcast.net>

[Donn Cave]
> ...
> I'm really delighted to hear about performance improvements in file
> object input in recent Python versions (in a message that would
> have been in this thread if Tim Peters used a newsreader instead
> posting to the list.)

Hmm.  Shows up fine in the newsreader I use -- this, like hearing about the
speedups for the first time from that msg, may be another symptom of General
Upgrade Fear <0.9 wink>.

> Mostly because it's a good thing in itself, assuming the numbers are
> there, but also because it shows that someone must have felt that
> performance was important enough to work pretty hard on it - that's a
> healthy sign.

Python-Dev was consumed by it at the time (December of 2000).

> By the way, I'll second Johannes Stiehler's recommendation of
> MxTextTools.  Definitely appropriate for SGML parsing, and much
> better than regexps for extensive parsing in my opinion - not
> just in terms of speed, but I suspect a more powerful way to
> describe text patterns than regexps.

Yes, it is.  "More convenient" is arguable, though -- there's a steep
learning curve, but then people often forget how hard it was to learn regexp
syntax and pragmatics too.





From mgilfix at eecs.tufts.edu  Mon Apr 29 11:14:08 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 29 Apr 2002 11:14:08 -0400
Subject: Lambda Scoping
In-Reply-To: <3ccd5f99.5559046@news.t-online.de>; from gerson.kurz@t-online.de on Mon, Apr 29, 2002 at 03:02:42PM +0000
References: <3ccd5f99.5559046@news.t-online.de>
Message-ID: <20020429111407.D12835@eecs.tufts.edu>

  Lambdas in python don't work exactly like lambdas that you would
expect in lisp. The scoping is a little different. To make this work,
you need to do something like this:

    lambda i=instance: i

  Which does an assignment to i when the lambda function is created
and gets carried around with the function.

  As for the merits of lambdas, well, you might want to read some
other threads being posted today :)

               -- Mike

On Mon, Apr 29 @ 15:02, Gerson Kurz wrote:
> The following code
> ...
> funcs = []
> for instance in range(10):
>     funcs.append( lambda: instance )
> print map(apply,funcs)
> 
> funcs = []
> for instance in range(10):
>     funcs.append( lambda instance=instance: instance )
> print map(apply,funcs)
> ...
> prints
> 
> [9, 9, 9, 9, 9, 9, 9, 9, 9, 9]
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> 
> which I sort-of understand (I figure, in "lambda: instance" instance
> is a lazy expression; "lambda instance=instance: instance" the
> expression itself - "instance" - is a lazy expression, too, but the
> argument list is not.) However, can anybody please explain this more
> detailed? 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (gerson.kurz)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From zopestoller at thomas-guettler.de  Thu Apr 18 02:44:31 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 18 Apr 2002 08:44:31 +0200
Subject: Mailman lists at Starship
References: <mailman.1019085516.2498.python-list@python.org>
Message-ID: <3CBE6B4F.3030506@thomas-guettler.de>

Gerhard H?ring wrote:

> Does anybody know who is the technial support contact ;-) for these?
> 
> I'm the list admin of
> http://starship.python.net/mailman/listinfo/python-de
> and I'm having some problems.
> 
> Gerhard
> 

Some days ago there was this message in python-announce:

"""
We plan on upgrading the mail server mail.python.org on Wednesday
27-Feb-2002, at 11am -0500 (EST, US/Eastern).  This will affect all
mail and mailing lists on the python.org and zope.org domains.

Downtime should be minimal; if all goes well, just 10-15 minutes.  You
shouldn't even notice it.  If you still cannot get email through by
noon our time, check the www.python.org home page for status, or email
Barry (barry at zope.com), Martijn (mj at zope.com), and Greg
(gward at mems-exchange.org).

-Postmaster @ {python,zope}.org
"""

thomas




From gerhard at bigfoot.de  Mon Apr  1 01:50:36 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Apr 2002 06:50:36 GMT
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com>
Message-ID: <slrnaaftai.8mf.gerhard@lilith.my-fqdn.de>

Robert Oschler wrote in comp.lang.python:
> What's so special about Mozilla

Short version:
Netscape 4.x sucks a**. Mozilla rocks.

Long version: Mozilla is a complete rewrite of a new HTML renderer
called Gecko, that is up to todays standards and actually understands
HTML 4.x, XHTML, (most of) CSS level 2, XML and does client-side
XSL-transformations. Netscape 4.x only understands technologies from
the last millenium (CSS level 1, HTML up to 4.x). Also, the
mailreader, newsreader and HTML editor were completely rewritten. So
see this as a new browser, not anything like that old beast that is
Netscape 4.x. There's a commercially branded and qa-ed version called
Netscape 6.x that also comes with lots of very important AOL software
;-)

> What's with all the stuff in your sig?  Especially the last one that
> looks like a LISP statement.  I'm feeling a bit of "sig envy".

Obfuscated Python - try it out in your Python interpreter :-)

Cheers,

Gerhard
--
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From aleax at aleax.it  Tue Apr 30 04:32:34 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 08:32:34 GMT
Subject: Third Draft: Pep for Deprecating Builtins
References: <mailman.1020129837.31112.python-list@python.org>
Message-ID: <CEsz8.80366$vF6.2381303@news2.tin.it>

Arthur Siegel wrote:

>>I'm not convinced Python 3 is a "new enough" language for this PEP's
>>ideas, although that would seem the obvious point for this sort of
>>tinkering
> 
> Perhaps I don't understand the transition path to new style classes.
> 
> Is the plan to have the 2 class structures co-exist indefinitely?
> That would seem strange and more than a little ugly.
> 
> But isn't the concept of deprecating old style classes near
> impossible, as a real world matter?

Why would it be?  The degree of compatibility between new-style classes and 
classic ones is amazing.  Try sticking a __metaclass__=type at the top of
your modules and see what little breaks -- basically only some rather 
obscure and advanced metaprogramming techniques, hardly make-or-break issues
for "real world matters".

I'm not into channeling GvR (you need the timbot for that), so I can't claim
I know what is The Plan, but a slow and gradual transition to "exclusive" 
use of new-style classes over a few releases seems perfectly feasible to 
me, given the high degree of compatibility GvR was able to achieve.

The quotes around "exclusive" are due to my doubts about your assertion
above, which translates into "having more than one kind of metaclass 
available seems strange and more than a little ugly".  Why couldn't
Python supply two or more metaclasses ready for use, just as it offers
both type and types.ClassType today?  I.e., why should all metaclasses
other than the built-in type be necessarily user-coded ones, or third
party extensions -- certainly there IS space for some in Python's large
standard library, no?

So, types.ClassType COULD stay around "forever" from my POV, even long after
the default metaclass has become type -- at worst, the way to enable the
alternative metaclass as the default one would be 2 lines at module scope:

import types
__metaclass__ = types.ClassType

If some few percents of code need this peculiar metaclass, or others yet,
why couldn't they conceptually be supplied in such wise...?


Alex



From sjmachin at lexicon.net  Sat Apr 20 18:36:07 2002
From: sjmachin at lexicon.net (John Machin)
Date: 20 Apr 2002 15:36:07 -0700
Subject: An extraordinarily stupid question
References: <3CC0D892.2060006@nyu.edu>
Message-ID: <c76ff6fc.0204201436.3b66cda@posting.google.com>

Shane Hoversten <srh232 at nyu.edu> wrote in message news:<3CC0D892.2060006 at nyu.edu>...
> Hi -
> 
> 	This is really dumb, but I just want to print out a character without a 
> newline and without a space behind it.  You can't do:
> 
> print c,
> 
> 	Because Python puts a space behind it.

Not part of the solution to your problem, just keeping the record
straight:

   Actually it puts a space in *front*.

> I can't find anywhere in the dox 
> that actually talk about "print" in any detail at all.

So that we can have more "findable" docs and more precise docs, would
you please let us know either what might have prevented you from
finding the section on the print statement in the language reference
manual, or if you found it, why you think it didn't 'talk about
"print" in any detail at all'?


From richard at electrophobia.com  Thu Apr 25 00:05:03 2002
From: richard at electrophobia.com (Richard Parker)
Date: Thu, 25 Apr 2002 04:05:03 GMT
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com>
Message-ID: <B8ECCE7D.F941%richard@electrophobia.com>

Paul Rubin at phr-n2002a at nightsong.com wrote on 4/24/02 8:05 PM:

> But HMAC's keys are already dependent (they're the same key xor'd with
> two different constants), so I'm not sure how concatenating something is
> really worse.  I'm back to wanting to re-read the HMAC paper, which I'll
> try to do sometime.

It is possible I am remembering the proof incorrectly, but as I recall the
security of HMAC was proved by first proving secure a construction with
independent keys and then showing that HMAC was secure with the additional
assumption that the compression function of the hash function is sufficient
to make the two derived keys indistinguishable from random.

> Currently I'm using
> 
> K1 = H('auth1' + K)
> K2 = H('auth2' + K)
> MAC = H(K1 + H(K2 + ciphertext))
> 
> which I hope is at least as good as HMAC (the keys are more
> independent) and it uses fewer interpreter operations (though more SHA
> calls).  It's speed isn't too bad, but I always like to look for
> improvements.

I like that idea.  I do, however, recommend that as a precaution you choose
values for the two constants that differ by more than one bit.  While more
expensive than HMAC in the theoretical sense, this HMAC variant looks secure
and I'm not surprised to hear that it is actually faster in Python.  I can't
see any attack that wouldn't also imply the ability to crack the hash
function, which is the same level of security offered by HMAC.  If you are
using the MAC more than once with same key you could improve performance by
only computing the values of K1 and K2 once (this optimization would also
apply to HMAC).

> Note that there's a trivial O(2**64) attack on my authentication since
> I'm truncating the MAC to 8 bytes.  So if a security difference
> between HMAC and what I'm doing needs more than O(2**64) work to
> exploit, it's not really useful to an attacker.  Is that a reasonable
> way to think of this?

Yes, it is certainly acceptable to perform that kind of analysis.  On the
other hand, I personally try hard to avoid cryptographic designs that
introduce dependencies that make it difficult to change parameters or
primitives without reanalyzing the entire design.

-Richard



From phd at phd.pp.ru  Thu Apr  4 03:06:30 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 4 Apr 2002 12:06:30 +0400
Subject: Portable runtime
In-Reply-To: <5.1.0.14.2.20020404071247.04bb6210@202.95.131.154>; from tripps81@yahoo.com on Thu, Apr 04, 2002 at 07:20:09AM +0700
References: <5.1.0.14.2.20020404071247.04bb6210@202.95.131.154>
Message-ID: <20020404120630.F14507@phd.pp.ru>

On Thu, Apr 04, 2002 at 07:20:09AM +0700, Tripp Scott wrote:
> I was just wondereing whether there is a Python binding to APR, 

   No.

> NSPR, or perhaps plans to sit CPython on top of one of these libraries?

   Currently ther are no, but patches will be gladly accepted!

> Btw, does Python emulate threads on platforms where threads are not available?

   No, Python uses pthreads. There are separate patches to do this
(uthreads).

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




From maxm at mxm.dk  Thu Apr 11 08:35:14 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 11 Apr 2002 14:35:14 +0200
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it>
Message-ID: <3CB58302.5090200@mxm.dk>

Alex Martelli wrote:
> Ante Bagaric wrote:

> A vastly superior way is
> to use sys.stdout.write("%s,"%i) -- you just need some care to
> show it off interactively because you need a '\n' at the end,
> so, for example:
> 
> 
>>>>if 1:
>>>
> ...   for i in range(5): sys.stdout.write("%s,"%i)
> ...   sys.stdout.write('\n')
> ...
> 0,1,2,3,4,


A lazy bumm like me would just write:

 >>> print ','.join([str(i) for i in range(5)])
0,1,2,3,4

btw: it would be nice if join() had a switch so that it would 
automatically do a str() on each object it joins.

like::

 >>> print ','.join(range(5), toStr=1)
0,1,2,3,4

I keep having to do this repeatedly in my code when formatting output.

Perhaps join() should even do this as the default, as it makes the most 
sense in the join() context.

like::

def newJoin(self, items):
     return self.join([str(i) for i in items])

Is there any case where it isn't applicable at all?

regards Max M



From andreas at mtg.co.at  Fri Apr 12 05:29:59 2002
From: andreas at mtg.co.at (Andreas Kostyrka)
Date: Fri, 12 Apr 2002 11:29:59 +0200
Subject: BUG: Python 2.1.1 nttplib
Message-ID: <20020412112959.73a8da53.andreas@mtg.co.at>

Hi!

In my Python 2.1.1 installation nntplib.py:NNTP.body
misses the second optional argument file as described in the
documentation.

Best regards,

Andreas Kostyrka




From timr at probo.com  Mon Apr 15 01:49:41 2002
From: timr at probo.com (Tim Roberts)
Date: Sun, 14 Apr 2002 22:49:41 -0700
Subject: problem with wxHtmlWindow
References: <a98rht$pok$1@wanadoo.fr>
Message-ID: <jdqkbu807tsccbisua5uidmgbm1lq9l1vg@4ax.com>

"manouche" <emmanuel at onda-latina.com> wrote:
>
>What are the libraries I have to import in order to use wxHtmlWindow?

from wxPython.html import *

This works; I've done it.

>I tried importing the libraries of the example
>
>from wxPython.wx import *
>from wxPython.html import *
>import wxPython.lib.wxpTag
>
>but my python shell says that:

This traceback doesn't make sense.  The final line does not reference
wxHtmlWindow.  Did you cut and paste this from Boa?  Could you possibly
have missed a couple of lines?

>Traceback (most recent call last):
>  File "D:\prog\boa1\wxApp1.py", line 22, in ?
>    main()
>  File "D:\prog\boa1\wxApp1.py", line 18, in main
>    application = BoaApp(0)
>  File "C:\PYTHON22\lib\site-packages\wxPython\wx.py", line 1587, in
>__init__
>    _wxStart(self.OnInit)
>  File "D:\prog\boa1\wxApp1.py", line 12, in OnInit
>    self.main = wxFrame1.create(None)
>  File "D:\prog\boa1\wxFrame1.py", line 6, in create
>  File "D:\prog\boa1\wxFrame1.py", line 23, in __init__
>  File "D:\prog\boa1\wxFrame1.py", line 19, in _init_ctrls
>    self.SetClientSize(wxSize(624, 642))
>NameError: global name 'wxHtmlWindow' is not defined
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From gh_pythonlist at gmx.de  Tue Apr  2 20:14:44 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 3 Apr 2002 03:14:44 +0200
Subject: relational database?
In-Reply-To: <23891c90.0204020644.3ea4bc40@posting.google.com>
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com>
Message-ID: <20020403011444.GA562@lilith.my-fqdn.de>

* Paul Boddie <paul at boddie.net> [2002-04-02 06:44 -0800]:
> "Ruediger Maehl" <ruediger.maehl_nospam at web.de> wrote:
> > SAPDB, a full featured database for free at www.sapdb.org
> > with Python interface.
> 
> I'm not sure I would second that recommendation unless SAP DB has
> changed dramatically over the past few weeks.

As far as I've seen, it's quite a complex piece of software. And the
code itself, the organisation of the code and the build process look
like they've come from a different universe. I'd have to think hard to
invent something similarly confusing.

> On UNIX systems, PostgreSQL and pgPySQL are possibly as easy, 

pyPgSQL is also quite easy to set up on Windows. If it fails, complain
to me. To run PostgreSQL itself on Windows, you'll need to download
Cygwin and it's PostgreSQL package, which will work for development
purposes, but the performance isn't comparable to running it on a real
Unix. Unfortunately, the PostgreSQL folks aren't interested to invest
the time for a native win32 port. But I'm sure if somebody is really
interested and funds such a port, it will be done.

> Another interesting issue that one might want to consider later on is
> Unicode support with database systems and Python DB-API modules.

Interesting indeed.

> Some
> modules don't seem to support Unicode (pyPgSQL, for example), whereas
> others apparently rely on the database system for such support to
> actually function (eg. mxODBC).

Last time I checked, mxODBC was the only one to support Unicode at all.
The reason I checked was that I want to add Unicode support to pyPgSQL.
A draft patch is available at pyPgSQL's Sourceforge patch manager.

I'll try to revive that effort soon on the pyPgSQL mailing list, as I
need some input for that.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 7.3 ?C      Wind: 2.9 m/s



From maxm at mxm.dk  Sat Apr 13 14:13:51 2002
From: maxm at mxm.dk (Max M)
Date: Sat, 13 Apr 2002 20:13:51 +0200
Subject: FileUpload via CGI by Python
References: <slrnabgabq.gjo.jarvin24@tutor.cc.tut.fi>
Message-ID: <3CB8755F.9050207@mxm.dk>

You can use the standard cgi module.

regards Max M

J?rvinen Petri wrote:
> Hey, i'm trying to do form that uploads file to server.
> 
> Is there any ready solutions?
> 
> - Petri



From just at xs4all.nl  Mon Apr  8 14:18:57 2002
From: just at xs4all.nl (Just van Rossum)
Date: Mon, 08 Apr 2002 20:18:57 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com> <a8slgr$pin$1@nyheter.chalmers.se>
Message-ID: <just-112AD2.20185708042002@news1.xs4all.nl>

In article <a8slgr$pin$1 at nyheter.chalmers.se>,
 jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen) wrote:

> In article <3CAF524F.63E46C81 at alcyone.com>,
> Erik Max Francis  <max at alcyone.com> wrote:
> >Jacob Hallen wrote:
> >
> >> Python is a language that is weakly typed, ...
> >
> >You mean dynamically typed, not weakly typed.
> 
> I have been brought up in an environment which called anything
> allowing mixed types in a list weakly typed.

The two are orthogonal; have a look at the bottom of this page:

  http://diveintopython.org/odbchelper_funcdef.html

Just


From johnroth at ameritech.net  Tue Apr  2 21:27:22 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 2 Apr 2002 21:27:22 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org>
Message-ID: <uakpqdot127bc7@news.supernews.com>

"Guido van Rossum" <guido at python.org> wrote in message
news:3CAA5777.66E7C2D5 at python.org...
> Alex Martelli wrote:
> > I _have_ seen old-time Fortranites rant quite convincingly against
that
> > useless, newfangled LOGICAL stuff.
>
> And that is exactly what many of the responses in this thread
> (including Laura's over-long FUD piece) sound like.
>
> I'm reading (at least skimming) everything, but the more I read the
> reactions, the more I believe that this is just the response I have to
> expect for *every* change to the language I propose, and the more I'm
> tempted to follow my gut instincts.  Long ago (right after lambda :-)
> I learned to say "no" to proposals that didn't feel right to me, even
> if I couldn't express very clearly why.  I think it's time I learn to
> say "yes" even if the community doesn't see the light yet.
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)

Right on. I've got a couple of minor issues, but
you're the language architect. You've turned out
a very good product to date; I'm more than willing
to go along and see how the next idea turns out,
even if I don't agree with you on the integer divide
thing.

John Roth




From drew_csillag at yahoo.com  Tue Apr  9 12:06:27 2002
From: drew_csillag at yahoo.com (Drew Csillag)
Date: Tue, 9 Apr 2002 12:06:27 -0400
Subject: [ANNOUNCE] SkunkWeb 3.2 Released!
Message-ID: <20020409120627.A1201@localhost.localdomain>

SkunkWeb 3.2 has just been released.

The release is available from the SkunkWeb home page at
	http://skunkweb.sourceforge.net

or directly from 
        http://prdownloads.sourceforge.net/skunkweb/skunkweb-3.2.tgz

What's new? (from 3.1.3)
------------------------------
* New VFS layer which makes it now possible to serve portions (or all)
  of your document root from regular filesystems, zip archives or tar
  archives 
* Works with Python 2.2
* Confirmed to work on Windows with Cygwin
* Many bugfixes
* New product service and a wizard tho produce them
* CGI adaptor 
* New demo
* If you have PIL install, you can have your hight and width
  attributes of image tags generated by <:img:> automatically filled
  in
* Many PyDO improvements, including a SAP driver
* Upgraded MySQL support


Documentation updates are in the works.

Very best wishes to all,

-Drew




From phd at phd.pp.ru  Tue Apr  9 14:21:40 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 22:21:40 +0400
Subject: yet another "Here's to 1.5.2" post
In-Reply-To: <upu18x4h3.fsf@online.no>; from syver-en+usenet@online.no on Tue, Apr 09, 2002 at 03:44:20PM +0000
References: <mailman.1018132589.8324.python-list@python.org> <tnys8.11$p56.2333@newsb.telia.net> <upu18x4h3.fsf@online.no>
Message-ID: <20020409222140.D27632@phd.pp.ru>

On Tue, Apr 09, 2002 at 03:44:20PM +0000, Syver Enstad wrote:
> Am I the only one in the entire world that thinks '\t'.join(mylist) is
> elegant and natural?

   No, this is ugly. One of the worst syntactic things in the lnaguge.

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




From phr-n2002a at nightsong.com  Wed Apr 24 08:37:32 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 05:37:32 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com>
Message-ID: <7xsn5l6zn7.fsf@ruckus.brouhaha.com>

Richard Parker <richard at electrophobia.com> writes:
> >> the secret prefix method to construct a MAC from a hash function, i.e.
> >> MAC(x) = H(K || x).  This method is generally considered to be insecure.
> >> Use the HMAC construction instead....
> 
> The secret prefix method is vulnerable to an "appending" attack.  An
> attacker given a message-MAC pair can compute, without the key, the correct
> MAC for a new message that has the original message as a prefix.  

Ouch!  That is nasty and I'd missed it.  Thanks!

How about if I just hash a second time, i.e. H(H(K || x)) or 
H(K || H(K || X)), rather than coding up the full HMAC machinery?
Although, HMAC in Python using string multiplication instead of loops
might not be so slow.  I'll see if I can code it that way.

> The portions of your code that use H() just as a hash function don't need to
> use HMAC.

Good point.  


From mcherm at destiny.com  Mon Apr  1 20:14:14 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 01 Apr 2002 20:14:14 -0500
Subject: "Don't Kill Opera" documentation release
Message-ID: <3CA905E6.5070109@destiny.com>

> I've just made an "emergency" release of the 2.2 documentation (based
> on the release22-maint branch in CVS.  If you use Opera 6.01, you can
> now feel a little safer on python.org.  ;-)
> 
> 
>   -Fred

Thank you very much for considering this worthy of "emergency" 
treatment. I'm a Mozilla user myself, but I think Opera is great, and 
shouldn't be needlessly broken.

-- Michael Chermside




From marklists at mceahern.com  Fri Apr 12 21:40:27 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 12 Apr 2002 20:40:27 -0500
Subject: fp/lambda question
In-Reply-To: <a97s9h$gfg$1@reader07.wxs.nl>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMENFKKAA.marklists@mceahern.com>

[Jozef]
> I am not really sure even what this means... does lambda create a new
> scope or something? Is there another way to write this?

What you wrote seems to work for me.  I'm using Python 2.2.1.  What version
are you using?  nested_scopes is new with 2.2, but available in 2.1 with
from future import, true?

// mark





From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 19:29:15 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 23:29:15 +0000 (UTC)
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com> <5hIt8.10742$b62.296832@news1.tin.it>
Message-ID: <slrnabereb.mfc.huaiyu@gauss.almadan.ibm.com>

Alex Martelli <aleax at aleax.it> wrote:
>Huaiyu Zhu wrote:
>
>> Alex Martelli <aleax at aleax.it> wrote:
>> 
>>>Personally, I'd rather see curry built-in -- it IS easy to write curry
>>>in Python 2.2, but it's such a wonderful idiom (and would replace so
>>>many lambda's;-) that I'd love to see it officially blessed.  Not that
>>>I expect it to be (and nor will my heart be broken because of this:-).
>> 
>> Do you have in mind a special syntax for curry?
>
>http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549
>
>see my curry-by-closure comments.  

Cool.  When do we expect it to occur in standard distribution?  :-)

I think that having positional
>args given at call-time concatenated after those given at curry-time
>("if any" in either case), and kw args given at call-time taking precedence 
>over those given at curry-time (ditto), is most natural and useful.
>I think the syntax you propose is equivalent (for positional args --
>you don't mention kw args).

Yours is more advanced.

Huaiyu


From m.faassen at vet.uu.nl  Tue Apr 16 05:36:16 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 16 Apr 2002 09:36:16 GMT
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <a9grag$b2e$1@newshost.accu.uu.nl>

phil hunt <philh at comuno.freeserve.co.uk> wrote:

> OK, I've decided to write a PEP. Preliminary to this, I'd like 
> people's comments.

> I wish to make the pas startement optional. That is, in any Python 
> program where there is a pass it can be left out and the meaning of 
> the program is unchanged.

> There are two rationales for this: convenience and beauty.

> Thoughts?

-1

I don't think it's convenient (what's an extra line, and it occurs very
rarely in my code..), and I don't think it's pretty either.
My eyes would trip over it as much as single line if:

if foo: bar

which I think is ugly. I need the visual hint of an indented block:

if foo:
    bar

the same holds for pass:

def foo():
   pass
def bar():
   pass

is a lot easier for my eyes to parse than:

def foo():
def bar():

which leaves me with little hint of what is going on. This looks like a single
block, not two, from a distance/in the corner of my eyes. I need to read the
text instead of just look at its shape in order to understand what is going
on.

'pass' is pretty rare, and it needs to stand out when it's around. If I'm
defining an interface where there is no method body, I'll just drop in a
docstring; a good idea for an interface in any case.

Regards,

Martijn 
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From aleax at aleax.it  Tue Apr 30 16:38:52 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 20:38:52 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020172854.26653.python-list@python.org> <Xns9200AC87B7A86duncanrcpcouk@127.0.0.1> <uctp0ugohd6ib1@news.supernews.com>
Message-ID: <whDz8.35683$8D3.1031080@news1.tin.it>

John Roth wrote:
        ...
> Doesn't matter. What we need is <sequence>.join(<string>)
> where the resulting string is the result of applying str() to each
> of the members of the sequence, with the <string> between
> each pair of members. In other words, .join() is a member
> function of all objects that have a sequence interface.

Yes, this is how just about every naive designer approaches the
question.

Sure makes me glad I'm using a language designed by a guy
that doesn't fall for such easy traps.

I've explained more than once on this forum why, given the fact
that for any x.y(z) Python fully polymorphizes on x but can only
polymorphize through "use via a standard protocol" on z, it would
be a *serious design error, substantially diminishing the overall
power of the language system* to architect it your way.  Google
will no doubt help you find my posts on this if you care.

It's no doubt an interesting issue, for those who care -- why is
a futile wish for this particular design error so widespread.  Oh well.


Alex



From grante at visi.com  Wed Apr  3 00:21:57 2002
From: grante at visi.com (Grant Edwards)
Date: Wed, 03 Apr 2002 05:21:57 GMT
Subject: chdir() problem
References: <b5b450db.0204022107.37bc0a25@posting.google.com>
Message-ID: <slrnaal4bl.n9a.grante@isis.visi.com>

In article <b5b450db.0204022107.37bc0a25 at posting.google.com>, joe wrote:

> within python 2.2, chdir() works fine.
> 
> however, i want to be able to change the "active" directory, so that when
> the script completes, the user is left in a target directory (which is not
> the original directory)

Can't be done.  The chdir() call only affects the process that's running the
Python interpreter.  The current directory of parent process is unaffected.

When the script ends, that process (with the new current directory)
vanishes.  The "user's" process (the parent of the Python process) can't
have it's environment changed by the child process that's running Python.

-- 
Grant Edwards                   grante             Yow!  Is it 1974? What's
                                  at               for SUPPER? Can I spend my
                               visi.com            COLLEGE FUND in one wild
                                                   afternoon??


From tim.one at comcast.net  Sat Apr 20 14:14:33 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 20 Apr 2002 14:14:33 -0400
Subject: Python's Lisp heritage
In-Reply-To: <a9s72u$gjj$1@slb3.atl.mindspring.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCOELNOPAA.tim.one@comcast.net>

[Andrew Dalke, on cons cells vs contiguous vectors for list storage]
> ...
> So algorithms which might run fast under Lisp do not under Python,
> and vice versa.  The decision for this difference was made after
> observations in ABC that most list use was through indexing, and not
> through a direct decision to be different from Lisp.

All lists in ABC were sorted -- you couldn't stop that.  If you wanted to
retain the order of insertion, you needed to insert explicit (index, value)
pairs, or build a dict explicitly mapping index to value.  ABC used balanced
binary trees (AVL style) under the covers for list storage.  So it was
neither Lisp-like nor Python-like in this respect.  It aimed at log N
worst-case time for all common list operations (whether indexing or
searching), but at the cost of log N expected-case time too.  Overall, ABC
didn't have "bad cases", but it was hard for anyone save a theoretician to
guess that since *everything* ran at bad-case speed <wink>.

doubting-that-cons-cells-were-even-considered-ly y'rs  - tim





From martin at v.loewis.de  Sat Apr 13 19:09:44 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 01:09:44 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com>
Message-ID: <m34rif19hj.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> You didn't read my post well enough-- we can already _DO_
> that with Python v2.2.x:

I read your post. You said that the idiom you bring up is not clear to
you; I agree - adding a "not" just to get a 0/1 value is not clear.

To me, using a function called "bool" is very clear.

> Everyone_is_entitled_to_have_my_opinion'ly yours,

Certainly. Did I suggest otherwise?

Regards,
Martin


From ASkwar at DigitalProjects.com  Thu Apr 25 11:42:01 2002
From: ASkwar at DigitalProjects.com (Alexander Skwar)
Date: Thu, 25 Apr 2002 17:42:01 +0200
Subject: Faster md5/1-way encryption?
In-Reply-To: <3CC7317B.3040208@rogers.com>
References: <20020424215752.GC15371@teich.Garten.DigitalProjects.com> <3CC7317B.3040208@rogers.com>
Message-ID: <20020425154201.GD21205@teich.Garten.DigitalProjects.com>

?Mike C. Fletcher? sagte am 2002-04-24 um 18:28:11 -0400 :
> Would likely be much better as:

That's not true.  It's not only much better, it's astonishngly,
extremely much better!  Instead of ~6000 seconds for a 190k line file, I
now only need about 6 seconds!

Thank you *VERY* much!

Alexander Skwar
-- 
How to quote:	http://learn.to/quote (german) http://quote.6x.to (english)
Homepage:	http://www.iso-top.de      |    Jabber: askwar at a-message.de
   iso-top.de - Die g?nstige Art an Linux Distributionen zu kommen
                       Uptime: 1 day 10 hours 41 minutes




From ljohnson2 at knology.net  Mon Apr  8 11:04:34 2002
From: ljohnson2 at knology.net (Lyle Johnson)
Date: Mon, 8 Apr 2002 10:04:34 -0500
Subject: swig-problem
References: <3cb1ae14$0$12336$9b622d9e@news.freenet.de>
Message-ID: <ub3cbpdmtl65f2@corp.supernews.com>

Use the "-o" flag for SWIG to change the output file name:

    swig -c++ -python -o Tree_wrap.cpp Tree.i

Hope this helps,

Lyle




From fperez528 at yahoo.com  Wed Apr 17 17:59:45 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 17 Apr 2002 15:59:45 -0600
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <Nefv8.30383$vF6.948633@news2.tin.it>
Message-ID: <a9krad$1j8$1@peabody.colorado.edu>

<posted & mailed>

Alex Martelli wrote:

> Your intuition, which led you astray here (Python does just
> what it should do), can be trained in several ways.  The
> works of J. L. Borges and I. Calvino, if you like fiction
> that's reasonably sophisticated but still quite pleasant,
> are good bets.  If you like non-fiction written by
> engineers fighting hard to dispell some of the errors of
> philosophers, Wittgenstein and Korzibsky are excellent.
> 
> I'm not kidding, but I realize that many Pythonistas don't
> really care for either genre.  In which case, this group
> and its archives, essays by GvR and /F, and the Python
> sources, may also prove interesting reading.
> 

Well, I certainly hope your post encourages some to get acquainted with two of  
the most elegant and fantastic writers of the 20th century. Unfortunately I 
don't read Italian so I need to read Calvino translated, but there are few 
joys in life which compare to reading Borges in his native Spanish 
(especially his poetry; by the way, I think much of his work is by now 
available in English). 

And I would argue that the writings from these two gentlemen _should_ appeal 
to the logically oriented mind of any programmer worth his salt.

You made my day with a reference here to some of my deepest 'intellectual 
loves' :)

Regards,

f.


From gmcm at hypernet.com  Sun Apr 28 10:34:00 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 28 Apr 2002 14:34:00 GMT
Subject: Help: py2exe and os.path
References: <JlPy8.72016$vF6.2149385@news2.tin.it>
Message-ID: <Xns91FE6B9995466gmcmhypernetcom@199.171.54.213>

Utente wrote:

[snip]

> py2exe works fine, but i get the following error (this is the only
> error in a long source!!!):
> 
> warning: py2exe:
> ***************************************************************
> **********
> warning: py2exe: * The following modules were not found:
> warning: py2exe: *   os.path

>From import's viewpoint, there is no module named os.path, so
don't import it.

When you import os, it locates the appropriate module for doing
path manipulations on your platform, and makes this available
as "path" (could be posixpath, macpath...). Since it's a global
of os, it's available as os.path.

-- Gordon
http://www.mcmillan-inc.com/


From jimmy at retzlaff.com  Mon Apr  1 02:28:29 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Sun, 31 Mar 2002 23:28:29 -0800
Subject: Call for kjBuckets.c [was Re: relational database?]
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F5612@MAYOR.retzlaff.com>

> Richard Jones [mailto:rjones at ekit-inc.com] wrote
>
> Does anyone have a copy of kjbuckets.c that they could send me?

Oleg Broytmann has it on a page mirrored in a couple of places. Look at
the bottom of either:

http://phd.by.ru/Software/Python/
or
http://phd2.chat.ru/Software/Python/

Jimmy



From gerhard at bigfoot.de  Sat Apr  6 19:14:30 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Sun, 7 Apr 2002 02:14:30 +0200
Subject: 2.2 schizo win32 linkage
In-Reply-To: <uTUaTJAJ83r8EwkA@jessikat.fsnet.co.uk>
References: <naBIBIAc1ur8EwVD@jessikat.fsnet.co.uk> <slrnaauuj6.tl.gerhard@lilith.my-fqdn.de> <uTUaTJAJ83r8EwkA@jessikat.fsnet.co.uk>
Message-ID: <20020407001430.GA1545@lilith.my-fqdn.de>

* Robin Becker <robin at jessikat.fsnet.co.uk> [2002-04-07 00:04 +0100]:
> In article <slrnaauuj6.tl.gerhard at lilith.my-fqdn.de>, Gerhard H?ring
> <gerhard at bigfoot.de> writes
> ...
> >
> >The best solution I can think of is for dependencies of an extension
> >module to place them in the same directory as the extension module:
> >this is where Windows searches first when trying to satisfy the DLL
> >dependency.
> so where should Python's tcl/tk dlls live. Other extensions also depend
> on them.

But these other extensions depend on tkinter, not on the tcl/tk dlls
directly, right?

These DLLs should go in the same directory as _tkinter.pyd, which should
be in site-packages.

Gerhard




From sholden at holdenweb.com  Thu Apr 25 06:23:04 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 25 Apr 2002 06:23:04 -0400
Subject: cgi script to study tarballs
References: <3CC7844D.ED73E83C@alum.mit.edu>
Message-ID: <IQQx8.11800$z55.3038@atlpnn01.usenetserver.com>

"Will Ware" <wware at alum.mit.edu> wrote ...

> This script is incredibly slow with large tarballs, but it
> does work.
>

I can see why you mention it's speed, but what makes you think we want to
know how well-hung your software is.

c.l.py-just-isn't-ready-for-that-ly y'rs  - steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From gcordova at hebmex.com  Thu Apr  4 10:55:53 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 4 Apr 2002 09:55:53 -0600 
Subject: Downloading pages from a Windows Intranet
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062334CFB61@mail.hebmex.com>

> 
> Trying to use urllib or urllib2 to get a page from our Intranet
> (Windows) gives me an Access Denied error.  The funny thing is that on
> some pages it works (?).  Any way, someone found a Perl module
> (Win32::Internet) that creates an object oriented interface to the
> Win32 Internet Functions (WININET.DLL).  With it we can get any page
> from our intranet, even asp pages.  Is there an equivalent module or
> functions that I could use from Python?  Or can anyone give me an
> example of how to work around this?  Or should I just settle for a
> Perl/Python solution?
>

This is really strange, because here at work urllib and urllib
work like a charm, I've used them to prove my point about python
being a powerful tool (instead of a curious toy) more than once,
by showing that very few python lines can create a powerful tool
if used properly.

So, unless you're using some kind of authentication which is not
part of the http protocol, we can't really help you much unless
you give us a bit more information.

Check your internet connection settings, show us a bit of code,
what error tracebacks do you see? etc.

Salutations.

-gustavo




From tdelaney at avaya.com  Mon Apr 29 20:43:37 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 30 Apr 2002 10:43:37 +1000
Subject: Comment on draft PEP for deprecating six builtins
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2AB@natasha.auslabs.avaya.com>

From: James J. Besemer [mailto:jb at cascade-sys.com]

> Incidentally, by my count, *I* am the only victim of an 
> unprovoked personal attack on this list today.

Damn! 

Alex - Your mother was a hamster, and your father smelled of elderberries.

Steve - I fart in your general direction.

Laura - Go away or I shall taunt you another time!

Tim Delaney




From sholden at holdenweb.com  Mon Apr 29 08:57:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 29 Apr 2002 08:57:16 -0400
Subject: input/output through socket or COM port
References: <3CCB1CE9.3E08F027@ix.netcom.com> <Xns91FE207CF4D44cliechtigmxnet@62.2.16.82> <3CCB557C.5BDB495E@ix.netcom.com> <Xns91FEB6CCA14DBcliechtigmxnet@62.2.16.82>
Message-ID: <Qtbz8.7690$M63.1803@atlpnn01.usenetserver.com>

The "standard reference" to this would be Gordon McMillan's webtorial[1]

have fun
 Steve


[1]    http://www.mcmillan-inc.com/sock1.html
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/


"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns91FEB6CCA14DBcliechtigmxnet at 62.2.16.82...
> Jive Dadson <jdadson at ix.netcom.com> wrote in
> news:3CCB557C.5BDB495E at ix.netcom.com:
>
> > Thanks a lot.  I snarfed the pserial zip file.  Do you know a similar
> > thing for
> > a TCP/IP host?
>
> s = socket.socket(...)
> s.connect(...)
> fi = s.makefile()
>
> --
> Chris <cliechti at gmx.net>
>





From op73418 at mail.telepac.pt  Mon Apr 22 16:09:42 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 22 Apr 2002 21:09:42 +0100
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <3CC45005.2E9C0F86@tundraware.com> <b8p8cu8f7u9lkbmpcrksjfcqp67nkoq3om@4ax.com> <aa1q95$eb5$1@panix1.panix.com>
Message-ID: <per8cu8p2uo3ir33dfg4vge0nr36vlolr1@4ax.com>

On 22 Apr 2002 16:02:45 -0400, aahz at pythoncraft.com (Aahz) wrote:

>In article <b8p8cu8f7u9lkbmpcrksjfcqp67nkoq3om at 4ax.com>,
>Gon?alo Rodrigues  <op73418 at mail.telepac.pt> wrote:
>>
>>Ahh, OK. Then I agree completely -- Mathematics *is* larger than its
>>would-be applications to the natural sciences. I think it was P. Erdos
>>who remarked once that all natural science is conceivably finite - e.g.
>>there would be a time when there would be nothing more to learn about
>>the universe - but mathematics is literally infinite.
>
>I'm not so sure about the latter assertion, and I'm *definitely* not
>certain about the assertion that natural science is finite.

The first assertion by Erdos is indeed problematic. The second is not.

Best,
Gon?alo Rodrigues


From siegfried.gonzi at kfunigraz.ac.at  Wed Apr 10 04:31:02 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 10 Apr 2002 10:31:02 +0200
Subject: using NumPy for histogram plot
References: <3cb358b4@news.isc.rit.edu> <a8vof9$rmt$2@peabody.colorado.edu>
Message-ID: <3CB3F846.DE4D79BB@kfunigraz.ac.at>

Fernando P?rez wrote:

 
> google(Konrad Hinsen Scientific Python)
> 
> It has a histogram implementation which is very nice. Incidentally, it can be
> plotted straight out with grace's grace.histoplot(), but I don't know if
> grace runs under Windows or not. But at least this will get you the
> histogram.

A scond histogram version:

http://groups.google.com/groups?q=histograms+group:comp.lang.python.*&hl=en&selm=m3n23pt9u6.fsf%40visar.physast.uga.edu&rnum=9

It just calculates the histogram values (it is quite the same as Hinson
his version; but the returned object is not that weird). I did not try
it but the histogram can possibly be displayed with the free high
quality plotting library DISLIN*.

S. Gonzi
*Concerning Windows: istalling DISLIN is a task of a few minutes and it
then just works properly from the start on.


From lac at strakt.com  Sat Apr  6 15:44:21 2002
From: lac at strakt.com (Laura Creighton)
Date: Sat, 6 Apr 2002 22:44:21 +0200
Subject: I need to go to Malmö tomorrow
Message-ID: <200204062044.g36KiLTk017922@ratthing-b246.strakt.com>

I haven't forgotten you all, but it will be a while before I get to replying.
Alex and I have code to write together as well.   I get a week of pair programming
with Alex!  Way Cool.

Laura




From dale at riverhall.NOTHANKS.co.uk  Mon Apr 22 13:19:09 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Mon, 22 Apr 2002 18:19:09 +0100
Subject: Encryption
Message-ID: <70h8cugniao40148h76fv4o7bgci2j3ddd@4ax.com>

We need to decrypt, on Win32, some data coming from a Linux system,
encrypted with mcrypt.

I'd like to do it in Python but a command line utility would be
acceptable.

I found the python-mcrypt module on Source Forge but it needs
compiling and I couldn't find all the include files. Also, there seems
to be an issue with Python >= 2.1. 

Mcrypt can use a number of encryption techniques and I'm not too
bothered which we use. Just as long as I can decypher it at our end.

Any suggestions?
--
Dale Strickland-Clark
Riverhall Systems Ltd



From joonas at olen.to  Fri Apr 12 07:05:43 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Fri, 12 Apr 2002 11:05:43 GMT
Subject: Communication between a Python process and a Python cgi script
References: <3CB5E224.4E3DC5BA@olen.to> <a94rsq$29e8$1@nntp6.u.washington.edu>
Message-ID: <3CB6C0A1.2030800@olen.to>

Donn Cave wrote:
> Quoth Joonas Paalasmaa <joonas at olen.to>:
> | I have a Python program started by root that does some magic stuff
> | every 10 minutes. So the program in an infinite loop of sleep, do stuff,
> | sleep, do stuff... How can I make a cgi-script to end the process's
> | sleep phase and make it to "do stuff" immediately. I tried signals, but
> | it is propably too risky because the process is run as root. Any
> | suggestions?
> 
> You could try a named pipe -
> $ mkfifo /tmp/fifo
> $ cat /tmp/ptest
> pipe = open('/tmp/fifo', 'r')
> while 1:
>     ping = pipe.read()
>     print 'ping', repr(ping)
> $ python /tmp/ptest &
> $ echo junk > /tmp/fifo
> ping 'junk\n'
> $ echo junk > /tmp/fifo
> ping 'junk\n'
> 
> Find a better name and place for the fifo, check that it will be
> writable by the CGI process.  Consider security issues, who will
> be able to do this - might be no compromise, yet you could have
> a "denial of service" kind of attack.

Thanks! This is exactly what I needed.

- Joonas



From tim.one at comcast.net  Tue Apr  2 16:29:41 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 02 Apr 2002 16:29:41 -0500
Subject: The Python Way
In-Reply-To: <20020402115427.GC1081@andy.tynant.ftech.net>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAEKGCOAA.tim.one@comcast.net>

[Andy Gimblett]
> So, is Aahz/Tim saying that Queue.Queue is "better" in general than
> threading.RLock ?  If so, why?

It's almost impossible to get yourself in trouble using Queue.  It's almost
(but not quite as) impossible *not* to get yourself in trouble rolling your
own protocols by hand via raw locks.  If you can't use a Queue, learn how to
use conditional variables correctly -- they support a protocol that's very
robust, but needs to be learned.  Locks are fine for pure, unconditional
mutual exclusion, but even then, sooner or later you're going to forget to
code the try/finally wrapper and end up baffled by deadlock.

it's-not-a-question-of-"better"-unless-you-assume-sanity-is-better-ly
    y'rs  - tim




From Randy.L.Kemp at motorola.com  Thu Apr  4 09:43:47 2002
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Thu, 4 Apr 2002 08:43:47 -0600 
Subject: Java to Python translation
Message-ID: <81F59ACE9063D511A5BB0002B328D22603050DD2@il93exm20.css.mot.com>

What the heck kind of application has 100 MB of Java code anyway? 

Boy, that is a good question.  Let's look at the size of some applications I have on my Windows 2000 machine:
JBoss-2.4.4_Tomcat-4.0.1  -  36.1 MB -- Jboss EJB server and Tomcat servlet container
Jonas244                  -  23.0 MB -- Jonas EJB server
mysql                     -  26.3 MB -- Mysql Database server
Resin-2.1.0               -  17.4 MB -- Resin servlet container    
Total space               - 102.8 MB
Now the sum of these products equals about 100 MB of Java + C code

-----Original Message-----
From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
Sent: Thursday, April 04, 2002 8:27 AM
To: python-list at python.org
Subject: Re: Java to Python translation


vio <vmilitaru at sympatico.ca> writes:
> > This makes no sense.  WHY do you want to migrate all that Java to Python?
> > Why not leave it in Java and keep using it?  That's what none of us
> > are understanding.  What do you GAIN from making an unmaintainable
> > machine translation?
> 
> Focus.

> Focus on one language and one environment. Which happens to be
> python. Migrating legacy code is a strategic decision. Obviously,
> the goal is to NOT end up with an unmaintaibable machine
> translation. In that case the "solution" becomes worse than the
> "problem" (one of python's big appeal is its ease of
> maintenance).

You won't get that from an automatic translation.  Part of the ease
of maintenance comes from the code for an app just being smaller in
Python than it would be in Java.  But a translation would be bigger
and not smaller.  And the code output will be awful, because the
translation system won't have any way to figure out what the programmer
was actually thinking.  Look at a Babelfish translation of French
into English to see what you'll be dealing with code-wise.  This
is what I mean when I say that the translation project might fail
disastrously, but if it succeeds, you'll have an even bigger disaster.

What the heck kind of application has 100 MB of Java code anyway?
It sounds like it may already be the output of some automatic process.






From andrew.henshaw at gtri.gatech.edu  Mon Apr  1 08:20:24 2002
From: andrew.henshaw at gtri.gatech.edu (Andrew Henshaw)
Date: Mon, 01 Apr 2002 08:20:24 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <uagnl8q1pf7930@corp.supernews.com>

Guido van Rossum wrote:

> I offer the following PEP for review by the community.  If it receives
> a favorable response, it will be implemented in Python 2.3.
> 
>
I weight this as -1 for inclusion in Python 2.3.  I'm reserving judgement 
for later adoption.  

I was hoping that the 2.3 release would not bring with it such a major new 
feature.  

Andrew Henshaw


From prova at qui.com  Sun Apr 28 09:37:11 2002
From: prova at qui.com (Utente)
Date: Sun, 28 Apr 2002 13:37:11 GMT
Subject: Still on py2exe
Message-ID: <bWSy8.29406$8D3.782924@news1.tin.it>

Excuse me for bothering...
i'm interested in py2exe, but i experience some strange (for me...)
behavior.
Everything works fine running the .pyw (or the .py) main file. I have some
modules, and the application is build with wxPython (no problems for py2exe
with wxPython anyway).
One of the modules reads and writes a dictionary on a file with the shelve
module. And it works fine.
When i make the "exe" with py2exe everything goes well, the application
runs, but when i try to load data (either with an absolute or a relative
path) it stops. The file exists, the application works fine when iterpreted
but stops whit no error messages or anything when executed from the exe.
I really don't understand why, and i don't see a solution for my problem.
Can anybody give some suggestion?

Thanks, Max




From aleax at aleax.it  Thu Apr 18 15:55:11 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 18 Apr 2002 19:55:11 GMT
Subject: why no 'length' method in sequences ?
References: <a9mqn3$5ao$1@aphrodite.grec.isp.9tel.net>
Message-ID: <zwFv8.36236$vF6.1097456@news2.tin.it>

Richard Gruet wrote:

> Hi all
> 
> I wonder why there is no 'length' (or 'len') method - or attribute - in
> the sequences types. Having to use the built-in function len() doesn't
> sound very Object-Oriented to me. I understand that the built-in function
> len() must be kept for compatibility reasons, but why not to add also the
> method len() to the sequence types ? Is there a reason ?

I would be a silly burden on people who want to implement a sequence
type, to make them define a method called len that is an exact synonym
for the existing special method __len__.  And it would be an intolerable
burden, conceptually and in terms of performance, if looking up an
attribute named 'len' performed strange black-magic contortions (even
worse if those contortions were specialized for only the cases in which
said attribute is meant to be called -- remember Python does not draw
deep distinctions between attributes meant to be called, and others).

It may not SOUND "very Object-Oriented", but having special methods
"clothed" by means of built-ins and operators is a clever and subtle
set-up, as I tried explaining in the other post to this thread.  And Python
has always preferred the steak to the sizzle, whenever choice was needed:
part of what makes it such a great language.


Alex



From ds_chitale at hotmail.com  Fri Apr 19 17:09:18 2002
From: ds_chitale at hotmail.com (Dhananjay Chitale)
Date: Fri, 19 Apr 2002 21:09:18 +0000
Subject: Urgent : Installation of Python 2.0 on sun Solaris
Message-ID: <F242dmahRaEryDGB30r000013ca@hotmail.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020419/5480d491/attachment.html>

From gustafl at algonet.se  Sun Apr 21 17:48:52 2002
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Sun, 21 Apr 2002 21:48:52 +0000 (UTC)
Subject: Pyxie, anyone?
Message-ID: <Xns91F7F276A63DAgustafl@195.100.94.182>

I have problems using the pyxie library for processing XML. This page

  http://pyxie.sourceforge.net/

says that "To run Pyxie, you need two things:

- A Python distribution
- The XML package for Python [...]"

I have the latest versions of both and still I get the error "ImportError: 
No module named pyxie" when I try to import the module. Has Pyxie been 
phased out from the XML package?

Gustaf


From johnroth at ameritech.net  Tue Apr  2 21:21:27 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 2 Apr 2002 21:21:27 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org>
Message-ID: <uakpepcn1iir15@news.supernews.com>

"Laura Creighton" <lac at strakt.com> wrote in message
news:mailman.1017755045.4308.python-list at python.org...
> Dear Guido:
>
> I would first like to thank you for giving us an opportunity to
respond.
> I have spent most of the weekend thinking and writing a reply to this,
> and I think that this has made me a better teacher.  For this I am
> grateful.  I realise that this is rather long, but I have condensed
> it substantially from my efforts earlier this weekend.   Thank you
> for taking the time to read this.
>
> Laura Creighton
>
> -----------------------------
>
I'm not going to quote this, as it's already too long. I read up to
the diatribe on the Swedish leap year, and I went AAAARRRRGGGH!

I really don't care if you want the silly function to return
a Boolean, the number of leap days in the year, or the name
of the idiot parliamentarian who decided to fix the calendar
so that 1712 was different from the rest of Europe.

HOWEVER, if you want true or false as the return
value, the name should be something like 'isLeapYear()'.
If you want the number of leap days in the year, then
it should be named something like 'numberOfLeapDays().'

Sorry to make this sound like a flame, but I try to
practice XP. The first time I ever come across a
function named "isSomething" that returns a conceptual
integer, it's going to get refactored, right then and
there to a name that actually expresses what it does.

John Roth




From erno-news at erno.iki.fi  Mon Apr  8 01:47:59 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 08 Apr 2002 08:47:59 +0300
Subject: purging file buffers - to make new data be written to file
References: <a8qhjp$7go$1@news6.svr.pol.co.uk>
Message-ID: <kuadseoi5s.fsf@lasipalatsi.fi>

In article <a8qhjp$7go$1 at news6.svr.pol.co.uk>, ddoc
<news at 92tr.freeserve.co.uk> writes:

| Can I do that from within a python script, or do I have to give a command 
| line argument to make Python start without buffers?

| On Linux.  The program is to save a memo in a slightly organised format, 
| but I find the file is not written until Python exits.

f.flush() will commit any data that python and the c library might
have not yet written to the operating system. this will let
other processes in the system see the new data.

if you want to make sure the data ends up on disk physically,
use os.fsync(f.fileno()).

  -- erno


From peter at engcorp.com  Tue Apr  9 01:02:35 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 09 Apr 2002 01:02:35 -0400
Subject: What has PEP 285 done to us?
References: <mailman.1018299317.16035.python-list@python.org>
Message-ID: <3CB275EB.48058A8B@engcorp.com>

Cliff Wells wrote:
> 
> This list is known for being particularly helpful and friendly.  Let's keep it
> that way.  If you want to argue technical issues and programming techniques,
> this is a good place.  If you want to fight and bicker, take it to the
> playground (or maybe Yahoo!'s Fight Room - the people there won't know what the
> hell a boolean is, but they'll be more than willing to exchange insults with
> you over it, because you don't know either, you moron ;)

+1 to that idea.

-Peter


From gbreed at cix.compulink.co.uk  Wed Apr  3 09:06:07 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Wed, 3 Apr 2002 14:06:07 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <ualr7hmlhkemab@news.supernews.com>
Message-ID: <a8f28f$7p9$1@thorium.cix.co.uk>

John Roth wrote:

> I don't think that's the case here. This proposal opens up a whole
> can of worms involving the concept of 'truth', that is, what the
> conditional and logical operators see as 'true' or 'false'.

Then no wonder it's so controversial!  This question has occupied 
philosophers for centuries, and is unlikely to be resolved in time for 
2.3.

                 Graham


From weeks at vitus.scs.agilent.com  Wed Apr 17 12:15:58 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Wed, 17 Apr 2002 16:15:58 +0000 (UTC)
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com>
Message-ID: <1019060158.737451@cswreg.cos.agilent.com>

dominikush at yahoo.com wrote:
: >>> dict = {'a':[1],'b':[2]}
: >>> list = dict.values()
: >>> dict['a'].append(3)
: >>> list
: [[1, 3], [2]]

: Who is wrong here: my intuition or Python (2.2)?

As others have said, Python is right on this.  However, Python aids you in
your confusion, as follows:

There should have been an operator in Python that says when objects are
conceptually the same.  For the moment, call it "eq".  The following would
hold:

    3 eq 3
    not (3 eq 4)
    a eq a		# PROVIDED a HAS A VALUE
    "foo" eq "foo"
    not ([] eq [])
    not ({} eq {})
    () eq ()

Furthermore, the behavior of "eq" for class instances should be user
controllable.  Typically, immutable objects would be "eq" when their parts
are "eq", while mutable objects would be "eq" only when one "is" the other.

Furthermore, two "eq" objects should hash to the same value.  So *any*
object could be a hash key.

Python almost has an "eq" operator.  "==" is like "eq" with two exceptions:

    lists
    dictionaries

Since "==" is so close to "eq", it is natural to think of "==" as meaning
conceptual sameness.  And if "==" meant conceptual sameness, then lists
would be immutable, similar to Lisp lists (in the absence of rplaca and
rplacd).  In that case, .append(3) would have been implemented by consing,
and your intuition would have been right.

I'm not saying that your mind did work this way.  But it might have.


Greg


PS: Interestingly enough, all the expressions involving "eq" above are true
on my system with "eq" replaced by "is".  But (I believe) the following are
implementation-dependent:

    3 eq 3
    "foo" eq "foo"
    () eq ()

And "eq" disagrees with "is" for (properly defined) immutable class
instances.  And, although I didn't mention it above, 3 eq 3.0 should
probably hold.  (Which, I should add, means that Guido was *conceptually*
on target with the "/" redefinition.)


From mkc+dated+1022375908.02c649 at mathdogs.com  Thu Apr 25 21:21:28 2002
From: mkc+dated+1022375908.02c649 at mathdogs.com (Mike Coleman)
Date: 25 Apr 2002 20:21:28 -0500
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com> <c76ff6fc.0204241806.22f3b0d8@posting.google.com>
Message-ID: <87y9fb5k6f.fsf@mathdogs.com>

sjmachin at lexicon.net (John Machin) writes:
> 1. Explain the utility of having a dict d such that d[foo] == foo, for
> *all* foo in d. If all you want is logically a set, not a mapping, do
> d[suffix] = 1.

Yeah, I was just generating an example.  I should have used '1'--the real
problem is exactly the same either way.

The closest thing to a solution here was to look at the semi-documented
'buffer' objects.  Thanks to who suggested that.

> 2. google("suffix tree"); google("suffix array")

Well, if a good suffix tree library existed, that'd be good, too.  The only
one I see right now looks like it has some fairly severe implementation
limits.  (Well, severe if you like your objects by the million...  ;-)

Mike


From opengeometry at NOSPAM.yahoo.ca  Tue Apr 23 13:23:20 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 23 Apr 2002 17:23:20 GMT
Subject: mass mailing
References: <mailman.1019544044.22867.python-list@python.org>
Message-ID: <aa45a8$7kjio$1@ID-99293.news.dfncis.de>

Jonathan Vanasco <jvanasco at hotmail.com> wrote:
> yeah.. thats what i've figured.. i've been playing around with it lately,
> trying to move some of perl's Mail::Bulkmail features into python, based
> on the mailman code...  i think what is going to end up happening, is
> sorting a list of emails by domain name, then processing the domains into
> chunks of 100 or less, and feeding those, via smtplib to postfix as a ,
> delimited list (or so the rfc specks say will work)
> 
> ideally, that will mean 1 envelope/message per chunk -- so my bandwidth
> and qfiles could go down from anywhere between 100 and 0 %
> 
> the extra coding is worth it though -- checking through the mailman
> archives, someone from gnu.org had major problems handling 50,000 emails.
> granted -- this project may never receive more than 100 emails.  but
> should it go higher, i want to be prepared.  and not recoding something
> under a shitty support rate vs an inflated development one :)

I assume you're the OP of another thread on similiar topics.  My news
server lost your original post there.  So, sorry if I'm repeating...

    - Python would be most useful in generating and formating email
      messages (ie. header and body).  
    - Once you have the messages that you want to send, you can use Mutt or
      Sendmail to send the email out.
    - You can limit to 100 simultaneous emails by using 'xargs -P 100' or
      something like that.

1.  I would do something like this.  Use Python to generate
	file.1, file.2, ..., file.1000
    which are emails that you want to send out.  To send them out
    sequentially,
	for i in file.*; do
	    sendmail -t < $i
	done
    and, to send 100 emails simultaneously,
	ls file.* | xargs -P 100 -n 1 mysender
    where 'mysender' is another script which actually calls Sendmail.  From
    top of my head, it would go like
	cat $1 | sendmail -t -f jvanasco at hotmail.com

2.  You can also dump all emails into a single file, and then send out 100
    messages at a time,
	formail -n 100 -s mysender < file.big
    where 'mysender' look like
	sendmail -t -f jvanasco at hotmail.com

3.  You can do this using Mutt directly.

There are so many different ways of doing this.  So, can you provide more
detail to your situation?

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8 CPU cluster, NAS, (Slackware) Linux, Python, LaTeX, Vim, Mutt, Tin


From philh at comuno.freeserve.co.uk  Wed Apr 24 09:55:48 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 24 Apr 2002 14:55:48 +0100
Subject: FROM DIFFICULTIES WITH LEARNING AT SCHOOL, TO COMMUNICATION PROBLEMS IN THE FAMILY.,
References: <17c6c5e5.0204230928.1d051796@posting.google.com>
Message-ID: <slrnacdeb4.12k.philh@comuno.freeserve.co.uk>

On 23 Apr 2002 10:28:27 -0700, Anne Mc Murray <anne_mc_murray at hotmail.com> wrote:
>One of my sons is reluctant to produce any good school work, and has
>turned into a frightful headache for the family's, despite being very
>intelligent and creative, and a good reader too.
>I've tried other alternatives, with psychologists, and
>psychopedagoges, even private teachers, but nothing seems to work.

Does it *matter* that his school work isn't good? IMO, if his work 
isn't marked as part of an exam assessment that will affect his 
career or further education choices, it isn't such a big deal.

>Please, has anyone a suggestion to make?

Try posting to an appropriate newsgroup?

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From gerhard at bigfoot.de  Wed Apr 24 05:48:12 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 24 Apr 2002 09:48:12 GMT
Subject: Accessing class-level variables
References: <mailman.1019621385.21532.python-list@python.org> <Xns91F9DA8F160A1RASXnewsDFE1@130.133.1.4> <mailman.1019624445.12502.python-list@python.org> <Xns91FA105A11462RASXnewsDFE1@130.133.1.4>
Message-ID: <slrnaccvoj.eo.gerhard@developer2.opus-gmbh.net>

In article <Xns91FA105A11462RASXnewsDFE1 at 130.133.1.4>, Philip Swartzleonard wrote:
> Er, i think you're trying too hard. +) Now that i realized what you 
> really want, I just tried this...:
> 
>>>> class A:
> 	d=5
> 
>>>> class B(A):
> 	pass
> 
>>>> B.d
> 5
>>>> class B(A):
> 	def p(self):
> 		B.d += 1
> 		print B.d
> 		print A.d
> 
>>>> B().p()
> 6
> 5
>>>> B().p()
> 7
> 5

Heh. Thanks for the example code. But my needs are weirder. I really do want to
access the class variables of a *derived* class from the *superclass* 8-) These
are available for an instance method (it doesn't matter where the instance
method is defined, be it in the superclass or in the derived class), by using
self.__class__. One more reason why Python's explicit 'self' is useful.

I also can't explain the behaviour your code shows using the Language
Specification. http://www.python.org/doc/current/ref/class.html doesn't say how
inheritance affects class variables. Anybody can explain me how this works?

Gerhard


From trentm at ActiveState.com  Mon Apr 29 14:27:32 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 29 Apr 2002 11:27:32 -0700
Subject: Python sous UNIX
In-Reply-To: <aak1q3$fgl$1@wanadoo.fr>; from infocap@wanadoo.fr on Mon, Apr 29, 2002 at 08:01:11PM +0200
References: <aak1q3$fgl$1@wanadoo.fr>
Message-ID: <20020429112732.A12315@ActiveState.com>

[infocap wrote]
> Bonjour ? tous,
> 
> Je travaille pour une entreprise d'?dition mais voil? que depuis quelques
> temps je m'int?resse beaucoup ? l'informatique et particuli?rement au
> langage de programmation PYTHON  sous UNIX et j'aimerai approfondir mes
> connaissances au travers de petits programmes int?ressants, alors, si
> quelqu'un peut satisfaire ma demande, je le remercie d'avance.

Je ne sais pas si l'anglais serais une barrier pour toi, mais il y a
plusiers petits example programmes dans la Python Cookbook.

    http://aspn.activestate.com/ASPN/Python/Cookbook/

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From jcollins at boulder.net  Tue Apr 30 23:20:00 2002
From: jcollins at boulder.net (Jeffery D. Collins)
Date: Tue, 30 Apr 2002 21:20:00 -0600
Subject: Palm
References: <B43D149A9AB2D411971300B0D03D7E8BF0A2BD@natasha.auslabs.avaya.com>
Message-ID: <3CCF5EE0.4010206@boulder.net>

The new home for Pippy can be found at http://pippy.sourceforge.net.

Jeff


Delaney, Timothy wrote:

>>From: Terje Johan Abrahamsen [mailto:terjej at mailandnews.com]
>>
>>Is it possible to make Python programs that run on the Palm OS?
>>Preferably stand alone programs, without a xxx kb interpreter?
>>    
>>
>
>http://www.python.org/download/download_other.html
>
>or
>
>http://www.google.com/search?hl=en&q=python+palm
>
>In particular, look at the second entry.
>
>Tim Delaney
>
>
>  
>







From martin at v.loewis.de  Sat Apr 13 20:00:56 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 02:00:56 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <roy-0BE2A8.19181913042002@news1.panix.com>
Message-ID: <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de>

Roy Smith <roy at panix.com> writes:

> Any operator can get replaced by a function call.  It's just a matter of 
> what's more convenient to write.

Indeed, and so it is desirable to have infix shorthand notations for
the frequent operations. However, as the number of infix operators
grows, the convenience loses: you now have to start remembering line
noise.

Regards,
Martin



From fredrik at pythonware.com  Wed Apr  3 16:35:37 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 03 Apr 2002 21:35:37 GMT
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <xIGq8.25783$VQ2.8306517@twister.socal.rr.com> <mailman.1017854591.19430.python-list@python.org>
Message-ID: <JAKq8.29518$n4.5871064@newsc.telia.net>

Rich Harkins wrote:

> Because of the is comparison even if "MYBLANKVALUE" is in d then this code
> will still work.  Just don't intern whatever "MYBLANKVALUE" is (I frequently
> use undefined="UNDEFINED").

bad idea.  Python automatically interns string literals that happen
to look like identifiers (see the "Intern selected string constants"
section in Python/compile.c), so this only works if you can guarantee
that your "fn" function is never called with a string literal.

undefined=["UNDEFINED"] is a bit safer, but you really should use
has_key or try/except for things like this.

</F>




From msoulier at nortelnetworks.com_.nospam  Tue Apr  9 18:00:57 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 9 Apr 2002 22:00:57 GMT
Subject: Emacs-like editor, scriptable with python?
References: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de>
Message-ID: <slrnab6p69.1du.msoulier@pmerd071.ca.nortel.com>

On Tue, 09 Apr 2002 23:26:38 +0200, Thomas Guettler
<pan-newsreader at thomas-guettler.de> wrote:
> Up to now I was happy with emacs because I never used
> to programm it with lisp. Now I need to and find it difficult.
> 
> Is there an emacs-like editor which is scriptable with python?

    Well, it's not Emacs-like, but Vim permits a python interpreter to be
built in. 

    http://www.vim.org

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From aahz at pythoncraft.com  Sat Apr 20 20:30:31 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Apr 2002 20:30:31 -0400
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com>
Message-ID: <a9t177$pi5$1@panix1.panix.com>

In article <roy-AAA3A1.19193820042002 at news1.panix.com>,
Roy Smith  <roy at panix.com> wrote:
>
>I want to look an oid up in my dictionary, and if it's not there, I want to 
>keep working my way up the tree until I find it.  Thus, if 
>'1.3.6.1.4.3.5.66.3.4' isn't a key, I want to try '1.3.6.1.4.3.5.66.3', and 
>then '1.3.6.1.4.3.5.66', and then '1.3.6.1.4.3.5', and so on until I find 
>it (if nothing else, '1' is guaranteed to be a key).
>
>What I've come up with uses recursion:
>
>    def getBaseObject (self, objects, oid):
>        try:
>           return objects[oid]
>       except KeyError:
>           return self.getBaseObject (objects, self.getParentOid (oid)) 

I'd write this as

    while not objects.has_key(oid):
        oid = self.getParentOid(oid)
    return objects[oid]
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From speeves at unt.edu  Mon Apr 22 20:55:45 2002
From: speeves at unt.edu (speeves)
Date: 22 Apr 2002 17:55:45 -0700
Subject: changes to win server registry with python script not appearing in gui tools
Message-ID: <d6351bca.0204221655.4f554f6e@posting.google.com>

Hi!

The title says it all...  Why don't the changes that I am making to
the registry appear in the gui tools?  I have completely automated
many of the gui-functions with a simple python script, but any query
to the registry must now come from the python scripts, and not the gui
tools, as they are never current anymore :(
(Even after rebooting)

Just wondering...  (Seems like a fluke in the win os to me, but...)

speeves
central Web Support
UNT


From eppstein at ics.uci.edu  Mon Apr 15 02:56:09 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 23:56:09 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <m38z7pmq4p.fsf@mira.informatik.hu-berlin.de>
Message-ID: <eppstein-18C460.23560914042002@news.service.uci.edu>

In article <m38z7pmq4p.fsf at mira.informatik.hu-berlin.de>,
 martin at v.loewis.de (Martin v. Loewis) wrote:

> Indeed, if you write
> 
> >>> 0x40000000L<<1
> 2147483648L
> 
> it works the way you expect it to work. Please see
> 
> http://www.python.org/peps/pep-0237.html
> 
> Python has currently completed phase A of the PEP, meaning that
> operations that previously raised an OverflowError now return a long;
> nothing else has changed.
> 
> Notice that your example is the historic meaning of shift; in phase
> B.1, it will change to the arithmetic meaning (multiplication with
> 2). Since this is a semantic change, you'll get a warning if you make
> use of a short int shift that loses a bit.

Thanks.  I notice the PEP doesn't actually say anything about signs, 
though, it just talks about losing bits, so I hope this example is 
covered.

Is there some way of getting the arithmetic meaning now, without a 
warning, using a __future__ import?

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From juenglin at informatik.uni-freiburg.de  Sat Apr 13 17:53:56 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 13 Apr 2002 23:53:56 +0200
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com>
Message-ID: <xduhemf464r.fsf@leto.informatik.uni-freiburg.de>

"Terry Reedy" <tejarex at yahoo.com> writes:

> > I understand now, that 'i+=1' is no real in-place operation
> 
> Nor is it a false in-place operation :<)
> 
> > but  just a shortcut for 'i=i+1' since int is immutable.
> 
> Correct.  This is true for all immutable types.
> 
> > For the sake of clarity, an 'in-place' operation should really
> > work in-place, don't you think?
> 
> Yes, but +=, etc, are not in-place operations, and are not advertised
> as such.
> 

Michael's answer to my initial mail suggested that __iadd__ stands
for an in-place version of add:

>> Shouldn't I find the special methods '__iadd__' et al, when
>> asking for the attributes of 'int'?
>
>ints are immutable, so they don't implement __iadd__.  lists 

I'm well confused now... 
Can anybody explain why the special method 'i.__iadd__' does not 
exist, while 'i+=1' is a legal expression ('i' being a int)?

Thanks so far,
Ralf




From fperez528 at yahoo.com  Thu Apr 25 02:56:50 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 25 Apr 2002 00:56:50 -0600
Subject: urllib.urlopen("file://...") fails
References: <m2k7qwjnma.fsf@jtan.com>
Message-ID: <aa89g0$pv8$1@peabody.colorado.edu>

Seth Delackner wrote:

> With python 2.2 on windows, with the file t:\tmp\package.html existing,
> 
>>>> import urllib
>>>> u = urllib.urlopen("file://t:/tmp/package.html")
> 
> Fails with the error below, but "file:t:/tmp/package.html" works. Is
> that even a valid url?
> 
>>>> u = urllib.urlopen("file:///t:/tmp/package.html")
                                ^

One / too many? You're directing it to \t:\.... which it obviously can't find.

I don't run windows so I can't test so I may be wrong. It just looks like that 
is your problem.

Cheers,

f.


From martin at v.loewis.de  Mon Apr  1 18:24:52 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 01:24:52 +0200
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com>
Message-ID: <m3u1qvvw6z.fsf@mira.informatik.hu-berlin.de>

Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:

> My problem is the following:
> 
> Let f be a file object. If I do
> 
> f.write("Mar?o 2002")
> 
> all goes well. But if I try

It depends on your meaning of "well". It writes 9 bytes into the file,
yes.

> f.write("Mar?o 2002" + march.Name())
> 
> where march.Name() returns a unicode string I get a unicode error. I
> tried converting both unicodes to strings via str but obviously I got an
> error (in the first string the culprit is the "?" character).
> 
> Can someone help me out here and show me the way to write these strings
> to the file?

If you want to write text into a file, you should know what encoding
you want to use in this file. Then make sure that all byte strings
written to the file indeed use that encoding.

In the specific case,

f.write("Mar?o 2002" + march.Name().encode("iso-8859-1"))

will do the right thing, assuming you want the text to be encoded in
that encoding.

HTH,
Martin



From max at malva.com.uaREMOVE.IT  Mon Apr 15 09:19:12 2002
From: max at malva.com.uaREMOVE.IT (Max Ischenko)
Date: Mon, 15 Apr 2002 16:19:12 +0300
Subject: C++ code generation from xml
References: <of1e9a.4vg.ln@cvs> <a9eekb$1p90$1@gavrilo.mtu.ru>
Message-ID: <g0ke9a.ufm.ln@cvs>

 Oleg Orlov11980884 wrote:

>>
>> Any suggestions?

> Try www.reportlab.co.uk/rl-cgi/viewcvs.cgi/rlextra/preppy/preppy.py

401 Authorization Required

Would you mind to exlain a bit?
Is it a some kind of C++ preprocessor?



-- 
Meanwhile, PDP-1 made his way deep into the core of the rm-star, using his
ability to manipulate label_t to slip from context to context undetected.


From phr-n2002a at nightsong.com  Tue Apr 16 03:54:58 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 16 Apr 2002 00:54:58 -0700
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk>
Message-ID: <7x662s5b8t.fsf@ruckus.brouhaha.com>

"Duncan Smith" <buzzard at urubu.freeserve.co.uk> writes:
> So what I'm looking for is speed, and some advice so that I don't end up
> trying too many alternatives. 

If you have to do something like that over and over for zillions of
huge files, you're best off writing in C and tuning carefully.

Regarding duplicates, maybe you can just sort the file with an
external sort utility, so the duplicates will all be next to each
other.  Then you don't have to mess with dicts.  I didn't examine your
code closely enough to figure out if that makes sense, so maybe it doesn't.


From borcis at geneva-link.ch  Thu Apr 11 08:33:32 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 11 Apr 2002 14:33:32 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <slrnab9s50.6g8.mlh@vier.idi.ntnu.no> <3cb57b6b$0$78750$edfadb0f@dspool01.news.tele.dk>
Message-ID: <3CB5829C.4030305@geneva-link.ch>

Anders J. Munch wrote:

> 
>>
>>>[] and {} are objects,
>>>
>>Yes, as are None and 0.
>>
> 
> The important part is that they are _mutable_ objects.  The identity of each
> instance is by itself important information.


But we all know that at the price of lots of gc, mutable object behavior 
can be closely mimicked with an implementation boiling down to immutable 
state objects constructed on the fly.

IOW you forgot the empty sequence () and the empty string "".


> As for intuitive ways of explaining concepts: Examples, examples, examples.
> Accept no substitute.


Learning from counter-examples requires appropriate algorithms, but if 
algorithms are no substitute to a pure examples diet, what substitutes 
will you provide to examples that are in effect counter-examples, or do 
you suggest we process them without algorithms ?

Cheers, Boris
--
"Art is a trigram"






From boris at cantal.lip6.fr  Tue Apr  9 08:08:12 2002
From: boris at cantal.lip6.fr (Boris Boutillier)
Date: Tue, 09 Apr 2002 14:08:12 +0200
Subject: changing local dictionnary for functions
References: <pan.2002.04.08.18.15.31.115122.19669@cantal.lip6.fr> <vIos8.3287$vT6.670@atlpnn01.usenetserver.com> <01zs8.62383$pT1.1793993@news1.tin.it>
Message-ID: <pan.2002.04.09.14.08.08.16606.31396@cantal.lip6.fr>

On Tue, 09 Apr 2002 12:04:44 +0200, Alex Martelli wrote:

> Steve Holden wrote:
> 
>> "Boris Boutillier" <boris at cantal.lip6.fr> wrote in message
>> news:pan.2002.04.08.18.15.31.115122.19669 at cantal.lip6.fr...
>>> Is there a way to pass a local dictionnary to a function as it is
>>> possible to do with a file with the execfile statement ?
>>>
>>> In the python reference manual, in the section 4.1 : Code
>>> Blocks,execution frames and namespace, there is a table that say that
>>> function body uses globals namespace of the containing block and a new
>>> locals namespace. And there is a note that say global and local can be
>>> overrident, how is it possible ? I think it would be the answer to my
>>> first question.
>> 
>> I suspect this is an arror in the documentation, since the namespaces
>> provided to a function call are (unless I am mistaken) determined
>> solely by the context of the call. Sorry.
>> 
>> If  I'm wrong, you can guarantee that three different people will point
>> this out within three minutes of my hitting the "send" key :-)
> 
> The only way that occurs to _me_ for "changing a function's
> dictionaries" would be to extract the code object from the function and
> build a new function (with module new) using that code object and
> whatever dictionary you want for globals.  Even so, the locals (not
> really a dictionary... they only halfway-masquerade as one through the
> built-in function of that name) would be created anew at each call to
> the functions, so it still wouldn't meet the request by Boris as I
> understand it.
> 
> 
> Alex
> 

Thanks all for you answer, effectively alex this does not meet exactly my
request but I'm looking in this direction, by doing some code
manipulation I'll suppress use of local variables by replacing load_fast
by load_global opcodes etc... Don't where this leads but I'll continue to
dig !

Boris


From joost_jacob at hotmail.com  Thu Apr 25 17:28:55 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 25 Apr 2002 14:28:55 -0700
Subject: console redirection to Tkinter
References: <3CC34237.CE6F21F1@club-internet.fr>
Message-ID: <13285ea2.0204251328.5e52a77c@posting.google.com>

Here is an example using popen2.  The echo command comes with Windows and linux.


from Tkinter import *
import popen2, sys

cmd = "echo 'hello world'"
fin, fout, ferr = popen2.popen3(cmd)
root = Tk()
t = Text(root)
t.insert('1.0', fout.readline()) 
t.pack()
root.mainloop()


From Oschler at earthlink.net  Tue Apr  2 10:50:19 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 2 Apr 2002 10:50:19 -0500
Subject: IDLE for SUSE Linux?
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com> <3CA8D724.336A22BA@3captus.com> <5waq8.5401$ku5.476012@e3500-atl2.usenetserver.com> <3CA98091.4B8C83A2@3captus.com>
Message-ID: <Eqkq8.10671$ku5.785587@e3500-atl2.usenetserver.com>

Bernard,

Just replying so other adventurers can find it on Gooja.  I installed the
Python demo-scripts package and the Tools directory just popped up with Idle
in it.

thx

Bernard Yue <bernie at 3captus.com> wrote in message
news:3CA98091.4B8C83A2 at 3captus.com...
>
> Have you installed python-tk?  That maybe the reason.  IDLE needs
> Tkinter.
>






From trentm at ActiveState.com  Fri Apr 19 11:16:19 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 19 Apr 2002 08:16:19 -0700
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
In-Reply-To: <1d9301c1e79b$3b6f6db0$475afea9@thomasnotebook>; from theller@python.net on Fri, Apr 19, 2002 at 02:10:43PM +0200
References: <1d9301c1e79b$3b6f6db0$475afea9@thomasnotebook>
Message-ID: <20020419081619.B8105@ActiveState.com>

[Thomas Heller wrote]
> "Trent Mick" <TrentM at ActiveState.com> wrote in message news:mailman.1019071474.3193.python-list at python.org...
> > We are pleased to announce that versions 2.1.3 and 2.2.1 of
> > ActivePython are now available for download from
> > http://www.ActiveState.com/ActivePython
> > Both of these are bugfix releases and supercede previous ActivePython
> > releases.
> 
> Are the HTMLHelp files available separately? Can they be installed
> separately?

No, we haven't separated the HTMLHelp files (actually .chm files, aren't
HTMLHelp file that old weenie .hlp file format?). The way the installer
is setup the "documentation" is a separate installable piece but the
"core" piece is set to "must be installed" -- so you are kind of stuck.
You *could* of course fake it if you *really* don't want the
ActivePython core: install it, make copies of the .chm's, uninstall it.


Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From shalehperry at attbi.com  Fri Apr 12 03:08:08 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 12 Apr 2002 00:08:08 -0700 (PDT)
Subject: emacs python mode completion
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBMEJPKKAA.marklists@mceahern.com>
Message-ID: <XFMail.20020412000808.shalehperry@attbi.com>

On 11-Apr-2002 Mark McEahern wrote:
> Python mode in emacs has the nice syntax coloring and all, but what would be
> really great is if I could press tab or something and have it complete
> statements.  I found this thread where Guido replied and pointed out
> rlcompleter and the like for the interactive interpreter:
> 
> 
> http://groups.google.com/groups?hl=en&selm=199801111855.NAA21082%40eric.CNRI
> .Reston.Va.US
> 
> Is there a tool for emacs that does Python command completion?
> 

Doing that would require a python interpreter unless you limit it to language
keywords and perhaps __builtins__.




From fperez528 at yahoo.com  Tue Apr 23 02:13:30 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 23 Apr 2002 00:13:30 -0600
Subject: rlcompleter not calling __getattr__ on [ ]
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org> <aa2460$534$1@peabody.colorado.edu> <mailman.1019533244.22752.python-list@python.org>
Message-ID: <aa2u72$k0a$1@peabody.colorado.edu>

holger krekel wrote:


> great. but i guess we should try to come up with a solution that
> allows
> 
>   a=[[].<tab>
> 
> and similar cases to work. Also
> 
>   a=().<tab>
> 
> should be allowed, too. I'll check that tommorow. Additionally
> i want to see if it is possible to

Currently ipython seems to do these just fine, after removing the following 
from the delimiters:

readline_remove_delims -/'"[]{}()

It completes everything I can see great, and even somewhat weird cases like

a=(4.5).<tab>

vs

a=().<tab>

work fine: the first calls the float methods, the second the tuple ones:


In [1]: r=(4.5).<TAB HERE>
(4.5).__abs__           (4.5).__getattribute__  (4.5).__pos__           
(4.5).__rpow__

[snip...]

In [1]: r=().<TAB HERE>
().__add__           ().__getattribute__  ().__le__            ().__reduce__
().__class__         ().__getitem__       ().__len__           ().__repr__
().__contains__      ().__getslice__      ().__lt__            ().__rmul__

[snip...]

> 
> a) (optionally) just return method names which don't start with '__'

I'm not sure I want to add this option (which would be easy). While true that 
they are often annoying, they also say a lot about the internal structure of 
any given object. I'm afraid of the 'million preferences trap', and I'm 
trying to maintain a reasonable limit on the total number of options 
provided, making 'sound' choices for default behavior.

> 
> b) return to 'readline' strings of the form:
>    
>    COMPLETION "docstring"
> 
>    as this would show you brief documentation for each method.

As I said, ipython's ? and @doc functions already do much of this for me, but 
it may be a nice addition.

Thanks for all the help,

f.


From maik.hertha at volkswagen.de  Wed Apr  3 02:11:54 2002
From: maik.hertha at volkswagen.de (Maik Hertha)
Date: Wed, 03 Apr 2002 09:11:54 +0200
Subject: How to wait only so long for a server to respond?
References: <nwtq8.11707$ml2.924462@newsread1.prod.itd.earthlink.net>
Message-ID: <3CAAAB3A.8020504@volkswagen.de>

Ralph wrote:

>Pythoners:
>
>I have some code that requests a response from a server on the Internet, 
>but I only want to wait so long for the server to respond.  How can I tell 
>python to wait x seconds/minutes, then move on to the next section of code 
>if the server has not responded?
>
>Thanks as always.
>
>Raplh
>
you should have a look on the signal module.
example is here: http://www.python.org/doc/current/lib/node214.html


-- 

mit freundlichem Gru? /
best regards

Maik Hertha

--------------------------------------------------- h+h
EBSP Anwenderbetreuung, +49 5361 9-74950
Volkswagen AG / Brieffach 1721 / D-38436 Wolfsburg
http://ebsp.wob.vw.de              
maik.hertha at volkswagen.de
---------------------------------------------------
hartmann+hertha
it (beratung / entwicklung / support)
http://www.hartmann-hertha.de
mhertha at hartmann-hertha.de
--------------------------------------------------- h+h




From eppstein at ics.uci.edu  Tue Apr 16 02:30:42 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 23:30:42 -0700
Subject: Does Python need a '>>>' operator?
References: <mailman.1018925938.16233.python-list@python.org> <a9g8ok$9a3$1@panix1.panix.com> <a9gcmq$bmb$1@nntp9.atl.mindspring.net>
Message-ID: <eppstein-07B485.23304215042002@news.service.uci.edu>

In article <a9gcmq$bmb$1 at nntp9.atl.mindspring.net>,
 "Andrew Dalke" <dalke at dalkescientific.com> wrote:

> It's a set whose complement (that is, the set containing everything
> which isn't in the original set) is finite.
> 
> For example, in integers the set
> 
>   ...-3, -2, -1, 0, 1, 3, 4, 5, 6, 7, ...
>                      ^^^ 2 is missing
> 
> is cofinite because its complement, which is {2}, is finite.
> 
> The set of even integers is not a cofinite subset of the integers
> because its complements, the odd integers, is also infinite.
> 
> They are relevant here because bits can only be constructed either
> explicitly (you need to say them) or though operators.  Operators
> like 'and' and 'or' don't make infinite sets, but 'not' does.
> This value cannot be represented directly as an infinite set of
> bits, but can be represented as a finite set of bits and a flag
> that says "use the complement."  There is no way to construct
> a bitlist using these operators (in finite time) which is not
> either finite or cofinite, which means there is a relatively
> easy way to store these calculations on a computer.

If you did extend long integers (interpreted as finite and cofinite 
bitstrings) to bitstrings that are neither finite nor cofinite, you 
would end up with an interesting mathematical set known as the 2-adic 
integers.  Two 2-adic numbers are close to each other if their 
difference is divisible by a high power of 2.  The ones for which the 
bitstring is periodic (except for a finite number of positions) can be 
thought of as representing rational numbers with odd denominator.  But 
the 2-adics also contain certain irrational numbers such as a square 
root of 17 (not equal to the real-number sqrt).

I imagine it wouldn't be especially difficult to implement the periodic 
bitstrings.  But beyond mathematical cuteness I'm not sure what good it 
would be...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From mlh at vier.idi.ntnu.no  Sun Apr  7 14:06:21 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 7 Apr 2002 18:06:21 +0000 (UTC)
Subject: Argument against iterable ints...
References: <mailman.1018117586.26522.python-list@python.org>  <just-D808D9.08280107042002@news1.xs4all.nl>
Message-ID: <slrnab12kt.cde.mlh@vier.idi.ntnu.no>

In article <just-D808D9.08280107042002 at news1.xs4all.nl>, Just van Rossum wrote:
>[Magnus Lie Hetland]
>> Just a situation that cropped up recently: While writing a little
>> generator for flattening nested iterables, I used
>>
>>   try: iter(foo)
>>   except: return foo
>>   else:
>>       ....
>>
>> (of course I could have used a for loop directly in the try clause --
>> that doesn't change anything.)
>>
>> If integers became iterable, this sort of code would suddenly become
>> infinitely recursive. Given the propensity for using exceptions in
>> this manner rather than type checking in Python, this seems like a
>> weighty argument to me...
>
>No matter how I dislike iterable ints, this argument stinks, as the 
>situation you describe already exists with strings: each element of a 
>string is itself a string and...

Yes. And I think that stinks. The fact that we have one such problem
in the language is no reason to add another.

Oh, well.

>Just

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From clspence at one.net  Thu Apr 25 18:14:12 2002
From: clspence at one.net (Chris Spencer)
Date: 25 Apr 2002 17:14:12 -0500
Subject: Python doesn't know how to do math?
Message-ID: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com>

	I tested this using the Python command interpreter on a Windows XP box
running:
Python 2.1 (#15, Apr 16 2001, 18:25:49) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> float(5)/float(100)
0.050000000000000003
>>> 5.0/100.0
0.050000000000000003

	Okay, I might not be a math genius, but this is most definately WRONG.
What gives?  Anyone have insight into this little "feature" of Python floating
point operations?

Chris.



From logiplexsoftware at earthlink.net  Fri Apr  5 19:08:59 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 5 Apr 2002 16:08:59 -0800
Subject: single-quoted string conversion to triple-quoted string
In-Reply-To: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>
Message-ID: <20020405160859.6f4e23e6.logiplexsoftware@earthlink.net>

On Sat, 06 Apr 2002 00:05:58 GMT
robin and jim wrote:

> Is there a way to convert a single quoted string to a triple quoted string?
> 
> For example, given:
> 
> ''' abc %d xyz '''
> 
> the result of ''' abc %d xyz''' % 10 is:
> 
> ' abc 10 xyz '
> 
> which is a single quoted string.
> 
> I would like to convert the result to a triple-quoted string.

What do you mean?  What is the difference between single- and triple-quotes in
this context?  

Python 2.2 (#1, Mar 15 2002, 11:52:42) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-81)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> """hello"""
'hello'
>>> "hello"
'hello'
>>> 

As you can see, the results are the same (a string).  Perhaps if you give an
idea of what you are trying to do...

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From hzhu at mars.localdomain  Fri Apr 19 02:30:57 2002
From: hzhu at mars.localdomain (Huaiyu Zhu)
Date: Fri, 19 Apr 2002 06:30:57 GMT
Subject: Assignment operator?
References: <a9mksa$8dqs$1@as201.hinet.hr> <WiBv8.15852$eS3.5828@atlpnn01.usenetserver.com> <a9mpq4$940c$1@as201.hinet.hr>
Message-ID: <slrnabvb0e.1jq.hzhu@mars.localdomain>

Ante Bagaric <abagaric_NOSPAM at rest-art.hr> wrote:

>> I think what you haven't understood is the update in-place semantics that
>> mutable object implementations *may* choose to adopt. Otherwise you seem
>to
>> have it down pat.
>
>Hmm, and the compiler doesn't know in advance what object a name references,
>so it has no choice but to flag '+=' as rebinding operation even if it is
>not
>the case. It doesn't sound nice, but I admit there's no obvious alternative
>that
>comes to mind. This whole namespace and local/global issue has an aura of
>'unfinishness' around it :)

At the time when it was introduced, there have been quite a lot of oposition
to the fact that its behavior depends on the mutability of the object,
something that is not always known in advance.  There were even several
proposals that suggested different spelling for "always rebind", "always in
place", and "depending on mutability".  In the end one spelling was adopted
for the current behavior.

IIRC, the suggested spellings include '+=', '+!' and '+:'.

Since then there have been many questions here about the exact behavior.
It is similar to the question about / behavior between ints and floats.  But
your observation of its interaction with scoping rules appears to be new.
This suggests there is a real benefit to have an "always in place" operator.


Huaiyu


From dalke at dalkescientific.com  Wed Apr 10 04:00:02 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Wed, 10 Apr 2002 02:00:02 -0600
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org> <zHOs8.201693$Yv2.65602@rwcrnsc54>
Message-ID: <a90rv2$fvh$1@slb4.atl.mindspring.net>

Emile van Sebille:
>Oooh... that's almost as much fun as saying 'xyzzy' the first time...

I just said it, but nothing happened here.

                    Andrew
                    dalke at dalkescientific.com






From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sat Apr 13 17:23:44 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sat, 13 Apr 2002 14:23:44 -0700
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com>
Message-ID: <3cb8a237@news.mhogaming.com>

"M. J. Mix" <marilyn.mix at omitron.com> wrote in message
news:7226483f.0204101216.2c8fcda4 at posting.google.com...
| Is anyone using Python as a spacecraft ground system
scripting
| language, or onboard (embedded) language?
|
| Pros and cons when compared with TCL or JAVAScript.  Just
| some real life, honest opinions.
|
| The usual issues: reliability and maintenance effort in
applications.
| Also, use as a scripting language by non-programmers. (I
saw the post
| on indenting, so I'm concerned about that.)

+++++++++++

With CPU power and memory getting less expensive, using a
scripting language in an embedded system is becoming more
feasible every day.  (Though this may not be true 'today',
as space craft are limited to the smaller "RAD-hard"
component sets.)  Python is quite suitable for embedded
systems work.  When compared to a compiled systems language
(like 'C'), some of the main differences will be:

  1) Slower execution (4..10 times slower than 'C' code)
  2) Larger memory requirements (dynamic objects instead of
statically allocated variables)
  3) Python uses 'late binding', so some errors will not be
found until run time (that could have been found at compile
time with a strongly typed language.)
  4) Objects are dynamically created at run time, so you
could have an "out of memory" surprise

If these limitations are not bothersome to your application,
then by all means-- set some engineers to work embedding
Python into your space craft!

The way Python uses indentation for delineation of code
blocks was "kind of weird" for me (at first anyway)-- but
after coding in Python for just an hour or so, I got used to
the idea very quickly, and now I find it to be a real
pleasure.  (I hate the colon at the end of a block header
though- I forget to type that a LOT!  I think it should be
optional, unless you have a statement on the same line as
the header.)

TCL and JAVAScript are very large (when compared to Python),
and would be too difficult to build an embedded version
for-- at least Python can be cut down to fit without too
much work.  (Of course, that depends on what services your
"embedded system" will provide to the Python byte-code
interpreter.)

I would be very careful what you allow with this system--
perhaps it would control only the non-mission critical
(application level) things on the spacecraft, (like pointing
the spy camera, zoom level, etc.)

One of the big advantages of using a scripting language is
that you can get instant feedback while using the system in
the 'interactive mode' (on the console.)  I would have to
ask you if this is what you intend to do with your
spacecraft-- and I assume you have solved the communication
delay problems...  If you do not intend to operate the
spacecraft in 'interactive mode', then why use a scripting
language at all?





From mertz at gnosis.cx  Mon Apr 22 11:54:01 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 22 Apr 2002 11:54:01 -0400
Subject: Science And Math Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
 <mailman.1019299357.11513.python-list@python.org>
 <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com>
 <Xns91F67C40EB0E0lumberjack@216.148.53.83>
 <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
 <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1>
 <3CC3A89C.BE2BF65@tundraware.com>
Message-ID: <ZIDx8kKkX0eO092yn@gnosis.cx>         

|> mathematics, including the study of algorithms, is a natural science.

Tim Daneliuk <tundra at tundraware.com> wrote previously:
|I disagree strongly...  All of mathematics is a formal construct of the
|human mind created with the intent of removing ambiguity and enhancing
|our ability to describe what we *think* and what we *observe*.

To my mind, Daneliuk present here the strongest argument that
mathematics (and therefore CS) *IS* indeed a science.

The wrinkle I would add to it is in describing exactly *which* science
mathematics is.  Rather than being something close to Physics, and
describing the behavior of the (large-scale) physical universe, I see
Mathematics as a special branch of Psychology.  Mathematics describes
the laws of thought, but at a level of abstraction that applies to any
thinking being.  If we find martians, or teach dolphins, or invent
computers sophisticated enough to "think", their thought patterns will
be constrained by the rules of Mathematics--even though they quite
likely will not be characterized by uniquely human Psychological
features like, say Oedipal complexes, structured 3-scale memory,
prototype-centered reasoning, and so on.

Alex may disagree with me here, but I think my opinion here is quite
close to the middle Wittgenstein, or the _Philosohy of Mathematics_ and
the _Blue and Brown Books_.  It is also very close to some of Godel's
remarks, in his philosophical moods.

Yours, Lulu...

--
 mertz@  _/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: \_\_\_\_    n o
gnosis  _/_/             Postmodern Enterprises            \_\_
.cx    _/_/                                                 \_\_  d o
      _/_/_/ IN A WORLD W/O WALLS, THERE WOULD BE NO GATES \_\_\_ z e





From neuhauser at mail.cz  Tue Apr 30 10:36:38 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 16:36:38 +0200
Subject: Python Webpages
In-Reply-To: <3CCEAE2B.8000300@gmx.net>
References: <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net> <23891c90.0204300033.16549c52@posting.google.com> <3CCEAE2B.8000300@gmx.net>
Message-ID: <20020430143638.GO325@roman.mobil.cz>

> From: Tom Chance <tomchance at gmx.net>
> Subject: Re: Python Webpages
> To: python-list at python.org
> Date: Tue, 30 Apr 2002 14:46:03 +0000
> 
> Sorry if I'm coming in at a tangent to this conversation, or repeating 
> something that's been said before, but I've been using Perl and the 
> HTML::Template module for ages now and it seems to me to be the natural 
> way to do things. You basically create a series of HTML templates, and 
> put fake HTML tags into them like:
> 
> <TMPL_VAR NAME=TITLE>
> 
> And then in your perl scripts you call a new HTML::Template object, give 
>  it the values to put into all of the fake HTML tags, and tell 
> HTML::Template to print out the page, like so:
> 
> use HTML::Template;
> my $template = HTML::Template->new(filename => 'template.html');
> $template->param(TITLE => 'Hello there!');
> print $template->output();
> 
> This lets you keep the HTML seperate from the logic of the script 
> (something which I find preferable to PHP's system, especially if the 
> script is vey complicated), 

    Pardon me, but what's the PHP's system? Perhaps Smarty:

    require("Smarty.class.php");
    $smarty = new Smarty;
    $smarty->assign("Title","My Homepage");
    $smarty->assign("Names",array("John","Gary","Gregg","James"));
    $smarty->display("index.tpl");

    I always thought Perl's system was

    print "Content-Type: text/html\n\n";
    print "<html><body>Hello world!</body></body>\n";

-- 
FreeBSD 4.4-STABLE
4:20PM up 1 day, 6:46, 11 users, load averages: 0.00, 0.03, 0.01




From martin at v.loewis.de  Mon Apr 15 10:10:38 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 16:10:38 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <m34ridmq00.fsf@mira.informatik.hu-berlin.de> <3cba7c41@news.mhogaming.com>
Message-ID: <m3g01xoxwh.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> | That's what I'm trying to tell you all the time: the
> | '>>>' operator is meaningless if it is defined as
> | "fill in zeroes". How does it know where to start
> | inserting zeroes?
> 
> No, it isn't meaningless:
> 
> The 'long' class has an internal representation for the
> long number.  The number of bytes that are currently
> being used to contain the number are known (internally)
> to the object.  The zeroes get shifted into the high bit
> of the number, no matter how many bytes are used to
> contain the number.

long integers are represented internally not in two's complement, but
with a sign and an absolute value. So right-shifting *already* shifts
in zeroes. That does not change the fact that -0xFFL >> 1 is -0x7FL -
even if the internal representation only uses a single byte.

> An int type simply shifts a zero into bit 31, so that
> this works the same as it will on a long.

No, that won't be the same: an int is internally represented with the
two's complement. Shifting in zeroes into the internal representation
would have different effect on ints and longs.

> The '>>' operator should also work the same for a long
> as it does for an int.

I agree, and I believe it already does.

> The '<<' operator should also work the same for a long
> as it does for an int.

I agree; that will need to change.

> The 'hex()' method should work the same for a long as
> it does for and int.

I agree; that will need to change.

> Why is this so confusing to you?  

I'm not confused. Why do you think I am?

Regards,
Martin


From claird at starbase.neosoft.com  Mon Apr 15 14:08:47 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 15 Apr 2002 13:08:47 -0500
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3cb8a237@news.mhogaming.com>
Message-ID: <C04BD5ACF621A707.9626610E9084F7B7.9644F5C8B424796F@lp.airnews.net>

In article <3cb8a237 at news.mhogaming.com>,
Ken Peek <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
			.
			.
			.
>component sets.)  Python is quite suitable for embedded
>systems work.  When compared to a compiled systems language
>(like 'C'), some of the main differences will be:
>
>  1) Slower execution (4..10 times slower than 'C' code)
>  2) Larger memory requirements (dynamic objects instead of
>statically allocated variables)
>  3) Python uses 'late binding', so some errors will not be
>found until run time (that could have been found at compile
>time with a strongly typed language.)
>  4) Objects are dynamically created at run time, so you
>could have an "out of memory" surprise
>
>If these limitations are not bothersome to your application,
>then by all means-- set some engineers to work embedding
>Python into your space craft!
			.
			.
			.
While this presentation is accurate, it's incomplete
in a way which calls out for comment.

I summarize my protest:  2), 3), and 4) are compari-
sons whose signs are indeterminate.  Yes, for example,
Python objects are dynamically created, and there is
the potential for "out of memory".  I have a personal
belief, however, that the memory analysis of general
C programs is no easier to compute than that of Python
programs.

More positively:  my personal belief is that it's
feasible to demonstrate that coding in a particular
style of Python makes its memory behavior at least
as deterministic as that of the corresponding C
programs.

Yes, one can construct examples of cosmetically
similar faulty C and Python source, where a C
compiler squawks, but a Python developer might
reasonably be expected to fail to notice the error
until a run-time exercise.  I claim this is an
egregiously artificial situation, though.  Alex
Martelli frequently writes on this subject.  I'll
summarize my own thoughts:
* wise Python developers use static analyzers;
* wise Python developers have good testing
  habits;
* Python source is only a fraction as long as
  C, so *of course* there are more errors for
  a C compiler to catch.

As to 2), there are abundant counterexamples of C
coders using such inefficient algorithms that they
require MORE memory than their Python-coding peers.
-- 

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


From mgilfix at eecs.tufts.edu  Thu Apr 18 11:26:47 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Thu, 18 Apr 2002 11:26:47 -0400
Subject: Pure python crypto
In-Reply-To: <7xofghe34j.fsf@ruckus.brouhaha.com>; from phr-n2002a@nightsong.com on Thu, Apr 18, 2002 at 03:02:20AM -0700
References: <mailman.1019024313.11744.python-list@python.org> <7x3cxubvgb.fsf@ruckus.brouhaha.com> <mailman.1019053415.9589.python-list@python.org> <7x662qqb3n.fsf@ruckus.brouhaha.com> <mailman.1019100335.10929.python-list@python.org> <7xofghe34j.fsf@ruckus.brouhaha.com>
Message-ID: <20020418112647.A2347@eecs.tufts.edu>

On Thu, Apr 18 @ 03:02, Paul Rubin wrote:
> Might as well.  My address is above.  What's the encryption speed?

  Haven't done any throughput testing on it yet. I just finished it
over the weekend but then exams took over :)

> > > FWIW, I coded RC4 in Python a while back and if I remember correctly,
> > > got about 100k/sec throughput on a P3-750.  That's sort of marginal
> > > for (say) sftp over ADSL.  Using the SHA-OFB hack that I described,
> > > I got about 400-500k/sec.  
> > 
> >  Do you still have a copy of this? I would be nice to assemble a suite
> > that can be used to create a pure python implementation of SSH (which
> > could also use wrappers to C implementations that exist as well).  I
> > wish I had a better idea of what was out there module-wise though.
> 
> I'll take a look, I probably still have it.  I only spent a few
> minutes coding it anyway, so can do it again if needed.  If coding RC4
> is an obstacle to you though, I don't think you're going to finish
> implementing SSH ;-).  Anyway, after coding RC4, I saw how much slower
> it was than SHA-OFB, so I stopped caring about it and may have trashed
> the implementation.

  Haven't looked into implemented RC4 yet :) I was looking at DES but
there are so many different ways to approach implementation...

> With luck, future versions of Python should include a DES/3DES module,
> so an SSH package can use that.

  Agreed. They do exist with C wrappers. I'll have to wait until after
exams though, so probably sometime in May before I really consider it.

              -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From starx at pacbell.net  Mon Apr  8 13:41:11 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 8 Apr 2002 17:41:11 GMT
Subject: executing an external program
References: <3CB14C40.402@beth.uniforum.org.za> <j4r8lqxy3d.fsf@informatik.hu-berlin.de>
Message-ID: <Xns91EA6CE4BFF3BRASXnewsDFE1@130.133.1.4>

Martin v. L?wis || Mon 08 Apr 2002 03:51:18a:

> mixo <mixo at beth.uniforum.org.za> writes:
> 
>> How can I continue execution of my script, and run only one instance
>> of the external program and keep track of what is doing so as to
>> avoid confusion?
> 
> On Unix, you can use os.fork/os.exec to create the subprocess
> asynchronously. You can then use os.wait to wait for the completion of
> the child process when you want to start the next process. For that,
> you have three modi operandi:
> 1. you invoke wait blockingly.
> 2. you invoke wait non-blockingly, and operate in a polling fashiong.
> 3. you install a SIGCLD handler, and start the next process when the
>    previous one terminates.

I don't really know anything about fork (or even if it works on other 
platforms), but you could also use threading. Create a thread that does 
nothing but call system() repetadly, and put the results on some kind of 
locked threadsafe variable/object (one of those queue.queue 
thigamajiggers?). Look at docs for the thread module if you think this 
path will help you.


-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From fperez528 at yahoo.com  Sun Apr 14 15:00:50 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sun, 14 Apr 2002 13:00:50 -0600
Subject: C API Q II : __main__.__dict__.keys() not same as dir()
References: <a9bl5l$2aq$2@green.tninet.se>
Message-ID: <a9cjmp$74u$1@peabody.colorado.edu>

Johan Hahn wrote:

> Anyway, for now I use the InteractiveInterpreter and do the input buffering
> in C and it works great. So great I decided to implement completion of
> Python commands, just like when you hit tab in Unix.
> 
> For example if you type...
>>>> import code
>>>> code.I
> ...and press <tab> you should be presented the list:
> ['InteractiveInterpreter', 'InteractiveConsole']
> 
> That went fairly easy too (thanks to Python), but one problem remains.
> 
> When you import a module using python shell, it appears in
> __main__.__dict__. In my program however, that is not the case. The module
> is added to sys.modules and can be used without any problem but I need to
> see it for my completion code to work.
> Here is an example of how it looks in the console:
> 

Note that you'll need to be able to specify in what namespace the readline 
completer does its evaluations. As supplied in py2.2, the standard 
rlcompleter module has __main__ hardwired into it. I submitted a patch to fix 
this which has been accepted for 2.3.

If you want it, you can download IPython from 
http://www-hep.colorado.edu/~fperez/ipython/ and grab the FlexCompleter.py 
file from it (it's the patched rlcompleter). Then subclass from it as needed 
(see the MagicCompleter class in iplib.py for an example). By the way, 
IPython is also a console based on InteractiveConsole but with a ton of 
improvements, some of those may be of use to your project. The code is LGPL, 
so feel free to use it.

Cheers,

f.


From pyth at devel.trillke.net  Mon Apr 22 13:13:13 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 22 Apr 2002 19:13:13 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <xVWw8.76795$T%5.8026@atlpnn01.usenetserver.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com> <mailman.1019415039.30680.python-list@python.org> <304d20df.0204220633.50bf23bf@posting.google.com> <mailman.1019491242.2145.python-list@python.org> <xVWw8.76795$T%5.8026@atlpnn01.usenetserver.com>
Message-ID: <20020422171313.GH27395@devel.trillke>

On Mon, Apr 22, 2002 at 12:28:39PM -0400, Steve Holden wrote:
> "holger krekel" <pyth at devel.trillke.net> wrote ...
> [ ... ]
> > If so, let me note that "execing a string" is what python does itself.
> > e.g. when you import a module the interpreter is execing a module-string
> or
> > code-object. It's some weird hack but the regular python way.
> >
> Well, there's the minor point that the efficiency of the interpreter's
> regular operation is hugely affected by the pre-compiled ".pyc" file that
> will frequently be created. When you import a module, therefore, usually you
> will be executing a code object.

agreed.

> Writing a module to "exec" one or other of multiple strings gets you pretty
> much the worst of all worlds -- when the (compiled) module is loaded, it
> will be executed as a code object, but then one selected version-specific
> string must still be compiled by an "exec" before being executed each time
> the module is run.

true. Upon first load of the module there is some overhead which 
the use of e.g. generators hopefully make up for :-)

> If some misguided user tries to take advantage of this "portability"
> technique by using the same directory for multiple versions of Python the
> situation gets a whole lot worse, as the ".pyc" files will be re-created
> each time the Python interpreter is different from the one that imported the
> module the last time!

Good point. Installing to the version specific "site-packages" probably helps 
and is not uncommon. After all i presume that most (not all!) modules are not
intended to be used from different python-versions at once. Wrong?

regards,
	holger




From nospam at bigfoot.com  Fri Apr 19 11:05:34 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 19 Apr 2002 17:05:34 +0200
Subject: How to save bitmap from MS Windows clipboard to xxx.png?
References: <a9p1e3$efb$1@news.nextra.cz>
Message-ID: <a9pb8b$2nlg$1@norfair.nerim.net>

Hi,

Have a look to the Python Imaging Library (AKA PIL). This should be a good
start for converting graphic formats

http://www.pythonware.com/products/pil/index.htm

And have a look to the Python Win32 extensions that may (?) provide
scripting access to the clipboard.

http://starship.python.net/crew/mhammond/

HTH

--Gilles

"Petr Prikryl" <Answer.via.news.please.prikryl at skil.nospam.cz> a ?crit dans
le message news: a9p1e3$efb$1 at news.nextra.cz...
> Hi,
>
> I am quite new to Python to now what was done already
> and what was not.
>
> What I need is to save the window snapshot (in MS
> Windows) into PNG file.  It just came to my mind that
> some Python utility may already exist for that task
> or that some PNG support is implemented.
>
[...]




From nas at python.ca  Sun Apr 14 18:35:19 2002
From: nas at python.ca (Neil Schemenauer)
Date: Sun, 14 Apr 2002 15:35:19 -0700
Subject: Neural Networks
In-Reply-To: <3CB9E3D9.121EA05C@ntlworld.com>; from a.clarke11@ntlworld.com on Sun, Apr 14, 2002 at 09:17:29PM +0100
References: <3CB9E3D9.121EA05C@ntlworld.com>
Message-ID: <20020414153519.A21599@glacier.arctrix.com>

tony clarke wrote:
> Does anyone know any sites, or have any code or other info about
> artificial intelligence. I am specifically interested in simple neural
> network programming, being a simple soul myself!

People seem to like this little module

  http://arctrix.com/nas/python/bpnn.py

for some reason.  There are more sophisticated implementations around
too.  Ask google.

  Neil




From ponderor at lycos.com  Tue Apr 30 11:10:50 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 30 Apr 2002 10:10:50 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr 30)
Message-ID: <CA21AB3426569724.5B895F6A06735602.F49F49A906FC38EA@lp.airnews.net>

QOTW:  "you really can't change anything without breaking something for
someone." -Tim Peters

"Later I will show a way to declare the types strongly and statically
enough to satisfy the most ardent Java or C# masochist." -Paul Prescod
http://www.xml.com/pub/a/2002/04/24/google.html

"BTW, this is one thing I love about the Python community.  Ask a
stupid question and you get good and helpful answers." -Chris Spencer


    NotImplemented vs. NotImplementedError 
        http://mail.python.org/pipermail/python-list/2002-April/100707.html
    
    Makin' like a beaver
        http://mail.python.org/pipermail/python-list/2002-April/101115.html
    
    Floating point frustrations revisited for x.00000000000000000n'th time:
        http://mail.python.org/pipermail/python-list/2002-April/100619.html
    
    SHA-based encryption function in Python discussed    
        http://mail.python.org/pipermail/python-list/2002-April/100314.html
    
    List begun to provide logic programming and constraint-propagation features in Python: 
        http://www.logilab.org/python-logic/
    
    Context sensitive help for prefered string notation
        http://mail.python.org/pipermail/python-list/2002-April/100151.html
    
    
========================================================================

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

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

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

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

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

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

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


-- The Python-URL! Team--

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

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


From billludden at email.msn.com  Mon Apr  8 11:08:25 2002
From: billludden at email.msn.com (billludden)
Date: Mon, 8 Apr 2002 09:08:25 -0600
Subject: Newbie:  How to sound an alarm
References: <ucFMmso3BHA.1332@cpimsnntpa03> <LU8s8.5403$d7.1425212@newssrv26.news.prodigy.com>
Message-ID: <#N$lg8w3BHA.1652@cpimsnntpa03>

Thanks you both for your responses.  Both very helpful.  Bill.




From noah at noah.org  Tue Apr 23 02:18:53 2002
From: noah at noah.org (Noah Spurrier)
Date: Mon, 22 Apr 2002 23:18:53 -0700
Subject: (Win32) Mapping WebDAV to Drive Letter
References: <3CC00F34.8000807@thomas-guettler.de> <jck8cukmp84qeae0qq3j0vnvgdou9jdrj7@4ax.com>
Message-ID: <0eb701c1ea8e$be050460$0200a8c0@saysyou>

Do you need a drive LETTER? Maybe a name in
"My Network Places" will work for you. If you called
it "foo" then you could refer to it with its UNC name
        \\foo\
It's pretty easy to do from the Windows 2000 GUI.
You just add it through "My Network Places" and use
the Wizard. Maybe there is some way to automate this.
        http://www.zope.org/Members/Noah/webdav

How to setup a Windows 2000 WebDAV folder
Step 1 - Open "My Network Places"
    Double-Click on the "My Network Places" icon on your desktop.
Step 2 - Start Web Folder Wizard
    Double-Click on the "Add Network Place" icon. This will start the Web
Folder Wizard.
Step 3 - Location
    Type http://www.MyZopeServer.com:8080/somefolder into the box and click
"Next".
Step 4 - Authentication
    Type in your User ID and Password into the appropriate boxes and then
click on "Ok". (This is the name from an acl_users file.)
Step 5 - Name your Web Folder
    Type a friendly name for your WebDAV folder into the box and then click
on the "Finish" button.
Step 6 - Finished
    Your Web Folder will now appear under "My Network Places".
    Double clicking the web folder will connect you to your Zope folder.

Yours.
Noah

----- Original Message -----
From: "Tim Roberts" <timr at probo.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, April 22, 2002 11:10 AM
Subject: Re: (Win32) Mapping WebDAV to Drive Letter


> Thomas Guettler <zopestoller at thomas-guettler.de> wrote:
> >
> >Does someone know how you could use the win32 api
> >with python to map a webdav server to a drive letter?
> >
> >There is a C library for the webdav part. But I have
> >no clue how you could map it to a drive letter.
>
> This requires an "installable file system driver", which is a kernel-mode
> component.  Unless you are a driver writer, believe me when I say you do
> NOT want to go there.
> --
> - Tim Roberts, timr at probo.com
>   Providenza & Boekelheide, Inc.
>





From sfconsulting at attbi.com  Mon Apr  8 11:51:48 2002
From: sfconsulting at attbi.com (Daniel Popowich)
Date: 8 Apr 2002 08:51:48 -0700
Subject: 2.2 descriptor API
References: <410890c4.0204040806.363f5728@posting.google.com> <3CADF520.C0264C0B@python.org>
Message-ID: <410890c4.0204080751.715ceed@posting.google.com>

Guido van Rossum <guido at python.org> wrote in message news:<3CADF520.C0264C0B at python.org>...
> Daniel Popowich wrote:
> > 1.  Why is the __get__ method called on a descriptor when it is
> >     referenced as a class attribute?
> 
> This was a requirement.  When you get a method out of a class in Python 2.1
> and before, it gets special treatment which gives you an "unbound method".
> The unbound method insists that you call it with an instance of the right
> class, in order to prevent mistakes.  The new descriptor API must be able
> to support similar behavior.  But it must also support other behavior, like
> class methods and static methods.
> 
> > 2.  What is the purpose behind the second argument; does someone have
> >     an example of how or when I would make use of it?
> 
> When __get__ is called on a class, the first argument is None, and the
> second argument is the class.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)



My questions had more to do with design than technical facts.  It
strikes me that the __get__ method is not so much the publication of a
new API, but rather the exposure of low-level, run-time interpreter
mechanics.  Essentially, you've exposed the necessary arguments so one
can call PyMethod_New (via the new module's instancemethod) to bind a
function to a class:

    class myclassmethod(object):
	def __init__(self, func):
	    self.func = func

	def __get__(self, obj, type=None):
	    return new.instancemethod(self.func, type, type.__class__)

But I question the utility of this for Joe Schmoe, python programmer.
If instead of using the descriptor api, the language had constructs
like the following to define methods:

    class Foo:
       # static method definition
       def static foo(*args): pass

       # class method definition
       def class bar(cls, *args): pass

       # instance method definition
       def baz(self, *args): pass

in other words, after "def", allow a qualifier (or perhaps a list of
qualifiers) before the function name, then the interpreter can mark
the functions appropriately and do the right magic when the
class/instance attribute is accessed (*).

I understand that a method like the current __get__ is necessary at
SOME level of the interpreter, I just question the utility of exposing
it.  What I would prefer seeing for __get__ is symmetry with __set__
and __delete__, that is, its signature being: __get__(self, obj) and
it is only called on references to bound instances.  

As for referencing class attributes: I think the interpreter should
always return the value of the attribute, doing magic on methods to
return the proper binding based on its definition.  For the
power-programmer (that rare individual who needs to program their own
version of staticmethod or classmethod) there's always the new module.

Thanks for listening,

Daniel Popowich


(*) this would also remove another blemish from the language: having
to define a function and LATER redefine to make it a static or class
method.  It's far from intuitive, not very readable (especially in the
case of lengthy methods) and, I imagine, difficult for introspection
tools (eg, doc generators) to recognize them.  In short, using the
descriptor API to create static/class methods is un-pythonic, a wart.


From philh at comuno.freeserve.co.uk  Sun Apr 14 14:01:22 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 19:01:22 +0100
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <m3662uy7h8.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnabjgvi.741.philh@comuno.freeserve.co.uk>

On 14 Apr 2002 11:06:43 +0200, Martin v. Loewis <martin at v.loewis.de> wrote:
>bokr at oz.net (Bengt Richter) writes:
>
>> >bool(a) ^ bool(b)
>> >
>> Hm, ^ should probably be != to return bools instead of integers?
>
>It turns out that ^ returns bool if both arguments are bool.

And what if one is bool and the other is an int?

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From huaiyu at gauss.almadan.ibm.com  Mon Apr 22 20:30:26 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 23 Apr 2002 00:30:26 +0000 (UTC)
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com>
Message-ID: <slrnac9ap2.9fe.huaiyu@gauss.almadan.ibm.com>

John Roth <johnroth at ameritech.net> wrote:
>
> The study of algorithms may very well
>be a branch of mathematics, but mathematics is not science.
>
>Science is, among other parts, the study of the natural universe
>and things in it. The study of algorithms is a study of things
>that have been created by the human mind.
	  
I hope the following terminology might be better:

Science is the study of objective descriptions of interesting phenomena.
Natural science is concerned with phenomena independent of human mind.
Mathematics is concerned with formal structures.  Science generally consists
of mathematics and natural sciences (ignoring social sciences here).

Before the invention of computers, the best device for generating
interesting formal structures is a live human with books, pens and lots of
scratch paper.  But mathematics is not about the peculiarities of specific
human thinking.  The word "formal" meant that two reasonably knowledgeable
mathematician would essentially agree on a conclusion if sufficient details
are spelt out.  Nowadays computers themselves generate sufficiently complex
and interesting formal structures that could not be generated by human brain
alone.  Study of algorithms is part of mathematics.

Objectivity is defined by observability, verifiability, refutability, etc.
That applies to mathematics as well as other branches of science.
Mathematicians work by making observations, conducting experiments,
proposing theories, exploring consequences and verifying results, just like
natural scientists do, except that their objects are formal structures.

The structures mathematicians study are discovered.  The symbolic languages
used to describe them are invented, but such inventions are guided by the
structures.

(The natural laws physicists study are discovered.  The instruments used to
observe and experiment are invented, but such inventions are guided by the
physical laws.)

The "unreasonably good applicability" of mathematics in natural sciences is
partly due to the fact that actual structures in the physical world often
give hints of interesting formal structures, partly due to the fact that
problems in natural sciences often influence mathematicians' sense of
"interesting".  However, this does not mean that mathematics is itself part
of natural sciences.

None of the above was intended to _define_ any term.  The intention is to
constrast various terms.

Huaiyu


From aahz at pythoncraft.com  Mon Apr  8 19:09:57 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Apr 2002 19:09:57 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <ub315bi2fkdube@news.supernews.com> <3CB1CA65 <ub3tb6o43pkgdc@news.supernews.com>
Message-ID: <a8t805$c6e$1@panix1.panix.com>

In article <ub3tb6o43pkgdc at news.supernews.com>,
John Roth <johnroth at ameritech.net> wrote:
>
>As a final note, how do I submit an acceptable PEP?
>That is, what are the restrictions on formatting it? If I
>have to format it myself like the rest of the PEPs and
>documentation, it's going to be a while.

http://www.python.org/peps/pep-0001.html
http://www.python.org/peps/pep-0009.html
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From pecora at anvil.nrl.navy.mil  Thu Apr  4 10:43:17 2002
From: pecora at anvil.nrl.navy.mil (Louis M. Pecora)
Date: Thu, 04 Apr 2002 10:43:17 -0500
Subject: CLAPACK Build on Mac w/ NumPy?  Help.
References: <030420021516186189%pecora@anvil.nrl.navy.mil> <mailman.1017875950.28270.python-list@python.org>
Message-ID: <040420021043171630%pecora@anvil.nrl.navy.mil>

In article <mailman.1017875950.28270.python-list at python.org>, eric
<eric at enthought.com> wrote:

> Hey Lou,
> 
> You might look at SciPy.  The linalg2 module (in CVS) uses f2py to wrap both
> Fortran Lapack and C lapack.  If linalg2 (its looking pretty good, but still
> under development) isn't suitable for your purposes, then at least they might
> provide some hints.
> 
> eric

Will check it out. Thanks.

-- 
-- Lou Pecora

  - My views are my own.


From gerhard at bigfoot.de  Mon Apr  1 00:42:13 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Apr 2002 05:42:13 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com>
Message-ID: <slrnaafpaa.8j8.gerhard@lilith.my-fqdn.de>

Gil wrote in comp.lang.python:
> I think my original install of Python1.5 with RedHat 7.1 is not
> playing nice with a later install of Python2.2. How do I troubleshoot
> this?

What does "python -V" say? 1.5.2 or 2.2?

What does "which python1.5" say? /usr/bin ?
What does "which python2.2" say? also /usr/bin?

If both python1.5 and python2.2 are in /usr/bin and "python -V" says
2.2, you could just "rm /usr/bin/python" and
"ln -s /usr/bin/python1.5 /usr/bin/python" to make 1.5.2 your default
Python. If you want to use Python 2.2 for your scripts later on, just
put:

#!/usr/bin/env python2.2

at the top instead.

Note that there are two flavours of the Python 2.2 RPMs - you'll want
the python2-... one in the future. It installs a "python2" instead of
a "python" symlink in /usr/bin in order to not confuse Redhat's really
clever software.

If the Redhat people cared for Python programmers, they'd simply not
use /usr/bin/python in their scripts but hard-code the required
version.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From jb at cascade-sys.com  Sun Apr 21 05:32:29 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 21 Apr 2002 02:32:29 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net>
Message-ID: <3CC2872D.EF791985@cascade-sys.com>

Andrew Dalke wrote:

> James J. Besemer:
> >Seems to me the more prominent genetic similarities are the overall
> > semantics: late binding, garbage collection, lists and sequences as
> > a way of life.
>
> Those are also present in Perl, Tcl, Smalltalk, APL, Mathematica,
> and many other languages.

All much (decades) newer than Lisp.

> Yes, Lisp was there first, but it's only an
> indirect influence on Python.

I hereby once and for ever more retract my casual suggestion that Lisp had or
may have had any direct influence on Python.  I was clearly in error, in part
having specifically mis-remembered some statements I read long ago by Guido,
which in retrospect I find I had completely backwards.

That being said, I guess I just don't get the strong reaction to this secondary
point.  I'm new here but I get the impression that "lisp" is a naughty word,
for some unfathomable reason (unless it's simply that it's syntax sucks).  For
me to say Python shares many concepts from Lisp is intended as praise.  You
react as if it's some kind of an insult.

Actually, it makes me think a little less of the language designers, if so many
good things from Lisp evidently were picked up by accident rather than as a
conscious design choice (if that's truly the case).

> >  Also, the fact that functions are values that may be free rebound
> > to other names.
>
> And Perl and Tcl and ...

And APL and others, but again Lisp was one of the principle origins of this
concept, decades before all the rest.

> >  MAP, APPLY, and especially LAMBDA also are pretty unique to Lisp.
> >  This is VERY like Lisp and very UNLIKE "Algol like" languages.
>
> I first learned about map through Perl.  That's also where I learned
> about anonymous functions, although Perl doesn't use the word 'lambda.'

IIRC, Larry Wall expressly acknowledges Lisp's contribution for these features
and others.

> They were added to Python just before the 1.0 release, as a contribution
> by someone other than Guido.  He's noted for saying way back in 1994
> ] To be honest, I wish I hadn't introduced lambda, map, filter and
> ] reduce -- they support a style that is inconsistent with the rest of
> ] Python. ...

That's disappointing to hear.  I thought including the ability to do some
functional-style programming was a great thing.  Somebody else here or in one
of the books I read recently I thought said support for functional programming
was a good thing.

> One long-time complaint about them is that they weren't lisp-like
> enough, for example, in how lambda handles scope.  Python used to
> be quite different than Lisp in how it did static scope.  But as I
> recall, others have pointed out that some lisp handle scoping
> differently than others, and closer to how Python does it.

Originally Lisp ONLY had dynamic scope, like Perl, which is weird.   Scheme
adopted lexical scope which is how Python 2.2 works by default.  In between,
yes, different Lisp implementations experimented with different approaches,
because dynamic scope is so weird and intrinsically slower than static.

> BTW, the first two features are best done in modern Python using
> list comprehensions and the f(*a, **kw) syntax.

"best" is a value judgment.  But it's true MAP could be defined as a user
function and apply appears to serve no actual value I can discern.

Lambda however seems useful, judging by the number of times it seems to show up
in code examples.

> Oh, I believe there's a lot of indirect influence on Lisp on Python,
> just as there is on most every other language.  Even the first language
> I wrote (a BASIC written in BASIC) likely had some influence from Lisp,
> and that was before I knew the language existed.

> I just argue that there are few direct influences.

Noted.

> I think you misunderstood my statement, but then again, I wasn't
> very clear.  Lisp's idea of lists is as a composition of pairs.  As
> such, certain operations like list indexing are (usually) slow, while
> list mutation is fast.  On the other hand, Python uses a contiguous
> block of memory, so indexing is fast but mutation (other than append)
> is usually slow.

I think you misunderstood my comment but it was about Lisp anyway so never
mind.

In any case, Python clearly made the CORRECT choice, as generally indexing
(access) is the more frequently used operation once a data structure is built
up.

And certain types of mutation, particularly append, do not necessarily have to
be slow.  I assumed the language inventors would provide some extra space at
the end of each list so that append could be almost instant (except for every
Nth one).  Same as hash arrays can be fairly efficient if you don't make them
too small.  But I thus far never looked into Python's implementation and now
you make me wonder.

> So algorithms which might run fast under Lisp do not under Python,
> and vice versa.

I can think of few examples where Python would necessarily be slower than Lisp.

Where Python suffers a disadvantage I suspect the difference has more to do
with implementation details, such as going with reference counting rather than
a higher performance garbage collection scheme.

> The decision for this difference was made after
> observations in ABC that most list use was through indexing, and not
> through a direct decision to be different from Lisp.

After this discussion I'm guessing a lot of the lisp heritage crept in
unnoticed via ABC, which has a number of the shared traits.  So I was right to
see the similarities but wrong to attribute it to any conscious choice.

> >> ] Python comes from the Algol family of languages.
> >
> >In syntax yes.  In semantics it seems quite the opposite.
>
> I've heard the statement from Lisp people that all other languages
> are an attempt to reproduce Lisp, badly.

I don't consider myself to be a "Lisp person."  I suppose it dates me that I've
simply been around long enough to have learned the language well enough to
appreciate that it deserves a lot of respect.  And when you look at the
totality of features, Lisp and Python have a tremendous amount in common.

That being said, I think Python is better than Lisp or Scheme in almost every
way.  The key innovation to my eye is the combination of a rich semantics of a
late binding language like Lisp (and a host of other more recent languages)
with an ALGOL like syntax.  Syntax was one of Lisp's worst characteristics and
the 40+ year old Algol if/then/else/while/do surface syntax has emerged as a
clear winner.

Here, I only object to saying Python comes from Algol because, other than the
surface syntax (and the connection is fairly loose there) the languages are
radically different.  To me, the "Algol family" all means early/static binding,
a small number of base types and arrays, maybe structs as the only user
abstraction.

> >> ] Common Lisp and Scheme.
> >> ]
> >> ] These languages are close to Python in their dynamic semantics,
> >
> >Which was my point.
>
> You said Python was a "Lisp derivitave" with the use of len(obj)
> based in "precedent dating back to the original language."  You
> also said "Python claims a Lisp heritage."

I didn't say "it was".  Rather my exact words were, "it seems" and "I expect".
Clearly I was admitting to some speculation on my part in the matter.

> I am not attempt to say that Python and Lisp don't have similar
> semantics.  What I am saying is that these specific points you
> made are not correct.

See my retraction above.  Despite all the nit-picking, you have agreed that
there are many similarities.

And I think my original comments about len() still hold weight.

> Python isn't a Lisp derivative, except in
> the sense that all modern languages have Lisp influence.

Not true.  Java and C++ are modern languages that have comparatively little in
common with Lisp.

> And that's not the sense you mean since you said "So from a linguistic
> heritage standpoint, it's actually rather odd that Python lacks
> some equivalent of Scheme's IF."

See retraction above.

Strike "from a linguistic heritage standpoint, it's".  Insert "I find it to
be".

> > I don't agree.  I think syntax and semantics are almost orthogonal,
> > with syntax being the less complicated, less important of the two.
> > The fact that Python and Scheme have similar semantics make them very
> > similar despite dramatic differences in their outer appearance.
>
> Spoken, by the way, like a Lisp programmer.

You don't have to resort to insults.  ;o)

> I found that language
> almost impenetrable because of the syntax and my several attempts to
> learn the language have failed.  I enjoy Python's semantics, which
> implies at least to me that syntax is very important.  Perl and Python
> classes also have similar semantics -- Perl got them from Python --
> but I couldn't figure out how to do OO programming in Perl because
> the syntax was so cumbersome.

You misread my statement.  I never said syntax wasn't important.  Rather I said
they were orthogonal, which means largely independent.  That is languages with
similar semantics might vary considerably in syntax. E.g., Lisp, Perl and
Python.  Similarly, languages may have very similar syntax and yet depart
considerably in semantics.  E.g., Java, C++ and C# all share a very similar
syntax although they have considerable differences in semantics.

Your anecdotes about lisp and Perl I think MAKE my point.  I agree that Lisp's
and Perl's syntax are significant barriers to understanding, at least for some
people.  That's why I LIKE Python so much.  It's the best way I've seen thus
far for interfacing to all that functionality.

> The only this I can add is that after watching many non-programmers
> (biologists, chemists, and physicists) learn Python, *none* have
> complained to me that list.len() doesn't work.

Well if they never complained then we'd be silly to listen to the complaints of
professional programmers who do find it odd.

Really, what's so wrong about adding a synonym?

> Instead, since your commentary is at least partially based on a
> view of history which is incorrect, I decided to comment on that
> instead.

Thanks for setting the record straight.  Sincerely and truly.

> I could also have pointed out that except for metaobjects (which in Python
> comes more from a Smalltalk heritage) the way Python has done classes
> has effectively not changed in the last 10 years, so another statement
> you made

> > over time Python has increasingly placed much greater emphasis on Objects

> is suspect.  But I don't have a firm understanding of the history of
> that aspect of Python's history to back up my belief.  Neither, I
> believe, do you.

Here I was talking about changes in the several years SINCE I've been an avid
user, so I was not speculating.  While it's true that some things have not
changed, there has been a lot of activity to change the language in a way that
places a greater emphasis on Objects.  Seems to me the entire effort to
consolidate "types" and "objects" is a tremendous amount of work to shift in
the OO direction.  Similarly, my original example, where "".split() used to be
an error but now it's ok -- why the hell did len() get left behind?  Now you
don't even have to import string to get at the string library, as most of the
functions are now part of the "string class/type."  And, oh yeah, Metaobjects
-- that counts too, right?  So I think it's perfectly valid to say Python's
emphasis on OO has increased over time.  Don't you agree?

I hope you don't take umbrage at anything I say here.  I think civil arguments
are a way for both parties to learn.  I certainly learned a few things, at
least about Python "heritage".

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From geoff at gerrietts.net  Mon Apr 22 21:01:28 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Mon, 22 Apr 2002 18:01:28 -0700
Subject: Proper syntax for adding methods to a class instance dynamically?
In-Reply-To: <aa2ahm$8pr$1@peabody.colorado.edu>
References: <aa2ahm$8pr$1@peabody.colorado.edu>
Message-ID: <20020423010128.GC13077@isis.gerrietts.net>

Quoting Fernando P?rez (fperez528 at yahoo.com):
> I'd like to know what the proper way to do this is. I've been using the new 
> module with new.instancemethod(), but the docstring for new isn't very 
> confidence inducing:
> 
> Is there a simpler, perhaps 'safer' way to achieve the same result? What I 
> want is to have code which runs after an instance of a certain class has been 
> created and which can add new methods to the class instance. I'd also like to 
> know how to add them to the whole class. I have some solutions in store but I 
> feel they're kind of kludgy.

I don't know much (read: anything) about instance methods, or how to
generate them, etc. So I can't help you much there. But, for adding a
method to a class, that's easy:

>>> class Spam:
...   pass
...  
>>> def eggs(self, stuff):
...   print self, stuff
... 
>>> Spam.eggs = eggs
>>> i = Spam()
>>> i.eggs("stuff")
<__main__.Spam instance at 0x81500e4> stuff


That'll work for a new-style class, too.

--G.

-- 
Geoff Gerrietts <geoff at gerrietts net>          http://www.gerrietts.net/
     "Now, now my good man, this is no time for making enemies." 
      --Voltaire, on his deathbed, when asked to renounce Satan




From tejarex at yahoo.com  Thu Apr 11 12:04:02 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Thu, 11 Apr 2002 16:04:02 GMT
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu> <3CB4B1D6.1030504@dittmar.net> <slrnab9kqn.i8b.huaiyu@gauss.almadan.ibm.com>
Message-ID: <Stit8.85858$r7.7473007@bin8.nnrp.aus1.giganews.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnab9kqn.i8b.huaiyu at gauss.almadan.ibm.com...
> >This makes it possible to delete items from the list as the access
is
> >back to front.
>
> I've seen this several times in the past, in the context of making
the loop
> counter correct.  But how does it fare in terms of performance?
Does the
> rest of the list get copied with each del?   This sounds like an
O(nm)
> operation, where n is the list length and m is the number of del's.

Yes.  To keep the operation O(n), scan the list and replace values to
be deleted with a delete flag (such as None) that does not otherwise
appear in the list.  Then filter out the delete flags with (unchecked)
filter(lambda x: x != flag, flagged_list) .

Terry J. Reedy





From mwh at python.net  Mon Apr 22 08:44:14 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 22 Apr 2002 12:44:14 GMT
Subject: curses mouse-events in linux console
References: <3CC3E067.6080107@pp.htv.fi>
Message-ID: <lk662jrjav.fsf@pc150.maths.bris.ac.uk>

janne halttunen <jhalttun at pp.htv.fi> writes:

> I'm on Mandrake Linux 8.2 with Python 2.2, and I have a problem of
> mouse-events not emitting properly in curses when on console with
> gpm. In X it works fine, the getch-method returning instantly on
> mouse-clicks, but on console it doesn't return until after pressing
> key on the keyboard.

Umm, does any curses application respond to the mouse on the console?
I thought it was an xterm-and-derivatives only thing, but I could well
be wrong.

Cheers,
M.

-- 
  > Emacs is a fashion statement.
  No, Gnus is a fashion statement.  Emacs is clothing.  Everyone 
  else is running around naked.
        -- Karl Kleinpaste & Jonadab the Unsightly One, gnu.emacs.gnus


From johnroth at ameritech.net  Mon Apr 22 09:07:59 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 22 Apr 2002 09:07:59 -0400
Subject: Interesting topics
References: <mailman.1019468028.28617.python-list@python.org> <alwyn-94496B.11363122042002@news-text.blueyonder.co.uk>
Message-ID: <uc82ov31i5lb9d@news.supernews.com>

"Alwyn" <alwyn at alwyn.demon.co.uk> wrote in message
news:alwyn-94496B.11363122042002 at news-text.blueyonder.co.uk...
> In article <mailman.1019468028.28617.python-list at python.org>,
>  Doru-Catalin Togea <doru-cat at ifi.uio.no> wrote:
> >
> > Just curious: I have come across some pretty weird statements in
emails,
> > like the two spam mails supposedly from Jessica F. In the last one,
one
> > can read:
> >
> > Who will we open after Bernadette kicks the pathetic star's boat?
> >
> > We fear the polite gardner.
> >
> > I wonder, are these statements computer generated?
>
> Yes. I believe the Hipcrime tool does this.
>
> > Or do they make sense
> > if you have seen a certain movie, a TV-show, or read a certain book?
I
> > don't want to hurt anybodys feelings, but TO ME, their being sent in
> > emails seems like the symptoms of a mental disorder of some kind.
>
> The point of the random 'gibberish' sentences is to make each message
> unique, so as to evade the spam detectors.

Since I haven't seen them, they obviously aren't getting past
my vendor's spam detectors.

John Roth





From gtaylor at lowebroadway.com  Wed Apr 24 04:41:11 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 24 Apr 2002 01:41:11 -0700
Subject: Python, PyQt make install problems on IRIX
References: <f0fd5987.0204220106.7673e345@posting.google.com> <3cc45c8d$0$307$e4fe514c@dreader1.news.xs4all.nl> <f0fd5987.0204230057.25fdc747@posting.google.com> <aa3r2a$tiabi$1@fido.engr.sgi.com> <3cc59530$0$41709$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <f0fd5987.0204240041.1a771008@posting.google.com>

Boudewijn Rempt <boud at valdyas.org> wrote in message news:<3cc59530$0$41709$e4fe514c at dreader4.news.xs4all.nl>...
> offer at sgi.com wrote:
> 
> > $ from gtaylor at lowebroadway.com -#196241 | sed "1,$s/^/> /"
> >>
> >>
> >>Irix certainly is'nt all that common, in a way I think that kind of
> >>helps people make installers, because there are not many hardware
> >>configurations to support or variations of the OS, like Linux, there
> >>are loads of variations on the same thing.
> > 
> > Hey, Irix is pretty common here :-)
> 
> I would think so :-).
> 
> > 
> > Anyway its my intention to ship PyQt in the next release of Irix Freeware
> > (Aug), so I'll keep an eye out for this problem (I have no idea what's
> > causing it).
> > 
> 
> That's very cool!

Hey, a man from the Hallowed City, I think shipping PyQt on the
Freeware CD is great, Python & QT seems to be very popular now. But
anyway, I tried running make install a bit at a time, and found out
that this one line is making the difference:

/bin/sh ..libtool --mode=install /usr/freeware/bin/install -c
libqtcmodule.la /usr/local/lib/python2.2/site-packages/libqtcmodule.la

I can run that line after a restart and PyQT programs will work, does
anyone know what it does that is not still effected after a restart?

Thanks

Garry


From roy at panix.com  Sat Apr 13 19:18:19 2002
From: roy at panix.com (Roy Smith)
Date: Sat, 13 Apr 2002 19:18:19 -0400
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de>
Message-ID: <roy-0BE2A8.19181913042002@news1.panix.com>

martin at v.loewis.de (Martin v. Loewis) wrote:
> One might question whether the bitwise operators should have been
> there in the first place - functions can do the same things just as
> well.

Any operator can get replaced by a function call.  It's just a matter of 
what's more convenient to write.


From shippy at nmt.edu  Fri Apr 26 15:00:50 2002
From: shippy at nmt.edu (Jeff Shipman)
Date: Fri, 26 Apr 2002 13:00:50 -0600
Subject: cgi module question
Message-ID: <3CC9A3E2.9060803@nmt.edu>

I'm using Python 1.5.2 and the cgi module. I'm
wondering if there's any way to recover the string
that was passed to the cgi module during a POST
operation (normally comes in through stdin)? It's
very important that I get the string in the exact
form that it was posted to the CGI.

-- 
Jeff "Shippy" Shipman     E-Mail: shippy at nmt.edu
Computer Science Major    ICQ: 1786493
New Mexico Institute of Mining and Technology
Homepage: http://www.nmt.edu/~shippy



From mtaylor at lowebroadway.com  Mon Apr  8 05:55:21 2002
From: mtaylor at lowebroadway.com (Moray Taylor)
Date: 8 Apr 2002 02:55:21 -0700
Subject: import and function question
Message-ID: <d906b2c1.0204080155.77b87631@posting.google.com>

Hi,

I have two questions..

a) Can I 'import' a module with the full path, i.e. I don't want to
add it to sys.path, just something like 'import /home/me/mod.py', is
this possible.
I have tried adding it to the path, importing, and immediatly removing
it from the path, but the next time I do an 'import' with a different
module of the same name, it picks up the old one, even though it isn't
in the sys.path.

b) I need to overwrite a function with a new one of the same name, but
'import'ed from a module. Is this possible, I really just need to
destroy this function before I import the new one, can this be done?

The reason I want to do these peculiar things, is for a plugin system,
where there are lots of plugin.py files, but in different directories.

Thanks for any help on this.

Moray


From niemeyer at conectiva.com  Tue Apr 30 18:59:07 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Tue, 30 Apr 2002 19:59:07 -0300
Subject: Why do class methods share mutable defaults between instances?
In-Reply-To: <000701c1f094$00688380$8210a8c0@singularity>
References: <000701c1f094$00688380$8210a8c0@singularity>
Message-ID: <20020430195907.A27672@ibook.distro.conectiva>

> class Demo:
>     def __init__(self,name,dict={},list=[]):
[...]

That's not a problem. It's a known behavior. The code "def func(...):" will
be executed just once. It means your dict={} will be executed just once
as well. To understand better, think about this:

def __init__(self, name, dict=func()):

Would you expect func to be executed each time you call the function?

> My intent was that each time Demo was instanciated a new dictionary would be 
> assigned to 'dict' if there was no corresponding argument.  However, the 
> behavior is as though the {} were replaced with a class constant.

It's not replaced by a class constant. It's evaluated, and the resulting
dictionary instance is assigned as the parameter default value.

> In order to avoid this behavior I had to reorganize Demo as follows:
> 
> class Demo2:
>     def __init__(self,name,dict=None,list=None):
>         if dict==None:
>             dict = {}

That's the right way to do it.

> I am wondering if this is a bug or a subtle usage error on my part.  I would 
> expect that syntax on functions would be valid for class methods as well, but 
> in this case you get drastically different behaviors.

I'm not sure what you mean by "different behaviors". That's the way
functions work as well. Have a look:

>>> def demo(a, b={}):
...   b[a] = 1
...   print b
... 
>>> demo("a")
{'a': 1}
>>> demo("b")
{'a': 1, 'b': 1}

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From frr at easyjob.net  Mon Apr 29 09:16:29 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Mon, 29 Apr 2002 15:16:29 +0200
Subject: CGI and client's IP?
References: <1a6qcuc92l8ac0tcegqj95fqemdctda21v@4ax.com> <mailman.1020077596.31223.python-list@python.org>
Message-ID: <pshqcugj7g51in405gjh68achp72koskmf@4ax.com>

On Mon, 29 Apr 2002 12:25:13 +0200, Gerhard H?ring <gh_pythonlist at gmx.de>
wrote:

>* Fernando Rodr?guez <frr at easyjob.net> [2002-04-29 11:58 +0200]:
>> Can I get the client's IP froma a cgi script? O:-)
> 
>This might do it: os.environ["REMOTE_ADDR"]
>
>It's instructive to look at the output of this little CGI script:

Very instructive. Thanks! :-)





-----------------------
Fernando Rodriguez


From jeanfrancoismenard at msn.com  Mon Apr  8 08:05:55 2002
From: jeanfrancoismenard at msn.com (=?ISO-8859-1?Q?Jean-Fran=E7ois_M=E9nard?=)
Date: Mon, 08 Apr 2002 08:05:55 -0400
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <3CB187A3.5080501@msn.com>

Harald Schneider wrote:
> Hi,
> 
> what are your experiences, using wxPython/wxWindows in large projects?
> 
> Large means in my case a sales system, based on MySQL.
> I'm thinking about implementing it in wxPython, but I'm not sure about the
> stability of
> wxWindows and the wxPython wrapper ...
> 
> The alternative way would be a project in Borland Delphi, which has proofed
> its
> stability over years ...
> 
> So what do you think about it!
> 
> Thanks a lot,
> Harald
> 
> 

I would suggest you have a look at Jython.  I just discovered it, and 
I'm amazed.  You can code user interface (Swing !) in Python, 
interactively!

You can import and use almost *any* java package, and most of the 
standards python libraries.  For database connectivity, you have the 
choice to use JDBC directly, or through the Python DB layer!

 From now on, I'll never look back at Java...



From starx at pacbell.net  Mon Apr 22 11:41:18 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 22 Apr 2002 15:41:18 GMT
Subject: Interesting topics
References: <mailman.1019468028.28617.python-list@python.org>
Message-ID: <Xns91F85897F9A6DRASXnewsDFE1@130.133.1.4>

Doru-Catalin Togea || Mon 22 Apr 2002 02:32:32a:

> 
> Hi all,
> 
> I have subscribed to several other lists besides this one, (not python
> lists) and I noticed that there are being posted quite interesting
> mails here. On other lists, even though unmoderated, people do not
> tolerate that many off-the-topic postings. 
> 
> Personally, I don't really care, YET. I have to delete over 150
> postings everyday anyways, so I've become really proficient at hitting
> the 'D' key which marks mails for deletion in PINE. But if the ratio
> of noise to signal increases too much on this list, it can become
> boring to remain on it.
> 

Remember, the mailing list you're on is tied directly into a usenet 
newsgroup called comp.lang.python. Newsgroups and the related software 
all scale a lot better to this kind of volume, with the ability to just 
read what you want and ignore the rest, and i'm not sure in general, but 
i'm pretty sure it's a lot faster than it would be to get all this 
through the little pipe from our crappy email servers here =).

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From pyth at devel.trillke.net  Sun Apr 21 14:49:20 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 21 Apr 2002 20:49:20 +0200
Subject: Modules that provide differing functionality amongst different Python versions...
In-Reply-To: <a9v02u$lal$1@panix3.panix.com>
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <a9ut8d$dql$1@panix3.panix.com> <mailman.1019412399.4283.python-list@python.org> <a9v02u$lal$1@panix3.panix.com>
Message-ID: <20020421184920.GI10574@devel.trillke>

On Sun, Apr 21, 2002 at 02:23:26PM -0400, Aahz wrote:
> In article <mailman.1019412399.4283.python-list at python.org>,
> holger krekel  <pyth at devel.trillke.net> wrote:
> >On Sun, Apr 21, 2002 at 01:35:09PM -0400, Aahz wrote:
> >> In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
> >> Allan Crooks <googlegroups at sixtyten.org> wrote:
> >>>
> >>>Another thought I've had is doing it by specifying, say for module
> >>>'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
> >>>what versions of Python it is running on, and then compiling all
> >>>modules it can deal with, and then importing all objects into it's
> >>>namespace.
> >> 
> >> Yup.  But do it with a package.
> >
> >But doesn't this lead to lots of duplicated -hard to maintain-  code?
> 
> Depends.  You can certainly construct your package to use common code.
> But if you're truly set on using the idioms of each Python version to
> the full extent while still providing backwards compatibility, you're
> going to have a problem with code duplication.  Period.

Depends. You can certainly construct your code so that e.g.
it either uses generators or fully-constructed lists. Python supports 
a nice abstraction:

for item in func():
	whatever_you_do_to(item)

func() could return a list or be a generator, you never know :-)
It's probably no coincedence that generators and fully constructed
lists are often exchangeable for 1:n relations (e.g. directory-lists).

filter(somefilter, func()) # generator or list? don't care :-)

But if you are truly set on using incompatible constructs  
*everywhere* for the fun of it, you are going to have a problem
with code duplication :-)

	Holger




From logiplexsoftware at earthlink.net  Fri Apr  5 18:19:42 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 5 Apr 2002 15:19:42 -0800
Subject: icotl and Pointers
In-Reply-To: <200204052206.g35M6X304288@catfish.lcs.mit.edu>
References: <200204052206.g35M6X304288@catfish.lcs.mit.edu>
Message-ID: <20020405151942.2ae62c0f.logiplexsoftware@earthlink.net>

On Fri, 05 Apr 2002 17:06:33 -0500
Umar Saif wrote:

> Hi,
> 	I am trying to write a piece of python code to control a sound device
(with 
> an oss driver).
>  The ioctls for the sound driver take in a pointer to an integer for the 
> argument -- I was wondering whether it is possible to do this in python ?
Or 
> writing a native library is the only way around this ?

This is from a fix sent to me for uspp by its author, Isaac Barona:

import struct
b = fcntl.ioctl(self.__handle, TERMIOS.TIOCINQ, struct.pack('i', 0))
n = struct.unpack('i', b)[0]

This is ugly but it works.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From jim at jim.jim  Wed Apr  3 22:39:14 2002
From: jim at jim.jim (jimh)
Date: Wed, 3 Apr 2002 19:39:14 -0800
Subject: cgi.py performance
Message-ID: <a8ght4$o8o$1@web1.cup.hp.com>

I am using cgi.py to handle the transfer of large incoming files.  Because
of the way that cgi.py handles multipart/form-data, the performance is VERY
poor.  I have rewritten read_lines_to_outerboundary() to process the
incoming data in large (binary) chunks and have gotten the performance of a
40MB file from 1:41 down to just 15 seconds(!).  However, there are some
assumptions that I have made about multipart/form-data format (even though I
have read the RFC) as well as the workings of FieldStorage (some parts are
still a little fuzzy to me).

I guess my first question should be: has something already been done to
address the performance of multipart/form-data?  And if not, could I get
what I have written reviewed?

Jim





From philh at comuno.freeserve.co.uk  Thu Apr  4 14:57:55 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Thu, 4 Apr 2002 20:57:55 +0100
Subject: Booleans, integer division, backwards compatibility; where is Python going?
Message-ID: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>

Recently on this newsgroup there's been a discussion about adding a 
bool datatype to Python; and Guido has decided to do this.

Some time ago, there was a rather contentious discussion about 
changing the semantics of the / operator. To be precise, what should 
it do when the arguments are integers: should 1/2 be 0 or 0.5?

In both cases, there are good arguments in favour of the change. 
(Though having said that, in both cases Python has done perfectly 
well for years without it). The problem with making these changes is 
that they can cause currently-working Python code to fail on new 
versions of the language.

Another problem is that when a language is constantly changing, it 
becomes necessary to constantly relearn it. Python has changed a lot 
in the last few years, for example new operators such as +=, and 
allowing users to subclass from builtin types, are big (and useful) 
changes. Python 1.5.1 was an immature language that could benefit 
from changing.

But is this changing going to be continual? Are the language 
developers constantly going to tinker with the language, out of a 
sense of fun? (Not that having fun is bad, it's just that I don't 
want to code against a continually moving target).

Is there a final destination in sight, which these changes are 
leading to?

(Note that I am not against changes to the standard library as 
opposed to the language itself; nor do I have problems with changes 
that do not break existing code).

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From huaiyu at gauss.almadan.ibm.com  Fri Apr 12 19:40:01 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 12 Apr 2002 23:40:01 +0000 (UTC)
Subject: How to avoid mutual imports
References: <slrnabeai4.lrg.huaiyu@gauss.almadan.ibm.com> <ac677656.0204121403.2bf3c8d8@posting.google.com>
Message-ID: <slrnabes2g.mfc.huaiyu@gauss.almadan.ibm.com>

Tom Good <Tom_Good1 at excite.com> wrote:
>huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote in message news:<slrnabeai4.lrg.huaiyu at gauss.almadan.ibm.com>...
>> I need advice on how to handle the following situation:
>> 
>> utils1.py:
>>     a
>> 	b
>>     c
>>     if __name__ == '__main__':
>> 	   from interface import data_initialization  # circular
>[snip]
>
>From your description I'm not sure what the problem is.  That import
>is not  actually circular because when utils1.py is imported from
>another module, ( _name__ == '__main__' )  will be false, so the
>import statement after it will not be executed.
>
>Am I missing something?

No, I was missing something.  I over-simplified the problem.  In fact there
are other things that depend on this and they were not in the __main__
condition originally.

But in any case, all the circular imports are now removed by either putting
them into function definitions or under the __main__ condition.  I had the
habit of putting all imports at the beginning of a module so they are
obvious.  Now they are generally either at the beginning of the module or
the beginning of __main__, or occasionally in some functions.

Huaiyu


From eppstein at ics.uci.edu  Thu Apr 25 22:21:55 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 25 Apr 2002 19:21:55 -0700
Subject: Python and Cocoa?
References: <mailman.1019784527.3611.python-list@python.org>
Message-ID: <eppstein-593D22.19215525042002@news.service.uci.edu>

In article <mailman.1019784527.3611.python-list at python.org>,
 Chris Watson <opsys at voodooland.net> wrote:

> I just ordered my first mac, a dual Ghz G4 with OS X. (I know ill hate myself 
> in july when Apple hopefully gets a faster bus that isnt in the stoneage and 
> DDR ram.) Are there any hooks into Cocoa with Python? I have been searching 
> google for a bit with not very either/or results. It seems someone was 
> working on MacPython to work with Cocoa but im not sure what stage thats at. 
> Does anyone know of anything useable to write Cocoa apps with Python?

I'd like to know too.
It is certainly possible to write Carbon apps.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From op73418 at mail.telepac.pt  Wed Apr  3 13:16:29 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 03 Apr 2002 19:16:29 +0100
Subject: __name__ on new-style classes
References: <stcfaugkbh0u65d1euuofhs2b69ej7i34h@4ax.com> <3CA7C873.D1AC908A@python.org> <jvogausg16i1vgp93qkbb8f9sb3nicgb7u@4ax.com> <3CAA619C.A38F2DC9@python.org> <t7cmau88q9knlq96ucseofcmtdbsn7viev@4ax.com>
Message-ID: <mkhmaucfkmgfgr5raq1c8hvqce8sgcgu9u@4ax.com>

On Wed, 03 Apr 2002 17:52:22 +0100, Gon?alo Rodrigues
<op73418 at mail.telepac.pt> wrote:

>On Wed, 03 Apr 2002 01:53:56 GMT, Guido van Rossum <guido at python.org>
>wrote:
>
>>"Gon?alo Rodrigues" wrote:
>>> 
>>> On Mon, 01 Apr 2002 02:35:47 GMT, Guido van Rossum <guido at python.org>
>>> wrote:
>>> 
>>> >It's a feature.  The reason for the change is that I've never seen an
>>> >example of a use for it (as opposed to changing an instance's __class__).
>>> >If you have one, let us know.
>>> 
>>> I bumped into this "feature" a few days ago trying to understand
>>> metaclasses in Python. And to understand them there is nothing better
>>> than actually coding one, so I'll explain my problem/solution and let
>>> you judge if this is a good use case for having a writable __name__.
>>> 
>>> My problem is modelling finite arithmetic in Python. Here by finite
>>> arithmetic I mean all the quotients of the integer ring Z. There is a
>>> countable number of them, one for each n > 1 (n=0 is Z and n=1 is the
>>> trivial ring {0}. If n is negative then it is the same as -n). If n is
>>> prime then we actually have a field.
>>> 
>>> The obvious way (to me, that is) to do this is to let each finite ring
>>> Z_n be a class, inheriting from int. Since there are an infinite number
>>> of these we have to have a metaclass, call it Field, that has as
>>> intances the Z_n rings, e.g. Z_n would be Field(n). The skeleton of this
>>> metaclass would be something like:
>>> 
>>> class Field(type):
>>>     """The Finite Field metaclass."""
>>> 
>>>     def __new__(cls, n):
>>>         if isinstance(n, int) and n > 1:
>>>             #The template class.
>>>             class Z(int):
>>>                 pass
>>> 
>>>             #Generate proper name for the instance class, e.g. "Z_n"
>>>             Z.__name__ = "Z_%d" % n
>>> 
>>>             #Return class.
>>>             return Z
>>>         else:
>>>             raise TypeError, "%s is not an integer > 1." % str(n)
>>> 
>>> If I cant write to __name__ then all the instances will have the same
>>> name... Not A Good Thing, IMHO.
>>
>>But you're doing the meta-instantiation all wrong.  Rather than using a
>>class
>>statement, you should be using cls.__new__(cls, "Z_%d"%n, (int,), {...}).
>>
>>--Guido van Rossum (home page: http://www.python.org/~guido/)
>
>Thanks for the correction. I should have figured that one out. Still,
>shouldn't it be
>
>type.__new__(cls, "Z_%d" % n, (int,), {...})
>
>(instantiating the type class not the cls class) Btw, could you tell me
>what is inherently wrong in generating classes with a class statement?
>Could the generated class somehow behave improperly?

Forget it. I've been reading the descintro and it becomes kinda obvious.
>
>All the best,
>Gon?alo Rodrigues

Thaks anyway,
Gon?alo Rodrigues



From pje at telecommunity.com  Wed Apr 24 08:24:44 2002
From: pje at telecommunity.com (Phillip J. Eby)
Date: Wed, 24 Apr 2002 08:24:44 -0400
Subject: Software Engineers Wanted - 2 positions - Boca Raton, FL
Message-ID: <MPG.17307c2b9af25858989688@news.supernews.com>

Hi all.  The position described below is open at Verio for two software 
engineers in Boca Raton, Florida, USA.  These are not contract positions, 
and you must be local or willing to relocate at your own expense.  Please 
e-mail a plain text resume (or URL to a web-based HTML one) along with 
salary requirements, or apply via the job postings at verio.com.  Thanks.


Key Responsibilities and time percentages thereof:

30% - Work with users to satisfy their needs, by establishing and 
refining requirements and developing solutions, preferably solutions that 

do not require any programming to be done.

15% - Participate in team decisionmaking, by asking intelligent 
questions and actively expressing the reasoning behind his or her views.

40% - Develop software with an active sense of ownership in the process 
and its results, on a variety of time scales from hours to weeks, either 
as independent tasks or as part of a group effort.

12.5% - Support our team's operators and administrators by providing
technical assistance on our software, or by serving as a standby system
administrator.  (The right candidate may at some point be eligible to 
take on-call duty and its associated pay, but this is not required or 
expected.)

2.5% - Handle unexpected, non-routine matters, exercising personal 
judgment in a way consistent with the existing values of the department 
and the business.


Essential Requirements

* Flexibility - should have experience working on a variety of kinds 
of software, preferably with a variety of tools

* Experience developing software in a fast paced, small group (2-10) of
developers, preferably with involvement in architectural decisions and 
the use of revision control software.  (Note: experience with larger 
teams is NOT a substitute for this requirement and may actually be a 
MINUS.)

* Ability and experience developing software for Unix-like platforms 
such as GNU/Linux

* Strong object-oriented development skills, relational database 
development experience, and knowledge of SQL.

* Verbal and written communication skills: should be clear, articulate, 
and persuasive, in writing and in speech.

* Analytical ability: should be able to tell the difference between 
the solutions people propose to fix their problems, and what their 
problems actually are.

* Must understand web application development; should be familiar with
HTML, JavaScript, and application server development such as servlets,
FastCGI, or other long-running server process approaches.


Non-Essential Requirements

* Python experience preferred, and you will have to learn it if you don't
know it.  But we'd much prefer a well-rounded developer who's new to
Python, over an experienced Python developer who's not well-rounded.

* Experience using "agile development" methods such as Extreme 
Programming or Scrum

* Experience in non-programming work, especially as a small business
manager or owner


Position Conditions

* Some short-term trips may be required for special projects

* Use of Python, CVS, and peer code reviews are mandatory. 


From mgilfix at eecs.tufts.edu  Fri Apr 19 01:24:24 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 19 Apr 2002 01:24:24 -0400
Subject: help splitting strings
In-Reply-To: <c76ff6fc.0204181617.3a79df1e@posting.google.com>; from sjmachin@lexicon.net on Thu, Apr 18, 2002 at 05:17:58PM -0700
References: <8mstbu0p2jlq1b9g1agr7mdcesanqm2evq@4ax.com> <mailman.1019147737.28803.python-list@python.org> <c76ff6fc.0204181617.3a79df1e@posting.google.com>
Message-ID: <20020419012424.B24410@eecs.tufts.edu>

  Ah. I wasn't aware of the CSV module and what it did.
Seems like good advice.

            -- Mike

On Thu, Apr 18 @ 17:17, John Machin wrote:
> Michael Gilfix <mgilfix at eecs.tufts.edu> wrote in message news:<mailman.1019147737.28803.python-list at python.org>...
> > You definitely want to use the re module and finall
> > here.  That way you can have tighter control over what
> > constitutes a field.  The difficulty here is crafting your
> > regular expression so it matches what you want. 
> > 
> 
> Sorry to be flatly contradictory, but you definitely *don't* want
> to use regular expressions to try to unravel CSV format, whether (de
> facto) standard or not. Therein lie both madness and slow processing.
> Take Emile's advice and use Dave Cole's CSV module. If you have data
> that it doesn't handle (i.e. not "standard") then post again or send
> me e-mail -- I am lashing together a module that handles non-standard
> variations of CSV format.
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (sjmachin)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From jon+usenet at unequivocal.co.uk  Mon Apr 15 11:35:03 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 15 Apr 2002 15:35:03 -0000
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org> <3cba62f8@news.mhogaming.com>
Message-ID: <slrnablsp5.f08.jon+usenet@snowy.squish.net>

In article <3cba62f8 at news.mhogaming.com>, Ken Peek wrote:
> This should be fixed so that decimal display of longs uses the '-'
> sign, but hex display does not (you would just get the number of hex
> digits in their raw "twos complement" form that the long object's
> hex display method "knows" there is...)

So you think that:

  long(hex(-1l), 16)

should not equal -1 but, err, some random choice of 15, 255, 4095,
65535, etc?


From mal at lemburg.com  Thu Apr 25 18:01:18 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 26 Apr 2002 00:01:18 +0200
Subject: configuring mxCGIPython
References: <NCBBLFCOHHDIKCAFGCFBEEKEKMAA.marklists@mceahern.com>
Message-ID: <3CC87CAE.6B3F2FD5@lemburg.com>

Mark McEahern wrote:
> 
> [M.-A. Lemburg]
> > No. Could be that libxmlparse is referenced by pyexpat which is part
> > of cgipython. On most systems the expat libs should be part
> > of the system, so setting LD_LIBRARY_PATH to e.g. /usr/lib
> > should help. If not, you can try getting that file from some
> > other source and install it in the same dir as cgipython.
> 
> That seems to be the case.  I found and uploaded:
> 
>   libxmlparse.so.1
>   libxmltok.so.1
> 
> Now, when I run the shell script that wraps cgipython, I get:
> 
>   'import site' failed; use -v for traceback
>   Traceback (most recent call last):
>     File "../../CGIPython/cgipython.py", line 27, in ?
>   IOError: [Errno 2] No such file or directory
> 
> Why would it be trying to do an import site?  Hmm.

Because Python always does on startup if not run with the
-S switch (this is not available in cgipython though).
Could you send the output of running cgipython
with the environment variable PYTHONVERBOSE defined to 1 ?

A simple work-around would be adding the site.py from the
standard Python lib -- this should really be included in
cgipython, though, and I think it is, so there must
be something in site.py which fails.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From offer at sgi.com  Mon Apr 29 10:43:47 2002
From: offer at sgi.com (offer at sgi.com)
Date: 29 Apr 2002 14:43:47 GMT
Subject: PyQt 3.1 / Sip unresovled symbol
Message-ID: <aajm73$v6cvi$1@fido.engr.sgi.com>

I'm trying to get PyQt compiled for the Freeware CD and am hitting an
unresovled symbol...

[ morgaine 6.5.16f-Apr01 ] python2
Python 2.1.1 (#1, Mar 25 2002, 11:44:01) [C] on irix6-n32
Type "copyright", "credits" or "license" for more information.
>>> import qt
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/freeware/lib/python2.1/site-packages/qt.py", line 46, in ?
    import libsip
ImportError: 216936:python2: rld: Fatal Error: unresolvable symbol in /usr/freeware/lib/python2.1/site-packages/libsip.so: __record_needed_destruction


Any ideas on where to look ? It seems moc related....


richard.

-- 

-----------------------------------------------------------------------
Richard Offer                         Technical Lead, Trust Technology.
"Specialization is for insects"
__________________________________________http://reality.sgi.com/offer/



From guido at python.org  Thu Apr  4 09:20:20 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:20:20 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xbsd0qpmn.fsf@ruckus.brouhaha.com> <3CABABCA.EB446FEE@python.org> <7x7kno5isl.fsf@ruckus.brouhaha.com>
Message-ID: <3CAC6205.39B5E01B@python.org>

> Guido van Rossum <guido at python.org> writes:
> > I wouldn't say "should", but it sure would be nice.  Unfortunately the
> > internal architecture of the array module doesn't make this easy; it
> > would probably require a major refactoring, or perhaps a less pervasive
> > switch to new-style classes so that subclassing can be used.
> 
Paul Rubin replied:
> 
> I'm not sure I understand the issue about new-style classes, but if
> old-style classes are going away then new-style has to be supported,
> right?

The array object is an old-style type object.  That's not the same as an
old-style class (it's totally unrelated).  But new-style type objects are
the same as new-style classes (that's what new-style classes are all
about) and it's easy enough to convert an old-style type to a new-style
type.  Then the new-style type will allow subtyping, so you can write a
subtype that implements bit elements.

> Anyway, I'll put it on my list, but don't hold your breath.

I won't, but I'll remind you that if you give, you'll get more back.

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


From dalke at dalkescientific.com  Mon Apr 22 14:34:06 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 22 Apr 2002 12:34:06 -0600
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <3CC45005.2E9C0F86@tundraware.com>
Message-ID: <aa1l4o$h74$1@slb5.atl.mindspring.net>

Tim Daneliuk:
>Well, OK, I'll say it more formally:  Let n denote the hightest dimension
>used by any of the natural sciences.  Mathematicians routinely operate
>in k dimensions where k >> n.   Acceptable?

No.  Hilbert spaces have an infinite number of dimensions.

Doing a quick google search and I found a book titled
  INFINITE DIMENSIONAL DYNAMICAL SYSTEMS IN MECHANICS AND PHYSICS

                    Andrew
                    dalke at dalkescientific.com





From peter at engcorp.com  Sun Apr 14 23:06:33 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 14 Apr 2002 23:06:33 -0400
Subject: Awesome wxPython IDE
References: <3cb9c75d.115396601@news.mweb.co.za>
Message-ID: <3CBA43B9.22850DF1@engcorp.com>

[dang, hit ctrl-Enter on previous post before complete]

sqrt_2 wrote:
> 
> For those of you who develop apps using wxPython there is this great
> product I know of.
> It is called Boa-Constructor and also incorporates Zope.
> Have a look at:
> http://boa-constructor.sourceforge.net

Of course, those products are well known around here.

What's less well known is the relationship between Boa-Constructor 
and Zope.  I strongly suspect there isn't one.

Can you elaborate on this supposed "incorporation of Zope
into Boa-Constructor"?

-Peter


From maxm at mxm.dk  Thu Apr 11 03:53:22 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 11 Apr 2002 09:53:22 +0200
Subject: Tuple comprehension
References: <mailman.1018502663.25698.python-list@python.org>
Message-ID: <3CB540F2.4020805@mxm.dk>

Sean 'Shaleh' Perry wrote:
> On 11-Apr-2002 Emile van Sebille wrote:
> 
>>Magnus Lie Hetland
>>
>>>As far as I can see, the PEP on dict comprehension is still open... I
>>>think both dictionary comprehension and tuple comprehension seem like
>>>natural parallels to list comprehension. But then I would, wouldn't I?
>>><wink>
>>>
>>Why not just tuple([my list comp]) ?
>>
> 
> because it feels wrong to create a (possibly large) list and then duplicate it
> into a tuple.  My mind sees that and thinks 'overhead'.

The natural way to implement it in practice is almost certainly just do 
it behind your back.

Append takes a mutable sequence.

The most effcient mutable sequence in Python is the List, so why should 
they make a "better" mutable sequence to support tuple comprehensions?

And if they did make a better mutable sequence should there really be 
two of them? That would be redundant code that would need to be 
refactored. Probably into the list class.

Then they would use the list class to implement tuple comprehensions...

 From here on my arguments gets a wee bit circular ...

regards Max M



From mertz at gnosis.cx  Mon Apr 29 13:37:21 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 29 Apr 2002 13:37:21 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org>
 <mailman.1019299357.11513.python-list@python.org>
 <a9s72u$gjj$1@slb3.atl.mindspring.net>
 <mailman.1019381679.10995.python-list@python.org>
 <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz>
 <mailman.1019543864.22119.python-list@python.org>
 <m2g01my7bw.fsf@mycroft.actrix.gen.nz>
 <mailman.1019798929.16905.python-list@python.org>
 <m2znzpwc7a.fsf@mycroft.actrix.gen.nz>
 <mailman.1020032874.18514.python-list@python.org>
 <tyf1ycyhmmn.fsf@pcitapi22.cern.ch>
Message-ID: <RTYz8kKkX0hM092yn@gnosis.cx>         

Jacek Generowicz <jacek.generowicz at cern.ch> wrote previously:
|> I feel the default S-expression syntax
|> represent a legitimate barrier to entry, however slight, to new
|> people approaching and learning the language for the first time.

|As your experience suggests, it is a very slight barrier. The
|multitude of people who have never made a sincere effort to overcome
|this barrier

Personally (as someone who has been paid to program maybe a dozen
different languages, over about two decades), I have found S-expressions
to be a *HUGE* barrier to entry from me programming Lisp.  Actually,
EVERYONE else I've spoken to who has tried to learn Lisp has expressed
the same sentiment (except those few in whom it "took").

Yeah, yeah... blah, blah..  I know the underlying semantics are not the
surface syntax.  And I know that macro packages can define other syntax
forms.  This information is *stunningly* irrelevant.  Every Lisp book
and tutorial I've tried to read, and almost every bit of sample code,
uses these same "meaningless" S-expressions.  You really cannot get away
from the parenthesis and prefix notation in practical terms, whatever
sophistical advocates claim to the contrary.

Yours, Lulu...

--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s





From cliechti at gmx.net  Sat Apr  6 15:05:46 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 6 Apr 2002 21:05:46 +0100
Subject: "Memory could not be Read/Written" crashes with C extension DLL... HELP PLEASE!
References: <5a4226f0.0204052223.75aa0318@posting.google.com> <5a4226f0.0204061049.c2b1026@posting.google.com>
Message-ID: <Xns91E8D6C93ED4cliechtigmxnet@62.2.16.82>

kevin at cazabon.com (Kevin Cazabon) wrote in 
news:5a4226f0.0204061049.c2b1026 at posting.google.com:

> Through some help, I've found that this section is causing the
> problem... If I simply use "return Py_BuildValue("lllO", pp1,
> inBufferSize, pp3, inBuffer);", it works fine with no crashes.
> 
> What is the difference though, shouldn't the code below work too?  Is
> there maybe something wrong in the PyTuple code?  (As mentioned
> before, it only crashed when called multiple times (40+ usually), but
> worked fine otherwise.
> 
> Kevin.
> 
>>   // build pp1, inBufferSize, pp3 into PyObjects for returning
>>   pLong1 = PyLong_FromLong(pp1);
>>   pLong2 = PyLong_FromLong(inBufferSize);
>>   pLong3 = PyLong_FromLong(pp3);
>> 
>>   // build pLong1, pLong2, pLong3, pString into a tuple to return
>> 
>>   pReturn = PyTuple_New(4);
>>   PyTuple_SetItem(pReturn, 0, pLong1); 
>>   PyTuple_SetItem(pReturn, 1, pLong2);
>>   PyTuple_SetItem(pReturn, 2, pLong3);
>>   PyTuple_SetItem(pReturn, 3, inBuffer);
>> 
>>   return pReturn;
>> }

PyTuple_SetItem is stealing a reference, but the reference to inBuffer
was only borrowed. There is your wrong count :-)

Python22/Doc/api/tupleObjects.html: """
int PyTuple_SetItem(PyObject *p,  int pos, PyObject *o) 
Inserts a reference to object o at position pos of the tuple pointed to by 
p. It returns 0 on success. Note: This function ``steals'' a reference to 
o.
"""

Python22/Doc/ext/parseTuple.html: """
Note that any Python object references which are provided to the caller are 
borrowed references; do not decrement their reference count! 
"""

chris

-- 
Chris <cliechti at gmx.net>



From bbollenbach at shaw.ca  Fri Apr 12 19:56:45 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Fri, 12 Apr 2002 23:56:45 GMT
Subject: ANN: Twisted 0.16.0: Licenses and Open Source don't conflict.
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <eGIt8.21408$de1.1000064@news3.calgary.shaw.ca> <MPG.172101b4b7f13f639896ae@news.nwlink.com>
Message-ID: <1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca>

> This doesn't change the fact, though, that Public Domain 
> software is Free Software in the FSF  sense of the word.  
> 
> (See http://www.fsf.org/philosophy/categories.html)

Interesting that in the page you cite, they state specifically the
opposite of what you're saying:

<begin gnu>
Public domain software
======================

Public domain software is software that is not copyrighted. It is a
special case of non-copylefted free software, which means that some
copies or modified versions may not be free at all.

Sometimes people use the term ``public domain'' in a loose
fashion to mean ``free'' or ``available gratis.'' However,
``public domain'' is a legal term and means, precisely, ``not
copyrighted''. For clarity, we recommend using ``public
domain'' for that meaning only, and using other terms to convey
the other meanings.
<end gnu>

I overlooked the specific legal meaning of "Public Domain" and therefore
was incorrect in the assertion that the releaser of such software "only
has their implicit copyright". 

However, I was still correct in saying that just because you release in 
the public domain doesn't mean you're releasing "free" or even more 
generally "open source" software, which was mainly the point I was
trying to convey.


--
Brad


From aleax at aleax.it  Tue Apr  9 12:55:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 16:55:24 GMT
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk> <13285ea2.0204090806.4a2d987b@posting.google.com>
Message-ID: <02Fs8.454$vF6.11198@news2.tin.it>

J.Jacob wrote:
        ...
> Is there a way to peek into the latest python-dev messages? I could
        ...
http://mail.python.org/pipermail/python-dev/2002-April/date.html

I don't think there's a way with granularity < 1 month.


Alex





From gh_pythonlist at gmx.de  Mon Apr  8 17:59:01 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 8 Apr 2002 23:59:01 +0200
Subject: CGI multiple select?
In-Reply-To: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>
References: <56ns8.37728$tg4.448055@vixen.cso.uiuc.edu>
Message-ID: <20020408215901.GA1871@lilith.my-fqdn.de>

* John <nobody at nobody.com> [2002-04-08 15:35 -0500]:

Do you own the nobody.com domain? If not, you're abusing their services.
If I were the domain owner, you'd get into trouble. If all you want is
not receive spam, please get yourself a throway account at a freemailer
instead.

> Hi,
> 
> In my html form, I have 'multiple select'
> 
> <select name=spred multiple>
> <option..>
> <option..>
> 
> This form invokes CGI routine, where I try to read selected entries from
> form["spred"]
> 
> My dilemma is here:
> 
> When only one entry is selected:
> form["spred"].value (will be the object selected)
> 
> When multiple entries were selected:
> for o in form["spred"]:
>   o.value (--> each of these will be the object selected)

Dunno. This works fine for me and it gets lists only - a one-element
list for one choice and an empty list if nothing is selected:

#!/usr/local/bin/python
import cgi
try:
    import cgitb
    cgitb.enable()
except ImportError:
    pass

# URL of this cgi script
MYURL = "/cgi-bin/test.py"

print "Content-type: text/html\n\n"

form = cgi.FormContentDict()
optionen = form.get("cities", [])

print """
<html>
<body>
<p>
<b>You selected these entries last time:</b>
%(optionen)s
</p>
<form action="%(myurl)s">
<select name="cities" multiple>
<option>Paris</option>
<option>Moscow</option>
</select>
<input type="submit>
</form>
</body>
</html>""" % {"myurl": MYURL, "optionen": optionen}

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.6 ?C      Wind: 0.7 m/s




From dominikush at yahoo.com  Wed Apr 17 10:03:30 2002
From: dominikush at yahoo.com (dominikush at yahoo.com)
Date: Wed, 17 Apr 2002 16:03:30 +0200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <3CBD7406.3050909@mxm.dk>
Message-ID: <3CBD80B2.91A7D596@yahoo.com>

> > Who is wrong here: my intuition or Python (2.2)? If it's
> > my intuition, how can I train my thinking about Python's
> > execution model, so that my intuition get's better ;-)
> 
> Your intuition!
> 
> Objects like lists gets copied by reference. So Python behaves exactly
> like it should.
> 
> Maybe it is clearer with another notation:
> 
>  >>> dict = {'a':<list ref1>,'b':<list ref2>}
>  >>> list = dict.values()
>  >>> list
> 
> [<list ref1>, <list ref2>]
> 
>  >>> dict['a'].append(3)
>  >>> dict
> 
> {'a': <list ref1>, 'b': <list ref2>}
> 
>  >>> list
> 
> [<list ref1>, <list ref2>]
> 
> regards Max M

I understand that! But my main objection remains: Why should
I care about the type of object being handled by a dictionary?
This could not only break unit tests but also make maintenance
of ``old'' code somewhat painful.

Regards,

Dominikus


From joonas at olen.to  Sun Apr  7 04:01:15 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: Sun, 07 Apr 2002 11:01:15 +0300
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com>
Message-ID: <3CAFFCCB.D7CE6C95@olen.to>

distrex wrote:
> 
> I understand most of python. But I still don't get modules. I know
> they are just basicaly a function but i don't know how to use them
> effectively and what each one does. Everything good in python you will
> at some point need to use a module right. If anyone can help me with
> this i'd be greatful. Info or websites would be nice.

Actually modules are not functions. For further information, refer to
the Modules-section of the Python Tutorial at
http://www.python.org/doc/current/tut/node8.html


From gregor at hoffleit.de  Fri Apr  5 12:14:53 2002
From: gregor at hoffleit.de (Gregor Hoffleit)
Date: Fri, 5 Apr 2002 19:14:53 +0200
Subject: sharp sl5000
In-Reply-To: <MMfr8.62526$S52.2173482@news2.tin.it>
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <4mer8.51712$pT1.1468090@news1.tin.it> <Xns91E77C8D2BE34PaCmAnRDLM@161.48.128.20> <MMfr8.62526$S52.2173482@news2.tin.it>
Message-ID: <20020405171453.GA18450@hal.mediasupervision.de>

* Alex Martelli <aleax at aleax.it> [020405 14:21]:
> Fred Pacquier wrote:
>         ...
> > the water. But just yesterday I read a review of the just-released (in the
> > US) Zaurus, and the last line mentions that Sharp is "thinking about
> > introducing the new Zaurus line in Europe"... next year. So if you want
> 
> Funny -- the developer edition WAS already for sale in UK & Ireland... which
> are part of Europe, of course.
> 
> > one soon you'll probably have to take your chances and order one overseas,
> 
> I was hoping to order one in UK or Ireland (no customs within the EU...).

It definitely sells in Germany: 699,- ? for the SL-5500G. There was an
advertisement yesterday from our local discounter in the newspaper.
Have a look at www.promarkt.de to confirm this; you should be able to
navigate to it using the pictures ;-)

    Gregor




From philh at comuno.freeserve.co.uk  Tue Apr 16 18:23:19 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Tue, 16 Apr 2002 23:23:19 +0100
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <slrnabmtk0.a8q.ak@ak.silmarill.org>
Message-ID: <slrnabp92n.ipr.philh@comuno.freeserve.co.uk>

On Tue, 16 Apr 2002 00:53:33 GMT, Andrei Kulakov <ak at silmarill.org> wrote:
>In article <slrnabmoci.dhk.philh at comuno.freeserve.co.uk>, phil hunt wrote:
>> 
>> OK, I've decided to write a PEP. Preliminary to this, I'd like 
>> people's comments.
>> 
>> I wish to make the pas startement optional. That is, in any Python 
>> program where there is a pass it can be left out and the meaning of 
>> the program is unchanged.
>> 
>
>[snip]
>
>I think the idea behind pass is that usually not having anything in
>a conditional is a mistake, so we want to be informed about it. Putting
>pass is like saying "I'm sure that's what I want."

I'm sure what I want, without needing to explicitly say it.



-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From logiplexsoftware at earthlink.net  Tue Apr 30 13:40:25 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 30 Apr 2002 10:40:25 -0700
Subject: remove read only flag
In-Reply-To: <20020430103320.49a5e140.logiplexsoftware@earthlink.net>
References: <XlAz8.52$FK2.1293@nreader2.kpnqwest.net>
	<20020430103320.49a5e140.logiplexsoftware@earthlink.net>
Message-ID: <20020430104025.6afb440e.logiplexsoftware@earthlink.net>

On Tue, 30 Apr 2002 10:33:20 -0700
Cliff Wells wrote:

> On Tue, 30 Apr 2002 17:18:47 GMT
> Francis Meyvis wrote:
> 
> > Hoi,
> > 
> > How can one remove the read-only flag from a file
> > (e.g. c:\temp\file.txt) using python in win32.
> 
> see os.chmod()
> 
> Hm.  I wonder how this maps to Windows permissions...

Well, it seems that os.chmod(filename, 666) works as expected (AFAICT).

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From pyth at devel.trillke.net  Sat Apr 27 16:23:26 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 27 Apr 2002 22:23:26 +0200
Subject: Update locals()
In-Reply-To: <20020427150607.A2052@unpythonic.net>
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net>
Message-ID: <20020427202326.GB16217@devel.trillke>

On Sat, Apr 27, 2002 at 03:06:08PM -0500, jepler at unpythonic.net wrote:
> On Sat, Apr 27, 2002 at 09:57:25PM +0200, holger krekel wrote:
> > Hello,
> > 
> > how can you dynamically update the locals() namebindings?
> 
> You don't.  See the library documentation:
>     locals()
> 	Return a dictionary representing the current local symbol table. Warning:
> 	The contents of this dictionary should not be modified; changes may not
> 	affect the values of local variables used by the interpreter.

Uff. Thanks for the pointer although it is dissappointing. I always
liked to assume that python does this decision at run time. Doing
it at compile time may have performance benefits but it prevents
some nice uses. 

> Whenever a name is not bound inside a function, it is a global (it could
> never have a definition inside the function, so if the name comes from
> anywhere, it's global).  If there's a "global" declaration, then it's
> global.  If there's an assignment, then it's a local -- but changing a key
> in locals still doesn't work (or is at least permitted not to work).
> 
> You probably need to rethink the way you're approaching the problem.

maybe you have an advice?

Consider you have a dictionary e.g. returned from a
re.match - object. Is there no better way than

def func(somedict):

    ...
        somedict['keyname1']
    ...

to use the info? I have lots of descriptive
Named Patterns (via '?P<name>') and i want to work with
them in the function without redundancies. I really would like
an automatic pattern instead of doing

keyname1=somedict['keyname1']
keyname2=somedict['keyname2']
...

or something similar. I also have other uses but this one is the
easiest to explain.

    holger




From tegbert at abl.med.utah.edu  Tue Apr  2 13:31:39 2002
From: tegbert at abl.med.utah.edu (tegbert)
Date: Tue, 2 Apr 2002 11:31:39 -0700
Subject: calling dll's functions with python
In-Reply-To: <8wmq8.52601$S52.1805120@news2.tin.it>
References: <8wmq8.52601$S52.1805120@news2.tin.it>
Message-ID: <Pine.LNX.4.10.10204021127250.1335-100000@abl.med.utah.edu>

Create Python native methods that "wrap" your DLL calls in a package that
you can import and then call from Python.  Check out Guido's documentation
on "Extending and Embedding" the Python interpreter.  It's in the standard
Python documentation package.

Tim

> On Tue, 2 Apr 2002, Giovanni Laudicina wrote:
>
> I'm a beginner with python and i did'n find this info on the FAQ and
> tutorials:
> How can I call a function included in a dll (I have only the .dll file, not
> the source code) ?
> Thanks to everyone,
> Giovanni Laudicina



From martin at v.loewis.de  Sun Apr 14 05:10:18 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 11:10:18 +0200
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com>
Message-ID: <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de>

grante at visi.com (Grant Edwards) writes:

> I agree.  Since the other boolean operators are "not" "and"
> "or", it would be a bit of a surprise that one doesn't use
> "xor". And having to convert a boolean to a "chunk of bits"
> first and do a bitwise xor on it is just plain ugly.

As Bengt points out, it is tricky to define this analogous to 'and'
and 'or'. Both have a shortcut meaning, evaluating the second one only
if the first does not determine the outcome. Furthermore, they both
return the last expression evaluated. This is not possible for 'xor':
you have to evaluate both, and there is no clear choice as to which
one is better as a result - in fact, the OP wants it to return bool.

So xor would be very much *unlike* 'and' and 'or'.

Regards,
Martin


From alf at logilab.fr  Fri Apr 26 10:05:48 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Fri, 26 Apr 2002 14:05:48 +0000 (UTC)
Subject: What is NotImplemented ?
Message-ID: <slrnacinls.j5.alf@virgo.logilab.fr>

A quick question, here. What is the built in NotImplemented object? 
And what can you do with it? I thought I could raise it as an 
exception, but this is not the case (I use NotImplementedError instead).

 
Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From peter at engcorp.com  Tue Apr  2 23:04:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 02 Apr 2002 23:04:09 -0500
Subject: Newbie: Classes and Lists
References: <8Guq8.1209$zT5.41108@news.itd.umich.edu>
Message-ID: <3CAA7F39.94FE7EA9@engcorp.com>

CGH wrote:
> 
> import random
> 
> class Chromosome:
> 
remove this:
>     data = []
> 
>     def __init__(self, length):
add this:
          self.data = []
>         for bit in range(length):
>             self.data.append(random.randrange(0,2,1))

> I want to define a population of 100 chromosomes. But instead of a list of
> Chromosome objects I seem to get one big (100 x 20) list of 0s and 1s.
> Clearly I've misunderstood something fundamental here. Any help would be
> appreciated.

Yes, you've misunderstood something fundamental: instance variables
instead of class variables.  Basically, your "data = []" was
creating a single field 'data' which was associated with the _class_,
not with a single instance.  When you referenced it with self.data,
Python implicitly used the class variable instead of complaining.
(Perhaps this violates Python's more regular rule of explicit being
better than implicit?  Not sure...)

You need to create instance variables like that by explicit
use of "self.data" (or whatever) in the __init__ method of
the class.  Then each instance gets its own self.data list.

-Peter


From fdrake at acm.org  Wed Apr  3 17:54:42 2002
From: fdrake at acm.org (Fred L. Drake)
Date: Wed,  3 Apr 2002 17:54:42 -0500 (EST)
Subject: [development doc updates]
Message-ID: <20020403225442.0343418EAD1@grendel.zope.com>

The development version of the documentation has been updated:

    http://www.python.org/dev/doc/devel/

Updated to include whatever Guido did about bool().




From martin at v.loewis.de  Mon Apr 15 10:02:13 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 16:02:13 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com>
Message-ID: <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> 2) Secondly, when we are printing integers in the hex format,
> we are not interested in the sign-- we are looking at the
> bit pattern.

Who is this "we" in this statement? When I print numbers in hex, I'm
interested in the hexadecimal representation if the number. For
*printing* it, I usually don't care what the internal representation
is.

> 3) With unification, the long and the int types should not
> print differently when calling the 'hex()' method.

And indeed, they won't.

> (The 'int' version of this method NEVER has, and shouldn't print a
> minus sign, and _I_ don't think the long version of 'hex()' should
> either.  This _is_ what needs to happen to have proper unification.

The PEP currently says differently: 

# In most cases the long int semantics will prevail; however, the
# trailing 'L' from long int representations will be dropped.

This clearly says that the long int semantics will prevail.

> 4) We are not using 1.5.2 anymore.  We are on 2.2.x now.
> You need to imprint on your brain that there now should be
> _NO_ visible difference (to the programmer) between the long
> and the int type.  In fact, the int and long representations
> are now just an implementation detail, and should be hidden,
> and transparent to the programmer.

I completely agree.

> Martin, why do you resist the unification of these two
> types for _ALL_ operations and methods?  

Why did you conclude that I resist? I'm just saying that the long
semantics will prevail, not the int semantics.

Regards,
Martin



From todd at slideburn.com  Fri Apr 12 16:53:50 2002
From: todd at slideburn.com (Todd Chapman)
Date: Fri, 12 Apr 2002 20:53:50 GMT
Subject: PyQt Installation Problem...
Message-ID: <3CB749CB.10701@slideburn.com>

Hello,

Just installed SIP and PyQt and have come across a slight twist.  I'm on 
  Red Hat 7.2 and after compiling, PyQt installs in the directory where 
it was extracted and not in the site-packages subdirectory of my python 
installation.  Looks like the files that were generated in the cong 
stage got located the right python version and directory, it just didn't 
go there (/usr/lib/python2.2/site-packages).  Any suggestions?

Todd



From dalke at dalkescientific.com  Mon Apr 15 15:16:13 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 15 Apr 2002 13:16:13 -0600
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <slrnabi0f4.15a.grante@tuxtop.visi.com> <m31ydiy7b9.fsf@mira.informatik.hu-berlin.de> <eppstein-7156A5.13521614042002@news.service.uci.edu> <3cba0068@news.mhogaming.com> <vTou8.20361$8W5.17808@nwrddc02.gnilink.net> <3cba4868@news.mhogaming.com>
Message-ID: <a9f9rk$h8s$1@slb2.atl.mindspring.net>

logistix:
>It gets a little trickier though.  Here's the real reason for short-circuit
>evaulation:
>
>>>> a = 1
>>>> a or c
>1
>>>> c or a
>Traceback (most recent call last):
>  File "<interactive input>", line 1, in ?
>NameError: name 'c' is not defined
>>>>

Ken Peek:
>An undefined variable is a programming disaster waiting to
>happen, and so should ALWAYS throw an exception if you try
>to access it before it is defined.

Perhaps logistix used a bad example.  I think the point he was
making is that 'and' and 'or' does short circuit logic, so the
full expression is not evaluated every time.  This is useful,
as for expressions like

  value = kwargs.get("-filename") or ask_user_for_filename()

where if the filename isn't specified, say, on the command line
then there's an interactive prompt to get it from the user.

This is useful since the other approach is

value = kwargs.get("-filename")
if not value:
  value = ask_user_for_filename()

which is longer so harder to read.  This is even more useful for
chains where

  value = f() or g() or h()

which I've used often for getting values which can be from multiple
sources (as with the chain command-line, environment variable, config
file, default setting).  The 'if' for this is 5 lines long.

Because of the short circuiting, 'and' and 'or' must done as operators
and not as function calls.  On the other hand, xor must evaluate both
terms to do the computation, so there is less need for xor to be
a new operator.

                    Andrew
                    dalke at dalkescientific.com





From ak at silmarill.org  Mon Apr 15 18:39:54 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Mon, 15 Apr 2002 22:39:54 GMT
Subject: Python newbie
References: <ab96a890.0204131438.5cce8f66@posting.google.com>
Message-ID: <slrnabmlpd.a8q.ak@ak.silmarill.org>

In article <ab96a890.0204131438.5cce8f66 at posting.google.com>, slack wrote:
> Hello,
> I'm interested in Python programming language since one around one
> year. I think it is an interesting language because it's very simple
> to use this language to program windows... We don't have to think
> about event handling instead of the purpose of the program itself...
> well... I don't want to explain why I find it interesting... I've
> installed a version of Python on Windows computers but I find it very
> slow compared to Linux version of Python. I've noticed that the first
> time I used Python on Windows. Well. This is not really the problem. I
> want to use Python on Windows but I don't know if I can convert .py
> files into .exe files? Is it possible thanks to some utility?
> 
> thank you.
> 
> slack
>
You don't need to make it into an exe file to use it on windows. You can
double click on a .py file to run it.


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From mal at lemburg.com  Fri Apr 26 13:55:42 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 26 Apr 2002 19:55:42 +0200
Subject: mx.BeeBase.BeeDict
References: <tYay8.8691$WJ4.11752@news1.bredband.com>
Message-ID: <3CC9949E.4A6F3C8D@lemburg.com>

Magnus Heino wrote:
> 
> Could anyone please show me an example usage of mx.BeeBase?
>
> Thanks!
> 
> [magnus at daysleeper magnus]$ python2.2
> Python 2.2.1 (#2, Apr 13 2002, 14:35:50)
> [GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> from mx.BeeBase.BeeDict import BeeStringDict
> >>> s=BeeStringDict('foo')
> >>> s['foo']='bar'
> >>> s['foo']
> 'bar'
> >>> s.commit()
> 2000 2002-04-26 13:33:54.74 [p4474] Committing all changes for "foo"...
> 2000 2002-04-26 13:33:54.74 [p4474]  writing key "foo":
> -------------------------------------------------------------------------------
> 'bar'
> -------------------------------------------------------------------------------
> not found; cc=-1
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/opt/lib/python2.2/site-packages/mx/BeeBase/BeeDict.py", line 1067,
> in commit
>     index[key] = address
> TypeError: record address must be an integer
> >>>

Please upgrade to the lastest egenix-mx-base beta which I've
announced on the egenix-users mailing list. What you are
seeing is the result of the file object methods returning
longs instead of integers and mxBeeBase in the GA vesion
of egenix-mx-base is currently not capable of handling these.

The beta is, though.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/




From elimli at yahoo.com  Sun Apr 21 14:24:36 2002
From: elimli at yahoo.com (eli li)
Date: Sun, 21 Apr 2002 20:24:36 +0200
Subject: constructors in python
References: <a9uoa0$t0r$1@news.iucc.ac.il> <3CC2ED37.3D9225B3@engcorp.com>
Message-ID: <a9usk3$rk$1@news.iucc.ac.il>

But what does happened when I forget (I'm just a man) to call father's c-or?
Semi-initialized object? C++ doesn't permit not to call the base c-or.

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CC2ED37.3D9225B3 at engcorp.com...
> eli li wrote:
> >
> > Consider, we have two classes: one derived from other. In derived class
c-or
> > how is it possible to call the base c-or? Simple call base.__init__
(...)?
>
> Assuming c-or means constructor:
>
> [untested code]
>
> class Base:
>    def __init__(self):
>        return
>
> class Derived(Base):
>    def __init__(self):
>        Base.__init__(self)
>        return
>
>
> -Peter




From cpbotha at i_triple_e.org  Mon Apr  1 14:04:16 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Mon, 01 Apr 2002 19:04:16 GMT
Subject: GUI programming
References: <mailman.1017437568.25402.python-list@python.org> <a84561$p5g$1@wanadoo.fr> <slrnaabbct.npb.cpbotha@crabtree.cpbotha.net> <4fsdauofmus28t630866pgf4cu3n6bp16m@4ax.com> <20020401020912.GE333@lilith.hqd-internal> <mailman.1017681605.21669.python-list@python.org>
Message-ID: <slrnaahc1v.2ge.cpbotha@crabtree.cpbotha.net>

In article <mailman.1017681605.21669.python-list at python.org>, Cliff Wells wrote:
> On Mon, 1 Apr 2002 04:09:12 +0200
> Gerhard H?ring wrote:
> 
>> * Dale Strickland-Clark <dale at riverhall.NOTHANKS.co.uk> [2002-03-31 12:31
> +0100]:
>> > "Charl P. Botha" <cpbotha at i_triple_e.org> wrote:
>> > 
>> > >With wxPython, one has to get, configure and build wxWindows first
> (and this
>> > >does take quite a while).  
>> > 
>> > Nonsense.
>> > 
>> > 1. Download
>> > 2. Run Setup
>> > 3. Use it.
>> > 
>> > The whole operation takes less than 5 minutes.
>> > 
>> > Well - it does on Windows, anyway.
>>  
>> Obviously you've never installed glib, gtk, wxGTK then wxPython all in
>> /usr/local on a Unix system. Not speaking all the dependencies if
> wxPython if
>> you really want to run all of the demos. This can easily take hours.
>> 
> 
> wxPython is remarkably easy to install on both Windows and RPM-based Linux
> systems.  I haven't tried it on anything else, but I doubt even a
> tarball-based install would be very difficult.
> 
> Building from CVS would be more difficult, but that's rarely necessary.

Try getting a running wxPython installation on SGI Irix or SUN Solaris for
instance.

Secondly, why would building from CVS be more difficult than building from
tarballs?  The tarballs also contain oodles of source that have to be
configured, built and installed.  If you haven't actually done this, you
should give it a try before actually commenting. :)

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From samschul at pacbell.net  Thu Apr  4 18:58:09 2002
From: samschul at pacbell.net (Samuel Schulenburg)
Date: 4 Apr 2002 15:58:09 -0800
Subject: Solution for   python22_d.lib
References: <f445a2fe.0204041111.66e3cced@posting.google.com> <m3n0wjdyzr.fsf@mira.informatik.hu-berlin.de>
Message-ID: <d7fcf8b4.0204041558.1ecf5f27@posting.google.com>

I have used the folloing technique to get around the need to have a
debug library for python. The following technique has been posted
previously by Tim Peters.



#ifdef _DEBUG
#undef _DEBUG           // Get the release library
#include <Python.h>
#include <graminit.h>   // used to embed python
#include <pythonrun.h>  // used to embed python
#define _DEBUG          // Switch back to debug mode for user code
#else
#include <Python.h>
#include <graminit.h>   // used to embed python
#include <pythonrun.h>  // used to embed python
#endif

Sam Schulenburg


From tim.one at comcast.net  Sun Apr 21 00:31:57 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 21 Apr 2002 00:31:57 -0400
Subject: How to make regexes faster? (Python v. OmniMark)
In-Reply-To: <kulmbimohe.fsf@lasipalatsi.fi>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEMMOPAA.tim.one@comcast.net>

[Erno Kuusela]
> it doesn't show up in many newsreaders, because your message lacks a
> "references" header. that's the usual mechanism that newsreaders use
> to do threading.

Sure.

> your message does have a "in-reply-to" header, which your newreader
> could be using. it's for mail and not news, so not standard.
>
> maybe mailman could be hacked to fake a "references" header based
> on "in-reply-to".

Barry won't do that. "Practicality beats purity" isn't one of Mailman's
principles, although I have evidence it's not unknown to Barry <wink>.





From gh_pythonlist at gmx.de  Tue Apr  2 12:40:18 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Tue, 2 Apr 2002 19:40:18 +0200
Subject: ANN: Twisted 0.16.0
In-Reply-To: <Pine.LNX.4.44.0204021043340.31094-100000@mercury.sabren.com>
References: <3CA83409.8080401@itamarst.org> <Pine.LNX.4.44.0204021043340.31094-100000@mercury.sabren.com>
Message-ID: <20020402174017.GA7571@lilith.hqd-internal>

* Michal Wallace <sabren at manifestation.com> [2002-04-02 10:55 -0500]:
> On Mon, 1 Apr 2002, Itamar Shtull-Trauring wrote:
>  
> > For more information on how Twisted helps you run licensed, unpirated
> > Open-Source software, visit:
> > http://twistedmatrix.com/documents/howto/register
> 
> Okay, I see this was posted yesterday (April Fools day), but
> the site's still up today. 
> 
> Is this for real?

Yes, unfortunately.

> If so, why? How can piracy be a problem for an open source project?

Twisted is no longer open source. They've changed the license to a
Shared Source one. I thought that was mentioned prominently on their
site?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 13.3 ?C      Wind: 3.5 m/s



From trentm at ActiveState.com  Wed Apr  3 17:44:32 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 3 Apr 2002 14:44:32 -0800
Subject: File opening
In-Reply-To: <a8frho$qqu$1@netnews.upenn.edu>; from zayats@blue.seas.upenn.edu on Wed, Apr 03, 2002 at 09:17:44PM +0000
References: <a8frho$qqu$1@netnews.upenn.edu>
Message-ID: <20020403144432.A15288@ActiveState.com>

[Salim Zayat wrote]
> Hey all.  I was wondering about opening files.  Now I know you need to 
> include the path to the file in the open statement.  Now the problems 
> arises here, though.  I am writing a module that reads in a config file 
> and creates buttons from it.  So if my module is run directly, the only 
> thing you need is 
> 
> > open('./buttons.config', 'r')
> 
> But if you wanna import the module, then the path to the open statement 
> needs to be full path to the file.  Is there a fix to this?

Structure your app to place the config files in a more appropriate place,
where "appropriate" depends on the platform. On *nix this should be:
    ~/.<appname>/buttons.config
for user-specific config info and/or:
    /etc/<appname>/buttons.config
for site-specific config info.

On Windows, the "appropriate" physical location depends the flavour of
Windows and the currently set locale, but can be determined with win32api
calls (requires the Win32 Python extensions). So, for example, I have a
little app called "go" which has a user-specific configuration file
"shortcuts.xml". My app has this little function to find that config file:


def getShortcutsFile():
    """Return the path to the shortcuts file."""
    if sys.platform.startswith("win"):
        base = os.path.join(
            shell.SHGetFolderPath(0, shellcon.CSIDL_APPDATA, 0, 0),
            "TrentMick", "Go")
    else:
        base = os.path.expanduser("~/.go")
    return os.path.join(base, "shortcuts.xml")


Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From philh at comuno.freeserve.co.uk  Sun Apr 14 14:30:42 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sun, 14 Apr 2002 19:30:42 +0100
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018757487.4937.python-list@python.org> <Xns91EFEE502A532JamesLLugojcom@216.148.53.85>
Message-ID: <slrnabjimi.741.philh@comuno.freeserve.co.uk>

On 14 Apr 2002 06:26:16 GMT, James Logajan <JamesL at Lugoj.com> wrote:
>Tim Peters <tim.one at comcast.net> wrote:
>> A crushing amount of debate over "stability" has gone by on Python-Dev
>> the last week, and the one thing that strikes me over and over is that,
>> with just a few minor exceptions, nobody volunteers to *do* anything
>> except tell other people what to do.  What are you willing to do to
>> make your desires a reality?  If it doesn't involve contributing work,
>> time or money, you just want a free ride.  But nobody owes anyone a
>> free ride, and open source can't change that fact of life either.
>
>Not sure, but I think you are close to describing what has been called the 
>tragedy of the commons.
>
>Back when I was paid to do Python work, I contributed all of US$50 (I 
>think; maybe it was $20) to the PSA (I think it was called) and tried to 
>get the company I was consulting for to kick in a couple thousand (or a 
>couple hundred at least). They never got around to that, alas.
>
>Unfortunately I'm not sure what would be gained by contributions to a 
>strategy that I have specifically stated I don't agree with. Further, it is 
>unclear how I can "contribute" to the slower pacing of releases that I'd 
>like to see.

One way to do that would be to fork the program and have a new
version of Python in which it would be guaranteed that changes would
not stamp on existing projects. This "Stable Python" would of course
gradually become incompatible with the existing "Unstable Python",
leading to duplication of effort and overall confusion, which is why 
I hope it doesn't happen.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From mixo at beth.uniforum.org.za  Tue Apr 16 07:59:42 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Tue, 16 Apr 2002 13:59:42 +0200
Subject: output formating
Message-ID: <3CBC122E.5090306@beth.uniforum.org.za>

I am looking for a way to do something like ('cal' command )

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

        January               February                 March
Su Mo Tu We Th Fr Sa   Su Mo Tu We Th Fr Sa   Su Mo Tu We Th Fr Sa
        1  2  3  4  5                   1  2                   1  2
  6  7  8  9 10 11 12    3  4  5  6  7  8  9    3  4  5  6  7  8  9
13 14 15 16 17 18 19   10 11 12 13 14 15 16   10 11 12 13 14 15 16
20 21 22 23 24 25 26   17 18 19 20 21 22 23   17 18 19 20 21 22 23
27 28 29 30 31         24 25 26 27 28         24 25 26 27 28 29 30
                                               31
.
.
.
.


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Put another way, how can I split ouput into columns, and rows?



From donn at u.washington.edu  Wed Apr 17 11:41:49 2002
From: donn at u.washington.edu (Donn Cave)
Date: 17 Apr 2002 15:41:49 GMT
Subject: File locking using fcntl.flock()
References: <mailman.1018927830.29158.python-list@python.org> <a9hnje$a2k$1@nntp6.u.washington.edu> <mailman.1019005652.25281.python-list@python.org>
Message-ID: <a9k53t$34ue$1@nntp6.u.washington.edu>

Quoth Richard Jones <rjones at ekit-inc.com>:
| On Wed, 17 Apr 2002 03:38, Donn Cave wrote:
...
| > I don't have Solaris, but I bet it doesn't have flock.
|
| It does, under the heading of "BSD compatibility".

For python, that would make 2 questions:  1) was python built with
the BSD compatibility, and 2) is flock() a bona fide implementation
or does it call fcntl.  The answer to one or both is apparently "no",
or you'd get flock semantics.

| That's almost a moot point though, because I also mentioned that I tried both 
| fcntl.lockf() ("This is essentially a wrapper around the fcntl() locking 
| calls.") and posixfile.lock() and neither of those worked either. The _real_ 
| problem here is that Solaris appears to allow a process to have two locks on 
| a single file in the same process. Things are fine if the lock is requested 
| from separate processes.

I think I explained this.  It isn't just Solaris.  Those functions will
have the same semantics on NetBSD etc.

| Thus in separate processes fcntl.lockf() and friends work fine. This is 
| troublesome because threads can't use file locks. Just gotta avoid threading 

That's a good idea in general, anyway.

	Donn Cave, donn at u.washington.edu


From martin at v.loewis.de  Mon Apr 15 02:31:59 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:31:59 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com>
Message-ID: <m34ridmq00.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> The other thing I was trying to say, is that if this _BUG_
> (gee- that wasn't so hard!) in Python is fixed, then we need
> a way to consistently do all of this bit shifting.  

Please refer to PEP 237: Python 2.2 has completed phase A, in which
the behaviour that David observed is not a bug; this will change in
phase B.1.

> BUT, since we NOW don't know where the high bit is (but the class
> that is managing the long object DOES), we need a '>>>' operator to
> consistently handle a logical right shift-- (i.e.: there is no way
> to code this behavior in Python.)

That's what I'm trying to tell you all the time: the >>> operator is
meaningless if it is defined as "fill in zeroes". How does it know
where to start inserting zeroes?

Regards,
Martin


From johnroth at ameritech.net  Wed Apr  3 06:45:39 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:45:39 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com> <uakogq1r0i0524@news.supernews.com> <7xpu1h4h1m.fsf@ruckus.brouhaha.com>
Message-ID: <ualqgql534bjab@news.supernews.com>

"Paul Rubin" <phr-n2002a at nightsong.com> wrote in message
news:7xpu1h4h1m.fsf at ruckus.brouhaha.com...
> "John Roth" <johnroth at ameritech.net> writes:
> > I absolutely agree. For me, the entire point is providing
> > a **standard** way of expressing **in the code** the
> > notion that a particular variable only takes two values:
> > True and False. This is such a common case that it deserves
> > to be promoted to be a part of the language.
>
> Why not ask for enumeration types then?

Enumeration types should be part of the set PEP
(I forget the number.) They go together well, in
that an enumeration is simply a set, and some languages
that support both concepts unify them.

However, supporting True and False as an
enumeration type is the same kind of hack as
using 0 and 1.

John Roth




From ngps at vista.netmemetic.com  Thu Apr 25 10:07:41 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 25 Apr 2002 14:07:41 GMT
Subject: sql but... without server
References: <7fvfcu0vkc610fma8545v0nhdm7c6hem7q@4ax.com>
Message-ID: <aa92jd$df$1@clematis.singnet.com.sg>

According to Johann  <programisci at murator.com.pl>:
> I  would like to use sql database with python but I do not want to
> install any sql server. 

Ask Google about Metakit and Metakit/SQL.

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



From liste_nospam_ at opikanoba.org  Wed Apr 17 03:24:11 2002
From: liste_nospam_ at opikanoba.org (=?windows-1252?Q?Fr=E9d=E9ric=20Laurent?=)
Date: Wed, 17 Apr 2002 07:24:11 GMT
Subject: amkCrypto installation on win32
Message-ID: <1103_1019028251@news.free.fr>

Hello,

I want to use RSA encoding/decoding, so I downloaded amkCrypto 0.1.3.
I use python 2.2 on windows XP.

Must I have to download and install swig in order to make amkCrypto work
(for the mxCrypto module) ?
I don't have any C++ compiler installed, and it seems to be quite complicated :
amkCrypto -> mxCrypto -> swig -> c++ and so on...
moreover the link http://starship.skyport.net/crew/amk/python/crypto.html, given
in the documentation, on the pycrypt package is broken...

So, I just want to generate a pblic/private key, crypt and decrypt a text
buffer... but it sounds to be not so simple to do this !

Did I miss something ???

thanks in advance

--
Fr?d?ric Laurent
http://www.opikanoba.org 




From mlh at vier.idi.ntnu.no  Thu Apr 11 22:27:41 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 02:27:41 +0000 (UTC)
Subject: Tuple comprehension
References: <mailman.1018541080.13442.python-list@python.org>
Message-ID: <slrnabchgs.c8l.mlh@vier.idi.ntnu.no>

In article <mailman.1018541080.13442.python-list at python.org>, Sean
'Shaleh' Perry wrote:
>> 
>> Ah...I think I understand your perspective a little better now. I never
>> think of a tuple as a list that's immutable, so I don't encounter very
>> often the same population problem. To me, lists become immutable when I
>> give them names in all caps, just like my variables become constants when
>> I name them that way.
>> 
>
>I do similar things, but using the language to help prevent mistakes is handy
>when I give people modules to use in their own code.
>

I think the only _real_ reason to have tuples at all is that they are
immutable (and therefore usable as dictionary keys). And a slight
efficiency advantage over lists, maybe...

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From TerryByrne1963 at yahoo.com  Mon Apr 29 10:18:24 2002
From: TerryByrne1963 at yahoo.com (Terry Byrne)
Date: 29 Apr 2002 07:18:24 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <aaevge$emr$1@peabody.colorado.edu>
Message-ID: <93d52e82.0204290618.349a0e5f@posting.google.com>

Fernando,

With pleasure. Here is some detail on the 'after', I hope not too
much! The before was simpler, just ran re.search on every line, had no
string.find test and no pass command...

A function does the logfile evaluation. The function generates an HTML
page describing the results and conditionally e-mails certain
managers, depending on the errors it finds. When calling the function,
one passes in the list of strings (the log file content, generated by
Python's file.readlines() method), a flag-string telling what type of
log file it is (so the function knows what error codes to look for),
and the filename of the log file.

def HTMLIfy(aryStrings, strType, strShortName):
  # Count important notifications
  #   ...
  # Code skipped for brevity
  #   ...
  for aLine in rfContent:
    #re expensive. apply conditionally
    if aType=='val':
      idx = -1
      idx = aLine.find('FrNfo:2401 ')
      if idx > -1:
        # The error number lets me know a record number follows in
parens
        # Generate HTML link to that record number.
        print '\n\tFound 2401'
        mo = re.search('FrNfo:2401 \(Record# ([0-9]+)\)', aLine,
re.IGNORECASE)
        # Next line shortened for brevity
        aLine=aLine[0:mo.start()-1]+'FrNFO:2401 (<a href="'+ ...
+mo.group(1)+'</a> '+aLine[mo.end()+1:]
        aLine += '<br>'
        rfHandle.write(aLine)
        aLine = None
        idx = -1
        pass
      idx = aLine.find('Error Validator:2420')
      if idx > -1:
        LINKS_VNOQUERYHITS += 1
        aLine += '<br>'
        rfHandle.write(aLine)
        aLine = None
        idx = -1
        pass
      idx = aLine.find('Error Validator:2412')
      if idx > -1:
        LINKS_VNOJUMPDEST += 1
        aLine += '<br>'
        rfHandle.write(aLine)
        aLine = None
        idx = -1
        pass

At first I was just running the re.search on each line. Time: ~ 5 hrs
for 60K line log file. Started testing with the string.find method to
make sure a certain error code appeared on the line before running the
"expensive" re.search, and that cut execution time from ~5 hrs to ~
1.5 hrs, good but "no cigar," as they say.

Then I realized that each line of content could be treated the same as
you'd treat a switch() structure in C, or a case structure in Pascal:
if one option is a "hit", then you needn't even check for all the
other options. So I started pass-ing to the next line of the content
whenever I got a "hit." That cut my time down to under a minute for a
60K-line log file. Like any good idea, it's truly simple!

Terry


From jimmy at retzlaff.com  Thu Apr 18 16:59:19 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Thu, 18 Apr 2002 13:59:19 -0700
Subject: PEP for an unrestricted __init__ ?!
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com>

> holger krekel [mailto:pyth at devel.trillke.net] writes:
> I'd would like to know if anybody thinks it's
> worthwile to start a formal PEP
...
> to bring full factory-pattern power to the __init__
> constructor of a class.

I believe that's exactly what __new__ is for in new-style classes. See
PEP 285 for an example of __new__ usage for exactly this purpose (in the
Specification section).

Jimmy




From lumber at jack.com  Tue Apr 30 00:59:48 2002
From: lumber at jack.com (Lumberjack)
Date: 30 Apr 2002 04:59:48 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Xns91FF62618790Flumberjack@216.148.53.85> <b03e80d.0204291719.1f36c42b@posting.google.com>
Message-ID: <Xns91FFDFB10888lumberjack@216.148.53.85>

sandskyfly at hotmail.com (Sandy Norton) wrote:
> I'm sure you know you can communicate 100% of your point with less
> than 1% of your sarcasm. I'd hate to see people like you dissuade
> future pep writers from articulating their ideas to the community.

Nothing will dissuade fools; certainly not sarcasm. But for the non-fools 
who inadvertently do foolish things, a sharp prick, metaphorically 
speaking, will train them to ponder more carefully any future ideas before 
presenting them publicly and wasting people's time. (I'd change my nom de 
guerre to Ivan Petrovich Pavlov, but I don't recall him being portrayed in 
any Monty Python skit.)

Besides, I'm a lumberjack and I'm okay.


From sholden at holdenweb.com  Tue Apr 16 15:38:14 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 15:38:14 -0400
Subject: Backing up mail in Outlook Express using Python
References: <mailman.1018975712.21736.python-list@python.org>
Message-ID: <D6%u8.352$hl2.107@atlpnn01.usenetserver.com>

"Noah" <noah at noah.org> wrote in message
news:mailman.1018975712.21736.python-list at python.org...
>
> I'd be interested to see that Delphi code.
> As far as I could tell the messages in OE
> are binary -- and not just simple binary
> headers around the message texts.
> Does your Delphi code use COM to get the
> messages out of Outlook?
>
> Yours,
> Noah
>
[ ... ]
>
> I did adapt some code from Delphi to read the mailbox files, and can let
you
> have that if it would help, but sounds to me like a file copy is all
you're
> looking for.
>
> regards
>  Steve
>
I have replied privately to Noah with a copy of the code, which can also be
found at

    http://www.holdenweb.com/Python/PDCode/dbxread.py

if anyone else is interested.

regards
 Steve






From huaiyu at gauss.almadan.ibm.com  Wed Apr 10 16:27:24 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 10 Apr 2002 20:27:24 +0000 (UTC)
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org> <Npor8.8281$q63.818@atlpnn01.usenetserver.com> <3CAE52CA.60F56E3D@python.org> <oAfs8.26107$Sw5.2561@atlpnn01.usenetserver.com> <3CB35101.AF4FA8F7@noaa.gov>
Message-ID: <slrnab9811.hm3.huaiyu@gauss.almadan.ibm.com>

Chris Barker <Chris.Barker at noaa.gov> wrote:
>I would love being able to do:
>
>if bool(anyobject):  #(or maybe istrue(), or something like that)
>	whatever
>
>What I don't like so much is:
>
>if anyobject:
>	whatever
>
>because unless anyobject is of a class I have defined myself, Python has
>decided ahead of time what an untrue (or nothing) value is for that
>object. 

If you do want Python's convention of untrue, then the above two do not make
much difference (if bool is implemented correctly).

If you want your own version of untrue, then you need

if my_predicate(anyobject):
    whatever

You still do not need

if bool(my_predicate(anyobject)):
    whatever

I don't see any situation that you need "if bool(some_expression)".

Huaiyu


From cameron at hooper.net  Thu Apr 25 21:32:04 2002
From: cameron at hooper.net (Cameron Hooper)
Date: Thu, 25 Apr 2002 21:32:04 -0400
Subject: Numeric question
Message-ID: <pan.2002.04.25.21.32.03.310021.2578@hooper.net>

Hi,

I hope someone can help me with this problem. It relates to the
Numeric extension. The code below will fail when calling the
'outerproduct()' procedure:

---------------------------------------------------------
# test.py

from Numeric import *

inputs = array( [[1., 1.],
                 [1., 0.],
                 [0., 1.],
                 [0.,0.]] )


binput = ones((4,1), Float)
inputs = concatenate( (inputs, binput), 1 )

x = inputs[0][:]
t = array([1.])

# this step fails
print outerproduct(x,t)

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

The error produced is:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/tmp/python-25603Vd", line 21, in ?
    print outerproduct(x,t)
  File "/usr/lib/python2.1/site-packages/Numeric/Numeric.py", line 359,
     in outerproduct return asarray(a).flat[:,NewAxis]*asarray(b).flat[NewAxis,:]
ValueError: flattened indexing only available for contiguous array

However, if I do not use the concatenate procedure to create the input
array then all works perfectly:

----------------------------------------------------------
# test2.py

from Numeric import *

inputs = array( [[1., 1.,1.],
                 [1., 0.,1.],
                 [0., 1.,1.],
                 [0.,0.,1.]] )


x = inputs[0][:]
t = array([1.])

print outerproduct(x,t)

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

The shape of inputs is identical under both programs: shape(inputs) =
(4,3). 

Can anyone see what's going wrong.

Thanks,

Cameron


From martin at v.loewis.de  Fri Apr 19 03:27:45 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Apr 2002 09:27:45 +0200
Subject: Compile error: Python 2.2.1 in Solaris 2.5.1, with GCC 2.95.3
References: <3CBF20BA.6FBEDF89@argo.es>
Message-ID: <m3wuv4172m.fsf@mira.informatik.hu-berlin.de>

Jesus Cea Avion <jcea at argo.es> writes:

> CAUTION: This patch is not intented to be included *as is* in future
> Python 2.2.2. That bugfix will be work for Python maintainers :-).
> Its only mission is to allow correct Python 2.2.1 compilation under
> my environment.

If you want your patches considered for inclusion into Python, please
post them to sf.net/projects/python.

> @@ -94,8 +94,8 @@
>  
>  # Symbols used for using shared libraries
>  SO=            @SO@
> -LDSHARED=      @LDSHARED@
> -BLDSHARED=     @BLDSHARED@
> +LDSHARED=      @LDSHARED@ 
> +BLDSHARED=     @BLDSHARED@ -Wl,--rpath -Wl,/usr/local/lib
>  DESTSHARED=    $(BINLIBDEST)/lib-dynload

That chunk is unacceptable, even if rewritten in a portable way. Why
is /usr/local/lib better than any other directory? Why not include,
say, /opt/gnome/lib, or /home/loewis/lib, by default, as well?

If you want an extra rpath, just set LD_RUN_PATH before building
Python. On our installation, we have tweaked gcc to always add
/usr/local/lib (and /usr/local/X11R6/lib) to every binary, by editing
the specs file.

> -       SunOS/5*) case $CC in
> -                 *gcc*)
> -                   if $CC -Xlinker -V 2>&1 | grep BFD >/dev/null
> -                   then
> +       SunOS/5*) 
>                         LINKFORSHARED="-Xlinker --export-dynamic"
> -                   fi;;
> -                 esac;;

You are using binutils 2.12, right? That problem has been fixed in the
Python CVS in a different way.

Regards,
Martin


From greg at cosc.canterbury.ac.nz  Mon Apr 15 01:27:09 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Mon, 15 Apr 2002 17:27:09 +1200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122>
Message-ID: <3CBA64AD.AE4C563D@cosc.canterbury.ac.nz>

Bengt Richter wrote:
> 
> On thinking about it, I think this is ugly, though probably
> convenient in the implementation. I.e., looking at the bits
> of the absolute value is not usually what you really want when
> you convert to hex, IMO.

The way I see it, if you're doing bit twiddling
using unified ints/longs, you should always be
working with *positive* integers, in which case
the hex representation will always be what
you want.

The only reason negative integers come into
the picture now is that setting bit 31 of a
non-long int has the side effect of making it
negative. Once ints and longs are unified,
this presumably will no longer happen.

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


From erno-news at erno.iki.fi  Thu Apr  4 07:58:02 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 04 Apr 2002 15:58:02 +0300
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com>
Message-ID: <kusn6blixx.fsf@lasipalatsi.fi>

In article <a8fba8$ln8$1 at panix2.panix.com>, roy at panix.com (Roy Smith)
writes:

| I've got a dictionary d, which may or may not have a value I'm looking
| for.  If it does, I want to do a few things.  If it doesn't, I want to
| do a few other things. 

this isn't suitable if you can have None values, but when
you don't, it can let you get away with writing key only once.

value = d.get(key)
if value is None:
    otherStuff()
else:
    stuff()

  -- erno


From perry at subdimension.com  Thu Apr 25 10:25:36 2002
From: perry at subdimension.com (Perry)
Date: 25 Apr 2002 07:25:36 -0700
Subject: ActiveX events in wxPython not working
Message-ID: <6ddcff55.0204250625.1992ce1b@posting.google.com>

Hi,

I'm using wxPython as an ActiveX (GUI) container. The OCX draws a
gantt
chart from supplied data. Everything works fine, appart events.

Here is the code:

...
from wxPython.lib.activexwrapper import MakeActiveXClass
import win32com.client.gencache
...
# makepy.py -o gchart.py
import gchart
# -- OR --
#try:
#    gchart = win32com.client.gencache.EnsureModule('{A4E79A23-C9E1-11CF-BDD7-02608C4302A9}',
0, 2, 8)
#except:
#    raise ImportError("Can't load OCX")

class MyEvents(wxWindow):
  def OnDiagramLClick(self, x, y, returnStatus):
      print "OnDiagramLClick"
  def KeyPress(self, KeyAscii):
      print "KeyPress"
  def OnSupplyTextEntryAsVariant(self, controlIndex, TextEntry,
returnStatus)
        print "OnSupplyTextEntryAsVariant"

class MyGanttPanel(wxPanel):
  def __init__(self, parent):
     wxPanel.__init__(self, parent, -1)

     sizer = wxBoxSizer(wxVERTICAL)
        
     GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
     gantt.VcGantt = GanttClass(self, -1, wxPoint(0,0),
style=wxMAXIMIZE)
     gantt.VcGantt.EnableSupplyTextEntryEvent = 1
        
     sizer.Add(gantt.VcGantt, 1, wxEXPAND)

     self.SetSizer(sizer)
     self.SetAutoLayout(true)


In the file generated by makepy.py I found the following lines,
which I guess explain how to write the functions.

# Handlers for the control
# If you create handlers, they should have the following prototypes:
# def OnHistogramRClick(self,
                        Histogram=defaultNamedNotOptArg,
                        x=defaultNamedNotOptArg,
y=defaultNamedNotOptArg,
                        returnStatus=defaultNamedNotOptArg):
# def OnTableWidth(self,
                   Table=defaultNamedNotOptArg, 
                   tableWidthRatio=defaultNamedNotOptArg, 
                   returnStatus=defaultNamedNotOptArg):


Can somebody pleas tell me what am I doing wrong.

TIA


From noah at noah.org  Sat Apr 13 03:27:52 2002
From: noah at noah.org (Noah)
Date: Sat, 13 Apr 2002 00:27:52 -0700
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
In-Reply-To: <a951vl$2a6s$1@nntp6.u.washington.edu>
Message-ID: <PAENINBOOODFDEPOKPKCCEAECLAA.noah@noah.org>

Ah... some things are  more clear now.
I did have something like this working before,
but it was running on Solaris. Now I'm working on OpenBSD.
What I have discovered is that apparently the ftp 
on OpenBSD writes the "Password:" message directly to the tty.
It also reads the password back direct from the tty device.
AFTER that it switches back to stdin and stdout.
FTP was never putting anything into the pipe
because it never got past the login. After this
the fix was simple. So now I have non-blocking 
pipes wrapped with a select and all is happy
(why do this? So I can control the timeout without
having to poll inside of a loop)

The quick fix was to make use of the fact that
the OpenBSD version allows you to invoke FTP with 
the password on the command line (which seems
oddly unsecure). Now it works perfectly.
Expect python here I come!

Long term, I would like something more portable,
so I guess I will have to start to look at that
pty stuff now. I'm trying to figure out the pty library,
but it lacks samples (except for test_pty.py)
and documentation on the library is slim. 
Hopefully pty will work under OpenBSD. I looked
at the source code and there is support for BSD
in there, but the comments say it was only tested
on SGI and Linux.

It would be nice to have a 100% Python expect.

Yours,
Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Donn Cave
Sent: Thursday, April 11, 2002 3:16 PM
To: python-list at python.org
Subject: RE: Nonblocking IO on popen pipe (isn't nonblocking)


Quoth "Noah" <noah at noah.org>:
| Hmm... This sounds suspicious... If FTP won't write its buffered
| output then no application could see it -- that includes the shell!

When you use ftp in a normal session, the shell doesn't see its
output.  You're interacting with the ftp process through the
tty driver.  This phenomenon occurs when output is a pipe, not
when it's a tty.

| Python shouldn't be any different in what it sees.
| If I start FTP from the shell, then how does FTP know 
| when to flush the buffer so that you can see it in a shell?

C stdio I/O, which ftp presumably uses, buffers its output differently
depending on the device.  If it sees that it's writing to a tty, then
it "line buffers" output.  Otherwise, it "block buffers".  That's why
the solution is a device that acts like a pipe but looks like a tty.
(Of course all this is assuming that ftp actually does use stdio.  It
doesn't have to, and I haven't checked.)

	Donn Cave, donn at u.washington.edu





From trentm at ActiveState.com  Mon Apr  1 20:43:52 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 1 Apr 2002 17:43:52 -0800
Subject: Anyone need a program written?
In-Reply-To: <002e01c1d74b$7f1842e0$445d4540@Dell2>; from brian@sweetapp.com on Fri, Mar 29, 2002 at 09:59:38AM -0800
References: <AE7F3B92-433D-11D6-ADF0-003065D61546@lanl.gov> <002e01c1d74b$7f1842e0$445d4540@Dell2>
Message-ID: <20020401174352.C5055@ActiveState.com>

[Brian Quinlan wrote]
> Craig wrote:
> > How about a Python substitute for make.
> 
> Trent Mick was working on something like this a billion years ago. Here
> is the sourceforge project url:
> http://sourceforge.net/projects/black
> 

And about a billion - 0.00001 years ago, it fizzled. Currently it lives only
in my dreams. You should check out SCons though:

    http://sourceforge.net/projects/scons



Trent

-- 
Trent Mick
TrentM at ActiveState.com



From tim.one at comcast.net  Sun Apr 28 21:49:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 28 Apr 2002 21:49:25 -0400
Subject: A simply newbie question about ndiff
In-Reply-To: <3cc4804f$0$15473$afc38c87@news.optusnet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELEPBAA.tim.one@comcast.net>

[Neville Franks]
> Well believe it or not but there are quite a few people using ED that
> have source files in excess of 70K [lines]. Not something I would
> advocate but ...
>
> On a more serious note I do have folks who compare large'ish data
> (ascii text) files. These are typically dumps of some sort. Apart from
> being large they can also get less than optimal results as they can
> have large numbers of repeated lines which throws my diff out.

I just checked in a new version of difflib.py's SequenceMatcher class that
does a dynamic analysis of which elements are so frequently repeated as to
constitute noise.  The innermost matching loop can skip those at first, and
that can yield an enormous speedup.  It looks much more effective than the
current IS_LINE_JUNK gimmick.  For example, I took a 100K-line C source
file, made a copy, changed a line in the middle, and it took less than 15
seconds to do the get_opcodes() bit, including the time to read the files
and break them into lines.  I still doubt it would be a heck of a lot faster
in C++, because this mostly exercises the speed of dicts, and Python's
string dicts are extremely efficient.  I can easily imagine it running much
slower in C++, though!

> I've organized a copy of ED for you to play with and would be
> interested to hear what you think of it's Diff capabilities. ED
> doesn't have Python support just yet, but it is planned. It does
> support some 34+ languages though. A Free Trial is available at
> www.getsoft.com for anyone else who may be interested.

Thanks!  I got around to grabbing a copy last night but haven't had time to
play with it yet.  I can testify that it installed without a hitch <wink>.





From nobody at nobody.com  Tue Apr 16 20:30:19 2002
From: nobody at nobody.com (John)
Date: Tue, 16 Apr 2002 19:30:19 -0500
Subject: list_a-list_b
Message-ID: <Sh3v8.41523$tg4.492644@vixen.cso.uiuc.edu>

Hi, I need a help in writing a function that does

list_a=[1,2,4], list_b=[2,4]
list_minus(list_a,list_b)
print list_a #will print out a=[1] after removing all elements ([2,4])
appearing in list_b

Thanks.





From phd at phd.pp.ru  Tue Apr  9 07:17:05 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 15:17:05 +0400
Subject: Linking against python library causes core dumps
In-Reply-To: <c9f472a0.0204090306.7fb516ea@posting.google.com>; from mparaz@yahoo.com on Tue, Apr 09, 2002 at 04:06:22AM -0700
References: <c9f472a0.0204090306.7fb516ea@posting.google.com>
Message-ID: <20020409151705.R8297@phd.pp.ru>

On Tue, Apr 09, 2002 at 04:06:22AM -0700, Miguel A Paraz wrote:
> I'm working on the Python interface to the FreeRADIUS AAA server
> <http://www.freeradius.org>.
> 
> When linked against the Python library libpython2.2.a, and my code, it
> fails like this:
> 
> #0  ptmalloc_init () at malloc.c:1713

   pthread malloc?

> #1  0x4026de64 in malloc_hook_ini (sz=21, caller=0x40270a34) at
> malloc.c:1856
> #2  0x40269f8d in __libc_malloc (bytes=21) at malloc.c:2798
> #3  0x40270a34 in __strdup (s=0x810ab28 "/usr/local/etc/raddb") at
> strdup.c:43
> #4  0x08058b23 in main (argc=1, argv=0xbffffa0c) at radiusd.c:550
> #5  0x40206507 in __libc_start_main (main=0x8058a9c <main>, argc=1, 
>     ubp_av=0xbffffa0c, init=0x8056d20 <_init>, fini=0x81098e0 <_fini>,
>     rtld_fini=0x4000dc14 <_dl_fini>, stack_end=0xbffff9fc)
>     at ../sysdeps/generic/libc-start.c:129
> 
> Where radiusd.c:550 is simply:
> 
>     radius_dir = strdup(RADIUS_DIR);
> 
> 
> At this point, no Python functions are even called yet.
> If I remove libpython from linking, it starts up fine.
> I'm stumped - I see no reason why simply linking a lib would mess up
> the malloc() system.

   Looks like a problem with threaded Python with non-threaded freeradius.
Or vice versa.

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




From tchur at optushome.com.au  Mon Apr 29 17:13:36 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Tue, 30 Apr 2002 07:13:36 +1000
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>
Message-ID: <3CCDB780.57B643AB@optushome.com.au>

PoulsenL at capecon.com wrote:
> 
> I have a Numeric array with missing elements from which I wish to calculate
> a histogram.  However, in order to do so I must first remove the 'None'
> values from the array.  I have a kludgy solution, but I was wondering if
> someone could point me in the right direction on what is the proper method
> of collapsing a masked array to just the non-masked values.

Loren,

Use the .compressed() method on the masked array. It is documented in
table 6 in the MA section of the NumPy documentation.

Tim C




From pyth at devel.trillke.net  Sat Apr 27 17:20:19 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 27 Apr 2002 23:20:19 +0200
Subject: local variable not found
In-Reply-To: <pan.2002.04.27.23.01.32.51309.30401@onesite.org>
References: <pan.2002.04.27.23.01.32.51309.30401@onesite.org>
Message-ID: <20020427212019.GD16217@devel.trillke>

On Sat, Apr 27, 2002 at 11:01:32PM +0200, Alain Tesio wrote:
> Hi,
> 
> I have a problem trying to use code compiled on the fly,
> python looks in the globals variables only.
> 
> Look at this example :
> 
> ===
> import codeop
> 
> def f():
> 	x=[1,2,3]
> 	eval(codeop.compile_command("y=x"))
> 	print y
> 
> f()

You have the same problem that i had (see the 10-minute
old thread about 'update locals()'.

Problem is that when python compiles your function
it doesn't look into the string, that's done at runtime.
Thats why 'print y' is assumed to refer to a global variable.

Now when you do 

> 	eval(codeop.compile_command("y=x"))

the assignment is made in your local scope.
This results in 

> NameError: global name 'y' is not defined

so much for the problem :-)

    holger




From guido at python.org  Sat Apr  6 21:07:11 2002
From: guido at python.org (Guido van Rossum)
Date: Sun, 07 Apr 2002 02:07:11 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <qjab$WAy$3r8EwH2@jessikat.fsnet.co.uk>
Message-ID: <3CAFAAA4.41EA7CF9@python.org>

Robin Becker wrote:
> 
> In article <mailman.1017873312.18135.python-list at python.org>, Guido van
> Rossum <guido at python.org> writes
> >Here's one final posting of PEP 285.  I've read (or at least skimmed)
> >all feedback that was on our news server by 9:30am EST on April 2nd,
> >plus anything mailed directly to me or to python-dev until now.
> 
> If this was a final post why are you now retiring to the dev list after
> another 5 days.
> 
> It seems pretty pointless to appeal to the masses if you wish only to
> have a resounding clap on the back and then when that doesn't happen to
> retreat to your friendly worshipers.
> 
> O Gods when Rome burns where are you?
> --
> Robin Becker

You just made my kill file.

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


From phil at river-bank.demon.co.uk  Mon Apr 29 06:13:22 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Mon, 29 Apr 2002 11:13:22 +0100
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net> <3CCCD547.6090405@SPAMnwinternet.com> <6x6z8.30735$8D3.846221@news1.tin.it>
Message-ID: <3CCD1CC2.B89602B4@river-bank.demon.co.uk>

Alex Martelli wrote:
> 
> <posted & mailed>
> 
> Jim wrote:
>         ...
> > to have sip and PyQt installed, but I assume you do, since PyKDE2 won't
> > build and install without them. Make sure, however, that they all match
> > in version (latest is sip3.1, PyQt3.1, PyKDE2-3.1 - the 3.1 is the *sip*
> > version *not* the KDE version).
> 
> Having KDE 2.2.2, Qt 3.0.3, PyQt 3.1, and Python 2.2.1, I'd like to take
> PyKDE for a spin, but can't figure out what to download from where, and/or
> how to configure it properly.  All my attempts so far (admittedly I haven't
> given it much focuses attention) end up with something like:
> 
> checking for Qt version >= 230 and < 233... configure: error: "incompatible
> Qt version"
> 
> and indeed I don't have such an old Qt (and can't/won't downgrade).  Is
> there a way to do what I'm looking for -- try some PyKDE on an installation
> with all the "latest stable releases" of the various other tools, as
> listed -- or do I just have to wait for a stable KDE 3 or whatever...?

First off, as far as I'm aware, KDE v2 requires Qt v2 irrespective of
PyQt/PyKDE versions.

Secondly, PyKDE2 is a step behind PyQt (the author, Jim Bublitz, has a
much bigger job than I have with PyQt) so you must have the older v3.0
installed. All the necessary code can be downloaded from
http://www.riverbankcomputing.co.uk/pykde/download.php.

Phil




From cpbotha at i_triple_e.org  Wed Apr 10 10:49:11 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Wed, 10 Apr 2002 14:49:11 GMT
Subject: Easter Eggs
References: <mailman.1018446981.22409.python-list@python.org>
Message-ID: <slrnab8kfs.n1u.cpbotha@crabtree.cpbotha.net>

In article <mailman.1018446981.22409.python-list at python.org>, 
Anthony Baxter wrote:
>>>> Kalle Svensson wrote
>> [Lulu of the Lotus-Eaters]
>> > 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
>> > 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
>> > magic command was.  Can someone remind me, please?
>> "import this", IIRC.
> 
> It's in 2.1.2, 2.1.3, 2.2, and 2.2.1.

Are you sure?

[dutidad:/home/cpbotha]% python
Python 2.1.2 (#1, Mar 16 2002, 00:56:55) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import this
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named this
>>>

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From logstx at bellatlantic.net  Sat Apr 13 14:51:58 2002
From: logstx at bellatlantic.net (logistix)
Date: Sat, 13 Apr 2002 18:51:58 GMT
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net> <3cb843d0$0$4068$9b622d9e@news.freenet.de> <82bscn4nvz.fsf@acropolis.localdomain>
Message-ID: <i7%t8.3632$bn2.1579@nwrddc01.gnilink.net>

>
> As far as I can see, which is admittedly not that far, interfaces
> don't really serve a useful purpose in Python. If you want
> polymorphism in statically typed OO languages like C++ and Java, you
> need to design your functions and methods to accept only objects
> descended from a known type. But in Python, the type need not
> (actually cannot) be specified at design time, so what purpose would
> an interface serve? Wouldn't it be better to just inherit
> _implementation_ when necessary, rather than to inherit pure
> interfaces?
>

Interfaces are important if you want to enforce a design-by-contract.
Granted, in Python you can query a class at runtime to see if it implements
a property or method, but this doesn't guarantee that the method does what
you want it to do semantically.  (Just because a function is named foo, it
doesn't mean it foos the way you want it to.)

Not to be a prick either, but if you're not using interfaces or ABC's, then
you're not taking advantage of the true goals of OO programming, you're just
warping the function names and data structures for your convienence.  The
whole point of OO is to deal through abstract interfaces that hide the
implementation details.




From pyth at devel.trillke.net  Mon Apr 29 18:11:22 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 00:11:22 +0200
Subject: Update locals()
In-Reply-To: <q7bz8.76797$vF6.2277789@news2.tin.it>
References: <20020427195725.GA16217@devel.trillke> <20020428205757.GH16217@devel.trillke> <02042900255602.25256@arthur> <mailman.1020042230.9019.python-list@python.org> <q7bz8.76797$vF6.2277789@news2.tin.it>
Message-ID: <20020429221122.GR16217@devel.trillke>

On Mon, Apr 29, 2002 at 12:36:38PM +0000, Alex Martelli wrote:
[#now posted, accidentally send via private mail previously]

> The ONLY specific functionality that exec-without-explicit-dicts
> (EWOED for short) "gives" you is exactly the ability to enable such
> uncontrolled namespace alteration.  And that is exactly what makes
> it an abomination.

In many cases i agree. You showed it to me. thank you.

But sometimes abominations are exactly what you want :-)
I hardly dare to give you the following example as you
probably are going to launch some 'cruise missiles' at me:

   exec bootstrap

This is one line in a *very* simple network server that receives code
(or strings) and executes them. It's called 'bootstrap' because
you can consider this network server to be totally naked (like
a computer before an Operation system starts up). No threads,
no security, no nothing, except a socket receiving a bootstrap.

The idea is that the naked system is 'protocol-less' and the
*client brings in the protocol*. The newly arrived code can even take
over the 'sock.accept' loop and impose security restrictions
(signed code etc.).  I can show you some pretty nifty
examples at EuroPython what can be done with it. It's ongoing
experimenting research, though.

Loosing the possibility to

(a) alter locals()/globals() AND

(b) loosing EWOED semantics

would make the *basic idiom* of this approach, allowing the client
side to introduce protocols and (completly) modify behaviour of
the server, more difficult. IMHO.

> I think I'll be at EuroPython, yes, although I don't understand what
> you can possibly mean by "on the screen" in this context

I don't know. Might as well be my fault. With 'screen' i meant
'computer screen' and that we would sit in front of it with
a keyboard, scrolling in source code, hacking on the keyboard
and such things :-)

regards and thanks for your patience, anyway,

    holger




From tim.one at comcast.net  Wed Apr 17 01:11:58 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 17 Apr 2002 01:11:58 -0400
Subject: Does Python need a semantics for and/or over generators ? was :
 'xor' operator?
In-Reply-To: <7934d084.0204151608.271b1737@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEOEOOAA.tim.one@comcast.net>

[Boris Borcic]
> Indeed, shouldn't two generators g1 and g2, when combined
> in an expression of the form e.g. "g1 and g2" or "g1 or g2",
> combine meaningfully to become a third generator ? Respectively
> one that returns a successive value of g2 for every value that
> g1 returns, until one of them returns IterationError,

The Icon language supports operators on generators, and that's a good place
to look for thought-out, battle-tested ideas. "Should" is too judgmental for
me, though <wink>.





From cimarron+google at taylors.org  Tue Apr  9 06:15:26 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 9 Apr 2002 03:15:26 -0700
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com>
Message-ID: <29e28c51.0204090215.38f92899@posting.google.com>

> I am wondering what book would be the best for a begginer? 

If you want the best overall value, I recommend three books: 
1. K&R 
2. "The Standard C Library" by P.J. Plauger, and 
3. "The Dictionary of Standard C" by Rex Jaeschke.  

Cim


From jb at cascade-sys.com  Fri Apr 26 02:04:27 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 23:04:27 -0700
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org> <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com> <3CC53718.A10FF2E6@cascade-sys.com>
Message-ID: <3CC8EDEB.3DDDE803@cascade-sys.com>

"James J. Besemer" wrote:

> Steve Holden wrote:
>
>> Well, on this particular group I feel you *are* in the minority.
>
> As I acknowledged up front.
>
>> Many c.l.py
>> readers are *not* "professional programmers", but chemists,
>> mathematicians,
>> biophysicists, and a host of others for whom programming is a means
>> to an
>> end. Many others are, like me, professional programmers who attempt
>> to
>> increase Python's popularity by assisting those less familiar with
>> programming concepts.
>
> Some of the better "professional programmers" that I have had the
> pleasure to work with or employ had formal training in chemistry,
> physics or math rather than "computer science".  Some of the worst
> programmers had CS degrees from big name schools and couldn't design
> their way out of a paper bag.  I think the scientists tend to come to
> the subject of programming with more maturity, discipline and focus
> than some with formal [sic.] training in CS.
>
> I think some people have a natural aptitude for programming and some
> people do not.  It has nothing to do with educational background.  The
> ones with the aptitude excel and the ones without tend to average or
> sub average work.  I also know that basic programming tasks may be
> completed by children, so the basic skills are not at all hard to
> master.
>
> I also have had experience coaching newbies in Python and other new
> languages and also I have spent time teaching programming to middle
> school kids.  At one point in my life, I worked supporting a bunch of
> EE researchers who were not programmers, so I think I actually relate
> to this community better than it may appear on the surface.
>
> I guess my main point, which I don't want to get lost is that we DON'T
> have to dumb down the language in order to help beginning users.
>
>> Furthermore, you express exactly the kind of attitude that gets
>> "professional programmers" a bad name. Let's remember our social
>> skills here
>> ;-). Remember, Python is *not* just for professional programmers, so
>> you
>> shouldn't write as though it were "your" language.
>
> I apologize for any offense I may have given.  I'm new here and I'm
> the type of person that tends towards direct talk.  I see in this
> social setting it distracts from my message rather than (as I am used
> to) lending clarity.  Please forgive me for coming on too strong.
>
> I certainly make no presumption that Python is "my" language.  If I
> were king, there is a very short list of relatively minor things I'd
> change immediately, adding a ternary operator being one of them.  And
> if a topic I care about comes up on the list (I did not start the
> thread) I feel I have every right to chime in with my opinion, humble
> or otherwise.
>
> More importantly, I don't think we have to make a choice to optimize
> only for "pros" or only for newbies.  Seems to me already Python is a
> large and in places a complicated language.  To it's credit, it's way
> easier to learn and use than Perl, which is the closest practical
> alternative around today.  There are and always will continue to be
> advanced topics that will take newbies time to absorb (possibly
> forever).  Given the overall complexity of the system, seems to me
> that little items such as a ternary operator are way down in the
> noise.
>
> Does anybody honestly think the addition of a ternary operator would
> be more confusing to newcomers than the business with lexical scope
> changing behavior over time?  Iterators?  Generators?  List
> comprehensions? Unifying types and classes?  Sockets, Select, Polling
> Objects, multi-threading and synchronization of concurrent
> programming?
>
> No.  Arguing that comparatively trivial extensions such as ternary
> choice will hurt beginners is pure rhetoric, simply superstitious
> behavior.  The conclusion is foregone and you use this specious
> argument to rationalize your prejudice.
>
> For the most part, beginners are focused on getting their own programs
> to work.  Like with Lexical scope, they can simply ignore the issue
> for the most part until they gain sufficient experience with the
> language (along with larger applications).  If they run into questions
> about constructs they do not understand, then they can look things up
> in the doc.  But seriously, is your typical newbie going to have more
> trouble understanding a simple inline choice operator or some code
> that relies on some subtle aspect of dynamic scoping?
>
> Seriously, would something like
>
>     x = ( if a>b : a else: b )    # trivial example, would be more
> complex in real life
>
> truly be all that much more confusing for your average newibe than
>
>     y = [ func( i ) for i in collection ]    # trivial example, would
> be more complex in real life
>
> Wouldn't the second, existing construct actually be MORE confusing?
>
> Ha!  I bet your mythical newbie would be LESS confused by the proposed
> ternary operator than by the recommended substitute:
>
>     x = ( a > b and [ a ] or [ b ])[ 0 ]
>
> It'll confuse almost everybody.
>
> We already added an "inline FOR", what could possibly be so bad about
> an "inline IF"?
>
> Finally, many the sophisticated concepts mentioned earlier I expect
> were dreamed up by people with specialized training in software.  Not
> to put anyone down but people with specialized training and experience
> in the field of programming and programming languages often are better
> able to articulate new features like this and figure out how best to
> fit them into the language.  Just a factual statement, no offense
> intended to anyone trained in different fields.
>
>> > So I am not swayed by arguments that :? confuses people.
>>
>> So, if one is confused, one should leave the job of programming to
>> the
>> professionals? That will help the programmer shortage.
>
> That's not at all what I said.
>
> Rather, (a) the average user is smart enough to NOT be confused, and
> (b) there likely is something to be learned from people with a lot of
> technical experience and expertise in the matter.
>
> You refer to all the scientists who use Python.  I'm sure they've all
> taken calculus, linear algebra and undoubtedly mastered lots more
> complicated subjects.  I find it impossible to believe that this
> august group could not in time come to master the syntax and semantics
> of such a simple, useful and not all that unfamiliar of a construct.
> If I was one of these mythical users, I'd be insulted that you presume
> I'm so stupid.
>
> I sense that this is a religious issue, so both parties will resort to
> elaborate arguments, well out of scope of the actual issue.  All
> bullshit aside, do you really and sincerely in your heart believe that
> this feature alone would "bloat" the language, actually harm the user
> community and materially detract from Python's growing "following?"
>
>> > But I think part of the problem
>> > with inline choice is that people who use it do not properly
>> > parenthesize and indent it to properly highlight what's going on.
>> > I always split the test and the two alternatives on separate
>> lines,
>> > with each of the latter two starting with the special operators.
>> >
>> >     def abs( n ):
>> >         return (( n > 0 )
>> >                       ? ( n  )
>> >                       : ( -n ))
>> >
>> I take it you're using some sort of "desirable" extension syntax
>> here?
>
> Yes.  Here, I'm attempting to employ C's :? operator within Python.
> I'm sorry, previous posts seemed to assume that people generally knew
> what it meant, so I apologize for relying on that evidently
> presumptuous presumption.
>
>> I completely fail to see how this is preferable to the instantly
>> comprehensible
>>
>>     def abs(n):
>>         if n > 0:
>>             return n
>>         else:
>>             return -n
>
> In this case you are right there is little advantage.  But I find
> cases frequently arise where the decision appears nested in a larger
> expression (other than the entire body of a function) in which return
> would not be a legal or meaningful alternative.  [For a change of
> pace, and a test of your intellectual honesty, you give an example and
> I'll translate it to existing Python.]
>
> In any language, you always can get by with additional declarations
> (functions and/or auxiliary variables).  To some, this improves the
> readability of the program.  To others, like myself, it obfuscates the
> meaning.  In any case, creating additional steps creates extra,
> unnecessary work for the computer and there are cases in inner loops
> where such differences in performance may be significant.  Since the
> extra program steps furthermore are unnecessary, it's doubly
> irritating to have to spell it out.  I.M.H.O.
>
> And to say there's already some other way to do it in Python is doubly
> disingenuous.  You can say that about just about half the features in
> the language -- Generators, Iterators and List Comprehensions.  Toss
> 'em out, no need.  Easily done another way.  Hell, by that rationale
> we can get rid of FOR and IF, as there's way to express everything
> they can do in terms of WHILE.
>
>> By the way, please note that there *is* a style guide for Python,
>> which you
>> will, find at
>>
>>     http://www.python.org/peps/pep-0008.html
>>
>> It unequivocally states that spaces should not be used after opening
>>
>> parentheses or before closing ones. So, while we're talking about
>> "proper"
>> usages, let's get them *all* right, shall we?
>
> Thanks for picking this tiniest of nits from my fur.
>
> I actually am familiar with the style guide and even how it's changed
> over time.  IIRC, it used to be 8 column tabs were god's inviolate
> creation (which I still tend to favor), with 4 ok if you did penance.
> Now 4 is standard and god's own are the also ran.  Who knows?
> Someday, spaces like I use them might become standard.  (Yeah, right
> ;o)
>
> For the record, while I whole heatedly agree with most of the things
> in the style guide, including many of Guido's personal prejudices, I
> do happen to disagree on a few points.  And, similarly, after over 25
> years of writing software for a living there is no point in trying to
> convince me there's a better way.  Besides, "A Foolish Consistency is
> the Hobgoblin of Little Minds" and "sometimes the style guide just
> doesn't apply."  Right?  Seems to me, little example snippets in an
> email is last place where the rules need to be strictly enforced
> (unless you're one of those Hobgoblins).
>
> Philosophically, I feel the clarity of written code is of paramount
> importance (aside from actually working as intended).  I adhere to
> Cornell CS Professor David Gries' advice to liberally use white space
> to improve the clarity of your code.  In one effective presentation he
> admonished his audience of professional programmers to "open up your
> code and and let it breathe."  He accompanied his advice with a very
> dramatic slight of hand with transparencies on the overhead projector
> where a compact snipped of sample code opens up with white space and
> indentation throughout.  It's always stuck with me and I have found
> that the extra spaces inside parentheses and square brackets do
> noticeably improve comprehension of existing code.  It's more work to
> put them there but to me it's worth it.  Quite simply it more clearly
> delineates the inside from the outside.  In ten years when all y'all's
> vision starts to fade, you may finally appreciate the wisdom of this
> advice.
>
> Things I particularly agree with in the style guide are (a) the
> importance of consistency varies depending on scope; (b) you generally
> need to apply some common sense; and (c) if you're working on someone
> else's code you should respect the style already present (unless
> you're willing to redo the whole thing properly).
>
>> > Thus although you can use and/or as a partial substitute, there
>> are
>> problems with the hack:
>>
>> Indeed. An experienced (professional?) programmer would, of course,
>> consult
>> the FAQ. See
>>
>>     http://www.python.org/cgi-bin/faqw.py?query=ternary?req=search
>>
>> for further discussion on this topic.
>
> The FAQ admits the very "flaw" and "ugly"ness (FAQ's term) with the
> alternatives.  I didn't need to look because I already was well
> acquainted with the issue.  Even the sample function they give as an
> alternative is not a complete equivalent.  The very fact that this
> topic is a Frequently Asked Question and one with such a lame answer I
> think emphasizes that it's a true WART on the language.
>
>> I agree with this. But don't try to suggest that there aren't
>> problems with
>> the ternary construct in C, most notably the omission of an "=" in
>> the test,
>> turning a test into an assignment.
>
> If you are referring to the confusion between "=" and "==" -- that
> problem is completely unrelated to the ternary construct and
> absolutely would not apply to Python extended only with a ternary
> operator.  Rather it is a problem with C's concept that assignment is
> an operator that may appear in any expression rather than a
> statement.  So long as assignment remains a statement, then there is
> no problem with omitted "=" that I am aware of.  Python could diagnose
>
>     a = if a = b : 0 else: 1        # syntax error!
>
> as an error just the same as
>
>     a = ( b = c )     # syntax error!
>
> is today.
>
> Then too, I'm crazy enough to count assignment operators as another
> one of Python's few serious omissions.  I wish to hell I could simply
> write:
>
>     while data = socket.recv( 1024 ):
>         process( data )
>
> The recent addition of file iterators helps some but that's only one
> case.  Are we going to add iterators to sockets?  If so, should it
> default to recvline() or some other operation?
>
> While it is true that assignment operators do cause some confusion, it
> nevertheless again is a simple concept, readily mastered by even
> novice programmers.  It's not brain surgery.  I've taught non
> programmers how to program in C and they tend to pick up on these
> several idioms right away.  Arrays, arrays vs. pointers and
> multi-dimensional arrays seemed to be the bigger language problem,
> followed quickly by tools and libraries particular to a particular lab
> environment.
>
> Furthermore, to the extent there is risk, there are express remedies.
> I worked for a while with a dialect of AWK.  AWK, like C, treated
> assignment as an operator.  To protect people from the '=' vs. '=='
> confusion, the AWK compiler adopted a convention that it would issue a
> warning for any top level assignments that appeared within an
> expression.  To suppress the warning, an additional level of
> parentheses was required.  This actually worked out quite well,
> stylistically, as the extra parentheses gave the reader a clear,
> visible hint that something unusual was going on.  It's a good
> convention which I retain even in C/C++, where it means nothing to the
> compiler but aids humans reading the code later on.
>
>> > (b) Since and/or lacks clarity it arguably is more likely to
>> > people than inline choice.  And whereas :? confuses beginners,
>> > and/or (since it's an idiom rather than a language construct)
>> > is more likely to confuse more advanced users.
>>
>> [I presume that should have been "...likely to confuse people ..."].
>
> Yes, thanks.
>
>> Frankly, I don't think a language that "confuses beginners" is
>> likely to
>> develop a huge following, since everyione is a beginner at the
>> start. I
>> think confusion of *any* kind is to be avoided, and Python does as
>> well as
>> most languages in this respect. I particularly dislike your
>> implication that
>> if anyone is to be confused it should be the beginners and not the
>> "professional programmers".
>
> Of course, that's not what I meant.  I'm saying that the and/or idiom
> likely will confuse more people (INCLUDING experts) than an explicit
> ternary operator that's a documented standard part of the language.
> Don't you agree?
>
>> > To me clarity of expression is the most important quality of a
>> well
>> > written program.
>
>> To me, clarity of expression is the most important quality of a well
>> written
>> program.
>
> I find it very encouraging that if you go high enough up the tree,
> there is some common ground; in fact we are absolutely agreed.
>
> Even if we disagree about the details and outcomes we both have the
> same objective in mind, which is a big head start.
>
>> Binary conditional evaluation ala ? is completely unnecessary. It's
>> eminently reasonable for programmers to feel a need for such an
>> addition to
>> Python, but once they have been reeducated the feelings will go
>> away.
>
> "Reeducated" raises for me the specter of the Khmer Rouge.  Violent
> revolution and anarchy.  Kill all the professionals and "reeducate"
> anybody who resists.  I'm sure that's not what you meant.
>
> Too, I am fairly confident this feeling will not go away.  The several
> gratuitous departures from C are a daily annoyance with each new
> Python module I write.
>
>> I disagree. The biggest issue is the bloat this will introduce into
>> the
>> language, and programs written in it, without the least advantage.
>> YAGNI.
>
> I think it's unreasonable to refer to this tiny, easily understood
> feature as 'bloat', especially given the scope of Python now and the
> myriad other features that already spilled over the dam.
>
>> Blerch. This is supposed to be easy to understand? In which
>> universe. I
>> submit it's completely unpythonic.
>
> Just an alternative.  I thought using the ":" to separate the question
> from the answers made a kind of sense.  But I'll have to defer to your
> expertise on what is or is not "Pythonic".
>
>> Well, there were problems with Algol 60's conditional expressions,
>
> Oh?  Something other than the dangling else ambiguity (which does not
> apply in this case)?
>
>> but I
>> agree that *in a free-form language* they are useful and fairly
>> natural.
>
> Wow, so you don't simply hate them altogether.
>
>> Python, of course, is not a free-form language, and so inline
>> conditionals
>> of any kind are a far less natural fit with the rest of the
>> language.
>
> Fascinating.  I'm not sure I'm getting this but it sounds like this is
> a fundamental point that makes a ternary operator good or bad.
>
> I learned yesterday that Guido says Python is an Algol derivative.
> Isn't that "free form"?
>
> Please define "free-form language", with emphasis on why Python is not
> one.  Then please explain how free-formed-ness has any bearing either
> way on allowing a proposed construct, which you otherwise admit may be
> "useful and fairly natural."
>
> I am truly eager to learn, master.
>
>> This isn't the first time this topic has been dragged out of
>> retirement, and
>> I don't suppose it will be the last.
>
> I expect it'll be a continual nuisance unless/until Python adopts the
> changes.  The majority of SW development done on Unix and on PCs is by
> C/C++ programmers and a large fraction of college grads come
> pre-programmed to know one or both languages.  To that huge body of
> prospective users, the several gross and unnecessary inconsistencies
> with C appear to be glaring errors.  I know a lot of people with that
> background who reject Python out of hand (in favor of, Bletch, Perl)
> simply because of those perceived inconsistencies.  Sad and unfair but
> true.
>
> I accept I'm probably pissing in the wind on all this but I think it's
> an important consideration when trying to attract a substantial
> "following".
>
>> It's probably obvious that your post
>> got up my nose -- I'm not normally so crotchety, so I guess I should
>>
>> apologise for the cranky nature of my reply.
>
> Well, you thought I was being uncivil when I did not intend to be and
> then your broke Miss Manners' first rule that rudeness (perceived or
> otherwise) does not justify rudeness in turn.  No problem.  I have no
> problem with criticisms about my ideas, or just about anything else
> short of personal insults.
>
> I have nothing but an abundance of goodwill for Python and it's users
> base.  That's why I'm here.  Any suggestions I may make are sincerely
> well intentioned, whether it may seem that way or not.
>
>> I shall not be contributing more to this thread.
>
> Ah, the old, chicken shit, take some pot shots and run away.....
>
> Worse: you wait until NOW to tell me!   You shoulda put this
> disclaimer at the top of your note -- woulda saved me a lotta time.
> ;o)
>
> But then too, in retrospect, I suppose you could do nothing but agree
> with everything I say in this reply, so I suppose there's really
> nothing more for you to add.   ;o)
>
> Seriously, I'm interested to hear your further views on the matter.
>
> Regards
>
> --jb
>
> --
> James J. Besemer  503-280-0838 voice
> http://cascade-sys.com  503-280-0375 fax
> mailto:jb at cascade-sys.com
>

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From theivo at abv.bg  Wed Apr 10 12:26:30 2002
From: theivo at abv.bg (I.C.)
Date: 10 Apr 2002 09:26:30 -0700
Subject: Transferring files using sockets
Message-ID: <6ac1ecad.0204100826.25217b0a@posting.google.com>

I am trying to transfer a file using the socket module, but I'm not
sure how to do this. Should I split the file into little(1KB) packages
and then transfer them, or is there a possibility to send the whole
file?

Thank you for the help !


From duncan at NOSPAMrcp.co.uk  Thu Apr 25 04:42:16 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 25 Apr 2002 08:42:16 +0000 (UTC)
Subject: weak references and threads
References: <3CC653B1.9060701@vip.fi> <Xns91FA17D7E4F8ARASXnewsDFE1@130.133.1.4>
Message-ID: <Xns91FAADABF8644duncanrcpcouk@127.0.0.1>

Philip Swartzleonard <starx at pacbell.net> wrote in
news:Xns91FA17D7E4F8ARASXnewsDFE1 at 130.133.1.4: 

> There is only one reason I know that you would want to use weak 
> references, and that is in the creation of advanced, customized data 
> structures. The point is to avoid something called 'circular
> references', which is probably best explained by an example. Let's
> take a realy basic class (this is right from my 2.2 IDLE prompt):
> 
I have used weak references to quite nice effect in some of my unit tests. 
The idea is that I create a weak value dictionary: that is a dictionary 
where the key is a normal dictionary key but the value is a weak reference. 
When the object referenced in the value disappears then the key is also 
automatically deleted from the dictionary.

In a few cases (especially when testing C extensions) I have wanted to be 
sure that the data structures all disappeared neatly when expected. The 
setup for the tests creates an empty dictionary, each object is added to 
the dictionary using a tuple (id, somestring) as the key, and the tearDown 
method for the tests checks that the dicionary is empty (and if not it 
prints out the keys of the items that still exist).

Actual access to the weak dictionary is wrapped up in a class to make it 
more obvious what is happening.

Another good use for weak references is for objects which can be shared 
(e.g. the flyweight pattern). Typically the state of such objects is 
determined by the arguments passed to the constructor. In this case you use 
a factory function (or define __new__ in a newstyle class) which checks a 
weak dictionary keyed on the constructor arguments to see whether a 
suitable object already exists. If so you return the object, if not you 
create a new one and add it to the dictionary before returning it.

I believe that the latest versions of wxPython do something similar. They 
have to wrap a Python object around each wxWindows object and previously 
you could end up with several different Python wrappers for the same 
wxWindows object. The latest versions use a weak dictionary to ensure that 
the same object is reused as long as it still exists.
 
-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From andreas at mtg.co.at  Wed Apr 10 17:05:35 2002
From: andreas at mtg.co.at (Andreas Kostyrka)
Date: Wed, 10 Apr 2002 23:05:35 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <Ohxq8.45137$pT1.1283197@news1.tin.it>
References: <mailman.1017466974.6751.python-list@python.org>
	<3CAA5777.66E7C2D5@python.org>
	<kupu1hple7.fsf@lasipalatsi.fi>
	<eppstein-CC3FF6.18410402042002@news.service.uci.edu>
	<Ohxq8.45137$pT1.1283197@news1.tin.it>
Message-ID: <20020410230535.674eef0a.andreas@mtg.co.at>

On Wed, 03 Apr 2002 06:27:58 GMT
Alex Martelli <aleax at aleax.it> wrote:

> of no use cases for it and is thinking about taking it away again (maybe
> some Smalltalkers will put together an impressive array of practical
> use cases... let's hope...!).  It works similarly:
> 
Ok, some very useful case, actually I've done this in pre-2.0 Python with some
kludges and some magic :)

class MagicDataBaseTable:
	...
	def newRow(klass,**params):
		...
	newRow=classmethod(newRow)
	...

class MyTable(MagicDataBaseTable):
	...

Now the newRow class method knows that it is being called upon MyTable and can be parametrized by
the table and column names of that table.

Andreas




From kybic at ieee.org  Wed Apr  3 06:03:20 2002
From: kybic at ieee.org (Jan Kybic)
Date: 03 Apr 2002 13:03:20 +0200
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at>
Message-ID: <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>

Hello,
        this is a post in favour of a faster Python:


> Or are there any hints that Numric is really way slower than Matlab?

In my field (signal and image processing) and on high-level tasks,
Python+Numeric runs at the same speed as the equivalent Matlab code.
I have one big application that I basically wrote twice for both
platforms and that's the conclusion I made.

On the other hand, all the calculation that you cannot express using
high-level primitives (using matrices, vectors,...) and has to be
performed with loops element by element, is painfully slow. Equally
slow in Python, as in IDL, Matlab,... It is so slow that it is
unusable, it is very difficult to develop an algorithm if it takes
days to process any reasonably sized input. My solution is to rewrite
the critical part of the code in C/C++, I did this for Matlab,
Python+Numeric, as well as IDL, but this is awkward - it forces you to 
work simultaneously with two languages and to deal besides the algorithm
itself with a lot of implementation and interface details.

In fact, it has become such a burden, that I am seriously considering
switching to another language. The only problem is that I still have
not found anything high-level, elegant, fast, and free at the same time.

It would be so much easier if I could just compile parts of the Python
code into the machine code. This would enable to use Python both for
the prototyping and for the production code.

One poster said that a true Python compiler is not feasible due to
Python's dynamic nature. My reponse to this is that I am quite willing 
to add (optional) typing information on variables to assist the
compiler to generate an efficient code. I have recently came across
the language Dylan which permits just that. I think this is a feasible
way of improving the Python performance.

I would love to hear your thoughts on the subject. 

Jan



-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/


From claird at starbase.neosoft.com  Sat Apr 27 13:07:57 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 27 Apr 2002 12:07:57 -0500
Subject: lint for python?
References: <roy-6B1ABC.08373127042002@news1.panix.com>
Message-ID: <9769F76E839E417F.CAC2C155CCCEF709.367C176DE375A23A@lp.airnews.net>

In article <roy-6B1ABC.08373127042002 at news1.panix.com>,
Roy Smith  <roy at panix.com> wrote:
			.
			.
			.
>So, my question is, is there such a thing as lint for python?  Something 
			.
			.
			.
<URL: http://www.unixreview.com/documents/s=2426/uni1018986621203/0204h.htm >
-- 

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


From huaiyu at gauss.almadan.ibm.com  Wed Apr 10 20:05:43 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 11 Apr 2002 00:05:43 +0000 (UTC)
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu> <3CB4B1D6.1030504@dittmar.net>
Message-ID: <slrnab9kqn.i8b.huaiyu@gauss.almadan.ibm.com>

Daniel Dittmar <daniel at dittmar.net> wrote:
>Using .remove () is similar to scanning the list, so it doesn't buy you anything in terms of performance.
>
>
>How about:
>
>for i in xrange (len (lst) - 1, -1, -1):
>     item = lst [i]
>     if item < 5:
>         print item
>         del lst [i]
>
>This makes it possible to delete items from the list as the access is 
>back to front.

I've seen this several times in the past, in the context of making the loop
counter correct.  But how does it fare in terms of performance?  Does the
rest of the list get copied with each del?   This sounds like an O(nm)
operation, where n is the list length and m is the number of del's.

Huaiyu


From johnroth at ameritech.net  Mon Apr  8 07:59:00 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 8 Apr 2002 07:59:00 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <eppstein-8A20F8.18322106042002@news.service.uci.edu> <ub0k49taktp243@news.supernews.com> <eppstein-CB6351.08534907042002@news.service.uci.edu>
Message-ID: <ub315bi2fkdube@news.supernews.com>

"David Eppstein" <eppstein at ics.uci.edu> wrote in message
news:eppstein-CB6351.08534907042002 at news.service.uci.edu...
> In article <ub0k49taktp243 at news.supernews.com>,
>  "John Roth" <johnroth at ameritech.net> wrote:
>
> > > By the way, this is conceptually very similar to ints (at least,
> > > nonnegative ints): ints answer a specific question: how many
things?
> > > They do not answer the question: which things are they.
> >
> > Not exactly. Consider the common case of 'take a number'
> > played at various burocracies. The manager can tell how many
> > there are by inspecting the next number on the roll, but each person
> > waiting has his own number, and they can be called out of order.
>
> I was talking about cardinal numbers, you're talking about ordinal
> numbers.
>
> Ordinal numbers in programming are often used for converting objects
> into positions to look up data in an array.  They're less often needed
> in Python because we have dictionaries instead.

Ints are neither cardinal nor ordinal. They just are. Cardinal and
ordinal are ways of analyzing the situation that may, or may not,
apply.

That said, I've come to the conclusion that Nicholas Wirth
had it right all along in Pascal, the way he handled enums and
sets. His style of enums can be used as indexes, and eliminate
the problem of relatively anonymous integers being used as
'magic numbers.'

Unfortunately, I don't see a PEP on enums. Maybe I should
dust off the rewrite of the set PEP I made that included them.

John Roth
>
> --
> David Eppstein       UC Irvine Dept. of Information & Computer Science
> eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/




From jwbaxter at spamcop.net  Mon Apr 15 23:47:28 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Mon, 15 Apr 2002 20:47:28 -0700
Subject: 2.1 vs. 2.2
References: <mailman.1018749868.11004.python-list@python.org>
Message-ID: <jwbaxter-D3B649.20472815042002@corp.supernews.com>

In article <mailman.1018749868.11004.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> "The Python interpreter is very stable"*30000

Perfectly happy here with Python 2.1 on Mac OS X (10.1.3, although 
Python was compiled on a much older version).

Rather pretty visual effect as the angled bands of light and dark passed 
by (in my 132x44 window).

I'm glad I didn't try it on my 56K frame connection to elsewhere, though.

  --John


From aahz at pythoncraft.com  Thu Apr 11 12:16:12 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Apr 2002 12:16:12 -0400
Subject: What has PEP 285 done to us?
References: <mailman.1018408041.12582.python-list@python.org> <slrnab7hn9.4tj.gerhard@lilith.my-fqdn.de> <a91mdd$ck$1@panix1.panix.com> <slrnab9jff.1kn.gerhard@lilith.my-fqdn.de>
Message-ID: <a94csc$fof$1@panix1.panix.com>

In article <slrnab9jff.1kn.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>Aahz wrote in comp.lang.python:
>> In article <slrnab7hn9.4tj.gerhard at lilith.my-fqdn.de>,
>> Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>>>
>>>Why would anybody care if you aren't man/woman enough to post with
>>>your real name?
>> 
>> Because many people, respectable people, *do* post to Usenet without
>> using their "real name".
>
>Yes, there are such people, but they are rare. For one of those, there
>come a hundred trolls, 1337 hax0rz and their like.

In my other home newsgroup, there are people named piranha, Piglet,
songbird, and Pyanfar.  It's a little more unusual in technical
newsgroups, but pseudonyms are a fine netnews tradition that I'll fight
to keep, particularly given how frequently I've had trouble even after
changing my name legally.  It wasn't so many years ago that someone tried
to keep me from voting in a Usenet CFV despite my years of activity on
the Net.

Judge by a person's writing, not zir name.  I mean, heck, I'd stack Kent
Paul Dolan against any of the modern trolls -- using a real name doesn't
mean diddlysquat.

>> What the heck is a real name, anyway?
>
>In western contries, firstname + lastname.

You're wrong about that.  My full legal name is "Aahz", and I'm in the
USA.  Of course, there's also the Gandhi quote.

>IMO using your real name is polite and whatever you say is not worth
>being said if you can't put your name under it.

You're entitled to your opinion, of course, but I'll post a disagreement
any time you mention it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I disrespectfully agree."  --SJM


From fperez528 at yahoo.com  Mon Apr 15 15:28:34 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 13:28:34 -0600
Subject: Assistance sought with mondo numerical app
References: <86hemc3h4y.fsf@haggis.aae.uiuc.edu>
Message-ID: <a9f9mq$lss$1@peabody.colorado.edu>

D. Michael McFarland wrote:

> Dear Fellow Pythonistas,
> 
>     A friend of mine is in the throes of finishing his
> Ph.D. dissertation, comparing experimental and numerical results for
> material damage caused by grinding.  For the numerical part, he's
> using a finite element library I wrote in Python.  This library made
> programming the tensor operations much easier than they might have
> been in, say, Fortran, but the end result takes rather longer to run
> than one could hope--about a day per problem case on a PII-450.  He
> has two machines of this class available, and I have two, and that
> might just be enough to let him make his deadline, but it occurred to
> me where to find more machines: this newsgroup.
> 
>     In short, we want your flops.

Have you looked at the possibility of rewriting some of the tight inner loops 
in C? These days using weave (http://scipy.org/) it's perfectly reasonable to 
get large speedups with very little work. I don't know your code so I don't 
know how realistic this is (and you guys seem to be under tight deadlines), 
but you can look at this pdf: 
http://www-hep.colorado.edu/~fperez/python_misc/python-c.pdf
for a simple example where the speedup is a factor of 25.

Just an idea. Sorry but the cycles on our machines are tied up 24/7 by our own 
numerical codes, so I can't help on that front.

Cheers,

f.


From phil at river-bank.demon.co.uk  Sun Apr 28 06:34:41 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Sun, 28 Apr 2002 11:34:41 +0100
Subject: PyQt 3.1 Compilation Problems
References: <3CCB01F5.1090607@its.caltech.edu>
Message-ID: <3CCBD041.F2A703CF@river-bank.demon.co.uk>

James Lamanna wrote:
> 
> I have QT 3.0.3 installed (using a Debian package).
> I've download and compiled libsip.
> However, when I try to compile PyQt-3.1 (for Qt 3.x), I get the
> following error on 'make install':
> 
> /bin/sh ../mkinstalldirs /usr/lib/python2.1/site-packages
>   /usr/bin/install -c -m 644 qt.py /usr/lib/python2.1/site-packages/qt.py
> make  install-data-hook
> make[3]: Entering directory `/home/james/PyQt-3.1/qt'
> (cd /tmp;
> PYTHONPATH=/usr/lib/python2.1/site-packages:/usr/lib/python2.1/site-packages
> /usr/bin/python -c "import qt")
> Traceback (most recent call last):
>    File "<string>", line 1, in ?
>    File "/usr/lib/python2.1/site-packages/qt.py", line 47, in ?
>      import libqtc
> ImportError: /usr/lib/python2.1/site-packages/libqtcmodule.so: undefined
> symbol: metaObject__C9QSGIStyle
> make[3]: *** [install-data-hook] Error 1
> make[3]: Leaving directory `/home/james/PyQt-3.1/qt'
> make[2]: *** [install-data-am] Error 2
> make[2]: Leaving directory `/home/james/PyQt-3.1/qt'
> make[1]: *** [install-am] Error 2
> make[1]: Leaving directory `/home/james/PyQt-3.1/qt'
> make: *** [install-recursive] Error 1
> 
> Any ideas how this would be resolved?

This is a similar bug in the Mandrake 8.1 Qt package. They removed the
the QSGIStyle and QCDEStyle classes for some reason, but didn't make the
corresponding change to qconfig.h.

Add the following to $QTDIR/include/qconfig.h and try again...

#define	QT_NO_STYLE_SGI
#define QT_NO_STYLE_CDE

Phil




From fperez528 at yahoo.com  Fri Apr 12 16:33:47 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 12 Apr 2002 14:33:47 -0600
Subject: emacs python mode completion
References: <mailman.1018619185.28393.python-list@python.org>
Message-ID: <a97gda$8n5$2@peabody.colorado.edu>

Mark McEahern wrote:

> [Stephen J. Turnbull]
>> Although it's not exactly what you what, something that works in all
>> modes is dabbrev (bound to M-/ in XEmacs by default).  This completes
>> words according to context.  So if I type M c M-/ here I get "McEahern."
> 

Yes, dabbrev is absolutely fantastic. I wish still that there were true 
namespace-based completion, but that would require some serious python-lisp 
hacking, as emacs would have to import your objects and parse your python 
code to know which attributes to complete.

Sounds like a nice project for a serious lisp hacker. Any takers :)

f.


From python at rcn.com  Mon Apr 29 03:42:03 2002
From: python at rcn.com (Raymond Hettinger)
Date: Mon, 29 Apr 2002 03:42:03 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it>
Message-ID: <aaitih$o2k$1@bob.news.rcn.net>

"Alex Martelli" <aleax at aleax.it>
> There are other obvious incompatibilities between math.pow and builtin
pow:
>
> >>> math.pow(2,3)
> 8.0
> >>> pow(2,3)
> 8

That's a problem whether or not this proposal is adopted.
Like 'open' in the os module, 'pow' is a high risk candidate from for being
clobbered by a 'from math import *'.

> But I double the suitable module is math, as this would
> introduce further incompatibilities one way or another.  Module operator
> might be most appropriate for both pow and divmod.

Good point.  I'll revise accordingly.

> I think a finer-grained mechanism might be preferable, and handled by
> site.py / siteconfigure.py, perhaps, much like today are handles such
> issues as sys.path normalization, and Unicode default-encoding.  The
> former may be a more successful example than the latter, so maybe the
> need might be best approached via an extension to the parsing and
> handling of .PTH files.
>
> Right now, .PTH files handle importing and path-setting needs of
> installed packages.  Having them handle such packages' builtins-needs
> as well seems a minor extension.  Should this slow start-up too much,
maybe
> some modest acceleration could be provided e.g. by sys.whatever helper
> functions.  No objections to a PYTHONBUILTINS equivalent of PYTHONPATH,
> in addition, for off-the-cuff needs.  Dunno 'bout the proposed flag: a
> binary on/off switch seems too coarse-grained to be useful.

Agreed.  We need some sort of undeprecate mechanism by the time the
functions disappear (around Py2.6 or so).  Site.py seems like a better
approach than I proposed.

> >     in perpetuity.  Like "from __future__" this mechanism can be used
for
> >     other deprecations, allowing the language to evolve without
> >     accumulating clutter.
>
> Exactly because of such "ongoing needs", the mechanism deserves careful
> design.

Yes!

> No doubt more builtin candidates for deprecation can be proposed.
> Take, for example, 'apply'.  What need does it serve, ever since the
> forms *args and **kwds started applying to actual arguments (i.e., at
> function-call time) as well as to formal arguments (i.e., at function
> definition time)?  I.e., since 2.0 (or 1.6, I don't recall).  Indeed
> that would have been my first candidate for built-in deprecation (also
> for alphabetical reasons:-).
>
> abs, buffer, chr, cmp, coerce, id, intern, oct, ord, are a few others
> that might be worth thinking about in terms of frequency of need --
> they're all needed some of the time, of course, but how often?  Often
> enough to be warranted as built-ins?  I surely use, e.g., map, more
> often than, e.g., cmp, coerce, or id.  Not to mention intern.  So why
> should map go away and the others remain?  This is not meant to be a
> rhetorical question, but to prompt active reflection/discussion/study.

I picked the first six candidates for re-assignment or elimination.
No doubt, there will be others.

Thank you for the insightful comments.  I'll revise the draft accordingly.


Raymond Hettinger




From claird at starbase.neosoft.com  Mon Apr 15 13:57:35 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 15 Apr 2002 12:57:35 -0500
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <3cb8a237@news.mhogaming.com>
Message-ID: <6869630C3D61331A.C957E867BC5C5F6E.6740371EBD87DEE7@lp.airnews.net>

In article <3cb8a237 at news.mhogaming.com>,
Ken Peek <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
			.
			.
			.
>TCL and JAVAScript are very large (when compared to Python),
>and would be too difficult to build an embedded version
>for-- at least Python can be cut down to fit without too
>much work.  (Of course, that depends on what services your
			.
			.
			.
?  Tcl's smaller'n Python, by most interpretations
I know.  Also, Tcl has a slightly better tradition
of supporting stripped down versions to fit in,
say, 100K, or 200K.

If someone truly wants to pursue this, I can pre-
sent more of the evidence.

If the differences in size between Python and Tcl,
whatever they are, are big enough to matter to you,
then you probably should be looking at Lua, REBOL,
or Forth, anyway.

I've completely lost track of what it takes to
implement JavaScript.  The interpreters used to be
huge--certainly larger than Python's.  I don't know
whether that remains true.
-- 

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


From maxm at mxm.dk  Fri Apr  5 09:07:57 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 16:07:57 +0200
Subject: file rewind?
References: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>
Message-ID: <3CADAFBD.7040204@mxm.dk>

joel8bit wrote:

> when I run the module from the comman line by itself, it counts the lines
> alright, but does not count the characters.  do I need to "rewind" the file
> reference after the first function call to countLines()?

I think you need the seek() method.

regards Max M



From kseehof at neuralintegrator.com  Fri Apr 12 07:51:24 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Fri, 12 Apr 2002 04:51:24 -0700
Subject: Looking for doh!.. simple sequence type
In-Reply-To: <3CB6BED8.2000502@mxm.dk>
Message-ID: <EFEFIHAINIINNELLHFAIOEOCCAAA.kseehof@neuralintegrator.com>

Max wrote:
> 
> Hmm ... I am trying to create a simple sequence type that I cant iterate 
> over like (Python 2.1)::
> 
> lst = Lst()
> for i in lst:
>      print i
> 
> And I thought I could do it like:
> 
> class Lst:
> 
>      def __len__(self):
>          return 4
> 
>      def __getitem__(self, idx):
>          return 42
> 
> But I have to write it like below to avoid an endless loop..
> 
> lst = Lst()
> for i in range(len(lst)):
>      print i,
> 
>  >>>42 42 42 42
> 
> I am shure it's so simple tha I will go Doh! .. But perhaps I haven't 
> had enough cofee yet. 'Cause I cannot seem to figure it out.
> 
> regards Max M
> 
The problem is that __getitem__ doesn't check length.  Instead, it
terminates on an IndexError exception.

class Lst:
    def __getitem__(self, idx):
        if idx>=4:
            raise IndexError
        else:
            return 42
    
lst = Lst()
for i in lst:
    print i
    
42
42
42
42




From robin at jessikat.fsnet.co.uk  Tue Apr 23 13:37:03 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 23 Apr 2002 18:37:03 +0100
Subject: Python call tree tool?
References: <O8RiIIA1cYx8EwCY@jessikat.demon.co.uk>
 <lkd6wqpe6m.fsf@pc150.maths.bris.ac.uk>
Message-ID: <7dD$gUA$uZx8EwWr@jessikat.demon.co.uk>

In article <lkd6wqpe6m.fsf at pc150.maths.bris.ac.uk>, Michael Hudson
<mwh at python.net> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> I searched in vain for a tool that can create a call tree or  inverted
>> call tree for python modules/packages. It seems inconceivable that the
>> Python community hasn't created such a tool so I must be looking in the
>> wrong places. 
>> 
>> Any one know where such a beast is?
>
>There's Tools/script/trace.py which might or might not do what you
>want.  Apparently it's in a bit of a sorry state, but I don't really
>know how/why.
>
>Cheers,
>M.
>
trace.py seems to trace the execution of code. What I want is the call
graph/inverted call graph a la CFlow so that I can determine which
functions/classes are not called or referred to in some set of modules.

This seems harder in Python as we can dynamically create references
using eval/exec etc, but it would help to narrow the search for dead
code if even a static analysis could be done. My present effort uses a
file system walk and an re match for the names declared in a fixed
module. That gives me stuff referenced in comments etc and is a bit
limited.

I guess I can use inspect to figure out the functions/classes that are
declared in an arbitrary module, but it seems harder to determine if
these things are actually referenced in other modules. For that I assume
I will have to do some code analysis. 
-- 
Robin Becker


From jb at cascade-sys.com  Tue Apr 30 22:13:58 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 19:13:58 -0700
Subject: Newbie: finding the key/index of the min/max element
References: <200204302300.g3UN0kkG002462@ratthing-b246.strakt.com>
Message-ID: <3CCF4F66.BA0A5DD4@cascade-sys.com>

Laura Creighton wrote:

> JB wrote:
> > The term derrives from the Sandinistas in Nicaragua, the Communist rebels who
> > fought against the US-installed conservative Somoza dictatorship (and others).
> >
> > The suffix is sometimes used to refere to any contrarian group, and in cases
> > simply any group.
>
> That is simple ignorant American narrow-minded foolishness talking.
> Normally, I would take extreme great care to not say anything so
> offensive, but I have a deep suspcion that James J Besemer is one
> of the 'can't take the heat' folks, and will learn to like a kinder
> and gentler place than the one he proposes to make for Raymond Hettinger
> if he finds out what it is like to live in the place he proposed.

I'm truly sorry that my sincere attempt to answer what I thought was a simple
question caused you to resort to open hostility.  I am truly impressed by your
vicisousness.

Once again by innocently describing a true relationship I have observed between two
items (words in this case) puts be at loggerheads with someone who knows a little
more than me on the subject.

> At any rate, idiot fool deluded with the idea that the American perspective
> is somehow comprehensive,

I did not presume to present an "American perspective" nor did I presume to be
comprehensive.  Nor did I suggest or imply that either was the case.

Without suggesting it's a good thing, I humbly submit that for many Americans they
do recognize the 'ista" mainly from the Sandinista context.  They don't do this
because of any political association or assumption of American pride but merely
because it's the first place or one of the few places most of them heard the term.

> 'ista' by those of us who speak Spanish means one hell of a lot
> broader than that.  And not the same broadness in every Spanish
> speaking nation.  My Spanish comes from where I lived in childhood,
> San Pedro Sula, Honduras, and thus is fairly close to that of
> Nicauragua (the country next door).  <except I lived in Cuenca Equador
> also, complicated stuff my childhood>.  What 'ista' means varies from
> sub-region to sub-region, and I get into confusion using it as I am
> used to when I speak with Spaniards, where the rules are different --
> but -- at any rate, in Honduras, it is used for everything from
> 'following the teaching of' to 'liking'.  You get 'istas' for every
> damn fad on the planet, every popular song, let alone songwriter, and
> just preferring the colour blue.

Thank you for adding something truly informative to the topic.

I submit that if you calm down and put your emotions aside there is not a lot of
what I said that is factually inconsistent with your reality.  The second half of
my explanation particularly seems to resonate with your definition.  In all
fairness, I think I was more incomplete than factually in error.  And based on your
description, this "ista" suffix seems to be a rather complex and subtle idiom that
you admit even confuses you sometimes.  I suspect most non-Spanish speaking people,
Americans or otherwise, would be ignorant of most of this nuance.

> All of this has been going on for vastly longer than Super-Power
> sponsored hell in small countries.

Nothing in my statement was intended to support or defend my country's behavior in
South America or anywhere else in the world.  From what I have learned over the
years, it's pretty much indefensible.  If you wanna talk politics, you may find we
have a lot more in common than you may imagine.

> The Super-Power sponsored hell was such a vileness, that nobody would
> be particularily interested in invoking that image.  Pythonista may
> resonate to ignorant Americans with only that sound, but the rest of
> us have a richer cultural experience in which to ground such an
> expression.  At least by God I hope so -- else we have finally
> reached the Nadir of Disneyism, where violent repression followed with
> anarchy, and the dying hopes and dreams of a generation are _glorified_
> because they are BRANDED.
>
> Yours most disgustedly,
> Laura Creighton

I did not intend to invoke any particular image, only to document a phonetic
relationship between words.  With you, I obviously struck a nerve.  Clearly in some
way you were deeply wounded by what all happend down there.  I certainly had no
intention of invoking a hell from your past.  I hope you will forgive me.

I also sense that you are pissed at me personally, which puzzles me just a little.
This is the second time you have said offensive things to me though I have never
directed any comments whatsoever at you.  Perhaps you have a chip on your shoulder
about all Americans or men or whatever.  Maybe I remind you of somebody you hated.
I don't know.  I like to think I have shown that I can roll with the punches and I
feel I've made a few new friends on this list (some not even Americans).  I
resolved this morning to try and be more civil, in large part due to your note
yesterday.  Not out of an inability to "take the heat" it but because you made a
good point which was hard to disagree with.  However, your continuing to attack me
like this is not a way to encourage me to try and be civil.  Your emotionalism left
you open for a large number of smart-ass remarks, which I respectfully and with
difficulty suppresed.

Finally, I am truly puzzled by your Raymond Hettinger reference.  Although I
strongly disagree with him, I thought my very sincere reply was respectful and
deferential rather than hostile, and I thought it was very much in line with much
of the nit-picking that goes on here.  I certainly don't wish him any ill will, and
I will cheerfully support any good proposals he puts forth.  But his justification
for his proposal strikes me as pretty empty, as I indicated.  How does one properly
register dissent in your world?  Am I just supposed to blithly agree with
everything that is said?  Seems that would certainly put me in a distinct
minority.  If I'm missing something I trust you will point it out.


Oop.  Gotta go fix dinner for the family.

Sincerely,

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From bernie at 3captus.com  Fri Apr  5 22:58:33 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Sat, 06 Apr 2002 03:58:33 GMT
Subject: ZPT : return a string HTML from a script python
References: <73f739e9.0204050117.36bde90b@posting.google.com>
Message-ID: <3CAE72A8.56F73821@3captus.com>

Marco wrote:
> 
> I have a ZPT and I want to print an HTML string (it's arrive from a
> script python) inside it.
> 
> This is a piece of ZPT code that call the script "pyPrint"
> 
> --   <span tal:content="python:here.pyPrint
> --        (folder=request['folder3'])">
> --   </span>
> 

Use 'structure' to render the return.

    <span tal:content="structure python:here.pyPrint
        (folder=request['folder3'])">
    </span>

You can find more example on
http://www.zope.org/Members/peterbe/DTML2ZPT.


> This is a script
> 
> --  message = "Result: <br>"
> --  for folder in folder:
> --      folder=container.pyString2list(string=folder.id)
> --      rootFolder = container.restrictedTraverse(['',]) # container
> --      folderDaCancellare = rootFolder
> --      folderPadre = None
> --      for item in folder[1:]:
> --          folderPadre = folderDaCancellare
> --          folderDaCancellare = folderDaCancellare[item]
> --      message = messaggio+folderDaCancellare.id
> --      if folderPadre != None and
> len(folderDaCancellare.objectValues()) == 0:
> --          folderPadre.manage_delObjects(folderDaCancellare.id)
> --          message = message + " cancellato<br>"
> --      else:
> --          messaggio = message + " non cancellato<br>"
> --
> --  return message
> 
> I want that the <br> tag will become a new line in ZPT page.
> 
> Sorry for my english

-- 
One OS to rule them all, one OS to find them,
One OS to bring them all and in the darkness bind them,
In the Land of Redmond, where the Shadows lie.


From pj at engr.sgi.com  Thu Apr 25 00:52:40 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 25 Apr 2002 04:52:40 GMT
Subject: stdin - seek() issue on Linux.
References: <mailman.1019617785.428.python-list@python.org> <1019625593.731162@yasure> <aa71nq$tukai$1@fido.engr.sgi.com> <1019703987.36802@yasure>
Message-ID: <aa822o$u2mgj$1@fido.engr.sgi.com>

Dann wrote:
|>  ... Whatever seek means in Python,
|> it inherits that meaning from C fseek ...

yes.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From greg at cosc.canterbury.ac.nz  Mon Apr 15 22:53:11 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Tue, 16 Apr 2002 14:53:11 +1200
Subject: Darwin + Tkinter + Aqua-Tk? (Re: Macintosh Development)
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz> <eppstein-A27785.23594014042002@news.service.uci.edu> <20020415144153325+0100@pc-62-30-160-65-hw.blueyonder.co.uk> <eppstein-7165E1.08012115042002@news.service.uci.edu>
Message-ID: <3CBB9217.8A54E272@cosc.canterbury.ac.nz>

David Eppstein wrote:
> 
> If you've installed a *different* version of Python (Unix Python, not
> MacPython), you can do this.  It doesn't come installed by default, and
> I have no idea whether it can do GUI stuff (unless you also install
> X-windows, or something).

In theory, it would seem that you could install Unix python
on Mac OS X together with the Aqua version of Tcl/Tk, and
run Tkinter stuff.

When I tried this recently, however, it didn't work. I
got a Tkinter window, but it was "dead" -- it didn't
respond to mouse clicks. Tcl applications run using
wish worked, but it wouldn't work from Python for
some reason.

I'd be interested if anyone happens to know how to
get this to work!

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


From cliechti at gmx.net  Wed Apr 17 18:37:02 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 18 Apr 2002 00:37:02 +0200
Subject: ActivePython 2.1.3 and ActivePython 2.2.1 released!
References: <mailman.1019071474.3193.python-list@python.org>
Message-ID: <Xns91F4680F3221cliechtigmxnet@62.2.16.82>

Trent Mick <TrentM at ActiveState.com> wrote in 
news:mailman.1019071474.3193.python-list at python.org:
> In addition, PyPPM (the Python Package Manager) enables you to easily
> install a number of other popular Python packages. For example, try
> this:
>     > pyppm install Numeric
> to install Numeric Python. Use this command to list all available
> packages:
>     > pyppm search

is that also available separately and can i set up my own packages and 
server?

chris 


-- 
Chris <cliechti at gmx.net>



From donn at u.washington.edu  Thu Apr 18 12:31:25 2002
From: donn at u.washington.edu (Donn Cave)
Date: 18 Apr 2002 16:31:25 GMT
Subject: UTC equivalent of time.mktime() ?
References: <NCBBLFCOHHDIKCAFGCFBMEJAKLAA.marklists@mceahern.com>  <mailman.1019062773.24881.python-list@python.org>   <a9kcnu$11v6$1@nntp6.u.washington.edu> <mailman.1019121337.10950.python-list@python.org>
Message-ID: <a9msct$aea$1@nntp6.u.washington.edu>

Quoth Graham Ashton <gashton at cmedltd.com>:
...
| I was using strptime() and mktime() to get seconds since the epoch in
| localtime. I was trying to work out if I could do anything simpler than
| this (in which two of the "-" signs should be "+" signs):
|
| >>> def local2utc(secs):
| ...     if time.daylight:
| ...         return secs - time.timezone - time.altzone
| ...     else:
| ...         return secs - time.timezone
|
| to get the seconds since the epoch into UTC. It's looking as though the
| answer (in general) is "no".

Yes, I mean, no.

	Donn Cave, donn at u.washington.edu


From mwh at python.net  Mon Apr  8 09:59:11 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 8 Apr 2002 13:59:11 GMT
Subject: New classes in 2.2
References: <un0wglgqf.fsf@online.no>
Message-ID: <u8z7yb83q.fsf@python.net>

Syver Enstad <syver-en+usenet at online.no> writes:

> I soon found out that super would (in addition to ordinary methods)
> only work for class data members and static functions, and not for
> class methods were they in my view would be most elegant.

Will do in 221.

Cheers,
M.

-- 
  ARTHUR:  Don't ask me how it works or I'll start to whimper.
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11


From babbouux at yahoo.fr  Tue Apr 23 13:24:48 2002
From: babbouux at yahoo.fr (barbara feledziak)
Date: Tue, 23 Apr 2002 17:24:48 GMT
Subject: Found a new interpreted language
Message-ID: <ANgx8.168342$K5.14142962@bin5.nnrp.aus1.giganews.com>

I am since a long time a python programmer

I find this product very interesting. I can execute interpreted code
directly on my PC microprocessor

http://www.hyperpanel.com/boot_direct.php3?ID=12740423







From kseehof at neuralintegrator.com  Tue Apr  9 10:52:56 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 09 Apr 2002 07:52:56 -0700
Subject: Embedding Python with C++
In-Reply-To: <000701c1dfb7$37955fb0$933d06d5@rocketship>
Message-ID: <EFEFIHAINIINNELLHFAIAEFFCAAA.kseehof@neuralintegrator.com>

>
> > merman wrote:
> > > Hi Ken,
> > >
> > > can I use your wizard with bcc55 too?
> > >
> > > o-o
> > >
> > > Thomas
>
> Ken wrote:
> > No, it's a plug-in for the Visual C++ IDE.
> >
> > However you should be able to compile the generated code
> > created by PyExtWizard with bcc.
> >
> > Much of the code generation functionality is independent
> > of Visual C++, so it would be a feasible project to make
> > PyExtWizard a multi-platform tool.  I will eventually
> > need to do this for my own needs.  The GUI for the tool
> > will be in wxPython.
> >
> > - Ken

merman wrote:
> ... thanx Ken - but I have VS7.NET (Beta2).
> There are too much failures if I compile - no chance.

Haven't tried it, but I doubt that VS7.NET supports VC6 wizards.
Both products are by Microsoft and are therefore incompatible :-)
If the old wizards don't work, it's useless to try to compile
PyExtWizard (see below).

> and with swig:
> Maybe is my Makefile not valid - yesterday I posted it to comp.lang.python
> (thread swig-problem from merman).
> I have try it with swig and bcc55 under win2k.

You won't get anywhere :-(  SWIG has nothing to do with anything here.

> Last error-messages from bcc55 are:
>
> Error: Unresolved external '_Py_InitModule4' referenced from
> C:\MYC\PY_EXT\TREE_
> WRAP.OBJ
> Error: Unresolved external '_PyModule_GetDict' referenced from
> C:\MYC\PY_EXT\TRE
> E_WRAP.OBJ
>
> ... and so on. Is there a linker-probem?
> I have no solution.
>
> o-o
>
> regards
>
> Thomas

I think you misunderstood me.  You won't be able to build
PyExtWizard with anything other than MS Visual C++ (MSVC++ 6.0).
It makes no sense to build it with anything else since it is a
plug-in wizard for MSVC++.  PyExtWizard is not a stand-alone
application.  As a plug-in, it enhances VC++ by allowing you to
create a new VC++ projects that implement python extension modules.

It makes no sense to use SWIG on PyExtWizard.  PyExtWizard is not
an API and it does not contain any C++ code that anyone would have
any reason to wrap.  The code generation functionality is implemented
in python and uses the capabilities of the MSVC++ wizard system.

When I said, "you should be able to compile the generated code
created by PyExtWizard with bcc", I meant that if you were to use
PyExtWizard with Visual C++ to generate a new python extension
module to your specifications, you would probably be able to compile
the resulting code with a different compiler (hopefully anywhere).

In other words, PyExtWizard is not portable at all, but it should
generate portable code (this hasn't been tested yet).  I guess you
could call it a "cross wizard" (Harry Potter on a bad hair day :-)

I alluded to my future intent to make PyExtWizard independent of
Visual C++, and therefore portable.  This would involve the following
steps:

1. Throw away -all- of the C++ code.  The C++ portion of PyExtWizard
serves only to interact with the MSVC++ IDE.

2. Replace the MSVC++ wizard functionality with python code.

3. Reimplement the GUI in wxPython.

- Ken Seehof





From eppstein at ics.uci.edu  Wed Apr  3 17:14:23 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Wed, 03 Apr 2002 14:14:23 -0800
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <Xns91E5CE0E5919Dcliechtigmxnet@62.2.16.82>
Message-ID: <eppstein-055127.14142303042002@news.service.uci.edu>

In article <Xns91E5CE0E5919Dcliechtigmxnet at 62.2.16.82>,
 Chris Liechti <cliechti at gmx.net> wrote:

> if d.has_key(key):
>      foo = stuff(d[key])
> else:
>      foo = otherStuff(key)

I've seen a couple responses suggesting this.

Is there a reason why in modern Python versions one should still use 
has_key?  I like the syntax "if key in d" better.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From thomas at gatsoft.no  Tue Apr 16 06:03:52 2002
From: thomas at gatsoft.no (Thomas Weholt)
Date: Tue, 16 Apr 2002 10:03:52 GMT
Subject: Backing up mail in Outlook Express using Python
Message-ID: <cGSu8.101$1kc.179320832@news.telia.no>

I need to backup all my mail stored in Outlook Express. Can I use Python to
do this?

Thomas




From bokr at oz.net  Wed Apr 10 17:51:11 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 21:51:11 GMT
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <3CB49ADD.89B242A4@sympatico.ca>
Message-ID: <a92c4f$ne7$0@216.39.172.122>

On Wed, 10 Apr 2002 16:04:45 -0400, "Colin J. Williams" <cjw at sympatico.ca> wrote:
[...html...]
____________________________________________________________________

Steve, 

Apologies for the HTML. 

Steve Holden wrote: 

  Sorry abt the format of this posting: OE foundit very difficult to remove 
  Colin's HTML... 

  "Colin J. Williams" <cjw at sympatico.ca> wrote in message 
  news:3CAF6EE6.7564CDF7 at sympatico.ca... 

  Guido asks seven questions: 
  [Colin provides seven answers] 

  > Perhaps we should also be rethinking the meaning of None. 

  Or the meaning of Life ... :-) 

  > Maybe it would be better to think of it as a NULL in relational 
  > algebra. 
  > Suppose we have a person record (p) with attributes age and sex. 
  > p.age= 22 and p.sex= F       (F asumed to be part of an enumeration) 
  > then we know something about this person. 
  > On the other hand, suppose we have p.age= None and p.sex= M 
  > We know that we have a male of unknown age. 
  > Suppose now that we ask whether this person's age is greater 
  > than 25: 
  >         if p.age > 25: ... 
  > Currently, the response is False, but really we don't know.

  > Thus, it might be better to think in terms of a three value 
  > system False, Unknown, True. 

  If you really wanted to go this way (and there are worse options), the 
  "logical" (no pun intended) value to use for Unknow id clearly None itself. 
  Relational systems mostly seem to return NULL as the result of operations on 
  NULL, whether they be comparisons or arithmetic operations. 
   

Yes, that would be one approach and is the value used in the examples below.  However, None is of NoneType. 
Perhaps it would be better to have a value of BoolType. 

    
  Note this would introduce three-valued logic into Python, which would be 
  helluva confusing to beginners, who tend to understand the law of the 
  excluded middle implicitly.

I wonder about the confusion.  It would perhaps add clarity to an aspect of 
Python. 

Let's look at some examples: 

a= -25 
b= None 
c= +25 
d= '0' 
print 'Simple comparisons' 
print '   a,   b,   c,   d:', `a`, `b`, `c`, `d` 
print 'a < b, a < c, a < d:', a < b, a < c, a < d 
print 'b < a, b < c, b < d:', b < a, b < c, b < d 
print 'c < a, c < b, c < d:', c < a, c < b, c < d 
print 'd < a, d < b, d < c:', d < a, d < b, d < c 

The result is: 
   a,   b,   c,   d: -25 None 25 '0' 
a < b, a < c, a < d: 0 1 1 
b < a, b < c, b < d: 1 1 1 
c < a, c < b, c < d: 0 0 1 
d < a, d < b, d < c: 0 0 0 

If the three value scheme were introduced, the results 
would be: 

a < b, a < c, a < d: None    1    1 
b < a, b < c, b < d: None None None 
c < a, c < b, c < d: 0    None    1 
d < a, d < b, d < c: 0    None    0 

Similarly, for logical operations: 

print 'Logical operations' 
print '0 or  None, 1 or  None:', `0 or  None`, `1 or  None` 
print '0 and None, 1 and None:', `0 and None`, `1 and None` 
print '              not None:', `not None` 
print 

The result is: 

Logical operations 
0 or  None, 1 or  None: None 1 
0 and None, 1 and None: 0 None 
              not None: 1 

With a three value scheme, the results would 
be the same for the first two, whereas the 
result of `not None` would itself be None. 

I would appreciate comments. 

Colin W. 

______________________________________________________________


HTML on usenet bugs me.

Regards,
Bengt Richter


From aleax at aleax.it  Thu Apr 11 07:39:55 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 11:39:55 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr>
Message-ID: <fCet8.7130$vF6.214745@news2.tin.it>

Ante Bagaric wrote:

> Is there a way to prevent the print statement to automatically add a space
> between two objects?
> 
>>>>print "a=",a
> a= 1
> 
> how to make that cursed blank character dissappear?
> Other than print "a=%d" % a because it wouldnt work for iteration of
> prints..
> 
> for i in range(5):
>     print i,
> 
> gives 0, 1, 2, 3, 4 and I want 0,1,2,3,4

Actually this gives 0 1 2 3 4 -- I don't know where you got the commas
in the output that you seem to have observed, maybe your print statement
was not quite as you show it here.


> Yes, I know it's possible to circumvent this behaviour, but I'm curiuous
> whether it can be done within the print statement alone.

The print statement sets the softspace attribute on sys.stdout when it
has just printed with a trailing comma, resets it when it emits a line
break.  If it finds the attribute set when about to continue emitting,
it outputs a space first.  The following superdirty trick works as
expected:

>>> import sys
>>> for i in range(5):
...   print setattr(sys.stdout,'softspace',0) or ("%s,"%i),
...
0,1,2,3,4,
>>>

because the expressions are evaluated BEFORE the print statement's
opcode occurs, thus before it checks sys.stdout.softspace.


Apart from showing off that you really know how print works, it
would of course be absurd to use such tricks in order to abuse yet
further the vastly-overused print.  A vastly superior way is
to use sys.stdout.write("%s,"%i) -- you just need some care to
show it off interactively because you need a '\n' at the end,
so, for example:

>>> if 1:
...   for i in range(5): sys.stdout.write("%s,"%i)
...   sys.stdout.write('\n')
...
0,1,2,3,4,
>>>

or, back to trick-land,

>>> for i in range(5):
...   sys.stdout.write("%s,%s"%(i,'\n'[i!=4:]))
...
0,1,2,3,4,
>>>

but you need no trick at all to use the write method in the
perfectly normal way within a script or module.


Alex



From aleax at aleax.it  Tue Apr 30 10:39:48 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 14:39:48 GMT
Subject: Global Variable
References: <aalpqc$h9s$1@pegasus.csx.cam.ac.uk>
Message-ID: <U0yz8.81624$vF6.2427699@news2.tin.it>

<posted & mailed>

Xiao-Qin Xia wrote:

> Hello,
> 
> If a module (for example, gmpy) is imported in the main.py, how can it be
> used by other modules (for example, mymodule) without being imported by
> each of the other modules?
> 
> #---------main.py---------
> import gmpy
> import mymodule
> ...
> 
> 
> #--------mymodule.py---------
> #I want use gmpy here, do I have to import it again? can main.py add gmpy
> into a global dict?

It is by far best to import gmpy into each module that wants to use it.

This is normal Pythonic use, has negligible runtime cost (since the FIRST
import does all the work and stashes the module object snugly into
sys.modules['gmpy'], and all other imports reuse this 'cached' object),
and enhances your program's clarity considerably.

That much being said, if you've made a bet about whether it could be
done -- it can.  Your "main module" would need to do:

import gmpy
import __builtin__
__builtin__.gmpy = gmpy

"et voila".

However, don't do it.  You'll be sorry if you start messing around with
__builtin__ for any but the most pressing of emergencies.


Alex



From bokr at oz.net  Thu Apr 11 21:38:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Apr 2002 01:38:58 GMT
Subject: File Size Mystery??
References: <3cb6358d.33229187@news>
Message-ID: <a95dri$mc1$0@216.39.172.122>

On Fri, 12 Apr 2002 01:18:56 GMT, brobbins333 at shaw.ca wrote:

>this is foofile.txt (created in Windows Notepad):
>ur75ht8495u7475ht89302oskv476830rj
 01234567890123456789012345678901234
           1         2         3
>
>and here is what Python tells me about it:
>
>
>>>> print os.path.getsize('foofile.txt')
>34
>>>> print len('foofile.txt')
                012345678901
                          1
>11
>
>
>os.path.getsize() returns the correct size of the file, 
>but len() does not. Why is there a discrepancy?
>
;-)

Regards,
Bengt Richter


From wilk at flibuste.net  Tue Apr 16 17:19:16 2002
From: wilk at flibuste.net (William Dode)
Date: Tue, 16 Apr 2002 23:19:16 +0200
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
	<3CBBD31D.AA2E653C@alcyone.com>
	<20020416115225.1d9fa2d2.wilk@flibuste.net>
	<3CBC0078.C10C5F4@alcyone.com>
	<20020416165356.3aeb88eb.wilk@flibuste.net>
	<3CBC6D12.6BE3A139@alcyone.com>
Message-ID: <20020416231916.239839f8.wilk@flibuste.net>

Le Tue, 16 Apr 2002 11:27:30 -0700
Erik Max Francis <max at alcyone.com> ?crivait:

> William Dode wrote:
> 
> > I mean that on the page of google they speak about api for java,
> > visual.net and perl, but not for python. I would like to see python
> > everywhere !
> 
> Well, that's because the Python interface was not done by them.
I said them, they was happy (and surprise) to know that it exists...

> 
> -- 
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.
> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.


-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From fredrik at pythonware.com  Tue Apr 30 10:01:42 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Apr 2002 14:01:42 GMT
Subject: Comment on draft PEP for deprecating six builtins
References: <007001c1efa7$04ede140$445d4540@Dell2> <mailman.1020106431.14597.python-list@python.org>
Message-ID: <atxz8.34775$n4.7371266@newsc.telia.net>

James J. Besemer wrote:
> This PEP is angles in the clouds stuff.  A lot of work for everybody and
> virtually zero practical benefit to anybody.

that's why you can safely ignore it.  you don't need to
feed the PEP trolls, you know.

</F>




From borcis at geneva-link.ch  Thu Apr 25 05:41:20 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Thu, 25 Apr 2002 11:41:20 +0200
Subject: Science And Math
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com> <mailman.1019498507.4893.python-list@python.org> <hup8cu8lrglr4e5tn1sn3c171bua1fsan2@4ax.com>
Message-ID: <3CC7CF40.9080904@geneva-link.ch>


Gon?alo Rodrigues wrote:

> On Mon, 22 Apr 2002 13:49:00 -0400, Lulu of the Lotus-Eaters
> 
>>It is obvious from my other little note in the thread that I feel
>>sympathy with the Intuitionist idea about mathematics.  Of course, this
>>sympathy is somewhat forclosed by my attachment to the Axiom of Choice.
> 
> If it were just the axiom of choice... 
> 
> These came about because these non-constructive proofs can yield some
> really hair-raising examples (an orange can be split and reassembled in
> such a way to yield two oranges, etc.)


BTW, I recently had a debate on Archimedes' Principle, that I started by 
focussing on it as "an example of a physics statement that can be 
entirely formulated without any maths" only to be attacked by a 
physicist implying that no word of it could be properly defined without 
math, and also implied that stronger results/principles/theorems should 
base weaker results. The whole with a lawyer-like booby-trapped rhetoric.

AC and the Banach-Tarski paradox (the strange consequence of AC you 
allude to in terms of "oranges") thus came in handy to win the debate 
(because it refuted the primacy of math in the way claimed by my 
contradictor, showing math did not per force provide an ultimate 
understanding of the concept of volume that is central to Archimedes' 
Principle).

And strangely enough, the name of "Banach-Tarski" admitted a parse that 
described the role I found for it : ((ban (act arse)) key)... I am still 
very much intrigued by that coincidence, and finding it suggests that 
what I called "booby-trapped lawyer-like rethoric" might show a 
structural similarity with the belief in AC. Even though this would not 
explain the coincidence away, but make it richer.

Regards, Boris Borcic
--
filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From marklists at mceahern.com  Sat Apr  6 14:06:01 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 6 Apr 2002 13:06:01 -0600
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204061839.g36Idfdj017724@ratthing-b246.strakt.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEHOKJAA.marklists@mceahern.com>

[Laura Creighton]
> My pleasure, especially since this is such a cool example.  You have
> just shot yourself in the foot, and do not know it.  When you have
> shipped your isValid function, across the world to your 10,000 clients,
> you will discover this.
>
> Your boss comes in and says, oh, Mark, we need a change to that
> isValid function.  We need to return 3 states, Valid, Invalid, and
> Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.
>
> What do you do now?

Get a different job?  ;-)

Seriously, if my boss knows which functions need to change and how, what am
I there for?

So let's say I start off with this function:

	def getValidity(user, password):
		return someInt

Now, I add a new return value to it for this new scenario.  I still have to
change the code that calls this function because it didn't previously know
about the new return value.

What have I gained aside from a useless rule that seems to promise me
use-me-and-you-won't-need-to-change-stuff, but can't really deliver?

Or maybe you're saying I'll have to change LESS stuff.  I don't think that's
what you're saying, else you'd have said it.  ;-)

I must confess, this feels like a shell game.  You want to deny the obvious
fact that at some point in my code I need to know whether the user's
password matches.  And it quite clearly either matches or it doesn't.  I
spell that boolean, I don't know about you.

You seem to be saying, "You *think* that's what you want, but later on you
might want other things."  Fine, I concede that point.  I'll cross that
bridge when I get to it.  I don't see how trying to cross it now--especially
when I have no clue as to the particular vector of change--is going to do me
any good.

If there's other stuff I need to do, I can easily write other code to do it.
And it may mean replacing or refactoring existing code.  Big deal.  Where's
the problem?  I just don't see the problem.  Much less the supposed
solution.

As you said, change is inevitable.  I don't have a problem with that.  I
still don't see what's wrong with using booleans.

> Do you now? Use Booleans when you want to return ON and OFF -- the
> state of that bit in memory, of something that is constrained by
> physical reality, or mathematical reality to only have 2 values, ever,
> cross my heart and the world will end if this ever ceases to be the
> case.

Such as whether the user's password is correct or not?

> Don't artificially limit yourself to 2 values because your
> vision at the time only extended to 2 values.  The one thing constant
> in this world is change.   If you don't need a third value, somebody
> else will, later.  Integers are your friends.

I'd say, don't artificially try to defer everything when your solution is
staring you in the face.

// mark





From max at alcyone.com  Mon Apr 22 13:34:55 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 22 Apr 2002 10:34:55 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com>
Message-ID: <3CC449BF.8030807@alcyone.com>

Brett g Porter wrote:

> In what ways are Scheme and Lisp only distantly related?

Probably in that the original poster likes Lisp (I presume he means 
Common Lisp) and dislikes Scheme, and wants to distance them as much as 
possible.

"Lisp" is generally the term applied to the broad category of Lisp-like 
languages, of which Scheme is most definitely an example.  Scheme is 
indeed one of the closer relatives to Common Lisp.

Here's a Web page outlining some of the differences between Scheme and 
Common Lisp:

	http://dept-info.labri.u-bordeaux.fr/~strandh/Teaching/MT

The differences can be significant from the low level view, but from the 
big picture they are very similar.
-- 
  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
  A new, virtual planet, every day.



From aleax at aleax.it  Sat Apr 27 18:40:47 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 27 Apr 2002 22:40:47 GMT
Subject: Update locals() / resolved
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org>
Message-ID: <PNFy8.28407$8D3.758650@news1.tin.it>

holger krekel wrote:
        ...
> ok. i found a nice workaround (if one dare call it that:-)
> 
> class dictlookup(type({})):
>     def __getattr__(self, name):
>         return self[name]
> 
> then i can do:
> 
> def func(...,somedict,...):
>     m = dictlookup(somedict)
>     ...
>        m.key1

I'd implement it quite differently, myself:

class dictlookup:
    def __init__(self, somedict):
        self.__dict__ = somedict


Alex



From johnroth at ameritech.net  Mon Apr 29 15:25:44 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 29 Apr 2002 15:25:44 -0400
Subject: comparing str's with ints
References: <20020429115856.R1840-100000@palanthas.neverending.org> <aak5n9$efr$1@slb6.atl.mindspring.net>
Message-ID: <ucr7gdps6vuh0a@news.supernews.com>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:aak5n9$efr$1 at slb6.atl.mindspring.net...
> Frank Tobin:
> >What in the world is supposed to be the semantics of comparing a
string
> >with an integer?  Seems like a type trap...
>
> As I understand the history, earlier versions of Python didn't support
> throwing exceptions from an internal cmp().  Instead, it used a
> well-defined but arbitrary comparison scheme to handle failures.
> Though that scheme kept being tweaked as mismatches were found between
> it and expectations.
>
> I agree that it's a type trap.
>

Exactly. As I understand it, the semantics of comparing
two dissimilar types are quite simple. They are supposed
to provide consistent results on any given release, but
not otherwise. Consistent simply means that the same
comparison should provide the same result, not that it
should mean anything to anyone other than the interpreter.

In other words, don't do that.

If you really, truely have to do Perl type comparisons
between integers and strings, you might consider
deriving your own integer and string types, and providing
a custom __cmp__() method.

John Roth





From st-newsgroups at thomas-guettler.de  Wed Apr 10 10:27:46 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Wed, 10 Apr 2002 16:27:46 +0200
Subject: Create Excel Files without COM
References: <3CB44516.60005@thomas-guettler.de> <slrnab8hp2.3r2.stefan.heimann@kunz.ratzer>
Message-ID: <3CB44BE2.6080509@thomas-guettler.de>

Stefan Heimann wrote:

> Thomas Guettler <st-newsgroups at thomas-guettler.de> wrote::
> 
> 
>>Hi!
>>
>>Is there something like WriteExcel from perl in python?
>>I need to generate an MS-Excel file. I don't want to use
>>COM although it will run on windows.
>>
>>The data is simple. Just text in columns and rows.
>>
> 
> What about CSV? 


CSV or the XML Version of an Excel Spreadsheet would be very easy
to produce. But it would be better if it would be .xls

  thomas





From jb at cascade-sys.com  Thu Apr 25 20:39:21 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 17:39:21 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com>
Message-ID: <3CC8A1B9.C2BDD833@cascade-sys.com>

Aahz wrote:

> Ewww!  Where did you get the idea that awk is a precursor of Python?

Look up "precursor" in the dictionary and you'll see it's an accurate statement.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From deokhwankim at hotmail.com  Wed Apr 24 00:26:25 2002
From: deokhwankim at hotmail.com (Deokhwan Kim)
Date: 23 Apr 2002 21:26:25 -0700
Subject: static method & class method
Message-ID: <320bded.0204232026.354c21e1@posting.google.com>

I have heard that static method and class method are available from
2.2.

But It seems to me that static method is redundant.  Is there anything
that static method can do but that class method can NOT do?  Or any
benefit?

Thank you for your reading.


From neuhauser at mail.cz  Sat Apr 27 19:14:48 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Sun, 28 Apr 2002 01:14:48 +0200
Subject: slices: foo[i:j:k] => TypeError
In-Reply-To: <fFDy8.70854$vF6.2116208@news2.tin.it>
References: <mailman.1019856470.541.python-list@python.org> <aCky8.66523$vF6.2038703@news2.tin.it> <mailman.1019924208.19241.python-list@python.org> <fFDy8.70854$vF6.2116208@news2.tin.it>
Message-ID: <20020427231448.GE82912@roman.mobil.cz>

> From: Alex Martelli <aleax at aleax.it>
> Subject: Re: slices: foo[i:j:k] => TypeError
> To: python-list at python.org
> Date: Sat, 27 Apr 2002 20:15:07 GMT
> 
> Roman Neuhauser wrote:
>         ...
> >     Ok. What I don't understand then is why is extended slicing
> >     described among primaries in the language reference. And I'm not
> 
> Because the parser needs to know about it (to support Numeric).  Indeed,
> it's not a syntax error that you're getting.
> 
> >     talking about a development version of the manual, I'm talking about
> >     the 2.1 version.
> 
> I think it was introduced in 1.5.2.
 
    quite some time, isn't it? :)

> >     What's the expected date for the builtin types to accept slice
> >     objects?
> 
> No idea.  The original intro was before PEPs, so there's not even a
> PEP to refer to, I believe.  You may want to ask on Python-Dev --
> if the BDFL OK's it for 2.3, I can write the code, if you or some other
> volunteer can handle the rest of the patch submissal process (sorry,
> can't volunteer more time than what it takes to write & debug the code).

    No, thanks. I'll stick with NotImplemented or sumsuch. I just found
    it interesting that the manual describes something that's not even
    meant to work.

    Thanks for the info!

-- 
FreeBSD 4.4-STABLE
1:13AM up 12 days, 14:25, 9 users, load averages: 0.00, 0.02, 0.00




From tismer at tismer.com  Mon Apr 29 06:44:47 2002
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 29 Apr 2002 12:44:47 +0200
Subject: Looking for GUI Form Builder for Python
References: <a9en77$4vf8@imsp212.netvigator.com> <3CBB41E5.8080901@csi.com>
Message-ID: <3CCD241F.10505@tismer.com>

Uwe Fechner wrote:
> Calvin wrote:
> 
>> Dear all,
>>  Can I build Python TK GUI interface using a form builder. Anyone know 
>> such
>> builder. I just want to build my interface in a visual way.
>> -Calvin
>>
>>
> 
> Hello,
> 
> I think, that there is only one good solution:
> 
> Boa Constructor:
> 
> http://sourceforge.net/projects/boa-constructor/
> 
> It is a little bit hard to install, (you should use the newest version 
> from cvs), but it's a really good IDE and open source.

While this is not Tk but wxPython, I want to second
this statement. Boa Constructor is great, and
getting better every day.
If you need help with the installation, drop me a note.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From cwood at austin.rr.com  Sun Apr 28 22:50:26 2002
From: cwood at austin.rr.com (Charlie Wood)
Date: Mon, 29 Apr 2002 02:50:26 GMT
Subject: Super-newbie syntax error question
References: <mailman.1020045350.28826.python-list@python.org>
Message-ID: <Sx2z8.6921$9F5.439339@typhoon.austin.rr.com>

I found the error. It was a space/tab issue with my text editor.

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in message
news:mailman.1020045350.28826.python-list at python.org...
> >
> >     f = file("/article.txt", "w")
> >     f.write('Category: %s\nTitle: %s\nDescription: %s\nLink: %s\n\n' %
> >       (self._category, self._title, self._descr, self._link))
> >     f.close()
> >
> >                 self._title = None
> >                 self._link = None
> >                 self._descr = ""
> >                 self._category = ""
> >
> >
> >
> > ...and here's the resulting error:
> >
> >
> >
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> >   File "C:\PYTHON22\lib\rss2html.py", line 46
> >     f = file("/article.txt", "w")
> >     ^
> > SyntaxError: invalid syntax
> >
>
> Should be:
>
> f = open() not file().
>
>




From barry at zope.com  Thu Apr  4 15:36:18 2002
From: barry at zope.com (Barry A. Warsaw)
Date: Thu, 4 Apr 2002 15:36:18 -0500
Subject: RELEASED Mailman 2.0.9
Message-ID: <15532.47426.82687.419071@anthem.wooz.org>

I've just released version 2.0.9 of Mailman, the GNU Mailing List
Manager.  Mailman is released under the GNU General Public License
(GPL).  Version 2.0.9 fixes a race condition that can cause mail to
occasionally be lost, among other useful fixes.  I recommend anybody
running any version of Mailman 2.0.x to upgrade to 2.0.9, the latest
stable release.  Details of what's changed are included below.

GNU Mailman is software to help manage electronic mail discussion
lists.  Mailman gives each mailing list a unique web page and allows
users to subscribe, unsubscribe, and change their account options over
the web.  Even the list manager can administer his or her list
entirely via the web.  Mailman has most of the features that people
want in a mailing list management system, including built-in
archiving, mail-to-news gateways, spam filters, bounce detection,
digest delivery, and so on.

Mailman is compatible with most web servers, web browsers, and mail
servers.  It runs on any Unix-like operating system.  Mailman 2.0.9
requires Python 1.5.2 or newer.  To install Mailman from source, you
will need a C compiler.

For more information on Mailman, including links to file downloads,
please see the Mailman web page: http://www.gnu.org/software/mailman

And its mirrors at:

    http://mailman.sourceforge.net
    http://www.list.org

Patches and source tarballs are available at

    http://sourceforge.net/project/showfiles.php?group_id=103

There are email lists (managed by Mailman, of course!) for both
Mailman users and developers.  See the web sites above for details.

Cheers,
-Barry

2.0.9 (02-Apr-2002)

    - Closed a race condition which could, under rare circumstances,
      cause the occasional message to get lost.

    - HTML escape message excerpts and headers on the admindb page so
      JavaScript and other evil tags can't mess up the display.

    - Some additional Python 2.2 compatibility fixes.

    - Unlink the footer logos so as not to bug the python.org and
      gnu.org maintainers as much. :(

    - Fix a crash in the DSN bounce detection module, which could
      cause some bounce messages to remain in the queue.

    - Add the RFC-2822 mandated Date: header on internally generated
      outgoing messages.  Not all MTAs add this field if missing
      (read: Qmail).




From phd at phd.pp.ru  Sun Apr 28 16:08:45 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 29 Apr 2002 00:08:45 +0400
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <uco40vl11ipj04@corp.supernews.com>; from rhhyatt@charter.net on Sun, Apr 28, 2002 at 08:11:36AM -0700
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <5yEp8.5051$hi7.18667@news-server.bigpond.net.au> <3CA7192C.E1C8974A@engcorp.com> <mailman.1017612966.7043.python-list@python.org> <uco40vl11ipj04@corp.supernews.com>
Message-ID: <20020429000845.C26722@phd.pp.ru>

On Sun, Apr 28, 2002 at 08:11:36AM -0700, Bob Hyatt wrote:
> I have had a fruitless search for Gadfly. As you noted, the website for
> Chordate is not responding.

   I am keeping (temporary) a copy of kjbuckets at
http://phd.pp.ru/Software/Python/misc/ and mirrors:
http://phd.by.ru/Software/Python/misc/
http://phd2.chat.ru/Software/Python/misc/
   I am looking for a permanent maintainer.

   Gadfly now lives in the Zope sources at www.zope.org.

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




From slackness at caramail.com  Mon Apr 29 03:21:02 2002
From: slackness at caramail.com (slack)
Date: 29 Apr 2002 00:21:02 -0700
Subject: Python newbie
References: <ab96a890.0204131438.5cce8f66@posting.google.com> <slrnabmlpd.a8q.ak@ak.silmarill.org>
Message-ID: <ab96a890.0204282321.6c22697e@posting.google.com>

Thank you but I know that. I would like my programs to be launched by
people who don't have Python interpreter on their machine. I think
that Py2exe (mentionned before by.. I don't remember, sorry, but
thanks) could resolve this problem. I'll try it.
I would like to know one more thing... Is it possible to use Python to
program for Windows CE devices??

Thanks


From op73418 at mail.telepac.pt  Mon Apr  1 18:50:59 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 02 Apr 2002 00:50:59 +0100
Subject: unicode string problems
References: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com> <m3u1qvvw6z.fsf@mira.informatik.hu-berlin.de>
Message-ID: <dfshau4s6cd8jsl1a5eeoic8dmedb2pkib@4ax.com>

On 02 Apr 2002 01:24:52 +0200, martin at v.loewis.de (Martin v. Loewis)
wrote:

>Gon?alo Rodrigues <op73418 at mail.telepac.pt> writes:
>
>> My problem is the following:
>> 
>> Let f be a file object. If I do
>> 
>> f.write("Mar?o 2002")
>> 
>> all goes well. But if I try
>
>It depends on your meaning of "well". It writes 9 bytes into the file,
>yes.

All goes well = it magically works without me having to give a second
thought of encodings ;-)

>
>> f.write("Mar?o 2002" + march.Name())
>> 
>> where march.Name() returns a unicode string I get a unicode error. I
>> tried converting both unicodes to strings via str but obviously I got an
>> error (in the first string the culprit is the "?" character).
>> 
>> Can someone help me out here and show me the way to write these strings
>> to the file?
>
>If you want to write text into a file, you should know what encoding
>you want to use in this file. Then make sure that all byte strings
>written to the file indeed use that encoding.
>
>In the specific case,
>
>f.write("Mar?o 2002" + march.Name().encode("iso-8859-1"))
>
>will do the right thing, assuming you want the text to be encoded in
>that encoding.

Thanks, this should solve it. Now back to coding (gonna be a long
night...)

>
>HTH,
>Martin

All the best,
Gon?alo Rodrigues



From joakim.storck at home.se  Tue Apr  9 13:41:26 2002
From: joakim.storck at home.se (Joakim Storck)
Date: 9 Apr 2002 10:41:26 -0700
Subject: xml encoding in minidom
Message-ID: <db8a9aa8.0204090941.3b518901@posting.google.com>

I've been playing around with the xml.dom.minidom, but I have some
problems with the document encoding. What I'm trying to do is
basically to take care of data from a web form using the cgi-module,
then insert data into a xml-document using the minidom. However, the
data does contain swedish characters (?, ?, ?) which messes things up
more than I could have imagined.

Since the minidom has no way of setting document encoding directly, i
tried the following:

doc = xml.dom.minidom.Document()
pi = doc.createProcessingInstruction('xml','version="1.0"
encoding="ISO-8859-1" ')
doc.appendChild(pi)

But then I get double headers:

<?xml version="1.0" ?>
<?xml version="1.0" encoding="ISO-8859-1" ?>

There is also some kind of problem with the minidom.toxml()-method.
>From what I've read in other postings it has to do with file encoding,
which seems to be set to UTF-8 by default.

Preferrably I'd like to take care of both problems by writing
something like this:

doc = xml.dom.minidom.Document()
doc.setEncoding('ISO-8859-1')

I'm no expert, neither at xml nor Python, but it seems I'm not the
only one who experienced these problems.

Thanks' for any suggestions!

Joakim


From ark at research.att.com  Wed Apr  3 14:20:19 2002
From: ark at research.att.com (Andrew Koenig)
Date: 03 Apr 2002 14:20:19 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <031001c1db3c$0460ad90$ced241d5@hagrid>
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
	<yu99wuvovkmm.fsf@europa.research.att.com>
	<00af01c1db2a$e7316d10$ced241d5@hagrid>
	<200204031619.g33GJXm02577@europa.research.att.com>
	<016b01c1db31$8da51010$ced241d5@hagrid>
	<yu99ofh0vg4x.fsf@europa.research.att.com>
	<031001c1db3c$0460ad90$ced241d5@hagrid>
Message-ID: <yu99d6xgd1xo.fsf@europa.research.att.com>

Fredrik> looks pretty consistent to me ;-)

It may look that way, but if the documentation is to be believed,
it's just happenstance.

Fredrik> (Python docs tend to use "true" to mean anything which isn't
Fredrik> "false", and where operator.truth returns the right thing.
Fredrik> this won't change post-PEP -- if you start interpreting
Fredrik> "return true" as "return True", you'll be asking for trouble)

Indeed.  In other words, the library does not have a consistent
convention for what values to return for answers to yes/no questions.
A particular implementation may happen to follow such a convention,
but it is under no obligation to do so.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark




From rjones at ekit-inc.com  Mon Apr  1 18:21:18 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 2 Apr 2002 09:21:18 +1000
Subject: restructured pythonpoint [was Re: The fastest XML parser around]
In-Reply-To: <UD5q8.874$aO4.4453379@news-text.cableinet.net>
References: <mailman.1017654423.12312.clpa-moderators@python.org> <a89pnc$7j9$1@panix1.panix.com> <UD5q8.874$aO4.4453379@news-text.cableinet.net>
Message-ID: <200204012321.XAA06709@crown.off.ekorp.com>

On Tue, 2 Apr 2002 09:00, Andy Robinson wrote:
> Not yet.  The simplest thing would be to add a 3-line hack to validate
> the document if pyRXP is present :-)
>
> I plan to rewrite pythonpoint using our XML toolkit - indeed, use it
> as the exercise in the manual - but we need to finalize two more components
> first:  a 'lazy accessor' giving friendly syntax for navigating the tree,
> and
> something to convert attributes to numbers/lengths/colors en masse.
> After that the Pythonpoint code could get a lot shorter.  Some broad
> plans are in the back of the pyRXP documentation, referenced from
> the web page my announcement pointed at.

Just as an aside - I wrote a quick'n'dirty ReStructuredText 
(http://structuredtext.sf.net/) to PythonPoint converter a few weeks ago. It 
makes writing the slides a _lot_ easier as long as you don't want to get too 
fancy (for example, don't indent a bullet list, 'cos the pypt renderer will 
barf trying to place the bullet point in an indented paragraph).

The stx-to-pypt converter is checked into the ReST_ sandbox under the 
richardj directory: 

http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/structuredtext/sandbox/richardj/


    Richard



Example:

ekit Website
------------

*Richard Jones, Anthony Baxter*

*ekit.com Technical Conference, March 2002*

History
-------

- Blah blah blah blah blah
- Blah blah blah blah blah
- Blah blah blah blah blah

ekit At A Glance
----------------

- Blah blah blah blah blah
- Blah blah blah blah blah
- Blah blah blah blah blah

Page Structure
--------------

Either in frames or in a single table (shown below)

.. image:: page_structure.png


[and so on]



From ping at lfw.org  Wed Apr  3 13:06:31 2002
From: ping at lfw.org (Ka-Ping Yee)
Date: Wed, 3 Apr 2002 12:06:31 -0600 (CST)
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <200204031712.g33HCaP22228@odiug.zope.com>
Message-ID: <Pine.LNX.4.33.0204031118110.16190-100000@server1.lfw.org>

On Wed, 3 Apr 2002, Guido van Rossum wrote:
> You misunderstand it.  Strings, lists, numbers and so on are still
> acceptable as truth values, and when you want to know whether x is
> true or false, you still have to say "if x:" -- never "if x == True:".

Yes yes.  I understand this part just fine.  It's not the *listening*
i'm concerned about -- 'if' takes care of that easily for me.  It hears
[], None, 0 as false and hears 'spam', {1: 2}, 47.3 as true.

It's the *speaking*.  When i want to say true or false, then there's
the dilemma.

I know, your answer is "you should always just say True or False",
and Mark McEahern said the same thing.  But this cannot be so in
practice: *everything* already returns 0 or 1.  (It may be possible to
get around this if we commit to changing the entire standard library
before releasing a version of Python with True and False, but alas,
this is not the only issue... read on.)

As long as True and False are somewhere represented as 0 and 1,
the values 0 and 1 will never lose legitimacy as booleans.  This
business with str() and/or repr() producing "0" or "1" for backwards
compatibility prevents us from considering 0 and 1 relegated to a
truly non-boolean status.

Consider this:

    >>> a = [0, False, 1, True]

    >>> print a
    [0, 0, 1, 1]

    >>> for x in a: print x
    0
    0
    1
    1

Good heavens!

What about this:

    >>> d = {}
    >>> d[0] = 'zero'
    >>> d[False] = 'false'
    >>> len(d)
    1 or 2?

Basically, the concept of having a permanently schizophrenic type
in the language scares me.  The above shows, i believe, that a
reasonable implementation must print True as True and False as False,
and never mention 1 or 0.  Moreover, as soon as you start sorting a
bag of objects, or keying dictionaries on objects, you are forced to
run into the distinction between 0 and False, and between 1 and True.

I'm not against the idea of booleans, of course -- but i do think
that halfway booleans are worse than what we have now.  And getting
to real booleans [*] involves real pain; it's just a question of
whether that pain is worth it.  Even if we get all the way there --
as in we manage to convert enough code and convince everyone to use
the new style -- i will never ever want "and" and "or" to return
booleans (i just hope that doesn't confuse anyone).


-- ?!ng

[*] By real booleans, i mean the following.  (Booleans would have
    to behave like this for me to consider them "good enough" to
    be better than what we have now.)

    >>> False, repr(False), str(False)
    (False, 'False', 'False')
    >>> True, repr(False), str(False)
    (True, 'False', 'False')
    >>> False + True
    TypeError...
    >>> False == None
    0
    >>> False == 0
    0
    >>> True == 1
    0
    >>> {0: 0, False: False, 1: 1, True: True}
    {0: 0, False: False, 1: 1, True: True}

... and probably

    >>> None < False < True < 0
    True

(Hee hee -- i suppose the fact that "boolean" starts with a "b"
gets us this for free.  But i wonder how many people are going
to be puzzled by True < 0?)





From mstern at aon.at  Mon Apr  1 11:54:19 2002
From: mstern at aon.at (Matthias Stern)
Date: Mon, 1 Apr 2002 18:54:19 +0200
Subject: [Newbie] embed c++ in Python
Message-ID: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at>

Hi!

I'm new to Python and I'm looking for a good tutorial how to embed a C/C++
function into Python. I found some tutorials but no one describes what I
need:

I want to know how to write a function which receives some variables as
argument changes them and sends the results back to Python. E.g.:

int doSomething(int *a, int *b) {
   :
   *a = 10;
   *b = 123;
   :
}

Can I do this in Python? I'd like to call the function doSomething() from
Python and receive the changed values a and b in Python again.

Is there a tutorial? Has someone got a simple example for this?

Thx, Mac




From boud at valdyas.org  Sun Apr 28 11:04:54 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sun, 28 Apr 2002 17:04:54 +0200
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <3ccc0fd3$0$37904$e4fe514c@dreader3.news.xs4all.nl>

Philipp Lenssen wrote:

> Is there a simple, open source text-editor project available in Python?

Do you want a simple editor to type python code in, or a simple editor
to embed in your own apps, or do you want an editor written in Python?

If the first -- anything will do. If you mean the second: all GUI toolkits
have what you need. As for the third: I've been working on an emacs-clone
written in Python, and I have some snippets of source available, if you want
them. It uses PyQt to actually put characters on screen, but the rest is
handled by Python.

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


From boud at valdyas.org  Mon Apr  1 16:36:41 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 01 Apr 2002 20:36:41 -0100
Subject: ANN: GUI Programming with Python using Qt
Message-ID: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl>

Well, I'd like to finally officially announce that my book, _GUI 
Programming with Python using the Qt Toolkit_ is now officially available 
on paper. Whew!

You buy it directly from the publisher's website: http://www.opendocs.org,

or, preferably, at your favourite local or on-line bookshop. For instance, 
Amazon lists it as:

http://www.amazon.com/exec/obidos/ASIN/0970033044/qid=1017685892/sr=8-1/ref=sr_8_1_1/103-3219259-3855046

(And I wonder how that's going to come across.) The ISBN is 0-97003300-4-4.

Anyway, I've written about 160.000 words of English prose and code, all 
about Qt, PyQt, sip, Qt Designer and BlackAdder. I cover both Qt 2 and Qt 3 
in a lot of detail.

All source code to the examples is included on a cd-rom, and Opendocs will 
shortly open a webforum where I'll post updated code and where people can 
discuss problems.

You can read the book online at: http://www.opendocspublishing.com/pyqt.

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


From cwood at austin.rr.com  Sun Apr 28 19:07:14 2002
From: cwood at austin.rr.com (Charlie Wood)
Date: Sun, 28 Apr 2002 23:07:14 GMT
Subject: Super-newbie syntax error question
Message-ID: <Cg%y8.5752$9F5.405701@typhoon.austin.rr.com>

I have some code that I've cobbled together, and have just introduced a
syntax error. I've read through the docs, done lots of trial and error, and
I'm left scratching my head. Any ideas?

Here's the code:

    def endElement(self, name):
        if self._parent == "item":
            if name == "title":
                self._title = self._text
            elif name == "link":
                self._link = self._text
            elif name == "description":
                self._descr = self._text
            elif name == "category":
                self._category = self._text
            elif name == "item":
                if not self._list_started:
                    self._out.write("\n")
                    self._list_started = 1

                self._out.write('%s: %s\n%s\n%s\n\n' %
                                (self._category, self._title, self._descr,
self._link))

    f = file("/article.txt", "w")
    f.write('Category: %s\nTitle: %s\nDescription: %s\nLink: %s\n\n' %
      (self._category, self._title, self._descr, self._link))
    f.close()

                self._title = None
                self._link = None
                self._descr = ""
                self._category = ""



...and here's the resulting error:



Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "C:\PYTHON22\lib\rss2html.py", line 46
    f = file("/article.txt", "w")
    ^
SyntaxError: invalid syntax



Your kindness is appreciated.





From aleax at aleax.it  Sun Apr  7 11:10:39 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 07 Apr 2002 15:10:39 GMT
Subject: Booleans, integer division, backwards compatibility; where is  Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <1018120666.804421@yasure> <3CAFA81B.B9E4AF9@python.org>
Message-ID: <PjZr8.57418$pT1.1647471@news1.tin.it>

<posted & mailed>

Guido van Rossum wrote:
        ...
> there (note that python-dev no longer has a closed-subscription policy).

How so?  One visits:

http://mail.python.org/mailman/listinfo/python-dev

enters email and passwords, hits subscribe, and gets this message:

"""
Python-Dev Subscription results

Subscription was deferred because subscriptions to Python-Dev require 
administrator approval.  Your request has been forwarded to the list 
administrator.  You will receive email informing you of the moderator's 
decision when they get to your request.
"""

"Requires administrator approval" may not be "a closed-subscription
policy" but it sure looks quite close.  Besides, the subscription does not
actually start (at least it doesn't seem to have -- I'm trying to follow
the thread by periodically hitting refresh on :
http://mail.python.org/pipermail/python-dev/2002-April/date.html
[quite unwieldy, no facilities to reply, etc] and I've noticed I cannot rely
on everybody CC'ing me when they post...).

If this is meant to be an open-subscription mailing list, I would suggest
this arrangement be changed.


Alex



From mgilfix at eecs.tufts.edu  Sun Apr 21 13:31:51 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sun, 21 Apr 2002 13:31:51 -0400
Subject: Python's Lisp heritage
In-Reply-To: <3CC2872D.EF791985@cascade-sys.com>; from jb@cascade-sys.com on Sun, Apr 21, 2002 at 02:32:29AM -0700
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <3CC2872D.EF791985@cascade-sys.com>
Message-ID: <20020421133151.A5380@eecs.tufts.edu>

  I agree that functional support is a good thing. Why can't a
language include support for multiple programming paradigms? When
I first started using python, I was really pleasing that I
had a powerful OOP paradigm as well as functional primitives.
filter/reduce/map sometimes fit certain tasks in ways that iteration
can make combersome. When used appropriately, they express a task in a
much more complex yet /clearer/ way.

  Each pardigm has its advantages and disadvantages. Isn't the best
language one which allows you to emply the right tool in the right
situation?

                        -- Mike

On Sun, Apr 21 @ 02:32, James J. Besemer wrote:
> > They were added to Python just before the 1.0 release, as a contribution
> > by someone other than Guido.  He's noted for saying way back in 1994
> > ] To be honest, I wish I hadn't introduced lambda, map, filter and
> > ] reduce -- they support a style that is inconsistent with the rest of
> > ] Python. ...
> 
> That's disappointing to hear.  I thought including the ability to do some
> functional-style programming was a great thing.  Somebody else here or in one
> of the books I read recently I thought said support for functional programming
> was a good thing.
> 
> > BTW, the first two features are best done in modern Python using
> > list comprehensions and the f(*a, **kw) syntax.
> 
> "best" is a value judgment.  But it's true MAP could be defined as a user
> function and apply appears to serve no actual value I can discern.
> 
> Lambda however seems useful, judging by the number of times it seems to show up
> in code examples.

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From phr-n2002a at nightsong.com  Tue Apr 23 01:49:26 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 22 Apr 2002 22:49:26 -0700
Subject: need encryption algorithm
References: <e2e9e807.0204221741.597acd60@posting.google.com> <e2e9e807.0204222041.5fa92f05@posting.google.com>
Message-ID: <7xy9ffx8ux.fsf@ruckus.brouhaha.com>

h.e.w.frater at cs.cf.ac.uk (Hugh) writes:
> I fully understand what everyone is going on about with regard to me
> saying that the quality of the encryption is not that important. The
> fact of the matter is, I'm only writing the software for my 1st year
> university project and in reality it's only a theoretical system, no
> one is actually going to use it. I'm using the exercise to learn
> python and basic database design although to be quite honest I loathe
> internet programming. Don't know why, just don't like it.

In that case just prefix the string with "EEEE" to signify that it's
encrypted, and remove the EEEE on "decryption".


From neal at metaslash.com  Tue Apr  9 17:06:43 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 09 Apr 2002 17:06:43 -0400
Subject: property() type that only protects set?
References: <3cb34fb8$1_4@goliath.newsgroups.com>
Message-ID: <3CB357E3.95EFC8E1@metaslash.com>

Brad Clements wrote:
> 
> Anyone have an implementation of property() that doesn't require a getter,
> only a setter (and optional del)
> 
> I only need to protect set operations, and want to allow get to go straight
> to __dict__
> 
> (or object when I'm using slots)

>From the docstring:

>>> print property.__doc__
property(fget=None, fset=None, fdel=None, doc=None) -> property attribute

fget is a function to be used for getting an attribute value, and likewise
fset is a function for setting, and fdel a function for del'ing, an
attribute.  Typical use is to define a managed attribute x:
class C(object):
    def getx(self): return self.__x
    def setx(self, value): self.__x = value
    def delx(self): del self.__x
    x = property(getx, setx, delx, "I'm the 'x' property.")
----------------------------------------------------------------------------

so use:

    x = property(fset=setx)

Neal


From robinjim at earthlink.net  Sat Apr  6 07:38:19 2002
From: robinjim at earthlink.net (robin and jim)
Date: Sat, 06 Apr 2002 12:38:19 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>
Message-ID: <%_Br8.20276$nt1.1661286@newsread2.prod.itd.earthlink.net>

Gee, I thought it would be obvious why I was asking such a question <wink>!

As unit tests run, using the unittest module shipped with Python 2.2,  the
default behavior is to display a one-line description of each test. The
default implementation displays the first line of the test method's
**docstring**, if available; otherwise the name of the test method (e.g.,
"test_my_regular_expression") is displayed.

Well, yours truly thought he would be clever and add some additional
information to this docstring as it became available at run time via the %
placeholder mechanism.  Alas, the result is a single-quoted string that
unittest does not recognize as a docstring so it displays the test method's
name (e.g., "test_my_regular_expression") instead. I, naively, think that if
the single quoted result can be re-converted to a triple-quoted string, it
will be recognized by unittest.

Hope this helps clarify what I'm asking -- and thanks for your interest.

"robin and jim" <robinjim at earthlink.net> wrote in message
news:GZqr8.20207$ml2.1517221 at newsread1.prod.itd.earthlink.net...
> Is there a way to convert a single quoted string to a triple quoted
string?
>
> For example, given:
>
> ''' abc %d xyz '''
>
> the result of ''' abc %d xyz''' % 10 is:
>
> ' abc 10 xyz '
>
> which is a single quoted string.
>
> I would like to convert the result to a triple-quoted string.
>
>
>




From jon+usenet at unequivocal.co.uk  Mon Apr  1 17:21:18 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 01 Apr 2002 22:21:18 -0000
Subject: Python embedded like PHP
References: <mailman.1015988122.18668.python-list@python.org> <359f85cd.0203191203.2a8e4bd@posting.google.com> <mailman.1016619871.27287.python-list@python.org> <7xk7s7mpiy.fsf@ruckus.brouhaha.com> <mailman.1016626151.9950.python-list@python.org> <slrna9haud.e6b.akuchlin@ute.mems-exchange.org> <slrnaa3bi8.lam.jon+usenet@snowy.squish.net> <23891c90.0203290655.25bbcab9@posting.google.com> <slrnaacob1.bu2.jon+usenet@snowy.squish.net> <23891c90.0203310820.60696fb1@posting.google.com>
Message-ID: <slrnaahnai.i1t.jon+usenet@snowy.squish.net>

In article <23891c90.0203310820.60696fb1 at posting.google.com>, Paul Boddie wrote:
> Template systems typically don't address these issues, whereas
> stand-alone validation systems address the nature of the input data
> but not how such data is "described" to the browser (and thus to the
> user). A coordinated approach which unifies input, output, and
> validation seems to be the logical next step in any significantly big
> or complex application.

Ah, I see. I must admit I have never seen a web application where
the benefits this would provide would outweigh the extra complexities
it would entail. From an HTML point of view, the only two input
attributes which could be magically generated are 'size' and
'maxlength'. 'size' is more dependent on the form design than the
input requirements, and hence it is not appropriate to base this value
on the input type, and the benefits of automagically setting maxlength
would seem to be extremely minor.

>From a coding point of view, I would think the benefits are fairly
small too. In most applications I can think of, input data is usually
being looked up from rather than entered into a database. For example,
if the user enters a username the question is not normally "is this a
valid username?", but "is this an existing username?", so no input
validation is required. I guess the common exception is email
addresses.

The nearest jonpy gets to this sort of thing is MultiForm, a class
for creating multi-page input forms. It provides a 'check' method
which you override to validate the input data and produce appropriate
error messages.

Cheers


Jon


From joost_jacob at hotmail.com  Mon Apr 29 10:24:41 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 29 Apr 2002 07:24:41 -0700
Subject: Tkinter Text widget reading input (how to do overwrite mode?)
Message-ID: <13285ea2.0204290624.4fbe8cba@posting.google.com>

Here is a question for somebody with experience with Tkinter Text widgets.
I have a Tkinter widget that inherits from Frame and it has a self.text
attribute of type Text.
To read user input in my widget (until the user presses RETURN) i add
the following two methods (this is part of my code):

    def readxy(self, x=0, y=1):
        where = `y`+'.'+`x`
        self.startUserInput = where
        self.userInput = StringVar()
        self.userInput.set('')
        self.text.mark_set(INSERT, where)
        self.text.bind('<Return>', lambda event, self=self: self._onEnter(event))
        self.text.focus()
        self.text.update()
        self.text.wait_variable(self.userInput)
        return self.userInput.get()
    def _onEnter(self, event):
        self.userInput.set(self.text.get(self.startUserInput, INSERT))
        self.text.mark_unset(INSERT)
        self.text.unbind('<Return>', '')

This works fine, but the user is typing in 'INSERT' mode.  That means that any
text on the line after the cursor is shifted right with each typed character.
My question is: how can i get an 'OVERWRITE' mode ?
Do i have to program that manually or is there a simple way to 'put a Text
widget in overwrite mode'.  Did somebody do this before?  Of course i searched
the net for an answer before asking it here on c.l.p.  I noticed that IDLE
does not have an 'OVERWRITE' mode either, nor does any other editor
implemented with Tkinter i could find on the net.

Joost Jacob
www.liacs.nl/home/jjacob/

I have a very small mind and must live with it.
                -- E. Dijkstra


From tanzer at swing.co.at  Tue Apr 16 02:35:08 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Tue, 16 Apr 2002 08:35:08 +0200
Subject: Does Python need a '>>>' operator?
In-Reply-To: Your message of "Mon, 15 Apr 2002 19:07:02 GMT."
             <qxFu8.18112$b62.514495@news1.tin.it>
Message-ID: <E16xMYW-0007Q5-00@tswings.swing.cluster>

Alex Martelli <aleax at aleax.it> wrote:

> `x` (backquotes) is another example, a silly duplication of repr.

Amen.

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





From tinu at email.ch  Thu Apr 11 03:17:59 2002
From: tinu at email.ch (tinu)
Date: 11 Apr 2002 09:17:59 +0200
Subject: Python 2.2.1: no readline support on Linux
Message-ID: <m3g022d7q0.fsf@yak.ethz.ch>

Dear Pythoneers

I just installed Python 2.2.1 on SuSE Linux 7.3, (standard
build from source) and lost all readlines support. The same
feature worked perfect with 2.1. Has anybody advice on this?

Thnx

Martin L?thi



From martin at v.loewis.de  Tue Apr 16 02:30:11 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 16 Apr 2002 08:30:11 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de> <a9frf6$d6u$0@216.39.172.122>
Message-ID: <m3662sm9zg.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> >> I would guess so. I do. But for language purposes my preference is
> >> to refer to an abstract bit pattern, not hardware (even though
> >> hardware mappings will exist and be useful).
> >
> >That is only meaningful for positive numbers - it just doesn't work
> >for twos complement.
> >
> I works fine the way I'm thinking of it ;-) You must be looking at
> it differently, and I'm not sure what view it doesn't work for.
> My view is an abstraction where the sign bit is indefinitely duplicated
> to the left, but since they're all the same, you can think of them
> as represented by a single sign bit. 

That part works fine, I agree. However, neither the traditional hex
representation (strip of leading zeroes), nor the "logical shift"
operator >>> (insert zeroes when right-shifting) is meaningful using
this mental model. For the former, a negative number gets an infinite
hex representation, for the latter, it is impossible to tell how many
of the infinite number of one bits should be cleared in a logical
shift.

Regards,
Martin


From sholden at holdenweb.com  Tue Apr 30 12:25:13 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 12:25:13 -0400
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de> <xduu1pt1iuf.fsf@leto.informatik.uni-freiburg.de> <vIxz8.81546$vF6.2424489@news2.tin.it> <aamebj$c2olc$1@ID-59885.news.dfncis.de> <1bzz8.82059$vF6.2440100@news2.tin.it>
Message-ID: <jCzz8.17994$xy4.10082@atlpnn01.usenetserver.com>

"Alex Martelli" <aleax at aleax.it> wrote ...
> Thomas Heller wrote:
>
> > "Alex Martelli" <aleax at aleax.it> wrote ...
> >> Ralf Juengling wrote:
> >>         ...
> >> > BTW: It is not easy to add __iter__ to list's meta-type, is it?
> >>
> >> list's "metatype" (metaclass) is the built-in type named 'type'.
> >> like for all other built-in types and new-style classes.  Adding
> >> __iter__ to type would thus make almost 'everything' iterable.
> >>
> > Correction: adding __iter__ to type would make instances of
> > type iterable (these things are normally called 'classes').
>
> Classes (new-style) and built-in types, as I said.  Maybe I wasn't
> as clear as needed to express once again the distinction between
> levels -- instances of a metaclass are types (classes), so thanks
> for pointing it out.
>
> > Not too useful, it seems.
>
> Particularly given that any class or type (or other callable) that
> is callable without arguments can be used, in the same was as
> Ralg, the OP, wanted to use an "iterable list", by passing it
> in as the first argument to the two-argument form of iter.  The
> second arg need only be any object that is sure to differ from
> what calling the class/type/other callable can result in, to
> ensure the resulting iterator is unbounded.
>

OK, let me say right at the outset that I admit this posting is a deliberate
contribute to this week's "lighten up" theme. Ignore the fact that I type on
a desk that was only last night moved down from the second floor office.
Ignore the twenty azaleas I planted last weekend. Clearly I'm in need of a
life, so to prove I have one please accept the following in the (not
entirely serious) spirit intended. Though I have followed this thread, I
only decided to finally post when I realised I had an opportunity to write
...

Alex:

This is true, but remember that Ralg clearly stated he was being finical.

What you say appears to mean is that iter(type, None), for example, should
return an unbounded iterator of type. I'd be interested in what you have to
say about this:

>>> int
<type 'int'>
>>> int()
0
>>> int(1)
1
>>> iter(int, None)
<callable-iterator object at 0x00779F58>
>>> x = iter(int, None)
>>> x
<callable-iterator object at 0x0077CC98>
>>> x()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: 'callable-iterator' object is not callable
>>> dir(x)
['__class__', '__delattr__', '__getattribute__', '__hash__', '__init__',
'__iter
__', '__new__', '__reduce__', '__repr__', '__setattr__', '__str__', 'next']

LOC counters might like to observe the same error message by entering

>>> iter(int, None)()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: 'callable-iterator' object is not callable
>>>

if-you-can't-call-a-callable-iterator-why-is-it-
        -called-a-callable-iterator-ly y'rs  - steve
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/





From robinjim at earthlink.net  Sun Apr  7 08:57:23 2002
From: robinjim at earthlink.net (robin and jim)
Date: Sun, 07 Apr 2002 12:57:23 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <0rsr8.1752$CU.3728@rwcrnsc51.ops.asp.att.net> <a8odr2$rdh$0@216.39.172.122>
Message-ID: <TmXr8.22750$nt1.1846121@newsread2.prod.itd.earthlink.net>

You understood exactly what I was hoping for.

 Thanks for your posting.


"Bengt Richter" <bokr at oz.net> wrote in message
news:a8odr2$rdh$0 at 216.39.172.122...
> On Sat, 06 Apr 2002 01:45:32 GMT, "Emile van Sebille" <emile at fenx.com>
wrote:
>
> >"robin and jim" <robinjim at earthlink.net>
> >> Is there a way to convert a single quoted string to a triple quoted
> >string?
> >>
> >> For example, given:
> >>
> >> ''' abc %d xyz '''
> >>
> >> the result of ''' abc %d xyz''' % 10 is:
> >>
> >> ' abc 10 xyz '
> >>
> >> which is a single quoted string.
> >>
> >> I would like to convert the result to a triple-quoted string.
> >>
> >>
> >
> >What do you want to do with it?  Once a string is created, it doesn't
> >matter which style quoting was used -- it's simply a string.
> >
> True, but notice:
>
>  >>> """ ' """
>  " ' "
>  >>> """ " """
>  ' " '
>  >>> """ ' " """
>  ' \' " '
>  >>>
>
> perhaps he would have liked """ ' " """ back instead of ' \' " '
> Seems like just as valid as switching the other quotes, e.g., a quick hack
> to simulate extending the quote-choosing of repr:
>
>  >>> def repr3(x):
>  ...     rep = repr(x)
>  ...     if rep.find('\\n')<0 and rep[1:-1].find("\\'")<0: return rep
>  ...     rep=rep[1:-1].replace('\\n','\n')
>  ...     rep=rep.replace("\\'","'")
>  ...     return '"""' + rep + '"""'
>  ...
>  >>> tests = [""" ' """, """ " """, """ ' " """, """Line 1
>  ... Line 2"""]
>  >>> for t in tests: print 'plain: %s\n repr: %s\nrepr3: %s\n'% (t,
repr(t), repr3(t))
>  ...
>  plain:  '
>   repr: " ' "
>  repr3: " ' "
>
>  plain:  "
>   repr: ' " '
>  repr3: ' " '
>
>  plain:  ' "
>   repr: ' \' " '
>  repr3: """ ' " """
>
>  plain: Line 1
>  Line 2
>   repr: 'Line 1\nLine 2'
>  repr3: """Line 1
>  Line 2"""
>
> Regards,
> Bengt Richter
>




From duncan at NOSPAMrcp.co.uk  Mon Apr 15 08:48:28 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 15 Apr 2002 12:48:28 +0000 (UTC)
Subject: Questions on the C Thread API
References: <mailman.1018871909.16619.python-list@python.org>
Message-ID: <Xns91F18C58A3AAAduncanrcpcouk@127.0.0.1>

Phil Thompson <phil at river-bank.demon.co.uk> wrote in 
news:mailman.1018871909.16619.python-list at python.org:

> Currently, any C++ method that might block effectively has a call to
> PyEval_SaveThread() before it and a call to PyEval_ReleaseThread() after
> it. Any C++ virtual method that has been re-implemented in Python first
> checks to see if it already has the global lock, and acquires it if it
> doesn't. The Python method is then evaluated, and the global lock
> released if it had previously been acquired. The check if the lock has
> already been acquired is necessary because I only release the lock on a
> small subset of C++ methods, ie. those that might block, which I do for
> reasons of efficiency.

PyEval_SaveThread and PyEval_ReleaseThread both release the 
interpreter lock. Did you mean PyEval_RestoreThread for the second one of 
these?

Anyway, when you are going to do something that might block you save the 
thread state and restore it later. I presume you save the state in some 
thread local storage because you could have several threads all suspended 
in this way.
> 
> All the above works fine, so long as all the C++ calls are made from the
> same thread. With the support of the Qt thread classes, that assumption
> is no longer valid.
> 
> Therefore, just before calling the Python code that re-implements a C++
> virtual method, the global lock has to be acquired, and the correct
> thread state restored. However, to prevent a deadlock the thread state
> can only be restored if it is not the current one, but there is no way
> to get the value of the current thread state without having acquired the
> global lock. Catch 22.

Get the saved thread state for the current thread. If you don't have one 
then by definition you already have the global lock. If you have one then 
call PyEval_RestoreThread to get the lock and set the state and clear your 
saved state to NULL. Remember whether or not you had a thread state around 
the callback and only save the state/release the lock if you previously 
restored a state.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From logiplexsoftware at earthlink.net  Wed Apr 10 17:46:00 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 14:46:00 -0700
Subject: Extracting from a list
In-Reply-To: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <20020410144600.27c850c8.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 16:35:10 -0400
Joel Bender wrote:

> Hi,
> 
> 
> I would like to process the elements of a list that match a condition, 
> then remove them.  I'll use integers in my example, but my list items 
> are actually objects.
> 
> What's the "right" way to do this?  For example:
> 
>     >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>     >>> dothese = [i for i in lst if i < 5]
>     >>> lst = [i for i in lst if i >= 5]
>     >>> for item in dothese:
>     ...     print item
> 
> I'd rather not scan the list twice, particularly if there's nothing to 
> do.  I'd rather not build another double-linked list.
> 
> Is there something like this?
> 
>     >>> for item in lst:
>     ...     if (item < 5):
>     ...         print item
>     ...         del lst[ xxxx ]
> 
> But the "trick" is figuring out what xxxx should be.  Using lst.remove() 
> would work if the internals used "is" and not "is equal to" and the 
> removal didn't modify the list where it would break the loop somehow.
> 
> I could rebuild the list for things that should not be processed:
> 
>     >>> newlst = []
>     >>> for item in lst:
>     ...     if (item < 5):
>     ...         print item
>     ...     else
>     ...         newlst.append(item)
>     >>> lst = newlst
> 
> But these seems like a lot of work when there's nothing in the list to 
> process.

How about:

lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
for item in lst:
    if item < 5:
        print item
lst = [i for i in lst if i >= 5]

This way you only process the list twice and don't need a redundant copy.  If
you are really concerned about processing the list the second time when there's
nothing to be done, you could use a flag:

lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
flag = 0
for item in lst:
    if item < 5:
        print item
        flag = 1
if flag:
    lst = [i for i in lst if i >= 5]

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From robin at jessikat.fsnet.co.uk  Wed Apr  3 07:06:26 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 3 Apr 2002 13:06:26 +0100
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <3CAA335F.4A17F6FF@alcyone.com> <uakogq1r0i0524@news.supernews.com>
 <7xpu1h4h1m.fsf@ruckus.brouhaha.com> <a9xq8.53684$S52.1860202@news2.tin.it>
Message-ID: <U73UpVACBvq8EwIz@jessikat.fsnet.co.uk>

In article <a9xq8.53684$S52.1860202 at news2.tin.it>, Alex Martelli
<aleax at aleax.it> writes
....
>constraint on a variable -- that's part of why Python is
>so GREAT.  I hope Python stays pythonic... there are so
>many languages which appear to be mostly about imposing
>constraints ("expressing constraints in the code"), why can't
>we keep ONE language whose point is instead that of
>letting us be happily productive...?-)
>
>
>Alex
>
The great designer has several times expressed his desire to cripple
python's more dynamic features. Many recent changes have been in the
wrong direction eg the restriction on imports in functions etc. 

I believe the fundamental choice here is more about style than substance
as it's clear this addition adds little and effects much. Of course
there are similarities to the 'only one way to do things' syndrome in
this proposal. 

If we slavishly give people dictator status we get dictators. GvR
designed a reasonable language and it has improved up to now. I'm fairly
sure adding this will not improve things.

GvR asked for opinions and got them. The most articulate and compelling
are against. Well we got told that these are FUD etc etc. My gut feeling
is that this is wrong and that lambda, map, etc are good. Perhaps I
should eat sour grapes for breakfast.
-- 
Robin Becker


From huaiyu at gauss.almadan.ibm.com  Wed Apr 10 20:51:45 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 11 Apr 2002 00:51:45 +0000 (UTC)
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <3CB49ADD.89B242A4@sympatico.ca> <a92c4f$ne7$0@216.39.172.122>
Message-ID: <slrnab9ngr.i8b.huaiyu@gauss.almadan.ibm.com>

On 10 Apr 2002 21:51:11 GMT, Bengt Richter <bokr at oz.net> wrote:
>On Wed, 10 Apr 2002 16:04:45 -0400, "Colin J. Williams" <cjw at sympatico.ca> wrote:
>[...html...]
>
>Logical operations 
>0 or  None, 1 or  None: None 1 
>0 and None, 1 and None: 0 None 
>              not None: 1 
>
>With a three value scheme, the results would 
>be the same for the first two, whereas the 
>result of `not None` would itself be None. 

Thanks for translating this beautiful proposal from the ugly HTML.  :-)

I had also proposed this as a way to properly define rich comparisons:

	a   b   a > b   a == b   a < b
	1   0   True    False    False
	0   0   False   True     False
	0   1   False   False    True
	0   1j  None    False    None
	
It is often necessary to compare objects in a set even though there is only
a partial order.  We often do not want exceptions when the order is not
defined, since comparisons are often imbedded in expressions but try/except
has to be statements.

On the other hand, a comparison operator with a total order is useful in
many situations even though this total order is not related to the _natural_
partial order.   For example, cmp(1, None) is well defined.  Why would
cmp(1, 1j) be an exception?   This is especially painful if all you want is
just a _unique_ order, whatever it is.

Anyone paying attention to such matters over time would discover that
allowing three valued logic would solve a supprisingly large number of
trivial but nasty problems.

However, I do not see an easy way to extract None from Python's pool of
objects denoting false.  If only it was not designed that way ...

Hopefully when IEEE floating point support becomes ubiquitous we could
redefine the comparison with NaN for this purpose.  (You might be able to
get at it now as NaN=(1e200j**2).imag, but the comparison is always False.)

Huaiyu


From philh at comuno.freeserve.co.uk  Mon Apr  8 06:45:54 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 8 Apr 2002 11:45:54 +0100
Subject: Lex and Yacc for Python
References: <slrnab19hf.23u.Alex@Tron.ddeus.lan> <m3vgb3s1tv.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnab2t72.2ot.philh@comuno.freeserve.co.uk>

On 07 Apr 2002 22:14:52 +0200, Martin v. Loewis <martin at v.loewis.de> wrote:
>Oleksandr Kozachuk <Alex at DDeus.de> writes:
>
>> I need something, comparable to Yacc/Lex. Or something for
>> Context-Free-Grammar. Can any one recommend me anything?
>
>I recommend YAPPS if you need a simple solutions to a simple problem,
>and SPARK if you need a complex solution to a complex problem. See
>
>http://www.python.org/sigs/parser-sig/

I too would recommend SPARK.

Never used YAPPS, so won't comment about it.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From aleax at aleax.it  Sat Apr  6 04:55:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 06 Apr 2002 09:55:24 GMT
Subject: Argument against iterable ints...
References: <slrnaasnpk.178.mlh@vier.idi.ntnu.no>
Message-ID: <gCzr8.54623$pT1.1549029@news1.tin.it>

Magnus Lie Hetland wrote:

> Just a situation that cropped up recently: While writing a little
> generator for flattening nested iterables, I used
> 
>   try: iter(foo)
>   except: return foo
>   else:
>       ....
> 
> (of course I could have used a for loop directly in the try clause --
> that doesn't change anything.)

It consumes the first item of foo if foo is an iterator, and thus:

try: for __ in foo: break
except: return foo
else:
    ...

has subtly different semantics from your code above.  I agree this
does not affect your main point (on which I agree with you).


Alex



From max at alcyone.com  Thu Apr  4 00:21:03 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 21:21:03 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <Pine.LNX.4.33.0204031648350.19271-100000@server1.lfw.org> <mailman.1017893479.16809.python-list@python.org>
Message-ID: <3CABE2BF.3A9D74B7@alcyone.com>

Andrew Koenig wrote:

> I would be very surprised if none of the Lisp family of languages
> has similar anomalies in the relationship between eq and equal.

Well, in Scheme it's left unspecified whether #f (the canonical false
value) is identical to () (the empty list).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From borcis at geneva-link.ch  Tue Apr  9 08:16:25 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Tue, 09 Apr 2002 14:16:25 +0200
Subject: [OT] Re: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>
Message-ID: <3CB2DB99.4080400@geneva-link.ch>

Magnus Lie Hetland wrote:


> 
>>python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
> 
> Hm. What programming language is this supposed to be?


    PythonWin 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] [...]

    >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
    'BUSH'


What's your problem ? Shows gregarious reflexes are easily captured

with simple models, and Python's worth as a modelling language in war 
and human rights related matters - just as has been demonstrated by the 
professor who used python to track war crime statistics in Kosovo.

The "python >>>" that prefixes the expression is there to nudge readers 
in less python-related contexts to instantiate the adequate shell to its 
prompt. (say, a redhat user who just knows python by name).

Regards, Boris
--
python >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")



From mlh at vier.idi.ntnu.no  Fri Apr  5 22:20:40 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 03:20:40 +0000 (UTC)
Subject: A faster Python?, Python compiler, Dylan,...
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1017935351.5088.python-list@python.org>
Message-ID: <slrnaasqc7.178.mlh@vier.idi.ntnu.no>

In article <mailman.1017935351.5088.python-list at python.org>, eric
wrote:
[snip]
>If this is the only criticism, than we're doing quite well. :-)

How about including some code from Python2C (or similar projects)...
Then you could write the code in the inline string with (possibly
restricted) Python syntax, have it translated to C and then inlined...
Then one *wouldn't* have to write things twice, and the criticism
disappears ;)

(Actually, I think it would be quite cool...)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From johnroth at ameritech.net  Tue Apr 30 14:36:56 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 30 Apr 2002 14:36:56 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de> <Xns92008AC5FF09Aduncanrcpcouk@127.0.0.1> <mailman.1020172854.26653.python-list@python.org> <Xns9200AC87B7A86duncanrcpcouk@127.0.0.1>
Message-ID: <uctp0ugohd6ib1@news.supernews.com>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
news:Xns9200AC87B7A86duncanrcpcouk at 127.0.0.1...
> Roman Neuhauser <neuhauser at mail.cz> wrote in
> news:mailman.1020172854.26653.python-list at python.org:
>
> >> I prefer writing
> >>    str.join(' ', aList)
> >> rather than
> >>    ' '.join(aList)
> >>
> >> When the string is a literal I find the former reads much better.
YMMV
> >
> >     IMO that's because join being a string method is just awkward.
How
> >     does this read:
> >     aList.join(' ')
> >     After all, you tell the list to give you it's concatenation,
using
> >     the argument as the glue.
>
> That doesn't work if you want to join a tuple, or an iterator or any
other
> sequence. Dont forget you can even join a string:
> >>> print str.join(',', 'abc')
> a,b,c

Doesn't matter. What we need is <sequence>.join(<string>)
where the resulting string is the result of applying str() to each
of the members of the sequence, with the <string> between
each pair of members. In other words, .join() is a member
function of all objects that have a sequence interface.

John Roth





From dalke at dalkescientific.com  Tue Apr 23 17:55:31 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 23 Apr 2002 15:55:31 -0600
Subject: String handling bug in Python
References: <3cc5d5c8$1@news.nz.asiaonline.net>
Message-ID: <aa4lif$i6h$1@slb6.atl.mindspring.net>

Colin Brown:
>>>> r'\'
>  File "<string>", line 1
>     r'\'
>        ^
>SyntaxError: invalid token
>>>>
>
>It appears as if the r option for strings does not handle a backslash as
the
>last string character properly.

That's the documented behaviour for raw strings, down to the exact
example you gave.  Why do you expect it to be otherwise?

http://python.org/doc/current/ref/strings.html
] When an `r' or `R' prefix is present, a character following a
] backslash is included in the string without change, and all backslashes
] are left in the string. For example, the string literal r"\n" consists
] of two characters: a backslash and a lowercase `n'. String quotes can
] be escaped with a backslash, but the backslash remains in the string;
] for example, r"\"" is a valid string literal consisting of two characters:
] a backslash and a double quote; r"\" is not a valid string literal (even
] a raw string cannot end in an odd number of backslashes).

                    Andrew
                    dalke at dalkescientific.com





From ykingma at accessforall.nl  Tue Apr  2 17:02:08 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Tue, 02 Apr 2002 23:02:08 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>  <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org>
Message-ID: <3CAA2A44.37BB218F@accessforall.nl>

Laura,

<snip>

> 
>         4) Should we strive to eliminate non-Boolean operations on bools
>            in the future, through suitable warnings, so that e.g. True+1
>            would eventually (e.g. in Python 3000 be illegal).  Personally,
>            I think we shouldn't; 28+isleap(y) seems totally reasonable to
>            me.
> 
> This is not an argument for allowing non-Boolean operations on bools();
> this is an argument for not writing functions that return Booleans.  Make
> them return numbers instead, so that you can use them as you did.  

I couldn't agree more.
The problem is that a lot of code has been written assuming 1/0 behaviour
of boolean values.

If there were a real boolean type in Python this backward
compatible behaviour can be implemented in the __add__, __radd__
and other special methods of the numeric classes (see the language reference on special
method names: http://www.python.org/doc/current/ref/specialnames.html).
They would work by using the value 1 or 0 when encountering a boolean value.

<snip>


>     Rationale
> 
>         Most languages eventually grow a Boolean type; even C99 (the new
>         and improved C standard, not yet widely adopted) has one.
> 
>         Many programmers apparently feel the need for a Boolean type; most
>         Python documentation contains a bit of an apology for the absence
>         of a Boolean type.
> 
> So fix the docs, don't change the code! <wink>.  I think the fact that
> in python control flow structures distinguish between Something and
> Nothing is one of the beauties and glories of the language, and you
> should delete any documentation that says otherwise.

The documentation for this is in the library reference, not in the language
reference.

> 
>         I've seen lots of modules that defined
>         constants "False=0" and "True=1" (or similar) at the top and used
>         those.  The problem with this is that everybody does it
>         differently.  For example, should you use "FALSE", "false",
>         "False", "F" or even "f"?  And should false be the value zero or
>         None, or perhaps a truth value of a different type that will print
>         as "true" or "false"?  Adding a standard bool type to the language
>         resolves those issues.
> 
> So would adding True and False to the __builtins__, and probably
> operator.truth as well, and then modifying PEP 8 saying to use the
> things if you actually have a need for True and False.  Then you could
> also get a much needed word in edgewise discouraging
> 
> if bool(x) == True:
> 
> or actually using True and False much, because there is usually a better
> more pythonic way to do what people used to other languages are accustomed
> to doing with booleans.  This is precisely what some people have said
> here: 'When I started using Python, I made True and False, but once I
> stopped trying to program in some other language using python, I
> stopped needing these things'.  (see the recent post by Don Garrett
> <garrett at bgb-consulting.com> in this thread for an example.)
> 
> This is what I have observed as well.  And I fear if you add these new
> types to the langauge people will never take this step.  The existence
> of the types in the language will discourage them from thinking that using
> True and False all the time is not pythonic.  It is nice that people are
> puzzled, wondering how all those python programmers live without a
> boolean type.  Eventually they puzzle it out.  This is not a bug, but
> a feature <wink>.

Thanks for putting this so explicitly.

This Something/Nothing feature is indeed very useful.
However, it is not part of the python language, it is part of the way in which
the standard types are implemented.

Whenever you derive a boolean from, ie. call __len__() or __nonzero() on, objects
of the standard types they act as Something/Nothing objects which allows one
to think like:

if something_there:
    ...
else: # nothing_there
    ...

A real boolean is required to control the conditional branch.
Since slices are also explicitly available, I'd like a real boolean
to be available. It fits the language.

> 
>         Here are some arguments derived from teaching Python.  When
>         showing people comparison operators etc. in the interactive shell,
> 
>         I think this is a bit ugly:
>             >>> a = 13
>             >>> b = 12
>             >>> a > b
> 
>             1
> 
>         If this was:
>             >>> a > b
> 
>             True
> 
>         it would require one millisecond less thinking each time a 0 or 1
>         was printed.
> 
> This is the basis of our disagreement.  I think that it is very, very
> important that much more than a millisecond be spent on this.  This is a
> fundamental python concept, which I want to teach.  This is precisely
> where you learn that python distinguishes between Something and Nothing,
> and if you have a problem seeing why this implies  a > b printing as 1,
> then you probably have a problem with the whole concept.  And making it
> return True is precisely what I never, ever, ever want a python learner
> to see.
> 

As I said, it's not in the language, it's in the standard types.
The millisecond is spent on the implicit derivation of a boolean
in the readers' head. The  interpreter does the same thing, although faster.

> 
> I think that a really strict boolean might be nice to have.  And if you
> ever write one, you will curse the day you let this hack be named bool.
> Replacing this int-with-a-hat with real bools will break so much
> code.
> 

After having my head set straight on this by Martin v. Loewis
(see the thread 'but not as int subtype') I concluded in favour
of PEP 285, with the remark that what can in the future
not be implemented using the special methods of bool and current
standard types should be deprecated at the same time the PEP
gets into force. The only example of such code I know of
is user implementations of __getitem__, ie. objects
indexable with square brackets, that are actually indexed with bool arguments.
When all such user code has been fixed, there is nothing in the way
of a real boolean type that is not an int-in-new-hat.


> I don't want to have to teach the pythonic meaning of True and False to
> people who already believe they know what True and False is.  This is
> going to be hell.  It is about as hard a thing as there is in teaching,
> getting around people's previous conceptions.

Starting with only the Something/Nothing background it would indeed be
problematic to teach about True/False. However an implied derivation of
of True/False from Something/Nothing is no problem, eg. 'if 1:' is easy to
teach.

In fact this could be seen as a good example of the usefulness of
object orientation: it allows separation of concerns to the extent
that the standard types are documented separately from the language.

<snip>

Regards,
Ype

-- 
email at xs4all.nl


From dmmcf at uiuc.edu  Mon Apr 15 15:18:21 2002
From: dmmcf at uiuc.edu (D. Michael McFarland)
Date: Mon, 15 Apr 2002 14:18:21 -0500
Subject: Assistance sought with mondo numerical app
Message-ID: <86hemc3h4y.fsf@haggis.aae.uiuc.edu>

Dear Fellow Pythonistas,

    A friend of mine is in the throes of finishing his
Ph.D. dissertation, comparing experimental and numerical results for
material damage caused by grinding.  For the numerical part, he's
using a finite element library I wrote in Python.  This library made
programming the tensor operations much easier than they might have
been in, say, Fortran, but the end result takes rather longer to run
than one could hope--about a day per problem case on a PII-450.  He
has two machines of this class available, and I have two, and that
might just be enough to let him make his deadline, but it occurred to
me where to find more machines: this newsgroup.

    In short, we want your flops.

    I'm writing in the hope that a few people with CPU time to spare
over the next couple of weeks will agree to run a case or two each.
Volunteers would need to be running a fairly recent Python and the
Numerical module, and install Neil Schemenauer's sparsemodule from

http://python.ca/nas/python/sparsemodule/

The output files are large, so some bandwidth is needed.  Unix or
something like it is the only OS we're prepared to run on, unless
someone wants to try to make our code run under MS Windows (I wouldn't
know where to begin, myself, but it might not be too difficult).

    I'm open to suggestions, but the simplest strategy seems to be to
mail source code, problem input files and a Makefile to control the
whole mess to each participant, with results to be FTPed back.

    Incidentally, we have always meant to release this code once it
was polished, and we still plan to do so.  But first this guy needs to
graduate, for which he needs to crunch a _lot_ of numbers.  He's
swamped with related work, and I feel like I got him into this, so I'm
appealing to the Python community on his behalf.  Any offers of
assistance or other advice will be gratefully received.

                              Best regards,
                              Michael

-- 
    D. Michael McFarland                      dmmcf at uiuc.edu
    Visiting Senior Research Scientist & Visiting Lecturer
    Department of Aeronautical and Astronautical Engineering
    University of Illinois at Urbana-Champaign


From tdickenson at devmail.geminidataloggers.co.uk  Mon Apr  8 06:42:59 2002
From: tdickenson at devmail.geminidataloggers.co.uk (Toby Dickenson)
Date: Mon, 08 Apr 2002 11:42:59 +0100
Subject: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> <marklists@mceahern.com>  of "Sat, 06 Apr 2002 11:57:07 MDT." <NCBBLFCOHHDIKCAFGCFBAEHKKJAA.marklists@mceahern.com> <mailman.1018118665.30693.python-list@python.org>
Message-ID: <lvs2bu81ti1og2p9v79p4vpu2k23vfk5rm@4ax.com>

Laura Creighton <lac at strakt.com> wrote:

>Your boss comes in and says, oh, Mark, we need a change to that
>isValid function.  We need to return 3 states, Valid, Invalid, and
>Valid-but-You-Don't-Have-Access-Because-You-Didn't-Pay-Your-Bill.

What objects do you return for those three states? You didnt say, but
from your previous posts I would guess you were intending something
like invalid: 0, valid: 1, unpaid:  2

Changing the old two-values function into this new three-valued
function would be a disservice to any code that previously used
"if user.isValid():" but wants to treat unpaid users like invalid
ones.

>If you don't need a third value, somebody 
>else will, later. 

You have to pray that, when they do, the third state is a subdivision
of the original 1/true/something state, rather than a subdivision of
the old 0/false/nothing state.

Toby Dickenson
tdickenson at geminidataloggers.com


From mkc+dated+1021521407.f909ec at mathdogs.com  Tue Apr 16 00:01:04 2002
From: mkc+dated+1021521407.f909ec at mathdogs.com (Mike Coleman)
Date: 15 Apr 2002 23:01:04 -0500
Subject: Why don't strings share data in Python?
Message-ID: <87d6x09trz.fsf@mathdogs.com>

Does anyone know why strings (i.e., those of length >1) don't share their data
in Python?  Since their immutable, it seems like this would be the obvious
thing to do.  So, for example, the space behavior of this code could be linear
rather than quadratic/horrific:

d = {}
for i in xrange(100000000):
    d[mybigstring[i:]] = mybigstring[i:]


Mike


From stojek at part-gmbh.de  Wed Apr 10 15:47:38 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Wed, 10 Apr 2002 19:47:38 GMT
Subject: mac-adress with python (windows)
Message-ID: <3cb493ed.40186328@news.easynews.net>

Hi,
following some links from this and other groups I had to
fight through pages and pages of C code and loooong 
threads. Then I found the following example in my
win32com docu:

>>>import pythoncom
>>>print pythoncom.CreateGuid()

<iid:{ECD43800-4CB9-11D6-ABD9-0050DA4545E7}>

with the last six bytes being the ethernet card ID.

Can anybody tell me if this works for all Window versions?
(I can only test for Win NT 4 SP6)

Thanks,
Marcus


From shalehperry at attbi.com  Wed Apr 17 09:44:39 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 17 Apr 2002 06:44:39 -0700 (PDT)
Subject: Counterintuitive Python behavior
In-Reply-To: <3CBD7406.3050909@mxm.dk>
Message-ID: <XFMail.20020417064439.shalehperry@attbi.com>

On 17-Apr-2002 Max M wrote:
> dominikush at yahoo.com wrote:
> 
>> Who is wrong here: my intuition or Python (2.2)? If it's
>> my intuition, how can I train my thinking about Python's
>> execution model, so that my intuition get's better ;-)
> 
> Your intuition!
> 
> Objects like lists gets copied by reference. So Python behaves exactly 
> like it should.
> 

Or from another side:

a = [...]
b = a

b[n] = ...

and a gets updated.  values() returns the contents of the dictionary so you are
seeing the same type of behaviour.




From ark at research.att.com  Tue Apr  9 16:56:52 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 9 Apr 2002 16:56:52 -0400 (EDT)
Subject: binding a reference to a variable
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com>
References: <NCBBLFCOHHDIKCAFGCFBGEBOKKAA.marklists@mceahern.com>
Message-ID: <200204092056.g39Kuq404692@europa.research.att.com>

Mark> [Andrew Koenig]

>> What I would really like is to be able to define a function called,
>> say, `set' such that set(x, 42) has the same effect as `x = 42'.

Mark> What about this?

Mark>   def set(name, value):
Mark>     globals()[name] = value

Mark>   set("x", 5)

Mark>   print x

Mark> The only difference is that you pass set a string representation of the
Mark> variable you want to set.  Close, but no cigar?

Close, but no cigar.  A few problems:

       It's restricted to global variables.

       It uses the relationship between variables and their names
       at run time, and I would prefer the mapping to be done at
       compile time.

Mark> There's also this recipe:

Mark>   http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/80373

Same problem

Mark> I'm curious why you need this?

I want to be able to stuff an object into a larger data
structure and have that object reach out into a variable and leave a
trail of breadcrumbs.




      





From mgilfix at eecs.tufts.edu  Fri Apr 19 18:59:12 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 19 Apr 2002 18:59:12 -0400
Subject: Pure python crypto
In-Reply-To: <20020417021703.H16954@eecs.tufts.edu>; from mgilfix@eecs.tufts.edu on Wed, Apr 17, 2002 at 02:17:03AM -0400
References: <20020417021703.H16954@eecs.tufts.edu>
Message-ID: <20020419185912.G29377@eecs.tufts.edu>

  For those of you who were interested, the pure python blowfish
module can be found here: http://bofh.concordia.ca/blowfish.py.

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From flognat at flognat.myip.org  Wed Apr 17 08:39:34 2002
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Wed, 17 Apr 2002 12:39:34 GMT
Subject: Anyone into win32com IE, fetching a picture.. 
Message-ID: <m38z7mv6rd.fsf@flognat.myip.org>

Hello!

Sorry for asking probably a more windows-like question here, than on
windows group.. but this is my home.. my.. well you know.. 

Basically I am using win32com and InternetExplorer to fetch pages, and
I would also like to fetch the pr0n^H^H^H^H ehm.. snake pictures, yeah
snake pictures from the pages and store them.. 

ie = win32com.client.Dispatch('InternetExplorer.Application')
ie.Navigate(url)
doc = ie.Document
while doc.readyState != 'complete': time.sleep(0.1)

Well now doc contains a bunch of info about the stuff in the url, if
the url was an image, I can get the size, name and so on, but I
haven't managed to get my hands on the data-part of the image.. Any
hints??

Yeah I know.. go urllib2 or similar.. but but.. now (using win32com) I
have cookies, https and so on working good out of the box..

Anyone who recognizes??

Anyone who have a good http/url library that can handle this?

        /Andy

-- 
 The eye of the beholder rests on the beauty!


From martin at v.loewis.de  Wed Apr 17 18:27:37 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Apr 2002 00:27:37 +0200
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de> <m3g01um27o.fsf@ferret.object-craft.com.au>
Message-ID: <m3ads2dkpy.fsf@mira.informatik.hu-berlin.de>

Dave Cole <djc at object-craft.com.au> writes:

> If you create a new thread state for the callback, what will happen to
> Python exceptions raised in the callback?

They propagate back to the bottom of the callback. If you don't treat
them special, deleting the thread state will also release the
exception object and traceback object. Assuming the thread callback
came "out of nowhere", there is nothing reasonable to do, anyway: you
might want to print the exception to stderr or something like that.

Regards,
Martin



From noah at noah.org  Thu Apr 11 16:56:49 2002
From: noah at noah.org (Noah)
Date: Thu, 11 Apr 2002 13:56:49 -0700
Subject: Nonblocking IO on popen pipe (isn't nonblocking)
In-Reply-To: <a94oq5$m06$1@nntp6.u.washington.edu>
Message-ID: <PAENINBOOODFDEPOKPKCEENHCKAA.noah@noah.org>

Hmm... This sounds suspicious... If FTP won't write its buffered
output then no application could see it -- that includes the shell!
Python shouldn't be any different in what it sees.
If I start FTP from the shell, then how does FTP know 
when to flush the buffer so that you can see it in a shell?
Shells essentially do a fork/exec to start an application.
This is no different than what I should be able to do in Python.
What I'm saying is that if FTP buffers the output and never
flushes it then there would be no way for the shell to see
it either. In fact, we see this all the time with newbie
Python applications that print in a loop, but never flush.
New users wonder why they can't see their output until the 
script is finished.

Yours,
Noah
-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Donn Cave
Sent: Thursday, April 11, 2002 12:40 PM
To: python-list at python.org
Subject: RE: Nonblocking IO on popen pipe (isn't nonblocking)


Quoth "Noah" <noah at noah.org>:
| Damn... The only problem I see with this explaination is
....
If you can explain how you do it in Perl, maybe someone can
tell you what you need in Python.  The buffering problem I was
talking about really has nothing to Python, it would be in ftp.
If ftp won't write its buffered output, then Perl, Python and C
are all in the same boat.

	Donn Cave, donn at u.washington.edu





From none at none.com  Tue Apr  9 12:00:29 2002
From: none at none.com (John)
Date: Tue, 9 Apr 2002 12:00:29 -0400
Subject: big ints & elliptic curves
Message-ID: <a8v338$4sl$1@watserv3.uwaterloo.ca>

Is there any good library for multi-precision arithmetics in Python? Also
looking for elliptic curve implementations.
Thanks




From synthespian at uol.com.br  Sat Apr 27 23:32:52 2002
From: synthespian at uol.com.br (synthespian)
Date: Sun, 28 Apr 2002 00:32:52 -0300
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <CmUw8.85199$XV5.7694212@bin4.nnrp.aus1.giganews.com> <3CC449BF.8030807@alcyone.com>
Message-ID: <pan.2002.04.28.00.32.39.955778.3624@uol.com.br>

On Mon, 22 Apr 2002 14:34:55 -0300, Erik Max Francis wrote:

> Brett g Porter wrote:
> 
>> In what ways are Scheme and Lisp only distantly related?
> 
> Probably in that the original poster likes Lisp (I presume he means
> Common Lisp) and dislikes Scheme, and wants to distance them as much as
> possible.
> 
> "Lisp" is generally the term applied to the broad category of Lisp-like
> languages, of which Scheme is most definitely an example.  Scheme is
> indeed one of the closer relatives to Common Lisp.
> 

	You go now and post that opinion on comp.lang.lisp and you're going to 
get beaten up till you drop.
	For one thing LISP is a huge language, with various programming
paradigms, and Scheme insists on being small and functional - what the
hell for, nobody at comp.lang.lisp seems to know.
	
	Cheers
	H
	


 
--------------------------------------------------------
_________   1001101010    | _________   Blah, blah,
| ^   ^ |  /1111100110    | | ^   ^ |  /blah, blah.
| O   O | / 0010100111    | | _   _ | /
|   ^   |/                | |   ^   |/
| ----- |                 | |  ---  |
\_______/                 | \_______/
                          |
________________________________________________________
Micro$oft-Free Human         100% Debian GNU/Linux
     KMFMS              "Bring the genome to the people!"


From tanzer at swing.co.at  Tue Apr  2 12:11:22 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Tue, 02 Apr 2002 19:11:22 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: Your message of "01 Apr 2002 17:57:45 +0200."
             <m3g02f76o6.fsf@mira.informatik.hu-berlin.de>
Message-ID: <E16sRoY-0003ux-00@tswings.swing.cluster>

martin at v.loewis.de (Martin v. Loewis) wrote:

> > 'if x == y:' should cause no warning, of course -- it's a general, perfectly
> > good polymorphic usage.  Why should the user have to silence warnings
> > or break the smoothness of polymorphism to ensure against x, or y, or
> > both being of this newfangled bool type?
>
> Indeed. I wonder how often the warning would trigger, though. I'd
> expect that people very rarely compare the result the got from a
> predicate to something else.

Except if one wants to compare the results of two or more predicates.
I'd expect such comparions to be not that rare -- I know that I
sometimes use such.

And Alex was talking about polymorphic usage -- by definition, you
can't know what types are compared there.

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




From aleax at aleax.it  Mon Apr 29 05:53:33 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 09:53:33 GMT
Subject: Comments on Version 2, Draft Pep for Deprecating Builtins
References: <aaiu6e$pfr$1@bob.news.rcn.net>
Message-ID: <xK8z8.75479$vF6.2259175@news2.tin.it>

Raymond Hettinger wrote:
        ...
>     Q.  Why wasn't zip() included.
>     A.  It doesn't take and apply a function argument.  Also, it is used
>     in
>         non-functional programming to neatly code lockstep iteration.
> 
>     Q.  Why wasn't apply() or oct() included?
>     A.  I picked six candidates for relocation or elimination based on
>         whether I could make a case for them.  No doubt there will be

I don't get it.  Isn't apply the EASIEST built-in to show up as "useless
today"?  It DOES "take and apply a function argument" if that's what you're
targeting (a slant which would make it pretty hard to understand what pow
and divmod -- also, btw, used in programming that has nothing to do with
FP -- have to do with the rest).

Maybe there would be benefit in separating concerns:

A. what mechanism[s] could/should be used to move current builtins to
   other modules (possibly including a 'deprecated' one) without code
   breakage.  It seems to me that the .PTH mechanism gives possibly the
   smoothest road to that end, be it by adding 'from x import y, z' to
   the constructs it supports (my preference) or by exploiting its
   existing 'import x' support by providing suitable modules for 'x'

B. what if any current built-in functions that accept a function argument
   should be deprecated or otherwise moved.  Here it seems to me that
   the case is strongest for apply, weakest for callable (it COULD move
   to a new module intended for feature-resting, perhaps with some of
   operator's current contents -- but, should it?), and classmethod,
   staticmethod and property (all new shiny thingies), pretty good for
   filter (always well replaceable by a LC), not too bad for map (mostly
   replaceable by LC+zip), dubious for reduce (I _am_ surprised that
   reduce(operator.add, seq) now seems to be slower than a loop of
   for x in seq: result+=x, but still the reduce form might be said to
   be a more direct expression -- arguable, I admit).

C. what about current built-in functions that do not necessarily take a
   callable argument -- for each such function F, _is_ F important
   enough to need to be a built-in?  Why isn't, e.g., 'intern' in
   module sys, or other internals-related module -- how often does it
   get called, compared to uses of, e.g., sys.exit or sys.stderr?


Alex



From cliechti at gmx.net  Wed Apr  3 18:39:39 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 4 Apr 2002 00:39:39 +0100
Subject: How best to write this try/except block?
References: <a8fba8$ln8$1@panix2.panix.com> <Xns91E5CE0E5919Dcliechtigmxnet@62.2.16.82> <eppstein-055127.14142303042002@news.service.uci.edu>
Message-ID: <Xns91E66E671A60cliechtigmxnet@62.2.16.82>

David Eppstein <eppstein at ics.uci.edu> wrote in news:eppstein-
055127.14142303042002 at news.service.uci.edu:

> In article <Xns91E5CE0E5919Dcliechtigmxnet at 62.2.16.82>,
>  Chris Liechti <cliechti at gmx.net> wrote:
> 
>> if d.has_key(key):
>>      foo = stuff(d[key])
>> else:
>>      foo = otherStuff(key)
> 
> I've seen a couple responses suggesting this.
> 
> Is there a reason why in modern Python versions one should still use 
> has_key?  I like the syntax "if key in d" better.

i like it also but its still very new (py 2.2)...

[Python22/Doc/lib/typesmapping.html]

chris

-- 
Chris <cliechti at gmx.net>



From Paul.Lydon at uk.experian.com  Thu Apr 18 03:59:44 2002
From: Paul.Lydon at uk.experian.com (Lydon, Paul)
Date: Thu, 18 Apr 2002 08:59:44 +0100
Subject: pre-PEP for optional 'pass'
Message-ID: <5DABF7424840D411BE9600508B6621F841C033@exchukthis02.experian.co.uk>


> -----Original Message-----
> From: mjackson at wrc.xerox.com [mailto:mjackson at wrc.xerox.com]
> Sent: 18 April 2002 02:48
> To: python-list at python.org
> Subject: Re: pre-PEP for optional 'pass'
> 
> 
> Erik Max Francis <max at alcyone.com> writes:
> > Gareth McCaughan wrote:
> > 
> > > (These days there ought to be a computer scientist in the
> > > mix too. Suggestions?)
> > 
> > "I'll just write a program to exhaustively find them all.  What's a
> > prime number again?"
> 
> "My program halted with no output, so there must not be any."

Obviously what is needed is an "isPrime()" function - now should that return
True, true, Yes, 1 .......

-----------------
Paul Lydon 
PC Build Group
Experian


=======================================================================
Information in this email and any attachments are confidential, and may
not be copied or used by anyone other than the addressee, nor disclosed
to any third party without our permission.  There is no intention to
create any legally binding contract or other commitment through the use
of this email.

Experian Limited (registration number 653331).  
Registered office: Talbot House, Talbot Street, Nottingham NG1 5HF




From bsturk at news1.news.adelphia.net  Mon Apr  8 02:22:41 2002
From: bsturk at news1.news.adelphia.net (Brian Sturk)
Date: Mon, 08 Apr 2002 06:22:41 GMT
Subject: pty/select/timers/SIGINT ?s w/ embedded python in vim
Message-ID: <slrnab2f49.359.bsturk@telengard.sturk.net>

I've written a python script for vim that allows you to run a shell
within a buffer using the pty module.  I've got most of what I wanted
to working but there are still 2 issues remaining that make using
it somewhat annoying.  I've been trying to get these working for
quite a while and haven't made any inroads on them at all.  I'm
hoping someone might have an idea or two...

If any of this is too vague or I'm not making sense I can post
code snippets or, you can check out the script:

    http://nh.ultranet.com/~bsturk/programs/vimsh.tar.gz

1.  Reading output from slow commands (ftp) or commands that have constant
    output until interrupted (ping):

    I need a way to be able to read output in the background, but
    unfortunately I can't use threads.  I also couldn't find a timer
    type module ( I did find Timer, which seems to use threads ).

    Here's an example scenario...

    The vim user executes a command, for instance, ftp.

    I use select to see if there's any output and depending on how
    long it takes I may or may not get something.
    
    If my timeout to select is short I sometimes miss the data but
    commands that aren't slow aren't hung up on on their last read.
    
    If the timeout is greater I have a better chance of reading
    in data from slow commands but other commands appear sluggish
    when reading the last bits from stdout.

    Maybe I'm approaching this the wrong way.  I had contemplated doing
    some kind of process watching so I could be more intelligent about
    whether to be reading or not.  I've run out of ideas on this one.

    I've got a workaround for this by mapping a key to check for
    more output.

2.  SIGINT handling:

    I wrap executing a shell command w/ a try, except KeyboardException and
    correctly catch the exception when the user presses ctrl-c.  I then
    send whatever key sequence the user has set for EINTR out to the
    running process, so something like this:

        os.write( self.ind, intr_key )

    self.ind in this case is the file descriptor returned from
    pty.fork( ), and intr_key might be something like '^C'.

    This works in that the executed command is interrupted but also
    seems to trigger another KeyboardException in the python script. 
    Does this have something to do w/ process groups or something like
    that?  I would have thought that only the process that 'received' it would
    'get' it.  What this is keeping me from doing is reading any left
    over data, like ping's statistics or a new prompt.

    Again the workaround is to use the key mapping to check for more
    output.

    Any help on these would be greatly appreciated.

-- 
.---------------------------------------------------,---------.  
| Brian Sturk - http://www.nh.ultranet.com/~bsturk  \  C/C++  | .>   )\,^a__
|--------------------------|  bsturk at nh.ultranet.com | Python |(  _ _)/ /-." ~
| http://www.telengard.com `-------------------------`--------| `( )_ )/
| Telengard Technologies Inc. -  NT/*nix UI & device drivers  |_<_s_<_s
'-------------------------------------------------------------'


From noah at noah.org  Mon Apr 29 22:33:36 2002
From: noah at noah.org (Noah)
Date: Mon, 29 Apr 2002 19:33:36 -0700
Subject: PythonCOM syntax question
In-Reply-To: <B0000071394@buford.turtleprod.com>
Message-ID: <PAENINBOOODFDEPOKPKCEEBECNAA.noah@noah.org>

Probably you need to add .Value() like this:
	AccountNumber	= buffer.Field("AccountNum").Value()

One thing I find very helpful is to use the
'COM Makepy utility'. I launch it from PythonWin
under the 'Tools' menu. Afterwards you can use the
Interactive Window to manually create your objects.
PythonWin will then do autocompletion on your COM objects.
I find this very helpful to identify what
fields and methods Python can see.

So you might type
	buffer.
and as soon as you hit the '.' you would like a list
of fields and methods that you can access.

Yours,
Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Greg Pierce
Sent: Monday, April 29, 2002 12:43 PM
To: python-list at python.org
Subject: PythonCOM syntax question


I'm trying to interact with the Navision Axapta ERP package via it's COM connector...and I'm having trouble with particular types of
calls.  I'm hoping someone can help me if there's some syntax problem I'm having.

I'm able to create the Axapta COM objects and can call certain methods on them, but not others.  I'm no COM expert, but there is a
difference in the method specs that I can see.  Those that I _cannot_ call all have a "get" in the API documentation for Axapta --
like this:

HRESULT get_Field( [in] VARIANT field, [out, retval] VARIANT *pVal );

Those that I _can_ call don't -- like this:

HRESULT Call(  [in] BSTR bstrMethod, [in, optional] VARIANT v1,  [out, retval] VARIANT *pResult);

Is there some different syntax I should need to use in Python to access those "get" methods?

This code in VB works ( where "buffer" is an already created COM object) :

MyValue = buffer.Call( "openBalanceMST" )
AccountNumber	= buffer.Field("AccountNum")

But in Python, the "call" line works, and then the "field" line returns a "com_error: (-2147352573, 'Member not found.', None,
None)"

Any ideas?  Did that explanation make sense?

Thanks,

g.









From mats at laplaza.org  Fri Apr  5 12:56:05 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Fri, 05 Apr 2002 17:56:05 GMT
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com>
Message-ID: <3cadd841.225619243@news.laplaza.org>

On 5 Apr 2002 02:05:29 -0800, alanmk at hotmail.com (Alan Kennedy) wrote:

:brian d foy wrote:
:> for our April 2002 issue (April Fool!), The Perl Review replaced
:> its Perl content with Python articles, so Python folks might want
:> to read this issue:
:> 
:>    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf
:
:Forgot to mention: Does this mean that the Golf competition is open to
:Python entries :-)
:
:The golf competition is a neat idea. Python hasn't completely
:suppressed the hacker in me :-)


Ugh, it's horrid, and seems very perl-ish.  Do something in the
smallest amount of space, readability/maintainability be damned.

We need something that emphasizes the beauty, simplicity, clarity of
the solution (some Japanese term, perhaps?)


Mats Wichmann



From id at c1299689-a.saltlk1.ut.home.com  Tue Apr  2 01:36:10 2002
From: id at c1299689-a.saltlk1.ut.home.com (id at c1299689-a.saltlk1.ut.home.com)
Date: Tue, 02 Apr 2002 06:36:10 GMT
Subject: (Help) Embedding in Python 1.5
References: <slrnaahu3v.21k.id@c1299689-a.saltlk1.ut.home.com> <mailman.1017719704.5493.python-list@python.org>
Message-ID: <slrnaaikb0.30h.id@c1299689-a.saltlk1.ut.home.com>

In article <mailman.1017719704.5493.python-list at python.org>, 
Gerhard H?ring wrote:
> * HH <henrykhecht at my-deja.com> [2002-04-02 00:16 +0000]:
>> I'm trying to embed some Python in some C. [...] Could someone supply
>> me with a modern (but not too modern, i.e. 2.x) example makefile, or
>> at least some advice?
> 
> Does my previous answer to a similar question help? URL is (long):
> ...
> I'd be very interested if this actually works for other people.

Short answer: no, but I did try =].  First problem is that google
mangled "get_config_var" -> "get config var", which confused me for a
full minute or so.  Second, PYFLAGS ends up being:

-lieee -ldl  -lpthread -lm -L -I/usr/include/python1.5 -lpython1.5

which is slightly worse than what I had by guessing
(get_config_var("LIBPL") returns null or None in my case).

I'm sorry your idea didn't work.  REALLY sorry, because I still don't
have any extra ideas of my own (beyond what I wrote in the original
post).

-- 
HH


From steingrd at student.matnat.uio.no  Sat Apr  6 12:39:53 2002
From: steingrd at student.matnat.uio.no (Steingrim Dovland)
Date: Sat, 06 Apr 2002 17:39:53 GMT
Subject: checking if a number is int
References: <Xns91E8C6EF779FEvarialgmxde@130.133.1.4>
Message-ID: <m3bscwhgkb.fsf@localhost.localdomain>

Robert Spielmann <varial at gmx.de> writes:

> is there a way in python to check if a number is integer? My problem is 
> the following: I need to know if sqrt(x) is an integer value or if there 
> is a fractional part.

You can do it using the modulo-operator %.  An integer value % 1 
will return 0 and a non-integer value % 1 will return something
greater than 0.

>>> x = 5
>>> y = 5.33333
>>> x % 1
0
>>> y % 1
0.33333

steingrim dovland

-- 
Here the sea spray give
I was with her
We're under the ship so get me over


From peter at engcorp.com  Wed Apr 10 20:46:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 20:46:31 -0400
Subject: Help with Python advocacy and marketing
References: <mailman.1018447948.26949.python-list@python.org>
Message-ID: <3CB4DCE7.4CB5A72F@engcorp.com>

"Stephan R.A. Deibel" wrote:
> 
> >> * We are collecting user-side advocacy quotes for use in press
> >>   releases announcing future versions of Python.  If you have
> >
> >Some companies might have difficulty allowing their name to
> >be used by "just anyone" in the context of Python advocacy.
> >Can you describe more about the way in which this information
> >will be distributed, who is represented (is this a Zope Corp
> >or PythonLabs sponsored endeavour?), and so on?  
> 
> The press releases will come from the Python Software Foundation (PSF),
> which is the non-profit that holds the intellectual property rights to
> Python.  I've volunteered to help produce and distribute the releases.

[snip remainder of thorough and excellent response]

Thanks very much Stephan.  Even knowing it's the PSF is likely
sufficient in our case.  I'll start looking into it for my
organization soon.

-Peter


From gtaylor at lowebroadway.com  Tue Apr 23 06:40:33 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 23 Apr 2002 03:40:33 -0700
Subject: Destroying instances with PyQT
Message-ID: <f0fd5987.0204230240.56804d3f@posting.google.com>

I have attached a QTimer to one of my objects, but I cannot seem to
get rid of it when I'm finished, which cause the whole object not to
be destroyed, normally my instances are erased simply by erasing the
Dictionary in which they reside, which has always worked before the
QTimer was added. In the QT manual it says you can destruct an
instance like this:

QTimer::~QTimer?()

How can I translate this into Python?

TIA

Garry


From boud at valdyas.org  Tue Apr 16 16:22:07 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 16 Apr 2002 22:22:07 +0200
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <8MVu8.26112$vF6.828988@news2.tin.it> <x7elhfpmvc.fsf@ra.iago.org> <E16xWz4-0004oN-00@mail.python.org> <mailman.1018987328.23993.python-list@python.org>
Message-ID: <3cbc884e$0$85430$e4fe514c@dreader3.news.xs4all.nl>

Cliff Wells wrote:


> I'm not too familiar with other GUI toolkits, but in wxPython it's
> possible to generate a custom wxEvent that is put on the GUI's event queue
> and processed
> like any other GUI event.  You can then create an event handler for that
> custom
> event and safely do your updates there.  I've found this to be the
> simplest way
> to interact with a GUI from a thread.  I'd be surprised if PyQt didn't
> support something similar.
> 

>From 3.2, if everything works as needed. I wonder how the wxPython 
developers solved the problem Phil has just solved...

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


From geoff at gerrietts.net  Fri Apr  5 19:27:43 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 5 Apr 2002 16:27:43 -0800
Subject: Booleans, integer division, backwards compatibility; where is Python going?
In-Reply-To: <a8lc0l$ikk$1@panix1.panix.com>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net> <mailman.1018036693.6922.python-list@python.org> <a8lc0l$ikk$1@panix1.panix.com>
Message-ID: <20020406002743.GB4774@isis.gerrietts.net>

Quoting Aahz (aahz at pythoncraft.com):
> In article <mailman.1018036693.6922.python-list at python.org>,
> Geoff Gerrietts  <geoff at gerrietts.net> wrote:
> >
> >The second is that I personally find little to admire in C++, but one
> >of the things I do admire is that Bjarne decided to take a couple
> >years off to let the compiler writers and application programmers
> >catch up. The result was a language that everyone more or less thought
> >they knew, a solid foundation for people to build on and move forward.
> 
> That was Python 1.5.2.

You're right of course. I'm in the singularly bad position of having
entered the python world in about September of 2000; my experience has
been pretty much 17 months of "trust no ground to be fixed". I pretty
much neatly missed the 15 or 16 months of relative quiet between.

That's bound to contribute to the colour of the lenses I view the
world with; bound to elevate change control on my list of priorities.

Thanks,
--G.

-- 
Geoff Gerrietts                <geoff at gerrietts net>
"I have read your book and much like it." --Moses Hadas




From bh at intevation.de  Tue Apr 30 06:04:08 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 30 Apr 2002 12:04:08 +0200
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org>
Message-ID: <6qvga9bizr.fsf@abnoba.intevation.de>

Roman Neuhauser <neuhauser at mail.cz> writes:

> > >> The problem with int.chr() is that a lot of people are going to get
> > >> confused when they find that "42.chr()" doesn't work (although "42
> > >> .chr()" does).
[Alex explains that 42. is a float]
>     Right, I didn't realize 42. was a valid float. The space would be a
>     reasonable (beit tedious) solution. As long as this is highlighted
>     in the docs, I don't see a problem.

Why should chr be an int method in the first place? It's not tied to
ints all that much[1]. IMO it sits somewhere between ints (including
int-like objects !) and strings so it shouldn't be a method of either
class. Leaving it a function is perfectly fine, therefore, although
putting it into the string module might be a good idea. The same would
apply to ord, IMO.

   Bernhard


[1] In fact, chr today accepts any object that can be converted
successfully to an int, so that chr(42.5) does work, but that's not
always desirable.


-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From erno-news at erno.iki.fi  Wed Apr 17 06:29:22 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 17 Apr 2002 13:29:22 +0300
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <7xhemb744m.fsf@ruckus.brouhaha.com> <Ms9v8.22165$b62.626694@news1.tin.it>
Message-ID: <kuu1qafwjh.fsf@lasipalatsi.fi>

In article <Ms9v8.22165$b62.626694 at news1.tin.it>, Alex Martelli
<aleax at aleax.it> writes:

| Paul Rubin wrote:
|| "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
||| Is it possible to have a safe eval() on the server-side even when the
||| string consists of submitted user data?
|| 
|| Well, look at the docs for the rexec (restricted execution) and Bastion
|| classes, but basically no, there's no safe way.  Rexec doesn't stop an
|| infinite loop from hosing your cpu.

| If you have to watch for that, forking and running the DoS risk in the
| child process can help -- the parent process needs to keep an eye
| over the resources (mostly, time) being used by the child process and
| be ready to kill it if need be.  Simple, it ain't.

you can use resource limits - then it's pretty simple. both cpu time
memory usage can be restricted using the resource module (remember to
use the hard limit).

  -- erno


From bkc at Murkworks.com  Tue Apr 30 09:46:10 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 30 Apr 2002 09:46:10 -0400
Subject: OT: Happy Birthday Larry Niven
Message-ID: <3cce9f3a$1_6@goliath.newsgroups.com>

I think Python is a language that Motie Engineers would use...


(See "The Mote in Gods Eye")

--
Novell DeveloperNet Sysop #5

_




-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From gashton at cmedltd.com  Thu Apr  4 10:10:07 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 04 Apr 2002 16:10:07 +0100
Subject: CGIHTTPServer and custom headers
In-Reply-To: <Pine.LNX.4.44L.0204041905150.13708-100000@ods.pp.ru>
References: <Pine.LNX.4.44L.0204041905150.13708-100000@ods.pp.ru>
Message-ID: <1017933007.10710.9.camel@tp6>

On Thu, 2002-04-04 at 16:06, Denis S. Otkidach wrote:
> 
> Have you tried Medusa
> (http://www.amk.ca/files/python/medusa-0.5.1.tar.gz)?

No, I'd forgotten all about it. Thanks for the tip.

-- 
Graham Ashton





From pyth at devel.trillke.net  Wed Apr 24 12:38:57 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 24 Apr 2002 18:38:57 +0200
Subject: Subclass of int and initialisation with not integer values
In-Reply-To: <pan.2002.04.24.18.26.26.813701.17049@lip6.fr>
References: <pan.2002.04.24.17.09.52.840883.17049@lip6.fr> <pan.2002.04.24.18.26.26.813701.17049@lip6.fr>
Message-ID: <20020424163857.GC20209@devel.trillke>

> In fact I found the answer, it is that __new__ is called before __init__
> and as a true int is return by new, __init__ is called too late.
> So i solved my problem by redefining __new__ instead of __init__.
 
> Sorry for this useless post.

far away from it. After all metaclasses are
just in the process of beeing understood by the wider
community (including myself) even though the python-gods
and demi-gods might smile at at us :-)

	holger




From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 17:10:47 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 23:10:47 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE> <7xhemb744m.fsf@ruckus.brouhaha.com>
Message-ID: <a9i41d$dqo$1@nets3.rz.RWTH-Aachen.DE>

"Paul Rubin" <phr-n2002a at nightsong.com> wrote in message
news:7xhemb744m.fsf at ruckus.brouhaha.com...
> "Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> writes:
> > Is it possible to have a safe eval() on the server-side even when the
string
> > consists of submitted user data?
>
> Well, look at the docs for the rexec (restricted execution) and Bastion
> classes, but basically no, there's no safe way.  Rexec doesn't stop an
> infinite loop from hosing your cpu.
>..

Thanks. I will check into that.

>
> Why do you want to eval strings from untrusted users?
>

Sorry, it's a bit of a longer story.

I have an XML-based (data oriented) language with some scripting mechanisms.
None of these include any kind of reading or writing in the file system, by
language definition. These XML files then get interpreted (currently by PHP,
VBScript, I'm adding Python) on the server-side. For the sake of argument
and to stick to my original problem, let's say I do more or less trust the
XML data, because I host it, and even if a stranger wants me to host his
data I could look through it.

Now. In the ASP/ VBScript version I keep all session data on the server side
and simply submit the session ID forth and back. The untrusted user on the
client side can alter the session ID, but not the actual session data.
However in the PHP version, and how I'm currently laying it out in the
Python port, the session data is put into hidden form variables. These can
be altered before send back by anybody knowing how to edit the XHTML (and
change the HTTP header if I'd check the referrer.) I still like this concept
currently because it allows me to not be dependent on HTTP header cookies
when I use the language internal session methods.

And session data is evaluated. E.g. the session data is player_has_item =
"true", the value could be changed to player_has_item =
"delete_all_files()", which of course results in unwanted side effects if
eval() is used. (Trouble is, player_has_item could be a number, a string, or
a boolean. Only in the VBScript port I do also store the variable type,
which is defined in the XML data.)

But maybe if there's no secure eval() I have to rethink the whole process.

(For more information on the project please see http://questml.com .)




From martinb at otenet.gr  Wed Apr 24 03:51:48 2002
From: martinb at otenet.gr (Martin Barnard)
Date: 24 Apr 2002 10:51:48 +0300
Subject: Killing all child processes?
In-Reply-To: <59aa3aef.0204231823.77932129@posting.google.com>
References: <59aa3aef.0204231823.77932129@posting.google.com>
Message-ID: <1019634727.5374.12.camel@toad>

Shouldn't you be looking into each child process having a handler to
pass on any SIG events it receives from STDIN (or parent)?
e.g. for each child process, take a PID and store it.
If parent receives a SIG event, then pass it down to child processes.
Each child process should have the same handler code in also (possibly
with a signal to parent if it quits, so that the list of child processes
is current for each parent).
Sorry if it sounds a little strange, I've just woken up, and am rambling
a little :)

Let's see if I can visualise:

   Parent_1
      |
  ____|____
 |         |
Child1   Child2
  |
Child3
 
Now, if I send a SIGKILL to Parent_1, it will pass on <SIGKILL> to
Child1 and Child2. Child1 will pass on <SIGKILL> to Child3 via the same
mechanism.

The only potential problem would be if a child branch died before hand,
and the PID was taken by another process that Parent_1 could kill (same
user).

Hope this helps, if not, just ignore my waffling.

Martin.

On Wed, 2002-04-24 at 05:23, Kevin wrote:
> Hi all,
> 
> I can't seem to get my python program to kill all the child prcoesses
> that it may fork. Is this possible?
> 
> I've researched usenet+the web and read subproc.py from
> ftp.python.org, but still no go.
> 
> My latest attempt was a python script that forks itself, uses setpgid,
> then the child does an execl(). The execl'd program creates some
> children. Meanwhile, the parent decides time is up on the child and
> tries to kill everything, but  my os.kill(pid,9) only succeeds in
> killing the child aka the execvp'd process.
> 
> There's another bug that I haven't yet addressed about the parent
> waiting around when the child doesn't write to either of the pipes.
> But that'd not my focus now, so don't worry about that; I just want
> all the children to die.
> 
> My full code it at the bottom. Hints, anyone? This is on Linux and
> FreeBSD.
> 
> Thanks,
> Kevin
> 
> ---Full code follows---
> ---- pcontrol.py below
> #!/usr/bin/python
> #
> # pcontrol.py
> #
> 
> import os
> import sys
> import signal
> 
> _alarm = "request timed out"
> 
> time = 5
> 
> cmd="sh1.sh"
> 
> ###########################################################
> 
> def alarm_handler(a , b):
>     global _alarm
>     signal.alarm(0)
>     raise _alarm
> 
> ###########################################################
> 
> # Okay, let's do this the hard way: create two unnamed
> # pipes for stdout/err. Then, fork this process, and the
> # child becomes the new command and shoves their output
> # onto the pipes, while the parent waits for the child to die.
> 
> # Create two named pipes
> child_out_r, child_out_w = os.pipe()
> child_err_r, child_err_w = os.pipe()
> 
> # Let's fork a child that gets replaced with the test process - krf
> child_pid = os.fork()
> 
> if (child_pid != 0):
>     # We're the parent
> 
>     os.close(child_out_w) # the parent won't read the pipes
>     os.close(child_err_w)
> 
>     os.setpgid(child_pid, 0)
> 
>     # Now do the timing...
>     signal.signal(signal.SIGALRM, alarm_handler)
>     signal.alarm(time)
> 
>     try:
>         child_pid, child_status = os.waitpid(child_pid, 0); # wait for
> child
>     except OSError:
>         print("OSError... did the executable time out?")
>         child_status = -1 # this is the same as a timeout for now
>     except _alarm:
>         print("Timed out")
>         # kill it!
>         os.kill(child_pid, 9)
>         child_status = -1 # timeout
> 
>     signal.alarm(0) # clean up alarm
> 
> else:
>     # We're the child
>     os.setpgid(0, 0) # now the child is it's group leader (?)
> 
>     # Redirect error to our pipes
>     os.close(child_out_r) # the child won't read the pipes
>     os.close(child_err_r)
>     os.dup2(child_out_w, sys.stdout.fileno())
>     os.dup2(child_err_w, sys.stderr.fileno())
> 
>     # replace the process
>     os.execl(cmd, cmd) #strange, but execl->execv needs a second
> arg...
>     # the child is no more... (assuming success)
>     
> print "Status = ", child_status
> print "Child stdout [",os.read(child_out_r,999),"]"
> print "Child stderr [",os.read(child_err_r,999),"]"
> 
> 
> ------- sh1.sh below
> #!/bin/sh
> 
> echo "This is processs (1) $$."
> sh2.sh &
> sh3.sh &
> sleep 600
> echo "Process (1) $$ exiting."
> 
> ------- sh2.sh below
> #!/bin/sh
> 
> echo "This is processs (2) $$."
> sleep 600
> echo "Process (2) $$ exiting."
> 
> ------- sh3.sh below
> #!/bin/sh
> 
> echo "This is processs (3) $$."
> sleep 600
> echo "Process (3) $$ exiting."
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-- 
Experience is that marvelous thing that enables you recognize a mistake
when you make it again.
		-- F. P. Jones





From donn at u.washington.edu  Tue Apr 23 12:35:30 2002
From: donn at u.washington.edu (Donn Cave)
Date: 23 Apr 2002 16:35:30 GMT
Subject: Python & autoftp
References: <53b7d218.0204230720.686e473f@posting.google.com>
Message-ID: <aa42gi$6jg$1@nntp6.u.washington.edu>

Quoth davies at jkevin.freeserve.co.uk (kevin):
| I'm pretty new to Python and I would like to develop a script in
| Python to automatically FTP. I have already used bash and expect and
| now would like to replace bash entirely with Python.
|
| Is there any merit in replacing bash with Python??

In general, each is superior in its own domain of application.
More specifically, if your program works fine now, it probably
won't work any better in Python, but you may learn something
about Python and UNIX internals while you're at it.

| Can anyone point me in the right direction.

The expect part is probably the hardest.  I assume you mean to
replace expect too, not just bash.  Look at openpty and forkpty
in the posix ("os") module, and see if you can get them to work
in a little test application (if you even have them.)  There
are also more complete expect solutions, if that's interesting
you might look in the Vaults of Parnassus exit from www.python.org.

Alternatively, whatever you're doing might be done more easily
by replacing ftp itself with ftplib module.  That will make the
expect part unnecessary.

	Donn Cave, donn at u.washington.edu


From nobody at nobody.com  Sat Apr  6 13:45:45 2002
From: nobody at nobody.com (John)
Date: Sat, 6 Apr 2002 12:45:45 -0600
Subject: Textarea
Message-ID: <OjHr8.36744$tg4.437615@vixen.cso.uiuc.edu>

Can we read file to display inside <textarea> and allow users to modify the
content and overwrite, using Python CGI? Thanks!




From robin at jessikat.fsnet.co.uk  Fri Apr 26 14:10:37 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 26 Apr 2002 19:10:37 +0100
Subject: reportlab and BaseDocTemplate
References: <20020426180122.3f7ce976.wilk@flibuste.net>
Message-ID: <y9Y9YKAdgZy8EwN2@jessikat.demon.co.uk>

In article <20020426180122.3f7ce976.wilk at flibuste.net>, William Dode
<wilk at flibuste.net> writes
>hello,
>
>I would like to use BaseDocTemplate to make report with different
>PageTemplate. But the documentation is not enough for me, where can i find
>some examples ?
>
>thanks
>
If you direct your mail to the reportlab users list
see http://two.pairlist.net/mailman/listinfo/reportlab-users 
you'll get a more reportlab aware audience.

Anyhow the docs/userguide directory contains a very large example ie the
userguide itself which uses the base doc template class. It might be
easier to look first at demos/odyssey/dodyssey.py which has a title
page, chapter page in single column and later pages in double column.
-- 
Robin Becker


From jeff at ccvcorp.com  Fri Apr  5 20:54:51 2002
From: jeff at ccvcorp.com (Jeff Shannon)
Date: Fri, 5 Apr 2002 17:54:51 -0800
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net>
Message-ID: <MPG.1717f5348f08f8ea98969d@news.nwlink.com>

In article 
<GZqr8.20207$ml2.1517221 at newsread1.prod.itd.earthlink.net>, 
robinjim at earthlink.net says...
> Is there a way to convert a single quoted string to a triple quoted string?
> 
> For example, given:
> 
> ''' abc %d xyz '''
> 
> the result of ''' abc %d xyz''' % 10 is:
> 
> ' abc 10 xyz '
> 
> which is a single quoted string.
> 
> I would like to convert the result to a triple-quoted string.

In Python, there is no internal difference between single-quoted 
and triple-quoted strings -- they're all just strings.  The 
triple-quote trick is just a way to allow string literals to more 
easily contain characters that would be difficult to include 
otherwise.  Notable among those characters are newlines, single 
quotes, and double quotes.  However, once they're parsed, there's 
no difference at all.

>>> triple = """This is a string with 
... a newline and a " quote."""
>>> double = "This is a string with \na newline and a \x22 
quote."
>>> single = 'This is a string with \na newline and a " quote.'
>>> triple == double
1
>>> triple == single
1
>>> double == single
1
>>> 

As you see, the only difference between quoting styles, is a 
matter of what's easiest to type/read.

-- 

Jeff Shannon
Technician/Programmer
Credit International


From robin at jessikat.fsnet.co.uk  Mon Apr  8 14:04:15 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 8 Apr 2002 19:04:15 +0100
Subject: PEP 285: Adding a bool type
References: <EFEFIHAINIINNELLHFAIMECBCAAA.kseehof@neuralintegrator.com>
 <mailman.1018281440.8404.python-list@python.org>
Message-ID: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk>

This article in salon 
        http://www.salon.com/tech/feature/2002/04/08/lehman/print.html

about Manny Lehman, my old department head, gives an insight into why
people like myself are often afraid of software change.

'''Included within the new set of laws are the Law of Continuing Growth
("The functional capability of E-type systems must be continually
increased to maintain user satisfaction over the system lifetime") and
the Law of Declining Quality ("The quality of E-type systems will appear
to be declining unless they are rigorously adapted, as required, to take
into account changes in the operational environment"). For added
measure, Lehman has also thrown in the Principle of Software
Uncertainty, which states, "The real world outcome of any E-type
software execution is inherently uncertain with the precise area of
uncertainty also unknowable."'''


-- 
Robin Becker


From siegfried.gonzi at kfunigraz.ac.at  Mon Apr 15 14:24:40 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 15 Apr 2002 20:24:40 +0200
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org>
Message-ID: <3CBB1AE8.E2843017@kfunigraz.ac.at>

Brian Quinlan wrote:

> Where need to link with Pythonxx.lib. But instead of this manual crap,
> you should just write a simple distutils setup script.


I ran into a similar problem: including an external C function in
Python. The SWIG manual is somewhat behind time and my first attempts
failed. I just then wrote to the developers (or mailinglist?) and they
explained me a "how to" with the assistance of distutils.

It was quite easy (even for me, and it was my first attempt ever, to
glue together applications). 

The distutils-documentation is quite vague in this respect; if I hadn't
contacted the authors of SWIG, I would'nt have contemplated to use
distutils in order to prepare my C function for including in Python.

Maybe it is a good idea to contact the SWIG adress directly (it is
likely that someone can help you).


S. Gonzi
[Someone interested in  the Bohren-Huffman Mie
(http://atol.ucsd.edu/%7Epflatau/scatlib/index.html) code? I made
successful the C wraper function for Python on Windows? It wouldn't be a
problem for me to give it to someone interested in or explain how I did
it]


From claird at starbase.neosoft.com  Thu Apr  4 18:22:36 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 4 Apr 2002 17:22:36 -0600
Subject: Eval only in homeopathic dosage (was: turning a string into an object name)
References: <1017881888.857285@DS9.ai-lab.fh-furtwangen.de> <IbXq8.48892$pT1.1388420@news1.tin.it> <3cac943f.132176618@mammoth.usenet-access.com> <MPG.17165ec2e3c10c6c98969a@news.nwlink.com>
Message-ID: <3D605F0E18400355.00C9CA937F5F0F26.B72EE95D42998046@lp.airnews.net>

In article <MPG.17165ec2e3c10c6c98969a at news.nwlink.com>,
Jeff Shannon  <jeff at ccvcorp.com> wrote:
			.
			.
			.
>If you have numerous objects that are conceptually similar, so 
>that it makes sense to name them sequentially, then you *really* 
>want to put them in a list or dictionary instead.
>
>The same thing is true of almost *every* use of exec or eval() -- 
>it seems handy, but if you look for it, there's a better solution 
>to your real problem.
			.
			.
			.
Of course.

The question I have is this:  do the (Common, let's say)
Lispers agree yet?  Their usual response to lots of situ-
ations is that Lisp has hygienic macros, and that makes
all the difference in the world.  Well, yeah, Lisp's 
great for definition of Little Languages, which are
always inherently desirable.  On the other hand, intro-
spective fiddling is always, *always* tricky, and never
helps maintenance of (end-user) applications.  We pro-
pellorheads lust for it, and of course it's crucial in
small doses, but it's simply a hazard beyond one or 
two molecules per program.

I'll make this concrete:  what's a modern example of 
use of a hygienic macro in an end-user application
that wouldn't truly be better off as a dictionary or
related data structure?
-- 

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


From johanfo at essay.org  Sun Apr 28 08:50:29 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Sun, 28 Apr 2002 14:50:29 +0200
Subject: Python Webpages
Message-ID: <aagr7h$5c9$1@maud.ifi.uio.no>

Hi, there.

I previously saw a posing named "Python vs Php" and sys atht
"... there are many packages you can download..."

However, only one package is mentioned: jonpy

What is the most popular and promising webpackages for python ?

--
JF?




From bokr at oz.net  Wed Apr 17 16:21:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 20:21:16 GMT
Subject: list_a-list_b
References: <mailman.1019022993.8455.python-list@python.org> <Xns91F32263AA919RASXnewsDFE1@130.133.1.4>
Message-ID: <a9klfs$n5t$0@216.39.172.122>

On 17 Apr 2002 10:21:46 GMT, Philip Swartzleonard <starx at pacbell.net> wrote:
[...]
>Tim Peters || Tue 16 Apr 2002 10:54:56p:
[...]
>> 
>> though.  Contrary to popular belief, there's no Fabulous Prize
>> awaiting those who minimize either vertical or horizontal space
>> <wink>. 
>
>Minimizing relative vertical space is great; that's why list 
>comprehensions are better than nested for loops for expressing simple 
>ideas ('i want var to be an x by y matrix of foozles' =), and why moving 
>stuff into subfunctions is nice. Horizontal space though... feh. Unless 
>you're using emacs default python mode, in a two or three level deep 
>looping structre in a function in a class, and you need to write 
>'self.data[x][y].variable = self.long_function_name( lots, of, arguments 
>)' ... =)

If you want the form but not the looong one-liner:

    self.data[x][y].variable = (
        self.long_function_name(
            lots,
            of,
            arguments 
        )
    )

Hm, this makes me wonder if '=' could introduce an optionally indented expression block,
using ident/dedent where the above is using = ( .. )

Regards,
Bengt Richter


From huaiyu at gauss.almadan.ibm.com  Wed Apr 17 14:07:44 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 17 Apr 2002 18:07:44 +0000 (UTC)
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <87vgaswc1s.fsf@tleepslib.sk.tsukuba.ac.jp> <7934d084.0204161757.72ff525f@posting.google.com> <87u1qatvc7.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <slrnabrefg.smi.huaiyu@gauss.almadan.ibm.com>

Stephen J. Turnbull <stephen at xemacs.org> wrote:
>Those discussing these things have long since published definitions,
>agreed to accept each others' definitions, _and they have also long
>since agreed that for the purposes of most discussions comparing the
>broad idea of open source with that of free software, "free software
>== open source software"_.  Huaiyu clearly found Tim Peters's
>contribution in that vein useful.  So his _question_, and the
>appropriate answer, were ambiguous.

Given various bits of information we have discovered, I think it is the OSI
"marketing message" that is in err.  (Not that we should have expected
differently from marketing.)  The FSF is quite accurate and honest in their
assessment of not only their own definitions, but other as well.

>Actually, I have to laugh at myself, because I was in fact wrong to
>assert that the sets of licenses/covered programs are the same, and
>Huaiyu led me to truth.  Believe me, I didn't plan that, but there is
>proof: I was seeking (despite being fairly sure I already had the
>answer) and I found.<wink>

It is very gracious of you to say that.  We are all seeking, in different
ways.  And I believe we all become a little wiser as result.

Huaiyu


From wilk at flibuste.net  Mon Apr  8 06:02:42 2002
From: wilk at flibuste.net (William =?ISO-8859-1?B?RG9k6Q==?=)
Date: Mon, 8 Apr 2002 12:02:42 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <20020408120242.41456ace.wilk@flibuste.net>

Le Mon, 8 Apr 2002 11:55:24 +0200
"Harald Schneider" <h_schneider at marketmix.com> ?crivait:

> Hi,
> 
> what are your experiences, using wxPython/wxWindows in large projects?
> 
> Large means in my case a sales system, based on MySQL.
> I'm thinking about implementing it in wxPython, but I'm not sure about
> the stability of
> wxWindows and the wxPython wrapper ...
> 
> The alternative way would be a project in Borland Delphi, which has
> proofed its
> stability over years ...
> 
> So what do you think about it!
I think i have the same question in my head !
More, i would like to know how is the stability of python on windows,
compare to java or delphi for example.

bye

-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From bbollenbach at shaw.ca  Wed Apr  3 19:25:33 2002
From: bbollenbach at shaw.ca (Brad Bollenbach)
Date: Thu, 04 Apr 2002 00:25:33 GMT
Subject: Behaviour of fileinput with stdin
Message-ID: <14Nq8.7009$%3.476438@news1.calgary.shaw.ca>

Given the following code:

#!/usr/bin/python
import fileinput
for line in fileinput.input():
    print line

saved as, foo.py...

$ ./foo.py
this is some input
this is more stuff
EOF <-- (as in, on this line the ONLY thing i do is hit Ctrl-D)

Why must I hit Ctrl-D /twice/ for it to "take" to the EOF? And why does
it merely print out what I've just entered, and then keep prompting me,
instead of ending immediately, as Perl's <> operator does?

e.g.
#!/usr/bin/perl
while (<>) { print }

saved as foo.pl

$ ./foo.pl
this is some text
this is some text   <-- immediately prints it back out
EOF <-- (same as above)
$

It was my understanding that fileinput.input() is the Pythonic Way of
doing Perl's <>. Can someone please explain the difference in behaviour
between the two? Specifically:

1.) Why doesn't fileinput.input() immediately print the lines?
2.) Why won't it "take" to the EOF until I hit EOF a second time?

--
Brad


From bvdpoel at uniserve.com  Wed Apr 17 20:12:28 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Wed, 17 Apr 2002 17:12:28 -0700
Subject: Problems with copies (and references) to lists
Message-ID: <3CBE0F6C.17CA6751@uniserve.com>

Took awhile, but I do have the following bit of code working as it
should. The doc string is misleading, but I'm not sure of how better to
describe it out of context... The important thing is, it is giving the
right result, and it is NOT changing the original list.

-------
def doubleUp(a, fact):
        """ Makes fact/2 copies of list [ [s, l, v,...] ].
                New copy has all values of l divided by fact/2 and
                copies of s set to s/2 and 50+(s/2).
		fact will always be a power of 2.
        """

        new = a
        while fact >1:
                b1 = []
                b2 = []
                for t in new[:]:
                        n=t[:]          # this prevents munging our
original
                        n[0] /= 2
                        n[1] /= 2
                        b1+=[n[:]]      # [:] forces new copy of n!
                        n[0] += 50
                        b2+=[n]
                fact /=2
                new = b1 + b2
        return new

a = [[0.0, 20.0, 66, 44],[50.0, 20.0, 55, 33]]
print doubleUp(a, 2)
print doubleUp(a,4)
print a

--------

But, I'm wondering if it is at all efficient. Or even good python. Seems
that the number of [:] needed just makes for ugly code.

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From phr-n2002a at nightsong.com  Mon Apr  1 19:32:29 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 01 Apr 2002 16:32:29 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA8E37A.E89CA002@noaa.gov>
Message-ID: <7x4rivj5ya.fsf@ruckus.brouhaha.com>

Chris Barker <Chris.Barker at noaa.gov> writes:
> By the way, here is an example of why the current sytem is bad:
> 
> string.find() returns the index of the found string, and 0 being a
> perfectly valid index, it can NOT be used to return an indication
> that the substring was not found, so -1 is returned instead. None
> could not be returned to mean not found, because both 0 and None
> would resolve as false.  I'd love to see False be returned, and then
> you could do:
> 
> if string.find(...):
> 	do something.

Are you proposing to start making 0 resolve as true?  If not, returning
False has the same problem as returning None.


From phr-n2002a at nightsong.com  Thu Apr  4 20:37:13 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 17:37:13 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85> <3CAC1EF9.C4A97594@alcyone.com> <Xns91E6ACE256C42JamesLLugojcom@216.148.53.96>
Message-ID: <7x7knnvsc6.fsf@ruckus.brouhaha.com>

James Logajan <JamesL at Lugoj.com> writes:
> x = ["is", "are"]
> y = {0: "", 1:"s"}
> 
> for n in range(3):
>    print "There %s %d item%s." % (x[n <> 1], n, y[n <> 1])
> 
> Will the behavior be backword compatible and equivalent or not? Simple 
> question I assume; waiting for an answer.

Yes, if I understand the PEP, that should still work.

This, however, won't work compatibly:

   print "Calm down, it's only %s's and %s's." % (1==1, 1==0)


From mstern at aon.at  Tue Apr  2 01:47:50 2002
From: mstern at aon.at (Matthias Stern)
Date: Tue, 2 Apr 2002 08:47:50 +0200
Subject: [Newbie] embed c++ in Python
References: <3ca8901c$0$27282$6e365a64@newsreader02.highway.telekom.at> <a8a62r$57u$1@peabody.colorado.edu>
Message-ID: <3ca95375$0$3698$6e365a64@newsreader02.highway.telekom.at>

> > I want to know how to write a function which receives some variables as
> > argument changes them and sends the results back to Python.

Thank you for the links I got (http://www.scipy.org/site_content/weave
 and http://www.acl.lanl.gov/siloon/ ). I've visited the sites and
discovered a new problem: Both solutions require some additional
stuff to be installed. Maybe it's much simpler to use them but it
is quite hard to add some additional stuff like this to our project. It's
a cross-platform project with many operating systems and we want as
little additional programs as possible.

So I have to ask again: is it possible to do this without any additional
installations ("pure" Python 2.2) ? (L ike the "spam" example on the Python
Homepage) Has anyone a simple example code or link for this?

Thx, Mac




From christophe.delord at free.fr  Thu Apr 18 15:54:24 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 18 Apr 2002 21:54:24 +0200
Subject: Counting the number of elements in a list - help.
References: <a9hu8e$2fg$1@lumberjack.rand.org>
	<MPG.1728bf07ecc579969896b3@news.nwlink.com>
Message-ID: <20020418215424.28033abb.christophe.delord@free.fr>

On Thu, 18 Apr 2002 12:32:13 -0700
Jeff Shannon <jeff at ccvcorp.com> wrote:

It could be shorter and smarter like this :

def InRange(sequence, index):
    return 0 <= index < len(sequence)

> In article <a9hu8e$2fg$1 at lumberjack.rand.org>, bmason at rand.org 
> says...
> > I need to know these things; may someone help me out?
> > 
> > 1) a function which returns the number of elements in a list
> > 2) a function which takes a list and an index and returns a zero when that
> > list index is out of range.
> 
> It seems that your problem has been solved, but... just 'cause...
> 
> Your second function could easily be written like this:
> 
> def InRange(sequence, index):
>     try:
>         sequence[index]
>     except IndexError:
>         return 0
>     else:
>         return 1
> 
> (Note that the 'return 1' isn't strictly required to be in an 
> else: clause here, but I think that using else, in this case, is 
> a clearer expression of intent.)
> 
> -- 
> 
> Jeff Shannon
> Technician/Programmer
> Credit International


-- 
Christophe Delord
http://christophe.delord.free.fr/


From fdrake at acm.org  Wed Apr  3 13:27:46 2002
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 3 Apr 2002 13:27:46 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <yu99ofh0vg4x.fsf@europa.research.att.com>
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com>
	<016701c1db1a$569543e0$0900a8c0@spiff>
	<yu99wuvovkmm.fsf@europa.research.att.com>
	<00af01c1db2a$e7316d10$ced241d5@hagrid>
	<200204031619.g33GJXm02577@europa.research.att.com>
	<016b01c1db31$8da51010$ced241d5@hagrid>
	<yu99ofh0vg4x.fsf@europa.research.att.com>
Message-ID: <15531.18850.911468.547377@grendel.zope.com>

Andrew Koenig writes:
 > The very first page I visited was the description of gc, the
 > garbage collector interface.  The first function that returns
 > a truth value there is described as
 > 
 >         isenabled()
 >             Returns true if automatic collection is enabled.

I spent a bit of time talking to Guido about the effect of PEP 285 on
the documentation.  My own preference is to do the same as the
examples you cite: use the English words "true" and "false", and not
give concrete typed values.

Guido would rather *not* have it this way, but rather document the
exact concrete return values.  He'd have me refer to "True" and
"False" in the same way we refer to "None", and refer to the return
type as "bool" (if indeed we're using the new bool type), or, if 1 and
0 are the return values, document them specifically as well.

For function arguments, he'd like to see truth values described as
Boolean, and the values would be the English "true" and "false".
Translations would be able to translate the English words, but not the
True and False values, which are language constants (as constant as
built-ins are).

Guido & I may clash over this a few times.  ;-(


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation




From bosahvremove at netscape.net  Wed Apr  3 13:09:03 2002
From: bosahvremove at netscape.net (Bo)
Date: Wed, 03 Apr 2002 18:09:03 GMT
Subject: is there gonna be a Core Python Programming 2e ?
References: <codq9uko6ueumep852po13m6pbtfs3r2ts@4ax.com> <a7jkm6$cnt$1@panix1.panix.com> <fcd9edc1.0204012323.4a88c6fe@posting.google.com>
Message-ID: <Xns91E58484660E9bosahvnetscapenet@66.185.95.104>

My favourite Python Book!

Any chance of getting a whack at the missing chapters :) Esp SQL and Win32?

My sympathies for the python version race problem. I can understand, as a 
consumer I wouldnt be eager to buy a new book for python 2.2 vs 2.1. At the 
time I bought it I was happy it encluded 2.0 elements.

2.4?, 3.0? either way I'll tell my friends.

Bo


From Dick.Zantow at lexisnexis.com  Wed Apr 17 10:32:46 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Wed, 17 Apr 2002 10:32:46 -0400
Subject: Classes Again !
References: <3cbd7527$1@news.comindico.com.au>
Message-ID: <a9k156$hoa$1@mailgate2.lexis-nexis.com>

"Graeme Matthew" <graeme.matthew at unite.com.au> wrote in message
news:3cbd7527$1 at news.comindico.com.au...
> Hi all  I am going mad again, but making progress, been on a problem for
1
> hour already and cannot see the cause ! please could someone help
>
> The getSubDocument method in my template class is failing suddenly , i
dont
> have a clue why ! thanks a mil
>
> ERROR:
>
> Traceback (most recent call last):
>   File "C:\www\cgi-bin\main.cgi", line 10, in ?
>     data = t.getSubDocument('C:/www/templates/template_org_query.html')
> AttributeError: Template instance has no attribute 'getSubDocument'
>

There are some errors in the code: "none" is not "None", and in the line
>         prog = re.compile('<body>(.*)</body>',re.DOTALL,re.IGNORECASE)
the comma before the second "re." should of course be a vertical bar (
'|' ). When I
corrected those, it seemed to work better.

> CGI SCRIPT BELOW
>
> #! C:/python22/python.exe
>
> import sys
> sys.path.insert(0,'C:/www/classes/')
> import Template
>
> t = Template.Template('C:/www/templates/template_main.html')
> t.loadTemplate()
> data = "hello mom"
> data = t.getSubDocument('C:/www/templates/template_org_query.html')
> t.replaceDirective('workbench',data)
>
> print "Content-type: text/html\n\n";
> print t.document()
>
> CLASS BELOW:
>
> import re
>
> class Template:
>
>     def __init__(self, filename):
>
>         self.filename = filename
>         self.subdoc = none
>
>     def loadTemplate(self):
>         f = open(self.filename,'rb')
>         self.data = f.read()
>         f.close
>
>     def replaceDirective(self, directive, data):
>         prog = re.compile('<!--\s*' + directive + '\s*-->.*?<!--\s*' +
> directive + '\s*-->',re.DOTALL | re.IGNORECASE)
>         self.data = prog.sub(data,self.data)
>
>     def document(self):
>         return self.data
>
>     def getSubDocument(self, fil):
>
>         f = open(fil,'rb')
>         self.subdoc = f.read()
>         f.close
>
>         prog = re.compile('<body>(.*)</body>',re.DOTALL,re.IGNORECASE)
>         result = prog.match(self.subdoc)
>
>         if result != none:
>             self.subdoc = result.group(0)
>         else:
>             self.subdoc = '<parse error: no data, is it the correct file
?'
>
>         return self.subdoc
>
>
>




From emile at fenx.com  Wed Apr 24 17:01:28 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 24 Apr 2002 21:01:28 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org>
Message-ID: <I2Fx8.46966$CH.7051@rwcrnsc52.ops.asp.att.net>

Oyvind Gronnesby
> * Cimarron Taylor
> |
> | Will this do what you want?
> | 
> | verboseGender = ('female','male')[gender=='m']
> 
> Wouldn't this break in a PEP 285 Python world?
> 


You're not suggesting there's only 1 True gender?  ;-))

--

Emile van Sebille
emile at fenx.com

---------



From starx at pacbell.net  Fri Apr 12 06:05:35 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 12 Apr 2002 10:05:35 GMT
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com>
Message-ID: <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4>

Seo Sanghyeon || Fri 12 Apr 2002 02:47:52a:

> Hello. I am new to Python. I have a question. I got confused
> by all those 'magic words' -- __init__, __class__, __doc__,
> __name__ and so on. Where can I find a documentation about
> these words? Or is there a list of 'magic words'?
> 
> Happy Python!
> Seo Sanghyeon

The first section of the index[1] for the language reference manual lists 
all of these things (the underscore section), with links. You can often 
find the relevant names for an object if you can find the right 
description of the object in the reference, e.g., [2] points near moudle, 
class, and instance information. Be warned, the language refernece can 
read like a brick if you aren't used to it, and it's easy to get confused.

The library reference[3] has some useful information near the beginning, 
in section 2. For exapmle, section 2.2.9.3 [4] gives all the names you 
need to do evil things to function objects.

If anyone has a better rundown... =)

Have fun.

[1]http://www.python.org/doc/current/ref/genindex.html
[2]http://www.python.org/doc/current/ref/types.html#l2h-74
[3]http://www.python.org/doc/current/lib/lib.html
[4]http://www.python.org/doc/current/lib/typesfunctions.html

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From PoulsenL at capecon.com  Tue Apr 30 11:13:31 2002
From: PoulsenL at capecon.com (PoulsenL at capecon.com)
Date: Tue, 30 Apr 2002 11:13:31 -0400
Subject: Remove masked elements from an array
Message-ID: <72F73B808A78D511A73600034771D9FF717CDD@dc_exchange2.howrey.com>

By the way it was indeed a ``Numeric'' array.  Thanks again.

-----Original Message-----
From: PoulsenL at capecon.com [mailto:PoulsenL at capecon.com]
Sent: Tuesday, April 30, 2002 11:06 AM
To: python-list at python.org
Subject: RE: Remove masked elements from an array


Just wanted to express my gratitude for everyone's input.  This community is
the main reason why I love to use Python.

Loren


-- 
http://mail.python.org/mailman/listinfo/python-list




From terabaap at yumpee.org  Tue Apr  9 12:51:49 2002
From: terabaap at yumpee.org (Manoj Plakal)
Date: Tue, 09 Apr 2002 10:51:49 -0600
Subject: IDE Wars, the prequel!
References: <mailman.1017880688.9896.python-list@python.org> <6RSq8.29207$0T6.8753904@e3500-atl1.usenetserver.com>
Message-ID: <3CB31C25.6090101@yumpee.org>

Robert Oschler wrote:
> Thanks, that clears it up for me.  Seeing as I'm going to use Zope, Quanta
> is starting to look very attractive since it does HTML editing, integrates
> well with Zope, and has facilities for Python.

        Wing IDE (wingide.com) is a pretty nice Python IDE and it
        has Zope support and a good debugger that can run
        GUI programs (wxPython, Qt, Tk, Gtk etc) without
        problems (unlike IDLE and PythonWin). I believe Komodo
        is no longer being actively developed?

        Manoj




From cgale1 at cox.net  Fri Apr  5 07:03:18 2002
From: cgale1 at cox.net (Van Gale)
Date: Fri, 05 Apr 2002 12:03:18 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <TdSq8.48142$pT1.1370343@news1.tin.it> <3CAD4CD8.2090700@mxm.dk> <7Adr8.1447$%A.217236@news1.west.cox.net> <Ster8.62071$S52.2165674@news2.tin.it> <3CAD8E0F.3080404@mxm.dk>
Message-ID: <aogr8.2659$%A.278202@news1.west.cox.net>

Max M wrote:

> ;-) <- smiley added this time around

Ok, you guys have me laughing!   Very clever, and I apologize for being too
serious :)

Van Gale





From siegfried.gonzi at kfunigraz.ac.at  Tue Apr  9 04:06:01 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Tue, 09 Apr 2002 10:06:01 +0200
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <mailman.1018334899.16298.python-list@python.org>
Message-ID: <3CB2A0E9.1AE9DF80@kfunigraz.ac.at>

Laura Creighton wrote:
> 
> > I know this is way off topic, but I figure the heavy hackers here will
> > understand my mindset better than those at comp.lang.C
> >     I have been doodling with python (nasty habit I know but I have to
> > do somthing with my snake) and think I may be ready to learn a second
> > language.  I am leaning towards C over C++ and Java even though I
> > loose objects it will ground me firmly in functional programing.
> 
> If you want to learn 'functional programming' then C is not the place.
> I recommend Haskell. 

If it is only for the sake of his learning experience: he should go with
Clean. Maybe I am a little bit biased but installing Clean (on Unix,
Macintosh and Windows) is matter of a few minutes.

If he goes with Haskell it is likely that the performance issue will
come in the way; and here Clean (without the need of C) really shines.
In Haskell it is possible too, but only with the help of a C compiler.

I am not sure whether they (Haskell) upgraded their monadic I/O stuff
but sooner or later he will mabye become scared of the monadic part.
Clean here has something like the operater "<<" (as in C++) and this
without side effects.

So, this only holds true for the learning part. If you want to use
functional programming in your work Clean will become more or less a
hindrance (I would say: There is some purity  --not only that there are
only 10 users in Europe-- but Clean is useless on a daily basis). 


>  However, if your real goal is
> to hack C, and all the above is just one misused word -- you want
> K & R.  Everybody wants K & R.  

Not really. This book is really,really boring.

And it is a "clash of paradigms" if you were used to functional (or
Python and the like) programming first. 

There is a good book out there: "Teach C++ yourself in 21 days". That
"21 days" is ridiculous but nonetheless the book teaches a good style
and the book includes a Microsoft C++ 6 compiler. The chapter about
input/output is a little bit scarce.


S. Gonzi


From peter at engcorp.com  Thu Apr 11 01:41:49 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 11 Apr 2002 01:41:49 -0400
Subject: Thanks and kudos to Python core team (was Re: Booleans, integer  
 division,backwards compatibility; where is  Python  going?)
References: <mailman.1018485447.10170.python-list@python.org> <3CB4EC5A.691F1F7B@engcorp.com> <Xns91ECDF3AC49F7lumberjack@216.148.53.85>
Message-ID: <3CB5221D.EE464B76@engcorp.com>

Lumberjack wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote:
> > I'm dismayed by comments attacking the core developers when made
> > by people who have not contributed more than I or my team have.
> > Beggars can't be choosers, and nobody likes a whiner...
> 
> This incorrectly labeled "troll" has made small monetary and other
> contributions to the advancement of Python. And a larger investment in
> using the language.

I suspect I've made a rather significantly larger investment in using the
language, and yet I'm willing to keep my yap shut most of the time
on language issues with which I have some disagreement because 
I haven't made any significant contributions.  It sounds like you
have not either.

In any case what's your point?  You have every right, and 
apparently the strong motivation, to fork off your own branch
of the language and continue development in your own fashion.
Surely that will let you demonstrate how to bring legitimacy
in ways the current project leads have not been able...

So I suggest you take my advice, and fork off.

-Peter


From jb at cascade-sys.com  Mon Apr 29 16:04:55 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 13:04:55 -0700
Subject: Comment on draft PEP for deprecating six builtins
References: <007401c1efb0$91d9d7e0$445d4540@Dell2>
Message-ID: <3CCDA767.21BDA0DD@cascade-sys.com>

Brian Quinlan wrote:

> Decide for yourself:
>
> 1 a : of or relating to an enemy <hostile fire>
>   b : marked by malevolence <a hostile act>
>   c : openly opposed or resisting <a hostile critic>
>       <hostile to new ideas>
>   d : not hospitable <a hostile environment>

Well, he's not an environment so (d) is out.

(c) could apply but it's really more of a metahpore and openly opposed or
resisting is legit, right?  Frank opinions are solicited and provided -- par
for the course -- the system works.

(a) doesn't apply, as he's doesn't give any indication that he's your enemy
that I can tell.  We're all in this together and he clearly is truly trying
to help.

That leaves (b).  I don't think (b) applies either, as "malevolence" means
wishing violent harm on someone, which was not expressed or implied anywhere
in his note.

Sarcasm hardly counts as malevolence or hostility.  Sarcasm actually can be
quite jovial, in my experience.  Close friends use it all the time, right?
It's akin to humor, though, sadly, easily misinterpreted in a text only
medium.

More importantly, my original point was -- pay close attention here -- if
you think Lumberjack was AT ALL "hostile" you must have led a Very Sheltered
life on the Internet.  I've seen hostile and it's Not Pretty but it's in a
completely different league than this.  After all, Lumberjack did NOT (!)
get into the PEP writers personal grooming habits, his sexual orientation,
number of partners, probable predisposition for bestiality or pedophilia,
his mother's mating or grooming habits, drug use or anything truly
'hostile,'  IMHO.  "Hostile" is a whole 'nother universe...

> I've observed in the past that people don't like to be attacked
> personally.

Well duh.

But I gotta say some people here act like kindergarden kids, "Boo Hoo Hoo,
mommy, mommy the bad man used Sarcasm."

> Then Lumberjack should attack the PEP, not the author e.g.
>
>    "How about you and other PEP writers take some responsibilities
>    for your fantasies for once? You up to helping thousands rewrite
>    their code? I thought no"
>
> Seems like an attack on PEP writers in general, not on the PEP itself.

I agree Lumberjack's post would have been completely above reproach and
otherwie Perfect if he left out this paragraph.

Nevertheless, rather than an attack on the PEP writer the sarcasm can bee
seen as a further illustration of why the PEP itself was so wrong-headed.
As originally proposed, I agree the near complete lack of credible
motivation for the PEP made it a "fantasy" (in a bad way).  And in this the
writer carelessly overlooked the responsibility to the tons of existing code
and to the user base (most of whom would simply be puzzled by the change).

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From boris.huart at mathworks.fr  Wed Apr 17 04:04:17 2002
From: boris.huart at mathworks.fr (Boris Huart)
Date: Wed, 17 Apr 2002 10:04:17 +0200
Subject: tcp/ip server
Message-ID: <3CBD2C81.B8F23E03@mathworks.fr>

Hi all,

I am developping a TCP/IP server in python 1.5.2 and I would like it to
send data (mainly floats) to a Java applet. Does anyone know how to make
it work out ? For the moment the server understands the requests (in
strings) from the applet but the
applet doesn't understand anything from the server...

TIA,

Boris



From Oschler at earthlink.net  Tue Apr  2 09:41:28 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 2 Apr 2002 09:41:28 -0500
Subject: IDLE for SUSE Linux?
References: <Yc0q8.1638$0T6.43440@e3500-atl1.usenetserver.com> <3CA8D724.336A22BA@3captus.com> <5waq8.5401$ku5.476012@e3500-atl2.usenetserver.com> <3CA98091.4B8C83A2@3captus.com>
Message-ID: <6qjq8.10106$ku5.757357@e3500-atl2.usenetserver.com>

Bernard,

Just checked, yes I did.

thx

Bernard Yue <bernie at 3captus.com> wrote in message
news:3CA98091.4B8C83A2 at 3captus.com...
>
> Have you installed python-tk?  That maybe the reason.  IDLE needs
> Tkinter.
>
>






From ajs at ix.netcom.com  Fri Apr 26 09:27:20 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Fri, 26 Apr 2002 09:27:20 -0400
Subject: Steve Holden wrote: [was Python-list digest, Vol 1 #10572 - 14 msgs]
Message-ID: <001d01c1ed26$26f1b740$0334fea9@carol>

Steve had written - 

>Many c.l.py readers are *not* "professional programmers", but 
>chemists, mathematicians, biophysicists, and a host of others 
>for whom programming is a means to an end. Many others are, 
>like me, professional programmers who attempt to
>increase Python's popularity by assisting those less familiar with
>programming concepts.

I think this is honestly (and revealingly) worded, represents 
increasing the most consistent thread of your posts, and 
is its own kind of problem.

The attempt to "increase Python's popularity by" becomes too 
transparent. Better a side effect, than a motive. Or play it
smart enough that it seems as such.

Reprimanding a long-time Python contributor for the wording
of a post that you fear will spread FUD was strange enough.
But there a consistency.

Steve - I bought and enjoy your book and do consider it a 
service to the non-professional community.

But the cheerleading and defensiveness on cl py has
in IMO gone a bit over the top, and is a service to
no one.

Accept, this is Usenet.  I can say dugapoopoo if I want. 

Art











From shalehperry at attbi.com  Fri Apr 12 03:33:02 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 12 Apr 2002 00:33:02 -0700 (PDT)
Subject: Beginner's performance problem
In-Reply-To: <memo.20020411143211.1400B@a.a>
Message-ID: <XFMail.20020412003302.shalehperry@attbi.com>

> 
> Aha! I'd put off reading up on the regular expression stuff until later as 
> I thought it just did searches, as opposed to also doing in-place 
> replaces. Half an hour spent on the HOW-TO and I'm a lot better informed. 
> Add a couple of r'\b' to the above, and I ought to be laughing. 
> 

test and triple test your \bfoo\b regex's.  They sometimes mean things you did
not realize.  For instance re.compile(r'\bengine') would match 'regex-engine'.




From logstx at bellatlantic.net  Sat Apr 27 16:17:02 2002
From: logstx at bellatlantic.net (logistix)
Date: Sat, 27 Apr 2002 20:17:02 GMT
Subject: handling different arguments depending on first argument
References: <3CCAF77B.3BEBF259@fxtech.com>
Message-ID: <2HDy8.6660$iJ.4510@nwrddc02.gnilink.net>

>>> def fun(required, *optional, **keywords):
...  print "Required => %s" % required
...  print "Optional => %s" % repr(optional)
...  print "Keywords => %s" % repr(keywords)
...
>>> fun(1,2,3,4,5)
Required => 1
Optional => (2, 3, 4, 5)
Keywords => {}

>>> fun(1,2,3)
Required => 1
Optional => (2, 3)
Keywords => {}

>>> fun(1,2,3,4, keyword1=4, keyword2=5)
Required => 1
Optional => (2, 3, 4)
Keywords => {'keyword2': 5, 'keyword1': 4}

--
-

"Paul Miller" <paul at fxtech.com> wrote in message
news:3CCAF77B.3BEBF259 at fxtech.com...
> I've written a C function (callable from Python) that can take a
> variable number of arguments, and the arguments after the first are
> dependent on what the first was. For example, the signatures for 2
> "modes" might look like this:
>
> my_func(MODE1, var1, var2, var3=None)
> my_func(MODE2, var1, var2=None)
>
> etc.
>
> Is it possible to handle this kind of different calling convention in
> pure Python? I'd like to write the equivalent of an "overloaded" Python
> object constructor, which takes various arguments depending on the
> first.




From ruediger.maehl_nospam at web.de  Wed Apr  3 01:44:55 2002
From: ruediger.maehl_nospam at web.de (Ruediger Maehl)
Date: Wed, 3 Apr 2002 08:44:55 +0200
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com>
Message-ID: <newscache$hzbztg$btb$1@www-neu.dzsh.de>

"Paul Boddie" <paul at boddie.net> schrieb im Newsbeitrag
news:23891c90.0204020644.3ea4bc40 at posting.google.com...
> I'm not sure I would second that recommendation unless SAP DB has
> changed dramatically over the past few weeks. I found that the
> installation was quite a bit more complicated than, for example, MySQL
> (at least on Windows) and upon issuing the arcane commands for

Yes, you need to get used to the commands. But by now there is a
database manager (client/server, web) available, which supports
you in daily routine.

> restarting database instances, the software would complain that the
> instances needed restarting, which was hardly very reassuring with
> respect to either the stability or the usability of the software.

Sorry, but I don't understand this statement.

> As for the Python DB-API module, I couldn't get any version of it I
> found to work with the database system. Admittedly, my objective was
> to get a JDBC connection up and running, which was eventually
> successful, and I therefore cut that particular side-investigation
> short.

I had no problem to use the native interface.

> Despite it's often-cited shortcomings, MySQL and the MySQLdb module
> are quite easy to set up. On UNIX systems, PostgreSQL and pgPySQL are
> possibly as easy, although one doesn't need to suffer a manual with
> PostgreSQL which consists of a single >1.5MB HTML file. I've also had
> success with Sybase Adaptive Server Anywhere and mxODBC (without a
> driver manager), but that software is commercially licensed and may
> not even be available any more.

Compared to SAPDB, MySQL and PostgreSQL are well developed
over a long time and there are more tools available. IIRC, MySQL has had
a 2GB limit for the size of a table and was not free and PostgreSQL
needed some Cygwin thing to run (on W*ndows). That were the reasons
for me to use SAPDB and it is in production for 14 months now without
any problems. Anyway, YMMV.

> [snip]
>
> Paul

R?diger




From m_mariappanX at trillium.com  Wed Apr 24 00:43:29 2002
From: m_mariappanX at trillium.com (Mariappan, MaharajanX)
Date: Tue, 23 Apr 2002 21:43:29 -0700
Subject: BLT/Python PMW problem
Message-ID: <53A7943A5BD8D411B6930002A5073155022D7480@bgsmsx90.iind.intel.com>

Hi Sridhar,

Do you able to import blt in python shell? 

First build BLT 2.4x, and then rebuild python with tkinter+blt module is
enabled in Modules/Setup file and retry this. I able to run blt widgets
sucessfully. But i'm not able understand what r u tring... by below lines.

>>>> from Tkinter import *
>>>> m=Tk() 
>>>> m.tk.call('package', 'require', 'BLT')


Maharajan 

>-----Original Message-----
>From: Sridhar Anandakrishnan [mailto:sak at essc.psu.edu]
>Sent: Tuesday, April 23, 2002 8:40 PM
>To: python-list at python.org
>Subject: BLT/Python PMW problem
>
>
>Hi, 
>
>When I try and load BLT from within Tkinter, I get this error:
>
>Python 2.2.1 (#1, Apr  9 2002, 13:10:27) 
>[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> from Tkinter import *
>>>> m=Tk() 
>>>> m.tk.call('package', 'require', 'BLT')
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>TclError: couldn't load file "/usr/lib/libBLT24.so":
>/usr/lib/libBLT24.so: undefined symbol: TclpAlloc
>>>> 
>
>Any idea what's up?  I don't get this problem if I do this 
>directly from
>Tcl, so it is some sort of path problem?  I don't know Tcl well enough
>to track this down - any help appreciated.
>
>(python 2.2.1, tcl/tk 8.3.4, blt 2.4x)
>
>Sincerely,
>Sridhar.
>-- 
>http://mail.python.org/mailman/listinfo/python-list
>




From grante at visi.com  Mon Apr 15 23:19:12 2002
From: grante at visi.com (Grant Edwards)
Date: Tue, 16 Apr 2002 03:19:12 GMT
Subject: Does Python need a semantics for and/or over generators ? was : 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <3CBABA71.3040909@geneva-link.ch> <YwCu8.32367$vm6.5103753@ruti.visi.com> <yNDu8.13611$Ji7.6515@atlpnn01.usenetserver.com>
Message-ID: <slrnabn631.1iq.grante@tuxtop.visi.com>

In article <yNDu8.13611$Ji7.6515 at atlpnn01.usenetserver.com>, Steve Holden wrote:
> "Grant Edwards" <grante at visi.com> wrote in message
> news:YwCu8.32367$vm6.5103753 at ruti.visi.com...
>> In article <3CBABA71.3040909 at geneva-link.ch>, Boris Borcic wrote:
>>
>> > For the shortcut semantics, we'd like to have :
>> >      a xor b
>> >
>> > behave exactly like :
>> >      (not b and a) or (not a and b)
>>
>> I don't understand how that's a shortcut (assuming that means
>> the same thing as short-circuit).  You have to evaluate both
>> operands of an xor operator.  There is no possibly way to
>> short-circuit the expression so that only one is evaluated.
>
> But they each might be evaluated *twice* in the given
> equivalent expression if the existing Boolean operators didn't
> short-circuit.

I see.   Of course if there _were_ an xor operator, that
wouldn't be a problem...

-- 
Grant Edwards                   grante             Yow!  I will establish
                                  at               the first SHOPPING MALL in
                               visi.com            NUTLEY, New Jersey...


From Scott.Daniels at Acm.Org  Sun Apr 14 17:32:27 2002
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Sun, 14 Apr 2002 21:32:27 GMT
Subject: currying (was: Re: A trivial question about print)
References: <a93oqk$jf6$1@bagan.srce.hr> <3CB58302.5090200@mxm.dk> <kTft8.6500$b62.159068@news1.tin.it> <Xns91EDDBF2DF1FAcliechtigmxnet@62.2.16.82> <3CB5FF36.6060300@mxm.dk> <Fgwt8.8607$b62.224911@news1.tin.it> <slrnabe72t.l8u.huaiyu@gauss.almadan.ibm.com> <3CB756E4.5030404@jerf.org> <slrnabesm3.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <1103_1018819947@news.dsl-only.net>

On Fri, 12 Apr 2002 23:50:27 +0000 (UTC), huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote:
> Jeremy Bowers <newsfroups at jerf.org> wrote:
> >Has anyone done any work on creating a more general curry function? One 
> >that might for instance curry the second of three positional arguments? 
> >Using the above definitions:
> >f2 = curry2(f, a, c)
> >f2(b)
> 
> In general you must have a way to indicate a subset.  Like
> g = curry([2, 3, 5], f, a1, a4, a6)
> g(a2, a3, a5)
> ...
I had kind of thought that the keyword args would take care of this.  
Sure there are cases where this wouldn't be true, but  I'd hate to have 
to read code that included ... curry([2, 3, 5], f, a1, a4, a6) ..., I'd have 
the devil's own time determining which args were specified.  IMHO the 
real questions are (1) front or back (default last or first), and (2) call-time 
vs. curry-time keyword superceding.  You of course know which I'd 
choose (the ones in the recipe).

On issue (1), providing finals args back to front interferes with keyword args, 
so I'd forget that.  Providing call-time before curry time in a "flexible" order is 
worse: you cannot read the curry-time args and know what is covered.  I 
concede most programming styles leave the "fiddley" args -- perfect curry 
fodder -- at the end of the argument list.  Even so, I'd prefer the earlier args be 
the curry args, since you more readily know what args are being consumed.
I'd argue in favor of making those "fiddley" args keyword args.

Issue (2) is more debatable: call-time seems to me more pythonic, but I can 
see logic for either side.  The keyword stuff can be seen as setting defaults
if call-time supercedes (which is useful in and of itself).  More than anything 
else, this is what tips me in favor of call-supercedes-curry.  Raising an 
exception on keyword conflict is another possibility, but this seems more 
like the strong typing / declared constants view: don't trust the user to do 
it right.

Of course I'd love to see curry show up in the built-ins (I always forget to 
import it before I use it).

As to Alex's comment on the recipe -- of course he's right; I just hadn't seen 
the possibility of curry by closure until nested scopes showed up after I
had written curry.

-Scott David Daniels
Scott.Daniels at Acm.Org




From BPettersen at NAREX.com  Fri Apr 26 17:23:49 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 26 Apr 2002 15:23:49 -0600
Subject: String handling bug in Python
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com>

> From: Tim Peters [mailto:tim.one at comcast.net] 
> 
> [Stephen Ferg]
> > ...
> > I would like to see Python support TRULY raw string 
> literals -- string 
> > literals that are not escaped in any way.  Perhaps they 
> could be added 
> > with a prefix of z (for zero escaping?), so that
> >
> >      z"\" == "\\"
> >      z"\'" == "\\\'"
> 
> The first one can't be spelled at all with a raw string 
> today.  The second one can be, and indeed the same way you 
> spelled it, as r"\'".  What do you do in your scheme if you 
> want a z-string to contain a single double-quote? 

I always thought that could be spelled r'"' or r'''"''', but I guess
there would still be a problem if you wanted all of ", ', ''', and """
in a string. I would think that would occur less frequently than \ at
the end of a string though (especially with windows path names). I feel
I must be missing something?

Confused'ly y'rs
-- bjorn




From phd at phd.pp.ru  Mon Apr  1 10:51:30 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 1 Apr 2002 19:51:30 +0400
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F5612@MAYOR.retzlaff.com>; from jimmy@retzlaff.com on Sun, Mar 31, 2002 at 11:28:29PM -0800
References: <A7B5C22F20EE3C4E95744E2F03F1887F5612@MAYOR.retzlaff.com>
Message-ID: <20020401195130.B13742@phd.pp.ru>

On Sun, Mar 31, 2002 at 11:28:29PM -0800, Jimmy Retzlaff wrote:
> > Richard Jones [mailto:rjones at ekit-inc.com] wrote
> > Does anyone have a copy of kjbuckets.c that they could send me?
> 
> Oleg Broytmann has it on a page mirrored in a couple of places. Look at
> the bottom of either:
> 
> http://phd.by.ru/Software/Python/
> or
> http://phd2.chat.ru/Software/Python/

   I am not a maintainer, just a holder. I rely on the help of people who
update the source and provide binaries.

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



From tdelaney at avaya.com  Mon Apr 29 04:37:26 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 29 Apr 2002 18:37:26 +1000
Subject: Comment on draft PEP for deprecating six builtins
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2A8@natasha.auslabs.avaya.com>

> From: Alex Martelli [mailto:aleax at aleax.it]
>         ...
> > I also personally don't think lamdas should be removed. 
> However, if a
> > lambda exceeds a (BDFL-pronounced) number of opcodes it 
> should be a syntax
> > error <0.5 wink>. This should reduce most lambdas to one or two
> > expressions.
> 
> All lambdas are alredy "reduced" to exactly one expression.  Somehow
> constraining that expression to be "simple" is a laudable 
> goal, but -- I
> dunno: it seems to me that putting further constraints, and moreover
> ones hard to verify precisely for a programmer who wants to 
> be thinking
> about Python source code, not about bytecodes that come out, is hardly
> the way of solving lambda's woes.

Indeed - I was in no way serious with the suggestion. However, I was
sideways-suggesting that lambdas should always be *short*.

Hell - I even used a <0.5 wink> which is a construct I never normally use
...

Tim Delaney




From claird at starbase.neosoft.com  Fri Apr 12 16:57:22 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 12 Apr 2002 15:57:22 -0500
Subject: Python and spacecraft onboard or ground control scripting
References: <7226483f.0204101216.2c8fcda4@posting.google.com> <7E283A6EA4F0DBD6.B68D98BE8C574F5D.CE8B287F1A68C959@lp.airnews.net> <LFGz2JAYZxt8EwR6@jessikat.demon.co.uk>
Message-ID: <459A680B92F06092.DD5B3EAA28701812.3A8CCB250DB66074@lp.airnews.net>

In article <LFGz2JAYZxt8EwR6 at jessikat.demon.co.uk>,
Robin Becker  <robin at jessikat.fsnet.co.uk> wrote:
>...
>well there's certainly been some efforts with Tcl eg
>
>s. "Tcl and Concurrent Object-Oriented Flight Software: Tcl on Mars",
>by David E. Smyth, Mars Pathfinder Flight Software Team, Jet
>Propulsion Laboratory.
>
>        Describes the use of Tcl on the "Mars Pathfinder" project, in
>        which Tcl (with several extensions) is used in several ways:
>        as the language for inter-object messages within the
>        object-oriented flight software; as the implementation
>        language for the monitor and control environment; and as the
>        initial implementation language for some of the flight
>        software itself.
>
>but I'm not sure that Tcl actually went to Mars.
>-- 
>Robin Becker

Folklore has it that Tcl *did* travel to Mars, and its
interpretive nature was crucial in responding to some
mission contingency; the control team needed to upload
new code during flight.  We ought to document this ...

While I understand the observations that Python, Tcl, 
and so on surely aren't safe enough for medical devices,
nuclear plants, air traffic control, military devices,
and so on, I have reasonably certain knowledge of instances
of each of these.
-- 

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


From andrewm at object-craft.com.au  Tue Apr 30 02:03:17 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Tue, 30 Apr 2002 16:03:17 +1000
Subject: Python vs. Perl, which is better to learn? 
In-Reply-To: Your message of "29 Apr 2002 22:11:31 MST."
             <df30afd4.0204292111.31e5914a@posting.google.com> 
Message-ID: <20020430060317.F3BE638F50@coffee.object-craft.com.au>

>I have a lot of experience in ANSI C/C++, I want to learn a new
>language and which it can help me in the work.
>I just want to clear which(python or perl) is more comfort for me?

I've been using (and abusing) perl for many years, and I'm still not
comfortable using it. Not only that, but because there are so many ways of
expressing things in perl, I often can't understand my own code 6 months
down the track (at least, at first glance). I'd go as far as to say the
sendmail.cf language is simpler, easier to learn and maintain than perl.

Python on the other hand is a lot of fun. Good designs tend to happen
more often and with less effort, and the language doesn't get in the way
(it works the way you expect it to work).

And, in fact, Python has changed (for the better) the way I code C - I
want the dynamicism of Python, so I spend more time coding infrastructure,
which pays off when the inevitable "can you make it do this" comes along.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/




From sdeibel at wingide.com  Wed Apr 10 09:12:00 2002
From: sdeibel at wingide.com (Stephan R.A. Deibel)
Date: Wed, 10 Apr 2002 09:12:00 -0400 (EDT)
Subject: Help with Python advocacy and marketing
In-Reply-To: <20020410063636.2732.6090.Mailman@mail.python.org>
Message-ID: <Pine.LNX.4.30.0204100849350.905-100000@pangolin.wingide.com>

>> * We are collecting user-side advocacy quotes for use in press
>>   releases announcing future versions of Python.  If you have
>>   a compelling statement about how Python has helped your
>>   project or business, this is a great way to help Python and
>>   get some media exposure in the process.  We can't guarantee
>>   your quote will be used in a press release, but with your
>>   permission it will be included on a Python advocacy website
>>   that is under development.
>
>Some companies might have difficulty allowing their name to
>be used by "just anyone" in the context of Python advocacy.
>Can you describe more about the way in which this information
>will be distributed, who is represented (is this a Zope Corp
>or PythonLabs sponsored endeavour?), and so on?  My company
>(not engcorp.com) has so far remained in the closet about
>its use of Python and although it would be nice to get
>permission to help with a little "XXXX company uses Python
>in the following ways..." I know I couldn't begin to get
>permission without more background.

The press releases will come from the Python Software Foundation (PSF),
which is the non-profit that holds the intellectual property rights to
Python.  I've volunteered to help produce and distribute the releases.

We know about company's policies on endorsing products but have been
successful with a number of prominent companies to obtain quotes
because they recognize Python advocacy is a good thing for its users,
and they get a little bit of free PR as part of the deal.

If you do ask about this at your company, make sure to stress that Python
is open source and not a commercial product, and that the PSF is
non-profit.

The final press release that contains each quote is always given to the
company's PR dept for approval before it's released.  I can furnish
a draft press release as example on request from those that might be able
to do a quote.

Once approved, the releases will be distributed to a range of media
contacts and business wire services.  The goal is writing to media
contacts is to try to get more wide-spread mention of Python in printed
and online technical periodicals.  The wire services categorize releases
and distribute them on a technical news feed, usually globally.  These get
monitored and searched by keyword, and similarly may result in additional
exposure in technical publications.

Finally, if approved by the company, the quote will be added to a page
on the python.org website.

Note that some of the other efforts I mentioned are not yet specifically
associated with the PSF.  I've set up pythonology.org to become an
advocacy site for Python, with the vision that it will later include the
Success Stories and Quick Start guides, but there's no requirement that
materials reside there if contributors want them to be elsewhere.

Hope that helps to clarify.

- Stephan

------------------------------------------------------------------------
Wing IDE for Python                          Archaeopteryx Software, Inc
www.wingide.com                              Take Flight!





From bokr at oz.net  Fri Apr 12 17:54:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Apr 2002 21:54:17 GMT
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com> <kZbt8.5601$b62.137821@news1.tin.it> <3cb5bd03$1_6@goliath.newsgroups.com> <slrnabcl3k.c8l.mlh@vier.idi.ntnu.no> <mailman.1018592665.14227.python-list@python.org> <3CB72CE6.DFAD5C10@python.org>
Message-ID: <a97l29$hnv$0@216.39.172.122>

On Fri, 12 Apr 2002 18:49:17 GMT, Guido van Rossum <guido at python.org> wrote:

>"David Mertz, Ph.D." wrote:
>> Unfortunately, there's more to it than this.  There still is--and
>> presumably always will be--a difference between "instances" and other
>> types of things.  For example, a list is a mutable sequence that has
>> some items in it.  In contrast, an instance object--whether new style or
>> old style (and w/ or w/o __slots__)--is a thing that has some
>> attributes.  The basic shape of the data will always be different.  The
>> type/class unification doesn't change that.
>
>Hm...
>
>So after
>
>    class C(list):
>        def appendmany(self, *args):
>            for arg in args:
>                self.append(arg)
>
>Do you consider a C instance to be a list or an instance?
>You might look at the presence of a __dict__ attribute
>storing instance attributes as the distinguishing factor.
>But then what would you do after
>
>    class D(list):
>        __slots__ = ["foo"]
>
>List or instance?  The shape of a D instance is the same
>as the shape of a list, but with an extra slot added.
>
ISTM what matters most of the time is how you can use it, not
what its exact parentage is. E.g., people (ab)use type(x) to
try to check if an argument to a function should be used one
way or another way. I'm wondering if we're heading towards
a hascapability(thing, cap) kind of general query, as a useful
standard set of capabilities objects can be endowed with evolves?

hasattr(c, '__getitem__') etc. can answer some of those questions,
but if you had an official enumerated set of capabilities, you could
have a __capabilities__ slot and check with a mask (e.g., against
constants in __builtins__, e.g., INDEXING).

For your class above people might write hascapability(c, INDEXING) or
c.__capabilities__&INDEXING rather than type(c) is list (which fails)
or isinstance(c, list) which might be unnecessarily narrow for the
generic capability (which a non-list-based custom object might support).

BTW, A standard capabilities/protocol query method might be something to
reserve a mask bit for, for custom extension of the concept.

Regards,
Bengt Richter


From tdelaney at avaya.com  Wed Apr  3 02:11:04 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 3 Apr 2002 17:11:04 +1000 
Subject: PEP 287: reStructuredText Standard Docstring Format
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A249@natasha.auslabs.avaya.com>

From: Paul Rubin [mailto:phr-n2002a at nightsong.com]
> Richard Jones <rjones at ekit-inc.com> writes:
> 
> That looks pretty similar to an Info file (texinfo output).  I think
> I'd rather type "@section{Overview}" than have to supply those special
> underscores directly.

The whole point is that this is the *input*, optionally to produce another
format, and is also human-readable in the source code (or any other text
file). Your statement above simply supports the use of ReST, as its native
format is "pretty similar" to texinfo output, which you have stated
previously that you prefer to the texinfo source.

> > Note that my manual Table of Contents at the top is now 
> > possible using a directive::
> > 
> >    .. contents:: **Table of Contents**
> > 
> > The output is at http://roundup.sf.net/doc/installation.html
> 
> So the ink isn't even dry on this PEP and already it's sliding down
> the slippery slope into explicit markup <wink>.

Huh? If you'd read the accompanying docs, you would know about the ReST
directives. I wouldn't call the .. contents:: directive "explicit markup"
considering that it inserts an entire table of contents with one short line
...

> I don't find Javadoc unreadable (it's just HTML) but it's certainly
> ugly and I'd rather read a formatted version.

Exactly! With ReST you can read either the original version, or a formatted
version, and *neither* is ugly. In fact, while I was reading the
accompanying docs I preferred to read the text (ReST) version rather than
the HTML version.

> > Again, have you _read_ the ReST documentation, specifically the
> > discussion of alternatives? It's extensive, and very well
> > written. It discusses many existing STX systems and their strengths
> > and weaknesses.

Personally, I got lost a few times, but it was my first reading of the docs,
and there is a lot of information there.

> I looked briefly at the ReST documentation linked from the PEP, but
> if I'm supposed to read it in detail to be able to use it, I might
> as well use a more traditional markup language.

I looked briefly at the Python documentation linked from the web site, but
if I'm supposed to read it in detail to be able to use it, I might as well
use a more traditional programming language.

I don't think I need to say more.

Tim Delaney



From aleax at aleax.it  Wed Apr 24 08:05:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 24 Apr 2002 12:05:46 GMT
Subject: Detecting a new style class
References: <mailman.1019552082.8541.python-list@python.org> <aa48nq$a61$1@panix1.panix.com> <mailman.1019640045.27779.python-list@python.org>
Message-ID: <ucxx8.18368$8D3.509045@news1.tin.it>

<posted & mailed>

[Not following c.l.py these days, but thanks to lac for pointing this 
out!}:

Graham Ashton wrote:
        ...
> Okay, good point. I'm using the Alex's Borg pattern in place of a
> singleton, as shared state will do quite nicely. I'm also trying to
> protect myself from getting caught out by new style classes (in other
> words it doesn't work -- in general -- for classes derived from, for
> example, dict).

Right: Borg in its simplest incarnation is based on an instance's state
being all kept in its __dict__, and a new-style instance may have
per-instance state elsewhere (e.g. by subclassing list or dict).

> An alternative question would be -- does anybody know how to do the Borg
> so that it works for new style classes?

http://www.aleax.it/Python/5ep.html explores the issue a bit and
settles on automatic delegation (DeleBorg) as the most solid
solution.  I'm still not quite convinced that it's the best solution,
but if you cannot constrain the classes to be Borg-ized, it's hard
to think of another.

If you CAN constrain the classes in question, then I think that
checking for the metaclass is the right approach.  E.g. if you only
want to accept a class K that's a "classic class", that's the same
as saying that type(K) must be types.ClassType -- not the
builtin 'type', or something yet more devious.


Alex



From theoryboy at my-deja.com  Sun Apr 21 10:43:54 2002
From: theoryboy at my-deja.com (Peter Saffrey)
Date: 21 Apr 2002 07:43:54 -0700
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com> <slrnac349u.ea.usenet@fohnhab.minus1.de>
Message-ID: <ced73313.0204210643.2d88ac82@posting.google.com>

Perhaps I should have said that I need this in an object-oriented
context. It doesn't seem to work using an anonymous lambda function to
a class method: example below.

Sorry, I didn't realise it would make any difference, and I was trying
to make my example concise.

Peter

#!/usr/bin/python

from Tkinter import *

class gui:
	def __init__(self):
		win = Frame()
		win.pack()
		Button(win, text='print 3',
command=lambda:self.printNumber(3)).pack()
		Button(win, text='print 5',
command=lambda:self.printNumber(5)).pack()

		win.mainloop()
		
	def printNumber(number):
	        print number
        
if __name__ == '__main__':
	test = gui()


From just at xs4all.nl  Wed Apr 17 15:00:47 2002
From: just at xs4all.nl (Just van Rossum)
Date: Wed, 17 Apr 2002 21:00:47 +0200
Subject: Trinary operator?
References: <mailman.1019066614.29792.python-list@python.org>
Message-ID: <just-4227D7.21004717042002@news1.xs4all.nl>

In article <mailman.1019066614.29792.python-list at python.org>,
 "Mark McEahern" <marklists at mceahern.com> wrote:

> > Not that it's a big deal since there's many ways to do that, but can I use
> > something like
> >     verboseGender = (gender == 'm') ? 'male' : 'female'
> > in Python?
> 
> People seem to use and/or for that:

But they shouldn't.

Just


From lenssen at hitnet.rwth-aachen.de  Wed Apr 17 16:09:02 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Wed, 17 Apr 2002 22:09:02 +0200
Subject: preg_replace?
Message-ID: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE>

I am porting a program that I already wrote in VBScript and PHP.
My current task: I need to replace everything inbetween square brackets with
"0". In VBScript I did this "by hand" (looping, inString and so on). In PHP
somebody helped me with the following solution:

    $s = preg_replace("%\[[^\]]*\]%", "0", $s);

Even though I've been reading through regex-tutorials in the meantime, I'm
still not familiar with it. But is there a simple rule how to convert the
above into Python code? Not only would it solve my current problem but I
already got some helpful regex's for PHP and could reuse them in the future.

Anyway, the best (or worst) I did so far is:
    pattern = re.compile(r'(\[[^\]]*\])')
    pattern.sub('0', s)

I'm not convinced this remotely resembles how it's done in Python; it
doesn't cause an error, doesn't do what I want either.

Here's my hoped before after:
    s = "Hello [test] world [bla bla]..." # before regular expression
    s = "Hello 0 world 0..." # after




From emile at fenx.com  Wed Apr 10 10:46:58 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 10 Apr 2002 14:46:58 GMT
Subject: Create Excel Files without COM
References: <3CB44516.60005@thomas-guettler.de> <slrnab8hp2.3r2.stefan.heimann@kunz.ratzer> <3CB44BE2.6080509@thomas-guettler.de>
Message-ID: <CfYs8.204364$Yv2.67473@rwcrnsc54>

Thomas Guettler"
> CSV or the XML Version of an Excel Spreadsheet would be very easy
> to produce. But it would be better if it would be .xls
>

If you tab delimit the data and name it myfile.xls, current versions of
excel will 'do the right thing'

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From logiplexsoftware at earthlink.net  Mon Apr 29 20:45:53 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 17:45:53 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <Xns91FFAE356DC8lumberjack@216.148.53.84>
References: <007001c1efa7$04ede140$445d4540@Dell2>
	<3CCD9649.98AED86F@cascade-sys.com>
	<mailman.1020120416.1514.python-list@python.org>
	<Xns91FFAE356DC8lumberjack@216.148.53.84>
Message-ID: <20020429174553.54923b75.logiplexsoftware@earthlink.net>

On 30 Apr 2002 00:07:17 GMT
Lumberjack wrote:

> No doubt you would stay friendly with someone who dislikes the color of the 
> electrical wire strung through the walls of your house and seriously 
> proposes (against your wishes) to level your house and rebuild it simply to 
> put in different colored wiring. And you are paying for all this, not them. 
> That is either being very gentlemanly or very stupid. However, I see no 
> reason follow your misguided attempt at suffering fools gladly.

He doesn't want to suffer fools, hence his remarks about people who flame
rather than give rational reasons for their position.  You have good reasons
for being against the PEP, so state them.  No reason to throw a tantrum.


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From claird at starbase.neosoft.com  Wed Apr 24 14:07:03 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2002 13:07:03 -0500
Subject: Finding scripts
References: <af916b72.0204240711.41f125f8@posting.google.com>
Message-ID: <09494EE987520275.3B1A5BCECE14B040.72143909819FBF9D@lp.airnews.net>

In article <af916b72.0204240711.41f125f8 at posting.google.com>,
Tom Hendricks <tchendricks at lucent.com> wrote:
>I am a newbie to the python world.  I am in the process of configuring
>about 300 switches with basically the same information except for IP
>address and VLAN information.
>
>It was suggested to me that I download python and use expect to create
>a script to perform this work more efficiently and faster.  This is
>great in theory if you have programmed before but I think it would be
>faster to manually configure the switches than to learn python at this
>time, short-term thinking, of course.
>
>Can someone direct me to where I might find scripts for configuring
>Enterasys Smartswitch 6000, Smartswitch 2200, and Matrix E7's?
>
>In the meantime, back to my Learn to Program using Python: Lesson 1,
>Getting Started by Richard Baldwin.
			.
			.
			.
It certainly is technically possible to achieve your
goal by working with Python and Expect.

I think someone did you a disservice, though, in sug-
gesting that combination.  More appropriate is simply
to work entirely in Expect <URL: http://
www-106.ibm.com/developerworks/library/l-sc1/?n-l-4112 >.
As fond as I am of Python, I think you've considerably
magnified your burden by attempting to put it in the
role you describe.

It is very unlikely that you will "find scripts for
configuring Enterasys ..." and the others.

Perhaps you would be best off to buy a tiny amount of
consulting--maybe a couple of hundred dollars--from
someone who can give you a few "starter" scripts spe-
cific to your equipment.

Thousands of Lucent employees have experience with
Expect (among other languages and tools).  Are none
of them in a position to help you?
-- 

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


From aleax at aleax.it  Fri Apr  5 04:35:59 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 09:35:59 GMT
Subject: Diff in lists
References: <3CAD6AB2.1D4E9846@ifi.uio.no>
Message-ID: <3eer8.51687$pT1.1467954@news1.tin.it>

Kristin Skoglund wrote:

> Hi
> 
> Is it possible to run a diff-function on lists in Python?

Sure, see the difflib module in the standard Python library,
http://www.python.org/dev/doc/devel/lib/module-difflib.html .

(Please don't post questions both here AND on help at python.org:
just-here is fine, unless you prefer privacy).


Alex



From joshualevy at yahoo.com  Tue Apr  2 16:05:24 2002
From: joshualevy at yahoo.com (Joshua Levy)
Date: 2 Apr 2002 13:05:24 -0800
Subject: portable subprocess with return value and path search
Message-ID: <5fa44575.0204021305.600a987@posting.google.com>

I'm trying desperately to get python to do the following
four things at the same time:
    1. Run a subprocess (like "java com.foo.bar arg")
    2. Get the process return value from that subprocess
    3. Work on both Windows and UNIX  (I'll settle for 
          post-Win2000 and Solaris/Linux.)
    4. Resolve the executable, based on the current path.
       (So I only need "java", and not "c:\jdk1.2.2\bin\java",
       in the example above.)
Os.system does not return values from Windows.  The versions of
Os.spawn on Widnows do not resolve executables.
Thanks in advance for any help.

Joshua Levy


From h_schneider at marketmix.com  Tue Apr  9 03:53:53 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Tue, 9 Apr 2002 09:53:53 +0200
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com> <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>
Message-ID: <a8u6cc$b78$01$1@news.t-online.com>

Thanks for your very detailed information. On what platform do you use your
software? I have to deal with a mixed environment: Win2K and Win98. Probably
Win98 will bring up some more probs ... :-/

Bye,
Harald

"Chris Spencer" <clspence at one.net> schrieb im Newsbeitrag
news:vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v at 4ax.com...
> Okay, I guess I should chime in here.
> I am a professional developer for a small software company.  My company
> has been using Python for its projects for the past 18 or so months.  We
have a
> whole product suite written in Python.  Until recently, this product was
95%
> console driven, even on our Windows port.  Customers have been asking for
a GUI
> to tie all the parts of our product together.
> For the past month, I have been developing a GUI Administrator program
> for our product using wxPython.





From gimbo at ftech.net  Mon Apr 22 08:41:26 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Mon, 22 Apr 2002 13:41:26 +0100
Subject: Science And Math  Was: Python's Lisp heritage
In-Reply-To: <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
References: <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <20020422124126.GD23706@andy.tynant.ftech.net>

On Mon, Apr 22, 2002 at 12:57:38PM +0100, Gon?alo Rodrigues wrote:

> You should understand that this is a philosophical viewpoint. The
> mathematical field is roughly divided in the Platonist field, those who
> believe that mathematical objects exist somewhere out there in space (to
> quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
> field of those who believe mathematics is essentialy a game we play on
> paper with symbols with no "real" meaning attached to it besides the
> convenience in describing natural phenomena.

To put this another way, ask yourself where you stand on this
question: is mathematics discovered or invented?

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From eppstein at ics.uci.edu  Mon Apr 15 02:59:40 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 14 Apr 2002 23:59:40 -0700
Subject: Macintosh Development
References: <bnEt8.130789$QB1.31528888@typhoon3.we.ipsvc.net> <3CBA5EDB.C60BF821@cosc.canterbury.ac.nz>
Message-ID: <eppstein-A27785.23594014042002@news.service.uci.edu>

In article <3CBA5EDB.C60BF821 at cosc.canterbury.ac.nz>,
 Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:

> MacPython has a facility for creating an "applet" from
> a Python script. This is a small Macintosh application
> that can be launched from the Finder. You can drop
> files on the applet, and their pathnames will appear in
> sys.argv.

Which is great for your own use -- it's what I do with Python programs 
that I'm going to run with much frequency.  But it's not so good for 
distributing programs to others, unless you feel comfortable telling 
them to install the appropriate version of Python before using your 
applet.

MacPython does also have a facility for creating a standalone 
application, but these are much bigger (several megabytes?) and I 
haven't been able to get this to work.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From JamesL at Lugoj.com  Fri Apr  5 21:29:38 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 6 Apr 2002 02:29:38 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org>
Message-ID: <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84>

Cliff Wells <logiplexsoftware at earthlink.net> wrote:
> Ah, you must have chose APL or Forth, since they're pretty much in the
> feature-lock stage.  It can't be C since they've also added the damn
> boolean. 

I am sorry, but no one is tinkering with C and releasing new versions at 
anything near the rate that Python versions are coming out. It is the high 
rate of language change that is one of the causes of my disaffection, not 
change per se. Python is over 10 years old I believe, and still the changes 
are being made at an alarming rate. I don't have all the dates, but 
consider the major releases:

1.0:     ?-?-?
...
1.4:     1996?-?-?
1.5:     1998?-?-?
1.5.2:   1999-04-13   (Final)
2.0:     2000-10-16   18 months after 1.5.2, ~2.5 years after 1.5.
2.1:     2001-04-17    6 months after 2.0
2.2:     2001-12-21    8 months after 2.1
2.3:     2002?-?-?     4 to 12 months after 2.2 if this year.
2.4...:  ?-?-?         Asymptotically approaching zero?

I exclude 1.6 because I believe it was an obligatory release. A newcomer 
will, on average, have a new version to contend with 4 to 15 months after 
their initial introduction. Veterans will need to contend with releases on 
approximately a yearly basis. I have difficulty understanding why Guido 
isn't exercising more discipline in the release schedule.

> Seriously though, if this is a big concern for you (and it might well
> be), target the 1.5.2 language spec and the chances of your code not
> working even in 2.2 are slim, so I'm not sure what would be compelling
> enough to make you abandon Python.

I tried that and that strategy _failed_. The chances of failure was 100%. 
The failure was due specifically to how (as far as I can tell) Tkinter 
handles UTF-8 sequences in Unicode aware versions of Python. So I am forced 
to put version checking into my code! That just added to my irritation. 
There are other _forward_ incompatibilities in Python, but it is the hidden 
ones that misguided programmers are writing into common libraries that make 
the proliferation of Python versions too irritating to deal with.


From anthony at interlink.com.au  Tue Apr  9 01:57:52 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Tue, 09 Apr 2002 15:57:52 +1000
Subject: SSL error: 2.1.1 works but 2.2 doesn't 
In-Reply-To: Message from Walter Miller <w3miller@yahoo.com> 
   of "Mon, 08 Apr 2002 22:54:25 MST." <20020409055425.53965.qmail@web21411.mail.yahoo.com> 
Message-ID: <200204090557.g395vqE02023@mbuna.arbhome.com.au>

>>> Walter Miller wrote
> Apologies if this has been discussed/solved already. 
> I looked for prior postings but haven't found a
> solution for fixing this error I'm getting with
> _socket.pyd on 2.2.

Does it work with 2.1.3? 2.2.1rc2? 


-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From skip at pobox.com  Tue Apr 16 13:17:40 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 16 Apr 2002 12:17:40 -0500
Subject: Brackets in brackets, braces in braces
In-Reply-To: <2e1a250.0204160730.637139f1@posting.google.com>
References: <2e1a250.0204160730.637139f1@posting.google.com>
Message-ID: <15548.23732.172340.995962@12-248-41-177.client.attbi.com>

    Arpad> How can I find strings [using regular expressions] ... Where in
    Arpad> the braces can be also another braces.

In general, you can't.  As Tim Peters and others have pointed out in the
past, regular expressions can't count.  You need a more powerful parser than
re's to properly handle nested constructs.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From claird at starbase.neosoft.com  Wed Apr 24 10:38:01 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Apr 2002 09:38:01 -0500
Subject: variable substitution
References: <e2e9e807.0204240519.688cf3cb@posting.google.com>
Message-ID: <214EF3CE5B8A4E58.715400CDD5BBE254.2689C5F83ABBB06D@lp.airnews.net>

In article <e2e9e807.0204240519.688cf3cb at posting.google.com>,
Hugh <h.e.w.frater at cs.cf.ac.uk> wrote:
>How do I get the %(variable)s thing to work. I need to substitute a
>variable into a print statement so I can send links with sessionId's
>attached back to the browser. I'm trying this at the moment, and It
>says invalid sysntax.
>print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
>Type</a>'
>I saw this in a document written by Guido that I found here:
>http://www.w3j.com/6/s3.vanrossum.html
>
>
>hugh

That document looks to me to have been mangled.  Do <URL:
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66018 >
and  <URL: http://
www.python.org/doc/1.5.2p2/tut/node9.html#SECTION009100000000000000000 >
help?

There ought to be a more canonical-but-readable Web reference
for what David Beazley calls the string "modulo operator"; I'm
open to suggestions.
-- 

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


From lac at strakt.com  Wed Apr  3 12:28:53 2002
From: lac at strakt.com (Laura Creighton)
Date: Wed, 03 Apr 2002 19:28:53 +0200
Subject: Threading problem 
In-Reply-To: Message from Laura Creighton <lac@strakt.com> 
   of "Wed, 03 Apr 2002 19:18:15 +0200." <200204031718.g33HIGrq029056@ratthing-b246.strakt.com> 
References: <a8f8f6$vdu$1@news.rchland.ibm.com>  <200204031718.g33HIGrq029056@ratthing-b246.strakt.com> 
Message-ID: <200204031728.g33HSrrq029129@ratthing-b246.strakt.com>

> Me
> Take out the self in threading.Thread().__init__(self, name=thName)

> Queue.Queue() was designed for exactly what you want to do.  Its dead
> easy to use, as opposed to managind your own List and locks.
> See http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965
> for an example of how to use it.  Or go read Aahz's tutorial.

Ick.  Take out the () as well.  Sorry about that.
Laura





From aleax at aleax.it  Sat Apr 13 11:48:32 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 15:48:32 GMT
Subject: cStringIO write
References: <QHxVGIALCDu8EwLs@jessikat.fsnet.co.uk>
Message-ID: <krYt8.13007$b62.354393@news1.tin.it>

Robin Becker wrote:

> Is it deliberate that 2.2 cStringIO only has a write method when created
> uninitialised? StringIO doesn't have this 'feature', but no mention is
> made of this in the docs.

Very deliberate, I'd say -- they're different types:

[alex at lancelot cb]$ python
Python 2.2 (#1, Dec 23 2001, 20:09:01)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import cStringIO
>>> readonly = cStringIO.StringIO('ciao')
>>> writeonly = cStringIO.StringIO()
>>> type(readonly)
<type 'cStringIO.StringI'>
>>> type(writeonly)
<type 'cStringIO.StringO'>
>>> print type(writeonly).__doc__
Simple type for output to strings.
>>> print type(readonly).__doc__
Simple type for treating strings as input file streams
>>>


And this is indeed covered in the docs:
"""
The following data objects are provided as well:

InputType
The type object of the objects created by calling
  StringIO with a string parameter.
OutputType
The type object of the objects returned by calling
  StringIO with no parameters.
"""

Maybe you could post to sf, either a tiny patch to the docs clarifying this 
(I think a "(read-only)" or something like that in the brief description of 
InputType might be enough?), or a vastly richer and more complex patch if
you think this isn't a documentation problem but rather than these types
should be merged into a single, more complicated one.


Alex



From niemeyer at conectiva.com  Tue Apr  2 09:17:45 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Tue, 2 Apr 2002 11:17:45 -0300
Subject: PEP 285: Adding a bool type
In-Reply-To: <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>
References: <mailman.1017486169.15848.python-list@python.org> <3CA5E2EA.1CADAF70@engcorp.com> <m3y9ga7z38.fsf@mira.informatik.hu-berlin.de> <mailman.1017673984.1773.python-list@python.org> <%C4q8.117999$Ou.44167@atlpnn01.usenetserver.com>
Message-ID: <20020402111745.A1433@ibook.distro.conectiva>

> Arguably, most of that code should have read
> 
>     if ret:
>         ...
> 
> rather than
> 
>     if ret == None:
>         ...
> 
> since the latter has the flavor of "if expr==True".
> 
> So, is breaking bad code more excusable than breaking good code ;-)?

It should be "if not ret", and most of the time this is a matter of
taste (some times you just have to do this, since you want to check for
"None", and not for 0). Nevertheless, this is a valid construct, and
not uncommon (indeed, I have no problems with this). If you don't
like this, don't use for yourself, but language design is much more than
"I don't like this so everyone using it should burn".

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]



From aahz at pythoncraft.com  Tue Apr 30 13:13:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 30 Apr 2002 13:13:01 -0400
Subject: Python vs. Perl, which is better to learn?
References: <mailman.1020147892.13578.python-list@python.org>
Message-ID: <aamjat$qku$1@panix1.panix.com>

In article <mailman.1020147892.13578.python-list at python.org>,
Andrew McNamara  <andrewm at object-craft.com.au> wrote:
>
>I'd go as far as to say the
>sendmail.cf language is simpler, easier to learn and maintain than perl.

.sigfile?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From martin at v.loewis.de  Mon Apr 15 02:24:53 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 15 Apr 2002 08:24:53 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122>
Message-ID: <m3d6x1mqbu.fsf@mira.informatik.hu-berlin.de>

bokr at oz.net (Bengt Richter) writes:

> If hex long representation were interpreted to represent actual twos
> complement bits, with the top hex digit 8-bit being extended indefinitely
> to the left, you could have hex(-1L) be represented as 0xFL (which would
> require that hex(15L) be represented as 0x0FL).

So you would you represent +7 and -7 in hex?

Regards,
Martin


From greg at cosc.canterbury.ac.nz  Fri Apr 26 00:28:23 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 26 Apr 2002 16:28:23 +1200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift>
Message-ID: <3CC8D767.599720@cosc.canterbury.ac.nz>

Julian Tibble wrote:
> 
> erm... If there is no copyright then there is no *need* for copyleft.

But if RMS would be happy with a world in which there
was simply no copyright, he wouldn't bother with the
GPL -- he could get the same effect just by putting
all the stuff he writes in the public domain.

The fact that he hasn't done that means he
*does* want *some* kind of intellectual property
laws.

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


From tejarex at yahoo.com  Mon Apr 22 15:26:44 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 22 Apr 2002 19:26:44 GMT
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org> <slrnac2v9d.kv6.jon+usenet@snowy.squish.net> <mailman.1019315200.29890.python-list@python.org> <slrnac326d.gba.jon+usenet@snowy.squish.net> <MPG.172df6156ea0ac7c9896bd@news.nwlink.com>
Message-ID: <UtZw8.49218$ro5.4506829@bin6.nnrp.aus1.giganews.com>

"Jeff Shannon" <jeff at ccvcorp.com> wrote in message
news:MPG.172df6156ea0ac7c9896bd at news.nwlink.com...
> Then you probably use a unixoid OS.  ;)  Since Windows uses \ as
> a path separator, any Windows path strings tend to have lots of
> \'s in them, and are thus good candidates for using raw strings.

Since Windows also allows / as a path separator both in the Start/Run
box and for API calls such as the one wrapped by open(), neither
r'string' nor \\ is really necessary for paths anymore than in *nix
(unless you are planning to display them to users for entry on a MSDOS
command line).

Terry J. Reedy





From sholden at holdenweb.com  Tue Apr 23 13:09:03 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 13:09:03 -0400
Subject: Newbie: Running a script on WindowsXP
References: <#dmZtdu6BHA.1188@cpimsnntpa03>
Message-ID: <5Bgx8.197$_i3.82@atlpnn01.usenetserver.com>

"billludden" <billludden at email.msn.com> wrote in message
news:#dmZtdu6BHA.1188 at cpimsnntpa03...
> Hi,
>
> I am trying to figure out how to run a Python21 script
> from the "Run" box on WindowsXP Home edition.
> When I run it, the Command Prompt box just briefly
> flashes on the screen and closes itself.  I have tried
> using the -i option like you would use in Windows 98
> but it doesn't seem to work.  I also tried appending
> Autoexec with:
>     SET PATH=.........;C:\PROGRA~1\PYTHON21
>
> Does anyone understand what I'm doing wrong?
> (incidentally, I had the same problem with "hello world"
>  when I tried it in C)
>
> Thank you!
> Bill
>

Probably you aren't giving the interpreter the correct full path to the
program file. On my Win2K machine,

    python -i C:/Steve/Projects/Python/test9.py

gives the expected behavior, and I see (last few lines of) the output in a
terminated command window. But

    python -i nosuchprogram.py

just flashes up a command window long enough to display (I am presuming)

    python: can't open file 'nosuchprogram.py"

which, sadly, only takes a few tens of milliseconds. Arguably this is a bug
in the response to the -i flag.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From aahz at pythoncraft.com  Tue Apr 16 16:48:18 2002
From: aahz at pythoncraft.com (Aahz)
Date: 16 Apr 2002 16:48:18 -0400
Subject: python threading
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com> <3cbc16f7$0$71662$e4fe514c@dreader4.news.xs4all.nl> <8MVu8.26112$vF6.828988@news2.tin.it> <mailman.1018978678.13870.python-list@python.org>
Message-ID: <a9i2mi$cib$1@panix1.panix.com>

In article <mailman.1018978678.13870.python-list at python.org>,
Justin Sheehy  <justin at iago.org> wrote:
>
>Unless I'm misunderstanding the question, threading.Event is the answer.
>
>I've wrapped Queues with Events before, and it makes things a little
>nicer.  It is quite straightforward to make a Queue that exposes a
>get method that will simply block until an Event is set by another
>thread or an optional timeout has elapsed.
>
>It doesn't solve the whole problem, but waiting on an Event is much
>nicer than a get/sleep loop.

Nicer how?  A timed-out Event *is* a get/sleep loop, and if you're not
using a timed-out Event, the blocking Queue.get() should suffice.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From imbosol at vt.edu  Tue Apr  9 02:18:10 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 9 Apr 2002 02:18:10 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <a8tov6$484$1@panix1.panix.com>
Message-ID: <231u8a.lc.ln@127.0.0.1>

Aahz wrote:
> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
> free to buy another book if K&R turns out to be too heavy for you right
> now, but recognize that you'll need it eventually.

Really?  I've been programming in C for ten years, and I've never even
seen the K&R book.  I think I learned it just fine.  Not that I have a
better suggestion.

(Of course, I rarely feel the need to use C at all these days.)


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From zopestoller at thomas-guettler.de  Fri Apr 12 08:05:45 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 12 Apr 2002 14:05:45 +0200
Subject: Variables common to all objects of a class
Message-ID: <3CB6CD99.70409@thomas-guettler.de>

Hi!

How do you store variables which should be common to all
objects of a class?

I would do it like this. Is this the right pyhton-way?

#foo.py
class_variable=0
def class_method():
     print "I work only on class variables"

class foo:
     object_variable=1
     def my_class_variable(self):
         print "cv", class_variable, "ov", self.object_variable


if __name__=="__main__":
     f=foo()
     f.my_class_variable()
     import foo
     foo.class_method()

thomas



From paul at paulrodger.com  Thu Apr 18 02:32:50 2002
From: paul at paulrodger.com (Paul Rodger)
Date: Thu, 18 Apr 2002 16:32:50 +1000
Subject: unhandled exceptions without a stack-trace
Message-ID: <20020418163250.A31150@flare.taz.wox.org>

Is there any way I can:

1) raise an exception ...
2) not catch that exception (I want python to abort) ...

without displaying the "Traceback (most recent call last):" stack trace?

-- 
paul at paulrodger dot com      http://paulrodger.com/pgpkey.html 
1024D/F0C363EF BC57 D4A4 AAB4 625F 61E8  D933 8F3D 04C5 F0C3 63EF




From hgs at dmu.ac.uk  Fri Apr 12 13:32:23 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Fri, 12 Apr 2002 18:32:23 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
Message-ID: <Pine.GSO.4.40.0204121815510.518-100000@neelix>

Trying to bring python up to date (working version is 1.5.2) on a
Sun Sparc Solaris2.7 system:

GNU Make version 3.79, by Richard Stallman and Roland McGrath.
Built for sparc-sun-solaris2.7

gcc --version  ==  2.95.3

Doing `gmake test` in the Python-2.2.1 directory gives:
        [...]
running build_scripts
not copying /home/hgs/Python-2.2.1/Tools/scripts/pydoc (up-to-date)
find ./Lib -name '*.py[co]' -print | xargs rm -f
./python -E -tt ./Lib/test/regrtest.py -l
Traceback (most recent call last):
  File "./Lib/test/regrtest.py", line 54, in ?
    import random
  File "/home/hgs/Python-2.2.1/Lib/random.py", line 76, in ?
    from math import log as _log, exp as _exp, pi as _pi, e as _e
ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/Modules/mathmodule.so: symbol PyExc_OverflowError: referenced symbol not found
gmake: [test] Error 1 (ignored)
./python -E -tt ./Lib/test/regrtest.py -l
Traceback (most recent call last):
  File "./Lib/test/regrtest.py", line 54, in ?
    import random
  File "/home/hgs/Python-2.2.1/Lib/random.py", line 76, in ?
    from math import log as _log, exp as _exp, pi as _pi, e as _e
ImportError: ld.so.1: ./python: fatal: relocation error: file /home/hgs/Python-2.2.1/Modules/mathmodule.so: symbol PyExc_OverflowError: referenced symbol not found
gmake: *** [test] Error 1

Could someone suggest how to proceed, please?
        Thank you,
        Hugh





From aahz at pythoncraft.com  Sun Apr 21 01:52:05 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 01:52:05 -0400
Subject: Mailman vs References: (was Re: How to make regexes faster? (Python v. OmniMark))
References: <mailman.1019363618.25367.python-list@python.org>
Message-ID: <a9tk25$sut$1@panix1.panix.com>

In article <mailman.1019363618.25367.python-list at python.org>,
Tim Peters  <tim.one at comcast.net> wrote:
>[Erno Kuusela]
>>
>> maybe mailman could be hacked to fake a "references" header based
>> on "in-reply-to".
>
>Barry won't do that. 

You sure?  Last time I talked to him about it a couple of years ago, he
said he'd at least consider it if I gave him a patch.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From pyth at devel.trillke.net  Mon Apr 29 21:02:19 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 03:02:19 +0200
Subject: Remove masked elements from an array
In-Reply-To: <TNECXS1Y4WQCBD0VU52GB4ZT93OMV.3ccde987@Egil>
References: <20020430002740.GT16217@devel.trillke> <TNECXS1Y4WQCBD0VU52GB4ZT93OMV.3ccde987@Egil>
Message-ID: <20020430010219.GV16217@devel.trillke>

This gets funny.

>> The first version uses filter.
>> 'filter' calls 'f' for every item in 'nums' and puts
>> the result into the new list iff it has a 'true' value.
>>
>> The second version uses list comprehension.
>> It basically does the same if you look closely
>> at it.
>
>*Don't* try this if zero is a valid value for your histogram:
>
>newnums = filter(None, nums)

This is a quite stretching forward declaration for 'this'.
"Don't try this" right after citing a sentence is rather
misguiding. Less ambivalent alternatives:

"Yes, but don't try this:"

"But don't try the following (if zero is ...)"

"Be careful with ..."

> [this is the line that you mentally and physically snipped]

No. Interpreted in a different and - dare i say - quite possible way. 
Unless you say that - by applying python's grammar - the ":" makes
it unambigous :-)

amazed-about-the-wonders-of-communication-ly yours Holger




From abagaric at rest-art.hr  Wed Apr 17 04:01:48 2002
From: abagaric at rest-art.hr (Ante)
Date: Wed, 17 Apr 2002 10:01:48 +0200
Subject: preferred syntax for list extend?
References: <3CBCC196.466476CE@uniserve.com>
Message-ID: <a9jaah$493b$1@as201.hinet.hr>

<bvdpoel at uniserve.com> wrote in message
news:3CBCC196.466476CE at uniserve.com...
>
> Just doing a bit of fooling around and I've found that if:
>
> q1=[1,2,3,4]
> q2=[5,6,7,8]
>
> then:
>
> q1 += q2
>
> is about 15% faster than
>
> q1.extend(q2)
>
> Mind you, my test was with a relatively short list, don't know if longer
> lists would act differently.
>
> I would have thought that using extend() would be preferred and faster?

Even if it was faster, I would still prefer +=.
foo.extend(yada) doesn't look beautiful to me at all.
After all, why do we have, e.g. == operator so we can say if a == b and
not a.equals(b)? Because the former is simply better looking, and to me the
same applies to += versus .extend() or even .append()






From aahz at pythoncraft.com  Sun Apr  7 10:46:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 7 Apr 2002 10:46:04 -0400
Subject: Incomplete draft: Namespaces
References: <a8nric$sbb$1@panix1.panix.com> <a8orkv$2cj$0@216.39.172.122>
Message-ID: <a8pm3c$99p$1@panix1.panix.com>

In article <a8orkv$2cj$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>On 6 Apr 2002 17:07:08 -0500, aahz at pythoncraft.com (Aahz) wrote:
>>
>>I've been wrestling with the way namespaces really work, and I've been
>>discussing it a bit on python-dev.  I'm moving the discussion here to
>>get some broader viewpoints.  Here's an incomplete draft of what I've
>>got so far:
>
>Thanks for a summary of notes on names, binding, etc., but so that
>I/we can respond more appropriately, what is your goal?
>I.e., to document things as they are, to explore the fringes
>in terms of what might be, to compare to other languages, or what?

My primary goal is to get things correct for my PyPerl tutorial at
OSCON.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From jcmason at student.math.uwaterloo.ca  Mon Apr 15 21:49:41 2002
From: jcmason at student.math.uwaterloo.ca (Joe Mason)
Date: 16 Apr 2002 01:49:41 GMT
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5 <3cbb7e33@news.mhogaming.com>
Message-ID: <a9fvvl$nr4$1@watserv3.uwaterloo.ca>

In article <3cbb7e33 at news.mhogaming.com>,
Ken Peek <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote:
>Quite frankly, I don't give a damn HOW the long (or int) is represented
>internally to the machine-- I simply ALWAYS want the same to interface to an
>object that behaves as if it is a twos-complement integer.  When I left shift
>it, I don't want the damn SIGN to change-- when I right shift it, I want to be
>able to CONTROL the sign bit that is shifted in, when I display the integer in
>decimal, I want a minus sign prepended to the string if the number is negative,
>and when I print out the number in hexadecimal format, I DON'T want "sign and
>magnitude"-- I simply want a string of hex digits with a '0x' prepended to them,
>representing the twos-complement form of the number, and I want 2*N hex digits
>EVERY TIME, where N is the current size of the integer (in bytes.)

As I see it, the problem here is an interface mismatch.  The fact that
integers are represented as twos-complement numbers is, most of the time,
an implementation detail.  As you point out, you don't care if it actually
*is* a bitfield, you just want the interface of one.  This is an interface 
that should definitely be available, but not necessarily for every integer:
most of the time, it's not important at all. 

So I'd argue that the ability to behave as a twos-complement bitfield is
not something a standard integer should support, and that what you need is
a bitfield object (which a constructor that takes an integer).  This could
be a new language construct or completely encapsulated in a module.  (I'd
lean towards a module, but I think that would mean that every bitwise logical
operator would become a method or function, and I can see a case being made
that keeping the operators available is important enough to be part of the
language.) 

Also, to be completely clean, any operators that currently have no use
except to treat integers as bitfields should be made to work only on
bitfield objects instead.

  x = bitfield(i)
  y = bitfield(j)
  z = x << 8 & y

or

  x = bitfield(i)
  y = bitfield(j)
  z = x.shiftleft(8).bitand(y)

All the arguments about whether doing things with sign extension is
violating the transparency of integers (in an 'it should just work' sense) 
now become invalid (or at least are confined to the constructor and any
methods to convert back to an int).  The bitfield contracts to work just
like a string of bits.

Obviously this is an unrealistic proposal: removing bitwise logical
operators isn't going to happen.  But I'm still interested, in case I
ever design a Python-like language (heh): from your point of view, would
this be acceptable or desirable?  Or would it be too much of a pain in the
ass?  (I don't work with bitfields much, but I can imagine that if you
commonly have to keep swapping back and forth between the two ways of
conceiving of a number, this would be a bad idea.)

Joe


From uwe at rocksport.de  Mon Apr  1 08:54:42 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 1 Apr 2002 13:54:42 GMT
Subject: wxPython  / wxSingleChoiceDialog ...
Message-ID: <a89or2$8dn6m$1@hades.rz.uni-sb.de>

Hi there,

I tried to manipulate wxSinglChoiceDialog so that the presented 
choices are printed with a fixed width font. I tried SetFont()
but nothing changed...

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt                       Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From johnroth at ameritech.net  Tue Apr 30 14:32:32 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 30 Apr 2002 14:32:32 -0400
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it> <mailman.1020156002.23125.python-list@python.org> <6qvga9bizr.fsf@abnoba.intevation.de> <mailman.1020162473.26988.python-list@python.org> <6qn0vlbcpk.fsf@abnoba.intevation.de>
Message-ID: <uctooia295ne2f@news.supernews.com>

"Bernhard Herzog" <bh at intevation.de> wrote in message
news:6qn0vlbcpk.fsf at abnoba.intevation.de...
> Roman Neuhauser <neuhauser at mail.cz> writes:
>
> > > From: Bernhard Herzog <bh at intevation.de>
> > > Why should chr be an int method in the first place? It's not tied
to
> > > ints all that much[1]. IMO it sits somewhere between ints
(including
> > > int-like objects !) and strings so it shouldn't be a method of
either
> > > class. Leaving it a function is perfectly fine, therefore,
although
> > > putting it into the string module might be a good idea. The same
would
> > > apply to ord, IMO.
> >
> >     Ugh, that looks to me as appropriate as join() being a string
> >     method / module function.
>
> Exactly. I still think that join as a function makes more sense than
as
> a string method and for the same reasons as stated for chr above. Even
> if join is implemented as a string method.

To me, .join() makes much more sense as a list or tuple method than
either. Semantically, .join() takes a sequence and transforms it into
a string, therefore it should be a method on the input type (the
sequence),
not on the output type (the string).

I'm against making such things global functions for a number of
reasons, not least of which is the potential combinatorial explosion.
Also, in a "pure" OO language, there would be very few global
functions; Python has way too many for me to consider it to be
good in that regard.

>
> > > [1] In fact, chr today accepts any object that can be converted
> > > successfully to an int, so that chr(42.5) does work, but that's
not
> > > always desirable.
> >
> >     Any object that can be converted to a character should have a
chr()
> >     method.
>
> Would the same reasoning apply to objects that can be converted to
ints,
> for example. I.e. should strings have a method .int() instead of int
> taking a string as argument (among other types)? (forgetting for the
> moment, that it's actually implemented through the __int__ method :) )

We've got a slightly different principle here; int() is a type
constructor,
while chr() is not. chr() creates a single character string. It is
explicitly
the inverse of ord().

> I see numbers and characters as quite different datatypes and a
> conversion between the two is right in the middle and neither more on
> the int side than on the str side.

True, but not very relevant. The basic principle is that a type
conversion method is either a method on the source type, or it is a
constructor
of the result type. In either case, it should have the result type's
name.
The second case is somewhat more difficult in Python, since it
does not have multiple constructors.

> The correspondence between characters depends on an encoding,
something
> which chr and ord do not take into account at all, for example, and
one
> couldn't expect that e.g. int knows about encodings, so that a chr
> method could take the encoding as argument (perhaps defaulting to
> ascii).

Now you're confusing two different things. The purpose of ord() and
chr() is to extract a character into an int so that it can then be
recreated.
These functions are holdovers from earlier languages, and fail miserably
if you want to do character set conversions.

I don't see any reasonable use for an encoding operand in ord().
Consider that strings should know their encoding. This isn't true
currently, but it really should be. In that case, <string>.ord()
produces
the integer index into the encoding, and <string>.encoding() produces
the object that defines the encoding. In the inverse case,
<int>.chr(<encoding>) produces a string with the character from that
encoding.

Character set conversions are best done by string to string functions
or methods. I don't see any benefit in trying to extend ord() and chr()
to handle character set conversions.

Also, your arguement depends on implementation considerations
(one couldn't expect that e.g. int knows about encodings.) Sometimes
you have to take these into consideration, but you should strive
to leave them out if at all possible.

> Of course, whether a conversion operation should be a method, a
function
> or a constructor is largely a matter of taste and is influenced by
> whatever other languages one is coming from. For someone with a
> C-background, characters are ints, so making chr an int method may
seem
> appropriate from that perspective.

It's only a matter of taste until you experiance the heartburn from
what you just ate. Making type conversions global functions simply
leads to a combinatorial explosion of global functions, something
that should be avoided if at all possible.

As far as methods versus constructors goes, conversion methods
chain nicely. Constructors don't. There are situations where one or
the other leads to clearer expression. I suppose that this means one
should do it both ways.

John Roth
>
>    Bernhard
>
> --
> Intevation GmbH                                 http://intevation.de/
> Sketch                                 http://sketch.sourceforge.net/
> MapIt!                                           http://www.mapit.de/




From tim.one at comcast.net  Mon Apr 15 22:56:36 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 15 Apr 2002 22:56:36 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <3CBB8C89.150B3E01@cosc.canterbury.ac.nz>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEJBOOAA.tim.one@comcast.net>

[Tim]
>> Viewing longs as characteristic bit vectors, they model finite
>> and co-finite sets.

[Greg Ewing]
> Well, I meant finite bit strings,
> which is what people twiddle most of the time.

I've got nothing against complements, and the complement of a finite set is
a co-finite set -- there's no problem here.

>> Only finite sets "look positive"; co-finite sets "look negative".  It
>> doesn't take more than the unary ~ operator to go from one kind to the
>> other

> Yes, that's a bit of a nuisance. Whenever you
> do a complement, you really need to specify how
> many bits you're complementing somehow,
> e.g.
>
>   y = ~x & 0x0ffffffff

Why?  If you want to *display* the result as a hex number, that's "the
usual" way to do it (it gets rid of the minus sign in all cases).  But in
the middle of computing bitset operations, it doesn't improve semantics,
speed or clarity.  It adds brittleness, though, by injecting a notion of
maximum size where such a notion isn't needed.

> The neatest solution

Sorry, I don't "a problem".

> would be to have a special fixed-length-bit-string type. Maybe if
> array.array grows a boolean type option, we could give it
> & | ~ << >> operators too.

Different issue, and talked about quite a bit on Python-Dev over the last
few weeks.  Probably the province of Numarray (array.array's code fights
bitstrings every step of the way -- byte is its smallest natural
granularity).





From cliechti at gmx.net  Sat Apr 27 13:56:46 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Apr 2002 19:56:46 +0200
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com>
Message-ID: <Xns91FDCB279C4FFcliechtigmxnet@62.2.16.82>

joost_jacob at hotmail.com (J.Jacob) wrote in 
news:13285ea2.0204270857.2f34f8a4 at posting.google.com:

> I have always liked the unix fortune program.
> Unfortunately nowadays some *n.x's come without fortune,
> and sometimes I have to work with Windows.  I wanted to be
> able to do:
>     C:\mydirectory> python -c "import fortune; fortune.quote()"
> printing a nice quote to stdout.
> The fortune module has to be able to locate a file with
> the name 'allfortunes'.  And you want everything to be
> portable.  And working with python 1.5.2 too.
> 
> The allfortunes file contains quotes separated by lines
> with a '%' as its first character.  I downloaded a couple
> of BSD fortune files and combined them into allfortunes.
> If you put the following fortune.py file with an
> allfortunes file into one of the directories in your
> sys.path, or in you current working directory, it will work.
> 
> This is a first hack, it is slow, the allfortunes file is
> not compressed, so comments are welcome, maybe you can
> suggest improvements?  I was trying to put the allfortunes
> file into a compressed string at the end of the fortune.py
> file, so you do not need a separate file anymore, but I
> did not see a best way to do it.  Please let me know if it
> does not work with your python distribution.
> 
> joost_jacob at hotmail.com
> 
> 
> 
> # file fortune.py
> import random, os
> 
> def randlistindex(yourlist):
>     "Return a random index number into yourlist"
>     return random.randint(0, len(yourlist)-1)

why not use random.choice(seq)?
quote docs:"""Return a random element from the non-empty sequence seq."""
 
> def quote():
>     """
>     Return a quote <string> from file allfortunes,
>     assume the file allfortunes is located in the same
>     directory as this module."""
>     datafile = os.path.join(
>       os.path.dirname(__import__(__name__).__file__),
>       'allfortunes' )
>     quotesfile = open(datafile)
>     quoteslist, quote = [], ''
>     for s in quotesfile.readlines():

readelines loads all into memory... either use xreadlines or the
new "for line in file" feature of py 2.2 when dealing with large files.

when you don't care to load everything into memory something like this 
might work:

>>> quoteslist = quotesfile.read().split('%')

and maybe applying strip to each quote to remove whitespaces from begin and 
end of quote.

>>> quoteslist = map(string.strip, quoteslist)


>         if s[0] == '%':
>             if quote:
>                 quoteslist.append(quote)
>                 quote = ''
>         else: quote = quote + s
>     if quote: quoteslist.append(quote)
>     print quoteslist[randlistindex(quoteslist)],
> 



-- 
Chris <cliechti at gmx.net>



From ponderor at lycos.com  Mon Apr 15 20:24:56 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 15 Apr 2002 19:24:56 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Apr 15)
Message-ID: <33A1FAA138114B2E.E4C1115DD899A291.4B00EFB2D9F60637@lp.airnews.net>

QOTW: "Fortunately, the very act of using Python grants you super
intelligence. Unfortunately it tends to erode unless you update
frequently." -Joshua Macey
        http://mail.python.org/pipermail/python-list/2002-April/097612.html

"Pyrex isn't designed to speed up Python code, but to bridge the
gap between Python and non-Python." - Greg Ewing, RE: ANN: Pyrex - a
language for writing Python extension modules 
    http://mail.python.org/pipermail/python-list/2002-April/097636.html

"There's no such thing as inexpensive patent litigation." -Alan Sutin, head
of IT and e-commerce practice for U.S. office of Greenberg Traurig LLP


    Help with Python advocacy and marketing 
        http://mail.python.org/pipermail/python-list/2002-April/097629.html
    
    Pymacs: an emacs-like editor which is scriptable with python. 
        http://mail.python.org/pipermail/python-list/2002-April/097592.html
    
    vi & vim Tab management tips
        http://mail.python.org/pipermail/python-list/2002-April/097821.html
    
    Google Web API's with Python, Thanks Mark Pilgrim & Simon Willison
        http://diveintomark.org/archives/2002/04/11.html
    
    Easter Eggshells: import-stuff-coffee-this-except-t-shirt-ni-jerk-xyzzy-syzygy-Lulu-started-it-ly ours
        http://mail.python.org/pipermail/python-list/2002-April/097610.html
    
    
========================================================================

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

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

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

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

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

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

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


-- The Python-URL! Team--

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

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


From bokr at oz.net  Mon Apr  1 06:38:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 1 Apr 2002 11:38:52 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <a89gsc$1t0$0@216.39.172.122>

On Sat, 30 Mar 2002 00:39:10 -0500, Guido van Rossum <guido at python.org> wrote:

>I offer the following PEP for review by the community.  If it receives
>a favorable response, it will be implemented in Python 2.3.
>
>A long discussion has already been held in python-dev about this PEP;
>most things you could bring up have already been brought up there.
>The head of the thread there is:
>
>    http://mail.python.org/pipermail/python-dev/2002-March/020750.html
>
>I believe that the review questions listed near the beginning of the
>PEP are the main unresolved issues from that discussion.
>
>This PEP is also on the web, of course, at:
>
>    http://python.org/peps/pep-0285.html
>
>If you prefer to look at code, here's a reasonably complete
>implementation (in C; it may be slightly out of date relative to the
>current CVS):
>
>    http://python.org/sf/528022
>
>--Guido van Rossum (home page: http://www.python.org/~guido/)
>
>PEP: 285
>Title: Adding a bool type
>Version: $Revision: 1.12 $
>Last-Modified: $Date: 2002/03/30 05:37:02 $
>Author: guido at python.org (Guido van Rossum)
>Status: Draft
>Type: Standards Track
>Created: 8-Mar-2002
>Python-Version: 2.3
>Post-History: 8-Mar-2002, 30-Mar-2002
>
>
>Abstract
>
>    This PEP proposes the introduction of a new built-in type, bool,
>    with two constants, False and True.  The bool type would be a
>    straightforward subtype (in C) of the int type, and the values
>    False and True would behave like 0 and 1 in most respects (for
>    example, False==0 and True==1 would be true) except repr() and
>    str().  All built-in operations that conceptually return a Boolean
>    result will be changed to return False or True instead of 0 or 1;
>    for example, comparisons, the "not" operator, and predicates like
>    isinstance().
>
>
>Review
>
>    Dear reviewers:
>
>    I'm particularly interested in hearing your opinion about the
>    following three issues:
>
>    1) Should this PEP be accepted at all.
Ok, in general, but see below.
>
>    2) Should str(True) return "True" or "1": "1" might reduce
>       backwards compatibility problems, but looks strange to me.
>       (repr(True) would always return "True".)
"True"
>
>    3) Should the constants be called 'True' and 'False'
>       (corresponding to None) or 'true' and 'false' (as in C++, Java
>       and C99).
True & False

>
>    Most other details of the proposal are pretty much forced by the
>    backwards compatibility requirement; e.g. True == 1 and
>    True+1 == 2 must hold, else reams of existing code would break.
>
>    Minor additional issues:
>
>    4) Should we strive to eliminate non-Boolean operations on bools
                                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>       in the future, through suitable warnings, so that e.g. True+1
>       would eventually (e.g. in Python 3000 be illegal).  Personally,
>       I think we shouldn't; 28+isleap(y) seems totally reasonable to
>       me.
ISTM that should be phrased differently. I.e., you don't do non-Boolean
operations on bools per se, you convert them first, perhaps implicitly.

IMO the rules for implicit conversions should be very well defined and
well documented in tutorials and docs, because they can be a source of
bugs as well as ability to express concisely.

28+isleap(y) IMO is implicitly 28+int(isleap(y)), and I'm ok with that,
knowing that isleap returns a bool, and that bool is a subtype of int
(and also if it isn't a subtype, for this aritmetic context, see the following).

If bool were a truly separate class, ISTM other issues would come up.
I.e., should there ever be implicit conversion from one class to another
if there is no subtype relationship? Probably not, unless the class
defines a custom operator for it.

You could do that, and it might be an interesting solution.

E.g., let object have a standard __bool__ method that could
be overridden. The default would be to return a bool suitable
for if-testing, as distinct from net value in logical expressions
like "x or y". I.e., x.__bool__() would determine if x was
logically true, but if it was, the value of "x or y" would be x,
not True.

In relational expressions, I expect the proposed subtyping to make

    0 == False => True
    1 == True  => True
    2 == True  => False

because what's really happening UIAM is effectively

    0 == int(False) => True
    1 == int(True)  => True
    2 == int(True)  => False

The results seem ok, but I don't like the semantics unless the
conversions are explicit. I don't think relational operators
should convert their arguments implicitly between types, and
that is one thing wrong with making bool a subtype of int.

The other implicit possibility would be even worse:

    bool(0) == False => True
    bool(1) == True  => True
    bool(2) == True  => True # bad if bool() were implicit, fine if explicit

explicit bool() would presumably call upon 0 .__bool__() and 1 .__bool__.
(note that as with __str__, the space is required).

I'd anticipate using an overridable .__bool__ in many ways, e.g.,

    def __bool__(self): return bool(self.__len__())
or
    def __bool__(self): return self._have_more_data()
etc.

The default obj.__bool__() would always be True where __bool__ was
not overridden, to match current if-tests results on instances, functions
objects, etc.

Where objects are derived from other builtins, __bool__ would be inherited,
according to those base definitions of course.

>
>    5) Should operator.truth(x) return an int or a bool.  Tim Peters
>       believes it should return an int because it's been documented
>       as such.  I think it should return a bool; most other standard
>       predicates (e.g. issubtype()) have also been documented as
>       returning 0 or 1, and it's obvious that we want to change those
>       to return a bool.

I think operator.truth(x) should return a bool, but Tim shouldn't worry
because bool will be implicitly converted to int where it's being used
legitimately (;-) as int (that's not to say str- and repr-related things
may not break).

In a context where either int 0/1 or True/False is legal unchanged,
I think no conversion should be forced, e.g.,

    d={}; d[1==1]='key is True'
but
    ['unused','key is 1'][1==1]

should get an implicit int(True).

I think arithmetic operator contexts should be allowed implicit int conversion,
but I don't think relational expressions qualify as arithmetic contexts for
that purpose, even if one of the compared items is an integer. With subtyping
int it is logical to do it, but I don't like that subtype choice, partly for
this reason.

>
>
>Rationale
>
>    Most languages eventually grow a Boolean type; even C99 (the new
>    and improved C standard, not yet widely adopted) has one.
>
>    Many programmers apparently feel the need for a Boolean type; most
>    Python documentation contains a bit of an apology for the absence
>    of a Boolean type.  I've seen lots of modules that defined
>    constants "False=0" and "True=1" (or similar) at the top and used
>    those.  The problem with this is that everybody does it
>    differently.  For example, should you use "FALSE", "false",
>    "False", "F" or even "f"?  And should false be the value zero or
>    None, or perhaps a truth value of a different type that will print
>    as "true" or "false"?  Adding a standard bool type to the language
>    resolves those issues.
>
>    Some external libraries (like databases and RPC packages) need to
>    be able to distinguish between Boolean and integral values, and
>    while it's usually possible to craft a solution, it would be
>    easier if the language offered a standard Boolean type.
>
>    The standard bool type can also serve as a way to force a value to
>    be interpreted as a Boolean, which can be used to normalize
>    Boolean values.  Writing bool(x) is much clearer than "not not x"
Yes.
>    and much more concise than
>
>        if x:
>            return 1
>        else:
>            return 0
Yes.
>
>    Here are some arguments derived from teaching Python.  When
>    showing people comparison operators etc. in the interactive shell,
>    I think this is a bit ugly:
>
>        >>> a = 13
>        >>> b = 12
>        >>> a > b
>        1
>        >>>
Yes, ugly, especially in that context. But it's the same for all the
bool-returning functions.
>
>    If this was:
>
>        >>> a > b
>        True
>        >>>
Good.
>
>    it would require one millisecond less thinking each time a 0 or 1
>    was printed.
>
>    There's also the issue (which I've seen puzzling even experienced
>    Pythonistas who had been away from the language for a while) that if
>    you see:
>
>        >>> cmp(a, b)
>        1
>        >>> cmp(a, a)
>        0
>        >>> 
>
>    you might be tempted to believe that cmp() also returned a truth
>    value.  If ints are not (normally) used for Booleans results, this
>    would stand out much more clearly as something completely
>    different.
This is good.
>
>
>Specification
>
>    The following Python code specifies most of the properties of the
>    new type:
>
>        class bool(int):
what if instead:
         class bool(object):
             def __int__(self):
                 if self:
                     return 1
                 else:
                     return 0
#
# ... and all the integer ops with a bool as subject that you
#     want to make legal/control...
#

             def __bool__(self):  # support if x ...: x and ... uses of x.__bool__()
                 return self      # probably optimized away in the implementation, but ...
>
>            def __new__(cls, val=0):
>                # This constructor always returns an existing instance
>                if val:
Noting that "if val" would probably compile to the usual
                 LOAD_FAST                0 (val)
                 JUMP_IF_FALSE
meaning the VM implementation of JUMP_IF_TRUE/FALSE will have to check
what kind of critter val is, and if it has a __bool__ method, to call that.
(I think, not having been in there yet.)

>                    return True
>                else:
>                    return False
>
>            def __repr__(self):
>                if self:
>                    return "True"
>                else:
>                    return "False"
>
>            __str__ = __repr__
>
>            def __and__(self, other):
>                if isinstance(other, bool):
>                    return bool(int(self) & int(other))
>                else:
>                    return int.__and__(self, other)
                     return other.__and__(other,self)
I.e., how about giving other a chance? If other calls on us at __int__, that would be ok,
but if it doesn't, we should not volunteer ourselves as int.
>
>            __rand__ = __and__
maybe not necessarily so for other?
>
>            def __or__(self, other):
>                if isinstance(other, bool):
>                    return bool(int(self) | int(other))
>                else:
>                    return int.__or__(self, other)
                     return other.__or__(other,self)
>
>            __ror__ = __or__
maybe not necessarily so for other?
>
>            def __xor__(self, other):
>                if isinstance(other, bool):
>                    return bool(int(self) ^ int(other))
>                else:
>                    return int.__xor__(self, other)
                     return other.__Xor__(other,self)
>
>            __rxor__ = __xor__
maybe not necessarily so for other?
>
>        # Bootstrap truth values through sheer willpower
>        False = int.__new__(bool, 0)
>        True = int.__new__(bool, 1)
         False = object.__new__(bool, 0)
         True = object.__new__(bool, 1)
I'm just copying, Don't know if this will work ;-)
>
>    The values False and True will be singletons, like None; the C
>    implementation will not allow other instances of bool to be
>    created.  At the C level, the existing globals Py_False and
>    Py_True will be appropriated to refer to False and True.
>
>    All built-in operations that are defined to return a Boolean
>    result will be changed to return False or True instead of 0 or 1.
>    In particular, this affects comparisons (<, <=, ==, !=, >, >=, is,
>    is not, in, not in), the unary operator 'not', the built-in
>    functions callable(), hasattr(), isinstance() and issubclass(),
>    the dict method has_key(), the string and unicode methods
>    endswith(), isalnum(), isalpha(), isdigit(), islower(), isspace(),
>    istitle(), isupper(), and startswith(), the unicode methods
>    isdecimal() and isnumeric(), and the 'closed' attribute of file
>    objects.
>
>    Note that subclassing from int means that True+1 is valid and
>    equals 2, and so on.  This is important for backwards
>    compatibility: because comparisons and so on currently return
>    integer values, there's no way of telling what uses existing
>    applications make of these values.
True, but they will either be legitimate or not. If implicit conversion from bool
were supplied for integer arithmetic contexts but not relational operators,
I wouldn't think much would break, though I admit not thinking very far on that.
What use of bools as integers do not use integer-requiring operations that would
implicitly call the bool.__int__ method?
>
>
>Compatibility
>
>    Because of backwards compatibility, the bool type lacks many
>    properties that some would like to see.  For example, arithmetic
>    operations with one or two bool arguments is allowed, treating
>    False as 0 and True as 1.  Also, a bool may be used as a sequence
>    index.
>
>    I don't see this as a problem, and I don't want evolve the
>    language in this direction either; I don't believe that a stricter
>    interpretation of "Booleanness" makes the language any clearer.

I think in the most general sense bool(x) asks whether x belongs
to the subset of the possible x instances defined by its class as having
True bool values. No one can second-guess the class on that, or should, IMO.
I think an overridable __bool__ method for the object class makes
sense to test this set membership. (Of course, in practice it's usually
membership in the set having False bool values that's tested,
and the result inverted). To me, the issue of using bools 'as ints' is
a matter of defining implicit context-sensitive promotion rules between
types, and that's separate from bool's value as a class that now can
reflect its semantics better, e.g., through str and repr, appropriately
separate from int.

>
>    Another consequence of the compatibility requirement is that the
>    expression "True and 6" has the value 6, and similarly the
>    expression "False or None" has the value None.  The "and" and "or"
>    operators are usefully defined to return the first argument that
>    determines the outcome, and this won't change; in particular, they
>    don't force the outcome to be a bool.  Of course, if both
>    arguments are bools, the outcome is always a bool.  It can also
>    easily be coerced into being a bool by writing for example
>    "bool(x and y)".
This all works fine with my object.__bool__ suggestion, I believe. See
more above.

>
>
>Issues
>
>    Because the repr() or str() of a bool value is different from an
>    int value, some code (for example doctest-based unit tests, and
>    possibly database code that relies on things like "%s" % truth)
>    may fail.  How much of a backwards compatibility problem this will
>    be, I don't know.  If we this turns out to be a real problem, we
>    could changes the rules so that str() of a bool returns "0" or
>    "1", while repr() of a bool still returns "False" or "True".
>
>    Other languages (C99, C++, Java) name the constants "false" and
>    "true", in all lowercase.  In Python, I prefer to stick with the
>    example set by the existing built-in constants, which all use
>    CapitalizedWords: None, Ellipsis, NotImplemented (as well as all
>    built-in exceptions).  Python's built-in module uses all lowercase
>    for functions and types only.  But I'm willing to consider the
>    lowercase alternatives if enough people think it looks better.
>
>    It has been suggested that, in order to satisfy user expectations,
>    for every x that is considered true in a Boolean context, the
>    expression x == True should be true, and likewise if x is
>    considered false, x == False should be true.  This is of course
That's equivalent to demanding that bool(x)==x should be True, which is
only true for bools. Someone was having braindead moment. (I'm not immune ;-)

>    impossible; it would mean that e.g. 6 == True and 7 == True, from
>    which one could infer 6 == 7.  Similarly, [] == False == None
>    would be true, and one could infer [] == None, which is not the
>    case.  I'm not sure where this suggestion came from; it was made
>    several times during the first review period.  For truth testing
>    of a value, one should use "if", e.g. "if x: print 'Yes'", not
>    comparison to a truth value; "if x == True: print 'Yes'" is not
>    only wrong, it is also strangely redundant.
I think it's a matter of saying that the relational operators won't
implicitly convert their arguments to bool, so x==y doesn't braindeadly
implement bool(x)==bool(y).

I think here is where subtyping from int may be a problem, because the
relational operators (I assume) will promote the subtype implicitly to
int, and give indications that are really not semantically meaningful,
like 2 > True => True. IMO, that should really raise an exception about
comparing incompatible types.

Actually, the relational operators seem to be pretty promiscuous 'way
beyond mere promotion of numeric related types ;-/

But it says in the docs that it may change re comparison of different
types. Maybe, like __bool__, objects should have an opportunity
to produce something for sorting and other ordering uses, e.g., __rank__ ?
(BTW is the 'consistent ordering' mentioned for comparisons really
a default def __rank__(self): return id(self) ? )
>
>
>Implementation
>
>    An experimental, but fairly complete implementation in C has been
>    uploaded to the SourceForge patch manager:
>
>    http://python.org/sf/528022
>
>
>Copyright
>
>    This document has been placed in the public domain.
>
>
>
>Local Variables:
>mode: indented-text
>indent-tabs-mode: nil
>fill-column: 70
>End:
>
Maybe this has been discussed before, but I would have expected some mention in the PEP
if so. HTH.

Regards,
Bengt Richter


From siddhartha1987 at hotmail.com  Tue Apr 30 16:59:30 2002
From: siddhartha1987 at hotmail.com (Hermann Hesse)
Date: Tue, 30 Apr 2002 13:59:30 -0700
Subject: inner classes
Message-ID: <F173Fm5PtAalwotWTXs00005c64@hotmail.com>

Michael Chermside wrote:

>Most of the time, this is the behavior that you want -- >inner classes in 
>Java were invented to solve problems that are really >easy to handle in 
>Python because of the existance of bound methods. (I >can explain what that 
>meant if you like.) But as you might expect, if >you actually DO want the 
>java-like behavior, Python allows it, you just >have to be explicit and 
>explain what you mean.

Can someone give an example of how you'd use bound methods to replace inner 
classes in python?



_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp.





From mlh at vier.idi.ntnu.no  Thu Apr 11 23:32:57 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 03:32:57 +0000 (UTC)
Subject: [OT] Re: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org>  <3CB2CB16.8080103@geneva-link.ch> <slrnab5jqc.9j.mlh@vier.idi.ntnu.no>  <3CB2DB99.4080400@geneva-link.ch> <slrnab9soc.6g8.mlh@vier.idi.ntnu.no>  <3CB5546D.8010802@geneva-link.ch>
Message-ID: <slrnabclb9.c8l.mlh@vier.idi.ntnu.no>

In article <3CB5546D.8010802 at geneva-link.ch>, Boris Borcic wrote:
>Magnus Lie Hetland wrote:
>
>>>
>>>   >>> filter(lambda W : W not in "ILLITERATE","BULLSHIT")
>>>   'BUSH'
>>>
>>>What's your problem ?
>> 
>> Nothing. Sorry -- thouhgt there was a syntax problem there. A bit
>> quick on the trigger ;)
>
>Revealing :)
>

Rub it in, why don't you ;)

(And I hope you're not talking politics here? I don't think I've
revealed any political attitudes in this... And no, the choice of the
word "trigger" does _not_ mean that I'm pro-handguns...)

Actually, in a brief moment I saw the '"ILLITERATE","BULLSHIT"' part
as a tuple -- some precedence confusion with the lambda, I guess ;)

>> 
>> (So, what you're basically saying is that Bush is the literate part of
>> bullshit? <wink>)
>
>
>More concretely, providing support to the idea that the "USA" in 
>"CRUSADE" explains the general pattern of some, er, crude presidential 
>choice of words.

Heh :)

>Blind-spots-are-consubstantial-to-visions-ly yours,
>
>Boris Borcic
>--
>"Hope achieves the square root of the impossible"

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From jcea at argo.es  Thu Apr 18 15:38:34 2002
From: jcea at argo.es (Jesus Cea Avion)
Date: Thu, 18 Apr 2002 21:38:34 +0200
Subject: Compile error: Python 2.2.1 in Solaris 2.5.1, with GCC 2.95.3
Message-ID: <3CBF20BA.6FBEDF89@argo.es>

I've detected two problems:

a) Problems importing "pthreads.h".

b) Problems detecting a correct value for "LINKFORSHARED".

c) Problems with libraries in "/usr/local/lib".

Here you have the patch.

CAUTION: This patch is not intented to be included *as is* in future
Python 2.2.2. That bugfix will be work for Python maintainers :-).
Its only mission is to allow correct Python 2.2.1 compilation under
my environment.

Before compile Python 2.2.1, you must do:

LD_LIBRARY_PATH=/usr/local/lib
export LD_LIBRARY_PATH

>>>>>

cvs server: Diffing .
Index: Makefile.pre.in
===================================================================
RCS file: /opt/src/cvsroot/python/Makefile.pre.in,v
retrieving revision 1.1.1.4
retrieving revision 1.1.1.4.2.1
diff -c -u -r1.1.1.4 -r1.1.1.4.2.1
cvs server: conflicting specifications of output style
--- Makefile.pre.in     2002/04/18 10:28:35     1.1.1.4
+++ Makefile.pre.in     2002/04/18 17:29:20     1.1.1.4.2.1
@@ -94,8 +94,8 @@
 
 # Symbols used for using shared libraries
 SO=            @SO@
-LDSHARED=      @LDSHARED@
-BLDSHARED=     @BLDSHARED@
+LDSHARED=      @LDSHARED@ 
+BLDSHARED=     @BLDSHARED@ -Wl,--rpath -Wl,/usr/local/lib
 DESTSHARED=    $(BINLIBDEST)/lib-dynload
 
 # Executable suffix (.exe on Windows and Mac OS X)
Index: configure
===================================================================
RCS file: /opt/src/cvsroot/python/configure,v
retrieving revision 1.1.1.7
retrieving revision 1.1.1.7.2.1
diff -c -u -r1.1.1.7 -r1.1.1.7.2.1
cvs server: conflicting specifications of output style
--- configure   2002/04/18 10:28:36     1.1.1.7
+++ configure   2002/04/18 12:16:59     1.1.1.7.2.1
@@ -3275,13 +3275,8 @@
                then
                        LINKFORSHARED="-Wl,--export-dynamic"
                fi;;
-       SunOS/5*) case $CC in
-                 *gcc*)
-                   if $CC -Xlinker -V 2>&1 | grep BFD >/dev/null
-                   then
+       SunOS/5*) 
                        LINKFORSHARED="-Xlinker --export-dynamic"
-                   fi;;
-                 esac;;
        esac
 fi
 echo "$ac_t""$LINKFORSHARED" 1>&6
Index: setup.py
===================================================================
RCS file: /opt/src/cvsroot/python/setup.py,v
retrieving revision 1.1.1.5
retrieving revision 1.1.1.5.2.1
diff -c -u -r1.1.1.5 -r1.1.1.5.2.1
cvs server: conflicting specifications of output style
--- setup.py    2002/04/18 10:28:37     1.1.1.5
+++ setup.py    2002/04/18 17:29:20     1.1.1.5.2.1
@@ -1,7 +1,7 @@
 # Autodetecting setup.py script for building the Python extensions
 #
 
-__version__ = "$Revision: 1.1.1.5 $"
+__version__ = "$Revision: 1.1.1.5.2.1 $"
 
 import sys, os, getopt
 from distutils import sysconfig
@@ -211,8 +211,8 @@
         # lib_dirs and inc_dirs are used to search for files;
         # if a file is found in one of those directories, it can
         # be assumed that no additional -I,-L directives are needed.
-        lib_dirs = self.compiler.library_dirs + ['/lib', '/usr/lib']
-        inc_dirs = self.compiler.include_dirs + ['/usr/include']
+        lib_dirs = self.compiler.library_dirs + ['/usr/local/lib/', '/lib', '/usr/lib']
+        inc_dirs = self.compiler.include_dirs + ['/usr/local/include', '/usr/include']
         exts = []
 
         platform = self.get_platform()

<<<<<

-- 
Jesus Cea Avion                         _/_/      _/_/_/        _/_/_/
jcea at argo.es http://www.argo.es/~jcea/ _/_/    _/_/  _/_/    _/_/  _/_/
                                      _/_/    _/_/          _/_/_/_/_/
PGP Key Available at KeyServ   _/_/  _/_/    _/_/          _/_/  _/_/
"Things are not so easy"      _/_/  _/_/    _/_/  _/_/    _/_/  _/_/
"My name is Dump, Core Dump"   _/_/_/        _/_/_/      _/_/  _/_/
"El amor es poner tu felicidad en la felicidad de otro" - Leibniz


From logiplexsoftware at earthlink.net  Mon Apr 29 15:48:07 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 29 Apr 2002 12:48:07 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <007401c1efb0$91d9d7e0$445d4540@Dell2>
References: <3CCD9649.98AED86F@cascade-sys.com>
	<007401c1efb0$91d9d7e0$445d4540@Dell2>
Message-ID: <20020429124807.2efdf0eb.logiplexsoftware@earthlink.net>

On Mon, 29 Apr 2002 12:03:35 -0700
Brian Quinlan wrote:

> > > Why the hostility?
> > 
> > I don't interpret this strongly worded and slightly sarcastic critique
> 
> > to be hostile.
> > 
> > If you call this hostile I don't think you know the meaning of the
> word.

[snip]
 
> Then Lumberjack should attack the PEP, not the author e.g. 
> 
>    "How about you and other PEP writers take some responsibilities 
>    for your fantasies for once? You up to helping thousands rewrite 
>    their code? I thought no"
> 
> Seems like an attack on PEP writers in general, not on the PEP itself.

I have to agree with Brian.  While Lumberjack's response did address issues
with the PEP, it was laced with personal insinuations about the PEP's author
(and PEP writers in general):

[Lumberjack]
"How about you and other PEP writers take some responsibilities for your
fantasies for once?"
"But no, you'd rather break code and make existing instructional material
obsolete"
"The [sic] would be in line with the contempt you seem to have for others.
"You think the world revolves around your notion of "rightness"?"
"I believe the organization you belong to has this chant: ..."

The overall tone was anger not reason.  Arguments stated in this fashion
deserve the same consideration they give to their opponent: none.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From db3l at fitlinxx.com  Tue Apr 30 00:20:36 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 30 Apr 2002 00:20:36 -0400
Subject: String handling bug in Python
References: <mailman.1020127367.12415.python-list@python.org>
Message-ID: <uznzlrf57.fsf@ctwd0143.fitlinxx.com>

"Delaney, Timothy" <tdelaney at avaya.com> writes:

> import os.path
> 
> path = '//computer/path/to/file'
> path = os.path.normpath(path)
> print path

Um, yeah, but why bother?  I suppose if I were going to present it to
a user but wanted the source to use forward slashes I might do this,
but it's already a Windows specific reference, so I'd probably just go
with backslashes in the code or just leave the forward slashes
unconverted.

Tim's point (I believe) was that in like 99.9% of the cases, you can
just pass the forward slash version of a filename (without any need to
go through the overhead of normalizing it) directly to the Win32 API.
E.g., stick with forward slashes for everything - no difference in
overhead and you avoid the potential to tickle the raw string
limitation.  I do it all the time myself, even in Windows specific
code, since that's the path notation I'm more familiar with.

But if I were in a case where I needed the backslashes, and doing a
hardcoded string like the above, I don't see the benefit to writing it
with a forward slash in the source only to incur runtime overhead to
flip the slashes?  Even if I needed to end in a single backslash ala
this thread :-)

As it turns out, your original "path" would still work fine in most
(maybe all) Win32 filesystem calls.  The problem UNCs tend to be for
(if I recall correctly) stuff like named pipes, and/or things like
service control manager calls or other APIs where one of the
parameters is just a remote machine name (e.g., not a file path) that
needs the \\ prefix.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From skip at pobox.com  Tue Apr 23 13:12:49 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 23 Apr 2002 12:12:49 -0500
Subject: file_ref = open(string_ref)
In-Reply-To: <1019579845.565223@cswreg.cos.agilent.com>
References: <1019579845.565223@cswreg.cos.agilent.com>
Message-ID: <15557.38417.154062.196957@12-248-41-177.client.attbi.com>

    Greg> Is this code
    Greg> C1:   file_ref = open(string_ref)

    Greg> better than this?:

    Greg> C2:   file = open(string)

    Greg> C1 is more accurate, given the view that data in Python are
    Greg> references to objects.  According to this view, the left-hand
    Greg> variable names a reference to a file object.

Two things come to mind:

    * everything's a reference, so "_ref" is redundant

    * "file" is a builtin in 2.2 ("file is open" returns True)

Personally I prefer

    f = open(string)

;-)

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From zen at shangri-la.dropbear.id.au  Wed Apr 24 04:06:37 2002
From: zen at shangri-la.dropbear.id.au (Stuart Bishop)
Date: Wed, 24 Apr 2002 18:06:37 +1000
Subject: Mass Mailing help needed (Paul Rubin)
In-Reply-To: <jrgx8.119$_i3.97@atlpnn01.usenetserver.com>
Message-ID: <3397E6FA-575A-11D6-B0C2-000393031882@shangri-la.dropbear.id.au>

On Wednesday, April 24, 2002, at 02:58  AM, Steve Holden wrote:

> "Graham Ashton" <gashton at cmedltd.com> wrote ...
>> On Tue, 2002-04-23 at 15:20, A.M. Kuchling wrote:
>>
>>> It's certainly possible to send multiple messages over a single SMTP
>>> connection,
>>
>> I must confess to not being that well up on the Python library with
>> regard to email messages. Are there any python modules that support
>> this?
>>
>
> The smtplib library is explicitly provided to allow your applications to
> make direct connections to SMTP servers.

And furthermore, using the Queue and Thread modules is pretty simple,
letting you open multiple connections to multiple upstream SMTP servers.
I tend to close and reopen them after I've hit an arbitrary threshold
of sent messages, mainly due to paranoia.

Calling sendmail directly can be good too, as you can specify synchronous
delivery on the command line (-O DeliveryMode=i I think...). This way
you can open 10 threads, each invoking sendmail in synchronous delivery 
mode,
and know you have 10 simultanous outgoing connections but keep an empty
queue. This isn't possible with some MTA's such as Postfix.  I
personally just hit my ISP's massive upstream SMTP servers as my tiny 
20,000
user mailouts are rather insignificant when compared with all the other 
users
on the cable network...

--
Stuart Bishop <zen at shangri-la.dropbear.id.au>
http://shangri-la.dropbear.id.au/





From borcis at geneva-link.ch  Mon Apr 29 16:54:03 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Mon, 29 Apr 2002 22:54:03 +0200
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org>
Message-ID: <3CCDB2EB.9E238601@geneva-link.ch>

James J. Besemer wrote:
> 
> > >From the comments, I've learned:
> >
> > 5.  oct() is another candidate based on the rarity of usage
> 
> Candidate perhaps but some people still like Octal.

Octal has some virtues, e.g. 365 == 0555 making it easy
to remember the number of days in a usual (non-leap) year :)

365 == 0555 == 012**2 + 013**2 + 014**2 == 015**2 + 016**2 -ly yours,

Boris Borcic
--
"On na?t tous les m?tres du m?me monde"


From philh at comuno.freeserve.co.uk  Sat Apr  6 12:40:27 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Sat, 6 Apr 2002 18:40:27 +0100
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96>
Message-ID: <slrnaaucob.3j0.philh@comuno.freeserve.co.uk>

On 5 Apr 2002 16:48:57 GMT, James Logajan <JamesL at Lugoj.com> wrote:
>philh at comuno.freeserve.co.uk (phil hunt) wrote:
>> Another problem is that when a language is constantly changing, it 
>> becomes necessary to constantly relearn it. Python has changed a lot 
>> in the last few years, for example new operators such as +=, and 
>> allowing users to subclass from builtin types, are big (and useful) 
>> changes. Python 1.5.1 was an immature language that could benefit 
>> from changing.
>
>Version 1.5.2 was sufficiently complete and long-lived enough that I 
>established it as the version I would develop against.
>
>> Is there a final destination in sight, which these changes are 
>> leading to?
>
>The answer to this isn't as interesting to me as _when_ will this tinkering 
>end? Something that isn't up for debate: because of the problems I have 
>actually encountered with making code work with the many extant versions of 
>Python, I will no longer be doing any new projects in the language and am 
>not recommending Python to anyone.

This is the situation I wish to avoid.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From aahz at pythoncraft.com  Mon Apr  1 20:25:13 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 20:25:13 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <a89qqt$b2a$1@panix1.panix.com> <Pine.SOL.4.44.0204011633190.12842-100000@death.OCF.Berkeley.EDU>
Message-ID: <a8b19p$cbh$1@panix1.panix.com>

In article <Pine.SOL.4.44.0204011633190.12842-100000 at death.OCF.Berkeley.EDU>,
Brett Cannon  <bac at OCF.Berkeley.EDU> wrote:
>
>And as for postponing it, I think that is a VERY good idea.  It would be
>nice to have a release where I don't need to think about using any
>__future__ imports.

This change will not use a __future__ directive.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From sholden at holdenweb.com  Tue Apr 16 17:22:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 16 Apr 2002 17:22:00 -0400
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org>
Message-ID: <UD0v8.1765$hl2.41@atlpnn01.usenetserver.com>

<sean.bowman at acm.org> wrote ...
> hello,
>
> I'm wondering if there's a way to optimize code generated using the
> compiler module.  for example, if I say::
>
>  import dis
>  from parser import compileast, expr
>
>  code = compileast(expr('x+3+7.'))
>  print dis.dis(code)
>
> I notice that the constants aren't folded.  is there a way to do this, or
> do I have to fiddle with the ast myself?
>
The interpreter does not perform constant folding, and it is not currently
intended that it should. If someone with sufficient knowledge, skill and
time wants to submit a patch, I am sure that will be regarded with nothing
but delight! :-)

regards
 Steve






From boud at valdyas.org  Tue Apr 16 15:58:52 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 16 Apr 2002 21:58:52 +0200
Subject: python threading
References: <BFFEB31BC23CCD44B8924E139E1D87E448EF1F@MAILBOX.nameconnector.com> <mailman.1018983512.22651.python-list@python.org>
Message-ID: <3cbc82dc$0$85459$e4fe514c@dreader3.news.xs4all.nl>

Alex Martelli wrote:

> On Tuesday 16 April 2002 20:39, Geoffrey Talvola wrote:
> ...
>> Don't most GUI toolkits have some way to post an event to the main GUI
>> thread's event loop from another thread?  wxPython has wxPostEvent()
>> which does this.
> 
> I don't know of any Tkinter or PyQt equivalent.  I hope somebody can
> enlighten me about those, as they're the toolkits I need to use.
> 

Phil has released the first release candidate for 3.2 today, which should
fix QThread support. Qt offers support for posting events: unsuprisingly 
called QThread.postEvent(). I'm going to test this as soon as I get PyQt 
built again, so that'll probably be Wednesday evening.

Then again, QThread is a different beast from threading.Thread, so you
probably can't just use QThread.postEvent() from a Python thread :-).

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


From guido at python.org  Thu Apr  4 09:35:26 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:35:26 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com> <016701c1db1a$569543e0$0900a8c0@spiff> <mailman.1017849614.417.python-list@python.org> <a8gq78$hg$0@216.39.172.122>
Message-ID: <3CAC658E.51B3A912@python.org>

Bengt Richter wrote:
> What if I wanted to subtype str and have a bool signifying URL-escaped or not,
> yet still have normally functioning __len__() etc.?

That's an absurd idea, right there with redefining + to mean *.

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


From Lutz.Schroeer at kybernetik-manufaktur.de  Fri Apr 26 15:59:54 2002
From: Lutz.Schroeer at kybernetik-manufaktur.de (Lutz Schroeer)
Date: 26 Apr 2002 19:59:54 GMT
Subject: Simple bar graphs
References: <Xns91FCD43535458Latzikatz@139.174.2.56> <aacas8$9d997$2@ID-99293.news.dfncis.de>
Message-ID: <Xns91FCE04392F6ELatzikatz@139.174.2.56>

William Park <opengeometry at NOSPAM.yahoo.ca> wrote in
news:aacas8$9d997$2 at ID-99293.news.dfncis.de: 

> I used 'gdchart' long ago.  It generates GIF directly.  But, if your
> requirement is simple enough, you can use HTML rules, say in table
> layout. 

Looked at that, too, good stuff, but unfortunately it's just a Python 
frontend to the GDChart package (IMHO), which is a C lib and needs 
compiling, which isn't permitted on my webspace for security reasons (says 
my provider).

Lutz


From phr-n2002a at nightsong.com  Fri Apr 12 23:30:22 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 20:30:22 -0700
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org> <7xy9fse6ol.fsf@ruckus.brouhaha.com> <slrnabf5r2.s6b.grey@teleute.dmiyu.org>
Message-ID: <7xr8lk46nl.fsf@ruckus.brouhaha.com>

Steve Lamb <grey at despair.dmiyu.org> writes:
> > The SL5500 is a high powered computer with a 32-bit CPU and 64 MB of
> > ram.  It runs Python just fine.  
> 
>     *Eyes Sharp's list of retailers for the SL-5500 a little closer.*
> Re-e-eaaaalllly.  Hmmmm..

The SL-5500 hardware is similar to an Ipaq 36/37xx plus it has a cute
little slide-out keyboard that's not all that useful.  I got the
developer version for $199.00 which was a good deal at the price.  I
wouldn't recommend it at $500 though.  I'll post a review to my website soon.


From sameer_ at email.com  Tue Apr 16 01:32:46 2002
From: sameer_ at email.com (sameer)
Date: 15 Apr 2002 22:32:46 -0700
Subject: python threading
Message-ID: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>

I am trying to run multiple threads in a wxFrame class to instantiate
different panels within it on Win 2k,  and well, it's a little hairy. 
Anyone got a good web source that discusses python threading in depth?

More so, my question is, how can I get the return value of a function
that I run on a seperate thread?  I don't see anything for this in the
threading module.

Thanks in advance for your help!


From aahz at pythoncraft.com  Sun Apr 21 13:35:09 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Apr 2002 13:35:09 -0400
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com>
Message-ID: <a9ut8d$dql$1@panix3.panix.com>

In article <304d20df.0204210629.3f2b4d58 at posting.google.com>,
Allan Crooks <googlegroups at sixtyten.org> wrote:
>
>Another thought I've had is doing it by specifying, say for module
>'x', having modules 'x_20', 'x_21', 'x_22', with x simply finding out
>what versions of Python it is running on, and then compiling all
>modules it can deal with, and then importing all objects into it's
>namespace.

Yup.  But do it with a package.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From flognat at flognat.myip.org  Tue Apr 16 05:04:21 2002
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Tue, 16 Apr 2002 09:04:21 GMT
Subject: Win32com usage of IE raises exception.. (how do I interpret win32com errs?)
References: <m33cxwxdae.fsf@flognat.myip.org>
Message-ID: <m3y9fovwtn.fsf@flognat.myip.org>

Duuh some googling of the newsgroups recommended reboot of the
computer, did that before, did it again.. and it worked *sigh*.. 

http://groups.google.com/groups?hl=en&selm=3C9E4F90.6050708%40skippinet.com.au

But the question still remains.. Anyone out there with a good
browser-class, programming/usage?? (is starting to experiment)

        /Andy

/ Andrew Markebo <flognat at flognat.myip.org> wrote:
| I am trying to reuse some win32com demo that was posted ages ago in
| this group, firing up ActiveState Python (i tried the python.org
| python+win32com, same 'problem), win32makepy, and the code:

[.. cut the rest of the article, google for it if you want to see it ..]

-- 
 The eye of the beholder rests on the beauty!


From klausman-un030402 at tuts.net  Wed Apr  3 03:08:18 2002
From: klausman-un030402 at tuts.net (Tobias Klausmann)
Date: 3 Apr 2002 08:08:18 GMT
Subject: duelling pythons
References: <1eed326b.0203312131.2ef13b62@posting.google.com> <mailman.1017695170.25885.python-list@python.org> <m3d6xibyfn.fsf@nbecker.fred.net> <mailman.1017770346.22212.python-list@python.org>
Message-ID: <a8ed9i$rnodf$1@ID-41653.news.dfncis.de>

Cliff Wells <logiplexsoftware at earthlink.net> wrote:
> On 01 Apr 2002 21:56:44 -0500
> nbecker at fred.net wrote:
> 
>> Keep lobbying Redhat.  The correct solution is for them to put
>> 
>> #! /usr/bin/python1.5
>> 
>> If they need a specific version.  I argued this with them already.
>> Keep up the pressure.  Maybe they will get it.
> 
> Absolutely agree.  Or better #!/usr/bin/python.rh with a link python.rh ->
> python1.5 so they can eventually upgrade without changing every shebang.

In a related matter: how common is /usr/bin/env ?
I really like that tool to run python wherever it is: 

#!/usr/bin/env python

As it is part of sh-utils, I guess it's pretty widely used.

Anybody know of specific environments where it is not commonly
available?

so long,
Tobias
 


From lac at strakt.com  Mon Apr 29 22:18:30 2002
From: lac at strakt.com (Laura Creighton)
Date: Tue, 30 Apr 2002 04:18:30 +0200
Subject: Comment on draft PEP for deprecating six builtins 
In-Reply-To: Message from "James J. Besemer" <jb@cascade-sys.com> 
   of "Mon, 29 Apr 2002 17:24:35 PDT."
Message-ID: <200204300218.g3U2IUkG030640@ratthing-b246.strakt.com>

> 
> 
> > [James J. Besemer]
> >
> > > I don't interpret this strongly worded and slightly sarcastic crit
> iqu=
> e to
> > > be hostile.
> 
> Fran=E7ois Pinard wrote:
> 
> > For me, sarcasm and strong wording are not welcome.  There are ways 
> to
> > disagree, and even disagree much, while staying friendly.
> 
> Americans tend to value straight talk.  I put it all under the categor
> y o=
> f
> effective communication.  Being French Canadian, this may be alien to 
> you.

No, there is nobody left on the planet, let alone in Canada who hasn't
already heard more than their fill of Americans touting rudeness as
effective communication.  
 
> Although I agree with you that sarcasm and rhetorical questions are no
> substitute for a logical argument, they can put an effective emphasis 
> on =
> an
> otherwise factual and logical post such as Logjam's (or whatever his n
> ame=
>  is).
> 
> Nor is Sarcasm or frank talk necessarily unfriendly.  My close friends
>  te=
> ase
> each other mercilessly.  After an exchange that grew a bit tense at po
> int=
> s,
> Steve Holden and I emerged better friends and better acquainted with e
> ach=
>  other
> than I think we would have otherwise.  Many people rely on close frien
> ds =
> to
> utter the bare truths that mere acquaintences would avoid.

Yes, this is a well-documented problem Americans have in participating
in international forums.  You need to learn that we are not your friends.
In nearly all countries, friendship takes a very long time to create.
It is not something which is freely there, all the time, and always to
be given to new acquaintances, and it is something which we do not desire
to hand out and have freely available in such a fashion.

When you treat us this familiarly, we take offense.  This is because only
small children and the mentally deficient behave this way around here.
Thus we think that you are being irritating for the purpose of being
irritating.  You will have an easier time of it, if you stop assuming
that your expression of American values is common to all Americans,
and that there is any reason for any of to actually care about American
values in the first place.

Laura Creighton




From mwh at python.net  Mon Apr  8 07:49:37 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 8 Apr 2002 11:49:37 GMT
Subject: 2.2.1c2 test_fcntl hangs
References: <mailman.1018152991.30437.python-list@python.org>
Message-ID: <ulmbybe3n.fsf@python.net>

Andrew MacIntyre <andymac at bullseye.apana.org.au> writes:

> On 6 Apr 2002, Jochen [iso-8859-1] K?pper wrote:
> 
> > With latest cvs release22-maint (April 6, 1am EST) test_fcntl hangs.
> > (I couldn't find any postings about this.  I think I have seen this
> > before myself, though...)
> 
> I don't believe there have been any recent changes that might have
> provoked this - at least I don't remember any checkins on the 2.2
> maintenance branch (or the head branch for that matter) that would seem
> plausible as causes (not that I'd rely too much on my memory either).

Nothing substantive has changed in Modules/fcntlmodule.c in *years* (a
large pile of constants have been added, the functions now take file
objects as well as file descriptors and something changed for os/2-emx
support -- but that's not on release22-maint).

> If you have succeeded in building from the 2.2 maintenance branch before,
> what was the checkout time of the last successfully built tree?
> 
> Have you changed anything else on your system in similar timeframe - such
> as glibc?
> 
> Please report this to the SF bug tracker.  I'd suggest assigning it to
> Michael Hudson, as 2.2.1 patch czar.

Indeed; posting bugs to comp.lang.python two days before release is a
staggeringly bad way of going about it.  Cc:ing me is ever so slightly
preferable, bugs on sf (assigned to me, in the 2.2.1 candidate group)
is better.

Though, unless the bug is clear and the fix is obvious, it ain't going
to get fixed at this stage.

Cheers,
M.

-- 
  Not only does the English Language borrow words from other
  languages, it sometimes chases them down dark alleys, hits
  them over the head, and goes through their pockets.   -- Eddy Peters


From aleax at aleax.it  Mon Apr 29 17:56:15 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 21:56:15 GMT
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu>
Message-ID: <3kjz8.32779$8D3.928578@news1.tin.it>

Paul Hughett wrote:
        ...
> Hence my question: Is there an established meaning within Python for
> slice objects with a negative increment?  E.g. slice(1, 10, -2) ?

Sure!

>>> import Numeric
>>> x=Numeric.array(range(20))
>>> print x[1:10:-2]
zeros((0,), 'l')
>>> print x[10:1:-2]
[10  8  6  4  2]

slice(1,10,-2) is exactly what x.__getitem__ gets in the first case (and
slice(10,1,-2) in the second case).  In the first case, since the step is < 
0 and the start index is already < the stop index, an empty array comes
out -- just as it would for x[10:1:2], of course (or x[10:1] for that 
matter).  In the second case, since here the start index is >= the stop
index, then the usual rule (start, start+step, start+2*step) also applies.


> If not, then I am inclined to define slice(i,j,-n) to mean the same as
> reverse(slice(i,j,n)); then slice(1,10,-2) would mean (9, 7, 5, 3, 1).
> This seems simpler for users than the alternatives that I've considered.

Sorry, but I disagree.  Even if your users aren't used to Numeric's slice
behavior, it's exactly what one would expect from other analogous
Python behaviors, e.g.:

>>> range(1,10,-2)
[]
>>> range(10,1,-2)
[10, 8, 6, 4, 2]

> The primary use case that I have in mind is a multi-dimensional array
> type in which it may be necessary to access a given dimension in
> either the forward or backward direction.  Under the proposed rule,
> elements i up to j in the forward direction would be i:j:1 or just
> i:j, while the same elements in the reverse direction would be i:j:-1.

Under Python's rule, they're j:i:-1 instead -- much better, imho.

> Even-numbered elements up to j in the forward direction would be 0:j:2
> and in the reverse direction would be 0:j:-2.

Wouldn't j:0:-2 be better?

> Any comments?

Take care not to let apparent "convenience" overrule simplicity,
regularity, and universal application of rules and analogies.  I think
that, even more than syntax cleanliness, this is what puts Python
apart from Perl, which always put great stock in "convenience".


Alex



From bmason at rand.org  Tue Apr 16 15:33:12 2002
From: bmason at rand.org (Bryce C. Mason)
Date: Tue, 16 Apr 2002 12:33:12 -0700
Subject: Counting the number of elements in a list - help.
Message-ID: <a9hu8e$2fg$1@lumberjack.rand.org>

I need to know these things; may someone help me out?

1) a function which returns the number of elements in a list
2) a function which takes a list and an index and returns a zero when that
list index is out of range.

I want to continue appending things to a list until the list reaches a
specific size.  For example:

while (list_size(list)<10):
    list.append("something")

anyone know of a "list size" function?  It has to be out there..hehe, i'm
just really green at python.  If I knew #2, I could build the function
myself, but #1 is preferable.

Help!

Bryce




From lumber at jack.com  Tue Apr 23 15:56:03 2002
From: lumber at jack.com (Lumberjack)
Date: 23 Apr 2002 19:56:03 GMT
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz>
Message-ID: <Xns91F9838BF5E61lumberjack@216.148.53.83>

"Petr Prikryl" <Answer.via.news.please.prikrylp at nospam_skil.cz> wrote:
> Being quite new to Python -- is there any preference of
> single or double quotes when writing strings?

Strings containing vowels should be quoted with single quotes, except on 
Thursdays. Otherwise use doublequotes except the day after a full moon, in 
which case you should use the triple double quote mechanism. These rules 
will stand you in good stead. UNDER NO CIRCUMSTANCES SHOULD YOU USE ONLY 
ONE OF THE QUOTE CHARACTERS THROUGHOUT YOUR CODE. If you do the Earth will 
swallow you up and you will be looked down upon by others as too simple 
minded to be a programmer.

> While I do not consider this a problem for me (I can sleep
> quite well), I think that proposing some prefered way
> could make the look of Python sources more similar
> each other, i.e. slightly more understandable.

If you are sleeping well then clearly you aren't doing things correctly. 
While nature should take its course, one should never leave these things to 
chance and one can never have too many proposals.


From sholden at holdenweb.com  Sun Apr 28 14:18:36 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 28 Apr 2002 14:18:36 -0400
Subject: portable fortune program
References: <13285ea2.0204270857.2f34f8a4@posting.google.com>
Message-ID: <z4Xy8.72324$z55.71416@atlpnn01.usenetserver.com>

"J.Jacob" <joost_jacob at hotmail.com> wrote in message
news:13285ea2.0204270857.2f34f8a4 at posting.google.com...
> I have always liked the unix fortune program.
> Unfortunately nowadays some *n.x's come without fortune,
> and sometimes I have to work with Windows.  I wanted to be
> able to do:
>     C:\mydirectory> python -c "import fortune; fortune.quote()"
> printing a nice quote to stdout.
> The fortune module has to be able to locate a file with
> the name 'allfortunes'.  And you want everything to be
> portable.  And working with python 1.5.2 too.
>
> The allfortunes file contains quotes separated by lines
> with a '%' as its first character.  I downloaded a couple
> of BSD fortune files and combined them into allfortunes.
> If you put the following fortune.py file with an
> allfortunes file into one of the directories in your
> sys.path, or in you current working directory, it will work.
>
> This is a first hack, it is slow, the allfortunes file is
> not compressed, so comments are elcome, maybe you can
> suggest improvements?  I was trying to put the allfortunes
> file into a compressed string at the end of the fortune.py
> file, so you do not need a separate file anymore, but I
> did not see a best way to do it.  Please let me know if it
> does not work with your python distribution.
>
> joost_jacob at hotmail.com
>
>
[Jacob's original program appears after these comments]

Jacob:

Well done! Utilities like this need to be more widely available. One of the
other problems you have solved (without realizing it, possibly) is that
Windows users can now get their fortune. I'm a strong believer that once
Windows users see how powerful open source applications are they'll be a lot
less worried about using GNU/Linux and the wonderful system tools that come
with them.

You have obviously already thought about speed: the more fortunes you have,
the more reading you do from filestore. With today's modern computers, and
infrequent fortune usage, this probably doesn't matter for many users. So,
you can just shrug your shoulders and ignore this suggestion, which has to
do with locating the individual fortunes with in the file. Alternatively,
you could ask yourself "If this were a fortune server on the web, how would
I give my server the best chances of scaling to 1,000 or 1,000,000 users?"
and start on a hunt for efficiency.

If there were a subsidiary data file (containing the starting position in
the file for each fortune in your original data file) you could try to read
this in, make a random choice from the resulting list of positions (using
choice(), as suggested by another poster), and only then use the fortunes
file. From there a single fortunes.seek()[1] will get you to the start of
the chosen fortune, and you can just read that one fortune and display it.

You might ask: "Where does the subsidiary data file come from?". The answer
is that your program will check for the presence of the fortune file as
always (no harm opening it right at the start, since you have to read it
later), and look in that directory for the subsidiary file. If no data file
is found, the program will attempt to create it if it doesn't already exist
(or if it's older than the fortune file). If you are on a Unix or cygwin
system you could use os.system()[2] and rely on the make(1)[4] command for
this portion of the process, at least at the start..

That way, as long as the file can be created it will be. You should be
careful to ensure your program still works even if it tries but cannot
create the file. This could happen if, say, only root could write into the
directory, and an ordinary user was running the program. Clearly it would be
a great idea to retain your existing code, or something very close to it, to
handle such exceptional circumstances.

What should the format of that file be? A pickle[3] containing a list of
(possibly) long integers would be simplest to create and use in Python.

If you wanted to use the Python source as a fortunes files, one way to do
this would simply be to build a list using code similar to the following
pseudo-code:

    fortunes = []
    fortunes.append("""I tell you three times it is 1""")
    fortunes.append("""We hold these 1s to be self-evident, that all men are
created equal, that they are endowed by their Creator with certain
unalienable Rights, that among these are Life, Liberty and the pursuit of
Happiness.""")
    fortunes.append("""A lie would have no sense unless the 1 were felt
dangerous.
- Alfred Adler""")
    fortunes.append("""The first and last thing required of genius is the
love of truth.
- Johann Wolfgang von Goethe""")

... and so on.

You can probably quite easily see how you could write a program to write
most of this code for you while reading to allfortunes file. Adding so much
code would, of course, make the Python module much bigger. If it were going
to be in continual use as a web server, then, you might want to write it as
some sort of long-running process that only has to start up once before
serving several thousand fortunes. That's getting a bit sophisticated,
though Python is easily capable of handling that sort of stuff[5].

I hope you don't in any way see this as critical of your efforts. As I said
at the start, I think your program is great. This being comp.lang.python you
may well see another set of postings with other, possibly even better, ideas
for improvement.

regards
--
[1]:    http://www.python.org/doc/current/lib/bltin-file-objects.html
[2]:    http://www.python.org/doc/current/lib/os-process.html
[3]:    http://www.python.org/doc/current/lib/module-pickle.html
        http://www.python.org/doc/current/lib/module-cPickle.html
[4]:    http://www.gnu.org/manual/make-3.79.1/html_mono/make.html#SEC1
[5]:    http://www.zope.com/

Steve Holden: http://www.holdenweb.com/ Python Web Programming:
http://pydish.holdenweb.com/pwp/

>
> # file fortune.py
> import random, os
>
> def randlistindex(yourlist):
>     "Return a random index number into yourlist"
>     return random.randint(0, len(yourlist)-1)
>
> def quote():
>     """
>     Return a quote <string> from file allfortunes,
>     assume the file allfortunes is located in the same
>     directory as this module."""
>     datafile = os.path.join(
>       os.path.dirname(__import__(__name__).__file__),
>       'allfortunes' )
>     quotesfile = open(datafile)
>     quoteslist, quote = [], ''
>     for s in quotesfile.readlines():
>         if s[0] == '%':
>             if quote:
>                 quoteslist.append(quote)
>                 quote = ''
>         else: quote = quote + s
>     if quote: quoteslist.append(quote)
>     print quoteslist[randlistindex(quoteslist)],





From peter at engcorp.com  Wed Apr 10 21:10:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 21:10:26 -0400
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu> <eppstein-40A0D7.17104110042002@news.service.uci.edu>
Message-ID: <3CB4E282.AAD468E4@engcorp.com>

David Eppstein wrote:
> 
> In article <a92hl1$1v4c at r02n01.cac.psu.edu>,
>  Rajarshi Guha <rxg218 at psu.edu> wrote:
> 
> > I got this code to work:
> >
> > liter = iter(l)
> > while(1):
> >   try:
> >      print liter.next(), liter.next()
> >   except StopIteration:
> >      break
> >
> > Works fine - is it an efficient way to handle this situation?
> > (I could arrange the data into tuples if necessary)
> 
> Does it work fine?  What if the list has an odd number of items -- the
> last item won't be printed.

It works perfectly, given the requirements specified.

If the customer wanted to handle lists with odd numbers of items
he would have specified the desired behaviour.  Or he will now. :-)

don't-implement-what-wasn't-asked-for-ly yr's,
-Peter


From peter at engcorp.com  Sun Apr  7 19:55:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 07 Apr 2002 19:55:38 -0400
Subject: Newbie:  How to sound an alarm
References: <ucFMmso3BHA.1332@cpimsnntpa03>
Message-ID: <3CB0DC7A.4483E476@engcorp.com>

billludden wrote:
> 
>      I would like to know how to make my box sound a tone or beep when my
> program finishes.
> I posted this query last week but I cannot access messages posted during
> that time (must be Windows).

Did you know you can find the archives for all posted messages by
going to groups.google.com?  In this case, the following search
pops up at least one response to your inquiry:

http://groups.google.com/groups?q=newbie+alarm&meta=group%3Dcomp.lang.python

-Peter


From pyth at devel.trillke.net  Thu Apr 25 11:13:05 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 25 Apr 2002 17:13:05 +0200
Subject: python and very large data sets???
In-Reply-To: <ad381f5b.0204250629.5144d196@posting.google.com>
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <uofg9huaf.fsf@ctwd0143.fitlinxx.com> <ad381f5b.0204250629.5144d196@posting.google.com>
Message-ID: <20020425151305.GB22555@devel.trillke>

On Thu, Apr 25, 2002 at 07:29:27AM -0700, Rad wrote:
> Thanks for all the suggestions and ideas, I'll try to answer to all
> your questions in this one post.
> 
> It is a longish term project but after a few initial data extractions
> I'll have some time to re-implement the procedures if there is a need
> to do so.

good :-)

> At this time I don't know for sure how many unique ID's are going to
> be, not less than 15 million I guess.

how lang are the ID's supposedly? Variable length, fixed length?

> Don't have any options on the way data is delivered, it's going to
> come on tape/s and all relevant files will be in fixed width format. 
> I am OK with that, shouldn't have any problems reading it in and
> storing it on the HD unless DDS starts clicking :(.  I have four 73 GB
> SCSI disks on the computer that will run win2k, 2 GB of RAM, dual xeon
> processor and three power units.

The power units are extremely important :-)

> However, as I just mentioned I don't know anything about MySQL (except that one
> has to pay for it)

where did you hear that? it's wrong as mysql is under an open/free
license. 

> having been learning Python in the last few weeks
> I kind of like it + I was encouraged by your posts and I'm now pretty
> convinced that Python2.2 is the way to go. 

Be careful. Some people pointed out that in the end it might
be a task which a database can handle more reliably. 

> I won't have to access data
> randomly, just pull out the data fitting certain criteria from
> original 4 files and create resulting files.  I might decide
> (depending on the size) to put those into a database though.
> I don't understand how I can interface Python to a database?  

it's not hard. Try looking for 'mysqldb python' or so on google.

> All four files have dates in them but they do come in YYYYMMDD format
> and I was planning to use string comparisons, same for the rest of
> data,I was thinking to treat it all as strings.

Huh? YYMMDD is easily convertible to a 32bit value. see "help('time')"
on the python-prompt.

> My initial plan is to either split files into smaller files that could
> fit into the memory or read chunks of data from original big file
> using file_object.redlines(some_size_that_ can_ fit_ in_ the_ memory).

Reading the file isn't the problem. See my code example in a previous
posting to see how you can read 'chunkwise'.

> I'm stuck with it (the project) and I'm looking forward to the
> challenge.

You should really start to generate some example files (made up
by you) and start to code *right now*. Don't believe these
academics who tell you that you absolutely need a four-year-plan
or a thesis on 'performance problems of todays computer systems'.
(although the latter might help :-)

Even if you later decide to use a mysql-db you still have to apply 
the same basic techniques of reading/writing the files, extracting 
ids and writing stuff. And you have to understand what exactly 
you want to do no matter if it is python or python+SQL or whatever.

have fun,

	holger




From robin at jessikat.fsnet.co.uk  Tue Apr  9 14:35:40 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 9 Apr 2002 19:35:40 +0100
Subject: PEP 285: Adding a bool type
References: <7ZG$xVAfuds8Ew+l@jessikat.demon.co.uk>
 <a8uh5p$nu7$1@thorium.cix.co.uk> <zpF9N5ATtts8Ewc0@jessikat.demon.co.uk>
 <a8v6bp$nc3$1@slb1.atl.mindspring.net>
Message-ID: <DHOR2LA8Rzs8Ew4y@jessikat.demon.co.uk>

In article <a8v6bp$nc3$1 at slb1.atl.mindspring.net>, Andrew Dalke
<dalke at dalkescientific.com> writes
>Robin Becker:
>>Scripts in sed or awk that are 20 years old still run because their
>>languages are stable.
>
>Just this morning I read a message from one mailing list which was
>trying to get their awk script working.  The problem was a difference
>between nawk and awk.  I myself recall prefering nawk over awk because
>it let me access multiple files at the same time.
>
>Since there are several different implementation of awk (awk, nawk,
>and gawk) each with different features, I wouldn't suggest using
>them as good examples.  Judging from the ChangeLog, gawk added some new
>features a year ago.  Eg,
>
.....
I think you make my point as the scripts I refer to didn't use any of
the nawk/gawk features so they still run. I'm sure that there are old
scripts which have been broken by recent sed/gawk/awk implementations. 

Nor do I claim that 1.4 scripts don't run in modern python. My worry is
that some don't. That worry makes every python script less certain.

Of course now you indicate that awk/gawk/sed et al have changed that
will make me doubt the certainty of those scripts.

The implementers of the new features broke things not the original
author, but the vandals get kudos and the innocent pay the bill.

Uncertainty will no doubt reach infinity as the first quantum computers
are being introduced at which fuzzy time no scripts will work.
-- 
Robin Becker


From cbbrowne at acm.org  Fri Apr 26 00:54:04 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: Fri, 26 Apr 2002 00:54:04 -0400
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz>
Message-ID: <m3sn5j83gz.fsf@chvatal.cbbrowne.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> Julian Tibble wrote:
>> 
>> erm... If there is no copyright then there is no *need* for copyleft.
>
> But if RMS would be happy with a world in which there
> was simply no copyright, he wouldn't bother with the
> GPL -- he could get the same effect just by putting
> all the stuff he writes in the public domain.
>
> The fact that he hasn't done that means he
> *does* want *some* kind of intellectual property
> laws.

Two fallacies here:

a) RMS would _not_ get the same effects via a "public domain" license.

   What he wants is for people to share code.  The "in the public
   domain" license certainly supports use; it has no inherent
   mechanism to encourage people to release code "in the public
   domain."

   The "quasi-viral nature" of the GPL is there to expressly encourage
   people to release code under the GPL.  That wouldn't be
   accomplished by "plain old P.D."

b) That "intellectual property" is a straightforward class of things
   that are logically grouped together.

   The only "common" thing between them is that lawyers have a
   specialty area they call "intellectual property."

   Copyright _isn't_ the same as patents which _aren't_ the same as
   trademarks, trade secrets, ad infinitum.  They're not intended to
   do similar things, they use different kinds of legal frameworks,
   they tend to be tied to different government departments, and
   generally function differently.
-- 
(reverse (concatenate 'string "gro.mca@" "enworbbc"))
http://www.cbbrowne.com/info/lisp.html
Walk softly and carry a megawatt laser.


From dale at riverhall.NOTHANKS.co.uk  Fri Apr 19 05:37:23 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Fri, 19 Apr 2002 10:37:23 +0100
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com> <mailman.1019068175.11920.python-list@python.org>
Message-ID: <99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com>

>
>Or it might be preferable to use a dictionary:

But several orders of magnitude more processing. Certainly not for
tight, efficient loops.

--
Dale Strickland-Clark
Riverhall Systems Ltd



From jb at cascade-sys.com  Sun Apr 28 17:29:25 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 28 Apr 2002 14:29:25 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9s72u$gjj$1@slb3.atl.mindspring.net> <mailman.1019381679.10995.python-list@python.org> <m2wuv0xr1u.fsf@mycroft.actrix.gen.nz> <mailman.1019543864.22119.python-list@python.org> <m2g01my7bw.fsf@mycroft.actrix.gen.nz> <mailman.1019798929.16905.python-list@python.org> <tyf1yd25zgs.fsf@pcitapi22.cern.ch> <mailman.1019840809.5035.python-list@python.org> <Kvvy8.51992$z55.49903@atlpnn01.usenetserver.com>
Message-ID: <3CCC69B5.64EC1805@cascade-sys.com>

Steve Holden wrote:

> Don't quote me out of context, dammit. Just because you didn't detect the
> humour didn't mean it wasn't there. And it waas "reeducated", not
> "reprogrammed to think Pythonically".

(a) Ironically, seriously and truly, you weren't the only person to make a
remark like that to me last week.  So I wasn't intentionally quoting you or
anybody specific.  I was merely paraphrasing my sense of the group consciousness
to the best of my recollection. Damn it.

(b) "Just because you didn't detect the humor didn't mean it wasn't there."
Frankly, I think the L. Ron Hubbard/reprogramming angle is way funnier than
being reeducated by the Khmer Rogue.  Damn it.

(c) You need to lighten up, damn it.  There's a whole vast world out there
beyond the kingdom of Python.  Damn it.

> People can also learn to ride bicycles where turning the handlebars to the
> left turns the wheel to the right.
>
> People can even learn to program in C, dammit.
>
> Your problem is you just won't take those damned glasses off again now
> you've got used to them.

What glasses?  Oh, you mean the example *I* used earlier about how people can
adapt to anything?  And how just because people can adapt (e.g., to Python)
doesn't mean it is flawless?

What's your point again?  Damn it.  ;o)

I've programmed in many different languages over the decades and even maintained
and implemented a few myself.  I think my perspective is at least as broad and
objective as anybody here.  Way more so than some I've seen in my short time
here.

You seem to be the one who is locked into just one way of viewing the world.

> True. But passion is a relative judgement. Programming languages and
> operating systems stopped being a religious issue for me many years ago, and
> so I tend to reserve passion for more appropriate activities.

Geeze you sure fooled me.   ;o)

I gotta say, though, if one more ankle biter sycophant quotes the scriptures
according to Guido to me, I think I'm gonna puke.

Elsewhere, Steve Holden adds:

> And I can say dugapoopoo too. I'll cut out the cheerleading and
> defensiveness immediately :-)

I'm not holding my breath.  ;o)

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From noah at noah.org  Wed Apr 24 03:59:58 2002
From: noah at noah.org (Noah)
Date: Wed, 24 Apr 2002 00:59:58 -0700
Subject: Killing all child processes?
In-Reply-To: <59aa3aef.0204231823.77932129@posting.google.com>
Message-ID: <PAENINBOOODFDEPOKPKCEEGFCMAA.noah@noah.org>

You want to kill the process group.
Practically the only change you need to make is to
call os.kill with the negative of the child pid.

So in the line that reads:
	  os.kill(child_pid, 9)
You need to change it to read:
        os.kill(-child_pid, 9)
That kills the given pid and every process that has a
process group id with the same number as the child_pid.

You ensured that all the children had the same group as
the first child (session leader) when you wrote this line:
	os.setpgid(0, 0)
But note that Python has a special call just for that:
	os.setpgrp() # Sets process session leader.

FYI, execlp() takes two arguments because the
first one is the path to the command to exec and the second
one is supposed to become the argv[0] of the command.
But it doesn't seem to do that. I'm not sure why. 

Yours,
Noah



-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Kevin
Sent: Tuesday, April 23, 2002 7:24 PM
To: python-list at python.org
Subject: Killing all child processes?


Hi all,

I can't seem to get my python program to kill all the child prcoesses
that it may fork. Is this possible?

I've researched usenet+the web and read subproc.py from
ftp.python.org, but still no go.

My latest attempt was a python script that forks itself, uses setpgid,
then the child does an execl(). The execl'd program creates some
children. Meanwhile, the parent decides time is up on the child and
tries to kill everything, but  my os.kill(pid,9) only succeeds in
killing the child aka the execvp'd process.

There's another bug that I haven't yet addressed about the parent
waiting around when the child doesn't write to either of the pipes.
But that'd not my focus now, so don't worry about that; I just want
all the children to die.

My full code it at the bottom. Hints, anyone? This is on Linux and
FreeBSD.

Thanks,
Kevin

---Full code follows---
---- pcontrol.py below
#!/usr/bin/python
#
# pcontrol.py
#

import os
import sys
import signal

_alarm = "request timed out"

time = 5

cmd="sh1.sh"

###########################################################

def alarm_handler(a , b):
    global _alarm
    signal.alarm(0)
    raise _alarm

###########################################################

# Okay, let's do this the hard way: create two unnamed
# pipes for stdout/err. Then, fork this process, and the
# child becomes the new command and shoves their output
# onto the pipes, while the parent waits for the child to die.

# Create two named pipes
child_out_r, child_out_w = os.pipe()
child_err_r, child_err_w = os.pipe()

# Let's fork a child that gets replaced with the test process - krf
child_pid = os.fork()

if (child_pid != 0):
    # We're the parent

    os.close(child_out_w) # the parent won't read the pipes
    os.close(child_err_w)

    os.setpgid(child_pid, 0)

    # Now do the timing...
    signal.signal(signal.SIGALRM, alarm_handler)
    signal.alarm(time)

    try:
        child_pid, child_status = os.waitpid(child_pid, 0); # wait for
child
    except OSError:
        print("OSError... did the executable time out?")
        child_status = -1 # this is the same as a timeout for now
    except _alarm:
        print("Timed out")
        # kill it!
        os.kill(child_pid, 9)
        child_status = -1 # timeout

    signal.alarm(0) # clean up alarm

else:
    # We're the child
    os.setpgid(0, 0) # now the child is it's group leader (?)

    # Redirect error to our pipes
    os.close(child_out_r) # the child won't read the pipes
    os.close(child_err_r)
    os.dup2(child_out_w, sys.stdout.fileno())
    os.dup2(child_err_w, sys.stderr.fileno())

    # replace the process
    os.execl(cmd, cmd) #strange, but execl->execv needs a second
arg...
    # the child is no more... (assuming success)
    
print "Status = ", child_status
print "Child stdout [",os.read(child_out_r,999),"]"
print "Child stderr [",os.read(child_err_r,999),"]"


------- sh1.sh below
#!/bin/sh

echo "This is processs (1) $$."
sh2.sh &
sh3.sh &
sleep 600
echo "Process (1) $$ exiting."

------- sh2.sh below
#!/bin/sh

echo "This is processs (2) $$."
sleep 600
echo "Process (2) $$ exiting."

------- sh3.sh below
#!/bin/sh

echo "This is processs (3) $$."
sleep 600
echo "Process (3) $$ exiting."





From fperez528 at yahoo.com  Mon Apr 29 21:34:25 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 29 Apr 2002 19:34:25 -0600
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com> <20020429174425.GN16217@devel.trillke> <mailman.1020115270.9666.python-list@python.org> <aaknh8$1c4$1@peabody.colorado.edu> <mailman.1020129172.25158.python-list@python.org>
Message-ID: <aaksgg$42j$1@peabody.colorado.edu>

James J. Besemer wrote:

>> No, please heed other advice and use the MA approach. Please note that I'm
>> not dissing this post on its own merits, it's a nice discussion, it just is
>> ill-adapted to numeric processing. Numeric provides its own fast way to
>> manipulate arrays with invalid data, by all means use it. After all, speed
>> is one of the big reasons why you use Numeric to begin with (you could do
>> most things in it with regular lists, if you had the age of the universe to
>> wait).
> 
> I would agree with you that it might be worth the extra trouble and effort
> to download, install and learn to use Numerics IF the user in question has a
> big, production application to run.
> 

Quoting the original post:

I have a Numeric array with missing elements from which I wish to calculate
a histogram.  However, in order to do so I must first remove the 'None'
values from the array.  I have a kludgy solution, but I was wondering if
someone could point me in the right direction on what is the proper method
of collapsing a masked array to just the non-masked values.

/quote

That's why I emphasized to him using Numeric's masked array facilities: 
_because_ he said he was using Numeric to begin with (unless his use of 
'Numeric array' with a capital N misled me, and he just means 'a list with 
numbers in it' ;). All in all I liked the thread because it exposed valuable 
alternative approaches to a problem, and that's always a good way to learn. 
But I assume that if the OP was using Numeric to begin with, performance 
probably is a concern to him in the first place.

Otherwise I was happy to read the various alternatives proposed.

cheers,

f.


From phlip_cpp at yahoo.com  Fri Apr 19 11:56:47 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 19 Apr 2002 08:56:47 -0700
Subject: Test-first and mock objects with PyOpenGL
Message-ID: <63604d2.0204190756.43adc6c1@posting.google.com>

Alpha OpenGL Engineers:

I have a large spike-solution ("quicky non-production quality code
base") of PyOpenGL, and I want to refactor it into something healthy.

I don't want to use GuruChecksOutput. This means I want tests that
reveal changes did not affect how the code call the GL libraries.

The standard way to do this is to provide a MockObject in place of
something, call the tested functions, let them call into the
MockObject, and test what they told it. If these values did not change
then the refactor did not break anything important.

However, I don't know if I have to Mock GL itself. Maybe I can replace
OpenGL's back-end with a Mock that renders not to the screen but into
my test code.

Who has done this and knows the pattern for it?

-- 
  Phlip


From hgs at dmu.ac.uk  Fri Apr 19 11:35:24 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Fri, 19 Apr 2002 16:35:24 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4lmbju3fx.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204191630380.24244-100000@neelix>

On 19 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:
>
> > I remembered: I forgot to uncomment all the places where the Tk stuff
> > lives.
> [...]
> > # *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
> > 	-L/usr/local/lib \
> [...]
> > ld.so.1: ./python: fatal: libtk8.3.so: open failed: No such file or directory
>
> That is an old Solaris problem: the dynamic loader searches on /lib
> and /usr/lib for shared libraries, no matter what directories where
> given with -L to the compiler. You have the following options (again
> all alternative):
>
> 1. Set LD_LIBRARY_PATH when running Python. This is most common, but
>    also ugliest, since it requires every Python user to set
>    LD_LIBRARY_PATH.

I did this (in my .login so it doesn't happen again!) and it worked.

> I'd personally go with option 2, and build _tkinter.so, or go with
> option 5.

Nearly all is well now:
Only one test failed: test_largefile.
20 tests were skipped but that would be less if I could enable that
regrtest.py thing you told me about.  Whereabouts is the correct place
to add the options to enable those tests?  Is it configure --something
or do I fix the resulting Makefile...
>
> HTH,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list
>
        Thank you,
        Hugh





From mertz at gnosis.cx  Mon Apr  8 17:27:09 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Mon, 08 Apr 2002 17:27:09 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
 <uavbgcabvl642b@news.supernews.com> <a8p2fb$3da$1@inputplus.demon.co.uk>
 <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk>
 <mailman.1018248742.29568.python-list@python.org>
 <slrnab2inr.8qu.mlh@vier.idi.ntnu.no> <Lpgs8kKkXgpE092yn@gnosis.cx>
Message-ID: <tsgs8kKkXU0F092yn@gnosis.cx>         

|the Lotus-Eaters wrote:
|>    fs = (can_be_moved, can_be_sized, can_be_stacked,
|>          can_be_coloured, can_be_hidden)
|>    major = len([1 for f in fs if f(o)]))

|from operators impor add
|fs = ...
|major = reduce(add, [f(o) for f in fs])

Well... my version is two bytes shorter :-) (more if you count the extra
import).

The real problem, however, with Magnus' version is that if the functions
return some overloaded true value rather than nice 0/1 booleans, his
doesn't work.  I started to try to do a reduce() version, but then felt
like I would have needed a filter() or map() to get cannonical 0/1
values.  My len() version handles functions that might return
empty/non-empty lists, other integers, etc.

Yours, Lulu...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From aahz at pythoncraft.com  Wed Apr 10 19:22:17 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 19:22:17 -0400
Subject: Building strings (was Re: Creating objects in thread created with subclass)
References: <mailman.1018401560.15646.python-list@python.org>
Message-ID: <a92hf9$k7v$1@panix1.panix.com>

In article <mailman.1018401560.15646.python-list at python.org>,
Nick Arnett <narnett at mccmedia.com> wrote:
>
>Ah.  Well.  Then as long as we're on the subject... I couldn't figure out a
>way to build an insert command for MySQLdb without it.  Given a list of
>tuples to insert, I need to give the database handle a statement that looks
>like this:
>
>self.dbh.execute("INSERT INTO Foo (fieldnames) VALUES (%s,[%s])",(data[1:]))
>
>The first tuple in the list of tuples is the fieldnames, which can vary.
>Thus, the number of "%s" occurrences in the statement has to be calculated.
>Is there a way to create this insertion statement without using exec?

fields = []
values = []
for name, value in data:
    fields.append(name)
    values.append("'" + SQLescape(value) + "'")

fields = string.join(',', fields)
values = string.join(',', values)
statement = "INSERT INTO Foo (%s) VALUES (%s)" % (fields, values)

Never underestimate the power of a for loop combined with lists.

(This should be a Cookbook recipe, but the site's down right now.)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"Playing rules lawyer on UseNet is like watching Captain Hook wanking."
--Brian Mailman, news.groups


From dalke at dalkescientific.com  Mon Apr 29 15:07:26 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 29 Apr 2002 13:07:26 -0600
Subject: comparing str's with ints
References: <20020429115856.R1840-100000@palanthas.neverending.org>
Message-ID: <aak5n9$efr$1@slb6.atl.mindspring.net>

Frank Tobin:
>What in the world is supposed to be the semantics of comparing a string
>with an integer?  Seems like a type trap...

As I understand the history, earlier versions of Python didn't support
throwing exceptions from an internal cmp().  Instead, it used a
well-defined but arbitrary comparison scheme to handle failures.
Though that scheme kept being tweaked as mismatches were found between
it and expectations.

I agree that it's a type trap.

                    Andrew
                    dalke at dalkescientific.com






From johnroth at ameritech.net  Mon Apr 15 08:47:17 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 15 Apr 2002 08:47:17 -0400
Subject: Does Python need an 'xor' operator?
References: <mailman.1018819588.582.python-list@python.org>
Message-ID: <ubliudkt3q6s6d@news.supernews.com>

"Tim Peters" <tim.one at comcast.net> wrote in message
news:mailman.1018819588.582.python-list at python.org...
> I doubt that a new "xor" operator will ever get in (compelling use
cases are
> conspicuous by absence), but anyone pushing for it had better push for
a new
> "eqv" operator too (that is, if you're going to claim that it's ugly
that
> x^y may not return one of its argument, surely it's just as ugly that
the
> same holds for x==y).
>

I don't see why, other than providing a 'complete' set of operators.
'and' and 'or' short circuit and return one of their two
operands. 'xor' obviously can't short circuit, but it can be defined
to return the 'true' operand, otherwise False, which seems to be
reasonable, and reasonably consistent with the way 'and' and 'or'
work.

'eqv', however, can neither short circuit, nor is there an obvious
selection of which operand to return when it is true.

Also, 'eqv' is not really a substitute for '=='. Eqv should
be: bool(x) == bool(y), if we follow the behavior of 'and' and 'or'.
In other words 6 eqv 7 is true, since bool(6) == bool(7).

John Roth
>
>




From margerum at yahoo.com  Tue Apr 30 18:19:13 2002
From: margerum at yahoo.com (Mike Margerum)
Date: 30 Apr 2002 15:19:13 -0700
Subject: writing palm pdb files
Message-ID: <dd4c6f61.0204301419.58b7af0b@posting.google.com>

Can someone kindly point me to some python routines to create palm
database PDB files on the desktop?

Thanks a lot,
Mike Margerum


From graeme.matthew at unite.com.au  Tue Apr 16 06:52:29 2002
From: graeme.matthew at unite.com.au (Graeme Matthew)
Date: Tue, 16 Apr 2002 20:52:29 +1000
Subject: Namespaces
Message-ID: <3cbc0275@news.comindico.com.au>

Hi, there can someone help, im currently designing an object model using
Python OO to feed my scripts for a management system

ive got an issue with namespaces

I have a basic Template class and the only way for me to instantiate it is
as follows:

#! C:\\python22\\python.exe

import sys

sys.path.append('C:\\www\\classes')
import Template

template = Template.Template()

print "done"

I cannot get it to instantiate like

template = Template()

any help would be nice thanks





From rdacker at pacbell.net  Tue Apr 30 17:12:05 2002
From: rdacker at pacbell.net (bob ackerman)
Date: Tue, 30 Apr 2002 14:12:05 -0700
Subject: import from a file in a subdirectory
In-Reply-To: <20020430193802.GF16217@devel.trillke>
Message-ID: <EC9CE326-5C7E-11D6-A62D-003065428126@pacbell.net>

On Tuesday, April 30, 2002, at 12:38  PM, holger krekel wrote:

> On Tue, Apr 30, 2002 at 12:27:03PM -0700, rdack wrote:
>> i have a script running in a directory.
>> i want to call a subroutine in a file in a subdirectory.
>>
>> in the first file i said  from subdir/ff.py import  subrtn
>> python doesn't like the slash.
>>
>> how do i do it?
>
> import sys
> sys.path.append('subdir')
>
> import ff
>
> --holger

yes. this works. thank you.





From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 21:26:57 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 18:26:57 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <3cba4cf7@news.mhogaming.com> <eppstein-9E09EB.21122714042002@news.service.uci.edu> <3cba66aa@news.mhogaming.com> <m3lmbpmqij.fsf@mira.informatik.hu-berlin.de> <3cba79ab@news.mhogaming.com> <m3lmbpoyai.fsf@mira.informatik.hu-berlin.de> <3cbaefb5@news.mhogaming.com> <a9fajp$76e$0@216.39.172.122> <m3662sk6xo.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cbb7e33@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3662sk6xo.fsf at mira.informatik.hu-berlin.de...
| bokr at oz.net (Bengt Richter) writes:
|
| > IMO 'internal representation' is a red herring. We have an abstract
| > concept of number representation (twos complement) that corresponds
| > to many hardware devices' usage, but it is the abstraction that is
| > important IMO, not the hardware.
|
| Yes, but this notion is inherently bound with the notion of "size" of
| an integer, atleast in the signed variant (for unsigned integers,
| there is no need for twos complement). If you create the complement,
| you have to know how many bits to invert.
|
| I absolutely agree with your (earlier) conclusion that bit arithmetic
| is best done with unsigned numbers.
|
| > When looking at integers cast as arrays of booleans, hex provides a
| > representation, but w.r.t. boolean values it is not very useful to
| > print it as base-16 integer with sign+absolute value.
|
| I also agree.
|
| > If I infer correctly, Ken is arguing that hex is more often used to
| > view the boolean aspect of integers than their numeric
| > aspect. That's true for me too.
|
| And so it is for me. Still, I disagree that a display of a hex number
| with a sign is illogical or incorrect. In fact, when talking about
| "unbounded" numbers, it is illogical to assume twos complement.
|
| > >Yes-- I read the PEP too-- when this happens, I will consider it a
| > >Python "_WART_", and will set about inventing ways around it for my
| > >own use.  I have forgotten more computer languages than most people
| > >will ever learn-- an I have NEVER found printing a hexadecimal
| > >number in sign and magnitude format to be useful...
| >
| > I'll second that.
|
| At a minimum, you need to propose an alternative behaviour then. I
| must say that I don't find the proposed behaviour of mandatory leading
| zeroes (the 0h proposal) very intuitive, or following tradition.
|
| Instead, there is a simple solution: use unsigned numbers if you want
| to represent bit arrays.
|
| > I would guess so. I do. But for language purposes my preference is
| > to refer to an abstract bit pattern, not hardware (even though
| > hardware mappings will exist and be useful).
|
| That is only meaningful for positive numbers - it just doesn't work
| for twos complement.
|
| Regards,
| Martin

Ahhhh!!!  Got'cha!!!!

Python DOESN'T HAVE UNSIGNED NUMBERS!!!

AND-- the PEP clearly states that (maybe in the future for efficiency) when the
internal representation of a long (most likely that has just been assigned to)
can be represented by an int, then the run-time system is at liberty to convert
it to an int.  Since there is no way to know this happened to or object, we
either have to make an overly complex "Rube Goldberg" type exploration, if we
want to be able to do a logical right shift of an int-long object.  THAT is the
reason for the '>>>' operator-- the designers of Java got rid of unsigned
integers, because many programming mistakes happened in 'C' with them in
existence-- this (in turn) necessitated the creation of a different operator for
'arithmetic right shift', and 'logical right shift'.

Quite frankly, I don't give a damn HOW the long (or int) is represented
internally to the machine-- I simply ALWAYS want the same to interface to an
object that behaves as if it is a twos-complement integer.  When I left shift
it, I don't want the damn SIGN to change-- when I right shift it, I want to be
able to CONTROL the sign bit that is shifted in, when I display the integer in
decimal, I want a minus sign prepended to the string if the number is negative,
and when I print out the number in hexadecimal format, I DON'T want "sign and
magnitude"-- I simply want a string of hex digits with a '0x' prepended to them,
representing the twos-complement form of the number, and I want 2*N hex digits
EVERY TIME, where N is the current size of the integer (in bytes.)

Is what I want so EVIL that Python can't do things that way?




From jepler at unpythonic.net  Mon Apr 15 23:29:05 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 15 Apr 2002 22:29:05 -0500
Subject: pre-PEP for optional 'pass'
In-Reply-To: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <20020415222905.B1046@unpythonic.net>

On Tue, Apr 16, 2002 at 12:26:10AM +0100, phil hunt wrote:
> 
> OK, I've decided to write a PEP. Preliminary to this, I'd like 
> people's comments.
> 
> I wish to make the pas startement optional. That is, in any Python 
> program where there is a pass it can be left out and the meaning of 
> the program is unchanged.

Actually, I'd rather have the 'yada-yada-yada' operator, which I believe is
also slated for inclusion in Perl6.  It's spelled '...' and is placed in the
'atom' production in the grammar:
    atom: ... | NAME | NUMBER | STRING+ | '...'

If the "..." part of an expression is ever reached during normal
evaluation, a NotImplemented exception is raised.

Thus, if you would write
    class MyPrototype:
	def __init__(self, x, y, z): ...

Attempting to create an instance of MyPrototype would raise
NotImplemented.  Also, if you have a complicated expression that you've only
partly finished, such as
    def xor(a,b): return (a or b) and not (a and ...)
this will raise NotImplemented when a is true, since the author hasn't
finished.

The new syntax
    [1, 2, 3, ...]
would create a generator that gives successive positive integer values.  
    [2, 3, 5, 7, 9, ...]
would generate successive primes.  However, if Python cannot guess what
series the leading items define, then it will raise NotImplemented when
the 'yada' term is reached.  Two examples of this might be
    ['99 bottles of beer on the wall', '99 bottles of beer', 
	    'take one down', ...]
    ['spam', 'spam', 'spam', 'spam', 'spam', ...]
though of course Python reserves the right to recognize new patterns in
bugfix releases.

Jeff
PS Hey, it's still April! 




From stom at hotpop.com  Thu Apr 25 01:30:21 2002
From: stom at hotpop.com (Stom Fairecroire)
Date: Thu, 25 Apr 2002 13:30:21 +0800
Subject: Python v. Euphoria
References: <394fcus92g8rt0e2ektsd36s9qu7cq2cij@4ax.com>
Message-ID: <u15fcu4p9l23par9hogki8j0l6mq1v5for@4ax.com>

On Thu, 25 Apr 2002 13:23:31 +0800, Stom Fairecroire <stom at hotpop.com>
wrote:


errrr - Hewlett, even
>Stom



From bruce_dodson at bigfoot.com  Fri Apr 26 19:47:58 2002
From: bruce_dodson at bigfoot.com (Bruce Dodson)
Date: Fri, 26 Apr 2002 20:47:58 -0300
Subject: Python.NET?  Jython.NET?  Ni!
Message-ID: <RGly8.3497$Z6.36288@nnrp1.uunet.ca>

Is ActiveState's Python for .NET dead?  I can find no reference to it on the
ActiveState page, other than an old research build for .NET Beta 2.

Seems to me the easiest place to start for a .NET version of Python would be
with Jython.  I haven't looked at the Jython source code, but can someone
who has done so comment?




From sholden at holdenweb.com  Mon Apr  8 10:46:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 8 Apr 2002 10:46:02 -0400
Subject: Newbie's 1st Web App decisions!
References: <a8i8gn$rbe$1@hfc.pacific.net.hk>
Message-ID: <p4is8.795$eJ6.210@atlpnn01.usenetserver.com>

"Scrumpy" <reply at in_newsgroup.tks> wrote in message
news:a8i8gn$rbe$1 at hfc.pacific.net.hk...
> I want to learn Python for use in creating web apps and possibly to create
> stand-alone apps with wxPython.
>
> To help you Python scholars understand my limited programming/scripting
> background I've;
> 1) Created apps within the following environments - dBASE IV, Lotus
> Approach 2 & MS Access (an old version) with limited use of their related
> languages and no standalone exes and;
> 2) Created/coded HTML4.01+CSS2 static freebie websites which included
pages
> of formatted data created by merging HTML templates with CSV databases.
>
> Mainly my background is in IT/Network mgmt.
>
[ ... ]
>
> As far as books are concerned (a must) I'm thinking of buying
> Web Programming in Python (this is written partly(?) by the folks from the
> Slither project).
> http://www.amazon.com/exec/obidos/ASIN/0130410659/qid=1017861581/sr=1-
> 3/ref=sr_1_3/104-5555059-2897528
>
I can recommend this one, as the author is a close personal friend ;-) But,
to set the record straight, apart from small acknowledged contributions, the
writing (and therefore the mistakes) are all mine apart from David Beazley's
contribution of the three XML chapters.

[ ... ]
>
> Webware seems to be one of the more respected projects and I get the
> impression that it improves the productivity/time to completion for people
> using it.
>
Webware is indeed a flexible and friendly framework, with much to recommend
it.

[ ... ]
>
> What I would be most appreciative of, is your helpful suggestions on how
to
> start in Python Web app development and on relevant books.
>
If you already have a server, just start writing CGI scripts for it. There
is a LOT of material on the web explaining how to do this.

Once you are familiar with web technologies, then pick up Webware or similar
and start to build applications using your framework of choice. Come to
c.l.py with questions!

> Finally (at last you cry!), although Zope is one way to go, as this is for
> me primarily a learning experience, I would prefer to stay closer to
Python
> and 'perhaps' enjoy better web app performance. Zope does seem to be
better
> suited for creating CMS apps but I am not creating a CMS app :)
>
I would tend to agree with you on this one. Although the Zope book list is
slowly growing, there's a lto to learn before Zope makes sense.


"Python Web Programming" is intended to explain enough of the technology, in
enough detail, that you can build web systems in Python without having to
use an external server. It doesn't try to argue that this is always the best
way to go, though.

regards
 Steve






From magnus at hetland.org  Tue Apr 16 08:50:43 2002
From: magnus at hetland.org (Magnus Lie Hetland)
Date: Tue, 16 Apr 2002 14:50:43 +0200
Subject: Deprecate input()
In-Reply-To: <1018897091.1149.281.camel@localhost>; from ianb@colorstudy.com on Mon, Apr 15, 2002 at 01:58:11PM -0500
References: <a9cv63$qhu$1@bob.news.rcn.net> <slrnablens.5k9.mlh@vier.idi.ntnu.no> <1018897091.1149.281.camel@localhost>
Message-ID: <20020416145043.F12735@idi.ntnu.no>

Ian Bicking <ianb at colorstudy.com>:
>
> On Mon, 2002-04-15 at 06:35, Magnus Lie Hetland wrote:
> > But (in the (very?) long run) I think that the name "input" should
> > then be used for what is now raw_input. I know -- breakage, and all
> > that... I just think it fairly odd to have *only* raw_input, since the
> > "raw" qualification then becomes fairly meaningless.
> 
> Why not just make input into what it should be, a moderately-evaluating
> raw_input... like:
[snip]
> Of course, it gets more complex if you allow lists, dictionaries, and
> tuples... which you probably should.  But that's certainly doable while
> still being safe.

A good idea, IMO, although I think such beasts as lists and tuples
might cause the semantics to become a bit muddled...

>   Ian
> 

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org




From phr-n2002a at nightsong.com  Thu Apr 25 08:15:09 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 25 Apr 2002 05:15:09 -0700
Subject: SHA-based encryption function in Python
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <1a517b5.0204241301.1e97c170@posting.google.com> <7x4ri0oh2j.fsf@ruckus.brouhaha.com> <3CC7889D.2080002@nowhere.org>
Message-ID: <7x8z7cvusy.fsf@ruckus.brouhaha.com>

Bryan Olson <fakeaddress at nowhere.org> writes:
> Let me see if I understand.  We start with a shared secret key called
> "key", which may be of any length.  Each message requires it's own nonce
> of 16 bytes....
> The ciphertext is the exclusive or of the plaintext and the running key.
> The nonce travels with the ciphertext.

Yes, this is correct.

> Now I have a few suggestions.
> 
> The system presents a cipher, a MAC, key derivations, and formats. I
> think these should be defined individually.  I agree with defining how
> to encrypt-and-sign with a master key, but I think the smaller elements
> should also be defined and exposed on their own.

I guess that could be worthwhile--any comments from comp.lang.python
readers who might actually want to use this thing?  So far, only the
sci.crypt crowd seems to be saying anything.

> I would separate the IV from the ciphertext, and make computing the MAC
> a separate function.  I see no need to include the nonce in computing
> the auth_key. In some protocols, such as SSL/TLS, we would send a stream
> in many sections.

I thought of adding a function that would allow encrypting in segments
but don't currently have applications that need that.  Remember that
the main goal of this code (and I didn't expect it to become such a
complicated project) was to just have something in portable Python
that could be used instead of the silly rotor module.  It's slow
enough that it would be painful to encrypt large files with it.  And
in an SSL-like protocol it still seems ok to encrypt-and-sign each
message.  (You couldn't implement real SSL with this module because of
the nonstandard cipher).

> Here's a candidate reference implementation.  Each function does one
> thing that I believe should be individually defined.  The code is
> written for lucidity and simplicity, not speed. I'm using Paul Rubin's
> latest (at the time of this writing) MAC, but really I recommend going
> with HMAC.

OK, I might reorganize the code somewhat, to supply those additional
functions.  I've also come up with what should be a pretty fast way to
do HMAC in Python (see my other followup), so I'll switch the MAC to HMAC.

Thanks as always.

Paul


From broadview at netsgo.com  Wed Apr 17 20:50:24 2002
From: broadview at netsgo.com (Haegi Kim)
Date: Thu, 18 Apr 2002 09:50:24 +0900
Subject: Type Conversion to COM Variant
Message-ID: <a9l482$mqo$1@news1.kornet.net>

Hi, experts,

I have a COM client which gets a reference to a COM server.
The COM client has a exec expression in order to execute a script in which
the reference to the COM server is used.
It works find with the COM server in InProcServer mode, but doesn't in
LocalServer and raises the following error.
----------------------------------------------------------------------------
-----------
exceptions.TypeError: Objects of type 'instance' can not be converted to a
COM VARIANT"
----------------------------------------------------------------------------
-----------
In what case the type conversion causes the error? And what should I do?
Thanks in advance.




From chris.gonnerman at newcenturycomputers.net  Sat Apr 13 01:15:31 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 13 Apr 2002 00:15:31 -0500
Subject: What is free software? [Re: Licenses and Open Source don't conflict.]
References: <mailman.1017656355.5887.python-list@python.org><12bb92e6.0204091223.738e0798@posting.google.com><i_nt8.15754$ir6.679622@news1.calgary.shaw.ca><GrAt8.11681$F43.8566@atlpnn01.usenetserver.com><eGIt8.21408$de1.1000064@news3.calgary.shaw.ca><MPG.172101b4b7f13f639896ae@news.nwlink.com><1vKt8.21374$Kq4.896208@news2.calgary.shaw.ca><MPG.17211dcc4bd939939896af@news.nwlink.com> <oq1ydk2u0c.fsf_-_@titan.progiciels-bpi.ca>
Message-ID: <005001c1e2aa$3d554280$0101010a@local>

----- Original Message -----
From: "Fran?ois Pinard" <pinard at iro.umontreal.ca>


> > > Public domain software is [...] a
> > > special case of non-copylefted free software, [...]
>
> [Jeff Shannon]
>
> > Which *does*, in fact, mean that it is free software.  Unless perhaps
> > free software means something other than free software?
>
> There are many definitions of free software, and some people fight over
> which definition is the real one.  Some even want monopoly of freedom! :-)

There is a specific disjunction between "free software" as in the GNU def
above, and "Free Software" (with caps as shown) as used by Stallman.  He
uses the second form as a sort of synonym for GPL/LGPL programs... and he
says he's trying to avoid confusion :-)

> How I understand the GPL is that a software is not free, if it can be
> used to create a bound between parties, in such a way that one of the
> parties has lost part of its freedom on that software through that bound,
> in particular the freedom to modify the software and/or share it around.

This is essentially what Stallman finds offensive.

> A public domain software can be modified by anyone, and the result can be
> made available to someone else under a binding license.  So, that software
> was indeed free, but that quality of being "free" was totally fragile...

Exactly.  MIT (X11) and BSD licenses exhibit similar fragility.

> The GPL wants to help programs at being free in more permanent ways.

This always leads to a flamewar :-(  There are those (uniformly programmers)
who hate the GPL because it restricts their ability to "hide" the source if
they incorporate GPL'd code in their non-GPL programs.  Frankly I don't
understand the problem; if programmer Paul wants to create a proprietary
program, naturally he should avoid programmer George's GPL library.  Yelling
about it is useless, as George has just as much right to restrict those
using
his library as Paul has to hide his source.  In my opinion Paul is more
"wrong"
(or at least restrictive) than George.

Another reason often given (by the misinformed) is that GPL code can somehow
"infect" non-GPL free code (BSD let's say); it goes something like this:
Bob
writes an excellent library and releases it under the BSD license.  George
has
his previously mentioned GPL'd library, which is also most cool.  Fred comes
along, and writes a program which uses both Bob's and George's libraries,
and
for some reason Fred actually incorporates the library source in his project
rather than expecting you to separately install it.  Fred must release his
program under the GPL (he has no choice in this matter), but this has *no
effect* on the license of Bob's library.  Of course, as a GPL product Fred
is
required to release the code he got from Bob even though Bob didn't require
that.

The assertion is that Bob's library source has been "infected" by the GPL,
but
unless Bob agrees to this, Fred has no right to change the license.  Again,
the
GPL covering Fred's code *forces* him to reveal Bob's code, even though Bob
doesn't require him to do so.  Extracting just Bob's code may be easy or it
may
be hard, but if Bob has written his comments well, a third (fourth?) party
should be able to find Bob's source on the Web and avoid the middleman.

But as I said, this always leads to a flamewar... and here it comes, I'm
sure.

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





From brian at sweetapp.com  Mon Apr 29 15:03:35 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 29 Apr 2002 12:03:35 -0700
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <3CCD9649.98AED86F@cascade-sys.com>
Message-ID: <007401c1efb0$91d9d7e0$445d4540@Dell2>

> > Why the hostility?
> 
> I don't interpret this strongly worded and slightly sarcastic critique

> to be hostile.
> 
> If you call this hostile I don't think you know the meaning of the
word.

Decide for yourself:

1 a : of or relating to an enemy <hostile fire> 
  b : marked by malevolence <a hostile act> 
  c : openly opposed or resisting <a hostile critic> 
      <hostile to new ideas> 
  d : not hospitable <a hostile environment>

> And I don't see how his offering a strong opinion stiffles discussion.

I've observed in the past that people don't like to be attacked
personally.
 
> I happen to agree with just about everything he said.
> 
> This PEP is angles in the clouds stuff.  A lot of work for everybody
and
> virtually zero practical benefit to anybody.

Then Lumberjack should attack the PEP, not the author e.g. 

   "How about you and other PEP writers take some responsibilities 
   for your fantasies for once? You up to helping thousands rewrite 
   their code? I thought no"

Seems like an attack on PEP writers in general, not on the PEP itself.

Cheers,
Brian





From robin at jessikat.fsnet.co.uk  Wed Apr 17 03:36:23 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 17 Apr 2002 08:36:23 +0100
Subject: Python 2.2 unwise.exe
References: <a9ids1$j54$0@216.39.172.122>
 <mailman.1019004272.10841.python-list@python.org>
Message-ID: <FWvAiLA3XSv8Ewwi@jessikat.fsnet.co.uk>

In article <mailman.1019004272.10841.python-list at python.org>, Tim Peters
<tim.one at comcast.net> writes
.....
>I know all that stuff, but it doesn't solve the problem.  Getting rid of the
>"delete in-use files" gimmick does.  I don't know whether that's the *only*
>problem here, though (Robin Becker seems to think not, but I don't believe
>everything I read on Usenet <wink>).
....
well actually I think your hypothesis is correct.
-- 
Robin Becker


From jb at cascade-sys.com  Sun Apr 21 03:07:14 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 21 Apr 2002 00:07:14 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1>
Message-ID: <3CC26522.447E4A10@cascade-sys.com>

Carl Banks wrote:

> Computer programming is not science.
>
> The study of algorithms, their properties, etc., is science.
> Implementing a quicksort algorithm is not science, but studying its
> properties is.
>
> As long as we restrict computer science to the study of computer
> algorithms and such, then computer science is science.  If we use
> computer science as a sexy synonym for computer programming, then it
> isn't.

I agree 100% with all the above.

And there are other software related endeavors that may fairly be considered
science besides the ones listed above.

I also agree with someone else's remark that a lot of the "measurements" promoted
in the guise of "science" are mere numerology.

> Don't even get me started on whether is computer engineering is
> engineering.

Depends on the nature of the work.

The vast bulk of digital design hardly qualifies, even if computers play a role
in the design.  However, I think the few and select teams of people who design
computers on a chip from whole cloth include individuals who may fairly be called
"computer engineers".  Some from-scratch OS development similarly can fairly be
called engineering.

What I count as "software engineering" in my own background includes mostly
"embedded" programming -- software tightly coupled with special purpose
hardware.  At times this kind of programming may fairly be characterized as real
engineering.  Timing constraints, system performance and how software affects
external events is central to the design task.  I worked for a while as a
contractor at one of HP's printer divisons.  Everything was picoliters, meters
per second, megabytes per second and microseconds.  I routinely used algebra,
trigonometry and physics in the software design or specifications.

But yeah, most computer programming (including a fair amount of embedded work) is
more of a routine task and can hardly be considered engineering.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From gh_pythonlist at gmx.de  Mon Apr  8 03:50:14 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 8 Apr 2002 09:50:14 +0200
Subject: IsList?
In-Reply-To: <slrnab2feb.atp.mlh@vier.idi.ntnu.no>
References: <9F9s8.37463$tg4.444710@vixen.cso.uiuc.edu> <slrnab2b3b.1gd.gerhard@lilith.my-fqdn.de> <slrnab2feb.atp.mlh@vier.idi.ntnu.no>
Message-ID: <20020408075014.GA1895@lilith.my-fqdn.de>

* Magnus Lie Hetland <mlh at vier.idi.ntnu.no> [2002-04-08 06:50 +0000]:
> In article <slrnab2b3b.1gd.gerhard at lilith.my-fqdn.de>, Gerhard
> =?iso-8859-15?Q?H=E4ring?= wrote:
> >John wrote in comp.lang.python:
> >> Hi, How can one know wheter var 'a' is list or not? Is there any
> >> IsList-like function?
> [snip lots of stuff]
> 
> If you want to check for sequences in general you can use the
> isSequenceType function:

See also this more recent post con operator.isSequence/Mapping/Type from
Tim Peters in 2001, which suggests Guidi still thinks they're bad:

http://groups.google.com/groups?selm=mailman.990826839.12032.python-list%40python.org

I also remember Alex Martelli argumenting against their use.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 1.9 ?C      Wind: 1.8 m/s




From dalke at dalkescientific.com  Sun Apr  7 12:02:37 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 7 Apr 2002 10:02:37 -0600
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com>
Message-ID: <a8pqkh$492$1@slb1.atl.mindspring.net>

John Roth wrote:
>People who were using bools for your Democrat or Republican
>example were simply using them incorrectly. In most states, there
>were way more than two choices: the Green party got enough votes
>to throw the election to Bush. If he hadn't run, Gore would have
>won!

Just as an off-topic reminder.  The difference between Bush and
Gore was less than 2,000.  Given the final Florida results of
    Bush    2,909,176, Republican
    Gore    2,907,451, Democrat
    Nader      96,837, Green
    Browne     18,856, Libertarian
    Buchanan   17,356, Reform
    Phillips    4,280, Official Constitution Party
    Hagelin     2,287, Natural Law

That means Hagelin, the *7th* place party, had enough votes to
"throw the election to Bush."

                    Andrew
                    dalke at dalkescientific.com






From owen at nospam.invalid  Tue Apr 16 13:17:30 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Tue, 16 Apr 2002 10:17:30 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk>
Message-ID: <a9hmba$2p26$1@nntp6.u.washington.edu>

In article <slrnabmoci.dhk.philh at comuno.freeserve.co.uk>,
 philh at comuno.freeserve.co.uk (phil hunt) wrote:

>OK, I've decided to write a PEP. Preliminary to this, I'd like 
>people's comments.
>
>I wish to make the pas startement optional. That is, in any Python 
>program where there is a pass it can be left out and the meaning of 
>the program is unchanged...

I'd go along with it and offer a third rationale:

The need for pass is a pain when commenting out entire blocks of code 
(e.g. while debugging). Instead of just commenting out the code, one 
must also add a "pass" statement, and then try to remember to take it 
out again when re-enabling the code.

In my opinion an optional pass is a very good thing -- sometimes you 
want to be able to assert "this block intentionally left blank", but 
sometimes it makes more sense to just accept a truly empty block.

Thanks for proposing this. I'm sure you'll get lots of horrified 
reactions (remember the fuss whenever someone suggests removing the 
required colon at the start of blocks? Forgetting that damn colon is 
still the most common mistake I make when coding Python, so I'd love to 
see it go, but it'll be a cold day in hell...)

-- Russell
-- 
Return      owen
address     astro
garbled     washington
in header   edu


From tim.one at comcast.net  Sun Apr 28 16:40:15 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 28 Apr 2002 16:40:15 -0400
Subject: Python's Lisp heritage
In-Reply-To: <aahhtq$7lq$1@slb4.atl.mindspring.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEKJPBAA.tim.one@comcast.net>

[synthespian, to Andrew Dalke]
> <sarcasm> No kidding! You can modify your classes?! Gee, I wonder if you
> can do * that * in C++ </sarcasm>
> But can you modify them * in runtime * ?! Without halting your
> application and, say, recompiling? <smirk>
> Answer: NO
> Only in LISP can you do that.
> That's why NASA is so fond of using LISP on their probes.
> And what's this discussion Python X LISP? I thought Peter Norvig's
> article already had cleared that up (that Python is a LISP from people
> who never understood CONS).

[Andrew Dalke]
> I'm not sure what your point is here.

That's because it's Usenet.  In person, it's very easy to tell when someone
is mooning you.

> In more recent Pythons (1.5 and newer, I think) you can do
> ...

He doesn't care, you know.

a-creative-application-of-a-banana-would-be-more-appropriate-ly y'rs  - tim





From erno-news at erno.iki.fi  Tue Apr 16 05:41:09 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Apr 2002 12:41:09 +0300
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com>
Message-ID: <ku4ricgeve.fsf@lasipalatsi.fi>

In article <87d6x09trz.fsf at mathdogs.com>, Mike Coleman
<mkc+dated+1021521407.f909ec at mathdogs.com> writes:

| Does anyone know why strings (i.e., those of length >1) don't share
| their data in Python?  Since their immutable, it seems like this
| would be the obvious thing to do.

in addition to reasons presented i other replies, this would
also result in much less obvious memory usage. saving a short
slice of a huge string would hang on to the huge string. java
does this, and i've heard people make less than flattering
comments about this feature.

  -- erno


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 01:46:04 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 22:46:04 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <3CBA64AD.AE4C563D@cosc.canterbury.ac.nz>
Message-ID: <3cba6963@news.mhogaming.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3CBA64AD.AE4C563D at cosc.canterbury.ac.nz...
| Bengt Richter wrote:
| >
| > On thinking about it, I think this is ugly, though probably
| > convenient in the implementation. I.e., looking at the bits
| > of the absolute value is not usually what you really want when
| > you convert to hex, IMO.
|
| The way I see it, if you're doing bit twiddling
| using unified ints/longs, you should always be
| working with *positive* integers, in which case
| the hex representation will always be what
| you want.
|
| The only reason negative integers come into
| the picture now is that setting bit 31 of a
| non-long int has the side effect of making it
| negative. Once ints and longs are unified,
| this presumably will no longer happen.

"presumably" ???

Well, WILL it, or WON'T it?  The functioning of ints and longs
should be completely transparent to the programmer, using ANY
operator, and the printing of the value in hexadecimal should
also not be different.

"presumably" just doesn't cut the mustard with me.  Unification
should unify everything-- not just "some" operations...

When you multiply an int that has the value: 0x40000000 by 2,
you get 0x80000000L --- which is just as it should be.  BUT,
when you now divide it by 2, it should change back to an int.
This should all happen automagically, "under the hood", without
the programmer even having to worry about it.

When you create a string of a long using the hex() method, the
'-' sign should not even be printed-- you should rather get the
true number of bytes (in twos-complement hexadecimal) that are
currently being used to contain the object.

These are some of the things that need to happen _IF_ we truly
want seamless unification of the int and long types.





From aleax at aleax.it  Wed Apr  3 01:18:46 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 06:18:46 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA335F.4A17F6FF@alcyone.com> <uakogq1r0i0524@news.supernews.com> <7xpu1h4h1m.fsf@ruckus.brouhaha.com>
Message-ID: <a9xq8.53684$S52.1860202@news2.tin.it>

Paul Rubin wrote:

> "John Roth" <johnroth at ameritech.net> writes:
>> I absolutely agree. For me, the entire point is providing
>> a **standard** way of expressing **in the code** the
>> notion that a particular variable only takes two values:
>> True and False. This is such a common case that it deserves
>> to be promoted to be a part of the language.
> 
> Why not ask for enumeration types then?

Enumeration types would be VERY nice to have.  But I
would NOT want x<y, x in y, x is y, and so on, to return
values from an enumeration type...

Besides, having enumerations, or booleans, would *NOT*
"express in the code the notion that a particular variable
only takes [N] values".  For that, you'd need some kind of
declaration.  Python does not have declarations: and THAT
is what this requirement of Mr Roth amounts to.  There is
no way in Python to "express in the code" any such
constraint on a variable -- that's part of why Python is
so GREAT.  I hope Python stays pythonic... there are so
many languages which appear to be mostly about imposing
constraints ("expressing constraints in the code"), why can't
we keep ONE language whose point is instead that of
letting us be happily productive...?-)


Alex



From anthony at interlink.com.au  Thu Apr 18 07:28:51 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Thu, 18 Apr 2002 21:28:51 +1000
Subject: Installation problem 
In-Reply-To: Message from Michael Hudson <mwh@python.net> 
   of "Thu, 18 Apr 2002 10:00:27 GMT." <lkads1mif8.fsf@pc150.maths.bris.ac.uk> 
Message-ID: <200204181128.g3IBSpm14087@localhost.localdomain>

>>> Michael Hudson wrote
> Oleg Broytmann <phd at phd.pp.ru> writes:
> >    It seems Python 2.2.1 has got a lot of compilation problems on different
> > versions of Solaris :(
> 
> Also that I go *no* such reports during the c1, c2 process :(
> Quite a few people got it to build and pass the test suite...

... including me, on systems at work (sunos5.7) and on the sourceforge
compile farm (sunos 5.8).

I'm mystified as to why it seems to be a problem platform...


-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From tmckernan at dyrectmedia.com  Thu Apr  4 11:01:47 2002
From: tmckernan at dyrectmedia.com (Tim McKernan)
Date: 4 Apr 2002 08:01:47 -0800
Subject: parameter limit in zope xml-rpc
Message-ID: <43077ab.0204040801.1f96a36a@posting.google.com>

I'm running Zope 2.4.1 and python 2.1, on an NT box.  I'm using Zope
and python as an xml-rpc server to pass functions on to a group of C
functions.  Based on the error message below it seems that there is
either 1) a limit to how many parameters can be sent over the wire
using xml-rpc, or 2) a limit to how many parameters zope will accept. 
Can anyone explain why I am seeing this?


1) I create a method called OfferAppts with 9 parameters.  I test it
with a Java xml-rpc client that sends in 9 parameters and it returns
an appropriate response from the server.

2) I create a method called OfferAppts with 10 parameters.  I test it
and with a Java xml-rpc client that sends in 10 parameters and this is
what the server returns:

  <h2>Zope Error</h2>
  <p>Zope has encountered an error while publishing this resource.</p>

  <p>
  <strong>Error Type: TypeError</strong><br>
  <strong>Error Value: too many arguments</strong><br>
  </p>


Thanks,
Tim


From bokr at oz.net  Mon Apr 15 13:35:25 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Apr 2002 17:35:25 GMT
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net> <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1>
Message-ID: <a9f30t$mmo$0@216.39.172.122>

On Mon, 15 Apr 2002 11:35:13 +0000 (UTC), Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote:

>Scott David Daniels <Scott.Daniels at Acm.Org> wrote in
>news:1104_1018820620 at news.dsl-only.net: 
>
>>> Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:
>>> > Thank you very much for pointers, but I can think of something not
>>> > listed there. For example, __all__ for "from module import *",
>>> > pickle protocols, __new__ for subclassing from a type and so on. Is
>>> > there a complete (or near-complete) list?
>>> ...
>> 
>> You could also take this lack of a list as a charter for your
>> collecting _all_ (not __all__) such names, and creating a document for
>> others who, like you want to know the list of such names.  This could
>> include package references if you so desire (others will complain if
>> you leave something out, providing you with more fodder for your
>> document). 
>> 
>
>To get you started, here is a list built by grepping over the Python 
>sources. Some of the names are probably red herrings, and there are 
>probably some used by libraries not in this list, but it should give you a 
>start:
>
[... grepped sorted list ...]

Cool. Do you feel motivated to do the same on the .../doc/* tree and
then do a diff? ;-)

Regards,
Bengt Richter


From cliechti at gmx.net  Tue Apr 30 19:41:03 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 May 2002 01:41:03 +0200
Subject: Bulding arrays from text file data.
References: <28bbdc67.0204301518.5e7c41c8@posting.google.com>
Message-ID: <Xns92011165F47Bcliechtigmxnet@62.2.16.82>

joe.woodward at padtinc.com (Joe Woodward) wrote in
news:28bbdc67.0204301518.5e7c41c8 at posting.google.com: 
> Does anyone know of a faster way to build an array out of text file
> with formatted data. What I have been doing works well, but now my
> files and 200MB and up. The following way builds a Numeric.array
> without knowing the size to start with. I can then resize it after the
> fact.
> 
> 
> yldpos=Numeric.array(map(float,re.split('\s+',string.strip(open('datafi
> le.txt').read())))) 

note that string.split() without args defaults to one or more whitespace 
too and i think its faster than an re.split

maybe it would also be faster if you read in blocks so that the strings 
which you split don't get too big. (maybe readlines(sizehint) with sizhint 
on 8192 or so)

don't know, but i would try it :-)

chris

-- 
Chris <cliechti at gmx.net>



From bhan at andrew.cmu.edu  Mon Apr 29 23:11:57 2002
From: bhan at andrew.cmu.edu (Benjamin Han)
Date: Mon, 29 Apr 2002 23:11:57 -0400
Subject: Newbie question: how to join a list of elements of user-defined types?
In-Reply-To: <mailman.1020132292.8898.python-list@python.org>
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org>
Message-ID: <200204292311.57363.bhan@andrew.cmu.edu>

On Monday 29 April 2002 10:02 pm, Cliff Wells wrote:
> On 29 Apr 2002 18:54:12 -0700
>
> Fortepianissimo wrote:
> > This question is so fundamental that I strongly suspect it's already
> > answered - ig that's the case, please forgive this poor newbie, for he
> > has done some work searching over Python FAQ and Google.
> >
> > Basically I want to have a subclass of list, like
> >
> > class MyList (list):
> >     ...
> >
> > then I want to
> >
> > l=MyList()
> > ... do stuff to fill l
> > s=" ".join(l)
>
> s = " ".join([str(i) for i in l])

This surely looks nicer. I'm just wondering how far Python has fared in terms 
of this concept of OO? This example seems to show a very different character 
of OO in Python, in that __str__ is not automatically called, which seems 
rather counter-intuitive.

Without private members, and with seemingly complex ways of dealing with class 
members and methods - so what's the state of OO in Python? Is it still 
rapidly evolving? How much is an effort in converting Python into C++? Anyone 
would like to comment on some of these questions, or point to some article, 
FAQ (esp. helpful from a C++ programmer's point of view)?

Thanks for all those replied,

Ben


From boud at valdyas.org  Fri Apr 12 01:09:32 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Fri, 12 Apr 2002 07:09:32 +0200
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
References: <oOjt8.8815$vF6.254710@news2.tin.it> <P4qt8.15576$Kq4.661039@news2.calgary.shaw.ca>
Message-ID: <3cb66c5c$0$43849$e4fe514c@dreader1.news.xs4all.nl>

Brad Bollenbach wrote:

> In article <oOjt8.8815$vF6.254710 at news2.tin.it>, Alex Martelli wrote:
>> Studying PyQt (latest releases of everything), I started noticing
>> segmentation faults as I quit my application and eventually whittled it
>> down back to...:
>> 
>> #!/usr/bin/env python2.2
> [snip strangeness]
> 
> For what it's worth, and if it's of any help to either martellibot or
> Phil, the code runs without error -- clicking Quit or closing via the
> "X" -- on Debian woody, Python 2.1, and with the python-pyqt package
> installed.

Yes, it's fairly recent behaviour, since it performs flawlessly with
the PyQt included with SuSE 73, but not with PyQt 3.1.

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


From tdelaney at avaya.com  Wed Apr  3 02:57:22 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 3 Apr 2002 17:57:22 +1000 
Subject: PEP 287: reStructuredText Standard Docstring Format
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A24A@natasha.auslabs.avaya.com>

> > Huh? If you'd read the accompanying docs, you would know 
> about the ReST
> > directives. I wouldn't call the .. contents:: directive 
> "explicit markup"
> > considering that it inserts an entire table of contents 
> with one short line
> 
> Of course it's explicit markup, regardless of what it does.

True. I was thinking that it's "easily readable markup", hence "not
explicit" ... which is of course wrong.

ReST has explicit markup, all the way through it (otherwise, how could it
work?). However, it is markup which leaves the source in an easily readable
form, whilst still being sufficiently expressive and unambiguous.

Tim Delaney



From rdsteph at earthlink.net  Fri Apr 26 22:57:34 2002
From: rdsteph at earthlink.net (Ron Stephens)
Date: Sat, 27 Apr 2002 02:57:34 GMT
Subject: Python.NET?  Jython.NET?  Ni!
References: <mailman.1019871772.28751.python-list@python.org>
Message-ID: <3CC9DB7F.3010306@earthlink.net>

Hooray Mark! ;-)))))



From BPettersen at NAREX.com  Sat Apr 27 00:06:12 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 26 Apr 2002 22:06:12 -0600
Subject: Python as scripting plug-in for large C++ project
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192B98369@admin56.narex.com>

I think I've finally convinced our CTO that we should use Python as the
scripting component for a large C++ project with some .NET/ASP thrown in
for good measure :-)

Now come the questions...

 - We extensively (exclusively) use our own basic datatypes,
   e.g. NInteger which in addition to keeping a value also
   stores whether it is NULL (think databases). How much work
   is it going to be to expose these to embedded Python?
   
 - Our object/relational mapping tool (written in Python :-)
   outputs ~100KLocs of C++ that defines an object tree that 
   ends in timeseries data ("vectors") or records. The code
   is highly optimized and makes sure we keep data integrity. 
   Would it be simpler to use SWIG/SIP/boost to expose this, 
   or would it be just as easy to have the mapping tool 
   generate the Python glue directly.

 - Are there any shortcuts, pitfalls, best practices etc. that
   we should follow. If anyone has experiences they would like
   to share I'd be most grateful.

Needless to say, this is seen as a high risk proposition by management,
but they seem to have confidence in me <wink>, and after four years of
programming Python I have confidence that I can pull this off ;-)

If it matters, we're going to standardize on the 2.2(.1) branch of
Python, and probably stay with it for 12-18 months. If the project is
successful we'll eventually have to ship it to 
customer sites (although we would probably go onsite for 
installation). Does that pose any additional wrinkles?

Any pointers to documentation other than the extending and 
embedding chapters would be helpful, as would pointers to any OS
projects that have allready done something similar...

-- bjorn




From fgeiger at datec.at  Mon Apr  1 03:07:51 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Mon, 1 Apr 2002 10:07:51 +0200
Subject: ActivePython 2.2 released!
References: <mailman.1017384524.14832.python-list@python.org> <a81lb4$cqs$1@rex.ip-plus.net> <mailman.1017560090.31342.python-list@python.org>
Message-ID: <a894gm$jla$1@rex.ip-plus.net>

Good news, thank you!

Franz

"David Ascher" <DavidA at ActiveState.com> schrieb im Newsbeitrag
news:mailman.1017560090.31342.python-list at python.org...
> > For 2.1.2: Will there be an update option in the installer? I've
installed
> > many external Python modules, which I do not want to reinstall when
updating
> > from 2.1.1 to 2.1.2, if somehow possible.
>
> Hmm.  I'll have to check, but I think that you should be able to
> uninstall 2.1, install 2.1.2 at the same location, and things should
> "just work".
>
> --david ascher
>




From netzapper at magicstar.net  Fri Apr  5 15:17:24 2002
From: netzapper at magicstar.net (A. Jones)
Date: Fri, 05 Apr 2002 20:17:24 GMT
Subject: file.getline()/cout problem
References: <3cae0463.46737479@mammoth.usenet-access.com>
Message-ID: <3cae0615.47170764@mammoth.usenet-access.com>

On Fri, 05 Apr 2002 20:13:43 GMT, netzapper at magicstar.net (A. Jones)
wrote:

>In trying to understand how file access might work in my instance, I
>wrote a small test program with the expectation that it would print
>off 35 lines of arbitrarily length of a file.

Oops... I guess the loop is for 36 copies... but, it doesn't matter...
it only gets to about 19.5 lines.


Netzapper/Aubrey Jones


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From syver-en+usenet at online.no  Fri Apr 12 09:51:30 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 12 Apr 2002 13:51:30 GMT
Subject: A simpler super.
Message-ID: <u662x3u1j.fsf@online.no>

This works on 2.2.1 (as far as I have tested) when yourself is a class and when
yourself is an instance. Since I have limited experience with the
innards of Python, it would be nice if anyone could critique the
implementation.

def Super(yourself):
    if isinstance(yourself, type): # for classes
        return super(yourself, yourself)
    return super(yourself.__class__, yourself)

-- 

Vennlig hilsen 

Syver Enstad


From pyth at devel.trillke.net  Fri Apr 19 15:58:53 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 19 Apr 2002 21:58:53 +0200
Subject: PEP for an unrestricted __init__ ?!
In-Reply-To: <MPG.1729fa43d442b9779896b6@news.nwlink.com>
References: <A7B5C22F20EE3C4E95744E2F03F1887F5647@MAYOR.retzlaff.com> <mailman.1019182657.12773.python-list@python.org> <MPG.1729fa43d442b9779896b6@news.nwlink.com>
Message-ID: <20020419195853.GJ28515@devel.trillke>

On Fri, Apr 19, 2002 at 10:57:19AM -0700, Jeff Shannon wrote:
> In article <mailman.1019182657.12773.python-list at python.org>, 
> pyth at devel.trillke.net says...
> 
> > But __new__ or __init__ still restrict you to return instances
> > of their respective class? This does sound rather sane,
> > but i imagined some other uses for an unrestricted __init__ (or __new__)
> > such as making functions-objects a special case of classes:
> > 
> >    def func(args):
> >       "compute some stuff
> >       return result
> > and
> >    class func:
> >        def __init__(self, args):
> >           "compute some stuff"
> >           return result
> 
> Well, you can already accomplish something like what you are 
> asking for, by defining a __call__() method for the class.  That 
> lets class instances be used as callable objects -- using 
> function syntax on them.

right. but the version with __call__ makes functions and classes
still look quite different. (Please note that i am not complaining. 
python *does* provide powerful concepts for classes and functions.)

maybe it's just me but i often pass functions around. and sometimes
i don't want to set all parameters from one place but from different
places. E.g. by providing two out of four needed parameters
i get back a function which only takes two arguments (the missing ones).
Actually i have written a 15-line function wrapper (as a class :-)
that provides exactly this behaviour. The problem is that i *have* to write
func(allarguments) () for the case where i have all parameters
in once place. And i rather want this behaviour:
'if not enough arguments, return a function with the so-far available
arguments as presets. otherwise perform computation and return.'

Fortunately, after digging more into PEP 253 i am quite convinced
that guido's time machine really works: my proposed behaviour really
seems to be what the __new__ concept provides (at least in my quick
tests).

regards,

	holger




From pyth at devel.trillke.net  Tue Apr 23 02:57:10 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 23 Apr 2002 08:57:10 +0200
Subject: rlcompleter not calling __getattr__ on [ ]
In-Reply-To: <aa2u72$k0a$1@peabody.colorado.edu>
References: <b5ddd682.0204180902.2866a9c0@posting.google.com> <a9mv1o$7jh$1@peabody.colorado.edu> <b5ddd682.0204220919.32ce27d@posting.google.com> <aa1hvn$o9i$1@peabody.colorado.edu> <mailman.1019499942.19137.python-list@python.org> <aa2460$534$1@peabody.colorado.edu> <mailman.1019533244.22752.python-list@python.org> <aa2u72$k0a$1@peabody.colorado.edu>
Message-ID: <20020423065710.GK27395@devel.trillke>

On Tue, Apr 23, 2002 at 12:13:30AM -0600, Fernando P?rez wrote:
> holger krekel wrote:
> 
> > great. but i guess we should try to come up with a solution that
> > allows
> > 
> >   a=[[].<tab>
> > 
> > and similar cases to work. Also
> > 
> >   a=().<tab>
> > 
> > should be allowed, too. I'll check that tommorow. Additionally
> > i want to see if it is possible to


> Currently ipython seems to do these just fine, after removing the following 
> from the delimiters:
> 
> readline_remove_delims -/'"[]{}()

does 

  b=[]
  a=a-b.<tab>

work then? And 

  [[].<tab> 
  

really works, too?

	holger




From neal at metaslash.com  Sun Apr 21 11:27:53 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Sun, 21 Apr 2002 11:27:53 -0400
Subject: TKinter Button widget
References: <ced73313.0204200634.251529e2@posting.google.com> <slrnac349u.ea.usenet@fohnhab.minus1.de> <ced73313.0204210643.2d88ac82@posting.google.com>
Message-ID: <3CC2DA79.26698885@metaslash.com>

Peter Saffrey wrote:
> 
> Perhaps I should have said that I need this in an object-oriented
> context. It doesn't seem to work using an anonymous lambda function to
> a class method: example below.

That isn't a class method (2.2 terminology).  Here's a version which works.
You can do somethings a bit different in 2.2, but this should work back
to 1.5.2 at least.

Changes are in call to lambda and printNumber() takes self.

Neal
--
from Tkinter import *

class gui:
    def __init__(self):
        win = Frame()
        win.pack()
        Button(win, text='print 3', command=lambda self=self:self.printNumber(3)).pack()
        Button(win, text='print 5', command=lambda self=self:self.printNumber(5)).pack()
        win.mainloop()

    def printNumber(self, number):
        print number

if __name__ == '__main__':
    test = gui()


From jessica at agozgi.sc.us  Tue Apr 23 11:25:57 2002
From: jessica at agozgi.sc.us (Jessica Felix)
Date: Tue, 23 Apr 2002 15:25:57 GMT
Subject: Art and all that Jazz:  are you ugly, I mean, attacking through outer powders
Message-ID: <comp.lang.python.d42583a5@agozgi.sc.us>

Some durable cups are bizarre and other brave dusts are clean, but will Abu dream that?

"ART AND ALL THAT JAZZ" is a gallery on the Healdsburg Plaza
that incorporates the two passions of its owner, Jessica Felix -
art and jazz.

Jessica is an artist whose medium is jewelry. She creates what she calls
"spirited art," specializing in images from indigenous cultures around the world.

http://www.artandallthatjazz.com/index-2.shtml

119A Plaza Street
Healdsburg, CA, USA 95448
1-707-433-7900

Occasionally, Kenny never attempts until Frank smells the think cloud tamely.

They are sowing near sweet, under clean, with sick jugs.





From kermit at thefrog.org  Wed Apr 10 15:59:01 2002
From: kermit at thefrog.org (GuySmiley)
Date: Wed, 10 Apr 2002 19:59:01 GMT
Subject: Extension Question
References: <cab7ec6a.0204100140.6b141d16@posting.google.com> <c76ff6fc.0204100629.55b2b2@posting.google.com>
Message-ID: <9Q0t8.89331$w7.7457248@bin6.nnrp.aus1.giganews.com>

Hi John,

Thanks for the hierarchy explanation. That's the bit I was missing. After
reading the example module and the example type, I had the impression a
central method dictionary was necessary for all types. DOH!

Thanks again,
sbp

"John Machin" <sjmachin at lexicon.net> wrote in message
news:c76ff6fc.0204100629.55b2b2 at posting.google.com...
> ics1010 at yahoo.com (sbp) wrote in message
news:<cab7ec6a.0204100140.6b141d16 at posting.google.com>...
> > In an extension I'm writing, I have a main import module which works
> > fine. I'm a little confused about trying to include an object type in
> > the same dll. From the import module's init I call an init function
> > for the type which calls Py_InitModule with the name and method array
> > for the type,
>
> Whoops ... Py_InitModule should be called once, to initialise the
> *module*, specifying the name and the table of module-level
> functions/methods.
> For old-style types, one or more of those functions will be
> constructors for instances of the type(s) you have defined in your
> module. For the moment, we'll draw a veil over the new-style types
> introduced in version 2.2.
>
> For the standard methods, you fill in functions in the slots of the
> type's descriptor table. For the methods that you invent, you need a
> tp_gettattr slot pointing to a function like the following, where spam
> is your type.
>
> static struct PyMethodDef spam_methods[] = {
>    {"add_eggs", (PyCFunction)spam_add_eggs, METH_VARARGS,
>     spam_add_eggs_doc},
>    /* etc */
>    {NULL, NULL}         /* sentinel */
> };
>
> static PyObject *
> spam_getattr(self, name)
>    spamobject *self;
>    char *name;
> {
>    return Py_FindMethod(spam_methods, (PyObject *)self, name);
> }
> /* There are more modern ways of doing this but if you can find the
> docs
>    tell the world :-)
> */
>
> So what happens when you do this:
>  >>> import mymodule
>  >>> spobj = mymodule.spam_factory(some_args) # your constructor
>  >>> spobj.add_eggs(2)
>
> is that Python finds the type of spobj, looks in the type's tp_getattr
> slot, calls your spam_gettatr function looking for "add_eggs"; this
> calls Py_FindMethod which rummages in the spam_methods table, finds
> that "add_eggs" is implemented by the spam_add_aggs C function, does a
> little more magic ("binding the method to the instance"), and then
> calls spam_add_eggs ...
>
> > but when I try to call a method from this array, I get
> > an exception stating there is no attribute for the method name I'm
> > calling. I have functions for the various type protocols (i.e.
> > attributes) which work fine.
>
> It is not apparent to me exactly what you mean by "functions for the
> various type protocols (i.e. attributes)" nor how this could "work
> fine" when you don't seem to be able to invoke a constructor to create
> an object.
>
> What source(s) of information have you been using to get so far??? I
> don't have any book other than the venerable 1st edition (1996) of
> "Programming Python" by Mark Lutz but this technique is in that book
> (see example 14-2 stacktyp.c) and I can't see how/why it wouldn't be
> in later books that have a serious chapter on extending.
>
> I suggest that you find some practical working examples of extension
> modules and follow their lead -- try (1) the Vaults of Parnassus (2)
> files in the modules and objects directory of a pre-2.2 Python source
> distribution e.g. listobject.c and arraymodule.c (3)
> http://object-craft.com.au/projects/csv/ which is a well-written
> not-too-large module.
>
> HTH,
> John
>




From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 23:23:08 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 20:23:08 -0700
Subject: Does Python need a '>>>' operator?
References: <mailman.1018828228.8551.python-list@python.org> <3cba62f8@news.mhogaming.com> <slrnablsp5.f08.jon+usenet@snowy.squish.net>
Message-ID: <3cbb996d@news.mhogaming.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnablsp5.f08.jon+usenet at snowy.squish.net...
| In article <3cba62f8 at news.mhogaming.com>, Ken Peek wrote:
| > This should be fixed so that decimal display of longs uses the '-'
| > sign, but hex display does not (you would just get the number of hex
| > digits in their raw "twos complement" form that the long object's
| > hex display method "knows" there is...)
|
| So you think that:
|
|   long(hex(-1l), 16)
|
| should not equal -1 but, err, some random choice of 15, 255, 4095,
| 65535, etc?

No, since we are "unifying" the int and long type, it is none
of your business HOW they are stored-- so we do away with the
'int()' and 'long()' functions...

BTW-- since -1 can be represented by one byte, my proposal is
for 'hex()' to print this as '0xff'...

I happen to like doing this extension stuff a byte at a time--
others have suggested that a nibble representation may be
better, and in that case, hex(-1) would print as '0xf'...

If we are using the "nibble at a time" method, the positive
number 15 (as in: hex(15)), would print as: '0x0f', because
we need another hex digit to convey the sign.  This could be
shifted left 3 more times without having to add another hex
digit to preserve the sign bit...

Now, if we have a 32 bit machine, and I want to bit-twiddle a
160-bit number (lets say for my new and improved random number
generator or something), and I need to right shift it LOGICALLY
(shifting a zero into bit 159), we would need a '>>>' operator
to do that (because the '>>' operator will shift right
ARITHMETICALLY, causing the sign bit to be copied into bit
159).  Remember, the machine may be STORING the number in some
bazaar format that has nothing to do with a binary twos
complement number, but that is of no matter to us-- as we are
always presented with a binary twos complement number as the
'interface' to the number.  The class that owns the object
would have a '>>>' method, and when we use the '>>>' operator,
that method would be called.  HOW it actually performs this
work for us is really none of our business.

See?





From physics at icecreamhasnobones.org  Fri Apr 19 16:02:46 2002
From: physics at icecreamhasnobones.org (n/a)
Date: 19 Apr 2002 13:02:46 -0700
Subject: What webhosting
References: <mailman.1019071656.4705.python-list@python.org>
Message-ID: <dbc49da.0204191202.464fb1c6@posting.google.com>

"A" <printers at sendme.cz> wrote in message news:<mailman.1019071656.4705.python-list at python.org>...
> Hi,
> Does anybody know about a webhosting provider ( with Python and 
> Telnet access) that does not remove a user's account if the user 
> sends occasionally some unsolicited emails?
> Thanks for reply
> Ladislav

Is ssh access ok instead of telnet ?
We will have to charge you a premium based on the potential work
load increase for dealing with complaints, is that ok ?
Also, you will be required to deliver thru our servers.
We can not protect your site if you send the emails from some other ip space.


From bernie at 3captus.com  Tue Apr  2 05:51:40 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Tue, 02 Apr 2002 10:51:40 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <3CA98D81.E515EBB@3captus.com>

Hi Guido,

Guido van Rossum wrote:
 
>     1) Should this PEP be accepted at all.
> 

Thought it was decided.  Since it is not.  I cast my vote as +1.

When I first saw the subject on python-dev, I don't like it, due to
backward compatabilty issue.  I change my mind, after reading the PEP
and most of the discussion.  The PEP actually keep the impact to
minimal.

I believe a computer language do need a boolean datatype.  It is needed
because some people see a half glass of water as half full and other see
it as half empty.  0 and 1 is not precise enough.  Acutally, I was
surprised to find that Python has None and Mapping type, but not Bool,
when I first pick up the language.  BTW, I vote for 'bool' as class
name.

>     2) Should str(True) return "True" or "1": "1" might reduce
>        backwards compatibility problems, but looks strange to me.
>        (repr(True) would always return "True".) 

str(True) should return 'True' because 2 is ture too.  But I can live
with '1' as well.

>     3) Should the constants be called 'True' and 'False'
>        (corresponding to None) or 'true' and 'false' (as in C++, Java
>        and C99).

'True' and 'False' for consistency.
 
>     Most other details of the proposal are pretty much forced by the
>     backwards compatibility requirement; e.g. True == 1 and
>     True+1 == 2 must hold, else reams of existing code would break.
> 

Don't like True + 1 == 2, but we are living in an imperfect world.

>     Minor additional issues:
> 
>     4) Should we strive to eliminate non-Boolean operations on bools
>        in the future, through suitable warnings, so that e.g. True+1
>        would eventually (e.g. in Python 3000 be illegal).  Personally,
>        I think we shouldn't; 28+isleap(y) seems totally reasonable to
>        me.
> 

I think we should if in Python 3000, x == False returns true when x is
false.

>     5) Should operator.truth(x) return an int or a bool.  Tim Peters
>        believes it should return an int because it's been documented
>        as such.  I think it should return a bool; most other standard
>        predicates (e.g. issubtype()) have also been documented as
>        returning 0 or 1, and it's obvious that we want to change those
>        to return a bool.
>

I prefer operator.truth(x) returns bool.
  

Bernie

-- 
In ancient England a person could not have sex unless you had consent of 
the King (unless you were in the Royal Family). When anyone wanted to
have a 
baby, they got consent of the King, the King gave them a placard that
they 
hung on their door while they were having sex. The placard had 
F.*.*.*. (Fornication Under Consent of the King) on it. 

Now you know where that came from.


From amuys at shortech.com.au  Thu Apr 25 19:25:28 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 25 Apr 2002 16:25:28 -0700
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com> <3CC6ED95.569DA16A@metaslash.com> <mailman.1019682346.19715.python-list@python.org> <aa7be1$i3v$1@panix1.panix.com>
Message-ID: <7934d084.0204251525.7eea458f@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<aa7be1$i3v$1 at panix1.panix.com>...
> In article <mailman.1019682346.19715.python-list at python.org>,
> holger krekel  <pyth at devel.trillke.net> wrote:
> >
> >I just don't happen to see the advantages of bringing a database into
> >the picture. It seems like a classical batch job and it 'random access
> >many times' is not needed, so why?
> 
> From the original post:
> 
>     Things would afterwards get more complicated cause I will have to
>     pullout ID's from "sub_file1", remove duplicate ID's create
>     "no_dup_sub_file1", match those to ID's in remaining 3 main files and
>     pullout data linked with those ID's.
> 
> This screams "*JOIN*" to me.  Now, if sub_file1 is less than 100MB,
> *maybe* Python can handle it.  IMO, that is true IIF the records are
> strictly fixed-length.  But IME requirements will change such that joins
> over larger and larger datasets will be needed -- and why re-invent a
> database that's designed precisely for this purpose?

It screams dbm to me, however given the nature of software to expand
it's requirements I agree.  MySQL is definately the approach I would
personally be taking.  It will almost certainly take less time to
learn basic SQL (and for this type of job, you don't need anything
fancy) then to write and debug what is really a mini-database.

I suggest you take a look at 

http://www.mysql.org 
 and 
http://www.dcs.napier.ac.uk/~andrew/sql/

Andrae Muys

P.S. Actually I prefer postgresql, but as in this case speed is likely
to be more useful then referential integrity, I favour MySQL.  If
however the database might end up being used in it's own right, I
would be more then willing to sacrifice the speed advantage of MySQL
for the referential, constraint, and transaction features of
Postgresql.


From niemeyer at conectiva.com  Sun Apr  7 05:47:06 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Sun, 7 Apr 2002 06:47:06 -0300
Subject: Stability and change
In-Reply-To: <3CAF18F8.56821C0B@python.org>
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org> <uDDr8.66496$S52.2342405@news2.tin.it> <just-89BA7A.17235906042002@news1.xs4all.nl> <3CAF18F8.56821C0B@python.org>
Message-ID: <20020407064706.C1600@ibook.distro.conectiva>

> Agreed.  I've moved this to python-dev; see you all there.

Bye.

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From aleax at aleax.it  Fri Apr  5 10:03:56 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 05 Apr 2002 15:03:56 GMT
Subject: newbie: declare function before definition?
References: <Pine.LNX.4.44.0204050953040.17812-100000@kramer.thekramers.net>
Message-ID: <w1jr8.52760$pT1.1488632@news1.tin.it>

David Kramer wrote:

> Disclaimer: 16 years programming experience, 1 week experience with
> Python.
> 
> It appears to me that in Python, functions need to be defined before they
> are called (physically in the file, I mean).  This means any piddly little

No, only before in the sense of TIME.  You need to execute a def statement
for the function before you execute a call to the function, in the clock
sense of 'before'.

> utility functions have to be before my main routine, and the reader has to
> wade through them before getting to the main routine.

Not at all, as long as the 'main routine' IS a routine, i.e., a function,
not just toplevel code of course.


> In some other languages, one can declare a function, and define it later,
> and the compiler is happy.  Is there a way to do this in Python?

Sure!  Make your 'main routine' into a routine called main, def it where
you want, call it at the end.


e.g.:

def main():
    print "I'm gonna call a function just for fun"
    function()
    print "there, told ya"

def function()
    print "I'm a tiny little function..."

main()


Alex



From gtaylor at lowebroadway.com  Mon Apr 22 05:06:32 2002
From: gtaylor at lowebroadway.com (Garry Taylor)
Date: 22 Apr 2002 02:06:32 -0700
Subject: Python, PyQt make install problems on IRIX
Message-ID: <f0fd5987.0204220106.7673e345@posting.google.com>

I have a very odd problem with QT 3, and PyQt 3.1, when I complete
make installing PyQT, my Python/QT programs run fine, but after a
restart, they do not. I can make install again, and then everything
will work again, but clearly I do not want to make install PyQt every
time I start the machine. The only thing I can think of is environment
variables, such as $PATH, but I have replicated the $PATH variables
from after the make install but no avail.

I think the worst thing about this problem is the error message, which
is simply 'Killed' and it does not dump core. I never had any such
problems on QT 2.x, and I don't have this problem on my Linux box. The
C++ demos with QT run fine, but no Python program which uses QT does,
so the problem appears to be with either SIP or PyQT, both of which I
have reinstalled.

BTW, I'm using an SGI O2, with IRIX 6.5, I used to have older versions
of PyQt, SIP, and QT installed, so I guess maybe the install of newer
versions did not fully overwrite the older ones, but that does not
really explain why it works after a make install, but not a restart,
because should not move around files, the only thing which I think
should change is env variables.

Has anyone else had this problem, or have any clues as to why it is
happening?

T a lot IA

Garry


From geoff at gerrietts.net  Wed Apr 10 17:15:55 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 10 Apr 2002 14:15:55 -0700
Subject: Indenting lines 400..457 help
In-Reply-To: <a929tk$ik2$1@pcls4.std.com>
References: <a929tk$ik2$1@pcls4.std.com>
Message-ID: <20020410211555.GC27359@isis.gerrietts.net>

Quoting Ned Harvey (rahvee at yaahhoo.com):
> For the life of me, I can't find any easy way to indent (or unindent) lines
> numbered 400..457 or whatever.  Does anyone have any good suggestions?
> 
> I'm usually a vi user, but I'll happily learn a command of some other
> format.  I've tried
>     :400,457s/(.*)/   \1/g
> and a ton of variations of the same idea, but it seems that the
> search/replace doesn't do what I expect it to.

In vim, the easiest way is to hit V, scroll down a few lines, then hit
the > key to shift everything right one indent level. If the indent
level is too large, :set shiftwidth=4. I also recommend :set
tabstop=8, :set softtabstop=4, :set expandtab.

In oldschool vi, try:
:400
57>>

Another common trick I use is:

:400,457s/^/    /

Best of luck,
--G.

-- 
Geoff Gerrietts             "Overspecialize and you breed in weakness.
<geoff at gerrietts net>     It's slow death." -- Maj. Motoko Kusanagi
http://www.gerrietts.net                          (Ghost in the Shell)




From gh_pythonlist at gmx.de  Fri Apr 12 19:45:40 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 13 Apr 2002 01:45:40 +0200
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
In-Reply-To: <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com>
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca> <slrnabeqkg.mfc.huaiyu@gauss.almadan.ibm.com>
Message-ID: <20020412234540.GC15144@lilith.my-fqdn.de>

* Huaiyu Zhu <huaiyu at gauss.almadan.ibm.com> [2002-04-12 23:15 +0000]:
> BTW, Python is OSS but not FS.

Stop it. Please don't spread that nonsense any further.

Quoting http://www.fsf.org/licenses/license-list.html

    """
    The License of Python 1.6a2 and earlier versions.  This is a free
    software license and is compatible with the GNU GPL.  Please note,
    however, that newer versions of Python are under other licenses (see
    below).

    The License of Python 2.0.1, 2.1.1, and newer versions.  This is a
    free software license and is compatible with the GNU GPL.  Please
    note, however, that intermediate versions of Python (1.6b1, through
    2.0 and 2.1) are under a different license (see below).
    [...]

    The License of Python 1.6b1 and later versions, through 2.0 and 2.1.
    This is a free software license but is incompatible with the GNU
    GPL.  The primary incompatibility is that this Python license is
    governed by the laws of the State of Virginia, in the USA, and the
    GPL does not permit this.
    """

Guido van Rossum invested *a lot* of time to clarify this issue with the
FSF. The issue was that they *thought* the Python 1.6 -> 2.0 licenses
were incompatible with the GPL because of the reason mentioned above.
Thank $DEITY this "problem" is solved now.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 9.6 ?C      Wind: 0.8 m/s




From mtaylor at lowebroadway.com  Fri Apr 19 08:00:45 2002
From: mtaylor at lowebroadway.com (Moray Taylor)
Date: 19 Apr 2002 05:00:45 -0700
Subject: Coloring/Styling QListViewItems with PyQT
Message-ID: <d906b2c1.0204190400.588f637f@posting.google.com>

Does anyone have any simple example code of how to color or style
(make text bold  or whatever) a QListViewItem with PyQT, I tried a
couple of times using QPainter, but only managed to color the whole
listview, any tips?

Thanks

Moray Taylor


From andreas at andreas-jung.com  Sun Apr 21 22:02:42 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Sun, 21 Apr 2002 22:02:42 -0400
Subject: word stemming / conflation
References: <3CC2C23B.DAF9057D@anansispaceworks.com> <003d01c1e9a1$7f7512f0$02010a0a@suxlap>
Message-ID: <004701c1e9a1$e87ff170$02010a0a@suxlap>

Sorry, that was the wrong URL. Take this one:

http://sourceforge.net/projects/pystemmer/

Andreas

----- Original Message -----
From: "Andreas Jung" <andreas at andreas-jung.com>
To: "Terry Hancock" <hancock at anansispaceworks.com>; <python-list at python.org>
Sent: Sunday, April 21, 2002 22:00
Subject: Re: word stemming / conflation


> I have written  wrappers for the snowball stemmers of
> Dr. Porter: http://pystemmer.sourceforge.net.
>
> Andreas
>
>
> ----- Original Message -----
> From: "Terry Hancock" <hancock at anansispaceworks.com>
> To: <python-list at python.org>
> Sent: Sunday, April 21, 2002 09:44
> Subject: word stemming / conflation
>
>
> > I was working on an algorithm to condense strings of
> > text into mnemonic labels (hopefully this will produce
> > more meaningful labels than simply assigning a number!).
> > I got the idea from the way later versions of
> > latex2html generate filenames.
> >
> > Anyway, after tinkering with this for awhile I've
> > discovered that it can be a bottomless pit! A little
> > web research reveals various MS / PhD papers for doing
> > this task (which apparently is called "conflation" as
> > well as "word stemming") in English and lots of other
> > languages. Looks like it's a lot harder than you
> > might at first think. I didn't find much in Python,
> > though.
> >
> > So, I'm wondering if there are some Python resources
> > (or better-yet already written modules) for doing
> > this sort of thing.  I don't really need a completely
> > robust system -- the occasional error will be quite
> > tolerable. So I'm more interested in quick and simple
> > solutions that work most of the time, rather than
> > the ultra-robust library-science type.
> >
> > Thanks for any information or leads you might be
> > able to suggest,
> >
> > Terry
> >
> > --
> > ------------------------------------------------------
> > Terry Hancock
> > hancock at anansispaceworks.com
> > ------------------------------------------------------
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> >
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>





From hughett at mercur.uphs.upenn.edu  Tue Apr 30 16:01:29 2002
From: hughett at mercur.uphs.upenn.edu (Paul Hughett)
Date: 30 Apr 2002 20:01:29 GMT
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu> <3kjz8.32779$8D3.928578@news1.tin.it> <aam88o$oec$1@netnews.upenn.edu> <g8yz8.81675$vF6.2428847@news2.tin.it>
Message-ID: <aamt6p$dfa$1@netnews.upenn.edu>

Alex Martelli <aleax at aleax.it> wrote:

:> Alex Martelli <aleax at aleax.it> wrote:

:>>> r=Numeric.array(range(13))

:>>> r[-3::]
: array([10, 11, 12])
:>>> r[-3::-1]
: array([10,  9,  8,  7,  6,  5,  4,  3,  2,  1,  0])
:>>>

Hmm.  Is the value of r[-1:-1:-1] equal to array([12, 11, 10, 9, 8, 7,
6, 5, 4, 3, 2, 1, 0]) or to an empty array?  It could plausibly be
either, depending on whether the middle -1 is interpreted as a lower
limit or as 1 from the end.

In any case, I will be adopting the Numeric convention.  As you point
out, having multiple conventions is definitely a bad idea.


Paul Hughett


From sholden at holdenweb.com  Wed Apr 24 09:32:48 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 24 Apr 2002 09:32:48 -0400
Subject: variable substitution
References: <e2e9e807.0204240519.688cf3cb@posting.google.com>
Message-ID: <kwyx8.807$z55.780@atlpnn01.usenetserver.com>

"Hugh" <h.e.w.frater at cs.cf.ac.uk> wrote ...
> How do I get the %(variable)s thing to work. I need to substitute a
> variable into a print statement so I can send links with sessionId's
> attached back to the browser. I'm trying this at the moment, and It
> says invalid sysntax.
> print '<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
> Type</a>'
> I saw this in a document written by Guido that I found here:
> http://www.w3j.com/6/s3.vanrossum.html
>

You need to add a "%" operator and put a dictionary on its RHS. The
parenthesised words are then used to look up the appropriate values in the
dictionary (technically, any mapping type will work). e.g.:

    print '''<a href = "addItemType.py?sessionID=%(sessionId)s">Add Item
Type</a>''' % {'sessionId': 'XXX1234'}

Of course, it's more useful of the dictionary is reusable, or if you are
substituting more than one value. Otherwise you could just use the shorter
forms, e.g.:

    print '''<a href = "addItemType.py?sessionID=%s">Add Item Type</a>''' %
'XXX1234'

Note that if there is more than one value to substitute, you have to use a
tuple on the RHS, as in

    print '''<a href = "addItemType.py?sessionID=%s">Add %s Item Type</a>'''
% ('XXX1234', 'My')

A few experiments with the interactive interpreter should set you straight.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From eppstein at ics.uci.edu  Sat Apr  6 21:32:21 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sat, 06 Apr 2002 18:32:21 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it> <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com>
Message-ID: <eppstein-8A20F8.18322106042002@news.service.uci.edu>

In article <uavbgcabvl642b at news.supernews.com>,
 "John Roth" <johnroth at ameritech.net> wrote:

> Here I thought that teaching about bools was quite easy.
> Bools answer a specific question: is it x? True or False.
> They do not answer the question: which X is it?
> 
> In other words, bools select one possible case out of
> however many there are; they are not a substitute for a
> universe which has only two cases.

By the way, this is conceptually very similar to ints (at least, 
nonnegative ints): ints answer a specific question: how many things?  
They do not answer the question: which things are they.

Taken to the illogical extreme, Laura's position (that we should use 
ints instead of bools so we can always answer which x it is) would also 
mean that we should use lists instead of ints so we can always answer 
which things they are.

Specifically, in her favorite example of isLeapYear, if she thinks it 
should return a count of leap days instead of a bool, maybe it would be 
even more appropriate to return a list of leap days.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From logiplexsoftware at earthlink.net  Wed Apr 17 14:16:28 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 17 Apr 2002 11:16:28 -0700
Subject: Trinary operator?
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE>
	<NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com>
Message-ID: <20020417111628.1efd1e2d.logiplexsoftware@earthlink.net>

On Wed, 17 Apr 2002 13:02:56 -0500
Mark McEahern wrote:

> > Not that it's a big deal since there's many ways to do that, but can I use
> > something like
> >     verboseGender = (gender == 'm') ? 'male' : 'female'
> > in Python?
> 
> People seem to use and/or for that:
> 
> >>> gender = 'm'
> >>> verbose = (gender == 'm' and 'male' or 'female')
> >>> verbose
> 'male'
> >>> gender = 'f'
> >>> verbose = (gender == 'm' and 'male' or 'female')
> >>> verbose
> 'female'

Or it might be preferable to use a dictionary:
 
>>> gender = 'f'
>>> verbose = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
>>> verbose
'female'
>>> gender = 'm'
>>> verbose = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
>>> verbose
'male'
>>> gender = 'x' 
>>> verbose = {'m': 'male', 'f': 'female'}.get(gender, 'unknown')
>>> verbose
'unknown'

IMHO, it's no less readable and a bit more flexible.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From rjones at ekit-inc.com  Thu Apr 18 00:37:26 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 18 Apr 2002 14:37:26 +1000
Subject: Hostname Resolution not working in Explorer
In-Reply-To: <3CBE4DAA.E710D618@engcorp.com>
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net> <3CBE4DAA.E710D618@engcorp.com>
Message-ID: <200204181437.26855.rjones@ekit-inc.com>

On Thu, 18 Apr 2002 14:38, Peter Hansen wrote:
> Gary Spivey wrote:
> > I have windows XP Home, and IE6. I have a zope server running on my box.
> > The server is running fine, and I can access if it I access
> >
> > http://127.0.0.1:8080/mange
> >
> > However, if I use
> >
> > http://localhost/mange
> >
> > the page isn't found. The same holds true if I use my hostname rather
> > than localhost - also, the actual IP address works.

This is a common problem with Windows - it doesn't know how to resolve 
"localhost" in some situations. Usually it coincides with you changing the 
network configuration (unplugging a network cable or something).

Repeat after me, Windows is dumb.


    Richard





From sakkinen at sakkinen.titu.jyu.fi  Tue Apr  9 02:44:06 2002
From: sakkinen at sakkinen.titu.jyu.fi (Markku Sakkinen)
Date: Tue, 9 Apr 2002 06:44:06 +0000 (UTC)
Subject: Inheritance workshop
References: <a7uhaj$8bd$2@mordred.cc.jyu.fi>
Message-ID: <a8u2jm$at4$1@mordred.cc.jyu.fi>

In comp.object Markku Sakkinen <sakkinen at cs.jyu.fi> wrote:

> Second and last call for contributions:
> 		The Inheritance Workshop at ECOOP 2002
> 		======================================
> ...

Well, this did not remain the last call after all:

Due to several requests and the recent problems of the ECOOP 2002
website, the deadline for submission of position papers has been
extended by one week.

  New deadline: Monday, April 15, 2002

Full information is available at the workshop web site:
  http://www.cs.auc.dk/~eernst/inhws/


From alexandre.herbiniere at caramail.com  Wed Apr 17 11:59:20 2002
From: alexandre.herbiniere at caramail.com (alexandre)
Date: 17 Apr 2002 08:59:20 -0700
Subject: need help to choose an appropriate database
Message-ID: <fac97694.0204170759.35b66092@posting.google.com>

I use zope to access to a database, but I have to choose it.
I mustn t store a lot of data so should i choose MySQL, PostGreSQL or
anther one ?
Which one is the most efficient ( speed...)


From kkto at csis.hku.hk  Mon Apr 29 02:42:36 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 29 Apr 2002 14:42:36 +0800
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <7iit6broo3.fsf@enark.csis.hku.hk>

>>>>> "Stephen" == Stephen J Turnbull <stephen at xemacs.org> writes:

    Stephen> The consumer protection argument is the OSI line (a special
    Stephen> case thereof), and rms don't have much truck with anything so
    Stephen> economical and relativist.

If you think it that way, fire up Emacs, and type C-h C-p.  It clearly tells
you otherwise.  In particular, RMS acknowledge that at some time the
government should do something like copyright, giving a limited monopoly in
order to give economic incentive to work.  He just argued that software
development is not an area in which this is a correct approach---because, in
order to do that, the general public have to give up too much.

So I would characterize OSI and free software by their point of attack.  The
OSS group mainly attacks from the side of software producers, trying to
convince them that open-source produces better software for smaller cost.
Free software community attack it from the users' side, telling consumers to
understand that they are supposed to have the right to see whatever inside
the program, and that it is a very important right that they should not sign
anything to give it up.  The only catch here is that in the eyes of free
software advocates, developers are themselves users of their software and
software produced by others.  The battle line of free software is simply
longer than that of OSS.

Currently, the OSS point of attack seems more effective.  But I like the
free software line of attack much better.  Users must not be at the mercy of
producers.

Regards,
Isaac.



From skip at pobox.com  Wed Apr 24 15:16:28 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 24 Apr 2002 14:16:28 -0500
Subject: Software Engineers Wanted - 2 positions - Boca Raton, FL
In-Reply-To: <MPG.17307c2b9af25858989688@news.supernews.com>
References: <MPG.17307c2b9af25858989688@news.supernews.com>
Message-ID: <15559.1164.465561.689559@12-248-41-177.client.attbi.com>

    Phillip> Hi all.  The position described below is open at Verio for two
    Phillip> software engineers in Boca Raton, Florida, USA.  

Phillip (and anyone else with Python-related jobs to announce),

If you haven't already done so, bop on over to

    http://www.python.org/Jobs.html

read the blurb at the start, scan the listings, and send jobs at python.org a
position announcement with the necessary info.  One of the python.org
webgnomes will get it up on the site within a day or so.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From lord_elendil at hotmail.com  Thu Apr 18 03:37:38 2002
From: lord_elendil at hotmail.com (Bryce C. Mason)
Date: Thu, 18 Apr 2002 07:37:38 GMT
Subject: My First .py
Message-ID: <6Juv8.134255$QB1.35129169@typhoon3.we.ipsvc.net>

Hello Python World!

I am happy to announce that my first python program is a success!  What I
was considering writing in C++ was made far easier in python.  I
successfully stripped screen output captured to a text file of important
components and put them in a better order.  My project leader will be
thrilled that he doesn't have to import the text file into word, make it
neat and line up, print out hundreds of pages, sort them, and then choose
which he wants.  The program does everything but choose which pieces to
keep.

I think I'll go get a book on how to let python retrieve information over
the net.  This would remove another step of him having to capture screen
output.  Anyone have a good suggestion for some examples of how python can
manipulate network resources?

Bryce Mason




From rpyron at alum.mit.edu  Sat Apr 13 00:19:43 2002
From: rpyron at alum.mit.edu (Robert Pyron)
Date: Sat, 13 Apr 2002 04:19:43 GMT
Subject: Tools/Scripts/texi2html.py
References: <g4t3bugaplcn104b10ip18nghfer4uia89@4ax.com>
Message-ID: <rccfbu0gbdfrq41866shbmc6jt9pj2a364@4ax.com>

On Mon, 08 Apr 2002 20:40:09 GMT, Robert Pyron <rpyron at alum.mit.edu>
wrote:

>I have made some changes to Tools/Scripts/texi2html.py that I think
>might be useful for others...
>
>I wanted to produce a Windows HTMLHelp (.CHM) file for MySQL.
>To that end, I made the following modifications:
>
>1. BUGFIX: In function makefile(), strip blanks from the nodename. 
>   This is necesary to match the behavior of parser.makeref() and 
>   parser.do_node().
>
>2. I modified the original texi2html code to add support for
>the following items:
>     @uref       URL reference
>     @image      image file reference
>     @multitable output an HTML table
>     accents     partial support for accents, to match MAKEINFO output
>
>3. I added a new command-line option, '-hh basename', to specify
>HTML Help output. This will cause three files to be created
>in the current directory:
>      `basename`.hhp  HTML Help Workshop project file
>      `basename`.hhc  Contents file for the project
>      `basename`.hhk  Index file for the project
>When fed into HTML Help Workshop, the resulting file will be
>named `basename`.chm.
>
>4. A new class, HTMLHelp, to accomplish item 3.
>
>My version of texi2html.py can be found at:
>http://www.dullandboring.com/pystuff/index.htm
>
>Bob Pyron

Update, 4/12/2002
A few more changes: 

BUGFIX fixed KeyError in end_ifset (well, I may have just made it go
away, rather than fix it) 

BUGFIX allow @menu and menu items inside @ifset or @ifclear 

add support for @vtable 

BUGFIX changed the way I am storing and outputting node information
for htmlhelp 

The program now handles documentation for MySQL (my original goal),
TexInfo, GCC, and glibc. 

By the way, I would be pleased and honored if my modifications were
to be included in the standard distribution. How do I submit them
for consideration?

Bob Pyron




From stojek at part-gmbh.de  Tue Apr 16 15:26:09 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Tue, 16 Apr 2002 19:26:09 GMT
Subject: C-extension not threadable?
Message-ID: <3cbc783d.39583390@news.easynews.net>

Hi,

I am using a C-extension in my code that works just fine. Now I have
added threading to keep the UI alive, and a callback function to
notify the UI when the C-function is finished. The thread is started
like this:


T=threading.Thread(group=None,target=ExtModule.CFunction,name=None,
                           args=(a,b,c,self.callback))
        T.start()

#-----------
    def callback(self,dummy):
        wxPostEvent(self,ResultEvent(1,'ready  %s'%(time.clock())))
        return 1


The C-extension does what it is supposed to do, the callback function
works, but the C-function does not start as a parallel thread but is
freezing my UI. Anything basic I miss here ?

Thanks
Marcus


From ChrisE at lantech.com  Tue Apr 23 10:54:42 2002
From: ChrisE at lantech.com (Christopher Encapera)
Date: Tue, 23 Apr 2002 10:54:42 -0400
Subject: changes to win server registry with python script not appeari
	ng in gui tools
Message-ID: <BC74F8F7E87DD511990C000103E2FEFA2FEB@XCHG>

You do mean regedit and regedt32 - or does gui tools refer to something
else...?



Christopher Encapera
Lantech.Com
chrise at lantech.com
502-267-4200


-----Original Message-----
From: speeves at unt.edu [mailto:speeves at unt.edu]
Sent: Monday, April 22, 2002 8:56 PM
To: python-list at python.org
Subject: changes to win server registry with python script not appearing
in gui tools


Hi!

The title says it all...  Why don't the changes that I am making to
the registry appear in the gui tools?  I have completely automated
many of the gui-functions with a simple python script, but any query
to the registry must now come from the python scripts, and not the gui
tools, as they are never current anymore :(
(Even after rebooting)

Just wondering...  (Seems like a fluke in the win os to me, but...)

speeves
central Web Support
UNT
-- 
http://mail.python.org/mailman/listinfo/python-list




From gashton at cmedltd.com  Thu Apr  4 09:20:44 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 04 Apr 2002 15:20:44 +0100
Subject: CGIHTTPServer and custom headers
Message-ID: <1017930044.10710.7.camel@tp6>

I tried to use the CGIHTTPServer module to implement a simple server for
running a simple CGI that I'm developing. I thought "Great, this can be
a nice stand alone little test environment, without the need for
apache".

It worked fine until I wanted to send the browser a temporary redirect
header (i.e. 302 response code), or indeed anything other than a 200
response code.

In other words, the CGIHTTPRequestHandler class has this little nugget
in it, irrespective of what headers the script is actually sending back
to the browser:

  self.send_response(200, "Script output follows")

Unsurprisingly, this means the browser ignores my redirects.

Now I know that the CGIHTTPServer module isn't supposed to be used for
production use, and I know I could patch it myself if it was important
enough to me, but right now what I'd like is a (fully) Python solution
that doesn't involve me touching apache.

Can anybody recommend the most straightforward approach? My app is dead
simple and doesn't need speed or scalability. I'm currently browsing the
twisted docs, and the asyncore docs. Is there an easier way?

Thanks (by the way, I'm on Linux).

-- 
Graham Ashton





From jacob at boris.cd.chalmers.se.cd.chalmers.se  Mon Apr  8 13:54:35 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 8 Apr 2002 17:54:35 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1018016611.6776.python-list@python.org> <a8nhfi$gbt$1@nyheter.chalmers.se> <3CAF524F.63E46C81@alcyone.com>
Message-ID: <a8slgr$pin$1@nyheter.chalmers.se>

In article <3CAF524F.63E46C81 at alcyone.com>,
Erik Max Francis  <max at alcyone.com> wrote:
>Jacob Hallen wrote:
>
>> Python is a language that is weakly typed, ...
>
>You mean dynamically typed, not weakly typed.

I have been brought up in an environment which called anything
allowing mixed types in a list weakly typed.

If you prefer to call it dynamically typed, I don't mind.

Jacob Hall?n

-- 


From pj at engr.sgi.com  Thu Apr  4 23:53:32 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 5 Apr 2002 04:53:32 GMT
Subject: "Don't Kill Opera" documentation release
References: <mailman.1017709803.3793.python-list@python.org> <a8j443$omncm$1@fido.engr.sgi.com> <mailman.1017978132.7545.python-list@python.org>
Message-ID: <a8jakc$oo7d9$1@fido.engr.sgi.com>

Skip asked:
|>     Paul> Oh - Opera support informed me in an email today that the
|>     Paul> next Beta should be out any day.
|> 
|> For what platform?


For Linux, I'd guess, since the email discussion involved
my report of a Bug in Beta 1 of Opera 6 for Linux.

The exact phrase was "planned release in a few days".
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From emile at fenx.com  Thu Apr 25 22:17:28 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 26 Apr 2002 02:17:28 GMT
Subject: Numeric question
References: <pan.2002.04.25.21.32.03.310021.2578@hooper.net>
Message-ID: <YM2y8.71958$HH5.5067@rwcrnsc51.ops.asp.att.net>

Cameron Hooper
> The error produced is:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "/usr/tmp/python-25603Vd", line 21, in ?
>     print outerproduct(x,t)
>   File "/usr/lib/python2.1/site-packages/Numeric/Numeric.py", line
359,
>      in outerproduct return
asarray(a).flat[:,NewAxis]*asarray(b).flat[NewAxis,:]
> ValueError: flattened indexing only available for contiguous array
>

This tells you the problem.  Something is not contiguous.

>>> t.iscontiguous()
1
>>> x.iscontiguous()
0

OK.  It's x.  So let's build a contiguous x and try it again:


>>> from Numeric import *
>>>
>>> inputs = array( [[1., 1.],
...                  [1., 0.],
...                  [0., 1.],
...                  [0.,0.]] )
>>>
>>>
>>> binput = ones((4,1), Float)
>>> inputs = concatenate( (inputs, binput), 1 )
>>>
>>> x = array(inputs[0][:])
>>> t = array([1.])
>>>
>>> # this step fails
... print outerproduct(x,t)
[[ 1.]
 [ 1.]
 [ 1.]]

Not anymore... ;-)

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From borcis at geneva-link.ch  Wed Apr 10 06:33:13 2002
From: borcis at geneva-link.ch (Boris Borcic)
Date: Wed, 10 Apr 2002 12:33:13 +0200
Subject: Something or Nothing^H^H^H^H^H^H^HEmpty
References: <3cb3fbc0$0$78808$edfadb0f@dspool01.news.tele.dk> <3CB427D3.7020404@geneva-link.ch> <3cb432ca$0$231$edfadb0f@dspool01.news.tele.dk> <a91itn$2fgp$1@news.nextra.cz>
Message-ID: <3CB414E9.BE1C1288@geneva-link.ch>

Petr Prikryl wrote:
> 
[...]
>   Operations and built-in functions that have a
>   boolean result always return 0 for false and 1
>   for true, unless otherwise stated. (Important
>   exception: the boolean operations "or" and
>   "and" always return one of their operands.)
> 
> In other words, it says that in the *boolean
> context* the named cases are *converted* to
> *false* or *true* values respectively.  So, there
> is nothing like...

IMO this means everywhere a boolean role is already given
by definition to a function result, 0 or 1 is returned,
as a matter of current design style. IOW the Pep's booleans
already existed, only they were implicit.

To me the shortcut behavior of "or" and "and" and the
way this allows them to return their operands is the
clincher against your theory.

> ::
> 
>   > {} is a dictionary-shaped nothing. [] is a
>   > list-shaped nothing. 0 is an integer-shaped
>   > nothing. 0.0 is a float shaped nothing.
> 
> Again, it only obscures the things.  The students
> should have asked "What do you mean by the
> *shaped nothing*?".  (At least for me, it is far
> more obscure than the simple notion of yes/no or
> true/false.)

Blind spot are shared simultaneously with visions,
that's why arguing from clarity/obscurity is a subtle
matter.

> Compare it with what the documentation already
> says (in different wording):

I don't agree your translation is perfectly faithful.

Regards, Boris


From sholden at holdenweb.com  Sat Apr 20 12:08:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 20 Apr 2002 12:08:31 -0400
Subject: Substitution with Hash Values
References: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com> <mailman.1019254077.7316.python-list@python.org> <slrnac1lbh.nma.jon+usenet@snowy.squish.net> <mailman.1019303438.1339.python-list@python.org>
Message-ID: <oqgw8.26239$T%5.21037@atlpnn01.usenetserver.com>

"holger krekel" <pyth at devel.trillke.net> wrote ...
> On Sat, Apr 20, 2002 at 02:41:53AM -0000, Jon Ribbens wrote:
> > In article <mailman.1019254077.7316.python-list at python.org>, holger
krekel wrote:
> > > re.sub('\{\{(.*?)\}\}',lambda x: dict[x.group(1)], str)
> >
> > Need a raw string there, i.e.
> >
> >   r"\{\{(.*?)\}\}"
>
> yes, i am sometimes confused by 'raw strings' and 'strings'.
> Could you help by saying why exactly it is needed here?
>
In raw string literals the backspaces are treated almost the same as other
characters, making it easier to generate strings whose values *contain*
backspaces, often needed in regular expressions.

>>> print "this is\na string"
this is
a string
>>> print r"this is\na string"
this is\na string

regards
 Steve






From sholden at holdenweb.com  Mon Apr 22 07:56:39 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 07:56:39 -0400
Subject: writing a body using email module
References: <508cb8ae.0204220123.493c35be@posting.google.com>
Message-ID: <wWSw8.72199$T%5.62931@atlpnn01.usenetserver.com>

"B" <martin_byrne2 at yahoo.co.uk> wrote in message
news:508cb8ae.0204220123.493c35be at posting.google.com...
> Hi,
>
> I've recently been looking at the relativly new email module. I've
> been able to create the email with to,from,subject and attachments but
> I cant for the life of me figure out how to write the body of the
> mail.
>
> Can anyone help me out with this?
>

Looks (from the docs, I've never used this library) like you need to call
the add_payload() method with your message string as the argument.

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From logiplexsoftware at earthlink.net  Fri Apr 26 17:41:47 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 26 Apr 2002 14:41:47 -0700
Subject: String handling bug in Python
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E20192215165@admin56.narex.com>
Message-ID: <20020426144147.6c004b45.logiplexsoftware@earthlink.net>

On Fri, 26 Apr 2002 15:23:49 -0600
Bjorn Pettersen wrote:

> > From: Tim Peters [mailto:tim.one at comcast.net] 
> > 
> > [Stephen Ferg]
> > > ...
> > > I would like to see Python support TRULY raw string 
> > literals -- string 
> > > literals that are not escaped in any way.  Perhaps they 
> > could be added 
> > > with a prefix of z (for zero escaping?), so that
> > >
> > >      z"\" == "\\"
> > >      z"\'" == "\\\'"
> > 
> > The first one can't be spelled at all with a raw string 
> > today.  The second one can be, and indeed the same way you 
> > spelled it, as r"\'".  What do you do in your scheme if you 
> > want a z-string to contain a single double-quote? 
> 
> I always thought that could be spelled r'"' or r'''"''', but I guess
> there would still be a problem if you wanted all of ", ', ''', and """
> in a string. I would think that would occur less frequently than \ at
> the end of a string though (especially with windows path names). I feel
> I must be missing something?

It seems if you imposed the limitation that the z-string not contain the quote
character used to delimit the string then you could simply use concatenation to
get around that limitation:

z"'" z'"' z'\'


-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From gustafl at algonet.se  Mon Apr 22 18:07:24 2002
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Mon, 22 Apr 2002 22:07:24 +0000 (UTC)
Subject: Pyxie, anyone?
References: <Xns91F7F276A63DAgustafl@195.100.94.182> <3CC36FD2.E98469F0@engcorp.com>
Message-ID: <Xns91F917C12553gustafl@195.100.94.182>

Peter Hansen <peter at engcorp.com> wrote:

> And the Pyxie library itself.

Oh, I read too fast. Thanks. :-)

Gustaf


From h.e.w.frater at cs.cf.ac.uk  Sat Apr 20 21:49:19 2002
From: h.e.w.frater at cs.cf.ac.uk (Hugh)
Date: 20 Apr 2002 18:49:19 -0700
Subject: dco2.so problem
Message-ID: <e2e9e807.0204201749.62f3b8c6@posting.google.com>

I can import DCOracle2 now that I've put dco2.so, DCOracle2.py and
__init__.py in my cgi-bin directory (probably not the right thing to
do, but it imports OK).
I now have another problem. When trying to do dbh =
DCOracle2.connect(connectstring) I get the following error in the
apache error log:

Traceback (most recent call last):
  File "/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/addItemType.py",
l
ine 12, in ?
    dbh = DCOracle2.connect('scm1hewf/*******@cs2000')
  File "/usr/local/apache/htdocs/cgi-bin/userbin/H.E.W.Frater/DCOracle2.py",
lin
e 93, in connect
    db = dco2.connect(u, p, dbname)
dco2.DatabaseError: (12154, 'Error while trying to retrieve text for
error ORA-1
2154')

Does anyone know what this means and how to fix it.

Hugh


From just at xs4all.nl  Mon Apr  8 11:18:35 2002
From: just at xs4all.nl (Just van Rossum)
Date: Mon, 08 Apr 2002 17:18:35 +0200
Subject: Carbonized embedded MacPython
References: <f69ccee4.0204080600.21d8ac87@posting.google.com>
Message-ID: <just-984F55.17183508042002@news1.xs4all.nl>

You're much better off asking such specialized questions on the 
MacPython mailing list:
   http://www.python.org/sigs/pythonmac-sig/

Just


In article <f69ccee4.0204080600.21d8ac87 at posting.google.com>,
 mattmsykes at yahoo.co.uk (Matt M Sykes) wrote:

> Hi,
> 
> I would like to embed Python into an already-existing Carbon
> application.
> 
> The application is quite large, so removing/replacing all the Carbon
> API calls would be infeasible at this point due to time constraints.
> That is, I must use Carbonized Python libs.
> 
> I downloaded and installed MacPython22full.bin which contains some
> Carbon libs.  I soon found that I didn't have any header files.
> 
> On to the source.  There are two places to obtain sources: python.org
> and the MacPython homepage, http://www.cwi.nl/~jack/macpython.html.
> Note ftp://ftp.cwi.nl/pub/jack/python/mac/GUSI2-for-MacPython22.sit is
> a broken link at the MacPython homepage.
> 
> The source at the MacPython homepage does not have a configure script
> nor a project file; how do I build it?  It does have *some* prj files
> but none for building python.  Hoping for the best, I tried using
> those header files to compile an embedded-python test program with my
> Carbon libs.  My machine promptly crashed when I ran it.  To me this
> means the header files don't match the libs I have.  The exact same
> code runs perfectly on Windows and FreeBSD.
> 
> Next: sources from the ftp site mentioned at python.org:
> ftp://ftp.python.org/pub/python/2.2.
> 
> This source appears to require the unix-type build environment on OS
> X; the instructions are to run the configure script and compile with
> the unix-type (GNU?) compiler using make.
> 
> From what I understand, I must use the same compiler (same run-time
> environment) to build both python and my target application if I wish
> to embed python in the target application.
> 
> For political and practical reasons beyond my control, the target
> application must use the Metrowerks compiler (I am not in charge of
> the Mac builds, I'm just helping with the embedded python side of it).
> 
> There is a mwerks directory in the sources, but the README there is
> clearly out of date as none of the files it mentions exist.  There is
> one project file called 'frozen.prj' which appears to be out of date
> also.  I am not convinced this was the source used to build the Carbon
> MacPython at the MacPython website.  I believe Jack did some magic
> which is not included here.
> 
> All I really need (I think) are the exact header files which were used
> to build the Carbon libs which are available at the MacPython website,
> as well as info on the precise build environment and settings.  The
> good news is that it says the Carbon libs were built with the
> Metrowerks compiler.
> 
> Is there anyone who has embedded Python into a Carbon app?  Is there
> anyone who is able to build Carbon-MacPython from sources (using the
> Metrowerks compiler)?
> 
> Any help would be appreciated.
> 
> --Matt


From s.ernst at web.de  Thu Apr  4 09:31:05 2002
From: s.ernst at web.de (Sascha Ernst)
Date: Thu, 4 Apr 2002 16:31:05 +0200
Subject: how to convert an integer to a string?
Message-ID: <a8ho5l$ubt$00$2@news.t-online.com>

thanx, sascha





From jb at cascade-sys.com  Tue Apr 30 18:38:14 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 15:38:14 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no> <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>  <mailman.1020033011.19748.python-list@python.org> <slrnactoru.pv.mlh@vier.idi.ntnu.no> <mailman.1020196975.26619.python-list@python.org> <fyEz8.4763$Au5.3848@atlpnn01.usenetserver.com>
Message-ID: <3CCF1CD6.BF418F27@cascade-sys.com>

Steve Holden wrote:

> > I realized just AFTER pressing send.  A little less coffee that day and I
> > may have spared myself that minor embarrassment.
> >
>
> Heh, heh, heh.

On the aforementioned and possibly several other matters.  ;o)

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From rjones at ekit-inc.com  Wed Apr  3 17:34:20 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 4 Apr 2002 08:34:20 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <hyAq8.29178$n4.5840455@newsc.telia.net>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <mailman.1017795008.700.python-list@python.org> <hyAq8.29178$n4.5840455@newsc.telia.net>
Message-ID: <200204032234.WAA29573@crown.off.ekorp.com>

On Wed, 3 Apr 2002 20:10, Fredrik Lundh wrote:
> Richard Jones wrote
>
> > Again, have you _read_ the ReST documentation, specifically the
> > discussion of  alternatives? It's extensive, and very well written.
>
> do you realize that the ReST documentation is much
> longer than the complete Ruby language reference?

I believe I was referring to a specific section of the documentation, which 
is doesn't really compare to the Ruby language ref. Perhaps if you compared 
it to a document comparing Ruby to other langauages, you might have an 
argument.


> I'd rather learn a new programming language than yet
> another markup language.  especially when the latter
> sucks.

What's the value in this statement? Nowhere does it state that it is 
mandatory that python programmers write their docstrings in ReST. Only that 
there will be tools available for those that do.

And what the hell's with "especially when the latter sucks"? Really...


    Richard




From mcfletch at rogers.com  Tue Apr 30 19:55:48 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 30 Apr 2002 19:55:48 -0400
Subject: Bulding arrays from text file data.
References: <28bbdc67.0204301518.5e7c41c8@posting.google.com>
Message-ID: <3CCF2F04.8010601@rogers.com>

Okay, off the top:

	string.split( string ) instead of re.split('\s+') should be significantly 
faster

	your string.strip( ) usage is causing almost the whole string to be 
copied if there is any whitespace at beginning/end.  Might be better 
(test it) to simply delete null strings from beginning/end of list if 
they exist. i.e.

		if strings and not strings[0]: del strings[0]
		if strings and not strings[-1]: del strings[-1]

	if you can process the file by lines, using xreadlines might help if the 
lines are sufficiently long and you're really running out of memory. 
trade off is lots of Python code getting run.

	minor note: shape can probably accept -1 as an argument saying "whatever 
fits in this dimension"
	
if none of those gets you in spitting range of acceptable, a scheme 
where you read in a large chunk (a few tens of megabytes of text), 
accumulate those results in a list, add the tail to the next few 
megabytes, then process that chunk (rinse, repeat), would likely be the 
next step in optimising.

HTH,
Mike
	

Joe Woodward wrote:
> Does anyone know of a faster way to build an array out of text file
> with formatted data. What I have been doing works well, but now my
> files and 200MB and up. The following way builds a Numeric.array
> without knowing the size to start with. I can then resize it after the
> fact.
> 
> 
> yldpos=Numeric.array(map(float,re.split('\s+',string.strip(open('datafile.txt').read()))))
> 
> yldpos.shape=(len(yldpos)/5,5)
> 
> 
> Thank you,
> 
> Joe Woodward 
> PHoenix Analysis & Design Technologies


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From gh_pythonlist at gmx.de  Mon Apr  1 23:59:19 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 2 Apr 2002 06:59:19 +0200
Subject: Extending Python on Windows
In-Reply-To: <3CA90BE1.9010703@cfl.rr.com>
References: <3CA90BE1.9010703@cfl.rr.com>
Message-ID: <20020402045919.GC1247@lilith.hqd-internal>

* Gregory Propf <gpropf at cfl.rr.com> [2002-04-01 20:39 -0500]:
> [trouble compiling extension modules]

Solution: distutils

A good starting point is Alex Martelli's recipe:

http://groups.google.com/groups?q=group:comp.lang.python+alex+distutils+recipe&hl=en&selm=9v57kr02q3i%40enews4.newsguy.com&rnum=1

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 6.5 ?C      Wind: 0.7 m/s



From patmiller at llnl.gov  Wed Apr 10 11:51:17 2002
From: patmiller at llnl.gov (Patrick Miller)
Date: Wed, 10 Apr 2002 08:51:17 -0700
Subject: A faster Python?, Python compiler, Dylan,...
Message-ID: <3CB45F75.9373D224@llnl.gov>

Chris writes:
> "Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> > I'm not sure that Pyrex will ever compile *all* valid Python
...
> It could be interesting to see how far you could get with just the subset
> you do compile, though. The Squeak people can compile their entire VM from
...
> So, you don't need to compile all of Python syntax to have a very
> interesting project on your hands.

With the SciPy weave.accelerate project, we're trying
two schemes to keep the whole language in play...

1) keep a pretty general reverse API (Python describing
a type's C/C++ API) which can include a generic PyObject*
type.  Then if w.a can't find an accelerated type description
it can still strip off some of the more expensive parts
of the interpretation (e.g. lower loop overhead, 
cut some of the function call cost, faster attribute
fetching, etc...)

2) Fall back to simply executing the original Python.  This
can mean that we can eventually write Python in Python
where some of the bits will compile down to C/C++ and
other more dynamic parts just don't.

My goal is to get down to a really slim compiled core that allows
much of the language to be implemented directly in Python.
This has big ramifications for Jython and Pippy.

Pat


-- 
Patrick Miller | (925) 423-0309 | patmiller at llnl.gov

Son, when you grow up you will know who I really am.
I am just a child like you who has been forced to act
responsibly.  -- Rod Byrnes




From marcelk at brick.bitpit.net  Thu Apr 11 16:30:01 2002
From: marcelk at brick.bitpit.net (Marcel van Kervinck)
Date: Thu, 11 Apr 2002 20:30:01 GMT
Subject: how to detect if a file is complete
References: <32f17666.0204040701.64dd0bfe@posting.google.com>
Message-ID: <63q49a.k6n2.ln@brick.bitpit.net>

harm <harm at laatje.dhs.org> wrote:
> Hi all,

> I want to copy some files from one dir to another using python. 
> The files are copied onto my linux box via a samba-share.

> I want to write a python script which checks a dir and copies all
> files to another dir. But there is the possibility that the file is
> not yet complete. (This are rather large Postscript files >100mb).
> Therefore I need to detect if the file is complete.

> The files are copied on the samba-share by Adobe Acrobat Distiller,
> therefore i cannot save some check file like $FILENAME.complete...

> Is there some other way (using python) to detect if the file is
> complete?

If you do not want a sloppy approach that involves
probing or running fuser, and if you control the writer,
then the default way is this: Copy the file under a dummy 
name first, and after completion rename it to its 
final name. Renaming is atomic, so by the time the
receiver sees the file, it will be complete.

	Marcel
--  _   _
  _| |_|_|
 |_    |_     marcelk at bitpit.net
         |_|  Marcel van Kervinck


From robinjim at earthlink.net  Sat Apr  6 10:45:54 2002
From: robinjim at earthlink.net (robin and jim)
Date: Sat, 06 Apr 2002 15:45:54 GMT
Subject: single-quoted string conversion to triple-quoted string
References: <GZqr8.20207$ml2.1517221@newsread1.prod.itd.earthlink.net> <%_Br8.20276$nt1.1661286@newsread2.prod.itd.earthlink.net> <yNCr8.179198$Yv2.61305@rwcrnsc54>
Message-ID: <SKEr8.20531$nt1.1685307@newsread2.prod.itd.earthlink.net>

Ahh very good!

One of my shortcomings as a Python programmer is failing to remember that
almost everything is accessible (e.g., the __doc__ attribute).  I'm much too
pre-conditioned, from other languages, to view such things as off-limits.

Thanks!

"Emile van Sebille" <emile at fenx.com> wrote in message
news:yNCr8.179198$Yv2.61305 at rwcrnsc54...
> "robin and jim" <robinjim at earthlink.net> wrote in message
> news:%_Br8.20276$nt1.1661286 at newsread2.prod.itd.earthlink.net...
> >
> > Gee, I thought it would be obvious why I was asking such a question
> <wink>!
> >
> > As unit tests run, using the unittest module shipped with Python 2.2,
> the
> > default behavior is to display a one-line description of each test.
> The
> > default implementation displays the first line of the test method's
> > **docstring**, if available; otherwise the name of the test method
> (e.g.,
> > "test_my_regular_expression") is displayed.
>
> Yes.  This comes from the __doc__ attribute of the method.
>
> >
> > Well, yours truly thought he would be clever and add some additional
> > information to this docstring as it became available at run time via
> the %
> > placeholder mechanism.  Alas, the result is a single-quoted string
> that
> > unittest does not recognize as a docstring so it displays the test
> method's
> > name (e.g., "test_my_regular_expression") instead. I, naively, think
> that if
> > the single quoted result can be re-converted to a triple-quoted
> string, it
> > will be recognized by unittest.
> >
>
> You could change the __doc__ attribute of the method before testing:
>
> import unittest
>
> class IntegerArithmenticTestCase(unittest.TestCase):
>     def testAdd(self):  ## test method names begin 'test*'
>         self.assertEquals((1 + 2), 3)
>         self.assertEquals(0 + 1, 1)
>     def testMultiply(self):
>         self.assertEquals((0 * 10), 0)
>         self.assertEquals((5 * 8), 40)
>
> if __name__ == '__main__':
>     testvals = [3,7,4,5]
>     IntegerArithmenticTestCase.__dict__['testAdd'].__doc__ = 'testing
> with %s' % testvals
>     unittest.main()
>
> HTH,
>
> --
>
> Emile van Sebille
> emile at fenx.com
>
> ---------
>
>




From mgilfix at eecs.tufts.edu  Wed Apr 17 02:13:39 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 17 Apr 2002 02:13:39 -0400
Subject: list_a-list_b
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEOHOOAA.tim.one@comcast.net>; from tim.one@comcast.net on Wed, Apr 17, 2002 at 01:54:56AM -0400
References: <eppstein-6EBA52.22290716042002@news.service.uci.edu> <LNBBLJKPBEHFEDALKOLCCEOHOOAA.tim.one@comcast.net>
Message-ID: <20020417021339.G16954@eecs.tufts.edu>

  I didn't make any promise of efficiency :) It was more of a way to
clarify the iteration someone else suggested. One nice thing about
using filter though and the implied iteration is that it becomes
much simpler to make comparisions of compound objects. I've become
a big fan of things like 'map' and 'filter', etc... I find they
often express a very elegant solution (not necessarily the most
efficient but since they are built-in functions - can they perform any
optimization?). Perhaps I just like functional environments. The sort
of comprehension primitives you liked from lisp but not using lisp.

                 -- Mike

On Wed, Apr 17 @ 01:54, Tim Peters wrote:
> [David Eppstein]
> > ...
> > Better is to convert b to a dictionary
> > (where are dict comprehensions when you need them):
> >
> >     dict_b = {}
> >     for x in list_b: dict_b[x] = 1
> 
> Looks clear enough to me, although
> 
>     for x in list_b:
>         dict_b[x] = 1
> 
> is clearer.  If you like, you can do this:
> 
>     dict([(x, 1) for x in list_b])
> 
> or this:
> 
>     dict(zip(list_b, [1]*len(list_b)))
> 
> or even:
> 
>     dict(zip(list_b, list_b))
> 
> since you don't really care what the dict's values are, just the keys.  I
> draw the line before:
> 
>     dict(zip(*[list_b]*2))
> 
> though.  Contrary to popular belief, there's no Fabulous Prize awaiting
> those who minimize either vertical or horizontal space <wink>.
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (tim.one)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 236 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020417/8fb03d14/attachment.sig>

From jbublitzNO at SPAMnwinternet.com  Mon Apr 29 10:09:44 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Mon, 29 Apr 2002 14:09:44 GMT
Subject: KDE+GNOME Bindings in SuSE 7.3
References: <3CCC818C.2070505@gmx.net> <3CCCD547.6090405@SPAMnwinternet.com> <3CCD3885.7020301@gmx.net>
Message-ID: <3CCD543B.7040202@SPAMnwinternet.com>

Tom Chance wrote:
> OK I did that and found bizarrely that the "print sys.path" statement 
> printed out directories in the /usr/local tree, whilst I'd been working 
> with python in the /usr tree, so I had somewhere along the line got two 
> installs?! No problemo though, I symlinked 
> /usr/local/lib/python/site-packages to /usr/lib/python/site-packages and 
> it imported the gnome modules without a hitch :)

Give a guy some help and he uses your competitor's product :)
PyGTK is very well done and I'd use it if I wasn't stuck on Qt/KDE.

> As for KDE, I didn't realise you needed to download those extra 
> packages! From a tutorial I'd read from the python site about oding 
> Python with KDE it made it seem like the KDEBindings package 
> automatically set-up all the python bindings, including modules, for 
> you. Oh well. I'm going to wait until either PyKDE2-3.1 comes out (it's 
> on 3.0 atm), or better still until the packages come out for KDE3.0 final.

Probably best to wait for the next release at this point.

> Thanks for the help! :)

I'm glad it worked out.

Jim




From maxm at mxm.dk  Tue Apr 23 16:15:36 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 23 Apr 2002 22:15:36 +0200
Subject: dictionary method question
References: <v6ix8.8822$d7.2764674@newssrv26.news.prodigy.com>
Message-ID: <3CC5C0E8.2050105@mxm.dk>

dsavitsk wrote:


> for some dictionary d is there any relationship between the returns from
> d.keys() and d.values()?  that is, to get them in some arbitrary but
> matching order, do i need to loop over d.items()

Actually it is guaranteed in the docs that if you call keys() and 
values(), without modifying the dict in-between, that they will have the 
same/corresponding order.

So:

keys = aDict.keys()
values = aDict.values()
items = [(keys[i], values[i]) in range(len(keys))]


is the same as:

items = aDict.items()

regards Max M



From aleax at aleax.it  Thu Apr 11 13:34:12 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 17:34:12 GMT
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
Message-ID: <oOjt8.8815$vF6.254710@news2.tin.it>

Studying PyQt (latest releases of everything), I started noticing 
segmentation faults as I quit my application and eventually whittled it 
down back to...:

#!/usr/bin/env python2.2

import sys
import qt
# from qt import *

class MyWidget(qt.QWidget):

    def __init__(self, parent=None, name=None):
        qt.QWidget.__init__(self, parent, name)

        self.quit = quit = qt.QPushButton("Quit", self, "quit")
        self.connect(quit, qt.SIGNAL("clicked()"), qt.qApp, 
qt.SLOT("quit()"))


a = qt.QApplication(sys.argv)

w = MyWidget()
a.setMainWidget(w)
w.show()
a.exec_loop()


With the latest releases of Qt / sip / PyQt and Python, this gives me a 
Segmentation Fault (Mandrake 8.1, btw) as I hit the Quit button.  No
fault if I kill the app with the cross in the upper right corner (KDE).

BUT -- if I uncomment the "from qt import *" (of which I shouldn't
be using any part, as I loathe said usage)... the Segmentation
Fault goes away!

I'm a bit puzzled -- there must be something strange in the order of
finalization going on.  I can't believe I have to "import *" to make
my applications terminate gracefully -- perhaps just one or two
things from the qt module might suffice?

If anybody has more insight or can suggest more graceful workarounds,
please let me know!


Alex



From juenglin at informatik.uni-freiburg.de  Mon Apr 22 15:51:45 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 22 Apr 2002 12:51:45 -0700
Subject: __iadd__ and fellows missing (Python 2.2)
References: <xdupu153vao.fsf@leto.informatik.uni-freiburg.de> <lkr8lkj4bz.fsf@pc150.maths.bris.ac.uk> <xdulmbr51x3.fsf@leto.informatik.uni-freiburg.de> <VqXt8.28409$%8.2434287@bin2.nnrp.aus1.giganews.com>
Message-ID: <efb63dac.0204221151.6d866042@posting.google.com>

"Terry Reedy" <tejarex at yahoo.com> wrote in message news:<VqXt8.28409$%8.2434287 at bin2.nnrp.aus1.giganews.com>...
> "Ralf Juengling" <juenglin at informatik.uni-freiburg.de> wrote in
> message news:xdulmbr51x3.fsf at leto.informatik.uni-freiburg.de...
> > I understand now, that 'i+=1' is no real in-place operation
> 
> Nor is it a false in-place operation :<)
> 
> > but  just a shortcut for 'i=i+1' since int is immutable.
> 
> Correct.  This is true for all immutable types.
> 
> > For the sake of clarity, an 'in-place' operation should really
> > work in-place, don't you think?
> 
> Yes, but +=, etc, are not in-place operations, and are not advertised
> as such.

As I found out today, these operations are advertised as in-place
operations in PEP 203 (see below). 
The discussion on whether in-place operations applied to immutables
should raise a type error or not already has taken place, I guess?

Cheers,
Ralf

[Cut out from PEP 203]

      The proposed patch that adds augmented assignment to Python
      introduces the following new operators:
      
         += -= *= /= %= **= <<= >>= &= ^= |=
      
      They implement the same operator as their normal binary form,
      except that the operation is done `in-place' when the left-hand
      side object supports it, and that the left-hand side is only
      evaluated once.
      
      They truly behave as augmented assignment, in that they perform
      all of the normal load and store operations, in addition to the
      binary operation they are intended to do. So, given the expression:
      
         x += y
      
      The object `x' is loaded, then `y' is added to it, and the
      resulting object is stored back in the original place. The precise
      action performed on the two arguments depends on the type of `x',
      and possibly of `y'.

      The idea behind augmented assignment in Python is that it isn't
      just an easier way to write the common practice of storing the
      result of a binary operation in its left-hand operand, but also a
      way for the left-hand operand in question to know that it should
      operate `on itself', rather than creating a modified copy of
      itself.


From max at alcyone.com  Wed Apr 17 21:29:01 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 17 Apr 2002 18:29:01 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com> <a9gbhj$dhm$1@panix1.panix.com> <slrnabs3mm.1kfm.Gareth.McCaughan@g.local>
Message-ID: <3CBE215D.F146ED46@alcyone.com>

Gareth McCaughan wrote:

> (These days there ought to be a computer scientist in the
> mix too. Suggestions?)

"I'll just write a program to exhaustively find them all.  What's a
prime number again?"

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From emile at fenx.com  Wed Apr 24 10:47:10 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 24 Apr 2002 14:47:10 GMT
Subject: Window problem in windows!
References: <mailman.1019653396.7431.python-list@python.org>
Message-ID: <Ozzx8.64548$HH5.4797@rwcrnsc51.ops.asp.att.net>

Stian Jenssen
> ---code---
>
>     command = "pause \r\n"
>     command += "pause"
>     print os.system(command)
>
> ----code--
>
>
> It pops up 2 window instead of 1, in a windows enviroment. The code
works
> greate on *nix box, what seems to be the problem.


Well, then it averages out properly.  Here it opens no window! ;-)

Which really just means to say that we'll need more info on how it is
you're trying to do this, or a complete example the exhibits the
problem.

--

Emile van Sebille
emile at fenx.com

---------



From bvdpoel at uniserve.com  Sat Apr 13 15:19:35 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Sat, 13 Apr 2002 12:19:35 -0700
Subject: Need customised sort
Message-ID: <3CB884C7.EE7E04D@uniserve.com>

Yeah, I thought this would be easy...

I have a list like this:

	[ [0, x, [0, y, [4, n] ...]

and I want to sort it a bit. The sort should only work on the first
element AND, retain the order of list elements if 2 adjacent elements
have identical first elements. So, in the above example, if x > y I
still want the list to be sorted with the x first (as passed).

Hope this makes sense and someone has a quick and dirty solution.

Oh, maybe this will help explain... I'm doing some midi work and the
list represents a list of midi events. The first element is an offset
into the track (not a delta midi) and the second is the midi event. If I
just use a builtin sort() my '0' events tend to move around, depending
on the 'event'....and this can cause problems. Example, I might want to
do a program change and a note on event at 0. I and to my list in the
right order, but I don't want the sort to move the note event in front
of the program change (mind you, I'm not sure if matters for the midi
sequencer...). The reason for the sort in the first place is that some
note events might get placed in the wrong order (a note off event for,
say 1999, might get placed after a note on for 1888).

Thanks.



-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From abf at fly.srk.fer.hr  Thu Apr 11 09:09:53 2002
From: abf at fly.srk.fer.hr (Ante Bagaric)
Date: 11 Apr 2002 13:09:53 GMT
Subject: A trivial question about print
References: <a93oqk$jf6$1@bagan.srce.hr> <fCet8.7130$vF6.214745@news2.tin.it>
Message-ID: <a941v1$1dr$1@bagan.srce.hr>

Alex Martelli <aleax at aleax.it> wrote:
>> how to make that cursed blank character dissappear?
>> Other than print "a=%d" % a because it wouldnt work for iteration of
>> prints..
>> 
>> for i in range(5):
>>     print i,
>> 
>> gives 0, 1, 2, 3, 4 and I want 0,1,2,3,4

> Actually this gives 0 1 2 3 4 -- I don't know where you got the commas
> in the output that you seem to have observed, maybe your print statement
> was not quite as you show it here.

um, yep... I don't know where I got those commas myself :)
I wasn't copying it from the actual output, I was producing it right out
of my memory and I guess I was a little too quick. No commas there.
But the issue was that cursed blank char anyway... :)



> The print statement sets the softspace attribute on sys.stdout when it
> has just printed with a trailing comma, resets it when it emits a line
> break.  If it finds the attribute set when about to continue emitting,
> it outputs a space first.  The following superdirty trick works as
> expected:

>>>> import sys
>>>> for i in range(5):
> ...   print setattr(sys.stdout,'softspace',0) or ("%s,"%i),
> ...
> 0,1,2,3,4,
>>>>

[...]

> but you need no trick at all to use the write method in the
> perfectly normal way within a script or module.


> Alex

Wow, pretty exhaustive answer. :)


Yet, I was thinking maybe there was something you could set in advance, to
disable printing the space; so that you could use the same, simple, print
syntax.

BTW, i never wrote a single line in Python but it looks darn attractive from
what I have read so far. Just that this print behaviour brings some
reminiscence to Fortran output formatting where some thing just happen and
there's nothing you can do about it :)
Sure in most cases this feature helps, you dont need zillion " " strings
between your variables :), but wouldnt it be nice if there was python-simple
way to just get around it?

(Not that all of this is oh-so-important, of course)

Ante


From programisci at NOSPAM.murator.com.pl  Thu Apr 25 09:00:48 2002
From: programisci at NOSPAM.murator.com.pl (Johann)
Date: Thu, 25 Apr 2002 15:00:48 +0200
Subject: sql but... without server
Message-ID: <7fvfcu0vkc610fma8545v0nhdm7c6hem7q@4ax.com>

I  would like to use sql database with python but I do not want to
install any sql server. 

I would like to use sql engine as a standalone file (e.g. dll). 

It is because I wanted to prepare a standalone application which run
from CD without any servers or network connection. 

Do anone have any experience with similar task?

--
Johann


From ods at fep.ru  Wed Apr 10 08:58:50 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Wed, 10 Apr 2002 16:58:50 +0400 (MSD)
Subject: Display current line number of running script
In-Reply-To: <Xns91EC58418DA12mdkmdkmdkmdk@130.133.1.4>
Message-ID: <Pine.LNX.4.44L.0204101658160.13708-100000@ods.pp.ru>

On 10 Apr 2002, mdk wrote:

m> Is there a function that would return the current line number
m> that the
m> running script is on?

E.g., inspect.getlineno(sys._getframe())





From hgs at dmu.ac.uk  Thu Apr 18 14:18:22 2002
From: hgs at dmu.ac.uk (Hugh Sasse Staff Elec Eng)
Date: Thu, 18 Apr 2002 19:18:22 +0100 (BST)
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <j4u1q955fd.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.GSO.4.40.0204181847040.16636-100000@neelix>

On 18 Apr 2002, Martin v. [iso-8859-1] L?wis wrote:

> So the property that is being reversed is 'build as static'.
>
> > So the fastest fix is for me to doengrade binutils?
>
> No. The fastest fix is to apply the patch below, and rerun configure.
>
> Regards,
> Martin
>
OK here is the result of building with the patch.

Script started on Thu Apr 18 18:23:55 2002
this is .kshrc
neelix hgs Python-2.2.1 322 $ gmake test
        [...]
building '_socket' extension
skipping /home/hgs/Python-2.2.1/Modules/socketmodule.c (build/temp.solaris-2.7-sun4u-2.2/socketmodule.o up-to-date)
gcc -shared build/temp.solaris-2.7-sun4u-2.2/socketmodule.o -L/usr/local/ssl/lib -L/usr/local/lib -lssl -lcrypto -o build/lib.solaris-2.7-sun4u-2.2/_socket.so
skipping 'dbm' extension (up-to-date)
building 'gdbm' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/gdbmmodule.c -o build/temp.solaris-2.7-sun4u-2.2/gdbmmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/gdbmmodule.o -L/usr/local/lib -lgdbm -o build/lib.solaris-2.7-sun4u-2.2/gdbm.so
WARNING: removing "gdbm" since importing it failed

    That is expected.
        [...]
building 'fpectl' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/fpectlmodule.c -o build/temp.solaris-2.7-sun4u-2.2/fpectlmodule.o
gcc -shared build/temp.solaris-2.7-sun4u-2.2/fpectlmodule.o -L/usr/local/lib -o build/lib.solaris-2.7-sun4u-2.2/fpectl.so
WARNING: removing "fpectl" since importing it failed

    Also expected

building '_tkinter' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -DWITH_APPINIT=1 -I/usr/openwin/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/_tkinter.c -o build/temp.solaris-2.7-sun4u-2.2/_tkinter.o
In file included from /usr/local/include/tk.h:83,
                 from /home/hgs/Python-2.2.1/Modules/_tkinter.c:49:

        [lots of warnings about X11 trimmed]

gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -DWITH_APPINIT=1 -I/usr/openwin/include -I. -I/home/hgs/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /home/hgs/Python-2.2.1/Modules/tkappinit.c -o build/temp.solaris-2.7-sun4u-2.2/tkappinit.o
In file included from /usr/local/include/tk.h:83,
                 from /home/hgs/Python-2.2.1/Modules/tkappinit.c:16:

        [lots more warnigns about X11 trimmed]

gcc -shared build/temp.solaris-2.7-sun4u-2.2/_tkinter.o build/temp.solaris-2.7-sun4u-2.2/tkappinit.o -L/usr/openwin/lib -L/usr/local/lib -ltk8.3 -ltcl8.3 -lX11 -o build/lib.solaris-2.7-sun4u-2.2/_tkinter.so
WARNING: removing "_tkinter" since importing it failed

    I wonder why this failed.  Explored below

running build_scripts
not copying /home/hgs/Python-2.2.1/Tools/scripts/pydoc (up-to-date)
./python -E -c 'import sys ; from distutils.util import get_platform ; print get_platform()+"-"+sys.version[0:3]' >platform
find ./Lib -name '*.py[co]' -print | xargs rm -f
./python -E -tt ./Lib/test/regrtest.py -l
test test_al skipped -- No module named al
test test_bsddb skipped -- No module named bsddb
test test_cd skipped -- No module named cd
test test_cl skipped -- No module named cl
test test_curses skipped -- Use of the `curses' resource not enabled
test test_dl skipped -- No module named dl
test test_gdbm skipped -- No module named gdbm
test test_gl skipped -- No module named gl
test test_imgfile skipped -- No module named imgfile
test test_largefile crashed -- exceptions.IOError: [Errno 27] File too large

    How much space does it want?  I don't know this one anyway.

test test_linuxaudiodev skipped -- No module named linuxaudiodev
test test_minidom skipped -- No module named pyexpat
test test_nis skipped -- No NIS master found for any map

    I didn't expect that.  We use nisplus....

test test_ntpath skipped -- No module named nt
test test_openpty skipped -- No openpty() available.
test test_pyexpat skipped -- No module named pyexpat
test test_sax skipped -- no XML parsers available
test test_socket_ssl skipped -- Use of the `network' resource not enabled

    What network resouce?

test test_socketserver skipped -- Use of the `network' resource not enabled
test test_sunaudiodev skipped -- no audio device found!

    Its a headless server, so I'd expect that.

test test_winreg skipped -- No module named _winreg
test test_winsound skipped -- No module named winsound
165 tests OK.
1 test failed:
    test_largefile
21 tests skipped:
    test_al test_bsddb test_cd test_cl test_curses test_dl test_gdbm
    test_gl test_imgfile test_linuxaudiodev test_minidom test_nis
    test_ntpath test_openpty test_pyexpat test_sax test_socket_ssl
    test_socketserver test_sunaudiodev test_winreg test_winsound
Ask someone to teach regrtest.py about which tests are
expected to get skipped on sunos5.
gmake: [test] Error 1 (ignored)
./python -E -tt ./Lib/test/regrtest.py -l

    [ Same messages again]

165 tests OK.
1 test failed:
    test_largefile
21 tests skipped:
    test_al test_bsddb test_cd test_cl test_curses test_dl test_gdbm
    test_gl test_imgfile test_linuxaudiodev test_minidom test_nis
    test_ntpath test_openpty test_pyexpat test_sax test_socket_ssl
    test_socketserver test_sunaudiodev test_winreg test_winsound
Ask someone to teach regrtest.py about which tests are
expected to get skipped on sunos5.
gmake: *** [test] Error 1
neelix hgs Python-2.2.1 323 $ exit

script done on Thu Apr 18 18:46:27 2002

So why did tkinter fail:

Script started on Thu 18 Apr 2002 19:06:22 BST
neelix hgs 11 %> ./python
Python 2.2.1 (#2, Apr 18 2002, 18:23:04) [C] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import tkine ter
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named tkinter
>>>
neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Thu 18 Apr 2002 19:07:09 BST

Hmm...

Script started on Thu 18 Apr 2002 19:11:03 BST
neelix hgs 11 %> ./try_tkinter.sh
#!/bin/sh -v
gcc -v -shared build/temp.solaris-2.7-sun4u-2.2/_tkinter.o build/temp.solaris-2.7-sun4u-2.2/tkappinit.o -L/usr/openwin/lib -L/usr/local/lib -ltk8.3 -ltcl8.3 -lX11 -o build/lib.solaris-2.7-sun4u-2.2/_tkinter.so
gcc: build/temp.solaris-2.7-sun4u-2.2/_tkinter.o: No such file or directory
gcc: build/temp.solaris-2.7-sun4u-2.2/tkappinit.o: No such file or directory
Reading specs from /usr/local/lib/gcc-lib/sparc-sun-solaris2.7/2.95.3/specs
gcc version 2.95.3 20010315 (release)

neelix hgs 12 %> exit
neelix hgs 13 %>
script done on Thu 18 Apr 2002 19:12:24 BST


No, I can't see why it never built that.

        Hugh





From gerhard at bigfoot.de  Thu Apr 25 00:20:58 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 25 Apr 2002 06:20:58 +0200
Subject: SHA-based encryption function in Python
In-Reply-To: <7x7kmwjx5h.fsf@ruckus.brouhaha.com>
References: <7x3cxlk7qb.fsf@ruckus.brouhaha.com> <B8EB9FE2.EF8C%richard@electrophobia.com> <7xpu0pij41.fsf@ruckus.brouhaha.com> <B8EBC8E9.EFA0%richard@electrophobia.com> <7xsn5l6zn7.fsf@ruckus.brouhaha.com> <B8EC0013.EFCE%richard@electrophobia.com> <7xadrsslye.fsf@ruckus.brouhaha.com> <B8ECB435.F916%richard@electrophobia.com> <7x7kmwjx5h.fsf@ruckus.brouhaha.com>
Message-ID: <20020425042057.GA960@lilith.my-fqdn.de>

* Paul Rubin <phr-n2002a at nightsong.com> [2002-04-24 20:05 -0700]:
> I'll experiment with some ways of coding HMAC to run fast in Python.
> The implementation that comes with the 2.2 library is pretty slow.

Hmm, as the one who submitted this code, I don't see much room for
improvement:

- the hexdigest method should really use binascii.hexlify(), but I
  didn't know about the existence of it at the time

- I don't see how to optimize _strxor

But then, I never benchmarked the module.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From dalke at dalkescientific.com  Tue Apr  2 23:05:45 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 2 Apr 2002 21:05:45 -0700
Subject: Static Variables
References: <a8dt35$a2i$1@slb2.atl.mindspring.net>
Message-ID: <a8dvbr$dot$1@slb4.atl.mindspring.net>

Jim Jacobs:
>Is it possible to simulate "C" style static variables in functions?  In
>methods?

Depends on which aspect you're looking for.  You probably want
something like:

_cache = {}
def spam(i):
  if i in _cache:
    return _cache[i]
  x = hairy_algorithm_with_big_pointy_teeth(i)
  _cache[i] = x
  return x

Notes on the differences:
  '_cache = {}' is executed once, but during module import.  In C,
all statics are initialized before main is called (or is that C++?).

  'static' is used in C functions as a way to hide scope.  That's
less needed in Python because module variable have module scope, while
C variables have "compilation unit" scope.

  If you wanted to change the actual value of '_cache' (that is,
replace it with a different reference) you would need to declare it
in the function as 'global'.

  Variables starting with an '_' are not imported during "import *"

  If you want more complicated functions, you can also create a
function object, like

class spam:
  def __init__(self):
    self._cache = {}
    # prefill the cache
    for i in range(10):
      self(i)
  def __call__(self, i):
    if i in self._cache:
      return _cache[i]
    x = hairy_algorithm_with_big_pointy_teeth(i)
    _cache[i] = x
    return x

spam = spam()

'static' does other things than this; is there something else you
are looking for?

                    Andrew
                    dalke at dalkescientific.com





From donn at u.washington.edu  Tue Apr 23 18:08:10 2002
From: donn at u.washington.edu (Donn Cave)
Date: 23 Apr 2002 22:08:10 GMT
Subject: non-blocking sockets
References: <Pine.GSO.4.30.0204222138510.19452-100000@eddie.cdf> <5a4226f0.0204231304.32e9676b@posting.google.com>
Message-ID: <aa4m0a$rpq$1@nntp6.u.washington.edu>

Quoth kevin at cazabon.com (Kevin Cazabon):
| Why not use a separate thread to monitor the socket, and post an event
| when data is read from the socket and available to the rest of the
| program?  This way, the socket thread can block all it wants, as its
| only job is to monitor for incoming data, and the rest of your program
| can happily do other stuff in the meantime.

Why not?
 - Threads aren't universally portable?
 - On platforms with threads, sockets don't universally support
   concurrent read and write?
 - the thread interlocks over the read buffer are more work than
   the select() single-threaded solution?
 - it's even more complicated if there are multiple inputs?
 - there's no obvious way to tell a thread to quit, when it's
   blocking on a socket read?
 - select() also allows the program to happily do other stuff as
   long as it doesn't need data, and it naturally supports multiple
   inputs?

There's probably more, as I haven't used threads this way enough to
have run into all the problems.  It can be done, and I'm sure if you
know what you're doing, it works out fine, but since you asked ...

	Donn Cave, donn at u.washington.edu


From zayats at blue.seas.upenn.edu  Wed Apr  3 16:17:44 2002
From: zayats at blue.seas.upenn.edu (Salim Zayat)
Date: 3 Apr 2002 21:17:44 GMT
Subject: File opening
Message-ID: <a8frho$qqu$1@netnews.upenn.edu>

Hey all.  I was wondering about opening files.  Now I know you need to 
include the path to the file in the open statement.  Now the problems 
arises here, though.  I am writing a module that reads in a config file 
and creates buttons from it.  So if my module is run directly, the only 
thing you need is 

> open('./buttons.config', 'r')

But if you wanna import the module, then the path to the open statement 
needs to be full path to the file.  Is there a fix to this?

Thanks!


Salim


From sholden at holdenweb.com  Wed Apr 17 17:56:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 17:56:42 -0400
Subject: Help me learn Python !
References: <4t7pbu4u83fe7et66lqdvhobq0hv975rn0@4ax.com> <3J%u8.868$hl2.215@atlpnn01.usenetserver.com> <pc2sbuo47o1dk9icpl37mols59sutprrbj@4ax.com>
Message-ID: <temv8.7286$Ip3.4698@atlpnn01.usenetserver.com>

"peter" <REMOVEpeterp100 at hotmail.com> wrote ...
> On Tue, 16 Apr 2002 16:19:16 -0400, "Steve Holden"
> <sholden at holdenweb.com> wrote:
>
> >"peter" <REMOVEpeterp100 at hotmail.com> wrote ...
> >> I'm using Windows 2000, and the Xitami webserver.
> >>
> >>
> >I don't ever remember having these problems on my Win98 system. Although
> >I've now got Win2k I have not yet tried any LRWP stuff. Hold on ...
> >
> >Nope, don't see any indentation errors -- are you sure you haven't
mangled
> >the code? Note, by the way, you need to change the call
> >
> >            self.sock.connect(self.host, self.port)
> >
> >to
> >
> >            self.sock.connect((self.host, self.port))
> >
> >before it will run - a change in the library brought it into line with
the
> >documentation, breaking code that took advantage of the opriginal bug.
For
> >the record, I just ran the code and everything works fine: I get five
> >responses from testlrwp (which responds at "/testapp1"), and then "The
URL
> >you asked for does not exist on this website". All as expected.
> >
> >Hope this helps.
> >
> >regards
> > Steve
>
> Thank you for you reply,
>
> I think my problem has something to do with this website.  As I copy
> and paste the code from here
[URL and code deleted to avoid anybody else gettijng sucked in]

Well, that would explain it. Whoever is running that web site wants to try
using their own ocde occasionally :-) -- they have completely removed all
indentation from the code!

>
> is there another source for these libraries.
>
Indeed there is. If you've downloaded Xitami and installed it in <XitamiDir>
then you will also have a <XitamiDir>/addons/lrwp directory, and the code is
there, correctly indented.

> Also, I'm using notepad and IDLE.
>
>The only thing you'll want to watch out for is that a lot of Python code
now uses four-space indentation. While you can mix tabs and spaces (the
interpreter treats tabs as taking the spacing to the next multiple of 8)
this is NOT recommended.

Hope you can now run LRWP processes successfully. Don't forget that
modification to the accept() call!

regards
 Steve







From neuhauser at mail.cz  Tue Apr 30 10:40:00 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 16:40:00 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <atxz8.34774$n4.7370773@newsc.telia.net>
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <roy-E96293.07453430042002@news1.panix.com> <mailman.1020168414.27104.python-list@python.org> <atxz8.34774$n4.7370773@newsc.telia.net>
Message-ID: <20020430144000.GP325@roman.mobil.cz>

> From: "Fredrik Lundh" <fredrik at pythonware.com>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Tue, 30 Apr 2002 14:01:42 GMT
> 
> Roman Neuhauser wrote:
> 
> > having spotty knowledge of python
> 
> that's pretty obvious from your posts.
> 
> maybe you should spend a little more time learning, and
> a little less time telling us about your "opinions" on issues
> you don't really understand?
> 
> further reading:
> 
> http://www.google.com/search?q=cache:opG_4HjTnbIC:www.theonion.com/onion3811/child-rearing_advice.html

    how does that relate to python or OO?

-- 
FreeBSD 4.4-STABLE
4:38PM up 1 day, 7:04, 12 users, load averages: 0.03, 0.04, 0.00




From never at mind.info  Sun Apr 28 19:39:04 2002
From: never at mind.info (G. Willoughby)
Date: Mon, 29 Apr 2002 00:39:04 +0100
Subject: Simple text-editor in Python?
References: <aacdeg$7g$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <aai11p$jh$1@newsg1.svr.pol.co.uk>

I've started to write a text editor in python for Win32, its inspired by
EditPlus for windows. its nearly finished, just some prefs stuff to add plus
syntax highlighting, line numbering, find dialog to get working but appart
from that its getting there! email me if u want the source to carry on with!
i've kinda left it on the back burner for a while.

G. Willoughby

email:     sab (at) freeuk (dot) com





From jon+usenet at unequivocal.co.uk  Mon Apr 15 11:54:44 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Mon, 15 Apr 2002 15:54:44 -0000
Subject: httplib drops ssl connection to MS IIS Server
References: <mailman.1018741166.19994.python-list@python.org>
Message-ID: <slrnabltto.f08.jon+usenet@snowy.squish.net>

In article <mailman.1018741166.19994.python-list at python.org>, Marco Caamano wrote:
> socket.sslerror: (5, 'EOF occurred in violation of protocol')

It is a bug in MS IIS, it just closes the SSL socket rather than going
through the SSL shutdown negotiation. I don't think you can work
around this using httplib.

This is discussed in SourceForge bug id 494762. That was closed
without fix so I have opened a new one 544234.


From dave at rotwang.freeserve.co.uk  Fri Apr  5 06:20:34 2002
From: dave at rotwang.freeserve.co.uk (Dave Berkeley)
Date: Fri, 5 Apr 2002 12:20:34 +0100
Subject: location of a module's directory
Message-ID: <a8k1cb$s9v$1@news6.svr.pol.co.uk>

Is there a way to discover where a module was loaded from?

I want to include a text configuration file with a .dll, and place them in
the same directory. If the module knows where it was loaded from it will be
able to find the config file.

If you do (Python 2.1.1)

>import sys
>print sys.modules

you can get a dictionary of loaded modules, and a description, which
includes the path the module was loaded from. However, I can find no way of
accessing this information programmatically.

Thanks

Dave Berkeley




From jlamanna at its.caltech.edu  Sat Apr 27 15:54:29 2002
From: jlamanna at its.caltech.edu (James Lamanna)
Date: Sat, 27 Apr 2002 12:54:29 -0700
Subject: PyQt 3.1 Compilation Problems
Message-ID: <3CCB01F5.1090607@its.caltech.edu>

I have QT 3.0.3 installed (using a Debian package).
I've download and compiled libsip.
However, when I try to compile PyQt-3.1 (for Qt 3.x), I get the 
following error on 'make install':

/bin/sh ../mkinstalldirs /usr/lib/python2.1/site-packages
  /usr/bin/install -c -m 644 qt.py /usr/lib/python2.1/site-packages/qt.py
make  install-data-hook
make[3]: Entering directory `/home/james/PyQt-3.1/qt'
(cd /tmp; 
PYTHONPATH=/usr/lib/python2.1/site-packages:/usr/lib/python2.1/site-packages 
/usr/bin/python -c "import qt")
Traceback (most recent call last):
   File "<string>", line 1, in ?
   File "/usr/lib/python2.1/site-packages/qt.py", line 47, in ?
     import libqtc
ImportError: /usr/lib/python2.1/site-packages/libqtcmodule.so: undefined 
symbol: metaObject__C9QSGIStyle
make[3]: *** [install-data-hook] Error 1
make[3]: Leaving directory `/home/james/PyQt-3.1/qt'
make[2]: *** [install-data-am] Error 2
make[2]: Leaving directory `/home/james/PyQt-3.1/qt'
make[1]: *** [install-am] Error 2
make[1]: Leaving directory `/home/james/PyQt-3.1/qt'
make: *** [install-recursive] Error 1

Any ideas how this would be resolved?
Thanks.

--James Lamanna





From aahz at pythoncraft.com  Fri Apr  5 12:18:42 2002
From: aahz at pythoncraft.com (Aahz)
Date: 5 Apr 2002 12:18:42 -0500
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
Message-ID: <a8km9i$7t6$1@panix1.panix.com>

In article <slrnaapc23.r7n.philh at comuno.freeserve.co.uk>,
phil hunt <philh at comuno.freeserve.co.uk> wrote:
>
>Recently on this newsgroup there's been a discussion about adding a 
>bool datatype to Python; and Guido has decided to do this.
>
>Some time ago, there was a rather contentious discussion about 
>changing the semantics of the / operator. To be precise, what should 
>it do when the arguments are integers: should 1/2 be 0 or 0.5?
>
>In both cases, there are good arguments in favour of the change. 
>(Though having said that, in both cases Python has done perfectly 
>well for years without it). The problem with making these changes is 
>that they can cause currently-working Python code to fail on new 
>versions of the language.

Note that in the case of division, the incompatible change has been
officially postponed to Python 3.0.  In both cases, the argument is that
Python has *not* worked "perfectly well"; both cases in current Python
generate ambiguities of the sort that Guido loathes.

Whether the removal of those ambiguities is worth the pain of change is
something that Guido has to rule on.  I'm a little disappointed that the
division case took so much heat before Guido modified things
sufficiently to handle the problems pointed out, but at this point, I
feel confident that the end product was at least not-bad decisions, even
if I am not wholly convinced that they were good decisions.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From sholden at holdenweb.com  Wed Apr 17 10:15:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 10:15:31 -0400
Subject: optimization with the "compiler" module
References: <mailman.1018990292.9573.python-list@python.org> <a9icab$phq$1@nntp9.atl.mindspring.net> <Xns91F32355A8FA9RASXnewsDFE1@130.133.1.4> <lJev8.12963$hl2.6959@atlpnn01.usenetserver.com>
Message-ID: <6ufv8.14393$hl2.2057@atlpnn01.usenetserver.com>

"Steve Holden" <sholden at holdenweb.com> wrote ...
> When you refer to an "upgrading function" I suspect you mean a coercion?
> There's no need for a coercion, since for any Strange instance s, s+x == s
> for any x (and, indeed, s+x is x as well)\

That last should, of course, have read

    (and, indeed, s+x is s as well)






From mwh at python.net  Tue Apr  2 14:19:19 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 2 Apr 2002 19:19:19 GMT
Subject: PyArg_BuildValue, how to return a PyString object?
References: <5a4226f0.0204012004.24bb7b00@posting.google.com> <lkelhya3rg.fsf@elios.maths.bris.ac.uk> <3CA9E603.F950315@noaa.gov>
Message-ID: <lk6639x615.fsf@elios.maths.bris.ac.uk>

Chris Barker <Chris.Barker at noaa.gov> writes:

> Michael Hudson wrote:
> 
> > No.  Rememebr to DECREF "O" args after calling Py_BuildValue --
> > something I'm always forgetting.
> 
> Can't you use "N" instead? That seems to have worked for me.

Oh yeah.  Thanks for that!

In my defence, I'll just claim that I only learnt the format codes
properly for PyArg_ParseTuple...

Cheers,
M.


From cstyphoon at yahoo.com.cn  Fri Apr  5 00:36:57 2002
From: cstyphoon at yahoo.com.cn (=?gb2312?q?=B3=DB=20=C0=EE?=)
Date: Fri, 5 Apr 2002 13:36:57 +0800 (CST)
Subject: I like python
Message-ID: <20020405053657.80679.qmail@web15002.mail.bjs.yahoo.com>

                              styphoon at yahoo.com.cn  
                              Shichuan Province China

Hello,comrades!
I am a highschool-student in Chinese, and I am fond of
computer, especial programming.But now I have some
difficulties.Are you ready to help me? If it is
possible,then I will tell you what difficulties I have
met.I know only a little English, and, do you know
Chinese? The next paragraph is written in Chinese, can
you understand? I wonder whether you can write to me
in Chinese? In  that way, I will be more grateful to
you.
  


 
??????????,????????,?????,??????????,???????????.????????,???????????????????????,??????.?????????,?????????????.
???,???????,??,??!
                          ???  ??
                         2002.4.5.
                         Best wishes!
             

_________________________________________________________
Do You Yahoo!? 
???????????????????????
http://cn.worldcup.yahoo.com/




From philh at comuno.freeserve.co.uk  Wed Apr 10 12:30:05 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Wed, 10 Apr 2002 17:30:05 +0100
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <orGs8.5669$d7.1571810@newssrv26.news.prodigy.com> <a8vo37$s8a$1@panix1.panix.com>
Message-ID: <slrnab8q4d.ec2.philh@comuno.freeserve.co.uk>

On 9 Apr 2002 17:56:55 -0400, Aahz <aahz at pythoncraft.com> wrote:
>In article <orGs8.5669$d7.1571810 at newssrv26.news.prodigy.com>,
>dsavitsk <dsavitsk at e-coli.net> wrote:
>>"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
>>news:slrnab617e.97a.philh at comuno.freeserve.co.uk...
>>>
>>> IMO, anyone for whom K&R isn't to their taste, probably won't make a
>>> very good programmer anyway.
>>
>>Yow! IMO, anyone who thinks that only one sort of person (a sort who
>>likes a particular book no less) makes a good programmer needs to relax
>>a bit in their judgmentalness.
>
>Well, Phil *should* have said "good C programmer",

That was sort-of what I meant.

C has a particular aesthetic, to do with sparseness and simplicity. 
K&R captures that aesthetic well, IMO.

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From pan-newsreader at thomas-guettler.de  Tue Apr  9 17:26:38 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Tue, 09 Apr 2002 23:26:38 +0200
Subject: Emacs-like editor, scriptable with python?
Message-ID: <pan.2002.04.09.23.26.38.473612.1108@thomas-guettler.de>

Up to now I was happy with emacs because I never used
to programm it with lisp. Now I need to and find it difficult.

Is there an emacs-like editor which is scriptable with python?

  thomas

-- 
Thomas Guettler <guettli at thomas-guettler.de>
http://www.thomas-guettler.de


From mlh at vier.idi.ntnu.no  Wed Apr 10 23:18:58 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 11 Apr 2002 03:18:58 +0000 (UTC)
Subject: Tuple comprehension
Message-ID: <slrnaba052.efc.mlh@vier.idi.ntnu.no>

It seems this has been brought up before, but I could only find two
posts about it (about a year old). The question is: Could we have
tuple comprehensions?

The argument I saw used earlier was based on the author's
preconceptions about tuples being primarily "records" and not
"containers". The only defining quality I see to tuples as opposed to
lists is that they are immutable. (Yes, I see the "recordness", but I
don't think it's an argument against tuple comprehension.)

Tuple comprehension could be useful when creating a tuple based on
another sequence, where you'd iterate over the values, doing something
to each of them. Of course you can do tuple([x for x in ...]). If
adding tuple comprehensions is too costly, then this is perfectly
viable. (I guess a tuple comprehension would have to create a list or
something similar internally anyway...)

As far as I can see, the PEP on dict comprehension is still open... I
think both dictionary comprehension and tuple comprehension seem like
natural parallels to list comprehension. But then I would, wouldn't I?
<wink>

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From tim.one at comcast.net  Sun Apr 14 17:32:30 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 17:32:30 -0400
Subject: Topy news - and a request for Python pretty-printing
In-Reply-To: <oqadsam0de.fsf@carouge.sram.qc.ca>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEBPOOAA.tim.one@comcast.net>

[Fran?ois Pinard]
> ...
> Would someone happens to know if there is a pretty-printer for Python
> source _lines_, where I would find some inspiration maybe?  I do not need
> an indenter, there is no problem there.  The problem is to split a long
> Python line into many lines which would nicely represent the original,
> that is, for example, for which indentation stresses operator precedence.
> Cleverly breaking long string constants is also an interesting problem!

Sorry, I haven't heard of such a beast.  The closest I've seen is the "block
comment" reflowing facilities in various Python editors.  For very long
lines with clear structure (lists, tuples and dict constructors; function
calls), I suppose you could do a lot worse than to feed them to Emacs,
injecting a newline after every comma.  Then they'll line up the way Guido
would type them by hand <wink>.






From logiplexsoftware at earthlink.net  Tue Apr 16 15:47:51 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 16 Apr 2002 12:47:51 -0700
Subject: python threading
In-Reply-To: <E16xWz4-0004oN-00@mail.python.org>
References: <6bd9f01b.0204152132.6a6ad09e@posting.google.com>
	<8MVu8.26112$vF6.828988@news2.tin.it>
	<x7elhfpmvc.fsf@ra.iago.org>
	<E16xWz4-0004oN-00@mail.python.org>
Message-ID: <20020416124751.0a6af80d.logiplexsoftware@earthlink.net>

On Tue, 16 Apr 2002 19:42:37 +0200
Alex Martelli wrote:

> On Tuesday 16 April 2002 07:35 pm, Justin Sheehy wrote:
> 	...
> > I've wrapped Queues with Events before, and it makes things a little
> > nicer.  It is quite straightforward to make a Queue that exposes a
> > get method that will simply block until an Event is set by another
> > thread or an optional timeout has elapsed.
> >
> > It doesn't solve the whole problem, but waiting on an Event is much
> > nicer than a get/sleep loop.
> 
> Yes, but a GUI thread can't do that either -- nor can a network one;
> they must be ready for user-input events and netword-data-arriving
> events too.  The problem is how to seamlessly merge waiting for
> THOSE kinds of things and for Python-thread-stuff kinds...

I'm not too familiar with other GUI toolkits, but in wxPython it's possible to
generate a custom wxEvent that is put on the GUI's event queue and processed
like any other GUI event.  You can then create an event handler for that custom
event and safely do your updates there.  I've found this to be the simplest way
to interact with a GUI from a thread.  I'd be surprised if PyQt didn't support
something similar.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From Oschler at earthlink.net  Mon Apr  1 02:04:00 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Mon, 1 Apr 2002 02:04:00 -0500
Subject: Python/Linux
References: <0UMp8.42204$1f1.3348597@e3500-atl2.usenetserver.com> <nfTp8.44157$1f1.3618974@e3500-atl2.usenetserver.com> <slrnaaftai.8mf.gerhard@lilith.my-fqdn.de>
Message-ID: <w20q8.1572$0T6.38985@e3500-atl1.usenetserver.com>

Gerhard,

Ok, I'll give it a whirl (Mozilla).  Best download link?  Hopefully got SUSE
binaries :).

Also, will it coexists with Konqueror (which shells Netscape I think) and
Netscape?

thx

Gerhard H?ring <gerhard at bigfoot.de> wrote in message
news:slrnaaftai.8mf.gerhard at lilith.my-fqdn.de...
> Robert Oschler wrote in comp.lang.python:
> > What's so special about Mozilla
>
> Short version:
> Netscape 4.x sucks a**. Mozilla rocks.
>






From aleax at aleax.it  Sat Apr 13 11:36:14 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sat, 13 Apr 2002 15:36:14 GMT
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it> <7xvgaw80nd.fsf@ruckus.brouhaha.com>
Message-ID: <OfYt8.16090$vF6.498361@news2.tin.it>

Paul Rubin wrote:

> Alex Martelli <aleax at aleax.it> writes:
        ...
> the arcsin, etc).  Using lambdas for the functions where there weren't
> ready-made built-ins seemed natural and obvious to me.  Would you really
> have written separate definitions for those functions?

I might -- and I might organize the whole structure as one dictionary of 
pairs (function, inversefunction) rather than two dictionaries of functions
(although I see where your organization might be handier, depending on how
the code is structured to use these tables).  Neither assertion implies that
I consider your choices "wrong"!  Just that we prefer different styles.  
E.g.:

>     unops = {'sqrt': math.sqrt,
        ...
>              'log': lambda x: math.log(x)/math.log(10),
        ...
>     inv_unops = {'sqrt': lambda x: x*x,
        ...
>                  'log': lambda x: 10.**x,

vs:

# unary functions that module math lacks
_1_over_ln_10 = 1.0 / math.log(10)
def log10(x): return math.log(x) * _1_over_ln_10
def pow10(x): return 10.0 ** x
def square(x): return x * x

# map name -> (function, inverse_function)
unops = { 'sqrt': (math.sqrt, square),
           'log': (log10, pow10),
        ...

I _think_ this arrangement is fractionally clearer than the lambda-based 
one.  Just my opinion, of course.


Alex



From jb at cascade-sys.com  Sun Apr 21 05:38:27 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 21 Apr 2002 02:38:27 -0700
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <m3bscdlwzc.fsf@chvatal.cbbrowne.com>
Message-ID: <3CC28893.C65118FF@cascade-sys.com>

Christopher Browne wrote:

> Well, after 40 years, we're still in the process of seeing features
> Lisp had years ago move into the mainstream.

Good point, excellent examples.

> C++ is
> newer than PL/1; is it _truly_ greatly superior, or is it just that
> it's well-marketed, while nobody's pushing PL/1?

I guess nobody is pushing it.  I understand IBM trademarked PL/1 thru PL/100 but we've yet to see
PL/2.

Then too, I thought PL/1 lacked classes/objects, which would give C++ a big edge.

> As a language, Java has little that _isn't_ in a lot of its
> predecessors.  It may be nicer than C in a lot of ways, but for it to
> have taken 30 years to do that seems rather wasteful.

It's a marketing gimick as much or more than a programming language.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From marklists at mceahern.com  Tue Apr 30 21:59:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 30 Apr 2002 20:59:31 -0500
Subject: crypto
In-Reply-To: <7x4rhshcyx.fsf@ruckus.brouhaha.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEIPKNAA.marklists@mceahern.com>

[Paul Rubin]
> Do you want to interoperate with some particular program like gpg?

I'm not sure, but I doubt I'll need to do that.

> If not, I have a Python library (pre-alpha, but you can try it out if
> you want) that does what you're asking.  Preliminary docs are at:
>
>   http://www.nightsong.com/phr/crypto/crypto.txt

Wow, I really like the philosophy behind this approach.  What kind of
license are you envisioning for this?  I'd be interested to try it out.

Thanks!

// mark





From aleax at aleax.it  Fri Apr 12 03:12:09 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 12 Apr 2002 07:12:09 GMT
Subject: try/except/else flexibility...
References: <slrnabcon3.g5r.mlh@vier.idi.ntnu.no>
Message-ID: <dNvt8.10463$vF6.308097@news2.tin.it>

Magnus Lie Hetland wrote:
        ...
>   try: foo + ''
>   else: bar()
> 
> and if I want something done only if it causes no exceptions, for
> instance pop off an element of a list unless it is empty, I could just
> do:
> 
>   try: somelist.pop()
> 
> Is this heresy? I see the harm in not specifying exactly which
> exceptions you're expecting, of course -- are there other arguments
> against this? (I'm sure that might be argument enough... Oh, well...)

It seems to me that forgetting or mis-aligning an except or finally
clause is an error that these language changes would make it harder
to diagnose.  I'm an enthusiast for "easy to ask forgiveness than
permission", but this way you would make it easiest and most
natural to not even ask permission.  I think this would take us
back too close to (say) the typical problem with C -- system calls
or library functions carefully giving out error codes, and bulldozing
application programmers totally ignoring those codes and doing no
error-checking at all.  One advantage of exceptions is that the
default is NOT to ignore -- and the ease to just put everything in 
a blind 'try:' might, I fear, claw back most of this advantage,


Alex



From huaiyu at gauss.almadan.ibm.com  Thu Apr 11 14:08:27 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 11 Apr 2002 18:08:27 +0000 (UTC)
Subject: Tuple comprehension
References: <mailman.1018540222.3383.python-list@python.org> <oGit8.8539$vF6.244299@news2.tin.it>
Message-ID: <slrnabbk8q.j2e.huaiyu@gauss.almadan.ibm.com>

On Thu, 11 Apr 2002 16:17:24 GMT, Alex Martelli <aleax at aleax.it> wrote:
>Tim Peters wrote:
>        ...
>> ...     for i in range(n):
>        ...
>> is one way to get the same end result without building a list in one gulp
>
>Actually, range DOES "build a list in one gulp" (but I guess you might
>use xrange instead:-).

One ancillary question: why xrange(n) is not iterator?

Huaiyu


From tim.one at comcast.net  Sun Apr 14 16:25:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 16:25:32 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <Xns91EFEE502A532JamesLLugojcom@216.148.53.85>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEBMOOAA.tim.one@comcast.net>

[Tim]
>> What are you willing to do to make your desires a reality?  If it
>> doesn't involve contributing work, time or money, you just want a
>> free ride.  But nobody owes anyone a free ride, and open source
>> can't change that fact of life either.

[James Logajan]
> Not sure, but I think you are close to describing what has been
> called the tragedy of the commons.

I think that's more about a finite resource getting destroyed when there's
no cost associated with using it.  The sheer number of people using Python
doesn't cost *us* anything apart from paying for the network bandwidth to
support the downloads, and we generally view more users as a very good
thing:  more testing, more acceptance, more contributions back to the
project (although you don't like the latter, because that means change too).

The problem is more that when someone wants something for nothing, they're
limited to taking what other people are willing to give away.  If what you
want is what people freely give to Python, that works great.

> Back when I was paid to do Python work, I contributed all of US$50 (I
> think; maybe it was $20) to the PSA (I think it was called)

Bless you.  It was probably $50 (unless you went for the student rate).

> and tried to get the company I was consulting for to kick in a couple
> thousand (or a couple hundred at least). They never got around to that,
> alas.

If they're feeling guilty yet <wink>, they can sponsor the PSF instead:

    http://www.python.org/psf/

> Unfortunately I'm not sure what would be gained by contributions to a
> strategy that I have specifically stated I don't agree with.
> Further, it is  unclear how I can "contribute" to the slower pacing of
> releases that I'd like to see.

I didn't ask you to contribute to the PSF, or to PythonLabs, I asked what
you were willing to do to make *your* desires a reality.  Python is the sum
total of what everyone else has given to achieve theirs, and that's how open
source works.  If you think, for example, that Paul Rubin and Donn Cave are
on the same page as you, what are the three of you willing to do -- with or
without Guido's help?  For example, fork off 1.5.2 and release a "Guaranteed
Unchanging in Jot or Tittle until Universal Heat Death" variant.  Guido
doesn't have time for that, and none of the current volunteers appear to
have any interest it.  If you say there's a market for it, do something
about it, or talk a group of like-minded people into doing it for you.
Charge money for it, if you like -- the Python license is very liberal, and
you take all of it without paying anyone a penny for it.

> I've overstayed my welcome here,

You haven't overstayed mine <wink>.

> but before I go I'd like to thank you for your efforts to advance
> Python and I wish you well in your endeavors.

Thanks, and to you too.  I expect there is agreement that it would be nice
if commercial users had a better way to proceed, but I don't see a way to
get there from here unless commercial users step up to the plate and do work
to make that happen.  The idea that Python will freeze, like sh or awk,
isn't going to happen.





From garrett at bgb-consulting.com  Mon Apr  1 14:30:36 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Mon, 01 Apr 2002 19:30:36 GMT
Subject: Moin/Boa
References: <mailman.1017439005.30794.python-list@python.org>
Message-ID: <3CA8B590.8E13E6DC@bgb-consulting.com>

  I run Moin under Apache, perhaps if you give us more details on the problem,
we can help!

Glenn Stauffer wrote:
> 
> Anyone have any experience running moin with the boa web server?  I'm running
> into a bit of a problem that I'm sure is a configuration issue, but I can't
> find anything in either set of docs to tell me what I need to do to fix it.
> 
> Glenn

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From joost_jacob at hotmail.com  Tue Apr  9 12:06:35 2002
From: joost_jacob at hotmail.com (J.Jacob)
Date: 9 Apr 2002 09:06:35 -0700
Subject: What has PEP 285 done to us?
References: <20020408134207.1af35256.logiplexsoftware@earthlink.net> <mailman.1018339297.32067.python-list@python.org> <3CB2AADF.9050908@mxm.dk>
Message-ID: <13285ea2.0204090806.4a2d987b@posting.google.com>

Max M <maxm at mxm.dk> wrote in message news:<3CB2AADF.9050908 at mxm.dk>...
> ...
> 
> So when Guido wrote that the post was an overlong piece of FUD I more or 
> less agreed. But I was rather surprised by his choice of words, and 
> found it to be a rather rude brush-of.

Ok that made me use google to try to look up what FUD means and it is:
"Fear, Uncertainty and Doubt".
I saw it before and knew it was supposed to be something bad but I was
always too lazy to look it up :)

> 
> It was hinted that discussions had taken place on python-dev, but I 
> don't subscribe to it so all I, and I guess most others, got was the 
> short reply from Guido.

Is there a way to peek into the latest python-dev messages? I could
not find a link at http://www.python.org/psa/MailingLists.html
Maybe if you subscribe to that mailing list (have to find their
address first) you get too much email to handle so being able to just
have a look at the lastest postings would be nice.


From grante at visi.com  Fri Apr 19 15:40:29 2002
From: grante at visi.com (Grant Edwards)
Date: Fri, 19 Apr 2002 19:40:29 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <NCBBLFCOHHDIKCAFGCFBCEJHKLAA.marklists@mceahern.com> <mailman.1019068175.11920.python-list@python.org> <99pvbuc5a668bqniibq10ussa9f7v3nq34@4ax.com> <mailman.1019233076.19513.python-list@python.org> <3CC0578E.822EB961@metaslash.com> <mailman.1019240861.31256.python-list@python.org>
Message-ID: <No_v8.45338$vm6.6047942@ruti.visi.com>

In article <mailman.1019240861.31256.python-list at python.org>, Cliff Wells wrote:

> Interesting.  It hadn't occurred to me that exceptions would be
> any faster than .get().  I didn't use a predefined dictionary
> since the OP was looking for an equivalent to C's ternary
> operator (which I took to mean "one line"), but that's clearly
> the most important optimization if speed is a concern (athough
> it's arguably more readable as well).

When somebody asks about a ternary operator they usually don't
really care about a ternary operator.  What they _really_ want
to do is to know how to solve in Python a problem that was
sovled in C by using a ternary operator.

-- 
Grant Edwards                   grante             Yow!  I'm encased in the
                                  at               lining of a pure pork
                               visi.com            sausage!!


From sjoshi at ingr.com  Fri Apr 26 17:12:16 2002
From: sjoshi at ingr.com (Sunit Joshi)
Date: 26 Apr 2002 14:12:16 -0700
Subject: How do I format this
Message-ID: <8f8ffe67.0204261312.77787e94@posting.google.com>

Hello
I have a format string which I'm formatting using a dict. The format
string is
"""
<td width="1\%">			
 <input type=radio value="%(fullname)s****%(isdir)s" name=radioName>
 <IMG height=17 hspace=2 src="ezmedia/window/filephotoicon.jpg"
width=19>
 </td>																				
 <td>
<div class="menucontents"><a href="%(fullname)s">%(name)s</a></div>
  </td>																				
 <td width="10\%">
<div class="menucontents">%(ext)s</div>				
 </td>																				
 <td width="15\%">
<div class="menucontents">%(size)s</div>			
 </td>																				
 <td width="22\%">						
 <div class="menucontents">%(modified)</div>
 </td>
</tr>
"""

The dictionary is 
{'isdir': 0, 'indent': 1, 'name': 'TonyFile.txt', 'isroot': 0,
'fullname': 'E:\\temp\\tony\\TonyFile.txt', 'modified': 'Fri Apr 19
09:02:17 2002', 'sizefm': ' Bytes', 'size': 20.0}

However it fails at <td width="1\%"> eventhough I'm escaping the '%'
with a message ValueError: unsupported format character '"' (0x22) at
index 17

Any ideas what I'm doing wrong here.

thanks
Sunit
sunitjoshi at netzero.net


From frr at easyjob.net  Tue Apr 23 13:15:00 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 19:15:00 +0200
Subject: Newbie: Running a script on WindowsXP
References: <#dmZtdu6BHA.1188@cpimsnntpa03>
Message-ID: <hj5bcuol0p4l0evef0uuucdeg91vo45nfg@4ax.com>

On Tue, 23 Apr 2002 10:54:07 -0600, "billludden" <billludden at email.msn.com>
wrote:

>Hi,
>
>I am trying to figure out how to run a Python21 script
>from the "Run" box on WindowsXP Home edition.
>When I run it, the Command Prompt box just briefly
>flashes on the screen and closes itself.  I have tried
>using the -i option like you would use in Windows 98
>but it doesn't seem to work.  I also tried appending
>Autoexec with:
>    SET PATH=.........;C:\PROGRA~1\PYTHON21

Type "python myscript.py" in the run box.




-----------------------
Fernando Rodriguez


From bokr at oz.net  Mon Apr  1 21:22:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 2 Apr 2002 02:22:12 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CA751EF.10F25F2D@ndh.net> <3CA75ECC.389A0186@alcyone.com> <3CA840A5.A522501D@ndh.net> <3CA8B4CC.88C35BB0@alcyone.com> <a8an7b$eme$0@216.39.172.122> <eppstein-250858.15583201042002@news.service.uci.edu>
Message-ID: <a8b4kk$je2$0@216.39.172.122>

On Mon, 01 Apr 2002 15:58:32 -0800, David Eppstein <eppstein at ics.uci.edu> wrote:

>In article <a8an7b$eme$0 at 216.39.172.122>, bokr at oz.net (Bengt Richter) 
>wrote:
>
>> Well, for types that are related in certain ways ... ;-)
>> 
>>  >>> 0 == 0.0
>>  1
>>  >>> 1 == 1.0
>>  1
>>  >>> 1 == 2.0
>>  0
>> 
>> That gets into what relational operators really should mean.
>> I think I exclude bool from the number tree, so I don't
>> like the effect of subtyping bool from int, even though
>> I support context-dependent implicit conversions to integer
>> (though that context should be defined by the client expression ops).
>
>If you called it "binary" instead of "bool" would it help you think of 
>it as being more like a number?
>
No. Because numbers are ordered, and bools are not. bools are selected
elements from an unordered set of two possibilities. The bool is one chosen
element, but the element can be anything that can be distinguished from the
other element of the set. They have names str(b) associated with them, but
they don't have an order. One is chosen to mean logical true and the other
logical false, but they are just tokens with no necessary order.

By putting the bool elements in a conventional order, you can extract a
number from the situation, i.e., the position of the element in the
(now) ordered set. I.e., 0 or 1 one comes from the position of the False or True
object in a conventionally ordered set which as python list would be [False,True].

Logically the numeric value of x's boolean value is [False,True].index(bool(x))
though I'd hate to have to write it that way ;-)

False and True refer to members of a set. The members are not identical, so
the pair actually permits a binary distinction without reference to order.

You can also have sets whose elements are numbers, and ordered, and [0,1]
is a choice with optimization possibilities ;-)

But bool could just as well be represented using any two distinguishable elements,
e.g., [123, 'x'] where False=123 and True='x', and you'd still get your desired
number by [False,True].index(bool(x)), just that now bool(x) would return 123 or 'x'
instead of Guido's singletons False and True. Note that if you used False=0 and True=1,
you could fool yourself into thinking that e.g., True = [False,True].index(bool(True))
was so in general, which it is not.

(Of course, literally using 123 and 'x' wouldn't give the right str,repr, etc.,
but maybe you could subtype 123 .__class__ and 'x'.__class__ to rig two singleton
objects that might do the job, if it was reeeaally important ;-) The point is that
they don't have to have anything to do with 0 and 1 in themselves as objects, and
they don't have an inherent order, whereas numbers 0 and 1 do have counting order.

That's what I think is the relationship between bools and ints, and why they're different.
A number/set theorist might correct my terminology, but I hope
my meaning is clear.

Regards,
Bengt Richter


From marklists at mceahern.com  Sat Apr 27 09:36:19 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 27 Apr 2002 08:36:19 -0500
Subject: name of the current function ?
In-Reply-To: <aae7kf$apv$06$1@news.t-online.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEOPKMAA.marklists@mceahern.com>

[Solveig Wiesmayr]
> I would like to get the name 'my_name' of the current function to pass it
> to some other (debugging-) function.

You'll be delighted to know that this simple google query:

  http://www.google.com/search?hl=en&q=current+function+name+python

returns this simple recipe, which seems to answer your question:

  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66062

Cheers,

// mark




From marklists at mceahern.com  Wed Apr  3 12:03:53 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 11:03:53 -0600
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204031645.g33Gjorq028972@ratthing-b246.strakt.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEPPKIAA.marklists@mceahern.com>

[Laura Creighton]
> That is more strongly worded then I would make it; I only think that
> deciding that a function should be compelled to return only one of
> two values requires long and careful deliberation.  This is because it
> is likely that some day you will wish that you could provide more
> information than this allows.  isinstance is a good example
> of this. I would be pleased if isinstance returned more useful info
> than just yes/no, such as : 0, no inheritance; 1: direct instance of the
> requested class; 2: instance of a subclass of the requested class.
> 90% of uses would still be yes/no, but this would be fine, and you would
> get the extra information when you needed it.

That's an eye-opening and thoughtful response.  All I'm saying is that this
design philosophy is completely orthogonal to the value of PEP 285.

Further, if I wanted such a function, I wouldn't call it isinstance().

Cheers,

// mark





From twisted at itamarst.org  Mon Apr  1 05:18:49 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Mon, 01 Apr 2002 13:18:49 +0300
Subject: ANN: Twisted 0.16.0
Message-ID: <3CA83409.8080401@itamarst.org>

Twisted is an event-based framework for internet applications. For more
information, visit:

    http://www.twistedmatrix.com

or join the list at:

    http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python


Twisted joins the fight against Software Piracy
===============================================

The main focus of this release is our new licensing scheme. The Business
Software Alliance may have put it best, in their page on anti-piracy
(http://www.bsa.org/usa/antipiracy/):

"Software pirates not only steal from the companies that make the software,
but with less money for research and development of new software, all users
are hurt.  That's why all software piracy - even one copy you make for a
friend, is illegal."

Software piracy is a serious crime, and one that the Open Source community
has been remarkably lax in pursuing and protecting against.  This is why
Twisted Matrix Laboratories is taking the forefront in Open Source software
registration technology.

In order to do your part to prevent the tragedy of unregistered, unlicensed
software, all you need to do is go to http://www.twistedmatrix.com/license
and enter your user information to obtain a license key.  You can provide us
with as much or as little information as you like!

For more information on how Twisted helps you run licensed, unpirated
Open-Source software, visit:
http://twistedmatrix.com/documents/howto/register


What's New in 0.16.0
====================

- Added Qt event loop support

- FTP client protocol support

- Reworked coil - a web-based configuration system for Twisted servers

- Win32 event loop can now run subprocesses

- Automatic log rotation for servers (and USR1 signal for manual rotation)

- Improved support for registering shutdown methods

- Better XML-RPC integration

- Bug fixes, small feature improvements, and more.


What is Twisted?
================

Twisted is an event-based framework for internet applications.  It includes
a web server, a telnet server, a multiplayer RPG engine, a generic client
and server for remote object access, and APIs for creating new protocols and
services. Twisted supports integration of the Tk, GTK+, Qt or wxPython event
loop with its main event loop. The Win32 event loop is also supported, as is
basic support for running servers on top of Jython.

Twisted currently supports the following protocols, all implemented in pure
python, most of them as both servers and clients:

     - FTP
     - HTTP
     - SOCKSv4
     - SMTP
     - IRC
     - telnet
     - POP3
     - AOL's instant messaging TOC
     - OSCAR, used by AOL-IM as well as ICQ (client only)
     - DNS
     - LDAP (client only)
     - finger
     - Echo, discard, chargen and friends
     - Twisted Perspective Broker

For more information, visit http://www.twistedmatrix.com, or join the list
at http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python




From hfoffani at yahoo.com  Fri Apr 12 11:20:16 2002
From: hfoffani at yahoo.com (Hernan M. Foffani)
Date: Fri, 12 Apr 2002 17:20:16 +0200
Subject: How to test tp_print?
Message-ID: <20020412112023.201$FO@news.newsreader.com>

I'm using something like this to test tp_print slots:

import os

fo = open("file.txt", "w")
print >> fo, "myInstance_1", "myInstance_2"
fo.close()
fo = open("file.txt")
if fo.read() != "myInstance_1 myInstance_2\n":
    print "tp_print doesn't follow the specs"

fo.close()
os.remove("file.txt")

Instead of the strings "myInstance_*" I have the
actual objects. This is just for simplicity.

My question is:
Is this code portable across different C stdio
libraries? I'm thinking in line delimiters here.
I suspect it is not :-(

Any other approach to this?

Regards,
-Hernan





From ark at research.att.com  Sat Apr  6 21:26:08 2002
From: ark at research.att.com (Andrew Koenig)
Date: Sun, 7 Apr 2002 02:26:08 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <3CABB156.1010401@nospam.attbi.com>
Message-ID: <yu99wuvkw8fz.fsf@europa.research.att.com>

djw> Pardon my ignorance (I've only been working with Python for a short
djw> while), but I don't understand why an idiom like:

djw> if x == True:

djw> would be/should be an error (or warning initially). I see C/C++ code
djw> like this all the time:

djw> bool x;
djw> x = Foo(Bar);
djw> if (x==true)
djw> {...

In C++, you've fixed the type of x when you wrote "bool x;",
so x can be only true of false.

The nearest equivalent in Python would be

        if bool(x) == True:

which has no particular problems aside from being needlessly verbose.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From arocas at alumni.uv.es  Wed Apr 10 10:05:58 2002
From: arocas at alumni.uv.es (Arnau Rossell=?US-ASCII?Q?=F3?= Castell=?US-ASCII?Q?=F3?=)
Date: Wed, 10 Apr 2002 16:05:58 +0200 (CEST)
Subject: Working with untrusted coders
Message-ID: <9739136773arocas@alumni.uv.es>

Hi,
I want to develop a MUD, and i wat thinking that python would be a good
language to make it, but such a MUD would have many coders whom i don't
trust completely, and i'd like to restrict his access. Python security
about private elements and others is merely a "don't step on the grass"
sign, everybody can change private functions and data in classes and even
modules. I'd like to know if there is some way of really securing python,
so parts of the code can't be overriden by the coders. Has anybody tried
it?

Regards,

Arnau




From gcordova at hebmex.com  Fri Apr 12 17:07:42 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 12 Apr 2002 16:07:42 -0500
Subject: Beginner's performance problem
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623376E815@mail.hebmex.com>

> 
> [Aahz]
> > I'm familiar with TVP, but what's TVM?
> 
> texturized vegetable manicotti, of course.  ;-)
> 
> // m
> 

Or a Totally Vicious Maneatingcow.

One can never know....

TGIF!!

-gustavo




From anthony at interlink.com.au  Tue Apr  9 02:10:15 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Tue, 09 Apr 2002 16:10:15 +1000
Subject: SSL error: 2.1.1 works but 2.2 doesn't 
In-Reply-To: Message from Walter Miller <w3miller@yahoo.com> 
   of "Mon, 08 Apr 2002 23:01:54 MST." <20020409060154.59291.qmail@web21410.mail.yahoo.com> 
Message-ID: <200204090610.g396AF802261@mbuna.arbhome.com.au>

>>> Walter Miller wrote
> I tried 2.2.1rc2 with the same results.  I haven't
> tried 2.1.3.

Log a bug on sourceforge.net - if you don't, it's unlikely to get
addressed. The chances of it being addressed for 2.2.1 are so close
to zero as to be ignored, but there will be a 2.2.2, so it'd be nice
to fix it for that.

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From michal at orfeus.bioinfo.pl  Sat Apr  6 11:04:39 2002
From: michal at orfeus.bioinfo.pl (=?Latin-2?Q?Micha=B3?= Kurowski)
Date: Sat, 6 Apr 2002 16:04:39 +0000 (UTC)
Subject: python versioning and external packages
Message-ID: <a8n6an$4q4$1@sunsite.icm.edu.pl>

Hi,

I've got problems with installing external packages on a system where
the only "python" is 2.2.1c1.

The binary "python2.2" has a link python and everything seems to be ok
but when it comes to installing an external package (pybligrapher)
I've got a nasty configure error - python not found, >= 1.5.2 is
required...

Seems ridiculous at a first sight but for someone who cannot consider
himself a autoconf guru it's a serious thing ;-).

Those anyone know a workaround ?

Many thanks, 

-- 
Michal Kurowski
<michal at orfeus.bioinfo.pl>


From Malachi at bardstowncable.net  Tue Apr 30 21:37:18 2002
From: Malachi at bardstowncable.net (Owen Marshall)
Date: Tue, 30 Apr 2002 21:37:18 -0400
Subject: Newbie: Working with large database -- help!
Message-ID: <WEHz8.1942$z65.211132@news.uswest.net>

Ok. First off, I am fairly new to Python. So don't flame me if there is a
very obvious solution =)

I am trying to work with a file -- for lack of better terms, a database. The
data is like this:

data|data|data|data

and so on as needed. My question: how do I extract each part of data from
between the " | " (none of the data has fixed lengths), and put it in a list
where I can do some work with it? Changing the database format itself is not
an option, though. Any ideas?

Owen T. Marshall




From dalke at dalkescientific.com  Sat Apr 20 16:05:33 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 20 Apr 2002 14:05:33 -0600
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83>
Message-ID: <a9shph$lhc$1@slb5.atl.mindspring.net>

Lumberjack wrote:
>Of course the
>"science" in "computer science" is pure bovine excrement. If you can't or
>wont measure it, it isn't science. "Computer religion" or "computer cults"
>would better capture the reality.

Is accounting science?  Measurement is part of its name, after all.

I'm reading a collection of essays by Tom DeMarco.  I just finished
one on measurements, like LOC and cyclomatic complexity, and how useless
they are as indicators of good or bad software development.

Just because you can measure it, doesn't mean it is science.  Numerology
may be more like it.

And there are measurements in CS.  One is big O notation, which predicts
the functional form for run-time.  And you forgot the obligatory quip
that fields with 'science' in the name, aren't.  :)

                    Andrew
                    dalke at dalkescientific.com





From zaka07 at hotmail.com  Wed Apr 24 12:41:22 2002
From: zaka07 at hotmail.com (Rad)
Date: 24 Apr 2002 09:41:22 -0700
Subject: python and very large data sets???
Message-ID: <ad381f5b.0204240841.52333a3e@posting.google.com>

I am preparing myself to work on extracting data from 4 text files
(fixed width format) which combined size is about 80GB.  Considering
deadlines, costs, and my limited programming knowledge I thought using
Python/Windows for the job would be the best option for me.  However,
I am worried about the speed in which Python (me and my hardware) will
be able to deal with these massive data sets but I am hoping that this
is still a quicker route then learning C.
I still haven't received above mentioned files so I can't test the
time needed to (for example) read a 15GB "file1", filter by few
variables, and write a resulting subset as a "sub_file1".  Things
would afterwards get more complicated cause I will have to pullout
ID's from "sub_file1", remove duplicate ID's create
"no_dup_sub_file1", match those to ID's in remaining 3 main files and
pullout data linked with those ID's.

I have a few weeks to prepare myself before data arrives and my
question is: am I going the right way about the project, is Python
(with humanly written code) capable of doing this kind of staff
relatively quickly?

Any help and suggestions would be greatly appreciated.

Thanks

P.S As you probably guessed I'm new to python


From fperez528 at yahoo.com  Sun Apr  7 20:02:11 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Sun, 07 Apr 2002 18:02:11 -0600
Subject: purging file buffers - to make new data be written to file
References: <a8qhjp$7go$1@news6.svr.pol.co.uk>
Message-ID: <a8qq6m$ooj$1@peabody.colorado.edu>

> On Linux.  The program is to save a memo in a slightly organised format,
> but I find the file is not written until Python exits.

You need to call your_file_object.close() for the buffer to be written to disk 
(technically, that triggers a system write call, which is itself delayed by 
filesystem write caching. But that's a minor point).

Cheers,

f. 



From max at alcyone.com  Sun Apr  7 21:05:47 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 07 Apr 2002 18:05:47 -0700
Subject: ACCEPTED: PEP 285
References: <mailman.1018222875.17263.python-list@python.org>
Message-ID: <3CB0ECEB.E9A3C179@alcyone.com>

Arthur Siegel wrote:

> But the whole thing risks coming off as a loyalty
> test of some sort.
> 
> Does it bother you that Paul Rubin is being
> "classified" by someone as close in to the
> inner circle as Aahz.

No, I haven't.  As a matter of fact, I have no idea what you're talking
about.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From tim.one at comcast.net  Mon Apr 22 18:57:15 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 22 Apr 2002 18:57:15 -0400
Subject: calendar.py doesn't use lists???
In-Reply-To: <9a6d7d9d.0204221405.7d394207@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEDKPAAA.tim.one@comcast.net>

[Aaron Watters]
> Hi guys.  I got an irksome problem.  It's possible that
> it's fixed in python 2.2.1, but even if it is, it irks me.

It's fixed in 2.2.1.

> I tried the following script:
>
> """
> import sys
> print sys.version
> if 1:
>     from calendar import month_abbr, month_name
>     # hey python guys, i want lists!!
>     for x in (month_abbr, month_name):
>         print type(x)
>     month_abbr = list(month_abbr)
>     month_name = list(month_name)
>     print month_abbr
>     print month_name
> """

...

> Using Python 2.2 I get this
>
> """
> 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)]
> <type 'instance'>
> <type 'instance'>
> [CRASH AND BURN... WINDOWS XP SUGGESTS I SEND A COMPLAINT TO MICROSOFT...]

So do I!  Microsoft should get many more complaints than it does.  In 2.2.1
it's hard to say what it will print, because it's become locale-dependent.
On my 100% US English box, I get this in 2.2.1:

<type 'instance'>
<type 'instance'>
['', 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
     'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
['', 'January', 'February', 'March', 'April',
     'May', 'June', 'July', 'August',
     'September', 'October', 'November', 'December']

> ...
> WHO'S GOOFY IDEA WAS THIS?

As with most goofy ideas, I believe it started with a foreigner <wink>.

> Seriously, why did you need to change these sequences from
> lists where they worked with my code, to somethingsimilartolists
> that broke my code?

They wanted locale-dependent results, including when changing locale
dynamically.  OTOH, the variables you imported aren't documented, so you you
were on shaky ground importing them anyway.

> Irked.  In general I've noticed a tendancy with recent python
> releases towards changing stuff for the fun of changing stuff,
> I'm sorry to say.

You'll have to take that up with the person who wanted this badly enough to
submit the patch.

> Please remember that when you fix something that isn't broken, you
> might just be breaking somebody else's code.

Well, they thought getting back English names was broken.  I don't judge it,
I just fixed it <wink>.

> Sorry to whine.

na-it's-always-nice-to-hear-from-you!-ly y'rs  - tim





From fperez528 at yahoo.com  Tue Apr 23 12:52:07 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 23 Apr 2002 10:52:07 -0600
Subject: file_ref = open(string_ref)
References: <1019579845.565223@cswreg.cos.agilent.com>
Message-ID: <aa43jd$a2l$1@peabody.colorado.edu>

Greg Weeks wrote:

> Is this code
> 
> C1:        file_ref = open(string_ref)
> 
> better than this?:
> 
> C2:        file = open(string)
> 

Definitely, but not for the reasons you state ;) Rather b/c with C2 you've 
just obliterated the new file builtin (synonym to open) and the string module 
name. So regardless of the IMHO needless distinction (in python, 'everything 
is a name which points to an object') between objects and references, I'd use 
C1 or something that doesn't use names of standard things like 'file' or 
'string' for my own variables.

Cheers,

f.


From jjb5 at cornell.edu  Wed Apr 10 19:20:49 2002
From: jjb5 at cornell.edu (Joel Bender)
Date: Wed, 10 Apr 2002 19:20:49 -0400
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <jjb5-22EDBB.19204910042002@newsstand.cit.cornell.edu>

I wrote:

> I would like to process the elements of a list that match a condition, 
> then remove them.

Outstanding answers and response, thank you!


Joel


From alwyn at alwyn.demon.co.uk  Thu Apr 18 12:42:03 2002
From: alwyn at alwyn.demon.co.uk (Alwyn)
Date: Thu, 18 Apr 2002 16:42:03 GMT
Subject: ranlib problem building 2.2.1 on Mac OS X
References: <m28z7l7hc4.fsf@javamonkey.com>
Message-ID: <alwyn-E3936E.17420018042002@news-text.blueyonder.co.uk>

In article <m28z7l7hc4.fsf at javamonkey.com>,
 Peter Seibel 
 <peter at localhost.i-did-not-set--mail-host-address--so-shoot-me> wrote:

> I'm trying to build Python 2.2.1 from the source tarball on Mac OS X.
> I ran ./configure and then make and the make dies with:
> 
> ranlib libpython2.2.a
> ranlib: warning: cannot infer operation mode from `libpython2.2.a'
> ranlib: you must specify a MODE
> Try `ranlib --help' for more information.
> make: *** [libpython2.2.a] Error 1

What version of Mac OS X is that? Are you sure you have the right 
Developer Tools for that version? 

I built Python 2.2.1 the other day from the standard distribution with 
no problems on OS X 10.1.3 with the December 2001 Developer Tools. 


Alwyn


From aahz at pythoncraft.com  Tue Apr  9 17:56:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 17:56:55 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <3CB2A0E9.1AE9DF80@kfunigraz.ac.at> <slrnab617e.97a.philh@comuno.freeserve.co.uk> <orGs8.5669$d7.1571810@newssrv26.news.prodigy.com>
Message-ID: <a8vo37$s8a$1@panix1.panix.com>

In article <orGs8.5669$d7.1571810 at newssrv26.news.prodigy.com>,
dsavitsk <dsavitsk at e-coli.net> wrote:
>"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
>news:slrnab617e.97a.philh at comuno.freeserve.co.uk...
>>
>> IMO, anyone for whom K&R isn't to their taste, probably won't make a
>> very good programmer anyway.
>
>Yow! IMO, anyone who thinks that only one sort of person (a sort who
>likes a particular book no less) makes a good programmer needs to relax
>a bit in their judgmentalness.

Well, Phil *should* have said "good C programmer", and with that
correction I completely agree with him.  There are some books that
manage to distill a quintessence, and K&R is one of that rare number.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From rxg218 at psu.edu  Thu Apr 11 16:10:17 2002
From: rxg218 at psu.edu (Rajarshi Guha)
Date: Thu, 11 Apr 2002 16:10:17 -0400
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com>
Message-ID: <a94qk5$22mg@r02n01.cac.psu.edu>

On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de Maria wrote:

> Is there a way to save in a file the status, objects, functions of
> interpreter
> in order to restore it after the interpreter has been closed?
> 
> If it is possible, one can work on a project stop and then continue
> later adding functions, objects and so on.

I have a related question - does the OP refer to object persistence or 
object serialization? Is there a difference between the two?

I looked at the shelve and pickle modules - the former uses a dbm database 
whereas the latter just dumps an ASCII representation of the object. The 
fact that shelve uses a dbm indicates that it should be more efficient. Is 
this true? When would I choose one over the other?

-- 
Rajarshi Guha
rajarshi at presidency.com


From goodger at users.sourceforge.net  Wed Apr  3 00:48:52 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Wed, 03 Apr 2002 05:48:52 GMT
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <mailman.1017794173.28885.python-list@python.org>
References: <mailman.1017794173.28885.python-list@python.org>
Message-ID: <B8D00218.20D57%goodger@users.sourceforge.net>

Paul Rubin wrote:
> Man, I thought it was an April fools joke when I saw it last night.

That's why I explicitly stated that it *wasn't* one. :-)

> All those `different` :kinds: 'of' "quoting" are much more confusing
> than explicit markup. So I think it's better to use explicit markup.

Single and double quotes are *not* significant in reStructuredText,
for the simple reason that they're commonly used in ordinary text.

The markup constructs used by reStructuredText were chosen from common
usage or designed with the goal of being "as simple and unobtrusive as
possible".  In reStructuredText, readability is a much more important
goal than writability.

For computers, explicit markup is definitely easier to deal with.
Having written a parser for implicit markup, I'm painfully aware of
this.  But for untrained humans, explicit markup is very difficult to
read -- all the markup is very distracting.

> The thing I like least about the PEP is that it's written like
> marketing material

PEPs *are* marketing tools.  They market proposals to the Python user
community and PythonLabs, lobbying for support.  They must be
persuasive, both technically and expositorially.

> --it sets out to push a particular solution and looks for things to
> bash about the alternatives rather than treating them evenhandedly
> and being willing to accept the idea that reStructuredText may not
> be the answer.

The PEP briefly examines alternatives proposed in the past, on the
Doc-SIG, in terms of "the generally accepted goals for a docstring
format".  I think you're reading too much into the prose as it stands.

This PEP toots its own horn.  Why should it toot somebody else's?

> Finally, among the explicit markup languages examined in the PEP,
> Texinfo should be included.

Texinfo is not included because it has never been seriously proposed
as a docstring format [#]_.  Looking at samples, I can see why.

.. [#] If you care to propose Texinfo as a standard docstring format,
   please join the Doc-SIG, make your case there, then write a PEP.

The entire premise of this PEP and of the Doc-SIG's (at least) 6-year
docstring markup odyssey is that Pythonistas want a markup that is
"readable in source form".  We're spoiled by Guido's programming
language design aesthetics, and that spilled over into the docstring
markup debate.  Better late than never.

-- 
David Goodger    goodger at users.sourceforge.net    Open-source projects:
 - Python Docstring Processing System: http://docstring.sourceforge.net
 - reStructuredText: http://structuredtext.sourceforge.net
 - The Go Tools Project: http://gotools.sourceforge.net



From guido at python.org  Mon Apr 15 21:48:56 2002
From: guido at python.org (Guido van Rossum)
Date: Tue, 16 Apr 2002 01:48:56 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <3CBB83B1.8EC2E2F6@python.org>

phil hunt wrote:

> IMO until Python developers adopt a policy of being very reluctant
> to do this, many IT managers will be wary of using it in anything
> other thsan toy projects, and Python's user base will not grow as
> quickly as (IMO) it should.

If in your opinion Python should grow faster, perhaps you could
help out by contributing rather than by telling us what to do...
(or proposing random language changes :-)

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


From SBrunning at trisystems.co.uk  Wed Apr 17 09:21:18 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 17 Apr 2002 14:21:18 +0100
Subject: shipping a python app
Message-ID: <31575A892FF6D1118F5800600846864DCBCFDB@intrepid>

> From:	roy at panix.com [SMTP:roy at panix.com]
> I've been given a task to write a small gui-based app which needs to
> run on both NT and Unix platforms.  About all I know about NT is how
> to spell it, and I've heard it's got some cool flight simulator games.
 
Have a look at Gordon McMillan's Installer
<http://www.mcmillan-inc.com/install1.html>.

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




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




From karsce at usa.net  Thu Apr  4 12:45:59 2002
From: karsce at usa.net (Subramaniam Karthik)
Date: Thu, 04 Apr 2002 10:45:59 -0700
Subject: Clipboard Monitor
Message-ID: <20020404174559.4216.qmail@uadvg137.cms.usa.net>

Hi, 
  I am writing a clipboard monitor for windows in python. my code looks
something like this.

class MyWindow:
    def __init__(self):
        message_map = {
            win32con.WM_DESTROY: self.OnDestroy,
            win32con.WM_DRAWCLIPBOARD: self.OnDrawClipboard,
            win32con.WM_CHANGECBCHAIN: self.OnChangeCBChain,
        }
        #   Register thw window class.
        wc = WNDCLASS()
        self.hinst = wc.hInstance = GetModuleHandle(None)
        . . .
        self.classAtom = RegisterClass(wc)      
        style = win32con.WS_OVERLAPPED | win32con.WS_SYSMENU
              style,0, 0, 150, 100,0, 0, self.hinst, None)
        self.hwnd = CreateWindow( self.classAtom, "Clipboard Monitor",  
              style,900, 50, 350, 200,0, 0, self.hinst, None)
        ShowWindow(self.hwnd,1)
        
        self.nextWnd = win32clipboard.SetClipboardViewer()

    def OnDestroy(self, hwnd, msg, wparam, lparam):
        win32clipboard.ChangeClipboardChain(self.hwnd, self.nextWnd)  
        DestroyWindow(self.hwnd)
        UnregisterClass(self.classAtom,GetModuleHandle
                             (None))                                
        PostQuitMessage(0)  #   Terminate the app.

    def OnChangeCBChain(self, hwnd, msg, wparam, lparam):
        SendMessage(self.nextWnd, msg, wparam, lparam)
        
    def OnDrawClipboard(self, hwnd, msg, wparam, lparam):
        SendMessage(self.nextWnd, msg, wparam, lparam)


def main():
    w = MyWindow()
    PumpMessages()

if __name__ == '__main__':
    main()

when i execute this program, i get the followinf error message.

File "C:\Program Files\karthik\Source files\test2.py", line 30, in __init__
    self.nextWnd = win32clipboard.SetClipboardViewer()
TypeError: SetClipboardViewer() takes exactly 1 argument (0 given)

Could anyone please tell me as to what i am doing wrong here?

Thanks

Karthik





From lenssen at hitnet.rwth-aachen.de  Tue Apr 16 16:46:50 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Tue, 16 Apr 2002 22:46:50 +0200
Subject: Safe eval?
Message-ID: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>

Is it possible to have a safe eval() on the server-side even when the string
consists of submitted user data?

Thanks.


I've been just discussing this in a PHP group, but now I want to port an
XML-based interpreter to Python too (see http://questml.com ), and the same
issues arise. Basically in PHP I didn't find a fail-safe solution; either
the configuration* is set to safe**, or it isn't, but lowering the
right-level for other scripts is not possible here***. Now, my current PHP
work-around is to replace certain characters and function names. That would
always be a solution in the end but it's not completely satisfying I think.

* on which most people like me don't have access on their hosted internet
server
** which would be globally OK for my needs, since I don't even want to write
to the file-system
*** it's possible in .NET from what I've seen.






From mgilfix at eecs.tufts.edu  Sat Apr 27 15:51:15 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sat, 27 Apr 2002 15:51:15 -0400
Subject: inline assignments in conditionals
In-Reply-To: <2143388891-BeMail@sugoi>; from jaf@lcsaudio.com on Sat, Apr 27, 2002 at 12:33:47PM -0700
References: <2143388891-BeMail@sugoi>
Message-ID: <20020427155114.H1588@eecs.tufts.edu>

On Sat, Apr 27 @ 12:33, jaf wrote:
> Writing things that way is really unnatural, at least for English
> speakers.  It makes the intent of the code harder to understand (at
> least for me).  For example, if you were talking to a friend, you
> woudn't say:
> 
>    "If today is Sunday, I'm going to go shopping"
> 
> but rather:
> 
>    "If Sunday is today, I'm going to go shopping"
> 
> and in any case, writing the constant first only protects in the case
> where one of the compared values is a constant.  If you are doing
> 
>    if (a == b):
> 
> then there is no way to protect yourself against actually writing '='.
> 
> Given that, I think not supportig inline assignment is a reasonable
> choice.  Perhaps another option would be to coin a synonymous operator
> that works in comparisons, but is distinct enough from '==' so as not
> to be used by accident?
> 
>    if (data gets socket.recv(1024)):
>       process(data)

  I'm +1 on this (Guido's gonna hate me I bet!). I recently found a
scenario (this was in C and then I thought about how I'd do it in
python) where I wanted to do an assignment only if I reached a certain
point in a bool statement, including short-circuiting.  It just made
my life a lot easier because I didn't have to include redundant
case-code. B-sides, sometimes, you wish you could compress the two
statements.

               -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From theller at python.net  Mon Apr 22 14:39:58 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 22 Apr 2002 20:39:58 +0200
Subject: Overriding methods in classes that use property()
References: <mailman.1019488863.14978.python-list@python.org>
Message-ID: <aa1le8$72vm7$1@ID-59885.news.dfncis.de>

"Edmund Lian" <elian at inbrief.net> wrote in message news:mailman.1019488863.14978.python-list at python.org...
> Hi,
>
> When I run the code appended below, class B inherits all the methods of
> class A, including the property() definition as I would expect.
> Unfortunately, class C also does the same; I cannot override class A's
> setVal method. Is this a bug, or do I have to do something different (and
> unexpected) with Python 2.2's new style classes?
>
> ...Edmund.
>
> class A(object):
>     def __init__(self):
>         self._val = None
>
>     def setVal(self, value):
>         self._val = value
>
>     def getVal (self):
>         return self._val
>
>     val = property(getVal, setVal)
>
> class B(A):
>     pass
>
> class C(A):
>     def setVal(self, value):
>         self._val = 'Huh?'
>
You can achive what you want with a metaclass, see code below.
Every time a new subclass of A is created, the __metaclass__
__init__ method is called, which retrieves the correct
getVal and setVal methods, and creates a property from them:

class A(object):
    class __metaclass__(type):
        def __init__(cls, name, bases, dict):
            cls.val = property(cls.getVal, cls.setVal)

    def __init__(self):
        self._val = None

    def setVal(self, value):
        self._val = value

    def getVal (self):
        return self._val

class B(A):
    pass

class C(A):
    def setVal(self, value):
        self._val = 'Huh?'


Regards,

Thomaas




From mhammond at skippinet.com.au  Sun Apr 28 19:45:41 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 28 Apr 2002 23:45:41 GMT
Subject: GetOpenFilename from python extension
References: <tczy8.79369$HH5.5211@rwcrnsc51.ops.asp.att.net>
Message-ID: <3CCC89BE.5010700@skippinet.com.au>

neal josephson wrote:
> I have a pretty simple python program that I want to add a windows file
> dialog box to - I tried writing a python extension that called
> GetOpenFilename but it just returns a DLG_INITIALIZE (or something like
> that) error.  Has anyone managed to get this to work?
> 

It can be done with win32all.

 >>> import win32ui
 >>> d=win32ui.CreateFileDialog(1)
 >>> d.DoModal()
1
 >>> d.GetPathName()
'G:\\src\\moz_play\\mozilla\\extensions\\python\\xpcom\\server\\loader.py'
 >>>

RTFM for more info on the params etc.

Mark.



From donnal at donnal.net  Mon Apr 22 09:55:43 2002
From: donnal at donnal.net (Donnal Walter)
Date: 22 Apr 2002 06:55:43 -0700
Subject: Programming metaphors (Was: Programming as literature.)
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <3CC3A2E0.B1AA0F70@tundraware.com> <Xns91F7EED51FA51lumberjack@216.148.53.83> <slrnac7qns.dh0.jajvirta@sirppi.helsinki.fi>
Message-ID: <918bc22f.0204220555.3f8456d@posting.google.com>

jajvirta at cc.helsinki.fi (Jarno J Virtanen) wrote in message news:<slrnac7qns.dh0.jajvirta at sirppi.helsinki.fi>...
> 22 Apr 2002 06:29:05 GMT Lumberjack wrote:
> > Tim Daneliuk <tundra at tundraware.com> wrote:
> >> Carl Banks wrote:
> >>> Computer programming is not science.
> >> 
> >> Agreed - it is craft.
> > 
> > Computer programming has all the properties of literature, including 
> > subjective analysis, the existence of "schools of thought," and genres, 
> > among other things.

> ... Software development is more of a 
> maintenance activity where bugs and shortcomings are fixed throughout
> the lifetime of a product. (Taken to the extreme, literally, it is
> actually only about fixing bugs and shortcomings (ie. adding features)
> which starts as soon as the initial release is made which is made
> very early in case of extreme programming.)
> 
> I find every other metaphor lacking something essential too.  One
> popular metaphor is building construction or engineering activity in
> general.  
>
> <snip>
>
> whether it's just that software development is such a young discipline
> or that there's some inherent factor that makes it uncertain and 
> undeterministic is beoynd my comprehension. 
> 
> Having said all that I'll just have to remind that metaphors are of
> course useful if they shed light on some aspect on the subject.  They
> just shouldn't be extended too far.  (Hmh, It seems that I've been
> stating the obvious all the time, well, that's just me ;-)

This is not a perfect metaphor either, but the one I like is
biological evolution. Organisms are made up of organs, which are made
up of cells, which are made up of proteins and other molecules. Cells
send messages to each other in the form of hormones, transmitters,
substrates, cofactors and so on. Cells exhibit encapsulation,
instantiation, inheritance, polymorphism, ... If a given protein or a
given cell type or a given species doesn't work properly in a
particular environment it often gets replaced by one that works
better. This is a bit more like the maintenance activity and bug fixes
you talked about.

Programming is considerably more deliberate than evolution, of course,
but evolution also captures some of the uncertainties and
non-deterministic aspects you mentioned as well.

Donnal Walter, MD, PhD
Arkansas Children's Hospital
donnal at donnal.net


From aleax at aleax.it  Thu Apr 11 11:55:25 2002
From: aleax at aleax.it (Alex Martelli)
Date: Thu, 11 Apr 2002 15:55:25 GMT
Subject: Readline on Window
References: <3CB5A6F9.2070205@thomas-guettler.de>
Message-ID: <Nlit8.7096$b62.171903@news1.tin.it>

Thomas Guettler wrote:

> I use:
> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> 
> C-a results in ^A. But I would like to go the beginning of the line

http://iacsolutions.com/projects/readline.html


Alex



From ianb at colorstudy.com  Mon Apr 15 17:13:22 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 15 Apr 2002 16:13:22 -0500
Subject: String interpolation question
In-Reply-To: <200204151450.22332.fperez528@yahoo.com>
References: <a9fcmo$npi$1@peabody.colorado.edu>
	<1018903571.1147.395.camel@localhost> 
	<200204151450.22332.fperez528@yahoo.com>
Message-ID: <1018905202.1147.440.camel@localhost>

On Mon, 2002-04-15 at 15:50, Fernando P?rez wrote:
> On Monday 15 April 2002 14:46, you wrote:
> > On Mon, 2002-04-15 at 15:19, Fernando P?rez wrote:
> > > but I don't want that because in my real problem, the string is long and
> > > I don't want the dangers of positionally matched %s arguments.
> >
> > Then do
> >   print somestring.replace("%", "%%") % y[i]
> >
> >   Ian
> 
> Sorry but I'm not sure I get your solution. If my print string is something 
> like
> 
> s='Lots of other stuff interpolated.... y[i]'
> 
> so that I need to make a call like
> 
> print s % locals()
> 
> how does your solution work?

You seemed to be concerned that there would be other %s's in the string,
so you didn't want to do positional substitution.  Actually, no that
doesn't work my way either, does it? :)  I wasn't thinking hard
enough.   Now I must copy the list to redeem my solution, or at least
disavow my last bit of stupidity :)
 
You could use another marker, like $$value, and do:

somestring.replace('%', '%%').replace('$$value', '%s') % y[i]

That's something to write in everywhere, no? :)

There has to be *some* way to delimit the position that needs to be
substituted -- whoever writes that string must be aware of that, no
matter if you are using string interpolation, % substitution, or a
complex templating language.

You could do something like:

templateRE = re.compile(r'\$(\w+)')
def subber(s, **kw):
    return templateRE.sub(lambda m, kw=kw: str(kw[m.group(1)]), s)

Then do:
print subber('The value is $yi', yi=y[i])

You could use $$ instead of $, or SUBTHIS[...], etc., depending on the
regular expression.  Being able to do $y[$i] is at least an order of
magnitude more programming.  Well, considering this solution is three
lines, it's probably two orders of magnitude harder.  And even then it
won't work right for everyone, as you found with Itpl.

  Ian






From merman at freenet.de  Sat Apr 13 14:21:59 2002
From: merman at freenet.de (merman)
Date: Sat, 13 Apr 2002 20:21:59 +0200
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net> <3cb843d0$0$4068$9b622d9e@news.freenet.de> <82bscn4nvz.fsf@acropolis.localdomain> <7xr8lja3l7.fsf@ruckus.brouhaha.com>
Message-ID: <3cb87520$0$140$9b622d9e@news.freenet.de>

"Paul Rubin" <phr-n2002a at nightsong.com> schrieb im Newsbeitrag
news:7xr8lja3l7.fsf at ruckus.brouhaha.com...
> Patrick W <xnexau at yahoo.com.au> writes:
> > Do you have a technical reason to actually want interfaces, or are you
> > simply experimenting with the concept?

> > As far as I can see, which is admittedly not that far, interfaces
> > don't really serve a useful purpose in Python. If you want
> > polymorphism in statically typed OO languages like C++ and Java, you
> > need to design your functions and methods to accept only objects
> > descended from a known type. But in Python, the type need not
> > (actually cannot) be specified at design time, so what purpose would
> > an interface serve? Wouldn't it be better to just inherit
> > _implementation_ when necessary, rather than to inherit pure
> > interfaces?
> >
> > I ask this not to be a prick, but to find out if I'm missing something
> > important.
>
> The purposes of interfaces in java is mostly to provide a form of multiple
> inheritance.  Java doesn't support multiple base classes, so interfaces
> are a workaround for that.  Python supports multiple base classes directly
> so it doesn't need a workaround.

Thanx for answers and sorry for my strange english ;-).

Note I'm a newbie:

1.) Yes, I want experimenting with the concept of course.

2.) Another reason is to accomplish a certain behaviour in the derived
class/ es.
I thought I can realize this goal either with abstract methods or
interfaces.

Is that an false idea?

o-o

Thomas










From loewis at informatik.hu-berlin.de  Mon Apr 15 07:21:12 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Apr 2002 13:21:12 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de> <3cba0944@news.mhogaming.com> <a9dbc9$bge$0@216.39.172.122> <eppstein-BBF667.20005414042002@news.service.uci.edu> <m38z7pmq4p.fsf@mira.informatik.hu-berlin.de> <eppstein-18C460.23560914042002@news.service.uci.edu>
Message-ID: <j43cxxnr6f.fsf@informatik.hu-berlin.de>

David Eppstein <eppstein at ics.uci.edu> writes:

> Is there some way of getting the arithmetic meaning now, without a 
> warning, using a __future__ import?

Not with a future import. You can convert all numbers to longs
explicitly; that will work in all Python versions.

Regards,
Martin


From JamesL at Lugoj.com  Sun Apr 14 03:15:17 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 14 Apr 2002 07:15:17 GMT
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org>
Message-ID: <Xns91F0276A4EF3JamesLLugojcom@216.148.53.85>

Okay, one more post to issue a mea culpa, among other things:

Tim Peters <tim.one at comcast.net> wrote:
> [Paul Rubin]
>> In that particular debate, a lot of people were asking for certain new
>> features to be LEFT OUT of Python. I don't understand how someone can
>> volunteer to contribute work, time, or money toward leaving something
>> out. 
> 
> That was barely mentioned in the "crushing amount of debate over
> 'stability' [that] has gone by on Python-Dev the last week", so I don't
> know what "that particular debate" means to you, but it surely doesn't
> refer to anything in the paragraph you were responding to.

Oops. I also misread Tim's paragraph, hence my other post in this thread 
that makes the same conceptual mistake. Sorry.

>  The debate
> on Python-Dev has about backward compatibility, and especially about
> producing bugfix releases for older releases for a much longer time. 
> Those things take work, and lots of it.  Note that James (to whom I was
> replying) was making the case for why commercial support is important
> to businesses -- nobody is going to buy a support contract that only
> guarantees not to add new features.

Just FYI: This has not been my experience with the motivations I've seen or 
why I've recommended for or against buying support contracts. Many firms 
I've worked with get support contracts to only get bug fixes _and_ priority 
response when problems happen. (E.g. Database systems like Oracle. 
Applications are designed assuming the existing feature set and behavior 
and it is nontrivial to rewrite a mission critical app to take advantage of 
some new feature.) Whether the support contract includes delivery of new 
releases was rarely a consideration. We never bought tools because of their 
future potential, we bought them because we thought they were relatively 
stable and would solve our problems then and there.

>  Quite the contrary, business wants
> bug fixes for older releases, and they *also* want new features in old
> releases, especially in the libraries (like, e.g., SSL support for
> Windows, and support for hot new protocols -- this is a fast-moving
> field, if you haven't noticed <wink>). 

Well, you know business requirements: the product should fit in a shirt 
pocket, have a 23 inch screen, run on a 9V battery for 3 years, and be 
capable of towing your car. ;-) This is why I now always demand a 
_prioritized_ list of requirements from marketing (and they can't all have 
the same priority!)


From sholden at holdenweb.com  Tue Apr 23 12:57:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 12:57:35 -0400
Subject: file_ref = open(string_ref)
References: <1019579845.565223@cswreg.cos.agilent.com>
Message-ID: <jqgx8.110$_i3.3@atlpnn01.usenetserver.com>

"Greg Weeks" <weeks at vitus.scs.agilent.com> wrote ...
> Is this code
>
> C1: file_ref = open(string_ref)
>
> better than this?:
>
> C2: file = open(string)
>
Well, let's first agree that all such matters are subjective and that beauty
is in the eye of the beholder, so there's room for disagreement. As long as
the disagreement is honest and respectful I'm happy to play ...

> C1 is more accurate, given the view that data in Python are references to
> objects.  According to this view, the left-hand variable names a reference
> to a file object.
>
"Data"? Really here we mean things like names, dictionary values and list
elements, right?

Because *all* names are references to objects, it seems reasonable to prefer
C2 to C1, in my (pragmatist's) opinion.

> I can't believe that C1 is superior to C2 or that the millions of lines of
> Python, Lisp, and Java code similar to C2 are in any way wrong.  So, in my
> opinion, our view of data is wrong.  To align our view of data to our
code,
> we need to replace the view
>
> V1: Objects are regions of memory.
>
> with:
>
> V2: Objects are addresses (of regions of memory).
>
Aaarrrggghhh!  No!!!!!!!!! Objects *are* regions of memory. Binding (e.g.
assignment, function declaration, et al) doesn't store an object, it stores
*a reference* to an object.

> If we adopt V2, then C2 (and all the code like it) is correct.  file is an
> object, string is an object, and even open is an object.  If we prefer V1,
> on the other hand, then we must consider C1 more correct.
>
Not the view I would take of correctness.

> The problem with V1 is not simply a mismatch between our concepts and our
> Lisp, Java, and Python code.  The sad fact is that the majority of C code
> that I've seen resembles C1.  This code is just as ugly and unnecessary in
> C as it would be in Python.
>
The difference is that in C, variables hold values. Assignment actually
stores an expression's value in a location. Some of the expressions are
explicitly cast as references, in which case the variable has to be
acknowledged as holding a reference.

That's (mostly) because C does not necessarily perform Python's implicit
dereferencing. In Python, when we use a name (said Humpty...) we accept
implicitly that we are referring to an object. In C, an experession might
evaluate to a type, or to a reference to a type (and, since references are
also type, references to references to ...).

There's no way in Python to store anything *but* references to objects, and
indeed the Python garbage collector actually collects things only when their
reference count reduces to zero.

> So, every now and then, I raise the "Objects are addresses" flag.
>
Don't. They aren't ;-). They are (dare I say it) *referred to* by their
addresses.

>
> Greg
>
>
> PS: Note that I don't proclaim "Objects are references".  The term
> "reference" is ambiguous and heavy with misleading connotation.
>
What else is an address but a reference? Objects aren't references, but we
use references to locate them.

> PS: In "Objects are addresses", "Objects" does not include the numerical
> values of Java or the numerical or structure values of C.  If this is a
> problem, then V1 has the same problem.
>
You have to say this because of a failure to distinguish beteen values and
references. In Java you have to discriminate carefully between, for example,
integers (which aren't objects) and Integers (which are). In Python, all
values are objects, so it doesn't matter as much.

> PS: What about C++.  Ucch, the worst.  C++ extravagently supports V1 by
> allowing "initialization by reference", whereby a variable is initialized
> to the same region of memory as another variable.  The result is
> considerably more complicated than Python, Lisp, and Java.  And it doesn't
> quite hang together, since there is no assignment by reference.

I'll leave this to Alex Martelli, who knows C++ fat better than I.

regards
 Steve
--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From maxm at mxm.dk  Thu Apr  4 04:01:25 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 04 Apr 2002 11:01:25 +0200
Subject: Absolute to relative URL?
References: <3CABFB6C.7050308@thomas-guettler.de>
Message-ID: <3CAC1665.8040209@mxm.dk>

Thomas Guettler wrote:


> Is there already a function in the standard python modules
> which returns the a relative url given two absolute urls?
> 
> Example:
> 
> relative_url("http://foo/a/b/c", "http://foo/a/d")
>  --> return: "../d"

I hope you mean that the result should be "../../d" ??

A function that generates a relative path given two absolute urls?

Not that I am aware of, but this problem was more fun than what I really 
should be working on today :-/, so here is an attempt:

regards Max M

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

from urlparse import urlparse, urljoin


def relativePath(url1,url2):
     "Returns relative path from url1 to url1"
     # get just path parts
     u1Parts = urlparse(url1)[2].split('/')[1:]
     u2Parts = urlparse(url2)[2].split('/')[1:]
     u1Len = len(u1Parts)
     u2Len = len(u2Parts)

##    # uncomment if simillar paths should return '.'
##    if u1Parts == u2Parts:
##        return '.' # that was easy

     diffStart = 0 # from where do the paths differ?
     for i in range(min(u1Len, u2Len)):
         if u1Parts[i] == u2Parts[i]:
             diffStart += 1

     result = []
     nDirsUp = len(u1Parts[diffStart:])
     result += ['..']*nDirsUp + u2Parts[diffStart:]
     return '/'.join(result)



url1 = "http://foo/a/b/c"
url2 = "http://foo/a/b/c"
print "Excpect: ''"
print repr(relativePath(url1,url2))
print ""

url1 = "http://foo/a/b/c"
url2 = "http://foo/a/d"
print "Excpect: '../../d'"
print repr(relativePath(url1,url2))
print ""

url1 = "http://foo/a/b/c"
url2 = "http://foo/a/b"
print "Excpect: '..'"
print repr(relativePath(url1,url2))
print ""

url1 = "http://foo/a/d"
url2 = "http://foo/a/b/c"
print "Excpect: '../b/c'"
print repr(relativePath(url1,url2))
print ""

url1 = "http://foo/a/b"
url2 = "http://foo/a/b/c"
print "Excpect: 'c'"
print repr(relativePath(url1,url2))
print ""


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

Excpect: ''
''

Excpect: '../../d'
'../../d'

Excpect: '..'
'..'

Excpect: '../b/c'
'../b/c'

Excpect: 'c'
'c'



From duncan at NOSPAMrcp.co.uk  Mon Apr 29 09:33:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 29 Apr 2002 13:33:31 +0000 (UTC)
Subject: Comment on draft PEP for deprecating six builtins
References: <aaim48$7fd$1@bob.news.rcn.net> <Hd6z8.74979$vF6.2240539@news2.tin.it> <mailman.1020083690.16055.python-list@python.org>
Message-ID: <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1>

Roman Neuhauser <neuhauser at mail.cz> wrote in 
news:mailman.1020083690.16055.python-list at python.org:

>     __builtin__.chr(int) -> int.chr()
>     __builtin__.abs(int) -> int.abs()
>     __builtin__.ord(str) -> str.ord()
>     etc
> 

The problem with int.chr() is that a lot of people are going to get 
confused when they find that "42.chr()" doesn't work (although "42 .chr()" 
does).

42 .__abs__() already works. :-)
-42 .__abs__() also works, but again the result might surprise some people 
if they expected it to be the same as int.__abs__(-42)

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From aleax at aleax.it  Tue Apr 16 06:30:48 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 16 Apr 2002 10:30:48 GMT
Subject: A little advice please? (Convert my boss to Python)
References: <a9fi0s$c9c$1@newsg1.svr.pol.co.uk> <7x662s5b8t.fsf@ruckus.brouhaha.com>
Message-ID: <s3Tu8.25135$vF6.807815@news2.tin.it>

Paul Rubin wrote:

> "Duncan Smith" <buzzard at urubu.freeserve.co.uk> writes:
>> So what I'm looking for is speed, and some advice so that I don't end up
>> trying too many alternatives.
> 
> If you have to do something like that over and over for zillions of
> huge files, you're best off writing in C and tuning carefully.

Not necessarily.  Python dictionaries are pretty amazing.  Duplicating
their functionality and speed is not just a question of "tuning
carefully".


> Regarding duplicates, maybe you can just sort the file with an
> external sort utility, so the duplicates will all be next to each
> other.  Then you don't have to mess with dicts.  I didn't examine your
> code closely enough to figure out if that makes sense, so maybe it
> doesn't.

Sorting is O(N log N).  Inserting N entries in a dictionary can be
pretty close to O(N), since entry insertion is darn close to an
amortized O(1).  Therefore, it's anything but obvious that sorting
should be a performance win.


Alex



From loewis at informatik.hu-berlin.de  Thu Apr 18 12:37:26 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 18:37:26 +0200
Subject: Python-2.2.1, Solaris7, make test fails...
In-Reply-To: <Pine.GSO.4.40.0204181557541.24683-100000@neelix>
References: <Pine.GSO.4.40.0204181557541.24683-100000@neelix>
Message-ID: <j4u1q955fd.fsf@informatik.hu-berlin.de>

Hugh Sasse Staff Elec Eng <hgs at dmu.ac.uk> writes:

>         [*shared*, *static* intereaved]
> 
> But to change the wording one needs to know which of the above
> properties are reversed, and how commenting out modules affects them in
> each case.

I'm not sure what your question is. The build process of Python works
that way:

1. Build all modules that are declared as static in Modules/Setup,
   combine them into libpythonxy.a, combine that into python.
2. Build all modules that are listed as shared in Modules/Setup.
3. Invoke setup.py. That builds all modules that
   a) are not builtin, and
   b) are not listed in Modules/Setup
   c) can be build on the target

So the property that is being reversed is 'build as static'.

> So the fastest fix is for me to doengrade binutils?

No. The fastest fix is to apply the patch below, and rerun configure.

Regards,
Martin

Index: configure
===================================================================
RCS file: /cvsroot/python/python/dist/src/configure,v
retrieving revision 1.279.6.4
retrieving revision 1.279.6.5
diff -u -r1.279.6.4 -r1.279.6.5
--- configure	22 Mar 2002 11:22:24 -0000	1.279.6.4
+++ configure	18 Apr 2002 14:55:05 -0000	1.279.6.5
@@ -3277,7 +3277,7 @@
 		fi;;
 	SunOS/5*) case $CC in
 		  *gcc*)
-		    if $CC -Xlinker -V 2>&1 | grep BFD >/dev/null
+		    if $CC -Xlinker --help 2>&1 | grep export-dynamic >/dev/null
 		    then
 			LINKFORSHARED="-Xlinker --export-dynamic"
 		    fi;;




From neal at metaslash.com  Fri Apr 19 10:17:39 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 19 Apr 2002 10:17:39 -0400
Subject: Python-2.2.1, Solaris7, make test fails...
References: <mailman.1019220338.385.python-list@python.org>
Message-ID: <3CC02703.DB2E7D8D@metaslash.com>

Hugh Sasse Staff Elec Eng wrote:
> 
> Now the build gives:

	[compiles & links]

> case $MAKEFLAGS in \
> *-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q build;; \
> *) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py build;; \
> esac
> ld.so.1: ./python: fatal: libtk8.3.so: open failed: No such file or directory
> 
> But

	[tcl/tk 8.3 is in /usr/local/lib]

Does LD_LIBRARY_PATH have /usr/local/lib in it?

If you do this:

	LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH
	export LD_LIBRARY_PATH

	(or use setenv ... for csh)

Does that work?

Neal


From steve.menard at polyester.com  Mon Apr 29 17:00:24 2002
From: steve.menard at polyester.com (Steve Menard)
Date: Mon, 29 Apr 2002 17:00:24 -0400
Subject: PyGTk and GTK+ 2.0 on win32
Message-ID: <Zpiz8.4518$1k5.841669@news20.bellglobal.com>

Hey there, I really hope someone can help me.

I would like to use Python with GTK 2.0. I have found a binary of PyGTK for
win32 at http://www.pcpm.ucl.ac.be/~gustin/win32_ports/, but looking inside
I can only find the binaries for GTK 1.3. I have also found the win32
binaries of GTK, but I am unsure if just switching the dll will work, since
the DLL names have changed.

Is there even a version of PyGTK which works against GTK 2.0?

Thanks in advance ....
                                          Steve




From cjw at sympatico.ca  Wed Apr 24 09:44:48 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Wed, 24 Apr 2002 09:44:48 -0400
Subject: Pyrex problems
Message-ID: <3CC6B6D0.F384850B@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020424/94ec0a21/attachment.html>

From phr-n2002a at nightsong.com  Wed Apr  3 04:46:25 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 03 Apr 2002 01:46:25 -0800
Subject: Java to Python translation
References: <3CAAC372.792A0F9A@sympatico.ca> <j44rit16wm.fsf@informatik.hu-berlin.de>
Message-ID: <7xbsd15d3i.fsf@ruckus.brouhaha.com>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
> > I need to translate a (rather voluminous 100MB) java program to
> > python. After a few searches on google, apparently the best tools
> > for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> > on the Lex-YACC HOWTO, I would be very interested in hearing of
> > others' successful experiences/feedback at this. Any pointers of
> > interest would similarily be appreciated.
> 
> Not that I can offer any inside - but I think many people would be
> interested to hear your results; whether the project fails or
> succeeds. I'm also sure some people might be interested in the
> converter itself. Please do consider publishing it, atleast if you
> have to give up, so that other might be able to continue.

I think for a project like that, even a success would be disastrous.
The resulting tools would be interesting to have, though.


From gashton at cmedltd.com  Tue Apr 23 11:42:01 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 23 Apr 2002 16:42:01 +0100
Subject: PEP 276 -- What else could iter(5) mean?
In-Reply-To: <3CC57670.9060308@earthlink.net>
References: 	<mailman.1015455997.18295.python-list@python.org><6qpu2gzktu.fsf@abnoba.inte
	 vation.de> <mailman.1015517623.2914.python-list@python.org>
	<u8g1dbkklg9i18@news.supernews.com>  <3CC57670.9060308@earthlink.net>
Message-ID: <1019576522.4104.3.camel@tp6>

On Tue, 2002-04-23 at 15:57, Charles Hixson wrote:

> And to me the meaning *is* blindingly obvious ... and different!
> 
> to me 5.iter should return an unlimitedly long string of 5's.

Not 0..5 then?

;)

-- 
Graham Ashton





From guido at python.org  Wed Apr 10 20:54:50 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 11 Apr 2002 00:54:50 GMT
Subject: 2.2 descriptor API
References: <410890c4.0204040806.363f5728@posting.google.com> <3CADF520.C0264C0B@python.org> <410890c4.0204080751.715ceed@posting.google.com>
Message-ID: <3CB4DF9B.143ADCDA@python.org>

[Daniel Popowich]
> > > 1.  Why is the __get__ method called on a descriptor when it is
> > >     referenced as a class attribute?

[Guido]
> > This was a requirement.[...]

[Daniel again]
> My questions had more to do with design than technical facts.  It
> strikes me that the __get__ method is not so much the publication of a
> new API, but rather the exposure of low-level, run-time interpreter
> mechanics.

Sure, but it's a very useful piece of mechanics.  E.g. in Zope 3,
new descriptor classes implementing custom __get__ methods are
under development that do things that Zope's master architect likes
to do.

It's no different than __getattr__ or __add__: it allows you to
write a class that does certain kinds of useful magic just like
Python's built-in classes do.

> But I question the utility of this for Joe Schmoe, python programmer.

Of course, if you don't need it, don't use it.  That's no reason
not to have it available in the language.

> If instead of using the descriptor api, the language had constructs
> like the following to define methods:
> 
>     class Foo:
>        # static method definition
>        def static foo(*args): pass
> 
>        # class method definition
>        def class bar(cls, *args): pass
> 
>        # instance method definition
>        def baz(self, *args): pass

When I added static and class methods and properties to Python 2.2,
I promised myself (and others) that I would eventually make up special
syntax.  But in the mean time, I implemented the underlying mechanism
and used it to provide provisional ways to spell the new features.

(The mechanism was easily decided on.  The new syntactic sugar
has to be carefully considered.)

> in other words, after "def", allow a qualifier (or perhaps a list of
> qualifiers) before the function name, then the interpreter can mark
> the functions appropriately and do the right magic when the
> class/instance attribute is accessed (*).

This and many different proposals have been discussed before.
Unfortunately I don't have the reference at hand; maybe some
other reader can help you out.

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


From ture at lysator.liu.se  Tue Apr 30 14:13:04 2002
From: ture at lysator.liu.se (Ture =?iso-8859-1?q?P=E5lsson?=)
Date: 30 Apr 2002 20:13:04 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <mailman.1020119995.30995.python-list@python.org> <slrnacrltr.6mq.jon+usenet@snowy.squish.net> <23891c90.0204300033.16549c52@posting.google.com> <3CCEAE2B.8000300@gmx.net> <slrnactf3d.lf5.jon+usenet@snowy.squish.net>
Message-ID: <upd6whnjgv.fsf@svartsmurf.lysator.liu.se>

(Following-up to a random article from the thread)

I suppose the W3C-me-harder way of doing this must be writing the
pages using some XML DTD conjured up to suit the problem at hand, and
converting them to (X)HTML using an XSLT stylesheet with extension
functions and/or elements written in Python for the things that cannot
be handled by XSLT. Has anybody actually done this? Is there any nice
software out there that could do it (i.e, an XSLT implementation with
hooks for writing extensions in Python)?

  -- T


From gh_pythonlist at gmx.de  Thu Apr  4 22:49:30 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Fri, 5 Apr 2002 05:49:30 +0200
Subject: Selecting cipher suites with socket.ssl
In-Reply-To: <84cff98e.0204041830.7c0d0171@posting.google.com>
References: <84cff98e.0204041830.7c0d0171@posting.google.com>
Message-ID: <20020405034930.GA15898@lilith.my-fqdn.de>

* Derek Thomson <derek at wedgetail.com> [2002-04-04 18:30 -0800]:
> Hi,
> 
> I've been using the socket.ssl method to create client-side SSL
> sockets. It works very nicely!
> 
> However, I'd like to know how I can specify which SSL cipher suites I
> allow i.e. the cipher suites that will be sent to the server in the
> cipher_suites field of the SSL ClientHello message.
> 
> I can't seem to find any documentation on this.
 
It's not possible. For everything but the most simple client-side SSL, I
recommend you use a different SSL implementation for Python. m2crypto
seems to be quite mature, but there are others, too:

http://www.informatik.fh-muenchen.de/~ifw00065/pyssl/index.html

If you're serious about SSL, I even recommend to avoid the built-in SSL
support in Python at all for now. It doesn't even actually check the
certificates, even if you supply them.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 3.2 ?C      Wind: 2.4 m/s




From theller at python.net  Fri Apr 26 18:26:42 2002
From: theller at python.net (Thomas Heller)
Date: Sat, 27 Apr 2002 00:26:42 +0200
Subject: [Crew] PSA revival - kind of?
References: <3CC9AA49.7050907@tismer.com> <mailman.1019850527.6618.python-list@python.org> <3CC9CBFE.527E8C28@python.org>
Message-ID: <aack7o$9hbet$1@ID-59885.news.dfncis.de>

"Guido van Rossum" <guido at python.org> wrote in message news:3CC9CBFE.527E8C28 at python.org...
> Thomas Heller wrote:
> > Do anything you want with my money (within resonable
> > bounds), but don't make me a member of another
> > organization. Please ;-)
>
> You can send your donation to the PSF.  Soon this will be tax-deductible
> (for US citizens).  http://www.python.org/psf/  Membership is free
> but only on invitation; but anyone can donate, much in the PSA spirit.
> We can talk about adding a starship account in the deal; Py subscriptions
> are not in the cards though.
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)

What are you talking about?
I'm already a PSF member, also I have a starship account.

This leaves the donation issue - email me privately about how
much money you need <wink>.

Thomas




From cliechti at gmx.net  Wed Apr  3 17:04:24 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 23:04:24 +0100
Subject: PyException
References: <1a27883b.0204031201.79639b2f@posting.google.com>
Message-ID: <Xns91E5EAE4DFBC3cliechtigmxnet@62.2.16.82>

ktrinh at allmerica.com (Khanh Trinh) wrote in 
news:1a27883b.0204031201.79639b2f at posting.google.com:

> The package was purchased from a vendor and they couldn't find the
> answer for this at all.
 
> Message: 
> Internal Error
> Traceback (innermost last):
>   File "<string>", line 18, in ?
> ImportError: No module named iclick
 
[very long traceback removed...]

have you checked the jython and java search path? it seems that jython does 
not find iclick (.jar or whatever) despite jython is started from within 
that module. try adding the directory where iclick.xxx resides to either 
the CLASSPATH environment variable if its a java extension or to PYTHONPATH 
if its a .py.

chris

-- 
Chris <cliechti at gmx.net>



From brian at sweetapp.com  Tue Apr 16 17:27:23 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Tue, 16 Apr 2002 14:27:23 -0700
Subject: Why don't strings share data in Python?
In-Reply-To: <a9i2cp$52m$1@watserv3.uwaterloo.ca>
Message-ID: <009b01c1e58d$7fb98a60$445d4540@Dell2>

Joe wrote:
> But, yeah, it'd be a little complicated.

Yep. I actually started to write a PEP about lazy slicing but decided
not to finish because it would very likely be rejected due to the
complexity of the implementation. Actually, I did get part way through
an implementation and it involved touching a lot of code.

PEP: XXX
Title: Lazy Slices
Version: $Revision: 1.0 $
Last-Modified: $Date: 2002/2/28 10:00:00 $
Author: brian at sweetapp.com (Brian Quinlan)
Status: Draft
Type: Standards Track
Created: 28-Feb-2002
Python-Version: 2.3
Post-History:


Abstract

    This PEP describes an approach which would improve the CPU and
    memory efficiency of generating large string and Unicode objects
    from slices.


Rationale

    Slicing is a powerful and commonly used Python idiom. However, it
    is not always efficient to use because the time needed to 
    generate a slice is proportional to the length of the generated
    object. This can force the programmer to use more complicated
    techiques to avoid the expensive duplication of the existing
    sequence data.
    
    Low-level languages use address arithmatic to manipulate sub-
    sequences in constant time e.g. 
    
        big_str = "Names:\nGuido\nTim\n..."
        if str[6:].find("Ame") != -1): ...

        Could be expressed as follows in C:

        char * big_str = "Names:\nGuido\nTim\n...";
        if (strstr(str + 6, "Ame") != 0) ...

    The Python implementation for string and Unicode objects should
    offer the same constant time efficiency when generating slices.


Implementation


    The most obvious implementation would involve changing the
    PyStringObject and PyUnicodeObject definitions to include a union
    similar to the following:    
    
        struct t_string {
            PyObject *ob_sinterned;
            int ob_size;
            char ob_sval[1];
        };
        
        struct t_slice {
            PyObject *ob_sparent;
            /* start index of the slice */
            int ob_start;
        };
        
        typedef struct {
            PyObject_HEAD
            long ob_shash;
            /* -1 indicates t_string, otherwise it indicates the 
                ending index of the slice */
            int ob_end;
        
            union {
                struct t_string;
                struct t_slice;
            };
        } PyStringObject;
    
    Whenever a slice is taken of a string of Unicode object, the
    new object would use the t_slice portion of the union and would
    retain a reference to the original object.
    
    The C and Python interfaces to the string and Unicode objects
    would remain unchanged.

    Think about these optimizations:
    
    - don't use lazy slices if the resulting object with be "small"
    (relative to the original or in absolute terms)
    - there could be a built-in function that changes the slice inplace 
    to become a first class string (ala intern)
    - if there is only one reference to a string, and it is a "small"
slice
    then maybe we should reclaim the memory


Copyright

    This document has been placed in the public domain.





From dalke at dalkescientific.com  Tue Apr  9 13:55:58 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 9 Apr 2002 11:55:58 -0600
Subject: True/False story... (PEP 285)
References: <a8u99o$2lcu$1@news.nextra.cz> <vHys8.62333$pT1.1792330@news1.tin.it> <3CB2E617.8050600@geneva-link.ch>
Message-ID: <a8vaba$cnv$1@slb3.atl.mindspring.net>

Boris Borcic wrote:
>Could be argued that there is no single natural mapping, for an
>alternate to the current one we could have
>
>True     == 1
>False    == -1
>not x    == -x
>x and y  == min(x,y)
>x or y   == max(x,y)
>x xor y  == -x*y

Quoting Hilbert,
   "It must always be possible to substitute 'table', chair' and 'beer mug'
    for 'point', line' and 'plane' in a system of geometrical axioms."

I personally started with a BASIC which had 0 for false and -1 (all bits
on) for true.

                    Andrew
                    dalke at dalkescientific.com





From timr at probo.com  Mon Apr  8 22:57:37 2002
From: timr at probo.com (Tim Roberts)
Date: Mon, 08 Apr 2002 19:57:37 -0700
Subject: Python server from inetd?
Message-ID: <iml4bukknqf9nqe13h7dp0qco3qj2r86rt@4ax.com>

I have a quick little socket server that works OK as a continuously running
app, listening for and handling connections.  However, it is a rarely used
service -- only a few times a week -- so it seems silly to have a dedicated
process for it.  Thus, I tried to port it to inetd.  I cannot seem to make
this work.  This is on a RedHat 6.1 Linux system with Python 2.1.

I thought all I needed to do was basically:

  #! /usr/bin/python
  import socket

  client = socket.fromfd( 0, socket.AF_INET, socet.SOCK_STREAM )
  while 1:
    bfr = client.read(1024)
    if not bfr:
      break
    ...

But this does not work.  The "read" call returns 0 immediately, over and
over, until the client exits.  I never get any data.

In inetd.conf, I used:
  nsdc  stream  tcp  nowait  root  /root/myservice.py  myservice

The connection is happening; I can fetch the client IP address and port
number, but I can't read any data.

Suggestions?  Anybody have a known-good simple inetd-launched socket
service in Python they could share?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Mon Apr 15 11:02:21 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Mon, 15 Apr 2002 08:02:21 -0700
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com> <m3ofgn1avt.fsf@mira.informatik.hu-berlin.de> <3cb8b86a@news.mhogaming.com> <a9aoqc$j94$0@216.39.172.122> <LHAu8.11094$Ji7.10494@atlpnn01.usenetserver.com>
Message-ID: <3cbaebd1@news.mhogaming.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:LHAu8.11094$Ji7.10494 at atlpnn01.usenetserver.com...

| "Bengt Richter" <bokr at oz.net> wrote ...
|
| Since in Boolean terms a xor b is equivalent to ((a and not b) or (b and not
| a)), maybe a xor b should return whichever of a or b is not false.
|
| >>> a = 42
| >>> b = 0
| >>> a ^ b
| 42
| >>> b ^ a
| 42
| >>>
|
| Since that's exactly what it appears to do, stop fiddling with what already
| works !

NO it DOESN'T!! ::

>>> a = "123"
>>> b = 0
>>> a ^ b
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: unsupported operand type(s) for ^: 'str' and 'int'
>>>

'xor' should return the string '123' in this case-- not an error.  In other
words, '^' deals with BITS (as it now does), and 'xor' deals with objects.

I need (in my work) the ability to diddle bits-- these operators need to "do the
good thing" when operating on ints (and longs).  I don't want these operators
overloaded to recognize other objects-- that is the purpose of 'not', 'or',
'and', (and would be operator 'xor').

The operator '!=' performs a similar operation, but always returns a boolean.
'xor' should return the non-false object if the other operand is false, and
false if both operands are either true or both operands are false.

The 3 different behaviors of '^', '!=', and 'xor' each have their place in
writing code...





From cheezus at wanadoo.nl  Mon Apr 29 15:13:08 2002
From: cheezus at wanadoo.nl (guyon moree)
Date: Mon, 29 Apr 2002 19:13:08 GMT
Subject: Asynchronous Socket with the ServerSocket module (??)
Message-ID: <8Xgz8.730$5y4.17760@castor.casema.net>

hi all,

i'm trying to write some sort of server program on linux. i stumbled into
the ServerSocket module and it does pretty much what i want. i create a
requesthandler class and pass it to the module and call server_forever() By
using ForkingTCPServer i am able to make it handle multiple connecties.


all fine...

but...

how can i keep a list of connections and pass for example messages from one
client to the other through the server?

thanx in advance, i am a newbie though.

regards,

Guyon




From aahz at pythoncraft.com  Wed Apr 10 15:58:07 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Apr 2002 15:58:07 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <231u8a.lc.ln@127.0.0.1> <a8vnt9$roa$1@panix1.panix.com> <dotv8a.un.ln@127.0.0.1>
Message-ID: <a925gf$luk$1@panix1.panix.com>

In article <dotv8a.un.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
>Aahz wrote:
>> In article <231u8a.lc.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
>>>Aahz wrote:
>>>>
>>>> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
>>>> free to buy another book if K&R turns out to be too heavy for you right
>>>> now, but recognize that you'll need it eventually.
>>>
>>>Really?  I've been programming in C for ten years, and I've never even
>>>seen the K&R book.  I think I learned it just fine.  Not that I have a
>>>better suggestion.
>> 
>> I'm of the opinion that if you used K&R, you'd agree with me.
>
>Well, since I have been programming in C for ten years, and have never
>needed K&R, I have already proven your original statments, that I
>"*MUST* get K&R. Period," and that I'll "need it eventually," wrong.
>
>So if I agreed with you, I'd be wrong.

Precisely.  Of course, there's no way to prove that without your trying
to read K&R, but note that several people in this thread have agreed
with me (even if they haven't phrased their opinions as vociferously).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From jamescalthoff at yahoo.com  Sat Apr  6 13:29:02 2002
From: jamescalthoff at yahoo.com (James Althoff)
Date: Sat, 6 Apr 2002 10:29:02 -0800
Subject: Argument against iterable ints...
Message-ID: <000b01c1dd98$edddfb40$54dac13f@Pacbell.net>

[Magnus Lie Hetland]
> Just a situation that cropped up recently: While writing a little
> generator for flattening nested iterables, I used
>
>   try: iter(foo)
>   except: return foo
>   else:
>       ....
>
> (of course I could have used a for loop directly in the try clause --
> that doesn't change anything.)
>
> If integers became iterable, this sort of code would suddenly become
> infinitely recursive. Given the propensity for using exceptions in
> this manner rather than type checking in Python, this seems like a
> weighty argument to me...

Magnus,

I don't see how the above is even recursive, much less infinitely recursive.
Could you elaborate?

Also, note that the final, rejected (finally rejected ;-) version of PEP 276
suggests an alternative to adding an iterator to class int (even though such
a simple method addition was what motivated the original idea).  The
alternative would be (would have been) to change the code in the for-loop
implementation to add a conceptual two-liner:

    if isinstance(iter_target,int):
        iter_target = xrange(iter_target)

Doing so would make
    for i in 5:
a simpler, less-cluttered spelling of
    for i in xrange(5):

and therefore would allow the "flatter is better than nested"
    for i in len(L):
as a substitute for the more nested, less flat
    for i in xrange(len(L)):

And it would prevent all problems with an integer iterator popping up in
undesirable circumstances -- since there would be no such iterator.

Jim





From stojek at part-gmbh.de  Fri Apr 12 10:49:42 2002
From: stojek at part-gmbh.de (Marcus Stojek)
Date: Fri, 12 Apr 2002 14:49:42 GMT
Subject: post event from C-extension
Message-ID: <3cb6f29f.23943234@news.easynews.net>

Hi,

my program has a couple of time consuming functions. With the help
of the wxPython Cookbook I use Threads to keep my GUI alive. When
finished, each Thread uses wxPostEvent to notify the application.

One function is a C-extension and I have no idea how to make this
thing sending an event. Any hint's, examples, links ?

Thanks
Marcus


From bokr at oz.net  Tue Apr 16 22:19:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 17 Apr 2002 02:19:58 GMT
Subject: Python 2.2 unwise.exe
References: <mailman.1019003794.8891.python-list@python.org>
Message-ID: <a9im4e$rf4$0@216.39.172.122>

On Tue, 16 Apr 2002 20:34:42 -0400, Tim Peters <tim.one at comcast.net> wrote:

>> 1. Boot the box.
>>
>> 2. Install 2.2.1.
>>
>> 3. Uninstall 2.2.1.
>
>[Bengt Richter]
>> What happens if you shut down and boot here?
>
>Nothing visible.  The entire \Python22 directory is gone (including
>WISE.EXE) immediately after completing step 3.  After a reboot, it's still
>gone <wink>.
>
Sure. But the point was to see if the deferred delete got flushed so that
reinstalling could succeed without thereafter still having a deferred
delete to get hammered with on the next reboot after that.

>> (I'm wondering if there's a deferred delete waiting for a reboot
>> which doesn't get cleared by the re-install at 4.)
>
>Of course.  But since the only relevant file in use at the time UNWISE.EXE
>is running *is* UNWISE.EXE, and Wise already got rid of it, it appears to be
>a Wise bug that it arranges to delete, at the next reboot, any file named
>\Python22\WISE.EXE the user happens to create after Wise already got rid of
>the original.
>
This is the delete which I wanted to see if it got flushed with the otherwise
unnecessary reboot at 3a.

If so, you could just specify reboot after uninstall before the next install as a fix,
I would think.

Regards,
Bengt Richter


From guido at python.org  Sat Apr  6 08:43:19 2002
From: guido at python.org (Guido van Rossum)
Date: Sat, 06 Apr 2002 13:43:19 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python 
 going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84>
Message-ID: <3CAEFC4E.2778AC55@python.org>

James Logajan wrote:
> I tried that and that strategy _failed_. The chances of failure was 100%.
> The failure was due specifically to how (as far as I can tell) Tkinter
> handles UTF-8 sequences in Unicode aware versions of Python. So I am forced
> to put version checking into my code! That just added to my irritation.
> There are other _forward_ incompatibilities in Python, but it is the hidden
> ones that misguided programmers are writing into common libraries that make
> the proliferation of Python versions too irritating to deal with.

I sent James a personal reply to this which I'd like to repeat here
(with typos fixed):

"""
I'm sorry that you had a bad experience when Unicode was introduced.
I could probably have done better (I didn't realize that Tcl had a
backwards compatibility feature to deal with Latin-1 strings).

About the change rate in general: it's hard to find the proper pace.
Most new features are requested by users (even the division change!)
and there's definitely a group of users who push for all sorts of new
features -- I say "no" a lot more than "yes"!  I try hard to stay
compatible (that's why the new bool type is not "pure"); I've
introduced things like __future__ statements and warnings that can
help introducing new features without breaking code.  Of course, there
are always tiny corners of the language (or library!) where things
break despite all good intentions, and given the large user base it's
inevitable that there's always *someone* who gets stuck with broken
code.

You clearly fall in the most conservative camp.  I just had a long
exchange with someone who pleaded strongly for moving forward at a
higher pace, breaking more existing code so that the perfect language
can be obtained sooner.  I will have nothing of that, but I hope you
understand my predicament: for every user like you who complains that
things change too fast, there's someone with an equally strong desire
for faster change.  I have to pick a middle ground (and it's easier to
say "don't upgrade" than "wait a year").  Since I'm a language
designer, I enjoy designing new features, but lately each new feature
has cost me probably 10 times more effort to make it backwards
compatible than to design the feature in the first place.  (Examples
besides bool: new-style classes, nested scopes, iterators and
generators all required significant efforts to avoid breaking old
code.)

I don't expect this to satisfy you or to change your mind; I don't
want it to sound like an apology (though it probably does); but I do
hope you understand the forces at work, and I hope you believe I'm
doing the best I can for backwards compatibility.
"""

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


From paolo at prosa.it  Thu Apr 18 05:52:41 2002
From: paolo at prosa.it (Paolo Andreetto)
Date: 18 Apr 2002 02:52:41 -0700
Subject: HTTPS: memory leakage
Message-ID: <1918e8fd.0204180152.26cb929b@posting.google.com>

Hi

It seems to me that HTTPS support of python2.1 causes a memory
leakage, but I can't understand the source of such a problem.
With this code:

import time,httplib,sys
    while 1:
	conn=httplib.HTTPS('<webserver address>')
	conn.putrequest('GET', '/index.html')
	conn.putheader('Accept', 'text/html')
	conn.putheader('Accept', 'text/plain')
	conn.endheaders()
	errcode, errmsg, headers = conn.getreply()
	print errcode # Should be 200
	f = conn.getfile()
	data = f.read() # Get the raw HTML
	print len(data)
	f.close()
	conn.close()
	time.sleep(1)

every 10 seconds, in the average, the process requires 4Kbytes and
this trand seems to be unbreakable.
This test run on a i386 with a debian (sid) distro, with kernel 2.4.18
and python 2.1.3. The host contacted was a i386 with a RedHat 7.2
distro, with apache and mod_ssl as supplied by the distribution.
Does it depend on the code or is there any problem in the system?

Thanks
Paolo


From mixo at beth.uniforum.org.za  Tue Apr  9 06:07:51 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Tue, 09 Apr 2002 12:07:51 +0200
Subject: Queus clarification
Message-ID: <3CB2BD77.9070708@beth.uniforum.org.za>

If 'Queue.put(some_method)' is called from some script,
does the script that called that 'Queue.put' have to wait
for it to finish execution?



From pobrien at orbtech.com  Tue Apr  2 11:45:42 2002
From: pobrien at orbtech.com (Patrick K. O'Brien)
Date: Tue, 2 Apr 2002 10:45:42 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>
Message-ID: <NBBBIOJPGKJEKIECEMCBCEBGMLAA.pobrien@orbtech.com>

[Laura Creighton]
[powerful prose snipped]

> I want to keep python out of the True and False business.  Python cares
> about whether a value is Something or Nothing.  This is beautiful, and
> _better_ than what the other languages do.

Wow. I'm almost speechless. You've presented a unique point of view with
such vividness that I'm quite literally stunned. If I could afford it I
would pay you to evaluate every one of my ideas before I made a decision.

Thank you for such a powerful argument. I can't predict whether it will
change the course of events, but it sure has opened my eyes to a different
perspective. I think it would be foolish to not ponder this issue for a
while before making any final decisions.

---
Patrick K. O'Brien
Orbtech




From sqrt_2  Mon Apr 15 12:28:39 2002
From: sqrt_2 (sqrt_2)
Date: Mon, 15 Apr 2002 16:28:39 GMT
Subject: Jython help pls
References: <3cb9c6c3.115241899@news.mweb.co.za> <Xns91F0D381895A5cliechtigmxnet@62.2.16.82> <3cb9ddfb$0$150$9b622d9e@news.freenet.de>
Message-ID: <3cbaff3c.195235684@news.mweb.co.za>

Thanks for the info. I seem to have figured out most of the bugs.
And yes the 'Java' was a case misspelling for 'java'. 8-)
Now I just need to get used to some syntax things etc...

On Sun, 14 Apr 2002 22:01:45 +0200, "merman" <merman at freenet.de>
wrote:

>
>"Chris Liechti" <cliechti at gmx.net> schrieb im Newsbeitrag
>news:Xns91F0D381895A5cliechtigmxnet at 62.2.16.82...
>> sqrt_2 wrote in news:3cb9c6c3.115241899 at news.mweb.co.za:
>> > Can anyone help with Jython compiler error messages. I find it
>> > difficult to understand what the errors are as I have only just
>> > started with this product.
>>
>> maybe you want to work through a python tutorial? the one one python.org
>> is for CPython but 90% or more will apply to Jython too.
>>
>> > Messages  like:
>> > 'NameError: Java'
>> note that python and java are case sensitive. you probly meant "from java
>> import ..." (small 'J')
>>
>> > What does that mean?? Is there a reasonable site where I can find
>> > explanations for common error messages? The Jython site FAQ is no help
>> > at all.
>>
>> The python docs should help
>>
>>
>>
>> --
>> Chris <cliechti at gmx.net>
>>
>
>
>... or try www.jython.org
>
>o-o
>
>thomas
>
>
>



From aahz at pythoncraft.com  Thu Apr 25 01:44:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Apr 2002 01:44:01 -0400
Subject: Python v. Euphoria
References: <394fcus92g8rt0e2ektsd36s9qu7cq2cij@4ax.com>
Message-ID: <aa8531$fho$1@panix3.panix.com>

In article <394fcus92g8rt0e2ektsd36s9qu7cq2cij at 4ax.com>,
Stom Fairecroire  <stom at hotpop.com> wrote:
>
>Many, many years ago I did a lot of programming in HPL (Hewlet Packard
>Language) then did lots using PICK basic. Now I want to start some
>very low key programming again - just for my own amusement. Anyone
>have any views on the pros and cons  of Python v. Euphoria ? Which
>would be the better option?

Based on a quick look at the Euphoria web site, Python would be a much
better option.  Euphoria is a proprietary product that is only available
on a few platforms; it looks like Euphoria programs tend to not be
cross-platform even for the available platforms.  Euphoria also looks a
bit uglier than Python, and it lacks some of Python's features, such as
threads.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

What if there were no rhetorical questions?


From matt at greenroom.com.au  Tue Apr  9 00:57:58 2002
From: matt at greenroom.com.au (Matthew Baulch)
Date: Tue, 09 Apr 2002 14:57:58 +1000
Subject: Embedding Python with C++
References: <GYts8.26982$2J2.6830478@typhoon.southeast.rr.com>
Message-ID: <3CB274D6.5060905@greenroom.com.au>

Chris Stoy wrote:

> Hey all,
> 
> I'm looking to embed Python as a scripting language in a game I am working
> on.  I want to be able to call C++ methods from Python, and call Python from
> C++ (for the usual code quick in Python, convert to C++ for speed later.)
> 
> I've looked into using the Boost.Python code and I'm fine with that.
> 
> My problem is, I have no clue how to compile Python into a static linked
> library for Windows.  Actually, I need it to be cross-platform, which is
> another issue.
> 
> I've never embedded Python before, and I'm fairly new to Python, so can
> someone please give me a pointer to how to do this?  What are the files I
> need to compile to build the static lib (using MS Visual Studio 6.0?) Also,
> I don't need all of the standard Python components (like networking.)  Can
> those be removed from the lib?
> 
> Thanks.
> 
> Chris.
> 


Look @ http://www.python.org/doc/current/api/api.html

It should solve most of your question

Cheers,
   -matt-



From guido at python.org  Wed Apr  3 20:22:46 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 01:22:46 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <7xbsd0qpmn.fsf@ruckus.brouhaha.com>
Message-ID: <3CABABCA.EB446FEE@python.org>

Paul Rubin wrote:
> 
> Guido van Rossum <guido at python.org> writes:
> >     True and False will properly round-trip through pickling and
> >     marshalling; for example pickle.loads(pickle.dumps(True)) will
> >     return True, and so will marshal.loads(marshal.dumps(True)).
> 
> It occurs to me that the array module should also support the new bool
> type (though the letters 'b' and 'B' are already used, so a different
> onewould have to be chosen).  Then at least we get an efficient way of
> handling bit vectors that we didn't have before.

I wouldn't say "should", but it sure would be nice.  Unfortunately the
internal architecture of the array module doesn't make this easy; it
would probably require a major refactoring, or perhaps a less pervasive
switch to new-style classes so that subclassing can be used.

Maybe you and a few others can get together and start working on the
code?  Or do I have to do everything myself?

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


From tim.one at comcast.net  Fri Apr  5 13:24:35 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 05 Apr 2002 13:24:35 -0500
Subject: Booleans, integer division,
 backwards compatibility; where is Python going?
In-Reply-To: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDOOLAA.tim.one@comcast.net>

[phil hunt, on language change]
> ...
> But is this changing going to be continual? Are the language
> developers constantly going to tinker with the language, out of a
> sense of fun? (Not that having fun is bad, it's just that I don't
> want to code against a continually moving target).
>
> Is there a final destination in sight, which these changes are
> leading to?

It's possible it's the same final destination that C, Fortran, Perl and Java
(for examples) are approaching with their own never-ending changes, but, if
so, I doubt we'll live long enough to celebrate arrival.  If you want a
change-free language, you're best off looking for a dead or dying language.
For example, Algol 60 hasn't changed in decades.  For that matter, Python
1.5.2 hasn't changed in years.  WRT any technology, you can jump off the
train any time you like.  Unless you pick your exit point very carefully,
though, you may end up alone in the middle of nowhere.

or-in-a-small-town-with-a-crate-full-of-8-track-tapes-ly y'rs  - tim





From stephen at xemacs.org  Sat Apr 13 03:33:53 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 13 Apr 2002 16:33:53 +0900
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1017656355.5887.python-list@python.org> <12bb92e6.0204091223.738e0798@posting.google.com> <i_nt8.15754$ir6.679622@news1.calgary.shaw.ca> <GrAt8.11681$F43.8566@atlpnn01.usenetserver.com> <mailman.1018626392.16328.python-list@python.org> <XSIt8.20985$Kq4.882398@news2.calgary.shaw.ca>
Message-ID: <87wuvc3vdq.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Brad" == Brad Bollenbach <bbollenbach at shaw.ca> writes:

    Brad> You misread, or misunderstood (or both) what I said. It's
    Brad> not "optional" to license Open Source/Free Software. It got
    Brad> to be *called* Open Source/Free _because_ of the license you
    Brad> chose for it.

I'd put it slightly differently: it can _only_ be called OSS/FS if you
apply a free license to it.

    Brad> If you have a piece of software with no license whatsoever
    Brad> attached to it, it's not proprietary, open source/free,
    Brad> shareware, "freeware", or whatever.

Wrong, at least in the U.S.  It's proprietary, because all rights are
implicitly reserved to the copyright holder.  Rick Moen says it better
than I can: http://www.linuxmafia.com/~rick/faq/#djb.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From marilyn.mix at omitron.com  Wed Apr 10 16:16:18 2002
From: marilyn.mix at omitron.com (M. J. Mix)
Date: 10 Apr 2002 13:16:18 -0700
Subject: Python and spacecraft onboard or ground control scripting
Message-ID: <7226483f.0204101216.2c8fcda4@posting.google.com>

Is anyone using Python as a spacecraft ground system scripting
language, or onboard (embedded) language?

Pros and cons when compared with TCL or JAVAScript.  Just 
some real life, honest opinions.

The usual issues: reliability and maintenance effort in applications. 
Also, use as a scripting language by non-programmers. (I saw the post
on indenting, so I'm concerned about that.)


From fgeiger at datec.at  Sat Apr 13 08:49:43 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Sat, 13 Apr 2002 14:49:43 +0200
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de>
Message-ID: <a999hl$nmd$1@rex.ip-plus.net>

You have to do this by yourself, i.e. there's no built-in support for this


class Abstract: # All subclasses of this have to conform to this
   def __init__(self):
      return

   def method(self):
      raise NotImplementedError # Oops, you are working with an Abstract
object or you have forgotten to override method()!


class Concrete:
   def __init__(self):
      Abstract.__init__(self)
      return

   def method(self):
      pass # Your stuff goes here
      return


a = Abstract()
a.method() # Will raise an error

c = Concrete()
c.method() # Will do, what you told it to do


Try Google to find more on this as this is frequently discussed here.

Best regards
Franz

"merman" <merman at freenet.de> schrieb im Newsbeitrag
news:3cb7f47e$0$360$9b622d9e at news.freenet.de...
> Hi,
>
> is there a way to simulate interfaces (like in java) in python?
> ... or abstract classes? (c++/ java)
>
> o-o
>
> thomas
>
>
>
>




From alx78 at hotmail.com  Wed Apr 17 13:53:34 2002
From: alx78 at hotmail.com (alex)
Date: 17 Apr 2002 10:53:34 -0700
Subject: Curses for windows.
Message-ID: <f8e5025.0204170953.570cfb43@posting.google.com>

Hi,

Where is "curses" module for python 2.2 in windows NT.

bye.


From max.wenzel at web.de  Tue Apr 23 10:03:04 2002
From: max.wenzel at web.de (Max)
Date: 23 Apr 2002 07:03:04 -0700
Subject: check, if an url exists
Message-ID: <2eea6430.0204230603.420cf7ee@posting.google.com>

Halli-Hallo 

I'm looking for a script, that checks an url, if exists or not.
Can anyone help me ?? 

Thanks a lot and many greetings,
Max


From pyth at devel.trillke.net  Wed Apr 24 11:07:23 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 24 Apr 2002 17:07:23 +0200
Subject: flatten tuple
In-Reply-To: <m23cxl8a8y.fsf@video.paradise.lost>
References: <m2adrvnxiv.fsf@mother.paradise.lost> <mailman.1019594863.1719.python-list@python.org> <m23cxl8a8y.fsf@video.paradise.lost>
Message-ID: <20020424150723.GA19893@devel.trillke>

On Wed, Apr 24, 2002 at 09:03:09AM -0500, John Hunter wrote:
> >>>>> "holger" == holger krekel <pyth at devel.trillke.net> writes:
> 
>     holger> Hello John, because i have related interests i have taken
>     holger> the time to write a very small (one screen) and robust
>     holger> more general 'SubstDict'. you can use it like this:
> 
> Thanks to all for really helpful suggestions.  Duncan's code worked
> great -- Bengt's even does the recursive check.  I have upgraded to
> 2.2, and will study your examples closely to gleam some knowledge from
> them.

feel free to ask.

	holger




From aleax at aleax.it  Fri Apr 26 18:31:29 2002
From: aleax at aleax.it (Alex Martelli)
Date: Fri, 26 Apr 2002 22:31:29 GMT
Subject: Removing dupes from a list
References: <20020426203946.GV21205@teich.Garten.DigitalProjects.com> <XFMail.20020426135630.shalehperry@attbi.com> <mailman.1019856418.32215.python-list@python.org>
Message-ID: <5zky8.66519$vF6.2038409@news2.tin.it>

Alexander Skwar wrote:

> ?Sean 'Shaleh' Perry? sagte am 2002-04-26 um 13:56:30 -0700 :
>> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52560
> 
> Thanks, but connections to activestate.com simply time out :(  However,
> a traceroute to aspn.activestate.com doesn't show any unusual long hops.
> But AS is reachable from other networks.  Strange.
> 
> Anyhow, seems like uniquer() from the 1st comment is what I'm looking
> for, because unique() also looses order just like my two "functions".

As uniquer's author, I'm glad that my concern about ordering helps
you:-).  And you should see what other gyrations we came up with
for the printed Cookbook... real neat, if I say so myself.  Hope I'll have
the time to put the best up on the online cb as soon as the printed
one's ready (...but I should really be working on the Nutshell instead
by that time...).


Alex



From tejarex at yahoo.com  Wed Apr 10 10:24:43 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 10 Apr 2002 14:24:43 GMT
Subject: RELEASED: Python 2.2.1
References: <mailman.1018435709.31373.python-list@python.org>
Message-ID: <LWXs8.208801$VJ1.16864889@bin3.nnrp.aus1.giganews.com>

"Michael Hudson" <mwh at python.org> wrote in message
news:mailman.1018435709.31373.python-list at python.org...
> We've released a bugfix release of Python, 2.2.1!

Hooray!  Just as soon as I send off my income tax forms....

> This release should be fully compatible with Python 2.2.  Everyone
> using 2.2 is recommended to upgrade to 2.2.1.
>
> On the 2.2.1 pages
>     http://www.python.org/2.2.1/

> This being a bugfix release, there are no exciting new features --
we just
> fixed a lot of bugs.  For a moderately complete list, please see:
>
>     http://www.python.org/2.2.1/NEWS.txt

Ah, but you did introduce one (unexciting?) new feature that users
should know about:

- Added new builtin function bool() and new builtin constants True and
  False to ease backporting of code developed for Python 2.3.  In 2.2,
  bool() returns 1 or 0, True == 1, and False == 0.

This will also make it easier to write code that looks forward to 2.3+
as long as one does not test type(True), `True`, and str(True) (and
ditto for False).

Terry J. Reedy






From terrybyrne1963 at yahoo.com  Tue Apr 30 10:11:44 2002
From: terrybyrne1963 at yahoo.com (Terry Byrne)
Date: 30 Apr 2002 07:11:44 -0700
Subject: Newbie tip: Prg speed from ~5 hrs to < 1 minute
References: <93d52e82.0204270901.3cf1ecd@posting.google.com> <aaevge$emr$1@peabody.colorado.edu> <93d52e82.0204290618.349a0e5f@posting.google.com> <c76ff6fc.0204291406.4bd58af5@posting.google.com>
Message-ID: <c9534282.0204300611.22b07d9b@posting.google.com>

John,

Re: pass, I stand corrected, of course. Thank you. It just happened to
appear to do exactly the thing I needed in that context. But in other
contexts, ...

As for assigning aLine=None, then trying to work with aLine, I must
plead accidental mis-paste. I threw my post together by copying and
pasting from numerous versions of the eventual product, that one got
in there by mistake!

Terry

sjmachin at lexicon.net (John Machin) wrote in message news:<c76ff6fc.0204291406.4bd58af5 at posting.google.com>...
> TerryByrne1963 at yahoo.com (Terry Byrne) wrote in message news:<93d52e82.0204290618.349a0e5f at posting.google.com>...
> 
> [big snip]
> > 
> > At first I was just running the re.search on each line. Time: ~ 5 hrs
> > for 60K line log file. Started testing with the string.find method to
> > make sure a certain error code appeared on the line before running the
> > "expensive" re.search, and that cut execution time from ~5 hrs to ~
> > 1.5 hrs, good but "no cigar," as they say.
> > 
> > Then I realized that each line of content could be treated the same as
> > you'd treat a switch() structure in C, or a case structure in Pascal:
> > if one option is a "hit", then you needn't even check for all the
> > other options. So I started pass-ing to the next line of the content
> > whenever I got a "hit."
> 
> Please re-read your source of information on the pass statement.
> "pass" does nothing. Zip. Zilch. It is only a place holder for cases
> where you want to do nix but Python syntax insists on a statement
> being present. Here are a couple of use cases:
> 
> class FubarError(Exception):
>    pass
> 
> if a and b or c and d:
>    pass
> else:
>    # Late at night, don't want to try inverting
>    # the above in my head, don't like not (...)
>    do_something()
> 
> You can experiment with pass's 'functionality', by putting something
> like the following in a file and running it from the command-line:
> 
> for i in range(5):
>    print i
>    pass
>    print 'Still here!'
>    pass
>    print 'Persistent little varmint ...'
>    pass
>    print 'Now do you believe Uncle John?'
> 
> Then change all occurences of 'pass' to '# pass' and see what happens.
> Then change all the occurrences of '# pass' to 'continue' and see what
> happens.
> 
> > That cut my time down to under a minute for a
> > 60K-line log file.
> 
> Let's have a look at the following snippet of your code:
> 
> >      idx = aLine.find('Error Validator:2420')
> >      if idx > -1:
> >        LINKS_VNOQUERYHITS += 1
> >        aLine += '<br>'
> >        rfHandle.write(aLine)
> >        aLine = None 
> >        idx = -1
> >        pass
> >      idx = aLine.find('Error Validator:2412')
> >      if idx > -1:
> 
> This just cannot possibly be working [or maybe this code isn't even
> executed at all!!!]. You bind aLine to None then attempt to execute
> idx = aLine.find('...') which will cause an exception:
> 
> AttributeError: 'NoneType' object has no attribute 'find'
> 
> ... unless of course previously you have done
> 
> None = ""
> 
> which needless(?) to say is something you should *never* do.
> 
> If you can't find out why this is happening yourself, it's time to
> head for the  tutor list.
> 
> All those lines which bind idx to -1 are pointless. Try changing all
> of them firstly to idx = -99999999 and secondly to idx = 99999999. If
> either of those makes any difference at all (except a tiny increase in
> the run-time) then something drastic has happened: Python is stuffed,
> your copy of Python is stuffed, ...
> 
> HTH,
> John


From andy47 at halfcooked.com  Mon Apr 22 21:32:30 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Tue, 23 Apr 2002 01:32:30 +0000 (UTC)
Subject: DBAPI
References: <87adrvambl.fsf@complete.org>
Message-ID: <Xns91F975562E137andy47halfcookedcom@203.109.252.31>

John Goerzen <jgoerzen at complete.org> wrote in 
news:87adrvambl.fsf at complete.org:

> Hi,
> 
> I have some DB-API questions.
> 
> I have used things like JDBC for Java and DBI for Perl.  To me, DBAPI
> for Python looks rather week.  For instance, if the code is to be
> truly portable across databases, I'd have to check the paramstyle for
> the driver and write each query 5 times!
> 
> Is there any "meta-DBAPI" module that I can use that would present me
> with a standardized interface, and then (for instance) convert my
> queries to whatever format the backend wants?
> 
> -- John
> 

As far as I am aware there is no meta-DBAPI. In my experience when people  
talk about standard database interfaces that enable them to write 
portable code they haven't actually tried to write any. SQL isn't 
standard so it is impossible to write database independent code for all 
but the most trivial of applications.

I agree that the DB-API is a little flexible and does result in a number 
of different styles, but it tries to guide rather than direct. Given the 
diversity of databases that are supported by Python modules it is hardly 
surprising that they all don't operate in exactly the same way. 

On the other hand it is very hard to write Python db modules that will 
work in exactly the same way on several databases anyway. The example you 
give of parameter style is a case in point, most Python db modules stick 
close to the implementation in the supplied database libraries rather 
than try and wrestle them into some abritrary standard.

The value of any meta-DBAPI is in direct proportion to the amount of 
effort it would take to change all of the existing db modules. Its not 
impossible, but its a lot of hard work and no one has seen the need to do 
this yet.

For more detailed discussion may I suggest moving to the DB-SIG mailing 
list, for details see - http://mail.python.org/mailman/listinfo/db-sig

Regards,
Andy
-- 
Contents free posts a speciality


From Chris.Barker at noaa.gov  Fri Apr  5 16:46:33 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 05 Apr 2002 13:46:33 -0800
Subject: wxPython Book ?
References: <11e94203.0204031526.f11bfe6@posting.google.com> <a8gr9l$c29$04$1@news.t-online.com>
Message-ID: <3CAE1B38.64308274@noaa.gov>


Harald Schneider wrote:
> > Is there a wxPython book available? that gives a depth tutorial?
> There is a book on the way:
> > > We're a computer book publisher seeking reviewers for a book proposal
> > > on wxPython we currently have in hand. 

That proposal was rejected. Hopefully, another will come along soon...

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From ufechner at csi.com  Mon Apr  8 17:02:43 2002
From: ufechner at csi.com (Uwe Fechner)
Date: Mon, 08 Apr 2002 23:02:43 +0200
Subject: Compatibilty table between BOA, wxPython, Python?
References: <11e94203.0204062147.5d2281b9@posting.google.com> <3CB0245A.3010808@csi.com> <11e94203.0204071640.103c4bb7@posting.google.com>
Message-ID: <3CB20573.6060006@csi.com>

Jose Rivera wrote:
> The problem is that I made a program with BOA 0.0.5, it worked ok...
> and now reports the next error, so I guess that the problem may be the
> compatibility with wxPython 2.3.2.1
> 
> Traceback (most recent call last):
>   File "wxapp1.py", line 24, in ?
>     main()
>   File "wxapp1.py", line 20, in main
>     application = BoaApp(0)
>   File "C:\TOOLS\Python22\lib\site-packages\wxPython\wx.py", line
> 1587, in __ini
> t__
>     _wxStart(self.OnInit)
>   File "wxapp1.py", line 11, in OnInit
>     self.main = wxFrame1.create(None)
>   File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 11, in
> create
>     return wxFrame1(parent)
>   File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 161, in
> __init__
>     self._init_ctrls(parent)
>   File "C:\TOOLS\MyDevelopments\InfoMan\wxFrame1.py", line 115, in
> _init_ctrls
>     self.statusBar1 = wxStatusBar(size = wxSize(685, 20), id =
> wxID_WXFRAME1STAT
> USBAR1, parent = self, name = 'statusBar1', style = 0, pos =
> wxPoint(0, 347))
>   File "C:\TOOLS\Python22\lib\site-packages\wxPython\stattool.py",
> line 55, in _
> _init__
>     self.this = apply(stattoolc.new_wxStatusBar,_args,_kwargs)
> TypeError: 'pos' is an invalid keyword argument for this function

Perhaps this is an answer to your question:

http://www.geocrawler.com/lists/3/SourceForge/2090/0/8325017/

Generally I think, the mailing list of the boa project is a better place
to ask boa related questions.

Regards:

Uwe



From ianb at colorstudy.com  Tue Apr 16 03:20:09 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 16 Apr 2002 02:20:09 -0500
Subject: Python webmail packages?
In-Reply-To: <7xsn5wjk85.fsf@ruckus.brouhaha.com>
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com>
	<7876a8ea.0204151614.138853be@posting.google.com>
	<mailman.1018918474.3622.python-list@python.org> 
	<7xsn5wjk85.fsf@ruckus.brouhaha.com>
Message-ID: <1018941610.1149.1133.camel@localhost>

On Tue, 2002-04-16 at 00:17, Paul Rubin wrote:
> Ian Bicking <ianb at colorstudy.com> writes:
> > BoboMail works okay.  It uses some portions of an old version of Zope
> > (hence it's named after that old version -- actually, it might be new
> > versions of those pieces, I don't know) -- but it's actually CGI-based. 
> 
> Hmm, how much code is involved?  I didn't want to deal with installing
> a huge package like Zope.

Installation isn't hard, like I said it's CGI.  It happens to reuse some
code from Zope, but all that code is contained in the package itself.

  Ian






From pete at dexterslab.de  Tue Apr 16 06:31:52 2002
From: pete at dexterslab.de (Henning Peters)
Date: Tue, 16 Apr 2002 12:31:52 +0200
Subject: xml2dict - dict2xml
Message-ID: <a9gudi$964$04$1@news.t-online.com>

hi NG,

i am searching for the best solution to convert a small xml fragment in a
python dictionary and vice versa.
speed is a very important factor, so i have tried the xml.parsers.expat
interface, but i did not found out, how to map a tree with it (i.e. to parse
only one branch).

this...

<cmd name="foo">
  <var>var1</var>
  <var>var2</var>
  <var type="hash">
    <var key="a">var1</var>
    <var key="b">var2</var>
  </var>
</cmd>

...should be converted to a dict like this:
{'cmd': 'foo', 'var':('var1', 'var2', {'a':'var1', 'b':'var2'})}

in most cases the xml data is very small (<<1kByte).
it is only necessary to be able to map strings (var), hashes (var
type="hash") and lists (var type="list")
all vars are encapsulated by a cmd tag. the cmd tag is the root tag
the way back should work similar...

what would you suggest is the best (and fast) way to do these conversions?
do you know whether there already exist such an approach?

bye,
henning





From pj at engr.sgi.com  Thu Apr  4 22:21:18 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 5 Apr 2002 03:21:18 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <aeei8a.i11.ln@127.0.0.1>
Message-ID: <a8j57e$omv31$1@fido.engr.sgi.com>

|> ... Kronicker Delta ...

sp?  Kronecker Delta

But why do you lament that '==' is no longer like the Kronecker Delta function?
That still seems like a fitting name to me.
-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From skip at pobox.com  Tue Apr  2 11:40:33 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 2 Apr 2002 10:40:33 -0600
Subject: Someone using PyGtk
In-Reply-To: <3ca9d7d1$1_2@news.uni-bielefeld.de>
References: <3ca9d7d1$1_2@news.uni-bielefeld.de>
Message-ID: <15529.57089.87974.950040@12-248-41-177.client.attbi.com>

    Mario> i wonder wether someone is using pygtk and if it is nice toolkit
    Mario> to learn....

    Mario> OK, what i'm looking for is an easy to learn (i'm not very
    Mario> experienced in gui-programming), quite fast(small) toolkit.

If you like Gtk, you'll like PyGtk, which exposes Gtk to Python
programmers.  PyGtk itself isn't big, but Gtk is pretty large.  Don't bother
with 1.2.  Go straight to 2.0, which was released recently.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)



From martin at v.loewis.de  Sun Apr 14 05:20:33 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 11:20:33 +0200
Subject: 2.1 vs. 2.2
References: <mailman.1018751366.25497.python-list@python.org> <YM9u8.14076$b62.391028@news1.tin.it>
Message-ID: <m3wuvaws9q.fsf@mira.informatik.hu-berlin.de>

Alex Martelli <aleax at aleax.it> writes:

> Not what I saw happen in the small set of languages whose standardization
> I had to follow.  With C++, in particular, the standard was so far from
> "allowing all variations that already existed" that, AFAIK, still today, 
> years later, no commercial implementation claims 100% standard compliance
> (something that I still find rather amazing).  Of course, the usefulness
> of that is as dubious as that of a standard which did allow everything.

As an co-author of a C++ compiler, I can assert that the C++ standard
very useful both to C++ users and C++ vendors. It is easy to analyse
g++ bug reports against the standard; it turns out that you can close
many reports as "not a bug" if you have the backup of the
standard. Users agree that they should not dispute the standard text
with their vendor (in the rare cases where they dispute the conclusion
that their code is wrong, they usually agree that they need to talk to
the committee, reporting defects etc).

For this kind of application of a standard, it is immaterial whether
g++ claims to implement the standard "in full". There are obvious
areas where g++ intends to comply, and deviations in that area are
considered as bugs, and there are areas where g++ admittedly (and in a
documented manner) omits features from the standard. Users generally
appreciate that distinction, and try to avoid unimplemented features,
or use official work-arounds.

Regards,
Martin


From logiplexsoftware at earthlink.net  Wed Apr 10 19:03:20 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 10 Apr 2002 16:03:20 -0700
Subject: Extracting from a list
In-Reply-To: <MPG.171e61f8eb76328a9896aa@news.nwlink.com>
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
	<Xns91EC96C0C8E9CRASXnewsDFE1@130.133.1.4>
	<MPG.171e61f8eb76328a9896aa@news.nwlink.com>
Message-ID: <20020410160320.1e1b025c.logiplexsoftware@earthlink.net>

On Wed, 10 Apr 2002 15:52:51 -0700
Jeff Shannon wrote:

> In article <Xns91EC96C0C8E9CRASXnewsDFE1 at 130.133.1.4>, 
> starx at pacbell.net says...
> > Joel Bender || Wed 10 Apr 2002 01:35:10p:
> > 
> > > I would like to process the elements of a list that match a condition, 
> > > then remove them.  I'll use integers in my example, but my list items 
> > > are actually objects.
> > 
> > Two approaches come to mind, both using while instead (Because we need 
> > to use the number instead of just the item):
> > 
> > i = 0
> > while i < len(list)
> >     	if list[i] < 5:
> >     	    	print list[i]
> >     	    	del list[i]
> >     	i += 1
> 
> This will *not* work as expected.
> 
> When you del list[I], then list[I+1] becomes list[I].  You then 
> increment I, and get the next item... except that the item you 
> get is the one that *used* to be list[I+2].  You skip an item 
> each time you delete an item.
> 
> >>> mylist = range(5)
> >>> mylist
> [0, 1, 2, 3, 4]
> >>> I = 0
> >>> while I < len(mylist):
> ... 	if mylist[I] < 3:
> ... 		print mylist[I]
> ... 		del mylist[I]
> ... 	I += 1
> ... 
> 0
> 2
> >>> mylist
> [1, 3, 4]
> >>> # note that 1 got skipped!
> 
> This is why it's always dangerous to modify the list you're 
> iterating over.  It can be done safely (see Daniel's back-to-
> front method), but it takes some work.

Although slightly modified it would work:

>>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>>> i = 0 
>>> while i < len(lst):
...     if lst[i] > 5:
...         print lst[i]
...         del lst[i]
...     else:
...         i += 1
... 
7
8
6
9
>>> lst
[3, 2, 5, 1, 0, 4]

I agree in principle that modifying a list that is being iterated over is
somewhat risky.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From duncan at NOSPAMrcp.co.uk  Thu Apr 18 04:21:30 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 18 Apr 2002 08:21:30 +0000 (UTC)
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org> <3CBBAB98.91FA1B17@alcyone.com> <a9gbhj$dhm$1@panix1.panix.com> <slrnabs3mm.1kfm.Gareth.McCaughan@g.local>
Message-ID: <Xns91F45DF1695BFduncanrcpcouk@127.0.0.1>

Gareth.McCaughan at pobox.com (Gareth McCaughan) wrote in 
news:slrnabs3mm.1kfm.Gareth.McCaughan at g.local:

> A mathematician, a physicist and an engineer set out to
> prove that all odd numbers are prime.
> 
> Mathematician: 3 is prime, 5 is prime, 7 is prime ...
> well, it's obviously a simple proof by induction. Trivial!
> 
> Physicist: 3 is prime, 5 is prime, 7 is prime, hmm,
> 9 isn't prime: experimental error. 11, 13, ... yeah,
> must be right.
> 
> Engineer: 3 is prime, 5 is prime, 7 is prime, 9 is prime,
> 11 is prime, 13 is prime, 15 is prime, ... [fades into distance]
> 
> (These days there ought to be a computer scientist in the
> mix too. Suggestions?)

Computer Scientist:
The computer scientist writes a program to test every odd number and see 
whether it is prime:
    i, allprime = 3, True
    try:
        while 1:
            if not isprime(i): allprime = False
            i += 2
    except OverflowError:
    	   pass # Ok, we tested everything we could
    print allprime and "Passed" or "Failed"

Still waiting for the results! (Python 2.2.1 of course)

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From brueckd at tbye.com  Wed Apr  3 12:33:42 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 3 Apr 2002 09:33:42 -0800 (PST)
Subject: Java to Python translation
In-Reply-To: <7xelhx3zz6.fsf@ruckus.brouhaha.com>
Message-ID: <Pine.LNX.4.33.0204030923150.5663-100000@ns0.tbye.com>

On 3 Apr 2002, Paul Rubin wrote:

> vio <vmilitaru at sympatico.ca> writes:
> > I need to translate a (rather voluminous 100MB) java program to
> > python. After a few searches on google, apparently the best tools
> > for this task are Lex-Yacc (Flex-Bison on linux). While I am working
> > on the Lex-YACC HOWTO, I would be very interested in hearing of
> > others' successful experiences/feedback at this. Any pointers of
> > interest would similarily be appreciated.
>
> That's crazy.  Don't even think about it.  If you want Python-Java
> integration, use Jython.

Actually, it might be really useful. There's a lot more Java programmers
out there (and a lot more money being spent on Java in general), so it
would be pretty cool to have a semi- or fully-automatic way to leverage
some of the work that is being done in Javaland.

Case in point: right now I'm using Jython for some Oracle database work
only so I can use the pure-Java Oracle JDBC driver (setting up Oracle on
Linux so I can use the Python DB API is a mess). There's nothing
Java-specific in the JDBC driver, but I really doubt Oracle is going to
release a Python one any time soon. A port from Java to Python would be
*very* straightforward - the only library support you'd need would be for
sockets and maybe locale stuff.

-Dave





From mwh at python.net  Mon Apr 15 07:24:49 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 15 Apr 2002 11:24:49 GMT
Subject: Pseudo-TTY (pty module) help!
References: <mailman.1018809928.2389.python-list@python.org>
Message-ID: <lky9fpgpyo.fsf@pc150.maths.bris.ac.uk>

"Noah" <noah at noah.org> writes:

> Does anyone have any experience or samples of use of the pty module?

Not really.  I've used os.openpty (or was it os.forkpty?) which are
what I think the pty module emulates when the above are not around (no
idea when this is, either).

Here's something that I got to work:

def su():
    import os, tty, time
    def p(d=0.2): time.sleep(d)
    pid, fd = os.forkpty()
    if pid == 0:
        os.execlp("su","su")
    else:
        p()
        print `os.read(fd, 1000)`
        p()
        os.write(fd, "XXXX\r")
        p(2)
        print `os.read(fd, 1000)`

I don't know if this helps!

The only other suggestions are to post some code and read Stevens'
APUE.

Cheers,
M.

-- 
  /* I'd just like to take this moment to point out that C has all
     the expressive power of two dixie cups and a string.
   */                       -- Jamie Zawinski from the xkeycaps source


From rjones at ekit-inc.com  Tue Apr  2 19:49:08 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Wed, 3 Apr 2002 10:49:08 +1000
Subject: PEP 287: reStructuredText Standard Docstring Format
In-Reply-To: <7xr8lx3c5n.fsf@ruckus.brouhaha.com>
References: <B8CEABB2.20C5A%goodger@users.sourceforge.net> <7xr8lx3c5n.fsf@ruckus.brouhaha.com>
Message-ID: <200204030049.AAA11432@crown.off.ekorp.com>

On Wed, 3 Apr 2002 09:37, Paul Rubin wrote:
> David Goodger <goodger at users.sourceforge.net> writes:
> > Here's a serious proposal, safe to post now that April Fool's is over.
> > Please read & comment.
>
> Man, I thought it was an April fools joke when I saw it last night.
> All those `different` :kinds: 'of' "quoting" are much more confusing
> than explicit markup.

Are you speaking from experience of using the "markup" (which tends to be 
extremely unobtrisuve) or are you just bashing? Given that 'single' and 
"double" quoting have no effect? And :colons: aren't used for general 
document quoting, but rather structural information? And that `backquoting` 
generally won't be used in most documents?

For a _practical_ example of the "markup", please see the Roundup 
documentation mentioned in my response. I think you'll find that the original 
text is perfectly readable and significantly easier to edit and maintain that 
any "explicit" markup language. Apologies for the long 
  
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/roundup/roundup/doc/installation.txt?rev=HEAD

Note that my manual Table of Contents at the top is now possible using a 
directive::

   .. contents:: **Table of Contents**

The output is at http://roundup.sf.net/doc/installation.html


>  So I think it's better to use explicit markup.
> The ultimate explicit system is tangle/weave but that's a bit extreme.
> Javadoc has been very successful in practice and you could do a lot
> worse than just copying it lock stock and barrel.

Javadoc. Bleah. Completely unreadable in its source form - which is in the 
code. You need to be able to read the documentation when viewing the code!


> The reStructuredText stuff looks like POD all over again.  If you want
> POD, I'd say just use POD, rather than coming up with yet another
> incompatible format.

Please, read the justification text, and the anlysis of existing formats. 
POD, while not mentioned, shares a lot of the shortcomings of other STX 
formats in that it was grown over time.


> The thing I like least about the PEP is that it's written like
> marketing material--it sets out to push a particular solution and
> looks for things to bash about the alternatives rather than treating
> them evenhandedly and being willing to accept the idea that
> reStructuredText may not be the answer.

Again, have you _read_ the ReST documentation, specifically the discussion of 
alternatives? It's extensive, and very well written. It discusses many 
existing STX systems and their strengths and weaknesses.


>  Javadoc and POD are
> widely-deployed, practical systems; a lot of people use them and like
> them and there are REASONS for that.

They are defacto standards. That's not always the sign of a well-designed 
system.


     Richard



From bhan at andrew.cmu.edu  Tue Apr 30 16:20:00 2002
From: bhan at andrew.cmu.edu (Benjamin Han)
Date: Tue, 30 Apr 2002 16:20:00 -0400
Subject: Newbie: finding the key/index of the min/max element
Message-ID: <200204301620.00255.bhan@andrew.cmu.edu>

In my quest of becoming a "Pythonistas" (where does this word come from BTW?), 
here comes my 2nd trivial question: how do you find the key/index of the 
max/min element in a sequence/map? I know you can call max(l) but this only 
gives you the element itself, not the key. I also know it's not that hard to 
code it up the hard way (for k in d.keys(): ... # do comparison), but is 
there a more elegant way of doing this?

Thanks,

Ben


From logiplexsoftware at earthlink.net  Tue Apr  2 16:56:49 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 2 Apr 2002 13:56:49 -0800
Subject: wxpython: problem with events and checklistbox
In-Reply-To: <3caa2be5.48482250@news.easynews.net>
References: <3caa2be5.48482250@news.easynews.net>
Message-ID: <20020402135649.264611bf.logiplexsoftware@earthlink.net>

On Tue, 02 Apr 2002 22:16:06 GMT
Marcus Stojek wrote:

> HI,
> 
> in my GUI I have a wxCheckListBox (named CheckList) and an fuction
> IfChecked(self,event) that process the checkevents (EVT_CHECLISTBOX).
> 
> Now, when unpickling all data at the start of my program,
> I want to check some items with CheckList.check(index,true)
> My problem is, that after each checking the function IfChecked is
> called. In the docs I found something about PopEventHandler, but this
> doesn't seem to solve my problem. Again, this should be an all day
> problem, but I have no idea where to look.

How about not calling EVT_CHECKLISTBOX until /after/ you load the
wxCheckListBox?

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From stever at bic.mni.mcgill.ca  Wed Apr 17 09:59:49 2002
From: stever at bic.mni.mcgill.ca (Steve ROBBINS)
Date: Wed, 17 Apr 2002 09:59:49 -0400
Subject: 2.0*4 ==> Bus error on IRIX 6.5
Message-ID: <20020417095949.A1486861@shadow.bic.mni.mcgill.ca>

Hello,

I just built python 2.1.3 on an IRIX 6.5 machine (Origin 200) using MIPSpro
compilers (Version 7.30) and the test suite fails.

It turns out that multiplying a float by an integer causes a
bus error:

  $ ./python
  Python 2.1.3 (#2, Apr 17 2002, 09:37:30) [C] on irix646
  Type "copyright", "credits" or "license" for more information.
  >>> 2.0*4 
  Bus error (core dumped)


[Oddly, it works if you multiply two floats or two integers]

>From the core dump, I see that the fault happens here:

  #0  0x1004fd7c in float_mul () at floatobject.c:382
  382             CONVERT_TO_DOUBLE(v, a);


I looked on the python.org web pages, FAQs, etc and didn't
find any similar problem reported.  I couldn't find the way to
search the python bug tracker, so I had a quick look at the
summary lines.  There was one about alignment that caught my eye
(#472568) but the workaround suggested there does not ameliorate
the situation.

Help?





From ak at silmarill.org  Sat Apr 20 19:57:01 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 20 Apr 2002 23:57:01 GMT
Subject: Disappearing global variables
References: <slrnac3lc2.3hm.ak@ak.silmarill.org> <_Mkw8.5895$Dm.409622@bin2.nnrp.aus1.giganews.com> <slrnac3sau.4fs.ak@ak.silmarill.org>
Message-ID: <slrnac4065.55n.ak@ak.silmarill.org>

In article <slrnac3sau.4fs.ak at ak.silmarill.org>, Andrei Kulakov wrote:
> In article <_Mkw8.5895$Dm.409622 at bin2.nnrp.aus1.giganews.com>, Terry
> Reedy wrote:
>> 
>> "Andrei Kulakov" <ak at silmarill.org> wrote in message
>> news:slrnac3lc2.3hm.ak at ak.silmarill.org...
>>> You can look at full program at silmarill.org/pybook/pybook.py -
>> it's
>>> a little over 500 lines.
>> 
>> can you produce a considerable shorter (minimal) version that exhibits
>> the same problem?
>>
> I figured this out.. I had a datafile that held a list of Book()
> instances; then I changed Book() class a little, so here I was trying to
> use old instances in the file where class was changed. I don't quite
> understand why it deleted other instances from global namespace. I guess
> if you do something like this, outcome is unpredictable :P.
>
No, that wasn't that.. it worked at first but now I'm getting the same
error again. Odd.



-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From aleax at aleax.it  Tue Apr  9 06:17:19 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 10:17:19 GMT
Subject: Queus clarification
References: <3CB2BD77.9070708@beth.uniforum.org.za>
Message-ID: <Pczs8.75824$S52.2691705@news2.tin.it>

mixo wrote:

> 
> If 'Queue.put(some_method)' is called from some script,
> does the script that called that 'Queue.put' have to wait
> for it to finish execution?

If you have created the queue object (which you should
NOT call Queue -- that's the name of the module and class!)
with a bounded size, and the queue is full, and then you
call the put method this way, then you're asking that your
thread be stopped until some other thread makes the
queue not-full.  If you don't want the calling thread
to wait, call put_nowait instead of put (clearest),
and then you'll get a Queue.Full exception (which of
course you can catch with a try/except) instead of having
your thread wait.

If your queue is of unbounded size, put will not have
to wait,


Alex



From Oschler at earthlink.net  Tue Apr  2 17:27:06 2002
From: Oschler at earthlink.net (Robert Oschler)
Date: Tue, 2 Apr 2002 17:27:06 -0500
Subject: Associate PY files with Python (KDE/Linux)
References: <uaqq8.12467$ku5.1078856@e3500-atl2.usenetserver.com>
Message-ID: <Beqq8.12484$ku5.1079965@e3500-atl2.usenetserver.com>

Sorry that was a dumb one, chalk it up to Linux newbie-ness.  The Browse for
app button was right on the KDE "Open-With" window.  Found Python and made
the association.  I wonder why some apps seem to be part of the KDE Start
Menu and others, like Python, aren't.  Curious

thx

Robert Oschler <Oschler at earthlink.net> wrote in message
news:uaqq8.12467$ku5.1078856 at e3500-atl2.usenetserver.com...
> I'm running KDE 2.x under SUSE 7.3 Pro.  I'd like to set up a file
> association so that when I click on a .py file in Konqueror it loads
Python
> with the py file ready to go.  Anyone know how?
>
> thx
>
>
>
>





From eppstein at ics.uci.edu  Thu Apr  4 12:07:55 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 04 Apr 2002 09:07:55 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com> <3CA9FD45.E6E33129@lemburg.com> <200204022145.g32LjAK20525@odiug.zope.com> <3CAB0BC1.20B90FED@lemburg.com> <200204031437.g33Eb2506920@pcp742651pcs.reston01.va.comcast.net> <3CAC149F.3F1BA77B@lemburg.com> <mailman.1017930740.20674.python-list@python.org>
Message-ID: <eppstein-AA64FA.09075504042002@news.service.uci.edu>

In article <mailman.1017930740.20674.python-list at python.org>,
 Guido van Rossum <guido at python.org> wrote:

> I don't think Boolean algebra is a big subject.  I bet that most
> people only learn about it explicitly when they learn computer
> programming.

There are advanced graduate math books on Boolean algebra.
For instance Halmos's _Lectures on Boolean Algebras_ or possibly 
Johnstone's _Stone Spaces_.  So the subject may be bigger than you think 
it is...

But these books' content is almost completely irrelevant for day-to-day 
computer programming.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From anthony at interlink.com.au  Tue Apr  9 11:32:24 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Wed, 10 Apr 2002 01:32:24 +1000
Subject: Booleans, integer division, backwards compatibility; where is Python going? 
In-Reply-To: Message from paul@boddie.net (Paul Boddie) 
   of "09 Apr 2002 06:26:24 MST." <23891c90.0204090526.18675d02@posting.google.com> 
Message-ID: <200204091532.g39FWOU09401@mbuna.arbhome.com.au>

>>> Paul Boddie wrote
> Perhaps there should be a community site which focuses on such news
> from the busy package maintainer's perspective, rather than the
> language architect's perspective. (Not to undermine the presumably
> excellent work already done in summarising the goings-on on the
> development list.)

Do amk's What's New articles not do this for you already?

Anthony





From mlh at vier.idi.ntnu.no  Thu Apr 11 22:25:43 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 12 Apr 2002 02:25:43 +0000 (UTC)
Subject: Tuple comprehension
References: <mailman.1018539023.28081.python-list@python.org>
Message-ID: <slrnabchd7.c8l.mlh@vier.idi.ntnu.no>

In article <mailman.1018539023.28081.python-list at python.org>,
brueckd at tbye.com wrote:
[snip]
>I use list comprehensions all the time, but I have trouble coming up with
>too many cases where tuple comprehensions would be useful. Maybe I'm
>missing a large set of potential uses?
[snip]

Well, at the moment I'm working with coordinates and various
aggregates (related to Anygui), and would like to have tuples as the
"base case" for those, e.g.

  foo.geometry == 10, 10, 100, 100

I might want to calculate that based on a rule where I have something
like 

  names == ['x', 'y', 'width', 'height']
  state = {'x': 10, 'y': 10, ...}

In this case, a tuple comprehension would be useful to me. Not
essential, of course, but then again, neither would a list
comprehension. It just seems natural to do

  geometry = (state[name] for name in names)

(I'm assuming that the added parentheses would signal the tuple
comprehension, just like the added brackets will signal generator
comprehension.)

> -Dave

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From neuhauser at mail.cz  Tue Apr 30 04:54:44 2002
From: neuhauser at mail.cz (Roman Neuhauser)
Date: Tue, 30 Apr 2002 10:54:44 +0200
Subject: Comment on draft PEP for deprecating six builtins
In-Reply-To: <o5ez8.77579$vF6.2303394@news2.tin.it>
References: <aaim48$7fd$1@bob.news.rcn.net> <mailman.1020083690.16055.python-list@python.org> <Xns91FF939823C7Dduncanrcpcouk@127.0.0.1> <mailman.1020089751.18057.python-list@python.org> <o5ez8.77579$vF6.2303394@news2.tin.it>
Message-ID: <20020430085444.GF325@roman.mobil.cz>

> From: Alex Martelli <aleax at aleax.it>
> Subject: Re: Comment on draft PEP for deprecating six builtins
> To: python-list at python.org
> Date: Mon, 29 Apr 2002 15:59:16 GMT
> 
> Roman Neuhauser wrote:
> 
> >> The problem with int.chr() is that a lot of people are going to get
> >> confused when they find that "42.chr()" doesn't work (although "42
> >> .chr()" does).
> > 
> >     why? letters aren't valid decimal digits, are they?
> 
> Decimal digits have nothing to do with it.  The lexer does maximal-munch,
> so it snags off "42." as one first valid token (a float), "chr" as a
> second one, and so on -- unless you separate the digits from the dot,
> e.g. with whitespace, this is more or less inevitable.  And making the
> lexer complicated/slower/less predictable to allow "calling methods on
> integer literals" without space doesn't appear an attractive trade-off.

    Right, I didn't realize 42. was a valid float. The space would be a
    reasonable (beit tedious) solution. As long as this is highlighted
    in the docs, I don't see a problem.

-- 
FreeBSD 4.4-STABLE
10:47AM up 1 day, 1:13, 11 users, load averages: 0.02, 0.01, 0.00




From theller at python.net  Mon Apr 15 13:41:17 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 15 Apr 2002 19:41:17 +0200
Subject: Magic words
References: <45e6545c.0204120147.3707d410@posting.google.com> <Xns91EE1FA914746RASXnewsDFE1@130.133.1.4> <45e6545c.0204130449.d6cfd69@posting.google.com> <Xns91EFC59F4C450RASXnewsDFE1@130.133.1.4> <1104_1018820620@news.dsl-only.net> <Xns91F17F7FFB78Fduncanrcpcouk@127.0.0.1>
Message-ID: <a9f3cc$2o0m2$1@ID-59885.news.dfncis.de>

"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message news:Xns91F17F7FFB78Fduncanrcpcouk at 127.0.0.1...
> Scott David Daniels <Scott.Daniels at Acm.Org> wrote in
> news:1104_1018820620 at news.dsl-only.net:
>
> >> Seo Sanghyeon || Sat 13 Apr 2002 05:49:17a:
> >> > Thank you very much for pointers, but I can think of something not
> >> > listed there. For example, __all__ for "from module import *",
> >> > pickle protocols, __new__ for subclassing from a type and so on. Is
> >> > there a complete (or near-complete) list?
> >> ...
> >
> > You could also take this lack of a list as a charter for your
> > collecting _all_ (not __all__) such names, and creating a document for
> > others who, like you want to know the list of such names.  This could
> > include package references if you so desire (others will complain if
> > you leave something out, providing you with more fodder for your
> > document).
> >
>
> To get you started, here is a list built by grepping over the Python
> sources. Some of the names are probably red herrings, and there are
> probably some used by libraries not in this list, but it should give you a
> start:
[snip list]

or use my (not yet announced) Python Index search:

http://starship.python.net/crew/theller/pyhelp.cgi?keyword=__%5Cw*__&version=current&regexp=on

Thomas




From leisinmr at purdue.edu  Thu Apr 11 18:15:36 2002
From: leisinmr at purdue.edu (Matt)
Date: 11 Apr 2002 15:15:36 -0700
Subject: ASCII to HEX conversion
Message-ID: <34e98ec2.0204111415.504d246f@posting.google.com>

I need to write a program that converts ASCII test into the
hexidecimal equivalent.  It would help if it could convert to binary,
too.

Example:

ASCII = "A"
HEX = 41

I tried using the hex() function, but it does not work.  If anyone
could tell me how to do this without making a lookup table of all the
ASCII values.


Matt Leisinger
leisinmr at purdue.edu


From maxm at mxm.dk  Tue Apr 30 03:59:33 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 30 Apr 2002 09:59:33 +0200
Subject: Newbie question: how to join a list of elements of user-defined
 types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com> <mailman.1020132292.8898.python-list@python.org>
Message-ID: <3CCE4EE5.2090202@mxm.dk>

Cliff Wells wrote:
> On 29 Apr 2002 18:54:12 -0700
> Fortepianissimo wrote:

>>Basically I want to have a subclass of list, like
>>
>>class MyList (list):
>>    ...
>>
>>then I want to
>>
>>l=MyList()
>>... do stuff to fill l
>>s=" ".join(l)
> 
> 
> s = " ".join([str(i) for i in l])

Oh the joy and fun of optimizing code snippets, map() is really nice in 
this case:

     s = " ".join(map(str,l))

regards Max M



From wolfson at uchicago.edu  Tue Apr 30 00:10:06 2002
From: wolfson at uchicago.edu (Ben Wolfson)
Date: Mon, 29 Apr 2002 23:10:06 -0500
Subject: Third Draft:  Pep for Deprecating Builtins
References: <aak1uq$bgt$1@bob.news.rcn.net> <mailman.1020108174.25482.python-list@python.org> <3CCDB2EB.9E238601@geneva-link.ch> <MNiz8.32691$8D3.925320@news1.tin.it>
Message-ID: <pan.2002.04.29.23.10.04.870331.4649@uchicago.edu>

On Mon, 29 Apr 2002 16:19:40 -0500, Alex Martelli wrote:

> I'm surprised you miss the main advantage -- 0666, the Number
> of the Beast in octal, translates to decimal 438, the A.D. in which
> Emperor Theodosius II promulgated the Theodosian Code, the
> wildly antisemitic laws which, among other things, forbid Christians
> to even give alms to Jews, forbid Jews to hold State office or build
> synagogues, condemned to death any Jew converting a Christian.
> 
> Surely this is therefore a number which is often handy in the everyday
> practice of programming, and even thinking of eventually making a
> non-builtin of the precious Python function that helps us work with a
> numbering system which has such wondrous properties is unholy.

I'm afraid, Alex, that your otherwise brilliant analysis is fundamentally
flawed:  the Number of the Beast, viz. 666, is 01232 in octal, 1232 being
the year in which rocketry was used successfully to repel Mongol invaders
at the battle of Kai-fung-fu.  Clearly, this proves that Edward Teller is
the devil's spawn.  It is enlightening to recall once again in this
context the name Lucifer, and to remember that the brightest blast is
often the most destructive.

Especially in these troubled times, we should be thankful for the oct()
builtin, as every time we use it we are moved yet more to pacifism!

-- 
BTR    
BEN WOLFSON HAS RUINED ROCK MUSIC FOR A GENERATION
 -- Crgre Jvyyneq


From Chris.Barker at noaa.gov  Tue Apr 16 13:12:14 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 16 Apr 2002 10:12:14 -0700
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
References: <mailman.1018760667.27824.python-list@python.org> <slrnabjj1b.741.philh@comuno.freeserve.co.uk> <23kobu49iibfommpcvoickvc9pj481jvbe@4ax.com>
Message-ID: <3CBC5B6D.70427378@noaa.gov>

maxx at easynerws.com wrote:
> Why?
> 
> At my company we are stuck in the realm of Windows, and the IT staff must
> constantly patch the webservers to keep up with security and bugfixes. A quick
> check of the IT media will show management's displeasure over this continual
> process, but it does not appear to have hampered the deployment of Windows
> servers.

If Python had Microsoft's other advantages in the Marketplace, it would
be a Whole lot more popular. Imagine if Microsoft decided to drop VBasic
and start selling MSPython!

> As mentioned in elsewhere this thread, Java and VB, have and are undergoing
> complete architecture restructuring, and they still are expanding their
> territories.

Again, with the marketing put into JAVA, Python would be very popular indeed.

Personally, I don't have any problem with the rate of change of Python.
I started a projecta  while ago with 2.0, upgraded to 2.1 because of
rich comparisons, adn have held there. However, I just installed 2.2 on
mac OS-X, ran the code there, and had no problems.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From dalke at dalkescientific.com  Tue Apr 30 20:25:53 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Tue, 30 Apr 2002 18:25:53 -0600
Subject: OT: Happy Birthday Larry Niven
References: <3cce9f3a$1_6@goliath.newsgroups.com> <3CCEBD17.5040903@jerf.org> <3ccf161d$1_7@goliath.newsgroups.com>
Message-ID: <aanctd$960$1@slb6.atl.mindspring.net>

Brad Clements:
>As I recall.. Moties are essentially genetically designed for their
>particular task. .As in the case of the vehicle drivers who swerved around
>each other and looked like they were going to crash, but didn't.
>
>If they used Perl, they would crash.
>
>Besides, Motie engineers could recombine essential elements quickly to suit
>their needs, in an efficient manner. That smacks of "batteries included"
and
>I'd say Perl code could not be recombined..

I thought Motie ships were wierdly designed.  Ahh, here's a quote from
"The Mote in God's Eye", p70 in my paperback copy

  ... but some of the intact mechanisms were just as odd.  There was no
  standardization of parts in the probe.  Two widgets intended to do
  almost the same job could be subtly different or wildly different.
  Braces and <<page break>> mountings seemed had carved.  The probe was
  as much a sculpture as a machine.
  ...
  Every nut and bold in that probe was designed separately.
  ...
  "...The Motie engineers made two widgets do one job, all right, but the
  second widget does two other jobs, and some of the supports are
  also bimetallic thermostats and thermoelectic generators all in one.
  Rod, I barely understand the words.  Modules: human engineers work in
  modules, don't they?"
    "For a complicated job, of course they do."
    "The Moties don't.  It's all one piece, everything working on
  everything else."

'Course, they ain't got nothing on Pak Protectors, who could single-
handedly build a fusion powered interstellar ship if given enough yams
to live on.

IMO, neither species would use Python.

                    Andrew
                    dalke at dalkescientific.com





From logiplexsoftware at earthlink.net  Tue Apr  2 12:46:24 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 2 Apr 2002 09:46:24 -0800
Subject: duelling pythons
In-Reply-To: <m3d6xibyfn.fsf@nbecker.fred.net>
References: <1eed326b.0203312131.2ef13b62@posting.google.com>
	<mailman.1017695170.25885.python-list@python.org>
	<m3d6xibyfn.fsf@nbecker.fred.net>
Message-ID: <20020402094624.5cdea6ca.logiplexsoftware@earthlink.net>

On 01 Apr 2002 21:56:44 -0500
nbecker at fred.net wrote:

> Keep lobbying Redhat.  The correct solution is for them to put
> 
> #! /usr/bin/python1.5
> 
> If they need a specific version.  I argued this with them already.
> Keep up the pressure.  Maybe they will get it.

Absolutely agree.  Or better #!/usr/bin/python.rh with a link python.rh ->
python1.5 so they can eventually upgrade without changing every shebang.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308



From max at alcyone.com  Wed Apr  3 14:26:15 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 03 Apr 2002 11:26:15 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com> <3CAB0015.6563C468@python.org>
Message-ID: <3CAB5757.6EFD1303@alcyone.com>

Guido van Rossum wrote:

> That's an interesting response, because it highlights what I think is
> the main misunderstanding about the PEP.  You should think of it a
> very *small* change!
	...
> This is long enough of a response.  To those people who complain that
> I don't take their long thoughts seriously, I can only respond that I
> thought long and hard about this PEP too, and am similarly
> disappointed
> that it faces so much opposition based upon what *might* happen.

Indeed, I'm really surprised by the length of the responses and the
dissent against this PEP.  It seems to me a very simple change, simply
helping to clarify and codify an existing practice.  It seems highly
unlikely that any significant amount of code will break; code that
_does_ break would seem to be doing un-Pythonic things that go against
many of the very arguments that are being used to shoot down this PEP
(e.g., Python truth is something-vs.-nothing, etc.).

The something-vs.-nothing objection seems to be missing the point.  The
addition of bool does not affect what truth and falsity mean in Python,
it just adds another explicit set of values for use when that is all you
mean to talk about.  This supports more self-documenting, and instantly
clear code -- if I see True instead of 1, I know exactly what is
expected and what the possible values are.  bool won't affect the
meaning of "truth" in Python at all, so these objections seem to be red
herrings.

I'm really surprised at the volume of objections to a change that can
only help make code clearer and will have very few compatibility
ramifications.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From opengeometry at yahoo.ca  Sun Apr 28 16:05:13 2002
From: opengeometry at yahoo.ca (William Park)
Date: Sun, 28 Apr 2002 16:05:13 -0400
Subject: Scripting lanuages for laboratory tests
In-Reply-To: <aagkfm$6kf$1@news5.svr.pol.co.uk>; from etienne@coetzeee.freeserve.co.uk on Sun, Apr 28, 2002 at 12:01:33PM +0100
References: <aagkfm$6kf$1@news5.svr.pol.co.uk>
Message-ID: <20020428160513.A4605@node0.opengeometry.ca>

On Sun, Apr 28, 2002 at 12:01:33PM +0100, Etta Makwetta wrote:
> To whom it may concern
> 
> I am tring to decide on which scripting language to use in one of our
> laborities during testing procedures, and I don't really know what the
> real advantages and disadvantages are.  I have made the following list,
> and would like to know if anybody knows of any obvoius other
> advantages/disadvantages that Python has compared to Tcl.  Any other
> comments about Ruby, Perl etc.  are also welcome.

Sorry to throw cold water, but it all depends what you want to do.
"Laboratory testing" can mean anything...  

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-cpu Cluster, Hosting, NAS, Linux, LaTeX, Python, vim, mutt, tin




From gerhard at bigfoot.de  Fri Apr 12 23:27:07 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 13 Apr 2002 03:27:07 GMT
Subject: Initilizing MD lists w/ objects.
References: <Xns91EEC9BB14699RASXnewsDFE1@130.133.1.4>
Message-ID: <slrnabf9cb.nub.gerhard@lilith.my-fqdn.de>

Philip Swartzleonard wrote in comp.lang.python:
> I'm looking for a reasonable way to initilize a multidimensional list with 
> a bunch of blank objects of some type (already been burned by using [obj
> ()] * 400 or similar =). This is the best i've come up with, but it seems 
> icky:
> 
>         self.data = [None]*40
> 
>         for x in range(40):
>             self.data[x] = []
>             for y in range(40):
>                 self.data[x].append(Vis_cell())

What about list comprehensions, like:

self.data = [[Vis_cell() for column in range(40)] for row in range(40)]

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From emile at fenx.com  Wed Apr 24 16:54:03 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 24 Apr 2002 20:54:03 GMT
Subject: Quick: Format Specs in %
References: <Xns91FA8B766EF98RASXnewsDFE1@130.133.1.4>
Message-ID: <LXEx8.46927$CH.7101@rwcrnsc52.ops.asp.att.net>

"Philip Swartzleonard" <starx at pacbell.net> wrote in message
news:Xns91FA8B766EF98RASXnewsDFE1 at 130.133.1.4...
>
> '%2f' % some_number pads the number with 2 blank spaces
> '%.2f' % some_number fixes the number to 2 decimal places
>
> the problem:
> '%2.2f' % some_number only does the second of those two things. I need
> both. I tried looking up the % operator in the index of the language
> reference, but couldn't find it.


Are you looking for %8.2f ?

--

Emile van Sebille
emile at fenx.com

---------



From sjmachin at lexicon.net  Tue Apr 30 09:19:09 2002
From: sjmachin at lexicon.net (John Machin)
Date: 30 Apr 2002 06:19:09 -0700
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org> <aact9h$9odgq$2@ID-99513.news.dfncis.de> <c3d20384.0204292330.ea14313@posting.google.com>
Message-ID: <c76ff6fc.0204300519.45f00bff@posting.google.com>

jidanni at hotmail.com (Dan Jacobson) wrote in message news:<c3d20384.0204292330.ea14313 at posting.google.com>...
> > This will help you get started:
> > 
> >         http://pythonzh.sourceforge.net/
> 
> I see nothing there.

Indeed. Norwegian Blue Parrot?


From cliechti at gmx.net  Fri Apr  5 17:02:12 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 5 Apr 2002 23:02:12 +0100
Subject: Newbie's 1st Web App decisions!
References: <mailman.1018028292.11988.python-list@python.org> <a8kuol$19l$1@hfc.pacific.net.hk>
Message-ID: <Xns91E7EA86BCD53cliechtigmxnet@62.2.16.82>

Scrumpy <reply at in_newsgroup.tks> wrote in
news:a8kuol$19l$1 at hfc.pacific.net.hk: 
...
> I only pointed out a personal preference i.e. I don't like to mix app
> logic and html layout in the same page, as can be seen in many pages
> with PHP. 

when the pages are static you could consider http://ht2html.sf.net
there you can write the body of the page in a .ht file and the ht2html tool 
wraps it up with nav and banner etc to a html.

the design is embodied in some python classes which you can adopt for youu 
own design.

it's very nice, i use it for some pages.
(and i have extened it to include py code in the html pages :-)

...
> Although it is nice to be told about this or that web app related
> project, I'm not in a position to know if it's better or worse than
> the alternatives. I'd really appreciate hearing (relatively) unbiased 
> comparisons between projects and perhaps being given a slight push
> towards one.

beeing unbiased is a bit difficult ;-) most of us had good or bad 
expeiences with a tool so you'll always get a hint in a certain direction.

i think you can gather some tips which tools are simple or complicated and 
then there is nothing better than trying them out yourself.

chris
-- 
Chris <cliechti at gmx.net>



From martin_byrne2 at yahoo.co.uk  Mon Apr 15 06:47:19 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 15 Apr 2002 03:47:19 -0700
Subject: KeyboardInterrupt errors in ASP pages
References: <508cb8ae.0204110045.431dadee@posting.google.com> <3CB550C9.1000902@mxm.dk> <3CB83C5C.9040704@skippinet.com.au>
Message-ID: <508cb8ae.0204150247.38738070@posting.google.com>

Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<3CB83C5C.9040704 at skippinet.com.au>...
> Max M wrote:
> > B wrote:
> > 
> >> I am using Python as the language for creating server side asp pages.
> >> The problem is that when using Response.Redirect(page), I seem to get
> >> the message below randomly but frequently enough to be a problem.
> >>
> >> Has anyone else found this problem and is there a way around it. Is it
> >> a bug in the Response.Redirect code?
> >>
> >> Python ActiveX Scripting Engine error '80020009'
> >> Traceback (innermost last): KeyboardInterrupt
> >> /Default.asp, line 0
> > 
> > 
> > As far as I understand it's a known bug, but without any known solution. 
> > I get it too at times.
> > 
> > I have just signed over a project to a customer on .asp/iis so I am a 
> > bit worried about it. But I don't really know what to do about it.
> > 
> > Maybe I should write a test harness to pound the IIS to see if I can 
> > provoke the error.
> 
> Yes please - that is exactly what I need.  I need some way to provoke 
> the error often enough that I can afford to "wait" for it in a debugger. 
>   Anything you can do to get a reliable repro would be a big help!
> 
> Mark.


Check out 
http://bugs.activestate.com/show_bug.cgi?id=10204

One of the entries,the third last one, is basically the same as whats
happening to me. Considering that the example has only
Response.Redirect in it, it would seem fairly certain that the
Response object is the culprit. However its not infeasible that
something deeper down is causing the problem which then manifests its
self as a keyboard interrupt during a Response.Redirect.


From sholden at holdenweb.com  Wed Apr 17 16:44:45 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 17 Apr 2002 16:44:45 -0400
Subject: preg_replace?
References: <a9kkpi$mep$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <2blv8.6760$Ip3.1344@atlpnn01.usenetserver.com>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote ...
> I am porting a program that I already wrote in VBScript and PHP.
> My current task: I need to replace everything inbetween square brackets
with
> "0". In VBScript I did this "by hand" (looping, inString and so on). In
PHP
> somebody helped me with the following solution:
>
>     $s = preg_replace("%\[[^\]]*\]%", "0", $s);
>
> Even though I've been reading through regex-tutorials in the meantime, I'm
> still not familiar with it. But is there a simple rule how to convert the
> above into Python code? Not only would it solve my current problem but I
> already got some helpful regex's for PHP and could reuse them in the
future.
>
> Anyway, the best (or worst) I did so far is:
>     pattern = re.compile(r'(\[[^\]]*\])')
>     pattern.sub('0', s)
>
> I'm not convinced this remotely resembles how it's done in Python; it
> doesn't cause an error, doesn't do what I want either.
>
> Here's my hoped before after:
>     s = "Hello [test] world [bla bla]..." # before regular expression
>     s = "Hello 0 world 0..." # after
>
Well, what's wrong with simply replacing the second statement with

    pattern.sub('[0]', s)

Seems like the simplest solution to my simple mind ...

regards
 Steve






From sholden at holdenweb.com  Sat Apr 27 12:31:15 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 27 Apr 2002 12:31:15 -0400
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org> <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com> <slrnacb5ko.4i6.akuchlin@ute.mems-exchange.org> <b0f083db.0204231422.785cb57c@posting.google.com> <7llx8.2088$0x3.1467@atlpnn01.usenetserver.com> <aaa5fu$2u4@netnews.hinet.net>
Message-ID: <SpAy8.57312$z55.12064@atlpnn01.usenetserver.com>

"Jeff Pitman" <bruthasj at yahoo.com> wrote in message
news:aaa5fu$2u4 at netnews.hinet.net...
> Steve Holden wrote:
>
> > "Jeremy Hylton" <jeremy at alum.mit.edu> wrote ...
> >> "A.M. Kuchling" <akuchlin at ute.mems-exchange.org> wrote ...
> >> > In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>,
> >> > Steve Holden wrote:
> >> > > Since urllib knows nothing of cookies, you will need to integrate
> >> > > some
> > sort
> >> > > of a cookie jar into the library, with a new API  for the clients
to
> >> > > retrieve and store the cookies.
>
> Or do it transparently.
>

Well, the intention was that cookie operation should be transparent, and
non-existent in the case where the default cookie jar was used. But you've
obviously divined my intent :-)

>
> >> > This is worthwhile, but I don't think it belongs in urllib.  It
> >> > belongs in a module or package of its own that provides general
> >> > Web-browser features such as cookies, remembering authentication
> >> > usernames and passwords, and a cache.  This package could then be
used
> >> > for implementing HTML-scraping scripts, spiders, or a Web browser.
>
> No, it belongs in urllib2 because it uses recursion to open websites that,
> for example, redirect from http://site/index.php to
> http://site/login_page.php to  https://site/login_page.php.  All I can say
> is "good luck!" trying to intervene outside of the package without
> rewriting AbstractHTTPHandler.
>
> >> I'm not sure what the difference between an HTTP client, like urllib
> >> or urllib2, and a Web-browser is.  Other than urllib's monolithic
> >> design, why wouldn't you want these sorts of features in the module?
>
> Exactly!
>
> > Personally I imagined passing a dictionary as an optional cookie jar
> > argument, keyed by (domain, path) tuples. The library code would update
> > this as dictated by its interactions with web sources.
>
> This is what I did:
>
>
http://sourceforge.net/tracker/index.php?func=detail&aid=548197&group_id=547
0&atid=305470
>
> And, right now, it is transparent to anyone using urllib2.  It uses a
> persistent Dict (within a script who imports urllib2) that keys on the
> hostname and stores a "Cookie" object.  This Cookie object is scraped
every
> time Cookies are sent from the server to the client.  The Cookie object is
> then consulted and its headers sent on each request to the server.
>
> Obviously this is v0.0.0.1, but I think it is a start of something
useable.
> I'm trying to create a library that "HTML-scrapes" websites and while
doing
> so I hit a brick wall with Cookies.  This library is going to be similar
to
> related perl scripts, except much cleaner.
>
> Sample screen-scrape:
>
>     ua = HTMLAgent( "http://www.yahoo.com/" )
>     ua.report()
>
>     form = ua.getFormByIndex( 0 )
>     form.fill( 'q', 'python' )
>     ua.submit( form )
>
>     ua.report()
>     ua.clickByName( 'Python Language Website' )
>
>     print "Now at", ua.location.geturl()
>
> So far so good with the library I've written so far. Except it can be slow
> as it uses minidom to parse the HTML.  And, I'm a newbie at this stuff so,
> I don't know how to limit the parse to only <form></form> tags in the
> screen-scrape process.
>
> I'll clean it up a little tonight and drop it somewhere if you want to
look
> at it.

Jeff:

I think this is excellent progress, and would love to see the code (though I
might not have time until Monday, so don't bust a gut).

If you want some code as a start for the cookies I have some lying around,
though it won't give you everything. It does, though, at least read & write
cookies. There's also an example using htmllib to extract table information
only, which might readilt yadapt to your needs (or at least give you food
for thought).

Mail me if you need either of those.

--

home: http://www.holdenweb.com/
Python Web Programming:
http://pydish.holdenweb.com/pwp/







From phlip_cpp at yahoo.com  Wed Apr 17 17:24:09 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 17 Apr 2002 14:24:09 -0700
Subject: Host a Gtk widget inside a Tk frame
Message-ID: <63604d2.0204171324.3b8c4d6f@posting.google.com>

Xists:

I have this friend who has written a large GUI using only Tkinter.

Then he discovered a must-have widget. But unfortunately this widget
binds to Gtk. (The Widget in question is VPython, but that's not
important right now.)

My friend thinks the best way would be to extract the X windows ID
from a Tk Frame, wrap this in some way that Gtk understands, and
present it to the Gtk control as its parent handle. The widget is
fairly stand-alone, so it expects only hosting behavior from its
parent. But my friend's customers expect a seamless interface.

How can I talk my friend out of this self-destructive behavior? (Or is
there actually a way to do that??)

-- 
  Phlip
   http://www.greencheese.org/EvolutionaryPsychology

"We may eventually come to realize that chastity is no more a virtue
than malnutrition."  -- Alexander Comfort


From alexis.layton at post.harvard.edu  Mon Apr  1 13:46:02 2002
From: alexis.layton at post.harvard.edu (Alexis Layton)
Date: Mon, 01 Apr 2002 18:46:02 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <010420021349434531%alexis.layton@post.harvard.edu>

I am generally in favor of PEP 285: Adding a bool type.

Like most, I believe the capitalized versions of the truth values
"True" and "False" should be used.

I believe str(True) should not be "1", but rather "True" or whatever
other localization should be in effect once python is properly
internationalized... (<wink>?)

repr(True) should always be "True".

As for int-compatibility, were we designing Python from the ground up
it might be the case that it should be designed as a pure truth type. 
But we are constrained by Pythons 1-2.2.x so True has to be
truly-comparable to 1.

-- 
Alexis Layton
alexis.layton at post.harvard.edu


From st-newsgroups at thomas-guettler.de  Wed Apr 10 09:58:46 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Wed, 10 Apr 2002 15:58:46 +0200
Subject: Create Excel Files without COM
Message-ID: <3CB44516.60005@thomas-guettler.de>

Hi!

Is there something like WriteExcel from perl in python?
I need to generate an MS-Excel file. I don't want to use
COM although it will run on windows.

The data is simple. Just text in columns and rows.

  thomas



From paul at boddie.net  Tue Apr  9 04:43:25 2002
From: paul at boddie.net (Paul Boddie)
Date: 9 Apr 2002 01:43:25 -0700
Subject: Newbie - web app deployment issues
References: <a8skcd$mr5$1@hfc.pacific.net.hk>
Message-ID: <23891c90.0204090043.287c317f@posting.google.com>

Scrumpy <reply at in_newsgroup.tks> wrote in message news:<a8skcd$mr5$1 at hfc.pacific.net.hk>...
> Hi all,
> 
> Thanks for your feedback to my earlier post "Newbie's 1st Web App 
> decisions!"

It probably looks like I'm tracking down all your posts to this
newsgroup, but then it's a solid topic you're bringing up and I can't
resist responding. ;-)

> As a Python newbie who wants to try developing web apps, I want to avoid as 
> much as possible, the need to deploy to specialized hosting environments. 
> For example, out of the many hosting companys in the world, only a very 
> small percentage provide Zope hosting. I haven't confirmed this but there's 
> probably a premium to pay for Zope hosting (which may be warranted but it's 
> not something I want to pay for at this point in time).

I've seen http://www.freezope.org, but that seems really to be for
experimentation only. However, you could always choose a very-low-cost
Web host which only provides static Web pages and then link to dynamic
content on FreeZope or other such sites.

> Some web hosting companys are more accomodating than others i.e. after 
> evaluation, some may add customer requested/coded python modules, others 
> will refuse and others will not install Python at all.

Some hosting companies will let you install whatever programs you
like, provided that they don't run as long-running processes. The
positive aspect of this is that provided you know the architecture of
the server, you can attempt to obtain binaries for Python and other
programs from various sources [1]. The downside is that any
restrictions on long-running processes rules out a number of the Web
frameworks available for Python.

> At least in the near future, hosting my own internet site or having a whole 
> server rented/co-located at an ISP/hosting company is not an option so I 
> have to rely on what is provided by, or will be accepted by hosting 
> companys on their existing web servers.

Yes, for experimentation or for having a low cost personal site, many
hosting options are too expensive or seem unjustifiable to people like
me. ;-)

> How did you get your Python based web app hosted i.e. what 
> modules/components/frameworks did you use to run your app, which were 
> allowed by a 'non-specialized' web host i.e. a host that probably runs 
> Apache and has Python and a few modules preinstalled?

I've recently installed Python on a FreeBSD host, along with the
MySQLdb module, and it is my intention to install one of the "more
embedded" Web frameworks as well. I think I may use the jonpy package
(http://jonpy.sf.net), although Albatross also seems worth looking at.

> I would be 'most' interested to hear from people using the project(s) 
> listed below but I would also appreciate feedback if you have a good 
> solution which doesn't include them.
> 
> The web app servers/frameworks that I am looking into in more detail are:
> 
> Slither http://slither.toolsofcomputing.com/
> Webware for Python http://webware.sourceforge.net/
> Skunkweb http://skunkweb.sourceforge.net/about.html

Webware and SkunkWeb use long-running processes, although there is an
option with Webware to start and stop the application server for each
request; naturally, this is hardly recommendable. I still haven't
looked into Slither yet.

Paul

[1] mxCGIPython is a good start for those wanting a binary, but I then
discovered the FreeBSD "ports" system (see http://www.freebsd.org) and
it's possible to find lots of binaries for download in .tar.gz format,
including various Python extension modules (MySQLdb, for example) -
it's very convenient indeed!


From marklists at mceahern.com  Wed Apr  3 13:26:32 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 3 Apr 2002 12:26:32 -0600
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: <Pine.LNX.4.33.0204031118110.16190-100000@server1.lfw.org>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEAGKJAA.marklists@mceahern.com>

[Ka-Ping Yee]
> What about this:
>
>     >>> d = {}
>     >>> d[0] = 'zero'
>     >>> d[False] = 'false'
>     >>> len(d)
>     1 or 2?

This returns:

  1

Just curious:  If you were using True and False as keys to a dictionary,
would you ever expect len(d) to return something > 2?

// mark





From bokr at oz.net  Tue Apr  9 05:07:13 2002
From: bokr at oz.net (Bengt Richter)
Date: 9 Apr 2002 09:07:13 GMT
Subject: calling class methods as functions
References: <3CB2A038.2468F842@gol.ge>
Message-ID: <a8ub01$o0c$0@216.39.172.122>

On Tue, 09 Apr 2002 10:03:04 +0200, Giorgi Lekishvili <gleki at gol.ge> wrote:

>Hello!
>
>I wonder if someone explaines tom how can I overcome the following
>problem in Python2.0 (I have to use this interpreter):
>suppose, we have 2 classes, B, and C.
>
>in class B we have a method "write"
>
>in class C we have a method "write"
>
>def write (self, arg):
>    import B
>    res=B.B.write(self, arg)    #res is a string being either "ok" or
>"failure"
>    del B
>    return res
>
>
>The interpreter gives an error, that class isntance must be given as 1st
>argument.
>
>What's wrong?
>
>Thanx,
>Giorgi
>
>PS. I know that the best solution is to make a baseclass and put the
>method "write" there. Of course. But I would like to know why this
>solution fails.
>
I can't tell what you mean by "this solution" unless you post actual code.
Just make a minimal example and show how it fails by copying a session
directly with copy/paste. Showing the call that gives the error.

Regards,
Bengt Richter


From johnroth at ameritech.net  Wed Apr  3 06:42:10 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 3 Apr 2002 06:42:10 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it>
Message-ID: <ualqa3n4dlof64@news.supernews.com>

"Alex Martelli" <aleax at aleax.it> wrote in message
news:IDAq8.45703$pT1.1295831 at news1.tin.it...
> Paul Rubin wrote:
>
> > Jeff Shannon <jeff at ccvcorp.com> writes:
> >> > I am opposed to the addition of the new proposed type on the
grounds
> >> > that it will make Python harder to teach both to people who have
never
> >> > programmed before and to people who _have_.
> >>
> >> At first glance, I was marginally in favor of this PEP.  After
> >> reading Alex Martelli's arguments, I was marginally opposed to
> >> it.  I think that Laura's comments are *very* convincing, and I'm
> >> now fairly solidly opposed to it.  Well said, Laura.
> >
> > I agree, I missed Laura's post earlier but it's terrific.
>
> Full agreement.  But not from Guido, who dubs it an "over-long FUD
> piece" -- all others who commented on Laura's post, whether they
> agreed with it or not, seemed to appreciate its worth.  I guess that
> tells us something about how worthwhile it is to spend hours studying
> and reflecting in depth upon issues, and composing beautiful posts
> on the subject, when arguments are requested: the message is quite
> clear, "if you don't agree with what he's already decided to do,
> don't bother, you'll only peeve him off".  Nice PR exercise...;-).

Maybe my post in response didn't propagate? As I said in that
post, the root of her problem is a poorly chosen name for the
function. If she paid attention to constructing the name properly,
it would come out as something like "numberOfLeapDays"
rather than "isLeapYear", and there would be **no**
instructional issue.

John Roth
>
> Oh well.
>
>
> Alex
>




From claird at starbase.neosoft.com  Fri Apr 26 15:41:04 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 26 Apr 2002 14:41:04 -0500
Subject: Simple bar graphs
References: <Xns91FCD43535458Latzikatz@139.174.2.56>
Message-ID: <6CE46680FDC41393.A2EE7694A5B94610.A362BEFCF5BCB5FA@lp.airnews.net>

In article <Xns91FCD43535458Latzikatz at 139.174.2.56>,
Lutz Schroeer  <Lutz.Schroeer at kybernetik-manufaktur.de> wrote:
>I'm looking for a _pure_ Python module which generates some very simple bar 
>graphs and outputs them as gif or jpg. I looked through Parnassus and other 
>sources but could not get enough information to make a decision.
			.
			.
			.
Would something closely related satisfy you?  There's
a "trick" for having the browser itself synthesize the
image.  Coincidentally, I'm just starting to document
an example implementation, but in Tcl <URL: http://
wiki.tcl.tk/barchart >.  If this model suffices for
you, I'll do a Python example, too.  It takes more time
to explain it than to code it.
-- 

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


From cbbrowne at acm.org  Sat Apr 20 20:24:39 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: Sat, 20 Apr 2002 20:24:39 -0400
Subject: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83>
Message-ID: <m3bscdlwzc.fsf@chvatal.cbbrowne.com>

In an attempt to throw the authorities off his trail, Lumberjack <lumber at jack.com> transmitted:
> aahz at pythoncraft.com (Aahz) wrote:
>> What's amazing about Lisp is that it's still in many ways on the cutting
>> edge of computer science.
>
> Perhaps you should name these "many ways". Taken at face value, your 
> statement pretty much claims that computer science hasn't progressed in 40 
> some years. I can at least agree with that statement. Of course the 
> "science" in "computer science" is pure bovine excrement. If you can't or 
> wont measure it, it isn't science. "Computer religion" or "computer cults" 
> would better capture the reality.

Well, after 40 years, we're still in the process of seeing features
Lisp had years ago move into the mainstream.

-> People are starting to get the idea that maybe computers are better
   at managing memory than people are, so that rather than doing it
   _completely_ manually using malloc()/free(), or getting "clever" by
   attaching it to virtual function classes, or getting "very slightly
   clever" by attaching reference counting to some allocations, maybe
   the environment should do generational GC for you.

-> Newer languages increasingly feature data structures like lists,
   hash tables, and such.

-> There has been a dramatic dearth of progression in the languages
   commonly used vis-a-vis syntax handling.  Algol (the _first_
   version) was of comparable sophistication, and likely _better_
   elegance, than most of the later languages based on it.

When looking at the more recent languages, while they may toss in
bigger libraries of "stuff," that's not evidence of them being
fundamentally wondrous progressions of the state of the art.  C++ is
newer than PL/1; is it _truly_ greatly superior, or is it just that
it's well-marketed, while nobody's pushing PL/1?

As a language, Java has little that _isn't_ in a lot of its
predecessors.  It may be nicer than C in a lot of ways, but for it to
have taken 30 years to do that seems rather wasteful.
-- 
(concatenate 'string "chris" "@cbbrowne.com")
http://www.cbbrowne.com/info/languages.html
Rules of the  Evil Overlord #128. "I will not  employ robots as agents
of  destruction  if  there  is  any  possible way  that  they  can  be
re-programmed  or if their  battery packs  are externally  mounted and
easily removable." <http://www.eviloverlord.com/>


From aleax at aleax.it  Tue Apr 30 10:47:40 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 14:47:40 GMT
Subject: Slice objects with negative increment
References: <aakcpj$5f$1@netnews.upenn.edu> <3kjz8.32779$8D3.928578@news1.tin.it> <aam88o$oec$1@netnews.upenn.edu>
Message-ID: <g8yz8.81675$vF6.2428847@news2.tin.it>

Paul Hughett wrote:

> Alex Martelli <aleax at aleax.it> wrote:
> : Paul Hughett wrote:
> :         ...
> :> Hence my question: Is there an established meaning within Python for
> :> slice objects with a negative increment?  E.g. slice(1, 10, -2) ?
> 
> : Sure!  [Use the same convention as Numeric]
> 
> Okay, I've looked up the convention for Numeric, but the description
> isn't entirely clear in the case of negative increments.  It states that
> the default values for start and stop when the increment is negative are
> "the beginning" and "the end" of the axis.  Suppose that the axis has
> n elements.  Does ::-k default to
> 
>    n:0:-k        which refers to element n, which doesn't exist
>    (n-1):0:-k    which skips element 0
>    n:-1:-k       which refers to element n, and apparently to -1 => n-1,
>    or
>    (n-1):-1:-k   which is not generally the reversal of 0:n:k for k > 1

>>> r=Numeric.array(range(13))
>>> r
array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12])
>>> r[::-1]
array([12, 11, 10,  9,  8,  7,  6,  5,  4,  3,  2,  1,  0])
>>> r[::-2]
array([12, 10,  8,  6,  4,  2,  0])
>>> r[::-3]
array([12,  9,  6,  3,  0])
>>>


> By the way, I don't see anything in "Numerical Python" about the
> meaning of a negative start or stop.  Do they follow the usual Python
> convention of meaning the nth element from the end?

>>> r[-3::]
array([10, 11, 12])
>>> r[-3::-1]
array([10,  9,  8,  7,  6,  5,  4,  3,  2,  1,  0])
>>>


Alex



From JamesL at Lugoj.com  Thu Apr  4 02:25:22 2002
From: JamesL at Lugoj.com (James Logajan)
Date: 4 Apr 2002 07:25:22 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org>
Message-ID: <Xns91E5EEE6FB57FJamesLLugojcom@216.148.53.85>

Guido van Rossum <guido at python.org> wrote:
[ ... ]
>     7) Should the name 'bool' be changed?
> 
>     => No.

Too bad. I believe his name was spelled Boole, not Bool or Boolean. You had 
a chance to properly honor him and didn't for some reason. Why drop a 
single letter of a dead man's name? This seems to be a pettiness that has 
gained universal appeal, alas. Shades of "creat" versus "create". Maybe 
there is some conspiracy against the letter "e" on the end of words?

[ ... ]
> Compatibility
> 
>     Because of backwards compatibility, the bool type lacks many
>     properties that some would like to see.  For example, arithmetic
>     operations with one or two bool arguments is allowed, treating
>     False as 0 and True as 1.  Also, a bool may be used as a sequence
>     index.

Just to be clear, what kind of object will this be when it is a key in a 
dictionary? If I have:

x = { 1: "I win", 0: "I lose" }

and do

x[2 == 1] = "Bummer"

what does the dictionary contain? Two entries or three? Is a bool value 
promoted to an int when used as a dictionary key?


From sholden at holdenweb.com  Mon Apr 22 09:51:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 09:51:38 -0400
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <20020419205101.17293.2839.Mailman@mail.python.org> <mailman.1019282437.10223.python-list@python.org>
Message-ID: <kCUw8.74285$T%5.6717@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> descended from the mountain to
confess...
> [...]
> I am probably in the minority here in that I have little patience for
people
> who are "confused" by or have difficulty with advanced language features.
> If you purport to be a professional programmer then you should know the
> base language cold.

Well, on this particular group I feel you *are* in the minority. Many c.l.py
readers are *not* "professional programmers", but chemists, mathematicians,
biophysicists, and a host of others for whom programming is a means to an
end. Many others are, like me, professional programmers who attempt to
increase Python's popularity by assisting those less familiar with
programming concepts.

Furthermore, you express exactly the kind of attitude that gets
"professional programmers" a bad name. Let's remember our social skills here
;-). Remember, Python is *not* just for professional programmers, so you
shouldn't write as though it were "your" language.

In summary: "professional" as a self-description leads people to expect
certain standards of behavior. Your income mightr derive from programming,
but that's only a *part* of being professional.

> So I am not swayed by arguments that :? confuses people.

So, if one is confused, one should leave the job of programming to the
professionals? That will help the programmer shortage.

> But I think part of the problem
> with inline choice is that people who use it do not properly
> parenthesize and indent it to properly highlight what's going on.
> I always split the test and the two alternatives on separate lines,
> with each of the latter two starting with the special operators.
>
>     def abs( n ):
>         return (( n > 0 )
>                       ? ( n  )
>                       : ( -n ))
>
I take it you're using some sort of "desirable" extension syntax here? I
completely fail to see how this is preferable to the instantly
comprehensible

    def abs(n):
        if n > 0:
            return n
        else:
            return -n

(except that this wouldn't confuse anyone). Although I'm told there's a
special place reserved in Hell for those who practice such abuse, it is of
course legal syntactically to rewrite this as

    def abs(n):
        if n>0: return n
        else: return -n

By the way, please note that there *is* a style guide for Python, which you
will, find at

    http://www.python.org/peps/pep-0008.html

It unequivocally states that spaces should not be used after opening
parentheses or before closing ones. So, while we're talking about "proper"
usages, let's get them *all* right, shall we?

> You'd never write if/then/else without proper indentation
>  (not even possible in Python ;o).  Same rules should apply
>  to complex expressions.
>
> Python's and/or semantics, although they may be employed
> to provide similar functionality, are more general, unrelated
> and in most cases an inappropriate substitute.
>
> Thus although you can use and/or as a partial substitute, there are
problems with the hack:

Indeed. An experienced (professional?) programmer would, of course, consult
the FAQ. See

    http://www.python.org/cgi-bin/faqw.py?query=ternary&req=search

for further discussion on this topic.
>
> Most importantly (a) the and/or idiom strays from the desired abstraction
and thus obscures the intent of the developer.
>
I agree with this. But don't try to suggest that there aren't problems with
the ternary construct in C, most notably the omission of an "=" in the test,
turning a test into an assignment.

> (b) Since and/or lacks clarity it arguably is more likely to
> people than inline choice.  And whereas :? confuses beginners,
> and/or (since it's an idiom rather than a language construct)
> is more likely to confuse more advanced users.

[I presume that should have been "...likely to confuse people ..."].
Frankly, I don't think a language that "confuses beginners" is likely to
develop a huge following, since everyione is a beginner at the start. I
think confusion of *any* kind is to be avoided, and Python does as well as
most languages in this respect. I particularly dislike your implication that
if anyone is to be confused it should be the beginners and not the
"professional programmers".
>
> (c) Worst of all, the idiom doesn't even work in all cases, thus admitting
the possibility of subtle, unintentional, seemingly intermittent errors,
e.g.:
>
>      x = ( len( s ) > 0
>                      and s[0]
>                      or -1 )     # FAILS!!  when s[0] == 0
>
Sorry, this is abuse. Also see the FAQ reference above.

> To me clarity of expression is the most important quality of a well
> written program.  Binary conditional evaluation ala ?: is indispensable.
>  It's eminently reasonable for programmers to feel a "need" for such
> an addition to Python.
>
To me, clarity of expression is the most important quality of a well written
program. Binary conditional evaluation ala ? is completely unnecessary. It's
eminently reasonable for programmers to feel a need for such an addition to
Python, but once they have been reeducated the feelings will go away.

> Biggest issue the syntax.

I disagree. The biggest issue is the bloat this will introduce into the
language, and programs written in it, without the least advantage. YAGNI.

> With ":" used the way it is in Python, C's :? syntax probably is not an
easy fit.  One possibility might be to reverse the two symbols:
>
>     x = ( len( s ) > 0
>             : s[ 0 ]
>             ? -1 )
>
> Having ":" separate the test expression from the two alternatives seems
sorta natural.
>
Blerch. This is supposed to be easy to understand? In which universe. I
submit it's completely unpythonic.

> Another alternative would be to borrow from Algol68 and (IIRC) Algol60.
They allowed if/then/else to be embedded within an expression.  Generalizing
if:/else: to return a value also seems somewhat natural, so perhaps that
could work.
>
>     x = ( if len(s) > 0:   s[ 0 ]
>                        else:    -1 )
>
> Since it's all within an expression, indentation doesn't matter.
>
> Just a thought....
>
Well, there were problems with Algol 60's conditional expressions, but I
agree that *in a free-form language* they are useful and fairly natural.
Python, of course, is not a free-form language, and so inline conditionals
of any kind are a far less natural fit with the rest of the language.

This isn't the first time this topic has been dragged out of retirement, and
I don't suppose it will be the last. It's probably obvious that your post
got up my nose -- I'm not normally so crotchety, so I guess I should
apologise for the cranky nature of my reply. I shall not be contributing
more to this thread.

maybe-i-should-have-another-cup-of-coffee-ly y'rs  - steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From dreed at capital.edu  Fri Apr 12 08:27:28 2002
From: dreed at capital.edu (Dave Reed)
Date: Fri, 12 Apr 2002 08:27:28 -0400
Subject: Python 2.2.1 on Solaris 2.5.1: compilation problem
In-Reply-To: <20020412144929.B12997@phd.pp.ru> (message from Oleg Broytmann on
	Fri, 12 Apr 2002 14:49:29 +0400)
References: <mailman.1018447222.24371.python-list@python.org> <uit6zvbur.fsf@python.net> <20020410192232.G19899@phd.pp.ru> <mailman.1018603073.26909.python-list@python.org> <lk4rihjgek.fsf@pc150.maths.bris.ac.uk> <20020412144929.B12997@phd.pp.ru>
Message-ID: <200204121227.g3CCRSd04250@localhost.localdomain>

> From: Oleg Broytmann <phd at phd.pp.ru>
> Date: Fri, 12 Apr 2002 14:49:29 +0400
> 
> On Fri, Apr 12, 2002 at 10:33:40AM +0000, Michael Hudson wrote:
> > > case $MAKEFLAGS in \
> > > *-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py -q build;; \
> > > *) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -O2' ./python -E ./setup.py build;; \
> > > esac
> > > running build
> > > running build_ext
> > > building 'struct' extension
> > > Segmentation Fault - core dumped
> > 
> > Yikes!  Can you dig into this?  I can't, not having access to a
> > relavent machine... (you did run make clean, didn't you?)
> 
>    I'll try. What is worse, Python 2.2 failed the same way :( It seems
> upgrading gcc from 2.8.0 to 2.95.3 did more harm than good :(((
> 
> Oleg.

Have you compiled earlier versions of Python on this system?

I compiled Python 2.1.1 on a Solaris 2.6 machine (don't remember which
version of gcc). The machine has since been upgraded to Solaris
2.7. I've successfully compiled python 2.2.1 on Solaris 2.7 using gcc
2.95.3.

% python2.2
Python 2.2.1 (#1, Apr 10 2002, 08:09:51) 
[GCC 2.95.3 20010315 (release)] on sunos5

Dave




From clspence at one.net  Mon Apr  8 15:43:17 2002
From: clspence at one.net (Chris Spencer)
Date: 8 Apr 2002 14:43:17 -0500
Subject: wxPython / Your Experiences in large project ???
References: <a8rp2u$ht8$02$1@news.t-online.com>
Message-ID: <vnq3buoibt8lbqlhi4vjoo4qobgfl6hi0v@4ax.com>

	Okay, I guess I should chime in here.
	I am a professional developer for a small software company.  My company
has been using Python for its projects for the past 18 or so months.  We have a
whole product suite written in Python.  Until recently, this product was 95%
console driven, even on our Windows port.  Customers have been asking for a GUI
to tie all the parts of our product together.
	For the past month, I have been developing a GUI Administrator program
for our product using wxPython.

Some statistics:
Code is hovering around 6000+ lines in 12 modules
Approx. 60% of that code is directly related to wxPython
Development time in man-hours: 280

Pros:
1.  It has incredible flexibility.  I've done some pretty weird stuff that I'd
plain just be afraid to try in MSVC++.
2.  Rapid development.  This project would've taken me nearly a year to do in
C++/MSVCRT.  At a ROUGH guess, it would've hovered around 30,000-40,000 lines.
3.  Robust library.  I used to dive into www.codeguru.com to grab source code
for custom controls so I could modify them to my needs.  I've felt no real
reason to "build" new controls to suit my needs with wxPython.
4.  Maintainability.  Those 6000 lines of code I've done are all pretty much in
the category of "self-documenting code".

Cons:
1.  BAD documentation.  They've taken the wxWindows library documentation, added
a few notes, and called it the wxPython documentation.  And the thing is, they
haven't kept it updated for more recent releases of wxWindows/wxPython, so
there's a lot of undocumentated features that just don't show up.  I had to
start using the wxWindows docs from the wxwindows site, and interpret (and
source-code dive) what a reasonable wxPython equivalent would be.  LOTS of
trial-and-error.

2.  This project would've taken half the time it eventually took if wxPython was
either documented better, or took greater pains to protect the pointers that
wxWindows sometimes uses.  I'm in Python fer chrissakes, I shouldn't care about
null pointers.  Yet the grid control kept GPF-ing on me for weeks until I
figured out that Python was "blowing away pointers" behind the scenes for
garbage collection.

3.  Bloat.  The support files alone for my program (the wxWindows .dll, the
wxPython .pyd files, etc) amount to about 6 megs of space.  The program's BASE
memory usage upon startup is 14 megs.  This might not seem so bad for a 6000+
line program until you realize that a *100* line program would STILL take up 14.
megs of memory.

4.  Did I mention that the documentation is both incomplete and odious?  I
should not have to constantly do a "find in files" to find out what parameters a
damned method call takes because the way wxPython implemented a method is
different than the wxWindows way and someone forgot to make a note of it in the
rubbish passed off as a "manual".

5.  I've had a REAL problem with conservative "window refreshes", where unless I
specifically tell each window to repaint itself, I get a lot of scroll bars
missing, a lot of windows being "grey" (shown without controls), Tab windows
without the tab headers, etc.  wxWindows/wxPython should be more aggressive
about repainting windows.  Even MCVCRT controls are better about repainting
themselves without having to be told that their render clipping planes are
dirty.

	The pros outweight the cons by a WIDE margin once you've climbed the
learning curve.  While on that slope, though, the lack of robust and complete
documentation will HURT you.

Chris.

On Mon, 8 Apr 2002 11:55:24 +0200, "Harald Schneider"
<h_schneider at marketmix.com> wrote:

>Hi,
>
>what are your experiences, using wxPython/wxWindows in large projects?
>
>Large means in my case a sales system, based on MySQL.
>I'm thinking about implementing it in wxPython, but I'm not sure about the
>stability of
>wxWindows and the wxPython wrapper ...
>
>The alternative way would be a project in Borland Delphi, which has proofed
>its
>stability over years ...
>
>So what do you think about it!
>
>Thanks a lot,
>Harald
>



From dopefiend at hotmail.com  Sat Apr  6 15:05:09 2002
From: dopefiend at hotmail.com (DopeFiend)
Date: Sat, 6 Apr 2002 21:05:09 +0100
Subject: fg News men photographs of film stars sunbathing nude or topless
 fgvq dt goe  rsvapvc
Message-ID: <a8nkek$d75$46@nsnmrro1-gest.nuria.telefonica-data.net>

The site with paparazzi photographs of actresses, sports stars, singers,
wrestlers and models caught on camera nude, topless, upskirt or nipple slip!

Follow my link to the site...

http://exposed.onehandsurfing.com/

Enjoy before they shut me down!





s ug  d dyfjarssd tfkgao ssa kssufu fso fgd rd daokfrdrs u  


From ak at silmarill.org  Mon Apr 15 14:20:19 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Mon, 15 Apr 2002 18:20:19 GMT
Subject: input() vs raw_input()
Message-ID: <slrnabm6im.a8q.ak@ak.silmarill.org>

input() must not be mentioned in any tutorials for new users. It belongs
somewhere in the depths of the thickest reference manual. Where do new
users *learn* about it? I don't remember seeing it in any tutorial and
yet it pops up all the time.


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From shalehperry at attbi.com  Sun Apr 28 21:53:58 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sun, 28 Apr 2002 18:53:58 -0700 (PDT)
Subject: Super-newbie syntax error question
In-Reply-To: <Cg%y8.5752$9F5.405701@typhoon.austin.rr.com>
Message-ID: <XFMail.20020428185358.shalehperry@attbi.com>

> 
>     f = file("/article.txt", "w")
>     f.write('Category: %s\nTitle: %s\nDescription: %s\nLink: %s\n\n' %
>       (self._category, self._title, self._descr, self._link))
>     f.close()
> 
>                 self._title = None
>                 self._link = None
>                 self._descr = ""
>                 self._category = ""
> 
> 
> 
> ...and here's the resulting error:
> 
> 
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "C:\PYTHON22\lib\rss2html.py", line 46
>     f = file("/article.txt", "w")
>     ^
> SyntaxError: invalid syntax
> 

Should be:

f = open() not file().




From cookedm+news at physics.mcmaster.ca  Thu Apr  4 14:11:37 2002
From: cookedm+news at physics.mcmaster.ca (David M. Cooke)
Date: Thu, 04 Apr 2002 14:11:37 -0500
Subject: Prototype in Python, Was: A faster Python?, Python compiler,
 Dylan,...
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53>
 <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr>
 <3CAAEF17.32333986@kfunigraz.ac.at> <lbh8z84nake.fsf@xihu.inria.fr>
 <15531.7864.698585.485436@12-248-41-177.client.attbi.com>
 <lbh7knolreu.fsf@xihu.inria.fr>
 <15531.9709.924953.953487@12-248-41-177.client.attbi.com>
 <mailman.1017850275.2134.python-list@python.org>
 <0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop>
 <mailman.1017906916.22194.python-list@python.org>
 <3CAC1456.C8A51DC9@kfunigraz.ac.at> <lbhlmc3iytx.fsf_-_@xihu.inria.fr>
Message-ID: <qnkhemr6zyu.fsf@arbutus.physics.mcmaster.ca>

At some point, Jan Kybic <kybic at ieee.org> wrote:

>> your experience gained before tell you that Python slows down
>> remarkably), wouldn't it be wise to implement the function in C or C++
>> from the start on?
>
> If you knew from the beginning what is to be implemented, of course.
> But in real word (especially in the research world), you do not know
> what is going to work. You try something based on your preliminary
> analysis and experience and it does not quite give you the result you
> wanted. So you keep tweaking the algorithm (and implementation) little
> by little, until you are satisfied. For this exploratory programming,
> Python is perfect. You can of course write it directly in C/C++ but if
> the edit/compile/run cycle takes 30s or even more for templated C++
> libraries, you are much less productive. Moreover, you write much more
> in C than in Python.  Once you are satisfied, you want your algorithm
> to run fast. So you rewrite everything into C/C++, having the Python
> implementation as a reference. Not difficult, just takes time.

And the Python code will likely be easier to verify. Once it's working
correctly, rewrite it in C/C++ and run tests to make sure the Python
version and the C version agree.

In general, I find if I'm calculating something which *must* work (or
else I make fool of myself :-), I'll do it twice in two different ways
and languages. It just gives you that extra level of confidence.

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke
|cookedm(at)physics(dot)mcmaster(dot)ca


From spw105 at york.ac.yk  Wed Apr 10 20:26:28 2002
From: spw105 at york.ac.yk (Ü)
Date: Thu, 11 Apr 2002 01:26:28 +0100
Subject: Beginnger Question
References: <c4a65c5f.0204090624.710228ff@posting.google.com> <57Ds8.284095$2q2.26316987@bin4.nnrp.aus1.giganews.com> <3CB3039C.3040108@greenroom.com.au> <a8v0mo$j50$1@newsg1.svr.pol.co.uk> <a914t4$g70$1@rzmux02.rz.uni-augsburg.de> <HCXs8.82135$l7.6504138@bin7.nnrp.aus1.giganews.com>
Message-ID: <a92l2a$90i$1@news6.svr.pol.co.uk>

> > >x = 2
> > >def wrong(num):
> > >    x = x+1
> > >    print x
>
> No, this should be
>
> def incx():
>   global x
>   x = x+1
>
> x = 2
> incx()
> print x
> # prints 3

Yup, I see where I went wrong now, but that *was* exactly what I was aiming
to do, at least... hehe




From alexandrecourbot at linuxgames.com  Sat Apr 13 10:06:16 2002
From: alexandrecourbot at linuxgames.com (Alexandre Courbot)
Date: 13 Apr 2002 16:06:16 +0200
Subject: Getting a Python library for cross-compiling with mingw32
Message-ID: <1018706814.23939.26.camel@progeny>

Hello everybody,

I'm currently porting my Linux app (which heavily uses the Python
library) to Windows, but I don't own Windows myself. So I'm
cross-compiling the stuff from my Linux box using mingw32.

The problem is, every Windows binary comes bundled into a .exe installer
that I can't use. I've performed some searches, but was unable to find
something suitable for my case (a simple zip or tar.gz containing the
Python archive would do it, but even that I can't find)

I then attempted to cross-compile Python itself. It looks like it's not
intented for that :) After disabling configure tests that didn't like
cross compilers, I had to twiddle even more to get a binary. Finally, I
managed to get a binary and a library, that even worked under Wine!
Unfortunately, at init time, when the os.py module is imported and
looking for the os in sys.builtin_module_names, it can't find anything.
Looks like I lack the Windows specific module in my binary. The
interpreter does work after that, but I'd like to have things done
properly.

Please note that I largely prefer having a version >= 2.2. Anyone knows
of a place which provides suitable binaries, or any Python hacker could
help me properly compiling my binary? It looks like the source
distribution wouldn't need big changes to be cross-compileable, which
would be nice. I'd be glad to help making the source distribution
cross-compileable, provided I can solve this issue and you guys think
the feature would be usefull.

Thanks for any help,
Alex.
-- 
http://www.gnurou.org





From robin at jessikat.fsnet.co.uk  Thu Apr 18 14:30:14 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 18 Apr 2002 19:30:14 +0100
Subject: pyRXP Memory Likage???
References: <c5f48de3.0204161113.14919d25@posting.google.com>
Message-ID: <Ql$1uLA2Cxv8EwhR@jessikat.demon.co.uk>

In article <c5f48de3.0204161113.14919d25 at posting.google.com>, Jose
Manuel Esteve <chema at imdea.com> writes
>I have been playing with pyRXP and is incredibly fast as compared with
>minidom, also the memory footprint is smaller for one file, however if
>you loop for parsing a lot of xml strings the memory usage gets higher
>and higher as compared to minidom
>
>import pyRXP
>xml='<start><tag1>text here</tag1></start>'
>P=pyRXP.Parser()
>for i in range(0,100000):
>    tup=P.parse(xml)
>
>will show the behaviour

OK I checked in a new version which fixes two leaks. The above seems to
work OK now, on my PC the memory fluctuates a little, but shows no
steady upward movement like before.
-- 
Robin Becker


From jcmason at student.math.uwaterloo.ca  Mon Apr 15 21:27:04 2002
From: jcmason at student.math.uwaterloo.ca (Joe Mason)
Date: 16 Apr 2002 01:27:04 GMT
Subject: Win32 Open Source or free .NET integration?
References: <RSJu8.3093$MM2.1226076@news1.news.adelphia.net>
Message-ID: <a9ful8$mra$1@watserv3.uwaterloo.ca>

In article <RSJu8.3093$MM2.1226076 at news1.news.adelphia.net>,
Michael Kelly  <mkelly2002 at adelphia.net> wrote:
>I haven't used .NET myself yet but another programmer
>using it raves about the back end integration among
>the languages.  Also I noticed ActiveState is doing
>a Visual Python that's supposed to integrate fully
>into .NET.  I know there are some open source project
>for .NET for Linux/unix but I was wondering if anyone
>has heard of a free Win32 implementation in the works?

The *nix versions (Mono's the only one I know about, but I think I recall
hearing of at least one other targetting BSD) will probably work under cygwin,
and might even be portable to Windows native.  Probably better to wait until 
they're close to being stable, or at least a fair ways along, to avoid
duplicating a lot of work.

Also, if anyone implements a clone of the .NET basic libraries, they should
be usable on the original .NET VM, no?  And the VM is probably the easiest
part due to the volume of the libraries.  So if there's another free VM
with core libraries, it shouldn't be too hard to get a completely free
version for Windows.  (Unless the libraries use a lot of native code, which
is possible.  I don't even know how/if .NET allows calls outside the VM.)

Joe 


From guido at python.org  Mon Apr  8 22:55:40 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 08 Apr 2002 22:55:40 -0400
Subject: RELEASED: Python 2.1.3
Message-ID: <200204090255.g392tex17317@pcp742651pcs.reston01.va.comcast.net>

Today, April 8th 2002, we've released Python 2.1.3, the latest version
of Python 2.1.  This is not the latest and greatest release, but is
for people who want or need to run Python 2.1.  Anthony Baxter was the
2.1.3 releasemeister, for which I thank him profusely.  Great job,
Anthony!

The new release is available here:

    http://www.python.org/2.1.3/

(Note that we no longer release files via SourceForge -- it's a bit of
a hassle, and few people seem to download their files from there.)

This release has a small number of critical bug fixes that have come 
up since 2.1.2 was released, in particular one bug that caused Zope 
to crash under certain obscure circumstances.


What's New in Python 2.1.3?
---------------------------

Core

- Fixed two bugs reported as SF #535905: under certain conditions,
  deallocating a deeply nested structure could cause a segfault in the
  garbage collector, due to interaction with the "trashcan" code;
  access to the current frame during destruction of a local variable
  could access a pointer to freed memory.

- SF #457466: popenx() argument mangling hangs python (win9x only).
  Under certain circumstances, using any of the popen calls on win9x
  would cause python to hang.

Library

- The xml.sax.expatreader.ExpatParser class will no longer create
  circular references by using itself as the locator that gets passed
  to the content handler implementation.  [SF bug #535474]

- A security hole ("double free") was found in zlib-1.1.3, a popular
  third party compression library used by some Python modules.  The
  hole was quickly plugged in zlib-1.1.4, and the Windows build of
  Python now ships with zlib-1.1.4.

Windows

- The installer now installs Start menu shortcuts under (the local
  equivalent of) "All Users" when doing an Admin install.

- If you do what it takes to compile SSL support for Windows, a bug in
  httplib.py prevented SSL from working correctly.  That bug has been
  fixed, but note that SSL support on Windows is still not enabled in
  this release.


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




From siegfried.gonzi at kfunigraz.ac.at  Tue Apr 16 05:10:01 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Tue, 16 Apr 2002 11:10:01 +0200
Subject: newbie-question: extending python with c++
References: <mailman.1018890724.31468.python-list@python.org> <3CBB1AE8.E2843017@kfunigraz.ac.at> <a9f8ru$lef$2@peabody.colorado.edu>
Message-ID: <3CBBEA69.7BA7EAAC@kfunigraz.ac.at>

Fernando P?rez wrote:

> Could you share that mini-howto here with the rest of us? I'm sure that more
> than one would find it useful.


I think it is not inappropiate to post here the instruction (thought I
got it due to private mail; but there is no private junk in it and hence
I believe the original author does not mind: author: Luigi Ballabio; I
augmented Luigi's B. stuff with the pointer stuff and some other
explanations).

The example is not really related to C++ but it demonstrates the recipe
well.

In the SWIG manual there is an example demonstration, which is as
follows:

A) The C code:

/* File : example.c */
/* A global variable */
double Foo = 3.0;

/* Compute the greatest common divisor of positive integers */
int gcd(int x, int y) {
  int g;
  g = y;
  while (x > 0) {
    g = x;
    x = y % x;
    y = g;
  }
  return g;
}

Save the above code to a file called "example.c".

B) You want to make an includable Python function of the above C code:
Construct an Python setup file:

from distutils.core import setup, Extension

setup(name         = "Example",
       description  = "Example Python extension",
       py_modules   = [],
       ext_modules  = [Extension("example",
                                
["C:/Wissenschaft/SWIG-1.3.11/ex/example_wrap.c",
				  "C:/Wissenschaft/SWIG-1.3.11/ex/example.c"],
                                 libraries = [],
                                 define_macros = [],
                                 include_dirs =
[C:/Wissenschaft/SWIG-1.3.11/ex/],
                                 library_dirs = [],
                                 extra_compile_args = [],
                                 extra_link_args = [])
                      ]
      )

This file called "setup.py" and should reside in the same folder as
example.c! Look carefully to the above scheme:

a) C:/Wissenschaft/SWIG-1.3.11/ex/ is my directory where example.c and
setup.py resides; yours can differ!

b) The stuff (right after "example") tells you where to find
"example_wrap.c" (which, see below, swig will create) and "example.c".
If lets say your example.c includes some header files, e.g.
"myMathHeader.h" then you would possibly include the C function here:
"MyMathHeader.c".

C) The SWIG specific stuff: Create first an file called "example.i",
where you specify your functions which you want to become visible (this
are then the functions which you can use in your Python environment):

/*File: example.i*/

%module example  
extern int gcd(int x, int y);
extern double Foo;


[example.i should reside in the directory where you store example.c and
setup.py]

D) Setup Microsoft Visual C++ (I am not sure how it works with other
compilers):
 
Your MVC++ installation created a file named vcvars32.bat which 
sets up the environment for using MVC++ from the command line. Run it 
before the next step, or add its contents to your autoexec.bat so 
that you can forget it ever after.

If you do not find vcvars32.bat easily: let Windows search for you and
change to that directoy (in your DOS command line window) and follow
point D), and type: vcvars32.bat


E) I have not set a gloabl variable for the SWIG and Python environment,
so I have always to change the directory in order to follow the next
steps (remember: C:/Wissenschaft/SWIG-1.311/ex/ is just my directory;
yours can differ): 

 1.1 swig -python C:/Wissenschaft/SWIG-1.3.11/ex/example.i
     [step 1.1 creates example_wrap.c] 
 1.2 I have to change to the Python directory:
 1.3 python C:/Wissenschaft/SWIG-1.3.11/ex/setup.py build
 If everything is okay you can move on:
 1.4 python C:/Wissenschaft/SWIG-1.3.11/ex/setup.py install
     [step 1.4 will put a file in the directory
/Python/Lib/site-packages

Now you can fire up your Python environment and include the functions:

from example import*
s = gcd(3,4)



Then I used the above scheme to create my  Mie calculation Python
function; I encounterd some problems:

a) I had to pass pointers to my C function, e.g.  c-function(float a,
float* p, float* d,...)

where p is an array pointer and d is just a pointer to a float; they all
become updated

b) In SWIG you realize this by including "-lpointer" :
		swig -python -lpointer example.i

This will make that you then can use some pointer-functions in your
Python environment after importing example.py.
 
c) If you want to use this c-function in Python, you would have to
create an array:

from c-function import*

p = ptrcreate("float",0,10)

creates an vector with 10 floating point (C floating points) values
initialzed to 0.

BUG: creating an arary with only 1 dimension results on my machine to
unforseeable idle crashes; I think there is maybe a SWIG bug. For
example I had to pass 4 differetn floating point pointers (float*
a,float* b,float* c,float* d). I did this as:

a = ptrcreate("float",0,1) and so on

This does not really work. I rewrote the C function and passed an array
which becomes updated with the 4 values: ptrcreate("float",0,4). I think
creating an array with dimension 1 is buggy (but you would have to play
around yourself in order to figure it out).


Your C function delivers you the pointer back and you can convert it to
a Python object (see also the SWIG manual):

def build_list(p,nitems):
	l = []
        for i in range(0,nitems):
        	l.append( ptrvalue(p,i) )
        return l

list_a = build_list(p,10)

and it is wise to dispose then the pointer in Python:  ptrfree(p)  

You can then work on with list_a.

I got the advise from the developers to use the dirty pointer stuff only
for quick hacks. Otherwise: you should invest time and try to use
typemap (see the SWIG documentation), which is much more elegant and
safe.


Thank you Mr. Gonzi for improving the world,
S. Gonzi


From phr-n2002a at nightsong.com  Thu Apr  4 22:55:00 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 04 Apr 2002 19:55:00 -0800
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <aeei8a.i11.ln@127.0.0.1> <a8j57e$omv31$1@fido.engr.sgi.com>
Message-ID: <7x1ydun6jv.fsf@ruckus.brouhaha.com>

pj at engr.sgi.com (Paul Jackson) writes:
> But why do you lament that '==' is no longer like the Kronecker
> Delta function?  That still seems like a fitting name to me.

Because the value is no longer 1 or 0.


From cliechti at gmx.net  Sun Apr  7 12:33:32 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Apr 2002 17:33:32 +0100
Subject: Can I print from Python?
References: <20020407103654.19127.00002640@mb-fm.news.cs.com>
Message-ID: <Xns91E9B2CEE1835cliechtigmxnet@62.2.16.82>

jimw1956 at cs.comnojunk (Jim Williams) wrote in
news:20020407103654.19127.00002640 at mb-fm.news.cs.com: 

> Hi all...I'm new to Python and really enjoying it so far, but I am
> somewhat perplexed that as far as I have seen it has no facilities for
> printing to a printer?  Surely this is possible. Someone please
> enlighten me...I have done a good bit of research and searching and
> can find no reference to how to do this. 

one neat solution is to use Reportlab to generate a PDF and you'll find an 
Acrobat reader on all big platforms today.

you can also look at some GUI systems. e.g. wxPython has some printing 
support etc. but when i need a printed output from a script, i generate a 
PDF.

chris

-- 
Chris <cliechti at gmx.net>



From grante at visi.com  Wed Apr  3 10:33:52 2002
From: grante at visi.com (Grant Edwards)
Date: Wed, 03 Apr 2002 15:33:52 GMT
Subject: How to check whether email /domain exists
References: <mailman.1017822968.14537.python-list@python.org>
Message-ID: <AhFq8.12288$vm6.1831818@ruti.visi.com>

In article <mailman.1017822968.14537.python-list at python.org>, A wrote:

> What is the simplest way how to check 
> whether a given email address( or domain for this email) exists?

The simplest way is using the function below:

def isEmailAddrValid(addr):
    return 1

It's small, fast, and generates no network traffic!  You don't
even have to be attached to a network to use it! And it's
almost as accurate as anything else you can do short of sending
an e-mail and requesting that the user reply.

On a more serious note, about the only thing you can really do
is to look up the MX record and see if the host it points to
has an SMTP server by connecting to port 25.

-- 
Grant Edwards                   grante             Yow!  Is this going to
                                  at               involve RAW human ecstasy?
                               visi.com            


From fperez528 at yahoo.com  Wed Apr 24 13:31:41 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 24 Apr 2002 11:31:41 -0600
Subject: python and very large data sets???
References: <ad381f5b.0204240841.52333a3e@posting.google.com>
Message-ID: <aa6q9k$sf6$1@peabody.colorado.edu>

Rad wrote:

> I am preparing myself to work on extracting data from 4 text files
> (fixed width format) which combined size is about 80GB.  Considering
> deadlines, costs, and my limited programming knowledge I thought using
> Python/Windows for the job would be the best option for me.  However,
> I am worried about the speed in which Python (me and my hardware) will
> be able to deal with these massive data sets but I am hoping that this
> is still a quicker route then learning C.
> I still haven't received above mentioned files so I can't test the
> time needed to (for example) read a 15GB "file1", filter by few
> variables, and write a resulting subset as a "sub_file1".  Things
> would afterwards get more complicated cause I will have to pullout
> ID's from "sub_file1", remove duplicate ID's create
> "no_dup_sub_file1", match those to ID's in remaining 3 main files and
> pullout data linked with those ID's.
> 
> I have a few weeks to prepare myself before data arrives and my
> question is: am I going the right way about the project, is Python
> (with humanly written code) capable of doing this kind of staff
> relatively quickly?
> 
> Any help and suggestions would be greatly appreciated.

May I suggest that you also pick up some basic (4 hrs worth) grep/awk 
knowledge? Sometimes I find that for quickly extracting a few tagged data 
fields from a large file, awk is much faster than python. I'll make the awk 
call in my python code and then keep operating for the more complex stuff in 
python.

I only use awk for _very simple_ pattern-based or column-based extractions, 
but for those tasks it's super-easy and it's quite fast (all C). Anything 
even minimally complicated I leave to the actual python code.

Just an idea. On a big problem, a single-tool mindset may be a bit 
constricting, give yourself some mental room.

good luck,

f.


From jb at cascade-sys.com  Sat Apr 20 01:58:15 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 19 Apr 2002 22:58:15 -0700
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <20020419205101.17293.2839.Mailman@mail.python.org>
Message-ID: <3CC10377.76F03AE3@cascade-sys.com>

> From: grante at visi.com (Grant Edwards)
>
> When somebody asks about a ternary operator they usually don't
> really care about a ternary operator.  What they _really_ want
> to do is to know how to solve in Python a problem that was
> sovled in C by using a ternary operator.

I couldn't disagree more.  Worse -- when you suggest rather "there's another way to do it," I'm reminded of Perl and have to stifle a gag reflex.

In decades of Software Engineering I've found inline if/then/else to be every bit as useful for expression evaluation as the if/then/else statements are for control flow.  If the capability is available in the language, it's often the clearest and most efficient way of expressing the desired behavior.  Think: binary choice.  What could
be more natural?

Dictionaries, arrays and case/switch statements are great for N-way decision making but to use them for binary choice seems a perversion.  Two-way expression branching actually is as or much more common than N-way and arguably deserves special recognition.

Python claims a Lisp heritage, which I imagine is why and/or work the way they do.   But for decision making, original Lisp most often used COND, which was an N-Way evaluation (and the only conditional construct in the language):

    (COND (( condition1 value1 ) ( condition2 value2 )...( true valueN )))

The condition of each successive sub list would be evaluated until one was found to be true then the corresponding value would be returned for the entire expression.  If no clause was true the system halted with a fatal runtime error.  Like most fully general things, COND actually was somewhat cumbersome to use.  Also, in original Lisp
AND and OR fully evaluated all arguments, so COND was the only way to express any kind of decision making.

The two-way choice was so common and vitally useful that Scheme (a successor to Lisp) adopted it as a standard form:

    (IF condition then_expr else_expr)

with the obvious meaning.  It's probably the most frequently used single construct in that language.  So from a linguistic heritage standpoint, it's actually rather odd that Python lacks some equivalent of Scheme's IF.

I am probably in the minority here in that I have little patience for people who are "confused" by or have difficulty with advanced language features.  If you purport to be a professional programmer then you should know the base language cold.  So I am not swayed by arguments that :? confuses people.  But I think part of the problem
with inline choice is that people who use it do not properly parenthesize and indent it to properly highlight what's going on.  I always split the test and the two alternatives on separate lines, with each of the latter two starting with the special operators.

    def abs( n ):
        return (( n > 0 )
                      ? ( n  )
                      : ( -n ))

You'd never write if/then/else without proper indentation (not even possible in Python ;o).  Same rules should apply to complex expressions.

Python's and/or semantics, although they may be employed to provide similar functionality, are more general, unrelated and in most cases an inappropriate substitute.

Thus although you can use and/or as a partial substitute, there are problems with the hack:

Most importantly (a) the and/or idiom strays from the desired abstraction and thus obscures the intent of the developer.

(b) Since and/or lacks clarity it arguably is more likely to people than inline choice.  And whereas :? confuses beginners, and/or (since it's an idiom rather than a language construct) is more likely to confuse more advanced users.

(c) Worst of all, the idiom doesn't even work in all cases, thus admitting the possibility of subtle, unintentional, seemingly intermittent errors, e.g.:

     x = ( len( s ) > 0
                     and s[0]
                     or -1 )     # FAILS!!  when s[0] == 0

To me clarity of expression is the most important quality of a well written program.  Binary conditional evaluation ala ?: is indispensable.  It's eminently reasonable for programmers to feel a "need" for such an addition to Python.

Biggest issue the syntax.  With ":" used the way it is in Python, C's :? syntax probably is not an easy fit.  One possibility might be to reverse the two symbols:

    x = ( len( s ) > 0
            : s[ 0 ]
            ? -1 )

Having ":" separate the test expression from the two alternatives seems sorta natural.

Another alternative would be to borrow from Algol68 and (IIRC) Algol60. They allowed if/then/else to be embedded within an expression.  Generalizing if:/else: to return a value also seems somewhat natural, so perhaps that could work.

    x = ( if len(s) > 0:   s[ 0 ]
                       else:    -1 )

Since it's all within an expression, indentation doesn't matter.

Just a thought....

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From brueckd at tbye.com  Wed Apr 10 13:35:38 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 10 Apr 2002 10:35:38 -0700 (PDT)
Subject: Transferring files using sockets
In-Reply-To: <15540.27738.975194.101310@beluga.mojam.com>
Message-ID: <Pine.LNX.4.33.0204101034250.24093-100000@ns0.tbye.com>

On Wed, 10 Apr 2002, Skip Montanaro wrote:

>
>     IC> I am trying to transfer a file using the socket module, but I'm not
>     IC> sure how to do this. Should I split the file into little(1KB)
>     IC> packages and then transfer them, or is there a possibility to send
>     IC> the whole file?
>
> It should be possible to just execute
>
>     sock.send(open("somefile").read())
>
> if sock is a connected stream socket.  You might want to break it down in
> chunks though, either to conserve local memory or provide transit feedback:
>
>     f = open("somefile")
>     while 1:
>         data = f.read(8192)
>         if not data:
>             break
>         sock.send(data)
>         sys.stderr.write(".")
>         sys.stderr.flush()
>     sys.stderr.write("\n")

Also, 'send' doesn't necessarily send the whole chunk of data; as a return
value it gives you the number of bytes actually sent. Python 2.2 and later
sockets have a sendall method you can use instead.

-Dave





From jeremy at alum.mit.edu  Tue Apr 23 18:22:39 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 23 Apr 2002 15:22:39 -0700
Subject: Fundamental problem with urllib...
References: <mailman.1019290298.32137.python-list@python.org> <yNUw8.74422$T%5.18813@atlpnn01.usenetserver.com> <slrnacb5ko.4i6.akuchlin@ute.mems-exchange.org>
Message-ID: <b0f083db.0204231422.785cb57c@posting.google.com>

"A.M. Kuchling" <akuchlin at ute.mems-exchange.org> wrote in message news:<slrnacb5ko.4i6.akuchlin at ute.mems-exchange.org>...
> In article <yNUw8.74422$T%5.18813 at atlpnn01.usenetserver.com>, 
> 	Steve Holden wrote:
> > Since urllib knows nothing of cookies, you will need to integrate some sort
> > of a cookie jar into the library, with a new API  for the clients to
> > retrieve and store the cookies.
> 
> This is worthwhile, but I don't think it belongs in urllib.  It
> belongs in a module or package of its own that provides general
> Web-browser features such as cookies, remembering authentication
> usernames and passwords, and a cache.  This package could then be used
> for implementing HTML-scraping scripts, spiders, or a Web browser.

urllib2 provides a more flexible framework for implementing
URL-loading programs, like a spider.  I think it would be helpful to
have the features you mention integrated into urllib2.

I'm not sure what the difference between an HTTP client, like urllib
or urllib2, and a Web-browser is.  Other than urllib's monolithic
design, why wouldn't you want these sorts of features in the module?

Jeremy


From marklists at mceahern.com  Mon Apr 15 19:36:59 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 15 Apr 2002 18:36:59 -0500
Subject: Going through folders...
In-Reply-To: <scbb7efb.022@Riksvasa.syd.kth.se>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEDNKLAA.marklists@mceahern.com>

[FREDRIK HULDTGREN]
> I want to go through a catalogue structure and make a textfile 
> that lists the files in each catalogue/undercatalogue. I know 
> that there is a function for doing this somewhere, I just can't 
> remember what it is right now...
> all help is appreciated...

import os
os.listdir(r"c:\temp")

Cheers,

// mark




From kseehof at neuralintegrator.com  Tue Apr  9 03:19:49 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 09 Apr 2002 00:19:49 -0700
Subject: Embedding Python with C++
In-Reply-To: <GYts8.26982$2J2.6830478@typhoon.southeast.rr.com>
Message-ID: <EFEFIHAINIINNELLHFAIOEEDCAAA.kseehof@neuralintegrator.com>

Chris Stoy wrote:
> Hey all,
>
> I'm looking to embed Python as a scripting language in a game I am working
> on.  I want to be able to call C++ methods from Python, and call
> Python from
> C++ (for the usual code quick in Python, convert to C++ for speed later.)
>
> I've looked into using the Boost.Python code and I'm fine with that.
>
> My problem is, I have no clue how to compile Python into a static linked
> library for Windows.  Actually, I need it to be cross-platform, which is
> another issue.
>
> I've never embedded Python before, and I'm fairly new to Python, so can
> someone please give me a pointer to how to do this?  What are the files I
> need to compile to build the static lib (using MS Visual Studio
> 6.0?) Also,
> I don't need all of the standard Python components (like networking.)  Can
> those be removed from the lib?
>
> Thanks.
>
> Chris.

Actually you don't need to turn your python code into a library
to access from C++.  That's not the best approach.  The entry
points into python are provided by python*.dll (in windows).

Instead, use PyObject_CallObject and associated supporting functions
to invoke callable objects (python functions and methods).  Read the
"Extending and Embedding" section in the main python documentation.
http://www.python.org/doc/current/ext/ext.html

I recommend making the main application python, and implement low
level functionality as extensions.  In my experience this works much
better than writing that main application in C++.

Implement your low-level C++ funcitonality as a set of extension
types that your main python application calls.

If your development enviroment is Windows, check out my extention
wizard.  It's a Visual C++ wizard for creating python extensions.
I use it in situations where SWIG is not useful (i.e when I am
writing new code and not just wrapping an existing library).

You can find PyExtWizard here:
http://www.neuralintegrator.com/downloads

- Ken Seehof





From guido at python.org  Thu Apr  4 09:27:51 2002
From: guido at python.org (Guido van Rossum)
Date: Thu, 04 Apr 2002 14:27:51 GMT
Subject: ACCEPTED: PEP 285
References: <mailman.1017873312.18135.python-list@python.org> <_vUq8.552234$pN4.39754796@bin8.nnrp.aus1.giganews.com>
Message-ID: <3CAC63C8.6B3A4D0B@python.org>

> "Guido van Rossum" <guido at python.org> wrote in message
> >     This PEP does *not* change the fact that almost all object types
> >     can be used as truth values.  For example, when used in an if
> >     statement, an empty list is false and a non-empty one is true;
> >     this does not change and there is no plan to ever change this.
> 
Terry Reedy wrote:
> almost? all?  I am under the impression that 'if x:' is universally
> valid, and consider that one of the good features of Python's prime
> dichotomy system.

It's possible for a class or type to explicitly resist being used
in a Boolean context:

    class C:
        def __nonzero__(self):
            raise TypeError, "I am *not* Boolean!"

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


From sholden at holdenweb.com  Wed Apr 10 11:18:07 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 10 Apr 2002 11:18:07 -0400
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com> <3CB2FE25.86A37EA1@imk.fraunhofer.de> <ub6dji5plutpc5@corp.supernews.com>
Message-ID: <SKYs8.19881$GM.6702@atlpnn01.usenetserver.com>

"DeepBleu" <DeepBleu at DeepBleu.org> wrote in message
news:ub6dji5plutpc5 at corp.supernews.com...
>
> "Joachim Kaeber" <joachim.kaeber at imk.fraunhofer.de> wrote in message
> news:3CB2FE25.86A37EA1 at imk.fraunhofer.de...
> > DeepBleu wrote:
> >
> > > Yes, I am using zipfile (from zipfile import ZipFile).  Sorry I stated
> zlib.
> > > Still with one text file in particular, WinZip reports "missing 2
bytes"
> > > DeepBleu
> >
> > is the script run on Windows? If yes, is the zipfile opened
> > in binary mode?
> >
> The script is running on Sun Solaris with Python 2.1
>
> >>>from zipfile import ZipFile
> >>>zip_archive = ZipFile("text1.zip", "w", 8)
> >>>zip_archive.write(PathToTextFile, os.path.split(PathToTextFile)[1])
> >>>zip_archive.close()
>
> Before starting this process, I am running Unix2DOS on the text file:
> Unix2Dos TextFile TextFile
>
> Reason is that the zip file will be opened later in windows (using WinZip
> for example).
> Hope this answers your question,

Do you close the zipfile after you've finished writing it?

regards
 Steve






From gmcm at hypernet.com  Fri Apr 12 08:49:10 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 12 Apr 2002 12:49:10 GMT
Subject: Include zlib.dll in McMillan Installer
References: <a93ld3$jb4$03$1@news.t-online.com> <Xns91ED608AADBDEgmcmhypernetcom@199.171.54.213> <a95vrj$4eh$02$1@news.t-online.com>
Message-ID: <Xns91EE59D52F1B2gmcmhypernetcom@199.171.54.214>

[posted and mailed]

Harald Schneider wrote:

> at first my compliments for your Installer :)

Thank you.
 
[snip]

> I use the Activestate distribution of Python 2.1 It seems, that
> here zlib.dll is still needed. It's located in \python\DLLs\zlib.dll

I have ActiveState Python2.1, build 210 (among others). No zlib.dll;
zlib.pyd depends only on python21.dll, MSVCRT.dll and KERNEL32.dll.
 
Still puzzled.

-- Gordon
http://www.mcmillan-inc.com/





From fredp at mygale.org.nospam  Wed Apr 10 11:37:38 2002
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 10 Apr 2002 15:37:38 GMT
Subject: Baffled by CGI cross-browser issues... help sought :)
Message-ID: <Xns91ECB3C2132FBPaCmAnRDLM@161.48.128.20>

Hi everyone,

The recent thread "Standalone CGI Server - Question for Group" taught me 
many things, and got me started on my first mini-project (alas :).

As a server I use Sandy Nortons's cute CGIHTTPServer trick, combined with 
Andrew Kuchling's mini-CGI-framework from Webtechniques 2/1998 
(http://www.newarchitectmag.com/archives/1998/02/kuchling/). Both make it 
easy to get the basics running on a local (Windows98SE + py 2.1) machine 
and experiment.

All went well at first (with the usual beginers' gotchas :), and I was 
quickly up to a half-dozen scripts working together. But now I've hit a 
wall I didn't think I'd find there : browser compatibility.

During development all testing was done with K-meleon (a lightweight 
Windows browser based on the Mozilla Gecko engine) and noticed no problems. 
Everything also works fine with Mozilla itself. But when I finally thought 
of testing my pages/scripts on more mainstream browsers there were nasty 
surprises :

* on Netscape 4.x things sort of work, except that each time I submit a 
form I get an error message like "A network error occurred while receiving 
data (connection reset by peer)". After dismissing the dialog the resulting 
page usually displays correctly (although sometimes truncated).

* under MSIE 5.5 it's worse, forms will display but submitting them never 
works, with a bewildering variety of failure modes : custom IE error pages 
(server not found or DNS error), blank pages, or more rarely terse messages 
from the http server (not executable, forbidden, etc.)

At this point I've spent more time tinkering and trying out wild guesses 
than it took me to actually write the initial scripts, and still no joy. It 
is dismaying to see something that simple and small break down that way...

If anyone has an idea of what and where to investigate I'd be grateful for 
that, as I'm nonplussed. The code is tiny and non-secret (although probably 
ugly as hell) so if anyone wants to take a look, that's easy too.

TIA,
fp

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


From jb at cascade-sys.com  Fri Apr 26 03:26:03 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 26 Apr 2002 00:26:03 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <3CC84749.1AA2A087@cascade-sys.com> <3CC8FD85.D28F6D4B@gras.de>
Message-ID: <3CC9010B.D901F927@cascade-sys.com>

Christian Tismer wrote:

> Sorry about that. I knew it of course, but the
> association with the well-known "Kernighan/Ritchie"
> Book made me write this mess.

No biggie.  Just thought you'd wanna know.

> BTW, AWK is more related to Perl than to Python.

I think they're all three pretty different and unrelated.

Awk and Perl do both use {} for grouping and share a common syntax for IF, FOR and
WHILE statements, but that's about it.  Python with it's punction on every symbol
and the lack of functions with named parameters is pretty out there on its own.

In any case, my only point was  that AWK preceeded Python in time, not that there
was any relationship between the two.

> I really loved that little language.

Yeah, me too, though nowdays I almost always reach for Python.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From mejiaspolonio at teleline.es  Fri Apr 19 14:22:07 2002
From: mejiaspolonio at teleline.es (CERAMICA MEJIAS POLONIO C.B.       (SPAIN))
Date: Fri, 19 Apr 2002 20:22:07 +0200
Subject: POTTERY
Message-ID: <006201c1e7cf$1de3cf00$15ef63d5@mio>



Dear Sirs

Our company?s name is "Cer?mica Mej?as Polonio". We manufacture hand-made pottery for decoration.


We would like to offer you our articles which you can see on our web-site www.myempresa.com/mejiaspolonio 

We are already exporting to some European and Asian countries where our articles are very successful due to their good finishing and difference in style.

We hope you like our articles. If you want to contact us you can do it at:

"Cer?mica Mej?as Polonio C.B."

C/Carrera Baja,22

14540 La Rambla (Cordoba)

Spain.

Phone/fax: +34 957 682 280

Translator?s contact: +34 650 06 02 79

e-mail: mejiaspolonio at teleline.es

web: www.myempresa.com/mejiaspolonio 

Waiting for your news. 

Yours Faithfully.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020419/0db9a365/attachment.html>

From perry at subdimension.com  Fri Apr 26 07:26:13 2002
From: perry at subdimension.com (Perry)
Date: 26 Apr 2002 04:26:13 -0700
Subject: ActiveX events in wxPython not working
References: <6ddcff55.0204250625.1992ce1b@posting.google.com> <mailman.1019763678.16478.python-list@python.org>
Message-ID: <6ddcff55.0204260326.46d2e567@posting.google.com>

Cliff Wells <logiplexsoftware at earthlink.net> wrote in message news:<mailman.1019763678.16478.python-list at python.org>...
> On 25 Apr 2002 07:25:36 -0700
> Perry wrote:
> 
> > Hi,
> > 
> > I'm using wxPython as an ActiveX (GUI) container. The OCX draws a
> > gantt
> > chart from supplied data. Everything works fine, appart events.
> > 
> > Here is the code:
> 
> [snip]
> 
> > class MyGanttPanel(wxPanel):
> >   def __init__(self, parent):
> >      wxPanel.__init__(self, parent, -1)
> > 
> >      sizer = wxBoxSizer(wxVERTICAL)
> >         
> >      GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
> >      gantt.VcGantt = GanttClass(self, -1, wxPoint(0,0), style=wxMAXIMIZE)
>        ^^^^^ Where is this instantiated?
> 
> [snip]
> 
> I don't see where the object "gantt" is being created.  You can't add an
> attribute to a non-existent object:
> 
> Python 2.2 (#1, Mar 15 2002, 11:52:42) 
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-81)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> f.a = 1 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> NameError: name 'f' is not defined
> >>> class foo: pass
>  ... 
> >>> f = foo()
> >>> f.a = 1
> >>> 
> 
> Are you running the program from a shell or by double-clicking its icon?  It
> may be that there are errors you aren't seeing if you aren't using a shell. 
> Other than that, I'm afraid my ActiveX knowledge is (happily) non-existent, so
> if this isn't your problem then I'm afraid I can't help much.

Solved !!!

Apparently a week ago when I read the doc string in the file
activexwrapper.py i was not paying enough attention. Another trip to this
file solved the problem.

GanttClass = MakeActiveXClass(gchart.VcGantt, eventObj=MyEvents)
                                              ^^^^^^^^
should be:

GanttClass = MakeActiveXClass(gchart.VcGantt, eventClass=MyEvents)


Now everything is working fine. :-)


From news at tylerdave.com  Thu Apr  4 16:48:32 2002
From: news at tylerdave.com (Dave Forgac)
Date: Thu, 4 Apr 2002 16:48:32 -0500
Subject: calldll
Message-ID: <a8ihlb$64l$1@nsnmpen2-gest.nuria.telefonica-data.net>

I am trying to make a python module similar to the perl module
Win32::SystemInfo.  The perl module uses a dll called cpuspd.dll.  I am
trying to use this same dll within python.  The function in the dll is
called 'GetCpuSpeed'.  I have not been able to find much documentation on
the subject but from what i can see, calldll seems to be my best bet.  Can
anyone point me in the right direction?

-Dave Forgac




From elimli at yahoo.com  Sun Apr 21 13:10:58 2002
From: elimli at yahoo.com (eli li)
Date: Sun, 21 Apr 2002 19:10:58 +0200
Subject: constructors in python
Message-ID: <a9uoa0$t0r$1@news.iucc.ac.il>

Consider, we have two classes: one derived from other. In derived class c-or
how is it possible to call the base c-or? Simple call base.__init__ (...)?




From sean.bowman at acm.org  Tue Apr 16 16:53:42 2002
From: sean.bowman at acm.org (sean.bowman at acm.org)
Date: Tue, 16 Apr 2002 15:53:42 -0500 (CDT)
Subject: optimization with the "compiler" module
Message-ID: <Pine.BSF.4.21.0204161551470.29317-100000@exigence.opensoftwareservices.com>

hello,

I'm wondering if there's a way to optimize code generated using the
compiler module.  for example, if I say::

 import dis
 from parser import compileast, expr
 
 code = compileast(expr('x+3+7.'))
 print dis.dis(code)

I notice that the constants aren't folded.  is there a way to do this, or
do I have to fiddle with the ast myself?

many thanks,
Sean





From eyal at hyperroll.com  Sat Apr  6 10:08:04 2002
From: eyal at hyperroll.com (Eyal Lotem)
Date: Sat, 06 Apr 2002 18:08:04 +0300
Subject: file rewind?
References: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>
Message-ID: <3caf0ef8@news.bezeqint.net>

joel8bit wrote:

> I have this code:
> 
> def countLines(file):
>     numLines = 0
>     for line in file.readlines():
>         print line
>         numLines = numLines + 1
> 
>     print numLines
> 
> def countChars(file):
>     numChars = 0
>     for line in file.readlines():
>         print line
>         for char in line:
>             print char
>             numChars += 1
>     print numChars
> 
> def test(name):
>     file = open(name, "r")
>     countLines(file)
>     countChars(file)
> 
> if __name__ == "__main__":
>     test("C:\\test.txt")
> 
> when I run the module from the comman line by itself, it counts the lines
> alright, but does not count the characters.  do I need to "rewind" the
> file reference after the first function call to countLines()?

Yes, you can do this with:
file.seek(0)

Also, I would suggest the alternate code:

        def countLines(file):
                return len(file.readlines())

        def countChars(file):
                return len(file.read())

Or if you want to be more efficient:

        def countChars(file):
                cur_offset = file.tell()        # remmember current position
                file.seek(0, 2)                 # seek to end of file
                file_len = file.tell()          # the end-position (number of bytes)
                file.seek(cur_offset)           # seek back to where we were
                return file_len



From martin at v.loewis.de  Thu Apr 11 12:37:56 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 11 Apr 2002 18:37:56 +0200
Subject: distutils
References: <GTet8.4573$Dw4.8174@news1.bredband.com> <j4sn62xx1y.fsf@informatik.hu-berlin.de> <Jift8.8$WJ4.85@news1.bredband.com>
Message-ID: <m38z7u9onv.fsf@mira.informatik.hu-berlin.de>

Magnus Heino <magnus.heino at pleon.sigma.se> writes:

> Well, thats valid for 
> 
> python2.2 setup.py install
> 
> while im doing 
> 
> python2.2 setup.py bdist_rpm
> 
> There isnt anything you can do in setup.py?

Probably, but I don't know what it is. You'll have to study bdist_rpm.py.

Regards,
Martin


From peter at engcorp.com  Wed Apr 10 21:57:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 10 Apr 2002 21:57:34 -0400
Subject: looping throuhg a list 2 items at a time
References: <a92f5j$250g@r02n01.cac.psu.edu> <mailman.1018480111.18619.python-list@python.org> <a92hl1$1v4c@r02n01.cac.psu.edu> <eppstein-40A0D7.17104110042002@news.service.uci.edu> <3CB4E282.AAD468E4@engcorp.com> <eppstein-5BEE17.18114410042002@news.service.uci.edu>
Message-ID: <3CB4ED8E.70804A2A@engcorp.com>

David Eppstein wrote:
> 
> In article <3CB4E282.AAD468E4 at engcorp.com>,
>  Peter Hansen <peter at engcorp.com> wrote:
> 
> > If the customer wanted to handle lists with odd numbers of items
> > he would have specified the desired behaviour.  Or he will now. :-)
> 
> I don't know, silently dropping the last item seems like it could be
> nonobvious enough to cause trouble later.  You don't think it should at
> least have a sanity check, causing more obvious breakage if things
> aren't as expected?  Something like:
> 
>     i = 0
>     while i < len(L):
>         print L[i],L[i+1]
>         i += 2
> 
> which will throw an exception rather than just ignoring the last item...

True, errors should never pass silently.  I was getting carried
away by the thought you were suggesting some kind of careful
test to handle odd-length lists in a reasonable way, rather
than just suggesting a noisy kind of failure would be best.
My bad.  Sorry.

-Peter


From Scott.Daniels at Acm.Org  Thu Apr 11 02:28:31 2002
From: Scott.Daniels at Acm.Org (Scott David Daniels)
Date: Thu, 11 Apr 2002 06:28:31 GMT
Subject: Function question
References: <zl9s8.37442$tg4.444316@vixen.cso.uiuc.edu> <slrnab68r0.dvl.huaiyu@gauss.almadan.ibm.com>
Message-ID: <1103_1018506511@news.dsl-only.net>

On Tue, 9 Apr 2002 17:22:41 +0000 (UTC), huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote:
> On Sun, 7 Apr 2002 23:55:47 -0500, John <nobody at nobody.com> wrote:
> >...What I want to do is:
> >I want to get 'a' dynamically..
> >But once a value is decided, I want a separate function (say, new_add) that
> >does
> >new_add(b) (equals to) add(a,b)
> >
> >That is, I want the function definition to be dynamic.
> >
> >Would that be possible? Any comment would be greatly appreciated.
> 
> The easiest way With Python 2.2:
> 
> >>> def f1(a):
> ....    def f2(b):
> ....       return a+b
> ....    return f2
> .... 
or check 
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549

for "curry ".

if you have a function qqq(a, b, ...), 
nf = curry(qqq,v) 
makes nf a function like qqq(b, ...).
The trickiness in curry is coping with some args without knowing
how many args the function is supposed to take.

-Scott





From mlh at vier.idi.ntnu.no  Fri Apr  5 21:36:36 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 6 Apr 2002 02:36:36 +0000 (UTC)
Subject: Argument against iterable ints...
Message-ID: <slrnaasnpk.178.mlh@vier.idi.ntnu.no>

Just a situation that cropped up recently: While writing a little
generator for flattening nested iterables, I used

  try: iter(foo)
  except: return foo
  else:
      ....

(of course I could have used a for loop directly in the try clause --
that doesn't change anything.)

If integers became iterable, this sort of code would suddenly become
infinitely recursive. Given the propensity for using exceptions in
this manner rather than type checking in Python, this seems like a
weighty argument to me...

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From eric at enthought.com  Fri Apr  5 22:43:26 2002
From: eric at enthought.com (eric)
Date: Fri, 05 Apr 2002 22:43:26 -0500
Subject: (no subject)
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1
Message-ID: <00e101c1dd14$5620a930$6b01a8c0@ericlaptop>

017935351.5088.python-list at python.org> <slrnaasqc7.178.mlh at vier.idi.ntnu.no>
Subject: Re: A faster Python?, Python compiler, Dylan,...
Date: Fri, 5 Apr 2002 21:39:55 -0500
Organization: enthought
MIME-Version: 1.0
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.50.4522.1200
X-Mimeole: Produced By Microsoft MimeOLE V5.50.4522.1200

Hey Magnus,

That is exactly what Pat Miller of LLNL is working on.  The function
weave.accelerate() (in CVS and quite limited right now) is aimed at this
purpose.  It is a little different than Python2C to satisfy some needs in the
scientific community.  Also, Greg Ewing's new Pyrex project is a good candidate
for this sort of thing -- hopefully we can learn from each other because Pyrex
looks very promising.

So, the short answer is "good idea -- we're working on it." :-)

eric


----- Original Message -----
From: "Magnus Lie Hetland" <mlh at vier.idi.ntnu.no>
Newsgroups: comp.lang.python
Sent: Friday, April 05, 2002 10:20 PM
Subject: Re: A faster Python?, Python compiler, Dylan,...


> In article <mailman.1017935351.5088.python-list at python.org>, eric
> wrote:
> [snip]
> >If this is the only criticism, than we're doing quite well. :-)
>
> How about including some code from Python2C (or similar projects)...
> Then you could write the code in the inline string with (possibly
> restricted) Python syntax, have it translated to C and then inlined...
> Then one *wouldn't* have to write things twice, and the criticism
> disappears ;)
>
> (Actually, I think it would be quite cool...)
>
> --
> Magnus Lie Hetland                                  The Anygui Project
> http://hetland.org                                  http://anygui.org






From huaiyu at gauss.almadan.ibm.com  Fri Apr 26 17:06:11 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 26 Apr 2002 21:06:11 +0000 (UTC)
Subject: Beyond __str__ and __repr__?  (Was Re: Python doesn't know how to do math?)
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com>
Message-ID: <slrnacjga3.2uv.huaiyu@gauss.almadan.ibm.com>

David Bolen <db3l at fitlinxx.com> wrote:
>instead.  The default output at the interpreter is to use repr() to
>print an object, which shows the highest precision possible to best
>assure that its output could be used to re-create the object.  The
>print statement will use str() which is designed for "prettier" output
>and will do some rounding.

The current arrangement between repr and str has some shortcomings:

- nested structures use repr for elements
- repr is constrained by the desire to be inverse of eval for simple types
- default repr for instances are optimized for object id, not content
- str cannot differentiate between 2 and "2"

It appears that another display level is needed, somewhat between repr and
str.  Let's call it 'disp' here.  It is intended to be used for display in a
controlled way, esp for nested structures.

Here's what I think it should do:

- disp(0.05)    -> "0.05", just like str(0.05).
- disp("abc")   -> "'abc'", just like repr('abc').
- disp([a,b,c]) -> '['+', '.join(map(disp, [a,b,c]))+']'.
- disp(obj)     ->  obj.__disp__().

Having and additional display method also reduces the need to override
__repr__, which is the source of very confusing diagnostics when not done
right (ie, when diagnostics are needed).

Huaiyu


From juenglin at informatik.uni-freiburg.de  Wed Apr  3 04:49:28 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 03 Apr 2002 11:49:28 +0200
Subject: PyInt_AsLong()
Message-ID: <xdu7knpaz87.fsf@leto.informatik.uni-freiburg.de>

Hi,

PyInt_AsLong() returns '-1? to indicate an error, this meets the convetion 
for PyXXX()-functions which return integers. But how do you distinguish an
'-1' indicating an error from a the legal long-value '-1'?

Cheers,
Ralf



From garrett at bgb-consulting.com  Tue Apr  9 13:00:23 2002
From: garrett at bgb-consulting.com (Don Garrett)
Date: Tue, 09 Apr 2002 17:00:23 GMT
Subject: pointer
References: <a8uukg$fuu$04$1@news.t-online.com> <mailman.1018364239.31845.python-list@python.org> <a8v36u$j3p$1@news.nsk.su>
Message-ID: <3CB31E5C.EE9E9BDA@bgb-consulting.com>


Andrey Panchenko wrote:
>> a = 1
> b = a
> a = 3
> 
> b must hold 3. But it's not true. IMHO it's better to say that the python
> names are like pointers only for mutable objects.
> (Caution - newbie opinion)
> 
> Andrey mailto:andrey at app.omsk.su

  The assignment a = 3 assigned a new reference, not a new value to an
existing integer.

  a = [1, 2]
  b = a
  a = [2, 3]

  Will leave b with the value [1, 2]. b referenced the same value as a, but
then a new reference was assigned to a, leaving b with the old value.

  BUT

  a = [1, 2]
  b = a
  a = a.append(3)

  Will leave b with the value [1, 2, 3], because b points to the same list as
a.

  All variables reference values, which is not quite the same thing as a
pointer. At least not exactly.

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc


From gimbo at ftech.net  Thu Apr  4 14:50:52 2002
From: gimbo at ftech.net (Andy Gimblett)
Date: Thu, 4 Apr 2002 20:50:52 +0100
Subject: how to detect if a file is complete
In-Reply-To: <20020404194611.GA11526@andy.tynant.ftech.net>
References: <32f17666.0204040701.64dd0bfe@posting.google.com> <20020404194611.GA11526@andy.tynant.ftech.net>
Message-ID: <20020404195052.GB11526@andy.tynant.ftech.net>

On Thu, Apr 04, 2002 at 08:46:12PM +0100, Andy Gimblett wrote:

> Alternative approach: use fuser to see what processes have the file

Sorry, just realised that if you hadn't come across fuser before, and
in the context of the original question, you might think I meant fuser
was something pythonic and go looking in the docs.  :-)  It's not,
it's a unix utility.

man fuser

:-)

-Andy

-- 
Andy Gimblett - Programmer - Frontier Internet Services Limited
Tel: 029 20 820 044 Fax: 029 20 820 035 http://www.frontier.net.uk/
Statements made are at all times subject to Frontier's Terms and
Conditions of Business, which are available upon request.




From peter at engcorp.com  Wed Apr  3 22:37:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 03 Apr 2002 22:37:02 -0500
Subject: OK
References: <e8d5abee.0204031816.2bfec035@posting.google.com>
Message-ID: <3CABCA5E.760E468A@engcorp.com>

huangcg wrote:
> 
> Is python out of fashion?

No, it's out of the past.  

(My other response would have been along the lines of "Huh?")


From fredrik at pythonware.com  Tue Apr  9 03:27:51 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 09 Apr 2002 07:27:51 GMT
Subject: html parsing
References: <3CB1E20A.B3300AA4@ncsa.uiuc.edu>
Message-ID: <XJws8.38226$l93.7594497@newsb.telia.net>

Randy Heiland wrote:
> I'm a newbie at parsing html in Python... could someone offer a snippet
> of code to:
> 1) specify a URL
> 2) parse the html and return the hyperlink associated with a unique
> anchor string

the htmllib-example-1 script almost does what you want.

    http://www.oreilly.com/catalog/pythonsl/chapter/ch05.html#t5

to read from an url, replace the "open" call with a
call to "urllib.urlopen".

to get the link for a given text, just look it up in the
anchors dictionary.

hope this helps!

</F>

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




From erno-news at erno.iki.fi  Wed Apr  3 11:38:53 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 03 Apr 2002 19:38:53 +0300
Subject: Threading problem
References: <a8f8f6$vdu$1@news.rchland.ibm.com>
Message-ID: <kuelhwpwiq.fsf@lasipalatsi.fi>

In article <a8f8f6$vdu$1 at news.rchland.ibm.com>, "Larry Whitley"
<ldw at us.ibm.com> writes:

|         threading.Thread().__init__(self, name=thName)

        threading.Thread.__init__(self, name=thName)

  -- erno



From ajs at ix.netcom.com  Sun Apr  7 22:21:06 2002
From: ajs at ix.netcom.com (Arthur Siegel)
Date: Sun, 7 Apr 2002 22:21:06 -0400
Subject: ACCEPTED: PEP 285
Message-ID: <000001c1dea4$5c9cfe80$aa44f6d1@ArtsPortable>

Well just to end this, with a few cut and pastes
that at least let *me* believe I'm not off the
wall.

We go from from Guido's conclusion that he saw no
"killer" reason not to go ahead with the PEP,
to Aahz remark:

"How could Guido have known there were no good arguments against, save by
proposing the PEP?"

Guido's wording I think actually being sensible,
and Aahz's obviously insulting to the many poeple
who voiced good reasonable objections that I don't
think *they* themselves would necessarily classify as "killer".

To the out of nowhere remark from Aahz:

"I think it might be possible to make changes in the way Guido-introduced
PEPs get presented to reduce the amount of shouting from people like Laura.
I hold no hopes that anything would get better results from
people like Arthur and Paul Rubin."

I've worked hard at being out of good graces with
folk like Aahz, who work too hard at their insider status.

Simply am confused how Paul made the grade.

Paul doesn't seem to care much, one way or other -
which makes him eminently more sensible than I.

Art










From brueckd at tbye.com  Tue Apr 23 10:12:38 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 23 Apr 2002 07:12:38 -0700 (PDT)
Subject: Why won't this POST
In-Reply-To: <8f8ffe67.0204222047.47d661ec@posting.google.com>
Message-ID: <Pine.LNX.4.33.0204230707370.26902-100000@ns0.tbye.com>

On 22 Apr 2002, Sunit Joshi wrote:

> Hello
> I'm trying to post data to IIS5.0 but keep getting error message about
> "host not found". The python script is in /cgi-bin/ProcessInput.py. I

> socket.error: host not found
[snip]
> req = HTTP('localhost/eubix')
> req.putrequest('POST' ,'/cgi-bin/ProcessInput.py') #Here's where it
> stalls

'localhost/eubix' isn't a valid hostname, hence the error. When you
construct the HTTP object you must pass in just the host name or IP
address (and port if you need to override the default):

req = HTTP('localhost')

Then in putrequest you need to specify the URL to get to the object on the
server that you're interested in. I'm not sure how you have things set up,
but I'm guessing it'll be either

req.putrequest('POST', '/cgi-bin/ProcessInput.py') # Like you have it

or

req.putrequest('POST', '/eubix/cgi-bin/ProcessInput.py')

Also note that this whole thing might be easier if you use urllib instead.

-Dave





From jb at cascade-sys.com  Tue Apr 30 00:45:46 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Mon, 29 Apr 2002 21:45:46 -0700
Subject: Remove masked elements from an array
References: <72F73B808A78D511A73600034771D9FF717CD8@dc_exchange2.howrey.com>
		<20020429174425.GN16217@devel.trillke>
		<mailman.1020115270.9666.python-list@python.org>
		<aaknh8$1c4$1@peabody.colorado.edu>
		<3CCDEF17.E4D1DC88@cascade-sys.com> <20020429181721.0da66800.logiplexsoftware@earthlink.net>
Message-ID: <3CCE217A.FFA76653@cascade-sys.com>

Cliff Wells wrote:

> Yes, but the OP clearly stated that he had a Numeric array that he wanted to
> filter the values from.  Other than that, no argument.

Thanks for the clarification.

I heard "numeric" array.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From andreas at andreas-jung.com  Mon Apr 22 06:24:53 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Mon, 22 Apr 2002 06:24:53 -0400
Subject: word stemming / conflation
References: <3CC2C23B.DAF9057D@anansispaceworks.com> <003d01c1e9a1$7f7512f0$02010a0a@suxlap> <004701c1e9a1$e87ff170$02010a0a@suxlap> <3CC2D120.EEF3A438@anansispaceworks.com>
Message-ID: <012101c1e9e7$f1766570$02010a0a@suxlap>

See snowball.sourceforge.net ;-)

-aj
----- Original Message -----
From: "Terry Hancock" <hancock at anansispaceworks.com>
To: "Andreas Jung" <andreas at andreas-jung.com>
Cc: <python-list at python.org>
Sent: Sunday, April 21, 2002 10:48
Subject: Re: word stemming / conflation


> Andreas Jung wrote:
> > http://sourceforge.net/projects/pystemmer/
> > From: "Andreas Jung" <andreas at andreas-jung.com>
> > To: "Terry Hancock" <hancock at anansispaceworks.com>;
<python-list at python.org>
> > Sent: Sunday, April 21, 2002 22:00
> > Subject: Re: word stemming / conflation
> >
> > > I have written  wrappers for the snowball stemmers of
>
> Actually I just found that after I posted -- I was about
> to post an "oh nevermind I found it post" :-D  Turns out
> I searched sourceforge for "stemming" not "stemmer". Looks
> like just what I was looking for, anyway.  Thank you for
> what looks like a lot of work!
>
> Ironically, I guess they need a better stemmer
> too, huh? ;-D
>
> Is "Snowball" any relation to "SNOBOL"?  I'm not really
> familiar with either, being mostly a scientific programmer
> (until recently, I guess -- I find myself doing a lot
> of information storage tasks now and less number
> crunching).
>
> Thanks!
> Terry
>
> --
> ------------------------------------------------------
> Terry Hancock
> hancock at anansispaceworks.com
> ------------------------------------------------------
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>





From aahz at pythoncraft.com  Tue Apr  9 17:53:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 9 Apr 2002 17:53:45 -0400
Subject: Off topic C question
References: <4a7fde99.0204081858.58baabdd@posting.google.com> <a8tov6$484$1@panix1.panix.com> <231u8a.lc.ln@127.0.0.1>
Message-ID: <a8vnt9$roa$1@panix1.panix.com>

In article <231u8a.lc.ln at 127.0.0.1>, Carl Banks  <imbosol at vt.edu> wrote:
>Aahz wrote:
>>
>> If you want to be a C programmer, you *MUST* get K&R.  Period.  Feel
>> free to buy another book if K&R turns out to be too heavy for you right
>> now, but recognize that you'll need it eventually.
>
>Really?  I've been programming in C for ten years, and I've never even
>seen the K&R book.  I think I learned it just fine.  Not that I have a
>better suggestion.

I'm of the opinion that if you used K&R, you'd agree with me.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"There are times when effort is important and necessary, but this should
not be taken as any kind of moral imperative."  --jdecker


From skip at pobox.com  Wed Apr 17 12:29:45 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 17 Apr 2002 11:29:45 -0500
Subject: Counterintuitive Python behavior
In-Reply-To: <1019060158.737451@cswreg.cos.agilent.com>
References: <3CBD1C02.8AB2CF62@yahoo.com>
        <1019060158.737451@cswreg.cos.agilent.com>
Message-ID: <15549.41721.616995.497834@12-248-41-177.client.attbi.com>

    Greg> There should have been an operator in Python that says when
    Greg> objects are conceptually the same.  For the moment, call it "eq".
    Greg> The following would hold:

    Greg>     3 eq 3
    Greg>     not (3 eq 4)
    Greg>     a eq a            # PROVIDED a HAS A VALUE
    Greg>     "foo" eq "foo"
    Greg>     not ([] eq [])
    Greg>     not ({} eq {})
    Greg>     () eq ()

Guido's time machine strikes again:

    >>> 3 is 3
    True
    >>> not (3 is 4)
    True
    >>> a = "sdfsdf"
    >>> a is a
    True
    >>> "foo" is "foo"
    True
    >>> not ([] is [])
    True
    >>> not ({} is {})
    True
    >>> () is ()
    True

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From mcherm at destiny.com  Sun Apr  7 12:23:43 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Sun, 07 Apr 2002 12:23:43 -0400
Subject: PEP 284, Integer for-loops
Message-ID: <3CB0728F.5050800@destiny.com>

Greg Ewing writes:
> To me, the lack of a step is a *feature*, not a bug!
> 
> Steps other than 1 in a range are confusing when
> you don't land exactly on the endpoints. Which
> endpoint do you miss? Do you even hit either of
> them? And it's even worse when the step is negative.
> Do the endpoints swap over? Does the closed/open
> status of the endpoints swap over? Do you miss
> the opposite endpoint if the step isn't -1?
> Etc.
> 
> So, I see *considerable* benefit from eliminating
> this degree of freedom from the range generation
> itself, and making the programmer be explicit on
> how to map a step-1 range onto something else if
> needed.

I agree. If you want a "generalized" loop, with the ability to have 
varied step size, forward and backward looping, flexible exit or start 
conditions (inclusive, exclusive, apply-arbitrary-test), or even 
complicated things like:

John Machin writes:
 > for k in 0,1,2, 5 step 5 until 25, k * 2 while k <=1024:

Then there's a PERFECT control structure for your purpose. It's called a 
"while" loop, and it requires you to be explicit, but is otherwise 
completely flexible.

Seems to me that the key problem to solve is the awkwardness of 
addressing the common case -- simplicity is nice!

-- Michael Chermside





From dalke at dalkescientific.com  Sun Apr  7 19:52:47 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 7 Apr 2002 17:52:47 -0600
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net> <3CB08659.F3ED147@geneva-link.ch>
Message-ID: <a8qm6t$4ic$1@slb3.atl.mindspring.net>

Boris Borcic:
>BTW, I've been wondering if all the problems with voting machines
>(if not Bush's election itself) should not be ascribed to the infamous
>Y2K bug : after all, if programmers had not been monopolized by the
>task of preventing the bug hitting, maybe some work force would have
>been applied to design and implement more modern voting machines.

Getting even more off-topic,

"more modern" does not necessarily imply "more usable"
"more modern" ==/==> "more auditable"
"more modern" ==/==> "more correct"

Here's some links:
      http://www.csl.sri.com/users/neumann/neumann.html#5
      http://mainline.brynmawr.edu/~rmercuri/notable/evote.html
      http://www.asktog.com/columns/042ButterflyBallot.html

Quoting from http://catless.ncl.ac.uk/Risks/21.14.html#subj1
] In the wake of the recent Presidential election problems, the
] knee-jerk reaction of "gee, can't we modernize and solve all
] this with electronic and/or Internet voting?" is predictable,
] but still wrongheaded.  The shining lure of these "hype-tech"
] voting schemes is only a technological fool's gold that will
] create new problems far more intractable than those they claim
] to solve.

Rebecca Mercuri (in the 2nd link, above):
] I am adamantly opposed to the use of fully electronic or
] Internet-based systems for use in anonymous balloting and vote
] tabulation applications.  The reasons for my opposition are
] manyfold, and are expressed in my writings as well as those of
] other well-respected computer security experts.  At the present
] time, it is my strong recommendation that all election officials
] REFRAIN from procuring ANY system that does not provide an
] indisputable paper ballot.

I don't think the lack of programmers due to working on Y2K
remediation had any effect on the correctness of voting machines.

But, did it delay the addition of bool to Python?

                    Andrew
                    dalke at dalkescientific.com





From kybic at ieee.org  Wed Apr  3 09:03:45 2002
From: kybic at ieee.org (Jan Kybic)
Date: 03 Apr 2002 16:03:45 +0200
Subject: A faster Python?, Python compiler, Dylan,...
In-Reply-To: <3CAAEF17.32333986@kfunigraz.ac.at>
References: <mailman.1017795845.3464.python-list@python.org> <r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at> <lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at>
Message-ID: <lbh8z84nake.fsf@xihu.inria.fr>

The following message is a courtesy copy of an article
that has been posted to comp.lang.python as well.

SM>Have you tried weave?  It's part of SciPy.  From the SciPy front page:
>    Weave 0.2.1 Now Available: weave is a new package that provides tools
>    for including C/C++ code within Python code.  The weave.inline()
>    function executes C code directly within Python, weave.blitz()
>    translates Python Numeric expressions to C++ for fast execution...

Basically weave permits you to put both the C code and the Python code
into one file. Not bad, but you still have to write the code twice.
Useful for short C fragments, not long ones.

SG> Dubois dubbed this "steering your code". I do not have got much
> experience at this field (foreign functions). But why do you think that
> steering code is in the long run not the panacea? 

This is the first time I hear this  expression ('steering your code').
My major complaint is that you need to write the same code twice -
first in Python (it is so much easier to develop and debug in Python)
and then in C/C++. Moreover, you need to add the stuff to convert
forward/backward from/to Python/C. 

> I think the question actually is more or less: would you be contented
> with a Python which produces compiled code which in turn executes 10
> times slower than an equivalent C code? Normally, Python executes code

No, why should the code be 10 times slower? Low-level objects
(numbers, arrays) can be manipulated exactly the same way in both
Python and C, so you should get the same speed. 

> And the next question: It is feasible to improve Python to that level?

Yes. If you add voluntary static typing, then in the section where the
compiler knows the type of every object, it can emit exactly the same
code as a C/C++ compiler would. In other section, we just go on
interpreting the bytecode.

> What about your experience with Scilab? Or even due to INRIA: OCaml?

I do not know SciLab. I am considering OCaml, I do not like the syntax too
much, and it seems there are no true optimizing compilers.

Jan

-- 
-------------------------------------------------------------------------
Jan Kybic <kybic at ieee.org>      Robotvis, INRIA, Sophia-Antipolis, France
       or <Jan.Kybic at sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845
                    http://www-sop.inria.fr/robotvis/personnel/Jan.Kybic/



From tejarex at yahoo.com  Sun Apr 21 04:01:03 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Sun, 21 Apr 2002 08:01:03 GMT
Subject: nested exceptions?
References: <roy-AAA3A1.19193820042002@news1.panix.com>
Message-ID: <3luw8.130248$K5.11342184@bin5.nnrp.aus1.giganews.com>

"Roy Smith" <roy at panix.com> wrote in message
news:roy-AAA3A1.19193820042002 at news1.panix.com...
>     def getParentOid (self, oid):
>         parts = oid.split ('.')
>         parentParts = parts[:-1]
>         return '.'.join (parentParts)

You can do this part faster by using '.'.rfind(oid) to find *last* '.'
and then slice out part you want.

Terry J. Reedy




From jb at cascade-sys.com  Thu Apr 25 23:38:55 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 20:38:55 -0700
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>
Message-ID: <3CC8CBCF.80767E56@cascade-sys.com>

Aahz wrote:

> >Aahz wrote:
> >>
> >> Ewww!  Where did you get the idea that awk is a precursor of Python?
> >
> > JB:
> >Look up "precursor" in the dictionary and you'll see it's an accurate
> >statement.
>
> www.dictionary.com provides definitions that agree with my reaction.

1. You are in error (or intellectually dishonest) to claim that "definitions"
(plural) support your reaction, as only one of the three comes even close.

2. You should have surmised that definition #2 (merely "a predecessor" without any
other relationship) is the meaning I intended.  For me and others I know, AWK once
was our preferred (the only) utility language and it is now has largely been
replaced by Python.  Thus, for some people anyway, AWK clearly is a precursor to
Python.

Since your own reference included an innocent, neutral definition, you are out of
line to complain, even if you originally were ignorant of the alternative meaning
before looking it up.

3. For your information, my Oxford American Dictionary ONLY includes neutral
"predecessor" meanings for precursor, without the additional nuance in
dictionary.com's first definition.  To be honest, I originally thought "precursor"
might be too strong a word so I looked it up before using it.   I go to these
great lengths and I STILL wasn't sufficiently humble and deferential for this
group.

4. Actually, this all makes me wonder if your web source is accurate in the first
place?  I've come to place a lot of trust in my well worn OAD.  Shit.  Now I'll be
pouring over different dictionaries all evening.

5. Furthermore, one could argue that even the first definition is not wholly
inaccurate, that AWK in some way "indicated" that Python was "something to come."

After all, AWK was one of the original "toolbox" or "utility" languages and one
could argue that AWK's guarded success "indicated" that other utility languages
eventually "were to come."

Or I could say that AWK curiously is one of the few languages I know of other than
Python that has exactly 2 scopes (local and global) instead of a more general
scheme (since remedied in Python).  Who knows?  Maybe AWK is where the BDFL got
the idea from in the first place.  (Just kidding!)

6. In any case, let me be perfectly clear: I certainly and specifically did not
mean to imply any "heritage" or ancestry between the two languages, only that AWK
came earlier in time.

7. That being said, AWK was and is a great little language.  I still use it
occasionally because it's simpler than Python for some little things.  And I still
have some AWK code from long ago that still serves useful purposes.  And who
knows?  If Kernighan and co. had kept adding to AWK there might never have been a
point to Python.

8. I really gotta say -- some of you guys really need to lighten up and/or get a
life.  There is, was and always will be a world aside from Python.

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From fredrik at pythonware.com  Wed Apr  3 11:16:17 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 3 Apr 2002 18:16:17 +0200
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <NCBBLFCOHHDIKCAFGCFBCEPEKIAA.marklists@mceahern.com><016701c1db1a$569543e0$0900a8c0@spiff> <yu99wuvovkmm.fsf@europa.research.att.com>
Message-ID: <00af01c1db2a$e7316d10$ced241d5@hagrid>

Andrew Koenig wrote:

> Specifically: If I want to write a function that answers a yes/no
> question, I have lots of possible ways of spelling yes (1, 2, "yes",
> and so on) and lots of possible ways of spelling no (0, {}, None,
> and so on).  There isn't a single preferred way.

if you study real python code, you'll find that there is.

</F>





From tkaczyka at wharton.upenn.edu  Sun Apr 14 17:01:26 2002
From: tkaczyka at wharton.upenn.edu (Amie Tkaczyk)
Date: Sun, 14 Apr 2002 17:01:26 -0400
Subject: case3-part4
Message-ID: <a9cqn6$878$1@netnews.upenn.edu>

I keep getting the message that solver cannot come up with a feasible
solution.  Why is this happening and how do we fx it?


Amie






From skip at pobox.com  Sat Apr 20 10:12:27 2002
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 20 Apr 2002 09:12:27 -0500
Subject: Fundamental problem with urllib...
In-Reply-To: <20020420080245.GA2245@vomjom>
References: <20020420080245.GA2245@vomjom>
Message-ID: <15553.30539.78811.331692@12-248-41-177.client.attbi.com>

    Jonathan> Urllib will always have problems with some sites.  Notably,
    Jonathan> sites with cookies that also redirect.

    Jonathan> This is certainly not possible without overriding some
    Jonathan> functions in FancyURLopener....  This isn't a big deal (I did
    Jonathan> it myself for my program), but this can be simply fixed by
    Jonathan> adding support for cookies within urllib.

    Jonathan> Shall I take it upon myself to make such a patch?

Please do.  You can submit patches here:

    http://sourceforge.net/tracker/?group_id=5470&atid=305470

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From mlh at vier.idi.ntnu.no  Sat Apr 27 17:57:06 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 27 Apr 2002 21:57:06 +0000 (UTC)
Subject: What is NotImplemented ?
References: <slrnacinls.j5.alf@virgo.logilab.fr>  <lkn0vq34fi.fsf@pc150.maths.bris.ac.uk>
Message-ID: <slrnacm7li.mru.mlh@vier.idi.ntnu.no>

In article <lkn0vq34fi.fsf at pc150.maths.bris.ac.uk>, Michael Hudson wrote:
>Alexandre Fayolle <alf at logilab.fr> writes:
>
>> A quick question, here. What is the built in NotImplemented object? 
>> And what can you do with it? I thought I could raise it as an 
>> exception, but this is not the case (I use NotImplementedError instead).
>
>I think you can return it from implementations of, say, __add__ to
>mean roughly "I don't know what to do here".
[snip]

But how is this different from raising NotImplementedError? (If that
doesn't work in this case, why doesn't it?)

>Returning NotImplemented from the __add__ makes Python go off and call
>the __radd__ method.
>
>HTH,
>M.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From johnroth at ameritech.net  Sun Apr  7 19:55:34 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 7 Apr 2002 19:55:34 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net>
Message-ID: <ub1mp3ssjc7da5@news.supernews.com>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:a8pqkh$492$1 at slb1.atl.mindspring.net...
> John Roth wrote:
> >People who were using bools for your Democrat or Republican
> >example were simply using them incorrectly. In most states, there
> >were way more than two choices: the Green party got enough votes
> >to throw the election to Bush. If he hadn't run, Gore would have
> >won!
>
> Just as an off-topic reminder.  The difference between Bush and
> Gore was less than 2,000.  Given the final Florida results of
>     Bush    2,909,176, Republican
>     Gore    2,907,451, Democrat
>     Nader      96,837, Green
>     Browne     18,856, Libertarian
>     Buchanan   17,356, Reform
>     Phillips    4,280, Official Constitution Party
>     Hagelin     2,287, Natural Law
>
> That means Hagelin, the *7th* place party, had enough votes to
> "throw the election to Bush."
>
>                     Andrew
>                     dalke at dalkescientific.com

Just as off-topic. I believe it was around half a million,
nationwide. US election laws are interestingly archaic,
and definitely off-topic,  however, the Green party
had enough votes in Florida to have won the odd
electoral seat if the electoral seats had been divided
by vote, not by winner take all.

Your comment about Hagelin is also incorrect. From
the name of the party, I would assume that all Hagelin
voters would have voted for Bush (or another right
wing candidate) if he had not run, so his presence would
have 'thrown the election to Gore,' or at least tended
that way by denying Bush votes.

The point, however, is that the original poster used
an example where there were only two parties as
her example of why boolean didn't work well, and then
added a typologically distinct third case (ballot spoiled
for a specific reason.)

My point was that it was a contrived example that had
no basis in reality (consequently, it's a good example
of bad design,) and the proposed extension was an
example of even worse design.

Claiming that a syntactic construct is bad on the basis
of bad designs is, frankly, a non-starter in my opinion.

John Roth
>
>
>
>




From bokr at oz.net  Thu Apr 25 14:12:29 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Apr 2002 18:12:29 GMT
Subject: Trinary operator?
References: <a9kci2$dv0$1@nets3.rz.RWTH-Aachen.DE> <29e28c51.0204201335.6e0856da@posting.google.com> <87k7qwrf0v.fsf@pvv.org> <mailman.1019682973.29139.python-list@python.org> <871yd4odap.fsf@pvv.org> <MPG.1730fb06f5bab3e99896c8@news.nwlink.com> <sepu0okxo8.fsf@pc970145.GermanLloyd.de> <MPG.1731e3fd71e8350b9896c9@news.nwlink.com>
Message-ID: <aa9gud$qsr$0@216.39.172.122>

On Thu, 25 Apr 2002 10:59:59 -0700, Jeff Shannon <jeff at ccvcorp.com> wrote:

>In article <sepu0okxo8.fsf at pc970145.GermanLloyd.de>, 
>hoel at germanlloyd.org says...
>> Jeff Shannon <jeff at ccvcorp.com> writes:
>> 
>> > ...The only effective difference between 1 and True,=20
>> > and 0 and False, is how it's written, but they add, subtract,=20
>> > multiply, and hash the same.
>> 
>> But there is a difference:
>> 
>> >python
>> Python 2.2.1 (#1, Apr 10 2002, 17:11:02)=20
>> [GCC 3.0.4] on sunos5
>> Type "help", "copyright", "credits" or "license" for more information.
>> >>> a =3D 1
>> >>> b =3D 1
>> >>> a is b
>> 1
>> >>> a =3D True
>> >>> a is b
>> 0
>> >>>=20
>
>Yes, True and False are singleton objects that have a different 
>identity than 1 and 0.  The 'is' operator tests object identity.  
>
>>>> a = []
>>>> b = []
>>>> a is b
>0
>>>> 
>
>This doesn't change the fact that a and b are equivalent, and I 
>would still say that there is no effective difference between a 
>and b, here.  Just different object IDs.
>
Another example:
 >>> a=1
 >>> b=1
 >>> a==b, a is b
 (1, 1)
 >>> a=123
 >>> b=123
 >>> a==b, a is b
 (1, 0)

Regards,
Bengt Richter


From wilk at flibuste.net  Tue Apr 16 10:53:56 2002
From: wilk at flibuste.net (William Dode)
Date: Tue, 16 Apr 2002 16:53:56 +0200
Subject: google apis
References: <20020416092517.5a9e5845.wilk@flibuste.net>
	<3CBBD31D.AA2E653C@alcyone.com>
	<20020416115225.1d9fa2d2.wilk@flibuste.net>
	<3CBC0078.C10C5F4@alcyone.com>
Message-ID: <20020416165356.3aeb88eb.wilk@flibuste.net>

Le Tue, 16 Apr 2002 03:44:08 -0700
Erik Max Francis <max at alcyone.com> ?crivait:

> William Dode wrote:
> 
> > thanks, i write now to google to ask why they didnt spoke about.
> 
> Huh?
hooo my bad english !
I mean that on the page of google they speak about api for java,
visual.net and perl, but not for python. I would like to see python
everywhere !

> 
> -- 
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Nationalism is an infantile sickness.
> \__/ Albert Einstein
>     Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
>  A new, virtual planet, every day.


-- 
William Dod? - Informaticien Ind?pendant
http://www.flibuste.net


From grante at visi.com  Thu Apr 18 06:57:25 2002
From: grante at visi.com (Grant Edwards)
Date: Thu, 18 Apr 2002 10:57:25 GMT
Subject: Reading from a binary file in a portable way
References: <mailman.1019112514.13299.python-list@python.org> <3cbe7518_5@news.newsgroups.com>
Message-ID: <slrnabtaaa.1gv.grante@daleth.visi.com>

On Thu, 18 Apr 2002 09:28:37 +0200, jb <egrossde at yahoo.de> wrote:
>> Why can't you open the file in binary mode under Linux/UNIX?
>> 
>> You may also find the 'struct' module useful.
>
>I used os.open under MS-WIN with the flag O_BINARY and this is not supported 
>under Linux.

Use the built-in open() with the "b" flag.

-- 
Grant



From starx at pacbell.net  Wed Apr 24 00:15:52 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 24 Apr 2002 04:15:52 GMT
Subject: Php vs Python
References: <mailman.1019615325.15799.python-list@python.org> <3CC62DBE.EC54FFE6@engcorp.com>
Message-ID: <Xns91F9D8871BC3ARASXnewsDFE1@130.133.1.4>

Peter Hansen || Tue 23 Apr 2002 08:59:58p:

> "Ramil D. Tolentino" wrote:
>> 
>> Could someone give me a run down between python and php. 
>> Which is easier to use? which is better or more powerful? Thanks
> 
> Uh, PHP is better but Python is more powerful?
> 
> Or was it the other way around?
> 
> (You really need to explain something about your intended use
> before the answers will mean anything to you.  What do you
> want to accomplish?)

About all that can be said otherwise is: "PHP was built to serve dynamic 
web-pages using inline scipt, pythong was built to be a good generic 
programming language perhaps with small size projects first in mind." 

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From mlh at vier.idi.ntnu.no  Mon Apr  8 02:50:51 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 8 Apr 2002 06:50:51 +0000 (UTC)
Subject: IsList?
References: <9F9s8.37463$tg4.444710@vixen.cso.uiuc.edu>  <slrnab2b3b.1gd.gerhard@lilith.my-fqdn.de>
Message-ID: <slrnab2feb.atp.mlh@vier.idi.ntnu.no>

In article <slrnab2b3b.1gd.gerhard at lilith.my-fqdn.de>, Gerhard
=?iso-8859-15?Q?H=E4ring?= wrote:
>John wrote in comp.lang.python:
>> Hi, How can one know wheter var 'a' is list or not? Is there any
>> IsList-like function?
[snip lots of stuff]

If you want to check for sequences in general you can use the
isSequenceType function:

>>> from operator import isSequenceType
>>> isSequenceType([1, 2, 3])
1
>>> isSequenceType((1, 2, 3))
1
>>> isSequenceType("1, 2, 3")
1
>>> isSequenceType(123)
0

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From ykingma at accessforall.nl  Wed Apr  3 17:54:00 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Wed, 03 Apr 2002 23:54:00 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it>
Message-ID: <3CAB87ED.C5567DBE@accessforall.nl>

Alex,

you wrote:
> 
> Ype Kingma wrote:
>         ...
> > The problem is that a lot of code has been written assuming 1/0 behaviour
> > of boolean values.
> 
> The "problem" is that behavior is VERY useful.  So, taking it out would
> reduce productivity (by a little) as well as making the language harder to
> teach (by a little or by a lot -- I used to think "by a little", Laura's
> post is making me thing it may well be "by a lot").
> 

I wasn't talking about removing this behaviour, although I don't like it.

I said:

	After having my head set straight on this by Martin v. Loewis
	(see the thread 'but not as int subtype') I concluded in favour
	of PEP 285, with the remark that what can in the future
	not be implemented using the special methods of bool and current
	standard types should be deprecated at the same time the PEP
	gets into force.

and:

	The only example of such code I know of
	is user implementations of __getitem__, ie. objects
	indexable with square brackets, that are actually indexed with bool arguments.
	When all such user code has been fixed, there is nothing in the way
	of a real boolean type that is not an int-in-new-hat.

by which I meant that _user_code_ implementations of __getitem__()
that will be receiving bool's when they are introduced might need a simple
fix.

About the remaining current 1/0 behaviour I said:

	If there were a real boolean type in Python this backward
	compatible behaviour can be implemented in the __add__, __radd__
	and other special methods of the numeric classes (see the language reference on special
	method names: http://www.python.org/doc/current/ref/specialnames.html).
	They would work by using the value 1 or 0 when encountering a boolean value.


> > This Something/Nothing feature is indeed very useful.
> > However, it is not part of the python language, it is part of the way in
> > which the standard types are implemented.
> 
> The Language Reference manual of the Python language disagrees
> with you.  E.g. at http://www.python.org/dev/doc/devel/ref/Booleans.html :
> 
> """
> In the context of Boolean operations, and also when expressions are
> used by control flow statements, the following values are interpreted
> as false: None, numeric zero of all types, empty sequences
> (strings, tuples and lists), and empty mappings (dictionaries).  All
> other values are interpreted as true.
> """

Thank you for pointing this out. I only read the contents of both parts
(the language ref and the standard library ref) and based my post on that.
I should have read the manual completely.

> This is the Python LANGUAGE we know and love -- the LANGUAGE
> Reference says so, and us Language Lawyers know that.  You want to
> change Python, that's one thing.  But don't tell us that this part of the
> Language Reference "is not part of the python language".

I do not want to change that part of the language. I tried to
say that the backward compatible 1/0 behaviour and
the Something/Nothing behaviour can be implemented in the special methods
of the standard types, even when a non int-a-new-hat bool would be added
to the language.

> > if something_there:
> >     ...
> > else: # nothing_there
> >     ...
> >
> > A real boolean is required to control the conditional branch.
> 
> If you say so.  In this case, since we DO "control the conditional branch"
> in the current version of Python, then we HAVE "A real boolean" whatever
> that may be (as defined by the language reference, see above).  Every
> value in Python is "A real boolean", since any can "control the conditional
> branch".

On which I also said:

	Since slices are also explicitly available, I'd like a real boolean
	to be available. It fits the language.

by which I meant:
Since a boolean is implicitly there, why not have it explicitly available.
Slices are much the same: they are 'almost' implicitly there,
but they are already explicitly available. (How often do you
actually need slices? I'd expect you to use bool more often
than that if you could.)

 
> > As I said, it's not in the language, it's in the standard types.
> > The millisecond is spent on the implicit derivation of a boolean
> > in the readers' head. The  interpreter does the same thing, although
> > faster.
> 
> There's no delay at all in my mind as I parse "if zebras:" -- there
> WOULD be if that had to carry excess verbiage (to get a "real",
> "pristine" boolean type -- which fortunately the PEP does not
> propose, thanks be).  There's no delay in the minds of students
> for which the Python something/nothing concept has 'clicked'.  I
> read Laura's point as it being crucial to make that concept click,
> and the existence of True/False being a substantial obstacle to
> making it click.

This 'concept click' comes from the consistency by which the standard
types map the Something/Nothing to the True/False needed as part
of the decision in the conditional jump in 'if' and 'while'.
They are so consistent that one can (and should) forget about what happens
in between and that it python decisions can be taught as Something/Nothing
without digressing to True/False.

> > After having my head set straight on this by Martin v. Loewis
> > (see the thread 'but not as int subtype') I concluded in favour
> > of PEP 285, with the remark that what can in the future
> > not be implemented using the special methods of bool and current
> > standard types should be deprecated at the same time the PEP
> > gets into force. The only example of such code I know of
> > is user implementations of __getitem__, ie. objects
> > indexable with square brackets, that are actually indexed with bool
> > arguments. When all such user code has been fixed, there is nothing in the
> > way of a real boolean type that is not an int-in-new-hat.
> 
> Do you propose to make your "real boolean type" non-hashable,
> so that it can't be used to index into a dictionary?  Can you please
> clarify the reason for *THAT*?

Off course I want the boolean constants to be hashable.
However indexing using boolean constants instead of
integers will break some user code that assumes it's argument is an
integer, see below.

> If you do NOT want to forbid usage of your newfangled booleans
> as keys into dictionaries, then you're at least using imprecise language
> in talking about "objects indexable with square brackets", since
> dicts are indeed thus indexable (and more generally so are all
> mappings).  Besides, would adict[True] be the very same dict entry
> as adict[1], or not?  If they aren't to be the same entry, then how do
> you propose to transition smoothly from a present in which they are
> to a future in which they are not?  If they are, you need hash(True)
> to be 1 and True==1 to hold (or else dicts would be seriously slowed
> down -- a pythonic nono! -- and user-coded mappings would have
> serious problems too), so SOME level of confusability must be kept.

In the other thread 'but not as int subtype', in the context
of user code implementations of __getitem__, I also said:

	Indexing a dict with both int and bool keys might involve some confusion
	when True is not 1 anymore. This case should be rare, however.

For standard types, ie. normal dictionaries there is no problem since
one can simply add the int() method call in the implementation of
__getitem__ of the standard dictionary for backwards compatibility.
User code trying to implement a dictionary will need a similar
fix as I pointed out above. I can only hope there is no more breakage in
user code.

> I think the best levels of confusability (interchangeability, polymorphism,
> name it as you will) between concepts are "all" and "none", in general.  If
> "all", the concepts are practically coincident even though there may be
> some _conceptual_ separation that has no reflection in practice; if "none",
> the concepts are utterly separate in practice even though there may be
> some _conceptual_ common ground.  Either of these is easy to teach.
> Middle-ground is harder to teach, though there may be very good (real,
> practical -- or even philosophical) reasons to pay this price in some cases.
> 
> I don't think "none" is feasible here, and I much prefer the current status
> quo of "all" anyway.  The PEP has a lot of confusability (thanks to
> inheritance and True==1, perhaps even to str(True) being '1' depending on
> how that's decided); your proposal has very little (but, I think, not
> "none" -- see above).  So there's a conceptual price to pay, as well
> as a pragmatic one (low for the PEP, high indeed for your proposal).  And
> I _still_ don't see that the claimed benefits would actually amount to a
> substatial hill of beans.  I have nothing against beans, particularly in my
> soup, but they seem to come too dear here -- for a price I might perhaps
> be willing to pay for good steak, if that.  High price, tiny benefit, my
> personal mental scales slant more and more towards a 'nay'.

I'm afraid I'm not understanding you completely, but this may due
to the moment of the day here.

My conclusion (proposal?) from the other thread was in favour of the PEP,
with the remark that some user code that would break when bool does not
inherit from int anymore should be deprecated and fixed, in order to allow for a 
non int-in-a-new-hat bool in the future.
I think all the python system code can be fixed to withstand the change from
bool as int-a-new-hat to bool 'not as int subtype'. By python system code
I mean the C code implementing CPython and the Java code implementing Jython.
(The special methods of the standard types in this code would need
to add int() calls and evt. type tests for bool to obtain the backward
compatible 0/1 behaviour of bool.)

> > In fact this could be seen as a good example of the usefulness of
> > object orientation: it allows separation of concerns to the extent
> > that the standard types are documented separately from the language.
> 
> But that's NOT what the Python Language Reference chooses to do
> for "Booleans" -- it documents them right smack in the middle of
> the Language Reference itself.  So?

Even though I still have not read the manual completely I would (and did)
expect it to be possible that this aspect of the standard types
be defined in the library reference instead of the in the language
reference.
However, if this were the case, Laura might not have come up with
the Something/Nothing python way, so hats off to the manual makers.

Regards & thanks,

Ype

-- 
email at xs4all.nl


From akuchlin at ute.mems-exchange.org  Wed Apr 24 11:17:36 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 24 Apr 2002 15:17:36 GMT
Subject: Using 2.2 stdlib modules in 2.1
References: <3cc6c34f.3125812@news.t-online.de>
Message-ID: <slrnacdj4e.6bl.akuchlin@ute.mems-exchange.org>

In article <3cc6c34f.3125812 at news.t-online.de>, Gerson Kurz wrote:
> Can I just distribute hmac.py with my BSD-licensed open source app, or

That would be OK.  A good approach might be to add a py21 subdirectory
to your software, and append it to sys.path if the Python interpreter
is version 2.1.  That means the system hmac.py will be used with
Python versions 2.2 and up, acquiring any bug fixes that happen to be
made to the module in future versions.

--amk                                                             (www.amk.ca)
Prying curiosity means death.
    -- H.P. Lovecraft, "The Rats in the Walls"


From stephen at xemacs.org  Wed Apr  3 03:41:33 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 03 Apr 2002 17:41:33 +0900
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <gIpq8.44728$pT1.1266378@news1.tin.it> <mailman.1017787206.10470.python-list@python.org>
Message-ID: <87k7rpyy0y.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "David" == David Ascher <DavidA at ActiveState.com> writes:

    David> Alex Martelli wrote:
    >> Seconded.  _I_ am going to steal it (if I get Laura's
    >> permission, and if the PEP is rejected) for the Nutshell, for
    >> sure...

    David> I like it too.  Explaining Pythonic truth always seemed a
    David> little convoluted, but Laura's description is right on the
    David> ball.

Yup, I got very uneasy after reading the PEP and Tim Peters's posts on
it, and tried to post something with very similar motivation a couple
of days ago.  I lacked the experience and vocabulary to express it.
Thanks, Laura!

    David> I'm not sure how it translates to other languages.

In Japanese, even better than in English (where None can mean "the
void before creation", the primordial Exception).

It's "yuumu" (U+6709 U+7121) == "is some", "is not some."


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From db3l at fitlinxx.com  Thu Apr 25 20:21:20 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 25 Apr 2002 20:21:20 -0400
Subject: Python doesn't know how to do math?
References: <1jvgcukdvfeakgq8lp284dofe3e8cnrla9@4ax.com> <uit6f1jwy.fsf@ctwd0143.fitlinxx.com> <gb4hcu4b55gq99blg08eudh14pm8ivju0k@4ax.com>
Message-ID: <ug01j1f9b.fsf@ctwd0143.fitlinxx.com>

Chris Spencer <clspence at one.net> writes:

> 	I want to thank everyone who responded to this.  I read the
> floating point tutorial and now understand the issues.  I wasn't
> really looking for that much precision anyways (2 decimal places was
> all I really need).  So now that I know, I shant ask again.

Note that using str() (itself, or via print, or the '%s' formatting
operator) will give you a reasonable presentation, but the amount of
precision may vary and not under your control.

One way to ask for specific precision is by using the '%f' formatting
operator, e.g.:

    >>> value = 5.0/100.0
    >>> repr(value)
    '0.050000000000000003'
    >>> str(value)
    '0.05'
    >>> '%f' % value
    '0.050000'
    >>> '%6.2f' % value
    '  0.05'
    >>> '%.2f' % value
    '0.05'
    >>> '%.4f' % value
    '0.0500'

> BTW, this is one thing I love about the Python community.  Ask a
> stupid question and you get good and helpful answers.

While declaring Python unable to do math as in your subject line may
have been a bit presumptuous, I don't think the question itself is
stupid (common maybe, thus the FAQ and documentation, but not stupid).
It can be very easy to overlook the binary nature of most floating
point systems nowadays until you get burned the first time.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From phr-n2002a at nightsong.com  Wed Apr 17 04:06:28 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 17 Apr 2002 01:06:28 -0700
Subject: Pure python crypto
References: <mailman.1019024313.11744.python-list@python.org>
Message-ID: <7x3cxubvgb.fsf@ruckus.brouhaha.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> writes:
>   Would anyone have any use for a pure python implementation of
> Blowfish? Or any other crypto algorithms for that matter.  As a matter
> of interest, I found myself implementing the blowfish module recently
> but was unsure if any audience would be interested. No C compiler
> is required, which is nice. Could use some testing though.

I wouldn't mind having it, but it wouldn't be THAT useful, as it
would be awfully slow, and I don't know of any important programs
to interoperate with that need Blowfish.

DES/3DES seems more important for interoperability purposes.  I coded
them in pure Python but made a mistake somewhere, so the test vectors
don't work correctly, and I never got around to debugging it.  I should
take a look at it again.

If you just want to encrypt in Python without custom C modules, after
a fair amount of experimentation the fastest way I found is to use the
SHA module in output feedback mode to generate a keystream from the
key and a random initialization string, and use the array module to
combine the keystream with the plaintext.  This works pretty well on
Un*x/Linux because the /dev/urandom device makes it easy to get the
initialization strings.  However, there's no good source of such
strings under Windows using the standard modules at the moment.



From maxm at mxm.dk  Fri Apr  5 16:55:36 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 23:55:36 +0200
Subject: Booleans, integer division, backwards compatibility; where is
 Python going?
References: <mailman.1018040171.16545.python-list@python.org>
Message-ID: <3CAE1D58.4050500@mxm.dk>

brueckd at tbye.com wrote:

>>because of the problems I have
>>actually encountered with making code work with the many extant versions of
>>Python, I will no longer be doing any new projects in the language and am
>>not recommending Python to anyone.

> So, what language did you move on to?

386 assembler probably. It's a little less portable, but the specs 
haven't changed in over a decade.

regards Max M



From bokr at oz.net  Wed Apr 10 18:10:15 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 22:10:15 GMT
Subject: Extracting from a list
References: <jjb5-B4241A.16351010042002@newsstand.cit.cornell.edu>
Message-ID: <a92d87$pvp$0@216.39.172.122>

On Wed, 10 Apr 2002 16:35:10 -0400, Joel Bender <jjb5 at cornell.edu> wrote:

>Hi,
>
>
>I would like to process the elements of a list that match a condition, 
>then remove them.  I'll use integers in my example, but my list items 
>are actually objects.
>
>What's the "right" way to do this?  For example:
>
>    >>> lst = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
>    >>> dothese = [i for i in lst if i < 5]
>    >>> lst = [i for i in lst if i >= 5]
>    >>> for item in dothese:
>    ...     print item
>
>I'd rather not scan the list twice, particularly if there's nothing to 
>do.  I'd rather not build another double-linked list.
>
You could do something like:
 >>> def prck(i):
 ...    if i>=5 : return 1
 ...    print i
 ...    return 0
 ...
 >>> lst  = [3, 2, 5, 1, 0, 7, 4, 8, 6, 9]
 >>> lst = [i for i in lst if prck(i)]
 3
 2
 1
 0
 4
 >>> lst
 [5, 7, 8, 6, 9]

Regards,
Bengt Richter


From phr-n2002a at nightsong.com  Wed Apr 24 19:12:33 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 24 Apr 2002 16:12:33 -0700
Subject: Twisted ircd is written in python
References: <slrnac9ak3.fho.ak@ak.silmarill.org> <aa7dn4$402$1@nsnmrro2-gest.nuria.telefonica-data.net>
Message-ID: <7xit6gzo66.fsf@ruckus.brouhaha.com>

"Raphael Ribeiro" <gusraan at terra.com.br> writes:
> > google for twisted matrix
> didn't find it :(

Sheesh, it's www.twistedmatrix.com.


From spivey at ieee.org  Thu Apr 18 00:43:08 2002
From: spivey at ieee.org (Gary Spivey)
Date: Thu, 18 Apr 2002 04:43:08 GMT
Subject: Hostname Resolution not working in Explorer
References: <edrv8.17246$L1.1349660@newsread2.prod.itd.earthlink.net>
Message-ID: <w9sv8.17352$L1.1367388@newsread2.prod.itd.earthlink.net>

Looks like I screwed up my first post ...

I have windows XP Home, and IE6. I have a zope server running on my box. The
server is running fine, and I can access if it I access

http://127.0.0.1:8080/manage

However, if I use

http://localhost:8080/manage

it doesn't work. Same holds true if I use xx.xx.xx.xx:8080/manage and
hostname:8080/manage.
The names don't work, the numbers do. And they all work just fine when
pinging from a command window. So, something between me, internet explorer,
and the name resolution isn't working.


"Gary Spivey" <spivey at ieee.org> wrote in message
news:edrv8.17246$L1.1349660 at newsread2.prod.itd.earthlink.net...
>
> I have windows XP Home, and IE6. I have a zope server running on my box.
> The server is running fine, and I can access if it I access
>
> http://127.0.0.1:8080/mange
>
> However, if I use
>
> http://localhost/mange
>
> the page isn't found. The same holds true if I use my hostname rather than
> localhost - also, the actual IP address works.
>
> I have made sure that the name entries exist in the
> window/system32/drivers/etc/HOSTS file, and I can ping both localhost and
my
> hostname from a command window (with the correct address resolution).
>
> So - why can I not get the resolution to work from within Internet
Explorer?
>
> Cheers
> Gary
>
>
>




From tundra at tundraware.com  Mon Apr 22 14:10:03 2002
From: tundra at tundraware.com (Tim Daneliuk)
Date: Mon, 22 Apr 2002 18:10:03 GMT
Subject: Science And Math  Was: Python's Lisp heritage
References: <20020420092101.19095.84453.Mailman@mail.python.org> <mailman.1019299357.11513.python-list@python.org> <a9rn4i$mm4$1@newsreader1.netway.at> <a9s9me$890$1@panix1.panix.com> <Xns91F67C40EB0E0lumberjack@216.148.53.83> <slrnac3nqj.1ec.grante@tuxtop.visi.com> <tnos9a.de.ln@127.0.0.1> <uc59vi7etg1j2c@news.supernews.com> <k5uu9a.621.ln@127.0.0.1> <3CC3A89C.BE2BF65@tundraware.com> <dit7cuo1nguiha34nq80m9q39igpjlr70d@4ax.com>
Message-ID: <3CC45005.2E9C0F86@tundraware.com>

"Gon?alo Rodrigues" wrote:
<SNIP>

> 
> You should understand that this is a philosophical viewpoint. The
> mathematical field is roughly divided in the Platonist field, those who
> believe that mathematical objects exist somewhere out there in space (to
> quote Sun Ra) in a Wonderful World of Platonic Ideas, and the formalist
> field of those who believe mathematics is essentialy a game we play on
> paper with symbols with no "real" meaning attached to it besides the
> convenience in describing natural phenomena.

Yup, I'm aware of this - I didn't want to digress more into philosophy
than absolutely necessary.  IIRC, the Enlightenment thinkers also viewed
mathematics as a branch of natural science and they were (mostly) 
Aristoteleans, right?

<SNIP>

> You should be careful when you say this. There is *virtually no*
> mathematical field that has not been applied.

True enough, but that is not a sufficient condition to argue that mathematics
is correspondent one-for-one to the natural world.  All I'm arguing here is that
mathematics embraces a problem set *larger* than just the natural world and
this cannot simply be a natural science.


<SNIP>

> This is plainly not true. String theory models live in 10 and 24
> dimensions. You would not want to disparage the biggest fad in current
> theoretical physics, would you?


Well, OK, I'll say it more formally:  Let n denote the hightest dimension
used by any of the natural sciences.  Mathematicians routinely operate
in k dimensions where k >> n.   Acceptable?

> 
> Even so let me reiterate, that almost everything that mathematicians
> have concocted is currently used in theoretical physics. To give you
> just an example, my own, I use homotopy and higher order category
> theory.
> 
> Even more when mathematics advances by theoretical physics promptings. I
> could give you numerous examples, but just remind you that quantum
> mechanics contributed a LOT (the celebrated Von Neumann, for example) to
> the growth of abstract mathematics (functional analysis, Hilbert spaces,
> operator algebras, etc.). Some of the more active work in mathematics
> today is also prompted by theoretical physics.
> 
> Somehow (I do not intend to make this more precise now) new mathematical
> formalisms and the discover of new laws in the Universe go hand in hand.


Again, my claim was not that mathematics cannot be/is not appliable, merely
that it is bigger than just its applications.

> >
> >I'm not choosing sides here on the math vs. science debate - both are important
> >artifacts of the human intellect.  I merely take umbrage with the notion that
> >mathematics is innately wired into the universe somehow.
> 
> See above to understand that what you are saying is really not that
> obvious.


Well almost nothing about epistemology is obvious.  As I said, I am willing
to be persuaded otherwise, but it sure looks to me like mathematics is the
construct I claim, not innate to the natural world...

-- 
------------------------------------------------------------------------------
Tim Daneliuk
tundra at tundraware.com


From Answer.via.news.please.prikryl at skil.nospam.cz  Thu Apr 11 04:21:27 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Thu, 11 Apr 2002 10:21:27 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><010420021349434531%alexis.layton@post.harvard.edu> <mailman.1018508732.24530.python-list@python.org>
Message-ID: <a93h3m$st0$1@news.nextra.cz>

"Andreas Kostyrka" wrote...
>
> Alexis Layton <alexis.layton at post.harvard.edu> wrote:
>
> > As for int-compatibility, were we designing Python
> > from the ground up it might be the case that it
> > should be designed as a pure truth type. But we
> > are constrained by Pythons 1-2.2.x so True has to
> > be truly-comparable to 1.

This was also the problem in C++ where the
comparability of bool and int had to be solved
basically for the same reasons.  In C++, new
built-in bool type is strict and still it is
possible to mix it with int.

The mechanism for mixing it is called *implicit
conversion* that is done for the recognized
*contex* of the expression.  In other words, when
boolean value is found where integer is expected,
the implicit conversion of true to 1 is done
(false to zero).  When the opposite is required,
the implicit conversion from zero to false or !=0
to true is done.  When the conversion is
suspicious (i.e. when the compiler guessed that
the value is used in some very unexpected way --
say (35*false) -- then the warning may appear).

This way, it is possible to introduce strict bool
and not to break the old code.

> Nope. It must be truly-comparable to any non-zero
> object. As it happens this is what Python uses now
> as definition of Truth :) False OTOH must
> truly-comparable to ["",u"",None,{},[]] (and
> whatever I've just forgot).
>
> I personally have often written code that uses one
> variable to double as "boolean indicator" and
> "value storage":
>
> founditem=None
> while not founditem:
> ....
> if ....:
> founditem="VALUE"
>
> print "Found item with value:",founditem

This code would be preserved if Python itroduced
the strict bool together with implicit conversion
of int to bool and bool to int.

See details at my original "True/False story...
(PEP 285)".  The HTML version of the original is
temporatily available at
http://www.skil.cz/test/TrueFalseStory.html

Petr
--
Petr Prikryl (prikrylp at skil dot cz)




From rjones at ekit-inc.com  Mon Apr 29 23:25:41 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Tue, 30 Apr 2002 13:25:41 +1000
Subject: Call for kjBuckets.c [was Re: relational database?]
In-Reply-To: <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com>
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <mailman.1020024596.32196. <jLbz8.7885$M63.7239@atlpnn01.usenetserver.com>
Message-ID: <200204301324.25823.rjones@ekit-inc.com>

On Mon, 29 Apr 2002 23:15, Steve Holden wrote:
> "Oleg Broytmann" <phd at phd.pp.ru> wrote in message
> news:mailman.1020024596.32196.python-list at python.org...
>
> > On Sun, Apr 28, 2002 at 08:11:36AM -0700, Bob Hyatt wrote:
> > > I have had a fruitless search for Gadfly. As you noted, the website for
> > > Chordate is not responding.
> >
> >    I am keeping (temporary) a copy of kjbuckets at
> > http://phd.pp.ru/Software/Python/misc/ and mirrors:
> > http://phd.by.ru/Software/Python/misc/
> > http://phd2.chat.ru/Software/Python/misc/
> >    I am looking for a permanent maintainer.
> >
> >    Gadfly now lives in the Zope sources at www.zope.org.
>
> Oleg, hi! I also remember seeing somewhere patches that moved gadfly to re
> usage? gadfly.sf.net is a placeholder, it seems to me.

I'm working on repackaging gadfly right now and will place the  results up on 
gadfly.sf.net. I've started with the Zope version, so the re conversion is 
done. I've also integrated Oleg's kjbuckets efforts. I'm almost there... just 
got a few more wrinkles  to sort out before I make the first release.


   Richard





From ngps at vista.netmemetic.com  Tue Apr 16 11:35:35 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 16 Apr 2002 15:35:35 GMT
Subject: httplib drops ssl connection to MS IIS Server
References: <mailman.1018741166.19994.python-list@python.org> <slrnabltto.f08.jon+usenet@snowy.squish.net>
Message-ID: <a9hgc7$r0d$1@coco.singnet.com.sg>

According to Jon Ribbens  <jon+usenet at unequivocal.co.uk>:
> In article <mailman.1018741166.19994.python-list at python.org>, Marco Caamano wrote:
> > socket.sslerror: (5, 'EOF occurred in violation of protocol')
> 
> It is a bug in MS IIS, it just closes the SSL socket rather than going
> through the SSL shutdown negotiation. I don't think you can work
> around this using httplib.

Nobody is accusing MS IIS of being bug-free, but I think in this instance
the bug is with Python's SSL. Here's a slightly-reworked version of the
OP's code that uses M2Crypto.httpslib.HTTPSConnection:

from M2Crypto import SSL, httpslib

def testurl3():
    ctx = SSL.Context()
    ctx.set_info_callback()
    print '\n\nUsing HTTPS request'
    body = '\202test'
    hs = httpslib.HTTPSConnection('209.154.200.218', 443, ssl_context=ctx)
    hs.set_debuglevel(4)
    hs.putrequest('POST', '/scripts/gateway.dll?Transact')
    hs.putheader('Content-Type', 'x-Visa-II/x-auth')
    hs.putheader('Content-Length', str(len(body)))
    hs.endheaders()
    hs.send(body)
    resp = hs.getresponse()
    print resp.read()
    return

if __name__ == '__main__':
    testurl3()

Running it gives the following output:

----------
Using HTTPS request
LOOP: SSL connect: before/connect initialization
LOOP: SSL connect: SSLv2/v3 write client hello A
LOOP: SSL connect: SSLv3 read server hello A
LOOP: SSL connect: SSLv3 read server certificate A
LOOP: SSL connect: SSLv3 read server done A
LOOP: SSL connect: SSLv3 write client key exchange A
LOOP: SSL connect: SSLv3 write change cipher spec A
LOOP: SSL connect: SSLv3 write finished A
LOOP: SSL connect: SSLv3 flush data
LOOP: SSL connect: SSLv3 read finished A
INFO: SSL connect: SSL negotiation finished successfully
send: 'POST /scripts/gateway.dll?Transact HTTP/1.1\r\n'
send: 'Host: 209.154.200.218:443\r\n'
send: 'Accept-Encoding: identity\r\n'
send: 'Content-Type: x-Visa-II/x-auth\r\n'
send: 'Content-Length: 5\r\n'
send: '\r\n'
send: '\x82test'
reply: 'HTTP/1.1 100 Continue\r\n'
header: Server: Microsoft-IIS/4.0
header: Date: Tue, 16 Apr 2002 15:42:50 GMT

ALERT: write: warning: close notify
----------

The "LOOP", "INFO", etc. stuff shows SSL protocol negotiation in action.
No protocol violation. 

M2Crypto also wraps OpenSSL. Does a better job than the current Python's,
if I do say so myself. ;-)


> This is discussed in SourceForge bug id 494762. That was closed
> without fix so I have opened a new one 544234.

Perhaps an interested party may wish to explore grafting M2Crypto's
ssl_read() and ssl_write() routines to Python's socket module. I am busy
bootstrapping my start-up, so I can't commit any effort to this, sorry.

Alternatively, just use M2Crypto. 


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



From juenglin at informatik.uni-freiburg.de  Tue Apr 30 08:34:49 2002
From: juenglin at informatik.uni-freiburg.de (Ralf Juengling)
Date: 30 Apr 2002 14:34:49 +0200
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com>
Message-ID: <xdur8kx1i1i.fsf@leto.informatik.uni-freiburg.de>

Erik Max Francis <max at alcyone.com> writes:

> Ralf Juengling wrote:
> 
> > So, what is the proper, 'pythonic' way of argument checking?
> 
> The "Pythonic" way of argument checking is to simply use it and catch
> the exceptions that will be thrown if it doesn't behave as planned.

Hm, you might risk to destroy something this way (i.e. when an exception
occurs to late). 

> 
> Depending on what you want, some of the is... functions in the operator
> module might be to your liking.

Thanks, I wasn't aware of these (looked for them in the type module).
*BUT*, these (eg. isSequenceType) work only for C extension types, right?

Ralf


From fperez528 at yahoo.com  Tue Apr  9 18:02:48 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 09 Apr 2002 16:02:48 -0600
Subject: using NumPy for histogram plot
References: <3cb358b4@news.isc.rit.edu>
Message-ID: <a8vof9$rmt$2@peabody.colorado.edu>

Matt Monacelli wrote:

> I am trying to create a histogram plot of some data and am having a tough
> time with the Numeric documentation on how to do this.  I am running on a
> Windows machine (not Unix).  I am also trying to display this plot within a
> wxPython GUI app.  THanks for any help...
> 
> Matt

google(Konrad Hinsen Scientific Python)

It has a histogram implementation which is very nice. Incidentally, it can be 
plotted straight out with grace's grace.histoplot(), but I don't know if 
grace runs under Windows or not. But at least this will get you the 
histogram.

Cheers,

f


From noah at noah.org  Fri Apr 12 02:29:32 2002
From: noah at noah.org (Noah)
Date: Thu, 11 Apr 2002 23:29:32 -0700
Subject: ASCII to HEX conversion
In-Reply-To: <34e98ec2.0204111415.504d246f@posting.google.com>
Message-ID: <PAENINBOOODFDEPOKPKCOEOACKAA.noah@noah.org>

Hex should work. What's the problem with it?
What you also need is ord() to convert a string character
into a number. Try this:
	ASCII = 'A'
	HEX = hex ( ord(ASCII) )
	print HEX
	# If you don't want the "0x" 
	# at the beginning then just clip it off:
	print HEX[2:]

Yours,
Noah

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Matt
Sent: Thursday, April 11, 2002 3:16 PM
To: python-list at python.org
Subject: ASCII to HEX conversion


I need to write a program that converts ASCII test into the
hexidecimal equivalent.  It would help if it could convert to binary,
too.

Example:

ASCII = "A"
HEX = 41

I tried using the hex() function, but it does not work.  If anyone
could tell me how to do this without making a lookup table of all the
ASCII values.


Matt Leisinger
leisinmr at purdue.edu





From phr-n2002a at nightsong.com  Fri Apr 12 21:19:38 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 12 Apr 2002 18:19:38 -0700
Subject: porting python to embeded devices
References: <_lvs8.33252$lU2.1209375@wagner.videotron.net> <7x1ydkikml.fsf@ruckus.brouhaha.com> <slrnabeqci.s6b.grey@teleute.dmiyu.org>
Message-ID: <7xy9fse6ol.fsf@ruckus.brouhaha.com>

Steve Lamb <grey at despair.dmiyu.org> writes:
> On 12 Apr 2002 16:04:50 -0700, Paul Rubin <phr-n2002a at nightsong.com> wrote:
> > I'd say it's hopeless or at least impractical.  Those processors are
> > much too constrained to run Python.  It would be like trying to run an
> > SQL database on them.  Python has many virtues but being able to run
> > on very tiny processors isn't one of them.
> 
>     What about something like the Zaurus SL-5500?

The SL5500 is a high powered computer with a 32-bit CPU and 64 MB of
ram.  It runs Python just fine.  The computers the OP was asking about
are 8-bit CPU's with a few dozen to a few hundred bytes of ram (well
maybe a few kbytes for some 68HC11 configurations).  There's a
tremendous difference.  While there are some C compilers that can
produce code for those cpu's, they are normally programmed in assembler.


From info at medocom.com  Sun Apr 28 08:49:11 2002
From: info at medocom.com (Gert-Jan Dorenbos)
Date: Sun, 28 Apr 2002 14:49:11 +0200
Subject: Newbie question to setting PYTHONPATH environment var on Linux
Message-ID: <3CCBEFC7.4010403@medocom.com>

I am trying to use Python 2.1 on Suse Linux, but experience some basic 
problems as I am new to both.

I managed to set the PYTHONPATH to my directory with own modules in the 
~/.bashrc directory. This works fine if I start Python from a shell.

However, when I run Idle this does not work and the interpretor 
apparently does not read the PYTHONPATH variable.

Any suggestions?



From jimd at vega.starshine.org  Wed Apr  3 06:00:28 2002
From: jimd at vega.starshine.org (Jim Dennis)
Date: 3 Apr 2002 11:00:28 GMT
Subject: UnixDatagramServer
Message-ID: <a8encc$2nqj$1@news.idiom.com>

 What part am I missing:

#!/usr/bin/env python2.2
import SocketServer, sys, os, socket

class USrvr(SocketServer.UnixDatagramServer): pass

class RHandler(SocketServer.BaseRequestHandler):
	num = 0 
	def handler(self):
		conn = self.request
		msg = [] 
		while 1:
			data = conn.recv(65535)
			if not data: break
			msg.append(data)
		self.messages.append(msg)
		self.wfile.write("%s\r\n" % self.num)
		self.num += 1


if __name__ == "__main__":

	try: 		os.unlink("/tmp/svr.socket")
	except: 	pass

	svr = USrvr("/tmp/svr.socket", RHandler)
	svr.serve_forever()


  I'm just trying to write a server (Unix domain sockets) that
  will be a blackhole for any text send to it and return an 
  incremented integer for every new connection/transaction.

  It runs, it seems to bind to the socket and accept connections 
  on it. (I can see the socket open in lsof and I can run the following
  commands from an interactive session:

from socket import *
a = socket(AF_UNIX, SOCK_DGRAM)
a.connect('/tmp/svr.socket')
a.send('tst')
3
a.recv(1)

   ... and it simply blocks.

   So, the question is how to I get a response back!




From tim.one at comcast.net  Mon Apr  8 21:53:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 08 Apr 2002 21:53:01 -0400
Subject: Norton problem
In-Reply-To: <A9rs8.186857$VJ1.15080324@bin3.nnrp.aus1.giganews.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEEHOMAA.tim.one@comcast.net>

FYI, Norton was impossible to deal with when this problem first surfaced,
and put quite a spin on the story.  I gave them a tiny self-contained C
program showing that *any* program calling Microsoft's popen() function
caused NAV to blue-screen crash in the same way.  Tech support didn't have a
clue what to do with the C code, or about what it implied.  They told me to
buy a new version of NAV and see whether it still happened.  Now it appears
they're trying to say it's a flaw in Python.  Oh well.

at-least-they-ignored-guido-too<wink>-ly y'rs  - tim





From shalehperry at attbi.com  Wed Apr 10 17:23:57 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 10 Apr 2002 14:23:57 -0700 (PDT)
Subject: Indenting lines 400..457 help
In-Reply-To: <a929tk$ik2$1@pcls4.std.com>
Message-ID: <XFMail.20020410142357.shalehperry@attbi.com>

On 10-Apr-2002 Ned Harvey wrote:
> For the life of me, I can't find any easy way to indent (or unindent) lines
> numbered 400..457 or whatever.  Does anyone have any good suggestions?
> 
> I'm usually a vi user, but I'll happily learn a command of some other
> format.  I've tried
>     :400,457s/(.*)/   \1/g
> and a ton of variations of the same idea, but it seems that the
> search/replace doesn't do what I expect it to.
> 
> Thanks for your help.
> 

as an ex-vi user I will resist the urge to point out how easy this is under
python mode in emacs (-:

you are thinking about the problem too hard.  What you want to do is simply add
four mor spaces to the beginning of the line right?

:400,457s/^/    /

voila.




From duncan at NOSPAMrcp.co.uk  Mon Apr 15 07:04:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Mon, 15 Apr 2002 11:04:31 +0000 (UTC)
Subject: fp/lambda question
References: <a97s9h$gfg$1@reader07.wxs.nl> <KyRt8.11643$b62.313960@news1.tin.it> <a9a7sa$qd3$1@reader05.wxs.nl>
Message-ID: <Xns91F179739F715duncanrcpcouk@127.0.0.1>

"Jozef" <jozef at jozef.joz> wrote in news:a9a7sa$qd3$1 at reader05.wxs.nl:

>> def relatively_prime(a, b): return gcd(a, b) == 1
> 
> Thats funny, I wanted to do that but I don't know how to use filter
> with a function that takes 2 variables. E.g., to get a list if all
> numbers less than 10 relatively prime to 3, I would write
> 
>      filter(relatively_prime, range(1, 10)) # aaah, where does the 3
>      go??? 
> 
If you want this readable then use a list comprehension:
   [ n for n in range(1,10) if relatively_prime(n, 3) ]

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bkc at Murkworks.com  Tue Apr 30 18:13:32 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Tue, 30 Apr 2002 18:13:32 -0400
Subject: OT: Happy Birthday Larry Niven
References: <3cce9f3a$1_6@goliath.newsgroups.com> <3CCEBD17.5040903@jerf.org>
Message-ID: <3ccf161d$1_7@goliath.newsgroups.com>

Guess I better go back and re-read the book.

As I recall.. Moties are essentially genetically designed for their
particular task. .As in the case of the vehicle drivers who swerved around
each other and looked like they were going to crash, but didn't.

If they used Perl, they would crash.

Besides, Motie engineers could recombine essential elements quickly to suit
their needs, in an efficient manner. That smacks of "batteries included" and
I'd say Perl code could not be recombined..

--
Novell DeveloperNet Sysop #5

_
"Jeremy Bowers" <newsfroups at jerf.org> wrote in message
news:3CCEBD17.5040903 at jerf.org...
> Brad Clements wrote:
> > I think Python is a language that Motie Engineers would use...
>
> Oh goodness, no. They'd definately *prefer* perl; Python is too damned
> clean and very, very human.
>
> Even perl's probably a little on the clean side, for them. Motie
> engineers would go beyond "There's more then one way to do it", they'd
> want "It does more then one thing at a time." They'd probably like a
> language where every *char* can do at least two or three things. My
> guess is assembler with a suitably kooky CPU.
>
> This is a Motie programmer at work:
> http://tuxedo.org/jargon/html/The-Story-of-Mel.html . See, he uses the
> same opcodes as both code *and* data at some points... just as an example.
>




-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From kamikaze at kuoi.asui.uidaho.edu  Wed Apr  3 21:24:55 2002
From: kamikaze at kuoi.asui.uidaho.edu (Mark 'Kamikaze' Hughes)
Date: 4 Apr 2002 02:24:55 GMT
Subject: GUI Programming with Python using Qt
References: <3ca8a8fe$0$45639$e4fe514c@dreader3.news.xs4all.nl> <1d6q8.118888$Ou.77410@atlpnn01.usenetserver.com> <3ca940d3$0$32314$e4fe514c@dreader4.news.xs4all.nl> <slrnaak6oe.s4j.mlh@vier.idi.ntnu.no> <3caabd6d$0$45613$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <slrnaanebn.1jcm.kamikaze@kuoi.asui.uidaho.edu>

Wed, 03 Apr 2002 10:24:59 -0100, Boudewijn Rempt <boud at valdyas.org> spake:
>>   http://www.opendocspublishing.com/pyqt/
> I've forwarded this to Joshua -- I haven't tried buying one, of course :-).

  It would also be very nice if the opendoc version had a nice single
download archive - I had to fiddle with wget and do a lot of sed work to
fix the URLs so I could read it on my local HD and not hit the site
every time I turn the page.

  I'd really MUCH rather be able to buy just a CDROM with the book in
clean HTML (go ahead and take a look at the generated HTML...  Evil!)
with all the illustrations (which aren't in the opendoc version), but
failing that, I'll probably pick up the dead-tree edition.

  So far it seems to be an excellent book - just last week I started
teaching myself PyQt, so it came out at the right time.  I've skipped
over all the BlackAdder stuff, but I may try that out next.

-- 
 <a href="http://kuoi.asui.uidaho.edu/~kamikaze/"> Mark Hughes </a>
"No one is safe.  We will print no letters to the editor.  We will give no
space to opposing points of view.  They are wrong.  The Underground Grammarian
is at war and will give the enemy nothing but battle." -TUG, v1n1


From hbl at st-andrews.ac.uk  Thu Apr 25 07:49:10 2002
From: hbl at st-andrews.ac.uk (Hamish Lawson)
Date: Thu, 25 Apr 2002 12:49:10 +0100
Subject: Dynamically changing an object's class - still to be allowed?
Message-ID: <3CC7ED36.9B1E57A0@st-andrews.ac.uk>

I came across a debate that ran on the list while Python 2.2 was being
developed that suggested (if I understood correctly) that Guido was
planning to limit the ability to dynamically change an object's class by
assigning to its __class__ attribute. It was indicated that these
limitations might be implemented for Python 2.2, but they don't appear
to have been; might they yet?

Hamish Lawson




From emile at fenx.com  Fri Apr  5 09:49:18 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 05 Apr 2002 14:49:18 GMT
Subject: file rewind?
References: <_7ir8.1063$c6.130376@typhoon.mn.ipsvc.net>
Message-ID: <OPir8.173032$ZR2.90014@rwcrnsc52.ops.asp.att.net>

joel8bit> I have this code:
>
> def countLines(file):
[snip]

While you're probably experimenting with various things, for files that
fit in available memory the specific results you're going for here can
be got at using:

def test(name):
    lines = len(open(name).readlines())
    chars = len(open(name).read())

or, depending on what you're after, using open(name, 'rb').

--

Emile van Sebille
emile at fenx.com

---------



From aleax at aleax.it  Tue Apr 30 17:02:24 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 30 Apr 2002 21:02:24 GMT
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com> <xdur8kx1i1i.fsf@leto.informatik.uni-freiburg.de> <3CCEDA38.2559BE98@alcyone.com>
Message-ID: <ADDz8.83454$vF6.2479949@news2.tin.it>

Erik Max Francis wrote:
        ...
> No, they work with any object:
> 
>>>> import operator
>>>> operator.isSequenceType([1, 2, 3])
> 1
>>>> class MyList(list): pass
> ...
>>>> operator.isSequenceType(MyList())
> 1

>>> import operator
>>> class X: pass
...
>>> operator.isSequenceType(X())
1
>>>


Alex



From robin at jessikat.fsnet.co.uk  Sat Apr  6 09:57:17 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 6 Apr 2002 15:57:17 +0100
Subject: Stability and change
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk>
 <mailman.1018047273.8088.python-list@python.org>
 <Xns91E7BCCD31889JamesLLugojcom@216.148.53.84> <3CAEFC4E.2778AC55@python.org>
 <uDDr8.66496$S52.2342405@news2.tin.it>
Message-ID: <hUlpCJANzwr8Ew0a@jessikat.fsnet.co.uk>

In article <uDDr8.66496$S52.2342405 at news2.tin.it>, Alex Martelli
<aleax at aleax.it> writes
><posted & mailed>
>
>Guido van Rossum wrote:
>        ...
>> About the change rate in general: it's hard to find the proper pace.
>        ...
>> for faster change.  I have to pick a middle ground (and it's easier to
>
>Linus Thorvalds seems to have done pretty well with picking TWO middle 
>grounds -- two parallel tracks for Linux, "stable" and "experimental".
>
>Without the former he'd have lost the commercial interests, and "normal
>people" who DO need SOME level of stability; without the latter, a crucial 
>hard-core of neophile hackers might have gone away in search of other, more 
>dynamic projects.  With both tracks, things seems to have been smoother.
>
>Why don't we start thinking of a similarly dual-tracked Python -- one
>track aiming at stability (in a "middle ground" sensible sense: strong
.....
>
>Alex
>
excellent suggestion
-- 
Robin Becker


From chris at cmb-enterprises.com  Tue Apr 30 01:28:13 2002
From: chris at cmb-enterprises.com (Chris)
Date: Tue, 30 Apr 2002 01:28:13 -0400
Subject: Python vs. Perl,  which is better to learn?
References: <df30afd4.0204292111.31e5914a@posting.google.com>
Message-ID: <chris-4E0B1A.01281230042002@corp.supernews.com>

In article <df30afd4.0204292111.31e5914a at posting.google.com>,
 xiaotsing_79 at hotmail.com (David) wrote:

> I have a lot of experience in ANSI C/C++, I want to learn a new
> language and which it can help me in the work.
> I just want to clear which(python or perl) is more comfort for me?

In terms of how it looks, Perl might look more comfortable, but Python 
is likely to function more similarly to C++, while C will have more in 
common with Perl.

Learn both.


From tl_news at nexgo.de  Fri Apr 19 02:33:52 2002
From: tl_news at nexgo.de (Tino Lange)
Date: Fri, 19 Apr 2002 08:33:52 +0200
Subject: Accessing mdb(access 2000) files on linux using python
References: <508cb8ae.0204170837.6f800a24@posting.google.com>
Message-ID: <qdevbug0bqopbve9v8tsr5elb5fo1gdl1j@4ax.com>

On 17 Apr 2002 09:37:03 -0700, martin_byrne2 at yahoo.co.uk (B) wrote:
>Just wondering if anyone knows how to manipulate mdb files stored on a
>linux webserver using python.Using win32com on windows is fine but on
>linux can i manipulate the database using pythons odbc connectivity or
>is there another way.

>Any suggestions, links or sample code would be of great help.

You can use the Access ODBC-Driver on the Widows box and export it to
your Linux machines using the (commericial) ODBC-ODBC-Bridge.
(http://www.easysoft.com/products/2002/main.phtml)

Then it's no proble at all to access the Aceess Files, Even Excel
Sheets and so on from Linux/Unix using for example mxODBC.

Hope this helps

Tino



From pyth at devel.trillke.net  Wed Apr 17 12:24:07 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 17 Apr 2002 18:24:07 +0200
Subject: pre-PEP for optional 'pass'
In-Reply-To: <20020417092115.GD2063@andy.tynant.ftech.net>
References: <mailman.1018972676.30534.python-list@python.org> <Xns91F35FF3186F5duncanrcpcouk@127.0.0.1> <20020417092115.GD2063@andy.tynant.ftech.net>
Message-ID: <20020417162407.GB17361@devel.trillke>

On Wed, Apr 17, 2002 at 10:21:15AM +0100, Andy Gimblett wrote:
> We're veering slightly from the original reason for this point, which
> was how to write stub functions/methods which will be implemented
> later.  The above certainly doesn't indicate any intent to actually
> implement the methods (in fact, on the contrary, it suggests they
> could be removed). 

Ok. this is the first reason i actually hear for leaving pass out.
(are there other ones?).

when i needed not-implemented classes i used this wrapper:

class NotImplemented:
   def __getattr__(self,name): return lambda *x,**y: None # or raise error

you can use it like this:

class HalfImplemented(NotImplemented):
	def implementedmethod1(self, ...): ...
	...

and the method calls that don't exist silently return None/1
or raise an error. that was a nice solution for me.

Really, i am still interested: what are (other) interesting
(convenience) reasons for making 'pass' optional?

	holger




From cliechti at gmx.net  Wed Apr  3 15:37:53 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Apr 2002 21:37:53 +0100
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it> <3CAA5777.66E7C2D5@python.org> <mailman.1017799668.13959.python-list@python.org> <7xy9g54h5q.fsf@ruckus.brouhaha.com> <3CAB0015.6563C468@python.org>
Message-ID: <Xns91E5DC39566ADcliechtigmxnet@62.2.16.82>

Guido van Rossum <guido at python.org> wrote in 
news:3CAB0015.6563C468 at python.org:

> Someone wrote:
> 
>> It's a cost/benefit thing.  Type/class unification may have cost more
>> in complexity than bools cost, but it has high benefits.  The bool
>> thing would a complicated change with limited benefit, so it's provoked
>> some sharp reactions.
> 
> That's an interesting response, because it highlights what I think is
> the main misunderstanding about the PEP.  You should think of it a
> very *small* change!
> 
> Some operations that conceptually return a truth value (like a==b)
> have to pick something to represent that result.  This is only needed
> when no other object among the operation's arguments is suitable to
> represent the result, for example in "not x".  While almost any value
> is usable to express a truth value, traditionally the Python core and
> standard library have consistently picked 0 to represent falsehood
> and 1 for truth.
> 
> A problem with this is that the intention isn't always clear.  Does a
> particular 1 represents a specific count or is it simply a
> representative of truth?  While ultimately this is a matter of
> documentation, having a notation helps clarify the matter (in cases
> where it is used).

i see that as a very good reason for having a bool type

> To solve the problem of clarifying intent, the PEP introduces two new
> values that the core will use in preference over 0 and 1, and standard
> names for them.  It makes them the only values of a new type, so that
> they can be recognized as the canonical representatives when
> necessary.  For backwards compatibility, these values act just like
> the integers 0 and 1 in non-Boolean contexts (except when printed or
> converted to a string).  For simplicity of implementation, and to
> guarantee backwards compatibility at the C level, the new type is
> implemented as a subtype of the built-in int type.

quote from the PEP
"""
    It has been suggested that, in order to satisfy user expectations,
    for every x that is considered true in a Boolean context, the
    expression x == True should be true, and likewise if x is
    considered false, x == False should be true.  This is of course
    impossible; it would mean that for example 6 == True and 7 ==
    True, from which one could infer 6 == 7.  Similarly, [] == False
    == None would be true, and one could infer [] == None, which is
    not the case.  I'm not sure where this suggestion came from; it
    was made several times during the first review period.  For truth
    testing, one should use "if", as in "if x: print 'Yes'", not
    comparison to a truth value; "if x == True: print 'Yes'" is not
    only wrong, it is also strangely redundant.
"""

here i disagree.

6 == True and 7 == True does _not_ infer 6 == 7. when using an 'if' nobody 
thinks that way neither:
>>>if 6: print "True"
>>>if 7: print "True"
both print True but nobody would think they're equal.

i think the concerns above come from the background of _knowing_ that bool 
is a subclass of int. but thats an only implementation detail. when you 
teach that True and False are the prefered values for Something/Nothing in 
a boolean context that is no problem. that their value is 1 resp. 0 in a 
numerical context is the most useful choice regardles of any inheritance 
from int.

i would rather tolerate a bit more broken code if only '6 == True' -> True 
beacuse i think that feels better and is better to understand for newbees 
and newcommers from other laguages.
its more tolerable to have a redundant way than having to explain the 
reasons why 'True == 1' but not 'True == 6' but both 1 and 6 are True when 
used with 'bool(x)' or 'if x:'.

> No user code needs to be changed unless it is sensitive to the string
> representation of Python objects.  There is no prejudice against user
> code that continues to use 0 and 1 (or something else) to represent
> truth values.

but there should be a clear recomendation. when it especialy useful to have 
a bool type for RPC etc. anyone should get used to it because he might use 
his code later in an RPC environment.

> In order to give people coming from (or going to) other languages an
> easy time getting used to the new type, it is called bool, and its
> values are called False and True.  But like so many things in
> Python[*], its bool type acts slightly different than the Boolean type
> in other languages.  That's fine -- other languages don't all agree on
> the details of Boolean types anyway, varying from very strict (Pascal,
> Java) to very loose (C, Lisp).

i like the name 'bool' its a clear name where you can catch the intended 
use of the type and you learn/use someting that will come up again an again 
in CS. and short is good as it saves typing but is still very readable.

> This is long enough of a response.  To those people who complain that
> I don't take their long thoughts seriously, I can only respond that I
> thought long and hard about this PEP too, and am similarly disappointed
> that it faces so much opposition based upon what *might* happen.

don't see that too negative. some posters may fear that their voice is not 
heard and write big words. it's also a rare oportunity that an aspect of 
the language is discussed in the open newsgroup/list in that way, so 
everyone wants to have said someting. the discussions about the PEPs are 
not that immediate/close to the real Python as when you ask "should this be 
accepted".
i guess when you apply a "median" filter on all the answers and weight if a 
reason is more "fear from the unknown" or more fundamental you will get 
valuable information from the discussion.

i still think that offering a PEP for discussion is a valuable thing and i 
appreciate it. i hope we will see it in again in the future.
it makes many of us think of "our" Python instead of just "that Python" (it 
brings in some emotion, an hence the hot discussion).

chris

PS: 
you should also note that the PEP has been updated and that many of the 
open points crystallize to clear statements.

> --Guido van Rossum (home page: http://www.python.org/~guido/)
> 
> [*] Take variables for example.  You can write "x = 1" in Python just
> as in C or Perl, and superficially it does the same thing.  But the
> meaning is profoundly different!  1 is an object, and x is a name
> binding for that object.  This goes on in Python all the time: other
> examples are x+y, for loops, even function argument default values.
> Booleans have their own twists and always had ("if x", "x and y").
> Python is not unique here either -- many language have different
> semantics underneath familiar notations, and this can represent a real
> breakthrough (like Smalltalk, which redefined x+y as sending the
> message "+y" to the object x).
> 



-- 
Chris <cliechti at gmx.net>



From zeitlin at seth.lpthe.jussieu.fr  Tue Apr 23 12:51:29 2002
From: zeitlin at seth.lpthe.jussieu.fr (Vadim Zeitlin)
Date: Tue, 23 Apr 2002 16:51:29 +0000 (UTC)
Subject: wxPython: wxDC problem
References: <9d8d2e93.0204230721.7ae6d8a0@posting.google.com>
Message-ID: <slrnacb48h.cp6.zeitlin@seth.lpthe.jussieu.fr>

On 23 Apr 2002 08:21:29 -0700, Carole Valentin <carole at autosim.no> wrote:
> I have a road map drawn in a wxDC and some cars in the same wxDC.

 I assume that this is a wxPaintDC or wxClientDC, i.e. its contents is drawn
on some window, isn't it?

> When the cars are moving, I would like to redraw only them and not the
> entire map.

 Simple Refresh() only the region of the window encompassing the old and new
car positions (alternatively you could call Refresh() twice for the old and
the new location but I'm not sure it is really more efficient). And, of
course, in your redrawing logic you should only update the objects
intersecting the update region of the window.

> Does the wxDC::SetClippingRegion method create an independant wxDC?

 No, it doesn't, but this is besides the point: wxWindows clips any redrawing
you do to the update region anyhow, so you just have to avoid refreshing too
much. Redrawing the unaffected objects doesn't hurt as much (i.e. at least it
wouldn't result in horrible flicker unlike refreshing the entire window each
time the car moves) but it's a good to avoid doing it nevertheless.

 Regards,
VZ
-- 
GCS/GM d? H+ s++:-- p2 au--- a- w+ v C+++ UBLS+++ P- L++ N++ E--- W++++ M? V--
-po+ R++ G`` !tv b+++ D--- e++++ u++ h--- f+ r++ n- y?


From aleax at aleax.it  Tue Apr  9 04:48:36 2002
From: aleax at aleax.it (Alex Martelli)
Date: Tue, 09 Apr 2002 08:48:36 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8sn8u$sj7$0@216.39.172.122> <Eeos8.74753$S52.2651788@news2.tin.it> <a8u9dv$lfo$0@216.39.172.122>
Message-ID: <EVxs8.75520$S52.2683315@news2.tin.it>

Bengt Richter wrote:
        ...
>>implementation.  One can of course do much better than that:
>>
>>        [x for x in (y,)][0]
>>
> Yes, I thought of that one too, while I was out ;-)
> 
> I think that one is actually _more_ "horrid" ;-) Why? Because it depends
> on what IMO should _not_ work and my example depends on what reasonably
> IMO _should_ work (if optimization issues did not take practical
> precedence). IMO letting a binding side effect escape the scope of a list
> comprehension is a flaw (and I'm guessing that it will be fixed). In

Sorry [not really:-)] but your O (and mine) does not matter as much as 
GvR's and the Language Reference Manual.

> contrast, locals() returns a purported dictionary of local bindings, and
> IMO it's reasonable to expect to be able to modify them through dictionary
> operations, unless warned off conspicously.

Ditto: the warning in the docs IS quite conspicuous.


> Or perhaps I should say was: I saw another post from Aahz, which said,
> """
>     I define "assignment" in Python as being precisely equal to "=".
> """
> So that changes the ball game and clinches it. He's perfectly right,

He's wrong only in his choice of pronoun -- it's not _him_ defining
what terms means in Python, but the Language Reference manual.

I always prefer to mention binding and rebinding (the EFFECT of
what we're doing), but when I mention the _tool_ often used to
perform such effect I call it assignment when it is, only.


Alex



From bogus@does.not.exist.com  Sun Apr 21 20:43:56 2002
From: bogus@does.not.exist.com (James Chapman)
Date: Mon, 22 Apr 2002 00:43:56 GMT
Subject: newbie dictionary question
References: <TiHw8.99936$CH1.86776@sccrnsc02> <3CC34519.E6C1063A@gmx.de>
Message-ID: <g1Jw8.89176$G72.64386@sccrnsc01>

Hi,

I'd preferably like to use something like this but my problem is precisely
this. I have for example to dictionaries as follows:

a= {4.95:100,4.97:90,4.99:95}
b={4.95:100,4.96:99}

and I would wonder if there is an elegent way to do something like:

c = a+b
c={4.95:200,4.96:99,4.97:90,4.99:95}

thanks,

"Wolfgang Grafen" <wolfgang.grafen at gmx.de> wrote in message
news:3CC34519.E6C1063A at gmx.de...
> "James Chapman <" schrieb:

> >>> d1={'a':1,'b':2}
> >>> d2={'c':3,'b':4}
> >>> d1.update(d2)
> >>> d1
> {'b': 4, 'c': 3, 'a': 1}
>
> Hope this is what you mean but consider that the value of 'b' in d1 will
be
> overridden by the value of 'b' in d2.
>
> wolfgang
>




From fperez528 at yahoo.com  Mon Apr 29 22:15:28 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 29 Apr 2002 20:15:28 -0600
Subject: Newbie question: how to join a list of elements of user-defined types?
References: <ef08387c.0204291754.1224fd7b@posting.google.com>
Message-ID: <aakutg$5gv$1@peabody.colorado.edu>

Fortepianissimo wrote:

> Basically I want to have a subclass of list, like
> 
> class MyList (list):
>     ...
> 
> then I want to
> 
> l=MyList()
> ... do stuff to fill l
> s=" ".join(l)
> 
> with the hope that s will be a nicely joined string with " " as the
> delimiter. Given my C++ background, I tried to define __str__/__repr__
> for both MyList and the user-defined types of the list members,
> assuming the type conversion would silently be invoked. But I was
> wrong!
> 
> Even trying this failed:
> 
> l=[1,2,3]
> s=" ".join(l)
> 
> and I thought built-in types all have default __str__ and __repr__?
> 
> What am I missing here? Many thanks to someone who would take time to
> rescue a poor soul.

Well, this quick solution seems to work. It's probably an ugly/inefficient 
one,  so wait for the gurus for a deep/elegant one ;) But in the meantime 
this may get you going:

In [25]: class mystr(str):
   ....:   def join(self,alist):
   ....:     template = ('%s'+self.__str__()) * (len(alist)-1) + '%s'
   ....:     return template % tuple(alist)
   ....:

In [26]: mystr(' ').join(l)
Out[26]: '1 2 3'

In [27]: l
Out[27]: [1, 2, 3]

cheers,

f


From tim.one at comcast.net  Sun Apr 14 18:59:19 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 14 Apr 2002 18:59:19 -0400
Subject: Do you QA your Python? Was: 2.1 vs. 2.2
In-Reply-To: <slrnabjj1b.741.philh@comuno.freeserve.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCOECFOOAA.tim.one@comcast.net>

[phil hunt]
> It is not new features that are the problem, it is new features that
> break existing code.

According to some people.  Other people have said new features are a problem
even when they're 100% backward compatible.

> IMO until Python developers adopt a policy of being very reluctant
> to do this,

Guido is very reluctant to break code, but it's not an absolute for him:
among other things, it competes with the desire to fix design mistakes in
the language, which is unusual for a language designer (not to realize that
some ideas were mistakes, but to have enough guts to repair them).  Guido is
also much more sympathetic to not breaking code that was in fact forbidden
by the docs than any C or C++ vendor would be.  The __future__ + warning
mechanism also gives explicit and controllable warnings about incompatible
changes for at least one release cycle before they happen.  Heck, I've spent
more time writing replies in this thread today than I spent adjusting to
Python changes in my own code since Python 0.9.6 was released.

> many IT managers will be wary of using it in anything other than toy
> projects,

Good for them <wink>.  Perhaps their nimbler colleagues will leave them in
the dust; perhaps not.

> and Python's user base will not grow as quickly as (IMO) it should.

We have no way to measure it.

> ...
>    New features => good
>
>    New features that break existing code => bad

New features => many new expenses and forward-compatibility headaches

Repairing flawed semantics => lower costs over time, breakage or not

are also valid points.  Good versus Evil only exists on Usenet <0.1 wink>.





From ophirb at hotmail.com  Mon Apr 15 16:11:43 2002
From: ophirb at hotmail.com (ophir)
Date: 15 Apr 2002 13:11:43 -0700
Subject: installation probs on SGI: _socket package removed.
Message-ID: <6a3d3529.0204151211.7f38f354@posting.google.com>

When building python on SGI IRIX646 I've notice that although the
_socket library compiled and linked ok, the 'make' process failed to
import _socket and removed it.
Following the advice I found here at the newsgroup I attempted to
manually compile and link _socket.so, then tried importing it from
python.
The results: An exception was thrown saying the symbol __eprintf was
not defined in the shared object.
Has anyone encountered this problem and have any help on this?
What library is the symbol __eprintf defined in?


From sholden at holdenweb.com  Mon Apr 22 15:28:55 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 22 Apr 2002 15:28:55 -0400
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com> <20020412101611.GB15466@andy.tynant.ftech.net> <20020412131633.GA17830@andy.tynant.ftech.net> <mailman.1018636586.2076.python-list@python.org> <1103_1019500994@news1.c2i.net>
Message-ID: <8yZw8.78960$T%5.6143@atlpnn01.usenetserver.com>

"Audun Tonnesen" <aut at c2i.net> wrote in message
news:1103_1019500994 at news1.c2i.net...
> I have problems with retrbinary. The call is like this:
> ftpobject.retrbinary("RETR <filename>, file.write(someData),[blocksize])
>
> My problem is the <someData> argument to file.write(...). How do I define
or bind the name to the actual datachunk?
>

Your code should read something like what follows:

import ftplib
myfile = open("theoutputfile.bin", "wb")

def writer(x):
    """This function is called with  x containing data each time
    a block of data becomes available from the ftplib.FTP object."""
    myfile.write(x)

ftpobject = ftplib.FTP( ... whatever ...)

ftpobject.retrbinary("RETR myfile.bin", writer)

myfile.close()
ftpobject.quit()

In other words, you pass a *function* to retrbinary(), and retrbinary()
calls the function with a single argument each time it has data to be
handled. That's why the documentation refers to it as a "callback".

See the book URL below for more reading on this ;-)

regards
 Steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/






From db3l at fitlinxx.com  Tue Apr  2 22:56:24 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 02 Apr 2002 22:56:24 -0500
Subject: The Python Way
References: <mailman.1017783076.29463.python-list@python.org>
Message-ID: <uhemtzb87.fsf@ctwd0143.fitlinxx.com>

Tim Peters <tim.one at comcast.net> writes:

> [Andy Gimblett]
> > So, is Aahz/Tim saying that Queue.Queue is "better" in general than
> > threading.RLock ?  If so, why?
> 
> It's almost impossible to get yourself in trouble using Queue.  It's almost
> (but not quite as) impossible *not* to get yourself in trouble rolling your
> own protocols by hand via raw locks. (...)

And heck, in most cases with your own locking, you're bound to wrap it
up in some sort of container class to centralize your management of
the lock and its resource, and waddaya know - that's pretty much what
Queue is doing for you!

Assuming of course that it's for synchronizing data movement, I'm
always more than happy enough to delegate responsibility for it to
Queue, as a polished class known to be solid.

I'll still use locks as mutexes when I'm really just controlling
access to shared state that doesn't actually flow or fit a
producer/consumer concept, but that's about it.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From peter at localhost.i-did-not-set--mail-host-address--so-shoot-me  Thu Apr 18 00:37:05 2002
From: peter at localhost.i-did-not-set--mail-host-address--so-shoot-me (Peter Seibel)
Date: Thu, 18 Apr 2002 04:37:05 GMT
Subject: ranlib problem building 2.2.1 on Mac OS X
Message-ID: <m28z7l7hc4.fsf@javamonkey.com>

I'm trying to build Python 2.2.1 from the source tarball on Mac OS X.
I ran ./configure and then make and the make dies with:

ranlib libpython2.2.a
ranlib: warning: cannot infer operation mode from `libpython2.2.a'
ranlib: you must specify a MODE
Try `ranlib --help' for more information.
make: *** [libpython2.2.a] Error 1

while trying to build this target:

# Build static library
# avoid long command lines, same as LIBRARY_OBJS
$(LIBRARY): $(LIBRARY_OBJS)
	-rm -f $@
	$(AR) cr $@ Modules/getbuildinfo.o
	$(AR) cr $@ $(PARSER_OBJS)
	$(AR) cr $@ $(OBJECT_OBJS)
	$(AR) cr $@ $(PYTHON_OBJS)
	$(AR) cr $@ $(MODULE_OBJS) $(SIGNAL_OBJS)
	$(AR) cr $@ $(MODOBJS)
	$(RANLIB) $@

RANLIB is defined as:

RANLIB=		ranlib

Any ideas?

-Peter


From skip at pobox.com  Thu Apr  4 22:43:00 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 4 Apr 2002 21:43:00 -0600
Subject: "Don't Kill Opera" documentation release
In-Reply-To: <a8j443$omncm$1@fido.engr.sgi.com>
References: <mailman.1017709803.3793.python-list@python.org>
        <a8j443$omncm$1@fido.engr.sgi.com>
Message-ID: <15533.7492.550606.250639@12-248-41-177.client.attbi.com>

    Paul> Could you explain easily what you had to change to the Docs in
    Paul> order to satisfy Opera 6.01?  As an Opera user, I am curious what
    Paul> was the bug in Opera that you were working around (or at least I
    Paul> assume that's the situation).

There was a bug in the Content-Type meta tag.  I don't remember exactly what
the construct was, but I believe it looked something like

    <meta http-equiv="Content-Type" content="text/html" charset=>

instead of

    <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">

It had no effect on (at least) the Linux version of Opera, but gave
Opera/Win users fits.  Not sure about Mac.

    Paul> Oh - Opera support informed me in an email today that the
    Paul> next Beta should be out any day.

For what platform?

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)




From sholden at holdenweb.com  Tue Apr 23 07:52:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 23 Apr 2002 07:52:38 -0400
Subject: Import conundrum
References: <b9039be.0204230038.3132cf96@posting.google.com>
Message-ID: <rYbx8.93075$T%5.49825@atlpnn01.usenetserver.com>

"Cliff" <pythonista at uk2.net> wrote ...
> If I put the following Python 2.1 script in a file called 'xml.py':
>
>    from xml.dom.minidom import parseString
>    dom = parseString('<?xml version="1.0" ?><test/>')
>    print dom.writexml(sys.stdout)
>
> and then try to run it on Windows NT I get the following:
>
>    Traceback (most recent call last):
>      File "C:\home\temp\xml.py", line 5, in ?
>        from xml.dom.minidom import parseString
>      File "C:\home\temp\xml.py", line 5, in ?
>        from xml.dom.minidom import parseString
>    ImportError: No module named dom.minidom
>
Its all to do with the import mechanism, and your confusion is
understandable when the documentation says """Details of the module
searching and loading process are implementation and platform specific. It
generally involves searching for a ``built-in'' module with the given name
and then searching a list of locations given as sys.path."""

> If I rename the file to something else, then everything is as expected.
> What I am interested to know is whether this is 'correct' behaviour, or am
> I doing something really stupid?
>
Well, it is quite normal for the import statement to look in the current
directory. If there's a file called xml.py, how is the interprpeter to know
that you don't want to use it as the basis of your import? I suspect you're
going to say "because xml is a module, not a package". Makes sense to me.

> I ask, because I would personally expect the Python interpreter to know
the
> difference between the file called 'xml.py' and the library path 'xml.dom'
> (which I would expect to be mapped to .\xml\dom on Windows).
>
When you get into the differences between modules and packages, which the
documentation is even more explicitly murky on: """The sequence of
identifiers up to the last dot is used to find a ``package'' ; the final
identifier is then searched inside the package. A package is generally a
subdirectory of a directory on sys.path that has a file __init__.py. [XXX
Can't be bothered to spell this out right now; see the URL
http://www.python.org/doc/essays/packages.html for more details, also about
how the module search works from inside a package.] """

> And there I was thinking I understood this Python stuff... ;-)
>
In this particular case you can be forgiven for not understanding it -- or,
at least, I have forgiven myself <wink> and feel you should do the same. It
doesn't help that the essay referenced in the previous paragraph is
currently offline, but that's just a net glitch.

If you've ever read any of the code that messes with import mechanisms, it's
funky in the extreme. The only reason we can't say whether it's broken is
because it's impossible to tell what "broken" means!

saying-'ni'-doesn't-always-get-you-there-ly y'rs  - steve
--

home: http://www.holdenweb.com/    book: http://pydish.holdenweb.com/pwp/







From pyth at devel.trillke.net  Tue Apr 30 15:47:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 30 Apr 2002 21:47:15 +0200
Subject: Update locals() / resolved
In-Reply-To: <20020430210938.B3173@idi.ntnu.no>
References: <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no> <mailman.1020029269.29740.python-list@python.org> <fjbz8.7570$M63.5673@atlpnn01.usenetserver.com> <slrnactp07.pv.mlh@vier.idi.ntnu.no> <20020430190553.GE16217@devel.trillke> <20020430210938.B3173@idi.ntnu.no>
Message-ID: <20020430194715.GG16217@devel.trillke>

On Tue, Apr 30, 2002 at 09:09:38PM +0200, Magnus Lie Hetland wrote:
> holger krekel <pyth at devel.trillke.net>:
> [snip]
> > > Thes names aren't reserved in any way...
> > 
> > right. i do it all the time for file,str,dict,string,list...
> > they are simply to short and descriptive to be left bound
> > to something i don't access directly.
> 
> Sure. I wasn't arguing with you here (believe it or not ;)
>
>
Sure. I wasn't arguing with you here (believe it or not ;)

    holger




From mixo at beth.uniforum.org.za  Mon Apr  8 03:52:32 2002
From: mixo at beth.uniforum.org.za (mixo)
Date: Mon, 08 Apr 2002 09:52:32 +0200
Subject: executing an external program
Message-ID: <3CB14C40.402@beth.uniforum.org.za>

I am trying to execute a program several times, each time
with different parameters. The program returns 0 if it
was successfully executed, and some other number if it
wasn't. The problem I have is that I can only run one instance
of the program, and I still need my script to continue running
even though the external program is still running. How can I
continue execution of my script, and run only one instance of
the external program and keep track of what is doing so as to
avoid confusion?



From frr at easyjob.net  Tue Apr 23 10:33:10 2002
From: frr at easyjob.net (Fernando Rodríguez)
Date: Tue, 23 Apr 2002 16:33:10 +0200
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com> <tXdx8.94618$T%5.47937@atlpnn01.usenetserver.com>
Message-ID: <a3sacukia5u7qptgaq98qq3nn40dcbk9hs@4ax.com>

On Tue, 23 Apr 2002 10:08:04 -0400, "Steve Holden" <sholden at holdenweb.com>
wrote:


>I'm confused about why you don't get a proper MySQLdb installation from the
>same file I used, though. When I ran it, it installed a MySQLdb directory
>ujnder <python>/Lib/site-packages, and I have no problems importing MySQLdb.
>
>In your "Add/Remove programs" control panel item do you see "Python 2.2
>MySQL-python-0.9.1"? Maybe a de-install/reinstall would help. My own system
>has no reference at all to CompatMySQLdb that I can see, but I notice that
>when I open the installer .exe file in WinZip there it is. Did you perchance
>treat the installer as an archive rather than running it?

I removed it and reinstalled it, and now I do have a MySQL directory. :-)))  
I'll try again with MySQL.





-----------------------
Fernando Rodriguez


From billdozier at hotmail.com  Fri Apr 26 10:45:45 2002
From: billdozier at hotmail.com (Bill Dozier)
Date: 26 Apr 2002 07:45:45 -0700
Subject: Initializing a list with copies
References: <xduy9fa4z2x.fsf@leto.informatik.uni-freiburg.de>
Message-ID: <199334c5.0204260645.3af87dc5@posting.google.com>

Ralf Juengling <juenglin at informatik.uni-freiburg.de> wrote in message news:<xduy9fa4z2x.fsf at leto.informatik.uni-freiburg.de>...


> >>> n

>  2

> >>> l

>  [[], []]

> >>> l[0].append(1)

> >>> l

> [[1], [1]]


This looks like the old "mutable initializer" thingie.

> >>> l = [[] for i in range(n)]

> Can this be done more elegant? 
"[[] for i in range(n)]" looks fine to me.

> By some expression that was similar to '[[]]*n' ? 
Do you mean, "with few characters to type?" That's not the same as
"with more elegance."

Bill


From sjmachin at lexicon.net  Mon Apr 29 04:58:54 2002
From: sjmachin at lexicon.net (John Machin)
Date: 29 Apr 2002 01:58:54 -0700
Subject: Super-newbie syntax error question
References: <mailman.1020045350.28826.python-list@python.org>
Message-ID: <c76ff6fc.0204290058.1b4f7dcf@posting.google.com>

[Charlie Wood]

I have some code that I've cobbled together ...
...and here's the resulting error:
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "C:\PYTHON22\lib\rss2html.py", line 46
     f = file("/article.txt", "w")
     ^
 SyntaxError: invalid syntax

****** You may like to put code that _you_ have cobbled together
somewhere else, reserving c:\python22\lib\ for code that the Python
development team has cobbled together :-)

["Sean 'Shaleh' Perry"]
> Should be:
> 
> f = open() not file().

****** You may like to read "What's New In Python 2.2"


From phr-n2002a at nightsong.com  Fri Apr 26 19:21:57 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 26 Apr 2002 16:21:57 -0700
Subject: [OT] What is Open Source?
References: <mailman.1019858688.3528.python-list@python.org>
Message-ID: <7x1yd2qc4q.fsf@ruckus.brouhaha.com>

Tim Peters <tim.one at comcast.net> writes:
> > All the Python PEP's seem to contain a disclaimer about being placed
> > into the public domain.
> 
> Yup, and PEP 1 tells PEP authors to do this.
> 
> > Do you think Python has a problem because of this?
> 
> Not at all.  Perhaps it carries no legal force, perhaps it does:  IANAL, but
> Python couldn't care less either way.  The only people it might bother are
> the overly paranoid, fearing that if they reuse a PEP as if it were truly
> public domain, one of the authors may come after them brandishing a lawyer's
> opinion that the "public domain" declaration was without force.  Other
> countries would probably laugh that out of court, but you never know what a
> US court might decide.

So we'l have to start issuing PEP's with the disclaimer "I found this
PEP in my grandma's old trunk in the attic, dated April 1, 1920, so
it is out of copyright now."


From martin at v.loewis.de  Thu Apr 18 02:41:41 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 18 Apr 2002 08:41:41 +0200
Subject: C-extension not threadable?
References: <3cbc783d.39583390@news.easynews.net> <m3u1qbeah8.fsf@mira.informatik.hu-berlin.de> <m3g01ulbav.fsf@ferret.object-craft.com.au> <m3r8leyy1f.fsf@mira.informatik.hu-berlin.de> <m3g01um27o.fsf@ferret.object-craft.com.au> <m3ads2dkpy.fsf@mira.informatik.hu-berlin.de> <m38z7lrfh7.fsf@ferret.object-craft.com.au>
Message-ID: <m3u1q9y0d6.fsf@mira.informatik.hu-berlin.de>

Dave Cole <djc at object-craft.com.au> writes:

> I think that it is preferable to keep the exception and propagate it
> back from the callback through the API to the thread which called the
> API.  I am not sure why discarding the exception would be a good idea.

That assumes that the call originated from Python. If some library has
created the thread (e.g. in response to an incoming network connection
which was accepted by the library), there won't be anybody to
propagate the exception to.

> IMO, from the point of view of the extension module user it should not
> be visible that there was some non-Python code between the Python code
> where the exception was raised and the Python code which called the
> module API.

Again, assuming there is Python code on both sides, and also assuming
it is possible to pass something through from one end to the
other. *If* that is possible, you can just as well pass the thread
state through. If you cannot pass the thread-state through, due to the
design of the library you are wrapping, tough luck: chances are low
that you can pass the exception back.

Regards,
Martin


From mkc+dated+1021692856.d0f633 at mathdogs.com  Wed Apr 17 23:34:58 2002
From: mkc+dated+1021692856.d0f633 at mathdogs.com (Mike Coleman)
Date: 17 Apr 2002 22:34:58 -0500
Subject: Why don't strings share data in Python?
References: <mailman.1018937610.11880.python-list@python.org> <87n0w37z0n.fsf@mathdogs.com> <kuy9fmgbpf.fsf@lasipalatsi.fi>
Message-ID: <873cxt8ysd.fsf@mathdogs.com>

Erno Kuusela <erno-news at erno.iki.fi> writes:
> reminds me of the buffer object. you can have a buffer object
> representing 'oba' if you do buffer('foobar', 2, 3). i believe this
> does what you awnt memory management wise. the re module can work on
> buffer objects, maybe some other things can too.

Good idea.  They look a bit undocumented, but they do look like a real
possibility for this.  Thanks!


From opengeometry at NOSPAM.yahoo.ca  Fri Apr 26 15:47:21 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 26 Apr 2002 19:47:21 GMT
Subject: Simple bar graphs
References: <Xns91FCD43535458Latzikatz@139.174.2.56>
Message-ID: <aacas8$9d997$2@ID-99293.news.dfncis.de>

Lutz Schroeer <Lutz.Schroeer at kybernetik-manufaktur.de> wrote:
> I'm looking for a _pure_ Python module which generates some very simple bar 
> graphs and outputs them as gif or jpg. I looked through Parnassus and other 
> sources but could not get enough information to make a decision.

I used 'gdchart' long ago.  It generates GIF directly.  But, if your
requirement is simple enough, you can use HTML rules, say in table layout.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-cpu Cluster, Hosting, NAS, Linux, LaTeX, Python, vim, mutt, tin


From Ken.Peek at SpiritSongDesigns.comNOSPAM  Sun Apr 14 18:55:55 2002
From: Ken.Peek at SpiritSongDesigns.comNOSPAM (Ken Peek)
Date: Sun, 14 Apr 2002 15:55:55 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com> <m3zo07yxap.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3cba0944@news.mhogaming.com>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3zo07yxap.fsf at mira.informatik.hu-berlin.de:

| > >>> b = -a
| > >>> print hex(b >> 1)
| > -0x4000000000000000L
| >
| > # !!! SURPRISE !!!
| > # (I expected -0xC000000000000000L !!!)
|
| Can you explain why you expected this? How could Python
| possibly know how many bits you expected to be in the
| internal representation?
|
| Can you elaborate? You explained that the >>> fills in
| zeroes for 32-bit numbers that have the highest bit
| set. With long integers, you don't need that: If you
| have a number that has the 2**31 bit set, it may still
| be a positive number, and right-shifting will fill in
| the zero bit. So no need for a >>> operator.

OK--

As part of the "Grand Unification of the 'int' and the
'long int' types", the "new Python way" (as provided to
us by Guido,) of viewing 'long ints' is that if they are
a positive number, then the 'long int' should behave in
a manner as if there is an infinite number of zeros
above the highest bit in the number, and if the 'long
int' is negative, then the 'long int' should behave as
if there is an infinite number of ones above the highest
bit (i.e.: in so called "two's complement" fashion-- the
same way a signed 'regular int' behaves.)

Thus-- my "expectation" is that 0x8000000000000000 >> 1
is: 0x4000000000000000 (and that is the way Python now
works for positive 'long ints'.)

Given the above description of the way NEGATIVE 'long
ints' should behave, my 'expectation' is that:
-0x8000000000000000 >> 1 _SHOULD_ return:
-0xC000000000000000,

(_NOT_ -0x4000000000000000 as Python _NOW_ behaves.)

The way Python _NOW_ works, is to give a different
behavior for the '>>' operator depending on whether you
are working on an 'int', or a 'long int'.  This was OK
before, because the 2 object types were treated
differently by the language.  I thought we were trying
_NOW_ (versions >= 2.2.x) to "unify" the behavior of
'ints' and 'long ints'.

Am I wrong?

Now, if I am right about this, then we don't really
know if we are right shifting an 'int', or a 'long int'
and so it would be 'useful' to have a '>>>' operator in
this case, to guarantee that zeroes are shifted into
the highest bit.  JAVA has this facility, why must
Python _NOT_ have it?





From netzapper at magicstar.net  Tue Apr  2 11:56:12 2002
From: netzapper at magicstar.net (A. Jones)
Date: Tue, 02 Apr 2002 16:56:12 GMT
Subject: solution found; Thank you!
References: <3ca96e30.52835380@mammoth.usenet-access.com> <j4adsm2l57.fsf@informatik.hu-berlin.de>
Message-ID: <3ca9e26c.82595499@mammoth.usenet-access.com>

On 02 Apr 2002 17:08:36 +0200, loewis at informatik.hu-berlin.de (Martin
v. =?iso-8859-1?q?L=F6wis?=) wrote:

>netzapper at magicstar.net (A. Jones) writes:
>
>> Any insight would be much appreciated.
>
>Seeking and text mode does not really mix well. When you write to a
>text file, the C library may expand or collapse newline characters,
>and perform other modifications. So if you write n bytes to a text
>file, that doesn't mean that n bytes end on disk.
>
>I recommend to open the file in binary.
>
>Regards,
>Martin
>

Thank you, you put me on the right track.  file("QOD.qdb", 'ab') is
working perfectly.

Your help was much appreciated, Martin


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From ianb at colorstudy.com  Mon Apr 29 21:02:50 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 29 Apr 2002 20:02:50 -0500
Subject: Python Webpages
In-Reply-To: <aakelj$do6$1@maud.ifi.uio.no>
References: <aagr7h$5c9$1@maud.ifi.uio.no>
	<lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no>
	<aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net>
	<23891c90.0204290051.3edcc62d@posting.google.com>
	<slrnacq7jk.gav.jon+usenet@snowy.squish.net>
	<mailman.1020090053.19284.python-list@python.org>
	<slrnacqrps.sdd.jon+usenet@snowy.squish.net>
	<mailman.1020102500.7281.python-list@python.org>
	<slrnacra60.v3g.jon+usenet@snowy.squish.net>  <aakelj$do6$1@maud.ifi.uio.no>
Message-ID: <1020128570.456.189.camel@localhost>

On Mon, 2002-04-29 at 16:41, Johan Fredrik ?hman wrote:
> > I must admit I have never yet seen a system which integrates HTML and
> > code in the same file without the file ending up an unreadable mess.
> 
> I just thought of great idea:  How about using one file and not mix html and
> python code at the same time ?
> Example:
> ==================================================
> <html>
>     <body>
>     %content%
>     </body>
> </html>
> 
> <!--PYTHON Code
> from smartTemplates import *
> 
> page.content =  "Isn't this cool ;)"
> 
> -->
> ==================================================

It's not WYSIWYG compatible, but the easy way to do something like this
is:

"""
<html>
...
<!--
"""

(and then some python code which uses __doc__ and your template system)

#--></body></html>


The first """ is going to be lost by the WYSIWYG editor.  The rest can
be saved...

  Ian






From sheershion at mailexpire.com  Wed Apr 24 08:30:15 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Wed, 24 Apr 2002 12:30:15 -0000
Subject: Python-list digest, Vol 1 #10572 - 14 msgs
References: <mailman.1019522021.8009.python-list@python.org>
Message-ID: <Xns91FA10A2EED17rcamesz@amesz.demon.nl>

Delaney, Timothy wrote:

> I would rewrite this as:
> 
> def both_ways (val, direction):
> 
>     if dir:
>         arg1, arg2 = None, val
>     else:
>         arg1, arg2 = val, None
> 
>     another_func(arg1, arg2, ...)
> 
> The thing is, the ternary operator is only more readable when
> dealing with a single condition and single results. Adding even a
> single expression in there starts making it unreadable. Any time
> things get more complicated, you need to move to the if () {} else
> {} format anyway. 

I'd still say this is more explicit and readable:


def both_ways (val, direction):
    if dir:
        another_func(None, val, ...)
    else:
        another_func(val, None, ...)


Of course, if the argument list for 'another_func' is very long the 
readability argument might swing the other way.


Robert Amesz


From tim.one at comcast.net  Sun Apr  7 15:08:48 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 07 Apr 2002 15:08:48 -0400
Subject: Booleans, integer division,
 backwards compatibility; where is  Python  going?
In-Reply-To: <PjZr8.57418$pT1.1647471@news1.tin.it>
Message-ID: <LNBBLJKPBEHFEDALKOLCKELMOLAA.tim.one@comcast.net>

[Guido]
> (note that python-dev no longer has a closed-subscription policy).

[Alex Martelli]
> How so?  One visits:
>
> http://mail.python.org/mailman/listinfo/python-dev
>
> enters email and passwords, hits subscribe, and gets this message:
> ...

It's not closed in the sense that the list admins have been instructed to
approve all subscription requests unless they have a good reason not to.
The list admins are human, though, and sometimes they don't get to pending
subscription requests for a few days.

BTW, I'm not a list admin for Python-Dev, so can't approve your request (or
I would have, instead of bothering to type all this).

> ...
> If this is meant to be an open-subscription mailing list, I would
> suggest this arrangement be changed.

It's not meant to be open or closed, but somewhere in between, favoring the
open end.  A Python bool isn't adequate to model it <wink>.





From gh_pythonlist at gmx.de  Sat Apr 13 15:25:47 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 13 Apr 2002 21:25:47 +0200
Subject: Getting a Python library for cross-compiling with mingw32
In-Reply-To: <1018706814.23939.26.camel@progeny>
References: <1018706814.23939.26.camel@progeny>
Message-ID: <20020413192547.GA5477@lilith.my-fqdn.de>

* Alexandre Courbot <alexandrecourbot at linuxgames.com> [2002-04-13 16:06
* +0200]:
> Hello everybody,
> 
> I'm currently porting my Linux app (which heavily uses the Python
> library) to Windows, but I don't own Windows myself. So I'm
> cross-compiling the stuff from my Linux box using mingw32.

The main problem is not cross-compiling - I've done that successfully
from Linux x86 to Linux-StrongARM some day during the 2.2 beta days. The
main problem is that mingw32 isn't supported in Python's autoconf-based
build process, yet.

> [...] I then attempted to cross-compile Python itself. It looks like
> it's not intented for that :) After disabling configure tests that
> didn't like cross compilers, I had to twiddle even more to get a
> binary.  Finally, I managed to get a binary and a library, that even
> worked under Wine!

Wow! Cool. But you know that the resulting binary isn't necessarily
binary compatible with the Standard win32 Python distribution, do you?

> [...] Please note that I largely prefer having a version >= 2.2.
> Anyone knows of a place which provides suitable binaries, or any
> Python hacker could help me properly compiling my binary?

Next time I reboot to Windows, I can send you my gcc import library for
the standard win32 Python. It's easy to create with Cygwin, btw. I'd
have to check if this is suitable for cross-compiling from a *nix,
though.

> It looks like the source distribution wouldn't need big changes to be
> cross-compileable, which would be nice.

No, but since distutils is part of the Python build process, this got
messier. In fact, in my cross-compiling effort, I had to replace the
StrongARM ./python binary temporarily with a x86 ./python binary of the
same version to have distutils build its extension modules. For
cross-compiling, I'd suggest to add a switch --with-native-python= to
the configure script.

> I'd be glad to help making the source distribution cross-compileable,
> provided I can solve this issue and you guys think the feature would
> be usefull.

It sure would be very useful. One point: doesn't the configure script
use "uname" at some point? This would need to be worked around, too. And
one of these days I should resume my efforts with adding mingw32 as a
supported platform for Python's autoconf.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 10.1 ?C      Wind: 2.9 m/s




From phd at phd.pp.ru  Tue Apr  9 14:33:38 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 9 Apr 2002 22:33:38 +0400
Subject: Sending a signal to a process started by another user
In-Reply-To: <3CB31C2C.9010503@olen.to>; from joonas@olen.to on Tue, Apr 09, 2002 at 04:53:58PM +0000
References: <3CB31C2C.9010503@olen.to>
Message-ID: <20020409223338.I27632@phd.pp.ru>

On Tue, Apr 09, 2002 at 04:53:58PM +0000, Joonas Paalasmaa wrote:
> How can I send a signal to a process started by another user. In this 
> case my intention is to send a signal 14 to a process started by user jp 
> from a cgi script. The script raises an "Operation not permitted"- 
> OSError perhaps because the script is run by user "apache" and the 
> process by user "jp". How can In make this work properly?

   Using some form of setuid programs.

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




From pyth at devel.trillke.net  Fri Apr 19 17:01:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 19 Apr 2002 23:01:28 +0200
Subject: Trinary operator?
In-Reply-To: <a9pvp0$kqj$1@nets3.rz.RWTH-Aachen.DE>
References: <mailman.1019240861.31256.python-list@python.org> <a9pvp0$kqj$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <20020419210128.GK28515@devel.trillke>

On Fri, Apr 19, 2002 at 10:47:09PM +0200, Philipp Lenssen wrote:

> (huge code fragments deleted :-)

really? wouldn't these few simple lines do what you/one want/s?

genderdict = { 'english': { 'm':'male', 'f':'female'},
               'german' : { 'm':'m?nnlich','f': 'weiblich'} }

return genderdict.get(language,{}).get(gender,'UNKNOWN')

or something like this? But maybe i am missing something...

	holger




From max at alcyone.com  Tue Apr  2 17:40:31 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 02 Apr 2002 14:40:31 -0800
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk> <200204021340.g32Dekrq025410@ratthing-b246.strakt.com>  
	            <3CA9FD45.E6E33129@lemburg.com> <mailman.1017784039.388.python-list@python.org>
Message-ID: <3CAA335F.4A17F6FF@alcyone.com>

Guido van Rossum wrote:

> I don't see the introduction of the bool type as painful at all.
> 
> I find getting it accepted painful though. :-(

I think of it in very simple terms:  True and False are just
self-documenting values that indicate that you're talking about a
Boolean value rather than something else -- and enabling the ability to
write self-documenting code is always good.  For backward compatibility
concerns, bool must be implemented as (effectively) a subclass of int. 
Very little code will break with the introduction of the PEP.  That is
all.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From benpark22 at yahoo.com  Wed Apr  3 13:30:14 2002
From: benpark22 at yahoo.com (Ben Park)
Date: 3 Apr 2002 10:30:14 -0800
Subject: List files recursively: use recursive function or os.path.walk()
Message-ID: <4cc96c48.0204031030.53bfb69f@posting.google.com>

I want to list all files recursively in a a directory. I can either
use a recursive function all or use os.path.walk(). I tested both, and
found that os.path.walk() was actually slightly slower.

Here is the result:
==============
recursive call to list 24530 files,  dtime= 3.30
use os.path.walk() to list 24530 files,  dtime= 3.54

On a linux Pentium machine. Python version 2.1

conclusions:
* recursive call is slightly faster than os.path.walk() .
* code is shorter using os.path.walk() .


Here is the code:
===============
#!/usr/bin/env python
# vim: set syntax=python:

import os, time, string

print ' '

#=================
def listAllFile1(fileNames9,dir,files):
#=================
  def f1(a,dir=dir): return os.path.join(dir,a)
  files2 = map(f1, files)
  fileNames9.extend(files2)

#=================
def listAllFile(dir,fileNames9):
#=================
  files = os.listdir(dir)

  def f1(a,dir=dir): return os.path.join(dir,a)
  files2 = map(f1, files)
  fileNames9.extend(files2)

  def f1(a): return os.path.isdir(a) and not os.path.islink(a)
  dirs1 = filter(f1, files2)

  for dir1 in dirs1:
    listAllFile(dir1, fileNames9)


dir0 = '/home/benpark/tmp'
fileNames9 = []
t0=time.time()
listAllFile(dir0,fileNames9)
print 'recursive call to list %d files,  dtime=%5.2f'
%(len(fileNames9), time.time()-t0)

fileNames9 = []
t0=time.time()
os.path.walk(dir0,listAllFile1,fileNames9)
print 'use os.path.walk() to list %d files,  dtime=%5.2f'
%(len(fileNames9), time.time()-t0)


From tim.one at comcast.net  Sat Apr 13 19:16:08 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 13 Apr 2002 19:16:08 -0400
Subject: 2.1 vs. 2.2
In-Reply-To: <3cb8a755$0$12714$9b622d9e@news.freenet.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEPEONAA.tim.one@comcast.net>

[Jens Baader]
> ...
> Another question: When will we see an official ISO/ANSI standard
> for Python?

When enough people who feel as you do volunteer all the time and money it
takes to produce an ISO/ANSI standard.  If you haven't been involved in that
process, I can tell you that the effort it requires is staggering.  I know
of no plans to do anything of the sort for Python.

> I dislike working with nonstandardized languages.

If ISO/ANSI standardization is a requirement for you, forget Python.





From sampat at removemiel.mot.com  Tue Apr  2 03:44:26 2002
From: sampat at removemiel.mot.com (Kuntal Sampat)
Date: Tue, 2 Apr 2002 14:14:26 +0530
Subject: Cache problem with urllib2?
References: <a893o4$s51@nntp.cig.mot.com>
Message-ID: <a8bqv5$s45@nntp.cig.mot.com>

Apparently, calling urllib.urlcleanup() right before using urllib2.urlopen()
has solved the problem.

"Kuntal Sampat" <sampat at removemiel.mot.com> wrote in message
news:a893o4$s51 at nntp.cig.mot.com...
> Hi,
> I have a program that uses urllib2 to open a URL and then do further
> processing on it based on the content.
> The page I access is updated about once a week. My script is a cron job
> under UNIX. It wakes up regularly, checks for the update and sends me an
> email if the site has been updated. But my script fails to get the updated
> page on time. At one time it was over a day before my script actually saw
> the updated page.
> At another time, I have observed that if I do a "force reload" on the
> browser, then the script starts working fine (i.e. it accesses the new
> page).
>
> My question is this: Does urllib2 maintain a cache? If so, how can I clear
> it everytime before my script runs.
>
> Much thanks.
> Kuntal Sampat.
>
>




From martin at v.loewis.de  Mon Apr  1 10:30:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 01 Apr 2002 17:30:23 +0200
Subject: PEP 285: Adding a bool type
In-Reply-To: <20020401121204.A9299@ibook.distro.conectiva>
References: <mailman.1017486169.15848.python-list@python.org>
	<3CA5E2EA.1CADAF70@engcorp.com>
	<m3y9ga7z38.fsf@mira.informatik.hu-berlin.de>
	<20020401121204.A9299@ibook.distro.conectiva>
Message-ID: <m3k7rr77xs.fsf@mira.informatik.hu-berlin.de>

Gustavo Niemeyer <niemeyer at conectiva.com> writes:

> I hope nobody is planning to change "return None" to "return False".
> This would break a lot of code doing "ret == None".

It depends on the function, I guess. If the function conceptually
returns a bool, the return value should not be None.

However, I haven't found a case where a function returns None, yet
conceptually returns a bool - atleast not in the standard library.

Regards,
Martin




From martin at v.loewis.de  Mon Apr  1 18:07:27 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 02 Apr 2002 01:07:27 +0200
Subject: Arabic bidirectionality in Tkinter
References: <a8alrr$sa$1@netnews.upenn.edu>
Message-ID: <m38z87xbkg.fsf@mira.informatik.hu-berlin.de>

zayats at blue.seas.upenn.edu (Salim Zayat) writes:

> Qt deals with this bidirectionality automatically, but Tkinter doesnt 
> appear to.  What can I do?

It appears that there are a few BIDI widgets for Tk out there, for
example

http://imagic.weizmann.ac.il/~dov/taved.html
http://www.technion.ac.il/~tzafrir/R2L/

Both are for Hebrew; I don't know how useful they are for Arabic.

If they won't work, your options are obvious:

- Fix Tk to implement BIDI support in the widgets (perhaps using
  FriBidi).
- use another widget kit.

Regards,
Martin



From jhalttun at pp.htv.fi  Tue Apr 23 10:26:07 2002
From: jhalttun at pp.htv.fi (janne halttunen)
Date: Tue, 23 Apr 2002 17:26:07 +0300
Subject: curses mouse-events in linux console
References: <3CC3E067.6080107@pp.htv.fi> <lk662jrjav.fsf@pc150.maths.bris.ac.uk> <3CC54204.5010100@pp.htv.fi> <lkofgapqyd.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3CC56EFF.8000001@pp.htv.fi>

Michael Hudson wrote:
> janne halttunen <jhalttun at pp.htv.fi> writes:
> 
> 
>>Michael Hudson wrote:
>>
>>>janne halttunen <jhalttun at pp.htv.fi> writes:
>>>
>>>
>>>>I'm on Mandrake Linux 8.2 with Python 2.2, and I have a problem of
>>>>mouse-events not emitting properly in curses when on console with
>>>>gpm. In X it works fine, the getch-method returning instantly on
>>>>mouse-clicks, but on console it doesn't return until after pressing
>>>>key on the keyboard.
>>>
>>>Umm, does any curses application respond to the mouse on the console?
>>>I thought it was an xterm-and-derivatives only thing, but I could well
>>>be wrong.
>>
>>Well, it was working in python 2.0 or 2.1 with previous Mandrake
>>Linuxes 8.1 or 8.0..  And it's working now, except for the triggering
>>of the event.
> 
> 
> Oh.  Can't really help you I'm afraid...

Mm, I guess most console-people really prefer keyboard over mouse 
anyway. :)

It would be nice to have as an option, though.

janne



From mats at laplaza.org  Mon Apr  8 12:18:55 2002
From: mats at laplaza.org (Mats Wichmann)
Date: Mon, 08 Apr 2002 16:18:55 GMT
Subject: sharp sl5000
References: <slrnaaqpvc.2vl.peter@vlad.seahaze> <Xns91E77A4404E25PaCmAnRDLM@161.48.128.20>
Message-ID: <3cb1c295.11651323@news.laplaza.org>

On 5 Apr 2002 09:58:34 GMT, Fred Pacquier <fredp at mygale.org.nospam>
wrote:

:Peter Hayes <peter at seahaze.demon.co.uk> said :
:
:> Does anyone know if there is a version of Python that runs on the new
:> Linux based Sharp SL5000 PDA?
:
:Yes, it was ported during the beta phase last year. I remember seeing the 
:announcement on the developer's list when it was on Yahoo Groups, but it 
:has moved and I don't know where it is hosted now.

SourceForge.

And as to the availability, Sharp has replied to this question with
some pablum about how each individual sales region (usually ==
country) has to commit to enough volume from the factory in order to
make the creature available.  That's happened in the US, and
UK/Ireland for the early-access model (5000) which is now no longer
available (I have no info on whether there's going to continue to be a
Developer box available), and for the released model, the 5500,
Germany has added.  And yes the whole business of local configuration
and packaging rules lies behind this.  You can't even buy one in
Canada yet.

I've been lurking on a lot of the forums recently and there are some
real problems with the Zuarus as a consumer-grade device. Programmers,
who don't mind rough edges generally, love it.  I have some worries
about its' survival in the marketplace.
Mats Wichmann



From guido at python.org  Mon Apr  1 16:21:39 2002
From: guido at python.org (Guido van Rossum)
Date: Mon, 01 Apr 2002 16:21:39 -0500
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
In-Reply-To: Your message of "Mon, 01 Apr 2002 09:14:59 +0200."
             <E16rw1r-00016o-00@tswings.swing.cluster> 
References: <E16rw1r-00016o-00@tswings.swing.cluster> 
Message-ID: <200204012121.g31LLdk00743@pcp742651pcs.reston01.va.comcast.net>

> Nice tease. But Guido's hammer with regard to future strictification
> of bool is exactly what makes me uneasy.

Can you elaborate that?  I thought I made it pretty clear I didn't
want to strictify bool (though I did open up the question as to
whether other people want that).  Is that what you call a hammer?

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



From distrex at hotmail.com  Sun Apr  7 23:20:50 2002
From: distrex at hotmail.com (distrex)
Date: 7 Apr 2002 20:20:50 -0700
Subject: modules
References: <d9ea2696.0204061049.551bd53b@posting.google.com> <a8q9bi$2vka$1@news.idiom.com>
Message-ID: <d9ea2696.0204071920.1b9d75c4@posting.google.com>

jimd at vega.starshine.org (Jim Dennis) wrote in message news:<a8q9bi$2vka$1 at news.idiom.com>...
> In article <d9ea2696.0204061049.551bd53b at posting.google.com>, distrex wrote:
> 
> >I understand most of python. But I still don't get modules. I know
> >they are just basicaly a function but i don't know how to use them
> >effectively and what each one does. Everything good in python you will
> >at some point need to use a module right. If anyone can help me with
> >this i'd be greatful. Info or websites would be nice.
> 
>  Modules are things that you import into your namespace.  The
>  process of import them (or import *from* them) executes any 
>  code in them.  Usually (almost always) a module will only contain
>  class and function definitions, perhaps some "members" (variables)
>  and an "if __name__ == '__main__':" suite.
> 
>  You can think of a module as a library of classes, functions and
>  variables.  As such, it would be inappropriate to create modules 
>  that executed "visible" statements (such as printing stuff) or made 
>  other modifications to the importer's state.  Basically modules 
>  should normally just execute whatever is necessary to define their
>  namespace --- which usually requires NO "execution" (in the traditional
>  sense) at all.  (The confusion is that python is a dynamic language,
>  so things like:
>  
>  	def foo(): pass
> 	# and
> 	class Bar: pass
> 
>  ... are actually executable statements in Python, where they are 
>  "definitions" or "declarations" in other languages.
> 
>  The most important thing to realize about Python modules is that they
>  are singletons.  They are only executed upon initial import for any given 
>  interpreter session.   Additional imports are not an error, and won't 
>  raise an exception, but they will not re-execute the module's code.
> 
>  That can be confusing because subsequent import statements can 
>  change change namespaces around.  I can do something like:
> 
>  	import bisect as bifoo
> 	from bisect import insort
> 	import bisect 
> 
>  ... and I've only execute the contents of bisect once, but I've
>  made subsequent changes to my namespace (making the bisect.insort()
>  available directly available without any qualifiers (in my globals()
>  dictionary). and making it accessible via the bifoo and bisect
>  namespaces (which is silly, and used just for an example).
> 
>  You might be aware that there are three sorts of modules in
>  Python.  There are binaries (compiled code, usually from C sources)
>  which appear as .so (shared object) files under Linux and UNIX, and
>  .DLL files under MS Windows (don't know for Macs).  There are Python
>  files (simple .py files in python) and there are packages which 
>  are directories that contain __init__.py files with assignments to
>  __all__ lists.  You don't have to know or care about these differences
>  among these when using the modules.  You just use import the same way 
>  regardless of the underlying implementation details.
> 
>  While *writing* your own modules you can start a module as a .py
>  and later refactor it into a package or rewrite it in C and compile
>  as your needs require.

Don't take this the wrong way but I am a newbie and didn't understand
86% of what you just said. All I got was modules are classes functions
and variables. and a few other things but that was to complicated for
me (the newbie)


From Answer.via.news.please.prikryl at skil.nospam.cz  Thu Apr 18 04:11:06 2002
From: Answer.via.news.please.prikryl at skil.nospam.cz (Petr Prikryl)
Date: Thu, 18 Apr 2002 10:11:06 +0200
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com>
Message-ID: <a9lv5e$t86$1@news.nextra.cz>

Greg Weeks wrote in the answer to the previous post...
>
> [...]
> There should have been an operator in Python that
> says when objects are conceptually the same.

The problem here is what does it mean
*conceptually*.  For me *conceptually* means that
type([]) == type([]), but also
type([]) == type([1, 2, 3]).  In other words,
the *list* as the abstraction is the concept.

> For the moment, call it "eq".  The following would
> hold:
>
>     3 eq 3
>     not (3 eq 4)
>     a eq a # PROVIDED a HAS A VALUE
>     "foo" eq "foo"
>     not ([] eq [])
>     not ({} eq {})
>     () eq ()
>
> [...]
>
> Python almost has an "eq" operator.  "==" is like
> "eq" with two exceptions:
>
>     lists
>     dictionaries
>
> Since "==" is so close to "eq", it is natural to
> think of "==" as meaning conceptual sameness.
> And if "==" meant conceptual sameness, then lists
> would be immutable, similar to Lisp lists (in the
> absence of rplaca and rplacd).  In that case,
> .append(3) would have been implemented by
> consing, and your intuition would have been
> right.

For me, nor ``==`` nor ``is`` means *conceptual
sameness*, as said on the top of this message.

There are two distinct things when comparing the
objects: the object content and the object
identity.  When comparing two objects' content,
the content is the same if the structure and
the related element values inside the compared
objects are equal.  In other words, there is no
need for checking the objects identity in such
case; however, when the objects are identical
(i.e. one object) then the comparison can be
optimized to early return of true value.

Whith respect to this the ``==`` should compare
the content of comparable objects and the ``is``
operator should compare whether the identity of
the objects is equal.  For me, Python does
exactly this.

> PS: Interestingly enough, all the expressions
> involving "eq" above are true on my system with
> "eq" replaced by "is".  But (I believe) the
> following are implementation-dependent:
>
>     3 eq 3
>     "foo" eq "foo"
>     () eq ()
>
> And "eq" disagrees with "is" for (properly
> defined) immutable class instances.  And,
> although I didn't mention it above, 3 eq 3.0
> should probably hold.  (Which, I should add,
> means that Guido was *conceptually* on target
> with the "/" redefinition.)

The Python Reference Manual (Release 2.2.1),
section *5.9 Comparisons* says:

    The operators is and is not test for object
    identity: x is y is true if and only if x and
    y are the same object. x is not y yields the
    inverse truth value.

>From this point of view, the ``3 is 3.0`` should
never hold because they are of different types
(they never can be the same object);  but it
does not make sense to ask so either -- see below.

The ``3 == 3.0`` holds, because there is one more
step of the implicit conversion from int to float
and comparison of two float values (correct me,
if I am wrong).

The ``[]``, ``{}``, and ``()`` are in fact
constructors of new objects of the related type,
while ``3`` or ``"foo"`` are literals that may or
may not produce new object -- see below.

So, the ``[] is []`` should never hold, because
we are creating two instances of the list
abstraction.  On the other hand, the ``[] == []``
should always hold, because the content of the
two empty lists is equal.

> Furthermore, two "eq" objects should hash to the
> same value.  So *any* object could be a hash key.

It would be easier to implement this if there was
no keys() method of the dictionary, i.e. when you
used your *any object* content only to generate
the hash value and the hash value was used as the
key.  But then, there would be no back way from
the hash to the key value.  Moreover, two different
keys may produce the same hash values and the
dictionary structure has to discover that kind of
conflict and to do some more steps to make the
different keys really different.

If you think more about it, you will find that
Python requirements are natural, the key must be
immutable and even something more...  The
immutable key (like a tuple) must not contain
references to mutable objects, because the
references are automatically replaced by the
value of the object when the hash value is to be
determined (correct me if I am wrong).

Brian Quinlan wrote in the answer iside this thread ...
>
> >     >>> 3 is 3
> >     True
>
> This is an implementation accident.

... On the other hand, what meaningful result one
should expect from ``3 is 3``?  For me, ``3 == 3``
makes sense, but to ask whether ``3 is 3``
(i.e. whether the objects are identical) is
questionable. The Python Reference Manual
(Release 2.2.1) explains:

   5.2.2 Literals

   [... string, numbers, ...]

   All literals correspond to immutable data
   types, and hence the object's identity is less
   important than its value. Multiple evaluations
   of literals with the same value (either the
   same occurrence in the program text or a
   different occurrence) may obtain the same
   object or a different object with the same
   value.

See you,
  Petr

--
Petr Prikryl (prikrylp at skil dot cz)




From aahz at pythoncraft.com  Mon Apr  1 09:28:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Apr 2002 09:28:45 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <idVp8.48143$S52.1619383@news2.tin.it>
Message-ID: <a89qqt$b2a$1@panix1.panix.com>

In article <idVp8.48143$S52.1619383 at news2.tin.it>,
Alex Martelli  <aleax at aleax.it> wrote:
>
>        ...

I've been reading this thread before deciding how to vote.  Before Alex's
post, I was somewhere between +0 and +1; now I'm -0.  I have also seen
some suggestions for postponing this change to 2.4; that gets +0 from me.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Why is this newsgroup different from all other newsgroups?


From wandersw at ligbr.com.br  Mon Apr  8 12:48:06 2002
From: wandersw at ligbr.com.br (Wander Stabenow Weberling)
Date: 8 Apr 2002 09:48:06 -0700
Subject: Python + MySQL + Qt/Gtk(or somethink like that)
Message-ID: <d9641a00.0204080848.4d80d2b0@posting.google.com>

Hi,

I'm brazilian programmer and I realy want to know how to use python
with MySQL and a GUI toolkit like Qt or Gtk. All tutorials that I
found in the web (maybe I dont know how to search) are very basical.

If you know any information (book, tutoria, site, or somethink else)
that could help me, please send me via e-mail (wandersw at ligbr.com.br).

Thanks by now.


From DavidA at ActiveState.com  Tue Apr  9 13:23:50 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Tue, 09 Apr 2002 10:23:50 -0700
Subject: IDE Wars, the prequel!
References: <mailman.1017880688.9896.python-list@python.org> <6RSq8.29207$0T6.8753904@e3500-atl1.usenetserver.com> <3CB31C25.6090101@yumpee.org>
Message-ID: <3CB323A6.B74F6235@activestate.com>

Manoj Plakal wrote:

> I believe Komodo is no longer being actively developed?

Nope!  We've had cutbacks in staff, true, but we're still working on
it.  A release candidate (1.2.7) is currently being tested by the folks
on the komodo-discuss mailing list, and it will be released soon.  It
fixes a bunch of annoying bugs, especially for those other languages
(Perl, PHP, and Tcl =).

Our internal version has grown some other features as well, but that
version won't ship for a while, since we want to bunch a set of new
features and enhancements in a "bigger" release based on customer
feedback.

Cheers,

-- David Ascher
   ActiveState




From merman at freenet.de  Mon Apr  8 10:58:28 2002
From: merman at freenet.de (merman)
Date: Mon, 8 Apr 2002 16:58:28 +0200
Subject: swig-problem
Message-ID: <3cb1ae14$0$12336$9b622d9e@news.freenet.de>

Hello pythonistas,

sorry for my strange english!

I'm a newbie and I have a problem with SWIG-1.3.11.
I use win2k with the Borland-Compiler bcc55 (commandline).

I want try to write a little python-extension in c++.

Here's is my Makefile:

--- Begin of
Makefile.tree-swig ---------------------------------------------------------
-------------------

SWIG        = swig
PY             = C:\Python22
PY_INC    = C:\Python22\include
CC            = bcc32
TARGET   = -eTreec.dll

#OPTIMIZE     = -O
#FORCE_CPP = -P

#..............................................................

#all: Treec.lib Tree.py
all: Treec.dll Tree.py

# wrapper + real class
Treec.dll: Tree_wrap.obj Tree.obj
 $(CC) $(TARGET) Tree_wrap.obj Tree.obj

#generated class wrapper module
Tree_wrap.obj: Tree_wrap.cxx Tree.h
 $(CC) -I$(PY_INC) -I$(PY) -c Tree_wrap.cxx

Tree_wrap.cxx: Tree.i
 $(SWIG) -c++ -python Tree.i

Tree.py: Tree.i
 $(SWIG) -c++ -python Tree.i

Tree.obj: Tree.cpp Tree.h
 $(CC) -c Tree.cpp

#.........................................................

cpptest:
 $(CC) -eGetDirs.exe TreeMain.cpp Tree.cpp

clean:
 del /P *.exe *.obj *.tds *.dll *.lib *.*~ *.cxx

--- end of
Makefile.tree-swig ---------------------------------------------------------
------------------

If I type 'make -f Makefile.tree-swig cpptest' the compiler works fine and
generates the 'GetDirs.exe'. So far so good!

But if I type this line: 'make -f Makefile.tree-swig' it appears the
following error-message:

Fatal F1003 c:\Borland\Bcc55\include\stdcomp.h 5: Error directive: Must use
C++
for STDCOMP.H
*** 1 errors in Compile ***

I think the problem is to say Swig it should generate the wrapper
'Tree_wrap.cpp' and not 'Tree_wrap.cxx' because bcc55-compiler don't know
the *.cxx-extension and therefore the failure above was generated?!?

How can I say SWIG it should generate wrappers mit *.cpp-extension and not
with the *.cxx-extension?

Anybody knows a solution?

It's lot of text - SORRY! But I could' t explain this better.

thanx Thomas








From comdog at panix.com  Fri Apr  5 09:36:32 2002
From: comdog at panix.com (brian d foy)
Date: Fri, 05 Apr 2002 08:36:32 -0600
Subject: _The Perl Review_ does Python
References: <040420021725172972%comdog@panix.com> <f72dac0d.0204050205.4cb3e8ac@posting.google.com>
Message-ID: <050420020836327939%comdog@panix.com>

In article <f72dac0d.0204050205.4cb3e8ac at posting.google.com>, Alan Kennedy <alanmk at hotmail.com> wrote:

> brian d foy wrote:
> > for our April 2002 issue (April Fool!), The Perl Review replaced
> > its Perl content with Python articles, so Python folks might want
> > to read this issue:
> > 
> >    http://www.theperlreview.com/Issues/The_Perl_Review_0_2.pdf

> Forgot to mention: Does this mean that the Golf competition is open to
> Python entries :-)

i think Python code would be at a disadvantage since the golf umpires 
count whitespace in the number of strokes :(.

-- 
brian d foy <comdog at panix.com> - Perl services for hire
The Perl Review - a new magazine devoted to Perl 
<http://www.theperlreview.com>


From python at dohao.org  Sat Apr 27 00:51:50 2002
From: python at dohao.org (Wenshan Du)
Date: 26 Apr 2002 21:51:50 -0700
Subject: zope and jboss or weblogic
Message-ID: <e786d63d.0204262051.14ebe470@posting.google.com>

hi,
   I like zope, I want to know what's the difference between zope and
jboss or weblogic? which is better? Can python & zope replace J2EE?


From cabrera at wrc.xerox.com  Fri Apr 12 08:59:54 2002
From: cabrera at wrc.xerox.com (jic)
Date: 12 Apr 2002 12:59:54 GMT
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com> <mailman.1018606659.23241.python-list@python.org>
Message-ID: <a96loa$e38$1@news.wrc.xerox.com>

In article <mailman.1018606659.23241.python-list at python.org>, gimbo at ftech.net 
says...
>
>On Thu, Apr 11, 2002 at 05:47:42PM +0000, jic wrote:
>
>>       ftp.storbinary('STOR ' + fname,f.read )
>
>The second parameter's wrong: you want to call f.read() to get the
>file's contents, but in fact you're just passing a reference to the
>file object's read method.
>
>Try:
>
>ftp.storbinary('STOR ' + fname, f.read())
>
>Hope this helps,

Andy,

I did try that and it didn't work.  I also tried f.read(1024), 
f.read(100000000) and still didn't work.

thanks for your help.

jos?






From sholden at holdenweb.com  Mon Apr  1 18:33:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 1 Apr 2002 18:33:16 -0500
Subject: readlines() problem, more info
References: <nd2q8.149996$ZR2.64791@rwcrnsc52.ops.asp.att.net>
Message-ID: <_76q8.118831$Ou.2001@atlpnn01.usenetserver.com>

"John Purser" <NO_SPAM_jmpurser2 at attbi.com> wrote in message
news:nd2q8.149996$ZR2.64791 at rwcrnsc52.ops.asp.att.net...
> Hello,
>
> I've got a little function:
>
> def FindLine(Record):
>      im1_Dif.seek(0)
>      temp = im1_Dif.readlines(Record)
>      lngCount = 0L
>      for line in temp:
>           print str(lngCount), '   ', line
>           lngCount += 1
>      test = im1_Dif.readline()
>      print test
>
>  That I'm using to trouble shoot a small script.  When I run this in the
> interactive mode it returns 136 lines to "temp" no matter what number I
pass
> to it.  Can anyone tell me why this is happening?
>
> I'm using Python 2.2 on Windows 2000.
>
So am I.

>>> import sys
>>> print sys.stdin.readlines.__doc__
Read until EOF using readline() and return a list containing the lines
        thus read. If the optional sizehint argument is present, instead of
        reading up to EOF, whole lines totalling approximately sizehint
bytes
        (possibly after rounding up to an internal buffer size) are read.

I suspect you are seeing the rounding up takinhg place.

regards
 Steve






From loewis at informatik.hu-berlin.de  Thu Apr 18 12:09:48 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Apr 2002 18:09:48 +0200
Subject: Finalizing C extension modules
References: <7kctbustl4vsrp54fk7i7s8iu3ecf90oko@4ax.com>
Message-ID: <j43cxt6l9v.fsf@informatik.hu-berlin.de>

Matthias Baas <baas at ira.uka.de> writes:

> Suppose I have a C/C++ module that allocates some resources in its
> init function, where would be the proper place to release them again
> when the Python interpreter is shutting down? Is there an equivalent
> to Py_Finalize() for extension modules? 

You can use Py_AtExit.

> Why is it that extension modules aren't unloaded?

There is no portable, platform-independent way to provide that
feature; platforms have strange restrictions for unloading of shared
libraries. Furthermore, it might lead to surprising crashes: Somebody
may still hold a reference into some piece of code of the module; if
the module has been unloaded, calling the function at the C level will
just crash.

> However, I will never call Py_DECREF() for that class object I was
> retrieving from the pure Python module. Is this ok? Or will this
> result in a memory leak? Where am I supposed to call Py_DECREF()?

This will result in a memory leak, but that is no problem: Once the
operating system process terminates, the operating system will reclaim
all memory of the process, and most other resources.

If you are worried that you should release the object, register a
function with Py_AtExit.

HTH,
Martin



From weeks at vitus.scs.agilent.com  Sat Apr  6 14:43:22 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Sat, 6 Apr 2002 19:43:22 +0000 (UTC)
Subject: yet another "Here's to 1.5.2" post
Message-ID: <1018131203.756898@cswreg.cos.agilent.com>

Actually, Python 1.5.2 was already a bit too large.  The functional
programming constructs (lambda, map, reduce, filter) are fluff.

List comprehension is cutesy and practically Perlish, an idiosyncrasy
reminiscent (in flavor and aroma) to Perl's ".." operator.

"print" has always been something of a blemish.  Adding more functionality
to it doesn't seem like a good idea.  The new syntax smells bad.

The so-called "nested scopes" added in 2.1 are a mistake.  They take away
i) a simple rule and ii) the ability to capture the scope at any point in a
program with two dictionaries.

With somewhat less confidence -- and really, what's the point of being sure
about this (like it matters!) -- I think the redefinition "/" and the
addition of a boolean type were bad choices.

Now, if only the energy that was put into adding the above blemishes had
been placed into making Python more robustly installable.  On my system
(HP-UX 10.2) with my reasonable efforts at building a system:

    in 1.5.2, there is no FCNTL module

    in 2.2, "configure" fails due to erroneous C pre-processor code

    in 2.2, "make test" fails for reasons I haven't had the time to pin down

I'll keep using Python, so long as Python fills the Perl niche better than
Perl.  But creeping TIMTOWTDI and installation problems make me worry,
worry, worry.


Greg


From tim.one at comcast.net  Mon Apr 15 00:19:52 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 15 Apr 2002 00:19:52 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <eppstein-BBF667.20005414042002@news.service.uci.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEDDOOAA.tim.one@comcast.net>

[David Eppstein]
> ...
> The reason I was getting into trouble anyway was because of the way
> right-shifting can change sign:
>
> >>> 0x40000000<<1
> -2147483648

That looks like a left shift to me.  Right shifting in Python never changes
"the sign".

> This is inconsistent with int-long unification.

Very badly so, yes.  It's the prime reason this *part* of unification was
delayed:  it's going to break code.

> In fact, I think it is a bug.

It's functioning as documented and as designed, so while you can call it "a
bug", it's not "a bug" under any meaning of the phrase a developer will
accept.  I think of it is a design error.  Unfortunately, for some kinds of
bit-fiddling code it's a *useful* design error, so we're tiptoeing on raw
eggs here.

  The answer should, of course, be 2147483648L, just as you would
> get by multiplying by 2 instead of shifting.

That is the goal.





From a.schmolck at gmx.net  Wed Apr 17 11:08:21 2002
From: a.schmolck at gmx.net (A.Schmolck)
Date: 17 Apr 2002 16:08:21 +0100
Subject: Unicode name questions
References: <15548.35951.239060.214140@12-248-41-177.client.attbi.com> <009901c1e58b$28a5bd90$445d4540@Dell2> <mailman.1018994314.14044.python-list@python.org> <m3y9fmvij0.fsf@mira.informatik.hu-berlin.de> <Xns91F325AA52316RASXnewsDFE1@130.133.1.4>
Message-ID: <yfsvgaqbbx6.fsf@black132.ex.ac.uk>

Philip Swartzleonard <starx at pacbell.net> writes:

[skipped bit about unicode character names]
> 
> Hm. I was browing through some of their code charts for the heck of it, 
> and I noticed that they listed certain kata/hiragana characters's primary 
> names as SI, TI, TU, etc. I've listened to a considerable amount of 
> japanese media, and AFAICT the correct romanization of those sounds would 
> be SHI, CHI, TSU, etc.

(bear in mind I'm not an expert either...)

Depends what you mean by "correct": the *pronounciations* are better
transcribed by SHI, TSU etc. but that romanization is more for the benefits of
foreigners than what japanese themselves resort to when they must use romanji,
by my experience. The japanese syllabic alphabet is ordered something like:

...
KA    KI      KU    KE    KO
SA    S(H)I   SU    SE    SO 
TA    TI(CHI) T(S)U TE    TO
...
MA    MI      MU    ME    MO
..


So TSU is really just an irregularly pronouced TU, if you want (there is no
"tu" sound in japanese) and since it's unambigous and shorter to type, that's
what japanese seem to be using when they have to use latin characters and
don't need to provide pronounciation hints for ignorant foreigners :)


> 
> But then they also have some other wierdness, e.g., listing VA, VI, VE, VO 
> as WA, WI, WE, WO + diacrit, where WI and WE are old, obsolite* characters 
> as mentioned in my dictonary, and to the best of my knowledge the methods 
> for doing V sounds are eihter old-style use of B sounds, or new-style use 
> of A,I,E,O + diacrit. And there's this 'katakana iteration symbol' (and 
> hira, and voiced of both) that i've never so much as heard of... :\
> 


I can't see what's weird here: the sounds "vi" and "wi" etc. which do not
occur in (modern) japanese are tranliterated either
'katakana-U-with-diacritical followed by small-katakana-i' (new style) or
"katakana BI" (old style) in loan words. But these transliterations are
tranliterations, i.e. some kludge to write "VI" when the language in fact has
no such thing -- rather than an *unique character* WI / VI. Thus they don't
belong to the japanese character table, they are comprised of symbols taken
from it.

But there *used to be* unique katakana characters WI / VI (WI with
diacriticals) which are no longer used in modern japanese, but they should
certainly be in the unicode table and WI, VI seem pretty appropriate names to
me!



alex


From oliver.pust at web.de  Thu Apr 11 05:22:46 2002
From: oliver.pust at web.de (Oliver Pust)
Date: Thu, 11 Apr 2002 11:22:46 +0200
Subject: initializing list with number of elements given by a variable
Message-ID: <3CB555E6.3080207@web.de>

Hi,

I need to create a list which number of elements is given by an integer 
variable.

So something like this:

 >>> length_of_list = 6
 >>> list = ????????? # what to place here, is the question
 >>> print list
 >>> [0,0,0,0,0,0]    # the resulting list should look like this

Can anybody help me here?

Thanks,
	Oliver



From h_schneider at marketmix.com  Mon Apr  8 02:26:51 2002
From: h_schneider at marketmix.com (Harald Schneider)
Date: Mon, 8 Apr 2002 08:26:51 +0200
Subject: Read SFF (Structured Fax Format)-Files
Message-ID: <a8rcrv$qgg$01$1@news.t-online.com>

Hi,

is there a module, which is capable of reading or converting SFF files to
JPG ?
PIL has no native support for that ...

All the best,
Harald




From eppstein at ics.uci.edu  Mon Apr 15 14:43:04 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 15 Apr 2002 11:43:04 -0700
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <roy-0BE2A8.19181913042002@news1.panix.com> <m3n0w7ywqv.fsf@mira.informatik.hu-berlin.de> <Prau8.17347$vF6.558738@news2.tin.it>
Message-ID: <eppstein-15B457.11430415042002@news.service.uci.edu>

In article <Prau8.17347$vF6.558738 at news2.tin.it>,
 Alex Martelli <aleax at aleax.it> wrote:

> Very well put.  Indeed, given the small amount of bit-twiddling in
> typical Python usage, I think & ~ | ^ are already overboard -- they
> make a lot of sense in C, a very low-level language that is used a
> LOT for bit-twiddling, but, in Python, named built-in functions might
> have been more appropriate in hindsight (fewer precedence levels to
> remember) -- if not, even, functions in a separate module.

Is it really appropriate to design a language so that only "typical" 
usages are simple, and everything else is made more difficult?

My most recent Python program (a testbed for some signal encoding 
schemes) made extensive use of both bit twiddling and simple generators.  
I'm very happy that Python was available for that purpose.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From greg at cosc.canterbury.ac.nz  Wed Apr 10 02:16:47 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 10 Apr 2002 18:16:47 +1200
Subject: Pyrex and weave.accelerate
References: <mailman.1017795845.3464.python-list@python.org><r1tq8.161131$af7.78459@rwcrnsc53> <3CAAB238.290440C5@kfunigraz.ac.at><lbhpu1hhwnb.fsf_-_@xihu.inria.fr> <3CAAEF17.32333986@kfunigraz.ac.at><lbh8z84nake.fsf@xihu.inria.fr><15531.7864.698585.485436@12-248-41-177.client.attbi.com><lbh7knolreu.fsf@xihu.inria.fr><15531.9709.924953.953487@12-248-41-177.client.attbi.com><mailman.1017850275.2134.python-list@python.org><0efa01c1db5c$72a7d380$6b01a8c0@ericlaptop> <lbh8z83kiez.fsf@xihu.inria.fr>  <mailman.1 <00e101c1dd14$5620a930$6b01a8c0@ericlaptop> <mailman.1018297042.2577.python-list@python.org>
Message-ID: <3CB3D8CF.FBB67880@cosc.canterbury.ac.nz>

Patrick Miller wrote:
> 
> It appears that the premise of Pyrex (assumption warning!
> I'm not reading Greg's mind!) is that C types have
> Python equivalents expressible through the API
> (e.g. short/int/long are expressible through
> the PyInt_AsLong and PyInt_FromLong methods).

Something like that. There are certain basic types
(int, float, char *) for which the correspondence
is obvious, and Pyrex handles them automatically.
For the rest, Pyrex makes it easy for you to write
your own code to handle them however you see fit.

> Weave.accelerate allows programmers to teach weave
> the API for a type and then have it build the
> extension module from that knowledge.

One day Pyrex may know more about some of the
builtin Python objects, so you can do things like
getting at the contents of an array.array directly.
I'm hoping this will fall out naturally once
extension types can be defined, by treating these
objects as sort of "extern extension types".

> Now, I get full fallback to Python (in case weave fails to compile)

Failing to compile isn't really an issue with
Pyrex. If your Pyrex module fails to compile,
you fix it so it does compile. :-)

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


From ole.fostad at netcom.no  Sat Apr 13 13:07:31 2002
From: ole.fostad at netcom.no (Ole Fostad)
Date: Sat, 13 Apr 2002 17:07:31 GMT
Subject: Post-installation with distutils windows installer
Message-ID: <1104_1018717651@news.netcom.no>

Hi everyone!

I wondered if there is a way to run a post installation script when using the distutils 
Windows installer?

Ole Fostad






From johanfo at essay.org  Mon Apr 29 19:38:43 2002
From: johanfo at essay.org (Johan Fredrik Øhman)
Date: Tue, 30 Apr 2002 01:38:43 +0200
Subject: Python Webpages
References: <aagr7h$5c9$1@maud.ifi.uio.no> <lEXy8.34518$n4.7236390@newsc.telia.net> <aahiq2$6t3$1@maud.ifi.uio.no> <aai5oo$99l$1@maud.ifi.uio.no> <slrnacpc38.obc.jon+usenet@snowy.squish.net> <23891c90.0204290051.3edcc62d@posting.google.com> <slrnacq7jk.gav.jon+usenet@snowy.squish.net> <mailman.1020090053.19284.python-list@python.org> <slrnacqrps.sdd.jon+usenet@snowy.squish.net> <mailman.1020102500.7281.python-list@python.org> <slrnacra60.v3g.jon+usenet@snowy.squish.net> <aakelj$do6$1@maud.ifi.uio.no> <slrnacrj8f.n56.jon+usenet@snowy.squish.net>
Message-ID: <aakli4$fdc$1@maud.ifi.uio.no>

Because it kind of separates the code from the design at the same time as
they are kept in the same place.
Funny, the template engine from Francos was _very_ similar.  What I think we
all can agree on is that heavily nested html and python, where the python
code is "hidden" in the html or the opposite, is not good.

BTW I'm installing jonpy now :)
--
JF?


"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote in message
news:slrnacrj8f.n56.jon+usenet at snowy.squish.net...
> In article <aakelj$do6$1 at maud.ifi.uio.no>, Johan Fredrik ?hman wrote:
> > I just thought of great idea:  How about using one file and not mix html
and
> > python code at the same time ?
> >
> ><html>
> ></html>
> >
> ><!--PYTHON Code
> > -->
>
> You could do that, but why on earth would you want to?




From demian0311 at yahoo.com  Fri Apr 12 22:15:49 2002
From: demian0311 at yahoo.com (demian neidetcher)
Date: 12 Apr 2002 19:15:49 -0700
Subject: ANN: prysm 1.0
Message-ID: <mailman.1018728965.20299.clpa-moderators@python.org>

prysm 1.0

http://prysm.sourceforge.net/
https://sourceforge.net/projects/prysm/

prysm is an excruciatingly simple log file colorizer implemented in
Python.

the itch that i am scratching is log file blindness. there are certain
log file events that are of great importance. putting a bunch of
characters or newlines in logging doesn't always get your attention
and when it's too obnoxious you have to go in and remove it from your
regular logging. with prysm you can set up rules to define the color
(foreground and background) specific lines will be displayed.

it's also a testament to how much python rules... less than 100 lines,
including the gpl header!  ...sa-weeeet.

send questions, comments, etc... to
demian0311 at users.sourceforge.net




From sholden at holdenweb.com  Tue Apr 30 18:02:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 30 Apr 2002 18:02:18 -0400
Subject: Stackless goes Limbo
References: <3CC6D07E.409@tismer.com> <mailman.1019758546.27031.python-list@python.org> <aa9n7h$4m9$1@panix1.panix.com> <mailman.1019781707.18338.python-list@python.org> <aaaaic$3b7$1@panix1.panix.com>  <mailman.1019792457.31426.python-list@python.org> <slrnacm8a7.mru.mlh@vier.idi.ntnu.no> <dgXy8.72409$z55.22493@atlpnn01.usenetserver.com>  <mailman.1020033011.19748.python-list@python.org> <slrnactoru.pv.mlh@vier.idi.ntnu.no> <mailman.1020196975.26619.python-list@python.org>
Message-ID: <fyEz8.4763$Au5.3848@atlpnn01.usenetserver.com>

"James J. Besemer" <jb at cascade-sys.com> wrote in message
news:mailman.1020196975.26619.python-list at python.org...
>
> Magnus Lie Hetland wrote:
>
> > James J. Besemer wrote:
> > >
> > >Made me look.
> >
> > Gullible, anyone? <wink>
>
> Yup nailed me good on that one.
>
> I realized just AFTER pressing send.  A little less coffee that day and I
may have
> spared myself that minor embarrassment.
>

Heh, heh, heh.
--

Steve Holden: http://www.holdenweb.com/ ; Python Web Programming:
http://pydish.holdenweb.com/pwp/







From fperez528 at yahoo.com  Tue Apr 30 23:23:02 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 30 Apr 2002 21:23:02 -0600
Subject: Any CD-ROM writer support in python?
References: <3CCF59F7.D5AA5EE4@attglobal.net>
Message-ID: <aann9v$p14$1@peabody.colorado.edu>

Pierre Rouleau wrote:

> I am looking for python based CD-ROM writer support.
> Does anyone know where I could find some python code or libraries to do
> that?

If you're under Linux, cdrecord is _the_ standard. And you don't want to 
rewrite that, trust me ;) Under Windows, no idea.

hth,

g


From menscher+python at uiuc.edu  Tue Apr  9 02:33:04 2002
From: menscher+python at uiuc.edu (Damian Menscher)
Date: Tue, 09 Apr 2002 06:33:04 GMT
Subject: multiple ranges to for statement?
Message-ID: <AWvs8.38021$tg4.452125@vixen.cso.uiuc.edu>

I'm hoping to do something along the lines of:

for i in range(a,b),range(c,d):
    stuff goes here

But of course that sets i to the *two* values: range(a,b) and
range(c,d).  Is there a simple way to get it to take on every value
from a to b and then every value from c to d?  I would think there
should be, but I'm a newbie and can't figure it out.

So far, the best I can think of would be:

values = range(a,b)
for i in range(c,d):
    values.append(i)
for i in values:
    stuff goes here

It just doesn't seem very clean, so I'm wondering if I'm missing
something obvious.

Damian Menscher
-- 
-=#| Physics Grad Student & SysAdmin @ U Illinois Urbana-Champaign |#=-
-=#| 488 LLP, 1110 W. Green St, Urbana, IL 61801 Ofc:(217)333-0038 |#=-
-=#| 1412C DCL, Workstation Services Group, CCSO Ofc:(217)244-3862 |#=-
-=#| <menscher at uiuc.edu> www.uiuc.edu/~menscher/ Fax:(217)333-9819 |#=-


From kaustubhkp at hotmail.com  Fri Apr 19 00:01:53 2002
From: kaustubhkp at hotmail.com (Kaustubh Kalgaonkar)
Date: Fri, 19 Apr 2002 09:31:53 +0530
Subject: about Blt configuration
Message-ID: <F122J23yqfTp2RgiGIG0001c8e9@hotmail.com>

Hi,
>I have python 2.2 installed on my machine.I have also tried it on
>python 2.1.1. I downloaded Pmw ran the demo the demo ran fine
>except of BLTgraph for which it said BLT pakage not installed so i
>downloaded one "blt24y-for-tcl83.exe" installed it in
>c:\Python22\tcl the installation went through but it is still
>giving same error . i am working with windows 2000 please tell me
>if any further configuration is required.
Thanks
Kaustubh

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp.





From pyth at devel.trillke.net  Sun Apr 28 17:26:57 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 28 Apr 2002 23:26:57 +0200
Subject: Update locals() / resolved
In-Reply-To: <slrnacop1t.dcb.mlh@vier.idi.ntnu.no>
References: <20020427195725.GA16217@devel.trillke> <20020427150607.A2052@unpythonic.net> <20020427202326.GB16217@devel.trillke> <mailman.1019941968.7326.python-list@python.org> <PNFy8.28407$8D3.758650@news1.tin.it> <slrnacop1t.dcb.mlh@vier.idi.ntnu.no>
Message-ID: <20020428212657.GI16217@devel.trillke>

On Sun, Apr 28, 2002 at 09:06:05PM +0000, Magnus Lie Hetland wrote:
> In article <PNFy8.28407$8D3.758650 at news1.tin.it>, Alex Martelli wrote:
> >holger krekel wrote:
> >        ...
> >> ok. i found a nice workaround (if one dare call it that:-)
> >> 
> >> class dictlookup(type({})):
> 
> Why use type({}) when you can use dict?

1.  because i didn't know. 

2. I remembered from PEP253 that objects and types 
   are/should be 'unified'. So using type({}) somehow 
   made sense to me because it expresses
   'this object inherits behaviour from the {}-type'

3. And 'dict' is something i often use in local definitions
   when working with dictionaries. 

holger




From stephen at xemacs.org  Thu Apr 18 03:04:39 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 18 Apr 2002 16:04:39 +0900
Subject: [OT] What is Open Source?
References: <mailman.1018662715.17645.python-list@python.org> <slrnabm3h5.nkd.huaiyu@gauss.almadan.ibm.com> <mailman.1018897470.30703.python-list@python.org> <899f842.0204170840.29e579ae@posting.google.com> <20020418002823.C4068@phd.pp.ru> <mailman.1019077897.8110.python-list@python.org>
Message-ID: <87r8ldqygo.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Cliff" == Cliff Wells <logiplexsoftware at earthlink.net> writes:

    Cliff> I suppose Microsoft's claim that the GPL is viral could be
    Cliff> better applied to their own source since merely *seeing* it
    Cliff> could affect your code whereas you have to actually *use*
    Cliff> GPL'd code to be affected.

I've not seen this before.  Please send it to rms.

gpl-is-viral-but-why-upset-rms-when-there-are-better-targets-ly y'rs

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.


From maxm at mxm.dk  Sat Apr  6 04:54:26 2002
From: maxm at mxm.dk (Max M)
Date: Sat, 06 Apr 2002 11:54:26 +0200
Subject: Booleans, integer division, backwards compatibility; where is
 Python  going?
References: <slrnaapc23.r7n.philh@comuno.freeserve.co.uk> <Xns91E75A62ACE63JamesLLugojcom@216.148.53.96> <mailman.1018047273.8088.python-list@python.org> <7x1ydt8ve7.fsf@ruckus.brouhaha.com> <3CAE6A2D.291C35F2@engcorp.com> <7xadshtrj1.fsf@ruckus.brouhaha.com>
Message-ID: <3CAEC5D2.3020902@mxm.dk>

Paul Rubin wrote:

> Stable doesn't just mean old scripts run without change in new
> interpreters.  It also means new scripts usually run without change in
> old interpreters.  That's what keeps breaking in Python.  

Naturally you would not expect to use something like Unicode in Python 
and then expect it to be backwards compatible. That would be like magic.

> That is true, but the marketing is written to make people think the
> evolution has ALREADY HAPPENED, and Python is sufficiently effective
> and strong to not need further rapid change.  C is not stagnant (C
> implementations keep improving) but it is mature and stable.  If I
> download an interesting new C program today, I can probably build and
> run it on my 5 year old Debian 1.x system without upgrading my C compiler.
> I generally can't do that with Python, because of all the new features
> that keep getting introduced.

C is a completely different beast than Python. As far as I can tell 
Python is a high level collection of "best practice" c libraries. It 
just happens to have an easy syntax on top to express your views.

So many of the things that you will do manually in c is done for you in 
Python. So because of the higher level of the language it is not 
possible to be as backwards compatible.

Best practice change as we get more experienced in programming. New 
patterns and solutions evolve. If we are to express them efficiently in 
Python, Python must change too.

Programming and the art of expressing ones ideas through code is really 
still to difficult and limited. So either somebody has to develop a 
completely new paradigm in programming, or we have to evolve the 
existing languages and keep building new constructs on top of them.

Shurely evolution is the only feasible way to do it, and I strongly 
prefer if it is my language of choice that evolves the best.

At some time there might be to big a schizm between Python and the best 
way to express all of the patterns being build on top of it. Then a new 
language will need to be developed and we can leave Python behind.

But at this point I cannot really see where this better language should 
be developed from other than Python.

> Most recent changes in Python are good (list comprehensions, x in dict,
> etc) but I'd rather that they appear only in major releases (2.0, 3.0, etc)
> which happen every 2 years or so, rather than in the minor releases
> that come out every few months.

Then you would probably have a buggy Python that couldn't be used for 
anything for a year due to bugs playing of on each other.

regards Max M



From martin at v.loewis.de  Sat Apr 13 19:49:02 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Apr 2002 01:49:02 +0200
Subject: Does Python need a '>>>' operator?
References: <3cb88ae7@news.mhogaming.com> <m3sn5z1axf.fsf@mira.informatik.hu-berlin.de> <3cb8bc26@news.mhogaming.com>
Message-ID: <m3zo07yxap.fsf@mira.informatik.hu-berlin.de>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> writes:

> The merging of the 'long int' and the 'int' type is
> precisely WHY there _should_ be a '>>>' operator.  Did you
> read ALL of my post?

Can you elaborate? You explained that the >>> fills in zeroes for
32-bit numbers that have the highest bit set. With long integers, you
don't need that: If you have a number that has the 2**31 bit set, it
may still be a positive number, and right-shifting will fill in the
zero bit. So no need for a >>> operator.

> 
> | One might question whether the bitwise operators should have been
> | there in the first place - functions can do the same things just as
> | well.
> 
> You _ARE_ joking about this, aren't you?

I'm half-serious - the language would not lose too much without
operators for those operations. I'm absolutely serious that I don't
want to see any further operators added for such artificial special
cases.

> Let's say I want to work with a 64-bit number:

What is a 64-bit number in Python?

> >>> b = -a
> >>> print hex(b >> 1)
> -0x4000000000000000L
> 
> # !!! SURPRISE !!!
> # (I expected -0xC000000000000000L !!!)

Can you explain why you expected this? How could Python possibly know
how many bits you expected to be in the internal representation?

> I think this should work the same way an 'int' does-- but I
> don't know how to do that without breaking older code...

Add a function

def rightshift(number, shiftwidth, numberwidth=32, expandsign=1)

That function will work for 32-bit numbers, 64-bit numbers, and
97-bit-numbers. No need for a new operator.

Regards,
Martin


From johnroth at ameritech.net  Sat Apr  6 10:27:25 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 6 Apr 2002 10:27:25 -0500
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org><2zRcLUAVjXp8EwVv@jessikat.fsnet.co.uk><mailman.1017755045.4308.python-list@python.org><uakpepcn1iir15@news.supernews.com> <mailman.1017865089.20676.python-list@python.org> <Npor8.8281$q63.818@atlpnn01.usenetserver.com>
Message-ID: <uau4k9h4lg0rd6@news.supernews.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:Npor8.8281$q63.818 at atlpnn01.usenetserver.com...
> "Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
> news:mailman.1017865089.20676.python-list at python.org...
> > On Tue, 2 Apr 2002 21:21:27 -0500
> > John Roth wrote:
> >
> > >
> > > "Laura Creighton" <lac at strakt.com> wrote in message
> > > news:mailman.1017755045.4308.python-list at python.org...
> > > > Dear Guido:
> > > >
> > > > I would first like to thank you for giving us an opportunity to
> > > respond.
> > > > I have spent most of the weekend thinking and writing a reply to
this,
> > > > and I think that this has made me a better teacher.  For this I
am
> > > > grateful.  I realise that this is rather long, but I have
condensed
> > > > it substantially from my efforts earlier this weekend.   Thank
you
> > > > for taking the time to read this.
> > > >
> > > > Laura Creighton
> > > >
> > > > -----------------------------
> > > >
> > > I'm not going to quote this, as it's already too long. I read up
to
> > > the diatribe on the Swedish leap year, and I went AAAARRRRGGGH!
> > >
> > > I really don't care if you want the silly function to return
> > > a Boolean, the number of leap days in the year, or the name
> > > of the idiot parliamentarian who decided to fix the calendar
> > > so that 1712 was different from the rest of Europe.
> > >
> > > HOWEVER, if you want true or false as the return
> > > value, the name should be something like 'isLeapYear()'.
> > > If you want the number of leap days in the year, then
> > > it should be named something like 'numberOfLeapDays().'
> >
> Just so John know he isn't the only person going AARRGGHH, I read this
went
> AARRGGHH myself. It should not be named "something like
> 'numberOfLeapDays()'". It should be named LeapDays(), or possibly
> LeapDayCount() if you also have a function that returns the actual
dates
> which ARE the leap days. Or, I suppose, you could call it
> "theNumberOfLeapDaysInTheSwedishYear()" ;-)

Any of these would be acceptable. In practice, if I was working
on a project with you, I would accept any of those suggestions.
My objection, pure and simple, was that isLeapYear() requires
a boolean return, and hence is not something that you do
arithmetic to, regardless of the fact that Python permits it.

> > > Sorry to make this sound like a flame, but I try to
> > > practice XP. The first time I ever come across a
> > > function named "isSomething" that returns a conceptual
> > > integer, it's going to get refactored, right then and
> > > there to a name that actually expresses what it does.
> >
> People who live in glass houses ...

Oh?

John Roth





From jb at cascade-sys.com  Thu Apr 25 20:36:54 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Thu, 25 Apr 2002 17:36:54 -0700
Subject: Dynamic Binding
Message-ID: <3CC8A126.595FDDC0@cascade-sys.com>

Paul Foley wrote:

> ? JB:
> ? The original semantics (as formally defined in Lisp itself) called
> for
> ? dynamic binding.  Most original implementations did it that way,
>
> ISTM reading somewhere that the dynamic scoping in early
> implementations was actually the result of an error;

Nah.  There were typos in the specs in the original 1.5 document but
they had nothing to do with dynamic vs. static.

The original, landmark Lisp 1.5 Programmer's Manual clearly and
explicitly defines the language semantics to have dynamic scope.  This
specification appears informally at the end of chapter I and with
greater detail in Appendix B, from which I summarize:

     "Apply" takes the arglist from a lambda form and pairwise
     associates each arg with the corresponding values given in the
     function call.  Then the resulting list of pairs is prepended
     to the "AList," the master symbol table of name-value pairs
     for the interpreter.  Then the expression is passed to "Eval"
     to reduce to a value.  In evaluating atoms, Eval searches the
     AList sequentially for each name and naturally retrieves the
     innermost binding.  Pairs may be put on the AList at the top
     level of the interpreter, resulting in global variables.

Furthermore, I SPECULATE that dynamic scope was motivated in part by the
very concept of lambda binding.  Mathematically and on paper there was
no distinction between static and dynamic.  But in an implementation,
lambdas had to be able to freely nest, with inner bindings taking
precedence over whatever outer bindings there may be in a given
circumstance.  That's how lambda notation worked and it's essentially
the DEFINITION of dynamic binding.  Functions were simply names bound to
a lambda form, so they worked the same way.

So I think it's fair to say that "dynamic scope" originally was an
intrinsic and fundamental part of Lisp.

Paul Foley still:

> Where do you think dynamic scope is called for?

I never intended to put myself into a position to defend dynamic scope.
My only point was that it's how Lisp originally worked.

However, to answer your question, dynamic scope may be appropriate when
simplicity of implementation is of utmost importance.  It's trivial to
implement; you can almost argue that it's elegant.  It gives you the
basics (globals that may be overridden by locals) in a fairly general
fashion.  The fact that callee functions can access caller functions'
variables is a side effect of distinctly dubious utility, though I bet
$1 Larry Wall has an example or two somewhere how to exploit it for some
nefarious purpose.

Naive implementors continually reinvent it precisely because it's easy
and obvious.  Only later do some discover it's weaknesses and sense the
need for something better.  I don't mean to suggest that McCarthy was
naive; 35-40 years ago he was forging the way, exploring virgin
territory.  Naive was Larry Wall inventing Perl decades later, initially
without any local variables, then adding dynamic scope ("local()") and
only some time after that adding a form of truly local variables
("my()").  Ugh.

Reminds me of a variation of a famous saying: Some people see farther
than others because they stand on the TOES of the giants who preceded
them.  ;o)

Tim Peters wrote:

> [James J. Besemer]
> > The original semantics (as formally defined in Lisp itself
> > [McCarthy]) called for dynamic binding.
>
> It seems this would have been a surprise to McCarthy <wink>:
>
>     http://www-formal.stanford.edu/jmc/history/lisp/node4.html
>
> See paragraph "d. Free variables".  When a user bumped into a function
> where
> the difference between dynamic and lexical scoping mattered, and
> lexical
> scoping was what they needed, "I must confess that I regarded this
> difficulty as just a bug".  The understanding of these issues was much
>
> muddier at the time.

I disagree with your interpretation.

Rather he is saying that the implementation DID use dynamic "scoping,"
that it worked just fine in most circumstances and that when people
encountered a situation where "lexical scoping" would help McCarthy
regarded their "difficulty" simply to be a bug.

I presume he means that the bug was in their code.  "It hurts when arg
names conflict with your globals?  Don't DO that."

Tim Peters again:

> Dynamic scoping is attractive
> for implementation reasons; everyone who succumbs eventually regrets
> it...

I think that's a fair statement.

Christopher Browne wrote:

> Most of the time, lexical scope is likely to be more useful.  But
> there are places where dynamic scope is to be preferred.

Agreed.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com



--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From bergeston at yahoo.fr  Thu Apr  4 09:51:25 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Thu, 4 Apr 2002 16:51:25 +0200
Subject: Java to Python translation
References: <mailman.1017831968.9815.python-list@python.org> <a8hop5$rq6ba$1@ID-135695.news.dfncis.de>
Message-ID: <a8hp99$rvseg$1@ID-135695.news.dfncis.de>

"Bertrand Geston" <bergeston at yahoo.fr> wrote in message
news:a8hop5$rq6ba$1 at ID-135695.news.dfncis.de...
>
> "Andrew McNamara" <andrewm at object-craft.com.au> wrote in message
> news:mailman.1017831968.9815.python-list at python.org...
> > >> > I need to translate a (rather voluminous 100MB) java program to
> > >> > python. After a few searches on google, apparently the best tools
> > >> > for this task are Lex-Yacc (Flex-Bison on linux). While I am
working
> > >> > on the Lex-YACC HOWTO, I would be very interested in hearing of
> > >> > others' successful experiences/feedback at this. Any pointers of
> > >> > interest would similarily be appreciated.
> > [...]
> > >I think for a project like that, even a success would be disastrous.
> > [...]
> >
> > Indeed.
> >
> > But I bet a complete re-write in python would less than 10MB, and take
> > less than a tenth of the time of the original project... 8-)
> >
> > The suggestion of using Jython would be good one - subsystems can be
> > rewritten and tested as time and resources allow.
> >
>
> To translate Java to Python (no idea if it is worthy in this case, but in
a
> more general perpective, this is a problem that makes sense IMHO) , I
wonder
> about this intermediate solution:
> 1. use a parser/generator to translate the Java code to Jython code (with
> eventually some generated 'util' Java classes for the calls to the JDK
> packages and other libraries)
By this I mean Jython classes that are Java specific. The purpose is to
isolate the Jython code that is not Python compliant. The rest of the code
should be Python portable code.

> 2. translate manually the Jython code to Python code if is is really
worthy.
>
> My guess is that step 2 could be a lot simpler to achieve than the whole
> translation and that step 1 can be done automatically (even if the
> development of the generator requires some investment that could be shared
> in an open source project) because :
> - Java and Python are not so far in their concepts,
> - Python is more dynamic than Java and I suppose that this makes it
simpler
> than the opposite (Jython)
> - part of the effort could help anyway (classes to manage the java method
> overridde in Python, ...)
>
> I have no experience in such area, any comments are welcome.
>
>
>
> > --
> > Andrew McNamara, Senior Developer, Object Craft
> > http://www.object-craft.com.au/
> >
>
>




From johnroth at ameritech.net  Sat Apr 13 21:03:25 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 13 Apr 2002 21:03:25 -0400
Subject: Does Python need an 'xor' operator?
References: <3cb88af1@news.mhogaming.com>
Message-ID: <ubhlb81tfqo23e@news.supernews.com>

"Ken Peek" <Ken.Peek at SpiritSongDesigns.comNOSPAM> wrote in message
news:3cb88af1 at news.mhogaming.com...
> I think an 'xor' operator is needed.  The 'xor' operator means
> 'logical-exclusive-or', and returns a 1 or 0.
>
> Here is some code that does what I want (in all the cases I can think
of) for
> 'xor':
>
> # logical-exclusive-or operator:
> def xor(a,b):
>     return (not a) ^ (not b)
>
> This will work, but I think it is less clear (in an expression) than:
'a xor b'.

Write a PEP. I've got no objection, but you will get
the usual slew of objections from people that think
that it is too minor to warrent breaking code by
introducing a new operator.

Also, write it in terms of the new bool type, which
returns the True or False instances.

John Roth
>
>
>




From tanzer at swing.co.at  Sun Apr 14 05:26:43 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Sun, 14 Apr 2002 11:26:43 +0200
Subject: Notes on 2.2's changes
In-Reply-To: Your message of "11 Apr 2002 15:14:58 GMT."
             <slrnabba3h.h8h.akuchlin@crystal.mems-exchange.org>
Message-ID: <E16wgHT-0003Py-00@tswings.swing.cluster>

akuchlin at mems-exchange.org wrote:

> I'd like to provide a point of view on how many changes 2.2 required
> in a sizable amount of code.
>
> The MEMS Exchange codebase I work with is, as of today, 171,437 lines
> of Python code.  Looking through the CVS logs, I find the following
> messages relating to Python 2.2:
(details snipped)
> This means that from my point of view Python *has* been managed for
> stability, and the job has been done reasonably well.

I'd like to reinforce that but I would use `very well` instead of
`reasonably well`.

I'd also like to say a loud and clear **Thank You** to Guido and
everybody helping him. You're doing a terrific job. IMHO, each new
Python version I used (I started in earnest with 1.5) was a definite
improvement over an already excellent language.

I'd like to share my experiences with change in Python. In the
following I'll talk about three applications comprising 391 Python
modules with 83,305 non-blank, non-comment lines (114,612 total
lines). Our development started in 1998 with a team of 2 programmers.
Since then, about 10 different people of very different qualifications
were involved in the development.

Due to various reasons, we often are forced to use an old Python
version. For instance, we shipped our tools with 1.5.2 until last fall
when we finally managed to switch to 2.1 (unfortunately, new Python
versions often appear just around the time when we ship major
upgrades -- must be bad timing on our part :-).

I don't remember any problems in changing from 1.5.2 to 2.1.

I just tried the three applications under 2.2.1. Two of them ran
without any change, for the third I had to fix two bugs (in two places,
the code called a method of a class it didn't inherit from). To be
honest, last summer I had to make a similar change to one of the other
applications to make it run with one of the early alpha versions of
2.2.

Apart from these necessary changes, Python 2.2 also gave 4 warnings
about `from ... import *` inside functions. Looking at those functions
it seems like I'll have to write a bug report -- the functions
shouldn't be part of the module where they live now. After refactoring
the function-level import won't be needed anymore.

Conclusion: Python 2.2 told me about a few errors in my application
some of which had been lurking for a long time.

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





From tdelaney at avaya.com  Tue Apr 30 20:24:00 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 1 May 2002 10:24:00 +1000 
Subject: Is 'isinstance()' the right thing?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A2B9@natasha.auslabs.avaya.com>

> From: Ralf Juengling [mailto:juenglin at informatik.uni-freiburg.de]
> 
> Erik Max Francis <max at alcyone.com> writes:
> 
> > Ralf Juengling wrote:
> > 
> > > So, what is the proper, 'pythonic' way of argument checking?
> > 
> > The "Pythonic" way of argument checking is to simply use it 
> and catch
> > the exceptions that will be thrown if it doesn't behave as planned.
> 
> Hm, you might risk to destroy something this way (i.e. when 
> an exception
> occurs to late). 

Perhaps - but you also have the same problem if something implements an
interface, but doesn't adhere to the contract that is stated for that
interface.

Your unit tests should catch these problems.

Tim Delaney




From sjmachin at lexicon.net  Mon Apr  8 18:06:47 2002
From: sjmachin at lexicon.net (John Machin)
Date: 8 Apr 2002 15:06:47 -0700
Subject: C and Python
References: <3CB19A7E.4090600@rug.ac.be>
Message-ID: <c76ff6fc.0204081406.7d8d425@posting.google.com>

Philippe Faes <Philippe.Faes at rug.ac.be> wrote in message news:<3CB19A7E.4090600 at rug.ac.be>...
> I want to use python for quick prototyping and convert to C/C++ later on.
> Suppose I write some modules (say 2) in python that call eachothers 
> functions.

Short answer: Don't do that.

Consider carefully why you think that the modules (no matter in what
language they are written) should call each other's functions. To me
this is prima facie evidence that some functions are in the wrong
module(s).

When you have such a loop in your call graph, you will be forced
*forever* to consider all the modules in the loop as one module,
because they are interdependent. If they *are* conceptually one
module, then smash them together now. Otherwise, untangle them now.

Try to arrange your modules in layers; for example, "appliance"
modules call "sub-assembly" modules which call "nuts/bolts/tools"
modules.

> Then I port one module to C/C++ and later the second module. 
> In the meanwhile I will want my hybrid project (python-C) to work.
> 
> My question:
> Does any of this already exist?

If you mean a collection of macros or some other band-aid for
supporting switching between C calls and Python calls, I doubt it.
Prevention is better than cure.

> Any hints?

You may have already been planning this: try to write or at least
design an application which makes substantial use of your modules,
while they are still in the Python stage. This would help expose any
kinks in your APIs.

HTH,
John


From joonas at olen.to  Wed Apr 10 05:38:01 2002
From: joonas at olen.to (Joonas Paalasmaa)
Date: 10 Apr 2002 02:38:01 -0700
Subject: Sending a signal to a process started by another user
References: <3CB31C2C.9010503@olen.to> <mailman.1018377258.31626.python-list@python.org>
Message-ID: <11aed8f9.0204100138.20c54afe@posting.google.com>

Oleg Broytmann <phd at phd.pp.ru> wrote in message news:<mailman.1018377258.31626.python-list at python.org>...
> On Tue, Apr 09, 2002 at 04:53:58PM +0000, Joonas Paalasmaa wrote:
> > How can I send a signal to a process started by another user. In this 
> > case my intention is to send a signal 14 to a process started by user jp 
> > from a cgi script. The script raises an "Operation not permitted"- 
> > OSError perhaps because the script is run by user "apache" and the 
> > process by user "jp". How can In make this work properly?
> 
>    Using some form of setuid programs.

Can you give me an example how to use setuid scripts with CGI and Python.


From gashton at cmedltd.com  Wed Apr 10 12:56:00 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 10 Apr 2002 17:56:00 +0100
Subject: how to read hardware info on Win NT
In-Reply-To: <31575A892FF6D1118F5800600846864DCC77B5@intrepid>
References: <31575A892FF6D1118F5800600846864DCC77B5@intrepid>
Message-ID: <1018457760.556.113.camel@tp6>

On Wed, 2002-04-10 at 15:50, Mark Gash wrote:
> import os
> os.system("ipconfig /all > c:\ipdetails.txt")

Only there are issues with os.system interacting strangely with GUIs on
win32 (IIRC) so that might not be ideal for production code. Luckily for
me, I've managed to forget the details....

-- 
Graham Ashton





From ics1010 at yahoo.com  Wed Apr 10 05:40:15 2002
From: ics1010 at yahoo.com (sbp)
Date: 10 Apr 2002 02:40:15 -0700
Subject: Extension Question
Message-ID: <cab7ec6a.0204100140.6b141d16@posting.google.com>

In an extension I'm writing, I have a main import module which works
fine. I'm a little confused about trying to include an object type in
the same dll. From the import module's init I call an init function
for the type which calls Py_InitModule with the name and method array
for the type, but when I try to call a method from this array, I get
an exception stating there is no attribute for the method name I'm
calling. I have functions for the various type protocols (i.e.
attributes) which work fine. If this makes any sense, is it possible
to have a type included within a module with the ability to have
object.function() where function() comes from the array used in the
init? If so, what's the general procedure?

Thanks,
sbp


From tejarex at yahoo.com  Mon Apr  1 01:56:56 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Mon, 01 Apr 2002 06:56:56 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017597529.7660.python-list@python.org> <a87ljh$ofu$1@panix1.panix.com> <3CA75F2F.198772EB@alcyone.com> <8NMp8.184567$2q2.15567243@bin4.nnrp.aus1.giganews.com> <3CA7D34D.F054FB95@alcyone.com>
Message-ID: <YwTp8.188110$2q2.16054941@bin4.nnrp.aus1.giganews.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3CA7D34D.F054FB95 at alcyone.com...
> Terry Reedy wrote:
>
> > Yesterday, I gave four reasons for prefering 'truth' (or something
> > else) to bool.   Please consider them.  Already, my prediction in
> > point D is being borne out in some subsequent posts by some.
> > Something about the word 'bool(ean)' sets off something in some
> > people's brain that makes them want cripple a formalized Python
truth
> > type by prohibiting currently valid and useful behavior.  I find
this
> > as offensive both to the spirit of Python and my programming
freedom
> > as the suggestion that sequence*int should be deprecated and then
> > prohibited because it violates some standard of 'proper' sequence
> > behavior.  Rather than have to put up with and have to argue
against
> > such suggestions for the rest of my life, I would rather switch to
a
> > different term that does not carry with it the baggage that 'bool'
so
> > obviously does.
>
> But "truth" is the wrong word; you mean truth _value_.  Calling the
type
> name "truth" gives the impression that the only value it can have is
> True :-).

OK.  I suggested binary and tval as two other alternatives.

> This seems a semantic argument;

Yes, I did give a semantic argument as one of four yesterday, but what
you quote above from today is a very pragmatic argument based on
obseerved current behavior and predicted future behavior.

> I don't see any compelling reasons for  calling it anything other
than the common name.

I read this proposal a couple of weeks ago and did not think twice
about 'bool' until I saw several people taking use of the common name
as reason to drastically change Python truth values to conform to
their idea/model of what the common name means.

> The reason It doesn't make sense
> to follow other languages' lead _for its own sake_, but it sure does
if
> that is by far and away the most common name for that type of thing.

The problem for some is the difference between the thing itself and
Python's Pythonic version that is something like that thing but
definately not a copy of the thing itself.  So my alternative to
changing the name is awareness that there is and will be this problem
and a suggestion that the documentation and presentation emphasize
that Python truth values are not the same as standard mathematical
Boolean values and are not intended to be and never will be (I hope).
(Anymore than Python lists exactly conform to certain other models of
what a computer language 'list' is 'supposed' to be.)

> Such values are called "Booleans" (an instantly recognizeable name
in
> computer science) far more often than they're called "truth
functional
> values," or "truth values," or (more to the point) "truths."

Given my Fortran (logical), Basic (?), and C (no official name)
pre-Python background, I admittedly find this less compelling than
someone coming from Pascal or whatever.


Terry J. Reedy






From cookedm+news at physics.mcmaster.ca  Tue Apr 23 19:34:59 2002
From: cookedm+news at physics.mcmaster.ca (David M. Cooke)
Date: Tue, 23 Apr 2002 19:34:59 -0400
Subject: using automake/autoconf to build extension module?
References: <mailman.1019509602.30761.python-list@python.org>
Message-ID: <qnkwuuyng4c.fsf@arbutus.physics.mcmaster.ca>

At some point, Skip Montanaro <skip at pobox.com> wrote:

> Can someone explain what automake incantations are necessary to build a
> Python extension module?  I'd prefer not to use distutils if possible, since
> the rest of the project is using automake/autoconf/libtool for build
> configuration.
>
> I tried the following:
>
>     lib_LTLIBRARIES = alfmodule.la
>     alfmodule_la_LDFLAGS = -module -avoid-version -export-symbols-regex initalf
>     alfmodule_la_SOURCES = alf.h alfmodule.c
>     alfmodule_CFLAGS = -I$(prefix)/include/python$(PYTHON_VERSION)
>
> but it blows up during the compile because it has no -I flag referencing the
> include directory.  (Obviously the last line above is incorrect in some
> respect.  It never gets into the cc command.)

Have a look at gnome-python. It uses autoconf/automake. There's an m4
macro included that does a better job than the one included in the
current automake for detecting python and getting the stuff you need.

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke
|cookedm(at)physics(dot)mcmaster(dot)ca


From phr-n2002a at nightsong.com  Wed Apr  3 01:34:05 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 02 Apr 2002 22:34:05 -0800
Subject: PEP 287: reStructuredText Standard Docstring Format
References: <mailman.1017794173.28885.python-list@python.org> <B8D00218.20D57%goodger@users.sourceforge.net>
Message-ID: <7x4rituw82.fsf@ruckus.brouhaha.com>

David Goodger <goodger at users.sourceforge.net> writes:
> > The thing I like least about the PEP is that it's written like
> > marketing material
> 
> PEPs *are* marketing tools.  They market proposals to the Python user
> community and PythonLabs, lobbying for support.  They must be
> persuasive, both technically and expositorially.

I hope PEP's are not marketing tools.  Marketing tools are material
written by vendors to get customers to buy products because they are
the vendors products, rather than because they are what the customer
needs.  If a PEP advocates a particular approach over other
approaches, it should come from having carefully compared the
approaches and finding that the PEP's approach is actually the best.
This PEP doesn't give the impression of careful comparison.

> The PEP briefly examines alternatives proposed in the past, on the
> Doc-SIG, in terms of "the generally accepted goals for a docstring
> format".  I think you're reading too much into the prose as it stands.

Since the PEP is being presented to the whole community and not just
to the Doc-SIG (I'd never heard of the Doc-SIG til just now), it
should make its case in detail rather than just giving the conclusion.
That includes the case for having structured doc strings at all--do
you have formatting tools that read these doc strings and produce
printed output like POD?

> This PEP toots its own horn.  Why should it toot somebody else's?

The whole notion of "tooting somebody else's horn" fits the marketing
notion that where a proposal comes from is more important than its
content.  Marketing material typically tries to sell products by
ignoring or obscuring competing products that might actually be
better.  PEP's shouldn't do that, because PEP's aren't marketing
material.  PEP's should examine all the evidence.

> > Finally, among the explicit markup languages examined in the PEP,
> > Texinfo should be included.
> 
> Texinfo is not included because it has never been seriously proposed
> as a docstring format [#]_.  Looking at samples, I can see why.

It seems to have the same advantages as Javadoc and POD, but be more
readable in source form, which is what PEP 287 wanted.

> .. [#] If you care to propose Texinfo as a standard docstring format,
>    please join the Doc-SIG, make your case there, then write a PEP.

I think it should be mentioned in PEP 287, which after all tries to
list all the reasonable alternatives.  

> The entire premise of this PEP and of the Doc-SIG's (at least) 6-year
> docstring markup odyssey is that Pythonistas want a markup that is
> "readable in source form".  We're spoiled by Guido's programming
> language design aesthetics, and that spilled over into the docstring
> markup debate.  Better late than never.

I've written some pretty big Texinfo documents and find Texinfo to be
readable enough in source form, though it looks better after
formatting.  reStructuredText looks pretty simliar to the output of a
Texinfo formatter--maybe Texinfo could be adapted to produce
reStructuredText instead of Info files.


From bokr at oz.net  Wed Apr 10 16:12:14 2002
From: bokr at oz.net (Bengt Richter)
Date: 10 Apr 2002 20:12:14 GMT
Subject: how to read hardware info on Win NT
References: <mailman.1018450401.5880.python-list@python.org>
Message-ID: <a926au$9pn$0@216.39.172.122>

On Wed, 10 Apr 2002 15:50:20 +0100, Mark Gash <mgash at trisystems.co.uk> wrote:
[...]
>>Thanks for the __winreg hint. Looks like I have to get some infos
>>about the registry because I have no idea how to find the right key.
>>
>>If anyone knows, please tell me.
>>
>>I need the id of my ethernet card for a license documentation tool we
                               ^^^^
It is possible to have more than one, so I assume you want the one
associated with the current host ip? I made a hack building on Mark's advice
which does this. (Of course it could break if ipconfig changes its output format).
See below.

>>want to write in Python. As many commercial softwares are using this
>>id for generating the license keys etc. I thought there should be
>>module in Python.  Maybe there is a modul for commercial distribution
>>of Python software?
>
>Marcus,
>With further investigation, I cannot find where the MAC address of the card
>is in registry (I assume that you require the MAC address when you say that
>you want the id of your Ethernet card).
>
>I am also assuming that you are using a win32 machine.
>
>There are two other ways that I can think of in order to get your info:-
>
>1. Run the command 'ipconfig /all' and pipe the results into a text file
>that you can then search i.e.
>	'ipconfig /all > c:\ipdetails.txt'
Using popen you can avoid having to deal with a file, which I did below.

>
>2. The nbtstat command will report the MAC address of a card, so that for
>example, you could use the command 'nbtstat -a <machinename>' and pipe the
>results out to a text file, i.e.
>	'nbtstat -a mypc > c:\nbtdetails.txt' 
>
>Once the correct command has been tested, then from your python script wrap
>it up in an os.system call.
>
>So, we will end up with something along the lines of :-
>
>import os
>os.system("ipconfig /all > c:\ipdetails.txt")
>
>or 
>
>import os
>os.system("nbtstat -a mypc > c:\nbtdetails.txt")
>
>Then you will have to add some string handling/searching to find the MAC
>address and process it as required.
>
>Hope this helps.
>

You could try this. It seems to work on my NT4:
(You might want to add some error handling though).
It obviously also depends on exact spelling and the IP Address
line coming after the Physical Address line, etc.

===< getMac.py >===
def getMAC():
    import socket, os
    curr_ip = socket.gethostbyname(socket.gethostname())
    ipconfdata= os.popen('ipconfig/all').readlines()
    for line in ipconfdata:
        if line.find('Physical Address') >=0:
            pa = line.split(':')[-1].strip()
        if line.find('IP Address') >=0:
            ip = line.split(':')[-1].strip()
            if ip == curr_ip: return pa, curr_ip

if __name__ == '__main__':
    print '%s is the physical address associated with %s' % getMAC()
===================

Regards,
Bengt Richter


From pinard at iro.umontreal.ca  Mon Apr  8 08:49:13 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 08 Apr 2002 08:49:13 -0400
Subject: Nested scopes hitch
In-Reply-To: <1018261543.690.5.camel@tp6>
References: <a8kn23$ikq$1@zeus.polsl.gliwice.pl>
	<MPG.17179f847bf41a6998969b@news.nwlink.com>
	<1018261543.690.5.camel@tp6>
Message-ID: <oqelhqqrsm.fsf@titan.progiciels-bpi.ca>

[Graham Ashton]

> I remember being slightly surprised by the fact that self finds class
> variables.  Is there any particular reason why, considering that the
> you can reference them as ClassName.var anyway?

Hello, Graham.  I often found it useful, in a hierarchy of classes, being
able to preset a default value in all objects by using a class variable in
a base class, then overriding this default at __init__ time (or otherwise)
in only a few of the derived classes.  This relies on the fact that `self',
everywhere, finds class variables when not overridden.

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




From bokr at oz.net  Sun Apr 21 16:52:43 2002
From: bokr at oz.net (Bengt Richter)
Date: 21 Apr 2002 20:52:43 GMT
Subject: Modules that provide differing functionality amongst different Python versions...
References: <304d20df.0204210629.3f2b4d58@posting.google.com> <0evu9a.621.ln@127.0.0.1> <mailman.1019413960.21990.python-list@python.org>
Message-ID: <a9v8qr$plg$0@216.39.172.122>

On Sun, 21 Apr 2002 20:31:24 +0200, holger krekel <pyth at devel.trillke.net> wrote:

>On Sun, Apr 21, 2002 at 02:12:16PM -0400, Carl Banks wrote:
>> Allan Crooks wrote:
>> > Hi,
>> > 
>> > I was wondering what the best approach is on providing libraries for
>> > varying versions of Python?
>> > (...)
>
>> # Import stuff for version 2.1
>> if sys.version[:2] >= (2,1):
>>     from mymodule2_1 import *
>> 
>> # Import for version 2.2
>> if sys.version[:2] >= (2,2):
>>     from mymodule2_2 import *
>
Wouldn't you want to import from 2.2 first (and perhaps use elifs)?
(though obviously the if-tests themselves must be lowest-version compatible
if done in that order).

>this solves part of the 'code duplication' problem. Might get a bit tricky
>if you just want to define a few additional methods for some classes
>(for example 'generator versions' of methods otherwise returning lists).
>
>Ah, and by the way how is
>
>	sys.version[:2] >= (x,y) 
>
>supposed to work? On my python 2.2 
>
>	sys.version[:2] == '2.'   is True
>
>am i missing something here?
>
>	holger
>
>
He probably meant

 >>> sys.version_info
 (2, 2, 0, 'final', 0)
 >>> sys.version_info[:2]
 (2, 2)

Although 1.5.2 doesn't have this AFAIK, so

 >>> import sys, string
 >>> string.split(sys.version)[0] >= '2.2'
 1

where

 >>> string.split(sys.version)[0]
 '2.2'

might be a better idea.

Regards,
Bengt Richter


From fig at monitor.net  Thu Apr 11 12:03:48 2002
From: fig at monitor.net (Stephen R. Figgins)
Date: Thu, 11 Apr 2002 16:03:48 -0000
Subject: Sound programming?
References: <3CB50DBC.1B0E533F@alum.mit.edu>
Message-ID: <slrnabbcv6.2er.fig@localhost.localdomain>

In article <3CB50DBC.1B0E533F at alum.mit.edu>, Will Ware asks:
> Are there any other options?

I mentioned a couple of others in an october article on the O'Reilly
Network.

  http://www.onlamp.com/pub/a/python/2001/10/25/pythonnews.html

You can also check out Pythonsound,

  http://pythonsound.sourceforge.net/

and a few other csound based modules at listed in the vaults,

 http://py.vaults.ca/apyllo.py?i=63131194


-Stephen


From jon+usenet at unequivocal.co.uk  Tue Apr 23 10:13:34 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 23 Apr 2002 14:13:34 -0000
Subject: Trouble with Myqsl
References: <geaacus7q84htfn8jbakfq4ouo28qnm3d9@4ax.com> <tncx8.93454$T%5.85685@atlpnn01.usenetserver.com> <uloacus4pml1r7f7eac9ts3u20oroj34qb@4ax.com>
Message-ID: <slrnacar0d.kd0.jon+usenet@snowy.squish.net>

In article <uloacus4pml1r7f7eac9ts3u20oroj34qb at 4ax.com>, Fernando Rodr?guez wrote:
>>curs.execute(
>>    """insert into customer_info (email, url, referrer, first_visit)
>>    values ( %s, %s, %s, %s)""",
>>    ('billg at microsoft.com', 'www.easyjob.net', 'www.google.com', '2001-03-07
>>12:32:00')
>>)
> 
> mysqldb.error: (1064, "You have an error in your SQL syntax near
> '@microsoft.com, www.easyjob.net, www.google.com, 2001-03-07 12:32:00)' at
> line 1")

Woah, you have *seriously* screwed something up there. The code shown
is correct. I suggest you post the earlier part of your code where
your are connecting to the database and creating 'curs'.

Are you sure you really are doing dbc.execute(foo, bar) and not
dbc.execute(foo % bar)?


From DeepBleu at DeepBleu.org  Mon Apr  8 12:20:28 2002
From: DeepBleu at DeepBleu.org (DeepBleu)
Date: Mon, 8 Apr 2002 11:20:28 -0500
Subject: problem zipping a text file with zlib
Message-ID: <ub3gp3lfn2rc3d@corp.supernews.com>

I am using a python script and the zlib module to zip two text files into
two separate zip files.  They are zipping fine.  However, when I use winzip
to open the two zip files, one of them is opening fine.  The other one is
reporting that it is an invalid archive with two missing Bytes.  The two
text files are just two regular text files.  What are the conditions that
cause a zip file to fail due to missing bytes?
Thanks,
DeepBleu




From netzapper at magicstar.net  Fri Apr  5 15:13:43 2002
From: netzapper at magicstar.net (A. Jones)
Date: Fri, 05 Apr 2002 20:13:43 GMT
Subject: file.getline()/cout problem
Message-ID: <3cae0463.46737479@mammoth.usenet-access.com>

In trying to understand how file access might work in my instance, I
wrote a small test program with the expectation that it would print
off 35 lines of arbitrarily length of a file.

However, it's reading off a certain amount, and then abruptly ending
in the middle of a line.  The loop continues to run, but the strings
are apparently empty (when I put in a constant to be printed after the
string, it printed 35 times).

I was under the impression that because of the limited scope of
functions, the sTeP and teststring variable would flush each time
(while the position in the file was recorded in the global testfile
variable), and so, the buffer and string would not overflow.  Was I
incorrect?

Below is the code:


#include <iostream.h>
#include <string.h>
#include <fstream.h>

ifstream testfile("QOD.qdb");

string stringtest(void) {
string teststring;
char sTeP [1024];
testfile.getline(sTeP, 1024);
teststring.append(sTeP);
return teststring;
}

int main ()
{
	char holder;
	cout << "This is a test: ";
	for(int r=0; r<=35;r++) {
	cout << stringtest();
	}
	return 0;
}


-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 100,000 Newsgroups - Ulimited downloads - 19 servers ==-----


From huaiyu at gauss.almadan.ibm.com  Wed Apr 10 21:04:02 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 11 Apr 2002 01:04:02 +0000 (UTC)
Subject: type() new style class instance says "class", not "ObjectType"
References: <3cb4bd45$1_1@goliath.newsgroups.com>
Message-ID: <slrnab9o7v.i8b.huaiyu@gauss.almadan.ibm.com>

On Wed, 10 Apr 2002 18:38:47 -0400, Brad Clements <bkc at Murkworks.com> wrote:
>>>> class x(object):
>...  pass
>...
>>>> y = x()
>>>> type(y)
><class '__main__.x'>

Is this a bug?  Is x a meta class?

>>> y.__class__ 
<class '__main__.x'>
>>> z = y()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: 'x' object is not callable

Is y a class?  I'm puzzled ...

>
>>>> type(x)
><type 'type'>

This is OK:

>>> isinstance (x, TypeType )
1

Huaiyu


From op73418 at mail.telepac.pt  Mon Apr  1 17:41:31 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 01 Apr 2002 23:41:31 +0100
Subject: unicode string problems
Message-ID: <b9ohaus72ktk4huk3o35cfubj60agvvlou@4ax.com>

My problem is the following:

Let f be a file object. If I do

f.write("Mar?o 2002")

all goes well. But if I try

f.write("Mar?o 2002" + march.Name())

where march.Name() returns a unicode string I get a unicode error. I
tried converting both unicodes to strings via str but obviously I got an
error (in the first string the culprit is the "?" character).

Can someone help me out here and show me the way to write these strings
to the file?

Thank for everything, best regards,
Gon?alo Rodrigues



From terabaap at yumpee.org  Tue Apr 30 23:47:13 2002
From: terabaap at yumpee.org (Manoj Plakal)
Date: Tue, 30 Apr 2002 22:47:13 -0500
Subject: Pychecker plugins?
Message-ID: <3CCF6541.50402@yumpee.org>

Pychecker gurus,

Are there any plans for allowing user-written plugins for Pychecker?

Right now, it seems that Pychecker traverses Python code and
performs a fixed list of checks. It would be cool to expose
a plugin API so that users can write their own specific
checkers e.g., to confirm with local style guides, and not
have to burrow into Pychecker source code to do this.

And perhaps this could open the door to more than just checking,
one can imagine plugins transforming the traversed code to
implement profiling, or pre-post-condition-checking etc.
Bordering on what one might do with metaclasses or aspect-oriented
programming. Having Pychecker be a generic Python code-object
traverser-cum-transformer would be powerful.

Manoj



From bvdpoel at uniserve.com  Tue Apr 16 20:28:06 2002
From: bvdpoel at uniserve.com (bvdpoel at uniserve.com)
Date: Tue, 16 Apr 2002 17:28:06 -0700
Subject: preferred syntax for list extend?
Message-ID: <3CBCC196.466476CE@uniserve.com>

Just doing a bit of fooling around and I've found that if:

	q1=[1,2,3,4]
	q2=[5,6,7,8]

then:
	
	q1 += q2

is about 15% faster than

	q1.extend(q2)

Mind you, my test was with a relatively short list, don't know if longer
lists would act differently.

I would have thought that using extend() would be preferred and faster?

-- 
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bvdpoel at uniserve.com
WWW:   http://users.uniserve.com/~bvdpoel



From jeremy at alum.mit.edu  Wed Apr  3 18:54:03 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 3 Apr 2002 15:54:03 -0800
Subject: Bastion Documentation Error?
References: <a8dsj3$1n4j$1@news.idiom.com>
Message-ID: <b0f083db.0204031554.48fe1bdd@posting.google.com>

jimd at vega.starshine.org (Jim Dennis) wrote in message news:<a8dsj3$1n4j$1 at news.idiom.com>...
> I'm a bit confused by the Bastion.Bastion() function.
>  From the documentation I read:
> 
>  """
>  Protect the object object, returning a bastion for the object. Any
>  attempt to access one of the object's attributes will have to be
>  approved by the filter function; if the access is denied an 
>  AttributeError exception will be raised. 
>  """
> 
>  ... but playing with it I find that it raises an AttributeError on 
>  any attempt to access any class or instance variables.  It always 
>  seems to call the filter function, then (regardless of the results) 
>  it seems to then check to see if the attribute is a method and block 
>  access even if the filter function returned a true value.
> 
>  If that is the intent (that only *methods* can be exposed through
>  regardless of the filter function) then I think this should be made 
>  more clear in the documentation.

It's a bug in the documentation.  Could you submit a patch with the
new words?  Or at least a bug report?

Jeremy


From phd at phd.pp.ru  Fri Apr 26 03:59:08 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 26 Apr 2002 11:59:08 +0400
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
In-Reply-To: <7xsn5jat20.fsf@ruckus.brouhaha.com>; from phr-n2002a@nightsong.com on Thu, Apr 25, 2002 at 11:10:47PM -0700
References: <mailman.1019526833.9592.python-list@python.org> <3CC61DCA.DAD9DD8E@cosc.canterbury.ac.nz> <slrnacgvik.irb.chasm@galileo.rift> <3CC8D767.599720@cosc.canterbury.ac.nz> <7xsn5jat20.fsf@ruckus.brouhaha.com>
Message-ID: <20020426115908.A24848@phd.pp.ru>

On Thu, Apr 25, 2002 at 11:10:47PM -0700, Paul Rubin wrote:
> As a free software author, I'll consider stopping using copyright if
> all the other software publishers out there are willing to do the
> same.  Until then, I'll stay with the GPL.

   I cannot say better! I fully and wholeheratedly agree!

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




From cabrera at wrc.xerox.com  Fri Apr 12 09:00:55 2002
From: cabrera at wrc.xerox.com (jic)
Date: 12 Apr 2002 13:00:55 GMT
Subject: ftplib and FTP Objects...
References: <a94i7u$6id$1@news.wrc.xerox.com> <a96goi$oe9$1@news1.wdf.sap-ag.de>
Message-ID: <a96lq7$e38$2@news.wrc.xerox.com>

In article <a96goi$oe9$1 at news1.wdf.sap-ag.de>, daniel says...


>It seems as if the FTP server doesn't know the PASV command. You could try
>def set_pasv(self, val):
>        '''Use passive or active mode for data transfers.
>        With a false argument, use the normal PORT mode,
>        With a true argument, use the PASV command.'''
>to use normal instead of passive mode.
>
>You should also pass the open file to storbinary instead of the read method.

Good call.  Never even thought about that.  I'll try it!  Thanks.

jos?



From nospam at bigfoot.com  Fri Apr 26 11:39:47 2002
From: nospam at bigfoot.com (Gillou)
Date: Fri, 26 Apr 2002 17:39:47 +0200
Subject: XML and Python
References: <3cc812a3$1_1@nopics.sjc>
Message-ID: <aabrt2$3im$1@norfair.nerim.net>

Yes,

The docs are not "newbie friendly", and you need to have some DOM and SAX
skills before going into pyXML docs.
Buying this book will save you lots of time if you're an XML newbie.

http://www.oreilly.com/catalog/pythonxml/



From emile at fenx.com  Fri Apr 26 22:13:08 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 27 Apr 2002 02:13:08 GMT
Subject: looping through a list
References: <e2e9e807.0204261611.7514b387@posting.google.com>
Message-ID: <UOny8.60700$CH.9783@rwcrnsc52.ops.asp.att.net>

Hugh writes:
> result = c.fetchone()

Try this in the interpreter.  I'd expect (but don't know for sure) that
you're getting a list back...

>   if (result != ""):

If so, you'll always pass through this...

... and if you're getting a string you could be saying ' if result:'

>     i=0
>     text = '<tr width = "80%" align = "center">'

you may want to be using a list here instead of a string.  appending to
lists is more efficient than concatenating strings.

>     while (i != len(result)+1):

this could be ' for returneditem in result:'    assuming a list result.

>       i = i + 1
>       text = text + '<td align = "center" valign =
> "top">'+str(result[int(i)])  +'</td>'

you certainly don't need int() here.

>     text = text + "</tr>"

this probably wants to be
text.append('''<td align = "center" valign ="top">%s</td></tr>''' %
returneditem)

>     page.generate("","",sessionID,text)
>   else:
>     text = '<h4>No items found</h4>'
>     page.generate("","",sessionID,text)
>
>
> At the moment, it keeps saying list index out of bounds and I don't
> know where I'm going wrong. I've tried all sorts of things in the
> While statement and still can't get it going.


You're probably going wrong by adding 1 in testing

while (i != len(result)+1):

as i subsequently indexes into result.  In python, indexes end one shy
of the length.

HTH,


--

Emile van Sebille
emile at fenx.com

---------



From lenssen at hitnet.rwth-aachen.de  Sun Apr 21 15:36:43 2002
From: lenssen at hitnet.rwth-aachen.de (Philipp Lenssen)
Date: Sun, 21 Apr 2002 21:36:43 +0200
Subject: Safe eval?
References: <a9i2kg$cg8$1@nets3.rz.RWTH-Aachen.DE>
Message-ID: <a9v4cs$fao$1@nets3.rz.RWTH-Aachen.DE>

"Philipp Lenssen" <lenssen at hitnet.rwth-aachen.de> wrote in message
news:a9i2kg$cg8$1 at nets3.rz.RWTH-Aachen.DE...
> Is it possible to have a safe eval() on the server-side even when the
string
> consists of submitted user data?
>..

Since some people pointed me to rexec, my current function looks like this:

    def safeEval(self, s):
        return rexec.RExec.r_eval(rexec.RExec(), s)

(That's the only way I got it to work since I didn't get the example at
http://www.python.org/doc/lib/node547.html .)

Now my question is, how do I allow my own functions* contained in the same
class? The eval-String contains these as "self.myFunction(1, 2)", which
works fine using normal eval().
I'm sure I have to do something with ok_builtin_modules, but I don't know
what.

Before I used rexec, I had a normal eval() but did the following before,
since somebody pointed out those strings (I guess with rexec it's not needed
anymore):
    s = string.replace(s, "__", "")
    s = string.replace(s, "file", "")
    s = string.replace(s, "eval", "")

Thanks for any help.


*For sample purpose, allowed functions are:
    qml_random(min, max), qml_upper(s), qml_lower(s)




From stian.jenssen at nrk.no  Wed Apr 17 07:21:24 2002
From: stian.jenssen at nrk.no (Stian Jenssen)
Date: Wed, 17 Apr 2002 13:21:24 +0200
Subject: client timeout!
Message-ID: <D186E3C450FED2119AFA00508B08736F040D996A@maexch05.nrk.no>

Hi all!

Have a little problem. 
Have a script that runs on a ms server. 
It parses other machines for information, but 
when it don't get an answer from the other end it halts.

I want to bypass this problem with a shorter timout, in the .lib file.

Have anybody had a similar problem.
What file do i have to modify (urllib.py, httplib.py or other) and how?

Stian






From andrew.gregory at npl.co.uk  Fri Apr 19 11:21:59 2002
From: andrew.gregory at npl.co.uk (Andrew Gregory)
Date: 19 Apr 2002 08:21:59 -0700
Subject: Keyboard interrupts in Windows Python
Message-ID: <2830c89c.0204190721.633cd445@posting.google.com>

I am using Python to control experiments from the interactive Window
of Pythonwin (via some simple library routines in C to write data to
interface chips). Sometimes I wish to interrupt an experiment before
it has finished, ideally by using Control C to raise an exception. But
Control C has no effect.
How can I overcome this problem? Is there another control key which I
should be using?

It would be equally acceptable to raise an exception by clicking on a
button - but I am hoping to avoid Windows programming and work
interactively for the moment.

To give an example, if I typed time.sleep(1000) how would I terminate
this command under Windows?

Andrew Gregory, National Physical Lab., UK.


From grante at visi.com  Mon Apr  1 14:48:07 2002
From: grante at visi.com (Grant Edwards)
Date: Mon, 01 Apr 2002 19:48:07 GMT
Subject: GUI programming
References: <mailman.1017687607.4322.python-list@python.org>
Message-ID: <XP2q8.8178$vm6.1392106@ruti.visi.com>

In article <mailman.1017687607.4322.python-list at python.org>, Gustavo Cordova wrote:

>> [..references to building wxGTK problems..]
>> All in all, this can be very frustrating to new users.
> 
> Extremely.
> 
> I gave up after the n-th iteration.

It took me 5 or 6 tries, but I did get it to build and install.
One of the things I found odd was that by default it would try
to build stuff (GL, etc.) for which my system had no support.

Those of use who are used to building stuff from source have
gotten used to some sort of auto-configure operation that
avoids having to go in and manually disable the extras for
which we don't have the appropriate libraries.

I also had problems due to the version of the GCC I had
installed, disagreement between versions of wxGTK and wxPython,
and probably something else I'm forgetting.

-- 
Grant Edwards                   grante             Yow!  TAILFINS!!...click...
                                  at               
                               visi.com            


From bac at OCF.Berkeley.EDU  Mon Apr  1 19:37:19 2002
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Mon, 1 Apr 2002 16:37:19 -0800
Subject: PEP 285: Adding a bool type
In-Reply-To: <a89qqt$b2a$1@panix1.panix.com>
References: <mailman.1017466974.6751.python-list@python.org>
 <idVp8.48143$S52.1619383@news2.tin.it> <a89qqt$b2a$1@panix1.panix.com>
Message-ID: <Pine.SOL.4.44.0204011633190.12842-100000@death.OCF.Berkeley.EDU>

On 1 Apr 2002, Aahz wrote:

> In article <idVp8.48143$S52.1619383 at news2.tin.it>,
> Alex Martelli  <aleax at aleax.it> wrote:
> >
> >        ...
>
> I've been reading this thread before deciding how to vote.  Before Alex's
> post, I was somewhere between +0 and +1; now I'm -0.  I have also seen
> some suggestions for postponing this change to 2.4; that gets +0 from me.

I personally had already posted a +0 vote, but after readings Alex's
wonderful post, I change mine to -1.  Perhaps booleans would work better
as a module in the standard library if people feel the need to have them.
Either that or we can start pushing operator.truth() more (which I didn't
even realize existed until this thread started).

And as for postponing it, I think that is a VERY good idea.  It would be
nice to have a release where I don't need to think about using any
__future__ imports.

-Brett C.


> --
> Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/
>
> Why is this newsgroup different from all other newsgroups?
>



From theller at python.net  Thu Apr  4 12:03:46 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 4 Apr 2002 19:03:46 +0200
Subject: NTLM authorization - compile into EXE
References: <9b309481.0204031300.70ce3c63@posting.google.com>
Message-ID: <a8i11o$sm9k9$1@ID-59885.news.dfncis.de>

"Marc Scheuner" <> wrote in message news:9b309481.0204031300.70ce3c63 at posting.google.com...
> Folks,
>
> I'm faced with the need to deal with an MS ISA proxy server. This
> proxy users NTLM Authorization, which basically no internet tool
> supports - except MS IE......
>
> I downloaded the Python 1.5.2 source from here:
> http://www.geocities.com/rozmanov/ntlm
>
> and I also downloaded Python 1.5.2 itself - everything works okay.
>
> But under Windows, having an extra command line window open just for
> this NTLM proxy thingie is a bit "annoying" - so I thought I'd compile
> it into an EXE, using Py2EXE. Downloaded all that stuff, too, and went
> through all the steps recommended (distutils stuff and the like) - but
> I still can't seem to compile that Python code into a Windows
> EXEcutable.
>
> The error message I get is very cryptic - at least to me as a Phyton
> newbie. Don't really know what it tries to tell me....
>
> ---------------
> File "C:\Program Files\Python\py2exe\resources\VersionInfo.py", line
> 56, in ?
>   raise ImportError, "Could not import StringTables, no unicode
> available"
> ImportError: Could not import StringTables, no unicode available
> ---------------

Marc,

looks like a bug in py2exe ;-)
This patch to py2exe\build_exe.py should fix the bug, can you try it?

Thanks,

Thomas

Index: build_exe.py
===================================================================
RCS file: /cvsroot/py2exe/py2exe/py2exe/build_exe.py,v
retrieving revision 1.4
diff -c -r1.4 build_exe.py
*** build_exe.py 20 Mar 2002 09:54:44 -0000 1.4
--- build_exe.py 4 Apr 2002 16:56:01 -0000
***************
*** 838,845 ****
          # "legalcopyright legaltrademarks originalfilename productname"
          # "productversion"

!         from resources.VersionInfo import VS_VERSIONINFO, RT_VERSION, \
!              StringFileInfo, VarFileInfo, VersionError
          from py2exe_util import add_resource

          md = self.distribution.metadata
--- 838,850 ----
          # "legalcopyright legaltrademarks originalfilename productname"
          # "productversion"

!         try:
!             from resources.VersionInfo import VS_VERSIONINFO, RT_VERSION, \
!                  StringFileInfo, VarFileInfo, VersionError
!         except ImportError, details:
!             self.warn("%s\n  No VersionInfo will be created" % details)
!             return delete
!
          from py2exe_util import add_resource

          md = self.distribution.metadata





From kseehof at neuralintegrator.com  Sat Apr  6 15:32:16 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 6 Apr 2002 12:32:16 -0800
Subject: PEP 285: Adding a bool type 
In-Reply-To: <200204061913.g36JDcdj017764@ratthing-b246.strakt.com>
Message-ID: <MEENLMFCHGAOOICLMCHHMECPDLAA.kseehof@neuralintegrator.com>

> Thank you very much for the reply.
> I just want to correct something you ascribed to me.
> 
> > I guess we'll have to agree to disagree here: I don't think that bools
> > are bad, either Python's variety or any other language's kind.  I've
> > seen your arguments, and they just don't convince me.  So be it.
> 
> I think there is a time and a place for bools --  Describing things
> that have two states that can never, ever, have 3 or more states.  
> These things are vanishingly close to non-existent, unless you are
> a mathematician in which case you may be playing with these sorts
> of things all the time.  What I am trying to do is save people
> like Mark McEahern from his isValid function, which returns whether
> a username is valid or not.  On the day his boss demands a third
> value - valid-but-I-won't-let-you-in-anyway-you-didn't-pay-your-bill,
> he will thank me.

There is nothing wrong with isValid being boolean.  A user name is
either valid or it is not valid.  Whether the user has paid his bill
or not should be a separate variable.  Stuffing billing information
into the isValid function is a horrible idea, and in an existing code
base will almost certainly cause major hard-to-find problems.

> That is what I mean by teaching people not to use bools -- teaching
> when not to use them. This is very hard to teach.  I know it because
> I have been doing it for years and years.   I even know _why_ it is
> hard to teach, but that doesn't make it any easier.
> 
> Laura

It sounds to me like it may be difficult to teach because you are
using the wrong definition of boolean.

Bad definition: "A boolean represents exactly two possible values"
Good definition: "A boolean represents something that is or is not true"

Any name that starts with "is" strongly implies a boolean because things
either is or they ain't. :-)  If something looks boolean but ends up
containing a 2, you will have bugs, guaranteed.  Even after you patch
up the bugs, your code will be very confusing. Teaching when not to
use booleans can't possibly be any more difficult than teaching people
when not to prefix a function or variable with "is".

How about my_animal==your_animal returning a 2 if the animals are
different species but have identical noses, and a 3 if they like
each other?

- Ken Seehof





From starx at pacbell.net  Thu Apr 25 13:33:18 2002
From: starx at pacbell.net (Philip Swartzleonard)
Date: 25 Apr 2002 17:33:18 GMT
Subject: What is prefered: 'string' or "string"? (Should be in PEP8 ?)
References: <aa3dti$n0$1@news.nextra.cz> <aa3fbk$pfb$1@swifty.westend.com> <aa5f7l$uck$1@news.nextra.cz> <jwbaxter-B5D4A7.19244624042002@corp.supernews.com> <aa8f00$50o$1@swifty.westend.com>
Message-ID: <Xns91FB6B9F020B5RASXnewsDFE1@130.133.1.4>

Philipp Lenssen || Thu 25 Apr 2002 01:33:56a:

> "John Baxter" <jwbaxter at spamcop.net> wrote in message
> news:jwbaxter-B5D4A7.19244624042002 at corp.supernews.com...
>>..
>> I use single quote (that is, apostrophe) to avoid a shift key
>> manipulation.  Unless I forget.  And of course if the string begins
>> with a shifted character, the manipulation is deferred, not avoided.
>>..
> 
> I need to press Shift to get a single quote. Or double-quote,
> greater-than, and $.
> I'm not sure what is exclusive german keyboard layout here.

Hm, well, single-quote is unshifted on a standard US keyboard. Double 
quote is shift + single quote. Greather-then is shit + period (and less 
than is shift + comma)). $ is shift + 4.

-- 
Philip Sw "Starweaver" [rasx] :: www.rubydragon.com


From aleax at aleax.it  Sun Apr 14 14:52:37 2002
From: aleax at aleax.it (Alex Martelli)
Date: Sun, 14 Apr 2002 20:52:37 +0200
Subject: PyQt requires a "Mandatory" 'from qt import *' ...?
In-Reply-To: <3CB9C4B4.DC454AA8@river-bank.demon.co.uk>
References: <oOjt8.8815$vF6.254710@news2.tin.it> <3CB9C4B4.DC454AA8@river-bank.demon.co.uk>
Message-ID: <02041420523700.28331@arthur>

On Sunday 14 April 2002 20:04, Phil Thompson wrote:
	...
> The fix has now been applied to CVS. I will probably release 3.2 in
> about a week.

My compliments for your speed!


Alex




From tim.one at comcast.net  Thu Apr  4 21:40:15 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 04 Apr 2002 21:40:15 -0500
Subject: [Python-Dev] ACCEPTED: PEP 285 (truth pickles)
In-Reply-To: <roy-857582.08132604042002@news1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAEOLAA.tim.one@comcast.net>

[Roy Smith]
> If I pickle a bool, and then restore it, do I get the same singleton
> back?

Yes.  However, booling a pickle is a different question entirely.

don't-even-think-about-it-ly y'rs  - tim




From mal at lemburg.com  Tue Apr  2 12:49:24 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 02 Apr 2002 19:49:24 +0200
Subject: relational database?
References: <pRzp8.4373$hi7.15765@news-server.bigpond.net.au> <newscache$d8oxtg$iv4$1@www-neu.dzsh.de> <23891c90.0204020644.3ea4bc40@posting.google.com>
Message-ID: <3CA9EF24.2B9A0C27@lemburg.com>

Paul Boddie wrote:
> 
> "Ruediger Maehl" <ruediger.maehl_nospam at web.de> wrote in message news:<newscache$d8oxtg$iv4$1 at www-neu.dzsh.de>...
> >
> > "Trung Hoang" <thoa0025 at mail.usyd.edu.au> schrieb im Newsbeitrag
> > news:pRzp8.4373$hi7.15765 at news-server.bigpond.net.au...
> > > What relational database is easy to setup and use with python? besides
> > > msql (im having some problems)
> >
> > SAPDB, a full featured database for free at www.sapdb.org
> > with Python interface.
> 
> I'm not sure I would second that recommendation unless SAP DB has
> changed dramatically over the past few weeks. I found that the
> installation was quite a bit more complicated than, for example, MySQL
> (at least on Windows) and upon issuing the arcane commands for
> restarting database instances, the software would complain that the
> instances needed restarting, which was hardly very reassuring with
> respect to either the stability or the usability of the software.

Oh, SAP DB is quite usable. In fact, we switched from MySQL
to SAP DB for most of our work. SAP DB is fast, full-featured,
open-source and very stable. It does have it's share of bugs, but
then so do MySQL and all the others.

> As for the Python DB-API module, I couldn't get any version of it I
> found to work with the database system. Admittedly, my objective was
> to get a JDBC connection up and running, which was eventually
> successful, and I therefore cut that particular side-investigation
> short.

You should probably try zxODBC which is a JDBC-based version
of mxODBC written by Brian Zimmer. It's part of Jython since
version 2.1.
 
> Despite it's often-cited shortcomings, MySQL and the MySQLdb module
> are quite easy to set up. On UNIX systems, PostgreSQL and pgPySQL are
> possibly as easy, although one doesn't need to suffer a manual with
> PostgreSQL which consists of a single >1.5MB HTML file. I've also had
> success with Sybase Adaptive Server Anywhere and mxODBC (without a
> driver manager), but that software is commercially licensed and may
> not even be available any more.
> 
> Another interesting issue that one might want to consider later on is
> Unicode support with database systems and Python DB-API modules. Some
> modules don't seem to support Unicode (pyPgSQL, for example), whereas
> others apparently rely on the database system for such support to
> actually function (eg. mxODBC). Some people may find this an
> irritation earlier than others, however.

mxODBC is certainly well alive. Note that Unicode support in mxODBC
is available both as emulation (mxODBC does the conversion to e.g.
UTF-8) and using native support (the ODBC driver has to support
Unicode).

MS SQL Server and MS Access both support native Unicode mode
via mxODBC.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/



From gangli at msn.com  Tue Apr  2 16:48:00 2002
From: gangli at msn.com (gang li)
Date: 2 Apr 2002 13:48:00 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org>
Message-ID: <9dfc16ab.0204021348.67599701@posting.google.com>

Although I always respect your brilliant ideal about Python language
design, I have to say no to PEP 285.

When I just started to use Python, I feel a little uncomfortable
missing TRUE/FALSE since I am mainly a C/C++ programmer.  However the
missing TRUE/FALSE turned into a great pleasure for me in just a few
days after I really speed up on Python.  I really like Python's way to
tell right/wrong, true/false, and full/empty
 Every object in Python
can have Boolean meanings (None/not None), not just Boolean type
object can. That is beauty of Python.  Like ancient Chinese tall.
"World born from nothing to have two parts, e.g. male-principle and
female-principle, that two parts produced the current world, so every
thing inheritances from one of the parts". The Boolean meaning is
coming with every object. We do not need Boolean type to get it. Laura
Creighton (lac at strakt.com) has a very good definition for it,
something/nothing.  Don't add Boolean to Python to make it look just
like another programming language

Guido van Rossum <guido at python.org> wrote in message news:<mailman.1017466974.6751.python-list at python.org>...
> I offer the following PEP for review by the community.  If it receives
> a favorable response, it will be implemented in Python 2.3.
> 
> A long discussion has already been held in python-dev about this PEP;
> most things you could bring up have already been brought up there.


From aleax at aleax.it  Wed Apr  3 08:38:53 2002
From: aleax at aleax.it (Alex Martelli)
Date: Wed, 03 Apr 2002 13:38:53 GMT
Subject: [Python-Dev] Re: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xpu1h4h1m.fsf@ruckus.brouhaha.com> <a9xq8.53684$S52.1860202@news2.tin.it> <U73UpVACBvq8EwIz@jessikat.fsnet.co.uk>
Message-ID: <NBDq8.46279$pT1.1312341@news1.tin.it>

Robin Becker wrote:
        ...
> The great designer has several times expressed his desire to cripple
> python's more dynamic features. Many recent changes have been in the
> wrong direction eg the restriction on imports in functions etc.

Any restriction on "from wherever import *" and other namespace-polluting
things is very much in the RIGHT direction -- and has nothing to do with
"crippling Python's more dynamic features".

While Python scales quite well up and down the range of program sizes,
the size ranges of most interest to me are scripts of a few tens of
SLOCs, modules of about 100-1000 SLOCs, packages of about 1000-5000
SLOCs, subsystems of a few tens of thousand SLOCs, applications of
about 100 to 500 thousand SLOCs.  SLOC=Source Line Of Code, roughly
the equivalent of a Python "logical line" with common sense adjustmenta
("if x: print y" being 2 SLOCs whether you break the line or not:-).

These are *Python* SLOCs, roughly 15 statements/FP by my estimate
(Capers Jones might disagree, as he estimates 21 statements/FP for
Perl, but I also evaluate Perl at roughly 15 stat/FP on my own code),
versus a range of around 40 to 120 statements/FP for C++ depending
on how the latter is used (C.J. estimates 53 for C++ and Java, 128 for C); 
let's say you need to multiply by a factor between 3 and 8 (most
likely around 5) to get roughly equivalent program/module size
ranges for "typical" C++ or Java.

Substantial reductions in dynamic functionality would substantially
reduce language-level (increase statements per FP).  But at the larger
size ranges (as opposed, maybe, to the small scripts), namespace
pollution is a SERIOUS problem.  eval or (shudder) exec without
specified dictionaries, 'from blah import *', tampering with
__builtins__ -- each is a hard-to-debug subtle interaction just
waiting to bite.  Python has never had some of the worst features
in this regard, such as the ability to modify existing built-in
classes a la Smalltalk/Ruby -- superficially very attractive, but
really requiring lots of programmer discipline to avoid.  Thanks be.


> I believe the fundamental choice here is more about style than substance
> as it's clear this addition adds little and effects much. Of course

Hard to disagree on this one (except maybe s/effect/affects/ --- a
subtle nuance of meaning, but an important one:-).

> there are similarities to the 'only one way to do things' syndrome in
> this proposal.

Nothing wrong with "one obvious way, and preferably only one".  It's
an ideal, of course; as long as you need to ensure that a+b and b+a
give the same results for numbers a and b, you can't really approach
that ideal all that closely, for example (well, one COULD add a
syntax constraint about additions of variables having to have the
variables in alphabetical order, I guess -- and maybe emit a warning
[suppressable on request] otherwise...).


> If we slavishly give people dictator status we get dictators. GvR
> designed a reasonable language and it has improved up to now. I'm fairly
> sure adding this will not improve things.

I'm also "fairly sure" on the specifics -- but I'm even surer that,
whatever the technical and conceptual issues in one specific case,
*overall* GvR has done (and will probably keep doing) a better job
than I could have done, or that any committee could have done.  So,
I'd rather see things continue this way, than have him give up the
decision-making burden to any committee or any other individual.

Even though some more consensus-management, diplomatic, and rhetorical
skills, might well help the medicine go down in a most delightful way.
Despite our bruised egos, and the justified bitter after-taste for being 
ASKED to provide arguments and then REPROACHED for having actually done so,
"the language's the thing" -- I'd rather have a great language and some
ego bruises, than a "just pretty good" language and a well-soothed ego:-)


> GvR asked for opinions and got them. The most articulate and compelling
> are against. Well we got told that these are FUD etc etc. My gut feeling
> is that this is wrong and that lambda, map, etc are good. Perhaps I
> should eat sour grapes for breakfast.

Our guts may not be the best guide at the moment:-).  And anybody who
things lambda (as present in Python) is GOOD thinks thoughts just too
different from mine to allow easy communication, anyway:-).


Alex



From gashton at cmedltd.com  Wed Apr 24 05:19:43 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 24 Apr 2002 10:19:43 +0100
Subject: Detecting a new style class
In-Reply-To: <aa48nq$a61$1@panix1.panix.com>
References: <mailman.1019552082.8541.python-list@python.org> 
	<aa48nq$a61$1@panix1.panix.com>
Message-ID: <1019639984.614.8.camel@tp6>

On Tue, 2002-04-23 at 19:21, Aahz wrote:
> In article <mailman.1019552082.8541.python-list at python.org>,
> Graham Ashton  <gashton at cmedltd.com> wrote:
> >
> >>>> class A(AbstractSingleton, object):
> >...   pass
> 
> Maybe you should tell us a bit more about what you're trying to do so
> that you needn't rely on crufty internal stuff to get the correct result.

Okay, good point. I'm using the Alex's Borg pattern in place of a
singleton, as shared state will do quite nicely. I'm also trying to
protect myself from getting caught out by new style classes (in other
words it doesn't work -- in general -- for classes derived from, for
example, dict).

So, until I need singletons that support subclasses of dict in all their
glory, I decided to do the "simplest thing that will possibly work" and
raise an exception if I forgot about the limitation (hence the need to
detect whether it's been inherited by a new style class).

An alternative question would be -- does anybody know how to do the Borg
so that it works for new style classes?

P.S. Thanks to those who replied to my earlier posts -- both solutions
work nicely.

-- 
Graham Ashton





From phr-n2002a at nightsong.com  Tue Apr 16 01:17:14 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 15 Apr 2002 22:17:14 -0700
Subject: Python webmail packages?
References: <7xr8lhb54s.fsf@ruckus.brouhaha.com> <7876a8ea.0204151614.138853be@posting.google.com> <mailman.1018918474.3622.python-list@python.org>
Message-ID: <7xsn5wjk85.fsf@ruckus.brouhaha.com>

Ian Bicking <ianb at colorstudy.com> writes:
> BoboMail works okay.  It uses some portions of an old version of Zope
> (hence it's named after that old version -- actually, it might be new
> versions of those pieces, I don't know) -- but it's actually CGI-based. 

Hmm, how much code is involved?  I didn't want to deal with installing
a huge package like Zope.


From jb at cascade-sys.com  Tue Apr 30 19:47:52 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Tue, 30 Apr 2002 16:47:52 -0700
Subject: Comment on Draft Pep ver 4 -- Psuedo Deprecations of Builtins
References: <aan5hv$aor$1@bob.news.rcn.net>
Message-ID: <3CCF2D28.B1A43D42@cascade-sys.com>

Raymond Hettinger wrote:

> Reading all of the posts on the deprecation pep shows that there is a
> tension between ability to remove cruft versus breakage of old programs.

I also think there is tension about the definition of "cruft" in and of itself.
Some people have strong view about what should stay or go but at the end of the
day it's all opinions.  I have not seen objective rational for most of the
proposed decisions (thus far in my short tenure here).

> My idea for getting the best of both is psuedo-deprecation.

This seems much more cautious and appropriate.  If there is some Liberal
editorial bias against using certain operators or in prefering some styles over
others, then that doc writer can start by changing his copies of introductory
documentation.

Old code, people who already mastered the old functions and people who for
whatever reason happen to prefer an unauthorized programming style can continue
without disruption.

Still, it seems to me to be a fairly pointless exercise.

> but apply(), oct() and hex() are prime candidates.  The input() function
> should
> still be considered for genuine deprecation and deletion, but that reflects
> my
> view of input() as a bug.

Apply() and input() make some sense to me but what's so awful about oct() and
hex()?  From the narrow perspective of formatting integers, they fit mentally
quite nicely with str(), even though I realize that str() is part of some
greater, more general mechanism.  They're convenient, arguably simpler and
clearer than e.g., ( "%x" % n ).  Even though they're redundant, I don't see the
harm in simply leaving the poor babies alone.

> Alex Martelli pointed out that this was the original concept of deprecation,
> "it's there but we're sorry it is".

And all you ultimately need to do to deprecate something is to SAY it's
deprecated.  It can stay deprecated forever, you don't ever actually have to
remove it, or even remand it to a seperate section of the doc.

Of course, that begs the bigger issue about what items, if any, should be
deprecated in the first place.  You seem particularly anxious to prune when I
fail to see the advantage.  If you were designing the language from scratch, it
would be a completely different matter but in practice, you're talking about
changes while the machine is running.

> I think there is a clear advantage for creating a strong, central core of
> what you need to know and de-emphasizing what was part of the past.

It's hard to disagree with such as noble and general statement but I don't see
how it relates to any of the particular changes you propose.

I don't see that any of the proposed changes make material progress towards or
away from "creating a strong central core" or why de-emphasizing the "past" is
such a good thing.

I certainly don't agree that there's "clear advantage" to dinking with the
builtins.  Rather there appears to be clear risk and unjustifiable cost.

It truly seems to me that for some reason you're really itching to make this
change and you're having trouble coming up with truly objective rationale for
doing so.

Again, when we're talking about a major improvement to the language, such as
lexical scope, then may well be worth a little pain.  But the fact that we have
a way to gracefully phase in vitally important major changes should not be
abused for gratuitous little tweaks.

> there is a powerful intellectual advantage to mentally whittling down
> the list of tools to a well integrated, orthogonal, best-of-class set.

Again you make a statement of personal opinion that is quite the opposite from
obvious.

If the number of basic "tools" was, say 10, and you were proposing to narrow the
list down to the best 7, there MIGHT arguably be clear advantage.  E.g., if this
changed crossed some theoretical threshold that genuinely made the whole list
easier for people to learn and remember.

However, when builtins already consists of scores of discrete functions and this
is in the context of a language with literally hundreds upon hundreds of other
fundamental operators, functions and datatypes to learn, you don't really
accomplish ANYTHING that I can see by dinking with one or two builtins.

It's even sillier when dozens of redundant forms remain in the language.  At the
end of the day, it's completely subjective whether any particular particle of
redundancy is good and which is bad.  List Comprehensions are kind'a cool but
they're completely redundant with other ways of doing things.  And for a lot of
newbies, a regular for loop will be a whole lot clearer.

A little bit of redundancy is no big deal.

I worry even more when there seems to be some hidden agenda that these changes
are to enforce certain styles of programming over others.  E.g., the prejudice
against FP expressed by some (not necessarily you).  Who's to say your view of
what's "right" for everybody is truly right?  What's the problem with a little
FP now and again?  Too, some people seem to confuse FP with deeply nested
expressions.

> The quality of programs will improve, newbies will learn faster, the
> language can advance, AND old programs will run fine.

<gentle sarcasm>
And we'll get better gas mileage and loose weight in the process.
</gentle sarcasm>

I mean really, listen to yourself.  You're seriously arguing that the relatively
inconsequential change of moving a couple "old" functions to a separate section
of the DOCUMENTATION will materially affect the quality of programs and newbie
learning and that not making this change is preventing Python from advancing.

I respectfully have to disagree.

Respectfully yours,

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From danb_83 at yahoo.com  Sat Apr  6 18:17:20 2002
From: danb_83 at yahoo.com (Dan Bishop)
Date: 6 Apr 2002 15:17:20 -0800
Subject: PEP 285: Adding a bool type
References: <mailman.1017523442.6833.python-list@python.org>
Message-ID: <ad052e5c.0204061517.dfc0b41@posting.google.com>

"Jimmy Retzlaff" <jimmy at retzlaff.com> wrote in message news:<mailman.1017523442.6833.python-list at python.org>...
> Greg Weeks wrote:
> > I would, though, be happy to see built-in variables bound to 0 and 1.
> 
> I assume you mean that the only change would be to implicitly have
> built-ins like:
> 
> False = 0
> True = 1
> 
> There is a trap with this approach in that some people will come to
> conceptualize False/True as being instances of a Boolean type, even
> though they are not. Consider code like:
> 
> if MyCoolFunction(12.3) == True:
> 	print 'cool'

People who write code like that (even in languages with a real boolean
type) deserve to have their programs not work.


From fperez528 at yahoo.com  Mon Apr 15 17:36:04 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 15 Apr 2002 15:36:04 -0600
Subject: String interpolation question
References: <a9fcmo$npi$1@peabody.colorado.edu> <mailman.1018904544.23487.python-list@python.org>
Message-ID: <a9fhad$qbt$1@peabody.colorado.edu>

Geoff Gerrietts wrote:

>> but I don't want that because in my real problem, the string is long and I
>> don't want the dangers of positionally matched %s arguments.
> 
> Given your constraints, I think I'd prefer:
> 
>>>> print "%(val1)s" % { 'val1': y[i] }
> 
> Or if you insist on using locals() (a habit I find at least as
> dangerous as positional arguments):
> 
>>>> z = y[i]
>>>> print '%(z)s' % locals()
> 6
> 
> On the other hand, if you're looking at a long y and multiple values
> of i (though your example provides no such hint), I'd prefer:
> 
> d = {}
> for indx in range(len(y))
>   d['y_%d'%indx] = y[indx]
> d.update(locals())
> 
> s = '%%(y_%d)s' % (i,)
> print s % d
> 
> If none of these solutions seem to satisfy your needs, maybe you
> should be a little more explicit about the problem domain. If one of
> the requirements is that it be a one-liner, I'm not sure if I can
> help.

Thanks for your suggestions. Basically imagine having many arrays and needing 
to generate output of the form (this is perl syntax, which for once is a 
million times cleaner than python for _this_ problem):

print "$i arr1[$i]=$arr1[$i] arr2[$i]=$arr2[$i] arr3[$i]=$arr3[$i] ....\n";

To me, the above line is clear, unambiguous and requires zero contortions to 
either write or read (by the coder). I have been so far unable to write in 
python anything that comes anywhere within even shooting distance (with an 
ICBM) of this (but I'd love to be proven wrong ;)

I appreciate your solutions, but they all require the building of artificial 
structures or many temp variables. This makes for an ugly mess of the code if 
I have say 15 arrays for which I need to generate output like the above, and 
what annoys me is that a single well-interpolated line would do it (if such a 
facility existed in our otherwise beloved snake).

Oh well, I guess I'll just write the ugly mess.

Thanks for the help though.

Cheers,

f.


From aleax at aleax.it  Mon Apr 29 04:19:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 29 Apr 2002 08:19:42 GMT
Subject: [OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)
References: <mailman.1019526833.9592.python-list@python.org> <mailman.1019861990.32102.python-list@python.org> <874rhv448q.fsf@tleepslib.sk.tsukuba.ac.jp> <7ielgzro7j.fsf@enark.csis.hku.hk>
Message-ID: <ym7z8.30848$8D3.850421@news1.tin.it>

Isaac To wrote:
        ...
> Currently, the OSS point of attack seems more effective.  But I like the
> free software line of attack much better.  Users must not be at the mercy
> of producers.

"Yes, but".  The "theory of public choice" explains why producers (a 
smaller, compact group) are most often more effective at lobbying for their 
perceived interests than consumers (a larger, diffuse group).  If each
of (e.g.) 100 producers stands to lose $10,000 from a new public policy,
and each of 100,000 consumers stands to gain $11, overall welfare would
be enhanced by the policy BUT each producer has a strong incentive to
spend to purchase the rent coming from blocking the new policy, while each
consumer will not, rationally, have incentive to devote the matter much
thought, attention, time, energy.  The policy likely won't come.


Which is why it's pragmatically important to lobby for pro-consumers
policies even when it has to be by flawed arguments (distasteful, but
real).  For example, defend free trade by the argument "unfortunately
we do have to open our markets because in return nation X will open
theirs -- if we put protectionist measures in place they'd retaliate".

As Ricardo noticed two centuries ago, this makes about as much sense
as "retaliating" against other nations lacking good harbors by blowing
up one's own perfectly good ports -- nation X's collective advantage
is still with free trade even if our nation puts huge barriers up.

But putting this in a tit-for-tat framing works well with human
biological predispositions towards competition and cooperation, and
ranges on the side of free trade the relatively small group of producers 
who stand to gain relatively large amounts by nation X's markets being
open, as well as the larger groups of customers who stand to gain
smaller amounts by OUR markets being open.  Such intellectually
unpleasant but politically pragmatic approaches are thus useful.


I'm not claiming the pro-producer arguments for free or open software
are as flawed as the "else they'll retaliate" ones for free trade,
just offering a parallel with issues that are well known to anybody
who's ever dipped into public-choice theory.  Consumers SHOULD not be
at producers' mercy, but to help ensure that, signing some producers
up on the same side as consumers' interests is generally crucial.


Alex



From disposable-20020423.1706 at zigabyte.com  Tue Apr 23 18:15:44 2002
From: disposable-20020423.1706 at zigabyte.com (Alex Ramos)
Date: 23 Apr 2002 15:15:44 -0700
Subject: help wanted: apache & mod_python application on Linux
Message-ID: <bb2beadc.0204231415.24d8e12@posting.google.com>

Please reply through this link:
http://www.hotdispatch.com/view-task-requester?ID=15601256&POSTID=3228568421

I'm trying to run BoboMail from http://sourceforge.net/projects/bobomail

I have already installed and configured Apache 1.3.24 (the old,
but required version), Python1.6, mod_python, and the app itself,
and now I'm just getting some strange Python errors. 

The task requires logging in to a shell account that I will
provide, and performing the work hands-on.

Clarifications: Tue, 23 Apr 2002 4:16am 
I have now upgraded to Python 2.2.1, but it did not make things any easier.
Now apache won't even start up... posix function openpty() not found.


From ralph at inputplus.demon.co.uk  Mon Apr  8 06:54:29 2002
From: ralph at inputplus.demon.co.uk (Ralph Corderoy)
Date: 8 Apr 2002 10:54:29 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CB0495F.7010507@mxm.dk> <a8pnot$eck$1@inputplus.demon.co.uk> <a8qn4c$fmc$0@216.39.172.122>
Message-ID: <a8rst5$8ec$1@inputplus.demon.co.uk>

Hi Bengt,

> You could write
> 
>       a = b + oneIfMovableZeroIfNot(c)
> 
> if you wanted to be conspicuously clear about your function's intent.

is_movable is clear.  one_if_movable_zero_if_not isn't since I have two
phrases to parse, and a too-long name.  And at the end I'd just think,
`Oh, is_movable in other words.'

> It would express the semantics better if these were true boolean
> functions returning True or False, and you wrote it as
> 
>     major = [can_be_moved(o), can_be_sized(o), can_be_stacked(o),
>              can_be_coloured(o), can_be_hidden(o) ].count(True)

I'd just dislike building an array and then grepping through it for no
good reason, IMHO, when a simple addition suffices.

> Personally, I prefer to read unambigous code, and not have to refer
> to supporting documentation that I may have to double check anyway by
> reading function source, if I'm doing a serious review, e.g., chasing
> a bug. The new bools will help, since the arithmetic values can only
> be 0 or 1, and that is unambigous where integer function return
> values are not.

I don't see how your `count' version above means you don't have to
either believe the documentation of the five functions, or, better,
examine their code.  It's clear the code is counting those that return
True.  But they may be returning absolutely anything, just as they may
be in mine when I add them together.

Cheers,


Ralph.



From martin_byrne2 at yahoo.co.uk  Thu Apr 11 04:45:28 2002
From: martin_byrne2 at yahoo.co.uk (B)
Date: 11 Apr 2002 01:45:28 -0700
Subject: KeyboardInterrupt errors in ASP pages
Message-ID: <508cb8ae.0204110045.431dadee@posting.google.com>

I am using Python as the language for creating server side asp pages.
The problem is that when using Response.Redirect(page), I seem to get
the message below randomly but frequently enough to be a problem.

Has anyone else found this problem and is there a way around it. Is it
a bug in the Response.Redirect code?

Python ActiveX Scripting Engine error '80020009' 

Traceback (innermost last): KeyboardInterrupt 

/Default.asp, line 0


From philh at comuno.freeserve.co.uk  Mon Apr  1 15:30:11 2002
From: philh at comuno.freeserve.co.uk (phil hunt)
Date: Mon, 1 Apr 2002 21:30:11 +0100
Subject: Debugging flags [was: Re: PEP 285: Adding a bool type]
References: <mailman.1017466974.6751.python-list@python.org> <7xhemy1jo4.fsf@ruckus.brouhaha.com> <slrnaac6kf.2gs.philh@comuno.freeserve.co.uk> <slrnaacj3a.2gs.philh@comuno.freeserve.co.uk> <lXYp8.108382$Ou.84158@atlpnn01.usenetserver.com>
Message-ID: <slrnaahgqj.cqs.philh@comuno.freeserve.co.uk>

On Mon, 1 Apr 2002 08:05:24 -0500, Steve Holden <sholden at holdenweb.com> wrote:
>"phil hunt" <philh at comuno.freeserve.co.uk> wrote ...
>[...]
>> I've gone through some of my herbivore code, looking to see where
>> I've used 0 or 1 as a boolean. Example 1:
>>
>> debug = 0  # debugging this module?
>>
>> Frankly, if that is unclear to anyone, they are not cut out to be a
>> programmer. Bear in mind that most non-programmers are aware of the
>> 0/1 paradigm for boolean values: it appears on much electrical
>> equipment.
>>
>I find it much more helpful to use
>
>    debug = (1, 2, 7, 93) # debugging parsing, tree build, code generation
>and error messages
>
>with corresponding tests like
>
>    if 7 in debug:
>        ...

I do that for logging. The debug flags are specific to a particular 
module: they switch debugging on or off for that module, which is 
all the granularity I need.


-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if


From smithj at nsidc.org  Mon Apr  8 13:11:24 2002
From: smithj at nsidc.org (Jeff Smith)
Date: Mon, 8 Apr 2002 11:11:24 -0600
Subject: Reading/writing binary data.
In-Reply-To: <20020408121155.A16826@wblogan.net>
Message-ID: <Pine.SGI.4.10.10204081108430.144263-100000@graupel.colorado.edu>

On Mon, 8 Apr 2002, Ben Logan wrote:

> Hello,
> 
> I've been searching the Python (1.5) docs, but am unable to find any
> functions to read/write binary data from/to a file.  Under the "open"
> function I see that you can append a "b" to the mode to specify binary
> mode, but that doesn't help me any.
> 
> For example, if I want to take a list of numbers like [64,66,67] and
> write them to a file as binary, how would I do that?  I also need to
> be able to control whether the data is written in big endian or little
> endian format.
> 
> BTW, I'm running Redhat 7.1 on an Intel processor.
> 
> Thanks,
> Ben
> 
> -- 
> Ben Logan: ben at wblogan dot net
> OpenPGP Key KeyID: A1ADD1F0
> 
> Overdrawn?  But I still have checks left!
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

The array module has everything you need. You can use the tofile and
fromfile methods to read and write binary data, the tolist and fromlist
methods to switch data to/from lists, and the byteswap method for swapping
between big- and little-endian.

Jeff Smith






From sjames at mindless.com  Mon Apr 22 19:47:42 2002
From: sjames at mindless.com (Steven James)
Date: Mon, 22 Apr 2002 23:47:42 GMT
Subject: Using a Python script in Zope
Message-ID: <yi1x8.52892$ro5.4759478@bin6.nnrp.aus1.giganews.com>

I have a script that I wrote in Python to parse a text file and return a
dictionary of the values that I get from it. I used file.readline()
extensively to get the values from the file, but I can't seem to get this
function to work in Zope, where I intend to use the script. Anyone have
ideas on another way to do it, or how to make this work?


From aahz at pythoncraft.com  Mon Apr 29 23:37:21 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Apr 2002 23:37:21 -0400
Subject: Equivalent in Python?
References: <f7jz8.18268$Wd2.881746@wagner.videotron.net>
Message-ID: <aal3hh$e86$1@panix3.panix.com>

In article <f7jz8.18268$Wd2.881746 at wagner.videotron.net>,
Vincent Foley  <vinfoley at iquebec.com> wrote:
>
>What's the Python equivalient of Perl's "a"++; ?

There isn't any direct equivalent.  Python strings don't have any kind
of increment magic associated with them, but you could create a class
that worked like this:

    s = AutoIncrementString('a')
    s.increment()                   # or even s.inc()

However, if you designed it do this:

    s = AutoIncrementString('a')
    s += 1

you'd get harsh comments from the vast majority of Python programmers.
Bad juju, mixing strings and numbers without explicit conversion.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From bac at OCF.Berkeley.EDU  Wed Apr  3 22:59:41 2002
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Wed, 3 Apr 2002 19:59:41 -0800
Subject: Catalog of Python Patterns? [LONG!]
In-Reply-To: <a8g97m$10jq$1@news.idiom.com>
References: <a8g97m$10jq$1@news.idiom.com>
Message-ID: <Pine.SOL.4.44.0204031954400.13020-100000@death.OCF.Berkeley.EDU>

Have you looked at the Python Cookbook
(http://aspn.activestate.com/ASPN/Cookbook/Python)?  There are several GoF
patterns already there and done by some of the top Python programmers out
there.  There are even some Python-skewed version of some of these
patterns (the Borg pattern by Alex Martelli comes to mind).

-Brett C.



On 4 Apr 2002, Jim Dennis wrote:

>
>  GoF Design Patterns in Python
>
>  [Warning LONG POST AHEAD!]
>
>  I'd like to see a canonical catalog of OODP in Python.
>
>  At a minimum it should list each of the 23 GoF patterns, and
>  at least one minimal Python implementation of that pattern.
>  It should also point out the "native" use of each pattern in
>  the Python core, standard modules or among popular publicly
>  available third-party modules, packages and applications.
>  Naturally it should also discuss alternatives (Borg vs. Singleton)
>  and issues (particularly cases where Pythonic idioms make
>  patterns trivial, for example).
>
>  I realize that there used to be a "Python-Patterns" SIG (now
>  "retired") and a Google search will reveal a number of articles
>  on the topic (such as the one by Vespe Savikko).  Unfortunately
>  none of these sources is comprehensive.  Not one of these articles
>  or web sites shows a list of all 23 GoF patterns with a minimal
>  Python implementation.  Also many of them are somewhat dated.
>  I'd like to focus on Python 2.2 features (nods and caveats regarding
>  older versions can be added later).
>
>  I'll be the first to admit that I'm not a "Patterns" expert.
>  That's why I'd like to see *simple* (as it brain-dead, dirt,
>  simple) examples.  I'm also not an experienced professional
>  OO programmer.  I consider myself to be a student in this field.
>  So all of my comments should be take with a large lump of salt,
>  perhaps with a whole 5 lb. salt lick (for you equestrians).
>
>  Here's a start.  (All I've done is copy the GoF 23 from the
>  inside cover of their book, and inserted my questions, comments
>  and suggestions.  Note that many of my comments are laced with
>  the phrase "I imagine" or qualifiers like "maybe" or "could" ---
>  like I said, I'm just studying this stuff and I have little
>  practical experience with it).
>
>  Who wants to contribute "text book" sample implementations?
>
>  Creational::
>    Abstract Factory:
>     I've never understood the difference between an "abstract factory"
>     and a "factory method."  Since Python doesn't really have
>     "abstract" versus "concrete" classes, perhaps the distinction is
>     meaningless.
>
>     Every Python class acts like a factory.  anObj = Foo() returns
>     a reference to a new "Foo" object.  Additionally Python allows
>     us to define normal functions which return object references.
>     Indeed there is no way to tell from this assignment whether Foo
>     is a class or a function.  It could easily be a function that
>     returns an object from a pool or one that returns a reference to
>     a singleton. (Of course having a function masquerading as a
>     "class" in this way would inelegant and might violate the
>     principle of least surprise for some of the programmers using
>     or extending this code).
>
>     In Python2.2 it's possible to create/over-ride a class method
>     with the reserved __new__ name so that we have greater control
>     over the new instance creation (instantiation).  Guido shows
>     how __new__ can be used to implement a Singleton. It should be
>     easy to use __new__ to implement object pools (is that a
>     non-GoF pattern?), or any of creational patterns. (see below)
>
> 	Clearly, for a number of patterns, one object will be acting
> 	as some form of dynamic interface for one or more objects
> 	"behind it."  (Decorator, Adapter, Bridge, Proxy, Facade,
> 	etc).  The __init__ method (constructor) may have to instantiate
> 	these back-end objects, storing references to them as instance
> 	members.  The resulting "front-end" object must then dispatch
> 	method invocations (and attempted member accesses) to the
> 	appropriate back-end object(s) according to the required semantics.
>
> 	Indeed this model seems to be a unifying theme among many of the
> 	GoF patterns.  A Decorator enhances a back-end object by adding
> 	it's to members or methods to it (like a wrapper) and dispatching
> 	"base" methods back into it's instance of the underlying object.
> 	Bridges and Adapters reformulate method calls as the pass them
> 	back.  A Proxy selectively rejects some method invocations, or
> 	over-rides them, to "protect" the underlying object in some way.
> 	A Facade allows a group of objects to appear as a single object;
> 	it dispatches methods to the appropriate back-end instances.
> 	(In any of these cases the "front-end" might also wrapper or
> 	over-ride the return values).
>
> 	Quite a bit of the motivation and inspiration for the GoF
> 	patterns seems to be in recognition of the limitations of
> 	inheritance and class hierarchies.  They talk quit a bit about
> 	how these patterns allow OO programmers to avoid common
> 	pitfalls that result from attempts to use inheritance for
> 	"has a" rather than "is a" relationships, for cases where
> 	objects need to interact with other objects or in ways OTHER
> 	than inheritance.
>
> 	[Perhaps I'm wrong about this, but this is my take on the GoF
> 	patterns. BTW: the term "delegation" might be useful here.]
>
>    Builder:
>      I guess that "Builders" would often be used for "Composite"
> 	 objects.
>
>    Factory Method:
>      Are there any cases in Python 2.2 where this would be implemented
> 	 through something other than some combination of code in __new__
> 	 and/or __init__?
>
>    Prototype:
>      Are there any examples of this in the Python distribution?
> 	 I guess the implementation would usually be to create a
> 	 _clone() or __clone() method which would instantiate a blank
> 	 object (clone = self.__class__()?) passing it appropriate
> 	 constructor arguments (which it must have cached in its own
> 	 __init__) and calling on whatever methods are necessary to
> 	 put the clone into the desired (mostly identical) state as
> 	 the prototypical instance.  Finally the clone() function
> 	 would return a reference to the newly cloned instance.
>
> 	 I suppose one could have a Prototype (possibly some
> 	 selected instances of a class) define the __call__ method to
> 	 perform the cloning.  This would make the prototype look like
> 	 a factory or a normal class object throughout the rest of the
> 	 program.  I could envision a "Prototype" function that would
> 	 convert (promote?) a "normal" instance variable into a
> 	 Prototype by causing it to bind its __call__ method to the
> 	 necessary cloning operations.  This might look a little like
> 	 the way that one defines a class or static method and "promotes"
> 	 using syntax of the form:  mymethod = staticmethod(mymethod).
> 	 (Obviously the same function could be used to create a
> 	 new Prototype from any extent instance: fooProto = Prototype(inst)
> 	 instead of myproto = Prototype(myproto); the only difference
> 	 is whether you bind the result to the old name or a new one).
>
> 	 One question that naturally comes up with regard to Prototypes
> 	 is this: How does Python's introspection features facilitate
> 	 the implementation of Prototypes.  How much of the cloning
> 	 operation can be standardized and encapsulated in our
> 	 (hypothetical) "Prototype()" function (or mix-in class?)?
> 	 [I can imagine all sorts of clever fiddling with self.__dict__
> 	 et al --- possibly using the deepcopy module/functions; but
> 	 the actual fiddling is beyond me at this point, so it's
> 	 best left to the imagination].
>
>    Singleton:
>      Alex Martelli has long been a proponent of the Borg pattern
>      as an alternative to the Singleton.  While his point is well
> 	 taken (the focus should be on behavior, not object identity)
> 	 I think of the Borg as a way to work around limitation in the
> 	 earlier versions of Python.  With the introduction of new
> 	 classes and the ability to create static methods and over-ride
> 	 __new__ there is no need to use the Borg.
>
> 	 Incidentally the Borg is yet another example of the delegation
> 	 theme that I discussed before.  It redirects all access to
> 	 any member of the collective to the "queen" (back-end) instance.
>
>  Structural::
>    Adapter:
>    Bridge:
>      Frankly I don't see much difference between these two patterns.
> 	 I guess anygui might use Adapters to provide one uniform
> 	 interfacee to GUI widgets while adapting method calls into the
> 	 various semantics of back-end GUIs and widget collections.
> 	 A Bridge does the same thing.  GoF makes the distinction on
> 	 "intent" (Adapters being used to access legacy or external
> 	 classes through a common API, Bridges being a design decision
> 	 to decouple one's own classes from their APIs; it sounds like
> 	 hair splitting to me -- the same implementation techniques
> 	 should work in either case.
>
> 	 I can imagine cases where an Adapter or Bridge might have to
> 	 also be a bit of a Decorator or a Facade. If the semantics of
> 	 one API differ from another, for example, one might have to
> 	 instantiatiate additional backend objects to supply enough
> 	 functionality to support the desired common interface.  Otherwise
> 	 your interface is constrained to a least common denominator
> 	 among the possible back-end classes.
>
> 	 In these (and all of the other "delegation" patterns) I can
> 	 imagine lots of tricky, clever, subtle (and "too" extremes
> 	 of each: as in "too tricky," "too clever," and "too subtle")
> 	 ways of using __getattr__ and __setattr__ methods and the
> 	 hasattr(), to avoid explicit enumeration of the back-end
> 	 methods for those cases where the backend does have the
> 	 same interface as the Adapter/Bridge.  In other words it
> 	 may be possible through __getattr__ and __setattr__ or
> 	 similar special Python methods to allow any of these delegate
> 	 Patterns to only over-ride the necessary functions while
> 	 transparently passing along any other methods (and transparently
> 	 passing back results or raising exceptions?).
>
> 	 What are some pointers and caveats about this sort of "dynamic
> 	 delegation" (sort of "inheritance from outside of the family
> 	 tree")?
>
>    Composite:
>      The GoF refers to Composites as "tree structures."  This strikes
> 	 me as odd since I can imagine many composites of objects that
> 	 aren't trees.  I can imagine chains, directed graphs, networks,
> 	 and pools of objects that are interconnected through instance
> 	 members.  I can also see where one might have class/static
> 	 members which track all of the extant instances (__new__ might
> 	 append to "__class__.__instancelist" and __del__ might perform
> 	 the remove.  Could __del__ "divert" the destruction, blank the
> 	 instance and add it to a pool?  Might one define a wrapper around
> 	 the del() built-in that would do something like:
>
> 		def del(arg):
> 			if hasattr(arg,'__pool__'): arg.__pool__()
> 			else:                     _del(arg)
>
> 	But I gather (from reading the GoF) that the object here
> 	(to excuse the pun) is to allow for a method to propagate
> 	transparently to any subset of the composite regardless
> 	of whether that subset effects the whole tree (invoke it
> 	on the root node) or an individual leaf (the degenerate
> 	case).  So I guess the core of a Python Composite with
> 	support for a given method's propagation might be:
>
> 		Class CompositeFoo:
> 			def __init__(self):
> 				self.children = []
> 			def propagated_method(self, *args, **kwargs):
> 				# do stuff to self
> 				if supposed_to_propagate:
> 					for i in self.children:
> 						i.propagated_method(args, kwargs)
>
> 	[Does that make sense?  BTW: I'm envisioning some cases
> 	where the propagation is intended to be limited or conditional]
>
> 	So, what are the caveats?  Are there any examples in Python
> 	itself?  Are there any coding conventions that might be
>     encapsulated into an abstract base class or a mix-in?
>
>    Decorator:
>    Facade:
>      As I've already said, Decorators and Facades, Adapters, Proxies
> 	 and Bridges (and Mediators and possibly even State objects all
> 	 seem to be forms of delegation (or collaboration?).  In all
> 	 cases there is a delegate (front end) object which is "used"
> 	 in our code and it instantiates one or more of the underlying
> 	 (back end) objects as its own member(s).  The delegate then
> 	 dispatches or relays method invocations back to the underlying
> 	 object(s) according the the delegate's semantics.
>
> 	 Am I missing something here?  Is this a gross
> 	 oversimplification?  Are there any differences among these
> 	 patterns that require special implementation (for Python)
> 	 or to the same language features and techniques apply to
> 	 all of these "delegations" (my term).
>
>    Flyweight:
>      I've read innumerable times that the Python interpreter
> 	 implements a set of static immutable objects for a small
> 	 range of integers (-1 or -2 through 100 or so?).  I guess
> 	 this is an example of Python's native use of the Flyweight
> 	 pattern.
>
> 	 Are there any others?  Would it be possible to have a
> 	 sort of "copy on write" (COW) Flyweight?  These would behave
> 	 like Singletons for read-only purposes, but then be promoted
> 	 to separate/new objects on any setattr?  What would be a
> 	 use for such a beast.
>
>    Proxy:
>      See my comments for Decorators, Facades, Adapters, and Bridges.
>
>
>  Behavioral::
>    Chain of Responsibility:
>      A cascaded of Decorators or a Chain of responsibility.
> 	 What's the difference.
>
>    Command:
>      Some of the GoF's comments on this sound like it would
> 	 be ideal for supporting AOP (aspect oriented programming).
>
>    Interpreter:
>      I just don't get this pattern.  I'll have to mull on it
> 	 for a bit.  Other the trite observation that Python *is*
> 	 an interpreter, are there any examples of the Interpreter
> 	 pattern in Python?
>
>    Iterator:
>      Iterators seem to be evolving into one of Pyton's most
> 	 visible and commonly used features.  In Python2.2 we
> 	 see automatic iteration over files, and dictionaries as
> 	 well as the traditional iteration over built-in sequences
> 	 (lists, tuples, strings).  We also see support for defining
> 	 iterators for our own classes.
>
> 	 It seems like iterators and generators are intertwined in
> 	 someway, but I don't understand it.  [So maybe I'm just
> 	 wrong and they only seem to be related because they're
> 	 described in the same docs, or something].
>
>    Mediator:
>      This seems like a sort of bi-directional Facade to me.
>
>    Memento:
>      Here's where Python's "encapsulation by courtesy" and lack of
> 	 enforced data hiding might make a GoF pattern particularly
> 	 easy; almost unecessary.  Alternatively I can imagine that
> 	 some implementation of the "Memento" pattern might involve a
> 	 coding convention where each class that supports it implements
> 	 an __memoize() method.  Might that return a pickle of self?
>
>    Observer:
> 	 I can imagine where some objects, including composite
> 	 objects might "register" with some sort of "registry" (sorry
> 	 for the redundancy).  I'm just not sure what sort of applications
> 	 one my find for this sort of "registry" even though I'm sure
> 	 it would be easy (almost trivial) to implement in Python.
>
> 	 I suppose such a "registry" makes quite a bit of sense in the
> 	 Observer pattern.  Although I gather that this would usually be
> 	 a instance registry in each of the "observer/dispatchers"
>
>    State:
>      I see this as similar to an Adapter except that it may
> 	 change which back end object(s) it dispatches to (or calls
> 	 upon or delegates for) might change dynamically during the
> 	 lifetime of a given instance.  For example I can imagine an
> 	 object that implements a simple list for some of its members
> 	 until the list gets beyond a given size, then it might
> 	 change that to a dictionary, or some sort of AVL tree or
> 	 more complex data structure at the object needs to scale
> 	 upwards.
>
>    Strategy:
>    Template Method:
>      Would some of the mix-ins from SocketServer count as
> 	 examples of the Strategy or Template Method patterns?
>
>    Visitor:
>      os.path.walk is an obvious example of a standard Python
> 	 module/function that uses Visitor.  (Obvious because it's
> 	 documented as the "visitor" function).
>
>      However, this is also one of the most confusing functions
> 	 in the standard Python library.  I seem to see  befuddled
> 	 posts about it in the c.l.py newsgroup at least once a month.
> 	 As the GoF point out, Visitors are often used on Composites.
>
> 	 It seems like generators might often be alternatives to
> 	 Visitors.  Rather than passing a function to each object
> 	 and saying "do this to yourself" it seems natural to
> 	 iterator over a set of object references and invoke the
> 	 methods (or even, horrors! play with its attributes) directly.
>
> 	 However, Python's ability to pass functions, objects and
> 	 closures around as first order data types seems to make
> 	 Visitors easy to implement --- if you can just understand
> 	 the scoping well enough for the Visitor to know what to
> 	 do (and who to do it to) when it gets there.
>
>



From max at alcyone.com  Tue Apr 16 00:42:00 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 15 Apr 2002 21:42:00 -0700
Subject: pre-PEP for optional 'pass'
References: <slrnabmoci.dhk.philh@comuno.freeserve.co.uk> <mailman.1018927651.27228.python-list@python.org>
Message-ID: <3CBBAB98.91FA1B17@alcyone.com>

jepler at unpythonic.net wrote:

>     [2, 3, 5, 7, 9, ...]
> would generate successive primes.

Pretty lousy code, since 9 is not prime.  :-)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Nationalism is an infantile sickness.
\__/ Albert Einstein
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From maxm at mxm.dk  Fri Apr  5 02:06:00 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 05 Apr 2002 09:06:00 +0200
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <3CAA2A44.37BB218F@accessforall.nl> <lFpq8.44725$pT1.1265790@news1.tin.it> <3CAB87ED.C5567DBE@accessforall.nl> <TdSq8.48142$pT1.1370343@news1.tin.it>
Message-ID: <3CAD4CD8.2090700@mxm.dk>

Alex Martelli wrote:


> When this PEP passes and 2.3 becomes the only Python
> release that my code must support I may have to switch to
> "return True" vs "return 1"


Oh stop whining!

Just do a search and replace in your codebase::

     s/1/True/gi

That should solve all those problems in a few seconds.

regards Max M



From st-newsgroups at thomas-guettler.de  Thu Apr 11 10:00:36 2002
From: st-newsgroups at thomas-guettler.de (Thomas Guettler)
Date: Thu, 11 Apr 2002 16:00:36 +0200
Subject: Python extensible Emacs clone ?
References: <3CB58717.4BF3035B@dynamic.de>
Message-ID: <3CB59704.5050103@thomas-guettler.de>

Matthias Hanitzsch wrote:

> Hallo,
> 
> is there a free Emacs clone that uses Python as an extension language ?
> 
> Matthias
> 
> 


You might want to try pymacs:

It enables you to script emacs with python

http://www.iro.umontreal.ca/~pinard/pymacs/

(See Thread "Emacs-like Editor" from April 9 2002)

  thomas



From maxm at mxm.dk  Mon Apr  8 09:12:18 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 08 Apr 2002 15:12:18 +0200
Subject: Discussion of Python midi library
Message-ID: <3CB19732.8060304@mxm.dk>

I am writing a Midi Module for Python.

I basically want it make it possible to:

     1) Read midi from a file

     2) transform the midi data

     3) Write midi to a file

     4) Play midifiles (especially on Windows)

The biggest challenge here is getting the internal representation of the 
midi data right.

The midi data format in it's raw format is too specialized to make it 
possible to load some midi-data and then do any kind meningfull 
transformations on it.

So generally I want to be able to load in the data into a Python 
specific objects representation that can be easily modified and then 
saved back out as a new midi file.

I am doing algorithmical composition hence the need.

Has anybody got an interrest in participating in a discussion about how 
such a module should be designed please contact me.

I allready have some code and some ideas.

regards Max M



From guido at python.org  Fri Apr  5 14:00:21 2002
From: guido at python.org (Guido van Rossum)
Date: Fri, 05 Apr 2002 19:00:21 GMT
Subject: 2.2 descriptor API
References: <410890c4.0204040806.363f5728@posting.google.com>
Message-ID: <3CADF520.C0264C0B@python.org>

Daniel Popowich wrote:
> 1.  Why is the __get__ method called on a descriptor when it is
>     referenced as a class attribute?

This was a requirement.  When you get a method out of a class in Python 2.1
and before, it gets special treatment which gives you an "unbound method".
The unbound method insists that you call it with an instance of the right
class, in order to prevent mistakes.  The new descriptor API must be able
to support similar behavior.  But it must also support other behavior, like
class methods and static methods.

> 2.  What is the purpose behind the second argument; does someone have
>     an example of how or when I would make use of it?

When __get__ is called on a class, the first argument is None, and the
second argument is the class.

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


From david at dataovation.com  Sat Apr  6 13:23:58 2002
From: david at dataovation.com (David McInnis)
Date: Sat, 6 Apr 2002 10:23:58 -0800
Subject: Modify Word Doc with Python
In-Reply-To: <3caf28ce$0$69917$7586b60c@news.frii.com>
Message-ID: <010d01c1dd98$385ef800$9865fea9@bigdaddy>

The only properties that I need to change at this time are the document
properties like Author, Company, Name, Title etc.  (File > Properties).

David

PS:  It was late at night.  I HAVE python installed on a linux box. :) 



-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Marc Poulin
Sent: Saturday, April 06, 2002 7:00 AM
To: python-list at python.org
Subject: Re: Modify Word Doc with Python

In article <mailman.1018081037.4211.python-list at python.org>, "David
McInnis" <david at dataovation.com> wrote:

> I need to alter the document properties of Word DOCS that our users

More info, please. What properties are you specifically trying to 
alter?
-- 
http://mail.python.org/mailman/listinfo/python-list





From aleax at aleax.it  Mon Apr  8 12:09:42 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 08 Apr 2002 16:09:42 GMT
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <a8q0g1$qu5$1@panix1.panix.com> <a8qqf7$p4h$0@216.39.172.122> <Epes8.59265$pT1.1705378@news1.tin.it> <a8sf0k$2a5$0@216.39.172.122>
Message-ID: <ahjs8.73608$S52.2597704@news2.tin.it>

Bengt Richter wrote:
    ... 
>>>>Assignments in expressions will never happen.
    ...
> But why "identical" ;-) If I can find another spelling that works
> everywhere (and it wouldn't surprise me if you had one in mind ;-)
> wouldn't that be interesting?

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66061

for example.  But, like your attempt, it's based on NOT being
an assignment (so Aahz's quote above remains perfectly right).

Not all "stashing a reference to this value off somewhere" are
assignments (some of the many who aren't can correctly appear
within expressions -- others can't).


Alex



From phil at river-bank.demon.co.uk  Mon Apr 15 11:52:58 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Mon, 15 Apr 2002 16:52:58 +0100
Subject: Questions on the C Thread API
References: <mailman.1018871909.16619.python-list@python.org> <Xns91F18C58A3AAAduncanrcpcouk@127.0.0.1>
Message-ID: <3CBAF75A.9E2F1BA4@river-bank.demon.co.uk>

Duncan Booth wrote:
> 
> Phil Thompson <phil at river-bank.demon.co.uk> wrote in
> news:mailman.1018871909.16619.python-list at python.org:
> 
> > Currently, any C++ method that might block effectively has a call to
> > PyEval_SaveThread() before it and a call to PyEval_ReleaseThread() after
> > it. Any C++ virtual method that has been re-implemented in Python first
> > checks to see if it already has the global lock, and acquires it if it
> > doesn't. The Python method is then evaluated, and the global lock
> > released if it had previously been acquired. The check if the lock has
> > already been acquired is necessary because I only release the lock on a
> > small subset of C++ methods, ie. those that might block, which I do for
> > reasons of efficiency.
> 
> PyEval_SaveThread and PyEval_ReleaseThread both release the
> interpreter lock. Did you mean PyEval_RestoreThread for the second one of
> these?

Yes I did.

> Anyway, when you are going to do something that might block you save the
> thread state and restore it later. I presume you save the state in some
> thread local storage because you could have several threads all suspended
> in this way.
> >
> > All the above works fine, so long as all the C++ calls are made from the
> > same thread. With the support of the Qt thread classes, that assumption
> > is no longer valid.
> >
> > Therefore, just before calling the Python code that re-implements a C++
> > virtual method, the global lock has to be acquired, and the correct
> > thread state restored. However, to prevent a deadlock the thread state
> > can only be restored if it is not the current one, but there is no way
> > to get the value of the current thread state without having acquired the
> > global lock. Catch 22.
> 
> Get the saved thread state for the current thread. If you don't have one
> then by definition you already have the global lock. If you have one then
> call PyEval_RestoreThread to get the lock and set the state and clear your
> saved state to NULL. Remember whether or not you had a thread state around
> the callback and only save the state/release the lock if you previously
> restored a state.

This is the bit I don't understand. How can I get the saved thread state
for the current thread? Do you mean I can use the Python API to get it
(in which case I get back to Catch 22)? Or do you mean I have to write
code that determines by other means what the current thread is and then
get the corresponding Python thread state?

Phil




From phr-n2002a at nightsong.com  Sun Apr 14 19:44:46 2002
From: phr-n2002a at nightsong.com (Paul Rubin)
Date: 14 Apr 2002 16:44:46 -0700
Subject: Deprecate input()
References: <a9cv63$qhu$1@bob.news.rcn.net>
Message-ID: <7xy9fplua9.fsf@ruckus.brouhaha.com>

"Raymond Hettinger" <python at rcn.com> writes:
> The help and tutor maillists indicate that the use of input() rather than
> raw_input() is widespread.  I think this function is so harmful and so
> likely to be used that it should be deprecated hasta pronto (right away).
> When really needed, eval(raw_input()) is easy enough to code and explicit
> about what it does.
> 
> <no wink -- this function needs to go away>

I agree with this.  Some other stuff like "smart cookies" also need to go away.


From cjw at sympatico.ca  Thu Apr 11 10:10:55 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Thu, 11 Apr 2002 10:10:55 -0400
Subject: Meaning of None - was PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <mailman.1017755045.4308.python-list@python.org> <MPG.1713b8a1bedfe05098968b@news.nwlink.com> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <3CAF6EE6.7564CDF7@sympatico.ca> <ZMZr8.1903$Sw5.1235@atlpnn01.usenetserver.com> <3CB49ADD.89B242A4@sympatico.ca> <a92c4f$ne7$0@216.39.172.122> <slrnab9ngr.i8b.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3CB5996F.A546DC29@sympatico.ca>


Huaiyu Zhu wrote:

>
> I had also proposed this as a way to properly define rich comparisons:
>
>         a   b   a > b   a == b   a < b
>         1   0   True    False    False
>         0   0   False   True     False
>         0   1   False   False    True
>         0   1j  None    False    None

This raises a different issue than was raised initially.
For the last case a > b and a < b are treated as
incompatible types and raise exceptions.

>
>
> It is often necessary to compare objects in a set even though there is only
> a partial order.  We often do not want exceptions when the order is not
> defined, since comparisons are often imbedded in expressions but try/except
> has to be statements.

Correct, but I was addressing the case where one might get
a NULL value from a database.  In this case, the NULL
typically indicates that a value is unknown, not available
or even inappropriate.

Suppose A= 25 and B=NULL and we:
assert A>B

Currently, no exception is raised.  This is correct when
we have certainty as to whether A>B, A==B or A<B.

However, when the value of B is currently unknown and
perhaps even unknowable, would it not be better to return
a value other than True or False which indicates this
uncertainty?

Colin W.

>
>
> On the other hand, a comparison operator with a total order is useful in
> many situations even though this total order is not related to the _natural_
> partial order.   For example, cmp(1, None) is well defined.  Why would
> cmp(1, 1j) be an exception?   This is especially painful if all you want is
> just a _unique_ order, whatever it is.
>
> Anyone paying attention to such matters over time would discover that
> allowing three valued logic would solve a supprisingly large number of
> trivial but nasty problems.
>
> However, I do not see an easy way to extract None from Python's pool of
> objects denoting false.  If only it was not designed that way ...
>
> Hopefully when IEEE floating point support becomes ubiquitous we could
> redefine the comparison with NaN for this purpose.  (You might be able to
> get at it now as NaN=(1e200j**2).imag, but the comparison is always False.)
>
> Huaiyu




From sjmachin at lexicon.net  Wed Apr 24 21:34:29 2002
From: sjmachin at lexicon.net (John Machin)
Date: 24 Apr 2002 18:34:29 -0700
Subject: replacing Chinese chars with their spellings
References: <m2g01lfx7i.fsf@jidanni.org>
Message-ID: <c76ff6fc.0204241734.40e57406@posting.google.com>

Dan Jacobson <jidanni at deadspam.com> wrote in message news:<m2g01lfx7i.fsf at jidanni.org>...
> Before I start learning python, here's what I want to do: I have a
> table of Hakka Chinese words and their pronunciations.  I scan a file
> and replace any Hakka ["big5" 2, 4, 6 ... byte long strings] there
> with their pronunciations.  If it were just one character [two byte]
> words I would use the "c2t" program.  Is there a template that munches
> forth in a file and replaces the longest match in a database
> before moving on?

I doubt there is such a "template". What you would need to do is
something like
this:

1. Set up Python dictionaries, one for possible Hakka word-length. So
if the longest Hakka word in your table is say 5 Chinese characters,
then you would need 5 dictionaries.
2. Read your whole file into a string.
3. Isolate runs of Chinese "vocabulary" characters (runs being
separated by punctuation etc).
4.
For each run:
   Set pointer to start of run
   For each pointer position inside run:
      For n in (5,4,3,2,1):
         Look up the next n characters in the n-char dict;
         if found, substitute and move your input pointer on
         by n*2 bytes
      If all five lookups fail, it means you don't have
      a 1-char dictionary entry for a single Chinese character;
      put out whatever error message or substitution
      you consider appropriate, and move on to the next
      character in the run.

Warning: this is a data processing approach, not a computation
linguistics approach. It is quite possible that it might produce
howlers of the sort that early machine translation efforts were
alleged to do:

"The spirit is willing but the flesh is weak" -> Russian equivalent of
"The vodka is fine but the meat has gone rotten".

Presumably the point of having a multi-character pronunciation table
is that it is possible that pronounce("xy") can be != pronounce("x") +
pronounce("y"). With careful thought, you may be able to remove
redundant entries from your more-than-one-char dicts, so that they
contain only the necessary exception cases -- but do try the basic
approach first.

As well as pronunciation dictionaries, might you not also need to
incorporate rules for such things as tone sandhi?


From weeks at vitus.scs.agilent.com  Sat Apr 20 13:57:30 2002
From: weeks at vitus.scs.agilent.com (Greg Weeks)
Date: Sat, 20 Apr 2002 17:57:30 +0000 (UTC)
Subject: Counterintuitive Python behavior
References: <3CBD1C02.8AB2CF62@yahoo.com> <1019060158.737451@cswreg.cos.agilent.com>
Message-ID: <1019325450.422371@cswreg.cos.agilent.com>

Some comments regarding "conceptual sameness":


From: "Petr Prikryl" <Answer.via.news.please.prikryl at skil.nospam.cz>

> There are two distinct things when comparing the objects: the object
> content and the object identity.

Neither of these concepts corresponds to conceptual sameness.  I haven't
defined "conceptual sameness", but it is a natural concept:

Suppose a Bank_account type has the single attribute "balance".  Consider

    my_account = Bank_account(50)	# INITIAL BALANCE OF $50
    my_account.deposit(50)		# BALANCE IS NOW $100
    your_account = Bank_account(100)	# INITIAL BALANCE OF $100

Do we have the same bank account?  Of course not.  Adding money to your
bank account is different from adding money to my bank account.  So, in
this case, conceptual sameness corresponds to sameness of identity.  The
same goes for lists and dictionaries, for the same reasons.

On the other hand, suppose we define our own Complex number class.

    x = Complex(3,4)
    y = Complex(3,0) + Complex(0,4)

Are x and y the same?  Of course.  There is only one 3+i4.  In this case,
conceptual sameness corresponds to sameness of content.

A final example is immutable strings, which we imagine to be implemented as
a character array plus start and stop indices.  (This implementation
permits a rapid substring operation.  Note that the three attributes are
not allowed to be modified after they are initialized.)  Consider:

    s1 = String(['c', 'a', 't'], 1, 3)
    s2 = String(['a', 't', 'e'], 0, 2)

Are these the same string?  Conceptually, yes.  When you are comparing
strings in your code, it will be conceptual sameness that you are
interested in.  But neither the identity nor the content of s1 and s2 are
the same.

So, conceptual sameness corresponds sometimes to sameness of identity,
sometimes to sameness of content, and sometimes to neither.  And (I
maintain), in ordinary user code, conceptual sameness is the concept of
*identity* that is most useful.  (I stress "identity" because there are
weaker forms of comparison that are sometimes useful.  But not *as*
useful.)

So Python is missing THE single most useful comparision operator.  But
there are workarounds.  "==" means conceptual sameness for all objects
other than lists and dictionaries, provided that __cmp__ is appropriately
defined.  Furthermore, if __hash__ is appropriately defined, any object
other than a list or dictionary can be a hash key.  So, except for lists
and dictionaries, everything's fine.


From: Philip Swartzleonard <starx at pacbell.net>
> >     a eq a          # PROVIDED a HAS A VALUE
> 
> Well, if a is defined, a has a value. So this is either true or a 
> NameError (?). Or do you mean if a is None, a eq a should be false?

No.  I was perhaps overfastidiously ruling out:

    >>> a is a
    Traceback (innermost last):
      File "<stdin>", line 1, in ?
    NameError: a


Regards,
Greg



From shalehperry at attbi.com  Fri Apr 19 17:56:13 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 19 Apr 2002 14:56:13 -0700 (PDT)
Subject: Substitution with Hash Values
In-Reply-To: <pan.2002.04.19.16.48.06.120768.1546@jonesj.com>
Message-ID: <XFMail.20020419145613.shalehperry@attbi.com>

On 19-Apr-2002 Jeff Jones wrote:
> I am a long time perl user who is trying to get a grip on python.  I
> really like what I have seen so far, but I have hit a little snag
> regarding substitution (like perl).
> 
> I use html templates for all of my CGI programming, and everything that needs
> to be replaced is contained within double curly braces ( {{value}} ).  In
> perl I use a hash to keep all of the variables that are going to replace
> the placeholders in the template, and replace them the following way:
> 
>       ....
>       s/{{(.*?)}}/$hash{$1}/g;
>       ....
> 
> That substiution is accomplished as I read in the template file.
> 
> Can anyone tell me how I might do this in python?  I have tried many
> different ways, but none seem to work.
> 

I presume you know that a dictionary in python is the equivalent of the perl
hash.

hash = {}
hash[key] = value
hash[other_key] = new_value
etc.

as for the substitution perl is a little better here because regex is part of
the language, but python can do it too with a little more typing.

import re
brace_re = re.compile(r'{{(.*?)}}')
....
match = brace_re.search(input)
if match:
    input = brace_re.sub(match.group(1), input)
# use input
....




From ods at fep.ru  Wed Apr 10 08:20:29 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Wed, 10 Apr 2002 16:20:29 +0400 (MSD)
Subject: PEP 282: A Logging System
Message-ID: <Pine.LNX.4.44L.0204101618300.13708-100000@ods.pp.ru>

I like the idea of logging system in standard Python library. The
interface described in PEP is very interesting. But there are things to
be discussed.

First it's very slow. Here are timings obtained from my test script
(http://ods.pp.ru/logging/logtest.py):
---
simplelog (log): 0.59
simplelog (ignore): 0.07
logging (log): 212.88
logging (ignore): 0.42
logging (log, no inspect/thread): 3.25
logging (ignore, no inspect/thread): 0.42
---
That's mainly due to unnecessary inspect calls. Vinay said that he fixed
this problem in upcomming release, but the module is still slow. I'm not
sure what we realy need: not-so-complicated but fast logging system or
extremly flexible and a bit slow? I guess it should be very fast at
least when logging is disabled.

May we should use lazy evaluation of some time-consuming properties
(filename/lineno, asctime)?

Default behavior of logging is to log all messages: if no handlers
found on first call, default is installed.  That's strange.  I guess it
should produce no log messages by default, i.e. no default handlers
should be installed.

I dislike the idea to import (and use for locking) thread module even
when application is not threaded. This is a general problem with many
modules. May be we need a atexit-like registry of functions that will be
called on first thread import?

The filter interface (object with filter method) is defined so that even
simple filters cannot defined inline. It can be changed to require any
callable object (instance with __call__ method or just an inlined
lambda). Vinay said that this make it less readable for newbies, but I'm
not sure he is right.

The most usefull filter (after simple setting the minimum log level) I guess should be a name filter. Something like this (I've replaced filter method with __call__):
def NameFilter:
    def __init__(self, name):
        self.name = name
        self.length = len(name)
    def __call__(self, record):
        return record.name.startswith(self.name) and \
                   (len(record.name)==self.length or
                    record.name[self.length:self.length+1]==',')






From gleki at gol.ge  Mon Apr  8 08:59:35 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Mon, 08 Apr 2002 14:59:35 +0200
Subject: __init__.py
Message-ID: <3CB19437.C1B2B3E9@gol.ge>

I wonder if an experienced developer in Python gives me a few hints on
how to make the initfile of a package.
I find the section in the tutorial
http://www.python.org/doc/current/tut/node8.html#SECTION008400000000000000000

not very much helpful:(

Thanx,
Giorgi



From gh_pythonlist at gmx.de  Wed Apr  3 05:38:48 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 3 Apr 2002 12:38:48 +0200
Subject: How to check whether email /domain exists
In-Reply-To: <20020403131722.I23871@phd.pp.ru>
References: <3CAADA7B.25443.509AE1@localhost> <20020403131722.I23871@phd.pp.ru>
Message-ID: <20020403103848.GD463@lilith.hqd-internal>

* Oleg Broytmann <phd at phd.pp.ru> [2002-04-03 13:17 +0400]:
> On Wed, Apr 03, 2002 at 10:33:31AM +0200, A wrote:
> > What is the simplest way how to check 
> >  whether a given email address( or domain for this email) exists?
> 
>    Send a mail and ask the user to reply or click an URL.
 
Or send him a HTML mail that has an img tag downloading an image from the net
(1 pixel transparent GIF is enough) >:-)

Which is a good reason to not enable HTML mail in your mail client at all.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 15.4 ?C      Wind: 6.2 m/s



From boud at valdyas.org  Tue Apr  9 11:29:11 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 09 Apr 2002 14:29:11 -0100
Subject: PyQt question (colors)
References: <3cb2d4e7_11@news.newsgroups.com>
Message-ID: <3cb2dff6$0$85444$e4fe514c@dreader3.news.xs4all.nl>

Janos Blazi wrote:

> If I say
> 
>     stat_lab1.setBackgroundColor(QColor("DarkGreen"))
> 
> and stat_lab is a QLabel, this has the required effect. But if I say
> 
>     pb.setBackgroundColor(QColor("DarkGreen"))
> 
> and pb is a QProgressBar object, so this has no effect at all. Can anybody
> tell me the reason for this? I use PyQt with Qt 2.3 on a MS-Win2000
> system.
> 

Using colors in Qt is quite complicated, unfortunately, and it has
become even more complicated in Qt 3, where setBackgroundColor is
so deprecated as to have disappeared from PyQt.

In this case I think you will really have to read Troll Tech's documentation 
to get a real understanding: it's not something that can be accurately
explained in a Usenet posting. The upshot is that what you perceive
to be the same thing -- the background of a label and a progressbar,
are in the Qt system of widget painting different things.

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


From bergeston at yahoo.fr  Wed Apr 10 05:25:40 2002
From: bergeston at yahoo.fr (Bertrand Geston)
Date: Wed, 10 Apr 2002 11:25:40 +0200
Subject: Easter Eggs
References: <fm2s8kKkXAiR092yn@gnosis.cx> <mailman.1018391839.6535.python-list@python.org>
Message-ID: <a910e5$vnkga$1@ID-135695.news.dfncis.de>

"Kalle Svensson" <kalle at gnupung.net> wrote in message
news:mailman.1018391839.6535.python-list at python.org...
> [Lulu of the Lotus-Eaters]
> > Who can tell me what's in there.  Someone on c.l.py recently let drop
> > the hint about finding the Zen of Python at the interactive prompt in
> > 2.2.  I went so far as trying the command in Python versions 1.51, 1.52,
> > 2.0, 2.1, 2.2 (it's only in 2.2)... but then promptly forgot what the
> > magic command was.  Can someone remind me, please?
>
> "import this", IIRC.

Really cool.
That reminds me that I always wonder about the meaning of this one:

Flat is better than nested.

Is it because of my OO background or do I miss something ?
Any explanation ?

TIA

B.

>
> Med st?rsta h?gaktning,
>   Kalle
> --
> Kalle Svensson (kalle at gnupung.net) - Laziness, impatience, hubris: Pick
two!
> English: http://www.gnupung.net/  Svenska:
http://www.lysator.liu.se/~kalle/
> Stuff: ["http://www.%s.org/" % x for x in "gnu debian python
emacs".split()]
>
>




From xnexau at yahoo.com.au  Sat Apr 13 22:42:58 2002
From: xnexau at yahoo.com.au (Patrick W)
Date: 14 Apr 2002 12:42:58 +1000
Subject: newbie-question: interfaces
References: <3cb7f47e$0$360$9b622d9e@news.freenet.de> <a999hl$nmd$1@rex.ip-plus.net> <3cb843d0$0$4068$9b622d9e@news.freenet.de> <82bscn4nvz.fsf@acropolis.localdomain> <i7%t8.3632$bn2.1579@nwrddc01.gnilink.net>
Message-ID: <82ads7q9u5.fsf@acropolis.localdomain>

"logistix" <logstx at bellatlantic.net> writes:

> Interfaces are important if you want to enforce a design-by-contract.
> Granted, in Python you can query a class at runtime to see if it implements
> a property or method, but this doesn't guarantee that the method does what
> you want it to do semantically.  (Just because a function is named foo, it
> doesn't mean it foos the way you want it to.)
> 
> Not to be a prick either, but if you're not using interfaces or ABC's, then
> you're not taking advantage of the true goals of OO programming, you're just
> warping the function names and data structures for your convienence.  The
> whole point of OO is to deal through abstract interfaces that hide the
> implementation details.


I agree that interfaces / abstract base classes are important in
statically typed OO languages, but I still don't see what purpose they
serve in dynamically typed languages like Python. It seems to me that
object orientation has a somewhat different 'flavour' in the two types
of languages, and one of the main differences lies in the purpose that
inheritance serves.

If we use C++ as an example of a statically typed OO language, you
might use inheritance for two reasons. One would be to specify an
interface (guaranteeing that any instance of a class derived from a
certain type will implement a certain method which can be called
without knowing the exact (sub)type of an object). Another purpose
is to simply inherit behaviour, which is basically a convenient code
reuse mechanism.

I can see good uses for implementation inheritance in Python, but not
for interface inheritance. In your remarks above, you imply that there
is a choice in Python between providing an abstract interface, or
doing explicit type checking. But it seems to me that this overlooks
the fact that Python methods and functions are similar to C++
templates. If an object supports a particular behaviour, regardless of
whether and where it inherits this behaviour from, you can invoke
it. If it doesn't, you can't. Since none of this must (or can) be
handled at "compile time" in Python, what would be the advantage of an
forcing objects to inherit from an abstract interface?

It would seem to me that this would even have a (slight)
disadvantage. Eg. by effectively hard-wiring a function or method to
deal with only objects that adhere to a certain _inherited_ "fooable"
interface, you prevent it from handling an object that is "fooable" in
its own right, without inheriting that behaviour from somewhere
else. (Which leads me to suspect that this is the whole point ;-)
Perhaps we're just on different sides of the fence as to whether this
is desirable or not?)

Cheers,
P.


From sholden at holdenweb.com  Thu Apr 11 12:41:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 11 Apr 2002 12:41:34 -0400
Subject: problem zipping a text file with zlib
References: <ub3gp3lfn2rc3d@corp.supernews.com> <mailman.1018294997.29344.python-list@python.org> <ub5s4paqolpo0a@corp.supernews.com> <3CB2FE25.86A37EA1@imk.fraunhofer.de> <ub6dji5plutpc5@corp.supernews.com> <SKYs8.19881$GM.6702@atlpnn01.usenetserver.com> <ubb56d6fvmpjf0@corp.supernews.com>
Message-ID: <93jt8.1445$F43.775@atlpnn01.usenetserver.com>

"DeepBleu" <DeepBleu at DeepBleu.org> wrote in message
news:ubb56d6fvmpjf0 at corp.supernews.com...
>
[ ... ]
> >
> > Do you close the zipfile after you've finished writing it?
> >
> Do you mean text1.zip (from the code above) as the zipfile to be closed?
> No, I am not!  I thought it was enough to close the ZipFile instance
> (zip_archive).  Since text1.zip makes its appearance (first used) in:
> >>>zip_archive = ZipFile("text1.zip", "w", 8)
> How do I close the text1.zip (if that is the one you are refering to)?
> DeepBleu
>
Ignore me. It was obvious in the code you gave that you are indeed correctly
closing the file after writing it.

Since the zipfile format is binary, you should *not* be running Unix2DOS to
convert it - you are probably messing up the file format by doing that, and
may well find everything works fine if you just transfer the file across to
Windows as a binary image.

Unix2DOS will translate LF in the input into CR/LF in the output. WinZip (or
whatever) won't expect that, and your file will be hosed. The file that
works possibly just doesn't compress down to something containing newlines.


regards
 Steve






From sjmachin at lexicon.net  Wed Apr 24 22:06:44 2002
From: sjmachin at lexicon.net (John Machin)
Date: 24 Apr 2002 19:06:44 -0700
Subject: Why don't strings share data in Python?
References: <87d6x09trz.fsf@mathdogs.com>
Message-ID: <c76ff6fc.0204241806.22f3b0d8@posting.google.com>

Mike Coleman <mkc+dated+1021521407.f909ec at mathdogs.com> wrote in message news:<87d6x09trz.fsf at mathdogs.com>...
> Does anyone know why strings (i.e., those of length >1) don't share their data
> in Python?  Since their immutable, it seems like this would be the obvious
> thing to do.  So, for example, the space behavior of this code could be linear
> rather than quadratic/horrific:
> 
> d = {}
> for i in xrange(100000000):
>     d[mybigstring[i:]] = mybigstring[i:]

You might like to:

0. Try this:

d = {}
for i in xrange(100000000):
   suffix = mybigstring[i:]
   d[suffix] = suffix

My guess is that this will create only one copy of each slice, instead
of two.

1. Explain the utility of having a dict d such that d[foo] == foo, for
*all* foo in d. If all you want is logically a set, not a mapping, do
d[suffix] = 1.

2. google("suffix tree"); google("suffix array")


From tejarex at yahoo.com  Tue Apr 16 23:15:22 2002
From: tejarex at yahoo.com (Terry Reedy)
Date: Wed, 17 Apr 2002 03:15:22 GMT
Subject: Simulating multi-dim array problem - OR - reference confusions
References: <3CBC991C.7090202@nyu.edu>
Message-ID: <eN5v8.60521$3L2.5219559@bin7.nnrp.aus1.giganews.com>

"Shane Hoversten" <srh232 at nyu.edu> wrote in message
news:3CBC991C.7090202 at nyu.edu...
> 1) Am I going to get into trouble assuming (using the example above)
that
> b is basically a reference to the sublist a[0]?  I'm not clear how
> sharing works in python.

Someone else answered this.  If still have ??s, ask again.

> 2) Is there a way I can print out the address of a "reference"?

id(ob) return unique int associated with ob, which in CPython happens
to be its address.  Useful for finding out what is going on with
references (learning).  Used in representation of some objects - which
you can do also.

Terry J. Reedy





From johnroth at ameritech.net  Mon Apr  8 08:05:30 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 8 Apr 2002 08:05:30 -0400
Subject: PEP 285: Adding a bool type
References: <mailman.1017466974.6751.python-list@python.org> <7xit79yxd5.fsf@ruckus.brouhaha.com> <IDAq8.45703$pT1.1295831@news1.tin.it> <ualqa3n4dlof64@news.supernews.com> <gJDq8.55255$S52.1904212@news2.tin.it>  <uau3rsge16bp1e@news.supernews.com> <mailman.1018111094.10533.python-list@python.org> <uavbgcabvl642b@news.supernews.com> <a8pqkh$492$1@slb1.atl.mindspring.net> <ub1mp3ssjc7da5@news.supernews.com> <a8qog4$900$1@slb1.atl.mindspring.net>
Message-ID: <ub31hhkk857a52@news.supernews.com>

"Andrew Dalke" <dalke at dalkescientific.com> wrote in message
news:a8qog4$900$1 at slb1.atl.mindspring.net...
> John Roth replied to a comment of mine.  As my response was even
> more off-topic, I sent email to him, which bounced, saying:
>
> ] The following destination addresses were unknown (please check the
> ] addresses and re-mail the message):
> ]
> ] <johnroth at ameritech.net>
> ]
> ] Please reply to techsupport at ameritech.net
> ] if you feel this message to be in error.
>
>
> John?  Did you know this?  (It doesn't look spam protected, after
> all.)  If you want the reply from me, could you send me a valid
> email address?
>
>                     Andrew
>                     dalke at dalkescientific.com

I haven't been at ameritech for over a year, which is why
any email responses from newsgroups bounce. I like it that way,
it keeps the spam harvesters at bay.

To respond to me, please send (and I've reversed the address,
for the above reason) to:

moc dot rjhtorhj at newsgroups

I may update my address at supernews, since my new e-mail
provider supposedly has industrial strength spam detection
and elimination.

John Roth
>
>
>




From -$P-W$- at verence.demon.co.uk  Fri Apr 19 07:54:48 2002
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: 19 Apr 2002 12:54:48 +0100
Subject: What webhosting
References: <mailman.1019071656.4705.python-list@python.org>
Message-ID: <a9p0i8$4d3$1@verence.demon.co.uk>

[news.admin.net-abuse.email added and followups set]

In article <mailman.1019071656.4705.python-list at python.org>,
A <printers at sendme.cz> wrote:
>Hi,
>Does anybody know about a webhosting provider ( with Python and 
>Telnet access) that does not remove a user's account if the user 
>sends occasionally some unsolicited emails?

Spam is bad. Don't do it.

n.a.n-a.e people: Heads up for you on Mr Ladislav here. See also
http://groups.google.com/groups?selm=mailman.1019071656.4705.python-list%40python.org&output=gplain
for the original article.

-- 
Paul Wright | http://pobox.com/~pw201 |


From pj at engr.sgi.com  Fri Apr  5 20:13:19 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 6 Apr 2002 01:13:19 GMT
Subject: Booleans, integer division, backwards compatibility; where is Python going?
References: <mailman.1018031177.21308.python-list@python.org> <c76ff6fc.0204051701.68c63ecb@posting.google.com>
Message-ID: <a8li3f$ovbmd$1@fido.engr.sgi.com>

|> what's this 8-track story?

Think music, not data.  8-track tapes were popular in
automobiles, before cassette tapes.

-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373