From maccgi@bellsouth.net  Fri Oct  1 00:23:54 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Thu, 30 Sep 1999 19:23:54 -0400
Subject: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1
In-Reply-To: <v04011703b41931a3507b@[128.101.80.22]>
References: <v04011703b41931a3507b@[128.101.80.22]>
Message-ID: <v04205501b419a0a09680@[209.214.79.127]>

At 10:31 -0500 09/30/1999, Brian J. Stankiewicz wrote:
>I know that this is a bit of a newbie question, but I am trying to install
>MacPython on my 8500 with OS 8.0 and I am having a difficult time figuring
>out how to get the Tk commands to work.  Here is a short session with the
>search paths included, but when I try to import Tkinter it tries to find
>_tkinter and cannot find the module.  I found _tkinter.ppc.slib and placed
>a re-named copy _tkinter in the extensions folder, but this didn't do the
>trick either.  Could someone please give me a hand on this?

I'm not sure what the preferred approach to this is, but I ran into 
the same thing and fixed it by using  EditPythonPrefs to add 
$(PYTHON):Extensions:Imaging:PIL: to my sys.path. I don't recall 
having to do that in the past, but at least it worked for me.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From btcoburn@oberlin.edu  Fri Oct  1 04:19:02 1999
From: btcoburn@oberlin.edu (Ben Coburn)
Date: Thu, 30 Sep 1999 23:19:02 -0400 (EDT)
Subject: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1
In-Reply-To: <v04205501b419a0a09680@[209.214.79.127]>
Message-ID: <Pine.OSF.3.96.990930231111.18449A-100000@mercury.cc.oberlin.edu>

On Thu, 30 Sep 1999, Richard Gordon wrote:

> At 10:31 -0500 09/30/1999, Brian J. Stankiewicz wrote:
> >I know that this is a bit of a newbie question, but I am trying to install
> >MacPython on my 8500 with OS 8.0 and I am having a difficult time figuring
> >out how to get the Tk commands to work.  Here is a short session with the
> >search paths included, but when I try to import Tkinter it tries to find
> >_tkinter and cannot find the module.  I found _tkinter.ppc.slib and placed
> >a re-named copy _tkinter in the extensions folder, but this didn't do the
> >trick either.  Could someone please give me a hand on this?
> 

I distinctly remember having to add $(PYTHON):Lib:lib-tk to the search
path when I installed 1.5.2c1. You should not have to move any files. 

                                         Regards, Ben Coburn

-----------------------------------------------------------------------
Ben Coburn     "Black holes are where God divided by zero." -Anon.
 btcoburn@oberlin.edu                        Offsite Paradox Associate
 Biochemistry & Biology Major                Arcosanti Paradox Project
 Oberlin College                     http://www.arcosanti.org/paradox/
-----------------------------------------------------------------------




From bstankie@eye.psych.umn.edu  Fri Oct  1 04:51:44 1999
From: bstankie@eye.psych.umn.edu (Brian J. Stankiewicz)
Date: Thu, 30 Sep 1999 22:51:44 -0500
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python Crashes
Message-ID: <v04011702b419df474d7d@[207.58.29.184]>

Is there a reason why running the code on pg. 23 of Learning Python crashes
Python?  I am able to get the dialog box up just fine, but when I hit the
button titled "Hello", the program completely freezes and I have to "Kill"
(Option-Apple-esc) out of Python.

It won't even let me copy out of the command window.  I am going to retype
the trace-back for everyone's pleasure:

>>> from Tkinter import *
>>> w=Button(text="Hello",command='exit')
>>> w.pack()
>>> w.mainloop()
Traceback (innermost last):
  File "<input>", line 1, in ?
  File "Zamboo:Applications:Programming:Python
1.5.2c1:Lib:lib-self.tk.mainloop(n)
  File "Zamboo:Applications:Programming:Python
1.5.2c1:Lib:lib-tk:Tkinter.py",line 75 in _exit
SystemExit: 0

The Tkwindow is blank, but it is still there.  Any ideas?

-Brian
*-----------------------------------------------*
| Brian J.Stankiewicz 		(o)612-624-7353	|
| 75 East River Rd. 		(f)612-626-2079 |
| Elliot Hall					|
| University of Minnesota 			|
| Minneapolis, Minnesota 55455			|
|-----------------------------------------------|
|http://vision.psych.umn.edu/www/people/brian	|
|	bstankie@eye.psych.umn.edu		|
*-----------------------------------------------*


From bstankie@eye.psych.umn.edu  Fri Oct  1 04:47:14 1999
From: bstankie@eye.psych.umn.edu (Brian J. Stankiewicz)
Date: Thu, 30 Sep 1999 22:47:14 -0500
Subject: [Pythonmac-SIG] Tk installation fix
Message-ID: <v04011700b419dcebbf75@[207.58.29.184]>

Thanks to Ben Coburn and Richard Gordon I was able to get Tk to work on my
mac.  I fixed the problem by placing the following path's in the Python
Preferences:

$(PYTHON):Lib:lib-tk
$(PYTHON):Extensions:Imaging:PIL

Is there a reason that these two paths are not automatically set during the
installation.  I was a bit frustrated when I couldn't get through the
example code in Chapter 1 of Learning Python (pg 23 has a Tk sample).

-Brian
*-----------------------------------------------*
| Brian J.Stankiewicz 		(o)612-624-7353	|
| 75 East River Rd. 		(f)612-626-2079 |
| Elliot Hall					|
| University of Minnesota 			|
| Minneapolis, Minnesota 55455			|
|-----------------------------------------------|
|http://vision.psych.umn.edu/www/people/brian	|
|	bstankie@eye.psych.umn.edu		|
*-----------------------------------------------*


From ita.mis@euro.apple.com  Fri Oct  1 11:09:54 1999
From: ita.mis@euro.apple.com (Maero Tasselli)
Date: Fri, 01 Oct 1999 12:09:54 +0200
Subject: [Pythonmac-SIG] python 1.5.2c1 performance
Message-ID: <37F48870.2FE528C@euro.apple.com>

Hi,

I've read some comments about the degradation in performance for python
1.5.2c1 over 1.5.2b1.

I never used the beta version but, for those interested, here is the
pystone benchmark comparing version 1.5.1 and 1.5.2c1:


PowerMac 9600/MP 112MB/RAM MACOS/8.6


Python 1.5.1 (20MB RAM partition)

Pystone(1.1) time for 10000 passes = 2.73333
This machine benchmarks at 3658.54 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.7
This machine benchmarks at 3703.7 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.73333
This machine benchmarks at 3658.54 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.65
This machine benchmarks at 3773.58 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.76667
This machine benchmarks at 3614.46 pystones/second
>>>


Python 1.5.2c1 (20MB RAM partition)

Pystone(1.1) time for 10000 passes = 2.75
This machine benchmarks at 3636.36 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.71667
This machine benchmarks at 3680.98 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.73333
This machine benchmarks at 3658.54 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.73333
This machine benchmarks at 3658.54 pystones/second
>>>
Pystone(1.1) time for 10000 passes = 2.71667
This machine benchmarks at 3680.98 pystones/second
>>>


Regards
Tasselli Marco



From jhrsn@pop.pitt.edu  Fri Oct  1 13:21:18 1999
From: jhrsn@pop.pitt.edu (Jim Harrison)
Date: Fri, 1 Oct 1999 08:21:18 -0400
Subject: [Pythonmac-SIG] Advanced Editor patch -- change line endings
Message-ID: <199910011216.IAA25128@post-ofc05.srv.cis.pitt.edu>

>OK, here's a patch (at http://www.strout.net/python/mac/) which 
>extends the IDE's editor in a couple of ways:

For newbies like me--remember to change the line endings to Mac after the 
files are downloaded, at least when downloading by dragging out of the 
Navigator window. BBedit or BBedit Lite works well for this. Bad things 
happen if you run these patches with unix line endings.

Once that was done...this is awesome! Many thanks to Joe, Just, Jack and 
others who have worked on this. Also, function? and inspect() are great 
for those of use who are climbing the learning curve.
Similar functionality with an option-click on the name in the IDE editor 
might be worth considering.

Jim Harrison

________________________________________________________________________
James H. Harrison, Jr., MD, PhD
Associate Director of Pathology Informatics, Department of Pathology
Faculty Member in Residence, Center for Biomedical Informatics
University of Pittsburgh Health System                     
C920 PUH, 200 Lothrop Street                                  
Pittsburgh, PA 15213

jhrsn@pop.pitt.edu | voice: 412-647-5529 | fax: 412-647-9588

"If you want sense, you'll have to make it yourself!!"-Norton Juster
________________________________________________________________________



From joe@strout.net  Fri Oct  1 15:50:19 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 1 Oct 1999 07:50:19 -0700
Subject: [Pythonmac-SIG] Advanced Editor patch -- change line endings
In-Reply-To: <199910011216.IAA25128@post-ofc05.srv.cis.pitt.edu>
References: <199910011216.IAA25128@post-ofc05.srv.cis.pitt.edu>
Message-ID: <v04210103b41a79d6b019@[198.202.70.213]>

At 8:21 AM -0400 10/01/99, Jim Harrison wrote:

>For newbies like me--remember to change the line endings to Mac after the
>files are downloaded, at least when downloading by dragging out of the
>Navigator window. BBedit or BBedit Lite works well for this. Bad things
>happen if you run these patches with unix line endings.

This is a good point.  There's a free tool called Cyclone which will 
do drag-and-drop conversion once you've configured it; I've put a 
note about it on my web page.

>Once that was done...this is awesome! Many thanks to Joe, Just, Jack and
>others who have worked on this. Also, function? and inspect() are great
>for those of use who are climbing the learning curve.

Good, glad it's proving useful.

>Similar functionality with an option-click on the name in the IDE editor
>might be worth considering.

That's an excellent idea.  OK, so the click modifiers we'll be having soon are:

option-click: look for docstring on the clicked function or class
command-click: copy the clicked chunk of text to the current cursor location

Sounds good to me!

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jack@oratrix.nl  Fri Oct  1 16:49:51 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 01 Oct 1999 17:49:51 +0200
Subject: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1
In-Reply-To: Message by "Brian J. Stankiewicz" <bstankie@eye.psych.umn.edu> ,
 Thu, 30 Sep 1999 10:31:07 -0500 , <v04011703b41931a3507b@[128.101.80.22]>
Message-ID: <19991001154951.C0E7A36E440@snelboot.oratrix.nl>

> I know that this is a bit of a newbie question, but I am trying to install
> MacPython on my 8500 with OS 8.0 and I am having a difficult time figuring
> out how to get the Tk commands to work.  Here is a short session with the
> search paths included, but when I try to import Tkinter it tries to find
> _tkinter and cannot find the module.  I found _tkinter.ppc.slib and placed
> a re-named copy _tkinter in the extensions folder, but this didn't do the
> trick either.  Could someone please give me a hand on this?

Brian,
if I remember correctly Tk is an optional component, you did install it, did 
you? It could well be that if you _don't_ install it only the _tkinter module 
is not included but the Tkinter and friends are, giving the impression that 
you do have Tk installed.

And where did you find the _tkinter.ppc.slb file? It should be in the PlugIns 
folder, if it was installed elsewhere that is the problem (mine, to be 
exact;-).
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From joe@strout.net  Fri Oct  1 16:51:39 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 1 Oct 1999 08:51:39 -0700
Subject: [Pythonmac-SIG] PyAdvancedEditor.py update
Message-ID: <v04210107b41a889a28bc@[198.202.70.213]>

Hi all,

	I've updated PyAdvancedEditor.  It now sports the following 
additional features:

Better Chunking: The standard WASTE word-finding function isn't very 
appropriate for Python code; it breaks on underscores, doesn't break 
on periods, etc. AdvancedEditor gives you more Python-savvy chunks of 
text when you double-click. In addition, you can double-click a 
grouping character (a parenthesis, square bracket, curly brace, or 
quotation mark) to select the entire range delimited by that 
character. Note that the finding of quoted strings may not work 
correctly when triple-quoted strings are involved.

Option Click -- Function Look-up: This feature looks up the function 
or class name you option-click on, without moving the cursor.

Command Click -- Insert Chunk: When you command-click, the chunk of 
text at the location clicked is copied to the current cursor 
location. It's equivalent to double-clicking, doing a Copy, putting 
the cursor back to its previous location, and Paste (except that the 
clipboard is not modified). Because it uses the same chunking 
algorithm as double-clicking, this works nicely with grouping 
characters (like parentheses) too.


	To get these, all you need to do is download 
PyAdvancedEditor.py, fix the line endings, drop it in the 
Mac:Tools:IDE folder, and run BuildIDE.py.  You do NOT need to 
reapply any patches.

Cheers,
-- Joe
,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jack@oratrix.nl  Fri Oct  1 17:00:59 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 01 Oct 1999 18:00:59 +0200
Subject: [Pythonmac-SIG] Advanced Editor patch
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Thu, 30 Sep 1999 13:49:32 -0700 , <v04210106b4197c8281a8@[198.202.70.213]>
Message-ID: <19991001160059.D5DE736E440@snelboot.oratrix.nl>

> Function Look-up:
> 
> This feature looks for doc strings for a function or class when you 
> type a question mark ("?") after the function name in an editor 
> window. For example, go into any IDE editor window at type "string?". 
> The string.__doc__ will be printed to the output window. It also 
> works for functions, e.g. "string.strip?".  When the query is 
> successful, the question mark is absorbed (not entered into the 
> editor). If you really need to type the question mark after a known 
> function, you'll have to type it first after a space, and then delete 
> the space.

Joe,
this is a great feature, but a rather hacky implementation, I think. Why not 
use command-questionmark or something like that? (Command-questionmark may be 
bad because it's bound to help) And also add it to the menu, what the command 
should do is assume the cursor in the frontmost window is at the end of (or in 
the middle of) the thing you want to look up.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Fri Oct  1 17:04:05 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 01 Oct 1999 18:04:05 +0200
Subject: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1
In-Reply-To: Message by Richard Gordon <maccgi@bellsouth.net> ,
 Thu, 30 Sep 1999 19:23:54 -0400 , <v04205501b419a0a09680@[209.214.79.127]>
Message-ID: <19991001160520.8968136E440@snelboot.oratrix.nl>

> I'm not sure what the preferred approach to this is, but I ran into 
> the same thing and fixed it by using  EditPythonPrefs to add 
> $(PYTHON):Extensions:Imaging:PIL: to my sys.path. I don't recall 
> having to do that in the past, but at least it worked for me.

Ah, luckily that's Justs fault not mine (and he isn't here to contradict 
me:-). For reasons you don't want to know there are two _tkinters in what I 
send to him, and the one from PIL is what should be installed in the PlugIns 
folder. Apparently the installer missed that this time and puts it in the PIL 
folder.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From joe@strout.net  Fri Oct  1 17:18:56 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 1 Oct 1999 09:18:56 -0700
Subject: [Pythonmac-SIG] Advanced Editor patch
In-Reply-To: <19991001160059.D5DE736E440@snelboot.oratrix.nl>
References: <19991001160059.D5DE736E440@snelboot.oratrix.nl>
Message-ID: <v0421010db41a8e5180a2@[198.202.70.213]>

At 6:00 PM +0200 10/01/99, Jack Jansen wrote:

>this is a great feature, but a rather hacky implementation, I think. Why not
>use command-questionmark or something like that? (Command-questionmark may be
>bad because it's bound to help)

Actually, it's not bound to anything as far as I can see.  Maybe that 
would be a good solution.

Another possibility would be something more analogous to what MCL 
does: do it whenever when you type an open-paren after a function 
name.  Let the paren go through in any case (making it therefore 
somewhat less hackish than the current interface), but print out a 
docstring when you can.  Naturally you'd want to be able to turn this 
feature off, so as to not clutter up your output window...

Hmm, nope, I like Jack's idea better.  Anybody else have an opinion on this?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From schliep@zpr.uni-koeln.de  Fri Oct  1 17:24:26 1999
From: schliep@zpr.uni-koeln.de (Alexander Schliep)
Date: Fri, 1 Oct 1999 18:24:26 +0200
Subject: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1
In-Reply-To: Jack Jansen <jack@oratrix.nl>
 "Re: [Pythonmac-SIG] Problems with Tk using MacPython vers 1.5.2c1" (Oct  1,  5:49pm)
References: <19991001154951.C0E7A36E440@snelboot.oratrix.nl>
Message-ID: <9910011824.ZM21520@Octopussy.MI.Uni-Koeln.DE>

On Oct 1,  5:49pm, Jack Jansen wrote:
> Brian,
> if I remember correctly Tk is an optional component, you did install it, did

Nope. Tk is part of the distribution (as it should be). Is that Tk 8.2 btw ?

> you? It could well be that if you _don't_ install it only the _tkinter module
> is not included but the Tkinter and friends are, giving the impression that
> you do have Tk installed.
>
> And where did you find the _tkinter.ppc.slb file? It should be in the PlugIns
> folder, if it was installed elsewhere that is the problem (mine, to be
> exact;-).

I had to add $(PYTHON):Lib:lib-tk for import Tkinter to work
(which is a bug, right?) and then $(Python):Extensions:Imaging:PIL
for _tkinter to be found.

Could the default Python Prefs please be changed such that Tk works out of
the box ?

Yours,
Alexander

-- 
Alexander Schliep                    schliep@zpr.uni-koeln.de
ZPR/ZAIK                             Tel: +49-221-470-6011 (w)
University of Cologne                FAX: +49-221-470-5160
Weyertal 80                          http://www.zpr.uni-koeln.de/~schliep
50931 Cologne, Germany               


From eddthompson@ssi.parlorcity.com  Fri Oct  1 17:52:50 1999
From: eddthompson@ssi.parlorcity.com (EuphoriaDJ)
Date: Fri, 1 Oct 1999 11:52:50 -0500
Subject: [Pythonmac-SIG] python mode for alpha
In-Reply-To: <199910010503.BAA17829@python.org>
Message-ID: <l03130300b41a96ede52f@[206.137.30.80]>

I was digging through pythonftp and saw a listing for a mode for alpha but
it is only the name not the file, does anyone know where I could get the
file?

An Elephant: A mouse built to government specifications
Never try to out stubborn a cat
Natural laws have no pity

TTFN
http://www.parlorcity.com/edd/index.html




From maccgi@bellsouth.net  Fri Oct  1 18:09:48 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Fri, 1 Oct 1999 13:09:48 -0400
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python
 Crashes
In-Reply-To: <v04011702b419df474d7d@[207.58.29.184]>
References: <v04011702b419df474d7d@[207.58.29.184]>
Message-ID: <v04205500b41a99cd445b@[209.214.79.127]>

At 22:51 -0500 09/30/1999, Brian J. Stankiewicz wrote:
>Is there a reason why running the code on pg. 23 of Learning Python crashes
>Python?  I am able to get the dialog box up just fine, but when I hit the
>button titled "Hello", the program completely freezes and I have to "Kill"
>(Option-Apple-esc) out of Python.
>
>It won't even let me copy out of the command window.  I am going to retype
>the trace-back for everyone's pleasure:

I get pretty much the same results including the freeze, but I do see 
a Hello button. I don't know what the story is here, but noticed that 
before running this, the author launches python with a startup script 
called runpy, so that may have something to do with it. I haven't 
experimented, but you might study the preceding pages to see what 
runpy is all about.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From joe@strout.net  Fri Oct  1 19:14:25 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 1 Oct 1999 11:14:25 -0700
Subject: [Pythonmac-SIG] Advanced Editor patch
In-Reply-To: <199910011729.NAA21902@post-ofc06.srv.cis.pitt.edu>
References: <199910011729.NAA21902@post-ofc06.srv.cis.pitt.edu>
Message-ID: <v04210102b41aa981e45e@[198.202.70.213]>

At 1:34 PM -0400 10/01/99, Jim Harrison wrote:

>When the ICeTEe extension is installed, command-clicks are trapped and
>the clicked text is examined for a URL. If present, the URL is fed to the
>browser indicated in Internet Config. This could cause a conflict with
>your plans for command-clicks in the IDE.

Oops, you're right, that's what it does.  Hmph, I prefer the 
insert-chunk function, but command-click is something of a standard. 
Though it makes more sense in things like newsreaders and email apps 
than in a code editor.

Glenn has suggested making these functions user-configurable, but I 
haven't thought of a good way to do that without increasing the 
complexity overmuch.

Well, I'm open to suggestions here... at a minimum, I'll set up a 
little dictionary at the top of PyAdvancedEditor.py which maps the 
various modifier combinations to actions, so that you can rearrange 
the mappings that way.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From bstankie@eye.psych.umn.edu  Fri Oct  1 22:00:21 1999
From: bstankie@eye.psych.umn.edu (Brian J. Stankiewicz)
Date: Fri, 1 Oct 1999 16:00:21 -0500
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python Crashes
In-Reply-To: <v04205500b41a99cd445b@[209.214.79.127]>
References: <v04011702b419df474d7d@[207.58.29.184]>
 <v04011702b419df474d7d@[207.58.29.184]>
Message-ID: <v04011700b41ad08826c5@[128.101.80.22]>

Hi Richard and MP-SIG,

Okay, I have done some more testing ont he problem and here are some more
symptoms that I am finding.  If I run one of the Demos
(Demo:tkinter:Matt:Canvas-demo-sipmple.py) from the IDE I get the tk window
displayed, but when I hit the "quit" button the Python completely freezes.
However, if I run the same program by simply double-clicking on the .py
file and running it in the Python interpretor only, it runs just fine and
will quit gracefully.

Are Richard and I the only ones suffering with this?  Is this a problem
with the IDE?

-Brian

At 1:09 PM -0400 10/1/99, Richard Gordon wrote:
>At 22:51 -0500 09/30/1999, Brian J. Stankiewicz wrote:
>>Is there a reason why running the code on pg. 23 of Learning Python crashes
>>Python?  I am able to get the dialog box up just fine, but when I hit the
>>button titled "Hello", the program completely freezes and I have to "Kill"
>>(Option-Apple-esc) out of Python.
>>
>>It won't even let me copy out of the command window.  I am going to retype
>>the trace-back for everyone's pleasure:
>
>I get pretty much the same results including the freeze, but I do see
>a Hello button. I don't know what the story is here, but noticed that
>before running this, the author launches python with a startup script
>called runpy, so that may have something to do with it. I haven't
>experimented, but you might study the preceding pages to see what
>runpy is all about.
>
>
>Richard Gordon
>--------------------
>Gordon Consulting & Design
>Database Design/Scripting Languages
>mailto:richard@richardgordon.net
>http://www.richardgordon.net
>770.971.6887 (voice)
>770.216.1829 (fax)

____________________________________________________________
Brian J.Stankiewicz, Ph.D.      |bstankie@eye.psych.umn.edu
75 East River Rd.               |
Department of Psychology        |
University of Minnesota         |(Off)612-624-7353
Minneapolis, Minnesota 55455    |(Fax)612-626-2079
________________________________|___________________________
	http://vision.psych.umn.edu/www/people/brian
____________________________________________________________


From maccgi@bellsouth.net  Fri Oct  1 23:03:03 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Fri, 1 Oct 1999 18:03:03 -0400
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python
 Crashes
In-Reply-To: <v04210102b41adea08adc@[198.202.70.213]>
References: <v04011702b419df474d7d@[207.58.29.184]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011700b41ad08826c5@[128.101.80.22]>
 <v04210102b41adea08adc@[198.202.70.213]>
Message-ID: <v04205502b41adf9dab68@[209.214.65.76]>

At 14:58 -0700 10/01/1999, Joseph J. Strout wrote:
>This is not too surprising.  The IDE has its own menu bar.  Tk has 
>its own menu bar.  When you run a Tk program from within the IDE, 
>Tk's menu bar clobbers that of the IDE.  When the Tk program quits, 
>there is no longer a valid menu bar, and the Mac goes off in search 
>of one and never comes back.  ;)  (Actually, it probably just shoots 
>itself.)

Oh yeah, I forgot about that, but the book stuff that Brian 
originally mentioned breaks regardless of how it's run, so there's 
something else going on with that as well.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From joe@strout.net  Fri Oct  1 22:58:20 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 1 Oct 1999 14:58:20 -0700
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python
 Crashes
In-Reply-To: <v04011700b41ad08826c5@[128.101.80.22]>
References: <v04011702b419df474d7d@[207.58.29.184]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011700b41ad08826c5@[128.101.80.22]>
Message-ID: <v04210102b41adea08adc@[198.202.70.213]>

At 4:00 PM -0500 10/01/99, Brian J. Stankiewicz wrote:

>Okay, I have done some more testing ont he problem and here are some more
>symptoms that I am finding.  If I run one of the Demos
>(Demo:tkinter:Matt:Canvas-demo-sipmple.py) from the IDE I get the tk window
>displayed, but when I hit the "quit" button the Python completely freezes.
>However, if I run the same program by simply double-clicking on the .py
>file and running it in the Python interpretor only, it runs just fine and
>will quit gracefully.

This is not too surprising.  The IDE has its own menu bar.  Tk has 
its own menu bar.  When you run a Tk program from within the IDE, 
Tk's menu bar clobbers that of the IDE.  When the Tk program quits, 
there is no longer a valid menu bar, and the Mac goes off in search 
of one and never comes back.  ;)  (Actually, it probably just shoots 
itself.)

Sadly, the IDE is not good for making stand-alone apps which have 
their own menu bar, or for doing just about anything with Tk.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From andres@corrada.com  Sat Oct  2 01:18:14 1999
From: andres@corrada.com (Andres Corrada)
Date: Fri, 01 Oct 1999 20:18:14 -0400
Subject: [Pythonmac-SIG] python mode for alpha
References: <l03130300b41a96ede52f@[206.137.30.80]>
Message-ID: <37F54DBE.656034B3@corrada.com>

I have the sit file for the alpha-mode if no one volunteers a site to download
the file from. The mode is not maintained anymore by its author but it does a
respectable job.

-------------------------------------------------------
Andres Corrada-Emmanuel    Email: andres@corrada.com
Owner                      http://www.corrada.com/mamey
Mamey                      Phone: (413) 587-9595
-------------------------------------------------------


From dozier@abs.net  Sat Oct  2 04:19:27 1999
From: dozier@abs.net (Bill Dozier)
Date: Fri, 01 Oct 1999 23:19:27 -0400
Subject: [Pythonmac-SIG] python mode for alpha
Message-ID: <199910020326.XAA13720@vwww1.abs.net>

> THIS MESSAGE IS IN MIME FORMAT. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--MS_Mac_OE_3021664768_7267989_MIME_Part
Content-type: text/plain; charset="US-ASCII"
Content-transfer-encoding: 7bit

Here is an Alpha python mode written by Howard Oakley 
(howard@quercus.demon.co.uk,  http://www.quercus.demon.co.uk). Since he
posted this to comp.lang.python, I don't imagine he'll mind me offering it
here.

To load it, open the file with Alpha and let the app install the mode.
--MS_Mac_OE_3021664768_7267989_MIME_Part
Content-type: text/plain; name="pythonMode.tcl";
 x-mac-creator="414C4641";
 x-mac-type="54455854"
Content-disposition: attachment
Content-transfer-encoding: base64
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--MS_Mac_OE_3021664768_7267989_MIME_Part--



From stank005@tc.umn.edu  Sat Oct  2 16:51:14 1999
From: stank005@tc.umn.edu (Kimlynnae Pleticha)
Date: Sat, 2 Oct 1999 10:51:14 -0500
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python Crashes
In-Reply-To: <v04210102b41adea08adc@[198.202.70.213]>
References: <v04011700b41ad08826c5@[128.101.80.22]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011700b41ad08826c5@[128.101.80.22]>
Message-ID: <v04011700b41bd755b787@[207.58.29.184]>

At 2:58 PM -0700 10/1/99, Joseph J. Strout wrote:
>At 4:00 PM -0500 10/01/99, Brian J. Stankiewicz wrote:
>
>>Okay, I have done some more testing ont he problem and here are some more
>>symptoms that I am finding.  If I run one of the Demos
>>(Demo:tkinter:Matt:Canvas-demo-sipmple.py) from the IDE I get the tk window
>>displayed, but when I hit the "quit" button the Python completely freezes.
>>However, if I run the same program by simply double-clicking on the .py
>>file and running it in the Python interpretor only, it runs just fine and
>>will quit gracefully.
>
>This is not too surprising.  The IDE has its own menu bar.  Tk has
>its own menu bar.  When you run a Tk program from within the IDE,
>Tk's menu bar clobbers that of the IDE.  When the Tk program quits,
>there is no longer a valid menu bar, and the Mac goes off in search
>of one and never comes back.  ;)  (Actually, it probably just shoots
>itself.)

Hmmm.  I have to admit that this IS surprising to me.  The IDE is a great
environment to help debug code in, but it seems to me that it should be a
bit more stable. Is there a way to create Tk code so that it doesn't
clobber the IDE menu bar when it quits?  Or is there a way to change the
code in the IDE so that when running other Tk code the menu bar will not
get clobbered?   What methods have others used for writing Python Tk code?

Thanks,
-Brian


From joe@strout.net  Sat Oct  2 20:53:10 1999
From: joe@strout.net (Joseph J. Strout)
Date: Sat, 2 Oct 1999 12:53:10 -0700
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python
 Crashes
In-Reply-To: <v04011700b41bd755b787@[207.58.29.184]>
References: <v04210102b41adea08adc@[198.202.70.213]>
 <v04011700b41ad08826c5@[128.101.80.22]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011702b419df474d7d@[207.58.29.184]>
 <v04011700b41ad08826c5@[128.101.80.22]>
Message-ID: <l03102802b41c11ac8220@[198.202.70.213]>

At 8:51 AM -0700 10/2/99, Kimlynnae Pleticha wrote:

>Hmmm.  I have to admit that this IS surprising to me.  The IDE is a great
>environment to help debug code in, but it seems to me that it should be a
>bit more stable.

Technically, this isn't an instability in the IDE -- it's a design
limitation, but not "instability" when you run code that deliberately
clobbers it (e.g., by overwriting its menus).

In my experience, it's Tk that's instable.  But I confess it's been a while
since I tried it.

>Is there a way to create Tk code so that it doesn't
>clobber the IDE menu bar when it quits?

I don't know.  Tk on the Mac is pretty nasty no matter how you look at it,
I think.

> Or is there a way to change the code in the IDE so that when running
>other Tk
>code the menu bar will not get clobbered?

Running Tk will probably always be hazardous, but I do wish you could more
easily write stand-alone Mac programs in Python with the IDE.  For this to
work, you'd have to do something like what the codewarrior IDE does: launch
the program as a separate process, and communicate with it via AppleEvents.

Seems like it wouldn't be that hard to add... we could put a "Run
Separately" option in the pop-up menu next to "Run As Main".  When checked,
it would launch a slave app, and pass it the code.  The really hard part
would be the debugger, though... hard to see how that would work.

> What methods have others used for writing Python Tk code?

You could use the BBEdit "Run As Python" plug-in with BBEdit or BBEdit
Lite.  This used to be included with MacPython; I'm not sure if it's still
there or not.

But why are you so determined to use Tk?  If your objective is to learn
Python, this is a painful way to do it.  What is it you really want to do?
If it's just to play around with graphics, I'd suggest you try PIDDLE
instead; it works perfectly well within the IDE.
(http://www.strout.net/python/piddle/)

Cheers,
-- Joe


,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'




From jack@oratrix.nl  Mon Oct  4 14:13:39 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 04 Oct 1999 15:13:39 +0200
Subject: [Pythonmac-SIG] MacPython and Tk
In-Reply-To: Message by "Brian J. Stankiewicz" <bstankie@eye.psych.umn.edu> ,
 Fri, 1 Oct 1999 16:00:21 -0500 , <v04011700b41ad08826c5@[128.101.80.22]>
Message-ID: <19991004131340.2E06436E440@snelboot.oratrix.nl>

Hmm. As various people have noted recently the quality-control on Tkinter with 
MacPython leaves quite a lot to be desired:
- The installer doesn't install _tkinter in the right place
- Tkinter programs do not work from the IDE
- Simple Tk programs from "Learning Python" hang your system.

Of all these only the second one (Tk and IDE don't like each other) is a known 
problem, and I think it is documented (but Just isn't around, or he'd know).

The general problem is that Just and me are the main people responsible for 
MacPython, and neither of us uses Tkinter anymore (at least, not on the mac). 
Combine this with the fact that we're both really so busy that even getting 
out Python distributions in the first place is a problem and it becomes clear 
why the Tk part of the distribution doesn't have the quality it should. This 
is especially bad because Tk is used in early examples, and so it'll be 
something that is exposed to people new to MacPython in a rather early stage 
(thereby making MacPython as a whole look bad).

What I'd really like is if someone stepped in and volunteered to maintain the 
mac version of the whole Tk stuff, ideally someone who uses Tkinter a lot 
anyway, so bugs are bound to show up quickly. And, of course, you'd need 
CodeWarrior to build the stuff, etc.

Is anyone willing to take up Tkinter maintainance?
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct  4 14:21:33 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 04 Oct 1999 15:21:33 +0200
Subject: [Pythonmac-SIG] "A GUI Test Session" from Learning Python Crashes
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Fri, 1 Oct 1999 14:58:20 -0700 , <v04210102b41adea08adc@[198.202.70.213]>
Message-ID: <19991004132134.1F87E36E440@snelboot.oratrix.nl>

> This is not too surprising.  The IDE has its own menu bar.  Tk has 
> its own menu bar.  When you run a Tk program from within the IDE, 
> Tk's menu bar clobbers that of the IDE.  When the Tk program quits, 
> there is no longer a valid menu bar, and the Mac goes off in search 
> of one and never comes back.  ;)  (Actually, it probably just shoots 
> itself.)

This is fixable, I did it for twit (the now-retired debugger of which some 
ideas have survived in IDE). Just keep two menu bars around, one to be 
installed when the user program runs and one for when IDE runs. IDE knows when 
it goes to user code (and when it comes back) so it can swap them at the right 
moment:
        def switch_to_app(self):
                self.dbg_menu_bar = Menu.GetMenuBar()
                Menu.SetMenuBar(self.app_menu_bar)
                Menu.DrawMenuBar()
                
        def switch_to_dbg(self):
                self.app_menu_bar = Menu.GetMenuBar()
                Menu.SetMenuBar(self.dbg_menu_bar)
                Menu.DrawMenuBar()

Now, if an application (like a Tkinter app) messes with the menu bar that 
shouldn't influence operation of IDE.

If anyone puts such code into the IDE please report back,
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From joe@strout.net  Mon Oct  4 18:07:44 1999
From: joe@strout.net (Joseph J. Strout)
Date: Mon, 4 Oct 1999 10:07:44 -0700
Subject: [Pythonmac-SIG] update to PyAdvancedEditor
Message-ID: <v04210105b41e84ed7f37@[198.202.70.213]>

I've updated PyAdvancedEditor.py again.  This update fixes a 
limitation in the selection of a quoted string by double-clicking one 
of the quotation marks; adds support for launching a URL by holding 
command+option while clicking; adds a little dictionary at the top of 
the file which defines what the modifier keys do, so you can easily 
change them; and also adds more color constants there so you can 
customize the syntax coloring.

To install, grab PyAdvancedEditor.py from 
http://strout.net/python/mac/, put it in your Mac:Tools:IDE folder, 
and re-run BuildIDE.py.

I still haven't changed the interface to the function look-up, but I 
probably will at some point as per Jack's suggestion.

Cheers,
-- Joe
,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From dozier@abs.net  Mon Oct  4 19:30:25 1999
From: dozier@abs.net (Bill Dozier)
Date: Mon, 04 Oct 1999 14:30:25 -0400
Subject: [Pythonmac-SIG] JPython and the IDE
Message-ID: <199910041834.OAA11887@vwww1.abs.net>

Hi,

I've pretty much decided on using JPython on the Mac so that I can do GUIs
without having to learn and wrap PowerPlant or wait for wxMac and hope that
it is complete and as easy to use.

The one catch that I run into is debugging. If I do straight Java, I can use
CW's debugger, if I do straight Python, I can use Just's IDE. However, I
want to use JPython....

Do you sirs think that using the IDE with JPython is doable? (I just tried
running it with JPython and it immediately complains about not finding the
MacOS module...I can't find it either...where is it?) If someone that knows
how the IDE is put together (Just, Joe, ...?) gives me a hint or two, I'll
try to make it work and then let everyone else in on how.

Thanks!

Bill


From cbarker@jps.net  Mon Oct  4 21:46:09 1999
From: cbarker@jps.net (Chris Barker)
Date: Mon, 04 Oct 1999 13:46:09 -0700
Subject: [Pythonmac-SIG] Writing  a Stand Alone GUI program with the IDE
References: <199910030501.BAA18436@python.org>
Message-ID: <37F91211.BE219A37@jps.net>

Hi,

As the discussion has shown, there are people that would like to be able
to use the IDE to develop stand alone Python programs.

It was pointed out the Tk has a lot of other limitations on the Mac, but
I'd like to see any alteration of the IDE allow ANY gui toolkit.
Personally, I'm holding out for WxPython. It is a pretty nice framework.

By the way, is anyone putting any effort into improving the basic
interpreter window? If it worked well, I would be a lot happier working
without the IDE. When I use Python on Linux, I usually edit the code in
emacs, and run it in the interpretter. I end up using "execfile" a lot.

Specific things:   Command line recall and editing!!!
                   cut and paste to the command line.
                   I'd love to see readline and rlcompeter functioning
as well

-Chris



-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------


From draayer@door.net  Tue Oct  5 07:40:39 1999
From: draayer@door.net (Dean E.Draayer)
Date: Tue, 5 Oct 99 01:40:39 -0500
Subject: [Pythonmac-SIG] AskString dialog id error
Message-ID: <19991005064738727.AAA44@mail.door.net@[208.247.125.80]>

In EasyDialogs.py in the definition of AskString, the following line
should be deleted:

    Line 75:  id = 257

Evidently, the id parameter is being initialized with a default argument,
so this line will clobber any user-supplied value for this argument.

Dean Draayer


From jack@oratrix.nl  Tue Oct  5 09:49:37 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Tue, 05 Oct 1999 10:49:37 +0200
Subject: [Pythonmac-SIG] JPython and the IDE
In-Reply-To: Message by "Bill Dozier" <dozier@abs.net> ,
 Mon, 04 Oct 1999 14:30:25 -0400 , <199910041834.OAA11887@vwww1.abs.net>
Message-ID: <19991005084938.2859A35BB1E@snelboot.oratrix.nl>

> Hi,
> 
> I've pretty much decided on using JPython on the Mac so that I can do GUIs
> without having to learn and wrap PowerPlant or wait for wxMac and hope that
> it is complete and as easy to use.
> 
> The one catch that I run into is debugging. If I do straight Java, I can use
> CW's debugger, if I do straight Python, I can use Just's IDE. However, I
> want to use JPython....

Forget about using the IDE: it is firmly embedded in the MacOS world, and none 
of the MacOS GUI calls are available in Java (and, hence in JPython).

Your best bet is to look at what other people use to debug JPython. After all, 
Java pretends to be a machine architecture, so debugging should be the same no 
matter whether your virtual machine runs on Mac, Windows or Unix. Idle is also 
out, I think, as it uses Tk for windowing, but there must be something out 
there that uses Java windowing code...
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Tue Oct  5 09:52:27 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Tue, 05 Oct 1999 10:52:27 +0200
Subject: [Pythonmac-SIG] Writing a Stand Alone GUI program with the IDE
In-Reply-To: Message by Chris Barker <cbarker@jps.net> ,
 Mon, 04 Oct 1999 13:46:09 -0700 , <37F91211.BE219A37@jps.net>
Message-ID: <19991005085228.0516035BB1E@snelboot.oratrix.nl>

> By the way, is anyone putting any effort into improving the basic
> interpreter window? If it worked well, I would be a lot happier working
> without the IDE. When I use Python on Linux, I usually edit the code in
> emacs, and run it in the interpretter. I end up using "execfile" a lot.

One of the reasons Just started with the IDE is that the interpreter output 
window is so lousy. And fixing it was easier in Python than in C.

--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From dmccarthy@ou.edu  Wed Oct  6 15:34:14 1999
From: dmccarthy@ou.edu (David McCarthy)
Date: Wed, 06 Oct 1999 14:34:14 +0000
Subject: [Pythonmac-SIG] Formerly stable app crashes my computer under 1.5.2
Message-ID: <37FB5DE1.9689343C@ou.edu>

I have made an application that uses tkinter.  It was
pretty stable when
run in the 1.5.1 interpreter but it now causes my
computer to crash when
I run it in 1.5.2.  The application itself seems stable
but the computer
crashes after I quit python and open a new application
or bring up a window
from another currently active application.  I have
MacsBug installed and it
gives me a message like the following:

PowerPC unmapped memory exception at FFDBE458
NQDFillCRect+00184

I have also had problems with the popup menus in my
application when the
Appearance control panel is turned on under 1.5.1.

Does anyone have a clue as to what might be going on?
I'm assuming
it might be a problem with tkinter
Thanks
Dave McCarthy




From jhrsn@pop.pitt.edu  Wed Oct  6 21:57:56 1999
From: jhrsn@pop.pitt.edu (Jim Harrison)
Date: Wed, 6 Oct 1999 16:57:56 -0400
Subject: [Pythonmac-SIG] IDE class/method popup?
Message-ID: <199910062055.QAA20754@post-ofc06.srv.cis.pitt.edu>

The popup menu at the bottom left of my IDE script window lists methods 
and classes with popout arrows next to the class names, presumably for 
their contained methods. However, no menus pop out from these arrows. I 
presume that this is a "feature to be implemented," but I'm just checking 
to make sure I don't have a problem with my installation (system 8.6, 
Powerbook G3 "Wallstreet").

Jim Harrison

________________________________________________________________________
James H. Harrison, Jr., MD, PhD
Associate Director of Pathology Informatics, Department of Pathology
Faculty Member in Residence, Center for Biomedical Informatics
University of Pittsburgh Health System                     
C920 PUH, 200 Lothrop Street                                  
Pittsburgh, PA 15213

jhrsn@pop.pitt.edu | voice: 412-647-5529 | fax: 412-647-9588

"If you want sense, you'll have to make it yourself!!"-Norton Juster
________________________________________________________________________



From joe@strout.net  Wed Oct  6 22:03:50 1999
From: joe@strout.net (Joseph J. Strout)
Date: Wed, 6 Oct 1999 14:03:50 -0700
Subject: [Pythonmac-SIG] IDE class/method popup?
In-Reply-To: <199910062055.QAA20754@post-ofc06.srv.cis.pitt.edu>
References: <199910062055.QAA20754@post-ofc06.srv.cis.pitt.edu>
Message-ID: <v04210118b421694e1e5a@[198.202.70.213]>

At 4:57 PM -0400 10/06/99, Jim Harrison wrote:

>The popup menu at the bottom left of my IDE script window lists methods
>and classes with popout arrows next to the class names, presumably for
>their contained methods.

Yep.

> However, no menus pop out from these arrows. I
>presume that this is a "feature to be implemented"...

Nope, it's a feature that's been implemented for several versions now 
(as long as I've been using the IDE, I think).

I have noticed that sometimes the method menus don't appear.  This 
happens rarely, and I haven't been able to spot any particular 
pattern to it.  I wonder if it has to do with low-memory conditions? 
Jim, if you increase your memory allocation to PythonIDE, does the 
problem disappear?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From PClaerhout@CREO.BE  Thu Oct  7 07:59:25 1999
From: PClaerhout@CREO.BE (Pieter Claerhout)
Date: Thu, 7 Oct 1999 08:59:25 +0200
Subject: [Pythonmac-SIG] IDE class/method popup?
Message-ID: <2B1262E83448D211AE4B00A0C9D61B03688563@MSGEURO1>

Hello,

I just had the same problem this week, and check
if you use spaces instead of tabs for indentation.
If you use spaces, the IDE will not show the
methods for the classes If you use tabs, it will
show them. I replaced all spaces in my script files
with tabs and it worked for me (Python 1.5.2b1).

Kind regards,


Pieter 

Pieter Claerhout - PClaerhout@creo.be
Response Center - Applications Support

Creo Europe - Excelsiorlaan 21 - 1930 Zaventem - Belgium
Tel: +32 (2) 711 14 00 - Fax: +32 (2) 720 96 71


-----Original Message-----
From: Joseph J. Strout [mailto:joe@strout.net]
Sent: Wednesday, October 06, 1999 11:04 PM
To: Jim Harrison; Python-SIG
Subject: Re: [Pythonmac-SIG] IDE class/method popup?


At 4:57 PM -0400 10/06/99, Jim Harrison wrote:

>The popup menu at the bottom left of my IDE script window lists methods
>and classes with popout arrows next to the class names, presumably for
>their contained methods.

Yep.

> However, no menus pop out from these arrows. I
>presume that this is a "feature to be implemented"...

Nope, it's a feature that's been implemented for several versions now 
(as long as I've been using the IDE, I think).

I have noticed that sometimes the method menus don't appear.  This 
happens rarely, and I haven't been able to spot any particular 
pattern to it.  I wonder if it has to do with low-memory conditions? 
Jim, if you increase your memory allocation to PythonIDE, does the 
problem disappear?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'

_______________________________________________
Pythonmac-SIG maillist  -  Pythonmac-SIG@python.org
http://www.python.org/mailman/listinfo/pythonmac-sig


From jhrsn@pop.pitt.edu  Thu Oct  7 13:41:01 1999
From: jhrsn@pop.pitt.edu (Jim Harrison)
Date: Thu, 7 Oct 1999 08:41:01 -0400
Subject: [Pythonmac-SIG] IDE class/method popup?
Message-ID: <199910071238.IAA28025@post-ofc05.srv.cis.pitt.edu>

>I just had the same problem this week, and check
>if you use spaces instead of tabs for indentation.
>If you use spaces, the IDE will not show the
>methods for the classes If you use tabs, it will
>show them. I replaced all spaces in my script files
>with tabs and it worked for me (Python 1.5.2b1).

That did it, thanks (version 1.5.2c1). This will make editing Linux 
python scripts interesting.

Jim

________________________________________________________________________
James H. Harrison, Jr., MD, PhD
Associate Director of Pathology Informatics, Department of Pathology
Faculty Member in Residence, Center for Biomedical Informatics
University of Pittsburgh Health System                     
C920 PUH, 200 Lothrop Street                                  
Pittsburgh, PA 15213

jhrsn@pop.pitt.edu | voice: 412-647-5529 | fax: 412-647-9588

"If you want sense, you'll have to make it yourself!!"-Norton Juster
________________________________________________________________________



From cmann@pobox.com  Thu Oct  7 15:08:29 1999
From: cmann@pobox.com (Chris,Mann)
Date: Thu, 07 Oct 1999 10:08:29 -0400
Subject: [Pythonmac-SIG] Build Application/Applet
Message-ID: <19991007140829.41F5B2F7AE@smtp.pobox.com>

Hi-
    I wrote a little script to parse email digests into a tab delimited
format.  If I save the script as an applet from the IDE everything is fine.
but if i use either the Build Application or Build Applet tools, I get a
error saying there's a syntax error in my script (this happens when the tool
is writing __main__).  any ideas on where to look for the problem?

I'm assuming if I can get Build Application to work, then people I give the
script to won't have to have Python installed, correct?

thanks
-chris


From joe@strout.net  Thu Oct  7 15:44:41 1999
From: joe@strout.net (Joseph J. Strout)
Date: Thu, 7 Oct 1999 07:44:41 -0700
Subject: [Pythonmac-SIG] IDE class/method popup?
In-Reply-To: <199910071238.IAA28025@post-ofc05.srv.cis.pitt.edu>
References: <199910071238.IAA28025@post-ofc05.srv.cis.pitt.edu>
Message-ID: <v04210102b42261c26940@[204.210.37.16]>

At 8:41 AM -0400 10/07/99, Jim Harrison wrote:
> >I just had the same problem this week, and check
> >if you use spaces instead of tabs for indentation.
>
>That did it, thanks (version 1.5.2c1). This will make editing Linux
>python scripts interesting.

You know, this issue has come up before.  It occurs to me that the 
IDE has a "scripts" menu in which you can place any Python script you 
want to be easily accessible.  Couldn't you write "Tabify" and 
"Detabify" scripts, and place them in this menu?  They'd be designed 
to act on the frontmost window (if you need help getting a reference 
this window or its data, I can probably help).  So when you suck your 
space-indented file down, you'd just "Tabify" it, and when you're 
ready to upload to Linux, you'd "Detabify" it (if you really feel the 
need -- personally, I just leave the tabs in).

HTH,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jack@oratrix.nl  Thu Oct  7 16:51:25 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Thu, 07 Oct 1999 17:51:25 +0200
Subject: [Pythonmac-SIG] Build Application/Applet
In-Reply-To: Message by "Chris,Mann" <cmann@pobox.com> ,
 Thu, 07 Oct 1999 10:08:29 -0400 , <19991007140829.41F5B2F7AE@smtp.pobox.com>
Message-ID: <19991007155232.E033E35BB1E@snelboot.oratrix.nl>

> Hi-
>     I wrote a little script to parse email digests into a tab delimited
> format.  If I save the script as an applet from the IDE everything is fine.
> but if i use either the Build Application or Build Applet tools, I get a
> error saying there's a syntax error in my script (this happens when the tool
> is writing __main__).  any ideas on where to look for the problem?

Well, a good place would be the line where it reports the syntax error:-)

Seriously: I assume that you don't get that info, for some reason. Try running 
your script with the interpreter (in stead of with the IDE), maybe it'll fail 
there too (and then, hopefully, with a linenumber).

> I'm assuming if I can get Build Application to work, then people I give the
> script to won't have to have Python installed, correct?

Right.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From cmann@pobox.com  Thu Oct  7 18:07:28 1999
From: cmann@pobox.com (Chris,Mann)
Date: Thu, 07 Oct 1999 13:07:28 -0400
Subject: [Pythonmac-SIG] Build Application/Applet
Message-ID: <19991007170730.A97D22F7B0@smtp.pobox.com>

> Well, a good place would be the line where it reports the syntax error:-)
>
> Seriously: I assume that you don't get that info, for some reason. Try running
> your script with the interpreter (in stead of with the IDE), maybe it'll fail
> there too (and then, hopefully, with a linenumber).

nope.  runs fine in both the IDE and the interpreter.  Even saves fine as an
applet from the IDE.  But won't be converted into an applet or application
by the two tools.  Just get the error message, "Syntax error in script
'Data:Tools:parse_digest.py'"
The .out window has a list of unused modules and then says, 'Adding
"__main__"' when I get this error.

Should I try running the Build_Application.py file in debug?

-chris



From cmann@pobox.com  Thu Oct  7 20:37:25 1999
From: cmann@pobox.com (Chris,Mann)
Date: Thu, 07 Oct 1999 15:37:25 -0400
Subject: [Pythonmac-SIG] Build Application/Applet
Message-ID: <19991007193725.88E142F7F1@smtp.pobox.com>

>> Hi-
>>     I wrote a little script to parse email digests into a tab delimited
>> format.  If I save the script as an applet from the IDE everything is fine.
>> but if i use either the Build Application or Build Applet tools, I get a
>> error saying there's a syntax error in my script (this happens when the tool
>> is writing __main__).  any ideas on where to look for the problem?

I figured out what the problem was.  One of the lines in my script had a tab
after the text but before the linefeed.  Once I removed that tab, everything
built fine.  Should this really affect building the application?

-chris



From draayer@door.net  Fri Oct  8 08:02:54 1999
From: draayer@door.net (Dean E.Draayer)
Date: Fri, 8 Oct 99 02:02:54 -0500
Subject: [Pythonmac-SIG] BBEdit "Run As Python" plug-in?
Message-ID: <19991008071025864.AAA299@mail.door.net@[208.247.125.122]>

>You could use the BBEdit "Run As Python" plug-in with BBEdit or BBEdit
>Lite.  This used to be included with MacPython; I'm not sure if it's still
>there or not.

I couldn't find it in my MacPython installation.  Does anybody know where
to find this plug-in?

Dean Draayer


From jack@oratrix.nl  Fri Oct  8 10:47:27 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 08 Oct 1999 11:47:27 +0200
Subject: [Pythonmac-SIG] BBEdit "Run As Python" plug-in?
In-Reply-To: Message by "Dean E.Draayer" <draayer@door.net> ,
 Fri, 8 Oct 99 02:02:54 -0500 , <19991008071025864.AAA299@mail.door.net@[208.247.125.122]>
Message-ID: <19991008094728.45E4B35BB1E@snelboot.oratrix.nl>

> 
> >You could use the BBEdit "Run As Python" plug-in with BBEdit or BBEdit
> >Lite.  This used to be included with MacPython; I'm not sure if it's still
> >there or not.
> 
> I couldn't find it in my MacPython installation.  Does anybody know where
> to find this plug-in?

Just discontinued it, as the IDE is supposed to do everything it did, and more.

I think the sources can still be found in the Mac:Contrib folder, but I'm not 
sure.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Fri Oct  8 10:50:56 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 08 Oct 1999 11:50:56 +0200
Subject: [Pythonmac-SIG] Build Application/Applet
In-Reply-To: Message by "Chris,Mann" <cmann@pobox.com> ,
 Thu, 07 Oct 1999 15:37:25 -0400 , <19991007193725.88E142F7F1@smtp.pobox.com>
Message-ID: <19991008095148.8385735BB1E@snelboot.oratrix.nl>

> >> Hi-
> >>     I wrote a little script to parse email digests into a tab delimited
> >> format.  If I save the script as an applet from the IDE everything is fine.
> >> but if i use either the Build Application or Build Applet tools, I get a
> >> error saying there's a syntax error in my script (this happens when the tool
> >> is writing __main__).  any ideas on where to look for the problem?
> 
> I figured out what the problem was.  One of the lines in my script had a tab
> after the text but before the linefeed.  Once I removed that tab, everything
> built fine.  Should this really affect building the application?

No, it definitely shouldn't. I've tried to reproduce the problem, but however 
many tabs I put after lines BuildApplet doesn't seem to care, so could you 
send me your sourcefile so I can have a look? Please send it binhexed, not 
straight ascii, as in this case we want to file to come across verbatim.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From orla_redbird@crosswinds.net  Fri Oct  8 11:28:27 1999
From: orla_redbird@crosswinds.net (Gordon Worley)
Date: Fri, 8 Oct 1999 06:28:27 -0400 (EDT)
Subject: [Pythonmac-SIG] Accessing front window in IDE
Message-ID: <l031303001e1d821bb50f@[10.50.227.8]>

Joesph Strout wrote:
>>>
to act on the frontmost window (if you need help getting a reference
this window or its data, I can probably help).  So when you suck your
<<<

Could you please write a short example of how to do this.  I've tried in
the past, but wasn't able to get very far.  I'm sure others have had or are
having a problem with this, so some help would be much appreciated.

- Gordon Worley
http://www.crosswinds.net/~orla_redbird/
mailto:orla_redbird@crosswinds.net




From joe@strout.net  Fri Oct  8 16:30:32 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 8 Oct 1999 08:30:32 -0700
Subject: [Pythonmac-SIG] Accessing front window in IDE
In-Reply-To: <l031303001e1d821bb50f@[10.50.227.8]>
References: <l031303001e1d821bb50f@[10.50.227.8]>
Message-ID: <v04210107b423bd87951b@[198.202.70.213]>

At 6:28 AM -0400 10/08/99, Gordon Worley wrote:

>Could you please write a short example of how to [get the frontmost window].

OK, here's a script that, when run, adds "# Spam!" to the end of the 
frontmost window if that's an editor window.  You'll note a couple of 
simpler demos, which simply inspect the window or the editor object, 
commented out.  Place this in your IDE Scripts folder, and get the 
IDE to rescan this folder with the "Python > Preferences > Set 
Scripts Folder" menu command.

----------------------------------------------------------------------
"""
This script illustrates how to access the contents of the frontmost window
in the Python IDE.  Handy for making little utilities for the "Scripts"
menu, like text reformatters and such.
"""

import W
import Win
import PyEdit
import Res
import PyAdvancedEditor

# get a reference to the Python object corresponding to the frontmost window
wid = Win.FrontWindow()
if wid and W._application._windows.has_key(wid):
	window = W._application._windows[wid]
else:
	window = None

# now, see if that's an editor window...
if window and isinstance(window, PyEdit.Editor):
	# yep, we have an editor window... do something with (or to) it!

	# demo 1: just inspect the window
	#inspect(window)

	# demo 2: inspect the actual editor
	#editor = window.editgroup.editor
	#inspect(editor)

	# demo 3: grab the text, and operate on it
	editor = window.editgroup.editor
	selStart, selEnd = editor.ted.WEGetSelection()
	text = editor.get()

	text = text + "\r# Spam!"

	editor.set(text)
	editor.ted.WESetSelection(selStart, selEnd)
	if isinstance(editor, PyAdvancedEditor.PyAdvancedEditor):
		editor.markDirty(0, len(text))
----------------------------------------------------------------------
In the place where I've done "text = text + "\r# Spam!"", you could 
instead do any manipulation on the text you want -- it's the full 
text in the editor, with lines separated by '\r' characters.  You 
could tabify, detabify, etc.  Or you could leave the text alone, and 
do something with the window -- e.g., if you're good with 
AppleEvents, you could write us a "Reveal in Finder" script that 
switches to the Finder and selects the file corresponding to this 
editor window.

If you do any of these, please post them here or on a web page so the 
rest of us can enjoy them too!

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jack@oratrix.nl  Fri Oct  8 16:53:13 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 08 Oct 1999 17:53:13 +0200
Subject: [Pythonmac-SIG] Accessing front window in IDE
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Fri, 8 Oct 1999 08:30:32 -0700 , <v04210107b423bd87951b@[198.202.70.213]>
Message-ID: <19991008155313.605A635BB1E@snelboot.oratrix.nl>

> At 6:28 AM -0400 10/08/99, Gordon Worley wrote:
> 
> >Could you please write a short example of how to [get the frontmost window].
> 
> OK, here's a script that, when run, adds "# Spam!" to the end of the 
> frontmost window if that's an editor window.

Hmm, it would be nice to have a bit of an API for this that IDE extension 
scripts could use. If there was an easier interface than mucking around with 
the Window, PyEdit and waste internals that might prompt more people to write 
nifty little utilities.

--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From joe@strout.net  Fri Oct  8 17:01:41 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 8 Oct 1999 09:01:41 -0700
Subject: [Pythonmac-SIG] Accessing front window in IDE
In-Reply-To: <19991008155313.605A635BB1E@snelboot.oratrix.nl>
References: <19991008155313.605A635BB1E@snelboot.oratrix.nl>
Message-ID: <v04210101b423c466e08c@[198.202.70.213]>

At 5:53 PM +0200 10/08/99, Jack Jansen wrote:

>Hmm, it would be nice to have a bit of an API for this that IDE extension
>scripts could use. If there was an easier interface than mucking around with
>the Window, PyEdit and waste internals that might prompt more people to write
>nifty little utilities.

That's a good point.  Seems like most of the time, you'd want to 
either change the text, or change the selection (e.g., if doing some 
sort of automatic search thing), or both.  That's pretty easy.

But then, what if somebody wants to write the "Reveal In Finder" 
script?  That requires access to the window path; text and selection 
range aren't enough.  So we add access to the path too; but what if 
there are other things we haven't thought of?

Still, I guess we could do better than the current state of things. 
Maybe something like this:

--------------------- hypothetical test code begins ---------------------
from FrontWinAPI import FrontWinAPI

# instantiate a front-window API object... automatically grabs text,
# selection range, path, etc.
win = FrontWinAPI()

# insert "Spam!" to the top, and also move the selection range accordingly
spam = "# Spam!\r"
win.text = spam + win.text
win.selStart = win.selStart + len(spam)
win.selEnd = win.selEnd + len(spam)

# that's it -- when win object dies, the front window is updated automagically
--------------------- hypothetical test code ends ---------------------

Is that the sort of thing you had in mind?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From cbarker@jps.net  Fri Oct  8 19:33:59 1999
From: cbarker@jps.net (Chris Barker)
Date: Fri, 08 Oct 1999 11:33:59 -0700
Subject: [Pythonmac-SIG] tab vs spaces: Linux and mac
References: <199910080502.BAA26631@python.org>
Message-ID: <37FE3917.87B945C7@jps.net>

This is a multi-part message in MIME format.
--------------CF330202E61B206B95B49730
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi,

I confronted this probelm, as I develop python on both Linux and the
Mac. My solution was to just stick with using ONLY tabs in all my python
code, on both platforms. I did this because using spaces on the mac is a
real pain, as most editors won't do anything automatically for you.
Frankly, I think is a major shame that Python can use a mixture of
spaces and tabs, as that can be a real nightmare, and doesn't fit at all
with the "what you see is what you get" philosphy of making indentation
part of the syntax , as what you see is only what you get if a tab
equals 8 spaces.

Anyway,

Emacs python mode can be set to use only tabs, and I got a handy script
from someone that will convert Python code to use any legal combination
of spaces and tabs. I've enclosed it here.

-have fun,
-Chris



-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------
--------------CF330202E61B206B95B49730
Content-Type: text/plain; charset=us-ascii;
 name="tabcleaner.py"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="tabcleaner.py"

#!/usr/bin/python

import tokenize
import string

TABSONLY = 'TABSONLY'
SPACESONLY = 'SPACESONLY'
MIXED = 'MIXED'

class PyText:
    def __init__(self, fnm, optdict):
        self.optdict = optdict
        self.fnm = fnm
    	self.txt = open(self.fnm, 'r').readlines()
    	self.indents = [(0, 0, )]
    	self.lnndx = 0
    	self.indentndx = 0
    def getline(self):
        if self.lnndx < len(self.txt):
            txt = self.txt[self.lnndx]
            self.lnndx = self.lnndx + 1
        else:
            txt = ''
        return txt
    def tokeneater(self, type, token, start, end, line):
        if type == tokenize.INDENT:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, start[0]-1)
            self.indents.append((lvl+1, start[0]-1,))
        elif type == tokenize.DEDENT:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, start[0]-1)
            self.indents.append((lvl-1, start[0]-1,))
        elif type == tokenize.ENDMARKER:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, len(self.txt))
    def split(self, ln):
        content = string.lstrip(ln)
        if not content:
            return ('', '\n')
        lead = ln[:len(ln) - len(content)]
        lead = string.expandtabs(lead)
        return (lead, content)
        
    def process(self):
        style = self.optdict.get('style', SPACESONLY)
        indent = string.atoi(self.optdict.get('indent', '4'))
        tabsz = string.atoi(self.optdict.get('tabs', '8'))
        print 'file %s -> style %s, tabsize %d, indent %d' % (self.fnm, style, tabsz, indent)
        tokenize.tokenize(self.getline, self.tokeneater)
        #import pprint
        #pprint.pprint(self.indents)
        new = []
        for (lvl, s, e) in self.indents:
            if s >= len(self.txt):
                break
            if s == e:
                continue
            oldlead, content = self.split(self.txt[s])
            #print "oldlead", len(oldlead), `oldlead`
            if style == TABSONLY:
                newlead = '\t'*lvl
            elif style == SPACESONLY:
                newlead = ' '*(indent*lvl)
            else:
                sz = indent*lvl
                t,spcs = divmod(sz, tabsz)
                newlead = '\t'*t + ' '*spcs
            new.append(newlead + content)
            for ln in self.txt[s+1:e]:
                lead, content = self.split(ln)
                #print "lead:", len(lead)
                new.append(newlead + lead[len(oldlead):] + content)
        self.save(new)
        #print "---", self.fnm
        #for ln in new:
        #    print ln,
        #print
                
    def save(self, txt):
        bakname = os.path.splitext(self.fnm)[0]+'.bak'
        print "backing up", self.fnm, "to", bakname
        #print os.getcwd()
        try:
            os.rename(self.fnm, bakname)
        except os.error:
            os.remove(bakname)
            os.rename(self.fnm, bakname)
        open(self.fnm, 'w').writelines(txt)

def test():
    tc = PyText('test1.py')
    tc.process()
    tc = PyText('test1.py')
    tc.process(style=TABSONLY)
    tc = PyText('test1.py')
    tc.process(style=MIXED, indent=4, tabs=8)
    tc = PyText('test1.py')
    tc.process(style=MIXED, indent=2, tabs=8)
 
def cleanfile(fnm, d):
    if os.path.isdir(fnm) and not os.path.islink(fnm):
        names = os.listdir(fnm)
        for name in names:
            fullnm = os.path.join(fnm, name)
            if (os.path.isdir(fullnm) and not os.path.islink(fullnm)) or \
            	os.path.normcase(fullnm[-3:]) == ".py":
            	cleanfile(fullnm, d)
        return
    tc = PyText(fnm, d)
    tc.process()
  
usage="""\
%s [options] [path...]
 options
  -T : reformat to TABS ONLY
  -S : reformat to SPACES ONLY ( -i option is important)
  -M : reformat to MIXED SPACES / TABS ( -t and -i options important)
  -t<n> : tab is worth <n> characters
  -i<n> : indents should be <n> characters 
  -h : print this text
 path is file or directory
"""
if __name__ == '__main__':
    import sys, getopt, os
    opts, args = getopt.getopt(sys.argv[1:], "TSMht:i:")
    d = {}
    #print `opts`
    for opt in opts:
        if opt[0] == '-T':
            d['style'] = TABSONLY
        elif opt[0] == '-S':
            d['style'] = SPACESONLY
        elif opt[0] == '-M':
            d['style'] = MIXED
        elif opt[0] == '-t':
            d['tabs'] = opt[1]
        elif opt[0] == '-i':
            d['indent'] = opt[1]
        elif opt[0] == '-h':
            print usage % sys.argv[0]
            sys.exit(0)
    if not args:
        print usage % sys.argv[0]
    for arg in args:
        cleanfile(arg, d)
    
    

--------------CF330202E61B206B95B49730--



From steele@cs.brandeis.edu  Fri Oct  8 21:06:03 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Fri, 8 Oct 1999 16:06:03 -0400
Subject: [Pythonmac-SIG] RE: IDE class/method popup?
References: <199910080502.BAA26631@python.org>
Message-ID: <008301bf11c8$9cab59f0$d22d70d8@lexeme.com>

Pieter Claerhout:
>I just had the same problem this week, and check
>if you use spaces instead of tabs for indentation.

Jim Harrison:
>That did it, thanks (version 1.5.2c1). This will make editing Linux
>python scripts interesting.

Joe Strout:
> You know, this issue has come up before.  It occurs to me that the
> IDE has a "scripts" menu in which you can place any Python script you
> want to be easily accessible.  Couldn't you write "Tabify" and
> "Detabify" scripts, and place them in this menu?  They'd be designed
> to act on the frontmost window (if you need help getting a reference
> this window or its data, I can probably help).  So when you suck your
> space-indented file down, you'd just "Tabify" it, and when you're
> ready to upload to Linux, you'd "Detabify" it (if you really feel the
> need -- personally, I just leave the tabs in).

<http://www.cs.brandeis.edu/~steele/sources/python.html> has a IDE patch
called PyIDESpacesToTabsPatch.py:  "make the IDE change spaces to tabs when
it opens a file, and back to spaces when it writes it out".  There are
installation instructions at the top of the file.

I've been using this with earlier versions of MacPython 1.5.2bx, to view and
edit files from linux and the rest of the world.  I haven't yet tested it
with 1.5.2c1, or updated it to use Joe Strout's patch excellent mechanism.
It assumes 4 spaces per tab stop.



From steele@cs.brandeis.edu  Sat Oct  9 12:09:40 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Sat, 09 Oct 1999 07:09:40 -0400
Subject: [Pythonmac-SIG] line endings
Message-ID: <199910091109.HAA00141@life.ai.mit.edu>

Joseph J. Strout writes:
>>OK, here's a patch (at http://www.strout.net/python/mac/) which
>>extends the IDE's editor in a couple of ways:

Jim Harrison writes:
> For newbies like me--remember to change the line endings to Mac after the
> files are downloaded, at least when downloading by dragging out of the
> Navigator window. BBedit or BBedit Lite works well for this. Bad things
> happen if you run these patches with unix line endings.

The 1.5.2c1 PythonIDE can also do this.  (It incorporates the
PythonIDE-crlf-patch.txt patch I sent mail about a few months ago.)  Open a
UNIX or DOS-style file with the IDE, and it will ask whether to convert the
line separators.  Confirm, and then save the file.

<http://developer.apple.com/technotes/tn/tn1176.html> says that the StdCLib
shared library has been changed for MacOS 9, such that "fgets now recognizes
line endings for Mac, DOS, and UNIX files."  I wonder if MacPython will
inherit this behavior?  (I don't know have access to MacOS 9, and I haven't
read the Python sources to see where file.readline() is coming from, and
whether the compiler uses this.)  If it does, then UNIX/DOS-originated
source files should compile under MacOS 9 (and anything else that uses
readline should start working on foreign text files too).  The editor will
still need special casing to treat foreign line separator sequences as line
separators, but if readline() and the compiler treat them the same, then the
editor should be changed to do the conversion silently a la BBEdit, Alpha,
and the Metrowerks IDE.


From hung_jung_lu@adventure.com  Sat Oct  9 17:21:56 1999
From: hung_jung_lu@adventure.com (Hung Jung Lu)
Date: Sat, 9 Oct 1999 09:21:56 -0700
Subject: [Pythonmac-SIG] Folder alias, time offset
Message-ID: <775C7FB8CE24D31195930008C745F54301CDE3@hermes.adventure.com>

Hi,

I am new here. Sorry if this has been asked before.

I have two questions regarding Mac Python (1.5.1).

(1) How can one make a folder alias? I know that
          macostools.mkalias(src, dst)
     makes a file alias, but it does not work for folders.

(2) I consistently see that there is a mismatch between MacOS
     filesystem time and the operating system time. To be precise,
     they mismatch by around 4 years. Does that have something 
     to do with MacIntosh's definition of epoch?
     (I believe that it is 1904/1/1, right?) Is this a know bug, or is
     there a reason behind it all? If this is an intended feature, then
     is there a standard way of converting time back and forth?
     (I typically get a filetime by using t=os.stat(filename)[8], it can
     then be views in formatted ASCII time by using 
          print time.ctime(t),
     and I see that the printed time is 4 years younger.)

regards,

Hung Jung



From richard@richardgordon.net  Sat Oct  9 19:50:59 1999
From: richard@richardgordon.net (Richard Gordon)
Date: Sat, 9 Oct 1999 14:50:59 -0400
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
Message-ID: <v04205502b4253c21df8d@[209.214.74.89]>

A friend of mine asked me to do an AppleScript droplet for him that 
would change the creator code of his gifs & jpegs from that of 
FireWorks or PhotoShop to the code for Picture Viewer. I know how to 
do that in AppleScript and probably know how to do it in MacPerl, but 
I decided to use Python instead because it's cool, etc.

Anyway, naturally I'm now stuck. I gathered that what I wanted to use 
was macfs.FInfo, but I have yet to figure it out. The Python 
documentation deals with this in Sec. 7.3 and says that the FInfo 
Objects include:

Creator
        The 4-character creator code of the file.

That's all well and good, but when I try to use this, I just get

>>> macfs.FInfo(thefile)
<FInfo object at 300e0c0>

so I guess that I am missing the point here.

I'd appreciate any explanations about this and, while we're at it, 
what's the deal with FInfoType?

Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From bsayer@home.msen.com  Sun Oct 10 02:45:29 1999
From: bsayer@home.msen.com (Ben Sayer)
Date: Sat, 9 Oct 1999 21:45:29 -0400
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
Message-ID: <19991009214529.026125@home.msen.com>

On Sat, Oct 9, 1999, Richard Gordon <richard@richardgordon.net> wrote:

>Anyway, naturally I'm now stuck. I gathered that what I wanted to use 
>was macfs.FInfo, but I have yet to figure it out. The Python 
>documentation deals with this in Sec. 7.3 and says that the FInfo 
>Objects include:
>
>Creator
>        The 4-character creator code of the file.
>
>That's all well and good, but when I try to use this, I just get
>
>>>> macfs.FInfo(thefile)
><FInfo object at 300e0c0>

You need to use the macfs module, create an FSSpec object for the file,
then use the SetCreatorType method of the file's FSSpec object. For example:

import macfs

offs = macfs.FSSpec('sample.jpg')
offs.SetCreatorType('JVWR', 'JPEG')

>I'd appreciate any explanations about this and, while we're at it, 
>what's the deal with FInfoType?

My documentation states that the Finfo method creates a zero-filled FInfo
object. This is clearly not what you want. I'm not sure what one would
use the FInfo object type for, but I'm new to Mac programming too.

--Ben



From maccgi@bellsouth.net  Sun Oct 10 04:23:30 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Sat, 9 Oct 1999 23:23:30 -0400
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: <19991009214529.026125@home.msen.com>
References: <19991009214529.026125@home.msen.com>
Message-ID: <v04205500b425b696a402@[209.214.65.76]>

At 21:45 -0400 10/09/1999, Ben Sayer wrote:
>You need to use the macfs module, create an FSSpec object for the file,
>then use the SetCreatorType method of the file's FSSpec object. For example:

Thanks. Somebody previously sent me some offlist help along the same 
lines and I also ran across FixCreator.py in the distribution which 
was readily adaptable to what I wanted to do.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From jack@oratrix.nl  Mon Oct 11 10:33:11 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 11 Oct 1999 11:33:11 +0200
Subject: [Pythonmac-SIG] Accessing front window in IDE
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Fri, 8 Oct 1999 09:01:41 -0700 , <v04210101b423c466e08c@[198.202.70.213]>
Message-ID: <19991011093312.1527A35BB1E@snelboot.oratrix.nl>

> >Hmm, it would be nice to have a bit of an API for this that IDE extension
> >scripts could use. If there was an easier interface than mucking around with
> >the Window, PyEdit and waste internals that might prompt more people to write
> >nifty little utilities.
> 
> That's a good point.  Seems like most of the time, you'd want to 
> either change the text, or change the selection (e.g., if doing some 
> sort of automatic search thing), or both.  That's pretty easy.
> 
> But then, what if somebody wants to write the "Reveal In Finder" 
> script?  That requires access to the window path; text and selection 
> range aren't enough.  So we add access to the path too; but what if 
> there are other things we haven't thought of?

More-or-less. I'm thinking along the lines of how other applications export 
functionality through AppleEvents. We would do it through Python objects, of 
course, but with a similar functionality. The object hierarchy would be 
something like
IDE_application
    frontwindow() -> IDE_window
    getwindowbyname(name) -> IDE_window
    openwindow(filename) -> IDE_window
    ...
IDE_window
    getselection() -> IDE_text
    gettextrange(from, to) ->IDE_text
    setselection(from, to)
    ...
IDE_text
    get() -> string
    set(string)
    setcolor(r, g, b)
    ...
	
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct 11 10:44:58 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 11 Oct 1999 11:44:58 +0200
Subject: [Pythonmac-SIG] line endings
In-Reply-To: Message by "Oliver Steele" <steele@cs.brandeis.edu> ,
 Sat, 09 Oct 1999 07:09:40 -0400 , <199910091109.HAA00141@life.ai.mit.edu>
Message-ID: <19991011094458.E77CF35BB1E@snelboot.oratrix.nl>

> <http://developer.apple.com/technotes/tn/tn1176.html> says that the StdCLib
> shared library has been changed for MacOS 9, such that "fgets now recognizes
> line endings for Mac, DOS, and UNIX files."  I wonder if MacPython will
> inherit this behavior?

Not directly, as MacPython uses the Metrowerks C library. I wouldn't be 
surprised if MW will nclude this behaviour as well, though, when StdCLib does.

--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct 11 10:50:16 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 11 Oct 1999 11:50:16 +0200
Subject: [Pythonmac-SIG] Folder alias, time offset
In-Reply-To: Message by Hung Jung Lu <hung_jung_lu@adventure.com> ,
 Sat, 9 Oct 1999 09:21:56 -0700 , <775C7FB8CE24D31195930008C745F54301CDE3@hermes.adventure.com>
Message-ID: <19991011095016.8FD9935BB1E@snelboot.oratrix.nl>

> (1) How can one make a folder alias? I know that
>           macostools.mkalias(src, dst)
>      makes a file alias, but it does not work for folders.

I'm surprised, I didn't know it didn't work for folders. If anyone has a patch 
please forward it to me too for inclusion in the source tree.

> (2) I consistently see that there is a mismatch between MacOS
>      filesystem time and the operating system time. To be precise,
>      they mismatch by around 4 years. Does that have something 
>      to do with MacIntosh's definition of epoch?
>      (I believe that it is 1904/1/1, right?) Is this a know bug, or is
>      there a reason behind it all? If this is an intended feature, then
>      is there a standard way of converting time back and forth?
>      (I typically get a filetime by using t=os.stat(filename)[8], it can
>      then be views in formatted ASCII time by using 
>           print time.ctime(t),
>      and I see that the printed time is 4 years younger.)

Aw, grmpf. Metrowerks keeps changing their idea of how to represent date/time, 
and MacPython gets hit by it time and again. There have been three epochs in 
use in the software MacPython uses:
- The Mac epoch (1/1/1904, visible with GetFInfo() and such calls)
- The Metrowerks C library epoch (1/1/1900)
- The CWGUSI epoch.

Apparently I've messed up in the conversion again (or maybe I should simply 
take out my workaround for MW's messup in a previous release). I'll have a 
look.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct 11 10:54:49 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 11 Oct 1999 11:54:49 +0200
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: Message by Ben Sayer <bsayer@home.msen.com> ,
 Sat, 9 Oct 1999 21:45:29 -0400 , <19991009214529.026125@home.msen.com>
Message-ID: <19991011095449.9022D35BB1E@snelboot.oratrix.nl>

> On Sat, Oct 9, 1999, Richard Gordon <richard@richardgordon.net> wrote:
> 
> >Anyway, naturally I'm now stuck. I gathered that what I wanted to use 
> >was macfs.FInfo, but I have yet to figure it out. The Python 
> >documentation deals with this in Sec. 7.3 and says that the FInfo 
> >Objects include:
> >
> >Creator
> >        The 4-character creator code of the file.
> >
> >That's all well and good, but when I try to use this, I just get
> >
> >>>> macfs.FInfo(thefile)
> ><FInfo object at 300e0c0>
> 
> You need to use the macfs module, create an FSSpec object for the file,
> then use the SetCreatorType method of the file's FSSpec object. For example:
> 
> import macfs
> 
> offs = macfs.FSSpec('sample.jpg')
> offs.SetCreatorType('JVWR', 'JPEG')

That works, but that doesn't use a FInfo object. It is the old method of 
setting creator/type, and while I would love to depracate it in favor of using 
FInfo it is used too often by too many people (including me:-)

> >I'd appreciate any explanations about this and, while we're at it, 
> >what's the deal with FInfoType?
> 
> My documentation states that the Finfo method creates a zero-filled FInfo
> object. This is clearly not what you want. I'm not sure what one would
> use the FInfo object type for, but I'm new to Mac programming too.

A FInfo object holds all Finder Information of a file: not only creator and 
type but also icon position, flags, etc. The calls are
  fsspec.GetFInfo() -> FInfo object for the file
  fsspec.SetFInfo(finfo) Sets finder info
Since occasionaly you need an empty finder info object there's also the 
FInfo() call to create an empty one.

All the finder information is in attributes of the finfo object.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From maccgi@bellsouth.net  Mon Oct 11 15:23:36 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Mon, 11 Oct 1999 10:23:36 -0400
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: <19991011095449.9022D35BB1E@snelboot.oratrix.nl>
References: <19991011095449.9022D35BB1E@snelboot.oratrix.nl>
Message-ID: <v04205507b427933889c4@[209.214.75.104]>

--============_-1272470664==_ma============
Content-Type: text/plain; charset="us-ascii" ; format="flowed"

At 11:54 +0200 10/11/1999, Jack Jansen wrote:
>That works, but that doesn't use a FInfo object. It is the old method of
>setting creator/type, and while I would love to depracate it in favor of using
>FInfo it is used too often by too many people (including me:-)

Well, I got past this stuff by butchering FixCreator.py from the 
distribution, but now I've got a slight mystery that I think has to 
do with recursive functions. When a search folder is selected, any 
subfolders within it are processed as well and I suppose that this 
would go on as deeply as they may be nested. That's fine, but I'm 
having trouble keeping a cumulative total of the files that have been 
analyzed to display in a dialog at the end of everything, e.g.:

# import various things

# set some variables for file types- we're converting gifs, etc, to 
PictureViewer ("ogle")

def walktree(name, change):
	if os.path.isfile(name):
		fs = macfs.FSSpec(name)
		cur_cr, cur_tp = fs.GetCreatorType()
		if (cur_cr != NEW) and (cur_tp == GIF or JPEG or TIFF):
			fs.SetCreatorType(NEW, cur_tp)
			macostools.touched(fs)
	elif os.path.isdir(name):
		files = os.listdir(name)
		global lcounter
		lcounter = 0
		for f in files:
			walktree(os.path.join(name, f), change)
			lcounter = lcounter + 1

def run(change):
	fss, ok = macfs.GetDirectory('Folder to search:')
	if not ok:
		sys.exit(0)
	start = time.time()
	walktree(fss.as_pathname(), change)
	stop = time.time()
	runtime = stop - start
	if runtime <= 1:
		theString = 'Finished analyzing %d files in 1 second. 
Pretty cool.' % runtime
	else:
		theString = 'Finished analyzing %d files in %d 
seconds. Pretty cool.' % (lcounter, runtime) # forgive the wrapping 
in Eudora
	EasyDialogs.Message(theString)


if __name__ == '__main__':
	run(1)

# EOF

Because walktree keeps calling itself every time a new folder is 
encountered, the lcounter variable keeps being reset to 0, so the 
final dialog just says how many files happen to have been in the last 
folder processed. I've tried to park the current count in another 
variable, but if it's at the top level, it isn't seen for some reason 
and if it's set up as a global within walktree and initialized to 0, 
it keeps going back to 0 even tho I try to accumulate in it with 
something like totcounter = totcounter + lcounter.

So, my question is where do you hide a cumulative counter when a 
function is recursive? Or do I need to take this whole thing apart 
and split walktree into 2 functions so it's no longer recursive in 
nature? Thanks.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)
--============_-1272470664==_ma============
Content-Type: text/enriched; charset="us-ascii"

At 11:54 +0200 10/11/1999, Jack Jansen wrote:

<excerpt>That works, but that doesn't use a FInfo object. It is the old
method of 

setting creator/type, and while I would love to depracate it in favor
of using 

FInfo it is used too often by too many people (including me:-)

</excerpt>

Well, I got past this stuff by butchering FixCreator.py from the
distribution, but now I've got a slight mystery that I think has to do
with recursive functions. When a search folder is selected, any
subfolders within it are processed as well and I suppose that this
would go on as deeply as they may be nested. That's fine, but I'm
having trouble keeping a cumulative total of the files that have been
analyzed to display in a dialog at the end of everything, e.g.:


<bold><fixed><smaller># import various things


</smaller></fixed></bold><fixed><smaller># set some variables for file
types- we're converting gifs, etc, to PictureViewer ("ogle")


<bold>def
</bold><color><param>BFFF,0000,0000</param>walktree</color>(name,
change):

	<bold>if</bold> os.path.isfile(name):

		fs = macfs.FSSpec(name)

		cur_cr, cur_tp = fs.GetCreatorType()

		<bold>if</bold> (cur_cr != NEW) <bold>and</bold> (cur_tp == GIF
<bold>or </bold>JPEG <bold>or</bold> TIFF):

			fs.SetCreatorType(NEW, cur_tp)

			macostools.touched(fs)

	<bold>elif</bold> os.path.isdir(name):

		files = os.listdir(name)

		<bold>global </bold>lcounter

		lcounter = 0

		for f <bold>in</bold> files:

			walktree(os.path.join(name, f), change)

			lcounter = lcounter + 1


<bold>def</bold><color><param>BFFF,0000,0000</param>
run</color>(change):

	fss, ok =
macfs.GetDirectory(<color><param>0000,7FFF,0000</param>'Folder to
search:'</color>)

	<bold>if not</bold> ok:

		sys.exit(0)

	start = time.time()

	walktree(fss.as_pathname(), change)

	stop = time.time()

	runtime = stop - start

	<bold>if</bold> runtime <<= 1:

		theString = <color><param>0000,7FFF,0000</param>'Finished analyzing
%d files in 1 second. Pretty cool.'</color> % runtime

	<bold>else</bold>:

		theString = <color><param>0000,7FFF,0000</param>'Finished analyzing
%d files in %d seconds. Pretty cool.'</color> % (lcounter, runtime) #
forgive the wrapping in Eudora

	EasyDialogs.Message(theString)


	

<bold>if</bold> __name__ ==
<color><param>0000,7FFF,0000</param>'__main__'</color>:

	run(1)


# EOF


</smaller></fixed>Because walktree keeps calling itself every time a
new folder is encountered, the lcounter variable keeps being reset to
0, so the final dialog just says how many files happen to have been in
the last folder processed. I've tried to park the current count in
another variable, but if it's at the top level, it isn't seen for some
reason and if it's set up as a global within walktree and initialized
to 0, it keeps going back to 0 even tho I try to accumulate in it with
something like totcounter = totcounter + lcounter.


So, my question is where do you hide a cumulative counter when a
function is recursive? Or do I need to take this whole thing apart and
split walktree into 2 functions so it's no longer recursive in nature?
Thanks.



Richard Gordon

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

Gordon Consulting & Design

Database Design/Scripting Languages

mailto:richard@richardgordon.net

http://www.richardgordon.net

770.971.6887 (voice)

770.216.1829 (fax)

--============_-1272470664==_ma============--


From jack@oratrix.nl  Mon Oct 11 15:46:48 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 11 Oct 1999 16:46:48 +0200
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: Message by Richard Gordon <maccgi@bellsouth.net> ,
 Mon, 11 Oct 1999 10:23:36 -0400 , <v04205507b427933889c4@[209.214.75.104]>
Message-ID: <19991011144649.45B1835BB1E@snelboot.oratrix.nl>

That's fairly easy to fix: move the initialization of lcounter to outside the 
function and move your global statement to the top of the function (I'm not 
sure whether the latter is really needed, but everyone puts global statements 
at the top of their functions).

And, while you're at it, in your first print statement replace "runtime" by 
"lcounter".

In general, by the way, once you need global variables you're on the road to 
making your program more complex and it's better to use objects and classes. 
This program would become something like:

class Walker:
	def __init__(self):
		self.lcounter = 0

	def run(self, change):
		self.change = change
		....

		self.walktree(fss.as_pathname)
		....

	def walktree(self, name):
		... # Replace change with self.change, walktree with
		    # self.walktree, lcounter with self.lcounter

if __name__ == '__main__':
	Walker().run(1)
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From maccgi@bellsouth.net  Mon Oct 11 16:06:11 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Mon, 11 Oct 1999 11:06:11 -0400
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: <19991011144649.45B1835BB1E@snelboot.oratrix.nl>
References: <19991011144649.45B1835BB1E@snelboot.oratrix.nl>
Message-ID: <v04205509b427abc84f28@[209.214.75.104]>

At 16:46 +0200 10/11/1999, Jack Jansen wrote:
>That's fairly easy to fix: move the initialization of lcounter to outside the
>function and move your global statement to the top of the function (I'm not
>sure whether the latter is really needed, but everyone puts global statements
>at the top of their functions).

Thanks. I'll play with that and see what happens.

>And, while you're at it, in your first print statement replace "runtime" by
>"lcounter".

Yeah, I saw that after I posted. The actual script is okay and I 
think that this got screwed up as I was editing the message.

>In general, by the way, once you need global variables you're on the road to
>making your program more complex and it's better to use objects and classes.
>This program would become something like:

Hmm, this will take a while to absorb, but thanks a lot.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From joe@strout.net  Mon Oct 11 18:52:28 1999
From: joe@strout.net (Joseph J. Strout)
Date: Mon, 11 Oct 1999 10:52:28 -0700
Subject: [Pythonmac-SIG] Accessing front window in IDE
In-Reply-To: <19991011093312.1527A35BB1E@snelboot.oratrix.nl>
References: <19991011093312.1527A35BB1E@snelboot.oratrix.nl>
Message-ID: <v04210101b427d40c89c4@[198.202.70.213]>

At 11:33 AM +0200 10/11/99, Jack Jansen wrote:

>More-or-less. I'm thinking along the lines of...
>IDE_application
>    frontwindow() -> IDE_window
>    getwindowbyname(name) -> IDE_window
>    openwindow(filename) -> IDE_window
>    ...
>IDE_window
>    getselection() -> IDE_text
>    gettextrange(from, to) ->IDE_text
>    setselection(from, to)
>    ...
>IDE_text
>    get() -> string
>    set(string)
>    setcolor(r, g, b)
>    ...

But then, how is this any different from mucking about with IDE 
widgets, which is what I thought we wanted to avoid?  The IDE objects 
already have methods like those above; apart from the names, I don't 
see much difference.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From joe@strout.net  Mon Oct 11 19:10:49 1999
From: joe@strout.net (Joseph J. Strout)
Date: Mon, 11 Oct 1999 11:10:49 -0700
Subject: [Pythonmac-SIG] How do you use macfs.FInfo?
In-Reply-To: <v04205507b427933889c4@[209.214.75.104]>
References: <19991011095449.9022D35BB1E@snelboot.oratrix.nl>
 <v04205507b427933889c4@[209.214.75.104]>
Message-ID: <v04210104b427d80077a5@[198.202.70.213]>

At 10:23 AM -0400 10/11/99, Richard Gordon wrote:

>So, my question is where do you hide a cumulative counter when a 
>function is recursive?

You return it.  E.g.:

	def process(spam):
		"process spam and all sub-spams, and return a total"
		total = 0
		... do some processing that increments localTotal...
		... and now, for sub-spams:
		for subspam in subspams(spam):
			total = total + process(subspam)

		# all done!  return the grand total
		return total

See how the total is a sum of any local increments, plus all the 
totals of the sub-spams?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jpaint@sculpin.com  Tue Oct 12 00:17:19 1999
From: jpaint@sculpin.com (Jay Painter)
Date: Mon, 11 Oct 1999 16:17:19 -0700 (PDT)
Subject: [Pythonmac-SIG] New OSAm + StuffIt Python Plugins
Message-ID: <Pine.LNX.3.96.991011161115.2488A-100000@server.sculpin.com>

I've uploaded a very revamped version of the OSAm module which fixes
various memory problems, and other nastyness.  This module allows you to
execute AppleScript from within Python (it also allows you to compile and
save scripts, and run pre-compiled scripts).

There's also a new module I wrote, StuffIt, which uses the StuffIt SDK to
expand/stuff .sit files as well as HQX encode and MacBin files.
Unfortunatly, the licencing on the SDK is such that I cannot distribute a
binary.  If you want to use this, you'll have to download the StuffIt SDK
and compile the my code aginst their headers and stub library.  I had to 
write this module because StuffIt's AppleScript interface is very broken,
and will crash the machine if you access it to many times.

Both these modules can be found at:
ftp://sculpin.com/python

Regards,
Jay Painter



From hung_jung_lu@adventure.com  Tue Oct 12 00:45:25 1999
From: hung_jung_lu@adventure.com (Hung Jung Lu)
Date: Mon, 11 Oct 1999 16:45:25 -0700
Subject: [Pythonmac-SIG] Folder alias, time offset
Message-ID: <775C7FB8CE24D31195930008C745F54301CDE7@hermes.adventure.com>

Hmm... it does not seem trivial to create a folder alias.
There is something special about the Alias Info, and
Apple really does not want people to play with aliases
from anything other than the Finder itself.

(I have played with changing alias creator to
'MACS' and type to 'fdrp' and setting the alias bit in
file info, and do everything else just like a file alias,
but it does not work.)

Hung Jung

		-----Original Message-----
		From:	Jack Jansen [mailto:jack@oratrix.nl]
		Sent:	Monday, October 11, 1999 2:50 AM
		To:	Hung Jung Lu
		Cc:	'pythonmac-sig@python.org'
		Subject:	Re: [Pythonmac-SIG] Folder alias, time
offset 


		> (1) How can one make a folder alias? I know that
		>           macostools.mkalias(src, dst)
		>      makes a file alias, but it does not work for folders.

		I'm surprised, I didn't know it didn't work for folders. If
anyone has a patch 
		please forward it to me too for inclusion in the source
tree.

		


From savageb@pacbell.net  Tue Oct 12 01:54:28 1999
From: savageb@pacbell.net (savageb)
Date: Mon, 11 Oct 1999 17:54:28 -0700
Subject: [Pythonmac-SIG] Python 1.5.2c says mxDateTime "too new"?
Message-ID: <0FJG00647T6VXS@mta1.snfc21.pbi.net>

Hello,

I am having a hard time getting mxDateTime to run with Python 1.5.2c1. Joe,
I seem to remember getting an error like this before, but I looked back
through the sig archives and didn't see any mention of it. I'm hoping you or
someone else knows what this means:

Python 1.5.2c1 (#56, Apr 12 1999, 14:19:52)  [CW PPC w/GUSI w/MSL]
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import DateTime
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "Henrietta:Utilities:Python 1.5.2c1:Extensions:DateTime:DateTime.py",
line 8, in ?
    from mxDateTime import *
  File "Henrietta:Utilities:Python 1.5.2c1:Extensions:DateTime:__init__.py",
line 7, in ?
    __package_info__ = """
ImportError: PythonCore: An import library was too new for a client.
>>>

-thx, Bob


From orla_redbird@crosswinds.net  Tue Oct 12 11:05:28 1999
From: orla_redbird@crosswinds.net (Gordon Worley)
Date: Tue, 12 Oct 1999 06:05:28 -0400 (EDT)
Subject: IDE API [was Re: [Pythonmac-SIG] Accessing front window in IDE]
Message-ID: <l031303011e1b001da698@[10.50.227.4]>

Joseph J. Strout wrote:
>>>
But then, how is this any different from mucking about with IDE
widgets, which is what I thought we wanted to avoid?  The IDE objects
already have methods like those above; apart from the names, I don't
see much difference.
<<<

Instead of developing a new API, why not just add some documentation to the
current code.  I know this really is Just's deal, but I for one would very
much like to see the W modules get some better descriptions.  Trying to
figure out what args that the callbacks on each type of widget takes is
quite a nusance to me, so even if we have to develop a wrapper module for W
and basic IDE functions, I think that it would be worth it for both the
hope of getting code documented and some easier to read function, class,
and method names (not that they aren't already pretty easy to read compared
to some modules in a few of those *other* languages).

- Gordon Worley
http://www.crosswinds.net/~orla_redbird/
mailto:orla_redbird@crosswinds.net




From just@letterror.com  Tue Oct 12 14:04:00 1999
From: just@letterror.com (Just van Rossum)
Date: Tue, 12 Oct 1999 15:04:00 +0200
Subject: [Pythonmac-SIG] Various IDE answers, comments
Message-ID: <l03102808b428ce1a7d92@[193.78.237.124]>

Sorry I didn't mix in with the discussions earlier: I was away from my
email for two weeks.

I've tried to digest all traffic, but I'm sure I've missed a couple of
points. Please email me privately if I failed to address some issues.

1. Class/method popup
Currently it indeed only works for files that use tabs for indentation.
This can easily be fixed separately from Oliver Steele's stuff, and I
probably will very soon. But...

2. Non-tab indentation
I might prefer to solve this as a file read/write filter which will
transform any indentation into tab based indentation upon read and
transform to the original style of indentation upon write. I would leave
any non-indenting whitespace as it is. Would this be acceptable or would
Oliver Steele's approach be preferred?

3. Printing
Although I appreciate the efforts to add printing code to the IDE, I will
not put those mods in the distribution until it supports the page setup and
print records properly. Most notably: paper size and from/to page number
fields.

4. Syntax coloring
Similarly, as long as the code isn't solid (like IDLE's) I won't put it
into the distribution, since it will be confusing for newbies if it screws
up. Also, there needs to be a user interface for setting the various
font/color properties. I meant to look into both issues long ago, but have
not been able to do so. Don't get me wrong: I think syntax coloring is very
important, and it's high on my wish list. I still hope to look into it
deeper soon-ish.

5. IDE API
There should indeed be a standard way to get to the frontmost script
editor. I think this is best done through the application object, something
like:

import W
app = W.getapplication()  # this already works
e = app.getfronteditor()

6. W documentation
I admit that W is poorly documented. My collegue Erik has built some
documentation for/from it for our RoboFog project. Please have a look at
http://www.petr.nl/robofog/reference/index.html (click "module W" in the
modules frame, other modules are RoboFog-specific) and let us know whether
it's any good. Besides that, I will attempt to add more docstrings.


Just




From jimmy@cs.cofc.edu  Tue Oct 12 17:23:27 1999
From: jimmy@cs.cofc.edu (James B. Wilkinson)
Date: Tue, 12 Oct 1999 12:23:27 -0400
Subject: [Pythonmac-SIG] Is the menu enabled?
Message-ID: <l03020904b42908b728bd@[153.9.17.27]>

Is there a way to get the toolbox to tell me whether a given menu is
enabled or not at a given time. I'd prefer not to have to redraw the menu
bar if I haven't actually made any changes in that property. Right now the
only thing I can think of is to keep an instance variable in the
application that gets set every time the enabled/disabled status of a menu
changes. That would probably work, but I worry about having the variable
somehow get out of step with what it's supposed to be tracking, and I'd
prefer to have direct access to the real truth.

I see that the Menu module implements the GetMenuHandle() function, but I
have no idea how to use it in Python. In C, I think I could use the handle
to get to the MenuInfo record and look at the enableFlags. How to do that
in Python is beyond me, and I don't know how to learn anything beyond what
I've done: type "import Menu" and "dir(Menu)". That tells me that
GetMenuHandle() is there, but not how to use it. Is there a good way to
answer questions like this, or do you just have to know?

Thanks.


-------------------------------------------------------------
Jimmy Wilkinson            | Perfesser of Computer Science
jimmy@cs.CofC.edu          | The College of Charleston
(843) 953-8160             | Charleston      SC        29424

If there is one word to describe me,
that word would have to be "profectionist".




From joe@strout.net  Tue Oct 12 17:51:06 1999
From: joe@strout.net (Joseph J. Strout)
Date: Tue, 12 Oct 1999 09:51:06 -0700
Subject: [Pythonmac-SIG] Is the menu enabled?
In-Reply-To: <l03020904b42908b728bd@[153.9.17.27]>
References: <l03020904b42908b728bd@[153.9.17.27]>
Message-ID: <v0421010bb42916e81690@[198.202.70.213]>

At 12:23 PM -0400 10/12/99, James B. Wilkinson wrote:

>I see that the Menu module implements the GetMenuHandle() function, but I
>have no idea how to use it in Python. In C, I think I could use the handle
>to get to the MenuInfo record and look at the enableFlags. How to do that
>in Python is beyond me, and I don't know how to learn anything beyond what
>I've done: type "import Menu" and "dir(Menu)". That tells me that
>GetMenuHandle() is there, but not how to use it. Is there a good way to
>answer questions like this, or do you just have to know?

Go it a step further.  Start with "inspect(Menu)" rather than 
"dir(Menu)", as you'll have easy access to the docstrings for each 
function this way.  Then do  "foo = GetMenuHandle()" (passing 
whatever it needs passed to get this step to work), and then 
"inspect(foo)", which will show you the methods and members of foo.

Now, foo is a resource object (resources and handles are the same 
thing in MacPython, for some odd reason).  You can get its data (it's 
just a member, IIRC).  That data is of the same format as documented 
in your favorite toolbox reference (i.e., whatever a MenuHandle 
refers to).  You can use the struct module to unpack this and get 
whatever you need out of it.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From savageb@pacbell.net  Tue Oct 12 19:06:20 1999
From: savageb@pacbell.net (Bob Savage)
Date: Tue, 12 Oct 1999 11:06:20 -0700
Subject: [Pythonmac-SIG] W & Piddle (was Re: IDE API)
Message-ID: <199910121806.LAA07011@smtp2.Stanford.EDU>

>From: Gordon Worley <orla_redbird@crosswinds.net>
>even if we have to develop a wrapper module for W
>and basic IDE functions, I think that it would be worth it for both the
>hope of getting code documented and some easier to read function, class,
>and method names (not that they aren't already pretty easy to read compared
>to some modules in a few of those *other* languages).
>

If some work went into a wrapper for W, would that lead to the ability to
use PiddleQD w/out the IDE?

-Bob


From joe@strout.net  Tue Oct 12 19:20:42 1999
From: joe@strout.net (Joseph J. Strout)
Date: Tue, 12 Oct 1999 11:20:42 -0700
Subject: [Pythonmac-SIG] W & Piddle (was Re: IDE API)
In-Reply-To: <199910121806.LAA07011@smtp2.Stanford.EDU>
References: <199910121806.LAA07011@smtp2.Stanford.EDU>
Message-ID: <v0421010eb4292bfc0ab3@[198.202.70.213]>

At 11:06 AM -0700 10/12/99, Bob Savage wrote:

>If some work went into a wrapper for W, would that lead to the ability to
>use PiddleQD w/out the IDE?

Not the sort of wrapper we were thinking of before.

It's probably quite possible to make piddleQD work without the IDE; 
but what we need isn't wrappers of existing code, so much as 
encapsulation.  The IDE classes, W classes, and Framework classes 
seem to be a bit intertwined, and it's not clear to me how to 
separate them.

What we'd need to do is load the Framework, and possible W, classes 
that are needed, and construct the piddleQD window on those.

But I'm curious: what are you doing with piddleQD that isn't more 
convenient to do in the IDE anyway?

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jpaint@sculpin.com  Wed Oct 13 00:55:42 1999
From: jpaint@sculpin.com (Jay Painter)
Date: Tue, 12 Oct 1999 16:55:42 -0700 (PDT)
Subject: [Pythonmac-SIG] OSAm Binary
Message-ID: <Pine.LNX.3.96.991012165333.6719A-100000@server.sculpin.com>

I uploaded a new ftp://sculpin.com/python/OSAm.hqx which has a built
plugin for Python 1.5.2c and a utility script (ascript.py) which is a
useful Python wrapper for the plugin, and also serves as an example for
using it.

Regards,
Jay Painter 



From A.M.INGRALDI@larc.nasa.gov  Wed Oct 13 16:29:04 1999
From: A.M.INGRALDI@larc.nasa.gov (Anthony M. Ingraldi)
Date: Wed, 13 Oct 1999 11:29:04 -0400
Subject: [Pythonmac-SIG] gnuplot on MacOS?
Message-ID: <199910131529.LAA26491@express.larc.nasa.gov>

On Thursday, September 30, 1999, Anthony M. Ingraldi wrote:

> 
> I'll probably contact the author of the Gnuplot package to see if he would
> be interested in incorporating the small mods into the Gnuplot distribution.
>

The author of Gnuplot has incorporated my changes to allow the package to
run on Mac OS.  They will be included in the next release.
--
   Tony Ingraldi
   A.M.INGRALDI@LaRC.NASA.GOV
   Phone : (757) 864-3039
   Fax   : (757) 864-7892



From joe@strout.net  Wed Oct 13 21:51:05 1999
From: joe@strout.net (Joseph J. Strout)
Date: Wed, 13 Oct 1999 13:51:05 -0700
Subject: [Pythonmac-SIG] [pythonpiddle] Re: bug in piddlePIL ? or PIL 1.0 final
In-Reply-To: <Pine.GSO.4.05.9910131128460.26106-100000@inky>
References: <Pine.GSO.4.05.9910131128460.26106-100000@inky>
Message-ID: <v04210107b42a9f7d48d7@[198.202.70.213]>

At 11:31 AM -0700 10/13/99, Robert Kern wrote:

>This is a known problem (Well, at least I knew it).  PIL's function
>arguments changed between the writing of piddlePIL and the release of PIL
>1.0.  piddlePIL will have to be rewritten for PIL 1.0.

Ugh.  This is a problem, because the version of PIL that comes with 
MacPython (the only flavor I really have access to) is 1.0b1. 
Apparently these function arguments changed between b1 and 1.0 final, 
because the test cases work just fine for me.

I'm CC'ing this to the Mac-Python list, hoping Jack or Just will tell 
me that they've got PIL 1.0 final for the Mac just about ready to go 
and will send it out next week.

Barring that, any ideas on how we can fix this in a way that will 
work on both 1.0 beta and 1.0 final (and presumably any version in 
between)?  At worst, we could check the version and have alternate 
code.

Chris wrote:

>I made one change to the end of piddlePIL's drawLine() flattening the
>tuples into one sequence:
>
>  self._pen.line( (x1,y1),(x2,y2) ) -> self._pen.line((x1,y1,x2,y2)).

To which I say, "Hmm."  This doesn't look like it addresses the 
traceback you found before, which was on "ink = 
self.im.draw_ink(ink)".  And you say that with this change, it all 
works except for the strings test -- what goes wrong with that one?

Thanks,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From jack@oratrix.nl  Thu Oct 14 10:08:57 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Thu, 14 Oct 1999 11:08:57 +0200
Subject: [Pythonmac-SIG] [pythonpiddle] Re: bug in piddlePIL ? or PIL 1.0
 final
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Wed, 13 Oct 1999 13:51:05 -0700 , <v04210107b42a9f7d48d7@[198.202.70.213]>
Message-ID: <19991014090858.5E08F35BB1E@snelboot.oratrix.nl>

> I'm CC'ing this to the Mac-Python list, hoping Jack or Just will tell 
> me that they've got PIL 1.0 final for the Mac just about ready to go 
> and will send it out next week.

Sorry, I won't have any time for this for (at least) the next month or so. 
But: everything is under CVS, so dropping in the new PIL distribution, 
building it and doing a distribution of that isn't all that much work....
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From steele@cs.brandeis.edu  Thu Oct 14 11:24:11 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Thu, 14 Oct 1999 06:24:11 -0400
Subject: [Pythonmac-SIG] space indented files
Message-ID: <199910141024.GAA16626@life.ai.mit.edu>

Just van Rossum <just@letterror.com> writes:
> 2. Non-tab indentation
> I might prefer to solve this as a file read/write filter which will
> transform any indentation into tab based indentation upon read and
> transform to the original style of indentation upon write. I would leave
> any non-indenting whitespace as it is. Would this be acceptable or would
> Oliver Steele's approach be preferred?

Let's call the approach that recognizes space-runs as equivalent to tabs,
"space recognition".  Let's call the approach that translates space-based
indentation into tab-based indentation, and back, "space translation".
"Space recognition" is what emacs does, and it's what you call "Oliver
Steele's approach" above.  "Space translation" is the method that you
"might prefer to solve this as".

I've actually implemented both.  SpaceTranslationPatch.py implements space
TRANSLATION:  it "makes the IDE change spaces in space-indented files to
tabs when it opens them, and back to spaces when it saves a file that was
converted this way."  PythonSpaceEditor.py implements space RECOGNITION:  it
"makes the IDE work with space-indented source files, by changing its
indentation-related commands to recognize that a sequence of spaces is
equivalent to a tab."

After using them both, I prefer the space translation approach, in
SpaceTranslationPatch.py.  (This is the same preference as Just's.)  I
initially implemented space recognition because I thought it would be good
to stick to a known design -- that is, to copy what emacs does, because it
works -- and because it's guaranteed to preserve the contents of the file
(opening and then saving a file with the space transformation approach may
change some spaces to tabs, if a file has both).  But it's too annoying on a
Mac to click in what looks like a tab stop and find yourself in the middle
of a set of spaces, or to use the left or right arrows or the backspace to
move what looks like one tab and find yourself moving only a space.  These
problems could be fixed with additional work, but at that point the text
presentation has diverged from the underlying data so much that I think it's
better just to change the underlying data instead, and keep the connection
between it and its presentation direct.  (Emacs users live with these
problems, but they seem worse in a Mac environment, maybe because I'm used
to a higher standard for direct manipulation)

SpaceTranslationPatch.py's algorithm is this:
(1) When an editor window is opened for a file, detect whether it contains
the string '\r    ' (a line that starts with four spaces).
(2) If it does, assume that it's a space-indented file.  If the file was
saved with Python font preferences, read the tab width from that.  Otherwise
if there are lines that start after eight spaces but not after four, assume
the tab stop is at 8 spaces.  Otherwise assume it's at four.
(3) Set the editor tab width to this.
(4) Replace all runs of tabstop*n spaces at the beginning of a line, with n
tabs.  Also, set a flag on the editor window, that marks that the conversion
was performed.
(5) When an editor window is saved, if this flag is set, replace all runs or
n tabs at the beginning of a line with tabstop*n spaces.  This eases
interoperability with other platforms that uses spaces to indent.

Remaining work:
- recognize files indented with a mixture of tabs and spaces
- test the heuristics on more files
- add a user indication and user control for the translation (I probably
won't do this one, but if it gets added to the IDE someone might want to do
this)


From steele@cs.brandeis.edu  Sat Oct 16 16:54:02 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Sat, 16 Oct 1999 11:54:02 -0400
Subject: [Pythonmac-SIG] CommandHistoryPatch
Message-ID: <199910161554.LAA18933@life.ai.mit.edu>

CommandHistoryPatch.py adds to the Python Interactive window of MacPython 
1.5.2c1 IDE two features that I've found useful in command shells and in
other IDEs.  (It probably works with earlier versions of MacPython, but I
haven't tested it with any versions other than 1.5.2c1.)

(1) Option-p (pi) replaces the current input text (everything after the
">>> " or "... " prompt) with the previous input text (the line or lines
after the previous ">>> " or "... " prompt).  Typing option-p repeatedly
cycles through all the previous inputs, from last to first, and then from
last to first again.

(2) Pressing the Return key to execute the entire input text, no matter
where in the text the selection is positioned.  For example, typing Return
when the insertion point is between the "+" and the "2" in the line
  >>> 1+2
prints the result "3", instead of ignoring the "2", asking for a
continuation line, and then displaying a message about a syntax error.

The patch can be downloaded from
<http://www.cs.brandeis.edu/~steele/sources/python.html>.

I recommend we think about including this in the IDE at some point, but it's
probably useful for folks to play around with it and weigh in on how well it
works for them first.  In particular, I'm unsure about option-p for the
"previous" command (it keeps you from typing pi, and the natural pair for
"next", option-n, doesn't easily produce a capturable keystroke), although
each alternative has its own issues (for example, cmd-p would work, but then
cmd-n is already taken...).


From joe@strout.net  Sat Oct 16 20:04:48 1999
From: joe@strout.net (Joseph J. Strout)
Date: Sat, 16 Oct 1999 12:04:48 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
In-Reply-To: <199910161554.LAA18933@life.ai.mit.edu>
Message-ID: <l03102802b42e7c379cf4@[24.30.159.231]>

At 8:54 AM -0700 10/16/99, Oliver Steele wrote:

>I recommend we think about including this in the IDE at some point, but it's
>probably useful for folks to play around with it and weigh in on how well it
>works for them first.

This sounds very cool.  I think this key mapping is similar to what MCL
uses, though I wouldn't swear by it -- I also wouldn't claim it's the best
possible.  Is it possible to detect control-up and control-down?  Or maybe
page up/down or home/end?  Or control-p and control-n?

>each alternative has its own issues (for example, cmd-p would work, but then
>cmd-n is already taken...).

Yeah, and cmd-p usually means "print" (as it does, for example, with the
printing patch installed).

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'




From savageb@pacbell.net  Mon Oct 18 06:15:30 1999
From: savageb@pacbell.net (savageb)
Date: Sun, 17 Oct 1999 22:15:30 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
Message-ID: <0FJS00A739A02G@mta3.snfc21.pbi.net>

> 
> This sounds very cool.  I think this key mapping is similar to what MCL
> uses, though I wouldn't swear by it -- I also wouldn't claim it's the best
> possible.  Is it possible to detect control-up and control-down?  Or maybe
> page up/down or home/end?  Or control-p and control-n?
>

Yes, TCSH seems to respond to simply using the up/down keys, but I think
ctrl-up/ctrl-down seems logical.

Bob


From cbarker@jps.net  Mon Oct 18 18:45:32 1999
From: cbarker@jps.net (Chris Barker)
Date: Mon, 18 Oct 1999 10:45:32 -0700
Subject: [Pythonmac-SIG] space indented files
References: <199910150501.BAA25996@python.org>
Message-ID: <380B5CBC.E6CA3F83@jps.net>

This is a multi-part message in MIME format.
--------------1B1EF91C859852EFAA66EB3D
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Oliver Steele wrote:
****************************************
Remaining work:
- recognize files indented with a mixture of tabs and spaces
- test the heuristics on more files
- add a user indication and user control for the translation (I probably
won't do this one, but if it gets added to the IDE someone might want to
do
this)
****************************************

This sounds great!

I don't know if it will be helpful, but I've enclosed a script I got
from someone that does tab-space-tab+space translation for you (I'd love
to credit the author, but s/he did't put a name in the file, and I don't
remember who i got it from). I suppose the tricky part is to figure out
what format was orignally used. I have to say, though, that I'd be just
as happy if tab+spaces was NOT preserved! using the two together is a
great way to invite problems.

Note also, that if the file uses a mixture of tabs and spaces, you have
to use a tabstop of 8 spaces, as this is what Python assumes.

-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------
--------------1B1EF91C859852EFAA66EB3D
Content-Type: text/plain; charset=us-ascii;
 name="tabcleaner.py"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="tabcleaner.py"

#!/usr/bin/python

import tokenize
import string

TABSONLY = 'TABSONLY'
SPACESONLY = 'SPACESONLY'
MIXED = 'MIXED'

class PyText:
    def __init__(self, fnm, optdict):
        self.optdict = optdict
        self.fnm = fnm
    	self.txt = open(self.fnm, 'r').readlines()
    	self.indents = [(0, 0, )]
    	self.lnndx = 0
    	self.indentndx = 0
    def getline(self):
        if self.lnndx < len(self.txt):
            txt = self.txt[self.lnndx]
            self.lnndx = self.lnndx + 1
        else:
            txt = ''
        return txt
    def tokeneater(self, type, token, start, end, line):
        if type == tokenize.INDENT:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, start[0]-1)
            self.indents.append((lvl+1, start[0]-1,))
        elif type == tokenize.DEDENT:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, start[0]-1)
            self.indents.append((lvl-1, start[0]-1,))
        elif type == tokenize.ENDMARKER:
            (lvl, s) = self.indents[-1]
            self.indents[-1] = (lvl, s, len(self.txt))
    def split(self, ln):
        content = string.lstrip(ln)
        if not content:
            return ('', '\n')
        lead = ln[:len(ln) - len(content)]
        lead = string.expandtabs(lead)
        return (lead, content)
        
    def process(self):
        style = self.optdict.get('style', SPACESONLY)
        indent = string.atoi(self.optdict.get('indent', '4'))
        tabsz = string.atoi(self.optdict.get('tabs', '8'))
        print 'file %s -> style %s, tabsize %d, indent %d' % (self.fnm, style, tabsz, indent)
        tokenize.tokenize(self.getline, self.tokeneater)
        #import pprint
        #pprint.pprint(self.indents)
        new = []
        for (lvl, s, e) in self.indents:
            if s >= len(self.txt):
                break
            if s == e:
                continue
            oldlead, content = self.split(self.txt[s])
            #print "oldlead", len(oldlead), `oldlead`
            if style == TABSONLY:
                newlead = '\t'*lvl
            elif style == SPACESONLY:
                newlead = ' '*(indent*lvl)
            else:
                sz = indent*lvl
                t,spcs = divmod(sz, tabsz)
                newlead = '\t'*t + ' '*spcs
            new.append(newlead + content)
            for ln in self.txt[s+1:e]:
                lead, content = self.split(ln)
                #print "lead:", len(lead)
                new.append(newlead + lead[len(oldlead):] + content)
        self.save(new)
        #print "---", self.fnm
        #for ln in new:
        #    print ln,
        #print
                
    def save(self, txt):
        bakname = os.path.splitext(self.fnm)[0]+'.bak'
        print "backing up", self.fnm, "to", bakname
        #print os.getcwd()
        try:
            os.rename(self.fnm, bakname)
        except os.error:
            os.remove(bakname)
            os.rename(self.fnm, bakname)
        open(self.fnm, 'w').writelines(txt)

def test():
    tc = PyText('test1.py')
    tc.process()
    tc = PyText('test1.py')
    tc.process(style=TABSONLY)
    tc = PyText('test1.py')
    tc.process(style=MIXED, indent=4, tabs=8)
    tc = PyText('test1.py')
    tc.process(style=MIXED, indent=2, tabs=8)
 
def cleanfile(fnm, d):
    if os.path.isdir(fnm) and not os.path.islink(fnm):
        names = os.listdir(fnm)
        for name in names:
            fullnm = os.path.join(fnm, name)
            if (os.path.isdir(fullnm) and not os.path.islink(fullnm)) or \
            	os.path.normcase(fullnm[-3:]) == ".py":
            	cleanfile(fullnm, d)
        return
    tc = PyText(fnm, d)
    tc.process()
  
usage="""\
%s [options] [path...]
 options
  -T : reformat to TABS ONLY
  -S : reformat to SPACES ONLY ( -i option is important)
  -M : reformat to MIXED SPACES / TABS ( -t and -i options important)
  -t<n> : tab is worth <n> characters
  -i<n> : indents should be <n> characters 
  -h : print this text
 path is file or directory
"""
if __name__ == '__main__':
    import sys, getopt, os
    opts, args = getopt.getopt(sys.argv[1:], "TSMht:i:")
    d = {}
    #print `opts`
    for opt in opts:
        if opt[0] == '-T':
            d['style'] = TABSONLY
        elif opt[0] == '-S':
            d['style'] = SPACESONLY
        elif opt[0] == '-M':
            d['style'] = MIXED
        elif opt[0] == '-t':
            d['tabs'] = opt[1]
        elif opt[0] == '-i':
            d['indent'] = opt[1]
        elif opt[0] == '-h':
            print usage % sys.argv[0]
            sys.exit(0)
    if not args:
        print usage % sys.argv[0]
    for arg in args:
        cleanfile(arg, d)
    
    

--------------1B1EF91C859852EFAA66EB3D--



From cbarker@jps.net  Mon Oct 18 18:48:22 1999
From: cbarker@jps.net (Chris Barker)
Date: Mon, 18 Oct 1999 10:48:22 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
References: <199910170501.BAA25937@python.org>
Message-ID: <380B5D66.46BD581F@jps.net>

Another great addition form Oliver Steele!

I've been waiting desperately for this! 

Why not bind "previous command" to the "up arrow"?
That's how it is at the Unix command line, and it's very handy, and
easy. I can't imagine what else you would want the up arrow for.

-Chris

-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------


From joe@strout.net  Mon Oct 18 18:54:08 1999
From: joe@strout.net (Joseph J. Strout)
Date: Mon, 18 Oct 1999 10:54:08 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
In-Reply-To: <380B5D66.46BD581F@jps.net>
References: <199910170501.BAA25937@python.org> <380B5D66.46BD581F@jps.net>
Message-ID: <v0421010fb4310eee10a5@[198.202.70.213]>

At 10:48 AM -0700 10/18/99, Chris Barker wrote:

>Why not bind "previous command" to the "up arrow"?
>That's how it is at the Unix command line, and it's very handy, and
>easy. I can't imagine what else you would want the up arrow for.

You use the up arrow for moving the cursor up, e.g., in combination 
with the shift key to select text for copying and such.

Granted, this isn't a terribly useful thing to do (compared to 
command history), and probably nobody would miss it too much.  But 
the console window is a text-editing window, like any other, and it 
might be hard to muck with its cursor interface in such a fundamental 
way.  Just guessing here, though.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From cbarker@jps.net  Mon Oct 18 21:04:22 1999
From: cbarker@jps.net (Chris Barker)
Date: Mon, 18 Oct 1999 13:04:22 -0700
Subject: [Pythonmac-SIG] "Build Applet" not Drag and drop anymore?
References: <199910170501.BAA25937@python.org>
Message-ID: <380B7D46.B718ABE4@jps.net>

Hi,

I'm using what I think is the latest version of MacPython (1.5.2c1)

BuildApplet and BuildApplication don't seem to be drag and dropable
anymore. I miss that. Is there a way I can change them to make it work?
(I do have argv emulation enabled)

-Chris


-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------


From maccgi@bellsouth.net  Mon Oct 18 21:19:25 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Mon, 18 Oct 1999 16:19:25 -0400
Subject: [Pythonmac-SIG] "Build Applet" not Drag and drop anymore?
In-Reply-To: <380B7D46.B718ABE4@jps.net>
References: <199910170501.BAA25937@python.org> <380B7D46.B718ABE4@jps.net>
Message-ID: <v04205500b4313013ca38@[209.214.71.70]>

At 13:04 -0700 10/18/1999, Chris Barker wrote:
>Hi,
>
>I'm using what I think is the latest version of MacPython (1.5.2c1)
>
>BuildApplet and BuildApplication don't seem to be drag and dropable
>anymore. I miss that. Is there a way I can change them to make it work?
>(I do have argv emulation enabled)

They are working for me, so I would assume that something is hosed 
with your local setup. I can't imagine what that would be and will 
look to the list gurus for a reasonable guess about what's wrong.


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From joe@strout.net  Mon Oct 18 21:33:32 1999
From: joe@strout.net (Joseph J. Strout)
Date: Mon, 18 Oct 1999 13:33:32 -0700
Subject: [Pythonmac-SIG] "Build Applet" not Drag and drop anymore?
In-Reply-To: <v04205500b4313013ca38@[209.214.71.70]>
References: <199910170501.BAA25937@python.org> <380B7D46.B718ABE4@jps.net>
 <v04205500b4313013ca38@[209.214.71.70]>
Message-ID: <v04210101b4313475568b@[198.202.70.213]>

At 4:19 PM -0400 10/18/99, Richard Gordon wrote:

>They are working for me, so I would assume that something is hosed 
>with your local setup. I can't imagine what that would be and will 
>look to the list gurus for a reasonable guess about what's wrong.

Have you tried rebuilding the desktop?  Often drag-and-drop apps that 
won't accept drops just means tha the Finder needs a kick in the 
pants...

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From savageb@pacbell.net  Tue Oct 19 02:02:55 1999
From: savageb@pacbell.net (savageb)
Date: Mon, 18 Oct 1999 18:02:55 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
Message-ID: <199910190102.SAA13420@mail-gw5.pacbell.net>

>From: Chris Barker <cbarker@jps.net>
>Subject: [Pythonmac-SIG] CommandHistoryPatch
>Date: Mon, Oct 18, 1999, 10:48 AM
>

> Another great addition form Oliver Steele!

I agree.

>
> Why not bind "previous command" to the "up arrow"?
> That's how it is at the Unix command line, and it's very handy, and
> easy. I can't imagine what else you would want the up arrow for.

Hmmm...well at the moment I tend to use the up arrow to move the cursor
around (I know, "Get a mouse!"), but perhaps I wouldn't be as inclined to do
that if I had a command history, since that is a big part of why I do it in
the first place. Would it really be so bad to make it ctrl-up/ctrl-down?

Bob


From maccgi@bellsouth.net  Tue Oct 19 04:29:12 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Mon, 18 Oct 1999 23:29:12 -0400
Subject: [Pythonmac-SIG] A General Thank You
Message-ID: <v04205509b431941c4ac6@[209.214.72.147]>

I just wanted to take a minute to offer a general thank you to 
everyone who is providing patches for the IDE and MacPython in 
general (OSAm is pretty cool if you are an applescript kind of guy, 
btw). This stuff has helped me a lot in day to day work and I am 
hoping that it will get even better. Of course, thanks are also due 
to Just, Jack, and Guido, but I suppose that is implicit in thanking 
anyone else.

So, that's my good deed for the day, now back to evil....

Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From jack@oratrix.nl  Tue Oct 19 11:22:40 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Tue, 19 Oct 1999 12:22:40 +0200
Subject: [Pythonmac-SIG] CommandHistoryPatch
In-Reply-To: Message by "savageb" <savageb@pacbell.net> ,
 Mon, 18 Oct 1999 18:02:55 -0700 , <199910190102.SAA13420@mail-gw5.pacbell.net>
Message-ID: <19991019102241.5B04535BB1E@snelboot.oratrix.nl>

> > Why not bind "previous command" to the "up arrow"?
> > That's how it is at the Unix command line, and it's very handy, and
> > easy. I can't imagine what else you would want the up arrow for.
> 
> Hmmm...well at the moment I tend to use the up arrow to move the cursor
> around (I know, "Get a mouse!"), but perhaps I wouldn't be as inclined to do
> that if I had a command history, since that is a big part of why I do it in
> the first place. Would it really be so bad to make it ctrl-up/ctrl-down?

I think in this case we can have our cake and eat it. Mark Hammond's PythonWin 
does more-or-less what you say here: with the arrow keys you can walk to a 
certain line. If you are on any line and you start typing[*] the line is 
copied to the bottom line of the screen (minus any prompts prefixing it), 
along with your cursor and your edits are being done. And, of course, if you 
want to use the mouse in stead of the cursor keys to position the cursor that 
also works.

[*] I think that you have to hit an extra return in PythonWin to do the 
copying, but I don't see a good reason for copying that behaviour.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From managan@llnl.gov  Tue Oct 19 16:48:33 1999
From: managan@llnl.gov (Rob Managan)
Date: Tue, 19 Oct 1999 08:48:33 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch
In-Reply-To: <19991019102241.5B04535BB1E@snelboot.oratrix.nl>
References: <19991019102241.5B04535BB1E@snelboot.oratrix.nl>
Message-ID: <v04210101b432429dde11@[128.115.232.99]>

>
>I think in this case we can have our cake and eat it. Mark Hammond's PythonWin
>does more-or-less what you say here: with the arrow keys you can walk to a
>certain line. If you are on any line and you start typing[*] the line is
>copied to the bottom line of the screen (minus any prompts prefixing it),
>along with your cursor and your edits are being done. And, of course, if you
>want to use the mouse in stead of the cursor keys to position the cursor that
>also works.
>
>[*] I think that you have to hit an extra return in PythonWin to do the
>copying, but I don't see a good reason for copying that behaviour.
>--
>Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
>Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++

The plotting program Igor does something similar. The up arrow 
selects a line at a time in the command history. Then an Enter copies 
any selection  to the input line. Just another idea on what may work.

*-*-*-*-*-*-*-*-*-*-**-*-*-*-*-*-*-*-*-*-*-
Rob Managan       <mailto://managan@llnl.gov>
LLNL                      ph:  925-423-0903
P.O. Box 808, L-095       FAX: 925-422-3389
Livermore, CA  94551-0808



From cbarker@jps.net  Tue Oct 19 18:26:00 1999
From: cbarker@jps.net (Chris Barker)
Date: Tue, 19 Oct 1999 10:26:00 -0700
Subject: [Pythonmac-SIG] Re: "Build Applet" not Drag and drop anymore?
References: <199910190502.BAA13814@python.org>
Message-ID: <380CA9A8.D2A74BDB@jps.net>

Hi folks,

>>They are working for me, so I would assume that something is hosed 
>>with your local setup. I can't imagine what that would be and will 
>>look to the list gurus for a reasonable guess about what's wrong.

>Have you tried rebuilding the desktop?  Often drag-and-drop apps that 
>won't accept drops just means tha the Finder needs a kick in the 
>pants...

Bingo!  I guess dispite what Apple wants you to think, there really is a
lot of tricky little stuff you have to learn on a Mac.

-Chris


-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------


From jack@oratrix.nl  Tue Oct 19 23:12:21 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Wed, 20 Oct 1999 00:12:21 +0200
Subject: [Pythonmac-SIG] Re: "Build Applet" not Drag and drop anymore?
In-Reply-To: Message by Chris Barker <cbarker@jps.net> ,
 Tue, 19 Oct 1999 10:26:00 -0700 , <380CA9A8.D2A74BDB@jps.net>
Message-ID: <19991019221227.E3DDDEA119@oratrix.oratrix.nl>

Recently, Chris Barker <cbarker@jps.net> said:
> >Have you tried rebuilding the desktop?  Often drag-and-drop apps that 
> >won't accept drops just means tha the Finder needs a kick in the 
> >pants...
> 
> Bingo!  I guess dispite what Apple wants you to think, there really is a
> lot of tricky little stuff you have to learn on a Mac.

I think I am to blame for this, because I see this problem (having to
rebuild the desktop to get an applet to work) a lot more with Python
than with anything else. I think I don't fully understand either how
to kick the finder after BuildApplet has created a new application or
exactly how BNDL resources, the Inited finfo bit and all these other
thingies work together.

So if someone wants to have a look at BuildApplet and friends and make 
a fool of me in public: by all means, be my guest:-)

--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 


From joe@strout.net  Wed Oct 20 01:10:14 1999
From: joe@strout.net (Joseph J. Strout)
Date: Tue, 19 Oct 1999 17:10:14 -0700
Subject: [Pythonmac-SIG] "import library too new"?
Message-ID: <v04210108b432b88e63d4@[198.202.70.213]>

I remember this came up before... but I can't remember the answer. 
Can somebody please tell me what this means:

>>> import QDRotate
Traceback (innermost last):
   File "<input>", line 1, in ?
ImportError: PythonCore: An import library was too new for a client.


I thought maybe it meant I have to rebuild the QDRotate shared lib, 
but I did that and still get this error.  It worked fine in 1.5.2b1, 
but 1.5.2c1 gives me this.  Can somebody give me a clue?

Thanks,
-- Joe
,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From maccgi@bellsouth.net  Wed Oct 20 02:22:17 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Tue, 19 Oct 1999 21:22:17 -0400
Subject: [Pythonmac-SIG] "import library too new"?
In-Reply-To: <v04210108b432b88e63d4@[198.202.70.213]>
References: <v04210108b432b88e63d4@[198.202.70.213]>
Message-ID: <v04205504b432c8da225d@[209.214.64.73]>

At 17:10 -0700 10/19/1999, Joseph J. Strout wrote:
>I thought maybe it meant I have to rebuild the QDRotate shared lib, 
>but I did that and still get this error.  It worked fine in 1.5.2b1, 
>but 1.5.2c1 gives me this.  Can somebody give me a clue?

I don't know that it's much of a clue, but you aren't the only one to 
run into this lately, viz.:

At 17:54 -0700 10/11/1999, savageb wrote:
>I am having a hard time getting mxDateTime to run with Python 1.5.2c1. Joe,
>I seem to remember getting an error like this before, but I looked back
>through the sig archives and didn't see any mention of it. I'm hoping you or
>someone else knows what this means:
>
>Python 1.5.2c1 (#56, Apr 12 1999, 14:19:52)  [CW PPC w/GUSI w/MSL]
>Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> import DateTime
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>  File "Henrietta:Utilities:Python 1.5.2c1:Extensions:DateTime:DateTime.py",
>line 8, in ?
>    from mxDateTime import *
>  File "Henrietta:Utilities:Python 1.5.2c1:Extensions:DateTime:__init__.py",
>line 7, in ?
>    __package_info__ = """
>ImportError: PythonCore: An import library was too new for a client.
> >>>
>
>-thx, Bob


Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From jack@oratrix.nl  Wed Oct 20 09:28:14 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Wed, 20 Oct 1999 10:28:14 +0200
Subject: [Pythonmac-SIG] "import library too new"?
In-Reply-To: Message by "Joseph J. Strout" <joe@strout.net> ,
 Tue, 19 Oct 1999 17:10:14 -0700 , <v04210108b432b88e63d4@[198.202.70.213]>
Message-ID: <19991020082815.2F65435BB1E@snelboot.oratrix.nl>

> I remember this came up before... but I can't remember the answer. 
> Can somebody please tell me what this means:
> 
> >>> import QDRotate
> Traceback (innermost last):
>    File "<input>", line 1, in ?
> ImportError: PythonCore: An import library was too new for a client.
> 
> 
> I thought maybe it meant I have to rebuild the QDRotate shared lib, 
> but I did that and still get this error.

That should have worked. If QDRotate was linked against PythonCore 1.5.2b1 and 
you re-link it against 1.5.2c1 it should now work. Are you sure (a) you're 
importing the relinked version and (b) you relinked against the new PythonCore?

As you apparently have CodeWarrior, could you try running it under the 
debugger? The debugger log window shows which code fragments are loaded in 
which order, maybe this gives a clue...
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From joe@strout.net  Wed Oct 20 16:47:59 1999
From: joe@strout.net (Joseph J. Strout)
Date: Wed, 20 Oct 1999 08:47:59 -0700
Subject: [Pythonmac-SIG] "import library too new"?
In-Reply-To: <19991020082815.2F65435BB1E@snelboot.oratrix.nl>
References: <19991020082815.2F65435BB1E@snelboot.oratrix.nl>
Message-ID: <v04210104b4339415f20e@[198.202.70.213]>

At 10:28 AM +0200 10/20/99, Jack Jansen wrote:

> > I thought maybe it meant I have to rebuild the QDRotate shared lib,
> > but I did that and still get this error.
>
>That should have worked...

You're right, it does.  I'd run into a bug in CodeWarrior Pro 5.2 -- 
I still don't understand it; it was looking in the wrong place for 
the PythonCore library.  Even after I deleted the old one, and 
removed and re-added the new one (to the project) a few dozen times, 
making sure it was in the path, etc., CodeWarrior remained hopelessly 
confused and gave me mysterious internal errors.

I finally found a work-around, which was to copy PythonCore into the 
project folder.  Why this works, yet another location also in the 
project paths does not, I have no idea.  But I finally got QDRotate 
rebuilt with the correct shlb and it now works under 1.5.2c1.

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From savageb@pacbell.net  Wed Oct 20 17:36:37 1999
From: savageb@pacbell.net (Bob Savage)
Date: Wed, 20 Oct 1999 09:36:37 -0700
Subject: [Pythonmac-SIG] "import library too new"?
In-Reply-To: <v04210104b4339415f20e@[198.202.70.213]>
Message-ID: <B4333DA4.71%savageb@pacbell.net>

>>> I thought maybe it meant I have to rebuild the QDRotate shared lib,
>>> but I did that and still get this error.
>> 
>> That should have worked...
> 
> You're right, it does.

Joe, is there a chance you could work some of that same magic on the
mxDateTime shared library? I'ld like to be able to use it with the new
version of MacPython.

Bob



From patrick@swdev.com  Thu Oct 21 06:11:32 1999
From: patrick@swdev.com (Patrick Curtain)
Date: Wed, 20 Oct 1999 22:11:32 -0700
Subject: [Pythonmac-SIG] Curriculum, Course Outlines?
Message-ID: <380EA078.4E30A82B@swdev.com>

Hi All!

I'm in the middle of putting together a course outline for teaching
programming and I'm wanting to use Python.

Does anyone have any materials?  Course suggestions?

Thanks for any help!
--p

--
Patrick Curtain, Husband & Father      ( i also write software )
503 781 6199    patrick@swdev.com  http://www.swdev.com/




From jack@oratrix.nl  Thu Oct 21 11:16:58 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Thu, 21 Oct 1999 12:16:58 +0200
Subject: [Pythonmac-SIG] Curriculum, Course Outlines?
In-Reply-To: Message by Patrick Curtain <patrick@swdev.com> ,
 Wed, 20 Oct 1999 22:11:32 -0700 , <380EA078.4E30A82B@swdev.com>
Message-ID: <19991021101659.41CEF35BB1E@snelboot.oratrix.nl>

> I'm in the middle of putting together a course outline for teaching
> programming and I'm wanting to use Python.
> 
> Does anyone have any materials?  Course suggestions?

Patrick,
unless you want to teach only Mac-specific things, which I can't imagine, I 
think you should also ask this in a general forum like python-list@python.org. 
There should be very little machine-dependent in a Python course, except for 
the first lesson (which will probably focus on IDE for the Mac, PythonWin for 
Windows and Emacs for unix).
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From michael.wolff@strath.ac.uk  Thu Oct 21 11:51:49 1999
From: michael.wolff@strath.ac.uk (m wolff)
Date: Thu, 21 Oct 1999 11:51:49 +0100
Subject: [Pythonmac-SIG] Zope et al
In-Reply-To: <19991021101659.41CEF35BB1E@snelboot.oratrix.nl>
References: Message by Patrick Curtain <patrick@swdev.com> ,	     Wed, 20
 Oct 1999 22:11:32 -0700 , <380EA078.4E30A82B@swdev.com>
Message-ID: <v03102801b4349f1e5ddc@[130.159.116.159]>

Hello.

Are there any initiatives to implement Zope on MacOS?

michael

right hemisphere spirit trapped inside a left brain mind.




From jack@oratrix.nl  Thu Oct 21 13:16:36 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Thu, 21 Oct 1999 14:16:36 +0200
Subject: [Pythonmac-SIG] Zope et al
In-Reply-To: Message by m wolff <michael.wolff@strath.ac.uk> ,
 Thu, 21 Oct 1999 11:51:49 +0100 , <v03102801b4349f1e5ddc@[130.159.116.159]>
Message-ID: <19991021121636.C679535BB1E@snelboot.oratrix.nl>

> Hello.
> 
> Are there any initiatives to implement Zope on MacOS?

As far as I know Zope heavily depends on thread support, so for the moment it 
is impossible to implement under MacPython.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jhrsn@pop.pitt.edu  Thu Oct 21 14:46:35 1999
From: jhrsn@pop.pitt.edu (Jim Harrison)
Date: Thu, 21 Oct 1999 09:46:35 -0400
Subject: [Pythonmac-SIG] Curriculum, Course Outlines?
In-Reply-To: <380EA078.4E30A82B@swdev.com>
Message-ID: <B434917A.64E%jhrsn@pop.pitt.edu>

on 10/21/99 1:11 AM, Patrick Curtain at patrick@swdev.com wrote:

> I'm in the middle of putting together a course outline for teaching
> programming and I'm wanting to use Python.
> 
> Does anyone have any materials?  Course suggestions?

Patrick-

If you haven't already, I'd recommend you take a look at Guido van Rossum's
proposal "Computer Programming for Everybody"
<http://www.python.org/doc/essays/cp4e.html>. He and collaborators propose
to create a curriculum and tools for teaching beginning programming in
secondary schools and college, using Python. The proposal has been funded by
DARPA and work is beginning this fall. It's possible you could collaborate,
or at least get some direction from the materials they are developing.

Best wishes,

Jim Harrison
________________________________________________________________________
James H. Harrison, Jr., MD, PhD
University of Pittsburgh Health System
C920 PUH, 200 Lothrop Street
Pittsburgh, PA 15213

jhrsn@pop.pitt.edu | voice: 412-647-5529 | fax: 412-647-9588

"If you want sense, you'll have to make it yourself!!"-Norton Juster
________________________________________________________________________



From joe@strout.net  Thu Oct 21 19:19:24 1999
From: joe@strout.net (Joseph J. Strout)
Date: Thu, 21 Oct 1999 11:19:24 -0700
Subject: [Pythonmac-SIG] change in cPickle format?
Message-ID: <v04210112b43508ff16cc@[198.202.70.213]>

I'm having trouble with a pickle file which was created under 
MacPython 1.5.2b1.  I can't seem to load it now, as shown here:

	>>> f = open(_fontprefix + ":metrics.dat")
	>>> cPickle.load(f)
	Traceback (innermost last):
	  File "<input>", line 1, in ?
	EOFError:

Has the cPickle format changed?  Or has this file simply gotten 
corrupted somehow?

PIDDLE users: is piddlePIL working for you?  If you comment out the 
except block at line 48 (substituting a dummy "finally" block 
instead), do you get an error like the above when you run the code?

Thanks for any insight in pinning this down...

-- Joe
,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From cmann@pobox.com  Thu Oct 21 20:02:12 1999
From: cmann@pobox.com (chris mann)
Date: Thu, 21 Oct 1999 15:02:12 -0400
Subject: [Pythonmac-SIG] problems with smtplib?
Message-ID: <B434DB73.D5E%cmann@pobox.com>

I'm having problems with the smtplib module on Mac.

if I run the following code on windows or linux it works fine, but fails on
mac:
import smtplib
s = smtplib.SMTP('smtp.myhost.com')
s.helo()

i get the following error:
Traceback (innermost last):
  File "<input>", line 1, in ?
  File "Dev:Scripting:Python 1.5.2c1:Lib:smtplib.py", line 290, in helo
    name=socket.gethostbyaddr(socket.gethostname())[0]
error: host not found

any ideas why this is only failing on Mac?

thanks
-chris



From just@letterror.com  Thu Oct 21 20:00:58 1999
From: just@letterror.com (Just van Rossum)
Date: Thu, 21 Oct 1999 21:00:58 +0200
Subject: [Pythonmac-SIG] change in cPickle format?
In-Reply-To: <v04210112b43508ff16cc@[198.202.70.213]>
Message-ID: <l03102808b4351325c368@[193.78.237.162]>

At 11:19 AM -0700 10/21/99, Joseph J. Strout wrote:
>I'm having trouble with a pickle file which was created under
>MacPython 1.5.2b1.  I can't seem to load it now, as shown here:
>
>	>>> f = open(_fontprefix + ":metrics.dat")
>	>>> cPickle.load(f)
>	Traceback (innermost last):
>	  File "<input>", line 1, in ?
>	EOFError:
>
>Has the cPickle format changed?  Or has this file simply gotten
>corrupted somehow?

Erm, shouldn't you open that file in binary mode?

Just




From joe@strout.net  Thu Oct 21 20:20:15 1999
From: joe@strout.net (Joseph J. Strout)
Date: Thu, 21 Oct 1999 12:20:15 -0700
Subject: [Pythonmac-SIG] change in cPickle format?
In-Reply-To: <l03102808b4351325c368@[193.78.237.162]>
References: <l03102808b4351325c368@[193.78.237.162]>
Message-ID: <v04210116b435179784dd@[198.202.70.213]>

At 9:00 PM +0200 10/21/99, Just van Rossum wrote:

> >	>>> f = open(_fontprefix + ":metrics.dat")
> >	>>> cPickle.load(f)
>Erm, shouldn't you open that file in binary mode?

Ah.  Umm.  You may have a point there.

I used to always use pickle, which stores its data in text files, and 
so I always open in text mode.  Converting to cPickle, I neglected to 
change this.  That probably explains the problem (though since the 
file was generated in text mode as well, the damage is probably 
already done -- I'll have to rebuild the file).

Thanks Just!

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'


From just@letterror.com  Thu Oct 21 20:37:25 1999
From: just@letterror.com (Just van Rossum)
Date: Thu, 21 Oct 1999 21:37:25 +0200
Subject: [Pythonmac-SIG] problems with smtplib?
In-Reply-To: <B434DB73.D5E%cmann@pobox.com>
Message-ID: <l03102809b4351aa78897@[193.78.237.162]>

At 3:02 PM -0400 10/21/99, chris mann wrote:
>I'm having problems with the smtplib module on Mac.
>
>if I run the following code on windows or linux it works fine, but fails on
>mac:
>import smtplib
>s = smtplib.SMTP('smtp.myhost.com')
>s.helo()
>
>i get the following error:
>Traceback (innermost last):
>  File "<input>", line 1, in ?
>  File "Dev:Scripting:Python 1.5.2c1:Lib:smtplib.py", line 290, in helo
>    name=socket.gethostbyaddr(socket.gethostname())[0]
>error: host not found
>
>any ideas why this is only failing on Mac?

Works for me. Can it be your network setup? Firewall?
Can you see which of the two socket calls fail?

Just




From cmann@pobox.com  Thu Oct 21 20:54:07 1999
From: cmann@pobox.com (chris mann)
Date: Thu, 21 Oct 1999 15:54:07 -0400
Subject: [Pythonmac-SIG] problems with smtplib?
In-Reply-To: <l03102809b4351aa78897@[193.78.237.162]>
Message-ID: <B434E79F.D71%cmann@pobox.com>

on 10/21/99 3:37 PM, Just van Rossum at just@letterror.com wrote:

> At 3:02 PM -0400 10/21/99, chris mann wrote:
>> I'm having problems with the smtplib module on Mac.
>> 
>> if I run the following code on windows or linux it works fine, but fails on
>> mac:
>> import smtplib
>> s = smtplib.SMTP('smtp.myhost.com')
>> s.helo()
>> 
>> i get the following error:
>> Traceback (innermost last):
>> File "<input>", line 1, in ?
>> File "Dev:Scripting:Python 1.5.2c1:Lib:smtplib.py", line 290, in helo
>> name=socket.gethostbyaddr(socket.gethostname())[0]
>> error: host not found
>> 
>> any ideas why this is only failing on Mac?
> 
> Works for me. Can it be your network setup? Firewall?
> Can you see which of the two socket calls fail?

Could be the firewall.  Or it could be the network setup, but not sure why
it would work on Windows, but not Mac.  They both get their ip address via
DHCP (the linux box has a static ip address)

It's failing in the socket.gethostbyaddr() call.  The socket.gethostname()
call returns the ip address of the machine.

thanks
-chris





From doug@sonosphere.com  Thu Oct 21 22:44:56 1999
From: doug@sonosphere.com (Doug Wyatt)
Date: Thu, 21 Oct 1999 17:44:56 -0400
Subject: [Pythonmac-SIG] problems with smtplib?
In-Reply-To: <l03102809b4351aa78897@[193.78.237.162]>
References: <l03102809b4351aa78897@[193.78.237.162]>
Message-ID: <v04210109b4353889cb15@[10.9.8.81]>

At 21:37 +0200 10/21/99, Just van Rossum wrote:
> At 3:02 PM -0400 10/21/99, chris mann wrote:
>>I'm having problems with the smtplib module on Mac.
>>
>>if I run the following code on windows or linux it works fine, but fails on
>>mac:
>>import smtplib
>>s = smtplib.SMTP('smtp.myhost.com')
>>s.helo()
>>
>>i get the following error:
>>Traceback (innermost last):
>>  File "<input>", line 1, in ?
>>  File "Dev:Scripting:Python 1.5.2c1:Lib:smtplib.py", line 290, in helo
>>    name=socket.gethostbyaddr(socket.gethostname())[0]
>>error: host not found
>>
>>any ideas why this is only failing on Mac?
>
> Works for me. Can it be your network setup? Firewall?
> Can you see which of the two socket calls fail?

I have the same problem.

>>> socket.gethostname()
'10.9.8.81'
>>> socket.gethostbyaddr(socket.gethostname())
Traceback (innermost last):
   File "<input>", line 1, in ?
error: host not found

My machine does not have a name known to the outside world because, 
indeed, I am behind an IP masquerading firewall.

So I changed smtplib.ehlo() as follows:

			name = socket.gethostname()
			try:
				name = socket.gethostbyaddr(name)[0]
			except:
				pass

Sorry, I should have posted this fix when I made it.

Doug

-- 
Doug Wyatt                             doug@sonosphere.com
http://www.sonosphere.com/


From jack@oratrix.nl  Thu Oct 21 23:05:12 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Fri, 22 Oct 1999 00:05:12 +0200
Subject: [Pythonmac-SIG] problems with smtplib?
In-Reply-To: Message by chris mann <cmann@pobox.com> ,
 Thu, 21 Oct 1999 15:54:07 -0400 , <B434E79F.D71%cmann@pobox.com>
Message-ID: <19991021220518.DD9FFEA119@oratrix.oratrix.nl>

As Doug explained the problem is that your IP address doesn't have a
reverse mapping in the DNS database. While it is probably a good idea
to submit a bugfix to Guido, as it isn't smtplib's primary task to
play Internet Police, it is something you should get your sysadmin to
fix. There are lots of SMTP servers who will do a lookup/reverse lookup on
the hostname you pass with HELO and treat your messages as spam if the 
reverse lookup fails or doesn't match. There are also lots of other
problems you'll probably experience from your machine (like not being
able to connect to ftp.cwi.nl to get the latest version of
MacPython:-). 
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 


From patrick@swdev.com  Fri Oct 22 08:32:45 1999
From: patrick@swdev.com (Patrick Curtain)
Date: Fri, 22 Oct 1999 00:32:45 -0700
Subject: [Pythonmac-SIG] Curriculum, Course Outlines?
References: <B434917A.64E%jhrsn@pop.pitt.edu>
Message-ID: <381012DF.3879CBC1@swdev.com>

Thanks Jim!  (And the others who wrote!)

Jim Harrison wrote:

> If you haven't already, I'd recommend you take a look at Guido van Rossum's
> proposal "Computer Programming for Everybody"
> <http://www.python.org/doc/essays/cp4e.html>.

I definitely did!  That was a big part of how I drove the decision for using
Python to bring in the fundamental concepts.

> It's possible you could collaborate, or at least get some direction from the
> materials they are developing.

Now that would be ideal.  I didn't really think before sending this to the Mac
list, sorry about that!  Just the list I follow most, myself.  :)

I'll check to see if it's being discussed on the c.l.p list and then post there
as well.

Thanks All!
--p

--
Patrick Curtain, Husband & Father      ( i also write software )
503 781 6199    patrick@swdev.com  http://www.swdev.com/




From jeffrey@Digicool.com  Fri Oct 22 15:31:22 1999
From: jeffrey@Digicool.com (Jeffrey P Shell)
Date: Fri, 22 Oct 1999 10:31:22 -0400
Subject: [Pythonmac-SIG] Zope et al
In-Reply-To: <19991021121636.C679535BB1E@snelboot.oratrix.nl>
Message-ID: <B435ED7A.2C17%jeffrey@Digicool.com>

on 10/21/99 8:16 AM, Jack Jansen at jack@oratrix.nl wrote:

> As far as I know Zope heavily depends on thread support, so for the moment it
> is impossible to implement under MacPython.

Doesn't Thread support in the MacOS continue to improve?  I would think that
it's especially part of Carbon.  What blocks Python threads for MacOS?

As far as Zope goes, we'd love to see an initiative to build it.  I
personally have had a hard enough time getting it to compile on MacOS X
Server, but that is mostly due to my lack of knowledge of the proper way of
compiling Python to do shared libraries in OpenStep.  (BTB, on Stepwise
(www.stepwise.com), they have a binary of Python 1.5.2.  No threads, and I
doubt it does shared library).  Doing it on Mac OS X DP1's unix-side didn't
work at all because Python wouldn't compile the way I wanted it to (none of
the configure scripts ever expected the 'uname' command to return "mac os"
:-)

-- 
Jeffrey P Shell, jeffrey@Digicool.com
http://www.zope.org/



From jimmy@CS.cofc.EDU  Fri Oct 22 16:41:21 1999
From: jimmy@CS.cofc.EDU (James B. Wilkinson)
Date: Fri, 22 Oct 1999 11:41:21 -0400
Subject: [Pythonmac-SIG] Accessing front window in IDE
Message-ID: <199910221541.LAA21424@stono.cs.cofc.edu>



From savageb@pacbell.net  Fri Oct 22 19:31:21 1999
From: savageb@pacbell.net (savageb)
Date: Fri, 22 Oct 1999 11:31:21 -0700
Subject: [Pythonmac-SIG] Zope et al
In-Reply-To: <B435ED7A.2C17%jeffrey@Digicool.com>
Message-ID: <B435FB88.A1F%savageb@pacbell.net>

Jeffrey P Shell <jeffrey@Digicool.com>
> 
> (BTB, on Stepwise
> (www.stepwise.com), they have a binary of Python 1.5.2.  No threads, and I
> doubt it does shared library).

Hmmm...Well, it's a start. :)

I do think the issue of MOSX is looming larger. I understand Jack and Just
have every intention of maintaining Carbon compliance, but what about some
of the low-level modules, e.g. macfs and QD? From what I have read, both
Standard File Services (meaning things like StandardGetFile), and QuickDraw
are going away (to be replaced by "Navigation Services" and "Quartz"
respectivly).

Bob (It's a big scary world out there)



From dillera@isc.upenn.edu  Fri Oct 22 19:56:22 1999
From: dillera@isc.upenn.edu (Andrew Diller)
Date: Fri, 22 Oct 1999 14:56:22 -0400
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <B434FEBC.1448%dillera@isc.upenn.edu>
Message-ID: <B4362B96.1655%dillera@isc.upenn.edu>

Subject: C module creation on MacPython

Hello:

Can I get some pointers on how C modules are created for MacPython?

Specifically, I'm looking to use the sgmlop module from the madscientist
(www.pythonware.com/madscientist) and its only available as c source.

What is needed to create a c module in MacPython?

Any help would be appreciated.





From doug@sonosphere.com  Fri Oct 22 20:31:19 1999
From: doug@sonosphere.com (Doug Wyatt)
Date: Fri, 22 Oct 1999 15:31:19 -0400
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <B4362B96.1655%dillera@isc.upenn.edu>
References: <B4362B96.1655%dillera@isc.upenn.edu>
Message-ID: <v04210100b4366a7d0a43@[10.9.8.81]>

At 14:56 -0400 10/22/99, Andrew Diller wrote:
> Subject: C module creation on MacPython
>
> Hello:
>
> Can I get some pointers on how C modules are created for MacPython?
>
> Specifically, I'm looking to use the sgmlop module from the madscientist
> (www.pythonware.com/madscientist) and its only available as c source.
>
> What is needed to create a c module in MacPython?
>
> Any help would be appreciated.

Brief recipe:

Start with xxmodule.c (in the Modules directory in the source 
distribution).  Modify as needed to define your module's functions 
and data types (refer to the embedding/extending docs).

Build a CodeWarrior shared library project that includes xxmodule.c 
and the PythonCore shared library from your system folder.  Name the 
output shared library "xx.ppc.slb".  Put it somewhere in the module 
search path.

In xxmodule.c, export the function "initxx" (where xx is the module 
name).  This is most simply done by bracketing it with #pragma export.

#pragma export on
extern "C" void		// I'm using C++ so extern "C" is neede
initxx()
{
	...
}
#pragma export off

Hope I didn't forget anything :)

Doug

-- 
Doug Wyatt                             doug@sonosphere.com
http://www.sonosphere.com/


From cwebster@nevada.edu  Fri Oct 22 23:02:02 1999
From: cwebster@nevada.edu (Corran Webster)
Date: Fri, 22 Oct 1999 15:02:02 -0700 (PDT)
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <B4362B96.1655%dillera@isc.upenn.edu>
Message-ID: <Pine.OSF.3.96.991022144226.4106K-100000@pollux.nevada.edu>

> Hello:
> 
> Can I get some pointers on how C modules are created for MacPython?
> 
> Specifically, I'm looking to use the sgmlop module from the madscientist
> (www.pythonware.com/madscientist) and its only available as c source.
> 
> What is needed to create a c module in MacPython?

Doug Wyatt gave some pointers for compiling with CodeWarrior.  If you
don't have CodeWarrior, you can download MPW from Apple's ftp site for
free and compile using that.

You'll need to extract the Include files from the standard mac
distribution (do a custom install from the Installer) or from the source
distribution.

You can then compile and link from the MPW command line, or for more
complex projects, set up a Makefile to build for you.  Typical compilation
commands for a single-file module might look like:

set Python "your path to python folder"

MrC "mymodule.c" -o mymodule.c.x -w off -i "{Python}Mac:Include:" -i
"{Python}Include:"

PPCLink -o mymodule.ppc.slb "mymodule.c.x"  -t 'shlb' -c 'Pyth' -xm s
"{SharedLibraries}InterfaceLib" "{SharedLibraries}MathLib"
"{SharedLibraries}StdCLib" "{PPCLibraries}StdCRuntime.o"
"{PPCLibraries}PPCCRuntime.o" "{PPCLibraries}PPCToolLibs.o"
"{Python}PythonCore" -export initMyModule

(you can break lines with the partial derivative symbol option-d)

You may need to include other libraries, or not include some of the ones
I've included, depedning on what your module does.  I think there should
be an option to turn off certain warnings in the second command, too, but
I'm too lazy to look it up.

Anyway, presuming correct compilation, you will then have a compiled
module that you can place somewhere that Mac Python will find it, and then
you can import it like any other module.

I'm sort of a newbie at this, so I hope I haven't left anything out - I
just hacked around until I got things to work with MPW.

Regards,
Corran




From cwebster@nevada.edu  Fri Oct 22 23:25:16 1999
From: cwebster@nevada.edu (Corran Webster)
Date: Fri, 22 Oct 1999 15:25:16 -0700 (PDT)
Subject: [Pythonmac-SIG] Pyml for Macintosh and related thoughts
Message-ID: <Pine.OSF.3.96.991022150849.4106M-100000@pollux.nevada.edu>

I've been playing with David Konerding's Pyml modules which provide acces
to Mathematica from Python and seem to have it compiling and working on
the mac (Python 1.5.2 with Mathematica 4.0).

Not wanting to duplicate effort, has anyone else done this previously?
And if not, is there interest in making the compiled module available for
others to download (I'd need to get in touch with the author to find the
best way to do this)?

This raises a more general point: should there be a repository of
extension modules compiled for Mac Python.  Some modules aren't going to
be compilable - like some of the Zope stuff which requires threads - but
an archive of those which can be, would be a boon for the community, and
aid in compatibility between platforms. 

I guess ideally, the module authors would maintain mac versions, but
that's probably unrealsitic given that most work in a Unix/Windows world.

Any thoughts?

Regards,
Corran



From jimmy@cs.cofc.edu  Sat Oct 23 04:10:09 1999
From: jimmy@cs.cofc.edu (James B. Wilkinson)
Date: Fri, 22 Oct 1999 23:10:09 -0400
Subject: [Pythonmac-SIG] parameters to Scrap.GetScrap( ??, ??)
Message-ID: <l03020904b436cd2e7321@[153.9.17.27]>

I'm trying to call Scrap.GetScrap, and I'm having trouble giving it
parameters that it likes.  I want the second parameter to be 'TEXT', but I
can't find a way to tell it that. I'm passing None for the first parameter,
thinking that that might give me the result of passing NIL (Inside Mac V
1). It's not complaining about that, so maybe it's ok.

Inside Mac says the second parameter should be ResType, a 4-byte string.
I've tried giving it 'TEXT', and I've tried converting those four bytes to
the equivalent integer. I don't know what to try next. Can somebody give me
a clue?

Here's what I get when I try it:

    length, offset = Scrap.GetScrap(None, 'TEXT')
TypeError: Resource required


Thanks.


-------------------------------------------------------------
Jimmy Wilkinson            | Perfesser of Computer Science
jimmy@cs.CofC.edu          | The College of Charleston
(843) 953-8160             | Charleston      SC        29424

If there is one word to describe me,
that word would have to be "profectionist".




From joe@strout.net  Sat Oct 23 04:52:29 1999
From: joe@strout.net (Joseph J. Strout)
Date: Fri, 22 Oct 1999 20:52:29 -0700
Subject: [Pythonmac-SIG] parameters to Scrap.GetScrap( ??, ??)
In-Reply-To: <l03020904b436cd2e7321@[153.9.17.27]>
Message-ID: <l03102801b436e11d20b2@[198.202.70.213]>

At 8:10 PM -0700 10/22/99, James B. Wilkinson wrote:

>    length, offset = Scrap.GetScrap(None, 'TEXT')
>TypeError: Resource required

I think it's your first resource that's the problem.  It should be a
resource handle -- I don't think Python's GetScrap supports the behavior of
peeking at the scrap without getting it.  So do

	hdl = Res.Resource('')
	length, offset = Scrap.GetScrap(hdl, 'TEXT')

or something like that (typing from memory here).

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'




From steele@cs.brandeis.edu  Sat Oct 23 12:58:36 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Sat, 23 Oct 1999 07:58:36 -0400
Subject: [Pythonmac-SIG] suggestion: use Internet Config for file creation types
Message-ID: <B4371920.9CA%steele@cs.brandeis.edu>

I think MacPython (both the IDE and the interpreter) should use Internet
Config, to compute the type and creator of files that they create.  This
would mean that double-clicking on a *.html file produced by a Python script
would open it with Netscape Navigator or IE, depending on how your IC
database is configured -- this is what I always want from scripts I run, and
I imagine it's desirable for most other people and appropriate for most
scripts as well.  It also means that files created by a Python script on a
foreign system and then copied to the Mac with Anarchie, IE, or PC Exchange,
have the same type and creator as files created by running that same script
on a Mac, which I think is a useful and intuitive invariant.

An alternative would be to change each script that creates a file to set the
type and creator, but (1) this distributes the work everywhere, and (2)
ghettoizes the Mac by (2a) making Mac programmers write scripts that are
unnecessarily OS-dependent, and (2b) keeping scripts from other platforms
from working as well on the Mac (as well as Mac-specific scripts, and as
well as those scripts do on platforms such as Windows, which use the file
extension to encode the type and creator).

I probably won't ever implement this (I swore off C some years ago, and I
hope to stay on the wagon), but I'm hoping that if this strikes anyone else
as a good idea, maybe someone will give a go at implementing it.



From steele@cs.brandeis.edu  Sat Oct 23 12:58:37 1999
From: steele@cs.brandeis.edu (Oliver Steele)
Date: Sat, 23 Oct 1999 07:58:37 -0400
Subject: [Pythonmac-SIG] CommandHistoryPatch update
Message-ID: <B43718E6.9CA%steele@cs.brandeis.edu>

Thanks for your suggestions, everyone!  I've changed the previous history
key to ctl-uparrow (option-p was indeed a holdover from my MCL days); added
ctl-downarrow for next history; and added the feature where typing on a
previous line appends that line to the current input.

Enter/Return just appends the text of the line the selection is on (so press
Return twice to execute it).  Any other non-navigation key appends the text,
repositions the selection, and then types the key.  This sounds complicated,
but I think it works well; try it, and tell me what you think.

The new patch is available from
<http://www.cs.brandeis.edu/~steele/sources/python.html>, as
<http://www.cs.brandeis.edu/~steele/sources/CommandHistoryPatch.py>.
PatchUtils.py does the right thing, so you can just drop
CommandHistoryPatch.py on the interpreter, and it will ask you if you want
to update the old one.

By the way, I extended (by subclassing) PatchUtils.Patch with a ReplaceLines
command, since I needed to replace multiple lines.  I wouldn't mind if this
code were moved into PatchUtils.py.



From joe@strout.net  Sat Oct 23 14:36:05 1999
From: joe@strout.net (Joseph J. Strout)
Date: Sat, 23 Oct 1999 06:36:05 -0700
Subject: [Pythonmac-SIG] CommandHistoryPatch update
In-Reply-To: <B43718E6.9CA%steele@cs.brandeis.edu>
Message-ID: <l03102801b43769d94723@[24.30.159.231]>

At 4:58 AM -0700 10/23/99, Oliver Steele wrote:

>Thanks for your suggestions, everyone!  I've changed the previous history
>key to ctl-uparrow (option-p was indeed a holdover from my MCL days); added
>ctl-downarrow for next history; and added the feature where typing on a
>previous line appends that line to the current input.

Sounds super!  I can't wait to try it!

>By the way, I extended (by subclassing) PatchUtils.Patch with a ReplaceLines
>command, since I needed to replace multiple lines.  I wouldn't mind if this
>code were moved into PatchUtils.py.

Makes perfect sense to me -- I just added patching functions in there on an
"as needed" basis.  I'll roll that code in ASAP.  But I have relatives
visiting this weekend, and shortly after they leave I'm off to Switzerland
for a few days, so it might be a week or two before I get to it.  Just ping
me again if I seem to forget.

Thanks for the cool patch, Oliver!

Cheers,
-- Joe

,------------------------------------------------------------------.
|    Joseph J. Strout           Biocomputing -- The Salk Institute |
|    joe@strout.net             http://www.strout.net              |
`------------------------------------------------------------------'




From jimmy@cs.cofc.edu  Sun Oct 24 04:12:00 1999
From: jimmy@cs.cofc.edu (James B. Wilkinson)
Date: Sat, 23 Oct 1999 23:12:00 -0400
Subject: [Pythonmac-SIG] parameters to Scrap.GetScrap( ??, ??)
In-Reply-To: <l03102801b436e11d20b2@[198.202.70.213]>
References: <l03020904b436cd2e7321@[153.9.17.27]>
Message-ID: <l03020900b437db4a6516@[153.9.17.27]>

Joe Strout opined:

>At 8:10 PM -0700 10/22/99, James B. Wilkinson wrote:
>
>>    length, offset = Scrap.GetScrap(None, 'TEXT')
>>TypeError: Resource required
>
>I think it's your first resource that's the problem.  It should be a
>resource handle -- I don't think Python's GetScrap supports the behavior of
>peeking at the scrap without getting it.  So do
>
>	hdl = Res.Resource('')
>	length, offset = Scrap.GetScrap(hdl, 'TEXT')
>
>or something like that (typing from memory here).

Pretty good memory, I reckon. I just cut that and pasted it in. Bingo!
Well, after I had the presence of mind to import Res. Right proud of that,
I am.

Is there some place I should be able to find stuff like that for myself? I
didn't see anything like that in libmac.pdf, and I haven't found any other
docs for MacPython.


Another question, just in case the answer to the last one is "No":

Here's the end of the traceback I get if I start up the program right after
zeroing the scrap:

    TE.TEFromScrap()
Mac OS Error: (-102, 'No object of that type in scrap')

This looks like the relevant line in macerrors.py:
noTypeErr	=	-102	#No object of that type in scrap

My code at that point is

	try:
		TE.TEFromScrap()
	except macerrors.noTypeErr:
		print "Caught it!"	#For debugging
		pass

If I change the except line to just "except:", then the program prints
"Caught it!" and keeps on running ok.

So I've gotten myself mystified again. I'd much prefer to have the specific
error named in the except line so I'll hear about it if some other error
comes along sometime. Unfortunately, I can't figure out what to call this
error that I want to catch. noTypeErr seems to be the obvious thing, and it
don't work.

Guidance gratefully received.


-------------------------------------------------------------
Jimmy Wilkinson            | Perfesser of Computer Science
jimmy@cs.CofC.edu          | The College of Charleston
(843) 953-8160             | Charleston      SC        29424

If there is one word to describe me,
that word would have to be "profectionist".




From jack@oratrix.nl  Mon Oct 25 11:30:43 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 25 Oct 1999 12:30:43 +0200
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: Message by Corran Webster <cwebster@nevada.edu> ,
 Fri, 22 Oct 1999 15:02:02 -0700 (PDT) , <Pine.OSF.3.96.991022144226.4106K-100000@pollux.nevada.edu>
Message-ID: <19991025103044.F358335BB1E@snelboot.oratrix.nl>

> > What is needed to create a c module in MacPython?
> 
> Doug Wyatt gave some pointers for compiling with CodeWarrior.  If you
> don't have CodeWarrior, you can download MPW from Apple's ftp site for
> free and compile using that.

Corran,
I'm very surprised that your recipe works, as it mixes MPW libraries (in the 
extension module) with CodeWarrior libraries (in PythonCore). I would expect 
all sort of trouble with I/O, malloc/free, and possibly loads of other places.

If you're actually using this method, could you post a bit of information on 
what works, what doesn't work and whether there are specific things you should 
do (like calling printf, I can imagine)?
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct 25 12:52:19 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 25 Oct 1999 13:52:19 +0200
Subject: [Pythonmac-SIG] suggestion: use Internet Config for file creation
 types
In-Reply-To: Message by Oliver Steele <steele@cs.brandeis.edu> ,
 Sat, 23 Oct 1999 07:58:36 -0400 , <B4371920.9CA%steele@cs.brandeis.edu>
Message-ID: <19991025115219.E778735BB1E@snelboot.oratrix.nl>

> I think MacPython (both the IDE and the interpreter) should use Internet
> Config, to compute the type and creator of files that they create.

I thought about this at some point in the past, but decided against it. It 
tightly couples extensions to filetypes, as is done on Windows, and I think 
this is a bad thing in general.

What I could live with is if you did this optionally. I.e. if you do
    import autotyper
    autotyper.install()
somewhere in your script you get this behaviour. If you want to implement this 
I'll gladly accept it and add it to the distribution.

Note, by the way, that you don't need to write any C for this: there is in 
Internet config interface available, and you can hook into the standard open 
call with something like
import __builtin__

_builtin_open = __builtin__.open

def _open_with_typer(*args):
	rv = apply(_builtin_open, args)
	# do typer magic
	return rv

def install():
	__builtin__.open = _open_with_typer

--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From jack@oratrix.nl  Mon Oct 25 12:56:01 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Mon, 25 Oct 1999 13:56:01 +0200
Subject: [Pythonmac-SIG] parameters to Scrap.GetScrap( ??, ??)
In-Reply-To: Message by "James B. Wilkinson" <jimmy@cs.cofc.edu> ,
 Sat, 23 Oct 1999 23:12:00 -0400 , <l03020900b437db4a6516@[153.9.17.27]>
Message-ID: <19991025115602.2091335BB1E@snelboot.oratrix.nl>

> Is there some place I should be able to find stuff like that for myself? I
> didn't see anything like that in libmac.pdf, and I haven't found any other
> docs for MacPython.

If someone wants to pick up working on MacPython documentation: please, be my 
guest. I know the docs are very incomplete and outdated, but I don't have time 
to work on it....

> Here's the end of the traceback I get if I start up the program right after
> zeroing the scrap:
> 
>     TE.TEFromScrap()
> Mac OS Error: (-102, 'No object of that type in scrap')
> 
> This looks like the relevant line in macerrors.py:
> noTypeErr	=	-102	#No object of that type in scrap
> 
> My code at that point is
> 
> 	try:
> 		TE.TEFromScrap()
> 	except macerrors.noTypeErr:
> 		print "Caught it!"	#For debugging
> 		pass

The values in macerrors.py are the values you will be passed in the *argument* 
to the exception. The exception you get for all (or, at least, almost all) 
toolbox errors is MacOS.Error. This exception is also known as TE.Error, 
Res.Error, etc etc etc.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From dillera@isc.upenn.edu  Mon Oct 25 15:48:38 1999
From: dillera@isc.upenn.edu (Andrew Diller)
Date: Mon, 25 Oct 1999 10:48:38 -0400 (EDT)
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <19991025103044.F358335BB1E@snelboot.oratrix.nl>
Message-ID: <Pine.OSF.4.10.9910251047090.29521-100000@falstaff.isc-net.upenn.edu>

Wow- I'm hoping that I can get MPW to work. We only have one copy of
Codewarrior here, and I was hoping to be able to use MPW to do some
things.

I was able to get sgmlop compiled and linked w/ codewarrior, now I'm going
to download MPW and give it a go.

I'll let the list know what happens.

thanks for the input.

-andy


- - - - - - - - - - - - - - - - - - - 
Andrew Diller   dillera@isc.upenn.edu
Information Systems and Computing
University of Pennsylvania
v: 3-7360 f: 8-9348 

On Mon, 25 Oct 1999, Jack Jansen wrote:

> 
> > > What is needed to create a c module in MacPython?
> > 
> > Doug Wyatt gave some pointers for compiling with CodeWarrior.  If you
> > don't have CodeWarrior, you can download MPW from Apple's ftp site for
> > free and compile using that.
> 
> Corran,
> I'm very surprised that your recipe works, as it mixes MPW libraries (in the 
> extension module) with CodeWarrior libraries (in PythonCore). I would expect 
> all sort of trouble with I/O, malloc/free, and possibly loads of other places.
> 
> If you're actually using this method, could you post a bit of information on 
> what works, what doesn't work and whether there are specific things you should 
> do (like calling printf, I can imagine)?
> --
> Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
> Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
> www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 
> 
> 



From maccgi@bellsouth.net  Mon Oct 25 15:58:30 1999
From: maccgi@bellsouth.net (Richard Gordon)
Date: Mon, 25 Oct 1999 10:58:30 -0400
Subject: [Pythonmac-SIG] Gadfly SQL on Mac?
Message-ID: <v04205500b43a16ca1143@[209.214.68.92]>

I just ran across Aaron Watters' Gadfly 
(http://www.chordate.com/kwParsing/gadfly.html) while looking for 
something else and am interested in playing with it, but thought I'd 
ask if anyone happens to know if this runs ok on Mac or not? Also, 
it's the only SQL engine that I've come across that seems to be 
entirely ram-based and while that should make it amazingly fast, I 
wonder how practical it would be to use unless you had hordes of 
memory to allocate to MacPython? Thanks.

Richard Gordon
--------------------
Gordon Consulting & Design
Database Design/Scripting Languages
mailto:richard@richardgordon.net
http://www.richardgordon.net
770.971.6887 (voice)
770.216.1829 (fax)


From savageb@pacbell.net  Mon Oct 25 20:55:42 1999
From: savageb@pacbell.net (Bob Savage)
Date: Mon, 25 Oct 1999 12:55:42 -0700
Subject: [Pythonmac-SIG] Gadfly SQL on Mac?
In-Reply-To: <v04205500b43a16ca1143@[209.214.68.92]>
Message-ID: <B43A03CD.118%savageb@pacbell.net>

Richard Gordon <maccgi@bellsouth.net>
> I just ran across Aaron Watters' Gadfly
> (http://www.chordate.com/kwParsing/gadfly.html) while looking for
> something else and am interested in playing with it, but thought I'd
> ask if anyone happens to know if this runs ok on Mac or not? Also,
> it's the only SQL engine that I've come across that seems to be
> entirely ram-based and while that should make it amazingly fast, I
> wonder how practical it would be to use unless you had hordes of
> memory to allocate to MacPython? Thanks.

Richard,

I did some work with Gadfly on my Mac this past year. I found it to run
perfectly fine with one caveat: eventually the in-RAM model came back to
bite me. I had four tables, each with about 35,000 rows which gave me an
out-of-memory error. I might have been able to tweak it so that it still
worked by dividing the databse up into four databases (one for each table)
but that would have limited my ability to fashion Queries on my data. Still
haven't come up with a good alternative. Also Aaron gave me an experimental
patch to use a B-tree lookup table, but, again, that really restricted my
ability to do queries (it worked great for "where somerow is 'something'"
type searches but not on inexact matches like greater than, or max, etc.)

I would say you should consider how much information you are going to throw
at it. If the number of total rows in the DB are going to approach (or
exceed) 140,000, you will probably need a different solution. If you can
structure your data so that it never gets near that amount, I would say
Gadfly is a pleasure to use.

Bob Savage



From cwebster@nevada.edu  Tue Oct 26 00:02:54 1999
From: cwebster@nevada.edu (Corran Webster)
Date: Mon, 25 Oct 1999 16:02:54 -0700 (PDT)
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <19991025103044.F358335BB1E@snelboot.oratrix.nl>
Message-ID: <Pine.OSF.3.96.991025144635.27550A-100000@pollux.nevada.edu>

> > > What is needed to create a c module in MacPython?
> > 
> > Doug Wyatt gave some pointers for compiling with CodeWarrior.  If you
> > don't have CodeWarrior, you can download MPW from Apple's ftp site for
> > free and compile using that.

FWIW, in my original post, I forgot to mention that you need to change the
:Mac:Include:config.h file to add support for forward static declarations 
if you are using MrC in MPW:

#ifdef __MRC__
#define BAD_STATIC_FORWARD
#endif

needs to be added at the appropriate point.

> Corran,
> I'm very surprised that your recipe works, as it mixes MPW libraries (in the 
> extension module) with CodeWarrior libraries (in PythonCore). I would expect 
> all sort of trouble with I/O, malloc/free, and possibly loads of other places.

I hadn't considered this possibility - and yes, there ar warnings about
symbols appearing in PythonCore plus some other library, but I had been
ignoring them for the large part, assuming that MPW would choose sensibly
between possible options.  As I said, I basically hacked around until I
got stuff to compile and run. 

I haven't really pushed these compiled modules to the limits, so I
wouldn't have noticed slow memory leaks or the like which would result
from malloc/free not working properly.

> If you're actually using this method, could you post a bit of information on 
> what works, what doesn't work and whether there are specific things you should 
> do (like calling printf, I can imagine)?

The main things I've tried have been pretty much pure C modules like
Marc-Andre Lemburg's various mx modules, so I/O and so forth haven't been 
major issues.

I just did a quick check with printf, and it appears that the order that
MPW links is important - if PythonCore is linked before StdCLib, then
printf appears to work fine; if linked in the other order then it simply
does nothing - no error, nothing printed to standard output, etc. which
makes sense because MPW StdCLib wouldn't know about SIOUX and would be
trying to use SIOW or the like, although it is interesting that it didn't
cause a crash.

Testing in the IDE, nothing appears either way - whether it was run in the
interactive window or as a script. 

StdCLib only appears to be required for linking by MPW if you are using
input/output in your module, so most of the modules I worked with weren't
using it at all, and so were using Python's versions of things like malloc
and free. 

I'm going to have to look at this a bit more closely, but my guess is that
you probably want PythonCore linked first so that you're using the same
versions of the basic C libraries.

The only other multiply defined symbols between PythonCore and the
standard MPW libraries are the cursor utilities, and only Mac-specific
modules would be using that.  So I suspect that most other modules will
compile and run fine as long as they aren't relying on OS specific
features.

The long and the short of it is that you seem to be able to get basic
stuff to work fine in MPW, but there may be subtle bugs in extreme
situations; stuff which requires I/O seems trickier, but with the right
linking order can probably be made to work.

I hope all this makes sense. I'd be more than interested in what more
experienced people have to say - my Macintosh coding experience is
limited.

Regards,
Corran



From jack@oratrix.nl  Tue Oct 26 10:34:46 1999
From: jack@oratrix.nl (Jack Jansen)
Date: Tue, 26 Oct 1999 11:34:46 +0200
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: Message by Corran Webster <cwebster@nevada.edu> ,
 Mon, 25 Oct 1999 16:02:54 -0700 (PDT) , <Pine.OSF.3.96.991025144635.27550A-100000@pollux.nevada.edu>
Message-ID: <19991026093446.7FDC535BB1E@snelboot.oratrix.nl>

Wow, I'm really impressed that building extensions with MrC seems to work so 
well!

What I would really like is to extend the plugins.html document to include 
instructions for building plugins using MrC, and having a Makefile for 
xxmodule.

The only point is that plugins.html uses the long-useless and untestable 
InterSLIP as its example. If someone could take the plugins documentation and 
rewrite it around an example interface that actually works on all Macintoshes 
that would be very helpful. Something like (parts of) the Shutdown Manager, or 
tGetDateTime or some such comes to mind.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen@oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 




From doug@sonosphere.com  Wed Oct 27 11:07:18 1999
From: doug@sonosphere.com (Doug Wyatt)
Date: Wed, 27 Oct 1999 06:07:18 -0400
Subject: [Pythonmac-SIG] macglue.h
In-Reply-To: <19991026093446.7FDC535BB1E@snelboot.oratrix.nl>
References: <19991026093446.7FDC535BB1E@snelboot.oratrix.nl>
Message-ID: <v04210100b43c7d88e04a@[10.9.8.81]>

I'm embedding Python in a Mac app and had to make a small tweak.

macglue.h isn't C++ aware, so I get link errors in my C++ app when 
trying to link to the C functions declared in the header.  Of course

#ifdef __cplusplus
extern "C" {
#endif

around the function prototypes solves everything.

Doug
who earlier spent 90 minutes learning the hard way that Py_None must 
be INCREF'd before returning it from a C function :)

-- 
Doug Wyatt                             doug@sonosphere.com
http://www.sonosphere.com/


From cwebster@nevada.edu  Wed Oct 27 19:39:13 1999
From: cwebster@nevada.edu (Corran Webster)
Date: Wed, 27 Oct 1999 11:39:13 -0700 (PDT)
Subject: [Pythonmac-SIG] C module creation on MacPython
In-Reply-To: <19991026093446.7FDC535BB1E@snelboot.oratrix.nl>
Message-ID: <Pine.OSF.3.96.991027112042.16167A-100000@pollux.nevada.edu>

> Wow, I'm really impressed that building extensions with MrC seems to work so 
> well!
> 
> What I would really like is to extend the plugins.html document to include 
> instructions for building plugins using MrC, and having a Makefile for 
> xxmodule.

Since there seems to be interest in this, I've written a HOWTO.  It's
currently available at:

http://www.nevada.edu/~cwebster/Python/MPWHOWTO/MPWHOWTO.html

with associated files (including a makefile for xxmodule) in the same
directory. You can download the whole lot as: 

http://www.nevada.edu/~cwebster/Python/MPWHOWTO/MPWHOWTO.sit.hqx

I invite comments and corrections; this is a first draft.  In particular,
I'm going to need to investigate further how GUSI connects in (I have not
been using it, but it has the potential to make porting from Unix
significantly easier if it can be linked in correctly).

I'm also likely to fiddle with the xxmodule make file - currently it's
just what CreateMake spits out, modified as described in the HOWTO; but it
would make more sense to re-write it by hand to illustrate how one might
write one from scratch.

If this, or a modified version, is appropriate to add to the documentation
for MacPython, I'd have no problem with that (in fact it'd be sorta cool
to contibute).

Regards,
Corran




From cbarker@jps.net  Thu Oct 28 21:10:50 1999
From: cbarker@jps.net (Chris Barker)
Date: Thu, 28 Oct 1999 13:10:50 -0700
Subject: [Pythonmac-SIG] Creator of text files created by Python scripts
References: <199907220503.BAA21467@python.org>
Message-ID: <3818ADCA.345FC440@jps.net>

Hi,

I've found that when I create a new text file in a script that I'm
running from the IDE, it gets created with creator type "PYTH". If I run
the same script by dropping it on the Python interpreter, it gets
created with creator "ttxt", which is not ideal, but is a better option.

having various text files around that are not python scripts, but have
creator type "PYTH" is a pain. Can I change this behavioe in the IDE?

-Thanks,
-Chris




-- 
Christopher Barker,
Ph.D.                                                           
cbarker@jps.net                      ---           ---           ---
http://www.jps.net/cbarker          -----@@       -----@@       -----@@
                                   ------@@@     ------@@@     ------@@@
Water Resources Engineering       ------   @    ------   @   ------   @
Coastal and Fluvial Hydrodynamics -------      ---------     --------    
------------------------------------------------------------------------
------------------------------------------------------------------------


From savageb@pacbell.net  Fri Oct 29 06:38:06 1999
From: savageb@pacbell.net (savageb)
Date: Thu, 28 Oct 1999 22:38:06 -0700
Subject: [Pythonmac-SIG] Creator of text files created by Python scripts
In-Reply-To: <3818ADCA.345FC440@jps.net>
Message-ID: <B43E80CE.D65%savageb@pacbell.net>

Chris, you can do a few things.

For one, if you are using the Interpretor and you prefer something other
than 'ttxt' you can change that by using the edit python prefs app (click
the "default startup options" button to find it).

If your python script is usable as a "double-clickable" script (or perhaps
something that you trigger through drag&drop) then you can actually drop the
script applet onto the edit python prefs and that should give you the
ability to change the creator and type codes for just that instance of the
interpretor - you can use this method to change the IDE as well.

Finally if you didn't want to screw up the default settings (like if you
want your scripts to still open in the IDE by default) you can set the
creator and type codes as part of the script using the macfs module (you can
look at mac/scripts/fixfiletypes.py as an example).

Good luck with it,

Bob

Chris Barker <cbarker@jps.net>
> I've found that when I create a new text file in a script that I'm
> running from the IDE, it gets created with creator type "PYTH". If I run
> the same script by dropping it on the Python interpreter, it gets
> created with creator "ttxt", which is not ideal, but is a better option.
> 
> having various text files around that are not python scripts, but have
> creator type "PYTH" is a pain. Can I change this behavioe in the IDE?