From paul.bridger at paradise.net.nz  Tue Jun  1 07:51:34 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Tue, 01 Jun 2004 17:51:34 +1200
Subject: [C++-sig] pyste defect: expanding typedefs in generated code
In-Reply-To: <40BA6985.3020306@esss.com.br>
References: <40B9B676.905@paradise.net.nz> <40BA6985.3020306@esss.com.br>
Message-ID: <40BC1966.3010707@paradise.net.nz>

Nicodemus wrote:
> Actually, I believe that now GCCXML exports the declarated type of a 
> typedef, instead of the expanded name. If that's true, only the file 
> declarations.py must be changed, and little at that.

I had a look at this, but didn't really understand what was going on 
sufficiently to fix the problem. From the looks of things I will be gradually 
becoming more acquainted with Pyste's inner child, hopefully at some point I 
will have an opportunity to fix this.

However, I have a fix for another issue I encountered. :)
What is annoying in this case is that I cannot work out the first cause of 
pystes failure. There is a class 'Vector3' that is fairly simple, and defines 
some static member data:

/* Vector3.h */
class Vector3
{
//...
// special points
static const Vector3 ZERO;
static const Vector3 UNIT_X;
static const Vector3 UNIT_Y;
static const Vector3 UNIT_Z;
//...
};

/* Vector3.cpp */
const Vector3 Vector3::ZERO( 0, 0, 0 );
//...

These are exported fine except for an additional 'empty' member. The 
generated code is:
.def_readonly("", &Ogre::Vector3::)
.def_readonly("ZERO", &Ogre::Vector3::ZERO)
.def_readonly("UNIT_X", &Ogre::Vector3::UNIT_X)
.def_readonly("UNIT_Y", &Ogre::Vector3::UNIT_Y)
.def_readonly("UNIT_Z", &Ogre::Vector3::UNIT_Z)

Which obviously fails to compile.
I would guess the original error is in the use of GCCXML output, but the best 
I can do (without a hint or two as to where to look) for a fix is to change 
ClassExporter.py:288 (in ExportVariables) from this:
if self.info[var.name].exclude:
	continue

to this:
if self.info[var.name].exclude or var.name == '':
	continue

Paul Bridger



From paul.bridger at paradise.net.nz  Tue Jun  1 08:21:26 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Tue, 01 Jun 2004 18:21:26 +1200
Subject: [C++-sig] pyste defect: expanding typedefs in generated code
In-Reply-To: <40BC1966.3010707@paradise.net.nz>
References: <40B9B676.905@paradise.net.nz> <40BA6985.3020306@esss.com.br>
	<40BC1966.3010707@paradise.net.nz>
Message-ID: <40BC2066.6030806@paradise.net.nz>

paul.bridger wrote:
> I would guess the original error is in the use of GCCXML output, but the 
> best I can do (without a hint or two as to where to look) for a fix is 
Here's my new best. IMO fixing the error while parsing GCCXML output is 
better, unless members without names have some special significance known at 
the class level.

GCCXMLParser.py:280, in GetMembers:
< if type(decl) in Class.ValidMemberTypes():
 > if type(decl) in Class.ValidMemberTypes() and decl.name != '':

Paul Bridger



From rwgk at yahoo.com  Tue Jun  1 10:29:55 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Tue, 1 Jun 2004 01:29:55 -0700 (PDT)
Subject: [C++-sig] Boost.Python on MacOS X...at all...
In-Reply-To: <F4827076-B333-11D8-9149-000D932927FE@opnet.com>
Message-ID: <20040601082955.34809.qmail@web20212.mail.yahoo.com>

--- Nick Bastin <nbastin at opnet.com> wrote:
> Has anybody succeeded on building Boost.Python on MacOS X at all?

Yes, all the time, but you need Python 2.3 or higher. The Python that comes
with OS 10.3 is fine. You also need the latest Apple compilers. Please look at
the Boost.Python FAQ for details.

> It 
> appears that the linker flags are wrong for creating the dylib (c++ 
> -shared is not correct on MacOS X).

I am using a SCons-based build system. I am not sure if the bjam system is
still in working condition under OS 10, but at one point not too long ago it
was.

Ralf



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From nbastin at opnet.com  Tue Jun  1 10:42:57 2004
From: nbastin at opnet.com (Nick Bastin)
Date: Tue, 1 Jun 2004 04:42:57 -0400
Subject: [C++-sig] Boost.Python on MacOS X...at all...
In-Reply-To: <20040601082955.34809.qmail@web20212.mail.yahoo.com>
References: <20040601082955.34809.qmail@web20212.mail.yahoo.com>
Message-ID: <AE5310FF-B3A7-11D8-B5DC-000D932927FE@opnet.com>


On Jun 1, 2004, at 4:29 AM, Ralf W. Grosse-Kunstleve wrote:

> --- Nick Bastin <nbastin at opnet.com> wrote:
>> Has anybody succeeded on building Boost.Python on MacOS X at all?
>
> Yes, all the time, but you need Python 2.3 or higher. The Python that 
> comes
> with OS 10.3 is fine. You also need the latest Apple compilers. Please 
> look at
> the Boost.Python FAQ for details.

Yes, I've done all of that.  Basically, from what I'm hearing on this 
list, multiple people have it working on OS X, but nobody uses bjam.  
Hardly a ringing endorsement for the stock build system.

I have been unable to make it build using the Python 2.3 that comes 
with MacOS X, or my 2.4 CVS tree, which is really what I'm shooting 
for.  Unfortunately, I don't know a lot about jam, but it appears that 
the tools file for darwin requires that python be in a framework build, 
and can't work off of a standard unix-style installation.  If anybody 
knows otherwise, or how to tweak it to be so, I'd really appreciate 
knowing how.  I'm trying to stage the build, rather than an install, if 
that makes any difference.

--
Nick




From dave at boost-consulting.com  Tue Jun  1 11:56:42 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 01 Jun 2004 05:56:42 -0400
Subject: [C++-sig] Re: std::string crash - SOLVED
References: <012001c444ab$3be53530$6b0aa8c0@vladimirpc>
	<5.1.0.14.0.20040528160225.038c3ec0@mail.chello.be>
	<3432.129.186.232.105.1085762663.squirrel@mail.eng.iastate.edu>
	<003b01c446d4$65178db0$6b0aa8c0@vladimirpc>
Message-ID: <usmdfeh45.fsf@boost-consulting.com>

"Vladimir Ignatov" <vignatov at colorpilot.com> writes:

> BTW I found that Boost.Python can't be compiled using my VC6 - some files
> are fails to compile (sorry, don't have a log handy). Others library are
> builds okay.  (VC7 build Boost.Python just fine).

Known problem; fixed in CVS lo these many months.

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




From paul.bridger at paradise.net.nz  Wed Jun  2 02:24:13 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Wed, 02 Jun 2004 12:24:13 +1200
Subject: [C++-sig] Pyste patch to generate default policy for virtual methods
Message-ID: <40BD1E2D.3060109@paradise.net.nz>

Nicodemus,

For virtual methods, Pyste doesn't
a) complain when they aren't given a policy, and
b) give them the default return_value_policy(copy_const_reference)
Here's a patch for that.
I also made Pyste insert comments in generated code to show where various 
elements are exported. I have found this useful in editing my pyste files, 
but I don't care if you don't want that for mainstream Pyste.

Paul Bridger

ClassExporter.py
275a276,277
 >             if constructors:
 >                 self.Add('inside', '// Additional constructors')
286a289,290
 >         if vars:
 >             self.Add('inside', '// Class Variables')
288c292
<             if self.info[var.name].exclude:
---
 >             if self.info[var.name].exclude:
355a360,361
 >         if methods:
 >             self.Add('inside', '// Methods')
365c371
<
---
 >
425a432
 >             self.Add('inside', '// Virtual Methods')
505a513,514
 >         if operators:
 >             self.Add('inside', '// Operators')
778,779c787,792
<         # Add policy to overloaded methods also
<         policy = self.info[method.name].policy or ''
---
 >         # warn the user if this method needs a policy and doesn't have one
 >         method_info = self.info[method.name]
 >         method_info.policy = exporterutils.HandlePolicy(method, 
method_info.policy)
 >
 >         # Add policy to overloaded methods also
 >         policy = method_info.policy or ''



From igor at pcigeomatics.com  Wed Jun  2 17:39:58 2004
From: igor at pcigeomatics.com (Igor Lapshin)
Date: Wed, 2 Jun 2004 11:39:58 -0400
Subject: [C++-sig] Linker errors with VC7.1 (.NET) and Python.Boost 1.31.0
Message-ID: <BF0BED91AA17F448A3A927857E11FC64CAD8F4@thor.pcigeomatics.com>

Hi, 

 

I'm trying to compile the following code within  MS Visual Studio IDE
.NET 2003. I compiled Boost 1.31.0 usign bjam, but I have the
requirement that all exposed into Python classes must be compiled within
IDE. "no_init" was specified because the original class has pure virtual
functions, but I get the same errors even without them. The original
class links and works fine on Linux/gcc, so I suppose I have problems
specific to VC7.1 and Windows.


class myclass 
{
    const char* name;

public:
    myclass();
    const char* getName();
};

myclass::myclass(): name("I'm a class") 
{}

const char *myclass::getName()
{ 
return name; 
}

using namespace boost::python;
BOOST_PYTHON_MODULE(testme)
{
  class_<myclass,boost::noncopyable>("myclass", no_init)
 .def("getName", &myclass::getName)
 ;
}

I specified libboost_python-vc71-mt-gd.lib as link library.

 

test_boost.obj : error LNK2019: unresolved external symbol
"__declspec(dllimport) void __cdecl
boost::python::objects::register_dynamic_id_aux(struct
boost::python::type_info,struct std::pair<void *,struct
boost::python::type_info> (__cdecl*)(void *))"
(__imp_?register_dynamic_id_aux at objects@python at boost@@YAXUtype_info at 23@P
6A?AU?$pair at PAXUtype_info@python at boost@@@std@@PAX at Z@Z) referenced in
function "void __cdecl boost::python::objects::register_dynamic_id<class
myclass>(class myclass *)"
(??$register_dynamic_id at Vmyclass@@@objects at python@boost@@YAXPAVmyclass@@
@Z)

 

Am I missing something? Do I always need to write a wrapper class for
"no_init"? 

Thanks, 

Igor

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040602/78a97008/attachment.htm>

From igor at pcigeomatics.com  Wed Jun  2 19:16:55 2004
From: igor at pcigeomatics.com (Igor Lapshin)
Date: Wed, 2 Jun 2004 13:16:55 -0400
Subject: [C++-sig] Linker errors with VC7.1 (.NET) and Python.Boost 1.31.0
Message-ID: <BF0BED91AA17F448A3A927857E11FC64CAD900@thor.pcigeomatics.com>

Hello all, 

I solved the problem: I added python_boost.lib library into the project.

I am still wondering why this library was not put into Boost/lib
directory and why my project worked on Linux without that library (I
also used non-standard build mechanism and only a single
boost_python-gcc-mt-d library)

Regards,

Igor

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040602/ff3b2ca3/attachment.htm>

From rutt at 168-226-10-180.speedy.com.ar  Fri Jun  4 16:16:17 2004
From: rutt at 168-226-10-180.speedy.com.ar (rutt at 168-226-10-180.speedy.com.ar)
Date: Fri, 04 Jun 2004 14:16:17 +0000
Subject: [C++-sig] software
In-Reply-To: <416AA6DCB8D1L6HC@python.org>
References: <416AA6DCB8D1L6HC@python.org>
Message-ID: <LLK4G1HK2K4LAK8E@168-226-10-180.speedy.com.ar>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://HEEHLM.biz/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://NANNEH.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


vrnnhidj xwrrg vdjyunu vtfwj clctknd ddqj lvujrl 
cztf ihaxbvg hrqje trotwrln dyiy icjkbtq 
kaqibmo ifh tjal zitrvwv ugtdt e ojxhv 




From paul.bridger at paradise.net.nz  Fri Jun  4 06:23:25 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Fri, 04 Jun 2004 16:23:25 +1200
Subject: [C++-sig] Pyste exception handlers
Message-ID: <40BFF93D.7030506@paradise.net.nz>

Nicodemus,

How would you recommend exception handler registration be implemented in Pyste?

A problem here would be what to do in the differing cases of:
- building a single pyste file into a module
: in this case specifying exception handlers in the pyste file is sufficient
- building many pyste files into a single module
: in this case you probably want exception handlers to be specified in a 
module-level file, or perhaps you could put them in an arbitrary pyste file

I think Pyste has some deficiencies in organising modules from multiple 
files. It is wonderful that this can be done, but I don't like that so much 
has to be specified on the command-line (and that the meaning of command-line 
options changes depending on --multiple or not).

I think a better organisation would involve a module-level file, which could 
then contain such things as exception handlers. This would also make building 
modules out of several pyste files cleaner. (A disadvantage is that it 
expands the scope of the tool into something of a build tool).
What do others think?

Paul Bridger



From AnToine at alexion.nl  Fri Jun  4 14:35:43 2004
From: AnToine at alexion.nl (AnToine van Maarle)
Date: Fri, 4 Jun 2004 14:35:43 +0200
Subject: [C++-sig] (no subject)
Message-ID: <F152FBD2CB1A1C4DB2E04BD0CBF31E220C7834@w2000srv.alexion.lokaal>

Hi folks,

I don't realy kno if this is the correct mailinglist, but as far as I can see you work with c++ and this error message is a c++ something, maybe one of you can help, please?

Thanx

AnToine

====================== error message ====================================

C:\Program Files\Alexion Software\Relation Manager\program>arm.exe 
Extracting binaries
C:\Program Files\Alexion Software\Relation Manager\program\python23.dll
Manipulating evironment
PYTHONPATH=C:/Program Files/Alexion Software/Relation Manager/program
importing modules from CArchive
iu
archive
Installing import hooks
out1.pyz
Running scripts
bin\buildstdapp\out1.pyz/fcntl:7: DeprecationWarning: the FCNTL module is deprecated; please use fcntl
Traceback (most recent call last):
  File "<string>", line 90, in ?
  File "bin\buildstdapp\out1.pyz/wxPython.wx", line 1957, in __init__
  File "<string>", line 66, in OnInit
  File "bin\buildstdapp\out1.pyz/pynx.gui.frame", line 97, in __init__
  File "bin\buildstdapp\out1.pyz/wxPython.stattool", line 227, in Realize
wxPython.wxc.wxPyAssertionError: C++ assertion "wxAssertFailure" failed in e:\Pr
ojects\wx2.4\src\msw\tbar95.cpp(582): invalid tool button bitmap
 RC: -1 from main
OK.
23:02:35: Debug: e:\Projects\wx2.4\src\msw\app.cpp(407): 'UnregisterClass(MDI parent)' failed with error 0x00000584 (de klasse heeft nog geopende vensters.).






From grafik.list at redshift-software.com  Fri Jun  4 16:53:58 2004
From: grafik.list at redshift-software.com (Rene Rivera)
Date: Fri, 04 Jun 2004 09:53:58 -0500
Subject: [C++-sig] (no subject)
In-Reply-To: <F152FBD2CB1A1C4DB2E04BD0CBF31E220C7834@w2000srv.alexion.lokaal>
References: <F152FBD2CB1A1C4DB2E04BD0CBF31E220C7834@w2000srv.alexion.lokaal>
Message-ID: <40C08D06.6000105@redshift-software.com>

AnToine van Maarle wrote:
> Hi folks,
> 
> I don't realy kno if this is the correct mailinglist, but as far as I
> can see you work with c++ and this error message is a c++ something,
> maybe one of you can help, please?

That looks like a wxWidgest error. Ask on the wx-users or wx-dev 
lists... See: wxwidgets.org for details.


-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq



From AnToine at alexion.nl  Fri Jun  4 18:38:15 2004
From: AnToine at alexion.nl (AnToine van Maarle)
Date: Fri, 4 Jun 2004 18:38:15 +0200
Subject: [C++-sig] help
Message-ID: <F152FBD2CB1A1C4DB2E04BD0CBF31E220C7840@w2000srv.alexion.lokaal>

ok

-----Oorspronkelijk bericht-----
Van: c++-sig-bounces at python.org
[mailto:c++-sig-bounces at python.org]Namens c++-sig-request at python.org
Verzonden: vrijdag 4 juni 2004 18:05
Aan: c++-sig at python.org
Onderwerp: C++-sig Digest, Vol 11, Issue 4


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. software (rutt at 168-226-10-180.speedy.com.ar)
   2. Pyste exception handlers (paul.bridger)
   3. (no subject) (AnToine van Maarle)
   4. Re: (no subject) (Rene Rivera)


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

Message: 1
Date: Fri, 04 Jun 2004 14:16:17 +0000
From: rutt at 168-226-10-180.speedy.com.ar
Subject: [C++-sig] software
To: C++-sig <c++-sig at python.org>
Message-ID: <LLK4G1HK2K4LAK8E at 168-226-10-180.speedy.com.ar>
Content-Type: text/plain

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://HEEHLM.biz/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://NANNEH.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


vrnnhidj xwrrg vdjyunu vtfwj clctknd ddqj lvujrl 
cztf ihaxbvg hrqje trotwrln dyiy icjkbtq 
kaqibmo ifh tjal zitrvwv ugtdt e ojxhv 




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

Message: 2
Date: Fri, 04 Jun 2004 16:23:25 +1200
From: "paul.bridger" <paul.bridger at paradise.net.nz>
Subject: [C++-sig] Pyste exception handlers
To: Development of Python/C++ integration <c++-sig at python.org>
Message-ID: <40BFF93D.7030506 at paradise.net.nz>
Content-Type: text/plain; format=flowed; charset=us-ascii

Nicodemus,

How would you recommend exception handler registration be implemented in Pyste?

A problem here would be what to do in the differing cases of:
- building a single pyste file into a module
: in this case specifying exception handlers in the pyste file is sufficient
- building many pyste files into a single module
: in this case you probably want exception handlers to be specified in a 
module-level file, or perhaps you could put them in an arbitrary pyste file

I think Pyste has some deficiencies in organising modules from multiple 
files. It is wonderful that this can be done, but I don't like that so much 
has to be specified on the command-line (and that the meaning of command-line 
options changes depending on --multiple or not).

I think a better organisation would involve a module-level file, which could 
then contain such things as exception handlers. This would also make building 
modules out of several pyste files cleaner. (A disadvantage is that it 
expands the scope of the tool into something of a build tool).
What do others think?

Paul Bridger



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

Message: 3
Date: Fri, 4 Jun 2004 14:35:43 +0200
From: "AnToine van Maarle" <AnToine at alexion.nl>
Subject: [C++-sig] (no subject)
To: <c++-sig at python.org>
Message-ID:
	<F152FBD2CB1A1C4DB2E04BD0CBF31E220C7834 at w2000srv.alexion.lokaal>
Content-Type: text/plain;	charset="iso-8859-1"

Hi folks,

I don't realy kno if this is the correct mailinglist, but as far as I can see you work with c++ and this error message is a c++ something, maybe one of you can help, please?

Thanx

AnToine

====================== error message ====================================

C:\Program Files\Alexion Software\Relation Manager\program>arm.exe 
Extracting binaries
C:\Program Files\Alexion Software\Relation Manager\program\python23.dll
Manipulating evironment
PYTHONPATH=C:/Program Files/Alexion Software/Relation Manager/program
importing modules from CArchive
iu
archive
Installing import hooks
out1.pyz
Running scripts
bin\buildstdapp\out1.pyz/fcntl:7: DeprecationWarning: the FCNTL module is deprecated; please use fcntl
Traceback (most recent call last):
  File "<string>", line 90, in ?
  File "bin\buildstdapp\out1.pyz/wxPython.wx", line 1957, in __init__
  File "<string>", line 66, in OnInit
  File "bin\buildstdapp\out1.pyz/pynx.gui.frame", line 97, in __init__
  File "bin\buildstdapp\out1.pyz/wxPython.stattool", line 227, in Realize
wxPython.wxc.wxPyAssertionError: C++ assertion "wxAssertFailure" failed in e:\Pr
ojects\wx2.4\src\msw\tbar95.cpp(582): invalid tool button bitmap
 RC: -1 from main
OK.
23:02:35: Debug: e:\Projects\wx2.4\src\msw\app.cpp(407): 'UnregisterClass(MDI parent)' failed with error 0x00000584 (de klasse heeft nog geopende vensters.).






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

Message: 4
Date: Fri, 04 Jun 2004 09:53:58 -0500
From: Rene Rivera <grafik.list at redshift-software.com>
Subject: Re: [C++-sig] (no subject)
To: Development of Python/C++ integration <c++-sig at python.org>
Message-ID: <40C08D06.6000105 at redshift-software.com>
Content-Type: text/plain; charset=us-ascii; format=flowed

AnToine van Maarle wrote:
> Hi folks,
> 
> I don't realy kno if this is the correct mailinglist, but as far as I
> can see you work with c++ and this error message is a c++ something,
> maybe one of you can help, please?

That looks like a wxWidgest error. Ask on the wx-users or wx-dev 
lists... See: wxwidgets.org for details.


-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq



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

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


End of C++-sig Digest, Vol 11, Issue 4
**************************************



From nick at rockstarvancouver.com  Sat Jun  5 02:02:39 2004
From: nick at rockstarvancouver.com (Nick Trout)
Date: Fri, 4 Jun 2004 17:02:39 -0700
Subject: [C++-sig] Embedding in .NET architecture
Message-ID: <911F8C8EB7A8084AAEDD55CEDC54D8F851FE2D@iggy.rockstarvancouver.com>


Hello!

I'm trying to embed Python into Windows Forms Application I'm working
on. .NET is relatively new to me and having just got over the Managed
objects vs unmanaged objects problems I'm now having trouble getting
Boost working.

My objective is to start up an application and then use Python scripts
to customise the application and do various processing. I want to expose
various features of the application to Python to enable this. Given what
I am trying to do I think it's more appropriate to embed than to extend.

I've been through the archives and looked at the embedding.cpp test
example that is in the Boost source but I can't seem to find any
information which fixes my problem. When I run the test suite the
embedding.cpp example works when bjam compiles and runs it. However,
when I copy the example into my code and run it throws and exception
which crashes the application.

I'm trying to get my application running using multithreaded DLL
support. I've tried linking Boost against libboost statically linked
libraries and against the boost_python dynamic libraries but I keep
getting exceptions of various kinds (TEH, missing types etc). I know
Boost works with MSVC 7.1 because other people seem to be using it and
all the tests suite worked for me. What I haven't been able to do is get
Boost.Python working in a MSVC app that uses managed objects. I'm
therefore assuming that I don't have the correct settings, am missing
some configuration #define or am just too much of a novice at the .NET
framework to understand what the problem is.

One thing I haven't tried, which is repeated on the mailing list is to
build my application using bjam. I haven't tried this as I don't know
how it would work with all the "assemblies" and stuff that goes into a
WFA. I really don't want to have to do this if I can avoid it.

Current problem I am having: When I run the embedded code in my app:

    if (python::handle_exception( test ))
    {
        if (PyErr_Occurred())
            PyErr_Print();
        return 1;
    }

The exception handler fails because apparently test provides an invalid
(NULL) pointer so f() has nothing to call in the try-catch handler.

If I call test() directly in:

<<<<
    // Register the module with the interpreter
    if (PyImport_AppendInittab("embedded_hello", initembedded_hello) ==
-1)
        throw std::runtime_error("Failed to add embedded_hello to the
interpreter's "
                                 "builtin modules");

    // Initialize the interpreter
    Py_Initialize();

    // Retrieve the main module
    python::object main_module = python::extract<python::object>(
        PyImport_AddModule("__main__")
    )();
    
    // Retrieve the main module's namespace
    python::object main_namespace(main_module.attr("__dict__"));
>>>>

Execution gets as far as "python::object main_module..." and I get the
following error:

<<<<
An unhandled exception of type 'System.TypeLoadException' occurred in
WorldEditor.exe

Additional information: Could not load type
boost.python.detail.new_non_null_reference_t from assembly WorldEditor,
Version=1.0.1616.28683, Culture=neutral, PublicKeyToken=null.
>>>>

I've tried to set all of the correct project settings. I have
asynchronous exeception handling turned on (/EHa) as opposed to the
default (/EHsc) and RTTI is on.

I hope you can help please. It may be obvious to a .NET expert. If not,
I'll probably give this another day and move to PyCXX (and then Lua!)

Nick





From bob at redivi.com  Sat Jun  5 02:22:06 2004
From: bob at redivi.com (Bob Ippolito)
Date: Fri, 4 Jun 2004 20:22:06 -0400
Subject: [C++-sig] Embedding in .NET architecture
In-Reply-To: <911F8C8EB7A8084AAEDD55CEDC54D8F851FE2D@iggy.rockstarvancouver.com>
References: <911F8C8EB7A8084AAEDD55CEDC54D8F851FE2D@iggy.rockstarvancouver.com>
Message-ID: <60403A5E-B686-11D8-BBA0-000A95686CD8@redivi.com>


On Jun 4, 2004, at 8:02 PM, Nick Trout wrote:

> I hope you can help please. It may be obvious to a .NET expert. If not,
> I'll probably give this another day and move to PyCXX (and then Lua!)

Uh, why not just try one of the Python .NET projects, like this: 
http://www.zope.org/Members/Brian/PythonNet/index_html

There may or may not be others, I have no experience with any of them.

-bob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2357 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040604/696b0b1c/attachment.bin>

From nick at rockstarvancouver.com  Sat Jun  5 02:24:49 2004
From: nick at rockstarvancouver.com (Nick Trout)
Date: Fri, 4 Jun 2004 17:24:49 -0700
Subject: [C++-sig] Embedding in .NET architecture
Message-ID: <911F8C8EB7A8084AAEDD55CEDC54D8F881FD9B@iggy.rockstarvancouver.com>



> Behalf Of Nick Trout
> I'm trying to embed Python into Windows Forms Application I'm working
> on. .NET is relatively new to me and having just got over the Managed
> objects vs unmanaged objects problems I'm now having trouble getting
> Boost working.

Aggggggghhhhhhhh.

And arrrghhhhgghhhh again.

You need to put: 

#pragma unmanaged

before the Boost headers and around any code that uses Boost.

Jeez. I'm glad .NET is making my life easier and me more productive.

Nick





From nick at rockstarvancouver.com  Sat Jun  5 02:28:06 2004
From: nick at rockstarvancouver.com (Nick Trout)
Date: Fri, 4 Jun 2004 17:28:06 -0700
Subject: [C++-sig] Embedding in .NET architecture
Message-ID: <911F8C8EB7A8084AAEDD55CEDC54D8F881FD9C@iggy.rockstarvancouver.com>


> Behalf Of Bob Ippolito
> On Jun 4, 2004, at 8:02 PM, Nick Trout wrote:
> 
> > I hope you can help please. It may be obvious to a .NET expert. If
not,
> > I'll probably give this another day and move to PyCXX (and then
Lua!)
> 
> Uh, why not just try one of the Python .NET projects, like this:
> http://www.zope.org/Members/Brian/PythonNet/index_html

Oh. Very interesting, I shall certainly look at it in more detail.
Thanks Bob!

> There may or may not be others, I have no experience with any of them.

I did come across IronPython, which is supposed to run off the .NET CLR.
Rather surprisingly it seems more efficient than the Python VM (except
for compilation/eval).

Nick




From bob at redivi.com  Sat Jun  5 03:06:44 2004
From: bob at redivi.com (Bob Ippolito)
Date: Fri, 4 Jun 2004 21:06:44 -0400
Subject: [C++-sig] Embedding in .NET architecture
In-Reply-To: <911F8C8EB7A8084AAEDD55CEDC54D8F881FD9C@iggy.rockstarvancouver.com>
References: <911F8C8EB7A8084AAEDD55CEDC54D8F881FD9C@iggy.rockstarvancouver.com>
Message-ID: <9C876726-B68C-11D8-BBA0-000A95686CD8@redivi.com>


On Jun 4, 2004, at 8:28 PM, Nick Trout wrote:

>
>> Behalf Of Bob Ippolito
>> On Jun 4, 2004, at 8:02 PM, Nick Trout wrote:
>>
>>> I hope you can help please. It may be obvious to a .NET expert. If
> not,
>>> I'll probably give this another day and move to PyCXX (and then
> Lua!)
>>
>> Uh, why not just try one of the Python .NET projects, like this:
>> http://www.zope.org/Members/Brian/PythonNet/index_html
>
> Oh. Very interesting, I shall certainly look at it in more detail.
> Thanks Bob!
>
>> There may or may not be others, I have no experience with any of them.
>
> I did come across IronPython, which is supposed to run off the .NET 
> CLR.
> Rather surprisingly it seems more efficient than the Python VM (except
> for compilation/eval).

IronPython will be much more useful when Jim decides to release it :)

-bob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2357 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040604/714300b0/attachment.bin>

From nick at rockstarvancouver.com  Sat Jun  5 04:36:11 2004
From: nick at rockstarvancouver.com (Nick Trout)
Date: Fri, 4 Jun 2004 19:36:11 -0700
Subject: [C++-sig] Embedding in .NET architecture
Message-ID: <911F8C8EB7A8084AAEDD55CEDC54D8F851FE2E@iggy.rockstarvancouver.com>



> On Behalf Of Bob Ippolito
> IronPython will be much more useful when Jim decides to release it :)

So, from what I can make out IronPython looks like it compiles into .NET
CLR VM code and PythonNet is Python, with an interface to .NET CLR Libs.
I can't see any features in IronPython that I need over PythonNet. 

IronPython is an "unreleased research prototype". So that solves that
problem of working out which to use!

PythonNet is a one way interface from Python to CLR libraries, which is
all I need for embedding. It does support access to PyObject* though so
you could delve into Python if need be. I think it uses unmanaged
objects.

IronPython, compiled in CLR code looks like it is completely managed and
closer integrated to the CLR model.

So I suppose I have to decide if, now I have Boost working, it's more
work to get PythonNet working and support it, rather than support Boost.

Nick





From texhax at fruitfly.BDGP.berkeley.edu  Sun Jun  6 04:51:35 2004
From: texhax at fruitfly.BDGP.berkeley.edu (texhax at fruitfly.BDGP.berkeley.edu)
Date: Sun, 06 Jun 2004 02:51:35 +0000
Subject: [C++-sig] software
In-Reply-To: <1JFLD3K1F922EEA5@python.org>
References: <1JFLD3K1F922EEA5@python.org>
Message-ID: <5DLLI3J59F3HJEBL@fruitfly.berkeley.edu>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://FLLMNA.info/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://NANNEH.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


hhfggpxl yowdk phwotay naxje bqtogqw nxcv matkpe 
ocvw ykpziez hoeth xxmouclx yayg lnjcqgz 
oqoobuo tit ppqm egthtff udkgr w tictz 




From Peter.Vandersteegen at intec.Ugent.be  Mon Jun  7 09:59:40 2004
From: Peter.Vandersteegen at intec.Ugent.be (Peter Vandersteegen)
Date: Mon, 07 Jun 2004 09:59:40 +0200
Subject: [C++-sig] problems compiling simple example in boost-python
 documentation/tutorial...
Message-ID: <6.1.0.6.0.20040607092433.026648c0@pop-tech.intec.UGent.be>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040607/6234f1f2/attachment.htm>

From g.simpson at zoo-tech.com  Mon Jun  7 12:48:26 2004
From: g.simpson at zoo-tech.com (Gareth Simpson)
Date: Mon, 7 Jun 2004 11:48:26 +0100 
Subject: [C++-sig] Tedious Newbie Question
Message-ID: <176AB0A0ADC4734B8FE888D53E75D9D5017EB1F1@s15shefsrv.zoodigitalgroup.com>

Apologies in advance if this is not the list to be asking these sort of
things.

I am trying to use Boost.Python to embed a Python library into my C++
application.  The library in question uses unicode strings internally and I
can't for the life of me work out how to get them out and into my C++.
Google is largely silent on the matter.

The below code snippet shows what I'm trying to do.

I have no problem using extract<> to get a python string into a std::string
but trying the same thing with unicode python strings and std::wstrings
throws an exception.

Clearly I'm going about it the wrong way - what should I be doing?

Thanks in advance

Gareth





#include "stdafx.h"

#include "boost/python.hpp"

int _tmain(int argc, _TCHAR* argv[])
{
	using namespace boost::python;
	Py_Initialize();

	object mainModule (borrowed(PyImport_AddModule("__main__")));
	object mainNamespace = mainModule.attr("__dict__");

	handle<>(PyRun_String( 
		"str1 = 'a string'\n"	// a simple string
		"str2 = u'a unicode string'\n", // a unicode string
		Py_file_input , mainNamespace.ptr(), mainNamespace.ptr() ));

	std::string str1  = extract<std::string>(mainNamespace["str1"]); //
this works fine
	std::wstring str2  = extract<std::wstring>(mainNamespace["str2"]);
// boom!

	Py_Finalize();

	return 0;
}



From AnToine at alexion.nl  Mon Jun  7 15:40:58 2004
From: AnToine at alexion.nl (AnToine van Maarle)
Date: Mon, 7 Jun 2004 15:40:58 +0200
Subject: [C++-sig] RE: C++-sig Digest, Vol 11, Issue 6
Message-ID: <F152FBD2CB1A1C4DB2E04BD0CBF31E220C7849@w2000srv.alexion.lokaal>

thnx

-----Oorspronkelijk bericht-----
Van: c++-sig-bounces at python.org
[mailto:c++-sig-bounces at python.org]Namens c++-sig-request at python.org
Verzonden: zaterdag 5 juni 2004 18:02
Aan: c++-sig at python.org
Onderwerp: C++-sig Digest, Vol 11, Issue 6


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. software (texhax at fruitfly.BDGP.berkeley.edu)


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

Message: 1
Date: Sun, 06 Jun 2004 02:51:35 +0000
From: texhax at fruitfly.BDGP.berkeley.edu
Subject: [C++-sig] software
To: C++-sig <c++-sig at python.org>
Message-ID: <5DLLI3J59F3HJEBL at fruitfly.berkeley.edu>
Content-Type: text/plain

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://FLLMNA.info/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://NANNEH.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


hhfggpxl yowdk phwotay naxje bqtogqw nxcv matkpe 
ocvw ykpziez hoeth xxmouclx yayg lnjcqgz 
oqoobuo tit ppqm egthtff udkgr w tictz 




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

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


End of C++-sig Digest, Vol 11, Issue 6
**************************************



From Peter.Vandersteegen at intec.Ugent.be  Mon Jun  7 16:23:51 2004
From: Peter.Vandersteegen at intec.Ugent.be (Peter Vandersteegen)
Date: Mon, 07 Jun 2004 16:23:51 +0200
Subject: [C++-sig] problems compiling simple example in boost-python 
	documentation/tutorial...
Message-ID: <6.1.0.6.0.20040607162218.00a249d0@pop-tech.intec.UGent.be>

Hello,

Sorry for the previous mail.  I didn't realize a html-email would be scrubbed.
Therefore the original text is sent in plain text

Peter...



As a boost-python beginner I've started exploring the tutorial.  I however 
had trouble compiling a very basic example from the tutorial:
http://www.boost.org/libs/python/doc/tutorial/doc/virtual_functions_with_default_implementations.html
Changing some lines in the tutorial code however solved the compiling 
problem.   Now my question is:
Does this relate to my compiler: intel 7.1, msvc7.0? Or something I 
overlooked in my code?

I've started with a pure virtual base class, and indeed the virtual 
function can be overriden in python and called with a function in c++
... (fantastic :) )  No problems at all...

I recieve however compiler-errors (explained after program) when I however 
try to compile the exact example from the tutorial ...(see link and 
following text)

// file wraptestpoly.cpp
#include <boost/python.hpp>
using namespace boost::python;

struct Base
{
virtual int f(void){return 0;}
}; //*

int call_f(Base& b) { return b.f(); }


struct BaseWrap : Base
     {
         BaseWrap(PyObject* self_)
             : self(self_) {}
         virtual int f() { return call_method<int>(self, "f"); }
         int default_f() { return Base::f(); }
         PyObject* self;
     };
BOOST_PYTHON_MODULE(wraptestpoly){
class_<Base, BaseWrap>("Base")  //*1
         .def("f", &Base::f, &BaseWrap::default_f)
def("call_f", call_f);
}
//end file
The error I recieve is:

D:\libraries\boost_1_31_0\boost/python/object/value_holder.hpp(134): error: 
no instance of constructor "BaseWrap::BaseWrap" matches the argument list 
argument types are: (PyObject *, boost::reference_wrapper<const Base>::type)

If I however  use
         class_<Base, BaseWrap,boost::noncopyable>("Base")
instead of //*1 --> no compiling problems.
Another option which works is adding a copy-constructor:
                BaseWrap(PyObject* self_, Base const& copy)  : Base(copy), 
self(self_) {}
in the wrapperclass.

Now my question is: what did I do wrong with this simple example?  Could 
this problem arise from an outdated tutorial?

thx a lot

Peter  




From Pere.Mato at cern.ch  Mon Jun  7 18:20:54 2004
From: Pere.Mato at cern.ch (Pere Mato Vila)
Date: Mon, 7 Jun 2004 18:20:54 +0200
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
Message-ID: <5BF9E41B6C9F2546B624F30FE840C122351BAF@cernxchg21.cern.ch>

I am having the following compilation warnings when including
"boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 and gcc 3.2. Is
anything I am doing wrong?


In file included from
/afs/cern.ch/sw/lcg/external/Python/2.3.3/rh73_gcc32/include/python2.3/P
ython.h:8,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/detail/wrap_python.hpp:121,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/detail/prefix.hpp:13,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/args.hpp:9,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python.hpp:12,
                 from test.cpp:1:
/afs/cern.ch/sw/lcg/external/Python/2.3.3/rh73_gcc32/include/python2.3/p
yconfig.h:847:1: warning: "_POSIX_C_SOURCE" redefined
In file included from /usr/include/limits.h:26,
                 from
/usr/local/gcc-alt-3.2/lib/gcc-lib/i686-pc-linux-gnu/3.2/include/limits.
h:132,
                 from
/usr/local/gcc-alt-3.2/lib/gcc-lib/i686-pc-linux-gnu/3.2/include/syslimi
ts.h:7,
                 from
/usr/local/gcc-alt-3.2/lib/gcc-lib/i686-pc-linux-gnu/3.2/include/limits.
h:11,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/detail/wrap_python.hpp:27,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/detail/prefix.hpp:13,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python/args.hpp:9,
                 from
/afs/cern.ch/sw/lcg/external/Boost/1.31.0-python233/rh73_gcc32/include/b
oost-1_31/boost/python.hpp:12,
                 from test.cpp:1:
/usr/include/features.h:131:1: warning: this is the location of the
previous definition



------------------------------------------------------------
Pere Mato  CERN, PH Department, CH 1211 Geneva 23, Switzerland
           e-mail: Pere.Mato at cern.ch    tel: +41 22 76 78696
           fax:  +41 22 76 79425        gsm: +41 76 48 70855

 



From dave at boost-consulting.com  Mon Jun  7 22:27:00 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 07 Jun 2004 16:27:00 -0400
Subject: [C++-sig] Re: problems compiling simple example in boost-python
 documentation/tutorial...
References: <6.1.0.6.0.20040607092433.026648c0@pop-tech.intec.UGent.be>
Message-ID: <uy8mzi06j.fsf@boost-consulting.com>

Peter Vandersteegen <Peter.Vandersteegen at intec.Ugent.be> writes:

> Hello, As a boost-python beginner I've started exploring the
> tutorial.  I however had trouble compiling a very basic example from
> the tutorial:
> http://www.boost.org/libs/python/doc/tutorial/doc/virtual_functions_with_default_implementations.html
> Changing some lines in the tutorial code however solved the
> compiling problem.  Now my question is: Does this relate to my
> compiler: intel 7.1, msvc7.0? Or something I overlooked in my code?
> I've started with a pure virtual base class, and indeed the virtual
> function can be overriden in python and called with a function in
> c++ ... (fantastic :) ) No problems at all...  I recieve however
> compiler-errors (explained after program) when I however try to
> compile the exact example from the tutorial ...(see link and
> following text)

The problem is that Boost.Python is trying to export Base's copy ctor
for wrapping return-by-value functions.  THat's why your addition of
noncopyable works.  Actually it seems like you've got everything
figured out below.

BTW, it would be good nettiquette to leave a blank line between the
paragraphs of your postings.

> // file wraptestpoly.cpp
> #include
> <boost/python.hpp>
> using namespace boost::python;
>
> struct Base
> {
> virtual int f(void){return 0;}
> }; //*
>
> int call_f(Base&amp; b) { return b.f(); }
>     
>     
> struct BaseWrap : Base
>     {
>         BaseWrap(PyObject* self_)
>             : self(self_) {}
>         virtual int f() { return call_method<int>(self, "f"); }
>         int default_f() { return Base::f(); }
>         PyObject* self;
>     };
>
> BOOST_PYTHON_MODULE(wraptestpoly){
>
> class_<Base, BaseWrap>("Base")  //*1
>         .def("f", &amp;Base::f, &amp;BaseWrap::default_f)
> def("call_f", call_f);
> }
> //end file
> The error I recieve is:
> D:\libraries\boost_1_31_0\boost/python/object/value_holder.hpp(134):
> error: no instance of constructor "BaseWrap::BaseWrap" matches the
> argument list argument types are: (PyObject *,
> boost::reference_wrapper<const Base>::type)
>
> If I however  use 
> <x-tab>        </x-tab> class_<Base, BaseWrap,boost::noncopyable>("Base")
> instead of //*1 --> no compiling problems.
> Another option which works is adding a copy-constructor:
> <x-tab>        </x-tab>       BaseWrap(PyObject* self_, Base const&amp; copy)  : Base(copy), self(self_) {}
> in the wrapperclass.
> Now my question is: what did I do wrong with this simple example?
> Could this problem arise from an outdated tutorial?

The tutorial example isn't outdated - it never worked.  We've seen
about 1 problem per month with the tutorial examples since the
tutorial came out.  Joel, why don't you make test cases for each of
them and put them in the regression test suite?


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




From dave at boost-consulting.com  Mon Jun  7 22:32:13 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 07 Jun 2004 16:32:13 -0400
Subject: [C++-sig] Re: Tedious Newbie Question
References: <176AB0A0ADC4734B8FE888D53E75D9D5017EB1F1@s15shefsrv.zoodigitalgroup.com>
Message-ID: <usmd7hzxu.fsf@boost-consulting.com>

Gareth Simpson <g.simpson at zoo-tech.com> writes:

> Apologies in advance if this is not the list to be asking these sort of
> things.

This is exactly the right list.

> I am trying to use Boost.Python to embed a Python library into my C++
> application.  The library in question uses unicode strings internally and I
> can't for the life of me work out how to get them out and into my C++.
> Google is largely silent on the matter.
>
> The below code snippet shows what I'm trying to do.
>
> I have no problem using extract<> to get a python string into a std::string
> but trying the same thing with unicode python strings and std::wstrings
> throws an exception.
>
> Clearly I'm going about it the wrong way - what should I be doing?

Probably you need to be using a newer version of Boost; we've
supported wstring conversions since Boost 1.31.0.

But we don't support Py_Finalize(); see
http://www.boost.org/libs/python/todo.html#pyfinalize-safety


> Thanks in advance
>
> Gareth
>
>
>
>
>
> #include "stdafx.h"
>
> #include "boost/python.hpp"
>
> int _tmain(int argc, _TCHAR* argv[])
> {
> 	using namespace boost::python;
> 	Py_Initialize();
>
> 	object mainModule (borrowed(PyImport_AddModule("__main__")));
> 	object mainNamespace = mainModule.attr("__dict__");
>
> 	handle<>(PyRun_String( 
> 		"str1 = 'a string'\n"	// a simple string
> 		"str2 = u'a unicode string'\n", // a unicode string
> 		Py_file_input , mainNamespace.ptr(), mainNamespace.ptr() ));
>
> 	std::string str1  = extract<std::string>(mainNamespace["str1"]); //
> this works fine
> 	std::wstring str2  = extract<std::wstring>(mainNamespace["str2"]);
> // boom!
>
> 	Py_Finalize();
>
> 	return 0;
> }

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




From wstwej03 at sneakemail.com  Tue Jun  8 01:28:33 2004
From: wstwej03 at sneakemail.com (Eric)
Date: Mon, 7 Jun 2004 16:28:33 -0700
Subject: [C++-sig] wrapping a function that returns a FILE *
In-Reply-To: <911F8C8EB7A8084AAEDD55CEDC54D8F851FE2E@iggy.rockstarvancouver.com>
Message-ID: <13267-06242@sneakemail.com>

Following the suggestion found here:
http://mail.python.org/pipermail/c++-sig/2002-October/002386.html

I tried to wrap a function that returns a FILE * so that python would treat
it as a builtin file object. However, it does not work. I get the following
error:

TypeError: No to_python (by-value) converter found for C++ type: struct
_iobuf *

I have also tried the variation found here:
http://mail.python.org/pipermail/c++-sig/2002-October/002373.html

but get then I get a segfault in python when I attempt to run the code that
accesses the FILE * functions.

Any further suggestions for things I can try?





From joel at boost-consulting.com  Tue Jun  8 03:47:21 2004
From: joel at boost-consulting.com (Joel de Guzman)
Date: Tue, 08 Jun 2004 09:47:21 +0800
Subject: [C++-sig] Re: problems compiling simple example in boost-python
	documentation/tutorial...
In-Reply-To: <uy8mzi06j.fsf@boost-consulting.com>
References: <6.1.0.6.0.20040607092433.026648c0@pop-tech.intec.UGent.be>
	<uy8mzi06j.fsf@boost-consulting.com>
Message-ID: <40C51AA9.8000200@boost-consulting.com>

David Abrahams wrote:

> The tutorial example isn't outdated - it never worked.  We've seen
> about 1 problem per month with the tutorial examples since the
> tutorial came out.  Joel, why don't you make test cases for each of
> them and put them in the regression test suite?

I will. In the meantime, I commited the fix to CVS.

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



From nick at rockstarvancouver.com  Tue Jun  8 03:52:34 2004
From: nick at rockstarvancouver.com (Nick Trout)
Date: Mon, 7 Jun 2004 18:52:34 -0700
Subject: [C++-sig] wrapping a function that returns a FILE *
Message-ID: <911F8C8EB7A8084AAEDD55CEDC54D8F881FDA8@iggy.rockstarvancouver.com>



> On Behalf Of Eric
> but get then I get a segfault in python when I attempt to run the code
> that
> accesses the FILE * functions.
> 
> Any further suggestions for things I can try?

Which platform is this for?

>From Python docs: Python/C API Reference Manual, 2. The Very High Level
Layer:

	"Note also that several of these functions take FILE*
parameters. On particular issue which needs to be handled carefully is
that the FILE structure for different C libraries can be different and
incompatible. Under Windows (at least), it is possible for dynamically
linked extensions to actually use different libraries, so care should be
taken that FILE* parameters are only passed to these functions if it is
certain that they were created by the same library that the Python
runtime is using."

There is a rift between .NET and the Python libraries.

Nick




From rwgk at yahoo.com  Tue Jun  8 12:25:41 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Tue, 8 Jun 2004 03:25:41 -0700 (PDT)
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <5BF9E41B6C9F2546B624F30FE840C122351BAF@cernxchg21.cern.ch>
Message-ID: <20040608102541.67426.qmail@web20208.mail.yahoo.com>

--- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> I am having the following compilation warnings when including
> "boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 and gcc 3.2. Is
> anything I am doing wrong?

Due to a Python requirement the boost/python.hpp header has to be included
before any system header. Please move the python.hpp include to the very top of
your file. Let us know if it doesn't help.
Ralf



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From g.simpson at zoo-tech.com  Tue Jun  8 13:11:27 2004
From: g.simpson at zoo-tech.com (Gareth Simpson)
Date: Tue, 8 Jun 2004 12:11:27 +0100 
Subject: [C++-sig] Re: Tedious Newbie Question
Message-ID: <176AB0A0ADC4734B8FE888D53E75D9D5017EB49A@s15shefsrv.zoodigitalgroup.com>


> 
> > I am trying to use Boost.Python to embed a Python library 
> into my C++
> > application.  The library in question uses unicode strings 
> internally and I
> > can't for the life of me work out how to get them out and 
> into my C++.
> > Google is largely silent on the matter.
> >
> > I have no problem using extract<> to get a python string 
> into a std::string
> > but trying the same thing with unicode python strings and 
> std::wstrings
> > throws an exception.
> >
> > Clearly I'm going about it the wrong way - what should I be doing?
> 
> Probably you need to be using a newer version of Boost; we've
> supported wstring conversions since Boost 1.31.0.


That's the version of boost I have.

It looks like it's trying to delete something it shouldn't (access
violation) somewhere in the bowels of the STL / boost::python

Interestingly if I go up the call stack to where the extract<std::wstring>()
is being called, the unicode string has actually arrived.


> 
> But we don't support Py_Finalize(); see
> http://www.boost.org/libs/python/todo.html#pyfinalize-safety
> 


Ah, what should I do instead?



From dave at boost-consulting.com  Tue Jun  8 14:05:18 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 08 Jun 2004 08:05:18 -0400
Subject: [C++-sig] Re: Tedious Newbie Question
References: <176AB0A0ADC4734B8FE888D53E75D9D5017EB49A@s15shefsrv.zoodigitalgroup.com>
Message-ID: <uekoq8dc1.fsf@boost-consulting.com>

Gareth Simpson <g.simpson at zoo-tech.com> writes:

>> 
>> > I am trying to use Boost.Python to embed a Python library 
>> into my C++
>> > application.  The library in question uses unicode strings 
>> internally and I
>> > can't for the life of me work out how to get them out and 
>> into my C++.
>> > Google is largely silent on the matter.
>> >
>> > I have no problem using extract<> to get a python string 
>> into a std::string
>> > but trying the same thing with unicode python strings and 
>> std::wstrings
>> > throws an exception.
>> >
>> > Clearly I'm going about it the wrong way - what should I be doing?
>> 
>> Probably you need to be using a newer version of Boost; we've
>> supported wstring conversions since Boost 1.31.0.
>
>
> That's the version of boost I have.
>
> It looks like it's trying to delete something it shouldn't (access
> violation) somewhere in the bowels of the STL / boost::python

Oh, in that case it's probably something in your own code that's
causing the crash.

> Interestingly if I go up the call stack to where the extract<std::wstring>()
> is being called, the unicode string has actually arrived.

;-) That should tell you something.

>> 
>> But we don't support Py_Finalize(); see
>> http://www.boost.org/libs/python/todo.html#pyfinalize-safety
>
> Ah, what should I do instead?

Nothing; just don't call it.

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




From g.simpson at zoo-tech.com  Tue Jun  8 14:39:10 2004
From: g.simpson at zoo-tech.com (Gareth Simpson)
Date: Tue, 8 Jun 2004 13:39:10 +0100 
Subject: [C++-sig] Re: Tedious Newbie Question
Message-ID: <176AB0A0ADC4734B8FE888D53E75D9D5017EB4DF@s15shefsrv.zoodigitalgroup.com>

> >
> > It looks like it's trying to delete something it shouldn't (access
> > violation) somewhere in the bowels of the STL / boost::python
> 
> Oh, in that case it's probably something in your own code that's
> causing the crash.
> 
> > Interestingly if I go up the call stack to where the 
> extract<std::wstring>()
> > is being called, the unicode string has actually arrived.
> 
> ;-) That should tell you something.
> 

Bah, it's this.  My bad.

http://mail.python.org/pipermail/c++-sig/2004-May/007324.html



From olivier.ravard at novagrid.com  Tue Jun  8 14:26:12 2004
From: olivier.ravard at novagrid.com (Olivier Ravard)
Date: Tue, 8 Jun 2004 14:26:12 +0200
Subject: [C++-sig] release compilation with bjam
Message-ID: <006101c44d53$c979bf00$3d521481@ravard>

Hi,

I compiled boost.python under windows with mingw.
I test some projects  and it works fine.

I build my own project with boost.build and it works but the compilation used
only the debug release (linking).
How should I do to compile with optimization and make a release in a 
non debug mode ?

Thanks.

O.R.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040608/ed181c4a/attachment.htm>

From Pere.Mato at cern.ch  Tue Jun  8 17:26:36 2004
From: Pere.Mato at cern.ch (Pere Mato Vila)
Date: Tue, 8 Jun 2004 17:26:36 +0200
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
Message-ID: <5BF9E41B6C9F2546B624F30FE840C1225FDB76@cernxchg21.cern.ch>

My test program to show the problem contains only one line:

#include "boost/python.hpp"

and therefore I can not move it before anything. 

Pere 

> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> Grosse-Kunstleve
> Sent: 08 June 2004 12:26
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> 
> 
> --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > I am having the following compilation warnings when including 
> > "boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 
> and gcc 3.2. 
> > Is anything I am doing wrong?
> 
> Due to a Python requirement the boost/python.hpp header has 
> to be included before any system header. Please move the 
> python.hpp include to the very top of your file. Let us know 
> if it doesn't help. Ralf
> 
> 
> 
> 	
> 		
> __________________________________
> Do you Yahoo!?
> Friends.  Fun.  Try the all-new Yahoo! Messenger. 
> http://messenger.yahoo.com/ 
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 



From nitin.shukla at pw.utc.com  Tue Jun  8 17:33:47 2004
From: nitin.shukla at pw.utc.com (Shukla, Nitin (Export))
Date: Tue, 8 Jun 2004 11:33:47 -0400 
Subject: [C++-sig] release compilation with bjam
Message-ID: <DCD1A6766747044281DB96489FD792CF045F2A2A@pusehe0k.eh.pweh.com>

You should first build the Boost library in the release mode. To do that you
should specify this command line argument "-sBUILD=release" while build
Boost. Again when building you own project specify this command line
argument to bjam and it would build the release version of your project.
Make sure that in the jamfile you specify the <library-path> to the release
version of Boost.Python dll. 
 
HTH
Nitin
 
 -----Original Message-----
From: Olivier Ravard [mailto:olivier.ravard at novagrid.com]
Sent: Tuesday, June 08, 2004 8:26 AM
To: Development of Python/C++ integration
Subject: [C++-sig] release compilation with bjam



Hi,
 
I compiled boost.python under windows with mingw.
I test some projects  and it works fine.
 
I build my own project with boost.build and it works but the compilation
used
only the debug release (linking).
How should I do to compile with optimization and make a release in a 
non debug mode ?
 
Thanks.
 
O.R.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040608/39c8a540/attachment.htm>

From dave at boost-consulting.com  Tue Jun  8 18:49:22 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 08 Jun 2004 12:49:22 -0400
Subject: [C++-sig] Re: release compilation with bjam
References: <DCD1A6766747044281DB96489FD792CF045F2A2A@pusehe0k.eh.pweh.com>
Message-ID: <uise2571p.fsf@boost-consulting.com>

"Shukla, Nitin (Export)" <nitin.shukla at pw.utc.com> writes:

> You should first build the Boost library in the release mode. To do that you
> should specify this command line argument "-sBUILD=release" while build
> Boost. Again when building you own project specify this command line
> argument to bjam and it would build the release version of your project.
> Make sure that in the jamfile you specify the <library-path> to the release
> version of Boost.Python dll. 

That's all correct, except that if you're using the Boost.Build
example structure in libs/python/example, the appropriate Boost
library will be built on demand.  You don't need to separately build
it first.

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




From olivier.ravard at novagrid.com  Tue Jun  8 19:04:33 2004
From: olivier.ravard at novagrid.com (Olivier Ravard)
Date: Tue, 8 Jun 2004 19:04:33 +0200
Subject: [C++-sig] Re: release compilation with bjam
References: <DCD1A6766747044281DB96489FD792CF045F2A2A@pusehe0k.eh.pweh.com>
	<uise2571p.fsf@boost-consulting.com>
Message-ID: <00ff01c44d7a$ab98c040$3d521481@ravard>


----- Original Message -----
From: "David Abrahams" <dave at boost-consulting.com>
To: <c++-sig at python.org>
Sent: Tuesday, June 08, 2004 6:49 PM
Subject: [C++-sig] Re: release compilation with bjam


> "Shukla, Nitin (Export)" <nitin.shukla at pw.utc.com> writes:
>
> > You should first build the Boost library in the release mode. To do that
you
> > should specify this command line argument "-sBUILD=release" while build
> > Boost. Again when building you own project specify this command line
> > argument to bjam and it would build the release version of your project.
> > Make sure that in the jamfile you specify the <library-path> to the
release
> > version of Boost.Python dll.
>
> That's all correct, except that if you're using the Boost.Build
> example structure in libs/python/example, the appropriate Boost
> library will be built on demand.  You don't need to separately build
> it first.
>
It works.
Thanks.

O.R.

> --
> Dave Abrahams
> Boost Consulting
> http://www.boost-consulting.com
>
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig




From nicodemus at esss.com.br  Wed Jun  9 01:15:17 2004
From: nicodemus at esss.com.br (Nicodemus)
Date: Tue, 08 Jun 2004 20:15:17 -0300
Subject: [C++-sig] Pyste exception handlers
In-Reply-To: <40BFF93D.7030506@paradise.net.nz>
References: <40BFF93D.7030506@paradise.net.nz>
Message-ID: <40C64885.1080009@esss.com.br>

Hi Paul,

paul.bridger wrote:

> Nicodemus,
>
> How would you recommend exception handler registration be implemented 
> in Pyste?
>
> A problem here would be what to do in the differing cases of:
> - building a single pyste file into a module
> : in this case specifying exception handlers in the pyste file is 
> sufficient
> - building many pyste files into a single module
> : in this case you probably want exception handlers to be specified in 
> a module-level file, or perhaps you could put them in an arbitrary 
> pyste file
>
> I think Pyste has some deficiencies in organising modules from 
> multiple files. It is wonderful that this can be done, but I don't 
> like that so much has to be specified on the command-line (and that 
> the meaning of command-line options changes depending on --multiple or 
> not).
>
> I think a better organisation would involve a module-level file, which 
> could then contain such things as exception handlers. This would also 
> make building modules out of several pyste files cleaner. (A 
> disadvantage is that it expands the scope of the tool into something 
> of a build tool).
> What do others think?


I'm not ignoring you, it's just that I haven't found time to answer this 
message... I hope I will be to answer it in the next few days.

Thanks!
Nicodemus.



From paul.bridger at paradise.net.nz  Wed Jun  9 02:07:38 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Wed, 09 Jun 2004 12:07:38 +1200
Subject: [C++-sig] Pyste exception handlers
In-Reply-To: <40C64885.1080009@esss.com.br>
References: <40BFF93D.7030506@paradise.net.nz> <40C64885.1080009@esss.com.br>
Message-ID: <40C654CA.9010600@paradise.net.nz>

Sweet. I'm not sulkily not using Pyste anymore, just fighting off a computer 
game addiction. :)

Paul.

Nicodemus wrote:

> Hi Paul,
> 
> paul.bridger wrote:
> 
>> Nicodemus,
>>
>> How would you recommend exception handler registration be implemented 
>> in Pyste?
>>
>> A problem here would be what to do in the differing cases of:
>> - building a single pyste file into a module
>> : in this case specifying exception handlers in the pyste file is 
>> sufficient
>> - building many pyste files into a single module
>> : in this case you probably want exception handlers to be specified in 
>> a module-level file, or perhaps you could put them in an arbitrary 
>> pyste file
>>
>> I think Pyste has some deficiencies in organising modules from 
>> multiple files. It is wonderful that this can be done, but I don't 
>> like that so much has to be specified on the command-line (and that 
>> the meaning of command-line options changes depending on --multiple or 
>> not).
>>
>> I think a better organisation would involve a module-level file, which 
>> could then contain such things as exception handlers. This would also 
>> make building modules out of several pyste files cleaner. (A 
>> disadvantage is that it expands the scope of the tool into something 
>> of a build tool).
>> What do others think?
> 
> 
> 
> I'm not ignoring you, it's just that I haven't found time to answer this 
> message... I hope I will be to answer it in the next few days.
> 
> Thanks!
> Nicodemus.
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 
> 



From jeff.holle at verizon.net  Tue Jun  8 15:35:56 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Tue, 08 Jun 2004 09:35:56 -0400
Subject: [C++-sig] Forward References
Message-ID: <40C5C0BC.6020203@verizon.net>

I want to express this issue to pyste users.
With a class that uses forward references to express things like 
bi-directional assocition, it is necessary to add line
in the associated pyste configure file like:
   declaration_code('#include "MyClass.h"\n')

If this isn't done, the compiler errors, with at least gcc, aren't very 
helpful.



From shin1_morita at yahoo.co.jp  Wed Jun  9 16:09:22 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Wed, 9 Jun 2004 23:09:22 +0900 (JST)
Subject: [C++-sig] call_f(Derived) did not match C++ signature
Message-ID: <20040609140922.46979.qmail@web2302.mail.yahoo.co.jp>

Hi, this is my first post.

I've just started learning Boost.Python.
I want to expose C++ abstract class and function that
receives its instance like this example in the tutorial:

http://boost.org/libs/python/doc/tutorial/doc/class_virtual_functions.html

But I have a problem ...
I compiled this example as module foo and ran test_foo.py
attached to this mail.
Then I had following errors:

====== BEGIN OUTPUT ======
*****************************************************************
Failure in example: foo.call_f(a)
from line #9 of test_foo
Exception raised:
Traceback (most recent call last):
  File "/home/shin/.local/lib/python2.3/doctest.py", line
442, in _run_examples_inner
    compileflags, 1) in globs
  File "<string>", line 1, in ?
ArgumentError: Python argument types in
    foo.call_f(Derived)
did not match C++ signature:
    call_f(Base {lvalue})
*****************************************************************
1 items had failures:
   1 of   4 in test_foo
***Test Failed*** 1 failures.

EXIT STATUS: 1
====== END OUTPUT ======

What's wrong?
Thanks.



__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: foo.cc
Type: application/octet-stream
Size: 411 bytes
Desc: foo.cc
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040609/cb9b3068/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_foo.py
Type: application/octet-stream
Size: 370 bytes
Desc: test_foo.py
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040609/cb9b3068/attachment-0001.obj>

From aashish at iastate.edu  Wed Jun  9 16:52:11 2004
From: aashish at iastate.edu (aashish at iastate.edu)
Date: Wed, 9 Jun 2004 09:52:11 -0500 (CDT)
Subject: [C++-sig] call_f(Derived) did not match C++ signature
In-Reply-To: <20040609140922.46979.qmail@web2302.mail.yahoo.co.jp>
References: <20040609140922.46979.qmail@web2302.mail.yahoo.co.jp>
Message-ID: <3061.129.186.232.105.1086792731.squirrel@mail.eng.iastate.edu>

I think what you are doing wrong is instead of passing the object of
Derived you are passing the class itself
do this:

from test_foo import *

class Derived(Base):
    def f(self):
        return 42

derived = Derived()
print derived.f()
print call_f(derived)
------------------------------------

and in C++

#include <boost/python.hpp>

using namespace boost::python;

struct Base{
		virtual int f() = 0;
};

int call_f(Base& b) {return b.f();}

struct BaseWrap : Base{
	BaseWrap(PyObject* self_) : self(self_){}
	int f() { return call_method<int>(self,"f") ;}
	PyObject* self;
};

BOOST_PYTHON_MODULE(test_foo) {
 class_<Base, BaseWrap, boost::noncopyable>("Base") ;
 def("call_f", call_f);
}
---------------------------------

if you look carefull the call_f takes the ref to Base ...

Hope this helps,
~aashish



>/doc/class_virtual_functions.html
>
> But I have a problem ...
> I compiled this example as module foo and ran test_foo.py
> attached to this mail.
> Then I had following errors:
>
> ====== BEGIN OUTPUT ======
> *****************************************************************
> Failure in example: foo.call_f(a)
> from line #9 of test_foo
> Exception raised:
> Traceback (most recent call last):
>   File "/home/shin/.local/lib/python2.3/doctest.py", line
> 442, in _run_examples_inner
>     compileflags, 1) in globs
>   File "<string>", line 1, in ?
> ArgumentError: Python argument types in
>     foo.call_f(Derived)
> did not match C++ signature:
>     call_f(Base {lvalue})
> *****************************************************************
> 1 items had failures:
>    1 of   4 in test_foo
> ***Test Failed*** 1 failures.
>
> EXIT STATUS: 1
> ====== END OUTPUT ======
>
> What's wrong?
> Thanks.
>
>
>
> __________________________________________________
> Do You Yahoo!?
> http://bb.yahoo.co.jp/
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


-----------------------------------------------
Aashish Chaudhary       | Email:
Research Assistant      | ashish_cy at yahoo.com
Iowa State University   | aashish at iastate.edu
Ames, Iowa - USA        | Ph: (515) 441-1178
-----------------------------------------------



From vmilitaru at sympatico.ca  Wed Jun  9 13:54:31 2004
From: vmilitaru at sympatico.ca (Vio)
Date: Wed, 09 Jun 2004 07:54:31 -0400
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <40C6FA77.9080708@sympatico.ca>

I have a basic dynamic lib "noddy.so" which I want to
1- embed inside my executable
2- import by embedded python interpreter using an API call.

I have two questions:

1- what would be the recommended technique to embed a binary file inside 
an executable?
My guess would be to add the binary "noddy.so" to the linking command, 
but then
how do I get access to that chunk of code in my source file? My goal 
would be to achieve something like

    int* noddy_so = <binary here> ... but then how do I include the 
binary code, or make the source code aware of it?

Using "#include" doesn't feel right, as this is a binary file, not 
ASCII. And loading the file at run-time isn't right either, I need to 
load it at compile time. And is "int*" the appropriate data type for 
binary data?

2- Let's imagine that I managed to load "noddy.so" into my source code 
somehow, and now have a pointer to it.
Any hints as to the  correct python API call to import a dynamic lib 
from a memory buffer? Could someone
tell me for example if the following makes sense for my purpose?

------------------------------
char *modulename="noddy", *filename="no_real_file";
PyObject *tmp;

/*Here try to translate the binary lib into a PyObject*/
tmp = Py_CompileString(noddy_so, filename, Py_file_input);
if (tmp == NULL)
      error("Can't compile module [tmp]");

/*Import the "noddy" module*/
pmod = PyImport_ExecCodeModule(modulename, tmp);
if (pmod == NULL)
      error("Can't load module [pmod]");
------------------------------

Regards,
Vio





From shin1_morita at yahoo.co.jp  Wed Jun  9 18:11:23 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Thu, 10 Jun 2004 01:11:23 +0900 (JST)
Subject: [C++-sig] call_f(Derived) did not match C++ signature
In-Reply-To: <3061.129.186.232.105.1086792731.squirrel@mail.eng.iastate.edu>
Message-ID: <20040609161123.71152.qmail@web2307.mail.yahoo.co.jp>

> if you look carefull the call_f takes the ref to
> Base ...

Thanks, but I believe that I did almost exactly the same
thing.
Please check out my codes attached before ...

I found the same report in The C++-sig Archives:

http://mail.python.org/pipermail/c++-sig/2004-March/006988.html

Is this already resolved?
Anyone knows?

By the way, this is my environment:

gcc-3.2.3 and gcc-3.4.0
python-2.3.4
boost-1.31.0


__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/




From aherrman at mi.fu-berlin.de  Wed Jun  9 18:11:26 2004
From: aherrman at mi.fu-berlin.de (aherrman at mi.fu-berlin.de)
Date: Wed,  9 Jun 2004 18:11:26 +0200
Subject: [C++-sig] Survey of IDE users and source code editors 
Message-ID: <1086797486.40c736aee0c60@webmail.mi.fu-berlin.de>


As part of a University study of Integrated Development Environment 
(IDE) users, we are conducting a survey to appraise the relative 
utility of different features of IDE source code editors, such as those 
included with Visual Studio and Eclipse as well as stand-alone editors 
like Emacs that support programming tasks. Our survey is designed to 
assess how often users employ editor features like syntax highlighting 
and code completion, and how satisfied users are with these features in 
their particular environments. 

If you or your project group have active experience using an IDE, we 
would like you to participate in our study. You can find the survey at 

http://physik.fu-berlin.de/~zabel/ide_editor_survey/ 

The survey should take no more than 20 minutes to complete. No personal 
information is required of participants. 

The survey period is two weeks (until June 22nd, 2004). Participants 
who opt to enter their e-mail addresses on the survey page will receive 
a copy of the findings of our study after the survey period. 

We look forward to hearing from you. 

Philipp Zabel, Minor Gordon and Antoinette Herrmann 
Freie Universitaet Berlin 
Institut fuer Informatik 
Takustr. 9 
14195 Berlin





From gwl at u.washington.edu  Wed Jun  9 18:32:47 2004
From: gwl at u.washington.edu (Graeme Lufkin)
Date: Wed, 09 Jun 2004 09:32:47 -0700
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <1086798766.17992.23.camel@montague.astro.washington.edu>

	I've done this myself; let me try and explain what I did.  First, write
the extension module as normal.  I used Boost.Python, but what's
important is having access to the 'extern "C" void
init<modulenamehere>()' function pointer.  
Second, instead of compiling your module into its own shared library
(.so file), simply link it in with your executable.  Using the
Boost.Build system, you can just add the .cpp files that make up your
module to the dependency list of the executable.
Third, before you create the embedded interpreter, make a call to
PyImport_AppendInittab().  For example, if my module is called funcs,
the init function is declared as 'extern "C" void initfuncs();', so I
call:
PyImport_AppendInittab("funcs", initfuncs);
then I call Py_Initialize().
Fourth, finally I can actually import the module into my embedded
interpreter:
PyRun_SimpleString("import funcs");

	So, if you follow my steps, you have only one binary file, the
executable.  This file contains code for the module, as well as the
regular C++ stuff of your program.  You embed the Python interpreter,
and import your module into its Python namespace.
	The embedding part of the Boost.Python tutorial is a bit lacking, and
some of the code snippets won't compile, but give it a try anyway, as
there's no other option ;)
	Hope this helps

> I have a basic dynamic lib "noddy.so" which I want to
> 1- embed inside my executable
> 2- import by embedded python interpreter using an API call.
> 
> I have two questions:
> 
> 1- what would be the recommended technique to embed a binary file
> inside 
> an executable?
> My guess would be to add the binary "noddy.so" to the linking command,
> but then
> how do I get access to that chunk of code in my source file? My goal 
> would be to achieve something like
> 
>     int* noddy_so = <binary here> ... but then how do I include the 
> binary code, or make the source code aware of it?
> 
> Using "#include" doesn't feel right, as this is a binary file, not 
> ASCII. And loading the file at run-time isn't right either, I need to 
> load it at compile time. And is "int*" the appropriate data type for 
> binary data?
> 
> 2- Let's imagine that I managed to load "noddy.so" into my source code
> somehow, and now have a pointer to it.
> Any hints as to the  correct python API call to import a dynamic lib 
> from a memory buffer? Could someone
> tell me for example if the following makes sense for my purpose?
> 
> ------------------------------
> char *modulename="noddy", *filename="no_real_file";
> PyObject *tmp;
> 
> /*Here try to translate the binary lib into a PyObject*/
> tmp = Py_CompileString(noddy_so, filename, Py_file_input);
> if (tmp == NULL)
>       error("Can't compile module [tmp]");
> 
> /*Import the "noddy" module*/
> pmod = PyImport_ExecCodeModule(modulename, tmp);
> if (pmod == NULL)
>       error("Can't load module [pmod]");
> ------------------------------
> 
> Regards,
> Vio
-- 
	- Graeme
	gwl at u.washington.edu
"This sentence contains exactly threee erors."




From aashish at iastate.edu  Wed Jun  9 18:40:25 2004
From: aashish at iastate.edu (aashish at iastate.edu)
Date: Wed, 9 Jun 2004 11:40:25 -0500 (CDT)
Subject: [C++-sig] call_f(Derived) did not match C++ signature
In-Reply-To: <20040609161123.71152.qmail@web2307.mail.yahoo.co.jp>
References: <3061.129.186.232.105.1086792731.squirrel@mail.eng.iastate.edu>
	<20040609161123.71152.qmail@web2307.mail.yahoo.co.jp>
Message-ID: <3512.129.186.232.105.1086799225.squirrel@mail.eng.iastate.edu>

Hi,

Yes I did check your code. Sorry to say that I didnt check it before.

Well its working on my environment:
  Windows XP
  Python 2.3
  boost-1.31.0

I am not sure what the problem is.

~aashish



> By the way, this is my environment:
>
> gcc-3.2.3 and gcc-3.4.0
> python-2.3.4
> boost-1.31.0
>
>
> __________________________________________________
> Do You Yahoo!?
> http://bb.yahoo.co.jp/
>
>
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>


-----------------------------------------------
Aashish Chaudhary       | Email:
Research Assistant      | ashish_cy at yahoo.com
Iowa State University   | aashish at iastate.edu
Ames, Iowa - USA        | Ph: (515) 441-1178
-----------------------------------------------



From dave at boost-consulting.com  Wed Jun  9 20:06:40 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 09 Jun 2004 14:06:40 -0400
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
References: <20040609140922.46979.qmail@web2302.mail.yahoo.co.jp>
Message-ID: <uk6ygfvwv.fsf@boost-consulting.com>

Shin-ichi MORITA <shin1_morita at yahoo.co.jp> writes:

> Hi, this is my first post.
>
> I've just started learning Boost.Python.
> I want to expose C++ abstract class and function that
> receives its instance like this example in the tutorial:
>
> http://boost.org/libs/python/doc/tutorial/doc/class_virtual_functions.html
>
> But I have a problem ...
> I compiled this example as module foo and ran test_foo.py
> attached to this mail.
> Then I had following errors:
>
> ====== BEGIN OUTPUT ======
> *****************************************************************
> Failure in example: foo.call_f(a)
> from line #9 of test_foo
> Exception raised:
> Traceback (most recent call last):
>   File "/home/shin/.local/lib/python2.3/doctest.py", line
> 442, in _run_examples_inner
>     compileflags, 1) in globs
>   File "<string>", line 1, in ?
> ArgumentError: Python argument types in
>     foo.call_f(Derived)
                 ^^^^^^^
> did not match C++ signature:
>     call_f(Base {lvalue})
             ^^^^
> *****************************************************************
> 1 items had failures:
>    1 of   4 in test_foo
> ***Test Failed*** 1 failures.
>
> EXIT STATUS: 1
> ====== END OUTPUT ======
>
> What's wrong?

Normally for me to help you, you'd need to post the code you used that
caused the problem, but in this case I think it's pretty clear that
when you wrapped Derived you didn't put bases<Base> in the 
class_< ... > argument list.

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




From rwgk at yahoo.com  Wed Jun  9 21:10:02 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Wed, 9 Jun 2004 12:10:02 -0700 (PDT)
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <5BF9E41B6C9F2546B624F30FE840C1225FDB76@cernxchg21.cern.ch>
Message-ID: <20040609191002.30354.qmail@web20211.mail.yahoo.com>

--- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> My test program to show the problem contains only one line:
> 
> #include "boost/python.hpp"
> 
> and therefore I can not move it before anything. 

Hm. What is your platform? Linux (which), Cygwin?


> > -----Original Message-----
> > From: c++-sig-bounces at python.org 
> > [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> > Grosse-Kunstleve
> > Sent: 08 June 2004 12:26
> > To: Development of Python/C++ integration
> > Subject: Re: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> > 
> > 
> > --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > > I am having the following compilation warnings when including 
> > > "boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 
> > and gcc 3.2. 
> > > Is anything I am doing wrong?
> > 
> > Due to a Python requirement the boost/python.hpp header has 
> > to be included before any system header. Please move the 
> > python.hpp include to the very top of your file. Let us know 
> > if it doesn't help. Ralf
> > 
> > 
> > 
> > 	
> > 		
> > __________________________________
> > Do you Yahoo!?
> > Friends.  Fun.  Try the all-new Yahoo! Messenger. 
> > http://messenger.yahoo.com/ 
> > 
> > _______________________________________________
> > 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



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From phil at freehackers.org  Wed Jun  9 21:57:02 2004
From: phil at freehackers.org (Philippe Fremy)
Date: Wed, 9 Jun 2004 21:57:02 +0200
Subject: [C++-sig] multiple to_python converters ?
Message-ID: <200406092157.02366.phil@freehackers.org>


Hi,

I have the case where I have two python modules, both declaring the same 
to_python converter. The modules may be used separately or together. The 
problem is that if used together, I get (you see me coming I guess) a 
RunTimeError. Only one to_python converter may be registerted at a time for 
a given type.

I looked in the archives and on google. I saw two references to discussions 
of this problem and even a possible solution to be included in boost. But 
checking the documentation today, I could not find any hint on how to solve 
this problem.

So, is the problem still here ?

	regards,

	Philippe
-- 
An open source OS for contactless Smart Cards
http://www.jayacard.org - Jayacard 



From vmilitaru at sympatico.ca  Wed Jun  9 22:16:16 2004
From: vmilitaru at sympatico.ca (Vio)
Date: Wed, 09 Jun 2004 16:16:16 -0400
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <40C77010.8000505@sympatico.ca>

Many thanks for your solution Graeme. Unfortunately, I do need to use 
".so" files
as I want to "extend dynamically" an embedded python interpreter, skipping
recompile each time. Yes, I've tried both Boost and SWIG to generate 
wrappers,
but right now I like to do things "manually" to understand what's going on
"under the hood".

More details: Imagine a situation where the ".so" files are delivered 
through
a network; the easy solution would be to save the downloaded binary stream
as a ".so" file to disk,
then simply run the python command "import <modulename>" and I'm done.
But I simply wish to avoid the "save to disk" step which seems 
superfluous in this
case, and feed the binary lib directly to the embedded python interpreter.

Exploring the API docs, it would seem that something like

    PyRun_AnyFile(FILE *fp, char *filename)

could do what I want. But I don't know if/how I can give it a memory 
buffer instead
of a FILE pointer, and also I'm too much of a c/c++ newbie right now to 
figure out
how to handle huge binary data (as I write this, I'm trying to make 
sense of
Stroustrup's C++ bible "the c++ prog. lang, ed. 2000"), so a simple task 
like putting
contents of a ".so" file into a memory buffer is a big challenge for me 
right now.

Anyway, thanks for your help,
Vio

>	I've done this myself; let me try and explain what I did.  First, write
>the extension module as normal.  I used Boost.Python, but what's
>important is having access to the 'extern "C" void
>init<modulenamehere>()' function pointer.  
>Second, instead of compiling your module into its own shared library
>(.so file), simply link it in with your executable.  Using the
>Boost.Build system, you can just add the .cpp files that make up your
>module to the dependency list of the executable.
>Third, before you create the embedded interpreter, make a call to
>PyImport_AppendInittab().  For example, if my module is called funcs,
>the init function is declared as 'extern "C" void initfuncs();', so I
>call:
>PyImport_AppendInittab("funcs", initfuncs);
>then I call Py_Initialize().
>Fourth, finally I can actually import the module into my embedded
>interpreter:
>PyRun_SimpleString("import funcs");
>
>	So, if you follow my steps, you have only one binary file, the
>executable.  This file contains code for the module, as well as the
>regular C++ stuff of your program.  You embed the Python interpreter,
>and import your module into its Python namespace.
>	The embedding part of the Boost.Python tutorial is a bit lacking, and
>some of the code snippets won't compile, but give it a try anyway, as
>there's no other option ;)
>	Hope this helps
>
>>/ I have a basic dynamic lib "noddy.so" which I want to
>/>/ 1- embed inside my executable
>/>/ 2- import by embedded python interpreter using an API call.
>/>/ 
>/>/ I have two questions:
>/>/ 
>/>/ 1- what would be the recommended technique to embed a binary file
>/>/ inside 
>/>/ an executable?
>/>/ My guess would be to add the binary "noddy.so" to the linking command,
>/>/ but then
>/>/ how do I get access to that chunk of code in my source file? My goal 
>/>/ would be to achieve something like
>/>/ 
>/>/     int* noddy_so = <binary here> ... but then how do I include the 
>/>/ binary code, or make the source code aware of it?
>/>/ 
>/>/ Using "#include" doesn't feel right, as this is a binary file, not 
>/>/ ASCII. And loading the file at run-time isn't right either, I need to 
>/>/ load it at compile time. And is "int*" the appropriate data type for 
>/>/ binary data?
>/>/ 
>/>/ 2- Let's imagine that I managed to load "noddy.so" into my source code
>/>/ somehow, and now have a pointer to it.
>/>/ Any hints as to the  correct python API call to import a dynamic lib 
>/>/ from a memory buffer? Could someone
>/>/ tell me for example if the following makes sense for my purpose?
>/>/ 
>/>/ ------------------------------
>/>/ char *modulename="noddy", *filename="no_real_file";
>/>/ PyObject *tmp;
>/>/ 
>/>/ /*Here try to translate the binary lib into a PyObject*/
>/>/ tmp = Py_CompileString(noddy_so, filename, Py_file_input);
>/>/ if (tmp == NULL)
>/>/       error("Can't compile module [tmp]");
>/>/ 
>/>/ /*Import the "noddy" module*/
>/>/ pmod = PyImport_ExecCodeModule(modulename, tmp);
>/>/ if (pmod == NULL)
>/>/       error("Can't load module [pmod]");
>/>/ ------------------------------
>/
>




From grafik.list at redshift-software.com  Thu Jun 10 01:37:25 2004
From: grafik.list at redshift-software.com (Rene Rivera)
Date: Wed, 09 Jun 2004 18:37:25 -0500
Subject: [C++-sig] importing embedded dynamic lib
In-Reply-To: <40C77010.8000505@sympatico.ca>
References: <40C77010.8000505@sympatico.ca>
Message-ID: <40C79F35.5080708@redshift-software.com>

Vio wrote:

> More details: Imagine a situation where the ".so" files are delivered 
> through
> a network; the easy solution would be to save the downloaded binary stream
> as a ".so" file to disk,
> then simply run the python command "import <modulename>" and I'm done.
> But I simply wish to avoid the "save to disk" step which seems 
> superfluous in this
> case, and feed the binary lib directly to the embedded python interpreter.

If I'm understanding you correctly you want to load a dynamic library 
straight from a stream, instead of from a file? If that's the case 
unless you decide you want to undertake rewriting the dynamic loader you 
won't be able to do it. The dlopen call, which is what loads the .so in 
Linux and others, doesn't support that. About the best you could try is 
to create a temporary pipe and feed the stream into that, and give the 
pipe to dlopen. But I would bet it's not going to work as the dynamic 
loader is likely to require random access to the data.

-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq



From s_sourceforge at nedprod.com  Thu Jun 10 04:17:41 2004
From: s_sourceforge at nedprod.com (Niall Douglas)
Date: Thu, 10 Jun 2004 03:17:41 +0100
Subject: [C++-sig] importing embedded dynamic lib
In-Reply-To: <40C77010.8000505@sympatico.ca>
Message-ID: <40C7D2D5.11462.75838EA@localhost>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 9 Jun 2004 at 16:16, Vio wrote:

> More details: Imagine a situation where the ".so" files are delivered
> through a network; the easy solution would be to save the downloaded
> binary stream as a ".so" file to disk, then simply run the python
> command "import <modulename>" and I'm done. But I simply wish to avoid
> the "save to disk" step which seems superfluous in this case, and feed
> the binary lib directly to the embedded python interpreter.

Why don't you just mount the remote file system (SMB or NFS) and run 
via that? I personally develop on Linux and FreeBSD using a source 
tree on a Windows share on a Win2k machine. It's a little slow, but 
it's manageable.

Cheers,
Niall





-----BEGIN PGP SIGNATURE-----
Version: idw's PGP-Frontend 4.9.6.1 / 9-2003 + PGP 8.0.2

iQA/AwUBQMfExcEcvDLFGKbPEQJZRQCg8OTo23lq6YtAY41rddFiPIecgPoAoJ22
zqAulJvda5rIZkglBEWseO92
=oaxw
-----END PGP SIGNATURE-----



From vmilitaru at sympatico.ca  Thu Jun 10 02:17:45 2004
From: vmilitaru at sympatico.ca (Vio)
Date: Wed, 09 Jun 2004 20:17:45 -0400
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <40C7A8A9.5050400@sympatico.ca>

Perhaps I didn't explain myself correctly. My question would come down to:
*what API call*
does the python interpreter use when I type at the python prompt:

 >>> import mysharedlib

where "mysharedlib" is file "mysharedlib.so" somewhere on the PYTHONPATH.

What I wish is to use *that* call, but change arguments as to substitute
to the [my guess] file pointer for a memory buffer pointer.
The memory buffer having the exact contents of file "mysharedlib.so", 
thus avoiding
a disk access. While the result here is not very sexy, this exercise 
becomes interesting
when the app receives "mysharedlib.so" over the network.
Once received over network, easy solution is to write "mysharedlib.so" 
to local disk,
then either do something like:

PyRun_SimpleString("import mysharedlib")
or
PyImport_ImportModule("mysharedlib")

I simply want to skip that. In still other words, I guess I'm looking
for something like

PyImport_ExecCodeModule(char *name, PyObject *co)

only I'd have to construct a PyObject out of my buffer, since that's what
PyImport_ExecCodeModule() wants as 2nd argument. But API docs say this 2nd
argument should be python bytecode data, so I don't know if it won't 
choke on
compiled ELF binaries instead, hence I'm not too sure about this 
particular hack.

Anyway, my gut guess is that I must find what calls python uses to import
shared libs, and somehow use that and substitute that file pointer to my 
buffer pointer.
Just a wild guess.

If someone has some ideas on this, I'd appreciate it.
Vio





From grafik.list at redshift-software.com  Thu Jun 10 05:51:57 2004
From: grafik.list at redshift-software.com (Rene Rivera)
Date: Wed, 09 Jun 2004 22:51:57 -0500
Subject: [C++-sig] importing embedded dynamic lib
In-Reply-To: <40C7A8A9.5050400@sympatico.ca>
References: <40C7A8A9.5050400@sympatico.ca>
Message-ID: <40C7DADD.2080907@redshift-software.com>

Vio wrote:
> Perhaps I didn't explain myself correctly. My question would come down to:
> *what API call*
> does the python interpreter use when I type at the python prompt:
> 
>  >>> import mysharedlib
> 
> where "mysharedlib" is file "mysharedlib.so" somewhere on the PYTHONPATH.

Like I said before it uses dlopen and related OS functions, see:

http://www.die.net/doc/linux/include/dlfcn.h
http://www.die.net/doc/linux/man/man3/dlopen.3.html
http://www.die.net/doc/linux/man/man3/dlsym.3.html
http://www.die.net/doc/linux/man/man3/dlclose.3.html

Or go to your Linux shell and type "man dlopen".

> What I wish is to use *that* call, but change arguments as to substitute
> to the [my guess] file pointer for a memory buffer pointer.

You can not. Unless you write a different *dynamic linker loader* 
(ld.so), see:

http://www.die.net/doc/linux/man/man8/ld.so.8.html

[snip]
> PyImport_ExecCodeModule(char *name, PyObject *co)
> 
> only I'd have to construct a PyObject out of my buffer, since that's what
> PyImport_ExecCodeModule() wants as 2nd argument. But API docs say this 2nd
> argument should be python bytecode data, so I don't know if it won't 
> choke on
> compiled ELF binaries instead, hence I'm not too sure about this 
> particular hack.

It will not work. Python doesn't load ELF, it only loads Python code. 
It's the *operating system* which is loading the ELF.

[snip]
> Just a wild guess.

Very wild ;-)

And like my sig below says... Don't assume.. that things work they way 
you want them to work.


-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq



From vmilitaru at sympatico.ca  Thu Jun 10 02:53:32 2004
From: vmilitaru at sympatico.ca (Vio)
Date: Wed, 09 Jun 2004 20:53:32 -0400
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <40C7B10C.7040701@sympatico.ca>

followup...

I think what I'm looking for is "PyCObject_Import()",
which I found in wxPython.h

------------------------
// This needs to be called in modules that make calls to any of the 
functions
// exported by the wxPython API. It sets a static pointer to a structure of
// function pointers located in the wx._core extension module.
static void wxPyCoreAPI_IMPORT() {
wxPyCoreAPIPtr = (wxPyCoreAPI*)PyCObject_Import("wx._core", "_wxPyCoreAPI");
if (! wxPyCoreAPIPtr)
wxPyCoreAPIPtr = (wxPyCoreAPI*)PyCObject_Import("_core", "_wxPyCoreAPI");
}
------------------------

wx._core referred here is file "_core.so". Unfortunately, PyCObject_Import()
does not seem to be documented in the Python API docs.

More digging required...
Vio





From vmilitaru at sympatico.ca  Thu Jun 10 03:30:47 2004
From: vmilitaru at sympatico.ca (Vio)
Date: Wed, 09 Jun 2004 21:30:47 -0400
Subject: [C++-sig] importing embedded dynamic lib
Message-ID: <40C7B9C7.90004@sympatico.ca>

>
>
>It will not work. Python doesn't load ELF, it only loads Python code. 
>It's the *operating system* which is loading the ELF.
>

I *finally* got it. (with convincing args from python sources)

>-- Grafik - Don't Assume Anything
>
But then that's why we're called newbies, as long as we don't understand 
how
the world turns, we assume Elvis still lives.
Thanks for your help!
Vio





From grafik.list at redshift-software.com  Thu Jun 10 06:45:26 2004
From: grafik.list at redshift-software.com (Rene Rivera)
Date: Wed, 09 Jun 2004 23:45:26 -0500
Subject: [C++-sig] importing embedded dynamic lib
In-Reply-To: <40C7B9C7.90004@sympatico.ca>
References: <40C7B9C7.90004@sympatico.ca>
Message-ID: <40C7E766.8070402@redshift-software.com>

Vio wrote:

>> -- Grafik - Don't Assume Anything
>>
> But then that's why we're called newbies, as long as we don't understand 
> how
> the world turns, we assume Elvis still lives.

Of course Elvis Costello lives ;-)

> Thanks for your help!

Your welcome.


-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq



From Pere.Mato at cern.ch  Thu Jun 10 09:35:31 2004
From: Pere.Mato at cern.ch (Pere Mato Vila)
Date: Thu, 10 Jun 2004 09:35:31 +0200
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
Message-ID: <5BF9E41B6C9F2546B624F30FE840C1225FDBBA@cernxchg21.cern.ch>

Linux Redhad 7.3 with gcc 3.2


> -----Original Message-----
> From: c++-sig-bounces at python.org 
> [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> Grosse-Kunstleve
> Sent: 09 June 2004 21:10
> To: Development of Python/C++ integration
> Subject: RE: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> 
> 
> --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > My test program to show the problem contains only one line:
> > 
> > #include "boost/python.hpp"
> > 
> > and therefore I can not move it before anything.
> 
> Hm. What is your platform? Linux (which), Cygwin?
> 
> 
> > > -----Original Message-----
> > > From: c++-sig-bounces at python.org
> > > [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> > > Grosse-Kunstleve
> > > Sent: 08 June 2004 12:26
> > > To: Development of Python/C++ integration
> > > Subject: Re: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> > > 
> > > 
> > > --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > > > I am having the following compilation warnings when including
> > > > "boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 
> > > and gcc 3.2.
> > > > Is anything I am doing wrong?
> > > 
> > > Due to a Python requirement the boost/python.hpp header has
> > > to be included before any system header. Please move the 
> > > python.hpp include to the very top of your file. Let us know 
> > > if it doesn't help. Ralf
> > > 
> > > 
> > > 
> > > 	
> > > 		
> > > __________________________________
> > > Do you Yahoo!?
> > > Friends.  Fun.  Try the all-new Yahoo! Messenger.
> > > http://messenger.yahoo.com/ 
> > > 
> > > _______________________________________________
> > > 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
> 
> 
> 
> 	
> 		
> __________________________________
> Do you Yahoo!?
> Friends.  Fun.  Try the all-new Yahoo! Messenger. 
> http://messenger.yahoo.com/ 
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 



From olivier.ravard at novagrid.com  Thu Jun 10 13:15:32 2004
From: olivier.ravard at novagrid.com (Olivier Ravard)
Date: Thu, 10 Jun 2004 13:15:32 +0200
Subject: [C++-sig] Compilation options
Message-ID: <000c01c44edc$3f0c0320$3d521481@ravard>

Hi,

I compiled some modules with boost.build and mingw tool.
How can I do to tell bjam to compile with the -O3 option ?

Thanks.

O.R.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040610/3ba96236/attachment.htm>

From dave at boost-consulting.com  Thu Jun 10 14:35:10 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 10 Jun 2004 08:35:10 -0400
Subject: [C++-sig] Re: Compilation options
References: <000c01c44edc$3f0c0320$3d521481@ravard>
Message-ID: <ud647d20x.fsf@boost-consulting.com>

"Olivier Ravard" <olivier.ravard at novagrid.com> writes:

> Hi,
>
> I compiled some modules with boost.build and mingw tool.
> How can I do to tell bjam to compile with the -O3 option ?
>
> Thanks.

Are you sure you don't just want a full release build?

  bjam -sTOOLS=mingw -sBUILD=release

If you really *only* want the -03 flag,

  bjam -sTOOLS=mingw "-sBUILD=<cxxflags>-O3"

but that's usually not what you want.


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




From olivier.ravard at novagrid.com  Thu Jun 10 15:14:03 2004
From: olivier.ravard at novagrid.com (Olivier Ravard)
Date: Thu, 10 Jun 2004 15:14:03 +0200
Subject: [C++-sig] Re: Compilation options
References: <000c01c44edc$3f0c0320$3d521481@ravard>
	<ud647d20x.fsf@boost-consulting.com>
Message-ID: <002f01c44eec$cd8f5e70$3d521481@ravard>


----- Original Message ----- 
From: "David Abrahams" <dave at boost-consulting.com>
To: <c++-sig at python.org>
Sent: Thursday, June 10, 2004 2:35 PM
Subject: [C++-sig] Re: Compilation options


> "Olivier Ravard" <olivier.ravard at novagrid.com> writes:
> 
> > Hi,
> >
> > I compiled some modules with boost.build and mingw tool.
> > How can I do to tell bjam to compile with the -O3 option ?
> >
> > Thanks.
> 
> Are you sure you don't just want a full release build?
> 
>   bjam -sTOOLS=mingw -sBUILD=release
> 
> If you really *only* want the -03 flag,
> 
>   bjam -sTOOLS=mingw "-sBUILD=<cxxflags>-O3"
> 
> but that's usually not what you want.
> 

In fact, I want a release build with the -O3 option...

> 
> -- 
> Dave Abrahams
> Boost Consulting
> http://www.boost-consulting.com
> 
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig



From rwgk at yahoo.com  Thu Jun 10 16:38:12 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 10 Jun 2004 07:38:12 -0700 (PDT)
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <5BF9E41B6C9F2546B624F30FE840C1225FDBBA@cernxchg21.cern.ch>
Message-ID: <20040610143812.99879.qmail@web20212.mail.yahoo.com>

I can reproduce the warning... but only after removing the -w from the g++
command line. 

With g++ I am always compiling with -w because I get useless warnings otherwise
that I cannot disable with other switches. I rely on EDG compilers to check my
code for standard conformance; for this purpose the EDG compilers generally do
a much better job than g++.

If you don't want to use -w (I don't know how to tell bjam to insert the -w)
you have to analyze how exactly the conflicting definitions of _POSIX_C_SOURCE
come about. Maybe we can do something with #ifdef in the right places. However,
I doubt it is worth the effort unless you can demonstrate the same problem on a
more recent platform with a more recent compiler.

Ralf

--- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> Linux Redhad 7.3 with gcc 3.2
> 
> 
> > -----Original Message-----
> > From: c++-sig-bounces at python.org 
> > [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> > Grosse-Kunstleve
> > Sent: 09 June 2004 21:10
> > To: Development of Python/C++ integration
> > Subject: RE: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> > 
> > 
> > --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > > My test program to show the problem contains only one line:
> > > 
> > > #include "boost/python.hpp"
> > > 
> > > and therefore I can not move it before anything.
> > 
> > Hm. What is your platform? Linux (which), Cygwin?
> > 
> > 
> > > > -----Original Message-----
> > > > From: c++-sig-bounces at python.org
> > > > [mailto:c++-sig-bounces at python.org] On Behalf Of Ralf W. 
> > > > Grosse-Kunstleve
> > > > Sent: 08 June 2004 12:26
> > > > To: Development of Python/C++ integration
> > > > Subject: Re: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
> > > > 
> > > > 
> > > > --- Pere Mato Vila <Pere.Mato at cern.ch> wrote:
> > > > > I am having the following compilation warnings when including
> > > > > "boost/python.hpp". I am using Boost 1.31.0 Python 2.3.3 
> > > > and gcc 3.2.
> > > > > Is anything I am doing wrong?
> > > > 
> > > > Due to a Python requirement the boost/python.hpp header has
> > > > to be included before any system header. Please move the 
> > > > python.hpp include to the very top of your file. Let us know 
> > > > if it doesn't help. Ralf



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From shin1_morita at yahoo.co.jp  Thu Jun 10 17:19:41 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Fri, 11 Jun 2004 00:19:41 +0900 (JST)
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
In-Reply-To: <uk6ygfvwv.fsf@boost-consulting.com>
Message-ID: <20040610151941.77721.qmail@web2306.mail.yahoo.co.jp>

> Normally for me to help you, you'd need to post the
> code you used that

I attached the code to my first posting ...
But next time I would write code in the message.

> caused the problem, but in this case I think it's
> pretty clear that
> when you wrapped Derived you didn't put bases<Base>
> in the 
> class_< ... > argument list.

No, I din not wrap "Derived".
"Derived" is defined in python code.

Writing this, I've just found the cause.
I exposed in foo.cc:


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


__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/




From Paul_Kunz at slac.stanford.edu  Thu Jun 10 17:46:04 2004
From: Paul_Kunz at slac.stanford.edu (Paul F. Kunz)
Date: Thu, 10 Jun 2004 08:46:04 -0700
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <20040610143812.99879.qmail@web20212.mail.yahoo.com>
	(rwgk@yahoo.com)
References: <20040610143812.99879.qmail@web20212.mail.yahoo.com>
Message-ID: <200406101546.i5AFk4C24429@libra3.slac.stanford.edu>

>>>>> On Thu, 10 Jun 2004 07:38:12 -0700 (PDT), "Ralf W. Grosse-Kunstleve" <rwgk at yahoo.com> said:

> If you don't want to use -w (I don't know how to tell bjam to insert
> the -w) you have to analyze how exactly the conflicting definitions
> of _POSIX_C_SOURCE come about. Maybe we can do something with #ifdef
> in the right places. However, I doubt it is worth the effort unless
> you can demonstrate the same problem on a more recent platform with
> a more recent compiler.

   I see the same warning with Red Hat Linux 9 and gcc 3.4.0.   Is
that recent enough?   Also see them when I build boost 1.31.0 with
bjam.   




From tbm at esperanto.org  Fri Jun 11 05:40:16 2004
From: tbm at esperanto.org (tbm at esperanto.org)
Date: Fri, 11 Jun 2004 03:40:16 +0000
Subject: [C++-sig] software
In-Reply-To: <5E595638E4887FJG@python.org>
References: <5E595638E4887FJG@python.org>
Message-ID: <04AL1E4G9H00HA0K@esperanto.org>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://GNNDJA.biz/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://FKEFCK.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


ekclrtpb juaxe cnrqzri edtbu dtbpfzu qcgi wxkgql 
ewor ogrhowj iyemq gexgyntf yecj kuwoqkt 
gqamhxy pvd elit iohyxlx cohmj d cbojt 




From shin1_morita at yahoo.co.jp  Thu Jun 10 17:51:54 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Fri, 11 Jun 2004 00:51:54 +0900 (JST)
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
Message-ID: <20040610155154.86156.qmail@web2307.mail.yahoo.co.jp>

Sorry, I sent incomplete message.

> Writing this, I've just found the cause.
> I exposed Base in foo.cc:

class_<Base, BaseWrap, boost::noncopyable>("Base",
no_init);

and defined Derived in test_foo.py:

class Derived(Base):
    def __init__(self):
        pass # because I exposed with no_init.
    def f(self):
        return 42

This is wrong.
Next I removed "no_init" and "__init__" above like this:

class_<Base, BaseWrap, boost::noncopyable>("Base");

class Derived(Base):
    # Base.__init__ will be called.
    def f(self):
        return 42

Finally I got *passed*.

Base.__init__ must be called to initialize Derived
instance correctly, is this right?
If so, I think that "no_init" is not for abstract base
class, but just for only noncreatable class.

Why dose this page in the tutorial uses "no_init" to
explain abstract base class?
http://boost.org/libs/python/doc/tutorial/doc/class_virtual_functions.html

Thanks.


__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/




From dave at boost-consulting.com  Thu Jun 10 18:43:57 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 10 Jun 2004 12:43:57 -0400
Subject: [C++-sig] Re: Compilation options
References: <000c01c44edc$3f0c0320$3d521481@ravard>
	<ud647d20x.fsf@boost-consulting.com>
	<002f01c44eec$cd8f5e70$3d521481@ravard>
Message-ID: <u659z9xde.fsf@boost-consulting.com>

"Olivier Ravard" <olivier.ravard at novagrid.com> writes:

> ----- Original Message ----- 
> From: "David Abrahams" <dave at boost-consulting.com>
> To: <c++-sig at python.org>
> Sent: Thursday, June 10, 2004 2:35 PM
> Subject: [C++-sig] Re: Compilation options
>
>
>> "Olivier Ravard" <olivier.ravard at novagrid.com> writes:
>> 
>> > Hi,
>> >
>> > I compiled some modules with boost.build and mingw tool.
>> > How can I do to tell bjam to compile with the -O3 option ?
>> >
>> > Thanks.
>> 
>> Are you sure you don't just want a full release build?
>> 
>>   bjam -sTOOLS=mingw -sBUILD=release
>> 
>> If you really *only* want the -03 flag,
>> 
>>   bjam -sTOOLS=mingw "-sBUILD=<cxxflags>-O3"
>> 
>> but that's usually not what you want.
>> 
>
> In fact, I want a release build with the -O3 option...

I believe the release build already uses -O3; you can add -d+2 to the
beginning of the bjam command-line to check.

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




From dave at boost-consulting.com  Thu Jun 10 18:45:35 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 10 Jun 2004 12:45:35 -0400
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
References: <uk6ygfvwv.fsf@boost-consulting.com>
	<20040610151941.77721.qmail@web2306.mail.yahoo.co.jp>
Message-ID: <uzn7b8iq8.fsf@boost-consulting.com>

Shin-ichi MORITA <shin1_morita at yahoo.co.jp> writes:

>> Normally for me to help you, you'd need to post the
>> code you used that
>
> I attached the code to my first posting ...

I'm blind!  Sorry!

> But next time I would write code in the message.

No, an attachment is fine, thanks.

>> caused the problem, but in this case I think it's
>> pretty clear that
>> when you wrapped Derived you didn't put bases<Base>
>> in the 
>> class_< ... > argument list.
>
> No, I din not wrap "Derived".
> "Derived" is defined in python code.
>
> Writing this, I've just found the cause.
> I exposed in foo.cc:

...continued

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




From dave at boost-consulting.com  Thu Jun 10 18:46:18 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Thu, 10 Jun 2004 12:46:18 -0400
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
References: <20040610155154.86156.qmail@web2307.mail.yahoo.co.jp>
Message-ID: <uu0xj8ip1.fsf@boost-consulting.com>

Shin-ichi MORITA <shin1_morita at yahoo.co.jp> writes:

> Sorry, I sent incomplete message.
>
>> Writing this, I've just found the cause.
>> I exposed Base in foo.cc:
>
> class_<Base, BaseWrap, boost::noncopyable>("Base",
> no_init);
>
> and defined Derived in test_foo.py:
>
> class Derived(Base):
>     def __init__(self):
>         pass # because I exposed with no_init.
>     def f(self):
>         return 42
>
> This is wrong.
> Next I removed "no_init" and "__init__" above like this:
>
> class_<Base, BaseWrap, boost::noncopyable>("Base");
>
> class Derived(Base):
>     # Base.__init__ will be called.
>     def f(self):
>         return 42
>
> Finally I got *passed*.
>
> Base.__init__ must be called to initialize Derived
> instance correctly, is this right?

Correct.

> If so, I think that "no_init" is not for abstract base
> class, but just for only noncreatable class.

Correct, though I hope to fix that one day.
>
> Why dose this page in the tutorial uses "no_init" to
> explain abstract base class?
> http://boost.org/libs/python/doc/tutorial/doc/class_virtual_functions.html

Because it's wrong, I guess :(

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




From shin1_morita at yahoo.co.jp  Thu Jun 10 19:16:12 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Fri, 11 Jun 2004 02:16:12 +0900 (JST)
Subject: [C++-sig] Re: call_f(Derived) did not match C++ signature
In-Reply-To: <uu0xj8ip1.fsf@boost-consulting.com>
Message-ID: <20040610171612.39624.qmail@web2301.mail.yahoo.co.jp>

Thanks David, ~aashish, and every one.
Now my problem has been resolved :-)


__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/




From rwgk at yahoo.com  Thu Jun 10 19:59:31 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 10 Jun 2004 10:59:31 -0700 (PDT)
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <200406101546.i5AFk4C24429@libra3.slac.stanford.edu>
Message-ID: <20040610175931.74078.qmail@web20206.mail.yahoo.com>

--- "Paul F. Kunz" <Paul_Kunz at slac.stanford.edu> wrote:
> >>>>> On Thu, 10 Jun 2004 07:38:12 -0700 (PDT), "Ralf W. Grosse-Kunstleve"
> <rwgk at yahoo.com> said:
> 
> > If you don't want to use -w (I don't know how to tell bjam to insert
> > the -w) you have to analyze how exactly the conflicting definitions
> > of _POSIX_C_SOURCE come about. Maybe we can do something with #ifdef
> > in the right places. However, I doubt it is worth the effort unless
> > you can demonstrate the same problem on a more recent platform with
> > a more recent compiler.
> 
>    I see the same warning with Red Hat Linux 9 and gcc 3.4.0.   Is
> that recent enough?   Also see them when I build boost 1.31.0 with
> bjam.   

Surely recent enough that there is hope that some developer will listen. Now we
need a volunteer to

1. Reduce the problem. I.e. strip down the includes to a minimum. I'd expect
the critial point to be in boost/python/detail/prefix.hpp. See the comment in
there.

2. Potentially follow up on this:

http://sourceforge.net/tracker/?func=detail&aid=663397&group_id=5470&atid=105470

My concerns were brushed away quite rudely and I am not keen to pick this up
again.

Ralf



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From Paul_Kunz at slac.stanford.edu  Fri Jun 11 01:27:57 2004
From: Paul_Kunz at slac.stanford.edu (Paul F. Kunz)
Date: Thu, 10 Jun 2004 16:27:57 -0700
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <20040610175931.74078.qmail@web20206.mail.yahoo.com>
	(rwgk@yahoo.com)
References: <20040610175931.74078.qmail@web20206.mail.yahoo.com>
Message-ID: <200406102327.i5ANRvb12047@libra3.slac.stanford.edu>

>>>>> On Thu, 10 Jun 2004 10:59:31 -0700 (PDT), "Ralf W. Grosse-Kunstleve" <rwgk at yahoo.com> said:

> Surely recent enough that there is hope that some developer will
> listen. Now we need a volunteer to

> 1. Reduce the problem. I.e. strip down the includes to a
> minimum. I'd expect the critial point to be in
> boost/python/detail/prefix.hpp. See the comment in there.

   Here goes...

foo.cxx ...
----
#include "boost/python.hpp"

int main ()
{
  return 0;
}
----

compile with

g++  -c foo.cxx -I/usr/local/include/boost-1_31 -I/usr/local/include/python2.3 -ftemplate-depth-60

Results are...

cd /u/ek/pfkeb/home/temp/
g++  -c foo.cxx -I/usr/local/include/boost-1_31 -I/usr/local/include/python2.3 -ftemplate-depth-60
In file included from /usr/local/include/python2.3/Python.h:8,
                 from /usr/local/include/boost-1_31/boost/python/detail/wrap_python.hpp:121,
                 from /usr/local/include/boost-1_31/boost/python/detail/prefix.hpp:13,
                 from /usr/local/include/boost-1_31/boost/python/args.hpp:9,
                 from /usr/local/include/boost-1_31/boost/python.hpp:12,
                 from foo.cxx:1:
/usr/local/include/python2.3/pyconfig.h:847: warning: `_POSIX_C_SOURCE' redefined
/usr/include/features.h:171: warning: this is the location of the previous definition

Compilation finished at Thu Jun 10 16:24:24



From rwgk at yahoo.com  Fri Jun 11 02:17:37 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 10 Jun 2004 17:17:37 -0700 (PDT)
Subject: [C++-sig] [PATCH] unused variables.
In-Reply-To: <Pine.LNX.4.44.0405251619300.15190-100000@pccms185.cern.ch>
Message-ID: <20040611001737.46277.qmail@web20205.mail.yahoo.com>

Done, finally!

--- Giulio Eulisse <eulisse at mail.cern.ch> wrote:
> Could someone please patch boost/boost.python as follows? Those unused
> variables generate a lot of warnings when full warning report is used. 
> Thanks.
> 
> Ciao,
> Giulio
> 
> Index: boost/python/class.hpp
> ===================================================================
> RCS file: /cvsroot/boost/boost/boost/python/class.hpp,v
> retrieving revision 1.83.2.2
> diff -u -r1.83.2.2 class.hpp
> --- boost/python/class.hpp????? 7 Jan 2004 14:08:10 -0000?????? 1.83.2.2
> +++ boost/python/class.hpp????? 25 May 2004 13:35:12 -0000
> @@ -556,7 +556,7 @@
> ???? template <class Fn, class Helper>
> ???? inline void def_default(
> ???????? char const* name
> -??????? , Fn fn
> +??????? , Fn /*fn*/
> ???????? , Helper const& helper
> ???????? , mpl::bool_<true>)
> ???? {
> Index: boost/python/detail/caller.hpp
> ===================================================================
> RCS file: /cvsroot/boost/boost/boost/python/detail/caller.hpp,v
> retrieving revision 1.21.2.1
> diff -u -r1.21.2.1 caller.hpp
> --- boost/python/detail/caller.hpp????? 6 Jan 2004 13:08:25 -0000??????
> 1.21.2.1
> +++ boost/python/detail/caller.hpp????? 25 May 2004 13:35:12 -0000
> @@ -82,7 +82,7 @@
> 
> ?template <class ArgPackage, class ResultConverter>
> ?inline ResultConverter create_result_converter(
> -??? ArgPackage const& args_
> +??? ArgPackage const& /*args_*/
> ?? , ResultConverter*
> ?? , ...
> ?)



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From rwgk at yahoo.com  Fri Jun 11 03:15:59 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 10 Jun 2004 18:15:59 -0700 (PDT)
Subject: [C++-sig] warning: "_POSIX_C_SOURCE" redefined
In-Reply-To: <200406102327.i5ANRvb12047@libra3.slac.stanford.edu>
Message-ID: <20040611011559.7330.qmail@web20208.mail.yahoo.com>

--- "Paul F. Kunz" <Paul_Kunz at slac.stanford.edu> wrote:
> >>>>> On Thu, 10 Jun 2004 10:59:31 -0700 (PDT), "Ralf W. Grosse-Kunstleve"
> <rwgk at yahoo.com> said:
> 
> > Surely recent enough that there is hope that some developer will
> > listen. Now we need a volunteer to
> 
> > 1. Reduce the problem. I.e. strip down the includes to a
> > minimum. I'd expect the critial point to be in
> > boost/python/detail/prefix.hpp. See the comment in there.
> 
>    Here goes...
> 
> foo.cxx ...
> ----
> #include "boost/python.hpp"

The Python developers will not be interested in this. If you absolutely have to
avoid -w, and if you cannot tolerate the warning, start with
boost/detail/prefix.hpp, the Python documentation and my bug report. The goal
would be to show that you need to include a certain system file before
including Python.h/pyconfig.h. Maybe you get lucky and you find a way to avoid
the warning with some clever code rearrangements in boost, maybe it is
impossible given the harsh approach of Python.h/pyconfig.h. In my experience
resolving this type of problem can easily absorb a few days worth of time. I
still believe it is not worth the effort.
Ralf



	
		
__________________________________
Do you Yahoo!?
Friends.  Fun.  Try the all-new Yahoo! Messenger.
http://messenger.yahoo.com/ 



From amber at processed.users.sf.meat.com  Mon Jun 14 01:35:39 2004
From: amber at processed.users.sf.meat.com (amber at processed.users.sf.meat.com)
Date: Sun, 13 Jun 2004 23:35:39 +0000
Subject: [C++-sig] software
In-Reply-To: <DDL8G80319AIH5FG@python.org>
References: <DDL8G80319AIH5FG@python.org>
Message-ID: <4H3L763G724J96LL@processed.users.sf.meat.com>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://FJGCNA.info/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://FKEFCK.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


auzqhxhz wbgdo rhwutwe rvvcm rqiftwf nvxv reunga 
rvta qntntai ncwgu gkmcqbyn xprq ezrxulx 
kbfaugi tig ndmz qpprnfp eyhix a gqldd 




From laurenz at t-online.de  Mon Jun 14 06:00:16 2004
From: laurenz at t-online.de (yolanda)
Date: Mon, 14 Jun 2004 04:00:16 +0000
Subject: [C++-sig] access information
Message-ID: <200406140406.i5E46126097381@mxzilla1.xs4all.nl>

Dear 
Great s,e,x ! Awesome videos! Enchanting stories!
On the best adu1t site Love For Lust

Special offer:
Buy V1a,gra and get Free Access

http://www.loveforlust.biz/?m=c++-sig at python.org






From pierre.barbier at cirad.fr  Mon Jun 14 09:44:00 2004
From: pierre.barbier at cirad.fr (Pierre Barbier de Reuille)
Date: Mon, 14 Jun 2004 09:44:00 +0200
Subject: [C++-sig] multiple to_python converters ?
In-Reply-To: <200406092157.02366.phil@freehackers.org>
References: <200406092157.02366.phil@freehackers.org>
Message-ID: <1087199040.734.1.camel@pbarbier>

Mmmh ... a runtime _error_ ? that's strange : I had this problem but it
was only a warning and my code worked properly (only one of the two
converter was actually used ... so if it's the same it's no problem). Or
perhaps your problem is due to having two libraries defining the same
symbols ! If that's the case, you should have a third library which
includes the common symbols and link this third library with the two
first.

Pierre

Le mer 09/06/2004 ? 21:57, Philippe Fremy a ?crit :
> Hi,
> 
> I have the case where I have two python modules, both declaring the same 
> to_python converter. The modules may be used separately or together. The 
> problem is that if used together, I get (you see me coming I guess) a 
> RunTimeError. Only one to_python converter may be registerted at a time for 
> a given type.
> 
> I looked in the archives and on google. I saw two references to discussions 
> of this problem and even a possible solution to be included in boost. But 
> checking the documentation today, I could not find any hint on how to solve 
> this problem.
> 
> So, is the problem still here ?
> 
> 	regards,
> 
> 	Philippe
-- 
Pierre Barbier de Reuille

INRA - UMR Cirad/Inra/Cnrs/Univ.MontpellierII AMAP
Botanique et Bio-informatique de l'Architecture des Plantes
TA40/PSII, Boulevard de la Lironde
34398 MONTPELLIER CEDEX 5, France

tel   : (33) 4 67 61 65 77    fax   : (33) 4 67 61 56 68 




From edelrio at b.maus.de  Tue Jun 15 03:47:25 2004
From: edelrio at b.maus.de (edelrio at b.maus.de)
Date: Tue, 15 Jun 2004 01:47:25 +0000
Subject: [C++-sig] access information
In-Reply-To: <F869GF0ELB2HDALB@python.org>
References: <F869GF0ELB2HDALB@python.org>
Message-ID: <72KE8K4HCKLJ07JD@b.maus.de>

Dear C++-sig
Great s,e,x ! Awesome videos! Enchanting stories!
On the best adu1t site Love For Lust

Special offer:
Buy V1a,gra and get Free Access

http://www.loveforlust.biz/?m=c++-sig at python.org


mdvakqwk cwmh iebxn pyf xezhtz 
ltoan dd idnaaacx m igrkegb ckzh 
kgokribr xewh tkhtl jfa ulrfpy 
htjga wl ptfziyxn v ewglftx ktqy 
kxojetvj irxz aczjb xzw cwcocv 
lmoek bb algihqta n tbqrlrp koth 
qtimvxda iecj ptuyz myn gagzlz 
oxnif gr oumlxmwc p gryaqyn omup 





From dave at boost-consulting.com  Mon Jun 14 17:58:19 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Mon, 14 Jun 2004 11:58:19 -0400
Subject: [C++-sig] Re: multiple to_python converters ?
References: <200406092157.02366.phil@freehackers.org>
Message-ID: <ubrjmjfms.fsf@boost-consulting.com>

Philippe Fremy <phil at freehackers.org> writes:

> Hi,
>
> I have the case where I have two python modules, both declaring the same 
> to_python converter. The modules may be used separately or together. The 
> problem is that if used together, I get (you see me coming I guess) a 
> RunTimeError. Only one to_python converter may be registerted at a time for 
> a given type.
>
> I looked in the archives and on google. I saw two references to discussions 
> of this problem and even a possible solution to be included in boost. But 
> checking the documentation today, I could not find any hint on how to solve 
> this problem.
>
> So, is the problem still here ?

That error was changed to a warning on 11-Sep-03 and is in Boost
version 1.31.0

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




From paul.bridger at paradise.net.nz  Tue Jun 15 02:09:04 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Tue, 15 Jun 2004 12:09:04 +1200
Subject: [C++-sig] Pyste: wrapper function for overloaded function problems
Message-ID: <40CE3E20.1020409@paradise.net.nz>

I am trying to expose a C++ member function with a bunch of default 
arguments, but I also want to provide a C++ wrapper for it.
The problem I've had is that Pyste still generates the overload machinery 
even though it's not wanted since my wrapper function takes different arguments.

Any hints? I'm going digging in the Pyste source.

Paul Bridger



From christophe.grimault at novagrid.com  Tue Jun 15 10:28:53 2004
From: christophe.grimault at novagrid.com (Christophe Grimault)
Date: Tue, 15 Jun 2004 10:28:53 +0200
Subject: [C++-sig] Size difference between .so and .dll files
References: <40CE3E20.1020409@paradise.net.nz>
Message-ID: <40CEB345.6010804@novagrid.com>

Hi all,

We have 2 versions of the same  software. One on linux RH9, and one  Win 
XP.
Both of them compile and work fine. They use several C++ files compiled with
boost to interact with Python. On windows, a dll with a size of a few 
hundreds
of ko. On linux, the size for the equivalent .so is a few Mo.

Have we missed something ? Any hints to reduce this size, if possible, 
are welcome !

Thanks in advance ...

C. Grimault




From rwgk at yahoo.com  Tue Jun 15 16:39:48 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Tue, 15 Jun 2004 07:39:48 -0700 (PDT)
Subject: [C++-sig] Size difference between .so and .dll files
In-Reply-To: <40CEB345.6010804@novagrid.com>
Message-ID: <20040615143948.43122.qmail@web20221.mail.yahoo.com>

--- Christophe Grimault <christophe.grimault at novagrid.com> wrote:
> We have 2 versions of the same  software. One on linux RH9, and one  Win 
> XP.
> Both of them compile and work fine. They use several C++ files compiled with
> boost to interact with Python. On windows, a dll with a size of a few 
> hundreds
> of ko. On linux, the size for the equivalent .so is a few Mo.
> 
> Have we missed something ? Any hints to reduce this size, if possible, 
> are welcome !

You seem to imply a size difference of three orders of magnitude (kB vs. MB).
In my experience it is more a factor of roughly 2-4. I never looked in detail,
but you can compare the sizes of the downloads posted here:

http://cci.lbl.gov/cctbx_build/

E.g. the Redhat 9 download is 6.8MB, the Windows download is 5.0MB. This
comparison is not a really good size-benchmark (gunzip vs. info-zip, Python
sources are the same in both downloads), but it gives you a ballpark estimate.

Could it be that you are comparing debug builds? Last time I did a Linux debug
build (ages ago...) I noticed that the object files are enormous. I didn't pay
attention to the sizes of the final .so files, though.

Ralf



		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail 



From s_sourceforge at nedprod.com  Wed Jun 16 03:04:32 2004
From: s_sourceforge at nedprod.com (Niall Douglas)
Date: Wed, 16 Jun 2004 02:04:32 +0100
Subject: [C++-sig] Size difference between .so and .dll files
In-Reply-To: <40CEB345.6010804@novagrid.com>
Message-ID: <40CFAAB0.32271.25FB68FE@localhost>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 15 Jun 2004 at 10:28, Christophe Grimault wrote:

> We have 2 versions of the same  software. One on linux RH9, and one 
> Win XP. Both of them compile and work fine. They use several C++ files
> compiled with boost to interact with Python. On windows, a dll with a
> size of a few hundreds of ko. On linux, the size for the equivalent
> .so is a few Mo.
> 
> Have we missed something ? Any hints to reduce this size, if possible,
> are welcome !

1. Use GCC 3.4. This should knock 50% off your binary.
2. Patch GCC with my patch at 
http://www.nedprod.com/programs/gccvisibility.html. With -
fvisibility=hidden this should knock a further few hundred Kb off 
your binary and produce better quality code too.

No matter what you do, the Linux binary will be at least 20% larger 
than your MSVC binary.

Cheers,
Niall





-----BEGIN PGP SIGNATURE-----
Version: idw's PGP-Frontend 4.9.6.1 / 9-2003 + PGP 8.0.2

iQA/AwUBQM+cocEcvDLFGKbPEQLS8wCbBUA+oewaC4wT2JlPiE3ZpB47xBUAoK6o
y6Gdwodyc/ngkqG3aWpsOLPC
=sbJi
-----END PGP SIGNATURE-----



From caustin at gmail.com  Thu Jun 17 10:12:37 2004
From: caustin at gmail.com (Chad Austin)
Date: Thu, 17 Jun 2004 03:12:37 -0500
Subject: [C++-sig] Making Boost.Python understand DerivedSafe
Message-ID: <2096514e0406170112773a0e76@mail.gmail.com>

(I am not yet subscribed to the list...  it seems mailman is having
trouble sending me the authorization.  Please CC me on replies.. 
Thanks.)

Hi all,

I just started using Boost.Python to embed Python in a project of
mine.  I spent a couple days evaluating, testing, and studying, but it
seems to have paid off!  Great job on making such a time-saving
library!

I am wrapping an internally refcounted class called CutScene. 
CutScene objects are referenced by RefPtr<CutScene> smart pointers,
typedef'd as CutScenePtr.  I export the class with:

class_<CutScene, CutScenePtr>("CutScene").def /*etc*/;

So far, so good.  The problem comes about when I want to pass a
CutScenePtr into object::operator() to call a function.  If I pass a
CutScenePtr object in, it says something about reject_raw_object_ptr:

d:\Projects\empyrean\empyrean\third-party-vc7\include\boost\python\converter\arg_to_python.hpp(244)
: error C2784: 'void
boost::python::converter::detail::reject_raw_object_ptr(T *)' : could
not deduce template argument for 'T1 *' from 'pyr::RefPtr<T>'
        with
        [
            T=pyr::CutScene
        ]

If I pass the result of CutScenePtr::get() into operator() I get a
runtime error about Boost.Python not knowing how to convert
RefDerivedSafe<T>* to a Python object.  This makes sense, because
RefPtr's implicit conversion operator, operator->, and get() all
return RefDerivedSafe<T>*. (RefDerivedSafe<T> is a special class
derived from T that prevents C++ clients from accidentally calling
delete, ref(), or unref() on a smart pointer. This techique is used in
both ATL and nsCOMPtr in Mozilla).

Here's my question:  Is there a way to tell Boost.Python that
ptr(someCutScenePtr) is valid and should be equivalent to
ptr(someCutScenePtr.get()) and/or that RefDerivedSafe<T>* can be
implicitly converted to T*?  I've tried several different things, and
nothing seems to work...

I'm currently using Boost 1.31 with both gcc 3.3.1 and VS.NET 2003.

Many thanks,
Chad



From christophe.grimault at novagrid.com  Wed Jun 16 18:16:37 2004
From: christophe.grimault at novagrid.com (Christophe Grimault)
Date: Wed, 16 Jun 2004 18:16:37 +0200
Subject: [C++-sig] Size difference between .so and .dll files
References: <40CFAAB0.32271.25FB68FE@localhost>
Message-ID: <40D07265.6080305@novagrid.com>



Niall Douglas wrote:

>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>On 15 Jun 2004 at 10:28, Christophe Grimault wrote:
>
>  
>
>>We have 2 versions of the same  software. One on linux RH9, and one 
>>Win XP. Both of them compile and work fine. They use several C++ files
>>compiled with boost to interact with Python. On windows, a dll with a
>>size of a few hundreds of ko. On linux, the size for the equivalent
>>.so is a few Mo.
>>
>>Have we missed something ? Any hints to reduce this size, if possible,
>>are welcome !
>>    
>>
>
>1. Use GCC 3.4. This should knock 50% off your binary.
>
I have tried the  gcc-3.4. Compared to gcc-3.2 used before, a .so 
typically goes from 5.4 Mo to 4.0 Mo.

>2. Patch GCC with my patch at 
>http://www.nedprod.com/programs/gccvisibility.html. With -
>fvisibility=hidden this should knock a further few hundred Kb off 
>your binary and produce better quality code too.
>
Thanks for the hint. I did not have the time to try yet !

>
>No matter what you do, the Linux binary will be at least 20% larger 
>than your MSVC binary.
>  
>
Last question. I do not compile my .so with bjam but with a separate 
Makefile, outside of the boost architecture. May be I'm also missing 
something in the Makefile that the Jamfile does !? I think i remember 
that, if I try to compile the same .so, inside the boost architecture, I 
get a big improvement. Uhm ?

Thanks a lot !!

>Cheers,
>Niall
>
>
>
>
>
>-----BEGIN PGP SIGNATURE-----
>Version: idw's PGP-Frontend 4.9.6.1 / 9-2003 + PGP 8.0.2
>
>iQA/AwUBQM+cocEcvDLFGKbPEQLS8wCbBUA+oewaC4wT2JlPiE3ZpB47xBUAoK6o
>y6Gdwodyc/ngkqG3aWpsOLPC
>=sbJi
>-----END PGP SIGNATURE-----
>
>_______________________________________________
>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/20040616/29d1b546/attachment.htm>

From timster at mygale.org  Thu Jun 17 18:32:52 2004
From: timster at mygale.org (timster at mygale.org)
Date: Thu, 17 Jun 2004 16:32:52 +0000
Subject: [C++-sig] software
In-Reply-To: <FL5JH891J1866HFC@python.org>
References: <FL5JH891J1866HFC@python.org>
Message-ID: <I9LIB0JH52H3H2E8@mygale.org>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.


Our site is  http://MALMIK.biz/OE017/?affiliate_id=233763&campaign_id=601

uns * ub - scribe http://FKEFCK.biz/diamondtron.php?affiliate_id=233763&campaign_id=601


jymjhxpv xgxel vhkqxnb creyw ytoxcth bwrt ttdpob 
vbxd lnbbodt ucotj fplnynjp wdya umcpozt 
cnudexw zak rjqt xfgkowt tvtni k eyjxa 




From hellfire at icsi.berkeley.edu  Fri Jun 18 23:45:48 2004
From: hellfire at icsi.berkeley.edu (hellfire at icsi.berkeley.edu)
Date: Fri, 18 Jun 2004 21:45:48 +0000
Subject: [C++-sig] software
In-Reply-To: <345CKIJJ17C664JH@python.org>
References: <345CKIJJ17C664JH@python.org>
Message-ID: <B6AHII2JF51E2HC5@icsi.berkeley.edu>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!

Our site is  http://ceazpzh.agelmkb.info/?U7qZWFUXxY_haoUdfchbodt


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.

Our site is  http://fgnexmkb.cgnacmg.info/?MviRixgPpkT9yggrmiud


lcpztmha akfrz iombaaj hmfyu oyvckzf rxpp amzzrx 
hdmp gtoxtjl yqiof mxrwlopx odda muklgzj 
oluqbnv udt pecd xldiljd orrjx b lhrdl 




From emily at t-online.de  Thu Jun 17 09:20:44 2004
From: emily at t-online.de (khoi)
Date: Thu, 17 Jun 2004 07:20:44 +0000
Subject: [C++-sig] access information
Message-ID: <200406170728.i5H7SHC3094908@mxzilla7.xs4all.nl>

Dear 
Great s,e,x ! Awesome videos! Enchanting stories!
On the best adu1t site Love For Lust

Special offer:
Buy V1a,gra and get Free Access

http://www.loveforlust.biz/?m=c++-sig at python.org






From nicodemus at esss.com.br  Wed Jun 16 19:45:04 2004
From: nicodemus at esss.com.br (Nicodemus)
Date: Wed, 16 Jun 2004 14:45:04 -0300
Subject: [C++-sig] Pyste: wrapper function for overloaded function problems
In-Reply-To: <40CE3E20.1020409@paradise.net.nz>
References: <40CE3E20.1020409@paradise.net.nz>
Message-ID: <40D08720.8080201@esss.com.br>

Hi Paul, sorry about the delay,

paul.bridger wrote:

> I am trying to expose a C++ member function with a bunch of default 
> arguments, but I also want to provide a C++ wrapper for it.
> The problem I've had is that Pyste still generates the overload 
> machinery even though it's not wanted since my wrapper function takes 
> different arguments.
>
> Any hints? I'm going digging in the Pyste source.


You want to take a look at the method ExportMethods in ClassExporter.py. ;)
Patches are welcome, as always!

Regards,
Nicodemus.



From dave at boost-consulting.com  Sat Jun 19 13:00:41 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Sat, 19 Jun 2004 07:00:41 -0400
Subject: [C++-sig] Re: Making Boost.Python understand DerivedSafe
References: <2096514e0406170112773a0e76@mail.gmail.com>
Message-ID: <uvfhn3j8m.fsf@boost-consulting.com>

Chad Austin <caustin at gmail.com> writes:

> (I am not yet subscribed to the list...  it seems mailman is having
> trouble sending me the authorization.  Please CC me on replies.. 
> Thanks.)
>
> Hi all,
>
> I just started using Boost.Python to embed Python in a project of
> mine.  I spent a couple days evaluating, testing, and studying, but it
> seems to have paid off!  Great job on making such a time-saving
> library!
>
> I am wrapping an internally refcounted class called CutScene. 
> CutScene objects are referenced by RefPtr<CutScene> smart pointers,
> typedef'd as CutScenePtr.  I export the class with:
>
> class_<CutScene, CutScenePtr>("CutScene").def /*etc*/;
>
> So far, so good.  The problem comes about when I want to pass a
> CutScenePtr into object::operator() to call a function.  If I pass a
> CutScenePtr object in, it says something about reject_raw_object_ptr:
>
> d:\Projects\empyrean\empyrean\third-party-vc7\include\boost\python\converter\arg_to_python.hpp(244)
> : error C2784: 'void
> boost::python::converter::detail::reject_raw_object_ptr(T *)' : could
> not deduce template argument for 'T1 *' from 'pyr::RefPtr<T>'
>         with
>         [
>             T=pyr::CutScene
>         ]

1. Show us more detail: more of the error messages, some sample code,
   etc.

2. As far as I can tell you have used boost::ptr around one of your
   RefPtr<CutScene>s, call it 'x':

         ptr(x)

   boost::ptr is reserved for real pointers.

> If I pass the result of CutScenePtr::get() into operator() I get a
> runtime error about Boost.Python not knowing how to convert
> RefDerivedSafe<T>* to a Python object.  This makes sense, because
> RefPtr's implicit conversion operator, operator->, and get() all
> return RefDerivedSafe<T>*. (RefDerivedSafe<T> is a special class
> derived from T that prevents C++ clients from accidentally calling
> delete, ref(), or unref() on a smart pointer

Bah!

        T* p = x.get();
        delete p;

> . This techique is used in both ATL and nsCOMPtr in Mozilla).

That doesn't make it safe ;->

> Here's my question:  Is there a way to tell Boost.Python that
> ptr(someCutScenePtr) is valid 

No!  Why do you want to use ptr?

> and should be equivalent to
> ptr(someCutScenePtr.get()) and/or that RefDerivedSafe<T>* can be
> implicitly converted to T*?  I've tried several different things, and
> nothing seems to work...

Maybe this helps?

http://www.boost.org/libs/python/doc/v2/register_ptr_to_python.html

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




From jeff.holle at verizon.net  Sat Jun 19 05:54:38 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Fri, 18 Jun 2004 23:54:38 -0400
Subject: [C++-sig] Added to a vector_indexing_suite container in python
Message-ID: <40D3B8FE.9070900@verizon.net>

I am using gcc 3.2 and boost 1_31_0.

In my application, I'm using python is a consumer of data created in C++ 
and housed in vectors wrapped
with vector_indexing_suite.

Now I want to create some Python scripts that are data producers for 
test purposes, but can't seem to be able to populate these containers.

I've noted that a 'append' method is exposed, trying this doesn't add to 
the container.
Attempting to use 'myContainer[0] = something' causes an illegal 
indexing exception on an empty container...

Am I missing something?

Do I have to add my own methods independent of the vector_indexing_suite?



From jeff.holle at verizon.net  Sat Jun 19 11:18:37 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Sat, 19 Jun 2004 05:18:37 -0400
Subject: [C++-sig] [Re: Added to a vector_indexing_suite container in python]
Message-ID: <40D404ED.6020905@verizon.net>

I suspect that my previous post concerning this problem does not provide 
adequite information.
So here is more.

Following is a grep of my files that highlite what I'm trying to wrap 
with pyste and how I'm doing it:

    typedef std::vector<CParameter> Parameters;
    MemberFunction.h:    Parameters& getParameters( void);
    MemberFunction.h:    const Parameters& getParameters( void) const;
    MemberFunction.h:UMLModel::Parameters&
    PygetParameters(UMLModel::CMemberFunction& self);
    MemberFunction.pyste:exclude(a.getParameters)
    MemberFunction.pyste:add_method(a,"PygetParameters")
    MemberFunction.pyste:set_policy(a.PygetParameters,return_value_policy(copy_non_const_reference))
    MemberFunction.pyste:rename(a.PygetParameters,"getParameters")

The PygetParameters function exists because pyste throws errors in 
encountering the overloaded pair that are in-place to facilitate const 
correct coding.

Note the method call policy that I'm applying, "copy_non_const_reference".

It appears that this policy causes a copy of the container in the 
process of calling "getParameters" in python.
I believe this because the following python script:
     aMethod.getParameters().append(aParm)
     print 'number of parameters is %d' % len(aMethod.getParameters())
prints:
    number of parameters is 0

I've added my own "push_back" method to this container in a way similiar 
to the iterator.cpp file.
The same results occur, which are explained if a container copy is going on.

It there a way to avoid this copy?
 





From kevin at kevinatkinson.com  Mon Jun 21 21:22:07 2004
From: kevin at kevinatkinson.com (kevin at kevinatkinson.com)
Date: Mon, 21 Jun 2004 15:22:07 -0400
Subject: [C++-sig] opencascade
Message-ID: <002b01c457c5$1378e380$73c809c0@CPQ30541773923>

This email is addressed to Niki Spahiev; it is in regard to this message on mail.python.org:

http://mail.python.org/pipermail/c++-sig/2003-March/003577.html

I am trying to do precisely the same thing:  generate boost.python wrappers using the metadata exposed in the cdl files of Open Cascade.  Did you have any luck doing this?

I ask because I am anxious to get my hands on a Python wrapper of Open Cascade, and at this point it's looking like I may have to do it myself.  I know there is a fellow named Sandor Racz who has made an incomplete wrapper for 5.0, but I'm not having much luck with it.

I would greatly appreciate any tips you might have in this regard.

Best,

Kevin Atkinson
Software Developer

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040621/02f6b1a1/attachment.htm>

From nickm at sitius.com  Tue Jun 22 00:01:57 2004
From: nickm at sitius.com (Nikolay Mladenov)
Date: Mon, 21 Jun 2004 18:01:57 -0400
Subject: [C++-sig] Re: Making Boost.Python understand DerivedSafe
References: <2096514e0406170112773a0e76@mail.gmail.com>
Message-ID: <40D75AD5.A7240E56@sitius.com>

Try to satisfy the detailed requirements for the "HeldType"
(http://www.boost.org/libs/python/doc/v2/class.html#class_-spec).
I think you'll have to define get_pointer(CutScenePtr).
And maybe specialize python::pointee.

HTH,

Nikolay

Chad Austin wrote:
> 
> (I am not yet subscribed to the list...  it seems mailman is having
> trouble sending me the authorization.  Please CC me on replies..
> Thanks.)
> 
> Hi all,
> 
> I just started using Boost.Python to embed Python in a project of
> mine.  I spent a couple days evaluating, testing, and studying, but it
> seems to have paid off!  Great job on making such a time-saving
> library!
> 
> I am wrapping an internally refcounted class called CutScene.
> CutScene objects are referenced by RefPtr<CutScene> smart pointers,
> typedef'd as CutScenePtr.  I export the class with:
> 
> class_<CutScene, CutScenePtr>("CutScene").def /*etc*/;
> 
> So far, so good.  The problem comes about when I want to pass a
> CutScenePtr into object::operator() to call a function.  If I pass a
> CutScenePtr object in, it says something about reject_raw_object_ptr:
> 
> d:\Projects\empyrean\empyrean\third-party-vc7\include\boost\python\converter\arg_to_python.hpp(244)
> : error C2784: 'void
> boost::python::converter::detail::reject_raw_object_ptr(T *)' : could
> not deduce template argument for 'T1 *' from 'pyr::RefPtr<T>'
>         with
>         [
>             T=pyr::CutScene
>         ]
> 
> If I pass the result of CutScenePtr::get() into operator() I get a
> runtime error about Boost.Python not knowing how to convert
> RefDerivedSafe<T>* to a Python object.  This makes sense, because
> RefPtr's implicit conversion operator, operator->, and get() all
> return RefDerivedSafe<T>*. (RefDerivedSafe<T> is a special class
> derived from T that prevents C++ clients from accidentally calling
> delete, ref(), or unref() on a smart pointer. This techique is used in
> both ATL and nsCOMPtr in Mozilla).
> 
> Here's my question:  Is there a way to tell Boost.Python that
> ptr(someCutScenePtr) is valid and should be equivalent to
> ptr(someCutScenePtr.get()) and/or that RefDerivedSafe<T>* can be
> implicitly converted to T*?  I've tried several different things, and
> nothing seems to work...
> 
> I'm currently using Boost 1.31 with both gcc 3.3.1 and VS.NET 2003.
> 
> Many thanks,
> Chad

-- 
Nikolay Mladenov
Sitius Automation Inc.
www.sitius.com




From Peter.Vandersteegen at intec.Ugent.be  Tue Jun 22 15:07:22 2004
From: Peter.Vandersteegen at intec.Ugent.be (Peter Vandersteegen)
Date: Tue, 22 Jun 2004 15:07:22 +0200
Subject: [C++-sig] Question: compiling library using <python/numeric.cpp>
 extension/wrapper 
Message-ID: <6.1.0.6.0.20040622141832.02667098@pop-tech.intec.UGent.be>

Hello,

I've recently discovered boost-python.   Compiling several examples gave no 
problem.
Working with lists etc also no problems.
Compiling a simple library using the <python/numeric.cpp> 
extension/wrapper  however
resulted in several questions...

Therefor I've one question:

Question1:
Is it correct to state the following?   (something I find confusing)
There are (at least) 2 ways to get a boost-layer between Numeric python and 
c++.  The first one is based on
<Numeric/arrayobject.h>, the other on <python/Numeric.hpp>...

* The project num_util 
(http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/) uses 
the fist method?  They use boost as wrapper, but for the array class, they 
directly use the python numeric array class and not the boost.python array 
class?
* Directly using <python/Numeric.hpp> seperates you from dealing with the 
'raw' python numeric array class?  Added bonus: no need to write 
import_array() in BOOST_PYTHON_MODULE...

Thus, is it correct to say that these are 2 different approaches to convert 
an array from python to c++ ?

thx a lot for reading so far

Peter






From herveus at t2.technion.ac.il  Tue Jun 22 15:25:11 2004
From: herveus at t2.technion.ac.il (herveus at t2.technion.ac.il)
Date: Tue, 22 Jun 2004 13:25:11 +0000
Subject: [C++-sig] software
In-Reply-To: <670I533K0IBL1IKI@python.org>
References: <670I533K0IBL1IKI@python.org>
Message-ID: <6FEBEFKL4I10EKDL@techst02.technion.ac.il>

Microsoft Windows XP Professional 2002 
Retail price: $270.99 Our low Price: $50.00 You Save: $220.00  
 
Adobe Photoshop 7.0  
Retail price: $609.99 Our low Price: $60.00 You Save: $550.00  
 
Microsoft Office XP Professional 2002  
Retail price: $579.99 Our low Price: $60.00 You Save: $510.00  
 
Adobe Illustrator 10
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  
 
Corel Draw Graphics Suite 11
Retail price: $270.99 Our low Price: $60.00 You Save: $210.00  

Delphi 7
Retail price: $404.99 Our low Price: $60.00 You Save: $335.00  
 

And more!!!

Our site is  http://axwny.lddekfan.info/?h0PmjyNQWlUGzhNtcemcyb


Why so cheap?

All the software is OEM- Meaning that you don't get the box and the
manual with your software. All you will receive is the actual
software and your unique registration code. 

All the software is in the English language for PC. Our offers
are unbeatable and we always update our prices to make sure we
provide you with the best possible offers. Hurry up and place
your order, because our supplies are limited.

Our site is  http://dpnlr.lddekfan.info/?h0PmjyNQWlUGzhNifhnxut


ehqojxok vkate cyqfwti mtvgq bmgzhnd ngtz etjztk 
txmq otnicbq xhiyv oxceaqkt qobb ovltraa 
ttnqvpi tyr rttp wmpmyjf nlgdh t lault 




From caustin at gmail.com  Wed Jun 23 02:06:18 2004
From: caustin at gmail.com (Chad Austin)
Date: Tue, 22 Jun 2004 19:06:18 -0500
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
Message-ID: <2096514e0406221706691db56c@mail.gmail.com>

Hi all,

Has anyone successfully used Boost.Python in SUSE 9.0 or SUSE 9.1?  In
9.0 I had to build and install boost 1.31 from source, but 9.1 comes
with a boost 1.31 package.  Either way, merely #including
<boost/python.hpp> causes an internal compiler error in a variety of
places.  Note: SUSE 9.0 uses gcc 3.3.1 (+ SUSE-specific patches?) and
SUSE 9.1 uses gcc 3.3.3 (+ patches again?).  If I build gcc 3.4.0 from
source, Boost.Python compiles, but does not link do to
incompatibilities in libstdc++.  I'm going to try Boost.Python from
CVS before anything else, but I thought I'd ask if anyone else has had
success or failure in this area.

Thanks,
Chad



From paul.bridger at paradise.net.nz  Thu Jun 24 00:31:50 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Thu, 24 Jun 2004 10:31:50 +1200
Subject: [C++-sig] Pyste: wrapper function for overloaded function problems
In-Reply-To: <40D08720.8080201@esss.com.br>
References: <40CE3E20.1020409@paradise.net.nz> <40D08720.8080201@esss.com.br>
Message-ID: <40DA04D6.9050800@paradise.net.nz>

Hi,

I've just done a diff of my Pyste vs. CVS. There are a few small improvements 
I've made.

Two in ClassExporter.py:
The problem described below was easy to fix:
ClassExporter.py: 375
             if method.minArgs != method.maxArgs:
# here pyste adds overload declarations and creates extra argument to def()
becomes
             if method.minArgs != method.maxArgs and not method_info.wrapper:
# so we nolonger do this if the method has a wrapper

This worked fine for me, but I am nolonger using (and hence testing) this 
change every day. I was writing function/method wrappers for functions that 
took a customised string implementation, but I stopped doing this when I 
found I couldn't write constructor wrappers (AFAIK). Now everything is 
consistently unwrapped.

A diff snippet for that:
375c381
<             if method.minArgs != method.maxArgs:
---
 >             if method.minArgs != method.maxArgs and not method_info.wrapper:


I also noticed that for inherited virtual methods, Pyste would not:
a) warn about lack of defined policies, and
b) register the default policy when it could

Here is the fix:

778,779c787,792
<         # Add policy to overloaded methods also
<         policy = self.info[method.name].policy or ''
---
 >         # warn the user if this method needs a policy and doesn't have one
 >         method_info = self.info[method.name]
 >         method_info.policy = exporterutils.HandlePolicy(method, 
method_info.policy)
 >
 >         # Add policy to overloaded methods also
 >         policy = method_info.policy or ''


There was also a change in GCCXMLParser.py. I posted to C++-Sig about this a 
little while ago:
http://aspn.activestate.com/ASPN/Mail/Message/c++-sig/2088336

GCCXMLParser.py
280c280
<             if type(decl) in Class.ValidMemberTypes():
---
 >             if type(decl) in Class.ValidMemberTypes() and decl.name != '':


Thanks.
Paul Bridger


Nicodemus wrote:
> Hi Paul, sorry about the delay,
> 
> paul.bridger wrote:
> 
>> I am trying to expose a C++ member function with a bunch of default 
>> arguments, but I also want to provide a C++ wrapper for it.
>> The problem I've had is that Pyste still generates the overload 
>> machinery even though it's not wanted since my wrapper function takes 
>> different arguments.
>>
>> Any hints? I'm going digging in the Pyste source.
> 
> 
> 
> You want to take a look at the method ExportMethods in ClassExporter.py. ;)
> Patches are welcome, as always!
> 
> Regards,
> Nicodemus.
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 
> 










From caustin at gmail.com  Thu Jun 24 02:26:09 2004
From: caustin at gmail.com (Chad Austin)
Date: Wed, 23 Jun 2004 19:26:09 -0500
Subject: [C++-sig] Re: Making Boost.Python understand DerivedSafe
In-Reply-To: <uvfhn3j8m.fsf@boost-consulting.com>
References: <2096514e0406170112773a0e76@mail.gmail.com>
	<uvfhn3j8m.fsf@boost-consulting.com>
Message-ID: <2096514e0406231726642a11a2@mail.gmail.com>

On Sat, 19 Jun 2004 07:00:41 -0400, David Abrahams
<dave at boost-consulting.com> wrote:
> 2. As far as I can tell you have used boost::ptr around one of your
>    RefPtr<CutScene>s, call it 'x':
> 
>          ptr(x)
> 
>    boost::ptr is reserved for real pointers.

Ah, I didn't know that.  Taking out the ptr call seems to work.  Now I
understand how HeldType works.  (Thanks Nikolay.)

> > If I pass the result of CutScenePtr::get() into operator() I get a
> > runtime error about Boost.Python not knowing how to convert
> > RefDerivedSafe<T>* to a Python object.  This makes sense, because
> > RefPtr's implicit conversion operator, operator->, and get() all
> > return RefDerivedSafe<T>*. (RefDerivedSafe<T> is a special class
> > derived from T that prevents C++ clients from accidentally calling
> > delete, ref(), or unref() on a smart pointer
> 
> Bah!
> 
>         T* p = x.get();
>         delete p;

Of course you can work around it.  :P  The point is, it does help
catch unintentional bugs, especially in code written by someone new to
refcounting/smart pointers or when refactoring.

> > . This techique is used in both ATL and nsCOMPtr in Mozilla).
> 
> That doesn't make it safe ;->

We can go back and forth for eternity arguing about C++ style, but
that doesn't get anywhere.  I really just wanted some help in getting
this existing codebase connected nicely to Python.

> > Here's my question:  Is there a way to tell Boost.Python that
> > ptr(someCutScenePtr) is valid
> 
> No!  Why do you want to use ptr?

I was under the impression that if you don't use ptr, a copy of the
object is given to Python.  From the discussion at:
http://www.boost.org/libs/python/doc/v2/callbacks.html  I guess it
just wasn't exactly clear how I should pass my own refcounting pointer
into Boost.Python and how that fit in with HeldType.

Chad



From Peter.Vandersteegen at intec.Ugent.be  Thu Jun 24 09:43:56 2004
From: Peter.Vandersteegen at intec.Ugent.be (Peter Vandersteegen)
Date: Thu, 24 Jun 2004 09:43:56 +0200
Subject: [C++-sig] Question: compiling library using <python/numeric.cpp>
 extension/wrapper 
Message-ID: <6.1.0.6.0.20040624094350.025c1630@pop-tech.intec.UGent.be>

Hello,

Sorry if you recieve this more than once ...

I've recently discovered boost-python.   Compiling several examples gave no 
problem.
Working with lists etc also no problems.
Compiling a simple library using the <python/numeric.cpp> 
extension/wrapper  however
resulted in several questions...

Therefor I've one question:

Question1:
Is it correct to state the following?   (something I find confusing)
There are (at least) 2 ways to get a boost-layer between Numeric python and 
c++.  The first one is based on
<Numeric/arrayobject.h>, the other on <python/Numeric.hpp>...

* The project num_util 
(http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/) uses 
the fist method?  They use boost as wrapper, but for the array class, they 
directly use the python numeric array class and not the boost.python array 
class?
* Directly using <python/Numeric.hpp> seperates you from dealing with the 
'raw' python numeric array class?  Added bonus: no need to write 
import_array() in BOOST_PYTHON_MODULE...

Thus, is it correct to say that these are 2 different approaches to convert 
an array from python to c++ ?

thx a lot for reading so far

Peter






From rwgk at yahoo.com  Thu Jun 24 16:42:41 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Thu, 24 Jun 2004 07:42:41 -0700 (PDT)
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <2096514e0406221706691db56c@mail.gmail.com>
Message-ID: <20040624144241.53669.qmail@web20207.mail.yahoo.com>

--- Chad Austin <caustin at gmail.com> wrote:
> Has anyone successfully used Boost.Python in SUSE 9.0 or SUSE 9.1?  In
> 9.0 I had to build and install boost 1.31 from source, but 9.1 comes
> with a boost 1.31 package.  Either way, merely #including
> <boost/python.hpp> causes an internal compiler error in a variety of
> places.  Note: SUSE 9.0 uses gcc 3.3.1 (+ SUSE-specific patches?) and
> SUSE 9.1 uses gcc 3.3.3 (+ patches again?).

I've used gcc 3.3.1 under Mandrake 9.2 and SunOS 5.9 without problems, but
didn't have the opportunity to try SUSE.

> If I build gcc 3.4.0 from
> source, Boost.Python compiles, but does not link do to
> incompatibilities in libstdc++.>

Does the link command fail, or is it a run-time failure (i.e. dlopen fails)?
If it is the latter you might have to update the LD_LIBRARY_PATH. This is what
I am using to activate gcc 3.4 in a new shell:

set path=(/usr/local_cci/gcc-3.4.0/bin $path)
if ( $?LD_LIBRARY_PATH ) then
  setenv LD_LIBRARY_PATH "/usr/local_cci/gcc-3.4.0/lib:${LD_LIBRARY_PATH}"
else
  setenv LD_LIBRARY_PATH "/usr/local_cci/gcc-3.4.0/lib"
endif

> I'm going to try Boost.Python from
> CVS before anything else, but I thought I'd ask if anyone else has had
> success or failure in this area.

The Boost 1.31 release has a subtle issue when compiling with gcc 3.4:

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14528

If you run into this use Boost CVS. As of yesterday it was in fantastic shape.

Ralf



		
__________________________________
Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
http://promotions.yahoo.com/new_mail



From paustin at eos.ubc.ca  Thu Jun 24 18:53:36 2004
From: paustin at eos.ubc.ca (Philip Austin)
Date: Thu, 24 Jun 2004 09:53:36 -0700
Subject: [C++-sig] Question: compiling library using <python/numeric.cpp>
In-Reply-To: <6.1.0.6.0.20040622141832.02667098@pop-tech.intec.UGent.be>
References: <6.1.0.6.0.20040622141832.02667098@pop-tech.intec.UGent.be>
Message-ID: <16603.1808.886523.566258@gull.eos.ubc.ca>

Peter Vandersteegen writes:
 > Hello,
 > 
 > Therefor I've one question:

[snip]

 > 
 > * The project num_util 
 > (http://www.eos.ubc.ca/research/clouds/software/pythonlibs/num_util/) uses 
 > the fist method?  They use boost as wrapper, but for the array class, they 
 > directly use the python numeric array class and not the boost.python array 
 > class?
 > * Directly using <python/Numeric.hpp> seperates you from dealing with the 
 > 'raw' python numeric array class?  Added bonus: no need to write 
 > import_array() in BOOST_PYTHON_MODULE...

That's right  -- also, by using callbacks into python, David was able
to make the same implementation work for both Numeric and Numarray
(the successor to Numeric, which is not completely backward
compatible).  We use num_util because (in decreasing order of importance):
i) we need to move large arrays between python and C++, which means
we need access to the PyArrayObject pointer.  Numeric.hpp only
supports copying your results into a new tuple or list to return
to python.  ii) the callback overhead can be substantial for some
of our applications.  iii) we have a quite a bit of legacy code developed
for our interface under boost v1, which predates Numeric.hpp.

Regards, Phil







From nickm at sitius.com  Thu Jun 24 23:57:07 2004
From: nickm at sitius.com (Nikolay Mladenov)
Date: Thu, 24 Jun 2004 17:57:07 -0400
Subject: [C++-sig] Re: opencascade
References: <002b01c457c5$1378e380$73c809c0@CPQ30541773923>
Message-ID: <40DB4E33.345F3EBD@sitius.com>

I am also interested in the answer of that one.

-- 
Nikolay Mladenov
Sitius Automation Inc.
www.sitius.com




From hoeppler at diener.iap.physik.uni-tuebingen.de  Fri Jun 25 11:29:31 2004
From: hoeppler at diener.iap.physik.uni-tuebingen.de (Chris Hoeppler)
Date: Fri, 25 Jun 2004 11:29:31 +0200 (CEST)
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <2096514e0406221706691db56c@mail.gmail.com>
Message-ID: <Pine.LNX.4.33.0406251047390.19144-100000@diener.iap.physik.uni-tuebingen.de>

On Tue, 22 Jun 2004, Chad Austin wrote:

> Hi all,
>
> Has anyone successfully used Boost.Python in SUSE 9.0 or SUSE 9.1?  In
> 9.0 I had to build and install boost 1.31 from source, but 9.1 comes
> with a boost 1.31 package.  Either way, merely #including
> <boost/python.hpp> causes an internal compiler error in a variety of
> places.  Note: SUSE 9.0 uses gcc 3.3.1 (+ SUSE-specific patches?) and
> SUSE 9.1 uses gcc 3.3.3 (+ patches again?).  If I build gcc 3.4.0 from
> source, Boost.Python compiles, but does not link do to
> incompatibilities in libstdc++.  I'm going to try Boost.Python from
> CVS before anything else, but I thought I'd ask if anyone else has had
> success or failure in this area.
>
> Thanks,
> Chad

Hi,

I've managed to successfully compile and use a small module on SuSE 9.0
with the following setup:

* gcc version 3.3.1 (SuSE Linux)

* boost_1_31_0 obtained from boost site and compiled from sources

* python 2.3.3 also compiled from sources.

* Module:

#include <boost/python/suite/indexing/vector_indexing_suite.hpp>
#include <boost/python/module.hpp>
#include <boost/python/def.hpp>
#include <boost/python/implicit.hpp>

#include "rpower.h"

using namespace boost::python;


BOOST_PYTHON_MODULE(fftw_ext)
{
    class_<std::vector<fftw_real> >("fftw_realVec")
        .def(vector_indexing_suite<std::vector<fftw_real> >())
    ;

    def("power_spectrum", power_spectrum);
}

where power_spectrum is a function defined in rpower.h with the following
signature:

void power_spectrum(std::vector<fftw_real> &in,
        std::vector<fftw_real> &out);

However, I've just tried to compile a test file with just one line
#include <boost/python.hpp>
as mentioned in your post quoted above, which also results in an internal
compiler error:

g++ -Wall -ftemplate-depth-100 -DBOOST_PYTHON_DYNAMIC_LIB -g -O0 \
-fno-inline -fPIC -DDEBUG -I/usr/local/include/python2.3 \
-I/usr/local/include/boost-1_31 -c -o test.os test.cpp

<snipped warning about _POSIX_C_SOURCE being redefined>
/usr/local/include/boost-1_31/boost/python/object_core.hpp:283: internal
compiler error: Segmentation
   fault

Any clues on how to proceed from here?

Thanks,
Chris




From MKhesin at liquidnet.com  Fri Jun 25 15:54:19 2004
From: MKhesin at liquidnet.com (Max Khesin)
Date: Fri, 25 Jun 2004 09:54:19 -0400
Subject: [C++-sig] boost.python crush
Message-ID: <4B6FB7F60D37D41188C300B0D022E0C0029F4390@exchange.lnholdings.com>

I have the following sample class exposed via boost.python:

class String
{
public:
    String(const std::string& v):m_val(v){}
    bool lessThan(const std::string& other)                  const{return
this->m_val<other;}
    bool greaterThan(const std::string& other)               const{return
this->m_val>other;}
    bool equals(const std::string& other)                    const{return
this->m_val==other;}
    bool regexMatch(const std::string& other)                const{return
this->m_val==other;}
    bool withinDistance(const std::string& other, int dist)  const{return
distance(this->m_val, other)<=dist;}       
private:
    std::string m_val;
};


BOOST_PYTHON_MODULE(test)
{
    class_<String>("String", init<const std::string&>())
        .def("lessThan",        &String::lessThan)
        .def("greaterThan",     &String::greaterThan)
        .def("equals",          &String::equals)
        .def("regexMatch",      &String::regexMatch)
        .def("withinDistance",  &String::withinDistance)
        ;
}

I am calling this from python:

>> import test
>> s = test.String('hallo')
>> s.lessThen('hello')   #### <<< CRUSH

the debugger seems to point to an access violation in std::string dtor. I
palyed around with making the argument a plain std::string (instead of const
std::string&) and const char* - both of those crush, albeit in a different
place (also dtors, from the looks of it)

Any idea?

thanks,

max
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.



From rwgk at yahoo.com  Fri Jun 25 15:55:25 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Fri, 25 Jun 2004 06:55:25 -0700 (PDT)
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <Pine.LNX.4.33.0406251047390.19144-100000@diener.iap.physik.uni-tuebingen.de>
Message-ID: <20040625135525.94532.qmail@web20205.mail.yahoo.com>

--- Chris Hoeppler <hoeppler at diener.iap.physik.uni-tuebingen.de> wrote:
> However, I've just tried to compile a test file with just one line
> #include <boost/python.hpp>
> as mentioned in your post quoted above, which also results in an internal
> compiler error:
> 
> g++ -Wall -ftemplate-depth-100 -DBOOST_PYTHON_DYNAMIC_LIB -g -O0 \
> -fno-inline -fPIC -DDEBUG -I/usr/local/include/python2.3 \
> -I/usr/local/include/boost-1_31 -c -o test.os test.cpp
> 
> <snipped warning about _POSIX_C_SOURCE being redefined>
> /usr/local/include/boost-1_31/boost/python/object_core.hpp:283: internal
> compiler error: Segmentation
>    fault
> 
> Any clues on how to proceed from here?

You have to strip down the code until you've isolated the construct that causes
the ICE. Then try to rearrange that construct until the ICE goes away. In the
early days of Boost.Python I've done this a million times, and I am sure David
did it a billion times. It always takes too long, is very frustrating, and
therefore requires a lot of determination. If this is not enough to scare you,
copy the content of boost/python.hpp into your source file, then comment out
half the includes, see if it still ICE's, if yes comment out half of the
remaining, if not put back half of the ones you commented out, and so on until
you've found that one include that makes a difference. Then copy that into your
source code and repeat the exercise until you found the piece of code that
makes the difference. Then go into a different trial-and-error loop rearranging
the code. While the first step requires mainly determination, you will need
some creativity here. Often my last resort was to ask David. Finally you have
to communicate the fix to the developers of the effected library (e.g. it could
be in boost::graph or boost::bind or, heaven forbid, in boost::mpl or ...) and
convince them to adopt your fix.

Easy!

Ralf



		
__________________________________
Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
http://promotions.yahoo.com/new_mail



From dave at boost-consulting.com  Fri Jun 25 16:04:20 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Fri, 25 Jun 2004 10:04:20 -0400
Subject: [C++-sig] Re: Making Boost.Python understand DerivedSafe
References: <2096514e0406170112773a0e76@mail.gmail.com>
	<uvfhn3j8m.fsf@boost-consulting.com>
	<2096514e0406231726642a11a2@mail.gmail.com>
Message-ID: <ud63nvimz.fsf@boost-consulting.com>

Chad Austin <caustin at gmail.com> writes:

>> 
>> Bah!
>> 
>>         T* p = x.get();
>>         delete p;
>
> Of course you can work around it.  :P  The point is, it does help
> catch unintentional bugs, especially in code written by someone new to
> refcounting/smart pointers or when refactoring.

Just to be clear, I have no problem with that part; it's having the
implicit conversion at all that I take issue with... but I see below
you don't really care what I think about this ;-> (I can't blame you).

>> > . This techique is used in both ATL and nsCOMPtr in Mozilla).
>> 
>> That doesn't make it safe ;->
>
> We can go back and forth for eternity arguing about C++ style, but
> that doesn't get anywhere.  I really just wanted some help in getting
> this existing codebase connected nicely to Python.

I hope you got the help you needed.

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




From caustin at gmail.com  Fri Jun 25 16:07:03 2004
From: caustin at gmail.com (Chad Austin)
Date: Fri, 25 Jun 2004 09:07:03 -0500
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <20040624144241.53669.qmail@web20207.mail.yahoo.com>
References: <20040624144241.53669.qmail@web20207.mail.yahoo.com>
Message-ID: <2096514e04062507072f0ff45@mail.gmail.com>

Hi Ralf,

On Thu, 24 Jun 2004 07:42:41 -0700 (PDT), Ralf W. Grosse-Kunstleve
<rwgk at yahoo.com> wrote:
> 
> --- Chad Austin <caustin at gmail.com> wrote:
> > Has anyone successfully used Boost.Python in SUSE 9.0 or SUSE 9.1?  In
> > 9.0 I had to build and install boost 1.31 from source, but 9.1 comes
> > with a boost 1.31 package.  Either way, merely #including
> > <boost/python.hpp> causes an internal compiler error in a variety of
> > places.  Note: SUSE 9.0 uses gcc 3.3.1 (+ SUSE-specific patches?) and
> > SUSE 9.1 uses gcc 3.3.3 (+ patches again?).
> 
> I've used gcc 3.3.1 under Mandrake 9.2 and SunOS 5.9 without problems, but
> didn't have the opportunity to try SUSE.

After some investigation, it appears SUSE applies some custom patches
to their gcc builds, causing including boost/python.hpp to ICE.  I've
submitted a support request to SUSE for this issue.  Installing a
compatible gcc from source into /usr/local appears to work around the
problem.

> > If I build gcc 3.4.0 from
> > source, Boost.Python compiles, but does not link do to
> > incompatibilities in libstdc++.>
> 
> Does the link command fail, or is it a run-time failure (i.e. dlopen fails)?
> If it is the latter you might have to update the LD_LIBRARY_PATH. This is what
> I am using to activate gcc 3.4 in a new shell:

I believe it was an ABI change in libstdc++ or something.  My boost
was compiled with gcc 3.3 and my app with 3.4.  I just started getting
random crashes.

> set path=(/usr/local_cci/gcc-3.4.0/bin $path)
> if ( $?LD_LIBRARY_PATH ) then
>   setenv LD_LIBRARY_PATH "/usr/local_cci/gcc-3.4.0/lib:${LD_LIBRARY_PATH}"
> else
>   setenv LD_LIBRARY_PATH "/usr/local_cci/gcc-3.4.0/lib"
> endif

Awesome, thanks.

> > I'm going to try Boost.Python from
> > CVS before anything else, but I thought I'd ask if anyone else has had
> > success or failure in this area.
> 
> The Boost 1.31 release has a subtle issue when compiling with gcc 3.4:
> 
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14528
> 
> If you run into this use Boost CVS. As of yesterday it was in fantastic shape.

Thanks for the heads up!  I did notice that gcc 3.4 is a lot more
strict about some things.  It seems to discover lots of subtle errors
that other compilers don't.  :)

Chad



From caustin at gmail.com  Fri Jun 25 16:10:44 2004
From: caustin at gmail.com (Chad Austin)
Date: Fri, 25 Jun 2004 09:10:44 -0500
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <Pine.LNX.4.33.0406251047390.19144-100000@diener.iap.physik.uni-tuebingen.de>
References: <Pine.LNX.4.33.0406251047390.19144-100000@diener.iap.physik.uni-tuebingen.de>
Message-ID: <2096514e040625071078b1a2f1@mail.gmail.com>

Hi Chris,

Thanks for letting me know how you've gotten it to work.  I'll try
including the individual Boost.Python headers instead of
<boost/python.hpp>.  As I mentioned in my response to Ralf, if you
compile a compatible gcc (3.3.1 for SUSE 9.0) and install it into
/usr/local, including <boost/python.hpp> works.

Chad

On Fri, 25 Jun 2004 11:29:31 +0200 (CEST), Chris Hoeppler
<hoeppler at diener.iap.physik.uni-tuebingen.de> wrote:
> 
> 
> On Tue, 22 Jun 2004, Chad Austin wrote:
> 
> > Hi all,
> >
> > Has anyone successfully used Boost.Python in SUSE 9.0 or SUSE 9.1?  In
> > 9.0 I had to build and install boost 1.31 from source, but 9.1 comes
> > with a boost 1.31 package.  Either way, merely #including
> > <boost/python.hpp> causes an internal compiler error in a variety of
> > places.  Note: SUSE 9.0 uses gcc 3.3.1 (+ SUSE-specific patches?) and
> > SUSE 9.1 uses gcc 3.3.3 (+ patches again?).  If I build gcc 3.4.0 from
> > source, Boost.Python compiles, but does not link do to
> > incompatibilities in libstdc++.  I'm going to try Boost.Python from
> > CVS before anything else, but I thought I'd ask if anyone else has had
> > success or failure in this area.
> >
> > Thanks,
> > Chad
> 
> Hi,
> 
> I've managed to successfully compile and use a small module on SuSE 9.0
> with the following setup:
> 
> * gcc version 3.3.1 (SuSE Linux)
> 
> * boost_1_31_0 obtained from boost site and compiled from sources
> 
> * python 2.3.3 also compiled from sources.
> 
> * Module:
> 
> #include <boost/python/suite/indexing/vector_indexing_suite.hpp>
> #include <boost/python/module.hpp>
> #include <boost/python/def.hpp>
> #include <boost/python/implicit.hpp>
> 
> #include "rpower.h"
> 
> using namespace boost::python;
> 
> BOOST_PYTHON_MODULE(fftw_ext)
> {
>     class_<std::vector<fftw_real> >("fftw_realVec")
>         .def(vector_indexing_suite<std::vector<fftw_real> >())
>     ;
> 
>     def("power_spectrum", power_spectrum);
> }
> 
> where power_spectrum is a function defined in rpower.h with the following
> signature:
> 
> void power_spectrum(std::vector<fftw_real> &in,
>         std::vector<fftw_real> &out);
> 
> However, I've just tried to compile a test file with just one line
> #include <boost/python.hpp>
> as mentioned in your post quoted above, which also results in an internal
> compiler error:
> 
> g++ -Wall -ftemplate-depth-100 -DBOOST_PYTHON_DYNAMIC_LIB -g -O0 \
> -fno-inline -fPIC -DDEBUG -I/usr/local/include/python2.3 \
> -I/usr/local/include/boost-1_31 -c -o test.os test.cpp
> 
> <snipped warning about _POSIX_C_SOURCE being redefined>
> /usr/local/include/boost-1_31/boost/python/object_core.hpp:283: internal
> compiler error: Segmentation
>    fault
> 
> Any clues on how to proceed from here?
> 
> Thanks,
> Chris
> 
> 
> 
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>



From rwgk at yahoo.com  Fri Jun 25 16:11:43 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Fri, 25 Jun 2004 07:11:43 -0700 (PDT)
Subject: [C++-sig] boost.python crush
In-Reply-To: <4B6FB7F60D37D41188C300B0D022E0C0029F4390@exchange.lnholdings.com>
Message-ID: <20040625141143.96422.qmail@web20224.mail.yahoo.com>

Is this under Windows? If yes you have to be very careful how you link. Certain
combinations of DLL's lead to obscure failures like yours. There was a thread
about this recently; should be in the archieve:

http://mail.python.org/pipermail/c++-sig/

Look for "windows", "dll", "debug".
If you are using a different platform post the exact details.
Ralf

--- Max Khesin <MKhesin at liquidnet.com> wrote:
> I have the following sample class exposed via boost.python:
> 
> class String
> {
> public:
>     String(const std::string& v):m_val(v){}
>     bool lessThan(const std::string& other)                  const{return
> this->m_val<other;}
>     bool greaterThan(const std::string& other)               const{return
> this->m_val>other;}
>     bool equals(const std::string& other)                    const{return
> this->m_val==other;}
>     bool regexMatch(const std::string& other)                const{return
> this->m_val==other;}
>     bool withinDistance(const std::string& other, int dist)  const{return
> distance(this->m_val, other)<=dist;}       
> private:
>     std::string m_val;
> };
> 
> 
> BOOST_PYTHON_MODULE(test)
> {
>     class_<String>("String", init<const std::string&>())
>         .def("lessThan",        &String::lessThan)
>         .def("greaterThan",     &String::greaterThan)
>         .def("equals",          &String::equals)
>         .def("regexMatch",      &String::regexMatch)
>         .def("withinDistance",  &String::withinDistance)
>         ;
> }
> 
> I am calling this from python:
> 
> >> import test
> >> s = test.String('hallo')
> >> s.lessThen('hello')   #### <<< CRUSH
> 
> the debugger seems to point to an access violation in std::string dtor. I
> palyed around with making the argument a plain std::string (instead of const
> std::string&) and const char* - both of those crush, albeit in a different
> place (also dtors, from the looks of it)
> 
> Any idea?
> 
> thanks,
> 
> max
> This email and any files transmitted with it are confidential and intended
> solely for the use of the individual or entity to whom they are addressed. If
> you have received this email in error please notify the system manager. This
> message contains confidential information and is intended only for the
> individual named. If you are not the named addressee you should not
> disseminate, distribute or copy this e-mail.
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 



		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail 



From MKhesin at liquidnet.com  Fri Jun 25 16:49:29 2004
From: MKhesin at liquidnet.com (Max Khesin)
Date: Fri, 25 Jun 2004 10:49:29 -0400
Subject: [C++-sig] boost.python crush
Message-ID: <4B6FB7F60D37D41188C300B0D022E0C0029F4392@exchange.lnholdings.com>

Thanks - my searching eventually yielded this:
http://www.boost.org/libs/python/doc/building.html#variants
with the appropriate VC6 instructions at the bottom. This took care of it.
m.

> -----Original Message-----
> From: Ralf W. Grosse-Kunstleve [mailto:rwgk at yahoo.com]
> Sent: Friday, June 25, 2004 10:12 AM
> To: Development of Python/C++ integration
> Subject: Re: [C++-sig] boost.python crush
> 
> 
> Is this under Windows? If yes you have to be very careful how 
> you link. Certain
> combinations of DLL's lead to obscure failures like yours. 
> There was a thread
> about this recently; should be in the archieve:
> 
> http://mail.python.org/pipermail/c++-sig/
> 
> Look for "windows", "dll", "debug".
> If you are using a different platform post the exact details.
> Ralf
> 
> --- Max Khesin <MKhesin at liquidnet.com> wrote:
> > I have the following sample class exposed via boost.python:
> > 
> > class String
> > {
> > public:
> >     String(const std::string& v):m_val(v){}
> >     bool lessThan(const std::string& other)                 
>  const{return
> > this->m_val<other;}
> >     bool greaterThan(const std::string& other)              
>  const{return
> > this->m_val>other;}
> >     bool equals(const std::string& other)                   
>  const{return
> > this->m_val==other;}
> >     bool regexMatch(const std::string& other)               
>  const{return
> > this->m_val==other;}
> >     bool withinDistance(const std::string& other, int dist) 
>  const{return
> > distance(this->m_val, other)<=dist;}       
> > private:
> >     std::string m_val;
> > };
> > 
> > 
> > BOOST_PYTHON_MODULE(test)
> > {
> >     class_<String>("String", init<const std::string&>())
> >         .def("lessThan",        &String::lessThan)
> >         .def("greaterThan",     &String::greaterThan)
> >         .def("equals",          &String::equals)
> >         .def("regexMatch",      &String::regexMatch)
> >         .def("withinDistance",  &String::withinDistance)
> >         ;
> > }
> > 
> > I am calling this from python:
> > 
> > >> import test
> > >> s = test.String('hallo')
> > >> s.lessThen('hello')   #### <<< CRUSH
> > 
> > the debugger seems to point to an access violation in 
> std::string dtor. I
> > palyed around with making the argument a plain std::string 
> (instead of const
> > std::string&) and const char* - both of those crush, albeit 
> in a different
> > place (also dtors, from the looks of it)
> > 
> > Any idea?
> > 
> > thanks,
> > 
> > max
> > This email and any files transmitted with it are 
> confidential and intended
> > solely for the use of the individual or entity to whom they 
> are addressed. If
> > you have received this email in error please notify the 
> system manager. This
> > message contains confidential information and is intended 
> only for the
> > individual named. If you are not the named addressee you should not
> > disseminate, distribute or copy this e-mail.
> > 
> > _______________________________________________
> > C++-sig mailing list
> > C++-sig at python.org
> > http://mail.python.org/mailman/listinfo/c++-sig
> > 
> 
> 
> 
> 		
> __________________________________
> Do you Yahoo!?
> New and Improved Yahoo! Mail - Send 10MB messages!
> http://promotions.yahoo.com/new_mail 
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
> 
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.



From apri7 at mail.ru  Sat Jun 26 15:27:07 2004
From: apri7 at mail.ru (Sasha Prikhod'ko)
Date: Sat, 26 Jun 2004 17:27:07 +0400
Subject: [C++-sig] Boost:   multithreading initiated by a module
Message-ID: <E1BeDD1-000MoB-00.apri7-mail-ru@f19.mail.ru>

Dear Python and Boost developers!
I encountered a problem with threading in Visual Studio build.
Thez a C++ project having in base
some special Python site-library.
A module start a thread, while a function is called from C++ project.
But thread strangely dies at once!

What is the reason?   --- wrong build, wrong integration, wrong call...

Moreover, I cannot certainly recocnize this trouble as serious
and tie with plenty of threading discussions... :)

Sincerely,
Alex




From dleary at gmail.com  Sat Jun 26 18:59:53 2004
From: dleary at gmail.com (Dusty Leary)
Date: Sat, 26 Jun 2004 12:59:53 -0400
Subject: [C++-sig] implicitly_convertible and char*
Message-ID: <81c509c04062609594758fdc4@mail.gmail.com>

Hello,

The implicitly_convertible sample at
http://boost.org/libs/python/doc/v2/implicit.html#implicitly_convertible-spec
works fine for me.

However, changing "int" to "char*" makes it stop working.  Compiles
fine, won't invoke at runtime.

>From the traceback:

Boost.Python.ArgumentError: Python argument types in
    __testmodule__.test_string(X)
did not match C++ signature:
    test_string(char *)

I am using MSVC7.1

Any info?



From wstwej03 at sneakemail.com  Sun Jun 27 05:42:48 2004
From: wstwej03 at sneakemail.com (Eric)
Date: Sat, 26 Jun 2004 20:42:48 -0700
Subject: [C++-sig] Help with shared_ptr
In-Reply-To: <40CE3E20.1020409@paradise.net.nz>
Message-ID: <22226-95764@sneakemail.com>

Hello all,

I've managed to get myself stuck again, and could use some pointers in how
to debug my way out of this situation. I have to kinds of objects that are
wrapped in a boost::shared_ptr and returned to python. They are both
registered with python like this:

    boost::python::register_ptr_to_python< boost::shared_ptr<ClassOne> >();
    boost::python::register_ptr_to_python< boost::shared_ptr<ClassTwo> >();

the strange thing is that the release function of ClassOne is called when
python loses all references to it, but ClassTwo's is never called. From
within python I have managed to ensure that ClassTwo is actually being
released, by using weakref as follows:

# win is my ClassTwo object
wr = weakref.ref(win)
assert wr()
del win
sys.exc_clear()
gc.collect()
assert not wr()


So now I am trying to debug this situation. Any help or suggestions are
greatly appreciated.

Thank you,

Eric




From faheem at email.unc.edu  Sun Jun 27 08:17:51 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Sun, 27 Jun 2004 06:17:51 +0000 (UTC)
Subject: [C++-sig] calling python function from C++ using call
Message-ID: <slrncdspkf.om1.faheem@Chrestomanci.home.earth>

Hi,

I'm trying to call the function `array' (from the python module
numarray.strings) from inside C++ code. This function initializes a
character array. It has the arguments

array(buffer=None, itemsize=None, shape=None, byteoffset=0,
bytestride=None, kind=CharArray)

I'm doing this as follows:

The C++ end looks like

****************************************************************
bcallback.cc
****************************************************************
...
object charr(PyObject* func, list buf, object itemsize, tuple sh)
{
  return boost::python::call<object>(func, buf, itemsize, sh);
}
...
****************************************************************

and the Python end looks like

****************************************************************
callback.py
****************************************************************
from bcallback import charr
import numarray.strings as numstrings

def retcharr(buffer, sh):
    return charr(numstrings.array, buffer, None, sh)
****************************************************************

This actually works. Namely:

In [2]: import callback

In [3]: callback.retcharr(['a', 'c', 'a', 'c', 'a', 'c', 'g', 't',
'g','t','g','t'], (6,2))
Out[3]:
CharArray([['a', 'c'],
           ['a', 'c'],
           ['a', 'c'],
           ['g', 't'],
           ['g', 't'],
           ['g', 't']])

However, I'm wondering if this is the best way of proceeding, or if
there is a better way.

Thanks in advance for comments.

                                                               Faheem.




From domma at procoders.net  Sun Jun 27 18:00:56 2004
From: domma at procoders.net (Achim Domma (Procoders))
Date: Sun, 27 Jun 2004 18:00:56 +0200
Subject: [C++-sig] Getter/Setter as Property with Pyste
Message-ID: <BPEBKLCAPNMHJOAHJOIHOELJCBAA.domma@procoders.net>

Hi,

I try to generate Code with Pyste. The wrapped classes have lots of getter
and setter functions. The generated code looks like this:

.def("borderColor", (void (Magick::Image::*)(const
Magick::Color&) )&Magick::Image::borderColor)
.def("borderColor", (Magick::Color (Magick::Image::*)()
const)&Magick::Image::borderColor)

Is there a simple way to tell Pyste to generate an add_property call for
'borderColor' instead of the two functions?

regards,
Achim




From achim.domma at syynx.de  Sun Jun 27 18:02:37 2004
From: achim.domma at syynx.de (Achim Domma)
Date: Sun, 27 Jun 2004 18:02:37 +0200
Subject: [C++-sig] Getter/Setter as Property with Pyste
Message-ID: <BPEBKLCAPNMHJOAHJOIHIELKCBAA.achim.domma@syynx.de>

Hi,

I try to generate Code with Pyste. The wrapped classes have lots of getter
and setter functions. The generated code looks like this:

.def("borderColor", (void (Magick::Image::*)(const
Magick::Color&) )&Magick::Image::borderColor)
.def("borderColor", (Magick::Color (Magick::Image::*)()
const)&Magick::Image::borderColor)

Is there a simple way to tell Pyste to generate an add_property call for
'borderColor' instead of the two functions?

regards,
Achim




From faheem at email.unc.edu  Sun Jun 27 20:58:48 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Sun, 27 Jun 2004 18:58:48 +0000 (UTC)
Subject: [C++-sig] Re: calling python function from C++ using call
References: <slrncdspkf.om1.faheem@Chrestomanci.home.earth>
Message-ID: <slrncdu678.om1.faheem@Chrestomanci.home.earth>

On Sun, 27 Jun 2004 06:17:51 +0000 (UTC), Faheem Mitha
<faheem at email.unc.edu> wrote:

> Hi,
>
> I'm trying to call the function `array' (from the python module
> numarray.strings) from inside C++ code. This function initializes a
> character array. It has the arguments

Actually, I do now have a question. I realised that it would be better
if the C++ code could call the array function directly. This would
involve embedding the interpreter, right? I'd appreciate a
confirmation of this.

I.e. I'd need to do something equivalent to 
 
from numarray.strings import array

in the C++ code.

I'll try to figure out how to do this from the tutorial and the
reference manual, but it would be much appreciated if som kind person
would just tell me what the syntax should look like for importing. :-)
Pretty please?

                                                          Faheem.




From dave at boost-consulting.com  Mon Jun 28 02:39:56 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Sun, 27 Jun 2004 20:39:56 -0400
Subject: [C++-sig] Re: implicitly_convertible and char*
References: <81c509c04062609594758fdc4@mail.gmail.com>
Message-ID: <un02oh5wj.fsf@boost-consulting.com>

Dusty Leary <dleary at gmail.com> writes:

> The implicitly_convertible sample at
> http://boost.org/libs/python/doc/v2/implicit.html#implicitly_convertible-spec
> works fine for me.
>
> However, changing "int" to "char*" makes it stop working.  Compiles
> fine, won't invoke at runtime.
>
>>From the traceback:
>
> Boost.Python.ArgumentError: Python argument types in
>     __testmodule__.test_string(X)
> did not match C++ signature:
>     test_string(char *)
>
> I am using MSVC7.1
>
> Any info?

Full source code of your example?
Perhaps the problem is that X isn't actually convertible to char*?

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




From faheem at email.unc.edu  Mon Jun 28 06:26:19 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Mon, 28 Jun 2004 04:26:19 +0000 (UTC)
Subject: [C++-sig] compile errors with embedding example from tutorial
Message-ID: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>

Dear People,

When trying to compile the following code (direct from sample code in
the turorial) I get these errors. I am using distutils and realise my
flags may not be exactly right, but the compiler seems to be saying it
cannot find the correct version of handler to use! This does not look
like it has anything to do with the correct flags, though I could be
wrong.

I'm trying to call the interpreter from within a extension module. Is
there any reason I can't do this? Thanks in advance.

                                                             Faheem.

building 'embed' extension
gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
-Wstrict-prototypes -fPIC -I/usr/include/python2.3 -c embed.cc -o
/tmp/temp.linux-i686-2.3/embed.o -Wl,-E
embed.cc: In function `void foo()':
embed.cc:35: error: no matching function for call to `
   boost::python::handle<PyObject>::handle(const char[73], int,
   PyObject*, 
   PyObject*)'
/usr/include/boost/python/handle.hpp:129: error: candidates are: 
   boost::python::handle<T>::handle(boost::python
   ::detail::borrowed_reference_t*)
   [with T = PyObject]
/usr/include/boost/python/handle.hpp:110: error:                 
   boost::python::handle<T>::handle(const boost::python::handle<T>&)
   [with T = 
   PyObject]
/usr/include/boost/python/handle.hpp:205: error:                 
   boost::python::handle<T>::handle() [with T = PyObject]
error: command 'gcc' failed with exit status 1
make: *** [embed.so] Error 1
make: Target `all' not remade because of errors.

Compilation exited abnormally with code 2 at Mon Jun 28 00:02:00


*************************************************************************
embed.cc
*************************************************************************
#include <boost/python.hpp>
#include <boost/scoped_ptr.hpp>
#include <python2.3/Python.h>
#include <pythonrun.h>
using namespace boost::python;

void foo()
{
  Py_Initialize();

  handle<> main_module(borrowed( PyImport_AddModule("__main__") ));
  
  handle<> main_namespace(borrowed( PyModule_GetDict(main_module.get()) ));
    
  handle<>( PyRun_String("hello = file('hello.txt', 'w')\n"
			 "hello.write('Hello world!')\n"
                           "hello.close()", Py_file_input,
                           main_namespace.get(), main_namespace.get()) );
  
    Py_Finalize();
}

BOOST_PYTHON_MODULE(embed)
{
  def("foo",foo);
}
*************************************************************************




From caustin at gmail.com  Mon Jun 28 06:46:06 2004
From: caustin at gmail.com (Chad Austin)
Date: Sun, 27 Jun 2004 23:46:06 -0500
Subject: [C++-sig] compile errors with embedding example from tutorial
In-Reply-To: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
Message-ID: <2096514e040627214639a53ca5@mail.gmail.com>

Hi Faheem,

I ran into that too.  In short, the line with "handle<>( PyRun_String"
isn't being understood how you want it too...  You can surround that
expression with parenthesis, and it should work:

(handle<>( PyRun_String("hello = file('hello.txt', 'w')\n"
                        "hello.write('Hello world!')\n"
                          "hello.close()", Py_file_input,
                          main_namespace.get(), main_namespace.get()) ));

Chad

On Mon, 28 Jun 2004 04:26:19 +0000 (UTC), Faheem Mitha
<faheem at email.unc.edu> wrote:
> 
> Dear People,
> 
> When trying to compile the following code (direct from sample code in
> the turorial) I get these errors. I am using distutils and realise my
> flags may not be exactly right, but the compiler seems to be saying it
> cannot find the correct version of handler to use! This does not look
> like it has anything to do with the correct flags, though I could be
> wrong.
> 
> I'm trying to call the interpreter from within a extension module. Is
> there any reason I can't do this? Thanks in advance.
> 
>                                                              Faheem.
> 
> building 'embed' extension
> gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
> -Wstrict-prototypes -fPIC -I/usr/include/python2.3 -c embed.cc -o
> /tmp/temp.linux-i686-2.3/embed.o -Wl,-E
> embed.cc: In function `void foo()':
> embed.cc:35: error: no matching function for call to `
>    boost::python::handle<PyObject>::handle(const char[73], int,
>    PyObject*,
>    PyObject*)'
> /usr/include/boost/python/handle.hpp:129: error: candidates are:
>    boost::python::handle<T>::handle(boost::python
>    ::detail::borrowed_reference_t*)
>    [with T = PyObject]
> /usr/include/boost/python/handle.hpp:110: error:
>    boost::python::handle<T>::handle(const boost::python::handle<T>&)
>    [with T =
>    PyObject]
> /usr/include/boost/python/handle.hpp:205: error:
>    boost::python::handle<T>::handle() [with T = PyObject]
> error: command 'gcc' failed with exit status 1
> make: *** [embed.so] Error 1
> make: Target `all' not remade because of errors.
> 
> Compilation exited abnormally with code 2 at Mon Jun 28 00:02:00
> 
> *************************************************************************
> embed.cc
> *************************************************************************
> #include <boost/python.hpp>
> #include <boost/scoped_ptr.hpp>
> #include <python2.3/Python.h>
> #include <pythonrun.h>
> using namespace boost::python;
> 
> void foo()
> {
>   Py_Initialize();
> 
>   handle<> main_module(borrowed( PyImport_AddModule("__main__") ));
> 
>   handle<> main_namespace(borrowed( PyModule_GetDict(main_module.get()) ));
> 
>   handle<>( PyRun_String("hello = file('hello.txt', 'w')\n"
>                          "hello.write('Hello world!')\n"
>                            "hello.close()", Py_file_input,
>                            main_namespace.get(), main_namespace.get()) );
> 
>     Py_Finalize();
> }
> 
> BOOST_PYTHON_MODULE(embed)
> {
>   def("foo",foo);
> }
> *************************************************************************
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>



From gaiacrtn at free.fr  Mon Jun 28 15:21:24 2004
From: gaiacrtn at free.fr (Baptiste Lepilleur)
Date: Mon, 28 Jun 2004 15:21:24 +0200
Subject: [C++-sig] pyste, patch to allow injecting code in class definition
Message-ID: <008301c45d12$d086b4a0$603f223e@lain>

    The attached patch allows the user to inject code in the class
definition code, just like declaration_code() & module_code(). Among other
things, this allows us to add properties when  pyste support fail (anonymous
union...).

    Example of usage:

Vector3 = Class( "Ogre::Vector3", "OgreVector3.h" )

class_code( Vector3, '.def_readwrite( "x", &Ogre::Vector3::x )' )

class_code( Vector3, '.def_readwrite( "y", &Ogre::Vector3::y )' )

class_code( Vector3, '.def_readwrite( "z", &Ogre::Vector3::z )' )


    Which produces:

class_< Ogre::Vector3 >("Vector3", init< >())

    .def(init< Ogre::Real, Ogre::Real, Ogre::Real >())

    .def_readwrite( "x", &Ogre::Vector3::x )

    .def_readwrite( "y", &Ogre::Vector3::y )

    .def_readwrite( "z", &Ogre::Vector3::z )

    ;

    Baptiste.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: class_code.patch
Type: application/octet-stream
Size: 1819 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040628/f31f2406/attachment.obj>

From faheem at email.unc.edu  Mon Jun 28 19:48:59 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Mon, 28 Jun 2004 17:48:59 +0000 (UTC)
Subject: [C++-sig] Re: compile errors with embedding example from tutorial
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
	<2096514e040627214639a53ca5@mail.gmail.com>
Message-ID: <slrnce0mgb.om1.faheem@Chrestomanci.home.earth>

On Sun, 27 Jun 2004 23:46:06 -0500, Chad Austin <caustin at gmail.com> wrote:

> Hi Faheem,

Hi Chad,

Thanks for replying.

> I ran into that too.  In short, the line with "handle<>( PyRun_String"
> isn't being understood how you want it too...  You can surround that
> expression with parenthesis, and it should work:
>
> (handle<>( PyRun_String("hello = file('hello.txt', 'w')\n"
>                         "hello.write('Hello world!')\n"
>                           "hello.close()", Py_file_input,
>                           main_namespace.get(), main_namespace.get()) ));

Thanks. Indeed it does. I hope the tutorial can be changed
appropriately, unless it already has been, of course. Any idea why
this is necessary? It isn't in the other two places <handle> is used.

I can load my module, but when I try to run the function, I get

NameError: name 'file' is not defined

Any idea about this?

If you know of other places the tutorial needs to be changed, I would
appreciate being informed of them.

I think it might be useful if I said explicitly what it is I am trying
to do here, and ask for suggestions.

I'm writing an extension module, and want to call a python function,
namely array from the numarray.strings module. This is part of the
Python Numarray library. The question is how to do this.

I have tried doing this with callbacks and it seems to work fine. The
trouble with callbacks is that it means I have to have a python
wrapper with an argument slot for every function I want to
import. This is clunky.

Suppose I have three functions, say f1, f2, f3.

Then I have f1 calling f2 calling f3. 

f1 -> f2 -> f3

Suppose f3 calls some python function (say `foo') from an external
module or something.

Then if I rewrite f3 using Boost.Python, I still need a wrapper, say
f1', to pass `foo' as an argument. Then I need

f1 -> f2 ->  f3' -> f3

I suppose I don't need `foo' passed down from f1 because if I import
foo into global namespace into the same module where I define f1, f2,
f3', then Python's scoping rules will ensure that if I use `foo' as an
argument to f3' then it will be picked up from the global namespace.

Furthermore, if I decide to rewrite f2 into Boost.Python, then I would
have to create another Python wrapper f2' for f2, and pass f3' as an
argument to f2' (ugh). So I'd get the following

f1 -> f2' -> f2 -> f3' -> f3 

I'd rather have 

f1 -> f2 -> f3

where f2 and f3 only exist as C++ functions. Much cleaner.

So I thought that perhaps I could imbed the interpreter inside (say)
f3, and persuade it to make `foo' available to me inside C++. I don't
know if this is even possible. I have seen no examples of this. In
fact, I have seen no examples of using the python interpeter inside
extension modules. In fact, I've been rather unsuccessful so far in
getting anything related to embedding to work.

Am I missing something? Thanks in advance for any information.

                                                            Faheem.




From jeff.holle at verizon.net  Mon Jun 28 06:05:01 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Mon, 28 Jun 2004 00:05:01 -0400
Subject: [C++-sig] Calling a python function from C++.
Message-ID: <40DF98ED.8070800@verizon.net>

I'm using boost 1_31_0 on Mandrake 10.0 Linux.

I've created a simple demonstration of calling a python function with 
C++ object args using python.boost.
It currently works with one exception.  In the face of a missing python 
script, it "aborts" instead of properly throwing an exception if I use 
the "handle<>" machinery that boost.python offers.  The following code 
fragment contains the working code with the stuff that should do the 
equivalent, but exibits this problem, commented out.

PyObject *    /* Returns a borrowed reference */
getPythonFunction(string moduleName,string functionName)
{
//    handle<> _module( PyImport_ImportModule(const_cast<char 
*>(moduleName.c_str())));
//    if ( _module.get() == NULL) throwException("The python module 
\"",moduleName.c_str(),"\" is not available",NULL);
//    handle<> _namespace(borrowed( PyModule_GetDict(_module.get())));
//    if ( _namespace.get() == NULL) throwException("Failed to acquire 
the dictionary of \"",moduleName.c_str(),"\"",NULL);
//    handle<> _function(borrowed(PyDict_GetItemString( 
_namespace.get(),functionName.c_str())));
//    if ( _function.get() == NULL) throwException("Failed to find 
needed function \"",functionName.c_str(),"\"",NULL);
//  return _function.get();
    PyObject *pModule = PyImport_ImportModule(const_cast<char 
*>(moduleName.c_str()));
    if ( pModule == NULL) throwException("The python module 
\"",moduleName.c_str(),"\" is not available",NULL);
    PyObject *pDict = PyModule_GetDict(pModule);
    if ( pDict == NULL) throwException("Failed to acquire the dictionary 
of \"",moduleName.c_str(),"\"",NULL);
    PyObject *pFunc = PyDict_GetItemString(pDict,functionName.c_str());
    if ( pFunc == NULL) throwException("Failed to find requested python 
function \"",functionName.c_str(),"\"",NULL);
    Py_DECREF(pModule);
    return pFunc;
}

In this form, this function throws a std::run_time exception which is 
successfully caught with a "catch (exception& x)" clause.  With the 
commented out section active, something is thrown, but it can only be 
caught with "catch(...)".

I'm wondering if anybody can explain this.

I'm also wondering if I could contribute this example to boost.python.  
The python script that I call is:
    import sys
    from Greeter import *

    def AlterGreeting( greeting):
      greeting.setLanguage("Python")
      return greeting.greet()

    if __name__ == '__main__':
      print 'testing module'
      greetingObj = Greeter("C++")
      greeting = AlterGreeting(greetingObj)
      print greeting

Note that the Greeter module is a C++ extension module also created with 
boost.python using pyste.
The C++ part of this example does the same thing that the main module 
specific section of this python script does.

If anybody wants this example, simple request it.  I'll provide the 
complete project, which uses bjam v1, in a tar ball.




From caustin at gmail.com  Mon Jun 28 22:56:13 2004
From: caustin at gmail.com (Chad Austin)
Date: Mon, 28 Jun 2004 15:56:13 -0500
Subject: [C++-sig] Re: compile errors with embedding example from tutorial
In-Reply-To: <slrnce0mgb.om1.faheem@Chrestomanci.home.earth>
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
	<2096514e040627214639a53ca5@mail.gmail.com>
	<slrnce0mgb.om1.faheem@Chrestomanci.home.earth>
Message-ID: <2096514e04062813561ceb1d79@mail.gmail.com>

On Mon, 28 Jun 2004 17:48:59 +0000 (UTC), Faheem Mitha
<faheem at email.unc.edu> wrote:
> 
> On Sun, 27 Jun 2004 23:46:06 -0500, Chad Austin <caustin at gmail.com> wrote:
> 
> > Hi Faheem,
> 
> Hi Chad,
> 
> Thanks for replying.
> 
> > I ran into that too.  In short, the line with "handle<>( PyRun_String"
> > isn't being understood how you want it too...  You can surround that
> > expression with parenthesis, and it should work:
> >
> > (handle<>( PyRun_String("hello = file('hello.txt', 'w')\n"
> >                         "hello.write('Hello world!')\n"
> >                           "hello.close()", Py_file_input,
> >                           main_namespace.get(), main_namespace.get()) ));
> 
> Thanks. Indeed it does. I hope the tutorial can be changed
> appropriately, unless it already has been, of course. Any idea why
> this is necessary? It isn't in the other two places <handle> is used.

I think it's something to do with C++ thinking you're trying to
declare a function or function pointer or something.  That's the best
I could come up with...  Any insight from others would be appreciated.
 :)

> I can load my module, but when I try to run the function, I get
> 
> NameError: name 'file' is not defined
> 
> Any idea about this?

Nope, sorry...  I had a problem similar to that in that __builtins__
wasn't getting set in my modules, but I was doing some weird stuff. 
You may want to at least check that __builtins__ is set in your
script.

> If you know of other places the tutorial needs to be changed, I would
> appreciate being informed of them.
> 
> I think it might be useful if I said explicitly what it is I am trying
> to do here, and ask for suggestions.
> 
> I'm writing an extension module, and want to call a python function,
> namely array from the numarray.strings module. This is part of the
> Python Numarray library. The question is how to do this.

Something like:

    object m = object(handle<>( PyImport_Import("numarray.strings") ));
    object result = m.attr("array")(arguments);

should work for you.

> I have tried doing this with callbacks and it seems to work fine. The
> trouble with callbacks is that it means I have to have a python
> wrapper with an argument slot for every function I want to
> import. This is clunky.
>
> Suppose I have three functions, say f1, f2, f3.
> 
> Then I have f1 calling f2 calling f3.
> 
> f1 -> f2 -> f3
> 
> Suppose f3 calls some python function (say `foo') from an external
> module or something.
> 
> Then if I rewrite f3 using Boost.Python, I still need a wrapper, say
> f1', to pass `foo' as an argument. Then I need
> 
> f1 -> f2 ->  f3' -> f3
> 
> I suppose I don't need `foo' passed down from f1 because if I import
> foo into global namespace into the same module where I define f1, f2,
> f3', then Python's scoping rules will ensure that if I use `foo' as an
> argument to f3' then it will be picked up from the global namespace.
> 
> Furthermore, if I decide to rewrite f2 into Boost.Python, then I would
> have to create another Python wrapper f2' for f2, and pass f3' as an
> argument to f2' (ugh). So I'd get the following
> 
> f1 -> f2' -> f2 -> f3' -> f3
> 
> I'd rather have
> 
> f1 -> f2 -> f3
> 
> where f2 and f3 only exist as C++ functions. Much cleaner.
> 
> So I thought that perhaps I could imbed the interpreter inside (say)
> f3, and persuade it to make `foo' available to me inside C++. I don't
> know if this is even possible. I have seen no examples of this. In
> fact, I have seen no examples of using the python interpeter inside
> extension modules. In fact, I've been rather unsuccessful so far in
> getting anything related to embedding to work.

I'm totally confused by everything you just said about callbacks, but
you should be able to easily call back into Python from C++ using
Boost.Python's object (and related subclasses) or even the native
Python/C API.  I don't see a reason importing another native module
from a native module would cause a problem, as long as you write your
C++ code in the same way you'd write the Python.

> Am I missing something? Thanks in advance for any information.

Hope that helps,

-- 
Chad Austin
http://aegisknight.org/



From dleary at gmail.com  Tue Jun 29 00:17:03 2004
From: dleary at gmail.com (Dusty Leary)
Date: Mon, 28 Jun 2004 18:17:03 -0400
Subject: [C++-sig] Re: implicitly_convertible and char*
In-Reply-To: <un02oh5wj.fsf@boost-consulting.com>
References: <81c509c04062609594758fdc4@mail.gmail.com>
	<un02oh5wj.fsf@boost-consulting.com>
Message-ID: <81c509c04062815172a69324a@mail.gmail.com>

full source code:
================================================
#define BOOST_PYTHON_STATIC_LIB

#include <boost/python.hpp>
using namespace boost::python;

struct X {
    std::string v;
    X(const std::string& s) : v(s) {}
    operator char*() const { return (char*)v.c_str(); }
};

void test_x(X& x) {
    char* x_string = x;
    printf("test_x: %s\n", x_string);
}

void test_string(char* s) {
    printf("test_string: %s\n", s);
}

BOOST_PYTHON_MODULE(test_string) {
    class_<X>("X", init<const std::string&>());
    def("test_x", test_x);
    def("test_string", test_string);
    implicitly_convertible<X, char*>();
}
================================================

interpreter:

>>> import test_string
>>> x = test_string.X("foo")
>>> test_string.test_x(x)
test_x: foo
>>> test_string.test_string("blah")
test_string: blah
>>> test_string.test_string(x)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
Boost.Python.ArgumentError: Python argument types in
    test_string.test_string(X)
did not match C++ signature:
    test_string(char *)




On Sun, 27 Jun 2004 20:39:56 -0400, David Abrahams
<dave at boost-consulting.com> wrote:
> 
> 
> Dusty Leary <dleary at gmail.com> writes:
> 
> > The implicitly_convertible sample at
> > http://boost.org/libs/python/doc/v2/implicit.html#implicitly_convertible-spec
> > works fine for me.
> >
> > However, changing "int" to "char*" makes it stop working.  Compiles
> > fine, won't invoke at runtime.
> >
> >>From the traceback:
> >
> > Boost.Python.ArgumentError: Python argument types in
> >     __testmodule__.test_string(X)
> > did not match C++ signature:
> >     test_string(char *)
> >
> > I am using MSVC7.1
> >
> > Any info?
> 
> Full source code of your example?
> Perhaps the problem is that X isn't actually convertible to char*?
> 
> --
> Dave Abrahams
> Boost Consulting
> http://www.boost-consulting.com
> 
> _______________________________________________
> C++-sig mailing list
> C++-sig at python.org
> http://mail.python.org/mailman/listinfo/c++-sig
>



From faheem at email.unc.edu  Tue Jun 29 00:30:33 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Mon, 28 Jun 2004 18:30:33 -0400
Subject: [C++-sig] Re: Calling a python function from C++.
In-Reply-To: <40DF98ED.8070800@verizon.net>
References: <40DF98ED.8070800@verizon.net>
Message-ID: <E1Bf4e3-0003M2-00@Chrestomanci>

In gmane.comp.python.c++, you wrote:
> I'm using boost 1_31_0 on Mandrake 10.0 Linux.
>
> I've created a simple demonstration of calling a python function with 
> C++ object args using python.boost.
> It currently works with one exception.  In the face of a missing python 
> script, it "aborts" instead of properly throwing an exception if I use 
> the "handle<>" machinery that boost.python offers.  The following code 
> fragment contains the working code with the stuff that should do the 
> equivalent, but exibits this problem, commented out.

Hi Jeff,

I'd be interested to see your code. I am trying (not very successfully
so far) to figure out this stuff myself, and any working examples are
helpful. It sounds like you have a better handle on this stuff (no pun
intended) than I do. :-)

Please email it to the my email address shown above. Thanks.

                                                              Faheem.




From faheem at email.unc.edu  Tue Jun 29 00:39:15 2004
From: faheem at email.unc.edu (Faheem Mitha)
Date: Mon, 28 Jun 2004 22:39:15 +0000 (UTC)
Subject: [C++-sig] Re: compile errors with embedding example from tutorial
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
	<2096514e040627214639a53ca5@mail.gmail.com>
	<slrnce0mgb.om1.faheem@Chrestomanci.home.earth>
	<2096514e04062813561ceb1d79@mail.gmail.com>
Message-ID: <slrnce17gj.om1.faheem@Chrestomanci.home.earth>

On Mon, 28 Jun 2004 15:56:13 -0500, Chad Austin <caustin at gmail.com> wrote:
> On Mon, 28 Jun 2004 17:48:59 +0000 (UTC), Faheem Mitha
> <faheem at email.unc.edu> wrote:
>> 
>> On Sun, 27 Jun 2004 23:46:06 -0500, Chad Austin <caustin at gmail.com> wrote:

> Nope, sorry...  I had a problem similar to that in that __builtins__
> wasn't getting set in my modules, but I was doing some weird stuff. 
> You may want to at least check that __builtins__ is set in your
> script.

How should I do that? I thought that

handle<> main_module(borrowed( PyImport_AddModule("__main__") ));
handle<> main_namespace(borrowed( PyModule_GetDict(main_module.get())));

should do the trick, though I don't really understand what these lines
do. I wonder if the fact that I am calling the interpreter inside an
extension module is causing any problems.
  
>> If you know of other places the tutorial needs to be changed, I would
>> appreciate being informed of them.
>> 
>> I think it might be useful if I said explicitly what it is I am trying
>> to do here, and ask for suggestions.
>> 
>> I'm writing an extension module, and want to call a python function,
>> namely array from the numarray.strings module. This is part of the
>> Python Numarray library. The question is how to do this.
>
> Something like:
>
>     object m = object(handle<>( PyImport_Import("numarray.strings") ));
>     object result = m.attr("array")(arguments);
>
> should work for you.

Ok. Thanks.

>> I have tried doing this with callbacks and it seems to work fine. The
>> trouble with callbacks is that it means I have to have a python
>> wrapper with an argument slot for every function I want to
>> import. This is clunky.

[stuff deleted...]

> I'm totally confused by everything you just said about callbacks, but
> you should be able to easily call back into Python from C++ using
> Boost.Python's object (and related subclasses) or even the native
> Python/C API.  I don't see a reason importing another native module
> from a native module would cause a problem, as long as you write your
> C++ code in the same way you'd write the Python.

Sorry I was unclear. I'd be willing to try to clarify if that would be
useful. I can certainly make callbacks work. The point is I have to
pass the relevant functions from Python, which imposes some overhead,
which I'd like to get rid of.
                                                            Faheem.




From caustin at gmail.com  Tue Jun 29 07:58:48 2004
From: caustin at gmail.com (Chad Austin)
Date: Tue, 29 Jun 2004 00:58:48 -0500
Subject: [C++-sig] Re: compile errors with embedding example from tutorial
In-Reply-To: <slrnce17gj.om1.faheem@Chrestomanci.home.earth>
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
	<2096514e040627214639a53ca5@mail.gmail.com>
	<slrnce0mgb.om1.faheem@Chrestomanci.home.earth>
	<2096514e04062813561ceb1d79@mail.gmail.com>
	<slrnce17gj.om1.faheem@Chrestomanci.home.earth>
Message-ID: <2096514e0406282258665b43c8@mail.gmail.com>

On Mon, 28 Jun 2004 22:39:15 +0000 (UTC), Faheem Mitha
<faheem at email.unc.edu> wrote:
> 
> On Mon, 28 Jun 2004 15:56:13 -0500, Chad Austin <caustin at gmail.com> wrote:
> > On Mon, 28 Jun 2004 17:48:59 +0000 (UTC), Faheem Mitha
> > <faheem at email.unc.edu> wrote:
> >>
> >> On Sun, 27 Jun 2004 23:46:06 -0500, Chad Austin <caustin at gmail.com> wrote:
> 
> > Nope, sorry...  I had a problem similar to that in that __builtins__
> > wasn't getting set in my modules, but I was doing some weird stuff.
> > You may want to at least check that __builtins__ is set in your
> > script.
> 
> How should I do that? I thought that
> 
> handle<> main_module(borrowed( PyImport_AddModule("__main__") ));
> handle<> main_namespace(borrowed( PyModule_GetDict(main_module.get())));
> 
> should do the trick, though I don't really understand what these lines
> do. I wonder if the fact that I am calling the interpreter inside an
> extension module is causing any problems.

Just try evaluating the script "import __builtin__" or "str(10)" and
see if it fails or not.  ;)

There should be no problem with calling interpreter from an extension
module.  Language bindings are a two-way street, so there necessarily
must be communication between the two languages.

-- 
Chad Austin
http://aegisknight.org/



From thor.arne.johansen at ibas.no  Tue Jun 29 10:38:55 2004
From: thor.arne.johansen at ibas.no (thor.arne.johansen at ibas.no)
Date: Tue, 29 Jun 2004 10:38:55 +0200
Subject: [C++-sig] Re: Calling a python function from C++.
Message-ID: <OF64EBC804.FAEFD609-ONC1256EC2.002EEDA0@no>


>In gmane.comp.python.c++, you wrote:
>> I'm using boost 1_31_0 on Mandrake 10.0 Linux.
>>
>> I've created a simple demonstration of calling a python function with
>> C++ object args using python.boost.
>> It currently works with one exception.  In the face of a missing python
>> script, it "aborts" instead of properly throwing an exception if I use
>> the "handle<>" machinery that boost.python offers.  The following code
>> fragment contains the working code with the stuff that should do the
>> equivalent, but exibits this problem, commented out.
>
>Hi Jeff,
>
>I'd be interested to see your code. I am trying (not very successfully
>so far) to figure out this stuff myself, and any working examples are
>helpful. It sounds like you have a better handle on this stuff (no pun
>intended) than I do. :-)
>
>Please email it to the my email address shown above. Thanks.
>
>                                                              Faheem.

A while back I posted some example code for embedding python and have
python 'scripts' manipulate
C++ objects. It might also be of interest:

http://mail.python.org/pipermail/c++-sig/2002-December/002833.html

--
Thor Arne Johansen
Dept. Manager R&D
Ibas AS




From dave at boost-consulting.com  Tue Jun 29 15:58:33 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 09:58:33 -0400
Subject: [C++-sig] Re: Calling a python function from C++.
References: <40DF98ED.8070800@verizon.net>
Message-ID: <uwu1q4gae.fsf@boost-consulting.com>

Jeff Holle <jeff.holle at verizon.net> writes:

> I'm using boost 1_31_0 on Mandrake 10.0 Linux.
>
> I've created a simple demonstration of calling a python function with
> C++ object args using python.boost.
> It currently works with one exception.  In the face of a missing
> python script, it "aborts" instead of properly throwing an exception
> if I use the "handle<>" machinery that boost.python offers.  The
> following code fragment contains the working code with the stuff that
> should do the equivalent, but exibits this problem, commented out.
>
> PyObject *    /* Returns a borrowed reference */
> getPythonFunction(string moduleName,string functionName)
> {
> //    handle<> _module( PyImport_ImportModule(const_cast<char
> *>(moduleName.c_str())));
> //    if ( _module.get() == NULL) throwException("The python module
> \"",moduleName.c_str(),"\" is not available",NULL);


> //    handle<> _namespace(borrowed( PyModule_GetDict(_module.get())));
> //    if ( _namespace.get() == NULL) throwException("Failed to acquire
> the dictionary of \"",moduleName.c_str(),"\"",NULL);
> //    handle<> _function(borrowed(PyDict_GetItemString(
> _namespace.get(),functionName.c_str())));
> //    if ( _function.get() == NULL) throwException("Failed to find
> needed function \"",functionName.c_str(),"\"",NULL);
> //  return _function.get();

Why struggle?
This should be:

object    /* Returns a borrowed reference */
getPythonFunction(string moduleName,string functionName)
{
    object _module(handle<>(
              PyImport_ImportModule(
                    const_cast<char*>(moduleName.c_str()))));

    return _module.attr(functionName.c_str());
}

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




From dave at boost-consulting.com  Tue Jun 29 17:51:48 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 11:51:48 -0400
Subject: [C++-sig] Re: compile errors with embedding example from tutorial
References: <slrncdv7fb.om1.faheem@Chrestomanci.home.earth>
	<2096514e040627214639a53ca5@mail.gmail.com>
	<slrnce0mgb.om1.faheem@Chrestomanci.home.earth>
	<2096514e04062813561ceb1d79@mail.gmail.com>
	<slrnce17gj.om1.faheem@Chrestomanci.home.earth>
Message-ID: <uhdsutl9n.fsf@boost-consulting.com>

Faheem Mitha <faheem at email.unc.edu> writes:

> On Mon, 28 Jun 2004 15:56:13 -0500, Chad Austin <caustin at gmail.com> wrote:
>> On Mon, 28 Jun 2004 17:48:59 +0000 (UTC), Faheem Mitha
>> <faheem at email.unc.edu> wrote:
>>> 
>>> On Sun, 27 Jun 2004 23:46:06 -0500, Chad Austin <caustin at gmail.com> wrote:
>
>> Nope, sorry...  I had a problem similar to that in that __builtins__
>> wasn't getting set in my modules, but I was doing some weird stuff. 
>> You may want to at least check that __builtins__ is set in your
>> script.
>
> How should I do that? I thought that
>
> handle<> main_module(borrowed( PyImport_AddModule("__main__") ));
> handle<> main_namespace(borrowed( PyModule_GetDict(main_module.get())));
>
> should do the trick, though I don't really understand what these lines
> do. 

What specifically are you confused about?  Did you check the docs?  
I don't know why you'd do it that way when you can write:

  object main_module(handle<>(borrowed( PyImport_AddModule("__main__"))));
  object main_namespace = main_module.attr("__dict__");

instead.

> I wonder if the fact that I am calling the interpreter inside an
> extension module is causing any problems.

Certainly not.  Calling the interpreter inside an extension module is normal.


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




From dave at boost-consulting.com  Tue Jun 29 17:54:55 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 11:54:55 -0400
Subject: [C++-sig] Re: implicitly_convertible and char*
References: <81c509c04062609594758fdc4@mail.gmail.com>
	<un02oh5wj.fsf@boost-consulting.com>
	<81c509c04062815172a69324a@mail.gmail.com>
Message-ID: <ubrj2tl4g.fsf@boost-consulting.com>

Dusty Leary <dleary at gmail.com> writes:

> full source code:
> ================================================
> #define BOOST_PYTHON_STATIC_LIB
>
> #include <boost/python.hpp>
> using namespace boost::python;
>
> struct X {
>     std::string v;
>     X(const std::string& s) : v(s) {}
>     operator char*() const { return (char*)v.c_str(); }
> };
>
> void test_x(X& x) {
>     char* x_string = x;
>     printf("test_x: %s\n", x_string);
> }
>
> void test_string(char* s) {
>     printf("test_string: %s\n", s);
> }

Oh!

char* is not a C++ string, and there's no built-in converter for that.
I note that you're casting away the const-ness of v.c_str() above;
that's very bad juju.  If you make char* into char const* everywhere,
your example should work.  Alternatively, you can also register a
custom from_python converter for char*... but I recommend against it.

>
> BOOST_PYTHON_MODULE(test_string) {
>     class_<X>("X", init<const std::string&>());
>     def("test_x", test_x);
>     def("test_string", test_string);
>     implicitly_convertible<X, char*>();
> }
> ================================================
>
> interpreter:
>
>>>> import test_string
>>>> x = test_string.X("foo")
>>>> test_string.test_x(x)
> test_x: foo
>>>> test_string.test_string("blah")
> test_string: blah
>>>> test_string.test_string(x)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> Boost.Python.ArgumentError: Python argument types in
>     test_string.test_string(X)
> did not match C++ signature:
>     test_string(char *)

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




From hoeppler at diener.iap.physik.uni-tuebingen.de  Tue Jun 29 18:01:42 2004
From: hoeppler at diener.iap.physik.uni-tuebingen.de (Chris Hoeppler)
Date: Tue, 29 Jun 2004 18:01:42 +0200 (CEST)
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <20040625135525.94532.qmail@web20205.mail.yahoo.com>
Message-ID: <Pine.LNX.4.33.0406291720550.1603-100000@diener.iap.physik.uni-tuebingen.de>

On Fri, 25 Jun 2004, Ralf W. Grosse-Kunstleve wrote:

> You have to strip down the code until you've isolated the construct that causes
> the ICE. Then try to rearrange that construct until the ICE goes away. In the
> early days of Boost.Python I've done this a million times, and I am sure David
> did it a billion times. It always takes too long, is very frustrating, and
> therefore requires a lot of determination. If this is not enough to scare you,
> copy the content of boost/python.hpp into your source file, then comment out
> half the includes, see if it still ICE's, if yes comment out half of the
> remaining, if not put back half of the ones you commented out, and so on until
> you've found that one include that makes a difference. Then copy that into your
> source code and repeat the exercise until you found the piece of code that
> makes the difference. Then go into a different trial-and-error loop rearranging
> the code. While the first step requires mainly determination, you will need
> some creativity here. Often my last resort was to ask David. Finally you have
> to communicate the fix to the developers of the effected library (e.g. it could
> be in boost::graph or boost::bind or, heaven forbid, in boost::mpl or ...) and
> convince them to adopt your fix.

Okay. Thanks for the encouragement, Ralf. Here's the result of my
'investigation':

On the following system:

    gcc version 3.3.1 (SuSE Linux 9.0)
    boost-1.31

if I try to compile <boost/python/detail/translate_exception.hpp> (copied
to test.cpp) with the following command line

    g++ -Wall -ftemplate-depth-100 -DBOOST_PYTHON_DYNAMIC_LIB -g -O0 -fno-inline -fPIC -DDEBUG -I/usr/local/include/python2.3
-I/usr/local/include/boost-1_31 -c -o test.os test.cpp

the compiler segfaults with the following message:
    /usr/include/g++/bits/basic_string.h:728: internal compiler error:
Segmentation fault

The relevant line in translate_exception.hpp triggering the segfault is
indicated below:

template <class ExceptionType, class Translate>
struct translate_exception
{
    typedef typename add_reference<
        typename add_const<ExceptionType>::type
    >::type exception_cref;

    inline bool operator()(
        exception_handler const& handler
      , function0<void> const& f
      , typename call_traits<Translate>::param_type translate) const
    {
        try
        {
            return handler(f);
        }
        catch(exception_cref e)
--------------^^^^^^^^^^^^^^^^
        {
            translate(e);
            return true;
        }
    }
};

FWIW, the compiler is happy with  "catch(exception_cref &e)", though I do
--------------------------------------------------------^
not know whether it would be okay to make that change...

Chris.




From dave at boost-consulting.com  Tue Jun 29 18:36:53 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 12:36:53 -0400
Subject: [C++-sig] Re: Boost.Python in SUSE 9.0 or SUSE 9.1?
References: <20040625135525.94532.qmail@web20205.mail.yahoo.com>
	<Pine.LNX.4.33.0406291720550.1603-100000@diener.iap.physik.uni-tuebingen.de>
Message-ID: <ulli6s4m2.fsf@boost-consulting.com>

Chris Hoeppler <hoeppler at diener.iap.physik.uni-tuebingen.de> writes:

> FWIW, the compiler is happy with  "catch(exception_cref &e)", though I do
> --------------------------------------------------------^
> not know whether it would be okay to make that change...

Sure; you can do that in your local copy.
-- 
Dave Abrahams
Boost Consulting
http://www.boost-consulting.com




From jeff.holle at verizon.net  Tue Jun 29 04:30:50 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Mon, 28 Jun 2004 22:30:50 -0400
Subject: [C++-sig] Re: Calling a python function from C++
Message-ID: <40E0D45A.4020001@verizon.net>

Dave wrote:

Why struggle?
This should be:

object    /* Returns a borrowed reference */
getPythonFunction(string moduleName,string functionName)
{
    object _module(handle<>( PyImport_ImportModule(const_cast<char*>(moduleName.c_str())))); <<-- line 63 in Main.cpp

    return _module.attr(functionName.c_str());
}

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

Always a good question.
The code doesn't compile as presented.
I added #include <boost/python/object.hpp> to my module, have completely opened the boost::python namespace.

Get the error:

    Main.cpp:63: error: variable declaration is not allowed here


If I just instanciate an empty object with its default constructor, the module compiles.

An underlining concern I have is error handling.  When PyImport_ImportModule fails, a NULL pointer is returned.
Seems like the underlining machinery in "handle<>" is intolerant of this.






From dleary at gmail.com  Tue Jun 29 22:05:42 2004
From: dleary at gmail.com (Dusty Leary)
Date: Tue, 29 Jun 2004 16:05:42 -0400
Subject: [C++-sig] Re: implicitly_convertible and char*
In-Reply-To: <ubrj2tl4g.fsf@boost-consulting.com>
References: <81c509c04062609594758fdc4@mail.gmail.com>
	<un02oh5wj.fsf@boost-consulting.com>
	<81c509c04062815172a69324a@mail.gmail.com>
	<ubrj2tl4g.fsf@boost-consulting.com>
Message-ID: <81c509c040629130556f7312d@mail.gmail.com>

> 
> Oh!
> 
> char* is not a C++ string, and there's no built-in converter for that.
> I note that you're casting away the const-ness of v.c_str() above;
> that's very bad juju.  If you make char* into char const* everywhere,
> your example should work.  Alternatively, you can also register a
> custom from_python converter for char*... but I recommend against it.
> 

yeah, I just wrote up the simple test like that because I was lazy and
didn't think it would make a difference.  the same problem occurs with
const char*

new source code:
===================================================
#define BOOST_PYTHON_STATIC_LIB

#include <boost/python.hpp>
using namespace boost::python;

struct X {
    std::string v;
    X(const std::string& s) : v(s) {}
    operator char const*() const { return v.c_str(); }
};

void test_x(X& x) {
    char const* x_string = x;
    printf("test_x: %s\n", x_string);
}

void test_string(char const* s) {
    printf("test_string: %s\n", s);
}

BOOST_PYTHON_MODULE(test_string) {
    class_<X>("X", init<const std::string&>());
    def("test_x", test_x);
    def("test_string", test_string);
    implicitly_convertible<X, char const*>();
}
===================================================

>>> import test_string
>>> x = test_string.X("foo")
>>> test_string.test_string(x)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
Boost.Python.ArgumentError: Python argument types in
    test_string.test_string(X)
did not match C++ signature:
    test_string(char const *)
>>>



From rwgk at yahoo.com  Tue Jun 29 22:56:54 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Tue, 29 Jun 2004 13:56:54 -0700 (PDT)
Subject: [C++-sig] Re: implicitly_convertible and char*
In-Reply-To: <81c509c040629130556f7312d@mail.gmail.com>
Message-ID: <20040629205654.59109.qmail@web20212.mail.yahoo.com>

I don't really feel competent in this matter, but I believe char const* and
const char* are two different things. Did you try it with "const char*"?

--- Dusty Leary <dleary at gmail.com> wrote:
> > 
> > Oh!
> > 
> > char* is not a C++ string, and there's no built-in converter for that.
> > I note that you're casting away the const-ness of v.c_str() above;
> > that's very bad juju.  If you make char* into char const* everywhere,
> > your example should work.  Alternatively, you can also register a
> > custom from_python converter for char*... but I recommend against it.
> > 
> 
> yeah, I just wrote up the simple test like that because I was lazy and
> didn't think it would make a difference.  the same problem occurs with
> const char*
> 
> new source code:
> ===================================================
> #define BOOST_PYTHON_STATIC_LIB
> 
> #include <boost/python.hpp>
> using namespace boost::python;
> 
> struct X {
>     std::string v;
>     X(const std::string& s) : v(s) {}
>     operator char const*() const { return v.c_str(); }
> };
> 
> void test_x(X& x) {
>     char const* x_string = x;
>     printf("test_x: %s\n", x_string);
> }
> 
> void test_string(char const* s) {
>     printf("test_string: %s\n", s);
> }
> 
> BOOST_PYTHON_MODULE(test_string) {
>     class_<X>("X", init<const std::string&>());
>     def("test_x", test_x);
>     def("test_string", test_string);
>     implicitly_convertible<X, char const*>();
> }
> ===================================================
> 
> >>> import test_string
> >>> x = test_string.X("foo")
> >>> test_string.test_string(x)
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> Boost.Python.ArgumentError: Python argument types in
>     test_string.test_string(X)
> did not match C++ signature:
>     test_string(char const *)
> >>>



	
		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail 



From mike at nospam.com  Tue Jun 29 23:34:40 2004
From: mike at nospam.com (Mike Rovner)
Date: Tue, 29 Jun 2004 14:34:40 -0700
Subject: [C++-sig] Re: Re: implicitly_convertible and char*
References: <81c509c040629130556f7312d@mail.gmail.com>
	<20040629205654.59109.qmail@web20212.mail.yahoo.com>
Message-ID: <cbsn9e$7oo$1@sea.gmane.org>

Ralf W. Grosse-Kunstleve wrote:
> I don't really feel competent in this matter, but I believe char
> const* and const char* are two different things. Did you try it with
> "const char*"?

Nope. That's the same (read from right to left).
Different things is ...* const (constant pointer itself).

Mike

PS. only ddd afaik treat them differently ;)






From rwgk at yahoo.com  Tue Jun 29 23:54:34 2004
From: rwgk at yahoo.com (Ralf W. Grosse-Kunstleve)
Date: Tue, 29 Jun 2004 14:54:34 -0700 (PDT)
Subject: [C++-sig] Re: Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <ulli6s4m2.fsf@boost-consulting.com>
Message-ID: <20040629215434.65607.qmail@web20212.mail.yahoo.com>

--- David Abrahams <dave at boost-consulting.com> wrote:
> Chris Hoeppler <hoeppler at diener.iap.physik.uni-tuebingen.de> writes:
> 
> > FWIW, the compiler is happy with  "catch(exception_cref &e)", though I do
> > --------------------------------------------------------^
> > not know whether it would be okay to make that change...
> 
> Sure; you can do that in your local copy.

David is telling me:

> > Hi David, I think he was asking if we can make that change in the main
version
.
> > I think we should unless you have objections. I can run tests on all our
> > platforms to make sure it works everywhere. Please let me know.

> Yes, I have objections, at least if you don't make the patch very
> precise and conditional.  It will break on many compilers.  The
> add_reference invocation above it should have already added a
> reference and many compilers (even most GCCs?) will choke on adding a
> reference to a type that is already a reference.

I.e. what we need is a very specific set of #ifdef's that will apply your patch
only under SUSE 9, and only if the compiler with SUSE's patches is used. Here
is an alias that is sometimes helpful in pin-pointing system specific macros:

alias whatdefg++ 'echo '"'"'int main(void){return 0;}'"'"' >
/var/tmp/tmp$$.cpp; g++ -E -v \!* /var/tmp/tmp$$.cpp; rm -f /var/tmp/tmp$$.cpp'

Please use BOOST_WORKAROUND. There are lots of examples in the boost code, e.g.
boost/python/object/value_holder.hpp.

Ralf



		
__________________________________
Do you Yahoo!?
Yahoo! Mail - Helps protect you from nasty viruses.
http://promotions.yahoo.com/new_mail



From paul.bridger at paradise.net.nz  Wed Jun 30 02:08:31 2004
From: paul.bridger at paradise.net.nz (paul.bridger)
Date: Wed, 30 Jun 2004 12:08:31 +1200
Subject: [C++-sig] Abstract classes
Message-ID: <40E2047F.7050107@paradise.net.nz>

Pyste was generating constructors defs for abstract classes. I've changed 
Pyste so that it will insert no_init (and emit no constructors) for abstract 
classes.

Is this the correct behaviour for an abstract class?
eg.
class_< Foo, boost::noncopyable>("Foo", no_init)

This from the FAQ states this quite clearly:
http://www.python.org/cgi-bin/moinmoin/boost_2epython_2fFAQ
>     * no_init means "do not try to create an instance of that Python object, create __init__ function that throws an exception"
>     * noncopyable means "do not try to register a converter which can convert C++ return values of that class to Python". 
> 
> When wrapping an abstract class, it's necessary to specify both. 

However, what's confusing me is that I thought specifying no_init would make 
the following incorrect:

class Bar(CppModule.Foo):
	def __init__(self):
		CppModule.Foo.__init__(self)

Is that correct?



From dave at boost-consulting.com  Wed Jun 30 02:49:13 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 20:49:13 -0400
Subject: [C++-sig] Re: Calling a python function from C++
References: <40E0D45A.4020001@verizon.net>
Message-ID: <uzn6looom.fsf@boost-consulting.com>

Jeff Holle <jeff.holle at verizon.net> writes:

> Dave wrote:
>
> Why struggle?
> This should be:
>
> object    /* Returns a borrowed reference */
> getPythonFunction(string moduleName,string functionName)
> {
>     object _module(handle<>( PyImport_ImportModule(const_cast<char*>(moduleName.c_str())))); <<-- line 63 in Main.cpp
>
>     return _module.attr(functionName.c_str());
> }
>
> ------------------------------------------------------------------------
>
> Always a good question.
> The code doesn't compile as presented.

Oh, it should; GCC bug it looks like.

  object
  getPythonFunction(string moduleName,string functionName)
  {
      object _module = object(
          handle<>(
              PyImport_ImportModule(const_cast<char*>(moduleName.c_str()))
          )
      );

      return _module.attr(functionName.c_str());
  }

or 

  object
  getPythonFunction(string moduleName,string functionName)
  {
      object _module((
          handle<>(
              PyImport_ImportModule(const_cast<char*>(moduleName.c_str()))
          )
      ));

      return _module.attr(functionName.c_str());
  }

works.

> An underlining concern I have is error handling.  When
> PyImport_ImportModule fails, a NULL pointer is returned.
> Seems like the underlining machinery in "handle<>" is intolerant of
> this. 

What makes you say that?

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




From dave at boost-consulting.com  Wed Jun 30 02:49:39 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Tue, 29 Jun 2004 20:49:39 -0400
Subject: [C++-sig] Re: implicitly_convertible and char*
References: <81c509c040629130556f7312d@mail.gmail.com>
	<20040629205654.59109.qmail@web20212.mail.yahoo.com>
Message-ID: <uu0wtoonw.fsf@boost-consulting.com>

"Ralf W. Grosse-Kunstleve" <rwgk at yahoo.com> writes:

> I don't really feel competent in this matter, but I believe char const* and
> const char* are two different things. Did you try it with "const char*"?

They are the same thing.

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




From jeff.holle at verizon.net  Tue Jun 29 12:22:37 2004
From: jeff.holle at verizon.net (Jeff Holle)
Date: Tue, 29 Jun 2004 06:22:37 -0400
Subject: [C++-sig] Re: Calling a python function from C++ (from Jeff Holle)
Message-ID: <40E142ED.2030603@verizon.net>

>
>
>> An underlining concern I have is error handling.  When
>> PyImport_ImportModule fails, a NULL pointer is returned.
>> Seems like the underlining machinery in "handle<>" is intolerant of
>> this. 
>  
>
>
>What makes you say that?
>
You are correct about the success in compiling your much cleaner 
example, and it works :-) , but has a flaw.

When I hide the needed python script, the call to PyImport_ImportModule 
returns NULL.
When this occurres, both your example and what I had created, but 
commented out throws something because I can catch it with 
"catch(...)".  I don't however know how to be more specific.  Any 
hints?  Whatever it is, it doesn't inherit from "exception".

If I was operating on windows, I would anticipate it was a "structured 
exception", or whatever Microsoft calls their propritary stuff.  Don't 
know of the equivalent on linux though.  Without the "catch (...)", the 
command line program exits with "Aborted".






From jbrandmeyer at earthlink.net  Wed Jun 30 04:35:43 2004
From: jbrandmeyer at earthlink.net (Jonathan Brandmeyer)
Date: Tue, 29 Jun 2004 22:35:43 -0400
Subject: [C++-sig] Boost.Python in SUSE 9.0 or SUSE 9.1?
In-Reply-To: <2096514e04062507072f0ff45@mail.gmail.com>
References: <20040624144241.53669.qmail@web20207.mail.yahoo.com>
	<2096514e04062507072f0ff45@mail.gmail.com>
Message-ID: <1088562942.3005.17.camel@illuvatar>

On Fri, 2004-06-25 at 10:07, Chad Austin wrote:

> After some investigation, it appears SUSE applies some custom patches
> to their gcc builds, causing including boost/python.hpp to ICE.  I've
> submitted a support request to SUSE for this issue.  Installing a
> compatible gcc from source into /usr/local appears to work around the
> problem.

A coworker has SUSE 9.0 installed on his laptop and ran into the same
issue a couple of weeks ago.  He is out of the country on business right
now, so I'm going on memory about what we did.

When he contacted Suse about the issue, they said that they would not do
anything about it, since he could install a working GCC from source.  He
used GCC 3.3.3 from gcc.org (it was the latest in the 3.3 series at that
time), rebuilt Boost and the extensions with that, and everything
worked.  Using the --program-suffix= and
--enable-version-specific-runtime-libs to GCC's configure ensured that
the new compiler didn't stomp on Suse's custom compiler.  Finally, we
used the -sGXX= option to bjam to point it at the new compiler.

> > > If I build gcc 3.4.0 from
> > > source, Boost.Python compiles, but does not link do to
> > > incompatibilities in libstdc++.>
> > 
> > Does the link command fail, or is it a run-time failure (i.e. dlopen fails)?
> > If it is the latter you might have to update the LD_LIBRARY_PATH. This is what
> > I am using to activate gcc 3.4 in a new shell:
> 
> I believe it was an ABI change in libstdc++ or something.  My boost
> was compiled with gcc 3.3 and my app with 3.4.  I just started getting
> random crashes.

Gcc 3.4 uses libstdc++.so.6 while GCC 3.3 and 3.2 use libstdc++.so.5. 
They don't mix in the same program.

HTH,
-Jonathan




From roman.yakovenko at actimize.com  Wed Jun 30 08:23:20 2004
From: roman.yakovenko at actimize.com (Roman Yakovenko)
Date: Wed, 30 Jun 2004 09:23:20 +0300
Subject: [C++-sig] Abstract classes
Message-ID: <2D89F6C4A80FA547BF5D5B8FDDD0452325386B@exchange.adrembi.com>

Paul wrote:

> Pyste was generating constructors defs for abstract classes. 
> I've changed 
> Pyste so that it will insert no_init (and emit no 
> constructors) for abstract 
> classes.
> 
> Is this the correct behaviour for an abstract class?
> eg.
> class_< Foo, boost::noncopyable>("Foo", no_init)
> 
> This from the FAQ states this quite clearly:
> http://www.python.org/cgi-bin/moinmoin/boost_2epython_2fFAQ
> >     * no_init means "do not try to create an instance of 
> that Python object, create __init__ function that throws an exception"
> >     * noncopyable means "do not try to register a converter 
> which can convert C++ return values of that class to Python". 
> > 
> > When wrapping an abstract class, it's necessary to specify both. 
> 
> However, what's confusing me is that I thought specifying 
> no_init would make 
> the following incorrect:
> 
> class Bar(CppModule.Foo):
> 	def __init__(self):
> 		CppModule.Foo.__init__(self)
> 
> Is that correct?

I think this is correct. Abstract class has constructors. 

Roman

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.707 / Virus Database: 463 - Release Date: 6/15/2004
 




From shin1_morita at yahoo.co.jp  Wed Jun 30 17:03:09 2004
From: shin1_morita at yahoo.co.jp (Shin-ichi MORITA)
Date: Thu, 1 Jul 2004 00:03:09 +0900 (JST)
Subject: [C++-sig] Abstract classes
In-Reply-To: <2D89F6C4A80FA547BF5D5B8FDDD0452325386B@exchange.adrembi.com>
Message-ID: <20040630150309.44127.qmail@web2303.mail.yahoo.co.jp>

Hi,

> Paul wrote:
> 
> > Pyste was generating constructors defs for
> abstract classes. 
> > I've changed 
> > Pyste so that it will insert no_init (and emit no 
> > constructors) for abstract 
> > classes.
> > 
> > Is this the correct behaviour for an abstract
> class?
> > eg.
> > class_< Foo, boost::noncopyable>("Foo", no_init)
> > 
> > This from the FAQ states this quite clearly:
> >
>
http://www.python.org/cgi-bin/moinmoin/boost_2epython_2fFAQ
> > >     * no_init means "do not try to create an
> instance of 
> > that Python object, create __init__ function that
> throws an exception"
> > >     * noncopyable means "do not try to register
> a converter 
> > which can convert C++ return values of that class
> to Python". 
> > > 
> > > When wrapping an abstract class, it's necessary
> to specify both. 
> > 
> > However, what's confusing me is that I thought
> specifying 
> > no_init would make 
> > the following incorrect:
> > 
> > class Bar(CppModule.Foo):
> > 	def __init__(self):
> > 		CppModule.Foo.__init__(self)
> > 
> > Is that correct?
> 
> I think this is correct. Abstract class has
> constructors. 
> 
> Roman

I think CppModule.Foo.__init__(self) simply throws an
exception (because of no_init).
Thus Bar cannot be instantiated.

For now, __init__ must be called to initialize Python
instance correctly, no matter whether the exposed class is
abstract or not:

   
http://mail.python.org/pipermail/c++-sig/2004-June/007522.html

shin


__________________________________________________
Do You Yahoo!?
http://bb.yahoo.co.jp/




From dave at boost-consulting.com  Wed Jun 30 18:30:33 2004
From: dave at boost-consulting.com (David Abrahams)
Date: Wed, 30 Jun 2004 12:30:33 -0400
Subject: [C++-sig] Re: Calling a python function from C++ (from Jeff Holle)
References: <40E142ED.2030603@verizon.net>
Message-ID: <usmcdhuty.fsf@boost-consulting.com>

Jeff Holle <jeff.holle at verizon.net> writes:

>>
>>
>>> An underlining concern I have is error handling.  When
>>> PyImport_ImportModule fails, a NULL pointer is returned.
>>> Seems like the underlining machinery in "handle<>" is intolerant of
>>> this. 
>>  
>>
>>
>>What makes you say that?
>>
> You are correct about the success in compiling your much cleaner
> example, and it works :-) , but has a flaw.
>
> When I hide the needed python script, the call to
> PyImport_ImportModule returns NULL.
> When this occurres, both your example and what I had created, but
> commented out throws something because I can catch it with
> "catch(...)".  I don't however know how to be more specific.  Any
> hints?  Whatever it is, it doesn't inherit from "exception".

It's "boost::python::error_already_set".  It means that Python knows
what the error is already.  If you're in an extension module, it's
handled automatically.  Otherwise, you can look at
www.boost.org/libs/python/test/embedding.cpp for examples of how to
get at this exception.

This is all in the documentation, BTW:

http://www.boost.org/libs/python/doc/v2/object.html#object-spec-ctors
http://www.boost.org/libs/python/doc/v2/errors.html

> If I was operating on windows, I would anticipate it was a "structured
> exception", or whatever Microsoft calls their propritary stuff.  

Never!

> Don't know of the equivalent on linux though.  Without the "catch
> (...)", the command line program exits with "Aborted".

'course it does.  That's C++. ;-)

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




From jeff.holle at verizon.net  Wed Jun 30 04:30:46 2004
From: jeff.holle at verizon.net (Jeffrey Holle)
Date: Tue, 29 Jun 2004 22:30:46 -0400
Subject: [C++-sig] Re: Calling a python function from C++ (from Jeff Holle)
In-Reply-To: <40E142ED.2030603@verizon.net>
References: <40E142ED.2030603@verizon.net>
Message-ID: <cbutac$gs6$1@sea.gmane.org>

I've explored whats actually going on some with my debugger and now see 
what is being thrown.
It is "error_already_set".  This is a trivial object (it has no 
attributes at all), and isn't much better than "catch (...)".
However, its name implies something.  Is there a meaningful error 
message set somewhere else that can be accessed within the catch clause 
of "error_already_set"?

Jeff Holle wrote:
>>
>>
>>> An underlining concern I have is error handling.  When
>>> PyImport_ImportModule fails, a NULL pointer is returned.
>>> Seems like the underlining machinery in "handle<>" is intolerant of
>>> this. 
>>
>>  
>>
>>
>> What makes you say that?
>>
> You are correct about the success in compiling your much cleaner 
> example, and it works :-) , but has a flaw.
> 
> When I hide the needed python script, the call to PyImport_ImportModule 
> returns NULL.
> When this occurres, both your example and what I had created, but 
> commented out throws something because I can catch it with 
> "catch(...)".  I don't however know how to be more specific.  Any 
> hints?  Whatever it is, it doesn't inherit from "exception".
> 
> If I was operating on windows, I would anticipate it was a "structured 
> exception", or whatever Microsoft calls their propritary stuff.  Don't 
> know of the equivalent on linux though.  Without the "catch (...)", the 
> command line program exits with "Aborted".




From jeff.holle at verizon.net  Wed Jun 30 08:39:46 2004
From: jeff.holle at verizon.net (Jeffrey Holle)
Date: Wed, 30 Jun 2004 02:39:46 -0400
Subject: [C++-sig] Re: Calling a python function from C++ (from Jeff Holle)
In-Reply-To: <cbutac$gs6$1@sea.gmane.org>
References: <40E142ED.2030603@verizon.net> <cbutac$gs6$1@sea.gmane.org>
Message-ID: <cbvbt6$k5r$1@sea.gmane.org>

Given what the name of this exception, I assume there is a source of 
information to obtain useful information about it.
To this end, I've dug a bit into the Python C API and found the added 
the following to my code example:

   catch (error_already_set& x) {
	PyObject *err_type,*err_value,*err_traceback;
	PyErr_Fetch(&err_type,&err_value,&err_traceback);
         cout << "something bad happened" << endl;
     }

While I see that err_type and error_value are not null pointers after 
the call to PyErr_Fetch, I don't know what to do with either to get at a 
useful message.
Can anybody provide any hints?

Jeffrey Holle wrote:
> I've explored whats actually going on some with my debugger and now see 
> what is being thrown.
> It is "error_already_set".  This is a trivial object (it has no 
> attributes at all), and isn't much better than "catch (...)".
> However, its name implies something.  Is there a meaningful error 
> message set somewhere else that can be accessed within the catch clause 
> of "error_already_set"?
> 
> Jeff Holle wrote:
> 
>>>
>>>
>>>> An underlining concern I have is error handling.  When
>>>> PyImport_ImportModule fails, a NULL pointer is returned.
>>>> Seems like the underlining machinery in "handle<>" is intolerant of
>>>> this. 
>>>
>>>
>>>  
>>>
>>>
>>> What makes you say that?
>>>
>> You are correct about the success in compiling your much cleaner 
>> example, and it works :-) , but has a flaw.
>>
>> When I hide the needed python script, the call to 
>> PyImport_ImportModule returns NULL.
>> When this occurres, both your example and what I had created, but 
>> commented out throws something because I can catch it with 
>> "catch(...)".  I don't however know how to be more specific.  Any 
>> hints?  Whatever it is, it doesn't inherit from "exception".
>>
>> If I was operating on windows, I would anticipate it was a "structured 
>> exception", or whatever Microsoft calls their propritary stuff.  Don't 
>> know of the equivalent on linux though.  Without the "catch (...)", 
>> the command line program exits with "Aborted".




From jeff.holle at verizon.net  Wed Jun 30 09:29:59 2004
From: jeff.holle at verizon.net (Jeffrey Holle)
Date: Wed, 30 Jun 2004 03:29:59 -0400
Subject: [C++-sig] Re: Accessing C++ values from within Python
In-Reply-To: <cbu8sl$jkr$1@sea.gmane.org>
References: <cbu8sl$jkr$1@sea.gmane.org>
Message-ID: <40E26BF7.6070809@verizon.net>

I'm sending the source of a simple project that does what I believe you 
are trying to do.

In Main.cpp, note the use of call, and the use of ref within it.
The former does all the "magic" of converting C++ objects to/from python.
The later causes the C++ object to be passed by reference, allowing the 
python script to alter C++ object attributes.


xin wrote:
> Hi.
> I have a setup such as the following:
> 
> class object {
> public:
>     void doSomthingElse();
>     float x;
>     float *y;
>     Somthing SomePyScript;
> };
> 
> int main()
> {
>     object *array = new object[10]; // size of array is only an example
>     for (int i = 0; i < 10; i++)
>     {
>         array[i].y = new float[array[i]]
>         RunPythonScript(array[i]);
>     }
>     return 1;
> }
> 
> Where:
>  - " Somting SomePyScript; " Is some reference to a python script that 
> is affiliated with this instance of the class (varies from instance to 
> instance)
>  - " RunPythonScript(array[i]); " Is some method of running the python 
> script that is affiliated with the instance array[i].
> 
> My intention is that the python scripts change values within the class 
> 'object'
> e.g. The python code sets 'x' to equal 0.9. And then sets each value of 
> y to equal a muliple of 9. (y[0] = 0, y[1] = 9, y[2] = 18, ... ,y[100] = 
> 900 etc...)
> 
> Is this at all possible? Is it possible without moving all the memory 
> from one place, to another area accessable by both python and then back 
> again? If so how would I go about doing it?
> 
> I've tried all sorts of methods of attempting this, but have had no 
> luck. Any help would be appreciated.
> 
> Thanks.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: extended_embeddedGreeter.tar
Type: application/x-tar
Size: 20480 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/cplusplus-sig/attachments/20040630/78f7b668/attachment.tar>