pyobjc-dev — CVS commit notices go here

Re: [Pyobjc-dev] Comparing int and floats fails?

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

Francesco Pierfederici wrote:

> for some weird reason, if all of v, z1 and z2 are floats, everything
> works as expected. If v is a int while z1 and z2 are floats, the
> second statement gets ALWAYS executed. I should add that z in an
> element of a Python array of unsigned ints.
> 
> Has anybody ever seen something like this?

int/float comparisons work as expected in Python. Are you sure the
objects you're comparing have the types you expect? Maybe one is a
string and not a number? Try putting "print type(v), type(z1), type(z2)"
somewhere.

Just

[Pyobjc-dev] Comparing int and floats fails?

[Francesco P. <fpi...@no...>]

Hi everybody!

I stumbled into a very weird behaviour and I do not know if it is a 
Python bug, an PyObjC bug, one of my bugs or any combination of the 
three.

So, I implement a mouseMoved_ function in a subclass of NSImageView in 
a Python script. In this function I have a comparison between the pixel 
value of the displayed image and two thresholds:

if (v >= z2):
   self.pixelField.setStringValue_ ('>= ' + str (z2))
   return
if (v <= z1):
   self.pixelField.setStringValue_ ('<= ' + str (z1))
   return
else:
   self.pixelField.setFloatValue_ (v)
   return

for some weird reason, if all of v, z1 and z2 are floats, everything 
works as expected. If v is a int while z1 and z2 are floats, the second 
statement gets ALWAYS executed. I should add that z in an element of a 
Python array of unsigned ints.

Has anybody ever seen something like this?

Cheers
Francesco

---
Francesco Pierfederici         <fpi...@no...>
NOAO/AURA Inc.                http://www.noao.edu/staff/fpierfed/
950 N. Cherry Ave.              Phone: +1 520 318 8402
Tucson,  AZ 85719 USA       FAX:    +1 520 318 8360

[Pyobjc-dev] Boost Your Car's Gas Mileage 27%+.....jeanette

[Jodie H. <we8...@ya...>]

FUEL SAVER PRO

This revolutionary device Boosts Gas Mileage 27%+ by helping fuel burn bet=
ter using three patented processes from General Motors.
www.iqkm.org?axel=3D49

PROVEN TECHNOLOGY
A certified U.S. Environmental Protection Agency (EPA) laboratory recently=
 completed tests on the new Fuel Saver. The results were astounding! Maste=
r Service, a subsidiary of Ford Motor Company, also conducted extensive em=
issions testing and obtained similar, unheard of results. The achievements=
 of the Fuel Saver is so noteworthy to the environmental community, that C=
ommercial News has featured it as their cover story in their June, 2000 ed=
ition.

Take a test drive Today - www.iqkm.org?axel=3D49






No more advertisements, thanks - http://www.evkm.org/out5s/pre-rem2e.asp






























h igb gvaw  xtsj 
n pqaw ktaybwjrmoby
ehk drneaqef egyebjxdifuju   l

[Pyobjc-dev] FINANCIAL Independence Is Calling You...%CUSTOM_MINE

[Helena H. <97h...@ya...>]

Got your attention?

Let me show you the way Financial Independence: www.ndid.biz?axel=3D49 









No more advertisements, thanks - www.aype.org/out5s/pre-rem2e.asp




























wi h

[Pyobjc-dev] The Next Boon On The Internet-Keyword Ownership

[Keyword P. C. <no...@re...>]

No text version was provided

[Pyobjc-dev] Including Python Modules in Frameworks

[Carlos P. <ca...@ci...>]

Hi,
How can we make sure that pyobjc sees python classes in embedded 
frameworks. This seems to work in the provided Cocoa-Python-Objc 
Application project template for Project Builder. However, when I build 
a new framework using an Xcode framework target and embed it into my 
application, the modules in the framework are not seen by __main__.py. 
It is of course possible to play with the import path myself as I have 
done. Is this necessary, though? I'm not sure how Resource folders of 
embedded frameworks are treated.

Carlos

[Pyobjc-dev] [ pyobjc-Bugs-857057 ] HelloWorld.py won't run under Mac OS X 10.3

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

Bugs item #857057, was opened at 2003-12-09 18:24
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=857057&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Richard Brooksby (rptb1)
Assigned to: Nobody/Anonymous (nobody)
Summary: HelloWorld.py won't run under Mac OS X 10.3

Initial Comment:
I downloaded and installed PyObjC 1.0 from sources as 
described in install.html.  I then tried to run the HelloWorld 
example, but it failed:

$ python HelloWorld.py 
Traceback (most recent call last):
  File "HelloWorld.py", line 78, in ?
    if __name__ == '__main__' : main()
  File "HelloWorld.py", line 41, in main
    NSApp.setDelegate_(delegate)
AttributeError: 'builtin_function_or_method' object has no 
attribute 'setDelegate_'


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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=857057&group_id=14534

[Pyobjc-dev] [ pyobjc-Bugs-857048 ] FieldGraph won't build under Mac OS X 10.3

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

Bugs item #857048, was opened at 2003-12-09 10:17
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=857048&group_id=14534

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
Summary: FieldGraph won't build under Mac OS X 10.3

Initial Comment:
I installed PyObjC 1.0 from sources under Mac OS X 10.3 
according to the instructions in "install.html".  I then tried 
to build the FieldGraph example using ProjectBuilder 2.1.  
It failed with the following errors:

Building target “FieldGraph” with build style 
“Development” (optimization:level ‘0’, debug-symbols:on) 
— (21 errors)
	    /usr/bin/gcc2  -c -F/Users/rb/src/pyobjc-1.0/
Examples/FieldGraph/build  -I/Users/rb/src/pyobjc-1.0/
Examples/FieldGraph/build/include  -arch ppc -fno-
common -fpascal-strings -O0 -Wmost -Wno-four-char-
constants -Wno-unknown-pragmas -pipe -g -precomp-
trustfile /Users/rb/src/pyobjc-1.0/Examples/FieldGraph/
build/FieldGraph.build/FieldGraph.build/
TrustedPrecomps.txt -Wp,-header-mapfile,/Users/rb/src/
pyobjc-1.0/Examples/FieldGraph/build/FieldGraph.build/
FieldGraph.build/FieldGraph.hmap       bin-python-main.m  
-o /Users/rb/src/pyobjc-1.0/Examples/FieldGraph/build/
FieldGraph.build/FieldGraph.build/Objects-normal/ppc/bin-
python-main.o 
bin-python-main.m: In function `pyobjc_main':
bin-python-main.m:30: parse error before `const'
bin-python-main.m:46: `envc' undeclared (first use in this 
function)
bin-python-main.m:46: (Each undeclared identifier is 
reported only once
bin-python-main.m:46: for each function it appears in.)
bin-python-main.m:48: `PYTHONPATH' undeclared (first 
use in this function)
bin-python-main.m:54: `childEnvp' undeclared (first use in 
this function)
bin-python-main.m:58: `pythonPathInWrapper' undeclared 
(first use in this function)
bin-python-main.m:79: `mainBundle' undeclared (first use 
in this function)
bin-python-main.m:94: `aBundle' undeclared (first use in 
this function)
bin-python-main.m:94: `bundleEnumerator' undeclared 
(first use in this function)
bin-python-main.m:96: `bundlePaths' undeclared (first use 
in this function)
bin-python-main.m:104: parse error before `*'
bin-python-main.m:105: `pythonBinPath' undeclared (first 
use in this function)
bin-python-main.m:107: parse error before `const'
bin-python-main.m:125: `nextFileName' undeclared (first 
use in this function)
bin-python-main.m:125: `possibleMainsEnumerator' 
undeclared (first use in this function)
bin-python-main.m:126: `mainPyPath' undeclared (first 
use in this function)
bin-python-main.m:133: `possibleMains' undeclared (first 
use in this function)
bin-python-main.m:134: parse error before `const'
bin-python-main.m:139: `childArgv' undeclared (first use 
in this function)
bin-python-main.m:142: `mainPyPathPtr' undeclared (first 
use in this function)
bin-python-main.m:147: `i' undeclared (first use in this 
function)
bin-python-main.m:156: `pythonBinPathPtr' undeclared 
(first use in this function)
		bin-python-main.m:30: parse error before 
`const'
		bin-python-main.m:46: `envc' undeclared (first 
use in this function)
bin-python-main.m:46: (Each undeclared identifier is 
reported only once
bin-python-main.m:46: for each function it appears in.)
		bin-python-main.m:48: `PYTHONPATH' 
undeclared (first use in this function)
		bin-python-main.m:54: `childEnvp' undeclared 
(first use in this function)
		bin-python-main.m:58: `pythonPathInWrapper' 
undeclared (first use in this function)
		bin-python-main.m:79: `mainBundle' 
undeclared (first use in this function)
		bin-python-main.m:94: `aBundle' undeclared 
(first use in this function)
		bin-python-main.m:94: `bundleEnumerator' 
undeclared (first use in this function)
		bin-python-main.m:96: `bundlePaths' 
undeclared (first use in this function)
		bin-python-main.m:104: parse error before `*'
		bin-python-main.m:105: `pythonBinPath' 
undeclared (first use in this function)
		bin-python-main.m:107: parse error before 
`const'
		bin-python-main.m:125: `nextFileName' 
undeclared (first use in this function)
		bin-python-main.m:125: 
`possibleMainsEnumerator' undeclared (first use in this 
function)
		bin-python-main.m:126: `mainPyPath' 
undeclared (first use in this function)
		bin-python-main.m:133: `possibleMains' 
undeclared (first use in this function)
		bin-python-main.m:134: parse error before 
`const'
		bin-python-main.m:139: `childArgv' undeclared 
(first use in this function)
		bin-python-main.m:142: `mainPyPathPtr' 
undeclared (first use in this function)
		bin-python-main.m:147: `i' undeclared (first 
use in this function)
		bin-python-main.m:156: `pythonBinPathPtr' 
undeclared (first use in this function)


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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=114534&aid=857048&group_id=14534

Re: [Pyobjc-dev] Accessing python classes in obj-c

[Jiva D. <ji...@de...>]

Got it.  Thanks for the clarification.

On Dec 8, 2003, at 2:09 PM, b.bum wrote:

> Compile time is a bit harder.   In particular, you can't refer to any 
> class implemented in python directly because there won't be a real 
> class object to link against as the source is compiled.   To solve 
> this problem, I have used an abstract superclass implemented in 
> Objective-C (a class that simply declares all the methods where the 
> implementation is stubbed out) and a concrete subclass implemented in 
> Python.  To the abstract superclass-- the objc class-- I have added a 
> factory method that instantiates and returns the concrete subclass.
>
> This is effectively exactly how class clusters work.
>
> Finally, you have to trigger the execution of a bit of python code 
> from somewhere in your code for all of this to work.   The easiest 
> solution would be to simply grab the main() function that 
> uses/initializes the embedded interpreter and use that within your 
> app.   Then your main.py could import whatever classes you need to 
> start the application and you could go from there.
>
> b.bum
--
Jiva DeVoe
jiva at devoesquared.com
http://www.devoesquared.com

Re: [Pyobjc-dev] Can't activate external app

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

On Dec 8, 2003, at 4:26 PM, Philip Brodd wrote:

> I'm trying to launch RealOne Player from my app and send it a URL.  If  
> RealOne is already running, this successfully causes it to play the  
> given audio stream:
>
> import RealOne_Player
> realplayer = RealOne_Player.RealOne_Player()
> realplayer.GetURL( "rtsp://blah/blah.rm" )
>
> I can't see how to launch it from my program though. On a wild guess I  
> tried realplayer.activate(), which gave the error "MacOS.Error: (-600,  
> 'no eligible process with specified descriptor')".

What exactly does this have to do with PyObjC?  PyObjC has no wrapper  
called "RealOne_Player".  That looks like a gensuitemodule wrapper or  
something, you should've specified where it came from.  I believe the  
name for the "launch" event is called run.. so maybe realplayer.run()  
works?

In any case, you may want to try doing this (with PyObjC).  It's  
another approach entirely, but will probably work perfectly fine for  
the simple use case you describe.

	from Foundation import  NSURL
	from AppKit import NSWorkspace

	NSWorkspace.sharedWorkspace().openURL_(NSURL.URLWithString_(u"rtsp:// 
blah/blah.rm"))

-bob

[Pyobjc-dev] Can't activate external app

[Philip B. <br...@ma...>]

I'm trying to launch RealOne Player from my app and send it a URL.  If 
RealOne is already running, this successfully causes it to play the 
given audio stream:

import RealOne_Player
realplayer = RealOne_Player.RealOne_Player()
realplayer.GetURL( "rtsp://blah/blah.rm" )

I can't see how to launch it from my program though. On a wild guess I 
tried realplayer.activate(), which gave the error "MacOS.Error: (-600, 
'no eligible process with specified descriptor')".

--
Phil

Re: [Pyobjc-dev] Accessing python classes in obj-c

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

On Dec 8, 2003, at 12:17 PM, Jiva DeVoe wrote:
> So I understand that I can create Python files and the templates 
> supplied with PyObjC automatically will load those Python files and do 
> what ever I want them to.  I also understand if I instantiate a python 
> class in python, and access it in obj-c, I will get a proxy object to 
> the python object.  However, what if I have an objective-c 
> application, and I want to write a portion of it in python.  So, in 
> other words, I want to be able to create a class in Python, and 
> instantiate an object in objective-c that is of that class.  How would 
> I do that? Would I have to use something like SWIG?

No SWIG necessary.

If you subclass an Objective-C class from Python, the resulting class 
will look/feel/work just like any other Objective-C class.   To ensure 
that selector mapping-- method mapping-- works correctly, you will need 
to make sure that method's in the Python implementation are correctly 
named (i.e. -setObject:forKey: is setObject_forKey_ and not 
setObjectForKey ).

Because Python is a "lazily linked" interpreted language, you have to 
ensure that any classes you want to use have been defined before they 
are used.   This causes two issues;  runtime and compile time.

Runtime is easy -- just make sure you import the class in Python before 
it is used.

Compile time is a bit harder.   In particular, you can't refer to any 
class implemented in python directly because there won't be a real 
class object to link against as the source is compiled.   To solve this 
problem, I have used an abstract superclass implemented in Objective-C 
(a class that simply declares all the methods where the implementation 
is stubbed out) and a concrete subclass implemented in Python.  To the 
abstract superclass-- the objc class-- I have added a factory method 
that instantiates and returns the concrete subclass.

This is effectively exactly how class clusters work.

Finally, you have to trigger the execution of a bit of python code from 
somewhere in your code for all of this to work.   The easiest solution 
would be to simply grab the main() function that uses/initializes the 
embedded interpreter and use that within your app.   Then your main.py 
could import whatever classes you need to start the application and you 
could go from there.

b.bum

[Pyobjc-dev] Accessing python classes in obj-c

[Jiva D. <ji...@de...>]

So I understand that I can create Python files and the templates 
supplied with PyObjC automatically will load those Python files and do 
what ever I want them to.  I also understand if I instantiate a python 
class in python, and access it in obj-c, I will get a proxy object to 
the python object.  However, what if I have an objective-c application, 
and I want to write a portion of it in python.  So, in other words, I 
want to be able to create a class in Python, and instantiate an object 
in objective-c that is of that class.  How would I do that? Would I 
have to use something like SWIG?
--
Jiva DeVoe
jiva at devoesquared.com
http://www.devoesquared.com

Re: [Pyobjc-dev] VTK bindings

[Carlos P. <ca...@ci...>]

> This is the problem: Use PyObject_CallFunction(vtkRenderPyClass, "O", 
> Py_None)  instead, CallObject expects an argument tuple (hence the 
> error message you were getting).
>
> BTW. The code below doesn't seem to use ``rendererAddr``.
>
> Ronald


Thanks you so much. This was in fact the problem. You found both the 
solution to the problem I had and the one I was about to have. The code 
now works. I can now wrap the other methods and things should be fine. 
The correct call was:
result = PyObject_CallFunction(vtkRendererPyClass, "s", rendererAddr);

Thank you to everyone who helped me throughout this thread.

Carlos

Re: [Pyobjc-dev] VTK bindings

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

On 6 dec 2003, at 20:40, Carlos Phillips wrote:

>> Do you parse the method arguments using PyArg_ParseTuple? See one of 
>> the method mappings in the Foundation and AppKit wrappers for an 
>> example of argument parsing.
>
>
> I have included my call_VTKView_renderer function below. I simply 
> ignore the arguments since the method renderer does not take any 
> arguments.
>
> Carlos
>
> static PyObject*
> call_VTKView_renderer(
>         PyObject* method, PyObject* self, PyObject* arguments)
> {
>   PyObject *result = 0;
>   VTKView *objc_self = 0;
>   vtkRenderer *renderer = 0;
>   const char *rendererAddr = 0;
>
>   // Get objc representation of self.
>   objc_self = PyObjC_PythonToId(self);
>
>   // Get the renderer
>   renderer = [objc_self renderer];
>
>   // Convert pointer to mangled form
>   rendererAddr = vtkPythonManglePointer(renderer, "p_vtkRenderer");
>
>   // Import the VTK module and get the vtkRenderer class
>   PyObject *vtk_module = PyImport_ImportModule("vtk");
>   PyObject *vtkRendererPyClass =
>     PyObject_GetAttrString(vtk_module, "vtkRenderer");
>
>   // Create a python object for renderer. If renderer has
>   // already been wrapped in python, then the existing python object
>   // is returned.
>   result = PyObject_CallObject(vtkRendererPyClass, Py_None);

This is the problem: Use PyObject_CallFunction(vtkRenderPyClass, "O", 
Py_None)  instead, CallObject expects an argument tuple (hence the 
error message you were getting).

BTW. The code below doesn't seem to use ``rendererAddr``.

Ronald

Re: [Pyobjc-dev] VTK bindings

[Carlos P. <ca...@ci...>]

> Do you parse the method arguments using PyArg_ParseTuple? See one of 
> the method mappings in the Foundation and AppKit wrappers for an 
> example of argument parsing.


I have included my call_VTKView_renderer function below. I simply 
ignore the arguments since the method renderer does not take any 
arguments.

Carlos

static PyObject*
call_VTKView_renderer(
         PyObject* method, PyObject* self, PyObject* arguments)
{
   PyObject *result = 0;
   VTKView *objc_self = 0;
   vtkRenderer *renderer = 0;
   const char *rendererAddr = 0;

   // Get objc representation of self.
   objc_self = PyObjC_PythonToId(self);

   // Get the renderer
   renderer = [objc_self renderer];

   // Convert pointer to mangled form
   rendererAddr = vtkPythonManglePointer(renderer, "p_vtkRenderer");

   // Import the VTK module and get the vtkRenderer class
   PyObject *vtk_module = PyImport_ImportModule("vtk");
   PyObject *vtkRendererPyClass =
     PyObject_GetAttrString(vtk_module, "vtkRenderer");

   // Create a python object for renderer. If renderer has
   // already been wrapped in python, then the existing python object
   // is returned.
   result = PyObject_CallObject(vtkRendererPyClass, Py_None);

   return result;
}

Re: [Pyobjc-dev] VTK bindings

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

On 6 dec 2003, at 18:04, Carlos Phillips wrote:

>
> On Dec 2, 2003, at 1:13 PM, Bob Ippolito wrote:
>
>>
>> On Dec 2, 2003, at 12:38 PM, Ronald Oussoren wrote:
>>
>>>
>>> On 2 dec 2003, at 16:23, b.bum wrote:
>>>
>>>> On Dec 2, 2003, at 1:12 AM, Carlos Phillips wrote:
>>>>> 2003-12-02 03:30:46.165 Simple1[1541] PyObjCPointer created: at  
>>>>> 0x6d16080 of type  
>>>>> {vtkRenderer={?={?={?=^^?i}C{vtkTimeStamp=^^? 
>>>>> L}^{vtkSubjectHelper}}^{vtkAssemblyPath}^{vtkPropCollection}iIff^{v 
>>>>> tkPropCollection}^{vtkActor2DCollection}^{vtkWindow}[3f][4f][2f][2f 
>>>>> ][2f][2i][2i][3f][3f][4f]}^{vtkCamera}^{vtkLight}^{vtkLightCollecti 
>>>>> on}^{vtkCullerCollection}^{vtkActorCollection}^{vtkVolumeCollection 
>>>>> }[3f]^{vtkRenderWindow}ffiii*{vtkTimeStamp=^^? 
>>>>> L}fii^^{vtkProp}i^^{vtkAssemblyPath}iii[6f]f}8@0:4
>>>>
>>>> Like Bob said, once a PyObjCPointer is created, you are generally  
>>>> out of luck.  PyObjC transparently bridged Objective-C.  In this  
>>>> case, you are asking it to transparently bridge a relatively  
>>>> complex C data type.
>>>
>>> The problem is that PyObjCPointer is a not really a wrapped pointer,  
>>> when the PyObjCPointer object is created the value pointed-to is  
>>> copied into an internal buffer. Most of the time that doesn't work,  
>>> vtkRenderer is probably a C++ class and the copied value is invalid.
>>>
>>> I'll put creation of a real pointer wrapper on my todo-list, a  
>>> generic version of the types used to wrapped some other pointer  
>>> types (see pointer-support.m).
>>>
>>> Is VTK itself wrapped? If so, you could use the functions in  
>>> pointer-support.m to convert vtkRenderer* to the Python wrapper from  
>>> the VTK wrapper.
>>
>> VTK itself does have python wrappers.. they probably work on OS X by  
>> now, but there were problems last time I tried (a long while ago, due  
>> to linking issues).
>>
>> -bob
>>
>
> I am trying to create a wrapper for the following methods in VTKView.
> -(vtkRenderer *)renderer;
> -(vtkRenderWindow *)renderWindow;
> -(vtkRenderWindowInteractor *)renderWindowInteractor;
>
> Each return value is  a C++ class which has a VTK python wrapper. In  
> VTK, there is a function available to convert pointers along with type  
> descriptions into python objects. Thus all I need to do is create a  
> wrapper for these methods which returns the appropriate python object  
> instead of the C++ object.
>
> To do this, I have created a module similar to the AppKit module which  
> installs a wrapper method with the following call.
>
> if (PyObjC_RegisterMethodMapping( classVTKView,
> 				    @selector(renderer),
> 				    call_VTKView_renderer,
> 				    PyObjCUnsupportedMethod_IMP) < 0)
>
> call_VTKView_renderer does the following:
> 1) Unwrap the VTKView python proxy. Call the unwrapped objc object  
> self_objc.
> 2) rend = [self_objc renderer]
> 3) Wrap rend using VTK function.
> 4) Return wrapped rend
>
> I use PyObjCUnsupportedMethod_IMP as the fourth argument to  
> PyObjC_RegisterMethodMapping but I'm not sure what this argument is  
> for.
>
> When I compile the new module and load it into python. I gain access  
> to the VTKView class and can instantiate it. When I call renderer() on  
> an instance, I get the following python error:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: argument list must be a tuple

Do you parse the method arguments using PyArg_ParseTuple? See one of  
the method mappings in the Foundation and AppKit wrappers for an  
example of argument parsing.


>
> Is this because I don't have a callPython argument in my  
> PyObjCUnsupportedMethod_IMP call? If so what needs to be put there?

You can leave the PyObjCUnsupportedMethod_IMP as it is, you only have  
to change that when you want to be able to override the method in  
Python.

Ronald

Re: [Pyobjc-dev] initWithBitmapDataPlanes?

[Francesco P. <fpi...@no...>]

OK,

After trying all possible combinations (in random order), I finally 
managed to make it work. It looks like I was not creating a 
NSAutoreleasePool and my code was just crashing, after trying to 
display an image.

Now it works, I hope.

BTW, PyObjC rocks! And it is quite fast! I am impressed. I will do some 
tests with real images (8K x 8K or bigger) to see how things work when 
handling quite some memory...

Thanks!!!
fra



On Dec 6, 2003, at 10:40 AM, b.bum wrote:

> I added a unit test that invokes the init...() method w/an array of 
> bytes as a grey scale image.  It passes.
>
> However, I'm not rendering anything so it may not explode until the 
> actual rendering happens....
>
> So, three things may be happening:
>
> - you are trashing memory or passing in fewer bytes that expected
>
> - it blows up when it is rendered, but the initialization is working 
> fine
>
> - python is freeing the data you passed to NSBIR before the NSBIR has 
> rendered the bytes
>
> In other words, I didn't actually find anything useful but hopefully 
> the above helps.   What is the last line of python that is executed?
>
> b.bum
>
---
Francesco Pierfederici         <fpi...@no...>
NOAO/AURA Inc.                http://www.noao.edu/staff/fpierfed/
950 N. Cherry Ave.              Phone: +1 520 318 8402
Tucson,  AZ 85719 USA       FAX:    +1 520 318 8360

Re: [Pyobjc-dev] initWithBitmapDataPlanes?

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

I added a unit test that invokes the init...() method w/an array of 
bytes as a grey scale image.  It passes.

However, I'm not rendering anything so it may not explode until the 
actual rendering happens....

So, three things may be happening:

- you are trashing memory or passing in fewer bytes that expected

- it blows up when it is rendered, but the initialization is working 
fine

- python is freeing the data you passed to NSBIR before the NSBIR has 
rendered the bytes

In other words, I didn't actually find anything useful but hopefully 
the above helps.   What is the last line of python that is executed?

b.bum

Re: [Pyobjc-dev] VTK bindings

[Carlos P. <ca...@ci...>]

On Dec 2, 2003, at 1:13 PM, Bob Ippolito wrote:

>
> On Dec 2, 2003, at 12:38 PM, Ronald Oussoren wrote:
>
>>
>> On 2 dec 2003, at 16:23, b.bum wrote:
>>
>>> On Dec 2, 2003, at 1:12 AM, Carlos Phillips wrote:
>>>> 2003-12-02 03:30:46.165 Simple1[1541] PyObjCPointer created: at  
>>>> 0x6d16080 of type  
>>>> {vtkRenderer={?={?={?=^^?i}C{vtkTimeStamp=^^? 
>>>> L}^{vtkSubjectHelper}}^{vtkAssemblyPath}^{vtkPropCollection}iIff^{vt 
>>>> kPropCollection}^{vtkActor2DCollection}^{vtkWindow}[3f][4f][2f][2f][ 
>>>> 2f][2i][2i][3f][3f][4f]}^{vtkCamera}^{vtkLight}^{vtkLightCollection} 
>>>> ^{vtkCullerCollection}^{vtkActorCollection}^{vtkVolumeCollection}[3f 
>>>> ]^{vtkRenderWindow}ffiii*{vtkTimeStamp=^^? 
>>>> L}fii^^{vtkProp}i^^{vtkAssemblyPath}iii[6f]f}8@0:4
>>>
>>> Like Bob said, once a PyObjCPointer is created, you are generally  
>>> out of luck.  PyObjC transparently bridged Objective-C.  In this  
>>> case, you are asking it to transparently bridge a relatively complex  
>>> C data type.
>>
>> The problem is that PyObjCPointer is a not really a wrapped pointer,  
>> when the PyObjCPointer object is created the value pointed-to is  
>> copied into an internal buffer. Most of the time that doesn't work,  
>> vtkRenderer is probably a C++ class and the copied value is invalid.
>>
>> I'll put creation of a real pointer wrapper on my todo-list, a  
>> generic version of the types used to wrapped some other pointer types  
>> (see pointer-support.m).
>>
>> Is VTK itself wrapped? If so, you could use the functions in  
>> pointer-support.m to convert vtkRenderer* to the Python wrapper from  
>> the VTK wrapper.
>
> VTK itself does have python wrappers.. they probably work on OS X by  
> now, but there were problems last time I tried (a long while ago, due  
> to linking issues).
>
> -bob
>

I am trying to create a wrapper for the following methods in VTKView.
-(vtkRenderer *)renderer;
-(vtkRenderWindow *)renderWindow;
-(vtkRenderWindowInteractor *)renderWindowInteractor;

Each return value is  a C++ class which has a VTK python wrapper. In  
VTK, there is a function available to convert pointers along with type  
descriptions into python objects. Thus all I need to do is create a  
wrapper for these methods which returns the appropriate python object  
instead of the C++ object.

To do this, I have created a module similar to the AppKit module which  
installs a wrapper method with the following call.

if (PyObjC_RegisterMethodMapping( classVTKView,
				    @selector(renderer),
				    call_VTKView_renderer,
				    PyObjCUnsupportedMethod_IMP) < 0)

call_VTKView_renderer does the following:
1) Unwrap the VTKView python proxy. Call the unwrapped objc object  
self_objc.
2) rend = [self_objc renderer]
3) Wrap rend using VTK function.
4) Return wrapped rend

I use PyObjCUnsupportedMethod_IMP as the fourth argument to  
PyObjC_RegisterMethodMapping but I'm not sure what this argument is  
for.

When I compile the new module and load it into python. I gain access to  
the VTKView class and can instantiate it. When I call renderer() on an  
instance, I get the following python error:

Traceback (most recent call last):
   File "<stdin>", line 1, in ?
TypeError: argument list must be a tuple

Is this because I don't have a callPython argument in my  
PyObjCUnsupportedMethod_IMP call? If so what needs to be put there?

Carlos

[Pyobjc-dev] Boost Your Car's Gas Mileage 27%+.....%CUSTOM_MINE

[Tommie L. <3nr...@ya...>]

FUEL SAVER PRO

This revolutionary device Boosts Gas Mileage 27%+ by helping fuel burn bet=
ter using three patented processes from General Motors.
www.eoei.com?axel=3D49

PROVEN TECHNOLOGY
A certified U.S. Environmental Protection Agency (EPA) laboratory recently=
 completed tests on the new Fuel Saver. The results were astounding! Maste=
r Service, a subsidiary of Ford Motor Company, also conducted extensive em=
issions testing and obtained similar, unheard of results. The achievements=
 of the Fuel Saver is so noteworthy to the environmental community, that C=
ommercial News has featured it as their cover story in their June, 2000 ed=
ition.

Take a test drive Today - www.eoei.com?axel=3D49






No more advertisements, thanks - http://www.5jzd.org/out5s/pre-rem2e.asp































tzxpnjw  aqzgonbzqb ztiwfcyvcvexnje 

Re: [Pyobjc-dev] initWithBitmapDataPlanes?

[Francesco P. <fpi...@no...>]

Hi,

Thanks for the quick reply!

I actually saw the test scripts but when I tried using the function, I  
got a sgefault :-( (that is why I thought that the function had not  
really been wrapped)

What I do is the following:

img_size = str (len (self.img_data)) + 'B'
temp_data = struct.unpack (img_size, self.img_data)
data = array.array ('B')
for c in temp_data:
     data.append (c)
  del (temp_data)
my_bitmap =  
NSBitmapImageRep.alloc().initWithBitmapDataPlanes_pixelsWide_pixelsHigh_ 
bitsPerSample_samplesPerPixel_hasAlpha_isPlanar_colorSpaceName_bytesPerR 
ow_bitsPerPixel_ (
     (data, None, None, None, None),
     width,
     height,
     8,
     1,
     NO,
     YES,
     NSCalibratedWhiteColorSpace,
     width,
     8)
my_image = NSImage.alloc ().init ()
my_image.addRepresentation_ (my_bitmap)

# display the image in the GUI
self.server.controller.imgView.setImage_ (my_image)


Am I trashing the memory somewhere? I guess so, but have no idea  
where...


Cheers
fra


On Dec 6, 2003, at 12:03 AM, b.bum wrote:

> On Dec 5, 2003, at 10:40 PM, Francesco Pierfederici wrote:
>> The reason I am asking is that I am writing an application that  
>> receives a raster through a socket connection. The raster is a  
>> (potentially large) array of bytes that I would like to display as a  
>> grey-scale bitmap. Hence the need for NSBitmapImageRep and its  
>> initWithBitmapDataPlanes:...., unless, of course, there is a better  
>> way of doing displaying that data, other than using a  
>> NSBitmapImageRep!
>
> It is already partially mapped, but the assumption is that you have --  
> effectively -- five data planes worth of color information passed in  
> as a tuple.
>
> I believe that this means it will just work in your case -- you would  
> need to pass a tuple like (greyScaleBytes, None, None, None, None) in  
> as the first argument.
>
> See Modules/Appkit/_AppKitMapping_NSBitmap.m for the mapping code.  It  
> is incomplete.
>
> Better yet, see -- Lib/AppKit/test/test_nsbitmapimagerep.py -- as it  
> contains a set of unit tests that are passing and do pass bitmap data  
> across the bridge.
>
> So -- maybe -- Guido's time machine has been effective once again for  
> your particular use case. ;-)
>
> b.bum
---
Francesco Pierfederici         <fpi...@no...>
NOAO/AURA Inc.                http://www.noao.edu/staff/fpierfed/
950 N. Cherry Ave.              Phone: +1 520 318 8402
Tucson,  AZ 85719 USA       FAX:    +1 520 318 8360

Re: [Pyobjc-dev] initWithBitmapDataPlanes?

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

On Dec 5, 2003, at 10:40 PM, Francesco Pierfederici wrote:
> The reason I am asking is that I am writing an application that 
> receives a raster through a socket connection. The raster is a 
> (potentially large) array of bytes that I would like to display as a 
> grey-scale bitmap. Hence the need for NSBitmapImageRep and its 
> initWithBitmapDataPlanes:...., unless, of course, there is a better 
> way of doing displaying that data, other than using a 
> NSBitmapImageRep!

It is already partially mapped, but the assumption is that you have -- 
effectively -- five data planes worth of color information passed in as 
a tuple.

I believe that this means it will just work in your case -- you would 
need to pass a tuple like (greyScaleBytes, None, None, None, None) in 
as the first argument.

See Modules/Appkit/_AppKitMapping_NSBitmap.m for the mapping code.  It 
is incomplete.

Better yet, see -- Lib/AppKit/test/test_nsbitmapimagerep.py -- as it 
contains a set of unit tests that are passing and do pass bitmap data 
across the bridge.

So -- maybe -- Guido's time machine has been effective once again for 
your particular use case. ;-)

b.bum

[Pyobjc-dev] initWithBitmapDataPlanes?

[Francesco P. <fpi...@no...>]

Hi,

I just started using PyObjC and, so far, I am realy impressed by it!

I have a question; are there plans to wrap 
initWithBitmapDataPlanes:.... (in NSBitmapImageRep)?

The reason I am asking is that I am writing an application that 
receives a raster through a socket connection. The raster is a 
(potentially large) array of bytes that I would like to display as a 
grey-scale bitmap. Hence the need for NSBitmapImageRep and its 
initWithBitmapDataPlanes:...., unless, of course, there is a better way 
of doing displaying that data, other than using a NSBitmapImageRep!


Thanks
Francesco


---
Francesco Pierfederici         <fpi...@no...>
NOAO/AURA Inc.                http://www.noao.edu/staff/fpierfed/
950 N. Cherry Ave.              Phone: +1 520 318 8402
Tucson,  AZ 85719 USA       FAX:    +1 520 318 8360

Re: [Pyobjc-dev] Projects and executables and stuff

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

On Dec 5, 2003, at 12:57 PM, Jiva DeVoe wrote:
>  Forgive my lack of knowledge, please see my question below:
> On Dec 5, 2003, at 12:53 PM, b.bum wrote:
>
>> A brief update of the current state of things:
>>
>> - Panther and Jaguar require different build styles (unless you have 
>> Frameworkized Python installed)
>>
>
>  isn't the Mac  Python which ships with Panther a framework python?

Yes -- which means that Panther and Frameworkized python build the same 
way.  The non-linkable Jaguar python requires the different build style 
(that includes the execve() silliness).

b.bum
(Who was being remarkably obtuse in the original email... sorry).