pyobjc-dev — CVS commit notices go here

[Pyobjc-dev] Re: NSAnyEventMask as unsigned?

[Bob I. <bo...@re...>]

On Oct 29, 2003, at 12:51 PM, Ronald Oussoren wrote:

> On 29 okt 2003, at 18:39, Bob Ippolito wrote:
>
>> On Oct 29, 2003, at 12:28 PM, Ronald Oussoren wrote:
>>
>>> Why is it a problem that NSAnyEventMask is -1? It is also -1 in ObjC 
>>> and therefore I suppose the conversion from Python to ObjC for 
>>> ``unsigned int`` values is too strict and doesn't accept -1 as a 
>>> valid value.
>>
>> It is 0xffffffffU in ObjC -- the fact that it has the same binary 
>> representation as the signed -1 is just a side effect.  The 
>> conversion from Python to ObjC is indeed too strict for this kind of 
>> conversion, but rightly so, if you agree with the direction that 
>> Python itself is moving.  For example, in Python 2.4, 0xffffffff 
>> without the L qualifier will become 0xffffffffL, instead of the 
>> current -1.
>
> You read the header file :-). The 'U' is bogus, it is inside an enum 
> definition and enums are signed integers. The following program should 
> make this clear:
>
> #include <stdio.h>
> #import <AppKit/AppKit.h>
>
> int main(void)
> {
>         printf("%s\n", @encode(__typeof__(NSAnyEventMask)));
> 		return 0;
> }
>
> This prints 'i', which is the type-code for 'int'.
>
> I don't like the hand-coded list, and therefore tried to be smart by 
> adding the type encoding for constants to the inttable structure which 
> didn't work because of the signedness of enums :-(. I'll update the 
> CodeGenerator scripts to detect '*Mask' as unsigned integers instead 
> of plain integers.

Ah yes, you are of course correct.  I'm not (yet) very familiar with 
the ObjC facilities for typing.  However, the real problem is that 
Python really sucks at integers (from a bit twiddling / talking to C 
perspective).. it's handling of them diverges more and more from the C 
way of doing it after every release.  In these cases, what you have is 
not even an "integer" in the intuitive sense, it's an integer length 
bit vector.. it's C's fault that it can't express this otherwise (not 
that Python has a built-in type that's any better at this).

In any case, because this is Python, any sort of bit flag or mask 
really needs to treated as if it were unsigned or else bad things start 
happening and warnings get thrown all over the place (or probably break 
in 2.4).

It would be awesome if you could write up some quick documentation as 
to how to run the code generators and how to make them respect manual 
overrides (for signatures, at least).  I took a quick look at them, but 
I didn't quite get it.

-bob

[Pyobjc-dev] Re: NSAnyEventMask as unsigned?

[Ronald O. <ous...@ci...>]

On 29 okt 2003, at 18:39, Bob Ippolito wrote:

> On Oct 29, 2003, at 12:28 PM, Ronald Oussoren wrote:
>
>> Why is it a problem that NSAnyEventMask is -1? It is also -1 in ObjC 
>> and therefore I suppose the conversion from Python to ObjC for 
>> ``unsigned int`` values is too strict and doesn't accept -1 as a 
>> valid value.
>
> It is 0xffffffffU in ObjC -- the fact that it has the same binary 
> representation as the signed -1 is just a side effect.  The conversion 
> from Python to ObjC is indeed too strict for this kind of conversion, 
> but rightly so, if you agree with the direction that Python itself is 
> moving.  For example, in Python 2.4, 0xffffffff without the L 
> qualifier will become 0xffffffffL, instead of the current -1.

You read the header file :-). The 'U' is bogus, it is inside an enum 
definition and enums are signed integers. The following program should 
make this clear:

#include <stdio.h>
#import <AppKit/AppKit.h>

int main(void)
{
         printf("%s\n", @encode(__typeof__(NSAnyEventMask)));
		return 0;
}

This prints 'i', which is the type-code for 'int'.

I don't like the hand-coded list, and therefore tried to be smart by 
adding the type encoding for constants to the inttable structure which 
didn't work because of the signedness of enums :-(. I'll update the 
CodeGenerator scripts to detect '*Mask' as unsigned integers instead of 
plain integers.

Ronald

[Pyobjc-dev] [ pyobjc-Bugs-832177 ] Reference leak in Modules/objc/objc-class.m

[SourceForge.net <no...@so...>]

Bugs item #832177, was opened at 2003-10-29 00:16
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832177&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: Reference leak in Modules/objc/objc-class.m

Initial Comment:
The PySequence_Fast is created but not decref'ed.. see patch 
below:

Index: Modules/objc/objc-class.m
=====================================
==============================
RCS file: /cvsroot/pyobjc/pyobjc/Modules/objc/objc-class.m,v
retrieving revision 1.44
diff -u -r1.44 objc-class.m
--- Modules/objc/objc-class.m   19 Oct 2003 16:53:35 -0000      
1.44
+++ Modules/objc/objc-class.m   29 Oct 2003 05:15:03 
-0000
@@ -251,6 +251,7 @@
                                PySequence_Fast_GET_ITEM(seq, i));
                        
Py_INCREF(PySequence_Fast_GET_ITEM(seq, i));
                }
+        Py_DECREF(seq);
        }
 
        real_bases = PyList_New(0);

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832177&group_id=14534

Re: [Pyobjc-dev] Building apps for 10.2+ without Xcode?

[Bob I. <bo...@re...>]

On Oct 28, 2003, at 10:32 PM, Zachery Bir wrote:

> Can we use bundlebuilder's buildapp to build 10.2 capable apps in 
> Panther? I accidentally initially released a version of my PyObjC app 
> built on a pre-release of Panther, and had to scramble to build it on 
> a Jaguar box. I'd like to be able to build apps (without resorting to 
> Xcode) that will run on both Jaguar and Panther using just buildapp. I 
> have both the 10.2 SDK and 10.3 SDK installed from the Xcode CD. Will 
> that be enough?

I'm not the expert on this, but I think that you need to install the 
Jaguar version of MacPython, and include the Jaguar version of PyObjC.. 
both of these run just fine on Panther.

-bob

[Pyobjc-dev] Building apps for 10.2+ without Xcode?

[Zachery B. <zb...@ur...>]

Can we use bundlebuilder's buildapp to build 10.2 capable apps in 
Panther? I accidentally initially released a version of my PyObjC app 
built on a pre-release of Panther, and had to scramble to build it on a 
Jaguar box. I'd like to be able to build apps (without resorting to 
Xcode) that will run on both Jaguar and Panther using just buildapp. I 
have both the 10.2 SDK and 10.3 SDK installed from the Xcode CD. Will 
that be enough?

Zac

[Pyobjc-dev] [ pyobjc-Bugs-832142 ] Mixing spaces and tabs for indentation is bad!

[SourceForge.net <no...@so...>]

Bugs item #832142, was opened at 2003-10-28 22:24
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832142&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: Mixing spaces and tabs for indentation is bad!

Initial Comment:
I'm not sure who is doing this.. but it's all over the PyObjC 
source tree.. please make sure not to use tabs, unless you 
ONLY use tabs.

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832142&group_id=14534

[Pyobjc-dev] [ pyobjc-Bugs-832135 ] Incorrect signature for NSTextViewDelegate message

[SourceForge.net <no...@so...>]

Bugs item #832135, was opened at 2003-10-28 22:10
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832135&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: Incorrect signature for NSTextViewDelegate message

Initial Comment:
Below is the correct signature.  Note that this part of the 
protocol is only available in 10.3 and later.

selector(None, selector='textView:completions:
forPartialWordRange:indexOfSelectedItem:',signature='@@:
@@{_NSRange=II}N^i')

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832135&group_id=14534

[Pyobjc-dev] [ pyobjc-Bugs-832025 ] more pydoc problems

[SourceForge.net <no...@so...>]

Bugs item #832025, was opened at 2003-10-28 17:58
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832025&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: more pydoc problems

Initial Comment:
[crack:~] bob% pythonw
Python 2.3 (#1, Sep 13 2003, 00:49:11) 
[GCC 3.3 20030304 (Apple Computer, Inc. build 1495)] on 
darwin
Type "help", "copyright", "credits" or "license" for more 
information.
>>> import objc  
>>> objc.__version__
'1.0'
>>> help()

Welcome to Python 2.3!  This is the online help utility.

If this is your first time using Python, you should definitely 
check out
the tutorial on the Internet at http://www.python.org/doc/
tut/.

Enter the name of any module, keyword, or topic to get help 
on writing
Python programs and using Python modules.  To quit this 
help utility and
return to the interpreter, just type "quit".

To get a list of available modules, keywords, or topics, type 
"modules",
"keywords", or "topics".  Each module also comes with a 
one-line summary
of what it does; to list the modules whose summaries 
contain a given word
such as "spam", type "modules spam".

help> modules spam

Here is a list of matching modules.  Enter any module name 
to get more help.

objc: cannot unmap an image containing ObjC data
Trace/BPT trap

Here's a traceback
Program received signal SIGTRAP, Trace/breakpoint trap.
0x9083bec0 in _objc_trap ()
(gdb) bt
#0  0x9083bec0 in _objc_trap ()
#1  0x9083be54 in _objc_fatal ()
#2  0x90837c84 in _objc_fatalHeader ()
#3  0x8fe18748 in __dyld_call_funcs_for_remove_image ()
#4  0x8fe136a8 in __dyld__dyld_unlink_module ()
#5  0x9003d7c8 in NSUnLinkModule ()
#6  0x95fcc560 in _PyImport_GetDynLoadFunc ()
#7  0x95fc1b6c in _PyImport_LoadDynamicModule ()
#8  0x95fbf7a4 in PyImport_ExecCodeModuleEx ()
#9  0x95fc16f0 in PyImport_Import ()
#10 0x95fa94a8 in PyEval_GetFuncDesc ()
#11 0x95fa6c64 in PyEval_EvalCode ()
#12 0x95fa7e30 in PyEval_EvalCodeEx ()
#13 0x95fa97dc in PyEval_GetFuncDesc ()
#14 0x95fa9580 in PyEval_GetFuncDesc ()
#15 0x95fa6c64 in PyEval_EvalCode ()
#16 0x95fa7e30 in PyEval_EvalCodeEx ()
#17 0x95fa97dc in PyEval_GetFuncDesc ()
#18 0x95fa9580 in PyEval_GetFuncDesc ()
#19 0x95fa6c64 in PyEval_EvalCode ()
#20 0x95fa9728 in PyEval_GetFuncDesc ()
#21 0x95fa9580 in PyEval_GetFuncDesc ()
#22 0x95fa6c64 in PyEval_EvalCode ()
#23 0x95fa7e30 in PyEval_EvalCodeEx ()
#24 0x95fa97dc in PyEval_GetFuncDesc ()
#25 0x95fa9580 in PyEval_GetFuncDesc ()
#26 0x95fa6c64 in PyEval_EvalCode ()
#27 0x95fa9728 in PyEval_GetFuncDesc ()
#28 0x95fa9580 in PyEval_GetFuncDesc ()
#29 0x95fa6c64 in PyEval_EvalCode ()
#30 0x95fa9728 in PyEval_GetFuncDesc ()
#31 0x95fa9580 in PyEval_GetFuncDesc ()
#32 0x95fa6c64 in PyEval_EvalCode ()
#33 0x95fa7e30 in PyEval_EvalCodeEx ()
#34 0x95f5f354 in PyFunction_SetClosure ()
#35 0x95f4a8d0 in PyObject_Call ()
#36 0x95f529e8 in PyMethod_New ()
#37 0x95f4a8d0 in PyObject_Call ()
#38 0x95f51da8 in _PyInstance_Lookup ()
#39 0x95f4a8d0 in PyObject_Call ()
#40 0x95fa9df0 in PyEval_GetFuncDesc ()
#41 0x95fa6d44 in PyEval_EvalCode ()
#42 0x95fa7e30 in PyEval_EvalCodeEx ()
#43 0x95f5f354 in PyFunction_SetClosure ()
#44 0x95f4a8d0 in PyObject_Call ()
#45 0x95f529e8 in PyMethod_New ()
#46 0x95f4a8d0 in PyObject_Call ()
#47 0x95f51da8 in _PyInstance_Lookup ()
#48 0x95f4a8d0 in PyObject_Call ()
#49 0x95fa9ba8 in PyEval_GetFuncDesc ()
#50 0x95fa9598 in PyEval_GetFuncDesc ()
#51 0x95fa6c64 in PyEval_EvalCode ()
#52 0x95fa7e30 in PyEval_EvalCodeEx ()
#53 0x95fa4734 in PyEval_EvalCode ()
#54 0x95fc85f0 in PyRun_FileExFlags ()
#55 0x95fc7354 in PyRun_InteractiveOneFlags ()
#56 0x95fc713c in PyRun_InteractiveLoopFlags ()
#57 0x95fc6fd4 in PyRun_AnyFileExFlags ()
#58 0x95fd1ec0 in Py_Main ()
#59 0x00003c78 in ?? ()
#60 0x00003aec in ?? ()

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=832025&group_id=14534

[Pyobjc-dev] ANN: iChat channel for MacPython and PyObjC

[Just v. R. <ju...@le...>]

There's a iChat channel called "MacPython" now. Hit cmd-G in iChat and
type the channel name. See you there!

Just

[Pyobjc-dev] [ pyobjc-Bugs-831929 ] Some unsigned int constants wrapped as signed integers

[SourceForge.net <no...@so...>]

Bugs item #831929, was opened at 2003-10-28 15:51
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=831929&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: Some unsigned int constants wrapped as signed integers

Initial Comment:
For example,
>>> from AppKit import * 
>>> NSAnyEventMask
-1

This should be 0xFFFFFFFFL

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=831929&group_id=14534

[Pyobjc-dev] [ pyobjc-Bugs-831774 ] NSDecimalNumber instances handled incorrectly

[SourceForge.net <no...@so...>]

Bugs item #831774, was opened at 2003-10-28 07:36
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=831774&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
Summary: NSDecimalNumber instances handled incorrectly

Initial Comment:
NSDecimalNumbers are not handled correctly going back to 
ObjC.

That is, while NSDecimalNumbers are correctly converted to 
python doubles, when going back to Objective-C, they are 
converted to NSNumbers instead. The problem with this is 
that NSNumber doesn't implement things like:

  - (NSDecimalNumber *) decimalNumberByMultiplyingBy: 
(NSDecimalNumber *);

  pi...@ma...

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=831774&group_id=14534

Re: [Pyobjc-dev] ANN: PyInterpreter 0.2

[Bob I. <bo...@re...>]

On Oct 28, 2003, at 10:10 AM, Ronald Oussoren wrote:

> On 28 okt 2003, at 8:11, Bob Ippolito wrote:
>
>> PyInterpreter is a set of PyObjC classes (and a nib) that makes it 
>> easy to put an interactive Python interpreter into just about any 
>> PyObjC application.  It supports the following features:
>> 	Colored stdout / stderr
>> 	raw_input() / stdin.readline() support via modal dialog (a la 
>> PythonIDE)
>> 	command history
>> 	completion (OS X 10.3 and later only)
>>
>> Feel free to use this in your own projects, I'm hoping to get this 
>> into the PyObjC Examples tree sometime in the near future.
>
> I like this, except the raw_input part. I "accidently" typed help() in 
> the interpreter, which mostly showed that the modal dialog is not a 
> good idea. A sheet would have been even worse :-).

It's the only way to do input.. unless you know some way to do 
stackless-like-things without stackless.. I don't know how to wrap the 
inner-interpreter with something that allows me to provide 
sys.stdin.readline() info asynchronously.  I sure can't do it with 
threads, unless the inner-interpreter always runs in a thread, and that 
will make everything so much more complicated and harder to test.

In any case, the readline support can be turned off pretty easily if 
you look at the source.. it's just one line that you'd change to None.

> W.r.t. getting this into the PyObjC examples: it might be more usefull 
> to add this, and more IDE-components (like a class browser), to 
> PyObjCTools as this would make it even easier to reuse these 
> components.

Probably, but you have the nib issue (bundlebuilder won't like that).

-bob

Re: [Pyobjc-dev] ANN: PyInterpreter 0.2

[Ronald O. <ous...@ci...>]

On 28 okt 2003, at 8:11, Bob Ippolito wrote:

> PyInterpreter is a set of PyObjC classes (and a nib) that makes it 
> easy to put an interactive Python interpreter into just about any 
> PyObjC application.  It supports the following features:
> 	Colored stdout / stderr
> 	raw_input() / stdin.readline() support via modal dialog (a la 
> PythonIDE)
> 	command history
> 	completion (OS X 10.3 and later only)
>
> Feel free to use this in your own projects, I'm hoping to get this 
> into the PyObjC Examples tree sometime in the near future.

I like this, except the raw_input part. I "accidently" typed help() in 
the interpreter, which mostly showed that the modal dialog is not a 
good idea. A sheet would have been even worse :-).

W.r.t. getting this into the PyObjC examples: it might be more usefull 
to add this, and more IDE-components (like a class browser), to 
PyObjCTools as this would make it even easier to reuse these 
components.

Ronald

[Pyobjc-dev] ANN: PyInterpreter 0.2

[Bob I. <bo...@re...>]

PyInterpreter is a set of PyObjC classes (and a nib) that makes it easy 
to put an interactive Python interpreter into just about any PyObjC 
application.  It supports the following features:
	Colored stdout / stderr
	raw_input() / stdin.readline() support via modal dialog (a la 
PythonIDE)
	command history
	completion (OS X 10.3 and later only)

Feel free to use this in your own projects, I'm hoping to get this into 
the PyObjC Examples tree sometime in the near future.

See http://pythonmac.org/wiki/PyInterpreter for a screen shot, 
download, and more information.

-bob

Re: [Pyobjc-dev] Interactive interpreter in PyObjC

[Bob I. <bo...@re...>]

On Oct 27, 2003, at 8:47 AM, Bob Ippolito wrote:

> I started working on an interactive in-process interpreter for PyObjC 
> applications.  Here's what it does so far:
> 	Simple coloring of sys.stdout and sys.stderr
> 	Supports unicode (but assumes non-unicode is utf8 for now)
> 	Command history (up key, down key)
> 	Limits editing and cursor position (but not selection) to ONLY the 
> active line of code
> 	All the nice things that editing in a NSTextView gives you
>
> Obvious future enhancements would be:
> 	Auto indent
> 	Syntax highlighting
> 	Completion
> 	sys.stdin (for raw_input and such -- might be too hard to do without 
> stackless, threads, or other nasty stuff)

Ok I fixed some nasty bugs (thanks Just), use this one now:  
http://undefined.org/python/PyInterpreter-0.1.tgz

I also added sys.stdin.readline support (which gives you raw_input as 
well) in a similar way to how the MacPython IDE does it: A modal 
NSPanel input dialog (yay).  It's ugly but it works.

-bob

[Pyobjc-dev] Interactive interpreter in PyObjC

[Bob I. <bo...@re...>]

I started working on an interactive in-process interpreter for PyObjC 
applications.  Here's what it does so far:
	Simple coloring of sys.stdout and sys.stderr
	Supports unicode (but assumes non-unicode is utf8 for now)
	Command history (up key, down key)
	Limits editing and cursor position (but not selection) to ONLY the 
active line of code
	All the nice things that editing in a NSTextView gives you

Obvious future enhancements would be:
	Auto indent
	Syntax highlighting
	Completion
	sys.stdin (for raw_input and such -- might be too hard to do without 
stackless, threads, or other nasty stuff)

It's at http://undefined.org/python/PyInterpreter-0.0.tgz right now, 
I'd like to add this to the PyObjC source tree if someone gives me 
commit access :)

-bob

[Pyobjc-dev] Traffic to your site

[<pr...@gl...>]

You are receiving this e-mail because you were added to
http://www.global-submit.com in the past. Here is the URL
that was submitted to our search engine database.
http://www.gnustep.org/

A successful website can always benefit from
more visitors/traffic. We all know that more
traffic adds up to more sales. Now is the 
time to promote your site for the holiday
season. 

If you think that you could benefit from more 
traffic to your web site, please visit my web 
site at http://william.global-submit.com
give me a call at 1 819 571 4943, 24
hours a day 7 days a week,  for more 
information.

Feel free to send an e-mail to 
wi...@gl...

We offer everything from Banner Advertising,
Safe E-mail Marketing to web site hosting and
development and search engine services

Thank you and have a nice day,

William Hanks
http://william.global-submit.com
wi...@gl...
Call 1 819 571 4943 at any time.





RMV MESSAGE ID : T885365R746775

If would like to be taken out of our database then please visit this web site.
http://www.global-submit.com/cgi-bin/clipout.cgi?T885365R746775

[Pyobjc-dev] About Your Success

[<pr...@gl...>]

You are receiving this e-mail because you were added to
http://www.global-submit.com in the past. Here is the URL
that was submitted to our search engine database.
http://www.gnustep.org/

A successful website can always benefit from
more visitors/traffic. We all know that more
traffic adds up to more sales. Now is the 
time to promote your site for the holiday
season. 

If you think that you could benefit from more 
traffic to your web site, please visit my web 
site at http://william.global-submit.com
give me a call at 1 819 571 4943, 24
hours a day 7 days a week,  for more 
information.

Feel free to send an e-mail to 
wi...@gl...

We offer everything from Banner Advertising,
Safe E-mail Marketing to web site hosting and
development and search engine services

Thank you and have a nice day,

William Hanks
http://william.global-submit.com
wi...@gl...
Call 1 819 571 4943 at any time.





RMV MESSAGE ID : T885365R746775

If would like to be taken out of our database then please visit this web site.
http://www.global-submit.com/cgi-bin/clipout.cgi?T885365R746775

Re: [Pyobjc-dev] List migration prep

[Ronald O. <ous...@ci...>]

I've filed an admin request for this at SF (#830606)

Ronald

Re: [Pyobjc-dev] PyObjC equivalent of ObjC's categories?

[Bob I. <bo...@re...>]

On Oct 26, 2003, at 9:17 AM, Jacob Kaplan-Moss wrote:

> For me, one of the most powerful tools that ObjC gives is the ability 
> to create "categories" on existing classes.  This allows extending 
> existing classes without actually needing to subclass them.  For 
> example (taken from Apple's AnimatedSlider sample code):
>
> @interface NSSlider (SliderAnimation)
>
> - (IBAction) animateToFloatValueFrom: sender;
> - (void) animateToFloatValue:(float) newValue;
>
> @end
>
> The above adds two instance methods to the NSSlider used in the 
> application.
>
> How would I go about doing the same thing in PyObjC?  Obviously I 
> could subclass NSSlider to add the new methods, but I guess I'm 
> looking for something that's a closer functional match to what the 
> ObjC code does.

from the objc module:

Help on built-in function classAddMethods:

classAddMethods(...)
     classAddMethods(targetClass, methodsArray)

     Adds methods in methodsArray to class.   The effect is similar to 
how categories work.   If class already implements a method as defined 
in methodsArray, the original implementation will be replaced by the 
implementation from methodsArray.

I don't believe this is very well known, I had the same question as you 
yesterday, but I found the answer by reading through the PyObjC source 
code.

-bob

Re: [Pyobjc-dev] PyObjC equivalent of ObjC's categories?

[b.bum <bb...@ma...>]

On Oct 26, 2003, at 6:17 AM, Jacob Kaplan-Moss wrote:
> For me, one of the most powerful tools that ObjC gives is the ability 
> to create "categories" on existing classes.  This allows extending 
> existing classes without actually needing to subclass them.  For 
> example (taken from Apple's AnimatedSlider sample code):

Categories can also lead to extremely hard to maintain code because 
they effectively break encapsulation within your code base.   In the 15 
years (jeez!) that I have been doing ObjC-- often by acting as a 
consultant to help others fix/maintain/extend their codebase-- abuse of 
categories has often been the source of serious pain and consternation 
(and billable hours).  Effectively categories allow you to use a class 
design where your functionality is not "below" the existing classes, 
but is effectively inserted throughout.

Powerful?  Yes.  Extremely.  Wonderfully so.  Use them with care.

PyObjC fully supports categories, but in slightly different-- and even 
more dangerous-- fashion.  Effectively you can stick any method-- from 
objc or python-- onto any class.

See:

 >>> import objc
 >>> help(objc.classAddMethods)
 >>> help(objc.classAddMethod)

classAddMethods takes an array of methods and sticks 'em on a class 
somewhere.   classAddMethod allows you to specify the selector used 
when the method is installed.  I.e. you could do:

def doFoo:(self, arg1):
	pass

objc.classAddMethod(Foundation.NSBundle, "handleRequest:", doFoo)

Like categories, the method will be immediately available across all 
instances -- preexisting or newly allocated.

b.bum

[Pyobjc-dev] PyObjC equivalent of ObjC's categories?

[Jacob Kaplan-M. <ja...@ja...>]

Hey all --

For me, one of the most powerful tools that ObjC gives is the ability 
to create "categories" on existing classes.  This allows extending 
existing classes without actually needing to subclass them.  For 
example (taken from Apple's AnimatedSlider sample code):

@interface NSSlider (SliderAnimation)

- (IBAction) animateToFloatValueFrom: sender;
- (void) animateToFloatValue:(float) newValue;

@end

The above adds two instance methods to the NSSlider used in the application.

How would I go about doing the same thing in PyObjC?  Obviously I 
could subclass NSSlider to add the new methods, but I guess I'm 
looking for something that's a closer functional match to what the 
ObjC code does.

TIA,

Jacob

[Pyobjc-dev] [ pyobjc-Bugs-830343 ] typo in PyObjCTools

[SourceForge.net <no...@so...>]

Bugs item #830343, was opened at 2003-10-25 23:40
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=830343&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: typo in PyObjCTools

Initial Comment:
__all__ and __doc__ say sheetEndMethod

Implementation is named endSheetMethod

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=830343&group_id=14534

[Pyobjc-dev] [ pyobjc-Bugs-830342 ] Incorrect signature for a (AppKit) NSNib selector

[SourceForge.net <no...@so...>]

Bugs item #830342, was opened at 2003-10-25 23:39
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=830342&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Bob Ippolito (etrepum)
Assigned to: Nobody/Anonymous (nobody)
Summary: Incorrect signature for a (AppKit) NSNib selector

Initial Comment:
This is correct:
setSignatureForSelector(
    "NSNib",
    "instantiateNibWithOwner:topLevelObjects:",
    "c@:@N^@")

topLevelObjects is an in->out NSMutableArray**

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=830342&group_id=14534

[Pyobjc-dev] Hey, who put that Python in my Xcode?

[Bob I. <bo...@re...>]

Ok.. that was easy.  I wrote an Xcode plugin that EMBEDS A PYTHON 
INTERPRETER INSIDE XCODE!

This is very interesting.. I'm going to play with this more and clean 
it up, but this would make it really easy to hack around Xcode 
internals (god bless Objective C) and add Python specific features, or 
any feature for that matter.  Our Next Generation MacPython IDE might 
actually *be* Xcode ;)

I've put something quick up at 
http://pythonmac.org/wiki/XcodeIntegration

-bob