From niki at vintech.bg  Mon Oct  1 11:36:51 2007
From: niki at vintech.bg (niki)
Date: Mon, 01 Oct 2007 12:36:51 +0300
Subject: [C++-sig] Announcing pybindgen
In-Reply-To: <7465b6170709261308h571c91c3ofe5837dea5cc3a1d@mail.gmail.com>
References: <a467ca4f0709260548j42e3bba8y2b0aaf3a43b42621@mail.gmail.com>
	<7465b6170709261308h571c91c3ofe5837dea5cc3a1d@mail.gmail.com>
Message-ID: <4700BFB3.2080608@vintech.bg>

Roman Yakovenko wrote:
>     Although I like Py++ a lot, I really hate Boost.Python.  Maybe it's
>     because of my lack of skill in understanding C++ templates, but I
>     feel like Boost.Python hides too many details underneath a layer of
>     cryptic to follow set of C++ templates, while I need to see the
>     generated code in order to check that memory management details are
>     being declared and implemented correctly. 
> 
> 
> :-), I think you are going to reinvent the wheel here. I guess there are 
> many reasons why memory management  is complex and you eventually 
> rediscover all of them

About reinventing the wheel: Pyrex/Cython has to deal with similar 
memory management problems. Maybe there are some parts to share?

HTH

Niki Spahiev



From ropoctl at gmail.com  Tue Oct  2 18:26:33 2007
From: ropoctl at gmail.com (Robert O'Callahan)
Date: Tue, 2 Oct 2007 11:26:33 -0500
Subject: [C++-sig] Py++ question
Message-ID: <3be982d00710020926v2b2885c4j68c4d2b25e3a9ab1@mail.gmail.com>

I am trying to use a generate_code.py created by the py++ gui, but I
get this error:

Traceback (most recent call last):
  File "generate_code.py", line 40, in ?
    export()
  File "generate_code.py", line 37, in export
    mb.write_module( os.path.join( settings.generated_files_dir,
settings.module_name + '.cpp') )
  File "/usr/lib/python2.4/site-packages/pyplusplus/module_builder/builder.py",
line 311, in write_module
    file_writers.write_file( self.code_creator, file_name,
encoding=self.encoding )
  File "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/__init__.py",
line 40, in write_file
    sf.write()
  File "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/single_file.py",
line 27, in write
    self.write_file( self.file_name, self.extmodule.create(),
encoding=self.encoding )
  File "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/writer.py",
line 124, in write_file
    f.write( unicode( fcontent_new, encoding ) )
TypeError: decoding Unicode is not supported

Should I be writing the code generator from scratch?

-Rob


From matthew.scouten at gmail.com  Tue Oct  2 18:47:41 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Tue, 2 Oct 2007 11:47:41 -0500
Subject: [C++-sig] Wrapping global data
Message-ID: <3dc9bcf00710020947r231719c9u469de01e580f00f2@mail.gmail.com>

OK, I can wrap a global function, I can wrap a member function, I can
wrap a member variable,
But I can't seem to figure out how to wrap a global variable. Maybe
I'm just suffering from
temporary stupidity.

def_readonly() does not exist,
def() results in hundreds of cryptic error messages that look like it
is trying to wrap a function.


From roman.yakovenko at gmail.com  Tue Oct  2 20:52:49 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 2 Oct 2007 20:52:49 +0200
Subject: [C++-sig] Py++ question
In-Reply-To: <3be982d00710020926v2b2885c4j68c4d2b25e3a9ab1@mail.gmail.com>
References: <3be982d00710020926v2b2885c4j68c4d2b25e3a9ab1@mail.gmail.com>
Message-ID: <7465b6170710021152o5bb101ffvfd1bb92852008990@mail.gmail.com>

On 10/2/07, Robert O'Callahan <ropoctl at gmail.com> wrote:
>
> I am trying to use a generate_code.py created by the py++ gui, but I
> get this error:
>
> Traceback (most recent call last):
>   File "generate_code.py", line 40, in ?
>     export()
>   File "generate_code.py", line 37, in export
>     mb.write_module( os.path.join( settings.generated_files_dir,
> settings.module_name + '.cpp') )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/module_builder/builder.py",
> line 311, in write_module
>     file_writers.write_file( self.code_creator, file_name,
> encoding=self.encoding )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/__init__.py",
> line 40, in write_file
>     sf.write()
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/single_file.py",
> line 27, in write
>     self.write_file( self.file_name, self.extmodule.create(),
> encoding=self.encoding )
>   File
> "/usr/lib/python2.4/site-packages/pyplusplus/file_writers/writer.py",
> line 124, in write_file
>     f.write( unicode( fcontent_new, encoding ) )
> TypeError: decoding Unicode is not supported
>
> Should I be writing the code generator from scratch?



What version of Py++ do you use?  I fixed such error few weeks or month ago.

Can you try SVN version?

P.S. You will have to get SVN version of pygccxml too, but you can stay with
your current version of GCC-XML

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071002/a5fd0972/attachment.htm>

From roman.yakovenko at gmail.com  Tue Oct  2 21:02:43 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 2 Oct 2007 21:02:43 +0200
Subject: [C++-sig] Wrapping global data
In-Reply-To: <3dc9bcf00710020947r231719c9u469de01e580f00f2@mail.gmail.com>
References: <3dc9bcf00710020947r231719c9u469de01e580f00f2@mail.gmail.com>
Message-ID: <7465b6170710021202p116a5474ye9ec490981361c51@mail.gmail.com>

On 10/2/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
>
> OK, I can wrap a global function, I can wrap a member function, I can
> wrap a member variable,


The answer is: it depends
* on global variable type
* whether you need to share it between Python and C++ code at the same time

for immutable types: http://boost.org/libs/python/doc/v2/scope.html

for shared class instances:
http://wiki.python.org/moin/boost.python/HowTo#head-85904d74f4f9f20c1380ede50e2eb95a4c333c37

 P.S. Google will point you to other resources

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071002/12289d78/attachment.htm>

From matthew.scouten at gmail.com  Tue Oct  2 22:20:45 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Tue, 2 Oct 2007 15:20:45 -0500
Subject: [C++-sig] Wrapping global data
In-Reply-To: <7465b6170710021202p116a5474ye9ec490981361c51@mail.gmail.com>
References: <3dc9bcf00710020947r231719c9u469de01e580f00f2@mail.gmail.com>
	<7465b6170710021202p116a5474ye9ec490981361c51@mail.gmail.com>
Message-ID: <3dc9bcf00710021320h5e36131w3878d3bd6ef3dc20@mail.gmail.com>

On 10/2/07, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 10/2/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
> > OK, I can wrap a global function, I can wrap a member function, I can
> > wrap a member variable,
>
> The answer is: it depends
> * on global variable type
> * whether you need to share it between Python and C++ code at the same time
>
> for immutable types:
> http://boost.org/libs/python/doc/v2/scope.html
>
> for shared class instances:
> http://wiki.python.org/moin/boost.python/HowTo#head-85904d74f4f9f20c1380ede50e2eb95a4c333c37
>
>  P.S. Google will point you to other resources
>

Ah. From
http://mail.python.org/pipermail/c++-sig/2003-June/004240.html
it looks like what I thought I wanted to do is not possible. What I
actually need to do is create a function that returns that variable in
question. I would have thought that exposing a global variable
directly would be a fairly obvious thing to do. Is there a reason that
it is not possible?


From roman.yakovenko at gmail.com  Wed Oct  3 07:54:08 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Wed, 3 Oct 2007 07:54:08 +0200
Subject: [C++-sig] Wrapping global data
In-Reply-To: <3dc9bcf00710021320h5e36131w3878d3bd6ef3dc20@mail.gmail.com>
References: <3dc9bcf00710020947r231719c9u469de01e580f00f2@mail.gmail.com>
	<7465b6170710021202p116a5474ye9ec490981361c51@mail.gmail.com>
	<3dc9bcf00710021320h5e36131w3878d3bd6ef3dc20@mail.gmail.com>
Message-ID: <7465b6170710022254r7e05482al559f96702ed484f3@mail.gmail.com>

On 10/2/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
>
>
>
> Ah. From
> http://mail.python.org/pipermail/c++-sig/2003-June/004240.html
> it looks like what I thought I wanted to do is not possible. What I
> actually need to do is create a function that returns that variable in
> question. I would have thought that exposing a global variable
> directly would be a fairly obvious thing to do. Is there a reason that
> it is not possible?
>

Hint: int is immutable type in Python

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071003/6ff85bc5/attachment.htm>

From jumpjoe at fastwebnet.it  Wed Oct  3 15:46:09 2007
From: jumpjoe at fastwebnet.it (Giovanni Beltrame)
Date: Wed, 3 Oct 2007 15:46:09 +0200
Subject: [C++-sig] Boost.Python type registry
Message-ID: <200710031546.09201.jumpjoe@fastwebnet.it>

Hello everybody,
I am trying to understand how the Boost.Python type registry works...

What I want to do is, given a python object (let's say a PyObject*), determine
its wrapper C++ type_info.
Theoretically, what I want to do is going name->PyObject->C++ type. Notice
that the traditional extract<T> won't work for me as I want to determine which 
type is being wrapped.

I know Boost.Python is doing this internally in some way, but I haven't 
figured it out yet.

Can anybody help?

Thanks in advance,
-- 
Giovanni Beltrame
TEC-EDM Research Fellow
European Space Agency


From duranlef at iro.umontreal.ca  Wed Oct  3 22:31:33 2007
From: duranlef at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Duranleau?=)
Date: Wed, 3 Oct 2007 16:31:33 -0400 (EDT)
Subject: [C++-sig] Boost.Python type registry
In-Reply-To: <200710031546.09201.jumpjoe@fastwebnet.it>
References: <200710031546.09201.jumpjoe@fastwebnet.it>
Message-ID: <Pine.LNX.4.63.0710031110530.10764@salade.iro.umontreal.ca>

On Wed, 3 Oct 2007, Giovanni Beltrame wrote:

> Hello everybody,
> I am trying to understand how the Boost.Python type registry works...
>
> What I want to do is, given a python object (let's say a PyObject*), determine
> its wrapper C++ type_info.
> Theoretically, what I want to do is going name->PyObject->C++ type. Notice
> that the traditional extract<T> won't work for me as I want to determine which
> type is being wrapped.
>
> I know Boost.Python is doing this internally in some way, but I haven't
> figured it out yet.
>
> Can anybody help?

AFAIK, there is nothing in the API to help you there. Boost.Python doesn't 
work directly this way internally. It provides a way to test if an 
instance holds a given type, but nothing to return the contained type 
info.

-- 
Francois Duranleau


From pvicenti at gmail.com  Thu Oct  4 00:14:48 2007
From: pvicenti at gmail.com (Paul Vicenti)
Date: Wed, 3 Oct 2007 19:14:48 -0300
Subject: [C++-sig] Extansion compilation in Linux
Message-ID: <531d9eef0710031514g3c49ebfax7b810af02f50f8f9@mail.gmail.com>

Hi,
I'm moving a project from Windows and I'm having problems to compile. I get a:
'undefined reference to 'PyThreadState_Get' and whatever I use in my
code of Python library.

In windows I include Python.h and it links with python library to
resolve all symbols.
Here, it looks that the include file is found (otherwise, I expect
'cannot open file Python.h') but the library seems not to be linked.

Any help for this?

thanks,
paul


From jumpjoe at fastwebnet.it  Thu Oct  4 11:55:38 2007
From: jumpjoe at fastwebnet.it (jumpjoe at fastwebnet.it)
Date: Thu, 4 Oct 2007 11:55:38 +0200 (CEST)
Subject: [C++-sig] R: Re:  Boost.Python type registry
Message-ID: <33471873.346151191491738471.JavaMail.defaultUser@defaultHost>



>AFAIK, there is nothing in the API to help you there. Boost.Python doesn't 
>work directly this way internally. It provides a way to test if an 
>instance holds a given type, but nothing to return the contained type 
>info.

Is there at least a way to get a raw pointer to the wrapped object from a 
PyObject? I think extract<> is doing it.

Cheers,
Giovanni Beltrame


From duranlef at iro.umontreal.ca  Thu Oct  4 16:45:27 2007
From: duranlef at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Duranleau?=)
Date: Thu, 4 Oct 2007 10:45:27 -0400 (EDT)
Subject: [C++-sig] Extansion compilation in Linux
In-Reply-To: <531d9eef0710031514g3c49ebfax7b810af02f50f8f9@mail.gmail.com>
References: <531d9eef0710031514g3c49ebfax7b810af02f50f8f9@mail.gmail.com>
Message-ID: <Pine.LNX.4.63.0710041044050.30926@salade.iro.umontreal.ca>

On Wed, 3 Oct 2007, Paul Vicenti wrote:

> Hi,
> I'm moving a project from Windows and I'm having problems to compile. I get a:
> 'undefined reference to 'PyThreadState_Get' and whatever I use in my
> code of Python library.
>
> In windows I include Python.h and it links with python library to
> resolve all symbols.
> Here, it looks that the include file is found (otherwise, I expect
> 'cannot open file Python.h') but the library seems not to be linked.
>
> Any help for this?

AFAIK, there is no auto-link feature with gcc. You need to tell it 
explicitly to link with the Python library by adding the swithc 
-lpython2.x (where x is the version you are using).

-- 
Francois Duranleau


From duranlef at iro.umontreal.ca  Thu Oct  4 17:13:55 2007
From: duranlef at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Duranleau?=)
Date: Thu, 4 Oct 2007 11:13:55 -0400 (EDT)
Subject: [C++-sig] R: Re:  Boost.Python type registry
In-Reply-To: <33471873.346151191491738471.JavaMail.defaultUser@defaultHost>
References: <33471873.346151191491738471.JavaMail.defaultUser@defaultHost>
Message-ID: <Pine.LNX.4.63.0710041045490.30926@salade.iro.umontreal.ca>

On Thu, 4 Oct 2007, jumpjoe at fastwebnet.it wrote:

>> AFAIK, there is nothing in the API to help you there. Boost.Python doesn't
>> work directly this way internally. It provides a way to test if an
>> instance holds a given type, but nothing to return the contained type
>> info.
>
> Is there at least a way to get a raw pointer to the wrapped object from a
> PyObject? I think extract<> is doing it.

Neither. The only thing extract<> is "extracting" is the source object's 
type name as exposed to Python (obj->ob_type->tp_name where obj is a 
PyObject*) when reporting errors. Although with that you can maybe work 
something by registering an "inverse" mapping from Python type name to C++ 
type_info, but you would need to explicitly add entries to that registry 
each time you expose a class.

-- 
Francois Duranleau


From nickm at sitius.com  Thu Oct  4 20:14:30 2007
From: nickm at sitius.com (Nikolay Mladenov)
Date: Thu, 04 Oct 2007 14:14:30 -0400
Subject: [C++-sig] R: Re:  Boost.Python type registry
References: <33471873.346151191491738471.JavaMail.defaultUser@defaultHost>
Message-ID: <47052D86.7795232C@sitius.com>

I think that you can search the registry for entry with m_class_object
that matches 
your PyObject's type or search the lvalue_chain(s) for a converter that
returns !=0.
But there is no api for that so far.

HTH,

Nikolay

jumpjoe at fastwebnet.it wrote:
> 
> >AFAIK, there is nothing in the API to help you there. Boost.Python doesn't
> >work directly this way internally. It provides a way to test if an
> >instance holds a given type, but nothing to return the contained type
> >info.
> 
> Is there at least a way to get a raw pointer to the wrapped object from a
> PyObject? I think extract<> is doing it.
> 
> Cheers,
> Giovanni Beltrame



From gtang at bcm.edu  Fri Oct  5 00:36:12 2007
From: gtang at bcm.edu (Grant)
Date: Thu, 04 Oct 2007 17:36:12 -0500
Subject: [C++-sig] inner class pickle issue
Message-ID: <fe3pss$8ks$1@sea.gmane.org>

I am trying to write pickle support for an inner class by using
boost.pyhton. But it always fail with this error message:

>>> pickle.dump(k,mm)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib64/python2.5/pickle.py", line 1362, in dump
    Pickler(file, protocol).dump(obj)
  File "/usr/lib64/python2.5/pickle.py", line 224, in dump
    self.save(obj)
  File "/usr/lib64/python2.5/pickle.py", line 331, in save
    self.save_reduce(obj=obj, *rv)
  File "/usr/lib64/python2.5/pickle.py", line 400, in save_reduce
    save(func)
  File "/usr/lib64/python2.5/pickle.py", line 295, in save
    self.save_global(obj)
  File "/usr/lib64/python2.5/pickle.py", line 748, in save_global
    (obj, module, name))
pickle.PicklingError: Can't pickle <class 'libpyUtils2.KaiserBessel'>: it's
not found as libpyUtils2.KaiserBessel

Can I pickle an inner class? Or only the top level class can be pickled?

Grant



From rwgk at yahoo.com  Fri Oct  5 01:14:20 2007
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 4 Oct 2007 16:14:20 -0700 (PDT)
Subject: [C++-sig] inner class pickle issue
Message-ID: <351729.46895.qm@web31111.mail.mud.yahoo.com>

> Or only the top level class can be pickled?

Yes.

Try this:

import pickle
class outer(object):
  class inner(object): pass
oi = outer.inner()
pickle.dumps(oi)

You'll get a similar error message:

pickle.PicklingError: Can't pickle <class '__main__.inner'>: it's not found as __main__.inner

Ralf







From jumpjoe at fastwebnet.it  Fri Oct  5 15:23:46 2007
From: jumpjoe at fastwebnet.it (jumpjoe at fastwebnet.it)
Date: Fri, 5 Oct 2007 15:23:46 +0200 (CEST)
Subject: [C++-sig] R: Re:  R: Re:  Boost.Python type registry
Message-ID: <25216082.396051191590626009.JavaMail.defaultUser@defaultHost>



>I think that you can search the registry for entry with m_class_object
>that matches 
>your PyObject's type or search the lvalue_chain(s) for a converter that
>returns !=0.
>But there is no api for that so far.

This is another interesting approach, but after a few tries, I discovered the 
registry is really hard to get by, as it is a static variable defined in a 
"hidden" function in an anonymous namespace in the registry.cpp file.
Any suggestions on how to extract it? As an "extern" of some kind?

Cheers,
Giovanni Beltrame


From duranlef at iro.umontreal.ca  Fri Oct  5 17:22:20 2007
From: duranlef at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Duranleau?=)
Date: Fri, 5 Oct 2007 11:22:20 -0400 (EDT)
Subject: [C++-sig] R: Re:  R: Re:  Boost.Python type registry
In-Reply-To: <25216082.396051191590626009.JavaMail.defaultUser@defaultHost>
References: <25216082.396051191590626009.JavaMail.defaultUser@defaultHost>
Message-ID: <Pine.LNX.4.63.0710051117450.31218@brocoli.iro.umontreal.ca>

On Fri, 5 Oct 2007, jumpjoe at fastwebnet.it wrote:

>> I think that you can search the registry for entry with m_class_object 
>> that matches your PyObject's type or search the lvalue_chain(s) for a 
>> converter that returns !=0. But there is no api for that so far.
>
> This is another interesting approach, but after a few tries, I 
> discovered the registry is really hard to get by, as it is a static 
> variable defined in a "hidden" function in an anonymous namespace in the 
> registry.cpp file. Any suggestions on how to extract it? As an "extern" 
> of some kind?

Without modifying the source code, there is no way. The only way to search 
the registry externally is via the boost::python::converter::query 
function, but then you need to provide a type_info, and to do this 
exhaustively is not really simpler than using an "inverse" registry like I 
mentionned.

-- 
Francois Duranleau


From rwgk at yahoo.com  Fri Oct  5 20:46:26 2007
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Fri, 5 Oct 2007 11:46:26 -0700 (PDT)
Subject: [C++-sig] R: Re:  R: Re:  Boost.Python type registry
Message-ID: <858615.18297.qm@web31112.mail.mud.yahoo.com>

Would it help you if you had an interface to extract an array of
type_info objects known to the registry?
If you send patches with docs I'll review them.
BTW: What you need seems very unusual. Could you explain the situation?
Did you consider alternative approaches?
Ralf

----- Original Message ----
From: Fran?ois Duranleau <duranlef at iro.umontreal.ca>
To: jumpjoe at fastwebnet.it; Development of Python/C++ integration <c++-sig at python.org>
Sent: Friday, October 5, 2007 8:22:20 AM
Subject: Re: [C++-sig] R: Re:  R: Re:  Boost.Python type registry

On Fri, 5 Oct 2007, jumpjoe at fastwebnet.it wrote:

>> I think that you can search the registry for entry with m_class_object 
>> that matches your PyObject's type or search the lvalue_chain(s) for a 
>> converter that returns !=0. But there is no api for that so far.
>
> This is another interesting approach, but after a few tries, I 
> discovered the registry is really hard to get by, as it is a static 
> variable defined in a "hidden" function in an anonymous namespace in the 
> registry.cpp file. Any suggestions on how to extract it? As an "extern" 
> of some kind?

Without modifying the source code, there is no way. The only way to search 
the registry externally is via the boost::python::converter::query 
function, but then you need to provide a type_info, and to do this 
exhaustively is not really simpler than using an "inverse" registry like I 
mentionned.

-- 
Francois Duranleau
_______________________________________________
C++-sig mailing list
C++-sig at python.org
http://mail.python.org/mailman/listinfo/c++-sig





From duranlef at iro.umontreal.ca  Sat Oct  6 03:29:07 2007
From: duranlef at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Duranleau?=)
Date: Fri, 5 Oct 2007 21:29:07 -0400 (EDT)
Subject: [C++-sig] R: Re:  R: Re:  Boost.Python type registry
In-Reply-To: <858615.18297.qm@web31112.mail.mud.yahoo.com>
References: <858615.18297.qm@web31112.mail.mud.yahoo.com>
Message-ID: <Pine.LNX.4.63.0710052115040.24656@salade.iro.umontreal.ca>

On Fri, 5 Oct 2007, Ralf W. Grosse-Kunstleve wrote:

> Would it help you if you had an interface to extract an array of
> type_info objects known to the registry?
> If you send patches with docs I'll review them.
> BTW: What you need seems very unusual. Could you explain the situation?
> Did you consider alternative approaches?

I don't know why the OP required this. I once thought it would be useful 
for implicit conversions to/from boost.function objects, but then I 
realized that even if you get the type_info, you still need to extract the 
object without having a (C++) type. So I ended up having to register 
manually an implicit conversion to/from boost.function for all exposed 
functors, which is not all that convenient. And I have yet to make this 
work for functions.

-- 
Francois Duranleau


From p.jaroszynski at gmail.com  Sat Oct  6 18:59:23 2007
From: p.jaroszynski at gmail.com (Piotr =?utf-8?q?Jaroszy=C5=84ski?=)
Date: Sat, 6 Oct 2007 18:59:23 +0200
Subject: [C++-sig] Multiple-inheritance problem
Message-ID: <200710061859.24313.p.jaroszynski@gmail.com>

Hello,

Seems that when one creates a Python class, which multiple-inherits from 
classes exposed by boost.python then it can be only converted(casted) to the 
first class in the inheritance list. Simple example code follows.

C++:
#include <boost/python.hpp>

namespace bp = boost::python;

struct A { };
struct B { };

void test_A(const A &) { }
void test_B(const B &) { }

BOOST_PYTHON_MODULE(multiple_inheritance)
{
    bp::def("test_A", &test_A);
    bp::def("test_B", &test_B);

    bp::class_<A>("A", bp::init<>());
    bp::class_<B>("B", bp::init<>());
}

Python:
class C(A, B):
    pass

test_A(C())
test_B(C())

And I get:
Traceback (most recent call last):
  File "./multiple_inheritance.py", line 10, in <module>
    test_B(C())
Boost.Python.ArgumentError: Python argument types in
    multiple_inheritance.test_B(C)
did not match C++ signature:
    test_B(B)

Am I missing something or is it a known problem? Any workaround if so?

-- 
Best Regards,
Piotr Jaroszy?ski


From jumpjoe at fastwebnet.it  Mon Oct  8 10:48:07 2007
From: jumpjoe at fastwebnet.it (Giovanni Beltrame)
Date: Mon, 8 Oct 2007 10:48:07 +0200
Subject: [C++-sig]  Re:  Boost.Python type registry
In-Reply-To: <858615.18297.qm@web31112.mail.mud.yahoo.com>
References: <858615.18297.qm@web31112.mail.mud.yahoo.com>
Message-ID: <200710081048.07356.jumpjoe@fastwebnet.it>

On Friday 05 October 2007 20:46:26 Ralf W. Grosse-Kunstleve wrote:

> BTW: What you need seems very unusual. Could you explain the situation?
> Did you consider alternative approaches?

The situation is definitely unusual, but I will try to explain.

I have a simulator that uses a C++ kernel and a bunch of simulation 
components. Now I want to wrap both kernel and components and add scripting 
and inspection functionality in Python. Python will be used to instantiate 
the simulation components and run what I will call the "architecture".
So far, so good, I have everything wrapped.

The problem is that I might want to debug the architecture. I was thinking of 
using GDB, but since components are instantiated in Python (note that I might 
have more objects that belong to one class), it is hard to get a reference to 
them. Setting a breakpoint on the source code is be easy, but putting a watch 
on a variable is much more tricky.

In the end what I would like to have is a sort of python-controlled GDB 
breakpoint mechanism, where I set a condition in python and C++ is stopped 
when the condition arises.
Getting a reference to the object/variable should be enough to put a watch in 
GDB (and here's why I'm trying to pass from PyObject to the C++ address).

Any suggestion is welcome!

-- 
Giovanni Beltrame
TEC-EDM Research Fellow
European Space Agency


From ndbecker2 at gmail.com  Tue Oct  9 21:02:27 2007
From: ndbecker2 at gmail.com (Neal Becker)
Date: Tue, 09 Oct 2007 15:02:27 -0400
Subject: [C++-sig] interaction of boost::shared_ptr with boost::python
Message-ID: <fegj8p$aci$1@sea.gmane.org>

A bit puzzled about this one.
struct X {
  X() { std::cout << "X()" << '\n'; }
  ~X() { std::cout << "~X()" << '\n'; }
};

boost::shared_ptr<X> make_x () {
  return boost::shared_ptr<X> (new X);
}

boost::shared_ptr<X> make_x_from_shared (boost::shared_ptr<X> x) {
  return boost::shared_ptr<X> (new X);
}


X* clone1 (X const& x) {
  return new X;
}

boost::shared_ptr<X> clone2 (X const& x) {
  return boost::shared_ptr<X> (new X);
}
BOOST_PYTHON_MODULE (stuff) {
  class_<X> ("X", no_init)
    .def ("__init__", make_constructor (make_x))
    .def ("__init__", make_constructor (make_x_from_shared))
    .def ("__copy__", &clone1, return_value_policy<manage_new_object>())
    .def ("__copy2__", &clone2)
    ;

  def ("clone1", &clone1, return_value_policy<manage_new_object>());
  def ("clone2", &clone2);


>>> x = X()
>>> clone2(x)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: No to_python (by-value) converter found for C++ type: boost::shared_ptr<X>





From ndbecker2 at gmail.com  Wed Oct 10 20:54:25 2007
From: ndbecker2 at gmail.com (Neal Becker)
Date: Wed, 10 Oct 2007 14:54:25 -0400
Subject: [C++-sig] No registered converter was able to produce a C++
	rvalue...
Message-ID: <fej76a$pup$1@sea.gmane.org>

I know this comes up frequently, but this one has me baffled.

u = array1_complex_int ((1,2))
TypeError: No registered converter was able to produce a C++ rvalue of type
std::complex<int> from this Python object of type int

But there is a converter, it can be invoked from the python side:
>>> complex_int (2)
<blitz.complex_int object at 0x2aaaaab69db8>

or like this:
u = array1_complex_int (complex_int(e) for e in (1,2))

It seems that extract<std::complex<int> > is not finding the conversion. 
Any hints?

The conversion is there, because there is this minimal class_:

  bp::class_<std::complex<int> > ("complex_int", bp::init<int,optional<int>
>())
    ;




From seefeld at sympatico.ca  Wed Oct 10 21:08:40 2007
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Wed, 10 Oct 2007 15:08:40 -0400
Subject: [C++-sig] PyObject_Call() wrapper ?
Message-ID: <470D2338.7020004@sympatico.ca>

I'm looking for the boost.python equivalent of the generic

callable(*args, **kwds)

i.e. a version of the call operator that takes an arguments vector
as well as a keyword dictionary, instead of individual arguments.
I can't find it. Is that not implemented ? May I submit
a patch for it ?

Thanks,
		Stefan


-- 

      ...ich hab' noch einen Koffer in Berlin...


From amohr at pixar.com  Wed Oct 10 22:28:58 2007
From: amohr at pixar.com (Alex Mohr)
Date: Wed, 10 Oct 2007 13:28:58 -0700
Subject: [C++-sig] PyObject_Call() wrapper ?
In-Reply-To: <470D2338.7020004@sympatico.ca>
References: <470D2338.7020004@sympatico.ca>
Message-ID: <470D360A.8040508@pixar.com>

> I'm looking for the boost.python equivalent of the generic
> 
> callable(*args, **kwds)
> 
> i.e. a version of the call operator that takes an arguments vector
> as well as a keyword dictionary, instead of individual arguments.
> I can't find it. Is that not implemented ? May I submit
> a patch for it ?

I implemented this by overloading unary * so that in boost python you 
could actually write what you wrote above.  David A. signed off on the 
idea so I think he'd welcome a patch.

Unfortunately I never got permission from my employer to post my patch.

I think it would be awesome if you did it -- I think these are the cases:

object(a1, a2 .. aN)
object(a1, a2 .. aN, *args)
object(a1, a2 .. aN, **kw)
object(a1, a2 .. aN, *args, **kw)

Alex


From steven.wyckoff at gmail.com  Wed Oct 10 23:53:06 2007
From: steven.wyckoff at gmail.com (Steven Wyckoff)
Date: Wed, 10 Oct 2007 15:53:06 -0600
Subject: [C++-sig] convert child class to child class {lvalue} ?
Message-ID: <fejhjp$vso$1@sea.gmane.org>

Hi all,

I am trying to wrap some C++ code that involves an object factory and 
manager and I seem to be stumped by how boost is converting types. I have a 
base class and a few classes that inherit from it and a factory to create 
instances of these classes. Using the factory I can successfully create 
objects of the proper type. Following the tutorial I have virtual functions 
behaving properly. However, if I try to call methods that only exist in a 
child class, I get an Argument Error:

ArgumentError: Python argument types in
    childobj.ChildOnly(childobj)
did not match C++ signature:
    ChildOnly(struct childobj {lvalue})

 I have tried writing a function that will cast the pointer to a child class 
pointer. This lets me call child only methods, but then i get the mirror 
error when I try to call base class methods. The confounding part is that it 
appears that the type information is correct in python.

If I create an instance without using the factory (child = childobj()) then 
everything works just fine.

I guess I could write converters to go both ways and just cast to whatever 
type I need for the current function call, but that is hardly ideal. If 
anyone has any suggestions about how I can get around this (maybe a type 
converter that I'm not aware of?) I would appreciate it.

Here is my code:


C++ code:
-----------------------------------------------------------------------
struct baseobj
{
    virtual std::string ID() {return std::string("BaseObj");}
    void SetVal(int v) {val = v;}
    int GetVal() {return val;}
    int val;
};

struct childobj : public baseobj
{
    virtual std::string ID() {return std::string("ChildObj");}
    std::string ChildOnly() {return "OK";}
};

baseobj* createobj(std::string kind)
{
    if(kind == "Base")
    {
        return new baseobj();
    }
    else if(kind == "Child")
    {
        return new childobj();
    }
    else
    {
        return NULL;
    }
}

childobj* CastToChild(baseobj* pobj)
{
    return static_cast<childobj*>(pobj);
}

struct BaseWrap : baseobj, boost::python::wrapper<baseobj>
{
    std::string ID()
    {
        if(boost::python::override f = this->get_override("ID"))
            return ID();
        return baseobj::ID();
    }
    std::string default_ID() {return this->baseobj::ID();}
};

BOOST_PYTHON_MODULE(extending)
{
using namespace boost::python;

class_<BaseWrap, boost::noncopyable>("baseobj")
.def("ID", &baseobj::ID, &BaseWrap::ID, "Returns the ID of the object")
.def("GetVal", &baseobj::GetVal)
.def("SetVal", &baseobj::SetVal)
;

class_<childobj, bases<BaseWrap> >("childobj")
.def("ChildOnly", &childobj::ChildOnly)
;

implicitly_convertible<childobj,baseobj>();
class_<objman>("objman")
.def("SetObj", &objman::SetObj)
.def("GetObj", &objman::GetObj, 
return_value_policy<reference_existing_object>())
;

def("create", &createobj, return_value_policy<manage_new_object>() );

def("callID", &callID );

def("CastToChild", &CastToChild, 
return_value_policy<reference_existing_object>());

}


Python:
----------------------------------------------------------------------------
>>> from extending import *
>>> base = create("Base")
>>> base.ID()
'BaseObj'
>>> child = create("Child")
>>> child.ID()
'ChildObj'
>>> child.ChildOnly()
Traceback (most recent call last):
  File "<pyshell#3>", line 1, in <module>
    child.ChildOnly()
ArgumentError: Python argument types in
    childobj.ChildOnly(childobj)
did not match C++ signature:
    ChildOnly(struct childobj {lvalue})
>>> castchild = CastToChild(child)
>>> castchild.ChildOnly()
'OK'
>>> castchild.ID()
Traceback (most recent call last):
  File "<pyshell#6>", line 1, in <module>
    castchild.ID()
ArgumentError: Python argument types in
    baseobj.ID(childobj)
did not match C++ signature:
    ID(struct BaseWrap {lvalue})
    ID(struct baseobj {lvalue})





From Chang_Chen at sonic.com  Thu Oct 11 07:44:52 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Thu, 11 Oct 2007 13:44:52 +0800
Subject: [C++-sig] Can boost python do these 2 things automatically?
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF0160033A@SHAEXC01.ent.sonic.com>

Hi All

 

I am new to boost python. I have 2 questions for using boost python to
wrap C++ API

 

Our C++ API looks like this:

 

class BImpl;

Class B

{

  public:

   B(){}

   Error DoSomething();

private:

  BImpl*; mpBImpl         

}

class A

{

  public:

   A(){}

   Error CreateB(B& bObj);

  private:

    ...

}

 

Our error handle strategy is returning an error code other than throwing
exception. This is because throwing an exception across the DLL boundary
is not safe.

 

I want to wrap A and B with same C++ public interface in python, but 

 

#1 I want to return real object rather than return an error code. i.e.
in python

 

aObject = A()

*bObject = aObject.CreateB()*

 

#2 I want to throw an *python* exception rather than return an error
code like c++.

 

I do not know how to do it in Boost Python.

 

Any suggestion?

 

Chang

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071011/f2b4d5fe/attachment.htm>

From Mark.English at rbccm.com  Thu Oct 11 09:44:35 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Thu, 11 Oct 2007 08:44:35 +0100
Subject: [C++-sig] convert child class to child class {lvalue} ?
Message-ID: <EE7D40281B517A41AF5A90847A61F279379655BC@SEW39051.oak.fg.rbc.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org [mailto:c++-sig-bounces at python.org]On
> Behalf Of Steven Wyckoff
> Sent: 10 October 2007 22:53
> To: c++-sig at python.org
> Subject: [C++-sig] convert child class to child class {lvalue} ?
> 
> 
> Hi all,
> 
> I am trying to wrap some C++ code that involves an object factory and 
> manager and I seem to be stumped by how boost is converting 
> types.
<snip>
> baseobj* createobj(std::string kind)
> {
>     if(kind == "Base")
>     {
>         return new baseobj();
>     }
>     else if(kind == "Child")
>     {
>         return new childobj();
>     }
>     else
>     {
>         return NULL;
>     }
> }
Haven't tested this myself, but try something like:
---C++ code---
boost::shared_ptr<baseobj> createobj(std::string kind)
{
    boost::shared_ptr<baseobj> ptrCreate;
    if(kind == "Base")
    {
        ptrCreate = new baseobj;
    }
    else if(kind == "Child")
    {
        ptrCreate = new childobj;
    }
    else
    {
        ptrCreate = NULL;
    }
    return ptrCreate;
}
________________________________________

This E-Mail (including any attachments) may contain privileged or confidential information.  It is intended only for the addressee(s) indicated above.

The sender does not waive any of its rights, privileges or other protections respecting this information.  

Any distribution, copying or other use of this E-Mail or the information it contains, by other than an intended recipient, is not sanctioned and is prohibited.

If you received this E-Mail in error, please delete it and advise the sender (by return E-Mail or otherwise) immediately. 

This E-Mail (including any attachments) has been scanned for viruses. 

It is believed to be free of any virus or other defect that might affect any computer system into which it is received and opened. 

However, it is the responsibility of the recipient to ensure that it is virus free. 

The sender accepts no responsibility for any loss or damage arising in any way from its use.

E-Mail received by or sent from RBC Capital Markets is subject to review by Supervisory personnel. 

Such communications are retained and may be produced to regulatory authorities or others with legal rights to the information.

IRS CIRCULAR 230 NOTICE:  TO COMPLY WITH U.S. TREASURY REGULATIONS, WE ADVISE YOU THAT ANY U.S. FEDERAL TAX ADVISE INCLUDED IN THIS COMMUNICATION IS NOT INTENDED OR WRITTEN TO BE USED, AND CANNOT BE USED, TO AVOID ANY U.S. FEDERAL TAX PENALTIES OR TO PROMOTE, MARKET, OR RECOMMEND TO ANOTHER PARTY ANY TRANSACTION OR MATTER.


From Chang_Chen at sonic.com  Thu Oct 11 10:58:30 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Thu, 11 Oct 2007 16:58:30 +0800
Subject: [C++-sig] Can boost python do these 2 things automatically?
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF016003FC@SHAEXC01.ent.sonic.com>

Hi All

I am new to boost python. I have 2 questions for using boost python to wrap C++ API

Our C++ API looks like this:

class BImpl;
Class B
{
? public:
?? B(){}
?? Error DoSomething();
private:
? BImpl*; mpBImpl???????? 
}
class A
{
? public:
?? A(){}
?? Error CreateB(B& bObj);
? private:
??? ...
}

Our error handle strategy is returning an error code other than throwing exception. This is because throwing an exception across the DLL boundary is not safe.

I want to wrap A and B with same C++ public interface in python, but 

#1 I want to return real object rather than retrieve object. i.e. in python, I would like users writing code like this:

	aObject = A()
	~bObject = aObject.CreateB()~

#2 I want to throw an *python* exception rather than return an error code like c++.

I do not know how to do it in Boost Python.

Any suggestion?

Chang


From j.reid at mail.cryst.bbk.ac.uk  Thu Oct 11 12:05:16 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Thu, 11 Oct 2007 11:05:16 +0100
Subject: [C++-sig] Raoul Gough's container suite with svn boost-trunk
	boost.python?
Message-ID: <feksho$e5u$1@sea.gmane.org>

Has anyone tried using Raoul's container suite with the svn boost-trunk?

I'm getting the following error:

C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(236) 
: error C2903: 'extract_return_type' : symbol is neither a class 
template nor a function template
 
C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(232) 
: while compiling class template member function 
'boost::python::detail::py_func_sig_info 
boost::python::detail::caller_arity<2>::impl<F,Policies,Sig>::signature(void)'
         with
         [
             F=boost::python::list (__cdecl 
*)(std::vector<biopsy::binding_hit> &,boost::python::indexing::slice),
 
Policies=boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>,
 
Sig=boost::mpl::vector3<boost::python::list,std::vector<biopsy::binding_hit> 
&,boost::python::indexing::slice>
         ]

So the problem seems to be there is no extract_return_type in 
boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>

If I remember correctly, I've installed Roman's version by copying it 
into boost/python/suite/indexing

John.



From meine at informatik.uni-hamburg.de  Thu Oct 11 13:01:12 2007
From: meine at informatik.uni-hamburg.de (Hans Meine)
Date: Thu, 11 Oct 2007 13:01:12 +0200
Subject: [C++-sig] convert child class to child class {lvalue} ?
In-Reply-To: <fejhjp$vso$1@sea.gmane.org>
References: <fejhjp$vso$1@sea.gmane.org>
Message-ID: <200710111301.12681.meine@informatik.uni-hamburg.de>

Am Mittwoch, 10. Oktober 2007 23:53:06 schrieb Steven Wyckoff:
> BOOST_PYTHON_MODULE(extending)
> {
> using namespace boost::python;
>
> class_<BaseWrap, boost::noncopyable>("baseobj")
> .def("ID", &baseobj::ID, &BaseWrap::ID, "Returns the ID of the object")
> .def("GetVal", &baseobj::GetVal)
> .def("SetVal", &baseobj::SetVal)
> ;
>
> class_<childobj, bases<BaseWrap> >("childobj")
> .def("ChildOnly", &childobj::ChildOnly)
> ;

I never used wrapper<>, but did you try bases<baseobj>?
I mean - childobj *is* derived from baseobj, not BaseWrap.

-- 
Ciao, /  /
     /--/
    /  / ANS


From nickm at sitius.com  Thu Oct 11 18:20:07 2007
From: nickm at sitius.com (Nikolay Mladenov)
Date: Thu, 11 Oct 2007 12:20:07 -0400
Subject: [C++-sig] Raoul Gough's container suite with svn
	boost-trunkboost.python?
References: <feksho$e5u$1@sea.gmane.org>
Message-ID: <470E4D37.5B9DE011@sitius.com>

This error is related to the documentation patch which I recently
submited.
I will look into it.

Nikolay

John Reid wrote:
> 
> Has anyone tried using Raoul's container suite with the svn boost-trunk?
> 
> I'm getting the following error:
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(236)
> : error C2903: 'extract_return_type' : symbol is neither a class
> template nor a function template
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(232)
> : while compiling class template member function
> 'boost::python::detail::py_func_sig_info
> boost::python::detail::caller_arity<2>::impl<F,Policies,Sig>::signature(void)'
>          with
>          [
>              F=boost::python::list (__cdecl
> *)(std::vector<biopsy::binding_hit> &,boost::python::indexing::slice),
> 
> Policies=boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>,
> 
> Sig=boost::mpl::vector3<boost::python::list,std::vector<biopsy::binding_hit>
> &,boost::python::indexing::slice>
>          ]
> 
> So the problem seems to be there is no extract_return_type in
> boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>
> 
> If I remember correctly, I've installed Roman's version by copying it
> into boost/python/suite/indexing
> 
> John.



From steven.wyckoff at gmail.com  Thu Oct 11 18:32:19 2007
From: steven.wyckoff at gmail.com (Steven Wyckoff)
Date: Thu, 11 Oct 2007 10:32:19 -0600
Subject: [C++-sig] convert child class to child class {lvalue} ?
In-Reply-To: <200710111301.12681.meine@informatik.uni-hamburg.de>
References: <fejhjp$vso$1@sea.gmane.org>
	<200710111301.12681.meine@informatik.uni-hamburg.de>
Message-ID: <felj67$7l3$1@sea.gmane.org>

Hans Meine wrote:
> Am Mittwoch, 10. Oktober 2007 23:53:06 schrieb Steven Wyckoff:
>> BOOST_PYTHON_MODULE(extending)
>> {
>> using namespace boost::python;
>>
>> class_<BaseWrap, boost::noncopyable>("baseobj")
>> .def("ID", &baseobj::ID, &BaseWrap::ID, "Returns the ID of the object")
>> .def("GetVal", &baseobj::GetVal)
>> .def("SetVal", &baseobj::SetVal)
>> ;
>>
>> class_<childobj, bases<BaseWrap> >("childobj")
>> .def("ChildOnly", &childobj::ChildOnly)
>> ;
> 
> I never used wrapper<>, but did you try bases<baseobj>?
> I mean - childobj *is* derived from baseobj, not BaseWrap.
> 

That was it! That was the sneaky, silly thing I was doing wrong that 
broke everything. Switching to bases<baseobj> seems to let boost cast 
properly when legal just as one would expect it should. Thanks for the help!

Steven



From seefeld at sympatico.ca  Thu Oct 11 19:09:21 2007
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 11 Oct 2007 13:09:21 -0400
Subject: [C++-sig] PyObject_Call() wrapper ?
In-Reply-To: <470D360A.8040508@pixar.com>
References: <470D2338.7020004@sympatico.ca> <470D360A.8040508@pixar.com>
Message-ID: <470E58C1.3040008@sympatico.ca>

Alex Mohr wrote:
>> I'm looking for the boost.python equivalent of the generic
>>
>> callable(*args, **kwds)

[...]

> I think it would be awesome if you did it -- I think these are the cases:

I just submitted a patch:

http://svn.boost.org/trac/boost/ticket/1316

Reviews are highly welcome, especially from Dave. :-)

> object(a1, a2 .. aN)
> object(a1, a2 .. aN, *args)
> object(a1, a2 .. aN, **kw)
> object(a1, a2 .. aN, *args, **kw)

While these could all be implemented, the only one directly supported
by the Python C API (PyObject_Call()) is the one with N==0, i.e.
object(*args, **kw), so this is what my patch includes. Anything else
requires to manually rebuild the argument vector before passing it down.

(As a cosmetic point:  I'm not sure whether the code layout is ideal.
It may be better to put the additions into a separate file. Dave ?)


Regards,
		Stefan

-- 

      ...ich hab' noch einen Koffer in Berlin...


From amohr at pixar.com  Thu Oct 11 19:33:41 2007
From: amohr at pixar.com (Alex Mohr)
Date: Thu, 11 Oct 2007 10:33:41 -0700
Subject: [C++-sig] PyObject_Call() wrapper ?
In-Reply-To: <470E58C1.3040008@sympatico.ca>
References: <470D2338.7020004@sympatico.ca> <470D360A.8040508@pixar.com>
	<470E58C1.3040008@sympatico.ca>
Message-ID: <470E5E75.9040309@pixar.com>

> I just submitted a patch:
> 
> http://svn.boost.org/trac/boost/ticket/1316
> 
> Reviews are highly welcome, especially from Dave. :-)

Cool!

>> object(a1, a2 .. aN)
>> object(a1, a2 .. aN, *args)
>> object(a1, a2 .. aN, **kw)
>> object(a1, a2 .. aN, *args, **kw)
> 
> While these could all be implemented, the only one directly supported
> by the Python C API (PyObject_Call()) is the one with N==0, i.e.
> object(*args, **kw), so this is what my patch includes. Anything else
> requires to manually rebuild the argument vector before passing it down.

I think they should all be implemented because boost.python's c++ api is 
meant to let you write c++ code that is as similar as possible to 
python.  I find this to be one of the great features of boost.python.

Alex


From nickm at sitius.com  Thu Oct 11 21:01:18 2007
From: nickm at sitius.com (Nikolay Mladenov)
Date: Thu, 11 Oct 2007 15:01:18 -0400
Subject: [C++-sig] Raoul Gough's container suite with svn
	boost-trunkboost.python?
References: <feksho$e5u$1@sea.gmane.org>
Message-ID: <470E72FE.48A19C6D@sitius.com>

I found it,

the postcall_override template needs to have the following line

typedef typename Policy::extract_return_type extract_return_type;

HTH,

Nikolay

John Reid wrote:
> 
> Has anyone tried using Raoul's container suite with the svn boost-trunk?
> 
> I'm getting the following error:
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(236)
> : error C2903: 'extract_return_type' : symbol is neither a class
> template nor a function template
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(232)
> : while compiling class template member function
> 'boost::python::detail::py_func_sig_info
> boost::python::detail::caller_arity<2>::impl<F,Policies,Sig>::signature(void)'
>          with
>          [
>              F=boost::python::list (__cdecl
> *)(std::vector<biopsy::binding_hit> &,boost::python::indexing::slice),
> 
> Policies=boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>,
> 
> Sig=boost::mpl::vector3<boost::python::list,std::vector<biopsy::binding_hit>
> &,boost::python::indexing::slice>
>          ]
> 
> So the problem seems to be there is no extract_return_type in
> boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>
> 
> If I remember correctly, I've installed Roman's version by copying it
> into boost/python/suite/indexing
> 
> John.



From nickm at sitius.com  Thu Oct 11 19:50:57 2007
From: nickm at sitius.com (Nikolay Mladenov)
Date: Thu, 11 Oct 2007 13:50:57 -0400
Subject: [C++-sig] Raoul Gough's container suite with svn
	boost-trunkboost.python?
References: <feksho$e5u$1@sea.gmane.org>
Message-ID: <470E6281.ACED5B86@sitius.com>

Hi John,

I could not find the class postcall_override. Can you give me a pointer?
By "Roman's version", do you mean the files that come with py++?

Nikolay

John Reid wrote:
> 
> Has anyone tried using Raoul's container suite with the svn boost-trunk?
> 
> I'm getting the following error:
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(236)
> : error C2903: 'extract_return_type' : symbol is neither a class
> template nor a function template
> 
> C:\Dev\ThirdParty\boost\boost-trunk\boost/python/detail/caller.hpp(232)
> : while compiling class template member function
> 'boost::python::detail::py_func_sig_info
> boost::python::detail::caller_arity<2>::impl<F,Policies,Sig>::signature(void)'
>          with
>          [
>              F=boost::python::list (__cdecl
> *)(std::vector<biopsy::binding_hit> &,boost::python::indexing::slice),
> 
> Policies=boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>,
> 
> Sig=boost::mpl::vector3<boost::python::list,std::vector<biopsy::binding_hit>
> &,boost::python::indexing::slice>
>          ]
> 
> So the problem seems to be there is no extract_return_type in
> boost::python::indexing::detail::postcall_override<boost::python::indexing::default_container_policies>
> 
> If I remember correctly, I've installed Roman's version by copying it
> into boost/python/suite/indexing
> 
> John.



From j.reid at mail.cryst.bbk.ac.uk  Fri Oct 12 08:49:44 2007
From: j.reid at mail.cryst.bbk.ac.uk (John Reid)
Date: Fri, 12 Oct 2007 07:49:44 +0100
Subject: [C++-sig] Raoul Gough's container suite with svn
	boost-trunkboost.python?
In-Reply-To: <470E6281.ACED5B86@sitius.com>
References: <feksho$e5u$1@sea.gmane.org> <470E6281.ACED5B86@sitius.com>
Message-ID: <fen5f4$qgb$1@sea.gmane.org>

Nikolay Mladenov wrote:
> Hi John,
> 
> I could not find the class postcall_override. Can you give me a pointer?
> By "Roman's version", do you mean the files that come with py++?
> 
Yes. If I remember correctly they were taken from the py++ SVN 
repository. I can try and check later.



From matthew.scouten at gmail.com  Fri Oct 12 18:18:30 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Fri, 12 Oct 2007 11:18:30 -0500
Subject: [C++-sig] wrapping a function that takes a void*
Message-ID: <3dc9bcf00710120918w639580cas574390c07b10d354@mail.gmail.com>

I am wraping a function that takes a void* as an argument.

class Special
{
	int fooRequest(void * userData = NULL);
	virtual void fooCallback(void * userData)
}

The intention is that the user data pointer will be returned to
'remind' the program what callback goes with what request.

my first attempt:

struct Special_wrapper: Special, wrapper<Special>
{
	int fooRequest(PyObject * userData = NULL);
}

int Special_wrapper::fooRequest(PyObject * userData)
{
	return  fooRequest((void*) userData)
}


and later

class<Special_wrapper>("Special")
	.def("fooRequest" ,&Special_wrapper::fooRequest	,(arg("userData") =
(void*)NULL))


I get a error the first line of which is:
C:\Documents and
Settings\mscouten\Desktop\boost_1_32_0\boost_1_32_0\boost/python/converter/arg_to_python.hpp(181)
: error C2027: use of undefined type
'boost::python::converter::detail::cannot_convert_raw_PyObject<T>'

(Full error text available on request)

Does any body have any recommendations about how I could wrap this.


From Mark.English at rbccm.com  Fri Oct 12 18:33:55 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Fri, 12 Oct 2007 17:33:55 +0100
Subject: [C++-sig] wrapping a function that takes a void*
Message-ID: <EE7D40281B517A41AF5A90847A61F279379655CD@SEW39051.oak.fg.rbc.com>

X-Replace-Address: mark.ignorethisbit.english at xxxrbccmxxx.ignorethisbittoo_and_removethosepreceding_xxxs_.com
> -----Original Message-----
> From: c++-sig-bounces at python.org [mailto:c++-sig-bounces at python.org]On
> Behalf Of Matthew Scouten
> Sent: 12 October 2007 17:19
> To: Development of Python/C++ integration
> Subject: [C++-sig] wrapping a function that takes a void*
> 
> 
> I am wraping a function that takes a void* as an argument.
Is an opaque pointer what you're looking for ?
http://www.boost.org/libs/python/doc/v2/return_opaque_pointer.html

Note there are some potential bugs in this. The PyTypeObject's name causes a crash I believe but I'm using boost 1.32, Python 2.3.6 so it may have been fixed by now in either of those.
For example using the example code:
>>> import opaque_ext
>>> p = opaque_ext.get()
>>> type(p) #Causes an access violation

Cheers,
mE
______________________________________________________________________

This email is intended only for the use of the individual(s) to whom it is addressed and may be privileged and confidential.
Unauthorised use or disclosure is prohibited.If you receive This e-mail in error, please advise immediately and delete the original message.
This message may have been altered without your or our knowledge and the sender does not accept any liability for any errors or omissions in the message.


From matthew.scouten at gmail.com  Fri Oct 12 18:44:59 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Fri, 12 Oct 2007 11:44:59 -0500
Subject: [C++-sig] wrapping a function that takes a void*
In-Reply-To: <EE7D40281B517A41AF5A90847A61F279379655CD@SEW39051.oak.fg.rbc.com>
References: <EE7D40281B517A41AF5A90847A61F279379655CD@SEW39051.oak.fg.rbc.com>
Message-ID: <3dc9bcf00710120944i72b3be5dva66634080d825a01@mail.gmail.com>

On 10/12/07, English, Mark <Mark.English at rbccm.com> wrote:
> X-Replace-Address: mark.ignorethisbit.english at xxxrbccmxxx.ignorethisbittoo_and_removethosepreceding_xxxs_.com
> > -----Original Message-----
> > From: c++-sig-bounces at python.org [mailto:c++-sig-bounces at python.org]On
> > Behalf Of Matthew Scouten
> > Sent: 12 October 2007 17:19
> > To: Development of Python/C++ integration
> > Subject: [C++-sig] wrapping a function that takes a void*
> >
> >
> > I am wraping a function that takes a void* as an argument.
> Is an opaque pointer what you're looking for ?
> http://www.boost.org/libs/python/doc/v2/return_opaque_pointer.html
>
> Note there are some potential bugs in this. The PyTypeObject's name causes a crash I believe but I'm using boost 1.32, Python 2.3.6 so it may have been fixed by now in either of those.
> For example using the example code:
> >>> import opaque_ext
> >>> p = opaque_ext.get()
> >>> type(p) #Causes an access violation
>
> Cheers,
> mE

No, I need the opposite of that. I have a reference to some python
object (any python object) that needs to be passed to a c++ function
expecting a (void *). Later, that object will be handed back to python
as a reference again.

It is fine to tell the python programmer that it is his problem to
prevent the object from being GCed in the meantime.


From Mark.English at rbccm.com  Fri Oct 12 18:47:49 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Fri, 12 Oct 2007 17:47:49 +0100
Subject: [C++-sig] wrapping a function that takes a void*
Message-ID: <EE7D40281B517A41AF5A90847A61F279379655CF@SEW39051.oak.fg.rbc.com>

X-Replace-Address: mark.ignorethisbit.english at xxxrbccmxxx.ignorethisbittoo_and_removethosepreceding_xxxs_.com
> No, I need the opposite of that. I have a reference to some python
> object (any python object) that needs to be passed to a c++ function
> expecting a (void *). Later, that object will be handed back to python
> as a reference again.

Ok. How about:
int Special_wrapper::fooRequest(boost::python::object objUserData)
{
	return  fooRequest((void*) objUserData.ptr());
}
______________________________________________________________________

This email is intended only for the use of the individual(s) to whom it is addressed and may be privileged and confidential.
Unauthorised use or disclosure is prohibited.If you receive This e-mail in error, please advise immediately and delete the original message.
This message may have been altered without your or our knowledge and the sender does not accept any liability for any errors or omissions in the message.


From matthew.scouten at gmail.com  Fri Oct 12 21:02:59 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Fri, 12 Oct 2007 14:02:59 -0500
Subject: [C++-sig] wrapping a function that takes a void*
In-Reply-To: <EE7D40281B517A41AF5A90847A61F279379655CF@SEW39051.oak.fg.rbc.com>
References: <EE7D40281B517A41AF5A90847A61F279379655CF@SEW39051.oak.fg.rbc.com>
Message-ID: <3dc9bcf00710121202h478284a7lcab1dfd1b54e3313@mail.gmail.com>

On 10/12/07, English, Mark <Mark.English at rbccm.com> wrote:
> X-Replace-Address: mark.ignorethisbit.english at xxxrbccmxxx.ignorethisbittoo_and_removethosepreceding_xxxs_.com
> > No, I need the opposite of that. I have a reference to some python
> > object (any python object) that needs to be passed to a c++ function
> > expecting a (void *). Later, that object will be handed back to python
> > as a reference again.
>
> Ok. How about:
> int Special_wrapper::fooRequest(boost::python::object objUserData)
> {
>         return  fooRequest((void*) objUserData.ptr());
> }

Ok, that looks sane.
two questions though:
1) I assume I get the 'object' back with ' object o(p) ' where is a
void pointer?

2) how do I handle the default argument?  the export used to be
.def("fooRequest" ,&Special_wrapper::fooRequest ,(arg("userData") =
(void*)NULL))
Can I get a None to use in place of that NULL?


From simon at dackbrann.net  Sun Oct 14 16:55:29 2007
From: simon at dackbrann.net (Simon Norberg)
Date: Sun, 14 Oct 2007 16:55:29 +0200
Subject: [C++-sig] access struct inside class
Message-ID: <47122DE1.3050801@dackbrann.net>

Hello,
im trying to get python to read and write to a struct inside a class
in the same way i do in c++
Customers hej;
hej.address.contact = "The Devil";

but i cant even get the code to compile, any suggestions?

#include <iostream>
#include <string>

using namespace std;
class Customers {
private:
struct s_address
{
string contact;
string street1;
string street2;
string city;
string province;
string postalCode;
string country;
};
public:
s_address address;
Customers(){}
~Customers(){}
};

#include <boost/python.hpp>
#include <boost/python/module.hpp>
#include <boost/python/class.hpp>
#include <boost/python/scope.hpp>
#include <boost/python/def.hpp>

using namespace boost::python;

BOOST_PYTHON_MODULE(test)
{
scope outer=
class_<Customers>("Customers")
;
class_<Customers::address>("address")
.def_readwrite("contact", &Customers::address.contact)
;
}

When i compile i get this error:
test.cpp: In function ?void init_module_test()?:
test.cpp:36: error: ?Customers::address? cannot appear in a 
constant-expression
test.cpp:36: error: template argument 1 is invalid
test.cpp:18: error: invalid use of non-static data member 
?Customers::address?
test.cpp:37: error: from this location

what i want to be able to do in python is
from test import *
cust = Customers()
cust.address.contact = "Hello"
print cust.address.contact


From nico_ml at mgdesign.fr  Mon Oct 15 19:09:53 2007
From: nico_ml at mgdesign.fr (Nicolas Lelong)
Date: Mon, 15 Oct 2007 19:09:53 +0200
Subject: [C++-sig] access struct inside class
References: <47122DE1.3050801@dackbrann.net>
Message-ID: <00d701c80f4e$3458a6f0$dd00a8c0@shreider>

Hello,

it should be

&Customers::address::contact

not

&Customers::address.contact

...


From grh at mur.at  Mon Oct 15 20:07:37 2007
From: grh at mur.at (Georg Holzmann)
Date: Mon, 15 Oct 2007 20:07:37 +0200
Subject: [C++-sig] datastructure to numpy and back
Message-ID: <4713AC69.4000804@mur.at>

Hallo!

I know this was discussed quite some times, but I did find that many 
workarounds and not really a common way how to do that with Boost.Python 
... so I am a little bit confused ...

My problem: I have a class with some (big) custom C++ matrix types and I 
want to write some getters/setters for them to/from numpy arrays, where 
I want to avoid copies of the data.

So is there a common way to do this ?
(e.g. num_util from 
http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/num_util_release2/Readme.html 
?)

Or can maybe someone link me to some docu or examples which already did 
this ? (it must be quite common ...)

Many thanks for any hint,
LG
Georg


From simon at dackbrann.net  Mon Oct 15 20:48:53 2007
From: simon at dackbrann.net (Simon Norberg)
Date: Mon, 15 Oct 2007 20:48:53 +0200
Subject: [C++-sig] access struct inside class
In-Reply-To: <00d701c80f4e$3458a6f0$dd00a8c0@shreider>
References: <47122DE1.3050801@dackbrann.net>
	<00d701c80f4e$3458a6f0$dd00a8c0@shreider>
Message-ID: <4713B615.2070805@dackbrann.net>

Hello

Thanks for ypur replay
whit your proposed change i get:

test.cpp: In function ?void init_module_test()?:
test.cpp:36: error: ?Customers::address? cannot appear in a 
constant-expression
test.cpp:36: error: template argument 1 is invalid
test.cpp:37: error: ?class Customers::address? is not a class or namespace
error: command 'gcc' failed with exit status 1
make: *** [default] Fel 1

one error down 3 more to go

Nicolas Lelong wrote:
> Hello,
>
> it should be
>
> &Customers::address::contact
>
> not
>
> &Customers::address.contact
>
> ...
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
>   



From candace_walker at naughtydog.com  Mon Oct 15 22:31:23 2007
From: candace_walker at naughtydog.com (Candace Walker)
Date: Mon, 15 Oct 2007 13:31:23 -0700
Subject: [C++-sig] Programmer Search @ Naughty Dog!
Message-ID: <002f01c80f6a$5b86a580$09c9a8c0@DAMLT1>

Hi Hamilton, 

My name is Candace Walker and I am the recruiter at Naughty Dog in Santa
Monica, CA. Naughty Dog is owned by Sony Computer Entertainment of America
and is developer of the Crash Bandicoot and Jak and Daxter franchises. I
wanted to reach out to you and let you Naughty Dog is active to staff up for
solid engineers. Naughty Dog has two teams right now, our Game Team and our
ICE Team. 

Our game team is working on a new original IP called, Uncharted: Drake's
Fortune that debuted at E3 of 2006. I have provided the game trailer, along
with some press about the game. 
Demo: 
www.naughtydog.com/comingsoon/index.html 
Press from: Uncharted: Drake's Fortune 
. "Naughty Dog's new franchise for the PS3 could be the biggest first-party
title of the year." -Gamespy 
. "Naughty Dog has always had a reputation for squeezing the utmost
performance and fidelity out of Sony's hardware, and it looks like that
tradition will continue with Uncharted." -Gamespot (Gamespot Preview) 

In addition to our game team, Naughty Dog is home to the ICE Team, one of
Sony's World Wide Studios central technology groups. The ICE Team focuses on
creating core graphics technologies for Sony's worldwide first party
published titles for the PLAYSTATIONR3, including low level game engine
components, graphics processing pipelines, supporting tools, and GUI
profiling and debugging tools. 

More recently the ICE Team's scope has expanded in order to build engine
components and tools for use by third party developers. The first two of
these are PLAYSTATIONREdge, a set of components that enables developers to
get better performance out of the PLAYSTATIONR3, and GCM Replay, which
provides analysis, debugging, and profiling for the RSXT. 

Please feel free to contact me directly if this is something that you or
someone you know may be interested in hearing more about. 



 

 

Candace Walker 

Recruiter

 <http://www.linkedin.com/in/candacerwalker>
http://www.linkedin.com/in/candacerwalker

 

Naughty Dog, Inc.

1601 Cloverfield Blvd. 

Suite 6000 North

Santa Monica, CA 90404

 <mailto:candace_walker at naughtydog.com> candace_walker at naughtydog.com

Phone: 310-633-9275

 <http://www.naughtydog.com/> www.naughtydog.com 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071015/225e7a52/attachment.htm>

From dave at boost-consulting.com  Mon Oct 15 21:43:54 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 15 Oct 2007 09:43:54 -1000
Subject: [C++-sig] Header error in boost/python/object/make_holder?
References: <ab0910290710101131q6497be23ke66fdf8cace8bcfa@mail.gmail.com>
Message-ID: <87bqb0p3id.fsf@grogan.peloton>


on Wed Oct 10 2007, "Michel Valstar" <Michel.Valstar-AT-imperial.ac.uk> wrote:

> I am trying to build code of a recently graduated student, who is now unavailable to us (read:
> ignores us). When running scons in the appropriate directory, a lot gets build, until I get the
> following error:
>
> /usr/include/boost/python/object/make_holder.hpp:81: error: 'allocate' is not a member of
> 'boost::python::objects::value_holder<std::pair<const std::basic_string<char, std::char_traits
> <char>, std::allocator<char> >, State> >'
> /usr/include/boost/python/object/make_holder.hpp:83: error: no matching function for call to
> 'boost::python::objects::value_holder<std::pair<const std::basic_string<char, std::char_traits
> <char>, std::allocator<char> >, State> >::value_holder(int*&)'
> /usr/include/boost/python/object/value_holder.hpp:44: note: candidates are:
> boost::python::objects::value_holder<std::pair<const std::basic_string<char, std::char_traits
> <char>, std::allocator<char> >, State> >::value_holder()
> /usr/include/boost/python/object/value_holder.hpp:44: note:                
> boost::python::objects::value_holder<std::pair<const std::basic_string<char, std::char_traits
> <char>, std::allocator<char> >, State> >::value_holder(const
> boost::python::objects::value_holder<std::pair<const std::basic_string<char, std::char_traits
> <char>, std::allocator<char> >, State> >&)
> scons: *** [trackerbase/trackerbasewrapper.os] Error 1
> scons: building terminated because of errors.
>
> This does not seem to be an error in his own code, rather an error in the boost.python code?
> This is the first time I am using boost, and, as usual, I need to get this program working
> yesterday.
>
> My machine is a recently installed Ubuntu Feisty (hence the need for re-compilation). I am using
> the latest version of Boost (boost_1_34_1) and python 2.5.1
>
> Any help would be appreciated.

I'll try.

First, please read the Note at
http://boost.org/libs/python/doc/building.html#no-install-quickstart

Also, please note that
http://boost.org/more/mailing_lists.htm#cplussig may be a better place
to ask this question.  I'm following up there.

I'm pretty sure this is an error in your student's code and the
compiler is giving you a poor error message.  Most probably

  std::pair<std::basic_string<char, std::char_traits<char>,std::allocator<char> >, State> 

is an incomplete type at the point where make_holder is being used.
However, guessing is really futile.  If my guess is wrong, please
prepare and post a minimal, complete example as described in
http://article.gmane.org/gmane.comp.python.c++/10237/match=minimal+complete+example+reproduce+jamfile

9 times out of 10, preparing a minimal example reveals the actual
problem to the person asking the question ;-)

Cheers,

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com



From thauk at novuscom.net  Mon Oct 15 23:47:31 2007
From: thauk at novuscom.net (Thomas Hauk)
Date: Mon, 15 Oct 2007 14:47:31 -0700
Subject: [C++-sig] Header error in boost/python/object/make_holder?
In-Reply-To: <87bqb0p3id.fsf@grogan.peloton>
References: <ab0910290710101131q6497be23ke66fdf8cace8bcfa@mail.gmail.com>
	<87bqb0p3id.fsf@grogan.peloton>
Message-ID: <39935981-EB2A-48B6-BCFD-9E761923A477@novuscom.net>

On Oct 15, 2007, at 12:43 PM, David Abrahams wrote:
> 9 times out of 10, preparing a minimal example reveals the actual
> problem to the person asking the question ;-)

Speaking of which, and I'm changing the subject rather selfishly, but  
I still haven't had an answer about what the bindings to:

	struct Foo
	{
		unsigned char data[4];
	};

should be... and I don't think that type of question to this  
newsgroup gets any more minimal than that. :)

T

--
Some people, when confronted with a problem, think "I know, I'll use  
regular expressions."  Now they have two problems.
   --Jamie Zawinski, in comp.lang.emacs



From Chang_Chen at sonic.com  Tue Oct 16 11:23:13 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Tue, 16 Oct 2007 17:23:13 +0800
Subject: [C++-sig] a question about boost python
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF01679716@SHAEXC01.ent.sonic.com>

Hi All

I have a set of functions which last parameters are out-parameter. i.e.

  void  function1(int  t1, float t2, string& t3) ;

I want to expose these functions in python, but such definition
(~def("function1", function1)~) is not what I want.

Because we have to write python code like this:

   strTest=''
   function1(1,2.0,strTest);

I prefer to writing python code like this

   strTest = funciton1(1,2.0)

Is there any idea to solve this question?

Thanks





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/e2bc054e/attachment.htm>

From roman.yakovenko at gmail.com  Tue Oct 16 12:21:31 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 16 Oct 2007 12:21:31 +0200
Subject: [C++-sig] access struct inside class
In-Reply-To: <47122DE1.3050801@dackbrann.net>
References: <47122DE1.3050801@dackbrann.net>
Message-ID: <7465b6170710160321h406f2605sf55a122f0b021db1@mail.gmail.com>

On 10/14/07, Simon Norberg <simon at dackbrann.net> wrote:
>
> Hello,
> im trying to get python to read and write to a struct inside a class
> in the same way i do in c++
> Customers hej;
> hej.address.contact = "The Devil";
>
> but i cant even get the code to compile, any suggestions?
>
> #include <iostream>
> #include <string>
>
> using namespace std;
> class Customers {
> private:
> struct s_address
> {
> string contact;
> string street1;
> string street2;
> string city;
> string province;
> string postalCode;
> string country;
> };
> public:
> s_address address;
> Customers(){}
> ~Customers(){}
> };
>

1. you cannot export variable, without exporting its type first: you have to
expose s_address class too, if you want to expose address mem. variable.

2. here is the code that was generated by Py++(
http://language-binding.net/), the only change I did for original code
was to change everything to
public.

#include "boost/python.hpp"

#include "1.hpp"

namespace bp = boost::python;

BOOST_PYTHON_MODULE(pyplusplus){
    { //::Customers
        typedef bp::class_< Customers > Customers_exposer_t;
        Customers_exposer_t Customers_exposer = Customers_exposer_t(
"Customers" );
        bp::scope Customers_scope( Customers_exposer );
        bp::class_< Customers::s_address >( "s_address" )
            .def_readwrite( "city", &Customers::s_address::city )
            .def_readwrite( "contact", &Customers::s_address::contact )
            .def_readwrite( "country", &Customers::s_address::country )
            .def_readwrite( "postalCode", &Customers::s_address::postalCode
)
            .def_readwrite( "province", &Customers::s_address::province )

            .def_readwrite( "street1", &Customers::s_address::street1 )
            .def_readwrite( "street2", &Customers::s_address::street2 );
        Customers_exposer.def( bp::init< >() );
        Customers_exposer.def_readonly( "address", &Customers::address );
    }
}


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/cd456629/attachment.htm>

From roman.yakovenko at gmail.com  Tue Oct 16 12:23:55 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 16 Oct 2007 12:23:55 +0200
Subject: [C++-sig] a question about boost python
In-Reply-To: <51C471E6CBEC8148A08E065690F4DBDF01679716@SHAEXC01.ent.sonic.com>
References: <51C471E6CBEC8148A08E065690F4DBDF01679716@SHAEXC01.ent.sonic.com>
Message-ID: <7465b6170710160323t6e4025d1u827042e325f31be6@mail.gmail.com>

On 10/16/07, Chang Chen <Chang_Chen at sonic.com> wrote:
>
>  Hi All
>
> I have a set of functions which last parameters are out-parameter. i.e.
>
>   void  function1(int  t1, float t2, string& t3) ;
>
> I want to expose these functions in python, but such definition (~def("
> function1", function1)~) is not what I want.
>
> Because we have to write python code like this:
>
>    strTest=''
>
>    function1(1,2.0,strTest);
>
> I prefer to writing python code like this
>
>    strTest = funciton1(1,2.0)
>
> Is there any idea to solve this question?
>
In short: to create another function, with the desired interface. In your
case you just don't have other choice - std::string is mapped to Python
string, which is immutable.

Read more here:
http://language-binding.net/pyplusplus/documentation/functions/transformation/built_in/output.html

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/b7b91725/attachment.htm>

From roman.yakovenko at gmail.com  Tue Oct 16 12:31:39 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 16 Oct 2007 12:31:39 +0200
Subject: [C++-sig] Header error in boost/python/object/make_holder?
In-Reply-To: <39935981-EB2A-48B6-BCFD-9E761923A477@novuscom.net>
References: <ab0910290710101131q6497be23ke66fdf8cace8bcfa@mail.gmail.com>
	<87bqb0p3id.fsf@grogan.peloton>
	<39935981-EB2A-48B6-BCFD-9E761923A477@novuscom.net>
Message-ID: <7465b6170710160331n7e00fc91q3725cbc52124541f@mail.gmail.com>

On 10/15/07, Thomas Hauk <thauk at novuscom.net> wrote:
>
> On Oct 15, 2007, at 12:43 PM, David Abrahams wrote:
> > 9 times out of 10, preparing a minimal example reveals the actual
> > problem to the person asking the question ;-)
>
> Speaking of which, and I'm changing the subject rather selfishly, but
> I still haven't had an answer about what the bindings to:
>
>         struct Foo
>         {
>                 unsigned char data[4];
>         };
>
> should be... and I don't think that type of question to this
> newsgroup gets any more minimal than that. :)
>


Next code was generated by Py++, may be it is not optimal but it does the
work. __array_1.pypp.hpp file is attached



#include "boost/python.hpp"

#include "__array_1.pypp.hpp"

#include "1.hpp"

namespace bp = boost::python;

struct Foo_wrapper : Foo, bp::wrapper< Foo > {

    Foo_wrapper(Foo const & arg )
    : Foo( arg )
      , bp::wrapper< Foo >(){
        // copy constructor

    }

    Foo_wrapper()
    : Foo()
      , bp::wrapper< Foo >(){
        // null constructor

    }

    static pyplusplus::containers::static_sized::array_1_t< unsigned char,
4>
    pyplusplus_data_wrapper( ::Foo & inst ){
        return pyplusplus::containers::static_sized::array_1_t< unsigned
char, 4>( inst.data );
    }

};

BOOST_PYTHON_MODULE(pyplusplus){
    { //::Foo
        typedef bp::class_< Foo_wrapper > Foo_exposer_t;
        Foo_exposer_t Foo_exposer = Foo_exposer_t( "Foo" );
        bp::scope Foo_scope( Foo_exposer );
        pyplusplus::containers::static_sized::register_array_1< unsigned
char, 4 >( "__array_1_unsigned_char_4" );
        { //Foo::data [variable], type=unsigned char[4]

            typedef pyplusplus::containers::static_sized::array_1_t<
unsigned char, 4> ( *array_wrapper_creator )( ::Foo & );

            Foo_exposer.add_property( "data"
                , bp::make_function(
array_wrapper_creator(&Foo_wrapper::pyplusplus_data_wrapper)
                                    , bp::with_custodian_and_ward_postcall<
0, 1 >() ) );
        }
    }
}


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/12a072d1/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: __array_1.pypp.hpp
Type: text/x-c++hdr
Size: 5152 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/12a072d1/attachment.hpp>

From patrick.hartling at gmail.com  Tue Oct 16 17:00:58 2007
From: patrick.hartling at gmail.com (Patrick Hartling)
Date: Tue, 16 Oct 2007 10:00:58 -0500
Subject: [C++-sig] Problems passing objects to function by const reference
Message-ID: <944f7d8e0710160800l631fc01dk936d0af20472f91f@mail.gmail.com>

In our uses of Boost.Python, we have run into a case where very simple code
fails to compile. The case that we have is that a class type has no public
destructor and instantiations of that type are passed by const reference into
a function. The Boost.Python code for exposing the function that accepts the
objects fail to compile because an attempt is made to call the non-public
destructor.

The attached code demonstrates what we are trying to do. The compile error
from GCC 4.1 when compiling against Boost 1.34.1 is as follows:

destroy_test.h: In static member function 'static void
boost::python::detail::value_destroyer<false>::execute(const volatile
T*) [with T = C]':
.../boost-1.34.1/include/boost-1_34_1/boost/python/detail/destroy.hpp:90:
  instantiated from 'void
boost::python::detail::destroy_referent_impl(void*, T& (*)()) [with T
= const C]'
.../boost-1.34.1/include/boost-1_34_1/boost/python/detail/destroy.hpp:101:
  instantiated from 'void
boost::python::detail::destroy_referent(void*, T (*)()) [with T =
const C&]'
.../boost-1.34.1/include/boost-1_34_1/boost/python/converter/rvalue_from_python_data.hpp:135:
  instantiated from
'boost::python::converter::rvalue_from_python_data<T>::~rvalue_from_python_data()
[with T = const C&]'
.../boost-1.34.1/include/boost-1_34_1/boost/python/converter/arg_from_python.hpp:108:
  instantiated from 'PyObject*
boost::python::detail::caller_arity<1u>::impl<F, Policies,
Sig>::operator()(PyObject*, PyObject*) [with F = void (*)(const C&),
Policies = boost::python::default_call_policies, Sig =
boost::mpl::vector2<void, const C&>]'
.../boost-1.34.1/include/boost-1_34_1/boost/python/object/py_function.hpp:38:
  instantiated from 'PyObject*
boost::python::objects::caller_py_function_impl<Caller>::operator()(PyObject*,
PyObject*) [with Caller = boost::python::detail::caller<void (*)(const
C&), boost::python::default_call_policies, boost::mpl::vector2<void,
const C&> >]'
generated.cpp:21:   instantiated from here
destroy_test.h:4: error: 'C::~C()' is protected
.../boost-1.34.1/include/boost-1_34_1/boost/python/detail/destroy.hpp:33:
error: within this context

I have tried to understand what Boost.Python is doing internally, and it
seems like a copy of the object is being made that is later cleaned up by
calling the destructor explicitly. I could definitely be wrong about that
assessment, though. Ultimately, what I don't understand is why anything needs
to be destroyed in this case.

 -Patirck


-- 
Patrick L. Hartling
http://www.137.org/patrick/
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: destroy_test.h
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/afb2c1f7/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: generated.cpp
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/afb2c1f7/attachment-0001.txt>

From grh at mur.at  Tue Oct 16 17:45:49 2007
From: grh at mur.at (Georg Holzmann)
Date: Tue, 16 Oct 2007 17:45:49 +0200
Subject: [C++-sig] datastructure to numpy and back
In-Reply-To: <4713AC69.4000804@mur.at>
References: <4713AC69.4000804@mur.at>
Message-ID: <4714DCAD.6050300@mur.at>

Hallo again !

I still did not get it working with boost (after quite some reading and 
searching) ... aren't there any examples out there how to do that ?

Or maybe is it better to use SWIG or directly the NumPy C-API for that ?

Thanks,
LG
Georg

Georg Holzmann schrieb:
> Hallo!
> 
> I know this was discussed quite some times, but I did find that many 
> workarounds and not really a common way how to do that with Boost.Python 
> ... so I am a little bit confused ...
> 
> My problem: I have a class with some (big) custom C++ matrix types and I 
> want to write some getters/setters for them to/from numpy arrays, where 
> I want to avoid copies of the data.
> 
> So is there a common way to do this ?
> (e.g. num_util from 
> http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/num_util_release2/Readme.html 
> ?)
> 
> Or can maybe someone link me to some docu or examples which already did 
> this ? (it must be quite common ...)
> 
> Many thanks for any hint,
> LG
> Georg
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig



From matthieu.brucher at gmail.com  Tue Oct 16 17:55:00 2007
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Tue, 16 Oct 2007 17:55:00 +0200
Subject: [C++-sig] datastructure to numpy and back
In-Reply-To: <4714DCAD.6050300@mur.at>
References: <4713AC69.4000804@mur.at> <4714DCAD.6050300@mur.at>
Message-ID: <e76aa17f0710160855x23c0ea6al20312ccd4d3315f8@mail.gmail.com>

Hi,

I've written a swig wrapper around a structure that could be used for
wrapping anything that needs to expose the __array_struct__ property (didn't
had the time to try Boost with Albert's wrappers). It is not available on
the net, but it is based on some mails that were sent here and on the Pyrex
example for the array interface.

Matthieu

2007/10/16, Georg Holzmann <grh at mur.at>:
>
> Hallo again !
>
> I still did not get it working with boost (after quite some reading and
> searching) ... aren't there any examples out there how to do that ?
>
> Or maybe is it better to use SWIG or directly the NumPy C-API for that ?
>
> Thanks,
> LG
> Georg
>
> Georg Holzmann schrieb:
> > Hallo!
> >
> > I know this was discussed quite some times, but I did find that many
> > workarounds and not really a common way how to do that with Boost.Python
> > ... so I am a little bit confused ...
> >
> > My problem: I have a class with some (big) custom C++ matrix types and I
> > want to write some getters/setters for them to/from numpy arrays, where
> > I want to avoid copies of the data.
> >
> > So is there a common way to do this ?
> > (e.g. num_util from
> >
> http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/num_util_release2/Readme.html
> > ?)
> >
> > Or can maybe someone link me to some docu or examples which already did
> > this ? (it must be quite common ...)
> >
> > Many thanks for any hint,
> > LG
> > Georg
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/1402f28a/attachment.htm>

From thauk at novuscom.net  Tue Oct 16 17:55:16 2007
From: thauk at novuscom.net (Thomas Hauk)
Date: Tue, 16 Oct 2007 08:55:16 -0700
Subject: [C++-sig] Header error in boost/python/object/make_holder?
In-Reply-To: <7465b6170710160331n7e00fc91q3725cbc52124541f@mail.gmail.com>
References: <ab0910290710101131q6497be23ke66fdf8cace8bcfa@mail.gmail.com>
	<87bqb0p3id.fsf@grogan.peloton>
	<39935981-EB2A-48B6-BCFD-9E761923A477@novuscom.net>
	<7465b6170710160331n7e00fc91q3725cbc52124541f@mail.gmail.com>
Message-ID: <5C34DA80-A480-4B0C-B801-5DC0853E350A@novuscom.net>

On Oct 16, 2007, at 3:31 AM, Roman Yakovenko wrote:
> Next code was generated by Py++, may be it is not optimal but it  
> does the work. __array_1.pypp.hpp file is attached

Thanks for the code, Roman.

I must say how surprised I am at the amount of code it takes to wrap  
a struct containing only an array of chars! Is this a limitation of  
the way C++ and Python get bound, or is this because of some  
inefficiency in Boost.Python? If the latter, I would hope that this  
becomes a new base case for improving the package.

Would it be easier to wrap this using another technology, like SWIG?

T

--
Some people, when confronted with a problem, think "I know, I'll use  
regular expressions."  Now they have two problems.
   --Jamie Zawinski, in comp.lang.emacs



From roman.yakovenko at gmail.com  Tue Oct 16 18:05:28 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 16 Oct 2007 18:05:28 +0200
Subject: [C++-sig] Header error in boost/python/object/make_holder?
In-Reply-To: <5C34DA80-A480-4B0C-B801-5DC0853E350A@novuscom.net>
References: <ab0910290710101131q6497be23ke66fdf8cace8bcfa@mail.gmail.com>
	<87bqb0p3id.fsf@grogan.peloton>
	<39935981-EB2A-48B6-BCFD-9E761923A477@novuscom.net>
	<7465b6170710160331n7e00fc91q3725cbc52124541f@mail.gmail.com>
	<5C34DA80-A480-4B0C-B801-5DC0853E350A@novuscom.net>
Message-ID: <7465b6170710160905h411712b4l722645a4e496b3e6@mail.gmail.com>

On 10/16/07, Thomas Hauk <thauk at novuscom.net> wrote:
>
> On Oct 16, 2007, at 3:31 AM, Roman Yakovenko wrote:
> > Next code was generated by Py++, may be it is not optimal but it
> > does the work. __array_1.pypp.hpp file is attached
>
> Thanks for the code, Roman.
>
> I must say how surprised I am at the amount of code it takes to wrap
> a struct containing only an array of chars! Is this a limitation of
> the way C++ and Python get bound,


I warned you about code size. First of all this is how Py++ generates the
code in this case. Second if you will use indexing suite V2, than the amount
of code will drop.

or is this because of some
> inefficiency in Boost.Python?


I didn't tested the code for efficiency, but it doesn't copy the array, just
provides access to the items. So I don't expect problems with it.


If the latter, I would hope that this
> becomes a new base case for improving the package.
>
> Would it be easier to wrap this using another technology, like SWIG?
>

Don't know never used it.


-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/1ffb3893/attachment.htm>

From matthew.scouten at gmail.com  Tue Oct 16 19:31:40 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Tue, 16 Oct 2007 12:31:40 -0500
Subject: [C++-sig] passing a vector of structs
Message-ID: <3dc9bcf00710161031w5d0c8213qf41095ae1f0f2551@mail.gmail.com>

I have a vector of 'struct foo' in c++ that needs to be passed to a
function python.

'struct foo' is an extremely simple structure containing 3 ints and a
trivial ctor.

so I create a foo_wrapper:

struct foo_wrapper : foo , wrapper<foo>
{
	foo_wrapper(const foo& f);
	foo_wrapper();
};

and export it:

class_<foo_wrapper>("foo")
		.def(init<>())
		.def_readwrite("iE", &foo::iE )
		.def_readwrite("bA", &foo::bA)
		.def_readwrite("bL", &foo::bL)
		;

next I export a vector of foo_wrappers:

class_< std::vector<foo_wrapper> >("VectorOfFoo")
		.def(vector_indexing_suite< std::vector<foo_wrapper>, true >() );

I pass the vector<foo> to a python function:

Traceback (most recent call last):
  File "C:\redacted", line 23, in <module>
    pc.RequestExchangeList()
TypeError: No to_python (by-value) converter found for C++ type: class
std::vector<struct foo ,class std::allocator<struct foo > >

I try to copy the elements to a new vector<foo_wrapper>, it seems to
work fine, untill I try to access an element of that vector and get:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 7, in OnExchangeList
TypeError: No to_python (by-value) converter found for C++ type:
struct TTExchangeStruct_wrapper

So what am I doing wrong? How am I supposed to pass a vector of
structs from c++ to python?


From roman.yakovenko at gmail.com  Tue Oct 16 20:48:45 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Tue, 16 Oct 2007 20:48:45 +0200
Subject: [C++-sig] passing a vector of structs
In-Reply-To: <3dc9bcf00710161031w5d0c8213qf41095ae1f0f2551@mail.gmail.com>
References: <3dc9bcf00710161031w5d0c8213qf41095ae1f0f2551@mail.gmail.com>
Message-ID: <7465b6170710161148w6823d529k7bffad2a9b4693ce@mail.gmail.com>

On 10/16/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
>
> I have a vector of 'struct foo' in c++ that needs to be passed to a
> function python.
>
> 'struct foo' is an extremely simple structure containing 3 ints and a
> trivial ctor.
>
> so I create a foo_wrapper:
>
> struct foo_wrapper : foo , wrapper<foo>
> {
>         foo_wrapper(const foo& f);
>         foo_wrapper();
> };


In this case you don't need to create wrapper

and export it:
>
> class_<foo_wrapper>("foo")
>                 .def(init<>())
>                 .def_readwrite("iE", &foo::iE )
>                 .def_readwrite("bA", &foo::bA)
>                 .def_readwrite("bL", &foo::bL)
>                 ;
>
> next I export a vector of foo_wrappers:
>
> class_< std::vector<foo_wrapper> >("VectorOfFoo")
>                 .def(vector_indexing_suite< std::vector<foo_wrapper>, true
> >() );
>
> I pass the vector<foo> to a python function:
>
> Traceback (most recent call last):
>   File "C:\redacted", line 23, in <module>
>     pc.RequestExchangeList()
> TypeError: No to_python (by-value) converter found for C++ type: class
> std::vector<struct foo ,class std::allocator<struct foo > >
>
> I try to copy the elements to a new vector<foo_wrapper>, it seems to
> work fine, untill I try to access an element of that vector and get:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File "<stdin>", line 7, in OnExchangeList
> TypeError: No to_python (by-value) converter found for C++ type:
> struct TTExchangeStruct_wrapper
>
> So what am I doing wrong? How am I supposed to pass a vector of
> structs from c++ to python?



You need to post small and complete example.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/bec249c4/attachment.htm>

From hans_meine at gmx.net  Tue Oct 16 22:37:04 2007
From: hans_meine at gmx.net (Hans Meine)
Date: Tue, 16 Oct 2007 22:37:04 +0200
Subject: [C++-sig] How about class_<...>.enable_copy() ?
In-Reply-To: <200707232257.05883.meine@informatik.uni-hamburg.de>
References: <200706081706.37240.meine@informatik.uni-hamburg.de>
	<200707231922.15574.meine@informatik.uni-hamburg.de>
	<200707232257.05883.meine@informatik.uni-hamburg.de>
Message-ID: <200710162237.08845.hans_meine@gmx.net>

Hi again!

Anyone cares to comment on my older posting?

On Montag 23 Juli 2007, Hans Meine wrote:
> On Montag 23 Juli 2007, Hans Meine wrote:
> > foo = Foo(other) # copy constructor as in C++ (1)
> > foo = copy.copy(other) # pythonic copying (2)
> > foo = copy.deepcopy(other) # deepcopying (3)
> >
> > (1) is easy to do by exporting the copy constructor
> >
> > (2) could be implemented by setting __copy__ to the above copy
> > constructor, but one would assume __dict__ to be (non-deep)copied, too.
> >
> > (3) definitely needs its own implementation, which needs to call deepcopy
> > recursively.
> >
> > I would still believe that an .enable_copy() would be very handy.  It
> > could possibly be controlled by an enum arg which of the above methods
> > should be exported.

Here is the updated, perfectly working code for (2) and (3), which I propose 
to be included in BPL.  Also, some things appear hackish to me -- I would be 
glad to get feedback on how to do this in a more "standard" way:

a) wrap a pointer to a newly created object in a bp::object -- there must be a 
shorter/better way than
  bp::detail::new_reference(typename manage_new_object::apply<T *>::type()(p)
no?

b) shouldn't a 64bit-safe version of
   inline id(object) { return (int)(copyable.ptr()); }
be added to boost::python?

c) is my solution of accessing copy/deepcopy improvable?  (I guess not - it 
looks already quite OK to me.)

d) would you use the visitor pattern, i.e. class_<T>::enable_copy for the 
above code?  AFAICS, there are two small advantages:
* both functions can be def'd in one line
* the type does not have to be repeated, but is fetched from the class_<> 
template instance

> Maybe someone is interested in turning my code into a standard BPL
> facility? (BTW: Is __dict__ created on demand?  Then one should probably
> add a guard to both my __copy__ and __deepcopy__.)

I'd like to get an authoritive answer on this - do BPL-exported objects always 
have a __dict__, or is it only created on demand/access?

> On Montag 23 Juli 2007, Hans Meine wrote:
> > Probably I should use PyImport_ImportModule or the new
> > boost::python::import.
>
> Speaking of the latter, shouldn't the python:: prefix be removed from the
> code within boost_1_34_0/libs/python/src/import.cpp ?  I recently had
> syntax errors from GCC-4.1.2 within kdelibs-3.5.7 with a similar case:
> > namespace KABC {
> >
> > class KABC_EXPORT Field
> > {
> > public:
> >   virtual QString value( const KABC::Addressee & );
>
> Here, I had to remove KABC:: in order to let it compile.  (Which I found
> surprising, but anyways.)

This is not very much related, but interesting.

-- 
Ciao, /  /                                                    .o.
     /--/                                                     ..o
    /  / ANS                                                  ooo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071016/c8d7f284/attachment.pgp>

From Chang_Chen at sonic.com  Wed Oct 17 05:38:23 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Wed, 17 Oct 2007 11:38:23 +0800
Subject: [C++-sig]  a question about boost python
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF01679895@SHAEXC01.ent.sonic.com>

I know we need define a wrapper, but our function signature is quite
regular. i.e. only the last parameter is output parameter, so I want to
write a template, so that such wrapper can be easily written.

Our function definition looks like:

    void function1(int  t1, float t2, string& t3) ;

I want to define a template:

template <typename Return, typename T1, typename T2, typename TOUT> 
 struct FunctionConvert< Return (*) ( T1 , T2, TOUT& ) >
{
	typedef functionT	Return (*) ( T1 , T2, TOUT& );
	static TOUT run(functionT P,const T1& t1, const T2& t2)
	{
		TOUT out;
		P(t1,t2,out);
		return out;
	}
}

But I do not know how to integrate it with boost python :(.

- Chang



From meine at informatik.uni-hamburg.de  Wed Oct 17 10:25:02 2007
From: meine at informatik.uni-hamburg.de (Hans Meine)
Date: Wed, 17 Oct 2007 10:25:02 +0200
Subject: [C++-sig] How about class_<...>.enable_copy() ?
In-Reply-To: <200710162237.08845.hans_meine@gmx.net>
References: <200706081706.37240.meine@informatik.uni-hamburg.de>
	<200707232257.05883.meine@informatik.uni-hamburg.de>
	<200710162237.08845.hans_meine@gmx.net>
Message-ID: <200710171025.02589.meine@informatik.uni-hamburg.de>

Am Dienstag, 16. Oktober 2007 22:37:04 schrieb Hans Meine:
> Here is the updated, perfectly working code for (2) and (3), which I
> propose to be included in BPL.  Also, some things appear hackish to me -- I
> would be glad to get feedback on how to do this in a more "standard" way:

Ooops - of course I indented to attach something..

-- 
Ciao, /  /
     /--/
    /  / ANS
-------------- next part --------------
A non-text attachment was scrubbed...
Name: foo_copying.cxx
Type: text/x-c++src
Size: 1227 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071017/58ba2df8/attachment.cxx>

From Chang_Chen at sonic.com  Wed Oct 17 10:30:01 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Wed, 17 Oct 2007 16:30:01 +0800
Subject: [C++-sig] How about class_<...>.enable_copy() ?
In-Reply-To: <200710171025.02589.meine@informatik.uni-hamburg.de>
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF016799A7@SHAEXC01.ent.sonic.com>


Test reply
Nothing...
-----Original Message-----
From: c++-sig-bounces at python.org [mailto:c++-sig-bounces at python.org] On
Behalf Of Hans Meine
Sent: Wednesday, October 17, 2007 4:25 PM
To: Development of Python/C++ integration
Subject: Re: [C++-sig] How about class_<...>.enable_copy() ?

Am Dienstag, 16. Oktober 2007 22:37:04 schrieb Hans Meine:
> Here is the updated, perfectly working code for (2) and (3), which I
> propose to be included in BPL.  Also, some things appear hackish to me
-- I
> would be glad to get feedback on how to do this in a more "standard"
way:

Ooops - of course I indented to attach something..

-- 
Ciao, /  /
     /--/
    /  / ANS


From matthew.scouten at gmail.com  Wed Oct 17 16:31:26 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Wed, 17 Oct 2007 09:31:26 -0500
Subject: [C++-sig] passing a vector of structs
In-Reply-To: <7465b6170710161148w6823d529k7bffad2a9b4693ce@mail.gmail.com>
References: <3dc9bcf00710161031w5d0c8213qf41095ae1f0f2551@mail.gmail.com>
	<7465b6170710161148w6823d529k7bffad2a9b4693ce@mail.gmail.com>
Message-ID: <3dc9bcf00710170731l4d0e578dw6ac9b4b79ca891aa@mail.gmail.com>

On 10/16/07, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> On 10/16/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
> > I have a vector of 'struct foo' in c++ that needs to be passed to a
> > function python.
> >
> > 'struct foo' is an extremely simple structure containing 3 ints and a
> > trivial ctor.
> >
> > so I create a foo_wrapper:
> >
> > struct foo_wrapper : foo , wrapper<foo>
> > {
> >         foo_wrapper(const foo& f);
> >         foo_wrapper();
> > };
>
> In this case you don't need to create wrapper
>
> > and export it:
> >
> > class_<foo_wrapper>("foo")
> >                 .def(init<>())
> >                 .def_readwrite("iE", &foo::iE )
> >                 .def_readwrite("bA", &foo::bA)
> >                 .def_readwrite("bL", &foo::bL)
> >                 ;
> >
> > next I export a vector of foo_wrappers:
> >
> > class_< std::vector<foo_wrapper> >("VectorOfFoo")
> >                 .def(vector_indexing_suite<
> std::vector<foo_wrapper>, true >() );
> >
> > I pass the vector<foo> to a python function:
> >
> > Traceback (most recent call last):
> >   File "C:\redacted", line 23, in <module>
> >     pc.RequestExchangeList()
> > TypeError: No to_python (by-value) converter found for C++ type: class
> > std::vector<struct foo ,class std::allocator<struct foo > >
> >
> > I try to copy the elements to a new vector<foo_wrapper>, it seems to
> > work fine, untill I try to access an element of that vector and get:
> >
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in <module>
> >   File "<stdin>", line 7, in OnExchangeList
> > TypeError: No to_python (by-value) converter found for C++ type:
> > struct TTExchangeStruct_wrapper
> >
> > So what am I doing wrong? How am I supposed to pass a vector of
> > structs from c++ to python?
>
>
> You need to post small and complete example.
>
>  --
> Roman Yakovenko

Right, OK:
TTExchangeStruct_wrapper is my foo_wrapper. I forgot to change that
when sanitizing and simplifying me example. I split out the problem
code and made a stand alone compilable example:

c++ & boost::python:

struct foo
{
	foo():one(1),two(2),three(3) {}

	int one;
	int two;
	int three;
};

//The std::vector Indexing Suite seems to want this
//I Don't know why.
bool operator==(foo lhs ,foo rhs)
{
	return &lhs == &rhs;
}

struct foo_wrapper : foo , wrapper<foo>
{
        foo_wrapper(const foo& f);
        foo_wrapper();
};

foo_wrapper::foo_wrapper(const foo& exch): foo(exch), wrapper<foo>(){}

foo_wrapper::foo_wrapper(): foo(), wrapper<foo>(){}

class busyBox
{
public:
	int callout();
	virtual void callback(std::vector<foo> vf) {};
};

int busyBox::callout()
{
	std::vector<foo> vf(10);
	callback(vf);
	return 0;
}

class busyBox_wrapper: public busyBox, public wrapper<busyBox>
{
public:
	busyBox_wrapper();
	busyBox_wrapper(const busyBox &);
	int callout();
	virtual void callback(std::vector<foo> vf);
	void default_callback(std::vector<foo> vf);
};

busyBox_wrapper::busyBox_wrapper() : busyBox(), wrapper<busyBox>(){}

busyBox_wrapper::busyBox_wrapper(const busyBox &bb ) : busyBox(bb),
wrapper<busyBox>(){}

int busyBox_wrapper::callout()
{
	return busyBox::callout();
}

void busyBox_wrapper::callback (std::vector<foo> vf)
{
	if (override callbackPy = this->get_override("callback"))
	{

		std::vector<foo_wrapper> vfw;
		for (unsigned int i = 0; i < vf.size(); ++i)
			vfw.push_back(vf[i]);

		//callbackPy(vf);
		callbackPy(vfw);

		return;
	}
	return busyBox::callback(vf);
}

void busyBox_wrapper::default_callback(std::vector<foo> vf)
{
	return busyBox::callback(vf);
}



BOOST_PYTHON_MODULE(BusyBox)
{

	class_<busyBox_wrapper>("busyBox")
		.def("callback",&busyBox_wrapper::default_callback)
		.def("callout", &busyBox_wrapper::callout);

	class_<foo_wrapper>("foo")
		.def(init<>())
        .def_readwrite("one", &foo::one )
        .def_readwrite("two", &foo::two)
        .def_readwrite("three", &foo::three)
        ;

	class_< std::vector<foo_wrapper> >("std::vectorOfFoo")
		.def(vector_indexing_suite< std::vector<foo_wrapper>, true>());
}

and here is the python code to invoke the error:

import BusyBox

class subBusybox(BusyBox.busyBox):
	def callback(self, vf):
		print [x for x in vf]
		

sbb = subBusybox()
sbb.callout()

and here it the error itself(it fails while invoking the 'print [x for
x in vf]' line ):

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 3, in callback
TypeError: No to_python (by-value) converter found for C++ type:
struct foo_wrapper



To reiterate my question: I need to pass a vector of structs from a
c++ function to a python function. What am I doing wrong?


From matthew.scouten at gmail.com  Wed Oct 17 22:53:31 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Wed, 17 Oct 2007 15:53:31 -0500
Subject: [C++-sig] passing a vector of structs
In-Reply-To: <3dc9bcf00710170731l4d0e578dw6ac9b4b79ca891aa@mail.gmail.com>
References: <3dc9bcf00710161031w5d0c8213qf41095ae1f0f2551@mail.gmail.com>
	<7465b6170710161148w6823d529k7bffad2a9b4693ce@mail.gmail.com>
	<3dc9bcf00710170731l4d0e578dw6ac9b4b79ca891aa@mail.gmail.com>
Message-ID: <3dc9bcf00710171353mbdfd3b7lcc04f93e3cf493bf@mail.gmail.com>

On 10/17/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
> On 10/16/07, Roman Yakovenko <roman.yakovenko at gmail.com> wrote:
> > On 10/16/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
> > > I have a vector of 'struct foo' in c++ that needs to be passed to a
> > > function python.
> > >
> > > 'struct foo' is an extremely simple structure containing 3 ints and a
> > > trivial ctor.
> > >
> > > so I create a foo_wrapper:
> > >
> > > struct foo_wrapper : foo , wrapper<foo>
> > > {
> > >         foo_wrapper(const foo& f);
> > >         foo_wrapper();
> > > };
> >
> > In this case you don't need to create wrapper
> >
> > > and export it:
> > >
> > > class_<foo_wrapper>("foo")
> > >                 .def(init<>())
> > >                 .def_readwrite("iE", &foo::iE )
> > >                 .def_readwrite("bA", &foo::bA)
> > >                 .def_readwrite("bL", &foo::bL)
> > >                 ;
> > >
> > > next I export a vector of foo_wrappers:
> > >
> > > class_< std::vector<foo_wrapper> >("VectorOfFoo")
> > >                 .def(vector_indexing_suite<
> > std::vector<foo_wrapper>, true >() );
> > >
> > > I pass the vector<foo> to a python function:
> > >
> > > Traceback (most recent call last):
> > >   File "C:\redacted", line 23, in <module>
> > >     pc.RequestExchangeList()
> > > TypeError: No to_python (by-value) converter found for C++ type: class
> > > std::vector<struct foo ,class std::allocator<struct foo > >
> > >
> > > I try to copy the elements to a new vector<foo_wrapper>, it seems to
> > > work fine, untill I try to access an element of that vector and get:
> > >
> > > Traceback (most recent call last):
> > >   File "<stdin>", line 1, in <module>
> > >   File "<stdin>", line 7, in OnExchangeList
> > > TypeError: No to_python (by-value) converter found for C++ type:
> > > struct TTExchangeStruct_wrapper
> > >
> > > So what am I doing wrong? How am I supposed to pass a vector of
> > > structs from c++ to python?
> >
> >
> > You need to post small and complete example.
> >
> >  --
> > Roman Yakovenko
>
> Right, OK:
> TTExchangeStruct_wrapper is my foo_wrapper. I forgot to change that
> when sanitizing and simplifying me example. I split out the problem
> code and made a stand alone compilable example:
>
> c++ & boost::python:
>
> struct foo
> {
>         foo():one(1),two(2),three(3) {}
>
>         int one;
>         int two;
>         int three;
> };
>
> //The std::vector Indexing Suite seems to want this
> //I Don't know why.
> bool operator==(foo lhs ,foo rhs)
> {
>         return &lhs == &rhs;
> }
>
> struct foo_wrapper : foo , wrapper<foo>
> {
>         foo_wrapper(const foo& f);
>         foo_wrapper();
> };
>
> foo_wrapper::foo_wrapper(const foo& exch): foo(exch), wrapper<foo>(){}
>
> foo_wrapper::foo_wrapper(): foo(), wrapper<foo>(){}
>
> class busyBox
> {
> public:
>         int callout();
>         virtual void callback(std::vector<foo> vf) {};
> };
>
> int busyBox::callout()
> {
>         std::vector<foo> vf(10);
>         callback(vf);
>         return 0;
> }
>
> class busyBox_wrapper: public busyBox, public wrapper<busyBox>
> {
> public:
>         busyBox_wrapper();
>         busyBox_wrapper(const busyBox &);
>         int callout();
>         virtual void callback(std::vector<foo> vf);
>         void default_callback(std::vector<foo> vf);
> };
>
> busyBox_wrapper::busyBox_wrapper() : busyBox(), wrapper<busyBox>(){}
>
> busyBox_wrapper::busyBox_wrapper(const busyBox &bb ) : busyBox(bb),
> wrapper<busyBox>(){}
>
> int busyBox_wrapper::callout()
> {
>         return busyBox::callout();
> }
>
> void busyBox_wrapper::callback (std::vector<foo> vf)
> {
>         if (override callbackPy = this->get_override("callback"))
>         {
>
>                 std::vector<foo_wrapper> vfw;
>                 for (unsigned int i = 0; i < vf.size(); ++i)
>                         vfw.push_back(vf[i]);
>
>                 //callbackPy(vf);
>                 callbackPy(vfw);
>
>                 return;
>         }
>         return busyBox::callback(vf);
> }
>
> void busyBox_wrapper::default_callback(std::vector<foo> vf)
> {
>         return busyBox::callback(vf);
> }
>
>
>
> BOOST_PYTHON_MODULE(BusyBox)
> {
>
>         class_<busyBox_wrapper>("busyBox")
>                 .def("callback",&busyBox_wrapper::default_callback)
>                 .def("callout", &busyBox_wrapper::callout);
>
>         class_<foo_wrapper>("foo")
>                 .def(init<>())
>         .def_readwrite("one", &foo::one )
>         .def_readwrite("two", &foo::two)
>         .def_readwrite("three", &foo::three)
>         ;
>
>         class_< std::vector<foo_wrapper> >("std::vectorOfFoo")
>                 .def(vector_indexing_suite< std::vector<foo_wrapper>, true>());
> }
>
> and here is the python code to invoke the error:
>
> import BusyBox
>
> class subBusybox(BusyBox.busyBox):
>         def callback(self, vf):
>                 print [x for x in vf]
>
>
> sbb = subBusybox()
> sbb.callout()
>
> and here it the error itself(it fails while invoking the 'print [x for
> x in vf]' line ):
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in <module>
>   File "<stdin>", line 3, in callback
> TypeError: No to_python (by-value) converter found for C++ type:
> struct foo_wrapper
>
>
>
> To reiterate my question: I need to pass a vector of structs from a
> c++ function to a python function. What am I doing wrong?
>

UPDATE: Well, with no reply from the list, I spent all day trying to
find my error. It turns out that I was temporarily stupid.

If the above lines are changed from:

class_< std::vector<foo_wrapper> >("std::vectorOfFoo")
                 .def(vector_indexing_suite< std::vector<foo_wrapper>, true>());

to:

class_< std::vector<foo> >("std::vectorOfFoo")
                 .def(vector_indexing_suite< std::vector<foo>, true>());

it fixes everything.


From matthew.scouten at gmail.com  Thu Oct 18 23:52:54 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Thu, 18 Oct 2007 16:52:54 -0500
Subject: [C++-sig] I can't get None
Message-ID: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>

I want to return None from a c++ function.
I have googled for how to do this with no good results.
How do I get an object with None in c++?


From nat at lindenlab.com  Fri Oct 19 00:01:30 2007
From: nat at lindenlab.com (Nat Goodspeed)
Date: Thu, 18 Oct 2007 18:01:30 -0400
Subject: [C++-sig] I can't get None
In-Reply-To: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
References: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
Message-ID: <4717D7BA.50109@lindenlab.com>

Matthew Scouten wrote:

> I want to return None from a c++ function.
> I have googled for how to do this with no good results.
> How do I get an object with None in c++?

That baffled me for a long time too. Try just object().


From amohr at pixar.com  Fri Oct 19 00:04:23 2007
From: amohr at pixar.com (Alex Mohr)
Date: Thu, 18 Oct 2007 15:04:23 -0700
Subject: [C++-sig] I can't get None
In-Reply-To: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
References: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
Message-ID: <4717D867.3040201@pixar.com>

> I want to return None from a c++ function.
> I have googled for how to do this with no good results.
> How do I get an object with None in c++?

boost::python::object none;

Alex



From seefeld at sympatico.ca  Fri Oct 19 00:06:37 2007
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Thu, 18 Oct 2007 18:06:37 -0400
Subject: [C++-sig] I can't get None
In-Reply-To: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
References: <3dc9bcf00710181452g252a364cw637141b093570b1@mail.gmail.com>
Message-ID: <4717D8ED.2020902@sympatico.ca>

Matthew Scouten wrote:
> I want to return None from a c++ function.
> I have googled for how to do this with no good results.
> How do I get an object with None in c++?

return boost::python::object();

HTH,
		Stefan

-- 

      ...ich hab' noch einen Koffer in Berlin...


From matthew.scouten at gmail.com  Fri Oct 19 18:38:09 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Fri, 19 Oct 2007 11:38:09 -0500
Subject: [C++-sig] stuffing a PyObject into a object
Message-ID: <3dc9bcf00710190938h4e641d9bwb7118c20c1cb707f@mail.gmail.com>

Right, so....
I've got a void* that I happen to know is really a PyObject*. (how do
i know that you ask? It was handed to me by python and tunneled
through a 3rd party c++ library that promised not to  mess with it.) I
need to make it back into a boost::python::object so I can hand it
back to python.

I googled for how to do this and came across:
http://osdir.com/ml/python.c++/2002-12/msg00286.html

where I found the following recommendation:

boost::python::handle<> h(boost::python::allow_null(PyIter_Next(x)));
if (!h.get()) break; // end of iteration
boost::python::object o(h);

I can understand this except for a couple of things,

What is the PyIter_Next there for? demonstration purposes? or is it an
essential part of the conversion? Same question for that whole second
line.

will 'allow_null' make a 'None' out of a null pointer? or do some
other crazy thing?

more to the point, will this code work:

object foo(void* callbackID == NULL)
{
	boost::python::handle<> h(boost::python::allow_null((PyObject*)callbackID))
	return boost::python::object o(h);
}

I am trying to understand what is going on, and not just fall into
guess and check programming.


From lloyd at fusion.net.nz  Sat Oct 20 01:43:44 2007
From: lloyd at fusion.net.nz (Lloyd Weehuizen)
Date: Sat, 20 Oct 2007 12:43:44 +1300
Subject: [C++-sig] stuffing a PyObject into a object
In-Reply-To: <3dc9bcf00710190938h4e641d9bwb7118c20c1cb707f@mail.gmail.com>
References: <3dc9bcf00710190938h4e641d9bwb7118c20c1cb707f@mail.gmail.com>
Message-ID: <47194130.9090609@fusion.net.nz>

Hi Matthew

I'm not sure why the code you found suggests a PyIter_Next, I've never 
used that myself. I'm guessing it was pulled from an iteration example 
at one point and lost its context.

I do my conversions like this:
python::object(python::handle<>(python::borrowed(x)))

 From what I can tell, allow_null does not return a python None, it 
simply allows null to pass through the system. After looking through the 
boost source, it looks like handle converts the NULL to a None when 
necessary (See handle.hpp:256 (get_managed_object)).

Cheers,
Lloyd

Matthew Scouten wrote:
> Right, so....
> I've got a void* that I happen to know is really a PyObject*. (how do
> i know that you ask? It was handed to me by python and tunneled
> through a 3rd party c++ library that promised not to  mess with it.) I
> need to make it back into a boost::python::object so I can hand it
> back to python.
> 
> I googled for how to do this and came across:
> http://osdir.com/ml/python.c++/2002-12/msg00286.html
> 
> where I found the following recommendation:
> 
> boost::python::handle<> h(boost::python::allow_null(PyIter_Next(x)));
> if (!h.get()) break; // end of iteration
> boost::python::object o(h);
> 
> I can understand this except for a couple of things,
> 
> What is the PyIter_Next there for? demonstration purposes? or is it an
> essential part of the conversion? Same question for that whole second
> line.
> 
> will 'allow_null' make a 'None' out of a null pointer? or do some
> other crazy thing?
> 
> more to the point, will this code work:
> 
> object foo(void* callbackID == NULL)
> {
> 	boost::python::handle<> h(boost::python::allow_null((PyObject*)callbackID))
> 	return boost::python::object o(h);
> }
> 
> I am trying to understand what is going on, and not just fall into
> guess and check programming.
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig


From tsandler at seas.upenn.edu  Sun Oct 21 21:26:28 2007
From: tsandler at seas.upenn.edu (ted sandler)
Date: Sun, 21 Oct 2007 15:26:28 -0400
Subject: [C++-sig] can't get Boost.Python tutorial's "Hello World" example
 to work (on linux)
Message-ID: <ffg955$7f8$1@ger.gmane.org>

Hi all,

I am running linux, and although the Boost.Python "Quickstart" example
works fine, I am not able to get the Boost.Python "Hello World" example
working.

The quickstart example is located here
http://www.boost.org/libs/python/doc/building.html.

The hello-world example is located here:
http://www.boost.org/libs/python/doc/tutorial/doc/html/python/hello.html.

There are problems with the documentation on the hello-world web page
(I guess it's out of date).  The example files in the Boost-distribution
are different from those described in the "hello-world" tutorial.
Specifically, the Boost distribution contains a "Jamroot" file while the
web page discusses both a "Jamfile" and a "Jamrules" file (whose
hyperlink is broken).

Anyway, when I cd to the "hello-world" example directory and type bjam
at the prompt, bjam fails with the error message I've listed below.  Any
help getting me up and running here would be great!

Thanks a bunch,
-Ted


Jamroot:17: in modules.load
rule python-extension unknown in module
Jamfile</home/ted/boost_1_34_1/libs/python/example/tutorial>.
/home/ted/boost_1_34_1/tools/build/v2/build/project.jam:312: in load-jamfile
/home/ted/boost_1_34_1/tools/build/v2/build/project.jam:68: in load
/home/ted/boost_1_34_1/tools/build/v2/build/project.jam:170: in project.find
/home/ted/boost_1_34_1/tools/build/v2/build-system.jam:237: in load
/home/ted/boost_1_34_1/libs/python/example/../../../tools/build/v2/kernel/modules.jam:261:
in import
/home/ted/boost_1_34_1/libs/python/example/../../../tools/build/v2/kernel/bootstrap.jam:132:
in boost-build
/home/ted/boost_1_34_1/libs/python/example/boost-build.jam:7: in module
scope



From Chang_Chen at sonic.com  Mon Oct 22 09:05:17 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Mon, 22 Oct 2007 15:05:17 +0800
Subject: [C++-sig] To Roman Yakovenko - a question about boost python
In-Reply-To: <51C471E6CBEC8148A08E065690F4DBDF01679895@SHAEXC01.ent.sonic.com>
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF016CC910@SHAEXC01.ent.sonic.com>

Hi Roman Yakovenko

Do you think what I think on out parameter is ok or not?

Any suggestion?

- Chang

-----Original Message-----
From: c++-sig-bounces+chang_chen=sonic.com at python.org
[mailto:c++-sig-bounces+chang_chen=sonic.com at python.org] On Behalf Of
Chang Chen
Sent: Wednesday, October 17, 2007 11:38 AM
To: c++-sig at python.org
Subject: [C++-sig] a question about boost python

I know we need define a wrapper, but our function signature is quite
regular. i.e. only the last parameter is output parameter, so I want to
write a template, so that such wrapper can be easily written.

Our function definition looks like:

    void function1(int  t1, float t2, string& t3) ;

I want to define a template:

template <typename Return, typename T1, typename T2, typename TOUT> 
 struct FunctionConvert< Return (*) ( T1 , T2, TOUT& ) >
{
	typedef functionT	Return (*) ( T1 , T2, TOUT& );
	static TOUT run(functionT P,const T1& t1, const T2& t2)
	{
		TOUT out;
		P(t1,t2,out);
		return out;
	}
}

But I do not know how to integrate it with boost python :(.

- Chang

_______________________________________________
C++-sig mailing list
C++-sig at python.org
http://mail.python.org/mailman/listinfo/c++-sig


From Chang_Chen at sonic.com  Mon Oct 22 13:08:14 2007
From: Chang_Chen at sonic.com (Chang Chen)
Date: Mon, 22 Oct 2007 19:08:14 +0800
Subject: [C++-sig] Wrapping out parameter for python access
Message-ID: <51C471E6CBEC8148A08E065690F4DBDF016CCA1D@SHAEXC01.ent.sonic.com>

Hi all

This is a classical issue in boost python. I.e. removeing an argument
from the function definition and adding the "returned", by the original
function, value to the return statement of the function-wrapper

For example :

The original function is:

inline void hello_world( std::string& hw ){
    hw = "hello world!";
}

And the wrapper is:

inline std::string hello_world_wrapper()
{
	std:string temp;
	::hello_world(temp); 
	Return temp;
}

Py++ can help us to generate such wrapper, see
http://language-binding.net/pyplusplus/documentation/functions/transform
ation/built_in/output.html 

But in this thread
(http://mail.python.org/pipermail/c++-sig/2005-June/009097.html ), Dave
provided a start to write a wrapping template.

In My case, only last parameter is out parameter and whether function
parameter count does not matter. The function signature looks like:

Return function (T1 t1, T2 t2, TOUT& out); 

How to write a template for such function and integrate into def?

Any suggestion?

-Chang


From grh at mur.at  Mon Oct 22 18:03:38 2007
From: grh at mur.at (Georg Holzmann)
Date: Mon, 22 Oct 2007 18:03:38 +0200
Subject: [C++-sig] datastructure to numpy and back
In-Reply-To: <e76aa17f0710160855x23c0ea6al20312ccd4d3315f8@mail.gmail.com>
References: <4713AC69.4000804@mur.at> <4714DCAD.6050300@mur.at>
	<e76aa17f0710160855x23c0ea6al20312ccd4d3315f8@mail.gmail.com>
Message-ID: <471CC9DA.2010907@mur.at>

Hallo!

> I've written a swig wrapper around a structure that could be used for 
> wrapping anything that needs to expose the __array_struct__ property 
> (didn't had the time to try Boost with Albert's wrappers). It is not 
> available on the net, but it is based on some mails that were sent here 
> and on the Pyrex example for the array interface.

Thanks, I looked now through quite some SWIG examples and I was able to 
get a matrix/array from numpy into C++ (with the very good documentated 
examples in numpy-svn/numpy/doc/swig).

However, I did not manage to get an array out of C++ to python without 
copying so far.
In the examples cited above it's only shown how to generate a new one 
dimensional array and return that.

Do you maybe have some links (from the mails here and on) where these 
features are explained ? Or modules which are examples for such use-cases ?

Thanks for any hint,
LG
Georg


From matthew.scouten at gmail.com  Mon Oct 22 19:56:13 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Mon, 22 Oct 2007 12:56:13 -0500
Subject: [C++-sig] Comparing to None on the c++ side
Message-ID: <3dc9bcf00710221056n4ca21944sea68faaa38960b1@mail.gmail.com>

Right, so...

Python just called a c++ function and passed an object of some sort. I
want to know if it happens to be a None, so I can take special
precautions to avoid disaster.

I am trying to use something like:

void foo(object o)
{
	if( object() == o )
	{
		//Make 3rdparty c++ lib call that prevents
		//Sky from falling
	}
	else
	{
		//tell 3rdparty c++ lib to
		//try to use o to prop up sky...
	}
}

This does not seem to be working like I expected (the sky falls anyway)

So me question:

is object() == o a valid and reliable way to test for None?
If not, how should I be doing it?

It is possible that the problem is in 3rdparty c++ lib or my
understanding thereof, but I am trying to eliminate the other
possibilities.


From shrikantvc at gmail.com  Thu Oct 25 08:52:24 2007
From: shrikantvc at gmail.com (Shrikant Chikhalkar)
Date: Thu, 25 Oct 2007 12:22:24 +0530
Subject: [C++-sig] failed in PyImport_Import()
Message-ID: <87de5f460710242352w5b57dc4djbcfba96ccf0f3578@mail.gmail.com>

Hi,

i am tring to call python function from 'C' program. i get NULL value return
by PyImport_Import()

following is the code. code is copied from
http://www.codeproject.com/cpp/embedpython_1.asp

following file is saved as

call_function.c

/************************************************************************************************************/

// call_function.c - A sample of calling
// python functions from C code
//
#include <Python.h>

int main(int argc, char *argv[])
{
    PyObject *pName, *pModule, *pDict, *pFunc, *pValue;

    if (argc < 3)
    {
        printf("Usage: exe_name python_source function_name\n");
        return 1;
    }

    // Initialize the Python Interpreter
    Py_Initialize();

    // Build the name object
    pName = PyString_FromString(argv[1]);

    // Load the module object
    pModule = PyImport_Import(pName);

    // pDict is a borrowed reference
    pDict = PyModule_GetDict(pModule);

    // pFunc is also a borrowed reference
    pFunc = PyDict_GetItemString(pDict, argv[2]);

    if (PyCallable_Check(pFunc))
    {
        PyObject_CallObject(pFunc, NULL);
    } else
    {
        PyErr_Print();
    }

    // Clean up
    Py_DECREF(pModule);
    Py_DECREF(pName);

    // Finish the Python Interpreter
    Py_Finalize();

    return 0;
}

/************************************************************************************************************/
python file saved as

py_function.py

/************************************************************************************************************/

'''py_function.py - Python source designed to '''
'''demonstrate the use of python embedding'''

def multiply():
    c = 12345*6789
    print 'The result of 12345 x 6789 :', c


    return c

/************************************************************************************************************/

$gcc -g call_function.c  -lpython2.5 -ocall_function
$./call_function py_function.py multiply

i get following output

ImportError: No module named py_function
Failed to load "py_function"

when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..

can any one please find out the error that i am doing.

-- 
Shrikant
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071025/e3ca2dfd/attachment.htm>

From gjcarneiro at gmail.com  Thu Oct 25 09:21:19 2007
From: gjcarneiro at gmail.com (Gustavo Carneiro)
Date: Thu, 25 Oct 2007 08:21:19 +0100
Subject: [C++-sig] failed in PyImport_Import()
In-Reply-To: <87de5f460710242352w5b57dc4djbcfba96ccf0f3578@mail.gmail.com>
References: <87de5f460710242352w5b57dc4djbcfba96ccf0f3578@mail.gmail.com>
Message-ID: <a467ca4f0710250021x50e0a9c8m5ad849e9b9e3d5b1@mail.gmail.com>

On 25/10/2007, Shrikant Chikhalkar <shrikantvc at gmail.com> wrote:
>
> Hi,
>
> i am tring to call python function from 'C' program. i get NULL value
> return by PyImport_Import()
>
> following is the code. code is copied from
> http://www.codeproject.com/cpp/embedpython_1.asp
>
> following file is saved as
>
> call_function.c
>
> /************************************************************************************************************/
>
>
> // call_function.c - A sample of calling
> // python functions from C code
>
> //
> #include <Python.h>
>
> int main(int argc, char *argv[])
> {
>
>     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
>
>     if (argc < 3)
>     {
>         printf("Usage: exe_name python_source function_name\n"
> );
>         return 1;
>     }
>
>     // Initialize the Python Interpreter
>     Py_Initialize();
>
>
> // Build the name object
>     pName = PyString_FromString(argv[1]);
>
>     // Load the module object
>     pModule = PyImport_Import(pName);
>
>
>     // pDict is a borrowed reference
>     pDict = PyModule_GetDict(pModule);
>
>     // pFunc is also a borrowed reference
>     pFunc = PyDict_GetItemString(pDict, argv[
> 2]);
>
>     if (PyCallable_Check(pFunc))
>     {
>         PyObject_CallObject(pFunc, NULL);
>     } else
>     {
>
>         PyErr_Print();
>     }
>
>     // Clean up
>     Py_DECREF(pModule);
>     Py_DECREF(pName);
>
>     // Finish the Python Interpreter
>
>     Py_Finalize();
>
>     return 0;
> }
>
> /************************************************************************************************************/
>
> python file saved as
>
> py_function.py
>
>
> /************************************************************************************************************/
>
> '''py_function.py - Python source designed to '''
> '''demonstrate the use of python embedding'''
>
> def multiply():
>     c = 12345*6789
>     print 'The result of 12345 x 6789 :', c
>
>
>
>     return c
>
>
> /************************************************************************************************************/
>
> $gcc -g call_function.c  -lpython2.5 -ocall_function
> $./call_function py_function.py multiply


Shouldn't this be:
   $./call_function py_function multiply

I.e., you the module name should not have a .py extension.

i get following output
>
> ImportError: No module named py_function
> Failed to load "py_function"
>
> when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
>
> can any one please find out the error that i am doing.
>
> --
> Shrikant
>
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
>


-- 
Gustavo J. A. M. Carneiro
INESC Porto, Telecommunications and Multimedia Unit
"The universe is always one step beyond logic." -- Frank Herbert
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071025/0fc3b1ea/attachment.htm>

From shrikantvc at gmail.com  Thu Oct 25 13:03:42 2007
From: shrikantvc at gmail.com (Shrikant Chikhalkar)
Date: Thu, 25 Oct 2007 16:33:42 +0530
Subject: [C++-sig] C++-sig Digest, Vol 51, Issue 25
In-Reply-To: <mailman.45.1193306412.32757.c++-sig@python.org>
References: <mailman.45.1193306412.32757.c++-sig@python.org>
Message-ID: <87de5f460710250403q311e8fd1q61de04dc7b935f88@mail.gmail.com>

Hi,
thanks for the reply.
that i have already tried. Still i am getting same error. could you please
give some pointer that where i am doing wrong.

thanks
Shrikant


On 10/25/07, c++-sig-request at python.org <c++-sig-request at python.org> wrote:
>
> Send C++-sig mailing list submissions to
>         c++-sig at python.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://mail.python.org/mailman/listinfo/c++-sig
> or, via email, send a message with subject or body 'help' to
>         c++-sig-request at python.org
>
> You can reach the person managing the list at
>         c++-sig-owner at python.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of C++-sig digest..."
>
>
> Today's Topics:
>
>    1. failed in PyImport_Import() (Shrikant Chikhalkar)
>    2. Re: failed in PyImport_Import() (Gustavo Carneiro)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Oct 2007 12:22:24 +0530
> From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> Subject: [C++-sig] failed in PyImport_Import()
> To: c++-sig at python.org
> Message-ID:
>         <87de5f460710242352w5b57dc4djbcfba96ccf0f3578 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
>
> i am tring to call python function from 'C' program. i get NULL value
> return
> by PyImport_Import()
>
> following is the code. code is copied from
> http://www.codeproject.com/cpp/embedpython_1.asp
>
> following file is saved as
>
> call_function.c
>
>
> /************************************************************************************************************/
>
> // call_function.c - A sample of calling
> // python functions from C code
> //
> #include <Python.h>
>
> int main(int argc, char *argv[])
> {
>     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
>
>     if (argc < 3)
>     {
>         printf("Usage: exe_name python_source function_name\n");
>         return 1;
>     }
>
>     // Initialize the Python Interpreter
>     Py_Initialize();
>
>     // Build the name object
>     pName = PyString_FromString(argv[1]);
>
>     // Load the module object
>     pModule = PyImport_Import(pName);
>
>     // pDict is a borrowed reference
>     pDict = PyModule_GetDict(pModule);
>
>     // pFunc is also a borrowed reference
>     pFunc = PyDict_GetItemString(pDict, argv[2]);
>
>     if (PyCallable_Check(pFunc))
>     {
>         PyObject_CallObject(pFunc, NULL);
>     } else
>     {
>         PyErr_Print();
>     }
>
>     // Clean up
>     Py_DECREF(pModule);
>     Py_DECREF(pName);
>
>     // Finish the Python Interpreter
>     Py_Finalize();
>
>     return 0;
> }
>
>
> /************************************************************************************************************/
> python file saved as
>
> py_function.py
>
>
> /************************************************************************************************************/
>
> '''py_function.py - Python source designed to '''
> '''demonstrate the use of python embedding'''
>
> def multiply():
>     c = 12345*6789
>     print 'The result of 12345 x 6789 :', c
>
>
>     return c
>
>
> /************************************************************************************************************/
>
> $gcc -g call_function.c  -lpython2.5 -ocall_function
> $./call_function py_function.py multiply
>
> i get following output
>
> ImportError: No module named py_function
> Failed to load "py_function"
>
> when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
>
> can any one please find out the error that i am doing.
>
> --
> Shrikant
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/e3ca2dfd/attachment-0001.htm
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Oct 2007 08:21:19 +0100
> From: "Gustavo Carneiro" <gjcarneiro at gmail.com>
> Subject: Re: [C++-sig] failed in PyImport_Import()
> To: "Development of Python/C++ integration" <c++-sig at python.org>
> Message-ID:
>         <a467ca4f0710250021x50e0a9c8m5ad849e9b9e3d5b1 at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> On 25/10/2007, Shrikant Chikhalkar <shrikantvc at gmail.com> wrote:
> >
> > Hi,
> >
> > i am tring to call python function from 'C' program. i get NULL value
> > return by PyImport_Import()
> >
> > following is the code. code is copied from
> > http://www.codeproject.com/cpp/embedpython_1.asp
> >
> > following file is saved as
> >
> > call_function.c
> >
> >
> /************************************************************************************************************/
> >
> >
> > // call_function.c - A sample of calling
> > // python functions from C code
> >
> > //
> > #include <Python.h>
> >
> > int main(int argc, char *argv[])
> > {
> >
> >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> >
> >     if (argc < 3)
> >     {
> >         printf("Usage: exe_name python_source function_name\n"
> > );
> >         return 1;
> >     }
> >
> >     // Initialize the Python Interpreter
> >     Py_Initialize();
> >
> >
> > // Build the name object
> >     pName = PyString_FromString(argv[1]);
> >
> >     // Load the module object
> >     pModule = PyImport_Import(pName);
> >
> >
> >     // pDict is a borrowed reference
> >     pDict = PyModule_GetDict(pModule);
> >
> >     // pFunc is also a borrowed reference
> >     pFunc = PyDict_GetItemString(pDict, argv[
> > 2]);
> >
> >     if (PyCallable_Check(pFunc))
> >     {
> >         PyObject_CallObject(pFunc, NULL);
> >     } else
> >     {
> >
> >         PyErr_Print();
> >     }
> >
> >     // Clean up
> >     Py_DECREF(pModule);
> >     Py_DECREF(pName);
> >
> >     // Finish the Python Interpreter
> >
> >     Py_Finalize();
> >
> >     return 0;
> > }
> >
> >
> /************************************************************************************************************/
> >
> > python file saved as
> >
> > py_function.py
> >
> >
> >
> /************************************************************************************************************/
> >
> > '''py_function.py - Python source designed to '''
> > '''demonstrate the use of python embedding'''
> >
> > def multiply():
> >     c = 12345*6789
> >     print 'The result of 12345 x 6789 :', c
> >
> >
> >
> >     return c
> >
> >
> >
> /************************************************************************************************************/
> >
> > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > $./call_function py_function.py multiply
>
>
> Shouldn't this be:
>    $./call_function py_function multiply
>
> I.e., you the module name should not have a .py extension.
>
> i get following output
> >
> > ImportError: No module named py_function
> > Failed to load "py_function"
> >
> > when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
> >
> > can any one please find out the error that i am doing.
> >
> > --
> > Shrikant
> >
> >
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> >
> >
>
>
> --
> Gustavo J. A. M. Carneiro
> INESC Porto, Telecommunications and Multimedia Unit
> "The universe is always one step beyond logic." -- Frank Herbert
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/0fc3b1ea/attachment-0001.htm
>
> ------------------------------
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
>
> End of C++-sig Digest, Vol 51, Issue 25
> ***************************************
>



-- 
Shrikant Chikhalkar
+91 9850991554
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071025/63f1ada0/attachment.htm>

From shrikantvc at gmail.com  Thu Oct 25 13:24:02 2007
From: shrikantvc at gmail.com (Shrikant Chikhalkar)
Date: Thu, 25 Oct 2007 16:54:02 +0530
Subject: [C++-sig] C++-sig Digest, Vol 51, Issue 25
In-Reply-To: <mailman.45.1193306412.32757.c++-sig@python.org>
References: <mailman.45.1193306412.32757.c++-sig@python.org>
Message-ID: <87de5f460710250424k1ef321aei273ae019035bf476@mail.gmail.com>

Hi,

I have also check that "py_function.py" is in the same directory.
i have some doubt
1) do we need to install "py_function" module before calling multiply
function from that. ( install means python -m  py_function)
2) individually multiply function is working fine on python prompt.
3) I am using python2.5 headers

I know some where i had make a mistake. please help me to locate that

thanks
Shrikant

On 10/25/07, c++-sig-request at python.org <c++-sig-request at python.org> wrote:
>
> Send C++-sig mailing list submissions to
>         c++-sig at python.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://mail.python.org/mailman/listinfo/c++-sig
> or, via email, send a message with subject or body 'help' to
>         c++-sig-request at python.org
>
> You can reach the person managing the list at
>         c++-sig-owner at python.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of C++-sig digest..."
>
>
> Today's Topics:
>
>    1. failed in PyImport_Import() (Shrikant Chikhalkar)
>    2. Re: failed in PyImport_Import() (Gustavo Carneiro)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Oct 2007 12:22:24 +0530
> From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> Subject: [C++-sig] failed in PyImport_Import()
> To: c++-sig at python.org
> Message-ID:
>         <87de5f460710242352w5b57dc4djbcfba96ccf0f3578 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
>
> i am tring to call python function from 'C' program. i get NULL value
> return
> by PyImport_Import()
>
> following is the code. code is copied from
> http://www.codeproject.com/cpp/embedpython_1.asp
>
> following file is saved as
>
> call_function.c
>
>
> /************************************************************************************************************/
>
> // call_function.c - A sample of calling
> // python functions from C code
> //
> #include <Python.h>
>
> int main(int argc, char *argv[])
> {
>     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
>
>     if (argc < 3)
>     {
>         printf("Usage: exe_name python_source function_name\n");
>         return 1;
>     }
>
>     // Initialize the Python Interpreter
>     Py_Initialize();
>
>     // Build the name object
>     pName = PyString_FromString(argv[1]);
>
>     // Load the module object
>     pModule = PyImport_Import(pName);
>
>     // pDict is a borrowed reference
>     pDict = PyModule_GetDict(pModule);
>
>     // pFunc is also a borrowed reference
>     pFunc = PyDict_GetItemString(pDict, argv[2]);
>
>     if (PyCallable_Check(pFunc))
>     {
>         PyObject_CallObject(pFunc, NULL);
>     } else
>     {
>         PyErr_Print();
>     }
>
>     // Clean up
>     Py_DECREF(pModule);
>     Py_DECREF(pName);
>
>     // Finish the Python Interpreter
>     Py_Finalize();
>
>     return 0;
> }
>
>
> /************************************************************************************************************/
> python file saved as
>
> py_function.py
>
>
> /************************************************************************************************************/
>
> '''py_function.py - Python source designed to '''
> '''demonstrate the use of python embedding'''
>
> def multiply():
>     c = 12345*6789
>     print 'The result of 12345 x 6789 :', c
>
>
>     return c
>
>
> /************************************************************************************************************/
>
> $gcc -g call_function.c  -lpython2.5 -ocall_function
> $./call_function py_function.py multiply
>
> i get following output
>
> ImportError: No module named py_function
> Failed to load "py_function"
>
> when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
>
> can any one please find out the error that i am doing.
>
> --
> Shrikant
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/e3ca2dfd/attachment-0001.htm
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Oct 2007 08:21:19 +0100
> From: "Gustavo Carneiro" <gjcarneiro at gmail.com>
> Subject: Re: [C++-sig] failed in PyImport_Import()
> To: "Development of Python/C++ integration" <c++-sig at python.org>
> Message-ID:
>         <a467ca4f0710250021x50e0a9c8m5ad849e9b9e3d5b1 at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> On 25/10/2007, Shrikant Chikhalkar <shrikantvc at gmail.com> wrote:
> >
> > Hi,
> >
> > i am tring to call python function from 'C' program. i get NULL value
> > return by PyImport_Import()
> >
> > following is the code. code is copied from
> > http://www.codeproject.com/cpp/embedpython_1.asp
> >
> > following file is saved as
> >
> > call_function.c
> >
> >
> /************************************************************************************************************/
> >
> >
> > // call_function.c - A sample of calling
> > // python functions from C code
> >
> > //
> > #include <Python.h>
> >
> > int main(int argc, char *argv[])
> > {
> >
> >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> >
> >     if (argc < 3)
> >     {
> >         printf("Usage: exe_name python_source function_name\n"
> > );
> >         return 1;
> >     }
> >
> >     // Initialize the Python Interpreter
> >     Py_Initialize();
> >
> >
> > // Build the name object
> >     pName = PyString_FromString(argv[1]);
> >
> >     // Load the module object
> >     pModule = PyImport_Import(pName);
> >
> >
> >     // pDict is a borrowed reference
> >     pDict = PyModule_GetDict(pModule);
> >
> >     // pFunc is also a borrowed reference
> >     pFunc = PyDict_GetItemString(pDict, argv[
> > 2]);
> >
> >     if (PyCallable_Check(pFunc))
> >     {
> >         PyObject_CallObject(pFunc, NULL);
> >     } else
> >     {
> >
> >         PyErr_Print();
> >     }
> >
> >     // Clean up
> >     Py_DECREF(pModule);
> >     Py_DECREF(pName);
> >
> >     // Finish the Python Interpreter
> >
> >     Py_Finalize();
> >
> >     return 0;
> > }
> >
> >
> /************************************************************************************************************/
> >
> > python file saved as
> >
> > py_function.py
> >
> >
> >
> /************************************************************************************************************/
> >
> > '''py_function.py - Python source designed to '''
> > '''demonstrate the use of python embedding'''
> >
> > def multiply():
> >     c = 12345*6789
> >     print 'The result of 12345 x 6789 :', c
> >
> >
> >
> >     return c
> >
> >
> >
> /************************************************************************************************************/
> >
> > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > $./call_function py_function.py multiply
>
>
> Shouldn't this be:
>    $./call_function py_function multiply
>
> I.e., you the module name should not have a .py extension.
>
> i get following output
> >
> > ImportError: No module named py_function
> > Failed to load "py_function"
> >
> > when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
> >
> > can any one please find out the error that i am doing.
> >
> > --
> > Shrikant
> >
> >
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> >
> >
>
>
> --
> Gustavo J. A. M. Carneiro
> INESC Porto, Telecommunications and Multimedia Unit
> "The universe is always one step beyond logic." -- Frank Herbert
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/0fc3b1ea/attachment-0001.htm
>
> ------------------------------
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
>
> End of C++-sig Digest, Vol 51, Issue 25
> ***************************************
>



-- 
Shrikant Chikhalkar
+91 9850991554
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071025/26af7201/attachment.htm>

From don.j.olmstead at gmail.com  Fri Oct 26 00:36:37 2007
From: don.j.olmstead at gmail.com (Don)
Date: Thu, 25 Oct 2007 22:36:37 +0000 (UTC)
Subject: [C++-sig] =?utf-8?q?Using_an_intrusive=5Fptr_with_a_noncopyable_m?=
	=?utf-8?q?ember_variable?=
Message-ID: <loom.20071025T195626-196@post.gmane.org>

Hello all,

Just starting to try using boost python for exposing some code I've written.
Essentially I have a class Object that has a protected constructor, that is
derived by various other classes. This class does its own reference counting via
intrusive_ptr. Because it does its own reference counting and for thread safety
I have a mutex that is entered when changing the reference count.

So the header file for the Object class, with unrelated functionality stripped
out is.

namespace Shade
{
	class Object
	{
		protected:
			Object();
		public:
			virtual ~Object();

		// Reference counting
		public:
			void incrementReferences();
			void decrementReferences();
			unsigned int getNumberOfReferences() const;
		private:
			/** Number of references to the Object */
			unsigned int numReferences;
			/**
			 * Mutex for the reference counting system.
			 *
			 * Assures that a pointer isn't accidently deleted or
			 * left dangling when accessed by multiple threads.
			 */
			boost::mutex mutexRefCount;
	} ;

	/** Smart Pointer for the Object class */
	typedef boost::intrusive_ptr<Object> ObjectPtr;

	#include "ShadeObject.inl"
}

void intrusive_ptr_add_ref(Shade::Object* object);
void intrusive_ptr_release(Shade::Object* object);

I expose it to Python using the following code

BOOST_PYTHON_MODULE(Shade)
{
	python::class_<Object, boost::noncopyable, ObjectPtr>("Object", python::no_init)
		// class .def's
	;
	// Uninteresting dummy test class
	python::class_<DummyObject, boost::noncopyable, DummyObjectPtr,
python::bases<Object> >("DummyObject");

}

So in C++ if I do

DummyObjectPtr test1 = DummyObject();
DummyObjectPtr test2 = test1;

std::cout << test1->getNumberOfReferences() << std::endl;

The console will say
2

Now if I do the same in Python

test1 = DummyObject()
test2 = test1
print test1.getNumberOfReferences()

The console will say
1

The Object wont attempt to delete itself until both test1 and test2 are set to
something else so Python must be keeping a reference count. It would just be
nice to know what the total number of references are across Python and C++.

Any ideas?



From steven.wyckoff at gmail.com  Fri Oct 26 06:49:49 2007
From: steven.wyckoff at gmail.com (Steven Wyckoff)
Date: Thu, 25 Oct 2007 22:49:49 -0600
Subject: [C++-sig] failed in PyImport_Import()
In-Reply-To: <87de5f460710242352w5b57dc4djbcfba96ccf0f3578@mail.gmail.com>
References: <87de5f460710242352w5b57dc4djbcfba96ccf0f3578@mail.gmail.com>
Message-ID: <ffrrjk$6av$1@ger.gmane.org>

Based on reading the code, I cannot see why it would be failing to 
create the module. Having gone through a similar process not too long 
ago, there are a couple gotchas that I ran into.

First is the .py in the module name: you need it in the file name, and 
not in the module name (just like importing within python). Another is 
that depending on your debugger, it may be running the program in a 
different location than the executable is located (for example VS does 
this I know). You had mentioned that your py_function works properly in 
a python prompt but just as a heads up, we had problems where a syntax 
error in the python file would cause cryptic module loading errors. You 
may also want to check that there is not an old version of the .pyc file 
sitting around confusing things.

It looks like you are using PyErr_Print() and getting some info back 
from it, I found that to be very helpful, even more so once the scripts 
were running.

Python should look for a module in the directory where it is being run 
from, however, if you think that the issue is that python is not finding 
your file, you can add an arbitrary directory on your computer to the 
module search path within your program:

PyRun_SimpleString("import sys");
PyRun_SimpleString("sys.path.append(\"your\path\here")");

You could also try using PyRun_SimpleString("import py_function") to try 
importing the module just to try to isolate the issue.

I hope that your problem is the same as one of ours and good luck!

Steve Wyckoff


Shrikant Chikhalkar wrote:
> Hi,
> 
> i am tring to call python function from 'C' program. i get NULL value 
> return by PyImport_Import()
> 
> following is the code. code is copied from 
> http://www.codeproject.com/cpp/embedpython_1.asp
> 
> following file is saved as
> 
> call_function.c
> 
> /************************************************************************************************************/ 
> 
> 
> // call_function.c - A sample of calling 
> // python functions from C code
> 
> // 
> #include <Python.h>
> 
> int main(int argc, char *argv[])
> {
> 
>     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> 
>     if (argc < 3) 
>     {
>         printf("Usage: exe_name python_source function_name\n"
> );
>         return 1;
>     }
> 
>     // Initialize the Python Interpreter
>     Py_Initialize();
> 
>     
> // Build the name object
>     pName = PyString_FromString(argv[1]);
> 
>     // Load the module object
>     pModule = PyImport_Import(pName);
> 
> 
>     // pDict is a borrowed reference 
>     pDict = PyModule_GetDict(pModule);
> 
>     // pFunc is also a borrowed reference 
>     pFunc = PyDict_GetItemString(pDict, argv[
> 2]);
> 
>     if (PyCallable_Check(pFunc)) 
>     {
>         PyObject_CallObject(pFunc, NULL);
>     } else 
>     {
> 
>         PyErr_Print();
>     }
> 
>     // Clean up
>     Py_DECREF(pModule);
>     Py_DECREF(pName);
> 
>     // Finish the Python Interpreter
> 
>     Py_Finalize();
> 
>     return 0;
> }
> 
> /************************************************************************************************************/ 
> 
> python file saved as 
> 
> py_function.py
> 
> /************************************************************************************************************/
> 
> '''py_function.py - Python source designed to '''
> '''demonstrate the use of python embedding'''
> 
> def multiply():
>     c = 12345*6789
>     print 'The result of 12345 x 6789 :', c
> 
> 
> 
>     return c
> 
> /************************************************************************************************************/
> 
> $gcc -g call_function.c  -lpython2.5 -ocall_function
> $./call_function py_function.py multiply
> 
> i get following output
> 
> ImportError: No module named py_function
> Failed to load "py_function"
> 
> when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
> 
> can any one please find out the error that i am doing.
> 
> -- 
> Shrikant
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig




From shrikantvc at gmail.com  Fri Oct 26 12:27:00 2007
From: shrikantvc at gmail.com (Shrikant Chikhalkar)
Date: Fri, 26 Oct 2007 15:57:00 +0530
Subject: [C++-sig] C++-sig Digest, Vol 51, Issue 26
In-Reply-To: <mailman.1710.1193311451.13604.c++-sig@python.org>
References: <mailman.1710.1193311451.13604.c++-sig@python.org>
Message-ID: <87de5f460710260327l1e3574c2s9234978669888ac8@mail.gmail.com>

I got solution for the problem.

using root credentials run following command on shell prompt
$python2 setup.py install

where setup.py would be
from distutils.core import setup
setup(name='xxx',
      version='1.0',
      py_modules=['py_function'],
      )

thanks
shrikant

On 10/25/07, c++-sig-request at python.org <c++-sig-request at python.org> wrote:
>
> Send C++-sig mailing list submissions to
>         c++-sig at python.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://mail.python.org/mailman/listinfo/c++-sig
> or, via email, send a message with subject or body 'help' to
>         c++-sig-request at python.org
>
> You can reach the person managing the list at
>         c++-sig-owner at python.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of C++-sig digest..."
>
>
> Today's Topics:
>
>    1. Re: C++-sig Digest, Vol 51, Issue 25 (Shrikant Chikhalkar)
>    2. Re: C++-sig Digest, Vol 51, Issue 25 (Shrikant Chikhalkar)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Oct 2007 16:33:42 +0530
> From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> Subject: Re: [C++-sig] C++-sig Digest, Vol 51, Issue 25
> To: c++-sig at python.org
> Message-ID:
>         <87de5f460710250403q311e8fd1q61de04dc7b935f88 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
> thanks for the reply.
> that i have already tried. Still i am getting same error. could you please
> give some pointer that where i am doing wrong.
>
> thanks
> Shrikant
>
>
> On 10/25/07, c++-sig-request at python.org <c++-sig-request at python.org>
> wrote:
> >
> > Send C++-sig mailing list submissions to
> >         c++-sig at python.org
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> >         http://mail.python.org/mailman/listinfo/c++-sig
> > or, via email, send a message with subject or body 'help' to
> >         c++-sig-request at python.org
> >
> > You can reach the person managing the list at
> >         c++-sig-owner at python.org
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of C++-sig digest..."
> >
> >
> > Today's Topics:
> >
> >    1. failed in PyImport_Import() (Shrikant Chikhalkar)
> >    2. Re: failed in PyImport_Import() (Gustavo Carneiro)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Thu, 25 Oct 2007 12:22:24 +0530
> > From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> > Subject: [C++-sig] failed in PyImport_Import()
> > To: c++-sig at python.org
> > Message-ID:
> >         <87de5f460710242352w5b57dc4djbcfba96ccf0f3578 at mail.gmail.com>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > Hi,
> >
> > i am tring to call python function from 'C' program. i get NULL value
> > return
> > by PyImport_Import()
> >
> > following is the code. code is copied from
> > http://www.codeproject.com/cpp/embedpython_1.asp
> >
> > following file is saved as
> >
> > call_function.c
> >
> >
> >
> /************************************************************************************************************/
> >
> > // call_function.c - A sample of calling
> > // python functions from C code
> > //
> > #include <Python.h>
> >
> > int main(int argc, char *argv[])
> > {
> >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> >
> >     if (argc < 3)
> >     {
> >         printf("Usage: exe_name python_source function_name\n");
> >         return 1;
> >     }
> >
> >     // Initialize the Python Interpreter
> >     Py_Initialize();
> >
> >     // Build the name object
> >     pName = PyString_FromString(argv[1]);
> >
> >     // Load the module object
> >     pModule = PyImport_Import(pName);
> >
> >     // pDict is a borrowed reference
> >     pDict = PyModule_GetDict(pModule);
> >
> >     // pFunc is also a borrowed reference
> >     pFunc = PyDict_GetItemString(pDict, argv[2]);
> >
> >     if (PyCallable_Check(pFunc))
> >     {
> >         PyObject_CallObject(pFunc, NULL);
> >     } else
> >     {
> >         PyErr_Print();
> >     }
> >
> >     // Clean up
> >     Py_DECREF(pModule);
> >     Py_DECREF(pName);
> >
> >     // Finish the Python Interpreter
> >     Py_Finalize();
> >
> >     return 0;
> > }
> >
> >
> >
> /************************************************************************************************************/
> > python file saved as
> >
> > py_function.py
> >
> >
> >
> /************************************************************************************************************/
> >
> > '''py_function.py - Python source designed to '''
> > '''demonstrate the use of python embedding'''
> >
> > def multiply():
> >     c = 12345*6789
> >     print 'The result of 12345 x 6789 :', c
> >
> >
> >     return c
> >
> >
> >
> /************************************************************************************************************/
> >
> > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > $./call_function py_function.py multiply
> >
> > i get following output
> >
> > ImportError: No module named py_function
> > Failed to load "py_function"
> >
> > when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
> >
> > can any one please find out the error that i am doing.
> >
> > --
> > Shrikant
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> >
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/e3ca2dfd/attachment-0001.htm
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Thu, 25 Oct 2007 08:21:19 +0100
> > From: "Gustavo Carneiro" <gjcarneiro at gmail.com>
> > Subject: Re: [C++-sig] failed in PyImport_Import()
> > To: "Development of Python/C++ integration" <c++-sig at python.org>
> > Message-ID:
> >         <a467ca4f0710250021x50e0a9c8m5ad849e9b9e3d5b1 at mail.gmail.com>
> > Content-Type: text/plain; charset="utf-8"
> >
> > On 25/10/2007, Shrikant Chikhalkar <shrikantvc at gmail.com> wrote:
> > >
> > > Hi,
> > >
> > > i am tring to call python function from 'C' program. i get NULL value
> > > return by PyImport_Import()
> > >
> > > following is the code. code is copied from
> > > http://www.codeproject.com/cpp/embedpython_1.asp
> > >
> > > following file is saved as
> > >
> > > call_function.c
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > >
> > > // call_function.c - A sample of calling
> > > // python functions from C code
> > >
> > > //
> > > #include <Python.h>
> > >
> > > int main(int argc, char *argv[])
> > > {
> > >
> > >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> > >
> > >     if (argc < 3)
> > >     {
> > >         printf("Usage: exe_name python_source function_name\n"
> > > );
> > >         return 1;
> > >     }
> > >
> > >     // Initialize the Python Interpreter
> > >     Py_Initialize();
> > >
> > >
> > > // Build the name object
> > >     pName = PyString_FromString(argv[1]);
> > >
> > >     // Load the module object
> > >     pModule = PyImport_Import(pName);
> > >
> > >
> > >     // pDict is a borrowed reference
> > >     pDict = PyModule_GetDict(pModule);
> > >
> > >     // pFunc is also a borrowed reference
> > >     pFunc = PyDict_GetItemString(pDict, argv[
> > > 2]);
> > >
> > >     if (PyCallable_Check(pFunc))
> > >     {
> > >         PyObject_CallObject(pFunc, NULL);
> > >     } else
> > >     {
> > >
> > >         PyErr_Print();
> > >     }
> > >
> > >     // Clean up
> > >     Py_DECREF(pModule);
> > >     Py_DECREF(pName);
> > >
> > >     // Finish the Python Interpreter
> > >
> > >     Py_Finalize();
> > >
> > >     return 0;
> > > }
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > python file saved as
> > >
> > > py_function.py
> > >
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > '''py_function.py - Python source designed to '''
> > > '''demonstrate the use of python embedding'''
> > >
> > > def multiply():
> > >     c = 12345*6789
> > >     print 'The result of 12345 x 6789 :', c
> > >
> > >
> > >
> > >     return c
> > >
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > > $./call_function py_function.py multiply
> >
> >
> > Shouldn't this be:
> >    $./call_function py_function multiply
> >
> > I.e., you the module name should not have a .py extension.
> >
> > i get following output
> > >
> > > ImportError: No module named py_function
> > > Failed to load "py_function"
> > >
> > > when i kdbg(debug) it i vet NULL value returned from
> PyImport_Import()..
> > >
> > > can any one please find out the error that i am doing.
> > >
> > > --
> > > Shrikant
> > >
> > >
> > > _______________________________________________
> > > C++-sig mailing list
> > > C++-sig at python.org
> > > http://mail.python.org/mailman/listinfo/c++-sig
> > >
> > >
> >
> >
> > --
> > Gustavo J. A. M. Carneiro
> > INESC Porto, Telecommunications and Multimedia Unit
> > "The universe is always one step beyond logic." -- Frank Herbert
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> >
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/0fc3b1ea/attachment-0001.htm
> >
> > ------------------------------
> >
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> >
> >
> > End of C++-sig Digest, Vol 51, Issue 25
> > ***************************************
> >
>
>
>
> --
> Shrikant Chikhalkar
> +91 9850991554
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/63f1ada0/attachment-0001.htm
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Oct 2007 16:54:02 +0530
> From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> Subject: Re: [C++-sig] C++-sig Digest, Vol 51, Issue 25
> To: c++-sig at python.org
> Message-ID:
>         <87de5f460710250424k1ef321aei273ae019035bf476 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
>
> I have also check that "py_function.py" is in the same directory.
> i have some doubt
> 1) do we need to install "py_function" module before calling multiply
> function from that. ( install means python -m  py_function)
> 2) individually multiply function is working fine on python prompt.
> 3) I am using python2.5 headers
>
> I know some where i had make a mistake. please help me to locate that
>
> thanks
> Shrikant
>
> On 10/25/07, c++-sig-request at python.org <c++-sig-request at python.org>
> wrote:
> >
> > Send C++-sig mailing list submissions to
> >         c++-sig at python.org
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> >         http://mail.python.org/mailman/listinfo/c++-sig
> > or, via email, send a message with subject or body 'help' to
> >         c++-sig-request at python.org
> >
> > You can reach the person managing the list at
> >         c++-sig-owner at python.org
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of C++-sig digest..."
> >
> >
> > Today's Topics:
> >
> >    1. failed in PyImport_Import() (Shrikant Chikhalkar)
> >    2. Re: failed in PyImport_Import() (Gustavo Carneiro)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Thu, 25 Oct 2007 12:22:24 +0530
> > From: "Shrikant Chikhalkar" <shrikantvc at gmail.com>
> > Subject: [C++-sig] failed in PyImport_Import()
> > To: c++-sig at python.org
> > Message-ID:
> >         <87de5f460710242352w5b57dc4djbcfba96ccf0f3578 at mail.gmail.com>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > Hi,
> >
> > i am tring to call python function from 'C' program. i get NULL value
> > return
> > by PyImport_Import()
> >
> > following is the code. code is copied from
> > http://www.codeproject.com/cpp/embedpython_1.asp
> >
> > following file is saved as
> >
> > call_function.c
> >
> >
> >
> /************************************************************************************************************/
> >
> > // call_function.c - A sample of calling
> > // python functions from C code
> > //
> > #include <Python.h>
> >
> > int main(int argc, char *argv[])
> > {
> >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> >
> >     if (argc < 3)
> >     {
> >         printf("Usage: exe_name python_source function_name\n");
> >         return 1;
> >     }
> >
> >     // Initialize the Python Interpreter
> >     Py_Initialize();
> >
> >     // Build the name object
> >     pName = PyString_FromString(argv[1]);
> >
> >     // Load the module object
> >     pModule = PyImport_Import(pName);
> >
> >     // pDict is a borrowed reference
> >     pDict = PyModule_GetDict(pModule);
> >
> >     // pFunc is also a borrowed reference
> >     pFunc = PyDict_GetItemString(pDict, argv[2]);
> >
> >     if (PyCallable_Check(pFunc))
> >     {
> >         PyObject_CallObject(pFunc, NULL);
> >     } else
> >     {
> >         PyErr_Print();
> >     }
> >
> >     // Clean up
> >     Py_DECREF(pModule);
> >     Py_DECREF(pName);
> >
> >     // Finish the Python Interpreter
> >     Py_Finalize();
> >
> >     return 0;
> > }
> >
> >
> >
> /************************************************************************************************************/
> > python file saved as
> >
> > py_function.py
> >
> >
> >
> /************************************************************************************************************/
> >
> > '''py_function.py - Python source designed to '''
> > '''demonstrate the use of python embedding'''
> >
> > def multiply():
> >     c = 12345*6789
> >     print 'The result of 12345 x 6789 :', c
> >
> >
> >     return c
> >
> >
> >
> /************************************************************************************************************/
> >
> > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > $./call_function py_function.py multiply
> >
> > i get following output
> >
> > ImportError: No module named py_function
> > Failed to load "py_function"
> >
> > when i kdbg(debug) it i vet NULL value returned from PyImport_Import()..
> >
> > can any one please find out the error that i am doing.
> >
> > --
> > Shrikant
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> >
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/e3ca2dfd/attachment-0001.htm
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Thu, 25 Oct 2007 08:21:19 +0100
> > From: "Gustavo Carneiro" <gjcarneiro at gmail.com>
> > Subject: Re: [C++-sig] failed in PyImport_Import()
> > To: "Development of Python/C++ integration" <c++-sig at python.org>
> > Message-ID:
> >         <a467ca4f0710250021x50e0a9c8m5ad849e9b9e3d5b1 at mail.gmail.com>
> > Content-Type: text/plain; charset="utf-8"
> >
> > On 25/10/2007, Shrikant Chikhalkar <shrikantvc at gmail.com> wrote:
> > >
> > > Hi,
> > >
> > > i am tring to call python function from 'C' program. i get NULL value
> > > return by PyImport_Import()
> > >
> > > following is the code. code is copied from
> > > http://www.codeproject.com/cpp/embedpython_1.asp
> > >
> > > following file is saved as
> > >
> > > call_function.c
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > >
> > > // call_function.c - A sample of calling
> > > // python functions from C code
> > >
> > > //
> > > #include <Python.h>
> > >
> > > int main(int argc, char *argv[])
> > > {
> > >
> > >     PyObject *pName, *pModule, *pDict, *pFunc, *pValue;
> > >
> > >     if (argc < 3)
> > >     {
> > >         printf("Usage: exe_name python_source function_name\n"
> > > );
> > >         return 1;
> > >     }
> > >
> > >     // Initialize the Python Interpreter
> > >     Py_Initialize();
> > >
> > >
> > > // Build the name object
> > >     pName = PyString_FromString(argv[1]);
> > >
> > >     // Load the module object
> > >     pModule = PyImport_Import(pName);
> > >
> > >
> > >     // pDict is a borrowed reference
> > >     pDict = PyModule_GetDict(pModule);
> > >
> > >     // pFunc is also a borrowed reference
> > >     pFunc = PyDict_GetItemString(pDict, argv[
> > > 2]);
> > >
> > >     if (PyCallable_Check(pFunc))
> > >     {
> > >         PyObject_CallObject(pFunc, NULL);
> > >     } else
> > >     {
> > >
> > >         PyErr_Print();
> > >     }
> > >
> > >     // Clean up
> > >     Py_DECREF(pModule);
> > >     Py_DECREF(pName);
> > >
> > >     // Finish the Python Interpreter
> > >
> > >     Py_Finalize();
> > >
> > >     return 0;
> > > }
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > python file saved as
> > >
> > > py_function.py
> > >
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > '''py_function.py - Python source designed to '''
> > > '''demonstrate the use of python embedding'''
> > >
> > > def multiply():
> > >     c = 12345*6789
> > >     print 'The result of 12345 x 6789 :', c
> > >
> > >
> > >
> > >     return c
> > >
> > >
> > >
> >
> /************************************************************************************************************/
> > >
> > > $gcc -g call_function.c  -lpython2.5 -ocall_function
> > > $./call_function py_function.py multiply
> >
> >
> > Shouldn't this be:
> >    $./call_function py_function multiply
> >
> > I.e., you the module name should not have a .py extension.
> >
> > i get following output
> > >
> > > ImportError: No module named py_function
> > > Failed to load "py_function"
> > >
> > > when i kdbg(debug) it i vet NULL value returned from
> PyImport_Import()..
> > >
> > > can any one please find out the error that i am doing.
> > >
> > > --
> > > Shrikant
> > >
> > >
> > > _______________________________________________
> > > C++-sig mailing list
> > > C++-sig at python.org
> > > http://mail.python.org/mailman/listinfo/c++-sig
> > >
> > >
> >
> >
> > --
> > Gustavo J. A. M. Carneiro
> > INESC Porto, Telecommunications and Multimedia Unit
> > "The universe is always one step beyond logic." -- Frank Herbert
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> >
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/0fc3b1ea/attachment-0001.htm
> >
> > ------------------------------
> >
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> >
> >
> > End of C++-sig Digest, Vol 51, Issue 25
> > ***************************************
> >
>
>
>
> --
> Shrikant Chikhalkar
> +91 9850991554
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://mail.python.org/pipermail/c++-sig/attachments/20071025/26af7201/attachment.htm
>
> ------------------------------
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>
>
> End of C++-sig Digest, Vol 51, Issue 26
> ***************************************
>



-- 
Shrikant Chikhalkar
+91 9850991554
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071026/48a8be03/attachment.htm>

From roman.yakovenko at gmail.com  Sun Oct 28 20:51:30 2007
From: roman.yakovenko at gmail.com (Roman Yakovenko)
Date: Sun, 28 Oct 2007 21:51:30 +0200
Subject: [C++-sig] Wrapping out parameter for python access
In-Reply-To: <51C471E6CBEC8148A08E065690F4DBDF016CCA1D@SHAEXC01.ent.sonic.com>
References: <51C471E6CBEC8148A08E065690F4DBDF016CCA1D@SHAEXC01.ent.sonic.com>
Message-ID: <7465b6170710281251l72fbcfc8sc65d5c6a613a1581@mail.gmail.com>

On 10/22/07, Chang Chen <Chang_Chen at sonic.com> wrote:
>
> Hi all
>
> This is a classical issue in boost python. I.e. removeing an argument
> from the function definition and adding the "returned", by the original
> function, value to the return statement of the function-wrapper
>
> For example :
>
> The original function is:
>
> inline void hello_world( std::string& hw ){
>     hw = "hello world!";
> }
>
> And the wrapper is:
>
> inline std::string hello_world_wrapper()
> {
>         std:string temp;
>         ::hello_world(temp);
>         Return temp;
> }
>
> Py++ can help us to generate such wrapper, see
> http://language-binding.net/pyplusplus/documentation/functions/transform
> ation/built_in/output.html
>
> But in this thread
> (http://mail.python.org/pipermail/c++-sig/2005-June/009097.html ), Dave
> provided a start to write a wrapping template.
>
> In My case, only last parameter is out parameter and whether function
> parameter count does not matter. The function signature looks like:
>
> Return function (T1 t1, T2 t2, TOUT& out);
>
> How to write a template for such function and integrate into def?
>
> Any suggestion?


I attach the complete solution to the problem including build script and
unittest.

The key for solution is to use boost::bind library.

P.S. anyone with good knowledge of MPL & PPL, and some free time :-),  is
welcome to improve the code

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071028/ce05c284/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: reducer.zip
Type: application/zip
Size: 1471 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071028/ce05c284/attachment.zip>

From dave at boost-consulting.com  Tue Oct 30 00:31:12 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 29 Oct 2007 19:31:12 -0400
Subject: [C++-sig] Comparing to None on the c++ side
References: <3dc9bcf00710221056n4ca21944sea68faaa38960b1@mail.gmail.com>
Message-ID: <87wst5a473.fsf@grogan.peloton>


on Mon Oct 22 2007, "Matthew Scouten" <matthew.scouten-AT-gmail.com> wrote:

> Right, so...
>
> Python just called a c++ function and passed an object of some sort. I
> want to know if it happens to be a None, so I can take special
> precautions to avoid disaster.
>
> I am trying to use something like:
>
> void foo(object o)
> {
> 	if( object() == o )
> 	{
> 		//Make 3rdparty c++ lib call that prevents
> 		//Sky from falling
> 	}
> 	else
> 	{
> 		//tell 3rdparty c++ lib to
> 		//try to use o to prop up sky...
> 	}
> }
>
> This does not seem to be working like I expected (the sky falls anyway)
>
> So me question:
>
> is object() == o a valid and reliable way to test for None?

I think so.  You might add some "printfs" to help you figure out
what's going on.  Remember, you can use boost::python::eval.

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com



From dave at boost-consulting.com  Tue Oct 30 00:39:17 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 29 Oct 2007 19:39:17 -0400
Subject: [C++-sig] stuffing a PyObject into a object
References: <3dc9bcf00710190938h4e641d9bwb7118c20c1cb707f@mail.gmail.com>
Message-ID: <87r6jda3tm.fsf@grogan.peloton>


on Fri Oct 19 2007, "Matthew Scouten" <matthew.scouten-AT-gmail.com> wrote:

> I can understand this except for a couple of things,
>
> What is the PyIter_Next there for? demonstration purposes? or is it an
> essential part of the conversion? Same question for that whole second
> line.

I think it's a mistake.

> will 'allow_null' make a 'None' out of a null pointer? or do some
> other crazy thing?

No, it makes a null handle.  It seems superfluous, since the object
ctor will throw an exception anyway if its argument is null.

> more to the point, will this code work:
>
> object foo(void* callbackID == NULL)
> {
> 	boost::python::handle<> h(boost::python::allow_null((PyObject*)callbackID))
> 	return boost::python::object o(h);
> }

If the is not pre-incremented for you, then

	boost::python::handle<> h((PyObject*)callbackID));
	return boost::python::object o(h);

would be correct.  Otherwise, you need

	boost::python::handle<> h(boost::python::borrowed((PyObject*)callbackID)));
	return boost::python::object o(h);

> I am trying to understand what is going on, and not just fall into
> guess and check programming.

Good call.

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com



From dave at boost-consulting.com  Tue Oct 30 00:41:57 2007
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 29 Oct 2007 19:41:57 -0400
Subject: [C++-sig] How about class_<...>.enable_copy() ?
References: <200706081706.37240.meine@informatik.uni-hamburg.de>
	<200707231922.15574.meine@informatik.uni-hamburg.de>
	<200707232257.05883.meine@informatik.uni-hamburg.de>
	<200710162237.08845.hans_meine@gmx.net>
Message-ID: <87lk9la3p6.fsf@grogan.peloton>


on Tue Oct 16 2007, Hans Meine <hans_meine-AT-gmx.net> wrote:

> Hi again!
>
> Anyone cares to comment on my older posting?
>
> On Montag 23 Juli 2007, Hans Meine wrote:
>> On Montag 23 Juli 2007, Hans Meine wrote:
>> > foo = Foo(other) # copy constructor as in C++ (1)
>> > foo = copy.copy(other) # pythonic copying (2)
>> > foo = copy.deepcopy(other) # deepcopying (3)
>> >
>> > (1) is easy to do by exporting the copy constructor
>> >
>> > (2) could be implemented by setting __copy__ to the above copy
>> > constructor, but one would assume __dict__ to be (non-deep)copied, too.
>> >
>> > (3) definitely needs its own implementation, which needs to call deepcopy
>> > recursively.
>> >
>> > I would still believe that an .enable_copy() would be very handy.  It
>> > could possibly be controlled by an enum arg which of the above methods
>> > should be exported.
>
> Here is the updated, perfectly working code for (2) and (3), which I propose 
> to be included in BPL.  Also, some things appear hackish to me -- I would be 
> glad to get feedback on how to do this in a more "standard" way:

Is there a reason to have a special interface for this?  Suppose we
just do it unconditionally unless the person wrapping passes
noncopyable to class_<...> ?

-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com



From amohr at pixar.com  Tue Oct 30 01:13:34 2007
From: amohr at pixar.com (Alex Mohr)
Date: Mon, 29 Oct 2007 17:13:34 -0700
Subject: [C++-sig] How about class_<...>.enable_copy() ?
In-Reply-To: <87lk9la3p6.fsf@grogan.peloton>
References: <200706081706.37240.meine@informatik.uni-hamburg.de>
	<200707231922.15574.meine@informatik.uni-hamburg.de>
	<200707232257.05883.meine@informatik.uni-hamburg.de>
	<200710162237.08845.hans_meine@gmx.net>
	<87lk9la3p6.fsf@grogan.peloton>
Message-ID: <4726772E.1060402@pixar.com>

> Is there a reason to have a special interface for this?  Suppose we
> just do it unconditionally unless the person wrapping passes
> noncopyable to class_<...> ?

Speaking of which, is there any good reason why boost.python shouldn't 
correspondingly infer noncopyability if class_metadata<...>::wrapped 
derives from boost::noncopyable?

Alex


From rwgk at yahoo.com  Tue Oct 30 01:35:31 2007
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Mon, 29 Oct 2007 17:35:31 -0700 (PDT)
Subject: [C++-sig] stuffing a PyObject into a object
Message-ID: <64255.35162.qm@web31106.mail.mud.yahoo.com>

> > I can understand this except for a couple of things,
> >
> > What is the PyIter_Next there for? demonstration purposes? or is it
 an
> > essential part of the conversion? Same question for that whole second
> > line.
>
> I think it's a mistake.

This is five years ago, but IIRC my reply was meant as food for thought.
Why does it have to be a two-stage procedure with the detour through handle<>?
My understanding was: it is meant to make you think about PyObject* = 0.

> > will 'allow_null' make a 'None' out of a null pointer? or do some
> > other crazy thing?
>
> No, it makes a null handle.  It seems superfluous, since the object
> ctor will throw an exception anyway if its argument is null.

I wanted to give a realistic example why you may care about PyObject* = 0.
We want to detect PyObject* = 0 without getting an exception (// end of iteration)
*and* we don't want to have the bare PyObject* floating around since we
may forget to decref later.

> > I am trying to understand what is going on, and not just fall into
> > guess and check programming.
>
> Good call.

Guess and check is a fantastic way to find out what is going on
(or as Germany's most famous poet, Goethe, put it: Probieren geht
ueber studieren).

Here is an example:

  void
  foo()
  {
    using namespace boost::python;
    PyObject* obj = 0;
    handle<> h(allow_null(obj));
    std::cout << "HELLO" << std::endl << std::flush;
  }

If you compile and run this function it will finish fine and show the HELLO.
Without the allow_null() I get this:

SystemError: NULL result without error in PyObject_Call

and no HELLO.

Ralf





From hans_meine at gmx.net  Tue Oct 30 08:19:02 2007
From: hans_meine at gmx.net (Hans Meine)
Date: Tue, 30 Oct 2007 08:19:02 +0100
Subject: [C++-sig] How about class_<...>.enable_copy() ?
In-Reply-To: <87lk9la3p6.fsf@grogan.peloton>
References: <200706081706.37240.meine@informatik.uni-hamburg.de>
	<200710162237.08845.hans_meine@gmx.net>
	<87lk9la3p6.fsf@grogan.peloton>
Message-ID: <200710300819.13004.hans_meine@gmx.net>

On Dienstag 30 Oktober 2007, David Abrahams wrote:
> > Here is the updated, perfectly working code for (2) and (3), which I
> > propose to be included in BPL.  Also, some things appear hackish to me --
> > I would be glad to get feedback on how to do this in a more "standard"
> > way:
>
> Is there a reason to have a special interface for this?  Suppose we
> just do it unconditionally unless the person wrapping passes
> noncopyable to class_<...> ?

No; that would be even better (I originally thought about that, too).
The only disadvantage I see is a bit more generated code, but I think that 
copying copyable objects is something that many users expect to be able to 
do.  And for now, I really think that the possibility to disable it using 
noncopyable is enough; otherwise one might introduce an API for that later.

-- 
Ciao, /  /                                                    .o.
     /--/                                                     ..o
    /  / ANS                                                  ooo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071030/16f6a311/attachment.pgp>

From bue.longbow at gmail.com  Mon Oct 29 19:46:47 2007
From: bue.longbow at gmail.com (Bue Vedel-Larsen)
Date: Mon, 29 Oct 2007 18:46:47 +0000 (UTC)
Subject: [C++-sig] Inheritance and __init__
Message-ID: <fg59qn$5hn$1@ger.gmane.org>

I'm having a strange problem with Boost::Python. Consider the following 
Python code:

class Base:
	def foo(self, str):
		print str
		
class Derived(Base):
	def __init__(self):
		self.foo('__init__')

d = Derived()

This works as expected: When run, "__init__" is printed. My problem is 
that if Base is in a C++ module the same code fails with an 
Boost.Python.ArgumentError. See end of post for example code that 
exhibit the error.

Can anyone point out if I'm doing something wrong? I'm not using any 
virtual methods, or any kind of advanced inheritance magic. Is this a 
bug in Boost::Python?

Thanks in advance,
Bue

The exact error is:

Boost.Python.ArgumentError: Python argument types in
    Base.foo(DerivedB, str)
did not match C++ signature:
    foo(class Base {lvalue}, class std::basic_string<char,struct 
std::char_traits<char>,class std::allocator<char> >)

Minimum example code:

#include <boost/python.hpp>
#include <string>
#include <iostream>

class Base {
public:
    Base() {}
    ~Base() {}

    void foo( const std::string& str ) {
        std::cout << "In C++: " << str << std::endl;
    }
};

BOOST_PYTHON_MODULE(test_module) {
    using namespace boost::python;

    (new class_< Base >( "Base", "Base class", init<>() ))
        ->def( "foo", &Base::foo, "method foo" )
    ;
}

void run_script( const std::string& script ) {
    if( PyRun_SimpleString( script.c_str() ) ) {
        PyErr_Print();
        PyErr_Clear();
    }
}

int main() {
    PyImport_AppendInittab( "test_module", &inittest_module );
    Py_Initialize();

    run_script( "from test_module import *\n" );

    // This works
    run_script( 
        "class DerivedA(Base):\n"
        "\tdef bar(self):\n"
        "\t\tself.foo('In Python: DerivedA.bar()')\n"
        "dA = DerivedA()\n"
        "dA.bar()\n"
        );

    // This doesn't work
    run_script( 
        "class DerivedB(Base):\n"
        "\tdef __init__(self):\n"
        "\t\tself.foo('In Python: DerivedB.__init__()')\n"
        "dB = DerivedB()\n"
        );

#ifdef WIN32
    system( "PAUSE" );
#endif
    return 0;
}



From hans_meine at gmx.net  Tue Oct 30 11:37:32 2007
From: hans_meine at gmx.net (Hans Meine)
Date: Tue, 30 Oct 2007 11:37:32 +0100
Subject: [C++-sig] Inheritance and __init__
In-Reply-To: <fg59qn$5hn$1@ger.gmane.org>
References: <fg59qn$5hn$1@ger.gmane.org>
Message-ID: <200710301137.42455.hans_meine@gmx.net>

On Montag 29 Oktober 2007, Bue Vedel-Larsen wrote:
> I'm having a strange problem with Boost::Python. Consider the following
> Python code:
>
> class Base:
> 	def foo(self, str):
> 		print str
>
> class Derived(Base):
> 	def __init__(self):
You must call Base.__init__ here to initialize the base class.
> 		self.foo('__init__')
>
> d = Derived()
>
> This works as expected: When run, "__init__" is printed. My problem is
> that if Base is in a C++ module the same code fails with an
> Boost.Python.ArgumentError. See end of post for example code that
> exhibit the error.
the BPL object is not initialized.

-- 
Ciao, /  /                                                    .o.
     /--/                                                     ..o
    /  / ANS                                                  ooo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071030/90bac390/attachment.pgp>

From Mark.English at rbccm.com  Tue Oct 30 11:38:30 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Tue, 30 Oct 2007 10:38:30 -0000
Subject: [C++-sig] Inheritance and __init__
References: <fg59qn$5hn$1@ger.gmane.org>
Message-ID: <EE7D40281B517A41AF5A90847A61F27937965615@SEW39051.oak.fg.rbc.com>

> -----Original Message-----
> From: c++-sig-bounces+mark.english=rbccm.com at python.org 
> [mailto:c++-sig-bounces+mark.english=rbccm.com at python.org] On 
> Behalf Of Bue Vedel-Larsen
> Sent: 29 October 2007 18:47
> To: c++-sig at python.org
> Subject: [C++-sig] Inheritance and __init__
> 
> I'm having a strange problem with Boost::Python. Consider the 
> following 
> Python code:
> 
> class Base:
> 	def foo(self, str):
> 		print str
> 		
> class Derived(Base):
> 	def __init__(self):
> 		self.foo('__init__')
> 
> d = Derived()
> 
> This works as expected: When run, "__init__" is printed. My 
> problem is 
> that if Base is in a C++ module the same code fails with an 
> Boost.Python.ArgumentError. See end of post for example code that 
> exhibit the error.
> 
<snip>
>     // This works
>     run_script( 
>         "class DerivedA(Base):\n"
>         "\tdef bar(self):\n"
>         "\t\tself.foo('In Python: DerivedA.bar()')\n"
>         "dA = DerivedA()\n"
>         "dA.bar()\n"
>         );
> 
>     // This doesn't work
>     run_script( 
>         "class DerivedB(Base):\n"
>         "\tdef __init__(self):\n"
>         "\t\tself.foo('In Python: DerivedB.__init__()')\n"
>         "dB = DerivedB()\n"
>         );
Does it make a difference if in the __init__ methods you call
"super(KlassName, self).__init__()" first ?
E.g.
"class DerivedB(Base):\n"
"\tdef __init__(self):\n"
"\t\tsuper(DerivedB, self).__init__()
"\t\tself.foo('In Python: DerivedB.__init__()')\n"

Cheers,
mE
________________________________________

This E-Mail (including any attachments) may contain privileged or confidential information.  It is intended only for the addressee(s) indicated above.

The sender does not waive any of its rights, privileges or other protections respecting this information.  

Any distribution, copying or other use of this E-Mail or the information it contains, by other than an intended recipient, is not sanctioned and is prohibited.

If you received this E-Mail in error, please delete it and advise the sender (by return E-Mail or otherwise) immediately. 

This E-Mail (including any attachments) has been scanned for viruses. 

It is believed to be free of any virus or other defect that might affect any computer system into which it is received and opened. 

However, it is the responsibility of the recipient to ensure that it is virus free. 

The sender accepts no responsibility for any loss or damage arising in any way from its use.

E-Mail received by or sent from RBC Capital Markets is subject to review by Supervisory personnel. 

Such communications are retained and may be produced to regulatory authorities or others with legal rights to the information.

IRS CIRCULAR 230 NOTICE:  TO COMPLY WITH U.S. TREASURY REGULATIONS, WE ADVISE YOU THAT ANY U.S. FEDERAL TAX ADVISE INCLUDED IN THIS COMMUNICATION IS NOT INTENDED OR WRITTEN TO BE USED, AND CANNOT BE USED, TO AVOID ANY U.S. FEDERAL TAX PENALTIES OR TO PROMOTE, MARKET, OR RECOMMEND TO ANOTHER PARTY ANY TRANSACTION OR MATTER.


From ndbecker2 at gmail.com  Tue Oct 30 12:54:19 2007
From: ndbecker2 at gmail.com (Neal Becker)
Date: Tue, 30 Oct 2007 07:54:19 -0400
Subject: [C++-sig] properties in python class that extend c++ base
Message-ID: <fg761b$4q7$1@ger.gmane.org>

I have a c++ class with a property:
class_<mod_t> ("mod"...
    .add_property ("SPS", &mod_t::SPS)

If I subclass this class in python:

class X (mod):

Then when python code attempts to access the property in the derived class,
it fails:
Boost.Python.ArgumentError: Python argument types in
    None.None(pi4_modulator)
did not match C++ signature:
    None(modulator<non_offset_modbase, double> {lvalue})

Is there a solution to this?



From nat at lindenlab.com  Tue Oct 30 13:06:04 2007
From: nat at lindenlab.com (Nat Goodspeed)
Date: Tue, 30 Oct 2007 08:06:04 -0400
Subject: [C++-sig] properties in python class that extend c++ base
In-Reply-To: <fg761b$4q7$1@ger.gmane.org>
References: <fg761b$4q7$1@ger.gmane.org>
Message-ID: <47271E2C.4000803@lindenlab.com>

Neal Becker wrote:

> I have a c++ class with a property:
> 
> If I subclass this class in python:
> 
> class X (mod):
> 
> Then when python code attempts to access the property in the derived class,
> it fails:
> Boost.Python.ArgumentError: Python argument types in
>     None.None(pi4_modulator)
> did not match C++ signature:
>     None(modulator<non_offset_modbase, double> {lvalue})

   Does the Python subclass __init__() call the base-class __init__()?


From Mark.English at rbccm.com  Tue Oct 30 13:16:18 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Tue, 30 Oct 2007 12:16:18 -0000
Subject: [C++-sig] properties in python class that extend c++ base
References: <fg761b$4q7$1@ger.gmane.org>
Message-ID: <EE7D40281B517A41AF5A90847A61F27947F8CE90@SEW39051.oak.fg.rbc.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Neal Becker
> Sent: 30 October 2007 11:54
> To: c++-sig at python.org
> Subject: [C++-sig] properties in python class that extend c++ base
> 
> I have a c++ class with a property:
> class_<mod_t> ("mod"...
>     .add_property ("SPS", &mod_t::SPS)
> 
> If I subclass this class in python:
> 
> class X (mod):
> 
> Then when python code attempts to access the property in the 
> derived class, it fails:
> Boost.Python.ArgumentError: Python argument types in
>     None.None(pi4_modulator)
> did not match C++ signature:
>     None(modulator<non_offset_modbase, double> {lvalue})
> 
> Is there a solution to this?
If you overrode __init__ in the subclass, did you remember to call the
Base class's __init__ ?
What's the held type of the class_ ?
________________________________________

This E-Mail (including any attachments) may contain privileged or confidential information.  It is intended only for the addressee(s) indicated above.

The sender does not waive any of its rights, privileges or other protections respecting this information.  

Any distribution, copying or other use of this E-Mail or the information it contains, by other than an intended recipient, is not sanctioned and is prohibited.

If you received this E-Mail in error, please delete it and advise the sender (by return E-Mail or otherwise) immediately. 

This E-Mail (including any attachments) has been scanned for viruses. 

It is believed to be free of any virus or other defect that might affect any computer system into which it is received and opened. 

However, it is the responsibility of the recipient to ensure that it is virus free. 

The sender accepts no responsibility for any loss or damage arising in any way from its use.

E-Mail received by or sent from RBC Capital Markets is subject to review by Supervisory personnel. 

Such communications are retained and may be produced to regulatory authorities or others with legal rights to the information.

IRS CIRCULAR 230 NOTICE:  TO COMPLY WITH U.S. TREASURY REGULATIONS, WE ADVISE YOU THAT ANY U.S. FEDERAL TAX ADVISE INCLUDED IN THIS COMMUNICATION IS NOT INTENDED OR WRITTEN TO BE USED, AND CANNOT BE USED, TO AVOID ANY U.S. FEDERAL TAX PENALTIES OR TO PROMOTE, MARKET, OR RECOMMEND TO ANOTHER PARTY ANY TRANSACTION OR MATTER.


From matthew.scouten at gmail.com  Tue Oct 30 16:58:05 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Tue, 30 Oct 2007 10:58:05 -0500
Subject: [C++-sig] Am I crazy or is it my computer?
Message-ID: <3dc9bcf00710300858j1de24addyfc9970340a76879d@mail.gmail.com>

This is diving me nuts. I need to triple check an assumption.

I am *extending* python using Boost::Python, by wrapping a complex c++
library that was written years ago with no thought of future use in
python.

The library works by a system of callouts and callbacks with extensive
multithreading.

There are a number of minor worker threads in the background, but
[ASSUMPTION ONE:] this should not matter because they never need to
use data or call code that is known to the interpretor.

There is also a callback thread on which the callbacks are called.
This is NOT a python thread, an interpretor thread, or ANY other
special kind of thread. It was not started through the C/Python API.
It is a system thread started through the WIN32 API in the byzantine
depths of the library. It has not been introduced to the interpretor
in any way.

ASSUMPTION TWO:
I can have that thread call into python code by using the
PyGILState_Ensure & PyGILState_Release pair of functions as described
in http://www.python.org/dev/peps/pep-0311/ and
http://docs.python.org/api/threads.html to acquire the Global
Interpretor Lock.

(Actually:  I am using Resource Acquisition Is Initialization idiom
with the following simple class:

class callback
{
public:
	inline callback()
	{
		state = PyGILState_Ensure();
	}

	inline ~callback()
	{
		PyGILState_Release(state);
	}

private:
	PyGILState_STATE state;
};

This may be ASSUMPTION THREE)

Are my assumptions valid?

This may seem like a simple question that I should be googling for,
but like I said I am triple checking some assumptions. I have been
trying to debug the same extremely bizarre  problem for a couple of
days now and it only shows up in the asynchronous callbacks.


From ndbecker2 at gmail.com  Tue Oct 30 19:01:39 2007
From: ndbecker2 at gmail.com (Neal Becker)
Date: Tue, 30 Oct 2007 14:01:39 -0400
Subject: [C++-sig] properties in python class that extend c++ base
References: <fg761b$4q7$1@ger.gmane.org>
	<EE7D40281B517A41AF5A90847A61F27947F8CE90@SEW39051.oak.fg.rbc.com>
Message-ID: <fg7ri4$o3l$1@ger.gmane.org>

English, Mark wrote:

>> -----Original Message-----
>> From: c++-sig-bounces at python.org
>> [mailto:c++-sig-bounces at python.org] On Behalf Of Neal Becker
>> Sent: 30 October 2007 11:54
>> To: c++-sig at python.org
>> Subject: [C++-sig] properties in python class that extend c++ base
>> 
>> I have a c++ class with a property:
>> class_<mod_t> ("mod"...
>>     .add_property ("SPS", &mod_t::SPS)
>> 
>> If I subclass this class in python:
>> 
>> class X (mod):
>> 
>> Then when python code attempts to access the property in the
>> derived class, it fails:
>> Boost.Python.ArgumentError: Python argument types in
>>     None.None(pi4_modulator)
>> did not match C++ signature:
>>     None(modulator<non_offset_modbase, double> {lvalue})
>> 
>> Is there a solution to this?
> If you overrode __init__ in the subclass, did you remember to call the
> Base class's __init__ ?
> What's the held type of the class_ ?

Thanks guys.  Yup, I had forgotten to call Base class __init__.



From joel at boost-consulting.com  Tue Oct 30 18:49:49 2007
From: joel at boost-consulting.com (Joel de Guzman)
Date: Wed, 31 Oct 2007 01:49:49 +0800
Subject: [C++-sig] can't get Boost.Python tutorial's "Hello World"
 example to work (on linux)
In-Reply-To: <ffg955$7f8$1@ger.gmane.org>
References: <ffg955$7f8$1@ger.gmane.org>
Message-ID: <47276EBD.5090007@boost-consulting.com>

ted sandler wrote:
> Hi all,
> 
> I am running linux, and although the Boost.Python "Quickstart" example
> works fine, I am not able to get the Boost.Python "Hello World" example
> working.
> 
> The quickstart example is located here
> http://www.boost.org/libs/python/doc/building.html.
> 
> The hello-world example is located here:
> http://www.boost.org/libs/python/doc/tutorial/doc/html/python/hello.html.
> 
> There are problems with the documentation on the hello-world web page
> (I guess it's out of date).  The example files in the Boost-distribution
> are different from those described in the "hello-world" tutorial.
> Specifically, the Boost distribution contains a "Jamroot" file while the
> web page discusses both a "Jamfile" and a "Jamrules" file (whose
> hyperlink is broken).
> 
> Anyway, when I cd to the "hello-world" example directory and type bjam
> at the prompt, bjam fails with the error message I've listed below.  Any
> help getting me up and running here would be great!

Seems the tutorial is outdated indeed. I'll need to update the
tutorial. In the meantime, see this thread:

     "Trouble with Boost.Python tutorial"

where Dave suggests: "I suggest you work with
http://www.boost.org/libs/python/doc/building.html until the
tutorial gets fixed up."

Regards,
-- 
Joel de Guzman
http://www.boost-consulting.com
http://spirit.sf.net


From bue.longbow at gmail.com  Wed Oct 31 11:45:56 2007
From: bue.longbow at gmail.com (Bue Vedel-Larsen)
Date: Wed, 31 Oct 2007 10:45:56 +0000 (UTC)
Subject: [C++-sig] Inheritance and __init__
References: <fg59qn$5hn$1@ger.gmane.org>
	<EE7D40281B517A41AF5A90847A61F27937965615@SEW39051.oak.fg.rbc.com>
Message-ID: <fg9md4$uil$1@ger.gmane.org>

"English, Mark" <Mark.English at rbccm.com> wrote in
news:EE7D40281B517A41AF5A90847A61F27937965615 at SEW39051.oak.fg.rbc.com: 
>> 
>> I'm having a strange problem with Boost::Python. Consider the 
>> following 
>> Python code:
>> 
>> class Base:
>>      def foo(self, str):
>>           print str
>>           
>> class Derived(Base):
>>      def __init__(self):
>>           self.foo('__init__')
>> 
>> d = Derived()
>> 
<snip>
>>     // This doesn't work
>>     run_script( 
>>         "class DerivedB(Base):\n"
>>         "\tdef __init__(self):\n"
>>         "\t\tself.foo('In Python: DerivedB.__init__()')\n"
>>         "dB = DerivedB()\n"
>>         );
> Does it make a difference if in the __init__ methods you call
> "super(KlassName, self).__init__()" first ?
> E.g.
> "class DerivedB(Base):\n"
> "\tdef __init__(self):\n"
> "\t\tsuper(DerivedB, self).__init__()
> "\t\tself.foo('In Python: DerivedB.__init__()')\n"

Amazing, adding "super(DerivedB, self).__init__()" or "Base.__init__
(self)" before self.foo('In Python: DerivedB.__init__()') makes it work.

But why is this necessary? The pure Python example works without 
explicitly calling the __init__ method of the Base class. Do my class_ 
definition of the module lack something?

Regards,
Bue



From Mark.English at rbccm.com  Wed Oct 31 12:07:33 2007
From: Mark.English at rbccm.com (English, Mark)
Date: Wed, 31 Oct 2007 11:07:33 -0000
Subject: [C++-sig] Inheritance and __init__
References: <fg59qn$5hn$1@ger.gmane.org><EE7D40281B517A41AF5A90847A61F27937965615@SEW39051.oak.fg.rbc.com>
	<fg9md4$uil$1@ger.gmane.org>
Message-ID: <EE7D40281B517A41AF5A90847A61F27947F8CE9E@SEW39051.oak.fg.rbc.com>

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Bue Vedel-Larsen
> Sent: 31 October 2007 10:46
> To: c++-sig at python.org
> Subject: Re: [C++-sig] Inheritance and __init__
> 
> >> class Base:
> >>      def foo(self, str):
> >>           print str
> >>           
> >> class Derived(Base):
> >>      def __init__(self):
> >>           self.foo('__init__')
> >> 
> >> d = Derived()
> >> 
> <snip>
> >>     // This doesn't work
> >>     run_script( 
> >>         "class DerivedB(Base):\n"
> >>         "\tdef __init__(self):\n"
> >>         "\t\tself.foo('In Python: DerivedB.__init__()')\n"
> >>         "dB = DerivedB()\n"
> >>         );
> > Does it make a difference if in the __init__ methods you call 
> > "super(KlassName, self).__init__()" first ?
> > E.g.
> > "class DerivedB(Base):\n"
> > "\tdef __init__(self):\n"
> > "\t\tsuper(DerivedB, self).__init__()
> > "\t\tself.foo('In Python: DerivedB.__init__()')\n"
> 
> Amazing, adding "super(DerivedB, self).__init__()" or 
> "Base.__init__ (self)" before self.foo('In Python: 
> DerivedB.__init__()') makes it work.
> 
> But why is this necessary? The pure Python example works 
> without explicitly calling the __init__ method of the Base 
> class. Do my class_ definition of the module lack something?
> 
> Regards,
> Bue
As far as I know, the Python code working is undefined behaviour.
Whenever overriding __init__() (especially for new style classes which
Boost Python class_ instances implicitly are), call the base class
__init__() passing through arguments as necessary.

If you don't do this, the __init__ of the base class won't get called
which for Boost Python is bad news.

Try this:
class Base(object):
    def __init__(self):
        super(Base, self).__init__()
        self.data = 1
class Derived(Base):
    def __init__(self):
        print 'I am derived'
>>> d = Derived() #Happy
I am derived
>>> d.data #Unhappy

Cheers,
mE
________________________________________

This E-Mail (including any attachments) may contain privileged or confidential information.  It is intended only for the addressee(s) indicated above.

The sender does not waive any of its rights, privileges or other protections respecting this information.  

Any distribution, copying or other use of this E-Mail or the information it contains, by other than an intended recipient, is not sanctioned and is prohibited.

If you received this E-Mail in error, please delete it and advise the sender (by return E-Mail or otherwise) immediately. 

This E-Mail (including any attachments) has been scanned for viruses. 

It is believed to be free of any virus or other defect that might affect any computer system into which it is received and opened. 

However, it is the responsibility of the recipient to ensure that it is virus free. 

The sender accepts no responsibility for any loss or damage arising in any way from its use.

E-Mail received by or sent from RBC Capital Markets is subject to review by Supervisory personnel. 

Such communications are retained and may be produced to regulatory authorities or others with legal rights to the information.

IRS CIRCULAR 230 NOTICE:  TO COMPLY WITH U.S. TREASURY REGULATIONS, WE ADVISE YOU THAT ANY U.S. FEDERAL TAX ADVISE INCLUDED IN THIS COMMUNICATION IS NOT INTENDED OR WRITTEN TO BE USED, AND CANNOT BE USED, TO AVOID ANY U.S. FEDERAL TAX PENALTIES OR TO PROMOTE, MARKET, OR RECOMMEND TO ANOTHER PARTY ANY TRANSACTION OR MATTER.


From neoguru at ioframe.com  Wed Oct 31 13:30:09 2007
From: neoguru at ioframe.com (neoguru)
Date: Wed, 31 Oct 2007 05:30:09 -0700 (PDT)
Subject: [C++-sig] C++ singleton
In-Reply-To: <0IOV00G2X0UGX6@cells-s02.cells.es>
References: <0IOV00G2X0UGX6@cells-s02.cells.es>
Message-ID: <13508067.post@talk.nabble.com>


Yo may find usefull information about Singletin here:
http://sourcemaking.com/design_patterns/singleton
-- 
View this message in context: http://www.nabble.com/-C%2B%2B-sig--C%2B%2B-singleton-tf442256.html#a13508067
Sent from the Python - c++-sig mailing list archive at Nabble.com.



From neoguru at ioframe.com  Wed Oct 31 13:30:56 2007
From: neoguru at ioframe.com (neoguru)
Date: Wed, 31 Oct 2007 05:30:56 -0700 (PDT)
Subject: [C++-sig] C++ singleton
In-Reply-To: <0IOV00G2X0UGX6@cells-s02.cells.es>
References: <0IOV00G2X0UGX6@cells-s02.cells.es>
Message-ID: <13508067.post@talk.nabble.com>


Yo may find usefull information about Singleton here:
http://sourcemaking.com/design_patterns/singleton
-- 
View this message in context: http://www.nabble.com/-C%2B%2B-sig--C%2B%2B-singleton-tf442256.html#a13508067
Sent from the Python - c++-sig mailing list archive at Nabble.com.



From bue.longbow at gmail.com  Wed Oct 31 13:57:44 2007
From: bue.longbow at gmail.com (Bue Vedel-Larsen)
Date: Wed, 31 Oct 2007 12:57:44 +0000 (UTC)
Subject: [C++-sig] Inheritance and __init__
References: <fg9md4$uil$1@ger.gmane.org>
	<EE7D40281B517A41AF5A90847A61F27947F8CE9E@SEW39051.oak.fg.rbc.com>
Message-ID: <fg9u48$tre$1@ger.gmane.org>

"English, Mark" <Mark.English at rbccm.com> wrote in
news:EE7D40281B517A41AF5A90847A61F27947F8CE9E at SEW39051.oak.fg.rbc.com: 

>> But why is this necessary? The pure Python example works 
>> without explicitly calling the __init__ method of the Base 
>> class. Do my class_ definition of the module lack something?
>> 
> As far as I know, the Python code working is undefined behaviour.
> Whenever overriding __init__() (especially for new style classes which
> Boost Python class_ instances implicitly are), call the base class
> __init__() passing through arguments as necessary.
> 
> If you don't do this, the __init__ of the base class won't get called
> which for Boost Python is bad news.
> 

Ah, ok. Thanks for clearing that up.

Regards,
Bue



From matthew.scouten at gmail.com  Wed Oct 31 21:11:27 2007
From: matthew.scouten at gmail.com (Matthew Scouten)
Date: Wed, 31 Oct 2007 15:11:27 -0500
Subject: [C++-sig] Am I crazy or is it my computer?
In-Reply-To: <3dc9bcf00710300858j1de24addyfc9970340a76879d@mail.gmail.com>
References: <3dc9bcf00710300858j1de24addyfc9970340a76879d@mail.gmail.com>
Message-ID: <3dc9bcf00710311311sb096aa1oc4b926b92f3aaaa8@mail.gmail.com>

Ok, I just found this:
http://www.boost.org/libs/python/doc/v2/faq.html#threadsupport

I seems to be saying that I can not do what I want with out patching
Boost::Python (or does this only apply to multiple interpretors? )
Is this still true (please tell me it's been fixed)?
if so, where can I get this patch and how do I apply it?

I asked about multi-threading and boost python in on this list a while ago,
and got no answer. I took this a a "go ahead, no problems". If I am being an
idiot, could you tell me so? Just flame away at me so I know I've been
stupid.  A lot of people here actually know what they are doing. I do not
have that luxury.  I am finding boost python extremely complex and
non-intuitive.  There is also a lack of good documentation (example code is
nice to have, but does not constitute proper documentation all by itself).





On 10/30/07, Matthew Scouten <matthew.scouten at gmail.com> wrote:
> This is diving me nuts. I need to triple check an assumption.
>
> I am *extending* python using Boost::Python, by wrapping a complex c++
> library that was written years ago with no thought of future use in
> python.
>
> The library works by a system of callouts and callbacks with extensive
> multithreading.
>
> There are a number of minor worker threads in the background, but
> [ASSUMPTION ONE:] this should not matter because they never need to
> use data or call code that is known to the interpretor.
>
> There is also a callback thread on which the callbacks are called.
> This is NOT a python thread, an interpretor thread, or ANY other
> special kind of thread. It was not started through the C/Python API.
> It is a system thread started through the WIN32 API in the byzantine
> depths of the library. It has not been introduced to the interpretor
> in any way.
>
> ASSUMPTION TWO:
> I can have that thread call into python code by using the
> PyGILState_Ensure & PyGILState_Release pair of functions as described
> in http://www.python.org/dev/peps/pep-0311/ and
> http://docs.python.org/api/threads.html to acquire the Global
> Interpretor Lock.
>
> (Actually:  I am using Resource Acquisition Is Initialization idiom
> with the following simple class:
>
> class callback
> {
> public:
>         inline callback()
>         {
>                 state = PyGILState_Ensure();
>         }
>
>         inline ~callback()
>         {
>                 PyGILState_Release(state);
>         }
>
> private:
>         PyGILState_STATE state;
> };
>
> This may be ASSUMPTION THREE)
>
> Are my assumptions valid?
>
> This may seem like a simple question that I should be googling for,
> but like I said I am triple checking some assumptions. I have been
> trying to debug the same extremely bizarre  problem for a couple of
> days now and it only shows up in the asynchronous callbacks.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20071031/e1ddf572/attachment.htm>

From rwgk at yahoo.com  Wed Oct 31 23:41:09 2007
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Wed, 31 Oct 2007 15:41:09 -0700 (PDT)
Subject: [C++-sig] Am I crazy or is it my computer?
Message-ID: <642889.6428.qm@web31107.mail.mud.yahoo.com>

I don't know too much about threading, but I think

- you still need Niall Douglas' patches
if your threads want
  to call back into Python
- it is fair to say that the real problem is "Python & threading",

  mainly because of the dreaded global interpreter lock (GIL)

Recently there has been some discussion on python-dev about getting

rid of the GIL, but I'm not holding my breath...


IIUC, if your threads never call back into Python you don't have to

deal with the GIL.
If you have to deal with the GIL I'm not surprised to see you
frustrated. This is something that was largely ignored in the
Boost.Python development (there is only so much you can hope to
achieve in a given amount of time...).



In my work I've consciously avoided threading. Under this simplification
I find Boost.Python well documented and intuitive :-)

Ralf



----- Original Message ----

From: Matthew Scouten <matthew.scouten at gmail.com>

To: Development of Python/C++ integration <C++-sig at python.org>

Sent: Wednesday, October 31, 2007 1:11:27 PM

Subject: Re: [C++-sig] Am I crazy or is it my computer?



 Ok, I just found this: http://www.boost.org/libs/python/doc/v2/faq.html#threadsupport



I seems to be saying that I can not do what I want with out patching Boost::Python (or does this only apply to multiple interpretors? ) 

Is this still true (please tell me it's been fixed)?

if so, where can I get this patch and how do I apply it?



I asked about multi-threading and boost python in on this list a while ago, and got no answer. I took this a a "go ahead, no problems". If I am being an idiot, could you tell me so? Just flame away at me so I know I've been stupid.  A lot of people here actually know what they are doing. I do not have that luxury.  I am finding boost python extremely complex and non-intuitive.  There is also a lack of good documentation (example code is nice to have, but does not constitute proper documentation all by itself).